1. Let ABC be a right-angled triangle with ∠B = 90◦ . Let I be the incentre of ABC.

Draw a line
perpendicular to AI atpI. Let it intersect the line CB at D. Prove that CI is perpendicular to AD
and prove that ID = b(b − a) where BC = a and CA = b.

Solution: First observe that ADBI is a cyclic

quadrilateral since ∠AID = ∠ABD = 90◦ . Hence
∠ADI = ∠ABI = 45◦ . Hence ∠DAI = 45◦ . But
we also have

∠ADB = ∠ADI + ∠IDB = 45◦ + ∠IAB

= ∠DAI + ∠IAC = ∠DAC.

Therefore CDA is an isosceles triangle with CD =

CA. Since CI bisects ∠C it follows that CI ⊥ AD.

This shows that DB = CA − CB = b − a. Therefore

AD2 = c2 + (b − a)2 = c2 + b2 + a2 − 2ba = 2b(b − a).

p
But then 2ID2 = AD2 = 2b(b − a) and this gives ID = b(b − a).
2. Let a, b, c be positive real numbers such that
a b c
+ + = 1.
1+a 1+b 1+c
Prove that abc ≤ 1/8.

Solution: This is equivalent to

X
a(1 + b)(1 + c) = (1 + a)(1 + b)(1 + c).

This simplifies to
ab + bc + ca + 2abc = 1
Using AM-GM inequality, we have

1 = ab + bc + ca + 2abc ≥ 4(ab · bc · ca · 2abc)1/4 .

Simplificaton gives
1
abc ≤ .
8
3. For any natural number n, expressed in base 10, let S(n) denote the sum of all digits of n. Find
all natural numbers n such that n = 2S(n)2 .

Solution: We use the fact that 9 divides n−S(n) for every natural number n. Hence S(n)(2S(n)−1)
is divisible by 9. Since S(n) and 2S(n) − 1 are relatively prime, it follows that 9 divides either S(n)
or 2S(n) − 1, but not both. We also observe that the number of digits of n cannot exceed 4. If n
has k digits, then n ≥ 10k−1 and 2S(n)2 ≤ 2 × (9k)2 = 162k 2 . If k ≥ 6, we see that

2S(n)2 ≤ 162k 2 < 54 k 2 < 10k−1 ≤ n.

If k = 5, we have
2S(n)2 ≤ 162 × 25 = 4150 < 104 ≤ n.
 Therefore n ≤ 4 and S(n) ≤ 36.
If 9|S(n), then S(n) = 9, 18, 27, 36. We see that 2S(n)2 is respectively equal to 162, 648, 1458,
2592. Only 162 and 648 satisfy n = 2S(n)2 .
If 9|(2S(n) − 1), then 2S(n) = 9k + 1. Only k = 1, 3, 5, 7 give integer values for S(n). In these cases
2S(n)2 = 50, 392, 1058, 2048. Here again 50 and 392 give n = 2S(n)2 .
Thus the only natural numbers wth the property n = 2S(n)2 are 50, 162, 392, 648.

4. Find the number of all 6-digit natural numbers having exactly three odd digits and three even
digits.

Solution: First we choose 3 places for even digits. This can be done in 63 = 20 ways. Observe

that the other places for odd digits get automatically fixed. There are 5 even digits and 5 odd digits.
Any of these can occur in their proper places. Hence there are 56 ways of selecting 3 even and 3
odd digits for a particular selection of place for even digits. Hence we get 20 × 56 such numbers.
But this includes all those numbers having the first digit equal to 0. Since we are looking for 6-digit
numbers, these numbers have to be removed from our counting. If we fix 0 as the first digit, we
have, 2 places
for even numbers and 3 places for odd numbers. We can choose 2 places for even
numbers in 52 = 10 ways. As earlier, for any such choice of places for even digits, we can choose
even digits in 52 ways and odd digits in 53 ways. Hence the number of ways of choosing 3 even and
3 odd digits with 0 as the first digit is 10 × 55 . Therfore the number of 6-digit numbers with 3 even
digits and 3 odd digits is

20 × 56 − 10 × 55 = 10 × 55 (10 − 1) = 281250.

5. Let ABC be a triangle with centroid G. Let the circumcircles of 4AGB and 4AGC intersect the
line BC in X and Y respectively, which are distinct from B, C. Prove that G is the centroid of
4AXY .

Solution: Let D be the midpoint of AB. Ob-

serve that DX · DB = DG · DA = DY · DC.
But DB = DC. Hence DX = DY . This
means that D is the midpoint of XY as well.
Hence AD is also a median of 4AXY . Now
we know that AG : GD = 2 : 1. If G0 is the
median of 4AXY , then G0 must lie on AD
and AG0 : G0 D = 2 : 1. We conclude that
G = G0 .

6. Let ha1 , a2 , a3 , . . .i be a strictly increasing sequence of positive integers in an arithmetic progression.

Prove that there is an infinite subsequence of the given sequence whose terms are in a geometric
progression.

Solution: Let ha1 , a2 , . . . , an+1 . . .i = ha, a + d, . . . , a + nd, . . .i be a strictly increasing sequence of

positive integers in arithmetic progression. Here a and d are both positive integers. Consider the
following subsequence:
ha, a(1 + d), a(1 + d)2 , . . . , a(1 + d)n , . . .i.
This is a geometric progression. Here a > 0 and the common ratio 1 + d > 1. Hence the sequence is
strictly increasing. The first term is a which is in the given AP. The second term is a(1+d) = a+ad
which is the (a + 1)-th term of the AP. In general, we see that
      
n n n
a(1 + d)n = a + d a+ ad + · · · + adn−1 .
1 2 n

2
Here the coefficient of d in the braces is also a positive integer. Hence a(1 + d)n is also a term of
the given AP.

3
 Regional Mathematical Olympiad-2023

1. Let N be the set of all natural numbers and S = {(a, b, c, d) ∈ N4 : a2 + b2 + c2 = d2 }. Find the
largest positive integer m such that m divides abcd for all (a, b, c, d) ∈ S.

