Physics II by Halliday Resnick
Physics II by Halliday Resnick
Physics II by Halliday Resnick
10
conventions for geometrical optics have been put on a more physical basis;
(3) attention has been called to examples or problems that contain impor-
tant new material by using italicized titles; and (4) we have further sys-
temized and simplified the notation in several places, particularly in wave
optics.
We continue to be indebted to Dr. Benjamin Chi for assistance in pre-
paring the illustrations and for his detailed criticism of the optics chapters.
One of us (DH) wishes to thank Professor G. Wataghin for courtesies
extended at the University of Turin while this revision was in progress.
Finally, we extend special thanks to those students and instructors who
have been kind enough to write to us about their experiences with the first
edition.
David Halliday
University of Pittsburgh
January 1, 1962 Pittsburgh, Pennsylvania, 15213
Robert Resnick
Rensselaer Polytechnic Institute
Troy, New York, 12181
Preface to First Edition
The time lag between developments in basic science and their application
to engineering practice has shrunk enormously in the past few decades.
The base of engineering, once largely empirical, is now largely scientific.
Today the need is to stress principles rather than specific procedures, to
select areas of contemporary interest rather than of past interest, and to
condition the student to the atmosphere of change he will encounter during
his career. These developments require a revision of the traditional course
in general physics for engineers and scientists.
The most frequent criticisms made in varying degrees of textbooks used
in such a course are these: (a) the content is encyclopedic in that topics are
not treated with sufficient depth, the discussions are largely descriptive
rather than explanatory and analytical, and too many topics are surveyed;
(b) the content is not sufficiently "modern," and applications are drawn
mostly from past engineering practice rather than from contemporary
physics; (c) the organization of the material is too compartmentalized to
reveal the essential unity of physics and its principles; (d) the approach is
highly deductive and does not stress sufficiently the connection between
theory and experiment. Of course, it is unlikely that a textbook will ever
be written that is not criticized on one ground or another.
In writing this textbook we have been cognizant of these criticisms and
have given much thought to ways of meeting them. We have considered
the possibility of reorganizing the subject matter. The adoption of an
atomic approach from the beginning or a structure built around energy in
ix
x PREFACE
its various aspects suggest themselves. We have concluded that our goals
can best be achieved by modifying the selection and treatment of topics
within the traditional organization. To shuffle freely the cards of subject
matter content or to abandon entirely a sequence which represents the
growth of physical thought invites both a failure to appreciate the New-
tonian and Maxwellian syntheses of classical physics and a superficial
understanding of modern physics. A solid underpinning of classical
physics is essential to build the superstructure of contemporary physics in
our opinion.
To illustrate how we hope to achieve our goals within this framework, we
present here the principal features of our book.
1. Many topics are treated in greater depth than has been customary
heretofore, and much contemporary material has been woven into the body
of the text. For example, gravitation, kinetic theory, electromagnetic
waves, and physical optics, among others, are treated in greater depth.
Contemporary topics, such as atomic standards, collision cross section,
intermolecular forces, mass-energy conversion, isotope separation, the Hall
effect, the free-electron model of conductivity, nuclear stability, nuclear
resonance, and neutron diffraction, are discussed where they are pertinent.
To permit this greater depth and inclusion of contemporary material, we
have omitted entirely or treated only indirectly much traditional material,
such as simple machines, surface tension, viscosity, calorimetry, change of
state, humidity, pumps, practical engines, musical scales, architectural
acoustics, electrochemistry, thermoelectricity, motors, alternating-current
circuits, electronics, lens aberrations, color, photometry, and others.
2. We have tried to reveal the unity of physics in many ways. Through-
out the book we stress the general nature of key ideas common to all areas
of physics. For example, the conservation laws of energy, linear momen-
tum, angular momentum, and charge are used repeatedly. Wave concepts
and properties of vibrating systems, such as resonance, are used in mechan-
ics, sound, electromagnetism, optics, atomic physics, and nuclear physics.
The field concept is applied to gravitation, fluid flow, electromagnetism,
and nuclear physics.
The interrelation of the various disciplines of physics is emphasized by
the use of physical and mathematical analogies and by similarity of method.
For example, the correspondences between the mass-spring system and the
LC circuit or between the acoustic tube and the electromagnetic cavity
are emphasized, and the interweaving of microscopic and macroscopic
approaches is noted in heat phenomena and electrical and magnetic
phenomena. We have tried to make a smooth transition between particle
mechanics and kinetic theory, stressing that, in their classical aspects, both
belong to the Newtonian synthesis. We have also sought a smooth transi-
tion between electromagnetism and wave optics, pointing frequently to the
Maxwellian synthesis.
PREFACE xi
36 INDUCTANCE 899
43 INTERFERENCE 1068
43-1 Young's Experiment 1068
43-2 Coherence 1074
43-3 Intensity in Young's Experiment 1077
43-4 Adding Wave Disturbances 1080
43-5 Interference from Thin Films 1083
43-6 Phase Changes on Reflection 1089
43-7 Michelson's Interferometer 1090
43-8 Michelson's Interferometer and Light Propagation 1092
44 DIFFRACTION 1099
44-1 Introduction 1099
44-2 Single Slit 1102
44-3 Single Slit-Qualitative 1105
CONTENTS xvii
44-4 Single Slit-Quantitative 1108
44-5 Diffraction at a Circular Aperture 1112
44-6 Double Slit 1115
46 POLARIZATION 1147
SUPPLEMENTARY TOPIC V 1
SUPPLEMENTARY PROBLEMS 9
xviii CONTENTS
APPENDICES 29
Silk
thread
1
Fo::-·
r2
F is the magnitude of the force that acts on each of the two charges a and b;
r is their distance apart. These forces, as Newton's third law requires, act
along the line joining the charges but point in opposite directions. Note that
the magnitude of the force on each charge is the same, even though the
charges may be different.
Coulomb also studied how the electrical force varied with the relative size
of the charges on the spheres of his torsion balance. For example, if we touch
a charged conducting sphere to an exactly similar but uncharged conducting
sphere, the original charge must divide equally between the spheres. By
such techniques Coulomb extended the inverse square relationship to
t.,.;-iiiin--Fiber
attractive; this corresponds to the fact that there are two kinds of electricity
but (apparently) only one kind of mass.
Our belief in Coulomb's law does not rest quantitativel y on Coulomb's
experiments. Torsion balance measurement s are difficult to make to an
accuracy of better than a few per cent. Such measurement s could not, for
example, convince us that the exponent in Eq. 26-1 is exactly 2 and not, say,
2.01. In Section 28-5 we show that Coulomb's law can also be deduced from
an indirect experiment which shows that the exponent in Eq. 26-1 lies
652 CHARGE AND MATTER Chap. 26
between the limits 2 ± 2 X 10- 9 • This has recently been extended by
Williams, Faller and Hill to 2 ± 3 X 10- 16 •
Although we have established the physical concept of electric charge, we
have not yet defined a unit in which it may be measured. It is possible to do
so operationally by putting equal charges q on the spheres of a torsion balance
and by measuring the magnitude F of the force that acts on each when the
charges are a measured distance r apart. One could then define q to have a
unit value if a unit force acts on each charge when the charges are separated
by a unit distance and one can give a name to the unit of charge so defined.*
For practical reasons having to do with the accuracy of measurements, the
unit of charge in the mks system is not defined using a torsion balance but is
derived from the unit of electric current. If the ends of a long wire are
connected to the terminals of a battery, it is common knowledge that an
electric current i is set up in the wire. We visualize this current as a flow of
charge. The mks unit of current is the ampere (abbr. amp). In Section 34-4
we describe the operational procedures in terms of which the ampere is
defined.
The mks unit of charge is the coulomb (abbr. coul). A coulomb is defined as
the amount of charge that flows through a given cross section of a wire in 1 second
if there is a steady current of 1 ampere in the wire. In symbols
q = it,
N m
No= M'
where No is Avogadro's number, m the mass of the coin, and M the atomic weight
of copper. This yields, solving for N,
In a 100-watt, 110-volt light bulb the current is 0.91 amp. The student should verify
that it would take 40 hr for a charge of this amount to pass through this bulb. ....
* This scheme is the basis for the definition of the unit of charge called the statcoulomb.
However, in this book we do not use this unit or the systems of units of which it is a part;
see Appendix L, however.
Sec. 26-4 COULOMB'S LAW 653
F = _1_2_1q2_ (26-3)
41rEo r 2
Certain equations that are derived from Eq. 26-3, but are used more often
than it is, will be simpler in form if we do this.
In the mks system we can measure q1 , q2 , r, and Fin Eq. 26-3 in ways that
do not depend on Coulomb's law. Numbers with units can be assigned to
them. There is no choice about the so-called permittivity constant Eo; it must
have that value which makes the right-hand side of Eq. 26-3 equal to the
left-hand side. This (measured) value turns out to be *
IJJli,, Example 2. Let the total positive and the total negative charges in a copper
penny be separated to a distance such that their force of attraction is 1.0 lb ( = 4.5
nt). How far apart must they be?
We have (Eq. 26-3)
F = _l_ q1q2_
41rEo r 2
This suggests that it is not possible to upset the electrical neutrality of gross objects
by any very large amount. What would be the force between the two charges if they
were placed 1.0 meter apart? ~
If more than two charges are present, Eq. 26-3 holds for every pair of
charges. Let the charges be q1 , q2 , and q3 , etc.; we calculate the force exerted
on any one (say q1 ) by all the others from the vector equation
From Eq. 26-3, ignoring the signs of the charges, since we are interested only in the
magnitudes of the forces,
= 1.8 nt.
The directions of F 12 and F 13 are as shown in the ~gure.
The components of the resultant force F 1 acting i:m q1 (see Eq. 26-4) are
= -(1.8nt)(cos30°) = -l.6nt.
The student should find the magnitude of F 1 and the angle it makes with the x-axis .
...
26-5 Charge Is Quantized
In Franklin's day electric charge was thought of as a continuous fluid, an
idea that was useful for many purposes. The atomic theory of matter, how-
ever, has shown.tJiat !!_uids themselves, such as water and air, are not con-
tinuous but are made up of atoms. Experiment shows that the "electric
fluid" is not continuous either but that it is made up of integral multiples
of a certain minimum electric charge. This fundamental charge, to which
Sec. 26-6 CHARGE AND MATTER 655
we give the symbol e, has the magnitude 1.60210 X 10-19 coul. Any physi-
cally existing charge q, no matter what its origin, can be written as ne where
n is a positive or a negative integer.
When a physical property such as charge exists in discrete "packets"
rather than in continuous amounts, the property is said to be quantized.
Quantization is basic to modern physics. The existence of atoms and of
particles like the electron and the proton indicates that mass is quantized
also. Later the student will learn that several other properties prove to be
quantized when suitably examined on the atomic scale; among them are
energy and angular momentum.
The quantum of charge e is so small that the "graininess" of electricity does
not show up in large-scale experiments, just as we do not realize that the air
we breathe is made up of atoms. In an ordinary 110-volt, 100-watt light
bulb, for example, 6 X 10 18 elementary charges enter and leave the filament
every second.
There exists today no theory that predicts the quantization of charge (or
the quantization of mass, that is, the existence of fundamental particles such
as protons, electrons, pions, etc.). Even assuming quantization, the classi-
cal theory of electromagnetism and Newtonian mechanics are incomplete
in that they do not correctly describe the behavior of charge and matter on
the atomic scale. The classical theory of electromagnetism, for example,
describes correctly what happens when a bar magnet is thrust through a
closed copper loop; it fails, however, if we wish to explain the magnetism of
the bar in terms of the atoms that make it up. The more detailed theories of
quantum physics are needed for this and similar problems.
Table 26-1
Fig. 26-4 An atom, suggesting the electron cloud and, above, an enlarged view of the
nucleus.
cloud of electrons; see Fig. 26-4. The nucleus varies in radius from about
1 X 10- 15 meter for hydrogen to about 7 X 10- 15 meter for the heaviest
atoms. The outer diameter of the electron cloud, that is, the diameter of the
the atom, lies in the range 1-3 X 10- 10 meter, about 105 times larger than
the nuclear diameter.
lllll- Example 4. The distance r between the electron and the proton in the hydrogen
atom is about 5.3 X 10- 11 meter. What are the magnitudes of (a) the electrical force
and (b) the gravitational force between these two particles?
Sec. 26-6 CHARGE AND MATTER 657
From Coulomb's law,
F = _1_ q1q2
e 41rEo r 2
F = Gm1~
g r2
The significance of Coulomb's law goes far beyond the description of the
forces acting between charged balls or rods. This law, when incorporated
into the structure of quantum physics, correctly describes (a) the electric
forces that bind the electrons of an atom to its nucleus, (b) the forces that
bind atoms together to form molecules, and (c) the forces that bind atoms 01
molecules together to form solids or liquids. Thus most of the forces of our
daily experience that are not gravitational in nature are electrical. A force
transmitted by a steel cable is basically an electrical force because, if we pass
an imaginary plane through the cable at right angles to it, it is only the
attractive electrical interatomic forces acting between atoms on opposite
sides of the plane that keep the cable from parting. vVe ourselves are an
assembly of nuclei and electrons bound together in a stable configuration by
Coulomb forces.
In the atomic nucleus we encounter a new force which is neither gravita-
tional nor electrical in nature. This strong attractive force, which binds
together the protons and neutrons that make up the nucleus, is called simply
the nuclear force. If this force were not present, the nucleus would fly apart at
once because of the strong Coulomb repulsion force that acts between its
protons. The nature of the nuclear force is only partially understood today
and forms the central problem of present-day researches in nuclear physics.
~ Example 5. What repulsive Coulomb force exists between two protons in a
nucleus of iron? Assume a separation of 4.0 X 10- 15 meter.
From Coulomb's law,
F = __1__ q1q2
4m,o r 2
(9.0 X 109 nt-m 2/coul2)(1.6 X 10- 19 coul) 2
(4.0 X 10- 15 meter)2
= 14 nt.
658 CHARGE AND MATTER Chap. 26
This enormous repulsive force must be more than compensated for by the strong
attractive nuclear forces. This example, combined with Example 4, shows that
nuclear binding forces are much stronger than atomic binding forces. Atomic binding
forces are, in turn, much stronger than gravitational forces for the same particles
separated by the same distance. <1111
The repulsive Coulomb forces acting between the protons in a nucleus make the
nucleus less stable than it otherwise would be. The spontaneous emission of alpha
particles from heavy nuclei and the phenomenon of nuclear fission are evidences of
this instability.
The fact that heavy nuclei contain significantly more neutrons than protons is still
another effect of the Coulomb forces. Consider Fig. 26-5 in which a particular atomic
species is represented by a circle, the coordinates being Z, the number of protons in
the nucleus (that is, the atomic number), and N, the number of neutrons in the nucleus
(that is, the neutron number). Stable nuclei are represented by filled circles and radio-
active nuclei, that is, nuclei that disintegrate spontaneously, emitting electrons or
a-particles, by open circles. Note that all elements (iron, for example, for which
Z = 26; see arrow) exist in a number of different forms, called isotopes.
Figure 26-5 shows that light nuclei, for which the Coulomb forces are relatively
unimportant,* lie on or close to the line labeled "N = Z" and thus have about equal
numbers of neutrons and protons. The heavier nuclei have a pronounced neutron
excess, U 238 having 92 protons and 238 - 92 or 146 neutrons. t In the absence of
Coulomb forces we would assume, extending the N = Z rule, that the most stable
nucleus with 238 particles would have 119 protons and 119 neutrons. However, such
a nucleus, if assembled, would fly apart at once because of Coulomb repulsion. Rela-
tive stability is found only if 27 of the protons are replaced by neutrons, thus diluting
the total Coulomb repulsion. Even in U 238 Coulomb repulsion is still very important
because (a) this nucleus is radioactive and emits a-particles, and (b) it may break
up into two large fragments (:fission) if it absorbs a neutron; both processes result in
separation of the nuclear charge and are Coulomb repulsion effects. Figure 26-5
shows that all nuclei with Z > 83 have no stable isotopes.
We have pointed out that matter, as we ordinarily experience it, is made up of
electrons, neutrons, and protons. Nature exhibits much more variety than this,
however. No fewer than 28 distinct elementary particles are now known, most of
them having been discovered since 1940, either in the penetrating cosmic rays that
come to us from beyond our atmosphere or in the reaction products of giant cyclotron-
like devices.
Appendix E, which lists some properties of these particles, shows that, like the more
familiar particles of Table 26-1, their charges are quantized, the quantum of charge
again being e. An understanding of the nature of these particles an<l of their relation-
ships to each other is perhaps the most significant research goal of modern physics.
* Coulomb forces are important in relation to the strong nuclear attractive forces
only for large nuclei, because Coulomb repulsion occurs between every pair of protons in
the nucleus but the attractive nuclear force does not. J n U 238 , for example, every proton
exerts a force of repulsion on each of the other Hl protons. However, each proton (and
neutron) exerts a nuclear attraction on only a small number of other neutrons and protons
that happen to be near it. As we proceed to larger nuclei, the amount of energy associated
with the repulsive Coulomb forces increases much faster than that associated with the
attractive nuclear forces.
t The superscript in this notation is the mass number A ( = N + Z). This is the total
number of particles in the nucleus.
CHARGE AND MATTER 659
150~~-----,,c~~....,----~--,.c--~---,,s-~~r--~--r;:~~--r-::~-::.-----r-:--~!'7"~--,
~
~
10
Fig. 26-5 The filled circles represent stable nuclei, the open ones radioactive nuclei.
This figure does not show the more recently discovered radioactive nuclei, which appear
throughout the chart, extending up to Z = 103.
660 CHARGE AND MATTER Chap. 26
26-7 Charge Is Conserved
When a glass rod is rubbed with silk, a positive charge appears on the rod.
Measurement shows that a negative charge of equal magnitude appears on
the silk. This suggests that rubbing does not create charge but merely
transfers it from one object to another, disturbing slightly the electrical
neutrality of each. This hypothesis of the conservation of charge has stood up
under close experimental scrutiny both for large-scale events and at the
atomic and nuclear level; no exceptions have ever been found.
An interesting example of charge conservation comes about when an
electron (charge = -e) and a positron (charge = +e) are brought close to
each other. The two particles may simply disappear, converting all their
rest mass into energy according to the well-known E = mc 2 relationship;
this annihilation process was described in Section 8-9. The energy appears
in the form of two oppositely directed gamma rays, which are similar in
character to X-rays. The net charge is zero both before and after the
event so that charge is conserved. Rest mass is not conserved, being
turned completely into energy.
Another example of charge conservation is found in radioactive decay, of
which the following process is typical:
u23s - Thz34 + He4. (26-5)
The radioactive "parent" nucleus, U 238 , contains 92 protons (that is, its
atomic number Z = 92). It disintegrates spontaneously by emitting an a-
particle (He4 ; Z = 2) transmuting itself into the nucleus Th 234 , with Z = 90.
Thus the amount of charge present before disintegration ( +92e) is the same
as that present after the disintegration.
A final example of charge conservation is found in nuclear reactions, of
which the bombardment of Ca44 with cyclotron-accelerated protons is
typical. In a particular collision a neutron may emerge from the nucleus,
leaving Sc 44 as a "residual" nucleus:
Ca44 +p ----; Sc44 + n.
The sum of the atomic numbers before the reaction (20 + 1) is exactly equal
to the sum of the atomic numbers after the reaction (21 + 0). Again charge
is conserved.
QUESTIONS
1. You are given two metal spheres mounted on portable insulating supports. Find a
way to give them equal and opposite charges. You may use a glass rod rubbed with silk
but may not touch it to the spheres. Do the spheres have to be of equal size for your
method to work?
2. A charged rod attracts bits of dry cork dust which, after touching the rod, often
jump violently away from it. Explain.
3. If a charged glass rod is held near one end of an insulated uncharged metal rod as in
Fig. 26-6, electrons are drawn to one end, as shown. Why does the flow of electrons cease?
l'here is an almost inexhaustible supply of them in the metal rod.
Chap. 26 PROBLEMS 661
PROBLEMS
1. Protons in the cosmic rays strike the earth's upper atmosphere at a rate, averaged
over the earth's surface, of 0.15 protons/cm 2-sec. What total current does the earth receive
from beyond its atmosphere in the form of incident cosmic ray protons? The earth's
radius is 6.4 X 106 !Jleters.
~ .. A point cha;ge of +3.0 X 10-5 coul is 12 cm distant from a second point charge of
-1.5 X 10-6 coul. Calculate the magnitude and direction of the force on each charge.
--8:-· Two similar balls of mass m are hung from silk threads of
length land carry similar charges q as in Fig. 26-7. Assume that
(J is so small that tan (J can be replaced by its approximate equal,
sine. To this approximation show that
x = (___<[}__ ) Ya
21rEomg
a a
+2q a -2q
Fig. 26-8
g = F/m. (27-1)
This is an example of a vector field. For points near the surface of the earth
the field is often taken as uniform; that is, g is the same for all points.
The flow of water in a river provides another example of a vector field,
i:·alled a flow field (see Section 18-7). Every point in the water has associated
with it a vector quantity, the velocity v with which the water flows past the
point. If g and v do not change with time, the corresponding fields are
described as stationary. In the case of the river note that even though the
water is moving the vector v at any point does not change with time for
,teady-flow conditions.
The space surrounding a charged rod is affected by the presence of the rod,
and we speak of an electric field in this space. In the same way we speak of
a magnetic field in the space around a bar magnet. In the classical theory
of electromagnetism the electric and magnetic fields are central concepts.
Before Faraday's time, the force acting between charged particles was
thought of as a direct and instantaneous interaction between the two
particles. This action-at-a-distance view was also held for magnetic and for
graYitational forces. Today we prefer to think in terms of electric fields as
iollmrn:
1. Charge q1 in Fig. 27-1 sets up an electric field in the space around itself.
This field is suggested by the shading in the figure; later we shall show how to
represent electric fields more concretely.
663
664 THE ELECTRIC FIELD Chap. 27
2. The field acts on charge q2 ; this shows up in the force F that q2 ex-
periences.
The field plays an intermediary role in the forces between charges. There
are two separate problems: (a) calculating the fields that are set up by given
distributions of charge and (b) calculating the forces that given fields will
exert on charges placed in them. We think in terms of
In Fig. 27-1 we can also imagine that q2 sets up a field and that this field acts
on q1 , producing a force -Fon it. The situation is completely symmetrical,
each charge being immersed in a field associated with the other charge.
If the only problem in electromagnetism was that of the forces between
stationary charges, the field and the action-at-a-distance points of view
would be perfectly equivalent. Suppose, however, that q1 in Fig. 27-1
suddenly accelerates to the right. How quickly does the charge q2 learn
that q1 has moved and that the force which it (q2 ) experiences must increase?
Electromagnetic theory predicts that q2 learns about q1 's motion by a field
disturbance that emanates from q1 , traveling with the speed of light. The
action-at-a-distance point of view requires that information about q1 's
acceleration be communicated instantaneously to q2 ; this is not in accord
with experiment. Accelerating electrons in the antenna of a radio transmitter
influence electrons in a distant receiving antenna only after a time l/ c where l
is the separation of the antennas and c is the speed of light.
Sec. 27-3 LINES OF FORCE 665
This equation instructs us to use a smaller and smaller test charge q0 , evaluating the
ratio F /qo at every step. The electric field Eis then the limit of this ratio as the size
of the test charge approaches zero.
-----------tj-i-------;--
----------11~.______
; __
LINES OF FORCE 667
Figures 27-4 and 27-5 show the lines of force for two equal like charges and
for two equal unlike charges, respectively. Michael Faraday, as we have
said, used lines of force a great deal in his thinking. They were more real
for him than they are for most scientists and engineers today. It is possible
to sympathize with Faraday's point of view. Can we not almost "see" the
charges being pushed apart in Fig. 27-4 and pulled together in Fig. 27-5 by
the lines of force? The student should compare Fig. 27-5 with Fig. 18-15,
which represents a flow field. Figure 27-6 shows a representation of lines of
force around charged conductors, using grass seeds suspended in an insulating
liquid.
Lines of force give a vivid picture of the way E varies through a given region of
space. However, the equations of electromagnetism (see Table 38-3) are written in
terms of the electric field strength E and other field vectors and not in terms of the
lines of force. The electric field E varies in a perfectly continuous way as any path in
the field is traversed; see Fig. 27-7a. This kind of physical continuity, however, is
different from the continuity of the lines of force, which has no real physical meaning.
To illustrate this point we assert that the lines of force need not be continuous but can
be drawn as in Fig. 27-7b. This figure agrees with the two properties of lines of force
laid down at the beginning of this section. This point of view has been especially
emphasized by Joseph Slepian. *
* See American Journal of Physics, 19, 87 (1951).
Fig. 27-6 Photographs of the patterns of electric lines of force around (a) a charged
plate (compare Fig. 27-2), and (b) two rods with equal and opposite charges (compare
fig. 27-5). The patterns were made by suspending grass seed in an insulating liquid.
, Courtesy Educational Services Incorporated, vVatertown, Mass.)
669
670 THE ELECTRIC FIELD Chap. 27
(a)
(b)
Fig. 27-7 (a) E varies continuously as we move along any path AB in the field set up by
q. (b) Lines of force need not be continuous.
27-4 Calculation of E
Let a test charge q0 be placed a distance r from a point charge q. The
magnitude of the force acting on q0 is given by Coulomb's law, or
The electric field strength at the site of the test charge is given by Eq. 27-2, or
F 1 q
E=-=~-· (27-4)
qo 41reo r2
Sec. 27-4 CALCULATION OF E 671
where r is the distance from the charge element dq to the point P. The re-
sultant field at P is then found by adding (that is, integrating) the field
contributions due to all the charge elements, or,
(27-7)
The integration, like the sum in Eq. 27-5, is a vector operation; in Example 5
we will see lrw such an integral is handled in a simple case .
.... Exampl4 3. An electric dipole. Figure 27-8 shows a positive and a negative
charge of eqdal magnitude q placed a distance 2a apart, a configuration called an elec-
-q
tric dipole. The pattern of lines of force is that of Fig. 27-5, which also shows an
electric dipole. What is the field E due to these charges at point P, a distance r along
the perpendicular bisector of the line joining the charges? Assume r » a.
Equation 27-5 gives the vector equation
E = E1 + E2,
where, from Eq. 27-4, *
1 q
E1 = E2 = - - - - - ·
41reo a 2 r2 +
* Note that the r's in Eq. 27-4 and in this equation have different meanings
672 THE ELECTRIC FIELD Chap. 27
The vector sum of E1 and E2 points vertically downward and has the magnitude
E = 2E 1 cos e.
From the figure we see that
a
cose = ---;:===
,v'a2 + r2
Substituting the expressions for E 1 and for cos e into that for E yields
E = ___'!:_ q a 1 2aq
47l'Eo (a2 + r2) Va2 + r2 47r€o (a 2 + r )%2
(27-8a)
The essential properties of the charge distribution in Fig. 27-8, the magnitude of
the charge q and the separation 2a between the charges, enter Eq. 27-8a only as a
product. This means that, if we measure E at various distances from the electric
dipole (assuming r » a), we can never deduce q and 2a separately but only the
product 2aq; if q were doubled and a simultaneously cut in half, the electric field at
large distances from the dipole would not change.
The product 2aq is called the electric dipole moment p. Thus we can rewrite this
equation for E, for distant points along the perpendicular bisector, as
1 p
E = - - 3· (27-8b)
47l'Eo r
The result for distant points along the dipole axis (see Problem 10) and the general
result for any distant point (see Problem 23) also contain the quantities 2a and q only
as the product 2aq ( = p). The variation of E with r in the general result for distant
points is also as 1/r3 , as in Eq. 27-8b.
The dipole of Fig. 27-8 is two equal and opposite charges placed close to each other
so that their separate fields at distant points almost, but not quite, cancel. On this
point of view it is easy to understand that E(r) for a dipole varies as 1/r3 (E1. 27-8b),
whereas for a point charge E(r) drops off more slowly, namely as 1/r2 (Eq. 27-4).
Example 4. Figure 27-9 shows a charge q1 ( = +1.0 X 10-6 coul) 10 cm from
a charge q2 ( = +2.0 X 10-6 coul). At what point on the line joining the two
charges is the electric field strength zero?
The point must lie between the charges because only here do the forces exerted by
q1 and q2 on a test charge oppose each other. If E 1 is the electric field strength due
to q1 and fa that due to q2, we must have
E1 = E2
1 q1 1 q2
or (see Eq. 27-4)
47!'1:o x2 47!'1:o (l - x) 2
Sec. 27-4 CALCULATION OF E 673
where x is the distance from q1 and l equals 10 cm. Solving for x yields
l 10 cm
x=-----
1 + V2 = 4.1 cm.
1+~
The student should supply the missing steps. On what basis was the second root of
the resulti{')uadratic equation discarded?
Example-_.~J \ Figure 27-10 shows a ring of charge q and of radius a. Calculate E
for points on the axis of the ring a distance x from its center.
Consider a differential element of the ring of length ds, located at the top of the
ring in Fig. 27-10. It contains an element of charge given by
ds
dq = q~
21ra
where 21ra is the circumference of the ring. This element sets up a differential electric
field dE at point P.
The resultant field E at Pis found by integrating the effects of all the elements that
make up the ring. From symmetry this resultant field must lie along the ring axis.
Thus only the component of dE parallel to this axis contributes to the final result.
The component perpendicular to the axis is canceled out by an equal but opposite
component established by the charge element on the opposite side of the ring.
Thus the general vector integral (Eq. 27-7)
E = fdE
becomes a scalar integral E = f dE cos e.
(J
dE _ 1 dq _ 1 (q ds) 1
- 41r1:o ~ - 41rfo 21r-;: a 2 x2 • +
x
From Fig. 27-10 we have cos (J = .
va2 + x2
Noting that, for a given point P, x has the same value for all charge elements and is
not a variable, we obtain
E = f dE cos 8 = f~- q ds
47rfo (21ra)(a 2 + x 2) v a2 + :c2
x
= ~- qx fds
41rfo (21ra)(a 2 + x2)% 3
·
The integral is simply the circumference of the ring ( = 21ra), so that
E = _1_ qx
41rfo (a 2 x 2 )% +
Does this expression for E reduce to an expected result for x O? For x » a we
can neglect a in the denominator of this equation, yielding
1 q
E~---·
- 41rfo x 2
This is an expected result (compare Eq. 27-4) because at great enough distances the
ring behaves like a point charge q.
Example 6. Line of charge. Figure 27-11 shows a section of an infinite line
of charge whose linear charge density (that is, the charge per unit length, measured
r
y
x
X ---------C:±··;..f-':::;-.:.+E-:-.-:-::£:+-.-=E=-:+::Z:-=:+E::-:-:-[t;:+:,:;:,:,.,+3--,-:+£:::-:;E+:Ej:·+E·-:·:,:-·+·-:J:·-:+C:·-·-·E+·::£·-:+·z,.:·:+E.~ E I
0 dx
in coul/meter) has the constant value;\, Calculate the field Ea distance y from the
line.
The magnitude of the field contribution dE due to charge element dq ( = A dx) is
given, using Eq. 27-6, by
dE = _1_ dq = _1_ ;\ dx .
41r1:o r 2 41r1:o y 2 x2 +
Sec. 27-5 A POINT CHARGE IN AN ELECTRIC FIELD 675
The minus sign in front of dEx indicates that dEx points in the negative x direction.
The x and y components of the resultant vector Eat point Pare given by
X=-OO
cos 8 dE.
Ex must be zero because every charge element on the right has a corresponding
element on the left such that their field contributions in the x direction cancel.
Thus E points entirely in the y direction. Because the contributions to Eu from the
right- and left-hand halves of the rod are equal, we can write
.[=
x=+oo
E = Eu = 2 0
cos edE.
Note that we have changed the lower limit of integration and have introduced a factor
of two.
Substituting the expression for dE into this equation gives
E = ~-
A f x=oo
cos8
dx
·
27rE:o x=O Y2 + X2
From Fig. 27-11 we see that the quantities 8 and x are not independent. We must
eliminate one of them, say x. The relation between x and 8 is (see figure)
x = y tan e.
Differentiating, we obtain dx = y sec 2 ed8.
Substituting these two expressions leads finally to
A 10=1r/2
E = ~- cos 8 d8.
271"«:oY o=o .
The student should check this step carefully, noting that the limits must now be on 8
and not on x. For example as x --, + oo, 8 --, 7r/2, as Fig. 27-11 shows. This
equation integrates readily to
The student may wonder about the usefulness of solving a problem involving an
infinite rod of charge when any actual rod must have a finite length (see Problem 15).
However, for points close enough to finite rods and not near their ends, the equation
that we have just derived yields results that are so close to the correct values that the
difference can be ignored in many practical situations. It is usually unnecessary to
solve exactly every geometry encountered in practical problems. Indeed, if idealiza-
tions or approximations are not made, the vast majority of significant problems of
all kinds in physics and engineering cannot be solved at all. <1111
P1
+ + + + + + + + + + + + + + +
This result also follows directly-from the work-energy theorem because a constant
force qE acts over a distance y.
Example 8. Deflecting an electron beam. Figure 27-13 shows an electron of mass
m and charge e projected with speed v0 at right angles to a uniform field E. Describe
its motion.
p
The motion is like that of a projectile fired horizontally in the earth's gravitational
field. The considerations of Section 4-3 apply, the horizontal (x) and vertical (y)
motions being given by
X = Vot
and
eE
Eliminating t yields y = --x2 (27-9)
2mvo 2
for the equation of the trajectory.
When the electron emerges from the plates in Fig. 27-13, it travels (neglecting
gravity) in a straight line tangent to the parabola at the exit point. We can let it fall
on a fluorescent screen S placed some distance beyond the plates. Together with
other electrons following the same path, it will then make itself visible as a small
luminous spot; this is the principle of the electrostatic cathode-ray oscilloscope.
Example 9. The electric field between the plates of a cathode-ray oscilloscope is
1.2 X 10 4 nt/coul. What deflection will an electron experience if it enters at right
angles to the field with a kinetic energy of 2000 ev ( = 3.2 X 10- 15 joule), a typical
value? The deflecting assembly is 1.5 cm long.
Recalling that Ko = 1mv 0 2 , we can rewrite Eq. 27-9 as
eEx 2
y=--·
4Ko
If x1 is the horizontal position of the far edge of the plate, y 1 will be the corresponding
deflection (see Fig. 27-13), or
eEx1 2
YI=~-
4Ko
(1.6 X 10-19 coul)(l.2 X 104 nt/coul)(l.5 X 10- 2 meter) 2
(4)(3.2 X 10- 15 joule)
= 3.4 X 10- 4 meter = 0.34 mm.
The deflection measured, not at the deflecting plates but at the fluorescent screen, is
much larger.
678 THE ELECTRIC FIELD Chap. 27
Example 10. A positive point test charge q0 is placed halfway between two equal
positive charges q. What force acts on it at or near this point P'?
From symmetry the force at the point is zero so that the particle is in equilibrium;
the nature of the equilibrium remains to be found. Figure 27-14 (compare Fig. 27-4)
z
shows the E vectors for four points near P. If the test charge is moved along the z axis,
a restoring force is brought into play; however, the equilibrium is unstable for motion
in the x-y plane. Thus we have the three-dimensional equivalent of saddle point
equilibrium; see Fig. 14-8. What is the nature of the equilibrium for a negative test
charge? ....
(b)
w= J i e
dW =
eo
T d8 = u,
where r is the torque exerted by the agent that does the work. Combining
this equation with Eq. 27-10 yields
U = l 80
e
pE sin 8 d8 = pE iOo
e
sin 8 d8
QUESTIONS
14. An electric dipole has its dipole moment p aligned with a uniform external electric
field E. (a) Is the equilibrium stable or unstable? (b) Discuss the nature of the equilib-
rium if p and E point in opposite directions.
PROBLEMS
1. Sketch qualitatively the lines of force associated with a thin, circular, uniformly
charged disk of radius R. (Hint: Consider as limiting cases points very close to the surface
and points very far from it.) Show the lines only in a plane containing the axis of the disk.
-~-~. (a) Sketch qualitatively the lines of force associated with these long parallel lines
;. ,of charge, in a perpendicular plane. Assume that the intersections of the lines of
charge with such a plane form an equilateral triangle and that each line of charge has
the same linear charge density "II. (coul/meter). (b) Discuss the nature of the equilib-
rium of a test charge placed on the central axis of the charge assembly.
3. In Fig. 27-4 consider any two lines of force leaving the upper charge. If the angle
between their tangents for points near the charge is e, it becomes () /2 at great distances.
Verify this statement and explain it. (Hint: Consider how the lines must behave both
close to either charge and far from the charges.)
4. Assume that the exponent in Coulomb's law is not "two" but n. Show that for
n >"' 2 it is impossible to construct lines that will have the properties listed for lines of force
in Section 27-3. For simplicity, treat an isolated point charge.
~ - What is the magnitude of a point charge chosen so that the electric field 50 cm
away has the magnitude 2.0 nt/coul?
6. Two equal and opposite charges of magnitude 2.0 X 10-7 coul are 15 cm apart. (a)
What are the magnitude and direction of Eat a point midway between the charges? (b)
What force (magnitude and direction) would act on an electron placed there?
~ - Two point charges are a distanced apart (Fig. 27-16). Plot E(x), assuming x = 0 at
the left-hand charge. Consider both positive and negative values of x. Plot E as positive
if E points to the right and negative if E points to the left. Assume q1 = + 1.0 X 10- 6
coul, q2 = +3.0 X 10-s coul, and d = 10 cm.
+q
a a
•
I d !• p
x-
i.ql x _q_2_ _ _ _ ~J +Q a
-Q
Fig. 27-16 . Fig. 27-17
(b) What is the direction of E? (c) Is it reasonable that E should vary as r- 2 here and ac
r- 3 for the dipole of Fig. 27-8?
12. (a) In Fig. 27-18 locate the point (or points) at which the electric fieldstrength iE
zero. (b) Sketch qualitatively the lines of force. Take a = 50 cm.
+q a -2q
IE a
~1
•
-5q +2q • -q a
+2q
Fig. 27-18 Fig. 27-19
13. What is E in magnitude and direction at the center of the square of Fig. 27-19?
Assume that q = 1.0 X 10-8 coul and a = 5.0 cm.
T4. Two point charges of unknown magnitude and sign are placed a distance d apart.
(a) If it is possible to have E = 0 at any point not between the charges but on the line
joining them, what are the necessary conditions and where is the point located? (b) Is it
possible, for any arrangement of two point charges, to find two points (neither at infinity)
at which E = O; if so, under what conditions?
15. A thin nonconducting rod of finite length l carries a total charge q, spread uniformly
along it. Show that Eat point Pon the perpendicular bisector in Fig. 27-20 is given by
E = _q~ 1
2noy v z2 + 4y2
Show that as l --> oo this result approaches that of Example 6.
p
Fig. 27-20
Chap. 27 PROBLEMS 683
16. A thin nonconducting rod is bent to form the arc of a circle of radius a and subtends
an angle Bo at the center of the circle. A total charge q is spread uniformly along its length.
Find the electric field strength at the center of the circle in terms of a, q, and Bo.
17. A nonconducting hemispherical cup of inner radius a has a total charge q spread
uniformly over its inner surface. Find the electric field at the center of curvature.
18. Electric quadrupole. Figure 27-21 shows a typical electric quadrupole. It consists
of two dipoles whose effects at external points do not quite cancel. Show that the value of
E on the axis of the quadrupole for points distant r from its center (assumer » a) is given
by
~----P
+q
l+P
}-2q
~-p
+q
Fig. 27-21
19. An electron is constrained to move along the axis of the ring of charge in Example 5.
Show that the electron can perform oscillations whose frequency is given by
This formula holds only for small oscillations, that is, for x « a in Fig. 27-10. (Hint:
Show that the motion is simple harmonic and use Eq. 15-11.)
20. For the ring of charge in Example 5, show that the maximum value of E occurs at
x = a/v2.
21. Consider the ring of charge of Example 5. Suppose that the charge q is not distri-
buted uniformly over the ring but that charge _q 1 is distributed uniformly over half the
circumference and charge q2 is distributed uniformly over the other half. Let qi + q2 = q.
(a) Find the component of the electric field at any point on the axis directed along the axis
and compare with the uniform case of Example 5. (b) Find the component of the electric
field at any point on the axis perpendicular to the axis and compare with the uniform case
of Example 5.
'22. A thin circular disk of radius a is charged uniformly so as to have a charge per unit
area of u. Find the electric field on the axis of the disk at a distance r from the disk.
684 THE ELECTRIC FIELD Chap. 27
23. Field due to an electric dipole. Show that the components of E due to a dipole are
given, at distant points, by
1 3pxy
Ex= - -
4Ho (x 2 +
y 2)%
1 p(2y 2 - x 2)
E =-- ,
Y 4Ho (x 2 + 2
y )%
where x and y are coordinates of a point in Fig. 27-22. Show that this general result
includes the special results of Eq. 27-8b and of Problem 10.
y p
I
+q IY
I
a +----1o"-----.:..:x_ ___._I _ _ _ _ x
a
-q
Fig. 27-22
24. What is the magnitude and direction of an electric field that will balance the weight
of (a) an electron and (b) an alpha particle?
25. A particle having a charge of -2.0 X 10-9 coul is acted on by a downward electric
force of 3.0 X 10-5 newton in a uniform electric field. (a) What is the strength of the
electric field'? (b) What is the magnitude and direction of the electric force exerted on a
proton placed in this field? (c) What is the gravitational force on the proton? (d) What is
the ratio of the electric to the gravitational forces in this case?
26. (a) What is the acceleration of an electron in a uniform electric field of 106 nt/coul?
(b) How long would it take for the electron, starting from rest, to attain one-tenth the
speed of light? (c) What considerations limit the applicability of Newtonian mechanics
Jo such problems?
"'·27 An electron moving with a speed of 5.0 X 108 cm/sec is shot parallel to an electric
field of strength 1.0 X 103 nt/coul arranged so as to retard its motion. (a) How far will
the electron travel in the field before coming (momentarily) to rest, and (b) how much
time will elapse? (c) If the electric field ends abruptly after 0.8 cm, what fraction of its
initial kinetic energy will the electron lose in traversing it?
28. An electron is projected as in Fig. 27-23 at a speed of 6.0 X 106 meters/sec and at
an angle e of 45°; E = 2.0 X 103 nt/coul (directed upward), d = 2.0 cm, and l = 10.0 cm
(a) Will the electron strike either of the plates? (b) If it strikes a plate, where does it do so?
29. Dipole in a nonuniform field. Derive an expression for dE /dz at a point midway
between two equal positive charges, where z is the distance from one of the charges, meas-
ured along the line joining them. Would there be a force on a small dipole placed at this
point, its axis being aligned with the z axis? Recall that E = 0 at this point.
T
d
_J_
Fig. 27-23
Chap. 27 PROBLEMS 685
30. Oil drop experiment. R. A. Millikan set up ar1 apparatus (Fig. 27-24) in which a
tiny, charged oil drop, placed in an electric field E, could be "balanced" by adjusting E
until the electric force on the drop was equal and opposite to its weight. If the radius of
the drop is 1.64 X 10-4 cm and E at balance is 1.92 X 105 nt/coul, (a) what charge is on
the drop? (b) Why did Millikan not try to balance electrons in his apparatus instead of
oil drops? The density of the oil is 0.851 gm/cm3 . (Millikan first measured the electronic
charge in this way. He measured the drop radius by observing the limiting speed that
the drops attained when they fell in air with the electric field turned off. He charged
the oil drops by irradiating them with bursts of X-rays.)
Microscope
Fig. 27-24 Millikan's oil drop apparatus. Charged oil drops from the atomizer fall through
the hole in the central plate.
31. In a particular early run (1911), Millikan observed that the following measured
charges, among others, appeared at different times on a single drop:
6.563 X 10-19 coul 13.13 X 10-19 coul 19.71 X 10-19 coul
8.204 X 10-19 coul 16.48 X 10-19 coul 22.89 X 10-19 coul
11.50 X 10-19 coul 18.08 X 10-19 coul 26.13 X 10-19 coul
What value for the elementary charge e can be deduced from these data?
32. An electric field E with an average magnitude of about 150 nt/coul points downward
in the earth's atmosphere. We wish to "float" a sulfur sphere weighing 1.0 lb in this field
by charging it. (a) What charge (sign and magnitude) must be used? (b) Why is the
experiment not practical? Give a qualitative reason supported by a very rough numerical
calculation to prove your point.
Gauss's Law
CHAPTER 28
Fig. 28-1 Two equal and opposite charges. The dashed lines represent hypothetical
closed surfaces.
The vectors E and ~S that characterize each square make an angle ewith
each other. Figure 28-2b shows an enlarged view of the three squares on the
surface of Fig. 28-2a marked x, y, and z. Nate that at x, e > go 0 ; at y, e =
go 0 ; and at z, e < go 0.
A semiquantitative definition of flux is
(28-1)
which instructs us to add up the scalar quantity E · ~S for all elements of area
into which the surface has been divided. For points such as x in Fig. 28-2 the
contribution to the flux is negative; at y it is zero and at z it is positive.
Thus if E is everywhere outward, e < go 0, E · ~S will be positive, and <FE for
the entire surface will be positive; see Fig. 28-1, surface S 1 . If E is every-
where inward, e > go 0 , E· ~Swill be negative, and <FE for the surface will be
negative; see Fig. 28-1, surface S2 . From Eq. 28-1 we see that the appro-
priate mks unit for <FE is the newton-meter 2 / coul.
The exact definition of electric flux is found in the diff""rential limit of Eq.
68 8 G AU ss 's LA W Chap. 28
e
e by an in te gr al over the surfac
rfac
e sum over the su
ds. Heplacing th
28el--1
yi
(2 8- 2)
divided
tl1e su rf ac e in question is to be is
at
hi s su rf ac e in te gral indicates th d th at th e sc al ar qu an ti ty E ·d S
T dS an surface.
elements of ar ea ke n for th e entire
into infinitesimal an d th e su m ta
for each element
to be evaluated
A.S
(x) (y ) (z )
(b )
t,
ot dc field. (b) Tl m ,e el em en
ic al su rf ac e ;m meraed ;n an de
-2 (a) A hy po
th et
e, shown enlarged
.
figar
of ea on this surfac
. 28
Sec. 28-1 FLUX OF THE ELECTRIC FIELD 689
The circle on the integral sign indicates that the surface of integration is a
closed surface.*
.... Example I. Figure 28-3 shows a hypothetical cylinder of radius R immersed in
a uniform electric field E, the cylinder axis being parallel to the field. What is <I> E
for this closed surface?
The flux cf,E can be written as the sum of three terms, an integral over (a) the left
cylinder cap, (b) the cylindrical surface, and (c) the right cap. Thus
=f (a)
E·dS +f (b)
E·dS +f (c)
E·dS.
For the left cap, the angle 8 for all points is 180°, E has a constant value, and the
vectors dS are all parallel. Thus
f (a)
E·dS = f E cos 180° dS
= -E dS f = -ES,
where S ( = 1rR 2) is the cap area. Similarly, for the right cap,
f (c)
E·dS = +ES,
the angle 8 for all points being zero here. Finally, for the cylinder wall,
f (b)
E·dS = 0,
because 8 = 90°, hence E·dS = 0 for all points on the cylindrical surface. Thus
The fact that cf>E proves to be zero in Example 1 is predicted by Gauss's law
because no charge is enclosed by the Gaussian surface in Fig. 28-3 (q = 0).
Note that q in Eq. 28-3 (or in Eq. 28-4) is the net charge, taking its alge-
braic sign into account. If a surface encloses equal and opposite charges,
the flux cf>E is zero. Charge outside the surface makes no contribution to the
value of q, nor does the exact location of the inside charges affect this value.
Gauss's law can be used to evaluate E if the charge distribution is so
symmetric that by proper choice of a Gaussian surface we can easily evaluate
the integral in Eq. 28-4. Conversely, if Eis known for all points on a given
closed surface, Gauss's law can be used to compute the charge inside. If E
has an outward component for every point on a closed surface, cf>E, as Eq.
28-2 shows, \vill be positive and, from Eq. 28-4, there must be a net positive
charge within the surface (see Fig. 28-1, surface 8 1 ). If E has an inward
component for every point on a closed surface, there must be a net negative
charge within the surface (see Fig. 28-1, surface 8 2 ). Surface 8 3 in Fig. 28-1
encloses no charge, so that Gauss's law predicts that cf>E = 0. This is con-
sistent with the fact that lines of E pass directly through surface 8 3 , the con-
tribution to the integral on one side canceling that on the other. What
would be the value of cf>E for surface 8 4 in Fig. 28-1, which encloses both
charges?
on the charge. The advantage of this surface is that, from symmetry, E must
be normal to it and must have the same (as yet unknown) magnitude for all
points on the surface.
In Fig. 28-4 both E and dS at any point on the Gaussian surface are
directed radially outward. The angle between them is zero and the quantity
E·dS becomes simply E dS. Gauss's law (Eq. 28-4) thus reduces to
eofE·dS = EofEdS = q.
Because Eis constant for all points on the sphere, it can be factored from
inside the integral sign, leaving
where the integral is simply the area of the sphere.* This equation gives
EoE (41rr 2 ) = q
1 q
or E = 41rEo r2. (28-5)
Equation 28-5 gives the magnitude of the electric field strength E at any
point a distance r from an isolated point charge q. The direction of E is
already known from symmetry.
Let us put a second point charge q0 at the point at which E is calculated.
The magnitude of the force that acts on it (see Eq. 27-2) is
F = Eq 0 .
Combining with Eq. 28-5 gives
1 qqo
F=--,
41re 0 r2
which is precisely Coulomb's law. Thus we have deduced Coulomb's law
from Gauss's law and considerations of symmetry.
Gauss's law is one of the fundamental equations of electromagnetic theory
and is displayed in Table 38-3 as one of Maxwell's equations. Coulomb's law
is not listed in that table because, as we have just proved, it can be deduced
from Gauss's law and from simple assumptions about the symmetry of E due
to a point charge.
It is interesting to note that writing the proportionality constant in Coulomb's law
as l/41rEo (see Eq. 26-3) permits a particularly simple form for Gauss's law (Eq. 28-3).
If we had written the Coulomb law constant simply ask, Gauss's law would have to
be written as (1/41rk)<'h = q. We prefer to leave the factor 41r in Coulomb's law so
that it will not appear in Gauss's law or in other much used relations that will be de-
rived later.
* The usefulness of Gauss's law depends on our ability to find a surface over which,
from symmetry, both E and e (see Fig. 28-2) have constant values. Then E cos e can be
factored out of the integral and E can be found simply, as in this example.
692 GAUSS'S LAW Chap. 28
28-4 An Insulated Conductor
Gauss's law can be used to make an important prediction, namely: An
excess charge, placed on an insulated conductor, resides entirely on its outer sur-
face. This hypothesis.was shown to be true by experiment (see Section 28-5)
before either Gauss's law or Coulomb's law were advanced. Indeed, the
experimental proof of the hypothesis is the experimental foundation upon
which both laws rest: We have already pointed out that Coulomb's torsion
balance experiments, although direct and convincing, are not capable of great
accuracy. In showing that the italicized hypothesis is predicted by Gauss's
law, we are simply reversing the historical situation.
Figure 28-5 is a cross section of an insulated conductor of arbitrary shape
carrying an excess charge q. The dashed lines show a Gaussian surface that
lies a small distance below the actual surface of the conductor. Although the
Gaussian surface can be as close to the actual surface as we wish, it is im-
portant to keep in mind that the Gaussian surface is inside the con-
ductor.
·when an excess charge is placed at random on an insulated conductor, it
will set up electric fields inside the conductor. These fields act on the charge
carriers of the conductor and cause them to move, that is, they set up internal
currents. These currents redistribute the excess charge in such a way that
the internal electric fields are automatically reduced in magnitude. Even-
tually the electric fields inside the conductor become zero everywhere, the
currents automatically stop, and electrostatic conditions prevail. This
redistribution of charge normally takes place in a time that is negligible for
most purposes. What can be said about the distribution of the excess
charge when such electrostatic conditions have been achieved"?
If, at electrostatic equilibrium, Eis zero everywhere inside the conductor,
it must be zero for every point on the Gaussian surface. This means that the
flux <l>E for this surface must be zero. Gauss's law then predicts (see Eq.
28-3) that there must be no net charge inside the Gaussian surface. If the
excess charge q is not inside this surface, it can only be outside it, that is, it
must be on the actual surface of the conditctor.
Silk thread
Fig. 28-6 The entire charge on the ball is transferred to the outside of the can. This state-
ment and the discussion of the first paragraph of Section 28-5 are strictly correct only if the
can is provided with a conducting lid which can be closed after the ball is inserted.
,&
<JFfififiidi~----------
M~
·', c
T I .
Eye I
I
I
E = -~i.2
41rEo r
where q is the total charge. Thus for points outside a spherically symmetric dis-
696 GAUSS'S LAW Cl)ap. 28
(a) (b)
Fig. 28-8 Example 2. A spherically symmetric charge distribution, showing two Gauss-
ian surfaces. The density of charge, as the shading suggests, varies with distance from
the center but not with direction.
tribution of charge, the electric field has the value that it would have if the charge
were concentrated at its center. This reminds us that a sphere of mass m behaves
gravitationally, for outside points, as if the mass were concentrated at its center. At
the root of this similarity lies the fact that both Coulomb's law and the law of gravita-
tion are inverse square laws. The gravitational case was proved in detail in Section
16-6; the proof using Gauss's law in the electrostatic case is certainly much simpler.
Figure 28-8b shows a spherical Gaussian surface of radius r drawn inside the charge
distribution. Gauss's law (Eq. 28-4) gives
in which q' is that part of q contained within the sphere of radius r. The part of q
that lies outside this sphere makes no contribution to Eat radius r. This corresponds,
in the gravitational case (Section 16-6), to the fact that a spherical shell of matter
exerts no gravitational force on a body inside it.
An interesting special case of a spherically symmetric charge distribution is a uni-
form sphere of charge. For such a sphere, which would be suggested by uniform
shading in Fig. 28-8, the charge density p would have a constant value for all points
within a sphere of radius R and would be zero for all points outside this sphere. For
points inside such a uniform sphere of charge we can put
or q' = q (iY
where !71"R 3 is the volume of the spherical charge distribution. The expression for E
then becomes
1 qr
E=------· (28-6)
471"EO R 3
Sec. 28-6 GAUSS'S LAW-SOME APPLICATIONS 697
This equation becomes zero, as it should, for r = 0. Note that Eqs. 28-5 and 28-6
give the same result, as they must, for points on the surface of the charge distribution
(that is, if r = R). Note that Eq. 28-6 does not apply to the charge distribution of
Fig. 28-8b because the charge density, suggested by the shading, is not constant in
that case.
Example 3. The Thomson atom model. At one time the positive charge in the
atom was thought to be distributed uniformly throughout a sphere with a radius of
about 1.0 X 10- 10 meter, that is, throughout the entire atom. Calculate the electric
field strength at the surface of a gold atom (Z = 79) on this (erroneous) assump-
tion. Neglect the effect of the electrons.
The positive charge of the atom is Ze or (79)(1.6 X 10- 19 coul). Equation 28-5
yields, for E at the surface,
1 q
E=---
4rrEo r 2
Figure 28-9 is a plot of E as a function of distance from the center of the atom, using
Eqs. 28-5 and 28-6. We see that E has its maximum value on the surface and de-
creases linearly to zero at the center (see Eq. 28-6). Outside the sphere E decreases
as the inverse square of the distance (see Eq. 28-5).
Example 4. The Rutherford, or nuclear, atom. We shall see in Section 28-7 that
the positive charge of the atom is not spread uniformly throughout the atom (see
Example 3) but is concentrated in a small region (the nucleus) at the center of the
atom. For gold the radius of the nucleus is about 6.9 X 10- 15 meter. What is the
electric field strength at the nuclear surface? Again neglect effects associated with
the atomic electrons.
The problem is the same as that of Example 3, except that the radius is much
smaller. This will make the electric field strength at the surface larger, in proportion
to the ratio of the squares of the radii. Thus
This is an enormous electric field, much stronger than could be produced and main-
tained in the laboratory. It is about 108 times as large as the field calculated in
Example 3.
1.2
::::,
0
~ 0.8
c: Fig. 28-9 Example 3. The electric
,....
c,:,
field due to the positive charge in a
8 0.4 gold atom, according to the (erro-
Pt:!.
neous) Thomson model.
0 .....
0 1.0 2.0 3.0
10
r, 10- meter
698 GAUSS'S LAW Chap. 28
Fig. 28-10 Example 5. An infinite rod of charge, showing a cylindrical Gaussian surface.
to the end caps and away from the plane. Since E does not pierce the cylindrical
surface, there is no contribution to the flux from this source. Thus Gauss's law,
eofE·dS = q
E = _!:___, (28-8)
21:0
Note that Eis the same for all points on each side of the plane; compare Fig. 27-2.
Although an infinite sheet of charge cannot exist physically, this derivation is still
useful in that Eq. 28-8 yields substantially correct results for real (not infinite) charge
sheets if we consider only points not near the edges whose distance from the sheet is
small compared to the dimensions of the sheet.
Example 7. A charged conductor. Figure 28-12 shows a conductor carrying on its
surface a charge whose surface charge density at any point is u; in general u will vary
from point to point. What is E for points a short distance above the surface?
--......_____
A
700 GAUSS'S LAW Chap. 28
The direction of E for points close to the surface is at right angles to the surface,
pointing away from the surface if the charge is positive. If E were not normal to the
surface, it would have a component lying in the surface. Such a component would
act on the charge carriers in the conductor and set up surface currents. Since there
are no such currents under the assumed electrostatic conditions, E must be normal
to the surface.
The magnitude of E can be found from Gauss's law using a small flat "pill box" of
cross section A as a Gaussian surface. Since E equals zero everywhere inside the
conductor (see Section 28-4), the only contribution to cI>E is through the plane cap
of area A that lies outside the conductor. Gauss's law
eofE·dS = q
Comparison with Eq. 28-8 shows that the electric field is twice as great near a
conductor carrying a charge whose surface charge density is u as that near a noncon-
ducting sheet with the same surface charge density. The student should compare
the Gaussian surfaces in Figs. 28-11 and 28-12 carefully. In Fig. 28-11 lines of force
leave the surface through each end cap, an electric field existing on both sides of the
sheet. In Fig. 28-12 the lines of force leave only through the outside end cap, the
inner end cap being inside the conductor where no electric field exists. If we assume
the same surface charge density and cross-sectional area A for the two Gaussian
surfaces, the enclosed charge ( = uA) will be the same. Since, from Gauss's law,
the flux cI>E must then be the same in each case, it follows that E ( = cI>E/ A) must
be twice as large in Fig. 28-12 as in Fig. 28-11. It is helpful to note that in Fig.
28-11 half the flux emerges from one side of the surface and half from the other,
whereas in Fig. 28-12 all the flux emerges from the outside surface. ~
Fig. 28-13 Experimental arrangement for studying the scattering of a-particles. Parti-
cles from radioactive source Sare allowed to fall on a thin metal "target" T; a-particles
scattered by the target through an (adjustable) angle 1> are counted by detector D.
The electrons in the gold atom, being so light, have almost no effect on the
motion of an oncoming a-particle; the electrons are themselves strongly de-
flected, just as a swarm of insects would be by a stone hurled through them.
Any deflection of the a-particle must be caused by the repulsive action of the
positive charge of the gold atom, which is known to possess most of the mass
of the atom.
Incident
a-particles
Fig. 28-14 The deflection of the incident a-particles depends on the nature of the nuclear
collision. (From Andrade, Scientific American, November 1956.)
702 GAUSS'S LAW Chap. 28
At the time of these experiments most physicists believed in the so-called
"plum pudding" model of the atom that had been suggested by J. J. Thom-
son (1856-1940). In this view (see Example 3) the positive charge of the
atom was thought to be spread out through the whole atom, that is, through.
a spherical volume of radius about 10- 10 meter. The electrons were thought
to vibrate about fixed centers inside this sphere.
Rutherford showed that this model of the atom was not consistent with
the a-scattering experiments and proposed instead the nuclear model of the
atom that we now accept. Here the positive charge is confined to a very
much smaller sphere whose radius is about 10-14 meter (the nucleus). The
electrons move around this nucleus and occupy a roughly spherical volume
of radius about 10-10 meter. This brilliant deduction by Rutherford laid
the foundation for modern atomic and nuclear physics.
The feature of the a-scattering experiments that attracted Rutherford's
attention at once was that a few a-particles are deflected through very large
angles, up to 180°. To scientists accustomed to thinking in terms of the
"plum pudding" model, this was a very surprising result. In Rutherford's
words: "It was quite the most incredible event that ever happened to me in
my life. It was almost as incredible as if you had fired a 15-inch shell at a
piece of tissue paper and it came back and hit you."
The a-particle must pass through a region in which the electric field
strength is very high indeed in order to be deflected so strongly.* Example
3 shows that, in Thomson's model, the maximum electric field strength is
1.1 X 10 13 nt/coul. Compare this with the value calculated in Example 4
for a point on the surface of a gold nucleus (2.3 X 1021 nt/coul). Thus the
deflecting force acting on an a-particle can be up to 108 times as great if the
positive charge of the atom is compressed into a small enough region (the
nucleus) at the center of the atoms. Rutherford made his hypothesis about
the existence of nuclei only after a much more detailed mathematical analysis
than that given here.
QUESTIONS
1. A point charge is placed at the center of a spherical Gaussian surface. Is <l>E changed
(a) if the surface is replaced by a cube of the same volume, (b) if the sphere is replaced by a
cube of one-tenth the volume, (c) if the charge is moved off-center in the original sphere,
still remaining inside, (d) if the charge is moved just outside the original sphere, (e) if a
second charge is placed near, and outside, the original sphere, and (f) if a second charge is
placed inside the Gaussian surface?
2. By analogy with <l>E, how would you define the flux <I>g of a gravitational field? What
is the flux of the earth's gravitational field through the boundaries of a room, assumed to
contain no matter?
3. In Gauss's law,
5. A surface encloses an electric dipole. What can you say about 'PE for this surface?
6. Suppose that a Gaussian surface encloses_ no net charge. Does Gauss's law require
that E equal zero for all points on the surface? Is the converse of this statement true, that
is, if E equals zero everywhere on the surface, does Gauss's law require that there be no net
charge inside?
7. Would Gauss's law hold if the exponent in Coulomb's law were not exactly two?
8. Does Gauss's law, as applied in Section 28-4, require that all the conduction electrons
in an insulated conductor reside on the surface?
9. In Section 28-4 we assumed that E equals zero everywhere inside a conductor.
However, there are certainly very large electric fields inside the conductor, at points close
to the electrons or to the nuclei. Does this invalidate the proof of Section 28-4?
10. It is sometimes said that excess charge resides entirely on the outer surface of a
conductor because like charges repel and try to get as far away as possible from one an-
other. Comment on this plausibility argument.
11. Is Gauss's law useful in calculating the field due to three equal charges located at
the corners of an equilateral triangle? Explain.
12. The use of line, surface, and volume densities of charge to calculate the charge
contained in an element of a charged object implies a continuous distribution of charge,
whereas, in fact, charge on the microscopic scale is discontinuous. How, then, is this pro-
cedure justified? ... _...
13. Is E necessarily zero inside a charged rubber balloon if the balloon is (a) spl:iJffcal or
(b) sausage-shapeg.? For each shape assume the charge to be distributed uniformly over
the surface. '
14. A spherical rubber balloon carries a charge that is uniformly distributed over its
surface. How does E vary for points (a) inside the balloon, (b) at the surface of the balloon,
and (c) outside the balloon, as the balloon is blown up?
15. As you penetrate a uniform sphere of charge, E should decrease because less charge
lies inside a sphere drawn through the observation point. On the other hand, E should
increase because you are closer to the center of this charge. Which effect predominates
and why?
16. Given a spherically symmetric charge distribution (not of uniform density of charge),
is E necessarily a maximum at the surface? Comment on various possibilities.
17. An atom is normally electrically neutral. Why then should an a-particle be de-
flected by the atom under any circumstances?
18. If an a-particle, fired at a gold nucleus, is deflected through 135 °, can you conclude
(a) that any force has acted on the a-particle or (b) that any net work has been done on it?
PROBLEMS
_l. Calculate 'PE through a hemisphere of radius R. The field of E is uniform and is
parallel to the axis of the hemisphere.
2. In Example 1 compute 'PE for the cylinder if it is turned so that its axis is perpendicu-
lar to the electric field. Do not use Gauss's law.
3. A plane surface of area A is inclined so that its normal makes an angle fJ with a uni-
fo.~m field of E. Calculate 'PE for this surface.
4. A point charge of 1.0 X 10-6 coul is at the center of a cubical Gaussian surface 0.50
meter on edge. What is 'PE for the surface?
704 GAUSS'S LAW Chap. 28
5. Charge on an originally uncharged insulated conductor is separated by holding a
positively charged rod nearby, as in Fig. 28-15. What can you learn from Gauss's law
about the flux for the five Gaussian surfaces shown? The induced negative charge on the
conductor is equal to the positive
charge on the rod.
/
--
->-&-"Gauss's law for gravitation"
....... .......
18
,,,,,...----- ___ 8
y _ __.,. . /
- 1-<1>, = - 1- -'g·dS = - m,
I / ..,.--- 84 ---- ..... 4,rG 4,rG 1'
/ 1 1/ 1,...-.81. . , -----
Conductor
8 '
1....---\ '
1
I I ;++ / :
r\ ~ ...+ r -:~
l __; I where m is the enclosed mass and
I
1
\ 1
\ ,_
', __ _________
, ........ ...,-,
__.
____ __
\. ...._ _
.,.,,/
__.. I I
I G is the universal gravitation
constant (Section 16-3). Derive
......
......
---- Newton's law of gravitation from
this.
Fig. 28-15 _.;r. Figure 28-16 shows a point
· charge of 1.0 X 10-7 coul at the
center of a spherical cavity of radius 3.0 cm in a piece of metal. Use Gauss's law to find
the electric field at point a, halfway from the center to the surface, and at point b.
/R An uncharged spherical thin metailic
/ shell has a point charge q at its center. Give
expressions for the electric field (a) inside the
shell, and (b) outside the shell, using Gauss's
law. (c) Has the shell any effect on the field
due to q? (d) Has the presence of q any effect
on the shell? (e) If a second point charge is
held outside the shell, does this outside charge
experience a force? (f) Does the inside
charge experience a force? (g) Is there a con-
tradiction with Newton's third law here?
,,,.,.,9: Two large nonconducting sheets of pos-
./ itive charge face each other as in Fig. 28-17.
What is Eat points (a) to the left of the sheets,
(b) between them, and (c) to the right of the
sheets? Assume the same surface charge
density u for each sheet. Assume a « l and Fig. 28-16
do not consider points near the edges.
(Hint: E at any point is the vector sum of the separate electric field strengths set up
by each sheet.) ~ a 4
+
+
!- a ~ ..
.
l l
+
+
+
+
+
+
J
+
+
+
+
+
10. Two large metal plates face each other as in Fig. 28-18 and carry charges with
surface charge density +<1 and -<1, respectively, on their inner surfaces. What is Eat
points (a) to the left of the sheets, (b) between them, and (c) to the right of the sheets.
Assume a « l and do not consider points near the edges.
11. Two large metal plates of area 1.0 meter2 face each other. They are 5.0 cm apart
and carry equal and opposite charges on their inner surfaces. If E between the plates is
55 nt/coul, what is the charge on the plates? Neglect edge effects. See Problem 10.
12. A thin-walled metal sphere has a radius of 25 cm and carries a charge of 2.0 X 10-7
coul. Find E for a point (a) inside the sphere, (b) just outside the sphere, and (c) 3.0 meters
from the center of the sphere.
13. A 100-ev electron is·fired directly toward a large metal plate that has a surface
charge density of -2.0 X 10-6 coul/meter2 • From what distance must the electron be
fired if it is to just fail to strike the plate?
·-··44. Charge is distributed uniformly throughout an infinitely long cylinder of radius R.
Show that E at a distance r from the cylinder
axis (r < R) is given by
pr
E=-
2eo
18. In Problem 17 a positron revolves in a circular path of radius r, between and con-
centric with the cylinders. What must be its kinetic energy K? Assume a = 2.0 cm,
b = 3.Q ..crn, and X = 3.0 X 10-8 coul/meter.
~ A long conducting cylinder carrying a total charge +q is surrounded by a conducting
cylindrical shell of total charge -2q. as shown in cross section in Fig. 28-22. Use Gauss's
Fig. 28-22
law to find (a) the electric field strength at points outside the conducting shell, (b) the
distribution of the charge on the conducting shell, and (c) the electric field strength in the
region between the cylinders. List the assumptions made in arriving at your answers.
20. A thin metallic spherical shell of radius a carries a charge qa. Concentric with it is
another thin metallic spherical shell of radius b (b > a) carrying a charge qb. Use Gauss's
law to find the electric field strength at radial points r where {a) r < a; (b) a < r < b; (c)
r > b. (d) How is the charge on each shell distributed between the inner and outer
surfaces of that shell?
[2l~'A small sphere whose mass mis 1.0 X 10-3 gm carries a charge q of 2.0 X 10-8 coul.
It hangs from a silk thread which makes an angle of 30° with a large, uniformly charged
sheet as in Fig. 28-23. Calculate the surface charge density u for the sheet.
+
.u
+
+
+
+ ()
+
+
+
+
+
+ m,q
+
+
Fig. 28-23
Chap. 28 PROBLEMS 707
22. Equation 28-9 (E = u / EO) gives the electric field at points near a charged conducting
surface. Show that this equation leads to a familiar result when applied to a conducting
sphere of radius r, carrying a charge q.
23. An a-particle, approaching the surface of a nucleus of gold, is a distance equal to
one nuclear radius (6.9 X 10-u meter) away from that surface. What is the electro-
static force on the a-particle and its acceleration at that point? The mass of the a-
particle, which may be treated here as a point, is 6.7 X 10-27 kg.
24. A gold foil used in a Rutherford scattering experiment is 3 X 10-0 cm thick.
(a) What fraction of its surface area is "blocked out" by gold nuclei, assuming a nuclear
radius of 6.9 X 10-15 meter? Assume that no nucleus is screened by any other. (b)
What fraction of the volume of the foil is occupied by the nuclei? (c) What fills all the
rest of the space in the foil?
25. The electric field components in Fig. 28-24 are Ex = bx72 , Ey = Ez = 0, in which
b = 800 nt/coul-m%. Calculate (a) the flux 'PE through the cube and (b) the charge within
the cube. Assume that a = 10 cm.
Fig. 28-24
Electric Potential
CHAPTER 29
WAB
VB - VA=--· (29-1)
qo
The work WAB may be (a) positive, (b) negative, or (c) zero. In these cases
the electric potential at B will be (a) higher, (b) lower, or (c) the same as the
electric potential at A.
The mks unit of potential difference that follows from Eq. 29-1 is the
joule/coul. This combination occurs so often that a special unit, the volt,
is used to represent it; that is,
q
710 ELECTRIC POTENTIAL Chap. 29
path-independence proof applies only to this special case, it illustrates the
general principles involved.
Point A in Fig. 29-1 may be taken as a defined reference point, and we
imagine a positive test charge q0 moved by an external agent from A to B.
We consider two paths, path I being a radial line between A and Band path
II being a completely arbitrary path between these two points. The open
arrows on path II show the electric force per unit charge that would act at
various points on a test charge q0 .
Path II may be approximated by a broken path made up of alternating
elements of arc and of radius. Since these elements can be arbitrarily small,
the broken path can be made arbitrarily close to the actual path. On path
II the external agent does work only along the radial segments because along
the arcs the force F and the displacement di are at right angles, F · di being
zero in such cases. The sum of the work done on the radial segments that
make up path II is the same as the work done on path I because each path
has the same array of radial segments. Since path II is arbitrary, we have
proved that the work done is the same for all paths connecting A and B.
Although this proof holds only for the special case of Fig. 29-1, the potential
difference is path-independent for any two points in any electrostatic field.
We discussed path independence in Section 8-2 for the general class of conserv-
ative forces; electrostatic forces, like gravitational forces, are conservative.
The locus of points, all of which have the same electric potential, is called
an equipotential surface. A family of equipotential surfaces, each surface
corresponding to a different value of the potential, can be used to give a
general description of the electric field in a certain region of space. We have
seen earlier (Section 27-3) that electric lines of force can also be used for this
purpose; in later sections (see, for example, Fig. 29-15) we explore the inti-
mate connection between these two ways of describing the electric field.
No work is required to move a test charge between any two points on an
equipotential surface. This follows from Eq. 29-1,
WAB
VB - VA=--,
qo
because W AB must be zero if VA = VB· This is true, because of the path
independence of potential difference, even if the path connecting A and B
does not lie entirely in the equipotential surface.
Figure 29-2 shows an arbitrary family of equipotential surfaces. The
work to move a charge along paths I and II is zero because all these paths
begin and end on the same equipotential surface. The work to move a
charge along paths I' and II' is not zero but is the same for each path because
the initial and the final potentials are identical; paths I' and II' connect the
same pair of equipotential surfaces.
From symmetry, the equipotential surfaces for a spherical charge are a
family of concentric spheres. For a uniform field they are a family of planes
at right angles to the field. In all cases (including these two examples) the
equipotential surfaces are at right angles to the lines of force and thus to E
Sec. 29-2 POTENTIAL AND FIELD STRENGTH 711
Fig. 29-2 Portions of four equipotential surfaces. The heavy lines show four paths along
which a test charge is moved.
(see Fig. 29-15). If E were not at right angles to the equipotential surface, it
would have a component lying in that surface. Then work would have to
be done in moving a test charge about on the surface. Work cannot be done
if the surface is an equipotential, so E must be at right angles to the surface.
There is a strong analogy between electrostatic forces and gravitational forces,
based on the fact that their fundamental laws are inverse square laws (see Eqs.
26-3 and 16-1) :
and F = Gm1m2.
g r2
Thus we can define the gravitational potential Vg (compare Eq. 29-2) from
w
Vg = - ,
m
where Wis the work required to move a test body of mass m from infinity to the point
in question. Gravitational equipotential surfaces can also be constructed; they prove
to be everywhere at right angles to the gravitational field strength vector g. For a
uniform gravitational field, such as that near the surface of the earth, these surfaces
are horizontal planes. This correlates with the facts that (a) no net work is required
to move a stone of mass m between two points with the same elevation and (b) the
same net work is required to move a stone along any path starting on a given hori-
zontal surface and ending on another.
WAB
VB - vA = - - = Ed. (29-4)
qo
This equation shows the connection between potential difference and field
strength for a simple special case. Note from this equation that another mks
unit for Eis the volt/meter. The student may wish to prove that a volt/
meter is identical with a nt/coul; this latter unit is the one first presented for
E in Section 27-2.
In Fig. 29-3 B has a higher potential than A. This is reasonable because
an external agent would have to do positive work to push a positive test
charge from A to B. Figure 29-3 could be used as it stands to illustrate the
act of lifting a stone from A to Bin the uniform gravitational field near the
earth's surface.
What is the connection between V and E in the more general case in which
the field is not uniform and in which the test body is moved along a path that
is not straight, as in Fig. 29-4? The electric field exerts a force q0 E on the test
charge, as shown. To keep the test charge from accelerating, an external
agent must apply a force F chosen to be exactly equal to -q 0 E for all posi-
tions of the test body.
If the external agent causes the test body to move through a displacement
dl along the path from A to B, the element of work done by the external agent
1
Fig. 29-3 A test charge qo is moved
from A to B in a uniform electric
d field E by an external agent that
exerts a force Fon it.
'V I
-'------+-·A E
Sec. 29-2 POTENTIAL AND FIELD STRENGTH 713
is F · di. To find the total work W AB done by the external agent in moving
the test charge from A to B, we add up (that is, integrate) the work contri-
butions for all the infinitesimal segments into which the path is divided.
This leads to
WAB JB
VB - VA= - - = - E·dl. (29-5)
qo A
00
E·dl. (29-6)
vB - vA = - L B B
E·dl = - LE cos 180° dl =LE dl.
B
Eis the constant for all parts of the path in this problem and can thus be taken out-
side the integral sign, giving
B
vB - vA = EL dl = Ed,
B ---+---+--~-+------1 --..._ C
Points Band C have the same potential because no work is done in moving a charge
between them, E and di being at right angles for all points on the line CB. In other
words, B and C lie on the same equipotential surface at right angles to the lines of
force. Thus
v B - vA = v c - vA = Ed.
This is the same value derived for a direct path connecting A and B, a result to be
expected because the potential difference between two points is path independent. ....
Sec. 29-3 POTENTIAL DUE TO A POINT CHARGE 715
29-3 Potential Due to a Point Charge
Figure 29-6 shows two points A and B near an isolated point charge q.
For simplicity we assume that A, B, and q lie on a straight line. Let us com-
pute the potential difference between points A and B, assuming that a test
charge q0 is moved without acceleration along a radial line from A to B.
In Fig. 29-6 E points to the right and di, which is always in the direction of
motion, points to the left. Therefore, in Eq. 29-5,
dl = -dr.
Fig. 29-6 A test charge qo is moved by an external agent from A to Bin the field set up
by a point charge q.
716 ELECTRIC POTENTIAL Chap. 29
Substituting this into Eq. 29-5 gives
leads to (29-7)
1 q
V=--· (29-8)
47l"Eo r
This equation shows clearly that equipotential surfaces for an isolated point
charge are spheres concentric with the point charge (see Fig. 2t)-15:r). A
study of the derivation will show that this relation also holds for points ex-
ternal to spherically symmetric charge distributions.
~ Example 3. What must the magnitude of an isolated positive point charge be
for the electric potential at 10 cm from the charge to be + 100 volts?
Solving Eq. 29-8 for q yields
Example 4. What is the electric potential at the surface of a gold nucleus'? The
radius is 6.6 X 10- 15 meter and the atomic number Z = 79.
The nucleus, assumed spherically symmetrical, behaves electrically for external
points as if it were a point charge. Thus we can use Eq. 29-8, or, recalling that the
proton charge is 1.6 X 10- 19 coul,
(29-9)
A GROUP OF POINT CHARGES 717
where qn is the value of the nth charge and r n is the distance of this charge
from the point in question. The sum used to calculate Vis an algebraic sum
and not a vector sum like the one used to calculate E for a group of point
charges (see Eq. 27-5). Herein lies an important computational advantage
of potential over electric field strength.
If the charge distribution is continuous, rather than being a collection of
points, the sum in Eq. 29-9 must be replaced by an integral, or
V = f dV = -41re1 fdq-,r
0
,vhere dq
is a differential element of the charge distribution, r is its distance
from the point at which V is to be calculated, and dV is the potential it
establishes at that point.
~ Example 5. What is the potential at the center of the square of Fig. 29-7?
Assume that q1 = + 1.0 X 10-s coul, qz = -2.0 X 10-s coul, q3 = +3.0 X 10-s
coul, q4 = +2.0 X 10-s coul, and a = 1.0 meter.
a p a
Fig. 29-7 Example 5
The distance r of each charge from Pis a/ y2 or 0.71 meter. From Eq. 29-9
v = L vn = _1_ q1 + q2 + qa + q4
n 47rEO r
(9.0 X 109 nt-m 2/coul2)(1.0 - 2.0 + 3.0 + 2.0) X 10-s coul
0.71 meter
= 500 volts.
Is the potential constant within the square? Does any point inside have a negative
potential? Can you sketch roughly the intersection of the plane of Fig. 29-7 with
the equipottntial surface corresponding to zero volts?
,;.,C-"''"". ·--~-;, _,. ·,
where (21ry)(dy) is the area of the strip: All parts of this charge element are the same
distance r' ( = V y2 + r 2) from axial point P so that their contribution dV to the
718 ELECTRIC POTENTIAL Chap. 29
dV = _I_ dq = _I <J'21ry dy
41re 0 r' 47re0 V y2 + r2
The potential Vis found by integrating over all the strips into which the disk can be
divided (Eq. 29-10) or
V = Jav = _<!__
2eo Jo
ra (y + r
2 2) -%y dy
= _<!__ ( ~ - r).
2eo
This general result is valid for all values of r. In the special case of r » a the quan-
tity V a2 + r2 can be approximated as
~ = r(1+ ;:y 2
= r(1 + i ~: + ···) ~ r + ~:,
in which the quantity in parentheses in the second member of this equation has been
expanded by the binomial theorem (see Appendix I). This equation means that V
becomes
where q ( = <J'1ra 2) is the total charge on the disk. This limiting result is expected
because the disk behaves like a point charge for r » a. ...._
by the arrow on the axis of symmetry of the molecule. In this molecule the effective
center of positive charge does not coincide with the effective center of negative charge.
It is precisely because of this separation that the dipole moment exists.
Atoms, and many molecules, do not have permanent dipole moments. However,
dipole moments may be induced by placing any atom or molecule in an external
electric field. The action of the field (Fig. 29-12) is to separate the centers of positive
and of negative charge. We say that the atom becomes polarized and acquires an
induced electric dipole moment. Induced dipole moments disappear when the electric
field is removed.
Electric dipoles are important in situations other than atomic and molecular ones.
Radio and radar antennas are often in the form of a metal wire or rod in which elec-
trons surge back and forth periodically. At a certain time one end of the wire or rod
will be negative and the other end positive. Half a cycle later the polarity of the ends
is exactly reversed. This is an oscillating electric dipole. It is so named because its
dipole moment changes in a periodic way with time.
1 ( -q - -
V=LVn=- 2q+ q- )
n 47!'eo r - a r r +a
q 2a 2
41reo (r - a)(r)(r + a)
Assuming that r » a allows us to put a = 0 in the sum and difference terms in the
denominator, yielding
1 Q
V=--,
47!'eo r 3
where Q ( = 2qa 2) is the electric quadrupole moment of the charge assembly of Fig.
27-21. Note that V varies (a) as 1/r for a point charge (see Eq. 29-8), (b) as 1/r2 for
a dipole (see Eq. 29-11), and (c) as 1/r3 for a quadrupole.
Note too that (a) a dipole is two equal and opposite charges that do not coincide in
space so that their electric effects at distant points do not quite cancel, and (b) a
·f----------..~
ql
l--r J
q2
Fig. 29-13
them, an external agent must do work that will be positive if the charges are
opposite in sign and negative otherwise. The energy represented by this
work can be thought of as stored in the system q1 + q2 as electric potential
energy. This energy, like all varieties of potential energy, can be trans-
formed into other forms. If q1 and q2 , for example, are charges of opposite
sign and we release them, they will accelerate toward each other, transform-
ing the stored potential energy into kinetic energy of the accelerating masses.
The analogy to the earth + stone system is exact, save for the fact that
electric forces may be either attractive or repulsive whereas gravitational
forces are always attractive.
VVe define the electric potential energy of a system of point charges as the
work required to assemble this system of charges by bringing them in from an
infinite distance. ,v
e assume that the charges are all at rest when they are
infinitely separated, that is, they have no initial kinetic energy. \
In Fig. 29-13 let us imagine q2 removed to infinity and at rest. The
* In mechanics the concept of potential energy (of compressed springs, falling masses,
etc.) is more commonly used than the concept of potential. In electrostatics the reverse is
true, electric potential being perhaps a more common concept than electric potential
energy. In what follows the student must be careful not to confuse these quite different
quantities, potential and potential energy.
Sec. 29-6 ELECTRIC POTENTIAL ENERGY 723
electric potential at the original site of q2 , caused by q1, is given by Eq. 29-8,
or
If q2 is moved in from infinity to the original distance r, the work required is,
from the definition of electric potential, that is, from Eq. 29-2,
(29-12)
Combining these two equations and recalling that this work W is precisely
the electric potential energy U of the system q1 +
q2 yields
U (= W) = _l_q1q2. (29-13)
4'1l"eo r12
The subscript of r emphasizes that the distance involved is that between the
point charges q1 and qz.
For systems containing more than two charges the procedure is to compute
the potential energy for every pair of charges separately and to add the
results algebraically. This procedure rests on a physical picture in which (a)
charge q1 is brought into position, (b) q2 is brought from infinity to its posi-
tion near q1 , (c) q3 is brought from infinity to its position near q1 and q2 , etc.
The potential energy of continuous charge distributions (an ellipsoid of charge,
for example) can be found by dividing the distribution into infinitesimal elements dq,
treating each such element as a point charge, and using the procedures of the preced-
ing paragraph, with the summation process replaced by an integration. We have
not considered such problems in this text.
~ Example 8. Two protons in a nucleus of U 238 are 6.0 X 10- 15 meter apart.
What is their mutual electric potential energy?
From Eq. 29-13
U = U12 +U +U
13 23
* We assume that the equipotentials are so close together that F is constant for all
parts of the path M. In the limit of a differential path (di) there will be no difficulty.
Sec. 29-7 CALCULATION OF E FROM V 725
where Fis the force that must be exerted on the charge to overcome exactly
the electrical force q0 E. Since F and q0 E have opposite signs and are equal in
magnitude,
These two expressions for the work must be equal, which gives
~v
or E cos O = - · (29-14)
~l
dV
Ez= - - · (29-15)
dl
There will be one direction I for which the quantity dV /dl is a maximum. From
Eq. 29-15, E1 will also be a maximum for this direction and will in fact be E itself.
Thus
(29-16)
The maximum value of dV /dl at a given point is called the potential gradient at that
point. The direction I for which dV /dl has its maximum value is always at right
angles to the equipotential surface, corresponding to() = 0 in Fig. 29-16.
If we take the direction I to be, in turn, the directions of the x, y, and z axes, we
can find the three components of Eat any point, from Eq. 29-15.
av av Ez = - - ·
av (29-17)
Ex= - ax; Ey = - ay; az
Thus if Vis known for all points of space, that is, if the function V(x, y, z) is known,
the components of E, and thus E itself, can be found by taking derivatives.*
* The symbol av /ax is a partial derivative. It implies that in taking this derivative of the
function V(x, y, z) the quantity xis to be viewed as a variable and y and z are to be regarded
as constants. Similar considerations hold for av/ ay and av /az.
726 ELECTRIC POTENTIAL Chap. 29
/
/
/
/
/
/
/
/
(a)
(b)
Sec. 29-7 CALCULATION OF E FROM V 727
(c)
Fig. 29-15 Equipotential surfaces (dashed lines) and lines of force (solid lines) for (a) a
point charge, (b) a uniform electric field, and (c) an electric dipole. In all figures there is a
constant difference of potential ~ V between adjacent equipotential surfaces. Thus from
Eq. 29-14, written for the case of() = 180° as ~l = -~ V /E, the surfaces will be relatively
close together where E is relatively large and relatively far apart where E is small. Simi-
larly (see Section 27-3) the lines of force are relatively close together where Eis large and
far apart where Eis small. See discussion and figures of Section 18-7 for other examples.
728 ELECTRIC POTENTIAL Chap. 29
V+~V
V-2~V
... Example 10. Calculate E(r) for a point charge q, using Eq. 29-16 and assuming
that V(r) is given as (see Eq. 29-8)
1 q
V=--·
41r1:o r
From symmetry, E must be directed radially outward for a (positive) point charge.
Consider a point Pin the field a distance r from the charge. It is clear that dV / dl at P
has its greatest value if the direction l is identified with that of r. Thus, from Eq.
29-16,
- - _q _!!:__
- 41r1:o dr r
(!) -- __1 !£.
41r1: r0 2
Example II. Efor a dipole. Figure 29-17 shows a (distant) point Pin the field
of a dipole located at the origin of an xy-axis system. Vis given by Eq. 29-11, or
V = _1_ p cos~.
41r1:o r 2
The result is
Sec. 29-7 CALCULATION OF E FROM V 729
We find Eu from Eq. 29-17, recalling that xis to be treated as a constant in this
calculation:
Ey --
av p (x2 + y2)% _ yJ(x2 + y2)"Y2(2y)
ay 41rt=o (x2+ y2)3
p x2 - 2y2
- 41rt=o (x2 + y2)%
Note that putting x = 0 describes points along the dipole axis (that is, the y axis),
and the expression for Ey reduces to
E - 2p __!___
Y - 41rt=o y 3
This result agrees exactly with that found in Chapter 27 (see Problem 10), for, from
symmetry, Ex equals zero on the dipole axis.
Putting y = 0 in the expression for Ey describes points in the median plane of the
dipole and yields
p 1
Ey=----,
41reo x3
which agrees exactly with the result found in Chapter 27 (see Example 3), for, again
from symmetry, Ex equals zero in the median plane. The minus sign in this equation
indicatef\ that E points in the negative y direction (see Fig. 29-10).
The component Ex is also found from Eq. 29-17, recalling that y is to be taken as
a constant during this calculation:
3p xy
41reo (x2 +
y2)~1
As expected, Ex vanishes both on the dipole axis (x = O) and in the median plane
(y = 0); see Fig. 29-10. "41
(}
Fig. 29-17 Showing a point P (} r
in the field of an electric dipole p. y
x
!~--------'----~x
730 ELECTRIC POTENTIAL Chap. 29
29-8 An Insulated Conductor
We proved in Section 28-4, using Gauss's law, that after a steady state is
reached an excess charge q placed on an insulated conductor will move to its
outer surface. We now assert that this charge q will distribute itself on this
surface so that all points of the conductor, including those on the surface and
those inside, have the same potential.
Consider any two points A and Bin or on the conductor. If they were not
at the same potential, the charge carriers in the conductor near the point of
lower potential would tend to move toward the point of higher potential. We
have assumed, however, that a steady-state situation, in which such currents
do not exist, has been reached; thus all points, both on the surface and inside
VI
~
0
>
i:t:)
0
r-i
~·
--
~
....
Q)
Q)
E
VI
8
0
>
i:t:)
0 4
r-i
~·
1 2 3 4
r, meter
(b)
Sec. 29-8 AN INSULATED CONDUCTOR 731
it, must have the same potential. Since the surface of the conductor is an
equipotential surface, E for points on the surface must be at right angles to
the surface.
We saw in Section 28-4 that a charge placed on an insulated conductor will
spread over the surface until E equals zero for all points inside. We now have
an alternative way of saying the same thing; the charge will move until all
points of the conductor (surface points and interior points) are brought to the
same potential, for if V is constant in the conductor then E is zero every-
where in the conductor (E 1 = -dV /dl).
Figure 29-18a is a plot of potential against radial distance for an isolated
spherical conducting shell of LO-meter radius carrying a positive charge of
1.0 X 10-6 coul. For points outside the shell V(r) can be calculated from
Eq. 29-8 because the charge q behaves, for such points, as if it were concen-
trated at the center of the sphere. Equation 29-8 is correct right up to the
surface of the shell. Now let us push the test charge through the surface,
assuming that there is a small hole, and into the interior. No extra work is
needed because no electrical forces act on the test charge once it is inside the
shell. Thus the potential everywhere inside is the same as that on the sur-
face, as Fig. 29-18a shows.
Figure 29-18b shows the electric field strength for this same sphere. Note
that E equals zero inside. The lower of these curves can be derived from the
upper by differentiation with respect tor, using Eq. 29-16; the upper can be
derived from the lower by integration with respect to r, using Eq. 29-6.
Figure 2~F18 holds without change if the conductor is a solid sphere rather
than a spherical shell as we have assumed. It is constructive to compare
Fig. 29-18b (conducting shell or sphere) with Fig. 28-9, which holds for a
nonconducting sphere. The student should try to understand the difference
between these two figures, bearing in mind that in the first the charge lies on
the surface whereas in the second it was assumed to have been spread uni-
formly throughout the volume of the sphere.
Finally we note that, as a general rule, the charge density tends to be high
on isolated conducting surfaces whose radii of curvature are small, and con-
versely. For example, the charge density tends to be relatively high on
sharp points and relatively low on plane regions on a conducting surface.
The electric field strength Eat points immediately above a charged surface
is proportional to the charge density u so that E may also reach very high
values near sharp points. Glow discharges from sharp points during thunder-
storms are a familiar example. The lightning rod acts in this way to
neutralize charged clouds and thus prevent lightning strokes.
We can examine the qualitative relationship between u and the curvature of the
surface in a particular case by considering two spheres of different radii connected
by a very long fine wire; see Fig. 29-19. Suppose that this entire assembly is raised
732 ELECTRIC POTENTIAL Chap. 29
Wire
to some arbitrary potential V. The (equal) potentials of the two spheres are, from
Eq. 29-8, *
where q1 is the charge on the sphere of radius R1 and q2 is the charge on the sphere of
radius R2.
The surface charge densities for each sphere are given by
and
Dividing gives
CT1 q1 Ri
CT2 = ~ Yil
Combining with Eq. 29-18 yields
which is consistent with our qualitative statement above. Note that the larger sphere
has the larger total charge but the smaller charge density.
The fact that er, and thus E, can become very large near sharp points is
important in the design of high-voltage equipment. Corona discharge can
result from such points if the conducting object is raised to high potential and
surrounded by air. Normally air is thought of as a nonconductor. However,
it contains a small number of ions produced, for example, by the cosmic rays.
A positively charged conductor will attract negative ions from the surround-
ing air and thus will slowly neutralize itself.
* Equation 29-8 holds only for an isolated point charge or spherically symmetric charge
distribution. The spheres must be assumed to be so far apart that the charge on either
one has a negligible effect on the distribution of charge on the other.
Sec. 29-9 THE ELECTROSTATIC GENERATOR 733
If the charged conductor has sharp points, the value of E in the air near the
points can be very high. If the value is high enough, the ions, as they are
drawn toward the conductor, will recC'ive such large accelerations that, by
collision with air molecules, they will produce vast additional numbers of
ions. The air is thus made much more conducting, and the discharge of the
conductor by this corona discharge may be very rapid indeed. The air
surrounding sharp conducting points may even glow visibly because of light
emitted from the air molecules during these collisions.
K = qV. (29-19)
The electron volt can be used interchangeably with any other energy unit.
