PROBLEMS AND SOLUTIONS I N

MATHEMATICS

Major American Universities Ph.D. Qualifying Questions and Solutions

PROBLEMS AND SOLUTIONS IN

MTHEMTICS
Compiled by:

Chen Ji-Xiu, Jiang Guo-Ying, Pan Yang-Lian, Qin Xe-Hu, Tong Yu-Sun, Wu Quan-Shui and Xu Sheng-Zhi

Edited by:

LI TA-TSIEN
Fudm University

orld Scientific
NewJersey London Singapore Hong Kong

Published by
World Scientific Publishing Co. Pte. Ltd.

5 Toh Tuck Link, Singapore 596224 USA ofice: Suite 202, 1060 Main Street, River Edge, NJ 07661 UX:office: 57 Shelton Street, Covent Garden, London WC2H 9HE

Library of Congress Cataloging-in-Publication Data Problems and solutions in mathematics I [edited by] Li Ta Tsien. p. cm. Includes bibliographical references. ISBN 9810234791 -- ISBN 9810234805 (pbk) 1. Mathematics -- Problems, exercises, etc. I. Li, Ta-chen. QA43.P754 1998 510,76--dc21 98-22020 CIP

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

First published 1998 Reprinted 2002,2003

Copyright 0 1998 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof; may not be reproduced in any form or by any means, electronic or mechanical, includingphotocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.

This book is printed on acid-free paper.

Printed in Singapore by Uto-Print

V

PREFACE

This book covers six aspects of graduate school mathematics: Algebra, Topology, Differential Geometry, Real Analysis, Complex Analysis and Partial Differential Equations. It contains a selection of more than 500 problems and solutions based on the Ph.D. qualifying test papers of a decade of influential universities in North America. The mathematical problems under discussion are kept within the scope of the textbooks for graduate students. Finding solutions to these problems, however, involves a deep understanding of mathematical principles as well as an acquisition of skills in analysis and computation. As a supplement t o textbooks, this book may prove to be of some help to the students in taking relevant courses. It may also serve as a reference book for the teachers concerned. It has to be pointed out that this book should not be regarded as an allpurpose troubleshooter. Nor is it advisable to take the book as an exemplary text and commit to memory all the problems and solutions and make an indiscriminate use of them. Instead, the students are expected to make a selective survey of the problems, take a do-it-yourself approach and arrive at their own solutions which they may check against those listed in the book. It would be gratifying to see that the students can work out the problems on their own and come up with better solutions than those provided by the book. If the students fail to do so or their solutions may turn out to be incomplete, it may reveal the inadequacy of their knowledge or approach, thus spurring them to greater efforts to promote their skills. The very purpose of the authors in writing the book is just to help the students to discover the truth by trial and error. This book was inspired by Professor K. K. Phua’s proposals. We are particularly grateful to him for his support. We also wish to thank Dr. Xu Peijun, Professors Zhang Yin-nan, Hong Jia-Xing and Chen Xiao-man for their painstaking efforts to collect test-oriented data. For selecting problems and providing solutions, we wish to acknowledge the following professors respectively: Wu Quan-shui (Part I), Pan Yang-lian (Part II), Jiang Guo-ying (Part

vi

111), Tong Yu-sun, Xu Sheng-zhi (Part IV), Chen Ji-xiu (Part V) and &in Tie-hu (Part VI). We are also indebted to Professor Guo Yu-tao for carefully reading and correcting the manuscript. Finally, we pay tribute t o Dr. Cai Zhi-jie for printing out the manuscript.

Li Ta-tsien Department of Mathematics Fudan University Shanghai 200433 China

vii

CONTENTS
Preface ................................................................ (v) Part I. Algebra ......................................................... (1) 1. Linear Algebra ................................................... (3) 2. Group Theory ................................................... (26)

3. Ring Theory ..................................................... (44) 4. Field and Galois Theory ......................................... (59) Part 11. Topology ...................................................... (81) 1. Point Set Topology .............................................. (83) 2. Homotopy Theory ............................................... (99) (118) 3. Homology Theory .............................................. Part 111. Differential Geometry ........................................ (151) 1. Differential Geometry of Curves ................................ (153) 2. Differential Geometry of Surfaces ............................... (171) 3. Differential Geometry of Manifold .............................. (194) Part IV. Real Analysis ................................................ (229) 1. Measurablity and Measure ...................................... (231) 2. Integral ........................................................ (256) 3. Space of Integrable Functions ................................... (283) 4. Differential ..................................................... (302) 5. Miscellaneous Problems ......................................... (322) Part V. Complex Analysis ............................................ (333) 1. Analytic and Harmonic Functions ............................... (335) 2. Geometry of Analytic Functions ................................ (360) 3. Complex Integration ............................................ (377) 4. The Maximum Modulus and Argument Principles ............... (413) 5. Series and Normal Families ..................................... (433) Part VI. Partial Differential Equations ................................ (455) 1. General Theory ................................................ (457) 2. Elliptic Equations .............................................. (472) 3. Parabolic Equations ............................................ (496) 4. Hyperbolic Equations ........................................... (513) Abbreviations of Universities in This Book ............................ (539)

Part I
Algebra

.

(a) Prove that m(t)divides ml(t).I. 1102 Let V be a finite dimensional vector space over a field K . that is. So c 3 5 dimR(V) = dimnR[X]/(f(X)) = degf(X). . there is a non-zero subspace W of V . W and V/W respectively. such that T ( W ) W . This is a contradiction. Then V E E [ X ] / Ifor some maximal ideal of B[X].such that T ( W )C W .Xi E B[X]and v E V It is clear that a subspace W of V is an B[X]-submodule of V if and only if T (W ) W . ml(t)..and rnz(t) be the minimal polynomial of S as linear transformation of V. Let S be a linear transformation of V into itself. W # V . Now suppose V is a simple IR[X]-module. Let m(t).D. (Indian u) Solution. W # V .3 Section 1 Linear Algebra 1101 Let V be a real vector space of dimension a t least 3 and let T E Endn(V). Hence V is not a simple B(X)-module. This implies that we have an irreducible polynomial f ( A ) with degree 2 3 in B[X]. SW C W ) . there exists an irreducible polynomial f(X) of B[X]such that I = (f(X)). r n 2 ( t ) . Let W be an invariant subspace of V (that is. Prove that there is a non-zero subspace W of V . Make V into an B[X]-module by defining for i=O w = T ( v ) for all w E V . Since R [ X ] is a P. Thus a.

As usual.Then we have m(t)= ml(t). rnl(t) divides m(t). Let m(X) be the minimal 3 . since m(t). (b) Since m(t)E AnnK[t]V G AnnK[t]W = (m1(t>). ) = Hence m(t)divides m l ( t ) m2(t). v . z = f ( T ) (x). Then it is clear that ( m ( t ) = A n n ~ p l V ’(. ml(t). mz(t). (a) Since m1(t). v c m1(t). Since ml(t)and mz(t)are relatively prime. V can be viewed as a module over the polynomial ring l R [ X ] via f ( X ) . T/ can be viewed as a K[t]-module via the linear transformation S .for any f(A) E lR[X] and z E V. m2(t)divides m(t). (Indiana) Solution. since {Taw I i 2 0) spans V by (b).m2(t). W = 0. Show that V doesn’t have proper T-invariant subspace. Then it is clear that m(t)= ml(t) = m2(t)= t 2 1. (c) Give an example of a case in which m(t)# ml(t). (c) Let W be a 2-dimensional vector space over the field Q of rational numbers and S : W -+ W be a linear transformation with minimal polynomial t2 1. (In i m a) d Solution.4 (b) Prove that if ml(t)and mZ(t)are relatively prime. Since W is an S-invariant subspace of V. m2(t).m2(t) divides m(t). Similarly. a cyclic module. Let V = W @ W and S : V --t V be the natural extenaion of S to V. ~ ( t m ( t ) E A n n ~ p l V (m(t)). + + 1103 Let V be a finite dimensional vector space over l and T : V -+ V be a R linear transformation such that (a) the minimal polynomial of T is irreducible and (b) there exists a vector u E V such that {TivI i 2 0) spans V.ml(t)and mz(t)are all monic polynomials. then m(t)= m1(t)* m2(t). Then we have V = lR[X] . So m(t)# ml(t)* m2(t) in this example. W is a K[t]-submodule of V.m l ( t ) ) A n n ~ p l W ) = and (m2(t)) = A%[t] V l W .

Jz . .5 polynomial of the linear transformation T : V Since m(A) is irreducible. . Show that A has n distinct eigenvalues in C if and only if A commutes with no nonzero nilpotent matrix. 1104 Let A be an n x n matrix with entries in C. ) where Bi(i 2 2) = 0 E Me. Then JIB' = BIJl and Bl(#0 ) is nilpotent. . . Sufficiency. A are distinct and .. . Suppose that A has n distinct eigenvalues A'. Then there exists an invertible n x n matrix P such that PAP-' = diag{ A 1 . B'. we have B # 0. Since All Az. .l . Let B' = diag(B1. where P is an n x n invertible matrix and J. (Inddana) Solution. But the nilpontency of B implies that P ' B P is nilpotent. . . . Then m(A) E Annwfxl(v). . .(C).. If A commutes with some nilpotent matrix B . Xa. V does not have proper T-invariant subspace.. Let B1 be the Jordan block of order e l . . Necessity.J t ) be the Jordan canonical form of A and PAP-' = diag(J1. Hence we have P-'BP = 0. Then B' .l B ' P . PAP-' = P A P . we may assume that el > 1 (If all the ei = 1. =v R[X]/(m(A)) R[A]* E 21 (we may assume that V # 0). ... An}. PAP-' = diag{X1.l B P is a diagonal matrix. we have t o show B = 0. So V is an irreducible R[A]-module. B . .. and A B = B A . . Jt). Taking B = P . Jz. then it is easy to see that A commutes with some nonzero nilpotent matrix)..Xn) commutes with P-'BP.. Thus. which is nilpotent. . A in C.. P . . Without loss of generality. Let diag(J1. We have to show that A commutes with some nonzero nilpotent matrix. Bz. So B = 0. with 0 on the diagonal. . Suppose that the characteristic polynomial of A has multiple roots.. is a Jordan block of order ei.. we have -+ V .

. . . .(X) are monic of positive degree and d. ..dl(X).D.(X)} where the d. .>o. the minimal polynomial m(X) of T is det(X1 . . So at least two of them coincide. . Here.I>~ o of V over F[X].(F[X]) has the form diag{l.> coincide.} of V over F . V = F[X].d. .} and thus induce a linear transformation . z l $ F [ X ] . . there exist z. U be a base for . z z ~ . . .. > are non-zero submodules { .zl has exactly two non-zero submodules. A on V .. Obviously.Let { U ~ .20 and { T Z 3 } . In general. matrix a normal form for X I . r 0 (Stanford) Solution.(X) if i 5 j.A ) .(z = 1 . s ) E V such that V = F [ X ] .(X) = det(X1. T E T }and { T * 8 } . A = ( u ~ ~ ) the ~ ~ of T relative to the base..(z) any z E V ..Show that if %'.A ) .. T : V -+ V a linear map such that the minimal polynomial of T coincides with the characteristic polynomial.) = where (d*(X)).A in M.{Ta?}. Show that A is diagonalizable. f ( A ) 6 F[X]. which is the square of an irreducible polynomial in K [ T ] . V can be viewed as a module over the polynomial ring F[X] simply by f(X).. the three subspaces generated by the sets {T'Z'}.T and 3 are any three non-zero vectors in V . + Hence s = 1 and v = F[X] z1 E?! f F[X]/(m(X)) is cyclic. .6 1105 Suppose V is a finite dimensional vector space over a field K . (HU7Wrd) .vn}.(X)(d. so m(X)= d. according to the assumptions. 2 . Let CT be a permutation on {q. v. . . l.z = f(T). . ~ F [ X ] . Ann(z. By the structure theory of finite generated modules over P. . 1106 Let V be a finite dimensional vector space over with basis {wl. U ~ . Since m(X) is the square of some irreducible polynomial. z . be .I. . then at least two of the three subspaces spaned by the sets (T*-?it}z>~. .

.. .s 1 (io = 0..dl(X).. and thus M is diagonaliz able. .2.(X)} be the normal form for X I . Suppose T is a non-singular linear transformation of V such that T-' = T2+T.1 is the minimal polynomial m(X) of T . . Thus X3 X2 . ' we assume that when u is expressed as the product of disjoint cycles (This decomposition may have 1-cycles)... . ~ ~ ~ . Now let n be the dimension of V over Q. .+ M = . Let Wj be the subspace of V generated by {Kj-l 1. The matrix of A relative to the base { V ~ . V . . @ Ws+l. . . X3 + X2 .--...d.7 Solution... . . V ~ . . ( H U 7 V d ) Solution. . V . without loss of generality.w2. . Hence.. . . is+l= n).. T is annihilated by the polynomial X3 X2 . . ) is . diag{l.. . .-.~ + 1 . vof~Wj overt.. . V ~ . .. . .. we may assume that CJ is the n-cycle (v1... Then M = i } diag{M1..wn).* . By re-ordering the elements 211.1 is irreducible over the field Q of rational numbers. . .1. 'u2. . .1.. . This completes the proof that A is diagonalizable. 1107 Let V be a finite dimensional vector space over the field of rational numbers.. } . Prove that 3 divides the dimension of V. 0 1 0 0 1 0 0 1 1 0 0 *. l. Since T-' = T2 T.-. .. .A A . . ~ ~ . 1.Ms+l} is the matrix of A relative to the base ( V ~ . .. . . . . vij} + for j = 1. A be the matrix of T relative t o + + + n-s some base of V . .. So it suffices to prove that every M i is diagonalizable. .. Mj be the matrix of Alwj : Let Wi -+ Wj relative to the base { ~ i ~ .. Then the Wj are invariant subspaces of A and V = W1 @ W2 @ . and prove that if dimV = 3. Obviously. 0 0 It is easy to see that the minimal polynomial of M is A" . then all such T's are similar. .u. .

.A ) = d l ( X ) .1) and the minimal polyE nomial m(X)of M is A" . M= I. 1. d .M ) . Let 1 : Fq -+ Fq be the Frobenius automorphism II(z) = z p .A ) ) = n..1 = det(XE . .8 where the &(A) are monk of positive degree and d i ( X ) l d j ( X ) if i 5 j. 1 0 a . . there exists a u E Fpsuch that {u. Suppose that 1 is diagonalizable as a linear map over Fp. . On the other . ..1 are in Fp. X3 rational canonical form for A (or 7') is T ' s are similar when dimV = 3. .1 has no multiple root.).1.. = &(A) = X3 + X 2 . II"-'(u)} * is 0 1 0 . II(u)... + X2 - 1).) (Hurvard) Solution. .. .A" . By the Normal Base Theorem. ( X ) . I12(u).I12(u). 1 II(u).. The normal form for X . 11: . 1.l"..112. . It follows that all the 1108 Let Fq be a finite field with q = p" elements. . . 3 .. It should be "n divides p .A is equivalent to diag(1.1...1. the matrix of 1 relative to the base {u. When II is considered as a linear map of Fq over Fp. Then m(X) = 1 A" . and all the roots of A" . The . If dimV = 3.II"-l(u)} is a base for Fq over F p .:. Since det(X1. It is wellknown that Fp C Fq is a Galois extension with Gal(Fq/Fp) = {1. Prove that II 1 considered as a linear map over Fp is diagonalizable if and only if n divides p" . By the irreduciblity of &(A) = m(X)= X3 + X2 ..IIn-1}. .. (Here is a misprint.11. we have &(A) = &(A) = .M is diag{ 1 . Thus we have proved that 3 divides the dimension of V .I. where p is a prime.. then X I . s = deg(det(XI .

ad)(.1.Lli)seelements are differentiable functions of real variable t . 1109 Let A(t) be a non-singular matrix r. .(A .1) . . So IT is diagonalizable as a linear map in over Fp. Let A'(t) denote the matrix formed by the derivatives of the elements. Hence M is similar to diag{ 1.1 has no multiple root and A" - 1 = (A .9 hand.' ) ~ } Mn(Fp). u ( ~ .a("-l)d) where d = and a is the generator of the group F.a d . .1. Thus n divides Conversely.. (A . Show that the derivative of the determinant det A satisfies Then and .\ . - 1 forms a subgroup of FJ = Fp\{O}. all the root of An p . if n divides p .a2d). . A" .. .

. Define an inner product on V by + bz + cz2 with real (a) Find a n orthonormal basis for V consisting of polynomials 4o(z). and 2. respectively.. A-’). having degree 0. Hence we have proved that d -(det A) = det A . . b.) .1. trace(A’ . and q52(z). 41(z). and c. (b) Use the answer to (a) to find the second degree polynomial that solves the minimization problem (Courant Inst.10 = det A1 + det A2 + . dt 1110 Let V be the vector space of polynomials p ( z ) = a coefficients a. + det A.

2).4i(. Hence P O ( % ) = $ z [:ix3 . 9 and 175 cg = 0. 2 ) .))= C l (i = 0 . (a) Since 1.) = z3 . 8 3 . (b) Let and QO(Z) = z3 -Po(. 1 and 2 respectively as usual. we can orthonormalization 1. By an easy calculation. Then for any p(x) E V .$a(z:)) = ( p o ( z ) .1. 1 .) such that ~ o ( z is orthogonal to V . we get co = 0.p o ( z ) .5 z ) 2 d x = -.)) =0 (i = 0.$i(.x 2 is a base of V . Obviously. 3 .x. rl ( .11 Solution.Zz.x.x 2 to get an orthonormal basis of V with degree 0. that is ) ( x 3 . c1 = and no(.

.... Xn Y = ( . z.. yn).. Suppose that all the entries of A. (a) Show that there exist numbers a and b so that det(A + m y * ) = u + bs.**. . and y are real.12 Thus. Yn y*=(y1... ann * ZlYl . X1Yn det a21 . * .. . . and 1111 Suppose that A is a n n x n matrix and that 2 = ( :). ”’ . 3 det(A + szy*) a11 an1 = det + .. when p ( z ) = gz. SXIYI + .)...... . (a) Directly. = det +s . [[ Un1 . * * aln ann + . .. ( c ) Is it true that a = 0 if det(A) = O? (Courant Inst.. a nn a2n +det anl .. + Sz1Yn SznY1 SXnYn .) Solution. is minimal.. (b) Show that if det(A) # 0 then a = det(A) and b = det(A) y*A-lz.

. ann + .. .. we have a = det(A) and + + b = det . a n n . * . . Ann All ) . . . (c) If det(A) = 0.. ...... ..). ( If det A # 0. n X n Yn n where A. . for any s... det(A + sxy*) = a bs as in (a). Ann = y*(det(A) .. . All . Aln ...... is the algebraic cofactor of A-l Thus 1 det(A) ___ aij.. anl *.. * * An1 Aln ..... . a = det(A) = 0.In . . .. An1 . y*A-lx....13 Hence det(A + sxy*) = a bs for some a and b...+ det n an-11 a n . .... ... A-')Z = det(A) ... . + det .. .... ... (b) Since for any s ..

and For any v E a". . v z . Define 1 sin(A) = A .l d i ag ( l . .. f n ) . Assume that XI = . . f z .. . e n ) be the standard Let n-cr. v 12-00 P .fn} is a basis of ker(T)... A... .Y.) = lirn X . . (Courant Inst. .. n+w n-cc orthonormal basis of (En. Define lirn T : C" + an by T ( v ) = lirn Anv. Then the matrix of T with respect to this basis is P ' . XI. V Z . dim(ker(T)) = n . .l ..a.. 0 } . Since A1. n-cc lim A'lw = = lim P d i a g { X .O)P'P'-l is the matrix of T with respect t o the basis ( f l .. f. diag{l.. v n ). f z . P isaninvertible matrix in illn(@) P .14 1112 Let A be an n x n real matrix with distinct (possibly complex) eigenvalues. 1113 Let A be a matrix. . .) Solution. . and corresponding eigenvector V ~ .-A3 3! 1 + -A5 5! * . w (Since lim (X.O. . X ~ } . .-. Let fn en Then ( f l . P . . and dim(Im(T)) = 1.. Hence {fl} is a basis of Im(T).. P ' .... . Prove that n-oo Anv exists..A2. e2..O) . lim Yn). P-' . . For .) is a basis o f a n and diag(l. 5 j 5 n. . . ...O.I A P = diag{A1.O . .. .. Find the dimensions of the kernel and 11-00 1 and that J A j l 1 for 2 < image of T and give basis for both.. {f2.Xn} .Xn aredistinct. . X ~ ....O) = P'-P'-l -diag(l..O.Az. Xz.. ... v. .. (el. Let P = ( v l .1.

7 and 1 -1 0 1 0 1 -1 0bviously.) Solution. (Courant Inst.L 3!T ) 3 . 1' )-'=( 1 ).[. Denote B = ( -3 Then B is similar to -") ..15 express sin(A) in closed form. 103 O ( 1 -1 ). Hence sin(A) - ( . ( 4 3o 0 ( 4 1 1 ) where .(4 0 10 ) .

Taking Y T = ( 1 .(B) be the largest eigenvalue of B . . this is.. -2 1 1 -2 All entries not shown are zero. (a) Show that (b) Use part (a) to show that A has all negative eigenvalues.I . ~ 9 ) I zi E IR} 5 Elo. 0 .2. (d) Let Vl = { ( O . . Y T B Y 5 0 and Y T B Y = 0 if and only if Y = 0.1 .B)Y >_ 0. (Courant Inst.I .. Show that A. which agrees with A except in the first row and column: 1 -3 -3 -2 B= 1 1 -2 1 1 *. (b) Since z T A z 5 0 for all z E lR9 and z T A z = 0 if and only if z = 0.YTY. > 1.. all the eigenvalues of A are negative.(B) > 1.YT(X. (c) Obviously. Then for any Y E Elo. (c) Let B be the following 10 x 10 tridiagonal matrix.. Let A. a . . *. z 1 . If B has more than two nonnegative eigenvalues. Then for any Y E V1. Let A1 and A 2 be two of them. Y T B Y 5 X. (d) Show that B has 9 negative eigenvalues. .. 0 .. Thus A . ... . 0 ) . we have Y T B Y = 5 5 X... . 1 A= **. we have y1.. (a) A direct verification shows that (a) is true.B is semi-definite positive.16 1114 Let A be the 9 x 9 tridiagonal matrix '-2 1 -2 1 1 . the symmetric matrix X. .. y2 E R I O such .) Solution.

. L ) z l 1x1 .eZ). z the subspace spaned by y1 and y2. P-lxI 1x1 - 2EV-IO) min If V is a 2-dimensional subspace.-. z + + + 1 there exists 0 # y E V n VZ.l V is also two-dimensional. . positive-definite. we have min 2EV-10) lTxl - - min Z€V--IO) IP'-ldiag(Xl. 2 ) . Az. Hint.A. yTBy = a?A1 u. * A 2 2 0. .-.A2. P-lxI IP-1x1 . there exists an n x n orthogonal matrix P such that P'TP = diag(A1.a contradiction.) Solution. Then min x €V .x1 when V ranges over all two dimensional subspaces of lR".. . (Courant Inst.. . Thus B has 9 negative eigenvalues..) 1x1 . You may wish to express T in a basis of eigenvectors. A. Then for any y = ulyl uzy2 E V2. Hence max v min Z€V-{O} w = m a x 1x1 v min Z€V-{O) Idiag(A1. Since dim V1+ dim V = 9 2 = 11 > 10..y2).) 1x1 Idiag(Al. y1 y1 = yTy2 = 1 and Byi = Aiyi (i = 1 . .).. > A. Show = by showing 5 and 2 separately. Since for any orthogonal matrix H and any y E B". ..Xn).17 T that y1ly2. Then we have yTBy < 0 (y # 0. 1115 Let T be a real symmetric. e 2 .. . By assumption. (Hyl = Iy1. P . Show that A2 ITXI = max min v 2EV-{O} 1x1 ' where V ranges over all two dimensional subspaces of Bnand 1x1 is the Euclidean norm of 2 . Let V = (yl. y E V1) and yTBy 2 0 (y E VZ). n x n matrix with distinct eigenvalues A 1 > A 2 > .. and Now let {el.. . V = (e1. en} be the standard orthonormal basis of Rn.. .A.to1 Idiag(Al.. . .

and Thus when V ranges over all two dimensional subspace of B"..-. .XI so. fz .2 Az. e.fi. for any two dimensional subspace V of 1R". Let (fl. .. . and if 0 Idiag(A1.18 = A2 (A. . . . A )..fz). > A 2 > 0). . 1x1 #x + azfz E V .fn} of IR" such that V = (fl. .. .. max min Z€V-{O) ITX . ez. . fn IT = Q(e1. x n is an orthogonal matrix. = alf1 Then Q = ( q i j ) . . we can find an orthogonal basis {fl. It follows that I v 1x1 On the other hand. lT.

r j m ( X ~ ) . ) n eigenvalues of m ( the sm= E + P + P 2 + . + z". (c) Use part (a) to compute the inverse of 1 A = [ ! i . be the n eigenvalues of P and ' . 1 0 0 (Courant Inst.P)-'ll 5 *.19 1116 For any n x n matrix P . ~ ~ ~ . A. * Then ~ m ( A l ) .) Solution. ~ are X . we claim that the norms of all eigenvalues of P are < 1. . (a) First. Prove that ll(I . Q-lpQ= be the Jordan form of P. + P" Since C k=O 00 IJPk(l co and the norms over vector space are all equivalent. = k=O (a) Prove that if k=O JIPklJJ00.' .P ) .. then (I-P)-' < (b) Assume that IlPll < 1 in a matrix norm induced by a vector norm.. . * ) rjm(z)= 1 + z + * . Denote ( A1 e f . < 00 k=O . Let A Az. . consider the sum R(P)= M c 00 Pk exist and R(P) ( I . r).

C m=O W is convergent (i = 1 .P). . that is.P)-' = (I-P)-'. llull < llull for the corresponding vector norm.P)-' = I + P ( I .u 0 for some vector u. m-+m lim Sm = lim ( I . Since m-w lim Q . Then llull = IlI-uII = IIPuII.O)(I. m-w 1 Q.P ) is invertible. for any non-zero vector u..~ ( lim S. Now suppose ( I .(I-P) rn-00 = I-P"+'. and we can write ( I . . Since llPll < 1.P"+')(I - PI-' I . Since S.llPull 5 llPll .1 . and (I.I J .P are non-zero.~ s . l l ~ l l =sup { #lv non-zero vector I . (b) By definition. Then we have .lim ~ + = n-rn = (I .) exists. Q = Q . . Thus all the eigenvalues of I . The matrix I . Hence pal < l ( 1 5 i 5 n). m-oo lim dm(X. Thus 00 ( l( I ) PI-' R ( P )= k=O Pk = ( I .P is therefore invertible.20 is convergent. 2 . = We must have u = 0. n ) .P)-'.

Then G' is isomorphic to a subgroup H of G L ( m . By Sylow's Theorem. Then G' is a pgroup. ( CoZumbia) Solution.Fpn)such that PHP-' U .N .F p n ) . . if we change the base of V via P . which are unipotent matrices. where m = dirnp. Thus. So 1117 Let G be a pgroup and G x V + V be a linear action of G on a finite vector space over a finite field Fan. the corresponding matrices of the linear transformations in G' are in U . there exists some P E GL. V . Using Sylow theory.(m.w --+ 0 * 01g E G}. prove that there exist a basis of V in which all the transformation v + a .21 Then A = I . Let G' = {V + V .Fan). v(a E G) are unipotent matrices. We fix a base of V over Fan. since there is a surjective homomorphism G -+ G' ( 5 EndF(V)). It is well known that m j=1 and U= is a Sylow psubgroup of GL(m..

. where S is diagonal and N is strictly triangular. let d l ( X ) . dimF F [ S ]< dimF On the other hand. ) (Stunfod) Solution. the dimension of the i vector spcae over F of matrices commutative with the matrix of S is v. all of whose (complex) eigenvalues are real and positive. dimF F [ S ] = degrn(X). (Stanford) + .j=1 + 1)nj. . there exists a t least one n x n real matrix B with B" = A . . Since the minimal polynomial of S is not equal to its characteristic polynomial. Let F [ S ]= {f(S): V --t V I f(X) E F [ X ] } . Make use of the Jordan form S N of a conjugate of A . Show that for any integer rn 2 1.&(A). Show that there is an endomorphism T of V so that T commutes with S but T is not a polynomial in S (T is a polynomial in S if T = aol+alS+a2S2+. So there is a linear transformation T of V such that T commutes with S but T is not a polynomial in S. 0bviously. S N = C 2s 2 j ( .22 1118 Let S be an endomorphism of a finite dimensional vector space V over a field F whose characteristic polynomial is not equal to its minimal polynomial. Hint.+akSk for some k and ai E F . N 2 C j=1 8 nj = dimF V. Then as F-vector space. 1119 Let A be an n x n real matrix.. . where r n ( A ) is the minimal polynomial of S . then by Frobenius theorem. . & ( A ) be the invariant factors # 1 of S and let n = degdi(X).

For any integer m and b 1 0 O b 1 2 1... Now let A be an n x n real matrix. .- ..23 Solution. A and Bm are similar. . XE-A and XE-B" have the same invariant divisors (1. Thus L : b " 0 Chb"-l b " Cibm-1 . chbm-l " b I* A = Q-lB"Q = (Q-lBQ)" for some invertible n x n real matrix Q. Hence P-lAP = (diag(B1.(Aa). there exists real matrix B. Then there exists an n x n invertible real matrix P such that F I A P= diag(J1.-2bm-n+2 0 .. As proved in the above.B2. any 1 5 i 5 k. . C.. .. . for any integer m 2 1.. J k ) where Ji ( 1 5 i 5 k) are Jordan blocks with diagonal elements real and positive. all of whose eigenvalues are real and positive.. . '..-lbm-n+l cibm-1 ..}. let b = m & B= 0 0 0 0 It is easy to see that *.. .Bk))m . J2. Suppose first that A= be a Jordan block with a real and positive... ~.* b O 1 b Bm= Obviously. ... 1. such that By = Ji (1 5 i 5 k)..

0 where B k 1120 Let V be a finite dimensional vector space over an algebraically closed field K . .24 and B1 B2 0 P-'] m = Bm. . and let T : V 4 V be a linear transformation. and (T . j ) ((A . and N ( T ) : V S V is nilpotent. . .* . (A-an)ebe the elementary divisors of VqX].. ( Stanford) Solution.~ = ~ i ) ~ ~ j ) . ... is nilpotent.)enl. . Viewing V as a module over the polynomial ring K[A] via T.where a1.j = K[A]wij. E K such that V is the direct sum of the V. * * * . (a) Show that there are 2'-invariant subspaces V . (X-a2)e=. a2. V and elements of a.a Z ) e z T . . (A-a.j E V and A n n ( ~ . we may write where all the w. Hint.ail) : V . ( T ) : V -+ V is diagonalizable. for some n 2 1. + V. . Denote V. N ( T )E K [ T ] such that T = S ( T ) + N ( T ) . ( A .) (b) show that there are polynomials S ( T ) .an are distinct and + By the structure theorem of finitely generated modules over PID.(X-al)el'l . Let (A al)ell.Then . . Use the Chinese Remainder Theorem. (A transformation N is nilpotent if N" = 0.

. (1 5 i 5 n).a 2 ) e 2 r 2 . Then S ( T ). + . Obviously. there exists some S(X) E K[X]such that S(X) ai( mod ( A - ai)eir. So the minimal polynomial m(X) I N ( X ) h for some positive integer h. ). -+ vj is nilpotent. Thus are pairwise relatively prime. Let N ( X ) = X . waj = Cya .a i l ) : V. Thus N ( T ) : V + V is nilpotent. Since ( A .al)e*'l (A .S(X). . T = S ( T ) N ( T ) and X .:j (a) is proved. . ( A . waj for all wij and it is easy t o see that S ( T ) : V -+ V is diagonalizable.a I N ( X ) for any i 1 5 i 5 n. by Chinese Remainder Theorem.25 The K j ' s are T-invariant subspace and ( T . Thus (b) is proved.

Let g E G. . i) If A 1 = A 2 . then N E G and NgN-'= -1 0) A l ( .26 SECTION 2 GROUP THEORY 1201 Let G be the group of real 2 x 2 matrices. let A be an invertible 2 x 2 matrix with real entries such that N=(-d)-?( if d = det A < 0. A 1 and A2 be the eigenvalues of g viewed in M2(@). Then A 1 A 2 = 1. g is similar to - in GL2(R). (Hamud) Solution. In case a). of determinant one. then A 1 = A 2 = 1 or A 1 = A 2 = -1. g = In case b) g = In case c). Describe the set of conjugacy classes of elements of G.) or NgN-'= (1 -1 ). .

then g is similar to ( 2 x".')I* ( 0' 21 ) ( 0' ) ( i11. So in this case. ) in A2 Xi [( i!i :)] [( i! . we have the conjugacy .2). M i f d = d e t M > 0 or N=(-d)-i( if det M = d -1 0) M ' < 0.27 Hence g is conjugate to conjugate to ( ( q' ) 1 1 ) or (1 -1 ) i n G . ) in G .') a21 ( E:: :tz ) E G.) or in G L ~ ( B ) . then we have and = 0. we have two conjugacy classes [ ( i1 yl')] and [ ( 0' 1. AS in case c).)]. Thus in case d ) . Then N E G and Thus g is conjugate to classes ( x". g is inGand ( i11' ) GL2(B). But (A i ) isnot in G. all = -a22 and = 1 which is a contradiction. for if A*( where A = d e t A = -a!l i :)A-'=( . we have two conjugacy classes In case d). ii) If X i # and E B (i = 1. g is similar to similar (conjugate) to is not conjugate to ( i11. There exists M E G L z ( R ) such that We take N = d-4 .) in G . Thus in case c).

we have the conjugacy classes . g is similar to -sin8 As it is well known. So in this case. B E M 2 ( B ) are similar in M2(~.i s i n 8 (since XI . But cos8 -sin$ and cos0 sin0 are not conjugate in G.28 iii) If XI # A 2 and X i E C\E (i = 1. we denote A1 = cos8 + i s i n 8 and = cos8 . - O i sin 8 ) in G L ~ ( C ) . As in the above. that two matrices A . X2 11 < 8 8 # 0.7 implies that A and B are similar in M2(nZ). we conclude that g is conjugate to cos$ -sin6 or sin8 cos6 -sin$ C O S ~ cosB sin8 in G. for if sin8 COSB -sin8 cos8 such that cos8 sin8 ) A-l = ( cos8 -sin0 det A = -(a:l + ail) = 1.2). A2 = 1 and trace ( g ) = A 1 + A 2 E E ) . where T.isino cos 8 sin8 cos8 ) in GL2(C). which is a contradiction. So there exists M E G L 2 ( E )such that cos8 sin0 MgM-I= and also similar to ( cos8 ( cos8.

Make use of Fqn. c) Determine the number of G-conjugates of U.) Xz and XI 3 XZ = 1 -sin6 c0s6 cos6 1 0 (61 1202 Let G = G L n ( F q ) . Thus we can conclude that the order of GL. the set of conjugate classes of elements of G is U { [ ( $ x". U = Un(Fq). Deduce that U is a Sylow psubgroup of G.q2 ways (no linear combination of the first two rows being allowed). the subgroup of upper triangular matrices with 1's on the diagonal a) Calculate the orders of G and U . the second row in qn . ) ] ~ A I .1 ways (a row of zeros not being allowed).the group of invertible n x n matrices over the finite field Fq with q = p " . p a prime.1 Hint.(Fq) is + . a) When forming a matrix in G L n ( F q ) . b) Deduce that every psubgroup of G is conjugate to a subgroup of U . (Columbia) Solution.we may choose the first row in qn .q ways (no multiple of the first row being allowed). E El " {[(i Y ) ]I b = * l } { [( )] < < .29 and cos6 sin6 -sin6 cos6 To sum up. e) Show that G contains elements of order qn . the third row in qn . d) Show that g E G has ppower order iff g = I N with N n = 0. and so on.

the second row in q n P 2 ways: the third rows in q n .I. n + We can define a map 19 : T + F. by mapping a matrix onto its principal diagonal. Uln ~ 2 3 ::: . we get gni = 0. + . x F..l . . . Then for any 1 5 i < j 5 n.. j E V such that g ( I E j j ) = u g where Eij is the matrix with a lone 1 in the (i.. So the order of T is b Hence the number of G-conjugates of U is i=l n (qi n .j = 0 (0 5 j < i 5 n). Hence the order of U n ( F q )is n (P'~ n SO U = Un(Fq) is a psubgroup of G = G L n ( F q ) .l ) / ( q- l)n. Denote h g h . So gn.or gUg-' = U . Then ( 9 ) is a psubgroup and ( 9 ) is conjugate to a subgroup of U by b)... d ) Suppose g E G has ppower order. = 0 if i < TI. if g E T: It is easy to see that g U g . x . 1 .j) place. j = 0 if i < n . Obviously Mn = 0 and g = I h .1. By this way.1)th row. . the number of G-conjugates of U is equal to [G : N G ( U ) ] where N G ( U ) is the normalizer of U in G. ..l M " h = 0. we get g n .. Conversely.l M h = I + N where N = h . . Since p j [G : U ] = .. there exists some 1 U12 0 u=[. U is a Sylow psubgroup of G. Suppose g = ( g i j ) n x n E N G ( U ) . So g N G ( U ) and we have N G ( U )= T . i=l b) It follows directly from Sylow's Theorem.. There exists an h E G such that h g h ..l = I + M where M is a matrix with zero on and below the diagonal.we may choose the first row in q n . c) By Sylow's Theorem.l ways.. of matrices ( a i j ) E G with a.. Checking the last row of g ( l + E i j ) and u g .30 When forming a matrix in U n ( F q ) . we get g i j = 0 if i > j Hence g E T = T n ( F q ) the set . matrix multiplication shows that 0 is a group epimorphism whose kernel is precisely U = Un(Fq).3 ways and so on.' & U (by matrix multiplication).l E U ..l ) . 0's elsewhere.l M h and N" = h . Then checking the ( n .

a l ( z ) = a z (over Fq) which is invertible. if g = I N with N" = 0. Thus g is similar (conjugate) to some matrix in U . has an element with order q" . (IHI . (T is injective. We fix a base of Fqn over Fq.1. ~ ( u = at. --f + 1203 [GI = IgEG U gHg-' 5 [G : N ( H ) ]. the Jordan canonical form for N is a matrix with zero on and below the diagenal. It follows that G contains element of order q" . So the order of g is ppower. we have a linear map al : Fqn FqR. It is well known that F.1.31 Conversely. ) Obviously. --$ G = Gn(Fq).1) + 1.. . then we can obtain a group homomorphism (T : FLf.. e) Consider the finite field Fqn as an extension of the field Fq and as a vector space over Fq. For any a(# 0) E Fqn.

2. then G I K N H . a K.. (1) = KOa K 1 a . . = G N is a composition series of G with composition factors H and H i / H i . K .l Ki/Ki-l (i = 1. K is not isomorphic to H . . f j . Then for any a E G . then (1) u H a H 1 a . On the other hand. H i I H N Ki ( i = O . (a) Let G = Q8 = { f l . Let H be a maximal subgroup of G. n ) and G / K . 1205 Prove that if G is a finitely generated infinite group then for each positive integer n. there exist subgroups Hi such that Hi I> H . (a) < G. n). Hence G is cyclic. l . n ) and Hi ~ H i + (i = O . . (a) Give an example to show that G / K need not be isomorphic to H . G has only finitely many subgroups of index n. & i . . This contradicts the maximality o f k .-1 u K .32 (b) First we claim that G is cyclic. . K a G and GIH is cyclic _-_of order 2. (Columbia) Solution. . . Now. .. we have H a G. I c } N K 4 (Klein four group). j . a H. aK n P 1a K U G is a normal series of G with factors KiIKi-1 (i = 1 . (b) Let (1) = KOa K1 4. * * * .1). . UEG 1204 Suppose that H and K are normal subgroups of a finite group G and that GIH = K. l Suppose H is simple. 3 t k } be the Hamilton’s Quaternion group and H = (i) = {3tl. it is easy to see that G is cyclic of prime power order since all of its maximal subgroup are conjugate.i. . so it is isomorphic to K . . (b) Show that if H is simple. Suppose that G is not cyclic.-l a H . n . = K be a composition series of . But G I K = { 1. 2 . K = (-1) = {*l} be the subgroups of G generated by i and -1 respectively. By the Jordan-Holder Theorem. we have the isomorphism G / K N H .. By (a) G = H . . . . Since GIH N K . Obviously.*i}.. Then we have G = U gHg-l by assumption. So (a) is contained in some maximal subgroup of G (since G is finite).

Hence G is a solvable group. Then G x G I H + G / H .. [G : H ] = n } is finite. and . there are only finite number of homomorphisms from G to S. G/HG has only finitely many subgroups of index n.lR). denotes the symmetric group on n letters. 1206 Let G= {( a!l ) la E lR*.33 Hint. Since G/HG is finite. the symmetric group on n letters. Let 0 be the map G + R* x lR* mapping a!l ) to its diagonal ( a . Let H be any subgroup of G such that [G : H ] = n. (Columbia) Solution. Thus G has only finitely many subgroups of index n. ( z H ) = ( g x ) . Show that G is solvable but not nilpotent. H defines an action of G on G I H . ( Co2umbia) Solution. This induces a homomorphism G + S. Since G is finite generated. where GIH denotes the set of all left cosets z H ( z E G ) . ~ ' ) Matrix multiplication shows that 0 is a homomorphism. E lR b 1 ( C_ GL(2. Let H 5 G and define HG = g€G g-lHg. Show that if H 5 G (finite index) then there exists a homomorphism of G/HG into S[G:XI. Suppose . and GIN E x IR* is also abelian.+))is an abelian subgroup of G. N = ker0 (21 (lR. The kernel of this action is HG = g-lHg. ker8= {( ) 16tnJN Obviously. n Hence the group GIHG is isomorphic to a subgroup of S. with HG as its kernel. g . . It follows that the set g€G {HGIH < G . where S.

So G is not nilpotent. that is. It follows that b = 0.p " . Hint. Let . then f ( X ) = f l ( X ) f 2 ( X ) . If T is semisimple. G = G L F ~ ( V Recall that IGI = (p" ).frn-l = 1.. (a) Let T E G and f ( X ) be the minimal polynomial of T over Fp. f k ( X ) for some distinct irreducible polynomials fi ( 1 5 i 5 k ) . n2 'nk.l ) .(G) C ( G / C i ( G ) ) = (Co(G)= 1 ) . that is. E. Find a Sylow psubgroup of G. Then IEl = p" and Ei is a field of order (1 5 pni. ( 1 5 i 5 k ) . So E contains an element ~i whose minimal polynomial over Fp is f i ( x ) . f(X) is separable (remark: here separability means that f(X) has no multiple roots). f i ( A ) ~ ( A P r n . Then X 1 f ( X ) .(G) = a * * = G $ where Ci+l(G)/C. Show that H can be simultaneously upper triangularized. - .l . y E IR. Let E be a splitting subfield of XPm .l ) . Prove that a transformation T E G is semisimple if and only if Tprn-l= 1 for some positive integer m.. 1207 Let p be a prime and let V be an n-dimensional vector space over Fp.fi(X). Thus T p r n . a = f l . x!!l So ay+ bx-' = xb+ya-l for any x Thus # 0.fi~(X)I(X~~-~-l) .p ) (p" . ) EG. Denote n = degfi(X) i and m = n1 .34 the center of G. A direct discussion as above shows that C ( G / C ( G ) ) ( 1 ) .. Then for any ( . (a) Recall that a linear transformation is called semisimple if its minimal polynomial is separable. (Indiana) Solution. Since ~i # 0 and 7. (b) Let H be a subgroup of G of order a power of p .= C. Since nilm. there is a basis of V with respect to which all of the elements of H are upper triangular.. Hence f(X) = fl(X).l)(pn. Hence = C1(G)= C2(G)= * * .l = 1. is isomorphic to a subfield of E .X over Fp and Ei = Fp[X]/(fi(X)) i 5 k ) .

1)' = -XP . Thus T is semisimple. This proves (b). Otherwise. (p" . So H K = K H is a subgroup of G. (a) Since K is the commutator subgroup of G. By Sylow's Theorem..we may choose the first row in pn-' ways. (b) Suppose that for any elements h E H and k E K holds hk = kh. (b) Show that there are elements h E H and k 6 K such that hk # kh.l 5 gKg-' = U. (p"-' . f ( X ) has no multiple roots. - . the subgroup of upper triangular matrices with 1's on the diagonal.p"-') n n-1 = p.p = pv (when forming a matrix in U ... for any h E H and k E K . ( c ) First we claim that H is not normal in G. Since )GI = (p" . Then f(X)I(P"'-' m 1).1). H is contained in some Sylow psubgroup K of G and K is conjugate t o U . G = H K is abelian. (b) Let U = U. (c) From (b) show that p divides q . Let G be a group of order p 3 .l)(pn . By Sylow Theorem. Then JUJ p"-l .1) .p " .1. K is normal in G and G / K is abelian.2 # 0. the second row in pn-2 ways and so on). Since K is abelian and H is abelian. Since (XP"'-' . 1208 Let p and q be distinct prime numbers. (Indiana) Solution. = . Hence there exists an element g in G such that g H g . IHJ = p 3 . hkh-lk-l E H n K = { e } and so hk = kh.I). Now let H be a subgroup of G of order a power of p. Thus we have proved that H can be simultaneously upper triangularized.2 .p ) * * * (p" . . Let H be a pSylow subgroup of G.(V). H n K = { e } and lHKl = p3 q = IGI. Since IK( = q . (a) Show that H is abelian and G = H K . q such that its commutator subgroup K is of order q. This contradicts the fact that the commutator subgroup K of G is of order q. U is a Sylow psubgroup of G. Hence G = H K and H 21 H / ( e ) = H / H n K 21 H K I K E G / K is abelian. ( p .35 Next suppose TPm-' = 1 for some positive integer m.

For example. is a group of order 3 . + 3f 3e. G also contains an element of order q . fz = 3el 3ez and f3 = 3el . by Sylow’s Theorem. Show G has a subgroup of order p q . By Cauchy’s Theorem. rPlqz and plrp . which is normal in G and of order p . A4 does not have a subgroup of order 6 . and g. q since H is normal in G. So it is Again by Sylow Theorem. Aq. subject to the relations 9e+3f+6g 3e = 0.36 the number of pSylow subgroups of G is greater than 1 and divides q . (Stanford) Solution. So G has a subgroup K of order q. Denote + + + + 9 3 6 . the alternating group of degree 4. = 0. (a) Let rp be the number of Sylow psubgroups of G. q be primes. Thus H . Let F be the free abelian g r o u p g e l $Zez $Ze3. 2 2 . plq . K be the subgroup of F generated by f1 = 9el 3ez 6e3. (b) What can you say if q = 2 (and p > q ) ? (Indiana) Solution. it is easy t o see that rp = 1. Then. p > q > 2 . Since p .1. q are primes and p > q > 2 . So G has only one Sylow psubgroup H . (b) If q = 2. q.3e2 6e3.3f+6g Give a decomposition of A as a direct sum of cyclic groups of prime order or infinite order.1. = 0. q . G may not contain a subgroup of order p . Let G be a group of order pq’. 1210 Let A be the abelian group on generators el f. Obviously A E F / K . 1200 (a) Let p . K is a subgroup of order p .

Suppose that M is invertible when viewed as a matrix of rational numbers.0} of M in M 3 ( Z ) and to find P = ( O1 0 -1 ) 1 1 0 and 1 Q = ( O0 -1 1 -2 -1 such that Q M P = diag{3.. $ Z e L and K = Zfl + Z f 2 +Zf3 = XCe.. f3)'.. e i . e$>' = P . (Stanford) . a) Show that 2"/MZn finite. (fi. Let (e. that there exists an n x n matrix N with rational entries so that M N equals the n x n identity matrix. is b) Show that the order of Zn/M57" is equal to the absolute value of the determinant of M . View M as an endomorphism of 2". @ Uel.e. 1211 Let M be an n x n matrix of integers.37 It is easy t o get the normal form diag{ 3.l ( e l . so A S r" F / K = 2 e i $Zek $57eL/Uei 57/(3) $2/(6) @Z 2 2 $ 2 3 $573 @ Saei 2 $2. fL Then F =Zei $Ze'. e 2 .0} ( P and Q are invertible matrices in M 3 ( Z ) ) .f 2 .6.6. i. f4)' = Q . e3)' and (fi.

eh. en)' = M P (ei. e n } be a base for the free module Z". . . . .. 1' 1 = Q . Q E M. a) It follows directly from b). ( .} where f. .. .. Obviously det D = dl . (fll f~. ([ + 2 .. .fn )I I = Q M P (ei. . It is also obvious from the facts that the map g : Zn/MZn Z " / l M p " .(Z) such that D = Q M P . e. .. 7 * * * . . . Since Q is invertible in M n ( Z ) . Let { e l ..)' Then {ei.(Z). . . f. (fi f~r : . n . . 2 .el. . 1 * * f n )'.(Z) be invertible matrices in M. .. .. e2.4) and (21.f. . el. . d. M Z n can be generated by {f. . = det Q . (ei.. . e2. - Hence 1212 Let D = a [ [ with E = *. } ... . . . where the di # 0 and dildi+l (i = 1 .l).Denote (ei. (Haward) + .. .. .ek} is another base ofZ".21).e.. .. . . where [MI is + the determinant of M .Solution. b) Let D = diag{dl. d z .. 6 ( r ) = M * X is injective.) / I = diag{dl.l). (el. ... Then M P is generated by { f l l f 2 .f )' = Q .dn} be the normal form of M in M.. . . fn)' = M . eh . e n ) . . dz. . det M . Then express G as a product of cyclic groups. Let (fl. d. (fil f. M * is the adjoint of M .ek)l = P-'(el. . and the fact p n / l M p n l = ( I det MI)"..det M . and P . . . det P = det M or . en)I .. f. 0bviously. I / fz..dz. ] Calculate the order of the additive group G = D2/K where K is the D-submodule of D2 generated by ( 2 t 11[ . . f n }. .ez.. (f. .. -. ..el.. .

e2 + (ez.1 -2 -1 -1 A= 1 1 -2 .(el . e2} is a generating subset of the D-submodule K of 0’.<ez) + (ez) + b(-el+ (el . Let { e l l e 2 } be a base of the free D-module 0’.e2 + + + (e2) + b(-2e1 + (-4 + Oezl + (-1 + t ) e ~ l . 2 1 e l + 21e2. - ( a b0[(2 ()el a(2el + < e l .21(el + 21(e2} is a generating +b(O e l subset of K as Z-module (additive group).2(e2. for A in hf6x4(Z) and to find invertible matrices P E &(Z) and Q E 1M4(Z) such that PA& = N . e2 + 21 .4ez + (a = a(21 .2(e2). (e2).e2 . For any a bt E D . It follows that D 2 / K -. {ell(el.e2 + 0 .(el .39 Solution.5(e2.l] 0 .0 . 21. 0 21 .(e2} i s a baseofD2asZ-module. -2el.2/(3) @Z/(21)and the order of D 2 / K is 63. <el + 21 .e2. e l + b()(2le1+ 21e2) + 0 . Then + 0’= Del @ Dez = Z e l @Z(el @Zez @Z<e2. Since ( is a root of the irreducible polynomial z2 z+ 1. . el 21.9 0 ~ 21 5 ~ . Denote 1 2 .e2 .+ 21 (el + 0 . and {(1+2<)e1+(-1+<)ez1(2+ ()el + (-4 + () . It is routine to get the normal form 1 0 0 0 N = [ 0 . It is easy t o see that {el + 2te1. D = Z [ t ] = Z@Z< (as additive group). .2el + ( e l 4e2 + 5e2.-el + (el . e2. + + (a bt)[(l+ = a ( e l + 2tel .e2 .e2 . 0 .5(ez).

if c > a > 1. Z )is the group of 2 x 2 matrices with integral coefficients and determinant = 1. let c = ha + e l . Similarly. a and c are relatively prime. then for some d' EZ. Since det M = ad . 1 5 c1 < a . Z ) with the form ( 1) is a prodct of the for some b'. for some b' E Z. we claim that matrices in S L ( 2 . It is easy to know that if a = 1.d' EZ. Suppose that M = Prove that S L ( 2 . (Stanford) Solution.there exist some q . Z ) is generated by ( 1 rtl ) ) ( i1 If both a and c are nonzero. If a Hence > c > 1.c'. and if c = 1.40 1213 where and S L ( 2 . .be = 1. a1 E Z such that 1 5 al < c and a = qc+al.

- 1215 Let P be a Sylow subgroup of a finite group G. Let N = { x E G I xPx-' = P } .. It is easy to check this by the property of the elementary matrices. it can be derived that M = is a product of the matrices in S L ( 2 . then y E N. d' E Z. ( i1 )) where b ' . So we have g = h . By Sylow Theorem. Hence h-'gP(h-'g)-' = P and h-'g E N . c ' . Z ) = generated by ( 1 3zl ) ( and 0 -1 ).41 According to the above discussions. Thus we have G = K . X ) with the form for some b'. c'. Hence gPg-l is also a Sylow subgroup of K . If N is the normalizer of P in G (i. So to prove that S L ( 2 . there exists h E K such that gPg-l = h P h . Let H be asubgroup of G. h-'g E K N . N = {g E G 1 gPg-' = P } then show that G = K . it suffices to prove that all are in (( 1 fl cc) . For any g E G. (Indiana) Solution. d ' E Z . (Indiana) .l .l = H . H 2 N . 1214 Let G be a finite group. Prove: If y E G such that y H y . N . This completes the proof. we have 9Pg-l C gKg-' 2 K since K is normal in G. K a normal subgroup of G and P a Sylow subgroup of K .e. N .

h-lyEH. The probability of G to be commutative is defined by P(G) = I{(a. So 1 y=h. + -) 4 =8 5 . If G is not commutative. obviously. prove that P ( G ) 6 5/8. P is a Sylow subgroup of H . Since yPy-' & yHy-' H. So there exists an h E H such that yPy-' = hPh-'.b) E G x G I ab = ba}J IG x GI (a) If G is not commutative. then. Hence h-'yP (h-ly)-' = P and h-'y N .b)E G x G I ab = bu}l = P(G) = L . (b) Give an example such that P ( G ) is exactly 5/8.42 Solution. yPy-' is also a Sylow subgroup of H . Since J{(a. Obviously. (Stanfod) Solution.N cH. (a) Let C(G) be the center of G and C G ( U= { b E G I ab = ba} be the ) centralizer of a in G for any a E G. IG/C(G)I L 4.(1 1 2 1 . 1216 Let G be a finite group.

3 5 l C ~ ( a )< 8.43 (b) Let G be the dihedral group 0 4 = { z * y j I 0 5 i 5 1 . .z2 = y4 = 1 zy = y3z}.C ( G ) . Then IGJ = 8 and C(G) = (1. 0 5 j 5 3. y}. For any a E G . It is easy to see that P ( G ) = l =/ 518. and so ~ C G ( U )4. obviously. .

\/---z].n&? is also a n irreducible divisor of 2 0 .G) zg. (1. b) By a). find all integer solutions to the equation y2 + 2 = x3. For any a = m nz/--2 E Z[z/--2].vI 5 1/2. If a = m nis an irreducible divisor of 2 0 .44 SECTION 3 RING THEORY 1301 a) Prove the ring Z [ n ] is Euclidean. So we have a map 6 :Z[-]* + Z?.\/---z where p and Y are rational numbers. Suppose that a.v)Q] +v E Z[] Q is in Z [] and r = a . Z [ G ] is a unique factorization domain. (Hurvard) Solution. b # 0 E a[-]. .bq = b(p .u ) = bq+r where Q = u + (v + v . a) It is readily known that Z [ n ] is a subring of a. we define 6 ( a ) = m2 2n2. Then w + + + a = b(p+vG) = b [ ( u + p. a H 6(a). We can find integers u and v such that I .v)Q.pl 5 1/2 u and l .u ) + b(v . 20 is not a prime element in Z[. ti = m . -1) is the set of units of Z .2112 5 6(b). [] Suppose (20. hence an integral ' domain. Then ab-' = p v.. Since = the integral divisor of yo + 2/--2 or yo can only be f l . Hence Z [ m is Euclidean. and if alyo then + + + + n. (a + t) 2* + 21v - wl2) < 6(b). Obviously T and 6(T) = 6 ( b ) * ( l p . we have (yo Q)(yo . In the ring Z [ G ] . b) Using a).yo) be an integer solution to the equation y2 2 = x3.

n E Z. .(q)2} elements in S. 2 (q)2} y .(R)*. .c3ly0 . On one hand for any i # j . n.p ) 2 = has exactly elements. ( q ) } 2 + 1303 For a ring R.2n3 = (3m2. has exactly elements.rn. exactly (a .2n2).. So without alyo loss of generality. .b2 I b E Z/@} has elements.T. Thus it is easy to conclude that the integer solution to the equation y 2 + 2 = z3 are y = -5 and x=3 { 1302 Z/pZsuch that a = b2 + c2.j # O ( O < i + j 5 p . T . it is trivial.(R) the ring of n x n matrices over R. {o. R* denotes its multiplicative group of units. yo is of the form ( m nJ--2)3. c E (Harvard) Solution. Let S be the set { b 2 I b E a/$}.S ) n S # 0.b2 = c 2 .45 - from which it follows that a31yo fl. and GL... .z.1/--2. the set a . i . there exist b.y 2 = i + j . . (b) Let J be a nilpotent ideal of R (J" = 0 for some m).0 < j 5 q.Thus 2 S' = { b 2 I b EZ/$} = { a .a E R*. (a) If a E R is nilpotent (am = 0 for some m) then 1 . Hence there exist 6.j < 9 if i > j . Since a .S C Z/pZ and Z/pZ has only p elements.(R) = M. 2 .. for any A2 E S where 9 < n < p . Hence + + + yo = m3 . Show that for any element a E Z/$. 9 G .0 I j 5 e. Now for any element a E Z/@. that is a = b2 c 2 .22.T.. 0 < p n 5 9 and A2 = -~ - n2 = ( n . When p = 2 . Then S = {o. On the ) are other hand.S := {a . ? 2 . c E a/@ such that a . Let p be a prime.6mn2 1 = 3m2n.1. M. : i 0 < i . Assume that p # 2..

.c E F. : F. There exsits a E 3.} is a pSylow subgroup { A = ( a i j )I = { A E M.(R) = I+M. .). T.s + 3. .(R) --t G L . b.". . ( J ) } (c) By calculating the possibility of the row vectors of A in GL3(Fq). with kernel ) (d) Show that U = {( l a b 0 1 c 0 0 1 ) a. % ( R / Jis surjective. 2 = diag{i. Such that Fi3 = ( a ) .T. (e) We consider the field extension 3 C 3. 1)) - IA =I mod M . Obviously.(J).s.4 6 Show that GL. . 3: H az is an invertible linear . is it easy to see that U is a pSylow subgroup of GL3(3.3.

1)(52. b ) of the Pell’s equation z2 ..2b2 = 1). relative to some base for Fq. and ” a = 2 4 . So if we take a. b = 2aobo) is another solution of z2. 2 ) is an integral solution of x 2 . = 2aP-1 . - P { A = ( a i j ) 3 x 3 E G&(Z/2U) 1 a ( A ) E U C GL3(Z/U)} where cr is the homomorphism G L 3 ( a / 2 u )-+ G L 3 ( Z / U ) . 3 .1) = 2 7 .1 .e. the order of A in GL3(Fq) q3 . v.1 . Thus P is a 5-Sylow subgroup of G L 3 ( Z / 2 U ) . = 3 . is ( f ) Let R = Z/2M and J = U/2U < R . a’ 1 = a. + so Let (GL3(Z/2U)I= 2 7 .1. for some integers ao. Let a = 2a’+ 1 and b = 2b’.Since . then b = b. we get an infinite set of integral solutions { ( a i . Then J 2 = 0 and % = R / J N Z / U . &om the above consideration. the kernel of the surjective homomorphism G&3(Z/2w) GL3(Z/5Z) 3 is I and + M 3 ( J ) . Let A E GL3(Fq)be the matrix of T.-lbi-l (i 2 l). Hence either both $ and a f 1 are square ’ numbers or a and ’ are square numbers.bi) 1 i 2 0) of the pell’s equation x 2 .47 transformation over Fq.2y2 = 1. . Then a must be b) odd and b even.31. Obviously. Obviously [ I M 3 ( J ) J= 59 IGL3(Z/U)I = 53(53 . 3 . b = f2aobo.1)(5 . bo = 2 and a. b E Z such that a2 .1./Fq. II = 512. a .2y2 = 1. 512 31. is an integral solution of x 2 . bo) is a solution of z2 .Then P/ ker cr ‘v U and JP] I kercr] .(a) = q3 . ( 3 .2y2 = 1 . + + .bo) is a solution of z2. say a’ = 2b.2y2 = 1 .2y2 = 1 .2y2 = 1. (Indiana) Solution. (ao. Then we have bf2 = So must be a square number. If $ and a ’ 1 are square numbers. = U 1304 Describe an infinite set of integral solutions ( a . ad. it is easy to see that if (ao.. bi = 2a. Suppose that (a. bo. then (al = 2 4 . By (b). s 3 .1 .2y2 = 1 (i.

(a) If R is an integral domain. (b) Since for any a E R. 1306 Let R be a commutative ring with 1.z = 0.. Let R be a commutative ring in which u p = a for all a E R.2y2 = 1.e. Char(R) = p and R ZlpZ. Conversely. prove that R is isomorphic (as a ring) to (b) In the general case (i.a u(aP-1. 1305 Let p be a prime. RIP .. b = k2aobo. (Indiana) Solution.c.p -z . By (a). then ( a = 2ag + 1 . b = 2 f 2 a o b o ) is an solution z2 . (Indiana) . u p = a. For any a (f 0) E R. prove that R is isomorphic (as a ring) to a subring of a direct product (not necessarily finite) of rings each of which is isomorphic to Z/pZ.2y2 = 1. R not necessarily an integral domain). (a) Suppose that R is an integral domain. from the solution ( 7 . the intersection of all prime ideals of R is 0.i. the nil radical of R is 0.bo) is an solution of ' .48 Remark.c. then (ao. since UP .70) of z2 . a is a root of the polynomial zP .c. ap-' = 1.bo) is an integer solution of z2 . (ascending chain condition) for finitely generated ideals then show that R satisfies the a. 5 ) of x 2 .c.d. Hence R is a field. Since p is a prime. that is. we get a solution (99.2y2 = -1. Give example (without proof) of such a ring which is an integral domain but not a p. if (ao. It follows that R has at most p elements. each of which is isomorphic to a/$.1) = 0.2y2 = -1 and a = 2 a i + 1.Z / p Z . For example. say a' = ug. RIP is an integral domain and for any a E RIP. If a' and are square numbers. a'+ 1 = 2bg. a - P For any prime ideal P of R.. If R satisfies the a. Hence = all prime . for all ideals. and for any a E R. Thus we have proved that R is isomorphic to a subring of a direct product of rings.a.2y2 = -1.

(1 + a ) ( l + b ) = 1 + (u + b+ab) 6 S. b E A.the polynomial ring over Z.l R is a local ring. (b) Show there is a one-to-one correspondence between the prime ideals of S . Show that if the induced map . we can construct a strictly ascending chain of finitely generated subideals of I . Let A be an ideal of R.c. but not a p.c. (a) Obviously. what is S . what can you say about the structure of S-lR? + + (Indiana) Solution. P n S # 0 if and only if P A # R here.d. X [ z ] . Hence R is a Noetherian ring.l R by (b). (a) Show that S = (1 a I a E A} is a multiplicatively closed subset of R.e. then for any ideal I of R. Obviously. for finitely generated ideals.c. For any 1+ a and 1+ b.l R and those prime ideals P of R such that P A # R. So S .c. the So every elements in S is invertible in R. (c) If A is contained in every maximal ideal of R. then S . 1 = 1 + 0 E S .49 Solution.l R ? (d) If A is a maximal ideal of R. for all ideals.i. then A J ( R ) . and let -+ N be an R-homomorphism.l R is isomorphic to RA. S-lR = R. 1307 Let R be a commutative ring. a . 1 is finitely generated. and it is easy to see that S .let M and N be R-modules.l A is the unique maximal ideal of S . If R satisfies the a. This is what we need to prove (b). (b) It is well known that there is an one-to-one correspondence between the prime ideas of S . It follows that S is a multiplicatively closed subset of R.l R and those prime ideals P of R such that P n S = 0. (d) If A is a maximal ideal of R. Hence Jacobson radical of R.c. i. the localization of R at the maximal ideal A. for all ideals.c.satisfies the a. + 1308 f :M Let I be a nilpotent ideal in a ring R.. (c) If A is contained in every maximal ideal of R . R satisfies the a. For otherwise.

* = I n . = 0. (Columbia) Solution. Obviously.1 E F [ t ] . L m={I4t>Ip. f ( M ) + I N = N . We define V.. It follows that I . by defining I f ( t ) l . Solution.V p ( f ( t ) ) .(f(t)) = 00 if and only if f ( t ) = 0. P ( t ) ) = 1 = (. Then. 1 (called padic absolute value) on F ( t ) . 1 is non-archimedean (If(t) W l P + Suppose F ( t ) is the completion with respect to 1 Ip. N / f ( M ) = I .N / f ( M ) = . we get an absolute value 1 . Is(t)lPH- 1 + m. Let R = { a ( t )1 a ( t )E F ( t ) . then f is surjective. Let p = p ( t ) = t .(t). ii) V. Explicitly construct one non-archimedean absolute value I I on F ( t ) .(f(t) + d t ) ) L fin(Vp(f(t)>. .50 - f :M / I M ---t NIIN (Harvard) is surjective. Since f :M/IM -+ N/IN is surjective. N / f ( M ) = 1 2 . Let t be transcendental over F . 4 l IP(t>lP) . : F ( t ) -+ I?i by %(O) = 00 and Vp(f(t)) = Ic if 0 # f ( t )= ~ ( t b) t )~c ( t ) . If is the completion of F ( t ) with respect t o 1 1. .P(t)). because I is nilpotent. So N = f ( M ) and f is surjective. Since I maXMt)l. c ( t ) E F [ t ]and * ( / ( b ( t ) .Ia(t)lp5 1) and M = { a ( t )E R I la(t)l < 1).dt))= V. Obviously. 1309 Let F be a finite field containing 5 elements. show that the set of a(t) E F ( t ) satisfying la(t)l 5 1 is a local ring. N / f ( M ) = * . I .(f(t).VP(dt))).(f(t)) + % ( S W and 4 v. N/f(M) = I N Hence + f(M)/f(M)= N / f ( M ) .where b ( t ) . = 2 . we have i) V.

. . . . . . b ( t > P ( t ) l P = l+)lPIP(t)lP> - . Since X I .X 2 . 2 .X .. . . .So. = UilY1 * * * U l m Y m fa + + + for any X . x .X. . + * * * gj for j = 1 .any For R is a subring of the field F ( t ) and M is an ideal of R.. .} is another generating subset of the ideal ( X I . . are algebraically independent over a . where b j l . .Y. n ). .2. Suppose that { Y l . and X I . X 2 . * * * .X . 2 .X . . . . and f j is a sum of monomials in ' Y 1 . . We We can write x. . ] ( StanfonE) Solution. a ( t )E R\M.X 2 .. . gj is a sum of monomials in so = 5(5 k=l uijbjk) x k + (terms in X I . .n). .of degree 2 2) j=1 ( i = 1 .X z . ) of the ring C [ X 1 . E a'.. . . . . . where a i l ..* . ' m aijbjk j=1 = 6ik (i. cannot be generated by fewer than n elements. . Thus R is a local ring with maximal ideal M . . . . Y2. a ( t ) ..51 and . In the same way.rn. 1310 Prove that the ideal generated by X I . in the polynomial ring a'[X1. have to show that m 2 n. . . . of degree two or greater..l E = F ( t ) is the inverse of a @ ) . we also have y j = bjiXi + bj2X2 + + bjnX. a i m E a . bj. k = 1..X .Y. X .l E R and a ( t ) is invertible in R. of degree two or greater. . .]. . we have la(t)lP 1 and Ia(t)-'lP= 1 where a ( t ) . a j 2 .

we have I1 n 12 = 1 1 I2 = (a") . and since ab E I J C N ( A ) . There exist a E I . = (Indiana) . b E J such that a + b = 1. By Chinese Remainder Theorem.52 1311 Let A be a commutative ring. " Then I1 and I2 are proper and I1 I2 = A . Let 1 1 = (a") and 12 = (b"). Since every prime ideal of A contains either I or J but no prime ideal contains both I and J . Find all z such that f(f(f(f(f(f(z)>>))>2. there exists an integer n such that (ab)" = an b = 0. 1. since + 1= (a + b ) 2 n E (a") + (b") = I1 + 12. Prove that A N A1 x A2 for some (nontrivial) rings A1 and A2. < 2. 1) -+ [0. ( Stanford) Solution. and let I and J be two (proper) ideals such that every prime ideal of A contains either I or J but no prime ideal contains both I and J . 1312 Define f : [0. (b") = (a"b") = 0 and thus A N A/I1 x A/I2 where A/I1 and A/I2 nontrivial. 1) by f(x)= { 2x. 22 - if 0 5 if 1 5 22 22 < 1.we have I J N ( A ) (the nil radical of A ) and I + J = A .

we have = n { u } .--) z if and only if z = = &. for any z 6 [0.[ n .[u])] = nu .) = (2 * (2 . Use this result to prove that T is an integral domain which is not a unique factorization domain. [u] denotes the largest integer We denote { u } = a . sin n2 0 . sinnz I n E nV}. f(f(f(f(f(f(. Define deg f(x) = N where U N or bN # 0. 0 5 k 5 62. Notice that 2 = (642) = 642 . Since E [0.n[u] . ( 2 2 ) . c o s n z .[n(u.[64 . cos nz + b. - 0 ) . it is easy t o see that N f ( z ) = a0 + n=l a.n[u]] nu . Show that deg f .)))))) z = { 2 6 4 = (642). cos nz + b.n .([nu] n [ u ] ) = = nu . Hence f(f(f(f(f(f(4 { n . For any real number a . (Columbia) Solution.53 Solution. Since for any real number a and positive integer n. 1313 Let T be the ring of all real trigonometric polynomials N f(z) = a0 + n=l a. By using the orthogonality of the set ( l . (2 .{ u } ] = n(u . [a].[nu .632 = [64z] = [63z + z] (to be an integer) if and only if 1 2 63 Thus f(f(f(f(f(f(z)).[u]) .[2z] = ( 2 2 ) 5 a (Gauss function). g = deg f + deg g. (2 * ( 2 . l). z] if and only if 63z = [64z].[nu]= { n u } .[u] here.{ u } } * * . 1). sin nz. Then it is clear that f ( z ) = 22 .

bncm 2 ) sinkx] . Hence the units of T are {*1}. dM = b N . CM.g(x) = 0 can happen only if either f(z)= 0 or g(z) = 0 by the above degree relation. d$ = . a N c M ~ b N d ~ The coefficients of c o s ( N + M ) x and s i n ( N + M ) x in f(z). g(z) = 1. ao. Hence deg(f(z). . Then we have a N = 0. then a N .e. . are zero. bN = 0. then the degree relation implies that d e g f ( z ) = 0 = degg(x).bN f ( z ) = a0 and g(z) = c g + n= 1 an cos nx + b. Since (an cos n x + bn sin n z ) (cm cos m x + dm sin m z ) + m-n=k c andm . a l . Since c h d& # 0 . If f(z). Let N -. sinmz. which is contrary to aN or bN # 0. . Thus we have proved that bN(c$ + d&) + = U N d M = 0. b l .if and only if all the coefficients of f(z).. So T is an integral domain.g(z) are and a N d M : b N C M respectively. . 2 and so aNCM = 0. Now f ( z ) . . If both of them are zero. aN and bo.b N C M .g(x)) = N + M = d e g f ( z ) + d e g g ( z ) . sin nz + c M m=l + cm cosmz + d.. i. Suppose that deg f(z)= N and deg g(z) = M .

outline a proof that B is Noetherian.. Let ~ 1 .u. . Since ) L _> K is a finite separable extension. a n the set of the conjugates of a. -+ T r L / ~ ( z y is non-degenerate since ' I ~ L I K 0. . 1 f f i s i n z are irreducible. .. we get TrL. for any z E B . E B . y) ) # be the dual basis of u l .. - + + + ) + . ( Stanfod) Solution. . Denote a = a 1 . + + Again by the degree relation. . 2 1 2. Hence B C Avl + Av2 . n ) = 1.55 Obviously. For any f(a)E L. a 2 . Now. Let p = Char(K) # and n = [ L : K ] . . we claim that the trace function ~ L I K 0. .cj # 0. Since A is Noetherian.clzn-l + c n ~ z . Then by using Newton's identities on the elementary symmetric polynomials. ( k i E B ) . . be a basis of L over K contained in B. it is easy to see that nLIK 0 # ( n ~ / K ( b = nb for any b E K ) . Then. . Hence T is not a unique factorization domain. in T we have cos 22 = (cos z sin z)(cos z . (the elements in L satisfying T T L / ~ ( u ~ = ~ . 1.a s i n z). j for all i. be the ~ + - minimal polynomial of a over K and j be the minimal positive integer such that p ] j and c j # 0. c.&j) = (-l)n+j+l * j .sin z) = (1 a s i n z)(1. v. v 6 . and cos z f sin z and 1 f d s i n z are not associates. u 2 . If p = 0 or ( p . + Av. The bilinear function ( 2 .Let zn . 1314 Let A be a Noetherian integral domain integrally closed in its field of fractions K . First. Let L be a finite separable field extension of K . Since xu. u2. = K [ a ]for some a E L. * knvn . be nL/K(f(a)) = i=l C f(ai). .* -u. If B is the integral closure of A in L . suppose that p # 0 and p I n. Thus cos 2% in T does not have an essentially unique factorization into irreducible elements. Ici = ~ L / K ( ~ u E ) A for any i.. . * -* ~ (-l). K Let 211. j). B is Noetherian. all the factors cosz f sin%. 3 has the form lclvl : k2v2 . Thus ~ L I # 0.

is a Noetherian ring. (a) Show that R. 03 It is easy to see that f = 0 " + + .. prove the converse. " Hence f is nilpotent. blbz . first we get that a0 is nilpotent. a0 a12 nilpotent and n=O c anzn is nilpotent. be the subring of the ring of polynomials F[x] consisting of polynomials f0 + f i x + fizz + . then f is nilpotent. bz. E I. (Stanford) Solution. in its field of fractions. + a k Z k Assume that is is nilpotent. (b) Show that the field of fractions of R. . Since A is commutative Noetherian. * . If I" = 0. If A is Noetherian. Since A is commutative. . suppose that A is Noetherian and all the elements ai are nilpotent. . and for n 2 1 let R. - 1316 Let F be a field. ak is nilpotent. . then all the elements aa E A are 00 n=O nilpotent... So. is F ( z ) . show that if f = c anzn is nilpotent. + f k x k E F [ z ] * such that f1 = fi = . Suppose that f = implies that a? = 0. = 0 for any b l . that is.56 1315 If A is a commutative ring and A[[%]] the ring of formal power series over is A . ao. By induction. . Hence 2 n=k+l is nilpotent and so ak+l is nilpotent. (Stanford) .e. then it is easy to see that f = 0.b. Let I be the ideal generated by {ai I 0 5 i < m}. I is finitely generated and nilpotent.b.and determine the integral closure of R. that if all the elements ai are nilpotent. i. Conversely. all the elements ai are nilpotent. = fn = 0... a i .

we may assume that ( p .1 + k * 4 E R. P . For any q / p E as ( q E Z.57 Solution.pl+qk . Hence F [ x ] is integral over R. p On the other hand.zn+' E R n [ X ] . It is clear that the field of fractions of R. Since F [ x ]is an integrally closed domian. R. a ( x ) / b ( z )is integral over F [ z ] . a ( z ) / b ( z )E F [ z ] .. in its field of fractions.) ( Stanford) Solution. On the other hand. Let R be a subrings of&. 1317 Describe all subrings of &. k E Z. E S). for any q / p E R. is a Noetherian ring by Artin-Tate Lemma (or Eakin's Theorem).Then + 1. S is a multiplicatively closed subset of Z containing 1. by Let Obviously. (a) Since F [ x ]i s Noetherian and s is finitely generated a &-module. Let Z s be the localization of Z at the multiplicatively closed subset S . q ) = 1 and pl q k = 1 for some I . Obviously. Hence Z s C R. is F ( z ) . It follows that R = Zs where 1 s={O#PEZ)-ER}. = q . (A subring contains 1 by definition. if a ( x ) / b ( x ) E F ( z ) is integral over R. (b') Observe that for any a ( x ) / b ( x )E F ( x ) . z is integral over &. Hence R C Zs. E R.. since z is a root of Xn+' . Hence F [ x ] is the integral closure of R. P P P So p E S and q / p E Zs. Then R _>a definition.

