This action might not be possible to undo. Are you sure you want to continue?
Comprehensive Introduction to
DIFFERENTIAL GEOMETRY
'A
VOLUME ONE Third Edition
MICHAEL SPIVAK
PUBLISH OR PERISH, INC.
Houston. Texas
1999
ACKNOWLEDGEMENTS
I am greatly indebted to
Richard S. Palais
without his encouragement these volumes would have remained a short set of mimeographed notes
and
Donald E. Knuth
without his TEX program they would never have become typeset books
PREFACE
these "notes" truly have become a book. At one time I had optimistically planned to completely revise all this material for the momentous occasion, but I soon realized the futility of such an under taking As I examined these five volumes, written so many years ago, I could . scarcely believe that I had once had the energy to learn so much material, or even recall how I had unearthed some of it. So I have contented myself with the correction of errors brought to my atten tion by diligent readers, together with a few expository ameliorations; among these is the inclusion of a translation of Gauss' paper in Volume Aside from that, this third and final edition differs from the previous ones only in being typeset, and with figures redrawn. I have merely endeavored to typeset these books in a manner befitting a subject of such importance and beauty. As a final note, it should be pointed out that since the first volumes of this series made their appearance in 1 970, references in the text to "recent" results should be placed in context
T
he preface to the first edition, reprinted on the succeeding pages, excused this book's deficiencies on grounds that can hardly be justified now that
2.
Priface to the First Edition
HOW THESE NOTES CAME TO BE and how they did not come to be a book For many years I have wanted to wri te the Great American Differential Today a dilemma confronts any one intent on penetratOn the one hand, one can
Geometry book.
ing the mysteries of differential geometry.
consult numerous classical treatments of the sUbject in an attempt to form some idea how the concepts within it developed. Unfortunately,
a modern mathematical education tends to make classical mathematical works inaccessible, particularly those in differential geometry. On the
other hand, one can now find texts as modern in spiri t, and as clean in exposition, as Bourbaki's Algebra. But a thorough study of these books
usually leaves one unprepared to consult classical works, and entirely ignorant of the relationship between elegant modern constructions and their classical counterparts. Most students eventually find that this
ignorance of the roots of the subject has its price  no one denies that modern defini tions are clear, elegant, and precise; it's just that it's impossible to comprehend how any one ever thought of them. And even after
one does master a modern treatment of differential geometry, other modern treatments often appear simply to be about totally different subjects. Of course, these remarks merely mean that no matter how well some of the present day texts achieve their objective, I nevertheless feel that an
introduction to differential geometry ought to have qUite different aims.
There are two main premises on which these notes are based. The first
premise is that it is absurdly inefficient to eschew the modern language of manifolds, bundles, forms, etc. , which was developed precisely in order to rigorize the concepts of classical differential geometry. Rephrasing everything in more elementary terms involves incredible
x
Priface to the First Edition
Tbe work
contortions which are not only unnecessary, but misleading.
of Gauss, for example, which uses infinitesimals throughout, is most naturally rephrased in terms of differentials, even if it is possible to rewrite it in terms of derivatives. For this reason, the entire first volume of these notes is devoted to the theory of differentiable manifolds, the basic language of modern differential geometry. This
language is compared whenever possible with the classical language. so that classical works can then be read. Tbe second premise for these notes is that in order for an introduction to differential geometry to expose the geometric aspect of the subject, an historical approach is necessary; there is no point in introducing the curvature tensor without explaining how it was invented and what it has to do with curvature. I personally felt that I could never acquire
a satisfactory understanding of differentiable geometry until I read the original works. The second volume of these notes gives a detailed exposition of the fundamental papers of Gauss and Riemann. Gauss I work is now available in English (General Investigations of Curved Surfaces; Raven Press). There are also two English translations of Riemann I s work, but I have provided a (very free) translation in the second volume. Of course, I do not think that one should follow all the intricacies of the historical process, with its inevitable duplications and false leads What is intended, rather, is a presentation of the sUbject along the lines which its development might have followed; as Bernard Morin said to me, there is no reason, in mathematics any more than in biology, why ontogeny must recapitulate phylogeny. When modern terminology finally is introduced. it should be as an outgrowth of this (mythical) historical development. And all the major approaches have to be presented. for they were all related to each other, and all still play an important role.
Priface to the First Edition
At this point I am reminded of a paper described in Littlewood I S Mathematician I s Miscellany. The paper began "The aim of this paper is to prove ... " and it transpired only much later that this aim was not achieved (the author hadn It claimed that it was). What I have outlined
xz
above is the content of a book the realization of whose basic plan and the incorporation of whose details would perhaps be impossible; what I have written is a second or third draft of a preliminary version of this book. I have had to restrict myself to what I could write and learn about within the present academic year, and all revisions and corrections have had to be made within this same period of time. Although I may some day be able
to devote to its completion the time which such an undertaking deserves, at present I have no plans for this. Consequently, I would like to make these notes available now, despite their deficiencies, and with all the compromises I learned to make in the early hours of the morning. Tbese notes were written while I was teaching a year course in dif ferential geometry at Brandeis University, during the academic year
a few graduate students. Most of them were familiar with the material in Calculus � Manifolds, which is essentially regarded as a prerequisite. An acquaintance
196970.
The course was taken by six juniors and seniors, and audited by
More precisely, the complete prerequisi tes are advanced calculus using linear algebra and a basic knowledge of metric spaces.
with topological spaces is even better, since it allows one to avoid the technical troubles which are sometimes relegated to the Problems, but I tried hard to make everything work wi thout it. The material in the present volume was covered in the first term, except for Chapter 10. which occupied the first couple of weeks of the second
necessary to take rest cures of nearly a week after completing Chapters 2,
3, and 7.
term, and Chapter 11, which was not covered in class at all. We found it
The same material could eaSily be expanded to a full year course
XlI Priface to the First Edition in manifold theory with a pace that few would describe as excessively I am leisurely. f or otherwise the second half of these notes would not have been written. grateful to the class for keeping up wi th my accelerated pace. whose expert knowledge saved me innumerable hours of labor. . I am also extremely grateful to Richard Palais.
. 6 20 2. . . I .. . . . the listing for each chapter gives some indication which topics are treated. . . . CHAPTER I. Vector bundles . . . . Addendum. . . . Coo functions Partial derivatives Critical points . . . . . The tangent bundle of a manifold Equivalence classes of curves. . . . THE TANGENT BUNDLE The tangent space of �n The tangent space of an imbedded manifold . . Immersion theorems Partitions of unity Problems CHAPTER 27 31 35 40 42 50 53 3. .. . . DIFF ERENTIAL STRUCTURES COO structures . . . . . . Orientation . . Equivalence of Tangent Bundles Problems . Xlll 63 67 71 75 77 82 84 89 95 . .TABLE OF CONTENTS Although the chapters are not divided into sections. . and derivations Vector fields . MANIFOLDS Elementary properties of manifolds Examples of manifolds Problems CHAPTER . . . and on what pages. .
. . INTEGRAL MANIFOLDS Prologue. Differential Equations Parameter Curves in Two Dimensions . .. classical integrability theorems Local Theory. . TENSORS The dual bundle The differential of a function Classical Multilinear functions Covariant and contravariant tensors Mixed tensors. V ECTOR FIELDS AND DIFF ERENTIAL EQUATIONS Integral curves Existence and uniqueness theorems . 2. 1 35 1 39 143 148 150 1 53 1 64 1 67 1 69 CHAPTER 6. Frobenius integrability theorem Global Theory Problems 179 190 194 198 CHAPTER 7. Forms .CHAPTER XlV Contents 4. DIFF ERENTIAL FORMS Alternating functions The wedge product . . Problems 20 1 203 207 210 215 218 225 227 . The local flow Oneparameter groups of diffeomorphisms Lie derivatives . Addendum Addendum Problems I. . . . . and contraction Problems versus 1 07 109 III modern terminology 1 15 1 17 121 127 CHAPTER 5. Brackets . . . . Differential of a form Frobenius integrability theorem (second version) Closed and exact forms The Poincare Lemma . .
. . . .. . . Addendum. . Oneparameter subgroups The exponential map Closed subgroups Left invariant forms Biinvariant metrics . 239 246 248 253 256 258 261 263 283 CHAPTER 9.. Left invariant vector fields Lie algebras . .. RIEMANNIAN METRICS Inner products . . . .Contents CHAPTER xv 8. . . .. de Rham cohomology Problems . . . Stokes' Theorem . Subgroups and subalgebras Homomorphisms. . The equations of structure . LIE GRO UPS Lie groups . . Geodesic completeness . Problems 37 1 374 376 379 380 382 384 39 1 394 400 402 406 . . Tubular Neighborhoods Problems . . .. . . . 30 1 308 312 316 323 334 341 344 348 CHAPTER 1 0. Riemannian metrics Length of curves The calculus of variations The First Variation Formula and geodesics The exponential map . Integrals over manifolds Volume elements . . . . INTE GRATION Classical line and surface integrals Integrals over singular kcubes The boundary of a chain Stokes' Theorem . . .
. . . Index of a vector field . . . . M aye rVietoris sequence for compact supports The exact sequence of a pair Poincare Duality . . . EXCURSION IN THE REALM OF ALGEBRAIC TOPOLOGY Complexes and exact sequences The MayerVietoris sequence Triangulations . . . . . · · To Chapters Problem . . . . 2 6 7. . . The Thorn class · · · · · · · · · . The Euler characteristic. . . .. 459 467 47 1 472 473 474 475 NOTATION INDEX . . 9. . 477 · 481 . PoincareHopf Theorem Problems APPENDIX A To Chapter Problems Problems To Chapter To Chapter 419 424 426 428 430 432 439 442 446 450 453 I . . . .CHAPTER XV! Contents I I. INDEX . 10 . . .
Comprehensive Introduction to DIFFERENTIAL GEOMETRY A VOLUME ONE .
.
To be precise. ric with which mention it. you can replace "metric space" by "topological space" in our definition. where � is the set of real x = (xl. . This example immediately suggests the next: any open . just �n itself. d(x. which should be consulted if one allows a manifold to be nonmetrizable.xn) with each E �. ] The second simplest example of a manifold is an open ball in �n. . x and some integer lent metric (one which makes it homeomorphic to �n with the usual metric). . for each � n we can take U to be all of �n. Whenever we speak of ffi. we shall assume that it has the "usual metric" T all ntuples he nicest example of a metric space is Eu�lidean n�space ffi. is also a manifold. . consisting of . Clearly. [If you know anything about topological spaces. a manifold is a metric space property: If 0 E �. in this case we can take Indeed. �n supplied with an equiva M is endowed plays almost no role. then there is some neighborhood U of U is homeomorphic to �n .xi)2 .CHAPTER 1 MANIFOLDS numbers.n.y) = �)yi . a hasty recollection of the definition shows that anything homeomorphic to a manifold is also a manifoldthe specific xE The simplest example of a manifold is. and we shall almost never U to be the entire open ball since an open ball in �n is homeomorphic to �n. this new definition allows some pathological creatures which are not metrizable and which fail to have other properties one might carelessly assume must be possessed by spaces which are locally so nice Appendix A contains remarks. =0 [=1 we will interpret �o as M with the following n=::O such that x E M. For n the single point A manifold is supposed to be "locally" like one of these exemplary metric spaces �n.n as a metric space. of course. x' unless another metric is explicitly suggested. supplementing various chapters.
then ¢(V) C �n is an open set containing ¢(x).2 subset Chapter 1 V of �n is a manifoldfor each x E V we can choose U to be some open E U C V. quite naturally. there is an open ball W with ¢(x) E We ¢(V). . Since ¢: V � �n is continuous. though by no means all. and thus open in M. and thus to �n. homeomorphic to W. of course. and U is a neighborhood of x (U contains V with x E V) which is homeomorphic to �n by a homeomor phism ¢: U � �n. which constitute mOst of this chapter. This compli cated little argument just shows that we can always choose the neighborhood U in our definition to be an open neighborhood. The open subsets of �n already provide many different examples of manifolds (just how many is the subject of Problem 24). it is. some preliminary remarks need to be made. the set ¢1 (W) is open in V. Exercising a mathematician's penchant for generalization. ball with x T u we immediately announce a proposition whose proof is left to the reader: An open subset of a manifold is also a manifold (called. Before proceeding to examine other examples. M. Thus x E ¢l(W) C V C U. an open submanifold of the original manifold). ConseIf is a point of a manifold some open set x quently.
3 V must be open. . then V C �n is open and f: V � �n is oneone and continu f(V) c �n is open. U M2 . we di with an equivalent metric Ji such that Ji(x . it begins to appear that. y) < 1 for = min(di. in fact. Theorem 1 is called "Invariance of Domain" . The quickest routes arc probably pro vided by Vick. for example.z) : z = M. An oldfashioned. (It follows that f(V) is open for any open V C V. left to the reader as an that n may depend on the point We next turn our attention to the integer 11 appearing in OUf definition. Notice V in our definition was not open). stated here without proof' ous. if M C �3 is O} U {(x. treatment may be found in Newman.y.) 1. then there would x. we need the following theorem. with metrics d. sO fI is continuous. Singular Homology Them]. The proof that the neighborhood easy exercise (it is also easy to see that if Theorem be an example where the V in our definition must be open is a simple deduction from Invariance of Domain. example.y. I ).y E Mi.z) : x = M={(x. we can define di or J di= i 1 + di M. THEOREM If . But to prove this. given can first replace each all x. and 11 = 1 for points in M2 • This M. For example. Homology Tlteor)1 and Massey. and M2. * All proofs require some amount of machinery. In general. and d2 .Manifolds With a little thought. Topology ojPla ne Sets. by the way. = 1 were false. and f is a homeomorphism. is an unnecessarily complicated device for producing one manifold from two. but pleasantly geometric. for it implies that the property of being a "domain" (a connected open set) is invariant under oneone continuous maps into �n. 0 and z=I} then we can choose n=2 for points in M.
the manifold is not connected. Consequently. For the proof �n is not homeomorphic to �m when of this intuitively obvious assertion we have recourse once again to Invariance of Domain.y) (we assume that which are). then there is a oneone continuous map from ffi. The further deduction. THEOREM. the resulting metric space is called the is a manifold. and M2 are manifolds. then X is acompact. and M2 are disjoint. If M. for if n > 111. however. PROOF For each x E X consider those numbers ball r > ° such that the closed {Y E X:d(x. which is a Odimensional manifold. we note that 11 �n. if not. uncountable discrete space is not acompact (it cannot be written as a countable ::I 111. In these examples.". This construction can be applied disjoint union of the metric spaces Mi. A disjoint union of manifolds M. n can work at a given point x E M. In particular. an union of compact subsets).4 Then we can define a metric Chapter 1 d(x. A manifold has dimension n or is ndimensional or is an nmanifold if it has dimension 11 at each point. a nonopen subset of is unique at each Consider once more a discrete space. sO is any discrete space {o ifx=y if x # y. Although different Il'S may be required at different points of a manifold it would seem that only one M. It is convenient to refer to the manifold M as Mn when we want to indicate that M has dimension n. is left to the reader. locally compact metric space. and M2 are open sets. that the n of our definition x E M. In the new space M. since a space with one point is a manifold. U M2 by if there is some i such that x. defined by the metric d(x.y)= I = {�i(X'Y) d on M = M.m into 2. The same phenomenon occurs with higherdimen sional manifolds. they can be replaced by new sets M. If X is a connected.y):::r} . M is clearly a manifold also. both M. Y E M otherwise i to any number of spaceseven uncountably many. The only compact subsets of such a space are finite subsets. as we see by taking a disjoint union of uncountably many manifolds homeomorphic to ffi. This unique n is called the dimension of M at x. As a first step. We will often need to know that this is the only way in which acompactness can fail to hold.
for some x.. proof is as follows..Y) S r + d(x"x2)}.y) sn} . The so the function r: X consequence Suppose A C . For each i there is a Yi E A such Zi is in the ball of radius r(Yi) with center Yi.y) r}. Let A' be the union of all closed balls of radius r(y) and center y. . some subsequence of the Yi. Then A' is also compact. since X is locally compact). for all y E A. then X is acompact. converges to some point Y E A. Now the closed ball B of radius �r(y) and Let that . {y E X: d(x"y) S r . .d(xj. > 0 is an interval.Y) S r} C {y E X:d(x2. n=1 U {y E X:d(x. which we might as well assume is the sequence itself. and X2 gives (2) � is continuous. this set includes all r > 0. then for each all such x E X define rex) to be onehalf the least upper bound of X= r.Z3. be a sequence in A'. The set of all such. since If not. If. � Z1.Z2. 00 The triangle inequality implies that {y E X: d(xj.Manifolds 5 is a compact set (there is at least one such r > 0.x 2)} C {y EX: d(x2 . so that which implies that S (I) Interchanging x. Since A is compact. This has the following important X is compact.
�r(x) �r(x) 2 > d (x. Moreover. r(y) =:: .  3 = 3 II which i s acompact. .y). Thus A' is compact. suppose that x is a point in the closure of A. Since X is connected. So the sequence Zi is eventually in the compact set B. or I ·dimensional sphere. Now let Xo E X and consider the compact sets A {xo} An+1 An'. it must be that X A. the limit point is actually in the closed ball of radius r (y) and center Y (Problem 10).2d(x y) > rex) � . we present the longpromised exam ples of manifolds. and consequently some subsequence converges. defined by Sl = = {x E �2: d(x.:. After this hassle with point·set topology. The only connected Imanifolds are the line � and the circle. I rex) . Their union A is clearly open. Then there is some Y E A with d(x. Yi) r (Yi)} are contained in B. so x E A. To see this.6 center y is compact. Since eventually the closed balls Chapter 1 ? Yi Y and since the function r is continuous.O) = I }. and A # 0 is ope and closed. S I. � I = = < By (I). {y E X : dey. It is also closed. y) �r(x). . This shows that if y E A n ) then x E A n'.
I) . . There is another way to prove this.sine) is a home omorphism. O) = I} Sl x . X Sl is called the ntorus..{CO. The 2sphere S2. . together with J it shows that Sl is indeed a manifold. 2). It is ob viously homeomorphic to a subset of �4. I) onto the line � x {I} C � x �. is our first example of a compact 2manifold or surface.. The pro jection P from the point (0.!<. it is always necessary to check that use of this notation is valid. n times {x E �n+1 : d(x. I)} � � x {I} explicitly. 2Jr) � Sl defined by J(e) = (cose.Jr) � Sl. We will often denote the point (cose..2]'[]. while Sl x Sl is commonly called "the torus".(Of course. I) may be taken care of similar ly. is also a homeomorphism. and it is also homeomorphic to a certain subset of �3 which is what most people have in mind when they speak of . on [0... or it suf fices to note that Sf is "homogeneous"there is a homeomorphism taking point into any other (namely. In particular . defined by the same formula. illustrated in (0. is a homeomorphism of Sl .sine) E Sl simply by e E [0. . by projecting onto � x {I}. it is even continuous.Manifolds 7 The function J: (0. The point (0. commonly known as "the sphere". though not oneone. better suited to generalization. I)} onto � x {I}: this is proved most simply by calculating P: Sl . an appropriate rotation of �2).) The function g: (Jr. From these few manifolds we can already construct many others by noting that if Mi are manifolds of dimensionlli (i = 1.2Jrl.!2.•   the above diagram. then MI x M2 is an (nl +n2) manifold. Considerations similar to these now show that the nsphere S" = is an nmanifold.{CO.
To provide a mOl'c explicit description of the 2holed torus. more precisely.8 Chapter 1 "the torus": This subset may be obtained by revolving the circle {(O.z) E �3 : (y_I)2 + Z2 al"Ound = 1/4} contained in {CO. has cOlnponents homeomorphic either to the torus or to the cylinder S' x �. a spare homeomorphic to a torus with a hole ('ut out. the region of the plane contained between two concentric circles.).y. the latter of which is also homeomorphic to the annulus..:) E �3 : J' > OJ. The next simplest compact 2manifold is the 2holed torus. The resulting surface. it is easiest to begin with a "handle". called a surface of revolution. we throw .
..h boundary piece of the sphere. ' " :: '=' There is one 2manifold of which most budding mathematicians make the acquaintance when they still know more about paper and paste than about . and which will be referred to as the boundary of the handle. . . and then identifying points on the boundary of the i. I ' . _____ . " I . _____ . It is homeomorphic to the space obtained by starting with the disjoint union of 11 handles and a sphere with n holes. " . The nholed torus may be obtained by repeated applications of this proce dure. " . : : . The 2holed torus may be obtained by piecing two of these together. '83 .9 away all the points on one side of a certain circle. C3EE1. which remains in OUf handle.h handle with corresponding points on the i. .. _____ . :: . " I .. : :. it is also described as the disjoint union of two handles with corresponding points on the boundaries "identified".
 o � .:e) 2cosB � . this boundary is homeomorphic to a circle.) x [I . I ) x (. as investigation of the paper model will show.I.10 � Chapter 1 metric spacesthe famous Mobius strip. the Mobius strip can also be described as [0.) x ( . I) instead .2".1. "identified". . We wish to identify each point x E S 2 with . cos � sine . not to two disjoint circles. . With our recently intro duced terminology. and ( I . 1 sin � ) . .. y ''''e have not yet had to make precise this notion of "identification". . " . we obtain the Mobius strip with a boundary.( 2sine . This can be described by analytically as the image in �3 of the function I: [0. which you "make" by giving a strip of paper a half twist before pasting its ends together.1) � �3 defined I(e. I) I) ° 1  � • .l) = 2 cos e + 1 cos circle of radius 2 .2".�cose.+ 1 cos � sine. If we define Ion [0. I) with (0. but our next example will force the issue.
p. The space which results. so that [p) = [p). there is no subset of �3 which represents it adequately.Manifidds 11 its antipodal point x E S2. the projective plane. 11'2. This just means that the open sets of 11'2 are of the form J(V) where V C S2 is an open set with the additional important property that if it contains p it also contains . We will postpone for a while the problem of defining the metric giving the distance between two points [p) and [q). The precise definition of 11'2 uses the same trick that mathematicians always use when they want two things which are not equal to be equal. for which J(p) = J(q) implies p = ±q. but we can easily say what the open sets will turn out to be (and this is all you need to know in order to check that 11'2 is a surface). is a lot harder to visualize than previous examples. The points of 11'2 are defined to be the sets {p. p} for p E S2. . We thus have a map J: S2 � 11'2 given by J(p) = [p). indeed. A subset V C 11'2 will be open if and only if JI (V) C S2 is open. We will denote this set by [p) E 11'2.
I)I)) ifs=OorI. s # 0. 1) (s.l) (0. we can make things easier for ourselves by first throwing away all points of S2 below the (x. JI (V) [0. since they are identified with points above the (x. J: [0. I) (s. and we must identify each P E Sl with p E Sl .I).I)) { res.1) s 0 all points E x denoted by or x � M given by if = V To get an idea of what 1l' 2 looks like.I)}. 1 J«s. { 0) ::: I}. [(0. which is homeomorphic to the disc D2 = x E �2 : d(x. Squaring things off a bit.1) {(O.12 Owpter 1 In exactly the same way.1) (1. this is the same as identifying points on the sides of a square according to the scheme shown below (points on sides with the same label are identified in such a way that the heads of the arrows are identified with each other).1) (1. The dotted lines in this picture are the key to understanding 1l'2 If we distort the  . so that the open sets of M are of the form J(V) where V is open and contains whenever it contains for = or 1. we could have defined the points of the Mobius strip M to be together with all sets There is a map and C M is open if and only if c x is open.(I.1)) [(1.1) (1.1) (s. )')plane. This leaves the upper hemisphere (including the bounding circle). V (s.1)). y)plane anyway.
is to be identified with the same thing at the lower right. a little experimentation wil1 convince you that this cannot be done (without having the two pieces of doth pass through each other).Manifolds 13 region between them a bit w e see that the front part o f B followed by the back part of A. can still be described mathematically in terms ( B 1 1 I . al though not homeomorphic to 11'2 . Th us to make a model of p2 we just have to sew a circular piece of cloth and a cloth Mobius strip together along their edges.. The subset of �3 obtained as the union of the Mobius strip and a disc. in reverse direction.. in other words. : u of p2 . it is locally oneone. that is. every point P E 11'2 has a neighborhood U on which J is oneone. we obtain a Mobius strip with B l " " A " a boundary (namely. both of which are circles. If this Mobius strip is removed. The projective plane is thus obtained from . we are left with two pieces which can be rearranged to form something homeomorphic to a disc. moreovel. which is a single circle). although J is not oneone. Unfortunately. � " ' A B the disjoint union of a disc and a Mobius strip with a boundary... There is dearly a continuous function f: p2 ?. at the upper left. Such a function J . 3 whose image is this subset.ffi. by identifying points on the boundary and points on the boundary of the disc. the dotted line.
The crosscap together with the disc at the bottom is a topologically immersed 1l'2 I��\ D'\jC � . . We can thus say that 1l'2 can be topologically immersed in �l.. .. with an extra dimension to play around v·. When the resulting figure is drawn up into 3space and the appropriate identifications are made we obtain the "crosscap".. The figures below show that the Mobius strip may be ob tained from an annulus by identifyjng opposite points of the inner circle. . this Can be done in the following way. In �4.. .. the disc can be added so as not to intersect the Mobius strip. explained in Chapter 2). ...14 Chapter 1 is caned a topological immersion (the single word "immersion" has a more spe cialized meaning. . .. . ... . although not topologically imbedded (there is no homeomorphism J from 1l'2 to a subset of �3). ... .. .) This inner circle can be replaced by a quadrilateral. . .. . . Another topological immersion of 1l'2 in �3 can be obtained by first immers ing the Mobius strip so that its boundary circle lies in a plane. however..rith. .. .. . . (This is also obvious from the fact that the Mobius strip is the projective plane with a disc removed. ...
