Crux v11n03 Mar
Crux v11n03 Mar
Crux v11n03 Mar
Mathematicorum
Published by the Canadian Mathematical Society.
http://crux.math.ca/
CRUX M A T H E M A T I C O R U M
V o l . 1 1 , No* 3
March 1985
Sponsored by
Carleton-Ottawa Mathematics Association Mathematique dsOttawa-€arl@to»
Publie par le College Algonquin, Ottawa
The assistance of the publisher and the support of the Canadian Mathematical Society*
the Carle ton University Department ©f Mathematics and Statistics , the University of Ottawa
Department of Mathematics, and the endorsement of the Ottawa Valley Education Liaison Council
are gratefully acknowledged.
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
CONTENTS
_ 69 -
- 70 -
THE OLYMPIAD CORNER: 63
M.S. KLAMKIN
The 1985 International Mathematics Olympiad will be held in Finland from June
29 to July 11. Each invited country can send a team of up to six students. For
further information, write to: The IMO-85 Organizing Committee, Department of
Mathematics, University of Helsinki, Hallituskata 15, SF-ooioo, Helsinki, Finland.
I now extend the list, begun last month, of problems proposed by various par-
ticipating countries (but unused) in past International Mathematics Olympiads. (I
am grateful to L.M. Glasser for translating the problems from Mongolia and the
U.S.S.R.) As usual, I solicit from all readers elegant solutions to these problems.
26. Proposed by Australia.
Let xi9x2>* .. 9 x denote n real numbers lying in the interval [0,1]. Show
that there is a number x e [0,1] such that
n
1 1
n.LA % 2
27 i Proposed by Australia.
Let ABCD be a convex quadrilateral, and let A ^ B ^ O ^ D x be the circum-
centers of triangles BCD, CDA, DAB, and ABC, respectively.
(i) Prove that either all of A l9 Bx,C l5 Dx coincide in one point, or else they
are all distinct. Assuming the latter case, show that A x and Cx are on opposite
sides of line B ^ , and that Bx and dl are on opposite sides of line A ^ . (This
establishes the convexity of quadrilateral AiBjCiDi.)
(ii) Let A 2 ,B 2s C2,D2 be the circumcenters of triangles B ^ D x , CxDxA l9 D ^ B x ,
and AxBxCx, respectively. Show that quadrilateral A2B2C2D2 is similar to quadri-
lateral ABCD.
28 1 Proposed by Belgium.
Determine all integers x such that
Xk + X3 + X2 + X + 1
is a perfect square.
29 i Proposed by Belgium.
Determine all integer solutions (x,y) to the Diophantine equation
. x3 - y3 - 2xy + 8.
- 71 -
30i Proposed by Brazil.
A box containsp white balls and q black bal1s^ and beside the box lies a
large pile of black balls. Two balls chosen at random (with equal likelihood) §pe
taken out of the box. If they are of the same color* a b U c k ball from the pile
is put into the box; otherwise, the white ball is put back Into the box. The pro-
cedure is repeated until the last two balls are removed from the box and one U s t
ball is put tn. What Is the probability that this last ball Is white?
AS = ^iS 2 C0S 2 ~.
33 • Proposed by Bulgaria.
Given are a circle r and a line I tangent to it at B. From a point A on r,
a line AP i I is constructed, with P e L If the point M is symmetric to P with
respect to AB, determine the locus of M as A ranges on r.
-1 -1 -1 -1
(a + ab a) + (a + b) -a
-1 -1 ~1
where x is that element for which x x = xx = e, the multiplicative identity.
36 i Proposed by Czechoslovakia.
Let
S = {-===== I rn9n p o s i t i v e i n t e g e r s } .
- 72 -
Show that for each (x9y) e S*S with x < y there exists z e S such that x < z < y.
37 g Proposed by Finland.
Four circles C9Ci3C29C3 and a line I are given, all in the same plane. The
circles C l 5 C 2 ,C 3 are all distinct, each touches the other two and touches also C and
I . If the radius of C is 1, determine the distance between its center and I.
3 88 Propos ed by Franc e.
Let the numbers be all positive and let
v
ls'k ~ v^i^2* • »w, , fc = 1,2,...,n.
Prove that
n
K
k=± fc=i *
39 , Proposed by Mongolia.
