Intro Special

University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln

Robert Katz Publications Research Papers in Physics and Astronomy
January 1964
An Introduction to the Special Theory of Relativity

Robert Katz
University of Nebraska-Lincoln, rkatz2@unl.edu
Follow this and additional works at: http://digitalcommons.unl.edu/physicskatz

Part of the Physics Commons
Katz, Robert, "An Introduction to the Special Theory of Relativity" (1964). Robert Katz Publications. 49.
http://digitalcommons.unl.edu/physicskatz/49
This Article is brought to you for free and open access by the Research Papers in Physics and Astronomy at DigitalCommons@University of Nebraska -
Lincoln. It has been accepted for inclusion in Robert Katz Publications by an authorized administrator of DigitalCommons@University of Nebraska -
Lincoln.
FOR STEVE AND JOHN
Published, copyright 1964 D. Van Nostrand Company, Inc.
Electronic edition published, copyright 2008 Robert Katz.

Preface
It is the purpose of this book to provide an introduction to

the Special Theory of Relativity which is accessible to any stu-
dent who has had an introduction to general physics and some
slight acquaintance with the calculus. Much of the material is at
a level suitable for high school students who have had advanced
placement in physics and mathematics. Since some of the exposi-
tion, particularly part of Chapter 6, is presented here in book
form for the first time, even terminal graduate students in physics
may find the material profitable.
The subject matter of the book has been taught, in the form
presented here, to first-year graduate students at Kansas State
University for many years. Much of it has sifted into courses at
lower level.
The typescript was read by Professors E. U. Condon, University
of Colorado, and Melba Phillips, University of Chicago. I wish
to thank them for their helpful criticism and valuable suggestions.
May 26, 1964 ROBERT KATZ
Manhattan, Kansas
s
Table of Contents
Preface 5
1 The Galilean Transformation 9
The Inertial Frame, 9; The Galilean Transformation, 11; The
Speed of Light, 17; The Special Theory of Relativity, 24; The
Rod Clock, 25
2 The Lorentz Transformation 29
The Lorentz Transformation, 29; Simultaneity and Time Se-
quence,31; Time Dilation, 33; Lorentz Contraction, 34; Vel-
ocity Transformations, 37; The Fizeau Experiment, 38; Aber-
ration, 40; Visual Appearance of Rapidly Moving Objects, 44;
Transformation of Acceleration, 48
3 Force and Motion 51
Introduction, 51; Newton's Second Law, 53; The Equivalence
Principle, 56; Transformation of Forces, 58
4 Energy and Momentum 61
Work, 61; Kinetic Energy (Nonrelativistic), 63; Kinetic Energy
(Relativistic), 64; Conservation of Linear Momentum, 68;
Center of Mass (Nonrelativistic), 71; Transformation of Mo-
mentum and Energy (Relativistic). 75; Center of Mass (Rela-
tivistic). 78
5 Some Relativistic Phenomena 83
Pair Production and Annihilation, 83; Compton Effect, 85;
Threshold for 7To Meson Production, 87; Doppler Effect, 89;
Relativistic Dynamics, 92; Gravitation and Light, 96
6 Relativity and Electromagnetism 103
Introduction, 103; The Lorentz Force, 105; Magnetization and
Polarization, 106; Transformations of Fields and Flux Densi-
ties, 108; Electromagnetic Induction, 110; Field of a Moving
Charge, 114; Transformation of Polarization and Magnetiza-
tion,117; The Unipolar Generator, 121; Postscript, 123
General Bibliography 127
Index 129
7
1 The Galilean
Transformation
I-I The Inertial Frame Does the earth move around the sun
or the sun around the earth? Do we describe the motion of the
moon by saying that it traces a nearly circular path around the
earth, or a somewhat more complicated path around the sun?
These questions are associated with the choice of a frame of
reference, a background against which we imagine motions to
be referred. The choice is not a simple one, since it has philo-
sophic as well as physical overtones. The central position ac-
corded to man in the universe led to the belief that the sun and
stars circulate about the earth, and in consequence, the universe
of Ptolemy (A.D. 127-151) demanded that heavenly bodies move
in paths compounded from circles, called epicycles. The path
of a point which moves on a small circle, whose center in turn
moves on a larger circle, is called an epicycle. In part the com-
position of the paths of heavenly bodies from circles arose from
the view that only the most perfect figure, the circle, was ap-
propriate to any object in the heavens. Copernicus' (1473-1543)
relocation of the origin of the coordinate frame to the sun led to
a new grouping of the earth and planets as a system of solar
satellites and resulted in a new set of physical laws, whose truth
was universal, valid on earth and in heaven. These were the laws
of Newton (1642-1727).
The philosophic implications of the relocation of the origin
of astronomical coordinates have been widely elaborated under
names like rationalism, or naturalism, or secularism. It is our
9
lOAN INTRODUCTION TO THE SPECIAL THEORY OF RELATIVITY
purpose here to emphasize the importance of the choice of a

coordinate frame in physics, though clearly the choice has far
broader implications in the affairs of men.
The logical choice of a coordinate frame to Newton was one
in which the stars were at rest. There are some physicists today
to whom this choice seems a proper one. A concept called Mach's
Principle, propounded by Ernst Mach (1838-1916), states that
inertia is entirely due to the mutual action of matter, so that force
is required to accelerate matter only because it is displaced rela-
tive to the other matter of the universe. More generally, physicists
place a weaker restriction on the choice of a coordinate frame:
the frame must be inertial. An inertial frame is one in which
Newton's first law is true.
For many centuries before Newton, inquiries into the way the
world was structured led to the belief that the natural state of
things on earth was a state of rest-objects left alone remained
at rest. To set them in motion and to keep them in motion was
thought to require the application of a force. Newton's formula-
tion gave equal importance to rest and to uniform motion in a
straight line, and asserted that either of these states was natural;
that is, either of these conditions would continue indefinitely if
no force was applied. This was in great contradiction to earlier
beliefs which supposed that force was required to sustain motion
as well as to initiate motion.
Newton's First Law states: A body at rest remains at rest, and
a body in motion will continue in motion with steady speed in a
straight line as long as no outside force acts on the body.
Today we accept the first law as axiomatic, and use it as a
recipe for finding a coordinate frame in which to describe physi-
cal systems. Unless otherwise stated all physical laws stated in this
book (and in most others) are with respect to an inertial frame.
The symbols which appear in physical equations represent quan-
tities which are usually measured by observers stationed in an
inertial frame.
A test for an inertial frame is simplicity itself. A frame of
reference may be thought of as a mesh of lines. In free space, in
THE GAlilEAN TRANSFORMATION 11
the absence of gravitational or other force fields, a particle set

down in an inertial frame may always be found there. If the
particle is set in motion, it will move with steady speed in a
straight line. Its coordinates, measured in this frame, will satisfy
the equation of a straight line, and its motion along this line
will be with constant speed. If the particle does not remain at
the point at which it is placed, or does not move in a straight line
with steady speed, then the frame is not inertial. A perfectly
smooth, level, spinning turntable on earth is not an inertial
frame. A particle placed on this turntable will remain at a single
point with respect to the earth as the turntable slips under it,
but it will describe a circular path with respect to the turntable.
Thus the turntable is not an inertial frame, for though no forces
act on the particle it is neither at rest nor moving in a straight
line with constant speed with respect to the turntable.
Experiment has shown that Newton's frame of reference, fixed
in the stars, is an inertial frame. A coordinate frame fixed in the
earth is not inertial, due to the earth's rotation about its axis, and
about the sun as well. For most practical purposes, and to the
accuracy with which many experiments are performed, the fact
that a frame fixed in the earth is not inertial does not alter the
outcome of experiments sufficiently to cause concern. In fact it
was only through the Foucault pendulum that the rotation of
the earth was first clearly demonstrated, in 1851. Foucault (1819-
1868) suspended a heavy pendulum bob by a long flexible wire,
and observed that the plane of vibration of the pendulum ro-
tated with respect to the floor. But there was no way the support
point could twist the plane in which the pendulum oscillated
once it had been started. He concluded that the floor was rotat-
ing beneath the pendulum. We infer from such an experiment
that the earth is not an inertial frame. There are everyday ex-
periences in which the distinction between inertial and non-
inertial frames must be made. One of these lies in the circulation
of the atmosphere, and another in the meaning of the term
centrifugal force.
12 The Galilean Transformation Through his studies of
12 AN INTRODUCTION TO THE SPECIAL THEORY OF RELATIVITY
projectile motion, Galileo (1564-1642) concluded that the motion

of a projectile launched from the ground at an arbitrary angle
could be derived from the motion of a projectile launched
straight up. A projectile thrown straight up from a uniformly
moving cart would be seen from the cart as moving straight up
and down, but the motion seen from the ground could be pre-
dicted by superimposing the motion of the cart onto that of the
projectile. These ideas of Galileo are the basis of the Galilean
transformation, through which we relate motions as they are
observed in two different inertial frames. We will see that Gali-
lean transformations must be superseded by a different transfor-
mation, the Lorentz transformation, when the two frames move
at great speeds with respect to each other. But for ordinary ex-
perience the Galilean transformation is much simpler and is
quite satisfactory.
Let us suppose that two sets of observers are available to study
a problem. One set is at rest with respect to the ground, and
makes its measurements with respect to a coordinate frame and
a set of clocks fixed in the ground and synchronized to read alike.
We call the set of observers and the observations they make the
unprimed set, and refer to their measured coordinates and time
as x, y, z, t, without primed superscripts. A second set of observers
is at rest with respect to a second coordinate frame, the primed
frame, which moves with constant velocity with respect to the
unprimed frame, as if the second set of observers were riding on
a large space platform. We will call the primed observers and
their observations the primed set. We will refer to their observa-
tions as x', y', z', t'.
We choose the two sets of coordinates to be parallel. The x'
axis is parallel to the x axis, the y' axis to the y axis, the z' axis
parallel to the z axis. Further, we choose the orientation of the
axes such that the motion of the primed frame is parallel to the
x axis. The primed frame moves in the +x direction with velocity
V with respect to the unprimed frame. For convenience we
choose the origins of the two sets of coordinates to coincide at
= =
times t t' 0, as in Fig. 1-2.1.
THE GALILEAN TRANSFORMATION 13
z z
FIG. 1-2.1
Suppose that a particle is located at a point P at a particular

time, and that measurements of the coordinates and the time
are made by the two sets of observers. The measured values can
be related by the equations
X' = x - Vt, (1-2.1a)
y' = y, (1-2.1b)
Zl = z, (1-2.1c)
and t' = t. (1-2.1d)
The equations 1-2.1 are the Galilean transformation equations,
which relate the observations of position and time made by two
sets of observers located on two different inertial frames, as
described.
The point P may be on the path of a projectile. To describe
the motion properly we need to say something about the velocity
and the acceleration of the projectile. This implies that we wish
to study differential increments of the coordinates in the two
frames. We take differentials of the transformation equations
1-2.1, recalling that V is constant, to find
dx' = dx - V dt, (1-2.2a)

dy' = dy, (1-2.2b)
dz' = dz, (1-2.2c)
and dt' = dt. (1-2.2d)
Now we divide each of the equations 1-2.2a-c by Eq. 1-2.2d to
find transformation equations for the components of the velocity.
We have
dx' dx
dt' = dt - V, or V",, = v'" - V. (1-2.3a)
&.:. = rJ1, or vII = VII' (1-2.3b)

dt' dt
dz' dz
- = -,
dt' dt
or V,, = v,. (1-2.3c)
These results may be put into vector form by multiplying each

of these equations by the appropriate unit vectors. A unit vector
is a dimensionless quantity which indicates direction. Thus the
unit vector 1.. is parallel to the x axis and of magnitude 1. Our
primed and unprimed coordinate axes are parallel so that 1.., =
1.., and so on. Thus if we multiply Eq. 1-2.3a by 1.., Eq. 1-2.3b
by I", and Eq. 1-2.3c by I, and add these equations, we find that
v""t.., + vllt ll + v"t" = v",t", + VI/tl/ + v,t, - Vt",
which may be condensed to read
v' = v-V. (1-2.4)
If we transpose the velocity of the primed frame with respect to
the unprimed frame, we obtain
v = v' + V. (1-2.5)
It is through Eq. 1-2.5 that the airspeed of an airplane is con-
verted to groundspeed, using knowledge of the velocity of the
air with respect to the ground. When we speak of windspeed we
generally mean the speed an air mass is moving with respect to
the ground. When we speak of the airspeed of an airplane, or its
cruising speed, we mean the speed with which the airplane moves
with respect to the air mass. For navigation with respect to the
earth the groundspeed is required, or more precisely, the velocity
with respect to the ground is required. The Galilean transforma-
tion makes air navigation possible.
In ordinary conversation we are sometimes careless about
specifying the frame to which a position or a motion is referred.
When we give the cruising speed of an airplane we may be vague
in noting that it is with respect to the air mass through which
the plane is moving. When we give the windspeed we may be
vague in noting that this is with respect to the ground. Consider
the question, with what speed is the air blowing past the wings
of an airplane whose cruising speed is 200 miles per hour (mph)
when it flies into a 50 mph headwind? By the definition of cruis-
ing speed the answer is 200 mph, when the airplane is operating
properly. If the unprimed coordinate frame is fixed in the ground
and the primed set is fixed in the air mass, then the design of the
airplane implies that its speed tI is 200 mph for proper operation.
If tI is very much less than 200 mph, the airplane falls down; if
tI is much greater than 200 mph, the wings fall off.
Neither the magnitude of the velocity nor the direction of the
path (the direction of the velocity vector) is the same in the two
frames. As a simple illustration consider the motion of a boat on
a river. Suppose that the water moves east at 3 mph and that a
boat heads north at 4 mph with respect to the water (see Fig.
1-2.2). If we take east as the positive x direction and north as
the positive y direction, then the motion of the boat with respect
to the primed frame (fixed in the water) is
V' = 4 mph Xl II'

Its motion with respect to the ground is the vector sum of the
motion of the boat with respect to the water, v', and the motion
of the water with respect to the ground, V (= 3 mph X 1",). Ac-
cording to Eq. 1-2.5, we have
V = (3 1% + 4 1~) mph.
v-5mph
v - 4mph.
lyL I....
Y-3mpb
FIG. 1-2.2
The boat moves at 5 mph with respect to the ground in a direc-

tion 53 north of east.
To examine the way in which accelerations transform in two
inertial frames, we may simply take time derivatives of Eq. 1-2.5,
bearing in mind that t = t'. Thus we may differentiate v' with
respect to t' and v with respect to t. Noting that V is constant, we
find
dv'= -dv,
-
dt' dt
or a' = a. (1-2.6)
In this section we have been careful to emphasize that the time
measured in both primed and unprimed frames has the same
value, that the operation of clocks does not appear to depend on
the speed with which they move. This view derives from ordinary
experience which is summarized in the Galilean transformation
equations. We have done this in anticipation of our study of
relativity where a distinction must be made between measure-
ments of time in two inertial frames. This distinction is one of
the key problems of the special theory of relativity. While the
Galilean transformation is extremely useful in the analysis of
physical systems, the work of Lorentz and Einstein at the begin-
ning of the present century showed it to be inadequate to describe
the behavior of physical systems when the translational velocity
V is very large, of the order of the speed of light in vacuum

(186,000 miles per second). When speeds approach the speed of
light, physicists have learned that they must use a different set
of transformation equations to relate motions in two inertial
frames. These are called the Lorentz transformation equations.
We will devote a great deal of attention to examining these
equations and their consequences.
Nevertheless, although the Galilean transformation is now
known to be inadequate to relate observations made on systems
moving at high speeds to observations of the same systems mov-
ing slowly, it is simpler and sufficiently accurate in the limit of
low velocities to be of great practical importance. Man has not
yet learned to move anything but elementary particles and per-
haps the nuclei of atoms at speeds approaching that of light.
Thus for most practical purposes and for virtually all of engineer-
ing the Galilean transformation yields sufficiently correct results.
1-2.1 A ship heads due north at 30 mph with respect to the sea. A wind
blows to the east at 40 mph with respect to the sea. Find the velocity of
the wind with respect to the ship, as it might be measured by an ob-
server on the ship. [50 mph, 37 south of east]
1-2.2 After going a mile upstream in a motor boat, a man accidentally
drops an oar overboard. He proceeds upstream for 10 minutes before
he misses the oar. He then turns round and retrieves the oar at the
point from which he started originally. If the boat travels at constant
speed with respect to the water, what is the speed of the current?
Work the problem twice, in the frame fixed in the banks, and in the
frame fixed in the stream. [3 mph]
1-2.3 A boy throws a stone to the east at 30 ft/sec from a northbound
train traveling at 88 ft/sec on straight tracks. The stone's initial velocity
is at 37 above the horizontal. The stone leaves the boy's hand at 10 ft
above the ground. Find the parametric equations of motion of the
stone with the point of projection as origin in a frame fixed in the
ground, and in a frame fixed in the train.
1.3 The Speed of Light In principle, measurement of the
speed of light is not different from the measurement of any other
speed. One measures the time required for light to travel over a
measured path. The difficulty arises from its great rapidity, so
that either long paths or short time intervals must be measured.

The first measurement of the speed of light was astronomical.
In 1675 Olaf Roemer (1644-1710), a Danish astronomer, ob-
served that a greater time elapsed between the eclipses of one of
Jupiter's moons during the season the earth is receding from
Jupiter, position 1 in Fig. 1-3.1, than during the season the
----~ -
Jupiter's orbit
@ Eorth', ,,~t
2
FIG. 1-3.1
earth is approaching Jupiter, position 2 in Fig. 1-3.1. He inter-

preted this discrepancy as due to the finite speed of light, asserting
that the light had to travel a greater distance between eclipses
when the earth was receding than when the earth was approach-
ing Jupiter. Roemer concluded that it took light 22 minutes to
cross the earth's orbit, a distance of about 3 X 108 km, giving a
speed of about 2.3 X 108 mjsec. Roemer's was a beautiful infer-
ence, giving a value within about 25 percent of the presently
accepted value of the speed of light c, for today we take
c = 2.997925 0.000003 X 108 m/sec.
An earlier attempt to measure the speed of light was made by
Galileo, who stationed two observers with lanterns a mile apart.

They were to signal each other by uncovering their lanterns ac-
cording to a prearranged schedule. While such a system might
have yielded a value for the speed of sound, it was unable to
respond with sufficient speed to determine the speed of light. Sub-
sequent experimenters replaced one of Galileo's obse"rvers with a
mirror and replaced his crude measurement of time by the rota-
tion of a toothed wheel. The toothed wheel was at once a shutter
to let light pass and a dock, for, at the proper speed of rotation,
light passing through one opening could reach the mirror and
return through the next opening in the wheel. Such an arrange-
ment was used much later by Fizeau (1819-1896) in 1849, with a
base line of 8.633 km from source to mirror. In Fizeau's experi-
ment the wheel had 720 teeth and rotated at 12.6 revolutions/
sec. More modern experiments form a pulse of light (or other
electromagnetic radiation) electronically, and detect it in a
similar way. Today it is possible to measure time intervals nano-
seconds (10- 9 ) long. But in principle the experimental measure-
ment of the speed of light is closely related to the system used by
Galileo.
Toward the end of the nineteenth century. as measurements
of the speed of light became more refined, the question arose-
What was the coordinate frame with respect to which the speed
of light was measured? Was the speed of light which was meas-
ured its speed with respect to the earth? to the stars? to Jupiter?
Was there a preferred medium for light, as air is a preferred
medium for sound?
It is clear that sound travels through air, and that the speed
of sound must be measured with respect to the air. Experiment
shows that the velocity of sound with respect to the earth can be
obtained by a Galilean transformation by adding the velocity of
sound with respect to the air to the velocity of the air with re-
spect to the earth. By analogy, physicists supposed that there was
a preferred medium for light which filled infinite space. This
medium was called the ether. The ether was thought to be
perfectly transparent to light, and to offer no resistance to the
passage of stars or the planets. Its function was to provide a

coordinate frame to answer the question-"With respect to what
is the velocity of light c?" Since the speed of the earth's orbital
motion is about 30,000 mjsec, it was supposed that sensitive
experiments might be devised to show that the speed of light
with respect to the earth would depend on the direction the light
traveled with respect to the earth's motion through the ether,
in accordance to the Galilean transformation.
Consider an experimental apparatus which might be devised
to study variations in the speed of light with respect to the earth,
as xelated to the motion of the earth through space. Such an
apparatus might compare the time for a light ray to pass down
a rod and be reflected back, when the rod was parallel to the
earth's motion, and then when the rodwas perpendicular to the
earth's motion.
Imagine an equipment consisting of a rod of length L provided
with a mirror at one end, and with a pulsed light source, a de-
tector, and a clock at the other end. In Fizeau's experiment the
detector was the eye, and the light pulser-clock was the toothed
wheel. With this idealized equipment, suitably scaled and suit-
ably fast, we might measure the speed of light.
When the rod is at rest with respect to the ether, the time for
the light to make a round trip is tJ.to, where
toto = 2L/c. (1-3.1a)

When the rod is moving through the ether with speed v and is
oriented parallel to its direction of motion, as in Fig. 1-3.2a, we
find the speed of light with respect to the rod by the Galilean
transformation, for this is the hypothesis to be tested. Then the
speed of light with respect to the rod is c - v when light is mov-
ing to the mirror, and c + v on the return trip. The elapsed time
for this trip is tot II' as given by
tot ll = L/(c - v) + L/(c + v),
or tot _2L 1 (t-3.tb)
II - c (1 - V 2/C 2)
THE GALILEAN TRANSfORMATION 21
t~4~C+V~c-r.. I
(a)
I t:
(b)
FIG. 1-3.2
Suppose the rod is moving through the ether with speed v,

Fig. 1-3.2b, the rod being oriented perpendicular to its direction
of motion. By the Galilean transformation the speed of light
with respect to the rod is (c2 - V 2)1/2 on both the outgoing and
the return trip, so that the elapsed time is At.L' given by
t:.t
.L
= 2
c (1
1
_ v2/c2) 1/2
(1 3 1 )
- C
If we now make use of the binomial theorem for (1 + x)m,