Thus (27s I S is a multiplicatively closed subset of Z} is the set of all subrings of& (S can be choosed as the complement in Z of the union of some prime ideals of a ) . .

u(zi))= ~ e . y 2 . Show that the polynomial - - - (x2 + 4) .' E . we have - - - ( . (b) Let { z l .16) . it is easy to see that is transitive. 4z2). . y E X and leaves all other elements y) fixed) is an element of G . . (g(zl). For any 1 < i 5 n.Y)(Y. Since G acts transitively on X . a prime number. G. Hence is an equivalence relation... (a) is a n equivalence relation. Define a relation on X by requiring x y if either x = y or the transposition (2. 2) Suppose f E &[z] is irreducible and has degree p .2 real roots and 2 complex roots. (b) If G acts transitively. (z2. we have n 2 m. Similarly. Thus { ~ ( z l ) . ( c ) If Card(X) is a prime number and if G acts transitively and contains a t least one transposition then G must be the whole permutation group of X. If f has exactly p . Since (x.4)(x2 . ( 2 1~ i ) g . Hence m 2 n. So all equivalence classes contain the same number of elements. on p symbols. - is reflexive and symmetric. # Y. 1) (a) By the defintion of -. ~ ~ ~ . (which interchanges z. Thus m = n. Show the following. ~(2.. x . y m } two equivalence classes of zn be determined by z = x1 and y = y1 respectively. and }{ y 1 . then all equivalence classes are distinct and contain the same number of elements.YY = ( z .there exists g E G such that g ( x ) = y. Solution. z ) .2 ( Cohmbia) is irreducible and determine its Galois group over &.)) are n distinct elements belonging to the equivalence class determined by y1 = g(x1). . . x 2 .SECTION 4 FIELD AND GALOIS THEORY 1401 1) Let X be a finite set and G a subgroup of the group of permutations of X . show that the Galois group of f over & is the symmetric group S.Y # z).Z)(x.

r p )is a splitting field of f(z) over & ( r i ) . Consider the conjugation automorphism onG. Let g(2) = (z2 + 4)2(z2. n. which is generated by all the transpositions.. since f(z) is irreducible and f ( r i ) = 0 = f ( r j ) .. m = 1 and n = Card(X). This maps f(z) to itself.. by (b). (z. on X . .l. f(z)= (z2 + 4)2(z2 . i.and the graph of y = g(z) has the form Fig.f4. where m is the number of distinct equivalence classes determined by -. Then q E Gal(E/&) and q(ri) = ~ j which shows G acts transitively .16) . . r p } . and also over Q ( r j ) . Since Card(X) is a prime.e. = rI(z . .2 is irreducible over &. r P )is a splitting field of f(z) over & contained in a.4)(z2. The real roots of g(z) are 0. Thus the conjugation interchanges r1 and r2 = Fl and fixed all other real roots.16). Let r1 and r2 be the two non-real roots of f(z). Thus for any z.Ti) i=l in So E = & ( T I . 2) Suppose - P f) ( . and n > 1. So Card(X) = m . By Eisenstein criterion.4)(x2 . @[XI. Since E =&(rl.60 (c) Suppose G acts transitively and contains a t least one transposition. For any r i . . say n. .)/& into&(rj)/& by sending ri to r j . Thus the Galois group G of f ( z ) over & is the symmetric group Sp on { T I .r p } of the (distinct) roots by q + q1x.y E X . It follows that G is the whole permutation group of X . all the equivalence classes contain the same number of elements. Hence the restriction of the conjugation to E is an element of G and it is a transposition. there exists an isomorphism of&(r. f2. Then. .this isomorphism can be extended to an automorphism q of El&. . r j E X. .1 + . z y. Identify G = Gal(E/Q) with a permutation group of the set X = { T I ..y)E G.

= ~ } is < p Let H =< C T ~>< G. ~ ~ ~ . = @(cP): InvH] = IHI = f and [Inv(H) : &] = e . { a ( < p ) .1 and put f = ( p . Hence the Galois group of f ( z ) over & is ST.e.61 Since I g ( n ) l > 2 for any odd integer n. g is a generator for (Z/pZ)*).. 1402 Let p be an odd prime.1a(base)for&(<p)/&.)/&) =< CT >= {c. Now for any i. a divisor of p . Define f -1 . Use the generator CT : Cp --+ ( C p ) g of the Galois group of&(Cp) over &. . a 2 ( ~ p ) .aP-’ = 11. Obviously. Hint.up-’.it is easy to see that f(z)= g(z) . Since H is normal in G. (Columbia) Solution.)g is an automorphism of &(GI over & and G = Gal(&(<.Fix e. Since g is a primitive root ( mod p ) . C pP . CT’. for any i. Inv(H)/& is a Galois extension and Gal(Inv(H)/&) 21 G / H .qa = C(<p)gej+i j =O Show that.2 has five real roots and two non-real roots. Let Cp be a primitive pth root of unity and g a primitive root ( mod p ) (i.. CT : Cp H (&.l)/e. vi generates a subfield of &(Cp) of degree e over &. Then J H J f. .

Show in succession that the polynomial A . 1403 Let K be a field and z be an element of an extension of K such that 2 is transcendental over K . and let H denote the subgroup of G consisting of the substitutions 2 + 2 b with b E K .yB E K ( y ) [ X ] not 0. Put G = Aut(K(z)/K). is a primitive element for Inv(H)/Q. Thus qi generates a subfield of &(&) of degree e .AB 6 K . Inv(H) =@(qi) for any i. It follows that 71.. and put y = A(z)/B(z). e. that z is is + . i.62 and f --I j=O f-I j=1 f-1 ..B) = 1..1 are distinct. (a) Let A . Hence j=O n (z q j ) is the minimal polynomial for qi over Q (for any - i). B E K [ X ] . gcd(A.e.

. Conclude that n.h(z) is a Galois extension of a given field L with lFq C L C Fq(x)if and only if L 2 l F q ( z ) .y B = 0 in K ( y ) [ X ] . (b) By (a).where K* .y B ( z ) = 0.y B is irreducible in A K [ y ] [ X= K [ X ] [ y ] .z)q+l/(z' .y B # 0. E K [ X ]such that A S + B T = 1 T in K [ X ] .y B is irreducible in K ( y ) [ X ] Thus the minimal .h(z). (c) By (b).y B is irreducible in K ( y ) [ X ]and that . c. I2 = {diag(a.yB and (c [ K ( z ): K ( y ) ]= deg(A . deg B } = 1. B ) = 1. there exist S.bc # 0. Again.a) 1 a E K*}. ] Thus A . K ( y ) = K ( z ) )if and only if max{ deg A. B ) = 1 in K [ X ] K ( y ) [ X ] ) . b. (c) Show that when K is infinite.deg B } . (b) Show that the elements of G are given by the fractional linear substitutions z -+a z + b/cz d with a . Hence y is transendental over K . Then It follows that degA = 0 and deg B = deg(A) = 0. d E K and ad . [ K ( z ): K ( y ) ]= max{deg A. that A . z must be algebraic over K . y = A ( z ) / B ( z )is a generator of K ( z ) / K (i.Suppose A .that y is transendental over K . ad .bc # 0.e. . or if and only if y has the form az+ b/cz +d. polynomial of z over K ( y ) is the monic polynomial which is a multiple in K ( y ) of A .z)qZ+l.63 algebraic over K ( y ) . which is contrary to A B 6 K . ( CoZumbia) Solution. Find the fixed field of H . Thus A . deg B}. Since A(z) . it is easy to see that G N G L z ( K ) / K *. which is contrary to the assumption. since gcd(A. (d) Show that when K = lFq and put + z (zq2 .q and the fixed field of G is n. (a) Since gcd(A. then ordG = q 3 .12. then G is infinite and the fixed field of G is K . If y is algebraic over K . Find the fixed field of H . z is algebraic over K ( y ) ..y B ) = max{degA.

X)+P Denote and (1 + x . deg B } = q3 . if there exists some y = f(z)/g(z)E Inv(G)\K.q . K C Inv(G). Then gcd(A. For any g :2 H az + b/cz + d .x)q+l/(XP . Hence AutK(y)K(z) is finite.X)q2-g. On the other hand. .X)n”-n A = (1 + XP-’ + X(P-1)2+ .4. Similarly. (xq2 X) - 1 (XP . it is routine to check g(z) = g((zq2 . Thus we have Inv(G) = K . K ( z ) / K ( y )is a finite dimensional simple extension. On the other hand. which is contrary to the facts that G AutK(y)K(z) and G is infinite. B ) = 1 and max{deg A. Thus Inv(G) = Fq(z) and Fq(z) is a Galois extension of a given field L with FP C L C F. = Hence F ( z ) c Inv(G). . + X(P-Q?)q+l n1 (XP . (ad .(z) : Inv(G)] = q3 .q .X)q2+1 = ((xq-x))q+l. (d) Suppose K = Fq. Obviously. so is G. + X(Q-1)Q)q+l B = (Xq. the fixed subfield of G. Then c and [F. then by (a). Hence we have [Fq(z): Fq(z)] = q3 . . .Z)Q+1/(2~2)q2+1) 2.(z) if and only if L _> F ( z ) . we have (XP2 .1 1 2 + . since H is also infinite.64 If K is infinite.+ x ( n . we have Inv(H) = K .bc # 0) in G. .

Y")(X" .y-')(x . G =< u .for example. ( X . and = earriIn.(z') = F q ( z q . E/Inv(G) is a Galois extension with Galois group G and [E : Inv(G)] = 2n. Since u ( Z ) = u(Yn + Y-") = Z and T(Z) = 2.C Y ) . Hence c and T are automorphisms of E/C. x + 1 = X2" . Y is a root of X2" .l .the Dihedral group of order 2n.ZX I. using 1403). Thus = Inv(H) = F.Y-") E F[X]. (b) Since X2" . we have Z E Inv(G) and F = C(Z) Inv(G). F = C ( 2 ) with Z = Y" + Y-". the Dihedral group of "' ~ order 2n. (a) Since Y is a generator of E over C.[X] and let z' = zq .X : q and B = 1 in F.z E Fq(z). >rx D. we have [Fq(z)Inv(H)] = q since IHI = q.there are unique homomorphisms u and T of E to E over C such that . ord(7) = 2 and Tu = u-T in G.(Y"+Y-") + 1 = (X" . (b) Show that Y is a root of a polynomial of degree 2n with coefficients in < F.2 .. Then z' E Inv(H) and by (a). Since ord(u) = n.. [Fq(z) : Fq(z')] q. and that the subgroup G of Aut(E) which these generate is isomorphic to D. It is readily verified that X2" -ZX+ 1 is irreducible over F (Z is transendental over C and X2" . (c) Show that EIF is a Galois extension with Galois group G.z). . (c) Obviously. 1404 Let E = C(Y) with Y an indeterminate.l . (Columbia) Solution.Y)(X .(Y-') * (X - y-1y-l) .65 Similarly.. Obviously. Let A = X .Z X 1 is irreducible in C[Z][X]. E is a splitting field of X2" .C"-'y) (X.(Y) = CY and T ( Y ) Y . (a) Show that there are unique automorphisms u and r of El@such that a ( Y ) = CY and T(Y) = Y .ZX 1 since + + + x2n - zx + 1 = ( X . = (T" = 1 = r 2 .

the fixed field of G in F . there are exactly 3 subfields lying strictly between K and F which correspond to the three proper . K C_ F is a Galois extension and [F : K ] = IGI = 4. field automorphisms of F ) given by ( ~ ( z ) 2 . Consider the elements (T. It follows that E / F is Galois and [E : F ] = 2n. + f ( t )= t4 . find a finite subset S of F such that K = &(S). (c) How many subfields lie strictly between K and F ? How many of these are Galois over K ? Justify your answers. (a) Since ( ~ ’ ( z ) = z.1) = z It is easy to see that ( z . Hence where K 4 is the Klein 4-group. Hence&(q) C F is a Galois extension and [ F : &(71)] = 4.66 in E . It follows that (c) By the fundamental theorem of Galois theory..7 in A u b ( F ) (i.qt2 + 1€&(V)[t]. 2. On the other hand. ?(z) = z and T ( T ( Z ) = = ( T T ( z ) . (b) If K is fixed field of G in F . Let (x11)2 5. (b) Obviously U ( Z . (Indiana) Solution.1) = 1 .. Then&(q) 2 K F . r 2 = 1 and UT = T U .x and ~ ( .1 (a) Find the subgroup G (of Aub(F)) generated by (T and 7.we have (T’ = 1. 1405 Let F = &(z) be the field of rational polynomials in one variable x over & (i. Then f ( t ) is irreducible in & ( q ) [ t and F is a splitting field of f ( t ) since ] in F [ t ] .z and = 7(z) = A.e. Denote 7 = (z .1)2 + E K .1)’ &. the quotient field of the polynomial ring&[z]). Hence we have F = Inv(G) and E / F is Galois with Galois group G.e.

) (.67 subgroups of G K4. all these 3 subfields are normal over K (finite dimensional and separable).F ) 1 = ( . u and the Let be elements of Autq(&(t)) given by u ( t )= and ~ ( t=) -t. CT3 ( t ) = and a4(t) = T 3 (E) -3 t. T 2 ( t ) = t .f . It follows that G is the fixed field of H . It is easy to see that r2 = 1.t 2 ) (2 . (b) Let H be the subgroup of G generated by u2 and T . T(T = C3T. Thus all these 3 subfields are Galois over K. Since all the subgroups of K4 are normal. Hence. So + . (Indiana) Solution.t ) ( z+ t ) (x z . = (2. u4 = 1 and isomorphic to the Dihedral group (b) Since D4. 1406 Consider &(t). = On the other hand. (a) Obviously. :) + & ( t 2 + $) of is irreducible over & (t2 $) and Q ( t ) is a splitting field over separable polynomial f(z). and obviously [&(t) : InvH] = (HI 4. Find a E @(t) such that & ( a )is the fixed field of H (Justify your answer). Let G be the subgroup of Aub(&(t)) generated by u and T . field of quotients of the polynomial ring @[t]. (a) Identify G. u 2 ( t ) u = = .

But fl are not roots of f(z).B. crPy = -1 and the discriminant A of f(2) is 34. a) If f(z)= x3 . let K = & ( a )where a is a root .32 1 is reducible over &. It follows that that is. a) Prove that f(x) is irreducible over &. Since f(2) = x3 .[& + + + . Obviously. b) Let . ' c) Find a generator of the Galois group Gal(K/&). Thus f(x) is irreducible over &. (Indiana) + Solution. Hint. This is a contradiction.rbe the other two roots of f ( 2 ) = z3 . InvH =& (t2+ $) 1407 Let & denote the field of rational numbers.I.32 + 1 in a splitting field of f(x) over K. Consider a .3a: 1 is monic. the rational roots must be integral factors of 1.3x 1. f ( ) = x3 . of.68 is a Galois extension and [&(t) :& (t2 + $)I = 4. f(x) has a root in &. cr + p + 7 = 0. then f ( z ) has a factor with degree one in that is. Hence /3 y = -a and .2. b) Prove that K/& is Galois.

2 are the roots of f(x).(Y = a2 . = -a2 . (Columbia) Solution. (a) Suppose that {fa = There exist some non-zero elements ( Y ~ . acting on K ( x ) by a E G sending z to z a. (b) As K-modules. (c> Suppose f ( x ) = a. E K such that + n Hence and i#i which is not true. K [ z ]and K ( z ) have what dimensions? (c) Let G denote the additive group of K.zn -& + an-lxn-l + + a12 + ' . (We may check a2 . it is easy to see that the Galois group Gal(K/&) F A3.69 It follows that (Y. 1 a E K } is linearly dependent over K . (T: + -A K=&(a)+K ak+(Y2-2 is a generator of Gal(K/&). Hence K / Q is Galois since Char(&) = 0. .* a0 E K[z]. Show that the G-orbit of f spans a finite dimensional K-module if and only if f E K[z]. . . . .K and some distinct eleE ments a l . ( Y . Thus { f a = I a 6 K } is linearly independent. . a. So K is a splitting field of f ( x ) . u2. Let f E K ( z ) . (b) dimK K[x] = dimK K ( z ) = 00.. (a) Show that the rational functions f a = ( a E K ) are linearly independent over K. Assume that K is infinite..(Y 2 and . .2 is a root of f ( x ) directly by using the hint). 1408 Let K be a field and 3: an indeterminate. c) From a) and b).

h ( z ) E K [ z ]and ( g ( z ) h ( z ) )= 1. .70 We claim that for any n 1 distinct elements a l .. there exist PI. " .. a. + Pn+l(z + Qn+1li for 1 5 i < n.'-. Let f ( x + a l ) . . . . +an)(ai E K ) span the subspace < {f(x+u) I u E K } >.../3n+l)K n t l to be the solution of the E equation system + + + + Then it is clear that i 1 1 a2 - 0 .. a 2 .. . suppose f ( z ) E K ( z ) .*f . Then for any a E K . ...+1 . we take (.+a) This shows that { f(x + a i ) 1 1 5 i 5 n + 1) spans the subspace generated by the G-orbit of f . .f ( z + a 2 ) .+Pn+l(~+an+l)~ + .*.. . . a n + l in K (note that K is infinite).f(x . . For any a E G. On the other hand. Hence = P l f ( z + a l )+ P a f ( z + a 2 ) + . a. . . . f(. . .Pn in K such that f(x+a) = 9(" +a) h(x n + u) g(x+ai) = -i=lP i y n h(x + ai) .f ( z a2).P 2 . { f ( z + a ) 1 a E K } spans a finite dimensional K-module. . f ( x a l ) . f ( x an+l)spans the subspace generated by the G-orbit of f . . a.. a. and the G-orbit o f f . .01. .)a =P ~ ( Z + alli + /32(z + a2)a + .+1 a: a. We write f(x) = g ( z ) / h ( x ) where g ( z ) . . . . 1 ant1 a1 .. = P ~ ( Z + $ I+)P~ ( x + a 2 ) " z and also (Z +. ./32. + P n + l f ( 2 + ~ 1 n + l ) .

Let a and a’ be roots of e(z) and e‘(z) in some extension field of E . Thus all the irreducible factors of f ( z ) over E are of the same degree. Hence we have an isomorphism 7 : F ( a ) -+ F(a’) which sends a to a (u E F ) and CY to a’. which is irreducible over F . CY and a’ are roots of f(z). since a and a’ are roots of the monic irreducible polynomial e(z) and e’(z) over E respectively. B () a separable irreducible polynomial over F . Then. Suppose e’(z) is another monk irreducible factor over E. u(e(z)) is an irreducible factor of f ( z ) over E . Hence 7 : F ( a ) -+ F ( a ’ ) can be extended to an isomorphism 7 of E(a) onto E(a’). For any (T E G = Gal(E/F).71 Since (g(z -+ a ) .h ( z + u)) = 1. Then. we can write E = F ( P ) . for any u E G. but :E E is in G. we have h(z +. We prove in the following that all monic irreducible factors of f ( z ) over E arise in this way. ( c ozum bia) Solution. 7 : E ( a ) + E(a’) is an isomorphism and ~ ( a )a’.a) for any a E F . is a root of g z . Since E / F is Galois.)I i=l n n h ( z +(. we still denote u to be the isomorphism E[z] ---t E[z] which extends (T on E and maps x to z. Obviously E is a splitting field of g(z). where .Show that all the irreducible factors of f ( z ) over E are of the same degree. Thus we have 1409 Let E be a finite Galois extension of F and let f(z) be an irreducible polynomial in F [ z ] . an easy discussion in a splitting field of h ( z ) over F will lead to degh(z) = 0. Since K is infinite.Let e(z) be a monic irreducible factor of f(z) over E. a(f(z)) = f(z). q(P) is a root of q ( q ( z ) ) = g(z). = we must have (vlE)(e(z)) = e’(z) and dege(z) = dege’(z). Then E(a) = F ( a ) ( P )and E(a’) = F ( a ’ ) ( P ) are splitting fields of g(z) over F ( a ) and F(a’) respectively. Since f ( z ) E F [ z ] . + . It follows that Now. Note that y ( P ) may not be P.

Hint. + + + + in K [ t ] . First prove that the zeroes of t62 .t -c = f ( t ) .t . Show that [ F :&I = 3. g ( t ) in K [ t ]where f ( t ) is a monic polynomial of degree n. Obviously. f ( t ) = (t .1.1. we obtain n . G is a subgroup of F" and G is cyclic of order 6 2 since (t6' It follows that . n wp) u = v(n .c and let u E E be a root of this polynomial. 1 5 n 5 p .u .c then so is u+ 1. Since p .u .1 in F . (Indiana) S o htion.c in K [ t ]is either irreducible or splits completely into p linear factors over K.in). Suppose that t p .1 over Zg. So we have n . u = (v . (b) Let G be all the zeroes of t62 . (b) Let F be the splitting field of the polynomial t62 . . Comparing the coefficients of the term of degree n .il)(t . Thus we have tp-t-c= (t-u-m) +. n w p = 1. . [ F :as]2 3. Show that the polynomial tP . the prime field of K . u)E K . Then for any m EZ. u il + i2 . Let E be a splitting field of tP .1 form a cyclic group G of order 62.t .1)' = 62t6' = 2t61 # 0. Hint. If u is a root of t P .. u E K .72 1410 (a) Let K be a field of characteristic p > 0.i2) * * * (t . in E K .u .t . u = 0 and there exist integers v and w such that w .

We have to show [ K : L] is a power of 2. such that L/& is algebraic and every finite field extension K / L . So. . Then E* is a cyclic group of order 53 . & C_ L a'.1 = 124.1. m where m is odd. is a product of irreducible polynomials of degree 2. Since L has no proper odd dimensional extension field. 1412 Let K be a field of characteristic p # 0. Let f(a:) be the minimal polynomial of o over L. (Indiana) Solution. when decomposed in nZ[a:]. For this purpose. = 1411 (a) Suppose you are given a field L . (a) Let K / L ( K C a ' ) be a finite field extension. we must have m = 1. f ( z ) .1. The set {a:" J a: E K } is a subfield of K that is denoted by KP (no proof required).73 Let E be a extension field of Z5 such that [E : Z5]= 3. If [L : KP] is finite. that is. Thus we have E N F and [F :as] 3. X Now for any finite field extension K I L . [ K : L] = [ L ( a ): L] = d e g f ( s ) is even. (b) Show that such a field L actually exists. K a'. and [K : L] = 2n. By Sylow's Theorem.1 splits in E and E is a splitting field of t62 . there exists some element a: E K such that K = L ( o ) by Primitive Element Theorem. Then all the elements of G' satisfies t6' . prove that it is a power of p . it is in fact a power of 2. Let GI be its unique subgroup of order 62. By (a). So t6' . Show that every finite field extension of L must in fact have degree equal to a power of 2. If K' is the corresponding subfield of K I L . Let G = GalK/L and JGJ= 2n . then [ K : K'] = 2" and [K' : L] = m. Then & 2 L & ( and L/& is algebraic. we may assume that K is Galois over L. and so K' = 1. (b) Let L be the field of real algebraic numbers. Then f(a:) has no real root since f(a:) is irreducible in L [ z ] . G has a subgroup H of order 2n. Hence. K 5 a' is of even degree. the subfield of 1R of numbers which are algebraic over &. (a) Let L be an intermediate field between Kp and K .

* .. .74 (b) A subset B of K is called pindependent if for any finite set bl..bp is irreducible in KP(b1.+ .-l)] ) i=l = p”. there exists a finite set of elements {bl. . Hence [KP(bl. b 2 . b 2 ..b2..b.bn) L. . bz) : KP(b1. : It follows that B is pindependent....bn} such that = KP(b1.. (Indiana) Solution. .bi-1) for any ! 1 5 i 5 n (KP(b1. b2. we can assume that bi $ KP(bl-. By Zorn’s Lemma... .} of distinct elements of B . Then for any finite set { b l .bi-l)[t]. Since Bi is apindependent subset. and bi @ KP(bl. there exists some i. Since . . b i : KP(bl. such that { b l . * * .b. tP . there exist pindependent subsets. For example... then K contains a maximal pindependent subset B.b2. i€I . .. [ K P ( b l . b. .. (b) If KP # K . b m )K P ) =pm. if 6 E K\Kp.. .. . .b. bi-1) = KP when i = 1). b m } C Bi. (a) If [L : KP] is finite. . . K contains a maximal pindependent subset.... Prove that if KP # K .. (c) Prove that the set B of part (b) satisfies K p ( B ) = K .. of distinct elements of B ..bz-l)] = p. Without loss of generality.. . . Let B = U Bi. . Now suppose that {Bi I i E I} is a chain of pindependent subsets of K .. It follows that [L: KP] 1 [KP(bl. B = { b } is a pindependent subset of K .bi-l). . b 2 ...*. . . .) : L] n = n [ K P ( b l .

bm-l)] = [KP(bl. . Since pp'" . we have f ( t ) I ( t P r n . b = b. b 2 . .. that is.. E contains a subfield L such that K & L E and L N K ( a ) .t over K .p = 0.b. in B U { b } .* . Let (Y be a root of the irreducible polynomial f ( t ) in some extension field of K. .}. then and [KP(bl. . (Indiana) Solution. The reason is that. the independency of B U { b } contradicts the maximality of B . Thus we have K P ( B ) = K . Suppose K p ( B ) c K .. ..b.. show that f divides t P r n . [KP(bl. Let b E K\Kp(B).75 (c) Let B be a maximal pindependent subset of K . CB and we still : KP] : KP(b1. . b 2 . b. 1414 Let K I F be a finite extension of fields and let L and E be intermediate fields. Prove that if p does not divide [E : F ] then E L. b m } . . .. 1413 Let K be a finite field with pr elements ( p a prime) and n be a positive integer...*"-l = p"* .b. Since m I n.. then {bl. If m is an integer which divides n and f(t) E K [ t ]is an irreducible polynomial of degree m. .* .b. .b. Then (El = prnand [E : K ] = n.) * 1 .-l) : KP] Since B C B U { b } .bz.) .. . Let E be a splitting field of t p r n . a prime.t . (Indiana) . Hence there exists an element p E L C_ E such that f(P) = 0. .. for any distinct elements b l . f ( t ) is the minimal polynomial of p.. . have [KP(bl.t ) . .-l} say.-. with E / F Galois and [K : L] = p .* and if b E {bl.p. if b @ { b l . Then IK(a)l = prm and [ K ( a ): K ]= m. bm) : KP] = p m .. Then B U { b } is pindependent.

Let i=O (1) Prove K is a field. (Indiana) Solution. Ki+l is (0 €6' 10" E Ki for some n > 0). It is clear that i=l u Ki is a subfield of a'. Let a E E be a root of f(z). L. L/L)I = IGal(E/L dividers IGal(E/F)I = [E : F ] .76 Solution. For any c E G a l ( E . L/L) = Gal(L(a)/L). Prove that deg(f) 2 5. since L Now suppose E prime. L = L(E) = L(a). Let f(z)be a separable irreducible polynomial over F such that E is its splitting field. If i the smallest subfield of @ containing the set 2 0. (2) Let f(z) E K [ z ]be irreducible. L = K . . we have N_ G a l ( E . (1) Since KO C_ K1 C_ C_ Ki C_ Ki+l C_ M * * a is a chain of subfields of a'. contrary to the assumption. And further. L and n E)I [K : L] = p . Then E = F ( a ) and E . Let E * L = E(L) L ( E ) be the composite of E and L in K . Then E . 1415 Let Ki be the subfields of a' defined as follows: KO = &. + it is clear that g (E ~ Aut(E/L n E ) . Hence E L is Galois over L. It follows that C E . Then E L = L(E) is a splitting field of f ( z ) over L. a p = IGal(E. Thus we have E C L. L/L) Gal(E/L n E ) .

Let a be an algebraic element in K . Contradicts the irreducibility of f(z).h ( t in F[t] and ) where g ( t ) is a irreducible polynomial in F[t] degree k. 1416 Let K be an extension field of Fp. By the formulas for the roots of quadratic. Suppose that f(t) is reducible in F[t] let f ( t ) = g ( t ) . if aPm-l I . for some n > O}. (Indiana) Solution. the field with p elements. for some positive integer m.x. where g ( n ) = p"-l. P. E Fp. Let n = [ F p ( a ) Fp]. a d n ) E Fp. then On the other hand. Let E = F(cr).hence in K [ z ] . Let E be a splitting field of F p ( a )C E . If m and n are coprime then show that f remains irreducible in F [t].1 (Indiana) Solution. : Thus [Fp(a) F] is the smallest positive integer m such that : E Fp. Since ( ~ ( g ( ~ ) ) P -= : l apn-l = 1.Then IFp(a)I = p" and upn-' = 1. zPm - so u is a root of m zp . Let f E K [ t ]be an irreducible polynomial of degree n. There exsits some i such that f(x) E Ki[x]. cubic and quartic equations and Ki+l 1 (0 E(C 10" E K . Suppose d e g f ( z ) 5 4. Prove that [Fp(a) Fp] is the smaIlest positive integer : m such that E Fp. 1417 Let F 2 K be a field extension of finite degree m. 1 5 k < n.(u) : F. Then [E : Fp] = m and Fp n = [F.]([EpP]= m. Let f(x) E K [ z ] be an irreducible polynomial with d e g f ( z ) > 1. f(z)splits in Ki+3[z].77 (2) (Remark: d e g f ( z ) may be 1).(&4)P-l = 1. Hence deg f ( z ) 2 5. of . Hence z over Fp and a E E .

since cr is a root of f ( t ) and f ( t ) is irreducible in K [ t ] .n ) = 1 and 1 5 Ic < n.-. (Stanford) Solution. which contradicts ( m . . then & E is a cyclic extension of order 32.78 g ( t ) is irreducible in where cr is a root of g ( t ) in some extension field of F . the cyclotomic field &(zn) over Q is a Galois extension and [&(zn): Q]= +(n) where z. Then [E : F ] = Ic since F [ t ] . . then & C E is a Galois extension and Gal(E/&) 1 .)/&) N Aut(G) where G is the cyclic group of order n. is an n-th primitive root of the unit. It is easy to see that IGal(&(td)/&)I = 4(n) and Gal(&(z.z . Aut(G) is cyclic of order n . When n = 2" and m 2 3. [E : K ] = [E : K(cr)].Zz @ . As in 1418. : Taking m = 7.It follows that n I Ic . if we take E = Inv(&). On the other hand. for any positive integer n. When n is prime. 1419 Find a Galois extension E over & with Gal(E/&) cyclic of order 32.m . By the Fundamental Theorem of Galois Theory. For any positive integer n. . then E is a Galois extension of & with Gal(E/&) cyclic of order 16. So if we take n = 17. 1418 Find a Galois extension E over & with Gal(E/&) cyclic of order 16.So [E : K ] = [E : F ] [ F : K ] = Ic .1.)/&) E Aut( ) where G is the cyclic group of order n. (Stanford) Solution. is an n-th primitive root of the unit.Z2. It is easy to see that IGal(&(zi)/&)I = +(n) and Gal(&(z. it is well known that Aut(G) -. Z p .m. d(n)is the Euler &function. : n. the cyclotomic field & ( z n ) over & is a Galois extension and [ & ( z n ) : &] = d(n) where z . [K(a!) K ]= [E : K(cr)]. 4(n) is the Euler $-fur Ition. E = & ( z 1 7 ) . Thus f ( t ) is irreducible in F [ t ] .

79 1420 Let E / F be a finite Galois extension. : E + E . ..@) in E.. and u 1 ( z ) . (E M n ( F ) ) . 3c.) = C g ( a ) where varies over G.. f(z) is also the minimal polynomial of the F-linear map M .+ . Let z be a primitive element of E / F . M ... is similar to diag{ul(z). To prove that it suffices to prove that for any 1 5 k 5 n . .z n ..say. Consider the F-linear map M a : E + E. z . Show that its trace is given by Trp(M. a has the form since { l . the matrix of M . g n ( ( z )are distinct.. we have - n i=l . .1. is the minimal polynomial of z over F. (z) = z . (Columbia) Solution.. Z ( z ) . G = Gal(E/F) and a E E .a.z).l } is a base for E / F . ~ . for any a E E .zn--l}. .(z) = a z .uz(. Obviously. Now. M . relative to the base (1.. .an(z)} in Mn(E). Since f(z)has distinct roots C T ~ ( Z )g. [E : F ] = n and Then . . Anyway.

n This completes the proof.1.and for any 1 5 i 5 n . .

Part I1 Topology .

.

but A U B is not path connected. If A and B are path connected. say U . 2102 Suppose that A and B are compact subspaces of spaces X and Y respectively. such that A n B # 8. By the same reason. 3 : Then A and B are path connected and A n B = A .sin -) I 0 < z 5 1). Prove that there are open sets U c X and V c Y such that . It means that A U B is connected. The following example shows that if A and B are path connected then A u B needs not to be path connected. such that U c f-'(-l).a3 SECTION 1 PO IN T S ET TOPOLOGY 2101 Let A and B be connected subspaces of a topological space X. f l is also constant. and that N is an open neighborhood of AxBcXxY. need A U B be path connected? (Indiana) Solution. Hence f is not surjective. we may assume ~ that A c f-'(-l).l). ~ We have f(zo) = -1. Let A = ((0. Without loss of generality. But since zo E B.O)) C R2 and 1 B = {(z. Since A is connected. Prove that A U B is connected. there is a point of B which belongs to U . f ( must be constant. Since f is continuous. Therefore we have B c f-'(-l). Let zo E AnB. Let f be any continuous map from A U B to So = {-1. there exists a neighborhood of zo.

Let' z be a point of A . y) belongs to A x B and N is an open neighborhood of A x B c X x Y . Let X* = X U {m}. c On the other hand. For any y E B. and consequently is also a normal space. there exists asubcover { U ( z i ) . with A compact and B closed. Since A is also compact.since (z. { U ( z ) . . Therefore the family of open sets {Vy(z). Does there exist a continuous function f : X [0.. there is asubcover {VYi(z). . . X is compact ' and Hausdorff.(z)..(z) x V.. Since B is compact. . It is easy to see that U and V are open q X and Y respectively and that A x B c U x V c N . U U U ( z j ) and V = j=1 q n V(zj). p } such that B = c U V. there exist open sets U . (Indiana) Solution. Since X is locally compact and Hausdorff. i=l P Let U ( z ) = fi Uyi(z) V(z) = 0 V. We denote by X" the one-point compactification of X. ' Then by the Urysohn Lemma there exists a continuous function X -+ [0.1] such that f l = 0 and f l = l? ~ ~ + ( Cincinnati) Solution. If X is compact.(z). Y )E V. 2103 j= 1 Let X be a locally compact Hausdorff space. .il .y E B } covers B. j = l . and I) n It is obvious that U ( z ) and V(z) are open i=l sets of X and Y respectively and that {z} x B i=l c U ( z )x V(z) cN.(4 cN. . Let A and B be disjoint subsets of X .1] ' F: .(z) c Y such that ( Z C . . ( z ) c X and V. then X is normal and the existence o f f is obvious. z E A } is an open cover of A . Hence we may assume that X is noncompact.84 Ax B cU x V cN. q } such that A Let U = sets of j=1 u U(zj). It is easy to see that A and F = B u{m} are two disjoint closed subsets of X .

the restriction o f f on X .t’) if and only if { x = x’ and t = t’ = 0 or y = y’ and t = t’ = 1. are required for this problem. Y is Hausdorff and f : X + Y is continuous and onto. then f is a closed map. If X and Y are topological spaces. (Indiana) Solution.t) is equivalent to (d. and S’ * So is homeomorphic to the unit sphere S2. f. (a) Let X be a topological space and be an equivalence relation on X. the join of X and Y is the quotient space * Y = ( X x Y x [ O ) 1])/ x -) where (z. By the quotient map ?r : X -+ X / we mean the map which assigns to x E X the equivalence -- - . (a) So * So is homeomorphic to the unit circle S’.y. (b) Show that if X is compact. - 2105 (a) Define quotient map. Define by X / the space of equivalence classes under -. y.85 such that f l = 0 and ~ the requirements that - f l = 1. What space ’ are they? (b) Describe X * So for a general space X . X * So is homeomorphic to the quotient space X x [0. (a) So * So and S * So are homeomorphic to familiar spaces. only the correct answers to the question asked.1]/ obtained from the cylinder X x [0.satisfies = 0 and f l = 1. (Indiana) Solution.1] by collapsing X x (0) and X x (1) to two points p and q respectively. ~ ~ 2104 - No proofs. Therefore. (b) Generally. (c) Show that i f f satisfies the condition of (b) then f is a quotient map.

Then V is an open neighborhood of C in Y . A is compact too. i is obviously a 1 . - - - - - - N 2106 Let f : X Y be a continuous function from a space X to a Hausdorff space Y . It follows from the continuity of f that f ( A ) is a compact subset of Y .1 map. and. the quotient space X / is compact. is obvious that It f-yc)= {n I 72 EX}. (Indiana) Solution. and let U be an open neighborhood of f-'(C) in X. then W is closed in X . and consequently is a closed set of Y because Y is a Hausdorff space. (b) Let A be a closed subset of X . Hence f is a closed map. . Then we define a map i : X/ -4 Y by i([x]) = f ( x ) . is a homeomorphism. f ( A ) is closed in Y . we see that ---f f-'(v) = x . W is compact too. r becomes a continuous map. Since Y is Hausdorff. Since f is onto. Since W c f-'(f(W)). therefore. It is clear that f = i o T . Take C = (1) E S1. f ( W ) is a compact set of Y . - (b) Let X = R and Y = S1. Hence i is a continuous 1 . Denote by [z] the equivalence class containing 2. More generally. there is naturally associated an equivalence relation on X such that z1 5 2 if and only if f(x1) = f(z2). Since f is a continuous function. (b) Give a counterexample to show that if X is not compact. If Y is homeomorphic to X / under the map i : X/ --+ Y and f = i o T then we call f a quotient map.f ( W ) . The quotient space X / may be topologized by defining a subset U c X/ to be open if and only if r . Let C be a closed subspace of Y . So f is a quotient map. Under this topology. then there need not be such a neighborhood V .U . Since X is compact. (a) Prove that if X is compact then there is an open neighborhood V of C in Y such that f-'(V) c U .' ( U ) is open in X .1 map from the compact space X / to the Hausdorff space Y . Since X is compact. By the result of (b). if f : X + Y is a continuous map.86 class containing x. Let V = Y . : X -+ Y is the continuous function defined f by j ( t ) = earit for t E R. (c) Let be the equivalence relation on X associated to the map f. (a) Let W = X .f-l(f(w))c x w = u.

It means that there exist two disjoint closed sets A and B in X such that they cannot be separated by disjoint open sets in X . Thus we see that R ( W ~ ) and R ( W ~ ) two disjoint open sets in Y and contain U and V respectively.' ( V ) are two nonempty disjoint closed sets in X .' ( C ) n €I in X . then.87 Let U = U Un be the open neighborhood of f . Let X be a connected space. n - one can easily prove that there does not exist such a neighborhood V. 2108 Let p : E B be a covering map with E locally path connected and simply connected.=(--+n. Let U and V be two nonernpty closed sets in Y such that U n V = 0. If yo $ U U V. 2107 Let X be a normal topological space and A c X a closed subspace. Then the quotient space Y obtained by collapsing A to a point yo is not regular. (a) Show that the quotient space Y obtained by collapsing A t o a point is normal. it is clear that there exist two disjoint open sets W1 and Wz in X containing R-'(U) and R-'(V) respectively such that WinA = 8 for i equal to 1 and 2. where 1 1 U. by the normality of X . because one can prove that the point yo and the closed set R ( B )in Y cannot be separated by disjoint open sets in ----f Y. Then R-'(U) and R . n n Since n-w lim lUnl = 2 = 0. --f . (b) Let X be a regular space which is not a normal space. let f : X -+B be a continuous map. (a) Let R : X Y be the identification map and yo E Y be the point which A collapses to. are If yo E U (or V).n+-). (b) Does this result hold if normality is replaced with regularity? (Indiana) Solution. we only need to take the sets W1 and Wz without the restrictions that Wi n A = 0.

x I s f 1 ( z )= f 2 ( z ) } . where bo = p ( e 0 ) . Z n } . . It is obvious that A is not empty. Let c ( z )-be the path component of %. i=l fi p ( c ( z ) )is an open neighborhood of z in X . i=l Choose another elementary neighborhood V(z) of z such that v ( ~cc ) np(w).X E A} is an open covering of X .l ( b o ) . It is obvious that each & x is open and R(z) n (x) = 0 for () i # j . Let A={ E . we see that A = X. we see that e l E p-l(b0).(z) n U i containing Zi. Prove that 2 is compact. 2109 Let p : X -+X be an n-sheeted covering projection. let p . Thus. n X is compact. Choose ix is a U . n . Let W ( z )be an elementary neighborhood.88 and f l . Then eo E p . For any point z E X.' ( W ( z ) ) is mapped topologically onto W ( z )by p . Therefore. W(x) is an path-connected open neighborhood of x such that each path component of p .(z). By the assumptions p : E -+ B is the universal covering map.' ( W ( z ) ) such that Zi E w.' ( z ) = ( 1 . (Indiana) Solution. 2 . f2 : X -+ g :E +E E be two lifts of f . E Ui. Since p is an open map. Let f2(20) = e l . E U such that Z . Since p f 1 = p f 2 = f .e. which means fi = g f l : X -+ E . Prove that there is a deck transformation such that fi = g f 1 . Take a point zo E X and let f l ( z 0 ) = eo E E . Thus there exists a deck transformation g such that g(e0) = e l .. by the connectedness of X . Suppose that (In dian a) Solution. - < 00. i. . Suppose that U = {UA. It is not difficult to prove that A is both-open-and-closed in X. g f l ( z 0 ) = f z ( 2 0 ) . Let - where w ( )a path component of p ..

Let U be any open neighborhood of 3:o in ( X . (Recall that 7 c U means that every set in the topology 7 is contained in U. It is obvious that h(z0) = zo and h ( U ) = U . . Show that if C is a connected subset of X that intersects both A and X . . Thus h is a homeomorphism from (X.. X i = 1. For any point 20 E X .U is also an open neighborhood of z o in ( X . Since 7 c U . We use the reductio ad absurdum. ' T ) is the identity map of X. I). Since C n A c U and C n ( X . i=l - 2110 Let 7 and U be two different topology on X such that X is compact and Hausdorff with respect t o both. Since X is compact. nIt follows that p-'(V(z)) c . and consequently X can be covered by a finite subcover of U. This contradicts the assumption. U ) denote the topological spaces of X with respect t o 7 and U respectively. Then it is easy to see is T(z)c E(z) c Ui for i = l . 7 ) . ) (Indiana) Solution. 6 Ui. (Recall that BdA = 2n (X.A ) . 2111 Let X be a topological space and let A c X. U ) to the Hausdorff space (X.89 Let P-'(v(zc>) = where that n (JW).U) ( X .can be covered by a finite number of V(zi). 7 ) and ( X . That is. U ) . Hence h is continuous at 2 0 .A . then C intersects BdA. . Suppose that C n BdA = 0.m. we have proved that for any point z E X there exists an elementary neighborhood V(z) of 3: such that p-l(V(z)) can be covered by afinite number of sets in U. continuous map.. i=l c(z) the path component ofp-l(V(z)) for any i.A ) .+ . Let ( X .) (Indiana) Solution. Then h is a 1 . . Prove that 7 p U. U which means U = 1.1 map from -+ the compact space ( X . We use the reduction to absurdity.)to (X. Suppose that 7 c U. We claim that h is a 7).A ) c V . h : (X.. Take U = C n x and V = C n ( X .

a ) > 0. . we see that both U and V are nonempty subset of C. u ) < E .2] c R and A = [0.(A) = {z E {z E A : d ( z . A which is a contradiction t o the connectedness of C. Since A is compact. a. show that C ( A ) is closed in X . E A such that d(z. If A is not compact. Must C ( A ) be closed . 1). Thus we have --+ which means 20 E C .< ( E . Then. Take X = [0. ( A ) . a ) ~ d ( z . we give a counterexample as follows. d ( z . Then C a ( A )= [0. 6 = E . + 4 6 which means that 06/4(zo) c O . ( A ) . .. ( A ) . and.d(z0.) 5 E. z o ) + d ( z o . (a) Let 20 be any point of O . z o as n + 00. A : d(z. -+ a for some a E A . we may assume that a.a)5 E for some a E A } . By the definition of O . a ).. But u nv = c n A n ( X - A ) = C n B ~ = 0. d ) be a metric space. for . that both U and V are open sets of C. i. So C .from the assumption. (b) If A is compact. For any subspace A E cX and real number > 0.d ) < E . (a) Prove that O . for a general subspace A of X ? (Indiana) Solution. Thus O . ( A ) . there exists a point a E A such that d(z0. for any 2 E 06/4(z0). each z there exists an a. ( A ) is an open subspace of X . It is clear that C = U U V and both U and V are closed subsets of C .e. consequently. 2112 Let ( X .(A) = C. let O. I g) . ( A ) is closed in X . without loss of generality. ( A ) such that z # zo for any n and 2. is not closed in X . By the definition of C E ( A ) .} in C . Then there exists a sequence {z. (b) Let zo be any cluster point of C e ( A ) . ( A ) is an open subspace of X . a ) < E for some a E A } .

(Indiana) Solution. Y is not a Hausdorff space. On the other hand.b. Thus A would be a nonempty subset of X . b) Give an example where U and V are not open in X and f is not continuous. { b . X = U U V . A X . Let Y = { a . This contracts the connectedness of X .' I u ( N ) and f . then there would exist a nonempty proper subset U of Y which is both-open-and-closed in Y. But U and V are open in X .Since flu and fv are continuous. therefore. { a . Since any neighborhood of the point a always contains the point c. then Y is Hausdorff.U is open in Y . If Y were not connected. C l . { c } . el.Prove or give counterexamples to the following assertions: (a) If X is Hausdorff. and f : X -+ Y be a function so that flu and flv are continuous. then Y is connected. b. a) This assertion is not correct. It means A # X . Since X is dense in Y and U is open in Y . respectively. But it is easy to see that the subspace { c } is dense in Y and is Hausdorff. Thus fF1(N) = f-lIv(N) uf-'Iv(N) . Since X is dense in Y and Y .U ) would be an open set of X . We give a proof to it as follows. c}. f . would be a nonempty open set of X. (Indiana) Solution. a) If U and V are open in X .A = X n (Y. The topology on Y is determined by the family of open sets 7 = { { a .U is open in Y . Let A = X n U . We give a counterexample as follows. (b) If X is connected.' I u ( N ) and f .' l v ( N ) are open in. because Y .l I v ( N ) are also open in X. show that f is continuous. 2114 Let X and Y be topological spaces.91 2113 Suppose that X is a dense subspace of a topological space Y. f . c ) .A is nonempty. which is both-open-and-closed in X. b) This assertion is true. X . a) Let N be an open set of Y. U and V .0).

= For otherwise.2].q . f : X +R is defined by f(x) = {x 2 XEU. 2115 Let q : X --t Y be a quotient space projection from a topological space X t o a connected topological space Y. f is continuous. but f is not continuous. V = q . Then q ( U ) n q ( V ) 0. U = [0.y) E V . a) Suppose that X is not connected. Thus X must be connected. y ) = 1 for (1. -() Since q is a quotient mapping.y) E R2 11 5 y 5 2. But X is not connected. which contradicts the connectedness of q'y. U and V such that X = UUV. XEV. 0) = x for ( x .Y Y ) = (q-l(y) n U ) u (q-l(y) n V ) . ) The topology of X is induced from the topology of R2. Let X = U U V . O ) E U and q ( 1 . b) Is X necessarily connected if the map q is not assumed to be a quotient mapping? Justify your assertion. For each y E Y.l ( y ) is connected. ( Coturnbia) Solution. a) Show that X is connected. the unit interval of R. b) The following example shows that the assumption that q is a quotient mapping is necessary for X to be connected. q ( U ) and q ( V ) are disjoint nonempty open subsets such that Y = q ( U ) U q ( V ) .which contradicts the connectedness of Y .21. q .92 is open in X .l ( q ( V ) )and U = q-l(q(U)). . Let Y = [0. Assume that q-l(y) is connected for each y EY. 1) and V = [l. and : X --+ Y be the map defined by q ( x . b) Let X = [0. and consequently. O ) E R2 10 5 and 2 5 1) v = ((1. where U = { ( x . let y E q ( U ) n q ( V ) . Therefore. Thus. there exist two disjoint nonempty open subsets of X .1]. Then q is continuous but is not a quotient mapping. Then f (u and flv are continuous. and it is obvious that q-l(y) n U and q-l(y) n V are both nonempty open subsets of q-l(y).

1} is the O-dimensional sphere with discrete topology.i. which means that p is a closed map.p ( U ) is an open set of X . p ( V ) is closed in X . = ) i=l n n Then W(z0) is an open neighborhood of z o in X . Let U c X x Y be a closed subset and zo E X . ' . and let 2 be a set such that Y is a subset of 2 and 2 is a subset of the closure of Y .(y) of X and open set V. Since Y is connected.e. (Y)z= Y n 2 = 2.(ZO) and (Wzo(y)x V. and it follows that f is not surjective. we get (Y)z f-'(l). Let f : 2 -+ So be any continuous map.(ZO) of Y such that ( Z O . ( Cincinnati) Solution. W(z0) C X . Show that p is a closed map. we may assume that f ( Y ) = {l}. Y c f-'(l).(~~). . and consequently. But c as well-known. there must exist a finite number of V. It means that there does not exist any continuous surjective map from 2 to So and therefore 2 is connected. 2117 Let Y be a connected subset of the topological space. i.p ( U ) . we have Y c 2 c F.. w ( ~ ~w~.(zo))nU = 0. According to the assumptions. .93 2116 Let X and Y be topological spaces..VYn (zo) such that n y = Let i=l u VYi(XO). Y ) E Wz0(y) x V.p ( U). we see that W(z0) n p ( U ) = 0. Since (W(z0)x Y )n U = 0.Then. Taking closures of these two sets with respect to 2 and noting that f-'(l) is a closed subset of 2. ( 2 0 . with Y compact. without loss of generality. Let p : X x Y -+ X be the usual projection onto the first factor. for any y E Y . Thus X . there exist an open set Wz.y) 6 U .e. Since U is closed.. Thus we have 2 c f-'(l). where So = {-1. (zo). Prove that 2 is connected. Since Y is compact. (Minnesotu) Solution.

a) Prove a metric space X has property S if X has a dense subset with property S. is connected.. consequently. Then. a both-open-and-closed subset of X\U. and. By the above fact. for any E > 0. Let X\C = U U V where U and V are two disjoint nonempty both-open-and-closed subsets of X\C. a) Suppose that X has a dense subset A which has property S. U is a connected subset of X . Thus W1 is a both-open-and-closed subset of V . Must X have property S? ( Cincinnati) Solution. b) Suppose X is a subset of a metric space. which contradicts that C is a component of X\A. has diameter < ~ / 2 . Now suppose that X\C is not connected. Since U is connected. Since C U U c X\A and CnU = 0.Since each V. Hence W1 is a nonempty both-openand-closed subset of (U U V )U (X\U) = X . We first prove that if X is a connected space. we may assume that A C V. Since A c X\C and A is connected.we see that C U U is a connected subset of X\A containing C . If C is a component of X\A. Then let U U V = W1 U W2 where W1 and W2 are two disjoint nonempty both-open-and-closed subsets of U U V. 2119 1) A metric space X has property S if for every E > 0 there is a cover of X by connected sets each of which has diameter < E . Hence X\C must be connected. and if V is a both-open-and-closed subset with respect to X\U. is also connected va . Suppose the closure of X has property S. we may assume that U c W. So the above statement is proved. there is a cover {Val& E I?} of A by connected sets such that each V.94 2118 Let A be a connected subspace of a connected set X . which contradicts the assumption that X is connected. Suppose that U U V is not connected. show that X\C is connected. ( Cincinnati) Solution. then U U V is connected. C u U is connected because C is connected. without loss of generality.

Then = R has property S. (i) X is shown in the Figure below. Fig. b) We give a counterexample as follows. s i n r / z ) 10 < z and 5 l} 6 = inf{z 1 (z7sin. ' Thus X has property S.y) 1-1 5 y u{(z12) 5 2) I 5 5 '1 {(llY)7O 5 y 5 2.lr/z) 10 < z 5 1) u ((0.2.rr/z) E A } . (i) Sketch X .} = 0.sin. c R 2 . but it is easy to see that X does not have property S. (ii) Let f : X -+ X be continuous. x 2120 Let X be the topologist's sine curve defined by X = {(z.95 and obviously has diameter < E. Since = X and = U Val we see that aEr {ValQ: E I} is a cover of X by connected sets each of which has diameter < E . Let R denote the euclidean real line. Show that either f ( X ) = X or else there exists 6 -- > 0 such that 3 3 < y 5 . . 22 (Toronto) Solution.1 (ii) Let A = f ( X ) n { ( z . X be the set of all rational numbers.

we see that f ( X ) = X.11.-- 3 3 < y 5 -} = 0. (ii) Show that T cannot be covered by 2 contractible open subsets. we claim that f ( X ) = X .I and U3 = X . U2 and U3 are contractible open subsets and T = U ~ u U .2 Thus let U1 = X . (See the Figure above. (i) It is well-known that the torus T can be identified to the quotient space of a square X obtained by identifying opposite sides of the square X according to the directions indicated by the arrows as shown in the Figure below. Hence in this case there exists a 60. using the fact that f ( X ) is path connected. such that Therefore it is easy to prove that the set {(O. UZ = X . f ( X ) is also path connected. If 6 > 0. we first note that since X is path connected. 2 2 2121 Let T = S1x S1denote the torus. n a Fig. 0 < 60 5 1. To prove it. 2.) Then U1. ( Toronto) Solution.96 I f 6 = 0.Y) n 10 < 2 < 6. U U 3 .Y) I -1 IY I 1) c f ( X ) * Once again. .{ u U b } . then it is obvious that f(x) {(z. (i) Show that T can be covered by 3 contractible open subsets.

= 0. since U is locally finite. . b) For any point p $! vp.b) is any open interval of R and f E F } . a) By definition. for each pi. for any subset K c J . Up. It contradicts the known fact that a. Hence is closed.. Then let V = W n V1 n .}aE~ an open cover of the space X. . U PEK r p is closed. Then there exists a unique topology 7 on the set S of which U is the topology subbase. only a finite number of Up for /3 E K . . . It is easy to see that 7 is the weakest topology on S amongst all those for which all members of F are continuous. . From the Van Kampen theorem it would follows that the fundamental group ?rl(T) = 0.’ ( ( a . there is an open u PEK neighborhood W of p which intersects only a finite number of sets in U . Since p $! r p . 2122 Let U = (U. n V..particularly. It is clear that V is an open neighborhood of p such that V n ( up) = 0. (Hama4 Solution. Prove that there exists a weakest topology in S amongst all those for which all members of F are continuous. Show further that the resulting topological space satisfies the Hausdorff separation axiom. Let U = { f . . Suppose that s1 and s2 are . U is said to be locally finite. be a) Give the definition for “U is locally finite”. say. nupi u u vp PEK P€K 2123 Let S be a set and let F be a family of real valued functions on S such that f ( s 1 ) = f ( s 2 ) for all f E F implies s1 = s2. b ) ) 1 ( a . Up. .97 (ii) Suppose that T could be covered by 2 contractible open subsets. ( Toronto) Solution. if each point p E X has a neighborhood which intersects only a finite number of U. there is an open neighborhood V . b) If U is locally finite show that.(T) x Z @Z.. of p such that V .

98 two distinct points of S. bl)) and Uz = f .7) such that si E U. So (S.b z ) ) are two disjoint open sets of (S.bi) for i = 1 .Hence we may take two open intervals (a1. Then n Ui = f-'((al. By the assumption there exists a t least a n f E F such that f(s1) # ~ ( s z ) .bl) ( U Z . is . 2 and (a1.7) Hausdorff.2.for i = 1. b z ) such that f ( S i ) E (ai. bz) = 0.bl) and ( a z .l ( ( a z .

Solution. where q is the homomorphism induced L .2. TI)). and let U = T . by the Van Kampen theorem.99 SECTION 2 HOMOTOPY THEORY 2201 a) Give generators and relations for the fundamental groups of the torus and of the oriented surface of genus 2.{y}. Under the identification the sides a and b each become circles which intersect in the point 20. we conclude that ?rl(T. I ) T is isomorphic to 7~1(U. as shown in Fig. b) Compute the fundamental group of the figure 8 and draw a piece of its universal covering space. modulo the smallest normal subgroup of ?rl(U.21) z1) containing the image +*(rl(UnV. Since V is simply connected.3 (b). Let V be the image of the interior of the square under the identification. we can present the torus T as the space obtained by identifying the opposite sides of a square. Let y be the center point of the square. a> As is well-known.

and. By the Van Kampen theorem we can prove that a l ( X .p2].3 (b).2. consequently. ~ where 6 is the equivalence class of a path d from 20 to 21. The following picture is a piece of its universal covering space.2.4 Let X denote the figure 8 space. and [ailpi] denotes the commutor aipiailpzrl.p1.p. where a l . Hence a l ( U . respectively. P are presented by circles a and b. Let U = X .zl) is an infinite cyclic group generated by 7 . Thus ?rl(T.2.aa. it is easy to see that x l ( U n V.{ p } .20) ’ is a free abelian group on two generators a and fl. respectively./31][a2.) b) Fig .ba.100 by the inclusion map q5 : U n V -+ U . where a and p are presented by circles a and b. bl. as shown in Fig. which are presented by circles a and b. 2 1 ) ) is just the commutator subgroup of 7rl(U. It is easily seen that a u b is a deformation retract of U . each level segment is mapped on the circle a . 1 ) is a free abelian group on 2 two generators a and p’. V = X . we see that rl(T. (See Fig.20) is a free group on two generators a and p.~ f l ‘ .{ q } . that q5+(y)= c y ’ p I ~ y ’ . It is also clear that T ~ ( U . z o ) is a free group on two generators a. where a . respectively.p2 with the single relation [a1. (See Fig. In a similar way.2. Changing to the base point 20. / 3 2 are presented by circles a l . The smallest normal subgroup of a l ( U .the equivalence class of a closed path c which circles around the point y once. 21) containing & ( a l ( U n V.2.~ . Z is )a free group on two generators a’ = F1a6 and p’ = C1p6. a2. respectively. p 1 . a 2 .21).) On the other hand.3 (a). we can see that the fundamental group of the oriented surface of genus 2 is a free group on four generators a1.4.5 Under the covering map T. it Fig .

f"(A) and such that the induced map y : S1+ X . which is an open subset of A. E f"(A) but 2. Since A is obviously compact and Hausdorff. where U and V are disjoint nonempty closed subsets of B. consequently. i) Since X is compact and A is closed. n times 00 f"(A) is connected. Thus it is easy to see that f"(A) is compact. For each positive integer n let fn(A) = ~ o ~ o .fn(A) ' (y'(s) = y(s) for s E S') is also nullhomotopic. Wl and Wz such that U c W1 and V c W2.B is open. such that x. Thus f N ( A )c W . Then B = U u V . Show that there exists a positive integer n such that y(S1) c X . closed subspace of a compact Hausdorff space X . o ~ ( A ) . which is a contradiction. and therefore. there exist disjoint open subsets of A . n=l n fn(A) m PYA) = ( f N W n Wl)u (fN(A) W2>. there would be a squence in A . is a compact subset..101 according to the direction indicated by the arrow. n which contradicts the fact that fN(A) connected. Suppose that B is not connected.W for every n. n=l ii) Suppose y : S1 X . and each vertical segment is mapped on the circle b according t o the direction indicated by the double arrows.B be a homotopy between y and the constant map. Noting that . . It is obvious that U and V are also closed subsets of A. We claim that there exists a positive integer N such that fN(A)c W . A is also compact. Therefore.. Since F(S1x I) is compact and X . It means that x. Noting that fn(A) is compact for every n and fn+'(A) c fn(A).W is closed in A and. 2202 Let A be a connected.+'(A) c (i) Show that B = n f"(A) c A for any n. (Indiana) Solution.B is a nullhomotopic map. A is a normal space. Since A . can see we that xo E B c W . closed and connected. E A . and suppose f : A -+ A is a continuous map. there exists at least a limit point 20 of the sequence { x C n } . Thus it follows that B c W1 U Wz = W . {xn}. @ W for every n. F(S' x I) is also closed in . we see that B = is a nonempty closed subset of A . and that A is compact. is ii) Let F : S1x I 3 X . For. otherwise.

such that pf = f . say f. 2203 Let RP" be the real projective n-space and T" be the rn-torus S1 x . c F(S1 x I ) c u c Particularly. It is well-known that R is a universal covering space of T". So homotopic and consequently f is also null-homotopic. x S1 ( m factors). Thus there exists a lifting of f . (Indiana) Solution. It is also well-known that " - ai(T") M Z e .T(z) tpo.t). there exists open sets U and V such that F(S1x I ) c U and B c V . (rn times) and that a l ( R P n ) M Z2. for any continuous map f : RP" -+ T". Define H : RP" x I -+ R" by - H(z.102 X . Therefore. cX . Let P : " R -+ T" be the universal covering map. . y(S') x . .fN(A). From the proof for i). it follows that there exists a positive N such that fN(A) V . Since X is normal.fN(A). the induced homomorphism f+ : a 1 ( R T ) -+ nl(T") is trivial. } x to a point S1c S1 x S1 . Let po be a fixed point of R".The remainder of ii) is obvious.t) (1 . Therefore. Then H is a homotopy between f and the constant map PO. Prove that any continuous map RP" -+ T m is null-homotopic. - + fl is null- 2204 Let X be the quotient space obtained by collapsing { p t ..

(See Fig.2. where S is the equivalence class of a path d connecting 20 and z1. It is easy to see that &(y) = lrl(u. 2205 (i) Suppose n 2 2. which is obtained by identifying the sides of a 2-gon. and V be the interior of the 2-gon. z l ) is isomorphic to the quotient group of a l ( U . by the Van Kampen theorem. Let y be the center of the 2-gon.103 Compute x1(X) and H . zl)is an infinite cyclic group on the generator y. The conclusion is that H i ( X ) is an infinite cyclic group for i equal to 0 . U .2. and V is simply connected.2. containing the 21) image q5. ( X ) .6. 2 and is zero otherwise.zl).V). 7 r l ( X . Therefore we conclude that . and U n V are path connected. It is easy t o see that u is a deformation ratract of U . Then. Does there exist a continuous map f : RP" -+ S1 which is not homotopic to a constant? (iii) Let T = S1 x S1 be the torus. Does there exist a continuous map f : S" + S1 which is not homotopic to a constant? (ii) Suppose n 2 2. Does there exist a continuous map f : T -+ S1 which is not homotopic to a constant? ( Toronto) . a F ig. 2 1 ) modulo the smallest normal subgroup of R ~ ( U . where y is the equivalence class of a closed path c which goes once round the point y.6) It is also clear that x l ( U n V. U = X . To compute H .(al(U n V. 1 .{y}. where q5* is the homomorphism induced by the inclusion map q5 : U n V -+ U .l(X) =z. V. al(U.6 X may be identified with the space shown in Fig. (Columbia) Solution. we may apply the Mayer-Vietoris sequence to the pair ( U . Thus. Therefore. U and V are open subsets. zl)is an infinite cyclic group on the generator S-luS. zl)). ( X ) .

. bo) (Indiana) in which the homomorphisms are i# and p # . (i) Let cr = [f] E ~ l ( B . 20 5 t l . e 2 n ). Hence f* is not trivial. -+ . prove that p# : ?rl(E. Define it f : T + S 1 by f ( p i t l . bo E B be a base point. has a lifting RP” -+ R such that a o f = f.rrl(B. e 2 x i t z ) = e27ritl E sl. (i) If F is path connected. b o )where f : I -+ B is a closed path a t bo . Solution.104 Solution. 2206 A continuous map of topological spaces: p : E B is called a fibration if it has the homotopy lifting property . Let p :E B be a fibration.l ( b o ) (the “fiber”). hence f is homotopic t o a o = C. f must be homotopic to a constant.eo) + Tl(E. ?must be homotopic to a constant map : S” -+ R. By the same argument as in (i). 7 c 7: + It is easy t o see that the induced homomorphism f* : H1(T) -+ H1(S1)maps one generator of H1(T) to the generator of H1(S1).that is. t z 5 1. there does not exist any continuous map f : Sn -+ S1which is not homotopic to a constant map. we see that there is a lifting of f. for any pair of continuous maps I --t G : X x I -+ B and h : X x (0) --t E such that p h = Glxx(oi there exists a continuous map H : X x I E such that H l x x i o ) = h and p H = G . Since R is contractible. = and f : S” --+ S1be a continuous map.e.any continuous map f : RPn -+ S1 induces a trivial homomorphism f* : x 1 ( R r n ) a l ( S 1 ) . i. which means that f is not homotopic to a constant. Since al(S”)= 0 for n 2 2. Let i : F E denote the inclusion map. Let G : I x I B be a continuous map defined by --f .eo) + “1(B. respectively. is exact. (ii) Prove that in general the 3-term sequence + --+ + --+ Tl(F.and another generator to zero. consequently. F = p .bo) is surjeceo) tive.and. which represents a. and eo E F .which is also a constant map from S” to S1. (i) Let a : R -+ S1be the universal covering map defined by ~ ( t )eZrrit. : S” --+ R such that T? = f. (iii) Denote T = S1 x S1 by T = ( e 2 r r i t l . (ii) Since a l ( R P n ) = Zz.

pg2 and pgz' are all constant path at bo. 1) = f ( t ) . Let C : I -+ E be a path in E defined by c ( t ) = H ( t . Thus f = g2 * c * g1* gF1 is a closed path a t e l .7 2207 Is the canonical map q : S2 --+ R P 2 (which identifies antipodal points of S2)nullhomotopic? Why or why not? (Indiana) . 0) = ( p T ) ( t ) . 0) = bo. Since F is path-connected. h . Let c ( 0 ) = e l and c(1) = e2. gl(1) = e l . On the other hand. It is obvious that p# .Then c1. Therefore there exists a continuous map H : I x I + E such that H / I x { o }= h and p H = G . c kerp#. It is easy to see that f is homotopic to C.2. suppose that G = [f] E kerp#.O) eo. ~ ( 1 = c i 1 ( 0 ) and c i 1 ( l ) = eo.{1} and c3 = H l { l } x ~ . Let c1 = H I { o i x r . b 1 . i#(a)= [pf] = [bo]. i.7.e. cg = H11. (See Fig. kerp# c imi#.eo). we may choose two paths g1 and 92 in F such that gl(0) = e2. where bo is the constant path a t bo. s ) G ( l . Particularly we have p H ( t . s ) E bo for any s. - [fi Fig.) Therefore. ~ ( 1 = cz(O). It is obvious that p h = GII. cg and c3 are paths in F and cl(0) = eo. s ) = f ( s t ) for ( s .Irl(F.{o}. t ) E I x I . It is clear that e l and e2 belong t o F. l I = b = [fl = a . G = = [c] = i#[c].2. Hence im. 0) = f(t). and h : I x (0) --+ E be the constant map such that h(t. 1). (ii) Let a = [f] E . there exists a continuous map H : I x I --+ E such that P H = G and H ( t . and G(0.105 G ( t . where gF1 is the inverse path of g2. Hence the sequence mentioned above is exact. Noting that pgl. l which means that p# is surjective.G(t. Then pc = f . By the homotopy lifting property. g2(0) = e l and gz(1) = e l . . ) ) Hence c = c1 x c2 x c i l is a closed path in F with base point eo. bg11 .. we have P# rfi = bfi = bg21 . Then there exists a homotopy G : I x I --+ B between pTand the constant path bo such that G(t.

o T .. Let T : B + B be a map so that T n = Id and so that T(b)= bfor some b E B.. Therefore. = Id. This is a contradiction. - - - - - 2209 Let T 2 = S1x S1. .) Show that there is a map ? : E --+ E so that p o T = T o p and Tn= Id. It is well-known that q is also the universal covering map from S 2 to R P 2 .= l T ( p ~ n . it follows that identity map Id : (E. We have T o p ( e o ) = T ( b ) = b. (Here T n = T o T o . Since I - T" = ( p T ) and . S1.106 Solution. Let i : X --t T 2be the inclusion map. Thus we conclude that q is not nullhomotopic.and therefore q has a lifting : S2 -+ S2which is also homotopic to the lifting of c . a constant map C : S 2 + S 2 .eo) + ( E . we have d e g c = degZ = 0.. Since ?rl(E)is trivial. Prove (Indiana) Solution. We use the reduction to absurdity.l = .- ~ " . = T~ O ) - P T " ( e 0 ) = T"-'(T(eo)) = T*-'(eo) = . But it is obvious that f is the identity map or the antipodal map from S 2 to S2. Then roi : X --+ X is . let X c T 2 be the subset S1 x { 1) U (1) x and that there is no retraction of T 2 to X . Suppose that q is nullhomotopic. by the uniqueness of lifting. Then q is homotopic to a constant map c : S 2 --+ R P 2 . n times.eo) such that p o ? = T o p.eo) is obviously a lifting of T n p a t the base point eo. On the other hand.. < 2208 Let p : E B be an universal cover with E and B path-connected and locally path-connected. denoted by r. there is a lifting of T o p . = T ( e 0 ) = eo. Suppose that there is a retraction of T 2 to X . On the other hand. we conclude that r?. Choose eo E p . + ( E . Hence degg = f l . T : ( E . . ---f - - (Indiana) Solution.l ( b ) and consider the map T o p : E + B . we see that @ is a lifting of T" o p at the base point eo.. eo) since T" = Id.