A country has n cities, any two of which are connected by a railroad. A
railroad worker has to travel on each line exactly once. If at any stop there is
a city he must reach but cannot (having already traveled on the line to or from that
city), then he can fly. What is the smallest number of plane tickets he must buy?
40 . Proposed by Poland
Let r be a unit circle with center 0, and let Pl5P2,...,P be points of r
such that n
OPi + 0P 2 + ... + OP = 0.
42 i Proposed by Rumania.
I f ( 1 + X + X2 + X3 + X^) = a 0 + CLiX + . . . + # 1 9 8 ^ 1 9 8
\
(i) determine the g.c.d. of the coefficients a 3 , a 8 , a 1 3 , ..., &1983'
( H ) show that 1 0 3 4 7 > a 9 9 2 > 10 3L+0 .
43 i Proposed by Spain.
Solve the equation
tan22r + 2 tan 2x tan 3x - 1 = 0.
- 73 -
44. Proposed by Sweden.
Let n be a positive integer having at least two distinct prime factors.
Show that there is a permutation (al,a2^ * * ,a ) of (i,2 9 . .. ,n) such that
n 2ua,
J k c o s - — - = o.
are tan A 9 tanB, tan C 9 where A 9 B 9 C are the angles of a triangle. Determine the
fourth root as a function of (only) p , q3 r, and s .
46 , Proposed by the U«S»S»B*
Let (p..) be a given m*n matrix with real entries, and let
n m
A. = Jp . . and 5. = Ip... (1)
and
0 < x < 1 => 0 < a < |, 0 < 3 < ^;
and in each case cos a = s i n g = Jl~-x2. In each case a l s o , 0 < a+3 ^ TT and
f(x2) - f(Xl)
< 1
x2 X-]
BM'xBN' J(BA+BC)x(BC+BD)
4
o,
where in line (1) we have used the given area ratios.
Thus B J M ' J N 1 are collinear and r = 1/2.
a + b
•2.
AC-EM-DB = CE-MD-BA,
and the desired result follows from the fact that IABI = I AC I
- 77 -
L\, [1984-: 14-1] From the 1981 Leningrad High School Olympiad,
x u u T*
Given that x + yJ = xr + y , where a: and y are natural numbers, prove that
x = y.
<=> (A - C)®(B - D) = 0
<=> the diagonals are perpendicular.
from which b - -2, then m = ±1 and the two common tangents are perpendicular.
- 78 -
7, [1984: 142] From the 1981 Leningrad High School Olympiad.
The angle bisectors AD, BE, CF of a triangle ABC concur in the incenter I.
If
£1 - §1 _ CI
ID " IE " IF'
prove that triangle ABC is equilateral.
a?te+l)te+2)...(a?+1980)(a:+1981) = 1
x 1 9 8 2 + ... + 1981!#,
2 3 + 3 4 = 89, a prime.
Finally* « = 1, and a = Z? = 2.
13. [1984: 142] From the 1981 Leningrad High School Olympiad.
In the convex quadrilateral ABCD9 /BAC = /CBD and /ACD = /BDA. Show that
2 2 2
IACI = IBCI f lADI .
- 80 -
Solution by M.S.K.
With the angles 0,a,3 as shown in the
figure, we note that /ABC = /ADC = 6 and then,
by the law of sines,
AC BC AD ,
sin 0 " sina " sin 3 " A> Sdy
"
~BD»ACsin0 = ~BD*BCsina + | B D « A D s i n 3 ,
from which
fr AC r BC n AD
AC*-r- = DBC»-=- + AAD*-y->
\\\% [1984: 142] From the 1981 Leningrad High School Olympiad.
Jb+c jc+a Ji+b ^a+b+c+1
Show that 2"'" + 2"'"- + 2"'" < T + 1 for all a9b9c > 0.
^a+b+c •a
obtained by dividing throughout by 2°""'^ and then setting x = 2 ,y = 2 , and
z = 2 , so that 0 < x9y9z < l.
Since F(x9y9z) is linear in each variable, it takes on its greatest and least
values at the ends of the intervals, that is, at the corners of the cube with ver-
tices (o,o,o), (i,o,o), ..., (1,1,0), ..., (1,1,1). Since this least value is
zero, F(x9y9z) > 0 for all points (x9y9z) in the interior of the cube.
J^, [1984: 142] From the 1981 Leningrad High School Olympiad.