Solution

Since d2 ≡ 0, 1 (mod 4) at most one of a, b, c is odd. Therefore 4 divides abcd. Also, if 3 does not
divide each of a, b and c then

d2 = a2 + b2 + c2 ≡ 1 + 1 + 1 ≡ 0 (mod 3).

Thus 3 divides abcd. Therefore 12 divides abcd and if m is the largest positive integer such that m
divides all abcd ∈ S then m = 12k for some positive integer k. But (1, 2, 2, 3) ∈ S and [Link] = 12.
Hence k = 1 and m = 12.

Remarks

The set S is infinite because (n, n + 1, n(n + 1), n2 + n + 1) ∈ S for every positive integer n.
2. Let ω be a semicircle with AB as the bounding diameter and let CD be a variable chord of the
semicircle of constant length such that C, D lie in the interior of the arc AB. Let E be a point on
AB such that CE and DE are equally inclined to the line AB. Prove that
(a) the measure of ∠CED is a constant;
(b) the circumcircle of triangle CED passes through a fixed point.

Solution

Construct the circle with AB as diameter and let this circle be Ω. Draw CK ⊥ AB with K on AB.
Let CK produced meet Ω again in P . Join EP . Observe that

∠DEB = ∠CEK = ∠P EK.

1
 Hence ∠P EK + ∠CEK + ∠CED = 180◦ . Therefore P , E, D are collinear. This shows that

∠CED = 2∠CP D

is a constant. If O is the centre of Ω then we get ∠COD = 2∠CP D = ∠CED. Hence the
circumcircle of triangle CED passes through O which is a fixed point.
3. For any natural number n, expressed in base 10, let s(n) denote the sum of all its digits. Find all
natural numbers m and n such that m < n and

(s(n))2 = m and (s(m))2 = n.

Solution

Let m < n be such natural numbers. Let

m = 10k−1 ak−1 + 10k−2 ak−2 + · · · + 10a1 + a0

be a k-digit number. Then we have

10k−1 ≤ m < n = s(m)2 = (ak−1 + ak−2 + · · · + a1 + a0 )2 ≤ 92 k 2 .

If k ≥ 5, this is not possible. Hence k ≤ 4.

If k = 4, then

m = 1000a3 + 100a2 + 10a1 + a0 < (a3 + a2 + a1 + a0 )2 ≤ 362 = 1296.

This shows that a3 = 1. In this case

m = 1000 + 100a2 + 10a1 + a0 < (1 + a2 + a1 + a0 )2 ≤ 282 = 784,

which is impossible. Hence m must be a 3-digit number. Again

m = 100a2 + 10a1 + a0 < (a2 + a1 + a0 )2 ≤ 272 = 729.

Hence a2 ≤ 7. If a2 = 7, then

m = 700 + 10a1 + a0 < (7 + a1 + a0 )2 ≤ 252 = 625,

which is not possible. Similarly, a2 = 6 gives

m = 600 + 10a1 + a0 < (6 + a1 + a0 )2 ≤ 242 = 576,

which again is impossible. If a2 = 5, we obtain the maximal digital sum 23 when a1 = a0 = 9.

Otherwise s(m) ≤ 22 and

m = 500 + 10a1 + a0 < n = s(m)2 ≤ 222 = 484.

Thus we can also rule out a2 = 5. Therefore a2 ≤ 4. This means, m ≤ 222 .

Now we can search which squares up to 222 admit an n such that m = (s(n))2 and n = (s(m))2 . The
first such square is m = 81 = 92 . But in this case n = s(m)2 = 81. But now m = n violating m < n.
The next square is m = 169 = 132 . In this case s(m)2 = 162 = 256 = n and s(n)2 = 132 = 169.
Thereafter, no square satisfies this. Thus we get the pair

(m, n) = (169, 256).

2
4. Let Ω1 , Ω2 be two intersecting circles with centres O1 , O2 respectively. Let l be a line that intersects
Ω1 at points A, C and Ω2 at points B, D such that A, B, C, D are collinear in that order. Let the
perpendicular bisector of segment AB intersect Ω1 at points P, Q; and the perpendicular bisector
of segment CD intersect Ω2 at points R, S such that P, R are on the same side of l. Prove that the
midpoints of P R, QS and O1 O2 are collinear.

Solution

Let the midpoints of segments P Q, O1 O2 , RS be denoted by X, Y, Z respectively.

We observe that B is the reflection of A in line P Q. Hence B is the orthocentre of ∆CP Q.
Hence, O1 X = BC/2. Similarly, O2 Z = BC/2.
By the S-A-S test, ∆XO1 Y ∼ = ∆ZO2 Y ; hence X − Y − Z with XY = Y Z.
The endpoints of segments P R, XZ, QS lie on parallel lines P Q and RS, so their midpoints are
collinear.

3
5. Let n > k > 1 be positive integers. Determine all positive real numbers a1 , a2 , . . . , an which satisfy
s
n n
X kaki X
= ai = n.
i=1
(k − 1)aki + 1 i=1

Solution 1

By A.M-G.M inequality we have

(k − 1)aki + 1  k(k−1) 1/k
≥ ai = ak−1
i
k
s
kaki √
which implies ≤ ai . Hence
(k − 1)aki + 1
s
n n n
X √ X kaki X
ai ≥ = ai = n.
i=1 i=1
(k − 1)aki + 1 i=1

But by Cauchy-Schwarz inequality we have

v
n u n
X √ u X
ai ≤ tn( ai ) = n.
i=1 i=1

Therefore s
n n n
X √ X kaki X
n≥ ai ≥ k
= ai = n
i=1 i=1
(k − 1)ai + 1 i=1
and hence equality holds everywhere which implies ai = 1 for i = 1, 2, . . . , n.