A . 2210 Let A c R3 be the union of the x and y-axis. Then we have Y = U U V .A ) .A . p l and p2 denote the points (1.: r l ( X ) -+ r l ( X ) is the identity homomorphism.6 = -2. Hence ~.p2) is a deformations retract of R3 . 0) respectively. It is easy to see that U has the .Z) I (Y2 + z2)(x2+ 2)= O}. O .y. p ) is not abelian. Take U = Y . and let p = ( O .1(X) is an non-abelian free group on two generators denoted by a and b. The Euler Characteristic K ( Y ) = 4 . b) Let e be a point on the arc ab different from a and b.107 the identity map. 1). Therefore we conclude that H1(R3 .{ p l .(-2) = 3.Therefore we have = which is a contradiction. (Indiana) Solution. O . It is well-known that r1(T2)is abelian and that . It is also clear that X . b) Prove that r l ( R 3.l .oi.{ c } and V = Y . Then it is easy to see that X . p z ) is homotopically equivalent to the space Y as shown in Fig. A = {(X.0) and ( . + Fig.0. z) I x2 z 2 = 1) be a circular cylindrical surface. Hence we have ab # ba and i*(a)i*(b) z*(b)i*(a).8.Y.2.{ e } .2.8 a) Y has a structure of a graph with 4 vertices and 6 edges.{pl. a) Compute H 1 ( R 3 . Hence the rank of H I ( Y ) is equal to 1 . Let X = {(x.A ) M H 1 ( Y ) z Y Z @YZ @Z.

that (R3 . Then let fl .A .g) + (Y.Since Since p is a regular covering space. : ( X .= r o f o : x Y. p ) is not abelian.z1) + need not be distinct from - 5 TO (?. Suppose that f : X -+ Y is a continuous function from the path-connected space X to Y with f ( z 0 ) = f(z1) = y. p*(. by the Van Kampen theorem we see that ?rl(Y) is a free group generated by ?rl(V) and ?rl(V). 2211 Let p : (?. Let fo(zl) = y. there is a deck transformation y such that y(y) = g. 9. (Indiana) . consequently. 2212 Let X be the identification space obtained from a unit 2-disk by identifying points on its boundary if the arc distance between them on the boundary circle is Compute the fundamental group of X.rrl(Y. that is.g) is a normal subgroup of 7rl(Y.y) be a regular covering space. : (X.g. + Hence 5 is the lifting o f f we want.9. Show that there is a second lifting. Therefore we conclude that ?rl(R3 . (A 5. ZO) --+ ( Y . p ) M ?rl(Y) is a free group on three generators and. and that there is a lifting of f .A .) (Indiana) Solution.108 same homotopy type as S1 and that V has the same homotopy type as the “eight figure space”.y). Since U n V is contractible.

Therefore r l ( U 121) is an infinite cyclic group on ' generator Z = y-'a'y. Fig. z o ) is an infinite cyclic group generated by u the closed path class of a. Since the circle a is a deformation retract of U . we may extend f ' .2.{y} and let V be the open disk. hence. We may consider f t o be a continuous nonconstant map f : a -+a. Take a point z1 E U n V as the base point.zl)is an infinite cyclic group generated by y o the closed path class of a closed path c which circles around the point y once. where y is the path class of a path d from z o to z1. It follows that rl(X) NNZG/U n =as.2. Since V is simply connected. r 1 ( X ) is isomorphic to the quotient group of r l ( U ) with respect to the least normal subgroup containing 4*(rl(U V ) ) . by the Van Kampen theorem. 2213 Sketch a proof of the Fundamental Theorem of Algebra (every nonconstant polynomial with complex coefficients has a complex zero) using techniques of algebraic topology. Then both U and V are path connected subsets of X and X = U U V . (See Fig. It is also clear that 4+(yc) 3Z.) It is clear that r l ( U n V. it is clear that wl(U.2. (Indiana) Solution. Note that If(z)l + 00 as JzI + 00. Hence the least normal subgroup containing = 4*(rl(U V ) ) is isomorphic to 3 2 . Let denote the complex plane and f (2) be a polynomial of positive degree with complex coeficients.9.9 Take a point y in the open disk.) Let U = X . (See Fig.9.where 4 : (U n V ) + rl(U)is n the homomorphism induced by the inclusion 4 : U n V -+ U .109 Solution.

we may prove the Fundamental Theorem of Algebra by means of the reduction to absurdity. Since U n V is contractible. + 2214 Let X denote the subspace of R3 that is the union of the unit sphere S2. we take U = X 1 V X 2 .2. Furthermore.110 to a map of the one-point compactification of ( X f : s2-+ s2 by setting f ( m ) = 00.) To compute ~ l ( X v X 2 ) .(4). (a) Compute the fundamental group of X . (b) Compute the integral homology groups of X . (See Fig. where 00 denotes the north pole.{ p } and V = X 1 V X2 . Then 1 we have U u V = X1 V X z . Noting the fact that if a continuous map f : S2 + S2 is not surjective then the degree o f f is zero.k > 0. then the degree of the extension f : S 2 -+ S 2 is equal t o k. a l ( X 1 v X 2 ) is a free product of the groups r l ( U ) and T ~ ( Vwith respect to ) the homomorphisms induced by the inclusion maps. (Indiana) Solution. It is obvious that . (4 Fig. and the portion.2. we may prove that if f is any polynomial of degree k > 0 then the degree of the extension f : S2 S 2 is still equal to k .10. Then we may first prove that if f ( z ) = z k . call it A. the unit disk D2 in the 2-y plane. by the Van Kampen.10 It is clear that X has the homotopy type of the one point union X 1 v X 2 where X 1 and X 2 are each homeomorphic to the union of the unit sphere and the portion of the z axis lying within S2. of the z axis lying within S 2 .

there exists a path C : [0. Thus f o (c-'hc) and a are homotopic.2. Hence - f* : m ( X . 20. ? r l ( X )is a free product generated by two generators. Show that is also surjective. 2 1 ) is surjective. (See Fig. For any [a] E ? r I ( Y .10. We can take as generators the closed path classes which are determined by the closed paths omp and omq respectively. It means that f o h and Zu?' = are homotopic. otherwise. Suppose that the induced homomorphism ---f is surjective.1] ---+ X such that c(0) = 20 and c(1) = 2 1 .111 Therefore. . (Indiana) Solution. f*([c-'hc]) = [a]. By the assumption. cac is a closed path at f ( z 0 ) .2 and are zero otherwise. n ( Y . consequently. Since X is path-connected. we conclude that i = 0.V). f ( z l ) )where a is a closed path at f ( ~ l ) . we see that Noting that H i ( X 1 ) M Hi(X2) which are infinite cyclic for i equal to 0. Then Z= f o c is a path connecting f(z0)and f(z1) --I in Y . and.) (b) Applying the Mayer-Vietoris sequence to the pair ( V .f ( 2 1 ) ) .2 1 Let X be a path-connected space. f : X Y a continuous function. and E X .1. there is a closed path h at 20 such that f*([h]) [Z?'].

it is not difficult to see. a m b Fig.) Since the automorphism group A ( X . suppose that f : X -+ Y is a homeomorphism. where both X and Y are connected.2. let p .11.2. But the spaces X and Y shown in Fig.{ f ( e O ) } is obviously not contractible..{ e o } is contractible and Y . rlX @ we see that 7rl(RP1x S2) x T1(RP2)@ “1(S2) a @ {O} M 2 =a. by considering the liftings of the circles a and b in X . By the formula . (Indiana) Solution. (See Fig. Otherwise. we come to a contradiction. but not necessarily homeomorphic.11.l ( z 0 ) = {eo. Then it is easy to see that they are homotopy equivalent to an 3-leaved rose G3.2. Prove that X and Y are homotopy equivalent. p ) F Z & and X is connected. in substance.112 2216 Let B denote the “figure eight space”. Then X . 2217 Calculate the fundamental group of the space R P 2 x S2.11 are not homeomorphic.{f(eo)}.2. X has only two different types as shown in Fig. .el}.{ e o ) is homeomorphic to Y . Let p : X -+ B and q : Y -+ B be 2-fold covering maps. (Indiana) Solution. that.11 For any 2-fold covering map p : X -+ B . Since X . r ( x Y )= m ( X > m ( Y > .

' ( X ) and *)) B is a path component of p . ker h is isomor*)/ phic to the group Z/6Z= Zg. and let p : 2 -+ 2 denote the covering space corresponding to the kernel of h.' ( X ) with a path component B of p . = ker(h).. *) Let 2 denote the figure 8 space.e4} of p-'(*) which corresponds to the elements {0. and p : 2 -+ 2 the *) corresponding cover. - - (Indiana) Solution. we see that the path component of p . e3} of p-'(*).e3.e4.' ( * ) = {eO. ( p * ( ~ 1 ( 2 .' ( Y ) . how many intersection points of A and B are there? (i.) If A is a path component of p . Then. what is the cardinality of the sat A n B?) (b) If G is a finite group. X and Y circles.ez.-+ Z/62be the homomorphism satisfying h ( z ) = 2 and *) h(y) = 3.e5}.12 (a) Let h : ~ ' ( 2 .2. Let x.-+ G a surjection.' ( X ) contains exactly the points {eO. We denote p .!?.e2. Thus ~ l ( 2 .113 2218 TI(.l ( * ) .' ( Y ) = p . it ) follows that the homomorphism h is surjective.4}of& respectively.y E be the elements in 2 defined by X . By the same reason. prove that the number of intersection points of a path component A of p . (a) Since h ( z ) = 2 and h(y) = 3 and T ~ ( Z . h : ~ ' ( 2 .' ( Y ) divides the order of G . Z = X V Y . where * denotes the vertex Fig.el. we conclude that A n B = {eo}. * is generated by 2 and y. . Y .e.' ( Y ) contains exactly the points {eo. Since p-'(x) n p . from h ( z ) = 2. Hence the covering space p : 2 -+ 2 is a 6-fold cover. we see that the path component of p .' ( * ) by p .2.

the fundamental group of X .{y}. (4 a d @ 2 0 a Fig. with proof. U . the homology of the universal cover of X . the corresponding cover p : 5 + 2 is a finite fold cover.13. the homotopy class of a closed path c which circles around the point y once (see Fig.zl) + r1(U. 2219 Let X be the result of attaching a 2-cell 0’to the circle S1by the map f : S1-+ S1given in terms of complex numbers by z -+ z6. with proof.2. zl). that wl(U nV. . 2 1 ) + . Then r generates a subgroup H1 of G and s generates a subgroup of H2 of G. Let 2 1 be a point in U n V . V .rrl(X. and U n V are arcwise connected. In a similar way as in (a). (a) Compute. (Indiana) Solution.) Applying the Van Kampen theorem.2. Then. and V is simply connected. it divides the order of G. Let V be the image of the interior of the hexagon under the identification.Irl(U. and let U = X . we see that the number of intersection points of A and B is equal to the number of elements in H1 n Ha. By Lagrange’s theorem. Suppose that h ( z ) = r and h(y) = s. as shown in Fig.13 Represent X as the space obtained by identifying the edges of a hexagon. (b) Compute.114 (b) Since h is a surjection and G is a finite group. we conclude that $1 : . Under the identification the edges a become a circle through the point 20.13.2. 21) is a n infinite cyclic group and generated by y. U and V are open subsets. It is clear that U n V has the same homotopy type with S1.21) is an epimorphism and its kernel is the smallest normal subgroup containing the image of the homomorphism 41: rl(u n v. Let y be the center point of the hexagon.

see ' we that H 2 ( 2 ) is a free Abelian group of rank 5.- ( 0. and T 2= S1x S1. Here ( f a - . 2 . Then the multiplication of [X. (a) We denote the homotopy class of a map f : X --+ S1by [f] and write s1 as s1= { e 2 = i o ECT I o 5 8 5 1).21)) is isomorphic t o U .115 and 4 are homomorphism induced by inclusion maps. z l ) is an infinite cyclic group generated by a' = ?-'a?. = I Z. g : X + S1is defined by g)(z) = f ( z ) g(z) for any z E X . It is obvious that dl(r) = ar6. H I ( % ) = 0 and H 2 ( 2 ) x H 2 ( X . If X is any topological space and S1 denotes the unit circle in the complex plane with its usual topology as a topological group with multiplication given by the multiplication of complex numbers. i = 0. is defined by [f] [g] = [f . Thus we conclude that ) (b) Since x is a finite 2-dimensional cw complex and T ~ ( X= its universal covering space X is a 6-fold covering space and is also a 2-dimensional CW complex. Thus ?rl(U.S1] of homotopy classes of maps from X to S1 inherits a natural group structure.lrl(U n V . (b) Compute this group explicitly for X = point. where 2 is the 1-skeleton of 2. 7 r l ( U . S (Indiana) Solution. But it is easy to see that X ( X ) = 1. Hence the smallest normal subgroup containing 41(. It is well-known that the Euler characteristic X ( 2 ) = 6 X ( X ) . You do not need to prove your assertions. (a) Define this group operation explicitly and indicate the group identity and how inverses are formed. Hi(%) i 2 3. 1 where It is obvious that the circle a is a deformation retract of U. X ' ) . where the S1] map f . g ] . consequently. S1. zo) is an infinite cyclic group generated by a and. So we conclude that . then it is known that the set [ X . where 7 is the homotopy class of a path d from zo to 21. From H o ( 2 ) M I . 2220 i = 1. hence X ( 2 ) = 6.

then [ X .9)2) (deg(f1 * Sl). where the map e : X + S1is defined by e ( z ) E eZni for any 3 E X.S 1 ] to Z @Z. S l ] x “l(S1)xa. M r2(S1)= 0.sl) = (deg(f * 911. respectively.S1] is trivial. The inverse of [ f ] is the homotopy class of a map f. denoted by f1. S1defined by t ) = f ( e 2 n i f 3 1e2xit+2xi(l-t)+ 1. Then let 4 : [X. Now we discuss the case of X = T 2 = S1x S1. g ( z ) is defined by the multiplication of complex numbers. it is easy to see that each homotopy class [ f ] can be represented by a map f : S2 -+ S1which maps the northpole N of S2 to the point po = eZUiof S1.S1] %Z@Z.degfz). deg(f2 . can easily prove that one [S’. We conclude that 3 [X. Note that f l can be extended to a map from X to S1. deg f 2 + deg 92) + (deg 91 deg 92) 1 = (deg fll deg f 2 ) 4(V1) + N g 1 ) - Therefore.92)) = (deg fl + degg1.S1]+Z@Z is defined by 4([f]) = (degf1. which is : defined by f(3:) = l / f ( z ) for any 3: E X . 4 is a homomorphism from [X. So the group [ X . For the case of X = S2. and cosequently q is an isomorphism. Then by the facts that S2 is simply connected and the universal covering space of S1. It is clear that 4 is a epimorphism. [sl) = = = 4Uf. one can easily prove that [S2.116 f ( z ) . 2. p o )and fZ(6) = f ( p O l e z U i e )for any e2nie E S1. Thus one can easily prove that 4 is a monomorphism. The identity of this multiplication is [el. deg(f . (b) If X = point. For the case of X = S1. R is contractible. . We have 4“fI . any S1] For map f from S1x S1-+ S1we define two maps f1 and f 2 from S1+ S1 by fl(6) = f ( e Z n i e . by still which is homotopic to f under the homotopy map F : X x I -+ ~ ( ~ 2 r r ie0 n i G .S1] obviously has only one element.

2 0 be T .Then exp(2) is the homotopy between f and c . If n l ( X . f is homotopic to a constant map. Suppose that ?rl(X) is finite. So the above ) argument still works in this case. X -+ R such that exp(7) = f . and the same conclusion holds. 20) were not finite. Since ?rl(S1)% Z and f * ( ~ l ( X ) is a finite ) subgroup of ?rl(S1).e. =0 Hence there exists a lifting of f .~ ( z o must have at least a limit point.20) is finite.. say Z. the universal covering space of S1. and f*(?rlT1(X)) for any map f : X -+ S1. show that ?rl(X. i.such that n(Z) = 20. locally path connected and the fundamental group of X is finite? (Indiana) Solution. Since R is simply connected.e. 7r1(X) = 0. then ~ ~ ~ ( would ) a closed set of infinite points in X. ?: 7 . we see that f * ( ~ l ( X )must be trivial. Let ?r : 2 -+ X be the universal cover of X. which is a contradiction. - Prove that if X is locally path connected and simply connected then every map X 4 S1is homotopic to a constant. Let exp : R -+ 6 denote the exponential covering map. Thus it is ) easy to see that x is not a local homeomorphism at Z. f is homotopic to a constant map Z.. (Indiana) Solution. Since 2 is compact. What can you say if we just assume that X is path connected. Denote the homotopy between and E by g. i. Since X is locally path connected and simply connected.117 2221 If X is a path-connected space whose universal cover is compact.

we show that 7 2 is injective. .. Therefore c = S. then the last row is also short exact.&(b) E kercl.h ( b ) = w 2 ( a ) .. (Hamud) Solution. Then since 2 ) c2(al(a2)) = Pl(EZ(a2)) = Pl(PZ(b)) = 0 and C 2 is injective. there is a b E B such that c = b1(b). Consider the following commutative diagram of abelian groups 0 1 0 ---t A3 62 1 B3 0 0 + 61 + 1 c 3 1 0 If all 3 columns and the first two rows are short exact. i. To prove the exactness at C. Hence 7 2 is injective. we have a l ( a 2 ) = 0. Thus there is = P262(a) . Since P2(62(a) .e. an a E A3 such that a z ( a ) = a2.(b) = 6162(u) = 0.Pz(b) = 0 and f ~ 2 is injective. Let c E C3 and 7 2 ( ~ = 0.118 SECTION 3 HOMOLOGY THEORY 2301 Prove the 3 x 3 Lemma. we see & ( a ) = b. By ) 3 the commutativity we see . Hence there is a n a2 E A2 such that ~ ~ ( a= Pz(b).Pz(b) = E z ( a 2 ) . Since b1 is surjective.b ) a2 E ker 01 = imcyz.

Then 72(61(b3))= E l h ( b 3 ) = &l(b). Let Xo = T2x T2.X 1 = R P 2 x R P 2 .E 1 & 2 ( a 2 ) = c. In a similar way we may prove that imy2 c ker 71.~ 2 ( u 2 E kerpl. since i and 2m 1 . Show examples of compact 4-manifolds with X = 0. Since a1 is surjective. Denote by u the connected sum of h h projective planes. 2302 Prove that if M is a compact manifold of odd dimension. . 2m+ 1 X ( M )= (-l)i dimHi(M. Therefore. Thus by the commutativity.3. X3 = U3 x U s . Therefore there is a b3 E B3 such ) that p 2 ( b 3 ) = b . Therefore X ( M ) = 0.) = i for i equal to 0. and X4 = S2x S2. then X ( M ) = 0. 3 and 4. ( Columbia) Solution.2.1. i=O By the PoincarC duality theorem. Thus the exactness a t C is proved. Thus. Then it is well-known that X ( U h ) = 2-h.1.119 Now we prove that ker y1 c imy2. there is an a2 E A2 such that al(u2) = ul. they appear in the sum with opposite signs.i have different parity. For any c E ker 7 1 .~ 2 ( u 2 ) . there is an b E B2 such that c = ~ l ( b ) Thus it is easy to see that Pl(b) E ker C1.Using the fact that + we see that X(X. . Hence c E imy2. since ~1 is surjective. M is 272-orientable.&).2. Suppose that d i m M = 2m + 1. for any i. and consequently that there is an a1 E A1 such that &(u1) = Pl(b). Since M is compact. Let X2 = U3 x 174. 2 We leave the proof of the exactness a t C1 to the reader. we have b .4.

(Illinois) Solution. Let pl and pa be the identification point respectively. Then U and V are open sets of E X .t . O ) =z3 (T. } for even n.s. By Mayer-Vietoris sequence. (Illinois) Solution. en}. s ) E ) i. s . s . Set U = E X . When n is even.. It is obviously to see that U and V are contractible spaces and X is a deformation retractor of U n V . = --t {C"In>O} C" C. we have where H q . So Zn(c) = { ( T . . ( c ) is isomorphic to Z $Z.{ p z } .{pl} and V = C X .s. The suspension E X of X is defined t o be the identification space obtained from X x [-1.(C) for all n.1] by identifying X x (-1) to a point and X x { 1) to another point. Z .t. ( T . ~ E Zn(c) if and only s = t .~) + n even (O. ( c ) = Z for even n. Compute H.~) + (T.s+t) nodd. Noting that imdn+l = {(O.l ( X ) denotes the reduced homology of X - 2304 Consider the chain complex c d. For example the sphere S" is the suspension of S"-' with the north and south poles corresponding to the two identification points. and E X = U UV. Compute the homology of E X in terms of the homology of X .s+t. : C.120 2303 Let X be a topological space.e. O .t)E C . we have H .-l defined by (s .

CB imf2. X') T fi* 0 --+ T fils'+ a : H2(D2. M z CBZ. s .we have Hn(c)= z . consider the following diagram 0 ---t HZ(X2) 5 H2(X2. ~ )E & ( c ) if and only if s t = 0. the 2sphere.S1) Hl(S1) + 0 The square is commutative and the level rows are exact. Therefore. Therefore.121 When n is odd. ( c ) M Z for any n.X1) 3 + H'(X1) % H1(X2) . So Zn(c) is isomorphic to Z $Z. It is well-known that H1(X1) = Z$Z. the image of one 2-cell under f1 is homeomorphic to S 2 . Let X2 be the space obtained by attaching two 2-cells to X I . Since H ~ ( xx') = imfl. H . 0 . + 2305 Construct a CW complex which has the following 2-homology groups: ( Coz~mbia) Solution. which has two generators a and b. Denote by X1 the "figure eight space" as shown below. . 0 )E C.+l = { ( ~ .HI(X2.} M Z. ~.and noting imd. ( T . one by the map f i : S' + 20 and the other by the map f2 : S' + a such that f2(z) = z2.

= 2!J. So(q) has J . we may conclude that the space X satisfies the requirements in the problem. = 0 and imf2Isi. n 2 0. Since H1(X2.1). It is also easy to see that H z ( X 2 ) =z.{ b } also has the homotopy type of S"(q).... 2307 Build a CW complex X by adding two 2-cells t o S1. ( Columbia) Solution. It is also clear that Sn(q 1) = U U V and U n V has the homotopy type of S n ( q .122 imfllsi. M 22 and kera. one by the map g1 : S 2 20 and the other by a map g2 : S2 -+ S 2 such that degg2 = 3. In fact S"(q + 1) = S"(q) v S". we have H1(X2)M Z $ Z 2 . where fip is the reduced homology group. V S" When n = 0.{ u } has the homotopy type of Sn(q) and V = Sn(q 1) . Then in a similar way as above. V S"). Now let X be the space obtained by attaching two 3-cells to X2. by the map z one and the other by the map z --+ z6.X1) = 0. Thus. What is the homology of this space? --+ z4 (Indiana) . Denote by S"(q) the space S" V S" V .. we may see that ima. V ) . when n > 0.(S" V S" V . --+ 2306 Compute H. z Y . q times q + 1 points.we may prove that gp(Sn(q)) is a free abelian group of rank q for p equal t o n and is zero otherwise. + + + Let a E S n ( q ) and b E S". by induction on q and the Mayer-Vietoris sequence of the pair ( U . Then H p ( S o ( q ) )is a free abelian group of rank q for p equal to 0 and is zero otherwise. for all p . It is easy to see that U = S n ( q 1) .

we have H 1 ( K ) M 2/22 M &. M Z @&. . we see that the image of ( f l S l ) * is 4Z.2. it follows that imd.KnA1)d. To compute H2(h). It is clear that j is an isomorphism. we see that H 2 ( K 2 ) imj2 M M ker 8 . imd2 M Z. is isomorphic to the subgroup 22 of H I ( @ ) . I 01 otherwise. KO) T f* T (f IS^)* H2(E2. is the boundary operator of the pair (Kn. i = 1.-l is the homomorphism induced by the inclusion map. where d. Noting that 1 ker d l % H 1 ( K 1 . is a n isomorphism.S1) 3 H l ( S 1 ) The square is commutative and it is well known that f. is 6Z. Then we have a chain complex K = {Cn(K).2.In the same way we see that the image of (glsi). K1) 2 HI(K1) 2 H I ( K 1 . K 2 is obtained from K 1 by attaching two 2-cells via the maps f and g as indicated in Fig.14. Since H ~ ( K ' . Noting f l s i is the map z -+ z4. : and C n ( K )-+ Cn-l(K) is defined to be the composition of homomorphisms. since j 2 is injective. ( X ) M H n ( K ) . K 1 is obtained from KO by attaching one l-cell. and that ker 3. Therefore. where C n ( K )= Hn(Kn.dn}. It is obvious that H n ( X ) = H. KM ) ~ imf. i = 2 . CB img. is a monomorphism and 8.14 X is a %dimensional CW complex with %skeleton K 2 = X . It is well-known that H ..(K) = 0 for n >_ 3.KO) M 2. l-skeleton K1 = S1 and O-skeleton KO = { p } . Let K" = K 2 for n 2 3. Thus. is isomorphic to Z @&.123 Solution. consider the following diagram: HZ(K1) --t HZ(K2) 2 H2(K2. Thus we conclude that ( zg3z2.Kn-l) and j. Hi(X) = i = 0. Fig.

where S 2 is the unit 2-sphere. Similarly we obtain Hz(X.0). ( X . Since the local homology groups are invariant under homeomorphism. Let h : X + X be any homeomorphism.((-11 0.(0)) M H. Let S = S 2 n X . otherwise. So H. X .(0)) for any q.O.((O.-I(U. Applying the Mayer-Vietoris sequence of the pair (W1.(U. To compute Z*(S).124 2308 Let X = {(z. y. Then S is a deformation retract of U .{XI) = Z @Z @Z.(0). Thus H .X . let Wl = .we see that b) Let z be any point of X different from the origin.X . b) Prove that any homeomorphism h : X Solution. a) Compute H . it follows that f(0) = 0.(f(0))). H 2 ( X . -1)) = s and W2 = s .(0)) M H2(X.z) E R3 I zyz = 0).O.y. (O. if z is in a coordinate axis. I). From the above results. x . . u .(U .(1. It is easy to see that U is contractible. ( X .(0)) g * ( S ) . where +X must leave the origin fixed. (Indiana) D3 = {(z. a) Let U = D3 n X .0. Wz).0)). x .(0)) = H. z ) E R3 I z2 + y2 + z 2 < 2).{ 0)).

{ p } ) -+ *.(D) f gq(S1X R1) -+ H g . RZ which is split exact. it follows that Hg(S1x R1)x H. ( R P 5 / R P 2 )x H q ( R P 5R P 2 ) .. (S1 = 1 .{ p } ) = otherwise.{ p } ) 3 Hq(S1x R1 .{ p } ) f gl(S1 x R1)+ 0 . ...(D . Since R P 5 is compact Hausdorff and R P 2 is a strong deformation retract of a compact neighborhood of R P 2 in R P 5 .{ p } ) @ @.{ p } ) x Hq(S1).{ p } . we see that H. the nontrivial part of the Mayer-Vietoris sequence is 0 + g1(D. S1x R1 . (Indiana) Solution.{ p } ) 3 x R1 .{ p } have the same homotopy type with S1. (Indiana) Solution. for any q.125 2309 a) Write down the Mayer-Vietoris sequence in reduced homology which relates the spaces S1x R1. Compute Hq ( R P 5 / R P 2 ) . and a disk D in S1 x R1 about the point p.sphere) b) Use a) to calculate the homology of S1x R1 . a) The Mayer-Vietoris sequence is -. 2310 Let the real projective plane R P 2 be embedded in the standard way in the real projective 5-space R P 5 . A - b) Since S1 x R1 and D . Thus we have %I iYi(S1 x R1 .+ H q ( D. Since D is contractible. where R P 5 / R P 2 is the space obtained from R P 5 by identifying R P 2 to a point.{ p } .l ( D .

*. 0. 2 2 . 0.(RP2) = otherwise. Y x R . 0.{ 0 } ) .O) E Y x R.it follows that R H.*.1. { z. otherwise. 0.3. from the exact sequence of the pair ( R P 5 . Q q Thus.Y x R . P 2 ) . otherwise.5. = 1. 2 3 ) E R 3 1 2 1 2 2 = 0 . 4 = 0. &. and H.2. 2 22 0 ) . i = 0. Solution. Let Y x R = { ( 2 1 .z2) E R2 I 2122 = 0 and 22 2 0). Let X be the space as shown in the following figure .126 It is well-known that H. Therefore Hi(Y x R .(RP5)= { z. - because U is contractible.5. q = 0.(0)) M Hi(U. q = 1. Then the point 0 = (0. 2311 Let Y = {(z1. U = D3 n (Y x R ) . = 0. To compute the local homology groups Hi(Y x R. U - (0)) M Hi-l(U . (Indiana) Prove that Y x R is not homeomorphic to R 2 . where D3 is an open ball centered a t 0. we take an open neighborhood of the point 0. &.{0}). (RP5/RP2) = a.

Let Y = X x (0) U A x [0.1] c Y .v. *.(0) and that H I ( X ) X Z @Z.1]and B = A x (1). Then yo is a strong deformation retract of ?r(U).?r(U)).B ) = H * ( Y . (Indiana) Solution. 1 of 1 X x [0. the inclusion map of pairs induces an isomorphism H * ( Y / B .V ) . for every n. Thus H z ( Y x R. Since the local homology groups are isomorphic under homeomorphism.it follows from the exact sequence that .yo). H * ( Y .-l(?r(U).?r(U)) H. Let V = A x [i. Let U = A x [$. Thus. 1 Prove that the reduced homology group g .U ) .Y x R . we have an isomorphism 11. Let ?r : Y -+ Y / B be the identification map and yo denote the point ?r(B)in Y / B . it follows that Y x R is not homeomorphic to R2. y o ) + + + H. ( X U C A ) is isomorphic to H.(0)) EZCBZ. in the exact sequence of the triple ( Y / B .R2 . H * ( Y .(Y/B.u . U ) M H * ( Y .(X. It is well-known that for any point p E R2 the local homology group H 2 ( R 2 . Hence.127 It is easy to see that X is a deformation retract of U . that is. X U C A is obtained from the subset X x (0) U A x [0.(?r(U).YO)+ H * ( Y / B . Since B is a strong deformation retract of U .* + Hn(?r(U). Y O ) = H*(Y/B.yo) = 0. A ) .( p } ) GZ.YO) * * * it follows that H.?r(U). By the excision property. 2312 Let A be a nonempty subset of X and X u C A be the union of X with the cone of A. Then X U C A = Y / B . 1 by identifying A x (1) to a point.

u - V )+ ( Y / B . Exactness. H. M H * ( Y . On the other hand.gives a homeomorphism of pairs n. A ) for every 2313 n.yo)H * ( Y / B )= G . ( X UCA).: (Y - v.T ( V ) T ( V ).yo). we have X x (0) H. from the exact sequence of the pair ( Y / B .128 Thus. B )M H * ( Y .(Y. (Here Hi denotes reduced singular homology. A ) . let Cf : C X -+ C Y be the map of suspensions induced by the map ( ~ . Excision.t) X x [0. For any topological space X let EX denote its (unreduced) suspension. it follows that H*(Y/B.. we have H * ( Y . s) if s = t = 0 or s = t = 1. and so an isomorphism of their homology groups. In a similar way the set T ( V )may be excised from the pair ( Y / B . - - (ii) Using only the Eilenberg-Steenrod axioms (Homotopy. where denotes the equivalence relation generated by requiring that (2.u V).n-(V)) to give an isomorphism Now the restriction of the map n. All of these combine to give an isomorphism H * ( Y / B .(V)).B ) . yo) Since Y admits obviously a strong deformation retraction onto which maps B onto A x { 0 } . x Therefore. Dimension) for a homology theory prove that Hi (EX) naturally isois morphic to Ei-l(X). t ) (f(z). ( X . -+ of (i) Prove that if f : X -+ Y is a homotopy equivalence then so is Cf.1]. .(X U C A ) is isomorphic to H . ( X .v. B ) M H .) t I f f : X + Y is a map.) (Indiana) . ) (y. (EX is the quotient space (X x [0.1]/ -.

.'(Z) means any point (z. (ii) Let 1 u = {rIx(z. U ) that fii(CX) M Hi(EX. U . It is also easy to see that H ( Z . Hence r?.t)1 t > -} 3 and 2 = {rIx(z. where.s) E EX x I . from the homolopy sequence of the pair (V. l). On the other hand. if Z = po or p l . for any (2. u . Thus we conclude that k i ( C X ) k*-l(X). respectively.O) = f(z) and G ( z .1) = g(z) for any z E X . Let f : X + Y be a continuous map. 3 It is obvious that U and V are both open sets of EX and that U and V are both contractible. and by po.z E X . lI. Then define : (X x I) x I -+ CY by c s) Let H : EX x I --t CY be a map defined by H ( Z .t ) 1 t 2 2/3}. and consequently it follows that if f : X + Y is a homotopy equivalence then so is Cf. By the excision property we have W = {IIx(z. where id : I + I is the identity map. (i) Denote by rIx : X x I + C X the quotient map.W ) we see that Hi(V.W). Let c U . 0) and (z. Then v It is easy to see that U .W has the same homotopy type of X . It follows from = the homology sequence of the pair (EX. By the definition we see that Cf = rIy 0 (f x id) 0 IIXl.i(U) = g.W ) M E&'(U .(V) 0 for all i. 1). 1) = Eg(Z). It is easy to check that H is well-defined and continuous. s) = E(II&'(%). Therefore Cf is homotopic to Cg. Suppose that g : X + Y is another continuous map which is homotopic to f by a homotopy G : X x I + Y such that G(z. pl the equivalence classes of (z.129 Solution. U ) . 0) = Cf(Z) and H ( E .O) or (z.t) 1 t < -}.

there is an a1 E Al such that y ( a 1 ) = /32(a2). i. It means that kerag = 0. Suppose that y and Po are surjective.) a) Prove that H 1 ( X ) = 0. there is a bo E Bo such that Po(b0) = b l .. Prove that a3 is injective. which means that a l ( a 1 ) .130 2314 Consider the following commutative diagram of abelian groups in which the row and column are exact sequences. Hence a3 = ~ 2 ( a 2 = a 2 a l ( a l ) = 0. .a2 = 0. and there is a bl E B1 such that P l ( b 1 ) = a(a1) . H 1 ( U ) = 0. Then from the exactness there is an a2 E A2 such that a z ( u 2 ) = a3. so 2315 Suppose that a topological space X is expressed as the union U U V of two open path-connected subspaces such that U n V is path connected. (Here H denotes singular homology.a 2 . Let a3 E A3 such that cyg(a3) = 0. a3 is ) injective. Hence by the commutativity we have a l ( a 1 ) . Once again by the exactness we have 0 = PlPo(b0) = P l ( b l ) . Since PO is surjective. (Indiana) Solution. Since y is surjective.e. and the inclusion U n V + V induces a surjective homomorphism H1(UnV) + H 1 ( V ) .a2 E kerP2.

It is obvious that U . + H ~ ( x 3 H ~ ( U v)% H ~ ( u CB H ~ ( v )H ~ ( x + H ~ ( U v). ) n ) 5 ) n Since U n V is path-connected.z) E s2I z > . So the homomorphism H z ( U n V ) + H z ( V ) induced by the inclusion map is surjective..131 b) Give an example t o show that the analogous assertion becomes false in general if H is replaced everywhere by Hz. 1 (Indiana) Solution. d* is surjective so that Therefore im& = 0. It means H 1 ( X ) = im$*. Eo(U n V )= 0. H 1 ( X ) = 0. V and U n V are path-connected and that H z ( U ) = H2(V)= 0. But H 2 ( X ) M Z.1 .e. 2316 Let D" = { z E R" 1121 5 1) denote the standard unit ball in Euclidean space Rn.1)-sphere.. By the hypothesis. y. the unit 2-sphere7 u = {(zclg. i.V ) A . b) Let X = S 2 . and denote the standard (n.5 ) and 1 1 v = {(z. Prove that f is surjective. . a) Consider the Mayer-Vietoris sequence of the pair ( U .l .. z ) E s2I z < -}. (Indiana) Solution. 2 Then X = U U V .9 . Suppose that f : D n --t D" is a continuous map such that the restriction of f to S"-l is a homeomorphism from S"-l to s n . Suppose that there is a point 20 E D n such that zo @ f(D"). we see that 20 E D" . By the assumption. We use the reduction to absurdity.

-'(S"-') is an isomorphism. and i. (Indiana) Solution.. and H 2 ( X 2 .e. Therefore g o i : Sn-' -+ Sn-' is a homeomorphism. : H2(X2. H 2 ( X ) is also a free abelian group of rank 2. i. which is a homeomorphism. : Hn-1(Sn-') H n .3 3 = 2. The Euler characteristic of X . + X ( X ) = rankHo(X) . X 1 )= Z $ Z @ Z . and no higher-dimensional cells. Let X k denote the k-skeleton of X .X ' ) --t H 1 ( X 1 .Compute the Euler characteristic of X and determine (with proof) all possibilities for H2(X). Hence (g 0 i)* : Hn-l(S"-') -+ H. X 1 )is free abelian. we see that H 3 ( X 3 . where d.XO).-1 : 9 . Thus because H 2 ( X 3 .132 Therefore Sn-' is a deformation retract of D" .X O ) } . H2(X) Z Z $Z. Then g = r o f is a continuous map from Dn to S"-l and gIs. and consequently (g o i)* is trivial. . X 2 ) 0. This is a contradiction. 2317 Let X be a connected CW complex with two 0-cells. But ( g o i)* = g* o i.l ( D n ) is --t obviously a trivial homomorphism because of Hn-l(Dn) = 0. By the assumption.X ' ) 3 H 1 ( X 1 ) 2.(X). Let i : S"-l -+ Dn be the inclusion map. Therefore = H 2 ( X ) M ker{d. three 1-cells. Let T : D" .(20). X ( X ) = 2 . Therefore rankHz(X) = 2.l -+ S"-' is just the restriction of f to Sn-'.rankHl(X) + rankH. is the composition of homomorphisms H 2 ( X 2 . Assume H1(X) x Z @Z/3. three 2-cells. H1(X1.( 2 0 ) -+ Sn-' be a retraction. On the other hand.

Thus we have d'+'(a) # 0. But it is well-known that H k ( X k . and by dk the composition of homomorphisms of pairs H k ( X k . Z )is not a nontrivial finite group and contradicts the hypothesis. Hk+l(X. X k . We choose a generator a in H k + l ( X k + l . Denote by Xk the Ic-skeleton of the C W complex X . H k ( X . X k ) . H ' + l ( X k + l . By the hypothesis. kerdk is trivial or free abelian. Z ) M ker dk/imd"' xk-2). Prove that if Hk(X.l . i Y k ( X . and Hk+l(X. We claim that &+'(a) Otherwise we would have imd'+l = 0. It follows that kerd"' = 0 and.Z) = 0. It means that H k ( X . consequently.1 ) It is well-known that 3 Hk-l(Xk-l) 5 H k T l ( X k . (Indiana) Solution.x k . Z )M ker dk c H k ( X k .Z) M ker dk+1/imdk+2.Z) = 0. and.X"-'). 2319 Let and .Z) is a nontrivial finite group then Hk+l(X. X k ) zz. Therefore # 0.l ) is a free abelian group.133 2318 Let X be a CW complex with exactly one k+l-cell. therefore.

{pl. Let 6 be the unit open ball centered at the point ( O . W2).2. 0.(O.134 Let X = A u B. 2 2 . other wise. 2.( O . 0)). 0.15) + 2: + 2: = 1). we see that Hi+l(U.O.O)) M H i ( U . It is obvious that U is contractible and that U .(0. By the excision property we have &(X.(0. 0.O.) Then W1 and W2 are open subsets of C U D and W u W2 = C u D.x - (O. 0. 0) in X . O .(O.O) has a deformation retract C U D.{p3. (Indiana) Solution. (See Fig. Therefore. 0.15 Let Wi = C U D .O)) for all i. Compute the relative homology groups H i ( X . we see that 2 = 0. Then U = 6n X is an open neighborhood of ( O . O .(0. 2: (See Fig.2.u . u . Fig. O . O .O)) = Hi(U . 2 3 .O.O)) .p4}.15. U . applying the Mayer-Vietoris 1 sequence to the pair (W1. O . O ) in R4. 0. From the homology sequence of the pair ( U . It is also clear that S1 is an deformation retract of W n Wz. 0. O . O .p2) and W2 = C U D . X . 0. where and D = { ( ~ 1 .O)) for all a. 2 4 )E R4 I 2 4 = 0. It is clear 1 that both C and D are homeomorphic to the unit sphere S2 and that D and C are deformation retracts of W1 and Wz respectively.

Hence we conclude that Hj(X.O.X . Z ) + H i ( Y . where R : R2 x S 2 -+ Y is the covering map. f(z)) be trivial.(O. must . Z ) -+ - + Hi(Y. we have It is clear that Rl(S1 x s1x S1) = “l(T2)@ “ l ( S 2 ) = a $a.O. Therefore. It is easy t o see that R2 x S 2 is the universal covering space of Y . Let Y = S 1 x S1x S2. I ) for all i different from 0 and 2.z) is finite. (Indiana) Solution. the homomorphism f* : R ~ ( Xz) --+ R ~ ( Y . Since ?rl(X.x) is finite.Z) is trivial for i different from 0 and 2. we may lift f t o a map : X -+ R2 x S 2 such that a o f = f . Hence we have - 7 f* = R. so is . 2320 Let X be a path connected and locally path connected space and let x E X .1 otherwise. Show that if al(X.Z) Hi(R2 x S2. But it is obvious that : H i ( X . Z ) for any i.135 for all i 2 0. 0 f* : H l ( X . then any continuous map f : X -+ Y induces the trivial map f* : H i ( X . 0)) = i=o. By the Kunneth theorem.

6’ya = 0.F 6’ are exact. we have (6’6. .&)(P. we have 6P = 7 a and &a’= 6’7.&a’a= 6’ya .t F Prove exactness of (Indiana) Solution. By the assumptions.a’a) 6’6P = .Therefore.136 2321 Call a commutative diagram of abelian groups “exact” if the sequence of groups and homomorphisms o + A (9)c ’‘ B c~ 5D It is an interesting fact that if A Z 6 - o is exact. B P1 and C + D 17 B S E 71 1 E D . then so is A C ’a l .+ E 1 E PI 6’6 C .

we have i m ( a . Prove that there exists a non-empty closed subset A of X such that f ( A ) = A .. By the assumptions. noting that ( e . Therefore zo 2 z1 = 220. 6 ( c ) )E ker(E . Thus. Give an example to show that compactness is essential for this assertion. consequently. Then there would exist an 2 0 E A such that z 5 zo for any z E A. . 20 1 = 521.a’). c ) E ker(y .137 It follows that im(P. The exactness at C @ E is proved. S’6(c) E ( e ) = 0. p ) = ker(y . Since f ( A ) = A iz xEA and zo E A . we see that there exists a b E B such that a’(b) = e and y(b) = 6 ( c ) . which is a contradiction. ( c . there exists an a E A such that a(. It is clear that { F n . Therefore. Hence we have (b. Then F is a non-empty closed subset of X. we see that the subset A = n=l n F. We claim that f ( A ) = A holds.S’).+ X be a continuous map. i.E ) . Let X = (0. Let F = f ( X ) .1] and f : X --+ X is defined by f(z) = for z E X.E ) c im(P. a’a)(u)E im(0. Let ( c .e.e.6) and im(a’.E ) . e) E ker(6’6 . In fact z o = sup z. and. a’a). Suppose that a non-empty closed A of X satisfies f ( A ) = A . i. n E Z+} is a sequence of non-empty closed subsets in X and that Since X is compact. there would exists an z1 E A such that f(z1) = 2 0 . 2322 Let X be a non-empty compact Hausdorff space and f : X . We give an example to show that compactness is essential for this assertion.y) = ker(c .. It means that ker(6’6 .6). (Indiana) Solution.) = b and p(a) = c .e ) = (0. We define 1 1 Fn+l = f(Fn) inductively for n E Z+. a’a).a’a) c ker(6’S . 00 is a non-empty closed subset of X .

( O . Then flx-(o. -1). Z )x H3(RP3. O ) ) x H 1 ( A U B ) =Z $Z $Z.O. we see that H1(X . l ) and V = A U B .O. 0. .it is obvious that f* is not an isomorphism. O . 0) is also a homeomorphism.Z) Z .O. Take U = A U B . r . (Indiana) Solution. (Indiana) Solution.o) : X .(O. 0 ) ) . Therefore. (a) Compute H1(X .138 2323 Recall that H 3 ( S 3 . f : S3 + S3 such that 7 r f = f .V). 0). O .(O. show that X is not homeomorphic to R2. Then it is easy to see that A U B is a deformation retract of X . Let 7~ : S3 + RP3 denote the universal covering map.Y. 0 ) + R2 . Let f be a function from S3 to RP3. M (RP" is n-dimensional real projective space.) Prove that there is no function f : S3 + RP3 inducing an isomorphism on the third homology. Applying the Mayer-Vietoris sequence to the pair ( U .(0.(O. b . = - I because deg f E Z. (b) Using part a . Then it is clear that the degree of 7r is equal to 2. It is well-known that S3 is the 2-fold universal covering space of RP3.o.> I ZY = 01. (a) In fact.d e g f . Denote by a and b the generators of H 3 ( S 3 ) and H3(RP3)respectively.(0. which . Since 7r1(S3)= { 0 } . Then we have f*(a)= 7r*(f*(a)) d e g f . 0 . (c) Prove or disprove: X is homotopy equivalent t o R2.f ( O . U U V = A U B and U n V has four path components which are all contractible. X = 7 ~ 1 7r2 where 7r1 is the plane y = 0 and 7r2 is the plane U z = 0. U and V are both contractible. there is a lifting of f. (b) Suppose that there exists a homeomorphism f : X + R2. Denote by A and B the unit circle in 7rl and a2 respectively. ( a ) = 2 . 2324 Let x = ((2.

X onto the plane 7r2. is a monomorphism. By the assumption. ( X ) -1?. 0)) to H1(R2 . d. M ker i.S n .f ( O .l ( B n ) = 0. Then F is a deformation retraction of homotopy equivalent to R 2 .A) a.l : S"-l -+ A is a homeomorphism. Thus we have keri. A ) = Z . Consider the following diagram H . f1.(X. and therefore It is well-known that 8: is an isomorphism and H n . Suppose that f : ( B " . (c) Let F : X x I +X be a map defined as follows. Show that if & ( X ) = 0 then H .-i(X) t f* a: t fl* i : + t f* Hn-l(B") Hn(B". ( X . Hence we have H . ( X ) = 0. Hence X is 2325 Let ( B " . .( O . = H .l ) in which the level rows are exact and the squares are commutative.l ( A ) = Z .A ) M imd.l ( S n . O .-i(A) i* ---t H.H. O .139 wouid induce an isomorphism from H1(X . n > 1.l ) be the standard ball and sphere pair in R". = Z. + H.S n . (Indiana) Solution. A ) is a continuous map and f l s n . Since H . But it is clear that It is a contradiction. .l ) -+ ( X . 0)). is an isomorphism. ( X .S"-1)+ H " .

{ p 1 } a n d V = S 2 x S 5 . 2327 Let A be a subspace of S 2 x S2 homeomorphic to the 2-sphere S2. (Indiana) Solution.{ p 2 } . 5 (b) Compute the homology of S2 x S5 with 2 points removed. and UUV = S 2 x S5. Then U . 2. . It is easy to see that S2VS5 a deformation retract of both U and V . It is easy to see that S 2 x S5 is homeomorphic to the space obtained by attaching a 7-cell to S2V S5. V) and using the fact that we conclude that i = 0. Applying is the Mayer-Vietoris sequence to the pair (U. i = 1. What can you say about the possibilities for the relative homology groups H . (b) Let X = S 2 x S 5 . x 1 E S 5 . Hence it is obvious that H i ( S 2 x S 5 ) is an infinite cyclic group for i equal to 0. and 7.{ p l . Denote by S2 V S5 the one point union of S2 and S5. U = S 2 x S 5 .2. ( S 2 x S2)are (no proof is required).6. By the Kiinneth formula we conclude that otherwise. V are both open sets of S2x S5 and UnV = X. Therefore the CW structure on S2 x S 5 has a 0-cell. a 2-cell. 5. State what the homology groups H . 21) E S2 x S5. and S5 is obtained by attaching a 5-cell to XI. which can be considered as the space obtained by attaching a 2-cell and a 5-cell to the point (20. (a) Let xo E S 2 . p 2 } . ( S 2 x S2.A)? (Indiana) Solution. otherwise. ez e z ez. and is zero otherwise. a 5-cell and a 7-cell. Then S2 is obtained by attaching a 2-cell t o 2 0 .5.140 2326 (a) Describe a CW structure on S2x S5 and use it t o compute the homology of s 2 x s .

X”) infinite cyclic for i equal to 1 and is zero is otherwise.A ) -+ H o ( A ) -+ Ho(S2 x S 2 ) -+ H o ( S 2 x S 2 . A ) = 0 for i equal to 0 and 1. Then the space X may be viewed as a space obtained from X* by attaching a 1-cell.. Since &(A) = 0.we see that H .1. (This is the union of two “cones” on X. ( S 2 x S 2 .X”) conclude that Hi(X) infinite we is cyclic for i equal to 0.141 From the homology sequence of the pair ( S 2 x S 2 . (Indiana) Solution. for all i 2 0.A ) M keri. A )-+0 0 + H 1 ( S 2 x S 2 . A )M H i ( S 2 x S 2 ) for i 2 4 and that 4 the nontrivial parts of the sequence are 0 and H 3 ( S 2 x S 2 . 2329 Define the “unreduced suspension” CX of a space X to be the quotient space of I x X obtained by identifying (0) x X to one point and (1) x X to one point. is the homomorphism induced by the inclusion map i : A S2 x S 2 . and H 2 ( S 2 x S 2 . It is clear that Hi(X. Thus from the homology sequence of the pair ( X . A ) -+0.2 and is zero otherwise. where i. A ) 4 H 2 ( A ) 5 H 2 ( S 2 x S 2 ) 5 H 2 ( S 2 x S 2 . a. - - 2328 Compute the homology groups of the space X obtained as the union of the 2-sphere S2 and the x-axis in R3.(X) H i + l ( C X ) .A ) M Z @Z/irni.using the above result and the fact that H . 4 (Indiana) ..) Show that there is a natural isomorphism ax : H. Let X’ = X .{the open interval (-1. By the exactness.l) in the x-axis}.(S2 S 2 . ( A ) M Hi(S2). x we have H3(S2 x S 2 . It is easy to see that X’ has the homotopy type of 5’. A ) . H.

) : V -. 2330 Denote by X the union of the torus S1x S1with the disc D 2 . The inverse of A * .(21).a D 2 ) 4 w H l ( d D 2 )-+ 0 .. Consider the Mayer-Vietories sequence of the pair (U. (a) Consider the homology sequence of the pair ( X . Take U = EX .(X) for all n 2 0. CTX : & i ( X ) -+ H i + l ( C X ) is just what we are looking for.. ( X . T 2 ) . H . 0 4 H 2 ( T 2 )5 H 2 ( X ) 1 H 2 ( X . Then U and V are open sets of E X . and U u V = EX.. (b) Is T 2 a retract of X ? Why or why not? (Indiana) Solution. d D 2 ) ) Z .-. + ~ ( U ~ 3 )H i + l ( u > ~ H i9 H(~~ > ( E x ) H i ( u n v ) . It is well-known that H . V s l +~ A ++ + - It is clear that U and V are contractible and that U n V has the homotopy type of X . ( X ) .) 20 Fig.( 2 0 ) and V = C X .where z f * is the homomorphism induced by the adjunction map f : D 2 -+ X .2. Let (0)’x X and (1) x X be identified to the points 20 and 2 1 respectively. T 2 ) M f * ( H 2 ( D 2 .. The naturality of ax can be derived from the naturality of the Mayer-Vietories sequence. where D 2 is attached to T 2 by identifying dD2 with a meridian curve S1x (20) in the torus. T 2 )= 0 for i # 2 and H 2 ( X .142 Solution. T 2 ) . (See below. Thus the nontrivial part of the sequence is as follows.16 (a) Calculate H. -+ a: H 1 ( T 2 )2 H 1 ( X ) -+ 0 Qt f* t fl* 0 + H2(D2. where 20 E S1. Thus the homomorphism A : k i + l ( C X ) + k i ( U n V ) is an isomorphism and we may obtain an isomorphism A* : H i + l ( C X ) --t & ..

This is a contradiction. Thus we see that H 1 ( X ) x Z @Z/Z M Z and H z ( X ) M H 2 ( T 2 )x Z . and let h : A -+ B be the surjection defined by h(1) = 1. = id : H 1 ( T 2 ) H 1 ( T 2 ) . (See Fig. In other words.143 It is easy to see that ima: = irnfl.with explanation. (Indiana) . Otherwise. the pull back of h via g is the group g * ( A )= { ( c .+ B 9 (b) Let A = Z / U . a )E C x A 1 g ( c ) = h ( a ) } . B = 2/Z. Let C = 2/82 and let g : C -+ B be the surjection defined by g(1) = 1. consequently. we have T* o i. Then r o i = i&z : T 2 -+ T 2 . there would exist a retraction map T : X . Hence we conclude that (b) We claim that T 2 is not a retract of X. Prove that the kernel i of g*(A)is isomorphic to the kernel of h.16) It is clear that f 1 * ( H l ( a D 2 ) )= 2 @ (0). Let i : T2+ X be the inclusion map.2. + (a) Let g*(h) : g * ( A ) + C be the homomorphism obtained by restricting the projection onto the first factor C x A -C to g * ( A ) . 9*( A ) c 1 g*(h) A l h . and.+ T 2 such that ~ ( =pi d p . and ker& = ker f 1 + . Identify the group g * ( A ) . We know that H 1 ( T 2 ) = Z @Z is generated by a and b. 2331 Given homomorphism h : A B and g : C -+ B . we have the following commutative diagram - because of H 1 ( T 2 ) = Z @I and H 1 ( X ) =Z.

Let F : kerg*(h) -+ ker h be defined by F(l. we see that kerg*(h) = {(l. B are open subsets in S3.2.5.4. (b) The homology of Y = R3 . ( C + - Noting that Ga(S3 . 0.2). (7. and. otherwise.0). and Hi(S3 . + H. C = l..B = S3 . Then A and a .0).2).C ) % Hi(S3 .) (Indiana) Solution. ( L 3 L (3..a) I a E kerh}.2}.{co}) % Hi(s3)A H ~ . a ) = a. We have the following Mayer-Vietoris sequence. (5.. (1.7}.2).O).6} and g-'(l) = {1. . (4.(7.1).a) = l. (a) Suppose that ( c . Thus it is clear that the group g*(A) consists of the following 16 elements: (O.3).1>..C ) @ Hi(S3 . Then g*(h)(c. u ) E kerg*(h). (b) In this case.. Assuming this fact compute (a) The homology of R3 .X .. .3}. It is easy to see that F is an isomorphism. (2. (2.R )~+ . I). i = 3. (5. (3.~ .3.0).{m}. (6. when C is a homeomorphic copy of the circle in R3.e. (a) Denote S3 = R3 U { }Let A = S3 .C. h-'(l) = {1.{ }= G@3) = 0 a ) for any i and that H i p 3 )= { z. therefore. and A U B = S 3 and A n B = R3 .+'(S3) 2 Hi(R3.. 2332 Recall that if C is a homeomorphic copy of the circle in S 3 . (0.C. (6.2).C. By the definition of g*(A). then H i ( S 3-C) is infinite cyclic for i equal to 0 or 1 and is zero otherwise.144 Solution. g-'(O) = {0.3). the identity of C.C ) = .. the one-point union of two circles.3). we have h-'(O) = {0. (4. I>. where X c R3 is a homeomorphic copy of the "figure-eight space" (i.

~ ~ From the result of (a).( ~ 2 ) .145 we can see that - = {: i = 0. which means imA = 0 too.U n V . Since the representative chain c of [c] can also be considered as a chain of (R3 .u )u ( R .by the definition of ++ we have [c] = + + ( [ c ] . Due to the above facts.(u nv). Let U = X . Hence we can consider the generator of H 2 ( S 2 ( r ) ) . Since U u V = X .v )= R~ . The claim is proved.1.2. .17 Represent X as shown in Fig. It is easy to see that iii(x3 (v n v))M iri(s2)>.17.V ) .X .V )5 i i 2 ( ~ u n V > 2 -U 3 -3 %Z$Z H 1 ( Y ) 2 ii1(R3 .2.x = ( R.u) n ( R -~v) ~ and ( R . R3 .V ) is o --+ ii2(y) i i 2 ( ~ 3 ) CB i i 2 ( ~ .0).as the generator of E2(R3 .[c]. %A7 @Z - MZ + A We claim that the homomorphism is an epimorphism.U ) and gi(R3.U ) @ ii1(R3 . Thus we get H 1 ( Y ) = Z@Z and a split exact sequence - - O + & ( Y ) 5ZCBZ5Z--+O. we have Y = R~ .{PI} and V = X .V ) --+ 0. Take a sufficiently large T > 0 such that the sphere S 2 ( ~belongs to R3 .U ) n (R3 . otherwise. Then S 2 ( r )is a ) deformation retract of R3 . we see that Zi(R3.U.V )is infinite cyclic for i equal to 1 or 2 and is zero otherwise.2. the nontrivial part of the Mayer-Vietoris sequence of the pair (R3 . Then we see that U and V have homotopy type of S 1 and that U n V is contractible.U n V). Fig.

It is known that if X c S3 is homeomorphic t o S1 then H . (Indiana) Solution. 0. . Therefore.B)= S 3 . H .(S1).X ) M H. 4.A ) @ &(s3 . this fact to compute the homology of S 3 .A ) M Hq(S3.So we conclude that H l ( Y ) = {:w Z7 I 2333 i 2 3. A - A - &(s3.(S1) = and that 4 # 1..A ) U ( S 3 . .(S3 . ( S 3 . Noting S3 . ~ Z. . 4 = 0. . S3 . q#3. * . in the Mayer-Vietoris sequence we have *.B ) . Using the fact that Z.B ) M H. i = 0. ( s ~ . i = 2. 4 = 1.Y )-.Y = ( S 3 .H . 4'1. H.A ) n ( S 3 .B) H . 4=2. ( s ~ ) . Denote Y by Y = A U B.Y where Y is a subspace Use of S3 homeomorphic to the disjoint union of two copies of S1.146 It follows that HZ(Y) =Z. where A n B = 8 and both A and B are homeomorphic to S1. .B are open sets of S3 and ( S 3 .ii.~ ( s . i=l.Y )= - { 2*Z. 3. ii.(S3) = we can easily see that { Z.(S3 .A and S3 .Y )5 .+l(~3) .

147 2334 (a) Sketch pictures of the universal covering of the one point union S1v S2 and of the connected 2-fold covering (no proofs required). (a) . and that S1V S2 is a deformation retract of U and V. Fig. i. the 2-fold covering map: z + z 2 from S1to S1. b) To compute the homology of the connected 2-fold covering space of S1v S 2 .2. is the exponential map: R + S1. Thus.18 The universal covering of the one point union S1VS2 is shown as in Fig.2. reis stricted on S1. Then it is easy to see that S1is a deformation retract of U n V .{the antipodal point of p l } and V = S 2 V S1V S2. (Indiana) Solution. restricted on each S2.2. The covering map T . we take U = S2 V S1V S2 .18. Fig.19 The covering map T . and. The connected 2-fold covering of S1v S2 is shown as follows. restricted on R. it is clear that . and.is the identity map.e... (b) Compute the homology of the connected 2-fold covering space of S1vS2. restricted on each S2.{the antipodal point of p 2 } (see the above figure). the homology of S2 V S1V S 2 . is the identity map..

Hence the homomorphism - - 4. o + H2(u) H2(v) H2(s2 s v s ~A )H ~ ( U v) @ 5 v 1 + n 5 &(U) @ &(V) 2 &(S2 v s1v S 2 ) + 0. : G1(u n v)+ H1(u)@ H1(v) is injective. Thus imA = 0. (Indiana) Solution. It means that H z ( S 2 V S 1 V S 2 ) ~ k e r A z i m $z H 2 ( U ) @ H z ( V ) = Z @ Z . i. 4 = 1121 0. : H1(U n V) -+ k 1 ( U ) and I . Fig. It is clear that - - H ~ ( v s1v s2) H ~ ( u ) kl(V)/im#.(S2) @ H. 2.(S2 - - v s1v S 2 ) = 4 = 1.(S1) = Therefore. MZ. 2335 Compute the homology of S1 x S1-point.20 It is easy to see that the space X = A u B is a deformation retract of the space H = S1x S1-point. It is easy to see that the inclusion-maps k : unv + U and I : U n V + V induce injective homomorphisms k. otherwise. we have H.148 and &(U) M &(V) x H. 4 # 172. the nontrivial part of the Mayer-Vietoris sequence is c Z. where A and B are each homeomorphic to S1and . s ~ M Hence..e. ker 4%= 0. . : H1(U n V) + H1(V). respectively.

is an isomorphism for all n. . It is clear that A and B are deformation retracts of U and V . ( u ) .(a’) = Pn(b). such that i. @ B. -+ A .149 A n B = {zo} as shown in Fig.(b) ..u E kerj.20 Choose points a E A and b E B such that u # z o and b # 2 0 .. Then 7 ’ o j L ( u ) E ker6. the rows are exact.(b) = j k ( u ) . The following sequence is exact: where A is defined by A(a’. Let U = X .there exists a b E B.{ a . Hn(S1 S1. i BL + An-l + * * * Write out the proof of exactness at Bh. ( d )= A(-a’..{ u } . such that jn(b) = 7 ’ o j . and that U n V = X .. Construct an exact sequence . Due to the exactness . and 7. We give the proof of exactness at Bh as follows. Thus Pn(b). we have j .lojA). It means that u = P. From the commutativity.(a’) .point) = H .2. ( X ) = H . -b).{ b } .P n ( b ) for any (a’. respectively.. j n ( b ) = jA(b). b ) = . . at C.u.i . @ B. 07.. 7. and consequently. Let u E ker(6. Applying Mayer-Vietoris sequence t o U and V. we have nL2. b } is contractible. . o@. and there exists an a’ E A. and let V = X . ( B ) = x 2336 Suppose the following diagram is commutative. ( A ) @ H . +A’. b) E A. (Indiana) Solution.

So k . is proved. from (l). g : X + 2 are two maps of X into contractible spaces Y and 2. V and U n V are open sets of M . ’ Since j ..+ H. . Thus. - - - - .(b) for some a E Ah and 6 E B. o 7 ’ o j . . u = ih(a’) . Prove that H. Suppose that u E imA. H.O) with f (z). . the equivalence relation is determined by (z. Thus the exactness a t B. . o Pn = 7. that is. 2 In the same way.~ (n v > u (1) II Since X x (0) = Y / is homeomorphic to Y and is a deformation retract of U . ( M ) Z k q . from the assumption that Y is contractible.3/4] and Y .3/4] = Y/-. where X x [0.1] = Z/ -’.1 ( U ) @ H q . X x I and Z by identifying each (z. Now we prove imA c ker(6.(U) H . It is obvious that U n V has the same homotopy type with X . II 1 I Let U = X x [0.(u n V ) + H J U ) @ I?p(v)+ H. we have H.e. Then we have M = U U V and U n V = X x (1/2.3/4).l ( V ) + ’.. In the same way. (Indiana) Solution. ( Y ) = 0. + A - H .(b) = 0.-1X. by Mayer-Vietoris sequence. we have the following exact sequence: -’ - * . o j .P.k .3/4] = Y denotes the disjoint union of X x [0. . Noting U . Let M be the mapping cylinder o f f and g . M is the identification space obtained from the disjoint union of Y . where the equivalence - - - - relation is determined by (z.(M) -+ H q .150 So ker(6. and each (z. ) C imA. ( U n V ) Z & ( X ) . let V = X x (f. 6 0 j.M Z H.. 0 jL(U) = -.l ( X ) . it is easy to see that 6 0 7’ . i. 0) II f ( z ) . o 7 ’ ojk).(V) E H. 1) g(z). o ih = 0 and j . 1) with g(z). .(Z) = 0. Suppose that X is a space and f : X + Y .

Part 111 Differential Geometry .

.

1 ii) From the hypothesis that k ( s ) 2 1 for all s. > N . y(e) = -p(e) cos e pye) sine. by Blaschke. ii) Assume further that a is a simple closed curve. of a ( s ) to be Now assume that the curvature k ( s ) 2 1 for all s. then the oval can be parameterized by 8 as follows .3. Define the diameter d. i) The following is an example of such an a ( s ) with k ( s ) 2 1 and d. Prove that d.153 SECTION 1 DIFFERENTIAL GEOMETRY OF CURVES 3101 Let a ( s ) be a closed plane curve.(e) = p(e)sin 6 p'(0) cos 8. (Indiana) Solution. + + . where the oval is counterclockwise orientated and 8 denotes the oriented angle from the z-axis to I. i) For any N E Z + sketch an example of such an a(. if we take the origin 0 as shown in the figure and denote by p ( 0 ) the distance from 0 to the tangent 1 a t the point (z. 2 is the best possible such constant). "t Fig. we know that the simple closed curve a is an oval.y) of the oval.) with d. For every oval. > N . 5 2 (or some other constant independent of a .

we can say function p ( 8 ) must be the solution to the following initial ODE origin.2 Now we can prove a more general result of Blaschke: Let two ovals C and C1 in a plane be internally tangent a t a point 0. by direct computation.(a). take the tangent point 0 of C and C1 as the common tangent line as the z-axis. Let p ( 8 ) and p l ( 8 ) be the of C and C1. Hence.154 p ( 8 ) is called the support function of the oval.1. Suppose that. a t every pair of points P and PI where C and C1 have the same tangent orientation.3.(8) of C.3 In fact. From this we can conclude that.(8) = ( p ( 8 ) + (8) ) .. the curvatures of C and C1 satisfy the inequality k l ( P l ) 5 k ( P ) . and their support functions that the support value problem of because p ’ ( 8 ) is exactly the distance from 0 to the normal line of C at point (. Fig.3. From the above. p ( 8 ) can be uniquely determined by k ( 8 ) = k. respectively. Then the domain encircled by C1 must contain the domain encircled by C. the relative curvature of the oval is k. . y(8)).. Fig.

T ( t ) = 0 and c ( t ) = -k. we assume that the parameter t is the arc length of the curve a. By the hypothesis. then owing to 0 5 4 5 8. Thus we can immediately obtain k ( t ) = constant. by the fundamental theorem of the theory of curves. N ( t ) = c ( t ) a ( t )for all t . where c is a smooth function. where 8 and its corresponding original 8 satisfy 8 + 8 = 27~.p ( 8 ) 2 0.155 Analogously. Determine all such curves. Noticing that every oval is the envelope of all its tangent lines. firstly. Hence. Secondly. we know that the curve a must be a circle (or a part of it). (Indiana) Solution. 2 0.can get we pl(8) . Then . Thus. we also have the similar expression of p l ( 8 ) . if 0 5 8 5 x. we see that the domain encircled by C must be contained in the domain encircled by C1. Differentiating both sides of the equality N ( t ) = c(t)cr(t)with respect to t . we have sin(8 . that is. If we take a circle with radius 1 and centered at a ( s g ) N ( s o ) as C1. for the curve C1. Suppose " the normal vector N ( t ) is proportional to the position vector. . The assertion of Blaschke is proved. p l ( 8 ) . by the Frenet formula. Therefore. we make the reflections of the oval C and C1 with respect t o their common tangent line a t 0 and reverse the orientation of the z-axis.p ( 8 ) 2 0.b). we have. then ii) follows immediately from the above assertion. we know that As for the sign of sin(8 . Therefore.4) - I - + 3102 Let a be a regular C curve in lR3 with nonvanishing curvature. we always have p l ( 8 ) 2p ( 8 ) . and take a as C . if x 5 8 5 2x. For convenience.

hws). Determine all triply ruled surfaces.w)du2 + 2 M ( u . Denote the Frenet frame by { T .v ) = N ( u . . (Indiana) Solution. then. there are three different asymptotic directions at every point of s. (Indiana) Solution. Noticing that the above equation is of 2nd order with respect to du : dw and it has two roots. M ( u . where L(u.dv) satisfies L(u. N ( u .156 3103 A surface S c lR3 is called triply ruled if a t every p E S we can find three open line segments L1.v) = 0 for all ( u . every point of S is a planar point. Lz. In other words. 3104 Let y : (a. a) Let a circular helix be parameterized as follows y(s) = (Tcosws.w) = M ( u . we obtain L(u. a) Show that a circular helix is an example of such a curve.v). N . a = 0. S must be a plane (or a part of it). both nonvanishing. Assume there exists a unit vector a E R3 with T .v)dudv + N ( u . v). v ) . by the hypothesis. B } . a = constant = cos a. If S is a triply ruled surface. Observe that every asymptotic direction ( d u . L3 lying in S such that L1 n L2 n L3 = { p } . Therefore.v)dv2 = 0. b) Show that N .Tsinws. v ) are the coefficients of the second fundamental form of S at point p ( u .b) 4 R3 be smooth with [y’l = 1 and curvature k and torsion 7. c) Show that k / r = constant = k t a n a .v).

T(s) f s i n a . two rays which start from the generator I . If we cut the cone along a generator 1 and develop it into a plane. l). Noticing that the central angle of the sector is f i r . then we can conclude that y never intersects itself. Thus. and the image of y on the sector must be one of the following three cases: a generator.b) 4 R3 be a C" curve parameterized by arc length. b ) . we may assume that the constant vector a = cosa. then T ( s ) . Assume k ( s ) # 0.a = 0 for all s E ( a . Hence T ( s )= (-rwsinws.0)) (Indiana) Solution. a = hw = constant. b ) . an obtuse angle. with curvature and torsion k ( s ) and ~ ( s ) . a straight line never intersecting the generator 1.k ( s ) / ~ ( s ) & t a n a . Then it is easy to verify that s is the arc length parameter.b ) . b) Differentiating T e a = constant with respect t o the arc length parameter s. d m . = 3105 Show that if y is a geodesic on the cone t = then y intersects itself at most a finite number of times. T(s ) ) N (s) . h. w = ( r 2+ h 2 ) . and the geodesic y becomes a straight line on the developed infinite sector. Differentiating the above equality with respect to 3 s. for all s E ( a .157 where T .a = 0.O. ~ ( s # 0 for all ) s E ( a . c) By the property of b). (z)y) E R2\{0. from which follows N ( s ) . we have 0 = (cos a k( s) f sin a . B ( s ) . and let T and N denote the unit tangent and normal vectors to . we obtain k ( s ) N ( s ).i are all constants. If we take a = (O. then the cone becomes an infinite sector without the vertex. 3106 Let y : (a.hw).rwcosws.

(b) From (a).r ~ ( s ) = const. namely -(1 . there exist constants T I .b). ds If we denote - T ( s )= a ( s ) T ( s ) b ( s ) B ( s ) . 1 r 1k( s ) r2k(s) + + ClT(9) = 1. ds dz ) Therefore. (Indiana) Solution.b). . The curve y is called a Bertrand curve if there exists a regular curve 7 : (a. r ( s ) ) N ( s ) b’(s)B(s). by (b). + C27(5) = 1. there exists a constant C such that r k ( s ) C T ( S = 1 for all s E ( a . dZ ds + + + Hence. In this case. namely r = const.= a’(s)T(s) ( a ( s ) k ( s ) b ( s ) . we have that r’(s) = 0.) (a) Prove that r is constant.r k ( s ) ) = const. from N ( s ) = fN(s) we know that a’(s) = b’(s) = 0. now we have - T ( s ) = -(1- ds ds rk(s))T(s) - ds -TT(S)B(S). then there exists a constant C such that r k ( s ) C T ( S= 1 for all s E ( a . Then. (a) From y(s) = y(s) + r ( s ) N ( s )it follows by differentiation that Taking inner products at both sides with N ( s ) = fN(s). (c) Suppose that y1 and 7 2 are the Bertrand mates of y. ds dz . b ) the normal lines of y and 7 a t s are equal.158 y.b). C . b) + IR3 such that for each s E ( a . + then by differentiating with respect to s . rg. 7 is called the Bertrand mate of y and we can write y(s) = y ( s ) + T N ( s )for some T = ~ ( sE) IR. ) (c) Prove that if 7 has more than one Bertrand mate. then y is a circular + helix. (b) Prove that if y is a Bertrand curve (with r as above).C2 such that for all s E ( a . (Note that s might not be an arc length parameter for 7. we obtain - k ( s ) N ( s ) .

t ( 8 ) . z(s)) = R2 with respect to s. Then we may assume that Z(S) = a ( s ) T ( s ) b ( s ) N ( s ) c ( s ) B ( s ) .159 Because the non-zero constants T I . the above system of linear algebraic equations has solution k ( s ) = const. T ( s ) = 0. Vs. Hence y must be ) a circular helix. which means that b ( s ) = -2 At last. differentiating . Assume that ~ ( s # ) 0 and k’(s) # 0 for all s.N ( s ) ) = 0. T ( s ) ) 0 with respect to s again. and a ( s ) . 3107 Let z(s) be a curve in R3 parameterized by arc-length. c ( s ) are suitable functions to be ascertained later. b ( s ) . where { T ( s ) . Differentiating ) ( z ( s ) . c(s) = . still differentiating (z(s). 1-2 are not equal. we obtain namely.B ( s ) } N is the Fkenet frame field along z(s). (Indiana) k4(s) S o ht ion. ~ ( s = const. N ( s ) ) = -& with respect to s. Show that a necessary and sufficient condition for z(s) to lie on a sphere is that 1 -+-.k’(s) k 2 ( S ) T ( S) * Conversely. we have ( z ( s ) . ( s ) . we obtain = + + 1 + k ( s ) ( 4 s ) . Differentiating (z(s).P ( S ) 1 +(s) P ( S ) = constant. Suppose that z(s) lies on a sphere centered at the origin. from which it follows that u ( s ) = 0.