Show that, for all positive x and y9
Solution by M.S.K.
More generally, for x,y^m,n > 0 consider
81 -
m . m _ m n]x_ m@ nj^o
x +y < x +
fy ^ Ix S
Since the case x = z/ is immediate, we can assume without loss of generality that
y = Xx with X > 1. The inequality then reduces to
(x « + l/» _ l ) ( x -l/„ . 1} ^ 0§
p p m I a?. )l/n
r
dido*• .a ) a. £ ) a* .
In closing,I would ask all readers submitting solutions to any of the problems
in this or in past columns to clearly identify the problems by stating their num-
bers and the place (year and page number) where they originally appeared.
Editorss 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.
SERGE
DE
NlMES
DENIM
maximize NIMES (where 1 = 1 ) , the city in southern France that gave its name to
denim cloth.
1023 B From a Trinity College3 Cambridge3 examination paper dated June 73 1901.
Show that, for n = 1,2,3,.,,,
n
v 2 3TT 3 1
Y Arctan r-o- = A r c t a n — Arctan ~—•-.
^ k2 *+ n n+1
102^8 Proposed by William Tunstall Pedoe3 student3 The High School of Dundee3
Scotland*
Prove that an odd number which is a perfect square cannot be perfect.
- 83 -
1025, Proposed by Peter Messer3 M»Dej Mequon3 Wisconsin.
A paper square ABCD is folded so that A
vertex C falls on AB and side CD is divided into
two segments of lengths I and m% as shown in the
figure* Find the minimum value of the ratio l/m.
abcd(- + r + - + J) = (a + b + c + d)2
a b e d
4-5 = 6 (mod 7) .
ap + bq < or.
- 84 -
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.
857, [1983: 179; 1984: 304; 1985: 20] Proposed by Leroy F6 Meyers^ The Ohio
State University.
(a) Given three positive integers, show how to determine algebraically (rather
than by a search) the row (if any) of Pascal's triangle in which these integers
occur as consecutive entries.
(b) Given two positive integers* can one similarly determine the row (if any)
in which they occur as consecutive entries?
(c)" The positive integer k occurs in the row of Pascal's triangle beginning
with l, k$ ... . For which integers is this the only row in which it occurs?
III. Comment on part (c) by David Singmaster3 Polytechnic of the South Bank3
London^ England.
In 1971, I examined N(a)9 the number of times the integer a occurs in Pascalfs
triangle [l]. I showed that N(a) = O(loga) and conjectured that N(a) was bounded.
The proposer's part (c)" asks to characterize those a for which N(a) = 2. Abbotts
Erdos, and Hanson [2] examined this and showed that the normal and average orders
of N(a) are both 2. (See [4] for definitions of normal and average orders.) The
normal order assertion implies that N(a) = 2 for almost all integers (in the usual
sense of density). Specificallys if g(x) is the number of integers a < x such
that N(a) > 25 then they showed easily that g(x) - OUx),
In [3], I reported on my search for repeated binomial coefficients. There are
infinitely many a with N(a) > 69 obtained by solving
n
(n+1) = ( )
leading to
F F - 1 k = F F - 1
*2i+2r2i+3 ±9 K
*2i 2i+3
where the F. are the Fibonacci numbers9 beginning with F 0 = 0. The only other non-
h8
trivial repeated binomial coefficients up to 2 are the following:
Note that the last equality is the case i = 1 of the above general solution. Note
also that #(3003) = 8, and 3003 is the only known a with N(a) > 8.
Perhaps some reader can investigate this question further.
REFERENCES
the X doubles as a multiplication sign. Find (a) the fewest HITS, and (b) the most
HITS, that the BOXER can deliver.
I. Solution by Edwin M. Klein3 University of Wisconsin-Whitewater.
Here are the 26 (with B0 > ER) computer-generated solutions to the alphametic,
listed in order of increasing HITS, so solutions 1 and 26 are the answers to parts
(a) and (b), respectively.