Thus a 10-gm object moving at 1000 cm/sec can be said to have a kinetic
energy of 3.1 X 10 18 ev. Most physicists would prefer to express this result
as 0.50 joule, the electron volt being inconveniently small. In atomic and
nuclear problems, however, the electron volt (ev) and its multiples the Mev
( = 10 6 ev), the Gev ( = 109 ev) and the Tev ( = 10 12 ev) are the usual units
of choice.
734 ELECTRIC POTENTIAL Chap. 29
~ Example 12. The electrostatic generator. Figure 29-20, which illustrates the
basic operating principle of the electrostatic generator, shows a small sphere of radius
r placed inside a large spherical shell of radius R. The two spheres carry charges q
and Q, respectively. Calculate their potential difference.
The potential of the large sphere is caused in part by its own charge and in part
because it lies in the field set up by the charge q on the small sphere. From Eq. 29-8,
V R = __
l
47!'Eo
('lR + i)
R
·
The potential of the small sphere is caused in part by its own charge and in part
because it is inside the large sphere; see Fig. 29-18a. From Eq. 29-8,
Silk thread
Insulator
Sec. 29-9 THE ELECTROSTATIC GENERATO~ 735
Insulating belt
Manhole
Movable platform
Lead shielding
Feet
Fig. 29-21 An electrostatic generator at MIT capable of producing 9-Mev protons. The
proton beam is accelerated vertically downward, being deflected into a horizontal plane
by the analyzing magnet shown at the bottom. (Courtesy of J. G. Trump.)
736 ELECTRIC POTENTIAL Chap. 29
From another I?oint of view, we note that since the spheres when electrically con-
nected. form a ~mgle conductor at electrostatic equilibrium there can be only a single
potential. This means that Vr - V R = 0, which can occur only if q = O.
In actual ele?trosta~ic genera~ors charge is_ carried into the shell on rapidly moving
belts made of msulatmg material. Charge 1s "sprayed" onto the belts outside the
shell by corona discharge from a series of sharp metallic points connected to a source
of moderately high potential difference. Charge is removed from the belts inside the
shell by a similar series of points connected to the shell. Electrostatic generators can
be built commercially to accelerate protons to energies up to 10 Mev, using a single
acceleration. Figure 29-21 shows a schematic diagram of an electrostatic generator
at MIT that can produce 9-Mev protons.
Generators can be built in which the accelerated particles are subject to successive
accelerations. An electrostatic generator at the University of Pittsburgh subjects
particles to three successive accelerations.
QUESTIONS
1. Are we free to call the potential of the earth + 100 volts instead of zero? What effect
would such an assumption have on measured values of (a) potentials and (b) potential
differences?
2. What would happen to a person on an insulated stand if his potential was increased
by 10,000 volts?
3. Do electrons tend to go to regions of high potential or of low potential?
4. Suppose that the earth has a net charge that is not zero. Is it still possible to adopt
the earth as a standard reference point of potential and to assign the potential V = 0 to it?
5. Does the potential of a positively charged insulated conductor have to be positive?
Give an example to prove your point.
6. Can two different equipotential surfaces intersect?
7. If E equals zero at a given point, must V equal zero for that point? Give some ex-
amples to prove your answer.
8. If you know E at a given point, can you calculate V at that point? If not, what
further information do you need?
9. If V equals a constant throughout a given region of space, what can you say about
E in that region?
10. In Section 16-6 we saw that the gravitational field strength is zero inside a spherical
shell of matter. The electrical field strength is zero not only inside an isolated charged
spherical conductor but inside an isolated conductor of any shape. Is the gravitational
field strength inside, say, a cubical shell of matter zero? If not, in what respect is the
analogy not complete?
11. How can you insure that the electric potential in a given region of space will have
a constant value?
12. An isolated conducting spherical shell carries a negative charge. What will happen
if a positively charged metal object is placed in contact with the shell interior? Assume
that the positive charge is (a) less than, (b) equal to, and (c) greater than the negative
charge in magnitude.
13. An uncharged metal sphere suspended by a silk thread is placed in a uniform ex-
ternal electric field E. What is the magnitude of the electric field for points inside the
sphere? Is your answer changed if the sphere carries a charge?
14. A charge is placed on an insulated conductor in the form of a perfect cube. What
will be the relative charge density at various points on the cube (surfaces, edges, corners);
what will happen to the charge if the cube is in air?
Chap. 29 PROBLEMS 737
PROBLEMS
1. An infinite charged sheet has a surface charge density u of 1.0 X 10-7 coul/meter2 .
How far apart are the equipotential surfaces whose potentials differ by 5.0 volts?
2. A charge q is distributed uniformly throughout a nonconducting spherical volume of
radius R. (a) Show that the potential a distance a from the center, where a < R, is
given by
q(3R 2 - a 2)
V= .
81rEoR 3
(b) Is it reasonable that, according to this expression, V is not zero at the center of the
sphere?
/---3:-A charge of 10-s coul can be produced by simple rubbing. To what potential would
such a charge raise an insulated conducting sphere of 10-cm radius?
4. Consider a point charge with q = 1.5 X 10-s coul. (a) What is the radius of an
equipotential surface having a potential of 30 volts? (b) Are surfaces whose potentials
differ by a constant amount (say 1.0 volt) evenly spaced in radius?
5. In Fig. 29-22, locate the points (a) where V = 0 and (b) where E = 0. Consider
only points on the axis and choose d = 1.0 meter.
fig. 29-22
6. In Fig. 29-22 (see Problem 5) sketch qualitatively (a) the lines of force and (b) the
intersections of the equipotential surfaces with the plane of the figure. (Hint: Consider
the behavior close to each point charge and at considerable distances from the pair of
charges.)
~-In Fig. 29-~derive an expression for VA - VB· Does your result reduce to the
expected answer when d = O? When q = O?
B ••---11~-------•A
q
(a)
----1-E- d ---------;.1
•
+q •
-3q
q
(b)
,.$. A point charge has q = + 1.0 X 10-6 coul. Consider point A which is 2.0 meters
distant and point B which is 1.0 meter distant in a direction diametrically opposite, as in
Fig. 29-24a. (a) What is the potential difference VA - VB? (b) Repeat if points A and B
are located as in Fig. 29-24b.
9. Calculate the dipole moment of a water molecule under the assumption that all ten
electrons in the molecule circulate- symmetrically about the oxygen atom, that the OH
distance is 0.96 X 10-s cm, and that the angle between the two OH bonds is 104 °. Com-
pare with the value quoted on p. 719; see Fig. 29-11.
738 ELECTRIC POTENTIAL Chap. 29
10. For the charge configuration of Fig. 29-25, show that V(r) for points on the vertical
axis, assuming r » a, is given by
V = _l _
41Teo
(g + 2qa)
r r2
.
Is this an expected result? (Hint: The charge configuration can be viewed as the sum of
an isolated charge and a dipole.)
-,--____.p
+q
+q
-q
Fig. 29-25
11. In a typical lightning flash the potential difference between discharge points is
about 109 volts and the quantity of charge transferred is about 30 coul. How much ice
would it melt at 0°C if all the energy released could be used for this purpose?
12. Calculate (a) the electric potential established by the nucleus of a hydrogen atom
at the mean distance of the circulating electron (r = 5.3 X 10-11 meter), (b) the electric
potential energy of the atom when the electron is at this radius, and (c) the kinetic energy
of the electron, assuming it to be moving in a circular orbit of this radius centered on the
nucleus. (d) How much energy is required to ionize the hydrogen atom? Express all
energies in electron volts.
13. What is the electric potential energy of the charge configuration of Fig. 29-7? Use
the numerical values of Example 5.
14. (a) A spherical drop of water carrying a charge of 3 X 10-11 coul has a potential of
500 volts at its surface. What is the radius of the drop? (b) If two such drops, of the
same charge and radius, combine to form a single spherical drop, what is the potential at
the surface of the new drop so formed?
15. If the earth had a net charge equivalent to 1 electron/meter2 of surface area, (a)
what would the earth's potential be? (b) What would the electric field due to the earth
be just outside its surface?
16. Figure 29-26 shows an idealized representation of a U 238 nucleus (Z = 92) on the
verge of fission. Calculate (a) the repulsive force acting on each fragment and (b) the
mutual electric potential energy of the two fragments. Assume that the fragments are
Fig. 29-26
Chap. 29 PROBLEMS 739
equal in size and charge, spherical, and just touching. The radius of the initially spherical
U 238 nucleus is 8.0 X 10-15 meter. Assume that the material out of which nuclei are
made has a constant density.
17. In the Millikan oil drop experiment (see Fig. 27-24) an electric field of 1.92 X 105
nt/coul is maintained at balance across two plates separated by 1.50 cm. Find the
potential difference between the plates.
/--Hl':'(a) Show that the electric potential at a point on the axis of a ring of charge of
radius a, computed directly from Eq. 29-10, is given by
V=-1- q
4no V x2 + a2
(b) From this result derive an expression for E at axial points; compare with the direct
calculation of E in Example 5, Chapter 27.
~ In Example 6 the potential at an axial point for a charged disk was shown to be
V = __!!__ ( V a 2 + r2 - r).
2eo
From this result show that E for axial points is given by
E---
u ( 1- r ) ·
- 2eo Va2 + r2
Does this expression for E reduce to an expected result for (a) r » a and (b) for r = O?
20. (a) Starting from Eq. 29-11, find the magnitude Er of the radial component of the
electric field due to a dipole. (b) For what values of e is Er zero?
21. Can a conducting sphere 10 cm in radius hold a charge of 4 X 10-6 coul in air
without breakdown? The dielectric strength (minimum field required to produce break-
down) of air at 1 atm is 3 X 106 volts/meter.
22. A Geiger counter has a metal cylinder 2.0 cm in diameter along whose axis is
stretched a wire 0.005 in. in diameter. If 850 volts are applied between them, what is the
electric field strength at the surface of (a) the wire and (b) the cylinder?
23. Two metal spheres are 3.0 cm in radius and carry charges of + 1.0 X 10-s coul and
-3.0 X 10-s coul, respectively, assumed to be uniformly distributed. If their centers
are 2.0 meters apart, calculate (a) the potential of the point halfway between their centers
and (b) the potential of each sphere.
-----24-:"' In Fig. 29-19 let R 1 = 1.0 cm and R2 = 2.0 cm. Before the spheres are connected
by the fine wire, a charge of 2.0 X 10-7 coul is placed on the smaller sphere, the larger
sphere being uncharged. Calculate (a) the charge, (b) the charge density, and (c) the
potential for each sphere after they are connected.
25. The metal object in Fig. 29-27 is a figure of revolution about the horizontal axis.
If it is charged negatively, sketch roughly a few equipotentials and lines of force. Use
physical reasoning rather than mathematical analysis.
Axis
Fig. 29-27
740 ELECTRIC POTENTIAL Chap. 29
26. Devise an arrangement of three point charges, separated by finite distances, that
has zero electric potential energy.
27. Derive an expression for the work required to put the four charges together as
indicated in Fig. 29-28.
+q -q
a ...
a a
-q a +q
Fig. 29-28
28. A gold nucleus contains a positive charge equal to that of 79 protons. An a-particle
(Z = 2) has a kinetic energy K at points far from this charge and is traveling directly
toward the charge. The particle just touches the surface of the charge (assumed spherical)
and is reversed in direction. Calculate K, assuming a nuclear radius of 5.0 X 10-15
meter. The actual a-particle energy used in the experiments of Rutherford and his col-
laborators was 5.0 Mev. What do you conclude?
29. What is the potential gradient, in volts/meter, at a distance of 10-12 meter from
the center of the gold nucleus? What is the gradient at the nuclear surface?
30. For the spheres of Fig. 29-19, what is the ratio of electric field strengths at the sur-
face?
31. (a) Through what potential difference must an electron fall, according to Newtonian
mechanics, to acquire a speed v equal to the speed c of light? (b) Newtonian mechanics
fails as v ---> c. Therefore, using the correct relativistic expression for the kinetic energy
K = m 0 c2 [
Vl -
1
(v/c) 2
- 1 J
in place of the Newtonian expression K = tm v
0 2, determine the actual electron speed ac-
quired in falling through the potential difference computed in (a). Express this speed
as an appropriate fraction of the speed of light.
32. Two thin insulated concentric conducting spheres of radii R1 and R2 carry
charges q1 and q2. Derive expressions for E(r) and V(r), where r is the distance from
the center of the spheres. Plot E(r) and V(r) from r = 0 tor = 4.0 meters for R1 = 0.50
meter, R2 = 1.0 meter, q 1 = +2.0 X 10-6 coul, and q2 = +1.0 X 10-6 coul. Compare
with Fig. 29-18.
33. Let the potential difference between the shell of an electrostatic generator and the
point at which charges are sprayed onto the moving belt be 3.0 X 106 volts. If the belt
transfers charge to the shell at the rate of 3.0 X 10- 3 coul/sec, what power must be pro-
vided to drive the belt, considering only electrical forces?
34. (a) How much charge is required to raise an isolated metallic sphere of LO-meter
radius to a potential of 1.0 X 10 6 volts? Repeat for a sphere of 1.0-cm radius. (b) Why
use a large sphere in an electrostatic generator since the same potential can be achieved
for a smaller charge with a small sphere?
35. An alpha particle is accelerated through a potential difference of one million volts
in an electrostatic generator. (a) What kinetic energy does it acquire? (b) What kinetic
energy would a proton acquire under these same circumstances? (c) Which particle would
acquire the greater speed, starting from rest?
Capacitors and Dielectrics
CHAPTER 30
30-1 Capacitance
In Section 29-3 we showed that the potential of a charged conducting
sphere, assumed to be completely isolated with no other bodies (conducting
or nonconducting) nearby, is given by
1 q
v +' = ---, (30-1)
41reo R
in which q is the charge on the sphere and R is the sphere radius. The sub-
script on Vindicates that we assume the charge to be positive. We represent
this potential in Fig. 30-1 by the line marked V +'· The line marked V 00
in that figure represents the potential of an infinitely distant reference posi-
tion; it has been assigned the value zero, following the usual convention.
V'
+
-----
V_'
Fig. 30-1 The potential difference between two conductors that carry constant, equal,
and opposite charges is reduced as the conductors are brought closer together.
741
742 CAPACITORS AND DIELECTRICS Chap. 30
Let us now imagine a second sphere of radius R, carrying a negative charge
-q and located a large distance (>>R) from the first sphere so that each may
still be considered to be electrically isolated. The potential of the second
sphere is given by
1 q
V_' = - ----, (30-2)
4'71"Eo R
and this quantity also is represented in Fig. 30- 1.
The potential difference V' between the two spheres is
1 2q
V' = V +' - V _, = - - - ·
4noR
This shows that V', the potential difference, and q, the magnitude of the
charge on either sphere, are proportional to each other. We may rewrite the
equation as
q = (2'71"E 0R) V' = C'V', (30-3)
in which the proportionality constant in parentheses is called the capacitance
of the two spheres and assigned the symbol C'.
Let us move the two spheres close together. The presence of each will now
spoil the spherical symmetry of the lines of force emanating from the other.
Lines from a given sphere, which, for large sphere separations, radiated uni-
formly in all directions to infinity, now terminate, in part, on the other
sphere. Under these conditions Eqs. 30-1 and 30-2 no longer apply, since
they were derived (see Section 29-3) on the assumption that spherical
symmetry existed, permitting the useful application of Gauss's law.
A positive charge brought near to an isolated object serves to raise the
potential of that object and a negative charge serves to lower it, as the
student can see by considering the work required to move a positive test
charge from infinity to points near such charges. Thus the potential of the
positively charged sphere will be lowered by the presence nearby of the nega-
tively charged sphere, from V +' to some lower value V + Similarly the
potential of the negative sphere will be raised from V _, to a higher value V -·
These new potentials are shown in Fig. 30-1, the potential changes for each
sphere being suggested by the vertical arrows.
From Fig. 30-1, although the charges on the spheres have not changed, it is
clear that the potential difference between the spheres has been considerably
reduced. Put another way, the capacitance of the system of two spheres (see
Eq. 30-3), defined from
q
c =-, (30-4)
v
has been made considerably larger than its initial value C' by bringing the
spheres closer together.
It is also possible to use Eq. 30-4 to define the capacitance of a single iso-
lated conductor such as a sphere. In such cases one may imagine that the
Sec. 30-1 CAPACITANCE 743
second "plate," carrying an equal and opposite charge, is a conducting sphere
of very large-essentially infinite-radius centered about the conductor.
The potential of this infinitely distant sphere, according to the usual conven-
tion for potential measurements, is zero. The capacitance of an isolated
sphere of radius R is given from Eqs. 30-4 and 30-1 as
q
C = - = 41r1: 0R.
v
The mks unit of capacitance that follows from Eq. 30-4 is the coul/volt.
A special unit, the farad, is used to represent it. It is named in honor of
Michael Faraday who, among other contributions, developed the concept of
capacitance. Thus
1 farad = 1 coul/volt.
The submultiples of the farad, the microfarad (1 µf = 10-5 farad) and the
micromicrofarad (1 µµf = 10- 12 farac!), are more convenient units in practice.
An analogy can be made between a capacitor carrying a charge q and a rigid con-
tainer of volume 'U containing µ moles of an ideal gas.
The gas pressure pis directly proportional toµ for a fixed temperature, according
to the ideal gas law ~Eq. 23-2)
µ = (:r) p.
For the capacitor (Eq. 30-4)
q = (C)V.
Comparison shows that the capacitance C of the capacitor, assuming a fixed tem-
perature, is analogous to the volume 'U of the container.
Note that any amount of charge can be put on the capacitor, and any mass of gas
can be put in the container, up to certain limits. These correspond to electrical
breakdown ("arcing over") for the capacitor and to rupture of the walls for the con-
tainer.
Figure 30-2 shows a more general case of two nearby conductors, which are
now permitted to be of any shape, carrying equal and opposite charges. Such
an arrangement is called a capacitor, the conductors being called plates. The
equal and opposite charges might be established by connecting the plates
momentarily to opposite poles of a battery. The capacitance C of any
capacitor is defined from Eq. 30-4 in which we remind the student that Vis
the potential difference between the plates and q is the magnitude of the charge
on either plate; q must not be taken as the net charge of the capacitor, which is
zero. The capacitance of a capacitor depends on the geometry of each plate,
their spatial relationship to each other, and the medium in which the plates
are immersed. For the present, we take this medium to be a vacuum.
Capacitors are very useful devices, of great interest to physicists and
engineers. For example:
1. In this book we stress the importance of fields to the understanding of
natural phenomena. A capacitor can be used to establish desired electric
field configurations for various purposes. In Section 27-5 we described the
deflection of an electron beam in a uniform field set up by a capacitor, al-
744 CAPACITORS AND DIELECTRICS Chap. 30
Fig. 30-2 Two insulated conductors carrying equal and opposite charges form a capacitor.
though we did not use this term in that section. In later sections we discuss
the behavior of dielectric materials when placed in an electric field (provided
conveniently by a capacitor) and we shall see how the laws of electromagne-
tism can be generalized to take the presence of dielectric bodies more readily
into account.
2. A second important concept stressed in this book is energy. By ana-
lyzing a charged capacitor we show that electric energy may be considered to
be stored in the electric field between the plates and indeed in any electric
field, however generated. Because capacitors can confine strong electric
fields to small volumes, they can serve as useful devices for storing energy.
In many electron synchrotrons, which are cyclotron-like devices for accele-
rating electrons, energy accumulated and stored in a large bank of capacitors
over a relatively long period of time is made available intermittently to
accelerate the electrons by discharging the capacitor in a much shorter time.
Many researches and devices in plasma physics also make use of bursts of
energy stored in this way.
3. The electronic age could not exist without capacitors. They are used, in
conjunction with other devices, to reduce voltage fluctuations in electronic
povver supplies, to transmit pulsed signals, to generate or detect electro-
magnetic oscillations at radio frequencies, and to provide time delays. In
most of these applications the potential difference between the plates will not
be constant, as we assume in this chapter, but will vary with time, often in a
sinusoidal or a pulsed fashion. In later chapters we consider some aspects
of the capacitor used as a circuit element.
Sec. 30-2 CALCULATING CAPACITANCE 745
where Vis the difference in potential between the plates. The integral may
Gaussian
surface
Fig. 30-3 A parallel-plate capacitor with plates of area A. The dashed line represents a
Gaussian surface whose height is hand whose top and bottom caps are the same shape and
size as the capacitor plates.
746 CAP A CITO RS AND DIELECTRICS Chap. 30
be taken over any path that starts on one plate and ends on the other because
each platf is an equipotential surface and the electrostatic force is path
independeht. Although the simplest path between the plates is a perpen-
dicular straight line, Eq. 30-6 follows no matter what path of integration
we choose.
If we substitute Eqs. 30-5 and 30-6 into the relation C = q/V, we obtain
q eoEA eoA
C = - = - = -· (30-7)
V Ed d
Equation 30-7 holds only for capacitors of the parallel-plate type; different
formulas hold for capacitors of different geometry.
In Section 26-4 we stated that e0 , which we first met in connection with
Coulomb's law, was not measured in terms of that law because of experi-
mental difficulties. Equation 30-7 suggests that e0 might be measured by
building a capacitor of accurately known plate area and plate spacing and
determining its capacitance experimentally by measuring q and V in the
relation C = q/V. Thus Eq. 30-7 can be solved for e0 and a numerical value
found in terms of the measured quantities A, d, and C; e0 has been measured
accurately in this way .
.... Example 1. The parallel plates of an air-filled capacitor are everywhere 1.0 mm
apart. What must the plate area be if the capacitance is to be 1.0 farad?
From Eq. 30-7
This is the area of a square sheet more than 6 miles on edge; the farad is indeed a large
unit.
Example 2. A cylindrical capacitor. A cylindrical capacitor consists of two
coaxial cylinders (Fig. 30-4) of radius a and band length Z. What is the capacitance
of this device? Assume that the capacitor is very long (that is, that l » b) so that
fringing of the lines of force at the ends can be ignored for the purpose of calculating
the capacitance.
gives EoE(21rr)(l) = q,
the flux being entirely through the cylindrical surface and not through the end caps.
Solving for E yields
q
E=--·
21rEorl
The potential difference between the plates is given by Eq. 29-5 [note that E and
dl ( = dr) point in opposite directions] or
C = !l = 2nol
V ln (b/a)
Like the relation for the parallel-plate capacitor (Eq. 30-7), this relation also depends
only on geometrical factors.
Example 3. Capacitors in parallel. Figure 30-5 shows three capacitors connected
in parallel. What single capacitance C is equivalent to this combination? "Equiva-
lent" means that if the parallel combination and the single capacitance were each
in a box with wires a and b connected to terminals, it would not be possible to dis-
tinguish the two by electrical measurements external to the box.
capacitors in parallel.
b 1 IC1 I
C2 Cs
The potential difference across each capacitor in Fig. 30-5 will be the same. This
follows because all of the upper plates are connected together and to terminal a,
whereas all of the lower plates are connected together and to terminal b. Applying
the relation q = CV to each capacitor yields
and
The total charge q on the combination is
q = q1 + qz + q3
= (C1 + C2 + C3)V.
The equivalent capacitance C is
V1 V2 V3
i~----1~
IC1 C2 I
Fig. 30-6 Example 4. Three capacitors in
-q ! +q -q ! +q -q +q senes.
L ____ _J
~-----v-----.a
a b
+++++ +++++
;w inmm•"-•nwm1 m
B
(a)
++ ++
(b)
Fig. 30-7 (a) Battery B supplies the same potential difference to each capacitor; the one
on the right has the higher charge. (b) Both capacitors carry the same charge; the one on
the right has the lower potential difference, as indicated by the meter readings.
C, will the induction between them remain the same? Will the relation of C and B
to A be unaltered, notwithstanding the difference of the dielectrics interposed
between them?
Since q is larger, for the same V, if a dielectric is present, it follows from the
relation C = q/V that the capacitance of a capacitor increases if a dielectric is
placed between the plates. The ratio of the capacitance with the dielectric * to
that without is called the dielectric constant K of the material; see Table 30-1.
Instead of maintaining the two capacitors at the same potential difference,
we can place the same charge on them, as in Fig. 30-7!:i. Experiment then
shows that the potential difference Va between the plates of the right-hand
capacitor is smaller than that for the left-hand capacitor by the factor 1/ K, or
Va= Vo/K.
We are led once again to conclude, from the relation C = q/V, that the effect
of the dielectric is to increase the capacitance by a factor K.
Table 30-1
PROPERTIES OF SOME DIELECTRICS
Dielectric
Dielectric Strength *
Material Constant (kv /mm)
Vacuum 1.00000 00
* This isthe maximum potential gradient that may exist in the dielectric without
the occurrence of electrical breakdown. Dielectrics are often placed between con -
ducting plates to permit a higher potential difference to be applied between them
than would be possible with air as the dielectric.
In such materials (called polar) the electric dipole moments p tend to align
themselves with an external electric field, as in Fig. 30-8; see also Section
27-6. Because the molecules are in constant thermal agitation, the degree of
alignment will not be complete but will increase as the applied electric field is
increased or as the temperature is decreased.
Whether or not the molecules have permanent electric dipole moments,
they acquire them by induction when placed in an electric field. In Section
29-5 we saw that the external electric field tends to separate the negative and
the positive charge in the atom or molecule. This induced electric dipole
moment is present only when the electric field is present. It is proportional
to the electric field (for normal field strengths) and is created already lined up
with the electric field as Fig. 29-12 suggests.
Let us use a parallel-plate capacitor, carrying a fixed charge q and not
connected to a battery (see Fig. 30-7b ), to provide a uniform external electric
field E 0 into which we place a dielectric slab. The over-all effect of alignment
and induction is to separate the center of positive charge of the entire slab
slightly from the center of negative charge. The slab, as a whole, although
remaining electrically neutral, becomes polarized, as Fig. 30-9b suggests.
The net effect is a pile-up of positive charge on the right face of the slab and of
negative charge on the left face; within the slab no excess charge appears in
any given volume element. Since the slab as a whole remains neutral, the
positive induced surface charge must be equal in magnitude to the negative
induced surface charge. Note that in this process electrons in the dielectric
-
-------\.:~----:;=;:::-----r-,-;-.>n----~
@ ~
~@ e---:
@
G,3
G{J)
(j)
(a) (b)
Fig. 30-8 (a) :VIolecules with a permanent electric dipole moment, showing their random
orientation in the absence of an external electric field. (b) An electric field is applied,
producing partial alignment of the dipoles. Thermal agitation prevents complete align-
ment.
752 CAPACITORS AND DIELECTRICS Chap. 30
E 0 =0 i-------N
E0 I
(a) (c)
Fig. 30-9 (a) A dielectric slab, showing the random distribution of plus and minus
charges. (b) An external field Eo, established by putting the slab between the plates of a
parallel-plate capacitor (not shown), separates the center of plus charge in the slab slightly
from the center of minus charge, resulting in the appearance of surface charges. No net
charge exists in any volume element located in the interior of the slab. (c) The surface
charges set up a field E' which opposes the external field Eo associated with the charges on
the capacitor plates. The resultant field E ( = Eo + E') in the dielectric is thus less than
Eo.
are displaced from their equilibrium positions by distances that are con-
siderably less than an atomic diameter. There is no transfer of charge over
macroscopic distances such as occurs when a current is set up in a conductor.
Figure 30-9c shows that the induced surface charges will always appear
in such a way that the electric field set up by them (E') opposes the external
electric field E 0 . The resultant field in the dielectric E is the vector sum of
E 0 and E'. It points in the same direction as E 0 but is smaller. If a d1'electric
is placed in an electric field, induced surface charges appear which tend to
weaken the original field within the dielectric.
This weakening of the electric field reveals itself in Fig. 30-7b as a reduc-
tion in potential difference between the plates of a charged isolated capacitor
when a dielectric is introduced between the plates. The relation V = Ed
for a parallel-plate capacitor (see Eq. 30-6) holds whether or not dielectric is
present and shows that the reduction in V described in Fig. 30-7b is directly
connected to the reduction in E described in Fig. 30-9. More specifically, if a
dielectric slab is introduced into a charged parallel-plate capacitor, then
Eo Vo
- K (30-9)
E Va
where the symbols on the left refer to Fig. 30-9 and the symbols VO and Va
refer to Fig. 30-7b. *
Induced surface charge is the explanation of the most elementary fact of
static electricity, namely, that a charged rod will attract uncharged bits of
* Equation 30-9 does not hold if the battery remains connected while the dielectric
slab is introduced. In this case V (hence E) could not change. Instead, the charge q on
the capacitor plates would increase by a factor K, as Fig. 30-7 a suggests.
Sec. 30-5 DIELECTRICS AND GAUSS'S LAW 753
Fig. 30-10 A charged rod attracts an uncharged piece of paper because unbalanced
forces act on the induced surface charges.
paper, etc. Figure 30-10 shows a bit of paper in the field of a charged rod.
Surface charges appear on the paper as shown. The negatively charged end
of the paper will be pulled toward the rod and the positively charged end will
be repelled. These two forces do not have the same magnitude because the
negative end, being closer to. the rod, is in a stronger field and experiences a
stronger force. The net effect is an attraction. A dielectric body in a uni-
form electric field will not experience a net force.
-r-q
(a)
(b)
Fig. 30-11 A parallel-plate capacitor (a) without and (b) with a dielectric. The charge q
on the plates is assumed to be the same in each case.
754 CAP A CITO RS AND DIELECTRICS Chap. 30
If no dielectric is present (Fig. 30-lla), Gauss's law (see Eq. 30-5) gives
t:ofE·dS = t:oEoA = q
q
or Eo=-· (30-10)
t:oA
If the dielectric is present (Fig. 30-llb), Gauss's law gives
Eo
E=-·
K
Eo q
E=-=-· (30-12)
K Kt:oA
Inserting this in Eq. 30-11 yields
q q'
(30-13a)
or (30-13b)
This shows correctly that the induced surface charge q' is always less in
magnitude than the free charge q and is equal to zero if no dielectric is present,
that is, if K = 1.
Now we write Gauss's law for the case of Fig. 30-llb in the form
(30-14)
q - q' again being the net charge within the Gaussian surface. Substituting
from Eq. 30-13b for q' leads, after some rearrangement, to
(30-15)
Sec. 30-5 DIELECTRICS AND GAUSS'S LAW 755
- d - 10-2 meter - · µµ ·
eof KE·dS = e E A = q,
0 0
V = -
u
L
E·dl = -
u
L
E cos 180° dl =
u
L
E dl = E 0(d - b) + Eb.
Numerically
V = (1.0 X 104 volts/meter)(5 X 10- 3 meter)
+ (0.14 X 104 volts/meter)(5 X 10- 3 meter) = 57 volts.
This contrasts with the original applied potential difference of 100 volts; compare Fig.
30-7b.
(f) Calculate the capacitance with the slab in place.
From Eq. 30-4,
C = !!___ = 8.9 X 10- 10 coul = 16 µµf.
V 57 volts
When the dielectric slab is introduced, the potential difference drops from 100 to 57
volts and the capacitance rises from 8.9 to 16 µµf, a factor of 1.8. If the dielectric
slab had filled the capacitor, the capacitance would have risen by a factor of K ( = 7.0)
to 62 µµf. _,..
The quantity in parentheses (see Eq. 30-12) is simply the electric field strength E in
the dielectric. The last term in Eq. 30-16 is the induced surface charge per unit area.
We call it the electric polarization P, or
q'
p = -· (30-17)
A
Sec. 30-6 THREE ELECTRIC VECTORS 757
(P= 0)
-q'
+q'
-q
(a)
'-----r-----1
D
(b)
Fig. 30-13 (a) Showing D, EOE, and P in the dielectric (upper right) and in the gap
(upper left) for a parallel-plate capacitor. (b) Showing samples of the lines associated with
D (free charge), EOE (all charges), and P (polarization charge).
The name is suitable because the induced surface charge q' appears when the dielectric
is polarized.
The electric polarization P can be defined in an equivalent way by multiplying the
numerator and denominator in Eq. 30-17 by d, the thickness of the dielectric slab in
Fig. 30-11,
q'd
P=-· (30-18)
Ad
The numerator is the product q'd of the magnitude of the (equal and opposite)
polarization charges by their separation. It is thus the induced electric dipole mo-
ment of the dielectric slab. Since the denominator Ad is the volume of the slab, we
see that the electric polarization can also be defined as the induced electric dipole
moment per unit volume in the dielectric. This definition suggests that since the
electric dipole moment is a vector the electric polarization is also a vector, its magni-
tude being P. The direction of Pis from the negative induced charge to the positive
induced charge, as for any dipole. In Fig. 30-13, which shows a capacitor with a
dielectric slab filling half the space between the plates, P points down.
We can now rewrite Eq. 30-16 as
q
A= EoE + P. (30-19)
758 CAPACITORS AND DIELECTRICS Chap. 30
The quantity on the right occurs so often in electrostatic problems that we give it the
special name electric displacement D, or
D =EoE + P
in which D = !L.
A
The name has historical significance only.
Since E and Pare vectors, D must also be one, so that in the more general case we
have
D = EoE + P. (30-21)
In Fig. 30-13 all three vectors point down and each has a constant magnitude for
every point in the dielectric (and also at every point in the air gap) so that the vector
nature of Eq. 30-21 is not very important in this case. In more complicated prob-
lems, however, E, P, and D may vary in magnitude and direction from point to point.
From their definitions we see the following:
1. D (see Eq. 30-20b) is connected with the free charge only. We can represent the
vector field of D by lines of D, just as we represent the field of E by lines of force.
Figure 30-13b shows that the lines of D begin and end on the free charges.
2. P (see Eq. 30-17) is connected with the polarization charge only. It is also pos-
sible to represent this vector field by lines. Figure 30-13b shows that the lines of
P begin and end on the polarization (.;!larges.
3. Eis connected with all charges that are actually present, whether free or polari-
zation. The lines of E reflect the presence of both kinds of charge, as Fig. 30-13b
shows. Note (Eqs. 30-17 and 30-20b) that the units for P and D (coul/meter 2 ) differ
from those of E (nt/coul).
The electric field vector E, which is what determines the force that acts on a suita-
bly placed test charge, remains of fundamental interest. D and P are auxiliary vec-
tors useful as aids in the solution of problems more complex than that of Fig. 30-13.
The vectors D and P can both be expressed in terms of E alone. A convenient
starting point is the identity
Comparison with Eqs. 30-12 and 30-20b shows that this, extended to vector form,
can be written as
D = KEoE. (30-22)
We can also write the polarization (see Eqs. 30-17 and 30--13b) as
P=~=!(1-D-
Since q/ A = D, we can rewrite this, using Eq. 30-22 and casting the result into vector
form, as
p = Eo(K - l)E. (30-23)
This shows clearly that in a vacuum (K = 1) the polarization vector P is zero.*
* Certain waxes, when polarized in their molten state, retain a permanent polarization
after solidifying, even though the external polarizing field is removed. Electrets, manu-
factured in this way, are the electrostatic analog of permanent magnets in that they
possess a gross permanent electric dipole moment. Materials from which electrets can be
constructed are called ferroelectric. Electrets do not obey Eq. 30~23 because they have a
nonvanishing value of P even though E = 0.
Sec. 30-6 THREE ELECTRIC VECTORS 759
Equations 30-22 and 30-23 show that for isotropic materials, to which a single
dielectric constant K can be assigned, D and P both point in the direction of E at any
given point.
The definition of D given by Eq. 30-22 allows us to write Eq. 30-15, that is, Gauss's
law in the presence of a dielectric, simply as
fD·dS = q, (30-24)
where, as before, q represents the free charge only, the induced surface charges being
excluded .
.... Example 6. In Figure 30-13, using data from Example 5, calculate E, D, and
P: (a) in the dielectric and (b) in the air gap.
(a) The electric field in the dielectric is calculated in Example 5 to be 1.43 X 10 3
volts/meter. From Eq. 30-22,
D = K(;oE
P = (;o(K - l)E
(b) The electric field E 0 in the air gap is calculated in Example 5 to be 1.00 X 10 4
volts/meter. From Eq. 30-22,
Do= KEoEo
Note that P vanishes outside the dielectric, D has the same value in the dielectric
and in the gap, and E has different values in the dielectric and in the gap. The student
should verify that Eq. 30-21 (D = e0E + P) is correct both in the gap and in the di-
electric.
It can be shown from Maxwell's equations that no matter how complex the prob-
lem the component of D normal to the surface of the dielectric has the same value on
each side of the surface. In this problem D itself is normal to the surface, there being
no component but the normal one. It can also be shown that the component of E
tangential to the dielectric surface has the same value on each side of the surface.
This boundary condition, like the one for D, is trivial in this problem, both tangential
components being zero. In more complex problems these boundary conditions on D
and E are very important. Table 30-2 summarizes the properties of the electric
vectors E, D, and P. <1111
760 CAP A CITO RS AND DIELECTRICS Chap. 30
Table 30-2
Associated Boundary
Name Symbol
with Condition
Gauss's law when dielectric media are present fD·dS = q Eq. 30--24
(q = free charge only)
that an external agent pulls electrons from the positive plate and pushes
them onto the negative plate, thus bringing about the charge separation;
normally the work of charging is done by a battery, at the expense of its store
of chemical energy.
Suppose that at a time ta charge q'(t) has been transferred from one plate
to the other. The potential difference V(t) between the plates at that
moment will be q'(t)/C. If an extra increment of charge dq' is transferred,
the small amount of additional work needed will be
dW = V dq = ( ~) dq'.
If this process is continued until a total charge q has been transferred, the
total work will be found from
W = I dW = i q q'
o C dq' =
1 q2
2C . (30-25)
w (= U) = tcv 2 . (30-26)
(30-27)
Although this equation was derived for the special case of a parallel-plate
capacitor, it is true in general. If an electric field E exists at any point in
space, we can think of that point as the site of stored energy in amount, per unit
volume, of !KE: 0 E 2 .
762 CAPACITORS AND DIELECTRICS Chap. 30
._ Example 7. A capacitor C1 is charged to a potential difference Vo. This charging
battery is then removed and the capacitor is connected as in Fig. 30-14 to an un-
charged capacitor C2•
(a) What is the final potential difference V across the combination?
Cz
Thus U is less than U0 ! The "missing" energy appears as heat in the connecting
wires as the charges move through them.
Example 8. A parallel-plate capacitor has plates with area A and separation d.
A battery charges the plates to a potential difference Vo. The battery is then dis-
connected, and a dielectric slab of thickness d is introduced. Calculate the stored
energy both before and after the slab is introduced and account for any difference.
The energy U o before introducing the slab is
Uo = !CoVo2•
After the slab is in place, we have
and thus U = i
2CV
2
= i
2K o
C (Vo)2
----; = ;1 U O·
The energy after the slab is introduced is less by a factor 1/K. The "missing" energy
would be apparent to the person who inserted the slab. He would feel a "tug" on the
slab and would have to restrain it if he wished to insert the slab without acceleration.
This means that he would have to do negative work on it, or, alternatively, that the
condenser + slab system would do positive work on him. This positive work is
w = uo - u = tc ov 02 ( 1- ~ )-
Sec. 30-7 ENERGY STORAGE IN AN ELECTRIC FIELD 763
Figure 30-15 shows how the forces that do this work arise, in terms of attraction be-
tween the free charge on the plates and the induced surface charges that appear on
the slab when it is introduced between the plates.
Fig. 30-15 When we introduce a dielectric slab into a charged capacitor, as shown, forces
arise which tend to pull the slab into the capacitor.
The energy density at any radius r is found from Eq. 30-27, assuming K = 1, or
The energy dU that lies in a spherical shell between the radii r and r + dr is
q2 dr
dU = (47rr 2 )(dr)u = -8 - 2 ,
7T'E:o r
where (47rr2)(dr) is the volume of the spherical shell. The total energy U is found by
integration, or
q2 J,oo dr
f
U = d U = 8no R r 2 = 87rE:oR.
q2
Note that this relation follows at once from Eq. 30-25 (U = q2/2C), where C (seep.
651) is the capacitance ( = 47T'E=oR) of an isolated sphere of radius R.
(b) What is the radius Ro of a spherical surface such that half the stored energy lies
within it?
In the equation just given we put
!.[! -
q2
-~
J,Ro dr-
2 ' - 871'€0 R r2
q2 q2 ( 1 1)
or
167T'E:oR = 87T'E:o R - R~ '
which yields, after some rearrangement,
Ro= 2R.
764 CAP A CITO RS AND DIELECTRICS Chap. 30
QUESTIONS
1. A capacitor is connected across a battery. (a) Why does each plate receive a charge
of exactly the same magnitude? (b) Is this true even if the plates are of different sizes?
2. Can there be a potential difference between two adjacent conductors that carry the
same positive charge?
3. The relation rr ex 1/R suggests that the charge placed on an
Foil isolated conductor concentrates on points and avoids flat surfaces,
---------'--=
where R = oo. How do we reconcile this with Fig. 30-3 in which
the charge is definitely on the flat surface of either plate?
4. A sheet of aluminum foil of negligible thickness is placed be-
tween the plates of a capacitor as in Fig. 30-16. What effect has it
Fig. 30-16 on the capacitance if (a) the foil is electrically insulated and (b) the
foil is connected to the upper plate.
5. Discuss similarities and differences when (a) a dielectric slab and (b) a conducting
slab are inserted between the plates of a parallel-plate capacitor. Assume the slab thick-
nesses to be one-half the plate separation.
6. An oil-filled parallel-plate capacitor has been designed to have a capacitance C and
to operate safely at or below a certain maximum potential difference V m without arcing
over. However, the designer did not do a good job and the capacitor occasionally arcs
over. What can be done to redesign the capacitor, keeping C and Vm unchanged and
using the same dielectric?
7. Would you expect the dielectric constant, for substances containing permanent
molecular electric dipoles, to vary with temperature?
8. What is your estimate of the amount by which the center of positive charge and the
center of negative charge are displaced in a situation like that of Fig. 30-9b? An intuitive
guess is all that is called for.
9. For a given potential difference does a capacitor store more or less charge with a di-
electric than it does without a dielectric (vacuum)? Explain in terms of the microscopic
picture of the Rituation.
10. An isolated conducting sphere is given a positive charge. Does its mass increase,
decrease, or remain the same?
11. A dielectric, slab is inserted in one end of a charged parallel-plate capacitor (the
plates being hori7ontal and the charging battery having been disconnected) and then re-
leased. Describe what happens. Neglect friction.
12. A capacitor is charged by using a battery, which is then disconnected. A dielectric
slab is then slipped between the plates. De-
scribe qualitatively what happens to the charge,
the capacitance, the potential difference, the
I
electric field strength, and the stored energy.
13. While a capacitor remains connected to
a battery, a dielectric slab is slipped between
the plates. Describe qualitatively what hap-
pens to the charge, the capacitance, the
potential difference, the electric field strength,
I
and the stored energy. Is work required to Fig. 30-17
insert the slab?
14. Two identical capacitors are connected as shown in Fig. 30-17. A dielectric slab is
slipped between the plates of one capacitor, the battery remaining connected. Describe
qualitatively what happens to the charge, the capacitance, the potential difference, the
electric field strength, and the stored energy for each capacitor.
15. Show that the dielectric constant of a conductor can be taken to be infinitely great.
Chap. 30 PROBLEMS 765
PROBLEMS
/
l. A potential difference of 300 volts is applied to a 2.0-µf capacitor and an 8.0-µf capaci-
tor connected in series. (a) What are the charge and the potential difference for each
capacitor? (b) The charged capacitors are reconnected with their positive plates together
and their negative plates together, no external voltage being applied. What are the charge
and the potential difference for each? (c) The charged capacitors in (a) are reconnected
with plates of opposite sign together. What are the charge and the potential difference for
each?
2. Calculate the capacitance of the earth, viewed as a spherical conductor of radius 6400
km.
){ A 100-µµf capacitor is charged to a potential difference of 50 volts, the charging
b~ttery then being disconnected. The capacitor is then connected, as in Fig. 30-14, to a
second capacitor. If the measured potential difference drops to 35 volts, what is the
capacitance of this second capacitor?
4. If we solve Eq. 30-7 for eo, we see that its mks units are farads/meter. Show that
these units are equivalent to those obtained earlier for eo, namely coul2 /nt-m 2.
5. Figure 30-18 shows two capacitors in series, the rigid center section of length b being "'
movable vertically. Show that the equivalent capacitance of the series combination is
independent of the position of the center section and is given by
eoA
C=--·
a-b
f i l_j
a
L-,....---
Fig. 30-18 Fig. 30-19
6. In Fig. 30-19 a variable air capacitor of the type used in tuning radios is shown.
Alternate plates are connected together, one group being fixed in position, the other group
being capable of rotation. Consider a pile of n plates of alternate polarity, each having an
area A and separated from adjacent plates by a distanced. Show that this capacitor has a
maximum capacitance of
(n - l)eoA
C= .
d
, 7.'A spherical capacitor consists of two concentric spherical shells of radii a and b,;,;
with b > a. Show that its capacitance is
ab
C = 41reo--·
b- a
8. A capacitor has square plates, each of side a, making an angle of () with each other
as shown in Fig. 30-20. Show that for small () the capacitance is given by
C = eoa 2 (l _a())
d 2d
766 CAP A CITO RS AND DIELECTRICS Chap. 30
(Hint: The capacitor may be divided into differential strips which are effectively in
parallel.)
Fig. 30-20
9. Suppose that the two spherical shells of a spherical capacitor have their radii ap-
proximately equal. Under these conditions the device approximates a parallel plate
capacitor with b - a = d. Show that the formula in Problem 7 does indeed reduce to
Eq. 30-7 in this case.
10. A parallel-plate capacitor has circular plates of 8.0-cm radius and 1.0-mm separation.
Wha,tcharge will appear on the plates if a potential difference of 100 volts is applied'?
/n. In Fig. 30-21 find the equivalent capacitance of the combination. Assume that
Ci = 10 .uf, C2 = 5 µf, Cs = 4 µ,f, and V = 100 volts.
12. In Fig. 30-21 suppose that capacitor Cs breaks down electrically, becoming equiva-
lent to a conducting path. What changes in (a) the charge and (b) the potential dif-
ference occur for capacitor C1 ?
I
1
v
C1
I
C2
I
j C31
Fig. 30-21
13. Find the effective capacitance between points x and y in Fig. 30-22. Assume that
C2 = 10 µ,f and that the other capacitors are all 4.0 µ,f. (Hint: Apply a potential difference
V between x and y and write down all the relationships that involve the charges and poten-
tial differences for the separate capacitors.)
Fig. 30-22
Chap. 30 PROBLEMS 767
14. If you have available several 2.0-µf capacitors, each capable of withstanding 200
volts without breakdown, how would you assemble a combination having an equivalent
capaci~nce of (a) 0.40 µfor of (b) 1.2 µf, each capable of withstanding 1000 volts?
VIn Fig. 30-23 find the equivalent capacitance of the combination. Assume that
Ci = 10 µf, C2 = 5 µf, C3 = 4 µf, and V = 100 volts.
I
1
v
l
l
C3
C2
I
Fig. 30-23
16. In Fig. 30-24 the battery B supplies 12 volts. (a) Find the charge on each capacitor
when switch 8 1 is closed and (b) when switch 8 2 is also closed. Take C1 = 1 µf, C 2 = 2 µf,
C3 = 3 µf, and C4 = 4 µf.
....______ ,1-B-----0
Fig. 30-24
<. 17. A parallel-plate capacitor is filled with two dielectrics as in Fig. 30-25. Show that
the capacitance is given by
Check this formula for all the limiting cases that you can think of. (Hint: Can you justify
regarding this arrangement as two capacitors in parallel?)
18. A parallel-plate capacitor is filled with two dielectrics as in Fig. 30-26. Show that
the capacitance is given by
768 CAPACITORS AND DIELECTRICS Chap. 30
Check this formula for all the limiting cases that you can think of. (Hint: Can you justify
regarging this arrangement as two capacitors in series?)
ft~ A dielectric slab of thickness b is inserted between the plates of a parallel-plate
capacitor of plate separation d. Show that the capacitance is given by
KEOA
C=----~
Kd - b(K - 1)
(Hint: Derive the formula following the pattern of Example 5.) Does this formula predict
the correct numerical result of Example 5? Does the formula seem reasonable for the
special cases of b = 0, K = 1, and b = d?
20. A slab of copper of thickness b is thrust into a parallel-plate capacitor as shown in
Fig. 30-27; it is exactly halfway between the plates. What is the capacitance before and
after the slab is introduced?
Fig. 30-27
21. For making a capacitor you have available two plates of copper, a sheet of mica
(thickness = 0.10 mm, K = 6), a sheet of glass (thickness = 2.0 mm, K = 7), and a slab of
paraffin (thickness = 1.0 cm, K = 2). To obtain the largest capacitance, which sheet
(or sheets) should you place between the copper plates?
22. In Example 5, suppose that the 100-volt battery remains connected during the time
that the dielectric slab is being introduced. Calculate (a) the charge on the capacitor
plates, (b) the electric field in the gap, (c) the electric field in the slab, and (d) the capaci-
tance. For all of these quantities give the numerical values before and after the slab is
introduced. Contrast your results with those of Example 5 by constructing a tabular
listing.
23. A parallel-plate capacitor has a capacitance of 100 µ,µ,f, a plate area of 100 cm 2, and a
mica dielectric. At 50 volts potential difference, calculate (a) E in the mica, (b) the free
charge on the plates, and (c) the induced surface charge.
24. Two parallel plates of area 100 cm 2 are each given equal but opposite charges of
8.9 X 10-7 coul. Within the dielectric material filling the space between the plates the
electric field strength is 1.4 X 106 volts/meter. (a) Find the dielectric constant of the
material. (b} Determine the magitude of the charge induced on each dielectric surface.
25. Hard rubber has a dielectric constant of 2.8 and a dielectric strength of 18 X 106
volts/meter. If it is used as the dielectric material in a parallel-plate capacitor, what
minimum area may the plates of the capacitor have in order that the capacitance be
7.0 X 10-2 µ,f and that the capacitor be able to withstand a potential difference of 4000
volts?
26. A cylindrical capacitor has radii a and b as in Fig. 30-4. Show that half the stored
electric potential energy lies within a cylinder whose radius is
r = Vab.
27. An isolated metal sphere whose diameter is 10 cm has a potential of 8000 volts.
What is the energy density at the surface of the sphere?
28. A parallel-plate capacitor has plates of area A and separation d and is charged to a
potential difference V. The charging battery is then disconnected and the plates are pulled
Chap. 30 PROBLEMS 769
apart until their separation is 2d. Derive expressions in terms of A, d, and V for (a) the
new potential difference, (b) the initial and the final stored energy, and (c) the work re-
quired to separate the plates.
29. Show that the plates of a parallel-plate capacitor attract each other with a force
given by
Prove this by calculating the work necessary to increase the plate separation from x to
x + dx.
30. In the capacitor of Example 5 the dielectric slab fills half the space between the
plates. (a) What per cent of the energy is stored in the air gaps? (b) What per cent is
stored in the slab?
31. A parallel-plate air capacitor has a capacitance of 100 µµf. (a) What is the stored
energy if the applied potential difference is 50 volts? (b) Can you calculate the energy
density for points between the plates?
32. For the capacitors of Problem 1, compute the energy stored for the three different
connections of parts (a), (b), and (c). Compare your answers and explain any differences.
33. A charge q is placed on the surface of an originally uncharged soap bubble of radius
Ro. Because of the mutual repulsion of the charged surface, the radius is increased to a
somewhat larger value R. Show that
q = [3s27r 2eopRoR(R2 + RoR + Rl)]Y2
in which p is the pressure of the atmosphere. Find q for p = 1.00 atm, Ro = 2.00 cm,
and R = 2.10 cm. (Hint: The work done by the bubble in pushing back the atmosphere
must equal the decrease in the stored electric field energy that accompanies the expansion,
from the conservation of energy principle.)
34. Two capacitors (2.0 µf and 4.0 µf) are connected in parallel across a 300-volt poten-
tial difference. Calculate the total stored energy in the system.
35. A parallel-connected bank of 2000 5.0-µf capacitors is used to store electric energy.
What does it cost to charge this bank to 50,000 volts, assuming a rate of 2¢/kw-hr?
36. In Fig. 30-21 find (a) the charge, (b) the potential difference, and (c) the stored energy
for each capacitor. Assume the numerical values of Problem 11.
37. In Fig. 30-23 find (a) the charge, (b) the potential difference, and (c) the stored energy
for each capacitor. Assume the numerical values of Problem 15.
Current and Resistance
CHAPTER 31
The current i is the same for all cross sections of a conductor, even though
the cross-sectional area may be different at different points. In the same way
the rate at which water (assumed incompressible) flmvs past any cross section
of a pipe is the same even if the cross sectiorr varies. The water flows faster
where the pipe is smaller and slower where it is larger, so that the volume
rate, measured perhaps in gal/min, remains unchanged. This constancy of
the electric current follows because charge must be conserved; it does not
pile up steadily or drain away steadily from any point in the conductor under
the assumed steady-state conditions. In the language of Section 18-3 there
are no "sources" or "sinks" of charge.
The existence of an electric field inside a conductor does not contradict
Section 28-4, in which we asserted that E equals zero inside a conductor. In
that section, which dealt with a state in which all net motion of charge had
stopped (electrostatics), we assumed that the conductor was insulated and
that no potential difference was deliberately maintained between any two
points on it, as by a battery. In this chapter, which deals with charges in
motion, we relax this restriction.
The electric field that acts on the electrons in a conductor does not pro-
duce a net acceleration because the electrons keep colliding with the atoms
(strictly, ions) that make up the conductor. This array of ions, coupled
together by strong spring-like forces of electric origin, is called the lattice
(see Fig. 21-5). The over-all effect of these collisions is to transfer kinetic
energy from the accelerating electrons into vibrational energy of the lattice.
The electrons acquire a constant average drift speed vd in the direction - E.
The analogy is to a marble rolling down a long flight of stairs and not to a
marble falling freely from the same height. In the first case the acceleration
caused by the (gravitational) field is effectively canceled by the decelerating
effects of collisions with the stair treads so that, under proper conditions, the
marble rolls down the stairs with zero average acceleration, that is, at con-
stant average speed.
Although in metals the charge carriers are electrons, in electrolytes or in
gaseous conductors they may also be positive or negative ions or both. A
convention for labeling the directions of currents is needed because charges
of opposite sign move in opposite directions in a given field. A positive
charge moving in one direction is equivalent in nearly all external effects to a
negative charge moving in the opposite direction. Hence, for simplicity and
algebraic consistency, we assume that all charge carriers are positive and we
draw the current arrows in the direction that such charges would move. If the
charge carriers are negative, they simply move opposite to the direction of
the current arrow (see Fig. 31-1). When we encounter a case (as in the Hall
effect; see Section 33-5) in which the sign of the charge carriers makes a
difference in the external effects, we will disregard the convention and take
the actual situation into account.
Current i is a characteristic of a particular conductor. It is a macroscopic
quantity, like the mass of an object, the volume of an object, or the length
of a rod. A related microscopic quantity is the current density j. It is a
772 CURRENT AND RESISTANCE Chap. 31
<~---E
Fig. 31-1 Electrons drift in a direction opposite to the electric field in a conductor.
j = i/A. (31-3)
The vector j at any point is oriented in the direction that a positive charge
carrier would move at that point. An electron at that point would move in
the direction - j.
The general relationship between j and i is that, for a particular surface in
a conductor, i is the flux of the vector j over that surface, or
i = fj·dS, (31-4)
where dS is an element of surface area and the integral is taken over the
surface in question. Equation 31-3 (written as i = jA) is a special case of
this relationship in which the surface of integration is a cross section of the
conductor and in which j is constant over this surface and at right angles
to it. However, Eq. 31-4 may be applied to any surface through which we
wish to know the current. Equation 31-4 shows clearly that i is a scalar be-
cause the integrand j ·dS is a scalar.
The arrow often associated with the current in a wire does not indicate that current
is a vector but merely shows the sense of charge flow. Positive charge carriers either
move in a certain direction along the wire or in the opposite direction, these two possi-
bilities being represented by+ or - in algebraic equations. Note that (a) the current
in a wire remains unchanged if the wire is bent, tied into a knot, or otherwise distorted,
and (b) the arrows representing the sense of currents do not in any way obey the laws
of vector addition.
. - i - 10 amp - 1300 ;· 2
JAl - A - 0.0079 in.2 - amp m.
The fact that the wires are of different materials does not enter into consideration
here.
Example 2. What is va for the copper wire in Example 1?
We can write the current density for the copper wire as 480 amp/cm 2. To compute
n we start from the fact that there is one free electron per atom in copper. The num-
ber of atoms per unit volume is dN 0 /M where dis the density, No is Avogadro's
number, and M is the atomic weight. The number of free electrons per unit volume
is then
dN 0 (9.0 gm/cm 3)(6.0 X 10 23 atoms/mole)(l electron/atom)
n =--=
M 64 gm/mole
= 8.4 X 1022 electrons/cm 3.
Finally, Va is, from Eq. 31-5,
j 480 amp/cm 2
Va= - = 22
ne (8.4 X 10 electrons/cm 3)(1.6 X 10-19 coul/electron)
= 3.6 X 10- 2 cm/sec.
It takes 28 sec for the electrons in this wire to drift 1.0 cm. Would you have guessed
that Va was so low? The drift speed of electrons must not be confused with the speed
at which changes in the electric field configuration travel along wires, a speed which
approaches that of light. When pressure is applied to one end of a long water-filled
tube, a pressure wave travels rapidly along the tube. The speed at which water moves
through the tube is much lower, however. <11111
774 CURRENT AND RESISTANCE Chap. 31
31-2 Resistance, Resistivity, and Conductivity
If the same potential difference is applied between the ends of a rod of
copper and of a rod of wood, very different currents result. The characteristic
of the conductor that enters here is its resistance. We define the resistance of
a conductor (often called a resistor; symbol ...1\/M-) between two points by
applying a potential difference V between those points, measuring the
current i, and dividing:
R = V /i. (31-6)
The resistivity of copper is 1.7 X 10-s ohm-m; that of fused quartz is about
10 16 ohm-m. Few physical properties are measurable over such a range of
values; Table 31-1 lists some values for common metals.
* These are materials whose properties (electrical in this case) do not vary with direction
in the material.
Sec. 31-2 RESISTANCE, RESISTIVITY, AND CONDUCTIVITY 775
Table 31-1
Temperature
Resistivity
Coefficient of Density, Melting
(at 20°C),
Resistivity,* gm/cm 3 Point, °C
ohm-m
a, per C 0
v
E=- and J
l A
E V/l
p = - = -·
j i/A
But V /i is the resistance R which leads to
l
R = p-· (31-9)
A
776 CURRENT AND RESISTANCE Chap. 31
V, i, and R are macroscopic quantities, applying to a particular body or extended
region. The corresponding microscopic quantities are E, j, and p; they have values at
every point in a body. The macroscopic quantities are related to each other by Eq.
31-6 (V = iR) and the microscopic quantities by Eq. 31-7, which can be written in
vector form as E = jp. •
The macroscopic quantities can be found by integrating over the microscopic
quantities, using relations already given, namely
(31-4)
The integral in Eq. 31-4 is a surface integral, carried out over any cross section of the
conductor. The integral in Eq. 29-5 is a line integral carried out along an arbitrary
line drawn along the conductor, connecting any two equipotential surfaces, identified
by a and b. For a long wire connected to a battery equipotential surface a might be
chosen as a cross section of the wire near the positive battery terminal and b might
be a cross section near the negative terminal.
The resistance of a conductor between a and b can be expressed in microscopic
terms by dividing the two equations, or
R = V_ab = - ibE·dl
'/,
If the conductor is a long cylinder of cross section A and length l and if points a and b
are its ends, the foregoing equation for R (see Eq. 31-7) reduces to
Figure 31-2 shows (solid curve) how the resistivity of copper varies with tempera-
ture. Sometimes, for practical use, such data are expressed in equation form. If we
are interested in only a limited range of temperatures extending, say, from Oto 500°0,
we can fit a straight line to the curve of Fig. 31-2, making it pass through two arbi-
trarily selected points; see the dashed line. We choose the point labeled To, p 0 in
the figure as a reference point, To being 0°0 in this case and p0 being 1.56 X 10-s
ohm-m. The resistivity p at any temperature T can be found from the empirical
equation of the dashed straight line in Fig. 31-2, which is
p = po[l + a(T - To)]. (31-10)
This relation shows correctly that p - po as T - T 0•
If we solve Eq. 31-10 for a, we obtain
_ 1 p - Po
a=- .
Po T - To
Comparison with Eq. 31-8 shows that a is a mean temperature coefficient of resistivity
for a selected pair of temperatures rather than the temperature coefficient of re-
sistivity at a particular temperature, which is the definition of a. For most practical
purposes Eq. 31-10 gives results that are within the acceptable range of accuracy.
T, °C
-200 0 200 400 600 800
8
E
I
E
.c
0
00 4
I
0
.....
c:i:
Fig. 31-2 The resistivity of copper as a function of temperature. The dashed line is an
ap,proximation chosen to fit the curve at the two circled points. The point marked To, p0
is chosen as a reference point.
778 CURRENT AND RESISTANCE Chap. 31
0.16
(/)
E
-§ 0.08 Fig. 31-3 The resistance of mer-
cury disappears below about 4 °K.
~
2 4 6
T, °K
The curve of Fig. 31-2 does not go to zero at the absolute zero of temperature, even
though it appears to do so, the residual resistivity at this temperature being 0.02 X
10-s ohm-m. For many substances the resistance does become zero at some low
temperature. Figure 31-3 shows the resistance of a specimen of mercury for temper-
atures below 6°K. In the space of about 0.05 K 0 the resistance drops abruptly to an
immeasurably low value. This phenomenon, called superconductivity,* was dis-
covered by Kamerlingh Onnes in the Netherlands in 1911. The resistance of ma-
terials in the superconducting state seems to be truly zero; currents, once established
in closed superconducting circuits, persist for weeks without diminution, even though
there is no battery in the circuit. If the temperature is raised slightly above the
superconducting point, such currents drop rapidly to zero.
0.6
a. .
Fig. 31-4 The current in a par- ~ 0.4 .•
ticular copper conductor as a func- .....-
tion of potential difference. This
conductor obeys Ohm's law.
0.2
o-
0 0.2 0.4 0.6
V, volts
80
a.
E
ro
c,:,
I Fig. 31-5 The current in a type
0
...... 2A3 vacuum tube as a function of
..... potential difference. This conduc-
20
tor does not obey Ohm's law.
4.0
10 20 30
V, volts
780 CURRENT AND RESISTANCE Chap. 31
Thermistors are often used to measure the rate of energy flow in microwave
beams by allowing the microwave beam to fall on the thermistor and heat it.
The rel&,tively small temperature rise so produced results in a relatively large
change in resistance, which serves as a measure of the microwave power.
Modern electronics, and therefore much of the character of our present tech-
nological civilization, depends in a fundamental way on the fact that many
conductors, such as vacuum tubes, crystal rectifiers, thermistors, and
transistors, do not obey Ohm's law.
We stress that the relationship V = iR is not a statement of Ohm's law. A
conductor obeys Ohm's law only if its V - i curve is linear, that is, if R is
independent of V and i. The relationship R = V / i remains as the general
definition of the resistance of a conductor whether or not the conductor
obeys Ohm's law.
The microscopic equivalent of the relationship V = iR is Eq. 31-7, or E
= jp. A conducting material is said to obey Ohm's law if a plot of E versus
j is linear, that is, if the resistivity pis independent of E and j. Ohm's law
is a specific property of certain materials and is not a general law of electro-
magnetism, for example, like Gauss's law.
A close analogy exists between the fl.ow of charge because of a potential difference
and the fl.ow of heat because of a temperature difference. Consider a thin electrically
conducting slab of thickness Lix and area A. Let a potential difference Li V be main-
tained between opposing faces. The current i is given by Eqs. 31-6 (i = V/R) and
31-9 (R = pl/A), or
. LiV LiVA
?,=-=--·
R pLix
In the limiting case of a slab of thickness dx this becomes
i = !A dV
p dx
dq dV
or - = - a A -- , (31-11)
dt dx
where a ( = 1/p) is the conductivity of the material. Since positive charge flows in
the direction of decreasing V, we introduce a minus sign into Eq. 31-11, that is,
dq/dt is positive when dV /dx is negative.
The analogous heat fl.ow equation (see Section 22-4) is
not simple, but when they have been carried out the agreement with the
experimental value of p has usually been good.
In a metal the valence electrons are not attached to individual atoms but
are free to move about within the lattice and are called conduction electrons.
In copper there is one such electron per atom, the other 28 remaining bound
to the copper nuclei to form ionic cores.
The speed distribution of conduction electrons can be described correctly
only in terms of quantum physics. For our purposes, however, it suffices to
consider only a suitably defined average speed ii; for copper ii = 1.6 X 108
cm/sec. In the absence of an electric field, the directions in which the elec-
trons move are completely random, like those of the molecules of a gas con-
fined to a container.
The electrons collide constantly with the ionic cores of the conductor,
that is, they interact ·.vith the lattice, often suffering sudden changes in speed
and direction. These collisions remind us of the collisions of gas molecules
confined to a container. As in the case of molecular collisions, we can des-
cribe electron-lattice collisions by a mean free path [, where [ is the average
distance that an electron travels between collisions.*
In an ideal metallic crystal at 0°K electron-lattice collisions would not occur,
according to the predictions of quantum physics, that is, l .- oo as T .- 0°K for
ideal crystals. Collisions take place in actual crystals because (a) the ionic cores at
any temperature Tare vibrating about their equilibrium positions in a random way,
(b) impurities, that is, foreign atoms, may be present, and (c) the crystal may contain
lattice imperfections, such as rows of missing atoms and displaced atoms. On this
view it is not surprising that the resistivity of a metal can be increased by (a) raising
its temperature, (b) adding small amounts of impurities, and (c) straining it severely,
as by drawing it through a die, to increase the number of lattice imperfections.
When an electric field is applied to a metal, the electrons modify their
random motion in such a way that they drift slowly, in the opposite direction
to that of the field, with an average drift speed vd. This drift speed is much
less than the effective average speed ii mentioned above (see Example 2).
Figure 31-7 suggests the relationship between these two speeds. The solid
lines suggest a possible random path followed by an electron in the absence of
an applied field; the electron proceeds from x toy, making six collisions on the
way. The dashed curves show how this same event might have occurred if
an electric field E had been applied. Note that the electron drifts steadily to
the right, ending at y' rather than at y. In preparing Fig. 31-7, it has been
assumed that the drift speed vd is 0.02ii; actually, it is more like 10-10ii, so
that the "drift" exhibited in the figure is greatly exaggerated.
The drift speed vd can be calculated in terms of the applied electric field E
and of ii and [. When a field is applied to an electron in the metal it will
experience a force eE which will impart to it an acceleration a given by
Newton's second law, eE
a=-·
m
* It can be shown that collisions between electrons occur only rarely and have little
effect on the resistivity.
782 CURRENT AND RESISTANCE Chap. 31
Fig. 31-7 The solid lines show an electron moving from x to y, making six collisions.
The dashed curves show what the electron path might have been in the presence of an
electric field E. Note the steady drift in the direction of - E.
Consider an electron that has just collided with an ion core. The col-
lision, in general, will momentarily destroy the tendency to drift and the
electron will have a truly random direction after the collision. At its next
collision the electron's velocity component in the - E direction will have
changed, on the average, by a(Z/v) where l/v is the mean time between
collisions. We call this the drift speed vd, or
Vd = a(J)v
= eE!.
mv
(31-13)
The electron's motion through the conductor is analogous to the constant rate of
fall of a stone in water. The gravitational force Fg on the stone is opposed by a
viscous resisting force that is proportional to the velocity, or
Fg =mg= "bv,
where bis a viscous coefficient (see Section 15-8). Thus the constant terminal speed
of the stone is
i-
-i
784 CURRENT AND RESISTANCE Chap. 31
Terminal a, connected to the positive battery terminal, is at a higher
potential than terminal b. If a charge dq moves through the box from a to b,
this charge will decrease its electric potential energy by dq Vab (see Section
29-6). The conservation-of-energy principle tells us that this energy is
transferred in the box from electric potential energy to some other form.
What that other form will be depends on what is in the box. In a time dt the
energy d U transferred inside the box is then
dU = dq Vab =i dt Vab·
dU
P = - = iVab· (31-15)
dt
If the device in the box is a motor, the energy appears largely as mechanical
work done by the motor; if the device is a storage battery that is being
charged, the energy appears largely as stored chemical energy in this second
battery.
If the device is a resistor, we assert that the energy appears as heat in the
resistor. To see this, consider a stone of mass m that falls through a height h.
It decreases its gravitational potential energy by mgh. If the stone falls in a
vacuum or-for practical purposes-in air, this energy is transformed into
kinetic energy of the stone. If the stone falls in water, however, its speed
eventually becomes constant, which means that the kinetic energy no longer
increases. The potential energy that is then steadily being made available
as the stone falls at constant speed appears as thermal energy in the stone
and the surrounding water. It is the viscous, friction-like drag of the
water on the surface of the stone that stops the stone from accelerating, and
it is at this surface that thermal energy appears.
The course of the electrons through the resistor is much like that of the
stone through water. The electrons travel with a constant drift speed va and
thus do not gain kinetic energy. The electric potential energy that they lose
is transferred to the resistor as heat. On a microscopic scale this can be
understood in that collisions between the electrons and the lattice (see Fig.
21-5) increase the amplitude of the thermal vibrations of the lattice; on a
macroscopic scale this corresponds to a temperature increase. This effect,
which is thermodynamically irreversible, is called Joule heating.
For a resistor we can combine Eqs. 31-15 and 31-6 (R = V/i) and obtain
either
(31-16)
vz
or P=-· (31-17)
R
Sec. 31-5 ENERGY TRANSFERS IN AN ELECTRIC CIRCUIT 785
Note that Eq. 31-15 applies to electrical energy transfer of all kinds; Eqs.
31-16 and 31-17 apply only to the transfer of electrical energy to heat energy
in a resistor. Equations 31-16 and 31-17 are known as Joule's law. This
law is a particular way of writing the conservation-of-energy principle for the
special case in which electrical energy is transferred into heat energy.
The unit of power that follows from Eq. 31-15 is the volt-amp. It can be
written as
1 volt-amp =
:,.._ 1
1 v~t-a ~ p ( 1 joule ) ( 1 c\ul )
1 v~t X 1 c~l 1 a~p X 1 sec
= 1 joule/sec.
The first conversion factor in parenthesis comes from the definition of the
volt (Eq. 29-1); the second comes from the definition of the coulomb. The
joule/sec is such a common unit that it is given a special name of its own, the
watt; see Section 7-7. Power is not an exclusively electrical concept, of
course, and we can express in watts the power ( = F·v) expended by an agent
that exerts a force F while it moves with a velocity v.
.... Example 5. You are given a 20-ft length of heating wire made of the special
alloy Nichrome; it has a resistance of 24 ohms. Can you obtain more heat by winding
one coil or by cutting the wire in two and winding two separate coils? In each case
the coils are to be connected individually across a 110-volt line.
The power P for the single coil is given by Eq. 31-17 :
V2 (110 volts) 2
P = ~R = 2 h = 500 watts.
4 oms
There are two "half-coils," so that the total power obtained by cutting the wire in
half is 2000 watts, or four times that for the single coil. This would seem to suggest
that we could buy a 500-watt heating coil, cut it in half, and rewind it to obtain
2000 watts. Why is this not a practical idea? <Ill!
786 CURRENT AND RESISTANCE Chap. 31
QUESTIONS
1. Name other physical quantities that, like current, are scalars having a sense repre-
sented by an arrow in a diagram.
2. What conclusions can you draw by applying Eq. 31-4 to a closed surface through
which a number of wires pass in random directions, carrying steady currents of different
sizes?
3. A potential difference V is applied to a circular cylinder of carbon by clamping it
between circular copper electrodes, as in Fig. 31-9. Discuss the difficulty of calculating
the resistance of the carbon cylinder, using the relation R = pL/ A.
Fig. 31-9
4. How would you measure the resistance of a pretzel-shaped conductor? Give specific
details to clarify the concept.
5. Discuss the difficulties of testing whether the filament of a light bulb obeys Ohm's
law.
6. Does the relation V = iR apply to nonohmic resistors?
7. The temperature coefficient of resistance of a thermistor is negative and varies
greatly with temperature. Account qualitatively for the shape of the curve of i versus 11
for the thermistor of Fig. 31-6.
8. A potential difference 11 is applied to a copper wire of diameter d and length l. What
is the effect on the electron drift speed of (a) doubling V, (b) doubling l, and (c) doubling d'?
9. If the drift speeds of the electrons in a conductor under ordinary circumstances are
so slow (see Example 2), why do the lights in a room turn on so quickly after the switch is
closed?
10. Can you think of a way to measure the drift speed for electrons by timing their
travel along a conductor?
11. Why are the dashed ·white lines in Fig. 31-7 curved slightly?
12. A current i enters the top of a copper sphere of radius Rand leaves at a diametrically
opposite point. Are all parts of the sphere equally effective in dissipating Joule heat?
13. What special characteristics must (a) heating wire and (b) fuse wire have? '
Chap. 31 PROBLEMS 787
14. Equation 31-16 (P = i 2R) seems to suggest that the rate of Joule heating in a
resistor is reduced if the resistance is made less; Eq. 31-17 (P = V 2 /R) seems to suggest
just the opposite. How do you reconcile this apparent paradox?
15. Is the filament resistance lower or higher in a 500-watt light bulb than in a 100-watt
bulb? Both bulbs are designed to operate on 110 volts.
16. Five wires of the same length and diameter are connected in turn between two points
maintained at constant potential difference. Will heat be developed at the fastest rate in
the wire of (a) the smallest or (b) the largest resistance?
PROBLEMS
1. A current of 5 amp exists in a 10-ohm resistance for 4 min. (a) How many coulombs
and (b) how many electrons pass through any cross section of the resistor in this time?
_,2:" A current is established in a gas discharge tube when a sufficiently high potential
difference is applied across the two electrodes in the tube. The gas ionizes; electrons move
toward the positive terminal and positive ions toward the negative terminal. What are the
magnitude and sense of the current in a hydrogen discharge tube in which 3.1 X 1018
electrons and 1.1 X 10 18 protons move past a cross-sectional area of the tubP each second?
3. A copper wire and an iron wire of the same length have the same potential difference
applied to them. (a) What must be the ratio of their radii if the current is to be the same?
(b) Can the current density be made the same by suitable choices of the radii?
4. A current i enters one corner of a square sheet of copper and leaves at the opposite
corner. Sketch arrows for various points within the square to represent the relative values
of j.,. Intuitive guesses rather than detailed mathematical analysis are called for.
/5. The belt of an electrostatic generator is 50 cm wide and travels at 30 meters/sec.
The belt carries charge into the sphere at a rate corresponding to 10-4 amp. Compute
the surface charge density on the belt.
6. A square aluminum rod is 1.0 meter long and 5.0 mm on edge. (a) What is the resist-
ance between its ends? (b) What must be the diameter of a circular LO-meter copper rod
if its resistance is to be the same?
7. A wire with a resistance of 6.0 ohms is drawn out so that its new length is three
times its original length. Find the resistance of the longer wire, assuming that the resis-
tivity and density of the material are not changed during the drawing process.
8. A copper wire and an iron wire of equal length l and diameter d are joined and a
potential difference V is applied between the ends of the composite wire. Calculate (a)
the electric field strength in each wire, (b) the current density in each wire, and (c) the
potential difference across each wire. Assume that l = 10 meters, d = 2.0 mm, and V =
100 volts.
9. A rod of a certain metal is 1.00 meter long and 0.550 cm in diameter. The resistance
between its ends (at 20°C) is 2.87 X 10-3 ohm. A round disk is formed of this same ma-
terial, 2.00 cm in diameter and 1.00 mm thick. (a) What is the resistance between the
opposing round faces? (b) What is the material?
10. Steel trolley-car rail has a cross-sectional area of 7.1 in. 2 What is the resistance of 10
miles of single track? The resistivity of the steel is 6.0 X 10-7 ohm-m.
11. (a) At what temperature would the resistance of a copper conductor be double its
resistance at 0°C? (b) Does this same temperature hold for all copper conductors, regard-
less of shape or size?
12. It is desired to make a long cylindrical conductor whose temperature coefficient of
resistance at 20°C will be close to zero. (a) If such a conductor is made by assembling
alternate disks of iron and carbon, what is the ratio of the thickness of a carbon disk to
that of an iron disk? Assume that the temperature remains essentially the same in each
disk. (b) What is the ratio of the rate of Joule heating in a carbon disk to that in an iron
disk?
788 CURRENT AND RESISTANCE Chap. 31
13. When a metal rod is heated, not only its resistance but also its length and its cross-
sectional area change. The relation R = pl/ A suggests that all three factors should be
taken into account in measuring p at various temperatures. If the temperature changes by
1.0 C 0 , what per cent changes in R, l, and A occur for a copper conductor. What con-
clusion do you draw? The coefficient of linear expansion is 1.7 X 10-5/C 0 •
14. The copper windings of a motor have a resistance of 50 ohms at 20°C, when the
motor is idle. After running for several hours the resistance rises to 58 ohms, What is
the te;nperature of the windings?
~ (a) Using data from Fig. 31-5, plot the resistance of the vacuum tube as a function of
applied potential difference. (b) Repeat for the thermistor of Fig. 31-6.
16. A small but measurable current of 10-10 amp exists in a copper wire whose diameter
is 0.10 in. Calculate the electron drift speed.
17. Heat is developed in a resistor at a rate of 100 watts when the current is 3.0 amp.
What is the resistance in ohms?
18. A potential difference of 1.0 volt is applied to a 100-ft length of #18 copper wire
(diameter = 0.040 in.). Calculate (a) the current, (b) the current density, (c) the electric
field strength, and (d) the rate of Joule heating.
19. The National Board of Fire Underwriters has fixed safe current-carrying capacities
for various sizes and types of wire. For #10 rubber-coated copper wire (wire diameter =
0.10 in.) the maximum safe current is 25 amp. At this current, find (a) the current density,
(b) the electric field strength, (c) the potential difference for 1000 ft of wire, and (d) the rate
of Joule heating for 1000 ft of wire.
20. A 500-watt immersion heater is placed in a pot containing 2.0 liters of water at 20°C.
(a) How long will it take to bring the water to boiling temperature, assuming that 80% of
the available energy is absorbed by the water? (b) How much longer will it take to boil
half the water away?
21. A nichrome heater dissipates 500 watts when the applied potential difference is 110
volts and the wire temperature is 800°C. How much power would it dissipate if the wire
temperature were held to 200°C by immersion in a bath of cooling oil? The applied poten-
tial difference remains the same; a for nichrome is about 4 X 10-4;c 0 •
22. A beam of 16-Mev deuterons from a cyclotron falls on a copper block. The beam is
equivalent to a current of 15 X 10-6 amp. (a) At what rate do deuterons strike the block?
(b) At what rate is heat produced in the block?
23. A "500-watt" heating unit is designed to operate from a 115-volt line. (a) By what
percentage will its heat output drop if the line voltage drops to 110 volts? Assume no
change in resistance. (b) Taking the variation of resistance with temperature into account,
would the actual heat output drop be larger or smaller than that calculated in (a)?
24. Show that p, the power per unit volume transformed into Joule heat in a resistor,
can be written as
p = j2p or
Electromotive Force
and Circuits
CHAPTER 32
o]: B R
li
-+------ i
Fig. 32-1 (a) A simple electric
(a) circuit and (b) its gravitational
analog.
''"'i@ ,':.,
a
"I i
h
'\\'' ~11, l ,
'' ,,;
(b)
end and leaves at its high-potential end. The seat must do an amount of
work dW on the (positive) charge carriers to force them to go to the point of
higher potential. The emf 8 of the seat is defined from
8 = dW/dq. (32-1)
The unit of emf is the joule/coul (see Eq. 29-1) which is the volt. We
might be inclined to say that a battery has an emf of 1 volt if it maintains a
difference of potential of 1 volt between its terminals. This is true only
under certain conditions, which we describe in Section 32-4.
If a seat of emf does work on a charge carrier, energy must be transferred
within the seat. In a battery, for example, chemical energy is transferred
into electrical energy. Thus we can describe a seat of emf as a device in
which chemical, mechanical, or some other form of energy is changed (re-
versibly) into electrical energy. The chemical energy provided by the
battery in Fig. 32-la is stored in the electric and the magnetic * fields that
surround the circuit. This stored energy does not increase because it is being
drained away, by transfer to Joule heat in the resistor, at the same rate at
which it is supplied. The electric and magnetic fields play an intermediary
role in the energy transfer process, acting as a storage reservoir.
Figure 32-lb shows a gravitational analog of Fig. 32-la. In the top figure
the seat of emf B does work on the charge carriers. This energy, stored
temporarily as electromagnetic field energy, appears eventually as Joule heat
in resistor R. In the lower figure the man, in lifting the bowling balls from
* A current in a wire is surrounded by a magnetic field, and this field, like the electric
field, can also be viewed as a site of stored energy (see Section 36-4).
Sec. 32-1 ELECTROMOTIVE FORCE 791
the floor to the shelf, does work on them. This energy is stored temporarily
as gravitational field energy. The balls roll slowly and uniformly along the
shelf, dropping from the right end into a cylinder full of viscous oil. They
sink to the bottom at constant speed, are removed by a trapdoor mechanism
not shown, and roll back along the floor to the left. The energy put into the
system by the man appears eventually as heat in the viscous fluid. The
energy supplied by the man comes from his own internal (chemical) energy.
The circulation of charges in Fig. 32-la will stop eventually if battery Bis not
charged; the circulation of bowling balls in Fig. 32-lb will stop eventually if
the man does not replenish his store of internal energy by eating.
Figure 32-2 shows a circuit containing two (ideal) batteries, A and B, a
resistor R, and an (ideal) electric motor employed in lifting a weight. The
batteries are connected so that they tend to send charges around the circuit
in opposite directions; the actual direction of the current is determined by B,
which supplies the larger potential difference. The energy transfers in this
~oott~ ·
Stored
electric
and
magnetic
field
energy
792 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
The chemical energy in B is being steadily depleted, the energy appearing in
the three forms shown on the right. Battery A is being "charged" while
battery B is being discharged. Again, the electric and magnetic fields that
surround the circuit act as an intermediary.
It is part of the definition of an emf that the energy transfer process be reversible,
at least in principle. The student will recall that a reversible process is one that
passes through equilibrium states; its course can be reversed by making an infinites-
imal change in the environment of the system; see Section 25-2. A battery, for
example, can either be on charge or discharge; a generator can be driven mechanically,
producing electric energy, or it can be operated backward as a motor. The (rever-
sible) energy transfers here are
electrical ;_:::: chemical
and electrical ;_:::: mechanical.
Joule heating is an electric energy transfer that is not reversible. We can easily heat
a conductor by supplying electric energy to it, but it is not possible to set up a
current in a closed copper loop by heating the loop uniformly. Because of this lack
of reversibility, we do not associate an emf with Joule heating.
or +e - ir - iR = 0.
The student should compare these equations with Fig. 32-3 b, which shows
the changes in potential graphically. In writing these equations, note that
we traversed rand R in the direction of the current and e in the direction of the
794 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
ii le :] it R
(a)
o------+- (:;
b r a R b
•I I ,N\ti
I I
• NNV'rI
I
•
I I
I I I
I I I I
I I I
.....
"' (:;
I I
0
>
I I
I
ro ir I
:p Va
c::
.....
(I)
0
c.. (,
f,- ir(=Vab)
vb
Fig. 32-3 A single-loop circuit. The rectangular block is a seat of emf with internal re-
sistance r. (b) The same circuit is drawn for convenience as a straight line. Directly
below are shown the changes in potential that one encounters in traversing the circuit
clockwise from point b.
emf. The same equation follows if we start at any other point in the circuit
or if we traverse the circuit in a counterclockwise direction. Solving for i
gives e
i =--- (32-3)
R +r
~ Example I. Resistors in series. Resistors in series are connected so that there
is only one conducting path through them, as in Fig. 32-4. What is the equivalent
resistance R of this series combination? The equivalent resistance is the single re-
sistance R which, substituted for the series combination between the terminals ab,
will leave the current i unchanged.
Applying the loop theorem (going clockwise from a) yields
-iR1 - iR2 - iR3 +8=0
or 'I,=
. e •
R1 + R2 + R3
Sec. 32-4 POTENTIAL DIFFERENCES 795
i~
a R1
t
Fig. 32-4 Example 1. Three resistors are connected
in series between terminals a and b. lr.
b R3
---i
which tells us that Vab, the potential difference between points a and b, has
the magnitude iR and that point a is more positive than point b. Combining
this last equation with Eq. 32-3 yields
R
Vab = e · (32-5)
R +r
To sum up: To find the potential difference between any two points in a cir-
cuit start at one point and traverse the circuit to the other, following any
path, and add up algebraically the potential changes encountered. This
algebraic sum will be the potential difference. This procedure is similar to
that for finding the current in a closed loop, except that here the potential
differences are added up over part of a loop and not over the whole loop.
The potential difference between any two points can have only one value;
thus we must obtain the same answer for all paths that connect these points.
796 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
If we consider two points on the side of a hill, the measured difference in
gravitational potential (that is, in altitude) between them is the same no
matter what path is followed in going from one to the other. In Fig. 32-3a
let us calculate Vab, using a path passing through the seat of emf. We have
Vb+ 8 - ir = Va
or (see also Fig. 32-3b)
Vab = Va - Vb= +0 - ir.
Again, combining with Eq. 32-3 leads to Eq. 32-5.
The terminal potential difference of the battery Vab, as Eq. 32-5 shows, is
less than 8 unless the battery has no internal resistance (r = 0) or if it is on
open circuit (R = oo); then Vab is equal to 8. Thus the emf of a device is
equal to its terminal potential difference when on open circuit .
.,_. Example 2. In Fig. 32-5a let 8 1 and 8 2 be 2.0 volts and 4.0 volts, respectively;
let the resistances r 1, r 2, and R be 1.0 ohm, 2.0 ohms, and 5.0 ohms, respectively.
What is the current?
(a)
---H b :,__
-1
~
0
> -2
m
~ -3 =
iR Vbc
~ -------,,.--""'-- - - - _ ___±
-4
-5
(b)
Fig. 32-5 (a) A single-loop circuit. (b) The same circuit is shown schematically as a
straight line, the potential differences encountered in traversing the circuit clockwise from
point a being displayed directly below. In the lower figure the potential of point a was
assumed to be zero for convenience.
Sec. 32-4 POTENTIAL DIFFERENCES 797
Emfs 81 and 82 oppose each other, but because 82 is larger it controls the direction
of the current. Thus i will be counterclockwise. The loop theorem, going clockwise
from a, yields
-82 + ir2 + iR + ir1 + 81 = 0.
The student should check that the same result is obtained by going around counter-
clockwise. He ilhould also compare this equation carefully with Fig. 32-5b, which
shows the potential changes graphically.
Solving for i yields
. 82 - 81 , 4.0 volts - 2.0 volts
i= =
R + r 1 + r 2 5.0 ohms+ 1.0 ohm+ 2.0 ohms
= 0.25 amp.
It is not necessary to know in advance what the actual direction of the current is.
To show this, let us assume that the current in Fig. 32-5a is clockwise, an assumption
that we know is incorrect. The loop theorem then yields (going clockwise from a)
- 82 - ir2 - iR - ir1 + 81 = 0
. Si - 82
or 1, = .
+ +
R r1 r2
Substituting numerical values (see above) yields -0.25 amp for the current. The
minus sign tells us that the current is in the opposite direction from the one we have
assumed.
In more complex circuit problems involving many loops and branches it is often
impossible to know in advance the correct directions for the currents in all parts of
the circuit. We can assume directions for the currents at random. Those currents
for which positive numerical values are obtained will have the correct directions;
those for which negative values are obtained will be exactly opposite to the assumed
directions. In all cases the numerical values will be correct.
Example 3. What is the potential difference (a) between points b and a in Fig.
32-5a? (b) Between points a and c?
(a) For points a and b we start at band traverse the circuit to a, obtaining
Vab ( = Va - Vb) = -ir2 + 82 = -(0.25 amp)(2.0 ohms)+ 4.0 volts
= +3.5 volts.
Thus a is more positive than band the potential difference (3.5 volts) is less than the
emf (4.0 volts); see Fig. 32-5b.
(b) For points c and a, we start at c and traverse the circuit to a, obtaining
Vac ( = Va - Ve) = +8 1 + ir1 = +2.0 volts + (0.25 amp)(l.O ohm)
= +2.25 volts.
This tells us that a is at a higher potential than c. The terminal potential difference
of Si (2.25 volts) is larger than the emf (2.0 volts); see Fig. 32-5b. Charge is being
forced through 8 1 in a direction opposite to the one in which it would send charge if
it were acting by itself; if 8 1 is a storage battery, it is being charged at the expense of
82.
Let us test the first result by proceeding from b to a along a different path, namely,
through R, r1, and 81. We have
Vab = iR + ir1 + 81 = (0.25 amp)(5.0 ohms)
+ (0.25 amp)(l.O ohm) + 2.0 volts = +3.5 volts,
which is the same as the earlier result. ....
798 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
32-5 Multiloop Circuits
Figure 32-6 shows a circuit containing two loops. For simplicity, we have
neglected the internal resistances of the batteries. There are two junctions,
b and d, and three branches connecting these junctions. The branches are
the left branch bad, the right branch bed, and the central branch bd. If the
emfs and the resistances are given, what are the currents in the various
branches?
We label the currents in the branches as i 1 , i 2 , and i 3 , as shown. Current
i 1 has the same value for any cross section of the left branch from b to d.
Similarly, i 2 has the same value everywhere in the right branch and i 3 in the
central branch. The directions of the currents have been chosen arbitrarily.
The careful reader will note that i 3 must point in a direction opposite to the
one we have shown. We have deliberately drawn it in wrong to show how
the formal mathematical procedures will always indicate this to us.
The three currents i 1 , i 2 , and i 3 carry charge either toward junction d or
away from it. Charge does not accumulate at junction d, nor does it drain
away from this junction because the circuit is in a steady-state condition.
Thus charge must be removed from the junction by the currents at the same
rate that it is brought into it. If we arbitrarily call a current approaching
the junction positive and the one leaving the junction negative, then
01
-----.0 ~
82
a b c
d
Sec. 32-5 MULTILOOP CIRCUITS 799
In multiloop circuits there is more than one loop, and the current in general
will not be the same in all parts of any given loop.
If we traverse the left loop of Fig. 32-6 in a counterclockwise direction,
the loop theorem gives
(32-6)
The right loop gives
(32-7)
These two equations, together with the relation derived earlier with the
junction theorem, are the three simultaneous equations needed to solve for
the unknowns i 1 , i 2 , and i 3 . Doing so yields
and (32-8c)
The student should supply the missing steps. Equation 32-8c shows that
no matter what numerical values are given to the emfs and to the resistances
the current i 3 will always have a negative value. This means that it will
always point up in Fig. 32-6 rather than down, as we deliberately assumed.
The currents i 1 and i 2 may be in either direction, depending on the particular
numerical values given.
The student should verify that Eqs. 32-8 reduce to sensible conclusions
in special cases. For R 3 = oo, for example, we find
. 81 - 82
i1 = i2 = and
R1 + R2
What do these equations reduce to for R 2 = oo?
The loop theorem can be applied to a large loop consisting of the entire
circuit abcda of Fig. 32-6. This fact might suggest that there are more equa-
tions than we need, for there are only three unknowns and we already have
three equations written in terms of them. However, the loop theorem yields
for this loop
which is nothing more than the sum of Eqs. 32-6 and 32-7. Thus this large
loop does not yield another independent equation. It will never be found in
solving multiloop circuits that there are more independent equations than
variables.
800 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
..... Example 4. Resistors in parallel. Figure 32-7 shows three resistors connected
across the same seat of emf. Resistances across which the identical potential dif-
ference is applied are said to be in parallel. What is the equivalent resistance R of
this parallel combination? The equivalent resistance is that single resistance which,
substituted for the parallel combination between terminals ab, would leave the
current i unchanged.
~ b
i
'!,
. = i1. + i2. + 1,3. = v(1
-
R1
+ -R21 + -R31) .
If the equivalent resistance is used instead of the parallel combination, we have
. v
i=-·
R
Combining these two equations gives
1 1 1 1
(32-9)
R= R1 + R2 + R3.
This formula can easily be extended to more than three resistances. Note that the
equivalent resistance of a parallel combination is less than any of the resistances that
make it up. "illl
the current i in the branch abed becomes zero. To balance the potentiometer in this way,
the sliding contact must be adjusted manually until the sensitive meter G reads zero.
* The voltmeter terminal marked + must be connected to the point of higher potential.
Otherwise the meter will deflect in a direction opposite to that intended.
802 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
~
ex
b
iol
The emf can be obtained from Eq. 32-10 if the current io is known becausef., can
be read from the calibrated markings along resistor R. However, it is standard
practice to replace 8., by a known standard emf Bs, and once again to adiust the
sliding contact to the zero-current condition. This yields-, assuming the current i 0
remains unchanged,
iof.R = e•.
Combining the last two equations yields
f,,
8., = Ba J.' (32-11)
which allows us to compare emfs with precision. Note that the internal resistance r
of the emf plays no role. In practice, potentiometers are conveniently packaged
units, containing a standard cell which, after calibration at the National Bureau of
Standards or elsewhere, serves as a convenient known standard seat of emf e•.
Switching arrangements for replacing the unknown emf by the standard and arrange-
ments for ascertaining that the current io remains constant are also incorporated
and the markings on resistor R, which is usually an array of several adjustable
resistors, are arranged to read 8., directly.
32-8 RC Circuits
The preceding sections dealt with circuits in which the circuit elements
were resistors and in which the currents did not vary with time. Here we
introduce the capacitor as a circuit element, which will lead us to the con-
cept of time-varying currents. In Fig. 32-10 let switch S be thrown to posi-
tion a. What current is set up in the single-loop circuit so formed? Let us
apply conservation of energy principles.
