Index

1 - NUMBER SYSTEM ......................................................................................................................................................... 1

2 - POLYNOMIAL.............................................................................................................................................................. 16
3 - QUADRATRIC EQUATION ........................................................................................................................................... 24
4 – LINEAR EQUATION .................................................................................................................................................... 26
5 - LOGARITHM ............................................................................................................................................................... 28
6 - MODULUS FUNCTION ................................................................................................................................................ 31
11 - CIRCLES..................................................................................................................................................................... 31
12 – SET THEORY ............................................................................................................................................................. 44
13 – ARITHMATIC PROGRESSION .................................................................................................................................... 46
14 - TRIANGLE ................................................................................................................................................................. 47
15 - INEQUALITY .............................................................................................................................................................. 61
16 - TRIGNOMETRY ......................................................................................................................................................... 62
17 – ALZEBRIC EQUATION ............................................................................................................................................... 66
18 – SURFACE AREA & VOLUME ..................................................................................................................................... 71
19 – PERIMETER & AREA ................................................................................................................................................. 72
20 – SEQUENCE & SERIES ................................................................................................................................................ 76
22 – CORDINATE GEOMETRY .......................................................................................................................................... 79
23 – PERMUTATION & COMBINATION ........................................................................................................................... 81
24 – STRAIGHT LINE ........................................................................................................................................................ 81
25 - POLYGON ................................................................................................................................................................. 82
26 - STATISTICS................................................................................................................................................................ 82
27 – BINOMIAL THEOREM .............................................................................................................................................. 84
29 – LINES & ANGLES ...................................................................................................................................................... 85
1 - NUMBER SYSTEM

Problem 7
 200 
What is the largest 2-digit prime factor of the integer  ? [1983]
 100 
Solution
 200  200!
Expanding the binomial coefficient, we get  = .
 100  100!100!
Let the prime be p; then 10 ≤ p < 100. If p > 50, then the factor of p appears twice in the denominator. Thus,
we need p to appear as a factor three times in the numerator, or 3p < 200. The largest such prime is 061 ,
which is our answer.

Problem 9
The numbers 1447, 1005, and 1231 have something in common. Each is a four-digit number beginning with
1 that has exactly two identical digits. How many such numbers are there ? [1983]

Solution
Suppose the two identical digits are both one. Since the thousands digits must be one, the other one can be in
only one of three digits,
11xy, 1x1y , 1xyl.
Because the number must have exactly two identical digits, x ≠ y, x ≠ 1, and y ≠ .1
Hence, there are 3.9.8 = 216 numbers of this form.
Suppose the two identical digits are not one. Therefore, consider the following possibilities,
1xxy, 1xyx, 1yxx.
Again, x ≠ y, x ≠ 1, and y ≠ 1. There are 3.9.8 = 216 numbers of this form as well.
Thus, the desired answer is 216 + 216 = 432 .

Problem 2
n
The integer n is the smallest positive multiple of 15 such that every digit of n is either 8 or 0. Compute .
15
[1984]
Solution
Any multiple of 15 is a multiple of 5 and a multiple of 3. Any multiple of 5 ends in 0 or 5; since n only
contains the digits 0 and 8, the units digit of n must be 0.

The sum of the digits of any multiple of 3 must be divisible by 3. If n has a digits equal to 8, the sum of the
digits of n is 8a. For this number to be divisible by 3, a must be divisible by 3. We also know that a > 0 since
n is positive. Thus n must have at least three copies of the digit 8. The smallest number which meets these
two requirements is 8880.
8880
Thus the answer is = 592 .
15

Problem 8
The sum of the following seven numbers is exactly 19: a1 = 2.56, a2 = 2.51, a3 = 2. 65, a1 = 2.56, a4 = 2.71,
a5 = 2.79, a6 = 2.81, a7 = 2.86, It is desired to replace each ai by an integer approximation Ai 1 ≤ i ≤ 7 , so
that the sum of the Ai's is also 19 and so that M, the maximum of the "errors" Ai − a i , the maximum
absolute value of the difference, is as small as possible. For this minimum M, what is 100M ? [1985]

Solution
If any of the approximations Ai is less than 2 or more than 3, the error associated with that term will be
larger than 1, so the largest error will be larger than 1. However, if all of the Ai are 2 or 3, the largest error

1
will be less than 1. So in the best case, we write 19 as a sum of 7 numbers, each of which is 2 or 3. Then
there must be five 3s and two 2s. It is clear that in the best approximation, the two 2s will be used to
approximate the two smallest of the ai, so our approximations are A1 = A2 = 2 and A3 = A4 = A5 = A6 = A7
= 3 and the largest error is |A2 – A2| =0.61, so the answer is 061.
Problem 10
How many of the first 1000 positive integers can be expressed in the form [2x] + [4x] + [6x] + [8x],
Where x is a real number, and [z] denotes the greatest integer less than or equal to z ? [1985]

Solution
We will be able to reach the same number of integers while x ranges from 0 to 1 as we will when x ranges
from n to n + 1 for any integer n
(Quick proof: [2(n + x)] + … = [2n + 2x] + … = 2n + [2x] …) .
Since [2.50] + [4.50] + [6.50] + [8.50] = 100 + 200 + 300 + 400, the answer must be exactly 50 times the
number of integers we will be able to reach as x ranges from 0 to 1, including 1 but excluding 0.
Solution 1
1
Noting that all of the numbers are even, we can reduce this to any real number x between 0 to , as this will
2
n n +1
be equivalent to to for any integer n (same reasoning as above).
2 2
So now we only need to test every 10 numbers; and our answer will be 100 times the number of integers we
can reach between 1 and 10.
We can now approach this by directly searching for the integers (this solution) or brute forcing all of the
cases (next solution):
1
We can match up the greatest integer functions with one of the partitions of the integer. If we let x = then
2
1
we get the solution ; now consider when x  : [2x] = 0, [4x] ≤ 1, [6x] ≤ 2, [8x] ≤ 3.
2
But according to this the maximum we can get is 1 + 2 + 3 = 6, so we only need to try the first 6 numbers.
1
• 1 : Easily possible, for example try plugging in x = .
8
1
• 2 : Also simple, for example using .
6
1
• 3 : The partition must either be 1 + 1 +1 or 1 + 2. If [4x] = 1, then x  , but then [8x] ≥ 2; not
4
possible; and vice versa to show that the latter partition doesn't work. So we cannot obtain 3.
1
• 4 : We can partition as 1 + 1 + 2, and from the previous case we see that works.
4
1
• 5 : We can partition as 1 +2 + 2, from which we find that works.
3
3
• 6 :We can partition as 1 + 2 + 3, from which we find that works.
8
Out of these 6 cases, only 3 fails. So between 1 and 10 we can reach only the integers 1, 2, 4, 5, 6, 10; hence
our solution is 6.100 = 600.

Solution 2
As we change the value of x, the value of our expression changes only when x crosses rational number of
m
the form , where n is divisible by 2, 4, 6 or 8.
n
m m
Thus, we need only see what happens at the numbers of the form = .
lcm (2, 4, 6,8) 24
This gives us 24 calculations to make; we summarize the results here:

2
 1 2
, →0
24 24
3
→1
24
4 5
, →2
24 24
6 7
, →4
24 24
8
→5
24
9 10 11
, , →6
24 24 24
12 13 14
, , → 10
24 24 24
15
→ 11
24
16 17
, → 12
24 24
18 19
, → 14
24 24
20
→ 15
24
21 22 23
, , →6
24 24 24
24
→ 20
24
Thus, we hit 12 of the first 20 integers and so we hit 50.12 = 600 of the first 1000.

Solution 3
n −1
 k
Recall from Hermite's Identity that   x + n  = [nx] . Then we can rewrite
k =0

 1  1  1  1  3  1  5
[2x] + [4x] + [6x] + [8x] = 4[x] +  x +  +  x +  + 2  x +  +  x +  +  x +  + 4  x +  +  x + 8 
 8  6  4  3  8  2
 2  3  5  7
+ x +  + 2 x +  + x +  + x +  .
 3  4  6  8

There are 12 terms here (we don't actually have to write all of it out; we can just see where there will be
duplicates and subtract accordingly from ). Starting from every integer x, we can keep adding to achieve
one higher value for each of these terms, but after raising the last term, we will have raised the whole sum
by 20 while only achieving 12 of those 20 values. We can conveniently shift the 1000 (since it can be
12
achieved) to the position of the 0 so that there are only complete cycles of 20, and the answer is .1000 =
20
600 .
Solution 4
Imagine that we increase x from 0 to 1. At the beginning, the value of our expression is 0, at the end it is 2 +
4 + 6 + 8 = 20. How many integers between 1 and 20 did we skip ? We skip some integers precisely at those
points where at least two of 2x, 4x, 6x and 8x become integers at the same time.

3
 1
Obviously, for x = and x = 1 all four values become integers at the same time, hence we skip three
2
1 3
integers at each of these locations. Additionally, for x = and x = the values 4x and 8x become integers
4 4
at the same time, hence we skip one integer at each of the locations.
Therefore for x  (0,1] we skip a total of 3 + 3 +1 + 1 = 8 integers. As in Solution 2, we conclude that we hit
12 of the integers from 1 to 20, and so we hit 50.12 = 600 of the first 1000.

Problem 14
In a tournament each player played exactly one game against each of the other players. In each game the
winner was awarded 1 point, the loser got 0 points, and each of the two players earned 1/2 point if the game
was a tie. After the completion of the tournament, it was found that exactly half of the points earned by each
player were earned against the ten players with the least number of points. (In particular, each of the ten
lowest scoring players earned half of her/his points against the other nine of the ten). What was the total
number of players in the tournament ? [1985]
Solution
Let us suppose for convenience that there were n + 10 players over all. Among the n players not in the
n n
weakest 10 there were   games played and thus   points earned. By the givens, this means that these n
2 2
n
players also earned   points against our weakest 10. Now, the 10 weakest players playing amongst
2
 10 
themselves played   = 45 games and so earned 45 points playing each other.
2
Then they also earned 45 points playing against the stronger n players. Since every point earned falls into
n
one of these categories, It follows that the total number of points earned was 2   + 90 = n 2 − n + 90 .
 2
 n + 10  ( n + 10 ) (n + 9)
However, there was one point earned per game, and there were a total of  = games
 2  2

played and thus

( n + 10 ) (n + 9) points earned. So we have n2 – n + 90 = ( n + 10 ) (n + 9) so 2n2 – 2n + 180 =
2 2
n2 + 19n + 90 and n2 – 21n + 90 = 0 and n = 6 or n = 15.
Now, note that the top n players got n(n – 1) points in total (by our previous calculation) for an average of n
– 1, while the bottom 10 got 90 points total, for an average of 9. Thus we must have n > 10, so n = 15 and
the answer is 15 + 10 = 025 .

Problem 2
Evaluate the product
( 5 + 6 + 7) (− 5 + 6 + 7) ( 5 − 6 + 7) ( 5 + 6 − 7) . [1986]

Solution
Simplify by repeated application of the difference of squares.
2 2 2
(( 6 + 7)2 − 5 )( 5 − ( 6 − 7 )
= (13 + 2 42 – 5) (5 – (13 – 2 42 ))
= ( 2 42 + 8)( 2 42 –8)
= ( 2 42 )2 – 82 = 104

Problem 5
What is that largest positive integer n for which n3 + 100 is divisible by n + 10 ? [1986]

4
Solution
If n + 10 | n3 + 100, gcd (n3 + 100, n + 10) = n + 10.
Using the Euclidena algorithm, we have gcd (n3 + 100, n + 10) = gcd(– 10n2 + 100, n + 10) = gcd(100n +
100, n + 10) = gced(–900, n + 10), so n + 10 must divide 900.
The greatest integer n for which n + 10 divides 900 is 890; we can double-check manually and we find that
indeed 900 | 8903 + 100.
In a similar manner, we can apply synthetic division.
n 3 + 100 900
We are looking for = n2 – 10n – 100 – .
n + 10 n + 10
Again, n + 10 must be a factor of 900  n = 890 .

Problem 6
The pages of a book are numbered 1 through n. When the page numbers of the book were added, one of the
page numbers was mistakenly added twice, resulting in an incorrect sum of 1986. What was the number of
the page that was added twice ? [1986]
Solution
n(n + 1)
Denote the page number as x, with x < n. The sum formula shows that + x = 1986 .
2
n(n + 1)
Since x cannot be very large, disregard it for now and solve + x = 1986 .
2
The positive root for n  3972  63.
Quickly testing, we find that 63 is too large, but if we plug in 62 we find that our answer is
62(63)
+ x = 1986  x = 033 .
2

Problem 8
Let S be the sum of the base 10 logarithms of all the proper divisors (all divisors of a number excluding
itself) of 1000000. What is the integer nearest to S ? [1986]
Solution
Solution 1
The prime factorization of 1000000 = 2656, so there are (6 + 1)(6 + 1) = 49 divisors, of which 48 are proper.
The sum of multiple logarithms of the same base is equal to the logarithm of the products of the numbers.
Writing out the first few terms, we see that the answer is equal to log 1 + log 2 + log 4 + …… + log
1000000 = log (2050)(2150)(2250) ……… (2656).
Each power of 2 appears 7 times; and the same goes for 5.
So the overall power of 2 and 5 is 7(1 +2 +3 +4 +5 +6) = 7.21 = 147.
However, since the question asks for proper divisors, we exclude 2656, so each power is actually 141 times.
The answer is thus S = log 21415141 = log 10141 = 141 .
Solution 2
Since the prime factorization of 106 is 26.56, the number of factors in 106 is 7.7 = 49. You can pair them up
into groups of two so each group multiplies to 106.
Note that log n + log (106/n) = log n + log 106 – log n = 6.
Thus, the sum of the logs of the divisors is half the number of divisors of 106.6 – 6
(since they are asking only for proper divisors), and the answer is (49/2).6 – 6 = 141.

Problem 10
In a parlor game, the magician asks one of the participants to think of a three digit number (abc) where a, b,
and c represent digits in base 10 in the order indicated. The magician then asks this person to form the
numbers (acb), (bca), (bac), (cab) and (cba), to add these five numbers, and to reveal their sum, N.
If told the value of N, the magician can identify the original number, (abc). Play the role of the magician
and determine the (abc) if N =3194. [1986]

5
Solution
Let m be the number 100a + 10b + c.
Observe that 3194 + m = 222(a + b + c) so
m  – 3194  – 86  136 (mod 222)
This reduces m to one of 136, 358, 580, 802.
3194 + m 3194
But also a + b+ c = > > 14 so a + b + c ≥ 15.
222 222
Only one of the values of m satisfies this, namely 358 .

Problem 1
An ordered pair (m, n) of non-negative integers is called "simple" if the addition m + n in base 10 requires
no carrying. Find the number of simple ordered pairs of non-negative integers that sum to 1492. [1987]
Solution
Since no carrying over is allowed, the range of possible values of any digit of m is from 0 to the respective
digit in 1492 (the values of n are then fixed). Thus, the number of ordered pairs will be (1 + 1)(4 + 1)(9 +
1)(2 + 1) = 2.5.10.3 = 300 .

Problem-3
By a proper divisor of a natural number we mean a positive integral divisor other than 1 and the number
itself. A natural number greater than 1 will be called nice if it is equal to the product of its distinct proper
divisors. What is the sum of the first ten nice numbers ? [1987]

Solution
Let p(n) denote the product of the distinct proper divisors of n. A number is nice in one of two instances:
1. It has exactly two distinct prime divisors.
If we let n = pq, where p, are the prime factors, then its proper divisors are p and q, and p(n) = p.q =
n.
2. It is the cube of a prime number.
If we let n = p3 with p prime, then its proper divisors are p and p2, and P(n) = p.p2 = n.
We now show that the above are the only two cases. Suppose that another nice number existed that does not
fall into one of these two categories. Then we can either express it in the form n = pqr (with p, q prime and r
> 1) or n = pe (with e ≠ 3).
In the former case, it suffices to note that p(n) ≥ (pr).(qr) = pqr2 > pqr = n.
In the latter case, then p(n) = 1.p.p2 ………pe = pe(e+1)2.
For p(n) = n, we need pe(e+1)2 = pe  e2 + e = 2e  e = 0, 3 (the case e = 0  n = 1 does not work).

Thus, listing out the first ten numbers to fit this form,
2.3 = 6, 23 = 8, 2.5 = 10, 2.7 = 14, 3.5 = 15, 3.7 = 21,
2.11 = 22, 2.13 = 26, 33 = 27, 3.11 =33. Summing these yields 182 .
Alternatively, we could note that n is only nice when it only has two divisors, which, when multiplied,
clearly yield n. We know that when the prime factorization of n = a1b1.a2b2.a3b3 …. .ambm, the number of
factors f(n) of n is
f(n) = (b1 + 1) (b2+ 1)(b3 + 1) ….. (bm + 1).
Since n is nice, it may only have 4 factors (1, n, p and q). This means that f(n) = 4. The number 4 can only
be factored into (2)(2) or (4)(1), which means that either b1 = 1 and b2 = 1, or b1 = 3. Therefore the only two
cases are n = pq, or n = p3.

Problem-7
Let [r, s] denote the least common multiple of positive integers r and s. Find the number of ordered triples
(a, b, c) of positive integers for which [a, b] = 1000, [b, c] = 2000, and [c, a] = 2000. [1987]

6
Solution
Solution 1
It's clear that we must have a = 2j5k, b = 2m5n and c = 2p5p for some nonnegative integers j, k, m, n, p, q.
Dealing first with the powers of 2: from the given conditions, max(j, m) = 3, max(m, p) = max(p, j) = 4.
Thus we must have p = 4 and at least one of m, j equal to 3. This gives 7 possible triples (j, m, p) : (0, 3, 4),
(1, 3, 4), (2, 3, 4), (3, 3, 4), (3, 2, 4), (3, 1, 4) and (3, 0, 4).
Now, for the powers of 5: we have max(k, n) = max(n, q) = max(q, k) = 3.
Thus, at least two of k, n, q must be equal to 3, and the other can take any value between 0 and 3. This gives
us a total of 10 possible triples: (3, 3, 3) and three possibilities of each of the forms (3, 3, n), (3, n, 3) and (n,
3, 3).
Since the exponents of 2 and 5 must satisfy these conditions independently, we have a total of 7.10 = 0.70
possible valid triples.

Solution 2
1000 = 2353 and 2000 = 2453. By looking at the prime factorization of 2000, c must have a factor of 24. If c
has a factor of 53, then there are two cases : either (1) a or b = 5323, or (2) one of a and b has a factor of 53
and the other a factor of 23. For case 1, the other number will be in the form of 2x5y, so there are 4.4 = 16
possible such numbers; since this can be either a or b there are a total of 2(16) – 1 = 31 possibilities. For case
2, a and b are in the form of 235x and 2y53, with x < 3 and y < 3 (if they were equal to 3, it would overlap
with case 1). Thus, there are 2(3.3) = 18 cases.
If c does not have a factor of 53, then at least one of a and b must be 2353, and both must have a factor of 53.
Then, there are 4 solutions possible just considering a = 2353, and a total of 4.2 – 1 = 7 possibilities.
Multiplying by three, as 0 ≤ c ≤ 2, there are 7.3 = 21. Together, that makes 31 + 18 + 21 = 070 solutions for
(a, b, c).

Problem-11
Find the largest possible value of k for which 311 is expressible as the sum of k consecutive positive integers.
[1987]
Solution
Let us write down one such sum, with m terms and first term n + 1.
1
311 = (n + 1) + (n + 2) + …… + (n + m) = m(2n + m + 1)
2
Thus m(2n + m + 1) = 2.311 so m is a divisor of 2.311. However, because n ≥ 0 we have m2 < m(m +1) ≤
2.311 so m < 2.311 < 36.
Thus, we are looking for large factors of 2.311 which are less than 36. The largest such factor is clearly 2.35 =
486; for this value of m we do indeed have the valid expression
311 = 122 + 123 + ….. + 607, for which k = 486 .

Problem-12
Let m be the smallest integer whose cube root is of the form n + r, where n is a positive integer and r is a
positive real number less than 1/000. Find n. [1987]

Solution
In order to keep m as small as possible, we need to make n as small as possible.
m = (n + r)3 = n3 + 3n2r + 3nr2 + r3.
1 1
Since r < and m – n3 = r (3n2 + 3nr + r2) is an integer, we must have that 3n2 + 3nr + r2 ≥ ≥ 1000 .
1000 r
This means that the smallest possible n should be quite a bit smaller than 1000. In particular, 3nr + r2 should
be less than 1, so 3n2 > 999 and n > 333 . 182 = 324 < 333 < 361 = 192, so we must have n ≥ 19. Since we
want to minimize n, we take n = 19. Then for any positive value of r, 3n2 + 3nr + r2 > 3.192 > 1000, so it is
1
possible for r to be less than . However, we still have to make sure a sufficiently small r exists.
1000

7
In light of the equation m – n3 = r(3n2 + 3nr + r2), we need to choose m – n3 as small as possible to insure a
small enough r. The smallest possible value for m – n3 is 1, when m = 193 + 1. Then for this value of m,
1 1
r= 2 < , and we're set. The answer is 019.
3n + 3nr + r 2
1000

Problem-9
Find the smallest positive integer whose cube ends in 888. [1988]
Solution
A little bit of checking tells us that the units digit must be 2. Now our cube must be in the form of (10k +2)3
; using the binomial theorem gives us 1000k3 + 600k2 + 120k + 8.
Since we are looking for the tens digit, mod 100 we get 20k + 8 ≡ 88 (mod 100).
This is true if the tens digit is either 4 or 9.
Casework:
• 4 : Then our cube must be in the form of
(100k + 42)3 ≡ 3(100k)(42)2 + 423 ≡ 200k + 88 (mod 1000).
Hence the lowest possible value for the hundreds digit is 4, and so 442 is a valid solution.
• 9: Then our cube is
(100k + 92)3 ≡ 3(100k)(92)2 + 923 ≡ 200k + 688 (mod 1000). The lowest possible value for the
hundreds digit is 1, and we get 192, which is our minimum.
The answer is 192.

Problem 15
In an office at various times during the day, the boss gives the secretary a letter to type, each time putting the
letter on top of the pile in the secretary's in-box. When there is time, the secretary takes the top letter off the
pile and types it. There are nine letters to be typed during the day, and the boss delivers them in the order 1,
2 , 3, 4, 5, 6, 7, 8, 9.
While leaving for lunch, the secretary tells a colleague that letter 8 has already been typed, but says nothing
else about the morning's typing. The colleague wonders which of the nine letters remain to be typed after
lunch and in what order they will be typed. Based upon the above information, how many such after-lunch
typing orders are possible? (That there are no letters left to be typed is one of the possibilities.) [1988]

Solution
Re-stating the problem for clarity, let S be a set arranged in increasing order. At any time an element can be
appended to the end of S, or the last element of S can be removed. The question asks for the number of
different orders in which the all of the remaining elements of S can be removed, given that 8 had been
removed already.
Since 8 had already been added to the pile, the numbers 1...7 had already been added at some time to the
pile; 9 might or might not have been added yet. So currently S is a subset of {1, 2, ...7} possibly with 9 at
the end. Given that S has k elements, there are k + 1 intervals for 9 to be inserted, or 9 might have already
been placed, giving k + 2 different possibilities.
7
7
Thus, the answer is    (k + 2) .
k =0  k 

= 1.2 + 7.3 + 21.4 + 35.5 + 35.6 + 21.7 + 7.8 + 1.9 = 704 .

Problem-1
Compute (31)(30)(29)(28) + 1 . [1989]

8
Solution
Solution 1
Let's call our four consecutive integers (n – 1), n, (n +1), (n + 2). Notice that (n – 1)(n)(n + 1)(n + 2) + 1 =
(n2 + n)2 – 2(n2 + n) + 1  (n2 + n – 1)2. Thus,
19 49
1+ = = (292 + 29 – 1 ) = 869 .
30 30
Solution 2
Note that the four numbers to multiply are symmetric with the center at 29.5. Multiply the symmetric pairs
to get 31.28 = 868 and 30.29 = 870.
868.870 + 1 = (869 − 1) (869 + 1) + 1 = 8692 − 12 + 1 = 8692 = 869 .
Solution 3
The last digit under the radical is 1, o the square root must either end in 1 or 9, since x2 = 1 (mod 10) means
x = ± 1. Additionally, the number must be near 29.30 = 80, narrowing the reasonable choices to 869 and
871.
Continuing the logic, the next-to-last digit under the radical is the same as the last digit of 28.29.3.31, which
is 6. Quick computation shows that 8692 ends in 61, while 8712 ends in 41. Thus, the answer is 869 .
Solution 4
Similar to Solution 1 above, call the consecutive integers
 3  1  1  3
 n −  ,  n −  ,  n +  ,  n +  to make use of symmetry. Note that n itself is not an integer - in this case,
 2  2  2  2
 3  3  1  1
n = 29.5. The expression becomes  n −  n +  n −  n +  + 1 .
 2  2  2  2
Distributing each pair of difference of squares first, and then distributing the two resulting quadratics and
5 25
adding the constant, gives n 4 − n 2 + . The inside is a perfect square trinomial, since b2 = 4ac. It's
2 16
2
 5 5
equal to  n 2 −  , which simplifies to n2 − . You can plug in the value of n from there, or further
 4 4
 1  1
simplify to  n −  n +  − 1 , which is easier to compute.
 2  2
Either way, plugging in n = 29.5 gives 869 .