(fhose who know about topological spaces will recognize it as a disguised case of the Urysohn Metrization Theorem. Consider the function J: S2 � �4 defined by J(x. it shows that things like ll'2 . and which the reader should peruse sometime before J't·acling Chapter 3. 'Ie) prove this. Clearly J(p) = J( pl· We maintain that J(p) = J(q) implies that p = ±q. =±x a+b+e ' a ( ) ( ) . 15 'J'he one gap in the preceding discussion is the definition of a metric for p2 . Roughly speaking. The missing metric can be supplied by an appeal to Problem 31.y. first of all (I) II' a.b.b.:.y. are.Manifold.+.xy.e yz= be xz=ac xy= abo #0. b e . whidI ought to be manifolds.z) = J(a.. so we also have hence (3) Using (2).z) = (yz. We have.:. this leads to (2) Now bx y= a ex Z= a (x +y+z)2 = x2 +y2+z2 + 2(xy+xz+yz) = 1+2(xy+xz+yz).e).xz. suppose that J(x. how('ver.) For the present.x2 +2y2+3z2).a+b+c=±x 1+. this gives a+ b+c= ±(x+y+z). we wi)) obtain our metric by a trick that simultaneously provides an imbedding of ll'2 in �4. which will Iatcr be used quite often.
we obtain y = ±b. J('l))' = = = = . Z = ±b.y. (4) also shows that the same sign holds in (7) and (8). = This I11ap is oneone and we can LIse it to define the metric in p2 : ci( [p]. Y Now (4) gives = = 3 ±b .' j([p) ) J(p). 0. ['IJ) d (j([p)).±I). Clearly.±I. = 0 or e = 0 and Thus x = (4) 0 also. ".O). so (5) gives = 3 . j(['l))) d (j(p).z) But y = z = 0 implies (by (I) again) that be = (±I. In this case we have proved our contention without even using the fourth coordinate of f. =3 _ _ y2 ) = = I I. b . = ±c Since J(p) J('l) precisely when p ±'l.y2 (7) 2y2 + 3 (1 y2 ."ith the same sign (which comes from (3)) holding for all three equations. Similarly. we can define j: ll'2 � �4 b.c) These equations clearly contradict = (O.O) or (O. If x then (I) would immediately give y = z = so that #0.O. = 0. _ b2 (8) z (this holds even if y = b = since then z. which completes the proof. (x. and we have (5) But (6) implies that yz = be 2y2 + 3 z2 2)·2 + 3z2 = 2b 2 + 3c2 (6) y2 + z2 b 2 + c2 and similarly for band e .O.c = ±I).16 Chapter 1 so x = ±a. 0. so b (a. Now suppose a = o.
g([x.xy) is a topological immersion of p2 in R 3 . Notice that the [0. representation.x'). or. (O. to a handle we can attach a projective space with a hole cut out. The closest we can come to picturing this is by drawing a cross cap sticking on a torus. what amounts to the same thing. a !\16bius strip.z)) (yz. we can create other surfaces in t he same way as the nholed torus. x' . By the way. = V"ith the new surface p2 at our disposal. it has a nicer.Manifolds 17 Then one can check that the open sets are indeed the ones described above. (I. For example. We can also join together a pair of projective planes with holes cut out. which amounts to sewing two Mobius strips together along their boundary. Consider the surface obtained from the square with identifications indicated below. and famous. it may also be obtained from the cylinder where I) x SI by identifying E I) X SI with is the reflection of x through a fixed diameter of the circle.xz.x) [0. the map ll'2 � �3 defined by the first 3 components of g: j.y. Although this can be pictured as two crosscaps joined together. The image in R3 is Steiner's "Roman surface".
below.x) �  . P3. forming the "Klein bottle". form a single circle. Xl). (O. to which il can then be joined. dividing the sides into thirds.18 Chapter 1 identifications on the square force PI. one of which is shaded. I) x Sl immediately suggests an immersion of the surface. Turning one end of the cylinder around and pushing it through itself orients the lefthand boundary so that is directly opposite (I. and P4 to be identified. P4} is a single point of our new space. The dotted lines Pl P. so that the set {PI. P2. P2. The description ill terms of [0. O B B A P' P4 Rearrangement of the two parts shows that this surface is precisely two Mobius strips with corresponding points on their boundary identified. P3. which separates the surface into two parts.
Manifold. a nonbounded compact manifold is calJed a "dosed manifold". for example. p} for p E S". then aM = 0.x") E�" :x" . the Mobius strip and the disc. we v. Notice that if M is a subset of�". We wi11 stick to the other terminology. we will mention the t'('bted family of "projective spaces". The set of all such x is called the boundary of M and is denoted by aM. the word "manifold" is used lor "manifoldwithboundary". we Can therefore distinguish those points x E M having a neighborhood homeomorphic to IHI". 19 Examples of higherdimensional manifolds will not be treated in nearly such d"tail.. Points no these "boundaries" do not have neighborhoods homeomorphic to ffi.. . it becomes bothersome to use this long designation. We have already discussed some spaces which are not manifolds only because they have a "boundary".:: 0 such that U is homeomorphic to either�" or IHI". 11 A point in a manifoldwIthboundary cannot have a neighborhood homeo morphic to both�" and IHI" (Invariance of Domain again).rill see later that the spaces pll for even n differ in a very important way from the same spaces for odd n.j1J sometimes use "bounded manifold" instead of "manifoldwithboundary". The (closed) halfspace IHI" is defined by !\ manifoldwithboundary is a metric space M with the following property: IHI" = {(xt. The description of the open sets in Il'" is precisely analogous to the description for 1l'2 Although these spaces seem to It)rln a family as regular as the family sn. but.::O}. then all points of M will be boundary points of M. . but \\. but they do have neighborhoods homeomorphic to an important subset of�". then aM is not necessarily the same as the boundary of M in the old sense (defined for any subset of�"). Often. If x E M. indeed. in addition to the family of nmanifolds S".1I. if M is a manifoldwithboundary of dimension < 11.. Projective nspace Il'" is defined as the rol1ection of all sets {p. If M is actually a manifold. . One further definition is needed to complete this introduction to manifolds. A manifold in our sense is then called "non bounded". then there is some neighborhood U of x and some integer . If manifoldswithboundary are studied as frequently as manifolds.
and Y is homeomorphic to X if and only if Y = U. (A path is a continuous image of [0. d) are also metrics and that they are equivalent to d (i.20 Chapter 1 PROBLEMS 1.. (e) Every manifold is locally connected.. Y. The space (X. (a) Every manifold is locally compact. and X. (a) Connectedness does not imply local connectedness. A space X is called locally connected if for each x E X it is the case that every neighborhood of x contains a connected neighborhood. then (X. and a connected manifold is pathwise connected.y) = 1 otherwise. Each X. for each i. with metrics d. where X = U. (b) The integer n in our definition is unique for each x. (c) X is locally connected if and only if components of open sets are open. (a) The neighborhood U in our definition of a manifold is always open. A difficult theorem states that every path contains an arc between its end points.d) � (X.. where the Y. (c) A connected manifold is arcwise connected. while d(x. is homeomorphic to X. y)== 3. (a) If U c � is an inten'al and J: U � � is continuous and oneone. Show that if d is a metric on X. are disjoint open sets and Y. for some i.(x. '" 2.11 but an arc is a oneone continuous image. but a direct proof of arcwiseconnectedness can be given for manifolds. and consequently homeomorphic to the di�joint union of its components.e.y) if x. 7. and d(x. which are open submanifolds. then I is either increasing or decreasing. for i E I. then the dimension of M at x is the same for all x E M. 0') An open subset of a locally connected space is locally connected. n Xj for i # j.joint union of its com ponents. the identity map I: (X. If (X. .d) is a homeomorphism). < I. 5. so every neighborhood of a point in a locally connected space contains an open connected neighborhood.) 4.d. 6. is an open subset of X. (a) A subset of an nmanifold is an nmanifold if and only if it is open.d) (or any space homeomorphic to it) is called the disjoint union of the spaces X. (b) Every manifold is locally pathwise connected.) are metric spaces.. (d) A locally connected space is homeomorphic to the &. X. then both d d!(I + d) and d min(I.y E X. (b) If M is connected.d) is a metric space. d.
then /(inside of A) inside of I( A) .{CO. Show that this last condition is actual1y unnecessary. 12. 0') The analogous condition is necessary for the Mobius strip.A (the "outside of A") is unbounded. Ie)!' 9.A has 2 components. = {x E �n : d(x..O) S I}. 11. . (a) Give an elementary proof that �t is not homeomorphic to �n for n > L (h) Prove directly from the GeneralizedJordan Curve Theorem that �m is not homeomorphic to ffi. 0') Check that 1l'2 is a surface.. (2) If ffi. which is discussed immediately afterwards. and show that it is a homeomorphism. A C is homeomorphic to snI and I: � IR:II is oneone and continuous (so that J is a homeomorphism on A). (First prove c. Every connected manifold (which is a metric space) has a countable base its topology. (a) The text describes the open subsets of 1l'2 as sets of the form I(V). (h) I f c �n is open. then V = V V 10. 14.1I for m ::I n.{(O.) (e) Pl'Ove Invariance of Domain. c �n is homeomorphic to Dn (a) One component of �n . (a) Compute the composition I = s' .. and a countable dense subset.p whenever it contains p. In the proof of Tlleorem 2.Manifold.O. � 13. and the other (til<' "inside of A") is bounded. Explain how the two cases differ. (c) The map I is a homeomorphism. show that the limit of a convergent subsequence or the Zi is actually in the closed ball of radius r(y) and center y. assume (I) (The GeneralizedJordan Cun'e Theorem) If A C �n is homeomorphic to snt.I)} � �nt. and A is the boundary of each. (b) Do the same for I: snt . (a) Check that the metric defined for 1l'2 gives the open sets described in the text. (b) The image I( ) is open. .1I  B B is connected. then �n . . 21 V 8. where V c S2 is open and contains . For this problem. I)} �t x {I} � �t explicitly for the map P on page 7.
A classical theorem of topology states that every compact surface other than S 2 is obtained by gluing together a certain number of tori and projective spaces. (b) Since we can consider snl C S n. and that all compact surfaceswithboundary are obtained from these by cutting out a finite number of discs.O) < J}. Show that �2 to the surface shown at the top of the next page. Show that pn .pnl is homeomorphic to interior Dn = {x E �n: d(x. a hole in a hole in a hole 17. C is homeomorphic . To which of these "standard" surfaces are the following homeomorphic? 15. and since antipodal points in snI are still antipodal when considered as points in sn. (a) Show that pI is homeomorphic to Sl .22 Chapter 1 16. we can consider pnI C pn in an obvious way. Let C c � C �2 be the Cantor set.
but not if n = I. it will be used in Problem 24. (c) This part requires some knowledge of topological spaces. (A) The infiniteholed torus: for all compact C. locally compact Hausdorff space X. Thus� has 2 ends.C. Let 8(X) be the set of all ends of a connected. the "right" one containing all numbers> some N. Consider the following three surfaces. � . An end of X is a function c which assigns to each compact subset C c X a lionempty component c(C) of X .1 23 (a) If C c� is compact. Define a topology on Xu 8(X) by dlOosing as neighborhoods Nc(co) of an end cO the sets Nc(co) = 19. then� . 18.{OJ is not homeomorphic to�m. the "left" component containing all numbers < some N. in such a way that C. This problem is a sequel to the previous one. c Cz implies e(Cz) C c(CIl.K is connected. Show that Xu 8(X) is a compact Hausdorff space.Manifold.{OJ does not "have one end" so�n . A locally compact (but noncompact) space X "has one end" if for every rompact C C X there is a compact K such that C e K e X and X .C has exactly 2 unbounded components. (b) Show that�n has only one end c for n > I. If c is an end of�. X has exactly one end c if and only if X "has one end" in the sense of Problem 18. or always the "right" one. locally connected. (b) �n . co(C) U {ends c: c(C) 20.C. show that c(C) is either always the "left" component of� . (a) �n has one end if n > I. and�n U 8(�n) for n > I? = co(C)}. More generally. What is� U 8(�).
(b) Surfaces (A) and (C) are homeomorphic! Hint: The region cut out by the lines in the picture below is a cylinder. and consider the region between the two pairs. Jll10l0' _ �JlllUlll �Il\ fiJllL � __ I _ e Ie . which occurs at the left of (A). while surface (B) does not. lH 1H 1\( 1\1 �1 ]l11'. Now draw in two more lines enclosing more holes.24 Chapter 1 (B) The doubly infiniteholed torus: • • • (C) The infinite jail cell window: � ••• l�\ r lti r lti r lti r= JJ!11'L :]1\I0110l10l1[ :]1101 01 Ok[ �!l t J!l t J!l t _ _ _ (a) Surfaces (A) and (C) have one end.
23. . . . 25 ···· • • • C) C) () () . For the purposes of this problem we will use a consequence of the Urysohn Metrization Theorem. then e(�2 .A and �2 . i. .• • infinite swiss cheese . .. . (a) If is a connected noncompact manifold. (b) If A c �2 is closed and totally disconnected (the only components of A are points). J j: M J: M x J(x ) n g: M(g(x). then (Compare with Problem 230. define Show that is closed.Manifolds 21. and a � � which = � � x (� x � x . j(x) .A) is homeomorphic to A...Bare nonhomeomorphic if A and B are nonhomeomorphic closed totally discon nected sets.e. . • • C') () () . (a) The three open subsets of �2 shown below are homeomorphic. () () : i i.A and �2 . . (b) The points inside the three surfaces of Problem 20 are homeomorphic.. We define A(n) inductively by A(I) = A' and A(n+l) = (A(n». if { n} is a sequence � co.. (b) There are only countably many nonhomeomorphic open subsets of �. which is eventually in the complement of every compact set. Hence �2 ..) (b) Given a homeomorphism � � x � x . then there is a continuous function � � such that "goes to co at co". . C)··" 24. J M M J: M j(M) M.. . (a) Every open subset of � is homeomorphic to the disjoint union of inter vals. that for any connected manifold there is a homeo morphism from to a subset of the countable product � x � x .. (c) The derived set A' of A is the set of all nonisolated points.J(x».j. � () () () () . .Bto be homeomorphic even though A and B are nonhomeomorphic closed subsets.j C) 22. (c) There are at most C nonhomeomorphic connected manifolds (where c = 2No is the cardinality of �). (a) It is possible for �2 . For each n there is a subset An of � such that A/n ) consists of one point. . ) by goes to co at co.
then aM is a closed subset of M and aM and M . (e) There are c nonhomeomorphic connected open subsets of �2 25.e) is on Steiner's surface.b. (c) If C.'h derived set is {cil. This is not necessarily true if M is not a closed subset.26 Chapter 1 '(d) There are c nonhomeomorphic closed totally disconnected subsets of �'.b. . is aM. < e3 < . i E J are the components of aM. and e) < e. (a. Hint: Let x = be/D. Hint: Let C be the Cantor set. 26. . If M C �n is a closed set and an ndimensional manifoldwithboundary. For each sequence n) < nz < . . etc. (b) If M is a manifoldwithboundary. a sequence of points in C.1/2) and (I /2. c) on Steiner's surface satisfies b'e2+a'e'+alb' (b) If = . abe. (a) Every point (a. one can add a set A i such that its n l1. then M .aM are manifolds. 27. e) E �3 : b'e'+ a'e'+alb' = abe} is the union of the Steiner surface and of the portions (co. . b.' .U.b. as a subset of �n .eI' C. then the topological boundary of M.co) of each axis. (c) The set {(a. and I' c J.e) satisfies this equation and 0 oft D = Jb'e'+a'e'+a'b'. then (a. is a manifoldwithboundary.. (a) A manifoldwithboundary could be defined as a metric space M with the property that for each x E M there is a neighborhood U of x and an integer 11 =:: 0 such that U is homeomorphic to an open subset of lHIn.
and 335. Indeed. . This is certainly not always 27 + . if "' : V � �n is another homeomorphism. but the notion of a differentiable function J : M � � does not. 215. we will denote the identity map from �n to �n by I.I : �n � � is also differentiable. 229. rather than by :n: (which will be used often enough in other contexts). On a general manifold M the notion of a continuous function J : M � � makes sense. 226. so that I i = Xi . however.. Unfortunately.I : �n � �n is differentiable. including some problems.CHAPTER 2 DIFFERENTIABLE STRUCTURES freshly sharpened. which the reader should bring along e are now ready to apply analysis to the study of manifolds. since Wsary tools of "advanced calculus". where differentia bility of functions can be defined. 225. then it is not necessarily true that J 0 ". it would seem reasonable to define J to be differentiable on U if J 0 </>1 : �n � � is differentiable. are contained in Chapters 2 and 3 of Calculus on Manifolds. 332. notably 29. This is the case despite the fact that M is locally like �n. The neces (x) we can expect J 0 1/1 1 to be differentiable for all J which make J 0 </>.' differ entiable only if </> 0 ". We will use freely the notation and results of these chapters. and U n V # 0. If U C M is an open set and we choose a homeomorphism </> : U � �n.
. it will suffice to consider homeomorphisms x: V � xCV) C �n onto open subsets of ffi. we will be interested almost exclusively in functions J : �" � �n which are Coo (that is. It is necessary to adorn our manifolds with a little additional structure. V ). namely. A particular member (x .28 Chapter 2 � the case. If V and V are open subsets of M. First of all. VVe can even imagine a mesh of coordinate lines on U. . The use of the letters X.. we wish to select a certain collection with the property that </>01/!1 is diff erentiable whenever </>. for these homeomorphisms. since x( V n V) and y( V n V) are open subsets of �" . two homeomorphisms x : V � xCV ) C �n and J' : V � reV) c �n are called COOrelated if the maps x ° y I : y(V n V) yo [ I : x(V n V) � � y(V n V) x(V n V) A family of mutually Coorelated homeomorphisms whose domains cover M is called an atlas for M. one need merely choose </> to be h o 1/!. This make sense. but a few refinements will be introduced along the way. . Also. if V n V = 0. where h : �n is a homeomorphism that is not diff erentiable. The only time this notation will be confusing (and it will be) is when we are referring to the manifold �n. for example. and is automatically true. the coordinates xl(p). there is no way out of this impasse. . )" etc. for the obvious reason that it provides a way of assigning "coordinates" to points on U. Moreover. . V ) of an atlas A is called a chart (for the atlas A). �n If we insist on defining differentiable functions on any manifold. xn (p ) to the point p E U. We will often mention the pair (x. . by considering the are Coo. it makes sense. sometimes we will use the words "differentiable': or "smooth" to mean Coo. Among all possible homeomorphisms from V C M onto �n. the precise nature of which is suggested by the previous discussion. which has "coordinates" xl(p). This is preCisely what we shall do. instead of x alone. instead of considering homeomorphisms from open subsets U of M onto �n. xn (p). ./. 1/! are in the collection. just to provide a convenient name for the domain of x . henceforth ad hered to almost religiousl)� is meant to encourage the casual confusion ofa point p E M with x(p) E �n. . each component function Ji possesses continuous partial derivatives of all orders). where it is hard not to lapse back into the practice of denoting points by x and )'. . or a coordinate system on V.
4. is an atlas of COOrelated charts on M.4. Indeed. there is a oneone onto function J : � � � such that namely.4. the obvious map J(x ) = x 3 • Thus (�. . •:. The advantage of this bigger atlas 'U is that the single word "chart". Although (�. The term actually used is x E 'U if and only if x 0 J E V.. where 'U (the "usual COOstructure for �n") is the maximal atlas containing {I}.n parallel to one of the axes. the identity J : �n � �n. together with all charts Coorelated to it. one can easily construct 'U from . (and one would be foolish not to do so once and for all).).. is contained in a unique maximal atlas A' for M. Let . We now define a Coo manifold (or differentiable manifold. or smooth manifold) to be a pair (M. 'U) and (�.4. 29 The simplest example of a manifold together with an atlas consists of �n with an atlas .Differentiable 5lructureJ inverse images under x of lines in ffi. Thus. so . we can adjoin any homeomorphism x : U � V with the property that x and XI are COO.. whose inverse is not Coo.. 'U) and (�. . then . If . PROOF.. and it is clearly the unique maximal atlas containing A ..' is an atlas. where .4. Another example is (�. is a maximal atlas for M..4.4. we can adjoin as many such x's as we likeit is easy to check that they are all COOrelated to each other. LEMMA.' are Coorelated.\"3 . 'U differs only superficially from . of only one map. if U and V are homeomorphic open subsets of �n..4.. V) are not the same. denotes something which must be described in cumbersome language if one can refer only to A Aside from this.. about the simplest example of a Coo manifold is (�n. We can easily make the atlas bigger. V) where V contains the homeomorphism x t? .' be thc set of all charts y which are Coorelated to all charts x E A It is easy to check that all charts in . V) are the sort of structures one would want to call "isomorphic". What has just been said for the atlas {I} applies to any atlas: I .4. 'U). when applied to this atlas..4.
. There are.>4' on any open submanifold N of f. we shall not come close to proving these assertions. This atlas may also be described in terms of the 2n homeomorphisms defined by f. : snI n {x E �n : x .I ) of snI are easily seen to be Coorelated. conversely.A) and (N. its inverse must also be continuous: so that a diff eomorphism is automatically a homeomorphism. 'Li). the definition of diffeo morphism would have had to be more complicated. . . But there are 28 diffeomorphism classes of differentiable structures on S7. and II is clearly a diffeomorphism also. the proof for �3 would ccrtainly be too difficult for inclusion here.(x) = g. two homeomorphic manifolds are necessarily diff eomorphic. x. the projections PI and P2 from the points (0. : sn . which are part of the field called "diff erential topology". 0. X. 2 A proof of the corresponding assertion for � is much harder. and the proof of the essential uniqueness of Coo structures on ffi.::: 5 requires very difficult techniques from topology. They therefore determine an atlac.. . . . . and over 16 million on S 3 1 . Consequently. It will always be understood that �n refers to the pair (�n.n for n . .I n {x E �n : x . unique differentiable structures I on sn for 11 ::: 6.I . up to diff eomorphism. > O} � �n. of course.+ I . . .. xn).(x) = (Xl. the atlas for M is sometimes referred to as the diff erentiable structure for M. . (perhaps most astonishing of all 4 has a diff is the quite recent discovery that � erentiable structure that is not diffeomorphic to the usual differentiable structure!) Other examples of differentiable manifolds will be given soon. In the case of spheres. which are Coorelated to P and P2. .30 Chapter 2 "diff eomorphic": two Coo manifolds (M. However. It is easy to see that a diffeomorphism must be continuous.I g. . we will suppress mention of the atlas for a differentiable manifold. I) and (0. Ifwe had not required our atlases to be maximal. This raises the natural question whether. but we can already describe a diff crentiable structure . . < O} � �nI . 'Li). . .the "usual Coo structure for snb. Normally. rather then differential geometry. and speak elliptically of "the diff erentiable manifold M". . 0. Later (Problem 924) we will be able to prove easily that � with any atlas is diffeomorphic to (�. :B) are diffeomorphic if there is a oneone onto function I : M � N such that x E :B if and only if x 0 I E A The map I is called a diffeomorphism.
a diff erentiable map is continuous. 0 J is (5) a diff erentiable function J: M � N is a diffeomorphism if and only if J is oneone onto and JI : N � M is diff erentiable. of course.). Clearly. Consider first the product M. this coincides with differentiability of the restricted map JIM' : M' � N. and hence J is differentiable if and only if J 0 XI is diff erentiable for each chart x.. J y JR:m is called differentiable at p E M if y o J 0 XI is diff erentiable at x(p ) for coordinate systems (x . x M2 of two diff eren . V) with p E V and J(p) E V.4. V ) and (y. V ) is a diffeomorphism from V to xCV ).4. If this is true for one pair of coordinate systems. . (3) a coordinate system (x. with V C N. V) in . Just as diffeomorphisms are analogues for Coo manifolds of homeomor phisms.' 31 a differentiable manifold (M.. the map y 0 J 0 XI : �n � �m is differentiable. eren A differentiable function J : M � � refers. the atlas . (I) a function J : �n � �m is differentiable as a map between Coo manifolds if and only if it is differentiable in the usual sense.. it is easily seen to be true for any other pair. It is easy to see that (2) a function J : M � �m is diff erentiable ifand only ifeach Ji : M � �m is diff eren tiabl e. to the usual diff tiable structure on �.Differentiable Structure. (4) a function J: M � N is differentiable if and only if each y i diff erentiable for each coordinate system y of N. The diff erentiable structures on many manifolds are designed to make certain functions differentiable.4. More particularly. there are analogues of continuous maps.' consists of all (x. V) for N. V ) for M and (y. We can thus define differentiability of J on any open subset M' e M. as one would suspect. A function J : M � N is called differentiable if for every coordinate system (x .