In time dt a charge dq ( = i dt) moves through any cross section of the
circuit. The work done by the seat of emf ( = e dq; see Eq. 32-1) must
equal the energy that appears as Joule heat in the resistor during time
e dq = i 2 R dt + d(;;)
or e dq = i2R dt + !!_ dq.
c
Dividing by dt yields
dq ·2 q dq
e-=iR+--·
dt c dt
But dq/dt is simply i, so that this equation becomes
e= iR + i. (32-12)
c
This equation also follows from the loop theorem, as it must, since the
loop theorem was derived from the conservation of energy principle. Start-
ing from point x and traversing the circuit clockwise, we experience an in-
crease in potential in going through the seat of emf and decreases in potential
in traversing the resistor and the capacitor, or
e- iR - !!._ = O
c '
which is identical with Eq. 32-12.
We cannot immediately solve Eq. 32-12 because it contains two variables,
q and i, which, however, are related by
. dq
i = -· (32-13)
dt
Substituting for i into Eq. 32-12 gives
dq q
S=R-+-· (32-14)
dt c
Our task now is to find the function q(t) that satisfies this differential equa-
tion. Although this particular equation is not difficult to solve, we choose
to avoid mathematical complexity by simply presenting the solution, which is
q = CS(l - e-t!Rc). (32-15)
We can easily test whether this function q(t) is really a solution of Eq. 32-14 by
substituting it into that equation and seeing whether an identity results. Differ-
entiating Eq. 32-15 with respect to time yields
dq (- ") _ f, -t/RC
dt - i - Re · (32-16)
Substituting q (Eq. 32-15) and dq/dt (Eq. 32-16) into Eq. 32-14 yields an identity,
as the student should verify. Thus Eq. 32-15 is a solution of Eq. 32-14.
804 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
<I)
.c 8
E
..Q
:::i
0
u
e
-~
E
C< 4
2 4 6 8 10
t, milliseconds
(a)
<I)
4
~
QJ
c.
E
-~
E
..... 2
4 6 8 10
t, milliseconds
(b)
Fig. 32-11 If, in Fig. 32-10, we assume that R = 2000 ohms, C = 1.0 µ,f, and e = 10
volts, then (a) shows the variation of q with t during the charging process and (b) the varia-
tions of i with t. The time constant is RC = 2.0 X 10-3 sec.
Figure 32-11 shows some plots of Eqs. 32-15 and 32-16 for a particular
case. Study of these plots and of the corresponding equations shows that
(a) at t = 0, q = 0 and i = S/R, and (b) as t ----,. ex:;, q ----,. Cs and i ----,. 0:
that is, the current is initially S/R and finally zero; the charge on the capaci-
tor plates is initially zero and finally CS.
Sec. 32-8 RC CIRCUITS 805
The quantity RC in Eqs. 32-15 and 32-16 has the dimensions of time
(since the exponent must be dimensionless) and is called the capacitative
time constant of the circuit. It is the time at which the charge on the capaci-
tor has increased to within a factor of (1 - e- 1 ) ( = 63%) of its equilibrium
value. To show this, we putt = RC in Eq. 32-15 to obtain
q = GS(l - e- 1) = 0.63GS.
Since GS is the equilibrium charge on the capacitor, corresponding tot ~ oo,
the foregoing statement follows .
.,_ Example 6. After how many time constants will the energy stored in the capaci-
tor in Fig. 32-10 reach one-half its equilibrium value?
The energy is given by Eq. 30-25, or
1 2
U = 2Cq'
the equilibrium energy U«) being (1/2C)(CS) 2 • From Eq. 32-15, we can write the
energy as
U = __!__ (CS)2(1 - e -t/Rc)2
2C
or U = U«i(l - e-t/Rc)2.
Putting U = !-U«) yields
! = (1 _ e-t/RC)2
and solving this relation for t yields finally
t = 1.22 RC = 1.22 time constants.
Figure 32-11 shows that if a resistance is included in the circuit the rate of increase
of the charge of a capacitor toward its final equilibrium value is delayed in a way
measured by the time constant RC. With no resistor present (RC = O), the charge
would rise immediately to its equilibrium value. Although we have shown that this
time delay follows from an application of the loop theorem to RC circuits, it is im-
portant that the student develop a physical understanding of the causes of the delay.
When switch S in Fig. 32-10 is closed on a, the resistor experiences instantaneously
an applied potential difference of S, and an initial current of S/R is set up. Initially,
the capacitor experiences no potential difference because its initial charge is zero,
the potential difference always being given by q/C. The flow of charge through the
resistor starts to charge the capacitor, which has several effects. First, the existence
of a capacitor charge means that there must now be a potential difference ( = q/C)
across the capacitor; this, in turn, means that the potential difference across the
resistor must decrease by this amount, since the sum of the two potential differences
must always equal S. This decrease in the potential difference across R means that
the charging current is reduced. Thus the charge of the capacitor builds up and the
charging current decreases until the capacitor is fully charged. At this point the
full emf Sis applied to the capacitor, there being no potential drop (i = 0) across the
resistor. This is precisely the reverse of the initial situation. The student should
review the derivations of Eqs. 32-15 and 32-16 and should study Fig. 32-11 with
the qualitative arguments of this paragraph in mind.
Assume now that the switch S in Fig. 32-10 has been in position a for a
time t such that t >> RC. The capacitor is then fully charged for all practical
purposes. The switch Sis then thrown to position b. How do the charge of
the capacitor and the current vary with time?
806 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
With the switch S closed on b, there is no emf in the circuit and Eq. 32-12
for the circuit, with 8 = 0, becomes simply
iR + '!_ = 0. (32-17)
c
Putting i = dq/dt allows us to write, as the differential equation of the circuit
(compare Eq. 32-14),
R dq +i = 0. (32-18a)
dt c
The solution is
(32-18b)
as the student may readily verify by substitution, q0 being the initial charge
on the capacitor. The capacitative time constant RC appears in this ex-
pression for capacitor discharge as well as in that for the charging process
(Eq. 32-15). We see that at a time such that t = RC the capacitor charge
is reduced to q0 e- 1 , which is 37% of the initial charge q0 .
The current during discharge follows from differentiating Eq. 32-18b, or
i = dq = - qo e-t/RC (32-19)
dt RC
The negative sign shows that the current is in the direction opposite to that
shown in Fig. 32-10. This is as it should be, since the capacitor is dis-
charging rather than charging. Since q0 = CB, we can write Eq. 32-19 as
'l = _ ! e-t/RC
R '
in which 8/R appears as the initial current, corresponding to t = 0. This
is reasonable because the initial potential difference for the fully charged
capacitor is 8.
The behavior of the RC circuit of Fig. 32-10 during charge and discharge can be
studied with a cathode-ray oscilloscope. This familiar laboratory device can display
on its fluorescent screen plots of the variation of potential with time. Figure 32-12
shows the circuit of Fig. 32-10 with connections made to display (a) the potential
difference V c across the capacitor and (b) the potential difference V R across the
resistor as functions of time. V c and V Rare given by
and VR = (R)i,
the former being proportional to the charge and the latter to the current.
Figure 32-13 shows oscillograph plots of V c and V R that result when, in effect,
switch Sin Fig. 32-10 is thrown regularly back and forth between positions a and b,
being left in each position for a time equal to several time constants. Intervals
during which the charge is building up are labeled ch and those during which it is
decaying are labeled dis.
RC CIRCUITS 807
VR (= iR)
0
1 GI,
1_
~I:
1---c-r> Vc(=H
T·
Fig. 32-12 The circuit of Fig. 32-10 with connections made to display the potential
variations across the resistor and the capacitor on a cathode ray oscilloscope.
a II II
(a)
rJ
• m ~
Ill II II
(b)
II II
I
(c)
Fig. 32-13 In Fig. 32-10 switch Sis thrown periodically, by electronic means, between
positions a and b. The variations with time of the potential differences across (a) the ca-
pacitor, and (b) the resistor are shown, as displayed on a cathode ray oscilloscope. (c) The
appearance of the screen when the oscilloscope is connected to display the sum of V R and
V c. (Courtesy E. K. Hege, Rensselaer Polytechnic Institute.)
808 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
The "charge" intervals in plot a (see Eq. 32-15) are represented by
V = (~) q = f;e-t!RC.
Note that the current, as indicated by plot b, is in opposite directions during the
charge and discharge intervals, in agreement with Eqs. 32-16 and 32-19.
In plot c in Fig. 32-13 the oscillograph has been connected to show the algebraic
sum of plots a and b. According to the loop theorem this sum should equal 8 during
the charge intervals and should be zero during the discharge intervals, when the
battery is no longer in the circuit, that is,
QUESTIONS
1. Does the direction of the emf provided by a battery depend on the direction of cur-
rent flow through the battery?
2. Figure 32-lb shows a gravitational analog of a single-loop electric circuit. Is the
source of "gravitational emf" in this figure reversible as far as energy exchanges are con-
cerned?
3. Discuss in detail the statement that the energy method and the loop theorem method
for solving circuits are perfectly equivalent.
4. It is possible to generate a 10,000-volt potential difference by rubbing a pocket comb
with wool. Why is this large voltage not dangerous when the much lower voltage pro-
vided by an ordinary electric outlet is very dangerous?
5. Devise a method for measuring the emf and the internal resistance of a battery.
6. A 25-watt, 110-volt bulb glows at normal brightness when connected across a bank
of batteries. A 500-watt, 110-volt bulb glows only dimly when connected across the same
bank. Explain.
7. Under what circumstances can the terminal potential difference of a battery exceed
its emf?
8. What is the difference between an emf and a potential difference?
9. In the flow of incompressible fluids, what are the analogies to (a) the loop theorem
and (b) the junction theorem?
10. Compare and contrast the formulas for the effective values of (a) capacitors and (b)
resistors, in series and in parallel.
11. Does the time required for the charge on a capacitor in an RC circuit to build up to a
given fraction of its equilibrium value depend on the value of the applied emf?
12. Devise a method whereby an RC circuit can be used to measure very high re-
sistances.
Chap. 32 PROBLEMS 809
PROBLEMS
12. In Fig. 32-16 imagine an ammeter inserted in the branch containing Ra. (a) What
will it read, assuming 0 = 5.0 volts, R 1 = 2.0 ohms, R 2 = 4.0 ohms, and Ra = 6.0 ohms?
(b) The ammeter and the source of emf are now physically interchanged. Show that the
810 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
ammeter reading remains unchanged. This reciprocity relationship holds for any circuit
that contains only one source of emf.
13. Two batteries of emf e and internal resistance r are connected in parallel across a
resistor Ras in Fig. 32-20b. (a) For what value of R is the power delivered to the resistor
a maximum? (b) What is the maximum power?
14. By using only two resistance coils-singly, in series, or in parallel-a student is able
to obtain resistances of 3, 4, 12, and 16 ohms. What are the separate resistances of the
coils?
_/15: The Wheatstone bridge. In Fig. 32-17 Rs is to be adjusted in value until points
a and b are brought to exactly the same potential. (One tests for this condition by mo-
mentarily connecting a sensitive meter between a and b; if these points are at the same
potential, the meter will not deflect.) Show that when this adjustment is made the fol-
lowing relation holds:
Unknown resistors (Rx) can be measured in terms of standards (Rs) using this device,
which is called a Wheatstone bridge.
Fig. 32-17
16. If points a and bin Fig. 32-17 are connected by a wire of resistance r, show that the
current in the wire is
e(R, - Rx)
i = (R + 2r)(R, + Rx) + 2R,Rx 1
where e is the emf of the battery. Assume that R 1 and R2 are equal (R1 = R2 = R) and
that Ro equals zero. Is this formula consistent with the result of Problem 15?
,,...ft- Two resistors, R1 and R2, may be connected either in series or parallel across a
(resistanceless) battery with emf e. We desire the Joule heating for the parallel combina-
tion to be five times that for the series combination. If R1 equals 100 ohms, what is R2?
18. Four 100-watt heating coils are to be connected in all possible series-parallel combi-
nations and plugged into a 100-volt line. What different rates of heat dissipation are
possible?
19. What is the equivalent resistance between the terminal points x and y of the circuits
shown in (a) Fig. 32-18a, (b) Fig. 32-18b, and (c) Fig. 32-18c? Assume that the re-
sistance of each resistor is 10 ohms.
Chap. 32 PROBLEMS 811
x
(a) (b) (c)
Fig. 32-18
_20. In Fig. 32-19 find the current in each resistor and the potential difference between
a and b. Put lh = 6.0 volts, 82 = 5.0 volts, 83 = 4.0 volts, R1 = 100 ohms, and R 2 = 50
ohms.
(a)
(b)
21. You are given two batteries of emf 8 and internal resistance r. They may be con-
nected either in series or in parallel and are used to establish a current in a resistor R, as in
Fig. 32-20. Derive expressions for the current in R for both methods of connection.
Which connection yields the larger current, considering various values of R / r?
22. (a) In Fig. 32-21 what power appears as Joule heat in R 1? In R 2? In R 3 ? (b)
What power is supplied by t; 1? By 82? (c) Discuss the energy balance in this circuit. As-
sume that 81 = 3.0 volts, 82 = 1.0 volt, R 1 = 5.0 ohms, R2 = 2.0 ohms, and R 3 = 4.0
ohms.
Fig. 32-21
812 ELECTROMOTIVE FORCE AND CIRCUITS Chap. 32
23. For manual control of the current in a circuit, a student uses a parallel combination
of variable resistors of the sliding contact type, as in Fig. 32-22, with R 1 = 20R 2 . (a) What
procedure is used to adjust the current to the desired value? (b) Why is the parallel com-
bination better than a single-variable resistor?
R R
(a) (b)
Fig. 32-23
.25. Resistance measurement. If meters are used to measure resistance, they may also
be connected as they are in Fig. 32-23b. Again the ratio of the meter readings gives only
an apparent resistance R'. Show that R' is related to R by
R = R' - RA,
in which RA is the ammeter resistance. Note that if RA « R, then R ""R'.
26. In Fig. 32-8 assume that 8 = 5.0 volts, r = 2.0 ohms, R1 = 5.0 ohms, and R2 =
4.0 ohms. If RA = 0.10 ohm, what per cent error is made in reading the current? Assume
that the voltmeter is not present.
27. In Fig. 32-8 assume that 8 = 5.0 volts, r = 20 ohms, R1 = 50 ohms, and R2 =
40 ohms. If Rv = 1000 ohms, what per cent error is made in reading the potential dif-
ferences across R 1? Ignore the presence of the ammeter.
Chap. 32 PROBLEMS 813
,2.s-:·(~) Find the three currents in Fig. 32-24. (b) Find Vab· Assume that R1 = 1.0
ohm, R 2 = 2.0 ohms, e1 = 2.0 volts, and 02 = es = 4.0 volts.
R1 a
R1
63
G1
Gz
R1
Fig. 32-24
.,,.,r_,,.,,,..""'
,29: How many time constants must elapse before a capacitor in an RC circuit is charged
to within 1.0 per cent of its equilibrium charge?
30. rn the circuit of Fig. 32-25 let i1, i2, and is be the currents through resistors R1, R2,
and Rs, respectively, and let Vi, V2, Vs, and V c be the corresponding potential differences
across the resistors and across the capacitor C. (a) Plot qualitatively as a function of time
after switch Sis closed the currents and voltages listed above. (b) After being closed for
a large number of time constants, the switch S is now opened. Plot qualitatively as a
function of time after the switch is opened the currents and voltages listed above.
Rs
Fig. 32-25
31. Show that the units of RC are indeed time units, that is, that 1 ohm X 1 farad = 1
sec.
32. Prove that when switch S in Fig. 32-10 is thrown from a to b all the energy stored
in the capacitor is transformed into Joule heat in the resistor. Assume that the capacitor
is fully.charged before the switch is thrown.
~ A 3.0 X 106-ohm resistor and a 1.0-µf capacitor are connected in a single-loop
circuit with a seat of emf with e = 4.0 volts. At 1.0 sec after the connection is made, what
are the rates at which (a) the charge of the capacitor is increasing, (b) energy is being stored
in the capacitor, (c) Joule heat is appearing in the resistor, and (d) energy is being delivered
by the seat of emf?
The Magnetic Field
CHAPTER 33
814
Sec. 33-1 THE MAGNETIC FIELD 815
feet of a current in a wire can be intensified by forming the wire into a coil of
many turns and by providing an iron core. Figure 33-2 shows how this is
done in a large electromagnet of a type commonly used for research involving
magnetism.
We define the space around a magnet or a current-carrying conductor as
the site of a magnetic field, just as we defined the space near a charged rod
as the site of an electric field. The basic magnetic field vector B, which we
define in the following section, is called the magnetic induction: * it can be
represented by lines of induction, just as the electric field was represented by
lines of force. As for the electric field (see Section 27-3), the magnetic field
vector is related to its lines of induction in this way:
cI>s = J
B·dS, (33-1)
in which the integral is taken over the surface (closed or open) for which cI>B
is defined.
* Magnetic field strength would be a more suitable name for B, but it has been usurped
for historical reasons by another vector connected with the magnetic field (see Section
37-7).
816 THE MAGNETIC FIELD Chap. 33
33-2 The Definition of B
This chapter is not concerned with the causes of the magnetic field; we
seek to determine (a) whether a magnetic field exists at a given point and
(b) the action of this field on charges moving through it. As for the electric
field, a particle of charge q0 serves as a test body. We assume that there
is no electric field present, which means that, neglecting gravity, no force
will act on the test body if it is placed at rest at the point in question.
Let us fire a positive test charge with arbitrary velocity v through a point
P. If a sideways deflecting force F acts on it, we assert that a magnetic field is
present at P and we define the magnetic induction B of this field in terms of
F and other measured quantities.
If we vary the direction of v through point P, keeping the magnitude of v
unchanged, we find, in general, that although F will always remain at right
angles to v its magnitude F will change. For a particular orientation of v
(and also for the opposite orientation -v) the force F becomes zero. We
define this direction as the direction of B, the specification of the sense of B
(that is, the way it points along this line) being left to the more complete
definition of B that we give below.
Having found the direction of B, we are now able to orient v so that the
test charge moves at right angles to B. We will find that the force Fis now
a maximum, and we define the magnitude of B from the measured magnitude
of this maximum force F .1., or
F ..L
B = --· (33-2)
qov
Let us regard this definition of B (in which we have specified its magnitude
and direction, but not its sense) as preliminary to the complete vector defini--
tion that we now give: If a positive test charge q0 is fired with velocity v through
a point P and if a (sideways) force F acts on the moving charge, a magnetic
induction B is present at point P, where B is the vector that satisfies the relation
F = q0 v x B, (33-3a)
v, q0 , and F being measured quantities. The magnitude of the magnetic
deflecting force F, according to the rules for vector products, is given by *
F = q0vB sin 8, (33-3b)
in which () is the angle between v and B.
Figure 33-3 shows the relations among the vectors. We see that F, being
at right angles to the plane formed by v and B, will always be at right angles
to v (and also to B) and thus will always be a sideways force. Equation 33-3a
is consistent with the observed facts that (a) the magnetic force vanishes as
v ~ 0, (b) the magnetic force vanishes if v is either parallel or antiparallel
to the direction of B (in these cases() = 0 or 180° and v xB = 0), and (c) if
v is at right angles to B (8 = 90°), the deflecting force has its maximum
value, given by Eq. 33-2, that is, F .1. = q0 vB.
* The student may wish to review Section 2-4, which deals with vector products.
Sec. 33-2 THE DEFINITION OF B 817
z
x
Fig. 33-3 Illustrating F = qov X B (Eq.
33-3a).
This definition of B is similar in spirit, although more complex, than the definition
of the electric field strength E, which we can cast into this form: If a positive test
charge q0 is placed at point P and if an (electric) force Facts on the stationary charge, an
electric field E is present at P, where E is the vector satisfying the relation
F = qoE,
q0 and F being measured quantities. In defining E, the only characteristic direction
to appear is that of the electric force FE which acts on the positive test body; the
direction of Eis taken to be that of FE. In defining B, two characteristic directions
appear, those of v and of the magnetic force FB; they prove always to be at right
angles.
In Fig. 33-4 a positive and a negative electron are created at point Pin a
bubble chamber. A magnetic field is perpendicular to the chamber, pointing
out of the plane of the figure (symbol O). * The relation F = q0 v x B
(Eq. 33-3a) shows that the deflecting forces acting on the two particles are
as indicated in the figure. These deflecting forces would make the tracks
deflect as shown.
The unit of B that follows from Eq. 33-3 is the (nt/coul)/(meter/sec).
This is given the special name weber/meter 2 , or tesla. Recalling that a
couljsec is an ampere,
1 nt 1 nt
1 weber/meter2 = - - - - - - - = - - -
coul (meter/sec) amp-m
An earlier unit for B, still in common use, is the gauss; the relationship is
1 weber/meter2 = 104 gauss.
The weber is used to measure cI>B, the flux of B; see Eq. 33-1.
The fact that the magnetic force is always at right angles to the direction
of motion Illeans that (for steady magnetic fields) the work done by this
* The syr.1bol © indicates a vector into the page, the X being thought of as the tail of
an arrow; the symbol O indicates a vector out of the page, the dot being thought of as the
tip of an arrow.
818 THE MAGNETIC FIELD Chap. 33
Fig. 33-4 A bubble chamber is a device for rendering visible, by means of small bubbles,
the tracks of charged particles that pass through the chamber. The figure is a photograph
taken with such a chamber immersed in a field of magnetic induction B and exposed to
radiations from a large cyclotron-like accelerator. The curved Vat point Pis formed by
a positive and a negative electron, which deflect in opposite directions in the magnetic
field. The spirals Sare the tracks of three low-energy electrons. (Courtesy E. 0. Lawrence
Radiation Laboratory, University of California.)
force on the particle is zero. For an element of the path of the particle of
length di, this work dW is FB·dl; dW is zero because FB and di are always
at right angles. Thus a static magnetic field cannot change the kinetic
energy of a moving charge; it can only deflect it sideways.
If a charged particle moves through a region in which both an electric field
and a magnetic field are present, the resultant force is found by combining
Eqs. 27-2 and 33-3a, or
(33-4)
This 1s sometimes called the Lorentz relation in tribute to H. A. Lorentz
Sec. 33-3 MAGNETIC FORCE ON A CURRENT 819
who did so much to develop and clarify the concepts of the electric and mag-
netic fields.
~ Example 1. A uniform field of magnetic induction B points horizontally from
south to north; its magnitude is 1.5 webers/meter 2 • If a 5.0-Mev proton moves
vertically downward through this field, what force will act on it?
The kinetic energy of the proton is
K = (5.0 X 106 ev)(l.6 X 10-19 joule/ev) = 8.0 X 10- 13 joule.
Its speed can be found from the relation K = imv 2 , or
F = ilB. (33-5)
The negative charges, which move to the right in the wire of Fig. 33-5,
are equivalent to positive charges moving to the left, that is, in the direction
of the current arrow. For such a positive charge the velocity v would point
to the left and the force on the wire, given by Eq. 33-3a (F = q0 v x B)
points up, out of the page. This same conclusion follows if we consider the
actual negative charge carriers for which v points to the right but q0 has a
negative sign. Thus by measuring the sideways magnetic force on a wire
carrying a current and placed in a magnetic field we cannot tell whether the
current carriers are negative charges moving in a given direction or positive
charges moving in the opposite direction.
Equation 33-5 holds only if the wire is at right angles to B. We can ex-
press the more general situation in vector form as
F = ii xB, (33-6a)
dF = i di xB. (33-6b)
• 0 0 0
• • 0 0
• • 0 0
dF sin(} dF /B
• 0
• 0
I
\ • 0
• •
I
• • • 0
. ·M . -
(}
\
\\
d(J
R
. 0 0
•
0
___ j __ \ ~ - - - -
>I O
I~ _j
• • • • • 0
• • • • • • •
F1 F3
• 0 0 0
• • • 0
• 0
• 0 0
,------------ -/B -,
Ix x x xilx x x I
x x x x
I F I
xxxxxxxxl
1 xxxx
lxxxxxxxxxxxxl
Ix x x x x x xix x x x x I
I I
Jxxxxxxxxxxxxl
L---------------~
Fig. 33-7 Apparatus used to measure B. The zero point of the balance is observed by
means of a light beam reflected from the mirror attached to the balanre beam.
822
Sec. 33-4 TORQUE ON A CURRENT LOOP 823
m of about 8.78 gm had to be added to the left balance pan to restore the original rest
point. The change in force when the current is reversed is 2F, and this must equal
the weight added to the left balance pan, or
mg= 2(9iaB) = 18iaB.
This gives
The net force on the loop is the resultant of the forces on the four sides of
the loop. On side 2 the vector I points in the direction of the current and has
the magnitude b. The angle between I and B for side 2 (see Fig. 33-Sb) is
F1 F1
x x x x x x x x x x x x
x x x x x x x x x x x x
1 n'
x x x x x x
~ x x
xxxx xxxxxxxxx n
:::Lxx
x x x x x x 8x xxx:::
xxxx
3
B
Fa B
a
(a) (b)
Fig. 33-8 A rectangular coil carrying a current i is placed in a uniform external magnetic
field.
824 THE MAGNETIC FIELD Chap. 33
90° - 8. Thus the magnitude of the force on this side is
F 2 = ibB sin (90° - 8) = 1:bB cos 8.
From the relation F = ii x B (Eq. 33-6a), we find the direction of F2 to be
out of the plane of Fig. 33-Sb. The student can show that the force F 4 on
side 4 has the same magnitude as F 2 but points in the opposite direction.
Thus F 2 and F 4 , taken together, have no effect on the motion of the loop.
The net force they provide is zero, and, since they have the same line of
action, the net torque due to these forces is also zero.
The common magnitude of F 1 and F 3 is iaB. These forces, too, are op-
positely directed so that they do not tend to move the coil bodily. As Fig.
33-Sb shows, however, they do not have the same line of action if the coil is
in the position shown; there is a net torque, which tends to rotate the coil
clockwise about the line xx'. The coil can be supported on a rigid axis that
lies along xx', with no loss of its freedom of motion. This torque can be
represented in Fig. 33-Sb by a vector pointing into the figure at point x' or
in Fig. 33-Sa by a vector pointing along the xx' axis from right to left.
The magnitude of the torque T' is found by calculating the torque caused
by F 1 about axis xx' and doubling it, for F 3 exerts the same torque about this
axis that F 1 does. Thus
r' = 2(iaB) (~) (sin 8) = iabB sin 0. ~
This torque acts on every turn of the coil. If there are N turns, the torque
on the entire coil is
r = Nr' = NiabB sin() = NiAB sin O, (33-7)
in which A, the area of the coil, is substituted for ab.
This equation can be shown to hold for all plane loops of area A, whether
they are rectangular or not. A torque on a current loop is the basic operating
principle of the electric motor and of most electric meters used for measuring
current or potential difference.
9ll- Example 3. A galvanometer. Figure 33-9 shows the rudiments of a galvanometer,
which is a device used to measure currents. The coil is 2.0 cm high and 1.0 cm wide;
it has 250 turns and is mounted so that it can rotate about a vertical axis in a uniform
radial magnetic field with B = 2000 gauss. A spring Sp provides a countertorque
that cancels out the magnetic torque, resulting in a steady angular deflection <J, cor-
responding to a given current i in the coil. If a current of 1.0 X 10-4 amp produces
an angular deflection of 30°, what is the torsional constant K of the spring (see Eq.
15-21)'?
Equating the magnetic torque to the torque caused by the spring (see Eq. 33-7)
yields
r = NiAB sin() = K<f,
or
NiAB sin8
K=----
<f,
(250)(1.0 X 10-4 amp)(2.0 X 10-4 meter2)(0.20 weber/meter 2)(sin 90°)
30°
= 3.3 X 10-s nt-m/deg.
Sec. 33-4 TORQUE ON A CURRENT LOOP 825
Scale
Note that the normal to the plane of the coil (that is, the pointer P) is always at
right angles to the (radial) magnetic field so that() = 90°. .....
T = p xE, (33-8)
where pis the electric dipole moment, and (b) it sets up a field of its own at
distant points, described qualitatively by the lines of force of Fig. 29-10
and quantitatively by Eq. 29-11. These two requirements are not in-
dependent; if one is fulfilled, the other follows automatically.
The magnitude of the torque described by Eq. 33-8 is
r = pE sin 0, (33-9)
where e is the angle between p and E. Let us compare this with Eq. 33-7,
the expression for the torque on a current loop:
U =
()
f r do =
()
f NiAB sin Odo = µB
i() sin Odo = -µB cos o,
.Jgoo .Jgoo goo
in which Eq. 33-7 is used to substitute for r. In vector symbolism this rela-
tion can be written as
U = -µ.·B, (33-12)
which is in perfect correspondence with Eq. 27-13, the expression for the
energy of an electric dipole in an external electric field,
u= -p·E.
.._ Example 4. A circular coil of N turns has an effective radius a and carries a
current i. How much work is required to turn it in an external magnetic field B from
a position in which 8 equals zero to one in which 8 equals 180°? Assume that N = 100,
a = 5.0 cm, i = 0.10 amp, and B = 1.5 webers/meter2 •
The work required is the difference in energy between the two positions, or, from
Eq. 33-12,
W = Uo= 1so 0 - Uo=o = (-µB cos 180°) - (-µB cos 0) = 2µB.
Sec. 33-5 THE HALL EFFECT 827
Butµ = NiA, so that
W = 2NiAB = 2N i('1ra 2)B
x x x x
x x
:,1 y x
(B :,1 y x
(B
x x x x
x x x x
(a) (b)
Fig. 33-10 A current i is set up in a copper strip placed in a field of magnetic induction
B, assuming (a) positive carriers and (b) negative carriers.
828 THE MAGNETIC FIELD Chap. 33
To analyze the Hall effect quantitatively, let us use the free-electron model of a
metal, the same model used in Section 31-4 to explain resistivity. The charge carriers
Mn be assumed to move along the conductor with a certain constant drift speed l'a·
The magnetic deflecting force that causes the moving charge carriers to drift toward
the right edge of the strip is given by qva xB (see Eq. 33-3a).
The charge carriers do not build up without limit on the right edge of the strip
because the displacement of charge gives rise to a transverse Hall electric .field Ey,
which acts, inside the conductor, to oppose the sideways drift of the carriers. This
Hall electric field is another manifestation of the Hall potential difference and is
related to it by
EH= Vxy/d.
Eventually an equilibrium is reached in which the sideways magnetic deflecting
force on the charge carriers is just canceled by the oppositely directed electric force
qEy caused by the Hall electric field, or
qEH + qva x B = 0,
which can be written
EH= -va xB. (33-13)
This equation shows explicitly that if EH and B are measured va can be determined
both in magnitude and direction; given the direction of va, the sign of the charge
carriers follows at once, as Fig. 33-10 shows.
The number of charge carriers per unit volume (n) can also be found from Hall effect
measurements. If we write Eq. 33-13 in terms of magnitudes, for the case in which
va and B are at right angles, we obtain EH = vaB. Combining this with Eq. 31-5
(vd = j/ne) leads to
j jB
EH= -B or n=--· (33-14)
ne eEH
The agreement between experiment and Eq. 33-14 is rather good for monovalent
metals, as Table 33-1 shows.
Table 33-1
Li 3.7 4.8
Na 2.5 2.6
K 1.5 1.3
Cs 0.80 0.85
Cu 11 8.4
Ag 7.4 6.0
Au 8.7 5.9
For nonmonovalent metals, for iron and similar magnetic materials, and for so-
called semiconductors such as germanium, the simple interpretation of the Hall effect
in terms of the free-electron model is not valid. A theoretical interpretation of the
Sec. 33-6 CIRCULATING CHARGES 829
Hall effect based on modern quantum physics gives a reasonable agreement with
experiment in all cm-es.
.... Example 5. A copper strip 2.0 cm wide and 1.0 mm thick is placed in a magnetic
field with B = 1.5 webers/meter 2 , as in Fig. 33-10. If a current of 200 amp is set
up in the strip, what Hall potential difference appears across the strip?
From Eq. 33-14,
jB
EH=-,
ne
Vxy . i i
but EH= - - and J = - = -,
d A dh
x x x x x x x x x
x x x x x x x x
x x x x x x
x x x x x x x x
x x x
~ x x
x r x x
x x x x x x
)""B
x x x x x x
x x x x x x
x x x x x x x x x x x x
830 THE MAGNETIC FIELD Chap. 33
path. From ~ewton's second law we have
mv 2 mv
qvB = - or r = -,
r qB
which gives the radius of the path. The three spirals in Fig. 33-4 show rela-
tively low-energy electrons in a bubble chamber. The paths are not circles
because the electrons lose energy by collisions in the chamber as they move.
The angular velocity w is given by v/r or, from Eq. 33-15,
v qB
w = - = -,
r m
Note that" does not depend on the speed of the particle. Fast particles move
in large circles (Eq. 33-1.5) and slow ones in small circles, but all require the
same time T (the period) to complete one revolution in the field.
The frequency " is a characteristic frequency for the charged particle in
the field and may be compared to the characteristic frequency of a swinging
pendulum in the earth's gravitational field or to the characteristic frequency
of an oscillating mass-spring sy::;tem. It is sometimes called the cyclotron
frequency of the particle in the field because particles circulate at this fre-
quency in the cyclotron.
~ Example 6. A 10-ev electron is circulating in a plane at right angles to a uniform
field of magnetic induction of 1.0 X 10- 4 weber/meter 2 ( = 1.0 gauss).
(a) What is its orbit radius?
~2:.
The velocity of an electron whose kinetic energy is K can be found from
v=
The student should verify that this yields 1.9 X 10 6 meters/sec for v. Then, from
Eq. 33-15,
r= ~ = (9.1 X 10- 31 kg)(l.9 X 106 meters/sec) = O.ll meter = 11 cm.
qB (1.6 X 10-19 coul)(l.O X 10-4 weber/meter 2)
(b) What is the cyclotron frequency? From Eq. 33-16,
" = qB = (1.6 X 10- 19 coul)(l.O X 10-4 weber/meter 2 ) = 2 .8 X 106 rev/sec.
21rm (21r)(9.1 X 10-3 1 kg)
(c) What is the period of revolution T?
point to the left at the position shown in the figure if the charge q were positive.
However, the charge is an electron, with q = - e, which means that v must point to
the right. Thus the charge circulates clockwise as viewed by an observer sighting in
the direction of B. _,..
Fig. 33-12 The University of Pittsburgh cyclotron. Note vacuum chamber V, magnet
frame F, magnetic pole faces P, magnet coils C, and the deuteron beam emerging into the
air of the laboratory. The rule is 6 ft long. (Courtesy A. J. Allen.)
direction is out of the plane of Fig. 33-13. The field is set up by a large
electromagnet, marked Fin Fig. 33-12. Finally, the space in which the ions
move is evacuated to a pressure of about 10- 5 mm-Hg. If this were not
done, the ions would continually collide with air molecules.
Suppose that a deuteron, emerging from the ion source, finds the dee that
it is facing to be negative; it will accelerate toward this dee and will enter it.
Once inside, it is screened from electrical forces by the metal walls of the
<lees. The magnetic field is not screened by the <lees so that the ion bends in
a circular path whose radius, which depends on the velocity, is given by
Eq. 33-15, or mv
r = --·
qB
After a time t 0 the ion emerges from the dee on the other side of the ion
source. Let us assume that the accelerating potential has now changed sign.
Thus the ion again faces a negative dee, is further accelerated, and again
describes a semicircle, of somewhat larger radius (see Eq. 33-15), in the dee.
Beam
Def le ctor
plate \
t: "Dee"
- ,.
r---
I
/ / i\
..., ~
"' ~
.8
,.. :: '"'''....
·c3
<J)
v ····· ·····
.....
0
•·
: L_
:
~ :; ""Dee"
= ········
Fig. 33-13 The elements of a cyclotron showing the ion source Sand the dees. The de-
flector plate, held at a suitable negative potential, deflects the particles out of the dee
system.
Sec. 33-7 THE CYCLOTRON 833
The time of passage through this dee, however, is still to. This follows because
the period of revolution T of an ion circulating in a magnetic field does not
depend on the speed of the ion; see Eq. 33-16. This process goes on until
the ion reaches the outer edge of one dee where it is pulled out of the system
by a negatively charged deflector plate.
The key to the operation of the cyclotron is that the characteristic fre-
quency v at which the ion circulates in the field must be equal to the fixed
frequency v 0 of the electric oscillator, or
V = VO,
This resonance condition says that if the energy of the circulating ion is to
increase energy must be fed to it at a frequency v 0 that is equal to the
natural frequency v at which the ion circulates in the field. In the same
way we feed energy to a swing by pushing it at a frequency equal to the
natural frequency of oscillation of the swing.
From Eq. 33-16 (v = qB/21rm), we can rewrite the resonance equation as
qB
- = VO, (33-17)
21rm
Once we have selected an ion to be accelerated, q/m is fixed; usually the
oscillator is designed to work at a single frequency vo. We then "tune" the
cyclotron by varying B until Eq. 33-17 is satisfied and an accelerated beam
appears.
The energy of the particles produced in the cyclotron depends on the radius
R of the dees. From Eq. 33-15 (r = mv/qB), the velocity of a particle cir-
culating at this radius is given by
qBR
v = --·
m
The kinetic energy is then
· q2B2R2
K = !mv 2 (33-18)
2m
~ Example 7. The University of Pittsburgh cyclotron has an oscillator frequency
of 12 X 106 cycles/sec and a dee radius of 21 in. (a) What value of magnetic induc-
tion B is needed to accelerate deuterons?
From Eq. 33-17, v 0 = qB/21rm, so that
B = 21rvom = (21r)(12 X 10 6/sec)(3.3 X 10- 27 kg) = 1. 6 webers/meter2.
q 1.6 X 10- 19 coul
Note that the deuteron has the same charge as the proton but (very closely) twice the
mass.
(b) What deuteron energy results?
From Eq. 33-18,
K = q2B 2R 2 = (1.6 X 10-19 coul)2(1.6 webers/meter2)2(21 X 0.0254 meter)2
2m (2)(3.3 X 10-27 kg)
-- (2 .8 X 10 -12 JOU
. 1e) ( 1. X 1 ev ) -
10 _ 19 joule - 17 Mev.
T
6
The cyclotron fails to operate at high energies because one of its assumptions, that
the frequency of rotation of an ion circulating in a magnetic field is independent of its
834 THE MAGNETIC FIELD Chap. 33
speed, is true only for speeds much less than that of light. As the particle speed
increases, we must use the relativistic mass min Eq. 33-16. The relativistic mass
increases with velocity (Eq. 8-20) so that at high enough speeds 11 decreases with
velocity. Thus the ions get out of step with the electric oscillator, and eventually
the energy of the circulating ion stops increasing.
Another difficulty associated with the acceleration of particles to high energies is
that the size of the magnet that would be required to guide such particles in a circular
path is very large. For a 30-Gev proton, for example, in a field of 15,000 gauss the
radius of curvature is 65 meters. A magnet of the cyclotron type of this size (about
430 ft in diameter) would be prohibitively expensive. Incidentally, a 30-Gev proton
has a speed equal to 0.99998 that of light.
Both the relativistic and the economic limitations have been removed by tech-
niques that can be understood in terms of Eq. 33-17 in which mis now taken to be
the relativistic mass, given by Eq. 8-20, or
mo
m= '
Vl - (v/c) 2
Table 33-2
THE CERN PROTON SYNCHROTRON
erator built at Geneva, Switzerland, by the European Council for Nuclear Research
(CERN), and embodying these principles.
Fig. 33-14 Electrons from the heated filament Fare accelerated by a potential difference
V and pass through a hole in the screen C. After passing through a region in which per-
pendicular electric and magnetic fields are present, they strike the fluorescent screen S.
836 THE MAGNETIC FIELD Chap. 33
Thus for a given electron speed v the condition for zero deflection can be
satisfied by adjusting E or B.
When Thomson's experiment is repeated today, usually as a student
experiment, the procedure* is: (a) note the position of the undeflected beam
spot, with E and B both equal to zero; (b) apply a fixed electric field E,
measuring on the fluorescent screen the deflection so caused; and (c) apply
a magnetic field and adjust its value until the beam deflection is restored to
zero.
In Section 27-5 we saw that the deflection y of an electron in a purely
electric field (step b), measured at the far edge of the deflecting plates, 1s
given by Eq. 27-9, or, with small changes in notation,
eEl2
y=--,
2mv 2
where v is the electron speed and l is the length of the deflecting plates; y is
not measurable directly, but it may be calculated from the measured dis-
placement of the spot on the screen if the geometry of the apparatus is
known. Thus y, E, and l are known; the ratio e/m and the velocity v are
unknown. We cannot calculate e/m until we have found the velocity, which
is the purpose of step c above.
If (step c) the electric force is set equal and opposite to the magnetic force,
the net force is zero and we can write (Eq. 33-19)
E
v = -·
B
Substituting this equation into the equation for y and solving for the ratio
e/m leads to
e 2yE
(33-20)
m = B 2 z2 '
in which all the quantities on the right can be measured. Thomson's value
for e/m was 1.7 X 10 11 couljkg, in excellent agreement with the modern
value of 1.75890 X 10 11 coul/kg.
QUESTIONS
1. Of the three vectors in the equation F = qv x B, which pairs are always at right
angles? Which may have any angle between them?
2. Why do we not simply define the magnetic induction B to point in the direction of the
magnetic force that acts on the moving charge?
3. Imagine that you are sitting in a room with your back to one wall and that an electron
beam, traveling horizontally from the back wall toward the front wall, is deflected to your
right. What is the direction of the field of magnetic induction that exists in the room?
* Thomson, no doubt for practical reasons, did not use this null method. He applied
the fields E and B separately and adjusted them to obtain the same beam deflection.
Equation 33-20 still holds. See Thomson's original paper: Philosophical Magazine,
44, 293 (1897).
Chap. 33 PROBLEMS 837
4. If an electron is not deflected in passing through a certain region of space, can we be
sure that there is no magnetic field in that region?
5. If a moving electron is deflected sideways in passing through a certain region of space,
can we be sure that a magnetic field exists in that region? .
6. A beam of protons is deflected sideways. Could this deflection be caused (a) by an
electric field? (b) By a magnetic field? (c) If either could be responsible, how would you
be able to tell which was present?
7. A conductor, even though it is carrying a current, has zero net charge. Why, then,
does a magnetic field exert a force on it?
8. Equation 33-11 (T = I' x B) shows that there is no torque on a current loop in an
external magnetic field if the angle between the axis of the loop and the field is (a) 0° or
(b) 180°. Discuss the nature of the equilibrium (that is, is it stable, neutral, or unstable?)
for these two positions.
9. In Example 4 we showed that the work required to turn a current loop end-for-end
in an external magnetic field is 2µB. Does this hold no matter what the original orienta-
tion of the loop was?
10. Imagine that the room in which you are seated is filled with a uniform magnetic
field with B pointing vertically upward. A circular loop of wire has its plane horizontal
and is mounted so that it is free to rotate about a horizontal axis. For what direction
of current in the loop, as viewed from above, will the loop be in stable equilibrium with
respect to forces and torques of magnetic origin?
11. A rectangular current loop is in an arbitrary orientation in an external magnetic
field. Is any work required to rotate the loop about an axis perpendicular to its plane?
12. (a) In measuring Hall potential differences, why must we be careful that points x
and yin Fig. 33-10 are exactly opposite to each other? (b) If one of the contacts is mov-
able, what procedure might we follow in adjusting it to make sure that the two points are
properly located?
13. A uniform magnetic field fills a certain cubical region of space. Can an electron be
fired into this cube from the outside in such a way that it will travel in a closed circular
path inside the cube?
14. Imagine the room in which you are seated to be filled with a uniform magnetic field
with B pointing vertically downward. At the center of the room two electrons are
suddenly projected horizontally with the same speed but in opposite directions. (a) Discuss
their motions. (b) Discuss their motions if one particle is an electron and one a positron.
15. In Fig. 33-4, why are the low-energy electron tracks spirals? That is, why does the
radius of curvature change in the constant magnetic field in which the chamber is im-
mersed?
16. What are the primary functions of (a) the electric field and (b) the magnetic field
in the cyclotron?
17. For Thomson's e/rn experiment to work properly (Section 33-8), is it essential that
the electrons have a. fairly constant speed?
PROBLEMS
_/l: The electrons in the beam of a television tube have an energy of 12 kev. The tube is
oriented so that the electrons move horizontally from south to north. The vertical com-
ponent of the earth's magnetic field points down and has B = 5.5 X 10-0 weber/meter2.
(a) In what direction will the beam deflect? (b) What is the acceleration of a given elec-
tron? (c) How far will the beam deflect in moving 20 cm through the television tube?
~fn a nuclear experiment a 1.0-Mev proton moves in a uniform magnetic field in a
circular path. What energy must (a) an alpha particle and (b) a deuteron have if they are
to circulate in the same orbit?
3. A wire 1.0 meter long carries a current of 10 amp and makes an angle of 30° with
a uniform magnetic field with B = 1.5 webers/meter2. Calculate the magnitude and di-
rection of the force on the wire.
838 THE MAGNETIC FIELD Chap. 33
/-4'<A wire of 60 cm length and mass 10 gm is suspended by a pair of flexible leads in a
magnetic field of induction 0.40 weber/meter2• What are the magnitude and direction
of the current required to remove the tension in the supporting leads? See Fig. 33-15.
x x x x x x
x x x x x x
x x x x x x
x x x x x x
~60cm-
Fig. 33-15
x x x x x x
x x x x x x
Hg
Fig. 33-17
Chap. 33 PROBLEMS 839
8. Figure 33-18 shows a wire ring of radius a at right angles to the general direction of
a radially symmetric diverging magnetic field. The magnetic induction at the ring is
everywhere of the same magnitude B, and its direction at the ring is everywhere at an
angle 8 ,vith a normal to the plane of the ring. The twisted lead wires have no effect on
the problem. Find the magnitude and direction of the force the field exerts on the ring if
the ring carries a current i as shown in the figure.
Fig. 33-18
9. A copper rod weighing 0.30 lb rests on two rails 1.0 ft apart and carries a current of
50 amp from one rail to the other. The coefficient of starting friction is 0.60. What is the
smallest magnetic field that would cause the bar to slide and what is its direction?
10. Figure 33-19 shows a wire of arbitrary shape carrying a current i between points a
and b. The wire lies in a plane at right angles to a uniform field of magnetic induction B.
Prove that the force on the wire is the same as that on a straight wire carrying a current i
directly from a to b. (Hint: Replace the wire by a series of "steps" parallel and perpen-
dicular to the straight line joining a and b.)
x x x x x x
b 10 cm
x x /,x x
/
x
/
x
rB
X/ x x
/
/
x }/ x x x x
a B
x x x x x x
x x x x x x 5cm
T = -1- £2'iB.
4'11'
840 THE MAGNETIC FIELD Chap. 33
14. Figure 33-21 shows a wooden cylinder with a mass m of 0.25 kg, a radius R, and a
length l of 0.1 meter with N equal to ten turns of wire wrapped around it longitudinally,
so that the plane of the wire loop contains the axis of the cylinder. What is the least cur-
rent through the loop that will prevent the cylinder from rolling down an inclined plane
whose surface is inclined at an angle () to the horizontal, in the presence of a vertical field
of magnetic induction 0.5 weber /meter 2, if the plane of the windings is parallel to the
inclined plane?
15. A current i, indicated by the crosses in Fig. 33-22, is established in a strip of copper
of height h and width w. A uniform field of magnetic induction Bis applied at right angles
to the strip. (a) Calculate the drift speed VJ for the electrons. (b) What are the magni-
tude and direction of the magnetic force F acting on the electrons? (c) What would the
magnitude and direction of a homogeneous electric field E have to be in order to counter-
balance the effect of the magnetic field? (d) What is the voltage V necessary between two
sides of the conductor in order to create this field E? Between which sides of the conductor
would this voltage have to be applied? (e) If no electric field is applied from the outside,
the electrons will be pushed somewhat to one side and therefore will give rise to a uniform
electric field Eu across the conductor until the forces of this electrostatic field EH balance
the magnetic forces encountered in part (b). What will be the magnitude and direction of
the field EH? Assume that n, the number of conduction electrons per unit volume, is
1.1 X 10 29 /meter 3 and that h = 0.02 meter, w = 0.1 cm, i = 50 amp, and B = 2 webers/
meter 2•
16. (a) Show that the ratio of the Hall electric field EH to the electric field E responsible
for the current is
EH B
-=-·
E nep
(b) What is the angle between EH and E? (c) Evaluate this ratio for the conditions of
Example 5.
17. A proton, a deuteron, and an a-particle, accelerated through the same potential
difference, enter a region of uniform magnetic field, moving at right angles to B. (a)
Compare their kinetic energies. (b) If the radius of the proton's circular path is 10 cm,
what are the radii of the deuteron and the a-particle paths?
18. A proton, a deuteron, and an a-particle with the same kinetic energies enter a re-
gion of uniform magnetic field, moving at right angles to B. Compare the radii of their
circul~J:,paths.
_.,..,,Hf'.'. An a-particle travels in a circular path of radius 0.45 meter in a magnetic field with
.B = 1.2 webers/meter 2 . Calculate (a) its speed, (b) its period of revolution, (c) its kinetic
energy, and (d) the potential difference through which it would have to be accelerated to
achieve this energy.
Chap. 33 PROBLEMS 841
20. An electron is accelerated through 15,000 volts and is then allowed to circulate at
right angles to a uniform magnetic field with B = 250 gauss. What is its path radius?
21. A beam of electrons whose kinetic energy is K emerges from a window at the end
of an accelerator tube. There is a metal plate a distance d from this window and at
right angles to the direction of the emerging beam. Show that we can stop the beam
from hitting the plate, if we apply a magnetic field B such that
B";>
: -. (2mK)Hi
~
ed
in which m and e are the electron mass and charge. How should B be oriented?
22. Show that the radius of curvature of a charged particle moving at right angles to a
magnetic field is proportional to its momentum.
23. What uniform magnetic field can be set up in space to permit a proton of speed
1.0 X 107 meters/sec to move in a circle the size of the earth's equator?
24. A deuteron in a large cyclotron is moving in a magnetic field with B = 1.5 webers/
meter 2 and an orbit radius of 2.0 meters. Because of a grazing collision with a target,
the deuteron breaks up, with a negligible loss of kinetic energy, into a proton and a neu-
tron. Discuss the subsequent motions of each. Assume that the deuteron energy is
shared equally by the proton and neutron at breakup.
25. A 2-kev positron is projected into a uniform field of induction B of 0.10 weber /meter 2
with its velocity vector making an angle of 89° with B. Convince yourself that the path
will be a helix, its axis being the direction of .B. Find the period, the pitch p, and the radius
r of the helix; see Fig. 33-23.
L /
/
/,.,..,,.---- --
I
v
I
I
I
__
/
//
_,,.,,...,,..,,.
---
Fig. 33-23
26. (a) In a magnetic field with B = 0.50 weber/meter2, for what path radius will an
electron circulate at 0.1 the speed of light? (b) What will its kinetic energy be?
842 THE MAGNETIC FIELD Chap. 33
27. Time-of-flight spectrometer. S. A. Goudsmit has devised a method for measuring
accurately the masses of heavy ions by timing their period of circulation in a known
magnetic field. A singly charged ion of iodine makes 7 rev in a field of 4.5 X 10-2 weber/
meter 2 in about 1.29 X 10-3 sec. What (approximately) is its mass in kilograms? Actu-
ally, the mass measurements are carried out to much greater accuracy than these approxi-
mate data suggest.
28. Mass spectrometer. Figure 33-24 shows an arrangement used by Dempster to mea-
sure the masses of ions. An ion of mass Mand charge +q is produced essentially at rest in
source S, a chamber in which a gas discharge is taking place. The ion is accelerated by
potential difference V and allowed to enter a field of magnetic induction B. In the field it
moves in a semicircle, striking a photographic plate at distance x from the entry slit and
being recorded. Show that the mass M is given by
B2q
M = - x2 •
8V
Fig. 33-24
29. Zeeman effect. In Bohr's theory of the hydrogen atom the electron can be thought
of as moving in a circular orbit of radius r about the proton. Suppose that such an atom
is placed in a magnetic field, with the plane of the orbit at right angles to B. (a) If the
electron is circulating clockwise, as viewed by an observer sighting along B, will the angular
frequency increase or decrease? (b) What if the electron is circulating counterclockwise?
Assume that the orbit radius does not change. [Hint: The centripetal force is now parti-
ally electric (FE) and partially magnetic (FB) in origin.]
30. In Problem 29 show that the change in frequency of rotation caused by the mag-
netic field is given approximately by
Be
~II=±--,
47r7n
Such frequency shifts were actually observed by Zeeman in 1896. (Hint: Calculate the
frequency of rotation without the magnetic field and also with it. Subtract, bearing in
mind that because the effect of the magnetic field is very small some--but not all-terms
containing B can be set equal to zero with little error.)
Chap. 33 PROBLEMS 843
31. The University of Pittsburgh cyclotron is normally adjusted to accelerate deuterons.
(a) What energy of protons could it produce, using the same oscillator frequency as that
used for deuterons? (b) What magnetic induction would be required? (c) What energy
of protons could be produced if the magnetic induction was left at the value used for deu-
terons? (d) What oscillator frequency would then be required? (e) Answer the same
questions for a-particles.
32. Estimate the total path length traversed by a deuteron in the University of Pitts-
burgh cyclotron during the acceleration process. Assume an accelerating potential be-
tween the <lees of 80,000 volts.
33. In a synchrocyclotron producing 400-Mev protons, what must the ratio be of the
oscillator frequency at the beginning of an accelerating cycle to that at the end? Such
a proton has a speed of 0.70c, where c is the speed of light.
34. A 10-kev electron moving horizontally enters a region of space in which there is a
downward-directed electric field of magnitude 100 volts/cm. (a) What are the magnitude
and direction of the (smallest) field of magnetic induction that will allow the electron to
continue to move horizontally? Ignore gravitational forces, which are rather small.
(b) Is it possible for a proton to pass through this combination of fields undeflected? If
so, under what circumstances?
35. An electric field of 1500 volts/meter and a magnetic field of 0.40 weber/meter2 act
on a moving electron to produce no force. (a) Calculate the minimum electron speed v.
(b) Draw the vectors E, B, and v.
Ampere's Law
CHAPTER 34
/
' ~
Fig. 34-2 Iron filings around a wire carrying a strong current. (Courtesy Physical
Science Study Committee.)
846 AMPERE'S LAW Chap. 34
tension of it made later by Maxwell, is one of the basic equations of electro-
magnetism (see Table 38-3).
We can gain an appreciation of the way Ampere's law developed histori-
cally by considering a hypothetical experiment which has, in fact, much in
common with experiments that were actually carried out. The experiment
consists of measuring B at various distances r from a long straight wire of
circular cross section and carrying a current i. This can be done by making
quantitative the qualitative observation of Fig. 34-1.
Let us put a small compass needle a distance r from the wire. Such a
needle, a small magnetic dipole, tends to line up with an external magnetic
field, with its north pole pointing in the direction of B. Figure 34-1 makes
it clear that B at the site of the dipole is tangent to a circle of radius r cen-
tered on the wire.
If the current in the wire of Fig. 34-1 is reversed in direction, all the com-
pass needles would reverse end-for-end. This experimental result leads to
the "right-hand rule" for finding the direction of B near a wire carrying a
current i: Grasp the wire with the right hand, the thumb pointing in the direction
of the current. The fingers will curl around the wire in the direction of B.
Let us now turn the dipole through an angle Ofrom its equilibrium position.
To do this, we must exert an external torque just large enough to overcome
the restoring torque T that will act on the dipole. T, 0, and B are related by
Eq. 33-11 (T = µ xB), which can be written in terms of magnitudes as
T = µ,B sin O (34-2)
and in which µ, is the magnitude of the magnetic moment of the dipole, 0
being the angle between the vectors µ and B. Even though we may not
know the value of µ, for the compass needle, we may take it to be a constant,
independent of the position or orientation of the needle. Thus by measur-
ing T and O in Eq. 34-2 we can obtain a relative measure of B for various
distances rand for various currents i in the wire. The experimental results
can be described by the proportionality
i
B o: -· (34-3)
r
We can convert this proportionality into an equality by inserting a propor-
tionality constant. As for Coulomb's law, and for similar reasons (see Sec-
tion 26-4), we do not write this constant simply ask but in a more complex
form, namely µ, 0 /21r, in which µ, 0 is called the permeability constant.* Equa-
tion 34-3 then becomes
µ,oi
B=--, (34-4)
21rr
Fig. 34-3 A circular path of integration surrounding a wire. The central dot suggests a
current i in the wire emerging from the page. Note that the angle between B and dI is
zero so that B·dl = B dl.
The left side of Eq. 34-5 can easily be shown to be f B·dl for a path
consisting of a circle of radius r centered on the wire. For all points on this
circle B has the same (constant) magnitude Band di, which is always tan-
gent to the path of integration, points in the same direction as B, as Fig.
34-3 shows. Thus
f B·dl = f B dl Bfdl
= = (B)(21rr),
which is Ampere's law. A host of other experiments suggests that Eq. 34-1
is true in general * for any magnetic field configuration, for any assembly
of currents, and for any path of integration.
In applying Ampere's law in the general case, we construct a closed linear
path in the magnetic field as shown in Fig. 34-4. This path is divided into
elements of length di, and for each element the quantity B·dl is evaluated.
Recall that B · di has the magnitude B dl cos () and can be interpreted as the
product of dl and the component of B ( = B cos (J) parallel to dl.
The inte-
* We must modify Eq. 34-1 if a time-varying electric field is present within the path
of integration. In this chapter we assume that if electric fields are present, they are
constant in magnitude and direction.
848 AMPERE'S LAW Chap. 34
gral is the sum of the quantities B · di for all path elements in the complete
loop. The term i on the right of Eq. 34-1 is the net current (strictly,
f j·dS) that passes through the area bounded by the closed path.
The permeability constant in Ampere's law has an assigned value of
µ0 = 411' X 10- 7 weber/amp-meter
Both this and the permittivity constant ( e0 ) occur in electromagnetic
formulas when the mks system of units is used.
The student may wonder why eo in Coulomb's law is a measured quantity, whereas
µo in Ampere's law is an assigned quantity. The answer is that the ampere, which is
the mks unit for the current i in Ampere's law, is defined by a laboratory technique
(the current bal,ance) that involves forces exerted by magnetic fields and in which this
same constant µ. 0 appears. In effect, as we show in detail in Section 34-4, the size of
the current that we agree to define as one ampere is adjusted so that µo may have
exactly the value assigned to it above. In Coulomb's law, on the other hand, the
quantities F, q, and rare measured in ways in which the constant eo plays no role.
This constant must then take on the particular value that makes the left side of
Coulomb's law equal to the right side; no arbitrary assignment is possible.
µoi
B=-· (34-4)
211'r
Sec. 34-2 B NEAR A LONG WIRE 849
Figure 34-5 shows a circular path of integration inside the wire. Symmetry
suggests that B is tangent to the path as shown. Ampere's law,
f B·dl = µ.oi,
11"T2
gives (B)(2'11"r) = µ.oio 11"R 2 •
since only the fraction of the current that passes through the path of integration is
included in the factor ion the right. Solving for B and dropping the subscript on the
current yields
At the surface of the wire (r = R) this equation reduces to the same expression as that
found by putting r = R in Eq. 34-4 (B = µoi/21rR).
850 AMPERE'S LAW Chap. 34
Example 2. Figure 34-6 shows a flat strip of copper of width a and negligible
thickness carrying a current i. Find the magnetic field at a distance R from the
center of the strip, at right angles to the strip.
x=O
Let us subdivide the strip into long infinitesimal filaments of width dx, each of
which may be treated as a wire carrying a current di given by i(dx/a). The field
contribution dB at point P in Fig. 34-6 is given, for the element shown, by the dif-
ferential form of Eq. 34-4, or
dB _ µo di _ µo i(dx/a)
- 271" ---;: - 27!" R sec((
= 2~R
µoi f sec dx
2 ().
Sec. ~4-3 MAGNETIC LINES OF INDUCTION 851
The variables x and 8 are not independent, being related by
x=Rtan8
or dx = R sec 2 () d8.
Bearing in mind that the limits on 8 are ±tan- 1 (a/2R) and eliminating dx from this
expression for B, we find
B = µ 0i f
R sec 2 8 d8
21raR sec 2 8
µ i
= ~ 0
I +tan-la/2R
d8 =
µoi a
-tan- 1 - ·
21ra -tan-'a/2R 1ra 2R
At points far from the strip, a/2R is a small angle, for which tan- 1 a ""a. Thus we
have, as an approximate result,
B "";~ (2~) = ; ; ~-
This result is expected because at distant points the strip cannot be distinguished
from a cylindrical wire (see Eq. 34-4). ....
),
--------------- J
Fig. 34-8 Lines of B near a long current-carrying wire immersed in a uniform external
field Be that points to the right. The current i is emerging from the page.
that, like stretched rubber bands, they represent the site of mechanical
forces. On this picture can we not visualize that the wire in Fig. 34-8 will be
pushed up? Today we use lines of induction largely for purposes of visualiza-
tion. For quantitative calculations we use the field vectors, describing the
force on the wire in Fig. 34-8, for example, from the relation F = ii x B.
In applying this relation to Fig. 34-8, we recall that B is always the
external field in which the wire is immersed; that is, it is Be and thus points
to the right. Since I points out of the page, the magnetic force on the wire
( = il x Be) does indeed point up. It is necessary to use only the external
field in such calculations because the field set up by the current in the wire
cannot exert a force on the wire, just as the gravitational field of the earth
cannot exert a force on the earth itself but only on another body. In Fig.
34-7, for example, there is no magnetic force on the wire because no external
magnetic field is present.
Faraday's idea of lines of induction was instrumental in overthrowing the older
action-at-a-distance theory of magnetic (and electric) attraction. Like many new
ideas, it was not immediately accepted. In 1851, for example, Faraday wrote:
I declare that I can hardly imagine anyone who practically and numerically
knows this agreement [with the action-at-a-distance theory] to hesitate an instant
in the choice between this simple and precise action, on the one hand, and anything
so vague as lines of force, on the other hand.
Sec. 34-4 TWO PARALLEL CONDUCTORS 853
(34-6)
The vector rule of signs tells us that Fb lies in the plane of the wires and points
to the left in Fig. 34-9.
We could have started with wire b, computed the field of induction which
it produces at the site of wire a, and then computed the force on wire a. The
ib
854 AMPERE'S LAW Chap. 34
force on wire a would, for parallel currents, point to the right. The forces
that the two wires exert on each other are equal and opposite, as they must
be according to Newton's law of action and reaction. For antiparallel cur-
rents the two wires repel each other.
This discussion reminds us of our discussion of the electric field between
two point charges in Section 27-1. There we saw that the charges act on
each other through the intermediary of the electric field. The conductors in
Fig. 34-9 act on each other through the intermediary of the magnetic field.
We think in terms of
current ~ field
current ~ current.
The attraction between long parallel wires is used to define the ampere.
Suppose that the wires are 1 meter apart (d = 1.0 meter) and that the two
currents are equal (ia = ib = i). If this common current is adjusted until,
by measurement, the force of attraction per unit length between the wires
is 2 X 10-7 nt/meter, the current is defined to be 1 ampere. From Eq. 34-6,
* Note that µ 0 appears in this relation used to define the ampere. As stated on page
848, µo is assigned the (arbitrary) value of 41r X 10-7 weber /amp-m, and the size of the
current that we define as 1 ampere is adjusted to give the required force of attraction per
unit length.
Sec. 34-4 TWO PARALLEL CONDUCTORS 855
B = Ba + Bb = µoi
-
(1 + -1- )·
-
271" x d- x
856 AMPERE' s LAw Chap. 34
This relationship does not hold for points inside the wires because Eq. 34-4 is not
valid there. <1111
•P
Fig. 34-13 A solenoid of finite length. The right end, from which lines of B emerge,
behaves like the north pole of a compass needle. The left end behaves like the south pole.
For points such as P in Fig. 34-12 the field set up by the upper part of
the solenoid turns (marked O) points to the left and tends to cancel the
field set up by the lower part of the solenoid turns (marked ®), which points
to the right. As the solenoid becomes more and more ideal, that is, as it ap-
proaches the configuration of an infinitely long cylindrical current sheet, the
field of induction at outside points approaches zero. Taking the external
field to be zero is not a bad assumption for a practical solenoid if its length is
much greater than its diameter and if we consider only external points near
the central region of the solenoid, that is, away from the ends. Figure 34-13
shows the lines of induction for a real solenoid, which is far from ideal in
that the length is not much greater than the diameter. Even here the spac-
ing of the lines of induction in the central plane shows that the external field
is much weaker than the internal field.
Let us apply Ampere's law,
f B·dl = µ.oi,
to the rectangular path abed in the ideal solenoid of Fig. 34-14. We write
the integral f B·dl as the sum of four integrals, one for each path segment:
f B·dl = ib +ic
B·dl B·dl + f B·dl + iaB·dl.
d
or (34-7)
Although Eq. 34-7 was derived for an infinitely long ideal solenoid, it
holds quite well for actual solenoids for internal points near the center of
the solenoid. It shows that B does not depend on the diameter or the length
of the solenoid and that B is constant over the solenoid cross section. A
solenoid is a practical way to set up a known uniform magnetic field for
experimentation, just as a parallel-plate capacitor is a practical way to set
up a known uniform electric field.
.... Example 5. A solenoid is 1.0 meter long and 3.0 cm in mean diameter. It
has five layers of windings of 850 turns each and carries a current of 5.0 amp.
(a) What is B at its center? From Eq. 34-7,
From symmetry the lines of B form concentric circles inside the toroid, as shown
in the figure. Let us apply Ampere's law to the circular path of integration of radius r:
f B·dl = µoi
or (B)(27rr) = µoioN,
where io is the current in the toroid windings and N is the total number of turns.
This gives
B = µo ioN_
27f r
In contrast to the solenoid, Bis not constant over the cross section of a toroid. Show
from Ampere's law that B equals zero for points outside an ideal toroid. .._
•b __ x'
------
•C
r dB
p
Fig. 34-16 The current element di establishes a magnetic field contribution dB at point P.
ment at the point in question. We found the field Eat that point by adding,
that is, by integrating, the field contributions dE for the entire distribution.
We now describe a similar procedure for computing Bat any point due to
an arbitrary current distribution. We divide the current distribution into
current elements and, using the law of Biot and Savart (which we describe
below), we calculate the field contribution dB due to each current element
at the point in question. We find the field B at that point by integrating
the field contributions for the entire distribution.
Figure 34-16 shows an arbitrary current distribution consisting of a cur-
rent i in a curved wire. The figure also shows a typical current element; it
is a length di of the conductor carrying a current i. Its direction is that of
the tangent to the conductor (dashed line). A current element cannot exist
as an isolated entity because a way must be provided to lead the current into
the element at one end and out of it at the other. Nevertheless, we can
think of an actual circuit as made up of a large number of current elements
placed end to end.
Let P be the point at which we want to know the magnetic induction dB
associated with the current element. According to the Biot-Savart law, dB
is given in magnitude by
µ 0 i dl sine
dB=- , (34-8)
41!' r2
where r is a displacement vector from the element to P and e is the angle
between this vector and di. The direction of dB is that of the vector dI x r.
In Fig. 34-16, for example, dB at point P for the current element shown is-
directed into the page at right angles to the plane of the figure. Note that
Eq. 34-8, being an inverse square law that describes the magnetic induction
due to a current element, may be viewed as the magnetic equivalent of
Coulomb's law, which is an inverse square law that describes the electric
field due to a charge element.
The law of Biot and Savart may be written in vector form as
µ 0 i dI x r
dB=-~~. (34-9)
411' r3
Sec. 34-6 THE BIOT-SAVART LAW 861
(34-10)
11
l_ : ~() '
dx
r x
11
11
11
11
Figure 34-17, a side view of the wire, shows a typical current element dx. The
magnitude of the contribution dB of this element to the magnetic field at P is found
from Eq. 34-8, or
dB = µoi dx sin () .
41r r2
The directions of the contributions dB at point P for all elements are the same,
namely, into the plane of the figure at right angles to the page. Thus the vector
integral of Eq. 34-10 reduces to a scalar integral, or
B = fdB = !!'.f!i
41r
fx=+"'
x=-oo
sin~ dx.
r
Now, x, (), and rare not independent, being related (see Fig. 34-17) by
r = Vx2 + R2
and sin () [ = sin (1r - 0)] = V R ,
x2 + R2
862 AMPERE'S LAW Chap. 34
so that the expression for B becomes
µoi f+oo R dx
B = 4; -oo (x2 + R2)%
µoi I X 1x=+oo
= 41rR (x 2 +R 2) 72 x=-oo
= µo £.
21r R
This is the result that we arrived at earlier for this problem (see Eq. 34-4). The law
of Biot and Savart will always yield results that are consistent with Ampere's law
and with experiment.
This problem reminds us of its electrostatic equivalent. We derived an expression
for E due to a long charged rod, using Gauss's law (Section 28-6); we also solved this
problem by integration methods, using Coulomb's law (Section 27-4).
Example 8. A circular current loop. Figure 34-18 shows a circular loop of radius
R carrying a current i. Calculate B for points on the axis.
dl
dB..L ___ dB
The vector di for a current element at the top of the loop points perpendicularly
out of the page. The angle Obetween di and r is 90°, and the plane formed by dl and r
is normal to the page. The vector dB for this element is at right angles to this plane
and thus lies in the plane of the figure and at right angles to r, as the figure shows.
Let us resolve dB into two components, one, dB II, along the axis of the loop and
another, dB ..l, at right angles to the axis. Only dB I contributes to the total induction
B at point P. This follows because the components dB11 for all current elements lie
on the axis and add directly; however, the components dB..l point in different direc-
tions perpendicular to the axis, and their resultant for the complete loop is zero, from
symmetry. Thus
where the integral is a simple scalar integration over the current elements.
For the current element shown in Fig. 34-18 we have, from the Biot-Savart law
(Eq. 34-8),
dB = l!.9'!. dl sin 90°.
41r r2
Sec. 34-6 THE BIOT-SAV ART LAW 863
We also have dB11 = dB cos a.
dB _ µoi cos a dl
Combining gives II - 41rr2
Figure 34-18 shows that r and a are not independent of each other. Let us express
each in terms of a new variable x, the distance from the center of the loop to the point
P. The relationships are
r = 2 yR +
x2 .
R R
= - =
and COSCI'.
r yR2 + x2 ·
Substituting these values into the expression for dB11 gives
µoiR
dB II = 41r(R2 + x2) % dl.
Note that i, R, and x have the same values for all current elements. Integrating this
equation, noting thatf dl is simply the circumference of the loop ( = 21rR), yields
(34-11)
...
If we put x » R in Example 8 so that points close to the loop are not
considered, Eq. 34-11 reduces to
µoi·R2
B=--·
2x3
Recalling that 1rR 2 is the area A of the loop and considering loops with N
turns, we can write this equation as
µ 0 (NiA) µo µ
B =- -
21r x 3 - 21r x3 '
whereµ is the magnetic dipole moment of the current loop. This reminds us
of the result derived in Problem 10, Chapter 27 [E = (1/2n 0 )(p/x 3 )], which
is the formula for the electric field strength on the axis of an electric dipole.
Thus we have shown in two ways that a current loop can be regarded as a
magnetic dipole: It experiences a torque given by T = µ. x B when placed in
an external magnetic field (Eq. 33-11); it generates its own magnetic field
given, for points on the axis, by the equation just developed.
864 AMPERE' s LAw Chap. 34
Table 34-1 is a summary of the properties of electric and magnetic dipoles.
Table 34-1
magnetic T = t,LXB
magnetic U = -1,L·B
~ Example 9. In the Bohr model of the hydrogen atom the electron circulates
around the nucleus in a path of radius 5.1 X 10- 11 meter at a frequency v of 6.8 X
10 15 rev/sec. (a) What value of Bis set up at the center of the orbit?
The current is the rate at which charge passes any point on the orbit and is given by
i = ev = (1.6 X 10-19 coul)(6.8 X 10 15/sec) = 1.1 X 10-3 amp.
Bat the center of the orbit is given by Eq. 34-11 with x = 0, or
B = µ,oiR2 = '!.oi
2(R2 + x2) u 2R
QUESTIONS
1. Can the path of integration around which we apply Ampere's law pass through a
conductor?
2. Suppose we set up a path of integration around a cable that contains twelve wires
with different currents (some in opposite directions) in each wire. How do we calculate i
in Ampere's law in such a case?
3. Apply Ampere's law qualitatively to the three paths shown in Fig. 34-19.
-, ,-,
I I
,_,
I
I
I
I
III •b
I
I I
I ,-,
,_,
II I
Fig. 34-19
PROBLEMS
1. A #10 bare copper wire (0.10 in. in diameter) can carry a current of 50 amp without
overheating. For this current, what is B at the surface of the wire?
2. A surveyor is using a compass 20 ft below a power line in which there is a steady
current of 100 amp. Will this interfere seriously with the compass reading? The hori-
zontal component of the earth's magnetic field at the site is 0.2 gauss.
3. A long straight wire carries a current of 50 amp. An electron, traveling at 10 7 meters/
sec, is 5.0 cm from the wire. What force acts on the electron if the electron velocity is
directed (a) toward the wire, (b) parallel to the wire, and (c) at right angles to the direc-
tions defined by (a) and (b)?
4. A conductor consists of an infinite number of adjacent wires, each infinitely long and
carrying a current i. Show that the lines of B will be as represented in Fig. 34-20 and that
B for all points in front of the infinite current sheet will be given by
B = }µ 0ni,
where n is the number of conductors per unit length. Derive both by direct application
of Ampere's law and by considering the problem as a limiting case of Example 2.
'-&. Figure 34-21 shows a hollow cylindrical conductor of radii a and b which carries a
current i uniformly spread over its cross section. (a) Show that the magnetic field B for
points inside the body of the conductor (that is, a < r < b) is given by
Check this formula for the limiting case of a = 0. (b) Make a rough plot of the general
behavior of B(r) from r = 0 tor ----> oo.
Chap. 34 PROBLEMS 867
6. A long coaxial cable consists of two concentric conductors with the dimensions shown
in Fig. 34-22. There are equal and opposite currents i in the conductors. (a) Find the
magnetic induction Bat r within the inner conductor (r < a). (b) Find B between the two
conductors (a < r < b). (c) Find B within the outer conductor (b < r < c). (d) Find B
outside the cable (r > c).
· 7. A long copper wire carries a current of 10 amp. Calculate the magnetic flux per
meter of wire for a plane surface S inside the wire, as in Fig. 34-23.
8. Two long, parallel #10 copper wires (diameter = 0.10 in.) carry currents of 10 amp
in opposite directions. (a) If their centers are 2.0 cm apart, calculate the flux per meter
that exists in the space between the axes of the wires. (b) What fraction of the flux in
(a) lies inside the wires? (c) Repeat the calculation of (a) for parallel currents.
9. A long wire carrying a current of 100 amp is placed in a uniform external magnetic
field of 50 gauss. The wire is at right angles to this external field. Locate the points at
wpich the resultant magnetic field is zero.
l,Q. Two long wires a distanced apart carry equal antiparallel currents i, as in Fig. 34-24.
Show that B at point P, which is equidistant from the wires, is given by
2µoid
B = -----
1r(4R2 + d2)
r
d R p
a a
11. Four long #10 copper wires are parallel to each other, their cross section forming a
square 20 cm on edge. A 20-amp current is set up in each wire in the direction shown in
Fig. 34-25. What are the magnitude and direction of Bat the center of the square?
12. In Problem 11 what is the force per meter acting on the lower left wire, in magnitude
and direction?
868 AMPERE'S LAW Chap. 34
,,,·13. Figure 34-26 shows a long wire carrying a cur-
rent of 30 amp. The rectangular loop carries a current
of 20 amp. Calculate the resultant force acting on
the loop. Assume that a = 1.0 cm, b = 8.0 cm, and
l = 30 cm.
14. Derive the solenoid equation (Eq. 34-7) starting
from the expression for the field on the axis of a circular
loop (Example 8). (Hint: Subdivide the solenoid into
0..
E 0.. a series of current loops of infinitesimal thickness and
ro E
0 ro integrate.)
(Y) 0
N 15. A square loop of wire of edge a carries a current i.
Show that the value of B at the center is given by
1 B=
2viµoi
1ra
.
wire is i, show that the magnetic induction at the center of the circle is given by
µoni
B = - tan (1r/n).
21ra
B=-
2µoi (l 2 + d 2) v, .
I
7r ld
!
(b) What does B reduce to for l » d? Is this a result
tliat you expect? Ip
21. Helmholtz coils. Two 300-turn coils are ar-
-T-
ranged a distance apart equal to their radius, as in
Fig. 34-29. For R = 5.0 cm and i = 50 amp, plot B
as a function of distance x along the common axis
over the range x = - 5 cm to x = +5 cm, taking
I
I
R
l
1
x = 0 at point P. (Such coils provide an especially
I
uniform field of B near point P.)
22. In Problem 21 let the separation of the coils be a I x
variable z. Show that if z equals R, then not only the R
first derivative (dB /dx) but also the second (d 2B /dx 2)
of B is zero at point P. This accounts for the uni-
Fig. 34-29
formity of B near point P for this particular coil sep-
aration.
23. A long "hairpin" is formed by bending a piece of wire as shown in Fig. 34-19. If a
10-amp current is set up, what are the direction and magnitude of B at point a? At
point b? Take R = 0.50 cm.
24. Calculate B at point P in Fig. 34-30. Assume that i = 10 amp and a = 8.0 cm.
25. A plastic disk of radius R has a charge q uniformly dis-
a/4 tributed over its surface. If the disk is rotated at an angular
frequency w about its axis, show that (a) the induction at the
l
a/4 center of the disk is
B = µowq
21rR
a
and (b) the magnetic dipole moment of the disk is
,...__________.J µ =
wqR 2
-4-·
e : - - a------:.
i-.e-l I
magnet and the coil. It makes no difference whether the magnet is moved
toward the coil or the coil toward the magnet.
The current that appears in this experiment is called an induced current
and is said to be set up by an induced electromotive force. Faraday was able
to deduce from experiments like this the law that gives their magnitude and
direction. Such emfs are very important in practice. The chances are good
that the lights in the room in which you are reading this book are operated
from an induced emf produced in a commercial electric generator.
In another experiment the apparatus of Fig. 35-2 is used. The coils are
placed close together but at rest with respect to each other. When the
switch S is closed, thus setting up a steady current in the right-hand coil, the
galvanometer deflects momentarily; when the switch is opened, thus in-
terrupting this current, the galvanometer again deflects momentarily, but in
the opposite direction. No gross objects are moving in this experiment.
In Faraday's words:
When the contact was made, there was a sudden and very slight effect at the
galvanometer, and there was also a similar slight effect when the contact with the
battery was broken. But whilst the voltaic current was continuing to pass through
the one helix, no galvanometrical appearances nor any effect like induction upon
the other helix could be perceived, although the active power of the battery was
proved to be very great . . . .
Experiment shows that there will be an induced emf in the left coil of
Fig. 35-2 whenever the current in the right coil is changing. It is the rate
at which the current is changing and not the size of the current that is significant.
.... Example 1. A long solenoid has 200 turns/ cm and carries a current of 1.5 amp;
its diameter is 3.0 cm. At its center we place a 100-turn, close-packed coil of diameter
2.0 cm. This coil is arranged so that B at the center of the solenoid is parallel to its
axis. The current in the solenoid is reduced to zero and then raised to 1.5 amp in
the other direction at a steady rate over a period of 0.050 sec. What induced emf
appears in the coil while the current is being changed?
The induction Bat the center of the solenoid is given by Eq. 34-7, or
The flux goes from an initial value of 1.2 X 10-5 weber to a final value of -1.2 X
10-5 weber. The change in flux Llcf>B for each turn of the coil during the 0.050-sec
Sec. 35-3 LENZ'S LAW 873
period is thus twice the initial value. The induced emf is given by
Fig. 35-3 If the magnet is moved toward the loop, the induced current points as shown,
setting up a magnetic field that opposes the motion of the magnet.
Let us apply Lenz's law to Fig. 35-3 in a different way. Figure 35-4
shows the lines of B for the bar magnet.* On this point of view the "change"
is the increase in <PB through the loop caused by bringing the magnet nearer.
The induced current opposes this change by setting up a field that tends to
oppose the increase in flux caused by the moving magnet. Thus the field
* There are two fields of Bin this problem-one connected with the current loop and one
with the bar magnet. The student must always be certain which one is meant.
Sec. 35-4 INDUCTION-A QUANTITATIVE STUDY 875
due to the induced current must point from left to right through the plane
of the coil, in agreement with our earlier conclusion.
It is not significant here that the induced field opposes the magnet field
but rather that it opposes the change, which in this case is the increase in
<l>B through the loop. If we withdraw the magnet, we reduce <l>B through the
loop. The induced field will now oppose this decrease in <l>B (that is, the
change) by re-enforcing the magnet field. In each case the induced field
opposes the change that gives rise to it.
fx x x x x x- x x x x YB
Ix x x Ix x x x ; 2 x x x I
1x x x x x x x x x i x x J
Ix x x ; x x x x x x x I
J F1 I
~ x x x x x x x x xi
~ x x x x x x x x x xi
I I
Ix x x x x x x x x i x x
:x x x x x x x F3 x x x I
[.:'.___ __'! _ ~ ~- _'!_ ~ _x_ .2..__X- ~
Fig. 35-5 A rectangular loop is pulled out of a magnetic field with velocity v.
876 FARADAY'S LAW Chap. 35
Note that the situation described by Fig. 35-5 does not differ in any essen-
tial particular from that of Fig. 35-4. In each case a conducting loop and
a magnet are in relative motion; in each case the flux of the field of the mag-
net through the loop is being caused to change with time.
The flux cf>s enclosed by the loop in Fig. 35-5 is
cf>s = Blx,
where lx is the area of that part of the loop in which Bis not zero. The emf
8 is found from Faraday's law, or
dcf>s d dx
8 = - - = - - (Blx) -Bl-= Blv (35-3)
dt dt dt '
where we have set -dx/dt equal to the speed v at which the loop is pulled
out of the magnetic field. Nate that the only dimension of the loop that
enters into Eq. 35-3 is the length l of the left end conductor. As we shall
see later, the induced emf in Fig. 35-5 may be regarded as localized here.
An induced emf such as this, produced by pulling a conductor through a
magnetic field, is sometimes called a motional emf.
The emf Blv sets up a current in the loop given by
. 8 Elv
i = - = -, (35-4)
R [(
where R is the loop resistance. From Lenz's law, this current (and thus 8)
must be clockwise in Fig. 35-5; it opposes the "change" (the decrease in
cf>s) by setting up a field that is parallel to the external field within the loop.
The current in the loop will cause forces Fi, F 2 , and F 3 to act on the three
conductors, in accord with Eq. 33-6a, or
F=ilxB. (35-5)
Because F 2 and F 3 are equal and opposite, they cancel each other; F 1 , which
is the force that opposes our effort to move the loop, is given in magnitude
from Eqs. 35-5 and 35-4 as
The agent that pulls the loop must do work at the steady rate of
B2l2v2
P = F1v = - - · (35-6)
R
From the principle of the conservation of energy, Joule heat must appear in
the resistor at this same rate. We introduced the conservation-of-energy
principle into our derivation when we wrote down the expression for the
current (Eq. 35-4); the student will recall that the relation i = 8/R for
Sec. 35-4 INDUCTION-A QUANTITATIVE STUDY 877
R R
which is indeed the expected result. This example provides a quantitative
illustration of the conversion of mechanical energy (the work done by an ex-
ternal agent) into electrical energy (the induced emf) and finally into thermal
energy (the Joule heating).
Figure 35-6 shows a side view of the coil in the field. In Fig. 35-6a the
coil is stationary; in Fig. 35-6b we are moving it to the right; in Fig. 35-6c
we are moving it to the left. The lines of induction in these figures repre-
sent the resultant fiel,d produced by the vector addition of the field B 0 due to
the magnet and the field Bi due to the induced current, if any, in the coil.
(a)
----------* (b)
-+---- (c)
Fig. 35-6 Side view of a rectangular loop in a magnetic field showing the loop (a) at rest,
(b) being pulled out, and (c) being pushed in.
878 FARADAY'S LAW Chap. 35
These lines suggest convincingly that the agent moving the coil always ex-
periences an opposing force.
~ Example 2. Figure 35-7 shows a rectangular loop of resistance R, width l, and
length a being pulled at constant speed v through a region of thickness d in which
a uniform field of induction B is set up by a magnet.
,---------1
I x x x x x I
rr:-:ix x~B
r
l
:x
I
xx x xi
I
Fig. 35-7 Example 2. A rectangular loop is
caused to move with a velocity v through a
L Ix xx x
.;>V
xi magnetic field. The position of the loop is
measured by x, the distance between the
I x x Ix x x : effective left edge of the field B and the right
~x_J I end of the loop.
I x
L
x x
_________
x x I
_J
----d---~
(a) Plot the flux <l>B through the loop as a function of the coil position x. Assume
that l = 4 cm, a = 10 cm, d = 15 cm, R = 16 ohms, B = 2.0 webers/meter2, and
v = 1.0 meter/sec.
The flux <I>B is zero when the loop is not in the field; it is Bla when the loop is
entirely in the field; it is Blx when the loop is entering the field and Bl[a - (x - d)J
when the loop is leaving the field. These conclusions, which the student should
verify, are shown graphically in Fig. 35-8a.
(b) Plot the induced emf f,,
The induced emf f, is given by f, = -d<I>B/dt, which can be written as
c, = _ d<J> B = _ d<J> B dx = _ d<J> B V
dt dxdt dx'
where d<I>B/dx is the slope of the curve of Fig. 35-8a. c.(x) is plotted in Fig. 35-8b.
Lenz's law, from the same type of reasoning as that used for Fig. 35-5, shows that
when the coil is entering the field the emf f, acts counterclockwise as seen from above.
Note that there is no emf when the coil is entirely in the magnetic field because the
flux <I>B through the coil is not changing with time, as Fig. 35-8a shows.
(c) Plot the rate P of Joule heating in the loop.
This is given by P = f, 2/R. It may be calculated by squaring the ordinate of the
curve of Fig. 35-8b and dividing by R. The result is plotted in Fig. 35-8c.
If the fringing of the magnetic field, which cannot be avoided in practice (see
Problem 34-28), is taken into account, the sharp bends and corners in Fig. 35-8 will
be replaced by smooth curves. What changes would occur in the curves of Fig. 35-8
if the coil were open circuited?
Example 3. A copper rod of length L rotates at angular frequency w in a uniform
field of magnetic induction B as shown in Fig. 35-9. Find the emf f, developed be-
tween the two ends of the rod.
If a wire of length di is moved at velocity vat right angles to a field B, a motional
emf de. will be developed (see Eq. 35-3) given by
de. = Bv dl.
Sec. 35-4 INDUCTION-A QUANTITATIVE STUDY 879
The rod of Fig. 35-9 may be divided into elements of length dl, the linear speed v of
each element being wl. Each element is perpendicular to B and is also moving in a
direction at right angles to B so that
S = f de = LL
Bv dl = £
L
B(wl) dl = !BwL2.
For a second approach, consider that at any instant the flux enclosed by the sector
aOb in Fig. 35-9 is given by
<l>B = BA = B(!L20),
880 FARADAY'S LAW Chap. 35
__,,.... --x
_...----
x x x --
x- ......._
/
/x x x x x x x
/
/ x x x x x x x x
/
/ x x x x x x x
I
/x x x x x x x x x x
I
x x x x x x x x x x x
I
Ix x x x x x x x x x x x x
I
Ix x x x x x x x x x x
I
Ix x x x x x x XO
Ix x x x x x x x x x x x x x x
I
\
x x I
\x x x x x x x x x x x x Bx x x x x/
\ I
\ x x x x x x x x x x x x x x x
I
\
\x x x x x x x x x x x x x x x I
\ I
\ x x x x x x x x x x x x x /
\ /
x x x x x x x x x x x
""' "' /
/
- ----- -
x x x x x x x /
"x x/
....... ......._ /
/
~ x x x x-
where !L20 can be shown to be the area of the sector. Differentiating gives
From Faraday's law, this is precisely the magnitude of 8, which agrees with the result
just derived. "41
To fix these ideas, consider Fig. 35-10, which shows a uniform field of in-
duction B at right angles to the plane of the page. We assume that B is
increasing in magnitude at the same constant rate dB/dt at every point.
This could be done by causing the current in the windings of the electro-
magnet that establishes the field to increase with time in the proper way.
The circle of arbitrary radius r shown in Fig. 35-10 encloses, at any in-
stant, a flux ipB· Because this flux is changing with time, an induced emf
given by 8 = -dipB/dt will appear around the loop. The electric fields E
induced at various points of the loop must, from symmetry, be tangent to
the loop. Thus the electric lines of force that are set up by the changing
magnetic field are in this case concentric circles.
If we consider a test charge q0 moving around the circle of Fig. 35-10, the
work W done on it per revolution is, in terms of the definition of an emf,
simply 8q0 . From another point of view, it is (q0 E)(21rr), where q0 E is the
force that acts on the charge and 27rf is the distance over which the force
acts. Setting the two expressions for W equal and canceling q0 yields
8 = E21rr. (35-7)
In a more general case than that of Fig. 35-10 we must write
8 = f E·dl. (35-8)
If this integral is evaluated for the conditions of Fig. 35-10, we obtain Eq.
35-7 at once. If Eq. 35-8 is combined with Eq. 35-1 (8 = -dipB/dt),
Faraday's law of induction can be written as
f E·dl = - dipB'
dt
(35-9)
.,,,..,1(-x-x---....
/ .......
/ ........
/ x x x '\,
I \
x x x x x \
I \
Fig. 35-10 The induced electric fields
{xxR~rx \
at four points produced by an in-
creasing magnetic field. I xE x x x x x x x,
\
\x xx x x rI
\ I
{ x x x /
'
.... ~
........
_____
x x
E
x ,.,-x
'/
/
,X.._B
882 FARADAY'S LAW Chap. 35
..... Example 4. Let B in Fig. 35-10 be increasing at the rate dB/ dt. Let R be
the radius of the cylindrical region in which the magnetic field is assumed to exist.
What is the magnitude of the electric field Eat any radius r? Assume that dB/dt =
0.10 weber/m 2-sec and R = 10 cm.
(a) For r < R, the flux <I>s through the loop is
<I>s = B(1rr 2 ).
Substituting into Faraday's law (Eq. 35-9),
,{: d<I>s
yE·dl = - -dt
d<I>s dB
yields (E)(21rr) = - -dt = - 2
(1rr) dt.
1 dB
Solving for E yields E= - -r-·
2 dt
The minus sign is retained to suggest that the induced electric field E acts to oppose
the change of the magnetic field. Note that E(r) depends on dB/dt and not on B.
Substituting numerical values, assuming r = 5 cm, yields, for the magnitude of E,
E = dB
!r -d = (!)(5 X 10-2 meter) (0.10 2weber) = 2.5 X 10-3 volt/meter.
t m -sec
(b) For r >R the flux through the loop is
This equation is true because B·dS is zero for those points of the loop that lie outside
the effective boundary of the magnetic field.
From Faraday's law (Eq. 35-9),
d<I>s dB
(E)(21rr) = - dt = - (1rR 2 ) dt.
1 R 2 dB
Solving for E yields E=-2rdt.
These two expressions for E(r) yield the same result, as they must, for r = R. Figure
35-11 is a plot of the magnitude of E(r) for the numerical values given. ~
6
....
Q)
+-'
Q)
E
:;::.- 4
0
>
,:,::,
I 2
0
,-t
~
0 ',<,
0 10 20 30 40
r, cm
Fig. 35-11 Example 4. If the fringing of the field in Fig. 35-10 were to be taken into
account, the result would be a rounding of the sharp cusp at r = R ( = 10 cm).
Sec. 35-5 TIME-VARYING MAGNETIC FIELDS 883
However, when changing magnetic flux is present, f E·dl is precisely not zero but
is, according to Faraday's law (see Eq. 35-9), -dif>B/dt. Electric fields associated
with stationary charges are conservative, but those associated with changing magnetic
fields are nonconservative; see Section 8-2. Electric potential, which can be defined
only for a conservative force, has no meaning for electric fields produced by induction.
Fig. 35-13 A 100-Mev betatron. M shows the magnet and D the region in which the
"doughnut" is located. (Courtesy General Electric Company.)
Sec. 35-6 THE BETATRON 885
Fig. 35-14 Cross section of a betatron, showing magnet M, coils C, and "doughnut" D.
Electrons emerge from the page at the left and enter it at the right.
(c) it keeps the radius of the orbit in which the electrons are moving a con-
stant; (d) it introduces the electrons into the orbit initially and removes
them from the orbit after they have reached full energy; and finally (e) it
provides a restoring force that resists any tendency for the electrons to leave
their orbit, either vertically or radially. It is remarkable that it is possible
to do all these things by proper shaping and control of the magnetic field.
The object marked D in Fig. 35-14 is an evacuated glass "doughnut"
inside which the electrons travel. Their orbit is a circle at right angles to the
plane of the figure. The electrons emerge from the plane at the left ( ·) and
enter it at the right (x). In the General Electric machine the radius of the
electron path is 33 in. The coils C and the 130-ton steel magnet shown in
Fig. 35-13 provide the magnetic flux that passes through the plane of this
orbit.
The current in coils C is made to alter periodically, 60 times/sec, to produce
a changing flux through the orbit, shown in Fig. 35-15. Here «l>B is taken as
positive when B is pointing up, as in Fig. 35-14. If the electrons are to
circulate in the direction shown, they must do so during the positive half-
cycle, marked ac in Fig. 35-15. The student should verify this (see Section
33-6). The electrons are accelerated by electric fields set up by the changing
flux. The direction of these induced fields depends on the sign of dif>B/dt
and must be chosen to accelerate, and not to decelerate, the electrons. Thus
only half the positive half-cycle in Fig. 35-15 can be used for acceleration;
it will prove to be ab.
The average value of dif>B/dt during the quarter-cycle ab is the slope of the
dashed line, or
1.8 weber
- - - - - = 430 volts.