3108 obtained by rotating the circle ((0. and @(t) the angle between the is normal of M and the principal normal of the curve at the point corresponding to the parameter t . Let M + c R3 be the torus (Indiana) Solution. where k ( t ) is the curvature of the curve. 1) (“top circle”). the curve c ( t ) is not a geodesic on M . z ( s ) . namely. Observe that the geodesic curvature of a curve on M can be computed as follows k . we have that vanishes identically because k’(s) # 0 and Then is a constant vector. which is parallel to the z-axis. z ) : (y 2)’ z’ = 1) around the z-axis. Is this curve a geodesic on M ? Explain without long computations. denoted by m. nor its principal normal. .m) = constant.m . because neither the top circle is a straight line. and let c ( t ) = (2cost. Then. 2 sint. Hence (z(s) .160 with respect to s. z(s) lies on a sphere centered a t m. = f k ( t ) sin€J(t). y. which is orthogonal to the z-axis. is parallel to the normal of M along c ( t ) .

then the corresponding normal curvature Ic. Local coordinate computations are not necessary. denoted by B . f'(zp)= 0. Let M be the surface in lR3 obtained by rotating the graph {(z. which means that y ( s ) B = constant.161 3109 Let y : [0. SO is a stationary point of y's tangent vector field. is a principal curvature. From il the Frenet formula we know that the binormal vector field of y.1] -+ lR3 be a C curve with 1il = 1 and nonvanishing curvature. Namely. Since every circle of latitude on M is a line of curvature. (a) Show that y lies in a plane. the curve {(z. Characterize in t e r m o f f the set of z such that && is a principal curvature of M a t (z. (a) 1. Thus. f(zp)). At P = ( z p 0. that is. f ( z ) ) : z E B}of f in the z z plane about the z axis. 3110 Let f : lR -+ lR be positive and smooth. is constant. the corresponding normal curvature where O(zp)is the angle between the normal of M and the principal normal of the circle of latitude at P. *$-J .e. y lies in a plane. that is. . -$(y(s)B) = 0. = 1 implies that the curve is parameterized by arclength s. 0. then it is an inflection point of the curve 7. f ( z ) ) . i. = implies that cos8(zP) = f l . (Indiana) Solution..O. (b) The vanishing curvature at point SO means ? ( S O ) = 0. Ic. Assume the torsion 7 = 0. " . Hint. If the point is the strict extreme value point of the tangent vector field.in the direction of the circle of latitude. Thus. (b) What happens if the curvature is allowed to vanish a t a point? (Indiana) Solution.f(z)) : z E IR} has a tangent parallel t o the z axis at P .

show that a ( s ) is a closed curve.) c R3 be a smooth curve parameterized by arclength.O. whereas the hypothesis of curvature k ( s ) > 0 implies that the plane R . Show that a ( s ) does not pass through 0 E lR3. If the position vector a(.) is always a linear combination of the binormal and normal vector B ( s ) . f(z)) is & is a principal curvature 3111 Let a(. then ( T ( s ) .) ii) If Icg(s) 1. Suppose the curve a(. we ) will get a ( s ) 5 0. Assume that the position vector a(. Integrating the obtained equality. (Indiana) Solution. Its curvature Ic a t P is just the other principal curvature of M . (Here k . Therefore. a ( s ) = const. Thus. a(. N ( s ) of a ( s ) . the trivial case. i) The hypothesis of torsion ~ ( s ) 0 implies that the curve a is a plane curve. we have ( a ( s ) .162 Besides.) is always a linear combination of the binormal and normal vectors. we conclude that the set of L such that f of M a t (z. if a ( s ) passes through the origin. T i) If k ( s ) > 0 and T ( S ) are curvature and torsion of a in E3.) E M 2 is parameterized by arclength. 3112 Let M 2 c IR3 be the cylinder z2 y2 = 1. (Indiana) + Solution. 0. the meridian passing P is also a line of curvature.)) 0. the second possible case is Therefore. show that a ( s ) is a closed curve (kg is the geodesic curvature in M).

p ( s 0 ) = (O. &A) where A = area of R. Y. then the corresponding plane curve of a . P ( S O ) = a ( s 0 ) and x(P(. we may assume that p(s) = (z(s).y.)) = a ( s > .z(s)(:9 d2 dX [& & + I & { &.f3’(s) E ” q 3 ( $ ) . the curve a c M n x must be closed.y(s). y(s). has the same geodesic curvature kg(s) 3 1. By x ( p ( s ) ) = a ( s ) . ds ’ ds ’ ds Hence Thus the problem of finding problem p(s) reduces to solving the following initial value .Y ) . i) That z-& = and z&.e.show that there exists a unique C curve p(s) E B3 such that . (Indiana) Solution. = (z.y E IR} C IR3 and ” a ( s o ) . where x(2.B’(s) E Tp(. Therefore. 0). So. denoted by &.O. & + &} + + ”> for suitable functions a ( s ) and b(s). i. z ( s ) ) where z ( s ) is unknown.) implies that d p ’ ( s ) = u(s)b(s) . iii) Show that if a ( s ) is a simple closed curve of length L bounding the region R in B2 then @(so + L ) . ii) If one develops M into a plane x. Then . d m - 1 which means that it is a circle with radius 3113 Let T be a two dimensional distribution in B3 defined by i) Show that T is not involutive. a must be closed. ii) Given a C curve a ( s ) E B2 = ((z.O).. . Thus the curvature of & is = 1 in x. are linearly independent shows that T is not involutive.163 where the curve a lies does not parallel the generating line of the cylinder M . ii) Let a ( s ) = (z(s).O) : z.

without loss of generality.O. Here s and L denote arclength and the length of y.z(so).e.where e(s) is a smooth function globally defined on y. iii) If P(s0 Q(S) is simple and closed. ii) Show that the total torsion of any closed curve y which lies on a sphere in lR3 must vanish.the unknown function z ( s ) can be and so is the space curve P ( s ) .Y(SO>. + i. N + case. i) Show that the angle through which a parallel normal vector field Y turns relative to the principal normal N along y is given by the total torsion of y. i. 3114 Call a normal vector field Y along a space curve y parallel if l is always i tangent t o y.y(so f L ) .164 When the given plane curve uniquely determined by Q(S) is C”. i. T N - = . where the sign is determined by the orientation of the curve a. respectively.fA).e. . and v ( s ) = cos O(s) .l ) 0. Thus we have ) I” r ( s) ds .P(s0) = (z(so + L ) ..TB)+ s i n e . (-ICT. B ( s ) . Therefore. then we easily have + L ) .cose k~ ( B .. i) The hypothesis that l is always tangent to y means that (v. and i i= hence. the normal vector field v has constant norm. (Indiana) Solution. . B ) B +case. which measures the angle between ~ ( s and N ( s ) . N ( s ) sinB(s) . N + (4 - T )c o s ~B..)sine. we may assume that (vI G 1. = (-sine. f ZdY) = (O.

. Suppose M and P intersect orthogonally.hence the segment is also a geodesic one. Let k ( s ) be the curvature of the intersection curve C = M n P . the smooth function e ( s ) . Show that the intersection curve (parameterized by arclength) is a geodesic on M . = ra T(S)dS +qo). yt overlaps with y = yo about po = yl.e(o) = 2n7F. along the segment where k ( s ) # 0. fi = the constant map into PO. i. Otherwise. the normal of M is parallel to the principal normal of C. where the integer n is just the degree of the map f.e. then C is naturally a geodesic on M . L] S1into R1. (Indiana) Solution. B ( L ). for every ytrwe have the corresponding map ft such that fo = f .L]can be regarded as a lift of a certain differentiable map f : [O. On the other hand. Let po E y and y contract to po smoothly on S 2 . fact that y is The closed means that ---f +)dS = e ( q . Furthermore. we may assume that for every t E [0.q o ) = J 0 L r(s)ds. Thus we get a family of curves { y t } .deg f1 = 0. 3115 Let M be a surface in El3 and let P be a plane.165 e(s) Noting that i/ is parallel to T . we have the relation qS) J. n = deg f. the unit normal vector field Y of the sphere along y satisfies the above mentioned hypothesis. Because the degree of a map is homotopically invariant. finally we have I” T(S)~S = 27r . Also. l). Therefore. . s E [O.1]. t E [0.we see that the above equality implies that = ~ ( s )and hence . ii) For any closed curve y which lies on a sphere in El3. If k 5 0.

Define a new curve 7 by setting. Hence k ( s ) = const. 0 3117 Suppose y(t) parameterizes a space curve with curvature function K ( t ) .166 3116 Let a(. where c E lR is an arbitrary fixed constant. . for each t E lR... (Indiana) Solution. and a ( s ) is a circle or part of it with radius 1 .e. we have which implies that f(s)= -k2(s). (Y(S)? (Indiana) Solution. differentiating with respect to s. Derive the curvature function for this new curve. then ( ( ~ ( s ) )= const. then a”(s) 0. Suppose that for some function f(s). If f ( s ) # 0. Further~ more.r ( t ):= cy(t/c). From 1 f ( s ) a ( s ) a ’ ( s ) -d(a’(s))2 = 0 = 2 we have f(s)d(a(s))2 = 0. k’(s) = 0. i.) = f ( s ) a ( s ) . a ( s ) is a spherical curve. i. T(S) = 0.) be a C2 curve in lR3 parameterized by arclength. a”(. If f ( s ) 0. What can you deduce about f(s).e. a ( s ) is a straight line.

ii) If the curvature k is allowed to vanish a t a single point c E ( a .N .] (Indiana) Solution.t E ( a . A counterexample is as follows. maybe.t E ( c . the connectedness of a and the smoothness of the Frenet frame field imply that T is unique. .f ( t ) . Obviously.O. the plane ?r : ( p . from which follows ( a ( t ( s ) )B ) = const = (a(t(so)). then. and a ( t ) . b ) . ) .a ( t ( s g ) ) B ) = 0 contains the curve a. according to the above discussion.Is ?r unique? ii) Is the conclusion of i) still true if the curvature is allowed to vanish at a single point c E (a. . (O. Of course. ?rl # x2. t ( s g ) = b. i) Let a be reparameterized by arclength s.b)? [Of course. B Therefore. swith t ( s 1 ) = a . suppose a = a ( t ( s ) ) . b ) must be on another plane a2.b) + lR3 be a smooth curve with nonvanishing curvature.167 Therefore. $ ( a l l ? )= 0. s E ( ~ 1 .b). but is assumed to be zero a t all other points. B } along a. The condition k # 0 means that we can ~) 0 define the F’renet frame field { T . f ( t ) ) if t > 0.0. a ( t ) . Then the hypothesis T implies that B is a constant vector field. the torsion is not defined at c. ( t . Hence. 0) if t = 0.0) where the function f is defined by . that is. Let a(t)= { if t < 0. i) Show that if the torsion of a vanishes identically then there is a plane ?r c lR3 containing a . namely. from y ( t )= y ’ ( t / c ) and y”(t) = y ” ( t / c ) / c we have 3118 Let a : ( a . a ( t )E ?r for all t E ( a . c ) must be on a plane T I . ( t .

and let s . by the Weingarten formula.. Let a be parameterized by arclength s .. E where s is arclength for a. = Show that every involute of a generalized helix a is a plane curve. 3120 Suppose that the unit normal vector to a surface M c lR3 is constant along c M . The curve . be parameterized by their arclength s and B s1 respectively. Differentiate the equation with respect t o s. p is a plane curve. Differentiate the obtained expression for . Using the Frenet formulas and the hypothesis (T(s).. we can ascertain that P ( s 1 ) = a ( s ) (so . P ( t ) lies on the line tangent to a a t a ( t )and (a’(t).) curve. u2). (Indiana) a regular curve a Solution. For convenience.a’(s)) const. let a and . we have. Then we may assume .f3’(t)) 0. that a plane contains it. s1 represent their corresponding points.. Firstly. the torsion of p vanishes everywhere. (u.B(sl) = a ( s ) X ( s ) T ( s ) . Noting that a being a generalized hrelix implies + + we obtain by straightforward calculation i.Tl(sI))= 0. and that the Gauss curvature of M vanishes a t each point of a. along a .e.= . duj o = -d= . (Recall that a is a generalized helix if for some constant vector u # 0. . and M be locally parameterized by X(ul.B successively. ..W d X n ds iJ=l 2 ds aut (Z) = -W(a’).) (Indiana) Solution.168 3119 Let a and p be two regular curves in lR3.s)T(s). a is an asymptotic (Note: A curve in a surface is called asymptotic if its acceleration is everywhere tangent to that surface.f3 is called an involute of a if for all t . Therefore.c ? . Deduce that in this case.

along a 2 2 duj duk = kgn x a' + (W(a'). we have d ( a ( s ) . n )= 0.a ( s o ) . ). (Include an explanation with your sketch. Secondly. by the Gauss formula.O) andP'(0) = (1.3. which means that the curve a is contained by a plane ( p .0). Hence a" is everywhere tangent to the surface.) (Indiana) Fig. ( a ( s ). namely.:2 ) ( + 2H(W(a').a')= 0. = where k.a ( s g ) . n )= 0.2HII K I = 0 it follows that along Q + K =. n). satisfies r" Jo L(s)ds m 3 ~ 1 2 .4 . along a . ifp's curvature function k ( s ) ( s = arclength) is odd. is geodesic curvature. from I11 . 3121 Sketch the closed regular plane curve p : [ . ( ( ~ ( s n) = const = (a(so).169 Further.T .n) = 0. Thirdly. and has the following graph. Therefore. noting that n is constant.a')n kgn x a'. T]-+ R2having p(0) = (0. we immediately deduce that.

3. This assertion follows from I” k(s)ds = 3a J.5.” g d s = 6 ( ~-) 6(0) x T .3.170 Solution.O). O ) . The image of the plane curve p is a “figure eight”.5 . Fig. too. which has z-axis as its tangent line a t (0. where 6 ( s ) is the oriented angle formed by . And y-axis is almost its “tangent” line at ( O .B’(s) and p’(0). as shown in Fig.

(B2. 2 3 ) E B3 : 2 : xi 22 = l}.Compute interior of Q J k.171 SECTION 2 DIFFERENTIAL GEOMETRY OF SURFACES 3201 i) Show that there exists a metric on the plane so that some geodesics are simple closed curves. the map x : ( S 2 \ { N } . For every p E R 2 . and B2 be the 21 o 2 2 plane. using the pull back of ds2 by (x-')*.) be a simple closed geodesic as described above and K ( z ) be the Gaussian curvature of the above metric at 2 E B2. which maps (21. we can obtain a 2-dimensional Riemannian manifold ( B 2(x-')*ds2).v).v) E B2. d s 2 ) -+ . we obtain 3 ) + + 21 = u2 + v2 + 1' 2u 22 = u2 + v2 + 1' 2v 23 = v2 . (Indiana) Solution. .1 u2+v2+1' 212 + Then the metric of S2\{N} can be expressed by ds2 = + dv2) (22 + 212 + 1 ) 2 ' 4(du2 Now.By direct calculation. i) Let S2 = { ( 2 1 . ii) Let a(. Here the integral is with respect to the Riemannian volume induced by the metric. Suppose that x : S2\{N} -+ B2 is the stereographic projection from the north pole of S 2 into the plane lR2. . then their images are simple closed geodesics on ( E 2(x-l)*ds2).its coordinates with respect to the z1 and 2 2 axes are denoted by (u.(x-')*ds2) is an isometry. Thus. 2 2 . Since all great circles which do not pass through the north pole are closed geodesics on S2\{N}. ~ 2 ~ 2 E S2\{N} to (u.

we immediately have K ( x ) = 2 T X ( D ) = 2T. interior of J ( I 3202 Let M 2 c lR3 be a surface containing x = 0. &)I = 4. iii) Because the eigenvalues of the Weingarten map L are f 5 .(O..172 ii) Denote by D the simply connected domain encircled by the simple closed geodesic a. kn(O. O . You may assume that the normal to the surface at x = 0 is ( O . i) By the Meusnier theorem. 1). By the relation among k ( O ) .(O. we know that the Gaussian curvature of M 2 at x = 0 is K ( 0 ) = -25. (Indiana) Solution. a is a normal section. &)and the geodesic curvature k. . then its curvature at z = 0 is L ( 0 ) = lk. i) What is the normal curvature of a a t z = O? ii) What is the geodesic curvature of a a t x = O? iii) What is the Gaussian curvature of M 2 at x = O? iv) Sketch the surface near x = 0. Assume that and ax 1 tangent to M a t z = 0. Then. &) we immediately obtain Ic. K ( 0 ) = det ( y4 ) = -25. or more directly.e. i. &)= 0. using the famous Gauss-Bonnet formula. and that in that basis the Weingarten map ax2 are L = ( 4 3 -4 ”) * Let Q be the curve (near z = 0) obtained by intersecting M with the z1z3 plane.(O. the normal curvature of Q at x = 0 is ii) Since a is obtained by intersecting M with 21x3 plane.

(Indiana) Solution. z = 2) and take (0. Therefore. y. z ) E M : z > 21.O. Orient M and evaluate + + + LzdxAdy+dw. &)= 4 > 0.(z2 y2)]dx A dy = + 12r d 1 2 ( 6 .(O. B . Choose the unit outward pointing normal as the orientation of M . -1) as its normal vector. we have k [ 6 . s / Fig. D = {(x. we know that x = 0 is a hyperbolic point of M 2 .3. Then.0. and w = yz2dx zzdy x2y2dz. Let P = {(z. z ) E E3 : z = 6 .173 iv) From iii).(zz y2)} (a paraboloid of revolution). z ) E E3 : x2 y2 _< 4 .6 3203 Let M = {(x. by the Stokes’ formula. Firstly. 1) and k. Noting that the normal to the surface at z = 0 is (O. Hence. we sketch the surface near x = 0 as follows.r2)rdr = 16n. y. we have that + l U p d w = 0. y.

(Indiana) Solution. . + + we have -~dw=-~4dyAdx+2dxAdy=-2 J. then y1 and 7 encircle a domain D such 2 that the Euler characteristic x ( D ) = 0 and d D = 71 uyz. 3205 Let D be the disk { ( x . J.& . Here we assume the Jordan curve theorem. noticing that dw = z 2 d y A d x 2yzd.0) = r2[1 . d y A d x = . Show that y1 and yz must intersect.O) E D. 0 ) be the usual polar coordinates and define a metric on D\{O} by + i} where h(r. r zdz A dy + dw = 8T. which contradicts the assumption of positive curvature. 3204 Let S be a surface diffeomorphic to the ordinary 2-sphere. If y1 and yz do not intersect.y ) E lR2 : x2 y2 < and let 0 = (0.174 Secondly. Suppose there is a C" metric of positive curvature on S for which there exist two simple closed geodesics y1 and 7 2 . Hence.z A d x + z d x A d y +zdz A dy + 2xy2dx A d z 2x2ydy A d z . x 2 )= ( r . Let ( z 1 . Applying the global Gauss-Bonnet formula to the domain D c S and noting that y1 and y2 are geodesics of S. we have K d c = 27rx(D) = 0.2r2 + r4 sin201'.

(b) Show that line segments in D which pass through the origin (suitably parameterized) are geodesics.+ 2lii2dxdy + li22.(1 . g12 = o(T). we see that &I = 1 + o ( r ) . y = r sin 8. .T sin adz). t 00. of I must be a continuous function of the coordinates. + Hence the metric of D\{O} may be rewritten as ds2 = dr2 + h(r. y) as the coordinates.2r2 + = [cos28 + (1 . Therefore. = giidx2 When r 0. we can express 2\{O} as the union of Il = { ( r . $& = o( 1). this metric can extend to a smooth metric az = % i +2 cos 8 sin 8[l . Hint.2r2 + 4r4 sin219)~ 8)dz2 sin2 on D .d~~.8)d82. Use a smooth coordinate system. 1 is a geodesic.175 (a) Prove that this metric extends to a smooth metric on D. Z22 = 1 + o ( r ) . Because 11 and 12 are geodesics by either metric.r sin 8d8 dy = sin 8dr T cos 8d8 + { dr = cos dx sin 8dy d8 = (cos 8dy . (Indiana) Solution.8)de2 + 4r4 sin2 8)2]dxdy +[sin2 8 + cos2 8( 1 .2r2 + 4r4 sin2 8)2]dy2 . on D\{O}. we know that both l1 and 12 are geodesics in D\{O}. { Then = r.8 = O 0 } and l2 = { ( r .8 = B0 + T } . So. and %o( l). From { follows dx = cos Bdr . as a whole line segment. 8 ) : 0 < T < :. the geodesic curvature k . e ) : 0 < r < :. Using ds2 = dr2 + h(r. Hence. (a) Set x = rCOS8. we know that kgIo = 0. If we use the extended metric. (b) For any line segment I in D . we can choose (x.

. 3207 Let H 2 = {(z. Using the result of i). &. But it contradicts the fact that . for minimal surfaces. Then. (b) Let V(0) = ( 0 . l ) be a tangent vector a t the point ( 0 . the boundary of the ball. Then. there must be at least one common point of M and the sphere with radius R and centered a t the fixed point. 6. the curvature of M n T is not less than that of S2na = S1. have the same tangent plane a t p . K ( p ) 2 &-.e. Show that the Gauss curvature of M at p is 2 ii) Show that there is no closed minimal surface in lR3. a t p . using the local canonical forms of the normal section curves M n 7~ and S 2 n7~ a t p . the Christoffel symbols). p is an elliptic point of M . . i) The surface M and the sphere S 2 . Ler R be the infimum of their radii. one concludes that the Gauss curvature of M a t p . y) (a) Compute the components of the Levi-Civita connection (i. one can easily show that. l ) E H 2 . Let V(t) = (a(t). Hence the Gauss curvature and of M at p is greater than or equal to ii) Suppose that M is a closed minimal surface in lR3. Suppose M is inside a ball of radius R and suppose a point p E M is on the boundary sphere. Let a be a normal plane of M and S 2 at p .176 3206 i) Let M be a surface (without boundary) in lR3. Then there is a family of spheres that contain the surface M inside and have a fixed center. is the usual inner product on B2.y > 0) be the upper half-plane in lR2 and let H 2 have the Riemannian metric g such that where (2.y) E lR2. (Indiana) Solution.b( t) )be the parallel transport of V(0) along the curve z = t .

Write a ( t ) = cosO(t). (a) The Gauss curvature K of S at P is defined by K = k l k 2 .u 2 ( t ) ) y) and the parallel transport vector field V ( t ) = ( a ( t ) . denote the curve ( 2 . Show that V ( t )makes an angle t with the direction of the y-axis in the clockwise direction. l ) . k2 are the two principal curvatures of S a t P.we have immediately 0 = -t.2. (Indiana) Solution. Thus. 3208 Consider the hyperboloid S of one sheet x 2 + y2 . in the direction of the circle of latitude.z 2 = 4.177 y = 1. b ( t ) = sinO(t) where O ( t ) is the angle V ( t ) makes with the z-axis.b ( t ) = sin@(t). by taking the outer unit normal vector field as the orientation of S. (a) From the hypothesis we have E = G = I F = 0 and hence ?I2 ' (b) For convenience. where k l . v 2 ( t ) ) . the normal section is a circle x 2 + y2 = 4.0). Noticing the above expression of is equivalent to rik. we see that the problem = -a(t) 9= b(t). 0 ) . Therefore. 2 . (b) Use the definition from part (a) to compute K at ( 0 . writing a ( t ) = cos@(t).1) = ( d ( t ) . . ( t ) ) = (wl(t). = ( t . the corresponding principal curvature Icl = -f. (a) Define the Gauss curvature K of S. (Indiana) Solution.v-yO)) = ( 0 .along the curve. Then b V ( t )must be the solution to the following initial value problem (vl(O). (b) At P = (0. Hint.

(z) = x E Y ( X ) will immerse M provided I E ~ is the reciprocal of neither principal curvature k l or k2 at x E M .z 2 = 4. ax aui i = 1. := P . Thus. and s1 is the spherical cap encircled by C. ( M ) at P. 3209 Calculate the total geodesic curvature of a circle of radius r on a sphere of radius R > r. then by the Rodriques equation we have on P . the Gauss curvature K = k l k 2 = -a.& k 1 ) ( 1 - Ek2)- ax dul x -# 0 au2 ax . thus the corresponding principal curvature k2 = Therefore. x ---Pc(z) -P.. : M R3defined by P. In this case.2. ( M ) -Pc(z) dU' a = (1 . locally. 3210 Let M be a compact surface embedded in R3. the total geodesic curvature is given by where C is the given circle. the normal section is a component of the hyperbola y2 . we take the lines of curvature as the parametric lines of M .(x) terms of k l and k 2 . 1 i. (Indiana) Solution. Show that the mapping P. with smooth unit normal vector field v. express the principal curvatures of M . If.& k i ) . By the Gauss-Bonnet formula.178 whereas in the direction of the meridian.(z) dU1 au2 a a = (1 . in ---f + (Indiana) Solution.

D .6 2 d z ax O=O. and the corresponding principal curvatures are ~ k .179 provided that I E ~ is the reciprocal of neither k1 or k2 at x E M .lco = (-do ds and .a . the corresponding part ---of 2) is and the corresponding points of A .y=a 1 dlnE 2~ aY 1 dlnG a ax whereas both 52are geodesics of k2. w E T(. C . B ( a . 3211 Let X : lR -+ l be a smooth function with compact support. which shows that the mapping P. C . .w ) for v . C ( a . Then. . we see that the u1 and u2 parametric lines are the lines of curvature l .y) in lR2. the surface is flat. and consider R the Riemannian surface obtained by equipping lR2 with a metric g of the form g(v. Suppose that. B .)lR2.a .. in lR2. (Indiana) Solution.1 and respectively. which is obviously by conformal to that of the Euclidean plane B 2 .sin 0 . w ) := ex(Y)(v.a ) . whose vertices are A ( . we can express the metric of the Riemannian surface k2 ds2 = ex(y)(dx2 dy2). Assuming the Gauss curvature K of this metric is everywhere non-negative. by the Liouville formula. D ( . immerses M into R3.a ) . we can see that the geodesic curvatures of the segments of coordinate curves i g and EE are + - 5. B . again by the Rodriques formula.. ( M ) . Applying the Gauss-Bonnet .y= -a k. Using the Descartes coordinates (x. respectively. Take a positive number a and consider a square domain D in IR2. D are A .-a). Furthermore. use the Gauss-Bonnet Theorem to deduce that in fact. a ) respectively.1.case + 2 O=n. in P . do 2dE 1 alnE aY coso+ 1 dlnG sin ' .

y=-a y=a which combining the above equality shows that % = 0 in B1. z ) : z 2 0). by the hypothesis K 2 0. letting a -+ 0.e.Therefore.y. dy2 - d2X namely ’ dY is a decreasing function..180 formula to 5 c IR2. R is flat.) is an asymptotic curve on M 2 . . Thus. -2 3212 Let M 2 c R3be a smooth compact surface such that M 2 c ((2. from the above integral equality we can obtain Because the number a is arbitrarily chosen. Assume that M 2 n ((2. z ) : z = 0) is a smooth curve a ( s ) . i) Show that a(. i.parameterized by arclength.y . we have -< 0 . ii) Show that a’(s) is always a principal direction. we have where the Gauss curvature Noting the area element d a = e’dxdy and the line element ds2 = ex(y)dz2 along i g and E E . it leads to Besides. K 0.

(b) Show that this is not possible if the surface is compact. then the set M 2 n {(z. z ) : z = 0) is tangent to M 2 . + X Fig. and length (ci) 0.z) : z = 0) is a tangent plane of the surface M 2 . we see that along a ( s ) .3.y. y.illustrated in Fig.a ' ( s ) .i. What kind of a set could M 2 n {(z.y. from which the Rodriques equation says that a ' ( s ) is always a principal direction with the principal curvature 0. on S.z) : z = 0) consists of elliptic or parabolic points of M 2 . iii) If we drop the assumption that M 2 n { ( z ly l z ) : t = 0) is a curve. i) The hypotheses imply that along the curve a ( s ) .} such that length (ri)-+ 0.y. Let Pi denote the points where C has horizontal tangents. which means that the tangent vector of a ( s ) for every s is an asymptotic direction. Then. 3.7. the unit normal vector field n ( s ) of M 2 is constant along a ( s ) . the plane {(z. Pi draw closed geodesics c. If there exist closed geodesics c. = . t) : z = 0) be? (Indiana) Solution. along a ( s ) . (a) Consider the following surface S of revolution generated by a C vi" brating curve C. 3213 (a) Construct an example of a non-compact C surface in " sequence of closed geodesics {c. rotating around the z-axis which is the asymptote of the curve. z ) : t = 0) is a curve. R3 with a (Indiana) Solution.O . Hence. a t which the plane {(z. = 0. y.181 iii) Let's now drop the assumption that M 2 n ((2.7 (b) Suppose S is a compact surface.. Thus. a ( s ) is an asymptotic curve on M 2 . ii) Noticing the above fact.i such . Therefore.

By the Gauss-Bonnet formula.r ( s ) ae’ e r. 0 ) + ( ~ ( s COSO..j .denotes the domain encircled by u. T ( S ) sin 8. 3214 Let ( r ( s ) .a .%+r2-=---+-. O . setting i --+ 00 in the above equality. Thus. z ( s ) ) . Ic = 1. is simply connected. when i is sufficiently large. then by the Gauss-Bonnet formula we have where a. we come t o a contradictory result 0 = 2 ~ . we have J I. u 2 ) . 3215 Let M 2 c R3 be an embedded compact surface of genus 2 1..T ( S ) T ’ ( S ) . we may suppose R. Because S is compact.ae 89 v8. On this surface ) compute the covariant derivatives V+& and V A & in terms of and 86 (Indiana) Solution. = Denoting (s..8~ 2 ~ a = . denotes the Christoffel symbol ( i . K d a = 2 r x ( M Z ) = 4 ~1( 9 ) _< 0.E~ a +-. - . I = d s 2 + r2(s)d02 Eds2 + Gd02. (Indiana) Solution. The direct computation gives & 6.e) = (ul.182 that length ( u i ) + 0.___ a G~ 8 rgl-+ 8s 21ae 2~ as 2 ~ a .we have - a vs. aas 22ae 2 a ~ e a y 2 . Show that the Gauss curvature of M must vanish somewhere on M ..a where as = l a a GI G2 r2. Consider be the surface of revolution (s.2). z ( s ) ) a unit speed curve in B3 with T ( S ) > 0.

F 2 d u d v we have immediately KdA = = T(U + TCOS u)dudv.M~ EG . 12T 12' dv cos udu = 0. Otherwise. and verify that KdA = 0 by explicit calculation. 0 < u. Therefore. M = 0. the continuous function K must vanish somewhere on M 2 . Secondly. compute its Gauss curvature function K. N = COSU. by the continuity. Finally.w) = ( ( u + r c o s u ) c o s w . y. G = (u + TCOSU)~. its Gauss curvature function K = L N . hence. ( u + TCOSU). we show that there exists another point Q with K(Q) < 0. there must be a point P with K ( P ) > 0. ( a + r c o s u ) s i n w .w < 27r.u ) ~ + z 2 = r 2 around the z-axis: T = {(z. by the connectedness. (Indiana) Solution.183 Firstly. ) ~ Parameterize this torus. r s i n u ) . Thus obtained torus-of-revolution T can be parameterized by sT X(u.4u2(z2+ y2) = 0). L = T. A straightforward computation gives the coefficients of its first and second fundamental forms E = r 2 . we would have a contradiction. F = 0.F 2 cos u T(U+ TCOSU)' Noting that the area element on T is dA = J E G . we claim that because of the compactness of M 2 . there exists a domain U of P such that KIu > 0. 3210 Consider the torus-of-revolution T obtained by rotating the circle ( z . . Z) : (z2 y2 + + z 2 + u2 - T ~ .

if IKI Kdv = 2 x x ( S 2 ) = 47r. Because the parallel translation in the plane coincides with the normal Euclidean one. “1 & . then 47r 5 Is IKldv 5 Area ( S 2 ) . By the Gauss-Bonnet formula. 5 1. for any Riemannian metric on S 2 . 5 1. we have J. (corresponding to the central angle 8 = $t in the domain D). 2 2 E be a curve on S 2 c Ill3. -sin(t). we obtain the result that. Thus. for a displacement t of a moving point p along z starting from 0. 0 < 8 < AT. the area of S 2 is not less than 47r. Compute the parallel 21 a (Indiana) Solution. 7 0 5 t < 27r T(L fils2. Let X o = translation of Xo along z ( t ) .. 1 -cos(t).8 = 27r - (i. Consider the cone that is tangent t o the unit sphere S 2 along the curve z. the oriented angle formed by the tangent vector with the parallel translation vector X ( t ) of XO is given by 27r .184 3217 Let z(t)= (. . where K is the (Indiana) Solution. 2 3219 Define “geodesics”. This cone minus one generator is isometric to an open set D c lR2 given in polar coordinates by 0 < p < +m. and characterize (with proof): i) All geodesics on the unit sphere S 2 c IR3. q) 3218 Show that for any Riemannian metric on S2 with II K Gauss curvature.

3220 Give an example (e. and kg is the geodesic curvature of a. du u +Cr*. On the contrary. z ) E R3 : z2 + y2 E 1). every great circle is a geodesic. or a straight generating line. n is the unit normal vector field of M along a .. k. every geodesic must become a straight line. Along a . ii) Since geodesics are intrinsic objects.k i = i.g.j ddsk = o -ds j. or a circle of latitude. draw a picture) to show that a connected surface can have two points which are not jointed by any geodesic. every geodesic on the cylinder must be a helix. i) On the unit sphere. Therefore. the principal normal vector N is parallel to n. a is a geodesic of M if and only if along a .2. that starts from a given point and is tangent t o a given direction. Let a ( s ) be a curve on a surface M parameterized by arclength. (Indiana) Solution. because along a .185 ii) All geodesics on the surface ((2. then if we develop the cylinder to a plane. we have d ( S ) = Ic(s)N(s) where u1. y. must be a great circle. E 0 or - d2ui ds2 .u2 are local coordinates on M . owing t o the uniqueness of the initial value problem every geodesic on M . Then. What is the usual topological hypothesis that prevents this problem? (Indiana) .

Then the surface K\{T} is a case in point.cosv) 1 . w) = -(cos Jz u . “Completeness” is the usual topological hypothesis that prevents this problem. i) Compute the metric [gjj] on the Clifford torus for a coordinate chart of the indicated type. cos where u and v are constrained to lie within intervals length < 2 ~ . (Indiana) Solution. iii) Deduce from the result of ii) that the Euler characteristic of a torus is zero.-(0. Let K be a plane and p . i) From x “---(-sinu. y two points in K. cosh(t) sin(8)) is a minimal surface in R3. (Indiana) Solution. Hint. Let T be an interior point of the line segment p y . 3221 Define “minimal surface in R3”.cosu.2 F M + G L = 0.0.186 Solution.-sinv.O. A minimal surface is a surface with mean curvature H=- 1E N .sin w). Calculation is not necessary here. cosh(t) cos(8).8) = (z. and prove that the catenoid.O). sin u. 3222 The “Clifford” torus in S3 can be parameterized using charts of the form 1 X ( u . 2 EG-F2 The straightforward calculation shows that the catenoid X ( Z . ii) Figure out the Clifford torus’ Gauss curvature function.is minimal. A 1 X “-a . w. obtained by rotating the graph of y = cosh(z) around the z-axis in R3.

f ( 0 . ii) The Gauss equation implies that the Clifford torus’ Gauss curvature function K 0.2) at ( O . O ) . we have Therefore. O . O ) ) E B3 are the roots of the following equation . O ) ) E R3 are the same as the eigenvalues of the Hessian matrix [ a 2 f / a z .g. Hint.187 we easily have 911 = 922 = -. i) Show that the principal curvatures of the graph z = f(Z1.22) at ( O . iii) Using the Gauss-Bonnet formula.. i) The hypothesis means that $ and vanish at (0. (Indiana) Solution. a 2 jat ( O . ] ii) Show that lR3 contains no compact embedded surface with strictly negative Gauss curvature at all its points. O . 2 1 g12 = 921 = 0. a minimum R or maximum) at the origin 2 1 = 2 2 = 0. Then. by straightforward calculation. . we see that the Euler characteristic of a torus T 3223 Let f : IR2 + l be a smooth function with a critical point (e.O). f ( O . the principal curvatures of the graph z = f(Z1. Look a t the “lowest” point on the surface.

and in terms of the local coordinates (z. (The existence of such a point follows from the compactness of the surface.O. { Y= 2 = 2(u. then T is the tangent plane of the surface a t p . Then. has smooth inverse { 2) =4ZlY) = v(2. ~ -+ ( z ( u .v) Y(U. Y)). z ) for U .~ ( o = upn M . 1) as the unit normal of the surface at p. for each coordinate neighborhood V c M of p E M ..Y>. y. v ) . ~ ) . such that on V. there exists a neighborhood U of p in R3. Using the local coordinates ( u . we can express this embedding as ( u . v ) ) . ) Noticing that we know from the implicit function theorem that there exists a neighborhood V. such that V = U M . if we take (O. it is easy to conclude that any normal curvature of the surface at p is greater than or equal t o zero. z ) = z . f ( 2 .188 ii) Let p ( z 0 . Thus the Gauss curvature of the surface a t p is not less than zero. we can find a neighborhood Up c U of p in lR3. the embedding can be expressed by ( 2 .Y) (z. z ( ~ . Therefore.v ) for V and (z. ) - . yo. = z ( u ( 2 . Y. (Indiana) Solution.y). y ( u . 3224 Let M be a 2-dimensional manifold smoothly embedded in lR3 with unit normal n. Since the inclusion map M + lR3 is an embedding. Y) Y)). By observing the normal section at p . c V of p . = Up n M . such that V.~ ( 2 . y . where f(z. a) Prove that for each p E M there exist an open neighborhood U p of p in lR3 and a smooth function F : Up -+ B2 such that F-l(O) = Up n M . zo) be the “lowest” point on the surface. y ) . b) Find F if M is the graph of a smooth function f : lR2 + lR.f ( z . setting F : Up ---t R3 by F ( z .y).) Since the surface is above the plane T : z = zo and p is the common point of R and the surface. we have F .

f ( z .then for each p E M . z ) = z .>> # 0. . where n(u. prove that is a diffeomorphism onto S2.and F : R3-+ R defined by F ( z . r(. [Recall that S C R3 is defined to be convex if for each two points of S the line segment joining these points lies in S. c) Assuming that 71 is one-to-one.> rank = (4% .P(. y ) . K ( P ) d P = 4?T. Xu x Xu Jacobi matrix the rank of the ( ffu ffv 3 /: ) = 2. = n ( u .which is a diffeomorphism. b) and c). define q ( p ) as the end point of the unit exterior normal n(p) after parallel translating it to the origin of E3..] a) Define the Gauss (or sphere) map 7) :M --$ s2= {(z. Assume that n is the exterior normal and that the Gauss curvature K of M is everywhere positive.+ S2 is a local diffeomorphism. Assume that M is the boundary of a bounded convex open set.. y. z ) : x 2 + y2 + z 2 = 1). . there exist a coordinate neighborhood C c M of p and a coordinate map h : C -+ U c B 2 .. a) For each p E M . d) Show that J.v) is the unit normal at X ( u . (You may assume b) and c) if you wish.) (Indiana) Solution. such that on U the Gauss map has the following expression q 0 h-'(u. For each p E M . You may use without proof the fact that M lies on one side of each of its tangent planes.). y . 3225 Let M be a 2-dimensional manifold smoothly embedded in R3 with unit normal n. 211. = K X .189 b) If M is the graph of a smooth function f : R2 -+ lR. 7) b) Prove that 7) is one-to-one. . We first prove that q : M . we can take Up = R3.v). x Xu and K > 0. Since nu x n. v) E C C M .

Then the Gauss-Bonnet formula gives 1. noticing that differentiability is a local property. it is easy to show that M is compact. From the above argument we know that there are neighborhoods C p . Since 7 : M -+ S 2 is a local diffeomorphism. : Cp 7 ( C p ) . q is an open map. q = (7o h-') o h is a diffeomorphism. 5. such that 7Ic. . closed and non-empty set of S 2 . Besides. Suppose that there are two distinct points P. in M we can choose C ~ . + - u = ( v I c p ) . set . Next. we see that the globally one-to-one. q l c .l is a local diffeomorphism. u n v = 0 and q ( U ) = q(V). surjective local diffeomorphism is naturally a global diffeomorphism. I . which implies q(M\U) = S2. C Q ) so small that C p n CQ = 0. Q E M such that q ( P ) = q ( Q ) . because M is compact and S2 is a Hausdorff space. : CQ -+ ~ ( C Qare diffeomorphisrns. in the neighborhood of h ( p ) . connected and oriented. on M . Thus. we show that q : M S 2 is surjective. Namely.1 ( 7 ( C P )n 7(CQ)). the implicit function theorem says that 7 o h-l has a smooth inverse on a neighborhood of h ( p ) (which may be smaller than U ) .hence 7 o h . where da.Q respectively. fi have local expressions Since Kda > 0. (by noting K > 0). da = Hence Kda = 4n.190 Thus. On the other hand. Therefore. we arrive at a contradiction.respectively. Now we prove that q : M ----t S2 is globally one-to-one by leading to a contradiction. 7 is a local diffeomorphism. The connectedness of S 2 implies that q ( M ) = S2. Thus q ( M ) is an open. Now take the inverse images of ~ ( C P ) ( C Q ) under nq q l ~ and q l cQ . Because P # Q. CQ of P. = (vIc*)-'(7(cP) v " rl(CQ))- Thus. 7 is also a closed map. In the end.

this problem is the famous Hadamard theorem. intrinsic quantities are those that are invariant under isometric correspondence. Using the Hopf’s rotation index theorem and the Gauss-Bonnet formula. they are extrinsic.is intrinsic. d) Assuming c). we know that the covering map q must be a homeomorphism and hence a global diffeomorphism. then the local diffeomorphism q : M -+ S2 is also a covering map. For example. Firstly. In other words. (Indiana) Solution. they are called “extrinsic quantities”. a) The terminology “intrinsic quantities” means those geometrical quantities that are definable only by the first fundamental form of M and its derivatives. c) Although the covariant derivative on M is defined by using the covariant derivative on R3. Aw J do so . we can simplify its proof as follows. Since M is compact and S 2 is connected. In fact. Otherwise. Using the theory of covering map. defined using the covariant derivative on R3. but extrinsic ones are not. Therefore. let C be a simple closed curve encircling a simply connected domain D where p lies. last local expression of the covariant derivative of M the involves the tangent vector field and the Christoffel symbols of M . b) The principal curvatures axe not intrinsic. we can show that the Gauss curvature of M a t p can be expressed by K ( p ) = lim D-+P -. because S 2 is simply connected. consider a plane and a cylinder. and M c R3 is connected and hence path connected. d) For each p E M . Further. discuss why the Gauss curvature of M is intrinsic.191 Remark. Let Aw denote the angle variance caused by a tangent vector after parallel translating it around C once. show that q : M --t S2 is a local diffeomorphism. a) Explain what is meant by intrinsic and extrinsic quantities on M . it is intrinsic. as the above. b) Are the principal curvatures intrinsic? c) Discuss why the covariant derivative on M . 3220 Let M be a 2-dimensional manifold smoothly embedded in R3with unit normal n.

we get a surface of revolution M 2 c R3. Thus the Euler characteristic number x ( M 2 )= 0. then the Gauss curvature is intrinsic.) (Indiana) z I I Fig. 3227 By revolving the curve y sketched below around the z-axis. by the famous Gauss-Bonnet formula. Compute JM K d A .3.8 Solution. show that the Gauss curvature of a ruled surface in Rfl is nowhere positive. because along these circles the normal vector of M 2 is parallel to the principal normal vector of the two circles respectively. where K is the Gauss curvature and dA the area form. (Indiana) . l). the boundary of M 2 consists of two circles which are just geodesics of M 2 . (Make sure to justify your answer. b) Show it for arbitrary n.192 Since parallel translation is intrinsic. 1)2-+ N such that for each fixed wo E (0. on M . wo) is a geodesic in N. The surface M 2 of revolution generated by the curve y is homeomorphic t o a section of a cylindrical surface. we have immediately K d A = 0. 3228 A surface C2 immersed in a Riemannian manifold N is said to be ruled if it can be parameterized near any point by a mapping X : (0. Therefore. Besides. a) When n = 3. too. the u-parameter curve X ( u .

F2 - b) Similar to the above. l ) . suppose C2 can be characterized locally by where. for convenience. However. we can easily deduce that the first coefficient of the second fundamental form of i)l(w). EG. C2 d 2X L = (n. 1Z(v)1 SE 1. Thus.(a’(v). a) Since every geodesic in N = R3 is a straight line.193 Solution. we have Noting that 1l’(v)12 . by a routine work. -) dU2 f0.Z’(w))2 = la’(w)x I’(w)I2 2 0.iw2 <o. which shows that the Gauss curvature of the surface K= LN .F2 EG. l ) is the arclength of the curve a ( v ) . and (a’(u).w) = a(w)+(uwhere Z(v) is the direction of the geodesic corresponding to w E ( 0 . we see F 0. then C2 can be characterized locally by X ( u . we may assume that v E ( 0 . we obtain the desired result. that the first fundamental form of C2 is ds2 = d U 2 + (1+ 2u(a’(w). through computation. .l(w)) Then.Z’(v)) + K = U2II’(V)12)dV2. Since the Gauss equation says -* dz’ it suffices to show that 2 0.

Y } be an orthonormal basis of R . Noticing that field.. E T'(M"). R ( X .Y)Z = V x V y Z . using the properties of covariant differentiation " and of the inner product. Y ) = R ( X .Y . we can conclude from straighforward computation that ( 1) + + is an orthogonal matrix. where R( ) is the Riemannian curvature tensor field of M " . define the Riemannian curvature operator by R ( X . Z .Y.( L X . If ( 2 . Explain why it is a tensor. y.Y ) Z = ( L Y . ii) For any vector fields X .W ) . (Indiana) Solution. Use this basis to define the sectional curvature K ( R ) and show that it depends only on R and not the particular basis chosen. Y . i) Let { X . f Z . Z.Z ) L Y . we can easily have R ( X . W ) W = R ( X . 2.. R(fX..W on M " .R ( X .194 SECTION 3 DIFFERENTIAL GEOMETRY OF MANIFOLD 3301 Let M" be a Riemannian manifold and R a 2-plane in T. then 2 = aX by.Y ) Z . Therefore.X . z. . Y . is dependent only on X. W)I. Z .V[X.Z ) L X . fW) = f R ( X . Y ) is another orthonormal basis of R . and R( ) is a 4th order covariant tensor ---- z.2. Using this or any other valid method. Y .Y. Y ) . in terms of local coordinate frame field. ) = R ( X . iii) Recall that if M" c En+' is a hypersurface.respectively. fY. W. Y . M. where L is the Weingarten map.W ) = ( R ( X .W ) . Thus. X . then R ( X . Y ) .Y]Z and the Riemannian curvature tensor field by R ( X . which proves that K ( R )depends only on R . W ) = R ( X . compute all sectional curvatures of the sphere {z E B4: 121 = 3). ? = cX d Y . one can show that V p E M " .y. Y . ii) Define the Riemann curvature tensor in terms of covariant differentiation.X .2. Y . i) The sectional curvature is defined by K ( R ) = .V y V x Z . Then for any C function f on M " .

X / 3 . Suppose F : lR3 + lR4 is a 1-1 C immersion in some neighborhood of C. Besides. X ) ( L X . - - - K(7r) = . one can easily show that V e i z= e i .Then by the original definition of : covariant differentiation. Y ).( R ( X . The image " F(C) is then a compact Riemannian submanifold of lR4 with boundary and therefore has a volume. { e i } be a local orthonormal Let frame field about 3 (as a point) of M n . if { X . Y } is any orthonormal basis of an arbitrary 2-plane T in T p ( M n ) . X ) ( L Y . the Weingarten map is L : X F-+ L ( X ) = . Hence. Therefore.1zl= 3 } . Hence. " iii) For the sphere M n = { z E lR4. Justify the following formula: .195 R( ) I p is a well-defined 4th order multilinear function. R( ) thus defined is a C tensor field. Y ) X . Y )= 1/9.V x g = . the inner product and covariant differentiation are all C". where V denotes the covariant differentiation in lR4. we can regard the position vector field 2 as the normal vector field of M n . Y ) = ( L X .( L Y . 3302 Let C = {z E lR3 : 0 5 3ci 5 1) be the unit cube in B3.

196

Let Au be the volume of the parallelopiped spanned by the vectors PzP;, P3Pl and in lR4. Then

0, Ax3 0. We take the principal is an infinitesimal when Ax1 -+ 0, Ax2 part of the infinitesimal Au as the volume element, namely, the volume element of F ( C ) is
+ ---f

Hence the desired formula follows immediately. Furthermore, if F ( x ) = ((1 XI)^, (1 ~ 2 )(1 , ~ 3 )(2 , ~ 3 ) ~then, by ~ ~ ), , denoting F(x) = F(z1,22,23) = (yl, y2, y3, ~ 4 ) we have that

+

+

+

F,

(&) =

2 ( 1 + ~ 1 ) -a- , aYl

= 2(1+x2)-,

a

aY2

= 2 ( 1 + ~ 3 ) a+ 3 ( 2 + ~ 3 ) dY3
a a Since - - and ayi %a

2

-.8
dY4

form an orthonormal frame, we obtain that

3303
There is no submersion from S3 into lR2.

(Indiana)

197

Solution. ( + ~ + Let S3 c E4be defined by { ( x ’ , x 2 , x 3 , x4) E E4; x ’ ) ~ ( x ~+) ( x 3 ) ) ” (2‘))” = 1). Observe the following three vector fields

where (xl, 2 , z 3 ,z‘) E S 3 . It is easy to see that they are nowhere-vanishing x tangent vector fields on S3. Since ( x ’ ) ~ (z2)>” ( 2 ’ ) ) ” ( x ~ = ~ then, ) 1, without loss of generality, we may assume x4 # 0. Hence, from the fact that

+

+

+

det

(

z2
x4
-23

-21

24

x3
24

-x2
21

)

= ( x ‘ ) ~ x4(x3)’ + ~

+

~ ( 2 ’ ~) ~ ( 2 ’ x4 ~ ~ )

+

=

# 0,

we know that X I ,X2, X3 are three linearly independent vector fields. Furthermore, by direct calculation, we obain

Now we suppose that there is such a submersion a : S3 --+ El2. Then, a : T p ( S 3 -+ T T ~ p ~ (is a surjection for every point p E S 3 . Thus, we may , ) E2) assume that, for example, a*Xg,7r*X3 are linearly independent, and 7rIT,X1 = aa,Xz b7r,X3 a t p . Then, on the one hand,

+

on the other hand,

Therefore,

a,X3 = -aba,Xz - b2a*X3
from which we get ab = 0 and -b2 = 1. Similarly, from [ X s , X 1 ] = 2x2 we can deduce that -a2 = 1 and ab = 0. They are all contradictory. So, there is no submersion from S3 into E z .

198

3304
Let M n and N k be Riemannian manifolds. Then M x N is naturally a Riemannian manifold with the product metric. If XI, . . .,x, are local coordinates on M and y1, * * * ,yk are local coordinates on N , the product metric in local coordinates ( X I, , x, , y1, . ,y k ) looks like

-

i) Let X be a vector field on M x N “along” M (i.e., in local coordinates no are present) and Y be a vector field along N . Show that DxY 0. ii) Show that a t z E M x N , some sectional curvatures always vanish. (e.g., product manifolds in the product metric never have strictly positive curvature.) (Indiana) Solution. i) Using the properties of the Riemannian connection on M x N , we only need to verify D e = 0. In fact, observing that the inverse matrix of g is of the form
yp and

&

&

and the first class of Christoffel symbols of M x N satisfy

we easily conclude that

Obviously, we also have D B y s azi = 0. a p

ii) By the definition of the curvature operator, using the result of i), we

= 0, Hence, for arbitrary X along M and Y This means that R along N , using the Cm(M x N)-linearity of the curvature operator, we have

(&,&)

R ( X ,Y, Y ) = 0. X,

Therefore, the sectional curvature determined by X and Y always vanishes. Obviously, here X may involve yp, and Y may involve 2,.

3305
Let F : M --t N be smooth, X and Y be smooth vector fields on M and N , respectively, and assume F,X = Y ; that is, that

a) Let w be a smooth 1-form on N . Define the Lie derivative Lyw of w with respect to Y . (If you use local coordinates, you must verify independence of choice of coordinates.) b) Prove that F*(Lyw)= Lx(F*w).

c) Let Z be a smooth vector field on M such that F,Z = W , where W is a smooth vector field on N . Show that L y W = F , ( L x Z ) . (Indiana) Solution. a) L y w is defined by Lyw = d ( Y [ w ) + Y Ldw, where the symbol ''Y denotes the interior product of a vector field with a form, for example, if R is a pform, then Y [R is a ( p - 1)-form defined by

Thus, Lyw is a well-defined 1-form on N . b) Let Z be a smooth vector field on it4 such that F,Z = W , where W is a smooth vector field on N . Then we have

200

c) Noticing that L y W = [Y,W]and [F,X,F,Z] = F , [ X , Z ] , we immediately obtain L y W = F , ( L x Z ) .

3306

Let
w =[

T,

8, z be the usual cylindrical coordinates in R3. Let
cos 8+5r2 sin28]d8+[-2r cos 8+6zr sin 8]dz+[4zr2sin 8]dr.

~ T sin 8+3z2r Z

Let the curve y be given in rectangular coordinates by
y ( t ) = ( z ( t ) , y ( t ) , z ( t )= (cost,sint,4sin5t+sin2tcosst), )

o 5 t 5 27F.
(Indiana)

Evaluate

s,, w .
f be defined by
f(r,8,Z ) = -2rz cOs8

Solution. Let a function

+ 3z2rsin 8,

(T,

8,

.>

EB ~ .

Then

f is a C” function in lR3 and we have

df = [ ~ T s i n 8 + 3 z 2 r ~ ~ ~ 8 ] d 8 + [ - 2 r ~ ~ ~ 8 + 6 ~ ~ s i n 8 cos8+3z2 z Z ] d z + [ - 2 sin81dr.
It is easy to see that f I-, is also a C” function, and by the invariance of the form of first order differentiation, the above expression of df is an exact l-form on the closed curve y. Noticing that y is a closed curve, we have, by using the Stokes’ theorem,
w

=

I

(w - d f ) =

I

5r2sin2 8d8 + (4zr2sin 8 + 22 cos 8 - 3z2 sin 8)dr

Without loss of generality, we may assume 0
i w = lr5sin28d8

5 8 < 27r and T = 1. Then
d8 = 57~.

=5

201

3307
Let M be a C" Riemannian manifold. Assume the theorem that there is a unique C" mapping V : X ( M ) x X ( M ) + X ( M ) denoted by V : ( X , Y )-+ VXY which has the following linearity properties: For all f , g E Cm(M)and X , X ' , Y, Y' E X ( M ) , we have

= f(vxy)+S(Vx,Y>, V x ( f Y gY') = f V x Y g v x y ' ( X f ) Y (Xg)Y', [ X ,Y ] = VXY - V Y X ,

VfX+gX, Y

+

+

+

+

X ( Y ,Y ' ) = (VXY,Y ' )

+ (Y,VXY').

a) Suppose a : [ q b ] --+ M is a smooth curve. Define what it means for a vector field Y on a to be parallel along a . Derive the differential equations that must be satisfied if Y is parallel along a. b) Let X , Y E X ( M ) . Let p E M and let a : [0, b] ---t M be an integral curve of X such that a(0) = p , da/dt = X ( a ( t ) ) . Show that

where P,,o,t : T,(o)M

+ T,(t)M

is the parallel transport along a. (Indiana)

Solution. a) As known, thus defined mapping V is the Riemannian connection. Using the properties of V , we can prove that V % Y is completely determined by a ( t ) and Y ( a ( t ) ) So V % Y is well-defined for every vector field along a. . Now we give a definition that a vector field Y on a is parallel along a , if and only if V % Y = 0, Vt E [a,b]. In order to derive the differential equations, we choose a coordinate neighborhood with local coordinates {xa}. Then

Denoting

202

we have, by the properties of V ,

Therefore, the desired differential equations are

b) Choose a basis { e l , - - - , e , } of T,(M), where n = d i m M . Let

Since P,,o,t is a parallel isomorphism for every t , then { e l ( t ) , . . , e,(t)} is a basis of T a ( t ) ( M ) .Hence we can denote

Then, by the properties of the connection V and by the fact that e i ( t ) is parallel along a ( t ) ,i = 1,.. + n, we immediately have ,

On the other hand, because

is equivalent to

ei = p ~ h , t ( e i ( t ) ) ,
we can write

203

3308
Let M points p , q dependent identically be a Riemannian manifold with the property that given any two E M the parallel transport of a vector from T p M t o T,M is inof the curve joining p and q. Prove that the curvature of M is zero, i.e., R ( X ,Y ) Z = 0 for all X , Y ,Z E X ( M ) . (Indiana)

Solution. For a n arbitrary point p E M , take a coordinate neighborhood D of p with local coordinates zl, . ,zn . Let

-

be n linearly independent vectors in T p M . Using the hypothesis that the parallel transport of a vector from T,M to T,M is independent of the curve joining p and q , then for q E D, we can transport every Vpi from T p M to T,M. Thus, we can define n linearly independent vector fields V1, . . . ,V , in D. Obviously, all V;’s are well-defined, and if we transport along special coordinate curves, then

namely
k wll,i(z) = 0 i, j , k = 1, * * * ,n.

Now, by the Ricci identity, we have
k k vil,jrn(zC)- v i l , m j ( z ) =

-

C v11(z>Rfjrn(z)= 0.
I

Noting the linear independence of

K’S, we immediately

have, in D,

Rljm= 0 k , l , j , m = l , - - . , n , k
i.e., the curvature of M is identically zero.

204

3309 Let M be an n-dimensional Riemannian manifold. Suppose there are n orthonormal vector fields XI,. . * , X, that commute with each other (i.e., [Xi, Xj] = 0; i,j = 1, .,n), show that the sectional curvature of M is identically zero. (Indiana)

Solution.
Set
k

Then, from (Xi,Xj) = 6;j it follows that
(vxkxiIxj)

+ (xi> 'Xkx'j)

= O!

i.e., T i i + rtj = 0; i,j,k = 1,. . ,n. On the other hand, from [ X i ,Xj] = 0 it follows that vx;xj - VX,Xi = [Xa,Xj] = 0, i.e., rk. - r!. = 0; i,j,k = 1,.. ,n. Namely, the Christoffel symbols '3 3% are antisymmetric with respect t o k,j, and symmetric with respect to i , j . Therefore,

rk. = -rJ
'3

ak

= -rii = rfj= rfk= -rk. = -r.. k
3'

a3

= 0; i, j , k = 1, . . , n. Hence the sectional curvature of M is which means identically zero.
3310
Let X be a smooth vector field on a Riemannian manifold M . The divergence of X, denoted by div(X), is defined by the function trace ( V X ) . i) If M is closed (i.e., compact without boundary), show that
/M

.

div(X)dv = 0.

ii) If M is compact with boundary dM, show that div(X)dw =
JM

1,

(X, N ) d s ,

205

where N is the outer normal vector field of a M . Hint. Consider w E h l ( M ) defined by w(Y) = ( X , Y ) , and try to use the Stokes' Theorem. (Indiana) S o ht ion. In fact, this problem is the famous Green theorem. The outline of the proof is as follows. Firstly, one can show that, for example, by means of the normal coordinate system about p E M , thus defined div(X) satisfies div(X)QM = d(i(X)Qm), where Q M is the volume form of M and i( ) is the interior product operator. Next, for p E d M , choose an oriented orthonomal frame field about p {el,...,e,} such that, at p , el = N p . Let { w l , - - . , w n } be the dual frame field of {el,. . , en}. Then the volume elements of M and d M are respectively

dv = n ~ ( p=) w
Observing that, a t p ,

~ A * * * A w ~ s,= R a M ( p ) d

= W 2

A*..Awn.

i(X)n,

= i(X)(w1 A . . . A w n ) = w1(x)w2 A . . A w n + (terms involve w ' )
=

(x,N ) Q ~ M (terms involve wl), +
d(i(X)QM) = J,,(X,
N)%M

and along d M w1 s 0, one can obtain, by Stokes' theorem,
/M

div(X)dv =

= J,,IX,

Wds.

If M is compact without boundary, the right hand side of the above formula vanishes naturally.
3311

Let w l , .. * , wk be one-forms. Show that {wi}fZl are linearly independent if and only if w1A w2 A . A wk # 0. (Indiana) Solution. Let wl,- ., wk be defined on an n-dimensional manifold with n 2 Ic. Suppose w1 A . . . A w k # 0. If wl, ,wk are linearly dependent, then without loss of generality, we may suppose that
1 9 . .

W k =Ulwl$"'+Uk-lw

k-1

206

with suitable functions a l ,
W1A
* * *

,a k - 1 .
* * *

Thus we have
(UlW'

A

WWk

W1A

A Wk-' A

+ . . . + ak-1Wk-l)

T

0,

which contradicts the above hypothesis. . On the contrary, if wl,.. , w k are linearly independent, then we can extend Thus them to a basis {W~,...,W~,W~'~,...,W~}.

w1A
implies that

... A w k A wk+l A ... A w k # o

3312
Let M be a Riemannian manifold. Let p E M (a) Show that there exists S > 0 such that exp, : Bb(0) c T p M -+ M is a diffeomorphism onto its image. (b) Show that there exists E > 0 such that exp,(B,(O)) is a convex set. Hint. let d ( z ) = distance from z to p . Show that d2 is convex in a neighborhood of p . (Indiana) Solution. (a) This is just the existence of the normal neighborhood of p . Use the fact that d exp, is nonsingular at p, and then the implicit function theorem. (b) The existence of convex neighborhoods is a classic result due to J. H. C. Whitehead. Refer to every standard textbook on differential geometry.

3313

Show that (i) A is a differentiable manifold.

207

(ii) A is a Lie group with the standard matrix multiplication as a product. (Indiana) Solution. Define a map F : Gl(2,LR) + Gl(1,LR) by F ( X ) = d e t X . Then F is a smooth homomorphism between Lie groups, and the rank of F is constant. Therefore, the kernel of F k e r F = F-'(l) = A is a closed regular submanifold of GZ(2, El) and thus a Lie group.

3314
(a) Let f be a smooth function on a Riemannian manifold M. Let gradf be the vector field defined by the equation

(gradf, w ) = dpf v , ~

w E TpM.

Let (z',. . . , 2") be local coordinates around p . Find the expression for gradf in terms of z', ,2". (b) For a vector field X define the divergence of X, div(X) as the trace of the operator Y + DyX where D is the Levi-Civita connection. Find the expression for the divergence of X in a local coordinate system (z', . . . , P). (c) Use (a) and (b) to find the expression for the Laplacian A in local coordinates, where A acting on a smooth function f is defined by Af = div(Vf). (Indiana) Solution. For convenience, we omit the suffix p . (a) Let

--

grad f = Then, from

c -.
1

a1 8 ax1

it follows that
k

208

Thus,

(b) Denote

x = c x i - a: . ax
i

Then, by the definition of divergence, we have

Af

= div(gradf) =

c
i

ik

af

3315 Let M be a compact connected Riemannian manifold without boundary. Let f be a smooth function satisfying Af = 0. Show that f = const. Hint. Use the definitions in Problem to show that div( f V f) = IV f 12+ f A f . (Indiana) Solution. For any function f , by straight calculation, we have div(fVf) = lVfI2

+ fAf.

Now, noticing the hypothesis of this problem, by Green’s theorem we have

J, lVf l2dW = 0,
which implies df = 0 everywhere, that is, f = const.

209

Suppose F : M N is a smooth map between differentiable manifolds, and is homotopically trivial. Show that in this case, F*w will be exact whenever w is a closed 1-form on N . (Note: F is homotopically trivial if it extends to a smooth mapping : M x [0,1] N such that B(z, = F ( z ) , for all z E M , 0) while F(z, 1) q E N ( q constant) for all z.) (Indiana) Solution. Consider the map G : M + { q } and the inclusion i : { q ) N . Then the map F : M -+ N is homotopic to the composition ioG. Furthermore, we know that F* and (i o G)* = G*o i* induce the same homomorphism
+
---f

-

3316

F** = ( i o G ) * * H 1 ( N , d ) . *

-

H 1 ( M , d).

Since i * ( Z 1 ( N d)) c Z1({q),d ) = 0, the induced homomorphism F** = (i o , G)** is a zero homomorphism, i.e., F * ( H 1 ( N , d ) )= 0. In other words, for every w E Z 1 ( N , d ) ,F*w E B 1 ( M , d ) ,namely, F*w is exact.

3317
Regard lR9 as the space of all 3 x 3 matrices with real entries. Does the subset C = {A E Rg: det(A) = 0) form a smooth submanifold of R9?

(Indiana) Solution. Observe that {A E lRg : rankA 5 l), the union of all axis in Rg, is closed. Then M := R9\{A E R9 : rankA 5 1) is an open submanifold of Rg. Define a map F : M + B1by F ( A ) = det(A), A E M . Noting that dF = ( A l l , A12, A13, A21,A22, A23, A31, A32, A33) where Aij is the algebraic complement of the corresponding entry aij of A, we see that rankdF = 1 on M . Therefore, F-l(O) = {A E IR9 : det(A) = 0) is a closed regular submanifold of M . Hence, it is also a submanifold of lR9.

210

3318

Let

Compute w , where S3 = { x E IR4 : 1x1 = l), oriented as the boundary of the unit ball (assume standard orientation on B'). (Indiana) Solution. Denote D4 = { x E IR4 : 1x1 5 1). Then, by the Stokes theorem we have

ss3

= J,,4w=J,4dw =

L

(dxl A d22 A d23 A d24
A dxl A d22 A d24)

+ d22 A dxl A d23 A d24
-. 2
x2

+dx3

=

J,. dxl

A d22 A d23 A d24 = v01(D4) =

3319

Prove that

is a three-dimensional submanifold of IR4.

(Indiana)
Solution. Define a map F from E4\{(0,, O , 0)) into B1by O

" Then F is a C map with ( 2 4 - 2 3 - 2 2 x l ) as its Jacobi matrix which has constant rank 1 on IR4\{(0, 0 , 0 , 0)). Thus
(21, 22,23,24)

: rank

Y) = ( f l .llY1I2. 0)) with dimension 3. 3320 Let vector fields XI.Y)). O . (Indiana) :X. X2 on lR4 be defined by x1= - a + 21-. there is not. a i) No. f3) = (11x112.y E lR3} and define a map F : R6-+ IR3 by F ( x .X2(P) E T. is orientable. for example. Y ) = 01.O.f 2 . ( X . Xl(P).we see that is also a three-dimensional submanifold of lR4.X2] = .(23 2 4 ) .O . & + 3321 Let M = { ( X . The reason is that the bracket [Xl.M2? ii) Is there a nonconstant function f in the neighborhood of 0 E lR4 such that X1f E 0 and X2 f O? (Indiana) Solution. ii) Yes. we can set f = 21x2 . It is easy to verify that the Jacobi matrix of the C map F " . Noting that E4\{(0. a axl a ax4 i) Is there a 2-dimensional submanifold M 2 of lR4 such that for each p E M2.Y E Solution. O. i) Identify lR6 with {(x. (x. llYll = 1.asJ does not satisfy the Frobenius condition. a ax2 ax3 x2=-+22--.21 1 is a regular submanifold of R4\{(0.y) : x. is a smooth compact embedded submanifold of IR6 and be identified with the unit tangent bundle of 5 ' . Y ) i) Show that M explain how M can ii) Show that M B3. 0)) is an open submanifold of B 4 .IIXII = 1.

the Hopf-finow theorem means that N is complete. if we regard z E S2 and y E T.& x $ is . a t z E from 4 to the unit tangent vector y E T . Therefore. E M . If y is a limit point of F ( M ) in N . ) au. M is orientable. where B’(6) = {v E . z has the local coordinates (ui.212 has constant rank three when fl = 11z1(2 1.ui) U. F ( M ) is also closed in N . and the image of its end point under F must be y. x I. for every (z.y) M . then we can identify M with the unit tangent bundle of S 2 . Hence.E . f2 = ( ( ~ ( =~ 1 and f3 = (z. y) E ) with E being a suitable positive real number and 4 is the map from (2.O) = M is a closed embedded submanifold of B6. By using identification of M with the unit tangent bundle of S 2 . 1. y) E M .. Because F is a local isometry.. Therefore.y)II = (llzllz Ilyl12)i = fi which means M is bounded in IR6..y) = = 1 0. take S > 0 so small that exp. x I. F-’(l. since II(z. Besides. for every (z. then there is a point z E M such that there exists a geodesic in N connecting F ( z ) and y.. ( S 2 ) . where I. . ( U . 0. 3322 Let F : M 4 N be a local isometry between connected Riemannian manifolds M and N ..(S2) with E 1)yI)= 1.) neighborhoods. Thus. we show that F is a covering map. and y is uniquely determined by the oriented angle 0. = (0. so is N and F is a covering map. the connectedness of N implies that F ( M ) = N . For every x’ E N . a coordinate neighborhood of (2. Thus.) n (Up x I p .dol x 4) form a covering of coherently oriented . has the following J acobian + + which means that { ( V a x I. If (U. too. 4 x 4. . Besides. S2 has a covering { ( U .. #Jp x $ p ) # 0. F is surjective. i.. (Indiana) Solution. . Naturally. Show that if M is complete. The local isometry of F and the completeness of M imply that the above geodesic can be uniquely lifted to a geodesic in M starting from z. y) to 0. : B’(S) -+ Bi is a diffeomorphism. since F maps every geodesic of M into a geodesic of N . # J ~ ) }of coherently oriented coordinate neighborhoods. M must be compact.e. then the transition function . ii) Because S 2 is orientable. F ( M ) is open in N . Next.

Let y be the unique geodesic in Bi connecting F ( z ) and .} c M and set Ba(S) = {v E T. f ( N ) : IvI < S} and Bi = {y’ E N : d ( y ’ .) 0 a is obvious. F-l(B. So.8.213 T . Because both F : B.. if there is z E B. Since F is a local isometry. i) Show that each component of X = {z E M : F ( z ) = z } is an embedded totally geodesic submanifold of M .). i. 2’. which . + Let F : M -+ M be an isometry of a Riemannian manifold M . exp. expXm also a diffeomorphism.l ( B i ) = U B ? . we claim that if Q # .” n B f .1] -+ M such that F ( T ( t ) ) = y ( t ) . are diffeomorphisms. Since F is a local isometry. z zp respectively. Hence F = exp.1] B. Hence F(y(1)) = z’ and y(1) = z for some a .-((M) lvl < 6) and Br = {y E M : d ( y . The uniqueness implies ya = yp.) c U B r . we have the following commutative diagram and expZe is surjective. . In other words. Firstly. z ’ ) < 5). Besides. Thirdly. In particular. o d F o (exp. z a ) < a}. V t . if z E F-’(B.. z = y a ( l ) = yp(1) = zcp. we claim : that Bi is an admissible neighborhood of z’ and F is a covering map. contradicts a # p. uBr c F-l(B. Therefore. . we have F(y. then there exist unique geodesics y. then 0 there is a unique geodesic y : [0. Secondly. then BF n Bf = 8. M is complete.. ya. are isometric. such that y(0) = F ( z ) and y(1) = 2’. we claim that F . we know that d F ..)-’ is a diffeomorphism. there exists a geodesic y : [0.“ -+ B. F o expZm = expzf odF is a diffeomorphism and hence exp.) = y = F ( y p ) . is is an immersion. In fact. F-l(z’) is discrete.e. and F : B p 6 B. Then. yp connecting z with . Thus the claim is proved. we say that for any a .yp are the lifts of y through z. Besides.. Denote F-l(z’) = {z. Otherwise. F : BO -+ Bi is a diffeomorphism. Since . L ( y ) = L(y) < S means that z = T ( 0 ) E B r . On the + other hand.