Solution 2
kbk
Claim: For any nonnegative real number b we have ≤ b.
(k − 1)bk + 1
This inequality holds iff
(b − 1)(kbk−1 − bk−1 + bk−2 + · + 1 ) ≥ 0.

Observe that if b ≥ 1 then

kbk−1 − bk−1 + bk−2 + · + 1 ≥ 0

and if b ≤ 1 then
kbk−1 − bk−1 + bk−2 + · + 1 ≤ 0.

Combining these two cases we obtain

(b − 1)(kbk−1 − bk−1 + bk−2 + · + 1 ) ≥ 0

which proves our claim.

By this claim we have s
kaki √
k
≤ ai
(k − 1)ai + 1
for i = 1, 2, . . . , n. The rest of the solution is the same as Solution 1.

4
6. Consider a set of 16 points arranged in a 4 × 4 square grid formation. Prove that if any 7 of these
points are coloured blue, then there exists an isosceles right-angled triangle whose vertices are all
blue.

Solution

Let us label the points as illustrated in the above diagram. We can consider the following cases:
Case 1: None of the central 4 points {F, G, J, K} is colored.
We can partition the remaining 12 points into the 3 sets {A, D, P, M }, {B, H, O, I}, {C, L, N, E}.
By PHP, at least 3 of the 7 colored points lie in the same set; forming a 45 − 45 − 90 triangle (an
isosceles right-angled triangle).
Case 2: At least one of the central 4 points is colored; WLOG let point F be colored.
Subcase 2.1: Points F, C are both colored.
Then, none of the points A, B, G, H, K can be colored, as each of them forms a 45 − 45 − 90 triangle
along with F, C. The remaining 9 points (out of which 5 are colored) can be partitioned into the
4 sets {E, I, J}, {D, O}, {L, M }, {N, P }. So by PHP, some set contains atleast 2 colored points,
which form a 45 − 45 − 90 triangle along with F .
Subcase 2.2: Point F, I are both colored. By symmetry, this is identical to subcase 2.1.
Subcase 2.3: Point F is colored, but neither C nor I is colored.
Then apart from C, F, I, the remaining 13 points (out of which 6 are colored) can be partitioned into
the 5 sets {A, B, E}, {G, J, K}, {D, O}, {H, N, P }, {L, M }. So by PHP, some set contains atleast 2
colored points, which form a 45 − 45 − 90 triangle along with F .

———-0———-

5
 RMO 2024
Official Solutions

Problem 1. Let n > 1 be a positive integer. Call a rearrangement a1 , a2 , . . . , an of 1, 2, . . . , n

nice if for every k = 2, 3, . . . , n, we have that a1 + a2 + · · · + ak is not divisible by k.
(a) If n > 1 is odd, prove that there is no nice rearrangement of 1, 2, . . . , n.
(b) If n is even, find a nice rearrangement of 1, 2, . . . , n.

Solution. For the first part, note that the given condition for k = n implies that the sum
a1 + a2 + . . . + an is not divisible by n. However, a1 , a2 , . . . , an is a rearrangement of 1, 2, . . . , n
n(n + 1)
so their sum is equal to 1 + 2 + . . . + n = which is divisible by n for odd n. Thus,
2
there cannot be any nice rearrangement of 1, 2, . . . , n for odd n.
For the second part, let n = 2m. We show that the sequence

2, 1, 4, 3, 6, 5, 8, 7, ....., 2m, 2m − 1

k(k + 1)
is a nice rearrangement of 1, 2, . . . , 2m. For k even, we have a1 +a2 +. . .+ak = which
2
is not divisible by k since (k + 1)/2 is not an integer. For k odd, we have a1 + a2 + . . . + ak =
k(k + 1)
+ 1 which is 1 more than a multiple of k, so it is again not divisible by k for
2
k > 1.

Problem 2. For a positive integer n, let R(n) be the sum of the remainders when n is
divided by 1, 2, . . . , n. For example, R(4) = 0 + 0 + 1 + 0 = 1, R(7) = 0 + 1 + 1 + 3 + 2 + 1 + 0 = 8.
Find all positive integers n such that R(n) = n − 1.

n
Solution. Let n > 8. The remainder when n is divided by some i satisfying 2 < i ≤ n is
(n − i). Adding, we get that

n ⌈n
2 ⌉−1
X X 1 l n m l n m  1 lnm
n − 1 = R(n) ≥ (n − i) = k= −1 ≥ ·4≥n
2 2 2 2 2
i=⌊ n
2 ⌋+1
k=1

This is a contradiction. So, we get that n ≤ 8. Now we can compute that R(1) = R(2) =
0, R(3) = R(4) = 1, R(5) = 4, R(6) = 3, R(7) = R(8) = 8. Therefore, the only solutions are
n = 1 and n = 5.

Problem 3. Let ABC be an acute triangle with AB = AC. Let D be the point on BC
such that AD is perpendicular to BC. Let O, H, G be the circumcentre, orthocentre and
centroid of triangle ABC respectively. Suppose that 2 · OD = 23 · HD. Prove that G lies on
the incircle of triangle ABC.