4.2 X 10-3 sec
886 FARADAY'S LAW Chap. 35
Acceleration
Deceleration
Counter-
clockwise Clockwise
rotation rotation
From Faraday's law (Eq. 35-1), this is also the emf in volts. The electron
will thus increase its energy by 430 ev every time it makes a trip around the
orbit in the changing flux. If the electron gains only 430 ev of energy per
revolution, it must make about 230,000 rev to gain its full 100 Mev. For an
orbit radius of 33 in., this corresponds to a length of path of some 750 miles.
The betatron provides a good example of the fact that electric potential has no
meaning for electric fields produced by induction. If a potential exists, it must be true
that, as Eq. 35-10 shows, f E·dl = 0 for any closed path. In the betatron, however,
this integral, evaluated around the orbit, is precisely not zero but is, in our example,
430 volts. It must not be thought, of course, that the betatron violates the conserva-
tion of energy principle. The gain in kinetic energy of the circulating electron (430
ev/rev) must be supplied by an identifiable energy source. It comes, in fact, from
the generator that energizes the magnet coils, thus providing the changing magnetic
field. The energy is transmitted to the electron through the intermediary of this
changing field.
x
1
v
x x x x x x
x x x x x x
y'
1---- - - - - --1
jx xx xx x 1
I x x x x x
I Fe_,
I x x + x x x
-v I
x x x x x
(b)
Fig. 35-16 A closed conducting loop in relative motion with respect to a magnetic
field. (a) An observer S, fixed with respect to the magnet that produces the field B,
sees the loop moving to the right. ( b) An observer S', fixed with respect to the loop,
sees the magnet moving to the left.
Figure 35-16 shows a closed loop which is caused to move at velocity v with
respect to a magnet that provides a uniform field Bin the region shown. We con-
sider first an observer, identified as S, who is at rest with respect to the magnet used
to establish the field B; see Fig. 35- l 6a. The induced emf in this case is called a
motional emf because the conducting loop is moving with respect to this observer.
Consider a positive charge carrier at the center of the left end of the conducting
loop. To observer S, this charge, constrained to move to the right along with the
888 FARADAY'S LAW Chap. 35
loop, is a charge q moving, for an instant at least when the loop has just been set
in motion, with a velocity v in the magnetic field B and as such it experiences a
sideways magnetic deflecting force given by Eq. 33-3a (F = qv X B). This
force causes the carriers to move upward along the conductor, so that they acquire
a drift velocity Vd also, as shown in Fig. 35-16a.
The resultant equilibrium speed of the carrier when the velocity v of the loop
has become constant is now V, which we find by adding v and va vectorially in
:Fig. 35-16a. The magnetic deflecting force Fm is, as always, at right angles to
the resultant velocity V of the carrier and is given by
Fm= qV X B (35-13)
Now Fm acting alone would tend to push the carriers through the left wall of the
conductor. Because this does not happen the conductor wall must exert a normal
force Non the carriers(see Fig. 35-16a)of magnitude such that va lies parallel to
the axis of the wire; in other words, N exactly cancels the horizontal component of
Fm, leaving only the component Fm cos e that lies along the direction of the con-
ductor. This component of force on the carrier is also cancelled out, this time
by the average force Fi associated with the internal collisions that the carrier
experiences as it drifts with (constant) speed Vd through the wire.
The kinetic energy of the charge carrier as it drifts through the wire remains
constant. This is consistent with the fact that the resultant force acting on the
charge carrier ( = Fm+ Fi+ N) is zero. The work done by Fm is zero because
magnetic forces, acting at right angles to the velocity of a moving charge, can do
no work on that charge. Thus the (negative) work done on the carrier by the
average internal collision force Fi must be exactly cancelled by the (positive)
work done on the carrier by the force N. In the last analysis N is exerted by the
agent who pulls the loop through the magnetic field, and the mechanical energy
expended by this agent appears as heat energy in the loop, as we have seen in
Section 35-4.
Let us then calculate the work dW done on the carrier in time dt by the force N;
it is
dW = N(v dt) (35-14)
in which v dt is the distance that the loop (and the carrier) have moved to the right
in Fig. 35-16a in time dt. We can write for N (see Fig. 35-16a and Eq. 35-13).
dW = (qBvd)(v dt)
= (qBv) (vd dt) = qBv dl (35-16)
in which dl( = vd dt) is the distance the carrier drifts along the conductor in
time dt.
The work done on the carrier as it makes a complete circuit of the loop is found
by integrating Eq. 35-16 around the loop and is
8=E·dl
which reduced to
8 = El (35-19)
in this case. This is so because no induced electric field appears in the upper and
lower bars, because of the nature of their motions, and none appears in the part
of the loop outside the magnetic field.
The emfs given by Eqs. 35-19 and 35-18 must be identical because the relative
motion of the loop and the magnet ii; identical in the two cases shown in Fig. 35-16.
Equating these relations yields
El= Blv,
E = vB. (35-20a)
In Fig. 35-16b the vector E points upward along the axis of the left end of the
conducting loop because this is the direction in which positive charges are observed
to drift. The directions of v and B are clearly shown in this figure. We see,
then, that Eq. 35-20a is consistent with the more general vector relation
E = v X B. (35-20b)
890 FARADAY'S LAW Chap. 35
We have not proved Eq. 35-20b except for the special case of Fig. 35-16; neverthe-
less it proves to be true in general, that is, no matter what the angle between v
and B.
An experiment carried out in 1926 by the German physicist Wilhelm Wien (1864-
1928) gives concrete support to our descriptions of the effect of the motion of an
observer on the nature of the electric and magnetic fields that he observes. The
radiations emitted from atoms vary slightly in wavelength if the atom is immersed
in either a magnetic or an electric field. The atom can thus be used as a probe
to examine the nature of such fields. Wien fired a beam of atoms with velocity
v through a magnetic field B. The radiations emitted were identical in the distribu-
tion of their wavelengths with those that would have been emitted by a resting
atom immersed in (a) the magnetic fa 11 B and (b) an induced electric field E given
by Eq. 35-20b.
We interpret Eq. 35-20b in the following way: Observer S fixed with respect to
the magnet is aware only of a magnetic field. The force to him arises from the
motion of the charges through B. Observer S' fixed on the charge carrier is aware
of an electric field E also and attributes the force on the charge (at rest with respect
to him initially) to the electric field. S says the force is of purely magnetic origin
and S' says the force is of purely electric origin. From the point of view of S,
the induced emf is given by §(v X B) · di. From the point of view of S', the
same induced emf is given by §E · di, where E is the ( induced) electric vector
that he observes at points along the circuit.
For a third observer S" who judges that both the magnet and the loop are
moving, the forc0 tending to move charges around the loop is neither purely
electric nor purely magnetic, but a bit of each. In summary, in the equation
F/q=E+vXB,
different observers form different assessments of E, B, and v but, when these are
combined, all observers form the same assessment of F /q and all obtain the same
value for the induced emf in the loop (this depends only on the relative motion).
That is, the total force is the same for all observers, but each observer forms a
different estimate of the separate electric and magnetic forces contributing to the
same total force .
.... Example 7. In Fig. 35-16 assume that B = 2.0 webers/meter2, l = 10 cm,
and v = 1.0 meter/sec. Calculate (a) the induced electric field observed by S',
and (b) the emf induced in the loop.
(a) The electric field, which is apparent only to observer S', is associated with
the moving magnet and is given in magnitude (see Eq. 35-20a) by
E = vB
= (1.0 meter/sec)(2.0 webers/meter 2 )
= 2.0 volt/meter.
8 = Blv
= (2.0 webers/meters 2) (1.0 X 10- 1 meter) (1.0 meter/sec)
= 0.20 volt.
Chap. 35 QUESTIONS 891
Observer S' would not regard the emf as motional and would use the relationship
8 = El
= 0.20 volt.
As must be the case, both observers agree as to the numerical vailue of the emf. <II(
QUESTIONS
1. The north pole of a magnet is moved away from a metallic ring, as in Fig. 35-17. In
the part of the ring farthest from the reader, which way does the current point?
Fig. 35-17
2. Eddy currents. A sheet of copper is placed in a magnetic field as shown in Fig. 35-18.
If we attempt to pull it out of the field or push it further in, an automatic resisting force
appears. Explain its origin. (Hint: Currents, called eddy currents, are induced in the
sheet in such a way as to oppose the motion.)
---
/
/
----- x x x ------ "'-- '--.
/
/
x
/
x x x x x
"x ~ B
I \
I x x x \
x x x x
I \
I x x x x x
I
I x x x x x
I
I
\ x x x x x
\
\ x x x x x I
\ I
\ x x x x x x /
"x /
/
""" " '-- '----- x x x/
---- ------- -----
Fig. 35-18
892 FARADAY'S LAW Chap. 35
3. Electromagnetic shielding. Consider a conducting sheet lying in a plane perpendicular
to a magnetic field B, as shown in Fig. 35-19. (a) If B suddenly changes, the full change
in Bis not immediately detected in region P. Explain. (b) If the resistivity of the sheet is
zero, the change is not ever detected at P.
Explain. (c) If B changes periodically at
high frequency and the conductor is made
of a material of low resistivity, the re-
gion near P is almost completely shielded
from the changes in flux. Explain. (d)
Is such a conductor useful as a shield
from static magnetic fields? Explain.
XX X•X
B
P Conducting xx:\: xx x
==*'"-~,#\¥%:%\\\\\'"'"* sheet
x :1:::
I
x
x,x
x x
-0-
I
·--·-
~d_j R
6. What is the direction, if any, of the conventional current through resistor R in Fig.
35-22 (a) immediately after switch Sis closed, (b) some time after switch S was closed, and
(c) immediately after switch S is opened. (d) When switch S is held closed, which end
of the coil acts as a north pole?
Chap. 35 QUESTIONS 893
----"v>
Fig. 35-23
8. If the resistance R in the left-hand circuit of Fig. 35-24 is increased, what is the
direction of the induced current in the right-hand circuit?
R A
9. In Fig. 35-25 the movable wire is moved to the right, causing an induced current as
shown. What is the direction of B in region A?
10. A loop, shown in Fig. 35-26, is removed from the magnet by pulling it vertically
upward. (a) What is the direction of the induced current? (b) Is a force required to remove
the loop? (c) Does the total amount of Joule heat produced in removing the loop depend
on the time taken to remove it?
Fig. 35-26
11. A magnet is dropped down a long vertical copper tube. Show that, even neglecting
air resistance, the magnet will reach a constant terminal velocity.
12. A magnet is dropped from the ceiling along the axis of a copper loop lying flat on the
floor. If the falling magnet is photographed with a time sequence camera, what differences,
if any, will be noted if (a) the loop is at room temperature and (b) the loop is packed in dry
ice?
894 FARADAY'S LAW Chap. 35
13. A copper ring and a wooden ring of the same dimensions are placed so that there is
the same changing magnetic flux through each. How do the induced electric fields in each
ring compare?
14. In Fig. 35-12 how can the induced emfs around paths 1 and 2 be identical? The
induced electric fields are much weaker near path 1 than near path 2, as the spacing of the
lines of force shows. See also Fig. 35-11.
15. In a certain betatron the electrons rotate counterclockwise as seen from above. In
what direction must the magnetic field point and how must it change with time while the
electron is being accelerated?
16. Why can a betatron be used for acceleration only during one-quarter of a cycle?
17. To make the electrons in a betatron orbit spiral outward, would it be necessary to
increase or to decrease the central flux? Assume that B at the orbit remains essentially
unchanged.
18. A cyclotron is a so-called resonance device. Does a betatron depend on resonance?
PROBLEMS
1. A hundred turns of insulated copper wire are wrapped around an iron cylinder of
cross-sectional area 0.001 meter 2 and are connected to a resistor. The total resistance in
the circuit is 10 ohms. If the longitudinal magnetic induction in the iron changes from
1 weber /meter 2 in one direction to 1 weber /meter 2 in the opposite direction, how much
charge flows through the circuit?
· 2. In Fig. 35-27 a closed copper coil with 100 turns and a total resistance of 5.0 ohms
is placed outside a solenoid like that of Example 1. If the current in the solenoid is changed
as in that example, what current appears in the coil?
Fig. 35-27
3. A circular loop of wire 10 cm in diameter is placed with its normal making an angle
of 30° v·ith the direction of a uniform 5000-gauss magnetic field. The loop is "wobbled"
so that its normal rotates about the field direction at the constant rate of 100 rev /min;
the angle between the normal and the field direction ( = 30°) remains unchanged during
this process. What emf appears in the loop?
4. A uniform field of induction B is normal to the plane of a circular ring 10-cm in
diameter made of #10 copper wire (diameter = 0.10 in.). At what rate must B change
with time if an induced current of 10 amp is to appear in the ring?
5. A uniform field of induction B is changing in magnitude at a constant rate dB /dt.
You are given a mass m of copper which is to be drawn into a wire ot radius r and formed
into a circular loop of radius R. Show that the induced current in the loop does not depend
Chap. 35 PROBLEMS 895
on the size of the wire or of the loop and, assuming B perpendicular to the loop, is given by
m dB
i = ~---,
47rp0 dt
appears in the loop. This is the principle of the commercial alternating-current generator.
(b) Design a loop that will produce an emf with GO = 150 volts when rotated at 60 rev/sec
in a field of magnetic induction of 5000-gauss.
fig. 35-28
9. A stiff wire bent into a semicircle of radius R is rotated with a frequency Pin a uni-
form field of induction B, as shown in Fig. 35-29. What are the amplitude and fre-
quency of the induced voltage and of the induced current when the internal resistance
of the meter Mis RM and the remainder of the circuit has negligible resistance?
x x x x x x
x x )( x x x
R
t
x > x ~x x
x x x x x x x
x x x x x x x
x x x x x x x
'--------i M i--------'
fig. 35-29
896 FARADAY'S LAW Chap. 35
10. A circular copper disk 10 cm in diameter rotates at 1800 rev /min about an axis
through its center and at right angles to the disk. A uniform field of induction B of 10,000
gauss is perpendicular to the disk. What potential difference develops between the axis
of the disk and its rim?
11. Figure 35-30 shows a copper rod moving with velocity v parallel to a long straight
wire carrying a current i. Calculate the induced emf in the rod, assuming that v = 5.0
meters/sec, i = 100 amp, a = 1.0 cm, and b = 20 cm.
12. A metal wire of mass rn slides without friction on two rails spaced a distance d
apart, as in Fig. 35-31. The track lies in a vertical uniform field of induction B. (a) A
constant current i flows from generator G along one rail, across the wire, and back down the
other rail. Find the velocity (speed and direction) of the wire as a function of time, as-
suming it to be at rest at t = 0. (b) The generator is replaced by a battery with constant
ernf 8. The velocity of the wire now approaches a constant final value. What is this ter-
minal speed? (c) What is the current in part (b) when the terminal speed has been reached?
13. In Fig. 35-32 the magnetic flux through the loop perpendicular to the plane of the
coil and directed into the paper is varying according to the relation
<f>s = 6t2 + 7t + 1,
where <f>R is in milliwebers (1 milliweber = 10-3 weber) and tis in seconds. (a) What is
the magnitude of the emf induced in the loop when t = 2 sec? (b) What is the direction
of the current through R?
x x x x x
x x
x x x x Ax x x x
D
x x x x x x x x
50cm
x x x x x x
c
R x x Bx x x x
14. In Fig. 35-33 a conducting rod AB makes contact with the metal rails AD and BC
which are 50 cm apart in a uniform magnetic field of induction 1.0 weber /meter 2 perpen-
dicular to the plane of the paper as shown. The total resistance of the circuit ABCD is 0.4
ohm (assumed constant). (a) What is the magnitude and direction of the emf induced in
the rod when it is moved to the left with a velocity of 8 meters/sec? (b) What force is re-
quired to keep the rod in motion? (c) Compare the rate at which mechanical work is done
by the force F with the rate of development of heat in the circuit.
15. A square wire of length l, mass rn, and resistance R slides without friction down
parallel conducting rails of negligible resistance, as in Fig. 35-34. The rails are connected
Chap. 35 PROBLEMS 897
to each other at the bottom by a resistanceless rail parallel to the wire, so that the wire
and rails form a closed rectangular conducting loop. The plane of the rails makes an
angle e with the horizontal, and a uniform vertical field of magnetic induction B exists
throughout the region. (a) Show that the wire acquires a steady-state velocity of magni-
tude
mgR sine
v = .
B 2t2 cos 2 e
(b) Prove that this result is consistent with the conservation-of-energy principle. (c)
What change, if any, would there be if B were directed down instead of up?
16. Prove that if the flux of magnetic induction through the coil of N turns of Fig.
35-35 changes in any way from <1>1 to <1>2, then the charge q that flows through the circuit
of total resistance R is given by
N(<1>2 - <1>1)
q=
R
~igure 35-36 shows a uniform field of induction B confined to a cylindrical volume
of radius R. Bis decreasing in magnitude at a constant rate of 100 gauss/sec. What is
the instantaneous acceleration (direction and magnitude) experienced by an electron
placed at a, at b, and at c? Assume r = 5.0 cm. (The necessary fringing of the field
beyond R will not change your answer as long as there is axial symmetry about a perpen-
dicular axis through b.)
// x
........ -------....
x x ....... ,
' "-
//
/x
/
/
x
x
x
x
t\ x
x R x
;>, x\
II ~
r \I
IX xx bx x xi
I r (B I
\x x x I x x x/
\ l I
\\ x x xa x x //
' '-
' x
.................. ______x x_....,,.,.,. /
/ /
Fig. 35-36
G = ~ ~~ R2 - GY.
/
/
/
/
----- --
x x x
......
'
' '-
I
I
/
x x x x x ""\
\
Ix x x x x x ,::"'x \
I
I B \
I
I x x x x x x I
I i I
\ R
\X x x x x x I
I
\
x x x x I
""'- /
---- x,,... /
Fig. 35-37
Show by graphical analysis that the relation B= 2BR is satisfied at the orbit radius,
E --,>.
I 1¥'(a
I I
I I
L _J
- q ..A
Fig. 35-38
Inductance
CHAPTER 36
36-1 Inductance
If two coils are near each other, a current i in one coil will set up a flux <I>B
through the second coil. If this flux is changed by changing the current, an
induced emf will appear in the second coil according to Faraday's law.
However, two coils are not needed to show an inductive effect. An induced
emf appears in a coil if the current in that same coil is changed. This is called
self-induction and the electromotive force produced is called a self-induced
emf. It obeys Faraday's law of induction just as other induced emfs do.
Consider first a "close-packed" coil, a toroid, or the central section of a
long solenoid. In all three cases the flux <I>B set up in each turn by a current
i is essentially the same for every turn. Faraday's law for such coils (Eq.
35-2) d(N<I>B)
8=----
dt
shows that the number of flux linkages N<I>B (N being the number of turns)
is the important characteristic quantity for induction. For a given coil,
provided no magnetic materials such as iron are nearby, this quantity is
proportional to the current i, or
N<I>B = Li, (36-2)
in which L, the proportionality constant, is called the inductance of the device.
From Faraday's law (see Eq. 36-1) the induced emf can be written as
d(N<I>s)
8=---- (36-3a)
dt
899
900 INDUCTANCE Chap. 36
8
Written in the form L=--, (36-3b)
di/dt
this relation may be taken as the defining equation for inductance for coils
of all shapes and sizes, whether or not they are close-packed and whether or
not iron or other magnetic material is nearby. It is analogous to the defining
relation for capacitance, namely
q
c =-·
v
If no iron or similar materials are nearby, L depends only on the geometry
of the device. In an inductor (symbol / OOOO\.) the presence of a magnetic
field is the significant feature, corresponding to the presence of an electric
field in a capacitor.
The unit of inductance, from Eq. 36-:3b, is the volt-sec/amp. A special
name, the henry, has been given to this combination of units, or
1 henry = 1 volt-sec/amp.
The unit of inductance is named after Joseph Henry (1797-1878), an Ameri-
can physicist and a contemporary of Faraday. Henry independently dis-
covered the law of induction at about the same time Faraday did. The units
millihenry (1 mh = 10-3 henry) and microhenry (1 µh = 10-6 henry) are
also commonly used.
The direction of a self-induced emf can be found from Lenz's law. Suppose
that a steady current i, produced by a battery, exists in a coil. Let us sud-
denly reduce the (battery) emf in the circuit to zero. The current i will
start to decrease at once; this decrease in current, in the language of Lenz's
law, is the "change" which the self-induction must oppose. To oppose the
falling current, the induced emf must point in the same direction as the
current, as in Fig. 36-la. When the current in a coil is increased, Lenz's
law shows that the self-induced emf points in the opposite direction to that of
the current, as in Fig. 36-lb. In each case the self-induced emf acts to op-
pose the change in the current. The minus sign in Eq. 36-3 shows that 8
and di/dt are opposite in sign, since Lis always a positive quantity.
(36-4)
Let us apply this equation to calculate L for a section of length l near the
center of a long solenoid. The number of flux linkages in the length l of the
solenoid is
NiJ>s = (nl) (BA),
where n is the number of turns per unit length, B is the magnetic induction
inside the solenoid, and A is the cross-sectional area. From Eq. :34-7, B is
given by
fB·dl = µoi,
i in Ampere's law is the total current that passes through the path of integration.
Solving for B yields
B = µ,oioN.
21rr
The flux cf> s for the cross section of the toroid is
cf>s f
= B·dS = .C(B)(h dr) = ib µ,;~~ h dr
µ,oioNhfb -
=~~-
dr= -µ,oioNh b
~ - n1 - ,
21r a r 21r a
where h dr is the area of the elementary strip shown in the figure.
The inductance follows from Eq. 36-4, or
36-3 An LR Circuit
In Section 32-8 we saw that if an emf 8 is suddenly introduced, perhaps
by using a battery, into a single loop circuit containing a resistor R and a
capacitor C the charge does not build up immediately to its final equilibrium
value ( = C8) but approaches it in an exponential fashion described by
Eq. 32-15, or
(36-6)
The delay in the nse of the charge is described by the capacitative time
constant Tc, defined from
Tc= RC. (36-7)
If in this same circuit the battery emf 8 is suddenly removed, the charge
does not immediately fall to zero but approaches zero m an exponential
fashion, described by Eq. 32-18b, or
(36-8)
The same time constant Tc describes the fall of the charge as well as its rise.
AN LR CIRCUIT 903
10
a S c R
(.,
-- i
di G -Rt/L
- =-e (36-11)
dt L
Substituting i and di/dt into Eq. 36-9 leads to an identity, as the student
can easily verify. Thus Eq. 36-10 is a solution of Eq. 36-9. Figure 36-5
shows how the potential difference V R across the resistor ( = iR; see Eq.
36-10) and V L across the inductor ( = L di/dt; see Eq. 36-11) vary with
time for particular values of e, L, and R. "The student should compare this
figure carefully with the corresponding figure for an RC circuit (Fig. 32-11).
We can rewrite Eq. 36-10 as
(36-12)
4
en
~
a,
c.
E
.!!!
.E
...;- 2
2 4 6 8 10
t, milliseconds
(a)
-
0
>
en
2 4 6 8 10
t, milliseconds
(b)
Fig. 36-5 If in Fig. 36-3 we assume that R = 2000 ohms, L = 4 henrys and e = 10
volts, then (a) shows the variation of i with t during the current buildup after switch Sis
closed on a, and (b) the variation of V L with t. The time constant is LIR = 2.0 X 10-3
sec.
will reach a value within 1/e (about 37%) of its final equilibrium value (see
Fig. 36-5).
If the switch S in Fig. 36-3, having been left in position a long enough
for the equilibrium current 8/R to be established, is thrown to b, the effect
is to remove the battery from the circuit. The differential equation that
906 INDUCTANCE Chap. 36
governs the subsequent decay of the current in the circuit can be found by
putting 8 = 0 in Eq. 36-9, or
di
L-+ iR = 0. (36-14)
dt
The student can show by the test of substitution that the solution of this
differential equation is
8
i = _ e-t/TL (36-15)
R
Just as for the RC circuit, the behavior of the circuit of Fig. 36-3 can be investi-
gated experimentally, using a cathode-ray oscilloscope. If switch S in this figure is
thrown periodically between a and b, the applied emf alternates between the values
8 and zero. If the terminals of an oscilloscope are connected across b and c in Fig.
36-3, the oscilloscope will display the waveform of this applied emf on its screen, as
in Fig. 36-6c. There is a practical problem that R and L for an actual coil are
physically inseparable; we are assuming an "ideal" resistor and an "ideal" inductor.
If the terminals of the oscilloscope are connected across the resistor, the waveform
displayed (Fig. 36-6a) will be that of the current in the circuit, since the potential
drop across R, which determines the oscilloscope deflection, is given by V R = iR.
During the intervals marked inc in Fig. 36-6. the current is increasing and the wave-
form (see Eq. 36-12) is given bv
V R ( = iR) = 8(1 - e-t/TL).
k-inc~dec~inc~dec-4-inc~
Fig. 36-6 Oscilloscope photograph showing the variation with time of (a) the potential
drop V R across the resistor, (b) the potential drop V L across the inductor, and (c) the
applied emf 8. During the intervals marked inc the current is increasing; during those
marked dee it is decreasing. Compare with Fig. 32-13. (Courtesy E. K. Hege.)
Sec. 36-4 ENERGY AND THE MAGNETIC FIELD 907
During the intervals marked dee, the current is decreasing and V R (see Eq. 36-15) is
given by
Note that both the growth and the decay of the current are delayed.
If the oscilloscope terminals are connected across the inductor, the screen will show
a plot of the potential difference across it as a function of time (Fig. 36-6b). While
the current is increasing, the equation of the trace (see Eq. 36-11) should be
When the current is decreasing, V Lis given in terms of the time derivative of Eq.
36-15 and is
VL
di
dt(= L-) =
-se-tfTL
.
Note that V L is opposite in sign when the current is increasing (di/dt positive)
and when it is decreasing (di/dt negative), as is true also for the induced emf SL[=
-L(di/dt) = -VLJ,
Examination of Fig. 36-6 shows that at any instant the sum of curves a and b
always yields curve c. This is an expected consequence of the loop theorem, as Eq.
36-9 shows.
ohms. If it is connected to a 100-volt battery, how long will it take for the current
to reach one-half its final equilibrium value?
The equilibrium value of the current is reached as t ---+ c,:i; from Eq. 36-12 it is
8/R. If the current has half this value at a particular time t 0, this equation becomes
50 henrys)
t 0 = 0.69r L = 0.69 ( 30 ohms = 1.2 sec.
uE = 1
zEo E2 ,
where Eis the electric field strength at the point in question. Although this
formula was derived for a parallel-plate capacitor, it holds for all kinds of
electric field configurations.
Energy can also be stored in a magnetic field. For example, two parallel
wires carrying currents in the same direction attract each other, and to pull
908 INDUCTANCE Chap: 36
them further apart work must be done. It is useful to think that this ex-
pended energy is stored in the magnetic field between and around the wires.
The energy can be recovered from the field if the wires are allowed to move
back to their original separation. In the electrostatic case the same argu-
ment was applied to the pulling apart of two unlike charges and in the gravi-
tational case to the pulling apart of two masses.
To derive a quantitative expression for the storage of energy in the mag-
netic field, consider Fig. 36-4, which shows a source of emf 8 connected to a
resistor R and an inductor L.
di
8 = iR + L-, (36-9)
dt
is the differential equation that describes the growth of current in this circuit.
We stress that this equation follows immediately from the loop theorem and
that the loop theorem in turn is an expression of the principle of conservation
of energy for single-loop circuits. If we multiply each side of Eq. 36-9 by i,
we obtain
di
Si = i 2 R + Li - , (36-16)
dt
which has the following physical interpretation in terms of work and energy:
1. If a charge dq passes through the seat of emf 8 in Fig. 36-4 in time dt,
the seat does work on it in amount 8 dq. The rate of doing work is (8 dq)/dt,
or 8i. Thus the left term in Eq. 36-16 is the rate at which the seat of emf
delivers energy to the circuit.
2. The second term in Eq. 36-16 is the rate at which energy appears as
Joule heat in the resistor.
3. Energy that does not appear as Joule heat must, by our hypothesis, be
stored in the magnetic field. Since Eq. 36-16 represents a statement of the
conservation of energy for LR circuits, the last term must represent the rate
dUB/dt at which energy is stored in the magnetic field, or
dUB di
- - = Li-· (36-17)
dt dt
We can write this as dUB = Li di.
Integrating yields
UB = ius ii
dUB = Li di = !Li2, (36-18)
Here the energy is stored in an electric field. In each case the expression for
the stored energy was derived by setting it equal to the work that must be
done to set up the field.
.... Example 3. A coil has an inductance of 5.0 henrys and a resistance of 20 ohms.
If a 100-volt emf is applied, what energy is stored in the magnetic field after the
current has built up to its maximum value 8/ R?
The maximum current is given by
. 8 100 volts
i = R= 20 ohms = 5 ·0 amp.
i = %(1 _ e-tlrL),
. (3.0 volts) 1)
i =
10 ohms (1 - e- = 0.189 amp.
At t = TL we have
di I
( 33.0h
volts ) _ 1 _
e - 0.37 amp sec.
dt .0 enrys
910 INDUCTANCE Chap. 36
From Eq. 36-17, the desired rate is
p _dUB_L_di
B - dt - 1, dt
= (3.0 henrys)(0.189 amp)(0.37 amp/sec)
= 0.210 watt.
Note that as required by the principle of conservation of energy (see Eq. 36-16)
Pi= PJ + PB,
or 0.567 watt = 0.357 watt + 0.210 watt
= 0.567 watt.
This equation gives the energy density stored at any point (in a vacuum or in
a nonmagnetic substance) where the magnetic induction is B. The equation
is true for all magnetic field configurations, even though it was derived by
considering a special case, the solenoid. Equation 36-19 is to be compared
with Eq. 30-27,
E2 ,
uE = 2I t:o,, (K = 1) (36-20)
which gives the energy density (in a vacuum) at any point in an electric
field. K ote that both UB and UE are proportional to the square of the appro-
priate field quantity, B or E.
The solenoid plays a role with relationship to magnetic fields similar to the
role the parallel-plate capacitor plays with respect to electric fields. In each
case we have a simple device that can be used for setting up a uniform field
ENERGY DENSITY AND THE MAGNETIC FIELD 911
f B·dl = µoi
or B = µoi.
21rr
Ampere's law shows further that the magnetic field is zero for points outside the
outer conductor (why?). Magnetic fields exist inBide each of the conductors but
we are concerned here only with energy stored between the conductors.
The energy density for points between the conductors, from Eq. 36-19, is
u = fdu = µoi2Z
41r
Ia
b dr
r
= µoi2Z ln ~'
41r a
QUESTIONS
1. Two coils are connected in series. Does their equivalent inductance depend on their
geometrical relationship to each other?
2. Is the inductance per unit length for a solenoid near its center (a) the same as, (b) less
than, or (c) greater than the inductance per unit length near its ends?
3. Two solenoids, A and B, have the same diameter and length and contain only one
layer of windings, with adjacent turns touching, insulation thickness being negligible.
Solenoid A contains many turns of fine wire and solenoid B contains fewer turns of heavier
wire. (a) Which solenoid has the larger inductance? (b) Which solenoid has the larger
inductive time constant?
4. If the flux passing through each turn of a coil is the same, the inductance of the coil
may be computed from L = N<PB/i (Eq. 36-4). How might one compute L for a coil for
which this assumption is not valid?
5. If a current in a source of emf is in the direction of the emf, the energy of the source
derreases; if a current is in a direction opposite to the emf (as in charging a battery), the
energy of the source increases. Do these statements apply to the inductor in Fig. 36-la
and 36-lb?
6. Show that the dimensions of the two expressions for L, N<P3/i (Eq. 36-4) and 8/
(di/dt) (Eq. 36-3b), are the same.
7. You are given N turns of wire connected in series. How should the turns be arranged
to obtain the maximum self-inductance?
Chap. 36 PROBLEMS 913
8. Does the time required for the current in a particular LR circuit to build up to any
given fraction of its equilibrium value depend on the value of the applied emf?
9. A steady current is set up in a coil with a very large inductive time constant. When
the current is interrupted with a switch, a heavy arc tends to appear at the switch blades.
Explain. (Note: Interrupting currents in highly inductive circuits can be dangerous.)
10. In an LR circuit like that of Fig. 36-4 can the self-induced emf ever be larger than
the battery emf?
11. In an LR circuit like that of Fig. 36-4 is the current in the resistance always the same
as the current in the inductance?
12. In the circuit of Fig. 36-3 the self-induced emf is a maximum at the instant the
switch is closed on a. How can this be since there is no current in the inductance at this
instant?
13. Give some arguments to show that energy can be stored in a magnetic field.
14. The switch in Fig. 36-3 is thrown from a to b. What happens to the energy stored
in the inductor?
15. In a toroid is the energy density larger near the inner radius or near the outer radius?
PROBLEMS
1. A IO-henry inductor carries a steady current of 2.0 amp. How can a 100-volt self-
induced emf be made to appear in the inductor?
,,%.'.Two inductances L 1 and L 2 are connected in series. and are separated by a large dis-
tance. (a) Show that the equivalent inductance Lis L1 + L2. (b) Why must their sep-
aration be large?
3. Show that if two inductors with equal inductance L are connected in parallel the
equivalent inductance of the combination 1s !L. The inductors are separated by a large
distance .
..4:/Two long parallel wires whose centers are a distance d apart carry equal currents in op-
posite directions. Show that, neglecting the flux within the wires themselves, the induc-
tance of a length l of such a pair of wires is given by
µold-a
L = -ln--,
7r a
where a is the wire radius. See Example 4, Chapter 34.
_,,,&:· A long thin solenoid can be bent into a ring to form a toroid. Show that if the
solenoid is long and thin enough the equation for the inductance of a toroid (see Example
1) reduces to that for a solenoid (Eq. 36-5).
6. A solenoid is wound with a single layer of #10 copper wire (diameter, 0.10 in.). It is
4.0 cm in diameter and 2.0 meters long. What is the inductance per unit length for the
solenoid near its center? Assume that adjacent wires touch and that insulation thickness
is negligible.
/7: The inductance of a close-packed coil of 400 turns is 8 mh. What is the magnetic
flux through the coil when the current is 5 X 10-3 amp?
8. A wooden toroidal core with'a square cross section has an inner radius of 10 cm and
an outer radius of 12 cm. It is wound with one layer of n8 wire (diameter, 0.040 in.;
resistance, 160 ft/ohm). What are (a) the inductance and (b) the inductive time
constant? Ignore the thickness of the insulation.
_.--9: The current in an LR circuit builds up to one--third of its steady-state value in 5.0
sec. What is the inductive time constant?
10. How many "time constants" must we wait for the current in an LR circuit to build
up to within 0.1 per cent of its equilibrium value?
11. The switch S in Fig. 36-3 is thrown from b to a. After one inductive time constant
show that (a) the total energy transformed to Joule heat in the resistor is 0.168S2rL/R
914 INDUCTANCE Chap. 36
and that (b) the energy stored in the magnetic field is 0.200S 2rL/R. (c) Show that the
equilibrium energy stored in the magnetic field is 0.500c 2rL/R.
12. Show that the inductive time constant TL can also be defined as tpe time that would
be required for the current in an LR circuit to reach its equilibrium value if it continued to
increase at its initial rate.
)-3: A 50-volt potential difference is suddenly applied to a coil with L = 50 mh and
R = 180 ohms. At what rate is the current increasing after 0.001 sec?
14. A coil with an inductance of 2.0 henrys and a resistance of 10 ohms is suddenly
connected to a resistanceless battery with S = 100 volts. At 0.1 sec after the connection
is made, what are the rates at which (a) energy is being stored in the magnetic field, (b)
Joule heat is appearing, and (c) energy is being delivered by the battery?
15. A coil with an inductance of 2.0 henrys and a resistance of 10 ohms is suddenly
connected to a resistanceless battery with 8 = 100 volts. (a) What is the equilibrium
current? (b) How much energy is stored in the magnetic field when this current exists in
the coil?
16. Prove that when switch Sin Fig. 36-3 is thrown from a to ball the energy stored in
the inductor appears as Joule heat in the resistor.
17. A circular loop of wire 5.0 cm in radius carries a current of 100 amp. What is the
energy density at the center of the loop?
18. What is the magnetic energy density at the center of a circulating electron in the
hydrogen atom (see Example 9, Chapter 34)?
~ A long wire carries a current i, uniformly distributed over a cross-section of the
wire. Show that the magnetic energy per unit length stored within the wire equals
µ 0i 2 /l61r. Note that it does not depend on the wire diameter.
20. Show that the self-inductance for a length l of a long wire associated with the flux
inside the wire only is µol/81r. Assume uniform distribution of the current, as in
Problem 19.
21. The coaxial cable of Example 5 has a = 1.0 mm, b = 4.0 mm, and c = 5.0 mm
(c is the radius of the outer surface of the outer conductor). It carries a current of 10 amp
in the inner conductor and an equal but oppositely directed return current in the outer
conductor. Calculate and compare the stored magnetic energy per meter of cable length
(a) within the central conductor, (b) in the space between the conductors, and (c) within
the outer conductor.
22. A length of #10 copper wire carries a current of 10 amp. Calculate (a) the magnetic
energy density and (b) the electric energy density at the surface of the wire. The wire
diameter is 0.10 in. and its resistance per unit length is 1.0 ohm/1000 ft.
-~·' What must be the strength of a uniform electric field if it is to have the same energy
density as that possessed by a 5000-gauss magnetic field?
Magnetic Properties
of Matter
CHAPTER 37
and nuclei that make up its atoms, it will be found that even these ele-
mentary particles are magnetic dipoles. Figure 37-3 contrasts the electric
and the magnetic characteristic of the free electron.
All electrons have a characteristic "spin" angular momentum about a
certain axis, which has the value of
Ls = 0.52723 X 10-34 joule-sec.
This is suggested by the vector Ls in Fig. 37-3b. Such a spinning charge can
be viewed classically as being made up of infinitesimal current loops. Each
such loop is a tiny magnetic dipole, its moment being given by (Eq. 33-10)
µ = NiA, (37-3)
where i is the equivalent current in each infinitesimal loop and A is the loop
area. The number of turns, N, is unity for each loop. The magnetic dipole
moment of the spinning charge can be found by integrating over the mo-
ments of the infinitesimal current loops that make it up; see Problem 2.
Although this model of the spinning electron is too mechanistic and is not
in accord with modern quantum physics, it remains true that the magnetic
dipole moments of elementary particles are closely connected with their
intrinsic angular momenta, or spins. Those particles and nuclei whose spin
angular momentum is zero (the a-particle, the pion, the 0 16 nucleus, etc.)
have no magnetic dipole moment. The "intrinsic" or "spin" magnetic mo-
(a) (b)
Fig. 37-3 (a) The lines of E and (b) the lines of B for an electron. The magnetic dipole
moment of the electron, µ.z, is directed opposite to the spin angular momentum vector, L 8 •
v = 2~~i 2~/P·
=
With this equationµ can be found from the measured quantities v, B, and I.
918 MAGNETIC PROPERTIES OF MATTER Chap. 37
Example 2. An electron in an atom circulating in an assumed circular orbit of
radius r behaves like a tiny current loop and has an orbital magnetic dipole moment *
usually represented by µz. Derive a connection between µ 1 and the orbital angular
momentum Lz.
Newton's second law (F = ma) yields, if we substitute Coulomb's law for F,
1 e2 mv 2
- - - =ma= - -
41T'E:o r 2 r
or v= (37-4)
The angular velocity w is given by
The current for the orbit is the rate at which charge passes any given point, or
The orbital dipole moment µ 1 is given from Eq. 37-3 if we put N = 1 and A = 1T'r 2 , or
(37-5)
Lz = (mv)r.
/J,z = Lz ( 2;)·
which shows that the orbital magnetic moment of an electron is proportional to its
orbital angular momentum.
For r = 5.1 X 10- 11 meter, which corresponds to hydrogen in its normal state,
we have from Eq. 37-5
e2 r-;;:-
µz = 4 ~;;m
= (1.6 X 10- 19 coul) 2 / 5.1 X 10- 11 meter
4 \J (11")(8.9 X 10- 12 coul2/nt-m 2)(9.1 X 10- 31 kg)
= 9.1 X 10-24 amp-m 2 • _,.
* This must not be confused with the magnetic dipole moment µs of the electron spin,
which is also present.
I
/
I
\
"
-"°"°
Fig. 37-4 Lines of B (a) for a bar magnet and (b) for a short solenoid. (c) Lines of E for an electric dipole. At large enough distances all
three fields vary like those for a dipole. The four dashed curves are interse<>tions with the plane of the figure of closed Gaussian surfaces.
Note that <f>B equals zero for (a) or (b). <f>E equals zero for surfaces like II in (c), which do not contain any charge, but <f>E is not zero for sur-
faces like I.
920 MAGNETIC PROPERTIES OF MATTER Chap. 37
37-2 Gauss's Law for Magnetism
Gauss's law for magnetism, which is one of the basic equations of electro-
magnetism (see Table 38-3), is a formal way of stating a conclusion that
seems to be forced on one by the facts of magnetism, namely, that isolated
magnetic poles do not exist. This equation asserts that the flux 'PB through
any closed Gaussian surface must be zero, or
where the integral is to be taken over the entire closed surface. We con-
trast this with Gauss's law for electricity, which is
The fact that a zero appears at the right of Eq. 37-6, but not at the right of
Eq. 37-7, means that in magnetism there is no counterpart to the free charge
q in electricity.
Figure 37-4a shows a Gaussian surface that encloses one end of a bar
magnet. Note that the lines of B enter the surface for the most part inside
the magnet and leave it for the most part outside the magnet. There is
thus an inward (or negative) flux inside the magnet and an outward (or
positive) flux outside it. The total flux for the whole surface is zero. In
Section 37-7 the interested student will find a fuller discussion of why the
lines of B for a bar magnet are drawn in the way shown.
Figure 37-4b shows a similar surface for a solenoid of finite length which,
like a bar magnet, is also a magnetic dipole. Here, too, 'PB equals zero.
Figures 37-4a and b show clearly that there are no "sources" of B; that is,
there are no points from which lines of B emanate. Also, there are no "sinks"
of B; that is, there are no points toward which B converges. In other words,
there are no free magnetic poles.
Figure 37-4c shows a Gaussian surface (I) surrounding the positive end of
an electric dipole. Here there is a net flux of the lines of E. There is a
"source" of E; it is the charge q. If q is negative, we have a "sink" of E
because the lines of E end on negative charges. For surfaces like surface II
in Fig. 37-4c for which the charge inside is zero, the flux of E over the surface
is also zero.
37-3 Paramagnetism
Magnetism as we know it in our daily experience is an important but
special branch of the subject called ferromagnetism; we discuss this in Sec-
tion 37-5. Here we discuss a weaker form of magnetism called paramag-
netism.
For most atoms and ions, the magnetic effects of the electrons, including
both their spins and orbital motions, exactly cancel so that the atom or ion
Sec. 37-3 PARAMAGNETISM 921
is not magnetic. This is true for the rare gases such as neon and for ions *
such as cu+, which makes up ordinary copper. These materials do not ex-
hibit paramagnetism. For other atoms or ions the magnetic effects of the
electrons do not cancel, so that the atom as a whole has a magnetic dipole
moment µ. Examples are found among the so-called transition elements,
such as Mn++, the rare earths, such as Gd+++, and the actinide elements,
such as u++++.
If a sample of N atoms, each of which has a magnetic dipole moment µ,
is placed in a magnetic field, the elementary atomic dipoles tend to line up
with the field. For perfect alignment, the sample as a whole would have a
magnetic dipole moment of Nµ. However, the aligning process is seriously
interfered with by the collisions that take place between the atoms if the
sample is a gas and by temperature vibrations if the sample is a solid. The
importance of this thermal agitation effect may be measured by comparing
two energies: one ( = JkT) is the mean kinetic energy of translation of a
gas atom at temperature T; the other ( = 2µ,B) is the difference in energy
between an atom lined up with the magnetic field and one pointing in the
opposite direction. As Example 3 shows, the effect of the collisions at
ordinary temperatures and fields is very great. The sample acquires a mag-
netic moment when placed in an external magnetic field, but this moment
is usually much smaller than the maximum possible moment N µ.
~ Example 3. A paramagnetic gas, whose atoms (see Example 2) have a magnetic
dipole moment of about 10-23 amp-m 2 , is placed in an external magnetic field of
magnitude 1 weber/meter 2 • At room temperature (T = 300°K) calculate and com-
pare Ur, the mean kinetic energy of translation ( = }kT), and UB, the magnetic
energy ( = 2µ,B) :
Ur = fkT = (t )(1.38 X 10-23 joule/°K) (300°K) = 6 X 10-21 joule,
UB = 2µ,B = (2)(10-23 amp-m 2 )(1 weber/meter 2 ) = 2 X 10-23 joule.
Because Ur equals 300 UB, we see that energy exchanges in collisions can interfere
seriously with the alignment of the dipoles with the external field. ....
If a specimen of a paramagnetic substance is placed in a nonuniform magnetic
field, such as that near the pole of a strong magnet, it will be attracted toward the
region of higher field, that is, toward the pole. We can understand this by drawing
an analogy with the corresponding electric case of Fig. 37-5, which shows a dielectric
specimen (a sphere) in a nonuniform electric field. The net electric force points to
the right in the figure and is
Fe = q(E + fiE) - q(E - fiE) = q(2fiE),
which can be written as
F dB)
- µ ( -- (37-8)
m - dx rnax
Thus, by measuring the magnetic force Fm that acts on a small paramagnetic speci-
men when it is placed in a nonuniform magnetic field whose field gradient (dB/dx)max
is known, we can learn its magnetic dipole moment µ. The magnetization M is
defined as the magnetic moment per unit volume, or
M=~,
v
where V is the volume of the specimen. It is a vector because µ., the dipole moment
of the specimen, is a vector.
In 1895 Pierre Curie (1859-1906) discovered experimentally that the magnetization
M of a paramagnetic specimen is directly proportional to B, the effective value of
magnetic induction in which the specimen is placed, and inversely proportional to
the temperature, or
B
M = C-, (37-9)
T
in which C is a constant. This equation is known as Curie's law. The law is physi-
cally reasonable in that increasing B tends to align the elementary dipoles in the speci-
men, that is, to increase M, whereas increasing T tends to interfere with this align-
ment, that is, to decrease M. Curie's law is well verified experimentally, provided
that the ratio B/T does not become too large.
E = E 0 - AE E = E 0 + AE
Fig. 37-5 A dielectric sphere in a nonuniform electric field.
DIAMAGNETISM 923
0.75
x
• 1.30° K
.. 2.00°K
"'
E
x 3.00°K
~ 0.50
• 4.21° K
~ -Theory
0.25
10 20 30 40
B/T, 10 3 gauss/degree
Fig. 37-6 The ratio M /M max for a paramagnetic salt (chromium potassium alum) in
various magnetic fields and at various temperatures. The curve through the experimental
points is a theoretical curve calculated ±rom modern quantum physics. (From measure-
ments by W. E. Henry.)
M cannot increase without limit, as Curie's law implies, but must approach a
value M max ( = µN /V) corresponding to the complete alignment of the N dipoles
contained in the volume V of the specimen. Figure 37-6 shows this saturation
effect for a sample of CrK(S04)2 · 12H20. The chromium ions are responsible for all
the paramagnetism of this salt, all the other elements being paramagnetically inert.
To achieve 99.5% saturation, it is necessary to use applied magnetic fields as high
as 50,000 gauss and temperatures as low as l.3°K. Note that for more readily
achievable conditions, such as B = 10,000 gauss and T = 10°K, the abscissa in
Fig. 37-6 is only 1.0 so that Curie's law would appear to be well obeyed for this
and for all lower values of B/T. The curve that passes through the experimental
points in this figure is calculated from a theory based on modern quantum physics;
it is in excellent agreement with experiment.
37-4 Diamagnetism
In 1846 Michael Faraday discovered that a specimen of bismuth brought near to
the pole of a strong magnet is repelled. He called such substances diamagnetic.
Diamagnetism, present in all substances, is such a feeble effect that its presence is
masked in substances made of atoms that have a net magnetic dipole moment,
that is, in paramagnetic or ferromagnetic substances.
Figures 37-7a and b show an electron circulating in a diamagnetic atom at angular
frequency w0 in an assumed circular orbit of radius r. Each electron is moving under
the action of a centripetal force FE of electrostatic origin where, from Newton's
second law,
FE = ma = mwo 2r. (37-10)
Each rotating electron has an orbital magnetic moment, but for the atom as a whole
the orbits are randomly oriented so that there is no net magnetic effect. In Fig.
37-7a, for example, the magnetic dipole moment l,Lz points into the page; in Fig.
924 MAGNETIC PROPERTIES OF MATTER Chap. 37
-e -e
FE FE
w~
B=O
~
(a) (b)
x x x FB x x x x x x x x x x
V2
x x x x x x x x x x
FB
x x x x x x x x x x
FE FE
x x x x x x x x x x x x x x
x x x x x x x x x x x x x x
x x x x x·-
~x
1 = w0 - Aw)
x x x x x
Ax 2 (= w0 + Aw)
x x x x x x x~x x x x x x x
B
(c) (d)
Fig. 37-7 (a) An electron circulating in an atom. (b) An electron circulating in the op-
posite direction. (c) A magnetic field is introduced, decreasing the linear speed of the
electron in (a), that is, v1 < vo. (d) The magnetic field increases the linear speed of the
electron in (b ), that is, v2 > vo.
37-7b it points out and the net effect for the two orbits sh01vn is cancellation. This
cancellation is also shown at the left in Fig. 37-8.
If an external field B is applied as in Fig. 37-7c and d, an additional force, given
by -e(v x B), acts on the electron. This magnetic force acts always at right angles
to the direction of motion; its magnitude is
FB = evB = e(wr)B. (37-11)
The student should show that in Fig. 37-7c FB and FE point in opposite directions
and that in Fig. 37-7d they point in the same direction. Note that since the centrip-
etal force changes when the magnetic field is applied (the radius can be shown
to remain constant), the angular velocity must also change; thus w in Eq. 37-11
differs from wo in Eq. 37-10.
Applying Newton's second law to Figs. 37-7c and d, and allowing for both direc-
tions of circulation, yields for the resultant forces on the electrons
FE± FB = ma = mw 2r.
Sec. 37-4 DIAMAGNETISM 925
Substituting Eqs. 37-10 and 37-11 into this equation yields
or w2 =F ( ~ ) w - wo 2 = 0. (37-12)
This quadratic equation can be solved for w, the new angular velocity. Rather
than doing this, we take advantage of the fact (presented without proof; see Problem
7) that w differs only slightly from wo, even in the strongest external magnetic fields.
Thus
w = wo 6.w + (37-13)
where 6.w <<wo. Substituting this equation into Eq. 37-12 yields
[wo2 + 2wo 6.w + (6.w)2] ± [{Jwo + {36.w] - wo 2 = 0,
where /3 is a convenient abbreviation for eB/m. The two terms w02 cancel each other;
the terms (6.w) 2 and {JAw are small compared to the remaining terms and may be set
equal to zero with only small error. This leads, as an excellent approximation, to
,,..__., 1 _ eB
6.w = =F 2/J - =F 2m (37-14)
eB
or, from Eq. 37-13, W = WO =f= --·
2m
Thus the effect of applying a magnetic field is to increase or decrease (depending on
the direction of circulation) the angular velocity. This, in turn, increases or decreases
the orbital magnetic moment of the circulating electron (see Example 2).
In Fig. 37-7c the angular velocity is reduced (because the centripetal force is
reduced) so that the magnitude of the magnetic moment is reduced. In Fig. 37-7d,
however, the angular velocity is increased so that the magnitude of µ 1 is increased.
These effects are shown on the right in Fig. 37-8, where it will be noted that the two
magnetic moments no longer cancel.
We see that if a magnetic field B is applied to a diamagnetic substance (zero net
magnetic moment in absence of applied field), a magnetic moment will be induced
whose direction (out of the plane of Fig. 37-7) is opposite to B; see also Fig. 37-8.
This is precisely the reverse of paramagnetism, in which the (permanent) magnetic
dipoles tend to point in the same direction as the applied field.
We can now understand why a diamagnetic specimen is repelled when brought
near to the pole of a strong magnet. If the pole is a north pole, there exists a non-
uniform magnetic field of induction with B pointing away from the pole. If a sphere
B=O
(a) (b)
926 MAGNETIC PROPERTIES OF MATTER Chap. 37
made of a diamagnetic material (bismuth, say) is brought near to this pole, the mag-
netization M that is induced in it points toward the pole, that is, opposite to B.
Thus the side of the sphere closest to the magnet behaves like a north pole and is
repelled by the nearby north pole of the magnet. For a paramagnetic sphere, the
vector M points along the direction of B and the side of the sphere nearest to the
magnet is a south pole, which is attracted to the north pole of the magnet.
~ Example 4. Calculate the change in magnetic moment for a circulating electron,
as described in Example 2, if a magnetic field of induction B of 2.0 webers/meter2
acts at right angles to the plane of the orbit.
We obtainµ from Eq. 37-3, or
37-5 Ferromagnetism
For five elements (Fe, Co, Ni, Gd, and Dy) and for a variety of alloys of
these and other elements a special effect occurs which permits a specimen
to achieve a high degree of magnetic alignment in spite of the randomizing
tendency of the thermal motions of the atoms. In such materials, described
as ferromagnetic, a special form of interaction called exchange coupling occurs
between adjacent atoms, coupling their magnetic moments together in rigid
parallelism.* Modern quantum physics successfully predicts that this will
occur only for the five elements listed. If the temperature is raised above a
certain critical value, called the Curie temperature, the exchange coupling
suddenly disappears and the materials become simply paramagnetic. For
iron the Curie temperature is 1043°K. Ferromagnetism is evidently a
property not only of the individual atom or ion but also of the interaction
of each atom or ion with its neighbors in the crystal lattice (see Fig. 21-5)
of the solid.
Figure 37-9 shows a magnetization curve for a specimen of iron. To obtain
such a curve, we form the specimen, assumed initially unmagnetized, into a
ring and wind a toroidal coil around it as in Fig. 37-10, to form a so-called
Rowland ring. When a current i is set up in the coil, if the iron core is not
present, a field of induction is set up within the toroid given by (Eq. 34-4)
(37-15)
* Exchange coupling, a purely quantum effect, cannot be "explained" in terms of
classical physics.
Sec. 37-5 FERRO MAGNETISM 927
1.0
0.8
)(
"'
E
:i 0.6
Fig. 37-9 A magnetization curve
for iron.
ES.
~ 0.4
0.2
2 4 6 8 10 12 14
B 0 , 10- 4 weber/meter2
where_ n is the number of turns per unit length for the toroid. Although
this formula was derived for a long solenoid, it can be applied to a toroid
if d « r in Fig. 37-10. Because of the iron core, the actual value of Bin the
toroidal space will exceed B 0 , by a large factor in many cases, since the
elementary atomic dipoles in the core line up with the applied field B 0 ,
thereby setting up their own field of induction. Thus we can write
(37-16)
where BM is the magnetic induction due to the specimen; it is proportional
to the magnetization M of the specimen. Often BM» B 0 .
The field B 0 is proportional to the current in the toroid and can be cal-
culated readily, using Eq. 37-15; B can be measured in a way that is de-
scribed below. An experimental value for BM can be derived from Eq.
d
928 MAGNETIC PROPERTIES OF MATTER Chap. 37
B = (i., At)R = qR ,
or
NA NA
in which R is the resistance of coil Sand is At is the charge q that passes through this
coil during time At. If a so-called ballistic galvanometer is connected to S, its deflection
will be a measure of the charge q. Thus it is possible to find B for any value of the
current i in the toroid windings. In practice, it is not necessary that the curve B(t)
in Fig. 37-lla be linear during the interval At.
* We ignore the minus sign because we are concerned only with the magnitude of 8.
Sec. 37-5 FERRO MAGNETISM 929
b
Fig. 37-12 A magnetization curve (ab) for a speci-
men of iron and an associated hysteresis loop
(ebcde).
t---1
0.01 mm
(a) (b) (c)
Fig. 37-15 (a) A boundary between two domains, with the magnetization in each domain
as shown by the white arrows. (b) If an external magnetic field pointing from left to right
is imposed on the specimen, the upper domain will grow at the expense of the lower. The
domain boundary will move down as the elementary dipoles reverse themselves. (c) The
process continues. The boundary has moved across a region in which there is a crystal
imperfection. (Courtesy H. J. Williams, Bell Telephone Laboratories.)
Sec. 37-6 NUCLEAR MAGNETISM 931
·••·• ·•·•·•· ···•· ······
••i
•••••
(a)
..•• ,. :: · : 11 •11·····
netism.
······ ······ "''
(c) . ······
······
.................. ······ ······
................. .... ······
1i1ii:!i!il
iii!!!
····· ······ ·····
•••••• ••••• •••••
............
.... ······
' ······ ..., .... ······
······ ·•···
•·•·•·•·• ·•·•·•·•·•·•••·•
.·•·.·.·..·.... ·..······· ······
····· ·•·······•·····•·•·•··· ···········
which Mn0 2 is an example, the exchange coupling to which we referred on page 834
serves to lock adjacent ions into rigid antiparallelism (see Fig. 37-16b). Such ma-
terials exhibit very little gross external magnetism. However, if they are heated
above a certain temperature, called the Neel temperature, the exchange coupling
ceases to act and the material becomes paramagnetic. In ferrimagnetic substances,
of which iron ferrite is an example, two different kinds of magnetic ions are present.
In iron ferrite the two ions are Fe++ and Fe+++. The exchange coupling locks the
ions into a pattern like that of Fig. 37-16c, in which the external effects are inter-
mediate between ferromagnetism and antiferromagnetism. Here, too, the exchange
coupling disappears if the material is heated above a certain characteristic tempera-
ture.
I I
(a) (b)
We focus our attention on the problem of measuring the magnitude µ of the mag-
netic moment of the proton. In principle, this can be done by placing a specimen
containing protons in an external field of magnetic induction B and by measuring
the energy ( = 2µB) required to turn the protons end for end. A rigorously correct
description of the procedures cannot be given without using quantum physics. The
description given, although based entirely on classical physics, nevertheless leads to
the correct conclusions.
Figure 37-17a shows a spinning proton with its axis making an angle() with a uni-
form external magnetic field B. Figure 37-17b shows a spinning top with its axis
making an angle() with a uniform external gravitational field g. In each case there is
a torque that tends to align the axis of the spinning object with the field. For the
proton (Eq. 33-11) it is given by
Tp = µB sine. (37-17a)
For the top it is given by Tt = mgr sin(), (37-17b)
where r locates the center of mass of the top and m is its mass.
In Section 13-2, we saw that the spinning top processes about a vertical axis with
an angular frequency given by
mgr
Wt=-,
Lt
in which Lt is the spin angular momentum of the top.
The proton, which has a quantized spin angular momentum Lp, will also precess
about the direction of the (magnetic) field because of the action of the (magnetic)
torque. The student should derive the expression for the frequency of precession,
being guided by the derivation of Section 13-2, but using the magnetic torque
(Eq. 37-17a) instead of the gravitational torque (Eq. 37-17b). The relation is
µB
w = --· (37-18b)
P Lp
I\
I ta
I \
Fig. 37-18 In the nuclear magnetic resonance method a small oscil- I \
lating magnetic field Bose is placed at right angles to a steady field B. I \
I B \
I \
can be used to measureµ. We place the spinning proton in a known field B, apply a
"perturbing field" at right angles to it, and vary the angular frequency w0 of this
perturbing field until resonance occurs. It is possible to tell when Eq. 37-19 is
satisfied because, at resonance, many spinning protons will tend to turn end for end
in the field, absorbing energy which can be detected by appropriate electronic tech-
niques.
Figure 37-19 is a schematic diagram of an experimental arrangement. The protons,
present as hydrogen nuclei in a small vial V of water, are immersed in a strong steady
magnetic field caused by the electromagnet whose pole faces N and S are shown. A
rapidly alternating current in the small coil C provides the (horizontal) weak, per-
turbing magnetic field Bose· This current is provided by a radio-frequency oscillator
Resonance
Oscillator detector
Frequency Oscilloscope
control knob
\
\
dl
1-+-~~-~ I
I dµ,
(b)
936 MAGNETIC PROPERTIES OF MATTER Chap. 37
is not valid when magnetic materials are present. Equation 37-21 predicts that, since
the right side is the same whether or not the core is in place, the induction B should
also be the same, a prediction not in accord with experiment.
We can increase B in the absence of the iron core to the value that it has ,vhen the
core is in place if we increase the current in the windings by an amount i 111 .0 • The
magnetization of the iron core is thus equivalent in its ejfect on B to such a hypothetical
current increase. We choose to modify Ampere's law by arbitrarily inserting a
magnetizing current term i 111 on the right, obtaining
If we give i111 a suitable value when the iron core is in place, it is clear that Ampere's
law, in this new form, can remain valid. It remains to relate this (largely hypo-
thetical) magnetizing current to something more physical, the magnetization M. *
Applying Eq. 37-22 to the iron ring of Fig. 37-2la yields
We can relate i111,o to the magnetization M if we recall (Eq. 33-10) that the magnetic
moment of a magnetic dipole in the form of a current loop is given by
µ = NiA,
where N is the number of turns in the loop, i is the loop current, and A is the loop
area. Let us use this equation to find what increase i111.o in current in the windings
around the slice of Fig. 37-2lb would produce a magnetic moment equivalent to that
actually produced by the alignment of elementary dipoles in the slice. We have
the quantity in the first parentheses on the right being the number of turns associ-
ated with the slice of thickness dl. This reduces to
(37-24)
We now choose to generalize from the special case of the Rowland ring by writing
Eq. 37-25 as
or
H = B ·- µoM
µo
fH·dI = i, (37-27)
which holds in the presence of magnetic materials and in whi1;n i is the true current only,
that is, it does not include the magnetizing current. This reminds us that the
electric displacement vector D permitted us to write Gauss's law for the case in which
dielectric materials are present, in a form involving free charges only, that is, not
polarization charges; see Table 30-2.
We state without proof (see Problems 12 and 13) that at a boundary between two
media (1) the component of H tangential to the surface has the same value on each
side of the surface * and (2) the component of B perpendicular to the surface has the
same value on each side of the surface. These boundary conditions are of great value
in solving complex problems.
To find Hin our Rowland ring, let us apply Ampere's law in the generalized form
of Eq. 37-27. We have
(H)(21rro) = N oio,
H = 2
No).
( 1rro io =
.
mo, (37-28)
in which n is the number of turns per unit length. Since we have not introduced
any information describing the core into Eq. 37-27, the value of H computed from
Eq. 37-28 is independent of the core material.
B can be measured experimentally by the method of Section 37-5 and M can then
be calculated from Eq. 37-26. The student should note in passing (see Eq. 37-15)
that the abscissa Bo in Fig. 37-9 is proportional to H ( = µ 0H), the ordinate being
proportional to B. Curves such as this and that of Fig. 37-12 are called B-H curves.
Let us assume that we have made measurements of H, B, and M for a wide variety
of magnetic materials, using either the technique described or an equivalent one.
For paramagnetic and diamagnetic materials we would find, as an experimental result,
that B is directly proportional to H, or
(37-29)
in which Km, the permeability of the magnetic medium, is a constant for a given tem-
perature and density of the material. Eliminating B between Eqs. 37-29 and 37-26
allows us to write
M = (Km - l)H, (37-30)
* Assuming; that there are no true current~ at the surface, as there are in the Rowland
ring of Fig. 37-21a, for example.
938 MAGNETIC PROPERTIES OF MATTER Chap. 37
(c) ~r: r~
B .uoH
M =O
(a) (b)
Fig. 37-22 (a) The lines of H and (b) the lines of B for a permanent magnet.
Note that the lines of H change direction at the boundary. The closed dashed
curves are paths of integration around which Ampere's law may be applied. The
relation B = µoH + µoM is shown to be satisfied for (c) a particular outside point p
and (d) a particular inside point q.
For a vacuum, in which there are no magnetic dipoles present to be aligned, the
magnetization M must be zero. Putting M = 0 in Eq. 37-26 leads to
The reversal of H at the boundary makes this possible. Note that M and H point
in opposite directions within the magnet. Table 37-1 summarizes the properties of
the three vectors B, H, and M.
Table 37-1
THREE MAGNETIC VECTORS
General relation among the three vectors B = µoH + µoM Eq. 37-26
.. Example 6. In the Rowland ring the (true) current io in the windings is 2.0 amp
and the number of turns per unit length (n) in the toroid is 10 turns/cm. B, measured
by the technique of Section 37-5, is 1.0 weber/meter 2 • Calculate (a) H, (b) M, and
(c) the magnetizing current i 111 ,0 both when the core is in place and when it is removed.
(d) For these particular operating conditions, what is Km?
(a) His independent of the core material and may be found from Eq. 37-28:
H = ni
= (10 turns/cm)(2.0 amp)
= 2.0 X 10 3 amp/meter.
940 MAGNETIC PROPERTIES OF MATTER Chap. 37
(b) M is zero when the core is removed. With the core in place, we may solve
Eq. 37-26 for M, obtaining for the magnitude of M,
M = B - µoH
µo
(1.0 weber/meter 2) - (41r X 10-7 weber/amp-m)(2.0 X 10 3 amp/meter)
(41r X 10-7 weber/amp-m)
= 7.9 X 10 5 amp/meter.
(c) The effective magnetizing current follows from Eq. 37-24:
7.9 X 10 5 amp/meter
1.0 X 103 turns/meter
= 790 amp.
An additional current of this amount in the windings would produce the same value
of B in the absence of a core as that obtained, by alignment of the elementary di-
poles, with the core in place.
(d) The permeability can be found from Eq. 37-29, or
B
Km= µoH
1.0 weber/meter 2
(41r X 10-7 weber/amp-m)(2.0 X 103 amp/meter)
= 397.
We emphasize that this value of Km holds only for the special conditions of this
experiment. ...ii
QUESTIONS
1. Two iron bars are identical in Rppearance. One is a magnet and one is not. How can
you tell them apart? You are not permitted to suspend either bar as a compass needle or
to use any other apparatus.
2. How could you reverse the magnetism of a compass needle?
3. Two iron bars always attract, no matter the combination in which their ends are
brought near each other. Can you conclude that one of the bars must be unmagnetized?
4. If we sprinkle iron filings on a particular bar magnet, they cling both to the ends and
to the middle. Sketch roughly the lines of B, both outside and inside the magnet.
5. The earth is a huge magnetic dipole. (a) Is the magnetic pole in the Northern
Hemisphere a north or a south magnetic pole? (b) ln the Northern Hemisphere do the
magnetic lines of force associated with the earth's magnetic field point toward the earth's
surface or away from it?
6. Cosmic rays are charged particles that strike our atmosphere from some external
source. We find that more low-energy cosmic rays reach the earth at the north and south
magnetic poles than at the (magnetic) equator. Why is this so?
7. How might the magnetic dipole moment of the earth be measured?
8. Give three reasons for believing that the flux <f>B of the earth's magnetic field is
greater through the boundaries of Alaska than through those of Texas.
Chap. 37 PROBLEMS 941
9. The neutron, which has no charge, has a magnetic dipole moment. Is this possible
on the basis of classical electromagnetism, or does this evidence alone indicate that classical
electromagnetism has broken down?
10. Is the magnetirntion at saturation for a paramagnetic substance very much different
from that for a saturated ferromagnetic substance of about the same size?
11. Explain why a magnet attracts an unmagnetized iron object such as a nail.
12. Does any net force or torque act on (a) an unmagnetized iron bar or (b) a permanent
bar magnet when placed in a uniform magnetic field?
13. A nail is placed at rest on a smooth table top near a strong magnet. It is released
and attracted to the magnet. What is the source of the kinetic energy it has just before
it strikes the magnet?
14. The magnetization induced in a given diamagnetic sphere by a given external
magnetic field does not vary with temperature, in sharp contrast to the situation in para-
magnetism. Is this understandable in terms of the description that we have given of the
origin of diamagnetism?
15. Compare the magnetirntion curves for a paramagnetic substance (Fig. 37-6) and
for a ferromagnetic substance (Fig. 37-9). What would a similar curve for a diamagnetic
substance look like? Do you think that it would show saturation effects in strong applied
fields (say 10 weber /meter2)?
16. Distinguish between the precession frequency and the cyclotron frequency of a
proton in a magnetic field.
17. In our discussion of nuclear magnetism we said that energy absorption occurs
because the dipoles are turned end for end. However, a given dipole might initially be
lined up either with the field or against it. In the first case there would be an absorption
of energy, but in the second case there would be a release of energy, each amount being 2µB.
Why do we observe a net absorption? These two events would seem to cancel.
18 . .Discuss similarities and differences in Tables 30-2 and 37-1.
PROBLEMS
__-Y.'The earth has a magnetic dipole moment of 6.4 X 1021 amp-m 2 • (a) What current
would have to be set up in a single turn of wire going around the earth at its magnetic
equator if we wished to set up such a dipole? (b) Could such an arrangement be used to
cancel out the earth's magnetism at points in space well above the earth's surface? (r)
On the earth's surface?
,:l: · · Assume that the electron is a small sphere of radius R, its charge and mass being
spread uniformly throughout its volume. Such an electron has a "spin" angular mo-
mentum L of 0.53 X 10-34 joule-sec and a magnetic moment µ of 9.3 X 10-24 amp-m 2 •
Show that e/m = 2µ/L. Is this prediction in agreement with experiment? (Hint: The
spherical electron must be divided into infinitesimal current loops and an expression for
the magnetic moment found by integration. This model of the electron is too mechan-
istic to be in the spirit of quantum physics.)
3. Calculate (a) the electric field strength and (b) the magnetic induction at a point
1.0 A (one angstrom unit) away from a proton, measured along its axis of spin. The mag-
netic moment of the proton is 1.4 X 10-25 amp-m2 .
4. A Rowland ring is formed of ferromagnetic material. It is circular in cross section,
with an inner radius of 5.0 cm and an outer radius of 6.0 cm and is wound with 400 turns
of wire. (a) What current must be set up in the windings to attain Bo = 2 X 10-4 weber/
meter2 in Fig. 37-9? (b) A secondary coil wound around the toroid has 50 turns and has
a resistance of 8.0 ohms. If, for this value of Bo, we have BM = 800B 0 , how much charge
moves through the secondary coil when the current in the toroid windings is turned on?
5. The dipole moment associated with an atom of iron in an iron bar is 1.8 X 10-23
amp-m 2 . Assume that all the atoms in the bar, which is 5 cm long and has a cross-sec-
942 MAGNETIC PROPERTIES OF MA'l'TER Chap. 37
tional area of 1 cm 2 , have their dipole moments aligned. (a) What is the dipole moment
of the bar? (b) What torque must be exerted to hold this magnet at right angles to an
external field of 15,000 gauss? The density of iron is 7.9 gm/cm 3 •
6. Can you give an explanation of diamagnetism based tin Faraday's law of induction?
In Figs. 37-7a and b, for example, what inductive effects can be expected as the magnetic
field is built up from zero to the final value B?
7. Prove that ~w «wo in Eq. 37-13.
8. Show that, classically, a spinning positive charge will have a spin magnetic moment
that points in the same direction as its spin angular momentum.
-- 9: Assume that the nuclei (protons) in 1 gm of water could all be aligned. What mag-
netic induction B would be produced 5.0 cm from the sample, along its alignment axis?
10. It is possible to measure e/m for the electron by measuring (a) the cyclotron fre-
quency 11e of electrons in a given magnetic field and (b) the precession frequency 11p of
protons in the same field. Show that the relation is
e lie µ.,
m 11pLa
Since µ. 8 and L. for the proton are accurately known, this experiment gives us our most
precise value of e/m today.
11. Dipole-dipole interactiori. The exchange coupling mentioned in Section 37-5 as
being responsible for ferromagnetism is not the mutual magnetic interaction energy be-
tween two elementary magnetic dipoles. To show this (a) compute B a distance a ( = 1.0
A) away from a dipole of moment µ. ( = 1.8 X 10-23 amp-m 2); (b) compute the energy
( = 2µ.B) required to turn a second similar dipole end for end in this field. What do you
conclude about the strength of this dipole-dipole interaction? Compare with the results
of Example 3. (Note: for the same distance, the field in the median plane of a dipole is
only half as large as on the axis; see Eq. 37-2.)
'"'12', .Boundary condition for B. Prove that at the boundary between two media the
normal component of B has the same value on each side of the surface. (Hint: Construct
a closed Gaussian surface shaped like a flat pillbox with one face in each medium and apply
Gauss's law for magnetism.)
13, Boundary condition for H. Prove that at the boundary between two media the
tangential component of H has the same value on each side of the surface, assuming that
there is no current at the surface. (Hint: Construct a closed rectangular loop, the two
opposite longer sides being parallel to the surface, with one side in each medium. Use
Ampere's law in the form that applies when magnetic materials are present.)
14. The magnetic energy density can be shown to be given in its most general form as
µ.B = !B·H.
Does this reduce to a familiar result for a vacuum?
Electromagnetic Oscillations
CHAPTER 38
38-1 LC Oscillations
The LC system of Fig. 38-1 resembles a mass-spring system (see Fig. 8-4)
in that, among other things, each system has a characteristic frequency of
oscillation. To see this, we assume that initially the capacitor C in Fig.
38-la carries a charge qm and the current i in the inductor is zero. At this
instant the energy stored in the capacitor is given by Eq. 30-25, or
1 qm2
UE = --· (38-1)
2 C
The energy stored in the inductor, given by
UB -_ 21£;2
•, (38-2)
is zero because the current is zero. The capacitor now starts to discharge
through the inductor, positive charge carriers moving counterclockwise, as
shown in Fig. 38-lb. This means that a current i, given by dq/dt and point-
ing down in the inductor, is established.
As q decreases, the energy stored in the electric field in the capacitor also
decreases. This energy is transferred to the magnetic field that appears
around the inductor because of the current i that is building up there. Thus
the electric field decreases, the magnetic field builds up, and energy is trans-
ferred from the former to the latter.
At a time corresponding to Fig. 38-lc, all the charge on the capacitor will
have disappeared. The electric field in the capacitor will be zero, the energy
stored there having been transferred entirely to the magnetic field of the
943
944 ELECTRO MAG NET IC OSCILLATIONS Chap. 38
(b) (d)
I
•• ••
UB
\
Orr,.)
_I a- I
.._______...... (f)
I_
UB UE
Fig. 38-1 Showing eight stages in a cycle of oscillation of an LC circuit. The bar graphs
below each figure show the stored magnetic and electric potential energy. The vertical
arrows on the induewr axis show the current. The student should compare this figure in
detail with Fig. 8-4, to which it exactly corresponds.
-
.-.
Co)
<::,,
(a)
II Time
'""
~
I
ace g ace gab
~I I I I I I I I I
(b) -~-~ ~ )
~ ~~Time
where im is the maximum current and qm is the maximum charge. Note that the
maximum current and the maximum charge do not occur at the same time but one-
fourth of a cycle apart; see Figs. 38-1 and 38-2. Solving for im and substituting
CVo for qm gives
Table 38-1
Mechanical Electromagnetic
1 q2
sprmg capacitor UE = --
2C
mass inductor
v = dx/dt i = dq/dt
some formal way like a spring and an inductor is like a mass and that certain
electromagnetic quantities "correspond" to certain mechanical ones, namely,
q corresponds to x,
i corresponds to v,
C corresponds to 1/k,
L corresponds to m.
Comparison of Fig. 38-1, which shows the oscillations of the LC circuit,
with Fig. 8-4, which shows the oscillations in a mass-spring system, indicates
how close the correspondence is. Note how v and i correspond in the two
figures; also x and q. Note, too, how in each case the energy alternates
between two forms, magnetic and electric for the LC system, and kinetic
and potential for the mass-spring system.
In Section 15-3 we saw that the natural angular frequency of oscillation
of an undamped mass-spring system is
/-1
w = 21rv = ~LC. (38-3)
We can test whether Eq. 38-6 is indeed a solution of Eq. 38-5 by substi-
tuting it and its second derivative in that equation. To find the second
derivative, we write
dq .
~ = i = -wqm sin (wt </>) +
(38-7)
dt
dzq
and -w 2 qm cos (wt + </>).
dt 2
Substituting q and d 2 q/dt 2 into Eq. 38-5 yields
1
--Lw 2 qm COS (wt + </>) + - qm COS (wt + </>) = 0.
c
Canceling qm cos (wt + <I>) and rearranging leads to
Thus, if w is given the constant value 1/yLC, Eq. 38-6 is indeed a solution
of Eq. 38-5. This expression for w agrees with Eq. 38-3, which was arrived
at by the method of correspondences.
The phase angle cp in Eq. 38--6 is determined by the conditions that
prevail at t = 0. If the initial condition is as represented by Fig. 38-la,
then we put </> = 0 in order that Eq. 38-6 may predict q = qm at t = 0.
What initial condition in Fig. 38-1 is implied if we select </> = 90°?
..... Example 2. (a) In an oscillating LC circuit, what value of charge, expressed in
terms of the maximum charge, is present on the capacitor when the energy is shared
equally between the electric and the magnetic field? (b) How much time is required
for this condition to arise, assuming the capacitor to be fully charged initially?
Assume that L = 10 mh and C = 1.0 ,uf.
(a) The stored energy and the maximum stored energy in the capacitor are,
respectively,
and
q = qm COS wt = V1 2 qm,
1 7r 7r
which leads to wt = cos- 1 --= = - or t = -·
y2 4 4w
The angular frequency w is found from Eq. 38-3, or
Fig. 38-4 The stored magnetic and electric energy in the circuit of Fig. 38-1. Note that
their sum is a constant.
Figure 38-4 shows plots of UE(t) and Us(t) for the case of q, = 0. Note
that (a) the maximum values of UE and Us are the same ( = qm 2 /2C); (b)
at any instant the sum of UE and Us is a constant ( = qm 2 /2C); and (c) when
UE has its maximum value, Us is zero and conversely. This analysis sup-
ports the qualitative analysis of Section 38-1. The student should compare
this discussion with that given in Section 15-4 for the energy transfers in
a mass-spring system .
.... Example 3. The LCR circuit. (a) Derive an expression for the quantity q(t)
for a single-loop circuit containing a resistance R as well as an inductance L and a
capacitance C. (b) After how long a time will the charge oscillations decay to half-
amplitude if L = 10 mh, C = 1.0 µf, and R = 0.1 ohm'?
(a) If U is the total stored field energy, we have, as before,
the minus sign signifying that the stored energy U decreases with time, being con-
verted to Joule heat at the rate i 2R. Combining these two equations leads to
which describes damped mass-spring oscillations. Once again the equations are
mathematically identical, the resistance R corresponding to the mechanical damping
constant b.
The solution of the LCR circuit follows at once, by correspondence, from the
solution of Eq. 15-37. It is (see Eqs. 15-38 and 15-39) for R reasonably small, and
for an initial condition in which the capacitor has a maximum charge
q = qme-Rtf 2L cos w't, (38-10)
where (38-11)
Note that Eq. 38-10, which can be described as a cosine function with an expo-
nentially decreasing amplitude, is the equation of the decay curve of Fig. 38-3. Note,
too (Eq. 38-11), that the presence of resistance reduces the oscillation frequency.
These two equations reduce to familiar results as R ----, 0.
(b) The oscillation amplitude will have decreased to half when the amplitude
factor e-Rtf 2L in Eq. 38-10 has the value one-half, or
, / 1 ( 0.10 ohm )2
w = '\J (10 X 10- henry)(l.O X 10- farad) - 2 X 10 X 10- henry
3 5 3
Note that the second term is rather small, so that in this case, as in many practical
cases, the resistance has a negligible effect on the frequency. The student should
show that 0.14 sec, the time at which the oscillations decrease to half-amplitude,
corresponds to about 220 cycles of oscillation. The damping is much less severe
than that illustrated in Fig. 38-3. ..._
952 ELECTROMAGNETIC OSCILLATIONS Chap. 38
38-4 Forced Oscillations and Resonance
Figure 38-5 shows an LCR circuit containing a sinusoidally varying emf S(t) given
by
8 = Sm cos w11t,
in which w" can be varied at will. The emf S(t) might be provided by a variable-
frequency oscillator. What amplitudes of electromagnetic oscillations are set up in
this circuit for various angular frequencies w" of the "driving force"?
The problem corresponds to that of forced oscillations in the damped mass-spring
system of Section 15-10. The differential equation describing that motion is
d 2x
m dt2 + b dx k
dt + X =
F
m
II
cos W t, (15-31)
where w" is the angular frequency of the external periodic driving force applied to the
system and Fm is its amplitude.
For the circuit of Fig. 38-5, the differential equation that follows from the cor-
respondences of p. 947 and the additional reasonable correspondence of 8 to F, is
L d2q
dt2 + R dqdt + (;1 q = C'
um COS W
II
t. (38-12)
This equation can also be derived by applying the energy conservation principle to
the circuit of Fig. 38-5.
f,(t)
We can use the correspondences of p. 947 to write down the solution to Eq. 38-12.
Starting from Eq.15-41 and making the appropriate substitutions, we obtain
q= 8; sin (w 11
t - </>), (38-13)
where
and </>, the phase angle between the "driving force" and the "response," is given by
Rw 11
"*'=cos
"" -1
0 .
As often as not, we are interested in the current i(t) in the circuit, rather than the
charge; the current corresponds to the velocity v(t) of the moving mass in Section
15-10. We can find i(t) by differentiating Eq. 38-13, or
i
. = -dq = w"Sm
-- ( ,, t - 'I'
COS W
"") = 1-m
. COS W
( ,, t - 'I'
"") .
dt G
Sec. 38-4 FORCED OSCILLATIONS AND RESONANCE 953
The amplitude im of the current oscillations is given, from these equations, by
(38-14)
Inspection of Eq. 38-14 shows that the current (not the charge; see Question 5) will
have its maximum amplitude when
w"L = _!__, (38-15a)
w"C
Comparison with Eq. 38-3 shows that the maximum amplitude of the current oscilla-
tions occurs when the frequency w" of the applied emf is exactly equal to the natural
(undamped) frequency w of the system.
At resonance (w" = w) the amplitude of the current oscillations is determined
entirely by the resistance; this follows by combining Eqs. 38-14 and 38-15a, or
. Sm
im = R (at resonance).
Figure 38-6 shows im as a function of w" for an oscillating LCR circuit containing
three different values of resistance. Note that the smaller the resistance, the sharper
the resonance curve. The sharpness of such a curve is measured by its half-width,
which is the difference between two frequencies, each of which corresponds to a cur-
rent amplitude of one-half the maximum current amplitude. The half-width of the
curve for R = 10 ohms in Fig. 38-6 is shown in that figure by the arrow marked Liw.
1.0
L = lOOµh
0.9 C = lOOµµf
E = lOmv
0.8
0.7 R = IO ohms
~ 0.6
~
.,.,
I 0.5
0
.....
-
!:: 0.4
0.3
0.2
0.1
Fig. 38-6 The current amplitude as a function of frequency in the circuit of Fig. 38-5.
The arrow marked Liw on the curve for R = 10 ohms is the half-width of that curve.
954 ELECTROMAGNETIC OSCILLATIONS Chap. 38
38-5 Lumped and Distributed Elements
In the oscillating mass-spring system the two kinds of energy involved
appear in separate parts of the system, the potential energy being stored in
the spring and the kinetic energy in the moving mass. An acoustic cavity
resonator, such as an organ pipe, is a mechanical oscillating system in which
the two forms of energy are not separated in space. Kinetic energy, asso-
ciated with moving air in the cavity, and potential energy, associated with
compressions or rarefactions of the air, can both be present throughout the
volume of the cavity. Such a cavity is an example of an oscillating system
with distributed rather than lumped (as in the mass-spring system) elements.
A similar distinction exists in electromagnetic systems. The LC circuit
of Fig. 38-1 is an example of lumped elements, in that the two kinds of
energy are stored in rather different places; the circuit is described by giving
the (lumped) system parameters Land C. Actually, in modern engineering
practice and in physics research electromagnetic systems with distributed
elements play a fundamental role.
Figure 38-7, a series of "snapshots" taken one-eighth of a cycle apart,
shows the pressure and velocity variations in the fundamental mode of a
particular acoustic resonator. There is a pressure node at the center and a
pressure antinode at each end. There is a velocity * node at each end and
a velocity antinode at the center. When the pressure variation is the great-
est, the velocity is zero (Figs. 38-7a and e). When the pressure is uniform,
the velocities have their maximum values (Figs. 38-7c, and g).
The energy in the acoustic resonator alternates between kinetic energy
associated with the moving gas and potential energy associated with the
compression and rarefaction of the gas. In Figs. 38-7c and g the energy is
all kinetic and in Figs. 38-7a and e it is all potential. In intermediate phases
it is part of each.
The kinetic energy of a small mass t::.m of the gas, which is moving parallel to the
cylinder axis with a speed Vg, is !t::.mvg 2, The kinetic energy density, that is, the
kinetic energy per unit volume, is
2 iv Vg
1 t::.m 2 1 2
UK = = zpoVg ,
in which t::. V is the volume of the gas element and po is the mean density of the gri,s.
The potential energy per unit volume in the gas, that is, the potential energy dens~ty,
associated with the compressions and rarefactions of the gas may be given by
Here B is the bulk modulus of elasticity of the gas and t::.p/Po, which is positive for a
compression and negative for a rarefaction, is the fractional change in gas density.
* The velocity of interest here is the directed velocity Vg of small volume elements of
the gas which are, however, large enough to contain a great number of molecules. The
thermal velocities of the molecules have no directional preference and are ignored.
Sec. 38-6 ELECTROMAGNETIC CAVITY OSCILLATOR 955
(c)
<+& . ···4.·.·.··,--i:,······i
~ =+~iti ~
~-
:;-r;-.. ~·:_:~. .:. ·~.. F.~
..~-:·_:~·-: . .·_~}.: .:
(b) ·.:::. :::: : UK I Up . ~> ::::···:':t>:: (d)
! \
(a) (e)
I
(h)
r:;..
~
:·..:~: .<J- .
,:,:... .......~·.· .-. .
<J,,- . . . ~ - -
:-:-:-·.
~·
Fig. 38-7 Showing eight stages in a cycle of oscillation of a cylindrical acoustic resonant
cavity. The bar graphs below each figure show the kinetic and potential energy. The
arrows represent the directed velocities of small volume elements of the gas. Compare
with Fig. 38-1.
The angular frequency of oscillation for the cavity of Fig. 38-7, in the fundamental
(or lowest frequency) mode shown in that figure, is found from
27rV 7rV
W1 = 271'V = -~ = T'
where vis the speed of sound in the gas and l is the length of the cavity. From Eq.
20-1 we may write v as yB 0 /p 0 • Note that in the above we have put 'A = 2Z, cor-
responding to the fundamental mode. What are the frequencies w2, wa, etc., of the
higher frequency modes?
(c)
.................
. . ... ... . ..
. . .- . .... . -.
x x x x
xxxxxxxx
xxxxxxxxx-xxxxxx -
UBg 111J1i1 UE
t \
(a)
~ ~(e)
~
L!, E
I
x
Ix
x x x x x"T/"
.. . . ... . . ..
x x
(h) (f)
(g)
xxxxxxxxxxxxxxx
xxxxxxxx
x x x x
. ... ... . .
................•
in the cavity, energy is stored in the electric field in an amount per unit
volume given by Eq. 30-27, or
1 E2 (38-16)
UE = zEo .
Figure 38-8 also shows, by the dots and crosses, how the magnetic field B
varies with time. The magnetic lines form circles about the cylinder axis.
Note that the magnetic field has a maximum value when the electric field is
zero, and conversely. At any point in the cavity energy is stored in the
magnetic field in an amount per unit volume given by Eq. 36-19, or
1 2
UB = -B. (38-17)
2µo
Thus, as in the LC circuit, energy is shuttled back and forth between the
electric and the magnetic fields. However, these fields no longer occupy
completely separate regions of space.
We state without proof that the angular frequency of oscillation for the electro-
magnetic cavity of Fig. 38-8 is, in the fundamental mode shown in that figure,
1.19c
W1 = --,
a
in which a is the cavity radius and c is the speed of electromagnetic radiations in
free space. We will see in Section 39-5 that c may be written in terms of electro-
magnetic quantities as 1/ ~ - As the field patterns of Fig. 38-8 suggest, the
resonant frequency of oscillation of the cavity, for the mode of oscillation shown,
depends only on the radius of the cavity and not on its length.
Table 38-2 summarizes some characteristics of the four oscillating systems that
we have discussed so far. For lumped systems it gives expressions for the two kinds
Table 38-2
FouR OscrLLATING SYSTEMS
IJJl,- Example 4. In the cavity of Fig. 38-8, what is the relationship between the
"average" value of E throughout the cavity, measured at the instant corresponding
to Fig. 38-8a, to the "average" value of B, measured at the instant corresponding to
Fig. 38-8c?
At the first instant the energy is all electric and at the second it is all magnetic.
The total energy U, found by integrating the energy density over the volume of the
cavity, must be the same at these two instants, or
U = f UE,mdV = f UB,mdV,
where dV is a volume element in the cavity and UE,m and UB,m are the maximum values
of UE and of UB at the site of this volume element; these maximum values occur one-
fourth of a cycle apart, as Fig. 38-8 shows. Substituting Eqs. 38-16 and 38-17 leads
to
I
foEm2 dV =
2
I
Bm2 dV
2µo
or
f
The quantity Em 2 dV can be written as Em 2V, where V is the cavity volume and
Em 2 is the average value of Em 2 throughout the cavity. Treating Bm in the same way
leads to
What is the total stored energy in the cavity under these conditions, assuming the
cavity to be 10 cm long and 3.0 cm in diameter?
The student will recall that in Example 6, Chapter 36, we showed the energy
density for a magnetic field of "ordinary" laboratory magnitude (say, 1 weber/meter2 )
to be enormously greater than that for an electric field of "ordinary" magnitude
(say, 10 5 volts/meter). This fact is consistent with the present example. <1111
E
.,,,..,,,,. .,..-- ------ -.... .............. + { -
/x x x x x,
/ // ',,.,/E + -
/ 'Ac- + -
/ X x x x,
I B ' + -
I \
/ xx xx xx\ + -
I r , -
+
Ix
I
x x x x< x x x\
J + -
I
IX X X X X X X X XI
I i + -- i
I I
+ -
\ R /
\xx xx x I + -
\ I
\
\ I
I
+ -
,x x x x x/
', B / + -
', ;/
+ -
_________
'~.... ......... x x x ,,..,,. .,,."'>""
-
+
(a) (b)
Fig. 38-9 (a) Showing the induced magnetic fields Bat four points, produced by a chang-
ing electric field E. The electric field is increasing in magnitude. Compare Fig. 35-10.
(b) Such a changing electric field may be produced by charging a parallel-plate capacitor as
shown.
Equation 38-20 asserts that a magnetic field (left term) can be produced by
a changing electric field (right term). The student should compare carefully
Fig. 35-10, which illustrates the production of an electric field by a changing
magnetic field, with Fig. 38-9a. In each case the appropriate flux <PB or <PE
is increasing. However, experiment shows that the lines of E in Fig. 35-10
are counterclockwise, whereas those of B in Fig. 38-9a are clockwise. This
difference requires that the minus sign of Eq. 38-19 be omitted from Eq.
38-20.
In Section 34-1 we saw that a magnetic field can also be set up by a cur-
rent in a wire. We described this quantitatively by Ampere's law:
in which i is the conduction current passing through the loop around which
the line integral is taken. Thus there are at least two ways of setting up
* Our system of units requires that we insert the constants eo and µo in Eq. 38-20. In
some unit systems they would not appear.
Sec. 38-7 INDUCED MAGNE'rIC FIELDS 961
(38-21)
or B = µ01:oR~ dE
2r dt
(r > R).
(b) Find Bat r = R for dE/dt = 10 12 volts/m-sec and for R = 5.0 cm. At r = R
the two equations for B reduce to the same expression, or
In all that follows we assume that no magnetic materials are present so that iM = 0.
962 ELECTROMAGNETIC OSCILLATIONS Chap. 38
ference is in part due to the fact that induced emfs can easily be multiplied by using
a coil of many turns. No technique of comparable efficiency exists for magnetic
fields. In experiments involving oscillations at very high frequencies dE/dt above
can be very large, resulting in significantly larger values of the induced magnetic
field. ~
q
<PE = -·
Eo
Therefore the displacement current in the gap is identical ,vith the conduc-
tion current in the lead wires.
* The word "displacement" was introduced for historical reasons that need not concern
us here.
t We may write this more generally, taking the presence of magnetic materials into
account, as
* Electrons, Waves and lv! essages, Hanover House, 1956. This book is recommended as
collateral reading in electromagnetism.
Table 38-3
°''°.s:,.
THE BASIC EQUATIONS OF ELECTROMAGNETISM (MAXWELL'S EQUATIONS)*
Text
Name Equation Describes Crucial Experiment
Reference
-
lctj
t"
lctj
Cl
Gauss's law for 1:0 fE·dS = q Charge and the electric field 1. Like charges repel and unlike charges attract, as Chapter 28 >-3
~
electricity the inverse square of their separation. 0
l'. A charge on an insulated conductor moves to its ~
0
outer surface. 'Z
~
H
Gauss's law for f B·dS = 0 The magnetic field 2. It is impossible to create an isolated magnetic Section 37-2 Cl
0
magnetism pole. Ul
Cl
H
t"
t"
Ampere's law (as ex- fB·dl The magnetic effect of a 3. The speed of light can be calculated from purely Section 39-5
~
tended by Maxwell) changing electric field or electromagnetic measurements. H
0
= µo ( Eo dcJ>E
dt + i ') of a current 3'. A current in a wire sets up a magnetic field near Chapter 34 ~
al
' the wire.
Faraday's law of in- f E·dl = - dPs The electrical effect of a 4. A bar magnet, thrust through a closed loop of Chapter 35
duction dt changing magnetic field wire, will set up a current in the loop.