The existence of Coo functions on a manifold depends on the existence of Coo functions on �n which are 0 oUlside of a compact set. we cali f differentiable when it can be extended to a differentiable function on an open neighborhood of lHln. . we first note that a Coo manifoldwithboundary can be defined in an obvious way. For each pair (XI. x M2 � Mi defined by 7[i(p" P2) Pi . It is only necessary to know when a map f : lHln � �n is to be considered differentiable. 29). We briefly recall here the necessary facts about such Coo functions (c.32 Chapter 2 tiable manifolds MI. x M2 which makes each 7[1 diff erentiable.f. and the two "projections" 7[i : M.culus on Malfif olds. The collection of these homeomorphisms can then be extended to a maximal atlas. there is a differentiable structure on ll'n which makes the map f: sn � ll'n (defined by f(p) [p) {p. . pg. where U does not contain . Consider any coordinate system (x. A "handle" is then a Coo manifoldwithboundary. so that f l U is oneone. It is easy to define a diff erentiable structure on M. Similarly. UI) of coordinate systems on Mi) we construct the homeomorphism = defined by Then we extend this atlas to a maximal one. one needs to know that there are lots of I hem. The map X 0 (f I U)' is a homeomorphism on leU) C ll'n. To deal with Coo functions effectively.p if it con tains p. and any two such are COOrelated. U) for sn. The details involved in this process are reserved for Problem 1 4.pi) diff erentiable. = = To obtain differentiable structures on other surfaces. Ca!. A differentiable structure on the 2holed torus can be obtained by "matching" the diff erentiable structure on two handles.
Then there is a Coo function f : M � [0.:::: 0. and (2) The function i s Coo.. The closure is called the support of f.Differentiable Structures 33 is Coo. e) and 0 elsewhere.I) . and � defined by /e je is Coo. it is positive on (e. Let C c U e M with C compact and U open. e) x . 8) and 0 (3) The function I : � � � defined by is Coo. there is a Coo function k : elsewhere. J(x) x ¢ (1. e(x+I>' x E (I. " x (e. Similarly. (Compare Case 2 of the proof of Theorem 15.. and denoted simply by support f (or sometimes supp f)· {x : /(x) oft O} 2. such that f = on C and support f e U. = 0 for all � defined by = o � which is positive on (0. h: � � hex) { eI /x' x oft x h(n>(O) j: � � { e(xI)' . LEMMA. j(xn ) = . increasing on (0. ]i[ +0 0= 1 x for E 2: 8.) 1) I . e On a Coo manifold M we can now produce many nonconstant Coo func tions.8).1) I � � (I) The function = � defined by 0 o =0 n. 8 > . il is 0 for x (4) The function g: �n � g(x) j(XI ) .
One remark is in order now. If ". < {3. so a CO manifold is just a manifold in the sense of Chapter I.* In f act. emellta1JI Diff ential Topolog)'_ 7 er . + J Pl p. a C� structure on M can always be extended to a Ccx structure in a unique way. choose a coordinate system (x. then the charts of a maximal C� atlas are all COrelated. Q' . . l:. we could have defined C' manifolds for each . 2: I. ).. For each p E C. Since C is compact. The symbol ew stands for analytic. for Q' = (J) the proof * For a proof see 1\. . e ) for some e > o. is Cr ifit has continuous partial derivatives up to order r ) . every Ccx structure contains a CfJ structure for each f3 > 0' . and the sum. . We can also define analytic manifolds (a function J : �n � � is analytic at a E �n if J can be expressed as a power series in the (xi . A "C o function" isjust a continuous function.. of the corresponding functions has support C U. Co manifolds for ". Then xC V) ::l (e. the proof of Lemma 2 produces an appropriate Co filllction J OIl a Ccx lnanifold./) which converges in some neighborhood of a). # co will hardly ever be mentioned again. where I is defined in (3). as in Lemma I. 2: o. and it is convenient to agree that r < 00 < (J) for each integer . not just for 00" . finitely many such neighborhoods cover C. Let J be the extended function. •: " r = By the way. + I . . . x (e. The converse of this trivial remark is a h ard theorem: For Q' � 1. J l + . . but this atlas can always be extended to a bigger atlas of COrelated charts.34 Chapter 2 PROOF. Of course.1unkres.ffi. On C it is positive. e ) x . for 0 :s. Thus. and is positive on a neighborhood of p whose closure is contained in U. The function g o x (where g is defined in (4)) is Coo on V. The function J can be P P constructed for each p.. This will not be proved here. Let J = I o (f + . the CO structUl"e (consisting of all homeomorphisms x: U � �n) being the largest. of course.. but it is unique up to diffeomot·phism..n ?.:::: 00. P p. (A function J : ffi. . V) with V c U and x(p ) = o. so on C it is 2: 8 for some 8 > o. the smaller structure is the "stronger" one. Clearly it remains Coo if we extend � U� it to be 0 outside of V. it is not unique.
. At this point classical notation for partial derivatives is systematically introduced. .. it is fitting that we begin differentiating them. .J(p) h ' {I (Joct!' (0).. with respect to a coordinate system (x. so it is worth recalling a logical notation for the partial derivatives of a function J: �n � �. .J(a) . . an) . ax' (p) = ax' p c. . h The Chain Rule states that if g : �m � �n and J: �n � �. then h.. � (p) = 8}� = 0 if i # j.J(a) the number J(a' . i= I X. then this partial derivative is just If x happens to be the identity map of �n.(J J: M � � and a coordinate system (x.J(g (a» . Dd (a). . 0» x 0 0 x. e) � M by (or simply aI aXl = D.". D. which . I = ' as an equation between functions). so it measures the rate change of J along the curve c.U). . . O lim Now. h.J(p) = aJlax' (p). If we .I (x(p) + (0.I ) define the curve Cj : (e. . ..(h» . ax} h� O r J(c. in fact it is just Notice that ifi = j ax' . then is the classical symbol for this partial derivative. . We denote by D.a' + h.Diff erentiable Structures 35 fails completely (and the result is falsean analytic function which is 0 on an open set is 0 everywhere).)(x(p» . With differentiable functions now at our disposal.U) we define aJ aJ D. .(h) = X..(f 0 x. . What we shall define are the partial derivatives of a dif ferentiable function J: M � �. for a function Dj (f o g) (a) = L D. . .
and this sometimes justifies the sloppiness involved in the definition of the polar coordinate system. y) cos e (x . e) = (r cos e . with its image being the set {r : r > O} x (0. eo + 2". is the use of the symbols alar and alae in connection with "polar coordinates".2". many results are essentially independent of which line is deleted.% (x . 2rr). y). �. y). y) = we can introduce a "coordinate system" P : A ?.y� ."raxis" 2" not restricted to the set A. y) E �2 : y = 0 and x ::: O} where r ex.). y) is restricted to lie in the appropriate interval (eo .� . (Of course. e (x.2 by p(x. One Can delete any ray other than L if e(x.36 Chapter 2 Another classical instance of this notation. yJ This really is a coordinate system on A in our sense.I (r.I is defined simply by p .) with x = r ex. y» . often not completely clarified. the polar coordinate system is often + (r. Jx2 + y2 and e(x.. r sin e). whose inverse p . On the subset A of �2 defined by A = �2 = �2 L _ {(x .. y) y = r ex. y) = (r(x..ffi. y) is the unique number in (0..) We havc really defined P as an inverse function. OJ "Oaxis" � +.. ... y) sin e(x.
y) = DI /(r' (p(x.y» · D.:. y) cos e(x. . y) appears because this curve . y) + Dzf(x. y» by the Chain Rule DI /(x.y)) . A similar computation gives C l _ _ cos8(x. so aJ a. DI J 0 P . y). This is to be expected. along a unit vector v = (cos e(x. of the curve C2 v·. and thus the (x. at the point (x.Y) = DI /(x.� / Xl' \ C2 . y) % direction. y).y) sin e(x. y)[r(x .. y» ( This formula just gives the value of the directional derivative of J at (x.y» + Dzf(P' (p(X. e) = = 37 J(r cose.DifJerentiable Structures From this formula we can compute aJlar explicitly: (f 0 p .'n9(X.sin e(x. y») + Dz f(x. y)) pointing outwards from the origin to (x. y). The factor " (x.(x. . y) · cose(x. "v' _ = ofa curve along the "raxis". is just a line in this direction.:.. sin e(x.I )(r. because CI . y). . . sin e(x.". . y) . DI [ rl )2 (p(x. The vector w = ( . the inverse image under P (x' Y).. y).).I ) ( P(x.'hich is the inverse image under P of a cun.e along the "eaxis". y»). r sin e).y) /8(x. y).. y) [r(x.) /1. cos e(x. y» is perpendicular to v. [r')' (p(x.
U) and (y. and suppressing the argument (x.. of course. y) denotes the identity coordinate system of �2 We have ax lar = COs e.+ ar ax ar ay ar aJ aJ ax aJ ay = + ae ax ae ay ali ' I n classical notation.38 Chapter 2 goes around a circle of that radius as e goes from 0 to 27[. etc. If (x. y) everyvvhere (thus writing an equation about functions).. then on U n V we have (I) PROOF It's the Chain Rule. aJ ay sin e aJ cos e. Using the notation aJlax for D1 !. if you just keep your cool: ar (p) = D.. V) are coordinate systems on M. I t is a pleasure to report that henceforth this may always be done: 3. Note that aJ/ae is independent of which line is deleted from the plane in order to define the function e unambiguously. the Chain Rule would always b e written i n this way.))(y(p)) i > = j =l I: Dj (f 0 xI )([x 0 yI )(y(p» ) . so Qur formulas can be put in the form  aJ aJ ax aJ ay = . (J 0 y . PROPOSITION. a ax ( rsm e) + / In particular. these formulas also tell us what axlar etc. we can write the above equations as �= ae ar aJ = aJ ax aJ cos e + . [x 0 yI )j (y(p» . are.I )(y(p» a I I = D ([J 0 [ ) 0 [x 0 y. and J: M � � is differentiable. where (x.y) times as fast as it should go in order to be used to compute the directional derivative of J in the direction w. D. etc. so it is going r(x.
and any coordinate system (x.(pl· . n x PROOF.I) (y(p» At this point we could introduce the "Einstein summation convention". and felicitous choice of notation. I 4.DifJerentiable Structures 39 = j= = :L Dj (J 0 [ I)(x(p» 1 1 j= aJ ax} :L � ( p ) . a)'i p ' I thus a ax j a = (p ) ax j P . often with hoards of indices being summed over. which appears both "above" (in axj lay') and "below" (in aJlaxj ). I won't use this notation because whenever I do. I soon forget I'm supposed to be summing. No tice that the summation in this formula occurs for the index j. There are scads of for mulas in which this happens. . and the convention is to omit the L sign completelydouble indices (which by luck. the operator f = alaxi lp satisfies f(Jg ) = J(p)f(g) + f(J)g(p ). the nature of things. U) and (x'. almost always occur above and below) being summed over. ax) ayi n · Di [Xj 0 y. U) with p E U. U') are two coordinate systems on M. We will often write formula (I) in the form a ay i = I: ayi axj . The operator taking J to aJlay'(p) is denoted by a .:. ay' P J= l ayi n I I: For later use we record a property of f = alaxilp: it is a "pointderivation". If (x. PROPOSITION. J= I n axj a here alay' is considered as an operator taking the function J to aJlay' . one can avoid what Elie Cartan has called the "debauch of indices". •:. and because by do ing things "right". g : M � �. the n matrix . Left to the reader. For any differentiable J.
� (P») clearly does not depend on the coordinate system (x. where v(Bn ) is the volume of Bn . however. which we merely state. is still a "small" subset of �'. On the other hand. In this case. V). To do this we need a lemma. in fact. it is called a regular point of f. so I is again constant. � �. . the rank of the m x 11 matrix (::. I(x. b). We want to define the same concept for a subset of a manifold.a1 [a. The most important theorem about critica1 points generalizes this fact. V) is a coordinate system around I(p). b) to be critical points. we will need some terminology. . It is possible for all points of the inten. y) = x. if q ¢ I(M) . of (closed Or open) rectangles with I and n=1 L v(Bn) < e. U) or (y. a function I: �. which in turn depends on a lemma from Calculus on Manif olds. If I : � � �. . It is called the rank of I at p.40 Chapter 2 is just the Jacobian matrix of x' 0 inverse is clearly XI at x(p). This is true. Other points in N are regular values. . for example. then x is a critical point of I if and only if j'(x) = O. its Now if I : M n � N m is Coo and (y. although this can happen only if I is constant On [a. � � has all points as critical values. Recall that a set A C ffi. may have all points as critical points without being constant. the value I(p) is called a critical value of f. thus q E N is a regular value if and only if p is a regular point of I for every p E lI (q). I = 0 everywhere. To state it. If p is a critical point of f. then D I = D. if p is not a critical point of I. in particular. If I : �. The point p is called a critical point of I if the rank of I at p is < 111 (the dimension of the image N). It is nonsingular.a nonvalue of I is still a "regular value".1l has "measure zero" if for every e > 0 there is a sequence B" B" B3. the image I(�') = � x to} C �.
by Lemma 6. then J(A) C N has measure 0. or has only countably many components. with A c Ui Vi. Using Lemma 6. COROLLARY. . Since J (Ui Vi ) is contained in the Ullion of at most countably many components of N. K on A for i. it is easy to see that if A C M has measure 0. Then IJ(x) . Conversely. If (y. Vi) with A c Ui Vi and each set xi (A n Vi) of measure 0. y E C. (. . This clearly implies that J(A) has measure ° if A does . then J(A) n V = Ui J(A n Vi) n v. = y 0 J o [l(x(A n Vi» . V ) is a chart on N. 6. We can assume that A is contained in a compact set C (since �n is a countable union of compact sets).:0 n2 Klx . then J(A) has measure o. LEMMA. If J : �n � �n is C1 and A C �n has measure 0.yl for all x. (But if M is the disjoint union of uncountably many copies of �. j = 1 . and A consists of one point from each component. . V). PROOF. such that each set xi (A n Vi) c �" has measure 0.y l for all X. Thus J takes rectangles of diameter d into sets of diameter ::. Lemma 6 thus implies another result: 7. Thus y(J(A) n V ) has measure 0. y E A. n2 K d. A subset A of a Coo nmanifold M has measure zero if there is a sequence of charts (XI. If J : M � N is a C I function between two nmanifolds and A e M has measure 0. then A does not have measure 0). . Vi). n . then x(A n V) C �n has measure ° for any coordinate system (x.J(y ) 1 ::. •:. LEMMA. Lemma 5 implies that there is some K such that IJ(x) . it follows that J(A) has measure 0. There is a sequence of charts (Xi . Let A C �n b e a rectangle and let J : A � � n b e a function such that I DjJi l ::. PROOF. Each set y(J(A n Vi) n V ) has measure 0. then it follows easily from Theorem 12 that A has measure 0. if this condition is satisfied and M is connected. n 2 Klx .Differentiable Structures 41 5 .J(y ) 1 .
. PROOF It clearly suffices to consider the case where M and N are �n .In. . V) such that y o J o x. O) . and its image could. (I) If J: M n � N m has rank k at p. . we can insure that J keeps the first k components of a point fixed. part (I) says that by first performing a diffeomorphism on �n. then there are coordinate systems (x. * For a proof. The stronger version ofSard 's Theorem . . . . since J may not even be oneone on ffi. O. . 9.42 Chapter 2 8. then there is some coor dinate system (x.14 of Calculus on Manif olds.k + t (a). . THEOREM. . XI k (a t . These diffcomorphisms on �n and �m are clearly necessary... . and M has at most countably many components. which we will never use (except once . . Remark: The special case M = � n . ". U) around p and some coordinate system (y. . . a n ) = (a t . . 0). If J : M � N is a C1 map between nmanifolds.t (a t . a . Theorem 8 is the easy case. and hence in a neighborhood of p. a n ) = (a t . . . Lectures 011 f'l" Diff erential Geomel1)'. If y is the identity of �m .k x to} c ffi. because some k x k submatrix of (a(yi 0 J)/axi) has nonzero determinant at p . . ". contain only points with first coordinate o. . . THEOREM (SARD 'S THEOREM). SC'T 1\. V) around J(p) with y 0 J 0 x t in the form yoJ0  I\1oreover. and the case m > n is the trivial case (Problem 20). m (a» . states' that the critical values of a C k map J: M n � N"' are a set of measure 0 if k 2: 1 + max (n . . It should be noted that J must have rank 2: k in some neighborhood of p. and then permuting the coordi nates in �m. for example. . which gives only local results.1i1IIor. in Problem 824). . U) and (y. More exact information Can be given when J actually has rank k in a neighborhood of p.n. for the present we are more interested in knowing what the image of J : M � N looks like lo�ally. . ak . in terms of the rank k of J at p E M. . (2) If J has rank k in a neighborhood of p. But this case is just Theorem 3. then the critical values of J form a set of measure 0 in N. given any coordinate system )" the appropriate coordinate system on N can be obtained merely by permuting the component functions of y. . Although Theorem 8 will be very important later on. . Topolog)' From the D{ff miiabtc Viewpoint or Sternberg. N = �m i s equivalent t o the general theo rem.
. . fl (p») .. x0 u (p).. . . .. __ ( a(v.. . . . � = I .an) = (a'. (I) Choose some coordinate system u around p. . { ya 0 f(q) = aar a = I. .n.. .. k r = k + I. . hence x.an). the lower square in the matrix . 43 fORn) JRn. . .. has the form in (I). . ). Since rank f = k in a neighborhood of p.(q) = a. . . 0 /» ) = ax} ." o a. .. k .an) = y 0 f(q) for q = x'(a'.. Condition (I) implies that (2) This shows that i s a coordinate system i n some neighbor hood of since (2) and the Inverse Function Theorem show that u ' is a diff eomorphism in a neighborhood of Now x = (x 0 u') 0 u p.DifJerelltiabLe Structures In part (2) we must clearly allow more leeway in the choice of Y. . since may not be contained in any kdimensional subspace of PROOF. k hence ur (q) = a r = k + I . ..an) means x(q) = (a'. . q = x'(a'.. . (I) det C au Define xa = ya 0 f xT = uT a = I. . . n. . (2) Choose coordinate systems x and v so that v 0 f 0 x . By a permutation of the coordinate functions ui and yi we can arrange that (ya .. . (a'. .ak. so y o f 0 x...
an) = y (al. . 0) . . Thus we can write r = k + l. Nm has rank m at p... v(q) = (bl. . . . . 1/rk+1 (a). . then for any coordinate system (y. bm) the Jacobian matrix y is a coordinate system in a neighborhood f(p yofox'(a'. (1) If m S n and f: Mn . . 1/rk+1 (a) _ 1fk+1 (al.. . . THEOREM. ak» by (3) = (a 1 . .an) = y o vI o v o j oxl(al. .. vk). around f(p). . m: V) U) . .44 Chapter 2 must vanish in a neighborhood of p. . Moreover.. . . 0 V I ... . " " 1/rm(a) _ 1fm(al.. .ak : Theorem 9 acquires a special form when the rank of f is n or 1 0.ak. . . . ak). 0. 1/rm(a» = (al. . .. . Define ya = va yr = vr _ 1fr for Since 0 (V i . . ak.. . • . m .. ... . there is some coordinate system (x. • .. . . around p with has nonzero determinant. ... so of ).
. fl (p») .an) (a'. Even if we clo not perform ¢' first. .. then for any coordinate p.. i t i s only necessary to observe that when k = In.. only the need be permuted. .. . C �)�(1/r:(7nIT{(f »:12 0 f(a'.an) (a' . . so A 1/1 is the desired If we are given a coordinate system x on ]Rn other than the identif we just define y. y = 0 y. This can be conected by another map on ]Rm. .. > II (y. U) around PROOF.. . .. ...bm) '" (X(¢I (bI . .. .an. .'" . . . .an. . .. � = l .. V) = . . the rank of equals n in some neighborhood of It is convenient to think of the case and N '" ]Rm and produce the coordinate system y for ]Rm when we are given the identity coordinate system for Part (2) of Theorem 9 yields coordinate systems for ]Rn and 1/1 for ]Rm such that u.O). . A(bI . . bm). f M '" ]Rn ]Rn. . to permute the yi in order to arrange that det f: Mn Nm has rank at p.an. 45 (I) This is practically a special case o f (I) i n Theorem 9. . 1/1 f o¢'(a'..O. .an) = A 1/1 f o¢'(b'.. .. . it is clearly unnecessary.. .. m.bn+'. 0). it is easily checked that A ljf is now the desired •:. a .bn). .O. Define A by x f 0 <� f x = ".. = = . bn)). 0) . .bn) '" ¢(a) '" A(b'.. C(Yaau.O..... (2) Since the rank of at any point must be :s n. .60 f p. .. . . in the proof of this case. .. . . 0) 0 y. .. .Differentiable Structures (2) If n :s m and system (x. the map f still takes ]Rn into the subset ¢ f(]Rn) which 1/1 takes to ]Rn {OJ ]Rmthe points of ]Rn just get moved to the wrong place in ]Rn {OJ. there is a coordinate system around f(p) with y o f ox'(a' . (CPI (bI .O.bn.. ..bn) for (b I . .O... 0 Then A 1/1 0 0 .O) (a I ..
Although the second and third immersions f31 and fh are oneone. it is case (2) which most interests us. The simplest example is the function f : > defined by f(x) = x3. Another example is IR IR x>O x < x=O o.46 Chapter 2 Although p is a regular point of f in case (I) of Theorem 10 and a critical point in case (2) (if 11 < 111 ). a differentiable map f need not be an immersion even if it is globally oneone. One can easily define a similar curve whose image looks like the picture below. Of course. their images are not homeomorphic IR . the curve itself manages to be differentiable by slowing down to velocity 0 at the point (0. and it is clear from Theorem 10(2) that an immersion is locally oneone (so it is a topological immersion. I g (x)l). A differentiable function f : M n > N m is called an immersion if the rank of f is 11. Three immersions of in 1R 2 are shown below. although its image is the graph of a nondiff erentiable function.(0) = o. at all points of M. A more illuminating example is the function h : IR > 1R2 defined by h(x) = ( g (x). On the other hand. with /. as defined in Chapter I). the dimension of the domain M. it is necessary that 111 ?: 11.0) .
g(q) = g(q'). In general.DifferenliabLe SlTucluTes 47 to 1R. . second time around shows that may even be a dense subset of Despite these complications. Coo �+ f Coo p V C p 0 i Vn . : > S X S M MI M. M VI = {q E M : yk+' (q) = . MI M MI M IR M M). such that nV OJ. . where { yr(q')) = 0ya (q) a = kl. > �!r! u. even if the oneone immersion f : P 7 is not a homeo morphism onto its image.+. MI M. with Mr) in aThus. I I. r i. MJ Mn VI MI E M" (y. . if N is the manifold neigh borhood of a point E M" then there is a function g on a neighborhood M of such that g = g on MI we can define ya (q' = = .n. . = yn(q) = this consequence Theorem = g : MisI an immediate(considered as aoffunction on 10(2). there is certainly some metric and some differentiable structure on f(P) which makes the inclusion map i : f(P) > an immer sion.. . The following picture. . indicating the image of an immersion fJ. a subset C with a differentiable structure (not nec essarily compatible with the metric inherits as a subset of is called an immersed submanifoId of if the inclusion map i : > is an immersion. l.k. if is a k dimensional immersed submanifold of and is a neighborhood in of a point P then there is a coordinate system V) of around p. Of course.
. Given P E P. yk l U. this is the only thing that can go wrong: 1 1 . . PROPOSITION. For example. as a map into M. we may not be able to define g on M. .' : fJ. then f is also Coo considered as a map into Mj • PROOF. Actually. c M is an immersed manifold. this cannot be done if g is one of the functions rf. . We want to show that j. J is a coordinate system of M. and (y ' I U" .' 0 f is Coo ifit is continuous. . on U. choose a coordinate system (y. The following figure shows that f might not even be continuous IR. If M. > M he the inclusion map. V) for M around f(p) such that U. = {q E V : yk+' (q) is a neighborhood of f(p) in M. Let i : M. = yn (q) = OJ .. with its Coo structure. f : P > M is a Coo function with f(P) e M" and f is continuous considered as a map into M] .48 Chapter 2 On the other hand. even if g is Coo on all of M. . C M is an immersed submanifold. = .(M) > One other complication arises with immersed submanifolds. If M. and f : P > M is a Coo function with f(P) e M" it is not necessarily true that f is Coo when considered as a map into MI .
' 12.' (y) is an (II . There is one way of getting submanifolds which is very important.' ( U.:.m)dimensional submanifold of M (or is empty). f : P 7 M which are homeomorphisms onto their image. = Left to the reader. . lS also a closed subset of M. defined as {x : Ixl 2 I}. but we first develop some of the machine. If j : M" > N has constant rank k on a neighborhood of j' (y). if y is a regular value of j : M" > N m . In particular. . •:. > M is an imbedding. It is to be hoped that however abstract the notion of Coo manifolds may appear.{OJ c )R". it is called a closed submanifold of M if M. C M is called simply a (COO) submanifold of M if the inclusion map i : M.' (y) is a closed submanifold of M of dimension n . ) C P is open. . PROOF. is open in M" this means that j.' C IR" .DifJerentiabLe Structures By assumption. then j . \� � • •• % submanifold . submanifolds of )R N will seem like fairly concrete objects. . . then j . Since all y j 0 j are Coo. Such an immersion f\1ost of these difficulties disappear when we consider oneone immersions is called an imbedding ("embedding" for the English). We will prove this fact only for compact manifolds. An immersed subman ifold M. i. as a special case. yk are a coordinate system on VI) the function f is Coo considered as a map into MI . so 0 j' i (open set) is an open set. .I 0 49 f is continuous. and gives the sphere S" . Now it turns out that ever (connected) Coo manifold can be imbedded in some )RN .y which would be used in . PROPOSITION. .k (or is empty). so that )' manifolds can be pictured as subsets of Euclidean space (though this picture is not always the most useful one). and y ' . Since U. Thus j takes some neighborhood of P E P into U.