Problem-3
n
Suppose n is a positive integer and d is a single digit in base 10. Find n if = 0.d25d25d25 … . [1989]
810
Solution
Repeating decimals represent rational numbers. To figure out which rational number, we sum an infinite

d 25 100d + 25
geometric series, 0.d25d25d25…. =  n
= .
i =1 1000 999
n 100d + 25 100d + 25 4d + 1
Thus = so n = 30 =750 .
810 999 37 37
Since 750 and 37 are relatively prime, 4d + 1 must be divisible by 37, and the only digit for which this is
possible is d = 9. Thus 4d + 1 = 37 and n = 750.

Problem-4
If a < b < c < d < e are consecutive positive integers such that b + c + d is a perfect square and a + b + c + d
+ e is a perfect cube, what is the smallest possible value of c ? [1989]
Solution
Since the middle term of an arithmetic progression with an odd number of terms is the average of the series,
we know b + c + d = 3c and a + b + c + d + e = 5c. Thus, c must be in the form of 3.x2 based upon the first

9
part and in the form of 52.y3 based upon the second part, with x and y denoting an integers. c is minimized if
it’s prime factorization contains only 3, 5, and since there is a cubed term in 52.y3, 33 must be a factor of c.
3352 = 675, which works as the solution.

Problem-5
Let n be the smallest positive integer that is a multiple of 75 and has exactly 75 positive integral divisors,
n
including 1 and itself. Find . [1990]
75
Solution
The prime factorization of 75 = 3152 = (2 + 1)(4 + 1)(4 + 1). For n to have exactly 75 integral divisors, we
need to have n = p1e1–1p2e2–1… such that e1e2 … = 75. Since 75|n, two of the prime factors must be 3 and 5.
To minimize n, we can introduce a third prime factor, 2. Also to minimize n, we want 5, the greatest of all
4 4 2 n 2 4 34 5 2
the factors, to be raised to the least power. Therefore, n = 2 3 5 and = = 16.27 = 432 .
75 3.52
Problem-6
A biologist wants to calculate the number of fish in a lake. On May 1 she catches a random sample of 60
fish, tags them, and releases them. On September 1 she catches a random sample of 70 fish and finds that 3
of them are tagged. To calculate the number of fish in the lake on May 1, she assumes that 25% of these fish
are no longer in the lake on September 1 (because of death and emigrations), that 40% of the fish were not in
the lake May 1 (because of births and immigrations), and that the number of untagged fish and tagged fish in
the September 1 sample are representative of the total population. What does the biologist calculate for the
number of fish in the lake on May 1 ? [1990]
Solution
Of the 70 fish caught in September, 40% were not there in May, so 42 fish were there in May. Since the
3 60
percentage of tagged fish in September is proportional to the percentage of tagged fish in May, = 
42 x
x = 840 .
(Note the 25% death rate does not affect the answer b/c both tagged and nontagged fish die.)

Problem-8
In a shooting match, eight clay targets are arranged in two hanging columns of three targets each and one
column of two targets. A marksman is to break all the targets according to the following rules:
(1) The marksman first chooses a column from which a target is to be broken.
(2) The marksman must then break the lowest remaining target in the chosen column.
If the rules are followed, in how many different orders can the eight targets be broken? [1990]
Solution
Suppose that the columns are labeled A, B and C. The question is asking for the number of ways to shoot at
the bottom target of the columns, or the number of ways to arrange 3 As, 3 Bs, and 2 Cs in a string of 8
8
letters. of the 8 letters, 3 of them are As, making   possibilities. Of the remaining 5, 3 are Bs, making
 3
 5 8  5
  possibilities. The positions of the 2 Cs are then fixed. Thus, there are   .   =560 ways of shooting
 3  3  3
all of the targets.
8
Alternatively, the number of ways to arrange 3 As, 3 Bs, and 2 Cs in a string of 8 letters is equal to
3!.3!.2!
= 560 .

Problem-11
Someone observed that 6! = 8.9.10. Find the largest positive integer n for which n! can be expressed as the
product of n – 3 consequtive positive integers. [1990]

10
Solution
Solution 1
(n − 3 + a)!
The product of n – 3 consecutive integers can be written as for some integer a. Thus, n! =
a!
(n − 3 + a)!
, from which it becomes evident that a ≥ 3. Since (n – 3 + a)! > n!, we can rewrite this as
a!
n!(n + 1)(n + 2)...(n − 3 + a)
= n!  (n + 1)(n + 2) … (n – 3 + a) = a !
a!
For a = 4, we get n + 1 = 4! So n = 23. For greater values of a, we need to find the product of a – 3
consecutive integers that equals a!. n can be approximated as a − 3 a ! , which decreases as a increases.
Thus, n = 23 is the greatest possible value to satisfy the given conditions.
Solution 2
Let the largest of the (n-3) consecutive positive integers be k. Clearly k cannot be less than or equal to n,
else the product of (n-3) consecutive positive integers will be less than n!.
Key observation: Now for n to be maximum the smallest number (or starting number) of the (n-3)
consecutive positive integers must be minimum, implying that k needs to be minimum. But the least k > n is
(n+1).
So the (n-3) consecutive positive integers are (5, 6, 7…, n + 1)
So we have (n + 1)! /4! = n! => n+1 = 24 => n = 23
Kris17
Generalization:
Largest positive integer n for which n! can be expressed as the product of (n-a) consecutive positive integers
= (a + 1) ! – 1
For ex. largest n such that product of (n-6) consecutive positive integers is equal to n! is 7! – 1 = 5039
Proof: Reasoning the same way as above, let the largest of the (n-a) consecutive positive integers be k.
Clearly k cannot be less than or equal to n, else the product of (n-a) consecutive positive integers will be less
than n!.
Now, observe that for n to be maximum the smallest number (or starting number) of the (n-a) consecutive
positive integers must be minimum, implying that k needs to be minimum. But the least k > n is (n+1).
So the (n-a) consecutive positive integers are (a+2, a+3, … n+1)
So we have (n+1)! / (a+1)! = n! => n+1 = (a+1)! => n = (a + 1)! – 1
Kris17

Problem-13
Let T = {9k : k is an integer, 0 ≤ k ≤ 4000}. Given that 94000 has 3817 digits and that its first (leftmost)
digit is 9, how many elements of T have 9 as their leftmost digit ? [1990]
Solution
Whenever you multiply a number by 9, the number will have an additional digit over the previous digit, with
the exception when the new number starts with a 9, when the number of digits remain the same. Since 94000
has 3816 digits more than 91, exactly 4000 – (3817 – 1) = 184 numbers have 9 as their leftmost digits.

Problem-5
Given a rational number, write it as a fraction in lowest terms and calculate the product of the resulting
numerator and denominator. For how many rational numbers between 0 and 1 will 20! be the resulting
product ? [1991]
Solution
a
If the fraction is in the form , then a < b and gcd(a, b) = 1.
b
There are 8 prime numbers less than 20 (2, 3, 5, 7, 11, 13, 17, 19), and each can only be a factor of one of a
or b. There are 28 ways of selecting some combination of numbers for a; however, since a < b, only half of
a 28
them will be between 0   1 . Therefore, the solution is = 128.
b 2

11
Problem-10
Two three-letter strings, aaa and bbb, are transmitted electronically. Each string is sent letter by letter. Due
to faulty equipment, each of the six letters has a 1/3 chance of being received incorrectly, as an a when it
should have been a b, or as a b when it should be an a. However, whether a given letter is received correctly
or incorrectly is independent of the reception of any other letter. Let Sa be the three-letter string received
when aaa is transmitted and let Sb be the three-letter string received when bbb is transmitted. Let p be the
probability that Sa comes before Sb in alphabetical order. When p is written as a fraction in lowest terms,
what is its numerator ? [1991]
Solution
Solution 1
Let us make a chart of values in alphabetical order, where Pa, Pb, are the probabilities that each string comes
b
from aaa and bbb multiplied by 27, and Sb denotes the partial sums of Pb (in other words Sb =  Pb ) :
n =1

String Pa Pb Sb
aaa 8 1 1
aab 4 2 3
aba 4 2 5
abb 2 4 9
baa 4 2 11
bab 2 4 15
bba 2 4 19
bbb 1 8 27

The probability is p =  Pa.(27– Sb), so the answer turns out to be

8.26 + 4.24...2.8 + 1.0 532
= , and the solution is 532 .
272 729
Solution 2
Let S(a, n) be the nth letter of string S(a). Compare the first letter of the string S(a) to the first letter of the
string S(b). There is a (2/3)2 = 4/9 chance that S(a, 1) comes before S(b, 1). There is a 2(1/3)(2/3) = 4/9 that
S(a, 1) is the same as S(b, 1).
If S(a, 1) = S(b, 1), then you do the same for the second letters of the strings. But you have to multiply the
4/9 chance that S(a, 2) comes before S(b, 2) as there is a 4/9 chance we will get to this step.
Similarly, if S(a, 2) = S(b, 2), then there is a (4/9)3 chance that we will get to comparing the third letters and
that S(a) comes before S(b).
So we have p = (4/9) + (4/9)2 + (4/9)3 =4/9 + 16/81 + 64/729 = 532 /729.

Solution 3
Consider n letter strings instead. If the first letters all get transmitted correctly, then the a string will be first.
Otherwise, the only way is for both of the first letters to be the same, and then we consider the next n – 1
letter string following the first letter. This easily leads to a recursion:
2 2 2 1 4 4
Pn = . + 2. . pn−1 = + pn−1 .
3 3 3 3 9 9
4 52 532
Clearly, p0 = 0  p1 =  p2 =  p3 = .
9 81 729

Problem-1
Find the sum of all positive rational numbers that are less than 10 and that have denominator 30 when
written in lowest terms. [1992]
Solution
Solution 1
4 r+2
There are 8 fractions which fit the conditions between 0 and 1: =
5 n − r −1

12
Their sum is 4. Note that there are also 8 terms between 1 and 2 which we can obtain by adding 1 to each of
19 49
our first 8 terms. For example, 1 + = . Following this pattern, our answer is 4(10) + 8(1 + 2 + 3 + … +
30 30
9) = 400 .

Solution 2
By Euler's Totient Function, there are 8 numbers that are relatively prime to 30, less than 30. Note that they
come in pairs (m, 30 – m) which result in sums of 1; thus the sum of the smallest 8 rational numbers
1
satisfying this is .8.1 = 4. Now refer to solution 1.
2

Problem-2
A positive integer is called ascending if, in its decimal representation, there are at least two digits and each
digit is less than any digit to its right. How many ascending positive integers are there? [1992]
Solution
Note that an ascending number is exactly determined by its digits: for any set of digits (not including 0,
since the only position for 0 is at the leftmost end of the number, i.e. a leading 0), there is exactly one
ascending number with those digits.
So, there are nine digits that may be used: 1, 2, 3, 4, 5, 6, 7, 8, 9. Note that each digit may be present or may
not be present. Hence, there are 29 = 512 potential ascending numbers, one for each subset of {1, 2, 3, 4, 5,
6, 7, 8, 9}.
However, we've counted one-digit numbers and the empty set, so we must subtract them off to get our
answer, 512 – 10 = 502 .

Problem-4
In Pascal's Triangle, each entry is the sum of the two entries above it. In which row of Pascal's Triangle do
three consecutive entries occur that are in the ratio 3 : 4 : 5 ? [1992]
Solution
n n n
In Pascal's Triangle, we know that the binomial coefficients of the nth row are   ,   ,....   . Let our row
0 1 n
n  n   n 
be the nth row such that the three consecutive entries are   ,   and  .
r  r + 1  r + 2
3 r +1
After expanding and dividing one entry by another (to clean up the factorials), we see that = and
4 n−r
4 r+2
= . Solving, n = 062 .
5 n − r −1

Problem-5
Let S be the set of all rational numbers r, 0 < r < 1, that have a repeating decimal expansion in the form
0.abcabcabc…. = 0.abc, where the digits a, b, and c are not necessarily distinct. To write the elements of S
as fractions in lowest terms, how many different numerators are required ? [1992]
Solution
We consider the method in which repeating decimals are normally converted to fractions with an example:
X = 0.176
 1000x = 176. 0.176
176
 999x = 1000x – x = 176  x =
999
Thus, let x = 0. abc
 1000x = abc. abc
 999x = 1000x – x = abc

13
 abc
x=
999
If abc is not divisible by 3 or 37, then this is in lowest terms. Let us consider the other multiples: 333
multiples of 3, 27 of 37, and 9 of 3 and 37, so 999 – 333 – 27 + 9 = 648, which is the amount that are
neither.
The 12 numbers that are multiples of 81 reduce to multiples of 3. We have to count these since it will reduce
to a multiple of 3 which we have removed from 999, but, this cannot be removed since the numerator cannot
cancel the 3. There aren't any numbers which are multiples of 372, so we can't get numerators which are
multiples of 37. Therefore 648 + 12 = 660 .

Problem-6
For how many pairs of consecutive integers in {1000, 1001, 1002, … 2000} is no carrying required when
the two integers are added ? [1992]
Solution
Solution 1
Consider what carrying means: If carrying is needed to add two numbers with digits abcd and efgh, then h
+ d ≥ 10 or c + g ≥ 10.6. Consider c  {0, 1, 2, 3, 4}. 1abc + 1ab(c + 1) has no carry if a, b  {0, 1, 2, 3, 4}.
This gives 53 = 125 possible solutions.
With  {5, 6, 7, 8}, there obviously must be a carry. Consider c = 9. a, b  {0, 1, 2, 3, 4} have no carry. This
gives 52 = 25 possible solutions. Considering b = 9,  {0, 1, 2, 3, 9} have no carry. Thus, the solution is 125
+ 25 + 6 = 156 .
Solution 2
Consider the ordered pair (1abc, 1abc – 1) where a, b and c are digits. We are trying to find all ordered pairs
where (1abc) + (1abc –1) does not require carrying. For the addition to require no carrying, 2a, 2b < 10, so
a, b < 5 unless 1abc ends in 00, which we will address later. Clearly, if c  {0, 1, 2, 3, 4, 5}. then adding
(1abc) + (1abc – 1) will require no carrying. We have 5 possibilities for the value of a, 5 for b, and 6 for c,
giving a total of (5)(5)(6) = 150, but we are not done yet.
We now have to consider the cases where b, c = 0, specifically when 1abc  {1100, 1200, 1300, 1400, 1500,
1600, 1700, 1800, 1900, 2000}.
We can see that 1100, 1200, 1300, 1400, 1500, and 2000 all work, giving a grand total of 150 + 6 = 156
ordered pairs.

Problem-12
In a game of Chomp, two players alternately take bites from a 5-by-7 grid of unit squares. To take a bite, a
player chooses one of the remaining squares, then removes ("eats") all squares in the quadrant defined by the
left edge (extended upward) and the lower edge (extended rightward) of the chosen square. For example, the
bite determined by the shaded square in the diagram would remove the shaded square and the four squares
marked by x. (The squares with two or more dotted edges have been removed form the original board in
previous moves.)

14
The object of the game is to make one's opponent take the last bite. The diagram shows one of the many
subsets of the set of 35 unit squares that can occur during the game of Chomp. How many different subsets
are there in all? Include the full board and empty board in your count. [1992]
Solution
By drawing possible examples of the subset, one can easily see that making one subset is the same as
dividing the game board into two parts.
One can also see that it is the same as finding the shortest route from the upper left hand corner to the lower
right hand corner; Such a route would require 5 lengths that go down, and 7 that go across, with the shape on
 12 
the right "carved" out by the path a possible subset. Therefore, the total number of such paths is   = 792
5
Links to Branches of Mathematics
This question is one case of the problem of counting order ideals or antichains in posets. Specifically, it ask
for the number of order ideals of the product poset of the chain of length 4 and the chain of length 6.
Problem-15
Define a positive integer n to be a factorial tail if there is some positive integer m such that the decimal
representation of m! ends with exactly n zeroes. How many positive integers less than 1992 are not factorial
tails ? [1992]
Solution
The number of zeros at the end of m! is
m  m   m   m   m 
f(m) =   +   +  + + + .....
 5   25  125   625   3125 
Note that if m is a multiple of 5,
f(m) = f(m + 1) = f(m + 2) = f(m + 3) = f(m + 4)
m m m m
Since f(m) ≤ + + + ..... = , a value of m such that f(m) = 1991 is greater than 7964. Testing
5 25 125 4
values greater than this yields f(7975) = 1991.
7975
There are = 1595 distinct positive integers, f(m), less than 1992. Thus, there are 1991 – 155 = 396
5
positive integers less than 1992 that are not factorial tails.

Problem-1
How many even integers between 4000 and 7000 have four different digits ? [1993]
Solution
The thousands digit is  {4, 5, 6}. If the thousands digit is even (4, 6, 2 pospossibilities), then there are only
10
− 1 = 4 possibilities for the units digit. This leaves possible digits for the hundreds and 7 for the tens
2
places, yielding a total of 2.4.8.7 = 448.
If the thousands digit is odd (5, one possibility), then there are 5 choices for the units digit, with 8 digits for
the hundreds and 7 for the tens place. This gives 1.5.8.7 = 280 possibilities. Together, the solution is 448 +
280 = 728 .

Problem-6
What is the smallest positive integer than can be expressed as the sum of nine consecutive integers, the sum
of ten consecutive integers, and the sum of eleven consecutive integers ? [1993]
Solution
Solution 1
Denote the first of each of the series of consecutive integers as a, b, c. Therefore, n = a + (a + 1) … (a + 8) =
9a + 36 = 10b + 45 = 11c + 55. Simplifying, 9a = 10b + 9 = 11c + 19. The relationship between a, b
suggests that is divisible by 9. Also, 10b – 10 = 10(b – 1) = 11c, so b – 1 is divisible by . We find that
the least possible value of b = 45, so the answer is 10(45) + 45 = 495.

15
Solution 2
Let the desired integer be n. From the information given, it can be determined that, for positive integers a, b,
c:
n = 9a + 36 = 10b + 45 == 11c + 55
This can be rewritten as the following congruences:
N  0 (mod 9)
N  0 (mod 10)
N  0 (mod 11)
Since 9 and 11 are relatively prime, n is a multiple of 99. It can then easily be determined that the smallest
multiple of 99 with a units digit 5(this can be interpreted from the 2nd congruence) is 495
Solution 3
9x = a
Let n be the desired integer. From the given information, we have 11 y = a , here, x and y are the middle
10 z + 5 = a
terms of the sequence of 9 and 11 numbers, respectively. Similarly, we have z as the 4th term of the
sequence. Since, a is a multiple of 9 and 11 it is also a multiple of lcm [9, 11] = 99. Hence, a = 99m, for
some m. So, we have 10z + 5 = 99m. It follows that 99(5) = 495 . is the smallest integer that can be
represented in such a way.

NUMBER THEORY

Problem -14
What is the largest even integer that cannot be written as the sum of two odd composite numbers? [1984]
Solution
Take an even positive integer x. x is either 0 mod 6, 2 mod 6, or 4 mod 6.
Notice that the numbers 9, 15, 21, ……., and in general 9 + 6n for nonnegative n are odd composites. We
now have 3 cases:
If x ≥ 18 and is 0 mod6, x can be expressed as 9 + (9 + 6n) for some nonnegative n. Note that 9 and 9 + 6n
are both odd composites.
If x ≥ 44 and is 2 mod 6, x can be expressed as 35 + (9 + 6n) for some nonnegative n. Note that 35 and 9 +
6n are both odd composites.
If x ≥ 34 and is 4 mod 6, x can be expressed as 25 + (9 + 6n) for some nonnegative n. Note that 25 and 9 +
6n are both odd composites.
Clearly, if x ≥ 44, it can be expressed as a sum of 2 odd composites. However, if x = 42, it can also be
expressed using case 1, and if x = 40, using case 3. 38 is the largest even integer that our cases do not cover.
If we examine the possible ways of splitting 38 into two addends, we see that no pair of odd composites add
to 38. Therefore, 038 is the largest possible number that is not expressible as the sum of two odd composite
numbers.

2 - POLYNOMIAL

Problem-5
Let an equal 6n + 8n. Determine the remainder upon dividing a83 by 49. [1983]
Solution
Solution 1
First, we try to find a relationship between the numbers we're provided with and 49. We realize that 49 = 72
and both 6 and 8 are greater or less than 7 by 1.
Expressing the numbers in terms of 7, we get (7– 1)83 + (7 + 1)83.
Applying the Binomial Theorem, half of our terms cancel out and we are left with 2(783 + 3403.781 + … +
83.7). We realize that all of these terms are divisible by 49 except the final term.

16
After some quick division, our answer is 035 .

Solution 2
Since (49) = 42 (the Euler's totient function), by Euler's Totient Theorem, a42  1 (mod 49) where gcd(a,
49) = 1.
8 + 6 14
Thus 683 + 883  62(42)–1 + 82(42)–1  6–1 8–1 +    035 (mod 49)
48 −1
• Alternatively, we could have noted that ab  ab(mod (n)) (mod n). This way, we have 683  683(mod 42) 
6– 1(mod 9), and can finish the same way.

Problem-15
Determine w2 + x2 + y2 + z2 if
x2 y2 z2 w2
+ + + =1
22 − 1 22 − 32 22 − 52 22 − 72
x2 y2 z2 w2
+ + + =1
42 − 1 42 − 32 42 − 52 42 − 72
x2 y2 z2 w2
+ + + =1
62 − 1 62 − 32 62 − 52 62 − 7 2
x2 y2 z2 w2
+ + + =1
82 − 1 82 − 32 82 − 52 82 − 7 2
Solution
Solution 1
x2 y2 z2 w2
Rewrite the system of equations as + + + =1.
t − 1 t − 32 t − 52 t − 7 2
This equation is satisfied when t = 4, 16, 36, 64, as then the equation is equivalent to the given equations.
After clearing fractions, for each of the values t = 4, 16, 36, 64, we have the equation
x2(t – 9)(t – 25)(t – 49) + y2(t – 1)(t – 25)(t – 49) + z2(t – 1)(t – 9)(t – 49) + w2(t – 1)(t – 9)(t – 25) = (t – 1)(t
– 9)(t – 25)(t – 49).
When t = 4, 16, 36, 64, a (t – 4)(t – 16)(t – 36)(t – 64) term can be subtracted from the right-hand side
because it equals 0. Thus we have the following equation which holds for t = 4, 16, 36, 64.
x2(t – 9)(t – 25)(t – 49) + y2(t – 1)(t – 25)(t – 49) + z2(t – 1)(t – 9)(t – 49) + w2 (t – 1)(t – 9)(t – 25) = (t – 1)(t
– 9) (t – 25) (t – 49) – (t – 4) (t – 16) (t – 36) (t – 64)
Each side of this equation is a polynomial in t of degree at most 3, and they are equal for 4 values of t (when
t = 4, 16, 36, 64). Therefore, the polynomials must be equal for all t.*
Now we can plug in t = 1 into the polynomial equation. Most terms drop, and we end up with
x2(–8)(–24)(–48) = – (–3)( –15)( –35)( –63)
so that
3.15.35.63 32.52.7 2
x2 = =
8.24.48 210
Similarly, we can plug in t = 9, 25, 49 and get
5.7.27.55 33.5.7.11
y2 = =
8.16.40 210
2
21.9.11.39 33.7.11.13
z = =
24.16.24 210
2
45.33.13.15 32.5.11.13
w = =
48.40.24 210
Now adding them up,
32.11.13(7 + 5) 32.11.13
z2 + w2 = =
210 28
2 2
32.5.7(5.7 + 3.11) 32.5.7.17
x +y = =
210 28

17
with a sum of
32 (3.11.13 + 5.7.17)
= 32.4 = 036 .
28
/*Lengthy proof that any two cubic polynomials in t which are equal at 4 values of t are themselves
equivalent: Let the two polynomials be A(t) and B(t) and let them be equal at t = a, b, c, d. Thus we have
A(a) – B(b) = 0, A(b) – B(b) = 0, A(c) – B(c) = 0, A (d) – B(d) = 0.
Also the polynomial A(t) – B(t) is cubic, but it equals 0 at 4 values of t. Thus it must be equivalent to the
polynomial 0, since if it were nonzero it would necessarily be able to be factored into (t – a)(t – b)(t – c)(t –
d) (some nonzero polynomial) which would have a degree greater than or equal to 4, contradicting the
statement that A(t) – (t) is cubic. Because A(t) – B(t) = 0, A(t) and B(t) are equivalent and must be equal for
all t.
Post script for the puzzled: This solution which is seemingly unnecessarily redundant in that it computes
x2, y2, z2 and w2 separately before adding them to obtain the final answer is appealing because it gives the
individual values of x2, y2, z2 and w2 which can be plugged into the given equations to check.