Q
:::;"
~
* Written on the assumption that no dielectric or magnetic material is present. "Cl
c,.,
00
Sec. 38-10 MAXWELL'S EQUATIONS AND CAVITY OSCILLATIONS 965
r--- h
-
·-
• I •
· ~-· •
r
. . . . . . . . ..•t • ... ..
E e • - •J•
. . . .I. .
--=---t>- - ~. --
x x x x x x x
x x x x x x x x x x x
X- x x x x x x x x x x x x x x x
(a) (b)
Fig. 38-10 Two cross sections of an electromagnetic resonant cavity at a phase of oscilla-
tion corresponding to Fig. 38-Sd. (a) The dashed circle is a path suitable for applying
Ampere's law. (b) The dashed rectangle is a path suitable for applying Faraday's law.
f E·dl = hE(r),
in which E(r) is the value of E at a radius r from the center of the cavity.
Note that E equals zero for the upper leg of the integration path (which
lies in the cavity wall) and that E and di are at right angles on the two side
legs. Combining these equations yields
1 d<l'>B
E(r) = - - - · (38-23)
h dt
Equation 38-23 shows that E(r) depends on the rate at which cf>B through
the path shown is changing with time and that it has its maximum value
when d<I>B/dt is a maximum. This occurs when Bis zero, that is, when Bis
changing its direction; the student will recall that a sine or cosi,1e is changing
most rapidly, that is, it has the steepest tangent, at the instant it crosses the
axis between positive and negative values. Thus the electric field pattern
in the cavity will have its maximum value when the magnetic field is zero
everywhere, consistent with Figs. 38-8a and e and with the concept of the
interchange of energy between electric and magnetic forms. The student
should demonstrate, by applying Lenz's law, that the electric field in Fig.
38-lOb points to the right as shown, if the magnetic field is decreasing.
Figure 38-lOa shows an end view of the cavity; the electric lines of force
are entering the page at right angles to the page and the magnetic lines form
clockwise circles. Let us apply Ampere's law in the form
'
(38-21)
966 ELECTROMAGNETIC OSCILLATIONS Chap. 38
to the circular path of radius r shown in the figure. ~ o charge is transported
through the ring so that the conduction current i in Eq. 38-21 is zero. The
line integral on the left is (B) (21rr) so that the equation reduces to
(38-24)
Equation 38-24 shows that the magnetic field B(r) is proportional to the
rate at which the electric flux cf>E through the ring is changing with time.
The field B(r) has its maximum value when dcf>E/dt is at its maximum; this
occurs when E = 0, that is, when E is reversing its direction. Thus we see
that B has its maximum value when E is zero for all points in the cavity.
This is consistent with Figs. 38-8c and g and with the concept of the inter-
change of energy between electric and magnetic forms. A comparison with
Fig. :"38-9a, which, like Fig. 38- lOa, corresponds to an increasing electric
field, shows that the lines of B are indeed clockwise, as viewed along the di-
rection of the electric field.
Comparison of Eqs. 38-23 and 38-24 suggests the complete interdepend-
ence of B and E in the cavity. As the magnetic field changes with time, it
induces the electric field in a way described by Faraday's law. The electric
field, which also changes with time, induces the magnetic field in a way
described by Maxwell's extension of Ampere's law. The oscillations, once
established, sustain each other and would continue indefinitely were it not
for losses due to Joule heating in the cavity walls or leakage of energy from
openings that might be present in the walls. In Chapter 39 we show that
this interplay of B and E occurs not only in standing electromagnetic waves
in cavities but also in traveling electromagnetic waves, such as radio waves
or visible light.
µo dcf>E µo .
B(r) = -- fo --- = -- ia.
21rr dt 21rr
This equation stresses that B in the cavity is associated with a displacement current;
compare Eq. 34-4. Applying the right-hand rule in Fig. 38-lOa shows that the dis-
placement current ia must be directed into the plane of the figure if it is to be associ-
ated with the clockwise lines of B that are present.
The displacement current is represented in Fig. 38-llb by the arrows that point to
the right and in Fig. 38-lla by the crosses that represent arrows entering the page.
Study of Fig. 38-11 shows that the current is continuous, being directed up the
Chap. 38 QUESTIONS 967
(a) (b)
Fig. 38-11 The cavity of Fig. 38-10 showing (a) the conduction current coming up the
walls and displacement current going down the cavity volume and (b) the displacement
current (black arrowheads) in the volume of the cavity and the conduction currents (white
arrowheads) at the walls. The arrows in each case represent current densities. Note
the continuity of current (conduction + displacement), that is, it is possible to form
closed current loops. (For a truly resistanceless cavity the conduction current would
lie entirely on the surface, requiring a modification of our definition of current density;
we leave this small complication for the student to consider.)
walls as a conduction current and then back down through the volume of the cavity as
a displacement current. If we apply Ampere's law as extended by Maxwell,
to the circular path of radius r 1 in Fig. 38-lla, we see that B at that path is due
entirely to the displacement current, the conduction current i within the path being
zero.
For the path of radius r 2, the net current enclosed is zero because the conduction
current in the walls is exactly equal and opposite to the displacement current in the
cavity volume. Since i equals ia in magnitude, but is oppositely directed, it follows
from Eq. 38-25 that B must be zero for all points outside the cavity, in agreement
with observation.
QUESTIONS
1. Why doesn't the LC circuit of Fig. 38-1 simply stop oscillating when the capacitor
has been completely discharged?
2. How might you start an LC circuit into oscillation with its initial condition· being
represented by Fig. 38-lc? Devise a switching scheme to bring this about.
3. In an oscillating LC circuit, assumed resistanceless, what determines (a) the fre-
quency and (b) the amplitude of the oscillations?
4. Tabulate as many mechanical or electric systems as you can think of that possess a
natural frequency, along with the formula for that frequency if given in the text.
5. Resonance in LC circuits, judged by Eq. 38-14 and Fig. 38-6, occurs when the fre-
quency w" of the "driving force" is exactly equal to the (undamped) natural frequency
of the LC circuit. In Section 15-10 we saw that resonance for mass-spring systems,
judged by Eq. 15-41 and Fig. 15-19, occurs when w" is close to, but not exactly equal to,
the (undamped) natural frequency of the mass-spring system. Is there a failure of the
principle of correspondence here?
968 ELECTROMAGNETIC OSCILLATIONS Chap. 38
6. Suppose that, in a low frequency oscillating LCR circuit, one could display on the
screen of a cathode-ray oscilloscope a potential difference proportional to the temperature
of the (specially designed) resistor as it changed with time in response to the hes.ting effect
of the oscillating current. What would the connection be between the frequency of this
temperature-time waveform and that of the oscillations of current, as measured by the
potential difference across the resistor?
7. Discuss the periodic flow of energy, if any, from point to point in an acoustic resonant
cavity.
8. Can a given circuit element (a capacitor, say) behave like a "lumped" element under
some circumstances and like a "distributed" element under others?
9. List as many (a) lumped and (b) distributed mechanical oscillating systems as you
can.
10. An air-filled acoustic resonant cavity and an electromagnetic resonant cavity of the
same size have resonant frequencies that are in the ratio of 106 or so. Which has the higher
frequency and why?
11. What constructional difficulties would you encounter if you tried to build an LC
circuit of the type shown in Fig. 38-1 to oscillate (a) at 0.01 cps, or (b) at 1010 cycles/sec?
12. Could an electromagnetic cavity of the type shown in Fig. 38-8 be designed to
operate at 60 cycles/sec? If so, give some details of its construction.
13. The "sharpness of tuning" of a copper electromagnetic cavity can be considerably
ipcreased by immersing it in liquid air. Explain. Be guided by Fig. 38-6, which shows
"tuning curves" for an LC circuit.
14. Electromagnetic cavities are often silver plated on the inside. Explain.
15. Why is Faraday's law of induction more familiar than its symmetrical counterpart,
Eq. 38-20?
16. Why is the quantity EO d4!E/dt referred to as a (displacement) current1
17. In Fig. 38-lc a displacement current is needed to maintain continuity of current in
the capacitor. How can one exist, since there is no charge on the capacitor?
18. Discuss the symmetries that appear between (a) the first two and (b) the second two
of Maxwell's equations.
19. At what parts of the cycle will (a) the conduction current and (b) the displacement
current in the cavity of Fig. 38-8 be zero?
20. Discuss the time variation during one complete cycle of the charges that appear
at various points on the inner walls of the oscillating electromagnetic cavity of Fig. 38-8.
PROBLEMS
l. You are given a 10-mh inductor and two capacitors, of 5.0- and 2.0-µ.f capacitance.
What resonant frequencies can be obtained by connecting these elements in various ways?
2. Given a 1.0-mh inductor, how would you make it oscillate at 1.0 X 106 cycles/sec?
3. An inductor is connected across a capacitor whose capacitance can be varied by
turning a knob. We wish to make the frequency of the LC oscillations vary linearly with
the angle of rotation of the knob, going from 2 X 105 cycles/sec to 4 X 105 cycles/sec as
the knob turns through 180°. If L = 1.0 mh, plot Casa function of angle for the 180°
rotation.
4. A 10-henry coil has a resistance of 180 ohms. What size of capacitor must be put in
series with it if the combination is to "resonate" when connected to a 60-cycle/sec outlet?
5. Derive the differential equation for an LC circuit (Eq. 38-5) using the loop theorem.
6. Derive Eq. 38-12, the differential equation for forced oscillations in an LCR circuit,
from the principle of conservation of energy.
7. Show that the q-oscillations in a damped LC circuit (see Eq. 38-10) reach half
their initial amplitude in a time given approximately by 0.69 (L/R).
8. A circuit has L = 10 mh and C = 1.0 µ.f. How much resistance must be inserted in
the circuit to reduce the (undamped) resonant frequency by 0.01 %?
Chap. 38 PROBLEMS 969
9. Suppose that in an oscillating LCR circuit the amplitude of the charge oscillations
drops to one-half its initial value after n cycles. Show that the fractional reduction in
the frequency of resonance, caused by the presence of the resistor, is given to a close
approximation by
w - w' 0.0061
--w-=~,
-1 R
in which </>=tan --·
2Lw'
Start from Eq. 38-10.
11. "Q" for a circuit. In the damped LC circuit of Example 3 show that the fraction
of the energy lost per cycle of oscillation, l}U /U, is given to a close approximation by
21rR/wL. The quantity wL/R is often called the "Q" of the circuit (for "quality"). A
"high-Q" circuit has low resistance, low fractional energy loss per cycle ( = 21r /Q) and
(see Fig. 38-6) a sharp resonance or "tuning" curve.
12. Show that the amplitude of the charge oscillations in an oscillating LCR circuit h1
given by
8m
~;====================
qm =
~( w"2L -iY + (w"R)2
in which w is the resonant frequency and /}w is the width of the resonance peak at i = Mm.
Note (see Problem 11) that this expression may be written as y'3/Q which shows clearly
that a "high-Q" circuit has a sharp resonance peak, that is, a small /}w/w.
15. A resistor-inductor-capacitor combination R1, Li, C1 (connected in series) exhibits
resonance at the same frequency as a second combination R2, L2, C2. If the two com-
binations are now connected in series, at what frequency would the whole circuit resonate?
16. In Example 5 show that the displacement current density ja is given, for r < R, by
dE
Jd = EO
dt'
17. Prove that the displacement current in a parallel-p~ate capacitor can be written as
dV
ia = C--·
dt
18. You are given a 1.0-µ.f capacitor. How would you establish an (instantaneous) dis-
placement current of 1.0 amp in the space between its plates?
19. In Example 5 how does the displacement current through a concentric circular loop
of radius r vary with r? Consider both r < Rand r > R.
970 ELECTROMAGNETIC OSCILLATIONS Chap. 38
20. In microscopic terms the principle of continuity of current may be expressed as
f (j + ja) ·dS = 0,
in which j is the conduction current density and ja is the displacement current density.
The integral is to be taken over any closed surface; the equation essentially says that
whatever current flows into the enclosed volume must also flow out. (a) Apply this equa-
tion to the surface shown by the dashed lines in Fig. 38-12 shortly after switch Sis closed.
(b) Apply it to various surfaces that may be drawn in the cavity of Fig. 38-11, including
some that cut the cavity walls.
Fig. 38-12
21. A cylindrical electromagnetic cavity 5.0 cm in diameter and 7.0 cm long is oscillating
in the mode shown in Fig. 38-8. (a) Assume that, for points on the axis of the cavity,
Em = 10 4 volts/meter. For such axial points what is the maximum rate (dE /dt)m at which
E changes? (b) Assume that the average value of (dE /dt)m, for all points over a cross
section of a cavity, is about one-half the value found above for axial points. On this
assumption, what is the maximum value of Bat the cylindrical surface of the cavity?
--:22:· Collect and tabulate expressions for the following four quantities, considering both
r < R and r > R. Copy down the derivations side by side and study them as interesting
applications of Maxwell's equations to problems having cylindrical symmetry.
(a) B(r) for a current i in a long wire of radius R (see Section 34-2).
(b) E(r) for a long uniform cylinder of charge of radius R (see Section 28-6; also Problem
14, Chapter 28).
(c) B(r) for a parallel-plate capacitor, with circular plates of radius R, in which E is
changing at a constant rate (see Section 38-7).
(d) E(r) for a cylindrical region of radius R in which a uniform magnetic field B 1s
changing at a constant rate (see Section 35-5).
Electromagnetic Waves *
CHAPTER 39
s
a b
~ 1____c...____
Fig. 39-1 An electromagnetic pulse can be sent along the coaxial cable by throwing
switch S from b to a.
971
speed c that will turn out to be exactly that of light, assuming a resistanceless
line. Figure 39-2a shows that the potential difference between the con-
ductors at a distance l along the line suddenly rises, at a time given by t =
l/c, from zero to a value determined by the battery emf. We can also con-
sider the variation of V with position x along the line at a given time t 1
after closing the switch. Figure 39-2b shows such an instantaneous "snap-
shot." It, too, suggests a traveling "wavefront" moving along the line at
speed c. At t = t 1 the signal has not yet reached points where x > ct 1 .
Fig. 39-2 (a) The variation with time of the potential difference between the conductors
of a coaxial cable at a distance l from the input end. (b) An instantaneous "snapshot" of
the pulse in the cable at a certain time t1. (c) The waveform if switch S in Fig. 39-1 is
periodically thrown between a and b. (d) The waveform if switch S is replaced by an
electromagnetic oscillator with a sinusoidal output.
Sec. 39-1 TRANSMISSION LINE 973
If switch Sis periodically thrown from b to a and back again, a wave dis-
turbance like that of Fig. 39-2c is propagated. This suggests that if the
battery and switch arrangement is replaced by an electromagnetic oscillator
with a sinusoidal output of frequency v a wave like that of Fig. 39-2d will
be propagated.
A traveling wave in a resistanceless transmission line will exhibit a wave-
length A given by
c
/\ = -·
v
Vo= V m sin wt
is applied between the terminals of a long resistanceless transmission line; the fre-
quency P ( = w/21r) is 3 X 109 cycles/sec. Write an equation for V(t) at a point P
which is 1.5 wavelengths down the line from the oscillator.
The general equation for a wave traveling in the x direction (see Eq. 19-10) is
- V m sin wt.
Thus VP is always equal in magnitude to Vo but is opposite in sign. What is the
wavelength in this example? ~
..... x xx xx . .... x xx x x
- <1---
i --I>
<1---
l <1---
--I>
~ t >Displacement
<1---
current
(c)
Fig. 39-3 (a) The electric and magnetic fields in a coaxial cable, showing a wave traveling
to the right at speed c. (b) A cross-sectional view at a plane through xx in (a); the wave is
emerging from the page. (c) Conduction current (open arrows) and displacement currents
(filled arrows) associated with the wave in (a); the arrows in each case represent current
density vectors.
{E'\
x x x • • •
x •
x x x • • •
x ...-----.-.0is
x x • •
x •
x
x
x
x
x
•
•
·~
•
•
x •
x x • •
x •
x x x • • •
Fig. 39-4 The area element Sin Fig. 39-3 (a)
enlarged and viewed from above, showing
the adjacent (a) electric and (b) magnetic
fields.
(b)
976 ELECTROMAGNETIC WAVES Chap. 39
rate, since Eat the element Mis at the very moment of reversing its direction as the
wave moves through. Thus the displacement current, which is given by
. d<!JE
ia = t:o--•
dt
also has its maximum value.
Figure 39-4b shows the magnetic field in the vicinity of the element of area. Let
us apply the generalized form of Ampere's law,
Example 3. Show that the conduction currents in Fig. 39-3c are appropriately
related to the magnetic field pattern.
/
/ -------- Conduction
/
/ ~i , current
I \
I \
I I Fig. 39-5 Example 3. The coaxial cable of
I \
I I Fig. 39-3b, showing the conduction currents
I I in the central and outer conductors. The
I wave is emerging from the page.
I
\
\
' "'
----~ ---------
Figure 39-5 shows a cross section of the cable at a plane through xx in Fig. 39-3a.
Let us apply Ampere's law,
Note that Bis related to the current in the central conductor by the usual right-hand
rule and that the expression for Bis that found earlier (Eq. 34-4) for a long straight
wire carrying a steady current.
Sec. 39-3 WAVEGUIDE 977
"Loudspeaker
Fig. 39-6 An acoustic transmission line, showing a sound wave traveling to the right.
The small arrows with filled heads show the directed drift velocities for small volume ele-
ments of the gar5. Compare Fig. 39-3a.
If we apply Ampere's law to the large ring of radius r', we must put the net conduc-
tion current equal to zero because the current in the outer conductor, shown by the
dots, is equal and opposite to that in the inner conductor. This means that B must
be zero for points outside the cable, in agreement with experiment. -11111
39-3 Waveguide
It is possible to send electromagnetic waves through a hollow metal pipe. We
assume that the inner walls of such a pipe, or waveguide as it is called, are resistance-
less and that the cross section is rectangular.
Figure 39-7 shows a typical electric and magnetic field pattern. We imagine that
a microwave oscillator is connected to the left end and sends electromagnetic energy
down the guide. Figure 39-7a shows a side view of the guide and Fig. 39-7b, a top
view; Fig. 39-7c shows the cross section. As for the coaxial cable, the field patterns
are such that E has no tangential component for any point on the inner surface of the
guide. The fields E and B are in phase, again like the coaxial cable.
As for all traveling waves, the angular frequency w of electromagnetic waves
traveling down a guide can be varied continuously. In a waveguide of given dimen-
sions, however, there exists, for every mode of transmission, that is, for every pattern
of E and B, a so-called cutoff frequency wo. A given guide will not transmit waves in a
given mode if their frequency is below the cutoff value for that mode in that guide.
The field patterns of Fig. 39-7 show the dominant mode for a rectangular guide; this
is the mode with the lowest cutoff frequency. Given the frequency w of electromag-
netic waves to be transmitted, it is common practice to select a guide whose dimen-
sions are such that w is larger than the cutoff frequency w0 for the dominant mode
978 ELECTROMAGNETIC WAVES Chap. 39
HIH
E\
~HtH
r a 1_1_
(a)
HtmH Hlt*tH 1Itl*tH7 (c)
I I
Xg
I I
I
I
I
ph
~-t
I I I I I
(b)
~~jf l iji 1 tijijt 1 .. 11M1+
1
I
1Jij~ --i>
Vph
a
t
Fig. 39-7 A waveguide, showing (a) a side view of the lines of E, (b) a top view of the
lines of B, and (c) a cross-sectional view of the lines of E. In (c) the wave is emerging from
the page. For simplicity the lines of Bare not shown in (a) and (c), nor are the lines of E
shown in (b).
but smaller than the cutoff frequencies of all other modes. Under these conditions
the dominant mode of propagation is the only one possible.
In a (resistanceless) coaxial cable the wave patterns travel at speed c. In the
acoustic transmission line of Fig. 39-6 (assumed "resistanceless") the waves also
travel at a speed v, which is the same as the propagation speed in an infinite medium.
In a waveguide, however, the speed is not c. In waveguides we must distinguish.
between (a) the phase speed Vph, which is the speed at which the wave patterns of
Fig. 39-7 travel, and (b) the group speed Vgr, which is the speed at which electromag-
netic energy or information-carrying "signals" travel along the guide. These speeds,
which are identical for electromagnetic waves in a coaxial cable and for acoustic
waves in a tube, are different for waves in a waveguide.
The phase speed is not directly measurable. The wave pattern is a repetitive struc-
ture, and there is no way to distinguish one wave maximum from another. The waves
can be observed to enter one end of the guide and to leave at the other, but there is no
way to identify a particular wave maximum so that its passage down the guide can
be timed. We can put a "signal" on the wave by increasing the power level of the
oscillator for a short time. This power pulse could be timed as it passes through
the guide, but there is no guarantee that it travels at the same speed as the wave
pattern and, indeed, it does not. The speed of such signals or markers is the speed
at which energy is propagated, that is, the group speed.
From Maxwell's equations it can be shown that the phase speed and the group
speed for the mode of Fig. 39-7 are
c
(39-1)
and (39-2)
in which a is the width of the guide and A the free-space wavelength. Note that as
a ---+ CXJ, which corresponds to free-space conditions, Vph = Vgr = c.
The phase speed Vph is greater than the velocity of light, the group speed Vgr being
correspondingly less. In relativity theory we learn that no speed at which signals or
energy travel can be faster than that of light. However, signals or energy cannot be
transmitted down a guide at speeds exceeding c; they travel with speed Vgr which is
always less than c so there is no conflict with the theory of relativity.
Sec. 39-4 RADIATION 979
The wavelength X in Eqs. 39-1 and 39-2 is the wavelength that would be measured
for the oscillations in free space, that is,
c
A=-, (39-3)
v
where c is the speed in free space and v is the frequency. For waves of a given fre-
quency, the wavelength exhibited in a guide (Ag) must differ from the free-space
wavelength A because the speed Vph has changed. The so-called guide wavelength Ag
is given by
A = Vph = Vph = A Vph.
g v c/'t. c
From Eq. 39-1 this yields
(39-4)
Thus the guide wavelength, which is the wavelength exhibited by the field patterns
in Fig. 39-7, is larger than the free-space wavelength .
.... Example 4. What must be the width a of a rectangular guide such that the
energy of electromagnetic radiation whose free-space wavelength is 3.0 cm travels
down the guide (a) at 95% of the speed of light? (b) At 50% of the speed of light?
From Eq. 39-2 we have
Vgr c
= 0.95c = ~l - (:aY·
Solving for a yields a = 4.8 cm; repeating for Vgr = 0.50c yields a = 1.7 cm.
This formula illustrates the cutoff phenomenon described above. If X = 2a, then
Vgr = 0 and energy cannot travel down the guide. For the radiation considered in
this example A = 3.0 cm, so that the guide must have a width a of at least ix 3.0 cm =
1.5 cm if it is to transmit this wave. The guide whose width we calculated in (a)
above can transmit radiations whose free-space wavelength is 2 x 4.8 cm = 9.6 cm
or less. <1111
39-4 Radiation
The acoustic transmission line of Fig. 39-6 cannot be infinitely long. Its
far end may be sealed by a solid cap or left open, or it may have a flange,
a horn, or some similar device mounted on it. If the far end is not sealed,
energy will escape into the medium beyond. This is called acoustic radiation.
In general, some energy will also be reflected back down the transmission
line. If acoustic radiation is desirable, the designer's .task is to fashion a
termination (that is, an "acoustic antenna") for the transmission line such
that the smallest possible fraction of the incident energy will be reflected
back down the line. Such a termination might take the form of a flared
horn. Acoustic radiation, of course, requires a medium such as air in order
to be propagated.
An electromagnetic transmission line such as a coaxial cable or a wave-
guide can also be terminated in many ways, and energy can escape from the
end of the line into the space beyond. In contrast to sound waves, a physical
980 ELECTROMAGNETIC WA YES Chap. 39
E
+Q
l j r
V+= Q V+++
i=0
v_ f-
p
p
t= 0 t=tT
(a) (b)
V+i f-
j '(
v_f
+
- j !p
v_
V+++
p
•
p
t=!T
(c)
t=iT
(d)
Fig. 39-9 Showing how radiation escapes from an oscillating electric dipole. The veloci-
ties of the charges, the electric dipole moment, the equivalent current, and the lines of E
are shown at four stages of oscillation, one-eighth of a period apart. The lines of B are not
shown. (Adapted from Attwood, Electric and Magnetic Fields, John Wiley and Sons,
third edition, 1949.)
Sec. 39-4 RADIATION 981
where w, the angular frequency, and k, the wave number, are related to the
frequency v and the wavelength")<.. by
and k = 21r/A.
982 ELECTROMAGNETIC WAVES Chap. 39
39-5 Traveling Waves and Maxwell's Equations
In earlier sections we have postulated the existence of certain magnetic
and electric field distributions, in resonant cavities, coaxial cables, and wave-
guides, and we have shown that these postulated distributions are consist-
ent with Maxwell's equations, as are the distributions of conduction and
displacement currents associated with the fields. The student who pursues
his studies of electromagnetism will learn how to derive mathematical ex-
pressions for E and B by subjecting Maxwell's equations to the boundary
conditions appropriate to the problem at hand. In this section we continue
our program by showing that the postulated patterns of E and B for a
traveling electromagnetic wave are completely consistent with Maxwell',s
equations. In doing so, we will be able to show that the speed of such waves
in free space is that of visible light and thus that visible light is itself an
electromagnetic wave.
If the observer at P in Fig. 39-9d is a considerable distance from the
source, the wavefronts described by the electric and magnetic fields that
reach him (see Fig. 39-10) will be planes and the wave that moves past him
will be a plane wave (see Section 19-2). Figure 39-11 shows a "snapshot"
of a plane wave traveling in the x direction. The lines of E are parallel to
the z axis and those of B are parallel to the y axis. The values of B and E
for this wave depend only on x and t (not on y or z). We postulate that
they are given in magnitude by
B = Bm sin (kx - wt) (39-6)
and E = Em sin (kx - wt). (39-7)
Figure 39-12 shows two sections through the three-dimensional diagram of
Fig. 39-11. In Fig. 39-12a the plane of the page is the xz plane and in
Fig. 39-11 A plane electromagnetic wave traveling to the right at speed c. Lines of B
are parallel to they axis; those of E are parallel to the z axis. The shaded rectangles on
the right refer to Fig. 39-12.
Sec. 39-5 TRAVELING WAVES AND MAXWELL'S EQUATIONS 983
a
z E
E,//~+~
. .. .
) • x x
x x' • • h x"' x
x x
x x .. • x
x
x
x
(a)
{E
c......
y B (B+dB
~ ~.. .
x x
x x x x
... .
x x ~ ~ x
h ~
.. . x
x
x
x
x
x
I
x
x
x
x
x
__J l-ax
(b)
Fig. 39-12 The wave of Fig. 39-11 viewed (a) in the xz plane and (b) in the xy plane.
Fig. 39-12b it is the xy plane. Note that, as for the traveling waves in a
coaxial cable (Fig. 39-3a) and in a waveguide (Fig. 39-7), E and B are in
phase, that is, at any point through which the wave is moving they reach
their maximum values at the same time.
The shaded rectangle of dimensions dx and h in Fig. 39-12a is fixed in
space. As the wave passes over it, the magnetic flux <flB through the rectangle
will change, which will give rise to induced electric fields around the rectangle,
according to Faraday's law of induction. These induced electric fields are,
in fact, simply the electric component of the traveling wave.
Let us apply Lenz's law to this induction process. The flux <flB for the
shaded rectangle of Fig. 39-12a is decreasing with time because the wave is
moving through the rectangle to the right and a region of weaker magnetic
fields is moving into the rectangle. The induced field will act to oppose this
change, which means that if we imagine that the boundary of the rectangle
is a conducting loop a counterclockwise induced current would appear in it.
This current would produce a field of B that, within the rectangle, would
point out of the page, thus opposing the decrease in <flB. There is, of course,
no conducting loop, but the net induced electric field E does indeed act
counterclockwise around the rectangle because E + dE, the magnitude of
Eat the right edge of the rectangle, is greater than E, the magnitude of E
at the left edge. Thus the electric field configuration is entirely consistent
with the concept that it is induced by the changing magnetic field.
984 ELECTROMAGNETIC WAVES Chap. 39
For a more detailed analysis let us apply Faraday's law of induction, or
see that the ratio of amplitudes is the same as the ratio of instantaneous
values, or
E = cB. (39-llb)
This important result will be useful in later sections.
We now turn our attention to Fig. 39-12b, in which the flux <I>E for the
shaded rectangle is decreasing with time as the wave moves through it.
According to Maxwell's third equation (with i = 0, because there are no
conduction currents in a traveling electromagnetic wave),
f B·dl =
d<l>E
µoEo - ,
dt
(39-12)
this changing flnx will induce a magnetic field at points around the periphery
of the rectangle. This induced magnetic field is simply the magnetic com-
ponent of the electromagnetic wave. Thus, as in the cavity resonator of
Section 38-10, the electric and the magnetic components of the wave are
intimately connected with each other, each depending on the time rate of
change of the other.
Comparison of the shaded rectangles in Fig. 39-12 shows that for each
the appropriate flux, <I>s or <I>E, is decreasing with time. However, if we pro-
ceed counterclockwise around the upper and lower shaded rectangles, we
see that f E·dl is positive, whereas f B·dl is negative. This is as it should
be. If the student will compare Figs. 35-10 and 38-9a, he will be reminded
that although the fluxes <I>s and <l>E in those figures are changing with time
in the same way (both are increasing) the lines of the induced fields, E and
B, respectively, circulate in opposite directions.
The integral in Eq. 39-12, evaluated by proceeding counterclockwise
around the shaded rectangle of Fig. 39-126, is
-h dB = µ 0 E0 ( h dx dd~)
or, substituting partial derivatives,
aB aE
- - = µoEo-· (39-13)
ax at
986 ELECTROMAGNETIC WAVES Chap. 39
Again, the minus sign in this equation is appropriate and necessary, for,
although B is increasing with x at the site of the shaded rectangle in Fig.
39-12b, E is decreasing with t.
Combining this equation with Eqs. 39-6 and 39-7 yields
-kBm cos (kx - wt) = -µoEowEm cos (kx - wt),
k 1
or (see Eq. 39-5) (39-14)
by a vector S, called the Poynting vector after John Henry Poynting (1852-
1914), who first pointed out its properties. We define S from
1
S = -ExB. (39-17)
µo
In the mks system S is expressed in watts/meter 2 ; the direction of S gives
the direction in which the energy moves. The vectors E and B ref er to
their instantaneous values at the point in question. If Eq. 39-17 is applied
to the traveling plane electromagnetic wave of Fig. 39-11, it is clear that
E x B, hence S, point in the direction of propagation. Note, too, that S
points parallel to the axis for all points in the coaxial cable of Fig. 39-3.
We get meaningful results if we extend the Poynting vector concept to
other electromagnetic situations involving either traveling or standing elec-
tromagnetic waves, as we will see in Examples 5 and 6. If we extend it to
circuit situations involving steady or almost steady currents and lumped
circuit elements, we are led to some interesting conclusions, which we ex-
plore in Problems 11, 17, and 18 .
.... Example 5. Analyze energy flow in the cavity of Fig. 38-8, using the Poynting
vector.
Study of Fig. 39-13 shows that when the energy is all electric (Figs. 39-13a and e)
it is concentrated along the axis, because this is the region in which E has its maximum
. . . .
111:.: ·:.:.:.:.:.:
(b) (d)
I \
(a) (e)
\ I
~
xx xxxx xxxx xx
x xx x x / x
xx ;xl ~
(h) (f)
. . ... ... . ..
................
(g)
Fig. 39-13 Example 5. Energy surges back and forth periodically between the central
region of the cavity and the region near the walls, as indicated by the Poynting vector S.
988 ELECTROMAGNETIC WA YES Chap. 39
value. When the energy is all magnetic (Figs. 39-13c and g), it is concentrated near
the walls. Thus the energy surges back and forth periodically between the central
region and the region near the walls. The figure shows, by the open arrows, the
direction of Sat various points in the cavity and at various times in the cycle. Note
that S equals zero for Figs. 39-13a, c, e, and g, which is appropriate because at these
instants of time the field configurations are momentarily stationary and energy
is not flowing. A pendulum bob at the end of its swing and at the bottom of its
trajectory forms a mechanical analogy. The student should verify from Eq. 39-17
that these arrows point in the correct directions. ....
Figure 39-14 can be used to derive the Poynting relation for the special case of a
traveling plane electromagnetic wave. It shows a cross section of a traveling plane
wave, along with a thin "box" of thickness dx and area A. The box, a mathematical
construction, is fixed with respect to the axes while the wave moves through it.
.. •
•
• •
•
x
'x I
r
x
x
X
x
•
• •
• •
• •
I x
x x
x
x
x
x
7
.
• • x
x
x • •
•
x x
x
x
-J k---dx
Fig. 39-14 A plane wave is traveling to the right at speed c; compare Fig. 39-12b. The
dashed rectangle in this figure represents a three-dimensional box, of area A and thickness
dx, that extends at right angles to the plane of the figure.
At any instant the energy stored in the box, from Eqs. 30-27 and 36-19, is
where A dx is the volume of the box and E and B are the instantaneous values of the
field vectors in the box.
Using Eq. 39-llb (E = cB) to eliminate one of the E's in the first term in Eq. 39-18
and one of the B's in the second term leads to
dU = [ !eoE(cB) + 2~ B 0
(~) J A dx
(µoeoc 2 + l)(EBA dx)
2µoc
dU = EBA dx
µoc
Sec. 39-6 THE POYNTING VECTOR 989
This energy dU will pass through the right face of the box in a time dt equal to
dx/c. Thus the energy per unit area per unit time, which is S, is given by
S = dU = EBA dx = J_ EB
dt A µoc (dx/c) A µo ·
This is exactly the prediction of the more general relation Eq. 39-17 for a traveling
plane wave.
This relation refers to values of S, E, and Bat any instant of time. We are usually
more interested in the average value of S, taken over one or more cycles of the wave.
An observer making intensity measurements on a wave moving past him would
measure this average value S. We can easily show (see Example 6) that Sis related
to the maximum values of E and B by
.... Example 6. An observer is at a great distance r from a point light source whose
. power output is P 0 • Calculate the magnitudes of the electric and the magnetic fields.
Assume that the source is monochromatic, that it radiates uniformly in all directions,
and that at distant points it behaves like the traveling plane wave of Fig. 39-11.
The power that passes through a sphere of radius r is (S) (47rr2 ), where S is the average
value of the Poynting vector at the surface of the sphere. This power must equal Po,
or
S = _I E2.
µoc
The average value of E 2 over one cycle is !Em 2 , since E varies sinusoidally (see Eq.
39-7). This leads to
or
= 240 volts/meter.
The relationship Em = cBm (Eq. 39-lla) leads to
_ Em _ 240 volts/meter _ _
Bm -
c
- 3 X 108 t I
me ers sec
- 8 X 10 7 weber/meter.2
990 ELECTROMAGNETIC WA YES Chap. 39
Note that Em is appreciable as judged by ordinary laboratory standards but that Bm
(= 0.008 gauss) is quite small. .....
QUESTIONS
1. In the coaxial cable of Fig. 39-1, what are the directions of the conduction current
(a) in the central conductor and (b) in the outer conductor, shortly after the switch is
thrown to position a? Consider points that have been reached by the wavefront of Fig.
39-2a and b and those that have not.
2. Compare a coaxial cable and a waveguide, used as a transmission line. Point out
both similarities and differences.
3. What is the relation between the wavelength in the cable and that in free space for a
coaxial cable?
4. Can traveling waves with a continuous range of wavelengths be sent down (a) a
coaxial cable and (b) a waveguide? Can standing waves with a continuous range of wave-
lengths be set up in a resonant cavity? Develop mechanical or acoustical analogies to
support your answers.
5. If a certain wavelength is larger than the cutoff wavelength for a guide in its dominant
mode, can energy be sent down it in any other mode?
6. Explain why the term fO dif>Eldt is needed in Ampere's equation to understand the
propagation of electromagnetic ,vaves.
7. In the equation c = 1 / ~ (Eq. 39-15), how can c always have the same value if
/.lo is arbitrarily assigned and fO is measured?
8. Is it conceivable that electromagnetic theory might some day be able to predict the
value of c (3 X 108 meters/sec), not in terms of µ 0 and fO, but directly and numerically
without recourse to any measurements?
9. What is the direction of the displacement current in Fig. 39--4? Give an argument
to support your answer.
10. In a coaxial cable is the energy transported in the conductors, through the agency
of the currents, or in the space between them, through the agency of the fields?
PROBLEMS
1. Using Gauss's law, sketch the instantaneous charges that appear on the conductors
of the coaxial cable of Fig. 39-3 and show that this pattern of charges is appropriately
related to the conduction currents shown in Fig. 39-3c.
2. For a rectangular guide of width 3.0 cm, plot the phase speed, the group speed, and
the guide wavelength as a function of the free-space wavelength. Assume the dominant
mode.
3. For a rectangular guirl.e of width 3.0 cm, what must the free-space wavelength of
radiation be if it is to require 1.0 µsec ( = 10-6 sec) for energy to traverse a 100-meter
length of guide? What is the phase speed under these circumstances?
4. Under what conditions will the guide wavelength in the guide of Fig. 39-7 be double
the free-space wavelength?
··. 5. How does the displacement current vary with space and time in a traveling plane
electromagnetic wave?
6. Prove that for any point in an electromagnetic wave such as that of Fig. 39-11 the
density of energy stored in the electric field equals that stored in the magnetic field.
7. A resonant cavity is constructed by closing each end of the coaxial cable of Fig. 39-3
with a metal cap. The cavity contains three half-waves. Describe the patterns of E and B
Chap. 39 PROBLEMS 991
that occur, assuming the same mode of oscillation as that shown in Fig. 39-3. (Hint:
Remember that E can have no tangential component at a conducting surface and that
Band E must be 90° out of phase.)
8. If a coaxial cable has resistance, energy must flow from the fields into the conducting
surfaces to provide the Joule heating. How must the electric lines of force of Fig. 39-3a
be modified in this case? (Hint: The Poynting vector near the surface must have a com-
ponent pointing toward the surface.)
9. What guide wavelength does 10-cm radiation (free-space wavelength) exhibit in a
rectangular guide whose width is 6.0 cm? Assume the dominant mode. What is the cutoff
wavelength for this guide?
10. Sketch five more figures to complete the sequence of Fig. 39-9, showing radiation
from an oscillating dipole. Include an indication of the lines of B in your drawings.
11. Figure 39-15 shows a long resistanceless transmission line, delivering power from a
battery to a resistive load. A steady current i exists as shown. (a) Sketch qualitatively
the electric and magnetic fields around the line, and (b) show that, according to the
Poynting vector point of view, energy travels from the battery to the resistor through the
space around the line and not through the line itself. (Hint: Each conductor in the line
is an equipotential surface, since the line has been assumed to have no resistance.)
f,1+
),
R~
--E
i
Fig. 39-15
12. Analyze the flow of energy in the waveguide of Fig. 39-7, using the Poynting
vector.
13. A cube of edge a has its edges parallel to the x-, y-, and z-axes of a rectangular
coordinate system. A uniform electric field E is parallel to the y-axis and a uniform
magnetic field B is parallel to the x-axis. Calculate (a) the rate at which, according to
the Poynting vector point of view, energy may be said to pass through each face of the
cube and (b) the net rate at which the energy stored in the cube may be said to change.
~4.: A #10 copper wire (diameter, 0.10 in.; resistance per 1000 ft, 1.00 ohm) carries a
current of 25 amp. Calculate E, B, and S for a point on the surface of the wire.
15. Sunlight strikes the earth, outside its atmosphere, with an intensity of 2.0 cal/cm 2-
min. Calculate Em and Bm for sunlight, assuming it to be a wave like that of Fig. 39-10.
16. A plane radio wave has Em "' 10-4 volt/meter. Calculate (a) Bm and (b) the in-
tensity of the wave, as measured by S.
· '·17- Figure 39-16 shows a cylindrical resistor of length l, radius a, and resistivity p,
carrying a current i. (a) Show that the Poynting vector Sat the surface of the resistor is
everywhere directed normal to the surface, as shown. (b) Show that the rate Pat which
energy flows into the resistor through its cylindrical surface, calculated by integrating the
Poynting vector over this surface, is equal to the rate at which Joule heat is produced;
that is,
-J S·dA = i 2R,
where dA is an element of area of the cylindrical surface. This shows that, according to
the Poynting vector point of view, the energy that appears in a resistor as Joule heat does
not enter it through the connecting wires but through the space around the wires and the
resistor. Note that S and dA are oppositely directed. (Hint: E is parallel to the axis
of the cylinder, in the direction of the current; B forms concentric circles around the
cylinder, in a direction given by the right-hand rule.)
992 ELECTROMAGNETI C WAVES Chap. 39
s s d
where ad is the volume of the capacitor and !~ 2 is the energy density for all points within
that volume. This analysis shows that, according to the Poynting vector point of view,
the energy stored in a capacitor does not enter it through the wires but through the space
around the wires and the plates. (Hint: To find S we must first find B, which is the
magnetic field set up by the displacement current during the charging process; see Fig.
38-9. Ignore fringing of the lines of E.)
Nature and Propagation of Light
CHAPTER 40
Frequency, cycles/sec
102 104 106 108 1010 1012 1014 1016 1018 1020 1022
106 104 102 1 10-2 10-4 10-6 10-8 10-10 10-12 10-14
Wavelength, meters
Fig. 40-1 The electromagnetic spectrum. Note that the wavelength and frequency
scales are logarithmic.
* For a report of electromagnetic waves with wavelengths as long as 1.9 X 107 miles
the student should consult an article by James Heirtzler in the Scientific American for
March 1962.
993
994 NATURE AND PROPAGATION OF LIGHT Chap. 40
80
->,
>
:;:::
'vi 60
c:
Q)
(J)
Q)
>
:;::: 40
"'
al
0:::
20
0
400 450 500 550 600 650 700
Wavelength, mµ
Fig. 40-2 The relative eye sensitivity of an assumed standard observer at different wave-
lengths for normal levels of illumination. The shaded areas represent the (continuously
graded) color sensations for normal vision.
"Light" is defined here as radiation that can affect the eye. Figure 40-2,
which shows the relative eye sensitivity of an assumed standard observer to
radiations of various wavelengths, shows that the center of the visible region
is about 5.55 X 10-7 meter. Light of this wavelength produces the sensa-
tion of yellow-green.*
In optics we often use the micron (abbr. µ) the millimicron (abbr. mµ),
and the Angstrom (abbr. A) as units of wavelength. They are defined from
1 µ = 10-5 meter
1 mµ = 10-9 meter
1 A = 10-10 meter.
Thus the center of the visible region can be expressed as 0.555 µ, 555 mµ, or
5550A.
The limits of the visible spectrum are not well defined because the eye
sensitivity curve approaches the axis asymptotically at both long and short
wavelengths. If the limits are taken, arbitrarily, as the wavelengths at
which the eye sensitivity has dropped to 1% of its maximum value, these
limits are about 4300 A and 6900 A, less than a factor of two in wavelength.
The eye can detect radiation beyond these limits if it is intense enough. In
many experiments in physics one can use photographic plates or light-sensi-
tive electronic detectors in place of the human eye.
* See "Experiments in Color Vision" by Edwin H. Land, Scientific American, May
1959, and especially "Color and Perception: the Work of Edwin Land in the Light of
Current Concepts" by M. H. Wilson and R. W. Brocklebank, Contemporary Physics,
December 1961, for a fascinating discussion of the problems of perception and the distinc-
tion between color as a characteristic of light and color as a perceived property of objects.
Sec. 40-2 ENERGY AND MOMENTUM 995
where E and Bare the instantaneous values of the electric and magnetic field
vectors.
Less familiar is the fact that electromagnetic waves may also transport
linear momentum. In other words, it is possible to exert a pressure (a
radiation pressure *) on an object by shining a light on it. Such forces must
be small in relation to forces of our daily experience because we do not
ordinarily notice them. The first measurement of radiation pressure was
made in 1901-1903 by Nichols and Hull in this country and by Lebedev in
Russia, about thirty years after the existence of such effects had been pre-
dicted theoretically by Maxwell.
Let a parallel beam of light fall on an object for a time t, the incident light
being entirely absorbed by the object. If energy U is absorbed during this
time, the momentum p delivered to the object is given, according to Max-
well's prediction, by
p
u
= - (total absorption), (40-2a)
c
where c is the speed of light. The direction of p is the direction of the in-
cident beam. If the light energy U is entirely reflected, the momentum de-
livered will be twice that given above, or
2U
p=- (total reflection). (40-2b)
c
In the same way, twice as much momentum is delivered to an object when
a perfectly elastic tennis baJl is bounced from it as when it is struck by a
perfectly inelastic ball of the same mass and speed. If the light energy U
is partly reflected and partly absorbed, the delivered momentum will lie
between U /c and 2U /c.
~ Example 1. A parallel beam of light with an energy flux S of 10 watts/cm 2 falls
for 1 hr on a perfectly reflecting plane mirror of 1.0-cm2 area. (a) What momentum
is delivered to the mirror in this time and (b) what force acts on the mirror?
(a) The energy that is reflected from the mirror is
C = (10 watts/cm 2)(1.0 cm 2)(3600 sec) = 3.6 X 104 joules.
The momentum delivered after 1 hr's illumination is
_ 2U _ (2)(3.6 X 104 joules) _ 2 4
P - -- -
_ k I
- . X 10 4 g-m sec.
c 3 X 10 8 meters/sec
* See "Radiation Pressure," G. E. Henry, Scientific American, p. 99, June 1957.
996 NATURE AND PROPAGATION OF LIGHT Chap. 40
(b) From Newton's second law, the average force on the mirror is equal to the
average rate at which momentum is delivered to the mirror, or
t 3600 sec
This is a small force.
Torsion
suspension
Fig. 40-3
Mirrors
Incident
light
beam ef/
M
M'
Sec. 40-2 ENERGY AND MOMENTUM 997
y
large thin sheet of a material of high resistivity as in Fig. 40-4. A small part of the
incident energy will be absorbed within the sheet, but most of it will be transmitted
if the sheet is thin enough.*
The incident wave E and B vary with time at the sheet as
E = Em sin wt (4v-3)
and B = Bm sin wt (40-4)
where E is parallel to the ±y axis and B is parallel to the ±x axis.
In Section 31-4 we saw that the effect of a (constant) electric force ( = -eE) on a
conduction electron in a metal was to make it move with a (constant) drift speed vd.
The electron behaves as if it is immersed in a viscous fluid, the electric force acting on
it being counterbalanced by a "viscous" force, which may be taken as proportional
to the electron speed. Thus for a constant field E, after equilibrium is established,
eE = bvd, (40-5)
where bis a resistive damping coefficient. The electron equilibrium speed, dropping
the subscript d, is thus
eE
v = -· (40-6)
b
If the applied electric field varies with time and if the variation is slow enough,
the electron speed can continually readjust itself to the changing value of E so that its
speed continues to be given essentially by its equilibrium value (Eq. 40-6) at all
times. These readjustments are more rapidly made the more viscous the medium,
just as a stone falling in air reaches a constant equilibrium rate of descent only rela-
tively slowly but one falling in a viscous oil does so quite rapidly. We assume that the
sheet in Fig. 40-4 is so viscous, that is, that its resistivity is so high, that Eq. 40-6
remains valid even for the rapid oscillations of E in the incident light beam.
As the electron vibrates parallel to the y axis, it experiences a second force due to
the magnetic component of the wave. This force F z ( = -ev x B) points in the z
direction, being at right angles to the plane formed by v and B, that is, the xy plane.
The instantaneous magnitude of F 2 is given by
e2EB
F-• = evB = - b- · (40-7)
* Some of the incident energy will also be reflected, but the reflected wave is of such low
intensity that it can be ignored in the derivation that follows; see Optics by B. Rossi,
Addison-Wesley Publishing Company, p. 411, 1957, from which this derivation is adapted.
998 NATURE AND PROPAGATION OF LIGHT Chap .. 40
Fz always points in the positive z direction because v and B reverse their directions
simultaneously; this force is, in fact, the mechanism by which the radiation pressure
acts on the sheet of Fig. 40-4.
From Newton's second law, Fz is the rate dpe/dt at which the incident wave de-
livers momentum to each electron in the sheet, or
dpe = e2EB. (40-8)
dt b
Momentum is delivered at this rate to every electron in the sheet and thus to the
sheet itself. It remains to relate the momentum transfer to the sheet to the absorption
of energy within the sheet.
The electric field component of the incident wave does work on each oscillating
electron at an instantaneous rate (see Eq. 40-6) given by
no work on the oscillating electron. Equation 39- llb shows that for a plane wave in
free space B and E are related by
E = Be.
Substituting above for one of the E's leads to
dUe e2 EBc
- - = ---· (40-9)
dt b
This equation represents the rate, per electron, at which energy is absorbed from the
incident wave.
Comparing Eqs. 40-8 and 40-9 shows that
dpe = ~ dU_!_.
dt c dt
dt = 2 it?:_!/_!_ dt
Integrating yields i t dpe
O dt C dt
O '
Ue
or Pe=--, (40-10)
c
where Pe is the momentum delivered to a single electron in any given time t and Ue
is the energy absorbed by that electron in the same time interval. Multiplying each
side by the number of free electrons in the sheet leads to Eq. 40-2a.
Although we derived this relation (Eq. 40-10) for a particular kind of absorber,
no characteristics of the absorber-for example, the resistive damping coefficient
b-remain in the final expression. This is as it should be because Eq. 40-10 is a
general property of radiation absorbed by any material.
* See "The Speed of Light," J. H. Rush, Scientific American, p. 67, August 1955.
Sec. 40-3 THE SPEED OF LIGHT 999
Toothed
wheel
M;no:f<:f= =========Lt2:::>t~;::Lt4==:
-1.- - - - - z - ~
s
Fig. 40-5 Fizeau's apparatus for measuring the speed of light.
8630 meters or .5.36 miles. When the light strikes mirror Af 1 again, some will
be transmitted, entering the eye of the observer through lens L 4 .
The observer will see an image of the source formed by light that has
traveled a distance 2Z between the wheel and mirror M 2 and back again. To
time the light beam a marker of some sort must be put on it. This is done by
"chopping" it with a rapidly rotating toothed wheel. Suppose that during
the round-trip travel time of 2Z/c the wheel has turned just enough so that,
when the light from a given "burst" returns to the wheel, point Fis covered
by a tooth. The light will hit the face of the tooth that is toward M 2 and
will not reach the observer's eye.
If the speed of the wheel is exactly right, the observer will not see any of
the bursts because each will be screened by a tooth. The observer measures
c by increasing the angular speed w of the wheel from zero until the image of
source S disappears. Let e be the angular distance from the center of a gap
to the center of a tooth. The time -needed for the wheel to rotate a distance
e is the round-trip travel time 2Z/c. In equation form,
e 2wl
or c = --· (40-11)
w c e
This "chopped beam" technique, suitably modified, is used today to measure
the speeds of neutrons and other particles .
..... Example 2. The wheel used by Fizeau had 720 teeth. What is the smallest angu-
lar speed at which the image of the source will vanish?
The angle eis 1/1440 rev; solving Eq. 40-11 for w gives
ce
w = - =
(3.00 X 10 8 meters/sec)(l/1440 rev)
12.1 rev/sec.
...,.
""'Ill
2l (2)(8630 meters)
The French physicist Foucault (1819-1868) greatly improved Fizeau's
method by substituting a rotating mirror for the toothed wheel. The Amer-
ican physicist Albert A. Michelson (1852-1931) conducted an extensive
series of measurements of c, extending over a fifty-year period, using this
technique.
Sec. 40-3 THE SPEED OF LIGHT 1001
Table 40-1
Uncer-
Speed,
Date Experimenter Country Method tainty,
km/sec
km/sec
which is the same relationship used for acoustic waves in closed pipes; n ( = 3 for Fig.
40-6) gives the number of half-waves contained in the cavity.
I I
~I . ~
l~-1
I
I EI
I
;
~
I
j~
(a)
Fig. 40-6 A resonant cavity con-
structed from a section of waveguide;
compare Fig. 39-7. For simplicity
the lines of E are not shown in (a) and
those of B are not shown in (b).
I f mn f ! t nn6 ! I fIDf f 1 (b)
I I
I I
Sec. 40-4 MOVING SOURCES AND OBSERVERS 1003
The procedure is to measure Ag for such a cavity, which has been tuned to resonance,
and then, using Eq. 39-4,
Ag= A ,
~l - (tJ2 (40-13)
calculate the free-space wavelength A. From the measured resonant frequency, the
speed c can be found from c = Av.
.... Example 3. Essen of the National Physical Laboratory in England made a
resonant cavity measurement of the speed of electromagnetic waves. His cavity was
made of a circular waveguide rather than a rectangular one; it can be shown that, for
the oscillation pattern used by him, the geometrical factor 2a in Eq. 40-13 must be
replaced by l.64062R, where R is the guide radius. The cavity radius was 3.25876
cm; the cavity length was 15.64574 cm and it proved to resonate at 9.498300 X 109
cycles/sec. At resonance it was determined that there were eight half-waves in the
cavity. What value of c results?
From Eq. 40-12, computing only an approximate result,
, _ 2l _ (2)(15.6 cm) _ 3
Ag - n - 8 - .90 cm.
Substituting into Eq. 40-13, suitably modified for a circular waveguide, yields
3.90 cm= / .
For practical reasons Essen analyzed his data in a more roundabout way than that
given. His final result, based on many measurements under different conditions and
carried to much greater accuracy than that illustrated in the above example, was
2.997925 X 108 meters/sec with an uncertainty of 0.000030 X 108 meters/sec. <111111
the speed of the earth through the ether always gave the result of zero.*
Physicists were not willing to believe that the earth was permanently at
rest in the ether and that all other bodies in the universe were in motion
through it. Other hypotheses about the nature of the propagation of light
also proved unsatisfactory for one reason or another.
Einstein in 1905 resolved the difficulty of understanding the propagation
of light by making a bold postulate: If a number of observers are moving
(at uniform velocity) with respect to each other and to a source of light and
if each observer measures the speed of the light emerging from the source,
they will all obtain the same value. This is the fundamental assumption of
Einstein's theory of relativity. It does away with the need for an ether by
asserting that the speed of light is the same in all reference frames; none is
singled out as fundamental. The theory of relativity, derived from this
postulate, has been subject to many experimental tests, from which agree-
ment with the predictions of theory has always emerged. These agreements,
extending over half a century, lend strong support to Einstein's basic
postulate about light propagation.
Figure 40-7 focuses specifically on the fundamental problem of light
propagation. A source of light, at rest in reference frame S', emits a light
pulse P whose speed v' is measured by an observer at rest in this same frame.
From the point of view of an observer in reference frame S, frame S' and its
associated observer are moving in the positive x direction at speed u. Ques-
tion: What speed v would observer S measure for the light pulse P? Ein-
stein's hypothesis asserts that each observer would measure the same speed
c, or that
v = v' = c.
v = v' + u. (40-14)
* See Section 43-7, which describes the crucial experiment of Michelson and Morley.
Sec. 40-4 MOVING SOURCES AND OBSERVERS 1005
a speed that is quite small compared to that of light. The body of experi-
mental evidence that underlies Eq. 40-14 thus represents a severely re-
stricted area of experience, namely, experiences in which v' << c and u « c.
If we assume that Eq. 40-14 holds for all particles regardless of speed, we
are making a gross extrapolation. Einstein's theory of relativity predicts
that this extrapolation is indeeJ not valid and that Eq. 40-14 is a limiting
case of a more general relationship that holds for light pulses and for ma-
terial particles, whatever their speed, or
v =
v' +u . (40-15)
1 + v'u/c2
Equation 40-15 is quite indistinguishable from Eq. 40-14 at low speeds,
that is, when v' « c and u « c; see Example 4.
If we apply Eq. 40-15 to the case in which the moving object is a light
pulse, and if we put v' = c, we obtain
c+u
v = = c.
1 + cu/c2
This is consistent, as it must be, with the fundamental assumption on which
the derivation of Eq. 40-15 is based; it shows that both observers measure
the same speed c for light. Equation 40-14 predicts (incorrectly) that the
speed measured by S will be c + u. Figure 40-8 shows that the (correct)
Eq. 40-15 and the (approximate) Eq. 40-14 cannot be distinguished from
each other at speeds that are small compared to the speed of light.
0.8
0.6 .,.g, ~
--
'-'
:::,
~
,.:
.§
'lt
~
~
-::5
0.4 ~
& J;>.....
t::
r.J
0.2
0
0 0.2 0.4 0.6 0.8 1.0
u/c
Fig. 40-8 The speed of a particle P, as seen by observer S in Fig. 40-7, for the special
case of v' = u. All speeds are expressed as a ratio to c, the speed of light. The vertical
arrow corresponds to 5 X 107 miles/hr.
1006 NATURE AND PROPAGATION OF LIGHT Chap. 40
IJJII- Example 4. Suppose that v' = u = 25,000 miles/hr. What per cent error is
made in using Eq. 40-14 rather than Eq. 40-15 to calculate v? The speed of light is
6.7 X 108 miles/hr.
Equation 40-14 gives
50,000 miles/hr
1.0000000014
This example shows that for speeds that are comparable to c, Eqs. 40-14 and 40-15
yield rather different results. A wealth of indirect experimental evidence points to
the latter result as being correct. ....
v' =
1
1
v-~-
+ u/v
l l. sound wave
2. observer fixed in medium
3. source receding from observer
(40-16)
In this equation vis the frequency heard when the source is at rest and v is
the speed of sound.
Sec. 40-5 DOPPLER EFFECT 1007
1. sound wave
[ 2. source fixed in medium (40-17)
3. observer receding from source
Even if the relative separation speeds u of the source and the observer are
the same, the frequencies predicted by Eqs. 40-16 and 40-17 are different.
This is not surprising, because a sound source moving through a medium in
which the observer is at rest is physically different from an observer moving
through that medium with the source at rest, as comparison of Figs. 20-10
and 20-11 shows.
We might be tempted to apply Eqs. 40-16 and 40-17 to light, substituting
c, the speed of light, for v, the speed of sound. For light, as contrasted with
sound, however, it has proved impossible to identify a medium of transmis-
sion relative to which the source and the observer are moving. This means
that "source receding from observer" and "observer receding from source"
are physically identical situations and must exhibit exactly the same Doppler
frequency. As applied to light, either Eq. 40-16 or Eq. 40-17 or both must
be incorrect. The Doppler frequency predicted by the theory of relativity
is, in fact,
1. light wave
1 - u/c
v . 2. source and observer (40-18)
Vl - (u/c) 2 separating
vI v(1-:} (40-17a)
and v' = v [ 1 - : + ~ (: Y
+ ··· l (40-18a)
The ratio u/c for all available monochromatic light sources, even those of
atomic dimensions, is small. This means that successive terms in these
1008 NATURE AND PROPAGATION OF LIGHT Chap. 40
(40-19)
(see Fig. 40-9b). It measures the extent to which the frequency of the light from
resting atoms fails to lie halfway between the frequencies v1 ' and v2 '. Table 40-2
shows that the measured results agree with the formula predicted by the theory of
relativity (Eq. 40-lSa) and not with the classical formula borrowed from the theory of
sound propagation in a material medium (Eq. 40-16a).
Ives and Stilwell did not present their experimental results as evidence for the
support of Einstein's theory of relativity but rather gave them an alternative the-
oretical explanation. Modern observers, looking not only at their excellent experi-
ment but at the whole range of experimental evidence, now give the Ives-Stilwell
experiment the interpretation we have described for it above.
100 v 30 kv
.-------11111 1111111111111111
Mirror 1
~2u
~
Acceleration
arc region region (a)
(b)
Table 40-2
I
I
Theoretical value according to I
classical theory (Eq. 40-l 6a) 1.67 2.26 2.90 3.94
I
Theoretical value according to the
theory of relativity (Eq. 40-18a) 0.835 1.13 1.45 1.97
* See Eq. 40-19; the table shows only part of the data taken by Ives and Stilwell.
The Doppler effect for light finds many applications in astronomy, where
it is used to determine the speeds at which luminous heavenly bodies are
moving toward us or receding from us. Such Doppler shifts measure only
the radial or line-of-sight components of the relative velocity. All galaxies *
for which such measurements have been made (Fig. 40-10) appear to be
receding from us, the recession velocity being greater for the more distant
galaxies; these observations are the basis of the expanding-universe concept.
(a) (b)
Fig. 40-10 (a) The central spot is a nebula in the constellation Corona Borealis; it is
130,000,000 light years distant.· (b) The central streak shows the distribution in wave-
length of the light emitted from this nebula. The two vertical dark bands show the
presence of calcium. The horizontal arrow shows that these calcium lines occur at longer
wavelengths than those for terrestrial light sources containing calcium, the length of the
arrow representing the wavelength shift. Measurement of this shift shows that the galaxy
is receding from us at 13,400 miles/sec. The lines above and below the central streak
represent light from a terrestrial source, used to establish a wavelength scale. (Courtesy
Mount Wilson and Mount Palomar Observatories.)
* See "The Red-Shift," Allen R. Sandage, Scientific American, p. 171, September 1956.
1010 NATURE AND PROPAGATION OF LIGHT Chap. 40
~ Example 6. Certain characteristic wavelengths in the light from a galaxy in
the constellation Virgo are observed to be increased in wavelength, as compared with
terrestrial sources, by about 0.4%. What is the radial speed of this galaxy with
respect to the earth? Is it approaching or receding?
If A is the wavelength for a terrestrial source, then
X' = 1.004X.
v' = 0.996v.
This frequency shift is so small that, in calculating the source velocity, it makes no
practical difference whether we use Eq. 40-16, 40-17, or 40-18. Using Eq. 40-17
we obtain
v' = 0.996v = v ( 1 - ~) ·
QUESTIONS
PROBLEMS
1. (a) At what wavelengths does the eye sensitivity have half its maximum value?
(b) What are the frequency and the period of the light for which the eye is most sensitive?
2. It has been proposed that a spaceship might propel itself in the solar system by radi-
ation pressure, using a large sail made of aluminum foil. How large must the sail be if the
radiation force is to be equal in magnitude to the sun's gravitational attraction? Assume
that the mass of the ship + sail is 100 slugs, that the sail is perfectly reflecting, and that
the sail is oriented at right angles to the sun's rays. The sun's mass is 1.97 X 10 30 kg.
'--3. Radiation from the sun striking the earth has an intensity of 1400 watts/meter2 •
Assuming that the earth behaves like a flat disk at right angles to the sun's rays and that
all the incident energy is absorbed, calculate the force on the earth due to radiation
pressure. Compare it with the force due to the sun's gravitational attraction.
4. Prove, for a plane wave at normal incidence on a plane surface, that the radiation
pressure on the surface is equal to the energy density in the beam outside the surface.
This relation holds no matter what fraction of the incident energy is reflected.
5. Prove, for a stream of bullets striking a plane surface at right angles, that the "pres-
sure" is twice the (kinetic) energy density in the stream above the surface; assume that the
bullets are completely "absorbed"
by the surface. Contrast this with
the behavior of light (Problem 4).
6. A small spaceship whose mass,
with occupant, is 100 slugs is drift-
ing in outer space, where no gravi-
tational field exists. If it shines a
searchlight, which radiates 10 4 watts,
into space, what speed would the
ship attain in one day because of
the reaction force associated with
the momentum carried away by the
light beam?
7. What is the radiation pressure
l.Ometer away from a 500-watt light
bulb? Assume that the surface on
which the pressure is exerted faces
the bulb and is perfectly absorbing
and that the bulb radiates uniformly
in all directions.
8. The uncertainty of the distance
to the moon, as measured by the
reflection of radar waves from it, is
about 0.5 mile. Assuming that this
uncertainty is associated only with
the measurement of the elapsed
time, what uncertainty in this time
is implied?
·· 9. Roemer's method for measur-
ing the speed of light consisted in Fig. 40-11
observing the apparent times of re-
volution of one of the moons of Jupiter. The true period of revolution is 42.5 hr. (a)
Taking into account the finite speed of light, how would you expect the apparent time of
revolution to alter as the earth moves in its orbit from point x to point y in Fig. 40-11?
(b) What observations would be needed to compute the speed of light? Neglect the mo-
tion of .Jupiter in its orbit. Figure 40-11 is not drawn to scale.
1012 NATURE AND PROPAGATION OF LIGHT Chap. 40
10. Suppose that light is timed over a 1-mile base line and its speed is measured to the
accuracy quoted on p. 1002. How large an error in the length of the base line could be
tolerated, assuming other sources of error to be negligible?
11. For what value of u/c does Eq. 40-17 differ from Eq. 40-18 by 1 %?
=-~~ The "red shift" of radiation from a distant nebula consists of the light (H.y), known
to have a wavelength of 4340 X 10-s cm when observed in the laboratory, appearing to
have a wavelength of 6562 X 10-8 cm. (a) What is the speed of the nebula in the line of sight
relative to the earth? (b) Is it approaching or receding?
· t3. The difference in wavelength between an incident microwave beam and one re-
flected from an approaching or receding car is used to determine automobile speeds on the
highway. (a) Show that if vis the speed of the car and 11 the frequency of the incident
beam, the change of frequency is approximately 2v11/c, where c is the speed of the electro-
magnetic radiation. (b) For microwaves of frequency 2450 megacycles/sec, what is the
change of frequency per mile/hr of speed?
14. Show that, for slow speeds, the Doppler shift can be written in the approximate
form
d>.. u
-=-,
>.. c
where .:l>.. is the change in wavelength.
15. The period of rotation of the sun at its equator is 24.7 days; its radius is 7.0 X 108
meters. What Doppler wavelength shifts are expected for characteristic wavelengths in
the vicinity of 5500 A emitted from the edge of the sun's disk?
16. An earth satellite, transmitting on a frequency of 40 X 106 cycles/sec (exactly),
passes directly over a radio receiving station at an altitude of 250 miles and at a speed of
18,000 miles/hr. Plot the change in frequency, attributable to the Doppler effect, as a
function of time, counting t = 0 as the instant the satellite is over the station. (Hint:
The speed u in the Doppler formula is not the actual velocity of the satellite but its com-
ponent in the direction of the station. Use the nonrelativistic formula (Eq. 40--16a) and
neglect the curvature of the earth and of the satellite orbit.)
17. A rocketship is receding from the earth at a speed of 0.2c. A light in the rocketshlp
appears blue to passengers on the ship. What color would it appear to be to an observer
on the earth? See Fig. 40-2.
18. In the experiment of Ives and Stilwell the speed u of the hydrogen atoms in a par-
ticular run was 8.61 X 105 meters/sec. Calculate L'.111 1, L'.1112, and L'.111/11, on the assump-
tions that (a) Eq. 40--18a is correct and (b) that Eq. 40--16a is correct; compare your
results with those given in Table 40--2 for this speed. Retain the first three terms only
in Eqs. 40--18a and 40--16a.
Reflection and Refraction-
Plane Waves
and Plane Surfaces
CHAPTER 41
Fig. 41-1 (a) A photograph showing reflection and refraction at an air-water interface.
(b) A representation using rays.
1013
1014 REFLECTION AND REFRACTION Chap. 41
The laws governing reflection and refraction can easily be found from ex-
periment:
1. The reflected and the refracted rays lie in the plane formed by the in-
cident ray and the normal to the surface at the point of incidence, that is,
the plane of Fig. 41- lb.
2. For reflection:
(41-1)
3. For refraction:
sin 01
-.-- = n21, (41-2)
sm 02
where n 21 is a constant called the index of refraction of medium 2 with respect
to medium 1. Table 41-1 shows the indices of refraction for some common
Table 41-1
Index
Medium of
Refraction
Water 1.33
Ethyl alcohol 1.36
Carbon bisulfide 1.63
Air (1 atm and 20°C) 1.0003
Methylene iodide 1.74
Fused quartz 1.46
Glass, crown 1.52
Glass, dense flint 1.66
Sodium chloride 1.53
Polyethylene 1.50-1.54
Fluorite 1.43
* Measured with respect to a vacuum. The index with respect to air (except, of
course, the index of air itself-see item 4) will be negligibly different in most cases.
1.480
1.475.
c:
21.470
u
Fig. 41-2 The index of refrac- e
tion of fused quartz at 18°C, ~ 1.465
with respect to a vacuum. 0
x
Q)
-g 1.460
1.455
1.450
3000 4000 5000 6000 7000 8000
Wavelength, A (in air at STP)
The law of reflection was known to Euclid. That of refraction was dis-
covered experimentally by Willebrod Snell (1591-1626) and deduced from
the early corpuscular theory of light by Rene Descartes (1596-1650). The
law of refraetion is known as Snell's law or (in France) as Descartes' law.
The laws of reflection and refraction can be derived from Maxwell's
equations, which means that these laws should hold for all regions of the
electromagnetic spectrum. Figure 41-4a shows an experimental setup for in-
vestigating the reflection of microwaves from a large metal sheet. Figure
41-4b shows the reading of the detector as a function of the angular position
of the mirror. The existence of a reflected beam at the proper angle confirms
the law of reflection for microwaves. There is ample experimental evidence
that Eqs. 41-1 and 41-2 correctly describe the behavior of reflected and
refracted beams in all parts of the electromagnetic spectrum.
It is common knowledge that a polished steel surface will form a well-defined reflected
beam if an incident beam falls on it, but a sheet of paper wi.11 reflect light
more or less in all directions (diffuse reflection). It is largely by diffuse reflection that
we see nonluminous objects around us. The difference between diffuse and specular
(that is, mirror-like) reflection is a matter of surface roughness; a reflected beam will
Fig. 41-3 Sunlight from a hole Fin a screen is refracted by prism ABC, forming a spec-
trum on screen MN; from Newton's Opticks (1704).
1016 REFLECTION AND REFRACTION Chap. 41
a~!~~~a ~M:tal
( </> Ym1rror
~ - - - - --
M.IC Incident ray I
I
oscillator "'-.,ip ara bo11c
. l
I
reflector I
Reflected ray t
Microwave~
(a) detector LJ
~100
"C
ro
....
Q)
....
+"'
0
(.)
60
Q)
~ 40
"C
Q)
>
:;:::; 20
ro
Q)
a::
10 20 30 40 50 60 70 80 90
ct,, degrees
(b)
Fig. 41-4 (a) An apparatus to test the law of reflection for microwaves. (b) A reflected
beam (for "}-. ~ 10 cm) appears for the expected orientation of the mirror.
be formed only if the average depth of the surface irregularities of the reflector is
substantially less than the wavelength of the incident light. This criterion of surface
roughness has different implications in different regions of the electromagnetic spec-
trum. The bottom of a cast-iron skillet for example is a good reflector for microwaves
of wavelength 0.5 cm but it is not a good reflector for visible light (that is, one cannot
shave by it).
A second requirement for the existence of a reflected beam is that the transverse
dimensions of the reflector must be substantially larger than the wavelength of the
incident beam. If a beam of visible light falls on a polished metal disk the size of a
dime, a reflected beam will be formed. However, if the same disk is placed in a beam
of short radio waves with, say, 'A. = 1.0 meter, radiation will be scattered in all
directions from it, and no well-defined unidirectional beam will appear. We inves-
tigate this phenomenon of diffraction in Chapter 44. The requirements that surfaces
be "smooth" and "large" also apply to the formation of refracted beams. If these
two requirements are not met the description of reflection and refraction in terms of
rays, whose behavior is governed by Eqs. 41-1 and 41-2, is not valid .
.... Example 1. Figure 41-5 shows an incident ray i striking a plane mirror MM'
at angle of incidence 0. Trace this ray.
The reflected ray makes an angle () with the normal at b and falls as an incident ray
on mirror M' M". Its angle of incidence ()' on this mirror is 7r /2 - fJ. A second
Sec. 41-1 REFLECTION AND REFRACTION 1017
reflected ray r' makes an angle 0' with the normal erected at b'. Rays i and r' are
antiparallel for any value of 0. To see this, note that
¢ = 1r - 20' = 1r - 2 (~ - 0) = 20.
Two lines are parallel if their opposite interior angles for an intersecting line (¢ and 20)
are equal.
Repeat the problem if the angle between the mirrors is 120° rather than 90°.
Example 2. An incident ray in air falls on the plane surface of a block of quartz
and makes an angle of 30° with the normal. This beam contains two wavelengths,
4000 and 5000 A. The indices of refraction for quartz with respect to air (nqa) at
these wavelengths are 1.4702 and 1.4624, respectively. What is the angle between
the two refracted beams?
From Eq. 41-2 we have, for the 4000-A beam,
sin 01 = nqa sin 02,
or sin 30° = (1.4702) sin 02,
which leads to 02 = 19.88°.
For the 5000-A beam we have
sin 30° = (1.4624) sin 02',
or 02' = 19.99°.
The angle !:i.O between the beams is 0.11 °, the shorter wavelength component being
bent through the larger angle, that is, having the smaller angle of refraction.
1018 REFLECTION AND REFRACTION Chap. 41
Example 3. An incident ray falls on one face of a glass prism in air as in Fig. 41-6.
The angle eis so chosen that the emerging ray also makes an angle e with the normal
to the other face. Derive an expression for the index of refraction of the prism
material with respect to air.
Note that L abc = a, the two angles having their sides mutually perpendicular.
Therefore
a = !ct>, (41-3)
The deviation angle i/; is the sum of the two opposite interior angles in triangle
aed, or
iJ; = 2ce - a).
Substituting !ct> for a and solving for e yields
e = !Cf+ ct>). (41-4)
At point a, e is the angle of incidence and a the angle of refraction. The law of
refraction (see Eq. 41-2) is
sin e = nga sin a,
in which nga is the index of refraction of the glass with respect to air.
From Eqs. 41-3 and 41-4 this yields
. f+ct>
--
Slil - =
. ct>
nga Sln 2
2
or
sin !(iJ;+ </>)
nga = sin (cp/2)
which is the desired relation. This equation holds only fore so chosen that the light
ray passes symmetrically through the prism. For this condition the deviation angle
i/; is a minimum; if e is either increased or decreased, a larger deviation will be pro-
duced. .,..
doing so. Fortunately, these and several other laws of optics can be derived
on the basis of a simpler but less comprehensive theory of light, put forward
by the Dutch physicist Christian Huygens in 1678. This theory simply
assumes that light is a wave rather than, say, a stream of particles. It says
nothing about the nature of the wave and, in particular-since Maxwell's
theory of electromagnetism appeared only after the lapse of a century-
gives no hint of the electromagnetic character of light. Huygens did not
know whether light was a transverse wave or a longitudinal one; he did
not know the wavelengths of visible light; he had little knowledge of the
speed of light. K evertheless, his theory was a useful guide to experiment for
many years and remains useful today for pedagogic and certain other prac-
tical purposes. We must not expect it to yield the same wealth of detailed
information that Maxwell's more complete electromagnetic theory does.
Huygens' theory is based on a geometrical construction, called Huygens'
principle, that allows us to tell where a given wavefront will be at any time
in the future if we know its present position; it is: All points on a wavefront
can be considered as point sources for the production of spherical secondary
wavelets. After a time t the new p0sition of the wavefront will be the surface of
tangency to these secondary wavelets.
We illustrate this by a trivial example: Given a wavefront (ab in Fig. 41-7)
in a plane wave in free space, where will the wavefront be a time t later?
Following Huygens' principle, we let several points on this plane (see dots)
serve as centers for secondary spherical wavelets. In a time t the radius of
these spherical waves is ct, where c is the speed of light in free space. The
plane of tangency to these spheres at time t is represented by de. As we
(a)
(b)
(c)
(d)
Sec. 41-4 HUYGENS' PRINCIPLE AND THE LAW OF REFRACTION 1021
words, f)i is the angle of incidence. The three wavefronts are related to each
other by the Huygens construction, as in Fig. 41-7.
In Fig. 41-8b a Huygens wavelet centered on point a will expand to
include point l after a time ).../c. Light from point pin this same wavefront
cannot move beyond the mirror but must expand upward as a spherical
Huygens wavelet. Setting a compass to radius A and swinging an arc about
p provides a semicirrle to which the reflected wavefront must be tangent.
Since point l must lie on the new wavefront, this tangent must pass through l.
Note that the angle Bi' between the wavefront and the mirror is the same as
the angle between the reflected ray and the normal to the mirror. In other
words, 81 ' is the angle of reflection.
Consider right triangles alp and a'lp. They have side lp in common and
side al ( = )...) is equal to side a'p. The two right triangles are thus congruent
and we may conclude that
as required by the law of reflection. If the student recalls that the Huygens
construction is three-dimensional and that the arcs shown represent segments
of spherical surfaces, he will be able to convince himself that the reflected
ray lies in the plane formed by the incident ray and the normal to the mirror,
that is, the plane of Fig. 41-8. This is also a requirement of the law of re-
flection; see p. 1014.
during this time. The refracted wavefront must be tangent to an arc of this
radius centered on h. Since c lies on the new wavefront, the tangent must
pass through this point, as shown. K ote that 82 , the angle between the
refracted wavefront and the air-glass interface, is the same as the angle
1022 REFLECTION AND REFRACTION Chap. 41
(a)
(b)
between the refracted ray and the normal to this interface. In other words,
82 is the angle of refraction. Note, too, that the wavelength in glass ()\ 2 ) 1s
less than the wavelength in air (;>.. 1 ).
For the right triangles hce and hce' we may write
A1
sin 81 = - (for hce)
he
. A2
and sm 82 = - (for hce').
he
Dividing and using Eq. 41-6 yields
a constant. (41-7)
Sec. 41-4 HUYGENS' PRINCIPLE AND THE LA w OF REFRACTION 1023
sin 81
-.-- = n21, (41-2)
sm 82
so that n 21 is now revealed as the ratio of the speeds of light in the two
media, or
(41-8)
(41-9)
fn which c is the speed of light in free space. The quantities (c/v 1 ) and
(c/v 2 ) (see Eq. 41-8) are the indices of refraction of medium 1 and of medium
2, respectively, with respect to a vacuum. Introducing the symbols n 1 and
n 2 for these quantities allows us to write the law of refraction as
(41-10)
If we assume that the medium above the glass in Fig. 41-9 is a vacuum
rather than air, the speed v1 becomes c and the wavelength, called ;\ 1 in
Fig. 41-9, assumes a value;\ that is characteristic of the wave in free space.
Equation 41-6 may thus be written
V2 ;\
;\2 = ; \ - = - · (41-11)
c n2
or (41-12)
Fig. 41-11 Showing the total internal reflection of light from a source S; the critical angle
is Oc,
Sec. 41-5 TOTAL INTERNAL REFLECTION
1025
For glass and air sin ec = (1.00/1.50) = 0.667, which yields ec = 41.8°.
Total internal reflection does not occur when light originates in the medium
of lower index of refraction.
Light can be "piped" from one point to another with little loss by allowing it to
enter one end of a rod of transparent plastic. The light will undergo total internal
Fig. 41-12 A bundle of tapered fibers (belcrw) is placed over the letter S. Above, with
the aid of a mirror, we see that the image, reduced in size, is transmitted to the top of the
bundle by total internal reflection in the individual fibers. (Courtesy Dr. N. S. Kapany,
Optics Technology, Inc.)
reflection at the boundary of the rod and will follow its contour, emerging at its far
end. Images may be transferred from one location to another, using a bundle of fine
glass fibers, each fiber transmitting a small fraction of the image.* Such bundles can
be made in which a (flexible) seven-foot length delivers half the energy entering it at
the far end. In a fiber in such a bundle a typical ray may undergo 48,000 reflections.
Most of the energy loss is due to absorption within the glass, reflection being almost
* See "Fiber Optics," by N. S. Kapany, Scientific American, November 1960.
1026 REFLECTION AND REFRACTION Chap. 41
Fig. 41-13 A photomicrograph of light emerging from the end of a bundle of fibers. The
fiber diameters approach the wavelength of light so that each fiber acts like an "optical
waveguide." We have here convincing visual evidence that light is an electromagnetic
wave. (Courtesy of Dr. N. S. Kapany, Optics Technology, Inc.)
truly total. Measurable amounts of light can be transmitted through single fibers
150 feet long.
Fiber optics techniques make possible many useful optical devices for transmitting
and transforming luminous images. Figure 41-12 shows a short fiber bundle con-
structed so that the fibers taper in diameter along its length. The wide end is shown
placed over the letter Sin the printed word "OPTICS." We see, with the aid of a
mirror placed above the bundle, that a letter S, reduced in size, has been transmitted
by tofal internal reflection through the bundle to its narrow end.
Figure 41-13 is an enlarged view of a cross section of such a fiber bundle in .which
the diameters of the individual fibers are made so small that they are of the order of
magnitude of the wavelength cf light. This condition violates the spirit of our
assumption of p. 1016, namely that the transverse dimensions of reflecting and
refracting surfaces would be large compared to the wavelength of light. Conse-
quently, a description of the reflect.ion and refraction of light in terms of rays, as in
Figs. 41-lb and 41-11, is not possible. Figure 41-13 is readily interpreted, however,
on the basis of the electromagnetic wave theory of light and provides convincing
pictorial supporting evidence for that theory. The fibers behave like waveguides *
* A dielectric rod can serve as a waveguide in the same way that a hollow metal pipe
can, and it has similar properties.
Sec. 41-5 TOTAL INTERNAL REFLECTION 1027
(see Section 39-3), and the patterns of darkness and light represent the distribution
of the E and B vectors for various modes of oscillation of the electromagnetic waves
traveling down the "guides."
(a) (b)
The angle 81 must be equal to or greater than the critical angle Oc where Oc is given
by Eq. 41-12:
. n2 1
Sln Oc = - = - ,
n1 n
in which, for all practical purposes, the index of refraction of air ( = n2) is set equal
to unity. Suppose that the index of refraction of the glass is such that total internal
reflection just occurs, that is, that Oc = 45°. This would mean
1
n = . 450 = 1.41.
sm
Thus the index of refraction of the glass must be equal to or larger than 1.41. If it
were less, total internal reflection would not occur.
. n2 1.33
sm Oc = n 1 = l. 50 = 0.887,
which corresponds to Oc = 62.5°. The actual angle of incidence ( = 45°) is less than
this so that we do not have total internal reflection.
There is a reflected ray, with an angle of reflection of 45 °, as Fig. 41-14b shows.
There is also a refracted ray, with an angle of refraction given by
~ 60f--~-+~~l""'?.?'4""""-:.:_::_4-~--+~~-+-~~t--~---+--J----i
:,,;;
~
~ 40t--~-+~~L....<LL----,-....it11,rLl.i_J_~--+~~-+-~~t--~-f------\----i
w
Reflected wave
0 0° 10° 20° 30° 40° 50° 60° 70° 80° 90°
Angle of incidence, ()
~
~ 40t--~-t-~~-t-~------i-~~4--it~~~~""-'-,--""""""'~
w
Reflected wave
Fig. 41-15 (a) The per cent of the energy reflected and refracted when an incident wave
in air falls on glass (n = 1.50). (b) The same for the incident wave in glass, showing total
internal reflection.
prediction for (a) a light beam in air falling on a glass-air interface and (b) a light beam
in glass falling on such an interface. Figure 41-15a shows that for angles of incidence
up to about 50°, less than 10% of the light energy is reflected. At grazing incidence,
however (that is, angles of incidence near 90°), the surface becomes an excellent
reflector. We are all familiar with the high reflecting power of a wet road for light
from automobile headlights that strikes near grazing incidence.
Figure 41-15b shows clearly that at a certain critical angle (41.8° in this case; see
Eq. 41-12) all the light is reflected. For angles of incidence appreciably below this
value, about 4% of the energy is reflected.
Figure 41-16 shows two fixed points A and Band a ray APB connecting them.*
The total length l of this ray is
l = y~2+ x2 + V b2 + (d - x) 2'
where x locates the point P at which the ray touches the mirror.
According to Fermat's principle, P will have a position such that the time of travel
of the light must be a minimum (or a maximum or must remain unchanged). Ex-
pressed in another way, the total length l of the ray must be a minimum (or a max-
imum or must remain unchanged). In either case, the methods of the calculus re-
quire that dl/dx be zero. Taking this derivative yields
t =
n1Z1 + n2l2 =
l
-·
c c
The quantity l ( = n 1Z1 + n2l 2 ) is called the optical path length of the ray. Equation
41-11 (written as A = n"'An) shows that the optical path length is equal to the length
* We assume that ray APB lies in the plane of the figure; see Problem 20.
1030 REFLECTION AND REFRACTION Chap. 41
that this same number of waves would have if the medium were a vacuum. The
optical path length must not be confused with the geometrical path length, which is
Z1 + Z2.
Fermat's principle requires that l be a minimum (or a maximum or must remain
unchanged) which, in turn, requires that x be so chosen that dl/dx = 0. The optical
pa th length is
l = n1li + n2Z2 = n1Va 2 x 2 + n2Vb 2 +(d - x)~ +--
Differentiating yields
dl
~
dx
= n1(!)(a 2 +x 2) - 1 i 2 (2x) + n2(!)[b 2 + (d - x) 2J- 1i 2 (2)(d - x)(-1) = 0
I
QUESTIONS
4. Would you expect sound waves to obey the laws of reflection and of refraction obeyed
by light waves?
5. A street light, viewed by reflection across a body of water in which there are ripples,
appears very elongated. Explain.
6. The light beam in Fig. 41-la is broadened on entering the water. Explain.
7. By what per cent does the speed of blue light in fused quartz differ from that of red
light?
8. Can (a) reflection phenomena or (b) refraction phenomena be used to determine the
wavelength of light?
9. How did Foucault's measurement of the speed of light in water decide between the
wave and the particle theories of light?
10. Why does a diamond "sparkle" more than a glass imitatiah cut to the same shape?
11. Is it plausible that the wavelength of light should change in passing from air into
glass but that its frequency should not? Explain.
12. How can one determine the indices of refraction of the media in Table 41-1 relative
to water, given the data in that table?
13. You are given a cube of glass. How can you find the speed of light (from a sodium
light source) in this cube?
14. Describe and explain what a fish sees as he looks in various directions above his
"horizon."
15. Design a periscope, taking advantage of total internal reflection. What are the
advantages compared with silvered mirrors?
16. What characteristics must a material have in order to serve as an efficient "light
pipe"?
17. Can the optical path length between two points ever be less than the geometrical
path length between those points?
PROBLEMS
1. One end of a stick is dragged through water at a speed v which is greater than the
speed u of water waves. Applying Huygens' construction to the water waves, show that
a conical wavefront is set up and that its half-angle a is given by
sin a = u/v.
This is familiar as the bow wave of a ship or the shock wave caused by an object moving
through air with a speed exceeding that of sound, as in Fig. 20---12 .
.-%.'·Prove that if a mirror is rotated through an angle a, the reflected beam is rotated
through an angle 2a. Is this result reasonable for a = 45°?
3. What is the speed in fused quartz of light of wavelength 5500 A; see Fig. 41-2?
4. The wavelength of yellow sodium light in air is 5890 A. (a) What is its frequency?
(b) What is its wavelength in glass whose index of refraction is 1.52? (c) From the results
of (a) and (b) find its speed in this glass.
5. The speed of yellow sodium light in a certain liquid is measured to be l.92 X 10 8
meters/sec. What is the index of refraction of this liquid, with respect to air, for sodium
light?
6. Suppose that the speed of light in air has been measured with an uncertainty of, say,
1 km/sec. In calculating the speed in vacuum, suppose that it is not certain whether n
for air is 1.00029 or 1.00030. (a) How much extra uncertainty is introduced into the
calculated value for c? (b) Estimate how accurately n should be known for this purpose.
~- In Fig. 41-6 show by graphical ray tracing, using a protractor, that if e for the
incident ray is either increased or decreased, the deviation angle f is increased. The
symmetrical situation shown in this figure is called the position of minimum deviation.
1032 REFLECTION AND REFRACTION Chap. 41
""8: A ray of light is incident normally on the face ab of a glass prism (n = 1.52), as
shown in Fig. 41-18. (a) Assuming that the prism is immersed in air, find the largest value
for the angle cp so that the ray is
totally reflected at face ac. (b)
Find cp if the prism is immersed
in water.
9. Show that for a thin prism
(¢ small) and light not far from
normal incidence (01 small) the de-
viation angle is independent of the
angle of incidence and is equal to
Fig. 41-18 (n - 1) cp (see Fig. 41-6).
10. A glass prism with an apex
angle of 60° has n = 1.60. (a) What is the smallest angle of incidence for which a ray can
enter one face of the prism and emerge from the other? (b) What angle of incidence would
be required for the ray to pass through the prism symmetrically, as in Fig. 41-6?
11. A 60° prism is made of fused quartz. A ray of light falls on one face, making an
angle of 45° with the normal. Trace the ray through the prism graphically with some
care, showing the paths traversed by rays representing (a) blue light, (b) yellow-green light,
and (c) red light. See Figs. 40-2 and 41-2.
--1?. Ptolemy, who lived at Alexandria toward the end of the first century A.D., gave
the following measured values for the angle of incidence 0 1 and the angle of refraction 02
for a light beam passing from air to water:
01 02 01 02
100 7°45 1 50° 35°0'
20° 15°30 1 60° 40°30 1
30° 22°30 1 70° 45°30 1
40° 29°0 1 80° 50°0 1
Are these data consistent with Snell's law; if so, what index of refraction results? These
data are interesting as the oldest recorded physical measurements.
----13, Prove that a ray of light
incident on the surface of a sheet of
plate glass of thickness t emerges
from the opposite face parallel to
its initial direction but displaced
sideways, as in Fig. 41-19. Show
that, for small angles of incidence
0, this displacement is given by
n - 1
x = te~-
n
Fig. 41-20
17. A drop of liquid may be placed on a semicircular slab of glass as in Fig. 41-21.
Show how to determine the index of refraction of the liquid by observing total internal
reflection. The index of refraction of the glass is unknown and must also be determined.
Is the range of indices of refraction that can be measured in this way restricted in any
sense?
Fig. 41-21
1034 REFLECTION AND REFRACTION Chap. 41
18. A light ray falls on a square glass slab as in Fig. 41-22. What must the index of
refraction of the glass be if total internal reflection occurs at the vertical face?
Fig. 41-22
19. A glass cube has a small spot at its center. What parts of the cube face must be
covered to prevent the spot from being seen, no matter what the direction of viewing?
What fraction of the cube surface must be so covered? Assume a cube edge of 1.0 cm
and an index of refraction of 1.50. (Neglect the subsequent behavior of an internally
reflected ray.)
20. Using Fermat's principle, prove that the reflected ray, the incident ray, and the
normal lie in one plane.
21. Prove that the optical path lengths for reflection and refraction in Figs. 41-16
and 41-17 are minima when compared with other nearby paths connecting the same two
points.
Reflection and Refraction-
Spherical Waves
and Spherical Surfaces
CHAPTER 42
Fig. 42-1 An attempt to isolate a ray by reducing the slit width a fails because of diffrac-
tion, which becomes more pronounced as a/}.. approaches zero.
Fig. 42-2 Diffraction of water waves at a slit in a ripple tank. Note that the slit width
is about the same size as the wavelength. (Courtesy of Educational Services Incorporated.)
(a)
(b)
(c)
Fig. 42-3 (a) The intensity of light diffracted from a slit of width a ,..__, 6 X 10-3 mm
and falling on a screen 50 cm beyond. (b) The slit width is reduced by a factor of two.
(c) The slit width is further reduced by an additional factor of seven. Note that secondary
maxima, made prominent in this case by deliberate overexposure, appear on either side
of the central maximum. These secondary maxima have been omitted from Fig. 42-1
for simplicity.
1038 REFLECTION AND REFRACTION-II Chap. 42
If the requirement for geometrical optics is not met, we cannot describe
the behavior of light by rays but must take its wave nature specifically into
account. This subject is called wave optics; it includes geometrical optics as
an important limiting case. We will treat wave optics in succeeding chapters.
behind the mirror that the object O is in front of it; I is called the image of 0.
Images may be real or virtual. In a real image light energy actually passes
through the image point; in a virtual image the light behaves as though it
diverges from the image point, although, in fact, it does not pass through
this point; see Fig. 42--4. Images in plane mirrors are always virtual. We
know from daily experience how "real" such a virtual image appears to be
and how definite is its location in the space behind the mirror, even though
this space may, in fact, be occupied by a brick wall.
Figure 42-5 shows two rays from Fig. 42_:4. One strikes the mirror at v,
along a perpendicular line. The other strikes it at an arbitrary point a,
making an angle of incidence e with the normal at that point. Elementary
* In our discussion of reflection from mirrors in Chapter 41 (see Fig. 41-8) we assumed
an incident plane wave; the incident rays are parallel to each other in that case.
Sec. 42-2 SPHERICAL WA VES~PLANE MIRROR 1039
geometry shows that the angles aOv and alv are also equal to fJ. Thus the
right triangles aOva and alva are congruent and
0 = -i, (42-1)
in which the minus sign is arbitrarily introduced to show that I and Oare on
opposite sides of the mirror. Equation 42-1 does not involve e, which means
that all rays striking the mirror pass through I when extended backward, as
we have seen above. Beyond assuming that the mirror is truly plane and
that the conditions for geometrical optics hold, we have made no approxi-
mations in deriving Eq. 42-1. A point object produces a point im?,ge in a
plane mirror, with o = -i, no matter how large the angle () in Fig. 42-5.
Because of the finite diameter of the pupil of the eye, only rays that lie
fairly close together can enter the eye after reflection at a mirror. For the
eye position shown in Fig. 42-6 only a small patch of the mirror near point a
is effective in forming the image; the rest of the mirror may be covered up
or removed. If the eye is moved to another location, a different patch of the
mirror will be effective; the location of the virtual image I will remain un-
changed, however, as long as the object remains fixed.
If the object is an extended source such as the head of a person, a virtual
image is also formed. From Eq. 42-1, every point of the source has an image
point that lies an equal distance directly behind the plane of the mirror.
Thus the image reproduces the object point by point.
Images in plane mirrors differ from objects in that left and right are inter-
changed. The image of a printed page is different from the page itself.