F ( z . d F ( ( ( s 0 ) )= ((so) and F is an isometry.(VnB(6)). lies in X . z ) = (z. Thus. d F ( j ( 0 ) ) = j(O). = 0). = y..e. and F be a reflection with respect to the plane z = 0. and F is an isometry. = 01 u {(z.(tv) connecting z and y. b ) -+ X parameterized by arclength is also a geodesic of M . y.. i.e. d F ( v ) = v. Let y : [0. y is a geodesic of M . Therefore. F is an isometry.e.214 Hint. Then. Use exponential coordinates. F ( y ) is also a shortest geodesic connecting F ( z ) = z and F ( y ) = y. Since F ( < ( s o ) ) = (‘(SO). In particular. Then M is a 2-dimensional manifold.On the other hand.y.. and x = {(2. ii) Let < < < < M = {(qy. Besides. b ) . Let the geodesic y : [0.. which means that y E X n Bg. we first assume that y = X n & and v E B(6) such that exp.i. In particular. ii) Give an example in which the components of X have different dimensions. X n B c exp.z ) . F ( y ) = F(y(1)) = y(1) = y. In order to prove the claim. ( M ) : IvI < 6) and B = {y E M : d ( z . For z E X c M . is naturally a geodesic of X. Since z. v E V which means that y E exp. ( M ) : d F ( v ) = v}. V is a subspace of T. If this is proved. we show that every geodesic y : ( a . we show that X has submanifold structures.(M).z> E IR3 : 2 = 0. let <(s) be a geodesic of M such that <(so) = SO). because exp. .i. we can assert that X has submanifold structures. (Indiana) Solution. then F ( < ) and are two geodesics of M which satisfy the same initial conditions. Hence.(V n B(6)). Thus.e.(VnB(G)) is obviously a submanifold of M .1] 4 M be the unique shortest geodesic y(t) = exp. For any SO E ( a .(tv). : B(6) -+ B is a diffeomorphism. z = 0) u {(z.y E X . Then we claim that XnBa = exp. v.v = y.1] 4 M be defined by y ( t ) = exp.z) ~ E 1123 : =o . where a 6 is so small that exp. y ) < 6}. Next. i) First.y. in a neighborhood of S O . Define V = {v E g T . ~ .(V n B ( 6 ) )c X n Ba. Because so is arbitrarily chosen. X is totally geodesic.y < 0}. set B(6) = {v E T . suppose that v E V n B ( 6 ) and g y = exp. From d F ( v ) = v follows d F ( T ( 0 ) )= j ( 0 ) . F ( < ) = <. namely.z) E ~3 : > o . exp.y. Thus the uniqueness implies F ( y ) = y.. Therefore. that F is an isometry implies F(y) = y.) E IR3 : y > 0 . i. y . <~ o . i(so) = ? ( S O ) .(V n B(6)).

Le.z1 < E for some z E X } . I ' For H k ( S 1 . := {z I/ u. Let 6 be the polar coordinate characterizing S1. E ) := {y E En: Iy .d) = {f E Coo(S1. w . without citing the de Rham Theorem.x + :2 . H1(S1. d ) = Zo(S'. Besides. define a function & Because R(0) = R(27r). (Indiana) Solution. set and E x. Do so directly.d)= 0 and S' is connected. since there are no non-vanishing k-forms on S'. then H * ( S ' . 3325 Let X denote a submanifold of Euclidean space E n .d) = {df I f E Coo(S1.. because 6 in usual sense is not a globally well-defined function on S'). For every w = g(O)d6 E C"(S'.215 3324 Compute the de Rham cohomology groups of the circle S'.d) = Z'(S'.R1)}.Let d6 be its dual non-vanishing 1-form on S'(Caution: Here d6 is only a formal symbol.d) = Cm(S1. d )= 0.d) = (Cd6 I C E Rl}Z R1. B1(S1. since Bo(S1. B ( X . n is globally well-defined on S1. i. k> 1.e.Cd6 = do. Hence. Therefore. denoting we see that w .A'(S1)). and it is not exact.R1)df = 0) E B . observe that Z'(S'.CdB is exact. we d) have Zk(S1. d ) .d)/B'(S'. IvI < E } . X . = 0 and hence H k ( S 1 ..v E N . Then is a non-vanishing vector field on S1. A'(S')).

e. i. e. ) Can you give conditions which imply equality? (Indiana) Solution. and 2 = 2 ( z . Here L z denotes the Lie derivative along z. t ):= 2 ( z . L Z Y ) = + (**) for all vector fields X and Y on M . t ) satisfies . t ) such that m i. If X has no boundary.216 Show that U. E ) . j ( z ) = g (&. z m . Y ) g ( X . E ) . - a . Call a vector field 2 on M a killing vector field if 2 generates a 1-parameter group of isometries of M . we have Lzg = 0 . In local coordinates (zl.l=l where g . then U. can show that V E X is a proper subset of B ( X .E l . Show that the two are not generally equal. Let X be an open line segment in E 2 . ii) Show that the expression (**) above is equivalent to where V denotes the Levi-Civita connection for ( M .g.y > > S(JhX. i) Show. (Indiana) Solution. . E )for all E . let the 1-parameter group of isornetries generated by a vector field 2 = C zi& i=l m be expressed by 2 = 2(z1.j=l m k. Then considering the boundary of X .X = + B ( X . Z ( g ( X . (Consider examples of 1-dimensional submanifold in E 2 . g ) .X c B ( X .. Setting y = z v immediately implies that V E Xc B ( X . the end points. &).. z m ) of ( M . 3326 Consider a Riemannian manifold ( M .that when 2 is Killing.9 ) .. either compact or complete.g ) . i.e.

(Indiana) Solution.217 Differentiating the obtained equality with respect to t and then setting t = 0. we obtain which can be written as k gkjz. + . Because the flow X t of X for every t is an isometry. + 3327 Let M 2 be a connected Riemannian manifold and X .Y)) g ( V z X .i + S i k zkj $9 ( -l . i) For every p E M .i V 0. From this follows what we desire in ii). or equivalently g (V a Z . or equivalently. we may assume that X = X I & and g 1 2 = (&. i. Conclude that ( X . Y ) g ( X . -l) j e er3 Z ) =o. take a local coordinate neighborhood about p such that the coordinate curves are the integral curves of X and their orthogonal trajectories. Noting that the Levi-Civita connection V satisfies Z ( g ( X . Y ) and ( Y . the vector field X should satisfy the Killing equation Xi.o X t . 2 . i) Show that the integral curves of X are curves of constant geodesic curvature. Assume that the flows Xtand Yt are isometries of M for all t. Y )are constant on M . ( X . Namely. j = 1 . ii) Assume that X and Y are linearly independent at all 2 E M 2and that their flows commute X t o Y. Y complete vector fields on M .i = 0. X I .j Xj. we easily obtain (w). = Y. &)= 0. V Z Y ) = and the Lie derivative satisfies L z X = [Z. X ) .

X = Y = G. i.Y 2 . 2 . we have locally X = X I & . we can make a suitable coordinate transformation and then.. =o -8x1 dxi . respectively. Ic.e. +.. that is equivalent to lay2 d ax1 d [ X .X ) = gll = const. dX1 &. we obtain 911% following coordinate transformation = 0. . which means that the metric of M 2about p can be written as ds2 = g11(x2)(dx1)2 &. (Y. o X t . along every integral curve of X . Then. ak = 0.Y ) = g12 = const. Then. Noting the expressions of X and Y . i.8x1 a x 2 dx2 dXl = O . ( X . Now. we can obtain = 0. then the vector field X = from the corresponding killing equation. take the integral curves of X and Y as the x1 and x2 coordinate curves. = const.Y ) = 922 = const. Y ]= -. if a adopting the original notations. we can compute the geodesic curvature of the xl-coordinate curve as follows: where 0 = 0. j = 2. we obtain -d=Yol2.g 2 2 ( ~ ~ ) ( d ~ ~ ) ~ . X1 = X1(xl). = Y. j . we obtain a922 -agll.. Again. ii) Analogously.e. using the Liouville formula. about p . i. all gij's are constants. Because their flows are commutative X t o Y.. Therefore.e.0 . we immediately have x ax2 Therefore.218 Taking i = 1. taking i = j = 1 and i = j = 2. ax -= 0. Ic = 1 . make the Hence If we still adopt the original notations. by the corresponding Killing equations. Y = Y2&. i. a ( X .

( v e .n. i .e j ) = (e: . } around p and then define (Vf)(p) = C ( e i f ) ( e i ) and (Af>(p> . they do not depend on the choice of orthonormal frame) and then show that (Indiana) Solution.e. e i ) f l = C [ e r ( e f f ) ( v e . . . choose an orthonormal frame field { e l . . j for suitable functions a!.e3) = S i j . . e i } be another orthonormal frame field around p . Then Green’s theorem implies .. through direct claculation.f A f. For any f E Cm(M).219 3328 Let M be a compact Riemannian manifold without boundary. e : ) .1n - e . may suppose that e i = C a ! e r i = l . . .. define O f E X ( M ) and Af E C “ ( M ) as follows: At any p E M . f ] . Then we Let { e . i i Hence Of and Af are all well defined. j = 1. .. Noting (ei .C [ e i ( e i f )..( v e i e i ) f I = i i Verify first that V f and A f are well defined (i. we can easily obtain i i C [ e i ( e i f ). Using the definition of the Laplacian here. . . we have j Using these equalities and the properties of Riemannian connection. . we have div( f V f ) = llVf112 . e .

b2 . then 21 = 2 2 = 2 3 = 2 4 = 0. = 2. 1.b = ( 2 E E4I b + 2. x : + 2 : = b or 2.b a manifold? Explain. we can show that they are all 2-dimensional submanifolds of R4. Using the theorem of closed regular submanifolds proved by rank theorem. b # 0. 3.f3 are two roots of the equation X 2 . then when b2 .b is a 1-dimensional submanifold of R 4 .o is a 0-dimensional manifold. b # 0. = 2}. If a = 0. Thus S0.220 3320 For what a .f3 = x i + x z .4a = 0 Sa.2 : = b . + 2. 3330 " " Let F : M -+ N be a C map between two C manifolds. Assume that F is onto. Denote a = xf + x i and . Then C Y ~ a and CY + p = b = means that a and . i . is Sa. If a # 0. x 3 = 2 4 = 0). . Let X be a smooth vector field on M . we can easily show that &. then = { Z E B4 1 x1 = 2 2 = 0 . b 2 0. 2. Therefore. (i) Show by an example that d F ( X ) may not be a vector field on N . Analogously.bX + a = 0. (Indiana) Solution.4a 2 0 is the prerequisite condition.4a > 0 and when b2 . If a = b = 0 .

if y1 # yz. # then.y E IR}. Y are related as X and Y in (ii) above. 3331 Let M" be a Riemannian manifold.. y(t) has the following expression - 21 1 t.N = {x : x E E}. (Indiana) Solution.221 (ii) Suppose Y = d F ( X ) is a smooth vector field on N . 1 and X z . (ii) Let a(. X Z l )= [Yl. one will obtain the desired equality.). d(F o a) (i) = ( d F o da) (g) = d F ( X ) = Y. along a . Show that d F ( [ X l .y) : z. (i) Let M = {(x. x and F : M + N be defined by F(z. XJ 1 x 3 0 (.) be an arbitrary integral curve of X . . .. Let (xl. z (iii) Suppose XI.x n ) be the local coordinates about p and set p : (xk. = (xz + y 2 ) & . Y. x2.yz)). we have da ($) = X .e.. by direct computation. If y ( t ) is the i-th coordinate curve that passes p . Then. d F ( X ) is not well defined in every point of N . Then. there is a unique vector field V f (called the gradient of f on M ) such that whenever y is a smooth curve in M with y(t) = p . Show that whenever f : M + lR is a smooth function.i 4. .y) = x. (Indiana) Solution. (iii) Using local coordinates. Show that F takes integral curves of X into integral curves of Y.. i.YZ]. by . Therefore. as a vector field.y l ) ) # d F ( X ( z . Hence. we have d F ( X ( x . which means that F takes integral curves of X into integral curves of Y .

. Then C z = 1. which means that u(x)is on the sphere of radius 2 centered at the origin in 45 nnxn -+ ii) Let i : S"-I -+ Rnbe the standard inclusion map. ii) Prove that cr is a local isometry (i.and define u : S n w 1 E n x n the map -+ as sending x = (21. 2.B ) := i=l j=1 a. .j b i j .) in S"-l to the symmetric matrix ~ ( x= I [ z i z j ] .u(x))= z. z i.. is (Indiana) Solution. ) Jz i) Show that u maps S"-' into the sphere of radius centered a t the origin in R" n. a .e. Let Sn-' be the unit sphere in IR". the pull-back via g of the dotproduct metric defined above on RnX" the standard one on S"-'). It is easy to verify that the above expression of V f (p)is independent of the choice of local coordinates. 5 i=1 1 " (u(x).. 3332 The space ELnxn of n x n real matrices forms an n2 dimensional Euclidean space. Hence .-. .222 we have the expression where gij's are the components of the matrix [gijl-' = [(&. in which the dot product between A = [a.. . n i) Let x = (zl. and r : an RnX" be defined by (21..zn) E S"-l.) -+ [yij] = -&[zizj].z.j]and B = [ b i j ] is given by n n ( A .j=1 c = 1 5. . Suppose that .

e2 (Indiana) .dn(w))= - 1 . 2 . where f : IR" IR is a smooth function. a metric of the form [sij] = e 2 f [ S i j ] . k E { 1 . c n i. i. n} = & ---f ii) Show that when n = 2.Then we have n n i=l i= 1 Therefore Hence we have (dn(v). the sectional curvature of M along an plane a t p is given by e1. Let ei = denote the standard coordinate basis vector fields.w).223 belong to Tx(Sn-').e. i) Show that for arbitrary indices i. and S i j is the Kronecker delta. Let M be the Riemannian manifold obtained by equipping B" with a metric conformal t o its usual one. j .j=1 (dXj + VjXi)(WiXj +WjXi) = 3333 c n i=l viwi = (v.. .

e. i) Show that X .t E IR implies [ X . . for all s. = yt o X.. Thus 1 [ X . i) Suppose that X. Y be complete vector fields on a manifold M and let X . W E R. Y ]= 0.Y]= 0. F = 0. Y ]= L X Y = lim -[Y .Since X ..-invariant. ii) Prove the converse to i). Yt be the flows induced by them. then Y is X. F. t E R. is a diffeomorphism for each s. o yt = yt o X . 3 4 0 s ii) Now suppose that [ X .Y = Y . i. o r. from the Gauss equation we obtain 3334 Let X . the complete vector field Y is F-invariant. Observe that for a diffeomorphism F : M -+M . e2 plane at p is just the Gauss curvature of M a t p .x.Y ]= 0. for all s. (Indiana) Solution. Noting that E = G = e 2 f . i) Set From gij = e 2 f & it follows that ii) The sectional curvature of M along an e l . if and only if F o yt = yt o F . Then .224 Solution.

Note that exptw. 4 can be expressed by and exp tv and exp tw can be denoted by respectively. X. . we obtain X . yn) be local coordinates of G1 and G2 about " the identities el and e2.z ) and (y'. 4(exp tv) = exp tw. . Y = Y . . . namely. we have (Indiana) Solution. we have for all s E R. ( X s * Y ) p f ( X o * Y ) p f= Y p f . Y is X. Let ( z l . Since 4 is a Lie group homomorphism. . = yt O X . respectively. namely. = Since f is arbitrary. Locally. t E l is a 1-parameter subgroup of G1 generated by R ' v E LG1. show that for all w E LG1. i=l axi we2 = a E w e -aye ' n e=1 . ox Hence 3335 If 4 : G1 -+ G2 is a Lie group homomorhism.for all s . 4(exp(u)) = exp(+*v).225 Hence for each p E M and any f E C"(p). then 4(exptv) is also a 1-parameter subgroup of G2 generated by a suitable w E LG2. . Assuming that wel " a =cvi-. Therefore.-invariant. . t E R.

22. -24. 3336 Consider the linearly independent vector fields r and v on whose values at x = ( 2 1 ..w. 2 3 .21..23) v. by left translation.v] = 0. Thus the Frobenius theorem guarantees that the rank-2 distribution defined by r and v in U is completely integrable. y 3 and Y4 I Z l Y l + 22y2 -22y1+ + 23Y3 + 24Y4 = 0. 23. = 2 (Z) (= I -=we. (-22. Taking t = 1 completes the proof. 2 4 ) E B4 are given by rx : U := B4\0 = = (21. ZlYZ . we have d $ ( v ) = w. 2 2 . i.24y3 + 23Y4 = 0. which are equivalent to . y2.226 we have t=O t=o Therefore.: a) Is the rank-2 distribution defined by these two vector fields in U completely integrable? b) Is the rank-2 distribution orthogonal to these two vector fields completely integrable? (Indiana) Solution. a 1 t=O aY" Furthermore.e. b) Construct the following linear algebraic equations about y 1 .24). d*w = w. Hence 4(exptw) = expt4. a) Direct calculation gives [r.

Thus the F’robenius theorem tells us that this distribution is not completely integrable.s+. 4 ) E B 4 \ 0 = U. ( 7 )) by setting.227 Because ( 2 1 . . By a long but straightforward calculation. 2 32. j = 1. + -(. by the vector fields Veiej. . and the frame-field is left-invariant. a) Show that any connection on G is competely determined by its effect on the frame field. 1 Veiej = . (2122 2223-22124.j = 1. we characterize the Levi-Civita connection on (G.without loss of generality. we obtain the following two linearly independent vector fields CY and p which define a rank-2 distribution orthogonal t o the above one and whose values at are given by = (2123+2224.. Therefore. ~ = -2(2: ] zi)v which does not belong to the distribution defined by CY and p.:). a) For arbitrary smooth vector fields X and Y on G. we obtain 2123 2223 +2224 . 2123. -k 2. (Indiana) Solution. ).n. .i. e j ] 2 for all i.2223 Setting and respectively. ~ 2 . o).[ e i .2223.2124 2124 2224 + 2123 . 2224 p = . + 3337 Let G be a Riemannian manifold with a global frame-field {ei}y=l. i. ( : . we may assume 2 1 # 0.n.)). we have [ c Y .. 0.we have - n n .e. b) Show that when G is a Lie group with a bi-invariant metric ( .

. e j ] .ej . 2 1 ..[ e i . we can well define x'e. we know that V x X = 0 is valid everywhere. if X = ei e j .228 Then. Thus.j=l n Thus defined operator V certainly satisfies all properties of a linear connection on G. = 0. Noting that g ( t ) is a geodesic of G and Xg(tl = we have obviously 9 y. e j ] . + + + On the other hand.(g)-ej + C zigveiej. let g ( t ) denote the unique 1-parameter subgroup of G such that g(0) = e .V e j e i= [ e i . i. because the metric of G is bi-invariant. then V e i + e j ( e i e j ) = 0 implies that Veiej Veje. Especially. where e is the unit element of G. vx. b) For every left invariant vector field X on G. we obtain V e i e j= . motivated by the properties of connection.x = - t=O Besides. the Levi-Civita connection V satisfies V. = X .

Part IV Real Analysis .

.

1 kn-1 lo*-1 + 4 1uR) and B = C u [$+. V i < n . Prove that S is Lebesgue measurable and find its measure. E LO. .1] be the set defined by the property that z E S if and only if in the decimal representation of z the first appearance of the digit 2 precedes the first appearance of the digit 3..+- kn-1 n = l kj#2. 10n 10 +. l ) Then 00 1 3n. kn-1 n=lki#2..3 10-1 + -.3 10n-1 kl + -. V i < n . For any z E [0.2 . p i ( z ) # 2 and pi(z) # 3).+ . * p s Let = (1) U{z u{ .k . (2) Vn. p n ( z ) # 2 and p n ( z ) # 3) and B = {z E [ O .) # 3). 10" l o +- 3 It follows that both A and B are measurable and therefore is S..-.1) there is a unique sequence {pn(z))r'l of integers satisfying the following properties (1) 0 Then 5 p n ( z ) 5 9. ( z ) 58 and (3) z = n=l . p i ( z ) # 2 and pi(. Since c y ) n . p n ( z ) = 2 . A = {z E [O.231 SECTION 1 MEASURABLITY AND MEASURE 4101 Let S c [0. 2 kl . 1) I h P n ( z ) # 2 and P() n .1) I 3 n . 3rn 2 n : p . # 3) E [0.+ .pn(z)= 2 . (Stanford) Solution. 1) I V n .

+ If T . - E = {F Show that E is not measurable. for z. S1is an abelian group under the binary operation ([z].1) n& are such that ( E [TI) n ( E [ s ] ) # 0 there are [z]. Thus [r] = [ s ] . If E is measurable then + + + T €[O. For any [z] E S1there is by the construction of E a unique element [y] E E such that z .1)W 00. [y]) H [z + y ] and the inverse operation. So we conclude that S1= U ( E + [TI). [y] E E y mod& and therefore such that [z] [T] = [y] [ s ] .l)W + m(S1) = T€ C [O. each ( E S1/ choose a representative f E S1. a contradiction.1) n& is such that r G z .y E IR and [ ] denoting the class in S1. Consider on S1the equivalence relation: [z] [y] # z . ( StanfoTd) Solution. 1) n&} is a partition of S1. [z] +-+[-. 1) m(E+[r]) = C T m(E). Otherwise m ( S 1 ) = contradiction.s < 1. + 2 4102 Define S1= IR/Z endowed with the natural Lebesgue measure.V[ E S'/ -}. s E [0. i. . I E <. Let m denote the natural Lebesgue measure on S1such that m(S1) = 1.y m o d Z .I. For -.232 and we have 1 m ( S )= m(A) m ( B ) = -. a w €P.y E &. It follows that { E + [TI 1 T E [0. too.e. and let E c S1be the set of these points. It follows that z [z] = [y] by the construction of E . Then [z] = [y] [r].where T E [0.y E &. If m ( E ) = 0 then m(S1) = 0.. which then implies that T = s since -1 < r .

B subsets of I such R that inf{ 1 . (ii) differences A\B. 2 Prove or disprove that p * ( AU B ) = p * ( A ) +p*(B). Let U = {Z E I d ( z .yI I z E A . Obviously. 4104 Let p* be the Lebesgue outer measure on IR and A . y E B } > 0. Let T = d ( A . F'uthermore. We have since the former is a ring containing D. By equality (2). We will show the equality. (Iowa) Solution.'D closed under finite intersection.233 4103 Let X be a set and D c ' P ( X ) . R = R since for any A E R. A )< T) 2 . D c a c P ( X ) such that R is closed under the following operations: (i) finite disjoint unions. x C R.A = A n A. Also we have since by the equality (1) the former is a ring containing 2). (Iowa) Solution. B c A.B ) > 0. let a be the smallest system. Denote ' by R the ring generated by D. Prove that a = R.

p ( B ) = sup{p(K) I K 5 B and K is compact}. that B is the Bore1 a-algebra.} be countably many measurable sets which are 00 mutually disjoint. Let {B. (Iowa) Solution. A C WI C U . inf(X+Y)=infX+infY.A U B inf(p(W1) C W} : inf{p(W) I W o p e n . is a decreasing sequence of sets with empty intersection then lim p(An) = 0. positive. Let A.234 Then U and V axe disjoint open sets containing A and B respectively. finitely additive measure on B. A U B C W C U U V ) E = + p ( W 2 ) I W open. Suppose moreover that p is regular. Finite additivity means that whenever { B i } is a finite collection of mutually disjoint measurable sets. Prove that p is countably additive if and only if it satisfies the following condition: If A. B C W2 C V} P * ( A )+ P * ( B ) . = i=n+l U Bi. that is for each B E B. p ) with a finite. 4105 a) Consider a measurable space (X. and that p is a finite. B C W2 2 V } i inf(p(W1) 1 open. finitely additive measure p. Then n=l n An = 0. X. a) The sufficiency. Prove that p is countably additive. then p(UBi) = C p ( B i ) . We have \ o o m / n . We have p * ( Au B ) = = = inf{p(W) I W open. T-00 b) Now that suppose X is a locally compact Hausdorff space.Y where the equality (1) follows from the following equality GR. A 5 cU } ~1 WI +inf{p(W2) I W2 open. positive.

which contradicts the fact that n=l 4106 For f : [0. n .1] -+ B. ( Indiana-Purdue) Solution.21 < -1. 12 E E . So n K. n For each n there is a compact Kn contained in A . such that which implies that and therefore i=l n Thus { i=l Ki I n E N}is a decreasing sequence of nonempty compact subsets n=l in the compact space K1. If(y) .--too lim p(A.f(z>II MIY. b) If /I is not countably additive. 03 # 0.235 Therefore Y is a measure.} of measurable sets with empty intersection such that .) = inf p ( A n ) > 0. The necessity is obvious. lf'(4I < MI. by a) there is a decreasing sequence {A. show that m(f(E)) = 0. Set Fn = I z E F . . 1 where M is any positive number. If m ( E ) = 0. Denote F ={ .21 if I Y . let E C { z I f'(z) exists}. It suffices to prove m ( f ( F ) )= 0.

There is an A n (n.236 Then F1 C F2 C * . lim For any E > 0.n + 1) C U C (n. take a sequence {In$} of open intervals such that For 2 . Prove that m ( A )= 0. and f(F)c uf(Fn) = n-cu f(Fn). b E IR.a < b. = 4107 Suppose A c IR is Lebesgue measurable and assume that for any a. . which implies that m ( f ( F ) )= 0. k ) ) m ( I n . k ) < ME- 5 M Thus m*(f(Fn))0.y E Fn f l I n $ .n open subset U in ( n .n + 1) such that + 1)) # 0. If m ( A ) # 0 there is an n such that m ( A n (n. we have Therefore 5 k m * ( f ( F nf' I n . (Iowa) Solution. + 1) n . .m ( F n ) = 0.

+ 1))< E. the total length of intervals removed in the n-th step is X ( l The claim holds for n = 1. etc. We are left with the set K x . Then the total length of intervals removed in the k 1-th step is + k i=l . There are at most countably many disjoint intervals ( u j . Assume that it holds for any n 5 k . bj)’s such that + Then A n ( n . For any n E W . and keep doing this ad infinitum.1] its middle part of length A. (Stanford) Solution. 1 Remove from each of them their middle parts of lengths X l l i j l . j We have which deduces that m ( A n (n.n 1)). n + 1) = U A n ( a j . Prove that the set Kx has Lebesgue measure zero.n a contradiction. Claim.237 where E < m(An (n. we are left with two intervals 1 and 12. b j ) . 4108 Choose 0 < X < 1 and construct the Cantor set Kx as follows: Remove from [0.

238 By induction the claim holds for any n E JV. (Iowa) Solution.O]) n+oo n 1 = 1 . from either A 7 . I. However. y E B. n+oo n and therefore p is the Lebesgue measure. It follows that g ( z ) 2 z . Y L0 <0 + Y. y ] = g(y) . Moreover. p ( z . y E IR with z < y. for any R z . we conclude that g ( z ) = zg(1). From the right-continuity of g. Does this imply that p is the Lebesgue measure? Justify your answer. 01. g(1) = p((0. Y we have g(z + Y) = g ( z ) + d Y ) . = P(Y. For any z E IR define + Then g : lR -+ l is nondecreasing and right-continuous. Y1. 11) .1 = 0.g(--)) = 1. p is the Lebesgue measure.P ( { W 1 = 1 . It follows that the Lebesgue measure of Kx is 1- c n=l X ( 1 . 4109 Suppose p is a positive Borel measure on lR such that (i) p([O.A y .lim (g(O) . or p(z z + Y1 = 4 0 . Yes. It follows that the measure p is induced by g.g(z).11) = P([O. and (ii) p ( E ) = p ( z E) for any Borel set E of 2R and every z E B. For any z .lim p((--.l]) = 1.

G. ndoo is a signed measure. 4111 Let X be Lebesgue measure on R.239 4110 Let U be a a-algebra of subsets of a set X and p n : U -+ lR be signed measures such that p ( E ) = lim p n ( E ) exists for every E E U. there is a Lebesgue measurable set A c E with X(A) = !. is a finite measure and pn’s are absolutely continuous with respect to Z . by the Vitali-Hahn-Saks Then k=l which shows that p is a generalized measure and therefore a signed measure. Prove that u .Show that for any Lebesgue measurable set E c 1R with X(E) = 1. Let. Given any mutually disjoint measurable sets En’s. (Iowa) Solution. for each E E U . Then . j (Iowa) .

4113 Let A c [O. co]. there is a point E E such that = f. c [-I. = 0. 12 E A}. Put A = E n (-co. A . 21. Show that there 1 exist two points z/ # z” in A with 2’ . Define the function f : IR f(2) = -+[0. Show that if r w for all n = 0 . . Denote all the rational numbers in [-1. 1 be a measurable set of positive measure. Z. + T. 20 lim f(z) = 0 and lim f(z) = 1. Thus .7 3 . Let S = span{ePa5 I n E lN u (0)) then S is a self-adjoint subalgebra of C([0.. co). . then .z” rational. .zo].) Since f(z0) x+-w . . ( Indiana-Purdue) Solution.z]). .. . . which then implies that p = 0.Y E E . . 2 E E. Denote A. It is continuous by the following inequality If(. S is dense in C([O. . Therefore for any f E C([O.240 Solution. co])separating the points of [0.m]) under the supremum norm topology.f(Y)I x+ w 5 1 2 .) = m(A) > 0. 1 by 1 X(E n (-m.. 2 .oo]).1] by rl. It follows from the Stone-Weierstrass Theorem. 4112 Let p be a complex Bore1 measure on [0.T. u (Iowa) Solution.YI. which is required. = {z Then m(A. . 1 .

E is measurable and since + ["3 + 2. 5 zl.. We will show next that * * - [O.T.l)\E. ... 3 31 - 3n-1 xn-l 7 = 2. n=l 00 Suppose that A. Thus - = 'I'm .= 2) then 0 1 a:€ I l).+1. Therefore there must be some m. Let n = min{k I 2 . Then C m(An) I m([-1. By equality ( l ) . Prove that E is Lebesgue measurable with Lebesgue measure 0. Also show that E + E = [0. (Iowa) Solution. n such that A n n Am # 0.1). for any 2 E [O. Take 2 E An n Am. 2 = n=l ? $ (where 0 I z. 4114 2.xn-lI 1 and therefore 2 21 + .21) n=l 00 = 3. I 2 and for any n there is an m 2 n such that 2 .-1 2 2 + -+ -3" . which contradicts m ( A ) > 0. n Am = 0 if n # m... =2 I1 + T.. .. = 0 or 1 . E c [0. Obviously.-+. there is at least one n such that 2.+ . 2 E A such that ' ' z =2 ' + r..1 .. 3"' 3 [? 2.1].+ + * a + - 3"-1 The converse inclusion is obvious. Then we c a n find z'..241 U An c [-1721. 1)\E = u { 00 Indeed.

(where 0 Define i]. < T. Take {r. [If I{. {zIg(z)= r } = {z [If gllco = .. Since E C [0.gllm 5 E (ii) for every T E B. Show that there exists g : I22 -+ El+ measurable such that (i) . and z . there is an m 2 n such that # 2). I g ( 4 = .sly. - Set g(z) = gl(z) . 1 gl(z). .}I = 0. 13. . (Indiana-Purdue) Solution. I 2. CF n= 1 03 5 z . = 2. L E. n= 1 This completes the proof. Set fl(z)= f z + arcctgz.gl)loo 5 E .glllcx. T. ~'$2'' Then z' = C 0 0 3 and . n E N. = Since \ ( r c U{zI z E Enlarcctgz = r.E + E [0. 1z . .1) say z = 2 . () ..+~ r . I r n . Denote () En = { z Set g1 = TnXE.T } \ = 0. for all n. llfl Then g ( )2a c c t g z . 4115 Let f : It2 -+ IR+ be measurable. limr. .=l 00 X" $ belong to E so that = z. n \ arcctgrc = T . For any z E [0. 2 ' ' = C . Obvior. \ { z g z = T } \ = 0.1 < fi(z) I ~n}. and let E > 0.242 m ( E )= 0. < . .} such that 0 < r1 < r2 < .arcctgz. We have g 2 0 and For every r E B. and for any n. = +m.arcctgz = T } TI. < E .

1]\K = p 2 1 ai=0.U p . . O . z++w + (Illinois) Solution. . .1] defined as follows f(z) = 0 if z is a point on the Cantor ternary set and f ( z ) = if 2 is in one of the complementary intervals of length 3 . Let K denote the Cantor ternary set.P . b) Evaluate 1 1 f(z)dz. a) Prove that f is measurable. we have which then is measurable. ) Show that (a) f ( z ) is a continuous function on R. * a p . and 4117 Let E be a Lebesgue measurable subset of lR with m ( E ) < 03 and let f ( ~= m ( ( ~2 ) n E ) . a 1 .1 1 .243 4116 Let f be the function on [0. h-0 lim m ( ( E+ h ) n E ) = m ( E ) . a 1 a 2 . We will show first that for any Lebesgue measurable set E . (Stanford) Solution.1 2 ) . Then [O. (b) lim f ( z ) = 0.2 uu (0 . . By the definition.

Let F = E’\E one has m(E ) < m( El) = lim m((E’ h-iO + h) n E’) = l i m m ( ( ( E + h ) n E ) u ( ( E + h ) n F ) U ( ( F + h ) nE ) G ( ( F + h ) n F ) ) h-0 G 5 h m ( ( E + h ) n E ) + 3.) = m ( E ) . then n 5 hm(En(E+h)) h+O <_ h+O lim m ( E n ( E + h ) ) 5 m ( E ) . If E is a compact set. so h-O lirn m ( ( E + h)nE ) = m(E). there is an increasing sequence { E n } of compact sets such that En C E and lirn m(E. an open set E’ of the above form such that E c E’ and m(E’\E) < E .@i)’s are finitely many mutually disjoint open intervals of finite length. h-+O It follows that m ( E ) = lim m ( ( E+ h ) n E ) .5 5 h+Om ( ( E +h ) n E ) + 3~ 5 m ( E )3 . n-+mh+O n+m lim m ( E n )= lim lirn m ( ( E n n+m h-10 h+O . ) o i )where ) r=l n (q. then there is. for any E > 0. n+m Then m(E) = 5 + h) n En) lim lim m ( ( E + h) n E ) = lim m ( ( E + h) n E ) . 3 ~ .G( a .244 If E is of the form . h-0 In general.

then for any n have n U ( z a i .245 so h-0 lim m ( ( E + h)nE ) = m(E).b i ) .1 < E .f(. n-cc lim m ( ( E z) X-++OO = = z-r+cc lim n-cc lim x++m + nE ) m ( ( E+ z) n E ) .bj)’s i<m u are mutually disjoint open intervals. (Iowa) Solution. z E 2R If(Y) .bi) n i=l <m and any z >0 we c ZB nB. (b) If E is compact.)I = Im(((E+ Y)\(E + + Y) ) l 5 m ( ( E+ Y)\(E + = m ( E (Y . If B is a bounded open set. The claim follows. Then + E ) n E = 0. In general there is an increasing sequence { E n } of compact sets such that En E and lim m ( E n )= m ( E ) .m ( ( E + z)\(E (a) For any y. the line y = mz has a non-void m 1 intersection with A x A . . Prove that cp(z) = J X & ) X A ( t ) d t is continuous a t z = 1.z)\E) m ( E (z .m((En+ n E n ) + m((En + n E n ) lim m ( ( ( E+ z) n E)\((En z) n E n ) ) + m( (En + z) n En) + Z) Z) 5 n-rn lim ( m ( ( E .y ) \ E ) = m ( E ).z)) n E ) m ( ~ ) + + + + M ( ( E + z)\(E + Y)) + + . Use this result to prove that there exists an E > 0 such that for any m E 2R with I .y)) n E ) -+ 0. say (a*. 4118 Let A c l be a set of positive R Lebesgue measure.m ( E (Y .)\(En + + z)) + m(E\En)) = 0.m ( ( E + (z . n E ) . where 2 5 m 5 03 and ( a j . then there is an T > 0 such that for any z > T . z b i ) (ai. (z as y -+ z.

) aJ = 0. By (2) we have Since . n n -.\K)) we have by (1) + P(Bn\K) In general. there is an increasing sequence {K.p ( z K n K ) I P(Z(B.} of compact sets such that U K . n Bn) .} of bounded open sets such that K = n=l n B. there is a decreasing sequence {B. 00 Since p ( z B .a) 5 ~ P ( z B ~ B ) 2-1 5 X’lim ~ l ( Z nB ) B 5~L(B). (1) where p is the Lebesgue measure. C A and p(A\ n=l 00 n= 1 u K.. If K is a compact set.246 We have p(B) = sup n<m a = 1 C(bi .

1 < and the line y = mnx has a void intersection with A x A . then [B]5 [ A ] . If there is a t o E (O. n-03 a contradiction. E B such that Imn .403 lim p ( A . ( A n A ) = 0. if [ B ] 5 [An] for some n..247 If there is no E > 0 with the property mentioned in the question. then 03 = . and therefore p ( A ) = lim p ( m n ( An A ) ) = 0. Then P is a partially ordered set: [-41 5 [B]if p(A\B) = 0. Let where is an equivalence relation: A B if p(A\B) = p(B\A) = 0. [B]= [A] since P(A\B) 5 C p(An\B) n=l 03 =0 .) Solution. Then P 5 t o and there is an increasing sequence { [ A . ] } of Q with { p ( A . then there is a measurable set B c A such that 0 < p ( B ) < p ( A ) . i.p(S)) not in the range of p.B = sup{p(A) I [A] E Q } . - u A n . For any [ B ] E Q. However. Let A = n=l - P = { A I A measurable and p ( A ) < t o } / -.e. then for 2 1 any n E mT there is an m. Given a totally ordered set Q of P .otherwise. 4119 Let p be a countably additive measure on a set S with p(S) < $00 that is without atoms. SO p ( A ) = /3 < t o and therefore [A] E 7'. p(S)]. if A is a measurable set with p ( A ) > 0. Prove that the range of p is the closed interval [0. (Courant Inst. ) = P 5 to. let . ) } increasing to p. this implies that m .

5 as above.otherwise [B]= [C]since n=l and ~ ( B \ C )= p ( B ) . by the assumption that there is a subset of COof Bo with 0 < ~ ( C O ) ( B o[CO] [Bo] [CO] [Bo]. It follows that [ B ]is a lower bound of S in R. <~ ). Show that any positive measure p on B which is finite a on compact sets is automatically regular. For any [Cl E S. [B]E R.p ( A ) and there is a decreasing sequence { [Eln]}of S with { p ( B . Therefore p ( B ) > t o . Given a = inf{p(B) I [B]E S}.248 It follows that [A] is an upper bound of Q in P.P(B) > O}/ -. Let n=l then So p ( B ) = a > 0. There is by Zorn’s Lemma a maximal element [Ao]of P.p ( A 0 ) ) .p ( B n C ) = -a = 0. p(S\A) > t o . contradiction. 5 and # a 4120 Let X be a compact Hausdorff space. It follows that p ( B ) > t o .p ( A 0 ) whenever B C S\Ao and p ( B ) > 0. (e.g. There is by Zorn’s Lemma a minimal element [Bo] R. ) } decreasing t o a. However. Let - R ={BI B c S\Ao. (Iowa) . if [Bn]I [C]for some n.p ( A 0 ) . then cr 2 t o . then [B]5 [C]. Consider the c-algebra 2 generated 3 by the compact G sets. Equip 72 with the partial order a totally ordered set S of R. Let where is defined as above.

each set in K is outer regular. If { E . there is a V in 0 such that X\U C V and p ( V ) < p(X\V) E . there are B1. Thus K\L is regular. M) + E. where the symbol 0 means disjoint union. Let a = { ¶=1 . Step 1. Step 2. E 0 such that E C Bi with i p(B.. 5 P(B\L) < P(K\L) p(K\L) < p ( K +E. regular Obviously For any E > 0. Also recall that E is outer regular if p ( E ) = inf{p(V) 1 E and E is inner regular if C V. For any E > 0 there are a B E 0 and an M E X such that K C B and p(B\K) < E and such that M C B\L and p(B\L) < p ( M ) E . E i = 1. 7 ~ . For any pair K . Then we have V). i= 1 n} is a finite class of mutually disjoint. C K\L. Then 7E is a ring such that 23 = By definition.. For any c > 0. . then U Ei is regular.\E. L in K .V open and V E 23) p ( E ) = sup{p(K) 1 K C E. K\L is regular. .K ? compact and K E B } . Let us preceed in five steps to show the regularity of p. Let us show each set U in 0 is inner regular. B. Consequently. Let K and 0 denote the classes of compact sets and open sets in 2 .249 Solution. respec3 tively. Recall that sets in 23 are called Baire sets and each compact Baire set K is a Ga set. . + + + K\L K nM c B\L. . .U (Ki\Li) I K i . I 1 5 i ? sets. .) < . Li E K}. 6 i=l Ei C ( Bi J i=l . . Then X\V 2 U and p ( U ) < p(X\V) E .

n=l n En is .. The inner regularity follows from n i= 1 Step 3. If { E n 03 In E IN} is an increasing sequence of regular sets then (J En is regular. . Let V = U1 Vn E 0. If { E n } is a decreasing sequence of regular sets. . En > 0 there are 03 V1..250 and which shows the outer regularity of the following inequalities u Ei. . then n= which shows the outer regularity of E . while the inner regularity follows from the following inequalities co Step 4. . E 0 such that C Vn and p(Vn) < p(En) + &. n=l Let E = u En. . then regular. Obviously 00 n=l p ( E ) I inf{p(V) I E For any E G V E O}. V.

S contains a. The outer regularity follows from the following inequalities. (Indianu) ..there is a K ..u(En) = p ( E n 5 V E 0) Step 5. By steps 1 and 2. E ?E such that n= 1 and which shows the inner regularity of E . By steps 3 and 4. 00 For any E > 0. n E lN. t monotone class. It 4121 Suppose that p is a nonnegative Bore1 measure on En. Let Assume f ( ~is finite for all ) T > 0 and assume Prove that p is identically 0. inf{p(V) I E V E 0} 5 inf inf{p(V) I En = inf. Let S = { E E 23 1 E is regular }.251 Let E = n En.S is follows that S = B. > n> .

Then T k i= 1 k i=l k = Z A ( C ( z $ T ) ) & 5 X(C(0. ) ' S are mutually disjoint. we get that p ( K ) = 0 and therefore p is identically 0. T ) c B ( z . . There exists an sequence { z k } of K such that k+oo lim f(zk) =a. It is easy to show that Assume without loss of generality that B(z.1))where B ( z .l) %k + z B(Xk31)k+m Hence . let s > 0 be such that K c C (0 .2s) and C(z*.252 Since C(z.. K c u k i=l C(z"7-) c C(0. i= 1 Letting E -+ 0. Prove that f attains its minimum on each compact set of IR". E Given a compact set K of IR".1 be a finite Bore1 measure defined on lRn.s). in K .. 1) denotes the open ball centered at z of radius 1. For any > 0 there exists an T > 0 such that g ( T ) < ( ~ T ) ~ EThere exist finitely many . zl. and define a function f on Enby f ( z ) = p ( B ( z . zk E IRn such that .&?Fir). Let K be any nonempty compact set of 1R" and let a = i n f f ( K ) . a s ) ) & . 4122 Let 1. (Indiana) Solution.

Thus.0.253 which shows that P ( B ( X . write z = 0.an-l 7: n = 1.}.. 3 ' 100 ' 100 (i) Compute the Lebesgue measure of E .a. 4123 Let E be the set of all numbers in [0.a1a2 . and therefore I # [O.an-17.Un-17< x < O. If ak = 0 for all k 3 : > n then = 0..= 0. l]\E. .1] which can be written in a decimal expansion with no sevens appearing.~~-18)~l.alaz.. (ii) Determine whether E is a Borel set..f(y)ln 5 e'zl+lyI 1 .= 0.ala2. . . ) Hence a is finite and f attains its minimum on K .. 2 (*> Show that if E c IR is a measurable set with m l ( E ) = 0 then m .2'. So E is Borel measurable and W m ( E )= 1 n=l 9"-l x -= 1 - 1 10" 1 = 0.) ..611a2 * * * Un-17 = O. 4124 Let f : IR -+ R" be a function such that for all z.yI. E. (Indiana) .. . It follows that 0.UlU2.-.2800.~~-lS. For any z E [0. .. 1 0 .Ulaz.. .l]\EU{(O.2699. 3 3 3 .aia2 * * * ~~-1699.. 1) = Q. .. y E R If(.. E E . . .. ( f ( E ) ) = 0. . The reverse inclusion is obvious. . E a contradiction. . Let n = min{k 2 1 I ak = 7)... - 1 27 28 (Indiana) Solution. .

T ) and rnl(U\E) < E . Write U = U(uilbi).e2'Ibi . a which implies that rn:(f(E)) = 0 and therefore f ( E ) is measurable. Let E 2 (-r. For any E > 0.bi)) i c ~(f(ai). (Indiana) Solution.T . (eZrlbi. Let fk = fXEn\u. 1 It follows that {fk} converges t o f in measure. Then condition (*) implies that f((ai. there is an open set U with X(U) < E such that f is continuous on E"\U (in the relative topology). Assume that r A = A for every nonzero rational number r.I 2 E l ) = m * ( { zE uk I If(x)I 2 E } ) 5 C. m * ( { zI lfk -f )(. where (ai. (Indiana) .bi)'s are mutually disjoint. Hence mn(f(E)) = 0. Let u k be an open set such that x ( u k ) < and f is continuous on R n \ u k . It follows that r n . Prove that either A or R\A has Lebesgue measure zero.. 4125 Let f : R" -+ lR be an arbitrary function having the property that for each E > 0. ( f ( ~ ) _< ) C C. Since the Lebesgue measure is complete f is measurable. Prove that f is measurable. 4126 T Let A c R be a Lebesgue measurable set and let r A = { r A I z E A } where is a real number. then f k is measurable.254 Solution. Assume without loss of generality that E is bounded. For any E > 0 there exists an open set U such that E U C ( .a i l ) ' ) .ail < CneZr&.r).

If m(lR*\B) = m(R\A) > 0 there exists a compact subset K of E*\B with positive Lebesgue measure. define function f : lR* + [0. For each measurable set E QJ then Y is the desired probability measure.255 Solution. Y 2 E a*.m). There exists a sequence {En}r=p=lmutually disjoint measurable sets of of finite and positive measure such that X = let n=l u En. It follows that m ( B ) = 0 and therefore m ( A ) = 0. T. . 4127 Let p be a c-finite measure on the measure space ( X . c )by m f(2) = 1. Then r B = B and T(R*\B) = R*\B for every nonzero rational number T . m ) ( ~ ( z = 1) such that p is absolutely ) continuous with respect to v . and v is absolutely continuous with respect to p. (Indiana) Solution. Let lR* = R\{O} and 3 = A\{O}. we conclude that m ( L . For any compact subset L of B . Prove that there exists a probability measure Y on ( X . Then f is continuous and for any nonzero rational number Hence f( z) = 0 for any z E R* Since . dY XK(Y)XL-l(Y-lz)-..l ) = 0 and therefore m ( L ) = 0.

l ) such that E > 0. It is not true. g : IR -+ IR is still integrable. There is a measurable set E of measure zero such that any x E (0.256 SECTION 2 INTEGRAL 4201 Prove or disprove that the composition of any two Lebesgue integrable functions with compact support f . (> = X { 0 } ( 4 and d . there is x EJ.. (Illinois) Solution. . since g o f (z) 5 1. m(J. 1)\E i s a Lebesgue point o f f . Prove that for every open interval I E ( 0 . let f. 2 ( 0 .l}(x). (Stanford) Solution. For example. i.. Then f and g are integrable functions with compact support. However.) < E .e. Assume that for any x E ( 0 . and fdm = 0. 1 = x{o. 4202 Let f E &(O. 1). l ) and every an open interval J. the function g o f is not integrable. l ) fdm = 0 .

. 03) is integrable (i. l ) such that z E J. Then W .e. Solution.e.. p ) be a positive measure space with p ( X ) < 03. i. m(J. Suppose that f is integrable. Show that a M measurable function f : X + [0..257 For any 2 E (0.e. f ( z ) = 0 a.. one has Jx f d p < CO) if and only if the series n=O c 00 I f (z) 2 n}) (Iowa) converges. l)\E and for any n there is an open interval Jn & ( 0 ..) < and It follows by equality (1) that f ( z ) = 0. Therefore 4203 Let (X.

taking value one on [-1. / p(t)dt > 0 4205 Let E be a Banach space. (Iowa) Solution. Then a contradiction occurs from the following limits JL% and Irn ( p ( t ) e t d t= . (a) Yes.1]. fdp = f (0) for all continuous f : B + (c of compact support? Justify.p ) a probability space. f d p = f’(0) + (c for all continuously differentiable f : R of compact support? Justify.258 Conversely. -+ E . 4204 (a) Is there a Borel measure p (positive or complex) on IR with the property that J . which shows that f is integrable. and f : X such that g o f is p-integrable for every g E E‘. ) (b) No. (X. let ‘p 2 0 be a continuously differentiable function of compact support. (b) Is there a Borel measure p (positive or complex) on B with the property that J. Let p ( E ) = x ~ ( 0for any Borel set E . If there were such a Borel measure.?r.

It follows that pn o f converges t o p o f everywhere and to h in measure and therefore h = p o f in L 1 ( X ) . If (b) does not hold. M . or (b) there exists an E E M such that (Iowa) Solution.e. T is bounded. L ( g ) = I g 0 fdp.259 Define L : E‘ -+ IR. p ) be a positive measure space with p(X) g be real-valued measurable functions with < co. p H p of.and let f and Show that either (a) f = g a. Define the linear operator T : E’ + L1(X). (Iowa) Solution. 4206 Let ( X . Since . It is ture that L E E”. We have So L is bounded. Assume that pn -+ ‘p in E’ and T(pn) + h in L1(X).. Does L E E”? Justify your answer. then for any E E M . By the closed graph theorem.

J. .A 1 5 E ~ } ' s . and S a closed set ina'.260 for any E E M . I f(z)@ S } has measure (Iowa) Solution. 4207 Let ( X .M . It follows that Therefore (a) holds. Show that { z E X 0. Suppose that 1 --JE ) P( E f@€S whenever E E M and p ( E ) > 0. one finds that a'\S is the union of countably many closed balls { z E C I Iz .If . For any E fdP = J. d P . Since a' is second countable. > 0 the sets are measurable.) = 0. From the equalities and we conclude that p ( E + ) = p ( E . p ) be a positive measure space.

not to S.1]. Show that where inf is extended over all measurable E c [0.) .X n I F E n ) ) # O * I ( But then E = {x E X & J”f d p belongs to { z E (J‘ I Iz . where 5 E ~ } ) . 0bviously. m) and let 0 < a 5 1.1] -+ (0. Take an N such that Suppose that E C [0. so m(E1) > $. EZ = E\E1. m ( E ) > a. Therefore 2 +).1] with m ( E ) 2 a . Thus . Denote El = E (f Then m(E2) < 5 . I 5 E ~ } .X.261 there is a t least an n such that P ( { ~ E Xf I ~ ) . 4208 Let f : [0.a contradiction.XnI I If(. (Indiana-Purdue) Solution.

4210 Let /I be a finite measure on lR.e. a measurable set E of measure zero such that for any t E B\E. n fn(t) = fl(t) -k C ( f k . with respect to the Lebesgue measure on lR.and define Show that f ( z ) is finite a.262 4209 Let {fn} be a sequence of real-valued functions in L 1 ( B )and suppose that for some f E L1(IR). n Therefore for any t E lR\E. by Levi’s Lemma. It follows that fn + f almost everywhere. (Indiana) Solution.l ) ( t ) k=2 converges.f k . Prove that Solution. (I Zin o is) l there is. Let . Since fn -+ f almost everywhere with respect to Lebesgue measure.

p ) be a positive measure space.e. [p].e.where dv(z) = 1 dx +22’ by h b i n i ’ s Theorem. without loss of generality. Suppose that fn -+f a. m] an integrable function. Assume. .and that Prove that (Iowa) Solution. f n : X + [0. i. The conclusion follows from that the measure v and the Lebesgue measure are equivalent. and f : X ---t [0.e.. with respect to the measure v. CQ] a sequence of integrable functions.263 then g E L 1 ( R . d v ) . the function is finite a. that p is totally a-finite. 4211 Let ( X .M .

Let d v ( 2 ) = g ( z ) d p ( z ) . Then v is a finite measure on X . Let Then g is integrable. Let d X n ( z ) 1f n ( z ) d p ( z ) . then by Fatou’s Lemma one has {s. f n d p } is bounded and therefore admits a convergent subsequence. Then 9 { $} . For any A E M . It follows that and therefore r r For any E > 0 there is a 6 > 0 such that Xn(E) < E whenever v ( E ) < 6.264 n=l where Xn’s are mutually disjoint measurable sets of finite and positive measure. 71 E m* Then An << v. {s. by the Vitali-Hahn-Saks Theorem. There is by Egroff’s Theorem an E with v ( E ) < 6 such that converges t o f uniformly on X\E. equivalent t o p. d p } is any such subsequence. If f n .

l ) . then (Iowa) Solution. + p(Xn) xx. finite measure on M . Conversely. The set function is of course a positive. Let W=C1 O0 n=l 2 1 . finite measure on M . and that P ( E ) = 0 ep ( E ) = 0 for E E M . (b) Show that if f is a complex function on X which is measurable with respect to M. positive measure on a u-algebra M in a set X.}be a disjoint sequence of M such that X = U X . (a) Let {X. ) > 0. P . and n=l 03 p ( X . Show also that r is a positive.265 4212 Let p be a a-finite. If p ( E ) = 0 then P(E) = 0. as desired. (a) Show that there exists W E L1(p) which takes its values in the open interval ( 0 .

e. Then fnW is integrable with respect n-oo to p and hence fn is integrable with respect to p. if f is integrable with respect to the measure p. there is a sequence of simple functions { f n } such that lfnl 5 f and lim fn = f . we see that fW is integrable with respect to p. . i. n In general. m+oo n-cc lim fnW = f W . Accordingly. Since and lim fn n+cc = f. . there is a sequence { fn} of simple functions integrable with respect to p such that Then fnW’s integrable with respect to p . n f =CaixEi! i=l where p ( E i ) i= 1 < +a.= i l . then both sxf d j i and sxfWdp equal to 2 aiii(Ei).l n . . if fW is integrable with respect to pl assuming without loss of generality that f is positive-valued. We have and therefore J.266 (b) If f is a simple function.fdF=J X fWdP Conversely.. .f is integrable with respect to p. and since are lim n-+w Jx\jnw-fmWldp= lim n-+w m+cc ~f~-fmldji=~.

4214 Let { fn} be a sequence of non-negative measurable functions in LP( R) for some 1 < p < 00.1] --t R such that (3) 0 5 g 5 1. IJ. There is a continuous function g : [0. (Iowa) Solution.fndml % hdm < E whenever m(E) < 6.1]. if : ( Indiana-Pvrdm) . Show that fn 4 f(P) and only if f 4 fP(L1). Then we have = n-+m lilfnXAdml It follows that Jiir J.11) and (ii) Ifn I h for all n. fndm + 0 for each Bore1 subset A [0. For such a 6 there are a compact set K and an open set U such that (1) K A C U and (2) m(U\K) < 6.Suppose that (i) fngdm -+ 0 for each g E C([O. For any E > 0. 1 and let (fn) be a sequence in 3 L 1 ( m )and h a non-negative element of L1(m). ( 4 ) g = 1 on K and (5) g = 0 outside U . fndm= 0.267 4213 Let m denote the Lebesgue measure on [O. Show that J. there is a 6 > 0 such that IE .

+ - Since IJu~ > 0.f I L Ifn . - PI 0.fP)/ I J tfi Jfp. Just as above. f > 0. Therefore Conversely. we have Since T:+X SO =f.fl which implies that IIE .268 Then f n 5 f and IL . take .+0.”+fP. 2 f. Suppose that J If.f (Ip -+ 0.” Then fn -f p l . it follows that + - Jfi For any c N1 such that .

Denote 7 = &/m(A)f. (a) Give an example where { f n } converges to f pointwise. such that m ( A ) < +m and J. Thus Jf. Take S > 0 such that < EP for n if m ( e ) < 6. 5 M . UP Ifn -flp) for any n >N. a measurable set. 0. f: > N1... < EP e if m ( e ) < S and n > N1. 4215 Let 1 5 p < 00. Take A .fllp f . Take N such that N 2 a n d m ( l f n . Vn .269 for n > N1 and any measurable set E . llfnllp and llfn .f l > q ) < S i f n > N . < f P &PP Then J. Wehave N1 (J.. All parts refer to Lebesgue measure on R.

1]. 4216 Suppose f n is a sequence of measurable functions on [0. Then gn 2 0.l f n .1] with 1' and f n --+ lfn(z)12dz 5 10 0 a.e. (c) Set gn = 2"(lfn Ip + Ifl") . [O. we have Therefore n-o3 I f n . Set n-cc fn =n ~ ( ~ .fllp 0 (Indiana-Purdue) Solution. Prove 1' lfnldz --t 0- . on [0. Using the Fatou lemma.fll = llfnll = 1.13.fJPdz = 0. (c) Show that if { f n } converges to f pointwise and llfnllp -+ IlfllP.f l P . llfllp 5 M . 1 1 and (b) We conclude that llfn . i ) . what can you conclude about Ilfllp? Justify this conclusion.270 (b) If { f n } converges to f pointwise and l ] f n l l p -+ M < $00. Then zE lim f n ( z )= f ( z ) = 0. By the Fatou's lemma. and n-+m lim g n ( z ) = 2 p + ' l f l p pointwise. then + - llfn . (a) Consider L[O.

27 1 Hint.1 I E ) < E and (2) f n converges to 0 uniformly on [O. and we have rl 4217 Let (X. Use Egorov and Cauchy-Schwartz.00). (Iowa) Solution. (Stanford) Solution.f p d p ) ’ = exp ( J . equality (1) holds. for any 0 < p < 1 and any 2 EX . Otherwise.e. 1 We have by the Cauchy-Schwartz Inequality and therefore Let E + 0. Show that i f f is an integrable function with values in [1. p ) be a probability space (a positive measure space with p ( X ) = 1).log f d p ) .M . For any E > 0 there is by Egorov’s Theorem a measurable set E c [0. 1]\E. It suffices to show that If f = 1 a.1] such that (1) m([O.. then lim P10 (J.

a) Since The series b) Let C n=2 00 ann-2 converges in L1[2. l ) decreasing to 0. 1+ ( f P n Pn - lim n00 ( log(1 + JJfP" J.m). . 4218 Let { a n } n > 2 be a sequence of real numbers with lan[ < logn. where { p n } is any sequence of ( 0 .(fP" - - 1)dP) J#" - 1)dP Pn = Llogfdp.272 By the Dominated Convergence Theorem we have n-co lim log(Jx f P n d P ) Pn = = n- lim 00 log(J. n fn(x) = -j+k-5. k=2 x 2 2. (Stanford) Solution. 00) b) Prove that where the sum on the right is the pointwise limit (you need not prove that this pointwise limit exists). Consider 00 n=2 a) Prove that this series converges in L1[2.

By the Riemann-Lebesgue Lemma. and let { c n } be a sequence in lR. 1 2 .273 Then (i) n-w fn(z) = lim 00 cc C aklc-" k=2 and (ii) Ifn(z)I 5 C lanln-". 2 Theorem. Show that where X is Lebesgue measure. H IakIk-" k=2 is integrable. we have = -X(S). and by the Dominated Convergence 4219 Let S be a bounded Lebesgue measurable set in IR. (Iowa) Solution. k=2 00 By a).

Y ?M2+Y211) If(. for ) b .274 4220 (a) Is the function Lebesgue integrable on the unit square 0 5 x 5 1. 0 5 y 5 l? (b) Compute the repeated integrals in the two orders.. Y)ldXdY rdrde drde (b) We have . (a) The function f is not integrable on the unit square. be explicit!) (Iowa) Solution. (c) Does the integral exist? (If you use a theorem.

By the Dominated Convergence Theorem + If I 4222 e t z f ( t ) is in Suppose that f is a measurable real valued function on l such that t R L 1 ( E )for all x E (-1. It is easy to show that for any x 20 2. Evaluate n-cc lim L n l n ( l + (!)’) dp. p ) be a measure space and f E L ( p ) .M .275 and 4221 Let ( X . Define the function H . (Iowa) Solution. In( 1 + x’) 5 It follows that for any n E lN the function n In( 1 !$) is dominated by and therefore integrable.l).

l). Next we show that dx E (-1.x ) ~ (x < z' < z ) = I(ett - e")tf(t)l. for any y ) > z.zlet"t2f(t). +m etztf(t)dt.Prove that 'p is differentiable on (. l.l). l ) . (Iowa) Solution.1 . Since for any x E (-1.z l e ' z X t f(t). t t Iy . < z < y) = \ e t z " t 2 f ( t ) ( z. (x 4 and t A 2 Iy .276 for all x E (. et"t2f(t) is integrable. we see that t follows that 4223 Evaluate . I). 2 0 and x < y < H < 0 and x < y < A' lzz. e"f(t)dt = -m J_. It kkz By the same method as above. For any fixed z E (-1. and we conclude that t H e " t f ( t ) is also in L1(E).1.

i=l )(l-&). First show that (1 E)n 5 e" for x + 2 0..(l-S Z! i=l + and n-cc lim ( 1 + . . Since for any n E IV and z 2 0. we have (I+ x --) n s e x . By the Dominated Convergence Theorem = brn e-"dx = 1. (Stanfod) Solution.. Hint.277 justifying any interchange of limits you use.)n =ex.

p ) be a probability space and f : X function. For the function f : X I x X z + B defined by . 4225 For i = 1. Let Mi = 2N (the galgebras of all subsets of nV) and let pi be the counting measure. Since l n t 5 t l n t for any t > 0.co) a measurable (Iowa) Solution. Under the condition that f In f is integrable we conclude that both f and In f are integrable. Prove or disprove: + [I. otherwise and 0 5 In f 5 f In f we see that both f and In f are integrable.2. and Indeed.278 4224 Let ( X .A. since ' '{ f ( x ) 5 el "f'ln f . let Xi = nV (the natural numbers).

Compute the iterated integrals and How do you reconcile your answers with Fubini's Theorem? (Iowa) Solution.j)is integrable and J. f ( i .f ( i .2--i. j ) d p z ( j )= -2-i + 2-i Therefore For any j E X 2 . j ) = 2). and for X > 0 define Cp(4 = P({Z E Q I f ( Z ) > XI) .p ) be any measure space. f ( i .. j ) is integrable and Since f is integrable ( C I f l ( i . j = i. For f E L1(Q. the two iterated integrals exist Li€N and coincide by the Fubini'i Theorem. i H = 0. j )= { 279 . p ) . otherwise. 2-i. For any i E X I . 4226 Let (0. j H f(i. j = is 1.

2 and II. A) H x ~ O .I).) + but on the other. are Bore1 measurable and that llflll = J m ( ' p ( A ) + +(A))dA. The measurablity of 'p and follows from that they are monotone. By Fubini's Theorem we have provided that either side exists. ~ ) ( l f l ( zA) is measurable. On one hand.280 and 4(A)= P ( { Z Show that the functions 'p E c I f(z) > -A>). 0 Hint. Set . As usual. X . . (Iowa) Solution. The function (z. 4227 Let f E L 2 ( 0 . it may be helpful to consider f positive first.

0 < h < 1 where c is a positive number independent of f. Imine) Solution.F ( z ) h l2 dz) 4228 Let f. l ] .281 Show that for each h. .g E L1(O. We have ( I J1-h 0 F(z + h) . (UC. y E [ O . 1) and assume that f(z)g(y) = f ( y ) g ( z ) for all z.

Y) I 0 5 2 < Y 511 (Iowa) and dA denotes the planar Lebesgue measure. .282 where A = {(Z.

we have + which shows that I ( p ) is a nondecreasing function of p E (1. . n 1 1 i=l If. The function I is strictly increasing if and only if f is not almost everywhere equal to the constant function. is a nondecreasing function of p E (1. .=.n}. . Show that for any [0. for i E { 1... For any p . .w)-valued measurable function f on X . . q E (1. w) with p < q . i = 1 . Under what circumstances is I ( p ) strictly increasing as a function of p? (Iowa) Solution.m).. let CY = and = then $ = 1.n. . let M such that p. > 1.. . fi E CPi ( p ) . . By Holder’s inequality. 4302 Let (X. * f n E Cp(/l)? (Iowa) Justify. m). must it be the case that fifi .p > 1 and be positive numbers c. p ) be a fixed measure space.283 SECTION 3 SPACE OF INTEGRABLE FUNCTIONS 4301 Let X be a positive measure space of total measure 1.

*. IIfnIlp. (Iowa) .1 - which shows that ppk c+k12p > Pl P k ? ' Pk+l PPk 1 pk+l?p < ' 9 1. fk belongs to 1 L Pk+:-P ( P ) and Then by the case n = 2 we conclude that (fl .)> = 00. and by H6lder's inequality. .. b. f i * . Yes. 1 o < -+. 4303 Let X = { a .and L"O(p)? Justify your answers. f k ) f k + l E LP and which completes the proof. (1) In case n = 2..+ -1= .. . . and let p be the measure determined by P ( { a ) ) = 0. P({b)) = 1 1 and P({.284 Solution.1 . lfilp E L"(p) and PI If2Ip E L F ( p ) .. Assume that the conclusion and the inequality (1) hold for n = k . c}.fnllp 5 llflllpl . What are the dimensions of the vector spaces L 1 ( p ) . Then for n=k+l. fn E LP(p). L z ( p ) . let M be the cr-algebra of all subsets of X . Moreover llfl. .-1.. By inductive assumption f i e .

Let f be a non-negative function in P ( E )for some 1 5 p T > 0. s > 0. Denoting we have 4305 T +s =p. we conclude that dimL'(p) = dimL2(p) = 1. c } we have dimLW(p) = 2. + < oc) and let .285 Solution. We have dimL1(p) = dimL2(p) = 1 and dimLW(p) = 2. Since two functions are equal in L"(p) if and only if they are identical on { b . 4304 lo Show that f d m = 0. Since each function f of L 1 ( p )or L 2 ( p )vanishes at c and two functions taking the same value at b coincide in L 1 ( p ) or L 2 ( p ) . ( Indiana-Purdue) Solution. Also let f h ( z ) = f ( z h).

Since we have . (i) Obviously follows that fh E Lp(R). Show that (IndianaPPurdue) Solution. (ii) Investigate what happens to 11 fifslll as fi E Ihl + co. From f[fS L f .286 (i) Show that fif s E L1(R). (Indiana-Purdue) Solution. (ii) We claim that lim lhl+m 11 f i f s I I 1 = 0. N ] . Therefore + 4306 Let f : R -+R+ be measurable. f" E L f . Now if Ihl > 2 N . and P +P = 1. Set p = 1 + f .N . Then p > 1 and f + & = 1. and let 0 < T < 00. then z h 4 E whenever z E E. For any E > 0. take N such that JEc f P < E where E = [ . it E L1(R).

287 which implies 4307 Let f be a bounded measurable function on (0. : Therefore .l). For any Then E > 0. llfllm = I I Ilf llm. Prove that (IZZinois) Solution.

Then f 2 1 2 is not integrable. 11). The minimum value of p for which f may fail to be in LP is 2.1] satisfying Determine those values of p . prove that s = {f E C[O. for any p E [l.let . For each n E IN. (Illinois) Solution.13.1. 1 . If 1 5 p < 2. 3 (Iowa) Solution. Let f(z)= However for any t > 0 & . It suffices to show that S is not closed in L1([O. 2).1 I llfllcc I 1) 1 is not compact in L1[O. - 1 (1+t)2 - < 2. 1 5 p < 00 for which f E LP and find the minimum value of p for which f may fail to be in P. 1 l+t 4309 With Lebesgue measure on [0. then f E LP. Indeed. m M it follows that f P is integrable.288 4308 Let f be a nonnegative measurable function on [0.

Since it admits no convergent subsequence.i. in the strong (i.(z) = sin(nz) for z E (0.2r) by Holder's inequality.2 1 .".--too lim (un. u .}rZl . (Stanford) Solution. norm) topology of 2. . n = 1 .}. defined by u. For any m. with inner product defined by 7) ( u . 2 ..w E L1(0.e. + 0 as n -+ 00 in the weak topology of X .271).w) = 0. for every w E 'H. i. E 'H. Show that (a) the set is closed and bounded.w) = 1 2= u(z)w(z)dz.~ is bounded. it is not compact. - Consider the elements u. .1]. - {u. n 2 1 with m # n which shows that {urn I m E IN} is closed.289 4310 Let 'H be Hilbert space L 2 ( 0 . the Hilbert space consisting of functions f E L2[0.e. (b) u.... the set {u. but not compact. (b) For any w E X .e. (a) Obviously. 4311 Let H1 be the Sobolev's space on the unit interval [0. By RiemannLebesgue Lemma.1 such that 1 . w E 'H.

(iii) Prove that unless f is constant almost everywhere. 2 a ) . (i) Give the Fourier expansion of fh.h). (ii) Find the &-norm llfh .fllz in terms of the a. Suppose For each h E lR define the function fh by f h ( z ) = f ( z . and h. 4312 Let f be a periodic function on lR with period 2 a such that f 1 [ 0 . (Stanford) . Since for any z E [O.290 n=-cc where are Fourier coeffents o f f .1] too the series n=-w C f^(n)e2ninzconverges to f(z) both in H1 and in L"O. Solution. Show that there exists constant C > 0 such that (Stanford) I l f I l P I Cllflll. It follows from (1) that IlfllP I Cllflll. 2 T ~ belongs to ~ ~ ( O .

I f f is not constant almost everywhere.291 Solution.}span a dense subspace of L 2 ( 0 . = 0.. Prove that for all 2 E [0.1]. 1 n = 1.f l l 2 = 0. Then 4313 Let f n ( x ) be an orthonormal family of functions in the Hilbert space L2(0. (i) We have n=-w +w n=-w (ii) We have n=-w (iii) I f f is constant almost everywhere.. then a. It follows that fh = f and 11th .. By Bessel’s inequality . there is an n # 0 such that an # 0. (Iowa) Solution.1). n # 0.1). and that this inequality is sharp (equality) if and only if {f.

We have where K = supp(f) and . 4314 I f f is a function on IR. (Stanfonl) Solution.fllL2(nz) = 0.1) while is a closed subspace of L2(0. Conversely. Then L t+O lim llft . containing span{x(o.z] 1 0 5 z 5 l}. y E [-1. it follows and therefore. then t-+O lim llft .1 b l Ia Ib 5 I} O is a dense subspace of L2(0.l).[-1.f n t l l 2 + llfnt . . Suppose that f : lR + a' be a continuous function with compact support. I~0] 5 z I 1) = s p a n { ~ ( ~ . l]} is compact.292 If {fn I R.then {fn} is a n orthonomal basis of L2(0.1] = { z -y 12 E K .1).l). In general.fllz 5 S(1lft . E nV} span a dense subspace of L 2 ( 0 . Then ft converges to f uniformly. {fn} span a dense subspace of L2(0.fllz. Prove that i f f is square integrable with respect to Lebesgue measure.1) and therefore (1) is sharp.fllz + llfnt fnllz) = 2llfn . let ft be the translate ft(z) = f ( t + z). since s ~ a n { x ( o . there is a sequence {fn} of continuous functions with compact support converging to f in L2(IR).fnll2 - + llfn .fllz) i t-0 2 l l f n W .

00. N 2 0.and let Show that for every E > 0. You may wish t o think about harmonic functions on the unit disk. we have . (Stanford) Solution. Hint. 4315 A trigonometric polynomial is a function p on lR of the form N n=-N for some Cn €a'.fll2 = 0. Then . Then where a ( d z ) is the Harr measure on S1such that a(S') = 27r. Let u be the harmonic extension o f f to the unit disc D = { z 1 IzI 5 1). there exists a trigonometric polynomial N n=-N such that Ilf and -PI103 <E ICnl 5 lan] for all n. t-0 lim lift . Regard f as a continuous function on the unit circle S1 of a'.293 Letting n . Suppose f is continuous and 21r-periodic on El.

