Crux v9n08 Oct
Crux v9n08 Oct
Crux v9n08 Oct
Mathematicorum
Published by the Canadian Mathematical Society.
http://crux.math.ca/
CRUX MATHEMATICORUn
Vol. 9 S No. 8
October 1983
Sponsored by
Carleton-Ottawa Mathematics Association Mathematique d'Ottawa-Carleton
Publie par le College Algonquin, Ottawa
The assistance of the publisher and the support of the Canadian Mathematical Olympiad
Committee, the Carleton University Departrent of Mathematics and Statistics, the University
of Ottawa Department of Mathematics, and the endorsement of the Ottawa Valley Education
Liaison Council are gratefully acknowledged.
ft * * ft & & * * £ ft j* ft ft ft it ft ft ft ft ft ft ft ft ft ft it ft it ft ft ft ft ft ft ft ft ft ft it
ftftftftftftftftftftftit ft ft ft ft it ft ft ft ft ft ft ft ft ft ft ft ft ft ft ft ft s',-ftftftft
CONTENTS
Boolean Inequalities from Lattices, Arrays, and Polygons . . . . J.L. Brenner 224
Fourth-Order Additive Digital Bracelets in Base Five . . . . Charles W. Trigg 231
Mathematical Clerihews . . . . . . . . Alan Wayne 235
The Olympiad Corners 48 , . . . M.S. Klamkfn 236
240
Hollywood Arithmetic
Problems - Problernes « . . 241
Introducing Maybe the Next Editor of Crux Mathematieorum 242
Solutions . . . . . . . . 243
The Puzzle Corner . . . . . . . . . . . Alan Wayne 256
- 223 -
- 224 -
{ ! - ( ! - pm)n} + {1 - (1 - qnf} + r
max{m n}
> < 1.
(1.51) (1 - pi x) + (1 - p 2 z ) + (1 - P3 ) ^ 2 + (I-P1-P2-P3) »
where l\ = n2n^ l2 = n^n-y Z 3 = n ^2., witfc equality if and only if nj = n 2 = n 3 = *•
Proof. Arrange n\n2n3 beads in an n^n2xnz rectangular parallelepiped with
n j planes of dimensions n2xn39 n2 planes of dimensions rc3xttl9 n 3 planes of dimensions
n\*n29 and a "fractured" dogleg of max {ni,n 2 ,n 3 } beads extending from one corner
bead in a direction with equal direction cosines 1/^ until i t reaches a face,
proceeding then in that face in a direction with equal direction cosines 1//2 until
i t reaches an edge, and then remaining in t h a t edge until i t reaches the bead in
the corner opposite the s t a r t i n g corner. (Of course, there is no "fracture" if
two of the n, are equal, and the "leg" is s t r a i g h t if all three of the n. are
equal.) Color each bead randomly with exactly one of four d i s t i n c t colors 1,2,3,4:
using color i with probability p . > 0 for i = 1,2,3, and using color 4 with pro-
b a b i l i t y r > 0. The following four events are mutually exclusive, and t h e i r
union is exhaustive if and only if n x = n2 = n3 = 1:
1.52. At l e a s t one of the n\planes of dimensions n2*n3 is e n t i r e l y of
color 1.
1.53. At l e a s t one of the n2 planes of dimensions n 3 x « ! is entirely of
color 2.
1.54. At l e a s t one of the n3 planes of dimensions n\yn2 is entirely of
color 3.
1.55. The dogleg is e n t i r e l y of color 4.
1
11
I t is easy to see that Pr(i.52) is the complement 1 - (1 - p j 1) of the
7 l nl
probability (1 - p\ ) that none of the n\ planes of dimensions n 2 >n 3 be e n t i r e l y
of color l , with similar results for PK1.53) and P r ( i . 5 4 ) . Obviously Pr(i.55) =
p max{n l 9 n 2 9 n 3 } 9 SQ t h f t t
3
V fA tA ln\^\ . max{ni,n 2 ,n 3 } . .