(1 + x)'" = 1 +;; +m(m2~ 1) x2
+ .
m(m - 1) ... (m - n + 2) ,,-1
(1-3.2)
(n - 1)! x ,
we may expand Eqs. 1-3.lb and c to the quadratic term to find
AlII = AloO + 2
V /C 2 + ) (1-3.3a)
and Atl. = Ato(l + !V 2/C 2+ ). (1-3.3b)
The differences in the elapsed time for the motion of light
through the ether in the three cases examined are in the second
order (quadratic terms) of vjc. That is, these times differ in the
difference between (vjC)2 and 1. For the velocities with which
the earth moves in its orbit, the quantity (vjC)2 is about 10- 8
We must measure t:..t II and t:..t l. to a precision of at least one part
in a hundred million if two separate measurements are to be
made and compared.
However, a way around the difficulty was conceived by Albert
A. Michelson (1852-1931), and in 1887 Michelson and Morley
performed an experiment which should have been able to detect
the motion of the earth through the ether, or put in another way,
should have been able to test the applicability of the Galilean
transformation to the motion of light. Michelson reasoned that a
direct comparison of the parallel and perpendicular transit times
might be made by using light waves as their own means of meas-
uring time. Suppose we combine the parallel and perpendicular
rods, as in Fig. 1-3.3. Then a beam of light from a source is
split into two by a lightly silvered mirror, half the light moving
B
~
/Beom splitter
'V ~
FIG. 1-3.3
down rod B and the other half down rod A. Upon reflection the
light returns down the rods and can be inspected by an observer
looking through the same mirror, but at 90 to the light source.
Since the light is a wave motion, the two beams start out in phase,
or in step. If the beams return in step, then the time required for
the two trips is the same, but if they return out of step the time
difference between the two trips is the time for half an oscillation.
If the light combines in step, the observer notes that the view is
bright, while if out of step the two waves cancel, and the view is
dark.
It is easy to see that this apparatus might have the necessary
precision. Yellow light, of wavelength 600 millimicrons (nanom-
eters), has a period of 2 X 10- 15 sec, so that the time for half an
oscillation is about 10- 15 sec. An interferometer in which the
length of each section is about 15 m yields an elapsed time of
10- 7 sec for the return trip. Thus a time measurement to half
the period of oscillation of a light wave could measure the transit
time to 10- 15/10- 7 , or to I part in 108 , as required. In the actual
experiment the path length employed was I I m.
Since there was no independent way of determining that the
length of the two paths was identical, the instrument was first
adjusted to yield a bright view to the observer and then was ro-
tated by 90. By this means any difference between the parallel
and perpendicular transit times could be detected.
Although the design of the instrument showed that it was
capable of detecting the time differences predicted by the Gali-
lean transformation, no shift in the appearance of the field of
view was detected which could be attributed to motion through
the ether. The experiment was a failure, but this failure was one
which in retrospect proved to be a great success.
The Michelson-Morley experiment was repeated many times,
by different observers, using apparatus constructed of different
materials, in different laboratories. Each time the experiment
failed to detect the required difference in time for the light to
traverse paths parallel and perpendicular to the earth's motion.
Physicists, disturbed by these results, tried to grasp at straws.
It was supposed that the ether was dragged (convected) along

the surface of the earth and for this reason no effect was found
in the Michelson-Morley experiment. Other experiments were
performed to detect the convected ether, without success.
An ingenious postulate proposed by H. A. Lorentz (1853-1928),
and known as the Lorentz contraction, was that moving objects
shrink in their direction of motion through the ether. If the
amount of shrinkage was properly chosen, the times of flight in
the two paths would be the same. Suppose that a rod at rest in
the ether has a length L o, and that a rod moving through the
ether with its length aligned perpendicular to its direction of
motion is unaltered, that is LJ. = L o Then Eq. 1-3.lc gives the
time t:.t J. correctly, provided that we replace L by L J. or by L o. If
a rod moving through the ether and aligned parallel to its mo-
tion shrinks, with its new length given by
(1-3.4)
we would find that
2Lo 1
At n = t:.tJ. = ~ (1 _ V 2 /C 2 )1/2 (1-3.5)
This result would account for the failure to detect a time differ-
ence in the two paths in the Michelson-Morley experiment.
These explanations had an unsatisfactory ring. Invented to ex-
plain away a negative experiment, they had no other detectable
consequence than to explain the effect for which they were cre-
ated.
But a train of thought was set into motion by the failure of the
Michelson-Morley experiment to find an ether, culminating in
the Special Theory of Relativity of Albert Einstein (1879-1956)
in 1905, a theory that no physical experiment could detect the
absolute motion of an inertial reference frame.
1-4 The Special Theory of Relativity In order to explain the
results of the Michelson-Morley experiment, we might propose
that nature behaves in a way we would not have anticipated, that
the Galilean transformation is incorrect, and that velocities do
not add vectorially (at high speeds). Following Einstein, we reason
that there is no purpose in postulating the existence of an ether

if there is no way to detect it; that is, if no physical experiment
can detect the absolute motion of an inertial reference frame.
Thus there is no purpose in speaking of a velocity of light except
with respect to the observer who measures it. We reason further
that no one observer has a special place, superior to all others.
The speed of light should have the same numerical value for all
observers, so long as they are all resident in inertial frames. We
propose, as a fundamental postulate of the relativity theory, that
physical law must have the same meaning in all inertial frames.
This postulate is called the postulate of covariance of physical
law.
As a first step in the application of the general principle, let
us suppose that the constancy of the speed of light is a physical
law; that is, the speed of light is the same in every inertial frame.
Then the speed of light is the same in every direction and does
not depend on the earth's speed through space or its speed with
respect to an undetectable ether.
While the relativity principle quickly dispels the need for a
convected ether and for the shrinkage of moving objects, it intro-
duces other questions needing further exploration, for the princi-
ple seems to contradict ordinary experience. We will see that the
relativity theory does not contradict ordinary experience. Rather
it contradicts extrapolations from ordinary experience which are
not ordinarily subjected to experimental test. While we know
that the Galilean transformation is true in ordinary experience,
we have no basis for assuming, without experimental test, that
it is true at speeds comparable to the speed of light. We will
require of statements of relativity that they agree with ordinary
experience in the limit of low velocities. Whenever an experi-
mental test has been made of deductions from the special theory
of relativity, these deductions have been shown to be quantita-
tively correct. In consequence physicists have great confidence in
relativity theory.
1-5 The Rod Clock One consequence of the special theory
of relativity that has aroused great popular interest lies in the
measurement of time. Moving clocks do not run at the same rate

as clocks fixed in the laboratory. We shall see that moving clocks
Tun slow.
Let us design a clock, taking account of our new view of the
speed of light. If the speed of light is constant, and has the same
value for observers in any inertial frame, then in principle we
have no need for independent measures of length and time. A
clock may be made from a rod, as in Fig. 1-3.2: a rod of length L
provided with a mirror at one end and provided at the other
end with a pulsed light source and a detector with a fast trigger
to the source. Such a system is an oscillator, or a rod clock. If the
light source is initially flashed, the light will travel down the
stick and back with speed c and be detected; the detector will
then trigger the light source to flash again. The time interval
between successive flashes is constant. In principle we may there-
fore define any two of the three quantities, the unit of length, the
unit of time, or the speed of light, but not all three. At present
it is simpler to define a fundamental unit of length and a funda-
mental unit of time, and to measure the speed of light, but there
is no basic reason why this must always be so.
Consider a moving rod clock, as in Fig. 1-5.1, oriented trans-
3
FIG. 1-5.1
verse to its direction of motion. Suppose that there is an observer

in the coordinate frame in which the rod is at rest. We call this
THE GALIlEAN TRANSFORMATION 27
the proper frame of the rod. An observer in the proper frame of

the rod will note that the light flashes in intervals of 2L/c, as-
suming the detector to be infinitely fast. An observer in the
laboratory notes that the path of the beam from the moving
source to the moving mirror is not L long, but is from the source
at position I to the mirror at position 2, and thence to the de-
tector at position 3. For him t:..t is the time from I to 2, so that the
interval between flashes is 2t:..t, according to
c2 lit2 = L2 +v 2
lit 2,
or lit = !:. 1
c (1 - v2 I c2 )1/2
Thus if 2t:..to is the interval between flashes as determined by a
proper observer (for whom the rod clock is at rest), we find that
an observer in the laboratory frame, for whom the rod clock is
moving with speed v transverse to the orientation of the rod,
measures the time interval between flashes to be 2M, and that
these two are related by the equation
(1-5.1)
In the literature of relativity it is customary to use the notation
vic = {J, and (1 - v2Ic 2)-1/2 = (1 - {J2)-1/2 = 'Y. (1-5.2)
Note that the relations between 13 and y resemble the relations
between sin 8 and sec 8 in a right triangle whose hypotenuse is
I, and where the side opposite the angle (J is of length p. Then
we have sin 8 = 13, sec 8 = y.
With the notation of Eq. 1-5.2, Eq. 1-5.1 becomes
lit = 'Y lito_ (1-5.3)
Now consider the meaning of the rod clock experiment. The
laboratory observer and the moving observer both have rod
clocks. Each observer knows his clock to keep correct time. Yet
whichever is considered to be the observer at rest notes that the
moving clock runs at an incorrect rate. Instead of measuring a
time interval between flashes of 2t:..t o which the rest clock reads,
28 AN INTRODUCTION TO THE SPECIAL THEORY OF RELATIVln
the moving clock is observed by the rest observer to take a time

2y.::ltobetween flashes. The speed of a moving object is always
less than c, so that y is always greater than 1.
An observer in an inertial frame, comparing two clocks, one
at rest with respect to his frame and the other moving with ve-
locity v with respect to his frame, must conclude that the inter-
val between flashes of the moving clock is longer than the cor-
responding interval of the rest clock. A clock consists of two parts,
one which beats off time intervals, and the other which counts
them. The counter in the laboratory which tallies the flashes of
the moving clock will show fewer counts than the counter which
tallies the flashes of the rest clock. The laboratory observer must
conclude that moving clocks run slow.
While we have here discussed a rod clock, no restriction to this
special model is implied. All clocks on a given inertial frame run
at the same rate, if they are running correctly. All moving clocks
run slow, whether these be rod clocks, or atomic clocks, or bio-
logical docks, or any other kind.
1-5.1 Find 'Y for fJ = S/5, 4/5, 5/U, 12/IS. Use the right triangle
relationships. If sin (J=fJ, then sec = 'Y. [1.2, 1.67, 1.08, 2.6]
(J
=
1-5.2 If fJ I - x, then 'Y- 2 = 2x - x 2 2x for small x or for fJ close
E!!
to I. Prove that the error in 'Y in making the approximation that x 2

may be neglected in comparison to 2x is -25x percent.
CHAPTER BIBLIOGRAPHY
P. Hume and D. Ivey, Frames of Reference (a film), Modern Learning
Aids, New York.
Measurement of the Speed of Light (a film), McGraw-Hill Book Com-
pany, Inc., Text Film Department, New York.
R. S. Shankland, "Michelson Morley Experiment," Am. J. Phys. 32,
16 (1964).
2 The Lorentz Transformation
2-1 The Lorentz Transformation We may consider a co-

ordinate transformation between two inertial frames as a dic-
tionary by whose means observers residing on the two frames can
make their meaning clear to each other. In ordinary experience
we need the assistance of the Galilean transformation to reconcile
the determination of velocities on two inertial frames. An ob-
server on the ground who measures the velocity of an airplane
with respect to his frame will arrive at an entirely different result
than the pilot reading his airspeed indicator and his compass
heading. If these observers communicate by radio, their results
are completely contradictory until the conflict is resolved through
the Galilean transformation.
Consider the requirements upon a set of transformation equa-
tions which will fulfill our needs:
(1) The transformation must be linear; that is, a single event
in one inertial frame must tra'nsform to a single event in another
frame, with a single set of coordinates.
(2) The transformation must approach the Galilean transfor-
mation in the limit of low speeds. Here low speeds mean low
compared to c, so that we wish to examine the transformation in
the limit f3 ~ O.
(3) The speed of light must have the same value, c, in every
inertial frame.
Just as the disturbance in a pond resulting from dropping a
pebble into the water is a system of circular ripples, so a flash of
light spreads out as a growing sphere. We may describe this
sphere, whose radius grows at speed c, by the equation
x2 + y2 + %2 = C2[2. (2-1.1)
29
Just as the spreading light flash forms a spreading sphere in the

unprimed frame, so it must also form a spreading sphere in the
primed frame. Suppose the flash of light takes place at the instant
when both t and t' are zero, and when the origins of the two
frames coincide. Then the equation of the sphere of light in the
primed frame must be
(2-1.2)
The coordinate transformation that satisfies these requirements
is called the Lorentz transformation, after its discoverer, and
forms the mathematical beginning of the special theory of rela-
tivity. The space and time coordinates of an event recorded in
two inertial frames whose axes are parallel, and whose origins
coincide at time zero (in both frames), are related by the equa-
tions
x' = 'Y(x - VI), and conversely x = 'Y(x' + VI'); (2-1.3)
y' = y, Y = y';
z' = z, z = z';
t' = 'Y(t - Vx/c 2
), t = 'Y(t' + Vx' /c 2
).
In these equations the primed frame is taken to be moving with

speed V in the +x direction with respect to the unprimed frame.
Conversely the unprimed frame may be taken to be moving with
speed V in the -x' direction with respect to the primed frame.
As indicated in Eqs. 1-5.2, we use the symbol y = (1 - V2 /C 2 )-1/2.
The Lorentz transformation fulfills the first two requirements
we have stated. It is linear, and in the limit V ~ 0, y ~ 1, the
Eqs. 2-1.3 become the Galilean transformation, Eqs. 1-2.1; the
Lorentz transformation yields the results of ordinary experience
in the limit of low velocities.
By direct substitution of Eqs. 2-1.3 into the equation of the
light sphere in the unprimed frame, Eq. 2-1.1, we find the equa-
tion of the light sphere in the primed frame.
Thus the Lorentz transformation fulfills requirement (3)
above. A pulse of light emitted at the coincident origin of the
THE LORENTZ TRANSFORMATION 31
two frames spreads out as a sphere in both frames. When ob-

servers in the unprimed frame translate their observations to
the language of the primed frame, the translated observation is
in agreement with observations of the primed observers, that a
light sphere spreads out from the origin of the primed coordi-
nates with speed c.
2-I.l By direct application of the Lorentz tranSiformation equations,
show that Eq. 2-1.1 is transformed into Eq. 2-1.2.
2-1.2 By taking differentials of the Lorentz transformation equations,
show that the quantity ds transforms to ds', where ds 2 = dx 2 + dy2 +
=
dz 2 - c2dt 2, and ds'2 dx'2 + d.,'2 + dz'2 - c 2dt'2.
2-2 Simultaneity and Time Sequence Suppose we now ex-

amine some implications of the Lorentz transformations. Let
there be as many observers in each inertial frame as are required.
Observers in each frame have access to identical meter sticks, and
to identical clocks. They use identical procedures in their meas-
urements. Each measured event is described at least by a set of
four coordinates, giving the position and time at which the
event took place.
In any inertial frame the clocks of all observers are synchro-
nized. They have been checked by stationing a third observer mid-
way between any two. He notes that the pair of clocks reads the
same. This procedure takes account of the finite speed of light,
and insures that there are no alterations in the clock's behavior,
as there might be if they were carried to a central timing station.
In anyone frame there is no doubt as to when two events are
simultaneous, wherever they occur. If two events occur at the
same time, the same reading of the clocks located at the places
where the events occurred assures us that the events were simul-
taneous.
Let us consider two events, I and 2, which are simultaneous in
=
the unprimed frame. Then t 1 t 2 , wherever the events took
place. But whether they are noted as simultaneous in the primed
frame depends on where the events took place. From the Lorentz
transformation, Eqs. 3-1.3, we note that
t\ = 'Y(t1 - VxI/c 2) and t'2 = 'Y(t2 - VX2/C2).

The difference between the time of the two events in the primed
frame is
V
t, 2 - t ,1 = 'Y
-; ( Xl - X2) . (2-2.1)
C
Two events simultaneous in the unprimed frame are observed to

be simultaneous in the primed frame only if they occur at the
same point, "1 = "2' The second event might appear earlier, or
later than the first, depending on the value of their" coordinates.
Suppose now that the event t 1 was first, and t 2 was second, in
the unprimed frame. The time difference between them may be
obtained from the above equations as
t'2 - t'l = 'Y [(t2 - t1) - ~ (X2 - Xl) 1 (2-2.2)
If the quantity in brackets is equal to zero, then the two events

are simultaneous in the primed frame. If greater than zero, then
the events are observed in the primed frame in the same order as
in the unprimed frame. If less than zero, then the events are
observed in reverse order. The last case can arise only if
X2 - Xl > C(t2 - tl),
that is, if the two events happen at such remote places that a
ray of light leaving event I could not have reached point 2 in
time to cause event 2. We say that the time sequence of two
events can only be inverted if they could not have been "causally
connected"; that is, if event I could conceivably have caused event
2 (by sending a radio signal, or by tripping a switch), then the
order of the events cannot be inverted. No signal can travel faster
than the speed of light, so that if a light beam could not bridge
the interval between the two events the second event could not
have had knowledge of the prior occurrence of the first.
2-2.1 Two light bulbs in the laboratory, one at " = 0 and the other
at "= =
10 km, are set to flash simultaneously at t O. Observers on a
magic carpet moving in the +" direction with speed ~ X 107 m/sec
observe the flashes. (a) What time interval do they note between flashes?
(b) Which bulb do they say goes off first? [(a) !1.~4 X 10- 9 sec (b) the
bulb at 10 km]
22.2 A long straight rQd is inclined at angle (J to the x axis. The rod
moves in the y direction with velocity V. (a) Find the velocity v with
which the point of intersection of the rod and the x-axis moves along
= =
the x axis. (b) If V 1010 em/sec. and (J 0.10 rad, what is the numeri-
cal value of v? (c) Does your result contradict the relativistic demand
that c is a limiting speed for all material particles? [(a) v = V cot (J (b)
1011 em/sec (c) no. for the point of intersection of the ruler with the x
axis is a mathematical point rather than a material particle.]
23 Time Dilation Suppose we examine, from the vantage
point of the laboratory (the unprimed frame), a clock which is
stationary in the primed frame. The primed frame is the proper
frame of the clock, for this is the frame in which the clock is at
rest. The clock then is moving with speed V in the +x direction
with respect to the laboratory.
Since the clock is in fixed x' position, the time interval between
two beats of the clock at 1'1 and 1'2 will appear in the laboratory
frame to be
t2 - tl = 'Y(t'2 - t'I), (2-3.1)
by application of the Lorentz transformation, Eqs. 21.3, for
X'2= X'I' We may therefore set X'2 = =
X'1 0 if we wish, for
nothing is changed if the clock is at the origin of the primed co-
ordinates.
This is the same result that we have obtained earlier for the
rod clock, 1-5, in a slightly different notation. But here the re-
sult does not depend on the detailed design of the clock.
One variety of subatomic particles, called mesons, decays
at an exponential rate such that lie of the original number re-
mains after 2.6 X 10- 8 sec, in a coordinate frame in which the
mesons are at rest. Beams of 71' mesons may be produced by bom-
barding the target of an accelerator with high energy protons; the
resulting mesons move out of the target at a speed which can be
O.99c. In their proper frame 71' mesons decay at their proper rate,
but laboratory observers measuring the decay rate of the moving
mesons must find that the meson docks run slow. From Eq. 2-3.1
the decay time will be measured in the laboratory to be 2.6 X

10- 8 X (1 - 0.99 2)-1/2 sec, or 1.87 X 10- 7 sec. In that time a
meson beam moving with a speed of nearly 3 X 108 m/sec will
move a distance of 56 m. Thus in 56 m the meson beam will
drop to l/e of its initial intensity. If this relativistic effect did
not exist, we would expect the intensity of the beam to drop
by l/e each 2.6 X 10- 8 sec in the laboratory, regardless of the mo-
=
tion. Since in this case y 7.18, we would here have expected
the beam to drop by e- 7 . 18 , or to drop to 1/1300 of its initial
intensity in 56 m. The difference between 1/1300 and l/e =
1/2.78 is readily detected. Time dilation is thus confirmed in
the laboratory, and is today routinely taken for granted in the
design of experiments which must be performed with particle
beams in high-energy accelerators.
Let us repeat once again what we mean by time dilation. We
mean that a proper time interval is dilated or expanded when
measured from some inertial frame other than the proper frame.
The proper frame is the frame in which the events took place
at a single point. There is no change in any proper frame by vir-
tue of its motion with respect to another frame. In fact it is one
of the basic principles of the special theory of relativity that there
is no physical effect which can be associated with the motion of
an inertial frame. Such a motion cannot be detected by measure-
ments wholly within that frame. It is impossible to build a com-
pletely self-contained speedometer which measures the motion of
an inertial frame, except by reference to a second frame.
2-3.1 A particle with a mean proper lifetime of I p.Sec moves through
the laboratory at 2.7 X 1010 em/sec. (a) What is its lifetime, as measured
by observers in the laboratory? (b) If it was manufactured in the target
of an accelerator, how far does it go, on the average, in the laboratory
before disintegrating? (c) Repeat the calculation of the preceding part
without taking relativity into account. [(a) 2.3 p.Sec (b) 6.2 X 104 em
(c) 2.7 X 104 em]
2-4 Lorentz Contraction Let us set up an imaginary experi-

ment to measure the length of a fixed rod and the length of a
moving rod. There is no problem in measuring the length of a
rod which is at rest in a coordinate frame. Observers in the frame

simply note the coordinates of the ends of the rod, at their con-
venience, and apply the Pythagorean theorem to determine the
length. Or, more simply, they might line up the rod with the x
axis, and take the difference in the x coordinates of the ends of
the rod as the length. But how do we measure the length of a
moving rod?
Suppose we put observers everywhere along the x axis in the
laboratory frame when the rod is moving in the +x direction
with speed V and is aligned parallel to the x axis. We ask the
observers to synchronize their watches, and then to give some sig-
nal if the end of the rod is at their coordinate position at a pre-
determined time. The difference in the coordinates of the two-
observers in the laboratory frame who signal that they see oppo-
site ends of the rod at the appointed time is to be taken as the-
length of the rod.
The results of the length measurement clearly depend on the
meaning of simultaneity. Depending on the choice of the recipe
for synchronizing watches, the rod could turn out to have any
length at all. The relativistic recipe has already been chosen, and
has been incorporated into the Lorentz transformation. We com-
pare the coordinates of the ends of the rod in the primed and un-
primed frames at the same laboratory (unprimed) time. From
Eqs. 2-1.3 we find
X'2 = 'Y(X2 - Vt) and X'l = 'Y(XI - Vt),
which we subtract:
X'2 - X'l = 'Y(X2 - Xl)' (2-4.1)
On the left of Eq. 2-4.1 we have the difference in the coordinates.
of the ends of the rod as determined in the primed frame, the
proper frame of the rod. On the right the difference in the co-
ordinates of the ends of the rod as determined in the laboratory-
frame appear. Let us make the identifications
x' 2 - X'l = L' and X2 - Xl = L,
so that we have
L' = "fL. (2-4.2)