Similarly, if a top is made to spin clockwise, the image, viewed in a vertical
mirror, will seem to spin counterclockwise. Figure 42-7 shows an image of
a left hand, constructed by using point-by-point application of Eq. 42-1; the
image has the symmetry of a right hand.*
* See "The Overthrow of Parity" by Philip Morrison, Scientific American, April, 1957,
for a discussion of the distinction in nature between right and left.
1040 REFLECTION AND REFRACTION-II Chap. 42
Fig. 42-6 A pencil of rays from O enters the eye after reflection at the mirror. Only a
small portion of the mirror near a is effective. The small arcs represent portions of
spherical wavefronts.
4
~
Fig. 42-8 Example 1. A person can view his +:
full-length image in a mirror that is only half 3 "So
his height. '<ii
I
2
1
0
~
Example 2. Two plane mirrors are placed at right angles, and a point object
O is located on the perpendicular bisector, as shown in Fig. 42-9a. Locate the images.
Images 11 and 12 are formed in mirrors ab and cd, respectively. There is also a third
image; it may be considered to be the image of 1 1 in mirror cd or the image of 1 2 in
mirror ab. The three images and the object O lie on a circle whose center is on the
line of intersection of the mirrors and whose plane is at right angles to that line.
In viewing 1 3 the light entering the observer's eye is reflected twice after leaving the
source. Figure 42-9b shows a typical bundle of rays. In viewing l 1 or l 2, the light
is reflected only once, as in Fig. 42-6. ..ill
* A spherical shell, viewed from inside, is everywhere concave; viewed from outside it is
everywhere convex. In this chapter concave and convex will always be judged from the
point of view of an observer sighting along the direction of the incident light.
1042 REFLECTION AND REFRACTION~n Chap. 42
-----
spherical convex mirror, form-
ing a virtual image at I, the
point from which they appear
to originate. Compare Fig.
42-10.
SPHERICAL WAYES-SPHERICAL MIRROR 1045
fore, we call the side of the mirror from which the light comes the R-side (for
real image). We call the back of the mirror the V-side (for virtual image),
because images formed on this side of the mirror must be virtual, no light
energy being present on this side.*
In the indented statement above we associated real images with positive
image distances. This suggests our first sign convention:
1. The image distance i is positive if the image (real) lies on the R-side
of the mirror, as in Fig. 42-10; i is negative if the image (virtual) lies on the
V-side, as in Fig. 42-12.
If the mirror in Fig. 42-10, which is concave as viewed from the direction
of the incident light, is made convex, the rays will diverge after reflection
and will form a virtual image, as Fig. 42-12 shows. Thus the indented
statement above suggests our second sign convention:
2. The radius of curvature r is positive if the center of curvature of the
mirror lies on the R-side, as in Fig. 42-10; r is negative if the center of curva-
ture lies on the V-side, as in Fig. 42-12.
The student should not commit these sign conventions to memory but
should deduce them in each case from the basic statement on p. 1044, using
Fig. 42-10 as a mnemonic aid.
For all cases in this book the object distance o is to be taken as positive. In
systems of two or more mirrors (or combinations of mirrors and refracting surfaces)
it is possible to arrange that converging light falls on the mirror. In such cases the
object is called virtual and the object distance o is negative; we limit our discussion
here to real objects .
..,.. Example 3. A convex mirror has a radius of curvature of 20 cm. If a point
source is placed 14 cm away from the mirror, as in Fig. 42-12, where is the image?
A rough graphical construction, applying the law of reflection at a in the figure,
shows that the image will be on the V-side of the mirror and thus will be virtual.
We may verify this quantitatively and analytically from Eq. 42-4, noting that r is
negative here because the center of curvature of the mirror is on its V-side. We have
1 1 2
-+-:=-
O i r
1 1 2
or
+14 cm +i = -20 cm'
which yields i = - 5.8 cm, in agreement with the graphical prediction. The nega-
tive sign for i reminds us that the image is on the V-side of the mirror and thus is
virtual. ....
When parallel light falls on a mirror (Fig. 42-13), the image point (real or
virtual) is called the focal point F of the mirror. The focal length f is the
distance between F and the vertex. If we put o ---,. oc:i in Eq. 42-4, thus
insuring parallel incident light, we have
i. = 1
2r =
f.
* This nomenclature may seem unnecessarily cumbersome at this point. We adopt it
because of the later extension of these ideas to refracting surfaces and lenses.
Fig. 42-13 (a) The focal point for a concave spherical mirror, showing both the rays
and the wavefronts. F and C lie on the R-side, the focal point is real, and the focal
length f of the mirror is positive (as is r). (b) The same, except that the incident light
makes an angle a with the mirror axis; the rays are focused at a point in the focal plane
F'F". (c) Same as (a) except that the mirror is convex; F and C lie on the V-side of the
mirror. The focal point is virtual and the focal length! is negative (as is r).
1046
Sec. 42-3 SPHEHICAL WA VES-SPHEHICAL MIRROR 1047
-1 + -1 = -,1 (42-5)
O i j
where f, liker, is taken as positive for mirrors whose centers of curvature are
on the R-sidc (that is, for concave, or converging mirrors; see Fig. 42-l:3a)
and negative for those whose centers of curvature arc on the V-sidc (that is,
for convex, or diverging mirrors; see Fig. 42-13c). Figure 42-l:3b shows an
incident plane ,vave that makes a small angle a with the mirror axis. The
rays are focused at a point in the focal plane of the mirror. This is a plane
at right angles to the mirror axis at the focal point.
Fig. 42-14 The image of an extended object in (a) a concave mirror and (b) a convex
mirror is located graphically. Any two of the three special rays shown are sufficient.
We now consider objects that are not points. Figure 42-14 shows a candle
in front of (a) a concave mirror and (b) a convex mirror. We choose to draw
the mirror axis through the foot of the candle and, of course, through the
center of curvature. The image of any off-axis point, such as the tip of the
candle, can be found graphically, using the following facts:
l. A ray that strikes the mirror after passing (either directly or upon being
extended) through the center of curvature C returns along itself (ray x in
Fig. 42-1:3). Such rays strike the mirror at right angles.
2. A ray that strikes the mirror parallel to its axis passes (or will pass when
extended) through the focal point (ray y).
1048 REFLECTION AND REFRACTIO:',l" ~u Chap. 42
3. A ray that strikes the mirror after passing (either directly or upon being
extended) through the focal point emerges parallel to the axis (ray z).
Figure 42-1.5 shows a ray (dve) that originates on the tip of the object
candle of Fig. 42-14a, is reflected from the mirror at point v, and passes
through the tip of the image candle. The law of reflection demands that
this ray make equal angles e with the mirror axis as shown. For the two
similar right triangles in the figure we can write
ce vc
bd vb
The quantity on the left (apart from a question of sign) is the lateral mag-
nification m of the mirror. Since we want to represent an inverted image by
a negative magnification, we arbitrarily define m for this case as - (ce/bd).
Since vc = i and vb = o, we have at once
m= (42-6)
0
This equation gives the magnification for spherical and plane mirrors under
all circumstances. For a plane mirror, o = - i and the predicted magnifica-
+
tion is 1 which, in agreement with experience, indicates an erect image the
same size as the object.
Images in spherical mirrors suffer from several "defects" that arise because the
assumption of paraxial rays is never completely justified. In general, a point source
will not produce a point image; see Problem 5. Apart from this, distortion arises
because the magnification varies somewhat with distance from the mirror axis, Eq.
42-6 being strictly correct only for paraxial rays. Superimposed on these defects are
diffraction effects which come about because the basic assumption of geometrical optics,
that light travels in straight lines, must always be considered an approximation.
Fig. 42-15 A particular ray for the arrangement of Fig. 42-14, used to show that the
lateral magnification mis given by -i/o.
Sec. 42-4 SPHERICAL REFRACTING SURFACE 1049
i/f
I \
3~-2.ox .
\ Concave mirror,
2
1
real invert d image
~ Ix ~
Convex mirror,
virtual erect image +---1
x-x
+0.25
I + 0·3 - ;+o.5o
--x~
I x +LO
I ) '
o/f
-3 -2 -1 0 1 2 3
Plane mirror
-1~y2.a
\ I
-2-x
!.+!.=!. +3.0 \ Concave mirror,
O i f virtual erect image
-3 '----X
+4.0
Fig. 42-16 A graphical representation of the mirror formula (Eq. 42-·5). The numbers
on the curves give the lateral magnifications (Eq. 42-6).
Figure 42-16 summarizes the predictions of the mirror equation (Eq. 42-5)
and of the lateral magnification equation (Eq. 42-6). For convex mirrors
the image is always virtual. For concave mirrors the image is virtual if the
object is inside the focal point (o/f < 1) and real otherwise (o/f > 1). As
o ---t OC:J for both types of mirror, the image approaches the focal point
(i/f ---t 1). A plane mirror (m = +1) is represented in the figure by a
point at the origin. It is appropriate that this point form the central point
of the lower-left branch of Fig. 42-16 because both convex and concave
mirrors approach plane mirrors as r is increased in magnitude, that is, as
f ---tOC) (or as o/f - 0 in Fig: 42-16, assuming a fixed object distance).
Fig. 42-17 Two rays from O converge after refraction at a spherical surface, forming a
real image at I.
The refracted ray intersects the axis at I. A ray from O that travels along
the axis will not be bent on entering the surface and will also pass through I.
Thus, for these two rays at least, I is the image of 0.
As in the derivation of the mirror equation, we use the theorem that the
exterior angle of a triangle is equal to the sum of the two opposite interior
angles. Applying this to triangles COaC and !Cal yields
81 = a+ /3
and (42-8)
As a is made small, angles /3, 'Y, 81 , and 82 in Fig. 42-17 also become small.
lVe at once assume that a, hence all these angles, are arbitrarily small, this
paraxial ray assumption was also made for spherical mirrors. Replacing
the sines of the angles by the angles themselves-since the angles are re-
quired to be small-permits us to write the law of refraction as
(42-9)
Eliminating 81 between this equation and Eq. 42-7 leads, after rearrange-
ment, to
(42-10)
Sec. 42-4 SPHEHICAL REFRACTING SURFACE 1051
In radian measure the angles a, (3, and 'Y in Fig. 42-17 are
av
0
av
(3 = - (42-11)
r
av
'Y r-..; - : - •
Only the second of these equations is exact. The other two are approximate
because I and O arc not the centers of circles of which av is an arc. However,
for paraxial rays (a small enough) the inaccuracies in Eq. 42-11 can be made
as small as desired.
Substituting Eqs. 42-11 into Eq. 42-10 leads readily to
n1 n2 n2 - n1
-+-:-=--- (42-12)
O i r
This equation holds whenever light is refracted from point objects at spheri-
cal surfaces, assuming only that the rays are paraxial. As with the mirror
formula, care must be taken to use Eq. 42-12 with consistent signs for o, i,
and r. Once again we establish our sign conventions by physical reasoning
from a particular case, that of Fig. 42-17:
In Fig. 42-17, in which light diverges from a real object, falls on a
convex refracting surface, and converges after refraction to form a real
image, the quantities o, i, and r in Eq. 42-12 have positive numerical
values.
Figure 42-17 was used to derive Eq. 42-12, and the student should asso-
ciate them in his mind as an aid in getting the signs correct. This basic
statement is quite similar to that which we made for mirrors on p. 1044.
We fix our attention on the side of the refracting surface from which the
incident light falls on the surface. In contrast to mirrors, the light energy
passes through a refracting surface to the other side, and if a real image is
formed it must appear on the far side, which we call the R-side. The side
from which the incident light comes is called the V-side because virtual
images must appear here. Figure 42-18 suggests this important distinction
between reflection and refraction.
In the indented statement above we associated real images with positive-
image distances. Thus we are led to the sign convention:
Reflected light
Fig. 42-18 Real images are formed on the same
(concave or side as the incident light for mirrors but on the
convex mirror)
opposite side for refracting surfaces and thin
V-side I R-side lenses. This is so because the incident light is
Incident light n1 I n2 reflected back by mirrors but is transmitted
I through by refracting surfaces.
I
I Refracted light
I
(concave or convex
refracting surface
or thin lens)
a virtual image, as Fig. 42-19 shows. Thus we are led to our second sign
convention:
2. The radius of curvature r is positive if the center of curvature of the
refracting surface lies on the R-side, as in Fig. 42-17; r is negative if the
center of curvature is on the V-side, as in Fig. 42-19.
The sign conventions for refracting surfaces are the same as for mirrors
(p. 1045), the fundamental difference between the two situations being
absorbed in the definitions of R-side and V-side in Fig. 42-18. This differ-
ence is easily remembered on physical grounds.
For all cases in this book the object distance o is to be taken as positive. In
systems of two or more refracting surfaces (or combinations of refracting surfaces
and mirrors) it is possible to arrange that converging light falls on the refracting
surface. In such cases the object is called virtual and the object distance o is negative;
we limit our discussions here to real objects.
Fig. 42-19 Two rays from O diverge after refraction at a spherical surface, forming a
virtual image at I.
Sec. 42-4 SPHERICAL REFRACTING SURFACE 1053
~ Example 4. Locate the image for the geometry shown in Fig. 42-17, assuming
the radius of curvature to be 10 cm, n2 to be 2.0, and n 1 to be 1.0. Let the object be
20 cm to the left of v.
From Eq. 42-12,
we have
1.0 + 2.0 = 2.0 - 1.0_
+20 cm i +10 cm
Note that r is positive because the center of curvature of the surface lies on the R-side.
This relation yields i = +40 cm in agreement with the graphical construction. The
light energy actually passes through I so that the image is real, as indicated by the
positive sign for i.
Example 5. An object is immersed in a medium with n 1 = 2.0, being 15 cm from
the spherical surface whose radius of curvature is -10 cm, as in Fig. 42-20; r is
negative because C lies on the V-side. Locate the image.
Figure 42-20 shows a ray traced through the surface by applying the law of refrac-
tion at point a. A second ray from O along the axis emerges undeflected at v. The
image I is found by extending these two rays backward; it is virtual.
From Eq. 42-12,
~.
Fig. 42-21 Refraction at a
plane surface at near-normal
incidence, showing a pencil of
rays and the corresponding
wavefronts entering the pupil.
(a) Source in air and (b) source
in water.
Sec. 42-5 THIN LENSES 1055
Figure 42-21 illustrates the situation graphically (a) for an object in air as seen from
below water and (b) for an object in water with air above. This shows that a diver,
looking upward at, say, an overhanging tree branch, will think it higher than it is by
the factor 1.33/1.00. Similarly, an observer in air will think that objects on the
bottom of a water tank are closer to the surface than they actually are, in the ratio
1.00/1.33. These considerations, being based on Eq. 42-12, hold only for paraxial
rays, which means that the incident rays can make only a small angle with the normal;
this angle has been exaggerated in the figure for clarity .. Note again that n1 is always
identified with the medium that lies on the side of the surface containing the incident
light. .....
Fig. 42-22 (a) Two rays from 0' intersect at I" (real image) after refraction at two spheri-
cal surfaces. (b) The first surface and (c) the second surface shown separately. The
quantities a and n have been exaggerated for clarity.
Let us now assume that the thickness l of the "lens" in Fig. 42-22 is so
small that it can be neglected in comparison with other linear quantities in
Sec. 42-5 THIN LENSES 1057
this figure (such as o', i', o", i", r', and r"). In all that follows we make this
thin-lens approximation. Putting l = 0 in Eq. 42-15 leads to
n 1 n - 1
-+-= - - - · (42-16)
i' i" r"
Adding Eqs. 42-13 and 42-16 leads to
~ +~
o' i"
= (n - 1) (~ -
r'
2-) ·
r"
Finally, calling the original object distance simply o and the final image
distance simply i leads to
~ +~
o i
= (n - 1) (~ -
r'
2-) ·
r"
(42-17)
This equation holds only for paraxial rays and only if the lens is so thin that
it essentially makes no difference from which surface of the lens the quanti-
ties o and i are measured. In Eq. 42-17 r' refers to the first surface struck
by the light as it traverses the lens and r" to the second surface.
The sign conventions for Eq. 42-17 are the same as those for mirrors and
for single refracting surfaces. Because the lens is assumed to be thin, we
ref er to the R-side and the V-side of the lens itself (see Fig. 42-18) rather
than those of its separate surfaces. The sign conventions then are the fol-
lowing:
1. The image distance i is positive if the image (real) lies on the R-side
of the lens, as in Fig. 42-23a; i is negative if the image (virtual) lies on the
V-side of the lens, as in Fig. 42-23b.
2. The radii of curvature r' and r" are positive if their respective centers
Fig. 42-23 Illustrations used to establish sign conventions for thin lenses.
1058 REFLECTION AND REFRACTION~II Chap. 42
of curvature lie on the R-side of the lens. They are negative if their centers
lie on the V-side. In Fig. 42-23a, r' is positive and r" is negative; in Fig.
42-23b, r' is negative and r" is positive.
Figure 42-24a and c shows parallel light from a distant object falling on
a thin lens. The image location is called the second focal point F 2 of the lens.
The distance from F 2 to the lens is called the focal length f. The first focal
point for a thin lens (F 1 in figure) is the object position for which the image
Fig. 42-24 (a) Parallel light passes through the second focal poiut F2 of a converging
lens. (b) The incident light makes an angle a with the lens axis, the rays being focused
in the focal plane F'F". (c) Parallel light, passing through a diverging lens, seems to
originate at the second focal point F 2• C' and C" are centers of curvature for the lens
surfaces; F 1 is the first focal point.
Sec. 42-5 THIN LENSES 1059
is at infinity. For thin lenses the first and second focal points are on opposite
sides of the lens and are equidistant from it.
The focal length can be computed from Eq. 42-17 by inserting o ------} oo
and i = f. This yields
! = (n - 1) (~ - --1:_) · (42-18)
f r' r"
This relation is called the lens maker's equation because it allows us to com-
pute the focal length of a lens in terms of the radii of curvature and the index
of refraction of the material. Combining Eqs. 42-17 and 42-18 allows us to
write the thin-lens equation as
1 1 1
- +- = -· (42-19)
O i f
Figure 42-24b shows parallel incident rays that make a small angle a with
the lens axis; they are brought to a focus in the focal plane F' F", as shown.
This is a plane normal to the lens axis at the focal point.
In Fig. 42-24a we note that all rays in the figure contain the same number
of waveler{gths; in other words, they have the same optical path lengths; see
Section 41-6. The optical path lengths are the same because the wavefronts
are surfaces over which the wave disturbance has the same constant value
and because all the rays shown pass through the same number of wavefronts.
IJI,, Example 7. The lenses of Fig. 42-24 have radii of curvature of magnitude 40 cm
and are made of glass with n = 1.65. Compute their focal lengths.
Since C' lies on the R-side of the lens in Fig. 42-24a, r' is positive ( = +40 cm).
Since C" lies on the V-side, r" is negative ( = -40 cm). Substituting in Eq. 42-18
yields
-1 =
f
(n - 1) (1- - -
r'
1)
r"
= ( 1.65 - 1) (+401cm - ~- 1
-40 cm
) ,
or f = +31 cm.
A positive focal length indicates that in agreement with Fig. 42-24a the focal point
F 2 is on the R-side of the lens and parallel incident light converges after refraction
to form a real image.
In Fig. 42-24c C' lies on the V-side of the lens so that r' is negative ( = -40 cm).
Since r" is positive ( = +40 cm), Eq. 42-17 yields
f = -31 cm.
A negative focal length indicates that in agreement with Fig. 42-24c the focal point
F 2 is on the V-side of the lens and incident light diverges after refraction to form a
virtual image. ...i
The location of the image of an extended object such as a candle (Fig. 42-
25) can be found graphically by using the following three facts:
1. A ray parallel to the axis and falling on the lens passes, either directly
or when extended, through the second focal point (ray x in Fig. 42-25).
2. A ray falling on a lens after passing, either directly or when extended,
1060 REFLECTION AND REFRAC'l'JON~II Chap. 42
through the first focal point will emerge from the lens parallel to the axis
(ray y).
3. A ray falling on the lens at its center will pass through undeflected.
There is no deflection because the lens, near its center, behaves like a thin
piece of glass with parallel sides. The direction of the light rays is not
changed and the sideways displacement can be neglected because the lens
thickness has been assumed to be negligible (ray z; see also Problem 13,
Chapter 41).
Figure 42-26, which represents part of Fig. 42-25a, shows a ray passing
from the tip of the object through the center of curvature to the tip of the
image. For the similar triangles abe and dee we may write
de de
ab ae
The right side of this equation is i/ o and the left side is -m, where mis the
Fig. 42-25 Showing the graphical location of images for three thin lenses.
Sec. 42-5 THIN LENSES 1061
which holds for all types of thin lenses and for all object distances .
.... Example 8. A converging thin lens has a focal length of +24 cm. An object is
placed 9.0 cm from the lens as in Fig. 42-25b; describe the image.
From Eq. 42-19,
1 1 1
;+i=r
1 1 1
we have -+-9-.0-cm-+ i = +24 cm'
which yields i = -14.4 cm, in agreement with the figure. The minus sign means that
the image is on the V-side of the lens and is thus virtual.
The lateral magnification is given by
i -14.4cm
m = - - = - - - - - = +1.6,
o +9.0 cm
again in agreement with the figure. The plus signifies an erect image.
Images formed by lenses suffer from defects similar to those discussed for
mirrors on p. 1048. There are effects connected with the failure of a point
object to form a point image, with the variation of magnification with dis-
tance from the lens axis, and with diffraction. For lenses, but not for mir-
rors, there are also chromatic aberrations associated with the fact that the
refracting properties of the lens vary with wavelength because the index of
refraction of the lens material does. If a point object on the lens axis emits
white light, the image, neglecting other lens defects, will be a series of colored
points spread out along the axis. We have all seen the colored images pro-
duced by inexpensive lenses. A great deal of ingenious optical engineering
goes into the design of lenses (more commonly lens systems) in which the
various lens defects are minimized. The lens surfaces are normally not
spherical.
1062 REFLECTION AND REFRACTION-II Chap. 42
i/f
1 _\
3--2.ox
\ Converging lens,
real inverted image
2 -1.0 x'-..._I
Diverging lens, -o;o-
virtual erect image - - 1
x-'
+0.25 +OJj"""--x~
I I +0.50
x
+1.0
o/f
\f
-3 -2 -1 0 2 3
-1 -y1.o
1 1 1
-2 +30, Converging lens,
-+-
O i
=-f virtual erect image
3 +4.0x
Fig. 42-27 A graphical representation of the thin-lens formula (Eq. 42-19). The num-
bers on the curves give the lateral magnifications (Eq. 42-20). Compare Fig. 42-16.
QUESTIONS
1. If a mirror reverses right and left, why doesn't it reverse up and down?
2. Is it possible to photograph a virtual image?
3. What approximations were made in deriving the mirror equation (Eq. 42-4):
1 1 2
-+-:-=-?
o i r
4. Can you think of a simple test or observation to prove that the law of reflection is
the same for all wavelengths, under conditions in which geometrical optics prevails?
5. Under what conditions will a spherical mirror, which may be concave or convex, form
(a) a real image, (b) an inverted image, and (c) an image smaller than the object?
6. An unsymmetrical thin lens forms an image of a point object on its axis. Is the image
location changed if the lens is reversed?
7. Why has a lens two focal points and a mirror only one?
Chap. 42 PROBLEMS 1063
8. Under what conditions will a thin lens, which may be converging or diverging, form
(a) a real image, (b) an inverted image, and (c) an image smaller than the object?
9. Does the apparent depth of an object below water depend on the angle of view of the
observer in air? Explain and illustrate with ray diagrams.
10. A skin diver wants to use an air-filled plastic bag as a converging lens for underwater
use. Sketch a suitable cross section for the bag.
11. What approximations were made in deriving the thin lens equation (Eq. 42-19):
1 1 1
-+-=-?
O i f
12. Under what conditions will a thin lens have a lateral magnification (a) of -1 and (b)
of +1?
13. How does the focal length of a glass lens for blue light compare with that for red
light, assuming the lens is (a) diverging and (b) converging?
14. Does the focal length of a lens depend on the medium in which the lens is immersed?
Is it possible for a given lens to act as a converging lens in one medium and a diverging
lens in another medium?
15. Are the following statements true for a glass lens in air? (a) A lens that is thicker
at the center than at the edges is a converging lens for parallel light. (b) A lens that is
thicker at the edges than at the center is a diverging lens for parallel light. Explain and
illustrate, using wavefronts.
16. Under what conditions would the lateral magnification, (m = -i/o) for lenses and
mirrors become infinite? Is there any practical significance to such a condition?
17. Light rays are reversible. Discuss the situation in terms of objects and images if
all rays in Figs. 42-10, 42-14, 42-17, 42-19, 42-23, and 42-25 are reversed in direction.
18. What significance can be given to the origin of coordinates in the graphical repre-
sentation of the thin-lens formula (Fig. 42-27)?
19. In connection with Fig. 42-24a, we pointed out that all rays originating on the same
wavefront in the incident wave have the same optical path length to the image point.
Discuss this in connection with Fermat's principle (Section 41-6).
PROBLEMS
1. Solve Example 2 if the angle between the mirrors is (a) 45°, (b) 60°, (c) 120°, the object
always being placed on the bisector of the mirrors.
2. Two plane mirrors make an angle of 90° with each other. What is the largest
number of images of an object placed between them that can be seen by a properly placed
eye? The object need not lie on the mirror bisector.
3. A small object is 10 cm in front of a plane mirror. If you stand behind the object,
30 cm from the mirror, and look at its image, for what distance must you focus your eyes?
4. A small object O is placed one-third of the way between two parallel plane mirrors
as in Fig. 42-28. Trace appropriate bundles of rays for viewing the four images that lie
closest to the object.
Fig. 42-28
1064 REFLECTION AND REFRAC'I'ION--II Chap. 42
5. Redraw Fig. 42-29 on a large sheet of paper and trace carefully the reflected rays,
using the law of reflection. Is a point focus formed? Discuss.
Fig. 42-29
6. Fill in this table, each column of which refers to a spherical mirror. Check your
results by graphical analysis. Distances are in centimeters; if a number has no plus or
minus sign in front of it, it may have either sign.
a b c d e f g h
f 20 +20 20
----
r -40 40
i -10 4
---
Real image? no
---
Erect image? no
7. A short linear object of length l lies on the axis of a spherical mirror, a distance o
from the mirror. (a) Show that its image will have a length l' where
l' = l (_f-)
0 -f
2
-
(b) Show that the longitudinal magnification m'( = l' /l) is equal to m 2 where mis the lateral
magnification discussed in Section 42-3. (c) Is there any condition such that, neglecting
all mirror defects, the image of a small cube would also be a cube?
8. As an example of the effect of relaxing the paraxial ray assumption find, analytically,
Chap. 42 PROBLEMS 1065
the apparent depth of a swimming pool having water (n = 1.33) of actual depth 8.0 ft (a)
when viewed at normal incidence and (b) when viewed at an angle of 30° from the normal.
9. A layer of water (n = 1.33) 2.0 cm thick floats on carbon tetrachloride (n = 1.46)
4.0 cm thick. How far below the water surface, viewed at normal incidence, does the
bottom of the tank seem to be?
~O. Fill out the following table, each column of which refers to a spherical surface
separating two media with different indices of refraction. Distances are measured m
centimeters.
a b c d e f g h
Real image?
Draw a figure for each situation and construct the appropriate rays graphically. Assume
a point object.
···11-... fill in this table, each column of which refers to a thin lens, to the extent possible.
Check .your results by graphical analysis. Distances are in centimeters; if a number
(except in row n) has no plus sign or minus sign in front of it, it may have either sign.
a b c d e f g h i
Type converging
------------ --- ------
f 10 +10 10 10
--------- ------------
i
------ --- ------------ ---
Draw a figure for each situation and construct the appropriate rays graphically. Assume
a finite object.
1066 REFLECTION AND REFRACTION-II Chap. 42
12. Define and locate the first and second focal points (see p.1058) for a single spherical
refracting surface such as that of Fig. 42-17.
13. A luminous object and a screen are a fixed distance D apart. (a) Show that a con-
verging lens of focal length f will
form a real image on the screen for
two positions that are separated by
d = VD(D - 4f).
(b) Show that the ratio of the two
image sizes for these two positions
is
f = hh .
h+h
16. Show that the distance between an object and its real image formed by a thin
converging lens is always greater than four times the focal length of the lens.
17. A double-convex lens is to be made of glass with an index of refraction of 1.50.
One surface is to have twice the radius of curvature of the other and the focal length is
to be 6.0 cm. What are the radii?
18. The formula
1 1 1
-+-=-
O i f
is called the Gaussian form of the thin lens formula. Another form of this formuia, the
Newtonian form, is obtained by considering the distance x from the object to the first focal
point and the distance x' from the second focal point to the image. Show that
xx' = f 2•
19. A parallel incident beam falls on a solid glass sphere at normal incidence. Locate
the image in terms of the index of refraction n and the sphere radius r.
20. An erect object is placed a distance in front of a converging lens equal to twice the
focal length Ji of the lens. On the other side of the lens is a converging mirror of focal
length h separated from the lens by a distance 2(!1 + f2). (a) Find the location, nature
and relative size of the final image. (b) Draw the appropriate ray diagram. See Fig.
42-31.
Fig. 42-31
Chap. 42 PROBLEMS 1067
+ '
Lens 1
It
Ii
Object
Fig. 42-32
23. An object is 20 cm to the left of a lens with a focal length of + 10 cm. A second lens
of focal length +12.5 cm is 30 cm to the right of the first lens. (a) Using the image formed
by the first lens as the object for the second, find the location and relative size of the final
image. (b) Verify your conclusions by drawing the lens system to scale and constructing
a ray diagram. (c) Describe the final image.
Interference
CHAPTER 43
Fig. 43-1 Showing how Thomas Young produced an interference pattern by allowing
diffracted waves from pinholes S1 and S2 to overlap on screen C.
x x x x
Fig. 43-2 Thomas Young's original drawing showing interference effects in overlapping
waves. Place the eye near the left edge and sight at a grazing angle along the figure.
(From Thomas Young, Phil. Transactions, 1803.)
1070 INTERFERENCE Chap. 43
tion of the wave; between them there is reinforcement. If a screen is placed
anywhere across the superimposed waves, we expect to find alternate bright
and dark spots on it. Figure 43-3 shows a photograph of such interjerence
fringes; in keeping with modern technique, long narrow slits rather than pin-
holes were used in preparing this figure.
Interference is not limited to light waves but is a characteristic of all wave
phenomena. Figure 43-4, for example, shows the interference pattern of
Fig. 43-4 The interference of water waves in a ripple tank. There is destructive inter-
ference along the lines marked "Line of nodes" and constructive interference between
these lines. (Courtesy Physical Science Study Committee.)
Sec. 43-1 YOUNG'S EXPERIMENT 1071
Fig. 43-5 Rays from S 1 and S 2 combine at P. The light falling on screen B has been
taken as parallel. Actually, D » d, the figure being distorted for clarity.
water waves in a shallow ripple tank. The waves are generated by two vibra-
tors that tap the water surface in synchronism, producing two expanding
spherical waves.
Let us analyze Young's experiment quantitatively, assuming that the
incident light consists of a single wavelength only. In Fig. 43-5 P is an
arbitrary point on the screen, a distance r 1 and r 2 from the narrow slits 8 1
and 8 2 , respectively. Let us draw a line from 8 2 to b in such a way that the
lines P8 2 and Pb are equal. If d, the slit spacing, is much smaller than the
distance D between the two screens (the ratio d/D in the figure has been
exaggerated for clarity), 8 2 b is then almost perpendicular to both r 1 and r 2 .
This means that angle 8 18 2 b is almost equal to angle PaO, both angles being
marked e in the figure. This is equivalent to saying that the lines r 1 and r 2
may be taken as parallel.
We often put a lens in front of the two slits, as in Fig. 43-6, the screen C
being in the focal plane of the lens. Under these conditions light focused at
P must have struck the lens parallel to the line Px, drawn from P through
the center of the (thin) lens. Under these conditions rays r 1 and r2 are
strictly parallel even though the requirement D » d is not met. The lens L
may in practice be the lens and cornea of the eye, screen C being the retina.
The two rays arriving at P in Figs. 43-5 or 43-6 from 8 1 and 8 2 are in
phase at the source slits, both being derived from the same wavefront in the
incident plane wave. Because the rays have different optical path lengths,
1072 INTERFERENCE Chap. 43
Fig. 43-6 A lens is normally used to produce interference fringes; compare with Fig. 43-5.
The figure is again distorted for clarity in that f » d in practice.
... Example 1. The double-slit arrangement in Fig. 43-5 is illuminated with light
from a mercury vapor lamp so filtered that only the strong green line (X = 5460 A) is
effective. The slits are 0.10 mm apart, and the screen on which the interference
pattern appears is 20 cm away. What is the angular position of the first minimum?
Of the tenth maximum?
At the first minimum we put m = 0 in Eq. 43-2, or
sin() = (m + f)X = (!)(546 X 10- 9 meter) = 0 _0027 _
d 0.10 X 10-3 meter
This value for sin () is so small that we can take it to be the value of 0, expressed in
radians; expressed in degrees it is O.16 °.
At the tenth maximum (not counting the central maximum) we must put m = 10
in Eq. 43-1. Doing so and calculating as before leads to an angular position of 3.8°.
For these conditions we see that the angular spread of the first dozen or so fringes is
small.
Example 2. In Example 1 what is the linear distance on screen C between ad-
jacent maxima?
If() is small enough, we can use the approximation
and Ym+I = (m + 1) XD
d.
Their separation b..y is found by subtracting:
XD
D..y = Ym+l - Ym = d
= (546 X 10-9 meter)(20 X 10- 2 meter) = 1.og mm.
0.10 X 10-3 meter
As long as() in Figs. 43-5 and 43-6 is small, the separation of the interference fringes is
independent of m; that is, the fringes are evenly spaced. Note that if the incident
light contains more than one wavelength the separate interference patterns, which
will have different fringe spacings, will be superimposed. ~
Equation 43-1 can be used to determine the wavelength of light; to quote Thomas
Young:
From a comparison of various experiments, it appears that the breadth of the
undulations [that is, the wavelength] constituting the extreme red light must be
supposed to be, in air, about one 36 thousandth of an inch, and those of the extreme
violet about one 60 thousandth; the mean of the whole spectrum, with respect to
the intensity of light, being one 45 thousandth.
1074 INTERFERENCE Chap. 43
Young's value for the average effective wavelength present in sunlight (1/45,000 in.)
can be written as 5700 A, which agrees rather well with the wavelength at which the
eye sensitivity is a maximum, 5550 A (see Fig. 40-2). It must not be supposed that
Young's work was received without criticism. One of his contemporaries, evidently
a firm believer in the corpuscular theory of light, wrote:
We wish to raise our feeble voice against innovations that can have no other
effect than to check the progress of science, and renew all those wild phantoms of
the imagination which Bacon and Newton put to flight from her temple. This
paper contains nothing that deserves the name of either experiment or discovery.
Needless to say, posterity has decided in favor of Young.
43-2 Coherence
Analysis of the derivation of Eqs. 43-1 and 43-2 shows that a fundamental
requirement for the existence of well-defined interference fringes on screen C
in Fig. 43-1 is that the light waves that travel from 8 1 and 8 2 to any point P
on this screen must have a sharply defined phase difference cp that remains
constant with time. If this condition is satisfied, a stable, well-defined
fringe pattern will appear. At certain points P, cp will be given, independ-
ent of time, by n1r where n = 1, 3, 5, ... so that the resultant intensity
will be strictly zero and will remain so throughout the time of observation.
At other points cp will be given by n1r where n = 0, 2, 4 ... and the result-
ant intensity will be a maximum. Under these conditions the two beams
emerging from slit 8 1 and 8 2 are said to be completely coherent.
Let the source in Fig. 43-1 be removed and let slits 8 1 and 8 2 be replaced
by two completely independent light sources, such as two fine incandescent
wires placed side by side in a glass envelope. No interference fringes will
appear on screen C but only a relatively uniform illumination. We can in-
terpret this if we make the reasonable assumption that for completely inde-
pendent light sources the phase difference between the two beams arriving
at P will vary with time in a random way. At a certain instant conditions
may be right for cancellation and a short time later (perhaps 10-s sec) they
may be right for re-enforcement. This same random phase behavior holds
for all points on screen C with the result that this screen is uniformly illu-
minated. The intensity at any point is equal to the sum of the intensities
that each source 8 1 and 8 2 produces separately at that point. Under these
conditions the two beams emerging from 8 1 and 8 2 are said to be completely
incoherent.
Note that for completely coherent light beams one (1) combines the
amplitudes vectorially, taking the (constant) phase difference properly into
account, and then (2) squares this resultant amplitude to obtain a quantity
proportional to the resultant intensity. For completely incoherent light
beams, on the other hand, one (1) squares the individual amplitudes to
obtain quantities proportional to the individual intensities and then (2) adds
the individual intensities to obtain the resultant intensity. This procedure
is in agreement with the experimental fact that for completely independent
light sources the resultant intensity at every point is always greater than
the intensity produced at that point by either light source acting alone.
Sec. 43-2 COHERENCE 1075
(\ (\ (\ (\ (\ (\ (\ (\ (\ (\
(b)
Fig. 43-8 A screen with two small holes is placed against the end of an optical maser.
(a) The light passing through the holes forms an interference pattern on a strip of photo-
graphic film placed at (b). The fact that the pattern is formed shows that the light
emitted from the maser is coherent across the beam cross section. At (b) is shown the
image as formed on a photographic strip exposed in this manner. At (c) is an intensity
plot of the film made by measuring the degree to which the film has been exposed. (Cour-
tesy of D. F. Nelson and R. J. Collins, Bell Telephone Laboratories.)
E = E1 + E2 (43-5)
where (43-6b)
Em, the maximum possible amplitude for Eo, is equal to twice the amplitude
of the combining waves ( = 2E0 ), corresponding to complete reinforcement.
The student should verify Eq. 43-6 carefully. The amplitude Eo of the
resultant wave disturbance, which determines the intensity of the inter-
ference fringes, will turn out to depend strongly on the value of 8, that is, on
the location of point P in Figs. 43-5 and 43-6.
In Section 19-6 we showed that the intensity of a wave I, measured per-
haps in watts/meter 2 , is proportional to the square of its amplitude. For
the resultant wave then, ignoring the proportionality constant,
Io ex E/. (43-7)
This relationship seems reasonable if we recall (Eq. 30-27) that the energy
density in an electric field is proportional to the square of the electric field
strength. This is true for rapidly varying electric fields, such as those in a
light wave, as well as for static fields.
The ratio of the intensities of two light waves is the ratio of the squares of
the amplitudes of their electric fields. If I O is the intensity of the resultant
wave at P and I O is the intensity that a single wave acting alone would
produce, then
Io=
I0
(EEo0) 2
(43-8)
* The electric field E in the light wave rather than the magnetic field B is normally
identified with the "wave disturbance" because the effects of Bon the human eye and on
various light detectors are exceedingly small. Radiation pressure (Section 40-2) is one such
effect. Note too that, although Eq. 43-5 should be a vector equation, in most cases of
interest the E vectors in the two interfering waves are closely parallel so that an algebraic
equation suffices.
Sec. 43-3 INTENSITY IN YOUNG'S EXPERIMENT 1079
Note that the intensity of the resultant wave at any point P varies from
zero [for a point at which <j, ( = 2/3) = 1r, say] to Im, which is four times the
intensity 10 of each individual wave [for a point at which <j, ( = 2/3) = 0, say].
Let us compute I e as a function of the angle 8 in Figs. 43-5 or 43-6.
The phase difference <j, in Eq. 43-4 is associated with the path difference
Sib in Fig. 43-5 or 43-6. If Sib is tA, <j, will be 1r; if S 1 b is A, <j, will be 21r,
etc. This suggests that
21r
<j, = - (d sin 8),
A
or, finally, from Eq. 43-6b,
1rd
~ = !cf> = - sin 8. (43-10)
A
This expression for /3 can be substituted into Eq. 43-9 for 18, yielding the
latter quantity as a function of 8. For convenience we collect here the ex-
pressions for the amplitude and the intensity in double-slit interference.
m = 0, 1, 2, ...
in Eq. 43-llb. From Eq. 43-llc this reduces to
which is the equation derived in Section 43-1 (Eq. 43-1). To find the in-
tensity minima we write
1rd sin O
- - - = (m + !)1r m = 0, 1, 2, ... (minima),
A
which reduces to the previously derived Eq. 43-2.
Figure 43-9 shows the intensity pattern for double-slit interference. The
horizontal solid line is 10 ; this describes the (uniform) intensity pattern on
the screen if one of the slits is covered up. If the two sources were incoherent
the intensity would be uniform over the screen and would be 210 ; see the
horizontal dashed line in Fig. 43-9. For coherent sources we expect the
1080 INTERFERENCE Chap. 43
~
"vi
I ~twio;r
!\--A-
- --4Io(= Im)
~
" " \ ~ \ / \ ~ \ / ~ L ~ Io(one source)
~-A~2To(t;,~o~~;~~;-
(jJ 57!" 471" 37!" 271" 7r O 7r 271" 371" 471" 57!"
E. l_ 2 ],._ 3 O l_ E._ l_
sin O (see Eq. 43- 11 c) 2 d d 2 aA d
A 1 A
2 d
1 A
2 ,I cf
A 3 A
2 d 2d 2 d
m(Eq.43-1) 2 1 0 1 2 maxima
m (Eq. 43-2) 2 1 0 0 1 2 minima
Fig. 43-9 The intensity pattern for double-slit interference. The heavy arrow in the
central peak represents the half-width of the peak. This figure is constructed on the
assumption that the two interfering waves each illuminate the central portion of the
screen uniformly, that is, / 0 is independent of position as shown.
E 1 = E 0 sin wt (43-3)
which have the same angular frequency wand amplitude E 0 but which have
a phase difference tj, between them. In this case the result (Eqs. 43-lla and
c) is easily obtained algebraically.
In later chapters we will want to add larger numbers of wave disturbances,
often an infinite number, with infinitesimal individual amplitudes. Since
analytic methods become more difficult in such cases we describe a graphical
method, illustrating it by rederiving Eq. 43-lla.
A sinusoidal wave disturbance such as that represented by Eq. 43-3 can
be represented graphically, using a rotating vector. In Fig. 43-lOa a vector
of magnitude E 0 is allowed to rotate about the origin in a counterclockwise
direction with an angular frequency w. Following electrical engineering
practice we call such a rotating vector a phasor. The alternating wave dis-
turbance E 1 (Eq. 43-3) is represented by the projection of this phasor on the
vertical axis.
A second wave disturbance E 2 , which has the same amplitude E 0 but a
Sec. 43-4 ADDING WAVE DISTURBANCES 1081
(a) (b)
(c)
1082 INTERFERENCE Chap. 43
Eo
[)
Eo ~
(b)
.E
Fig. 43-12 Example 3. Four wave dis-
turbances are added graphically, using
the method of phasors.
Fig. 43-14 Interference by reflection from a thin film, assuming an extended source S.
different paths in going from P to the eye, one being reflected from the upper
surface of the film, the other from the lower surface. Whether point a ap-
pears bright or dark depends on the nature of the interference between the
two waves that diverge from a. These waves are coherent because they both
originate from the same point Pon the light source.
If the eye looks at another part of the film, say a', the light that enters the
eye must originate from a different point P' of the extended source, as sug-
gested by the dashed lines in Fig. 43-14.
For near-normal incidence (0 ,...._, 0 in Fig. 43-14) the geometrical path dif-
ference for the two rays from P will be close to 2d. We might expect the re-
sultant wave reflected from the film near a to be an interference maximum if
the distance 2d is an integral number of wavelengths. This statement must
be modified for two reasons.
First, the wavelength must refer to the wavelength of the light in the film
An and not to its wavelength in air A; that is, we are concerned with optical
path lengths rather than geometrical path lengths. The wavelengths A and
An (see Eq. 41-11) are related by
An = A/n. (43-12)
To bring out the second point, let us assume th~t the film is so thin that 2d
is very much less than a wavelength. The phase difference between the two
Sec. 43-5 INTERFERENCE FROM THIN FILMS 1085
waves would be close to zero on our assumption, and we would expect such a
film to appear bright on reflection. However, it appears dark. This is clear
from Fig. 43-13, in which the action of gravity produces a wedge-shaped
film, extremely thin at its top edge. As drainage continues, the dark area
increases in size. To explain this and many similar phenomena, we assume
that one or the other of the two rays of Fig. 43-14 suffers an abrupt phase
change of 1r ( = 180°) associated either with reflection at the air-film interface
or transmission through it. As it turns out, the ray reflected from the upper
surface suffers this phase change. The other ray is not changed abruptly in
phase, either on transmission through the upper surface or on reflection at
the lower surface.
In Section 19-9 we discussed phase changes on reflection for transverse
waves in strings. To extend these ideas, consider the composite string of
Fig. 43-15, which consists of two parts with different masses per unit length,
stretched to a given tension. If a pulse moves to the right in Fig. 43-15a,
approaching the junction, there will be a reflected and a transmitted pulse,
-I>
Initial I\ <}- ---!>
Final
Fig. 43-15 Phase changes on reflection at a
(a)
junction between two stretched composite
strings. (a) Incident pulse in heavy string ---{>
the reflected pulse being in phase with the incident pulse. In Fig. 43-15b
the situation is reversed, the incident pulse now being in the less massive
string. In this case the reflected pulse will differ in phase from the incident
pulse by 1r ( = 180°). In each case the transmitted pulse will be in phase with
the incident pulse.
Figure 43-15a suggests a light wave in glass, say, approaching a surface
beyond which there is a less optically dense medium (one of lower index of
refraction) such as air. Figure 43-15b suggests a light wave in air approach-
ing glass. To sum up the optical situation, when reflection occurs from an
interface beyond which the medium has a lower index of refraction, the re-
flected wave undergoes no phase change; when the medium beyond the inter-
face has a higher index, there is a phase change of 1r. * The transmitted wave
does not experience a change of phase in either case.
* These statements, which can be proved rigorously from Maxwell's equations (see also
Section 43-6), must be modified for light falling on a less dense medium at an angle such
that total internal reflection occurs. They must also be modified for reflection from
metallic surfaces.
1086 INTERFERENCE Chap. 43
We are now able to take into account both factors that determine the na-
ture of the interference, namely, differences in optical path length and phase
changes on reflection. For the two rays of Fig. 43-14 to combine to give a
maximum intensity, assuming normal incidence, we must have
2d = (m + t)An m = 0, 1, 2, ....
These equations hold when the index of refraction of the film is either greater
or less than the indices of the media on each side of the film. Only in these
cases will there be a relative phase change of 180° for reflections at the two
surfaces. A water film in air and an air film in the space between two glass
plates provide examples of cases to which Eqs. 43-13 and 43-14 apply.
Example 5 provides a case in which they do not apply.
If the film thickness is not uniform, as in Fig. 43-13, where the film is
wedge-shaped, constructive interference will occur in certain parts of the
film and destructive interference will occur in others. Lines of maximum
and of minimum intensity will appear-these are the interference fringes.
They are called fringes of constant thickness, each fringe being the locus of
points for which the film thickness dis a constant. If the film is illuminated
with white light rather than monochromatic light, the light reflected from
various parts of the film will be modified by the various constructive or de-
structive interferences that occur. This accounts for the brilliant colors of
soap bubbles and oil slicks.
Only if the film is "thin," which implies that d is no more than a few wavelengths
of light, will fringes of the type described, that is, fringes that appear localized on the
film and associated with a variable film thickness, be possible. For very thick films
(say d ,.._., 1 cm), the path difference between the two rays of Fig. 43-14 will be many
wavelengths and the phase difference at a given point on the film will change rapidly
as we move even a small distance away from a. For "thin" films, however, the phase
difference at a also holds for reasonably nearby points; there is a characteristic "patch
brightness" for any point on the film, as Fig. 43-13 shows. Interference fringes can
be produced for thick films; they are not localized on the film but are at infinity. See
Section 43-7 .
A = 8500 A (minima).
m
Maxima and minima occur for the following wavelengths:
Only the maximum corresponding tom = l lies in the visible region (see Fig. 40-2);
light of this wavelength appears yellow-green. If white light is used to illuminate the
film, the yellow-green component will be enhanced when viewed by reflection.
Example 5. N onrefiecting glass. Lenses are often coated with thin films of trans-
parent substances like MgF 2 (n = 1.38) in order to reduce the reflection from the
glass surface, using interference. How thick a coating is needed to produce a mini-
mum reflection at the center of the visible spectrum (5500 A)?
We assume that the light strikes the lens at near-normal incidence (8 is exaggerated
for clarity in Fig. 43-16), and we seek destructive interference between rays rand r 1 •
Equation 43-14 does not apply because in this case a phase change of 180° is asso-
ciated with each ray, for at both the upper and lower surfaces of the MgF 2 film the
reflection is from a medium of greater index of refraction.
There is no net change in phase produced by the two reflections, which means that
the optical path difference for destructive interference is (m + i)A (compare Eq.
43-13), leading to
2dn = (m + })A m = 0, 1, 2, ... (minima).
Solving ford and putting m = 0 yields
d = (m +!)A=~= 5500 A = lOOO A.
2n 4n (4)(1.38)
Fig. 43-16 Example 5. Unwanted reflections from glass can be reduced by coating the
glass with a thin transparent film.
1088 INTERFERENCE Chap. 43
Example 6. Newton's rings. Figure 43-17 shows a lens of radius of curvature R
resting on an accurately plane glass plate and illuminated from above by light of
wavelength A. Figure 43-18 shows that circular interference fringes (Newton's rings)
appear, associated with the variable thickness air film between the lens and the plate.
Find the radii of the circular interference maxima.
Here it is the ray from the bottom of the (air) film rather than from the top that
undergoes a phase change of 180°, for it is the one reflected from a medium of higher
refractive index. The condition for a maximum remains unchanged, however (Eq.
43-13), and is
2d = (m+ !)>- m = 0, 1, 2, ... , (43-15)
the index of refraction of the air film being assumed to be unity. From Fig. 43-17
we can write
d = R - VR2- r2 = R - R [1- (i )2] . Yz
If r/R << 1, the square bracket can be expanded by the binomial theorem, keeping
only two terms, or
I ( r
d= R - R [1- 2 R )2 + .. .J"-' 2R.
r2
Combining with Eq. 43-15 yields
Fig. 43-18 Example 6. Newton's rings. (Courtesy of Bausch and Lomb Optical Co.)
which gives the radii of the bright rings. If white light is used, each spectrum com-
ponent will produce its own set of circular fringes, the sets all overlapping. ....
Fig. 43-19 (a) A ray is reflected and refracted at an air-glass interface. (b) The optically
reversed situation; the two rays in the lower left must cancel.
r12 2E and r12t12E. Ray t12E, identified by the triple arrows, is also reflected and re-
fracted, producing the rays of amplitudes t12t21E and t12r21E as shown. Note that r12
describes a ray in medium 1 reflected from medium 2, and r 21 describes a ray in
medium 2 reflected from medium 1. Similarly, t12 describes a ray that passes from
medium 1 to medium 2; b describes a ray that passes from medium 2 to medium 1.
The two rays in the upper left of Fig. 43-19b must be equivalent to the incident ray
of Fig. 43-19a, reversed; the two rays in the lower left of Fig. 43-19b must cancel.
This second requirement leads to
rd12E + t12r21E = 0,
or
This result tells us that if we compare a wave reflected from medium one with one
reflected from medium 2, they behave differently in that one or the other undergoes a
phase change of 180°. We must rely on experiment to show that, as we pointed out
earlier, the ray reflected from the more optically dense medium is the one that ex-
periences the phase change of 180°.
are reflected at each of these mirrors and are sent back along their directions
of incidence, each wave eventually entering the eye E. Since the waves are
coherent, being derived from the same point on the source, they will inter-
fere.
If the mirrors M 1 and M 2 are exactly perpendicular to each other, the effect
is that of light from an extended source S falling on a uniformly thick slab of
air, between glass, whose thickness is equal to d 2 - d 1 . Interference fringes
appear, caused by small changes in the angle of incidence of the light from
different points on the extended source as it strikes the equivalent air film.
For thick films a path difference of one wavelength can be brought about by a
very small change in the angle of incidence.
If 1l1.2 is moved backward or forward, the effect is to change the thickness
of the equivalent air film. Suppose that the center of the (circular) fringe
pattern appears bright and that M 2 is moved just enough to cause the first
bright circular fringe to move to the center of the pattern. The path of the
light beam striking M 2 has been changed by one wavelength. This means
(because the light passes twice through the equivalent air film) that the
mirror must have moved one-half a wavelength.
The interferometer is used to measure changes in length by counting the
number of interference fringes that pass the field of view as mirror M 2 is
moved. Length measurements made in this way can be accurate if large
numbers of fringes are counted.
Fig. 43-20 Michelson's interferometer, showing the path of a particular ray originating
at point P of an extended source S.
1092 INTERFERENCE Chap. 43
Michelson measured the length of the standard meter, kept in Paris, in
terms of the wavelength of certain monochromatic red light emitted from a
light source containing cadmium. He showed that the standard meter was
equivalent to 1,553,163.5 wavelengths of the red cadmium light.
Physicists have long speculated on the advantages of discarding the
standard meter bar as the basic standard of length and of defining the meter
in terms of the wavelength of some carefully chosen monochromatic radia-
tion. This would make the primary length standard readily available in
laboratories all over the world. It would improve the accuracy of length
measurements, since one would no longer need to compare an unknown ob-
ject with a standard object (the meter bar), using interferometer techniques,
but could measure the unknown object directly and in an absolute sense, using
these techniques. There is the additional advantage that if the standard
meter bar were destroyed it could never be replaced, whereas light sources
and interferometers will (presumably) always be available.
In 1961 such an atomic standard of length was adopted by international
agreement. Quoting from an article * describing the event:
The wavelength of the orange-red light of krypton-86 has replaced the platinum
iridium bar as the world standard of length. Formerly the wavelength of this light
was defined as a function of the length of the meter bar. Now the meter is defined
as a multiple (1,650,763.73) of the wavelength of the light.
The light from krypton-86 was used in preference to that from cadmium or
other sources because it produces sharper interference fringes in the inter-
ferometer over the long optical paths sometimes used in length measurement.
Fig. 43-21 The "ether" is streaming with velocity - u through Michelson's interferom-
eter. The wave speeds shown are on the basis of the (incorrect) ether hypothesis.
rowing a boat a distanced downstream and the same distance upstream; the second
corresponds to rowing a boat a distanced across a stream and back.
On the ether hypothesis the speed of light on the path MM1 is c + u; on the return
path M 1M it is c - u. The time required for the complete trip is
d 2c 2d 1
l1 = - - +c--d-u =
c+u
d2- - -2
c - u c 1 - (u/c) 2
Assuming u/ c << 1, we can expand the quantities in the square brackets by using the
binomial theorem, retaining only the first two terms. This leads to
= 2d
c
{!2 ('.!!:)c 2
} = du2.
c3 (43-16)
1094 INTERFERENCE Chap. 43
Now let the entire interferometer be rotated through 90°. This will interchange
the roles of the two light paths, MM 1M now being the "cross-stream" path and
MM2M the "down- and upstream" path. The time difference between the two waves
entering the eye is also reversed; this changes the phase difference between the com-
bining waves and alters the positions of the interference maxima. The experiment
consists of looking for a shift of the interjerence fringes as the apparatus is rotated.
The change in time difference is 2i:it, which corresponds to a fringe shift of 2i:it/T
where T ( = A/c) is the period of vibration of the light. The expected maximum shift
in the number of fringes on a 90° rotation (see Eq. 43-16) is
i:iN = !~ =
T
2i:itc = 2d
"'A X
(!±)c 2
• (43-17)
i:iN = 2d
A
(!±)
c
2
= (2)(11 meters) (I0- 4) 2 = 0.4.
5.9 X 10- 7 meter
Even though a shift of only about 0.4 of a fringe was expected, Michelson and Morley
were confident that they could observe a shift of 0.01 fringe. They found from their
experiment, however, that there was no observable f ririge shift!
The analogy between a light wave in the supposed ether and a boat moving in
water, which seemed so evident in 1881, is simply incorrect. The derivation based on
this analogy is incorrect for light waves. When the analysis is carried through on
Einstein's hypothesis, the observed negative result is clearly predicted, the speed of
light being c for all paths. The motion of the earth around the sun and the rotation
of the interferometer have, in Einstein's view, no effect whatever on the speed of the
light waves in the interferometer.
It should be made clear that although Einstein's hypothesis is completely con-
sistent with the negative result of the Michelson-Morley experiment this experiment
standing alone cannot serve as a proof for Einstein's hypothesis. Einstein said that
no number of experiments, however large, could prove him right but that a single
experiment could prove him wrong. Our present-day belief in Einstein's hypothesis
rests on consistent agreement in a large number of experiments designed to test it.
The "single experiment" that might prove Einstein wrong has never been found.
Chap. 43 QUESTIONS 1095
QUESTIONS
PROBLEMS
Fig. 43-22
,,,'3. A double-slit arrangement produces interference fringes for sodium light (X = 5890 A)
that are 0.20° apart. What is the angular fringe separation if the entire arrangement is
immersed in water?
4. A double-slit arrangement produces interference fringes for sodium light (X = 5890 A)
that are 0.20° apart. For what wavelength would the angular separation be 10% greater?
&, A thin flake of mica (n = 1.58) is used to cover one slit of a double-slit arrangement.
The central point on the screen is occupied by what used to be the seventh bright fringe.
If X = 5500 A, what is the thickness of the mica?
6. Sodium light (X = 5890 A) falls on a double slit of separation d = 2.0 mm. D in
Fig. 43-5 is 4 cm. What per cent error is made in locating the tenth bright fringe if it is
not assumed that D » d?
,,7'., Sodium light (X = 5890 A) falls on a double slit of separation d = 0.20 mm. A thin
lens (f = + 1.0 meter) is placed near the slit as in Fig. 43-6. What is the linear fringe
separation on a screen placed in the focal plane of the lens?
Chap. 43 PROBLEMS 1097
··. 8. S 1 and S2 in Fig. 43-23 are effective point sources of radiation, excited by the same
oscillator. They are coherent and in phase with each other. Placed 4.0 meters apart,
they emit equal amounts of power in the form of LO-meter wavelength electromagnetic
waves. (a) Find the positiorn, of the first (that is, the nearest), the second, and the third
maxima of the received signal, as the detector is moved out along Ox. (b) Is the intensity
at the nearest minimum equal to zero? Justify your answer.
r 82
d
l ____ =t------X
Fig. 43-23
9. One of the slits of a double-slit system is wider than the other, so that the amplitude
of the light reaching the central part of the screen from one slit, acting alone, is twice that
from the other slit, acting alone. Derive an expression for Io in terms of e, corresponding
to Eqs. 43-llb and c.
10. Show that the half-width D.0 of the double-slit interference fringes (see arrow in Fig.
43-9) is given by
'A
D.0 =-
2d
if e is small enough so
that sine ,...., e.
11. Find the sum of the following quantities (a) by the vector method and (b) analyti-
cally:
Y1 = 10 sin wt
Y1 = 10 sin wt
Fig. 43-24
17. A thin film 4 X 10-5 cm thick is illuminated by white light normal to its surface.
Its index of refraction is 1.5. What wavelengths within the visible spectrum will be
intensified in the reflected beam?
18. White light reflected at perpendicular incidence from a soap film has, in the visible
spectrum, an interference maximum at 6000 A and a minimum at 4500 A, with no mini-
mum in between. If n = l.33 for the film, what is the film thickness, assumed uniform?
19. A plane wave of monochromatic light falls normally on a uniformly thin film of oil
which covers a glass plate. The wavelength of the source can be varied continuously.
Complete destructive interference of the reflected light is observed for wavelengths of
5000 and 7000 A and for no other wavelengths in between. If the index of refraction of
the oil is l.30 and that of the glass is 1.50, find the thickness of the oil film.
20. In Example 5 assume that there is zero reflection for light of wavelength 5500 A
at normal incidence. Calculate the factor by which the reflection is diminished by the
coating at 4500 and at 6500 A.
21. A plane monochromatic light wave in air falls at normal incidence on a thin film of
oil which covers a glass plate. The wavelength of the source may be varied continuously.
Complete destructive interference in the reflected beam is observed for wavelengths of
5000 and 7000 A and for no other wavelength in between. The index of refraction of glass
is 1.50. Show that the index of refraction of the oil must be less than 1.50.
22. White light reflected at perpendicular incidence from a soap bubble has, in the visi-
ble spectrum, a single interference maximum (at A = 6000 A) and a single minimum at the
violet end of the spectrum. If n = 1.33 for the film, calculate its thickness.
23. If mirror M 2 in Michelson's interferometer is moved through 0.233 mm, 792 fringes
are counted. What is the wavelength of the light?
24. A thin film with n = 1.40 for light of wavelength 5890 A is placed in one arm of a
Michelson interferometer. If a shift of 7.0 fringes occurs, what is the film thickness?
25. (a) What is the wavelength of the orange-red line of krypton-86 in angstrom units?
Take 1 A = 10- 10 meter exactly. (b) Does this question really make sense, in view of the
fact that the meter is defined in terms of this wavelength? Explain.
26. A Michelson interferometer is used with a sodium discharge tube as a light source.
The yellow sodium light consists of two wavelengths, 5890 and 5896 A. It is observed
that the interference pattern disappears and reappears periodically as one moves mirror
M2 in Fig. 43-20. (a) Explain this effect. (b) Calculate the change in path difference
between two successive reappearances of the interference pattern.
Diffraction
CHAPTER 44
44-1 Introduction
Diffraction, which is illustrated in Fig. 42-3, is the bending of light around
an obstacle such as the edge of a slit. 'Ne can see the diffraction of light by
looking through a crack between two fingers at a distant light source such
as a tubular neon sign or by looking at a street light through a cloth um-
brella. Usually diffraction effects are small and must be looked for carefully.
Also, most sources of light have an extended area so that a diffraction pat-
tern produced by one point of the source will overlap that produced by
another. Finally, common sources of light are not monochromatic. The
patterns for the various wavelengths overlap and again the effect is less
apparent.
Diffraction was discovered by Francesco Maria Grimaldi (1618-1663),
and the phenomenon was known both to Huygens (1629-1695) and to New-
ton (1642-1727). Newton did not see in it any justification for a wave
theory for light. Huygens, although he believed in a wave theory, did not
believe in diffraction! He imagined his secondary wavelets to be effective
only at the point of tangency to their common envelope, thus denying the
possibility of diffraction. In his words:
And thus we see the reasons why light . . . proceeds only in straight lines in
such a way that it does not illuminate any object except when the path from the
source to the object is open along such a line.
Fresnel (1788-1827) correctly applied Huygens' principle (which is called
the Huygens-Fresnel principle in Europe) to explain diffraction. In these
early days the light waves were believed to be mechanical waves in an all-
1099
1100 DIFFRACTION Chap. 44
pervading ether. \Ve have seen (Section 39-5) how Maxwell (1831-1879)
showed that light waves were not mechanical in nature but electromagnetic.
Einstein (1879-1955) rounded out our modern view of light waves by elim-
inating the need to postulate an ether (see Section 40-4).
Figure 44-1 shows a general diffraction situation. Surface A is a wavefront
that falls on B, which is an opaque screen containing an aperture of arbitrary
shape; C is a diffusing screen that receives the light that passes through this
aperture. This pattern of light intensity on C can be calculated by sub-
dividing the wavefront into elementary areas dS, each of which becomes a
source of an expanding Huygens' wavelet. The light intensity at an arbi-
trary point Pis found by superimposing the wave disturbances (that is, the
E vectors) caused by the wavelets reaching P from all these elementary
radiators.
The wave disturbances reaching P differ in amplitude and in phase be-
cause (a) the elementary radiators are at varying distances from P, (b) the
light leaves the radiators at various angles to the normal to the wavefront
(&ee p. 1020), and (c) soni.e radiators are blocked by screen B; others are not.
Diffraction calculations-simple in principle-may become difficult in prac-
tice. The calculation must be repeated for every point on screen Cat which
we wish to know the light intensity. We followed exactly this program in
calculating the double-slit intensity pattern in Section 43-3. The calcula-
tion there was simple because we assumed only two elementary radiators,
the two narrow slits.
Figure 44-2a shows the general case of Fresnel diffraction, in which the
light source and/ or the screen on which the diffraction pattern is displayed
are a finite distance from the diffracting aperture; the wavefronts that fall
on the diffracting aperture in this case and that leave it to illuminate any
point P of the diffusing screen are not planes; the corresponding rays are
not parallel.
Sec. 44-1 INTRODUCTION 1101
Fig. 44-2 (a) Fresnel diffraction. (b) Source S and screen Care moved to a large dis-
tance, resulting in Fraunhofer diffraction. (c) Fraunhofer diffraction conditions produced
by lenses, leaving source S and screen C in their original positions.
1102 DIFFRACTION Chap. 44
Fig. 44-2c. The first of these converts the diverging wave from the
source into a plane wave. The second lens causes plane waves leaving
the diffracting aperture to converge to point P. All rays that illuminate
P will leave the diffracting aperture parallel to the dashed line Px drawn
from P through the center of this second ( thin) lens. We assumed
Fraunhofer conditions for Young's double-slit experiment in Section 43-1
(see Fig. 43-5).
Although Fraunhofer diffraction is a limiting case of the more general
Fresnel diffraction, it is an important limiting case and is easier to handle
mathematically. This book deals only with Fraunhofer diffraction.
44-2 Single Slit
Figure 44-3 shows a plane wave falling at normal incidence on a long
narrow slit of width a. Let us focus our attention on the central point P 0 of
screen C. The rays extending from the slit to P 0 all have the same optical
path lengths, as we saw in Section 42-5. Since they are in phase at the
plane of the slit, they will still be in phase at P 0 , and the central point of
the diffraction pattern that appears on screen Chas a maximum intensity.
Fig. 44-3 Conditions at the central maximum of the diffraction pattern. The slit extends
a distance above and below the figure, this distance being much greater than the slit
width a.
Sec. 44-2 SINGLE SLIT 1103
We now consider another point on the screen. Light rays which reach P 1
in Fig. 44-4 leave the slit at an angle () as shown. Ray r 1 originates at the
top of the slit and ray r 2 at its center. If() is chosen so that the distance bb'
in the figure is one-half a wavelength, r 1 and r 2 will be out of phase and will
produce no effect at P 1 . * In fact, every ray from the upper half of the slit
will be canceled by a ray from the lower half, originating at a point a/2 below
the first ray. The point P 1 , the first minimum of the diffraction pattern,
will have zero intensity (compare Fig. 42-3).
The condition shown in Fig. 44-4 is
a A
- sin() = -,
2 2
or a sin() = A. (44-1)
As we stated earlier (see Fig. 42-1), the central maximum becomes wider as
the slit is made narrower. If the slit width is as small as one wavelength
(a = }..), the first minimum occurs at() = 90°, which implies that the central
maximum fills the entire forward hemisphere. We assumed a condition ap-
* Whatever phase relation exists between r1 and r2 at the plane represented by the
sloping dashed line in Fig. 44-4 that passes through b' also exists at P1, not being affected
by the lens (see Section 42-5).
1104 DIFFRACTION Chap. 44
or 2 (44-3)
Li¢ = ( ; ) (Lix sin fJ),
1106 DIFFRACTION Chap. 44
Fig. 44-6 A slit of width a is divided into N strips of width ~x. The insert shows condi-
tions at the second strip more clearly. In the differential limit the slit is divided into an
infinite number of strips (that is, N ----+ co) of differential width dx. For clarity in this
and the following figure, we take N = 18.
where Lix sin() is, as the figure insert sho,vs, the path difference for rays origi-
nating at the top edges of adjacent strips. Thus, at P, N vectors with the
same amplitude LiE0 , the same frequency, and the same phase difference Li</)
between adjacent members combine to produce a resultant disturbance. We
ask, for various values of Li</) [that is, for various points Pon the screen, cor-
responding to various values of e (see Eq. 44-3)], what is the amplitude Ee
of the resultant wave disturbance? We find the answer by representing the
individual wave disturbances LiE0 by phasors and calculating the resultant
phasor amplitude, as described in Section 43-4.
At the center of the diffraction pattern () equals zero, and the phase shift
between adjacent strips (see Eq. 44-3) is also zero. As Fig. 44-7a shows,
the phasor arrows in this case are laid end to end and the amplitude of the
resultant has its maximum value Em. This corresponds to the center of the
central maximum.
Sec. 44-3 SINGLE SLIT-QUALITATIVE 1107
(a)
@ I
(d)
1108 DIFFRACTION Chap. 44
tensity. The amplitudes must be squared to obtain the corresponding rela-
tive intensities (see Eq. 43-7).
Ee = 2R sin!.
2
Em
<p = -·
R
\a
a\
\
<P
\
\ R
\
\
\ Fig. 44-8 A construction used to
R \ calculate the intensity in single-slit
\
\ diffraction. The situation corre-
\ sponds to that of Fig. 44-7.
\
\
\
\
\
\
Sec. 44-4 SINGLE SLIT-QUANTITATIVE 1109
Em. c/J
Combining yields Eo = -sin-,
cp/2 2
sma
or Eo = Em--, (44-4)
a
cp
in which a=-· (44-5)
2
From Fig. 44-6, recalling that cp is the phase difference between rays from
the top and the bottom of the slit and that the path difference for these rays
is a sin (J, we have
or cp = ( 2; ) (a sin 0).
Combining with Eq. 44-5 yields
cp ~a .
a= - = -sm8. (44-6)
2 A
Equation 44-4, taken together with the definition of Eq. 44-6, gives the
amplitude of the wave disturbance for a single-slit diffraction pattern at any
angle 0. The intensity Io for the pattern is proportional to the square of the
amplitude, or
Io = Im ( si: a Y· (
44-7)
For convenience we display together, and renumber, the formulas for the
amplitude and the intensity in single-slit diffraction.
sma
[Eq. 44-4] Eo = Em-- (44-Sa)
a
single-
[Eq. 44-7] Io= Im (sin
-- ay slit (44-Sb)
a
diffraction
~a
[Eq. 44-6] a (= !c/J) = -sin() (44-Sc)
A J
Figure 44-9 shows plots of Io for several values of the ratio a/A. Note that
the pattern becomes narrower as a/A is increased; compare this figure with
Figs. 42-1 and 42-3.
Minima occur in Eq. 44-Sb when
a= m~ m = 1, 2, 3, .... (44-9)
1110 DIFFRACTION Chap. 44
1.0
0.8
->,
.vi a=X
0.6
0.4
-c:
Q)
c:
Q)
>
:;:::;
<ti
ai
0:::
0.2 (a)
15 10 5 0 5 10 15
(J, degrees
15 10 5 0 10 15
(J, degrees
a= lOX
(c)
15 10 5 0 5 10 15
(J, degrees
Fig. 44-9 The relative intensity in single-slit diffraction for three values of the ratio a/'A.
The arrow in (b) shows the half-width !1(J of the central maximum.
I = I [sin (m + !)11'] 2 ,
o m (m + !)11'
Io 1
which reduces to
Im - (m+!)211'2
This yields, form = 1, 2, 3, ... , Io/Im = 0.045, 0.016, 0.0083, etc. The successive
maxima decrease rapidly in intensity.
Example 4. Width of the central diffraction maximum. Derive the half-width
fl()of the central maximum in a single-slit Fraunhofer diffraction (see Fig. 44-9b).
The half-width is the angle between the two points in the pattern where the intensity
is one-half that at the center of the pattern.
Point x in Fig. 44-9b is so chosen that Io = !Im, or, from Eq. 44-8b,
! = (sin ax) 2 •
2 O'.x
This equation cannot be solved analytically for ax, It can be solved graphically,
as accurately as one wishes, by plotting the quantity (sin ax/ax) 2 as ordinate versus
ax as abscissa and noting the value of ax at which the curve intersects the line "one-
half" on the ordinate scale (see Problem 5). However, if only an approximate answer
is desired, it is often quicker to use trial-and-error methods.
We know that a equals 11' at the first minimum; we guess that ax is perhaps 11'/2
( = 90° = 1.57 radians). Trying this in Eq. 44-8b yields
!y_ = [sin (1r/2)] 2 = 0 4
Im 1r/2 ..
This intensity ratio is less than 0.5, so that ax must be less than 90°. After a few more
trials we find easily enough that
ax = 1.40 radians = 80°
O'.x = 11"a
~
· ()
Sln x = 1.40 ,
. 8x = l. 40t.. = l. 40 = 0 .0892 .
sin
11"a 511'
The half-width tl8 of the central maximum (see Fig. 44-9b) is given by
tl8 = 2Bx = 2 sin- 1 0.0892 = 2 X 5.1 ° = 10.2°,
which is in agreement with the figure.
1112 DIFFRACTION Chap. 44
44-5 Diffraction at a Circular Aperture
Diffraction will occur when a wavefront is partially blocked off by an
opaque object such as a metal disk or an opaque screen containing an aper-
ture. Here we consider diffraction at a circular aperture of diameter d, the
aperture constituting the boundary of a circular lens.
Our previous treatment of lenses was based on geometrical optics, diffrac-
tion being specifically assumed not to occur. A rigorous analysis would be
based from the beginning on wave optics, since geometrical optics is always
an approximation, although often a good one. Diffraction phenomena would
emerge in a natural way from such a wave-optical analysis.
Figure 44-10 shows the image of a distant point source of light (a star)
formed on a photographic film placed in the focal plane of a converging lens.
It is not a point, as the (approximate) geometrical optics treatment suggests,
but a circular disk surrounded by several progressively fainter secondary
rings. Comparison with Fig. 42-3c leaves little doubt that we are dealing
with a diffraction phenomenon in which, however, the aperture is a circle
rather than a long narrow slit. The ratio d/'A, where dis the diameter of the
lens (or of a circular aperture placed in front of the lens), determines the scale
of the diffraction pattern, just as the ratio a/>.. does for a slit.
Analysis shows that the first minimum for the diffraction pattern of a
circular aperture of diameter d, assuming Fraunhofer conditions, is given by
A
sin(} = 1.22 -· (44-10)
d
This is to be compared with Eq. 44-1, or
A
sin(} = -,
a
which locates the first minimum for a long narrow slit of width a. The factor
1.22 emerges from the mathematical analysis when we integrate over the
elementary radiators into which the circular aperture may be divided.
energy is concentrated into a smaller diffraction disk but because the larger
lens collects more light. Thus fainter objects, for example, more distant
stars, can be seen.
To reduce diffraction effects in microscopes we often use ultraviolet light,
which, because of its shorter wavelength, permits finer detail to be examined
than would be possible for the same microscope operated with visible light.
We shall see in Chapter 48 that beams of electrons behave like waves under
some circumstances. In the el,ectron microscope such beams may have an
effective wavelength of 0.04 A, of the order of 105 times shorter than visible
light (A,...._, 5000 A). This permits the detailed examination of tiny objects
like viruses. If a virus were examined with an optical microscope, its struc-
ture would be hopelessly concealed by diffraction.
1rd
where {3 = - sin O (44-13)
X
in which d is the distance between the center-lines of the slits.
The intensity for the diffracted wave from either slit is given by Eqs. 44-8b
and c, or, with a small change in nomenclature,
le,dif = lm,dif ( ~
sin a)
2
' (44-14)
,ra
where a= -sin 0. (44-15)
X
-
"in
c
Q)-
c
Q)
(a)
>
~
Q)
15
I
10
I
5
rr 0
I
5 10
I
15
8, degrees
/
/
Z'
' \
-
'in
c
Q)
c a= 5X (b)
Q)
I >
:;::::; \
ro
Q)
/
c:::
"
.......
15 10 5 0 5 10 15
8, degrees
I
/
'
-
\
I >, \
I
-
"in
c
Q)
I .S \
I \ (c)
I \ a= lOX
I \
15 10 5 0 5 10 15
8, degrees
Fig. 44-12 Interference fringes for a double slit with slit separation d = 50X. Three
different slit widths, described by a/X = 1, 5, and 10 are shown.
Sec. 44-6 DOUBLE SLIT 1117
r r r r
~ .?;-
'iii
c
2c
(!)
>
:;:;
ro
<ii
0:::
(a)
I I I \ I I I
15 10 5 0 5 10 15
8, degrees
(b)
8, degrees
~
VI
c
.....
(I)
.!::
(!)
.2:
.....
ro
<ii
0:::
(c)
15 10 5 0 5 10 15
8, degrees
Fig. 44-13 (a) The "interfere nce factor" and (b) the "diffracti on factor"
in Eq. 44-16
and (c) their product; compare Fig. 44-12b.
Fraunho fer diffracti on are focused at the same spot (see Fig. 44-5). Because
the two diffract ed waves are coheren t, they will interfer e.
The effect of interfer ence is to redistri bute the availabl e energy over the
screen, produci ng a set of fringes. In Section 43-1, where we assume d a << >..,
the availabl e energy was virtuall y the same at all points on the screen so
that the interfer ence fringes had virtuall y the same intensit ies (see Fig. 43-9).
If we relax the assump tion a« 'J\, the availabl e energy is not uniform over
the screen but is given by the diffract ion pattern of a slit of width a. In this
1118 DIFFRACTION Chap. 44
ti(a)
Fig. 44-14 (a) Interference fringes for a double-slit system
in which the slit width is not negligible in comparison to the
wavelength. The fringes are modulated in intensity by the
diffraction pattern of a single slit. (b) If one of the slits is
covered up, the interference fringes disappear and we see
the single slit diffraction pattern. (Courtesy G. H. Carra-
gan, Rensselaer Polytechnic Institute.)
(b)
case the interference fringes will have intensities that are determined by the
intensity of the diffraction pattern at the location of a particular fringe.
Equation 44-16 is the mathematical expression of this argument.
Figure 44-12 is a plot of Eq. 44-16 for d = 50t- and for three values of
a/t-. It shows clearly that for narrow slits (a = t-) the fringes are nearly
uniform in intensity. As the slits are widened, the intensities of the fringes
are markedly modulated by the "diffraction factor" in Eq. 44-16, that is,
by the factor (sin a/a) 2 .
Equation 44-16 shows that the fringe envelopes of Fig. 44-12 are precisely
the single-slit diffraction patterns of Fig. 44-9. This is especially clear in
Fig. 44-13, which shows, for the curve of Fig. 44-12b, (a) the "interference
factor" in Eq. 44-16 (that is, the factor cos2 /3), (b) the "diffraction factor"
(sin a/a.)2, and (c) their product.
If we put a = 0 in Eq. 44-16, then (see Eq. 44-15) a = 0 and sin a/a"'
a-->0
a/a = 1. Thus this equation reduces, as it must, to the intensity equation
for a pair of vanishingly narrow slits (Eq. 44-12). If we put d = 0 in Eq.
44-16, the two slits coalesce into a single slit of width a, as Fig. 44-15 shows;
d = 0 implies {3 = 0 (see Eq. 44-13) and cos2 {3 = 1. Thus Eq. 44-16 re-
duces, as it must, to the diffraction equation for a single slit (Eq. 44-14).
Figure 44-14 shows some actual double-slit interference photographs.
The uniformly spaced interference fringes and their intensity modulation
by the diffraction pattern of a single slit is clear. If one slit is covered up,
as in Fig. 44-14b, the interference fringes disappear and we see the diffrac-
tion pattern of a single slit .
._ Example 6. Starting from the curve of Fig. 44-12b, what is the effect of (a)
increasing the slit width, (b) increasing the slit separation, and (c) increasing the
wavelength?
(a) If we increase the slit width a, the envelope of the fringe pattern changes so
that its central peak is sharper (compare Fig. 44-12c). The fringe spacing, which
depends on d/t., does not change.
(b) If we increased, the fringes become closer together, the envelope of the pattern
remaining unchanged.
(c) If we increase A, the envelope becomes broader and the fringes move further
apart. Increasing )\ is equivalent to decreasing both of the ratios a/t- and d/t-. The
general relationship of the envelope to the fringes, which depends only on d/ a, does
not change with wavelength.
Sec. 44-6 DOUBLE SLIT 1119
where D is the distance of the screen from the slits. Substituting yields
The distance to the first minimum of the envelope is determined by the diffraction
factor (sin a/a) 2 in Eq. 44-16. The first minimum in this factor occurs for a = 1r.
From Eq. 44-15, •
sin 8 = aX = ~ = 480 X 10-9 meter = 0 _024 _
1ra a 0.02 X 10- 3 meter
This is so small that we can assume that 8 "' sin 8 "' tan 8, or
There are about ten fringes in the central peak of the fringe envelope.
Example 8. What requirements must be met for the central maximum of the
envelope of the double-slit Fraunhofer pattern to contain exactly eleven fringes?
The required condition will be met if the sixth minimum of the interference factor
(cos2 fJ) in Eq. 44-16 coincides with the first minimum of the diffraction factor
(sin a/a) 2•
The sixth minimum of the interference factor occurs when
in Eq. 44-12.
The first minimum in the diffraction term occurs for
a = 7r.
fJ d 11
~ = - = -·
a a 2
This condition depends only on the slit geometry and not on the wavelength. For
long waves the pattern will be broader than for short waves, but there will always be
eleven fringes in the central peak of the envelope. <11111
QUESTIONS
1. Why is the diffraction of sound waves more evident in daily experience than that of
light waves?
2. Why do radio waves diffract around buildings, although light waves do not?
3. A loud-speaker horn has a rectangular aperture 4 ft high and 1 ft wide. Will the
pattern of sound intensity be broader in the horizontal plane or in the vertical?
4. A radar antenna is designed to give accurate measurements of the height of an air-
craft but only reasonably good measurements of its direction in a horizontal plane. Must
the height-to-width ratio of the radar reflector be less than, equal to, or greater than unity?
5. A person holds a single narrow vertical slit in front of the pupil of his eye and looks
at a distant light source in t,he form of a long heated filament. Is the diffraction pattern
that he sees a Fresnel or a Fraunhofer pattern?
6. In a single-slit Fraunhofer diffraction, what is the effect of increasing (a) the wave-
length and (b) the slit width?
7. Sunlight falls on a single slit of width 104 A. Describe qualitatively what the result-
ing diffraction pattern looks like.
8. In Fig. 44-5 rays r1 and r3 are in phase; so are r2 and r4. Why isn't there a maximum
intensity at P2 rather than a minimum?
9. Describe what happens to a Fraunhofer single-slit diffraction pattern if the whole
apparatus is immersed in water.
10. Distinguish clearly between(), a, and</> in Eq. 44-8c.
11. Do diffraction effects occur for virtual images as well as for real images? Explain.
12. Do diffraction effects occur for images formed by (a) plane mirrors and (b) spherical
mirrors? Explain.
13. If we were to redo our analysis of the properties of lenses in Section 42-5 by the
methods of geometrical optics but without restricting our considerations to paraxial rays
and to "thin" lenses, would diffraction phenomena, such as that of Fig. 44-10, emerge
from the analysis? Discuss.
14. Distinguish carefully between interference and diffraction in Young's double-slit
experiment.
15. In what way are interference and diffraction similar? In what way are they dif-
ferent?
16. In double-slit interference patterns such as that of Fig. 44-14a we said that the
interference fringes were modulated in intensity by the diffraction pattern of a single slit.
Could we reverse this statement and say that the diffraction pattern of a single slit is
intensity-modulated by the interference fringes? Discuss.
Chap. 44 PROBLEMS 1121
PROBLEMS
1. In a single-slit diffraction pattern the distance between the first minimum on the
right and the first minimum on the left is 5.2 mm. The screen on which the pattern is
displayed is 80 cm from the slit and the wavelength is 5460 A. Calculate the slit width.
"2. A plane wave (A = 5900 A) falls on a slit with a = 0.40 mm. A converging lens
(f = +70 cm) is placed behind the slit and focuses the light on a screen. What is the linear
distance on the screen from the center of the pattern to (a) the first minimum and (b) the
second minimum?
3. A single slit is illuminated by light whose wavelengths are Aa and Ab, so chosen that the
first diffraction minimum of Aa coincides with the second minimum of Ab, (a) What
relationship exists between the two wavelengths? (b) Do any other minima in the two
patterns coincide?
4. (a) Show that the values of a at which intensity maxima for single-slit diffraction
occur can be found exactly by differentiating Eq. 44-8b with respect to a and equating to
zero, obtaining the condition
tan a = a.
(b) Find the values of a satisfying this relation by plotting graphically the curve y = tan a
and the straight line y = a and finding their intersections. (c) Find the (nonintegral)
values of m corresponding to successive maxima in the single-slit pattern. Note that the
secondary maxima do not lie exactly halfway between minima.
5. In Example 4 solve the transcendental equation
! = (sin ax) 2
2 ax
graphically for ax, to an accuracy of three significant figures.
6. (a) In Fig. 44-7d, why is E9, which represents the first maximum beyond the central
maximum, not vertical? (b) Calculate the angle it makes with the vertical, assuming the
~it to be divided into infinitesimal strips of width dx .
.,Z, What is the half-width of a diffracted beam for a slit whose width is (a) 1, (b) 5, and
(c) io wavelengths?
8. (a) A circular diaphragm 0.60 meter in diameter oscillates at a frequency of 25,000
cycles/sec in an underwater source of sound for submarine detection. Far from the source
the sound intensity is distributed as a Fraunhofer diffraction pattern for a circular hole
whose diameter equals that of the diaphragm. Take the speed of sound in water to be
1450 meters/sec and find the angle between the normal to the diaphragm and the direc-
tion of the first minimum. (b) Repeat for a source having an (audible) frequency of 1000
cycles/sec.
9. The two headlights of an approaching automobile are 4 ft apart. At what maximum
distance will the eye resolve them? Assume a pupil diameter of 5.0 mm and A = 5500 A.
Assume also that diffraction effects and not retinal structure limit the resolution.
10. The wall of a large room is covered with acoustic tile in which small holes are drilled
5.0 mm from center to center. How far can a person be from such a tile and still distin-
guish the individual holes, assuming ideal conditions? Assume the diameter of the oupil
to be 4.0 mm and A to be 5500 A.
11. (a) How small is the angular separation of two stars if their images are barely
resolved by the Thaw refracting telescope at the Allegheny Observatory in Pittsburgh?
The lens diameter is 30 in. and its focal length is 46 ft. Assume A = 5000 A. (b) Find the
distance between these barely resolved stars if each of them is 10 light years distant from
the earth. (c) For the image of a single star in this telescope, find the diameter of the
first dark ring in the diffraction pattern, as measured on a photographic plate placed at the
1122 DIFFRACTION Chap. 44
focal plane. Assume that the star image structure is associated entirely with diffraction
at the lens aperture and not with (small) lens ·'errors."
12. Find the separation of two points on the moon's surface that can just be resolved
by the 200-in. telescope at Mount Palomar, assuming that this distance is determined by
diffraction effects. The distance from the earth to the moon is 240,000 miles.
13. Construct qualitative vector diagrams like those of Fig. 44-7 for the double-slit
interference pattern. For simplicity, consider d = 2a (see Fig. 44-15). Can you interpret
the main features of the intensity pattern this way?
Fig. 44-15
14. Suppose that, as in Example 8, the envelope of the central peak contains eleven
fringes. How many fringes lie between the first and second minima of the envelope?
15. Ford = 2a in Fig. 44-15, how many interference fringes lie in the central diffraction
envelope?
16. If we put d = a in Fig. 44-15, the two slits coalesce into a single slit of width 2a.
Show that Eq. 44-16 reduces to the diffraction pattern for such ~slit.
17. (a) Design a double-slit system in which the fourth fringe, not counting the central
maximum, is missing. (b) What other fringes, if any, are also missing?
Gratings and Spectra
CHAPTER 45
45-1 Introduction
In connection with Young's experiment (Sections 43-1 and 43-3) we
discussed the interference of two coherent waves formed by diffraction at
two elementary radiators (pinholes or slits). In our first treatment we as-
sumed that the slit width was much less than the wavelength, so that light
diffracted from each slit illuminated the observation screen essentially uni-
formly. Later, in Section 44-6, we took the slit width into account and
showed that the intensity pattern of the interference fringes is modulated
by a "diffraction factor" (sin a/a) 2 (see Eq. 44-16).
Here we extend our treatment to cases in which the number N of radia-
tors or diffracting centers is larger-and usually much larger-than two.
We consider two situations:
1. An array of N parallel equidistant slits, called a diffraction grating.
2. A three-dimensional array of periodically arranged radiators-the
atoms in a crystalline solid such as NaCL In this case the average spacing
between the elementary radiators is so small that interference effects must
be sought at wavelengths much smaller than those of visible light. We
speak of X-ray diffraction.
In each case we distinguish carefully between the diffracting properties
of a single radiator (slit or atom) and the interference of the waves dif-
fracted, coherently, from the assembly of radiators.
Fig. 45-1 An idealized diffraction grating containing five slits. The slit width a is shown
for convenience to be considerably smaller than A, although this condition is not realized
in practice. The figure is distorted in that f is much greater than d in practice.
tion grating. As for a double slit, the intensity pattern that results when
monochromatic light of wavelength>.. falls on a grating consists of a series of
interference fringes. The angular separations of these fringes are determined
by the ratio A/ d, where d is the spacing between the centers of adjacent slits.
The relative intensities of these fringes are determined by the diffraction pat-
tern of a single grating slit, which depends on the ratio A/ a, where a is the
slit width.
Figure 45-2, which compares the intensity patterns for N = 2 and N = 5,
shows clearly that the "interference" fringes are modulated in intensity by
a "diffraction" envelope, as in Fig. 44-14. Figure 45-3 presents a theoretical
calculation of the intensity patterns for a few fringes near the centers of the
patterns of Fig. 45-2. These two figures show that increasing N (a) does
not change the spacing between the (principal) interference fringe maxima,
.i::'
"vi
c 100 -
a.> 80 N=2
c 60
> 40
a.>
-
20
.;:::;
"' 0
a, I I I', I I l.1 I I I I I I I Ll...L_LJ...L.[...Li...J',W
e, degrees
(b)
'.
0
N=5
d h _h n
I
d
I
10
I
±:I +d
40
,J
Fig. 45-3 Calculated intensity patterns for (a) a two-slit and (b) a five-slit grating for
the same value of d and JI.. This figure shows the sharpening of the principal maxima
and the appearance of faint serondary maxima for N > 2. The letters on the five-slit
pattern refer to Fig. 45-5. The figure assumes slits with a « 'A so that the principal
maxima are of uniform intensity.
This equation shows specifically that if we increase N for a given >.. and d,
~
E2m
(a)
(d)
Fig. 45-4 Drawings (a) and (b) show conditions at the central principal maximum for a
two-slit and a nine-slit grating, respectively. Drawings (c) and (d) show conditions at the
minimum of zero intensity that lies on either side of this central principal maximum. In
going from (a) to (c) the phase shift between waves from adjacent slits changes by 180°
(!'>.</> = 21r/2); in going from (b) to (d) it changes by 40° (!'>.</> = 21r/9).
Sec. 45-2 MULTIPLE SLITS 1127
E 8(= Em) \
(a) \
o·
~
(360°)
(i)
(240°)
d
I
I
''
,[µ&
(e)
I / ~q:_ (280°)
E9
(j)
I E9 (230°)
I I
--a2_0.'._
I
(b)
(330°) (f) I
(270°)
"'" I
/~
l E8 (k)
I (216°)
I
I I E9 \
I \
\
(c) --~ (g) \
1.90· (260°)
'-
(300°)
\
\
-~ E9 '-
(l)
(200°)
/
/ --~ (190°)
(d)
(n)
(288°) (180°)
/ ~180°
Eo
Fig. 45-5 The figures taken in sequence from (a) to (n) and then from (n) to (a) show con-
ditions as the intensity pattern of a five-slit grating is traversed from the central principal
maximum to an adjacent principal maximum. Phase differences between waves from
adjacent slits are shown directly or, when going from (n) to (a), in parentheses. Principal
maxima occur at (a), secondary maxima at, or near, (h) and (n), and points of zero intensity
at (d) and (k). Compare Fig. 45-3b.
~e0 will decrease, which means that the central principal maximum becomes
sharper.
We state without proof,* and for later use, that for principal maxima other than
the central one (that is, for m ,t: 0) the angular distance between the position em
of the principal maximum of order m and the minimum that lies on either side is
given by
")..
~em=----
Nd cos &m
(any principal maximum). (45-3)
For the central principal maximum we have m = 0, em = 0, and ~&m = ~&0, so that
Eq. 45-3 reduces, as it must, to Eq. 45-2.
The origin of the secondary maxima that appear for N > 2 can also be
understood using the phasor method. Figure 45-Sa shows conditi9ns at
the central principal maximum for a five-slit grating. The vectors are in
* See Problem 15.
1128 GRATINGS AND SPECTRA Chap. 45
phase. As we depart from the central maximum, e in Fig. 45-1 increases
from zero and the phase difference between adjacent vectors increases from
21r .
zero to !::.¢ = - d sm e. Successive figures show how the resultant wave
A
amplitude Ee varies with !::.¢. The student should verify by graphical con-
struction that a given figure represents conditiens for both !::.¢ and 21r - !::.¢.
Thus we start at !::.¢ = 0, proceed to !::.¢ = 180°, and then trace backward
through the sequence, following the phase differences shown in parentheses,
until we reach !::.¢ = 360°. This sequence corresponds to traversing the
intensity pattern from the central principal maximum to an adjacent one.
Figure 45-5, which should be compared with Fig. 45-3b, shows that for
N = 5 there are three secondary maxima, corresponding to!::.¢ = 110°, 180°,
and 250°. The student should make a similar analysis for N = 3 and should
show that only one secondary maximum occurs. In actual gratings, which
commonly contain 10,000 to 50,000 "slits," the secondary maxima lie so
close to the principal maxima or are so reduced in intensity that they cannot
be distinguished from them experimentally.
namely
d sin 8 = m>.. m=0,1,2 ... ,
where d is the distance between the rulings and the integer m is called the order of
the particular principal maximum. Essentially all gratings used in the visible
spectrum, whether of the transmission type, as in Fig. 45-6, or the reflection type, are
phase gratings.