1. 46x23 = 1058 10. 58x37 = 2146
2. 57x24 = 1368 11. 56x39 = 2184
3. 49x32 = 1568 12. 65x38 = 2470
4, 69x23 = 1587 13. 65x48 = 3120
5. 42x38 = 1596 14. 82x45 = 3690
6. 52x34 = 1768 15. 64x58 = 3712
7. 58x34 = 1972 16. 85x46 = 3910
8. 52x38 = 1976 17. 68x59 = 4012
9, 54x39 = 2106 18. 76x53 = 4028
- 86 -
19. 79x52 = 4108 23. 74x68 = 5032
20. 78x59 = 4602 24. 87x62 = 5394
21. 92x53 = 4876 25. 85x74 = 6290
22. 86x57 = 4902 26. 95x72 = 6840
Editorfs comment.
This is the sort of problem that a computer can solve at the flick of a bit,
and most solvers availed themselves of this fact (or at least sent in only answers,
however obtained). The few who chose to slug it out by brute force for the edifi-
cation of the editor managed to find few labour-saving finesses. They won the match,
but they ended up exhausted on the mat, almost too weak to claim the fruits of
victory.
* * a
a In 8 2 1
s ~ 50
for all triangles.
(b) As stated in the proposal, this question is trivial and the answer is no.
It suffices to have P(t) sufficiently close to Q(t) for all t to ensure that the
midpoint M of PQ moves in the same sense as P and Q 9 In each of the two cited
examples, "PQ bisects the area" and "PQ bisects the perimeter", the correspondence
between P and Q is an involution, and perhaps the proposer tacitly assumed this
condition. But even with this additional condition the answer is still no. We
give three examples which shed some light on this matter.
With the same notation as in the figure, we assume that when P is at A,B,D,
then Q is at D,C,A, respectively. Let p and q be the distances, measured along the
perimeter in the counterclockwise sense, from A to P and Q, respectively, where
Also solved by J.TS GROENMAN, Arnhem, The Netherlands (part (a) only); RICHARD
I. HESS, Rancho Palos Verdes, California; WALTHER JANOUS, Ursulinengymnasium, Inns-
bruck , Austria; and the proposer.
Figure 2
II. Solution to parts (b) and (o) by Leon Bankoff3 Los Angeles3 California.
(b) One of the standard properties of an ellipse is that any tangent makes
equal angles with the focal radii at the point of contact. Accordingly if, as
shown in Figure 2, H 2 is the reflection of H in side BC, then 0H X n BC = L, the
point of contact of the ellipse with side BC. The straightedge and compass con-
struction of L is therefore obvious- Similarly, 0H 2 n CA = M and 0H 3 n AB = N §
where H 2 and H 3 are the reflections of H in CA and AB f respectively.
- 90 -
(c) As noted earlier in this journal [1977: 114], the concurrency of AL 3 BM f CN
is an immediate consequence of Brianchon's Theorem, when ABC is considered as a
degenerate hexagon. D
Figure 2 illustrates the relationship between the circumcircle, the nine-point
circle, and the ellipse with foci H and 0 inscribed in acute triangle ABC*
The nine-point circle turns out to be the auxiliary circle of the inscribed
ellipse, whose major axis PQ is therefore equal to the circumradius of the triangle.
The reflections of H in the sides of the triangle lie on the circumcirole, thus
suggesting an alternate method for locating the points L,M,N. The nine-point circle,
centered at N f , the midpoint of OH, with radius half that of the circumcircle, can
be drawn, and it passes through the feet of the three altitudes of ABC; these alti-
tudes can then be extended to H 1 ,H 2 s H 3 on the circumcircle and, as before, we have
OHj n BC = L, etc.
Also solved by LEON BANKOFF, Los Angeles, California (also part (a)); JORDI DOU,
Barcelona, Spain? ROLAND H. EDDYf Memorial University of Newfoundland; J9T* GROENMAN,
Arnhem, The Netherlands; DAN PEDOE, University of Minnesota; BASIL C. RENNXE, James
Cook University of North Queensland, Australia; and the proposer (also parts (b) and
(c)).
REFERENCE
1. Luigi Cremona9 Elements of Projective Geometry^ Dover9 New York, 19605 pp.
260-261.
& it &
X y z x y z
s f
[MCB ] = — 0 0 1 [MC B] = x y 0
z+x s+a? x+y a?+j/'
X 0 z 0 1 0
With these and similar r e s u l t s , equation ( l ) in the proposal becomes
xy yz zx _ zx xy yz
z+x x+y y+z ~ x+y y+z z+x9
(y - 2 ) ( s - ^ ) ( x - j/) = 0.