Solution. Let I be the incenter of △ABC. First note that O, G, H, I all lie on AD since it
is simultaneously the perpendicular bisector of BC, the A−altitude, the A− median and
the angle bisector of ∠BAC.
Suppose the reflection of H across BC is M . Then M lies on the circumcircle of △ABC
as well as lies on the angle bisector of ∠BAC, so it is the midpoint of arc BC not containing
A. Then, we note that ∠M BI = ∠M IB, so M B = M I. Combining with M B = M C, we
have that M is the circumcenter of △BIC.
23
Now, let the circumradius of △ABC be R, let OD = x, HD = y. Then we have x = y.
2
2
Also, R = OM = OD + DM = OD + HD = x + y. Thus, y = R. This implies that
25

1
 48 16
AD = 2R − y = R. Now, recall that G divides AD in the ratio 2 : 1, so GD = R.
25 25
A
Also, we have △M DB ∼ △M BA since the angle
at M is common and ∠M BD = ∠M AB, both
equalling ∠BAC/2. Therefore, M B 2 = M D · M A,
and hence
4 2 2
M I 2 = M D · M A = y · 2R = R =⇒ M I = R. O
25 5
G
8
Thus, ID = R, which combined with GD =
25 I
16
R implies that GI = ID is equal to the inra-
25 H
dius, proving that G lies on the incircle. D
B C
M

Remark. A student well-versed in trigonometry may readily obtain cos A = 23/25 by

observing that OD = R cos A and HD = 2R cos B cos C = 2R cos2 (90◦ − A/2) = R(1 − cos A).
Now GD = AD/3 = (OA + OD)/3 = 16R/25 and r = AD/(1 + csc(A/2)) = 48R/(25 × 6) = 8R/25
whence GI = ID = r and the conclusion follows.

Problem 4. Let a1 , a2 , a3 , a4 be real numbers such that a21 + a22 + a23 + a24 = 1. Show that
1
there exist i, j with 1 ≤ i < j ≤ 4, such that (ai − aj )2 ≤ .
5

Solution 1. Let m be the minimum of |ai − aj | over all 1 ≤ i < j ≤ 4. Without loss
of generality, we may assume that a1 ≤ a2 ≤ a3 ≤ a4 . Then aj − ai ≥ (j − i)m for all
1 ≤ i < j ≤ 4. Thus,
X X
(ai − aj )2 ≥ (j − i)2 m2 = 20m2 .
1≤i<j≤4 1≤i<j≤4

On the other hand,

X
(ai − aj )2 = 4(a21 + a22 + a23 + a24 ) − (a1 + a2 + a3 + a4 )2 ≤ 4.
1≤i<j≤4

Thus, 20m2 ≤ 4 =⇒ m2 ≤ 1/5.

1
Solution 2. Suppose |ai − aj | > √ for all 1 ≤ i < j ≤ 4. Then if x, y are respectively the
5
3
maximum and minimum among the ai , then x − y > √ . Suppose u, v are the other two
5
2 2 1 2
ai apart from x, y. Then using a + b ≥ (a − b) , we have that
2
 
1 1 1 9 1
1 = x2 + y 2 + u2 + v 2 ≥ (x − y)2 + (u − v)2 > + =1
2 2 2 5 5

which is a contradiction.

Remark. There is another solution involving casework where the cases involving the
number of positive and negative ai are distinguished. We exclude it for brevity.

Problem 5. Let ABCD be a cyclic quadrilateral such that AB is parallel to CD. Let O be
the circumcentre of ABCD, and L be the point on AD such that OL is perpendicular to
AD. Prove that
OB · (AB + CD) = OL · (AC + BD).

Solution 1. Let K be the foot of perpendicular from O onto BC. Note that ABCD is a
isosceles trapezium, therefore AC + BD = 2AC. We have that L and K are the midpoints
of AD and BC respectively, therefore LK = (AB + CD)/2. Also OB = OA. Thus it suffices
OA OL
to prove that = .
AC LK

2
 A B

L K

D C

Now ∠AOL = ∠ACD = ∠BDC = ∠COK. Thus, ∠AOC = ∠LOK. Also note that
OL = OK since distance from center to two equal chords is the same. Thus, △AOC
and △LOK are isosceles triangles with ∠AOC = ∠LOK, hence they are similar, which
immediately implies the desired.

Solution 2. As before, note that ABCD is an isosceles trapezium. Let the intersection
of AC and BD be E, the foot of perpendicular from A onto CD be P , and let AP = h.
Let ∠BDC = ∠ACD = x Then ∠BEC = 2x. Let the radius OB = R. Thus, [ABCD] =
1 1
AC 2 sin 2x = (AB + CD) · h. Now, note that OL = R cos x by considering △AOL, and
2 2
h = AC sin x. Therefore, (AB + CD) · h = AC 2 sin 2x = h · 2 · AC cos x. Hence

OL AB + CD
= cos x =
OB 2AC
which finishes the problem since AC = BD.

Problem 6. Let n ≥ 2 be a positive integer. Call a sequence a1 , a2 , · · · , ak of integers an

n-chain if 1 = a1 < a2 < · · · < ak = n, and ai divides ai+1 for all i, 1 ≤ i ≤ k − 1. Let f (n) be
the number of n-chains where n ≥ 2. For example, f (4) = 2 corresponding to the 4-chains
{1, 4} and {1, 2, 4}.
Prove that f (2m · 3) = 2m−1 (m + 2) for every positive integer m.