Figure 45-7 shows a simple grating spectroscope, used for viewing the
spectrum of a light source, assumed to emit a number of discrete wavelengths,
or spectrum lines. The light from source Sis focused by lens L 1 on a slit S 1
placed in the focal plane of lens L 2 . The parallel light emerging from col-
limator C falls on grating G. Parallel rays associated with a particular in-
terference maximum occurring at angle 8 fall on lens L 3 , being brought to a
focus in plane F-F'. The image formed in this plane is examined, using a
magnifying lens arrangement E, called an eyepiece. A symmetrical inter-
Fig. 45-7 A simple type of grating spectroscope used to analyze the wavelengths of the
light emitted by source S.
1130 GRATINGS AND SPECTRA Chap. 45
ference pattern is formed on the other side of the central position, as shown
by the dotted lines. The entire spectrum can be viewed by rotating tele-
scope T through various angles. Instruments used for scientific research or
in industry are more complex than the simple arrangement of Fig. 45-7.
They invariably employ photographic or photoelectric recording and are
called spectrographs. Figure 47-12 shows a small portion of the spectrum of
iron, produced by examining the light produced in an arc struck between iron
electrodes, using a research type spectrograph with photographic recording.
Each line in the figure represents a different wavelength that is emitted from
the source.
Grating instruments can be used to make absolute measurements of wave-
length, since the grating spacing din Eq. 45-1 can be measured accurately
with a traveling microscope. Several spectra are normally produced in such
instruments, corresponding tom = ± 1, ±2, etc., in Eq. 45-1 (see Fig. 45-8).
This may cause some confusion if the spectra overlap. Further, this multi-
plicity of spectra reduces the recorded intensity of any given spectrum line
because the available energy is divided among a number of spectra.
This disadvantage of the grating instrument can be overcome by shaping the
profile of the grating grooves so that a large fraction of the light is thrown into a
particular order on a particular side (for a given wavelength). This technique, called
blazing, so alters the diffracting properties of the individual grooves (by controlling
their profiles) that the light of wavelength X diffracted by a single groove has a sharp
peak of maximum intensity at a selected angle() ( ~ 0).
Light can also be analyzed into its component wavelengths if the grating
in Fig. 45-7 is replaced by a prism. In a prism spectrograph each wavelength
Fig. 45-8 Example 1. The spectrum of white light as viewed in a grating instrument
like that of Fig. 45-7. The different orders, identified by the order number m, are shown
separated vertically for clarity. As actually viewed, they would not be so displaced.
The central line in each order corresponds to }.. = 5500 A.
Sec. 45-3 DIFFRACTION GRATINGS 1131
In the same way the angle for the 7000-A line is found to be 12.8°, and the entire
pattern of Fig. 45-8 can be calculated. Note that the first-order spectrum (m = 1) is
isolated but that the second-, third-, and fourth-order spectra overlap.
Example 2. A diffraction grating has 104 rulings uniformly spaced over 1 in.
It is illuminated at normal incidence by yellow light from a sodium vapor lamp.
This light contains two closely spaced lines (the well-known sodium doublet) of
wavelengths 5890.0 and 5895.9 A. (a) At what angle will the first-order maximum
occur for the first of these wavelengths?
The grating spacing d is 10-4 in., or 25,400 A. The first-order maximum corre-
sponds tom = 1 in Eq. 45-1. We thus have
. -1 ':!!'.~ - . -1 (1)(5890 A) = . -1 0 232 = 13 30
8 -- sm d - sm 25 400 A sm · · ·
'
(b) What is the angular separation between the first-order maxima for these lines?
The straightforward way to find this separation is to repeat this calculation for
A = 5895.9 A and to subtract the two angles. A difficulty, which can best be ap-
preciated by carrying out the calculation, is that we must carry a large number of
significant figures to obtain a meaningful value for the difference between the angles.
To calculate the difference in angular positions directly, let us write down Eq. 45-1,
solved for sin 8, and differentiate it, treating 8 and A as variables:
. 8 m"\
sm =a
1132 GRATINGS AND SPECTRA Chap. 45
If the wavelengths are close enough together, as in this case, d").. can be replaced by
LlA, the actual wavelength difference; d{) then becomes .6.(), the quantity we seek. This
Here ").. is the mean wavelength of two spectrum lines that can barely be
recognized as separate and .6.").. is the wavelength difference between them.
The smaller .6.").. is, the closer the lines can be and still be resolved; hence
the greater the resolving power R of the grating. It is to achieve a high
resolving power that gratings with many rulings are constructed.
The resolving power of a grating is usually determined by the same con-
sideration (that is, the Rayleigh criterion) that we used in Section 44-5 to
determine the resolving power of a lens. If two principal maxima are to be
barely resolved, they must, according to this criterion, have an angular
separation .6.() such that the maximum of one line coincides with the first
minimum of the other; see Fig. 44-11. If we apply this criterion, we can
show that
R = Nm, (45-6)
where N is the total number of rulings in the grating and m is the order.
As expected, the resolving power is zero for the central principal maximum
(m = 0), all wavelengths being undeflected in this order.
Let us derive Eq. 45-6. The angular separation between two principal maxima
whose wavelengths differ by .6."' is found from Eq. 45-4, which we recast as
N = : = l~OO = 330.
This is a modest requirement.
The resolving power of a grating must not be confused with its dispersion.
Table 45-1 shows the characteristics of three gratings, each illuminated with
light of;\ = 5890 A, the diffracted light being viewed in the first order (m =
1 in Eq. 45-1).
Table 45-1
d, D
Grating N e R
A 10-3 degrees/ A
The student should verify that the values of D and R given in the table
can be calculated from Eqs. 45-4 and 45-6, respectively.
For the conditions of use noted in Table 45-1, gratings A and B have the
same dispersion and A and C have the same resolving power. Figure 45-9
shows the intensity patterns that would be produced by these gratings for
two incident waves of wavelengths ;\ 1 and >. 2 , in the vicinity of;\ = 5890 A.
Grating B, which has high resolving power, has narrow intensity maxima
and is inherently capable of distinguishing lines that are much closer together
in wavelength than those of Fig. 45-9. Grating C, which has high dispersion,
produces twice the angular separation between rays >. 1 and >. 2 that grating B
does.
1134 GRATINGS AND SPECTRA Chap. 45
(a)
~
~
~L
---~l
!\ Grating
C X2
I\ (c)
0 0, degrees
------- fl(} - - - - -
25.5°
Fig. 45-9 The intensity patterns for light of wavelengths X1 and X2 near 5890 A, incident
on the gratings of Table 45-1. Grating B has the highest resolving power and grating C
the highest dispersion.
lili1- Example 4. The grating of Example 1 has 8000 lines illuminated by light from
a mercury vapor discharge. (a) What is the expected dispersion, in the third order,
in the vicinity of the intense green line (X = 5460 A)? Noting that d = 31,700 A,
we have, from Eq. 45-1,
. -1 mX - . -1 (3)(5460 A) - . -1 0 517 - 31 lo
0 -- sm d - sm 31 700 A - sm · - · ·
'
From Eq. 45-4 we have
D = d c:s O = (31,700 A:(cos 31.1 o) = 1.1 X 10-4 radian/ A = 6.3 X 10-3 deg/ A.
(b) What is the expected resolving power of this grating in the fifth order? Equation
45-6 gives
R = Nm = (8000)(5) = 40,000.
Thus near X = 5460 A a wavelength difference ~X given by Eq. 45-5, or
~X = ~ = 5460 A = 0 14 A
R 40,000 . '
can be distinguished.
Fig. 45-10 X-rays are generated when electrons from heated filament F, accelerated by
potential difference V, are brought to rest on striking metallic target T. Wis a "window"
-transparent to X-rays-in the evacuated metal container C.
for example, Eq. 45-1 shows that the first-order maximum occurs at
mA (l)(lA)
() = sin-1 - = sin- 1 4 = sin- 1 0.33 X 10-4 = 0.002°.
d 3Xl0 A
This is too close to the central maximum to be practical. A grating with
d"' A is desirable, but, since X-ray wavelengths are about equal to atomic
diameters, such gratings cannot be constructed mechanically.
In 1912 it occurred to the German physicist Max von Laue that a crystal-
line solid, consisting as it does of a regular array of atoms, might form a
natural three-dimensional "diffraction grating" for X-rays. Figure 45-11
shows that if a collimated beam of X-rays, continuously distributed in wave-
length, is allowed to fall on a crystal, such as sodium chloride, intense beams
corresponding to constructive interference from the many diffracting centers
Diffracted beams
X-ray beam
(continuous in
wavelength)
Fig. 45-12 Laue X-ray diffraction pattern from sodium chloride. A crystal of ordinary
table salt was used in making this plate. (Courtesy of W. Arrington and J. L. Katz,
X-ray Laboratory, Rensselaer Polytechnic Institute.)
by solids. The model represents the unit cell for sodium chloride. This is
the smallest unit from which the crystal may be built up by repetition in
three dimensions. The student should verify that no smaller assembly of
atoms possesses this property. For sodium chloride the length of the cube
edge of the unit cell is 5.62737 A.
Each unit cell in sodium chloride has four sodium ions and four chlorine
ions associated with it. In Fig. 45-13 the sodium ion in the center belongs
entirely to the cell shown. Each of the other twelve sodium ions shown is
shared with three adjacent unit cells so that each contributes one-fourth of
an ion to the cell under consideration. The total number of sodium ions is
then 1 + -!(12) = 4. By similar reasoning the student can show that al-
though there are fourteen chlorine ions in Fig. 45-13 only four are associated
with the unit cell shown.
The unit cell is the fundamental repetitive diffracting unit in the crystal,
corresponding to the slit (and its adjacent opaque strip) in the optical dif-
Fig. 45-13 A model showing how Na+ and c1- ions are stacked to form a unit cell of
NaCl. The small spheres represent sodium ions, the large ones chlorine. The edge a0
of the (cubical) unit cell is 5.62737 A.
1138 GRATINGS AND SPECTRA Chap. 45
• •
• •
• • • • •
• • • • •
• • • • •
(a) (b)
Fig. 45-14 (a) A section through a crystal of sodium chloride, showing the sodium and
chlorine ions. (b) The corresponding unit cells in this section, each cell being represented
by a small black square.
* For some directions in which a beam might be expected to emerge, from interference
considerations, no beam will be found because the diffracting characteristics of the unit
cell are such that no energy is diffracted in that direction. Similarly, in optical gratings
some lines, permitted by interference considerations, may not appear if their predicted
positions coincide with a null in the single-slit diffraction pattern (see Fig. 44-12).
(a)
(b)
Figure 45-16c shows an incident wave striking the family of planes, a single
member of which was considered in Fig. 45-16b. For a single plane, mirror-
like "reflection" occurs for any value of e, as we have seen. To have a con-
structive interference in the beam diffracted from the entire family of planes
Incident wave
x
\
Fig. 45-16 (a) A section through the
NaCl unit cell lattice of Fig. 45-14b.
The dashed sloping lines represent an
v\
I
arbitrary family of planes, with inter- h I
Incident wave
\
~
\
•
\
• Diffracted
wave
(c)
1142 GRATINGS AND SPECTRA Chap. 45
in the direction 0, the rays from the separate planes must reinforce each other.
This means that the path difference for rays from adjacent planes (abc in Fig.
45-16c) must be an integral number of wavelengths or
2d sin() = m">.. m = 1, 2, 3, .... (45-8)
This relation is called Bragg's law after W. L. Bragg who first derived it.
The quantity din this equation (the interplanar spacing) is the perpendicular
distance between the planes. For the planes of Fig. 45-16a analysis shows
that d is related to the unit cell dimension a0 by
(45-9)
Diffracted beams are possible at() = 12.6° (m = 1), () = 25.9° (m = 2), () = 40.9°
(m = 3), and () = 60.7° (m = 4). Higher-order beams cannot exist because they
require sin() to exceed unity. Actually, the odd-order beams (m = 1, 3) prove to
have zero intensity because the unit cell in cubic crystals such as NaCl has diffracting
properties such that the intensity of the light scattered in these orders is zero (see
Problem 26). <11111
* In practice, calcite (CaC0 3) proves to be more useful as a standard crystal for a num-
ber of technical reasons.
Chap. 45 QUESTIONS 1143
If p is the measured density of NaCl, we have, for the unit cell of Fig.
45-13, * recalling that each unit cell contains four NaCl "molecules,"
m 4mNaCl
p =- =
V a0 3
Here mNaCI, the mass of a NaCl molecule, is given by
M
-,
No
where M is the molecular weight of NaCl and N 0 is Avogadro's number.
Combining these two equations and solving for a0 yields
-(4M)li
ao - --
NoP
which permits us to calculate a 0 • Once a 0 is known, the wavelengths of
monochromatic X-ray beams can be found, using Bragg's law (Eq. 45-8).
QUESTIONS
1. Discuss this statement: "A diffraction grating can just as well be called an inter-
ference grating."
2. For the simple spectroscope of Fig. 45-7, show (a) that 8 increases with Xfor a grating
and (b) that 8 decreases with X for a prism.
3. You are given a photograph of a spectrum on which the angular positions and the
wavelengths of the spectrum lines are marked. (a) How can you tell whether the spectrum
was taken with a prism or a grating instrument? (b) What information could you gather
about either the prism or the grating from studying such a spectrum?
4. Assume that the limits of the visible spectrum are 4300 and 6800 A. Is it possible
to design a grating, assuming that the incident light falls normally on it, such that the
first-order spectrum barely overlaps the second-order spectrum?
5. (a) Why does a diffraction grating have closely spaced rulings? (b) Why does it
have a large number of rulings?
6. The relation R = Nm suggests that the resolving power of a given grating can be
made as large as desired by choosing an arbitrarily high order of diffraction. Discuss.
7. Show that at a given wavelength and a given angle of diffraction the resolving power
of a grating depends only on its width W ( = Nd).
8. According to Eq. 45-3 the principal maxima become wider (that is, b..8m increases) the
higher the order m (that is, the larger 8m becomes). According to Eq. 45-6 the resolving
power becomes greater the higher the order m. Explain this apparent paradox.
9. Is the pattern of Fig. 45-12 more properly described as a diffraction pattern or as an
interference pattern?
10. For a given family of planes in a crystal, can the wavelength of incident X-rays be
(a) too large or (b) too small to form a diffracted beam?
11. If a parallel beam of X-rays of wavelength)'.. is allowed to fall on a randomly oriented
crystal of any material, generally no intense diffracted beams will occur. Such beams
* This relation cannot be written down unless it is known that the structure of NaCl is
cubic. 'fhis can be determined, however, by inspection of the symmetry of the spots in
Fig. 45-12; the wavelength of the X-rays need not be known.
1144 GRATINGS AND SPECTRA Chap. 45
appear if (a) the X-ray beam consists of a continuous distribution of wavelengths rather
than a single wavelength or (b) the specimen is not a single crystal but a finely divided
powder. Explain.
12. Why cannot a simple cube of edge a 0 /2 in Fig. 45-13 be used as a unit cell for sodium
chloride?
13. How would you measure (a) the dispersion D and (b) the resolving power R for
either a prism or a grating spectrograph.
PROBLEMS
f
/
/ Given a grating with 4000 lines/cm, how many orders of entire visible spectrum
(4000-7000 A) can be produced?
2. Derive this expression for the intensity pattern for a three-slit "grating":
X
6.£) =--·
-3.2d
(b) Compare this with the expression derived for the two-slit pattern in Problem 10, Chap-
ter 43. (c) Do these results support the conclusion that for a fixed slit spacing the inter-
ference maxima become sharper as the number of slits is increased?
4. Using the result of Problem 2, show that a three-slit "grating" has only one secondary
maximum. Find its location and its relative intensity.
5. A grating designed for use in the infrared region of the electromagnetic spectrum is
"blazed" to concentrate all its intensity in the first order (m = 1) for X = 80,000 A. If
visible light (4000 A < X < 7000 A) were allowed to fall on this grating, what visual ap-
pearance would the diffracted beams present?
6. The central intensity maximum formed by a grating, along with its subsidiary
secondary maxima, can be viewed as the diffraction pattern of a single "slit" whose width
is that of the entire grating. Treating the grating as a single wide slit, assuming that
m = 0, and using the methods of Section 44-4, show that Eq. 45-2 can be derived.
7. A grating has 8000 rulings/in. For what wavelengths in the visible spectrum can
fifth-order diffraction be observed?
8. A diffraction grating has 5000 rulings/in., and a strong diffracted beam is noted at
{) = 30°. (a) What are the possible wavelengths of the incident light? (b) How could you
identify them in an actual case?
9. A diffraction grating 2.0 cm wide has 6000 rulings. At what angles will maximum-in-
tensity beams occur if the incident radiation has a wavelength of 5890 A?
10. Assume that the limits of the visible spectrum are arbitrarily chosen as 4300 and
6800 A. Design a grating that will spread the first-order spectrum through an angular
range of 20 °.
11. A grating has 8000 rulings/in. and is illuminated at normal incidence by white
light. A spectrum is formed on a screen 30 cm from the grating. If a 1.0-cm square hole
is cut in the screen, its inner edge being 5.0 cm from the central maximum, what range of
wavelengths passes through the hole?
Chap. 45 PROBLEMS 1145
)2:" Assume that light is incident on a grating at an angle if; as shown in Fig. 45-17.
show that the condition for a diffraction maximum is
d(sin if; + sin e) = m">,. m = 0, 1, 2, ....
Only the special case if; = 0 has been treated in this chapter (compare Eq. 45-1).
Fig. 45-17
13. Show that in a grating with alternately transparent and opaque strips of equal
width all the even orders (except m = 0) are absent.
14. A transmission grating with d = 1.50 X 10-4 cm is illuminated at various angles
of incidence by light of wavelength 6000 A. Plot as a function of angle of incidence
(0 to 90°) the angular deviation of the first-order diffracted beam from the incident direc-
tion.
15. Derive Eq. 45-3, that is, the expression for f:j.Om, the angular distance between a
principal maximum of order m and either adjacent minimum.
16. A three-slit grating has separation d between adjacent slits. If the middle slit is
covered up, will the half-width of the intensity maxima become broader or narrower?
See Problem 3 and also Problem 10, Chapter 43 .
. 17. A grating has 40,000 rulings spread over 3.0 in. (a) What is its expected dispersion
D for sodium light(">,. = 5890 A) in the first three orders? (b) What is its resolving power
in these orders?
18. In a particular grating the sodium doublet (see Example 2) is viewed in third order
at 80° to the normal and is barely resolved. Find (a) the grating spacing and (b) the total
width of the rulings.
19. A source containing a mixture of hydrogen and deuterium atoms emits a red doublet
at ">,. = 6563 A whose separation is 1.8 A. Find the minimum number of lines needed in a
diffraction grating which can resolve these lines in the first order.
20. Show that the dispersion of a grating can be written as
tan e
D=--·
h
21. A grating has 6000 rulings/cm and is 6.0 cm wide. (a) What is the smallest wave-
length interval that can be resolved in the third order at").. = 5000 A? (b) For this wave-
length and this grating, can the resolution be improved? How?
22. Light containing a mixture of two wavelengths, 5000 A and 6000 A, is incident
normally on a diffraction grating. It is desired (1) that the first and second principal
maxima for each wavelength appear ate ~ 30°, (2) that the dispersion be as high as pos-
sible, and (3) that the third order for 6000 A be a missing order. (a) What is the separation
1146 GRATINGS AND SPECTRA Chap. 45
between adjacent slits? (b) What is the smallest possible individual slit width? (c) Name
all orders for 6000 A that actually appear on the screen with the values chosen in (a) and
(b).
23. Light of wavelength 6000 A is incident normally on a diffraction grating. Two
adjacent principal maxima occur at sin () = 0.2 and sin () = 0.3, respectively. The fourth
order is a missing order. (a) What is the separation between adjacent slits? (b) What is the
smallest possible individual slit width? (c) Name all orders actually appearing on the
screen with the values chosen in (a) and (b).
24. An optical grating with a spacing d = 15,000 A is used to analyze soft X-rays of
wavelength >,. = 5.0 A: The angle of incidence () is 90° - 'Y, where 'Y is a small angle.
The first order maximum is found at an angle() = 90° - 2(3. Find the value of (3.
25. Monochromatic X-rays (A = 1.25A) fall on a crystal of sodium chloride, making
an angle of 45° with the reference line shown in Fig. 45-18. The planes shown are
those of Fig. 45-16a, for which d = 2.52A. Through what angles must the crystal be
turned to give a diffracted beam associated with the planes shown? Assume that the
crystal is turned about an axis that is perpendicular to the plane of the page. Ignore
the possibility (see Problem 26) that some of these beams may be of zero intensity.
Incident
beam
direction
- - - - - _i_
-----t----- T
----,-----
I d
1
-----,------
1
Fig. 45-18
46-1 Polarization
Light, like all electromagnetic radiation, is predicted by electromagnetic
theory to be a transverse wave, the directions of the vibrating electric and
magnetic vectors being at right angles to the direction of propagation instead
of parallel to it as in a longitudinal wave. The transverse waves of Figs. 46-1
and 39-11 have the additional characteristic that they are plane-polarized.
This means that the vibrations of the E vector are parallel to each other for
all points in the wave. At any such point the vibrating E vector and the di-
rection of propagation form a plane, called the plane of vibration; in a plane-
polarized wave all such planes are parallel.
Antenna) t Anten~n-a---~
Fig. 46-2 The vectors E in the transmitted wave are parallel to the axis of the receiving
antenna so that the wave will be detected. If the receiving antenna is rotated through
90° about the direction of propagation, no signal will be detected.
in that the elementary radiators, that is, the atoms and molecules, act inde-
pendently. The light propagated in a given direction consists of independent
wavetrains whose planes of vibration are randomly oriented about the direc-
tion of propagation, as in Fig. 46-3b. Such light, though still transverse,
is unpolarized. The random orientation of the planes of vibration produces
symmetry about the propagation direction, which, on casual study, conceals
the true transverse nature of the waves. To study this transverse nature, a
way must be found to unsort the different planes of vibration.
Fig. 46-3 (a) A plane-polarized transverse wave moving toward the reader, showing only
the electric vector. (b) An unpolarized transverse wave viewed as a random superposition
of many plane-polarized wavetrains. (c) A second, completely equivalent, description of
an unpolarized transverse wave; here the unpolarized wave is viewed as two plane-polar-
ized waves with a random phase difference. The orientation of the x and y axes about the
propagation direction is completely arbitrary.
1152 POLARIZATION Chap. 46
Fig. 46-7 Two square sheets of Polaroid are laid over a book. In (a) the axes of polar-
ization of the two sheets are parallel and light passes through both sheets. In (b) one
sheet has been rotated 90° and no light passes through. The book is opened to an illus-
tration of the Luxembourg Palace in Paris. Malus discovered the phenomenon of polar-
ization by reflection while looking at sunlight reflected off the palace windows through a
calcite crystal.
than those so far described (see Section 46-3). Equation 46-1 describes
precisely the lack of symmetry about the propagation direction that must be
exhibited by plane-polarized transverse waves. Longitudinal waves could
not possibly show such effects. Interestingly enough the human eye, under
certain conditions, can detect polarized light.*
._ Example 1. Two polarizing sheets have their polarizing directions parallel so
that the intensity Im of the transmitted light is a maximum. Through what angle
must either sheet be turned if the intensity is to drop by one-half?
From Eq. 46-1, since I = !Im, we have
!Im = Im cos 2 fJ
The same effect is obtained no matter which sheet is rotated or m which direc-
tion. .....
Historically polarization studies were made to investigate the nature of
light. Today we reverse the procedure and deduce something about the
nature of an object from the polarization state of the light emitted by or
scattered from that object. It has been possible to deduce, from studies of
* The so-called Haidinger's brushes; the interested student is referred to Concepts of
Classical Optics, John Strong, W. H. Freeman & Co., 1958.
Sec. 46-3 POLARIZATION BY REFLECTION 1153
the polarization of light reflected from them, that the grains of cosmic dust
present in our galaxy have been oriented in the weak galactic magnetic field
( ""'2 X 10-4 gauss) so that their long dimension is parallel to this field.
Polarization studies have shown that Saturn's rings consist of ice crystals.
The size and shape of virus particles can be determined by the polarization of
ultraviolet light scattered from them. Much useful information about the
structure of atoms and nuclei is gained from polarization studies of their
emitted radiations in all parts of the electromagnetic spectrum. Thus we
have a useful research technique for structures ranging in size from a galaxy
(,...._,10+ 20 meters) to a nucleus (,...__,10- 14 meter). Polarized light also has
many practical applications in industry and in engineering science.
Fig. 46-8 For a particular angle of incidence Op, the reflected light is completely polar-
ized, as shown. The transmitted light is partially polarized.
1154 POLARIZATION Chap. 46
Fig. 46-9 Polarization of light by a stack of glass plates. Unpolarized light is incident
on a stack of glass plates at Brewster's angle 8p. (Polarization in the plane of the page is
shown by the short lines and polarization normal to the page by the dots.) All light re-
flected out of the original ray is polarized normal to the page. After passing through
several reflecting interfaces, the light in the original ray no longer contains any appreciable
component polarized normal to the page.
or (46-2)
where the incident ray is in medium one and the refracted ray in medium
two. This can be written as
tan Op = n, (46-3)
.... Example 2. We wish to use a plate of glass (n = 1.50) as a polarizer. What is the
polarizing angle? What is the angle of refraction?
From Eq. 46-3,
Op = tan- 1 1.50 = 56.3°.
The angle of refraction follows from Snell's law:
(1) sin Op = n sin Or
~ Double Refraction
In earlier chapters we assumed that the speed of light, and thus the index
of refraction, is independent of the direction of propagation in the medium
and of the state of polarization of the light. Liquids, amorphous solids such
as glass, and crystalline solids having cubic symmetry normally show this
behavior and are said to be optically isotropic. Many other crystalline solids
are optically anisotropic (that is, not isotropic).*
Solids may be anisotropic in many properties. Mica cleaves readily in one
plane only; a cube of crystalline graphite does not have the same electric
resistance between all pairs of opposite faces; a cube of crystalline nickel
magnetizes more readily in certain directions than in others, etc. If a solid
is a mixture of a large number of tiny crystallites, it may appear to be iso-
tropic because of the random orientations of the crystallites. Powdered
* Many transparent amorphous solids such as glasses and plastics become optically
anisotropic when they are mechanically stressed. This fact is useful in engineering design
studies in that strains in gears, bridge structures, etc., can be studied quantitatively by
building plastic models, stressing them appropriately, and examining the optical anisotropy
that results, using polarization techniques. The interested student should consult "Photo-
elasticity," a chapter by H. T. Jessop in Vol. 6 of the Encyclopedia of Physics, edited by
H. Flugge (1958), Springer Verlag, Berlin.
1156 POLARIZATION Chap. 46
Fig. 46-10 Two images, one polarized 90° relative to the other, are formed by a calcite
crystal. The book on which the crystal is lying is Huygens' Opera Reliqua, wherein the
phenomenon of birefringence is discussed.
mica, for example, compacted to a solid mass with a binder, does not exhibit
the cleavage properties that characterize the crystallites making it up .
.Figure 46-10, in which a polished crystal of calcite (CaC0 3 ) is laid over
some printed letters, shows the optical anisotropy of this material; the image
appears double. Figure 46-11 shows a beam of unpolarized light falling on a
calcite crystal at right angles to one of its faces. The single beam splits into
two at the crystal surface. The "double-bending" of a beam transmitted
through calcite, exhibited in Figs. 46-10 and 46-11, is called double refraction.
If the two emerging beams in Fig. 46-11 are analyzed with a polarizing
sheet they are found to be plane-polarized with their planes of vibration at
right angles to each other, a fact discovered by Huygens in 1678. Huygens
used a second calcite crystal to investigate the polarization states of the
beams labeled o and e in the figure.
If experiments are carried out at various angles of incidence, one of the
beams in Fig. 46-11 (represented by the ordinary ray, or o-ray) will be
found to obey Snell's law of refraction at the crystal surface, just like a ray
passing from one isotropic medium into another. The second beam (repre-
sented by the extraordinary ray, or e-ray) will not. In Fig. 46-11, for ex-
ample, the angle of incidence for the incident light is zero but the angle of
refraction of the e-ray, contrary to the prediction of Snell's law, is not. In
general, the e-ray does not even lie in the plane of incidence.
This difference between the waves represented by the o- and e-rays with
respect to Snell's law can be explained in these terms:
Sec. 46-4 DOUBLE REFRACTION 1157
Fig. 46-11 A beam of unpolarized light falling on a calcite crystal is split into two beams
which are polarized at right angles to each other.
1. The a-wave travels in the crystal with the same speed v0 in all directions.
In other words, the crystal has, for this wave, a single index of refraction n 0 ,
just like an isotropic solid.
2. Thee-wave travels in the crystal with a speed that varies with direction
from v0 to a larger value (for calcite) Ve. In other words, the index of refrac-
tion, defined as c/v, varies with direction from n 0 to a smaller value (for cal-
cite) ne.
The quantities n 0 and ne are called the principal indices of refraction for the
crystal. Problem 7 suggests how to measure them. Table 46-1 shows these
indices for six doubly refracting crystals. For three of them the e-wave is
slower; for the other three it is faster. Some doubly refracting crystals (mica,
topaz, etc.) are more complex optically than calcite and require three prin-
cipal indices of refraction for a complete description of their optical proper-
ties. Crystals whose basic crystal structure is cubic (see Fig. 45-13) are
optically isotropic, requiring only one index of refraction.
The behavior for the speeds of the two waves traveling in calcite is sum-
marized by Fig. 46-12, which shows two wave surfaces spreading out from
an imaginary point light source S imbedded in the crystal. The a-wave sur-
1158 POLARIZATION Chap. 46
Table 46-1
PRINCIPAL INDICES OF REFRACTION OF
SEVERAL DO UBL y REFRACTING CRYSTALS
(For sodium light, ;\ = 5890 A)
Crystal Formula no ne ne - no
Fig. 46-12 Huygens' wave surfaces generated by a point source S imbedded in calcite.
The polarization states for three o-rays and three e-rays are shown by the dots and lines,
respectively. Note that in general (ray Sb) the bars representing the polarization direc-
tion are not perpendicular 't'o the e-rays.
Sec. 46-4 DOUBLE REFRACTION 1159
Figure 46-13, which shows a typical calcite crystal that may be obtained
by cleavage from a naturally occurring crystal, shows how to locate the
optic axis. The edges of calcite crystals may have any lengths but the
angles at which the edges intersect always have one or another of two values,
78° 13' or 102° 21'. The optic axis is found by erecting a line at either of
the two corners where three obtuse angles meet (the "blunt" corners), mak-
ing equal angles with the crystal edges. Any line in the crystal parallel to this
line is also an optic axis.
We can use Huygens' principle to study the propagation of light waves in
doubly refracting crystals. Figure 46-14a shows the special case in which
unpolarized light falls at normal incidence on a calcite slab cut from a crystal
in such a way that the optic axis is normal to the surface. Consider a wave-
fron t that, at time t = 0, coincides with the crystal surface. Following
Huygens, we may let any point on this surface serve as a radiating center
for a double set of Huygens' wavelets, such as those in Fig. 46-12. The
plane of tangency to these wavelets represents the new position of this wave-
front at a later time t. The incident beam in Fig. 46-l 4a is propagated
through the crystal without deviation at speed v0 • The beam emerging
from the slab will have the same polarization character as the incident beam.
The calcite slab, in these special circumstances only, behaves like an iso-
tropic material, and no distinction can be made between the o- and the
e-waves.
Figure 46-14b shows two views of another special case, namely, unpolarized
incident light falling at right angles on a slab cut so that the optic axis is
parallel to its surface. In this case also the incident beam is propagated
without deviation. However, we can now identify o- and e-waves that travel
through the crystal with different speeds, v0 and Ve, respectively. These
waves are polarized at right angles to each other.
Some doubly refracting crystals have the interesting property, called dichroism,
in which one of the polarization components is strongly absorbed within the crystal,
1160 POLARIZATION Chap. 46
the other being transmitted with little loss. Dichroism, illustrated in Fig. 46-15,
is the basic operating principle of the commercial Polaroid sheet. The many small
crystallites, imbedded in a plastic sheet with their optic axes parallel, have a polar-
izing action equivalent to that of a single large crystal slab.
Fig. 46-14 Unpolarized light falls at normal incidence on slabs cut from a calcite crystal.
The Huygens' wavelets are appropriate sections of the figure of revolution about the optic
axis represented by Fig. 46-15. (a) No double refraction or speed difference occurs.
(I>) No double refraction occurs but there is a speed difference. (c) Both double refraction
and a speed difference occur. (d) Same as (c) but showing the polarization states and the
emerging rays.
Sec. 46-4 DOUBLE REFRACTION 1161
Note that ray xa represents the shortest optical path for the transfer of light
energy from point x to thee-wavefront. Energy transferred along any other
ray, in particular along ray xb, would have a longer transit time, a conse-
quence of the fact that the speed of e-waves varies with direction.* Figure
46-14d represents the same case as Fig. 46- l 4c. It shows the rays emerging
from the slab, as in Fig. 46-11, and makes clear that the emerging beams are
polarized at right angles to each other, that is, they are cross-polarized.
We now seek to understand, in terms of the atomic structure of optically anisotropic
crystals, how cross-polarized light waves with different speeds can exist. Light is
propagated through a crystal by the action of the vibrating E vectors of the wave on
the electrons in the crystal. These electrons, which experience electrostatic restoring
forces if they are moved from their equilibrium positions, are set into forced periodic
oscillation about these positions and pass along the transverse wave disturbance that
constitutes the light wave. The strength of the restoring forces may be measured
by a force constant k, as for the simple harmonic oscillator discussed in Chapter
15 (see Eq.
In optically isotropic materials the force constant k is the same for all directions
of displacement of the electrons from their equilibrium positions. In doubly re-
fracting crystals, however, k varies with direction. For electron displacements that
lie in a plane at right angles to the optic axis k has the constant value k 0 , no matter
how the displacement is oriented in this plane. For displacements parallel to the
* The student who has not previously read Section 41-6 on Fermat's principle may care
to do so now.
1162 POLARIZATION Chap. 46
(a) (b)
optic axis, k has the larger value (for calcite) ke.* Note carefully that the speed
of a wave in a crystal is determined by the direction in which the E vectors vibrate
and not by the direction of propagation. It is the transverse E-vector vibrations
that call the restoring forces into play and thus determine the wave speed. Note too
that the stronger the restoring force, that is, the larger k, the faster the wave. For
waves traveling along a stretched cord, for example, the restoring force for the trans-
verse displacements is determined by the tension F in the cord. Equation 19-12
shows that an increase in F means an increase in the wave speed v.
Figure 46--16, a long weighted "tire chain" supported at its upper end, provides a
one-dimensional mechanical analogy for double refraction. It applies specifically
too- and e-waves traveling at right angles to the optic axis, as in Fig. 46--14b. If the
supporting block is oscillated, as in Fig. 46--16a, a transverse wave travels along the
chain with a certain speed. If the block is oscillated lengthwise, as in Fig. 46--16b,
another transverse wave is also propagated. The restoring force for the second wave
is greater than for the first, the chain being more rigid in the plane of Fig. 46--16b
than in the plane of Fig. 46--16a. Thus the second wave travels along the chain with
a greater speed.
In the language of optics we would say that the speed of a transverse wave in the
chain depends on the orientation of the plane of vibration of the wave. If we oscillate
the top of the chain in a random way, the wave disturbance at a point along the chain
can be described as the sum of two waves, polarized at right angles and traveling
with different speeds. This corresponds exactly to the optical situation of Fig. 46--14b.
* For doubly refracting crystals with ne > n 0 (see Table 46-1) k for displacements
parallel to the optic axis is smaller than for those at right angles to it. Also, for crystals
with three principal indices of refraction, there will be three principal force constants.
Such crystals have two optic axes and are called biaxial. The crystals listed in Table 46-1
have only a single optic axis and are called uniaxial.
Sec. 46-5 CIRCULAR POLARIZATION 1163
For waves traveling parallel to the optic axis, as in Fig. 46-14a, or for waves in
optically isotropic materials, the appropriate mechanical analogy is a single weighted
hanging chain. Here there is only one speed of propagation, no matter how the upper
end is oscillated. The restoring forces are the same for all orientations of the plane
of polarization of waves traveling along such a chain.
These considerations allow us to understand more clearly the polarization states
of the light represented by the double-wave surface of Fig. 46-12. For the (spherical)
o-wave surface, the E-vector vibrations must be everywhere at right angles to the
optic axis. If this is so, the same force constant ko will always be operative, and the
o-waves will travel with the same speed in all directions. More specifically, if we
draw a ray in Fig. 46-12 from S to the o-wave surface, considered three-dimension-
ally (that is, as a sphere), the E-vector vibrations will always be at right angles to
the plane defined by this ray and the optic axis. Thus these vibrations will always
be at right angles to the optic axis.
For the (ellipsoidal) e-wave surface, the E-vector vibrations in general have a
component parallel to the optic axis. For rays such as Sa in Fig. 46-12 or for the
e-rays of Fig. 46-14b, the vibrations are completely parallel to this axis. Thus a
relatively strong force constant (in calcite) k. is operative, and the wave speed v.
will be relatively high. For e-rays, such as Sb in Fig. 46-12, the parallel component
of the E-vector vibrations is less than 100%, so that the corresponding wave speed
will be less than v.. For ray Sc, in Fig. 46-12, the parallel component is zero, and
the distinction between o- and e-rays disappears.
Plane-polarized
light
(a)
(b)
EE
2
3
EB z
4
EB5
EE
6
EE
7
EE
8
Fig. 46-18 (a) Two plane-polarized waves of equal amplitude and at right angles to each
other are moving in the z direction. They differ in phase by 90°; where one wave has
maximum values, the other is zero. (b) Views of the resultant amplitude of the approach-
ing wave as seen by observers located at the positions shown on the z axis. Note that
each observer will see the resultant vector rotate clockwise with time.
The wave component whose vibrations are at right angles to the optic axis (the
o-wave) can be represented as
Ex = (Eo sin 45°) sin (wt - 90°) = - ~Eo cos wt = -Em cos wt,
the 90° phase shift representing the action of the quarter-wave plate. Note that Ex
ly \y
I
I
I
(a) (b)
Fig. 46-20 Plane-polarized light falls from behind on a quarter-plate oriented so that
the light emerging from the page is circularly polarized. In this case the electric vector
Em rotates clockwise as seen by an observer facing the light source.
Sec. 46-7 SCATTERING OF LIGHT 1167
reaches its maximum value one-fourth of a cycle later than Ey does, for, in calcite,
wave Ex (the o-wave) travels more slowly than wave Ey (thee-wave).
To decide the direction of rotation, let us locate the tip of the rotating electric
vector at two instants of time, (a) t = 0 and (b) a short time t1 later chosen so that wti
is a small angle. At t = 0 the coordinates of the tip of the rotating vector (see Fig.
46-20a) are
and
Figure 46-20b shows that the vector that represents the emerging circular polarized
light is rotating clockwise; by convention such light is called right-circularly polarized,
the observer always being considered to face the light source.
If the plane of vibration of the incident light in Fig. 46-20 is rotated through
±90°, the emerging light will be left-circularly polarized. ~ -·-.
is transferred to the object, where U is the amount of absorbed energy and w the angu-
lar frequency of the light. The student should verify that the dimensions in Eq.
46-5 are consistent.
Fig. 46-21 Light is polarized either partially (c and d) or completely (b) by scattering
from a gas molecule at a.
Sec. 46-8 DOUBLE SCATTERING 1169
* See Scientific American, July 1955, and Bees: Their Vision, Chemical Sense, and
Language, K. von Frisch, Cornell University Press, 1950.
1170 POLARIZATION Chap. 46
Second-scattered
wave, maximum
intensity
wave, zero
intensity
Incident
unpolarized
wave
Fig. 46-22 A double-scattering experiment used by Barkla to show that X-rays are
transverse waves.
In his later studies the student will learn that beams of particles such as electrons,
protons, and pions can be viewed as waves. Scattering (including double scat-
tering) techniques are often used to investigate the polarization characteristics of
such beams.
QUESTIONS
1. Why do sunglasses made of polarizing materials have a marked advantage over those
that simply depend on absorption effects?
2. Unpolarized light falls on two polarizing sheets so oriented that no light is trans-
mitted. If a third polarizing sheet is placed between them, can light be transmitted?
3. Can polarization by reflection occur if the light is incident on the interface from the
side with the higher index of refraction (glass to air, for example)?
4. Is the optic axis of a doubly refracting crystal simply a line or a direction in space?
Has it a direction sense, like an arrow? What about the characteristic direction of a
polarizing sheet?
5. Devise a way to identify the polarizing direction of a sheet of Polaroid.
6. If ice is doubly refracting (see Table 46-1), why don't we see two images of objects
viewed through an ice cube?
7. From Table 46-1, would you expect a quarter-wave plate made from calcite to be
thicker than one made from quartz?
8. Does the e-wave in doubly refracting crystals always travel at a speed given by
c/ne?
9. In Fig. 46-14a and b describe qualitatively what happens if the incident beam falls
on the crystal with an angle of incidence that is not zero. Assume in each case that the
incident beam remains in the plane of the figure.
10. Devise a way to identify the direction of the optic axis in a quarter-wave plate.
11. If plane-polarized light falls on a quarter-wave plate with its plane of vibration
making an angle of (a) 0° or (b) 90° with the axis of the plate, describe the transmitted
light. (c) If this angle is arbitrarily chosen, the transmitted light is called elliptically
polarized; describe such light.
12. What would be the action of a half-wave plate (that is, a plate twice as thick as a
Chap. 46 PROBLEMS 1171
quarter-wave plate) on (a) plane-polarized light (assume the plane of vibration to be at 45°
to the optic axis of the plate), (b) circularly polarized light, and (c) unpolarized light?
13. You are given an object which may be (a) a disk of grey glass, (b) a polarizing sheet,
(c) a quarter-wave plate, or (d) a half-wave plate (see Question 12). How could you
identify it?
14. Can a plane-polarized light beam be represented as a sum of two circularly polarized
light beams of opposite rotation? What effect has changing the phase of one of the cir-
cular components on the resultant beam?
15. How can a right-circularly polarized light beam be transformed into a left-circularly
polarized beam?
16. Could (a) a radar beam and (b) a sound wave in air be circularly polarized?
17. A beam of light is said to be unpolarized, plane-polarized, or circularly polarized.
How could you choose among them experimentally?
18. A parallel beam of light is absorbed by an object placed in its path. Under what
circumstances will (a) linear momentum and (b) angular momentum be transferred to the
object?
19. When observing a clear sky through a polarizing sheet, one finds that the intensity
varies by a factor of two on rotating the sheet. This does not happen when one views a
cloud through the sheet. Can you devise an explanation?
PROBLEMS
1. Unpolarized light falls on two polarizing sheets placed one on top of the other. What
must be the angle between the characteristic directions of the sheets if the intensity of the
transmitted light is (a) one-third the maximum intensity of the transmitted beam or (b)
one-third the intensity of the incident beam? Assume that the polarizing sheet is ideal,
that is, that it reduces the intensity of unpolarized light by exactly 50%.
2. An unpolarized beam of light is incident on a group of four polarizing sheets which
are lined up so that the characteristic direction of each is rotated by 30° clockwise with
respect to the preceding sheet. What fraction of the incident intensity is transmitted?
3. Describe the state of polarization represented by these sets of equations:
4. (a) At what angle of incidence will the light reflected from water be completely
polarized? (b) Does this angle depend on the wavelength of the light?
5. Calculate the range of polarizing angles for white light incident on fused quartz.
Assume that the wavelength limits are 4000 and 7000 A and use the dispersion curve of
Fig. 41-2.
6. A narrow beam of unpolarized light falls on a calcite crystal cut with its optic axis
as shown in Fig. 46-23. (a) Fort = 1.0 cm and for (Ji = 45°, calculate the perpendicular
distance between the two emerging rays x and y. (b) Which is the o-ray and which the
e-ray? (c) What are the states of polarization of the emerging rays? (d) Describe what
1172 POLARIZATION Chap. 46
happens if a polarizer is placed in the incident beam and rotated. (Hint: Inside the crystal
the E-vector vibrations for one ray are always perpendicular to the optic axis and for the
other ray they are always parallel. The two rays are described by the indices n 0 and ne;
in this plane each ray obeys Snell's law.)
Fig. 46-23
7. A prism is cut from calcite so that the optic axis is parallel to the prism edge as shown
in Fig. 46-24. Describe how such a prism might be used to measure the two principal
indices of refraction for calcite. (Hint: See hint in
Problem 6; see also Example 3, Chapter 41.)
8. How thick must a sheet of mica be if it
is to form a quarter-wave plate for yellow light
(X = 5890 A)? Mica cleaves in such a way that the
appropriate indices of refraction, for transmission
at right angles to the cleavage plane, are 1.6049
and 1.6117.
9. Prove that two plane-polarized light waves of
equal amplitude, their planes of vibration being at
right angles to each other, cannot produce inter-
ference effects. (Hint: Prove that the intensity of
the resultant light wave, averaged over one or more
Fig. 46-24
cycles of oscillation, is the same no matter what
phase difference exists between the two waves.)
10. Show that in a parallel beam of circularly polarized light the angular momentum
per unit volume Lv is given by
p
Lv = -,
we
where Pis the power per unit area (watts/cm2, say) of the beam. Start from Eq. 46-5.
11. Assume that a parallel beam of circularly polarized light whose intensity is 100 watts
is absorbed by an object. At what rate is angular momentum transferred to the object?
If the object is a flat disk of diameter 5.0 mm and mass 1.0 X 10-2 gm, after how long a
time (assuming it is free to rotate about its axis) would it attain an angular speed of 1.0
rev /sec? Assume a wavelength of 5000 A.
Light and Quantum Physics
CHAPTER 47
(47-1)
1174 LIGHT AND QUANTUM PHYSICS Chap. 47
~ 301--~~~-+-~~~-+--'~~~-+-~~~-1-~~~--l
I
IN
E
..g
~
;
~201--~~-++-~~~-+-~--'',--+-~~~-1-~~~--1
e5
Fig. 47-1 The spectral radiancy of tungsten at 2000°K. The dashed curve refers to a
cavity radiator at the same temperature. One micron (1 µ) = 10-6 meter = 10 4 A.
The radiancy CR can be interpreted as the area under the plot of%. against X.
In Fig. 47-1 this area-and thus CR-is 23.5 watts/cm 2 • The student should
note the formal similarity between such curves and the Maxwell speed dis-
tribution curve of Section 24-2.
For every material there exists a family of spectral radiancy curves like
that of Fig. 47-1, one curve for every temperature. If such families of
curves are compared, no obvious regularities stand out. A quantitative
understanding in terms of a basic theory presents serious difficulties. Fortu-
nately, it is possible to work with an idealized heated solid, called a cavity
radiator. Its light-emitting properties prove to be independent of any par-
ticular material and to vary in a simple way with temperature. In much the
same way it proved convenient earlier to deal with an ideal gas rather than to
analyze the properties of the infinite variety of real gases. The cavity
radiator is the ideal solid as far as its light-emitting characteristics are con-
cerned. We shall describe in the next two sections how the theoretical study
of cavity radiation in 1900 by the German physicist Max Planck (1858-
1947) laid the foundations of modern quantum physics.
47-2 Cavity Radiators
Let us construct a cavity in each of three metal blocks through the walls of
which a small hole is drilled. Let the blocks be made of any suitable materi-
als; for example, tungsten, tantalum, and molybdenum. Let each block be
Sec. 47-2 CAVITY RADIATORS 1175
1. The radiation from the cavity interior is always more intense than the
radiation from the outside wall. Comparison of the two curves in Fig. 47-1
makes this clear for tungsten. For the three materials given, at 2000°K the
ratio of the radiancy for the outside surface to that for the cavity is 0.259
(tungsten), 0.212 (molybdenum), and 0.232 (tantalum).
2. At a given temperature the radiancy of the hole is identical for all three
radiators, in spite of the fact that the radiancies of the outer surfaces are
different. At 2000°K the cavity radiancy (that is, the hole radiancy) is
90.0 watts/ cm 2 •
3. In contrast to the radiancy of the outer surfaces, the cavity radiancy
CRc varies with temperature in a simple way, namely as
(47-2)
140 .
fl
I
120
\
<J)
100 \
:!:
c
:::,
\
c::,
Fig. 47-2 The spectral radiancy for cavity
radiation at three different temperatures.
-:e ....
(ti
(ti
-
80
60 \
\
~
,<
40
"-
' y::
v
-,,.•4000° K
3000°K
~
t--
~
"
~2000°K
20 I >-
['-... ~
/
I
,V II -~
1 2 3 4
Wavelength, microns
1176 LIGHT AND QUANTUM PHYSICS Chap. 47
through the cylinder wall. It is clear from the photograph that the radiancy
of the cavity interior is greater than that of the cavity walls.
Many of the facts just given about cavity radiation can be deduced from
Fig. 47-4, which shows two cavities made of different materials, of arbitrary
shapes, and with the same wall temperature T. Radiation, described by
CRA, goes from cavity A to cavity B and radiation described by CRB moves
in the opposite direction. If these two rates of energy transfer are not equal,
one end of the composite block will start to heat up and the other end will
start to cool down, which is a violation of the second law of thermodynamics.
(Why?) Thus we must have
(47-4)
where CRc describes the total radiation for all cavities.
Not only the total radiation but also the distribution of radiant energy
with wavelength must be the same for each cavity in Fig. 47-4. This can
be shown by placing a filter between the two cavity openings, so chosen that
it permits only a selected narrow band of wavelengths to pass. Applying
the same argument, we can show that we must have
(47-5)
where CR;>,,c is a spectral radiancy characteristic of all cavities.
>,
u
c:
.!!!
""O
~
~
0Q,)
CL.
(/)
0 1 2 3 4 5 6
Wavelength, microns
Fig. 47-5 The circles show the experimental sp~ctral radiancy data of Coblentz for
cavity radiation. The theoretical formulas of Wien and Planck are also shown, Planck'i,
providing an excellent fit to the data.
In 1900 Max Planck pointed out that if Wien's formula were modified in
a simple way it would prove to fit the data precisely. Planck's formula, an-
nounced to the Berlin Physical Society on October 19, 1900, was
C1 1
(RA= - --- (47-6)
A5 ec2/AT -1
This formula, though interesting and important, was still empirical at that
stage and did not constitute a theory.
Planck sought such a theory in terms of a detailed model of the atomic
processes taking place at the cavity walls. He assumed that the atoms that
1178 LIGHT AND QUANTUM PHYSICS Chap. 47
make up these walls behave like tiny electromagnetic oscillators, each with
a characteristic frequency of oscillation. The oscillators emit electromag-
netic energy into the cavity and absorb electromagnetic energy from it.
Thus it should be possible to deduce the characteristics of the cavity radia-
tion from those of the oscillators with which it is in equilibrium.
Planck was led to make two radical assumptions about the atomic oscil-
lators. As eventually formulated, these assumptions are the following:
These assumptions were radical ones and, indeed, Planck himself resisted
accepting them wholeheartedly for many years. In his words, "My futile
attempts to fit the elementary quantum of action [that is, the quantity h]
somehow into the classical theory continued for a number of years, and they
cost me a great deal of effort."
Consider the application of Planck's hypotheses to a large-scale oscillator
such as a mass-spring system or an LC circuit. It would be a stoutly de-
fended common belief that oscillations in such systems could take place with
any value of total energy and not with only certain discrete values. In the
decay of such oscillations (by friction in the mass-spring system or by re-
sistance and radiation in the LC circuit), it would seem that the mechanical
or electromagnetic energy would decrease in a perfectly continuous way and
not by "jumps.'' There is no basis in everyday experience, however, to dis-
miss Planck's assumptions as violations of "common sense," for Planck's
constant proves to have a very small value, namely
* Later developments show that the correct formula for a harmonic oscillator is
E = (n + !)hv.
This change makes no difference to Planck's conclusions, however.
Sec. 47-4 PHOTOELECTRIC EFFECT 1179
Ill- Example 1. A mass-spring system has a mass m = 1.0 kg and a spring constant
k = 20 nt/meter and is oscillating with an amplitude of 1.0 cm. (a) If its energy is
quantized according to Eq. 47-7, what is the quantum number n? (b) If n changes by
unity, what fractional change in energy occurs?
(a) From Eq. 15-8 the frequency is
1 /k 1 /20 nt/meter
.,, = 271" '\J m = 2 '\J
71" l.O kg = 0.71 cycles/sec.
On the basis of his two assumptions, Planck was able to derive his radiation
law (Eq. 47-6) entirely from theory, receiving a Nobel prize for this accom-
plishment in 1918. His theoretical expressions for the hitherto empirical
constants c1 and c2 were
he
and
where k is Boltzmann's constant (see Section 23-5) and c is the speed of light.
By inserting the experimental values for c1 and c2 , Planck was able to derive
the values of both hand k. Planck described his theory to the Berlin Physi-
cal Society on December 14, 1900. Quantum physics dates from that day.
Planck's ideas soon received re-enforcement from Einstein, who, in 1905,
applied the concepts of energy quantization to a new area of physics, the
photoelectric effect.
Before discussing this effect, it is important to realize that although Planck
had quantized the energies of the oscillators in the cavity walls he still treated
the radiation within the cavity as an electromagnetic wave. Einstein's analy-
sis of the photoelectric effect first pointed out the inadequacy of the wave
picture of light in certain situations.
Fig. 47-6 An apparatus used to study the photoelectric effect. V cannot only be varied
continuously but can be reversed in sign by a switching arrangement not shown.
Fig. 47-7 Some data taken with the apparatus of Fig. 47-6. The applied potential
difference Vis called positive when the cup B in Fig. 47-6 is positive with respect to the
photoelectric surface A. In curve b the incident light intensity has been reduced to one-
half that of curve a.
3.0.
0
...
en
~ 2.0
ro
:;::::
A plot of Millikan's
Fig. 47-8
measurements of the stopping po- ...
c:
Cl)
0
c..
tential at various frequencies for Ill)
sodium. The cutoff frequency vo c:
·a.
c..
1.0
is 4.39 X 1014 cycles/sec.
.8
(/)
0
4.0 6.0 8.0 10.0 12.0
Frequency, 1014 cycles/sec
1182 LIGHT AND QUANTUM PHYSICS Chap. 47
1. Wave theory suggests that the kinetic energy of the photoelectrons
should increase as the light beam is made more intense. However, Fig. 47-7
shows that K max ( = eV 0) is independent of the light intensity; this has been
tested over a range of intensities of 107 .
2. According to the wave theory, the photoelectric effect should occur for
any frequency of the light, provided only that the light is intense enough.
However, Fig. 47-8 shows that there exists, for each surface, a characteristic
cutoff frequency v0 . For frequencies less than this, the photoelectric effect
disappears, no matter how intense the illumination.
3. If the energy of the photoelectrons is "soaked up" from the incident
wave by the metal plate, it is not likely that the "effective target area" for
an electron in the metal is much more than a few atomic diameters. Thus,
if the light is feeble enough, there should be a measurable time lag (see Ex-
ample 2) between the impinging of the light on the surface and the ejection
of the photoelectron. During this interval the electron should be "soaking
up" energy from the beam until it had accumulated enough energy to escape.
However, no detectable time lag has ever been measured. This disagreement is
particularly striking when the photoelectric substance is a gas; under these
circumstances the energy of the emitted photoelectron must certainly be
"soaked out of the beam" by a single atom.
.... Example 2. A metal plate is placed 5 meters from a monochromatic light source
whose power output is 10- 3 watt. Consider that a given ejected photoelectron may
collect its energy from a circular area of the plate as large as ten atomic diameters
(10- 9 meter) in radius. The energy required to remove an electron through the metal
surface is about 5.0 ev. Assuming light to be a wave, how long would it take for such a
"target" to soak up this much energy from such a light source?
The target area is 1r (10- 9 meter)2 or 3 X 10- 18 meter 2 ; the area of a 5-meter
sphere centered on the light source is 41r (5 meters)2 ,..._, 300 meters 2 • Thus, if the
light source radiates uniformly in all directions, the rate P at which energy falls on
the target is given by
3 X 10-18 meter 2 ) .
P = (10- 3 watt) ( 300 meters 2 = 10-23 Joule/sec.
Assuming that all this power is absorbed, we may calculate the time required from
t= ( 5_ ev
10- Joule/sec
23
) (1.6 X 10-
1 ev
19
joul~) ,..._, 20 hr.
stein's hypothesis suggests that light traveling through space behaves not like
a wave at all but like a particle. Millikan, whose experiments verified
Einstein's ideas in every detail, spoke of Einstein's "bold, not to say reckless,
hypothesis.''
Applying the photon concept to the photoelectric effect, Einstein wrote
hv = Eo + Kmax (47-11)
where hv is the energy of the photon. Equation 47-11 says that a photon
carries an energy hv into the surface. Part of this energy (E 0 ) is used in
causing the electron to pass through the metal surface. The excess energy
(hv - E 0 ) is given to the electron in the form of kinetic energy; if the electron
does not lose energy by internal collisions as it escapes from the metal, it
will exhibit it all as kinetic energy after it emerges. Thus K max represents
the maximum kinetic energy that the photoelectron can have outside the
surface; in nearly all cases it will have less energy than this because of internal
losses.
Consider how Einstein's photon hypothesis meets the three objections
raised against the wave-theory interpretation of the photoelectric effect. As
for objection 1 (the lack of dependence of Kmax on the intensity of illumina-
tion), there is complete agreement of the photon theory with experiment.
Doubling the light intensity merely doubles the number of photons and thus
doubles the photoelectric current; it does not change the energy ( = hv) of the
individual photons or the nature of the individual photoelectric processes
described by Eq. 47-11.
Objection 2 (the existence of a cutoff frequency) follows from Eq. 47-11.
If K max equals zero, we have
hv0 = E 0 ,
which asserts that the photon has just enough energy to eject the photo-
electrons and none extra to appear as kinetic energy. This quantity E 0 is
called the work function of the substance. If v is reduced below v0 , the indi-
vidual photons, no matter how many of them there are (that is, no matter
how intense the illumination), will not have enough energy to eject photo-
electrons.
Objection 3 (the absence of a time lag) follows from the photon theory
because the required energy is supplied in a concentrated bundle. It is not
spread uniformly over a large area, as in the wave theory.
Although the photon hypothesis certainly fits the facts of photoelectricity,
it seems to be in direct conflict with the wave theory of light which, as we
have seen in earlier chapters, has been verified in many experiments. Our
modern view of the nature of light is that it has a dual character, behaving
like a wave under some circumstances and like a particle, or photon, under
others. We discuss the wave-particle duality at length in Chapter 48. Mean-
while, let us continue our studies of the firm experimental foundation on
which the photon concept rests.
1184 LIGHT AND QUANTUM PHYSICS Chap. 47
Let us rewrite Einstein's photoelectric equation (Eq. 47-11) by substitut-
ing eV 0 for Kmax (see Eq. 47-9). This yields, after rearrangement,
h E0
V 0 = -v - ~· (47-12)
e e
Thus Einstein's theory predicts a linear relationship between VO and v, in
complete agreement with experiment; see Fig. 47-8. The slope of the experi-
mental curve in this figure should be h/e, or
h ab 2.20 volt - 0.65 volt
.~--------- = 3.9 X 10-15 volt-sec.
e be (10 X 10 14 - 6 X 10 14 ) cycles/sec
We can find h by multiplying this ratio by the electron charge e,
h = (3.9 X 10-15 volt-sec) (1.6 X 10-19 coul) = 6.2 X 10-34 joule-sec.
From a more careful analysis of this and other data, including data taken
with lithium surfaces, Millikan found the value h = 6.57 X 10-34 joule-sec,
with an accuracy of about 0.5%. This agreement with the value of h derived
from Planck's radiation formula is a striking confirmation of Einstein's pho-
ton concept.
~ Example 3. Deduce the work function for sodium from Fig. 47-8.
The intersection of the straight line in Fig. 47-8 with the horizontal axis is the cut-
off frequency vo. Substituting these values yields
Eo = hv0 = (6.63 X 10- 34 joule-sec)(4.39 X 10 14 cycles/sec)
= 2.92 X 10-19 joule = 1.82 ev.
antenna, radiate electromagnetic waves that again have this same frequency
v. Thus, on the wave picture the scattered wave should have the same fre-
quency v and the same wavelength X as the incident wave.
Compton* was able to explain his experimental results by postulating
that the incoming X-ray beam was not a wave but an assembly of photons
of energy E ( = hv) and that these photons experienced billiard-ball-like col-
lisions with the free electrons in the scattering block. The "recoil" photons
emerging from the block constitute, on this view, the scattered radiation.
Since the incident photon transfers some of its energy to the electron with
* P. W. Debye simultaneously and independently offered the same interpretation.
Primary
<P= oo
<P = 45°
Fig. 47-10 Compton's experimental results. The solid
vertical line on the left corresponds to the wavelength )..,
that on the right to )..'. Results are shown for four different
angles of scattering <p. Note that the Compton shift .6.).. for
'P = 90° is h/moc = 0.242 A.
0.750
A,A
1186 LIGHT AND QUANTUM PHYSICS Chap. 47
y y
Photon Electron
-~ x
A v =0
Before After
which it collides, the scattered photon must have a lower energy E'; it must
therefore have a lower frequency v'( = E' /h), which implies a larger wave-
length A'(= e/v'). This point of view accounts, at least qualitatively, for the
wavelength shift LlA. Notice how different this particle model of X-ray
scattering is from that based on the wave picture. Now let us analyze a
single photon-electron collision quantitatively.
Figure 47-11 shows a collision between a photon and an electron, the elec-
tron assumed to be initially at rest and essentially free, that is, not bound to
the atoms of the scatterer. Let us apply the law of conservation of
energy to this collision. Since the recoil electrons may have a speed v that
is comparable with that of light, we must use the relativistic expression
for the kinetic energy of the electron. From Eqs. 47-10 and 8-21 we may
write
hv = hv' 2 + (m - mo)e ,
in which the second term on the right is the relativistic expression for the
kinetic energy of the recoiling electron, m being the relativistic mass and m 0
the rest mass of that particle. Substituting e/A for v (and e/A' for v') and
using Eq. 8-20 to eliminate the relativistic mass m leads us to
he = he+ m e2 ( 1 _ 1) (47-13)
A A' 0
Vl - (v/e) 2 ·
This conclusion, that the momentum of a photon is given by h/A, may also
be deduced from the theory of relativity.
For the electron, the relativistic expression for the linear momentum is
given by Eq. 9-13, or
m 0v
Pe= ·
Vl - (v/c) 2
We can then write for the conservation of the x component of linear mo-
mentum
h h m0v
- = - cos cp
A A'
V +
1 - (v/c)
2 cos (} (47-15)
h m 0v .
0 = - sin cp - sm (}. (47-16)
A' Vl - (v/c) 2
Our immediate aim is to find .:1A ( = ").. ' - A), the wavelength shift of the
scattered photons, so that we may compare it with the experimental results
of Fig. 47-10. Compton's experiment did not involve observations of the
recoil electron in the scattering block. Of the five collision variables (A, ")...', v,
cp, and 0) that appear in the three equations (47-13, 47-15, and 47-16) we
may eliminate two. We chose to eliminate v and 0, which deal only with the
electron, thereby reducing the three equations to a single relation among the
variables.
Carrying out the necessary algebraic steps (see Problem 15) leads to this
simple result:
h
.:1A (= A' - A) = - (1 - cos cp). (47-17)
m 0c
Thus the Compton shift .:1A depends only on the scattering angle cp and not
on the initial wavelength A. Equation 47-17 predicts within experimental
error the experimentally observed Compton shifts of Fig. 47-10. Note from
the equation that .:1A varies from zero (for cp = 0, corresponding to a "graz-
ing" collision in Fig. 47-11, the incident photon being scarely deflected) to
2h/m0 c (for cp = 180°, corresponding to a "head-on" collision, the incident
photon being reversed in direction).
It remains to explain the presence of the peak in Fig. 47-10 for which the
wavelength does not change on scattering. This peak can be understood as
resulting from a collision between a photon and electrons bound in an ionic
core in the scattering block. During photon collisions the bound electrons
behave like the free electrons that we considered in Fig. 47-11, with the ex-
ception that their effective mass is much greater. This is because the ionic
core as a whole recoils during the collision. The effective mass M for a carbon
scatterer is approximately the mass of a carbon nucleus. Since this nucleus
1188 LIGHT AND QUANTUM PHYSICS Chap. 47
contains 6 protons and 6 neutrons, we have approximately that M = 12 X
1840m0 = 22,000m 0 . If we replace m 0 by Min Eq. 47-17, we see that the
Compton shift for collisions with tightly bound electrons is immeasurably
small.
As in the cavity radiation problem (see Eq. 47-7) and the photoelectric effect (see
Eq. 47-11), Planck's constant h is centrally involved in the Compton effect. The
quantity h is the central constant of quantum physics. In a universe in which h = 0
there would be no quantum physics and classical physics would be valid in the sub-
atomic domain. In particular, as Eq. 47-17 shows, there would be no Compton effect
(that is, ~A = 0) in such a universe.
~ Example 4. X-rays with A.= 1.00 A are scattered from a carbon block. The
scattered radiation is viewed at 90° to the incident beam. (a) What is the Compton
shift~;\? (b) What kinetic energy is imparted to the recoiling electron?
(a) Putting~ = 90° in Eq. 47-17, we have, for the Compton shift,
h
~A = - (1 - cos 'P)
moc
6.63 x 10- 34 joule-sec (1 - cos goo)
(9.11 X 10-31 kg)(3.00 X 10 8 meters/sec)
= 2.43 X 10- 12 meter = 0.0243 A.
(b) If we put K for the kinetic energy of the electron, we can write Eq. 47-13 as
he he
~ = ;\' + K.
Since A' =A+ ~A, we obtain
he he
A A+ ~A+ K,
which reduces to
hc~A
K = A(A + ~A)
(6.63 X 10-34 joule-sec)(3.00 X 10 8 meters/sec)(2.43 X 10- 12 meter)
(LOO X 10-10 meter)(l.00 + 0.024) X 10-10 meter
= 4.73 X 10- 17 joule = 295 ev.
The student may show that the initial photon energy E in this case ( = hv = he/A) is
12,400 ev so that the photon lost about 2.3% of its energy in this collision. A photon
whose energy was ten times as large ( = 124,000 ev) can be shown to lose 23% of its
energy in a similar collision. This follows from the fact that ~A does not depend on
the initial wavelength. Hence more energetic X-rays, which have smaller wave-
lengths, will experience a larger per cent increase in wavelength and thus a larger per
cent loss in energy. ...
Wavelength, A
3450 3500
Fig. 47-12 A small portion of the spectrum of iron, in the region 3400 to 3500 A.
and atoms in an electric arc struck between iron electrodes. We shall see
that Planck's quantization ideas, suitably extended, lead to an understand-
ing of line spectra also. The prototype for the study of line spectra is that of
atomic hydrogen; being the simplest atom it has the simplest spectrum.
Line spectra are common in all parts of the electromagnetic spectrum.
Figure 47-13 shows a spectrum of the ,y-rays (XI"./ 10- 12 meter) emitted
from a particular radioactive nucleus, an isotope of mercury. Figure 47-14
shows a spectrum of X-rays (X r-,..; 10-10 meter) emitted from a molybdenum
target when struck by a 35-kev electron beam. The sharp emission lines are
superimposed on a continuous background.
Figure 47-15 shows a spectrum associated with the molecule HCl. It
occurs in the infrared, with A 10-5 meter. This is an absorption spectrum
I"./
300
,J
250
c::
.E
200 r
.
Fig. 47-13 A wavelength plot for a gamma ray .!:: 150
emitted by the nucleus Hg 198 . (From data by Du- ~ :::,
Mond and co-workers.) u
0
100 i I
!
50
I \
i \
·-·-•,,,._.. . . ·:-r-·
\
0
3.10 3.05 3.00
-12
Wavelength (uncorrected), 10 m
1190 LIGHT AND QUANTUM PHYSICS Chap. 47
40
-
~
~ electrons. Note the two sharp
.iii 20 lines rising above a broad con-
-
c Q) tinuous base. The wavelength
c of the most intense line is 7.1 X
10 10-11 meter or 0.71 A. (From
data by Ulrey.)
Wavelength, A
100
~ 80
c:
0 60 i.
:.::::;
a.
...... 40 • I
Cl)
0
..c 20 • /I I II n
<
0 1.0 lJ'" u ' '"" uVII\.
3.8 3.7 3.6 3.5 3.4 3.3
Wavelength, microns
Fig. 47-15 An absorption spectrum of the HCl molecule near A = 3.5 X 10-6 meter =
3.5 µ. (From data by E. S. Imes.)
s::
0
:g_
Fig. 47-16 An oscilloscope trace show- £
..c
ing one strong line and four weak lines (IJ
Frequency
b
vo-s,~
l
ooov
-uv"' r '\ ( o·o
--, ,}
fl"O Fig. 47-17 A portion of the ab-
sorption spectrum of the protons
-
"iii '
c in molecular hydrogen at A ,..__, 43
Q)
where hv is the quantum of energy carried away by the photon that is emitted
from the atom during the transition.
To learn the allowed frequencies predicted by Eq. 47-18, it is necessary to
know the energies of the various stationary states in which a hydrogen atom
can exist. This calculation was first carried out by Bohr on the basis of a
specific model for the hydrogen atom put forward by him. Bohr's model
was highly successful for hydrogen and had a tremendous influence on the
Fig. 47-18 The spectrum of hydrogen. It consists of a number of series of lines, three
of which are shown. Within each series the spectrum lines follow a regular pattern, ap-
proaching a so-called series limit at the short-wave end of the series.
1192 LIGHT AND QUANTUM PHYSICS Chap. 47
further development of the subject; it is now regarded as an important pre-
liminary stage in the development of a more complete theory of quantum
physics.
Let us assume that the electron in the hydrogen atom moves in a circular
orbit of radius r centered on its nucleus. We assume that the nucleus, which
is a single proton, is so massive that the center of mass of the system is es-
sentially at the position of the proton. Let us calculate the energy E of such
an atom.
Writing Newton's second law for the motion of the electron, we have (using
Coulomb's law)
F = ma,
e2 v2
or --=m-·
4'll"E0 r 2 r
e2
U = V(- e) = - - - , (47-20)
4'll"Eor
where V ( = e/411"Eor) is the potential of the proton at the radius of the elec-
tron.
The total energy E of the system is
e2
(47-21)
Since the orbit radius can apparently take on any value, so can the energy
E. The problem of quantizing E reduces to that of quantizing r.
Every property of the orbit is fixed if the radius is given. Equations 47-19,
47-20, and 47-21 show this specifically for the energies K, U, and E. From
Eq. 47-19 we can show that the linear speed v for the electron is also given in
terms of r by
v= \ ) ~ ·
te2 (47-22)
v
~·o = - = (47-23)
211"r
Sec. 47-8 THE HYDROGEN ATOM 1193
= mv = ~ (47-24)
p
'14;;;;;
The angular momentum Lis given by
(47-25)
which tells how r is quantized. Substituting Eq. 47-27 into Eq. 47-21 pro-
duces
me 4 1
E = - -~ - n = 1, 2, 3, ... , (47-28)
8Eo2h2 n2
which gives directly the energy values of the allowed stationary states.
Figure 47-19 shows the energies of the stationary states and their as-
sociated quantum numbers. Equation 47-27 shows that the orbit radius
increases as n 2 . The upper level in Fig. 47-19, marked n = ex:;, corresponds
to a state in which the electron is completely removed from the atom (that
is, E = 0 and r = ex:;). Figure 47-19 also shows some of the quantum jumps
that take place between the different stationary states.
Combining Eqs. 47-18 and 47-28 allows us to write a completely theoreti-
cal formula for the frequencies of the lines in the hydrogen spectrum. It is
v - ~(2-j2 - ~)
8Eo2h3 k2
(47-29)
in which j and k are integers describing, respectively, the lower and the
upper stationary states. The corresponding wavelengths can easily be
1194 LIGHT AND QUANTUM PHYSICS Chap. 47
Fig. 47-19 An energy level diagram for hydrogen showing the quantum number n for
each level and some of the transitions that appear in the spectrum. An infinite number
of levels is crowded in between the levels marked n = 6 and n = oo. Compare this figure
carefully with Fig. 47-18.
Sec. 47-8 THE HYDROGEN ATOM 1195
Table 47-1
Quantum Number
Name
of Wavelength, A
series
j (lower state) k (upper state)
1 2 1216
1 3 1026
Lyman 1 4 970
1 5 949
1 6 940
1 00 912
2 3 6563
2 4 4861
Balmer 2 5 4341
2 6 4102
2 7 3970
2 00 3650
3 4 18751
3 5 12818
Paschen 3 6 10938
3 7 10050
3 8 9546
3 00 8220
1J11,, Example 5. Calculate the binding energy of the hydrogen atom (the energy
binding the electron to the nucleus) from Eq. 47-28.
The binding energy is numerically equal to the energy of the lowest state in Fig.
47-19. The largest negative value of E in Eq. 47-28 is found for n = 1. This yields
me 4
E= ---
8Eih2
(9.11 X 10-31 kg)(l.60 X 10- 19 coul) 4
(8)(8.85 X 10- 12 coul 2 /nt-m 2)2(6.63 X 10-34 joule-sec) 2
-2.17 X 10-18 joule = -13.6 ev,
which agrees with the experimentally observed binding energy for hydrogen.
1196 LIGHT AND QUANTUM PHYSICS Chap. 47
47-9 The Correspondence Principle
Although all theories in physics have limitations, they usually do not
break down abruptly but in a continuous way, yielding results that agree less
and less well with experiment. Thus the predictions of Newtonian mechanics
become less and less accurate as the speed is made to approach that of light.
A similar relationship must exist between quantum physics and classical
physics; it remains to find the circumstances under which the latter theory
is revealed as a special case of the former.
The radius of the lowest energy state in hydrogen (the so-called ground
state) is found by putting n = 1 in Eq. 47-27; it turns out to be 5.3 X 10-11
meter. If n = 10,000, however, the radius is (10,000) 2 times as large or 5.3
mm. This "atom" is so large that we suspect that its behavior should be
accurately described by classical physics. Let us test this by computing the
frequency of the emitted light on the basis of both classical and quantum
assumptions. These calculations should differ at small quantum numbers
but should agree at large quantum numbers. The fact that quantum physics
reduces to classical physics at large quantum numbers is called the correspond-
ence principle. This principle, credited to Niels Bohr, was very useful during
the years in which quantum physics was being developed. Bohr, in fact,
based his theory of the hydrogen atom on correspondence principle argu-
ments.
Classically, the frequency of the light emitted from an atom is equal to v0 ,
its frequency of revolution * in its orbit. This can be expressed in terms of a
quantum number n by combining Eqs. 47-23 and 47-27 to obtain
me4 2
Vo=----· (47-30)
8t:02h3 n3
me4
v = 8t:02h3
[ 1 1
(n - 1)2 - ;)i
J
me4 [ 2n - 1 J (47-31)
= 8Eo 2 h3 (n - 1) 2n 2 •
* Integral multiples of the frequency also exist but may be ignored without affecting the
present argument.
Chap. 47 QUESTIONS 1197
Table 47-2
THE CORRESPONDENCE PRINCIPLE AS APPLIED TO THE
HYDROGEN ATOM
Frequency of
Frequency of I Transition to
Quantum Revolution in
Next Lowest Differ-
Number, Orbit
State ence, %
n (Eq. 47-30)
(Eq. 47-31)
cycles/sec
cycles/sec
2 8.20 x 10 14 24.6 x 10 14 67
5 5.26 x 10 13 7.38 x 10 13 29
10 6.57 x 10 12 7.72 x 10 12 14
50 5.25 x 10 10 5.42 x 10 10 3
100 6.578 x 109 6.677 x 109 1.5
1,000 6.5779 x 106 6.5878 X 10 6 0.15
10,000 6.5779 x 10 3 6.5789 x 10 3 0.015
QUESTIONS
1. "Pockets" formed by the coals in a coal fire seem brighter than the coals themselves.
Is the temperature in such pockets appreciably higher than the surface temperature of an
exposed glowing coal?
2. The relation R = <rT 4 (Eq. 47-2) is exact for true cavities and holds for all tempera-
tures. Why don't we use this relation as the basis of a definition of temperature at, say,
100°C?
3. Do all incandescent solids obey the fourth-power law of cemperature, as Eq. 47-3
seems to suggest?
4. A hole in the wall of a cavity radiator is sometimes called a black body. Why?
5. It is stated that if we look into a cavity whose walls are maintained at a constant
temperature no details of the interior are visible. Does this seem reasonable?
6. How can a photon energy be given by E = hv (Eq. 47-10) when the very presence of
the frequency v in the formula implies that light is a wave?
7. In the photoelectric effect, why does the existence of a cutoff trequency speak in favor
of the photon theory and against the wave theory?
8. Why are photoelectric measurements so sensitive to the nature of the photoelectric
surface?
9. Does Einstein's theory of photoelectricity, in which light is postulated to be a pho-
ton, invalidate Young's interference experiment?
10. List and discuss carefully the assumptions made by Planck in connection with
the cavity radiation problem, by Einstein in connection with the photoelectric effect, and
by Bohr in connection with the hydrogen atom problem.
11. In Bohr's theory for the hydrogen atom orbits, what is the implication of the fact
that the potential energy is negative and is greater in magnitude than the kinetic energy?
12. Can a hydrogen atom absorb a photon whose energy exceeds its binding energy
(13.6 ev)?
1198 LIGHT AND QUANTUM PHYSICS Chap. 47
13. Discuss Example 1 in terms of the correspondence principle.
14. According to classical mechanics, an electron moving in an orbit should be able to do
so with any angular momentum whatever. According to Bohr's theory of the hydrogen
atom, however, the angular momentum is quantized according to L = nh/2r. Reconcile
these two statements, using the correspondence principle.
PROBLEMS
1. At what wavelength does a cavity radiator at 6000°K radiate most per unit wave-
length? Solve either analytically or graphically.
2. Using Fig. 47-1, verify that the emissivity of tungsten at 2000°K is 0.259.
3. A cavity radiator at 6000°K has a hole 0.10 mm in diameter drilled in its wall. At
what rate do photons in the range 5500 to 5510 A escape from this hole?
4. A cavity whose walls are held at 4000°K has a circular aperture 5.0 mm in diameter.
(a) At what rate does energy in the visible range (defined to extend from 0.40 to 0.70 µ.)
escape from this hole? (b) What fraction of the total radiation escaping from the cavity
does this represent? Solve either analytically or graphically.
5. Show that Wien's law (p. 1177) is a special case of Planck's law (Eq. 47--6) for short
wavelengths or low temperatures.
6. Solar radiation falls on the earth at a rate of 2.0 cal/cm2-min. How many photons/
cm2-min is this, assuming an average wavelength of 5500 A?
7. A spectral emission line, important in radioastronomy, has a wavelength of 21 cm.
To what photon energy does this correspond?
8. The energy required to remove an electron from sodium is 2.3 ev. Does sodium show
a photoelectric effect for orange light, with JI. = 6800 A?
9. Light of a wavelength 2000 A falls on an aluminum surface. In aluminum 4.2 ev
are required to remove an electron. What is the kinetic energy of (a) the fastest and (b)
the slowest emitted photoelectrons? (c) What is the stopping potential?. (d) What is the
cutoff wavelength for aluminum?
10. The work function for a clean lithium surface is 2.3 ev. Make a rough plot of the
stopping potential Vo versus the frequency of the incident light for such a surface.
11. In Example 2 suppose that the "target" is a single gas atom of 1.0 A radius and that
the intensity of the light source is reduced to 10-5 watt. If the binding energy of the most
loosely bound electron in the atom is 20 ev, what time lag for the photoelectric effect is
expected on the basis of the wave theory of light?
12. A 100-watt sodium vapor lamp radiates uniformly in all directions. (a) At what
distance from the lamp will the average density of photons be 10/cm3? (b) What is the
average density of photons 2.0 meters from the lamp? Assume the light to be mono-
chromatic, with JI. = 5890 A.
13. Show, by analyzing a collision between a photon and a free electron (using rela-
tivistic mechanics), that it is impossible for a photon to give all of its energy to the free
electron. In other terms, the photoelectric effect cannot occur for completely free elec-
trons; the electrons must be bound in a solid or in an atom.
14. Calculate the per cent change in photon energy for a Compton collision with ,pin
Fig. 47-11 equal to 90° for radiation in (a) the microwave range, with JI. = 3.0 cm, (b)
the visible range, with JI. = 5000 A, (c) the X-ray range, with JI. = 1.00 A, and (d) the
gamma ray range, the energy of the gamma-ray photons being 1.0 Mev. What are your
conclusions about the importance of the Compton effect in these various regions of the
electromagnetic spectrum, judged solely by the criterion of energy loss in a single Compton
encounter?
15. Carry out the necessary algebra to eliminate v and fJ from Eqs. 47-13, 47-15, and
47-16 to obtain the Compton shift relation (Eq. 47-17).
16. (a) Using Bohr's formula, calculate the three longest wavelengths in the Balmer
series. (b) Between what wavelength limits does the Balmer series lie?
Chap. 47 PROBLEMS 1199
17. In the ground state of the hydrogen atom, according to Bohr's theory, what are (a)
the quantum number, (b) the orbit radius, (c) the angular momentum, (d) the linear
momentum, (e) the angular velocity, (f) the linear speed, (g) the force on the electron, (h)
the acceleration of the electron, (i) the kinetic energy, (j) the potential energy, and (k) the
total energy?
18. How do the quantities (b) to (k) in Problem 17 vary with the quantum number?
19. How much energy is required to remove an electron from a hydrogen atom in a state
with n = 8?
20. A hydrogen atom is excited from a state with n = 1 to one with n = 4. (a) Cal-
culate the energy that must be absorbed by the atom. (b) Calculate and display on an
energy-level diagram the different photon energies that may be emitted if the atom returns
to its n = 1 state. (c) Calculate the recoil speed of the hydrogen atom, assumed initially
at rest, if it makes the transition from n = 4 ton = 1 in a single quantum jump.
21. A hydrogen atom in a state having a binding energy (this is the energy required to
remove an electron) of 0.85 ev makes a transition to a state with an excitation energy
(this is the difference in energy between the state and the ground state) of 10.2 ev. (a)
Find the energy of the emitted photon. (b) Show this transition on an energy-level dia-
gram for hydrogen, labeling the appropriate quantum numbers.
22. Show on an energy-level diagram for hydrogen the quantum numbers corresponding
to a transition in which the wavelength of the emitted photon is 1216 A.
23. A neutron, with kinetic energy of 6.0 ev, collides with a resting hydrogen atom in
its ground state. Show that this collision must be elastic (that is, kineti.c energy must
be conserved).
24. Apply Bohr's theory to singly ionized helium, that is, to a helium atom with one
electron removed. What 1·8lationship exists between this spectrum and the hydrogen
spectrum?
25. Using Bohr's theory, calculate the energy required to remove the electron from
singly ionized helium.
26. Positronium. Apply Bohr's theory to the positronium atom. This consists of a
positive and a negative electron revolving around their center of mass, which lies halfway
between them. (a) What relationship exists between this spectrum and the hydrogen
spectrum? (b) What is the radius of the ground state orbit? (Hint: It will be necessary
to analyze this problem from first principles because this "atom" has no nucleus; both
· particles revolve about a point halfway between them.)
27. Muonic atoms. Apply Bohr's theory to a muonic atom, which consists of a nucleus
of charge Ze with a negative muon (an elementary particle with a charge of -e and a mass
m that is 207 times as large as the electron mass; see Appendix F) circulating about it.
Calculate (a) the radius of the first Bohr orbit, (b) the ionization energy, and (c) the wave-
length of the most energetic photon that can be emitted. Assume that the muon is
circulating about a hydrogen nucleus (Z = 1).
28. If an electron is rotating in an orbit at frequency vo, classical electromagnetism
predicts that it will radiate energy not only at this frequency but also at 2vo, 3vo, 4vo, etc.;
see footnote on p. 1104. Show that this is also predicted by Bohr's theory of the hydrogen
atom in the limiting case of large quantum numbers.
29. In Table 47-2 show that the quantity in the last column is given by
~F
::en~G
beam)
Fig. 48-1 The apparatus of Davisson and Germer. Electrons from filament Fare accel-
erated by a variable potential difference V. After "reflection" from crystal C they are
collected by detector D.
In 1926 Elsasser pointed out that the wave nature of matter might be
tested in the same way that the wave nature of X-rays was first tested,
namely, by allowing a beam of electrons of the appropriate energy to fall on
a crystalline solid. The atoms of the crystal serve as a three-dimensional
array of diffracting centers for the electron "wave"; we should look for strong
diffracted peaks in certain characteristic directions, just as for X-ray diffrac-
tion.
This idea was tested by C. J. Davisson and L. H. Germer in this country
and by G. P. Thomson in Scotland.* Figure 48--1 shows the apparatus of
* For an historical account of Thomson's researches see "Early Work in Electron
Diffraction," Sir George Thomson, p. 821, Am. J Phys., December, 1961.
1202 WAVES AND PARTICLES Chap. 48
35 40 45 50 55 60 65 70 75
Kinetic energy, ev
Fig. 48-3 The strong diffracted beam at <f> = 50° and V = 54 volts arises from wave-like
"reflection" from the family of atomic planes shown, for d = 0.91 A. The Bragg angle
(} is 65°. For simplicity, refraction of the diffracted wave as it leaves the surface is ignored.
Sec. 48-2 ATOMIC STRUCTURE AND STANDING WAY"1S 1203
For the conditions of Fig. 48-3 the effective interplanar spacing d can be
shown by X-ray analysis to be 0.91 A. Since cp equals 50°, it follows that
O equals 90° - ! X 50° or 65°. The wavelength to be calculated from Eq.
48-2, if we assume m = 1, is
").. = 2d sine = 2(0.91 A) (sin 65°) = 1.65 A.
The wavelength calculated from the de Broglie relationship A = h/p is, for
54-ev electrons (see Example 1), 1.64 A. This excellent agreement, com-
bined with much similar evidence, is a convincing argument for believing that
electrons are wave-like in some circumstances.
Not only electrons but all other particles, charged or uncharged, show
wave-like characteristics. Beams of slow neutrons from nuclear reactors
are routinely used to investigate the atomic structure of solids. Figure 48-4
shows a "neutron diffraction pattern" for finely powdered lead.
The evidence for the existence of matter waves with wavelengths given
by Eq. 48-1 is strong indeed. Nevertheless, the evidence that matter is
.....
<J)
::,
500
c 400
.E
.... 300
<J)
a.
<JI 200
c
.... 100
0
.....
::,
<J) 0
z 19° 25° 31 ° 37° 43° 49° 55° 61° 67°
20, degrees
Fig. 48-4 A diffraction pattern for powdered lead, using a monochromatic neutron beam
from a nuclear reactor. The peaks represent "Bragg reflections" from the various atomic
planes, IJ being the corresponding "Bragg angle."
composed of particles remains equally strong; see Fig. 10-10. Thus, for
matter as for light, we must face up to the existence of a dual character;
matter behaves in some circumstances like a particle and in others like a wave.
(a)
~\
Fig. 48-6 (a) A stretched string of length l
clamped between rigid supports. (b) A
particle of mass m and velocity v confined to
move between rigid walls a distance l apart.
19-11--1> (b)
m v
1206 WAVES AND PARTICLES Chap. 48
Fig. 48-7 Three quantized modes of vibration for the string of Fig. 48-6a. The figure
also represents three of the quantized wave functions for the particle of Fig. 48-6b. The
ordinate is a displacement amplitude in the first case and a wave-function amplitude in
the second.
2Z
A=-·
n
Replacing Aby h/p (see Eq. 48-1) leads to
·nh
p = --, (48-5)
2Z
which shows that the linear momentum of the particle is quantized. The
momentum p ( = mv) is related to the energy E (which is entirely kinetic
and is equal to !mv2 ) by
p = -v'w. (48-6)
Combining Eqs. 48-5 and 48-6 leads to the quantization condition for E,
or
h2
E = n2 - - n = 1, 2, 3, .... (48-7)
8ml2
The particle cannot have any energy, as we would expect classically, but
only energies given by Eq. 48-7.
The matter wave is described, in strict analogy with Eq. 48-4, by
'11" =
n1rx]
['11" m sin -Z- cos wt n = 1, 2, 3, .... (48-8)
Figure 48-7 can serve equally well to show how the amplitude of the standing
matter waves for the ~tates of motion corresponding ton = 1, 2, and 3 varies
throughout the box. We see clearly in this problem how the act of localizing
or bounding a particle leads to energy quantization.
~ Example 2. Consider an electron (m = 9.1 X 10-31 kg) confined by electrical
forces to move between two rigid "walls" separated by 1.0 X 10-9 meter, which is
about five atomic diameters. Find the quantized energy values for the three lowest
stationary states.
From Eq. 48-7, for n = 1, we have
_ h2 2 _ 2 (6.6 X 10-34 joule-sec)2
E - n 8ml 2 - (l) (8)(9.1 X 10- 31 kg)(l.O X 10-9 meter) 2
= 6.0 X 10-20 joule = 0.38 ev.
The energies for the next two states (n = 2 and n = 3) are 2 2 X 0.38 ev = 1.5 ev and
3 2 X 0.38 ev = 3.4 ev.
Example 3. Consider a grain of dust (m = 1.0 µgm = 1.0 X 10-9 kg) confined to
move between two rigid walls separated by 0.1 mm ( = 10-4 meter). Its speed is only
10-5 meter/sec, so that it requires 100 sec to cross the gap. What quantum number
describes this motion?
The energy is
E ( = K) = imv 2 = i(10- 9 kg)(l0- 6 meter/sec) 2
= 5 X 10-22 joule.
1208 WAVES AND PARTICLES Chap. 48
Solving Eq. 48-7 for n yields
n -- v
- ~ E l-h -- v
8m1c., - I( 8) (10 -9 k g )( 5 X 10
-22 . 1)
JOU e.
( 10-4 meter
6 .6 X _ .
10 34 JOU 1e-sec
)
= 3 x 1014.
Even in these extreme conditions the quantized nature of the motion would never be
apparent; we cannot distinguish experimentally between n = 3 X 10 14 and n = 3
X 10 14 + 1. Classical physics, which fails completely for the problem of Example 2,
works extremely well for this problem. .,..
48-4 Meaning of '11
Max Born first suggested that the quantity '11 2 at any particular point is a
measure of the probability that the particle will be near that point. More ex-
actly, if a volume element dV is constructed at that point, the probability
that the particle will be found in the volume element at a given instant is
'112 dV. This interpretation of '11 provides a statistical connection between
=38e,
1~ -
0 ~ x
--
1z Fig. 48-8 The "probability functions"
for three states of motion of the parti-
cle of Fig. 48-6b, along with the corre-
sponding quantized energies for the
conditions of Example 2. The horizon-
tal lines show the predictions of classi-
cal mechanics, in which the probability
function is cons.tant for all positions of
the particle.
I""' ~ = 3
~=3•"
O x l
the wave and the associated particle; it tells us where the particle is likely to
be, not where it is.
For the particle confined between rigid walls the probability that the par-
ticle will lie between two planes that are distance x and x + dx from one wall
(see Fig. 48-8) is given by
n1r
'11 2 dx = 'Vm 2 sin2 - x cos 2 wt dx.
l
Since we are more interested in the particle's spatial location than in its time
behavior, we average '1' 2 over one cycle of the motion. This is equivalent
to replacing cos 2 wt by its average value for one cycle, namely one-half, or
- n1rx
'112 = 1-'¥ 2 sin2 - - n = 1, 2, 3, .... (48-9)
2 m l
Figure 48-8 shows '11 2 for the three stationary states corresponding to
n = 1, 2, and 3. Note that for n = 1 the particle is more likely to be near
Sec. 48-4 MEANING OF '1i° 1209
the center than the ends. This is in sharp contradiction to the results of
classical physics, according to which the particle has the same probability of
being located anywhere between the walls, as shown by the horizontal line
in Fig. 48-8.
The problem of a particle confined between rigid walls has little real applica-
tion in physics. We would prefer to illustrate the wave mechanics of Schro-
dinger by applying it to a more experimentally realizable situation, such
as the hydrogen atom. Only mathematical complexity prevents us from
doing this. We state without proof that when this problem is solved by
wave mechanics the motion of the electron in the ground state of the atom,
defined by putting n = 1 in Eq. 47-28, is described by the following wave
function,
'1i" = f _~
'\j 1ra3
e-r/a COS wt '
(48-10)
h2eo
where a=---·
1rme2
Averaging over the time (that is, replacing cos wt by!) yields for the probability P(r)
2
-- 4r 2
P(r) = -3 e-2r1a.
a
Figure 48-9 shows a plot of this function. Note that the most probable location for
the electron corresponds to the first Bohr radius. Thus in wave mechanics we do not
say that the electron in the n = 1 state in hydrogen goes around the nucleus in a
circular orbit of 0.529 A radius but only that the electron is more likely to be found at
this distance from the nucleus than at any other distance, either larger or smaller. <111111
Fig. 48-10 An incident beam of electrons is diffracted at the slit in screen A, forming a
typical diffraction pattern on screen B. If the slit is made narrower, the pattern becomes
wider.
Sec. 48-5 THE UNCERTAINTY PRINCIPLE 1211
Our inherent inability to describe the motions of electrons in a classical way finds
expression in the uncertainty principle, enunciated by Werner Heisenberg in 1927.
To formulate this principle, consider a beam of monoenergetic electrons of speed v0
moving from left to right in Fig. 48-10. Let us set ourselves the task of measuring
the position of a particular electron in the vertical (y) direction and also its velocity
component Vy in this direction. If we succeed in carrying out these measurements
with unlimited accuracy, we can then claim to have established the position and
motion of the electron (or one component of it at least) with precision. However, we
shall see that it is impossible to make these two measurements simultaneously with un-
limited accuracy.