Remembering that 'Y is at least I, we find that the measured
length of the rod in the laboratory frame, L, is less than its proper
length L' by the factor 'Y' This effect is today called the Lorentz
contraction, for the formula with which Lorentz attempted to
explain the negative result of the Michelson-Morley experiment is
precisely the same as we have found here, though the interpreta-
tion of the formula is now different from his.
Notice that we do not require that rods shrink in their direc-
tion of motion for a proper observer, for whom the rod is at
rest, as Lorentz proposed. Relativity requires that the contraction
is related to the act of measurement. The way in which we set up
to measure the length of a moving rod determines that we will
measure a shorter length than the rod length. We would not have
come to such a conclusion if the speed of light were infinite, for
then p would be zero and the Lorentz transformation would be
identical to the Galilean transformation. It is the finite speed of
light which gives rise to these results.
Is the moving rod really contracted in its direction of motion?
Is time really dilated? These questions depend on what is meant
by really. In physics what is real is identical with what is meas-
ured. There is no way to assign properties to a rod or to a clock
or to an electron except through measurement. In this sense the
phenomena we have discussed as time dilation and the Lorentz
contraction are real. But a physicist tries to unite his observations
into a concise set of ideas. He measures the lifetime of 7r mesons
at rest and in motion and finds it appropriate to combine these
measurements into a proper lifetime, for all otheilifetimes are
easily calculable from this one. In general it is the proper length
or the proper lifetime which one assigns to the meson or the rod.
But in the laboratory moving mesons are really alive long after
their twins, born at the same time, and at rest, are really dead.
2-4.1 A rod has a length of I m. When the rod is in a satellite moving
with respect to the earth at O.99c, what is the length of the rod as de-
termined by an observer in the satellite? [I m]
THE LORENTZ TRANSFORMAnON 37
2-4.2 A vector represented in coordinate form in the primed frame is
given as 8 I"" + 6 I",. Find its representation in the unprimed frame,
if the primed frame moves at V = 0.75e I. with respect to the unprimed
frame. [5.3 I", + 6 I,,]
2-4.3 A thin rod of length L o when measured by a proper observer is
moving at 0.75e with respect to a second observer in a direction at 37
to its own length. What is the length L of the rod as measured by the
second observer and his assistants? [0.8 L o]
25 Velocity Transformations To find the Lorentz trans-
formations of the velocities we employ the same technique we
have used before, in the case of the Galilean transformation, 1-2.
By taking differentials of Eqs. 2-1.3 we find
dx' = -y(dx - V dt), and dx = -y(dx' + V dt'); (2-5.1a)
dy' = dy, dy = dy'; (2-S.1b)
d:l = dz, dz = dz'; (2-5.1c)
dt' = -y(dt - V dx/e2 ) , dt' = -y(dt' + V dx' /e 2). (2-5.1d)
When we divide each of the first three equations in each column
by the fourth equation of that column, we find
U'" = U'" - V, U'",+ V

and U" = U'V; (2-5.2a)
1 _ U",V
1+-"'-
e2 2 e
U' U" . u, U'".' (2-5.2b)
".' = -y ( 1 -
V",V)'
-
2
1/ = -y(1 + u;:~y
e
. b b'
In these equations we have written V. = dt' V'.. = d? and so
on. V. is the x component of the velocity of the particle as
measured in the unprimed frame. V'tJJis the x' component of the
velocity of the particle as measured in the primed frame. The sub-
script y,z refers to either the y or the z component. In Eqs. 2-5.2
the primed frame is moving in the +x direction with respect to
the unprimed frame with speed V.
Note that Eqs. 2-5.2 reduce to the Galilean velocity transforma-
tions of Eqs. 1-2.3 in the nonrelativistic limit f3 ~ O. But there

is a significant difference at large velocities. 1 the moving object
is a photon moving in the primed frame with speed c in the +x
direction, and if the primed frame moves with speed c with re-
spect to the unprimed frame, then
Vz = e
--2
+ e = e.
1+:-
e2
The speed of light is c in all inertial frames, whatever their rela-
tive speed. By the Galilean transformation we would have ex-
pected V., = 2c, but we know that this is not applicable. Regard-
= =
less of the value of VJ if V'., c, then V., c. Of course this is
how it must be, for the condition that the speed of light is c
in all inertial frames is built into the Lorentz transformation.
25.1 (a) Two particles come toward each other, each with speed
O.ge with respect to the laboratory. What is their relative speed?
(b) Two particles are emitted from a disintegrating source, each mov-
ing with speed O.ge with respect to the source. What is their speed rela-
tive to each other? [(a) and (b) O.995e]
2-5.2 A particle has velocity V' = 3 1"" + 4 111' + 12 l z ' mjsec in a
coordinate frame which itself moves in the +x direction with respect to
the laboratory at V = O.8e. Find V in the laboratory frame. [V =
2.4 X 108 1., + 2.4 111 + 7.2 l z ) mjsec]
2-6 The Fizeau Experiment How can we detect evidence of
the velocity transformation formulas in the laboratory? We have
already seen that the speed of light in vacuum is c in every iner-
tial frame, but what about its speed in a medium, such as water?
The speed of light in a medium must clearly be with respect to
a coordinate frame fixed in the medium, for the very structure
of the medium, the position of its atoms and molecules, provides
a preferred reference frame. The speed of light in a medium is
less than CJ and the index of refraction for a particular wave
length is n = clv, a number generally greater than one.
If the medium is in motion with respect to the laboratory, at
speed V, then the speed of light with respect to the laboratory U
wiII be different from its speed with respect to the medium U' ,
according to Eqs. 2-5.2. If the light moves parallel to the motion
of the medium and in the same direction, then its speed with
respect to the laboratory is increased to Up, while if the light
moves antiparallel to the motion of the medium its speed with
respect to the laboratory is reduced to U a , both given from Eqs.
2-5.2 as
U'+ V U'- V
Up = U'V and Ua = U'V (2-6.1)
1+7 1--
e2
Consider the time difference between the transit times of two
beams moving in opposite directions through a moving medium
for a distance L as measured in the laboratory. The beams will
take times whose difference !:i.t is given by
!:i.t = .!:.... _ .!:.... = L

Up Ua
(1 +U' U'V/e
+V
1-U' U'V/e
2
- V
_
2
). (2-6.2)
Suppose we have beams of yellow light moving in opposite di-

rections through flowing water. As a practical case let us take the
path to be L = =
15 m, U' 0.7e (for the refractive index of water
= =
is about 1.5), and V 3 mjsec lO-se, for the steady flow of
water in a tube. \\le find
!:i.t = 15m (1 + 0.7 X 10- 8

_ 1 - 0.7 X 10-8)
+ 1O- e
0.7e 8 0.7e - 10- 8e
= 2.1 X ~~~ m/sec (2 X 0.7 X 10- 8

)
= 10-15 sec.
This time is about the time for half an oscillation of yellow light,
as we saw in 1-3 in our discussion of the Michelson-Morley ex-
periment. In an analogous way a beam of light can be split and
the separate parts sent two ways around a tube of flowing water,
as in Fig. 2-6.1. If the system is inspected when the water is at rest,
then a readily detectable difference in the appearance of the
image may be observed when the flow is turned on, in quantita-
-e
, 1
, ~
I'
1
b !4"
III;
j "
T ...
FIG. 2-6.1 Fizeau experiment. Light from a source is split by a lightly
silvered mirror at m. One beam palles through the flowing fluid in a
direction parallel to the flow, and the other in a direction anti parallel
to the flow.
tive agreement with our calculations. This experiment was first

performed by Fizeau (1819-1896) in 1851.
27 Aberration In 1726 Bradley first observed that distant
stars lying perpendicular to the plane of the earth's orbit describe
a small, nearly circular, annual orbit, 41" in angular diameter.
As shown in Fig. 2-7.1, the telescope must be pointed away from
the normal to the orbit, into the direction of the earth's motion,
as though to let the light fall through. This phenomenon is
" star, at rest In

primed frame
,
v (Apparent star
motion)
(a)
, ,. .,'
11' 11'
II 11
~,.-~
~
(b) te)
FIG. 2-7.1
wholly understandable in terms of the velocity transformation

equations.
If the proper frame of the star is the primed frame, the compo-
nents of the velocity of the starlight reaching the earth in this
frame are V'II = C, V':J) = 0, V'z = 0, to a sufficient approximation.

When the earth is at position A, Fig. 2-7.1a, and moving with
speed V with respect to the sun, then according to our notation
we think of the primed frame as moving in the +x direction with
speed V with respect to the unprimed frame. From Eqs. 2-5.2 we
find the components of the speed of light in the unprimed
(earth's) frame as
U. = V, VII = _q,-l = -(c2 - V 2 )1/2, V. = O. (2-7.1)
The resultant of these components is c, Fig. 2-7.1c. The direction
of the resultant is the direction in which the light will be seen to
come when the earth is at A. At B the tilt of the telescope must
be as indicated, differently from its orientation at A, and again
so as to "catch the light."
From Fig. 2-7.1 c we have
sin ep = vic. (2-7.2)
Thus the earth's speed can be measured through the stellar aber-
ration provided that we know the speed of light. Taking c =
= =
3 X 108 m/sec and ep 20.5", we find V 30 km/sec. Aberration
furnishes independent proof that the earth revolves about the
sun, and not the sun around the earth. Knowing the earth's speed
and the length of the year, we can find the circumference of the
earth's orbit, and thus the distance from the earth to the sun.
Aberration is also of importance in high-energy physics. Rap-
idly moving protons in the cosmic rays enter the earth's atmos-
phere and are decelerated, radiating light. Particles accelerated
in high-energy accelerators move in curved paths, being cen-
tripetally accelerated in this deflection; they thereby radiate elec-
tromagnetic energy. In both cases this radiation is predominantly
in the forward direction of the motion, as seen in the laboratory.
This bunching of the radiant energy in the forward direction
results from aberration, and is known as the headlight effect.
Suppose a source of radiation radiates energy uniformly in all
directions in its proper frame (the primed frame), as shown in
Fig. 2-7.2. The source moves in the +x direction with respect to
y' ( ') ( )
V a O.9c
(a)
b'
,~I~
Q- -~
~~ c'
(') ( )
Proper Loboratory
frome from.
(b)
FIG. 2-7.2 Headlight efled [After V. F. Weisskopf, Physics Today 13,
24 (1960)].
the laboratory frame (the unprimed frame) with speed V. Radia-

tion emitted at an angle 8' with respect to the x' axis will be
observed to make an angle 8 with respect to the x axis. We find
this angle by noting that V'z = c cos 8' and V", = c cos 8, and
substituting these values into Eq. 2-5.2a for the transformation of
the x component of the velocity, which gives
cos 8' + {3
cos 8 = 1 + {3 cos (/ (2-7.3)
On Fig. 2-7.2b we have mapped corresponding rays in the

=
primed and unprimed frames, when f3 0.9, using Eq. 2-7.3.
When 8' is 0 or 71', then 8 has the same values, but when 8' =
= = =
71'/2 we find cos 8' 0 so that cos 8 0.9 and 8 25, and
so on. The rays are strongly concentrated in the forward direction
with very little radiation coming off in the backward direction,
as seen in the laboratory. It is this effect that we have called the
headlight effect. The bunching of radiation in the direction of
motion may be observed in the radiation from accelerated
charged particles in the machines of high energy physics, where
it is called synchrotron radiation. The radiation may be observed
as visible light by looking into the direction of motion of the
beam.
27.1 At what speed does an observer have to move toward a star in
order that half the radiation emitted by the star is concentrated within
a cone subtending an angle of 0.01 radian. Approximate cos 6 as I -
62/2 + .... [0.9995c]
27.2 Prove that the aberration angle in Fig. 2-7.1 would be given by
tan 6 = vic if the Galilean transformation were correct.
2S Visual Appearance of Rapidly Moving Objects When

we discuss the effect of the Lorentz transformation we must be
clear aC>out the meaning of "observing" and "seeing." Up to this
point we have been careful to use the word observe, as in con-
nection with the Lorentz contraction, instead of the word see.
Let us consider in detail the manner in which laboratory ob-
servers determine the length of a moving rod, or the time interval
between flashes of a moving clock. First we must station a set of
observers everywhere in the laboratory frame. Of course, if the
motion is along the x axis, we only need observers along the "
axis. Then we instruct these observers to note their coordinates
and synchronize their clocks. If we wish to measure the length of a
moving rod, we instruct the observers to note whether or not an
end of the rod was at their coordinate position at a predeter-
mined time. The difference in the coordinates of the two observers
who are coincident with an end of the rod is defined as the
length of the moving rod. If we wish to measure the interval be-
tween flashes of a moving clock, as the rod clock, we ask labora-
tory observers to note whether they were coincident with the mov-
ing clock at the time it flashed, and if so what was the reading
on their own clock at the time the moving clock flashed. The dif-
ference between the times read by two laboratory observers is
the laboratory time interval which corresponds to the time inter-
val on the moving clock. This is the meaning of the operations
in the experimental determination of the Lorentz contraction
and of the time-dilation phenomena. This is the meaning of the
word observe.
By seeing, or by photographing, we mean something else. The
difference is most clear in connection with photography. When
we take a photograph, there is only one lens. When we see, there
is only one observer. The difference between observing and
seeing is the difference between an infinity of observers and one
observer.
In the operations we call observing we have taken care that
there is no time required for the passage of a light signal from
the event to the laboratory observer who detects it. The operation
of seeing is quite different. The observer is at one location in the
laboratory frame, and the events he notes may have taken ages
to reach him. A telescope camera photographing distant stars
during a solar eclipse may have its shutter open a very short time.
The light which enters the shutter may come from distant stars,
having been emitted millions of years ago, and during the same
exposure the shutter may admit light from the solar corona
emitted 500 sec ago, and even light from an adjacent clock
emitted nanoseconds ago. We may imagine the camera lens or the
eye to be at the center of a collapsing spherical shell, which col-
lapses at the speed of light and collects events as it goes. If an
event takes place in the time interval during which the shell
passes, information about the event is swept along by the shell
and will be recorded by the camera.
In spite of inferences that might be drawn incorrectly from a
casual inspection of the Lorentz contraction, a careful considera-
tion of the picture-taking process leads to the conclusion that
pictures of rapidly moving distant objects are undistorted in
shape. But the picture taken in the laboratory from one direction
will be the same as obtained in the proper frame, but from a
different vantage point, as we might expect from our discussion
of aberration in 27.
Suppose a cube of edge 1 is moving with speed V in the +x
v
.. Ca)
II
dl(j
AID
(b)
FIG. 2-8.1 Visual appearance of a rapidly moving cube: (a) the mov-
ing cube; (b) a proper photograph from point P; (c) the laboratory
photographer~s picture from P; (d) the laboratory photographer's inter-
pretation of his picture. In (a) the point P' is the point from which the
proper photographer would obtain the same picture as (c), which the
laboratory photographer took from P [After V. F. Weisskopf, Physics
Today 13,24 (1960)].
direction, and that it is photographed from a distant point P (see

Fig. 2-8.la). A camera in the laboratory frame and one in the
proper frame will take different pictures, even when both pictures
are taken from the same point. The two pictures are shown in
Fig. 2-8.lb and c. In the laboratory frame the camera collects
light from the back edge EF as well as from the front face ABeD,
for light emitted from the back edge at a time lie before light
from the front face was emitted will arrive at the camera at the
same time as light from the front face, for all this light will be
swept up in the same collapsing light shell. At that earlier time
the rear edge was at E'F', displaced from the present position EF
by Vlle. By the Lorentz contraction the front face will be con-
tracted so that the edge AB will be measured to be of length 1
(I - V2Ie2)1/2. The laboratory photographer's picture will be as
in Fig. 2-8.lc instead of Fig. 2-8.1b, which is the picture taken by
a proper photographer from point P. But the laboratory pho-
tographer will infer that the cube has rotated through an angle
ep = cos- 1 p, as shown in Fig. 2-8.ld. Note that this is the same
picture that a proper photographer would have taken from a
different orientation, as at P' in Fig. 2-8.la.
Although the discussion above was based on a cube, it is clear
that the apparent rotation must be the same for all rapidly mov-
ing distant objects. Photographs of rapidly moving distant ob-
jects will not show the Lorentz contraction. A spherical rocket
ship will look spherical, as if it were photographed from the side
in the proper frame. The picture that one would obtain in the
proper frame at an angle ()' with V is the same picture which
would be taken in the laboratory at an angle () with V, as in Eq.
2-7.3 for the aberration of light. In the figure we have taken
() = =
7T/2, so that from Eq. 2-7.3 cos ()' -p and ()' is greater than
7T12, as shown.
2-8.1 A small object is in the form of a sphere of radius R in its proper

frame. Determine the shape of the particle (a) as measured (observed)
by a set of observers in the laboratory frame, and (b) as photographed
by a photographer in the laboratory frame when the particle moves at
V = 0.5c in the +x direction with respect to the laboratory. [(a) el-
lipsoidal: 4x 2/3 + y2 + z2 = R2 (b) spherical]
2-8.2 A cube oriented with its edges parallel to the coordinate axes
=
moves in the +x direction at V 0.6c, and a picture is taken when the
cube passes the origin by a distant observer located along the -y axis.
At what angle with respect to V would a photographer in the proper
frame have to be located to take the same picture? [127]
29 Transformation of Acceleration If we take differentials

of the velocity transformation equations (2-5.2), we find that
dU'. V + U'. !::' dU'~, (2-9.1a)

dUx = 1 + VU'./c 2 (1 + VU'./C 2)2 c2
dU dU'".. U'".. V dU' (2-9.1 b)
".' = -y(1 + VU'.,/c 2) -y(1 +VU'./C 2)2 ~ "."
We will restrict ourselves to the case where the particle is in-
stantaneously at rest in the proper frame; that is,
U'. = U'" = U'. = 0,
but the instantaneous acceleration is not necessarily zero in the
proper frame. In this case Eqs. 2-9.1 become
dUx = dU'.,(l - Jl'l/c 2) = dU'.-y-2, (2-9.2a)
and dU" . = dU'" .-y-l. (2-9.2b)
To find the transformation equations for the acceleration we
must divide Eqs. 2-9.2 by an appropriate equation relating time
differentials in the two systems. From the Lorentz transformation
equations (2-1.3) we have
t = -y(t' - Vx' / c2),
and, taking differentials,
dt = -y(dt' - V dx' / c2),
or dt = -y dt' (1 _c!::' dt'dX').
2
dx'
Since we have set ourselves the limitation that V'a: = dt' = 0, the
preceding equation becomes
dt = -y dt'. (2-9.3)
Dividing each of Eqs. 2-9.2 by 2-9.3, we obtain the transformation
equations for the acceleration as
a
a:
= dUx
dt
= -y3
ldU'., =
dt'
'V- 3
,
a'
."
(2-9.4a)
_ dUll . _ 1- dU'lI . _ - 2 ' (2-9.4b)

all,' - ~ - 'Y 2 dt' - 'Y all
Equations 2-9.4 are transformation equations for the accelera-
tion in the special case that the particle is instantaneously at rest
(but may be accelerating) in the primed frame. We will have no
need in this book for the more general equations of transforma-
tion which could be derived from Eqs. 2-9.1.
In the Galilean transformation the acceleration of a particle
is the same in all inertial frames. This is not true at high ve-
locities where the Lorentz transformation applies. Accelerations
in one frame differ in magnitude and direction from the accelera-
tions measured in another inertial frame. This is due to the
Lorentz contraction and to time dilation.
If we make use of the binomial expansion of (I + x)m, as in
Eq. 1-3.2, we find that
'Y- 2 = (1 V2/ c2 ),
V
2
and 'Y- 3 = (1 V2/ c2 ) 3/2 = 1 - 23 ~ + ....
Thus to terms of the second order in V je the Galilean transfor-
mation is correct. So long as we can neglect V2je2 compared to I
we need not concern ourselves with differences in the accelera-
tion as measured on different inertial frames. These terms do
become important in our analysis of the behavior of electrons
and protons in electric and magnetic fields.
It is sometimes convenient to rewrite Eqs. 2-9.4 so as not to
make specific reference to an axis system. We replace the sub-
script x by one which indicates that we are considering a com-
ponent parallel to the motion of the' particle as observed from
the laboratory frame. We replace the subscripts y,z by one which
indicates that the component is perpendicular to the motion.
Thus we may write
all = a'II'Y- 3, (2-9.5a)
and a.L = a'.L'Y- 2 (2-9.5b)
29.1 A particle experiences an acceleration of a' = (3 1",. + 4 Ill' + 12
I z .) cm/sec2 in a coordinate frame in which it is instantaneously at rest.

What is its acceleration as measured in the laboratory, if the instantane-
=
ous proper frame of the particle is moving at V 0.98c I,r with re-
=
spect to the laboratory? [a (0.024 I", + 0.16 II/ + 0.48 I z ) cm/sec 2 ,
=
a 0.51 cm/sec2 ]
2-9.2 A particle experiences an acceleration of a' = (3 I",. + 4 11/')
cm/sec 2 in a coordinate frame in which it is instantaneously at rest.
Find (a) the magnitude of the acceleration, and (b) the direction it
makes with the x' axis. The particle is observed to move in the +x
direction at 0.98c in the laboratory frame. Find (c) the acceleration as
measured in the laboratory, and (d) the angle the acceleration makes
with the x axis in the laboratory frame. [(a) 5 cm/sec2 (b) 37 (c) 0.038
cm/sec 2 (d) 81.4]
R. P. Durbin, H. H. Loar, and W. W. Havens, "The Lifetimes of ... +
and ...- Mesons," Phys. Rev. 88, 179 (1952).
D. H. Frisch and J. H. Smith, "Measurement of Relativistic Time Dila-
tion Using p. Mesons," Am. J. Phys. 31, 342 (1963).
D. H. Frisch, J. H. Smith, and F. L. Friedman, Time Dilation (a film),
Educational Services Inc., Watertown, Mass.
J. Terrell, "Invisibility of Lorentz Contraction," Phys. Rev. 116, 1041
(1959).
V. F. Weisskopf, "The Visual Appearance of Rapidly Moving Objects,"
PhySICS Today 13,24 (1960).
.) Force and Motion
3.1 Introduction For many centuries motions have been

analyzed in terms of forces. Long before Galileo and Newton,
motions were classified as natural or violent. It was thought that
natural motions took place without the application of an outside
force, while violent motions required the application of a force.
Different conditions seemed to prevail on earth and in heaven.
For heavenly bodies the natural state seemed to be circular mo-
tion, for wherever one looked there were circles: the sun was
circular, the moon was bounded by circles, the planets were circu-
lar, and the stars seemed to describe circular paths. On the
surface of the earth, and for horizontal motions, a state of rest
seemed the natural state, for objects came to rest when left by
themselves, while vertical motions like those of falling bodies
seemed natural ones.
The analysis we call Newtonian mechanics is based on a similar
conceptual structure. Motions are analyzed in terms of forces.
But the concept of a natural motion is somewhat more abstract.
Newton's first law expresses the motion a body will describe if
left alone, the natural motion, as one either of rest or of uniform
motion in a straight line. According to Newton there is one
universal natural state, equally valid on earth and in heaven,
equally valid for vertical and for horizontal motions. In New-
tonian mechanics no force is required to maintain an object in
uniform motion in a straight line. It is the departure from uni-
form motion in a straight line that requires a force. Recall that
for earlier formulations a force was required for departure from
a state of rest.
Developments in physics in the past 50 or 75 years have limited
51
the realm of validity of Newtonian mechanics and of the Gali-

lean transformation. Nevertheless, for ordinary experience with
particles we ordinarily encounter (having masses greater than a
microgram, dimensions greater than a micron, and speeds less
than a megameter/sec) this formulation describes the world
with sufficient precision.
The discrepancy between Newtonian mechanics and nature
is remedied by a theory called Quantum Mechanics in the realm
of small objects, and by the Special Theory of Relativity in the
realm of high speeds. Both of these theories are so constructed
that their predictions agree with those of Newtonian mechanics
in the appropriate limit. And this is as it must be, for any physi-
cal theory must provide an accurate and concise description of
experience, or it never finds acceptance. When new theories are
born that seem to upset older ideas, their predictions must al-
ways be in agreement with the older ideas in the realm in which
the older theories are already proven accurate. Such a view was
first expressed by Niels Bohr (1885-1962) in 1923. Known as the
correspondence principle, it may be stated as: Different physical
theories must yield convergent descriptions of the same phenome-
non in their overlapping realms of validity. For example, the
momentum p of a particle of mass m moving with velocity v is
given by relativity theory as
p = m"{v. (3-1.1)
Prior to the special theory of relativity the momentum of a
particle had been given by
p = my. (3-1.2)
Deductions from Eq. 3-1.2 had always been correct until the
advent of radioactivity made it possible to perform experiments
at speeds close to c. Then it was found that the relativistic form
of the expression for momentum was in good agreement with
experiment, while the nonrelativistic form of Eq. 3-1.2 was not.
But in the nonrelativistic limit found by setting c ~ 00, we see
that the two expressions agree, as required by the correspondence
principle.
FORCE AND MOTION 53
Some authors interpret Eqs. 3-U and 3-1.2 to imply that m~
mentum is always expressed as the product of mass by velocity,
but that the mass m of a moving particle is different from its
mass mo when it is at rest. This interpretation is summarized in
the equations
p = mv (3-1.3)
and m = IDo'Y = ma/V1 - 02/C2. (3-1.4)
Calculations resulting from Eqs. 3-1.3 are clearly identical with
those arising from Eq. 3-U, and the choice of interpretation is
one of taste. In this book no distinction will be made between
rest mass and moving mass. The symbol m will always signify the
mass detennined when the particle is at rest in its inertial frame,
such as by comparison with a standard mass through use of an
inertial balance or a beam balance. Eq. 3-U will always be used
to describe the momentum in the relativistic case, with Eq. 3-1.2
as its nonrelativistic limit.
32 Newton's Second Law Newton's second law states: If-a
net force acts on a body, the momentum of the body will be
changed; the rate of change of the momentum is proportional to
the net force and is in the direction of that force.
The second law may be stated mathematically as
F=k~ (3-2.1)
dt'
where k is a scalar constant of proportionality. In the unit sys-
tems generally encountered in physics the constant k is defined
as I, a dimensionless constant, and Eq. 3-2.1 becomes
F=~. (3-2.2)
dt
In the nonrelativistic limit the momentum is given by Eq. 3-1.2
so that
dv
F=m-=ma (3-2.3).
dt
is the nonrelativistic form of Newton's second law.
At speeds comparable to c it is not possible to make the ap-

proximation that momentum is the product of mass times veloc-
ity. Instead we must use the correct relativistic form of the mo-
mentum as given in Eq. 3-1.1 and proceed from Newton's second
law as given in Eq. 3-2.2. Repeating these equations for clarity,
we have
F= t!.e-
dt
mv
and p = m-yv = .
vt - V 2/C 2
Carrying out the indicated differentiation of p with respect to

time t (which is measured in the laboratory frame), we find
d d-y
F = - m-yv = mv -
dt dt
+ m-y dv
-.
dt
(3-2.4)
The first term of Eq. 3-2.4 is in the direction of the velocity vec-
tor. The second term contains the derivative of the velocity vec-
tor with respect to time. Since the change in the velocity vector
may be in any direction, we may resolve that change into a com-
ponent perpendicular to the velocity and a component parallel
to the velocity. Writing the perpendicular component as the per-
pendicular component of the acceleration; that is, writing
(~~t = a~,
we find F~ = m-ya~. (3-2.5)
Before analyzing the component equation describing the com-
ponent parallel to the velocity, let us consider the operations in-
volved in computing ~:. This will involve the computation of
d~. Now v2 = v 'v, the scalar product of the vector v WIt