Since C([O. 1 ) I f ( z ) = 0 or 1. Let 1 1 S = {f E L”([0. (ii) False.11) is separable while Lw([O. (i) False. 4316 Prove or disprove the following statements: (i) the set of continuous functions on the interval [0. -< 2 5 -. 11).l]) (with Lebesgue measure) and = 1. 3 n+1 n Then as a subset of L”([O. 13).1 ) is not 3 dense in Loo([O.P {gn}T=l a sequence in Lq([O. (Iowa) Solution. 11) ( Stanfo rd) Solution. since the distance between any two elements in S is 1. However. If lim f n = f in l.f(0) = 0). (ii) Lw([O. is a separable metric space. Indeed. n E m. since .S is of cardinal N. l]). S is not separable. fngn -+ f g in measure.11) is not.294 There is an N such that where r is fixed so that Put N n=-N which is required. C([O.1] is dense in Loo([O.is it true that f n g n n-w n-w -+ f g in measure? Justify.l]). 4317 Let be a sequence in LP([O. where {fn}r=l + and lim gn = g in L Q . Yes.

iff . = d(g. d(f. Obviously. h E S. (Here. and that if and only if fn -+ f in measure. define Show that d ( f . For f ..g) = 0 iff g) f = g in S.g) is a metric on S.e. fn --+ f in this metric (Stanford) Solution. g . we have & - 0 a.1 For f E S. S = {(equivalence class) of measurable complex functions on X}. as usual.e. two measurable complex functions are equivalent if they agree a.295 we have for any E >0 where A is the Lebesgue measure. 4318 Let X be a measure space with measure p and suppose that p ( X ) < Let 00. g ) = p(f . f) 2 0 and d(f.

Prove that there is a constant M such that . If fn + f in measure.s which then implies that g "2. then -%0 in measure by the equality -f 1+f -+ { 2 EX I 1 + Ifn . If d(fn.296 Therefore d is a metric on S.f) -+0. 4319 Let g be a measurable function such that lfgl < 00 for every f E P ( B ) (fixed p >_ 1). * .* L lsnfl 5 . then - -d(fn. It follows that the sequence of bounded linear n-tm functionah f H J f(z)g.(z)dz on P ( I Rconverges to f H . ( st u n f o d ) Solution. . 5 n and 11 2 5 n. By the Banach-Steinhaus Theorem.f ) -+ 0. Define for each n E B a measurable function gn V gn(2) = { g!x" 5 hfl 5 if Ig(z)I else. lim gn f = gf. ) J pointwise. . f ++ f(z)g(z)dz is continuous on P ( R ) and therefore there is an h E L q ( R )such that .fl Ifn - fl > &} = - { 2 E XI Ifn - f l 2 l--E '} (O<&<l). Then gn E Lq(IR) and for any f E L p ( l R ) . By the Dominated Convergence Theorem d(fn. Moreover. 191fl .s all f E Lp(IR). f(z)g(z)dz. h E L q ( R ) .f) & 1+E + 0 as n ---t m..5 Isfl.

g E L 2 ( R ) . or the real numbers modulo 2a). a) Show that f * g E L o o ( R ) . Let for any n E Z.x.Do not neglect to check the measurablity of f *!l. We have ~ ~is an orthonormal basis of } ~ ~ a Therefore {f * k I k E L 2 ( S 1 ) }contains L2( S1). . The convolution of two functions on S1 is f *g(a) = - 2'.define the convolution f * g by f * 9 = J_.297 4320 Let S1 denote the unit circle (the set of complex numbers with modulus one. {xn I n €27) and therefore is dense in 4321 For functions f .Show that the linear space { f dense in L2(S1). where the integral is with respect to Lebesgue measure. too f(y)g(z .Y P Y . *k 1 k E L2(S1)} is (Iowa) Solution. I'" f(%7(ct - wo. Suppose that f is an element of L2(S1) with the property that its Fourier coefficient is non-zero for all n E Z.(O) = elne then { L2(S1).

f * g(z) is well defined. Since l(f * 9 n . 4322 I f f E L 1 ( R ) .9 / 1 2 we see that h = f + 0. Tfis bounded. (Iowa) Solution. Hint. By the Closed Graph Theorem. By (l).f * g)(z)l I llfllzllgn . Since by Holder's inequality If(y)g(z .define + (1) + i ([r If(y)l2dY) (Jt" Id.Y)ldY II l f l l 2 l l s l l 2 . --oo - YVdY) . closed graph theorem. Show that Tf is a bounded linear operator. b) Let { g n } be a sequence of L2(IR) such that gn --t g and Tfgn-+ h in L 2 ( R ) .f * g is bounded and therefore belongs to L m ( B ) . Indeed. you need not check this. Define Tf : L 2 ( R )-+ L2(BZ)by T f ( g ) f * g .298 b) Suppose that f has the property that for all g E L2(BZ) the convolution f * g is also an element of L2(IR). I r+m I = 0. = It is evident that Tf is a linear operator. a) We will show that f * g is uniformly continuous. 2 E lR * g = T f g .

( Stanford) (1) In general. have we 4323 Let f : [0. ?(() Solution. co) be an essentially bounded function. For any a with 0 < < Ilflloo.1] -+ [0. (Y let . say +0 as + 00. there is a sequence {fn} of functions of the form (1) such that Then Letting n -+ +a. If f is simple. for any f E L1(IR). Show (Indiana) Solution.299 Prove that. llflloo > 0.

we get that 4324 0 . M . Let 1 < p < Suppose that { f k } is a sequence in p ( x )such that ~upIIfkI(~ 00 and < k fk(Z) n+m lim = f ( z ) exists for pa . For any k E N . also -fill > 0. Hence Letting a 1 llflloo. then X(E. . I} of {lfkl Ik E m}. p ) be a measure space for which p ( X ) < 00. 2. If on the contrary k+m lim Ilfk-flll llfk denoted {fk}.) > 0. = [O. 0 Let ( X . Prove that (Indiana) Solution. and Lp(x) reflexive there exists a subsequence is converging to If1 weakly in I.e . such that lim k-m > 0 there exists asubsequence of {fk}. the Dominated Convergence Theorem.l]\E. Since { l f k l } is bounded in P ( X ) {(fk.p(X). by = 0.300 and F. where X is the Lebesgue measure.

Then Letting E + 0. a. .301 By the Vitali-Hahn-Saks Theorem. For such S > 0 there exists by Egroff’s Theorem a set E such that p ( E ) < S and {fk} converges to f uniformly on X\E. for any that for k = 1. . we get that a contradiction..2. E > 0 there exists a 6 > 0 such r r whenever p ( E ) < 6.

. Observe that (*) is equivalent to (Iowa) Solution. y). So lim g(z.. v) 5 g(z.) L g(Y1 Z)(Y - ( for all 2.f’(z) = 4(z) whenever # is continuous at z. v 5 y g(u.e. u 5 z. Y E (071)). since by (l). z) then 4(z) 4 is nondecreasing and L !J(Y. y) Y+X # 4(z) exists except in a countable set. Again by (1) f is Lipschitz and therefore absolutely continuous. i.13 -+ lR and g : [0.f(.302 SECTION 4 DIFFERENTIAL 440 1 Let f : [0. However. y) = +(z) exists except in a countable Y-+X + set and that for 0 5 z 5 y 5 1 we have Hint. Let #(z) = g(z.Z) L y-’ 4(Y) L g(z1y) L By (1) we have an inclusion (1) while the latter set is a t most countable. 1 + lR satisfy the inequality 3 f(Y) .1] x [O. (*) Assume also that g is nondecreasing in each variable. Show that lim g(z.

with for all 2. 1 1 f) ( . 5 M.1]. Since I ‘ z I f() such that and = f(0) + i= 2 f’(x)dz. by the Mean Value Theorem This completes the proof.1] -+ R is Lipschitz. Conversely. then f is differentiable almost everywhere and moreover for any x E [O. If f is Lipschitz.1] there is a sequence { g n } of continuous functions Ign(z)l rl IM. if and only if there is a sequence of continuously differentiable functions f n : [O. (11) fn(x) + f ( x ) for all 1 E [0. There are several different ways to do this problem. 1 1 For any n E A . define T which is required. E [O. one is to use the Fundamental Theorem of Calculus.x E [0.y E [0. (Stanford) Solution.1]. 11. .303 4402 Prove that a function f : [0. : Hint. 1 -+ R such that 1 (I) [fA(x)l I M for all x E [O.

1 . SR .1] such that (a) f(z)= n=l C 00 fn(z) converges for every 2 E [O. (Indiana-Purdue) Solution.1]. 3 Show that f is absolutely continuous on [0. (ii) f ( z ) -+ 0 as 121 + 00. Then 1 It follows that f is absolutely continuous. 4404 Assume that f E A C ( I ) for every I c R. (nlinois) Solution.304 4403 Let {fn} be a sequence of absolutely continuous real-valued functions on [0.for every I c R. Since f E A C ( I ) . Let n=l in L1\O. If both f and f’ are in L1(R) show that (i) f’ = 0. 1 .

4405 Let { f n } C AC([O. there exists g E L1 such that f : + g. Since fn E AC( [0. 0 f’(z)dz + I’” f ’ ( z ) d z = 0. lim x-+w sox’ ( t ) d t exists.f(O) = -f(o). . Then Moreover. l]). In the same way. 1 ( Indiana-Purdue) Solution.1 ) such that fn + f uniformly on [0. Since x++w f E L1.305 Since f’ E L1.show that there is f E AC([O.fn(0) = 0 for every n. 11). Thus So there exists an f E C([O. must have lim f ( z ) = 0. Therefore we f’(z)dz = Iim f(z) . fn(z) = 1’ fXW. we have lim f(z)= 0 and x-+-w Thus L f‘(z)dz = J_. L’ which implies f E AC( [0.l]).1]. which means f x-+w x++w lim f(z)exists.13.}is Cauchy ( L l ) .1 ) such that 1 f n -+ f uniformly on [0. If {f.

be open subsets of [0. 2 1 < 2 2 . For any 2 1 . lGnl I -.1)1 I Therefore we find a Lipschitz constant 1 $. I f ( 4 . Take N such that f 5 < 6.306 4406 (a) Assume that f E A C ( I ) and f’ E L”(I). 2 (b) (1) =+ (2) obviously. Take N 2 0..21) < 5.2 E I . such that Then N l f ( z 1 ) .b j ] } c I with (Indiana-Purdue) Solution. (1) f : I + R is Lipschitz (2) E > 0 3 35 = S ( E ) > 0 such that { I j = [ a j . Take { c j } such that x1 = cg < c1 < .2 1 2 6.f(. 4407 Let Gn. (b) Show that the following two statements are equivalent. (a) For any zl.. 2 .2 1 < 5.f(zl)l < 1. Then Suppose that 22 . Suppose that 22 . I. bj] c IIj I 5 6. we give the following fact. 2 2 . E I .c.1] such that GI 3 G 2 3 * * * . So we have 4< N(z2 . < C N = 2 2 . n = 1 . (2) 3 (1). . Show that f is Lipschitz. Take 6 > 0 such that If(bj)-f(aj)l 5 E for any Ij = [ u j .-1 < 6. c. :2n 1 Let . .

a < b.f(x”)I Gn = 0. Then GK # 0.f(x’)l = C IG.2’1. there is k’ > K .x 6) c GK.307 Show that AC([O.a i ) < 6. Set 6 = &. Take a. Gkt # 0. (ii) Suppose that there exists no such that Gn = 2‘‘ 2 2 8 for n 2 n o . x’. b E (x .x”1. + a . suppose that for any natural number K . then < & + N C(bi U i ) < & + .x‘’]~5 no1x” 1 no Conversely. i=l n=l K n=l K 1 = Klb. Then .1] iff there exists no such that (Indiana-Purdue) E > 0. n 2 no. 6 > 0 such that (x . If(x”) . (i) For any (9 f 5 MIX’. Take x E GK. n [x’.1] and C ( b i .6. .11).al. take N such that C N+1 M & < 5. If {(ailbi)} is a sequence of disjoint open intervals in [0. (ii) If(x’) . b.= &. Solution. for > x’. x” E [0. & - 2 which means f E AC([O.S. x Then + 6) c GK = K Gi.11).

308 4408 For what values ‘a’ and ‘b’ is the function (i) of bounded variation in (-1. f is Then 1 E 1 1 . since f is continuously differentiable on (0. Indeed. 1 and 1 E > 0.1]. for any of bounded variation on [ E . (ii) defferentiable a t ‘0’.let us first establish the following equality provided that either side is finite. The function f is of bounded variation if and only if To show (l).l). (Iowa) Solution.

b > 0. Case 111. b < 0. Then sup 0<€<1 If (&)I < 00 iff a 2 0. there is a 6 that -1bl2 Let N = > 0 such that 0 < x 5 6 implies d3 . Then b1 If a + b < 0. 0<€<1 00 E Case 11. since and . for Since a + 6 = 0 and a + b > 0 imply respectively that and ' the function f is integrable on (0. 1& + 1.2 I4 -.> Ibl xb .1]. If a + b 5 0. Then sup O<E<' If(&)I < iff a + b 2 0.:l!al € dx < oc). b = 0.309 Case I. while a 2 0 implies that sup O<e<l J' lf'(x)ldx = sup J' .

all 2 2 .) f is absolutely continuous.)ld. Since by the condition that If(.27r] and f’(z) = 0 a. . b) Let f be a real valued function of bounded variation on [ 0 . = +m. (ii) f is differentiable a t 0 if and only if { %>l or { b<o a > 1.27r] satisfying . (Stanford) Solution.e.yI.e.310 I’ and If’(. then and and therefore f’ is integrable. a) The conclusion is ture.. 4409 Prove or disprove a) Let f be a real function on [0. we have rl If a + b > 0.27r]. on [0. 2 ~ with f’(z) = 0 a.f(y)l <: 1 . ] Then f must be constant. f is constant. Then f must be constant.y E [0.

e. Show a1 b) if a+h h-+O f(z)dz = c. a) For any E > 0 there is a continuous function : lR -+ lR such that x-L f(t)dt = j z f+ ( t ) d t j z f. that f and g are real-valued.? State explicitly the theorems you use. Since the function is of bounded variation. a+t g(z)dz +c a. g E L1(IR).31 1 b) is false. Let f(x) = 0 on [0. 27~1. T ) and f ( x ) = 1 on [T. Then f is of bounded variation with f’(z) = 0 a. c ) if then there are constants a . 4410 Let f . (S t a n f o d ) Solution. c such that f ( u + t )= J. without loss of generality.e.e. Assume. it is differentiable almost everywhere by Lebesgue’s Theorem.. Can you deduce that f’(x) = g(z) a. but f is not constant.( t ) d t -cc --oo .

. f ( z ) d x = c. c) By a). we see that and therefore b) We have by a) h-0 lim f(' a + h.f ( x ) d Z h = f ( a + t ) .312 We have Letting c t 0. t E IR.c a. there is an a E BZ and a c such that I ra+h J a k.e. .

a. 1. (1) Prove that the following two statements are equivalent: (a) There is an f E L2(a.b).e. (Iowa) . It follows that f’(z) = g(z) a.b) such that F(z)= lz fdm ( m the Lebesgue measure). < 2 of ( a . (2) Show that the smallest constant M in (b) is equal to Ilfll. (b) There is an M > 0 such that for any finite partition zo < z 1 < .e.e.b) -+ Dl be measurable. t E IR. 4411 Let F : ( a .. we have = 0.’..313 By b) and by the assumption.. .

. . (1) (a) 3 (b). a t most countable). for any mutually disjoint intervals (a.b ) . 2 E (a.b ) E such that ---f F(x) = LZ f(t)dt + c. o ( a i j . We will show that if E l . Since a F is absolutely continuous.. and therefore there is an integrable function f : ( a . say Ei = 3EAi . .we have on)of ( a .314 Solution. then for any E > 0 If Ei's are open sets of ( a .En are any mutually disjoint measurable sets. By the Cauchy-Schwarz inequality (b) 3 (a) Obviously.b). & ) (A. n E lN. .b). .

. . . . there is for each i an increasing sequence { E .315 then by inequality (2) I 2 00 It:. Let E = {z ’ . f I 2 . . . En.a i j 5 M.Enj and passing to the limits we see that (3) holds in general. In general. I 65 f ( z ) < 1f ( z )5 and E. j ) = m ( E . ) . . and lim p ( E . . . . If Ei’s are compact sets there is for each i a decreasing sequence { E i j } E 1 of open sets of ( a . Applying (3) to the 3-00 mutually disjoint compact sets Elj. To show that f is square integrable. j } E 1of compact sets such that j=1 u Eij 00 E. it suffices to show that 00 where En = {z In 5 If(z)12 < n + 1). = {x Then M I --< -6}. . b) such that Ei = j=1 Eij and Eil n E k l = 0 (i # k ) . . Applying (3) to the mutually disjoint open sets Elj.=I j € A i P i j . Enj and passing to the limits we see that (3) holds for E l .

b )by (3).b].11 f E C[a.b ) there is a fll2 ---f 0. so We have Ilfnll2 = Ilf112. It follows that sequence {fn}rz1 b] such that 11 of C[a.b ) . (2) Define. IlfII = nlimO llfnll = 40 llfll = Ilfll2 for any f E C[a. For any E (zip'. then 11. . This completes the proof. for each f E L 2 ( a .316 5 M.b]. For any fn - f E L 2 ( a . Moreover 5 l l f l l n by (1). there is is a norm on L 2 ( a .a ) ) such that + Then Illf1I2. z i ) (xn = a k ( b .

.1]. where X the Lebesgue measure. p(E)= 0 It follows that p -4 - e 2t = 0 a. M the c-algebra of Borel subsets of X. Let a ( t ) = t 2 . We have and 4413 Let X be the Lebesgue measure on R2 and R defined by: 2 /I and v the Borel measures on . t E E e X(E) = 0 e q5(E) = o . t E E @ X(E) = 0 e 3 t 2 = 0 a. A. Justify.Y and q5 on M by 1 E Ida and $ ( E )= IdP. Does 4~ exist? Does exist? Compute the value of the Radon-Nikodym d? derivatives that exist. (Iowa) Solution.e.317 4412 p(t) = t 3 and define measures p ( E )= Let X = [0. 2 p ( E )= and ] 2tdt r E + ( E )= We have 1 E 3t2dt. For any Borel subset E of X .e.

It is true that v << p and it is false that p << v. however.pn) n A ) = 0. -= n=l c " 1 $~(n. m].2n)n A ) = 0 and therefore p ( ( n . Since and we have " 1 du n=l and dv d X Therefore (1) holds. i.. Is p If so. m]. 4414 Let p be the Lebesgue measure on [0.in]. i. (Iowa) Solution. find the corresponding derivatives. let A = 21 and then v ( A )= 0 but p ( A ) = We show next that (g.318 Is p << v or u << p? Find the corresponding derivaties if they exist. X((n. Define for any Lebesgue measurable subset E of [0.e. << p2 or/and p2 << pl? (Iowa) . Because p ( A ) = 0 implies for all n E N . v ( A ) = 0.

p << p2 does not hold. that q5(tg(. continuously differentiable function with a R R bounded derivative.39 1 Solution. t E IR. 4415 Let C#J : l -+ l be a bounded. and assume g E L(lR).1)) = 1.1)) = 0 and p([O.)) E L(lR) for all t E R). Define f(t) = J dJ(tg(z))dz. Since r 1 we have It follows that both p1 << p2 and p 2 << p1. l R a) What additional assumption on 4 will insure that f is well defined? (i. .e.. Moreover From p2([0.

We will show that Indeed.} such that 0 < a. l ) with the property that lim f k ( Z ) = 1 for almost every z E (0. z there exists an s with 0 < s < 1 such that Since b. There are sequences {a.e.} and {b. a. lim (bn . t E R (s # t).. and The conclusion follows from the Dominated Convergence Theorem. (Indiana) Solution. i.l).320 b) Under the additional assumption in a) above. < b. (Indiana) Solution.a.fi(z) = O for almost every lim k-rn k+oo z E (0. show that f is differentiable. Prove that . f’(t) exists for all t E IR.) = 1 . 4416 Let ( f k } denote a sequence of nondecreasing functions defined on ( 0 . Indeed. for any t . for any s. I). Assume 4(0) = 0 then f is well defined. n-co < 1.

l).) = lim k-r 00 fk(bn) =1 It follows that lim k-rm f. is integrable and I’ ks Hence k+co f i ( z ) d z = 0.321 and k-+m lim fk(a. . lim fL(z) = 0 for almost every z E (0.

.I for every finite subset {!I. Sn}c S.).". We have u ma- 00 : = 1 z € UI%P. S n (-n.pnTU I:. -i) be finite and n ) and must S is countable. ( Indiana-Purdue) Solution. Show that S is countable. For any r UIa\UI. be a collection of closed intervals in R. there must be a (Y such that m x E la. .+S. 4502 Let { I a } . . x can not be in 1.?. Obviously u1 : aEr is an open set. . Since S n (i. Show that is countable.322 SECTION 5 MISCELLANEOUS PROBLEMS 4501 Let S c R be a set of real numbers with the property that IS1+. n=l where cyn. . Take n such that but : .Q E r. . 51 (Indiana-Pardae) Solution.

Moreover.is irrational}. Therefore a a n which implies uIa\ u1 : a a is countable. P n )then x . 6 Icy. ( st anford) Solution. Let {Ui I i E I } be such an infinite set.} o f Z such that n-m 2 1 lirn (kn.which is a contradiction. Indeed for any x E A . c ( a n . lim sin(nx) = 1.27r) n-m I . I E a } is a dense subgroup of lR. Let 2 A = {x E (0. 7F Then A is a measurable set of measure 27r.323 If I . 4504 Prove that for almost every x E [0. Then ai # aj i whenever i # j . It follows that the map.21. open sets in a separable metric space X is countable.) = -. mutually disjoint. 4503 Show that any infinite set of non-empty. Thus we must have x = on or P. For each i E I there is an ai E D such that a E Ui...where D is a countable dense subset of X . for any x E A . (Stanford) Solution.21 I Ic.27r] n+co lim sin(nz) = 1. since {Ic: . i H ai of I to D is injective and therefore I is countable. 7~ 2 . there are sequences {kn} and {I.

( Indiana-Purdue) Solution. Suppose that there is a sequence of odd polynomials pn with p .1] iff f(0) = 0. {k.} either increasing to t 0 0 or decreasing to -m.2 1 I k. + f .2 1 . Show that there is a sequence of odd polynomials p n ( z ) with pn -+ f uniformly on [O. 4505 Let f E C ( I ) . n-w . ( Indiana-Purdzle) Solution. + 1)w) = 1.324 Since f 4 { k : . Show that there exists a sequence of polynomials { p n } such that pn + f uniformly on I and p l ( z ) 5 p 2 ( z ) 5 . . ~ )= 1.I EZ}. For any n.} admits a subsequence {k. If lirn Ic. + 1))= 2 and therefore n-+m lirn sin(-3kAz) = lirn sin(-3khx n-cc + 2(31. take a polynomial pn such that 4506 Let f E C([O.'for every z E I . = $00 then n-co n-cc lim sin(khz) = lim sin(khz .11). n+cc Otherwise lirn (-3kA) = +co and n-cc lirn ((-3kh)n+cc 2 T 1 + 2(31. Then f ( O ) = lim pn(O) = 0.

I]. with the supremum norm.e.1 1 1 1 r+’ is a contracting mapping on C[O. i. ex + zf(x2). there is only one function f E C[O. 4507 a) Show that the mapping I : C[O. Set 1 pn(x) = -[pn(x) 2 -~n(-x)]. Then f is a continuous function on [-1. Then Pn is an odd polynomial for any n. f(z) = es + f J.1]. (1) .l]). a) For any f and g in C([O.0]. b) Show that there exists one and only one smooth function f on [0.1]. b) By the Banach’s Theorem.1 + C[O. (Stanford) 1. 12 f(t)dt.1 such 1 that I(f) = f. We have uniformly on [0. 11. Take a sequence of polynomials p .1] satisfying the conditions: { f(0)( z=) = &f Solution.325 Conversely. + f uniformly on [-1. suppose that f(0) = 0. ( z ) with p . Set f ( z ) = -f(-z). 2 E [-1..

E L1 such that (i) llKnlll 5 M < 00.+ o o . . . I ( g ) = g . n = 1. * f -+ f uniformly.)l < 4M & E R any IyI < 6.. . lKnl Show K . (ii) ~ K . If g is another smooth function satisfying (2). }. 4508 Given f : R -+ R bounded and uniformly continuous and {K . Thus Take 6 > 0 such that lf(z . Irvine) Solution. K . . For the above 6. take Nz such that .3. . by (ii) there is an N1 such that If n > N1. . 3 . n = 1 .+ l a s n .g = f. 2 . then i. by the uniqueness of f. (iii) J12:121.e. For any E > 0..2.61 -+ O as n + 00 for all 6 > 0. Take M1 > 0 such that If(z)l < M I for all z E R.. (UC.326 By (1) f is smooth and satisfies the conditions: f(0) = 1.. .Y) holds for all 3: - f(.

E < 1.N2). co) be upper semicontinuous and define m : R [-oo. & 4 4M 2 Set N = max(N1. (b) Show: If I is an open interval contained in S.m ( z ) 2 1). (c) Show that S is nowhere dense. there exists E > 0 such that f(z0) - m(z0) < 1 .&z . It follows from (l).oo) m ( z )= liminf f(y).327 & < . Let S = { z I f(z) ..(2) that < holds for all z E R if n & & -+-=& 2 2 >N 4509 Let f : R + (-03. Irvine) Solution.. by Y-X -+ (a) Show S is closed. (UC. (a) For any zo E S".+ M . . then m ( z ) = -aon I. .

Define another metric d o n X by It is easy to show that the identity mapping of X is a homeomorphism of ( X . we can take an open interval I2 satisfying 7 2 c I1 and f(z) < -2 for z E 7 2 . 5 6 S which contradicts % E I c S. 4510 Prove that any topological metric space is homeomorphic to a bounded metric space.e. . we obtain a sequence of open intervals { I n } such that 7.328 Take 6 > 0 such that for 2 E O(z0. In the 1 same way.d ) is a metric space.o) + .there must be F E O(z0.< 1 2 2 & & +& = 1 if z E O(zo. Z ) . . () (c) Suppose that S is not nowhere dense. there is an open interval I c S. By the definition of rn(z). So S" is open. As in (a). f(z)< -n for z E 7 Obviously f ( z ) = --oo for 2 E n .m(..S).m(2)< f(20) + . c I n . Therefore m ( z )= liminf f ( y ) Y+Z 2 rn(zo).l . (Stanford) Solution.6) such that f(5)< rn(z0) Therefore f: .for 2 E O(z0. we can find 6 > 0 such that m ( z ) > rn(z0).. ? which is a contradiction. (b) Suppose that there is zo E I such that rn(z0) > -m.f(. Suppose that ( X . .6). Al = {Z E I.) < -1). i. . Denote i + i. 6) c I . A1 is a non-empty open set. i. By induction.& & holds for z E O(z0.e. m ( z )= -m for z E I . 2 E S". From (b).d ) to ( X . Take an open interval I1 such that 7 c A l ..6). Since S is closed. Thus f) ( .m(5)< 1.

4512 Let H be a Hilbert space and let f : H function such that f ( z n ) -+ oa whenever llznll minimum. i.. . (ii) Give a n example of a complete metric space M and an isometry A on M . too. then a = lim f ( z n . Prove that f attains a (SVNY. (i) Since A-' = A 2 . y). ~ ( z .z) = (e. + ---t B be a continuous convex 00. defined by A ( z . There is a sequence {zn} of H such that n-co lim f(zn)= a .329 4511 Let M be a metric space with distance function d. (ii) Let M be the 2-dimensional sphere M = { ( z .}F=~ that such k-rco l/znk (1 ---f +m. ~ 1+ z2 = 13. Therefore A is an isometry and therefore continuous. A o A o A = Id and d ( A z . ) = +00. having exactly two fixed points ( 0 .e. (i) Show that A is a continuous bijective isometry. which has exactly two fixed points. suppose A : M + M is . a distance nonincreasing periodic map of order 3. Let (Y = inf f ( H ) . (Stanford) Solution. z). l ) and (0. -1). is a periodic map of order 3. z) E C x 2R The isometry A : M -+M I 1 .*'z. A-' is distance nonincreasing. z) E M . n there is a subsequence {Z~. periodic of order 3.Ay) 5 d ( z . If SUP llznll = $00. Stony Brook) Solution.

Recall that the adjoint A* is the unique bounded linear operater on H such that ( A z . on lR.e. since llA*11 = - SUP IlYll51 llA*~ll = I l Y l l 5 1 114151 SUP SUP I(x7A*y)I SUP SUP I(Az.B. exists in L1(BZ). Then n-+m {fn} converges to f in measure and therefore by F. So sup llzn[l < +m. (You may wish to use (a) prove this. I k > N } . A*y) for all z.11 Therefore and for any y E cov{z. f ( x n k ) < 0.) (Stanford) Solution.330 a contradiction. there is a subsequence { z n k } For any p > a there is an N E lT such that for any k > N . 1 k >N} W = coV{znk I k > N } 11. h Since X E n * cov{z. (a) Show that IIA*ll = IIAll. It follows that a is finite and f attains its mininum at z. ( StunfoTd) Solution. Prove that A is closed in the norm topology of L1(lR). 4514 Let A be a bounded linear operator on Hilbert space H . If {fn} is a sequence of A such that f = lim f. (b) Show that AA* . on aZ. 4513 fzI A is the subset of L1(BZ)consisting of all functions f satistying I ( ) 5 1 a.e.A*A cannot be the identity on H. It follows that I ( ) 5 fzI 1 a.Riesz Theorem there is a subsequence {fnk} converging to f almost everywhere. So A is closed. f ( y ) < . where IlAll is the norm of A.. y ) = (z. ball of H . By the weak compactness of the closed converging weakly to a point z E H . one has f ( z ) 5 p..y E H .y)I llYll51 114151 5 IlAll .. (a) Indeed.

.33 1 and a fortiori IlAll = 11A*\\.A*A = I then = + It follows that I)A))2 JJAJ12 1. 4515 Suppose that A . AA* . Since = m(A)m(B) > 0. we will show that Indeed equalities (1) follow from the following inequalities If for some bounded linear operator A on H . B c 0. Assume without less of generality that A and B are compact sets. (b) First. with m(A) > 0. Show that IR are Lebesgue measurable. a contradiction. (Indiana) Solution. m ( B ) > contains an interval of positive Lebesgue measure.

4516 Let X be a compact Hausdorff topological space and let p be a finite regular Bore1 measure on X . .) p(X\X.. .. Then (1) i=l is for f = f on X. . n-co < i. Is it true that if f : X -+ IR is p-measurable then there exists a sequence {fn}~'l of continuous real-valued functions such that lim f = f a.e. Then the conclusion follows from the following inclusion { Z I m((z .. For any n E l there is a compact subset F.) f is continuous on X. Let X . of X such that p(X\F. [p]? Justify. = 00 u Fi.332 there exists an zo E IRsuch that m((zo-B)nA) > 0. and (3) p (2 each n E lN a real-valued continuous function f : X .B ) n A > 0) is nonempty and open.) < and f is continuous on F. N Yes. Then lim f = f a. Since z H m((z-B)nA) is continuous. the set { z I m((z .e.B ) n A ) > 0) CA + B. . -+ l R such that b]. n-0 (Iowa) Solution.

Part V Complex Analysis .

.

. if false. and continuous on with If(iy)I E 1. z n has radius of convergence exactly R. Then f(z) . (a) if f(z) = u iv is continuous at z = 0. If the assertion is true. (b) False. Rez > 0 } .g ( z ) is not identically zero in R. quote a relevant theorem or reason. = vy and 1-ly = -21. (b) if f(z) is analytic in R and has infinitely many zeros in R . give a counterexample or other justification. Because lim G = it follows from 1. then f'(0) exists. If neither of f and g is identically zero in R . a n-+m . then If(z)I 2 1 ( z E R).2. (e) if C a n z n has radius of convergence exactly R. which is analytic in R. (c) True.and has zeros z = 1 n = 1.--. (d) False. A counterexample is f ( z . . uy. + m. and the zeros have no limit point in R. But f is not identically zero in R. (d) if f(z) is analytic in R = { z . but f'(0) doesn't exist. y ) = Riemann equations a t z = 0. A counterexample is f(z) = e". A counterexample is f(z) = sin&. then f E 0. But f ( z ) is not bounded in R. g ( z ) E 0 in R. vy exist at z = 0 with u. f is analytic in R = { z : IzI < l}. (f) sin f i is an entire function. (e) True.. then either f E 0 or g 0.335 SECTION 1 ANALYTIC AND HARMONIC FUNCTIONS 5101 True-False. a t z = 0. then C n 3 a . continuous on with If(iy)I 5 1 (-co < y < +co). (Indiana-Purdue) Solution. (c) if f and g are analytic in R and f(z) . and the partials u. 6. f satisfies Cauchy(a) False.v. then both f and g have a t most countably many zeros in R.

1+az hence . prove that f is analytic there. (a) It is well known that the Laplacian can be written as [). z = 0 is a branch point of sin &.(’fI and that which implies that f ( z ) is analytic in R. then F ( z ) is analytic in 1. When 1-7iz IzI = 1. What can you conclude if this holds with equality. (Stunf07-d) Solution. (b) Suppose that f is analytic with If(z)I < 1 in IzI < 1 and that f(i-u) = 0 where u is a complex number with 0 < la1 < 1.( < 1). (c) Determine all entire function f that < lf(z)I. If both f ( z ) and z f ( z ) are harmonic in a domain Q. Actually. 5102 (a) Let f(z) be a complex-valued function of a complex variable. sin f i is not analytic at z = 0. Show that If(0)l 5 u2.336 (f) False.

f(2) I I < 1. hence P+iQ=--i- is analytic on G. (Indiana) Solution. ay2 We also have ap . we know that f has no zero in a‘. which implies that fIIl is also an entire function. (a) Show that .+ .i% is an analytic function on G. Take z = 0. where c and c’ are constants and JcI< 1. (a) Let 2 i e Because u is a harmonic function.=aQ -- ay ax axay azay = 0. I $$I + % 5103 Let G be a region in a‘ and suppose u : G -t lR is a harmonic function.’ z-u 1-Tiz -z + a . then for any closed curve c c G. we have ap ax aQ ay - aZu -+ ax2 ~ = 0. we obtain logf(z) = cz d. Hence f (z) = c c’e‘”. au ax au ay aZU azu So P ( x . we obtain If(0)l I bI2. elel which is equivalent to When it holds with equality. (b) Show that u has a harmonic conjugate on G if and only if has a primitive (anti-derivative) on G. we have F ( z ) f ( z ) = a-. the integral 2 i g .)[ 5 1 for IzI < 1.337 which implies that IF(. y) and Q ( x . It follows from < 1 that = c.. Integrating on both sides. (b) If has a primtive.y ) satisfy the Cauchy-Riemann equations. 1+zz (c) From lf’(z)l < lf(z)I.

.au _ . then dv = -dx av + -dy aV ax aY G . if u has a harmonic conjugate v on G. and the integral is taken along any curve connecting zo and z in G .iau Because u and v are harmonic functions. On the contrary.i aY has a primitive sz: k 5104 Suppose that u and v are real valued harmonic functions on a domain R such that u and v satisfy the Cauchy-Riemann equations on a subset S of R which has a limit point in R. ay ax For any closed curve c C = ld(u Hence + iv) = 0.__ ax dy’ -. Hence we can define a single-valued function where zo.ik and f2 = aY aY are analytic functions on a. Because av . fi = . + (Indiana-Purdue) Solution. y) is a harmonic conjugate of u(x. % . Prove that u iv must be analytic on 0. z E G.- av au ay ax’ we know that v(x.338 It follows that -Kdx dU du +z d y =0 holds for any closed curve c c G. 2 . y) on G. The reason lies on the fact that the real and imaginary parts of fi and f2 satisfy the Cauchy-Riemann equations respectively (see 5103 (a)). we have = --dx dU dU + -dy.

Show that f is constant. Suppose that f(z f(z + 1).i a u = if 2 = i ( g .i] 2 0 for z E [0.f ( z ) is real and + i) .339 By the assumption of the problem.1] c a‘ be the unit square.i $ ) bu & when z E S. It follows from fi = i f 2 for z E R that for z E Q. au av _.- ax dy’ du -dy av __ ax when z E S c Q.1]. by the uniqueness theorem of analytic functions. and let f be holomorphic in a neighborhood of Q . (Indiana) Solution. Because f is holomorphic on the closed unit square Q.1] x [0.f ( z ) is real and 2 0 for z E [O. Hence fi = & . 5105 Let Q = [0. we know that f1 = if2 holds for all z E Q. which implies that u + iw is analytic on Q. Because the subset S has a limit point in Q. we have As . by Cauchy integral theorem.

which is holomorphic in a and satisfies ‘ lf(z)I I %${If (. Hence f ( z ) can be analytically extended to a double-periodic function by f(z) = f ( z 5 1 and + 1) = f ( z + i ) . Because G ( z )has no zeros in S. and on sn ({z = 0) u { z = 1)) we have v = 0 and dz = 0.prove that f = 0 everywhere on S. 5106 Let f be continuous on the closure 3 of the unit square s = { z = z + i y E @ ’ : 0 < 2 < l . + f(z)dz = (udz . . Let f = u i w . we obtain that f ( z + i ) = f ( z ) for 0 5 z 5 1and f ( l + y i ) = f ( y i ) for 0 5 y _< 1. Apply the maximum modulus principle to both G ( z ) and l/G(z).vdy) + i(wdz + u d y ) . which implies that G ( z ) is a constant of modulus 1. On 3 n ({y = 0) U {y = 1)) we have u = 0 and dy = 0.Then G(z) is analytic in S and IG(z)I = 1 when z E 85’. For z E we choose the integral path on dS and consider the differential form f(z)dz in the integral.and if If = 0 on S n ({z = 0) U ( 2 = l}). o < y < l}. where the integral is taken along any curve in which has endpoints 0 and z. Then F ( z ) is analytic in S and continuous on S. It follows from G(z) = eF(’) that F ( z ) is also a constant. Let G ( z ) = e F ( * ) . by comparing the real and imaginary parts in the above identity. Hence we obtain Re(f(z)dz) = 0 on dS which implies ReF(z) = 0 when z E dS. Define F ( z ) = f(z)dz.340 for 0 5 z f(1+ Yi) . so l/G(z) is also analytic in S and 11/G(z)l = 1 when z E dS. we obtain IG(z)I 1 for z E S.)I1 < This shows that f ( z ) must be a constant. (Indiana) Solution.f ( Y 4 2 0 for 0 I y 5 1. and let f be analytic on S. If Rf = 0 on sn ({y = O } U {y = l)). Hence f ( z ) = F’(z) G 0. then si as.

A “partial fractions” expression of $ has the form: . z 2 . (b) When IzI 5 1.4 in { z : I z J5 l}. + 5108 Let P ( z ) be a polynomial of degree d with simple roots z1.lz4 + 2 + + z~ 2 4 z2 144 - lz1 3 . we obtain that f ( z ) has no singular point in { z : IzI 5 l}. . . (a) As 241 lim(z . (b) Show that the function f is holomorphic on an open set containing the closed disk { z : IzI 5 1). zd.lz12 - lz1 1 0. and the equalities hold if and only if z = 1. hence f ( z ) is holomorphic on an open set containing { z : IzI 5 I}. &. which shows that z = 1 is the only zero of z 4 + z 3 z2 + z .4 with residue equal t o z Hence when c = f ( z ) is holomorphic in a neighborhood of z = 1.4 22 1 + + 22 + z .4)’ 24 23 23 lim 2-1 4z3 + 3z2 + 2z + 1 1 1 10’ we know that z = 1 is a simple pole of a 4 + a s + 21 + 2 . . . lz4 + z3 + + z2 .+ +1 + z . we have A. By (a). (Iowa) Solution.1) = lim = - 2 4 1 (z4 .41 2 4 .341 5107 (a) Find the constant c such that the function 1 f(z)= z4+z3+22+2-4 -- C z-1 is holomorphic in a neighborhood of z = 1.

342 (a) Give a direct formula for cn in terms of P.) (b) Let Then f ( z ) is analytic o n c and lim f ( z ) = 0. (4 c n = lim z-+.a t z = z. f(z) is I 4 0 0 identically equal to zero.Z1)n.. Then the Laurent expansion of 2 a t z = z1 is P ( Z ) Hence 1 the form: has P(.) . P(.zn - 1 P(z) P’(zn)’ which is the residue of -2. By Liouville’s theorem.zm z .) Solution. (b) Show that 2 really has a representation of the form (*). p. () (c) Give a formula similar to (*) that works when z1 = z2 but all other roots are simple. hence (c) Denote the Taylor expansion of P ( z ) at z = z1(= z2) by P(2)= c 00 a n ( z . (Courant Inst.

If f ( a D ) C IR. It is obvious that g ( z ) is analytic on D and g ( a D ) C lR. and let f be an analytic function i n C \ (a) f j Show that there exist functions j=1 is analytic in C\Kj. then B1 is real and - Az2 + Bz + z ( z . B = -(ZAl + a z l ) + B1(1+ (at2)E lR. show that there is a complex constant A and a real constant B such that Az2 + Bz + z (Indiana) f(.. . By the reflection principle. Suppose g ( z ) B1. Assume K1 = { z . .... Kn be pairwise disjoint disks in C.~1 < + ~ 1 C ) C\ UKj j= 1 n * . ( a .Aizz 1-z z. fn such that (b) f ( z ) = C fj(z) for z €C\ U Kj.343 5109 Suppose f is meromorphic in a neighborhood of D (D = { ( z (< 1)) whose only pole is a simple one a t z = a E D. fz.ZB1.Tiz) * Solution. 5110 Let K1.a ) ( l . K2. Assume that the residue o f f at z = a is Al.a ) ( l . TI < )Z .z l ( 5 rl}. . Define g(z) = f ( z ) - -- Ai z-a -. j=1 n (Indiana) Solution. g ( z ) can be extended to an analytic function on the Riemann sphere hence g ( z ) must be a constant. and j=1 U Kj. n n f1. such that C1 = { Z.) = ‘(z . . Choose c1 > 0 sufficiently small.Tia)’ where A = 2 1 .

such Laurent expansion 0 Set f2(z) = k=-m uk ( z .344 In C1.z1)5 fl(z) is analytic in @\K1.221 5 7-2). Because f(z) K2. fl(z)-f2(z) has an analytic continuation to in @\ f(z)-fl(z)-f2(z) is analytic ij j=3 ~ j . is analytic in@\KZ.f ( z ) has the Laurent expansion f(z) = c +m up(z -.l ( z ) . consisting . f ( z ) .I)? k=-m Set fl(Z) = c 0 k=-oo u p ( z . Choose ~2 n = { z .1 times.~ (2) 2 ) f2(z) ~ . @ u {m}}. Set Then we have n j=1 where fj(z) is analytic in@\Kj. and the above identity holds for z E @\ U Kj.fl(z)fZ(Z) . Iz .f n .fi(z) has an analytic continuation j=2 to K1.fl(z) is analytic in @\ Assume K 2 u Kj. we get a function f ( z ) . > 0 sufficiently small. Because f ( z ) . which is analytic in @\Kn. j=1 n 5111 Recall that a divisor Df of a rational function f(z) on @ is a set { p E of zeros and poles p of f(z) (including the point co). Repeat the above procedure n .* * * .

Stony Brook) Solution. By the property of rational functions. n) be all the zeros and poles of f(z) with multiplicities npirespectively.m) be all the zeros and poles of g ( z ) m with multiplicities mqjrespectively. let q j ( j = 1 . . C npi= 0. First we assume that the point and g can be represented by 00 is not a zero or a pole o f f or g. we have - 1 . . . Let f and g be two rational functions with disjoint divisors.p i ) n p i and m Then n n n m and m m m n . Let pi (i = 1. .345 counted with their multiplicities np E 2. 2 . . then f n f ( z ) = A JJ(z . then we have j=1 mqj= 0.2. It should be noted that p. Prove that (SUNY. i=l n Similarly. is a zero of f when npi> 0 and a pole of f when npi < 0.

+ + . = without loss of generality.346 In case the point 00 is a zero or a pole of f or g. Then n-1 00 f(z) = A n ( z i=l and g(z) m =B n(~ m - 4j)"'j. j=1 Since C mpj = 0. Hence n.) j=1 = g(m) = B . (b) Find the radius of convergence R of the Taylor series T ( z ) of f ( z ) at -4 3i. 5112 Let f ( z ) be the "branch" of logz defined off the negative real axis so that f( 1) = 0. we may assume p . simplifying the (a) Find the Taylor polynomial of f of degree 2 a t -4 coefficients.1 i=l n-1 m and i = l j=1 which completes the proof of the problem. we may assume that g(p. 3i.

arcsin -). we have = l0g[(-4 + 3i) + (Z + 4 .arcsin -) + 5 -4+3i 1 z+4-3i -5 .3il < 5}. Because log z has only z = 0 and z = 00 as its branch points. Hence + + + + + + . and describe the relationship between f and T at such points.Imz < 0 } . 5 4 = -- + 3i 25 ' 25 24i 1250 ' and c2 =- + (b) Denote the Taylor series of f(z) at -4 3i by T ( z ) . while f(z) in R is the continuation of log z a t -4 3i in the slit plane @\(-m.3i( < 5.* Hence the Taylor polynomial of co f of degree 2 at -4 + 3i is + c1(z + 4 .4 + 3 i I ( ) 2 +. Hence R = 5. and has no other singular point.. T ( z ) # f(z). It is because T ( z ) in R is the continuation of logz at -4 3i in the disk { z : Iz 4 .. is this something special to this example. co = log 5 c1 where 3 + i(7r . (Minnesota) Solution.3i)] z+4-3i log( -4 + 3i) + log 1 + [3 z-4+ 4+-3ii 3 log 5 + i(7r . If there are no such points.347 (c) Identify on a picture any points z where T ( z ) converges but T ( z ) # f(z).5. (c) Denote the shaded domain shown in Fig. When z E Cl = { z : Iz 4 .01. or a general impossibility? Explain and/or give examples.3 4 2 .1 by a. the radius of convergence R of T ( z ) is equal to the distance between z = -4 3i and z = 0. (a) When z is in the neighborhood of zo = -4 f ( z ) = logz + 3i.3i) + cg(z + 4 .

2 Solution.41 < 4) so that f ( z ) = z i ( z + 1)i in A and f ( z ) is positive for 0 < 2 < 8.). We have Arargf(z) 1 1 1 + -Ararg(z + 1) = -(27~)+ -2( ..2 by I?. i.l(argf(z)+Arargf(zr)).2 ~ ) 3 2 Hence . and denote the change of $ ( z ) when z goes along r from the start point to the end point by A. Then g(z) = Ig(z)leia%dz) 1 = -Arargz 3 = ~() I.e. An analytic function g in A is obtained from f by analytic continuation along the path starting and ending at z = 4 (see Fig.2).5. Denote the closed path in Fig.5. T(z) = f ( z ) + 27ri. f.$(. Express g in terms of f .1 5113 Let f be the analytic function defined in the disk A = { z : Iz . Fig.348 the difference is 2ai. (Indiana) Fig.5.5.

(a) It is known that z = 0 and z = 00 are the branch points of function &. f i is changed to -fi. 2 . 4x. .) are poles of f ( z ) .l ) ~ ( e -= ~ e-nn). } .e& sin(-&) - e f i . Let r = { z : IzI = T } . sin f i which is still f(z). . .=25 (Indiana) Solution. . -e-fi f(z) = (a) Where is sin& * f single-valued and analytic? (b) Classify the singularities o f f . Hence f ( z ) is single-valued and analytic inC\{z = n2r2 n = 1 .349 5114 Define eJ. . 2 . and when z goes along I’ once in the counterclockwise sense. . When z is in the small neighborhood of z = 0. while f(z) is changed to e . f(z) can be represented by 03 - 2 n =c -zn O c 00 > O n z n n=O (%+I)! which implies that z = 0 is a removable singular point of f ( z ) . : (b) We have which shows that z = n2r2are simple poles of f ( z ) with residues 2 n ~ ( . Hence z = 0 and z = 00 are no longer the branch points of f(z). It is obvious that z = n2R2 (n = 1 .e-J. . (c) Evaluate f(z)dz.6 .

Which of these branches are (is) the derivative of a single-valued holomorphic function in 0. It is ' known that f and g can be represented by and - g(z) - 4- 1 - . (c) f ( z ) has only one pole z = 7r2 in the disk { z : IzI < 25). L < 2 5 5115 Let fl be the plane with the segment (-1 which of the multi-valued functions (a) f(z) = 1 5 z 5 1 .350 As to z = 00.r#J(z)the change of r#J(z)when z goes continuously along I counterclockwise once.-arg(l 2 +z) 0 -27r - 1 -arg(l 2 (-1 (-1 - z)] =0 1 1 Ar[--arg(l + z ) . y = 0) inside I? . Let I? be an arbitrary simple closed curve in a. Hence f(z)dz = 2 ~ i R e s ( fr 2 ) .z ) ] = 2 2 5 z 5 1 . it is the limit point of the poles of f(z).-arg(l . y = 0) deleted. Why? (Indiana-Purdue) Solution. For (b) g ( z ) = can we choose single-valued branches which are holomorphic in fl.{-+log(l+z)-+log(l--l)} Because and 1 Ar[argz . and hence is a non-isolated singular point of f(z). and denote by A. y = 0) not inside r 5 z 5 1 .

1 = 14. Hence loge15 = loge A r l o g z . Justify your answer. z 23 zi It follows that f(z)dz = 0 and lrg(z)dz = f 2 a . 1. (b) Define the lift mapping by w = logz which lifts the unit circle y one-toone onto a segment with length 2ir on the imaginary axis of w-plane.+ . we consider the Laurent expansion of f and g about z = 03: s .*2. . ' Arargz must be 4a. we consider the integrals 1 f(z)dz and 1 g(z)dz. Hence log el5 = 1+ (14 4ai) = 15 4ai.. Find log el5. But neither of the branches of g is the derivative of a single-valued holomorphic function in a. Because e E (e'. If the segment (-1 5 z 5 . (b) Let y denote the unit circle. (a) The set {ee+ie I 6 E IR} U (0) is a spiral which intersects the positive real axis at {e2nn : n = 0.*1. el5 E (e4A. By lifting the integration to a n appropriate covering space. f(z) g(2) = *i(l z - -)-5 22 1 1 = f i (1 + a2 + 7 + * * 7 a4 . i = *-(1- -)-T 22 1 1 1 b3 b5 =fi . sr ( 1 szt + 5116 (a) Let D c C be the complement of the simply connected closed set (eetie 1 6 E R}U (0).351 we have A r f ( z ) = 0 and Apg(z) = 0.. If the segment (-1 5 z 5 1.+ T +. + + + . Hence we obtain that both of the single-valued branches of f are the derivatives of single-valued holomorphic functions in Q. and each of f and g has two single-valued branches. The single-valued branch of logz is defined by loge = 1. e2=). In order t o know which of f and g has a single-valued primitive in R. Let log be a branch of the logarithm on D such that loge = 1. Hence both f ( z ) and g ( z ) have singlevalued branches which are holomorphic in Q. and A r l o g IzI = 15 . Because . where the integral is taken along any curve connecting zo and z in 0.. give a precise meaning to the integral J'(1og ~ ) ~ and find all possible values which can be assigned to it. d z (HUT7NZrd) Solution. where I is a continuous ' curve connecting z = e and z = el5 in D and A r log z is the change of log z when z goes continuously along r from z = e to z = el5.y = 0) is not inside I?. a ) . we know that when I connects e and el5 in D. e6=). oriented counterclockwise. and the primitives are f(z)dz c.y = 0) is inside I?. It follows that A r l o g z = Aplog IzI + iArargz.) . it is obvious that f (z)dz = 0 and Sr g(z)dz = 0.

Pare two real constants. Note that the Laplacian A=---+---=4ax2 ay2 a2 a2 a2 azaz7 and we obtian where t = IzI. + . Because f ( I .47r2) . we have reason to assume that the function f (with real variable t ) has continuous derivatives f’(t) and f”(t). where t can be any real number. has a single-valued harmonic conjugate in C\O. Since log IzI has no single-valued harmonic conjugate inC\O. and the solution is f ( t ) = Q log t + 0. z E C\O.1) have a single valued harmonic conjugate? (Indiana) Solution. it must be a constant. This differential equation is easy to solve. we know that when f(1).(2wew)lit+ Z U ) + 2 1 t i(t+la) = -47~(t+ ?r + i)eit = 47r(t + 7~ + i ) e i ( t + 7 r ) . Hence we have J(W2dz 7 = lt i(t+2*) w2ewdw= (w2ew)lit i(t+2x) - 21t i(t+27r) weWdw ewdw i(t = eit(-47rt . + 5117 Find the most general harmonic function of the form f(lzI). in C\O is Q log IzI 0. which implies that the set of values being assigned to the integral J (log ~ ) ~ d z 7 is a spiral (4-43 i)ezs : s E R}. Hence the most general harmonic function of the form f(1).i(t+ 27r)).352 both the starting point of y and the single-valued branch of log z on y can be arbitrarily chosen. Which of these f(I. where (Y. the segment on w-plane can be denoted by [it. 1 ) is harmonic.

.3 by D . 2 . . By the symmetry of 0 on the rest of the boundary by domain D. and the ten segments of the boundary by I l l I z .5. Denote the interior domain of the pentagram shown in Fig. 110. Let u be the harmonic function in the interior of the pentagram which has boundary values 1 on the two segments shown and 0 on the rest of the boundary.3 Then denote the harmonic function on D with boundary values 1 on l k and u k ( z ) . . . . 10. we have It follows from in .5. k = 1 .. (Stanford) Solution. Fig.353 5118 Consider the regular pentagram centered at the origin in the complex plane. . . What is the value of u at the origin? Justify your claim. put in order of counterclockwise.

does this implies that h is harmonic on G? (Iowa) Solution. q p l ] is harmonic. because z = 0 and z = 1 are branch points of Remark. Y>O lim z=x+yi-+x d m = -4-i. A counterexample is h ( z ) = Red-. The answer is No.354 5119 Suppose G is a region in C. [0.1] c G. dmlz=z lim z=x+yi+z &z-ij=&cT)i. Suppose that u and v are harmonic in D = { z : IzI < 1) and continuously differentiable on D u y. We know that a variation of Cauchy-Riemann equations for f = u + i * and g = v + iv* are u du r-=dr du* ae -- au dU* ae . and h ( z ) is continuous o n C . Let u* be a conjugate harmonic function of u in D and v* be a conjugate harmonic function of v in D. where the single-valued branch of d m is chosen by = 4. is d m .' r -d . 0 5 2 5 1.l] holomorphic. If the boundary values satisfy u = v on y and the radial derivatives satisfy = on y. h l . (Indiana) Solution. Since d m is analytic in C\[O. If the problem is changed to h : G -+ C is continuous and hl~\[o. and h : G + lR is continuous. But h ( z ) is not harmonic on&. then h must be holomorphic on G. When 0 5 2 5 1. 11. h ( z ) is harmonic there. prove that u = v in D. 5120 Let 7 be an arc of the unit circle. Y<O Hence h ( z ) = 0 when z = 2.

Then for z E y. + + 5121 Use conformal mapping to find a harmonic function U ( z ) defined on the unit disc { Iz1 < 1) such that r+l- lim U(reie)= +1 f o r O < d < a -1 for 7r < 8 < 27r. It is easy to know that w = is a conformal mapping of the unit disc D = { z : IzI < 1) onto the upper half plane H = {w : Imw > 0). Hence -s i U ( z ) = u(-i-) z-1 zS1 = -arg(-) T 2 z+l z-1 .355 and It follows from the continuous differentiability of u and v on D U y that u* and v* can be continuously extended to D U y and then are also continuously differentiable on D U y. such that f = g on y. we may assume that u * ( z o ) = v*(zo) = 0. It is well known that u(w) = targw .) Solution.2. = lz :dB = v*(z). Without loss of generality. Give the correct determination of any multiple-valued functions appearing in your answer. . where D* = {z : IzI > 1). Let zo be a fixed point on y. Let F = f . Since F = 0 on y.and for z E y denote the subarc of y from zo to z by y z .1 is a harmonic function in H and assume $1 on the negative real axis and -1 on the positive real axis. Hence we obtain two functions f = u iu* and g = v iv* which are analytic in D and continuous on D Uy. we obtain F 0 on D U y U D*. (Courant Insf. F can be analytically extended to an analytic function on D u y U D*.g. Then by the reflection principle. The boundary correspondence is that the negative real axis {w : --oo < w < 0) corresponds to the arc I’l = { z = eie : 0 < 8 < T ) and the positive real axis {w : 0 < w < +a} corresponds to the arc rz = { z = eie : T < d < 2 ~ } . which implies u = v in D.

then w(z) is the conjugate harmonic function of u ( z ) .dz u. but may be not single-valued. Define + f ( z ) = (u(2) + i. R 2-1 5122 Determine all continuous functions on { z E C : 0 < 1z1 5 l} which are harmonic on { z : 0 < IzI < 1) and which are identically 0 on { z EC : IzI = 1).{2arg(C . Let * d u = -u.. Suppose u ( z ) is a continuous function on (0 < IzI 5 1) which is harmonic on (0 < IzI < 1) and identically zero on { I z I = 1).argc) . (Minnesota) Solution. is a harmonic function in D = { z : I1 < 1) with the boundary values $1 on t and -1 on I'2.where the single-valued branch of arg( is defined by arg( %)lr=o = a. where A is a real number not necessarily zero. Remark.log 2.z ) . Denote w(z) = Jz * d u . This problem can be solved directly from the Poisson formula as follows: 2) 1 = -A..1 A - 2 z+l -arg-2.(Z)) .dy and A = h z l = r * d u . 2R A .

Using Jensen's formula. counted with multiplicity. n..zn is an analytic function in 1.n ) such that ( u j ( 5 t .1. = -bn for n = 0. Consider P ( z ) = e f ( ' ) .g ( z ) ) is identically zero on (1. for n = 1 . and IF(z)I = 1 on (1. = -bn and - n+cc lim = 0. uj .357 then f ( z ) is a single-valued analytic function on (0 identically zero on { IzI = 1). . . 5123 (a) Let f ( z ) be a holomorphic function in the disc IzI 5 T whose zeros in this disc are given by a l l a2. show that . and b-.--too < Iz1 < 1) and Ref(z) is n=-m 5 - u..1 = l}. 2.g ( ' ) which is analytic and does not assume zero in { IzI < l}. let n(t)be the number of (j= 1 .zn be the Laurent expansion of f ( z ) on (0 n-cc < IzI < lim = 0 and lim 0 . Then f ( z ) - g(z) = n=O c. and If(0)l = 1. where b-. 2 . Define g(z) = n=-m C 00 bnzn. + . satisfying 6-.2. . . 5 1. Jensen's formula states that Prove this.( < 1) and Re(f(z) . When IzI = 1. we finally obtain + +cc u ( z ) = Re n=-m b. a . . = a _ . by the maximum and minimum modulus principles. 2 . -. = 1). then . From the above discussion. Let f ( z ) = l}. . it follows from Rebo = 0 and -1 n=-cc 03 03 n=l n=l 00 n=l that Reg(z) = 0. Then g ( z ) is an analytic function on (0 < IzI < +m}. hence f ( z ) = g ( z ) ia. .zn A + -log 21r Iz]. we have F ( z ) i~ e i a . Suppose further that luj I < T for all j = I. . (b) With the hypotheses and notations of (a).

By the mean value theorem of harmonic functions. n(t)we have (b) It is obvious that log 2 = la3 I h:j. By the definition of the function which shows that the identity holds. . (c) Apply the identity in (b). ~ 1 5 r } . which implies that log IF(. we have & Lzx 0<8<2R log Jf(Re2')ldB= R Denote max log If(Re")J by M ( R ) . $. (a) Let then F ( z ) is holomorphic and has no zero in the disc ( 1 . 0<852x (d) What can be said about the zeros of bounded holomorphic functions in the unit disc? (Hamard) Solution.( 5 r } . we obtain n ( r )5 M(R)/log R -.)/ is harmonic in 1.358 (c) For r < R deduce an estimate on n ( r ) in terms of max log If(Reie)l.

359 (d) Let f ( z ) be a bounded holomorphic function in { z : IzI < 1).luj I) is convergent. we obtain n which implies that the series C (1 . We know that f(z) can have countably many zeros. we assume For any n. we Since f(z) is bounded. Suppose z = 0 is a zero of f ( z ) of multiplicity m 2 0 with = a. j=1 00 . ~ 1 = T } . and let the other zeros be ordered by 0 < lull 5 la21 I -. Apply Jensen's formula in (a) to F(z) = with 0 < T < 1 such that there is no zero of f on ( 1 . Obviously Ian1 -+ 1.and hence Let T -+ I.. we can choose T such that T > lun[..

give an explicit map. z + 1 : { z : IzI < 1) z-1 ---f (21 : Imzl < O}. {zi : Imzl w 1 < 0) ---t {z2 : 1.~21< 1 and Imzz > O}.SECTION 2 GEOMETRY OF ANALYTIC FUNCTIONS 5201 Find a one-to-one holomorphic map from the unit disk slit disk {IwI < 1) . (Illinois) Solution. (b) Is it possible to require that f(1) = 0 and f(2) = $? If yes. Then w = 43 0 4 2 o 41(z) = f ( z ) is a o n e . < l} ont ( S U N Y . 1(. (a) = f l ( z ) = z: = . explain why not. Hence w < 5 . 5202 (a) Find a function f that conformally maps the region { z : largzl < 1) one-to-one onto the region {w : 1wI < l}. 43(22) = 22” : {z2 : 1221 < 1 and Imz2 > 0} + {w : Iw( < l}\{w : Imw = 0.1)}.O 5 Rew < 1).b o n e holomorphic map from the unit disk {IzI < l} onto the slit disk {IwI < 1}\{[0. if No. 5 l o g r (log 1 = 0) is a conformal map of { z : largzl < l} onto {< : Re( > 0}.1)). Show that the function you have found satisfies the required conditions. and w = f 2 ( ( ) = is a conformal map of {< : Re< > 0) onto { w : II < l}. Stony B Solution. We construct the map by the following steps: 2 1 = 41(z) = -2.{[0.

~ ): 2 (1) Find one 1-1 onto conformal map f that sends the open quadrant > 0 and y > 0) onto the open lower half disc {(z.0 < b < 1) conformally on the full disk {IwI < 1) by means of a function . . 0 & ( z ) . d E IR. 5204 a Map the disk { Izl < 1) with slits along the segments [a. Let w = 4 2 ( ( ) = (=a <+ dm+(. c .361 is a conformal map of { z : largzl < 1) onto {w : IwI < 1) with f(1) = 0 and f(2) = 2q--1. Then w = F(6)= foY-'(G) is a conformal map 6 of { G : 1 < 1) onto {w : II < 1) with F ( 0 ) = 0. Then w = 4 2 o &(z) = d n + z 2 . 2r+1 -. (1) Let C = 41(z) = z 2 . (2) I f f is an arbitrary conformal map satisfying the condition of (l). d where 1 = . which can be represented by $([) = where a. where d 1 =. It map of {C = < + iq : 7 > 0) onto {w = u + iw : u2 + v2 < 1 and w < 0). s.then 4T1 o f o 4F1(C) is a conformal map of the upper half plane onto itself.' ( w ) 6 1 w is a conformal map of {w : II < 1) onto {W : 11 < l} with F ( 0 ) = 0.y) : x 2 y2 < 1 and y < 0).i a a is z=e 4 2 + iy : 2 > required conformal map. = ._ (b) Suppose W = f ( z ) is an arbitrary conformal map of {z : largzl < 1) onto {W : 11< 1) with f(1) = 0. (2) Find all such f .11. [-1.i is a conformal a. we have both [ F ( G ) 5 1 and I F ( w ) l 5 Iw(. It is a conformal map of { z = 0 and y > 0) onto {C = iv : 77 > 0). ( Toronto) + Solution. 5203 ( ( 2 . -b] (0 < < 1. b.)( = lY(z)I for every z E { z : (argzl < 1). Hence f can be written as 4 2 o II. and W = F(w)= f o f .bc > 0. Since - - - we cannot require that f(2) = - i. ad . ( 6 1 which implies that If(. By w 6 Schwarz's lemma.

f'(0) corresponding to the slits. and 4 2 (2) and it is easy to know that ( 4 2 0 41)'(0) = -(a 1 + 1 + b + -) b < 0.) = -1. ( i ) z 1 = 4 1 ( z ) = z + i : { I z I < 1}\{[a. We construct the conformal mapping by the following steps..a+~]}. > 0) {w : ll < 1). > 0.i..rs+i that 44(i) = 0 and d k ( i ) = -f.. It has the point (iii) z3 = 43(~2) = correspondences fi : C\[O. let A= L-4 &+&' -i. 1 + -+ #1(1)= 2 and &(-1) = -2. Compute f'(0) and the lengths of the arcs (Haruard) Solution. B= 2-4' 6 --t &+4 We also know that +i(-l)= (iv) w = 44(z3) = ~3: { z 3 : Imz3 .1]U[-1. 4 l ( a ) = a +. (ii) z2 = 4 2 ( 4 = the point correspondences z i + ( b + ") -Zl+(aP+): @\{[-6 . > 0).-b]}+~\{[-b-~. + ) It has a . It has the point correspondences #q(O) = 00.. 42(. 41(b) = -b . +m) -+ (z3 : Imz3 and For the convenience of computation. It is obvious w .a + i]} C\[O.362 w = f(z) with f(0) = 0.

we know that f maps the unit disk with slits [-1.E } and let fir be Let 0 < E < a.363 Now we define w = f(z) = 4 4 o 4 3 o 4 2 o & ( z ) .= 4arctg B+z B A+&- 5205 denote the arc {eit : E 5 t 5 27r .arg= 4arctgA = 4arctg A+z A-i T+& B+i B-2 5 4 and I2 = xg- .arg- B-i 1 $4 = 4arctg. let the complement of yE in the Riemann sphere. From the above discussion. > I}.1 conformally 1 onto the unit disk with f(0) = 0 and What correspond to the slits are the arc with endpoints containing point z = -1 and the arc with endpoints and point z = 1.. -b] and [a. (Stanford) Solution. f(0) = 0. f'(0) > 0. The lengths of the two arcs are and containing 2 I1 A-i A+i = arg. We are going to find the map f by the following steps: . describe the part of the unit disc that f maps onto 1(. I f f is the conformal map of the unit disk onto a.

a ) of D. + Fig. (1)Suppose = g ( z ) is another conformal map of G onto a circular disk D. is uniquely determined by G and a. then @ maps { z : IzJ < l} conformally onto D 1 with @(O) = 0.4 Let + ( z ) = $3 o $2 o # q ( z ) . such that f ( a ) = 0 and If'(a)l = 1 for some point a E G. which is C } c. such that g ( a ) = 0 and Ig'(a)l = 1. 5206 Suppose that w = f ( z ) maps a simply conncted region G one-to-one and conformally onto a circular disk DT with center w = 0. radius r . @ ( z ) can be extended t o a conformal map by of fie onto {C : ICI < 1).and the symmetric domain of D1 with respect t o I.( > 1) is D2.which is bounded by circular arcs I. (1) Prove that the radius r = r ( G . Hence the conformal map f in the problem is nothing but the inverse of @. a ) if G is the region between the hyperbola zy = 1 ( z > 0. arg$$(e-T*) = f 5 . where D1 is a domain bounded by { = eie. W(0) > 0. (Indianu ) Solution. > l}. C . is D 2= (1 < 11 l}\n1. and an circular arc I. y > 0) and the positive axes. and the domain f maps onto 1. with center C = 0 and radius T I . Since the symmetric domain of 1.. then w = F ( C )= f o g-'(C) is a conformal map of {C : 1" < r 1 ) onto {w : Iwl < r } + c .5. (2) Determine r ( G . we know that 1.4). with q53(e-+a) = 0. After considering the boundary correspondence.( < 1) with respect to arc { z = eit : It( < E } is 1(.5. orthogonal t o = 1) and connects points e f * and e .364 (shown in Fig.T a in {/<I 5 1). and {C = eie : (01 5 i}. the reflection principle. corresponds t o the arc { z = eit : It(< E } under the map @.. and if a = 1 i. 5 B 5 27r .

if T ( z ) = -.($ + i)l with 43(ie%") 0. then w = f ( z ) : G f ( a ) = 0 . z2 E a' are two distinct fixed points for T . Irj5i(iegx)I = 2 = d%lret=* Define f ( z ) = 4 3 .z 1 z = c-.z2 (b) Use (a) to find an expression for T " ( z ) . 2-3 1 . -.i = 1. Show that there exists a constant c such that 3 T ( z ). with 41(1+ 22 f) = $ + i. (a) Assume that z1. i. 3 . > 01. Apply Schwarz's lemma to F(C) and we have IF'(0)l 5 hence TI 5 T . 11#:(1+ f). Id. T ( z i ) = zi. For the same reason. a ) = * ---f {w : JwI < -&}. we have . 2 .e. l#l z1 = & ( z ) = z2 : G + { q : 0 < Imzl < 2}. with 5207 Let T ( z ) = be a Mobius transformation. n = 1 . In other words. Because the cross ratio is invariant under Mobius transformations. (2) We construct a conformal map of G onto a circular disk D. + (22 : Imzz with 42(: + i) = icy.' ( w ) and we have T 5 T I . (a) Let a E a' be a point different from z1.2.z2 -21 ./ = A... 4 2 o 41(z). = 4 2 ( z 1 ) = eHZ1: ( 2 1 : o < Imzl < 2 ) =fey. . If'(a)I = 1. which implies T I = T . z2.365 with F ( 0 ) = 0 and IF'(0)l = = 1. T is uniquely determined by G and a. apply Schwarz's lemma to F .32 (Iowa) Solution. in the following steps: 2. z T ( z ) . Hence r ( G .

(Humurd) Solution.(2" .2 2 - T " . CT"-l(z) . It follows from T"(z)-l 2-1 = 2"Tn(z)+l z+l that T"(z) = (2" (2" + 1). When -a2 < X < -b2.. the curves form .3 2 we obtain T ( z ). - 2 z -z1 z -zZ (b) Since T " ( z ) = T(T"-l(z)). the curves form a family of hyperbolas.. while when X > -b2. (a) Without loss of generality.366 which is Denoting T ( a ).Y2 b2+X -1 form a family of confocal conics".z1 .z1 = czT"-2(z) . a .(2n . .l ( z ) .1) + 1) .-a . then T ( a ) = 5. hence c = 5-1 : 2-1 5+1 2+1 s.22 T"-2(z) .z 1 T ( z ).1)z' 5208 (a) Justify the statement that "the curves 22 a2+X +-. (c) Show that the transformation 20 = $(z+:) carries straight lines through the origin and circles centered at the origin into a family of confocal conics. (b) Prove that such confocal conics intersect orthogonally.22 z .~ - c.z1 Cn .31 . we obtain that z = f l are two fixed points of T.. if a t all. T ( a ). easy to have it is T"(3) 31 T " ( z ). 2 . Choose a = 2.z1 = c.3 2 .z2 When T ( z )= by solving the equation = z. we assume a > b > 0. = z .

It follow-from and that = O. Noting that the tangent vector of L1 at (xo. where A1 # X 2 . -) 1 . &).yo) is t i = (&. the tangent vector of L2 at (xo.367 a family of ellipses.T sin 8. (b) Suppose (xO.+X2 - u2+Ai Y2 =1 b2+Xi and L 2 : .1.) and which implies that the confocal conics intersect orthogonally. When -a2 <A <4 2 . we have (u2 Y. Suppose the focuses of the conics are ( ~ c ( X ) .) + ( b 2 + A l ) ( b 2 + A. if at all. . *). (c) Let z = Tez'.A () When A = J(a2 + A) + [-(b2 + A)] = m. 4 + Al)(U2 + A. Hence the curves 22 a 2 + A Y2 +-=1 b2+A form a family of confocal conics. > -b2. and w =u + iv = -1z + 1 ( -) 2 Z = 1 -(T 2 1 i + -) cos 8 + -(r 2 T .yo) is the intersection point of L 1 : . O ) .+X-2- a2+ A2 b2+ Y2 A2 .yo) is 7 2 = ($&.

It is easy to know that eYc +1 is a conformal mapping of the domain {C : lReCl < 1) onto the unit disk {w : lwl < 1) with g(0) = 0. and if IRef(z)l < 1 for all z E D(0. the focuses of the hyperbolas are ( f l . the focuses 1 w = z(z + ). 5209 If f : D(0. prove that (Indiana) Solution. Because a(r :)2 . Hence w = F ( z ) = g o f ( z ) is analytic in .1 carries straight lines through the origin and circles centered at the origin into a family of confocal conics.0).:)' of the ellipses are ( f l .The image of straight lines through the origin is which are hyperbolas in w-plane. + = 1.l) = { z : IzJ < 1) --fa' is an analytic function which satisfies f(0) = 0. Because cos2 e + sin2 e = I. l).$ ( r . Hence the transformation 0). The image of circles centered at the origin is which are ellipses in w-plane.