II {1 - (1 - p . i) } + r l z 6
< 1,
t=l
- 226 -
and (1.51) follows from the fact that r = I-P1-P2-P3; the inequality degenerates
to equality i f and only i f n 1 = n 2 = n 3 = i . •
Finally, we state the generalization announced earlier.
1.6. THEOREM. Let k £ 2 be a given positive integer, and suppose that, for
i = 1929 ...sk, we have k veal numbers n. > 1 and k real numbers p. > 0 with
S = I p . £ 1. If N = Jin. and I. = N/n., then
k n
I (1 - pH)l > k -1 + ( i - s ) m a x { ^ } ,
%
i=i
with equality if and only if all the n. = 1.
The proof of this theorem in full generality requires calculus.
Figure 2.1
+ (l-,l)(1-,I-1)f...(l-,1« + 1 ) r t
) a
+ ( l - r l H l - r l - 1
) \ . . a - r i a + 1 ) + k
) Z
> 2 + s2l~\
In particular,
+ (l-,I)(l- ?
l
- 1
) 2
...(l- ?
4
« t 3 >
) 2
a i ( l + 3 ) 8
+ (l-,«)(l.I.^) ...(l-P )
> 2 + S 2 Z -S
and, if I = l (mod 4) and a = l ( Z - i ) ,
(2.21a) f(p) + /(?) + f(p) > 2 + s2Z""\
1-p I ^
>-1 - pZ-l ^> ... ^£ .1 - p
4(3Z+l)
H
,
and the number of factors in each term of (2.23) is the same as in /(p). •
The assertion
1 I 7 I 1 I
(2.24) (1 - y) + (1 - a1) + (1 - r) >2
is valid and is stronger than (2.23). But to obtain it a new type of diagram must
be used. This is discussed in the next section.
3. Special results obtainable from square arrays.
In this section, straightforward applications of the formula for the sum of
the probabilities of mutually exclusive events are used to obtain new inequalities.
3.1. THEOREM. If p^ > 0 for i = 1,2,3,4- and lp^ < 1, then
2 2 2
(3.2) (1 - p\) + (1 - p\) + (1 - p\) > 2,
3 3 33
(3.3) (1 - pi) + (1 - pi) + (1 - p \ ) + (1 - pi) > 3S
(3.4) (1 - pi) + (1 - p 2 ) + (1 - pi) > 2 9
(3.5) (1 - p i ) 5 + (1 - pi)5 + (1 - p s ) 5 + (1 - pi)5 > 3.
- 229 -
Proof of (3.2). Take four beads (numbered 1,2,3,4) and color them (with
probabilities pi, p 2 , P%) with colors <?i, o2, o3. The following events are mu-
tually exclusive:
1, 2, 3; I 1, 4 9 7; 1, 5, 9; 1, 6, 8;
4, 5, 6; 2, 5, 8; 2, 6, 7; 2, 4, 9;
7, 8, 9- 3, 6, 9. I 3, 4 5 8. | 39 5, 7.
The patterns for (3.4) and (3.5) should now be clear. In fact, the idea is
effective for four terms (like (3.5)) when the exponents are odd, and for three
terms (like (3.4)) when the exponents are even. To follow the idea, note three
things: (i) the rows of the second array are the columns of the first array; (ii)
the rows of the third array are the (broken) positive diagonals of the first array;
(iii) the rows of the fourth array are the (broken) negative diagonals of the
first array. D
The results that this method can deliver are limited. The inequality
2 2
2 2
(3.6) (1 - p \ ) + (1 Pi) + (1 P3> + (1 Ph) > 3
"Is true, but not obtainable from a picture with beads. Here is the impossibility
argument. The terms p2. imply that the sets being colored must have two beads each.
The terms (1 - pZ) imply that there are two sets of two beads each. But a total
of four beads can be partitioned into two such sets inonly three ways.
Proof. There are 21 intersection points marked with open or solid circles.