Solution. We will prove that for any two distinct primes p, q, that f (pm · q) = 2m−1 (m + 2)
for all integers m ≥ 1. Suppose n = pm · q, and let {a1 , a2 , · · · , ak } be a n-chain. Then
ai divides ai+1 implies that ai+1 /ai = pbi · q ci , where bi , ci are non-negative integers for
i = 1, · · · , k − 1. Note that ai+1 > ai implies that bi and ci cannot be simultaneously 0.
Now, we have b1 + . . . + bk−1 = m and c1 + . . . + ck−1 = 1. Thus, exactly one of the ci will
be equal to 1, and that implies that at most one of the bi can be 0.
Recall that a composition of m is a sequence of positive integers adding to m. Cor-
responding to any l-length composition x1 , . . . , xl of m, we will get exactly 2l + 1 many
n-chains. l of them are obtained by setting bi = xi for all i and choosing one of c1 , . . . cl to
be 1, and rest to be 0. The other l + 1 chains of length l + 1 are obtained by choosing some
1 ≤ j ≤ l + 1, then setting cj = 1, bj = 0, bi = xi for all i < j, and bi = xi−1 for all i > j.
This can be done in various ways as follows:

First way: it is well known that there are m−1 l−1 compositions of m with l parts. There-

3
fore, we need
m   m−1
X 
X m−1 m−1
(2l + 1) = (2l + 3)
l−1 l
l=1 l=0
m−1
X m − 1 m−1
X m − 1
=2 l· +3
l l
l=0 l=0
m−1   !
X m−2
=2 (m − 1) · + 3 · 2m−1
l−1
l=1
= 2(m − 1)2m−2 + 3 · 2m−1 = 2m−1 (m + 2).

Second way: We will show that the total number of compositions of m is 2m−1 and the
sum of the number of parts over all compositions of m is 2m−2 (m + 1) via direct bijections.
This finishes the problem, since we get the sum of (2l + 1) over all compositions to be
2 · 2m−2 (m + 1) + 2m−1 = 2m−1 (m + 2).
For the first one, consider sequences of 0’s and 1’s such that there are exactly m 1’s,
no two 0’s are adjacent and the sequence begins and ends with a 1. Then we can choose
whether or not to insert a 0 in the m − 1 spaces between the 1’s, hence there are 2m−1
possible ways to do it.
For the second one, we consider the above sequences but we put a single 0 at the end,
and we also select a special 0. Then we can choose the special 0 first. If this is the last
0 then we get 2m−1 choices for the other zeroes, and if not then we have m − 1 choices
for the special 0, and then 2m−2 choices for the other spaces. This totals to 2m−2 (m + 1)
ways.

Remark: There are other solutions involving induction using recursions of the form
X X
f (2m · 3) = f (2l · 3) + f (2l )
0≤l<m 0≤l≤m

or f (2m · 3) = 2(f (2m−1 · 3) + f (2m ) − f (2m−1 )). Again, we omit them for the sake of brevity.

4
 Mϖ Kϖ ϖ ϚϘϙϟ
RKLϢ ϛ 6€=̱ ϘϠ ŷ IMϔ ϚϘϙϟ

̋FDͦ PϢ
Ԣ 3%N3N$=M ϐ̋3R J ˿΁G K͟ϑ L 8€D NF$ 3̨ FǨB FS͕ S̲™

Ԣ ΁NM O€ þ3M NF$ 3̨ FǨB S̲™

Ԣ RJ þˎͳ 3$ :OI Dͤ ™

Ԣ RJ þˎ IMIM €3ͳ 3$ Sͮ™ ̎E3BK €3Ϣ 102
Ԣ SM þˎ 3 SN F GɈ$ R$ P΂ 3Mͤ™ þˎ R€ƣ 3 RlϕRl ʽ$4 3Mͤ™

ϙϖ KF N̒: ̌3 AOB 3 3&A ̋DL ΃ S̲ :& 180◦ R$ 3K S̲ M P ϔ ̸ AOB ȬM ̋FE¡̈MB 3&AL U$ù K͟ϔ 3
JBM ̍I͎ ̴ S̲™ þKA 3$ RC ̋D4 ̌3ϔ 3$ON ΁NM O€ þ3M 3 þL&5 3MB$ ΃ ϔ P R$ 5^MB$ ΃ 3 R$ M̱44€?
CD 3̨ M8F 3%R$ 3Mͤ5$ B̌3 C E¡ϕM̱4 OA GM ̞˳B S&ϔ D E¡ϕM̱4 OB GM ̞˳B S&ϔ M CP : P D = 1 : 2
S&™
Ϛϖ ̋D4 3̨ RK3MA

a3 + (a + 1)3 + (a + 2)3 + (a + 3)3 + (a + 4)3 + (a + 5)3 + (a + 6)3 = b4 + (b + 1)4

3 GA͌3ͳ a, b K͟ 3& SN FS͕ S̲™

ϛϖ KF N̒: ̌3 P (x) = x2 + 12 x + b O€ Q(x) = x2 + cx + d D& I΃GD Sͮ ̒:F3$ 5A3 Oˮ̌O3 Sͮ M RJ
Oˮ̌O3 x 3$ ̒N P (x)Q(x) = Q(P (x))™ R ̚ˮ̏C K͟ RK3MA P (Q(x)) = 0 3$ RJ Oˮ̌O3 SN VB
3̨̒: ™
Ϝϖ KF N̒: ̌3 3B͗L BN K͟ n2 3ϕO5¡ ϐþȓ$3 3 U$ùHN ϙ S̲ϑ ̋D 5 Sͮ ̒:F3$ 3͟û (i, j) Sͮϔ :S 1 ≤ i ≤
n, 1 ≤ j ≤ n™ þȓ$3 O5¡ 3& R þ3M R$ M€5 R$ JMF S̲ ̌3 :I J 1 ≤ i < j ≤ n O 1 ≤ k < l ≤ n S&ϔ B&
F BFͳ O5Ͷ 3$ M€5 ̍JɈ Sͳ ̒:F3$ 3͟û (i, k), (j, k), (j, l) Sͮ™ R ̚ˮ̏C K͟ ɓFBK Oː3 M€5ͳ 3̨ R€ƣ VB
3̨̒: ™
ϝϖ KF N̒: 3̨ Ω 3 OȆ S̲ M AB 3 :O S̲ :& ̌3 ˄R FS͕ S̲™ KF N̒: ̌3 Γ1 M̱4 AB 3̨ 3 BMl
3 OȆ S̲ :& M̱4 AB 3& C GM ˷P¡ 3MB S̲ M OȆ Ω 3& D GM JBM R$ ˷P¡ 3MB S̲™ R BMS KF N̒:
̌3 Γ2 M̱4 AB 3̨ ̴ RM BMl 3 OȆ S̲ :& M̱4 AB 3& E GM ˷P¡ 3MB S̲ M OȆ Ω 3& F GM JBM R$ ˷P¡
3MB S̲™ KF N̒: ̌3 M̱4 DC OȆ Ω R$ X ̸= D GM ̍KNB S̲ M M̱4 F E OȆ Ω R$ Y ̸= F GM ̍KNB S̲™
þK̓AB 3̈M ̌3 XY OȆ Ω 3 ˄R S̲™
Ϟϖ KF N̒: ̌3 þȓ$3 Oˮ̌O3 R€ƣ x, y, z R€ƣ 1 R$ Ic S̲™ þK̓AB 3̈M ̌3Ϣ
x+1 y+1 z+1 x−1 y−1 z−1
+ + ≤ + +
y+1 z+1 x+1 y−1 z−1 x−1