Solution 2
As in Solution 1, we have
(t – 1) (t – 9) (t – 25) (t – 49) – x2(t – 9) (t – 25) (t – 49) – y2(t – 1) (t – 25) (t – 49) – z2(t – 1) (t – 9) (t – 49)
– w2(t – 1) (t – 9) (t – 25) = (t – 4) (t – 16) (t – 36) (t – 64)
Now the coefficient of t3 on both sides must be equal. Therefore we have
1 + 9 + 25 + 49 + x2 + y2 + z2 + w2 = 4 + 16 +36 +64  x2 + y2 + z2 + w2 = 036 .

Problem-7
Assume that a, b, c and d are positive integers such that a5 = b4, c3 = d2, and c – a = 19. Determine d – b.
[1985]
Solution
It follows from the givens that a is a perfect fourth power, b is a perfect fifth power, c is a perfect square and
d is a perfect cube. Thus, there exist integers s and t such that a = t4, b = t5, c = s2 and d = s3. So s2 = t4 = 19.
We can factor the left-hand side of this equation as a difference of two squares, (s – t2) (s + t2) = 19. 19 is a
prime number and s + t2 > s – t2 so we must have s + t2 = 19 and s - t2 = 1. Then s = 10, t = 3 and so d = s3 =
1000, b = t5 = 243 and d – b = 757.

Problem-11
The polynomial 1 – x + x2 – x3 + … + x16 – x17 may be written in the form a0 + a1y + a2y2 + … + a16 y16 +
a17y17, where y = x + 1 and the ai’s are constants. Find the value of a2. [1986]
Solution
1 − x18 1 − x18
Using the geometric series formula, 1 – x + x2 + … – x17 = = .
1+ x y
1 − ( y − 1)18
Since x = y – 1, this becomes 3
. We want a2, which is the coefficient of the y term in
y
– (y – 1)18 (because the y in the denominator reduces the degrees in the numerator by 1). By the binomial
 18 
theorem that is (–1).( –1)   = 816 .
3
Solution 2
Again, notice x = y – 1. So
1 – x + x2 + … – x17 = 1 – (y – 1) + (y – 1)2 – (y – 1)3 +… – (y – 1)17 = 1 + (1 – y) + (1 – y)2 + (1 – y)3 …
+ (1 – y)17

We want the coefficient of the y2 term of each power of each binomial, which by the binomial theorem is
 2  3  17   18 
  +   + … +   . The Hockey Stick Identity gives us that this quantity is equal to   = 816 .
 2  2 2 3

18
Problem-5
Find 3x2y2 if x and y are integers such that y2 + 3x2y2 = 30x2 + 517. [1987]
Solution
If we move the x2 term to the left side, it is factorable:
(3 x2+1)(y2 – 10) = 517 – 10
507 is equal to 3*132. Since x and y are integers, 3 x2 + 1 cannot equal a multiple of three. 169 doesn't work
either, so 3 x2 + 1 = 13, and x =  2. This leaves y2 – 10 = 39, so y =  7. Thus, 3x2y2 = 3× 4 × 49 = 588 .

Problem-8
8 n 7
What is the largest positive integer n for which there is a unique integer k such that   ?
15 n + k 13
[1987]
Solution
Solution 1
Multiplying out all of the denominators, we get:
104(n + k) < 195n < 105(n + k)
0 < 91n – 104k < n + k
6 7n
Since 91n – 104k < n + k, k > n. Also, 0 < 91n – 104k, so k < . Thus, 48n < 56k < 49n.
7 8
k is unique if it is within a maximum range of 112, so n = 112.

Solution 2
Flip all of the fractions for
15 k + n 13
> >
8 n 7
105n > 56(k + n) > 104n
49n > 56k > 48n
Continue as in Solution 1.

Problem-14
(104 + 324)(224 + 324)(344 + 324)(464 + 32)(584 + 324)
Compute . [1987]
(44 + 324)(164 + 324)(284 + 324)(404 + 32)(524 + 324)
Solution
The Sophie Germain Identity states that a4 + 4b4 can be factorized as
(a2 + 2b2 – 2ab)(a2 + 2b2 + 2ab).
Each of the terms is in the form of x4 + 324. Using Sophie-Germain, we get that
x4 + 4.34 = (x2 + 2.32 – 2.3.x)( x2 + 2.32 + 2.3.x) = (x(x – 6) + 18) (x(x + 6) + 18) .
[(10(10 − 6) + 18)(10(10 + 6) + 18][22(22 − 6) + 18(22(22 + 6) + 18]...[(58(58 − 6) + 18)(58(58 + 6) + 18)]
[(4(4 − 6) + 18)(4(4 + 6) + 18][16(16 − 6) + 18)(16(16 + 6) + 18]...[(52(52 − 6) + 18)(52(52 + 6) + 18)]
[(10(4) + 18)(10(16) + 18][22(16) + 18(22(28) + 18]...[(58(52) + 18)(58(64) + 18)]
[(4(−2) + 18)(4(10) + 18][16(10) + 18)(16(22) + 18]...[(52(46) + 18)(52(58) + 18)]
58(64) + 18 3730
Almost all of the terms cancel out! We are left with = = 373 .
4(−2) + 18 10

Problem-15
Squares S1 and S2 are inscribed in right triangle ABC, as shown in the figures below. Find AC + CB if area
(S1) = 441 and area (S2) = 440. [1984]
A A

S2
S1
B A B
C A C 19
Solution

A A
T1 r3
A A S2
S1 T4
r2 T5 A
A B
A A B C A
C
Because A
A all the triangles in the figure are similar to triangle ABC, it's a good idea to use area ratios. In the
T T 441 T T 441 440
diagram above, 1 = 2 = . Hence, 1 = 2 = and T4 = T2 .
T3 T4 440 T3 T4 440 441
Additionally, the area of triangle ABC is equal to both T1 + T2 + 441 and T3 + T4 + T5 + 440.
Setting the equations equal and solving for T5,
T T
T5 = 1 + T1 – T3 + T2 – T4 = 1 + 1 + 2 . Therefore, 441T5 = 441 + T1 + T2.
441 441
However, 441 + T1 + T2 is equal to the area of triangle ABC! This means that the ratio between the areas T5
and ABC is 441, and the ratio between the sides is 441 = 21.
As a result, AB = 21 441 = AC 2 + BC 2 .
We now need (AC)(BC) to find the value of AC + BC, because AB2 + 2(AC)(BC) = (AC + BC)2.
1
Let h denote the height to the hypotenuse of triangle ABC. Notice that h − h = 440 .
21
(The height of ABC decreased by the corresponding height of T5)
Thus, (AB)(h) = (AC)(BC) = 22.212 .
Because AB2 + 2(AB)(BC)=(AC+BC)2=212.222, AC + BC = (21)(22) = 462.

Problem-2
For any positive integer k, let f1(k) denote the square of the sum of the digits of k. For n ≥ 2, let fn(k) = f1(fn–
1(k)). Find f1988 (11). [1988]
Solution
We see that f1 (11) = 4
f2 (11) = f1 (4) = 16
f3 (11) = f1 (16) = 49
f4 (11) = f1 (49) = 169
f5 (11) = f1 (169) = 256
f6 (11) = f1 (256) = 169
Note that this revolves between the two numbers. f1988(169) = 169  f1988(11) = 169 .

Problem-13
Find a if a and b are integers such that x2 – x – 1 is a factor of ax17 + bx16 + 1. [1988]
Solution
Solution 1
Let's work backwards! Let F(x) = ax17 + bx16 + 1 and let P(x) be the polynomial such that
P(x)(x2 – x – 1) = F(x).
First, clearly the constant term of P(x) must be –1. Now, we have
(x2 – x – 1)(c1x15 + c2x14 + … + c15x–1), where c15 is some coefficient. However, since F(x) has no x term, it
must be true that c15 = 1.
Let's find c14 now. Notice that all we care about in finding c14 is that (x2 – x – 1)(… + c14x2 + x – 1) =
something + 0x2 + something. Therefore, c14 = – 2. Undergoing a similar process, c13 = 3, c12 = – 5, c11 = 8,
and we see a nice pattern. The coefficients of P(x) are just the Fibonacci sequence with alternating signs!
Therefore, a = c1 = F16, where F16 denotes the 16th Fibonnaci number and a = 987 .

20
Solution 2
Let Fn represent the nth number in the Fibonacci sequence. Therefore,
x2 – x – 1 = 0  xn = Fn(x), n  N  xn+2 = Fn+1 . x + Fn, n  N.
The above uses the similarity between the Fibonacci recursive definition,
Fn+2 – Fn+1 – Fn = 0, and the polynomial x2 – x – 1 = 0.
0 = ax17 + bx16 + 1 = a (F17. x + F16) + b(F16 . x + F15) + 1 
(aF17 + bF16) .x + (aF16 + bF15 + 1) = 0, x  Q 
aF17 + bF16 = 0 and aF16 + bF15 + 1 = 0 
a = F16, b = – F17  a = 987 .

Solution 3
We can long divide and search for a pattern; then the remainder would be set to zero to solve for a. Writing
out a few examples quickly shows us that the remainders after each subtraction follow the Fibonacci
sequence. Carrying out this pattern, we find that the remainder is (F16 +F17a)x + F15b + F16a + 1 = 0. Since
the coefficient of x must be zero, this gives us two equations, F16b + F17a = 0 and F15b + F16a + 1 = 0.
Solving these two as above, we get that a = 987 .
There are various similar solutions which yield the same pattern, such as repeated substitution of x2 = x + 1
into the larger polynomial.

Solution 4
The roots of x2 – x – 1 are (the golden ratio) and 1 – . These two must also be roots of ax17 + bx16 + 1.
Thus, we have two equations: a17 + b16 + 1 = 0 and a(1 – )17 + b(1 – )16 +1 = 0.
a ( 17 − (1 −  )17 ) b( 16 − (1 −  )16 )
Subtract these two and divide by 5 to get + = 0.
5 5
 n − (1 −  ) n )
But the formula for the nth fibonacci number is (You may want to research this). Thus, we
5
have 1597a + 987b = 0, so since 1597 and 987 are relatively prime, and the answer must be a positive
integer less than 1000, we can guess that it equals 987 .

Problem-2
Find the value of (52 + 6 43 )3/2 – (52 – 6 43 )3/2 . [1990]
Solution
Suppose that 52 + 6 43 is in the form of (a + b 43 )2. FOILing yields that 52 + 6 43 = a2 + 43b2 + 2ab
43 .
This implies that a and b equal one of  1,  3. The possible sets are (3, 1) and (–3, –1); the latter can be
discarded since the square root must be positive.
This means that 52 + 6 43 = ( 43 + 3)2. Repeating this for 52 – 6 43 , the only feasible possibility is (
43 – 3)2.
Rewriting, we get ( 43 + 3)3 – ( 43 – 3)3.
Using the difference of cubes, we get that
[ 43 + 3 – 43 + 3][(43 + 6 43 + 9) + (43 – 9) + (43 – 6 43 + 9)]
= (6)(3.43 + 9) = 828 .

Problem-1
Find x2 + y2 if x and y are positive integers such that
xy + x + y = 71
x2y + xy2 = 880. [1991]
Contents
• 1 Problem
• 2 Solution
21
 ➢ 2.1 Solution 1
➢ 2.2 Solution 2
• 3 See also
Solution
Solution 1
Define a = x + y and b = xy. Then a + b = 71 and ab = 880. Solving these two equations yields a quadratic:
a2 – 71a + 880 = 0, which factors to (a – 16) (a – 55) = 0. Either a = 16 and b = 55 or a = 55 and b = 16. For
the first case, it is easy to see that (x, y) can be (5, 11) (or vice versa). In the second case, since all factors of
16 must be ≤ 16, no two factors of 16 can sum greater than 32, and so there are no integral solutions for (x,
y). The solution is 52 + 112 = 146.
Solution 2
Since xy + x + y + 1 = 72, this can be factored to (x + 1)(y + 1) = 72. As x and y are integers, the possible
sets for (x, y) (ignoring cases where x > y since it is symmetrical) are (1, 35), (2, 23), (3, 17), (5, 11), (7, 8).
The second equation factors to (x + y)xy = 880 = 24.5.11. The only set with a factor of 11 is (5, 11), and
checking shows that it is our solution.

Problem-4
How many ordered four-tuples of integers (a, b, c, d) with 0 < a < b < c < d < 500 satisfy a + d = b + and bc
– ad = 93 ? [1993]
Solution
Solution 1
Let k = a + d = b + c so d = k – a, b = k – c.
It follows that (k – c)c – a(k – a) = (a – c)(a + c – k) = (c – a)(d – c) = 93.
Hence (c – a, d – c) = (1, 93), (3, 31), (31, 3), (93, 1).
Solve them in terms of c to get
(a, b, c, d) = (c – 93, c – 92, c, c + 1),
(c – 31, c – 28,c, c + 3),
(c – 1, c + 92, c, c + 93), (c – 3, c + 28, c, c + 31).
The last two solutions don't follow a < b < c < d, so we only need to consider the first two solutions.
The first solution gives us c – 93 ≥ 1 and c + 1 ≤ 499  94 ≤ c ≤ 498, and the second one gives us 32 ≤ c ≤
496. So the total number of such four-tuples is 405 + 465 = 870 .

Solution 2
Let b = a + m and c = a + m + n.
From a + d = b + c, d = b + c – a = a + 2m + n.
Substituting b = a + m, c = a + m + n, and d = b + c – a = a + 2m + n
Into bc – ad = 93, bc – ad = (1 + m)(1 + m + n) – a(a + 2m + n) = m(m + n) = 93 = 3(31)
Hence, (m, n) = (1, 92) or (3, 28).
For (m, n) = (1, 92), we know that 0 < a < a + 1 < a + 93 < a + 94 < 500, so there are 405 four-tuples.
For (m, n) = (3, 28), 0 < a < a + 3 < a + 31 < a + 34 < 500, and there are 465 four-tuples. In total, we have
405 + 465 = 870 . four-tuples.

Solution 3
Square both sides of the first equation in order to get b and ad terms, which we can plug 93 in for.
(a + d)2 = (b + c )2  a2 + 2ad + d2 = b2 + 2bc + c2  2bc – 2ad = a2 – b2 + d2 – c2
We can plug 93 in for bc – ad to get 186 on the left side, and also observe that a – b = c – d after rearranging
the first equation. Plug in c – d for a – b.
186 = (c – d) (a + b) + (d –c) (d + c)  186 = – (d – c) (a + b) + (d – c) (d + c)  186
Now observe the possible factors of 186, which are 1.186, 2.93, 3.62, 6.31. (d – c) and (d + c – a – b) must
be factors of 186, and (d + c – a – b) must be greater than (d – c). 1.186 work, and yields 405 possible
solutions. 2.93 does not work, because if c – d = 2, , then a + b must differ by 2 as well, but an odd number
93 can only result from two numbers of different parity. c – d will be even, and a + b will be even, so c + d –
(a + b) must be even. 3.62 works, and yields 465 possible solutions, while 6.31 fails for the same reasoning
above. Thus, the answer is 405 + 465 = 870 .

22
Problem-5
Let P0(x) = x3 + 313x2 – 77x – 0 8. For integers n ≥ 1, define Pn(x) = Pn – 1(x – n). What is the coefficient of
x in P20(x) ? [1993]
Solution
P20(x) = P19(x – 20)
= P18((x – 20) – 19)
= P17(((x – 20) – 19) – 18)
= ……..
Notice that = P0(x – (20 + 19 + 18 + …. + 2 + 1)).
Using the formula for the sum of the first numbers,
20(20 + 1)
1 + 2 + … + 20 = = 210.
2
Therefore, P20(x) = P0(x – 210).
Substituting x – 210 into the function definition, we get P0(x – 210) = (x – 210)3 + (x – 210)2 – 77(x – 210) –
8.
We only need the coefficients of the linear terms, which we can find by the binomial theorem.
 3
• (x – 210)3 will have a linear term of   2102x =630.210x.
1
 2
• 313(x – 210)2 will have a linear term of – 313.   210x = – 626.210x
1
• – 77(x –210) will have a linear term of – 77x.
Adding up the coefficients, we get 630.210 – 626.210 – 77 = 763 .

Problem-5
Given a positive integer n, let p(n) be the product of the non-zero digits of n. (If n has only one digits, then
P(n) is equal to that digit.)
Let S = p(1) + p(2) + p(3) + … + p(999)
What is the largest prime factor of S ? [1994]
Solution
Solution 1
Suppose we write each number in the form of a three-digit number (so 5  005), and since our p(n) ignores
all of the zero-digits, replace all of the 0s with 1s. Now note that in the expansion of
(1 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9)(1 + 1 + 2 + 3 + … + 9)(1 + 1 + 2 + 3 + … + 9)
we cover every permutation of every product of 3 digits, including the case where that first 1 represents the
replaced 0s. However, since our list does not include 000, we have to subtract 1. Thus, our answer is the
largest prime factor of
(1 + 1 + 2 + 3 + … + 9)3 – 1 = 463 – 1 = (46 – 1)(462 + 46 + 1) = 33.5.7. 103 .
Solution 2
Note that p(1) = p(11), p(2) = p(12), p(3) = p(13), …. P(19) = p(9), and
p(37) = 3p(7). So
p(10) + p(11) + p(12) + ….. + p(19) = 46, p(10) + p(11) + …. + p(99) = 46*45 = 2070.
We add
p(1) + p(2) + p(3) + ….. + p(10) = 45 to get 2115. When we add a digit we multiply the sum by that digit.
Thus 2115.(1 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9) = 2115.46 = 47.45.46.
But we didn't count 100, 200, 300, ... , 900. We add another 45 to get 45.2163. The largest prime factor of
that is 103 .

23
3 - QUADRATRIC EQUATION

Problem-3
What is the product of the real roots of the equation x2 + 18x + 30 = 2 x 2 + 18 x + 45 ? [1983]
Solution
If we expand by squaring, we get a quartic polynomial, which isn't very helpful.
Instead, we substitute y for x2 + 18x + 30 and our equation becomes y = 2 2 y + 15 .
Now we can square; solving for y, we get y = 10 or y = – 6. The second solution is extraneous since
2 y + 15 is positive. So, we have y = 10 as the only solution for y. Substituting x2 + 18x + 30 back in for y,
x2 + 18x + 30 = 10  x2 + 18x + 20 = 0.
By Vieta's formulas, the product of the roots is 202 .
Problem
A machine shop cutting tool is in the shape of a notched circle, as shown. The radius of the circle is 50
cm, the length of AB is 6 cm, and that of BC is 2 cm. The angle ABC is a right angle. Find the square of the
distance (in centimeters) from B to the center of the circle.

O
C
B

Solution
Solution 1
Because we are given a right angle, we look for ways to apply the Pythagorean Theorem. Let the foot of the
perpendicular from O to AB be D and let the foot of the perpendicular from O to the line BC be E. Let OE =
x and OD = y. We're trying to find x2 + y2.

O D
C
E B
R

Applying the Pythagorean Theorem, OA2 = OD2 + AD2 and OC2 = EC2 + EO2.
Thus, ( 50)2 = y 2 + (6 − x)2 , and ( 50)2 = x 2 + ( y + 2)2 . We solve this system to get x = 1 and y = 5,
resulting in 12 + 52 = 026 .
Solution 2
Drop perpendiculars from O to AB(T1), M to OT1(T2) and M to AB(T3). Also, draw the midpoint M of AC.
Then the problem is trivialized. Why?

24
 M
T2
T3
A B
T1

O
First notice that by computation, OAC is a 50 − 40 − 50 isosceles triangle; thus AC = MO. Then,
notice that  MOT2 =  T3MO =  BAC. Thus the two blue triangles are congruent.
So, MT2 = 2, OT2 = 6. As T3B = 3, MT3 = 1, we subtract and get OT1 = 5, T1B = 1. Then the Pythagorean
Theorem shows OB2 = 026 .

Problem-4
1 1 2
Find the positive solution to + 2 – 2 = 0. [1990]
x − 10 x − 29
2
x − 10 x − 45 x − 10 x − 69
Solution
We could clear out the denominators by multiplying, though that would be unnecessarily tedious.
To simplify the equation, substitute a = x2 – 10x – 29 (the denominator of the first fraction). We can rewrite
1 1 2
the equation as + − = 0 . Multiplying out the denominators now, we get: (a – 16)(a – 40) + a(a
a a − 16 a − 40
– 40) – 2(a)(a – 16) = 0. Simplifying, – 64a + 40 × 16 = 0, so a = 10. Re-substituting, 10 = x2 – 10x – 29 
0 = (x – 13)(x + 3). The positive root is 013 .

Problem-7
Find A2, where A is the sum of the absolute values of all roots of the following equation:
91
X = 19 + [1991]
91
19 +
91
19 +
19
19 +
91
19 +
x
Solution
19
Let f(x) = . 19 + . Then x = f(f(f(f)(f(x)))), from which we realize that f(x). This is because if we expand
x
ax + b
the entire expression, we will get a fraction of the form on the right hand side, which makes the
cx + d
equation simplify to a quadratic.
As this quadratic will have two roots, they must be the same roots as the quadratic f(x) = x.
The given finite expansion can then be easily seen to reduce to the quadratic equation x2 – 19 x – 91 = 0.
19 + 383
The solutions are x = .
2
Therefore, A = |x+| + |x–| = 383 . We conclude that A2 = 383.

25
Problem-8
For how many real numbers a does the quadratic equation x2 + ax + 6a = 0 have only integer roots for x ?
[1991]
Solution
Solution 1
By Vieta's formulas, x1 + x2 = – a where x1, x2 are the roots of the quadratic, and since x1 , x2 are integers, a
must be an integer. Applying the quadratic formula,
−a a 2 − 24a
X=
2
Since – a is an integer, we need a 2 − 24a to be an integer (let this be ): b2 = a2 – 24a. Completing the
square, we get
(a – 12)2 = b2 + 144
Which implies that b2 + 144 is a perfect square also (let this be c2). Then
c2 – b2 = 144  (c + b)(c – b) = 144
The pairs of factors of 144 are
(1, 144), (2, 72), (3, 48), (4, 36), (6, 24), (8, 18), (9, 16), (12, 12), ; since c is the
average of each respective pair and is also an integer, the pairs that work must have the same parity. Thus
we get 10 pairs (counting positive and negative) of factors that work, and substituting them backwards
show that they all work.
Solution 2
Let x2 + ax + 6a = (x – s)(x – r).

sr + 6s + 6r = 0
Vieta's yields s + r = – a, sr = 6a. sr + 6 s + 6r + 36 = 36
( s + 6)(r + 6) = 36
Without loss of generality let r ≤ s.
The possible values of (r + 6, s + 6) are:
(– 36, –1), (– 18, –2), (– 12, –3), (– 9, –4), (– 6, –6), (1, 36), (2, 18), (3, 12), (4, 9), (6, 6)  10 values of a.

4 – LINEAR EQUATION

Problem-10
A1 walks down to the bottom of an escalator that is moving up and he counts 150 steps. His friend, Bob,
walks up to the top of the escalator and counts 75 steps. If A1's speed of walking (in steps per unit time) is
three times Bob's walking speed, how many steps are visible on the escalator at a given time ? (Assume that
this value is constant.) [1987]
Solutions
Solution 1
Let the total number of steps be x, the speed of the escalator be e and the speed of Bob be b. In the time it
took Bob to climb up the escalator he saw 75 steps and also climbed the entire escalator. Thus the
contribution of the escalator must have been an addition x – 75 steps. Since Bob and the escalator were both
moving at a constant speed over the time it took Bob to climb, the ratio of their distances covered is the
e x − 75
same as the ratio of their speeds, so = .
b 75
Similarly, in the time it took A1 to walk down the escalator he saw 150 steps, so the escalator must have
e 150 − x e 150 − x
moved 150 – x steps in that time. Thus = or = .
3b 150 b 50
e x − 75 150 − x
Equating the two values of we have = and so 2x – 150 = 450 – 3x and 5x = 600 and x =
b 75 50
120, the answer.