Place beads at the 21 points, and color the beads (at random) with colors ^ with
respective probabilities p. (1 * * * 5). The reader is invited to invent five
mutually exclusive events and thus derive (4.3). The fifth event, with probability
p7., is suggested by the seven solidly marked Intersection points. D
In an analogous fashion, Figure 4.4 leads to the next theorem.
THEOBEM 4.5. If pi > 0 for i = 1,2,...,5 and ZPi = 1, then
(4.6) (l - p\Ki - PV" + (i - P?)(i - pi? + <* " P ! > ( I " P S ) 2 ( 1 - P3)2
.13
+ (1 -p&U *plfV -Plf > 3 tps .
5^5
This result (4.6) is not a corollary of either of the inequalities j ( i - PiY > 3,
7 7
r d - p .) > 3 of Section 3.
5. Conclusion.
In this set of two articles (the first appeared in [1983: 128]), Boolean argu-
ments are used to derive some inequalities in which the parameters m, », I.. ^ are
Integers. No essentially analytical arguments are used. It is interesting that
the same inequalities are valid when the parameters are real. This extension will
be carried out in a more technical article; in it several other sharp inequalities
are obtained. The article will appear in 1985 in J. Math. Anal. & Appl.
Note that many corollaries of the theorems of this article can be written down
One of them is:
5.1. COROLLARY to ( 3 . 2 ) . If a,b,e > 0, then
A: 0000''0000
As written, in rows of thirteen elements, the sum of the digits in every column
of the two bracelets is zero. In the successive 78-digit quadrants of H9 the digit
sums are 4, 2, l, and 3, In J9 the successive quadrant sums are 2, 1, 3, and 4.
Two more appearances of the ubiquitous B.
In Table 2 are recorded the frequencies of occurrence of the five digits in
the successive four quadrants of H and J. In each case, like frequencies of non-
zero digits occur on diagonals slanting down toward the right.
In Table 3, the digits of each of n and j are consecutively arranged in
thirteen 24-digit rows. There the sum of the digits in each column is zero.
- 233 -
Table 2: Frequency of digits in fourth-order bracelets
Digits Digits
Quadrants 0 1 3 4 2 Quadrants 0 1 3 4 2
*1 18 17 16 15 12 Jl 13 14 21 20 10
H2 18 12 17 16 15 J2 13 10 14 21 20
#3 18 15 12 17 16 ^3 13 20 10 14 21
Hk 18 16 15 12 17 ^ 13 21 20 10 14
Bracelet H Bracelet J
It follows that the sum of the digits at the vertices of any regular thirteen-
gon inscribed in H or J will be zero.
The persistent B appears in the fifth row of #, and diagonally downward to
the right in alternate columns of J in the form of triads of like digits interwoven
with quartets of the same digits. The fifth row of H also contains two consecutive
sets of four consecutive digits in increasing order of magnitude; and the twelfth
row contains the five consecutive digits in decreasing order of magnitude.
The recordings of the two bracelets in three rows of 104 elements start out as
0 0 0 1 1 2 4 3 0 4 1 3 3
4 2 0 4 0 1 0 0 1 2 3 1 2
1 3 0 0 4 2 1 2 4 4 1 1 0
and
J: 1 1 1 1 4 2 3 0 4 4 1 4 3
4 1 2 2 4 4 2 2 2 0 1 0 3
0 3 2 2 2 4 0 3 4 1 3 1 4
.The sum of the digits in each of the columns above is zero. I t follows that the
sum of the digits at the vertices of any equilateral triangle inscribed in H or J
w i l l be zero. Indeed, the vertex d i g i t sum w i l l vanish for any inscribed regular
n-gon where n is any factor of 312 = 2 3 -3«i3.