Thus M(a:,y,s) lies on the locus if and only if at least two of x9y9z are equal The
required locus is therefore
I n m n n9
where l9m3n are the lines containing the medians through A,B,C, respectively.
Also solved by JORDI DOU* Barcelona, Spain? J*T. GROENMAN, Arnhemf The Nether-
lands; WALTHER JANOUS, Ursulinengymnasium* Innsbruck, Austria; DAN PEDOE, University
of Minnesota? and the proposer*
* * *
it follows that quadrilateral BMNC is cyclic, and hence that /BMC = /BNC0
- 92 -
Also solved by JORDI DOU# Barcelona, Spain? RICHARD I. HESS, Rancho Palos
Verdes, California; WALTHER JANOUS^ Ursulinengymnasium, Innsbruck, Austria; and
the proposer*
ft ft ft
ft
906 i [1984: 19] Proposed by E.J. Eckevt and P.d. Vestergaard3 Institute of
Electronic Systemss Aalborg University Centre3 Denmark*
Let P denote the set of rational points on the unit circle c, that is s the
set of all points (r/t9 s/t) where r9s9t are integerss t > 0 9 and r2+s2 = t2B It
is known that P is dense in C« Let T be the subset of P for which t is prime* Is
T dense in CI
Editorfs comment.
No solution was received for this problem, which therefore remains open. For
a proof that P is dense in C9 see Problem 109 [1976: 81].
ft ft ft
x = q . - j (mod q2).
<J V
increasing order), and so x+j is not the sum of two squares of integers. Hence
each of the k positive integers
is not expressible as the sum of two squares of integers5 and all positive integers
congruent to these modulo (<?i^2- • °<7T, ) 2 are
likewise not so expressible.
- 93 -
For k = l,2 5 9 . 9 9 7 9 the smallest fc-tuples of consecutive positive integers that
are not sums of two squares of integers are
3
6 7
21 22 23
21 22 23 2*4
75 76 77 78 79
91 92 93 9694 95
REFERENCE
We show that cos A cos B cos C ^ 0, If cos A = 0 S for example, then X = 0, so also
cos B = cos C = 0 and A+B+C = 37r/2f a contradiction. Now, from (2), X * 0 and
- 94 -
iPx tan A tan B tan C t$n A + tan B + tan C -
(_) = —-— = . = -z—. —. . =x say.
k(a + 3 + y) = k3a3Y*
so k2 = (a+3+Y)/a3Y anc
I
P f g(g+B+Y) »a/2
T / . /f eB(a+g+Y)
/ ^ a ^ ^f/ 2 „ri wY(a+g+y)
« ^ e ^ iY/2
\'
max \(a+B)(a+Y)) * \(3+Y)(B+a)/ UY+a)(Y+3)J
Editorfs comment.
The proposer noted that a closely related problem appears without solution in
Hardy [2], where it is credited to Math, Tripos. 1935. This problem asked only to
show that (3) holds when p is a maximum.
REFERENCES
909 i [1984: 201 Proposed by Stan Wagon3 Smith College3 Northampton, Massa-
chusetts.
For which positive integers n is it true that, whenever an Integer's decimal
expansion contains only zeros and ones, with exactly n ones, then the integer is
not a perfect square?
4251 2 = 18071001
24499 2 = 600201001
124499 2 = 15500001001
2S1249 2 = 79101000001
375501 2 = 141001001001
425501 2 = 181051101001
781249 2 = 610350000001
10257492 = 1052161011001
145262492 = 211011910010001
167687512 = 281191010100001
389757492 = 1519109010111001
434742512 = 1890010500011001
1552807492 = 24112111010001001
2M49499992 = 60000502010100001
2847807492 = 81100075001001001
100503755012 = 101010047711101001001
316242187512 = 1000091211611100000001
3165297807492 = 100191102101010011001001
5000000000000000012 = 250000000000000001000000000000000001
Comments were received from FRIEND H. K1ERSTEAD, JR.^ Cuyahoga Falls, Ohio;
and the proposer.
- 96 -
910 i C1984: 20] Proposed by 0. Bottema3 Del ft 3 The Netherlands*
Determine the locus of the centers of the conies through the incenter
and the three excenters of a given triangle.
Editorfs comment.