Unfortunately. . Now M is the union for i > I of the "annular" regions Ai Vi . U C2 U C3 U . Clearly C.lhich refines (9 and which is locally finite. Let V_. each contained in some member of (9. A cover (9 is called locally finite if every p E M has a neighborhood W which intersects only finitely many sets in (9 . • • • with M C. we obtain open sets Vi. Then V. If (9 is a cover of a space M. U C2 has an open = neighborhood U2 with compact closure.2 . I f (9 i s a n open cover o f a manifold M . then there is a n open cover (9' of M which is locally finite and which refines (9. By Theorem 1. we can obviously cover Ai by a finite number of open sets. C3 . and each contained in Vi = Ui+ 1 . with compact closure. compact and U. C Uj+1 . has an open neighborhood V. 1 3. THEOREM.Ui. <. . We can also choose these open sets to be diffeomorphic to In this way we obtain a cover (9' v. PROOF We can obviously assume that M is connected. j = IRn. Moreover.50 Chapter 2 the general case. since we will need it later on anyway. and hence is M.Ui . since a point in Vi is not in V for j 2: 2 + i. with U. Since each Ai is compact. Continuing in this way. = Vo = 0. whose union contains all Ci .. . a cover (9' of M is a refinement of (9 (or "refines (9 ") iffor every V in (9' there is some V in (9 with V C V (the sets of (9' are "smaller" than those of (9)a subcover is a very special case of a refining cover.l . there are many definitions and theorems involved. .2. we can choose all members of (9' to be open sets diffeomorphic to IRn. there are compact sets C" C2 .
Then it is possible to choose. .. .V]. 15. = 1 for all P E M (this sum is really a finite sum in some . then intersects only finitely many members of (9. Let W be an is a closed set contained in and M = i. We can clearly assume that M is connected. one for each in (9. Let (9 Then is a closed set contained in and M = open set with C C C Now C] V] .. THEOREM. V{ UVC2. by (I)). Let (9 be an open locally finite cover of a manifold M. Let be an . Let (9 be an open locally finite cover of a manifold M. }. ' . Un+i U�+j P E by cannot possibly eliminate p.. . open set with c c c Continue in this way. <. C] U V2 U V3 U = ) = V'V {V]. For any p E M there is a largest n with p E because (9 is locally finite.DijjewztiabLe Structures 51 Notice that i f (9 i s an open locally finite cover o f a space M and C M is compact.V2. U V� U (Vn+] U V. Now P E it follows that since replacing C2 V2 V2V2. C C 14. for each in (9.. THEOREM (THE SHRINKING LEMMA). V{UV2U " ' . . C] V{ UJV]. (2) "L <Pu (p) u neighborhood of p. U V3 U n V{ U V2 U·· .(U2 U V3 U . V3 . This shows that an open locally finite cover of a connected manifold must be countable Qike the cover constructed in the proof of Theorem 13). V' V' V PROOF. . V2. an open set with is also an C in such a way that the collection of all open cover of M. I ] .. . Then there is a collection of functions <pu : M . ..+2 U··· ). such that Coo V (1) support <Pu C V for each V. [0. V{ .
C. COROLLARY. where since it is a finite sum in a neighborhood of each point. Choose the U' as in Theorem function 1/Iu : M > [0.52 PROOF. THEOREM. then there is an imbed . Let C C C C C.(p) # O} is locally finite. 1 ] 14. Since the U ' cover M . I] such that (I) the collection of sets {p o ¢. • . If (9 is any open cover of a manifold M. Chapter 2 Case 1. = Coo (2) L ¢. If M" is a compact cling ! : M > for some N.(p) 1 for all p E M. Ue(') L 1/Iu Case 2. it is easy to see that support f e u. and ¢u(p) = 0 when U C Va . : M > [0. I]} satisfying (1) and (2) is called a partition of u nity. V } has an open locally finite refinement (9 to which Case 1 a applies. This case can be proved in the same way. Since (9' = { U E (9 : U C Up for some pl. C U. clearly ! (p) = I for all p E Using the fact that (9 is locally finite. jRN Coo manifold. 1 7 . provided that Lemma 2 is true for c U e M with closed (but not necessarily compact) and U open. �. : M > [0. then there is a collection of functions ¢. VeC}' L ¢U. Each U in (9 has compact closure. General case. != L u ¢u(P) = I for all p. (A collection {¢. (3) for each i there is a U E (9 such that support ¢. it is called subordinate to (9. 1 6. clearly Coo Define Ue(') L 1/Iu > 0 ¢U = everywhere. Cover Mwith open sets Va having compact closure and contained in M The open cover {Up.) It is now fairly easy to prove the last theorem of this chapter. if it satisfies (3). : This sum is Coo. Apply Lemma 2 to U' c U e M to obtain a which i s I o n U ' and has support C U. But this is a consequence of Case 1: For each p E choose an open set Up C U with compact closure.
This is an immersion. show that all charts in ... Moreover.. This shows that we must have q E Vi . JR. ... (a) All Coo functions are continuous. Vl ). Define f : M . the N x II Jacobian matrix It is also oneone.. and on Vi. There are a finite number of coordinate systems (Xl . as claimed. (Xk. 211 + I . PROBLEMS 1. . .Ai are COOrelated. (b) Show similarly that aM is welldefined for a Coo manifoldwithboundary M. Vri ' Xi (P) = Vri ' Xi (q) . where VIi = I. . There is some i such that p E Vj. Vkl with M = Vl U · · · U Vk . 4... For suppose that f(p) = f(q)... in fact.. 1] which are 1 on Vj and have support C Vi. and functions Vri : M .1. [0. (b) A function f : M . (a) If M is a metric space together with a collection of homeomorphisms x: U 7 whose domains cover M and v·"hich are Coorelated. so also Vri(q) = I . How many distinct Coo structures are there on 1R? (There is only one up to diffeomorphism.Di/Ji:renliabLe S'lruclures 53 PROOF. (b) In the proof of Lemma I. jRn 3. Then Vri(p) = I.. by JRN . (a) Show that being COOrelated is not an equivalence relation.) . that is not the question being asked. so p = q) since Xi is oneone on Vi . we can always choose N = ar ( axf ) contains the 11 x 11 matrix (ax�) axf ' . •:. show that the 11 at each point is unique wi/houl using Illvariance of Domain. and the composition of Coo functions is Coo. because any poiIlt p is in V! for some i. N is Coo if and only if g 0 f is Coo for every Coo function g : N . Choose V: as in Theorem 14. 2. Problem 333 shows that. where N = lIk + k.
h) ... M.. defined by Ji x h(Pl ....... M is an imbedding. (a) Show that (MI 7. 6.. a map / : N . PI ... then there is a unique Coo structure on aM such that the inclusion map i : aM .e. MI and 1f2 0 / : N . q E M.' is maximal for N if . 9.. Let g : ..4. h to Coo functions in a neighborhood of then Djg and Djh are the same at points of x {OJ (so we can speak of Dj / at these points). (b) It is even possible to choose c to be oneone. (d) If Ji : N . M is Coo if and .. 1 ] ...4. .. Check that the two projections PI and P2 on homeomorphisms Ji and g. (a) If N C M is open and .. V) in A (c) Show that the inclusion i : N . Ml x M2...4. with U C N. x M2) X M3 is diffeomorphic to MI x (M2 x M3) and that MI x M2 is diffeomorphic to M2 x MI . If / : 1HI" . If M is a Coo manifoldwith boundary. show that . i. p snl p are Coo related to the 2n p2 p sn snIP'n IRn lHIn IR IR lHIn.. is locally Coo. Ml x M2 differentiable for all fit E Ml .... (c) More generally. i s locally C oo (every point has a neighborhood on which / is COO).) (b) If / : ... M is Coo... show thaI h( P2) ). M2 are Coo.... IRn1 1 2... and that .. the Coo structure we have defined for Ml x M2 is the only one with this property..... 2).... M2 . (b) The diffcrentiable structure on Ml x M2 makes the "slice" maps 8.. (a) If U C p IP'n i s open and / : U . M. (a) If M is a connected Coo manifold and . is maximal for M. [pl...... ( PI . h) : N . U) in .) IR lHIn.' can also be described as the set of all (x I V n N. (b) Show that .4. p? (pt}. then there is a Coo curve c: [0.4. 1 0 . are Coo (i = 1 ... M is Coo. h E M2. (iit .. Show that / : only if / 0 g : . Compare the rank of / and the rank of / 0 g . has two extensions g.Ai consists of all (x. M with c (O) = and c ( I ) = q .. then / is Coo. (Obvious. P2 ) = (Ji is Coo and determine its rank in terms ofthe ranks of ji.. .. V n N) for (x... /1 x 12 : Nl x N2 . then / is Coo.. P2) of Ml . / can be extended to a Coo function on a neighborhood of (Not so obvious.54 Chapter 2 5.. Ml x M2 is Coo if and only if the compo sitions 1fl 0 / : N .. Ml x M2 at in terms of the ranks of Ji at For Ji : N... 1 1 ... can one determine the rank of (fJ .. Moreover... be the map ...' is the unique atlas with this property.
The closure S of the final resulting figure is known as Alexander� Homed S plwe..Differentiable Structures 55 13.y) C ]R 2 _ 1HJ 2 for y = (x. 0.I)dimen sional (differentiable) submanifold. This figure may be extended by putting V Mn V.S. ]R2 such that (I) f(x. < . Show that S is homeomorphic to S 2 (Hint: The additional points in the closure are homeomorphic to the Cantor seLl If is the unbounded component of ]R3 . V = V. O) (2) f(x. (a) There is a map f : ]R2 . 0) for all x. (3) f(x. but au # boundary (b) Consider the figure shown below. and then repeating this construction indefinitely.) V if V smaller copies of the two parts of S' into the regions indicated by arrows. O) < I or I < d(x.O) < 2). then U is a manifoldwithboundary. y) c 1HJ 2 for y :0: 0... 14. so part (a) is true only for differentiable submanifolds. then S boundary but U is not a 2dimensional manifoldwithboundary. (a) Let C be an open set such that boundary is an (n . (It is well to bear in mind the following example: = {x E ]R" : d(x. Show that U is an IIdimensional manifold withboundary.
fJ). such that = for alI E and a similar diffeomorphism V x [0. . I). Show that there is a unique Coo structure a: V . (c) Now suppose that there is a neighborhood of in and a diffeo morphism x [0. x x f define a Coo structure on P. aN aM a(p) V (p...O) Mp aM. nn2 JR ann22. (We will be able to prove later that such diffeomorphisms always exist). V) f(p).V U fllVJR" aM xi VV aM. I) induced by and is a diff eomorphism. .I .56 (4) Chapter 2 restricted to the upper halfplane or the lower halfplane is Coo. (b) Suppose and are Coo manifoldswithboundary and is a diff eomorphism. on P such that the inclusions of and are Coo and such that the map from V to x ( . . (x. I). V) a coordinate system around with and (y 0 n = n we can define a n = V n homeomorphism from V C P to by sending to 1HI" by and V to the lower halfplane by the reflection of y. but itself is not Coo. Show that the resulting Coo manifolds are diffeomorphic. (d) By using two different pairs define two diff erent Coo structures on considered as the union of two copies of with corresponding points on identified. Show that this procedure does not f M N M UJ N f: aM .. V U aM M a N fJ (a. aMfJ: . aN M N p aM x aM f(x) aN. f(V aM) aN. but that the diff eomorphism ca"not be chosen arbitrarily close to the identity map. . Let P = be obtained from the disjoint union of and by identif ying E with E If is a coordinate system around E and (y.
} be a countable cover of V. C = and 11 (0). . I: JR JR I(x) = { �1/x x > 0o x > CooCoo. for any closed C (b) Let {V. Let is Coo.  a function with 11 (0). I for some coordinate system taking an open subset of > [0. Ji of all orders :::: i. Coo I C = ({a JR": lal ) x MJ. I] be a function with ° on Vi and Functions like J.1 /x' is used just to get a function which is and e. . Coo J. Cl = C1 C2 CM.C. = X . Can the inclusion be an imbedding? Are the projections on each factor maps? (b) If and are manifoldswithboundary. . of order 1 . where each Vi is of the form < E I}  onto ° on Let Vi . Show that the function Coo Coo Coo M NM N Coo is (the formula e. 2.I 17.I /1xl could be used just as well). . . . I] such that C 11 (0) and C 11 (I). =C M JR". > aJ. once you 11 (0) and know the trick.Differentiable Structures 57 15.: M will be called mixed partials of ai = Show that sup of all mixed partials of Ji . The proof turns out to be quite easy. . I = �A I L a2i i=l 18. Lemma 2 (as addended by the proof of Theorem 15) shows that if and C2 are disjoint closed subsets of then there is a function > [0. a2J. Consider the coordinate system (yl. (a) Find a structure on 1HI1 x 1HI1 which makes the inclusion into JR2 a map. 2 Coo > x C1 I: M ° for < 0. y2) yl(a.b) = a y2 (a. construct a structure on x sucl1 that all the "slice maps" (defined in Problem 8) are 16. Actually. . axj ' axj axk ' J.b) = a b. defined by :s (a) It suffices to find.. + for JR2 defined by . we can even find with = 11 (1). C M M.
f: [0. A n •41+3) An •41+4 are contained in A n l . and 0 (provided [0 2 1 . write each 1.58 Chapter 2 (a) Compute from the definition. (a) The lower left square is A n . Define x []J [EJ 16 15 2 13 3 14 12 9 8 10 6 7 6 7 10 5 8 9 4 3 II I 4 5 I 2 II 12 13 E x . A n •41 +2 . a subdivision of . 2"'. .:) + fe (Hii ) ]' a/ax in terms of alar alae. square A n . (First compute from this). Notice that # on and i. 1] . (b) The upper left square is A n• 22n. [0. k +l have a common side. a2/ax2 f: Nm Cl aI. A n .. for II = and 3. 1] by the condition kf(l) An. the operator alai depends in terms of polar coordinates. 1] [0.: (rf. 1] (d) Squares An •41+ 1 .... .k for all 2""2'1I � I :S 2k2n · Show that f is continuous. The following pictures show.l+ l .2. (b) Also compute it from Proposition 3 (to find of and y 2) . in terms yl 19.. . not just on even though af/ayl (a. The numbering is determined by the following conditions: [0 .1] 1. I . k and An . �[f. x into 22n squares. I] [0. A n . If M n . onto [0. /ayj . and not oneone. is and m > II. 1 ]. Compute the "Laplacian" = 1 1 . . I .k is labeled simply k.b) yI I y af/ayl af/aI /. Answer: then compute 20. (c) Squares A n. then f(M) has measure that M has only countably many components)..
define f(p/2n) E 59 ]R2 as shown below. after renumbering coor 25. . [0. Ixl) . For each partition P "" {Io. .Differentiable Structures 22.) . i=1 k C 24. (b) Every submanifold of ]Rn is locally of tllis form. and that its image will not have ]Rn be continuous. . x E ]R) is not the image of any immersion of ]R into ]R2 ]Rk is open and f. . Id of i(c. What does the inside of this curve look like? 23. (a) If V C Coo. C M can be made into a kdimen sional submanifold of M if and only if around each point in M. define > f is uniformly continuous. 1] [0. 1]. a set M. f(p» E ]R" . (a) If M is a manifold. can be made into a closed submanif if and only such old coordinate systems exist around every point of M. p E V) is a submanifold of ]Rn. P » ) is bounded above (with length equal to sup{i(c . (b) The subset M. Show that the image of a rectifiable curve has measure O. 1]. A = f(O)�"" g . dinates. (b) Consider the homeomorphic image of S ' obtained by adding. so that it has a continuous extension ]R2.C(li. V) on M such that M. You will need . Show that g is oneone. there is a coordinate system (x. 1 ] '"' B = f(l) (a) Show that > measure 0 if the shaded triangles are chosen correctly. n V = {p : x k+ ' (p) = . For p/2n E [0. [0. a semicircle with diameter the line segment AB. . V > ]Rnk is then the graph of f = { (p. Let c . P) = The curve is rectifiable if {i(c. "Ld(C(li). Coo if {(x. P» )). below the image of g. = xn (p) = 0). (Neither Theorem 9 nor l O is quite strong enough. . The set 26.
41). Prove Proposition 12: If I : Mn . Coo Coo 31... (f) If M is a manifold. then II (y) is a (closed) submanifold of M of dimension n . then I is onto. Theorem l O is es sentially Theorem 213 of Calculus on Manifolds. Y is proper if II is compact for every compact C Y. comparison with the implicit function theorem will show how some information has been allowed to escape... Z]) = (yz. N (a) L(f) = 0 if and only if I is proper. . 30.. A continuous function I: ... y.. The limit set L (f) of I is the set of all y E Y such that y = lim I(xn) for some sequence x" X . Geometry and the Imagination. pg. There is a way of immersing IPz in IR 3 . N N N 28. Show that g fails to be an immersion at 6 points (the image points are the points at distance ± 1 /2 on each axis)... 3 1 7321. IRz with I(IR) closed. pp. I] is in only finitely many members of the cover. 1R.xz. the function I can be made if M is a manif old. and I : M . (b) If M and are nmanifolds with M compact and connected. but L(f) # 0. I] which is not locally finite but which is "point finite": every point of [0... (e) A submanifold M. (c) Prove the Shrinking Lemma when (9 is a (not necessarily countable) point finite cover of any space. xy) defined in Chapter I. (b) I( ) c Y is closed if and only if L(f) C (c) There is a continuous I: IR ... is an immersion.) 27. See Hilbert and CohnVossen. C X (C) X X X f(X)... (b) Prove the Shrinking Lemma when the cover (9 is pointfinite and countable (notice that localfiniteness is not really used). M is proper. C M is a closed submanifold if and only if the inclusion map i : M.. (d) A oneone continuous function I: .60 Chapter 2 the implicit function theorem (Calculus on Manifolds. 29...k (or is empty). there is a proper map I : M .. consider collections e . Y is a homeomorphism (onto its image) if and only if L(f) n I(y) = 0. (You will need Zorn's Lemma.. Let I : IPz . • • • E with no convergent z subsequence.. whose image is the Steiner surface. known as Boy's Surface.. 1R 3 be the map g([x .. (a) Find a cover of [0. has constant rank k on a neigh borhood of I'(Y). (a) An immersion from one nmanifold to another is an open map (the image of an open set is open). X 3 .
(a) The set of all nonsingular n x /1 matrices with real entries is called GL(/1. (e) Using the formula 1/Iij (A) = L. (c) Show that O(n) is compact.IR) is called the orthogonal group 0(11). is any neighborhood. Using the formula for D(det) in Calculus on Manifolds . show that for A E O(n ) the matrix 32. and that 1/1 0 RA = 1/1 for all A E O(n). is the subgroup of all matrices with det = I.I. (d) For any A E GL(/1. However. since it is an open subset of IRn'. . A" where At is the transpose of A. or unimodular group. 33. IR). 1R) of dimension n 2 . (b) The symmetric n x n matrices may be thought of as IRn (n + ' ) /2 Define 1/1 : GL(n . It is a manifold. The special linear group SL(n . I). IR). (c) This is false if IR is replaced by a disconnected manifold of pairs (U. Coo N S'. = (0. U ') where U E 19 . U :J M. Show that RA is a diffeomorphism. Coo ( a1/lU (A) ax kl ) has the same rank as ( a1/lij I axkl ( ) ). U') E e. 1R) > GL(n . and f : M. f can always be extended to a function in a neighborhood of M. and 1/Iij the n(l1+ 1)/2 component functions of 1/1. C M is a closed submanifold. > IR is then there is a function J: M > IR with J = f on M" and with support J c U. U' C U. the general linear group. together with all other U E 19 covers the space. Show that A E O(n ) ifand only if the rows [or columns] of A are orthonormal. 24. 1R). f has no continuous extension to a map from ]R2 to but the proof requires some topology. aik ajk k (A = (a/j »). in fact. and use partitions of unity). S I . (a) If M. By applying the chain rule. show that S L(n . define RA : GL(II. (b) This is f alse if M = IR and M. = = and f = identity. M. in this case.) Coo.) Conclude from Proposition 12 that O(n) is a submanifold of GL(n. The subgroup 1/1'(1) of GL(II. (extend locally. 1R) > (symmetric matrices) by 1/I(A) = A . (Here xkl are the coordinate functions in IRn 2 . Coo N. 1R) is a closed submanifold of GL(n .Differentiable Structures 61 Remark: It is also false if M = 1R2 . pg. 1R). and the union of all U' for (U . 1R) by RA(B) = BA .
Hint: If h denotes the k x k identity matrix. Let M(m . l1.Ao are < e . The group O(n)n SL(n.Ao are < e. then X has rank k if and only i f D = CAJ B.1)/2. . k ) there are permutation matrices P and Q such that PXoQ = ( AO Co BO Do (b) There is some e > 0 such that A is nonsingular whenever all entries of A . k) the set (a) For every Xo E M(m . then ( � �) (d) M(m .:s m. where the entries of A . n . x 1 aj l au k=i#J k=j#i k =i=j 2aj/ o otherwise.62 Chapter 2 show that a 1fr ij .n) is a submanifold of dimension k(m + n . or the rotation group R(n). k) C M(m. n) denote the set of all m of all m x n matrices of rank k . (c) If PXQ = ) ' where Ao is k x k and nonsingular.) Conclude that O(n) has dimension n (n . n . and M(m . (f) Show that det A = ±I for all A E O(n). 34. n. n matrices. 1R) is called the special orthogonal group SO(n).k) for all k .(A) = axkl Show that the rank of this matrix is n(n + 1)/2 at I (and hence at A for all A E O(n).
But there are Amany situations where we would like to picture this same arrow as starting at a different point P E JR": For example.CHAPTER 3 THE TANGENT BUNDLE point v E JR" is frequently pictured as an arrow from 0 to v. and the "velocity vector" or "tangent vector" c'(t) of the curve c is customarily pictured as the arrow from crt) to crt) + c'(t). . crt) + c'(t ) 63 . but the line between crt) and c(t) + c'(t) is tangent to the curve. . . . suppose c: JR > JR" is a differentiable curve. c"'(t» is just a point of JR". Then c'(t) = (cl'(t).
v) for v E 1R" by 1R"p . we define the "projection" map 1r: 1R" x JR" . in conformity with this notation. like v. the "tangent space of 1R"". We will often denote (p.1 (p) may be pictured as all arrows starting at p. it can be pictured more geometJ·ically as a particular subset of 1R" x 1R".64 Chapter 3 To give this picture mathematical substance. the one visualizable case occurring when 11 = 1.. 1R" by ]( a .. b) = a. elements of TIR" are called "tangent vectors" of 1R". At times. we will denote the set of all (p. The set ]( . we simply describe the "arrow" from p to p + V by the pair (p. This fibre can be made into a . The set of all such pairs is just 1R" x 1R". Alternately. which we will also denote by TIR". For any tangent vector v. To recover the first member of a pair v E TIR". it is more convenient to denote a member of TIR" by a single Jetter. v) E TIR" by vp ("the vector V at p"). the point Jr ( v) is "where it's at". This picture gives rise to some TJRn JRn _ _ ]( '_ terminologywe call ](1 (p) the fibre over l  p p.. v).
v) (j) (p.The Tangent Bundle 65 vector space in an obvious way: we define (The operations U 1( . Suppose that g : JRm > JRk is another differentiable function.p p p ing diagram "commutes" (the two possible compositions from TlRn to jRm are equal). the follow . Since f (v) is defined to be a vector E JRmJ( ) . J. v) = (p. and p E JRn. however. built into it. g. This is not the only reason for defining f in this particular way. it can be written . then the linear trans formation Df(p) : JRn > JRm may be used to produce a linear map from JRn > JRmJ( ) defined by (j) ) p p Vp "'" [ Df(p)(v)lJ(p) ' This map. (I) By our definition. and 0 . . the symbol f. denotes the map f : TJRn > TJRm which is the . so that.' (p). has the map f. v + w) a o (p. Usually we will just use ordinary + and · instead of If f : JR" > JRm is a differentiable map. whose apparently anomalous features will soon b e justified. .p union of all f.a · v). x respectively. = f 0 1(. . should really b e thought o fa s defined on and JR TJRn. Thus. peJRfI x (j) and 0 (p. ( [ Df(p)(v)lJ(p) } = (Dg(f(p)}(Df(p)(v» }g(f(p))' This looks horribly complicated. using (I). D (g 0 f)(p) = Dg(f(p» 0 Df(p). 1( 0 f.' (p) 1( . W) = (p. is de noted by f . as well as all maps Df(p). but. by the chain rule.
If g: JR" ..66 Chapter 3 thus we have This relation would clearly fall apart completely if I. c"'(I» c(I) .) = [DC(I)(I)]c(l) = (c" (I). The "tangent vector" of a curve c : JR . then g 0 c is a curve in JRm. rather than Df. 0 I..(vp) were not in JRmJ(p) . in terms of I. /'(I» c(I). J ) is the "unit" tangent vector of 1R at t . .. Henceforth. JR : 1'(1) then C'(I)C(I) = (I . .] Notice that the tangent vector of c at 1 is the same as where I t = (t . The tangent vector of c at 1 may be defined as g.. = (g 0 fl •. c. this vector lies along the tangent line to the graph of 1 at (1. . we will state almost all concepts about Jacobian matrices. I... it is merely an elegant restatement of the chain rule. also. 1(1 )) for I: JR . JR" can be defined in terms of this concept. The tangent .. ..(I.. JRm is differentiable... like rank or singularity. c'(I)c(I) E JR"C(I) ' [If c happens t o be of the form c(l ) = (1. with our present definition of I. /(1 » .
) = g..I ) .X(p) . (x 0 y. (tangent vector of c at I) . (i 0 x .I o x o y.(c.(I. = (i o x . IR N .. (lRnx(p) is an ndimen sional subspace of IR N .. IR x( p) . This subspace doesn't depend on the coordinate system x) for if y is another coordinate system..(I.I ) . and is an isomorphism (with inverse (y o x . I = (i 0 x . Consequently.I ) .. n n 'y(p) : IR y(p) .I ).(p). Consider now an ndimensional manifold M and an imbedding i : M . then (i o y I ).The Tangent Bundle 67 vector of g 0 C at t is (g 0 C).I ) � �� � �. Suppose we take a coordinate system (x.» v crt) = g. U ) around p. 0 (x 0 y .) ..I is a map from IRn to IR N with rank n. Then i 0 X .
68 Chapter 3 There is another way to see this. If c : (e. Thus.i) = p EM U (M. We can define a "projection" map n : T(M. We will denote this ndimensional subspace by (M.J ) * (lRnX(p) ) ' Moreover.I (p) has a vector space structure also. Consider first the manifold M = S ' and the inclusion i : S ' > jR 2 . and every diff erentiable curve in i(M) is of this form (Proof?). 0 x .sine) passes through every point of S ' . i)p con n(v) = p if V E (M. Beyond this we have to look a little more carefully at some specific examples. cos e) # O. every vector in this subspace is the tangent vector of some a) since every vector in jRnx(p) is the tangent vector of some curve c.sin e.i) > M by As in the case of TlRn . then a = .I 0 c is a curve in jR N which lies in . The Curve c(e) = (cos e. e) > jRn is a curve with c(O) = x(p). our ndimensional subspace is just the set of all tangent vectors at i(p) to differentiable curves in i(M). i)p C i(M) x jR N C TjR N . which justifies the picture we have drawn. For each p = (cos e.i)p. sin e) E S ' . . each "fibre" 7f . Then (S ' . i)p' We now want to look at the (disjoint) union T(M. Now so the tangent vector of every a is in (i o x. and c'(e) = (sine.(M). let up = (. cose)p (it clearly doesn't matter which of the infinitely many possible e's we choose).
). up = (p. We Can therefore define a homeomorphism I. would be a collection of nonzero tangent vectors. then each fibre has a vector space structure in a natural way. S2 fi. If we define the "fibres" of to be the sets (p). SS22 2 ]R2. i) fi .O in the set of vectors T(S'.. . clearly I. one at each point of which varied continuously. which makes the following dia· [rr'(a. S2.. and the inclusion i : Now consider the manifold M = C 1R 3 • In this case there is 1/0 map fz : X 1R with the properties of the map If there were. S' x ( ) T(S ' .).The Tangent Bundle 69 sists of all multiples of the vector up . for a fixed vector v #. .b) a] = T(S2. ft rr' rr'' 1R ' by fi ). : i) gram commute. then. Commutativity of the diagram means that fi takes fibres into fibres. restricted to a fibre is a linear isomorphism on to the image. It is a wellknown (hard) theorem of topology that this is impossible (you can't comb the hair on a sphere). . i) . .S ' x IR' � S.
1 )(2rr. The impossibility of choosing nonzero dashed vectors continuously clearly shows that there is no way to map T(M. 0)(2.0. Notice that 1 ( 0. = while = [� (O. which it can't (by an easy theorem of topology).O)J [ aal l [ DI(O.0.0. 1 ) (0. f).70 Chapter 3 There is another example where we can prove that no appropriate homeo morphism T(M. 0)(0. The same is true for all multiples of 1 ( 0. to be precise.1) = (2 cos8 + 1 cos � cos8.0) ) . 1 ( 0. 0): Ifwe could.0) ).M is the image of the map I: [0. 0) of M. for some continuous function A : [0.0. 2 sin 8. shown * as dashed arrows in the picture.0) = ( 1 . 0) i . This function would have to be nonzero everywhere and also satisfy A(21r) == A(O). i) > M X 1R 2 exists. 0) (2.0. 21r] x (I.. 1 sin n . 21r I r�� is a tangent vector. the vector (8. the particular subset of IR 3 defined in Chapter 1.0) = ( .0) . then each vector would be J (2. 2 sin 8 + 1 cos � sin 8. 21r] > 1R.  .1 . without appealing to a hard theorem of topology. At each point p = (2cos 8. 2 sin8. %2. 0 .0) (2. fibre by fibre. The map f will just be the inclusion M > 1R 3 where M is a Mobius strip. 0) This means that we can never pick nonzero dashed vectors continuouslY on the set of all points (2cos 8 . I) > 1R 3 defined by 1(8.0) . 0 .0) ) = * (21r. 1 )(8.
.The Tangent Bundle 71 homeomorphically onto M x JR2• We thus have another case where T(M. the part of T(M. which make each fibre n:I (p) into an ndimensional vector space over lR.. This additional feature qualifies T(M. there is a neighborhood V of p and a homeomor phism I : n:I (V) . For any imbedding i: M .. 0) . fibre by fibre. however. i) over V. such that the following "local triviality" condition is satisfied: For each p E B.. homeomorphically onto V x JR". i) is always simple locally: if (x.. E.I (V).. the structure of T(M. E. V x JR" which is a vector space isomorphism from each n:I (q) onto q x JR".. 0: JR x E . n:. with Ell (n:I (p) x n:I (p») c n:I (p) and 0 (JR x n:I (p») c n:I (p). JRN... V) is a coordinate system on M. In standard jargon. for all q E U. can always be mapped. Ell .. V = x JR" (p. .. v) . (3) Ell and 0 are maps Ell : p EB U n:I (p) x n:I (p) .. In fact. T(M. for each p E V.i) is "locally trivial". (2) n: : E .. i) does not "look like" a product M x JR". i) to be included among an extremely important class of structures: An IIdimensional vector bundle (or nplane bundle) is a fivetuple where (I) E and B are spaces (the "total space" and "base space" of t respectively)... � = (E.. B is a continuous map onto B. then n: .. we can therefore define by f: f (mp(vx(p) ) n:I (V) . the fibre where the abbreviation mp has been introduced temporarily.. B .
and a .n(x). :::: �z) if there is a homeomorphism h : E.' (p) and a E 1R. we will denote (j)(v. > Ez which takes each fibre ".) The bundles T(S z . : E > B. . This is called the trivial Ilplane bundle over X and will be denoted by . 0) automatically gives rise to a bundle �IA over any subset A C B. v) by v + w. : E.72 Chapter 3 Because this local triviality condition really is a local condition. and the obvious vector space structure on each fibre. but there is an even simpler example of a nontrivial bundle. i) are not trivial.n(IRn). The bundle T(S ' . = " .. (j). W E . The "tangent bundle" TlRn is just . The Mobius strip itself (not T(M. or even denote the bundle by E alone. i) considered before is equivalent to .' (p) isomorphically onto "z' (p). The map h is called an equivalence.'(S'). > B and �z = "z : Ez > B are equivalent (�. I] x IR by identifying (O.n(B) is called trivial. For vectors v. respectively. . and we shall usually refer simply to a bundle . w) and 0(a... for M can be obtained !i·om [0. v or av.. to be precise. ."nJ1sidered as a Idimensional vector bundle over S l .a) with ( I .. each bundle � = (E. Notation as cumbersome as all this invites abuse. : X x 1R" > X the projection on the first factor. . while S' can be obtained from {O. a). Equivalence is here a technical term: Two vector bundles �. The simplest example of an Ilplane bundle is juSt X x IRn with . A bundle equivalent to . i) and T(M. (The local triviality condition for a bundle � just says that �IU is trivial for some neighborhood U of p. B. i») can be . I } .
a)}) = {O. such that (I) the following diagram commutes jR 1(.a) = I for 0 < I < I and 1( {(0. Thus f may be restricted to be an isomorphism on fibres and f to be the identity. f: jRk jRjRll . fl. The relations between some of these cases are considered in the problems. > If M" c are submanifolds. . a). fixed bundles over various spaces. depending on whether one is consider ing all bundles at once. The analogue of a homomorphism is the following. An equivalence is obviously the analogue of an isomorphism. Z jRk k l jRl TjR TjR jRk * There are actually several possible choices. or a homeomorphism. > Bz. to �z is a pair of continuous maps (j. (2) j : 1(. then f . I }. The diagram above illustrates local triviality near the point {O. EI BI � Bz . This surely shows that M is not a trivial bundle. > 1( . the map 1( is defined by 1( I. fl is a bundle map fi'om . Since s must be ° somewhere.I U(p)) is a linear map. Such a map corresponds to a continuous function s : [0. and the map f satisfies f(M) C N. or a fixed base space with varying bundles. (I. I] > with S(O) = S(I).' A bundle map from �. ' (p) 1 � Ez l1(Z to for any differentiable The pair U . i : 'M > and Nm C and j : N > are the inclusions. I } of Sl Suppose that s : S' > M is a continuous function with 1( 0 s = identity of M (such a function is called a section). I] by identifying ° with I .The Tangent Bundle 73 [0. with j : E. the section s must be ° somewhere (that is. > Ez and f : B. s(8) E 1(1 (8) must be the ° vector for some 8 E S').
I I We can therefore put all these maps together. Elements of are of the form 0 ). M N. That is the T(M. while elements of If we map T(M. j). while and are two imbeddings of in and respectively.i)p k MIRIRI. T(M.i) TM. N. the dependence of on is al most illusory. this construction could be generalized much further.j)IV. j)p are of the form (j 0 x 1).i)p T(M.i) T(M. T(M. we just consider the function j 0 f 0 > and extend it locally to The case which we want to examine most carefully is the simplest: where = N and f is the identity. M. we could abbreviate to if we agreed that really denotes an equivalence class of bundles. then (M.j)p x. just remember that v E is the tangent vector of a curve c in so f (v) is the tangent vector of the curve . i) T(N.j). j). V) is another coordinate system. (v) E map from to Actually. (w) for we obtain a bundle map from to which is obviously an equivalence. T(N.i) T(N. The map > induced on fibres is independent of the coordinate system for if (y. In fact. T(M..i(N) M N i T(M. i)p j (i xI M for WE WE IRnx( IRnx(p)' p). (w) T(M. and obtain an equivalence from to In other words.j). rather than one bundle. T(M. (M. to the case where and are merely imbeddings of two abstract manifolds and and f : > is Coo.74 Chapter 3 takes to to see this. i) TMi . it would have sufficed to begin with a Coo function f : > since f can be extended to locally.i)I V T(M.I : i(M) N. In this way we obtain a bundle f o e in and consequently f. IRk IRk i j M i.
in some natural way. 0 f . and to each map f : M > N a bundle map (f. = g. since the maps 1 * : TV + TV and i. _ _ _ . and it is undoubtedly ugly.. : TM > TM is the identity. em (JRm) 1 1 (3) If U c M is an open submanifold. ea . * vVhen using the notation h. where i : U > M is the inclusion. (I) If 1 : M Coo TJRn � TJRm 1m In en (JRn) old f. : TU > TN is just the restriction of f•. TJRn will be different from our old defini tion (namely. it must be understood that the symbol "/" r lly refers to a triple (f. What we would like to do is to get a single bundle for each M.' TU � TM)IU/c ( _ _ _ '" i. The identity map I of U to itself and the inclusion map i : U + M have to be considered as different. . we can. If • g : N > P. I .. .. then (g 0 f) . Can we do this? Yes. so in stating our result . TM TU (flU). but the elements of TM will each be large equivalence classes. : TU > TM are certainly different (they map TU into two different sets). such that: Coo > M is the identity. then TU is equivalent to (TM)IU. and so will f for f : JRn > JRm. When we do. i) has. THEOREM. �TM/f. and for f: M > N the map (flU). . TN PROOF The construction of TM is an ingenious. We will obtain a single bundle for TM. en (JRn»).. It is possible to assign to each nmanifold M an nplane bun dle TM over M. sub terfuge. then 1. (2) There are equivalences In : TJRn > en (JRn) such that for every function f : JRn > JRm the following commutes.. More precisely. there is an equivalence TU :::: (TM)IU such that the following diagrams commute. though quite natural. N) where f : M > N. f).The T angent Bundle 75 sort of thing an algebraist might do. M. which has all the properties any one of these particular bundles T(M.. precisely we will write I'old J/' when necessary.
and this would be an equivalence (with inverse (xI). maps . the equivalence class of (x. we would have a map x. We now have a bundle . W E we define if (a) is satisfied. Then (x. JRn. so we want I (A) to be open for every open A C V x and thus we want any union of such sets to be open. v]p. We define a vector space structure on .. This condition makes perfect sense without any mention of bundles. I = (y 0 xI). and TM is defined to be the set of all tangent vectors at all points p E M.I (p) by Mp. v]p = [x.I (p) by the formulas [x. in conformity with the notation p. w]p = [x. IRn. Ix Ix : . so we leave this one part of the proof to Problem l. w) for some w E We can easily figure out what the relationship between v and w would be. v]p + [x. If JRn.I (p) isomorphicaIly onto {p) x If y is another coordinate system. since x. (a) It is easy to check (using the chain rule) that p is an equivalence relation. JRn . v + w]p a · [x. : TM > M.. V) is a coordinate system.. We will denote the fibre .I ) (x (p»)(v). 0 x. since (x(p). to some (x(p).. because D(y o xI)(x(p» is an isomorphism from to Our definition of TM provides a oneone onto map (b) JRn JRn . namely [x.. v).. v) is taken to (y(p). It is the clue which enables us to now define TM. and (y 0 xI). There is a metric with exactly these sets as open sets. but it is a little ticklish to produce. Since TV should be essentially (TM)IV.. JRn JRn). v) p (y. If x and y are coordinate systems whose domains contain p.(e) would be (y(p). then y. We want this to be a homeomorphism. we would have if JRn. is supposed to be the old (y 0 XI). These equivalence classes will be called tangent vectors at p. the map " takes p equivalence classes to p.76 Chapter 3 The construction is much easier to understand if we first imagine that we al rearfy had our bundles TM. v). v) will be denoted by [x. Here v is just an element of (and every v would occur. : TV > T(x(V» . a . v]q � (q.I (p) would be taken by x. w) w = D(y o x.. I (V) > V x JRn. (x. and v.). this definition is independent of the particular coordinate system x or y. w) by y. though TMp might be better. v]p. and T(x(U» should essentially be x(V) x a point e E .
. D (y 0 I 0 xJ )(x(p») (v)]J(p)' I. with c (o) = p. i). i lp . each defined on some interval around 0. V) is a coordinate system around p. IRk i. V) are coordinate systems around p and I(p) . the only difference is that each equivalence class for M contains some extra members. will be the elements of our new bun dle. we define (c) N. we next ask how fortuitous this was. D (i 0 x ')(x( p») (v)) E ( M . In II. ([x . i ). for all p E M. though perhaps confusing to the novice. IR IRn. the fibre of TV over p E V is almost exactly the same as the fibre of TM over p. and I is the identity coordinate system of then If . •:. V) is a coordinate system around p. which is equivalent to T(M.The Tangent Bundle 77 I : M > and (x. V) and respectively. and we proceed to define two different such bundles. the bundle 1C : TM > M will be called the tangent bundle of M. e) > M. [y. we define In. if (x. since in M there are more coordinate systems around p than there are in U C M. In II I = by (c) CJ '1 C2 if and only if: The equivalence classes. > is an imbedding. : M IRk . D(i 0 x J ) (x( p»)(V)] i (p) the composition is easily seen to be an equivalence. i) will play no further role in this storythe abstract substitute TM will always be used instead. To prove (2). to prove commutativity of the diagram. . (y. v]p) = O f course. Condition (3) is pI"actically obvious also. In fact. If (x . it must b e checked that this definition is independent o f x and y (the chain rule again). ( i ( p) . then TM is equivalent to T(M. In fact. mapping to have the same derivative at O. Condition (I) of our theorem is obvious. For I : M > there is a map h taking the '1 equivalence class N x 0 CI and x a C2. T'M. V]p) = [I. IRn Ix Henceforth. it is trivial. Having succeeded in producing a bundle over each M.([x . Can one find other bundles with the same properties? The answer is yes. For the first example. But T(M. we define to be where I is the identity map of and is defined in (b). we consider curves c : (e.
If J is defined only in a neighborhood of p. h corresponds to J•. We define a tangent vector at p to be a linear operator i which operates on all functions J and which is a "derivation at p": xl o y. In the second example. Coo i (fg) = J(p)i(g) + g (p)i(f). (2) gi (O) = D. if J = g in a neighborhood of O. we can already see that this example is "really the same" as TM [x. things are not so simple. . y' (0) = V. Then there are Coo functions gi Coo function in a convex open neighborhood U of 0 : U > with JR . LEMMA... There is a function h : M > with h (p) = 1 and support h C (0).78 Chapter 3 (/.xn) for x E U.) For x E U. . with support h C JI (0). and define i(f) as i(fh). xi dt. = :s .xn) = L:7= l xigi (X I . we may use this trick to define i (f): choose h to be 1 on a neighborhood of p. This comes out of the following. • JRn .f(O). since U is convex. Then n J(x) = J(x) . . (The second condition actually follows from the first. For these operators. this is defined for 0 :s t I.g) = i (f) . We have already seen that the operators i = have this property. clearly i (f) = i (g) if J = g in a neighborhood of p. suppose that J = 0 in a neighborhood of p. then O = i(f . vlp corresponds to: the 'J' equivalence class of where is a curve in with y JRn under this correspondence. Let J be a in (I) J(x l .J(O) = Jo I hx'(t)dt = l0 I L Di J(tX) . let hx(t) J(tx). but it is not a priori clear what its dimension is. Then a/axi lp JI Coo JR 0 = i (O) = i (fh) = J(O)i(h) + h(O)i(f) = 0 + i(f). . Without bothering to check details. . 2. . with J(O) = O.) of c to the J equivalence class of J 0 c. Thus.i(g). This condition is actually true for any derivation i. r i=1 Therefore we can let g (x) = J� Di J(tx) dt :PROOF. The set of all such operators is a vector space. For.
It is a simple exercise to use the coordinate system to transfer this result from to . Then (Ii denotes the i(f) = i(f . THEOREM. Hence i(c) = c . they are clearly linearly inde pendent. U) is a coordinate system around p. l) = l · i(I) + l · i(l). . In fact. •:./(0)) = e(£) igi ) coordinate function) 1= 1 = I )i (Ii )gi (0) + I i (O)i(gi)] 1= 1 ith PROOF. [x.The Tangent Bundle 79 3. Notice that i(l) = i(l .. and any derivation . Given / on U. Consider the case where M = JRn and p = o. a]p. . .. Assume U is convex. then Mn span this vector space. a�n Ip (so i is determined by the numbers i(xi »)./(0). if (x. a bundle constructed from all derivations at all points of is "really the same" as We can let correspond to the formula a/a Ii lo IRn M x M TM. so i(l) = o. From Theorem 3 we can see that. This shows that span the vector space. for the function / .e can be written a�I Ip . choose gi as in Lemma 2. once again. i(l) = 0 for any constant function c on U. The set of all linear derivations at p E is an ndimen sional vector space.
We wilJ denote it by the suggestive symbol dt 10 de l · . we wilJ not hesitate to write v(f) for a differentiable function f defined in a neighborhood of p. a tangent vector is often most easily described by telling what derivation it corresponds to.a) 1=1 axi n a L1 a1. and this is precisely the equation which says that p (y . ayj a' (p). I [x. that is. and the map f. If e : JR M is a differentiable curve.a]p most easily analyzed from the relation (f. en (JRn) .=1 a a!. TJRn 0 f. . the map which corresponds to can be defined as follows: (x. is often j . then > dt 10 d I for I. b ) . between E v Mp x ap [f.80 derived in Chapter 2. and to is calJed the tangent vector to c at 10. It is easily checked that under this correspondence. In fact.v)(g) = v(g fl. . then Notice that if denotes the identity coordi� ate system on P corresponds to when we IdenllfY wIth We will usualJy make no distinction whatsoever between a tangent vector and and the linear derivation it corresponds to... � . t i . 0 It is customary to denote the identity coordinate system on write JRI by this is a basis for JRlo. shows that Chapter 3 if and only if bJ = L. (i)](g) = i(g fl..IP 1= ax consequently. JRn.
. . . (x. Since locally looks like x clearly is itself a manifold. is simply the map x. . This expression shows that y. If (y. I n . (x. so it has been quarantined in an Addendum to this chapter. Then the map v � (x l (Jr(v» . .. xn (v)) E ]R2n is a homeomorphism from (TM)IV to x(V) ]Rn .a) = (I I .xn (Jr(v)). Di (y1. TM Coo(TM) I V v a i by Xi (v) . V) is another coordinate system. ax L. moreover. and quite a mess to prove.xl(v). ..a l . . L. I TM. . and Let us denote x x p Jr( ). a natural way to put a structure on If > then every element E is is a chart on uniquely of the form "I" dl will often stand for de dl I ' I ]R. but it is a little delicate to state. As you might well expect. . = � a' ay}i (p) = � a'. . . The tangent bundle of a manifold has a little more structure than an arbitrary nplane bundle. L:7=1 ai Di(yl xI )(I). it is no accident that our second and third examples turned out to be "really the same" as There is a general theorem that all "reasonable" examples will have this property. .I (l..I is Coo. . 0 = 0 0 0 then.x). L:7=1 ai Di (yn x. . an ) E ]R2n . 1=1 1=1 This shows that if (I.The Tangent Bundle 81 This symbol will be subjected to the standard abuses one finds (unexplained) in calculus textbooks: the symbol de the subscript now denoting a particular number E as well as the identity coordinate system. . . = V 0 b1. TM TM. . . when we identifY TV with V ]Rn in the standard way.I (I). there is. oxI )(x(p)). (x Jr. This map. then y.a) (y ox .I )(t)). x: V ]Rn CooTM M. as we have already seen.).). V ]Rn.
The functions ai are continuous or Coo if and only if X: U > uous or Coo. Recall that a section of a bundle 1C : E > B is a continuous function B > E such that 1C o S = identity of B. we define a new vector field TM is contin X+Y by Similarly. It follows that 1C : E > B is Coo. and the vector X(p) is of ten denoted by Xp (sometimes X may be used to denote a single vector. the local trivializations x* are Coo. Vector fields are customarily denoted by symbols like X. In general. (fX ) (p) = f(p)X(p).. if f: M JR. Y. or Z. U). TM for all pE U. a vector bundle 1C : E > B is called a Coo vector bundle if E and B are Coo manifolds and there are Coo local trivializations in a neighborhood of each point. then for any coordinate system (x. If X and Y are two vector fields. .82 Chapter 3 We thus have a collection of COOrelated charts on which can be ex tended to a maximal atlas. s: M there is no vector field on S2 which is everywhere nonzero. in some If we think of as the set of derivations. We have shown that there do not exist two vector fields on the Mobius strip which are everywhere linearly independent. With this Coo structure. IRn. M TM M. we define the vector field fX by > (X + Y)(p) = X(p) + Y (p) . for Coo vector bundles we can also speak of Coo sections. we have Mp) . A section of is called a vector field on for submanifolds of a vector field may be pictured as a continuous selection ofarrows tangent to The theorem that you can't comb the hair on a sphere just states that TM.
L. then X (f) is Coo for every Coo function I. thus X is a "derivation" of the ring !F. On U we can write a/axi now denoting the vector field the symbol I: M IR is a Coo function. and I are Coo. X(f + g) = X (f) + X(g) X(fg) = IX(g) + gX(f). It is not hard to check that if X is a Coo vector field. Y. Let !F denote the set of all Coo functions on M. If > a newf unction > > > . iflocally n a X(p) = L a'.The Tangent Bundle 83 Clearly X + Y and IX are Coo if X. Clearly. if X (f) is Coo for every Coo function I. then X is a Coo vector field (since X (xi ) = a i l. then we can define X (f): M IR by letting X operate on I at each point: X(f)(p) = Xp(f). In fact. The reason for this is that if A : !F !F is any deriva tion. then A = X for a unique Coo vector field X. a Coo vector field X is identified with the derivation X . (p) axi I ' P 1= 1 then . Conversely. al X(f) = � a ' axi' 1= ) which is a sum of products of Coo functions. and X is a vector field. indeed. we clearly must define Xp(f) = A (f)(p) . Of ten. and the operator Xp thus defined is a derivation at p . We have just seen that a Coo vector field X gives rise to a function X: !F !F.