- 'h'Itse If .
dt
The scalar product of two vectors A and B is written as A Band
is equal to the product of the magnitude of the two vectors and
=
the cosine of the angle between them; thus AB AB cos 8. We
FORCE AND MOTION 55
must make use of a formula for the differentiation of the scalar

product of two vectors, which is easily derived (Problem 3-2.1):
d(AB) = (dA)B + A.(dB). (3-2.6)
Thus d(vv) = 2v(dv),
-dvdt = 2v dv
2
and - = 2v a.
dt
If we write a as the vector sum of its components parallel to the
velocity and perpendicular to the velocity, the scalar product
va may be expressed as the product of the speed v by the paral-
lel component of the acceleration a II' so that
dv 2
7ft = 2vall' (3-2.7)
Applying this result to Eq. 3-2.4, we have

F II = mv(-i)(l - v2/c2)-S/2(-2vall/c2) + m'Ya".
When we collect terms and clear fractions, the equation becomes
F II = m'Y 3a ll' (3-2.8a)

For comparison we repeat the equation for F .1'
FJ. = m'YaJ.' (3-2.8b)
In Eqs. 3-2.8 the coefficients of a are sometimes given special
names. Thus the coefficient of a II' my3, is called the longitudinal
mass, while the coefficient of a .1' my, is called the transverse mass.
The fact that the coefficients of the acceleration are different
for the longitudinal and transverse component equations implies
that at high speeds the acceleration vector is not parallel to the
force vector.
Equation 3-2.8b, predicted by Einstein in 1905, was tested and
confirmed by A. H. Bucherer in 1909, in an experiment in which
f3 rays (electrons) emitted in radioactive decay were deflected by
a magnetic field, where the force on a moving charged particle
is always perpendicular to its direction of motion. Both Eqs. 3-2.8
have subsequently been confirmed many times, and their ac-
curacy is now implicit in the design of charged partide ac-

celerators.
3-2.1 The scalar product obeys the associative and distributive laws of
arithmetic. Thus, for example, (A + B)'C = AC + BC. From the
basic definition of a derivative as a differential quotient, show that
=
d(A'B) (dA)'B + A(dB).
3-3 The Equivalence Principle If we observe that an isolated
object of mass m behaves as if it is experiencing a force F (say.
because it causes a supporting spring to be deflected), then we
find that we can make two possible statements about it. Either
the object is in a gravitational field of field intensity g such that
=
F mg. or it is being accelerated (along with its supporting
spring and the frame to which the spring is attached) with ac-
=
celeration a such that F ma (in the low velocity limit). It is
even possible that the object is experiencing some combination
of gra'\litational field and accelerated motion, as in the case of
a mass hanging from the ceiling of an elevator which is accelerat-
ing upward. But we cannot distinguish between the two causes.
We cannot tell how much of the deflection of the spring is due
to the acceleration of the elevator and how much is due to the
gravitational field. In order to separate the two effects we would
first have had to observe the deflection of the spring when the
elevator was at rest. or when it was moving at constant speed in
a straight line.
This observation. and similarly based experiments performed
with great precision, have led to the concept that accelerated
coordinate frames and gravitational fields are equivalent, a view
summarized in the equivalence principle, which states that it is
impossible to distinguish between an accelerated coordinate
frame and a gravitational field. To quote Einstein. "This assump-
tion of exact physical equivalence makes it impossible for us
to speak of the absolute acceleration of the system of reference,
just as the usual theory of relativity forbids us to talk of the
absolute velocity of a system." The equivalence principle lies at
the foundation of the general theory of relativity.
FORCE AND MOTION 57
The special theory of relativity advises us that it is impossible
for us to build an entirely self-contained speedometer that
measures an absolute velocity. We must measure velocities with
respect to some coordinate frame. We cannot have a velocimeter
which is wholly built in, which makes no contact with an ex-
ternal frame of reference. In an automobile the speedometer
measures motion with respect to the road; in an airplane the air-
speed indicator measures speed with respect to the surrounding
air mass; in a space ship the navigational instruments measure
speed with respect to a planet or the sun.
In the same way the general theory of relativity and the
equivalence principle advise us that it is impossible to build a
wholly self-contained accelerometer. When accelerometers are
used in inertial navigation, as in submarines and satellites, knowl-
edge of the gravitational field must be supplied as independent
input information, for the device itself cannot tell whether it is
being accelerated, or is in a gravitational field, or both.
One interpretation of the word weight is the upward force
exerted on an object by its supports. Thus we are "heavier" when
we stand on the floor of an elevator which is accelerating upward,
and we are "weightless" when the floor of the elevator is falling
at the same acceleration as we would experience in free fall.
Let us consider the weight of a moving object. What is the
weight of a puck which slides horizontally on a smooth floor at
speed v? We interpret the word weight as meaning the force the
floor exerts on the puck, and interpret the problem from the
viewpoint of a set of observers on an inertial frame who see the
equivalent problem, that of a puck sliding on the floor of an
elevator which is accelerating upward in field free space. For
this situation must be identical with the real one where the puck
slides on the fixed floor in a uniform gravitational field. By
Eq. 3-2.8b the inertial observers know that the floor must exert
an upward force on the puck equal to myg, and they, as referees,
declare that this is the weight of the moving puck. Thus we
must conclude that a body moving horizontally in a uniform ver-
tical gravitational field is heavier than it was at rest. If the weight

of the body at rest is mg, then when it is moving horizontally
with speed v, its weight is myg = mgjVl - v:!/c:!.
3-4 Transformation of Forces According to the Galilean
transformation, the acceleration of a particle is the same as
measured in either of two inertial frames. From Eq. 1-2.6 we have
a = a'.
If we multiply both sides of the equation by m, we find
rna = rna'.
Making use of the nonrelativistic form of Newton's second law
in Eq. 3-2.3, we write for each of the two inertial frames that
F = rna and F' = rna'.
Thus we infer that the force acting on a particle is the same in
both frames, that is,
F = F', (3-4.1 )
and we would conclude that the force acting on a particle is
measured to have the same value in every inertial frame so long
as the Galilean transformation is valid.
The situation is different at high speeds, where the Lorentz
transformation is correct. The transformation equations for the
acceleration are given by Eqs. 2-9.5 in the special case that the
particle is instantaneously at rest in the primed frame. These
equations are
(3-4.2)
As before, we imply that the particle is instantaneously at rest in
the primed frame but is accelerating with acceleration a' with
respect to that frame, that the primed frame is moving with
speed v in the +x direction with respect to the unprimed or
laboratory frame, and that observers in the laboratory frame
measure the acceleration to be a with respect to the laboratory
(unprimed) frame.
FORCE AND MOTION 59
If we multiply each of Eqs. 3-4.2 by m, we find

m'Y 3a II = ma'lI (3-4.3a)
and m'Y2a.1 = ma'.1' (3-4.3b)
Let us now compare these results with Eqs. 3-2.8, which we write
as
F II = m'Y 3a ll
and F.1 = m'Ya.1
when referring to the laboratory frame in which the particle is
moving with speed v in the +x direction, and as
F'II = ma'lI
and F'.1 = ma'.1
when referring to the proper frame of the particle, in which the

particle is instantaneously at rest, so that y' =:: 1. Substituting
these equations into Eq. 3-4.3, we find
F II = F'II and 'YF.1 = F'.1' (3-4.4)
The relativistic theory tells us that the forces on a particle are
not the same when observed from different inertial frames, con-
trary to our intuition and to ordinary experience. But then, we
have seen that we cannot infer what to expect at high velocities
from our ordinary experience.
The distinctions that observers in different inertial frames
must make in regard to the forces acting on rapidly moving par-
ticles play a very important role in our understanding of elec-
tricity and magnetism, and particularly in the interplay between
electric and magnetic forces as observed in different frames.
Once again we note that in the nonrelativistic limit C~ 00,
we find y ~ I, and the relativistic force transformation Eqs.
3-4.4 approach the nonrelativistic Eqs. 3-4.1.
3-4.1 The mass of the electron is 9.1 X 10- 28 gm. (a) Find the accelera-
tion of an electron in its proper frame (in which it is instantaneously at
rest) if it is acted on by a force of F' =:: 9.1 X 10- 4 l ...dynes. (b) What
force does an observer in the laboratory frame believe to be acting on
the electron, if he observes it to be moving at v = 0.9998c 1.. with
respect to the laboratory frame? (c) What acceleration does the labora-
tory observer compute the electron to have, by use of the relativistic
relations between mass. and acceleration? (d) What acceleration does
the laboratory observer find for the electron if he translates the accelera-
tion noted by the proper observer through use of the Lorentz transfor-
mation equations for acceleration? [(a) 1024 l z ern/sec2 (b) 1.8 X 10- 5
I z dynes (c) 4 X 1020 I z cmjsec2 (d) 4 X 1020 cm/sec2 ]
3-4.2 The mass of a proton is 1.67 X 10- 27 kgm. The proton experi-
ences a force of (1.67 X 10-9 N) X (I",. + I z.) in its proper frame, in
which it is instantaneously at rest. (a) Find the acceleration of the pro-
ton in this frame. (b) Find the angle between force and acceleration in
this frame. As measured by a laboratory observer for whom the proton
=
is moving with speed corresponding to fJ 0.98 in the +x direction,
what is (c) the force on the proton, (d) the acceleration of the proton,
and (e) the angle between the force and the acceleration? [(a) 1018
mjsec2 X (I",. + I z.) (b) 0 (c) 1.67 X 1O- 9N X (I.. + 0.2 I z) (d) 1018
m/sec2 X (1/125 I.. + 1/25 I z (e) 67.4]
A. Einstein, "On the Generalized Theory of Gravitation," Scientific
American 182, 13 (1950).
H. Bondi, "Relativity," Progress in Physics 22, 97 (1959).
4 Energy and Momentum
4.1 Work The concept of physical work as the result of the
action of a force through a distance developed in mechanics and
has had important repercussions in all of physics, as well as in
other sciences. While the concept of work has physiological ori-
gins and is related to the feeling of tiredness, physical work is
defined in such a way that tiredness is not a measure of work.
If a constant force F is moved through a displacement s, we
say that physical work W (a scalar) has been done, according to
the equation
W = Fs = Fscos9. (4-1.1)
The work is the product of the magnitude of the force times the
magnitude of the displacement by the cosine of the angle between
them (when they are drawn from a common origin). Alterna-
tively, since the product F cos (J is the component of the force
parallel to the displacement, F II , we could rewrite Eq. 4-1.1 as
W = FilS. (4-1.2)
Note that no work is done by any force, however great, unless
there is a displacement. No work is done by any force unless
there is a component of the force in the direction of the displace-
ment.
In the event that the force is variable we first express the work
done when the force is moved through a small displacement ds
as dW, so that
dW = Fds, (4-1.3)
and then integrate, to find the work done as the variable force
traverses a particular path (see Fig. 4-1.1), so that
61
w = /''' Fds. (4-1.4)

'1
When work is done on an object, we find that there are changes
in the speed with which it is moving, or in its shape, or in its
temperature, or in its magnetization, or in some other physical
FIG. 4-1.1 If the force is different along different parts of the path, we
must sum small increments of work, or must integrate, to find the work
done in travening the path from point .. to point 5
property. We sum up these observations by saying that the work

done produces a change in the energ;y of the body. If the change
is' one of speed, we call the change one of kinetic energ;y. If the
change is one of position or of configuration, we call the change
one of potential energ;y. Changes in temperature are associated
with internal energ;y; changes in magnetization are associated
with magnetic energ;y.
Work and energy are measured in many different kinds of
units. Since the concept is initially a mechanical one, the funda-
mental units of energy are based on its definition in mechanics.
Thus we have the erg as the work done by a force of I dyne mov-
ing through a displacement of I cm. The joule is the work done
by a force of I newton passing through a displacement of I meter.
Other units of energy find wide use.
The electron volt is the energy acquired by an electron in
passing through a potential difference of 1 volt. For our purposes
it is sufficient to state that
1 electron volt = 1.602 X 10- 12 erg = 1.602 X 10-19 joule.

ENERGY AND MOMENTUM 63
The calorie is the heat required to raise the temperature of

I gram of water from 14.5 to 15.5C. For our purposes it is suffi-
cient to state that
1 calorie = 4.185 X 107 erg = 4.185 joules.
We will soon see that matter at rest in free space has energy
in the form of mass, and that this energy, measured in units of
mass, may be expressed in units of mechanical energy according
to the equation
E = mc2 (4-1.5)
One atom of a substance whose atomic weight is 1 (1/12 the mass
of a 0 2 atom) has a mass of (6.025 X 1023)-1 gm. The conversion
of this mass unit to energy units is often expressed in million
electron volts (MeV) as
1 atomic mass unit (amu) = 931.5 MeV.
4.2 Kinetic Energy (Nonrelativistic) If a force F acts on a
particle of mass m through a displacement ds, the work done by
the force on the particle is given by Eq. 4-1.3 as dW = Fds, and
we must integrate the elements of work over the path of the
particle to find the total work done on it, as indicated in 4-1.
In general the force is the rate of change of the momentum.
In the nonrelativistic case this statement reduces to the form
dv
F=m-
dt
When a particle moves with velocity v, the displacement ds
that it experiences in a time dt may be expressed as
ds = v dt.
We may combine these statements in the definition of work as
dW = Fds = (m ~;) . (v dt),

and canceling the dt which appears in both numerator and de-
nominator, we find
dW = m (dv) v.
Integrating, we have
W = m Jvdv. (4-2.1)
We have already seen that we may express v2 as v'v, so that
d(v 2) = d(v'v) = vdv + dvv = 2vdv.
Substituting this result into Eq. 4-2.1, taking the initial velocity
to be 0 and the final velocity to be v, and inserting the limits of
integration. we obtain
W = (u ~ d(v2) = ! mv2.
Jo
2 2
We say that a particle acquires kinetic energy T equal to W
as the result of the action of the force through a distance, and
that the kinetic energy of the moving particle is given by the
equation
T = !mv 2 (4-2.2)
The kinetic energy may also be expressed in terms of the mag-
nitude of the momentum; since in the nonrelativistic case p mv,=
we have
T = P:....
2m
(4-2.3)
4.3 Kinetic Energy (Relativistic) Recall that the definition

of work in 4-1 did not depend in any way on the speed of the
particle. Thus we may use the same procedure as before to find
an expression for kinetic energy at high speeds. We repeat the
initial steps of the development of 4-2, but at the appropriate
place we must insert the relativistic expression for momentum
instead of the nonrelativistic one. Thus we begin with
dW = Fds
and note that
ds = vdt,
but now we make use of the relativistic form of Newton's second
law, which states that
F = tiE = dm-yv.
dt dt
The work done in displacing the particle from rest at position
o to a final velocity vat position s is given by
W = Jo(. Fds = Jo(. (dm-yV)

--;It . (v dt)
= ~. [m-y (dv)'v + (d-y)mvv].

Now
d-y = d(l - v2/e 2)-1/2 = (1 - V2/C 2)-3/2 vdv e-2,
as we have seen. Using this result, and recalling that
vdv = ! d (~),
e2 2 c2
we may substitute into the preceding equation for W to find
W = ~ ~. (1 _ ::/e2)1/2 + (1 _ ::/e2)3/2) d (~),

which can be simplified to
W = me 2 (' d (~)
2 Jo (1 - v/e 2 2 )3/2
Performing the integration, we find
W = [(1 _:~~e2)1/2J:
When we substitute in the limits, we find the work done to be
(4-3.1)
Since the second term on the right in Eq. 4-3.1 contains no
velocity, we label the term rest energy, or mass energy, and say
that an object of mass m has total energy E when it is moving
with velocity v, given by
E = m-ye2 (4-3.2)
The difference between the total energy E and the rest energy
me2 is the kinetic energy T. The kinetic energy T is the energy
acquired by the particle as a result of the work done on it in
raising its speed from 0 to v.
Accordingly, the kinetic energy T of a particle of mass m mov-
ing with velocity v is given by the equation
T = E - mc2 = mc 2('Y - 1). (4-3.3)
In the limit of low velocities, where v is much less than c} we
may expand y as a series to find
(4-3.4)
so that
T = mc2(1 + Hv 2/C 2) - 1) = !mv2
As demanded by the correspondence principle, our two expres-
sions for kinetic energy agree in the limit of low velocities.
To find a relation between energy and momentum we note
that
E = m'Yc2 and p = m'Yv,
so that
2 - p2C2 = m2'Y2c4(1 - v2/C 2) = m2c4,
or +
E2 = p2C2 m2c4 (4-3.5)
In Eq. 4-3.5, E and me2 are often expressed in MeV, so that a
convenient unit of momentum is MeV Ie. Incidentally, Eq. 4-3.5
is often remembered through the analogue of a right triangle, in
which E is the hypotenuse.
It is remarkable that the relativistic relationship between en-
ergy and momentum is valid for all particles-even those of zero
mass, like neutrinos or photons.
Electromagnetic radiation (light, radio, y rays) can be described
as a wave motion or as a particle motion. The wave description
utilizes such terms as wavelength '\, say. the distance between
two successive crests of the wave, and the frequency v, the number
of vibrations per second. These quantities are related by the
equation
).v = c, (4-3.6)
all quantities being measured in vacuum. The particle descrip-

tion speaks of electromagnetic radiation as made up of photons
which have discrete amounts of energy. The two descriptions are
related through the equation
E = hv, (4-3.7)
which gives the energy of a single photon of radiation whose
wave description assigns to it a frequency v. Photons of light
have zero mass. Nevertheless the relation between their energy
and their momentum is given by Eq. 4-3.5, for the momentum of
a photon, p, is given by
hv h
p=-=-. (4-3.8)
c A
The direction of the momentum of the photon is identified with
the direction in which the wave is propagated.
In Eqs. 4-3.7 and 4-3.8 the symbol h represents a fundamental
physical constant called Planck's constant, after its discoverer,
Max Planck (1858.1947), which has the value
h = 6.625 X 10-27 erg sec.
Equation 4-3.8 was applied to the behavior of particles by
Louis de Broglie (b. 1892) to yield a wavelength to be associated
with moving particles, as
A = hlp, (4-3.9)
an equation which is now known as the de Broglie equation.
The constant h determines the realm in which classical me-
chanics must blend into quantum mechanics, just as the constant
c determines when classical mechanics must blend into relativity.
The very small numerical value of h implies that particles of
ordinary experience will demand an almost infinitesimally short
wavelength to describe their motion, so short as to be undetect-
able, as for a child's marble whose mass is about 1 gm which is
moving at a speed of a micron per second. These wave properties
are displayed for particles whose mass is of the order of that of a
proton or less. Again, the wave picture of light is to be modified
according to the size of h, for this numerical value determines

whether there is detectable energy in a single photon, or whether
there is detectable momentum in a single photon. If a single
photon cannot be detected, we should never arrive at a particle
conception for light.
4-4 Conservation of Linear Momentum If a single particle
has no force acting on it, there will be no change in its momen-
tum. We would therefore say that the momentum of a single
particle is constant if there is no external force acting on it, and
following the discussion of 3-2, this statement is independent
of the speed of the particle.
Suppose that we have a system of particles, m I , m2 . . . mIl'
FIG. 4-4.1
which act on each other, and experience forces from sources

outside the collection as well, as in Fig. 4-4.1. We will call the
forces which the particles exert on each other internal forces,
and forces which arise from outside the collection will be called
external forces. Let us further suppose that the forces which the
particles of the system exert on one another are equal and oppo-
site, in conformity with the concept of the equality of action and
reaction which is called Newton's Third Law. We use a double
subscript notation to represent internal forces and a single sub-
script notation to represent external forces. Thus, F i ; means the
force exerted by particle i on particle j, while F, means the net
external force on particle i. We apply Newton's second law to
each of the particles of the collection, in turn, as
F1 +F 21 + Fi1 + F..1 =~,

dt
doi
Fi + FH + F... = :...1:..:,
dt
~.. + F 1. . . . . + F i . . . . . + Fn-1,.. = ~.
Let us add these equations, making note of the fact that F'J =
-F ji, so that the sum of the internal forces is zero, to find that
F1 + F2 + F.. + F.. = ded (pI + P2 + p... + p..).