26
Solution 2
Again, let the total number of steps be x, the speed of the escalator be e and the speed of Bob be b (all "per
unit time").
Then this can be interpreted as a classic chasing problem: Bob is "behind" by x steps, and since he moves at
x 75
a pace of b + e relative to the escalator, it will take = time to get to the top.
b+e e
x 150
Similarly, Al will take = time to get to the bottom.
3b − e e

From these two equations, we arrive at

ex 2ex 6ex (ex) − (6ex) 5x
150 = = 2.75 = = = =  600 = 5x  x = 120 , where we have
3b − e b + e 3b + 3e (3b − e) − (3b + 3e) 4
a c ac
used the fact that = = (the proportion manipulations are motivated by the desire to isolate x,
b d bd
prompting the isolation of the 150 on one side, and the fact that if we could cancel out the b’s, then the e’s in
the numerator and denominator would cancel out, resulting in an equation with x by itself).

Problem-6
It is possible to place positive integers into the vacant twenty-one squares of the 5 × 5 square shown below
so that the numbers in each row and column form arithmetic sequences. Find the number that must occupy
the vacant square marked by the asterisk (*). [1988]
*
74

186

103

Solution
Solution 1 (specific) – Let the coordinates of the square at the bottom left be (0, 0 ), the square to the right
(1, 0), etc. Label the leftmost column (from bottom to top) 0, a, 2a, 3a, 4a and the bottom-most row (from
left to right) 0, b, 2b, 3b, 4b. Our method will be to use the given numbers to set up equations to solve for a
and b, and then calculate (*).
4a *
3a 74
2a 186
a 103
0 b 2b 3b 4b
We can compute the squares at the intersections of two existing numbers in terms of a and b; two such
equations will give us the values of a and b. On the fourth row from the bottom, the common difference is 74
– 3a, so the square at (2, 3) has a value of 148 – 3a. On the third column from the left, the common
difference is 103 – 2b, so that square also has a value of 2b + 3(103 – 2b) = 309 – 4b. Equating, we get 148
– 3a = 309 – 4b  4b – 3a = 161.
Now we compute the square (2, 2). By rows, this value is simply the average of 2a and 186, so it is equal to
2a + 186
= a + 93. By columns, the common difference is 103 – 2b, so our value is 206 – 2. Equating, a + 93
2
= 206 – 2b  a +2b = 113.
Solving 4b – 3a = 161
3 + 2b = 113

27
Gives a = 13, b = 50. Now it is simple to calculate (4, 3). One way to do it is to see that (2, 2) has 206 – 2b =
106 + 186
106 and (4, 2) has 186, so (3, 2) has = 146.
2
(3, 0) + (3, 4)
Now, (3, 0) has 3b = 150, so (3, 2) =  (3, 4) = * = 142.
2
Solution 2 (general)
First, let a = the number to be placed in the first column, fourth row. Let b = the number to be placed in the
second column, fifth row. We can determine the entire first column and fifth row in terms of a and b :

4a
3a
2a
a
0 b 2b 3b 4b

Next, let a + b + c = the number to be placed in the second column, fourth row. We can determine the entire
second column and fourth row in terms of , , and c :
4a 4a + b + 4c
3a 3a + b + 3c
2a 2a + b + 2c
a a+b+c a + 2b + 2c a + 3b + 3c a + 4b + 4c
0 b 2b 3b 4b
We have now determined at least two values in each row and column. We can finish the table without
introducing any more variables:
4a 4a + b + 4c 4a + 2b + 8c 4a + 3b + 2c 4a + 4b + 16c
3a 3a + b + 3c 3a + 2b + 6c 3a + 3b + 9c 3a + 4b + 12c
2a 2a + b + 2c 2a + 2b + 4c 2a + 3b + 6c 2a + 4b + 8c
a a+b+c a + 2b + 2c a + 3b + 3c a + 4b + 4c
0 b 2b 3b 4b

We now have a system of equations.

3a + b + 3c = 74
2a + 4b + 8c = 186
a + 2b + 2c = 103
Solving, we find that (a, b, c) = (13, 50, – 5). The number in the square marked by the asterisk is 4a + 3b +
12c = 142 .

5 - LOGARITHM
Problem -1
Let x, y and z all exceed , and let w be a positive number positive number such that logxw = 24, logyw = 40
and logxyzw = 12. Find logzw. [AIME 1983]

Solutions
Solutions 1
The logarithmic notation doesn't tell us much, so we'll first convert everything to the equivalent exponential
expressions.
X24 = w, y40 = w, and (xyz)12 = w. If we now convert everything to a power of 120, it will be easy to isolate
z and w.
x120 = w5, y120 = w3, and (xyz)120 = w10
With some substitution, we get w w z = w10 and logzw = 060 .
5 3 120

28
Solutions 2
Applying the change of base formula,
log w log x 1
Logxw = 24  = 24  =
log x log w 24
log w log y 1
Logyw = 40  = 40  =
log y log w 40
log w log x + log y + log z 1
Logxyzw = 12  = 12  =
log xyz log w 12
Therefore,
log z 1 1 1 1
= − − = .
log w 12 24 40 60
Hence, logzw = 060 .

Problem-5
Determine the value of ab if log8 a + log4 b2 = 5 and log8 b + log4 a2 = 7. [AIME 1984]

Solutions
Solutions 1
log a 2log b log ab3
Use the change of base formula to see that + = 5 ; combine denominators to find that =
log 8 log 4 3 log 2
log a 3 b
5. Doing the same thing with the second equation yields that = 7.
3 log 2
This means that log ab3 = 15 log 2  ab3 = 215 and that log a3b = 21 log 2  a3b = 221.
If we multiply the two equations together, we get that a4b4 = 236, so taking the fourth root of that, ab = 29 =
512 .

Solutions 2
We can simplify our expressions by changing everything to a common base and by pulling exponents out of
In a 2 In b In b 2 In a
the logarithms. The given equations then become + = 5 and + = 7 . Adding the
In 8 In 4 In 8 In 4
1 2 12
equations and factoring, we get ( + ) (In a + In b) =12 . Rearranging we see that In ab = .
In 8 In 4 1 2
+
in 8 In 4
12 12 In 2
Again, we pull exponents out of our logarithms to get In ab = = = 9 In 2 .
1 2 1
+ +1
3in 2 2 In 2 3
In ab
This means that = 9 . The left-hand side can be interpreted as a base-2 logarithm, giving us ab = 29 =
In 2
512 .

Solutions 3
This solution is very similar to the above two, but it utilizes the well-known fact that log mk n k = log m n .
Thus,
Log8 a + log4 b2 = 5  log 23 ( 3 a )3 + log 22 b 2 = 5  log 2 3 a + log 2 b = 5  log 2 3 ab = 5 .
Similarly, log8 b + log4 a2 = 7  log 2 3 ba = 7 .
Adding these two equations, we have
4 4 3
12
log 2 a 3 b 3 = 12  ab = 2 4
= 29 = 512 .

29
Problem-3
Find (log2 x)2 if log2 (log8 x) = log8 (log2 x). [AIME 1988]
Solutions
Raise both as exponents with base 8:
8log 2(log8 x) = 8log8 (log2 x)
23log2 (log8x) = log 2 x
(log8 x)3 = log 2 x
 log 2 x 
  3 = log 2 x
 log 2 8 
(log2x)2 = (log28)3 = 27 .
A quick explanation of the steps: On the 1st step, we use the property of logarithms that a logax = x. On the
2nd step, we use the fact that k loga x = log a xk. On the 3rd step, we use the change of base formula, which
log k b
states log a b = for arbitrary .
log k a

Problem 6
The graphs of the equations
y = k, y = 3x + 2k , y = − 3x + 2k ,
are drawn in the coordinate plane for k = – 10, – 9, – 8, ….., 9, 10.
2
These 63 lines cut part of the plane into equilateral triangles of side . How many such triangles are
3
formed? [AIME 1994]
Solutions
We note that the lines partition the hexagon of the six external lines into disjoint unit regular triangles, and
forms a series of unit regular triangles along the edge of the hexagon.

20
Solving the above equations for k = ±10, we see that the hexagon in question is regular, with side length
3
. Then, the number of triangles within the hexagon is simply the ratio of the area of the hexagon to the area

30
of a regular triangle. Since the ratio of the area of two similar figures is the square of the ratio of their side
lengths, we see that the ratio of the area of one of the six equilateral triangles composing the regular hexagon
2
 20 / 3 
to the area of a unit regular triangle is just   = 100. Thus, the total number of unit triangles is 6 ×
 2/ 3 
100 = 600. There are 6.10 equilateral triangles formed by lines on the edges of the hexagon. Thus, our
answer is 600 + 60 = 660 .

6 - MODULUS FUNCTION

Problem-2
Let f(x) = |x – p | + |x – 15| + |x – p – 15|, where p ≤ x ≤ 15. Determine the minimum value taken by f(x) for
x in the interval 0 < x ≤ 15. [1983]
Solution
It is best to get rid of the absolute value first.
Under the given circumstances, we notice that |x – p| = x – p, |x – 15| = 15 – x, and |x – p – 15| = 15 + p – x .
Adding these together, we find that the sum is equal to 30 – x, of which the minimum value is attained when
x = 15 .
Edit: |x – p – 15| can equal 15 + p – x or x – p – 15 (for example, if x = 7 and p = – 12, x – p – 15 = 4.
Thus, our two "cases" are 30 – x (if x – p ≤ 15) and x – 2p (if x – p ≥ 15).
However, both of these cases give us 15 as the minimum value for f(x), which indeed is the answer posted
above.

Problem-4
Suppose that |xi| < 1 for i = 1, 2, ….., n. Suppose further that |x1 + |x2| + …. + |xn| == 19 + |x1 + x2 + … + xn|.
What is the smallest possible value of n ? [1988]
Solution
Since |xi| < 1 then |x1 + |x2| + …. + |xn| == 19 + |x1 + x2 + … + xn| < n
So n ≥ 20. We now just need to find an example where n = 20 :
19 19 19 19 19
suppose x2k – 1 = and x2k – 1 = – ; then on the left hand side we have +− + ... + − = 20
20 20 20 20 20
 19 
  = 19.
 20 
19 19 19
On the right hand side, we have 19 + − + ... − = 19 + 0 = 19, and so the equation can hold for N =
20 20 20
020 .

11 - CIRCLES
Problem-11
The length of diameter AB is a two digit integer. Reversing the digits gives the length of a perpendicular
chord CD. The distance from their intersection point H to the center O is a positive rational number.
Determine the length of AB. [1983]

31
 C

A B
H O

D
Solution
AB 10x + y CD 10y + x
Let AB = 10x + y and CD = 10y + x. It follows that CO = = and CH = = .
2 2 2 2
 10x + y   10y + x 
2 2
9
Applying the Pythagorean Theorem on CO and CH, OH =   −  = .11(x + y)(x − y)
 2   2  4
3
= 11(x + y)(x − y) .
2
Because OH is a positive rational number, the quantity 11(x + y)(x − y) cannot contain any square roots.
Either x – y or x + y must be 11. However, x – y cannot be 11, because both must be digits. Therefore, x + y
must equal eleven and x – y must be a perfect square (since x + y) (since x + y > x – y). The only pair (x, y)
that satisfies this condition is (6, 5) so our answer is 065 .

Problem-13
In the adjoining figure, two circles with radii 6 and 8 are drawn with their centers 12 units apart. At P, one of
the points of intersection, a line is drawn in such a way that the chords QP and PR have equal length. (P is
the midpoint of QR) Find the area of the square with a side length of QP. [1983]

R
12

Solution
Solution 1
First, notice that if we reflect R over P we get Q. Since we know that R is on circle B and Q is on circle A,
we can reflect circle B over P to get another circle (centered at a new point C with radius 6) that intersects
circle A at Q. The rest is just finding lengths:
Since P is the midpoint of segment BC, AP is a median of triangle ABC. Because we know that AB = 12, BP
= PC = 6, and AP = 8, we can find the third side of the triangle using Stewart's Theorem or similar
approaches. We get AC = 56 .

32
So now we have a kite AQCP with AQ = AP = 8, CQ = CP = 6, and AC = 56 , and all we need is the length
of the other diagonal PQ. The easiest way it can be found is with the Pythagorean Theorem. Let 2x be the
length of PQ. Then
36 − x 2 + 64 − x 2 = 56 .
65
Doing routine algebra on the above equation, we find that x2 = 2 2
, so PQ = 4x = 130 .
2
Solution 2

Q
M
P
N
R
A 12 B 6

Draw additional lines as indicated. Note that since triangles AQP and BPR are isosceles, the altitudes are
also bisectors, so let QM = MP = PN = NR = x.
AR BR
Since = , triangles BNR and AMR are similar. If we let y = BN, we have AM = 3BN = 3y.
MR NR
Applying the Pythagorean Theorem on triangle BNR, we have x2 + y2 = 36. Similarly, for triangle QMA, we
have x2 + 9y2 = 64.
7 65
Subtracting, 8y2 = 28  y2 =  x2 =  QP2 = 4x2 = 130 .
2 2

Solution 3
Let QP = PR = x. Angles QPA, APB, and BPR must add up to 180º. By the Law of Cosines, APB = cos–1(–
11/24). Also, angles QPA and BPR equal cos–1(x/16) and cos–1(x12).
So we have cos–1(x/16) + cos–1(–11/24) = 180 – cos–1(x/12).
Taking the cos of both sides and simplifying using the cosine addition identity gives x2 = 130.

Solution 4
Observe that the length of the area where the two circles intersect can be found explicitly as 2. Let QP = PR
= x, then the power of point R with regards to the larger circle gives x(2x) = 10(26)  x2 = 130 .

Problem-14
The adjoining figure shows two intersecting chords in a circle, with B on minor arc AD. Suppose that the
radius of the circle is 5, that BC = 6, and that AD is bisected by BC. Suppose further that AD is the only
chord starting at A which is bisected by BC. It follows that the sin e of the minor arc AB is a rational
m
number. If this fraction is expressed as a fraction in lowest terms, what is the product mn ? [1983]
n

33
 B
M

C
A

Solution
Let A be any fixed point on circle O and let AD be a chord of circle O. The locus of midpoints N of the
chord AD is a circle P, with diameter AO.
Generally, the circle P can intersect the chord BC at two points, one point, or they may not have a point of
intersection. By the problem condition, however, the circle P is tangent to BC at point N.
Let M be the midpoint of the chord BC. From right triangle OMB,
BM 3
OM = OB2 − BM 2 = 4. Thus, tan  BOM = = .
OM 4
Notice that the distance OM equals PN + PO cos AOM = r(1 + cos AOM) (Where r is the radius of circle P).
OM 2OM 8 3
Evaluating this, cos  AOM = –1= –1= –1= .
r r 5 5
1 − cos AOM
2
5 2 − 32 4
From cos  AOM, we see that tan  AOM = = =
cos AOM 3 3
Next, notice that  AOB =  AOM –  BOM. We can therefore apply the tangent subtraction formula to
4 3
−
tan AOM − tan BOM 3 4 7 7 7
obtain , tan AOB = = = . It follows that sin AOB = = ,
1 + tan AOM.tan.AOM 1 + 4 . 3 24 7 + 24
2 2 25
3 4
resulting in an answer of 7.25 = 175 .
Problem-6
Three circles, each of radius 3, are drawn with centers at (14, 92), (17, 76), and (19, 84). A line passing
through (17, 76) is such that the total area of the parts of the three circles to one side of the line is equal to
the total area of the parts of the three circles to the other side of it. What is the absolute value of the slope of
this line ? [1984]

34
Solution

A = (14, 92)

M = (16.5, 88)

C = (19, 84)

B = (17, 76)

The line passes through the center of the second circle; hence it is the circle's diameter and splits the circle
into two equal areas. For the rest of the problem, we do not have to worry about that circle.
Solution 1
Draw the midpoint of AC (the centers of the other two circles), and call it M. If we draw the feet of the
perpendiculars from A, C to the line (call E, F), we see that AEM  CFM by HA congruency; hence M
 19 + 14 84 + 92   33 
lies on the line. The coordinates of M are  ,  =  ,88  .
 2 2   2 
 
 88 − 76 
Thus, the slope of the line is   = – 24, and the answer is 024 .
33
 − 17 
 2 
Remark: Notice the fact that the radius is 3 is not used in this problem; in fact changing the radius does not
affect the answer.

35
Solution 2
Define E, F to be the feet of the perpendiculars from A, C to the line (same as above). The equation of the
line is y = mx + b; substituting y = 76, x = 17 gives us that b = 76 – 17m, so the line is y = mx + (76 –
17m). AE = CF by the HA argument above and CPCTC, so we can use the distance of a point to a line
formula and equate.
m(14) − 92 + (76 − 17m) m(19) − 84 + (76 − 17m)
=
m +1
2
m2 + 1
– 3M – 16 = – 2m + 8
m = – 24
And | – 24 | = 24.

Problem-9
In a circle, parallel chords of lengths 2, 3, and 4 determine central angles of and radians,
respectively, where <If cos , which is a positive rational number, is expressed as a fraction
in lowest terms, what is the sum of its numerator and denominator ? [1985]
Solution

2
3

4


All chords of a given length in a given circle subtend the same arc and therefore the same central angle.
Thus, by the given, we can re-arrange our chords into a triangle with the circle as its circumcircle.

36
 2


 4

2+3+ 4 9 5 3 1 3
This triangle has semiperimeter so by Heron's formula it has area K = . . . = 15 .
2 2 2 2 2 4
The area of a given triangle with sides of length a, b, c and circumradius of length R s also given by the
abc 6 3 8
formula K = so = 15 and R = .
4R R 4 15
Now, consider the triangle formed by two radii and the chord of length 2. This isosceles triangle has vertex
angle so by the Law of Cosines,
2R 2 − 4 17
22 = R2 + R2 – 2R2 cos   cos  = = and the answer is 17 + 32 = 049 .
2R 2 32

Problem-2
Ten points are marked on a circle. How many distinct convex polygons of three or more sides can be drawn
using some (or all) of the ten points as vertices ? [1989]
Solution
Any subset of the ten points with three or more members can be made into exactly one such polygon. Thus,
we need to count the number of such subsets. There are 210 = 1024 total subsets of a ten-member set, but of
 10   10   10 
these   = 1 have 0 members,   = 10 have 1 member and   = 45 have 2 members. Thus the answer
0 1 2
is 1024 – 1 – 10 – 45 = 968.

Problem-12
A regular 12-gon is inscribed in a circle of radius 12. The sum of the lengths of all sides and diagonals of the
12-gon can be written in the form a + b 2 + c 3 +d 6 , where a, b, c and d are positive integers. Find a
+ b + c + d. [1990]

37
Solution 1

The easiest way to do this seems to be to find the length of each of the sides and diagonals. To do such, draw
the radii that meet the endpoints of the sides/diagonals; this will form isosceles triangles. Drawing the
altitude of those triangles and then solving will yield the respective lengths.
6− 2
• The length of each of the 12 sides is 2.12 sin 15. 24 sin 15 = 24 sin(45 – 30) = = 6( 6 –
4
2)
• The length of each of the 12 diagonals that span across 2 edges is 2.12 sin 30 = 12 (or notice that the
triangle formed is equilateral).
• The length of each of the 12 diagonals that span across 3 edges is 2.12 sin 45 = 12 2 (or notice that
the triangle formed is a 45 – 45 – 90 right triangle).
• The length of each of the 12 diagonals that span across 4 edges is 2.12 sin 60 = 12 3 .
• The length of each of the 12 diagonals that span across 5 edges is 2.12 sin 75 = 24 sin (45 + 30) = 24
6+ 2
= 6 ( 6 + 2 ).
4
• The length of each of the 6 diameters is 2.12 = 24.

Adding all of these up, we get

12 [6( 6 – 2 ) + 12 + 12 2 + 12 3 + 6 ( 6 + 2 )] + 6.24
= 12 (12 + 12 2 + 12 3 +12 6 ) + 144 = 288 + 144 2 + 144 3 + 144 6.
Thus, the answer is 144.5 = 720 .

Solution 2
A second method involves drawing a triangle connecting the center of the 12-gon to two vertices of the 12-
gon. Since the distance from the center to a vertex of the 12-gon is , the Law of Cosines can be applied to
this isosceles triangle, to give:
a2 = 122 + 122– 2.12.12.cos 
a2 = 2.122 – 2.122 cos 
a2 = 2.122(1 – cos )
a2 = 12 2 . 1 − cos 
There are six lengths of sides/diagonals, corresponding to
 = 30º , 60 º, 90 º, 120 º, 150 º, 180 º
38
Call these lengths a1 , a2 , a3, a4, a5, a6, from shortest to longest. The total length  that is asked for is = 12
(a1 + a2 + a3 + a4 + a5) + 6a6, noting that a6 as written gives the diameter of the circle, which is the longest
diagonal.
 = 12 [12 1 − cos150º ( 1 − cos30º + 1 − cos30º + 1 − cos90º ) + 1 − cos120º + 1 − cos150º )]

3 1 1 3
 = 144 2 ( 1− + 1 − + 1− 0 + 1 + + 1+ ) + 144
2 2 2 2
 3 2 6 3
 = 144 2  1− + +1+ + 1+  + 144
 2 2 2 2 
 
To simplify the two nested radicals, add them, and call the sum x :
3 3
x = 1− + 1+
2 2

Squaring both sides, the F and L part of FOIL causes the radicals to cancel, leaving 2 :
 3  3
x2 = 2 + 2  1 −  1 +
 
 2  2 
 
3
x2 = 2 + 2 1 −
4
1
x2 = 2 + 2
4
1
x2 = 2 + 2 .
2
x= 3.
Plugging that sum x back into the equation for , we find
 2 6 
= 144 AB  +1+ + c  + 144
 2 2 

= 144 + 144 2 + 144 3 + 144 6 + 144.

Thus, the desired quantity is 144.5 = 720 .

Problem-11
Twelve congruent disks are placed on a circle C of radius 1 in such a way that the twelve disks cover C, no
two of the disks overlap, and so that each of the twelve disks is tangent to its two neighbors. The resulting
arrangement of disks is shown in the figure below.
The sum of the areas of the twelve disks can be written in the from (a – b c ), where a, b, c are positive
integers and is not divisible by the square of any prime. Find a + b + c. [1991]

39
 1

Solution
We wish to find the radius of one circle, so that we can find the total area.
Notice that for them to contain the entire circle, each pair of circles must be tangent on the larger circle.
Now consider two adjacent smaller circles. This means that the line connecting the radii is a segment of
length 2r that is tangent to the larger circle at the midpoint of the two centers. Thus, we have essentially have
a regular dodecagon whose vertices are the centers of the smaller triangles circumscribed about a circle of
radius 1.
We thus know that the apothem of the dodecagon is equal to 1. To find the side length, we make a triangle
consisting of a vertex, the midpoint of a side, and the center of the dodecagon, which we denote A, M and O
respectively. Notice that OM = 1, and that OM A is a right triangle with hypotenuse OA and m  MOA =
15º. Thus AM = (1) tan 15º = 2 – 3 , which is the radius of one of the circles. The area of one circle is thus
(2 – 3 )2 = (7 – 4 3 ), so the area of all 12 circles is (84 – 48 3 ), giving an answer of 8 + 48 + 3 =
135 .

Problem-14
A hexagon is inscribed in a circle. Five of the sides have length 81 and the sixth, denoted by AB , has length
31. Find the sum of the lengths of the three diagonals that can be drawn from A. [1991]
Solution

E D
8
1
8 E 8

1 z 1
y
F E C
E
E 8 x
8
1
3 1
E
A B E
1
Let x = AC = BF, y = AD = BE, and z = AE = BD. E
40
Ptolemy's Theorem on ABCD gives 81y + 31.81 = xz, and Ptolemy on ACDF gives x.z + 812 = y2.
Subtracting these equations give y2 – 81y – 112.81 = 0, and from this y = 144. Ptolemy on ADEF gives 81Y
+ 812 = z2, and from this z = 135. Finally, plugging back into the first equation gives x = 105, so x + y + z =
105 + 144 + 135 = 384 .