ϙ
Regional Mathematical Olympiad-2019 problems and solutions
19
1. Suppose x is a nonzero real number such that both x5 and 20x + are rational numbers. Prove
x
that x is a rational number.

Solution:Since x5 is rational, we see that (20x)5 and (x/19)5 are rational numbers. But
 5 
194
 
19 19 1
(20x)5 − = 20x − (20x)4 + (203 · 19)x2 + 202 · 192 + (20 · 193 ) 2 + 4 .
x x x x
Consider
194
 
1
T = (20x)4 + (203 · 19)x2 + 202 · 192 + (20 · 193 ) 2 + 4
x x
4 2
   
19 19
= (20x)4 + 4 + 20 · 19 (20x)2 + 2 + (202 · 192 ).
x x
Using 20x + (19)/x is rational, we get
2
192

19
(20x)2 + = 20x + − 2 · 20 · 19
x2 x
is rational. This leads to
2
194 192

(20x)4 + = (20x)2 + 2 − 2 · 202 · 192
x4 x
6 0. We conclude that 20x−(19/x) is a rational
is also rational. Thus T is a rational number and T =
number. This combined with the given condition that 20x + (19/x) is rational shows 2 · 20 · x is
rational. Therefore x is rational.
2. Let ABC be a triangle with circumcircle Ω and let G be the centroid of triangle ABC. Extend AG,
BG and CG to meet the circle Ω again in A1 , B1 and C1 , respectively. Suppose ∠BAC = ∠A1 B1 C1 ,
∠ABC = ∠A1 C1 B1 and ∠ACB = ∠B1 A1 C1 . Prove that ABC and A1 B1 C1 are equilateral
triangles.

Solution:
 Let ∠BAA1 = α and ∠A1 AC = β. Then ∠BB1 A1 = α. Using that angles at A and B1 are same,
we get ∠BB1 C1 = β. Then ∠C1 CB = β. If ∠ACC1 = γ, we see that ∠C1 A1 A = γ. Therefore
∠AA1 B1 = β. Similarly, we see that ∠B1 BA = ∠A1 C1 C = β and ∠B1 BC = ∠B1 C1 C = δ.
Since ∠F BG = ∠BCG = β, it follows that F B is tangent to the circumcircle of 4BGC at B.
Therefore F B 2 = F G · F C. Since F A = F B, we get F A2 = F G · F C. This implies that F A is
tangent to the circumcircle of of 4AGC at A. Therefore α = ∠GAF = ∠GCA = γ. A similar
analysis gives α = δ.
It follows that all the angles of 4ABC are equal and all the angles of 4A1 B1 C1 are equal. Hence
ABC and A1 B1 C1 are equilateral triangles.
3. Let a, b, c be positive real numbers such that a + b + c = 1. Prove that
a b c 1
+ 2 + 2 ≤ .
a2 + b3 + c3 b + c3 + a3 c + a3 + b3 5abc

Solution:Observe that

a2 + b3 + c3 = a2 (a + b + c) + b3 + c3 = (a3 + b3 + c3 ) + a2 (b + c) ≥ 3abc + a2 b + a2 c.

Hence
a 1
≤ .
a2 + b3 + c3 3bc + ab + ac
Using AM-HM inequality, we also have
3 1 1 25
+ + ≥ .
bc ca ab 3bc + ca + ab
Thus we get  
a 1 1 3 1 1
≤ ≤ + + .
a2 + b3 + c3 3bc + ab + ac 25 bc ca ab
Similarly, we get  
b 1 3 1 1
≤ + +
b2 + c3 + a3 25 ca ab bc
and  
c 1 3 1 1
≤ + +
c2 + a3 + b3 25 ab bc ca
Adding, we get
 
a b c 5 1 1 1
2 3 3
+ 2 3 3
+ 2 ≤ + +
a +b +c b +c +a c + a3 + b3 25 ab bc ca
1
= .
5abc

4. Consider the following 3 × 2 array formed by using the numbers 1, 2, 3, 4, 5, 6:

   
a11 a12 1 6
a21 a22  = 2 5 .
a31 a32 3 4

2
 Observe that all row sums are equal, but the sum of the squares is not the same for each row.
Extend the above array to a 3 × k array (aij )3×k for a suitable k, adding more columns, using the
numbers 7, 8, 9, . . . , 3k such that

k
X k
X k
X k
X k
X k
X
2 2
a1j = a2j = a3j and (a1j ) = (a2j ) = (a3j )2 .
j=1 j=1 j=1 j=1 j=1 j=1