Since a l l possible d i g i t quartets appear in A9 H, and J, in any clockwise
matching of two of these bracelets, the sums of the corresponding digits w i l l form
a bracelet wherein the law of formation holds, so i t must be a repeated A or some
orientation of H or Jo For example, an H (or J) added to another H (or j) that
has been rotated through 180° forms a bracelet of zeros. The operation H + J9
with the f i r s t d i g i t of J matched with the third d i g i t of H (from Table l ) , then
with the fourth d i g i t , and so on, until i t has been matched with every d i g i t of #,
produces a bracelet of 312 bracelets beginning, in order, with
H J J H H J J H J J J
- 234 -
Table 3: Rearranged fourth-order bracelets
H: 0 0 0 1 1 2 4 3 0 4 1 3 3 1 3 0 2 1 1 4 3 4 2 3
2 1 3 4 0 3 0 2 0 0 2 4 1 2 4 1 3 0 3 2 3 3 1 4
1 4 0 4 4 2 0 0 1 3 4 3 1 1 4 4 0 4 2 0 1 2 0 3
1 1 0 0 2 3 0 0 0 3 3 1 2 4 0 2 3 4 4 3 4 0 1 3
3 2 4 2 14 13 4 2 0 4 0 10 0 12 3 12 3 4 0
4 1 4 4 3 2 3 2 0 2 2 1 0 0 3 4 2 4 3 3 2 2 0 2
1 0 3 1 0 4 3 3 0 0 1 4 0 0 0 4 4 3 1 2 0 1 4 2
2 4 2 0 3 4 4 1 2 1 3 2 3 4 2 1 0 2 0 3 0 0 3 1
4 3 1 4 2 0 2 3 2 2 4 1 4 1 0 1 1 3 0 0 4 2 1 2
4 4 1 1 0 1 3 0 4 3 0 2 4 4 0 0 3 2 0 0 0 2 2 4
3 1 0 3 2 1 1 2 1 0 4 2 2 3 1 3 4 1 4 2 1 3 0 1
0 4 0 0 4 3 2 4 3 2 10 1 4 1 1 2 3 2 3 0 3 3 4
0 0 2 1 3 1 2 2 3 3 0 3 4 0 2 4 0 1 2 2 0 0 4 l'O 0 0 1
J: 111 2 0 4 2 3 4 3 2 2 1 3 3 4 1 1 4 0 1 1 1 3
1111 4 2 3 0 4 4 1 4 3 2 0 4 4 0 3 1 3 2 4 0
4 0 3 2 4 4 3 3 4 4 4 0 2 0 1 3 1 0 0 4 0 4 3 1
3 13 3 0 2 3 3 3 1 0 2 1 4 2 4 1 1 3 4 4 2 3 3
2 0 3 3 3 4 3 3 3 3 2 1 4 0 2 2 3 2 4 1 0 2 2 0
4 3 4 1 2 0 2 0 4 1 2 2 4 4 2 2 2 0 1 0 3 4 3 0
0 2 0 2 4 3 4 3 4 4 0 1 4 4 4 3 0 1 3 2 1 2 3 3
4 2 2 1 4 4 1 0 4 4 4 2 4 4 4 4 1 3 2 0 1 1 4 1
2 3 0 1 1 0 2 4 2 3 1 0 1 0 2 3 1 1 2 2 1 1 1 0
3 0 4 2 4 0 0 1 0 1 2 4 2 4 2 2 0 3 2 2 2 4 0 3
4 1 3 1 4 4 2 1 1 3 2 2 3 0 2 2 2 1 2 2 2 2 3 4
1 0 3 3 2 3 1 4 0 3 3 0 1 2 1 4 3 0 3 0 1 4 3 3
1 1 3 3 3 0 4 0 2 1 2 0 0 3 0 3 1 2 1 2 1 1 0 4'1 1 1 2
MATHEMATICAL CLERIHEWS
William Rowan Hamilton, Charles Lutwidge Dodgson (Lewis Carroll)
A toping and untrammelled son Gave lectures in a clerk's apparel,
Of the Hibernians, Was fond of logic, pun and riddle,
Devised quaternions. And also Alice Pleasance Liddell.