Problem-9
Trapezoid ABCD has sides AB = 92, BC = 50, CD = 19, and AD = 70, with AB parallel to CD. A circle with
m
center P on AB is drawn tangent to BC and AD. Given that AP = , where m and n are relatively prime
n
positive integers, find m + n. [1992]
Solution
Solution 1
Let AB be the base of the trapezoid and consider angles A and B. Let x = AP and let h equal the height of the
trapezoid. Let r equal the radius of the circle.

Then
r h r h
sin A = = and sin B = = . (1)
x 70 92 − x 50
Let z be the distance along AB from A to where the perp from D meets AB.
44710959
Then h2 + z2 = 702 and (73 – z)2 + h2 = 502 so h = .
146
11753 161
We can substitute this into (1) to find that x = = and m + n = 164.
219 3
Remark: One can come up with the equations in (1) without directly resorting to trig. From similar
triangles, h/r = 70/x and h/r = 50/(92 – x).
This implies that 70/x = 50/(92 – x), so x = 161/3.

Solution 2
7r 50r 120r
From (1) above, x = and 92 – x = . Adding these equations yields 92 = .
h h h
70r 7 120r 7 161
Thus, x = = . = .92 = , and m + n = 164 .
h 12 h 12 3
We can use (1) from Solution 1 to find that h/r = 70/x and h/r = 50/ (92 – x ).
This implies that 70/x = 50/(92 – x) so x = 161/3.
Solution 3
Extend AD and BC to meet at a point X. Since AB and CD are parallel, XCD XAB. If AX is further
extended to a point A´ and XB is extended to a point B´ such that A´ B´ is tangent to circle P, we discover
that circle P is the incircle of triangle XA´B´. Then line XP is the angle bisector of AXB. By homothety, P
is the intersection of the angle bisector of XAB with AB. By the angle bisector theorem,
AX XB AX XD XB XC AD BD 7
=  − = −  = =
AP BP AP AP BP BP AP PB 5
Let 7a = AP, then AB = 7a + 5a = 12a.
7 92  7 161
AP = (AB) = = . Thus, m + n = 164. \
12 12 3

Solution 4
 19 + 92 
The area of the trapezoid is   , where is the height of the trapezoid.
 2 
Draw lines CP and BP. We can now find the area of the trapezoid as the sum of the areas of the three
triangles BPC, CPD and PBA.
1
[BPC] = .50.r (where r is the radius of the tangent circle.)
2

41
 1
[CPD] = .19.r
2
1
[PBA] = .70.r
2
[BPC] + [CPD] + [PBA] = 60r +
19h
= [ABCD] =
(19 + 2 ) h
2 2
60r = 46h
23h
r=
30
h h
From Solution 1 above, =
70 70
23h 161
Substituting r = , we find x = , hence the answer is 164 .
30 3

Problem-9
Two thousand points are given on a circle. Label one of the points 1. From this point, count 2 points in the
clockwise direction and label this point 2. From the point labeled 2, count 3 points in the clockwise direction
and label this point 3. (See figure.) Continue this process until the labels 1, 2, 3 … , 1993 are all used. Some
of the points on the circle will have more than one label and some points will not have a label. What is the
smallest integer that labels the same point as 1993 ? [1993]

3 1
Solution
1
The label 1993 will occur on the (1993)(1994) (mod 2000)th point around the circle. (Starting from 1) A
2
1 1
number n will only occupy the same point on the circle if (n)(n + 1)  (1993)(1994) (mod 2000).
2 2
Simplifying this expression, we see that (1993)(1994) – (n)(n + 1) = (1993 – n)(1994 + n)  0.
Therefore, one of 1993 – n or 1994 + n is odd, and each of them must be a multiple of 125 or 16.
For 1993 – n to be a multiple of 125 and 1994 + n to be a multiple of 16, n  118 (mod 125) and n  6
(mod 16). The smallest n for this case is 118.
In order for 1993 – n to be a multiple of 16 and 1994 + n to be a multiple of 125, n  9 (mod 16) and n  6
(mod 125). The smallest n for this case is larger than 118, so 118 is our answer.
Note: One can just substitute 193  – 7 (mod 2000) and 1994  – 6 (mod 2000) to simplify calculations.

Problem-13
Jenny and Kenny are walking in the same direction, Kenny at 3 feet per second and Jenny at 1 foot per
second, on parallel paths that are 200 feet apart. A tall circular building 100 feet in diameter is centered
midway between the paths. At the instant when the building first blocks the line of sight between Jenny and
Kenny, they are 200 feet apart. Let t be the amount of time, in seconds, before Jenny and Kenny can see
each other again. If t is written as a fraction in lowest terms, what is the sum of the numerator and
denominator ? [1993]
Solution
Solution 1
Consider the unit circle of radius 50. Assume that they start at points (– 50, 100) and (– 50, – 100). Then at
time t, they end up at points (– 50 + t, 100) and (– 50 + 3t, – 100). The equation of the line connecting these
points and the equation of the circle are
100 5000
y=– x + 200 – (1)
t t
502 = x2 + y2 (2)

42
When they see each other again, the line connecting the two points will be tangent to the circle at the point
x 100
(x, y). Since the radius is perpendicular to the tangent we get − = − or xt = 100y.
y t
xt 5000
Now substitute y = into (2) and get x =
100 1002 + t 2
xt 160
Now substitute this and y = into (1) and solve for t to get t = .
100 3
Finally, the sum of the numerator and denominator is 160 + 3 = 163 .
Solution 2
Let A and B be Kenny's initial and final points respectively and define C and D similarly for Jenny. Let O be
the center of the building. Also, let X be the intersection of AC and B. Finaly, let P and Q be the points of
tangency of circle O to AC and BD respectively.

D
O

A X
P C
B
 CD 
From the problem statement, AB = 3t, and CD =t. Since ABX ~ CDX, CX = AC.   = 200.
 AB − CD 
 t 
  = 100.
 3t − t 
Since PC = 100, PX = 200.
OP 50 1
So, tan (OXP) = = = .
PX 200 4
Since circle O is tangent to BX and AX, OX is the angle bisector of BXA.
1
2.  
2 tan(OXP) 4 = 8
Thus, tan (BXA) = tan(2OXP) = = .
1 − tan (OXP)
2
1
2
15
1−  
4
8 160
Therefore, t = CD = CX.tan(BXA) = 100. = , and the answer is 163 .
15 3

Problem-15
Let CH be an altitude of ABC. Let R and S be the points where the circles inscribed in the triangles ACH
and BCH are tangent to CH . If AB = 1995, AC = 1994, and BC = 1993, then RS can be expressed as m/n,
here and are relatively prime integers. Find m + n. [1993]

43
Solution

1994
1993

R
S

H A
B
1995
From the Pythagorean Theorem, AH2 + CH2 = 19942, and (1995 – AH2) + CH2 = 19932.
Subtracting those two equations yields AH2 – (1995 – AH)2 = 3987.
1995 3987
After simplification, we see that 2*1995AH –19952 = 3987, or AH = + .
2 2*1995
1995 3987
Note that AH + BH = 1995. Therefore we have that BH = – .
2 2*1995
3987
Therefore AH – BH = .
1995
AH + CH − AC CH + BH − BC
Now note that RS = |HR – HS|, RH = , and HS = .
2 2
Therefore we have,
AH + CH − AC − CH − BH + BC AH − BH − 1994 + 1993
RS = =
2 2
1992 332
Plugging in AH – BH and simplifying, we have RS = =  332 + 665 = 997 .
1995*2 665

12 – SET THEORY

Problem-12
For {1, 2, 3, …. , N } and each of its non-empty subsets, an alternating sum is defined as follows. Arrange
the number in the subset in decreasing order and then, beginning with the largest, alternately add and
subtract succesive numbers. For example, the alternating sum for {1, 2, 4, 6, 9}is 9 – 6 + 4 – 2 + 1 = 6 and
for {5} it is simply 5. Find the sum of all such alternating sums for N = 7. [1983]
Solution
Solution 1
Let S be a non- empty subset of {1, 2, 3, 4, 5, 6}.
Then the alternating sum of S plus the alternating sum of S with 7 included is 7. In mathematical terms, S +
(S  7) = 7. This is true because when we take an alternating sum, each term of S has the opposite sign of
each corresponding term of S 7.

44
Because there are 63 of these pairs, the sum of all possible subsets of our given set is 63 * 7. However, we
forgot to include the subset that only contains 7, so our answer is 64.7 = 448 .

Solution 2
Consider a given subset T of S that contains 7; then there is a subset T´ which contains all the elements of T
except for 7, and only those. Since each element of T´ has one element fewer preceding it than it does in T,
their signs are opposite; so the sum of the alternating sums of T and T´ is equal to 7. There are 26 subsets
containing 7, so our answer is 7 * 26 = 448 .

Problem-12
Let the sum of a set of numbers be the sum of its elements. Let S be a set of positive integers, none greater
than 15. Suppose no two disjoint subsets of S have the same sum. What is the largest sum a set S with these
properties can have? [1986]
Solution
The maximum is 061 , attained when S = {15, 14, 13, 11, 8}. We must now prove that no such set has sum
at least 62. Suppose such a set S existed. Then S must have more than 4 elements, otherwise its sum would
be at most 15 + 14 + 13 + 12 = 54 if it had 4 elements.
 6  6  6
But also, S can't have at least 6 elements. To see why, note that at least 1 + 6 +   +   +   = 57 of its
 2  3  4
subsets have at most four elements (the number of subsets with no elements plus the number of subsets with
one element plus the number of subsets with two elements plus the number of subsets with three elements
plus the number of subsets with four elements), so each of them have sum at most 54. By the Pigeonhole
Principle, two of these subsets would have the same sum, a contradiction to the givens.
Thus, S must have exactly 5 elements. S contains both 15 and 14 (otherwise its sum is at most 10 + 11 + 12
+ 13 + 15 = 61). It follows that S cannot contain both a and a – 1 for any a ≤ 13, or the subsets {a, 14} and
{a – 1, 15} would have the same sum. So now S must contain 13 (otherwise its sum is at most 15 + 14 + 12
+ 10 + 8 = 59), and S cannot contain 12, or the subsets {12, 15} and {13, 14} would have the same sum.
Now the only way S could have sum at least 62 = 15 + 14 + 13 + 11 + 9 would be if S = {15, 14, 13, 11, 9}.
But the subsets {9, 15} and {11, 13} have the same sum, so this set does not work, a contradiction to the
givens. Therefore no S with sum at least 62 is possible and 61 is indeed the maximum.

Problem-13
Let S be a subset of {1, 2, 3, …. , 1989} such that no two members of S differ by 4 or 7. What is the largest
number of elements S can have? [1989]
Solution
We first show that we can choose at most 5 numbers from {1, 2, …. , 11} such that no two numbers have a
difference of 4 or 7. We take the smallest number to be 1, which rules out 5, 8. Now we can take at most one
from each of the pairs: [2, 9], [3, 7], [4, 11], [6, 10].
Now, 1989 = 180.11 + 9, but because this isn't an exact multiple of 5, we need to consider the last 9
numbers.
Now let's examine {1, 2, ….. , 20}. If we pick 1, 3, 4, 6, 9 from the first 11 numbers, then we're allowed to
pick 11 + 1, 11 + 3, 11 + 4, 11 + 6, 11 + 9. This means we get 10 members from the 20 numbers. Our answer
is thus 179.5 + 10 = 905 .

Problem-10
The sets A = {z : z18 = 1} and B = {w : w48 = 1} are both sets of complex roots of unity. The set C = {zw : z
A and w B} is also a set of complex roots of unity. How many distinct elements are in C ? [1990]
Solution
Solution 1
The least common multiple of 18 and 48 is 144, so define n = e2i/144. We can write the numbers of set A as
{n8, n16, … n144} and of set B as {n3, n6, … n144}. nx can yield at most 144 different values. All solutions for
zw will be in the form of n8k1+3k2 .
8 and 3 are relatively prime, and it is well known that for two relatively prime integers a, b the largest
number that cannot be expressed as the sum of multiples of a, b is a.b – a – b. For 3, 8, this is 13; however,

45
we can easily see that the numbers 145 to 157 can be written in terms of 3, 8.
Since the exponents are of roots of unities, they reduce mod 144, so all numbers in the range are covered.
Thus the answer is 144 .

Solution 2
 2 k1   2 k 2 
The 18 and 48th roots of 1 can be found by De Moivre's Theorem. They are cis   and cis  
 18   48 
respectively, where cis = cos + i sin and k1 and k2 are integers from 0 to 17 and 0 to 47, respectively.
k k   8 k1 + 3 k 2 
zw = cis  1 + 2  = cis  .
 9 24   72 
ince the trigonometric functions are periodic every 2, there are at most 72.2 = 144 distinct elements in C.
As above, all of these will work.

Problem-8
Let S be a set with six elements. In how many different ways can one select two not necessarily distinct
subsets of S so that the union of the two subsets is S ? The order of selection does not matter; for example,
the pair of subsets {a, c}, {b, c, d, e, f} represents the same selection as the pair {b, c, d, e, f}, {a, c}.
[1993]
Solution
Solution 1
Call the two subsets m and n. For each of the elements in S, we can assign it to either m, n, or both. This
gives us 36 possible methods of selection. However, because the order of the subsets does not matter, each
possible selection is double counted, except the case where both m and n contain all 6 elements of S. So our
36 − 1
final answer is then + 1 = 365 .
2
Solution 2
6
Given one of   subsets with k elements, the other also has 2k possibilities; this is because it must contain
k
6
6
all of the "missing" n – k elements and thus has a choice over the remaining k. We want  k  2 k
= (2 + 1)6
k =0  
729 − 1
= 729 by Binomial Theorem. But the order of the sets doesn't matter, so we get + 1 = 365 .
2

13 – ARITHMATIC PROGRESSION

Problem-1
Find the value of a2 + a4 + a6 + a8 + ….. + a98 if a1, a2, a3 …. is an arithmetic progression with common
difference 1, and a1 + a2 + a3 + …. + a98 = 137. [1984]
Solution
Solution 1
One approach to this problem is to apply the formula for the sum of an arithmetic series in order to find the
value of a1, then use that to calculate a2 and sum another arithmetic series to get our answer.
A somewhat quicker method is to do the following: for each n ≥ 1, we have a2n – 1 = a2 – 1.
We can substitute this into our given equation to get (a2 – 1) + a2 + (a4 – 1) + a4 + ….. + (a98 – 1) + a98 = 137.
137 + 49
The left-hand side of this equation is simply 2(a2 + a4 + … + a98) – 49, so our desired value is =
2
093 .
Solution 2
If a1 is the first term, then a1 + a2 + a3 + … + a98 = 137 can be rewritten as:

46
 97.98
98a1 + 1 + 2 + 3 + …. + 97 = 137  98a1 + = 137
2
Our desired value is a2 + a4 + a6 + … + a98 so this is:
49a1 + 1 + 3 + 5 + … + 97
which is 49a1 + 492.
137 97.49
So, from the first equation, we know 49a1 = − . So, the final answer is:
2 2
137 − 97(49) + 2(49) 2
= 093 .
2
Solution 3
A better approach to this problem is to notice that from a1 + a2 + … a98 = 137 that each element with an odd
subscript is 1 from each element with an even subscript.
137 − 97
Thus, we note that the sum of the odd elements must be . Thus, if we want to find the sum of all of
2
137 − 97
the even elements we simply add 49 common differences to this giving us + 49 = 93 .
2

Problem-4
Let S be a list of positive integers - not necessarily distinct - in which the number 68 appears. The arithmetic
mean of the numbers in S is 56. However, if 68 is removed, the arithmetic mean of the numbers is 55.
What's the largest number that can appear in S ? [1984]
Solution
Suppose S has n members other than 68, and the sum of these members is s. Then we're given that
s + 68 s
= 56 and = 55. Multiplying to clear denominators, we have s + 68 = 56n + 56 and s = 55n so 68 =
n +1 n
n + 56, n = 12 and s = 12.55 = 660. Because the sum and number of the elements of S are fixed, if we want
to maximize the largest number in S, we should take all but one member of S to be as small as possible.
Since all members of S are positive integers, the smallest possible value of a member is 1. Thus the largest
possible element is 660 – 11 = 649 .

Problem-7
If the integer k is added to each of the numbers 36, 300, and 596, ne obtains the squares of three consecutive
terms of an arithmetic series. Find k. [1989]
Solution
Call the terms of the arithmetic progression a, a + d, a + 2d, making their squares a2 , a2 + 2ad + d2 , a2 + 4ad
+ 4d2 . We know that a2 = 36 + k and (a + d)2 = 300 + k, and subtracting these two we get 264 = 2ad + d2
(1). Similarly, using (a + d)2 = 300 + k and (a + 2d2)2 = 596 + k, subtraction yields 296 = 2ad + 3d2 (2).
Subtracting the first equation from the second, we get 2d2 = 32, so d = 4. Substituting backwards yields that
a = 31 and k = 925.

14 - TRIANGLE

Problem-3
A point P is chosen in the interior of ABC such that when lines are drawn through P parallel to the sides of
ABC, the resulting smaller triangles t1, t2 , and t3 in the figure, have areas 4, 9, and 49, respectively. Find
the area of ABC. [1984]

47
 A

t1 t2

t3
B C

Solution
By the transversals that go through P, all four triangles are similar to each other by the AA postulate. Also,
note that the length of any one side of the larger triangle is equal to the sum of the sides of each of the
absin C
corresponding sides on the smaller triangles. We use the identity K = to show that the areas are
2
proportional (the sides are proportional and the angles are equal) Hence, we can write the lengths of
corresponding sides of the triangle as 2x, 3x, 7x. Thus, the corresponding side on the large triangle is 12x,
and the area of the triangle is 122 = 144 .

Problem-6
As shown in the figure, triangle ABC is divided into six smaller triangles by lines drawn from the vertices
through a common interior point. The areas of four of these triangles are as indicated. Find the area of
triangle ABC. [1985]

C
A
84

35
40 30
A B
A A
Solution
Let the interior point be P, let the points on BC , CA and AB be D, E and F, respectively. Let x be the area
of AP and y be the area of CPD. Note that APF and BPF share the same altitude from P, so the
ratio of their areas is the same as the ratio of their bases. Similarly, ACF and BCF share the same
altitude from C, so the ratio of their areas is the same as the ratio of their bases. Moreover, the two pairs of
40 124 + x
bases are actually the same, and thus in the same ratio. As a result, we have: = or equivalently
30 65 + y
372 + 3x = 260 + 4y and so 4y = 3X + 112.
Y x + y + 84
Applying identical reasoning to the triangles with bases CD and BD , we get = so that 3y = x
35 105
+ y + 84 and 2y = x + 84. Substituting from this equation into the previous one gives x = 56, from which we
get y = 70 and so the area of ABC is 56 + 40 + 30 + 35 + 70 + 84 = 315.

Problem-9
In ABC, AB = 425, BC = 450, and AC = 510. An interior point P is then drawn, and segments are drawn
through P parallel to the sides of the triangle. If these three segments are of an equal length d, find d.
[1986]

48
Solution
Solution 1 (mass points)

X Z
45 42
P
0 5

Y
C A
51
Construct cevians AX, BY and CZ through P. Place masses0 of x, y, z on A, B and C respectively; then P has
mass x + y + z.
CP d
Notice that Z has mass x + y. On the other hand, by similar triangles, = .
CZ AB
x+y d
Hence by mass points we find that = .
x + y + z AB
y+z d z+x d
Similarly, we obtain = and =
x + y + z BC x + y + z CA
Summing these three equations yields
d d d x+y y+z z+x 2x + 2y + 2z
+ + = + + = = 2.
AB BC CA x + y + z x + y + z x+y+z x+y+z
Hence,
2 2 10
d= = =
1 1 1 1 1 1 1 1 1
+ + + + + +
AB BC CA 510 450 425 85 90 102
10 10 10
= = = = 306 .
1 1 1  1 1 35 3 10
 + + . +
5  17 18  102 5 306 306 306
Solution 2

E D´

450 P 425
E´ D

C A
F 510 F´

49
Let the points at which the segments hit the triangle be called D, D´, E, E´, F, F´ as shown above. As a result
of the lines being parallel, all three smaller triangles and the larger triangle are similar (ABC ~ DPD´ ~
PEE´ ~ F´PF). The remaining three sections are parallelograms.
Since PDAF´ is a parallelogram, we find PF´ = AD, and similarly PE = BD´.
So d = PF´ + PE = AD + BD´ = 425 – DD´.
Thus DD´ = 425 – d. By the same logic, EE´ = 150 –d.
Since DPD´ ~ ABC, we have the proportion:
425 − d PD 510 6
=  PD = 510 – d = 510 – d
425 510 425 5
17
Doing the same with PEE´, we find that PE´ = 510 – d.
15
6 17  50 
Now, d = PD + PE´ = 510 – d + 510 – d  d   = 1020  d = 306 .
5 15  15 
Solution 3
Define the points the same as above.
Let [CE´ PF] = a, [E´EP] = b, [BEPD´] = c, [D´PD] = d, [DAF´P] = e and [F´D´P] = f.
The key theorem we apply here is that the ratio of the areas of 2 similar triangles is the ratio of a pair of
corresponding sides squared.

Let the length of the segment be x and the area of the triangle be A, using the theorem, we get:
2 2 2
c+e+d  x  b+c+d  x  a +b+f  x 
=  ,= =  , =  adding all these together and using
A  BC  A  AC  A  AB 
f +d+b  1 1 1 
a + b + c + d + e + f = A we get + 1 = x2 *  + + 
 BC AB2 
2 2
A AC
Using corresponding angles from parallel lines, it is easy to show that ABC F´PF, since ADPF´ and
CFPE´ are parallelograms, it is easy to show that FF´ = AC – x.
f  AC − x  
2 2
x 
Now we have the side length ratio, so we have the area ratio =  = 1 −  , by symmetry, we
A  AC   AC 
2 2
d  x  b  x 
have = 1 −  and = 1 − 
A  AB  A  BC 
 x  x2
 (1 − 2) * AB = 0
x
Substituting these into our initial equation, we have 1 +  1 − AB  –
cyc AB2
=0 1+
cyc

2
= x answer follows after some hideous computation.
1 1 1
+ +
AB BC CA

Problem-15
Let triangle ABC be a right triangle in the xy-plane with a right angle at C. Given that the length of the
hypotenuse AB is `60, and that the medians through A and B lie along the lines y = x + 3 and y = 2x + 4
respectively, find the area of triangle ABC. [1986]
Solution
Translate so the medians are y = x, and y = 2x, then model the points A : (a, a) and B : (b, 2b). B : (b, 2b). (0,
0) is the centroid, and is the average of the vertices, so C : (– a – b, – a – 2b)
AB = 60 so
3600 = (a – b)2 + (2b – a)2
3600 = 2a2 + 5b2 – 6ab (1)
AC and BC are perpendicular, so the product of their slopes is -1, giving
 2a + 2b   a + 4b 
  = – 1
 2a + b   a + 2b 

50
 15
2a2 + 5b2 = – ab (2)
2
800
Combining (1) and (2), we get ab = –
3
Using the determinant product for area of a triangle (this simplifies nicely, add columns 1 and 2, add rows 2
3
and 3), the area is ab , so we get the answer to be 400.
2

Problem-9
Triangle ABC has right angle at B, and contains a point P for which PA = 10, PB = 6, and APB = BPC =
CPA. Find PC. [1987]
A

P
C
B
Solution
Let PC = x. Since APB = BPC = CPA, each of them is equal to 120º. By the Law of Cosines applied to
1
triangles APB, BPC and CPA at their respective angles P, remembering that cos 120º = – , we have
2
AB2 = 369 +100 + 60 = 196, BC2 = 36 + x2 + 6x, CA2 = 100 + x2 + 10x
Then by the Pythagorean Theorem, AB2 + BC2 = CA2 , so
x2 + 10x + 100 = x2 + 6x + 36 + 196 and 4x = 132  x = 033

Problem-7
In triangle ABC, tan CAB = 22/7, and the altitude from A divides BC into segments of length 3 and 17.
What is the area of triangle ABC ? [1988]
Solution
A

C 3 D 17 B
Let D be the intersection of the altitude with BC , and h be the length of the altitude. Without loss of
17 3
generality, let BD = 17 and CD = 3. Then tan DAB = and tan CAD = .
6 h
Using the tangent sum formula,
tan CAB = tan (DAB + CAD)
22 tan DAB + tan CAD
=
7 1 − tan DAB.tan CAD
17 3
+
h h , 22 20h
= = 2
 17  3 7 h − 51
1−  .
 h h
2
0 = 22h – 140h – 22.51
0 = (11h + 51)(h – 11)

51
 1
The positive value of h = 11, so the area is (17 + 3). 11 = 110.
2

Problem-12
Let P be an interior point of triangle ABC and extend lines from the vertices through P to the opposite sides.
Let a, b, c and d denote the lengths of the segments indicated in the figure. Find the product abc if a + b+ c =
43 and d = 3. [1988]

d d

a b
d P
A B

Solution
Solution 1
Call the cevians AD, BE, and CF. Using area ratios (PBC AND ABC have the same base), we have:
d [PBC]
=
a + d [ABC]
Similarily,
d [PCA] d [PAB]
= and =
b + d [ABC] c + d [ABC]
d d d [PBC] [PCA] [PAB] [ABC]
Then, + + = + + = =1
a +d b+d c + d [ABC] [ABC] [ABC] [ABC]
d d d
The identity + + = 1 is a form of Ceva's Theorem.
a +d b+d c+d
Plugging in d = 3, we get
a 3 3
+ + =1
a +3 b+3 c+3
3 [(a + 3)(b + 3) + (b + 3)(c + 3) + (c + 3)(a + 3)] = (a + 3)(b + 3)(c + 3)
3(ab + bc + ca) + 18(a + b + c) + 81 = abc + 3(ab + bc + ca) + 9(a + b +c) + 27
9(a + b + c) + 54 = abc = 441 .