Solution:Consider the following extension:

 
1 6 3 + 6 4 + 6 2 + (2 · 6) 5 + (2 · 6)
2 5 1 + 6 6 + 6 3 + (2 · 6) 4 + (2 · 6)
3 4 2 + 6 5 + 6 1 + (2 · 6) 6 + (2 · 6)
of  
1 6
2 5 .
3 4
This reduces to  
1 6 9 10 14 17
2 5 7 12 15 16 .
3 4 8 11 13 18
Observe

1 + 6 + 9 + 10 + 14 + 17 = 57; 12 + 62 + 92 + 102 + 142 + 172 = 703;

2 + 5 + 7 + 12 + 15 + 16 = 57; 22 + 52 + 72 + 122 + 152 + 162 = 703;
3 + 4 + 8 + 11 + 13 + 18 = 57; 32 + 42 + 82 + 112 + 132 + 182 = 703.

Thus, in the new array, all row sums are equal and the sum of the squares of entries in each row
are the same. Here k = 6 and we have added numbers from 7 to 18.
5. In a triangle ABC, let H be the orthocenter, and let D, E, F be the feet of altitudes from A, B, C to
the opposite sides, respectively. Let L, M, N be midpoints of segments AH, EF, BC, respectively.
Let X, Y be feet of altitudes from L, N on to the line DF . Prove that XM is perpendicular to
MY .

Solution:Observe that AF H and HEA are right-angled triangles and L is the mid-point of AH.
Hence LF = LA = LE. Similarly, considering the right triangles BF C and BEC, we get N F =
N E. Since M is the mid-point of F E it follows that ∠LM F = ∠N M F = 90◦ and L, M, N are
collinear. Since LY and N X are perpendiculars to XY , we conclude that Y F M L and F XN M are
cyclic quadrilaterals. Thus

∠F LM = ∠F Y M, and ∠F XM = ∠F N M.

3
 We also observe that CF B is a right triangle and N is the mid-point of BC. Hence N F = N C.
We get
∠N F C = ∠N CF = 90◦ − ∠B.
Similarly, LF = LA gives
∠LF A = ∠LAF = 90◦ − ∠B.
We obtain

∠LF N = ∠LF C + ∠N F C = ∠LF C + 90 − ∠B = ∠LF C + ∠LF A = ∠AF C = 90◦ .

In triangles Y M X and LF N , we have

∠XY M = ∠F Y M = ∠F LM = ∠F LN,

and
∠Y XM = ∠F XM = ∠F N M = ∠F N L.
It follows that ∠Y M X = ∠LF N = 90◦ . Therefore Y M ⊥ M X.

6. Suppose 91 distinct positive integers greater than 1 are given such that there are at least 456 pairs
among them which are relatively prime. Show that one can find four integers a, b, c, d among them
such that gcd(a, b) = gcd(b, c) = gcd(c, d) = gcd(d, a) = 1.

Solution:Let the given integers be a1 , a2 , . . . , a91 . Take a 91 × 91 grid and color the cell at (i, j)
black if gcd(ai , aj ) = 1. Then at least 2 × 456 = 912 cells are colored black. If di is the number of

4
 P
black cells in the ith column, then di ≥ 912. Now,
 !2 
91   91 91
X di 1 1 X X
≥  di − di 
1
2 2 91 i=1 i=1
91
! 91 !
1 X X
= di di − 91
2 × 91 i=1 i=1
1
≥ × 2 × 456 × (2 × 456 − 91)
2 × 91
 
91
>
2
 
91
Since there are only distinct pairs of columns, there must be at least one pair of rows (u, v)
2
that occur with two distinct columns s, t. Thus (u, s), (u, t), (v, s) and (v, t) are all black. Thus if the
integers corresponding to the columns u, v, s, t are a, c, b, d respectively, then gcd(a, b) = gcd(b, c) =
gcd(c, d) = gcd(d, a) = 1.

———-0———-

5
 Regional Mathematical Olympiad-2018

Solutions
1. Let ABC be a triangle with integer sides in which AB < AC. Let the tangent to the circumcircle
of triangle ABC at A intersect the line BC at D. Suppose AD is also an integer. Prove that
gcd(AB, AC) > 1.

Solution: We may assume that B lies between C and D. Let AB = c, BC = a and CA = b. Then
b > c. Let BD = x and AD = y. Observe thast ∠DAB = ∠DCA. Hence 4DAB ∼ 4DCA. We
get
x c y
= = .
y b x+a
Therefore xb = yc and by = c(x + a). Eliminating x, we get y = abc/(b2 − c2 ).
Suppose gcd(b, c) = 1. Then gcd(b, b2 − c2 ) = 1 = gcd(c, b2 − c2 ). Since y is an integer, b2 − c2
divides a. Therefore b + c divides a. Hence

a ≥ b + c.

This contradicts triangle inequality. We conclude that gcd(b, c) > 1.

2. Let n be a natural number. Find all real numbers x satsfying the equation
n
X kxk n(n + 1)
2k
= .
1+x 4
k=1

Solution: Observe that x 6= 0. We also have

n n
n(n + 1) X kxk X k|x|k
= ≤ .
4 1 + x2k 1 + x2k
k=1 k=1
n
X k
= 1
k=1 |x|k
+ |x|k
n
X k n(n + 1)
≤ = .
2 4
k=1

Hence equality holds every where. It follows that x = |x| and |x| = 1/|x|. We conclude that x = 1
is the unique solution to the equation.
3. For a rational number r, its period is the length of the smallest repeating block in its decimal
expansion. For example, the number r = 0.123123123 · · · has period 3. If S denotes the set of all
rational numbers r of the form r = [Link] gh having period 8, find the sum of all the elements of
S.