ALAN WAYNE, Holiday, Florida
- 236 -
THE OLYMPIAD CORNER: 48
M.S. KLAMKIN
Later on in this column I will give solutions to several problems proposed
here earlier. But first I give two new problem sets. The first consists of
the problems set at the final round of the 1982 Swedish Olympiad. The second
is a set of problems proposed in the March 1983 issue of Kozgpiekolai Matematikai
Lapok (Hungarian Mathematical Journal for Secondary Schools). As usual, for all
of these problems I solicit elegant solutions, which should be sent directly to
me at the address given at the end of this column.
*
x2 - lx2l = (x - Car])2
are there in the interval l < x < Nl
2, Let a9b9a be positive numbers. Prove that
3, Suppose one can find a point P in the interior of the quadrilateral ABCD
such that the four triangles PAB, PBC, PCD, and PDA have the same area.
Show that P is on one of the diagonals AC or BD.
4, In the triangle ABC the sides are AB = 33 cm, AC = 21 cm, and BC = m cm,
where m is an integer. It is possible to find a point D on AB and a
point E on AC such that
AD = DE = EC = n cm,
where n is an integer. What values can m take?
5, In an orthonormal coordinate system one considers the points (x9y)9
where x and y are integers with l < x < 12, l < y < 12. Each of these
144 points is coloured red, white, or blue. Show that there is a rectangle with
sides parallel to the axes and having all its vertices the same colour.
F, 2 4 1 2 , We have d cartons and one box. The cartons are numbered from 1
to d and some of them (maybe all) contain balls. We want to
collect the contents of all the cartons in the box. Carton i may be emptied
if it contains exactly i balls, and this is done in such a way that one ball is
placed into the box and the remaining i-\ balls are placed one by one into the
cartons l, 2, ..., i-i, respectively. For which n is it possible to place n balls
into an appropriate number of cartons so that they all could be collected?
F, 2 4 1 3 , Given is a square ABCD. Find the locus of the points P for which
PA + PC = /2 max {PB, PD}.
P, 375, Does there exist a function f:B + R such that 11m fix) = °° and
^1n(1n(.f!?L)...)r °
holds for all n (where n is the number of logarithm functions in the denominator)?
- 238 -
P, 376, Determine the sizes of those circles whose interior points can be
painted with two colours in such a way that the endpoints of all
line segments of unit length have different colours.
Determine P(n+l).
2, [1982: 270] Let (a\9 a2, ..., a , ..,) be a sequence of positive real num-
2
bers such that a < a - a ^ for all n. Show that a < l/n for all n.
n n n+1 w
Solution by K.S, Murray3 Brooklyn, N.Y.
The defining relation shows that we must have a2 < aid-ai). Thus
- 239 -
0 < ai < l and a2 ^ n < J.
We now use a „ < a (±-a ), n = 2,3,4,... and proceed by induction. Assume that
n+l n n
a^ < i/fc for some & > 2. Since #(!-#) is increasing in the interval [o,J], we have
3, C1982: 270] A round track has n fueling stations (some possibly empty) con-
taining a combined total of fuel sufficient for a car to travel once around
the track. Prove that, irrespective of the initial distribution of fuel among the
stations, it is always possible for a car with an empty tank to start from one of
the stations and complete a round trip without running out of fuel on the way.
Comment.
See Crux 354 [1979: 57] for a solution and for references to other appearances
of this problem by various proposers. However, the version given here, from a 1964
Peking Mathematics Contest, predates all the other references known to us at this
time.
I, [1983: 137] Which of (17091982! ) 2 and 17091982 1 7 0 9 1 9 8 2 is greater?
Solution by Noam D. Elkies3 student, Columbia University,
The result follows by setting n = 17091982 in the inequality
(n\)2 > nn9 n > 2, (1)
18-3, C1983: 107, 141] Show how to construct a triangle having its vertices
on three given skew lines so that the centroid of the triangle coincides
with a given point.
II. Solution by Howard Eves, University of Maine.
Denote the three mutually skew lines by Z, m, n, and the given point by G, as
shown in the figure. Let i' be the map of I under the homothety of center G and
- 240 -
ratio - 1 . Let p be the plane midway between the plane through m parallel to n and
the plane through n parellel to m\ this plane is the locus of midpoints of a l l
segments joining points of m to points of n. Let p cut V in A'. Let A'G cut I
in A. Let the line of intersection of the plane determined by A* and m with the
plane determined by A' and n cut 777 in B and n in C. Then ABC is the sought
triangle.