Its sense is nicely conveyed by the definition in The American Heritage DictiollaT ': } "To understand profoundly through intuition or empathy". . . Vn) and ( 'l . purportedly as a word from the Martian language. vn) belongs will be denoted by . and thus with IRn \ then the orientation preserving and orientation reversing maps are disjoint open subsets of the set of all nonsingular maps (those with det # 0). The relation of being equally oriented is clearly an equivalence relation. . . . by Robert A. this clearly makes no sense unless we supply V and W with more structure.e. in particular on the tangen t bundle of a manifold. and you should not read further until you grok it. For the present. To provide this extra structure. [VI . . . There is no way to pass continuously between these two groups: if we identify linear maps IRn 7 IRn with n x n matrices.) = v'. The problem becomes more acute if we want to define orientation preserving isomorphisms between two diff erent (but isomorphic) vector spaces V and W. .. . V (VI. The terminology "orientation preserving" is a bit strange. if I is orientation preserving) and oppositely oriented if det A < o. If JL denotes one (VI. since we have not yet defined anything called "orientation". (VI. x n . One basic theme in all these chap ters is that any structure one can put on a vector space leads to a structure on any vector bundle. . . then . (VI . Heinlein in his pop science fiction novel Stranger in a Strange Land. . . so that if JL is an orientation of V. . . A simple example of the latter is the map I : IRn > IRn defined by I (x ) = (x l . we will discuss just one new concept about manifolds. The nonsingular linear maps I : V > V from a finite dimensional vector space to itselffall into two groups. . . j=J n = * A cult word of the sixties) "grok" was coined. v'n) equally oriented if det A > 0 (i.l . _ xn ) (reflection in the hyperplane xn = 0). Vn) E JL if and only if . which is being preserved. vn] . and those with det I < 0. .. . those with det I > 0. .84 Chapter 3 The tangent bundle is the true beginning of the study of differentiable mani folds. The class to which . . We call . 'n ) for V determine an isomorphism I : V > V with I(v. [VI .' The next few chapters constitute a detailed study of this bundle. . Vii L Qji Vj. . . linear transformations in the first group are called orientation preserving and the others are called orientation reversing. vn) and (V'l . the matrix A = (aij) of I is given by the equations V. . Either of these two equivalence classes is called an orientation for V. which arises in this very way from the notion of "orientation" in a vector space. . vn] = JL. . divid ing the colJection of all ordered bases into just two equivalence classes. . . . we note that two ordered bases .
..I (U) U x lRn . . . en (X) X x IRn [(x. together with orientations. . (x. an orientation of E is defined to be a collection of orientations (p) which satisfy the following "compatibility condition" for for any open connected set If t : > given the standard orientation. fL X IR fLp rr. 1R2. then is either orientation preserving or orientation reversing on each fibre. .The Tangent Bundle 85 VI WI Examples of equally orien ted ordered bases in 1R. . = For the trivial bundle we can put the "standard orientation" on each fibre If > is an equiva lence. f(vn)] [VI . and 1R' orientation of V. ed. . for if we define the functions : > by f: [f(VI )' . (x. . ··. . an isomorphism V > W is called orientation preserving (with respect to and vI if = v whenever = if this holds for any one it clearly holds for all.I (U) U x IRn is an equivalence. en] then det (aij) : > is continuous and never o. v) are two ndimensional vector spaces. j=1 U C B: rr.) = I:>ji (X) . and (W.. then (' automatically satisfies the same condition.I IRn fL) fL. vn ). Now if (V . [eI . and t' : > is another equivalence. . . . . f: en (X) en (X) X f aij X IR n f(x. (VI . and is connected. If 1r: E > a non trivial nplane bundle. en )] {x} x IRn . Vn ] fL.e. . since rr. the other will be denoted by and the orientation for will be called the "standard orientation". and the fibres of U x IRn are fL B is Notice that if this condition is satisfied for a certain t. ej ). then ( is either orientation preserving or orientation reversing on all fibres.
s(p) fL is M.} but not every bundle has an orientation.fL) . I] x { . fLp fL fL (O. For example. fLp . For example. TIRn :::: en(IRn). TM. we call itself orientable or nonorientable de pending on whether is orientable or nonorientable. and nonorientable oth erwise.. This definition can be applied. To see this we standard orientation. has no orientation.fL) where is an orientation for �. sn.I IRn is M fL s: M [s(p)] fLp.I . I } with identified with ( I .86 Chapter 3 > : x x is an equivalence. TM TM (M.. an orientation of is also called an orientation of and an oriented manifold is a pair where an orientation for The manifold is orient able. since on which we have the C also orientable. considered as a Idimensional bundle over SI . TM IRn M.I. For. we had compatible orientations we could define a section Sl > by choosing to be the unique vector E A n . it has another orientation = { .I (p) with = A bundle is called orientable ifit has an orientation. although the Mobius strip has no nonzero section. If I'01. the Mobius strip. The sphere s(p) fLp . in particular. we can pick two vectors from each fibre so that the totality A looks like two sections. we can let A be [0.1 U IRn U IRn U B. I]. This shows that the orientations define an orientation of E if the compatibility condition holds for a collection of sets which cover If a bundle E has orientation = {fLp }. I] x [. then A just looks like the boundary of the Mobius strip obtained from [0. an oriented bundle is just a pair (�.a) a). to the tangent bundle of a Coo manifold In this case.
x = . by considering the "antipodal map" defined by = This map is just the restriction of a linear map defined by the same formula.i.0») and their negatives. 1)(0. S2 JL is the standard orientation of S2 Jr : > are oppositely oriented.I . The map ( is just � when is identified with a subspace of { x The map Uj) E Pl the bases is orientation reversing. (A(p). 0) i . A(v» .. since it contains the E B. this shows that T(S I S I ) is also trivial.VI . ·· . v) p I IR'A Vi (p. S2p 11"2 A : S2 > S2 A: IR3 > IR3 A. the restriction �IB' Mobius strip (for any orientable bundle � = to any subset B' C B is also orientable). if then we could simply choose The projective plane must be nonorientable also. 2 sin e. Nonorientability of can be seen in another way. «0. " Vn.(vI l.: S2p > S2A p) (p. Vn. . so for VI .Ip we can define (VI . The Mobius strip is the simplest example of a nonorientable 2manifold. Since TS I is trivial. .I } thus defined is called the "standard arien tarian" of Sn. [A.(snIp) E TIRnp (Problem 21).O) } c there are continuously varying vectors but that it is impossible to choose continuously from among the dashed vectors for = 1. so if = '' I t i 2" wp S M w [v . and consequently orientable.. A.: Vip > (Mi )p is an isomorphism and the subspaces Vip vary continuously (Problem 26). If we had orientations = and wp otherwise.i. The orientation JL {JLp : P E sn .I the vector w = Pp E en (IRn ) :::: TIRn is not in i. A (p) p. w ] JL 11"2 p p p p vp .I ) E JLp if and only if (w. This can be seen by noting that for any two manifolds MI and M2 the fibre (MI M2 )p of T(MI M2) can be written as VI p V2p where (Jri). V2] JLp . The torus Sl S I is another example of an orientable manifold. (vnI l) is in the standard orientation of IRnp . For tlle imbedding of considered pre\�ously we have already seen that on the Ell M M (0. . x x x Any nholed torus is also orientablethe proof is presented in Problem 1 6. which also discusses the tangent bundle of a manifoldwithboundary. .I E sn. A subset = { (2 cos e. This shows that if and E E then [vt.The Tangent Bundle 87 note that for each p E Sn.V2] JLA(p)' Thus the map A : S2 > S2 is . JLp P E S.
.. Conversely. because it is just the condition that 0 is orientation preserving. . given A' we can orient the fibres of in such a way that is orientation preserving. these same arguments show that is orientable for n odd and nonorientable for n even. since jj.. (x.4. the determinant condition will be very important later on. U) E M. the map would then Vip } on be orientation preserving with respect to jj. axj > TM (x.' cover M M det on .88 Chapter 3 "orientation reversing" (the notion of an orientation preserving or orientation reversing map makes sense for any imbedding of one oriented manifold into another oriented manifold of the same dimension). According to this definition.'. and obtain an orientation of Although our original definition is easier to picture geometrically. In this case. For projective 3space the situation is just the opposite. There is a more "elementary" definition of orientability.. 11"2 f A p [pl. (I) the domains of all (2) for all and V) . From this fact it follows easily that is not orientable: If had an orientation v = {V[pl } and g : is the map P � then we could define an orientation by requiring g to be orientation preserving.' as the collection of all for whieh x is orientation preserving (when x is given the standard orientation).4.4.' of the atlas . If g : map � we obviously can define orientations vp for by requiring g to be orientation preserving. for sueh that 11"2 S2n 11"2 s > f: M N > [p].. In general. U) (y. which is impossible. which does not use the tangent bundle of at all.:IRnTMI U T(x(U)) :::: x(U) IRn (y xI ).4. 11"3 A : S3 S 3 > S3 11"3 11"3 > pn M An orientation JL of allows us to distinguish the subset . = JL or JL. the antipodal map is the is orientation preserving. E ( ayi ) 0 U n v. is orient able if there is a subset . Condition (2) holds.: T(x(U)) T(x (U)) TMI U x* TM.. > > (x.4. U) x(U) x.
eMl PROOF. a Ullique functor from the category of Coo manifolds and Coo maps to the category of bundles and bundle maps which is naturally equivalent to (en. and a bundle map (jil> for each Coo map M . Those who have been through this sort of rigamarole before know (i... for certain equivalences (3) of Theorem I. (2) of Theorem I. T'M � T'N leN The details of this proof are so horrible that you should probably skip it (and you should definitely quit when you get bogged down). N satisfying f) f' (I) of Theorem I... is. * Functorites will notice that Theorems I and 4 say that there . N. the welcome symbol . 4.. T'U '" (T'M)IU.The Tangent Bundle 89 ADDENDUM EQ UIVALENCE OF TANGENT BUNDLES The fact that all reasonable candidates for the tangent bundle of M turn out to be essentially the same is stated precisely as follows. then both (TM)IU and (T'M)IU "look like" x (U) x so there ought to be a map taking the fibres of one to the fibres of the other. the idea behind the proof is simple enough. JRn. THEOREM'.e. T'M TM � TN f' M . and to the restriction of the functor on open submanifolds. for certain equivalences 1m. old f ) on * Euclidean spaces. If (x. those for whom this sort of proof is a new experience should find it painful and instructive.:. then there are equivalences such that the following diagram commutes for every Coo map eM ' TM ... Nevertheless. What we have to hope is that our conditions on TM and T'M make them "look alike" in a sufficiently strong way for this idea to really work out. U) is a chart on M... have faith) that it's going to work out. occurs quite a ways on. up to natural equiv alence.. Ifwe have a bundle T'M over M for each M.
1 Similarly. ® Ik' ® Ie 0 Ie . which are denoted by the same symbol '" . Then fJx. e"(IR")ly(V) Ie <D Ij.TV � T(y(V)) =. To compare ax and ay. (TlR")ly(V) '" IY l. This will be done in three stages. using equivalence ". consider the following diagram. Two of them.). Let ax denote the composition (/nlx (U» 0 '" 0 x. are the equivalences mentioned in condition (3).1 0 ax : (TM)IU > (T'M)IU is an equivalence. U).TU � T(x(U)) � (TIR")lx(U) ' "lx(Ul. so it takes the fibre of TM over p isomorphically to the fibre of T'M over p for each p E U. Our main task is to show that this isomorphism between the fibres over p is independent of the coordinate system (x.90 Chapter 3 ax = Let (x. 0 (". (I) Suppose V maps C U is open and y = xIV We will need to name all the inclusion i : U > M I : V > M j : V > U k : y(V) > x(U) . we can define fJx . U) be a coordinate system on M.' for T'. e"(IR")lx(U) (1) (V (TM)I V =. (TM)IU ""'=". Then we have the following string of equivalences.
so that it is also true for T'M. . the inclusions x(U) and y(V) come into play. (2). and (3). and the one on the right commutes because ° j I. To see this for square (D. > > i. i e� TV Square ® commutes because k ° y = x ° Square ® commutes for the same reason as square (D. JRn JRn This means that for P E V. we enlarge it. the isomorphism between the fibres over p is the same as Clearly the same is true for and since our proof used only properties (I). not the explicit construction of TM. it will be proved for TM (where it is actually obvious). (2).1 oay = fJx. for every p E V. (II) We now need a Lemma whieh applies to both TM and T'M.1 oax ay fJx fJy. as shown below.The Tangent Bundle 91 Each ofthe four squares in this diagram commutes. Thus on the fibres over p . and (3). fJy. Chasing through diagram (1) now shows that the following commutes. The two triangles on the left commute by con dition (3) for TM. ax. Again. i = (TM)IU (TM) I V 1e �Ti "Z ™� I . using only properties (I). Square @ obviously commutes.
(TJRm) IB � em(JR)mI B PROOF Case 1. Consider the JRn and j: B JRm are the inclusion maps. j/' and JRn JRm > . and f : A B is Coo. A C JRn f. p. following diagram. There is a map j : JRn JRm with j f on A . en (JRn) laid j. where i : A > > > 1 101d f• = . A. For each we want to show that two maps are the with j = f on an same on the fibre over Now there is a map j : open set where We then have the following diagram." are equal on Case 2. If and following diagram commutes. since the maps "old "old f. then the > TA =. . B C JRm are open.92 Chapter 3 LEMMA. (TJRn) IA � en (JR)n IA TB =. This implies that the two compositions TA =. General case. (TJRn ) I A � en (JRn) I A � en (JRn ) old j. p E A' C A . A'. p E A. Everything in this diagram obviously commutes. where every :::: comes from the fact that some set is an open submanifold of another. em (JRm) and are equal and this proves the Lemma in Case I.
. = � 0 /C . Thus it suffices to prove J* 0 i* = v 0 JL 0 K. TA +. it suffices to prove that v �l TA' TA y A 7 IRn T lRn :::: (A) (/C) � (V) . � . this does commute. and ® obviously commute. @./ � / (TA)IA' '" @ e @ e Boxes (D. in which j: is the inclusion map. we imbed it in a larger diagram. (T lRn) I A o A o i* = v O IL O K. . � . and other maps have also been named. (TlRn)IA �8n(lR)nIA > i. (J O i To prove that A 0 i. . . and @ commutes by Case 1. since v is oneone. for ease of reference.The Tangent Bundle 93 and i: A' A is the inclusion map. which amounts to proving commutativity of the following diagram. :::: (T lRn) IA' Je (�) Since j 0 i is just the inclusion of Y T lRn A' in IRn . To see that square @ (which has a triangle within it) commutes. . (2) / / .
T(y(U)) =. l nlx(U). •:.. and on A' we can replace "old h" by "old J/'. (TJRn)lx(U) v. as well as to y and ylV n v. = f� oeM is > Exactly the same result holds for T': left as a masochistic exercise for the reader. with the composition TA =.l o ay fJxl o ay (3) (TM)IU "":::. the two com Commutativity of diagram (2) shows that the composition (III) Now suppose (x. (T JRm)1 B � em (JRm)1 B coincides. . V) and (y. it is clearly an equivalence eM. (TJRn) I A � en (JRn) I A old J.94 Chapter 3 positions are equal in a neighborhood of any p E A. em (JRm)IB. because part (I) applies to x and xiV n V. Diagram (3) thus shows that fJy = old (y 0 X. In other words. The proof that eN 0 f. 0 fJx.TU � . Now that we have a welldefined bundle map TM T'M (the union of all fJx. V) are any two coordinate systems with p E To prove that and induce the same isomorphism un the fibre of at p. old (y o [I). on the subset (TA) I A'./ (y o [I). we have the following diagram. which proves the Lemma. we can assume without loss of generality that V = V. and are thus equal. (TlRn)ly(U) / nly(U). e n (JRn)ly(u) The triangle obviously commutes.I). and the rectangle commutes by part (II). e n (JRn)lx (U) T(x(U)) =. Assuming V = V. The desired result fJyI 0 ay = fJx . TM fJy.1 0 ax follows immediately. Vn TA � TB =.I o ax).
(c) There is a sequence of continuous functions Ji : A [0. . Never theless. and let V2 JR .j : M [0. .Aj) of part (b) with A.j. q E xj' ( Bj). G2 .. q E M there is a Vi and Vj with E Vi and q E Vj and open sets Bj C Xi(Vi) and Bj C Xj(Uj) so that p PE XiI (Bi). G3.q) = iI: 2i d (Gi(P). with X 2 : V2 JR defined by x2 (a) = a.q) may be ° for p # q) such that p is a metric on each Vi and each Xi is a homeomorphism: (b) If A C JRn is open. .j in a single sequence GJ. can find on M a pseudometric p (a function p: M M JR with all properties for a metric except that p(p. by showing that every neighborhood of ° would have to intersect every neighborhood of * . Show that p is actually a metric on M. Vi)} a sequence ofoneone functions Xi : Vi JRn Vi C M and X (Vi) open in JRn. A 2 . b e such a sequence for each open set Xi(Vi) . a # 0. let d be a bounded metric on JR.C.. It would seem that M ought to have a metric which makes each Vi open and each Xi a homeomorphism. A3. 1. (e) Suppose that for every p . Actually. and Xi' ( Bi) nXjI (Bj) 0. (d) Let fi. then there is some Ji with Ji ( p ) ¢ fi(A nC) .The Tangent Bundle 95 M be any set. this is not quite true: (a) Let M = JR U {*}. where * ¢ JR.). (P) . Let with > PROBLEMS 0 > = = > > we x > > > = • • • = .{ ° i. j 1 . which "separates points and closed sets": if C is closed and p E A . and define p on M by 00 1 p(p.j (P) P E Vii· P¢V Arrange all gi. then there is a sequence A I . Gi(q»). =1 Show that p is the required pseudometric. C Aj. with support Ji c A. Show that there is no metric on M of the required sort. . such that each Xj Xi' : Xi(Vj n Vj) Xj(Vi n Vj) is continuous. Let VI JR and XI : VI JR be the identity. 3. 1]. Hint: First arrange in a sequence all pairs (Ai. . and { (Xi. 2. 1] by _ f g. . . of open subsets of A such that every open subset of A is a union of certain Ai 'S.* X2 (*) = 0. Define gj.{O} U {*}.
E2... 7. bundles over the following base spaces: (i) the disjoin t union of twO circles. but weakly equivalent. B. (Ii) a figure eight c::><::::) . while g is an isomorphism 4. V IRn . A weak equivalence between two bundles over the same base space B is a bundle map (]. 3. .. (c) Show that locally every IIplane bundle has 11 linearly independent sections. E with the vector of is a section. 6.. i. w]q � (q... • 8... Vi ) with M = Ui Vi . (a) Check that p is an equivalence relation on the set of pairs (b) Check that the definition of is independent of the coordinate systems x and y which are used. f B2 fl . SII which are everywhere linearly independent. show that ] = g 0 h where g and h are contin uous maps such that h takes fibres linearly to fibres.e.I (B) for B C V IRn open. (a) Suppose (x.. · .. and is a homeomorphism of B ontO itself.I (V) . is a homeomorphism for a collection (Xi .. Find two inequivalent. on cach fibre..96 Chapter 3 V) and (y. V IRn .I (p) s: s(p) p SJ (p). (b) Show that an nplane bundle � is trivial if and only if there are II sections E SI ... Show that in the definition of a bundle map. x x x x a homeomorphism . (b) Show that if there is a metric on TM such that Ix. (c) Conclude from Problem I that there is a metric on TM which makes each Ix 2. v) .. w). Sn (P) (x.. the map 0 . Show that in the definition of an equivalence it suffices to assume that the map E I 7 E2 is continuous. are linearly independent for all E B... . B . where ] is an isomorphism on each fibre.I (p) . giving rise to two Ix : . maps on TM.. f: BI .. then all Ix are homeomorphisms. v). fl 5. continuity of follows automatically from continuity of ] : E I . . (To prove the inverse continuous..I ( V n V) the sets of the form Ix. V) are two coordinate systems. [y .. (c) Check the remaining details in Theorem l... Given a bundle map (]..I (A) for A C V IRn open are exactly the sets of the form Iy. f.. . note that locally it is just a map x x V IRn .. . Iy : . [x. Show that in .. : E ... V IRn).I ( V) . . v]q � (q. (a) Show that for any bundle . . (iii) the torus.
the same must be true p of :F / W. show that x l + W. and that they represent linearly independent elements of V/W. v] (0) v. (c) Conclude that (V/W) * has dimension ce = 2 e• 14. x ::: Show that all f. .a/ax. JR" [x. 1 3. space (:F / W) * . If x is a coordinate system with X(p) 0. . f* [x. If g : JR JR is Coo show that g(x) g(O) + g' (O)x + x2h(x) for some Coo function h: JR R 1 2. Show that x f (x) /x 2 lim +o exists for all f E W.a]p . (a) Show that the correspondence between and equivalence classes of curves under which p corresponds to the "j equivalence class of .:: o.g E :Fp. . as the next problem shows. Show Jet e: :F that . W E V we will denote by E Vw the vector corresponding to (w. are in V. v). (x) = 0 1 x . If V is a finite dimensional vector space over JR.e has a unique extension to a derivation. for v. (b) Show that under the correspondence the map can be defined by TM Vw > = > > = > = = = > > . L. = x 9.The Tangent Bundle 97 xl . The situation is quite different for C functions.. If f: M N and f* is the 0 map on each fibre. 1 1. (b) For O < e < I .lp ' f* [f* (il](g) erg 0 fl. and p JR be a linear operator with eUg) 0 for all f. (a) Let :Fp be the set of all Coo functions f: M JR with f(p) 0..a. and let W be the subspace generated by all products.o y for y a curve in with y ' = makes correspond to In. (b) Let W be the vector subspace of :Fp generated by all products fg for f. 10. let { x +E O f. p (c) Since (:F / W) * has dimension n dimension of M. As in the case of JR". then f is constant on each component of M. . p x"I + W is a basis for :Fp/ W (use Lemma 2). define a Coo structure on V and a homeomorphism from V V to TV which is independent of choice of bases. g E :F Show that the vector space of all derivations at p is isomorphic to the dual p.. (a) Let V be the vector space ofall C1 functions f: JR JR with frO) = 0.
so is a coordinate system around then directions". i* (8M)p > v) lHIn E points inward. (a) If is a manifoldwithboundary.t ). (a) A map is an immersion if and only if is oneone on each fibre of TM. .v" . (Compare with Problem 233(d). f: M N f p M f. the still has tangent vectors "pointing in all vectors still run through IRn . (b) Let a E IRn1 {OJ C A tangent vector in is said to point "in ward" if.. fog g f M M. the tangent bundle TM is defined exactly as for elen.ents of are p equivalence classes of pairs (x. n n g(p. then aM has a unique orientation aJL such that = JLp for = aJL if and only if every outward pointing W E (d) If JL is the usual orientation of show that aJL is (I)" times the usual orientation of IR 1 = a lHl .: Mp NJ(p). i. it is where a is the inclusion. TlHIn i: M M lHIna en(lHIn). A vector which is not in E is said to point "inward" if x. Although x takes a neIghborhood of E a onto rather than IRn.. TM U TN > . Show that this subspace does not depend on the choice of X. f o g f.(V) lHInx(p) M [v" .) 16. under the identification of with the vector is (a.vn_. . [w.) (e) Suppose we are in the setup of Problem 214. in fact. v Mp p M U lHIn p.t) '" (/(p). TP g: aM x [0. lHIn . If E a and x : > > f.98 Chapter 3 15. (e) Show that if has an orientation JL. (The reason for this choice will become clear in Chapter 8.] N x [0. Show that this definition does not depend on the coordinate system x.l ] ( )p Mp. Vn.\IRnIX(p» C Mp vn v Mp x lHIn. p M lHIn.(aM)p . Mp v) . then the rank of equals the rank of at g(a). Define I) a Show that is obtained from I) by x. the rank of at E is the rank of the linear transformation (b) If at a = where is a diffeomorphism. i. > is a subspace. . i. . More generally. where > O.