-1 < Imz functions f : R -+a' such that < If1 l}. which implies and the supremum is attained by f ( z ) = e"fo(z). By Schwarz's lemma. and let F be the family of all analytic < 1 on R and f(0) = 0. Find (Indiana) Solution. Because < 1. IF(C>l ICI. For any analytic function w = f ( z ) : fl --+a' such that If1 < 1 and f ( 0 ) = 0.369 D(0.we have eT+1 E . 7r 4 5210 Let R = { z E a'. where 8 is a real number. By Schwarz's lemma.)[ IF'(0)l 5 1. we consider the composite function w = F(C) = f o f{'(C). F(C) is analytic in the unit disk such that IF(C)I < 1 and F ( 0 ) = 0. I Choose (0 = w. The equality holds if and only if F(C) = e"C.l) and satisfies F ( 0 ) = 0 and IF(. we have it follows from f(0) = 0 that If'(0)l I -. is a conformal mapping of fl onto the unit disk with the origin fixed. .

Prove that If’(0)l 5 4. Describe E explicitly. F ( i ) can take any value in the 1+c is a 5 f}. Let R = C\{W = u + iv : u 2 0 and w = 0). we obtain If’(O>l L 4. It follows from 1 f ( z ) I < 1 1 If(z)I that f ( 0 C R. By Schwarz’s lemma. ) Set g(w) = (i f t = 2). IzI < 1) such that f ( 0 ) = -1. which implies I F ( $ ) ]5 It should be noted that when f changes in P. Let f E P and define {f(i) < = F ( z ) = f2(z). on D with g o f ( 0 ) = 0 and 1g o f ( z ) I < 1. (ii) f ( 0 ) = 1 + i.2i + 2i’ f2(z) Then F is a holomorphic function on the unit disc with F ( 0 ) = 0 and IF(z)I < 1. w=-1 I(s O f)’(O)lL 1. s) . 1 < (Indiana) Solution. 5212 Let P be the set of holomorphic function f on the open unit disc so that (i) Both the real and imaginary parts of f ( z ) are positive for IzI < 1.370 5211 Let f be an analytic function on D = { z . (Minnesota) Solution. Because w = 23 1-c (that is the inverse of C = disc {C : i. and suppose that 1 f ( z ) I < 1+ I f ( z ) I whenever J z J 1. By Schwarz’s lemma. Then g o f ( z ) is an analytic function + + + E. Let E = : f E P}. we have IF(z)I 5 121.

find sup If(8)l. If we denote the two roots of p2 . fE3 sup If(8)l = 4 - a. Define w = + ( z ) = 2(5+ :). Then F ( z ) = f o 4 ( z ) = f ( 2 ( 5 is analytic in D and satisfies F ( 0 ) = 0 and IF(z)I 5 1.-r2sin28 8 2 4 . Your answer should be an explicit number. By Schwarz's lemma.$1 5 arcsin $.p~--psind+4<0). 2 5 3 1 l r 1 4 20 E = {f(-) : f E P } = {reio : 1 . p 2 ( 4 ) where pl(q5) p z ( 4 ) and Iq5 .371 conformal mapping of { C : 5 :} onto {w : I .4.&)I 54- a. (Indiana) Solution. the set E can also be represented by < + 5213 Let 3 If F is the family of all analytic function on R such that If1 5 1 in R and lim f ( w ) = 0. W+CC R = {w = u + i w : u -2 + 52 w2 > l}. .-1 < -arcsin .2 5 3 + 4 5 0). f€3 and you should prove your assertion. it is easy to know that w = 4 ( z ) is a conformal map of D = { z : IzI < 1) onto R with d(0) = 00 and 4(4 = 8.y p s i n 4 4 = 0 by PI(+).)I 5 1. This upper bound can be reached if we let f = q5-l which belongs to family 7 and satisfies 4-l(8) = 4 So we obtain a. + 4)) a) IF(.$i w ' l the set : f E P} is equal to {f2(i) 10.. 3 5 $}. Hence If(8)l = IF(4 . we obtain that {w:Iw--z~< Hence -}={w=pe*+: 8 3 lr 4 20 Id--1<arcsin-. r 4 .

Since A = (d . t E [0. f is injective).a ) 2 + 4bc is equal to 4bc + 4a2 < 0. which is equivalent to cz2 + (d . z E 51. L : z ( t ) = zl +t(zz . (Indiana) Solution.. one in the upper-half plane and the other in the lower-half plane. which implies . So f has a unique fixed point inside D. we have ---f Hence > 0 for z E Q. Thus we have a + d = 0 and the inequality a d . Since Ref‘(z) Jif‘(z(t))dt # 0. + bc # 0. Since f is a conformal map of D onto itself. Hence ad-bc > 0 and a2+bc = d2+bcimpies f = id. c.372 5214 Let D be the upper-half and let f # id be a conformal map of D onto itself such that f o f = id. Then (a2 s.a ) z . + bc)z + b(a + d) + d)z + d2 + bc * f(z)= C(U d2 If a + d # 0. which contradicts the condition f # id. we know that f ( z ) = z has two conjugate roots.b = 0. Prove that f has a unique fixed point inside D. 1 3 is the line segment connecting z1 and 22. It follows from f o f = id that b(a + d) = c ( a + d) = 0 and a2 + bc = + 5215 Let fl be a convex. open subset of a‘ and let f : R a‘ be an analytic function satisfying Ref’(z) > 0. then b = c = 0. b. L c il.bc > 0. S t o n y Brook) Solution. Prove that f is one-to-one in 51 (i.bc > 0 can be written as bc a2 < 0. Now we consider the equation f ( z ) = = z . Since R is convex.e. ( S U N Y .z l ) . d E IR and ad . it can be written as f ( z ) = where a. we know that f ( z 1 ) # f ( z 2 ) whenever z1 # 2 2 . Let z1 # z2 be two arbitrary points in R.

(a) Let z1.22 z .n . . It follows that 1 1 P'k) + 1 (logP(z))' = -. .2. Because P ' ( z ) can also be written as nu. we have n + > 1.z 2 ) .-+ . It follows from the univalence of P ( z ) in { IzI < 1) that P'(z) = nu.l z " .. In other words. . the roots of P ' ( z ) are all situated outside the open unit disk. Stony Brook) Solution. . . . + P ( z ) z . By assumption. ~ n-1 j=1 n1 4 lal' 5 1. + 2 u ~ z + u l# 0 for all z E 1. show that the same is true for the roots of dP/dz.2 2 ) ( z . 2 . . IzI < 1 and a1 = 1.-l). . Rezj > 0 ( j = 1. ... (b) For any non-vanishing polynomial P ( z ) . z 2 . Let z1. z z .( < 1). z. then [nu. .z.. 5 ( S U N Y . . and P ( z ) = u ( z . ( z .n). Since a1 = 1.[ 1. (a) If all roots of P ( z ) lie in H .373 5216 Show that if the polynomial P ( z ) = a.. use the result in (a) to show that the convex hull of the roots of P ( z ) contains the roots of d P / d z .) Solution..z1)(z .^^-^ + (n. then Izj 2 1 for j = 1 .-1 be the roots of P ' ( z ) . ( z .zl)(z . by comparing the constant terms. .. let H = { z : Rez > 0). .z. .~ + . not identically zero. (Courant Inst. j=1 zj = al. 5217 Let P ( z ) be a polynomial on the complex plane.).zn un-1zn-' alz ao. is one-to-one in the unit disk + + + n-1 (-l)n-lna. z. . we obtain Iul = n.1.l ) ~ . be the zeros of P ( z ) .z. .21 z .

Re& < C j=1 n & < 0. which shows can not be zero on { z : Rez 5 0).z j ) ) < 0. and I is a directed straight line passing through two zeros zk and zi such that the other zeros are on the right side of I (including on I ) . After considering all the directed straight lines passing through two of the zeros of P ( z ) such that the other zeros are on the right side of the line. (a) Show that there exists on { z : IzI = 1) with < (b) Characterize those functions with (Minnesota) Solution. (a) Let f ( z ) = too n=-cu C bnzn. (b) Let z1. Hence Re < arg(z-zj) < $. . . z 2 . z. Denote the intersectional angle from the positive direction of the imaginary axis to I by 8. 5218 Let f ( z ) be a Laurent series centered a t 0. p(.374 When z E { z : Rez 5 0}. When z is on the left side of I. we have Re{e-ae(z . then 0. then Hence . Hence which shows that the zeros of P ’ ( z ) do not lie on the left side of I . . convergent inC\{O}.) or equivalently. with residue b at z = 0. . we obtain that the zeros of P’(z) lie on the convex hull of the zeros of P ( z ) . be the zeros of P ( z ) .

<-l = ( b .P I = Ib . then which is a contradiction.we obtain f(z) = z EC\{O}. . where D = { z : IzI < 1). 5219 Assume f is analytic in a neighborhood of 0. Hence there exists ( with = 1 such that (b) If max ICI=1 If(<) - = Ib . p.l)ei+(e).f’(z) # 0. f maps D into D .1)holds for all C with I(I = 1. and f maps d D into a l l . Apply the discreteness of zeros for analytic functions to f ( z ) . and hence f ( 0 .11. It follows fiom that which implies that (p(8)= -8. ( a ) Show that Vz E aD. where ( = eie and 4(8) is a continuous real-valued function.c-l = 7 holds on {C : 1[1 = 1). Let f(() . it follows from that If(C) . b-1 p.<-l1 < J b . (b) Show that -&[argf(e”)] > 0 for 8 in R.375 If If() .1 1 holds for all C with I(I = 1.

~ 0 ) 2 + .wo = ( z . where zo E d o . g ( e i 6 ) < 0 implies that R > 1 when r < 1. that Ararg(f(z) . Since R = 1 when r = 1.376 (c) Assume that f(0) = f'(0) = 0 and f l a ~ a two-to-one map from a D is onto dD. which implies that f ( z ) has two zeros (counted by multiplicity) in D.W O ) = nArarg(z . But from (a) it is impossible.zoI = r ) . g(z) = b o + b l ( z . Show that f(z) # 0 whenever 0 < IzI < 1. 2 3 which implies that f(z) assumes values outside the disk when z E I. Let I be an arc in ' defined by I = { z E % : ) z . Let f(z0) = wo E all. It is ' a contradiction to the fact that f maps D into D. Hence we obtain 2 2 7-s Rg (c) Because f l a ~ a tweto-one map from 8 0 onto a D . &Alzl=largf(z) = is 2.z o ) " g ( z ) ( n >_ 2). . Then where n 2 2. Since f(0) = f'(0) = 0.. If g ( e i e )1 0 . and w = f(z) = Rei@. . ' and denote by Ar#(Z) the change of # ( z ) when z goes along the arc I in ' the counterclockwise sense.z o ) + b z ( z . It follows from that Ararg(z .A variation of the Cauchy-Riemann equations for analytic function w = f(z) is Since f maps a D into a l l .zo) > and Ig(z) . Hence f(z) has no zero in (0 < I1 < I}.> K . then = at point ei6. This is also impossible. It is demanded that r is sufficiently small such when z E I?. (b) Let z = rei6. we know that % ( e i 6 ) = 0. . t . (a) Assume f'(z0) = 0. = = = 0. with bo # 0. z = 0 is a zero of f of multiplicity m = 2.bol < f ( z ) . which implies that g ( e i 6 ) = f'(ei0) = 0. Hence f'(z) # 0 for all 2 € aD. ( Indiana-Purdue) Solution. it follows from = aR that %(e i6 ) < 0.. If g ( e i e ) < 0 .ZO) 3T n+ Arargg(z) > .

. Function e e z is analytic in { z : 0 around z = 0 is: < IzI < +co}.l)! 5302 Evaluate where y is the positively oriented circle { Jz1= 1). (Indiana) .-.. + By the residue theorem. ( n.377 SECTION 3 COMPLEX INTEGRATION 5301 Evaluate the integral (Indiana) Solution. and its Laurent expansion The coefficient of the term $ in the above development ~ is 1 1 + 1 + 2! + . we obtain 1 +.= e.

sin z =~ i .. .. Hence we obtain a = 2. Function F ( z ) = ( $ 5 ) c o s z is analytic in { z : 0 < IzI Laurent expansion around z = 0 is: + < +a}.23 + $25 1 - .-...378 Solution..) (: coszdz (Indiana) Solution.)cosz= 2+('-2)P+(. its and F(z) = (. By the .e.2aiRes 5303 For what value of a is the function f ( z )= single-valued? LZ(. 1 1 1 2 t4!z4-. + 5) +.. It is obvious that is analytic in { z : 0 < IzI 5 l}. we have p in the above development ( 3 1O ) .)Z+. = 1 The necessary and sufficient condition for f ( z ) to be single-valued is that the residue of F ( z ) a t z = 0 is zero. I ~ Hence the coefficient of the term residue theorem. and with z = 0 as a pole. The Laurent expansion of around z = 0 can be obtained as follows: & - 1 sin3 z - 1 (z - k. the coefficient of the term in the above development is zero. . i..+. is $. 1 U a ( .

It is easy to see that when t -+ 1. For any r < e .IzI < 2). Show that . dt It follows from let + ztl 2 et . By Weierstrass theorem.rt and the convergence of the integral that is uniformly convergent in any compact subset of { z : IzI < e}.et -- A (t .1)2’ where A = $ cos 1. let IzI 5 T . Consider the integral et + z t dt.et dt. + What is the largest possible P so that h ( z ) is analytic for IzI < P? ( Indiana-Purdue) Solution. Hence the largest possible P is equal t o e. cos(t2) et . which implies that the integral is divergent. 5305 Let f ( z ) be analytic in S = { z €67. Hence P can not be larger than e . we know that h ( z ) is analytic in { z : IzI < e } .379 5304 Define h ( z ) = L m ( l zte-t)-le-t cos(t2)dt. When z = -e. et .

1 ( S U N Y . 1 . Stony Brook) Solution. It is easy to see that It follows from the above three equalities that 5300 'u Suppose that the real-valued function u is harmonic in the disk {IzI is its harmonic conjugate and u(0) = v(0) = 0. Then f ( z ) is analytic in { z : IzI < 2}.380 (Iowa) Solution. Let f ( z ) = u ( z ) iw(z). Show that < 2}. and we have + . dz dz z where y(t) = eZnit.t E [0.

prove that > 0 for all points (Indianu ) .IzI 5 11.l) = { z . and if Ref(z) /2T 0 e-nie[Ref (eie)]dO. w(z) = that - 22 5307 Let f be an analytic fqnction on an open set containing D(0. z E D ( 0 .381 It follows from u ( z )= and f(z) +m 2 -f f (> . l). (a) Prove that d”f(0) = dzn (b) If f(0) = 1.

we have = 2(n!). Hence (b) Because Ref(z) is harmonic on D(0. (a) Assume that 00 k=O we have By Cauchy Integral Formula. kI'" Noting that Ref(eis) Ref(e")de = Ref(0) = 1. 2 0. l). . by the mean-value formula of harmonic functions.382 Solution.

It follows from f ( z >= 00 C anzn n=O that f’(2) = c 00 nunzn-l. (0 < T < 1).14)-1. = 1 f)odz.. f() For n > 1. (Stanford) Solution. we choose T = 1 . where e is the base of natural logarithms. . It is obvious that lull = I’0l I 1 < e . 5309 Let f = u + iw be an entire function. show that the coefficients in the expansion n =O satisfy lunl < e for n 2 1. n=l where nu. : (1 + -)-I n-1 1 < e.383 5308 I f f is analytic in the unit disk and its derivative satisfies IfWI I (1.

Let R be an arbitrary positive number. F ( z ) must be a constant.-(ua(z)-v2(z))-Ziu(L)v(r) Then F ( z ) is an entire function with IF(z>I = e -(. ( Stanford) Solution.'(. we obtain that a. which implies that f ( z ) is a constant. = 0. then f must be a constant. Letting R -+ +m. (b) Let For any integer n > h. B . (a) Let F ( z ) = e .)-. . (b) Show that if If(z)I 5 A+BIzlh for all z € @ w i t h some positive numbers A . 5310 Let f be an entire function that satisfies IRe{f(z)}I 5 lzln for all z . Then it follows from Schwarz's theorem that when IzI < R.f 2 ( z ) = .384 (a) Show that if u 2 ( z ) 2 wz(z) for all z EC."(. where n is a positive integer. which implies that f ( z ) is a polynomial of degree bounded by h. Show that f is a polynomial of degree at most n. 5 A+BR~ R" . h. (Indiana) Solution.)) 5 1. then f ( z ) is a polynomial of degree bounded by h. By Liouville's theorem.

-3R" * 2n + IIm{f(O))l = 3R" If(. First we have the following complex forms of Green's formula: .)l L 1 + IIm{f(O))lr which implies that there exist constants A . Let +B where Hence when k > n. (Iowa) Solution. which shows that f(z) is a polynomial of degree at most n.385 Especially when IzI = +. 5311 Compute the double integral 1J . B such that If(z>I 5 AlzY holds for all z EC. cos z d x d y where D is the disk given by { z = x + i y E a' : x 2 + y 2 < 1).

(Indiana) Solution. (a) B Y n=l we have . 1 sin zd( -) Z 5312 n=O be analytic in D = (1.sin z -dz = R.386 The problem can be solved directly by either one of the above two forms: o r JJDcoszdxdy = sin zd? = . (b) Prove that for z E D. (a) Express A in terms of the coefficients an. < 1) and assume that the integral is finite.

( Stanford) Solution. Let the Laurent expansion of f in {z : 0 < IzI < l} be 00 From . Is 0 a removable singularity? Proof or counterexample.387 Noting that we obtain that . The answer to the problem is Yes. we have 5313 Let f be analytic in (0 < IzI < 1) and in L2 with respect to planar Lebesque measure. 1 00 00 (b) By Cauchy’s inequality.

21r 1' la. while the right side of the inequality is finite. and then . . must be zero when n 5 -1. Otherwise. the left side of the above inequality will tend to infinity. let E -+ +0. 12r2n+1dr< Then a. which leads to a contradiction. Hence n=O which shows that z = 0 is a removable singularity of f . 5314 Evaluate the integral (Indiana) Solution. a = rei+.388 we have Let E < 1 be a small positive number. Let z = peie.

5315 Evaluate by the method of residues. Denote It is obvious that I ( u ) is an analytic function in { u : la1 lr > 1).When r < p. Then we have I(a) = dx 2dx dx . (Columbia) Solution. When r > p.

=l + + 22 - idz 2(2u + l ) z + 1' = - iz. we may assume that ll > 1.=l (z - idz - z1)(z -.27T z2) z1 . It should be noted that branch of is also analytic in { u : la1 > l}.390 Let z = eisl then dx cosx and = = z dz iz ' + 2-1 2 ' = Denote the two roots of z 2 . 1221 < 1.z2 7T 27r d(Zl -t 2 2%& * ~ ~ -) ~ 42122 . and the d m should be chosen by a 5316 r g d w Consider the function (a) Use the residue theorem to find an explicit formula for f(z) = 1 2n g(z. 1 = 0 by z1 and z2. (b) Integrate the Taylor expansion n=O term by term to find the coefficients in the Taylor expansion .2(2u 1). Since z1 -z2 = 1.6)de when ( z ( < 1. By the residue theorem we have zl iz. - 2dm' 0.

zk k=O 00 where Uk 2r = (-1)' sink ode. Then . and the single-valued branch by 4 I r = ~ = 1. Because IC1 . z k . (Courant Inst. C21 = 1.27r].391 (c) Verify directly that (a) and (b) agree when IzI < 1.l ) " C I f . m in 1. k=O (1.1 ) ' s i n k 8 .e-iO sin8 = 22 - C2 -1 2ic ' and where of d ci = $(d=--l). It is easy to obtain that uzn-l = 0 and 2R (c) In order to verify that (a) and (b) agree when IzI function f(z) in (a) into a power series: < 1.e) = C ( . Since the series converges uniformly for all 8 E [0. (b) It follows from I sin81 5 1 and 00 IzI < 1 that g(z. < 1). and f(z)=$' (g(-l)ksink8.( (2 < 1) is defined =i(-dn-l).i0 . (a) Let c = ele. the integration with respect to 8 can be taken term by term.) Solution.y . we develop the f(z) = 2-K ~ 00 - dC-7- 27r(l- z"-i = 2 7 r . Z 2 " n=O .

(UC. ' .5. ' Define f ( z ) = e-" and choose the contour of integration I = r3 u r4 as shown in Fig. show that exists and is independent of a. rl U I'2 U Fig. by Cauchy integral theorem. Irvine) Solution.392 we know that the results in (a) and (b) agree when J z J< 1.5. First we have It follows from the existence of hllRe-x2dz R that exists.5 As f ( z ) is analytic inside I. 5317 If a is real.5.

we have f(z)dz = 2~iRes(f.6. and select the integral contour r as shown in Fig.' ey2+2Ryi = 0.393 -ie-R2 J.6 Let f ( z ) = &. dY it follows from the facts that e-R2 -+ 0 ( R + CQ) and 5318 Let n 2 2 be an integer. Letting R -+ 00. f ( z ) has one simple pole z = eEi inside I'. . Compute (Iowa) Solution.e:*). 0 R Fig . By the residue theorem.5.5.

7 Define f(z)= e .7. 3 j=1 .5. (Minnesota) Solution. B caus is analytic on I' and inside I by Cauchy integral theorem. and choose the contour of integration f(z) = have e-IZ r= 3 rj j=1 as shown in Fig.5.z 2 . Fig.394 and 5319 Evaluate JD- cos (x') dx with full justification. we ' .

we make a change of variable by w = z2. L m ( c o s z2 f(z)dz = I" / 0 e-x2dz + 2 ---f fi when R -+00.O 5 argw 5 -}.sin z2)dz = 2 ' which implies 5320 Evaluate (Iowa) .395 For the integral of f ( z ) on r2. 2 7F By Jordan's lemma. we have For the integral of f ( z ) on r3. we have It is well known that l. then where 72 = { w :lwl= R2. Hence we obtain by letting R co that + sin x2)dz + i 00 & (cos z2 .

8 Define and select the integral contour as shown in Fig.5.5. Because f(z) is analytic inside r.8. where It is easy to see that lim R-r 00 I' iReief (Reie)d6= 0 and . by Cauchy integral theorem.396 Solution. Fig.

Fig. we know that Res(f. ( S U N Y . .397 Since the Laurent expansion of f about z = 0 is 00 where a-1 = -2i. Since f(z)logz is holomorphic inside the contour I?.=-I = T . 2 7r 5321 Let f ( z ) be holomorphic in the unit disk IzI 5 1.f ( z )logzdz = 0.5. Letting E -+ 0 and R + ca. J.5.9 ' Let the contour of integration I be shown as in Fig. Stony Brook) Solution.we obtain dx = -. 0) = -22. The branch of log is chosen to be real for positive z. and the singlevalued branch of logz be chosen by argzl. by Cauchy integral theorem.9. Prove that where respective integration goes along the straight line from 0 to 1 and along the positively oriented unit circle starting from the point z = 1.

(Hamud) Solution. I 2r log la + be'@]& = Re = R e ( 2 ~ l o g a )= 2rlog lal. First we assume la1 > lbl. and then the multi-valued analytic function log(a bz) has single-valued branch on { z : IzI 5 1). 5322 where a and b are complex constants. . then d4 = and + g. and the value of logz at the starting point z = 1 is defined as 0. Take ei@ = z . we obtain where the integration contour IzI = 1 has starting point and end point z = 1.398 where It is easy t o see that Letting E + 0. not both equal to zero.

399 When la1 < lbl. as shown in Fig.5. let b = ae". we have In the case la1 = lbl. () = log( 1 z +z) 7 where the single-valued branch is defined by log(1 z ) lz=o= 0. Because + s . by Cauchy integral theorem.10 In order t o evaluate the integral we define f. f(z)dz = 0. Uy. Choose a contour of integration = r.5. Then Fig. we have J-R I*log 11+ ei@1d4 = Re J-R IR log(1 + e*+)d4 .10. Since f ( z ) is analytic on I' and inside r.

5.5.11. By the residue theorem. The single-valued branch of logz is chosen by argzI.'" Evaluate log la + beiGldq5 = 2 7 max(l0g la[.400 Hence we obtain J. .-1 = ?r.. we have J.log lbl}. f(z)dz = 2?riRes(f. Fig.11 Let and select the integral path I? as shown in Fig. ~ 5323 (Iowa) Solution. -l).

ice"f (&eie)d6. AS E -+ 0 and R + 00.'" iReief(Reie)d6 = 0 and In order to find Res(f.ai.-l). Comparing the imaginary parts on the two sides of the above identity. we obtain . -1) = 27ri + 2 2 .401 where 2X iRegef (Reie)d6 + (logx + 2ai)' dx rO + It is obvious that 1. R-rm lim J. we consider the Laurent expansion of f about z = -1: = where a-1 n=-3 c a n ( z + lyl = 1 . it turns out that -4ai log x + 4a2 dx = 2ai + 2 2 . Hence 2aiRes( f .

5.12. (a) Fig.) Solution.12 Define 1 The single-valued branch for log(z + 1) is chosen by log(z + l ) l z = o = 0.402 5324 Evaluate the following integrals: +im at (a) S-im ( t 2 .4 ) l o g ( r + l ) (the integration is over the imaginary axis). and the contour I of integration is shown in Fig.5. &dz for CY in the range -1 < QI < 2. we have f ( z ) d z = 2 ~ i R e sf(. 2). where and Because . As f ( z ) is analytic on and ' inside J? except a simple pole at z = 2. (b) (Courant Inst.

2 log3 Fig. I F where and Res(f.4)log(z + 1) = -(14 Ti -).13 Define The single-valued branch for zm is chosen by argzIr=z.5. +ice dz ( z 2 . As f ( z ) is analytic on and inside r except . a simple pole at z = e:'..- e f i 3e +a .5. and the contour 'I of integration is shown in Fig.403 and by letting E -+ 0 and R --+ 00.1 - 3e f ( 2 .13. we have f ( z ) d z = 2 ~ i R e s ( f e:')).efa) f ( z ) z+e:' .efa) = lim ( z .m ) i ' . we obtain ./..O = 0. .

by letting E 0 and R obtain dx = 5325 Show that J.5. i ) .14 where . Let ' where (argza)z=z>O = 0. for -1 00 Xu 7r( 1 . f(Reie)iReied8 = 0 2a - when a < 2 and lilil ' -+ f(&eie)iseied8= 0 + co. Fig. By the residue theorem.14. 5.we 7r when a > -1. we have f ( z ) d z = 2aiRes(f(z).a ) ( 1 + x 2 ) 2 d x = 4 cos( y )' < a < 3.404 Because lirn R U C 2z . a # 1. and select the integral path I as shown in Fig. What happens if a = l? Solution.

405 + J. we obtain XQ a(1-a) 2 e 1" 2 .a ) x y d X= 4 cos( y) ' X T(1-a) - 1 5326 (a) Prove that converges if 0 <a < 1 .a(1 . 2 " al -a) ( .O iReief (Re")dB = 0 . and from a > -1 that Letting E + 0 and R --$ 00." iEeie f (&eie)d8 and Res(f(z). It follows from a < 3 that lim RC . I.+ (1 when a = 1.i) = lim [ ( ~ 2'2 2 ta i)2 + 1'-9 - e"i 2 . When a # 1.

406

(b) Use complex integration to show that
?fa x P a cos x d x = sin - . I( -a ? 2

+ 1).
(Hamad)

S o h tion.

(4

1

03

eixx-udx =

(

1

J!

x - cos x d x ~

+

x P a cos x d x )

+i

03

x - sin x d x . ~

It follows from a < 1 that
z - cos x d x ~

is convergent. It is also obvious that

lLA

cos x d x

1 LA I
5 2,
sinxdx

5 2,

xPa is monotonic decreasing and
lim x - =~O X-++03 for
(Y

> 0.
x - sin x d x are ~

By Dirichlet’s criterion, we know that 2 - O cos x d x and e i z x - a d x is convergent when 0 also convergent. Hence

s;”

< a < 1.

(b)

zRD
iC o

c

R

Fig.5.15 Let f(z) = z - a e - z , and the contour of integration r is chosen as shown in Fig.5.15. The single-valued branch of f ( z ) on I? is definde by z-aIz=2>o > 0.

407

By Cauchy integral theorem,

+L
0

i&e"f(&eie)de 0. =

It follows from -a

< 1 that
l i i

L

0

ice'e f (&eie)d8 0, =

and from -a

> 0 and Jordan's lemma that

im l iReief (Reie)de= 0.
a

.* 1

Letting

E -+

0 and R -+

00,

we have

Multiplying both sides by e y i , and comparing the imaginary parts, we obtain

I"

x - cos xdx = sin -I?( ~ 2

T-a

--a

+ 1).

5327
Use a change of contour to show that

dt 7
provided that -a and p are positive. Define the left side as a limit of proper integral and show that the limit exists. (Courant Inst.)

Solution. Since

408

is monotonic with respect to x and
1 lim -= 0, x p

x-++a,

+

the convergence of the integral

dx
follows from Dirichlet's criterion. Define

and choose the contour of integration I' = I'l U I'2 U

I ' 3

as shown in Fig.5.16.

Fig.5.16 By Cauchy integral theorem, we have

=

1

R

--ax

'(

-

x2+pz

+
r

l2 1
f(z)dz-

R

-axi

b

d

x = 0.

It follows from Jordan's lemma that

Letting R --+

00

and considering the real part in the above identity, we obtain

409

5328

(a) Let c be the unit circle in the complex plane, and let f be a continuous a'-valued function on c . Show that

is a holomorphic function of z in the interior of the unit disk. (b) Find a continuous f on c which is not identically zero, but so that the associated function F is identically zero. (Minnesota) Solution. (a) Let zo be an arbitrary point in the unit disk. Then 1 - lzol = p > 0. Choosing 6 > 0 such that 6 < p, we prove that

has a power series expansion in { Iz - zo I 5 6). It is clear that

when Iz - z0I 5 6 and continuous on c . Thus

C E c.

We can also assume

I f (<)I 5
1

M because f is

fo=
c-z

(C - 20) - ( z - zo)

f( 0

-

f 0 -. (

< - zo

1- C - 2 0 z--lo

As

and
n=O

(f)n is convergent, the series C
n=O

00

( ) z "

converges uni-

formly for all

E c . Hence termwise integration is permissible, and we obtain

410

Since zo is arbitrarily chosen in the unit disk, F f ( z ) is holomorphic in { IzI (b) Take f(C) = f = 1). Then

(Icl

< 1).

F f ( z )=
In fact,

/
C

1 -< d

=

(- 1

-

t)

1 dC = ;(2ai

-

2ai) = 0.

can be taken as -for any positive integer n and fixed zo E { z : IzI < 1). When E c ,

f(C)

c

5329

Let [a,b] be a finite interval in B and define, for z in

D = C - [a,bll

f(z) =

J." tA. -z

Show that f(z) is analytic in D. Given c, a < c < b, calculate the limit of f ( z ) as z tends to c from the upper half plane and as z tends to c from the lower half plane. (UC, v i n e ) I Solution. For any zo E D , choose 6 > 0 sufficiently small such that { z : Iz - 201 5 6 } n { z = a : + i y : y = O , a L z I , b ) = 8 . Whenlz-zol < b , a < t < b , w e h a v e 1 1 t-z (t-zo)-(z-zo)
--

-

-. 1
t--0

1 1-- t - % a

00

n=O

and the series converges uniformly for t with a I t ,

5 b.

Hence

41 1

holds for z E { z : Iz -201 < b } , which implies f(z)is analytic in { z : )z - zoJ < 6). Since zo is an arbitrary point in D ,we obtain that f(z) is analytic in D. For z E D, f ( z ) can also be represented explicitly by

where the single-valued branch is defined by arg r = s o > b = 0. Let rl and r be two continuous curves connecting z = xo > b and z = c in the upper 2 half plane and the lower half plane respectively. Then the limit of f ( z ) as z tends t o c from the upper half plane is log - +iAr,arg-

[::I

a-b
z--a

- log

1-

c-b

c--a

+ *i,

while the limit of f ( z ) as z tends to c from the lower half plane is

5330 For each z E U = { z : Imz

> 0)

define

g(z) =

C l ,t-a
1
sin2 t

dt.

Determine which points a E B have the following property: there exist E > 0 and an analytic function f on D ( u , E such that f(z) = g ( z ) for all z E U n )

D(-a,E l .
(Indiana)

Solution. Let I be the half unit circle in the lower half plane whose direction is ' defined from point z = -1 to point z = 1, and define a function

It follows from the Cauchy integral theorem that when z E U ,f ( z ) 5 g ( z ) . With a similar reason as in problem 5328, f ( z ) is analytic in the complement of I?. Hence we obtain that for any a E a, # f l , there exists E > 0 a

412
( E < min{la-ll, la+ll}) such that f ( z ) is analyticin D ( u , E= { z : Iz-al < E } ) and f ( z ) = g ( z ) for all z E U n D(u,E). When a = 3 ~ 1such a f(z) does not exist. The reason is as follows: As ,

--

sin2t - sin2t - sin2z t-z t-z

+-sin2z' t-z

where sin2 t--sina t--x know that

-x

is an analytic function of two variables for ( t ,z ) E a' x a', we
h ( z )= 2 ~ i

' 'J-'

sin2 t - sin2 z dt t-z

is analytic for z E C. But sin2z dt = 2Ti

-Il'

sin2z z -1 dlog(t - z ) = 2Ti log z+l'

which has branch points z = f l , hence g ( z ) can not be analytically continued to D ( & l , & ) .

413

SECTION 4 THE MAXIMUM MODULUS AND ARGUMENT PRINCIPLES
5401

Let u €67, JuJ5 1, and consider the polynomial
U P ( z ) = - + (1 - ("I.> 2

- -z
a 2

2

.
(Indiana)

Show that ( P ( z ) (5 1 whenever

12)

5 1.

Solution.
~ ( z )=

=
When IzI = 1,

+ (1- IaIz)z - a 2 --z 2 2 l a z [ ( l - la12)+ -(- - izz)]. 2 2
U

-

-

U a u a Re( - - Z z ) = Re[- - (Cz)] = Re[- - -1 = 0, z Z z z

IIm(-- - az)I _< 21al.
2

a

Hence when ( z (= 1,
l a lp(z)12 = (1 - Ia12)>" (Im[-( - - ~ i z ) ] ) ~ 2 2 <_ (1 - 2[u(' + la14) [aI2= 1 - luI2 + laI4 5 I.

+

+

By the maximum modulus principle, IP(z)( 5 1 whenever IzJ 5 1.

5402

Let f be holomorphic in the unit disk 1(. < l}, continuous in 1.( 5 1) and If(.)/ = 1 whenever (zJ= 1. Prove that f is a rational function. ( S U N Y , Stony Brook)

414

Solution. If f ( z ) has infinite many zeros, by the isolatedness of the zeros of holomorphic functions, the zeros must have limit points on the boundary of the unit disk. But it will violate the fact that f is continuous in ( 1 . ~ 1 5 1) and If ( . ) / =1 whenever IzI = 1. Hence f has only finite zeros in the unit disk. Denote all these zeros by zlr 22,. . . ,z,, multiple zeros being repeated, and define

Then F ( z ) is holomorphic in 1.( < 1). continuous in 1(. 5 1) and IF(z)I = 1 when I z J = 1. By the maximum modulus principle, IF(z)I 5 1 in (1. 1 5 1). Since F ( z ) has no zero in (IzI 5 1) -L- is also holomorphic in 1.( < l}, F(z)

= 1 when IzI = 1. Application of the yields IF(z)I 2 1 in 1(. 5 1). Hence maximum modulus principle to IF(z)I = 1 holds in ( 1 . ~ 1 5 l}, which implies F ( z ) = eia with Q: a real number. So we obtain
continuous in 1) and

(1.~1 5

I&JI

&

5403

Let f be a continuous function on = { z : IzI 5 1) such that f is analytic in U. If f = 1 on the half-circle y = (eie : 0 5 8 5 a}, prove that f = 1 everywhere in (Indiana) Solution. Define F ( z ) = (f(z) - l ) ( f ( - z ) - l), then F ( z ) is also continuous on V and analytic in U. When z E d U , we have either f ( z ) - 1 = 0 or f(-z) - 1 = 0. Hence F ( z ) = 0 holds for all z E which implies either f ( z ) - 1 0 or 0. Since f(z) - 1 0 is equivalent to f(-z) - 1 0, we obtain f(-z) - 1 f ( z ) 1 for all z E V . Remark. The condition that “f = 1 on the half-circle y” can be weakened t o that “f = 1 on an arc y = (eie : 0 5 8 5 f } , where n is a natural number”. In this case, the proof is the same except that F ( z ) is defined by

v

v.

v,

~ ( z = ( f ( z > - l)(j(ze:z) )

-

l ) ( f ( z e + i ) - ~ ) . . . ( f ( z e ? ~ a ) - 1).

2n- 1

415

5404

Let S denote the sector in the complex plane given by S = { z : -5 < argz < $}. Let 3 denote the closure of S. Let f be a continuous complex function on 3 which is holomorphic in S. Suppose further (1) If(z)I 5 1 for all z in the boundary of S; (2) ~ f ( z i y ) ~ e f i for all z i y E S. 5 Prove that If(z)J 5 1 for all z E S. (SVNY, Stony Brook) Solution. Let F ( z ) = e - ' " f ( z ) , where E > 0 is an arbitrary fixed number. Then F ( z ) is also continuous on 3 and analytic in S. When z is on the boundary of S, J F ( z ) I = e-'"If(z)I 5 1. When IzI -+ $00 < argz < %), IF(z)I 5 e-€" . e f i + 0. By the maximum modulus principle, we have IF(z)I 5 1 for all z E S, which implies If(z)I 5 lea'[ = ear for all z E S. Because E > 0 can be arbitrarily chosen, letting E -+ 0, we obtain lf(z)I 5 1 for all z E S.

+

+

(-4

5405

Let K be a compact, connected subset of@ containing more than one point and let f be a one-to-one conformal map of@\K onto A = { z : J z J 1) with < f(00) = 0. If p is a polynomial of degree n for which Ip(z)I 5 1 for z E K , prove that M z ) I 5 If(z)I-" for z EG\K.

(Indiana) Solution. Because f is a one-to-one conformal map of@\K onto A with f(w) = 0, it has a simple zero a t z = 00. Since p is a polynomial of degree n, it has a pole of order n a t z = M. Hence the function F ( z ) = p(z)fn(z) is analytic in G\K which contains point z = 00. As f ( z ) maps@\K onto A = { z : IzJ< l}, we have lim lf(z)J = 1. Together with Ip(z)I 5 1 for z E K, we know that the
%-+K

limit of IF(.)\ when z tends to K can not be larger than 1. Apply the maximum modulus principle to F ( z ) o n @ \ K , we obtain IF(z)I 5 1 for z E G \ K , which implies Ip(z)I 5 If(z)I-" for all z E@\K.

416

5406

Suppose f and g (non-constant functions) are analytic in a region G and continuous on the closure of the region. Assume that E is compact. Prove that ( g ( achieves its maximum value on the boundary of G. (Iowa) Solution. Assume that 191 achieves its maximum value c ( c > 0) at zo E we prove that if zo E G , then f and g must be constants. Let If(zo)I = f(zo)e*+’, Ig(zo)I = g(zo)ei+2.

If1 +

If1 +

z,

Then for fixed

4 1

and

42,

~ ( z = f(z)ei+l )

+ g(z)e”Z

is analytic in G and continuous on

G. It follows from
= ~f(zo)I

IF(.)I
~ ( 2 0 )

5 If(.>I+ Is(z)I L c ,
= f ( z o ) e i + l + g(zO)ei”
~ ( z= f ( z ) e i + l )

+ Ig(zo)I = c

and zo E G that

+ g(z)ei+2

must be the constant c . Without loss of generality, we assume that f is not a constant, and try to lead to a contradiction. Since the image of an open set { z : Iz - 201 < 6) c G under f is an open set which contains point f(zo), f(z) assumes all the values f ( z ) = f(z0) &ei+ for small E > 0 and 0 5 4 < 2 r in { z : Iz - z0I < 6). Then 7 when 4 4 # 0, ?r, we have 1

+

+

lf(.)I + lg(z)l

= =
-

lf(.)I + Ic - f ( z ) e i 9
lf(z0)

+

+ ~c- f(zo)ea+l

- &ei(+++l)l

l,cei(++h) + f(zo)e”l f(zo)ei+l

I + l&ei(+++l) - g(zo)ei@21

>

+ g(zO)ei+Z = c,
If1 + lgl.
Hence f must

which contradicts that zo is a maximum value point of be a constant, which also implies g is a constant too.

417

5407

Suppose f(z) is an entire function with

Show that f(z) is identically 0.
(Iowa)

Solution. For any R > 0, consider function
g(z)

= (Z - Ri>(z R i ) f ( z ) .

+

When IzI = R, and Imz 2 0, denote by 0 the angle between the line perpendicular to the imaginary axis and the line passing through z and Ri. Then 0 5 0 5 5, and

When Iz( = R, and Imz < 0, denote by 8 the angle between the line perpendicular to the imaginary axis and the line passing through z and -Ri. Then 050< and

t,

It follows from the above discussion that when J I R, z=

By the maximum modulus principle, when IzI < R,

Now fixing z, and letting R -+ +m, we obtain f(z) = 0. Since R can be arbitrarily large, we have f(z) = 0 for all t EC.

418

5408

Suppose f is analytic on

(2;

0 < IzI

< 1) and

Show that f E 0.
(Indiana)

Solution. Denote the Laurent expansion o f f on { z ; 0

< IzI < 1) by

where

It follows that

When n < 0, letting r + 0, we have
a, = O

( n = -1,-2,...),

which implies z = 0 is a removable singularity of f . In other words, f can be extended to an analytic function of the unit disk. = 0 when IzI = 1. By the maximum modulus principle, we Since log 121 obtain f 0.

5409

Let f be an analytic function on D = { z : IzI < l},f ( D ) E D and f(0) = 0. (a) Prove that f ( - z ) I 5 2)zI2 for all z in D and if equality occurs for some non-zero z in D , then f(z) = eiaz2.

).(fI

+

we have + I I f(z) + f ( . Because If(z)I < 1 for z E D . If equality occurs for some non-zero z in D .419 (b) Prove that (Indiana) Solution.z ) = 2eioz2. and when z tends to dD. is a real constant.) f(--z)I 5 21zI2 holds for all z E D . the other coefficients must be zero. a4 = a6 = . where a. which implies f ( z ) = e i a z 2 . By the maximum modulus principle.. we have Since a2 = e i a . limit of the Hence can not be larger than 2. then F(O) = 0.z ) = 2eiaz2 = eio. is analytic in D . = 0. Let M n=l it follows from that a2 f ( z >+ f ( . (a) Let F ( z ) = f(z) + f(-z).. If(. (b) .

Let R be a sufficiently large real number such that R curve I’ on the right half-plane. ?r Take a closed ?r I’ = { z = t + i y : t = 0. (Iowa) Solution. X > 1. If the answer is yes. Since G ( z )has exactly one zero inside I. Because R can be arbitrarily large. prove that f ( z ) has a fixed (Rutgers) Solution. . When IzI = 1.Z. it follows from RouchC’s theorem that ’ F ( z ) has exactly one zero inside I?. is the point necessarily real? Justify. where > 2A. IF(z) .e-” and G ( z )= X When z E I?.G(z)l = le-”l 5 1 < IG(z)I. Hence f(z) takes value X exactly once in the right half plane. Prove or disprove: f ( z ) takes the value X exactly once in the right half-plane.z .420 5410 I f f is analytic and lf(z)I point. 2 2 Define F ( z ) = X .G ( z ) /= lf(z)I < 1 = IG(z)l. 5411 Let f ( z ) = z + e .< argz 5 -}. Let F ( z ) = f(z) . < 1 on { z : IzI 5 l}. Since G ( z )has only one simple zero in { z : IzI < 1))we conclude that f ( z ) . . which implies that f ( z ) has a fixed point in { z : IzI < 1). By RouchC’s theorem) F ( z ) and G ( z ) have the same number of zeros in { z : IzI < 1). IF(z) . -R 5 y 5 R} u { z : Izl = R.’ . -. F ( z ) has exactly one zero in the right half-plane.z has one zero in { z : IzI < l}.z and G ( z )= -z.

. Suppose that and Find the location of the zeros o f f in the open unit disc {z : II z < 1). Assume f is non-zero on the unit circle {z : II = 1). are the zeros of f ( z ) in {z : II . z being repeated. such that F ( z 0 ) = 0. We have F ( x ) = X . where g(z) is analytic and has no zero in {z : Il z 5 1).42 1 Take z = x 2 0.e-". x-++w lim F ( z ) = -m.x . Then < l). multiple zeros f(z) = !dz) j=1 nc. (Iowa) Solution. 0 < xo < $00. the point z in the right half-plane such that f ( z ) = X is necessarily real. there must exist 20. which is a real-valued function of real variable x. 5412 Suppose f is analytic in a region which contains the closed unit disc {z : ( z ( 5 1). Let C denote z the unit circle traversed in the counterclockwise sense.Zj). We have . Since F ( x ) is continuous and F(O) > 0. In other words.z.z2. Assume z1. +. n .

.y = 0) U { z : II = R. i. z1.422 It follows from (1) that n = 2.2 = ki. Hence z = 4 ~ 2 the only zeros of f(z) in the are 5413 (a) How many roots does this equation z4+z+5=0 have in the first quadrant? (b) How many of them have argument between f and $? (Indiana-Purdue) Solution.z1)(z . Then for f(z) = ( z . lr Choose a closed curve ' = I { z = z +iy.z z ) g ( z ) . U{z=z+iy:z=O. and which show that unit disc.e.O<yL . (a) Let R be sufficiently large such that when IzI = R. f ( z ) has two zeros in the unit disk.O 5 z 5 R.O 5 argz 5 -} z 2 R}. Set f ( z ) = z4 +z +5 and g(z) = z4 + 5.

mgf(z) denotes the change of argf(z) when z goes continuously from Re<' to R along I3 It is obvious that i '. We also know that Ar. the numbers of the zeros of f and g inside I' are equal. f has also one zero inside I Noting that R can be arbitrarily large. i) f (a$*) E {w : ?r -E < argw < where E > 0 is very small. . we know that ' .argf(z) = Ar. where Ar. By RouchC's theorem.< argz 4 - A 5 -}.g(z)I < Is(z. Since g has only one zero inside I?.argz4 = A. 2 A It is easy t o see that Imf(z) > 0 when f(0) = 5. Set f(z)=z4+z+5 and 9is approximately and r3 = { z : IzI = R.) .arg (1 + F) . Ap. z4 z + +5 = 0 has one root in the first quadrant. and f ( ~ E {w : o < argw < E } .)l holds when z E I?. (b) Let R be sufficiently large such that when la1 = R.argz4 + Ar.423 It is obvious that If(. zero. while .

2 x 5414 Prove that the equation sin z = z has infinitely many solutions in C. Let f(z) = sinz . then f ( z ) can be written as For any fixed natural number n. Let I' = I'l U r2 U r is taken once counterclockwise. the number of the roots of f ( z ) = 0 inside equal to is Hence has exactly one root in the domain {z :- f(z) = z4 z + 5 = 0 x 4 + < argz < -}.424 is very small. choose a positive number t >> logn and a closed contour I' = rl U I'2 U r3 U r4 in the counterclockwise sense. it follows 3 from the above discussion that By the argument principle. where .z and z = z + i y . (Indiana) Solution.

Assume that f ( z ) = sinz . f(rz) {w = u + i v = with the direction upwards. This problem can also be proved by Hadamard’s theorem. . with the direction downwards.2 7 ~ v ) 0) ~= with the direction from the right to the left. Since n is arbitrarily chosen. By Hadamard’s theorem. .(et 2 . we conclude that sinz = z has infinitely many solutions in c.425 Then we consider the image of r under w = f (2): f(rl) {w = u + i~ = : -2(n +1 ) 5~u 5 . : t = -2(n+ i f(I’3) ~ ) T . f ( z ) = sinz . and denote all the zeros by z1.t ) lies in the annulus starting from w = -2(n + 1)n + i in the counterclockwise sense.z has one zero inside the contour I?. z z .z has only finite zeros in (X. * . Hence the winding number of f(r)around w = 0 is 1.e-t ) .zg) and g( z) is a polynomial. o 5w 5 1 . By the argument principle. multiple zeros being repeated. Remark.z n . f ( z ) can be written as f(z) = e g ( z ) k ) ) P where P(Z> n = k=l .

or both f(r1) and f ( r 2 ) are {w : IwI = 5) in the clockwise sense. we obtain that Imu < 0 in the former case and that Imu > 0 in the latter case.z = e a r + b p ( z ) . . (b) I f f is meromorphic in the annulus. f has no zero in D. (a) Let D = { z : 1 < I1 < 2) and d D = rlU r2. In the latter case. By letting y + +oo and y -+ -oo respectively. and I'2 = { z : IzI = 1) is in the clockwise sense. The following is a counterexample. In other words.where rl = { z : IzI = 2) t is in the counterclockwise sense. Let z = z+iy and z be fixed.we know that both f(I'1) and f(rz) must be {w : lwl = 5) in the counterclockwise sense. (b) If f is meromorphic in D.426 It is obvious that f ( z ) is an entire function of order X = 1. which shows by the argument principle that f has at least two zeros in D. is the statement in part (a) still true? (Stanfod) Solution. Because f is non-constant analytic in D and If1 = 5 when z E dD. This contradiction implies that sin z = z has infinite many solutions in C.argf(z) 2 1. Let g(C) be a conformal map of {(=<+irl:$+$<l. Show that f has at least two zeros. where T-03 which implies that g ( z ) must be a polynomial of degree 1. Hence we have sin z . and comparing the increasing order on both sides.argf(z) 2 1 and &Ar. It might occur that f(rl) and f(r2)are two subarcs of {w : IwI = 5). Hence &Ap. the statement in (a) is not true. 5415 (a) Let f be a non-constant analytic function in the annulus (1 < IzI < 2) and suppose that If I = 5 on the boundary.

Then is a non-constant meromorphic function in D with f has no zero in D. where rl = { Z = TP: o 5 8 L ) .427 onto {w : IwI > 5) with the normalization g(0) = M.. e . and aZn # 0. Suppose that there is no real number x such that P ( z ) = 0. ( a 2 n ~ 2 n ( (ao+alz+.. and suppose that Prove that P has exactly TI roots (counted with multiplicity) in the open upper half plane { z E 67 : Imz > 0). and r2= { z = X + i y : -T 5 T.y = 01. But Let n be a positive integer.+azn_1z2"-11. 5416 I f I = 5 when t E LID. > Take a closed contour I? = I'l U I'2 in the counterclockwise sense. Then the number of zeros of P ( z ) inside I is equal to ' It is already known that . (Indiana) Solution. 0 where each aj €67. Let T > 0 be sufficiently large such that when IzI = T . g'(0) > 0. and let P be a polynomial of exact degree 2n: P ( Z )= a0 + alz + a2z2 + + aZntzn.

+--.arg 2a (u2nz2n) =n we obtain that P has exactly n roots (counted with multiplicity) in the open upper half plane. * . (a) Let 1 F(C) = f(-) = 1 + c c 1 1 1 + 3c2+ .+gcn. 5417 Consider the function 1 1 1 f(z) = 1 + .) Solution.+ -- z 2!z2 1 1 + . From . .428 We also have d log P(z) = G A r .e3+ * .zn n! (a) What does the integral count? (b) What is the value of the integral for large n and fixed r? (c) What does this tell you about the zeros of f ( z ) for large n? (Courant Inst. argP(z) 1 Note that and 1 -Ap.

when n is sufficiently large. When n is sufficiently large. which is just the number of zeros of f ( z ) in 1(. and that there is a unique. (c) From the above discussion. . F(C) converges to ec uniformly in any compact subset of@. we obtain {/(I < :} for fixed T and large n. simple solution z1 of the equation f(z) = w in 1(. there is no zero of f ( z ) in { IzI > T } . Since ec has no zero in a'.ec I < m I 1ecI for ICI = : which implies the numbers of zeros for F(C) and e in . ICl=? then m > 0. < R}. (b) When n -+ 00. if the integer n is sufficiently large.429 we know that the negative of 11 represents the number of zeros of F(C) in (1 < :}. all the n zeros of f ( z ) are in { IzI 5 T } . Let min leCl = m. lF(C) . 5418 (a) Suppose that f ( z ) is analytic in the closed disk IzI 5 R. C are equal. the equation has exactly one solution with ( z ( < 2. Show that this solution is given by the formula W (b) Show that. > T}. In other words. we conclude that for any fixed T > 0.

< 2). Ifn(z) .) 1 1 Solution.= -.w = ( z . (a) Let f ( z ) . When n is sufficiently large. z > 0 sufficiently small such that when E re. the equation z = 1 + (5)” has exactly one solution (denoted by z p ’ ) in 1(.z ~ ) l o g & ( ~ ) ]= .(z) is a continuous real-valued function for 1 5 2 5 $. Then f 1 f’(z) = [log(f(z) . we choose E = 2 . (c) f.w)]’ [log(z .1). we have fn (1) < 0 and fn > 0. E Hence fn(z) and g ( z ) have the same number of zeros in { IzI < 2 .zi)&(z).E L Ig(z>I. It follows from (z) that .430 (c) If z1 is the solution in (b).g ( z ) l = 11 5 z n = (1 - 5)” < 1 . and the number is 1. n-cc 2 (Courant Inst.& ’ ( z ) = ’ -w z-zi Q(z)* + +- Hence (b) Let fn(z) = z .E } . Since E can be arbitrarily small. Hence we have z?’ E (1.1 and rE = For fixed large n.€1. show that lim (zl. $). where Q ( z ) is analytic and has no zero in { IzI < R}.

(Indiana) Solution. If. for example. Since 7 . and the directions of 7 and 7 2 are both in the counter1 clockwise sense.ny2 is not homologous to zero (mod R). Now we claim that the case that = f(-. Since f is analytic in R and bounded by lf(z)I < 1.43 1 which implies 5419 Let 1 1 R = D(O. which is a contradiction. Let where E > 0 is small. . 1 1 1 f(-1 = f(--) = 5. because otherwise. f * yl and 1 f *y2 are also not homologous to zero (mod R). 2 2 we assume that z = f is a zero of f(z) .) will not happen. Thus we obtain either f(-) 1 1 =2 2' f(--) = -- 1 2 1 2 . As E tends to zero.l)\{-. Hence we obtain f(*-) = f2. --}.$ of order m. f(rl) and 1 f(y2) will tend to either 20 = 5 or 2u = -.f of order n and z = -f is a zero of f(z) . f * 7 or f * y2 1 will be homologous to zero (mod R). then f * y is not homologous to zero (mod R).nr2) 1 2 1 f(i) is homologous to zero (mod Q). then f * (my1 . 2 2 Find all analytic functions f : S2 + R with the following property: if y is any cycle in R which is not homologous to zero (mod R). the points z = =kf must be the removable singularities o f f .7 2 are not homologous to zero (mod R). while my1 .

Thus we conclude that the functions which satisfy the requirements of the problem are f(z) = z and f(z) = -2. which implies that f ( z ) = z. 2 2 1 1 we consider the function F(z)= fk).) - -- z . 1 2 f(--) = -.t z 1 which is analytic in D(0.l) and satisfies IG(z)I 5 1. It follows from G ( . = i)= 1 we consider the function f(z)+i G(z) = 1 + if(.51 1 1.432 or f(-) 1 2 = --.i ) = -1 that G(z) -1.t . = . which implies that f(z) = -z.l) and satisfies IF(z)I 5 1.i f ( z ) -2 1.zz which is also analytic in D(0. It follows from F ( that F ( z ) 1. 1.

.433 SECTION 5 SERIES AND NORMAL FAMILIES 5501 Let 00 nO = have a radius of convergence T and let the function f ( z ) t o which it converges t have exactly one singular point zo.zo 00 is convergent a t z = zo. which implies It follows that . Prove that lim an/an+l = zo. n-+w (Indiana) Solution. which is a simple pole. Hence the power series F(z) = n=O A Cantn. on I1 = T . Assume that the residue of f ( z ) at zo is A . the Taylor expansion of F ( z ) a t z = 0 has a radius of convergence larger than T . In other words. and define F(z)= f ( z )- A -.zo z Then F ( z ) is analytic on { z : 1zJ5 T } .z .

-c. (2) Show that this series converges to log(1 . t E ( 0 . Differentiating term by term. (2) Let f(z) = - n>l c n> 1 - zn n 1. n-+co % + I an 5502 (1) Show that the series n> 1 is convergent for 1 # a E C with la/ = 1. which shows that . by Dirichlet’s criterion we know that converge. 2 ~ we have ) and Because both C for 1 # a n>l tends to zero monotonically.( < 1). 2 ~ )then .=-c an n>l n>l cos nt + i sin nt n For t E ( 0 . we have -1 1-Z .a ) for such a.C $ is convergent and C n>l n>l E C with la1 = 1.434 and and we obtain lim -= zo. (1) Let a = eit. (Minnesota) Solution.

e i t ) = log(1 .. z = 1 is a regular point of f. First of all. Prove that the boundary of the disk is the natural boundary of the function. for (a1 < 1. .435 Integrating both sides on the above identity.C -n an n> = = -C. l ) there exists a small real number 6 > 0 such that the power series expansion of f at point 3 is convergent at z = 1 6. Assume the proposition is false.z ) . i. Suppose the series : is + .e.a). then for fixed 2 E ( 0 . we prove the following proposition: If the radius of convergence of n =O is equal to 1 and an 2 0 for all n. z = Teat where 0 < T < 1. we obtain f(z) = log(1. 5503 Consider a power series O01 C -nz n . I n= 1 Show that the series converges to a holomorphic function on the open unit disk centered at origin. then a = 1 is a singular point of f(z). Let Q = eit.= n> 1 r+l- eint lim-Cr-+1n>l (Teit)n 1 n lim log(1- T e i t ) = log(1. It follows from Abel's limit theorem that . 0 < t < 2a. ( Columbia) Solution.

By the above proposition.Ic + 1 )an(z .". c .. n=O which contradicts the statement that the radius of convergence of n=O is equal to 1. &I n. Noting that when z = 1 6 the right side in the above identity is a convergent double series with positive terms. .k + k=On=k is convergent at z = 1 6 . a ( .436 Thus 00 w w x b k ( z -z)k = k=O n(n . we assert that when z = 1 6 . Now we return to the power series It follows from that the radius of convergence of " 1 -zn! n n=l is equal to 1. z = 1 is a singular point of F ( z ) . and hence the order of summation can be changed. .1) . + + n=O W k=O . For any natural numbers p and q .z ) k z n .

the boundary of the unit disk is also the natural boundary of the function c C 00 $zn! although the series is absolutely n=l and uniformly convergent on the closure of the unit disk. f n ( z ) = einaz. Since the set { e . (Indiana-Purdue) Solution.and if If(z)I = 1.437 Since z = 1 is a singular point of " 1 I -zn. 5504 z E Suppose f is analytic in U = 1(. In the case when f ( z ) = eiaz. By Schwarz's lemma... we have [). Then . prove that either g ( z ) = 0 or g ( z ) = z . z = e-?4..2.(fI for all z E U . Since f n ( z ) is convergent. By the above discussion. In other cases.p r i : p .i is a singular aq point of F ( z ) . In other words. we conclude that the unit circle {IzI = 1) is the natural boundary of F ( z ) .1 for some z # 0. q = 1. then f(z) = einz where a is a real number. which implies that f ( z ) = z and g ( z ) = t. we obtain a = 0. < 1) with f(0) = 0 and If(z)I U . If the sequence { f n } is defined by composition < 1 for all and f n ( z >4 g(z) for all z E U .. it follows from f(0) = 0 and If(z)l < 1 that 5 IzI for all z E U . Remark.} is dense on {It1= l}. Let 0 < r < 1. n n=p it is also a singular point of F ( z e 9 " ' ) .

z0I < T } .f(z>I< m 5 lf(z>I holds on {z : Iz . fn must have a zero in { z : Iz . Ifn(z>. we know that for sufficiently large n. As {fn} converges to f(z) uniformly on compact subsets of D . such that f(z) # 0 when z E {Z : 0 < IZ 5 T } C D. then either f is identically zero in D or f(z) # 0 for all z E D. we know from Weierstrass’ theorem that f is analytic on D . Hence fn(z) converges to zero uniformly in-{lzl 5 arbitrarily chosen. there exists T > 0. Since zo is a zero of f ..z0I = T } . . Since 0 < T < 1 is 5505 Let be a sequence of analytic functions in a domain D which converges uniformaly on compact subsets of D t o a function f on D . we have .. but has a zero point zo E D . It follows from RouchC’s theorem that fn and f have the same number of zeros in { z : Iz . Suppose f is pot identically zero. (UC. Since the zeros of a non-zero analytic function are isolated. (a) First of all.438 For all z E { IzI 5 T } ..z0I < T } .(fI on { z : Iz . we obtain g ( z ) = 0 for all z E U . which is a contradiction to the assumption that fn(z) # 0 for all z E D.z01 = T}. (b) If each fn is one-to-one on D . show that f is either constant or one-toone on D . T}. {fn}r=l ZOI Let m be the minimum value of /). Irvine) Solution. (a) Prove that if fn(z) # 0 for all n 2 1 and z E D . Then m > 0. .

It follows from RouchC’s theorem E { z : Iz . (Indiana) . and let { f n } be a sequence of analytic automorphisms of D such that n-+m lim f n ( a ) = b E 8D for some point u E D. such that fn(Z{) . which implies fn(z{) = fn(zi) ( : # 24).z2 E D (z1 # zz). Prove that n+m lim f n ( z ) = b for every z E D.)I = Ifn(z) .z 2 ) = T } . I(fn(z) ..a and f(z) .a = 0 and fn(zi) .al { z : Iz .a ) holds on that f n (z) .f(z)I < m L If(.221 5 T } . such that f(z1) = f(z2) (denote it by u ) . such that and f ( z ) .u = 0.a # 0 in {z : 0 < ( z .439 (b) Suppose f is not a constant. Then there exist zl. and is not one-to-one on D.211 <T} and z: E { z : (z . when n is sufficiently large. Let rn be the minimum value of If(.a have the same number of zeros in {z : ) z.z 2 ( = T } . Choose T > 0 sufficiently small. In other words. Then m > 0.a1 on { z : IZ .z11 < T } and { z : (z .z l ( 5 T } U {z : 0 < Iz .221 < T } respectively.(f(z). 5506 Let D c a! be a bounded domain.z2( < T } . This is a contradiction to the assumption that f n is one-to-one on 2 D. there exists 2 : .zll = T or Iz .) . With the same reason as in (a).zll = T or (z .) .

Since { f n k } converges uniformly to f on { z : ( z . we have . 5507 Which of the following families are normal. f(z) is a non-constant analytic function of D . 1 1 on { z : Iz . there is a subsequence { f n k ( z ) ) converging uniformly on compact subsets of D to f(z). there exists a subsequence of {fn(ao)} converging to bo # b.b has zero(s) in { z : Iz .it follows from f(0) = 0 and diam f ( D ) 5 2 that lf(z)I 5 2.a = T } .bl : )z .a1 = T } > 0. diam g ( D ) 2 1). which shows that 7 is normal. ao # a. when Ic is sufficiently large. which obviously satisfies the conditions that f(z) is analytic in D and f(0) = 0. Here the diameter of a set S is diam S = sup{lz . and which is compact? Justify your answers.Re{g} > 0. If {fn(ao)} does not converge t o b. (Indiana) S o htion. Then there exists a subsequence { f n k } converging uniformly in compact subsets of D to f(z). For any two fixed points z.m n+ca lim fn(ao) = bo # b. Without loss of generality. then m = min{lf(z) .440 Solution. C E S}. (a) For any f E F. Let {fn} be a sequence of functions in F. which is a contradiction to the fact that fnk does not assume the value b E d D in D because f n k is an automorphism of D . Let T be sufficiently small such that f ( z ) .g(O) = l. (a) 3 = {f : f is analytic in D .CI : z . By Rouche’s theorem. we assume lim fn(u) = b 6 d D . Since { f n ( z ) }is a normal family. f(0) = 0. Because f(u) # f ( a o ) .C E D .a = T } .a1 < T } . fnk(z) .u [ 5 T } c D . diam f ( D ) 5 2) (b) B = { g : g is analytic in D. nP . Take a0 E D .b has no zero in { z : 0 < I z .

hence f(z) E 7 .} converges to R = {w : ( w .441 because diam f n . 5508 Suppose that 1 5 p < 00 and c 2 0 is a real number.31 < 1) U {w : IImwl < -.(z) # 0 for all z E D . gL(0) > 0. Since diam g(D) = g ( z ) does not belong to the family 0. by Caratheodory's theorem.} with respect to w = 1. we have G(z) # 0 for all z E D . ( D ) 5 2. and it is obvious that g n satisfies all the conditions required by the family Q. First we can choose a sequence of functions g n ( z ) in Q as follows: g n ( z ) is a conformal mapping of D onto R. If G(z) is a constant. We choose a compact subset K c D such that z. It follows from the uniform convergence of {f.1 < -} U {w : I . Hence we can define an analytic function g ( z ) = -log G(z). which shows that family Q is normal. Then for Gn(z) = e-gn('). 1 < Rew < 3) w 1 w 4 1 1 n satisfying gn(0) = 1.) = -logGn. by Hurwitz's theorem.which shows that F is also compact. Because the domain sequence {a.( .(z) converges uniformly in compact subsets of D to g ( z ) . By the Riemann mapping theorem. we have IGn(z)l < 1. where the single-valued branch is chosen by g(0) = . and we conclude that g n. Let 7 be the set of all analytic functions f on { IzI < 1) such that . Since z . which shows that is not compact. then the constant is e-l because ~ ( 0 = lim ) e-gn(') n-vw = e-1.log G(0) = 1.. we obtain diam f ( D ) 5 2.C E K . if G(z) is non-constant analytic. such a mapping gn exists and is unique. a} i. Hence there exists a subsequence {Gnk} converging uniformly in compact subsets of D to a function G(z) which is either a constant or a non-constant analytic function in D .} on K that If(. c E D can be arbitrarily chosen.f(0l 52.) . } be any sequence of functions in Q. { g n ( z ) } converges uniformly in compact subsets of D to g ( z ) which is a conformal mapping of D onto 52. since G.1 < 1 which is called the kernel of {R. (b) Let { g . But family Q is not compact. = {w : I .

Let 1 . We prove the assertion by contradiction. while the right side of the inequality is a constant. Izn . Show that 3 is a normal family. B y Hence where -+-=l.201 Cauchy integral formula. P q 1 1 Then As n + co. the left side of the above inequality tends to infinity. It suffices to prove that the functions in 3 are uniformly bounded on every compact set of {zI < 1). Then when n is sufficiently large. then there exist zn E D . (Illinois) Solution.f n E F such that zn --+ zo E D and fn(Zn) + 00. < r.442 sup O<r<l /'^ o If(reioe)lPde 5 c. The contradiction implies that 3 is a normal family. If it is not the case.lzol = 3r. .

f(z) # 0 for 0 < IzI < r 5 R. and that z = g(w)is the unique solution of that tends to zero with w.w = 0 that tends t o zero with w. f'(0) # 0. which implies that g(w)is holomorphic in {w : II w < m}. and apply this to find the explicit (b) Find the Taylor's expansion of g(w). Hence tf'(t)dt g(w)= GJ.p n=O t f ' ( t ) dt) w n . ( t ).443 5500 (a) Let f be holomorphic for 121 < R and satisfy f(0) = 0. define a holomorphic function of w for 1I w < m = m p If(pe")l. Solution. Let C be the circle IzI = p where p < T . Show that 1 tf'(t)dt g(w)= 2?ri f(t)-w J. series expansion of the root of the equation z3+ 32 .w f 1 =5(&J. . (a) It follows from IwI <m= rp lf(pe")l n that when t E C.

which shows that f(z) and f ( z ) .444 Let r be the image of C under f where C is the circle { z : IzI = p } taken once counterclockwise. Since z = 0 is the only simple zero of f in { z : IzI < p } . then f(t). (b) Let f(z) = z3 32. + then where . As the constant term in the Taylor expansion of g(w) is which is obviously zero. we know that f(z) = w has a unique solution in { z : 1x1 < p } . Denote the unique solution by z l . 9 @.w have the same number of zeros in { z : IzI < p } .w = (t .w).w Hence 1 Q’(t> + log Q(t)]’ = -+ t -z i Q(t) which shows that g(w) is just the unique solution of f(z) = w.zi)Q(t) where Q(t) is analytic and has no zero in {t : I1 t < p } .w)]’ = [log(t .O) = n(r. and * f’(t) = pog(f(t) . we assert that the unique solution g(w) tends to zero together with w .z1) f(t) . Because lwl the winding number < m = min If(pe”)l.

. * .. By Mittag-Leffler's theorem. when IzI 5 N .3 ..2 (:)k-2 k + '-2kk .l .-2.445 After some computation.-N+lareitssimple poles with residue n at z = -n. So f ( z ) is analytic in 1. 5 N } to a function 2N-1 n=l which is analytic in { IzI <N}.( < N } \ { .2 . + .. Prove in detail that your function has all the required properties. -2k t 5510 Find an explicit formula for a meromorphic function f whose only singularities are simple poles at -1. Because N can be chosen arbitrarily large.:) . . it is obvious that f ( z ) has all the required properties of the problem. . n 2 2N. ..... we construct For any natural number N . z2 is a meromorphic function whose only singularities in { IzI < N } are simple poles at z = -1. (Illinois) Solution.N 1 with residue n at z = -n. . we obtain a2k-1 a2k = 0 and = - zl 1 33k-2 32k-lt2k-1 [ 2 '-2k . a n d z = -1.1 (5)k-'] dt (3ck-2c". -2.with residue n a t z = -n. 2N2 Hence converges uniformly in { IzI In addition. .N + 1 } .. . . -2. .

. IPn(z) < E holds for all z E (1 < Izl < 2).we have 5 $1 a). explain why not. (b) Yes.1 < 2 } . The contradiction follows by taking z = 0 in the inequality. let (z( R and choose N 5 + > R2. give an explicit formula for g.1 is an analytic function in maximum modulus principle that 1. It is easy to see that .. Then when n 2N .2. it follows from the holds for all z E { Izl < 2 ) .3.( < 2}. if No. The function g can be chosen as where an = f i ( 1 i).} such that P n ( z ) --i uniformly on the annulus 1 < IzI < 2? If Yes.446 5511 (a) Does there exist a sequence of polynomials {P. -)? If Yes. (b) Does there exist an entire function g whose zero-set is {+( 1+ i) : n = 0.1. there exists N > 0 such that when n > N . then for any E E (0. give an explicit formula for the Pn . if No.} such that P n ( z ) -+ uniformly on (1 < 1. Multiply both sides by )zI2. explain why not. (a) No. For any R > 0. If there exists asequence of polynomials {P. (Illinois) Solution. 5 Because z 2 P n ( z ).

we see that g(z) is a n entire function with the required zereset.a2. which implies m is analytic in { z : IzI < R } . Then Fn(z) are entire functions with no zeros. we have -1 5 Refn(z) 5 1 + l o g 2 . and its zeros in { z : 1z1< R} are 0... and prove your assertion. For each positive integer n there exists an entire function fn such that (Indiana) Solution..al.ak . False..1 5 IC < $). we know that is analytic in {z : IzI < R}. If for each positive number n there exists an entire function fn such that then for 15 IzI 5 2. and -5 e 1 ~ ~ ~ ( zeRefm(z) < 2e > l = - .447 Because C n=N 00 5 converges. Define Fn(z) = efn('). Since R can be arbitrarily large. Hence g(z) is analytic in { z : IzI < R}. We prove the assertion by contradiction. (q 5512 State whether the following statement is True or False.

and there exists a subsequence {Fnk(z)} converging locally uniformly to an analytic function F ( z ) in { z : IzI < 2). (a) Give a single-valued definition for zi in G. Since IFn(z)l 2 for 1 5 IzI 5 2. (b) Why should there exist a sequence of polynomials P. For example.(z) = z* for all t in G? (c) Can the polynomials be chosen so that there exists a constant M with IPn(z)l M for all z E G and all n? Justify your answer. t contradiction to the fact that F ( z ) has no zero in { z : IzI < 2). Hence {F. (a) zi is defined by eil0gz. Because the complement of K. we know by Runge's theorem that there exists a sequence of polynomials which converges . is connected and contains z = 00. a single-valued branch of zi in G can be defined by argzlo<2<l= 0. This is a 5513 Let G = D\(-1. F ( z ) cannot be identically zero. By the maximum modulus principle. single-valued branch of logz can be chosen. and by Hurwitz's theorem F ( z ) has no zero in { z : 1x1 < 2).(z)} is a normal family in { z : ( z ( < 2).48 4 for 15 IzI 5 2.0]. which implies that F ( z ) = cyz with = 1 in { z : I1 < 2). But we have for 1 5 IzI < 2. where D = { z : IzI < 1). IF. Then K. In domain G. C Kn+lr and n-rco lim K. = G . such that n-rm lim P. (b) Choose where n > 2.(z)I 5 2e for IzI 5 2. 5 (Indiana) Solution.