I f V lies in p, any point on V may serve as point A ' , and the problem has
i n f i n i t e l y many solutions. I f V is parallel to p, the problem has no solution.
Comment by M.S.K.
I also received a solution equivalent to the above from Esther Szekeres (New
South Wales, Australia). These two solutions were submitted in response to my
request for a synthetic solution, following an analytic solution I had given
earlier [1983: 14-1].
Editor's note. All communications about this column should be sent to Pro-
fessor M.S. Klamkin, Department of Mathematics, University of Alberta, Edmonton,
Alberta, Canada T6G 2G1.
A A A
HOLLYWOOD ARITHMETIC
Item by Leonard Lyons in the September 1976 Reader's Digest, page 90: An
actress who was offered a co-starring role with Zero Mostel declined, because
Mostel is too overwhelming. "I'd be lost working with him. One plus Zero would
still be Zero."
& ft :'c
- 241 -
PROBLEMS--PROBLEMES
Problem proposals and solutions should be sent to the editor, whose address
appears on the front page of this issue. Proposals should, whenever possible, be
accompanied by a solution, references, and other insights which are likely to be
of help to the editor. An asterisk (*) after a number indicates a problem submit-
ted without a solution.
Original problems are particularly sought. But other interesting problems
may also be acceptable provided they are not too well known and references are
given as to their provenance. Ordinarily, if the originator of a problem can be
located, it should not be submitted by somebody else without his permission.
To facilitate their consideration, your solutions, typewritten or neatly hand-
written on signed, separate sheets, should preferably be mailed to the editor before
March 1, 1984, although solutions received after that date will also be consid-
ered until the time when a solution is published.
SOLUTIONS
No problem is ever permanently closed. The editor will always be pleased
to consider for publication new solutions or new insights on past problems.
721, [1982: 77; 1983: 87] Proposed by Alan Wayne, Pasco-Hernando Community
College, New Port Rickey, Florida.
A propos of the editor's comment following Crux 611 [1982: 30], verify that,
with decimal integers,
(a) uniquely, TRIGG is three times WRONG;
(b) independently, but also uniquely, WAYNE is seven times RIGHT.
I l l , Letter to the editor by Stewart Metchette, Culver City, California.
Per your comment to Problem 721 [1983: 89]:
I hope digitomaniacs continue their e f f o r t s , even at the risk of repeating
what has been done before. For example, you quote Victor Thebault's work [1] on
the 2-factor digital products: i f your quote is accurate, his solution l i s t for
the i o - d i g i t case is incomplete; i f inaccurate, you have done him a disservice.
There are an additional 9 and 6 solutions (for a total of 13 and 9) for
A-BCDE = FGHIJ and AB-CDE = FGHIJ,
respectively. These are listed by Madachy [2] and attributed by him to Charles
L. Baker.
You go on to say that "finding all products containing each of the digits
exactly once... has already been done." In Madachy's book, he mentions that
some 3-factor solutions might be possible, but knew of none.
Inspired by Madachy's comment, I ran a computer check and reported the results
in [ 3 ] :
A-BC-DE = FGHI, 12 solutions;
A«BC«DE = FGHIJ, lo solutions;
A«BC«DEF = GHIJ, 2 solutions.
REFERENCES
and
The r e s u l t i n g
* ~ s i n f = cota
X - COS 2w
and
show that, for fixed co and variable t , the locus of (x9y) is a segment PQ of
- 245 -
length 4 whose midpoint (cos 2OJ, sin 200) lies on the unit circle y with centre at
the origin 0. The endpoints of PQ correspond to t = -OJ±|; their coordinates are
and
x~ = cos 203 + 2sino), z/g = sin 203 + 2 cos 03.