To measure y we block the beam with an absorbing screen A in which we put a slit
of width fj.y. If an electron gets through the slit, its vertical position must be known
to this accuracy. By making the slit narrower, we can improve the accuracy of this
vertical position measurement as much as we wish.
Since the electron is a wave, it will undergo diffraction at the slit, and a photo-
graphic plate placed at B in Fig. 48-10 will reveal a typical diffraction pattern. The
existence of this diffraction pattern means that there is an uncertainty in the values of
Vy possessed by the electrons emerging from the slit. Let Vya be the value of Vy that
corresponds to an electron landing at the first minimum on the screen, marked by
point a and described by a characteristic angle ea. We take Vya as a rough measure of
the uncertainty /j.vy in v11 for electrons emerging from the slit.
The first minimum in the diffraction pattern is given by Eq. 44-2, or
. e
sm a = fj.y.
;\
If we assume that ea is small enough, we can write this equation as
;\
ea !'-' fj.y. (48-11)
This is so far beyond the possibility of measurement (a nucleus is only about 10-15
meter in diameter) that we can assert that for heavy objects like bullets the un-
certainty principle sets no limits whatever on our measuring procedures. Once again
the correspondence principle shows us how quantum physics reduces to classical
physics under the appropriate circumstances. .._
The uncertainty relation shows us why it is possible for both light and matter to
have a dual, wave-particle, nature. It is because these two views, so obviously op-
posite to each other, can never be brought face to face in the same experimental
situation. If we devise an experiment that forces the electron to reveal its wave
character strongly, its particle character will always be inherently fuzzy. If we
modify the experiment to bring out the particle character more strongly, the wave
character necessarily becomes fuzzy. Matter and light are like coins that can be
made to display either face at will but not both simultaneously. Niels Bohr first
pointed out in his principle of complementarity how the ideas of wave and of particle
complement rather than contradict each other.
QUESTIONS
1. How can the wavelength of an electron be given by A = h/p when the very presence
of the momentum pin this formula implies that the electron is a particle?
2. How could Davisson and Germer be sure that the "54-volt" peak of Fig. 48-2 was a
first-order diffraction peak, that is, that m = 1 in Eq. 48-2?
3. In a repetition of Thomson'r- experiment for measuring e/m for the electron (see
Section 33-8), a beam of 104-ev electrons is collimated by passage through a slit of width
0.50 mm. Why is the beam-like character of the emerging electrons not destroyed by
diffraction of the electron wave at this slit?
4. Why is the wave nature of matter not more apparent to our daily observations?
5. Apply the correspondence principle to the problem of a particle confined between
rigid walls, showing that those features which seem "strange" (that is, the quantization of
energy and the nonuniformity of the probability functions of Fig. 48--8) become undetect-
able experimentally at large quantum numbers.
6. In the n = 1 mode, for a particle confined between rigid walls, what is the probability
that the particle will be found in a small volume element at the surface of either wall?
7. A standing wave can be viewed as the superposition of two traveling waves. Can
you apply this to the problem of a particle confined between rigid walls, giving an interpre-
tation in terms of the motion of the electron?
8. What is the physical significance of the wave function 'Ir?
9. Why does the concept of Bohr orbits violate the uncertainty principle?
10. How can the predictions of wave mechanics be so exact if the only information we
have about the positions of the electrons is statistical?
11. Make up some numerical examples to show the difficulty of getting the uncertainty
principle to reveal itself during experiments with an object whose mass is about 1 gm.
12. Figure 48-8 shows that for n = 3 the probability function for a particle confined
between rigid walls is zero at two points between the walls. How can the particle ever
move across these positions? (Hint: Consider the implications of the uncertainty prin-
ciple.)
13. The uncertainty principle can be stated in terms of angular quantities (compare
Eq. 48-15) as
where t.L is the uncertainty in the angular momentum and ilcf, the uncertainty in the
angular position. For electrons in atoms the angular momentum has definite quantized
values, with no uncertainty whatever. What can we conclude about the uncertainty in the
angular position and about the validity of the orbit concept?
1214 WA YES AND PARTICLES Chap. 48
PROBLEMS
V-1 Introduction
In Chapter 39 we sought to make the existence of electromagnetic waves plausi'-
ble by showing that such waves are consistent with Maxwell's equations as
expressed in Table 38-3. Here we seek to start from Maxwell's equations and
derive from them a differential equation whose solutions will describe electro-
magnetic waves. We will show directly that the speed c of such waves is given by
Eq. 39-15, or c = 1 / ~ .
We followed a similar program in Supplementary Topic III for mechanical
waves on a stretched string. Starting from Newton's laws of motion we derived
a differential equation (Eq. III-1) whose solutions (Eq. III-2) described such
waves. We showed further that the speed v of these waves is given by Eq. III-3,
orv = ~ -
In Table 38-3 we wrote Maxwell's equations as
Eo :f E · dS = q, (V-1)
:f B · dS = 0, (V-2)
Only if p is a constant over all parts of the volume can we factor it out and write
p = m/r.
To carry out our program it is desirable to recast Maxwell's equations in the
form of equalities that apply at each point in space rather than as integrals that
apply to various regions of space. In other words, we wish to convert Maxwell's·
equations from the integral form of Eqs. V-1 to 4 into differential form. We will
then be able to relate E and Bat a point to the charge density and current density
at that point.
Ex= - - ,
av --,
av and Ez =
av
ax ay az
E= - . av- + J
(1 . av- + kav)
- · (V-5)
ax i)y az
E = -VV,
V =
.a + J.a
I -
a
- + k ~• (V-6)
ax ay i)z
This operator is useful in dealing with scalar and vector fields (see Sections 16-8
and 18-7 for examples of such fields). Given any scalar field'+' we may form a
vector field, called the gradient of'+' and written as grad'+' or V'+', simply by apply-
ing the operator V to '+'· Given a vector field U = Uxi Uyj + +
U.k we may
apply the operator V to it in two different ways. One way is to take the dot
product of V and U, yielding the scalar field called the divergence of U and written
SUPPLEMENTARY TOPIC V 3
as div U or V · U. The other way is to take the cross product of V and U, yielding
the vector field called the curl of U and written curl U or V X U. These opera-
tions may be summarized as
grad if;
. af + J. af
= Vif; = 1 -
af
- +k -
ax ay az
div u = v . u = aux + au y+ au z
ax ay az
auz auy)
curl u = v x u = i ( ay - az
+ j (aux - auz)
az ax
auy- aux)
+ I.. ( -ax - ay
Note that grad if; and curl U are vectors, whereas div U is a scalar. The student
can gain some familiarity with these operations by the following exercises: (1)
prove that curl (grad if;) = 0 and (2) prove that div (curl U) = 0.
Another frequently occuring operator is 'V 2 ("del squared"). It is simply
V · V, or, as the student can show from Eq. V-6,
a2 a2 a2
v2 = v . v = -- + ~ + ~·
ax 2 ay 2 az 2
When we apply 'V 2 to a scalar field if;, we obtain
(V-10)
In this section we show how to cast the first two of Maxwell's equations (Eqs.
V-1, 2) into differential form. Let us apply Eq. V-1 to a differential volume ele-
ment shaped like a rectangular parallelepiped and containing a point P at (and
near) which an electric field exists (see Fig. V-la). PointP is located at x, y, z in
the reference frame of Fig. V-lb and the edges of the parallelepiped have lengths
dx, dy, and dz.
We can write the surface area vector for the rear face of the parallelepiped as
dS = - i dy dz. The minus sign enters because dS is defined to point in the
direction of the outward normal, which is defined by -i. For the front face we
have dS = +i dy dz.
4 SUPPLEMENTARY TOPIC V
4=]1dz
~
• P(x,y,z)
O>-------y
dy
(a) x (b)
Fig. V-1
If the electric field at the rear face is E, that at the front face, which is a distance
dx away from the rear face, is E +(
aE/ ax) dx, the latter term representing the
change in E associated with the change dx in x.
The flux through the entire surface of the parallelepiped is .rf E · dS and the
contribution to this flux due to these two faces alone is
= dx dy dz ( -aE · i) = dx dy dz - a (E • i)
ax ax
= dx dy dz-·
aE,,
ax
With similar contributions from the other four faces the total electric flux becomes
...,,r E · dS = dx dy dz (a~
-
ax
+ -a~
ay
+ -a~)
az
·
From Eq. V-8 we may write this as
Now the right-hand side of Eq. V-1, which gives the charge enclosed by the
surface, may be written in general as q = Jp dr and, in particular, for the differ-
ential volume element at P, as
q = p dx dy dz, (V-13)
where p is the charge per unit volume at P. Substituting Eqs. V-12 and 13 into
Eq. V-1 and canceling the common factor dx dy dz, we have finally
di~ I dy
•P
0,1-------y
x
(a) (b)
Fig. V-2
V-2b and the sides of the rectangle, which is parallel to the x-y plane, have lengths
dx and dy. Going around the path, as shown by the arrows, we have
§ B · di = B · ( -j dy) (rear side)
+ B · ( +i dx) (left side)
§ B · di = dx dy (aB · j - aB ·
ax ay
i)
= dx dy [_i!__
ax
(B · j) - ~ ay (B · i) J
= dx dy (aBy _ aB.,)· (V-16)
ax ay
Now in the right-hand side of Eq. V-3 i is the current enclosed by the path and
d<PE/dt is the change in electric flux through the enclosed surface. Hence, if
J is taken to represent the current density and dS( = k dx dy) the surface area
vector, we can write
i = J · dS = J · (k dx dy) = dx dy J z (V-17)
and
d<PE aE aE
- = - · dS = - · (k dx dy)
dt at at
or
d<PE = aE. dx d . (V-18)
dt at Y
6 SUPPLEMENTARY TOPIC V
Substituting Eqs. V-16, 17, and 18 into Eq. V-3 and canceling the common factor
dx dy, we get
__!__ (aBy _ aBx) = Jz + eo aEz (V-19)
µo ax ay at
We could have proceeded exactly as above if we had started with a rectangle
parallel to the y-z plane or one parallel to the z-x plane. Each rectangle would
have given us a different component of an arbitrarily oriented differential surface
at P. Equation V-19 is obviously the z-component equation corresponding to
Eq. V-3. If we multiply it by k and add to it the two similar vector equations,
which may be obtained by cyclically permuting x, y, z and i, j, k, corresponding
to the x-component and y-component equations, we obtain
eo div E = p, (V-22)
div B = 0, (V-23)
curl B = µo(J + eo aE/at), (V-24)
which are four coupled partial differential equations. They apply at each point
of space in an electromagnetic field.
We will now derive the wave equation for electromagnetic waves in free space.
In free space, the charge density p and the current density J are zero, so that the
Maxwell equations there become
div E = 0,
div B = 0,
curl E = - aB/at,
and curl B = µoeo aE/at.
aB a
curl curl E = - curl - = - - curl B.
at at
SUPPLEMENTARY TOPIC V 7
But, from above, curl B = µ 0t:o( aE/ at), so that
v' 2E = µot:o - ·
a2 E (V-28a)
at 2
The student, proceeding as above, should be able to show that B satisfies the same
equation, or
(V-28b)
Equations V-28 are the equations of electromagnetic wave motion. Being vector
equations, they are equivalent to six scalar equations, one for each component of
E and of B.
There are many solutions of Eqs. V-28, corresponding to different kinds of
electromagnetic waves~plane, spherical, and cylindrical waves being three
examples. Let us consider a solution in which two components of E and two of
B vanish, that is, in which
E,, = Ez = 0 and B,, = By = 0.
Equations V-28 are satsified for these assumptions. For the nonvanishing com-
ponents, EY and Bz, Eqs. V-28 reduce to (see Eq. V-11)
a2Ey a2Ey + a2Ey a2Ey
ax2 + ay2 az2 = µot:o ~ (V-29a)
and
a2Bz a2Bz a2Bz a2Bz
ax2 + -ay2 + az2 = µoEo at2 (V-29b)
(V-29c)
c = w/k = 1/Veoµo,
which (see Eq. 39-15) gives the speed of electromagnetic waves in free space.
Supplementary Problems
Chapter 26
I. Two free point charges +q and +4q are a distance l apart. A third charge is so
placed that the entire system is in equilibrium. Find the location, magnitude, and
sign of the third charge.
2. If the balls of Fig. 26~7 are conducting, what happens to them after one is dis-
charged? Find the new equilibrium separation.
3. Two identical conducting spheres, having charges of opposite sign, attract each
other with a force of 0.108 nt when separated by 0.5 meter. The spheres are connected
by a conducting wire, which is then removed, and thereafter repel each other with a
force of 0.036 nt. What were the initial charges on the spheres?
4. A particle of charge -q and mass m moves in a circular orbit about a fixed charge
+Q. (a) Show that the "distance cubed ex: period squared" law,
ra = Qq 1'2
1671' 3Eom
is satisfied. Note that the proportionality constant depends on the property (q/m) of
the orbiting particle. (b) What is the corresponding situation when the force is gravi-
tational rather than electrical?
5. An electron is projected with an initial speed of 3.24 X 10 5 meter /sec directly
toward a proton which is essentially at rest. If the electron is initially a very great
distance from the proton, at what distance from the proton is its speed instantaneously
equal to twice its initial value? (Hint: Use the work-energy theorem.)
6. A "dipole" is formed from a rod of length 2a and two charges, +q and -q. Two
such dipoles are oriented as shown in Fig. 26-9, their centers being separated by the
-q
r-2a~e
e +q -q
r-2a~e
e +q
J_
( _- R ------------3"-J
Fig. 26-9
9
10 SUPPLEMENTARY PROBLEMS
distance R. (a) Calculate the force exerted on the left dipole. (b) For R » a, show
that the magnitude of the force exerted on the left dipole is approximately given by
3p2
F=--
21rEoR4
Chapter 27
1. A uniform vertical field E is established in the space bet,rnen two large parallel
plates. In this field one suspends a small conducting sphere of mass m from a string of
length l. Find the period of this pendulum when the sphere is given a charge +q if
the lower plate is charged positively; is charged negatively.
2. A charge q = 3.0 X 10- 6 coul is 30 cm from a small dipole along its perpendicular
bisector. The magnitude of the force on the charge is 5.0 X 10- 6 nt. Show on a
diagram (a) the direction of the force on the charge, (b) the direction of the force on the
dipole, and (c) determine the magnitude of the force on the dipole.
3. A thin glass rod is bent into a semicircle of radius R. A charge +Q is uniformly
distributed along the upper half and a charge - Q is uniformly distributed along the
lower half, as shown in Fig. 27-25. Find the electric field E at P, the center of the
semicircle.
+ + + + + + + + + +
4. A "semi-infinite" insulating rod (Fig. 27-26) carries a constant charge per unit
length of A, Show that the electric field at the point P makes an angle of 45° with the
rod. This result is independent of the distance R.
5. Find the frequency of oscillation of an electric dipole, of moment p and rotational
inertia I, for small amplitudes of oscillation about its equilibrium position in a uniform
electric field of strength E.
6. An electric dipole of moment p is placed parallel to an electric field line along the
y-axis in a nonuniform electric field (Fig. 27-27). The magnitude of the field E varies
uniformly along the y-direction as shown. (a) Show that the magnitude of the force
on the dipole is p dE/dy. (b) What is the direction of the force?
-qe e +q
R
p
7. One type of "electric quadrupole" is formed by four charges located at the vertices
of a square of side 2a. Point P lies a distance R from the center of the quadrupole on a
line parallel to two of sides of the square as shown in Fig. 27-28. For R » a, show that
SUPPLEMENTARY PROBLEMS 11
the electric field at P is approximately given by
3(2qa 2)
E=--·
41rEoR 4
Chapter 28
I. It is found experimentally that the electric field in a large region of the earth's
atmosphere is directed vertically down. At an altitude of 300 meters the field is 60
nt / coul and at an altitude of 200 meters it is 100 nt / coul. Find the net
amount of charge contained in a cube 100 meters on edge located between 200 and
300 meters altitude. Neglect the curvature of the earth.
2, Suppose that an electric field in some region is found to have a constant direction
but to be decreasing in strength in that direction. What do you conclude about the
charge in the region?
3. Two concentric conducting spherical shells have radii R1 = 0.145 meter and
R2 = 0.207 meter. The inner sphere bears a charge -6.00 X 10- 2 coul. An electron
escapes from the inner sphere with negligible speed. Assuming that the region between
the spheres is a vacuum, compute the speed with which the electron strikes the outer
sphere.
4. (a) Two identical nonconducting spheres have radius r and are fixed with their
centers a distance R > 2r apart. If each sphere has a total charge q uniformly dis-
tributed on its surface, what is the magnitude of the electric force that either sphere
exerts on the other? (b) Suppose instead that the spheres are conductors, the same
total charge on each still being q. Will the electric force that either sphere exerts
on the other in this case be greater than, less than, or equal to the force in case (a)?
Explain.
5. The spherical region a < r < b carries a charge per unit volume of p = A/r,
where A is constant. At the center (r = 0) of the enclosed cavity is a point charge Q.
What should the value of A be so that the electric field in the
region a < r < b has constant magnitude?
6. A solid insulating sphere carries a uniform charge per unit vol-
ume of p. Let r be the vector from the center of the sphere to a
general point P within the sphere. (a) Show that the electric field
at P is given by E = pr /3eo. (b) A spherical "cavity" is removed
from the above sphere, as shown in Fig. 28-25. Using superposition
concepts show that the electric field at '111 points within the cavity Fig. 28-25
is E = pa/3e 0 (uniform field), where a is the vector connecting the
center of the sphere with the center of the cavity. Note that both these results are
independent of the radii of the sphere and the cavity.
Chapter 29
1. What is the charge density on the surface of a conducting sphere of radius 0.15
meter whose potential is 200 volts?
2. Two identical conducting spheres of radius r = 0.15 meter are separated by a
distance a = 10.0 meters. What is the charge on each sphere if the potential of one is
+1500 volts and if the other is -1500 volts?
3. Two conducting spheres, one of radius 6.0 cm and the other of radius 12.0 cm,
each have a charge of 3 X 10- 8 coul and are very far apart. If the spheres are con-
nected by a conducting wire, find (a) the direction of motion and the magnitude of the
charge transferred and (b) the final charge on and potential of each sphere.
12 SUPPLEMENTARY PROBLEMS
4. In the rectangle shown in Fig. 29-29, the sides have lengths 5.0 cm and 15.0 cm,
q1
q1 r
= -5.0 X 10- 6 coul and q2 = +2.0 X 10- 6 coul. (a) What is the electric potential
l
A at corner B? At corner A? (b) How much work is involved
in moving a third charge q3 = +3.0 X 10- 6 coul from B to A
B ~ q along a diagonal of the rectangle? (c) In this process, is
2 external work converted into electrostatic potential energy or
Fig. 29-29 vice versa? Explain.
5. Show that the potential energy of an assembly of four
charges, each of charge q, in the configuration of a triangular pyramid, ao on an edge,
is 6 X (q 2 /ao)/4?reo.
6. Three charges of +o. l coul each are placed on the corner of an equilateral tri-
angle, 1.0 meter on a side. If energy is supplied at the rate of 1.0 kw, how many days
would be required to move one of the charges onto the midpoint of the line joining the
other two?
i. Show, using the fact that an electrostatic field is a conservative field, that one
cannot create an electric field in which all the lines of force are straight parallel lines
whose density (number per unit cross-sectional area) changes in a direction at right
angles to the lines of force.
8. Two line charges are parallel to the z-axis; one, of charge per unit length +>.., is a
distance a to the right of this axis, and the other, of charge per unit length - >.., is a
distance a to the left of this axis (the lines and z-axis being in the same plane). 8ketch
some of the surfaces of equipotential.
9. A particle of (positive) charge Q is assumed to have a fixed position at P. A
second particle of mass m and (negative) charge -q moves at constant speed in a
circle of radius r1, centered at P. Derive an expression for the work W that must be
done by an external agent on the second particle in order to increase the radius of the
circle of motion, centered at P, to r 2 . Express W in terms of quantities chosen from
among m, ri, r2, q, Q, and eo only.
10. A particle of charge Q is kept in a fixed position at a point Panda second particle
of mass m, having the same charge Q, is initially held at rest a distance r1 from P. The
second particle is then released and is repelled from the first one. Determine its
velocity at the instant it is a distance r2 from P. Let Q = 3.1 X 10- 6 coul, m =
2.0 X 10- 5 kg, ri = 9.0 X 10- 4 meter, and r2 = 25 X 10- 4 meter.
11. A particle of mass m, charge q > 0, and initial kinetic energy K is projected
(from "infinity") toward a heavy nucleus of charge Q, assumed to have a fixed position
V (volts)
a h
-5 x (meters)
-12
Fig. 29-30
SUPPLEMENTARY PROBLEMS 13
in our reference frame. (a) If the "aim is perfect," how close to the center of the
nucleus is the particle when it comes instantaneously to rest? (b) With a particular
imperfect aim the particle's closest approach to the nucleus is twice the distance deter-
mined in part (a). Determine the speed of the particle at this closest distance of
approach.
12. The electric potential varies along the x-axis as shown in the graph of Fig. 29-30.
For each of the intervals shown (ignore the behavior at the end points of the intervals)
determine the x-component of the electric field and plot Ex vs. x.
13. A charge per unit length 'A is distributed uniformly along a straight-line segment
of length L. (a) Determine the electrostatic potential (chosen to be zero at infinity) at
a point P a distance y from one end of the charged segment and in line with it (see Fig.
29-31). (b) Use the result of (a) to compute the component of the electric field inten-
sity at P in the y-direction (along the line). (c) Determine the component of the elec-
tric field intensity at P in a direction perpendicular to the straight line.
0 L x
14. On a thin rod of length L lying along the x-axis with one end at the origin (x = 0),
as in Fig. 29-32, there is distributed a charge per unit length given by 'A = kx, where k is
a constant. (a) Taking the _electrostatic potential at infinity to be zero, find Vat the
point Pon the y-axis. (b) Determine the vertical component, Ey, of the electric field
intensity at P from the result of part (a) and also by direct calculation. (c) Why
cannot Ex, the horizontal component of the electric field at P, be found using the result
of part (a)?
Chapter 30
I. Show that the capacitance of two oppositely charged metal spheres of the same
radius, when far apart, is one-half the capacitance of one isolated sphere.
2. Two metallic spheres, radii a and b, are connected by a thin wire. Their separa-
tion is large compared with their dimensions. A charge Q is put onto this system.
(a) How much charge resides on each sphere? (b) Apply the definition of capacitance
to show that the capacitance of this system is C = 41rto (a + b).
3. N identical spherical drops of liquid are charged to the same potential V. One
large drop is formed by combining these. Show that the potential of the large drop is
N,iV.
4. Charges q1, q2, qs are placed on capacitors of capacitance C1, C2, Cs respectively,
arranged in series as shown in Fig. 30-28. Switch S is then closed. What are the final
charges qi', q2', qs' on the capacitors?
5. When switch S is thrown to the left in Fig. 30-29, the plates of the capacitor of
capacitance C1 acquire a potential difference V 0. C 2 and Cs are initially uncharged.
14 SUPPLEMENTARY PROBLEMS
The switch is now thrown to the right. What are the final charges q 1, q 2 , q 3 on the
corresponding capacitors?
6. A geiger tube is made of two long concentric metal cylinders with a gas of dielectric
constant K between them. Neglecting end effects, use Gauss' law to calculate the
capacitance of this configuration. The center rod has a radius a, the surrounding tube
a radius b, and the length l » b.
7. A parallel plate capacitor has plates of area 0.12 m 2 and a separation of 1.2 cm.
A battery charges the plates to a potential difference of 120 volts and is then discon-
nected. A dielectric slab of thickness 0.4 cm and dielectric constant 4.8 is then placed
symmetrically between the plates. In terms of to (a) find the capacitance before the slab
is inserted. (b) what is the capacitance with the slab in place? (c) what is the free charge
q before and after the slab is inserted? (d) determine the electric field strength in the
space between the plates and dielectric. (e) what is the electric field strength in the
dielectric? (J) with the slab in place what is the potential difference across the plates?
(g) how much external work is involved in the process of inserting the slab?
Chapter 31
I. A steady beam of alpha particles (q = 2e) traveling with constant kinetic energy
20 Me V carries a current 0.25 X 10- 6 ampere. (a) If the beam is directed perpendicu-
lar to a plane surface, how many alpha particles strike the surface in 3.0 sec? (b) At
any instant, how many alpha particles are there in a given 20-cm length of the beam?
(c) Through what potential difference was it
necessary to accelerate each alpha particle
from rest to bring it to an energy of 20 Me V?
2. List similarities and differences between
the flow of charge and the flow of a fluid;
between the flow of charge and the conduction
of heat.
3. Explain -why the momentum which con-
duction electrons transfer to the ions in a
./
i
metal conductor does not give rise to a result-
ant force on the conductor.
Fig. 31-10 4. A resistor is in the shape of a truncated
right circular cone (Fig. 31-10). The end
radii are a and b, the altitude isl. If the taper is small, we may assume that the current
density is uniform across any cross section. (a) Calculate the resistance of this object.
(b) Show that your answer reduces to p(l/A) for the special case of zero taper (a= b).
5. Conductors A and B, having equal lengths of 40 meters and a cross-sectional area
of 0.10 m 2 , are connected in series. A potential of 60 volts is applied across the terminal
points of the connected wires. The resistances of the wires are 40 and 20 ohms respec-
tively. Determine: (a) the resistivities of the two wires: (b) the magnitude of the elec-
SUPPLEMENTARY PROBLEMS 15
tric field in each wire; (c) the current density in each wire; (d) the potential difference
applied to each conductor.
6. A 1250-watt radiant heater is constructed to operate at 115 volts. (a) What will
be the current in the heater? (b) What is the resistance of the heating coil? (c) How
many kilocalories are created in our hour by the heater?
7. An iron wire (diameter 1 mm, length 10 cm) is placed in an evacuated chamber.
Estimate the equilibrium temperature of the wire if it carries a current of 10 amp.
Assume that all heat transfer is by radiation and that the surface of the wire radiates
according to Eq. 47-2. Take the temperature of the chamber walls to be 27° C. List
any additional assumptions that you use.
Chapter 32
1. Two batteries having the same emf e but different internal resistances ri and r 2
are connected in series to an external resistance R. Find the value of R that makes the
potential difference zero between the terminals of the first battery.
2. The section of circuit AB (see Fig. 32-26) absorbs power P = 50.0 watts and a
current i = 1.0 amp passes through it in the indicated direction. (a) What is the
potential difference between A and B? (b) If the element C does not have internal
resistance, what is its emf? (c) What is its polarity?
c
(&', r)
Fig. 32-26
3. Two light bulbs, one of resistance R and the other of resistance r( < R), are con-
nected (a) in parallel and (b) in series. Which bulb is brighter?
4. N identical batteries of emf e and internal resistance r may be connected all in
series or all in parallel. Show that each arrangement will give the same current in an
external resistor R if R = r.
5. Twelve resistors, each of resistance R ohms, form a cube (see Fig. 32-27). (a)
Find RAs, the resistance of an edge. (b) Find Rec, the equivalent resistance of a face.
(c) Find RAc, the equivalent resistance of the body diagonal.
6. What internal connections (resistors only) can account for the fact that there are
2 ohms resistance between any pair of the terminals on a box with N terminals? (See
Fig. 32-28.)
16 SUPPLEMENTARY PROBLEMS
R2
(a) (b)
Chapter 33
U
I. Particles 1, 2, and 3 follow the paths shown in Fig. 33-25 as they pass through the
x x: x
x x
x x x magnetic field there. What can one conclude about
each particle?
2. (a) What speed would a proton need to circle
x x x x the earth at the equator, if the earth's magnetic
1 field is everywhere horizontal there and directed
x x x x x X along longitudinal lines. Take the magnitude of the
x x x earth's magnetic field to be 0.41 X 10- weber /meter
4 2
x x x
at the equator. (b) Draw the velocity and magnetic
x x induction vectors corresponding to this situation.
x 3. An electron is accelerated through a potential
difference of 1000 volts and directed into a region
x x between two parallel plates separated by 0.02 meter
2
with a voltage difference of 100 volts between them.
Fig. 33-25 If the electron enters moving perpendicular to the
electric field between the plates, what magnetic field
is necessary perpendicular to both the electron path and the electric field so that the
electron travels in a straight line?
SUPPLEMENTARY PROBLEMS 11
4. A positive point charge Q travels in a straight line with constant speed through an
evacuated region in which there is a uniform electric field E and a uniform magnetic
field B. (a) If Eis directed vertically up and the charge travels horizontally from north
to south with speed v, determine the least value of the magnitude of B and the cor-
responding direction of B. (b) Explain why Bis not uniquely determined when E and
v alone are given. (c) Suppose the charge is a proton which enters the region after
having been accelerated through a potential difference of 3.10 X 10 5 volts. If E =
1.90 X 10 5 volt/meter, compute the value of Bin part (a). (d) If in part (c) the elec-
tric field E is turned off, determine the radius r of the circle in which the proton now
moves.
5. A certain galvanometer has a resistance of 75.3 ohms; its needle experiences a full-
scale deflection when a current 1.62 X 10- 3 amp passes through its coil. (a) Determine
the value of the auxiliary resistance required to convert the galvanometer into a volt-
meter that reads 1.000 volt at full-scale deflection. How is it to be connected? (b)
Determine the value of the auxiliary resistance required to convert the galvanometer into
an ammeter that reads 0.0500 amp at full-scale deflection. How is it to be connected?
6. In a Hall effect experiment a current of 3.0 amp lengthwise in a conductor 1.0 cm
wide, 4.0 cm long, and 10- 3 cm thick produced a transverse Hall voltage (across the
width) of 1.0 X 10- 5 volt when a magnetic field of 1.5 weber /meter 2 passed perpendicu-
larly through the thin conductor. From these data, find (a) the drift velocity of the
charge carriers and (b) the number of carriers per cubic centimeter. (c) Show on a
diagram the polarity of the Hall voltage with a given current and magnetic field direc-
tion, assuming the charge carriers are (negative) electrons.
7. (a) What is the cyclotron frequency of an electron with an energy of 100 ev in the
earth's magnetic field of 1.0 X 10- 4 webers/meters 2 ? (b) What is the radius of curva-
ture of the path of this electron if its velocity is perpendicular to the magnetic field?
Chapter 34
I. Eight wires cut the page perpendicularly at the points shown in Fig. 34-32. A
wire labeled with the integer k (k = 1, 2, . . . 8) bears the current kio. For those with
odd k, the current flows up out of the page; for
those with even k it flows down into the page.
Evaluate §B · di along the closed path shown in
the direction indicated by the single arrowhead.
2. Two long straight wires pass near one another
at right angles. If the wires are free to move,
describe what happens when currents are sent
through them.
3. Suppose, in Fig. 34-25, that the currents are all
in the same direction. What is the force per meter
(magnitude and direction) on any one wire? In the
analogous case of parallel motion of charged parti- Fig. 34-32
cles in a plasma this is known as the pinch effect.
4. A long straight conductor has a circular cross section of radius R and carries a
current i. Inside the conductor there is a cylindrical hole
of radius a whose axis is parallel to the conductor axis at
a distance b from it. Use superposition ideas, and obtain
an expression for the magnetic induction B inside the hole.
5. A straight conductor is split into identical semicircular
Fig. 34-33 turns as shown in Fig. 34-33. What is the magnetic field
at the center C of the circular loop so-formed?
6. (a) A current i flows in a straight wire of length L in the direction shown in Fig.
34-34a. Starting from the Biot-Savart law, determine the resulting magnetic induc-
18 SUPPLEMENTARY PROBLEMS
tion (Bp, BQ, BR, Bs, respectively-direction and magnitude in each case) at each of the
four points P, Q, R, S (all coplanar with the wire). (b) Using the results of part (a),
compute the magnetic induction B (magnitude and direction) resulting at the point T
?S
L
t I
I
I
I
~a-.-ec---a~
...------.---------<;>
I I
j
I
: R I
--------0 I
1
I
-<---L ~ I
I
I
L 2a
j
L
-L~
:
I
I
--------0
Q j
I
t 6p I ~---2a---~
(a) (b)
Fig. 34-34
from the current flowing as indicated in the six-sided rectilinear closed loop shown in
Fig. 34-34b. (Everything drawn is meant to lie in the same plane and all angles are
90°.)
7. Use the Biot-Savort law to calculate the magnetic induction B at C, the common
center of the semicircular arcs AD and HJ, of radii R2 and R1 respectively, forming
o,:R,
part of the circuit A DJ HA carrying current i, as shown in Fig. 34-35.
I ' I
I ', () I
I ',
A H C J D R1 C
8. Compute the magnetic induction B at C, the common center of the circular arcs
of the circuit carrying current i, the arcs cutting a sector of angle 8, as shown in Fig.
34-36.
9. (a) A long wire is bent into the shape shown in Fig. 34-37, without cross-contact
[}
at P. Determine the magnitude and direction
of B at the center C of the circular portion when
the current i flows as indicated. (b) The circu-
)> )>
i lar part of the wire is rotated without distortion
p about its (dashed) diameter perpendicular to
Fig. 34-37
the straight portion of the wire. The magnetic
moment associated with the circular loop is
now in the direction of the current in the straight part of the wire. Determine Bat
C in this case.
Chapter 35
1. Show that emf has the same dimensions as time rate of change of magnetic flux.
2. In Fig. 35-39, l = 2.0 meters and v = 50 cm/sec. Bis the earth's magnetic field,
directed perpendicularly out of the page and having a magnitude 6.0 X 10- 5 webers/m 2
at that place. The resistance of the circuit ADCB, assumed constant (explain how
this may be achieved approximately), is R = 1.2 X 10-5 ohms. (a) What is the emf
SUPPLEMENTARY PROBLEMS 19
induced in the circuit? (b) What is the electric field strength in the wire AB? (c)
What force does each electron in the wire experience due to the motion of the wire in
the magnetic field? (d) What is the magnitude and direction of the current in the wire?
(e) What force must an external agency exert in order to keep the wire moving with this
constant velocity? (!) Compute the rate at which the external agency is doing work.
(g) Compute the rate at which electrical energy is being converted into heat energy,
• A • • • • • •
t'x
1--. •
v
• • • •
D
•
l
l__ .
,·
• • • • • •
•B • • • c.
B
3. Figure 35-40 shows two loops of wire having the same axis. The smaller loop is
above the larger one, a distance x which is large compared to the radius R of the larger
loop. Hence, with current i flowing as indicated in the larger loop, the consequent
magnetic induction is nearly constant throughout the plane area 1rr 2 bounded by the
smaller loop. Suppose now that xis not constant but is changing at the constant rate
dx/dt = v. (a) Determine the magnetic flux across the area bounded by the smaller
loop as a function of x. (b) Compute the emf generated in the smaller loop at the
instant when x = NR. (c) Determine the direction of the induced current flowing in
the smaller loop if v > 0.
4. A closed loop of wire consists of a pair of equal semicircles, radius 3. 70 cm, lying
in mutually perpendicular planes. The loop was formed by folding a circular loop
along a diameter until the two halves became perpendicular. A uniform magnetic
field B of magnitude 760 gauss is directed perpendicular to the fold diameter and makes
equal angles (45°) with the planes of the semicircles as shown in Fig. 35-4la. (a) The
magnetic field is reduced at a uniform rate to zero during a time interval 4.50 X 10- 3
sec. Determine the magnitude of the induced emf and the sense of the induced current
in the loop during this interval. (b) How-would the answers change if B is directed as
shown in Fig. 35-41b, perpendicular to the direction first given for it but still per-
pendicular to the "fold-diameter?"
v
B
x
x
(a) (b)
transferred through the resistor between t = t 1 and t = t2. Show that your answer is
proportional to the difference <l>B(t2) - <l>B(ti), and is otherwise independent of the
manner in which Bis changing. (b) Suppose the change in flux, <I>s(t 2) - <l>B(t 1) is zero.
Does it then follow that no joule-heating occurred dur-
ing this time interval?
.,.-----,:~ . 6. An electromagnetic "eddy current" brake consists
of a disk of conductivity rr and thickness t rotating
about an axis through its center with a magnetic field
B applied perpendicular to the plane of the disk over a
Fig. 35-43 small area a 2 (see Fig. 35-43). If the area a 2 is at a
distance r from the axis, find an approximate expression
for the torque tending to slow down the disk at the instant its angular velocity equals w.
Chapter 36
1. Two equal solenoids (inductance L) are wired in series. (a) If they are far apart,
show that the equivalent inductance is 2L. (b) If they are "close wound," show that
the equivalent inductance is either zero or 4L depending on the direction of the windings.
(Hint: They form a single solenoid.)
2. In the circuit shown in Fig. 36-8, 8 = 10 volts, R1 = 5.0 ohms, R2 = 10 ohms,
and L = 5.0 henries. For the two separate conditions (I) switch S just closed and
(II) switch S closed for a very long time, calculate (a) the current i1 through R1,
(b) the current L2 through R2, (c) the current i through the switch, (d) the voltage
across R2, (e) the voltage across L, and (J) di2/dt.
_(,,t R1 i2 t L
L
R2
~
377
R = - In
21r
(b)
-
a
ohms.
(The quantity 377 ohms is sometimes called the impedance of free space.)
5. Each item (a) coulomb-ohm-meter/weber, (b) volt-second, (c) coulomb-ampere/
farad, (d) kilogram-volt-meter 2 / (henry-ampere) 2 , (e) (henry /farad/·~ is equal to one
of the items in the following list: meter, second, kilogram, dimensionless number,
newton, joule, volt, ohm, watt, coulomb, ampere, weber, henry, farad. Give the
equalities.
SUPPLEMENTARY PROBLEMS 21
Chapter 37
I. A total charge q is distributed uniformly on a dielectric ring of radius r. If the
ring is rotated about an axis perpendicular to its plane and through its center at an
angular speed w, find the magnitude and direction of its resulting magnetic moment.
2. (a) What is the magnetic moment due to the orbital motion of an electron in an
atom when the orbital angular momentum is one quantum unit ( = h = 1.05 X 10- 34
joule-sec). (b) The intrinsic spin magnetic moment of an electron is 0.928 X 10- 23
amp-meter 2 . What is the difference in the magnetic potential energy U between the
states in which the magnetic moment is aligned with and aligned in the opposite direc-
tion to an external magnetic field of 1.2 weber /meter 2 ? (c) What absolute temperature
would be required so that the energy difference in (b) would equal the mean thermal
energy k'l'/2?
3. An electron travels in a circular orbit about a fixed positive point charge in the
presence of a uniform magnetic field B directed normal to the plane of ~0.
its motion. The electric force has precisely N times the magnitude of
the magnetic force on the electron. (a) Determine the two possible
angular speeds of the electron's motion. (b) Evaluate these speeds
numerically if B = 4.27 X 10 3 gauss and N = 100.
4. A simple bar magnet is suspended by a string as shown in Fig. S
37-23. If a uniform magnetic field B directed parallel to the ceiling
is then established, show the resulting orientation of string and magnet.
5. An iron magnet containing iron of relative permeability 5000 N
has a flux path 1.0 meter long in the iron and an air gap 0.01 meter long Fig. 37-23
each with cross-sectional areas of 0.02 meter. 2 What current is neces-
sary in a 500 turn coil wrapped around the iron to give a flux density in the air gap of
1.8 webers/meter 2 ?
Chapter 38
1. Initially the 900 µf capacitor is charged to 100 volts, and the 100 µf capacitor is
uncharged in Fig. 38-13. (a) Describe in detail how one may charge the 100 µf capaci-
Fig. 38-14
Ir
J__ c
Fig. 38-1 S
f
Fig. 39-18
SUPPLEMENTARY PROBLEMS 23
a < r < b. (b) Calculate the Poynting vector S for a < r < b. (c) By suitably
integrating the Poynting vector, show that the total power flowing across the annular
cross section a < r <bis 8 2 /R. Is this reasonable? (d) Show that the direction of S
is always from the battery to the resistor, no matter which way the battery is connected.
7. A long hollow cylinder (radius R, length l) carries a uniform charge per unit area
of u on its surface. An externally applied torque causes the cylinder to rotate at con-
stant acceleration w(t) = ed. (a) Find B within the cylinder (treat it as a solenoid).
(b) Find E at the inner surface of the cylinder. (c) Find S at the inner surface of the
cylinder. (d) Show that the flux of S entering the interior volume of the cylinder is
Chapter 40
I. A plane electromagnetic wave, with wavelength 3.0 meters, travels in free space
in the +x-direction with its electric vector E, of amplitude 300 volts/meter, directed
along the y-axis. (a) What is the frequency v of the wave? (b) What is the direction
and amplitude of the B field associated with the wave? (c) If E = Em sin (kx - wt),
what are the values of k and w for this wave? (d) What is the time averaged rate of
energy flow per unit area associated with this wave? (e) If the wave fell upon a per-
fectly absorbing sheet of area A, what momentum would be delivered to the sheet per
second and what is the radiation pressure exerted on the sheet?
2. Show that eoE X B has the dimensions of momentum/volume. (The Yector
eoE X B may be used to compute the momentum stored in the fields in the same man-
1 1 .
ner that the scalar - e0 E 2 + - B 2 may be used to compute the energy stored m the
2 2µo
fields.)
3. A particle in the solar system is under the combined influence of the sun's gravi-
tational attraction and the radiation force due to the sun's rays. Assume that the
particle is a sphere of density 1.0 gm/cm 3 and that all of the incident light is absorbed.
(a) Show that all particles with radius less than some critical radius, Ro, will be blown
out of the solar system. (b) Calculate Ro. (c) Does Ro depend on the distance from
the earth to the sun? (See the appendices for the necessary constants.)
Chapter 41
I. When an electron moves through a medium at a speed exceeding the speed of
light in that medium, it radiates electromagnetic energy (the Cerenkov effect, see Sec-
tion 20-7). What minimum speed must an electron have in a liquid of refractive index
1.54 in order to radiate?
2. Assume that the index of refraction of the earth's atmosphere varies, with dtitude
only, from the value one at the edge of the atmosphere to some larger value at the
surface of the earth. (a) Neglecting the earth's curvature, show that the apparent
angle of a star from the zenith direction is independent of how the refractive index of
the atmosphere varies with altitude and depends only on the value of n at the earth's
surface. (Hint: Compare a uniform atmosphere with one consisting of layers of
increasing refractive index.) (b) How does the earth's curvature affect the analysis?
3. A point source is 80 cm below the surface of a body of water. Find the diameter
of the largest circle at the surface through which light can emerge from the water.
4. A pole extends 2.0 meters above the bottom of a swimming pool and 0.5 meters
above the water. Sunlight is incident at 45°. What is the length of the shadow of the
pole on the bottom of the pool?
5. A given monochromatic light ray, initially m air, strikes the 90° prism at P (see
Fig. 41-24) and is refracted there and at Q to such an extent that it just grazes the
right-hand prism surface after it emerges into air at Q. (a) Determine the index of
24 SUPPLEMENTARY PROBLEMS
Fig. 41-24
v -
Fig. 41-25
- -
-
refraction, relative to air, of the prism for this wavelength in terms of the angle of
incidence 61 which gives rise to this situation. (b) Give a numerical upper bound for
the index of refraction of the prism. (c) Show, by a ray diagram, what happens if the
angle of incidence at P is slightly greater than 6 1, is slightly less than 61.
6. Two perpendicular mirrors form the sides of a vessel filled with water, as shown in
Fig. 41-25. A light ray is incident from above, normal to the water surface. (a) Show
that the emerging ray is parallel to the incident ray. Assume that there are two reflec-
tions at the mirror surfaces. (b) Repeat the analysis for the case of oblique incidence,
the ray lying in the plane of the figure. (c) Using three mirrors, state and prove the
three-dimensional analog to this problem.
Chapter 42
I. How many images of himself can an observer see in a room whose ceiling and two
adjacent walls are mirrors? Explain.
2. A thin flat plate of glass is a distance b from a convex mirror. A point source of
light Sis placed a distance a in front of the plate (see Fig. 42-33) so that its image in the
partially reflecting plate coincides with its image in the
.L. J\ I
directly in front of the eye to examine an object. Draw a
ray diagram showing the final image on the retina of the eye.
(b) The smallest distance at which a normal eye can focus
Fig. 42-33 upon an object is 25 cm. Use this result with your ray dia-
gram to prove that the (angular) magnification of a simple
magnifier is 25 If+ 1, ,vhere f is the focal length of the magnifier lens.
4. Two thin lens, one having f = + 12.0 cm and the other f = -10.0 cm, are sepa-
rated by 7.0 cm. A small object is placed 43.5 cm from the center of the lens system on
the principal axis first on one side and next on the other side. Find the location of the
final image in each case.
5. Show that a thin converging lens of focal length f followed by a thin diverging lens
of focal length -f will bring parallel light to a focus beyond the second lens provided
that the separation of the lenses L satisfies O < L < f. (b) Does
this property change if the lenses are interchanged? (c) What
happens when L = O?
6. A concave and a convex lens are cut out of a plane parallel
block of glass as shown in Fig. 42-34. Discuss the geometry of a
beam of parallel rays incident (a) on the concave lens and (b) on
the convex lens as the distance between lenses is increased from
contact- to a large separation.
7. An object is placed 1.0 meter in front of a convergent lens, of
focal length 0.5 meter, which is 2.0 meters in front of a plane
mirror. (a) Where is the final i~age, measured from the lens, Fig. 42-34
SUPPLEMENTARY PROBLEMS 25
that would be seen by an eye looking toward the mirror through the lens? (b) Is the
final image real or virtual? (c) Is the final image erect or inverted? (d) What is the
lateral magnification?
Chapter 43
1. In a double-slit arrangement the distance between slits is 5.0 mm and the slits are
1.0 meter from the screen. Two interference patterns can be seen on the screen, one
due to light of 4800 A and the other 6000 A. What is the separation on the screen
bet,veen the third-order interference fringes of the two different patterns?
2. In Young's interference experiment in a large ripple tank (see Fig. 43-4) the
coherent vibrating sources are placed 12.0 cm apart. The distance between maxima
2.0 meters away is 18.0 cm. If the speed of ripples is 25.0 cm/sec, find the frequency
of the vibrators.
3. One slit of a double-slit arrangement is covered by a thin glass plate of refractive
index 1.4, and the other by a thin glass plate of refractive index 1. 7. The point on the
screen where the central maximum fell before the glass plates were inserted is now occu-
pied by what had been the fifth bright fringe before. Assume :\ = 4800 A and that the
plates have the same thickness t and find the value of t.
4. In Fig. 43-25, the source emits monochromatic light of wavelength:\. Sis a very
narrow slit in an otherwise opaque screen I. A plane mirror, whose surface includes
the axis of the lens shown, is located a distance h below S. Screen II is at the focal
plane of the lens. (a) Find the condition for maxima and minima brightness of fringes
Region A
*
Source
Region B
I II
Fig. 43-25
on screen II in terms of the usual angle 0, the wavelength:\, and the distance h. (b) Do
fringes appear only in region A (above the axis of the lens), only in region B (below the
axis of the lens), or in both regions A and B? Explain. (Hint: Consider the image of
S formed by the mirror.)
5. Let fi(t) = Ai sin (wt +
<J,1) and f2(t) = A2 sin (wt +
<J,2). Suppose we want to
calculate the time average of their product:
-l
1'
lT Ji
t=O
(t)h (t) dt, 1'
271"
w
(Such a problem might arise if Ji represents a current and ha voltage; fih would then
be the instantaneous power.) Show that this average is equal to one-half the dot
product of the corresponding phasors.
6. Consider the problem of determining the sum
(a) Show that the sum may always be written in the form
Water
Fig. 43-26
film where one observes the third blue region from the outside of the drop? (c) Why do
the colors gradually disappear as the oil thickness becomes larger?
Chapter 44
I. It can be shown that, except for (} = 0, a circular obstacle produces the same
diffraction pattern as a circular hole of the same diameter. Furthermore, if there are
many such obstacles located randomly, then the interference effects vanish leaving only
the diffraction associated with a single obstacle. (a) Explain why one sees a "ring"
around the moon on a foggy day. (b) Calculate the size of the water droplet in the air
if the bright ring around the moon appears to have a diameter 1.5 times that of the
moon.
2. A monochromatic beam of parallel light is incident on a "collimating" hole of
diameter x » X. Point P lies in the geometrical shadow region on a distance screen, as
shown in Fig. 44-15a. Two obstacles, shown in Fig. 44-15b, are placed in turn over the
collimating hole. A is an opaque circle with a hole in it and B is the "photographic
negative" of A. Using superposition concepts, show that the intensity at Pis identical
for each of the two diffracting objects A and B (Babinet's principle).
3. An astronaut in a satellite claims he can just barely resolve two point sources on
the earth, 100 miles below him. What is their separation, assuming ideal conditions?
Take X = 5500 A, and the pupil diameter to be 5.0 mm.
4. Under ideal conditions, estimate the linear separation of two objects on the planet
Mars which can just be resolved by an observer on earth (a) using the naked eye,
SUPPLEMENTARY PROBLEMS 27
Screen
,+.
I
-x
-t
-
A B
(a) (b)
Fig. 44-15
(b) using the 200-in. Mt. Palomar telescope. Use the following data: distance to
Mars = 50 million miles; diameter of pupil = 5.0 mm; wavelength of light = 5500 A.
5. A double-slit system (slit separation d, slit width a) is driven by two loudspeakers
as shown in Fig. 44-16. By use of a variable delay line, the phase of one of the speakers
Variable
Audio delay
source
line
Fig. 44-16
may be varied. Describe in detail what changes occur in the intensity pattern at large
distances as this phase difference is varied from zero to 21r. Take both the interference
and diffraction effects into account.
Chapter 45
I. A diffraction grating has a large number N of slits, each of width d. Let I max
denote the intensity at some principal maximum, and let h denote the intensity of the
kth adjacent secondary maxima. (a) If k « N, show from the phasor diagram that,
approximately, hi I max = ll(k + {-) 21r 2 . (Compare this with the single-slit formula.)
(b) For those secondary maxima which lie roughly midway between two adjacent
principal maxima, show that roughly hi I max = llN 2 • (c) Consider the central
principal maximum and those adjacent secondary maxima for which k « N. Show
that this part of the diffraction pattern quantitatively resembles that of one single slit
of width Nd.
2. With light from a gaseous discharge tube incident normally on a grating with a
distance 1.732 X 10- 4 cm between adjacent slit centers, a green line appears with sharp
maxima at measured transmission angles e = ± 17.6°, 37.3°, -37.1 °, 65.2°, and -65.0°.
(a) Compute the wavelength of the green line that best fits the data.
3. A narrow beam of monochromatic light strikes a grating at normal incidence and
produces sharp maxima at the follmving angles from the normal: 6°40', 13°30', 20°20',
35°40'. No other maxima appear at any angle between 0° and 35°40'. The separa-
tion between adjacent slit centers in the grating is 5.04 X 10- 4 cm. (a) Compute the
wavelength of the light used. (b) Make the most complete quantitative statement that
can be inferred from the above data concerning the width of each slit.
28 SUPPLEMENTARY PROBLEMS
4. Two spectral lines have wavelengths ).. and ).. + Li).., respectively, where Li).. « )...
Show that their angular separation MJ in a grating spectrometer is given approximately
by LiO = Li)../V (d/rn) 2 - ).. 2 , where d is the separation of adjacent slit centers and rn is
the order at which lines are observed. Notice that the angular separation is greater in
the higher orders.
5. A diffraction grating is made up of slits of width 3000 A with a 9000 A separation
between centers. The grating is illuminated by monochromatic plane waves, ).. =
6000 A, the angle of incidence being zero. (a) How many diffracted lines are there?
(b) What is the angular width of the spectral lines observed, if the grating has 1000 slits?
(c) How is the angular width of the spectral lines related to the resolving power of the
grating?
6. A diffraction grating has a resolving power R = )../Li).. = Nm. (a) Show that the
corresponding frequency range, Liv, that can just be resolved is given by Liv = c/Nm)...
(b) From Fig. 45-1, show that the "times of flight" of the two extreme rays differ by an
amount Lit = Nd sin (} /c. (c) Show that (Liv) (Lit) = 1, this relation being independent
of the various grating parameters.
7. Consider an infinite two-dimensional square lattice as in Fig. 45-14b. One inter-
planar spacing is obviously ao itself. (a) Calculate the next five smaller interplanar
spacings by sketching figures similar to Figure 45-16a. (b) Show that your answers
obey the general formula
d = ao/Vh 2 + k2
where h, k are both relatively prime integers (no common integer factor other than
unity).
Chapter 46
1. A beam of light is a mixture of plane polarized light and randomly polarized light.
When it is sent through a Polaroid sheet, it is found that the transmitted intensity can
be varied by a factor of five depending on the orientation of the Polaroid. Find the
relative intensities of these two components of the incident beam.
2. It is desired to rotate the plane of polarization of a beam of plane polarized light
by 90°. (a) How might this be done using only Polaroid sheets? (b) How many sheets
are required in order that the total intensity loss is less than 5 % ? Assume that each
Polaroid sheet is ideal.
3. A sheet of Polaroid and a quarter-wave plate are glued together in such a way
that, if the combination is placed with face A against a shiny coin, the face of the coin
can be seen when illuminated with light of appropriate wavelength. When the com-
bination is placed with face A away from the coin, the coin cannot be seen. Which
component is on face A and what is the relative orientation of the components?
4. A beam of right circularly polarized light is reflected from a mirror. (a) Is the
reflected beam right or left circularly polarized? (b) Has the direction of the associated
linear momentum of the light changed? (c) Has the direction of the associated angular
momentum of light changed? (d) Describe the reaction "felt" by the mirror.
Definition of Standards and
Fundamental and Derived
Physical Constants*
APPENDIX A
29
30 APPENDIX A
DEFINITION OF STANDARDS
Abbre-
Standard viation Equivalent
* Same units and power often as the computational value. The numbers in parentheses are the standard
deviation uncertainties in the last digits of the quoted value.
Miscellaneous Terrestrial Data
APPENDIX B
APPENDIX C
Planet Mercury !".l Venus \;1 Earth EB, e, o Mars d" Jupier '21 Saturn 1;, Uranus ©, 1t£ Neptune tJJ Pluto E
----
Mean diameter
km 5,000 12,400 12,742 6,870 139,760 115,100 51,000 50,000 12,700?
Earth diameters 0.39 0.973 1.000 0.532 10.97 9.03 4.00 3.90 0.46
(.,)
.c,. Volume (earth
volumes) 0.06 0.92 1.00 0.15 1,318 736 64 39 0.10
Mass (earth
0.04 0.82 1.00 0.11 318.3 95.3 14.7 17.3
masses)
Density (earth
---~-
I I;!
densities) 0.69 0.89 1.00 0.70 0.24 0.13 0.23 0.29
Mean density
gm/cm 3 3.8 4.86 5.52 3.96 1.33 0.71 1.26 1. 6 ?
Surface gravity
(earth's) 0.27 0.86 1.00 0.37 2.64 1.17 0.92 1.44 ?
Velocity of
escape, km/sec 3.6 10.2 11.2 5.0 60 36 21 23 11?
Length of day
(earth days) 58.6d 30d? ld 1d37m23• 9h55m 10h38m 10.7h 15.8h ?
Planet Mercury 1;! Venus 9 Earth, EB, e, o Mars d" Jupiter '2l Saturn I? Uranus e , )!{ Neptune IV Pluto E
Period, days
87.97 224.70 365.26 686.98 4,332.59 10,759.20 30,685.93 60,187 .64 90,885
Inclination of
equator to orbit - oo? 23°27' 25°12' 3°7' 26°45' 98.0° 29° ?
Oblateness 0.00 0.00 >'l
1/296 1/192 1/15.4 1/9.5 1/14 1/45 ? :r:t'1
Atmosphere,
[f1
main con-
stituents none N2, C02, A N2, 02 N2, C02, H20 CH4, NHa CH4, NHa CH4, NHa CH4, NHa none ~ i,:j
Maximum sur- [f1
~
face tempera- [f1
ture, °K 700 700 350 320 153 138 110? 90? 80? ~
~
Distance from
Sun, 10 6 km 58 108 149 228 778 1426 2869 4495 5900
The Sun 0 329,390 earth masses, mean density 1.42, mean diameter 1,390,600 km, surface gravity 28 (earth's).
The Moon :] 0.01228 earth masses, mean density 3.36, mean diameter 3,476 km, surface gravity 0.17 (earth's), distance from earth
38 X 10 4 km.
a Adapted from Payne-Gaposchkin and Handbook of Chemistry and Physics.
w
U'I
Periodic Table of the Elements
APPE NDIX D
to have a mass of (exactly) 12 amu. For
Atomic weights are expressed in atomic mass units (amu), one atom of the isotope C being defined
12
unstable elements the mass number of the most stable or best known isotope is given in brackets.
Group-+ v VI VII VIII 0
I II III IV
Period Series 2 He
lH 4.0026
1 1 1.00797
7N 80 9F lONe
3 Li 4 Be 5B 6C 20.183
2 2 9.0122 10.811 12.01115 14.0067 15.9994 18.9984
6.939 18 A
11 Na 12 Mg 13 Al 14 Si 15 P 16 S 17 Cl
3 3 32.064 35.453 39.948
22.9898 24.312 26.9815 28.086 30.9738
23 V 24 Cr 25 Mn 26 Fe 27 Co 28 Ni
c., 19K 20 Ca 21 Sc 22 Ti 58.9332 58.71
c,,.
4 40.08 44.956 47.90 50.942 51.996 54.9380 55.847
39.102 36 Kr
4 30 Zn 31 Ga 32 Ge 33 As 34 Se 35 Br
29 Cu 79.909 83.80
5 63.54 65.37 69.72 72.59 74.9216 78.96
41 Nb 42 Mo 43 Tc 44 Ru 45 Rh 46Pd
37 Rb 38 Sr 39 Y 40 Zr 106.4
6 87.62 88.905 91.22 92.906 95.94 [99] 101.07 102.905
85.47 54 Xe
5 49 In 50 Sn 51 Sb --52T e 53 I
47 Ag 48 Cd 131.30
7 107.870 112.40 114.82 118.69 121.75 127.60 126.9044
- - 55 Cs 56 Ba 57-71 72 Hf 73 Ta 74W 75 Re 76 Os 77 Ir 78 Pt
180.948 183.85 186.2 190.2 192.2 195.09
8 132.905 137.34 e
Lanthanid 178.49
6 series•
84Po 85At 86 Rn
79Au 80 Hg 81 Tl 82 Pb 83 Bi [222]
9 196.967 200.59 204.37 207.19 208.980 (210] (210]
- -- --
87 Fr 88 Ra 89
7 10 (223] [226] Actinide
seriesb
63 Eu I 64 Gd I 65 Tb I 66
60 Nd I 61 Pm I 62 Sm I 151.96
57 La I 58 Ce I 59 Pr I 144.24 Dy I 67 Ho j 68 Er I 69 Tm I 70 Yb I 174,97
• Lanthanid · . I 138.91 157.25 158.924 162.50 164.930 167.26 168.934 173.04
71 Lu
APPENDIX E
I Averaged Typical
Family Particle S . Strange- No. of
Symbol Mass pm nessb Charge Antiparticle 0 lifetime, decay
name name particles
seconds products
neutrino
---- -
Muon Muon µ- 206.77 1 - -e µ+ 2 2.212 X 10- 6 e- +Pe+ Pµ
2
family
1 -
-
Muon's I'µ O(?) 2 0 I'µ 2 Infinite -
neutrino
Mesons Pion 11"+ 273.2 0 0 +e 11"- 3 2.55 X 10-s µ+ + v,.
11"0 264.2 0 0 0 11"0 1.9 x 10- 16 'Y + 'Y
Lambda Ao 2182.8 1
2 -1 0 Ao 2 2.51 x 10- 10 p + 7r- ;,..
particle "d
"d
t?'J
z
Sigma par- ~+ 2327.7 2
1 -1 +e ~+
-
8. 1 x
10- 11 n + 7r+ s
ticle ~- 2340.5 2
1
-1 -e ~- 6 1.6 x
10- 10 n +1r- ~
l'."l
~o 2332 1 -1 0 ~o about 10- 20 Ao+'
2
• Adapted and modified from The World of Elementary Particles, by Kenneth W. Ford.
b This is a "quantum number" whose assignment permits an understanding of the inter-relationships of the particles.
O Antiparticles have the same mass
and spin as the particles but their charges and strangeness numbers are opposite in sign.
d The K 0 meson has two different lifetimes: all other particles have only one.
Symbols, Dimensions, and Units
for Physical Quantities
APPENDIX F
All units and dimensions are in the mksq (rationalized) system. The primary
units can be found by reading kilograms for Jl!I, meters for L, seconds for T, and
coulombs for Q. The symbols are those used in the text.
In practice, Q is defined in terms of Af, L, and T. However, the addition of Q to
the traditional iVl, L, and T enables us to avoid the use of fractional exponents in
dimensional considerations. The term 'rationalized' simply means that a factor
1/41r is separated out of Coulomb's law in order to remove the factor 41r that would
otherwise appear in many other formulas in electricity.
radian
Angular frequency and speed w r-1 radians/ sec
Angular momentum L llJL2T- 1 kg-m 2 /sec .
Angular velocity (I) r-1 radians/sec
Area A,S £2 meter 2
Displacement r, d L meter
39
40 APPENDIX F
APPENDIX G
Conversion factors for common and not-so-common units may be read off
directly from the tables below. For example, 1 degree = 2.778 X 10- 3 revolu-
tions, so 16.7° = 16.7 X 2.778 X 10- 3 rev. The mks quantities are capitalized
in each table.
PLANE ANGLE
0 I II
RADIAN rev
SOLID ANGLE
41
APPENDIX G
42
LENGTH
AREA
-
1 SQUARE METER = 1 10 4 10.76 1550 1.974 X 10 9
1 square centimeter = 10-4 1 1.076 X 10- 3 0.1550 1.974 X 10 5
1 square foot = 9.290 X 10- 2 929.0 1 144 1.833 X 10 8
1 square inch = 6.452 X 10- 4 6.452 6.944 X 10- 3 1 1.273 X 10 6
1 circular mill = 5.067 X 10- 10 5.067 X 10- 6 5.454 X 10- 9 7.854 X 10- 1 1
VOLUME
METER 3 cm 3 1 ft 3 in. 3
1 U. S. fluid gallon = 4 U. S. fluid quarts = 8 U. S. pints = 128 U. S. fluid ounces = 231 in. 3
1 British imperial gallon = the volume of 10 lb of water at 62° F = 277.42 in. 3
1 liter = the volume of 1 kg of water at its maximum density = 1000.028 cm 3
CONVERSION FACTORS 43
MASS
Note: Those quantities to the right of and below the heavy lines are not mass units at all but are often
used as such. When we write, for example,
this means that a kilogram is a mass that weiohs 2.205 pounds. Clearly this "equivalence" is approximate
(depending on the value of o) and is meaningful only for terrestrial measurements. Thus, care must be
employed when using the factors in the shaded portion of the table.
gm KG slug amu
DENSITY
Note: Those quantities to the right or below the heavy line are weight densities and, as such, are dimen-
sionally different from mass densities. Care must be used. (See note for mass table.)
1728
TIME
SPEED
FORCE
Note: Those quantities to the right of and below the heavy lines are not force units at all but are often
used as such, especially in chemistry. For instance, if we write
we mean that a gram-mass experiences a force of 980.7 dynes in the earth's gravitational field. Thus, care
must be employed when using the factors in the shaded portion of the table .
. dyne NT lb pd!
1 poundal = 0.1383
PRESSURE
inch of NT/
atm dyne/cm 2 cm Hg lb/in. 2 lb/ft 2
water METER 2
• Where the acceleration of gravity has the standard value 9.80665 meters/sec 2 •
--
1 British thermal unit = 1 I 1.055 777.9 3.929 1055 252.0 2.930 6.585 6.585
x 1015
x 1010 X 10- 4 X 10- 4 X 10 21
10-1 6.242 6.242
1 erg = I 9.481 1 7.376 3.725
x 10-14
2.389 2.778
x 10-14 0
X 10- 11 x 10-8 X 10- 8 X 10 11 X 10 5 0
1.285 1.356 1 5.051 1.356 0.3239 3.766 8.464 8.464
z
-"'1
1 foot-pound =
X 10- 3 X 10 7 X 10- 7 X 10- 7 x 1018 x 1012 t,j
i:i:i
[fl
1 horsepower-hour = 2545 2.685 1.980 1 2.685 6.414 0.7457 1.676 1.676 H
x 1013 X 10 6 X 10 6 X 10 5 x 1025 x 1019 0
z
1 JOULE= 9.481 101 0.7376 3.725 1 0.2389 2.778 6.242 6.242 >zj
X 10- 4 X 10- 7 X 10- 7 x 1018 x 1012 >
0
1 calorie = 3.968 4.186 3.087 1.559 4.186 1 1.163 2.613 2.613 8
X 10- 3 X 10 7 X 10- 6 X 10- 6 x 1019 x 1013 0
i:i:i
[fl
1 kilowatt-hour = I 3413 3.6 2.655 1.341 3.6 8.601 1 2.247 2.270
x 1013 X 10 6 X 10 6 X 10 5 x 1025 x 1019
10-6
1 electron volt = I 1.519 1.602 1.182 5.967 1.602 3.827 4.450 1
x 10-22 X 10- 12 x 10-19 x 10-26 x 10-19 X 10- 20 x 10-26
-
1 million electron volts = I 1.519 I 1.602 1.182 5.967 1.602 3.827 4.450 106
POWER
ELECTRIC CHARGE
1 abcoulomb (1
emu) = 1 2.778 X 10- 3 10 1.036 X 10- 4 2.998 X 10 10
1 ampere-hour = 360 1 3600 3.730 X 10- 2 1.079 X 10 13
1 COULOMB= 0.1 2.778 X 10- 4 1 1.036 X 10- 5 2.998 X 10 9
1 faraday = 9652 26.81 9.652 X 10 4 1 2.893 X 10 14
1 statcoulomb (1
esu) = 3.336 X 10- 11 9.266 X 10- 14 3.336 X 10- 10 3.456 X 10- 15 1
ELECTRIC CURRENT
ELECTRIC RESISTANCE
CAPACITANCE
INDUCTANCE
I
ab henry HENRYS µh mh stathenry
MAGNETIC FLUX
MAGNETIC INDUCTION B
WEBER/
gauss kiloline/in. 2 METER 2 milligauss
= TESLA
MAGNETOMOTIVE FORCE
1 abampere-turn = 1 10 12.57
1 AMPERE-TURN = 0.1 1 1.257
1 gilbert = 7.958 X 10- 2 0.7958 1
I
1 pragilbert = 4, amp-turn 1 esu = 2.655 X 10- 11 amp-turn
AMP-
amp-
abamp-turn/cm amp-turn/in. TURN/ oersted
turn/cm
METER
1 abampere-turn per
centimeter = 1 10 25.40 1000 12.57
1 ampere-turn per centi-
meter = 0.1 1 2.540 100 1.257
1 ampere-turn per inch = 3.937 X 10- 2 0.3937 1 39.37 0.4947
1 AMPERE-TURN per
METER= 0.001 0.01 2.540 X 10-2 1 1.257 X 10- 2
1 oersted = 7.958 X 10- 2 0.7958 2.021 79.58 1
0
~
0
Delta ~ 0 Pi IT 7r
Epsilon E E Rho p p
Zeta z r Sigma ~ q-
Eta H 1] Tau T T
Theta e () l ,a. Upsilon T v
Iota I Phi <I> cf,, <p
Kappa K IC Chi x x
Lambda A )\ Psi '1'" iJ;
Mu M µ Omega Q w
49
Mathematical Formulas
APPENDIX
Quadratic formula
-b + Vb 2 - 4ac
If ax 2 + bx +c= 0, then x = - - - - - - -
2a
sin()
x
= '!!__ cos() =-
r r
y x
tan() =- cot() = -
x y
r r
sec() = - csc () = -
x y
Fig. App. I
Pythagorean theorem
x2 + y2 = r2
Trigonometric identities
sin 2 () + cos = 1
2 () sec 2 () - tan 2 () = 1 csc 2 () - cot 2 () = 1
sin (a ± /3) = sin a cos /3 ± cos a sin /3
cos (a ± /3) = cos a cos /3 + sin a sin /3
t an (a R) = -
tJ + tan a ± tan /3
-----
- 1 + tan a tan /3
sin 2() = 2 sin () cos ()
cos 2() = cos 2 () - sin 2 () = 2 cos 2 () - 1 = 1 - 2 sin 2 ()
. ei8 _ e-i8 ei8 + e-i8
sm() = - - - - cos()=----
2i 2
e±;9 = cos () ± i sin ()
Taylor's series
x2 :r3
f(xo + x) = f(xo) + .f'(xo)x + f"(xo) -2! + f"'(xo) _:3!____ +
50
MATHEMATICAL FORMULAS 51
Series expansions (these expansions converge for -1 < x < 1, except as noted)
1
-- =
l+x
1 - x +x 2 - x3 + ··
_ ;-- x x2 x3
vl+x=l+---+-- · ·
2 8 16
1 x 3x 2 5x 3
---==1--+---+
v1 + x 2 s rn
x2 x3
ex = 1 + x + - + - + (-oo<x<oo)
2 6
. x3 xo
smx=x--+- - (-oo<x<oo)
6 120
x2 x4
cos x = 1 - - + - - (-oo<x<oo)
2 24
x3 2x 0
tanx=x+-+-+ (-71'/2 < x < 71'/2)
3 15
n n(n - 1)
(X + y)n = xn + _ xn-Iy + xn-2y2 +
1! 2!
In what follows, the letters u and v stand for any functions of x, and a and mare
constants. To each of the integrals should be added an arbitrary constant of
integration. A Short Table of Integrals by Peirce and Foster (Ginn and Co.) gives
a more extensive tabulation.
dx
1. - = 1 1.fdx=x
dx
d du
2. - (au)= a - 2. f au dx = a f u dx
dx dx
d du dv
3. - (u + v) = - +- 3. f (u + v) dx = f u dx +f v dx
dx dx dx
d
4. - xm = mxm-l
dx
4. f xm dx = ::\ (m -:;,!= -1)
d
5. -ln x = -
dx
1
x 5. f~ = ln JxJ
d
dx
dv
dx
du
6. -(uv) = u - + v -
dx
6. f u dv dx = uv -
dx .
f v du dx
dx
d
-/,-ex=ex
dx
7. f ex dx = ex
d
dx
.
11. - arcsm x = _ 1
1
-v 1 - x2
11. fV dx
1 - x2
= arcsin x
d
12. - arcsec x =
dx
_1
xv x 2
1
- 1
12. fV x
dx
x2 - 1
= arcsec x
d .
13. - cos x = - sm x 13. J sin x dx = - cos x
dx
Vector products
Let i, j, k be unit vectors in the x, y, z directions. Then
i · i = j · j = k · k = 1, i · j = j · k = k · i = 0,
i X i = j X j = k X k = 0,
ix j = k, j xk = i, k xi= j.
Any vector a with components a.,, ay, a. along the x, y, z axes can be written
a = a.,i + ayj + a.k.
Let a, h, c be arbitrary vectors with magnitudes a, b, c. Then
a X (h + c) = a Xh +a X c
(sa) X h = a X (sh) = s(a X h) (s a scalar).
Let B be the smaller of the two angles between a and h. Then
a · h = h · a = axbx + ayby + a.b. = ab cos B
i j k
a xh = -h xa = ax ay a. = (aybz - byaz)i + (a.bx - b.a,,)j
b,, by b. + (a.,by - bxay)k
la X hi = ab sin B
a · (h X c) = h · (c X a) = c · (a X h)
a X (h X c) = (a· c)h - (a· h)c
Values of Trigonometric Functions
APPENDIX J
TRIGONOMETRIC FUNCTIONS
53
54 APPENDIX J
B = µ 0 (H + M)
becomes !!c = (
41r) (~ H
c2 41r
+ cM)
or
in Gaussian form. Symbols used in this book that are not listed below remain
unchanged. The quantity c is the speed of light.
58
THE GAUSSIAN SYSTEM OF UNITS 59
In addition to casting the equations in the proper form it is of course necessary
to use a consistent set of units in those equations. Below we list some equivalent
quantities in mks and Gaussian units. This table can be used to transform units
from one system to the other.
Quantity Symbol Mks system Gaussian system
Length l 1 meter 10 2 cm
Mass m 1 kg 10 3 gm
Time t 1 sec 1 sec
Force F 1 newton 10 5 dynes
Work or Energy W,E 1 joule 10 7 ergs
Power p 1 watt 107 ergs/sec
Charge q 1 coulomb 3 X 10 9 statcoul
Current i 1 ampere 3 X 10 9 statamp
Electric field strength E 1 volt/meter i X 10- 4 statvolt/cm
Electric potential v 1 volt tio statvolt
Electric polarization p 1 coul/meter 2 3 X 10 5 statcoul/cm 2
Electric displacement D 1 coul/meter 2 121r X 10 5 statvolt/cm
Resistance R 1 ohm t X 10- 11 sec cm- 1
Capacitance c 1 farad 9 X 10 11 cm
Magnetic flux q>B 1 weber 10 8 maxwells
Magnetic induction B 1 tesla = 1 weber/ 10 4 gauss
meter 2
Magnetic field strength H 1 amp-turn/meter 41r X 10- 3 oersted
Magnetization M 1 weber/meter 2 1/41r X 10 4 gauss
Inductance I 1 henry t x 10- 11
All factors of 3 in the above table, apart from exponents, should be replaced by
(2.997925 ± 0.000003) for accurate work; this arises from the numerical value of
the speed of light. For example the mks unit of capacitance ( = 1 farad) is actu-
ally 8.98758 X 10 11 cm rather than 9 ( = 3 2 ) X 10 11 cm as listed above. This
example also shows that not only units but also the dimensions of physical quan-
tities may differ between the two systems. In the mks system (see Appendix F)
the dimensions of capacitance are M- 1L- 2 T 2 Q2 ; in the Gaussian system they are
simply L, the Gaussian standard unit of capacitance being 1 cm.
The student should consult Classical Electromagnetism, p. 611, by J. D. Jackson
(John Wiley and Sons, 1962) for a fuller treatment of units and dimensions.
Answers to Odd-Numbered
Problems
Chapter 26
1. 0.82 amp. 3. ±2.4 X 10-8 coul.
5. 2.3 X 10-s coul. 7. (a) Q = -2-y2q,
(b) no.
9. (a) 5.7 X 10 13 coul, 11. 3.8 X 10-5 coul, 1.2 X 10-5 coul.
(b) no,
(c) 660 tons.
13. (b) Along the body diagonal. 15. (a) 6.3 X 10 11 ,
(b) 7.3 X 10-13 ,
17. (a) 510 nt,
(b) 7. 7 X 1028 meters/sec 2•
Cha{er 27
.!:? -11
5. .fr X 10 coul.
9. (a) The larger charge produces a field of 13 X 104 nt/coul at the site of the small!
the smaller charge produces a field of 5.3 X 104 nt/coul at the site of the larger.
(b) 1.1 X 10-2 nt, repulsive .
. 11. E lies in the median plane and points radially away from the charge axis.
13. 1.0 X 105 nt/coul, pointing up.
17. E = q/8noa 2, pointing along the axis of symmetry and away from the hemisphe
1 qx
21. (a) E = - , 25. (a) 1.5 X 103 nt/coul,
41rfo (a2 + x2) %
(b) 2.4 X 10- 16 nt (up),
(b) E = _1_ (q1 - q2)a. (c) 1.6 X 10- 25 nt,
21r 2 fo (a 2 + x 2) % (d) 1.5 x 10 10 •
27. (a) 7.1 cm,
(b) 2.9 X 10-8 sec,
(c) 11 %.
dE Sq
29. -d = - - l3 , where l is the distance between the charges; yes.
z 11"€0
Chapter 28
1. 1rR 2E (an expected result). 3. EA cos o.
7. (a) 4.0 X 106 nt/coul, 9. (a) E = ,, / eo, to the left,
(b) E = 0. (b) E = 0,
(c) E = <T/eo, to the right.
11. 4.9 X 10-10 coul. 13. 0.44 mm.
19. (a) E = _q_ (radially inward), where l is the length of the cylinders,
21rtolr
(b) -q on inner surface and -q on outer surface,
Chapter 29
1. 0.89 mm. 3. 900 volts.
5. (a) Between the charges a distance of 25 cm from +q and outside the charges a
distance of 50 cm from +q.
(b) Outside the charges a distance of 140 cm from +q.
q d
7. VA - V s = - - - - - -
21reo a(a d) +
9. 1.9 X 10-29 coul meter; text value (0.61 X 10-29 coul meter) is lower and correct; the
assumptions made in this problem are oversimplified.
11. 99 tons. 13. -6.4 X 10-7 joule.
15. (a)-0.12 volt, 17. 2900 volts.
(b) 1.8 X 10-8 nt/coul, radially inward.
21. No. 23. (a) -180 volts,
(b) +2900 volts and -9000 volts.
27. -0.2lq 2/ eoa.
29. (a) 1.1 X 10 17 volts/meter,
(b) 4.6 X 10 21 volts/meter, assuming a nuclear radius of 5 X 10-15 meter.
31. (a) 2.6 X 105 volts, 33. 9.0 kw.
(b)
v5 c =
3 0.745 c.
35. (a) 3.2 X 10-13 joule,
(b) 1.6 X 10-13 joule,
(c) proton.
Chapter 30
1. (a) q2 = qs = 4.8 X 10-4 coul, 3. 43 µµf.
V2 = 240 volts,
Vs = 60 volts;
(b) q2 = 2.0 X 10-4 coul,
qs = 7.7 X 10-4 coul,
V2 = Vs = 96 volts;
(c) q2 = qg = zero,
V2 = Vs = zero.
11. 3.2 µf. 13. 4 µf.
15. 7.3 µf. 21. Mica.
23. Assuming K = 5.4:
(a) 104 volts/meter,
(b) +5.0 X 10-9 coul, on the positive plate,
ANSWERS TO ODD-NUMBERED PROBLEMS 63
(c) -4.1 X 10-9 coul, next to the positive plate.
25. 0.63 meter2 • 27. 0.11 joule/meter3 •
31. (a) 1.3 X 10-7 joule, 33. 7.0 X 10-5 coul.
(b) no.
35. 7.0f 37. (a) q1 = q2 = 3.3 X 10-4 coul,
qa = 4.0 X 10-4 coul;
(b) V1 = 33 volts,
V 2 = 67 volts,
Va = 100 volts;
(c) U1 = 5.4 X 10-3 joule,
U2 = 10.9 X 10-3 joule,
Ua = 2.0 X 10-2 joule.
Chapter 31
1. (a) 1200 coul, 3. (a) 2.4, iron being larger;
(b) 7.5 X 1021 electrons. (b) no.
5. 6. 7X 10-5 coul/meter2• 7. 54 ohms.
9. (a) 2.2 X 10-1 ohm, 11. (a) 260°C,
(b) nickel (p = 6.8 X 10-11 ohm -meter). (b) yes.
13. 0.39% (p), 17. 11 ohms.
0.0017% (Z),
0.0034% (A).
19. (a) 4.9 X 106 amp/meter 2, 21. 620 watts.
(b) 8.4 X 10-3 volt/meter,
(c) 2 .6 volts,
(d) 65 watts.
23. (a) 8. 7%,
(b) smaller.
Chapter 32
1. 1.1 X 104 joules. 3. -0.62%.
7. (a) 990 ohms, 9. vd - v. = 1.3 volt.
(b) 9.9 X 10-4 watt.
11. (a) 120 ohms, 13. (a) R = !r,
(b) i1 = 0.051 amp, (b) FJ'l /2r.
i2 = ia = 0.019 amp,
i4 = 0.013 amp.
17. 38 ohms. 19. (a) 10 ohms,
(b) 14 ohms,
(c) 10 ohms.
.
21. series: i
.= 2e
---·
R + 2r
parallel: i = _f,_,_ •
2R +r
The series arrangement always gives the larger current.
23. (a) Put R 1 roughly in the middle of its range; adjust current roughly with R 2 ; make
fine adjustment with R 1•
(b) Relatively large percentage changes in R 1 cause only small percentage changes in
the resistance of the parallel combination, thus permitting fine adjustment. The
ratio is 1 :21.
27. 2.7%. 29. 4.6 time constants.
33. (a) 9.6 X 10- 7 coul/ sec,
(b) 1.1 X 10-5 watt,
(c) 2.7 X 10-5 watt,
(d) ~ SI v 1 0-6 watt.
64 ANSWERS TO ODD-NUMBERED PROBLEMS
Cbapter33
1. (a) East,
(b) 6.3 X 1014 meters/sec 2,
(c) 3.0 mm.
3. 7.5 nt, perpendicular to the wire and to B.
5. See Appendix F. 7. 3.8 coul.
9. 450 gauss, perpendicular to the rod and inclined at an angle of 31 ° to the normal to
the plane of the tracks.
11. 4.3 X 10-3 nt-meter. The torque vector is parallel to the long side of the coil and
points down.
15. (a) 1.4 X 10-4 meter/sec, 17. (a) Kp = Kd = fKa,
(b) 4.5 X 10-23 nt. (down), (b) Rd= 14 cm.
(c) 2.8 X 10-4 volt/m~ter (down), (c) Ra= 14 cm.
(d) 5.7 X 10-6volt(top +,bottom-),
(e) same as (b).
19. (a) 2.6 X 107 meters/sec,
(b) 1.1 X 10-7 sec,
(c) 14 Mev,
(d) 7.0 X 106 volts.
23. 1.6 X 10-s weber /meter 2, horizontal and at right angles to the equator.
25. T = 3.6 X 10-10 sec, 27. 2.11 X 10-25 kg or 127 proton masses.
p = 0.17 mm,
r = 1.5 mm.
29. (a) Increase, 31. (a) 8.5 Mev,
(b) decrease. (b) 0.80 weber/meter2 ,
(c) 34 Mev,
(d) 24 me/sec
(e) 34 Mev, 1.6 webers/meter 2,
34 Mev, 12 me/sec.
33. 1.4. 35. 3800 meters/sec.
Chapter34
1. 7.9 X 10-3 weber/meter 2•
3. (a) 3.2 X 10-15 nt, parallel to current;
(b) 3.2 X 10-15 nt, radially outward if v is parallel to the current;
(c) zero.
7. 1.0 X 10-5 weber/meter.
9. B = 0 along a line parallel to the wire and 4.0 mm from it. If the current is horizontal
and points toward the observer and the external field points horizontally from left
to right, the line is directly above the wire.
11. 8.0 X 10-5 weber/meter 2, up. 13. 3.2 X 10-3 nt, toward the long wire.
23. (a) 1.0 X 10-3 weber/meter2 , out of figure;
(b) 8.0 X 10-4 weber/meter2, out of figure.
27. (a) 9.4 X 10-5 weber/meter2,
(b) 1.5 X 10-5 nt-meter.
Chapter 35
1. 2.0 X 10-2 coul. 3, Zero.
9. Vm = 1r2BR 2,,, 11. 3.0 X 10-4 volt.
i,,, = 1r2BR 2,,/RM
13. (a) 3.1 X 10-2 webers/sec, 17. (a) 4.4 X 107 meters/sec 2, to the right;
(b) left to right. (b) zero;
(c) 4.4 X 107 meters/sec 2, to the left.
ANSWERS TO ODD-NUMBERED PROBLEMS 65
Cllapter36
1. Let the current change at 10 amp/sec. 7. 1.0 X 10-7 weber.
9. 12 sec. 13. 27 amp/sec.
15. (a) 10 amp, 17. 0.63 joule/meter3 •
(b) 100 joules.
21. (a) 2.5 X 10-6 joule/meter, 23. 1.5 X 108 volt/meter.
(b) 14 X 10-6 joule/meter,
(c) 0.8 X 10-5 joule/meter.
Chapter 37
l. (a) 5.0 X 107 amp, 3. (a) 1.4 X 1011 volts/meter,
(b) yes, (b) 2.8 X 10-3 weber /meter2•
(c) no.
s. (a) 7.6 amp meter 2, 9. 7.5 X 10-6 weber /meter2•
(b) 11 nt-meter.
1 i. (a) 1.8 webers/meter2,
(b) 6.5 X 10-23 joule.
Chapter 38
1. 600, 710, 1100, and 1300 cycles/sec. 13. (a) 224 rad/sec.
(b) 207 or 241 rad/sec.
dE
19. id = EQ1r1° 2 -
dt
(r < R),
dE
id = Eo1rR2 -
dt
(r > R).
21. (a) 1.4 X 1014 volts/meter-sec,
(b) 9.9 X 10-6 weber/meter2 •
Chapter 39
3. 5. 7 cm, 9.0 X 108 meters/sec 9. 18 cm, 12 cm.
(=1.7c).
13. (a) a 2EB/µo
for faces parallel to xy-plane, zero for others;
(b) zero.
15. 1000 volts/meter,
3.4 X 10-6 weber/meter 2•
Chapter 40
1. (a) 5100 A and 6100 A,
(b) 5.5 X 10 14 cycles/sec and 1.8 X 10-15 sec.
3. F,ad = 6.0 X 108 nt, 7. 1.3 X 10-7 nt/meter2 •
Fgr = 3.6 X 10 22 nt.
11. 0.13. 13. (b) 7.3 (cycles/sec)/(miles/hour).
15. 3.8 X 10-2 A. 17. Yellow-orange.
Chapter 41
3. 2.05 X 108 meters/sec. 5, 1.56.
15. (b) 0.17.
19. Cover the center of each face with a circle of radius 0. 45 cm. The fraction covered
is 0.64.
Chapter 42
1. (a) 7, 3. 40 cm.
(b) 5,
(c) 2.
66 ANSWERS TO ODD-NUMBERED PROBLEMS
19. Assuming the light is incident from the left, a distance of - r(n - 2 )2 to the right of
the right edge of the sphere. 2( n - 1)
21. (a) R' is negative and R" is positive,
(b) i = - - r - ,
2n+l
virtual and erect.
(c)
23. (a) Coincides in location with original object and is enlarged 5.0 times,
(c) virtual and inverted.
Chapter43
1. Slit separation must be 0.034 mm. 3. 0.15°.
5. 6.4 X 10-3 mm. 7. 3.0 mm.
Chapter44
1. 0.17 mm.
3. (a) >-a = 2>..,,,
(b) coincidences occur when mb = 2m4 •
5, 79.7°. 7. (a) 52.9°,
(b) 10.2°,
(c) 5.1 °.
9. 9100 meters. 11. (a) 0.16 sec of arc,
(b) 7.4 X 107 km,
(c) 2.2 X 10- 2 mm.
15. 3. 17. (a) Must have d = 4a,
(b) Every fourth fringe is missing.
Chapter45
1. Three complete orders.
5. The intensity would be concentrated near the twentieth order for blue and the eleventh
order for red. The orders would overlap to such an extent as to appear almost white.
ANSWERS TO ODD-NUMBERED PROBLEMS 67
7. All wavelengths shorter than 6300 A.
9. 0°, ±10°, ±21 °, ±32°, ±45°, and ±62°.
11. 5200 A to 6200 A. 17. (a) 0.0032°/A,
0.0077°/A,
0.024°/A;
(b) 40,000,
80,000,·
120,000.
19. 3600 lines.
21. (a) 4.6 X 10-2 A.
(b) No. The resolution could normally be improved by going to a higher order
diffraction, but in this case m = 3 is the highest order that can exist (assuming
that the light falls on the grating at right angles).
23. (a) 6 X 104 A,
(b) 1.5 X 104 A,
(c) 0, 1, 2, 3, 5, 6, 7, 9. The tenth order is at 8 = 90°.
25. 3°, 38 ° (clockwise); 27. Yes, n = 3 for X = 1.29 A,
15°, 31° (counterclockwise). n = .4 for X = 0.97 A.
Chapter 46
1. (a) ±55°, ±125°,
(b) ±35°, ±145°.
3. Assuming that a right-handed coordinate system is used,
(a) circular, counterclockwise as seen facing the source;
(b) elliptical, counterclockwise as seen facing the source, major axis of ellipse along
y.= x;
(c) plane, along y = -x.
5. 55°30' to 55°46'. 11. 2.7 X 10-14 kg-meter 2 /sec2,
2.1 hr.
Chapter47
l. 4.8A. 3. 2.1 X 1015 photons/sec.
7. 5.9 X 10-6 ev. 9. (a) 2.0 ev,
(b) zero,
(c) 2.0 volts,
(d) 3000 A.
11. 100 years. 17. (a) 1,
5.3 X 10-11 meter,
(b)
1.1 X 10-34 joule-sec,
(c)
2.0 X 10-24 kg-meter /sec,
(d)
4.1 X 1016 radians/sec,
(e)
(J) 2.2 X 106 meters/sec,
(g) 8.2 X 10-s nt,
(h) 9.0 X 1022 meters/sec 2
(i) + 13.6 ev,
(j) -27.2 ev,
(k) -13.6 ev.
19. +0.21 ev. 21. 2.6 ev.
25. +54 ev, assuming the electron to be in its ground state initially.
27, (a) 2.6 X 10-13 meter,
(b) 2800 ev,
(c) 4.4 A.
68 ANSWERS TO ODD-NUMBERED PROBLEMS
Chapter 48
1. (a) 1.7 X 10-35 meter.
3. (a) 3.3 X 10-24 kg-meter/sec, for each;
(b) 38 ev for the electron and 6200 ev for the photon.
5. (a) Higher orders cannot exist for this accelerating potential and for these planes.
(b) 59°; the crystal must be rotated with respect to the incident beam to satisfy
Bragg's law for this new wavelength.
7. 1.5 A. 9. (a) 0.20,
(b) 0.40,
(c) 0.33,
(d) 0.33.
11. 0.32. 15. 6.6 X 10-23 kg-meter/sec.
Answers to Odd-Numbered
Supplementary Problems
J.~q, g.q·
Chapter 27 81reo ri r2
qQ
II. ( a ) - - ·
I. 21r 21r 41reoK
g--E g+-E
m m
(b) ~ ·
(b) -
x L
.
41r,o y(y + L)
Chapter 28 (c) Zero.
I. +3.54 X 10- 6 coul.
Chapter 30
3. 1.98 X 10 7 meters/sec.
Q C1C2 + C1Ca CV
5. q1 = 1 O,
5. A = - ·
21ra 2 C1C2 + C1C3 + C2C3
q2 = q3 =
Chapter 29
I. 1.2 X 10- 8 coul/meter 2 .
69
70 ANSWERS TO ODD-NUMBERED SUPPLEMENTARY PROBLEMS
7. (a) lOcofarad,
(b) µoi ~ ' out of page at an
(b) 13.6co farad, 211'R
(c) 1200Eo coul, 1200co coul,
(d) 10 4 volts/meter, angle of 18° with page.
(e) 2.1 X 10 3 volts/meter,
(f) 88 volts, Chapter 35
(g) 19 X 10 3co joules. µoi11'r 2 R 2
3. (a)
2x 3 '
Chapter 31
3µ01rir 2v
I. (a) 2.3 X 10 12 , (b) - - - ·
2R 2N 4
(b) 5.0 X 10 3,
(c) 10 7 volts. (c) In the same sense as the current
in the larger loop.
5. (a) PA = .10 ohm-meter, PB
.05 ohm-meter, -1
(b) EA = 1.0 volt/meter, EB = .50
5. (a) q = R [<1>B(t2) - <l>B(ti)],
3.
3 Y2 µoi , toward center of square.
2 Chapter 39
411'a 271'€0
I. (a) In (b/a)
5. zero.
7. µoi
- (- - -
4 R1 R2
1 1) .
, mto page. (b)
µo In (b /a)
271'
Chapter 44
Chapter 40
I. (a) The droplets act as diffracting
I. (a) 1.0 X 10 8 cycles/sec,
obstacles producing the observed
(b) 1.0 X 10- 5 weber/meter 2,
ring.
(c) 2.1 meter- 1 , 6.3 X 10 8 rad/sec,
(b) .20 mm.
(d) 120 watts/meter 2 ,
3. 71 ft.
(e) 4.0 X 10- 7 A nt, 4.0 X 10- 7 nt/
meter 2 • Chapter 45
3. (b) About 6 X 10-7 meter (com-
parable with the wavelength of 3. (a) 5860 A.
light). (b) Since the fourth maximum is
(c) No. missing, the slit width must lie
between 1.00 and 1.69 microns.
Chapter 41 5. (a) One line to each side of the cen-
tral maximum.
I. 1.95 X 10 8 meters/sec. (b) 1.44 X 10- 4 rad.
3. 180 cm.
5. (a) v~i_+_s-in_2_0_1 1
(c) b.O = -
R
(b) 0
Chapter 46
Chapter 42 I. 67% polarized, 33% unpolarized.
I. 6. 3. Side A has the Polaroid, and the
5. (b) No. Polaroid axis must be at 45° to the
(c) Rays pass undeviated through principal axes of the quarter-wave
system. plate.
Index
180° = 1r rad
1 radian = 57.3° = 0.159 rev
1 slug = 32.2 lb (mass) = 14.6 kg
l kilogram = 2.21 lb (mass)
l pound (mass) = 0.454 kg
1 atomic mass unit = 1.66 X 10- 27 kg
l meter = 3\J.4 in. = 3.28 ft; l inch = 2.54 cm
1 mile = 5280 ft = 1.61 km
1 angstrom unit = 10- 10 meter = 0.1 mµ
1 millimicron = 10- 9 meter
1 liter = 61.0 in. 3
1 ft 3 = 28.3 Ii
l day = 86,400 sec
1 year = 3.16 X 107 sec = 365 days
l mile/hr = 1.47 ft/sec = 0.447 meter/sec
1 pound = 4.45 nt; 1 newton = 0.225 lb
1 atmosphere = 29.9 in.-Ilg = 76.0 cm-Hg = 1.01 X 10 5 nt/meter 2
1 Btu = 778 ft-lb = 252 cal = 1060 joules
1 calorie = 4. rn joules; 1 joule = 0.2:39 cal = 2. 78 X 10-7 kw-hr
1 electron volt = 1.60 X 10- 19 joule
l horsepower = 550 ft-lb/sec = 746 watts
1 weber /meter 2 = 1 tesla = 10 4 gauss
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: