Scribd Logo
Upload

Welcome to Scribd!

  • Romanian Mathematical Competitions1999 PDF

    Uploaded by

    Kanchit Saeho
    186 views33 pages

    Original Title

    Romanian Mathematical Competitions1999.pdf

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    186 views33 pages

    Romanian Mathematical Competitions1999 PDF

    Uploaded by

    Kanchit Saeho

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 33
    MIRCEA BECHEANU BOGDAN ENESCU R.M.C. ROMANIAN MATHEMATICAL COMPETITIONS 1999 Societatea de Stiinte Matematice din Romania MIRCEA BECHEANU University of Bucharest Dept. of Mathematics Bucharest, Romania BOGDAN ENESCU National College “B.P. Hasdeu” Buziu, Romania ROMANIAN MATHEMATICAL COMPETITIONS 1999 ©1999, Societatea de Stiinje Matematice din Romania ‘Foate drepturile sunt rezervate Socetiti de Stinte Matematice din Romania, ‘Aceasti Iucrare nu poate fi tradusi sau copiatd, partial sau in intregime, fri permisiunea scrisi a editorult ‘All rights reserved, This work may not be translated or copied in whole or in part without the written permission of the publisher. Distributia se face prin: Distributed by Societatea de Stiinte Matematice din Romania Str. Academici nr.14, 70109, Bucuresti, Romania e-mail: socmat@math.math.unibue.ro ISBN 973-0-00632-6 ‘Tehnoredactor: Bogdan Enescu ‘Tipeirit la tipografia EDITURTI PARALELA 45 Pitesti, str. Fratii Golesti, nr.29, 0300 tel-fax: +(40) 48-645846 SECTION 1 The 50" National Mathematical Olympiad Final round Alba Iulia, April 1999 PROPOSED PROBLEMS 7" Form 7.1. Determine the side’s lenghts of a right triangle if they are integer numbers and the product of leg’s lenghts equals three times the perimeter, : (er) 7.2, Let a, b, c be non zero integers, a #c, such that a ate co 4d? Prove that a” +6? +c? cannot be a prime number, (Stefan Smarandache) 7.3. Let ABCD be a convex quadrilateral with ZBAC = ZCAD, ZABC = ZACD, (ADB a) AB-DE=BC-CE; b) AC? + (AD-AF + AB-AE). E}, (AB > (DC = {F}. Prove that: (Sorin Peligrad) 7.4, In the triangle ABC, let De(BC), Ece(AB), EF || BC, Fe(AC), EG || AD, Ge(BC) and M,N be the midpoints of (AD) and (BC), respectively. Prove that: b) the midpoint of [FG] lies on the line MN. (Eugen Jecan) 8" Form B.A. Let P(x) = 2x° -3x? +2, and the sets A={ P(r) |nEN, 7 1999}, B={ p*+1 |peN}, C= 742 | geN} Prove that the sets AAB and ANC have the same number of stements (Eugen Paltanea, fon Chesca) ohne Vn 8.2. a) Let n > 2 be a positive integer and ¥.% J positive real numbers. Assuming that aba bh ey 2m EY tet had prove that mpm ence le eee a by) Leta, b, cbe positive real numbers such that ab + be +ba <3abe. Prove that : a tbite Zatbte, (Romeo Hlie) 8.3. Let ABCDA'B'C'D' be a right parallelepiped, E and F the projections of A on the lines A'D, A'C, respectively, and P, Q the projections of B’ on the lines A’C' and A’C. Prove that a) the planes (AEF) and (B'PQ) are parallel; sles AEF and B’PQ are similar. » themansis (Sorin Peligrad) 8.4, Let SABC be a regular pyramid, O the center of basis ABC, and M the midpoint of (BC]. If Ne[SA] such that SA = 25-NS and SOMMN={P}, AM=2-SO, prove that the planes (ABP) and (SBC) are pespendiclas (Stefan Smarandache) 9" Form 9.1, Let AD be the bisector of angle A of the triangle ABC. One considers the points M,N on the half-lines (AB and (AC, respectively, such that “MDA = 0, denote by 1(a,5) the positive root of the equation (a+b)x7-2(ab—Dx — (a+b) = 0. Let M={ (@,b)|@b and 1(a,b) < Vab } Determine, for (a, 6)0, g'(a)>0. Prove that there exists ce[a,b] such that the following equality holds: F(b)~ Fa) g(b)- g(a) Fat EO tO - pogo (Cristinel Mortici) 11.4. The function f:IR—->R is derivable and x) x f@) “(Sr (x). for every real number x Prove that fis a polynomial function, with deg(f) < 1. (Mihai Piticari, Sorin Radulescu) 9 12" Form 12.1, Find all continuous functions f R->[1,20) for which there exists aeR and keN such that: Aeyfl2x)...fm) San’, for every xeR and ne’ (Radu Gologan) 12.2. If Gis a finite group, denote by n(G) the number of its elements, ‘and by s(G) the number of its subgroups. Determine whether the following statements are true or false a) Forevery a> 0, there is a finite group G with "= 0, there is a finite group G with. o >a. sl (Barbu Berceanu) 42.3. Let f:ROR be a monotone function and a, 6, ¢ deR with a #0, 0, Moreover, for every x<[R, the following equalities hold: [OP fat = cx b, and re fat =ex+d Prove that f's a polynomial function, with deg(/)=1 (Mihai Piticari) 12.4, Let A be a commutative ring, with O41, having no zero divisors, and A[X] the associated ring of polynomials. For every integer 7, n=2, we define @:A[X]->AIX1, ,(0=/". Suppose the set Metn| neZ, n22, puis an endomorphism of A[XI} is non-empty. Prove that there exists an unique prime number p>0 such that MED PP Ph (Marcel Tena) SECTION 2 Selection examinations for the 40” IMO PROPOSED PROBLEMS First selection examination, Alba Iulia , April 17", 1999 Problem 1. a) Show that it is possible to choose one number out of any 39 consecutive positive integers, having the sum of its digits divisible by 11 ) Find the first 38 consecutive positive integers none of which having the sum of its digits divisible by 11 (Gh. Eckstein) Problem 2. Let ABC be a acute triangle. The (interior) bisectors of angles “B, £C meet the opposite sides in L, M, respectively. Prove that there is a point K in the interior of the side BC such that the triangle KLM is equilateral if and only if ZA=60° (Romeo Ilie) Problem 3. Show that for any positive integer n, the number 5,2(2 Tym (201) pars antl) 5 0 2 Qn is the sum of two consecutive perfect squares. (Dorin Andrica) 4 Problem 4, Show that for all positive real numbers ¥,,%,,...4%,, which satisfy x,x, ..x, =1, the following inequality holds: 1 1 1 + + + <1 n-ltx, 14x, n-1+x, (Gh. Eckstein) 1B Second selection examination, pril 25", 1999 Problem 5. Let x,,*,,....%, be distinct positive integers. Prove that 2n+1 3 Sade ete MG ex te) (Laurenfiu Panaitopol) Problem 6. Let ABC be a triangle, H its orthocenter, O its circumcenter and R its circumradius. Let D, E, F be the reflections of ‘A, B, C across BC, CA, AB, respectively. Show that D, E, F are collinear if and only if OF! RK. {Proposed problem by France, IMO 798) Problem 7. Prove that for any integer n,n = 3, there exist positive integers @,,a,,...,q, in arithmetic progression, and 1 positive integers bays, in geometric progression, such that: b, 1999! (Ciprian Manolescu) 4 Problem 9, Let O. A, B, C be variable points in the plane such that OA=4, OB=2y3 and OC=¥22 . Find the maximum value of the area of ABC. (Mihai Baluna) Problem 10. Determine all positive integers » for which there exists an integer a such that 21 a9. (39" IMO, Jury) Fourth selection examination, Bucharest , May 22™ , 1999 Problem 11. Let a, 7 be integer numbers, p a prime number such that p>laltl. Prove that the polynomial (X)=X"+aX+p cannot be represented as a product of two integer polynomials. (Laurengiu Panaitopol) Problem 12. Two circles intersect in points A and B. A line / that contains the point A intersects again the circles in the points C, D, respectively. Let M, N be the midpoints of the ares BC and BD, which do not contain the point A, and let K be the midpoint of the segment CD. Show that ZMKN=90°. (D. Tereshin) Problem 13. Let 7 > 3 and Aj, A2,...., An be points on a circle. Find the greatest number of acute triangles that can be considered with vertices in these points. (Gh. Eckstein) Fifth selection examination, Bucharest , May 23", 1999 Problem 14. The patticipants to an intemational conference are native and foreign scientists. Each native scientist sends a message to a foreign Scientist and each foreign scientist sends a message to a native scientist. There are native scientists who did not receive a message. Prove that there exists a set § of native scientists such that the outer $ scientists are exactly those who received messages from those foreign scientists who did not received messages from scientists belonging to S (Radu Niculescu) Problem 15. Let X bea finite set with» elements and Aj, Az,...,Am be threo-element subsets of X, such that |AMAJsI, for every i#j. Prove that there exists ACX with |A> [y2n1], such that none of the A’'s is a subset of A. oe Problem 16. A polyhedron P is given in the space, Find whether there exist three edges in P which can be the sides of a triangle. Justify your answer! (Barbu Berceanu) The 50" National Mathematical Olympiad Final round Alba Iulia, April 1999 SOLUTIONS 7” Form 7.1. Determine the side’s lenghts of a right triangle if they are integer numbers and the product of leg’s lenghts equals three times the perimeter. oy Solution. Let a, b, c be the lenghts of triangle’s sides. We have a =b*4", and bc = 3( atbte). Let P=atbte. Then bc=3P and B+e*=(b+c)-2be=(P-a)* It follows that a= P’+a?-2aP-6P, so P=2a+6, that is, a=b-+c-6. We have then: Bir be" 2bo-12b—-126+36 <> be-6b-6or18=0 > 2 -6(c-6)=18 Analyzing the ways in which 18 can be written as a product of integers, we find the following solutions of the problem: (a,b,c) €{(25,7,24),25,24,7),17,8,15),(17,15,8), (15,9,12), (15,12,9)} 7.2. Let a, b, ¢ benon zero integers, a c, such that a ath oF Prove that a? +5? +c? cannot be a prime number (Stefan Smarandache) +b? Trg Since ave, it follows that b*=ae and therefore: a +b?+0%= a tacte"= a"+2acte-b'= (ate)*-b*= (at+e-byato+b). Now, clearly, a’+67+c">3, so, if.a°+6"+c” is a prime number, then only four cases are possible: Solution. The equality “= is equivalent to (a-c)(b*-ac)=0. c 19 (o?-+b*+c*), and, finally: Inthe first two cases we are lead to: a@xb?+c?Ya+-c)+1=0, or (@-1)’H(e-1)*+b"=1, hence a=c=1 In the other cases we obtain. (a+ P+(e+1)+b7=1, hence a= —1. But a=e is a contradiction. 73, Let ABCD be a convex quadrilateral with 2BAC = ZCAD, ZABC = ZACD, (ADA(BC =(E}, (AB. 9 DC = {F}. Prove that: a) AB-DE=BCCE; b) acted (AD-AF + AB-AB). (Sorin Peligrad) Solution, a) Since ACE is an exterior angle for the triangle ABC, we have ACE=/ABC+ZBAC, hence ZACE=ZACD+ZCAD. Fr It follows that ZCAD = ZDCE, so the triangles CED and AEC are CE _DE jimi —==—— , But ZBAC = ZCAE, so we similar. From this, we obtain AE CE get ac. 48 Using the equalities above, we obtain AB-DE = have that AD < AE, so AB-AD < AB-AE. We get that AC’ < AB-AE and, in a similar way, AC’ < AD-AF. By adding these inequalities, we obtain the requested result. 7.4. In the triangle ABC, let De(BC), Ee(AB), EF || BC, Fe(AC), EG || AD, Ge(BC) and M.N the midpoints of (AD) and (BC), respectively. Prove that: 4) the midpoint of [FG] lies on the line MN. (Eugen Jecan) Solution, a) From the hypothesis, since EF|BC, we have that EF _AE similarly, 22 = BE it follows that 2&4 £5 _) BC AB AD AB BC AD b) Let H= EFMAN and Ie(BC) such that HII|AD. Let O=HIOMN and Je(BC) such that FIAD, A, B G NItp JC ‘Now, since EF||BC and N is the midpoint of BC, it follows that H is the midpoint of EF. We have that HI|AD and M is the midpoint of AD, so 0 is the midpoint of HI, hence O is the center of b) Since ZBAC = ZCAD and ZABC = ZACD, it follows 44 parallelogram EFIG, therefore OcGF and GO-OF that the triangles ABC and ACD are similar, hence AC’=AB-AD. We 20 21 8" Form 8.1. Let P(x) = 2x° -3x? +2, and the sets: A={ Pf) |neN, n< 1999}, B={ p'+1 | peN}, C=E @+2\qeN} Prove that the sets AMB and ANC have the same number of elements, (Bugen Paltinea, Jon Chesc) Solution. Observe that P(x)=(x-I'(2x+1)+1, so, if neN and P(n)eB, we can find peN such that (n=? GasD4tl= p? +1, which shows that either ™=1 or 2+] is an odd square. In the second case, there exists KEN such that (2n+1)=(2k+1) hence n= 2 +2k But since 2374231 < 1999 < 23274232, we derive that . ANB= P(n)|n= 2 +2k,0 2 be a positive integer and ¥,34,%.Ja--0%nIn positive real numbers, Assuming that Rpt Xy tex, SM TED be tN De prove that x % Xn bee no Yn b) Let a, b, ¢ be positive real numbers such that ab + be +ba <3abe. Prove that ata te te, S 22 a +b¥+0° Satbte. (Romeo Lie) Solution. a) Suppose, by way of contradiction, that wyteytete, ee MoM Jn Since Rte tte ERT tty by adding these inequalities we obtain 26, eotmea[n eDhea(ne2}e on yak x Ye y, It is known that if a > 0, then a+222, hence from the inequality a above we derive 2(x, +2 +...+%,)>2(% +2) +. x,), Which isa contradiction. b) If ab + be +ba < 3abe then ditstes On the other a*b te hand, a+b+ 3 ote TTT the AM-HM inequality ) Tit abe sothat aebrez>— petetet Tilia be abe Now, we can use a) taking x1=a, x=, xs=a, y, = The conclusion is obvious. 8.3, Let ABCDA’B'C'D' be a right parallelepiped, E and F projections of A on the lines A’D, A'C, respectively, and P, Q the projections of B' on the lines A’C’ and A‘C. Prove that: 23 6) the planes (AEF) and (B'PQ) are parallel, @) the triangles AEF and B'PQ are similar (Sorin Peligrad) 9” Form 9.1, Let AD be the bisector of angle A of the triangle ABC. One Solution. a) Observe that. AE (A'DC) and EF .L A’C. Inthe plate | considers the points M, N on the halflines (AB and (AC, I , , Fespectively, |L A'C and B'Q_L A’C, s0 it follows that EF||B’Q. (A'DO) we have EF the planes (AEF) and Ina similar way we (B'PQ) are parallel i) Clearly, ZEFA = ZPQB' (the sides of the angles ave parallel), On the other hand, the triangles AEF and B'PQ are right triangles. The conclusion is obvious. deduce that AF||PQ, hence 8.4. Let SABC be a regular pyramid, O the center of basis ABC, and Ni he midpoint of {BC]. If Ne{SAJ such that SA = 256 and SOnMN-{P}, AM=2-SO, prove thatthe planes (ABP) and (SBC) are perpendicular. | (Stefan Smarandache) | Solution. Since , it follows that | sO ] AO? +80" _ 25 SA_5_,SN_1 AO +50 >. ently, “* =2 and ===! Sor 9 Consonant sO 3 and 50 “15 1 SP. SP = oS a sa Jso?+ Ao? so * + so 5 N We obtain that xe and since ZNSP = ZOSA, it follows that SP SA the triangles NSP and OSA are similar. But MN LSA and SO L AM, tance P is the orthocenter ofthe triangle SAM. Finally, AP 4. SMI and TBC LAP imply thatthe planes (ABP) and (SBC) ate perpendicular 24 such that ZMDA = 2B and = Prove that: ENDA = £C. Let ADOMN={P} AD’ = AB-AC-AP. (Nicolae Oprea) Solution. From the hypothesis it follows trian; th ACD are similar, hence AD’=AN-AC a: he ingles AND and A PNV¢ Similarly, AD?=AM-AB, so we obtai > , . jin AD‘=AB-AC-AM-AN. Sit wel ave fe * Prove that AD* = AB-AC.AP, it is sufficient to show that AP=AM. jut the quadrilateral AMPN is cycl ZA+ZMPN=180°, therefore the triangles APM and AND ae similar From here we derive the requested equality. = 9.2, For a,b > 0, denote by 1(a,b) the positive root of the equation P(x)=(a+by*-2(ab—I)x ~ (a+b) = 0. Let M= { (a,b) {a #b and 1(a,6) < Vab }. Determine, for (0,5)eM, the minimum value of 1(a,5). (Dorel Mihet) 25 Solution, Denote by 1,8) the other root of the equation. Since we have (a,b) (a,b) <0, it follows that (a,b) < 0. Therefore: ob) ab > PULGBY 20 <> LEVI, hence (a,5)eM if and only if ab > 1 The clements of M for which (a,b) = Jab have the form (a2) » with a = 1, and (ok) uber min_1(a,b)<1. On the a a Ga ab | <9 y(a,b)eM, so 1,5) 2 1, W(a.b)eM. a+b It follows that min 1(a,b)=1 (adem other hand, P(1) =-2 Alternative solution. If we denote by a =tan(a) and b = tan(f ), with 2, Be(0, m/2), then a short computation shows that apm LF, hence (a,b)cM if and only if tan 2 5 < fantaytan(A) , which is 1-7(4,0) 1+0(a,b) equivalent to cos(a +B) < 0, or <0. But this leads to ta,byz 1,hence min t(a,b)=1 anes 9.3, In the convex quadrilateral ABCD the bisectors of angles A and C intersect in I. Prove that ABCD is circumscriptible if and only if S[AIB] + S[CID] =S[AID}+S[BIC], ( S[XYZ] denotes the area of the triangle XYZ). (Vasile Pop, 1. Gavrea) Solution. It is known that the quadrilateral ABCD is circumscriptible if and only if AB+CD = AD+BC. 26 Suppose ABCD is circumscriptible. Then I will be its incenter. If we denote by r the radius if the incircle, then we have: S[AIB}+S{CID] = (AB+CD) = (AD+BC) = S[AID}+S{BIC). Conversely, suppose that S[AIB] + S[CID] ~S[AID]+S[BIC}. We denote by x =d(I, AB) (the distance from Ito AB) and y =d(I, BC) Then x =d({, AD) and y =d(I, CD). It is sufficient to prove that x=y. The equality S[AIB] + S[CID] =S[AID}+S[BIC] is equivalent to ABx +CD-y = AD-x #BC-y, or x(AB-AD) = y(BC-CD). If AB=AD, then BC = CD and it follows that AB+CD = AD+BC. Suppose AB > AD. Then BC > CD and we consider the points A’e(AB) and C’<(BC) such that AD = AA’ and CD = CC’. We have that AAIA’= AID and ADCI = AC'IC, hence ID = IA! and ID = IC’. We derive that IA’ = IC’ ‘The equality x(AB-AD) = y(BC-CD) becomes S[A’IB] = S[C'TB], or IA“-IB-sin(A'IB) = IC’-IB-sin(C'IB), which means that ZA'IB = ZCTB. It follows that the triangles A'B and C'IB are congruent, hence ZIBA’ = ZIBC’ Finally, from this we obtain that x = y. 7 9.4, a) Let a,bER, a 4 é 2P -4* <1 (Vasile Berinde) Solution, First, observe that x =— and y =— * verity the three required relations. We will prove that there are no otter solutions. Suppose there exists another solution, with x > ; Then, from the first relation we deduce that y < : These two inequalities lead to log vlog <2, which contradicts the second relation Analogously, if we suppose that x < + then y >t and the third relation is not verified 10.2. On the sides (AB), (BC), (CD) and (DA) of the regular tetrahedron ABCD, one considers the points M, N, P, Q, respectively. Prove that MN-NP-PQ.QM > AM-BN-CP-DQ. (Nicolae Musuroia) Solution. By the cosine faw in the triangle MBN, it results: MN’ = MB’+ BN’-MB-BN > MB-BN. Similarly, we obtain: NP* > CN-CP, PQ’ > DP-DQ, MQ’>AQ-AM. 29 By multiplying these inequalities we obtain: MNNP2PQ?MQ? = MB-BN.CP-CN-DP-DQ-AQ:AM | On the other hand, we have: 'AM-BN.CP-DQ = BM-CN-DP-AQ, (ehis equality resuts by applying Menelau’s theorem in the triangles OBC with ths secant M-N-T and ACD with the secant Q-P-T) It follows that: MN.NP-PQ-QM 2 AM-BN-CP-DQ. 40,3, Let 2, 22 be the roots of the equation az’ +b2+c=0, with a,b,ceC, a40, and w;, W2 the roots of the equation (atd)z" +(b+b)z+ @+0)=0 Prove that if |2;{<1 and | za{<1, then |wsl=lw2l-1 (Marcel Chirité) Solution. Observe that a+7%#0, otherwise |@}=1|-1¢|, hence we lel have |2221°)5)=! | which contradicts the hypothesis, a Now, w1 is a root of the second equation, hence: (a+eynp +O +b), +@ rep=0 30 It follows that: . @ +c)w? +6 +b)H, +(a+e)=0 and since w; #0, we obtain: , at plug esr Orb—+Gre)=0, L which means that —-is also a root of the second equation. 1 Now, if w) = = 1= Gothen bei=L and, similarly, pal=1. Lf we = 1 wi= =, and i 1 and by direct computation one can show that this contradicts the hypothesis, 104, a) Let m3... %p Me Vos--0Jy ERs" , such that: iti) O< ay, 0, for every k=l, 2... 1-1. It follows that: Bm NO wi “2 Xn SZ (8,2 — Sp aEn = =~S\(2,~2:)- Sy 4) Sys ~ 20) Spy S04 with equality if $= 0, V k=1, 2 This implies that x, =ys, W k=l, 2, b) We can assume, with no loss of generality that O PL, V k=1, 2.2.05 ‘We now apply the property a) for the numbers a1,az,...,@qand 1,2,2%,...,2"", It is easy to check that the conditions (i) and (ii) are fulfilled, hence 32 11° Form 11,1, Let A be a 2x2 matrix with complex entries and C(A) = { Be otl(©) | AB = BA} Show that {det (4+B)| > det), VBeC(A), if and only if A’=0. orel Miher) | Solution, Let 4, and a, be the roots of the polynomial P(A)=det(A-Al), Then we have (A-Arh3)(A~Asls)0. Since Ah and —Aal2 belong to C(A), it follows that = det (4-Ah) | > [AP 71,2, that is A, and 4 are zero. Clearly, it follows that 4-0. Conversely, assuming that A?=0, we will prove the equality det(4+B)=det(B), for every BeC(A). If Bis invertible, then for every xeC we have 0+ det(-B?)= det(x°A?-B?)= det( xA+B) det( xA-B). It follows that the polynomial (x)= det( x4+B) is nonzero for every complex x, whence it is constant, But then Q(0)=Q(1), which | means det( A+B)= det B. i If B is not invertible, one can find a sequence of real numbers Grats With x90, such that det( x/r+B) # 0, for each 21. We have then det( AtxslytB)= det( xnl+B), 33 for each n 21, and letting n—»<0, we obtain det( A+B)= det B. 11.2. Let k be a positive integer, 2), z> ..,24 distinct complex numbers, and 1), us,...24; complex numbers such that the set (a, =m27 +20 +...0427 [2 EN") has a finite number of elements. Prove that one can find peN’, such that a,=dy.), ¥neN", (Vasile Pop) Solution. Since a, takes a finite number of values, it follows that considering all sequences of k consecutive terms of (a), eventually two of these sequences coincide, that is, one can find p, geN such that py Margi tyes Qgik-1~Apigser- We obtain zi (2-1) + yeh (zh I) +... mz (2p -D=0 uyz\(2? -1) + u,29" 2h - 1) +... uyzf "(2p -D 4,28 "\(2f -1) =0 ma ef 1) tye GE —+ Considering the above equalities as a linear homogenous system with ~1 as unknowns, its determinant is Looky =2f ~ ug eet 428 a 1 1 waft ze" =[he Tle-s)+0 . we wae BL Wot 34 hence the system has only the trivial solution 24= x= ... follows that 2? = for every positive integer 7 = x0. It 1, for every iI, 2,...,& and, consequently, ay“aysp, 113. Let fg:[a,b]>R two derivable functions, having increasing derivatives, with f(a)>0, g(@)>0. Prove that there exists c<{a,b] such that the following equality holds: (Cristinel Mortici) Solution. Using Lagrange’s theorem, we obtaiih the existence of the points c), ¢> €(@.b) such that: LO-L0 4 £O-2@) b-a = F'(q) anc ‘The functions fan g’are increasing and have the intermediate values property (Darboux), whence they are continuous. Therefore the function F'fa,b]R, Fex)=ftedg(e)~ £0) is continuous. It is easy to check that F(a)20 and F(6)<0, so there exists a point ce(a,b) such that F(¢)=0. = 8) 114, The function /:R-B is derivable and st= AS)4r0, for every real number x. Prove that fis a polynomial function, with deg(f} < 1 (Mihai Piticari, Sorin Radulescu) 35 Solution. We fix x > 0 and we define the set Me={t|120 and f()=f)}. Clearly, M is bounded from below, so let fo=inf M. Since /”is obviously continuous, if we take a sequence (Fy)nar with fy->fo, it follows that /,)2/éo) and, on the other hand /(x)= fi), whence F=f), which means that f¢M. We want to prove that 4-0. Suppose, by way of contradiction, that f9> 0. But then, since we get (from Lagrange’s theorem) the existence of a point ce (0/2, t) such that /(2)= ls) =f), which is a contradiction (ceM and o 0. Analogously, we obtain fe)= 10), for every x < 0. Now, since /” is a constant function, it follows that fis a polynomial function and deg(/) < 1. 