. to z i .. . a2 1 The singular part of f at z = n (n = 0. If there exists M with IPn(z)l5 M for all z E G and all n. which implies that z' can be extended to a single-valued analytic function in D.449 uniformly on K . 5514 (a) Prove that (b) Use this to show that acotaz = Z l o o + c2z 22 . .2. with enough terms to find the values of n=l C 5 and C 00 00 1 n=l (Hamad) Solution. (c) Develop a cot az in a Laurent series about the origin directly and by use of (b). It is obvious impossible. * . ( z ) which converges uniformly on compact subsets of D to an analytic function f ( z ) in D . Hence {Pn(Z).. we can find a polynomial Pn(z) such that for all z E K.} converges to zi uniformly on n compact subsets of G. It follows that {Pn(z)} is a normal family in D . hence zi = f ( z ) for z E G.n2 ' n=l Justify your steps. fl.) is . IPn(z)l 5 M for all z E D and all n.Now we consider (2-n)2. = 1. (a) Let f(z) = G. In other words. then because Pn(z) are continuous on D. and there exists a subsequence P. (c) No. f2. the series - . so the contradiction is obtained. Since Pn(z) converges t o zi in G.

holds for n 2 2N and n n=-2N 5 -2N. It follows from the convergence of C 5 n=2N 00 and C 5 that C ~~ M -cQ n=-m & is analytic in Because N can be arbitrarily large.450 For any natural number N and la1 5 N. Imt+f00 As the convergence of is uniform for lIm4 2 1. . As f(z) and are both periodic functions with period equal to 1. the limit of the series for Imy -+ fco can be obtained by taking the limit in each term and the limit is also zero. we restrict z in the strip { z :0 < Rez 5 1). It is obvious that lim f ( z ) = 0. Hence g ( z ) is a bounded entire function. Let Then g ( z ) is an entire function. we obtain the result that 00 1 n=-cc is a meromorphic function which has the same singulaxities as f(z).

&2. Let 22 1 " F(z)= . have G'(z) = 0 which implies that G = c we ( c is a constant). It is obvious that the constant must be zero. . + n=l c 22 . Hence we obtain 7rcot?fz= -+c1 " 22 z2 n=l .)is the series 1 " n=l - A.n2 + Then G ( z ) is an entire function. .n2' . it follows from the fact that F ( z ) and 1 ."+ E n 22 n=l are both odd functions that c = 0. we obtain T2 1 " n=l 1 1 .. we consider Now 22 With similar discussion to that in (a).45 1 which implies that g ( z ) is a constant. Thus we obtain the identity (b) Let F ( z ) = 7rctg7rz. we know that is a meromorphic function which has the same singularities as F ( z ) . The singular part of F at z = n (n = 0.G(z).G'(z) Comparing this identity with (l).n2 + c. f l . For z2 n=l . Differentiating both sides of the above identity.

.tl) 9 * .z2.-23 . does there exist an entire function f with f(zj) = aj? Can you write this function down? - (Hamad) .2 and g(z) Let f and g be polynomials.. f of degree . ....zn). 45 (3) It follows from (2) and (3) that around the origin. (a) Show that (b) Show that there exists a polynomial of degree 5 n .. = ( z . we can also obtain 5515 Let z1. Zn be distinct complex numbers.2 with f (zj) = uj if and only if n (c) Given a sequence of complex numbers z1. Take z = 0. . such that lzn 1 + 00.452 (c) The Laurent expansion of 7rcot 7rz about the origin is j cot l . rz x . 5 n . . we obtain After differentiating (4) on both sides.-z 1 z 7r2 3 7r4 .( z .

Z n ) is of degree n. .an are n complex numbers such that n we construct the function f ( z ) by For each j . n).( z . (a) Take R sufficiently large such that Zl.zj . -dz -dz = lim = 0.2.. E 1(. < RS. u 2 .z 1 ) . j=1 (b) If f ( z ) is a polynomial of degree 1. .(2 . then by (a). . and the coefficient of zn-' is 1. Because g ( z ) = ( z .z.z j . we have 5 n .Z.. .453 Solution.l ) ( z . --z ) g( z .2.q ) ( z .z 2 ) * .1."'. .2 with f(zj) = uj (j = If u l .) is a polynomial of degree n . ... we obtain ?#=O. ( z .Z2.. . Since . while f ( z ) is of degree 5 n .Z j + l ) .( z .

we can construct an entire function g ( z ) with simple zeros z l . . In other words. - where 7. It is easy to see that while for k # n. f(z) is a polynomial of degree 5 n . z2. f ( z ) is an entire function.2. Hence > 0. f ( z ) satisfies the condition f(zj) = aj ( j = 1.I -+ 00.2. Since R can be formly for all IzI 5 R. z2. n=l < R}. such that . .’ajz j ) g ( dz) ~ ( z .1 I -+00.I > 2R 2 00 1 u n ( a ) converges uni5 R when n 2 N .aj. is chosen such that when IzI 5 k$. -. In other words. which implies that f ( z ) is an entire function satisfying the required condition.( holds for all IzI arbitrarily large. uk(zn) = 0.n). Then we define - a. there exists N > 0 such that Iun(z>l 5 Iz. by the Weierstrass theorem about the canonical product of entire functions. Because lim z+rj -. (c) For the given sequence z1. while for k #j.Zj) . Because Iz. so that f(z) is analytic in 1.. for any R when n 2 N .454 the coefficient of zn-l of f(z) is zero..

Part VI Partial Differential Equations .

.

b) Show that u E C:(R") implies (Iowa) Solution. Show that ( 2 .457 SECTION 1 GENERAL THEORY 6101 a) Let A = (a. . = - .j)&=l be a real matrix. -) is the dot product of vectors in R" and wn is the area of the unit sphere in En. lxl<l Vi # j and Therefore. Ax)dx = Wn L<l n(n + 2) trace A. a) It is not difficult to verify that aijxjxjdx = 0. Here (..trace A Wn 1 n n(n + 2) trace A.

the conclusion is true for u E Ci(IRn). . IR Take a function p E C r ( l R )such that J. T' = 0 if and only if T(q5')= 0.. g d x = 0. R (Indiana) Solution.e.pdx = 1. we get immediately As Cr(lRn)is dense in Hi(IRn).By applying Green's formula. Then it is easy to verify that for any q E C r ( I R ) 5 satisfies the condition ? J. Vq5 E C r ( B ) . i. show that there is a number a such that T(q5)= / aq5dx for all q5 E C r ( l R ) . 6102 Let T be a distribution on IR and suppose that T' = 0 on IR. It is easy to verify that a function q5 in Cr(nZ) is the derivative of a function in C r ( B )if and only if q5dx = 0.458 b) First let u E Cr(IRn). Show that T = const.

Then we get immediately + + 6104 Let u be defined for f E D((0.b] c ( 0 . It is clear that u. Show that lim u ( z ) = 0.l)).l)) by Determine whether u is a distribution on (0. In fact.a / 2 E L2(Rn) if s > n/2. l ) such that SUPPfk c [a.l)) such that fk + 0 in D ( ( 0 . vk: .bl.459 Hence T($) 0 and = 6103 Let u E Hs(En) with s > n/2. defined above. we have 6 E L 1 ( R n ) . is a linear functional in D((0. Let {fk} be a sequence in D((0. l)). 14+w ( cincinnat2) Solution. l). (Indiana) Solution. And it is clear that (1 1[ 1 2) . u E H s ( R n ) if and only if (1 lt12)'/2G E L 2 ( R n ) . as Ic -+ 00. Then we have a compact interval [a. and support your answer. Therefore. We show first that 6 E L 1 ( B n )if u E H s ( E n ) with s > n/2.

o uniformly on [u.#dxdy . that is J. 1 N Then we have Therefore. Let N be a positive integer such that . XY -2 ( . u#. ) 5 belong to WIJ’(B)? (Iowa) Solution.and is a distribution on (0. Let uy(x7 y.460 and jp)-. in the sense of distributions. + y2)3/2 * It is clear that us.dxdy =- J.uy E L1(B). b] as for any nonnegative integer n. -+ oo.l)). 6105 Let B = {(x.l). u is continuous in D ( ( 0 . u.< a. y) I x2 For which p 2 1 does the function + Y2 < 1.

(Indiana) Solution. 0 in the above inequality.. we denote the ball centred at the origin with the By and S. Let RE = B\BE for any E C r ( B ) . u E W1ip(B)for 1 5 p < 2. Letting E -. u~cos(?~'. v. Then we have < 1. = aB.dzdy = - lc ux4dxdy + 1. Similarly L u = 2ly. Denote by v = v(x) the unit normal to C pointing into R 2 . and Let C be a smooth hypersurface dividing R" into two disjoint open regions R2.461 and r JB u+ydxdy = .. And we can verify that for which means that u @ W1yP(B) p = 2. we get &u = u. Derive a necessary and sufficient condition in terms of v +. ' BY It is not difficult to verify that + J.JB uy#dxdy r for all radius E E Cr(B).C ( *I in the classical (C')sense where b is a continuous function. B.. x)ds for 1 5 p < 2. 6106 Q . Therefore. Assume furthermore that v lies in both Co(?&) and C0((n2)(but not necessarily Co((IRn)!!) with v+(xo) lim v(x). v is a distribution solution of (*) in En and only if if . vu-(xo) lim v(x) x-xo x-DO +En1 XEn* for each xo E C . Suppose v : R" -+ R" solves the equation divv(x) = b(x) for x E R" .u4.and v for ZI to be a distribution solution of (*) on all of R".

. #dx. V+dx = - +(v+ ..v-) .y)): vxuu~2Ly-vu.. 2). and U1 and U2 are (different) constant vectors.462 - b#dx. V+dx = .(IRn). V#dx = IRn V# E CF(lRn). By Green’s formula. If U has curl zero in lR2 in the sense of distributions. (a) Let u.. (Indiana) Solution..v E Lloc(B2). J. (b) Suppose U takes the form 2 2 1 2 where 5 1 and a are open sets such that a U n = B2.V-) . we have -Ln 2.y).J. vds = 0. (a) Give a definition of what it means for a (not necessarily continuous) vector field U to have curl zero in the sense of distributions. Recall the definition of the curl operator in two dimensions acting on a vector field U(z.y) = (u(z.v . describe as completely as possible the curve r. Hence the equation (1) is equivalent to #(v+ .I’ z d a l n a 5 2 2 is a smooth curve.Y = o 6107 on C.v-) uds + lRn divv . Therefore v is a distribution solution t o (*) on all of B* if and only if (w+ . V# E C. Then curl U = 0 . V#dx - J.. 2.v(x.

Here S is the Schwartz class of rapidly decreasing functions. v< E {< : < R}.-(lR”. U2 = (uz. 4 - Construct a function X R E Cp(lRn) such that XR(<) = 1. there exists a positive constant R such that {< : P(i<) = 0) c {< : 1 < R). (Indiana) Solution. v4 E C. WI). Therefore. I’ is a straight line defined by (u1 . where Then curl U = 0 if and only if u1.v2)dy = 0. and assume that the set {z : p ( i z ) = 0) is bounded. that and E . where S’(W) denotes the space of tempered distributions on B”. 6108 Let p be a polynomial in n variables.463 in the sense of distributions is defined as (b) Let 171 = (u1. u2 and 712 are constants. we have J.6 E S. Integrating by parts.(ul . and 6 is the Dirac distribution defined by 6 4 = 4(0). Prove that there exists a tempered distribution E on lRn such that p ( D ) E . w1.u2)dz + (v1 . As the set {< : p ( i < ) = 0) is bounded. Set It is clear that E S ’ ( R n )n Cm(Rn).vz).v2)4dy = 0.%)4dz + (v1 .

From the transformed equation. we get p(D)E. we can see that suppG = { 0).e. ..-(nZ"). are constants. are constants. Hence is a polynomial. we get - C < P ( g ) G ( [ )= 0. P(D)= C. i.D" bllm where a denotes a multi-index Q = (a1.(i<>" o for all E # .) and the C..' ( Z ) E S'(nZn).) is a multi-index and a.^(€I = + c a. G ( t ) must be finite linear combinations of the derivatives of the Dirac delta function S(<) at { 0 } . Let G(E) be the Fourier transform of u. prove that u must be a polynomial. . l4lN where a = (a1. Let P ( D ) be an m t h order linear constant coefficient partial differential operator on En.464 It is easy to see that p ( i € ) Z ( € ). (Here S' denotes the space of tempered distributions on an. .e. .Transforming the equation. 6109 s.1 E C.) (Indiana) Solution..~ 2 .a. I4<m If u E S' satisfies P ( D ) u = 0.(0). Therefore. Suppose ~(i<) E ~.S(")(<).6 E where E = F .a2. Therefore..a. i. .

suppose that q5n in D.we have Then it is not difficult to verify that Hence t+O+ lim - & e-x2/4t -6 . (d) Calculate the Fourier transform of the distribution (Indiana) Solution. Let supp4.a ..465 6110 in D'. (c) Calculate rigorously the derivative of log 1x1 and express the result in terms of your answer to (b). 4 E D.e. a ) Vn. Then we have -+ 0 . Further. as (b) We define the distribution follows: 5 by its Cauchy Principle Value &: The linearity of the functional Vp: is obvious. (b) Show how to define as a distribution (i. c ( . define the Cauchy Principle Value and show that it is in D').( 1. e-xaf4tdx = 1 Vt For any > 0. (a) It is well-known that 5.

we have Here. 2 (d) By F ( f ) we denote the Fourier transform of a tempered distribution f For any 4 E S.:. Therefore. we have applied the integral equality 'J-00t 'IF -sin(zt)dt = 1 for z > 0.DcIu(c) 2 €mn .we have V p i Hence -&log 1 1 = V. we obtain 6111 a) Let Lu(2) = bI<m c a. ' ( c ) For any 4 E D . is a distribution in D .466 Therefore.

u.467 where the abs are constants. where S(z) is the Dirac function.u. b) We need only to show that In fact . Simplify your result to obtain D’Alembert’s solution of this problem. O ) = g(z).t).. What is a fundamental solution of L? When is it unique? b) If I+denotes the Heaviside step function Show that 1 F ( z . 0) = f(z).)1. = #J(z. u ( z . < 2 < 00. (Indiana) Solution. and a is a multi-index. on E 2 . t ) = -I+(t)l+(t. -00 (f. Let L* be the formal adjoint of L defined by L*V(C) = lallm c (-l)~aba(aaV(z)). t > 0 -00 < 2 < 00 U t ( 2 .g. c) Express the solution of the Cauchy problem utt .. a) A distribution E E D’(nZn) is called a fundamental solution of the differential operator L if L E ( 2 ) = S(z). If L*S is dense in S. where S is the space of rapidly decreasing functions. then the operator L has at most one fundamental solution in S‘. 2 is a fundamental solution of the wave operator W ( u )= utt . #J sufficiently nice) in terms of the above fundamental solution.

the solution of the original Cauchy problem can be obtained as .468 The solution of the Cauchy problem is given by Therefore.

6112 Let u be the solution of the initial problem where summation is understood in the pde. respectively. Assume that there is a constant C such that (Indiana) Solution.469 This is just the D’Alembert’s formula when q5 2 0. where G and denote the Fourier transforms of u and q5 with respect to z.0) = w. and 4 is a C” function on IR3 with compact support. qt. it yields that 3 %t) = Therefore. we have Set 77 = tit in the integral in the right side of the above equality. It is reduced to . By applying Fourier transform to the problem.0) = 0. By solving the above transformed problem. q t . we have ztt+ i t 1 4 6= 0.

(Indiana) Solution. So we have And then . m 4 And when t > 5 .1 < r. if I&(O)l = 2a # 0. then there exists a positive constant r such that I l2 a. given in the problem.t)I2dz = I&0124- Here 3 denotes the Fourier Transform of 4. K E E3. consider the solution u of the SchrGdinger Equation Show that lu(z. It can be verified that the Poisson's formula for the Cauchy problem of the heat equation applies to the Schr6dinger equation.470 This implies that In fact.we have This contradicts the condition satisfied by the solution u. 6113 Given E S (rapidly decreasing functions) over B1.

we get Here the Fourier transform of 4 is defined by .47 1 Set x = 2&.

pn-1U2(x0). Show that does not exist.472 SECTION 2 ELLIPTIC EQUATIONS 6201 Let u $ 0 satisfy u E C2(Rn). Applying the meanvalue property for the harmonic function u. = 0 on Au R". (Indiana) Solution. is the surface area of the unit sphere in En. n . There exists a point xo E En such that u ( x o ) # 0. we have where w . z-z*(=p Then we have u2(x)dx wnu2(xo)Jlo pn-'dp w 2 = -u n ( x 0)T n . Schwarz's inequality gives that is JT u 2 ( x ) d S z >_ w.

Show that u is then harmonic in R. u. we have where This implies that for any compact set K dist(K. an). Set where C is a constant such that For x E R. i.(R). 6202 Let u E Co(R) be weakly harmonic in an open set R. +u uniformly in K . Therefore u is harmonic in R.-(f% < dist(z. .473 for any T > 0. It is well-known that c R. The conclusion follows as T + 00. Therefore.. (Iowa) Solution. as E -+ 0. is harmonic in K if E < u. set PE(Y) = E-P Then it is easy to see that provided E 2-Y ( ) 7 P4Y) E C.e. the relation uAddx = 0 holds for all 4 E C. an).

y) E . as a function on y). y ) is continuous at z = z0 uniformly with respect to y near yo. harmonic in z E lR2 and continuous in y E BIZ. B4. yo) E B4and R such that lf(z. Then there exists a constant M > 0 lR4. Let (zo. These estimations imply that f ( z . Prove that u E 0 on z 2 0. 1 2 5 M.474 6203 Let f (2.yo) E B4follows immediately. .. . (Indiana) . 2 0 of En. V(z. Then the conclusion that f ( z .y)I > 0 be given. For z. we have By Green's formula.y0l2 < 2R2. 6204 Suppose u is a function defined on nZn such that (i) u is bounded and continuous on the half-space xC (ii) u = 0 on z = 0. (Indiana) Solution. y ) is continuous at (zo. i. Applying the mean-value theorem to the harmonic function z. (iii) u is harmonic in z > 0. we get we have the same estimation.z0l2 + Iy .e. .Show that f is continuous in (2. y) be a locally bounded function.

) (Indiana) Solution.475 Solution. < O where x’ = (xl.-1) and w. Therefore.bounded. > 0) and continuous on y the boundary alR. VX. where X’ = ( ~ 1 .-**. < 0 as a n odd function of u(x‘.e--. u is bounded and harmonic in B”. It is not difficult to verify from the above formula that and . By the uniqueness of the bounded solution to the Dirichlet problem (3). satisfies j = l . Redefine the function u in the half-space x. Thanks to the Liouville’s theorem... If u.xn-l). * . . = 0).-x~).(.2n) 2: . and let g(x) be a compactly supported. By using the uniqueness theorem of the Dirichlet problem and the Poisson’s formula of the Dirichlet problem in a ball. (Warning: you must T justify any assumptions of regularity you make on the solution u. = -u(x’. = 0 ) . = {x : x. ~ ~I‘ 1 )([1. we can verify that the redefined function u is harmonic in any ball centred a t the origin in lRn. we get u s 0 immediately.. n 1.* . is the surface measure of the unit sphere in B”. we can show that the unique bounded solution to the problem (3) is given by e. are bounded in show that its “tangential” derivatives magnitude by maxlVgl on all of B = {x : xn > 0). C1function on alR: = {x : x. -= . 6205 Let u(x) be C 2 on the half-space B = {x : x.

( Cincinnati) Solution. _ . we have uAudxdy . IyI u = 0 for 1x1 = 1. . au au = 0 ax ay for IyI = 1 has at most one solution. We need only prove that the problem Au = 0 for 1x1 < 1. It is easy to see that lXl uAudxdy = 0. we have which are just we want to show.-au = 0 for IyI = 1 ax ay has the unique solution u 0. au . y) is a function of two variables x and y. < 1. < 1. (Here u = u(x.476 Therefore. IyI u = 0 for 1x1 = 1.. 6206 Show that the problem Au = -1 for 1x1 < 1. and is a classical solution).lY I < 1 Integrating the above integral by parts.- < 1.

and that u can be chosen to be realvalued. . we obtain (Vu12dw= a2 IuI2dv. (a) Show that a is a real number. and suppose nontrivial solution exists to CY is such that a { -V2u = a 2 u in R.1) . Show that r (Indiana-Purdue) Solution. which implies that a is a real number.1) .u2(-1.477 = -(u2(1. u=O on a R . -1)) Then we get u f 0 immediately. (a) Multiplying the equation in the problem (*) by the complex conjugate of the solution u. and integrating the resulting equation in R. Then it is clear that u can be chosen to be real-valued. (b) Suppose a1 # a2 are such that nontrivial real solutions u1 and u2 satisfy (*).u y 1 . 6207 Let R be a bounded normal domain in lR3. -1) 1 2 + u2(-1.

Let u be subhamonic in the sense of (i).478 (b) Suppose that equation satisfied by u1 u 1 and u2 are chosen to be real-valued. we get immediately which implies the conclusion of the problem. Let u be a continuous function on a. Vu2dv = a: and From the above integral equalities. respectively. (Indiunu) Solution. we can show that u can not take its maximum in a unless u is a constant. one has u 5 h in B.r ) centered a t x with radius r satisfies B ( z . By a standard argument. (i) For every x E R and r > 0 such that the ball B ( x .r ) cc R one has (Here un denotes the surface area of the unit sphere in Enand B ( z . and integrating the resulting equations in R. we Vul . 6208 Let fl c B“ be an open set.r ) cc R means that the closure of B is contained within R. Multiplying the -V2u1 = a9u1 in R and the equation satisfied by u 2 -v2u2 = aiu2 in by u2 and can obtain u1.) (ii) For every ball B cc fl and every harmonic function h E C o ( B )such that u 5 h on d B .Prove the equivalance of these two notions of “subharmonic”. .

YI = 7-1. :Y E lVU(. 6209 . we have au n -(z) =dXi in the wn r n where w. c) Show that.h is also subharmonic in the sense of (i). L -SUP{lU(Y)l 4s) 0) (Indiana) Solution. and the mean-value theorem. Show 2 that IV4z)l L. Then w = u . a) The conclusion is clear. where B is the open ball centered at x of radius r . we have This completes the proof. By using the Green's formula. r ) .479 Let h be the function given in (ii). the above equality can be rewritten as . Hence that 5 0 on 8 B implies that w 5 0 in B. a) Show that is harmonic in R for each i. /n is the volume of the unit ball. { From the definition of subhamonicity in the sense of (ii).)l where d ( z ) = dist(z. let u be subhamonic in the sense of (ii). Conversely. is the surface area of the unit sphere in En. SUP~lU(Y)l 1% : 3 : n . h=u on a B ( z . r ) .Let R E B" be an open set. b) Applying the mean-value theorem for the harmonic function ball B . if R is bounded and n E R. hence w. b) Let B c a. We define the function h as follows: Ah = 0 in B ( z . and let u be harmonic in R and continuous in R. 8 0 ) .

For this ordinary differential equation. ds2) be a fixed positive constsant.sin(kr). By the fundamental solution cos k r .cos(kr). it is not difficult to verify that & where T. Then the spherically symmetric solutions depending only on T satisfy -(TU) d2 dr2 + k2TU = 0. (Indiana) Solution. T r Let u be a solution to the equation A u + k2u = 0. 6210 Let R C lR3 be bounded and open with smooth boundary. Let u E C1(n)n C2(R) solve A u + k 2 u = 0 in R (k > 0).480 And therefore. .2'1. we have c) The conclusion can be obtained from the result of the part b) immediately. Let T = 1x1.O. By B ( z o . Applying the above integral formula in .R) we denote the ball centered at zo with radius R. Let R < dist(zO. Derive an appropriate mean-value property for the solution. = ( z . there are two linearly independent solutions: 1 1 u = . First we look for the fundamental solutions of the equation.

E aB(zO. Prove that its spherical mean. = 0.R2) satisfy AU = u + 1. it is easy to verify that Subtracting the equation (2) from the equation (l).R). zo) 1 47Frxo. u(z)dS.. Txox - 4.we obtain u(x0) = a . 2 g(z.rrRsin(kR) J BB(xo.o.zO)= cos(kT. we get Now we look for a function g(z. z E B ( z o .cot(kR)1 sin(kr.R). defined by . For this function.R) This is a mean-value theorem for equation A u 6211 + k2u = 0. which satisfies (A.481 the domain B ( z o .+ k2)g(z.) rxox satisfies the above conditions. Let u E c2(.z0) = cot(kR) sin( kr.o. R). It is easy to see that the function g(z.zo).o.).

B ( 0 . r )denotes the ball of (Indiana) 1 =21r L B ( 0 . 6212 Let B = {(z. Solution. we get Then it follows that = r(w(r) 1). Set T = 1 1 Then we have 2. Here. and Applying Green's formula. Therefore + The proof is completed. dSy the element of arc length. l ) u(rw)dS. 1 u=O onaB . Prove that the problem Au= in Jm B. W(T) # 0. for z radius T .y) I x2 + y2 < 1).also satisfies ( l ) .482 and v(0) = u(O).

This implies that f E H . (gradv) + Xuvdx = where X > 0. w E L2(B) and f = -a v . ( Cincinnati) Solution. Hence the problem admits a unique weak solution u E Hip). (b) Will the conclusion of part (a) be valid if X = O? Explain. - . gvda v v E w'q(n) Vu.a w + ax dy in the sense of distributions (see the solution to the problem 6105). = h[(gradu) (gradv) v)1 - + Xuvldz. (Iowa) Solution. Then by applying Cauchy inequality and the trace theorem in Sobolev's spaces. fvdx 1. Let X v= d 3 w = &5qj5 and f = 1 &G-p' It is easy to see that v . (c) Soppose that f . and (1) L ( g r a d u ) .v E W'y2((n). g and 80 are sufficiently smooth.483 has exactly one weak solution u in HA(B). (a) Define the inner product in W172(f2) by (u. 6213 (a) Let R be a bounded domain in lRn with smooth boundary f E L 2 ( f i ) .and g E L2(dR) be given. we get + J. Show that the problem (1) has a unique solution u E W1!2(R). And denote the corresponding norm in W't2((n)by 11 ( ( 1 . Consider the problem aQ. write problem (1) in the classical form.' ( B ) .

z n . (b) By the classical form of the problem. z n ) = (x'. an + 6214 Let R be a unit ball in IR" centered at zero. the classical form of the problem (1) is Xu = f in R. Consider the following problem Find u E Ht(R) such that lVu.1 on R n {zn = 0).= u Inn(z. where IR" 3 2 = (21.).. ( Cincinnati) . x.. Denote u+ = u Inn~X. -Au dU . Thanks to the Riesz theorem. V#€H. Show that Au* = 0 and that du- ax. Then u is a solution to the problem (1).l . Integrating by parts. Therefore. WX'.Vq5dx = a . F(v)= fvdx +lngvdo is a bounded linear functional in W'?'((n). Prove the existence of a unique solution of that problem.44 8 where C is a constant.= g on dR. there exists a function u E W'?'(R) such that F ( v ) = (u. and assume that U* are regular enough. V' ).o} and u. g and dR are sufficiently smooth. Jnn {X . The uniqueness is clear. azn du+ . Therefore. v v E W'J(R). (c) Suppose that f .<o}. it is easy to see that the conclusion of the part (a) is false if X = 0 even for smooth f and g.1(R). =o} .

n{X.. > 0) and R. The proof is completed. = 0).= R equation by parts.a x . Integrating the above - l+ 1 Au+ . n { x .V 4 d x = where R+ = R n {z.. According to the Riesz theorem. ""'> / dx' = au+ ax. + L- - J. nn{ x .(Q). =1 on Q n {x.11 is the norm in Hi(R). 4 d x + nn{ xn=o} (E. . v4 E m Q ) . This proves the existence of the solution to the problem. there exists 1 (%4)1 u E Hi(R) such that = / nn{ I=o} . Assume that u* are regular enough. The uniqueness of the solution to the problem is clear. auax. Write the equation in the following form Vus V 4 d x Vu. we can get I Cll4ll1. < 0). we get immediately that AU* = o in a&. Define the scalar product in Hi(R) by By the Cauchy inequality and the trace theory in Sobolev spaces.4dx - l- Au. ax. we have + 1. From the above equation. =o} 4dx'. V4 E H.=o} 4d~'. .485 Solution. : where I(. Mx'.

show that 2 inf u an (essinf) ( Cincinnati) Solution.u.V 4 d ~2 0 inf u n V 4 E W. Let an d(z) = max{l Then it can be verified that . 4 E W. a.e. in R. From the above inequality and (I). we obtain Hence 4 =0 This implies that u 2 I a.e.486 6215 J. in R.VU .’”(R) and Now we have Vu V4dx = - J. an Assume that inf u = I > -00.e. b) Let u E W1?’((n)satisfy (I) above. 0). a) This is a corollary of the conclusion in b) of this problem.’”(R). b) If inf u = -00. the proposition is true. IV412dx 5 0. in R . Show that u 2 0 a. By the PoincarC inequality with the above estimate. 4 2 0. it holds that IV+(2dx= 0.

we have . -Au. w E W292(R)n W. respectively.nw. Multiplying the first equation and the second equation by u.n11w. and then integrating the resulting equations in R. in R.v. and integrating it on R. w strongly in L2(R). we obtain {u. as n ( Cincinnati) Solution.dx . Vw. Multiplying the first equation by v.487 6216 Consider functions u. both equations above can be written as Vu. -+ = w. Therefore.} is bounded in L2(R) and w. + +oo. we get By using Green's formula and noting that u.11i2 Then we have - = 0. -+ o strongly in ~ ~ ( 0 ) .. E W. . = 0...'72((n). w. Prove that '11. and w.'12(R). that such -Awn + u .

and Av. This implies that {un} is bounded in W.dx. +0 strongly in L’(R).12(s2).Suppose uk E cm(n) satisfies Auk +uk = f and has the property that for some positive number M .dx.Vw Vv. ~ ~ ~ ~ v Vu. we can get . 6217 Assume R c Rn is a bounded open set with smooth boundary and let f E cF(R). Multiply the first equation by Av. Vundx. we can get -J. This gives us the conclusion of the problem.IIVu..2. = 0.dx - . Prove that there exist a function u E C-(R) satisfying Au u = f and a subsequence {ukj} such that ukj converges uniformly to u on each compact subset of R. finally (4) By similar consideration.Il$ = LwAu. . By noting (3) and (4). it follows from the above equality that Av.488 By noting (1) and (2). . (Indiana) + . Vv.. respectively. J. and then integrating the resulting equations. uk(x)2dx 5 M for k = 1.dx = - J.Au..~ ~ A . it follows from above equality that Multiplying the first equation and the second equation by Au..AunAvndx + nllVun11i2 IIVUn11ia Hence + nllVvn11iz = k VW. Av.

2 . by C we denote various constants. v x E 0 1 . and therefore in D'(R). r ) we denote the ball with the radius T (< R) centered at z. It is clear that v k satisfies A V k + V k = 0 in R and Hereafter. where W k and z k are defined by and . we find that Integrating the above equality from Here R can be chosen so small that Ivk(z)I 1 h(r)dr # 0. By the regularity theorem of the elliptic equations. aB(X. " ' .u1. Therefore. k = 1 . we have u E Cm(R). Let R l be a subdomain of R such that 2 CC R and R = dist(dQR1) 1 fixed. Set V k = u k . see the solution to the problem 6210). Set vk =w k + z k . where h(r) is a continuous function of 0 to R with respect to T . For any z E 21.r) T. s there exists a subsequence { u k j } such that u k j converges to u weakly in L 2 ( R ) . we have + h(T)vk(z) = 1 vk(Y)dSy.489 Solution. by B ( z . By the mean-value theorem for the equation Au u = 0 (for the case n = 3. Then Au + u = f in the sense of the distribution. As { u k } i bounded in L2(n). it holds that : (1) 5 c. Now we show the second conclusion.

And by applying the mean-value property for the harmonic solutions.e. in R}. 5 c. and A = {v E Hi(R) : h l 5 v 5 h2 a. we prove the second conclusion of the problem immediately. By applying the solution formula of the Dirichlet's problem for the Poisson's equation. we can obtain from the above estimation IVzk(x)) 5 C in any compact subset of R1.e.2. . V k = 1. are given smooth functions.respectively. vx € ill. 6218 Let R be a bounded Lipschits domain in B". (Iowa) Solution. VEd Define the norm of w in Hi(R) as .. i.. n-+w lim I ( u n )= inf I(w). h2 : R + . * * . k = 1. where hl. we can get from the estimation (1) lVWk(Z)( <_ It is clear that Igix 1(1 Z. (a) Let {un} be the minimizing sequence of I(w) in A. Then by using the Ascoli-Arzela theorem. k) a 1 c.2. (a) Show that if A # 8. wed (b) Show that L V U V ( W u)dx 2 0 VV E A.a .lR. . there exists u E A satisfying I(u) = minI(v).

* Then F ( t ) has the minimum at t = 0. of {u. be a bounded domain with smooth boundary. This implies r Vu V(v . we get u E A and that by the weak lower semicontinuity of IIwII1.e.u)E A. there exists a subsequence Iull. And there exists a subsequence of {u.}.such that u + u a.denoted still by {u. This implies that I(u) = min I ( w ) . and in H.}. Let r + v E A.}. . wed The conclusion of (a) is proved.~(R) u ---t u in L ~ ( R ) . 0 5 t 5 1. Therefore we have + F’(0) 2 0. n > 1.491 Then we get the boundedness of { I n l } Therefore. Let u E C1@) be harmonic in R. for simplicity we still denote it by {u.such that u -u . w F ( t ) = J. - 6219 Let R c IR”.}. The proof is completed. . in R.u))dz. V(u t ( w . .u)dx 2 0. (b) It is easy to verify that u t ( v . Therefore.u)) V(u+ t(w .

For the function @2 .) b. @I =u = on dQ.. you may cite. Prove that ma IVuJ= rnax ( V u ( .e. a.u.2 .2 0 ) This inequality implies the conclusion of the part b.201 I u ( x ). @I(.u(z0) 1% .u ) 2 0 and @2 in R . From the maximum principle. any standard maximum principles you are using. Suppose there exist functions @ I . ( ~ m 2 l ) . n an i. we have A(Q2 . @ 2 E C 2 ( Qn C1(t) such that ) A@I 5 0. we have 5 u(z). without proof.492 a. it holds that sup(a.j=l b.u = 0 on 8Q.u)= 0. it holds that @2(2) vx E s2% .).u) = rnax(a2 . u ( z )5 Therefore. The conclusion is an immediate corollary of the following inequality: n i. (Indiana) Solution. . %. Prove that m3x l v u l s rnax rnax lVaI1. -x n an (Note: In both parts a and b of this question. rnax l n an an AQ2 2 0. () Similarly.@2(z0) ) x . vz E s2.

the only solution of (1) is u (1) 0. (b) Show that if X > 0 is sufficiently small.0 5 u 5 1.e.u) in R.493 6220 Let R be an open subset of Rn. -Au(z') 2 0. For the maximum point xo.j=l u2izj +~ u ~ I V 0. u = 0 on dR. Suppose u E C2((n)is a solution of the equation Au = u3 with the property that lVu(z)I 5 1 for each z E dR. i. then there exists xo E R such that u takes maximum at to and u ( 2 ) > 1. and dR sufficiently smooth. Prove that IVu(z)I 5 1 for all x E R. do. it is clear that Au(zo) 5 0.. where X > 0 is a parameter. (a) Prove that for any solution u of (l). Set w = IVu12. w takes its maximum on (n a t some point on dR implies that w 5 1 in R. (a) If u 2 1 is false. Hence w 5 1 on 6221 Consider (a nonlinear) equation -Au = Xu2(1 . It is easy to see that w E C2(R)n C'((n) and n Aw = 2 3. . (Indiana) Solution. ~ ~ 2 U Therefore. ( Cincinnati) Solution.

Integrating by parts and taking the boundary conditions into account. (Indiana-Purdue) Solution. (b) Multiplying the equation by u and integrating the resulting equality on R. Hence u 2 0.u) 2 0. As Xu2(1 . we have u E 0 immediately. dn is the trivial one u E 0. dU . we have r 0 = JnA(Au)uda . it holds that Applying PoincarC inequality.= 0 on dR.494 This contradicts the assumption that x u y 1 . u can not take minimum in R unless u f 0. If X < C-l. 8222 Let R be a bounded normal domain in IRn with smooth boundary dQ. Assume u E C4((n).u3(1 By using the conclusion of (a). Show that the only solution of the boundary value problem for the biharmonic equation: A(Au)=O i n Q u = 0 on dR.U)(ZO) < 0. we obtain from the above inequality where C is a constant. Hence u 5 1. we can get J. 1vul2dz = x J. Let ?i' be the exterior unit normal on dQ.

Therefore. Taking into account u = 0 on immediately. u is harmonic in f2.495 = lA~1~dx. af2. we get u 0 .

Show that there exists a constant C such that for all -00 < 2 < 00.. J 2 m -m By the Cauchy inequality. continuous. t > 0. (-0O IR1: u(2. The solution to the problem is given by the Poisson’s formula u(2.W d y . (Indiana) Solution. 03 lf(”)I2d2 < 0O.t)l c I t1/4‘ .t) = .t > 0) (-0O < 2 < m). and satisfies J_. we have Set y = 2 &q immediately + in the second integral of the above inequality. where f is bounded.O) = f ( 2 ) < 2 < 0O.r nf ( y ) e .496 SECTION 3 PARABOLIC EQUATIONS 6301 Consider the Cauchy problem for the Heat Operator in ut = u. Then we get lu(2.

497 6302 Let R c Bn be an open set with smooth boundary and suppose u E Cm(R x [0. Multiplying the equation by u and then integrating the resulting equation on R. t)l&(n)? E > 0. we can get The PoincarC inequality gives IIU(*. For any 5 CIIVU(*. we have . Prove that $ ( t ) + 0 as t 4 +m. it holds that Taking E = 1 / C .t)2dz. Assume that Let $ ( t )= 1 n u(z.t)ll:z(n) where C > 0 is a constant. m)) is a solution of the equation with u = 0 on a R x [O. 00). (Indiana) Solution. we obtain Solving this differential inequality.

. .O) = f ( z ) ( 2 > 0) U ( 0 . < 2 < 00. t ) with respect to z. This completes the proof. d) Show how to solve the problem > 0 .. construct explicitly a fundamental solution for this problem.t > 0) 4 2 .Show that t-a. . Transforming the following Cauchy problem with the initial data S(z): ut = u . lim u(2. t ) = 0 (t > 0) ut = u . (2 by using a) and an appropriate extension of f. a) Using the Fourier Transform.0) = f(z).t)ll&n) . and write down the solution u.--) 0 as t + +m. b) State a maximum principle for this problem.t ) = 0. 6303 Consider the Cauchy problem = u.t) be the Fourier transform of u ( z . (-00 < 2 < 00.0) we get . ( = 6(2). t > 0) u(z. a) Let G(<. Why is the Tychanov non-uniqueness example possible? c) Suppose in addition that f E L 1 ( B ) . ut where f is bounded and continuous.It is not difficult to verify that if l/f[-. 4 2 . (Indiana) Solution.t > O ) .

and satisfies the heat equation in lR x (O. 4 2 . < 0.t) = 4(y)e-Wcly. 2 2 > 0.499 The solution to the transformed problem is given by G(<. -f(-z). t ) E lR x (O. c) The conclusion is clear. we can get the explicit representation for the solution u ( z . 2n-21 b) The maximum principle can be stated i ~ follows: Suppose that u ( z . Then it holds that XER J inf f ( z ) 5 u ( z . t ) 5 sup f(z) for ( z . Then by solving the following Cauchy problem: ut = u x x (-cm < 2 < cm.T].T]with the initial data f ( 2 ) .t > O ) . XER The Techanov non-uniqueness example is possible for some unbounded solutions.) = { f(x>.t ) s is bounded and continuous on lR x [O.t ) = e-tzt. we get the solution to the initial-boundary problem 6304 Consider the problem .T]. t ) to the general Cauchy problem with the initial data f ( z ) u(z. d) Defined an odd function 4(z) as follows: 4(. Then the inverse Fourier transform of 6 gives the fundamental solution t o the From this fundamental solution. O )= 4(z).

we can obtain and Adding the double of the second equality to the first one. .(z. we have 1' 1: tlut(2. .t u(2. we can find that Then by using the Cauchy inequality.. we have And therefore.O) = uo(z) E > 0) L2(E). it holds that We can also get the result by the method of energy integrals. Find a bound on ': 1 Ju. t ) u t ( 2 .t ) d z d t .(z.Ut = u.t )12dzdt = I'L: tu.t)I2dz t > 0. in (Indiana) Solution. (z E IR. and integrating the resulting equations in x (0. From the Poisson's formula for the Cauchy problem of the heat equation. Multiplying the heat equation by u and tut respectively. T).

ut + a(2.(z. t ) E C o ( n ~with .501 otherwise an approach to the function U O ( X ) by a sequence of functions in Cr(En) be used.Let u ( z . T ) ~ ~Ld_zlu. can From the above two equalities. we have A/m 2 -w oo ~ u ( ~ .. t ) with z E R and t < 2 Let B1 denote the open line-segment ' . forming the top of ~ R T and let B = BRT\B~.t ) u = f(2. 2 ) u. t> ..T)I2dz +T Therefore. we get an estimation 6305 For the indicated domain R (interior of the parabola) let RT denote the points ( z . ut E C o ( R U B1) be a solution of ~ %x .

= 0. Suppose that u assumes its negative minimum at ( . f 5 0 in RT..O) = f(z)..we can say that every non-constant solution assumes its positive maximum (if there exists one) on Ba. Hence u.6. (Indiana-Purdue) . Show that if a < 0..(2 0 . u.. What can you say if a < 0. > + This contradicts f(zo.tO) 0.ut(z0.tO) 2 0..f E Co(D).(zO. UP(. t o ) E RT U B1.C. I.502 with a.t) + (0 5 z 5 A and t > 0).to)u(zo.t)= 0. G O .C.t o )5 0. Use the sequence { f n ( z ) = 2 sin ( y z ) } and an appropriate space of conn+l tinuous functions to decide whether this problem is well-posed. then every non-constant solution u assumes its negative minimum (if one exists) on B2. 6306 Consider the boundary-value problem B. f 2 0 in RT? (Iowa) Fig.1 Solution.tO) 0. ut = 0 u(0. f 2 0 in RT.tO) a(zO.t 0 ) . If a < 0. (Here a < 0 should be assumed in RT u B1). Then we have u(zO.t0) 0 < 5 and u. u(z. ut(z0.

(t).(-. By G(<.503 Solution.t ) we denote the Fourier transform of u with respect to 2.x . - a. f n ( z ) is the eigenfunction of the corresponding two-point boundary value problem. = uo(2). (You should be able to find u explicitly.(z)T. . for all f E C[O. This implies that the problem is not well-posed in the space of continuous functions with the norm defined above. t)ll + 00 as n + 00. at . By setting u = f.t > 0) Assume uo E S and find u. When n is even integer.) d at 2 --(<G) a at = -u -I-. 0) az. if not. we can find the solution u to the problem with the initial data f(z) = fn(z) .r. leave your result in the form of a one-dimensional integral.) (Indiana) Solution.) = i-F(u.r]. (x E t . u u(z. One is required to solve u = u . 6307 In Probability Theory one encounters the Ornstein-Uhlenbeck process. is clear that It But for any fixed t > 0 IIu. It is easy to see that F(zu. in which the particles of Brownian motion are subjected to an elastic force.

504 Then the transformed equation and initial condition are in the following form qt. set T 1 = -(I . and let h be a fixed point IR". u(<. Therefore. we can obtain . a2 1 ' 4 E D(nZ"). . 6308 Let L be the usual heat operator L = -a let at (: .2 t ) . t ) = et-3(e2'. we can find a convenient transformation of variables to the problem.t z ) and F-l(e-li(e2t-l)€2) = 1 - d m e 2(e-- 2 . 2 t = e-tx. we get the solution to the problem Remark.l)caiio(et<). From the above solution formula. The above solution formula can be obtained easily from the Poisson's formula for the Cauchy problem of the heat equation.e . In fact. I -+.+of ax.. Solving the above Cauchy problem for the first order linear partial differential equation.. It is not difficult to verify that F-'(iio(ett)) = e . Consider the problem . Then the problem is transformed into the following 7 =0 : 21 = Uo(t).0) = G o ( t ) .t u o ( e .1) .

but you need not prove that it represents a smooth solution.e-"n€)G = 0. Let G( t ) be the Fourier transform of u with respect to 2.t)llL~(R")< e(1-R2)tl141~L2(Rn).h . < E En. . u(2.t ) . t ) to the problem This is just the conclusion of the problem. Show then that the solution u satisfies Il"('. (Indiana) Solution. 2 E B" 2 E B".) b. t ) of the above problem in terms of the initial data. € ) ) t . (You should prove that your u is welldefined.t ) = &-(()(y -i h . a.505 ( L u ) ( z .t > 0. t 2 E En L which admids the following solution qt. t )= u(2 . Transforming (. for an unknown function u ( z . t ) . the above problem we obtain { O(<.> 0. a. ( l€12-exP( The inverse Fourier transform of G gives us the explicit representation for the solution u ( z .O) = m. Assume that 4 is the Fourier transform of a function which vanishes in the ball of radius R 2 0 centered a t the origin. Derive an explicit representation for a solution u ( z . i& + ([(I2 .O)= 4(2).

By using the Green's formula and noting the boundary condition.506 6309 Let R c IR" be a bounded domain and assume u ( x . (Iowa) Solution. Prove that we can find a constant C such that E(1) = I u 2 ( x .e..1 ) d x 5 C regardless of the initial value u ( z . we get from above euualitv --E(t) 2 dt I d where E ( t ) = & u 2 d x . t ) 2 0 is a function with u E C2(o [O. oo)). Using above estimation.0). The Hijlder inequality gives i. Multiplying both sides of the equation by u and integrating on f .which solves the equation x U+. . - AU = -U 4 (heat conduction with heat loss due to radiation) with the boundary condition u Ian= 0. + 1 IVu12dz = - 1 u5dz. we get from (1) I d --E(t) 2 dt 5 -JRJ-+Eqt).we have 2 utudx - Au u d x = - - I u5dx.

T]such that u takes the maximum a t ( z o .to) -sinu(zo. h continuous and g.T) x with g. B ( O .A+u)(zo. + if u(zO. we have E-f(t) It follows that 2 E-+(O) + 3(R(-. T ) x (0) W O .tO) = -(u3 sin(u))(zo. . It is clear that ut(zo.T]. -u3(z0. 6310 Let satisfy ut u3 u = h.u. then there exists (zo.tO) 1.to) 0 2 and Az(zo. if u( zo. t o ) and u(z0. to) > 1.t o ) 2 1. r ) x [O. B(0. we must have But from the equation we get (ut . TI. (Indiun u) Solution. Therefore. TI. LO. This completes the proof. = < 0.t.to) This contradicts (1).to) 5 0. u= 9.507 The above inequality can be written as Integrating the above inequality from 0 t o t . Prove that maxu < 1 on B ( 0 .tO) -1 . h < 1. If the conclusion of the problem is false. + + sin(u) = A.sin 1 < 0.T ) x (O.to) E B(0.T ) x (0.

508 6311 Let u E C2(R x (0. TI. if u take the positive maximum a t ( z o . By applying Green's formula.and integrating the resulting equation on 0.to) 5 0. t > 0.yn) 5 + IlfllLl(n).T]. where R is a smooth bounded domain in En. IIUIIL-(n)(t) L llfllL-(n) for all t E (0.to) 0. u=o z E d R .(n) for all t E (0. the following bound also holds: I I 4 4 ( n . T)) Co(nx [O. udx - u4dx. (c) Prove that the solution t o this boundary value problem is unique. t > 0: u(z.T].O) = f ( z ) 2 E a. (d) Show that for f 6 L4(R) n H1(R).Au)(xO.u3 2 E R . we obtain from above integral equality Then we get the conclusion immediately. (a) Multiplying the equation by u. ( t ) + II~llLl(n)(t) Ilfll.T]) a solution to the problem n be { (b) Prove that ~t = AU . we get Au . (Indiana) Solution. t o )E R x (O. In fact. TI. (b) First we show that u can not take the positive maximum in R x (O. Hence (ut . then we must have ut(zo. (a) Prove that IIUllLz(n)(t)5 IlfllL.to) 2 0 and Au(zo. 2 .

The w solves the following problem: { Wt = AW. we can get Il4lL. x E R. we get 1 utdx = J.u'dx) dt 4 5 J. This implies that u1 = u2.tO) < 0. we have Il. vt E ( 0 7 1 for above problem.and the uniqueness of the solution is proved.(u: + ~ w = 0. In a similar way. w ( z .1 3 ~ ~ I V u (.u2. and w = u1 . 0) = 0.llL-(s2)(t> L llfllL-(ct).l u i d x 5 0 d = . ~ x 5 (1 J. vt E (0. 1 + u ~ )2 w .T]. IVul'dx) 5 0.Tl. By a similar deduction as done in (a). t > 0.Au .J. we can show that u can not take the negative minimum in R x (O.This contradicts the inequality -u3(xO. we have dt ( / . we can obtain f and (a Lu'dx) + dt (1 J.u3utdx and utu3 = J. t > 0. u 4 d x + J. . (c) Let u1 and 212 be both solutions t o the boundary value problem.J.d u'dx <_ 0. Adding the both inequalities above. (d) Multiplying the equation by ut and u3 respectively. ~ x E a R . Therefore. and integrating the resulting equations.u'dx. Au .utdx . u3dx .(ct)(t) 50. x E $2. lVu12dx) = . By applying Green's formula to the first terms in the right side of the both equalities above.

. that u and w are continuous on and satisfy the condition u 5 w on the parabolic boundary n x [O. co). implies the conclusion of (d). 6312 For each p E En. ~ ) E and wt 2 aij(Vw)vzizj for ( z . co). Assume A E C ' ( B n ) . combining with (a) if necessary. Let R C B" be a bounded open set. t )R x ( 0 . we can obtain a differential inequality for w as follow: n n ij=l k=l By applying the maximum principle t o the above differential inequality. we find w 5 o on I x [0. Prove (from first principles) the following comparison principle: Let u and v be C2 solutions t o the inequalities ~t 5 aij(Vu)uzizj for ( ~ . Set w = u . This is the result we need. co) (Indiana) Solution. Suppose.v and Then from the given differential inequalities. respectively. furthermore. t ) E R x (0. A(p) be a real positive definite n x n matrix with let entries aij(p).510 This differential inequality.

(1) then ( z .j=l n a i j ( z . T I . z .t . t . T and E I u(zo. m)).to) R x (0. it must hold that Wt 5 0. u .j=l n for every t E R~. z ) are continuous functions of their arguments and assume that C ~ i j ( z . First we prove that if w E C2(ax [O. z .TI where Q = (R x (0)) U (dR x [O.In fact. p ) t i tLj x ( z . t . let u E C2(nx [0. i. u ) w Z i > 0. ~C~b i ( t . Suppose u satisfies the inequality ut - C i. ~ u ) w .m)) satisfies wt n n C a i j ( z . t . Q .t o )< min u. u.j=l i=l at ( q t )E 0 x ( O . Now suppose that u takes the minimum a t (%'. t .u)uxi o >_ i=l n for z E R. t ) = min u(z. t ) .T]). t > 0. z . -.t .{ o ) . (Indiana) Solution. p ) 1 t 1 2> o i. t ) Q 0x P. where X is a positive function. Assume u i j ( z . a t the minimum point (z. vu)uZixj - C b i ( z . Prove that for any T >0 -min u(z. z . t . p ) and b i ( z . wxi = 0 and which contradicts (1).T]. t )can not be the minimum point of w in R x (O.511 6313 For R c Rnbounded and open.

TI. w = u . t . .512 Take a > 0 large enough so that a u l l ( z .u. t )E s2 x [ o . T ] . t .V u ) > 0. Vu) + b l ( z . This is in contradiction with w having minimum point in R x (0.&eaxl for ( z . Then takes its minimum at some point in R x (O.u.T].if However. for the function w we have E >O is sufficiently small.

= $(tL 0) 1 -?(<I I cK + c21<)22 = 0.^(. it holds with a positive constant M that Set 7 = c<t in the integrand..n arbitrary. u(2. G(<. %(t. +.O) Show that the LP norm of . Therefore. > n is bounded by a constant (Indiana) Solution.O)= 0.t)at time t for p times t l-nlp. = 9(z). t ) be the Fourier transform (with respect to z) of the solution to the n-dimensional wave equation: Utt .513 SECTION 4 HYPERBOLIC EQUATIONS 6401 Assume that g is smooth and in L 1 ( R f l ) .t) = sin(cl<lt).c2AU = 0. ?([) is bounded in E n . we have I and Then g E L1(IR") implies that 2tt K O ) = 0. Transforming the equation and the initial conditions. Let G ( < . Then we get .

Show that .zn). (Iowa) Solution.. i.---.514 p > n ensures the convergence of the integral. = Gl(€).e. = Go(€). utt + l€12G(t. 6402 Consider the Cauchy problem for the wave equation Assume U O . E (a) Use the Fourier-transform t o show that the total energy at time t is constant in time.t) = 0. (a) Taking the Fourier-transforms of the wave equation and of the initial conditions with respect to the variable 2 = (zl. U ~ S. Solving the above initial problem of the ordinary differential equation. . q€. we get < %(€) qt.we have -..t>= Go(€) cos(ltlt) + -sin( I€lt) It I . . E ( t ) Eo. 0) G(€. (b) Let Eo denote the constant energy in (a). The above estimation gives the conclusion of the problem. 0) where = (€l.[n).

+ u. + %.. z E R3...O) z E B3... . = $(z). 6403 Consider the initial problem for the wave equation in three space variables: u..55 1 Integrating the above equation on En and noting that we obtain Then the Plancherel theorem gives us Combining this with (a). Utt = 0. u(z. .O)= qqz).. we finish the proof of (b).t> 0. Ut(2. 2 E B3.

6404 Let R be the upper-half space in lRn R = {(21.516 (a) Write down the formula for the solution of the problem. for 1 1 < 1. O<t<l. 11 2 ly-zl=lO Similarly.and $(z) = { 0. . where k is a constant. 0. we can verify that u vanishes when t = 10 for 1 1 5 7 or 1 1 2 13. (4 (b) By applying the domain of dependence for the solutions t o the wave equation. y Therefore J q5(Y)dS = Iy-zI= 10 J II. (Your answer should be explicit. t ) for all t 2 0. it holds that 2 IyI 2 3 for all y E {y : I . 2<t.2 3 > 0).(Y) = 0. no integrals). find the set of points in lR3 where you are sure that u vanishes when t = 10. we can show that the above equality holds when (c) It is easy to verify that 2 13. when 1 1 5 7. (Indiana-Purdue) Solution.22. (b) If q5 and II. vanish outside a ball of radius 3 centered a t the origin. 0. (c) If 4 vanishes everywhere in R3.23) E lR3.21 = 10). for 15 1x1 5 2. 2 k . 2 2 In fact. find u ( 0 . 0. for 2 < 121.

23)7 23 23 23 23 23 in E3 2 0.x * ~ ~ . (a) Write down a formula for the value u ( x . 2 0. ) given by the above formula satisfies the wave equation and the t initial conditions. Ys=-~/ta-(Yl-zl)a-(Yz-za)a -Y3)dS = 0. t > O . Then the value u(x.t ) at (2. . = ~ t ~ . E aa. < 0. t ) is in R. X E R .~ y l .O) = d(X). (2.= 0. -231. x t E R for (Indiana-Purdue) Solution. it is easy t o see that u(x.t 2 0. --+(% 2 2 .. the domain of dependence for the value u ( x ..) is given by t If both 4 and II. The value u(x. at2 2 U(X. -4(x1. respectively.~Y 3 W -x ~ -/ +(Yl. (b) Find a formula for the value u ( z . E R. t ) of the solution at ( x . z2. when t < 23. t ) = 0. J y .517 Consider the initial-boundary value problem azu azu azu a z U -+ -+ -. 0) = +(x). ax. t ) . ‘$(xl.m. When 2 3 = 0.Y2. ~Y2. ax: ax. (a) When x E R and 0 < t < 2 3 . ) . W ( X .( y+ l( Y l . x E R. satisfy certain compatibility conditions. @(z)= -x3). 2 2 . 2 3 ) .t ) of the solution at all t > 0. < 0. ) is given by the formula for the Cauchy t problem of the equation: (b) Define *(x) and @(x) as both odd functions with respect t o as follows: 4(21.

(i) For t E [O.t) satisfies the boundary condition on = 0.t) be a solution of class C2 to the Cauchy problem with compact support. evaluate lim E ~ ( u ( t ) ) .T]. have we ER-T+t(u(t)) = (u. 23 Therefore. (iv) Let 7~ = 2. 0 for 11 2 2 k.) (Indiana) Solution. on the set T < t . # 0 and suppose that $ ( ~ 1 ~ 2 2 Show that u satisfies ) . ) (ii) When n = 3." IVuj2)dSdp. (Here T = 1. 6405 Let u(2. u(z. when 23 = 0. where 2: = for some constant c. we have J Jy-xl=t @(y)ds = 0.4k.518 In a similar way. + . t+O" (iii) Do (ii) above when n = 1. Define the local energy by (i) Use the energy indentity to show that Conclude that the total energy E W ( u ( t ) is conserved.

(ii) From the formula of the solution to the problem it is easy to verify E ~ ( u ( t =)0 if t > R ) + a.T] It is clear that the total energy E.(u(t)) equals to E. we can calculate that Hence ER-T(U(0)) 5 ER-T+t(U(t)) 5 ER(u(t))Similarly. we can show that for t E [O.519 By the above expression. .(u(O)).

1x1 > 4k.ll.Izl-2k - . the solution is given by If 1x1 < t .1x1 + 2 k < d3 V t .J Therefore. we obtain t .IzI .4k. Hence (iii) By the formula it is easy to evalute t-oo lim ER(u(t))= 0.520 where a is the radius of a ball in which supp4 U supp.4k and IyI I k . is contained. we have IY - 4 I 1 + 2k.2 k ) ' I It is easy to verify that . 4 1 and hence 1 J W J(t + IzI + 2 k ) ( t . (iv) When n = 2. t . it is easy to see that Hence the formula of the solution can be rewritten as When 1x1 < t .

oo)). we can conclude that Thus we obtain a differential inequality for E ( t ) d -E(t) dt I CE(t).521 6406 domain and assume u ( z . have we Here we have applied the Green's formula: By the Poincar6 inequality. . Multiplying the equation by ut and integrating the resulting equation with respect to x in 0.which solves the equation Let R c B" be a bounded with the boundary condition = 0. t ) is a function with u E C2(ax [0. (Iowa) Solution. With prove that there is a C < 00 such that for all t 2 0. Then E ( t ) I exp(Ct)E(O) is an immediate consequence of the above differential inequality.

vt > 0. (b) What is its energy at time t = 10. Find the spherically symmetric solution to the problem. Then w solves the following problem: aZu 2au r ar { w. u(z. wt = -f’(r). the equation can be written as -+ -. T > 0.O) = u. w = 0. and then we can get (b) The energy of the problem is given by E ( t )= 2 It is easy to verify that 1 L. where T’ = z:+zg+zi. . E ( t ) = E(O). t = 0.t > 0. + (ut” IVul2)dz.O) = 2 E B3. z E r B3.(z. = wtt. and f(q) is a C3function on the real line with compact support such that where f’(q) is the derivative of f with respect to q.= U t t . r = 0. -m. (Indiana-Purdue) Solution. w = f ( r ) .t > 0. ar2 Set w = T U ..522 6407 (a) Find the explicit form of the solution of the initial value problem v2u= U t t . Solving this problem. In the symmetric case. r 2 E B3. fo.

. we have It is easy to verify that The area of the intersection {y E R3.O) = g(z). From the Poisson’s formula. Prove that there exists a constant the given data such that C depending only on (Indiana) Solution. Ut(2.523 And therefore = 2F 7 6408 Consider the wave equation in R3: utt . v x E R3. where g E Cr(R3). u(z.O)= 0 .Au = 0 for x E B 3 . Let R and M be positive constants such that and Ig(z)l 5 M . . t > 0.X I = t } n BR Iy 5 The area of aBR.

we obtain 1 lu(z.t) E En+1.t } .5 t o .524 Therefore.t > 0.t . x (Indiana) Solution. Multiplying the wave equation by ut. we can obtain Set 1 E(7)= 5 lr(u.)be a solution of the equation utt -Au = 0. M . Let = {(Z. we have .. 4 r R 2 = .t)l 5 .”+ IVuI2)(x. MR2 47rt t ’ 6409 Let u E C2(. It is not difficult to verify that Therefore.t) : 0 5 t 5 t o and 11 5 t o .Rn+l(x. x Prove that u vanishes in the conical region D = {(x. and integrating the resulting equation with respect to x in R.O) : 1 1 5 t o } in the plane t = 0. t = T } 1 2 1 for 0 5 r 5 to. Suppose t) also that u = ut = 0 on the ball B = {(z..T ) ~ z .

. By using Green's formula. t Ut(Z. u=O 6410 Let U(Z. > 0 ) . /ltnL8u 2 dx = 0.. gdx = 0. then I. Audx = 0. This implies that vo 5 7. Let R > 0 such that It is not difficult to verify that for any t > R..O) u(z. we get the following equality from the equation J. (Indiana) Solution. .525 It is clear that E ( 0 ) = 0. Hence E ( 7 ) = 0. 0) = f(z). Therefore. t ) be a solution of the Cauchy problem utt = AU (Z E B3. g E Cr(nZ3). Assume that f .5 to. Prove that if = dx). in D. we have uttdx = J.

JlzIRs X b) Show that there is a positive constant C such that 1 t .iax)t (/ t- R< ( C<t+R 3l ax) l I 2 for any t > R.gdx # 0. 0) 0) a) Use Fourier transform to show that there is a positive constant c such that u2 dx = c. . .. = 0.526 and by the initial conditions.. ut(x.m a x l u l ? CI for all sufficiently large t. by Cauchy inequality it follows that 21ax J R S 5 (J.. a) Let 6 denote the Fourier transform of u with respect to the variable x = (XI. (Indiana) Solution. zn). 6411 Consider the solution of the wave equation Utt = nu (x E R3. In the other side.t > 0) with data u(x. . This contradicts (1) provided that J.By taking the Fourier transforms of the wave equation and the initial conditions. we can get . = +(x) E Cp.

+ c c t+b and t.’. Then we have This contradicts (a).>l-+.21. t . t ) with respect to the variable 2 = (21. 400. )< la(t. If the conclusion is false. b) Let Q ( t )be the support of u ( z .m a x ~ u ( z .z3) for t > 0. then there exists a sequence { i n }such that t. a s n .)l-+t. .tn)l I IQ(t. ~ 2 This gives Iu(2. . The proof of (b) is completed.527 Therefore we have Then we get where if $0.

t ) = 0. l4<$t 4 3 0. E CF(R3). we conclude that if t > R and 11 2 < t . when t > A.t)dz.t)l = O(t-') Ixl9t as t -+ o) c.t ) d z = 0. c ( 2 f IR3. then u ( 2 . the solution to the 2-dimensional problem is given by u(21. t )= - . Therefore. Determine the large time J b) Let n = 2. we have 9 sup Iu(z. it holds that +.R. u4(2. J u4(2. q5.11. From the Huygens's principle for the %dimensional wave eauation. which implies that t ) = 0. vlzl< et. (Indiana) Solution. 2 2 . 11. a) Assume that R > 0 is so large that suppq5 u supp11. Let 8 behavior of the integral < 1. lxl<@t when t > __ 1-8' R b) From the Poisson's formula.528 6412 Consider the Cauchy problem for the wave equation a) Let n = 3. E C F ( R 2 ) . (21< R}.Show that for any 1 < 1.

if f > 0 for all z.'u. j~~ 1 .where h E L1(B3). t E R.. U ) =j v). The proof is completed. v. 6413 The linear transport equation is following PDE for a scalar function f of seven variables z. 0 j given.( B t + R)2. 2 When t > 5. 1 1 < R } .t)dv+ oo. Show that supp = o(t-3) as t X (Indiana) .): v ft + v . Consider the initial-value problem for this. E IR3. Suppose that 0 5f 0 (z. i. V x f = 0. v.v) 5 h(z). then the same is true for f at all later times .t) = J. (c. we have dt2. where C is a constant. b) Define the local density p by p(. f(Z. t (z E B3. let f ( c .. clear that if it is IzJ5 B t and Iyl 5 R then JY ..e.zI2 2 t Therefore. v.(y.X I I et +R and d2 . a) Show that.529 Let R > 0 be so large that suppllr C { z E IR2.

the forward characteristic curve departing from t = 0. = -xy ( x > 0) u=5 o n x y = 1. v. (2) From (1) and (2). The system of the characteristic equations for the problem is given by .= 0. a) The characteristic curves of the equation are defined by dx -= v dt and along which it holds that dt Therefore. (h diana) Solution.t) It is clear that f 2 0 i f f ? 0. we have 0 = JR3 f ( x . b) From (3).v)dv O This completes the proof.Solution.tv. we get the solution to the initial-value problem 0 f (2.tv.v). if a solution exists. x = a is x = (Y df . + tv. 6414 Solve the Cauchy problem xu. =f ( x . + yuu.

.= -xy ds du s=O:x=a. we have immediately and 21- du du d = . F. The initial data gives C = diately u = -1 43cxy. + 6415 Compute an explicit solution u(x.O) = z2. we get vt + 22121. Set v = u .O) = 22. t ) to the initial-value problem Ut + = 0. The characteristic curve of the quasilinear equation (1) is given by dx . with the initial conditions _dx dY z= y u .= 2v. Differentiating the equation and the initial condition with . where C is a constant. Then we obtain imme- This implies that + u + x y = c. .~ e 'ds = _ -ds (ae'y ds 1 -u2 2 + u). dt . respect to 2. (Y Fkom the above Cauchy problem.531 ds --. U(X. = 0. (Indiana) Solution. W(2. 1 u=5. y=-.

dt From the above characteristic equations. giving explicitly the discontinuity curve and the Rankine-Hugoniot jump condition along it. 05z< 1 . u(z. dt along which u is a constant. < 0. z = P is given by dx =u. + u = uO(P). du _ . 1-2. and P < 0.e.O)= Show that the continuous solution exists only for a finite time. 0. we can obtain immediately v=Hence u= 4t 1' 22 22 4t 1' + + 6416 Solve the following initial value problem. t = 0 : u = u"(p). The characteristic curve departing from t = 0. 2>1. .t 2 > 0. dt where u o ( z )= u(x. 1-p. __ {:: 2 E IR. i.= 0.0 . we have 2 = uO(P)t p. o<p<1. t = o : x = p . 21 = { P> 1. ( Indiana-Purdue) Solution.532 along which it holds that dv . u t + uus = 0.O). and find the discontinuous entropy solution. From (1) and (2). P < 0. 1 .. p > 1.

.e.= -(u+ dt 2 Noting u+ = 0. ' t<X<l. the discontinuous curve is dt .t 1-t 2-t .- and hence x < t.2).2 When t = 1.2 .2' 1 x=-(t+1). 2 > 1. Fig.6. 1. dx 1 .1-x 1-t 1-t' Therefore we get the continuous solution to the problem u = 1.. we have + u-). 1 - dx _Therefore.= 1. the Rankine-Hugoniot condition along the discontinuous curve is given by dt i. u. 0 5 5 1. For t 2 1. the continuous solution to the problem blows up.For 0 < t < 1. 0. 2 t l l .6. 1-1 for 0 5 t < 1 (see Fig. we have p=.

u)--bu=O dX at at4 aU forO<t<T. t ) = U o ( Y . where b is a constant and a is smooth. ii) Show that. aebt 5 u ( z . if 0 5 t l < t2 < T .O) where uo and a are smooth.t)be smooth solution of dU at 8 + a(u)-U = 0 dX = UO(X). . b) Let u(X. y ) is the characteristic through (Y. U(X.e bt v)- dX = 0. 0) ( & + h 3 ( z ) ) l z = y ) 1' t + i) Use this to compute the largest T for which a smooth solution can exist up to time T. 3 > z ( t + 1). 2 < ' ( t + l). for 0 < t < T.534 and the entropy solution to the problem is given by 2 0. then u z ( 4 t . if z ( t . 6417 a) Let u ( z . t ) be a smooth solution of -+a(z. then 00 00 (Indiana) Solution.Y).O). Show that if a then 5 u ( z .0) 5 p for all X. T ) . Then v satisfies the following equation dV at + U(X. a) Let u = ebtv. Show that. t ) 5 pebt for all x and all t E (0.

From the above expression.535 The solution v maintains a constant along a characteristic of the above equation. Y) = a(uo(y)). we can get the Y following conclusions easily. Therefore. a Then it follows from above problem that Therefore. 0) is given by the following Cauchy problem: &. I?”). we have This is just that we want to show.= a(x. If m >_ 0.O). Hence Differertiating above equality with respect t o y. ~ ) .Y)) { = a’(uo(y))ub(y).O) 5 v(z. Y ) = Y. . b) u(x. a Differentiating this problem with respect to y. t ) is a constant along the characteristic z = x(t. FZ(0. it holds that infw(x. y). i) Let m = inf a‘(uo(y))ub(y). y) through (y. { 4 0 .t)5 supv(z. we get g (g+. defined by the equation . the Cauchy problem admits a global smooth solution in ( 0 . X 2 dx dt This gives the estimation in the problem immediately.Y) = 1. we have The characteristic x(t.

T]. - 44YI s))dyds (2') + J_.T]) which are compactly supported and vanish on {t = 7) '. u(z. the largest T for which the problem admits a smooth solution in (0. (Indiana) Solution.y) in the integral. we obtain 00 00 This completes the proof of the problem. y). t ) and &z(t. on x [0. T) is given by mT+ 1 = 0. Consider the following Cauchy problem: 4t + 4.z). t )E P ( I R x [O.T]) defined to be a weak solution of (2) iff is lT : / 00 4 Y . A function u E P ( I R x [O. . 0418 (Uniqueness of weak solutions) Consider the Cauchy problem 'Zlt + 1.(z(t.. . ii) For any fixed t E (0.7']. '1 = 0.uO(Y)4(Y? O P Y = 0 for all test functions 4 ( z . on on IR x [O. IR x {T}. we have 00 Taking the expressions of u. If uo(z) E 0.)(-4t(Y. IR x (0). = f ( t .. prove that the only LP solution of (2') is u G 0.T) = 0. T).O) = uo(z). y) into account.536 If -co < m < 0. on 4(. by the change of variables z = z(t.

TI). ut u ( z .537 where f E C m ( Bx [0. provided uo(x) 0.)f(Y. b) Show that when b $ 0.is dense in L * ( B x [O. one still has the bound T V ( u ( * . 6419 Consider the Cauchy problem for a function u = u(x. As the space Cr(B2 x [0.~ . l . then TV(u(. where x E t B and t > 0. .t). . .. 0) = uo(x) E C(B2).a * . .l). Therefore. Hence we have u 0. We divide the line t = T into subintervals by the points * * . zo..t)u. - .T) . . Suppose that the backward characteristic curve through (xi. (Indiana) Solution. + a(z.^) = uo(ti).)) each t > 0 (where T V = “total variation”)..T]) with a compact support.TI).T]).. the equality (*) is valid for all functions f E L q ( B x [0.* .z . we have 6’J_+_m U(Y.5)dYd5 =0 (*I for all function f E C m ( Bx [O. Then u is a constant along a characteristic curve. ) Then we have xi.x . 2-1.= b(x.’ TV(b(*.T].. + J.. . . dt a) Suppose that b 0. O . n . t for a) Show that if b 0. * .= u(2. meets the line t = 0 at ( < i .T]) which is compactly supported. . zl. The characterstic curves of the equation are given by the equation dx .. It is not difficult to show that the above Cauchy problem admits a solution ~5 E Cm(Ex [O. dt along which the solutions to the equation satisfy du . ) )5 TV(uo(. O ) (i = n.. t ) .T]) with a compact support. ) ... consisting of all Cm-functions with compact support in E x [O. l .. where q = p / ( p . = b(x..t).. 5 TV(u0) t)) 5))ds..

538 Therefore. we get TV(U(. VT > 0. s)ds.T)) 5 TV(UO(')). the above inequalities.. we have u ( z i . When b $ 0. we get . ~ ) From . T ) = uo(ti) + Ir 0 b(zi(s). where by z = Z i ( t ) we denote the characteristic curve through ( z i . b) Let { z i } and {ti} be given as in a).

University of Cincinnati Columbia University Courant Institute of Mathematical Sciences. New York University Harvard Illinois Indiana Indiana-Purdue Iowa Minnesota Rutgers Stanford SUNY. Stony Brook Toronto Harvard University University of Illinois (Urbana) Indiana University (Bloomington) Indiana University-Purdue University (Indianapolis) University of Iowa (Iowa City) University of Minnesota (Minneapolis) Rutgers University Stanford University State University of New York (Stony Brook) University of Toronto University of California (Irvine) UC.539 Abbreviations of Universities in This Book Cincinnati Columbia Courant Inst. Irvine .

Sign up to vote on this title
UsefulNot useful