. there is one on S2 0.g.i)_p.i)p with (p. show that P has an orientation which agrees with J1. a JL) (aN. b a (bl + b2) = a . and N is IJ X Let f be a diffeomorphism from M to N which is orientation preserving on one copy of and orientation reversing on the other. b) = (Aa) . I) the vector field looks like the foIlowing picture 17.a. I)}. a · b (a l + a2) b a l b + a2 . . That is. b) . (g) Suppose M is with two holes cut out. and there are no zero divisors: . and v on M C P and N C P.The Tangent Bundle 99 by identifYing v E (aM)p S2 with (. v) (S2. bl + a b2 A(a . v) E (S2. Show that identifying  TIP'2 is homeomorphic to the space obtained from T(S2. Suppose we have a "multiplication" map from ]Rn x ]Rn to ]Rn that makes ]Rn into a (nonassociative) division algebra...B. Show that such a vector field can be picked so that near (0. What is the resulting manifold P? Sl [0. Although there is no everywhere nonzero vector field on S 2 . 18. . (p. .0.(v) E (aN)J(p) ..0. which is diffeomorphic to ]R2. • • (a. (a "magnetic dipole"): 19.i) by € {to. b a (Ab) for A E ]R a · (1 .0) a = • = • = .I ).av) is orientationreversing. > (f) If M and N have orientations JL and v and f: (aM . Sl .
b) · (e. . [0.0.J " /l = . a = O). (b) If a # then a . . (Recall that 1 (f.0). (b) Show that the resulting bundle is orientabJe if and only if T is orientation preserving. otherwise it suffices to assume the existence of a unique b with a . Tv). g)'(t) = (f'(t) '. see R. 20. Hint: An arc e from Coo . ' " . b b . nonexistence of zero divisors immediately implies that for a # ° there is some b with ab and b' with b'a = If the multiplication is associative it follows easily that b b'. 0) = 2. using methods of algebraic topology. (d) Multiplication by a is continuous. where ' denotes the transpose.g '( t J'). v) 21. Show that for p E S2 . = ( 1. (a) Every point in snI is a . pg. where T : ]Rn ]Rn is a vector space isomorphism. and for we can use "complex multiplication". > (1. Arner. [Incidentally. . Let M be a manifold. for I.". a . so that we always have multiplicative inverses.d) (ae . The tangent bundles TS 3 and TS7 are both trivial.0. = /l = 2 = 0. Palais. e'(O» ) = ° for all curves e with e ( O) = p and 1c (t)1 = for all t . . i)p are linearly independent.(S 2p) by showing that the inner product (p . (a) Consider the space obtained from I] x ]Rn by identifying with ( I . el . Show that this can be made into the total space of a vector bundle over Sl (a generalized Mobius strip). (0. e2 . Show that M is orientabJe. we can use ordinary multiplication. Monthly 75 (1968).. or 8. complexes.. g(t» ) + (j(t). (e) TSn1 is trivial. 366368. Conversely.0. en are linearly independent. .100 Chapter 3 (For example. the vector pp E ]R3p is not in i.. F. Multiplications with the required properties on ]R4 and ]Rs are provided by the "quaternions" and "Cayley numbers") respectively. 4. f Math. then the projection of a . For a simple proof.. en be the standard basis of ]Rn.) Let eJ . this condition implies that there are no zero divisors jf the multiplication is associative. S.bd. . 23. .ad + be). . (a.) 22. . It is a classical theorem that the reals. The Classification o Real Division Algebras. (f) TlP'n1 is trivial. (c) If p a · el E snI. that n = I. Adams has proved. the quaternions are not commutative and the Cayley numbers are not even associative. el for a unique a E ]Rn. . a · en on (snl. = (1. and quaternions are the only associative examples. see Cai£ulus on Manifolds. J. Suppose that (TM)IA is trivial whenever A C M is homeomorphic to S l .
f) a bundle map which is an isomorphism on each fibre. A vector space structure can be defined on > > > > X X (f. Thus one can "transport" the orientation of Mpo to Mp . define 1C' : E' Y by 1C' (y. The next two problems deal with important constructions associated with vector bundles.e) = e. e) = y. (e) If � is orientable. This bundle is denoted by f*(�). then f*(�) is also orientable. (a) Given an Ilplane bundle � = 1C : E and an mplane bundle ry = 1C' : E' let E" c E x E' be the set of all pairs (e. It is called the Whitney sum � Ell ry of � and ry. (a) Suppose � = 1C : E is a bundle and f : Y is a continu ous map. f(e)) E E'.e) with /(y) = 1C(e). 23. and define f : E' E by f(y. Show that 1C': E' Y is a bundle. together with Prob lem 29. then (f g )*(�) '" g*(f* (m . First consider pairs c. Let 1C"(e. Y. Show that 1C" : E" is an (11 + m)plane bundle. (b) Suppose we have another bundle f' = 1C" : E" Y and a bundle map > > X > > a X > B B > B 24. Show that �" '" �' = f*(�). Hint : Map e E E" to (1C"(e).The Tangent Bundle 101 po E M to p E M is contained in some such A so (TM)lc is trivial. Since 1C : E is a continuous map from a space to the base space of �. e') with 1C(e) = 1C'(e'). possibly quite messy. Show that if � is not orientable. Remark: Using results from the Addendum to Chapter 9. case can be treated by breaking up c into small pieces contained in coordinate neighborhoods. then i*(�) '" �IA. show that f*(� Ell ry) '" f*(�) Ell f*( ry). by using the vector space structure on 1C1 (f(y» . we can conclude that a neighborhood of some Sl C M is nonorientable if M is nonorientable. and is called the bundle induced (from �) by f. The general. (f) Give an example where � is nonorientable. then 1C*(�) is not orientable. (g) Let � = 1C : E be a vector bundle. but f*(�) is orientable. the symbol 1C*(�) makes sense. c' which meet only at po and p. f) from �" to � which is an isomorph�sm on each fibre. e') = 1C(e) = 1C'(e'). > B > B > B. the fibre of � Ell ry over p is the direct sum 1C1 (p) Ell 1C'I (p). and ( j. It must be checked that this is independent of the choice of c. . Let E' c Y x E be the set of all (Y. (b) If f: Y > B. (c) If g : Z (d) If A c and i : A is the inclusion map.
a�n /J the form of Y* 0 (x*).s orientable. and each fibre 1C 1 is a Coo submanifold of E. the orientation for (TM)v can be defined as of x.e2) (1CI (eI l. .·arh point. then T(M x N) :::: 1CM*(TM) Ell 1CN TN .) A different proof that TM is orientable is given in Problem 29. show that � ry is orientable. (b) The Osection of E is a submanifold.x). (Here is a more conceptual formulation: for v E TM.102 = Chapter 3 orientable. Problem 5). and use this to show that Ell is always orientable. and ry is nonorientable.. (c) Given bundles �i 1Ci : Ei > Bi. (f) If � is orientable. a ()o1C) /" a�I /" .e. (h) If is a "figure eight" (c. show that � ry is also non = X � � � X Ell Ell Ell � 25. (e) If � and ry are orientable. show that � ry :::: Ll* (� x ry). (g) Define a "natural" orientation on V EEl V for any vector space V. . (c) If M x N is orientable..1C2 (e2))· Show that this is a bundleI �I x �22 over BI x B2 . U) be a coordinate system on M with X (p) 0 and let v E Mp I:7= 1 ai a/axi ip · Consider the curve c in TM defined by c rt ) = v + a�i /p . (a) If M and N are Coo manifolds. i. . . (d) If Ll: B > B x B is the "diagonal map". Ll(X) = (x. define 1C : E x E > BI X B2 by 1C(eJ. (a) Let be = [a()01C) /" . (b) If M and N are orientable. Show that the Jacobian matrix of *( ) y* 0 (x*) I is of the form This shows that the manifold TM is alw.!). and 1CM [or 1CN] : M x N > M lor N] is the projection on M [or N]. .f.. then 1C* has maximal rank at .(p) 26. 27. carried diffeomorphically onto B by 1C. I .I shows that this orientation is independent of the choice (x. then both M and N are orientable. find two nonorientable Iplane bundles and ry over such that Ell ry is also nonorientable. then M x N is orientable. the bundle T(TM) is (lrientable. (a) If 1C : E > M is a COO vector bundle. 28.
0. (d) If n : E M is not orientable.I).. .. .. A sequence of bundle maps EI E2 l. (a) If T : IRn IRn is a linear transformation. and induction on II. 30.. w) for a unique T*v). the map w . the adjoint of T. Using part (a).( Chapter I I). (This is why the proof that the Mobius strip is a nonorientable manifold is so similar to the proof that the Mobius bundle over Sl is not orientable. . 0:: 2 (b) SO(l) is a point. n) i v ' 29. > > 31. . is orientable if the other two are. E3 with f g = identity of B is exact if at each fibre it is exact as a sequence of vector space maps. w) = (v. so it is connected. Tw) (for each v. then each bundle E. I v ' de a L (a) If � = n : E sequence > B is a o > Hint: (I) An element of the IOtal space of n*(�) is a pair of points in the same fibre. > > > an exact (e. I. then the manifold E is not orientable. (v. For II define f: SO (n) > snI by f(A) = A (po) . which determines a tangent vector of the fibre.. show that there is n*(�) TE n*(TB) O.. and then show that f. (T*v. (c) T(TM) is always orientable.. T* : IRn IRn. Tw) is linear. is defined by (T*v. this informa tion will all be important in Chapter 10. This problem requires some familiarity with the notion of exact sequences (c.The Tangent Bundle 103 Show that (b) Find a curve whose tangent vector at 0 is a/a(X .I (p) is home fomorphic to SO (II . Show that I (po) is homeomorphic to SO (II ..n* X). In addition to being used in Problem 32. (c) Show that 0 (11) has exactly two components. Show that f is continuous and open. (2) Map X E (TE) e to E2 (b) If 0 EI E3 0 is exact. (a) Let Po E snI be the point (0. I). . show that the matrix of T* is the transpose A t .. prove that SO(II ) is connected for all II 0::.) > > > > Coo vector bundle. > The next two Problems contain mOre information about the groups intro duced in Problem 233.. = a di (O) = ax. If A is the matrix of T with respect to the usual basis.I) for all p E SnI . so it is w .
A ) > (y.) (b) A linear transformation IRn > IRn is selfadjoint if so that = = for all W E IRn. Two continuous functions f Ji : X > Y are called homotopic if there is a. If Df(O) is nonsingular. 122). (b) Suppose f : IRn > IRn is Coo and /(0) = 0. (Remember that A is symmetric. (Notice that this also gives us another way of finding the dimension of O(n). (x) = H(x.{O J . The notation f : (X. I] > Y such that Ji(x) = H (x . v) T v (Tv. v) = 32. ) converges. 1R) is homeomorphic to O(n) x IRn (n +' /2 and has exactly two com ponents. while f(IRn .{OJ ) > (IRn . B)) if there is an H as above such that each H.104 Chapter 3 standard basis. Conclude that a nonsingular linear transformation (IR" . The functions H. A) > (Y. w) (v. (e) Show that a positive semidefinite A can be written as A B2 for some B. : X > Y defined by H.{OJ ) > (IRn. (g) The matrices A. (c) A selfadjoint (or the corresponding symmetric A) is called positive semi definite if ?: 0 for all E IRn. B). 1 . 1R) is continuous and H: 1R" x [0. show that f : (IR" . Show that C can be chosen orthogonal. = Ji . At = A. It is a standard theorem that a symmetric A can be written as CDC' for some diagonal matrix D (for an analytic proof. (Tv. IRn . means that f : X > Y and f( A ) c B. B). IRn {OJ ). (d) Show that At . are homotopic as maps from (IRn. B) homotopic (as maps from (X. Hint: Use the Schwarz inequality. {A : det A > OJ and {A : det A < OJ.{OJ ) is T:   T .IR" . A (n)2 . E O(n) and A2 is positive definite. and use part (e). a continuous function H : X x [0. by showing that eigenvectors for distinct eigenvalues are orthogonal. Hint: If A (n) > A and A (n ) = A (" ) . A is always positive semidefinite.{OJ ) with det > 0 is homotopic to the identity map. show that H is continuous. Ji : (X. see Calculus on ManijiJlds. then is selfadjoint if and only if A is symmetric. and positive definite if > 0 for all Show that a positive definite A is nonsingular. A ) > (Y. We call /0. . If A is the matrix of with respect to the (Tv. ) 01) GL(Il. t ) may be thought o fa s a path of functions from Ho = fa to H. : (X. then some subsequence of {A (n) . so that Ho and H. IR) can be written uniquely as A = A. for A C X and B e Y. A 2 where A.) (f) Show that every A E GL(Il . t ) = A(I)(X ) . Tw) T v. IRn . The map H is called a homotopy between fa and Ji .IR" {OJ ) to (IRn. A. v # o. (a) If A: [0. .{OJ) C 1R" . IRn . 1 ] > IRn is defined by H(x. I] > GL(Il. Hint: Consider At . T: T T T* . pg. i) i = 0. A) to (Y. and A2 are continuous functions of A.
By a chord of M we mean a point of ]RN of the form p . Thus. then there is a collection e ofcharts whose domains cover M such that for every (x. ]Rn .The Tangent Bundle 105 homotopic to D/(O) : (]Rn . f o /z : (]Rn . . show that if M has such a collection e of homeomorphisms. Assuming this result. ]Rn _ {Oj) . ]Rn . (e) Notice that the condition on y o x. Let C V be the open ball with center 0 and radius r. Show that if M is orientable. (ii) no tangent plane Mp contains v. we will say that f is orientation preserving at p if T_ J(p ) 0 f 0 Tp is orientation preserving at O.{O)).O). where U. then precisely one of f and T 0 f is orientation preserving at O.I in part (d) makes sense even if y 0 x. If f: U > V is a p homeomorphism. then for any Coo structure on M the tangent bundle TM is orientable. : B.q for p. . let T ]Rn > ]Rn be Tp (q) = p + q. Check that this does not depend on the choice of C V. the map y o x . V) in e. > 2n + I. we can define M to be orientab!e if there is a collection e of homeomorphisms x : U 7 whose domains cover M.I such that M x M and . 33.I is not differentiable. To prove that this definition agrees with the old one we need a fact from algebraic topology: If f : ]Rn > ]Rn is a homeomorphism with /(0) = 0 and T : IRn 7 IRn is T(x 1 . . (d) For p E ]Rn . Let Mil C ]RN be a Coo ndimensional submanifold. Hint: Define H(x. and let be the homeomorphism h : ]Rn > B. (a) Prove that if N (i) no chord of M is parallel to v. _x n ). use Lemma 2. t ) = f{lx) for 0 < t S 1 and H(x . x n. .{OJ). ]Rn . V C ]Rn are open.l . if M is any (not necessarily differentiable) manifold. JRII (Xl . (c) Let U be a neighborhood of 0 E ]Rn and f: U > ]Rn a homeomorphism with frO) = O.{O)) > (]Rn .I is orientation preserving at X(p) for all p E U n V. . . q E M. . To prove continuity at points (x. ]Rn . U) and (y. xn ) = . ]Rn _ {O)) > (]Rn . We will say that f is orientation preserving at 0 if f 0 h is homotopic to the identity map I : (]Rn . B. Hint: Consider certain maps from appropriate open subsets of TM to S N . then there is a vector v E S N . such that e satisfies the condition in part (d). O) = Df(O) (x).{OJ) > (]Rn . h(x) = then (� arctan IXI) x.I .
jRN jRN. A much harder J·esult of Whitney shows that f Mn can actually be imbedded in jR2n (H. Whitney. of Math. 645680). Di ferential Topology). there is a different m sort of argument to prove that a notnecessarilycompact nmanifold M can be imbedded in some jRN (in fact. which gives the same result even for noncompact manifolds (H. 37 (1935).106 Chapter 3 > (b) Let jRN. 17ze se/finl£rsections ofa smooth llmanifold in 2ns pace. Then we may show that M imbeds in jR2n +J using essentially the argument above. if M is compact. given by Problem ·230 (compare Guillemin . lntmduction IJJ Di erentiable Manifolds and Sternberg. then 1C I M is an imbedding. . of Math. Whitney. Di erentiable manifolds. In particular. 45 (1944). :nd Pollack. 220246).J C jRN be the subspace perpendicular to V. (c) Every compact Coo Ildimensional manifold can be imbedded in jR2n + J . Lectures on Di ff f f erential Geometo'. In Munkres. Ann. ff Ann. together with the exis tence of a proper map f : M > jR. Note: This is the easy case of "''bitney's classical theorem. and 1C . Elementary Diif lltial Topo/lJgy. Show that 1C I M is a oneone immersion. Proofs may be found in Auslander and MacKenzie. with N = (n + 1)2).J the corresponding projection.
Vn . defined by (f a f. and the isomorphism from V to V* obtained by sending Vi to V'i is not independent ofthe choice of basis (consider what happens if Vl is replaced by 2 vl l. for (f' fl ' = I v' and > a a g: > g)' g' The dimension of V' is the same as that of V. In fact. then I v· is the identity map of V' and if U V. not just on Vi alone. It is clear that if I v : V V is the identity. The linear function V'i depends on the entire set Vl . we replace each fibre (p) by some other vector space. then (f = f ' . . . if v E V.CHAPTER 4 TENSORS Acommon feature.l ) . Vn is a basis for V. Recall that V' denotes the vector space of all linear functions A : V If f : V 7 is a linear transformation. . then A(V) 107 0 for all A E V'. .l W' > W > JR . then the elements V *i E V*. . . if VI. which implies that . then there is a linear transformation f' : V· defined by rr. = If V" (A) = 0 for every A E V'. On the other hand. . = I w· . > W a are easily checked to be a basis for V'. . These simple remarks already show that f' is an isomorphism if f : V is. we can define v" E V" = (V')' unambigu ously by for every A E V'. and then fit all these new vector spaces together to form a new vector bundle over the same base space. (j'A)(V) = A(jV). for finite dimensional V. ll the constructions on vector bundles carried out in this chapter have a In each case. The simplest case arises when we replace each fibre V by its dual space V'.
= . It is called the natural isomorphism from V to V'*.108 v = O. Actually.) Now let � = 1f : E + B be any vector bundle. We construct the equivalence by mapping the fibre V y's of � over p to the fibre V" of � •• over p by the natural isomorphism. (We first pick an isomorphism from (�n ) . Chapter 4 Thus the map v >+ v " is an isomorphism from V (0 V" . The lack of a natural isomorphism from V to V' prevents us from constructing an equivalence between �. * (�')' is alwa equivalent to �. we can prove that there is no natural isomorphism from V to V'. formulated only after the term had long been in use. Let E' = B to take each [1f1 (p)]* to p. to �n. If you B into a vector bundle. Once the meaning is made precise.) + 1f'1 (p). the bundle �. and �. once and for all. and define the function 1f' : E ' and I : 1f1 (U) U X �n is a trivialization. we will see later that in "most" cases �* is equivalent to �. then we can define a function + + peB U [1fI (p)]'. by We can make 1f': E' requiring that all such [ ' be local tnvializations. this is true merely because the two vector spaces have the same dimension. it gives us an isomorphism At first it might appear that �. for the present. If U c B. since each 1f1 (p) is isomorphic to However. of �. '" �. in the obvious way: since the map [ restricted to a fibre. the dual bundle �. is an isomorphism. readers may ponder this question for themselves. In contrast. (Problem 6 gives a precise meaning to the word "natural".
When this construction is applied to the tangent bundle TM of M. Recall that X is This means that df (� IJ (�I J t fl = CO (f) = = (f 0 C)' (IO) or to (f o e) d(f(e(t» ) dl I.. If f: M � is a Coo function. the cotangent bundle T*M is actually a Coo vector bundle: since two trivializations x and y* of TM are Coorelated. * * * We can thus define Coo. ' (in fact. * the same is clearly true for x. The seclion df is called the differential of f. �* �: � � P >+ (J(p)(s(p» s(p) E 1C 1 (p) (J(p) E 1CH (p) = 1C 1 (p)* . it will appear obvious that this map is indeed an equivalence. denoted by T*M. where e(lo) = p. then w(X) is the Coo function p >+ w(p)(X(p» . sections of T OM. as well as continuous. if is TM). is called the cotangent bundle of M.)I = y 0 (X )I ). to . ' and y. the re sulting bundle. Suppose. This function will be denoted simply by (J(s). that dc/dI ll". Like TM. Even if can be pictured geometrically (e. then a Coo section df of T*M can be + defined by df(p)(X) = X(f) for X E Mp. �* . in particular. Yo ' 0 (X . If w is a Coo section of T*M and X is a Coo vector field. the fibre of T*M over p is (Mp)*. is a section of then we can define a function from B to � by � �* �* .Tensors 109 think about how is constructed.g. there is seldom a operates on If s is a section of and (J geometric picture for Rather.
. dt T *M over U. . . alaxnlp of Mp.. We can also write W(p) = L Wi (p) dx i (p). then the dxi are sections of Applying the definition. In fact.=1 n W = L w. Thus dx'(p). The section w is continuous or Coo if and only if the functions Wi are. . . . . dxi . This means that every section w can be expressed ulliquely on as V for certain functions Wi 011 U. this equation takes the nice form df dl (�) If (x. we obtain a classical formula: .. \. . .. . The section df must have some such expression.IrO Chapter 4 Adopting the elliptical notations de dl for dt r ' de l dg(l) dl for g' (I).=1 if c define sums of sections and products of functions and sections in the obvious way ("pointwise" addition and multiplication). U) is a coordinate system. . we see that = d(f(e(t» ) . dxn(p) isjust the basis of M/ dual to the basis alax' lp.
and it became clear that they must be distinguished from the tangent vectors a/ax" Once this realization came. like dc/dt . because true results were ob tained when these infinitely small quantities were divided into each other (pro vided one did it in the right way).=1 X. no matter how obscurely expressed. Classical differential geometers (and classical analysts) did not hesitate to talk about "infinitely small" changes dx i of the coordinates x i . it was only a matter of making new definitions. . If (X. df must be a function of this change. af . In short. i. which means that df should be a function on tangent vectors. The dx i themselves then metamorphosed into functions. which preserved the old notation. which became tangent vectors. df(p)(Xp) = Xp(f) = " a' . Since df is supposed to be the infinitesimal change of f under an infinitesimal change of the point. then on U we have P ROOF If Xp E Mp is then Thus n . this point of view was in n af .e. No one wanted to admit that this was nonsense.. a tangent vector. like df. except for quotients of infinitely small quantities. U) is a coordinate system and f is a Coo function. THEOREM.(p) L a . Eventually it was realized that the closest one can come to describing an infinitely small change is to describe a direction in which this change is supposed to occur. one can usually see that. =L axi (p) dx 1=1 •:.TeIlS01J 111 I . all classical notions involving infinitely small quantities became functions on tangent vectors. ' (p)(Xp) .just as Leibnitz had. Looking back at the classical works fi·om OUf modern vantage point. and waiting for everybody to catch up.
Then 1 0 C(I) = j(x l 0 e(I).) de al dl (.==1 " al = " �(c(I» . . (Xi 0 e)'(I) L 1=1 axl af = � ax' (C(I» L. In fact. say X i X i (I). =j = I 0 c : � + �.1I . dl L. . ... CLASSICAL FORMULATION = Let I be a function on M. It is the closest approach in classical analysis to the realiza tion of as a function on tangent vectors.(This equation signifies that true results are obtained by dividing by dt again. . . 1 (1) l(x l (I). Then I becomes a function of I.=1 . the diff erential d was usually introduced in the following way: l Let I be a function of the X l . which suppresses the curve C... no matter what tilef unctions X i (I) are.x". . . as we shall point out once again in Chapter 7.) = L: � (c(I» dl ax' .. where M be a curve. . " ( ) Since every tangent vector at of the form dc/dl. say I I(XI.=1 d(f(c(I» ) dl Multiplying by . . we have C(I) is dl af dl = � axl dx" L ..=1 (f 0 C)' (I) ax' dl (The classical notation. . .. = j = l oxI). (Xi 0 e)' (I) . x"(I» . . is still used by physicists. ..1 12 Chapter 4 some sense the one always taken by classical geometers. dx i (e(I» dl dC · dx i dl dl gives 1=1 af dl = � ax' dx" L. We now have We now have dl = � af dX i .x" 0 C(I» .) or = L: D.=1 Consequently. . . x"). c: � + j lffi.](x(e(I» ) . . and x a coordinate system (so that I oX for some function on namely Let MODERN FORMULATION Let X i be functions of t.
i. Nowadays such situations are always distinguished by calling the things which go in the same direction cccovariant" and the things which go in the opposite direction "con travariant". naturally enough. then there is a map J : TM + TN. and which contravariantit's just the opposite of what it logically ought to be. and then operate on it by w. Since J p is a linear * * transformation between two vector spaces. we have a map J p : Mp + Nf(p) . Notice that we cannot put all J* together to obtain a bundle map from T *N p to T *M. we will continually examine the classical way of expressing all concepts which we introduce. should equal J P2 ' On the other hand. Then we can define a section � of T *M bundle. (The complex symbolism tends to hide the simple idea: to operate on a vector. we push it over to N by J .. that a map J: M + N pro duces a map J going in the same direction on the tangent bundle and a map * J* going in the opposite direction on the cotangent bundle.e. There is no corresponding way of trans ferring a vector field X on M over to a vector field on N. the CCtranslation" of classical terminology becomes only a little more difficult than the translation of the German in which it was written. * for each P E M. the same q E N may be /(Pt) for more than one Pi E M. And no one has had the gall or authority to reverse terminology so sanctified by years of usage. it gives rise to a map Strict notational propriety would dictate that this map be denoted by but everyone denotes it simply by (/*p)* . So it's very easy to remember which kind of vector field is covariant. Suppose w is a section of T as follows: �(p) = w(/(p)) 0 J*p. in f act. roughly.) This section � is * denoted. we can say. while a section of T *M is called a covariant vector field. Despite these diff erences.Tensors 113 In preparation for our reading of Gauss and Riemann. and it just happens to have reversed this: a vector field is called a contravariant vector field. Recall that if J : M + N is Coo. and there is no reason why J p. . by J *w. After a while. we can * * do something with the cotangent bundle that we could not do with the tangent *N. Classical terminology used these same words.
(ax'j laxl) is the constant I h is can be seen directly from the fact that the matrix matrix 0 X. .=1 n " W = L WIi dxII .) = el. an). Comparing (*) with D (x' = x' n I dxIj = L � dXI . + anvn) = (a l . " aax. . x'j = L7=1 aijx" for certain au .j 1=1 dxi erentials "change in the same way" as from Theorem I.n x(v. any combination xi.1 14 Chapter 4 The rationale behind the classical terminology can be seen by considering coordinate systems on which are linear transformations.b) If is another such coordinate system. . so the x coordinate system is just an "oblique Cartesian coordinate system". bv. if then x lffi. In this case. x(a l vl + . . Notice that if we also have . . then Clearly so x' aij = ax'j laxl .I ) o XI . .=1 is also called "covariant". . n W = L Wi dxi . COllsequently. _  1' P I x(p) = (a. we see that the diff the coordinates hence they are cccovariant".
a function is multilinear if is linear for each choice of is clearly a vector space. V. . 1=1 These expressions can always be remembered by noting that indices which are summed Over always appear once "above" and once "below" . sections of T *M and TM. . . If Vb " " Vm are vector spaces. Xn . Vk _l . respectively. .. Substituting 115 into the first expression for w and comparing coefficients with the second. given two expressions � a1 L 1=1 a a ax. . Vk+I . . we find that n I W = . . . which is a warning that worse things are to COme. . . Af ter this was firmly established. . XII used to be denoted by XI .l lj = '"" i aaxlj . . this vector space will be denoted . This suggested subscripts Wi for covariant vector fields and superscripts at for contravariant vector fields. If V t+ T(VI . . . .) Covariant and contravariant vector fields. . the functions a'j must satisf y a n . The set of all such T VI . . i.. Vm = V.a X for a vector field. . Vrn. . = � 'I axIl . . are also called covariant and contravariant tensors (or tensor fields) of order I . the indices on the x's were shifted upstairs again to make the summation convention work out. .e. (Coordinate func tions X l .. . Vk_l .. Vk+ h " " Vrn) VI . i n terms o f the Wi . . Wi ax" j L ax'j 1=1 La On the other hand. . We begin with some worse algebra.=1 L.Tensors then we can express the W'.