Since PComr) = Q(cu), i t follows that, as 03 varies, P and Q describe the same curve.
This curve, shown in the figure, is a 3-cusped hypocycloid, or deltoid, generated
by a fixed point on a circle y of radius 1 which rotates without slipping inside a
circle $ of radius 3 with centre at the origin 0. The interior and boundary of the
deltoid constitute the desired locus. Its area is 2TT and i t s perimeter is 16.
(b) I f A,B,C are the angles of a triangle (possibly degenerate), then we have
the additional restrictions
- 246 -
With these restrictions and for fixed cu, the segment PQ is either vertical or else
i t has a positive slope cotco. The coordinates of the lower endpoint P are obtained
by setting t = o. They are
so, as co varies, P describes the segment RT shown in the figure. The coordinates
of the higher endpoint Q are obtained by setting t = -OJ+|. They are
As co varies, Q describes the arcs RS and ST of the deltoid. The required locus
consists of the interior and boundary of the mixtilinear triangular lamina RST
(shaded in the figure). The segment RT contains the images of a l l degenerate
triangles, the arcs RS and ST the images of a l l isosceles triangles, the point S
is the image of the equilateral triangle, and a l l scalene triangles are mapped
into the interior of the lamina.
Editor's comment.
Referring to the figure, suppose the tangent to the deltoid at T meets the
arc RS in U. The proposer showed that the segment TU contains the Images of a l l
right-angled triangles, so U is the image of the isosceles right-angled triangle;
and that the parts of the shaded lamina above and below TU contain the images of
a l l acute-angled and a l l obtuse-angled triangles, respectively.
Suppose a l l triangles are labelled so that A £ B < C, and suppose the tangent
to the deltoid at S meets the segment RT in V, As Murty showed earlier in this
journal [1982; 64-65], the points of segment VS correspond to triangles in which
B = TT/3; the points of the shaded lamina above and below VS correspond to the
triangles with B > TT/3 and B < TT/3, respectively; and the point V corresponds to
the degenerate triangle with angles o, TT/3, and 2TT/3,
I f TU n VS = W, then W corresponds to the triangle with angles TT/6, TT/3, and
TT/2. Finally, R and T correspond to the degenerate triangles with angles o, o, TT
and o, TT/2, TT/2, respectively.
s'« * &
- 217 -
760, C1982: 175] Proposed by Jordi Dou, Barcelona, Spain,
Given a triangle ABC, construct with ruler and compass, on AB and AC
as bases, directly similar isosceles triangles ABX and ACY such that BY = CX.
Prove that there are exactly two such pairs of isosceles triangles.
X(£+2ap, a+2bp)
and
Y(c+2ap9 a+2cp)9
where p is an arbitrary constant. (As a check, we find that each of the angles
BAX and CAY then equals Arccos l / / l + 4 p z . )
Now BY = CX i f and only i f
(-2&+c+2ap)2 + (a+2cp)2 = (-2c+b+2ap)2 + (a+2bp)2,
an equation equivalent to
4(£+c)p2 + 16ap - 3(b+c) = 0.
1523
1523
1523
1523
6514*
II. Solution to part (b) by the proposer.
The alphametic is equivalent to the following system of equations, which are
written in base ten:
2S + I = R + 8 e l f (1)
2N + o1 = So2, (2)
I + W + o2 = 8c? 3 , (3)
2E + Z + o3 = V. (4)
2E + Z + l = V; (4')
I + W + c2 = 8; (3')
I
S =^ " + 4*1. (!')
Editor's comment,
Feder noted that part (a) also has a unique solution in base ten:
1329
1329
1329
1329
5316 *
762, C1982: 209, 278] (Corrected) Proposed by J.T. Groenman3 Arrihem, The
Netherlands.