Thus the vector sum of the external forces acting on a collection
of particles is equal to the rate of change of the total linear mo-
mentum of the collection.
If we write
F = I:.F i and P = I:.pi,
dP
then F = dt' (4-4.1)
In the special case that F = 0 we see that P is constant. Thus,

if a system of particles is acted on by zero net external force, the
total linear momentum of the system is constant. This statement
is called the law of conservation of linear momentum.
The law of conservation of linear momentum has been studied
for systems ranging in size from nuclear to astronomical, in col-
lisions between protons, between protons and electrons, and in
everyday experience. No case of its violation has ever been de-
tected.
The concept of a closed system, with the associated ability to
ignore internal forces in dealing with momentum conservation,
gives this concept great applicability.
We may consider the oars of a rowboat and the occupants and
the water around it all part of a system. If the entire system is
initially at rest, its total momentum must remain zero regardless
of the manner in which the occupants manipulate the oars. The
oars scoop small volumes of water backward, and the boat ac-
quires equal forward momentum. Similarly, an airplane acquires
forward momentum by giving air backward momentum, through
the agency of a propeller or a jet engine. A jet airplane has readily
available mass in the surrounding air which it can eject back-
ward, but a rocket ship in outer space must carry along all the
mass which it subsequently ejects as momentum. If a rocket ship
is initially at rest in outer space, the total linear momentum of
the rocket, its exhaust gases, and all its parts must forever be
zero, so long as the rocket ship is acted on by no external forces.
The law of conservation of linear momentum is today regarded
as having universal validity, independent of its derivation from
Newton's laws of motion, though we have here based our discus-
sion on these laws. Any proper formulation of physics must yield
the result that momentum is conserved in closed systems (which
experience no external force), for all deductions which have been
made from the momentum conservation law have found experi-
mental verification, and all pertinent experiments have been
explained by it.
The neutrino was postulated by W. Pauli (1900-1958) in 1931

as a particle of zero charge which could conserve missing energy
and angular momentum (spin) in f3 decay. By 1950 experiment
had shown that the postulated particle could also account for the
missing linear momentum in f3 decay. A decade later the neutrino
was detected and its inferred properties verified.
=
In the preceding development the assertion that F ij -F ji im-
plies that the forces between particles are propagated with infi-
nite speed. To apply this development to the relativistic case,
where the finite speed of propagation of the force fields must be
taken into account, it is customary to attribute momentum to the
field. In this way we arrive at the concept of the momentum in
the electromagnetic field in classical electrodynamics. In 1901
the momentum of electromagnetic waves was verified experi-
mentally by Nichols and Hull, in an experiment which showed
that light exerted a force on a surface on which it fell, in quanti-
tative agreement with the momentum conservation principle.
Today, quantum electrodynamics speaks of electromagnetic radi-
ation as being propagated by photons which have both energy
and momentum, as we will see in our discussion of the Compton
effect, in 5-2.
4-4.1 A positive pion (-Il'+) at rest decays to a positive muon (p,+) and
a neutrino (v). The kinetic energy of the muon has been measured to
be T,. = 4.1 MeV. The mass of the muon is known from other experi-
ments to be 105.7 MeV. Find the mass of the pion. Do this nonrela-
tivistically, and then repeat your calculation relativistically. The ac-
cepted mass of the pion is 139.6 MeV.
4-5 Center of Mass (Nonrelativistic) It is convenient to de-
fine the position vector of the center of mass R of a collection of
particles by the equation
MR = Limiri, (4-5.1)
where M = Lim,
is the total mass of the collection of particles. By differentiating
both sides of Eq. 4-5.1 and noting that
FIG. 4-5.1
dR dr.
-dt =V and -dt = vI,
we find
(4-5.2)
Eq. 4-5.2 tells us that the total momentum of the collection of
particles is the same as the momentum of an imaginary particle
whose mass is the mass of the entire collection, which moves with
the center of mass.
Comparing Eq. 4-5.2 with Eq. 4-4.1, we see that the center of
mass of a collection of particles moves in response to the total
force on the collection just as a single particle would, for we have
dP
F= di' (4-5.3)
If a shrapnel shell bursts in mid-air, the bits of shrapnel fly

about under the influence of internal forces from the explosive,
but the center of mass of the collection of particles continues on
as a projectile, as though the shell never did burst. Of course, the
aerodynamic forces on the explosive gases and the shell fragments
are diff~rent than before, so that the trajectory of the center of
mass is modified from what it would have been if there were no
explosion.
If a rocket ship could be suspended in outer space and its
engines were then started, the center of mass of the particles
which were initially part of the ship would remain at rest for-
ever, so long as no external forces act on any parts of the ship.
It is useful to consider the energy of a collection of particles in
relation to the center of mass. Suppose that the position vector

of the center of mass is R and the position vector of the i'th
particle is r;, when measured from an arbitrary origin, but the
position vector of the i'th particle is r'; when measured from an
origin at the center of mass, as shown in Fig. 4-5.1. We have
ri = R + r'i. (4-5.4)
Differentiating this equation with respect to t, we find
v; =V + v';.
The kinetic energy of the i'th particle is
Ti = tmivi2 = tmiP + tmiv'i2 + miv'i'V,
as measured in the laboratory frame.
Let us add together the kinetic energy of all the particles of
the system to get the total kinetic energy T of the entire system.
We find
T = Li T i = HLi mi)V2 + Ld miv'i2 + [1e (L miT'i)]' V.

The bracketed part of the last term on the right is zero, for, from
Eq. 4-5.1, the part of this term in parentheses serves to define the
position of the center of mass in the primed frame, in which the
center of mass is at the origin, and has position vector zero. We
conclude that the last term on the right is zero and that the total
kinetic energy of the system is
(4-5.5)
As stated by Eq. 4-5.5, the total kinetic enerI!:Y of a collection of
particles about an arbitrary origin is equal to the kinetic enerI!:Y
of its center of mass plus the kinetic enerI!:Y of the collection with
respect to the center of mass.
In the absence of external forces the momentum of the center
of mass is constant, and so therefore must the energy of the center
of mass be constant. The kinetic energy contained in the motion
about the center of mass may vary; that is, the mechanical mo-
tion about the center of mass need not be conserved as mechani
cal kinetic energy.
We call the kinetic energy about the center of mass the

internal kinetic enerf5Y of the collection. The internal kinetic
energy of an ideal gas is characterized by its temperature. For one
mole of ideal gas the total internal kinetic energy is given by
T = % R(), where R is the universal gas constant and () is the
temperature in degrees Kelvin. We know that the temperature of
a real gas may change even though no external forces act on it,
as the gas expands and some of the internal kinetic energy is
exchanged for internal potential energy, as in the case of an ex-
panding refrigerant.
We say that the collisions between particles are perfectly elas-
tic when the internal kinetic energy of the particles (about their
center of mass) is the same before the collision as immediately
after. We say that the collisions between particles are inelastic
when the internal kinetic energy is not conserved. When two cars
or two lumps of putty collide, some of the internal kinetic energy
may go into deforming the objects, doing work against the forces
which hold the body together. A collision is said to be perfectly
inelastic if the bodies move along together with the same velocity
after the collision, as in the case of two balls of putty; that is,
the internal kinetic energy is zero after the collision if the colli-
sion is perfectly inelastic. The collision between two atoms is
inelastic if one or both of the atoms is raised to an excited state
from which it may subsequently radiate light. Thus we are sure
that the atoms in a neon sign are making inelastic collisions, for
otherwise they would not glow.
In all cases, regardless of whether the collisions are elastic or
inelastic, the momentum and energy of the center of mass are
conserved. For this reason physicists often consider collision proc-
esses in the center of mass frame, the frame in which the center
of mass is at rest, for in this frame the total momentum of the
system of particles is zero, and only the internal kinetic energy
of the system need be considered.
To discuss the motion of a collection of particles in the center
of mass frame we would first find the velocity of the center of
mass from Eq. 4-5.2. We would then apply the Galilean transfor-
mation to find the motion of each of the particles of the system

in the cen ter of mass frame, and from this we would find the
momentum and the kinetic energy of each of the particles in this
frame.
4-5.1 A uniform cylinder 8 ft long which weighs 32 lb is subdivided by
a thin internal partition into halves each 4 ft long. The left side A con-
tains 32 Ib of compressed gas. while the right side B is evacuated. The
cylinder rests on its side on a smooth horizontal table. The partition
bursts. When the gas has come to equilibrium inside the cylinder,
(a) what distance has the cylinder been displaced, and (b) with what
velocity is it moving? [(a) I ft (b) 0]
4-5.2 Prove that in an elastic collision between two identical bodies,
one of which is initially at rest, the angle between their velocities after
collision is ,../2, except for the case of central impact. Do this two ways.
Work the problem entirely in the laboratory frame; then work in the
center of mass frame and translate the result to the laboratory frame.
(Nonrelativistic case only)
4-5.3 The energy levels of a hydrogen atom permit it to absorb 10.2
eV of energy as internal energy. (a) What is the smallest kinetic
energy a hydrogen atom must have in order that it may collide inelasti-
cally with a second hydrogen atom at rest? (b) What is the smallest
energy an electron may have in order that it may collide inelastically
with a hydrogen atom at rest? [(a) 20.4 eV (b) 10.2 eV]
4-5.4 Two identical cars collide head on in a perfectly inelastic colli-
sion. In case (a) one car is moving at speed v and the other car is at rest.
In case (b) both cars are moving with the same speed v in opposite di-
rections. Assuming that the damage done is directly proportional to
the energy dissipated inelastically, how much more damage is done in
case (b) than in case (a)? [4X]
4.6 Transformation of Momentum and Energy (Relativistic)
In the proper frame in which a particle is at rest its momentum
is zero, and its energy is mc2 Since the proper frame is moving
with velocity V with respect to the laboratory, the momentum of
the particle as measured in the laboratory frame is given by pz =
= =
myoV, PII,Z 0, and its energy is given by E mYoc2 , where Yo =
(1 - V2/ C2)-1/2.
We might find transformation equations for momentum and
energy by applying the transformation equations for the velocity,

Eqs. 2-5.2, to the equations for the momentum and energy. But
it is more instructive to follow a different procedure.
Recall that the Lorentz transformation equations give time a
role analogous to that of a position coordinate. For this reason
we speak of relativity as being a four-dimensional study. A rela-
tivistic displacement vector must have four components, one for
each dimension of the relativistic space, rather than simply three
components corresponding to the three dimensions of ordinary
space vectors. The relativistic dispIacement is the prototype for
a vector quantity in relativity, just as the three-dimensional dis-
placement is the prototype for a vector quantity in everyday
experience. We will see that the fourth component of the mo-
mentum vector whose three space components constitute the
momentum is Ejc2
Recall that a light flash at the origin, at the instant the primed
and unprimed origins coincide, spreads out as a sphere in both
primed and unprimed frames, as in 21. The equation
x2 + +
y2 Z2 = c2t 2
describes the light sphere in the unprimed frame, and a similar
equation describes it in the primed frame.
A photon on the light sphere has both energy and momentum,
and since the photon is massless these are related, from Eq. 4-3.5,
as
pz2 + PlI 2 + p.2 = Ellc2
in the unprimed frame, and by a similar equation in the primed
frame.
For both sets of equations to be simultaneously true in all in-
ertial frames, momentum and energy must transform in the same
way as spatial displacement and time. Thus we may find the
transformation equations for momentum and energy by replac-
ing x by Pte, Y by PlI' Z by P., and t by Ejc2 in the Lorentz trans-
formation equations, as given in Eqs. 2-1.3. Thus we have
P'II. = PII., PII . = P'II .;

E = 'Yo(E' + VP'.,)
In these equations Yo relates to the motion of the frames in the
=
usual way, for yo-2 I - V 2/c2. The velocity of the particle in
the unprimed frame is v; its momentum in this frame is p myv, =
=
where y-2 I - V2/C2 The velocity of the particle in the primed
frame is v'; its momentum in the primed frame is pi = myv',
=
where y-2 I - v'2/C2.
If these equations are used to transform energy and momen-
tum, then the relativistic relation between energy and momen-
tum given in Eq. 4-3.5 as
W = p2C2 + m c'
2 (4-6.2)
is true for all particles in all inertial frames. Indeed, the
transformation equations were derived on the assumption that
this was so. We may verify that on substitution of Eq. 4-6.1 into
Eq. 4-6.2 we achieve the equation
(4-6.3)
We say that Eq. 4-6.2 is relativistically covariant, for this equa-
tion retains the same form and the symbols retain the same
meaning in all inertial frames.
Consider the relationship between the conservation of linear
momentum and the conservation of energy. In nonrelativistic
mechanics these are independent relationships. But in relativity
the two conservation principles are really one, for momentum
and energy are intimately related through the transformation
equations of 4-6.1. Momentum and energy form a four-vector,
just as do displacement and time. If the total momentum of a
collection of particles is constant in every inertial frame, then
the total energy of that collection is also constant in every iner-
tial frame. This is the immediate implication of the transforma-
tion equations, 4-6.1. Thus if we say ~PID is constant and ~P'ID is

constant, then it follows that ~E is constant and ~E' is constant,
simply by summing the transformation equation for PID for all of
the particles of the collection.
4-6.1 Derive transformation equations for momentum and energy by
applying Eqs. 2-5.2 to Eqs. 3-U and 4-3.2.
4-6.2 Verify the transformation from Eq. 4-6.2 to 4-6.3.
4-6.3 Repeat problem 4-5.2 for the relativistic case.
4-6.4 Show that the speed of a particle is given by v = dEjdp.
4-6.5 Find the greatest angle through which a nonrelativistic a partiCle
can be deflected by striking a stationary proton. [14]
4-6.6 Prove that if kinetic energy is conserved in the collision between
two bodies in one inertial frame, then it is conserved in every inertial
frame. Note that momentum is always conserved in collisions, but that
kinetic energy is only conserved in elastic collisions.
4-7 Center of Mass (Relativistic) Let us develop our argu-
ments about the center of mass and the center of mass frame in a
manner analogous to our development of these ideas in the non-
relativistic case. Suppose we define the center of mass frame as
the inertial frame in which the total momentum of the collection
of particles is zero. But the procedure is somewhat more compli-
cated at high speeds than it was in the nonrelativistic case.
We first find the total momentum of a collection of particles
as observed in the laboratory frame. Let us choose the x direction
as the direction of this momentum vector. Then the y and z
components of the total momentum of the collection of particles
are zero, as seen in the laboratory frame. From the transforma-
tion equations these components of the momentum are zero in
all other inertial frames moving at constant speed in the x
direction with respect to the laboratory frame.
The total x momentum, P'ID' in the center of mass frame is
zero. It is given by
0= pt.. = Edp'..)i = 'Yo [Ei (P..)i - ~ Ei EiJ.

from the transformation equation for P'z in Eqs. 4-6.1. Writing

P for the total momentum of the collection in the laboratory
frame, and E for the total energy in the laboratory frame. we
note that
Li (pz)i = P and Li E i = E.
Solving the preceding equation, we find
c2
V= Jj;P' (4-7.1)
In Eq. 4-7.1 we have written the result in vector form, for the
directions of both quantities are evidently the x direction, but
we no longer need to preserve the manner in which we set up the
coordinate frame. The center of mass frame is moving with
velocity V with respect to the laboratory frame. as given by Eq.
4-7.1.
As in the nonrelativistic case, we imagine that there is a par-
ticle of mass M located at the center of mass. As observed in the
laboratory frame, the momentum of this particle would be set
equal to the total momentum of the particles of the collection.
Thus we have
P = M'YoV, (4-7.2)
so that the mass of the imaginary particle is obtained by substi-
tuting Eq. 4-7.2 into Eq. 4-7.1, as
c2
V = Jj;M'YoV,
or
M = ~.2 (4-7.3)
'YoC
The mass of the equivalent particle is not simply the sum of the
masses of the particles of the collection, but the energy of the
equivalent particle in the laboratory frame is equal to the total
energy of the collection in the laboratory frame. The equivalent
particle has the same momentum as the total momentum of the
collection and the same energy as the total energy of the collec-
tion.
In the center of mass frame the equivalent particle is at rest,

so that Eq. 4-7.3 becomes
E
M=-,2 (4-7.4)
c
where we now must interpret E as the total energy of the collec-
tion of particles in the center of mass frame.
The center of mass frame might be thought to be the proper
frame of the collection of particles. Thus if a tank of gas is at
rest in the laboratory, then the laboratory frame is the frame in
which the center of mass of the collection is at rest, and is the
proper frame for the gas and its container.
If energy is added to the collection of particles by any means
whatever, then we find, from Eq. 4-7.4, that the equivalent mass'
of the collection is also changed according to
ilE
ilM=- (4-7.5)
c2
In the discussion which led to Eq. 4-7.5 we only considered the
rest energy of individual particles and the energy due to their
motion. Experiment has shown that all forms of energy-rest
energy, kinetic energy, and potential energy-affect the equiva-
lent mass of a collection of particles, and Eq. 4-7.5, which defines
the conversion of mass to energy, is a completely general relation-
ship. When a neutron and proton form a stable configuration of
negative potential energy called a deuteron, the mass of the col-
lection is lower than the summed masses of the separate particles
by the amount of the negative potential energy. When a gas is
heated, the molecules of gas acquire kinetic energy, and the mass
of the collection is increased.
The conclusion of Eq. 47.5, that energy may be given to a
system thereby increasing its mass, or that energy may be taken
from a system thereby decreasing its mass, has now been verified
by many different experiments.
A system consisting of two particles, an electron and a positron,
may give up all of its mass in an annihilation process in which
two photons are created.
A system consisting of many protons and neutrons (the nucleus

of an atom) may alter its configuration and form subsystems
(fission fragments). The mass of the fragments is less than the
mass of the parent. To conserve energy the fission fragments must
acquire kinetic energy equivalent to the mass difference. This is
the internal kinetic energy (energy about the center of mass)
which appears as "heat" in nuclear reactors or in atomic bombs.
As we have here defined mass, there can be no mass change
of a fundamental particle, for the mass of a particle is to be
determined when that particle is at rest, as in a beam balance, or
in an inertial balance. The mass of a system of particles may be
altered when that system gains or loses energy. In addition, the
kinetic energy of a rapidly moving particle may be converted to
radiation, or may impart mass to a system, as in the creation of
pions by rapidly moving protons. In all cases which involve
measurable mass changes Eq. 4-7.5 has been verified quantita-
tively.
4-7.1 Neutrons of 3 MeV kinetic energy are scattered by protons. Find
the kinetic energy of the recoil protons at (a) 0 and (b) 45. Angles are
measured relative to the initial direction of the neutron. [(a) 3 MeV
(b) 1.5 MeV]
4-7.2 Radium disintegrates into radon by the emission of an a par-
ticle. Find the kinetic energy of (a) the a particle and (b) the recoil
nucleus. Atomic masses are: radium 226.10309 amu; radon 222.09397
amu; helium 4.00388 amu. [(a) 4.80 MeV (b) 0.08 MeV]
4-7.3 Show that the recoil energy of a nucleus of mass M is hvj2Mc 2
times the energy of an emitted photon. Find the recoil energy of a
CS 137 nucleus when a "I ray of 662 KeV is emitted. [1.70 eV]
4-7.4 Is it possible for a photon to interact with an electron at rest and

give all of its energy to the electron as kinetic energy?
4-7.0 What fraction of the energy of a rapidly moving proton is not

available for inelastic interactions in proton-proton collisions when the
target proton is at rest in the laboratory and the energy of the accelerator
is (a) 3 GeV (b) 7 GeV (c) 25 GeV (d) 200 GeV (e) 1000 GeV? REF.:
G. K. O'Neil, Science 141, 679 (1963). [(a) 0.6 (b) 0.72 (c) 0.8 (d) 0.92
(e) 0.96]
4-7.6 If a proton accelerator were designed with two storage rings in

which two beams of the same energy could be directed toward each
other, all of the kinetic energy with respect to the laboratory would be
available for inelastic collisions, for the laboratory is the center of mass
frame. What energy would be required of a stationary target machine
to produce the same available interaction energy as a colliding pair of
stored beams with an energy of (a) 3 GeV (b) 25 GeV (c) 31 GeV? REF.:
G. K. O'Neil, Science 141, 679 (1963). [(a) 31 GeV (b) 1360 GeV
(c) 2040 GeV]
4-7.7 A chemical balance customarily weighs to a sensitivity of about
0.1 mg. How much energy in ergs would be liberated in a chemical
reaction if the change in mass were detectable in a weighing? [9 X 1016
ergs]
4-7.8 The energy required to ionize a helium atom is 24.6 eV. If
helium gas is bombarded with charged particles, at what incident energy
will ionization of helium just be observed if the projectiles are (a) a
particles (b) protons (c) electrons? [(a) 49.2 eV (b) 30.8 eV (c) 24.6 eV]
W. Bertozzi, The Ultimate Speed, An exploration with high energy
electrons (a film), Educational Services Inc., Watertown, Mass.
W. Bertozzi, "Speed and Kinetic Energy of Relativistic Electrons," Am.
J. Phys. 32,551 (1964).
P. Morrison, "The Neutrino," Scientific American 194,59 (1956).
R. E. Pollock, "Resonant Detection of Light Pressure by a Torsion
Pendulum in Air-An Experiment for Underclass Laboratories,"
Am. J. Phys. 31, 901 (1963).
F. Reines and C. L. Cowan, Jr., "The Neutrino," Nature 178, 446
(1956).
5 Some Relativistic Phenomena
5-1 Pair Production and Annihilation If energy could be
converted from one form to another without inhibition, we
might wonder how any sort of stability might be achieved in the
universe. Clearly there must be some restraint on the free inter-
change of energy. In the formulation of physics such restraints
often take the form of conservation principles. One of these is
the law of con$ervatjon of charge, which states that the total elec-
tric charge content of the universe must remain constant. Thus
a single electron cannot vanish, giving its energy to photons. But
a positron is a positively charged particle, identical in every other
way to the electron. An electron-positron pair may vanish or be
created without violating the charge conservation principle.
The mass of an electron, expressed in energy units, is 0.5108
MeV. An electron-positron pair at rest in the laboratory may
annihilate with the emission of a pair of photons of equal en-
ergy which move in opposite directions, so as to conserve linear
momentum. The energy of each photon is given by the ex-
pression
hv = mc 2 = 0.5108 MeV. (5-1.1 )
Thus the wavelength of annihilation radiation is
~ = h/mc. (5-1.2)
Npte that the electron and positron are themselves stable par-
ticles.
Positrons are emitted in the radioactive decay of some arti-
ficially radioactive nuclei, such as in the reaction 29CU64 ~
2sNi64 + p+. When positrons from such a source impinge on a
place where there are many electrons, as in a block of copper,
83
they rapidly come to rest. The range of these positrons in copper

is about 300,.... On measurement we find that monochromatic
annihilation radiation is produced whose wavelength is 0.2426 X
10-9 em. Other experiments have shown that two photons are
emitted simultaneously in opposite directions.
In studying the inverse process we find that a single photon
cannot materialize an electron-positron pair in free space, regard-
less of its energy. Energy conservation alone is not enough.
Momentum needs to be conserved also, and momentum could
not be conserved in such a process. We may easily see this by
considering the process of pair production in the center of m~ss
frame of the newly created electron-positron pair. The momen-
tum of the pair is zero in this frame, but since the velocity of
light is c in every frame, the momentum of the photon cannot
be zero.
Momentum conservation forbids pair production in free space,
but pair production can take place in the neighborhood of a
Molybdenum
Kaline
Scottered
ot900
FIG. 5-2.1 Wavelength alterations of the modified line in the Compton

effed at 90 and 135. The unmodified line is also shown [After A. H.
Compton, Am. J. Phys. 29, 817 (1961)].
SOME RELATIVISTIC PHENOMENA 85
third particle whose recoil can conserve momentum. Pair pro-

duction commonly takes place in the field of a nuCleus. The very
large mass of a nucleus enables it to conserve momentum though
moving at low speeds, so that the energy it requires is below the
limit of measurement.
To conserve energy the pair production process cannot take
place at photon energies less than the mass of the particles. We
Primary photon
(0)
S<Ot~
p=~,
8
Recoil electron
4>.
p.~ Primary photon
(b)
FIG. 5-2.2
observe that the threshold energy for pair production is 1.02