Solution: Let us first count the number of elements in S. There are 108 ways of choosing a block
of length 8. Of these, we shoud not count the blocks of the form abcdabcd, abababab, and aaaaaaaa.
There are 104 blocks of the form abcdabcd. They include blocks of the form abababab and aaaaaaaa.
Hence the blocks of length exactly 8 is 108 − 104 = 99990000.
 For each block abcdef gh consider the block a0 b0 c0 d0 e0 f 0 g 0 h0 where x0 = 9−x. Observe that whenever
[Link] gh is in S, the rational number 0.a0 b0 c0 d0 e0 f 0 g 0 h0 is also in S. Thus every element [Link] gh
of S can be uniquely paired with a distinct element 0.a0 b0 c0 d0 e0 f 0 g 0 h0 of S. We also observe that

[Link] gh + 0.a0 b0 c0 d0 e0 f 0 g 0 h0 = 0.99999999 = 1.

Hence the sum of elements in S is

99990000
= 49995000.
2

4. Let E denote the set of 25 points (m, n) in the xy-plane, where m, n are natural numbers, 1 ≤ m ≤ 5,
1 ≤ n ≤ 5. Suppose the points of E are arbitrarily coloured using two colours, red and blue. Show
that there always exist four points in the set E of the form (a, b), (a + k, b), (a + k, b + k), (a, b + k)
for some positive integer k such that at least three of these four points have the same colour. (That
is, there always exist four points in the set E which form the vertices of a square and having at
least three points of the same colour.)

Solution: Name the points from bottom row to top (and from left to right) as Aj , Bj , Cj , Dj , Ej ,
1 ≤ j ≤ 5.

Note that among 5 points A1 , B1 , C1 , D1 , E1 ,

there are at least 3 points of the same colour,
say, red. (This folllows from pigeonhole prin-
ciple.) We consider several cases: (the argu-
ment holds irrespective of the colour assigned
to the other two points.)
(I) Take three adjacent points having the same
colour. (e.g. A1 , B1 , C1 or B1 , C1 , D1 .) The
argument is similar in both the cases. If
A1 , B1 , C1 are red then A2 , B2 , C2 are all blue;
otherwise we get a square having three red
vertices. The same reasoning shows that A3 , B3 , C3 are all red. Now A1 , C1 , A3 , C3 have all red
vertices.
(II) Three alternate points A1 , C1 , E1 which are red: Then A3 , C3 , E3 have to be blue; otherwise,
we get a square with three red vertices. Same reasoning shows that A5 , C5 , E5 are red. Therefore
we have A1 , E1 , A5 , E5 have red colour.
(III) Only two adjacent points having red colour: There are three sub cases.
(a) A1 , B1 , D1 red: In this case A2 , B2 are blue and therefore A3 , B3 are red. But then B1 , D1 , B3
are red vertices of a square.
(b) B1 , C1 , E1 are red. This is similar to case (a).
(c) A1 , B1 , E1 are red. We successively have A2 , B2 blue; A3 , B3 red; A4 , B4 blue; A5 , B5 red. Now
A1 , E1 , A5 are the red vertices of a square.
These are the only essential cases and all other reduce to one of these cases.
√
5. Find all natural numbers n such that 1 + [ 2n] divides 2n. (For any real number x, [x] denotes the
largest integer not exceeding x.)
√
Solution: Let [ 2n] = k. We observe that x − 1 < [x] ≤ x. Hence
√ √
2n < 1 + k ≤ 1 + 2n.

2
 Divisibility gives (1 + k)d = 2n for some positive integer d. Therefore we obtain
√ 2n √
2n < ≤ 1 + 2n.
d
√
The first inequality gives d <
2n < 1 + k. But then
√
2n ( 2n)2 k2 1
d= = ≥ = (k − 1) + > k − 1.
1+k 1+k 1+k k+1
We thus obtain k − 1 < d < k + 1. Since d is an integer, it follows that d = k. This implies that
n = k(k + 1)/2. Thus n is a triangular number. It is easy to check that every triangular number is
a solution.

6. Let ABC be an acute-angled triangle with AB < AC. Let I be the incentre of triangle ABC, and
let D, E, F be the points at which its incircle touches the sides BC, CA, AB, respectively. Let BI,
CI meet the line EF at Y, X, respectively. Further assume that both X and Y are outside the
triangle ABC. Prove that
(i) B, C, Y, X are concyclic; and
(ii) I is also the incentre of triangle DY X.

Solution:
(a) We first show that BIF X is a cyclic quadrilateral. Since ∠BIC = 90◦ + (A/2), we see that
∠BIX = 90◦ −(A/2). On the otherhand F AE is an isosceles triangle so that ∠AF E = 90◦ −(A/2).
But ∠AF E = ∠BF X as they are vertically opposite angles. Therefore ∠BF X = 90◦ − (A/2) =
∠BIX. It follows that BIF X is a cyclic quadrilateral. Therefore ∠BXI = ∠BF I. But ∠BF I =
90◦ since IF ⊥ AB. We obtain ∠BXC = ∠BXI = 90◦ .
A similar consideration shows that ∠BY C = 90◦ . Therefore ∠BXC = ∠BY C which implies that
BCY X is a cyclic quadrilateral.

(b) We also observe that BDIX is a cyclic

quadrilateral as ∠BXI = 90◦ = ∠BDI and
therefore ∠BXI + ∠BDI = 180◦ . This gives
∠DXI = ∠DBI = B/2. Now the concyclic-
ity of B, I, F, X shows that ∠IXF = ∠IBF =
B/2. Hence ∠DXI = ∠IXF . Hence XI bi-
sects ∠DXY . Similarly, we can show that Y I
bisects ∠DY X. It follows that I is the incen-
tre of 4DY X as well.

———-0———-