ABC is a triangle with area K and sides aJb9o in the usual order. The i n -
ternal bisectors of angles A,B,C meet the opposite sides in D,E,F, respectively,
and the area of triangle DEF is K\
(a) Prove that
3abc < K^_ < 1
hia^b'^o6) " K " 4'
CAFE] AF-AE be
K " AB^SC ~ (a+b)(a+c)'
Since fc+t?+6|9i, we have b+c+6 = 13 or 91, so b+c = 7 or 85, and the corresponding
values of be are 12 and 2250. Only the pair (b+c9 be) = ( 7 , 12) is s a t i s f a c t o r y ,
and i t leads to the unique solution {b,<?} = {3,4}.
Also solved by W.J. BLUNDON, Memorial University of Newfoundanld (part (a)
also); JORDI DOU, Barcelona, Spain; G.C. GIRI, Midnapore College, West Bengal,
India (part (a) only); RICHARD I. HESS, Rancho Palos Verdes, California (part (b)
only); ROBERT S. JOHNSON, Montreal, Quebec (part (b) only); M.S. KLAMKIN,
- 251 -
University of Alberta (part (b) also); KESIRAJU SATYANARAYANA, Gagan Mahal Colony,
Hyderabad, India; and the proposer.
REFERENCE
1 1 q
m n p
REFERENCE
765, C1982: 210] Proposed by K.P. Shum and R.F. Turner-Smith, The Chinese
University of Bong Kong.
- 252 -
If n is a given positive integer, find all solutions 6 e [O,2TT) of the equa-
tion
cosne + sinne = 1.
(The trivial case n = 2 may be omitted.)
Solution by M.S. Klamkin, University of Alberta (revised by the editor).
For n = l, we have
( o, | 5 for odd n;
0
= (^
7r 9 for even n 2
' °' 1' ' T **
Combining the cases n = l and n > 3, we find that (4) gives all the solutions
for all n * 2. D
As a generalization, we will characterize all r-dimensional unit vectors
tt = (tti, w 2 , ..., u ) and ^ = (i?i, t>2. ...» ^ )>
where m > o and n > l are integers. (The proposed problem corresponds to the case
r = 2, m = 0.)
Case 1: m - o. Here, of course, the pair (#,£) satisfies (5) if and only
if -t satisfies
- 253 -
H=l. (6)
and
\v. | = 1 and v. = o for i * io (9)
for some i0 e {i,2,...,r>. The vectors £ whose components satisfy (8) and (9) are
characterized by
n
v- = ±1 and u . = o f o r t * £n» i f i s even.
i v) c= i). do)
\t=l / t=l £=1
If the pair ($9$) satisfies (5), then equality must hold in (10), and hence
2
u.^ = kv.t and u.v.
t -z,
= kv»9i i = l,2,...,r,
for some constant k. But (5) holds only for k - i, so the only solutions are the
pairs ($$) with £ = £.
- 254 -
Case 3: m > l, n > 2. Here we will use the fact that, for a l l vectors ft and £,
r r
z
l ufjl
^ ^ s I v (= 1) (ID
i~-± *=1
Now the pair (#,£) satisfies (5) if and only if equality holds in (11), that is,
if and only if
u.vt > o for i = 1,2,...,v (12)
and
i and U. = v. = o for i * i§ (13)
W W
for some £ 0 e U,2,...,r}. The vectors £ and £ whose components satisfy (12)
and (13) are characterized by
u. = v. - 0 for i * i0 and
Editor's comment.
Several solutions involved lengthy calculations (some with complex numbers)
resulting from the use of Heron's formula. And several solvers, trusting souls
that they are, blithely set out to calculate Kl without first assuring themselves
that there was a triangle Tx .
An interesting fact brought out by the above solution is that Tx is similar
to the pedal triangle T\. Eves noted the following special results: if p = 2i?,
then K* = K; and if p = r, the inradius of T, then Kl = iff. Finally, Klamkin
showed that Rx, the circumradius of T\ is not constant for all positions of P
- 256 -
on y', but that
REFERENCES
1. David C. Kay, College Geometry, Holt, Rinehart and Winston, New York,
1969, p. 270.
2. Roger A. Johnson, Advanced Euclidean Geometry (Modem Geometry) , Dover,
New York, i960.
* ft ft
1 = 2