MeV, in agreement with conservation principles.
5-2 Compton Effect In 1922 A. H. Compton (1892-1962)
utilized the concept of an elastic collision between a photon and
an electron to explain his observations of the scattering of x-rays.
Compton found that x-rays scattered from a block of graphite
were not all of the same wa,velength as the incident radiation,
as shown in Fig. 5-2.1. We apply the momentum conservation
conditions, using the notation of Fig. 5-2.2 but taking E and p
as the energy and momentum of the recoil electrOQ, to obtain
(5-2.1 )
or
h2e2
P2e2 - -2 (A 2
= A'2
A
+ A'2 - 2AA' cos ."
A..) . (5-2.2)
The energy conservation condition yields

he he
E - + me 2,
= -X - A' (5-2.3)
he
or E = - (A' - A)
X'A
+ me '2 (5-2.4)
which may be squared to give

2 2
E2 = h e (A'2 + X2) 2 2
_ 2h e + 2hme 3 (X' - A) + 24
X'2X2 X'X X'A me.
Making use of the relativistic relation between energy and mo-
mentum, Eq. 4-3.5, we find that
h
A' - X = - (1 - cos 4. (5-2.5)
me
The quantity h/mc = 0.02426 X 10- 8 cm is known as the Comp-
ton wavelength, for all photons scattered through Tr/2 have at
least this wavelength. It is also the wavelength of annihilation
radiation, Eq. 5-1.2.
The Compton effect is the fundamental process through which
electromagnetic radiation is elastically scattered by matter. The
radiation interacts with electrons in this scattering process. When
the recoil energy of the electron is sufficiently great that the elec-
tron is disengaged from the parent atom or the parent lattice,
then the electron is free, and a shift of wavelength of the scat-
tered radiation occurs. But even at relatively high x-ray energies
the recoil momentum is taken up by the entire atom or the
entire lattice a significant part of the time. For this reason the
scattered radiation is sometimes unaltered in wavelength, since
a heavy atom or a lattice can absorb an appreciable amount of
momentum without acquiring significant kinetic energy. When
light or radio waves are scattered by matter, the recoil energy
of the electron is so small that we must think of it as bound.
The recoil momentum is always absorbed by the parent atom or
lattice, and there is no detectable shift in wavelength in elastic
scattering processes at low energy.
At low energy an appropriate analysis may be made by consider-
ing that the eRect of the passing electromagnetic wave on an
electron is through its oscillating electric field. The electron vi-
brates at the frequency of the driving field, and radiates energy
as a result of its acceleration. The radiated energy is of the same
frequency as the incident field. This approach to the scattering
problem was made by J. J. Thomson and is known as Thomson
scattering, a form of analysis which may be regarded as the low-
energy limit of the Compton effect.
5-2.1 Determine the change in wavelength produced when an x-ray
photon is scattered by an electron through an angle of (a) 90 (b) 180.
[(a) 0.242 X 10- 8 em (b) 0.0484 X 10- 8 em]
5-2.2 X-rays of wavelength 0.5 X 10- 8 em are scattered by free elec-
trons in a block of carbon, through 90. Find the momentum of the
(a) incident photons, (b) scattered photons, (c) recoil electrons. and
(d) the energy of the recoil electrons. [(a) 1.32 X 10- 18 gm em/sec
(b) 1.26 XIO-18 gm em/sec (c) 1.82 X 10- 18 gm em/sec (d) 1.92 X 10- 9
erg]
5-3 Threshold for 7To Meson Production In the inelastic
collisions encountered in high-energy physics, new particles are
produced by the conversion of internal kinetic energy into
matter. Because of the requirements of the law of conservation
of momentum, a bombarding particle must have considerably
more kinetic energy in the laboratory frame than is superficially
implied by the mass of the newly manufactured particle. This
is because the momentum and energy of the center of mass must
be .conserved. Only the internal energy in a collision may be
converted into matter; this is the kinetic energy of the particles
in the center of mass frame.
Let us consider the energy threshold for the production of
uncharged 7To mesons, whose mass is 264.4 times the mass of an
electron, in the collisions of protons on protons. The reaction

may be described symbolically as
p+ p -+ p+ p + '11"0.
In the center of mass frame all of the product particles will be

at rest at threshold; that is, they will have no internal kinetic
energy-the collision is perfectly inelastic, with all the kinetic
energy being converted to another form.
In the laboratory frame the particles will all be moving with
the same velocity V as the center of mass at threshold. To find
V we note that before the collision only the proton from the
accelerator is moving. The proton of the target, say liquid hy-
drogen, is at rest. The total energy E and momentum p before
collision are
and
p = mp-yv,
where
-y-2 = 1 - V2/C 2
From Eq. 4-7.1 the velocity V of the center of mass is given by
c2 -y
V=:E P =-y+1 v .
The center of mass motion (and the motion of the reaction prod-
ucts at threshold) is characterized by Yo, where
-Yo-2 = 1 - V 2/C 2 = 1 +2 -y'

The energy of the reaction products in the laboratory frame
after the collision is given by
(E1ab)after = (2m p + m...)c2-yo.
Equating the energy of the system in the laboratory frame before
and after the collision, we find
mpC2(-y + 1) = (2m p + m...)c2 Ci -yy.

Squaring both sides of the equation and clearing terms, we find

2(1 + -y) = (2 + m.fm )2, p
or -y = H2 + m./m )2 - p 1.
By Eq. 4-3.3, the kinetic energy of the incident proton at this
value of y is
2
T = mpe2(-y _ 1) = mp c (4
2
+ 4 m.,
m
+ m.,:)
m
- 2mpe2.
p p
Simplifying, we find
T = mc
-,
(2 + 2mm.,).
2
p
(5-3.1)
Since m p = =
1836 me and m ...o 264 me> we find that the threshold
in kinetic energy at which an accelerated proton beam will begin
to produce 71"0 mesons in a hydrogen target will be T = 2.24
=
m ...oc2 301 MeV. The energy required in the beam is close to
2t times the mass energy of the 71"0 meson.
5.3.1 Find the threshold for (a) the reaction
p+p-+Ao+Xo.
the neutral lambda (Ao) particle and its antiparticle each have a mass
of 2182 me; (b) the production of a proton-antiproton (p) pair in the
reaction
p +- p -+ p + p + p + p,
where accelerated protons strike a hydrogen target and convert their
kinetic energy into the proton-antiproton pair. [(a) 763 MeV (b) 5.6
GeV]
5-4 Doppler Effect Again the universality of the Lorentz trans-
formation equations may be illustrated by applying the relativis-
tic momentum-energy transformation, Eqs. 4-6.1, to the behavior
of a photon.
Suppose a photon is observed in a frame in which the source
of the light is at rest. We call this the primed frame. In the
primed frame the frequency of the light is V, and the light is
observed to make an angle 8' with the x' axis. The momentum
and energy of the photon may then be described by the equations
p",, hll'
= -cos6,
,
(5-4.1)
c
, hll'. 6'
PII=-sm,
c
p'. = 0,
and E' = hll'.
Let us find the momentum and energy of the photon as meas-
ured in the laboratory frame by use of Eqs. 4-6.1. In the labora-
tory frame the photon would be measured to have frequency v
and to be moving with an angle 6 with respect to the x axis so
that
p", = -hll - cos 6' + V -hll') ,

cos 6 = 'Yo (hll' (5-4.2)
c c c'
. 6 = -hll'.
PII = -hll Sin sin 6' ,
c c
P. = 0,
E = hll = 'Yo ( hll' + V h~
-; cos 6' ) .
Simplifying these equations, we find

11 cos 6 = 'YOII' (cos 6' + VI c), (5-4.3a)
11 sin 6 = 11' sin 6', (5-4.3b)
and JI = 'YoII' (1 + ~ cos 6')- (5-4.3c)
If we divide Eq. 5-4.3a by 5-4.3c, we find
cos 6' + Vic

cos 6 = 1 + (Vic) cos 6" (5-4.4)
We have encountered Eq. 5-4.4 before, in the relativistic expres-
sion for aberration, Eq. 2-7.3, which was derived from the veloc-
ity addition formulas.
We have achieved an additional result as well. Not only will a
light ray be observed to travel in different directions in two
different inertial frames, but the frequency of the light will be

measured differently by observers in the two frames. In the
proper frame of the source the light will have the same fre-
quency when proper observers measure it, whatever their posi-
tion. In the laboratory frame, however, the frequency will de-
pend on the angle () which the light makes with the direction in
which its source is observed to move. This is known as the
Doppler effect, originally discovered in sound, and interpreted
as due to the motion of the source or the observet with respect
to the medium through which the sound was transmitted, the air.
Even though there is no preferred medium for the transmission
of light through a vacuum, nevertheless we still find a relativistic
Doppler effect, due to time dilation and to the Lorentz contrac-
tion.
If in the laboratory frame the light is seen from 1r/2 to the
=
direction of motion, then () 1r/2, and from Eq. 5-4.4, cos ()' =
- VIe. Putting this result into Eq. 5-4.3c, we find that
'Yov = v'. (5-4.5)
This is called the transverse Doppler effect, for the observation
is transverse to the direction of motion.
If the light is observed in the direction of motion, then () = O.
From Eq. 5-4.4 we find ()' = 0, and
_ ,(1 + VicI C)1I2.'

II - II 1 _
V (5-4.6)
that is, the light is shifted to the blue. When ()' = 1r, the plus and
minus signs in Eq. 5-4.6 are interchanged, and the light is shifted
to the red. This is known as the longitudinal Doppler effect.
By means of the longitudinal Doppler effect the speed of rota-
tion of distant double stars may be determined, for when one
star is approaching the earth the other is receding, and the spec-
tral lines of light from hydrogen atoms in the star pair become
measurably broadened as their frequency is Doppler shifted.
Even though isolated atoms emit light which is nearly mono-
chromatic, the light from a hot gas of atoms will be more diffuse
92 AN INTRODUCTION TO THE ltPECIAL THEORY OF RELATIVITY
in frequency because of the Doppler shifts due to the motion of

atoms toward and away from the spectrometer. This is known
as the Doppler broadening of spectral lines.
54.1 The wavelength of the a line of hydrogen is 6561.01 AO (lAO =
10- 8 cm). What wiII be the measured wavelength of the Her. line in the
absorption spectrum of a receding star moving at 3 X 106 m/sec away
from the earth? [6627AO]
54.2 Calculate the width of Her. lines due to Doppler broadening at

6000 0 K. Assume that the maximum speed with which hydrogen atoms
approach or recede from the spectrometer is estimated from the kinetic
theory of gases to be given by the equation tmv2 = IkT (k = 1.3~ X
10- 23 joule/molecule _OK). [0.3AO]
54.3 The H" lines measured on earth from opposite ends of the sun's
equator differ in wavelength by 0.091A 0. If the solar diameter is 865,000
miles, find the period of the rotation of this material of the sun, assum-
ing rotation to be the cause of the effect. [25 days]
54.4. An electron-positron pair annihilates in free space, emitting two
photons, each of energy 0.5108 MeV, in opposite directions in the
=
proper frame. If this frame moves at V 0.9c with respect to the
laboratory, and if the photons are emitted at an angle", with respect to
V, find the angle made by each of the photons with V and the energy
of each of the two photons as determined in the laboratory, for", (a)=
0 (b) 37 (c) 53 (d) 90. (e) If an apparatus is set up in the laboratory
which is capable of detecting two photons of the same energy coming
from different directions and of determining whether they were emitted
simultaneously, for what photon energy (determined in the laboratory)
should we expect to measure coincidences in the laboratory? [(a) 0',
2.21 MeV; 180, 0.12 MeV (b) 19, 2.1 MeV; _60, 0.24 MeV (c) 21,
1.8 MeV; -47',0.46 MeV (d) 90', 1.16 MeV; _90', 1.16 MeV (e) 1.16
MeV]
5.5 Relativistic Dynamics We recall that there are many in-

stances in which the time rate of doing work is of great impor-
tance. This quantity is called the power, expressed in units of
work divided by units of time-as ergs/sec, in the cgs system of
units, or as watts (= joules/sec) in the mks system of units.
The average power is defined as the work done, aw, divided
by the time interval in which it is done, or
- .1W
<P = .1t.
Applying the definition of work as the scalar product of force and
displacement, we have
- F.1s
<P = --:1t'
which becomes
<P = Fv (5-5.1)
in the limit of short times. Thus, the instantaneous power de-
livered by a force F to a body moving with velocity v is the
scalar product of the force and the velocity.
We are now able to complete the extension of Newton's second
law to rapidly moving systems. In our initial discussion of the
second law in Chapter 3, we took the momentum to be a space
=
vector, p myv, with three space components. Later, in 4.6, we
saw that momentum and total energy together make up a four-
vector whose components are
PI = p" = m",/v", (5-5.2a)
P2 = Pl/ = m",/vl/, (5-5.2b)
Pa = P. = m",/v., (5-5.2c)
P4 = E/c 2 = m",/. (5-5.2d)
Following the pattern of Newton's second law, where the space
parts of the equations related the force to the time rate of change
of the momentum, we ask: What is the time rate of change of
the fourth component of the momentum equal to? The time rate
of change of energy is power. We therefore infer that the missing
fourth component equation from the relativistic form of New-
ton's second law is that the power is equal to the time rate of
change of the energy, in the form
<P = Fv = did m",/c 2 (5-5.3)
Let us write the four component equations of Newton's second

law. We have
d
F. = de m'Yv"" (5-5.4a)
d
F" = de m'Y v", (5-5.4b)
d
F. = de m'Y v., (5-5.4c)
d
Fv = - m'Yc 2 (5-5.4d)
dt
In Eqs. 5-5.2 a particle of mass m moves with velocity v with
respect to an inertial frame, say the laboratory frame. In this
frame F is the force on the particle and t is the time measured
by laboratory observers using their laboratory clocks.
We wish to put Eqs. 5-5.4 in covariant form, in which every
quantity is a four-vector, or an invariant quantity which has the
Same meaning in every inertial frame. First we note that the
fourth component of the momentum.is not myc2 , but according
to Eq. 5-5.2d is my. Therefore we first divide Eq. 5-5.4d by c2
The next step is to multiply each equation by y, for reasons that
will be made clear soon, to obtain
d
'Y F",." . = l' de m'Y v"""
from Eqs. 5-5.4 a-c and
l' d
-Fv = '"I -m'"
c2 dt "
from Eq. 5-5.4d.
We recall that moving clocks run slow by a factor y (5-1). If
a time interval dt is measured by a laboratory clock, laboratory
observers will note that the time interval on a clock moving
through the laboratory will be dT in the same period, related by
the equation
dt = 'YdT, (5-5.5)
which we have seen before, in another notation, as Eq. 2-3.1. The
equations take on a simpler form if, instead of measuring time on
the laboratory clocks, we measure it on a succession of clocks in

different inertial frames, each of which is selected at different
points along the path of the particle to be the proper clock for the
particle at that instant. The time T will always be the instantane-
ous proper time, measured on a clock which is instantaneously at
rest with respect to the particle.
Let us substitute for dt into the above equations, from Eq.
5-5.5, to obtain
d
'YF., = dT m'Y v." (S-S.6a)
(S-S.6b)
d
'YF = dT m'Y v., (S-S.6c)
'Y
-Fov d
= -m'Y. (S-S.6d)
c2 dT
Equations 5-5.6 differ from Eqs. 5-5.4 only in form, through
algebraic manipulation, but there is a significant difference. Every
quantity on the right-hand side of the equation is either inde-
pendent of the inertial frame of the observer, as is dT, or is a com-
ponent of a four-vector whose transformation from one frame to
another inertial frame is accomplished by the Lorentz transfor-
mation equations.
This implies that the quantities on the left-hand side of the
equations also are components of a four-vector, which we call the
Minkowski force, after Hermann Minkowski (1864-1909), who
first discovered this formulation of relativistic dynamics. We find
it attractive to write Newton's second law in this generalized rela-
tivistic form as
(5-5.7)
where K, is the i'th component of the generalized force whose

four components are expressed as
K 1 = 'YF" (5-5.8a)
K2 = -yF7I , (5-5.8b)
K 3 = -yF" (5-5.8e)
-y
K. = """iFv. (5-5.8d)
c
Equation 5-5.7 is covariant. It has the same form in every in-
ertial frame. Further, the components of K and p measured in
one inertial frame are related to the components of K' and p'
(the same quantities measured by observers in another inertial
frame) by the Lorentz transformation.
The expression of the generalized force now brings the number
of relativistic four-vectors we have seen to a total of three. We
first saw the Lorentz transformation in connection with the posi-
tion four-vector whose components were x, y, z, and t. Next we
displayed the momentum four-vector, and finally the generalized
force four-vector.
In general, it is the attempt of the special theory of relativity
to express all physical vector quantities in four-vector form, for
in this form we can see clearly how to relate measurements made
of these quantities in different inertial frames. This means of
expression often displays hitherto unsuspected relatedness, as we
saw in connection with the union of momentum and energy in
the momentum four-vector. In all cases the components of a four-
vector must transform exactly as x, y, z, and t do in the Lorentz
transformation.
5-5.1 Show that the application of the Lorentz transformation to the
Minkowski force, Eqs. 5-5.8, yields results in agreement with Eqs. 3-3.4.
5-6 Gravitation and Light We may understand some of the
influence of gravitation on light through our studies of the
equivalence principle in 3-3 and our studies of the Doppler
effect in 5-4. To do so we must extend the results of the Special
Theory of Relativity (which deals with observations made on
different inertial frames) to observations made on accelerated
coordinate frames. We assert that:
1) Acceleration in itself does not affect the rate of a standard
dock or the length of a standard measuring stick, or if it does,
the effect may be disregarded in any measurements we are yet
able to perform; and
2) Acceleration does not affect the speed of light in vacuum.
We believe these assertions to be reasonable because we would
expect that any effect of acceleration would depend on its
magnitude. We can always reduce the acceleration to some low
value where the assertions would be true.
These assertions amount to the statement that we will regard
the Lorentz transformation as instantaneously correct when we
use it to compare observations made on an accelerated frame with
those made on an inertial frame. Observations of length, time,
and the velocity of light made on an inertial frame which is
instantaneously at rest with respect to the accelerated frame will
agree with similar observations made on the accelerated frame.
Suppose that the accelerated frame is moving in the +x direc-
tion with respect to an inertial frame with acceleration a, and
that there is a source S at the origin of this frame and a detector
D at some coordinate position x in this frame, along its x axis.
A pulse of light emitted from the source at t = 0 is received by
=
the detector at t xle later, at which time the detector is moving
=
at at axle faster than the source was moving when the light
was emitted.
Observers located at the detector in the accelerated frame must
agree with observers on an inertial frame instantaneously at rest
with respect to them about measurements of length and time,
and so they must also agree about measurements of frequency.
At the instant of detection these observers agree that the source
is receding from the detector with speed axle, and that the angle
between the direction of motion of the source and the detected
photon is 7f'. If the proper frequency of the photon is vo, as
measured at the source, observers at the detector must find the
frequency Doppler shifted to the red to some new value v.
In the notation of Eqs. 5-4.3 and 5-4.4, the angle (J =.". =()',
=
the proper frequency of the photon is Yo v', and the detected
frequency is v. Note that the speed of the proper frame with
=
respect to the laboratory frame is V axle and that Yo-2 =1 -
V21c2. Then according to Eq. 5-4.3c we have
II = 1101'0(1 - Vic)
(1 - Vic)
= 110 (1 - V2 I c2)1/2
_ [1 +- VVicI C]1/2
- 110 1
= 110(1 - 2Vlc + )1/2

= 110(1 - Vic + )
II = 110(1 - axlc 2 + ). (5-6.1)
Thus in an accelerated coordinate frame, light moving in the
direction of the acceleration will be shifted to the red. By the
same argument light moving opposite to the direction of the
acceleration will be shifted to the blue.
By the equivalence principle, an accelerated frame is equiva-
lent to a uniform gravitational field. Thus Eq. 5-6.1 is applicable
to the case where there is a source at the origin and a detector at
x, on an inertial frame in a uniform gravitational field of magni-
tude g = a, but directed in the -x direction. When a light source
which emits photons of frequency Vo is placed at the bottom of a
mine shaft, it will be measured, at the top of the shaft, to have
emitted light of frequency v, given by
II = 110(1 - gxlc 2). (5-6.2)
This result, which is known as the gravitational red shift, has
been observed in sunlight and in a shaft 22 meters high by means
of a very sharp source of radiation and a correspondingly sharp
spectrometer. The observation was made by Pound and Rebka
[Phys. Rev. Letters 3, 439 (1959)] by exploiting the recoilless
emission and absorption of nuclear y rays, in the Mossbauer
effect, discovered in 1958.
As noted from the top of a mine shaft, vibrating atoms at the
bottom of the shaft vibrate more slowly (glow redder) than vi-
brating atoms at the top. Ob~ervers at the top of a shaft must
note that clocks at the bottom run slow. Since this is a relativistic
phenomenon and has nothing to do with the construction of a
particular atom or clock, we must conclude that all clocks deep
in a mine shaft run more slowly than clocks at the top of the
shaft. This observation includes biological clocks as well as other
kinds. .
While the gravitational red shift may be regarded as being well
substantiated experimentally, a second gravitational effect on
light, the deflection of a ray of light in a gravitational field, has
not yet been fully confirmed quantitatively. Nevertheless, it is
instructive to examine the behavior of a light ray in a uniform
gravitational field by application of the equivalence principle
and aberration. Once again we proceed by studying the accel-
erated coordinate frame.
We wish first to study the path of a ray of light in an acceler-
ated frame (the unprimed frame) which moves with acceleration
a and instantaneous velocity V in the +x direction with respect
to an inertial frame (the primed frame) in whiCh the source was
at rest at the instant the light was emitted. Equivalently,. we may
say that the primed frame moves with respect to the unprimed
frame with speed - V at the time t when we examine the motion
of the light. By Eq. 5-4.4, the light will be seen to move in two
different directions in the two frames, as given by
- Vic
cos 8 = cos 8' . (5-6.3)
1 - !:::cos8'
c
A ray of light directed along the y' axis of the inertial frame
.continues to move in that direction with velocity c, and the angle
8' is always 7r12. But this is not the case in the accelerated frame.
for as the frame moves with increasing speed with respect to the
inertial frame, the angle 8 which the ray makes with the x axis
of the accelerated frame is given by
cos 8 = -Vic, (5-6.4)
and since
V = at,
we find that
cos (J = -at/c. (5-6.5)
(Since the aberration effect is of first order in V /c, and any effects
of time dilation are of second order in V /c, we need not concern
ourselves with any differences between time measured in the
accelerated frame and time measured in the inertial frame.)
It'
II
.,,/2
(a) (b)
FIG. 5-6.1 (a) The path of a ray of light in an inertial frame is a
straight line in field free space. Here the primed frame is the frame in
which th, source was instantaneously at rest at the instant the light
was emitted. (b) The path of the ray of light observed in the accelerated
frame (unprimed) is curved, deflected from the y axis in a direction
opposite to the acceleration. (On an inertial frame in a uniform gravi-
tational field the light is deflected in the direction of the field.)
Even though the velocity of light in the accelerated frame is a

constant, c, nevertheless the path of a ray of light is not a straight
line, but is curved, as in Fig. 5-6.1. To find the path of the ray
we note that
sin (J = (1 - V2/C2)1I2, and
(1- V2/ C2)1/2
tan (J = _ Vic = _(C 2/V2 - 1)112 = -(c2/a 2t 2 - 1)112.
Thus the slope of the light ray is negative.