12" Form 12.1, Find all continuous functions f° R->[1,20) for which there exists aeIR and ke N such that: Seoyfl2x)...flrre) Sant, for every xeR and nN”, (Radu Gologan) Solution. The condition fry{2x)...Am) S an* can be written equivalently: Vlog (Ux) 0 and x = adn, we obtain S04] 0 was arbitrarily chosen, so it follows that log fx)=0, Vx20, that is x)=1, Vx20. In a similar way we can prove that fx)=1, Vx<0. 37 12.2. 1fG isa finite group, denote by (G) the number ofits elements, and by s(G) the number of its subgroups. Determine whether the following statements are true or felse: ©) For every a> 0, there is a finite group G with 0, there is a finite group G with ">a 5 (Barbu Berceanu) Solution. Both statements are true: a) IfG and H are finite groups, itis easy to see that s(GxH) > s(G) sD. Indeed, if G' and H’ are subgroups of G and H, respectively, then clearly G’xH’ is a subgroup of GxH. Now, let’s consider the group of Klein K, that is, the group having the following table e fe [u [vy [w e le |x [v [w u_iu le [w |v viv iw fe [wv w_iw iv [w fe This group has 5 subgroups: {¢}, {eu}, {ev}, {ew}, teuvw) Denoting by K"=KXKx...xK (7 times), we have: mk") (2 y -(4 MK") KK) AS and since (3) +0, we are done b) Observe that if p is a prime number, the group (Z,, +) has only two subgroups: {0} and itself, hence n(Z,)/s(Z,)=p/2. Clearly, for every positive a we can find a prime number p such that p > 2a. 38 12.3, Let f:R->R be a monotone function and a, b, c, deR with a #0, ¢#0. Moreover, for every xe R, the following equality holds: LY pears ac+b, and [" pdt =ce+d Prove that /'is a polynomial function, with deg()=1 (Mihai Piticari) Solution. Let us consider the function F:R->R, F(x)=| f(@adt ° Then F verifies: F(c+V3)- F(x) =ax+b and F(x+V2)-F(x)=er+d,WreR ‘We can find two polynomial functions of second degree fi and f; such that fi(x +13) — fi(x) =ax-+b and f,(¢+V2)- fC) =ert+d WreR, From these equalities we can derive that the functions gi=F-f,, i=1,2, are periodical functions, with periods J/3 andV/2 , respectively. It follows that /i(x) - (x) +2,(x)~ g(x) =0, VxeR, But since gy and 2 are continuous and periodical, they are bounded, hence fi -f; must be constant, It follows that gi(x)= gx(x)*c, for some ceR, hence g1 also has the period V2. It is easy to derive from here that g; and g> are both constant. We obtained that F(x) = mx" +nx+ p, for some m,n, peR. It is known that the function F is derivable in each point x in which f is continuous, and that in these points we have F'(x)=2mx+n =fix). The set of such points is dense in R (being monotone) and , on the other hand, if two monotone functions coincide on a dense subset of I, they must be equal. In conclusion, flx)=2mxtn, Vice R; which ends the proof. 12.4, Let A be a commutative ring, with 01, having no zero divisors, and A[X] the associated ring of polynomials. For every integer 7, 122, we define 9: A[X]>ALX], 9,7". Suppose the set 39 M={n| neZ, n>2, g, is an endomorphism of A[X]} is non-empty. Prove that there exists an unique prime number p>0 such that M=OD FP Pied (Marcel Tena) Solution. We have 9,(/8)=9,(/)9,(8).¥ f, SeA[X]. This shows that 9, is an endomorphism of the ring A[X] if and only if g, (f+g) = vat yrig), or, equivalently, (rg) "=f"+g", Vf geA[X]. Since M is nonempty, there is a n for which this holds. Taking f=g ~ 1, we obtain (2"-2) -1=0. Hence the set of integers & =2 for which & -1=0 is also nonempty. Let p be minimal element of this set. We prove that p isa prime. Indeed, if p=ab, with 1< a

    20, the remainders (100-4), r/100-k+1),..., r(99) cover the last two rows of. the matrix, so we can choose one such that r(m-1)+r(100-kti) is divisible by 11. If k < 20, then 39-k > 20, and we apply the same argument for the second part of the list. b) Any 38 consecutive numbers with the requested property have the form: 100(m-1)+81, 100(m-1)+82,..., 100(m-1)+99, (19 numbers) followed by 100m, 100nr+1,..., 100nr+18 (19 numbers). Since r(81), 1(82),..., (99) cover all remainders except 8 and r(0), r(1),..., 78) cover all remainders except 10, we see that we must have r(m-1)=3 and r(m)=1, This is possible only if m-1=10x+9, with r(x)=5 and r(c+1)=1 Analogously, x=10y+9, with r(y)=7 and r(y+1)=1. For this, we must have y=10z+9, with r(z)=9 and r(z+1)=1 and z=10w+9, with r(#)=0 and r(w+1)=1. The least w with such property is, obviously w=0, hence m-1=9999 and the numbers are: 999981, 99982,....,1000018, Problem 2. Let ABC be a acute triangle. The (interior) bisectors of angles 2B, ZC meet the opposite sides in L, M, respectively. Prove that there is a point K in the interior of the side BC such that the triangle KLM is equilateral if and only if 2A=60° (Romeo lie) 44 Solution. Let us first notice that if M is a point on the bisector of the angle XOY and PeOX, QeOY such that MP=MQ, then either ZOPM=20QM, or ZOPM+ZOQM=180° (see figure below). x x Returning to the problem, let’s suppose such a point K exists. Then M lies on the bisector of A ZACB and =ML=MK, hence ZALM=2BKM or ZALM+ZBKM=180° L But in the last case, it M follows thatthe quadrilateral MKCL is cyclic, so ZC=180°-ZKML=120°, which is a contradiction since the triangle ABC is acute. So, we have ZALM=2BKM, and, in a similar way, ZLKC=2LMA. In the triangle AML, we have ZAML+ZALM+ZA=180°. On the other hand, BKM+ZLKC+ZMKL=180°. We derive that ZA=ZMKL=60° Conversely, suppose ZA=60°. Let Ke(BC) such that MK LBL. Since BL is the bisector of 2B, it follows that BL is the perpendicular bisector of MK, so LM=LK. Denote by I the intersection point of BL and CM, Clearly, AI bisects ZA and a short computation shows that 45 2Mi 0°, hence the quadrilateral AMIL is cyclic. Then 0 such that S,=(k-1)+#, or, equivalently, 2W-Ue+1-5,=0. The discriminant of this equation is A=4(25,-1), and, after usual computations, we obtain n+ et aa ey3) tal $ ‘After solving the equation, we find that pa 2 he BP ay" yt Therefore, it is sufficient to prove that is an integer number. Let us denote y=(1+V3)" +(-V3)', where m is a positive integer Clearly, Eq is an integer. We shall prove that ssid Ey. For Eq2, E\=2, E2=8, the assertion is true. Moreover, the numbers Eq satisfy the relation’ Eegi= Dig? 2E ya The property now follows by induction. 46 Problem 4, Show that for all positive real numbers x,,,,...,%,, which satisfy 1,2) ...x, =1, the following inequality holds: 1 1 yt n-1+x, <1 n-1+x, n-1+% i (Gh. Eckstein) " Solution: Let us note x, =a, for K-1,2,...,2. Then ayay...a,=1 and | wo have: iy Mp Aes (n—Daz" mI at apt +.4aty +ani+.tapt @-DSar" at. n=l | Alternative solution. Let P(x) be the polynomial i PO) = (+H) HH) LTH) ‘Then, the inequality becomes ay P'(n-1) s P (n-1). | Let us denote a; = S; (ty ,..., %» ) the fundamental symmetric polynomials in x; ,..., X, , Which are also the coefficients of P: PO) Hx" 4x" +. $a, eta, Hence P'(x)= nx"! +(n-Nax"? +...+a,, and the inequality (1) becomes: n(n-1)"" +a,(n-1)"" +a,(n-2)(n-D <(Qn-l)" +a,(n- I)" +a,(n-1)"? +..44,,2-) +4, After reducing similar terms, we obtain: oa Fit dy ys 47 (n=) s(n 1)"a, +2(n-1)"a, + + +(0-In-3)a, ,+(2-2Da, +a, When x; =.) =...=%,= 1 we have equality in the inequality above and also in the given inequality, that is: 8) opr so-0n(? sae yr { ) +(n- 2? }() n Using AM-GM inequalities, we have: @ so the right-hand side of (2) is greater than the right-hand side of (3) Second selection examination, Bucharest , April 25", 1999 Problem 5. Let x,,x,,....X, be distinct positive integers. Prove that 2n+l Max t..¢a2 > (x $x, +..4%,) ‘(CLaurengiu Panaitopol) Solution. We may suppose x1< x3<...< xq. Then 1S x1, 25 ¥2,...5 7S Xp Let xi=k+ay, for R=1,2,...,2. Since xps-%y 21, it follows that aya> as The given inequality can be rewritten a follows: xe +a,) >(n ye +Onrd Qa, £a;-0 for 1.2... then we get an equality. Otherwise, we obtain the equivalent inequality 330g? +69 ka, 13 2 nt) DK +On+D Va, and, after canceling similar terms, “ “ Lai + % ka, 2(2n+ og a Using Tehebyshet's inequality we have: vee 6) ka, 2 61S = n 6g 30s) a, >On Sa . ne a ci hence shal + She, > anode Thus, the inequality is yrs and the equality holds if and only ifthe given numbers are 1,2,... 49 Problem 6. Let ABC be a triangle, H its orthocenter, O its ‘Lemma: Let a, b be positive real numbers and k be a positive integer. circumcenter and R its circumradius. Let D, E, F be the reflections of Then A, B, C across BC, CA, AB, respectively. Show that D, E, F are =P aCe tks? collinear if and only if OH=2R 5 < see (Proposed problem by France, IMO °98) ifand only if Solution. Let G be the centroid of the triangle ABC, and A’,B' and C’ dxtk+l>e be the midpoints of BC, CA and AB, respectively. Let A"B"C” be the 6 triangle for which A, B and C are the midpoints of BC", C"A" and AB’, respectively. Then G is the centroid and H is the circumcenter of the triangle AYB"C”. Let D’, E’ and F’ denote the projections of O The proof is obvious . ‘The given problem consists in proving that the lower bound of | the function baste aa? on the lines B’C”, C"A" and A”B", respectively. PC px yy.9%,) = ; Consider the homothety h with -center G and ratio -1/2. It Rpm tte ‘maps A, B, C, A", B" and C” into A’, B', C’, A, B and C, respectively. over the set of ordered systems of positive integers 0 <2: <..<2y, is Note that A’D'LBC which implies AD:A’D'=2:1-GA:GA’ and equal to Qn+)) ZDAG=2D'A'G. We conclude that A(D)=D’ and, similarly, h(E) (F)-F’. Thus, D, E and F are collinear if and only if D’, E’ and F’ are . _ (n+l) collinear. Now, D’, E' and F" are the projections of O on the sides First, remark that /(J,2,...y)= “J. The second step Buc CA" and AB", respectively. By Simson’s theorem, they are consists in applying the lemma for an ordered system of the form collinear if and otily if © lies on the circumcircle of the triangle Sp << ay Sat] << ehetL, A"B"C". Since the circumradius of A”B"C” is 2R, O lies on the where 2x22, and taking =x? +..4234(¢-+)? +.+@+k-+0?, ,“itcumeitele ifand only if OH=2R. bax, bate, tt ++ tk ED. It follows that, PO posX Xo VS AU pogkpptheone RED Problem 7, Prove that for any integer n, = 3, there exist 7 positive integers a,,0,,...,, in arithmetic progression, and 7 positive integers » By induction it follows that the lower bound is realized on a sequence of the form x+1, x+2, ..., x4, Applying once again the lemma, we 4,,,,.-.,5, in geometric progression, such that: obtain: b, (UXHL...xtN-D Give an example of two such progressions having at least five terms. From fll,2,..n) Qn+l) (Mihai Balun’) . the required result follows. 50 31 . ! Solution. Let (1-4) wtb HED, 4 Observe m) om Dim id that B, >is, fork zs Then, fork ha it follows: 11 1\1 k+l B, aie +fayte tbat * G 3 Z)b et The last inequality is a consequence of the condition m > n? and the well-known inequality Ur wea ad pt ty AaB "a 12 *33 ke 223 k-L ke k i which holds for k 22. i If we define ack, 1< ka, it follows that: By Ay 1 and a, a+r, a+2r, s a+3r, a+4r be the arithmetic progression. Note that r=a,-a, bs —bs=t* -$F, hence 3s°(1? —s*)>3r>1(F-s*). It follows 3s” >?” > (s41)", so § = 2. Taking s =2, 1=3, a =r =17, we obtain the progressions: 16<17<24<34<36<51<54<68<81<85, 52 Third selection examination, Bucharest , May 15", 1999 Problem 8. Let a be a positive real number and (x,)na1 a sequence of real numbers such that xy=a and 2(n+2x, -Lie. for all n21 ‘Show that there exists a positive integer an such that x, 19991. (Ciprian Manolescu) Solution. We will prove by induction that Son? Lhe ti For n=l, the definition gives x» 2 3x, > x1. Suppose x,.1 >) Ay, and we shall prove the corresponding formula for X».2: Kyun 2M DN pap— Do = 8+ Dg +2 pas De > a a >(tDE 1 +25 Ley =D t Dee a a From this, we deduce Xp > DR > py a By induction: x,,,>"x,>n(r—lx,,>-.> nla. Obviously, for sufficiently large n, we have nla > 19991 Problem 9. Let 0, A, B, C be variable points in the plane such that OA=4, OB=24/3 and OC=22. Find the maximum value of the area of ABC. (Mihai Baluna) Solution. We look for a tetrahedron MNPQ with the following properties: H=prreoM, HN=4, HP=2/3, HQ=V22 and MN, MP, 33 MQ are pairwise perpendicular. If such a tetrahedron exists, then, taking DO.L(ABC), with DO=MH, we get: vol [ABCD ]=(2/3)area [ABD}-CD < (1/6)-AB-BC-CD= =(1/6)-MN-MP-MQ=vol [MNPQl, hence max(area[ABC})=area[NPQ} It remains to find MNPQ. Denote MH=x. It follows that MN=Vx7 +16, MH=¥x +12, MQ=vx? +22, NR=(1/4)(x7+16) (where R-NHMPQ) and, since MR is an altitude in the right triangle MPQ, we have MR-PQ=MP-MQ, so Flee +16y Weise Vera wwe obtain the equation (? -49)8y +2) = (4y-4)(4y +6), which is equivalent to (y-6)(4y? +y-2)=0. We must have y>4, therefore the only solution is y =6, hence x =¥B. It is easy to check that by taking MN=/24, MP=/20, ‘MQ= 30 and MN.LMP1MQLMN, we get the required tetrahedron. Since area [NPQI=(MN-MP-MQ)/(2:MH)=15y2 , this is the required maximum @? 416) Replacing "+16 Alternative solution. We may assume the segment OA fixed and look for the positions of the points B.C on two circles with center O and radii V12,J22, respectively, such that area(ABC) has maximum value (see fig. 1). First, we remark that in such a triangle ABC, the line OA is perpendicular on the side BC. That is a consequence of the following: given a segment BC and a circle of center O, the point of the circle at a maximal distance to the line BC is one of the endpoints of the diameter perpendicular to BC (see fig.2). ‘Thus, we have to consider a variable line /, perpendicular to OA, which intersects the two circles in variable points B, C, 34 respectively and find the position of / for which area(ABC) has maximum value, It is simple to work in coordinates figure 2 ‘The line / has the equation @ yet where rc[-V12, 12]. Considering the altitude AD of the triangle ABC and by symmetry with respect to the line y = 0, it is sufficient to take re[-V12 0]. ‘The equations of the circles are respectively Q) xty=12, @) way? = 22. From the intersections (1) and (2), (1) and (3) we obtain respectively B(~ Vi2—/7,t) and C(--¥22-°, 1). ‘Then area (aBoma-nol2—P 412-7) =f. We study the variation of the function . The derivative is: gy ag a er ~4t LO=>6 ne22-P +Vi2 a is | Its roots are given by the equation’ yo2- P)I2-7) =P — 41 1-424? — 33 +66) = 0. This equation has a negative root = -2 and two positive roots which cannot be considered. Hence, ¢) has'a maximum value for ¢ = —2 and f-2) = 15 V2 Problem 10. Determine all positive integers » for which there exists an integer a such that QL | a9. 39" IMO, Jury) Solution. First we prove that if 2”-1 divides m’+9 for some integer m, then m must be a power of 2. if this is not true, then m has an odd divisor / > 3, and since 2'-1 divides 2"-1, it also divides m’+9. Now, for every 1 > 3 we have 2'-1=-1 (mod 4), and so 2'-] has a prime divisor p=-1 (mod 4). Suppose p # 3. As it divides m?+3* we have, by Fermat's theorem, Lem? a(n)??? =€-9)79? =-1 (mod p) Since this is impossible, if / exists, then p has to be 3. This also leads to a contradiction: if / is odd, then 2' =~1#1 (mod 3) Conversely, we will prove that every power of 2 is a solution of the problem. Observe that ifm =2", then we have 2n—1=3(2? +1) +1). 41) So if 2"-1 divides m?9, then so does 2” +1, for each / =1,2,....k-1. Now, it easy to check that if a # 6, then 2” +1and 2” +Lare relatively prime, By thr Chinese remainder theorem, since the numbers 2” +1are pairwise relatively prime, there exists a positive integer c such that c=2” (mod 2°" +1) , foreach /=1,2,..., k-2. Hence c?+1=0 (mod 2” +1) for / =1,2,...,4-2, and so 2"-1 divides (3c)"+9. Jn conclusion, 2”-1 divides m’+9 for some integer m if and only if n is a power of 2 56 Fourth selection examination, Bucharest , May 22" , 1999 Problem 11, Let a, 7 be integer numbers, p a prime number such that p>laltl. Prove that the polynomial f(X}=X"+aX+p cannot be represented as a product of two integer polynomials. ; (Laurentiu Panaitopol) Solution. Let z be a complex root of the polynomial, We shall prove that >I. Suppose |z| <1 . Then, since 2”+az = ~p, we deduce: p= |2"taz\ = Ele” Fal < le" }Hal = Hal, which contradicts the hypothesis. ‘Now, suppose f=gh is a decomposition of f in integer polynomials. Then p=g(0)A(0). This gives \g(O)=1 or \A()IH1. Suppose [g(0)[=1. If 21, 22... Z are the roots of g then they are also roots of and we have: 1= (gOE| 2122 which is a contradiction. 