Solution 2
Let the labels A, B, C be the weights of the vertices. First off, replace d with 3. We see that the weights of
the feet of the cevians are A + B, B + C, C + A. By mass points, we have that:

a B+C b C+A c A+B

= = =
3 A 3 B 3 C
If we add the equations together, we get
a + b + c A 2 B + A 2 C + B2 A + B2 C + C 2 A + C 2 B 43
= =
3 ABC 3
If we multiply them together, we get
abc A 2 B + A 2C + B2 A + B2C + C2 A + C2 B + 2ABC 43 49
= = +2=
27 ABC 3 3
49
Multiplying both sides by 27, we get that abc = 27. = 441 .
3

52
Problem-10
Let a, b, c be the three sides of a triangle, and let be the angles opposite them. If a2 + b2 = 1989c2,
cot 
find . [1989]
cot  + cot 
Solution
Solution 1
c
We can draw the altitude h to c, to get two right triangles. cot  + cot  = , from the definition of the
h
2
2A c
cotangent. From the definition of area, h = , so cot + cot = .
c 2A
Now we evaluate the numerator:
cos 
cot  =
sin 
From the Law of Cosines and the sine area formula,
1988c2
cos  =
2ab
2A
sin =
ab
cos  1988c2
cot  = =
sin  4A
1988c 2
cot  1988
Then = 4A = = 994 .
cot  + cot 
2
c 2
2A
Solution 2
cos  cos  sin  cos  + cos  sin 
cot + cot = + =
sin  sin  sin  sin 
sin( +  ) sin 
= =
sin  sin  sin  sin 
By the Law of Cosines,
a2 + b2 – 2ab cos= c2 = 1989c2 – 2ab cos ab cos= 994c2 
cot  cot  sin  sin  cos  sin  sin  ab ab 994c2
Now = = = cos  = . = 994 .
cot  cot  sin  sin 2  c2 c2 ab

Solution 3
994c2
Use Law of cosines to give us c2 = a2 + b2 – 2ab cos () or therefore cos() = .
ab
csin(a)
Next, we are going to put all the sin's in term of sin (a). We get sin() = .
a
994c
Therefore, we get cot() = .
bsin a
Next, use Law of Cosines to give us b2 = a2 + c2 – 2ac cos().
a 2 − 994c 2
Therefore, cos() = .
ac
bsin(a)
Also, sin() = .
a
a 2 − 994c 2
Hence, cot() = .
bc sin(a)

53
 b 2 − 994c2
Lastly, cos() =
bc
b 2 − 994c 2
Therefore, we get cot() = .
bc sin(a)
994c
cot( ) b sin a
Now, = 2 .
cot(  ) cot( ) a − 994c 2 + b 2 − 994c 2
bc sin(a)
994c*bcsin a
After using a2 + b2 = 1989c2 , we get = 994 .
c2 sin a

Problem-15
Point P is inside ABC. Line segments APD, BPE, and CPF are drawn with D on BC, E on AC, and F on
AB (see the figure below). Given that AP = 6, BP = 9, PD=6, PE=3, and CF=20, find the area of ABC.
[1989]

D
E
P

A B
F
Solution
Solution 1
Because we're given three concurrent cevians and their lengths, it seems very tempting to apply Mass points.
We immediately see that wE = 3, wB = 1, and wA = wD =2. Now, we recall that the masses on the three sides
of the triangle must be balanced out, so wC = 1 and wF = 3.
Thus, CP = 15 and PF = 5.
Recalling that wC = wB = 1, we see that DC = DB and DP is a median to BC in BCP. Applying Stewart's
Theorem, BC2 + 122 = 2(152 + 92), and BC = 6 13 . Now notice that 2[CBCP] = [ABC], because both
triangles share the same base and the hABC = 2h BCP.
Applying Heron's formula on triangle BCP with sides 15, 9 and 6 13 , [BCP] = 54 and [ABC] = 108 .
Solution 2
Using a different form of Ceva's Theorem, we have
y 6 3 y 1
+ + =1 =
x+ y 6+6 3+9 x+y 4
Solving 4y = x + y and x + y = 20, we obtain x = BP = 15 and y = FP = 5.
Let Q be the point on AB such that FC||QD.
Since AP = PD and FP||QD, QD = 2FP = 10. (Midline Theorem)
FC
Also, since FC||QD and QD = , we see that FQ = QB, BD = DC, etc. (Midline Theorem)
2
1 1
Similarly, we have PR = RB ( = PB = 7.5 ) and thus RD = PC = 4.5.
2 2

54
PDR is a 3 – 4 – 5 right triangle, so  PDR (ADQ) is 90º.
1
Therefore, the area  ADQ= .12.6 = 36
2
3
Using area ratio, ABC = ADB × 2 = (ADQ × ) × 2 = 36.3 = 108
2

Solution 3
Because the length of cevian BE is unknown, we can examine what happens when we extend it or decrease
its length and see that it simply changes the angles between the cevians. Wouldn't it be great if it the length
of BE was such that APC = 90º ? Let's first assume it's a right angle and hope that everything works out.
Extend AD to Q so that PD = DQ = 6. The result is that BQ = 9, PQ = 12, and BP = 15 because CDP  
BDQ.
Now we see that if we are able to show that BE = 20, that is PE = 5, then our right angle assumption will be
true.
Apply the Pythagorean Theorem on APC to get AC = 3 13 , so AE = 13 and CE = 2 13 .
Now, we apply the Law of Cosines on triangles CEP and AEP.
Let PE = x. Notice that CEB = 180º – ABE and cos CEB = – cos AEB, so we get two nice equations.
81 = 52 + y2 – 2y 13 cos CEF 36 = 13 + y2 + y 13 cos CEF
Solving, y = 5 (yay!).
1 1
Now, the area is easy to find. [ABC] = [AQB] + [APC] = (9)(18) + (6)(9) = 108
2 2

Problem-7
A triangle has vertices P – (– 8, 5), Q = (–15, – 19) and R = (1, –7). The equation of the bisector of  P can
be written in the form ax + 2y + c = 0. Find a + c. [1990]
P

Q
Solution
Use the distance formula to determine the lengths of each of the sides of the triangle. We find that it has
lengths of side 15, 20, 25 indicating that it is a 3 – 4 – 5 right triangle. At this point, we just need to find
another point that lies on the bisector of .

55
Solution 1

P´

Use the angle bisector theorem to find that the angle bisector of  P divides QR into segments of length
25 15 25 15
=  x= , .
x 20 − X 2 2
QP ' 5  5xR + 3xQ 5yR + 3yQ 
It follows that = , and so P´ =  ,  = (– 5, –23/2).
RP ' 3  8 8 
−11
The desired answer is the equation of the line PP´. PP ´ has slope , from which we find the equation to
2
be 11x + 2y + 78 = 0. Therefore, a + c = 089 .

56
Solution 2

P
A

R
P
A

Extend PR to a point S such that PS = 25. This forms an isosceles triangle PQS. The coordinates of S, using
the slope of PR (which is –4/3), can be determined to be (7, – 15). Since the angle bisector of P must touch
the midpoint of QS  (–4, – 17) we have found our two points. We reach the same answer of 11x + 2y + 78
= 0.
Solution 3

P
A

R
P
A
P´

By the angle bisector theorem as in solution 1, we find that QP´ = 25/2. If we draw the right triangle formed
by Q, P´, and the point directly to the right of Q and below P´, we get another 3 – 4 – 5 (since the slope of
QR is 3/4). Using this, we find that the horizontal projection of QP´ is and the vertical projection of QP´ is
15/2.
 15   23 
Thus, the angle bisector touches QR at the point  −15 + 10, −19 +  =  −5, −  , from where we continue
 2  2
with the first solution.

57
Problem-2
Rectangle ABCD has sides AB of length 4 and CB of length 3. Divide AB into 168 congruent segments
with points A = P0, P1, …. . P168 = B, and divide CB into 168 congruent segments with points C = Q0, Q1,
… , Q168 = B. For 1 ≤ k ≤ 167, draw the segments Pk Q k .

Repeat this construction on the sides AD and CD , and then draw the diagonal AC . Find the sum of the
lengths of the 335 parallel segments drawn. [1991]
Solution

D C
Q1
1Q2
1


3
4
Q166
1Q167
4
A … 1
P1 P2 P166 P167 B
4
1 1
The length of the diagonal is 32 + 4 2 = 5 (a 3–4–5 right triangle). For each k, Pk Q k is the hypotenuse of a
k
3 – 4 – 5 right triangle with sides of 3. , 4.
168
  . 18
 6 ( ) ( ) 2  BC a + b 1 81 − 1 81x − 1 40.41 a +b
( )
2 2 2 2

( )
2 1 2 29 1600 167

− (81 − 1). . 81 − 81x 1+ k +5

2 2
3 3 . 18 3 sin ( )
2 3  c 421 4 2 2 9 81 (a − 9) + b
2 2
k =1

k
. Thus, its length is 5. .
168
 167 k  5 167
The sum we are looking for is 2.   5.  +5= k +5.
 k =1 168  84 k =1

5 168 + 167
Using the formula for the sum of the first natural numbers, we find that the solution is . +6+
84 2
5.167 + 5 = 840 .

Problem-13
Triangle ABC has AB = 9 and BC : AC = 40 : 41. What's the largest area that this triangle can have?
[1992]
Solution
Solution 1
First, consider the triangle in a coordinate system with vertices at (0, 0), (9, 0) and (a, b). Applying the
a 2 + b2 40
distance formula, we see that = .
(a − 9) + b
2 2 41

58
We want to maximize b, the height, with 9 being the base.
3200
Simplifying gives – a2 – a 1600 = b2 .
9
To maximize b, we want to maximize b2. So if we can write: b2 = – (a + n)2 + m, then m is the maximum
value of b2 (this follows directly from the trivial inequality, because if x2 ≥ 0 then plugging in a + n for x
gives us (a + n)2 ≥ 0).
3200 1600 2 1600 2
b2 = – a2 – a + 1600 = – (a + ) + 1600 + ( ).
9 9 9
2
 1600  1600 40 40.41
b≤ 1600 +   = 40 1 + = 1681 = .
 9  81 9 9
1 40.41
Then the area is 9. . = 820 .
2 9
Solution 2
Let the three sides be 9, 40x, 41x, so the area is (
1
4
)
812 − 81x 2 (81x 2 − 1) by Heron's formula. By AM-GM,

81x − 1
2

(812 − 81x 2 ) (81x 2 −1) ≤ 2 , and the maximum possible area is

1 812 − 1 1
. = (81 − 1) (81 + 1) = 10.82 = 820 .
4 2 8
4 205
This occurs when 812 – 81x2 = 81x2 – 1  x = .
9

Problem-14
In triangle ABC, A´, B´, and C´ are on the sides BC, AC, and AB, respectively. Given that AA´, BB´, and
AO BO CO AO BO CO
CC´ are concurrent at the point O, and that + + = 92, find . . . [1992]
OA' OB' OC' OA' OB' OC'
Solution
Using mass points, let the weights of A, B and C be a, b, and c respectively.
Then, the weights of A´, B´ , and C´ are b + c, c + a, and a + b respectively.
AO b + c BO c + a CO a + b
Thus, = , = , , and = .
OA' a OB' b OC' c
AO BO CO b+c c+a a+b
Therefore: . . = . .
OA' OB' OC' a b c
2abc + b c + bc + c a + ca + a b + ab
2 2 2 2 2 2
bc(b + c) ca(c + a) ab(a + b) b+c c+a
= = 2 + + + = 2 + + +
abc abc abc abc a b
a+b
c
AO BO CO
=2+ + + = 2 + 92 = 094 .
OA' OB' OC'

Problem-10

In triangle ABC, angle C is a right angle and the altitude from C meets AB at D. The lengths of the sides of
ABC are integers, BD = 293 and cosB = m/n, where m and n are relatively prime positive integers. Find m
+ n. [1994]

59
Solution

BC 293
Since ABC ~  CAB, we have =  BC2 = 293 AB. It follows that 292 | BC and 29 | B, so BC
AB BC
and AB are in the form 292 a and 29a2, respectively.
By the Pythagorean Theorem, we find that AC2 + BC2 = AB2  (29a2)2 + AC2 = (29a2)2, so 29a | AC.
Letting b = AC/29a, we obtain after dividing through by (29a2), 292 = a2 – b2 = (a – b)(a + b).
AC
As a, b  Z, the pairs of factors of 292 are (1, 292)(29, 29); clearly b =  0, so a –b = 1, a + b = 292.
29a
1 + 292
Then, a = = 421.
2
BC 292 29
Thus, cos B = = 2
= , and m + n = 450 .
AB 29a 421

Problem-15
Given a point P on a triangular piece of paper ABC consider the creases that are formed in the paper when
A, B, and C are folded onto P. Let us call P a fold point of  ABC if these creases, which number three
unless P is one of the vertices, do not intersect. Suppose that AB = 36, AC = 72, and  B = 90º. Then the

area of the set of all fold points of  ABC can be written in the form q  – r s , where q, r and s are
positive integers and s is not divisible by the square of any prime. What is q + r + s ? [1994]

Solution
Let OAB be the intersection of the perpendicular bisectors (in other words, the intersections of the creases) of
PA and PB , and so forth. Then OAB, OBC, OCA are, respectively, the circumcenters of PAB, PBC, PCA.
According to the problem statement, the circumcenters of the triangles cannot lie within the interior of the
respective triangles, since they are not on the paper. It follows that  APB,  BPC,  CPA > 90º; the locus

of each of the respective conditions for P is the region inside the (semi)circles with diameters AB , BC , CA .
We note that the circle with diameter AC covers the entire triangle because it is the circumcircle of  ABC,
so it suffices to take the intersection of the circles about AB, BC. We note that their intersection lies entirely
within  ABC (the chord connecting the endpoints of the region is in fact the altitude of  ABC from B.
Thus, the area of the locus of P (shaded region below) is simply the sum of two segments of the circles. If

we construct the midpoints of M1 , M2 = AB , BC and note that  M1 BM2 ~  ABC, we see that these

segments respectively cut a 120º arc in the circle with radius 18 and 60º arc in the circle with radius 18 3 .

60
 C
A

P
A

B A

A A
The diagram shows P outside of the grayed locus; notice that the creases [the dotted blue] intersect within
the triangle, which is against the problem conditions. The area of the locus is the sum of two segments of
two circles; these segments cut out 120º, 60º angles by simple similarity relations and angle-chasing.
1
Hence, the answer is, using the C ab sin C definition of triangle area,
2
 2    2 
1 2
( ) 1
( )
2 2

 3 .18 2
− .18 sin  +  . 18 3 − . 18 3 sin = 270 – 3 , and q + r + s = 597 .
2 3  6 2 3 

15 - INEQUALITY

Problem-10
Mary told John her score on the American High School Mathematics Examination (AHSME), which was
over 80. From this, John was able to determine the number of problems Mary solved correctly. If Mary's
score had been any lower, but still over 80, John could not have determined this. What was Mary's score?
(Recall that the AHSME consists of 30 multiple choice problems and that one's score, s, is computed by the
formula s = 30 + 4c – w, where c is the number of correct answers and w is the number of wrong answers.
(Students are not penalized for problems left unanswered.) [1984]
Solution
Let Mary's score, number correct, and number wrong be s, c, w respectively. Then
s = 30 + 4c – w = 30 + 4(c – 1) – (w – 4) = 30 + 4(c + 1) – (w + 4).
Therefore, Mary could not have left at least five blank; otherwise, 1 more correct and 4 more wrong would
produce the same score. Similarly, Mary could not have answered at least four wrong (clearly Mary
answered at least one right to have score above 80, or even 30.)
It follows that c +w ≥ 25 and w ≤ 3, so c ≥ 23 and s = 30 + 4c – w ≥ 30 + 4(23) – 3= 119.
So Mary scored at least 119. To see that no result other than 23 right/3 wrong produces 119, note that s = 119
 4c – w = 89 so w  3 (mod 4). But if w = 3, then c = 23, which was the result given; otherwise w ≥ 7 and
c ≥ 24, but this implies at least 31 questions, a contradiction. This makes the minimum score 119 .

61
Problem-3
A tennis player computes her win ratio by dividing the number of matches she has won by the total number
of matches she has played. At the start of a weekend, her win ratio is exactly .500. During the weekend, she
plays four matches, winning three and losing one. At the end of the weekend, her win ratio is greater than
.503. What's the largest number of matches she could've won before the weekend began ? [1992]
Solution
n 1 n +3 503
Let n be the number of matches won, so that = , and > . Cross multiplying, 100n + 3000
2n 2 2n + 4 1000
988
> 106n + 2012, and n < . Thus, the answer is 164 .
6

16 - TRIGNOMETRY

Problem-13
Find the value of 10cot(cot–13 + cot–17 + cot–113 + cot–121). [1984]
Solution
Solution 1
We know that tan(arctan(x)) = x so we can repeatedly apply the addition formula, tan (x + y) =
tan(x) + tan(y)
. Let a = cot–1(3), b = cot–1(7), c = cot–1(13), and d = cot–1(21).
1 − tan(x) tan(y)
1 1 1 1
We have, tan(a) = , tan(b) = , tan(c) = , tan(d) = ,
3 7 13 21
1 1
+ 1
So, tan(a + b) = 3 7 = and
1 2
1−
21
1 1
+
tan(c + d) = 13 21 = 1
1 8
1−
273
1 1
+
so, tan((a + b)) + (c + d) = 2 8 = 2.
1 3
1−
16
3
Thus our answer is 10. = 15.
2
Solution 2
xy − 1
Apply the formula cot–1x + cot–1y = cot–1 repeatedly. Using it twice on the inside, the desired sum
x+y
becomes cot(cot–1 2 + cot–1 8).
This sum can then be tackled by taking the cotangent of both sides of the inverse cotangent addition formula
shown at the beginning.
Solution 3
On the coordinate plane, let O = (0, 0), A1 = (3, 0), A2 = (3, 1), B1 = (21, 7), B2 = (20, 10), C1 = (260, 130),
C2 = (250, 150), D1 = (5250, 3150), D2 = (5100, 3400), and H = (5100, 0).
We see that cot–1( A2 OA1) = 3, cot–1( B2 OB1) = 7, cot–1( C2 OC1) = 13, and cot–1( D2 OD1) = 21.
5100 3
The sum of these four angles forms the angle of triangle OD2H, which has a cotangent of = ,
3400 2
3
which must mean that cot(cot–13 + cot–17 + cot–113 + cot–121) = .
2

62
 3
So the answer is 10 *   = 015 .
2
Solution 4
 b
Recall that cot–1  = – tan–1and that arg(a + bi) = tan–1 .
2 a
Then letting w = 1 + 3i, x = 1 + 7i, y = 1 + 13i andz = 1 + 21i, , we are left with
    
10 cot  − arg w + − arg x + − arg y + − arg z  = 10 cot (2 – arg wxyz)
2 2 2 2 
= – 10 cot(arg wxyz).
Expanding wxyz, we are left with
(1 + 3i) (1 + 13i) (1 + 7i) (1 + 21i) = (1 + 16i – 39) (1 + 28i – 147) = (16i – 38) (28i – 146) = 4(8i – 19) (14i
850 850 1275 3
– 73) = 4(1275 – 850i) – 10 cot tan–1 – = 10 cot tan–1 = 10 = 10 = 15.
1275 1275 850 2

Problem-3
If tan x + tan y = 25 and cot x + cot y = 30, what is tan(x + y) ? [1986]
Solution
Since cot is the reciprocal function of :
1 1 tax x + tan y
cot x + cot y = + = = 30
tan x tan y tan x.tan y
tax x + tan y 25 5
Thus, tan x . tan y = = =
30 30 6
Using the tangent addition formula:
tax x + tan y 25
tan (x + y) = = = 150.
1 − tan x.tan y 1 − 5
6
Problem-6
Two skaters, Allie and Billie, are at points A and B, respectively, on a flat, frozen lake. The distance between
A and B is 100 meters. Allie leaves A and skates at a speed of 8 meters per second on a straight line that
makes a 60º angle with AB. At the same time Allie leaves A, Billie leaves B at a speed of 7 meters per
second and follows the straight path that produces the earliest possible meeting of the two skaters, given
their speeds. How many meters does Allie skate before meeting Billie ? [1989]

60º
B
100
Solution
Label the point of intersection as C. Since d = rt, AC = 8 and BC = t. According to the law of cosines,

63
 8t 7t

60 º
A B
100
(7t)2 = (8t)2 + 1002 – 2.8t .100.cos 60 º
b
0 = 15t2 – 800t + 10000 = 3t2 – 160t + 2000 A
160  160 − 4.3.2000
2
100
t= = 20, .
6 3
Since we are looking for the earliest possible intersection, 20 seconds are needed. Thus, 8.20 = 160 meters
is the solution.

Problem-4
1
How many real numbers satisfy the equation log2 x = sin (5x) ? [1991]
5
Solution

y
A

0 1 2 3
x
A A A A
A

2
The range of the sine function is –1 ≤ y ≤ 1. It is periodic (in this problem) with a period of .
5
1 1
Thus, –1 ≤ log2 x ≤ 1, and –5 ≤ log2 x ≤ 5. The solutions for occur in the domain of ≤ x ≤ 32.
5 32
When x > 1 the logarithm function returns a positive value; up to x = 32 it will pass through the sine curve.
There are exactly 10 intersections of five periods (every two integral values of x) of the sine curve and
32
another curve that is < 1, so there are .10 – 6 = 160 – 6 = 154 values (the subtraction of 6 since all the
2
“intersections” when x < 1 must be disregarded). When y = 0, there is exactly touching point between the

64
 1 
two functions:  , 0  . When y < 0 or x < 1, we can count 4 more solutions. The solution is 154 + 1 + 4 =
5 
159 .

Problem-9
22 m m
Suppose that sec x + tan x = and that csc x + cot x = , where is in lowest terms. Find m + n.
7 n n
[1991]
Solution
Solution 1
Use the two trigonometric Pythagorean identities 1 + tan2 x = sec2 x and 1 + cot2 x = csc2 x.
22
If we square the given sec x = – tan x, we find that
7
2
 22   22 
sec x =   – 2   tan x + tan2 x
2
 7   7 
2
 22  44
1=   – tan x
 7  7
435
This yields tan x =
308
m
Let y = . Then squaring,
n
csc2 x = (y – cot x)2  1 = y2 – 2y cot x.
1 308
Substituting cot x = = yields a quadratic equation:
tan x 435
0 = 435y2 – 616y – 435 = (15y – 29)(29y + 15).
29
It turns out that only the positive root will work, so the value of y = and m + n = 044
15
Solution 2
Recall that sec2 x – tan2 x = 1, from which we find that sec x – tan x = 7/22. Adding the equations
sec x + tan x = 22/7
sec x – tan x = 7/22
together and dividing by 2 gives sec x = 533/308, and subtracting the equations and dividing by 2 gives tan x
= 435/3080. Hence, cos x = 308/533 and sin x = tan x cos x = (435/308)(308/533). Thus, csc x = 533/435
and cot x = 308/435. Finally,
841 29 29
csc x + cot x = = ,
435 15 15
so m + n = 044.