A related problem stated and proved in Milne [5]: The locus of the foci of all
parabolas touching the three sides of a triangle is the circumcircle of the triangle,
REFERENCES
9111 [1984: 53.1 Proposed by Allan Wm. Johnson Jr,3 Washington3 D«C&
Solve the f o l l o w i n g synonymical alphametic \n the smallest possible base:
HARD
SHARP.
HARSH
- 97 -
Solution by Friend E9 Kierstead3 Jr«3 Cuyahoga Falls9 Ohio.
From the column sums we have
D + P = H + bi9
2R = S + bj - i9
2A = R + bk - j 9
2H = A + bm - k9
S = H - m9
where b > 6 is the base and i9j3k9m are the carries. It is immediately obvious
that 777 = 1, so H = S+lo Then it is easily shown that
7S = (2>-2)(j+2fc+iO - I.
S = j + 2k + 4, R = 5j +fe+ 2, A = 2j + 5fc + 1.
S = 5, H = 69 R = 7, A = 3, {D9 P} = {2, 4 h
sin (x+y) + sin (y+z) + sin (z+w) + sin (w+x) z sin 2x + sin 2y + sin 2z + sin 2w,
2fc3TT < Z+W < (2fc3+l)7T9 2fei5.Tr < W+X < ( 2 ^ + 1 )TT,
where ki3k2^k39k^ are integers. In particular* it always holds if x^y^z^w > 0 and
x+y+z+w = 180°, with equality when x = y = z = w.
Let the angles x», i = l s 2 98 ..9n (with x A = xx) be such that
^ ^ ^+l ^
sin (xi+x2) + sin (^2+^3) + ... + sin (x +xi) > sin 2a: 1 + sin 2x2 + ... + sin 2x . (
n
2 \ sin2j?a = )> (sin 2 ^ + sin 2x2) -2 1 sin (xi+x2) cos (xi~x2)
i=l cyclic cyclic
and (l) follows, with equality when each x.-x. is an integral multiple of 2TT. D
If all x. > 0 and xi+x2+.*.+x = 180°9 there is the following geometric inter-
u TL
This inequality appears in the authorss paper [l], and no doubt has also appeared
much earlier. It also follows immediately that
REFERENCE
1 e 1
a n — -= £ - cote.
I — ttan
fc=l 2 2
S e t t i n g 0 = TT/2 9 we then o b t a i n
Tim S = -.
REFERENCE
1. E.W. Hobson, A Treatise on Plane and Advanced Trigonometry* Dover, New Yorks
1957s pa 129.
A A A
«r% *t ?t>
REFERENCE
CRUX MATHEMATICORUM
A Call for Editors
The mathematical community has been saddened to learn of the recent death of
Fred Maskell, who served as Managing Editor of Crux Mathematicorum for many years.
The Canadian Mathematical Society extends sympathy to Mr. Maskell's family and many
friends while acknowledging with sincere thanks the tremendous contribution both he
and Editor Leo Sauve have made to mathematics through their service with Crux.
Crux Mathematicorum seems now to have reached a crossroads in its existence
and the C„M»S9 has been asked to assume responsibility for Its publication. This Is
a unique problem-solving journal with an international reputation devoted to mathe-
matics at all levels and it would well complement the existing research publications
of the C«M.S9 The level of financing required is considerable, however. In addi-
tion, the Society is reluctant to begin a new volume year without the assurance that
it will have editors for years to come. Editors of the Canadian Journal of Mathe-
matics and the Canadian Mathematical Bulletin now serve five-year terms.
Editing a publication such as Crux Mathematicorum requires special and diverse
talents. An editor must have the inclination and ability to compile articles, pro-
blems and letters into a monthly issue of interest to a wide audience of mathematics
educators. Besides the amalgamation of submitted solutions to the same problem,
editorial comments and historical information are often necessary. In the future
operation of the journal, It is envisaged that word processing facilities at the
editorss institution would be virtually essential, thereby enabling the monthly dis-
patching of camera-ready copy to a managing editor!s office in Ottawa.
Suggestions concerning the future of Crux Mathematicorum and/or applications
for the editorship (commencing January 1986) should be sent as soon as possible to:
Professor Marcel Deruaz, Chairman,
CMS Ad Hoc Committee on Crux Mathematicorum,
Department of Mathematics,
University of Ottawa,
Ottawa, Ontario,
KIN 9B4
fc & &