2u>| zillza)---] zal > 1, Problem 12. Two circles intersect in points A and B. A line 7 that contains the point A intersects again the circles in the points C, D, respectively. Let M, N be the midpoints of the arcs BC and BD, which do not contain the point A, and let K be ‘the midpoint of the segment CD. Show that 7MKN=90°. . (D. Tereshin) Solution. Let N, be the symmetric point, of N with respect to K. Then the triangles KCN and KDN are congruent, Thus CN=ND and ZN{CK=ZNDK~180°-Z ABN. Cc Observe that we also have ZMCK=180°-ZABM. Adding these last two equalities we obtain ZN;CM=ZMBN. Using M the obvious equalities CM-MB and BN-ND=CN,, we obtain that the triangles MBN and NiCM are congruent, thus MNi=MN. In the isosceles triangle NiMN, the line MK is a median, hence it is also the perpendicular bisector of the segment NNj Problem 13. Let > 3 and Ay, Az,...., A,be points on a circle. Find the greatest number of acute triangles that can be considered with vertices in these points. (Gh, Eckstein) Solution, We can assume the points Ai, A2,..., A, are ordered counterclockwise on the circle. We will consider Ay1=A1, An2=A2,. Denote by 4,4; the are of the circle starting from A, and ending in A; (in the counterclockwise direction). We call an arc A:A, obtuse if m(A,Aj)2180°. Obviously, m(4,A,)+m(4,A,)=360°, hence at least one of the arcs 4,4; and A,A, is obtuse. Let x, be the number of the obtuse ares AA; having s-1 points in the interior. Since m(A,4;.,)+m(A;-.4)=360°, we deduce that « Xe— Xp 2M, for every s =1,2,..., m1. The equality holds only if there are no diametrically opposite points A;, Aj... ‘We will count the number of non-acute triangles AAA, Clearly, such a triangle has exactly one angle corresponding to an obtuse are. For each obtuse arc having s—1 points in the interior, there are n-s—1 non-obtuse triangles 4,4,A, for which ZA, corresponds to the arc AA,, namely those with 4,elnt 4,4. It follows that the number Nofnon-acute triangles is N=x(n-2)+4,(0-3)+ 4, 5-24 9 14% 0. By regrouping terms and using (*) we obtain 3 N20 FH) FT Gg $82) + $= Oyen * Maya) 2 2 132) .mo-Bto-D erfl42+. 42 ( 2 8 ifn is odd and 58 =4 N20-(%,) +3) +1-Ga tp) + HF Gena 4 Xig-aya) + n=2 n_nn-2)" oftvae teste nm if nis even. Equalities in the above inequalities are obtained only if there are no diametrically opposite points and x, =0 for k < n/2. This happens, for instance, when m(Ai42) = m(4oAs) =...= m(AoAr) =360°/n, if m is odd and when m(A;A2) = m(42A3) m(AyiA,) = 360°n+6 and m(A,A,) = =360°%n-(n-1)s, with 0 <6 <360%’, ifn is even. In conclusion, the number requested is Dn if nis odd ” - 2 (2) ~) (a-2)nqa +2) if nis even. 24 39 Fifth selection examination, Bucharest , May 23", 1999 Problem 14. The participants to an intemational conference are native and foreign scientists, Each native scientist sends a message to a foreign scientist and each foreign scientist sends a message to a native scientist. There are native scientists who did not receive a message. Prove that there exists a set S of native scientists such that the outer S scientists are exactly those who received messages from those foreign scientists who did not received messages from scientists belonging to $ (Radu Niculescu) Solution. Let 4 be the set of native scientists and B be the set of foreign scientists, Let f: A-»B and g :B->A be the functions defined as follows: f(a) is the foreign scientist receiving a message from a and (6) is the native receiving a message from b. It is given that g is not a surjective function. We have to prove that there exists a subset Sc A such that A-S=g(B-/(s)). Let P(A) be the set of parts of A and h : P(4) P(A) defined as follows: A(X)= A-g(B-/(X)), for every XC A. The function h is increasing. Indeed, if XcY, then (X) c f(¥), so B~ f(Y) < Bf), g(B- f1)) © g(B-fX)), A-gB- fO0) < A-g(B-fY)) and, finally, AMO hY) Let M= {XA |H(X) CX }. The set M is nonempty since clearly A © M. Let S= [|X . From the definition of S' we have that xewt (5) < S. From the monotony of hit follows that (((S)) < h(S). Thus 1(S)eM and therefore h(S)=S. 60 Problem 15. Let X be a finite set with n elements and Aj, Az,...,An be three-element subsets of X, such that |AMAJsI, for every i#/. Prove that there exists ACX with |Al>|V2n], such that none of the A's is a subset of A. cers) Solution. Take a subset Ac X with a maximum number of elements which contains no A;’s and set k=| A |. Notice that k > 2. By the maximality of A, for every xeX-A there exists i=i(x), 1 Ajgl=2. Denote AM Ayai=Le Since | A/~4, £1, for every #4, all the Z,’s are distinct. Therefore, the mapping which sends x to L, is one-to-one. So we find: noks(S)n-tsAoD este anual te] Problem 16, A polyhedron P is given in the space. Find whether there exist three edges in P which can be the sides of a triangle. Justify your answer! (Barbu Berceanu) Solution. The answer is afirmative. We will need the following Lemma. Let x, 2,.... n be positive real numbers such that x I, then x, xand xy form an isosceles triangle ) and x3.+%/$ x1, for i~3,..., n. Hence the numbers 2,...x, verify the hypothesis of the lemma. Consider now the two faces of the polyhedron sharing the side of length xy. We have then or ee yA, tote, ey with i, 2 be prime number such that 3 divides p-2. Let S={y2°-1 | x, y are integers, 0< xy

    You might also like