Solution 3 (least computation)

1 sin x 22 1 cos x
By the given, + = and + = k.
cos x cos x 7 sin x sin x
Multiplying the two, we have
1 1 1 22
+ + = k
sin x cos sin x cos x 7
Subtracting both of the two given equations from this, and simpliyfing with the identity
sin x cos x sin 2 x + cos 2 x 1
+ = = , we get
cos x sin x sin x cos x sin x cos x
22 22
1= k– – k.
7 7

65
 29
Solving yields k = , and m + n = 044
15
Solution 4
x x sin x
Make the substitution u = tan (a substitution commonly used in calculus). tan = , so
2 2 1 + cos x
1 + cos x 1 m 1 + sin x
csc x + cot x = = = . sec x + tan = ,
sin x u n cos x
Now note the following:
2u
sin x =
1+ u2
1− u2
cos x =
1+ u2
Plugging these into our equality gives:
2u
1+
1 + u 2 = 22
1− u2 7
1+ u 2

1 + u 22 15 m 29
This simplifies to = , and solving for u gives u = , and = .
1− u 7 29 n 15
Finally, m + n = 044.

Solution 5
We are given that
1 + sin x 22 1 + sin x 1 − sin x 1 − sin 2 x cos 2 x
=  . = =
cos x 7 cos x 1 − sin x cos x(1 − sin x) cos x(1 − sin x)
cos x 1 + 7sin x 22 − 22sin x 222 − 7 2 2.22.7
= or equivalently, cos x = =  sin x = 2 2  cos x = 2 2 .
1 − sin x 22 7 22 + 7 22 + 7
Note that what we want is just
2.22.7
1+ 2
1 + cos x 22 + 7 2 = 22 + 7 + 2.22.7 =
2 2
(22 + 7) 2 22 + 7 29
= = =  m + n = 29 + 15 = 044 .
sin x 22 − 7
2 2
22 − 7
2 2
(22 − 7)(22 + 7) 22 − 7 15
222 + 7 2

17 – ALZEBRIC EQUATION

Problem-1
n
Let x1 = 97, and for n > 1, let x n = . Calculate the product x1x2x3x4x5x6x7x8. [1985]
x n −1
Solution
n
Since x n = . xn . xn–1 = n. Setting n = 2, 4, 6 and 8 in this equation gives us respectively
x n −1
x1x2 = 2 , x3x4 = 4, x5x6 = 6 and x7x8 = 8 so x1x2x3x4x5x6x7x8 = 2.4.6.8 = 384 .
Notice that the value of x1 was completely unneeded!

Problem-1
12
What is the sum of the solutions to the equation 4
x= ? [1986]
7− 4 x

66
Solution
Let y = 4 x .
Then we have y(7 – y) = 12, or, by simplifying, y2 – 7y + 12 = (y – 3)(y – 4) = 0.
4
This means that x = y = 3 or 4.
Thus the sum of the possible solutions for x is 44 + 34 = 337.

Problem-4
Determine 3x4 + 2x5 if x1, x2, x3, x4, and x satisfy the system of equations below. [1986]
2x1 + x2 + x3 + x4 + x5 = 6
x1 + 2x2 + x3 + x4 + x5 = 12
x1 + x2 + 2x3 + x4 + x5 = 24
x1 + x2 + x3 + 2x4 + x5 = 48
x1 + x2 + x3 + x4 + 2x5 = 96
Solution
Adding all five equations gives us 6(x1 + x2 + x3 + x4 + x5) = 6(1 + 2 + 4 + 8 + 16)
so x1 + x2 + x3 + x4 + x5 = 31. Subtracting this from the fourth given equation gives x4 = 17 and subtracting
it from the fifth given equation gives x5 = 65, so our answer is 3.17 + 2.65 = 181

Problem-4
x
Find the area of the region enclosed by the graph of |x – 60| + |y| = . [1987]
4
Solution

5 3
y = 4 × – 60 y = 4 × + 60

45 50 55 60 65 70 75 80 85
A A A A A A A A

x
Since |y| is nonnegative, ≥ |x – 60|. Solving this gives us two equations:
4
x x
≥ x – 60 and – ≤ – 60. Thus, 48 ≤ x ≤ 80.
4 4
The maximum and minimum y value is, when |x – 60| = 0, which is when x = 60 and y = 15. Since the
x
graph is symmetric about the y-axis, we just need casework upon x. > 0, so we break up the condition |x
4
– 60| :
3
• x – 60 > 0. Then y = x + 60
4
67
 5
• x – 6 < 0. Then y = x – 60
4
The area of the region enclosed by the graph is that of the quadrilateral defined by the points (48, 0), (60,
15), (80, 0), (60 –15).
1
Breaking it up into triangles and solving, we get 2. (80 – 48)(15) = 480 .
2

Problem-8
Assume that x1 , x2 , …. , x7 are real numbers such that
x1 + 4x2 + 9x3 + 16x4 + 25x5 + 36x16 + 49x7 = 1
4x1 + 9x2 + 16x3 + 25x4 + 36x5 + 49x16 + 64x7 = 12
9x1 + 16x2 + 25x3 + 36x4 + 49x5 + 64x16 + 81x7 = 123.
Find the value of 16x1 + 25x2 + 36x3 + 49x4 + 64x5 + 81x16 + 100x7 . [1989]
Solution
Solution 1
Notice that because we are given a system of 3 equations with 7 unknowns, the values (x1, x2, … , x7 ) are
not fixed; indeed one can take any four of the variables and assign them arbitrary values, which will in turn
fix the last three.
Given this, we suspect there is a way to derive the last expression as a linear combination of the three given
expressions.
Let the coefficent of xi in the first equation be y2i ; then its coefficients in the second equation is (yi + 1)2
and the third as (yi + 2)2. We need to find a way to sum these to make (yi + 3)2 [this is in fact a specific
approach generalized by the next solution below].
Thus, we hope to find constants a, b, c satisfying ay2i + b(yi + 1)2 + c(yi + 2)2 = (yi + 3)2.
FOILing out all of the terms, we get
[ay2 + by2 + cy2] + [2by + 4cy] + b + 4c = y2 + 6y + 9.

Comparing coefficents gives us the three equation system:

a +b+c=1
2b + 4c = 6
b + 4c = 9
Subtracting the second and third equations yields that b = – 3, so c = 3 and a = 1. It follows that the desired
expression is a.(1) + b.(12) + c.(123) = 1 – 36 + 369 = 334 .

Solution 2
Notice that we may rewrite the equations in the more compact form as:
7 7 7 7

 i2 x i = c1 ,
i =1
 (i + 1)2 x i = c2 ,
i =1
 (i + 2)2 x i = c3 ,
i =1
and  (i + 3)
i =1
2
x i = c4

where c1 = 1, c2 = 12, c3k = 123 and c4 is what we're trying to find.

7
Now consider the polynomial given by f (z) : =  (z + i)
i =1
2
x i (we are only treating the xi as coefficients).

Notice that f is in fact a quadratic. We are given f(0), f(1), f(2) as c1, c2 , c3 and are asked to find c4 . Using
the concept of finite differences (a prototype of differentiation) we find that the second differences of
consecutive values is constant, so that by arithmetic operations we find c4 = 334.
Alternatively, applying finite differences, one obtains
 3  3  3
c4 =   f (2) –   f (1) +   f (0) = 334
 2 1  0

Solution 3
Notice that 3(n + 2)2 – 3(n + 1)2 + n2 = (n + 3)2
I'll number the equations for convenience
x1+ 4x2 + 9x3 + 16x4 + 25x5 + 36x6 + 49x7 = 1 (1)
4x1+ 9x2 + 16x3 + 25x4 + 36x5 + 49x6 + 64x7 = 12 (2)

68
9x1+ 16x2 + 25x3 + 36x4 + 49x5 + 64x6 + 81x7 = 123 (3)
16x1+ 25x2 + 36x3 + 49x4 + 64x5 + 81x6 + 100x7 = 1 (4)

Let the coefficient of xi in (1) be n2. Then the coefficient of xi in (2) is (n + 1)2 etc.
Therefore, 3* (3) – 3 * (2) + (1) = (4)
So (4) = 3 * 123 – 3 * 12 + 1 = 334 .

Solution 4
Notice subracting the first equation from the second yields 3x1 + 5x2 + .. + 15x7 = 11 .
Then, repeating for the 2nd and 3rd equations, and then subtracting the result from the first obtained
equation, we get 2x1 + 2x2 + … + 2x7 = 100. Adding this twice to the first obtained equation gives difference
of the desired equation and 3rd equation, which is 211. Adding to the 3rd equation, we get 334 .

Problem-15
Find ax5 + by5 if the real numbers a, b, x and y satisfy the equations ax + by = 3, ax2 + by2 = 7, ax3 + by3 =
16, ax4 + by4 = 42. [1990]
Solution
Solution 1
Set S = (x + y) and P = xy. Then the relationship
(axn + byn)(x + y) = (axn + 1 + by n + 1) + (xy)(axn – 1 + by n – 1)
can be exploited:
(ax2 + by2)(x + y) = (ax3 + by3) + (xy) (ax + by)
(ax3 + by3)(x + y) = (ax4 + by4) + (xy) (ax2 + by2)
Therefore: 7S = 16 + 3P
16S = 42 + 7P
Consequently, S = – 14 and P = – 38.
Finally:
(ax4 + by4)(x + y) = (ax5 + by5) + (xy)(ax3 + by3)
(42)(S) = (ax5 + by5) + (P)(16)
(42)(– 14) = (ax5 + by5) + (– 38)(16)
ax5 + by5 = 20 .
Solution 2
A recurrence of the form Tn = ATn – 1 + BTn – 2 will have the closed form Tn = axn + by2 , where x, y are the
values of the starting term that make the sequence geometric, and a, b are the appropriately chosen constants
such that those special starting terms linearly combine to form the actual starting terms.
Suppose we have such a recurrence with T1 = 3 and T2 = 7.
Then T3 = ax3 + by3 = 16 = 7A + B, and T4 = ax4 + by4 = 42 = 16A + 7B.
Solving these simultaneous equations for A and B, we see that A = – 14 and B = 38.
So, ax5 + by5 = T5 = – 14(42) + 38(16) = 20 .

Problem-15
n
For positive integer n, define Sn to be the minimum value of the sum 
k =1
(2k − 1) 2 + a 2 k ,

where a1, a2, …. , an are positive real numbers whose sum is 17. There is a unique positive integer n for
which Sn is also an integer. Find this n. [1991]
Solution
Solution 1
Interpret the problem geometrically. Consider right triangles joined at their vertices, with bases a1, a2, …..
an and heights 1, 3, …. 2n – 1. The sum of their hypotenuses is the value of Sn. The minimum value of Sn,
then, is the length of the straight line connecting the bottom vertex of the first right triangle and the top
2 2
 n   n 
vertex of the last right triangle, so Sn ≥  (2k − 1)  +   ak 
 k =1   k =1 
Since the sum of the first n odd integers is n2 and the sum of a1 , a2 , … , an is 17, we get Sn ≥ 17 2 + n 4 .

69
If this is integer, we can write 172 + n4 = m2, for an integer m. Thus, (m – n2)(m + n2) = 289.1 = 17.17 =
1.289. The only possible value, then, for m is 145, in which case n2 = 144, and n = 012 .
Solution 2
The inequality
2 2
 n   n 
Sn ≥   (2k − 1)  +   a k  is a direct result of the Minkowski Inequality. Continue as above.
 k =1   k =1 
Solution 3

Let ai = (2i – 1) tan I for 1 ≤ i ≤ n and 0 ≤ i < .
2
n n
We then have that Sn = 
k =1
(2k − 1) 2 + a 2 k =  (2k − 1)
k =1
2
sec  k

12 12 
  50
22
Sn + 17 n2
Note that that Sn + 17 =  .
n2 Sn − 17
2k − 1
n

 sec
k =1 k + tan  k

Note that for any angle , it is true that sec  + tan  and sec  – tan  are reciprocals.
n n
2k − 1
We thus have that Sn – 17 =  (2k − 1)(sec  k − tan  k ) =  .
k =1 k =1 sec  k + tan  k

By the AM-HM inequality on these n2 values, we have that:

Sn + 17 n2
  S2n ≥ 289 + n4
n2 Sn − 17
This is thus the minimum value, with equality when all the tangents are equal. The only value for which
289 + n 4 is an integer is n = 12 (see above solutions for details).

Problem-7
For certain ordered pairs (a, b) of real numbers, the system of equations
ax + by = 1
x2+ y2 = 50
has at least one solution, and each solution is an ordered pair (x, y) of integers. How many such ordered
pairs (a, b) are there ? [1994]
Solution
x2 + y2 = 50 is the equation of a circle of radius 50 , centered at the origin. The lattice points on this circle
are (±1,±7), (±5,±5) and (±7, ±1).
ax + by = 1 is the equation of a line that does not pass through the origin. (Since (x, y) = (0, 0) yields a(0) +
b(0) = 0  1).
So, we are looking for the number of lines which pass through either one or two of the 12 lattice points on
the circle, but do not pass through the origin.
It is clear that if a line passes through two opposite points, then it passes through the origin, and if a line
passes through two non-opposite points, the it does not pass through the origin.
 12 
There are   = 66 ways to pick two distinct lattice points, and thus 66 distinct lines which pass through
2
12
two lattice points on the circle. However, = 6 of these lines pass through the origin.
2
Since there is a unique tangent line to the circle at each of these lattice points, there are 12 distinct lines
which pass through exactly one lattice point on the circle.

70
Thus, there are a total of 66 – 6 + 12 = 072 distinct lines which pass through either one or two of the 12
lattice points on the circle, but do not pass through the origin.

18 – SURFACE AREA & VOLUME

Problem-2
When a right triangle is rotated about one leg, the volume of the cone produced is 800cm3. When the
triangle is rotated about the other leg, the volume of the cone produced is 1920cm3. What is the length (in
cm) of the hypotenuse of the triangle? [1985]
Solution
Let one leg of the triangle have length and let the other leg have length b. When we rotate around the leg
1
of length a, the result is a cone of height a and radius b, and so of volume ab2 = 800. Likewise, when
3
1
we rotate around the leg of length b we get a cone of height b and radius a and so of volume ba2 =
3
1
a 3
 ba 2 1920 12 12
1920. If we divide this equation by the previous one, we get = = = , so a = b .
b 1  ab2 800 5 5
3
1  12 
Then   b  b 2 = 800so b3 = 1000 and b = 10 so a = 24. Then by the Pythagorean Theorem, the
3  5 
hypotenuse has length a 2 + b 2 = 026 .

Problem-11
Ninety-four bricks, each measuring 4´´ × 10´´ × 19´´ are to stacked one on top of another to form a tower 94
bricks tall. Each brick can be oriented so it contributes 4´´or 10´´ or 19´´ to the total height of the tower.
How many different tower heights can be achieved using all ninety-four of the bricks? [1994]
Solution
Solution 1
We have the smallest stack, which has a height of 94 × 4 inches. Now when we change the height of one of
the bricks, we either add 0 inches, 6 inches, or 15 inches to the height. Now all we need to do is to find the
different change values we can get from 940's, 6's, and 's.
Because 0, 6, and 15 are all multiples of 3, the change will always be a multiple of 3, so we just need to find
the number of changes we can get from 0's, 2's, and 5's.

From here, we count what we can get:

0, 2 = 2, 4 = 2 + 2, 5 = 5, 6 = 2 + 2 + 2, 7 = 5 + 2, 8 = 2 + 2 + 2 + 2, 9 = 5 + 2 + 2, ….
It seems we can get every integer greater or equal to four; we can easily deduce this by considering parity or
using the Chicken McNugget Theorem, which says that the greatest number that cannot be expressed in the
form of 2m + 5n for m, n being positive integers is 5 × 2 – 5 – 2 = 3.
But we also have a maximum change (94 × 5), so that will have to stop somewhere. To find the gaps, we can
work backwards as well. From the maximum change, we can subtract either 0's, 3's, or 5's. The maximum
we can't get is 5 × 3 – 5 – 3 = 7, so the numbers 94 × 5 – 8 and below, except 3 and 1, work. Now there
might be ones that we haven't counted yet, so we check all numbers between 94 × 5 – 8 and 94 × 5.94 × 5 –
7 obviously doesn't work, 94 × 5 – 6 does since 6 is a multiple of 3, 94 × 5 – 5 does because it is a multiple
of 5 (and 3), 94 × 5 – 4 doesn't since 4 is not divisible by 5 or , 94 × 5 – 3 does since 3 = 3, and 94 × 5 – 2
and 94 × 5 – 1 don't, and 94 × 5 does.
Thus the numbers 0, 2, 4 all the way to 94 × 5 – 8, 94 × 5 – 6, 94 × 5 – 5, 94 × 5 – 3, and 94 × 5 work. That's
2 + (94 × 5 – 8 – 4 + 1) + 4 = 465 numbers. That's the number of changes you can make to a stack of bricks
with dimensions 4 × 10 × 14, including not changing it at all.

71
Solution 2
Using bricks of dimensions 4´´ × 10´´ × 19´´ is comparable to using bricks of dimensions 0´´ × 6´´ × 15´´
which is comparable to using bricks of dimensions 0´´× 2´´× 5´´. Using 5 bricks of height 2´´ can be
replaced by using 2 bricks of height 5´´ and 3 bricks of height 0´´.
It follows that all tower heights can be made by using 4 or fewer bricks of height 2´´. There are 95 + 94 + 93
+ 92 + 91 = 465 ways to build a tower using 4 or fewer bricks of height 2´´.
Taking the heights mod 5, we see that towers using a different number of bricks of height 2´´ have unequal
heights. Thus, all of the 465 tower heights are different.

19 – PERIMETER & AREA

Problem-4
A small square is constructed inside a square of area 1 by dividing each side of the unit square into n equal
parts, and then connecting the vertices to the division points closest to the opposite vertices. Find the value
1
of n if the area of the small square is exactly . [1985]
1985

A 1/n B
A A A
1
A

D C
A A
Solution
Solution 1
The lines passing through A and C divide the square into three parts, two right triangles and a parallelogram.
Using the smaller side of the parallelogram, 1/n, as the base, where the height is 1, we find that the area of
1
the parallelogram is A = . By the Pythagorean Theorem, the longer base of the parallelogram has length
n
 n −1 
2
1 A 1
= 1 +
2
 = 2n 2 − 2n + 1 , so the parallelogram has height h = = . But the
 n  n 2n 2 − 2n + 1
height of the parallelogram is the side of the little square, so 2n2 – 2n + 1 = 1985. Solving this quadratic
equation gives n = 032 .

72
Solution 2

A 1/n B
A A A
1
A
E G
F

D H C
A A
1 A
Surrounding the square with area are 4 right triangles with hypotenuse (sides of the large square).
1985
1
Thus, X + = 1, where X is the area of the of the 4 triangles. We can thus use proportions to solve this
1985
1 1
GF CG n 1985
problem. =  1985 = n  BE = .
BE CB BE 1 1985
BE EC 1985
Also, =  EC = (n − 1)
1 n −1 1985
n
1
Thus, 2(BE) (EC) + =1
1985
2n2 – 2n + 1 = 1985
n (n – 1) = 992
Simple factorization and guess and check gives us 032 .

Problem-12
Rhombus PQRS is inscribed in rectangle ABCD so that vertices P, Q, R, and S are interior points on sides
AB , BC , CD , and DA respectively. It is given that PB = 15, BQ = 20, PR = 30 and QS = 40. Let m/n, in
lowest terms, denote the perimeter of ABCD. Find m + n. [1991]

Solution

A P
15 B
b
A
S
A 20
O
A Q
A

D R C
A A A 73
Solution 1
Let O be the center of the rhombus. Via parallel sides and alternate interior angles, we see that the opposite
triangles are congruent ( BPQ   DRS,  APS   CRQ). Quickly we realize that O is also the
center of the rectangle.
By the Pythagorean Theorem, we can solve for a side of the rhombus; PQ = 152 + 202 = 25. Since the
diagonals of a rhombus are perpendicular bisectors, we have that OP = 15, OQ = 20.
Also,  POQ = 90º, so quadrilateral BPOQ is cyclic. By Ptolemy's Theorem, 25.OB = 20.15 + 15.20 = 600.
By similar logic, we have APOS is a cyclic quadrilateral. Let AP = x, AS = Y. The Pythagorean Theorem
gives us x2 + y2 = 625. (1). Ptolemy’s Theorem gives us 25.OA = 20x + 15y. Since the diagonals of a
1 4
rectangle are equal, OA = d = OB, and 20x + 15y = 600 (2). Solving for y, we get y = 40 – x.
2 3
Substituting into (1),
2
 4 
x2 +  40 − x  = 625
 3 
5x – 192x + 1755 = 0
2

192  192 2 − 4.5.1755 117

x= = 15,
10 5
We reject 15 because then everything degenerates into squares, but the condition that PR  QS gives us a
117 44
contradiction. Thus x = , and backwards solving gives y = . The perimeter of ABCD is 2
5 5
 117 44 
 20 + 15 + +  , and m + n = 677 .
 5 5 

Solution 2
From above, we have OB = 24 and BD = 48. Returning to BPQO, note that  PQO   PBP  ABD.
Hence,  ABC ~  OQP by AA similarity. From here, it's clear that
AB OP AD 15 144
=  =  AD = .
BD PQ 48 25 5
AB IQ AB 20 192
Similarly, =  =  AB = .
BD PQ 48 25 5
 192 192  672
Therefore, the perimeter of rectangle ABCD is 2 (AB + AD) = 2  +  = .
 5 5  5
Solution 3
The triangles QOB, OBC are isosceles, and similar (because they have  QOB =  OBC).
BQ OB
Hence =  OB2 = BC.BQ
OB BC
The length of OB could be found easily from the area of BPQ :
OB 2BP.BQ
BP . BQ = .PQ  OB =  OB = 24
2 PQ
44
OB2 = BC . BQ  242 = (20 + CQ).20  CQ =
5
2
 44  177
From the right triangle CRQ we have RC = 25 –    RC =
2
.
 5  5
We could have also defined a similar formula: OB2 = BP . BA, and then we found AP, the segment OB is
tangent to the circles with diameters AO, CO.
The perimeter is
44 + 177 672
2(PB + BQ + QC + CR) = 2(15 + 20 + ) =  m + n = 677.
5 5
Solution 4

74
For convenience, let  PQS = . Since the opposite triangles are congruent we have that  BQR = 3, and
a a
therefore  QRC = 3 – 90. Let RC = a, then we have sin (3 – 90) = , or – cos 3 = . Expanding
25 25
4
with the formula cos 3 = 4 cos3 – 3 cos, and since we have cos = , we can solve for a. The rest
5
then follows similarly from above.

Solution 5
We can just find coordinates of the points. After drawing a picture, we can see 4 congruent right triangles
with sides of 15, 20, 25, namely triangles DSR, OSR, OQP and BQP.
Let the points of triangle DSR be (0, 0) (0, 20) (15, 0). Let point E be on SR , such that SE = 16 and ER = 9.
Triangle DSR can be split into two similar 3-4-5 right triangles, ESD and EDR. By the Pythagorean
Theorem, point D is 12 away from point E. Repeating the process, if we break down triangle DER into two
more similar triangles, we find that point E is at (9.6, 7.2).
By reflecting point D = (0, 0) over point E = (9.6, 7.2), we get point O = (19.2, 14.4). By reflecting point D
672
over point O, we get point B = (38.4, 28.8). Thus, the perimeter is equal to (38.4 + 28.8) × 2 = , making
5
the final answer 672 + 5 = 677.

Problem-14
A rectangle that is inscribed in a larger rectangle (with one vertex on each side) is called unstuck if it is
possible to rotate (however slightly) the smaller rectangle about its center within the confines of the larger.
Of all the rectangles that can be inscribed unstuck in a 6 by 8 rectangle, the smallest perimeter has the form
N , for a positive integer N. Find N. [1993]

Solution
Answer: 448.
Solution: Put the rectangle on the coordinate plane so its vertices are at (± 4, ± 3), for all four combinations
of positive and negative. Then by symmetry, the other rectangle is also centered at the origin, O.
Note that such a rectangle is unstuck if its four vertices are in or on the edge of all four quadrants, and it is
not the same rectangle as the big one. Let the four vertices of this rectangle be A (4, y), b(–x, 3), C(–4, –y)
and D(x – 3) for nonnegative x, y. Then this is a rectangle, so OA = OB, or 16 + y2 = 9 + x2 , so x2 = y2 + 7.