SOME lELATIVISTIC PHENOMENA 101
Since the speed of light is c, the x component of its velocity is

e", = e cos 8 = - V,
and the y component is
ell = e sin 8 = e(l - f'J/e 2)1/2 = e(l - a2t 2/e 2)!/2.
Then
The path of the light ray may be described through the equations
x= Je", dt = - JV dt = - Jat dt = - !at 2

(S-6.6a)
and y = Je(l - a t /2e dt""" et
2 2 2) (S-6.6b)
for small a, or where at is much less than c.

To the degree of approximation shown, the light ray follows a
parabolic path, as observed in the accelerated (unprimed) frame.
By the equivalence principle the behavior of the light ray in
the accelerated frame is equivalent to the behavior of the ray in
an inertial frame placed in a uniform gravitational field. Thus,
in first approximation, a ray of light emitted in a uniform gravi-
tational field in a direction perpendicular to the field follows the
same path we would expect nonrelativistically for a ball projected
with velocity c horizontally in a vertical gravitational field.
A ray of light which follows a straight-line path in field free
space will be deflected in the direction of the gravitational field
intensity when the ray of light is observed from an inertial frame
in a uniform gravitational field. The deflection of the path is
through the aberration phenomenon, but the result obtained is
as if the photon experienced free fall.
A. A. Bartlett, "Compton Effect," Am. J. Phys. 32,120 (1964).
A. J. F. Boyle and H. E. Hall, "The Mossbauer Effect," Progress in
Physics 25, 442 (1962).
A. H. Compton, "The Scattering of Xrays as Particles," Am. J. Phys. 29,

817 (1961).
K. W. Ford, The World of Elementary Particles, Blaisdell Publishing
Co., New York, 1963.
D. H. Frisch and A. M. Thorndike, Elementary Particles, Momentum
Book I, D. Van Nostrand Company, Inc., Princeton, N.J., 1964.
H. Lustig, "The Mossbauer Effect," Am. J. Phys. 29, 1 (1961).
6 Relativity and
Electromagnetism
6.1 Introduction The original structure of the Faraday-
Maxwell theory of electricity and magnetism was built on ex-
perimental observation based on the use of charged test bodies as
the probe for examining the electric field, and the compass
needle as the probe for examining the magnetic field. As an
example, the magnetic effect of current was discovered by Oersted
through the deflection of a compass needle on initiation of cur-
rent flow in a nearby wire. In the present discussion we will
retain that structure, and will imagine that we have at. our dis-
posal magnetic poles, and currents of magnetic poles through
which we may generate fields or probe fields. By this means we
expose the fundamental conceptual structure and symmetry of
electricity and magnetism, and can develop its inherent rela-
tivity. Nevertheless, we must understand that magnetic effects are
generally regarded as produced only by moving charges, since
no free magnetic pole has been discovered.
The mathematical formulation is most conveniently expressed
if we use the vector product, or cross product, notation. The
vector product C of two vectors, A and B, is itself a vector normal
to the plane of A and B. The sense of C is given by a right-hand
The basic outline of the treatment developed here has been published
in the American Journal of Physics, Vol. 110, pp. 41-44 (1962), as "The Mag-
netic Pole in the Formulation of Electricity and Magnetism" by Robert
Katz. Some of the material has been developed in fuller detail in an intro-
ductory text entitled Physics by H. Semat and R. Katz, Holt, Rinehart and
Winston (1958). See also "Magnetic Monopoles" by K. W. Ford, Scientific
American 209, 122 (1963).
103
screw rule. If a right-handed screw is rotated in the sense which

rotates vector A into the direction of vector B when these are
drawn from a common origin, the point of the screw advances in
the direction of vector C, as in Fig. 6-1.1. The magnitude of C
FIG. 6-1.1 The vector product, or cross product, of two vectors A and
B is the vector C, written as A X B = C. The magnitude of C is given
by C = AB sin (J, and the direction of C is perpendicular to the plane
formed by A and B pointing in the direction of advance of a right-hand
screw turned so as to advance from A to B. The angle (J is the angle
between A and B when these are drawn from a common origin.
is equal to the product of the magnitudes of the two vectors

A and B and the sine of the angle between them (when they are
drawn from a common origin). Thus
C = AB sin 8. (6-1.1)
The order of the factors in the vector product is of some im-
portance. While the magnitude of A X B is the same as the
magnitude of B X A, the two vectors are oppositely directed, so
that
A X B = -B X A. (6-1.2)
The vector product C is perpendicular to each of the two vectors
RELATIVITY AND ELECTROMAGNETISM 105
A and B separately, for it is perpendicular to their common

plane.
We have already noted the principle of conservation of charge,
which states that the total charge content of the universe is con-
stant: charge is neither created nor destroyed. We extend this
concept to assert that the charge of a charged body is the same
for every inertial observer. In a similar way the theory of elec-
tricity and magnetism is structured on the concept of conserva-
tion of magnetic poles. We will assume that the pol~ strength of
a magnetic pole is the same for all observers in all inertial
frames, without regard to the speed with which it is moving.
6-2 The Lorentz Force Let us write the Lorentz force on a
charged particle as
F = q(E + v X B) (6-2.1a)
and the Lorentz force on a magnetic pole as
F = qm(H - v X D). (6-2.1b)
These equations are written in the mks system of units. The
force F is in newtons. The electric charge q is in coulombs. The
magnetic pole strength qm is in webers. The field quantities are
point functions in a particular inertial frame; that is, they de-
pend on the choice of the inertial frame in which they are
measured as well as on the point in space at which they are
measured. This is clear because we have to distinguish between
rest and motion of the probe charge or probe pole in determining
the field vectors E, B, H, and D.
The electric field intensity E, at a particular point in space
and with respect to a particular inertial frame, is the force per
unit charge on a positive probe charge at rest in that frame, and
is expressed in units of newtons/coulomb. The magnetic field
intensity H, at a particular point in space and with respect to a
particular inertial frame, is the force per unit pole on a north
probe pole at rest in that frame, and is expressed in units of
newtons/weber. The magnetic flux density B is expressed in
webers/m2 and the electric flux density (electric displacement)
D is expressed in coulombs/m2 The flux densities Band D may
be determined by observing the force on moving probe particles

of appropriate character. A moving charge will permit the de-
termination of B, while a moving pole will permit the determina-
tion of D.
The meaning of the field vectors remains the same in vacuum
or in a material medium, as described above. The probe charge
or probe pole is intended to be of small charge or pole strength
so as not to alter the system it probes when it is used to measure
the field. In a medium, the fields vary rapidly in the vicinity of
atoms. When we wish to determine the rapid variations in the
field, we imagine that we probe with a physically small probe,
and refer to the field so determined as the microscopic field.
When we are interested in average properties of the field and do
not wish to experience the violent fluctuations as we pass into
volumes occupied by individual atoms, we imagine that the
probe is physically large and ghost-like so that it can encompass
tens or hundreds of atoms at one instant. The field experienced
by the large ghost-like probe is called the macroscopic field.
Regardless of whether we are interested in the macroscopic or
microscopic fields, the sense of what we mean by the field quanti-
ties remains the same, and is appropriate to the Lorentz force
equations.
62.1 An airplane making magnetic surveys carries a field-measuring de-

vice on a long cable. The earth's magnetic field H is measured to have
=
a horizontal component of 0.18 0ersted. (1 NjWb 411" X 10- 3
0ersted). If the airplane flies at 600 mph through a thunderhead where
the vertical component of the electric field E is 11 X 105 NjC, what is
the maximum percentage error in the measured magnetic field due to
the passage of the magnetometer through the electric field? For pur-
poses of calculation assume that the field is measured by determining
the force on the magnetic pole of a compass needle. [0.005 percent]
6-3 Magnetization and Polarization It is customary to as-

cribe the effect of a medium on the field to its polarization P or
its magnetization M, the electric or the magnetic dipole moment
per unit volume.
An electric dipole consists of a pair of equal positive and nega-
tive charges separated by a distance s. If the displacement vector

from negative to positive charge is s, the electric dipole moment
p of the charge pair is
p = qs. (6-3.1)
When such an electric dipole is placed in a uniform electric
field, the net force on the dipole is zero, but the torque on it is
G = P X E. (6-3.2)
A magnetic dipole may be imagined to be a pair of equal
north (+) and south (-) poles, separated by a distance s. If the
displacement vector from south to north pole is 5, the magnetic
dipole moment of the pair is
m = q".s. (6-3.3)
When such a magentic dipole is placed in a uniform magnetic
field, the net force on the dipole is zero, but the torque on it is
G= m X H. (6-3.4)
Conversely, if a particle or a bit of matter experiences.a torque
when it is placed in an electric field, we say that it has a net
electric dipole moment; and if it experiences a torque when
placed in a magnetic field, we say that it has a net magnetic
moment. We may conceive of an experiment by which the torque
on a bit of uniformly polarized material of volume AT placed in
a magnetic field is determined, and is analyzed to yield a measure
of the polarization through the equation
G = (P.ih) X E. (6-3.5)
Similarly, the magnetization of a bit of matter may be found
through an analogous experiment and the equation
G = (MAT) X H. (6-3.6)
The polarization is measured in units of electric dipole mo-
ment per unit volume, or coulombs/m 2, while the magnetization
is measured in magnetic dipole moment per unit volume, or
webers/m2
In introductory texts in physics it is shown that the several

electric vectors are related through the equation
D = EoE + P, (6-3.7a)
and that the several magnetic vectors are related through the
equation
B = Poll + M. (6-3.7b)
The quantity EO is called the permittivity of free space, and has
the value
EO = 8.85 X 10-12 C2/N-m 2 = (3611" X 109)-1 C 2/N_m 2
The quantity 1'-0 is called the permeability of free space, and has
the value
P-o = 411" X 10-7 Wb 2 /N-m 2.
The product Eol'-o is the reciprocal of the square of the velocity of
light in vacuum; that is,
EoP-o = c-2 = (3 X 108 m/sec)-2.
Since vacuum contains neither electric nor magnetic dipoles,
the polarization and magnetization of vacuum are zero, and Eqs.
6-3.7 reduce to
D = EoE (6-3.8a)
and B = Poll. (6-3.8b)
6-4 Transformations of Fields and Flux Densities We call
on a result from relativistic mechanics which we have already
derived in Eq. 3-4.4. If a force F acts on a particle in the labora-
tory frame in which it moves with velocity v, observers on the
proper frame of the particle will measure the force on the par-
ticle to have a different value, F'. The components of these forces
parallel and perpendicular to the velocity v are given by
F II = F'II (6-4.1a)
and 'YF.I. = F'.I.' (6-4.1 b)
where 1'-2 = 1 - V2/C 2
Some writers prefer an asymmetric treatment in which Eq. 6-3.4 is written

as G = m X JoH and Eq. 6-3.7b is written asB = JoH +~. These differ-
=
ences may be reconciled by setting m ~ and M = P<N.
To apply Eqs. 6-4.1 to the electromagnetic field and flux density

vectors we imagine that we probe the fields at a particular point
in space by means of a probe particle, first electric and then mag-
netic. We shall determine the fields in the laboratory frame and
in the proper frame by application of the Lorentz force equa-
tions in each frame, and shall relate our determinations through
the force transformation equations.
Laboratory observers find the force on the charge q moving
with velocity v with respect to the laboratory frame to be
F = q{E + v X B), (6-4.2a)
with component equations taken parallel and perpendicular to
the velocity as
F II = qE II (6-4.2b)
and F.L = q{E.L + V X B). (6-4.2c)
Proper observers who are at rest with respect to the charge
know that there can be no force on the charge due to a magnetic
flux density in their frame, because the charge is at rest. Accord-
ing to these observers the force on the charge must be entirely
due to an electric field E', from the equation
F' = qE', (6-4.3a)
with component equations
F'II = qE'1i (6-4.3b)
and F'.L = qE'.L' (6-4.3c)
By means of Eqs. 6-4.1, the dictionary through which forces
in the two inertial frames are related, we find that
E'II = Ell and E'.L = "Y{E.L +V X B). (6-4.4a)
If we multiply both sides of each of the equations of 6-4.4a by
EO and apply Eq. 6-3.8, which relates D to E and B to H, we find
D'II = D II and D'.L = "Y{D.L + EolJ.QV X H). (6-4.4b)
With the assumption of the Lorentz force on a moving pole, a
parallel argument leads to the transformation equations for
Hand B. We obtain
H'II = H II and H'J. = ,),(HJ. - v X D), (6-4.4c)

B'II = B II and B'J. = ,),(BJ. - EO/JOV X E). (6-4.4d)
Equations 6-4.4b and 6-4.4d were derived from 6-4.4a and
6-4.4c for the vacuum case. But in any inertial frame the meaning
of the fields and flux densities is the same whether in vacuum or
in the medium. The transformation equations remain correct
whether the fields and flux densities are measured in vacuum or
in a medium. We will see later in 6-7 that these transformation
equations may be applied to Eqs. 6-3.7 to yield transformation
equations for magnetization and polarization.
Since we may regard either frame as the proper frame in which
the probe charge and pole may be placed at rest, we may readily
derive the inverse transformations by interchanging the prime
and unprime superscripts and replacing v by -v. We then have
Ell = E'II' EJ. = ')'(E'J. - v X B'); (6-4.5a)
D II = D'II' DJ. = ')'(D'J. - EoJLoV X H'); (6-4.5b)
H II = H'II' HJ. = ')'(H'J.+ v X D'); (6-4.5c)
B II = B'II' BJ. = ')'(B'J. + EO/JOV X E'). (6-4.5d)
The transformation equations 6-4.4 and 6-4.5 relate the fields
and flux densities in the laboratory (unprimed) frame and in a
second frame (the primed frame) which is moving with velocity
v with respect to the laboratory frame.
These equations provide the basis for vague statements which
are sometimes encountered, such as "motion with respect to an
electric field makes a magnetic field" or "A moving magnetic
field makes an electric field:'
6-4.1 In the n = =
I Bohr orbit of hydrogen (r 0.529A 0) find the
electric flux density at the position of the electron (a) in the laboratory
frame (D) and (b) in the proper frame of the electron (D'). Find the
magnetic field intensity at the position of the electron (c) in the labora-
tory frame (H), and (d) in the proper frame (H,) of the electron. The
speed of motion of the electron may be obtained from the quantum
condition on angular momentum mVT = hj21r. [(a) 0.72 X 10- 18 Cjm 2
(b) 0.72 X 10- 18 Cjm 2 (c) 0 (d) 1.6 X 10- 12 NjWb)]
6-5 Electromagnetic Induction There are two aspects to
electromagnetic induction, as reflected in the transformer and in
the generator (dynamo). A changing flux through a loop of wire

generates the electromotive force of the transformer. While this
effect may be conceived as the electric field due to a displacement
current of magnetic poles (following 6-6), we shall devote no
space to its exposition. The generator effect is due to the motion
of the wires of the generator with respect to a static magnetic
field, and is best understood in relativistic terms as due to the
transformation of electromagnetic fields and flux densities.
Suppose a wire moves through a magnetic field of flux density
B. In the frame of the laboratory the electric field intensity is
zero, but in the frame of the wire the electric field intensity E'
given by Eq. 6-4.4a is
E' = 'YV X B. (6-5.1)
If a wire of length s slides over a fixed conductor a consisting
of two parallel tracks which are electrically connected at one
end, through a magnetic field in the laboratory frame, as in
Fig. 6-5.1, current flows in the circuit. The electromotive force e
induced in this process is the work done in carrying a unit posi-
tive probe charge around a closed loop which is partly at rest
in the laboratory frame, and partly at rest in the moving frame,
for the loop is made up of the fixed conductor a and the moving
conductor s. Integrating around the loop, we have
e= f Eds = fa Eds + /, E' ds = /, 'YV X Bds. (6-5.2)

It is customary to ignore the factor y in introductory treatments,
for the velocity is assumed small compared to CJ and to evaluate
the integral as
e = 'YBsv :: Bsv (6-5.3)
in the geometry of Fig. 6-5.1.
Since the motion of a wire through a unifonn magnetic field
is the prototype for the design of electric generators, we see that
the commercial generation of electric power depends on the in-
herent relativistic character of electricity and magnetism.
The effect of the motion of the wire through the field has
sometimes been attributed to "flux cutting." It is difficult to
conceive of the Faraday disk or its linear counterpart, the con-

ducting bar moving along its length in a magnetic field, in flux-
cutting terms. The Faraday disk is shown in Fig. 6-5.2a, and its
linear counterpart in Fig. 6-5.2b. These devices are sometimes
classed as homopolar generators.
X8 x X8 x x IX
x x x x
x x x x x x
X8 X X8 X X 8X
FIG. 6-5.1 A current , flows in a closed circuit made up of a wire s
which slides over parallel tracks through a region of flux density B
(measured in the frame in which the tracks are at rest). An electromo-
tive force is induced in the circuit. The wire is sometimes called the "seat
of emf," for the only contribution to the emf is in that portion of the
circuit containing the moving wire. Of course, the wire does not slide
by itself. An outside agency must exert a force in the direction of v to
keep the wire sliding with constant speed, and to provide the mechani-
cal energy which is converted to electrical energy.
As in section 5-6, we will assert that acceleration does not

affect the validity of the Lorentz transformations and their conse-
quences. We will treat both cases illustrated in Fig. 6-5.2 in the
same manner. The magnetic flux density has been determined
to be B and the electric field to be zero in the laboratory frame.
At a point P fixed to the moving conductor, we have, by Eq.
6-4.4a, that the electric field E' is given by
E' = ,\,V X B. (6-5.1)
Thus, if the bar in Fig. 6-5.2b is of width 1, the potential differ-

ence (emf) across the edges of the bar is given by
e= "'(vBl. (6-5.4)
In the case of the disk, the instantaneous proper frame of a
B
.
B

B

B B

B
8
(a)
8

8 8
... 8

8 8
.pe. v
..
8 8
8
8
+ . . . . . .
8 8
(b)
FIG. 6-5.2 (a) Faraday disk generator. A copper disk rotating in a uni-
form Reid generates a potential difference between brushes on the axle
and the rim. (b) Linear counterpart of the Faraday disk. A long bar
moves through a uniform magnetic Reid.
point P which is T distant from the center is moving with speed

=
v wT with respect to the laboratory frame. The electric field is
directed radially outward. Concentric cylinders centered on the
axle are equipotential surfaces. The electric field has magnitude
E' = ,wrB (6-5.5)

and the potential difference (emf) is
[; = JwBr dr = wBr /2 2 (6-5.6)
in the limit of low speeds.

6-5.1 A wire of length s moves with velocity v at right angles to a
magnetic field of flux density B, the wire being perpendicular to both
v and B. Find the force on a charge q at rest in the wire, as determined
by (a) a Jaboratory observer and (b) an observer in the frame of the
wire. The proper observer (at rest in the frame of the wire) attributes
the force to an electric field E. (c) Give the value of this electric fu~ld.
(d) What is the value of the emf across the ends of this wire as deter-
mined by the proper observer? [(a) qv X B (b) '"tqv X B (c) '"tv X B
(d) '"tvsB].
6-5.2 A wire 200 em long is moved in the +x direction at a speed of
50 m/sec. so that the wire is always parallel to the z direction. The mag-
= = =
netic field has components BIlJ 0.4 Wb/m 2 , B II 0.8 Wb/m 2 , B z -0.6
Wb/m2. Find the emf induced in the wire. [80 volts, in the +z direc-
tion]
6.6 Field of a Moving Charge At low speed a charged body
generates an electric flux density D and a magnetic field intensity
H given by
(6-6.1 a)
and (6-6.1 b)
Equation 6-6.lb is equivalent to the Biot-Savart law for the mag-

netic field due to a current element, as we may see by replacing
qv by i dl.
Similarly, a slowly moving magnetic pole generates a magnetic
flux density B and an electric field intensity E given by
B = qm 1, (6-6.1c)
4111'2
and E = -v xB = (6-6.1d)
An equation equivalent to the Biot-Savart law for the electric

field due to an element of pole current may be obtained by re-
placing qmv by im dI.
Equations 6-6.1 are valid in vacuum or in a homogeneous me-
dium, in the reference frame in which the particle velocity is v
and where lr is a unit vector directed from the particle to the
field point P located r away.
To find the field equations valid at arbitrary speeds we apply
the relativistic transformations to the Coulomb field of a charged
particle at rest in its proper frame. Let us call the direction of
motion of the particle in the laboratory frame the x direction.
Then the proper (primed) observer notes the electric flux density
to have cOII1ponents
D'", = qx'/4rr'3, D'" = qy'/41rr'3, D'. = qz'/4rr'3, (6-6.2)
and the magnetic field to be zero.
An observer in the laboratory (unprimed) frame notes that the
particle is moving with velocity v, and applies Eqs. 6-4.5 to
translate the proper field components into laboratory field com-
ponent values. He finds
D", = D'"" D" = 'YD'", D. = 'YD'., (6-6.3a)
H", = H''" = 0 = (v X D)""
H" = 'Y(v X D')" = (v X D)",
H. = 'Y(v X D'). = (v X D).,
or H = v X D, (6-6.3b)
in agreement with the relationship between Hand D at low
speeds given in Eq. 6-6.1 b.
To express the fields in terms of the position of the particle
in laboratory coordinates we must take note of the Lorentz con-
traction. If the charged particle is at the common origin of both
frames, a field point whose coordinates in the proper frame are
x', y', z' will have laboratory coordinates x, y, z, related through
the equations
x' = "'IX, y' = y, t = z, (6-6.4)
and the quantity
r'2 = X'2 + y'2 + Z'2
transforms into
S2 = 'Y 2X2 + y2 + Z2 = 'Y 2r2 (1 - fJ2 sin 2 6).
With these results we find from Eq. 6-6.3
D" = 'Yqxj41rs3, D II = 'Yqy j41rs 3, D. = 'Yqz/41rs3,
which may be expressed in terms of rand 6 as
D = qt. (6-6.5a)
41r'Y 2r2 (1 - fJ2 sin 2 6)1/2
and H = v X D. (6-6.5b)
The electric field of a rapidly moving charged particle is not
spherically symmetric. The fields ahead of the particle (6 = 0)
and directly behind it (8 = 'IT) are reduced, while the field is in-
creased in the transverse direction, as shown in Fig. 6-6.1.
When a relativistic charged particle passes through matter, an
FIG. 6-6.1 Electric field Intensities at fixed distance from a rapidly

moving charge, detennined in the laboratory frame h = 10).
atomic electron experiences a sudden thrust to the side as the
charged particle goes by. In the limit f3 ~ 0 and y ~ I we find
. that Eqs. 6-6.5 approach the nonrelativistic forms given in Eqs.
6-6.1.
Similar results may be obtained for the fields of a rapidly mov-
ing pole.
=
66.1 A proton moves through the laboratory at 'Y 10. What is the
electric field intensity E at points I millimicron from the proton (a)
ahead, (b) at 30, (c) at 60, (d) at 90 to the proton~s motion? [(a)
Eoo =1.4 X 107 N/C (b) E3O' =
1.8 Eo' (c) E 60 0 =
22 Eoo (d) Egoo =
1000Eoo.]
66.2 Two electrons move abreast down the axis of a linear accelerator
= =
in parallel paths at fJ 0.995 ('Y 10), separated transversely by one
millimicron. (a) Find the force between them in their proper frame.
(b) Find the force between them in the laboratory frame. Do this two
ways. First, transform the result of part (a) to the laboratory frame.
Next compute the fields in the laboratory frame, and then apply the
Lorentz force equations to find the force in the laboratory frame. Note
that a smaller force is experienced in the laboratory frame than in
the proper frame. In the laboratory beams of rapidly moving particles
spread apart more slowly than beams of slowly moving charged parti-
cles through this effect, called relativistic focussing. [(a) 2.3 X 10- 10 N,
repulsive (b) 2.3 X 10- 11 N, repulsive]
6-6.3 Prepare polar plots of E (at 1 mIL) vs. 8 for electrons whose
laboratory energy is (a) mc2 (b) 10 mc2 (c) 100 mc2 Plot E(O) as the
radial distance from the origin at different angles 0 with the direction
of motion in the laboratory frame.
6-7 Tranformation of Polarization and Magnetization We
may find transformation equations for magnetization and polar-
ization by applying Eqs. 6-3.7, which relate the field and flux
density to the properties of the medium, to the transformation
equations for E, B, D, and H which we have already derived.
We find
D'.L = 'Y(D.L + Eop,oV X H).
Now
D'.L = EoE'.L + P'.L
and
and
E'.L = ')'(E.L + V X B)
so that
Eo')'(E.L +v X B) + P'.L = ')'(EoE.L + P.L + EoIJ.oV X H).
Thus
P'.L = ')'(P.L - EOV X (B - IJ.oH),
or
(6-7.1a)
Similarly
P'II = P II , (6-7.1 b)
M'.L = ')'(M.L + IJ.oV X P), (6-7,lc)
M'II = Mil' (6-7.1d)
In the usual way, the inverse transformation is obtained by
interchanging prime and unprime superscripts and by replacing
v by -v. We have
P.L = ')'(P'.L + EOV X M'), (6-7.2a)