B
b
A
A
C
A
D
D´

Reflect D across the side of the rectangle containing C to D´(–8 – x, – 3). Then BD´ =
(−8 − x − (−x)2 + (3 − (−3))2 ) = 10 is constant, and the perimeter of the rectangle is equal to 2(BC + CD´).
The midpoint of BD ' is (– 4 – x, 0), and since –4 < – 4 = – x and – y ≤ 0, C always lies below BD ' .
If y is positive, it can be decreased to y´ < y. This causes x to decrease as well, to x´, where x´2 = y´2 + 7 and
x´ is still positive. If B and D´ are held in place as everything else moves, then C moves (y – y´ ) units up
and (x – x´) units left to C´, which must lie within  BCD´ . Then we must have BC´ + C´D´ < BC + CD´,
and the perimeter of the rectangle is decreased. Therefore, the minimum perimeter must occur with y = 0, so
x= 7.
By the distance formula, this minimum perimeter is

75
  2 
2  4 − 7 + 32 + (4 + 7) 2 + 32  = 4
 
( 8−2 7 + 8+ 2 7 )
=4 ( 7 −1 + 7 + 1 = 8 7 =) 488 .

Solution 2
Note that the diagonal of the rectangle with minimum perimeter must have the diagonal along the middle
segment of length 8 of the rectangle (any other inscribed rectangle can be rotated a bit, then made smaller;
this one can't because then the rectangle cannot be inscribed since its longest diagonal is less than 8 in
length). Then since a rectangle must have right angles, we draw a circle of radius 4 around the center of the
rectangle. Picking the two midpoints on the sides of length 6 and opposite intersection points on the
segments of length 8, we form a rectangle. Let a and b be the sides of the rectangle. Then ab = 3(8) = 24
since both are twice the area of the same right triangle, and a2 + b2 = 64. . So (a + b)2 = 64 + 2(24) = 112, so
2 (a + b) = 488 .

20 – SEQUENCE & SERIES

Problem-5
A sequence of integers a1, a2, a3, … is chosen so that an = a n – 1 – a n – 2 for each n ≥ 3. What is the sum of
the first 2001 terms of this sequence if the sum of the first 1492 terms is 1985, and the sum of the first 1985
terms is 1492? [1985]
Solution
The problem gives us a sequence defined by a recursion, so let's calculate a few values to get a feel for how
it acts. We aren't given initial values, so let a1 = a and a2 = b. a2 = b. Then a3 = b – a, a4 = (b – a) – b = – a, a5
= – a – (b – a) = – b, a6 = – b – (– a) = a – b, a7 = (a – b) – ( – b) = a, and a8 = a – (a – b) = b.
Since the sequence is recursively defined by the first 2 terms, after this point it must continue to repeat.
Thus, in particular aj+6 = aj for all j, and so repeating this times, aj+6 = aj for all integers n and j.
Because of this, the sum of the first 1492 terms can be greatly simplified: 1488 = 6.248 is the largest
multiple of 6 less than 1492, so
1492 1488 247 6

 ai = (a1489 + a1490 + a1491 + a1492 )

i =1
+  ai = (a1 + a 2 + a 3 + a 4 ) =
i =1
 a
n = 0 j=1
6n + j = (a + b+ (b − a) + (− a)) +
247 6 6

 a
n = 0 j=1
j = 2b − a , where we can make this last step because a
j=1
j = 0 and so the entire second term of our

expression is zero.
Similarly, since 1980 = 6.360,
1985 1980

 a i = (a1 + a 2 + a 3 + a 4 + a 5 ) +
i =1
a
i =1
i = a + b+ (b− a) + (− a) + (− b) = b − a
2001 1988
Finally, a
i =1
i = a1 + a 2 + a 3 +  a i = a + b + (b− a) = 2 b
i =1

Then by the givens, 2b – a = 1985 and b –a = 1492 so b = 1985 – 1492 = 493 and so the answer is 2.493 =
986.

Problem-13
The numbers in the sequence 101, 104, 109, 116, … are of the form an = 100 + n2, where n = 1, 2, 3…. For
each n, let dn be the greatest common divisor of an and an + 1. Find the maximum value of dn as n ranges
through the positive integers. [1985]

76
Solution
Solution 1
If (x, y) denotes the greatest common divisor of x and y, then we have dn = (an , an + 1) = (100 + n2 , 100 + n2
+ 2n + 1). Now assuming that dn divides 100 + n2 , it must divide 2n + 1 if it is going to divide the entire
expression 100 + n2 + 2n + 1.
Thus the equation turns into dn = (100 + n2, 2n + 1). Now note that since 2n + 1 is odd for integral n , we
can multiply the left integer, 100 + n2 , by a multiple of two without affecting the greatest common divisor.
Since the n2 term is quite restrictive, let's multiply by 4 so that we can get a (2n + 1)2 in there.
So
dn = (4n2 + 400, 2n + 1) = ((2n + 1)2 – 4n + 399, 2n + 1) = (– 4n + 399, 2n + 1).
It simplified the way we wanted it to! Now using similar techniques we can write
d n = (–2 (2n + 1) + 401, 2n + 1) = (401, 2n + 1).
Thus dn must divide 401 . for every single n. This means the largest possible value for dn is 401, and we
see that it can be achieved when n = 200
Solution 2
We know that an = 100 + n2 and an + 1 = 100 + (n + 1)2 = 100 + n2 + 2n + 1. Since we want to find the GCD
of an and an + 1 , we can use the Euclidean algorithm:
an + 1 – a n = 2n + 1
Now, the question is to find the GCD of 2n + 1 and 100 +n2 . We subtract 2n + 1 100 times from 100 + n2 .
This leaves us with n2 – 200n. We want this to equal 0, so solving for n gives us n = 200. The last remainder
is 0, thus 200 * 2 + 1 = 401 is our GCD.

Problem-7
The increasing sequence 1, 3, 4, 9, 10, 12, 13 ….. consists of all those positive integers which are powers of
3 or sums of distinct powers of 3. Find the 100th term of this sequence. [1986]
Solution
Solution 1
Rewrite all of the terms in base 3. Since the numbers are sums of distinct powers of 3, in base 3 each number
is a sequence of 1s and 0s (if there is a 2, then it is no longer the sum of distinct powers of 3). Therefore, we
can recast this into base 2 (binary) in order to determine the 100th number. 100 is equal to 64 + 32 + 4, so in
binary form we get 1100100. However, we must change it back to base 10 for the answer, which is 3 6 + 35 +
32 = 729 + 243 + 9 = 981 .
Solution 2
Notice that the first term of the sequence is 1, the second is 3, the fourth is 9, and so on. Thus the 64 th term
of the sequence is 729. Now out of 64 terms which are of the form 729 + S, 32 of them include 243 and
32 do not. The smallest term that includes 243, i.e. 972, is greater than the largest term which does not, or
854. So the 95th term will be 972, then 973, then 975, then 976, and finally 981 .

Problem-13
In a sequence of coin tosses, one can keep a record of instances in which a tail is immediately followed by a
head, a head is immediately followed by a head, and etc. We denote these by TH, HH, and etc. For example,
in the sequence HHTTHHHHTHHTTTT of 15 coin tosses we observe that there are five HH, three HT, two
TH, and four TT subsequences.
How many different sequences of 15 coin tosses will contain exactly two HH, three HT, four TH, and five
TT subsequences ? [1986]
Solution
Let's consider each of the sequences of two coin tosses as an operation instead; this operation takes a string
and adds the next coin toss on (eg, THHTH + HT = THHTHT).
We examine what happens to the last coin toss. Adding HH or TT is simply an identity for the last coin toss,
so we will ignore them for now. However, adding HT or TH switches the last coin. H switches to T three
times, but T switches to H four times; hence it follows that our string will have a structure of THTHTHTH.
Now we have to count all of the different ways we can add the identities back in. There are 5 TT
subsequences, which means that we have to add 5 T into the strings, as long as the new Ts are adjacent to
existing Ts. There are already 4 Ts in the sequence, and since order doesn’t matter between different tail flips

77
this just becomes the ball-and-urn argument. We want to add 5 balls into 4 urns, which is the same as 3
 5 + 3  2 + 3
dividers; hence this gives   = 56 combinations. We do the same with 2 Hs to get   = 10
 3   3 
combinations; thus there are 56.10 = 560 possible sequences.

Problem-1
The increasing sequence2, 3, 5, 6, 7, 10, 11, …. consists of all positive integers that are neither the square
nor the cube of a positive integer. Find the 500th term of this sequence. [1990]
Solution
Because there aren't that many perfect squares or cubes, let's look for the smallest perfect square greater than
500. This happens to be 232 = 529. Notice that there are 23 squares and 8 cubes less than or equal to 529,
but 1 and 26 are both squares and cubes. Thus, there are 529 – 23– 8 + 2 = 500 numbers in our sequence less
than 529. Magically, we want the 500th term, so our answer is the smallest non-square and non-cube less
than 529, which is 528 .

Problem-6
Suppose r is a real number for which
 19   20   21   91 
 r + 100  +  r + 100  +  r + 100  + ... +  r + 100  = 546
Find [100r] . (For real x, [x] is the greatest integer less than or equal to x .) [1991]

Solution
 91 
There are 91 – 19 + 1 = 73 numbers in the sequence. Since  r + can be at most 1 apart, all of the
 100 
546
numbers in the sequence can take one of two possible values. Since = 7R35, the numbers must be
73
either 7 or 8. As the remainder is 35, 8 must take on 35 of the values, with 7 being the value of the remaining
73 – 35 = 38 numbers. The 39th number is 19 + 39 – 1 = 57, and so
 57 
8  r +  
 100 
743 744
Solving shows that r< , so [100r] = 743.
100 100

Problem-8
For any sequence of real numbers A = (a1, a2, a3, …….0), define A to be the sequence (a2 – a1, a3 – a2, a4 –
a3, ……..), whose nth term is an + 1 – an. Suppose that all of the terms of the sequence  ( A) are , and that
a19 = a92 = 0. Find a1. [1992]
Solution
Solution 1
Note that the s are reminiscent of differentiation; from the condition (A) = 1, we are led to consider the
d2A
differential equation = 1.
dn 2
1
This inspires us to guess a quadratic with leading coefficient 1/2 as the solution; an = (n – 19)(n – 92) as
2
we must have roots at n = 19 and n = 92.
1
Thus, a1 = (1 – 19)(1 – 92) = 819 .
2
Solution 2
Let  = A, and nA =  ( (n – 1)A).

78
  n − 1  n − 1 3
 an +  an +    an + ……
2
Note that in every sequence of ai, an = 
 1   3 
 n − 1  n − 1 2  n − 1 3
Then an = a1 +   a1 +    a1 +    a1 + ……
 1   2   3 
Since a1 = a2 – a1,
 n − 1  n − 1  n − 1
an = a1 +   (a2 – a1) +   .1 = a1 + n(a2 –a1) +  
 1   2   2 
 18 
a19 = 0 + a1 + 18(a2 – a1) +   = 18a2 – 17a1 + 153
2
 91 
a92 = 0 = a1 + 91(a2 – a1) +   = 91a2 – 90a1 + 4095
2
Solving, a1 = 819 .

Solution 3
The sequence  (A) is the second finite difference sequence, and the first k – 1 terms of this sequence can
be computed in terms of the original sequence as shown below.
a3 + a1 – 2a2 = 1
a4 + a2 – 2a3 = 1
ak + ak–2 – 2ak = 1
ak+1 + ak–1 – 2ak = 1
Adding the above k – 1, equations we find that
(ak + 1 – ak) = k – 1 + (a2 – a1). (1)
We can sum equation (1) from k =1 to 18, finding
18(a1 – a2) – a1 = 153. (2)
We can also sum equation (1) from k = 1 to 91, finding
91(a1 – a2) – a1 = 4095. (3)
Finally, 18.(3) – 91.(2) gives a1 = 819 .
Kris17
Problem-1
The increasing sequence 3, 15, 24, 49, …… consists of those positive multiples of 3 that are one less than a
perfect square. What is the remainder when the 1994th term of the sequence is divided by 1000? [1994]
Solution
One less than a perfect square can be represented by n2 – 1 = (n + 1)(n – 1). Either n + 1 or n – 1 must be
divisible by 3. This is true when n ≡ – 1, 1 ≡ 2, 1(mod 3).
n ≡ – 1, 1 ≡ 2, 1 (mod 3).
1994
Since 1994 is even, n must ≡ 1 (mod 3). It will be the = 997th such term, so
2
n = 4 + (997 – 1).3 = 2992. The value of n2 – 1 = 29922 – 1 (mod 1000) is 063.

22 – CORDINATE GEOMETRY

Problem-2
What is the largest possible distance between two points, one on the sphere of radius 19 with center (– 2, –
10, 5) and the other on the sphere of radius 87 with center (12, 8,– 16) ? [1987]
Solution
The distance between the two centers of the spheres can be determined via the distance formula in three
dimensions:
(12 − (−2))2 + (8 − (−10))2 + (−16 − 5)2 = 142 + 182 + 212 = 31.

79
The largest possible distance would be the sum of the two radii and the distance between the two centers,
making it 19 + 87 + 31 = 137 .

Problem-2
During a recent campaign for office, a candidate made a tour of a country which we assume lies in a plane.
On the first day of the tour he went east, on the second day he went north, on the third day west, on the
n2
fourth day south, on the fifth day east, etc. If the candidate went miles on the nth day of this tour, how
2
many miles was he from his starting point at the end of the 40th day ? [1993]
Solution
On the first day, the candidate moves
[4(0) + 1]2/2 east, [4(0) + 2]2/2 north, [4(0) + 3]2/2 west , [4(0) + 4]2/2 south, and so on.
The E/W displacement is thus
9
(4i + 1)2 9 (4i + 3)2
12 – 32 + 52 …… + 372 – 392 =  − .
i =o 2 i =o 2
Applying difference of squares, we see that
9
(4i + 1)2 9 (4i + 3)2 9
(4i + 1 + 4i + 3)(4i + 1 − (4i + 3))

i =o 2
− 
i =o 2
= 
i =o 2
9

=  −(8i + 4) .
i =o
9
(4i + 2) 2 9 (4i + 4) 2 9
The N/S displacement is  − =  −(8i + 6)
i =o 2 i =o 2 i =o

Since 5 3 = 45, the two distances evaluate to 8(45) + 10.4 = 400 and 8(45) + 10.6 = 420.
By the Pythagorean Theorem, the answer is 4002 + 4202 = 29.20 = 580 .

Problem-8
The points (0, 0), (a, 11) , and (b, 37) are the vertices of an equilateral triangle. Find the value of ab.
[1994]
Solution
Consider the points on the complex plane. The point b + 37i is then a rotation of 60 degrees of a + 11i about
the origin, so:
1 3i 
(a + 11i)(cis 60º) = (a + 11i)  +  = b + 37i.
 2 2 

Equating the real and imaginary parts, we have:

a 11 3
b= –
2 2
11 a 3
37 = +
2 2
Solving this system, we find that a = 21 3 , b = 5 3 .
Thus, the answer is 315 . A
Note: There is another solution where the point b + 37i is a rotation of – 60 degrees of a + 11i ; however, this
triangle is just a reflection of the first triangle by the y-axis, and the signs of a and b are flipped. However,
the product ab is unchanged.

80
23 – PERMUTATION & COMBINATION

Problem-1
One commercially available ten-button lock may be opened by depressing -- in any order -- the correct five
buttons. The sample shown below has {1, 2, 3, 6, 9} as its combination. Suppose that these locks are
redesigned so that sets of as many as nine buttons or as few as one button could serve as combinations. How
many additional combinations would this allow ? [1988]

1 6
2 7
3 8
4 9

5 10

Solution
 10 
Currently there are   possible combinations. With any integer x from 1 to 9, the number of ways to
5
9
10 
choose a set of x buttons is    .
k =1  k 
n
n
Now we can use the identity    = 2n.
k =0  k 

So the number of additional combinations is just

 10   10   10 
210 –   –   –   = 1024 – 1 – 1 – 252 = 770 .
 0   10   5 

24 – STRAIGHT LINE

Problem-11
 
Lines 1 and  2 both pass through the origin and make first-quadrant angles of and radians,
70 54
respectively, with the positive x-axis. For any line , the transformation R() produces another line as
follows:  is reflected in 1, and the resulting line is reflected in  2.
19
Let R() = R() and R(n)() = R(R(n – 1)()). Given that  is the line y = x , find the smallest positive
92
integer for which R(m)() = . [1992]
Solution
Let  be a line that makes an angle of  with the positive x-axis. Let  be the reflection of  in 1, and let
 be the reflection of  in 2.

The angle between and 1 is  − , so the angle between 1 and  must also be . Thus,  makes
70
   
an angle of −−  = −  with the positive x-axis.
70  70  35

81
    
Similarly, since the angle between  and  is  −   − , the angle between  and the positive x-axis is
 35  54
       8
−  −  −  = − + = + .
54   35  54  27 35 945
8
Thus, R() makes an +  angle with the positive x-axis.
945
8
So R(n)() makes an +  angle with the positive x-axis.
945
8m
Therefore, R(m)() =  iff is an integral multiple of . Thus, 8m ≡ 0 (mod 945).
945
Since gcd(8, 945) = 1, m ≡ 0 (mod 945), so the smallest positive integer m is 945 .

25 - POLYGON

Problem-3
59
Let P1 be a regular r gon and P2 be a regular s gon (r  s  3) such that each interior angle of P1 is as
58
large as each interior angle of P2. What's the largest possible value of s ? [1990]
Solution
(n − 2)180
The formula for the interior angle of a regular sided polygon is .
n
(r − 2)180
r 59 58(r − 2) 59(s − 2)
Thus, = . Cross multiplying and simplifying, we get = .
( s − 2)180 58 r s
s
Cross multiply and combine like terms again to yield 58rs − 58.2s = 59rs − 59.2r
 118r – 116s = rs.
Solving for r, we get .
116s
r= .
118 − s
r  0 and s  0, making the numerator of the fraction positive. To make the denominator positive, s  118;
the largest possible value of is 117 .

26 - STATISTICS

Problem-11
A sample of 121 integers is given, each between 1 and 1000 inclusive, with repetitions allowed. The sample
has a unique mode (most frequent value). Let D be the difference between the mode and the arithmetic mean
of the sample. What is the largest possible value of [D] ?
(For real x, [x] is the greatest integer less than or equal to x.) [1989]
Solution
Let the mode be x, which we let appear n > 1 times. We let the arithmetic mean be M, and the sum of the
numbers  x be S.
S + xn S  12 − n 
Then D = |M – x | = = − x
121 121  121 

82
As S is essentially independent of x, it follows that we wish to minimize or maximize x(in other words, x =
1, 1000). Indeed, D(x) is symmetric about x = 500.5; consider replacing all of numbers xi in the sample with
1001 – xi, and the value of D remains the same.
So, without loss of generality, let x = 1. Now, we would like to maximize the quantity
S  12 − n  S +n
−  (1) = −1
121  121  121
S contains 121 – n numbers that may appear at most n – 1 times. Therefore, to maximize S, we would have
1000 appear n – 1 times, 999 appear n – 1 times, and so forth. We can thereby represent S as the sum of n –
121 − n 
1 arithmetic series of 1000, 999, …. , 1001 –  .
 n − 1 
121 − n 
We let k =  , so
 n − 1 
 k (1000 + 1001 − k ) 
S = (n – 1)   + R(n)
 2
where R(n) denotes the sum of the remaining 121 – (n – 1)k numbers, namely
R(n) = (121 – (n – 1)k)(1000 – k).
12 − n
At this point, we introduce the crude estimate[1] that k = , so R(n) = 0 and 2S + 2n = (121 – n)
121
 121 − n   120 
 2001 −  + 2n = (120 – (n – 1))  2001 − 
 n −1   n −1 
Expanding (ignoring the constants, as these do not affect which yields a maximum) and scaling, we wish
36
to minimize the expression 5(n − 1) + .
n −1
36 36
By AM-GM, we have 5(n − 1) +   5(n − 1). , with equality coming when
n −1 n −1
36
5(n – 1) = , so n – 1  3.
n −1
Substituting this result and some arithmetic gives an answer of 947 .
In less formal language, it quickly becomes clear after some trial and error that in our sample, there will be n
values equal to one and n – 1 values each of 1000, 999, 998… .
It is fairly easy to find the maximum. Try n = 2, which yields 924, n = 3, which yields 942, n = 4, which
yields 947, and n = 5, which yields 944. The maximum difference occurred at n = 4, so the answer is 947.
Notes
• ^ In fact, when n = 2, 3, 4, 5 (which some simple testing shows that the maximum will occur
121 − n
around), it turns out that is an integer anyway, so indeed
n −1
121 − n  121 − n
K=  = .
 n − 1  n −1

Problem-3
The table below displays some of the results of last summer's Frostbite Falls Fishing Festival, showing how
many contestants caught n fish for various values of n.
n 0 1 2 3 … 13 14 15
Number of contestants who caught n fish 9 5 7 23 … 5 2 1

In the newspaper story covering the event, it was reported that

(a) the winner caught 15 fish;
(b) those who caught 3 or more fish averaged 6 fish each;
(c) those who caught 12 or fewer fish averaged 5 fish each.
What was the total number of fish caught during the festival? [1993]

83
Solution
Suppose that the number of fish is x and the number of contestants is y. The y – 21 fishers that caught 3 or
x − 19
more fish caught a total of x – (0(9) + 1(5) + 2(7)) = x – 19 fish. Since they averaged 6 fish, 6 = x
y − 21
– 19 = 6y – 126.
Similarily, those whom caught 12 or fewer fish averaged 5 fish per person, so
x − (13(5) + 14(2) + 15(1)) x − 108
5= =  x – 108 = 5y – 40.
y −8 y −8
Solving the two equation system, we find that y = 175 and x = 943 , the answer.

27 – BINOMIAL THEOREM

Problem-3
Expanding (1 + 0.2)1000 by the binomial theorem and doing no further manipulation gives
1000  1000  1000  1000 
 (0.2) + … + 
0 2 1000
  (0.2) +   (0.2)1 +   (0.2)
 0   1   2  1000 
1000 
= A0 + A1 + A2 + …. + A1000. where Ak =   (0.2) for k = 0, 1, 2, …, 1000.
k

 k 
For which k is Ak the largest? [1991]
Solution
Solution 1
N N! ( N − k + 1)! k
Let 0 < x < 1. Then we may write Ak =   x k = xk = x .
k k !( N − k )! k!
Taking logarithms in both sides of this last equation and using the well-known fact
Log(ab) = log a + log b (valid if a, b > 0), we have
 ( N − k + 1)! k   k ( N − j + 1)! x  k
 ( N − j + 1)! x 
Log (Ak) = log  x  = log = log    log 
= .
 k!   j =1 j!  j =1  j! 
( N − j + 1) x
Now, log(Ak) keeps increasing with k as long as the arguments > 1 in each of the log[] terms
j!
(recall that log y < 0 if 0 < y < 1).
 ( N + 1) x 
Therefore, the integer k that we are looking for must satisfy k =  , where [z] denotes the greatest
 1 + x 
integer less than or equal to z.
In summary, substituting N = 1000 and x = 0.2 we finally find that k = 166.
Solution 2
We know that once we have found the largest value of k, all values after Ak are less than Ak. Therefore, we
are looking for the largest possible value such that:
1k 1000  1k +1 1000 
.  . 
5  k  5  k +1 
1k
Dividing by gives:
5
1000  1 1000 
   . 
 k  5  k +1
We can express these binomial coefficients as factorials.
1000! 1 1000!
 . We note that the 1000! can cancel.
(1000 − k )!.(k )! 5 (1000 − k − 1)!.(k + 1)!

84
Also, (1000 – k)! = (1000 – k)(1000 – k – 1)! Similarly, (k + 1)! = (k + 1)k!
Canceling these terms yields,
1 1 1
 .
(1000 − k ) 5 (k + 1)
Cross multiplying gives:
5k + 5 > 1000 – k  k > 165.8
Therefore, since this identity holds for all values of k > 165.8, the largest possible value of k is 166 .

29 – LINES & ANGLES

Problem-14
A beam of light strikes BC at point C with angle of incidence  = 19.94º and reflects with an equal angle of
reflection as shown. The light beam continues its path, reflecting offline segments AB and BC according to
the rule: angle of incidence equals angle of reflection. Given that  = /10 = 1.994º and AB = AC determine
the number of times the light beam will bounce off the two line segments. Include the first reflection at C in
your count. [1994]
A

B  
C
Solution
At each point of reflection, we pretend instead that the light continues to travel straight.

A
•
B  
• •
•
•
•
•

Note that after k reflections (excluding the first one at C) the extended line will form an angle k at point .
For the kth reflection to be just inside or at the point C, we must have
180 − 2
k  180 – 2  k  = 70.27.

180 − 2 
Thus, our answer is, including the first intersection,   = 071 .
  

85