P II = P'II, (6-7.2b)
M.L = ')'(M'.L - IJ.oV X P'), (6-7.2c)
Mil = M'II' (6-7.2d)
Thus a moving electret which has only permanent polariza-
tion in its proper frame may appear to laboratory observers to
be magnetized. A moving magnet appears to be polarized.
We may readily interpret these results in elementary terms if
we note that the magnetic moment of a current loop is given by
m = IJ.oiA, (6-7.3a)
and the electric dipole moment of a pole current loop is given by
(6-7.3b)
where A is the area of the loop. The direction of the moment is
given by a right-hand rule for Eq. 6-7.3a and a left-hand rule for
Eq. 6-7.3b. The magnetic moment of a current loop may be
derived by calculating the torque on a loop of current in a uni-
form magnetic field. We recall that the magnetic force on a cur-
rent element may be obtained from the Lorentz force on a mov-
ing charge by replacing qv by i dl. The magnetic moment may

then be obtained by comparison of this result with Eq. 6-3.4 for
the torque on a magnetic dipole. The electric dipole moment
of a pole current loop may be obtained through use of an anal-
ogus recipe, with imdl replacing qmv in the Lorentz force equa-
tion, and the result of the torque calculation compared to Eq.
6-3.2.
Let us now examine a moving ribbon of polarized material of
proper polarization pI (a ribbon electret), moving parallel to its
length and polarized perpendicular to its direction of motion, as
shown in Fig. 6-7.la.
fp, fp, fp, [p, i~ fp, , f~

-
p
i
v
(0)
'"
'"
M M M M M M
I' hI' /' ;t -"
-"
_/-,
I
(b)
FIG. 6-7.1
In the laboratory frame, from Eq. 6-7.2 we find

p = -yP' (6-7.4a)
and M = -'YJl.oV X P'. (6-7.4b)
The ribbon appears to be differently polarized in the laboratory

frame, and magnetized as well. Note that the difference in polari-
zation vanishes in the limit y ~ 1, but that the magnetization
remains, for this is a first-order effect.
These results may be easily understood from elementary con-
siderations. A polarized ribbon may be thought of as a collection
of large dipoles, whose negative charges are all along the bottom
edge of the ribbon and whose positive charges are at the top
edge. The motion of positive charge to the right constitutes a flow
of current to the right, but the motion of negative charges to
the right constitutes a flow of current to the left. We may there-
fore imagine the ribbon to be replaced by two currents. For
purposes of calculation we add pairs of opposing currents cross-
ing from the top edge of the ribbon to the bottom edge, and now
conceive of replacing the moving ribbon by a stationary ribbon
of current loops, as in Fig. 6-7.lb. By the right-hand rule we note
that the magnetization is into the paper.
A ribbon of width h and thickness g contains volume fgh in a
length f. The total dipole moment contained in this volume is
Pfgh. If ,\ is the edge polarization charge per unit length, then we
have
Xfh = Pfgh,
or X = Pg.
The current flow is the charge per second flowing past a point
in space. If the ribbon moves at speed v, then
i = Xu = Pgu.
The magnetic moment in the section is the product of the mag-
netization times the volume, and is also given by Eq. 6-7.3a so
that
Mfgh = p,oPgufh,
or M = p,oPu.
This result IS ill agreement with our transformation equation
6-7.4b in the limit of low velocities. Since we would expect the
polarization to be the same in b,oth frames, this conclusion from

Eq. 6-7.4a is of no surprise.
In precisely the same way we imagine a ribbon of magnetized
material to consist of pole currents along the edges of the ribbon,
and analyze the problem as one of electric dipole moment arising
from pole current loops. A laboratory observer measuring a mov-
ing strip of magnetic ribbon finds it polarized as well as magnet-
ized.
6-8 The Unipolar Generator In July IS51, Faraday entered
into his diary a discussion of an experiment in which a current
was induced in a wire which made sliding contact with points
near the axis and the surface of a rotating cylindrical magnet, as
shown schematically in Fig. 6-S.la.
Just as in the case of the Faraday disk we shall examine the
generator and its linear counterpart, schematized in Fig. 6-S.lb,
at the same time. As before, we shall assert that acceleration does
not affect the validity of the Lorentz transformations and their
consequences.
We idealize the magnet and ribbon counterpart by asserting
that they are so constructed that we may neglect H' as very small,
and that they are uniformly magnetized to a magnetization M',
so that B' within the magnetized material is also uniform and
equal to M', by Eq. 6-3.7b.
We may therefore consider that in the proper frame of the
ribbon there is magnetization M' and flux density B'. In the
laboratory, from Eqs. 6-4.5, the magnetic flux density is
B = -yB' (6-S.1a)
and the electric field iIlt~nsi[y is
E = --yv X B'. (6-S.1b)
Let us briefly revert to the elementary idea that the emf is the
open circuit potential difference across the terminals of a source,
just as in a battery.
If the thickness of the ribbon is g, then the potential difference
(emf) across its face is of magnitude
(a)
P
1 P
,1IM'~ 11M')~~'
/P /P / P1 / / /P /Py
v
s' s' s' s' s' s'
P J"P P I ~
IM,)./,;1IM,)/1IM
P 'P 'P
1
)/
v -
lb)
FIG. 6-8.1
e = 'YvB'g (6-8.2a)
We may analyze the behavior of the ribbon in another way.
From Eq. 6-7.2a moving magnetization appears in the laboratory
frame as polarization, with
(P = 'YEoV X M/.
From electrostatic theory, there is a potential discontinuity across
the opposite faces of a dipole layer of dipole moment per unit
area <P, given by <P/ EO' If the polarization of a ribbon of thickness

g is P, then its dipole moment per unit area is
(J> = Pg,
so that
e= "(vM'g, (6-8.2b)
which is the result of Eq. 6-8.2a, if we note that in this case
M' = B'.
The same analysis applies to the rotating magnet of Fig. 6-8.1a.
Here v = (() x r, and again as in Eq. 6-8.1b
E = -"(("I X r) X B'. (6-8.1b)

The electric field is radial. We may drop the vector notation. We
note that cylinders concentric with the axis are equipotential
surfaces, and that the potential difference between axis and sur-
face is given by
e = foR wrB' dr = wB'R2/2, (6-8.3)
in the limit of small speeds. The potential difference (emf) be-

tween the axis and surface of the cylinder is equal to the flux
within the rotating cylinder which would be cut by a radial wire
in one second.
6-9 Postscript Unlike classical mechanics, classical electro-
magnetic theory was only clarified, not essentially modified by the
special theory of relativity. In fact, relativity theory grew out of
electrodynamics, and Einstein's original paper on the subject was
entitled "On the Electrodynamics of Moving Bodies." As a result
of the complete inner consistency between them, the connections
between electromagnetism and relativity may be approached in a
variety of ways. One is to follow the historical path of examining
the experimental evidence for the incompatibility of electro-
magnetism and Newtonian mechanics. Through this avenue the
relativistic modifications of mechanics which reconcile the diffi-
culties may be found. Such a procedure has been followed in
many older texts and treatises on classical electricity and magnet-
ism, which may be found among the references at the end of

this chapter.
While the concept of a magnetic pole was used freely in the
work of Maxwell, his predecessors, and his followers, the subse-
quent discovery of the electron and the failure of experimental
physics to discover the corresponding particle, the free magnetic
pole, has led many physicists to avoid the use of the pole concept
in their exposition of electromagnetic theory (see Coulomb's Law
Committee report). In this case all field quantities are traced to
electric charges and currents. The magnetic effect of currents
may then be derived from the Lorentz transformation and Cou-
lomb's law. A full treatment of the transformation of the sev-
eral electromagnetic field quantities by this procedure may re-
quire considerable mathematical sophistication. Whatever ex-
pository path is chosen, the destination is the same, for the field
transformation equations depend on Maxwell's equations, and
not on the manner of their exposition. References have been pro-
vided which treat some of these questions by alternate proce-
dures.
E. Amaldi, G. Baroni, H. Bradner, H. G. deCarvaIho, L. Hoffman,
A. Manfredini, and G. Vanderhaeghe, Search for Dirac Magnetic
Poles, CERN 63-13 (1963).
R. Becker and F. Sauter, Electromagnetic Theory and Relativity, Blais-
dell Publishing Co., New York, 1964.
S. Devons, "The Search for the Magnetic Monopole," Science Progress
51,601 (1963).
B. Dibner, Oersted and the Discovery of Electromagnetism, Blaisdell
Publishing Co., New York, 1962.
R. P. Feynman, R. B. Leighton, and M. Sands, The Feynman Lectures
on Physics, Addison-Wesley Publishing Co., Inc., Reading, Mass.
F. Gutmann, "The Electret," Rev. Mod. Phys. 20, 457 (1948).
J. D. Jackson, Classical Electrodynamics, John Wiley and Sons, Inc.,
New York, 1962.
E. C. Kemble, Chairman, Coulomb's Law Committee, "The Teaching
of Electricity and Magnetism at the College Level," Am. ]. Phys. 18,
I and 69 (1950).
J. C. Maxwell, A Treatise on Electricity and Magnetism (2 vol.) (re-
print), Dover Publications, Inc., New York, 1954.
W. K. H. Panofsky and M. Phillips, Classical Electricity and Magnetism,
Addison-Wesley Publishing Company, Inc., Reading, Mass., 1955.
E. M. Pugh and E. W. Pugh, Principles of Electricity and Magnetism,
Addison-Wesley Publishing Company, Inc., Reading, Mass.
W. T. Scott, "Resource Letter FC-I on the Evolution of the Electro-
magnetic Field Concept," Am. J. Phys. 31, I (1963).
F. S. Shire, Classical Electricity and Magnetism, Cambridge University
Press, Cambridge, 1960.
A. Sommerfeld, Electrodynamics (Lectures on Theoretical Physics,
Vol. 3) (reprint), Academic Press, New York (1964).
E. Whittaker, History of the Theories of Aether and Electricity (2 vol.)
(reprint), Harper Torchbook, New York.
General Bibliography
M. Born, Einstein's Theory of Relativity (reprint), Dover Publications,
Inc., New York, 1962.
E. U. Condon and H. Odishaw, Handbook of Physics, McGrawHilI
Book Company, Inc., New York, 1962.
A. Einstein, H. A. Lorentz, H. Minkowski, and H. Weyl, The Principle
of Relativity (reprint), Dover Publications, Inc., New York, 1923.
G.Holton, "Resource Letter SRT-l on Special Relativity Theory,"
Am. J. Phys. 30, 462 (1962). Selected reprints of journal articles from
this bibliography are available as a reprint book on special relativity
theory from the American Institute of Physics, 335 E. 45 St., New
York 17, N.Y.
G. Joos, Theoretical Physics, Hafner Publishing Co., New York, 1950.
L. Landau and E. Lifshitz, The Classical Theory of Fields, Addison-
Wesley Publishing Company, Inc., Reading, Mass., rev. 2d ed., 1962.
C. Moller, The Theory of Relativity, Oxford University Press, Oxford,
1934.
W. Pauli, Theory of Relativity, Pergamon Press, New York, 1958.
H. Semat and R. Katz, Physics, Holt, Rinehart and Winston, New York,
1958.
R. S. Shankland, "Conversations with Albert Einstein," Am. J. Phys.
31, 47 (1963).
127
Index
Aberration, 40, 101 Compton wavelength, 86
Acceleration, transformation of, Conservation, of charge, 83, 105
Galilean, 16 of energy, 78, 86
Lorentz, 48 of linear momentum, 68, 85
Accelerated coordinate frames, 56, of magnetic poles, 105
96 Coordinate frame, accelerated, 96
Air navigation, 15 center of mass, 74, 78
Annihilation radiation, 83 inertial, 10
Atomic weight, 63 COPERNICUS (14731543), 9
Correspondence principle, 52
Covariance, 25, 77
Binomial expansion, 21
BOHR, NIELS (1885.1962), 52
BRADLEY, JAMES (1693.1762), 40 Dynamo, III
Bucherer experiment, 55 deBrogIie equation, 67
DEBROGLIE, LOUIS (b. 1892), 67
Dipole moment, 107
Calorie, 63
Doppler effect, broadening of spec-
Causal connection, 32
tral lines, 92
Center of mass, nonrelativistic, 71
longitudinal, 91
relativistic, 77
transverse, 9I
Center of mass frame, nonrela-
tivistic, 71
relativistic, 77 EINSTEIN, ALBERT (1879-1956), 24,
Centrifugal force, II 123
Charge conservation, 83, 105 Elastic collisions, 74
Clock, rod, 25 Electrical generator, dynamo, II I
Closed system, 70 homopolar, 112
COMPTON, A. H. (1892-1962),85 unipolar, 112
Compton effect, 85, 87 Electromagnetic induction, 110
129
130 INDEX
Electromotive force, III Gravitational deflection of light,

seat of, 112, 121 100
Electron volt, 62 Gravitational red shift, 98
Energy, 62
conversion to mass, 80, 87-89 Headlight effect, 42
kinetic, nonrelativistic, 64 Homopolar generator, 112
kinetic, relativistic, 63
potential, 80 Inelastic collisions, 74
Energy-momentum relation, rela- Inertial frame, 10
tivistic, 66, 75 Interferometer, 23
Epicycle, 9 Internal energy, 62
Equivalence principle, 56, 96 Internal force, 69
Erg, 62
Joule, 62
Ether, 19
Jupiter, 18
Excited state, 74
External force, 69 Kinetic energy, internal, 74
nonrelativistic, 63
Faraday disk, 113 of and about center of mass, 73
FARADAY, MICHAEL (1791-1867), relativistic, 64
103, 113, 121
Field of a moving charge, 114-116 Light, speed of, 17
Field intensity, electric (E), 105 Linear momentum, conservation
magnetic (H), 105 of,68
FIZEAU, A. H. L. (1819-1896), 19,40 Longitudinal mass, 55
Flux density, electric (D), 105 LORENTZ, H. A. (1853-1928), 24
magnetic (B), 105 Lorentz contraction, 24, 36
Force four-vector, 96 Lorentz force, 105
Force transformation, nonrelativis- Lorentz transformation, 17, 29, 30
tic, ~8
relativistic, 58, 59 MACH, ERNST (1838-1916), 10
FOUCAULT, J. B. (1819-1868), 11 Mach's principle, 10
Foucault pendulum, II Mass, change of, 80
Four-vector, displacement, 30, 96 longitudinal, 55
force, 96 moving, 53
momentum-energy,77 rest, 53
Frame of reference, accelerated, 96 transverse, 55
center of mass, 74, 78 Mass-energy conversion, 80, 83
inertial, 10 Macroscopic field, 106
Magnetization, 106
Galilean transformation, II transformation of, 117
GALILEO (1564-1642), 12 MAXWELL, JAMES C. (1831-1879),
General relativity, 56, 96 103, 124
INDEX 131
Mesons, decay of, 33 Proper frame, 27

production threshold, 87 Proper lifetime, 36
Microscopic field, 106 PTOLEMY (A.D. 127-151),9
MICHELSON, ALBERT A. (1852-
1931),22
Quantum mechanics, 53
Michelson-Morley experiment, 22
Minkowski force, 95
MINKOWSKI, HERMAN (1864-1909), Relativistic dynamics, 92
95 Relativity, general, 56, 96
Momentum, 52 Rest energy, 65
Momentum-energy relationship, Rod clock, 25
relativistic, 66 ROEMER, OLAF (1644-1710),18
Momentum-energy transformation,
relativistic, 75
Scalar product, 54
Mossbauer effect, 98
Scattering of x-rays, Compton, 85
Moving charge, fields of, 114-116
Thomson, 87
Seeing, 45
Natural state, 10,51
Simultaneity, 31
Newtonian mechanics, 51
Speed of light, 17
NEWTON, ISAAC (1642-1727), 9
Speedometer, 34
Newton's first law, 10
second law, 54
third law, 69 Time dilation, 33
Nichols-Hull experiment, 71 Time sequence, 30
Neutrino, 71 THOMSON, J. J. (1856-1940),87
Thomson scattering, 87
Pair production, 83 Transformation equations, non-
Particle production thresholds, pi relativistic, for acceleration, 16
meson, 87 for coordinates (and time), 13
lambda-antilambda pair, 89 for force, 58
proton-antiproton pair, 89 for velocity, 14
PAULI, WOLFGANG (1900-1958), 71 Transformation equations, rela-
Pi meson, lifetime, 33 tivistic, for acceleration, 48
production threshold, 87 for coordinates (and time), 30
Permeability of free space, 108 for electromagnetic fields and
Permittivity of free space, 108 flux densities, 108
PLANCK, MAX (1858-1947), 67 for energy and momentum, 75
Planck's constant, 67 for force, 58
Polarization, 106 for magnetization and polariza-
transformation of, 117 tion, 117
Potential energy, 62 for velocity, 37
Power, 92 Transverse mass, 55
132 INDEX
Unipolar generator. 121 Violent state. 51
Unit vector. 14 Visual appearance of moving ob
jects.44
Vector product, 103
Velocity, transformation of. Gali- Weight, of a moving object, 57
lean, 14 of a resting object. 57
Lorentz, 37 Work,61
Errata in AN INTRODUCTION TO THE SPECIAL THEORY OF RELATIVITY
Robert Katz Van Nostrand Momentum .9
Note -1 means 1 line from the bottom, +6 means 6 lines from top
p ~a answers to 1-5.1 1 z 1.2~, 1.67, 1.08, 2.6~
P 31 -1 Eqs. 2-1.3
P 34 +2 1.84 xlO- 7 sec
p 34 +8 y. 7.0~
P 38 Problem 2-5.1 b) Two particles are emitted in opposite

directions from a stationary source.
P 44 Problem 2-7.1 answer 0.99995c
p 47 +7 edge ~
----~:..2 dU~
P 48
P 50
Eq. 2-9.1 b dU y,z
Problem 2-9.2 b) 53
. c) 0.16
P 87 Problem 5-2.1 a) 0.242xlO -9 - em

-9
P 87 Problem 5-2.2 d) 1. !.2x10 erg
P 88 -1 m c 2 (y + 1)
p
( }c2(~)ill
P 89 +11 T 2.07
P 89 +12 2.07 x m11 c 2 280 MeV
o
p 89 Problem 5-3.1 a) 774 MeV b) 5.61 GeV
P 92 Problem 5-4.2 0.54

A
P 92 Problem 5-4.4 a) 0, 2.23 MeV 180, 0.12 MeV
b) 8.75, 2.06 MeV -69.08, 0.33 MeV
c) 13.09, 1.80 MeV 49.29, 0.54 MeV
d) 25.84, 1.17 MeV
e) 1.17 MeV
P 96 Problem 5-5.1 Eq. 3-!.4

P 110 Problem 6-4.1 a) 4.55 c/m 2 b) 4.55 C/m 2
P 118 +6 add final paren ) ,
P 117 Problem 6-6.1

Intro Special

Uploaded by

Copyright:

Available Formats

Intro Special

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Intro Special

Uploaded by

Copyright:

Available Formats

University of Nebraska - Lincoln

DigitalCommons@University of Nebraska - Lincoln

An Introduction to the Special Theory of Relativity

Follow this and additional works at: http://digitalcommons.unl.edu/physicskatz

Published, copyright 1964 D. Van Nostrand Company, Inc.

Electronic edition published, copyright 2008 Robert Katz.

It is the purpose of this book to provide an introduction to

purpose here to emphasize the importance of the choice of a

the absence of gravitational or other force fields, a particle set

projectile motion, Galileo (1564-1642) concluded that the motion

Suppose that a particle is located at a point P at a particular

dx' = dx - V dt, (1-2.2a)

&.:. = rJ1, or vII = VII' (1-2.3b)

These results may be put into vector form by multiplying each

V' = 4 mph Xl II'

The boat moves at 5 mph with respect to the ground in a direc-

V is very large, of the order of the speed of light in vacuum

that either long paths or short time intervals must be measured.

earth is approaching Jupiter, position 2 in Fig. 1-3.1. He inter-

Galileo, who stationed two observers with lanterns a mile apart.

passage of stars or the planets. Its function was to provide a

toto = 2L/c. (1-3.1a)

Suppose the rod is moving through the ether with speed v,

If we now make use of the binomial theorem for (1 + x)m,

It was supposed that the ether was dragged (convected) along

that there is no purpose in postulating the existence of an ether

measurement of time. Moving clocks do not run at the same rate

verse to its direction of motion. Suppose that there is an observer

the proper frame of the rod. An observer in the proper frame of

the moving clock is observed by the rest observer to take a time

to I. Prove that the error in 'Y in making the approximation that x 2

2-1 The Lorentz Transformation We may consider a co-

Just as the spreading light flash forms a spreading sphere in the

In these equations the primed frame is taken to be moving with

two frames spreads out as a sphere in both frames. When ob-

2-2 Simultaneity and Time Sequence Suppose we now ex-

t\ = 'Y(t1 - VxI/c 2) and t'2 = 'Y(t2 - VX2/C2).

Two events simultaneous in the unprimed frame are observed to

t'2 - t'l = 'Y [(t2 - t1) - ~ (X2 - Xl) 1 (2-2.2)

If the quantity in brackets is equal to zero, then the two events

the decay time will be measured in the laboratory to be 2.6 X

2-4 Lorentz Contraction Let us set up an imaginary experi-

rod which is at rest in a coordinate frame. Observers in the frame

L' = "fL. (2-4.2)

U'" = U'" - V, U'",+ V

tions of Eqs. 1-2.3 in the nonrelativistic limit f3 ~ O. But there

!:i.t = .!:.... _ .!:.... = L

Suppose we have beams of yellow light moving in opposite di-

!:i.t = 15m (1 + 0.7 X 10- 8

= 2.1 X ~~~ m/sec (2 X 0.7 X 10- 8

tive agreement with our calculations. This experiment was first

" star, at rest In

wholly understandable in terms of the velocity transformation

frame are V'II = C, V':J) = 0, V'z = 0, to a sufficient approximation.

the laboratory frame (the unprimed frame) with speed V. Radia-

On Fig. 2-7.2b we have mapped corresponding rays in the

2S Visual Appearance of Rapidly Moving Objects When

direction, and that it is photographed from a distant point P (see

2-8.1 A small object is in the form of a sphere of radius R in its proper