) ® T = S. . . Vk+i ). so we can define nfold tensor products unambiguously. ® T. . . Vk . . this tensor product operation is itself multilinear. etc. and S E r i (V) we can define the "tensor product" T ® S E rk+i (V) by T ® S(v" . . · · . ® T S. these maps can 1C'' (U) + U X �n k . . then the isomorphisms Ip : I: 1C' (U) 1C' (p) + + {pI U X �n X �n yield isomorphisms Ifwe choose an isomorphism r k ( �n) be put together to give a map ' I : + �nk once and for all. . In particular. U c B and is a trivialization. Vk +' . . O n the other hand. then there is a linear transformation f* : rm(w) + rm (V). (S. . Notice that r ' ( V) = V*. . If f: V + W is a linear transfor mation. Vk +i ) = T(v" . . Vk ) . which thus has dimension nk We can use this new algebraic construction to obtain a new bundle from any vector bundle � = 1C : E + B. S(Vk+ " . . . T ® S is not S ® T. . . if VI .116 Chapter 4 by rm(v). . (S ® T) ® U = S ® (T ® U). We let E' O fcourse. Vn is a basis for V and v *I . defined completely analogously to the case m = I: are easily scen to be a basis for r k (V). v *n is the dual basis for V* = r'(V). then the elements For T E r k (V). S. + + and let If = peB U rk (1C' (p» . . .
ik are.. . and multilinearity of ®./. aXi. and a section is called a covariant tensor field of order k. X'" . . dxn (p) A (p) = L AI) . . .. Xk )..XI . .ik dX11 ® . . il.. . . . Xk ) Xk ) = aA(XI . A(Xh .: A = L A II . n are a basis for Tk (Mp).. .lk (P) dxi' (p) ® .ik ::. . Xk) = A(XI . then the kfold tensor products dXil (p) ® . . . ® dxlk (p). . ® dX lk (p) E Tk (Mp) I ::.···. Xi + X'I . . . .. we just use equation (**) on page 1 1 4. . . Tk (�) �*k T (TM) (Mp) *. il . A covariant tensor field A of order k can just be thought of as an operation A on k vector fields XI . . . . so that TM. il.. . . . . . . ® dX ik now denotes a section of Tk (TM) . . . .Tensors 117 We make H ' : E ' + B into a vector bundle by requiring that all such / ' be local trivializations. Thus. . . t" . . The section A is continuous or Coo if and only if the functions A i . . . . If (x. . .. . . U) is a coordinate system. . For the case of the bundle is called the bundle of covariant tensors of order k.k or simply Ifwe also have then (the products are just ordinary products of functions).··. • • . . The bundle is the special case k = I .. Xk Notice that A is multilinear on the set V of Coo vector fields: A(XI . on U every covariant tensor field A of order k is a basis for can be written dx l (p). . . .i/. ® dXlk .. which yields a function: .: where dX il ® .Xk ) + A(XI . To derive this equation. . . . . .
A : V x · · · x V + F ". . then there is a unique tensor field A with A = A.. 2. . Xk)(P)· = l (p)A(p)(Xdp). . i. Xd p» We are finally ready for another theorem. If P ROOF Note first that if v E Mp is any tangent vector. XI .. . \vhere each at now denotes a constant function and f is a Coo function with I(p) = I and support I C V. . if U) is a coordinate system and is linear over F. . . . Now if v" . THEOREM. it is actually linear over the Coo I Ullclions F. . . . (p) for each we claim that i. Xt (p). In fact. . Xk E V we clearly must define The problem is to prove that this is welldefined: If XI (p) = Y. . . f(p)X. Xk)(P) = A (p)(Xd p). Vk E Mp are extended to vector fields X" . if I is Coo. . one that is used over and over.' k limes (x. . then for we have A(X" . .1 18 Chapter 4 rvforeover.=1 n a I on U outside V. . because A is defined "pointwise". .e. . 1X" . . (p). . then we can define v = L al axl p ' . . . . · · · . Xd p» = I(p) · A(X" . . . then there is a vector Geld X E V with X(p) = v. . . . . .
) For simplicity. = fA(Y). it obviously suffices to show that A(X)(p) X = L. Because of Theorem 2. take the case k = 1 (the general case is exactly analogous). nor will we use the symbol A any longer. and support 1=1 � a .. = 0 if g C V. in particular. Let ( x .b' a x ' V) be a coordinate system around p. so f A(X) evaluating at p gives X(p) = o. •:. AUX) = AUY) A(X)(p) (2) To prove the result. A(Y) (p). Note that Theorem 2 applies. so that on V we = A(Y)(p). � bl (p) . to use the in terminology. Let f be a Coo function with f(p) = 1 and support f C V. we will never distinguish between the tensor field A and the operation A.. so that by (I).. A since bl (p) (g � ) a 1 = (p) o. can write = (The map A "lives at points".Tensors 1 19 (I) Suppose first that X = Y in a neighborhood V of p . the cotangent bundle: a function fi·om V + 'F which is linear over 'F comes fiom a covariant vector field w . then is a welldefined Coo vector field on all of A(X)(p) = M which equals X on V. to the case k = 1. Just as with covariant vector fields) a Coo map f : M + N gives a map f* taking covariant tensor fields A of order k on N to covariant tensor fields f* A of order k on M: . The proof that A(X)(p) = A(Y)(p) when X(p) = Y(p) is in two steps. Then fX = fY. where all b' (p) = o. If g is 1 in a neighborhood V of p. Now A(Y)(p) = = 0. where T k (TM) = T*M.
. We could also use the notation '1k(TM).j. thus. Now an element v ofa vector space V can be thought of as a linear function v : V* + �. . there is an analogue of Theorem 2._ ' aX'Jk aX'}1.: Ah . if and are covariant tensor fields of orders k and I. .. Recall that a contravariant vector field is a section X of TM. A A(p) a. .h� <8> .IP <8> . then we can define a new covariant tensor field of order k + I: A B (A <8> B)(p) = A(p) <8> B(p) Mp x .··. . .. . . aX}k ax}' h. j(.k (p) . .jk (remember that each alax j Ip operates on Mp*). <8> __ . that allows us to dispense with the notation A.···. . then we easily compute that A.···. a A = L A.P" " Pk = L jl. a A(p) = L A}. . j(. if only for the sake of completeness we should define contravariant tensor fields. So each Xp E Mp.IP aX}k ax}'. Wk into a function: A Naturally.. A contravariant tensor field of order k is just is a klinear function a section of the bundle T k (T*M).120 Chapter 4 A <8> B (operating on Moreover. each on Mp* . or simply a A = Lh. respectively..jk A contravariant tensor field A of order k can be considered as an operator taking k covariant vector fields WI . ·· h __ <8> . proved exactly the same way.. " <8> _a_. If we have another such expression.. In local coordinates we can write k + 1 times). we just define V(A) to be A(V). x Mp Although covariant tensor fields will be our main concern. · ·}. " <8> . h. and to identify contravariant tensor fields of order k with operators on k covariant vector fields that are linear over the Coo f unctions F . if we use TdV) to denote all klinear functions on V*.
' (p» .A) = A (S(V» .Tensors 121 Finally. ' (V). given by Generally speaking.' ( V) denote all bilinear functions T: V x V· + R A vector bundle � = 1C : E + B gives rise to a vector bundle 7. the correspondence S >+ 5 from End( V) to 7. obtained by replacing each fibre 1C . although they should be kept to a minimum. we are ready to introduce "mixed" tensor fields. T: V x V · + �. we can take the trace of the corresponding S : V + V. Since both End(V) and 7.' ( V) is linear and one one. sections of 7.' (TM) are called tensor fields.A) = A(V). this map is an isomorphism. Moreover. for 5 = implies that A (S(V» = for all A. let 7. which implies that S(v) = for all v. In particular.' (1C . Given S. .' ( V) have dimension n'. There are all sorts of algebraic tricks one can play with 7. If V is a vector space. given a bilinear e(v. The inverse. Notice that each S E End( V) gives rise to a bilinear 5 E 7. is not so easy to describe. for each v the vector S(v) E V is merely determined by describing the action ofa A on it according to (*). we consider a special case first.' ( V) makes the identity map I : V + V in End( V) correspond to the CCevaluation" map e : V x V· + � in 7/ ( V) 0 0 0. It is not hard to check that this isomorphism of End(V) and 7. To make the intro duction less painful. In particular. 5: by the formula Vx v· + �. 5(V. our isomorphism can be used to transfer any operation from End( V) to 7/ (V). however. V is a basis of V and n T = L T/v·i <8> Vj .j . this number is called the contraction of T. If v" . certain ones are quite important.' ( V). Let End( V) denote the vector space of all linear transformations T : V + V ("endomor phisms" of V). . i.' (�). .' (p) by 7. covariant of order I and contravariant of order I .
given a tensor field A.122 then we can find the matrix A = in terms of the T/. then 7.) These identifications and operations can be carried out. I L a axj ' II (contraction of A) = The general notion of a mixed tensor field is a straightforward generalization. i=1 = II � trace A (p) if we consider A (p) E End(rrl (p» . v*j) T/ . A= A J dX '. defined by S(vt) L aji Vj . which is a section of we can consider each A (p) as an endomorphism of Mp. fibre by fibre. each section A gives rise to af unction (contraction of A) : B defined by p r+ contraction of A (p) = + 2 T L T/. More over. = Chapter 4 (aij) of S.1 (TM). J. If in a coordinate system . obtained by replacing each fibre rrI (p) by End(rrI (p» . aj/ v*j (S(v/» T(v/. and we obtain a function "contraction of A". each A (p) is an endomorphism of rr1 (p). j=1 = II Thus contraction of = = (The term CCcontraction" comes from the fact that the number of indices is con tracted from to 0 by setting the upper and lower indices equal and summing. In this case. . <8> i i. In particular. in fact. a section A of Ti (�) can just as well be considered as a section of the bundle Elld(�). Thus. in any fibre bundle 7i (�). 1=1 1/ ( V) T: V 'v' k times <8> • • • <8> V x V* '. Define to be the set of all (k + I)linear L A.' i limes <8> • • • <8> V* + R . covariant of order and contravariant of order 1 .
.i��. ax'P. . Classical differential geometry books are filled with monstrosities like this equation.·.: then axil axl ax'PI A'g::: g� = L Af(.·.. n' of of oP . In every classical diff erential geometry book. i. Here is an important example... h.j � ax'p . Consequently..k (�) . ·..ijr . A (�) A= L il k . all important tensors are actually defined by defining the nHI "HI x x'.. one will find the following assertion: "The Kronecker delta is a tensor. the classical definition of a tensor field is: an assignment of functions to every coordinate system so that (*) holds between the functions assigned to any two coordinate systems and (!) Or even.. jl.: jl. it is asserted that if one chooses the same functions for each coordinate system." In other words. . J. aX. axh i( /. x.···.k (TM) (�)..: � . cca set of functions which changes according to (*)".. . ·. ..j ax. In fact.al ' " ax'ctkJ.e. i. covariant of order k and contravariant of order I.lh/. Locally. for .Tensors 123 Every bundle gives rise to a bundle Sections of are called tensor fJelds. and if A' = L il. in classical diff erential geometry.a axj et L this is certainly true.ijr • . or simply of type an abbreviation that also saves everybody embarrassment about the use of the words "covariant" and "contravariant".. a tensor field of type can be expressed as � J. Ai.'\'o. then (*) holds. nk+l functions in terms of the coordinate system and then checking that (*) holds.'.
) a collection of functions one for each coordinate system.) LO! dx'(p) <8> a!j IP (v. classically..124 Chapter 4 a " A Le 0f.j j = L. what this equation shows is that '.J.j j b i. .. i.. a I " n Thus isjuSI the evaluation map Mp x Mp' + �. P=I = + �. dxl <8> axj = From our point of view..0. E Mp' with the expressions J.) i.J is a certain tensor field. sat isf ying A(p) AI. independent of the choice of the coordinate system x To identify the mysterious map A(p). in a similar manner... Given a tensor.e. _ L.. it isjust the identity map. The contraction of a tensor is defined. Mp x Mp' we consider = v E Mp and J. aib.(v). considered as an endo morphism of Mp. L bp dXP (p). i=1 = J..J. then A (p)(v.
. . VI . A2) = contraction of: .ax i . The revolution in the modern approach is tllat the set of all vectors is made into a bundle.=1 n so that this sum is a welldefined function.ja ) L L L: ax">: ax a=1 i. .. . . Ifwe consider each A(p) E T.= This is precisel the definition we adopted when we defined tangent vectors as y equivalence classes [x. x' ' aJ (v.yv = L AaPvya P " J1. L. .} n ax = L A !. L ax'a. the functions (�) we are laking B(p) n BJ1. A2.j a=1 ax = L A{a) i. a n to each coordinate system numbers a l ) .j = L AL . (Mp) to be B(p)(v) . defined as the sum of the diagonal entries of its matrix. This calculation tends to obscure the one part which is really necessaryverification of the fact that the trace of a linear transformation. A) >+ A (p)(v. a 'n assigned to satisfy x. is independent of the basis with respect to which the matrix is written.A) . Incidentally. so that vector fields can be defined as sections. and that all other alp.(p)... then 2 E T. a=1 "transform correctly" if the A�r do. V2.Tensors we note that 125 a=1 � A'aa = � (" Aj' � aX.3 (Mp ). a vector at a single point p is classically just an such that the assignment of n TZumbers a l . ax/ja . a tensor of type can be contracted with respect to any pair of upper and lower indices. For example.. . A) . V2. ax'j = ") a'. . A). . While a COlltravarianl vector field is classically a set of n functions which "transforms in a certain way". n . rather than as equivalence classes of sets offunctions.
or as equivalence classes of curves. and hence must come from some A . As we shall see. The tangent bundle itselfwas almost a victim of the excesses of revolutionary zeal. We won't go quite that far. invariant definitions of all the important tensors in diff erential geometry are made by means of Theorem 2. but we will give an "invariant" definition (one that does not use a coordinate system) of any tensors that are defined. We seldom define A (p) directly. . For a long time. which miraculously turns out to be linear over the Coo functions F. the party line held thaI TM must be defined either as derivations. The modern revolt against the classical point of view has been so complete in certain quarters that some mathematicians will give a three page proof that avoids coordinates in preference to a three line proof that uses them. the return to the old definition was influenced by the "functorial" point of view of Theorems 31 and 34. At the appropriate time we will discuss whether or not this is all a big cheat. instead we define a function A on vector fields. Unlike the "Kronecker delta)' and contractions.126 Chapter 4 types of tensors are constructed from this bundle. such invariant definitions are usually not so easy to come by.
(c) Show that U*dyj)(p) = t a(y. For v E Mp. • • Wn for V are equally oriented. 1=1 0 f = identity.···. V) are coordinate p and f(p). Jn • • • + 3. Vn and WI .(v) = df(v)f(p) E �f(p)· v*). . . . ) w* .g: M f: M dx" f* ( . then 1 . express j�(L7=1 a i alax i l p ) in terms of the alay j lp . .j" dyh <8> <8>dyj.. L. more generally. show that 4 . show that f. ah . � be Coo. w*) . . . U) and (y. Let Mn systems around + + (Proposition 23 is the special case (b) Show that a a(y f) a f* ( axi Ip) = � ajXI. is orientable. . (a) If g: N �. dxl (p).) (d) Express in terms of the 2. (a) Show that i fthe ordered bases VI .. .(p) · ifJ If(p) ' L. and suppose that (x. (b) Show that a bundle � is orientable if and only if �. Let + dUg) = fdg+g dj.) } 1 . . .x� f) (p) .Tensors PROBLEMS 127 f: Nm. respectively. If f. . then the same is true of the bases n and n for V*. . Y j=1 and. v* N are Coo. .
. Ci = 1. (vi) I f a j AiA j i s a n invariant for A i a n arbitrary vector.j = 0 for all vectors Ai such that Ai fJi = 0. . n . . check them. show that if bars + casr = 0. An "invariant" is just a (welldefined) function. or b = c and ars is skewsymmetric. Il indicates the component and ex for ex = I. det(AaI') oJ 0. . In each case. the other sets are com ponents of a vector field.�fJ. Hints and answers are given at the end. any vectorfield A satisfy i ing Ai fJ.. and these vectors are independent. if Vi is defined [cJ (iii)] by a ij Aali v j = 0. then a. . (v) If aij and bij are components of a tensor field. so are aij + b lj . then any vectorfield Ai is expressible in the f orm where the a's are invariants. if Qij)J). so Ai jLi means L7= 1 Ai JLi . and by definition 5. . i (iii) If fJ. .ij = a'ji for the coordinates in any other coordinate system. then a ij + aj i are the i components of a tensor. n the vector. and consequently i 1 a ij + aj = (fJj!Jj + fJj (Ji ). i . in particular. . If aij and bkl are components of a tensor field. after (xiii). then aij + Qj = O.) = 0. 1. using the classical methods. then (alj . Remember that the summation convention is always used.is an invariant and either Ai or fJi are the components i of an arbitrary [covariant or contravariant] vector field. and then translate the problem and solution into modern terms. . that is. so are aiJ bki .128 Chapter 4 (i) If the quantity ).1 which satisfy the equation. where fJi is a given covariant vector.: (viii) If ars are the components of a tensor and b and c are invariants.1 and fJi Vi oJ 0. then either b = c and ars is symmetric. (ii) If Aali are the components of n vector fields [in an nmanifold] . The following statements and problems are all taken from Eisenhart's clas sical work Riemallnian Geomet�). i (vii) If a ij AiJ.} fJ.= 0 is expressible linearly in terms of II . where i for i = I. . . .. . (ix) By definition the rallk of a tensor of the second order aij is the rank of the matrix (aU). n . . Show that the rank is invariant under all transformations of coordinates. .1 independent vector fields i Aali for Ci = . .are the components of a given vectorfield.(Jj )�i� j = 0 is satisfied by every vector field �i. (iv) If aij = a ji for the components of a tensor field in one coordinate system.
. w) + T (w. . 'I ···'p then Of.:·. Z). .. (vii) Given w [with w(p) oJ for all pj . where a is an invariant. .is a given vector. let T' (v. . T(v + w.and writing v uniquely as vo+cYo for Vo E ker fJ. (Begin by choosing Yo complementary to ker fJ. then T(Z. ip if j) .such that HINTS AND ANSWERS. . (ix) T: V x V + � corresponds to T: V + V' [where T(v)(w) = T(v.i�p are the components of a tensor in all coordinate systems. . (iii) Given w [with w (p) oJ for all pj . v) . where ai and bj are the components of two vectors.) (vi) For T: V x V + �. For. Then T + T' is determined by S(v) = T(v. v) + T(v. . . For example. . i (xii) I f a ij Ai = aAj holds for all vectors A i such that p iAi = 0 . (xi) Show that the tensor equation aij Ai = aAj. v + w) = T(v. .l. . If O"(Z) = T( Y.a8lj)Ai = 0. . there are everywhere linearly indepen dent vector fields X" . jp is a n even permutation o f i) . . w). The rank of T may be defined as the rank of T (consider the matrix of t with respect to bases vI . (xii) Let V = p' . . . Vn and v* ) .) . show that for the symmetric tensor aibj + ajbi the rank is two. jp) oJ {i" . . . . . . . . v) + T (w. . Xn _. . . can i be written in the form (a j . ip} i f j) . . Show also that a ij = 8 ija. choose Y complementary to kerw at all points. v) . . Similarly. . . v) = for all v implies that T + T' = 0. on S' x � there is an w such that ker w (p. . is one. = ° 1 I 1 if ja = jp for some a oJ fJ Or ia = ip for SOme a oJ fJ or if U" . jp is an odd permutation of iI. w)] . T(v. ..Tensors 129 (x) Show that the rank of the tensor of components aibj. ° ° ° M T(v) = av + fJ(v)y for all v . ip (i) w is determined if w(X) is known for all X. I) consists of vectors tangent to S' x {I}. w) = T(w. and vice versa. . there is a y complementary to ker fJ. v* n) . (This is true only locally. if the equation is to hold for an arbitrary vector Ai . 1f T : V + V and fJ. Z) = w(Z)O"(Z)!w(y) for all vector fields Z. where fJ.E V * and T(v) = av for all v E kerfJ. which span ker w at each point. then (xiii) If j 8 h . .
such that F( I v) = I F(V) and F(g 0 f) = F(g) 0 F(v). . may also be defined . (d) If F is any functor and f : V isomorphism. Vp . . the following diagram commutes: V � V** f 1) Show that there do nOI exist isomorphisms i v : V lowing diagram always commutes. F(V) = V*'. A covarianlfunctor from (finite dimensional) vector spaces to vector spaces is a function F which assigns to every vector space V a vector space F( V) and to ev ery linear transformation f : V + W a linear transformation F(f) : F( V) + F(W). covariant in some and contravariant in others. except that FU) : F(W) + F(V) and F(g 0 f) = FU) 0 F(g). . . Show that for any linear transformation f : V + W. ' Ak ) = T(AI 0 f. V � V* 1** . (f) The "T k functor". . 6. . F(V) = V*. FU) = f'* is a functor. FU) = f * is a contravariant functor. .130 Chapter 4 0 : V x . . . . F(V) = T k ( V). !" . W � W** l 1 + V* such that the fol 7. . W � W* Hinl: There does not even exist an isomorphism i : � diagram commute for all linear f : � + �. FU) = f is a functor. · · . . Ap) = det(Ai (Vj» . + W is an isomorphism. (c) The "T functor". . F(V) = 1k ( V) = Tk ( V*). Functors of morc than one argument. f l fr + �* which makes the (a) The "identity functor". F(V) = V. (e) The "dual functor". x V* + � by O(VI . . Ak 0 f) is a functor. . . x V x V* p times x (xiii) Define p times . (b) The "double dual functor". . F U) = f * is a contravariant functor. then FU) is an A conl>ava/ian/f ullclor is defined similarly. (a) Let iv : V + V** be the "natural isomorphism" iV (V)(A) = A(V). k F U) ( T)(AI> .
show that the following diagram commutes . (e) The functor F(V) = V* is continuous. . (The bundle T*M is a special case of the construction in (d). Show that a different choice of bases leads to a homeomorphic metric on Hom(V. we can identify Hom( V. W). Then F(A) : F(�n) + F(�"). . Choos ing a basis for V and W. See the next two problems for other examples. The bundles 'Jjk (M) are all special cases. . once and for all. . an isomorphism F(�n ) + �k Then F(A) can be considered as a map h(A) : �k + �k Show that Iz : GL(n. �) is a homomorphism. and such that to every trivialization I : 1C1(U) + U X �n (c) The functor 7J (V) = 7 1 ( V*) = V** is continuous. Choose. .1 (p» . �) we can consider it as a map A : �n + �n. (a) Let F be a functor from Y". Call F continuous if this map is always continuous [using the metric in part (a)] . as well as an example of a functor which is not continuous. 9. e ) be the standard basis of �n . (b) How does the homomorphism h depend on the initial choice of the isomor phism F(�") + �k ? (c) Let v = (V I . Show that if � = 1C : E + B is any vector bundle. . W) to Hom(F(V). and show how to construct a bundle F(�).) (d) DefIne a continuous contravariant functor F. then there is a bundle F(�) = 1C' : E' + B for which 1C'1 (p) = F(1C. and consequently give it the metric of �nm . the class of ndimensional vector spaces. wn ) be ordered bases of V and let e = (e l . �) + GL(k .Tensors 131 corresponds a trivialization 8. W) with the m x n matrices.) Generally. W ) denote all linear transformations f rom V t o W. and F is continuous. to yk Given A E GL(n. F(W» . (The bundle Ti (TM) is just a case of the construction in (b). . (b) A functor F gives a map from Hom (V. . vn ) and w = (w" . . the same construction can be used when F is a functor of several arguments. If e + v denotes the isomorphism n taking ei to Vi. . (a) Let Hom(V. .
Show that � is an equivalence relation. and define yn F. this means that �k h(A) �k [ // F(e + v) F(e + w) also commutes.132 where A = Chapter 4 (a. �) is any homomorphism. We will denote the equivalence class of (v. Then Fh : + is a noncontinuous functor. choose ordered bases V . and that everyequivalence class contains exactly one element (v.q' E �k . Show that this is a welldefined linear transformation. (e) Show that the operations [v. j =1 . q') are welldefined operations making the set of all equivalence classes into a kdimensional vector space Fh ( V). (g) Let Ci: � + � be a noncontinuous homomorphism (compare page 380).q .vj . that Fh is a functor. + such that the homomorphism defined there is a functor Fh in part (a) is equal to h.q] = [v.�) + GL(k . W E f(v. = L'j�. h (A)(q)]. qd + [V .wj.Vj . W. F(V) n After identifying F(�n) with �k . THEOREM. define ( V . aq] � (w . q) if q = h(A)q' where W. aj.q) by [v. q]. �) = � be h (A) = Ci (det A). This suggests a way of proving the following. : yn yk (d) For q. aj. + q2] a · [v. If fl : GL(n. �) + GL( I . q] = [w . and f: V + W. and let h : GL(n. yn y ' . q2] = [v. define A by (f) If V.j) is defined by Wi = L aj..) = L'j�. (f)[v .. q) for a given v. (A) = h(A) when we identify Fh (�n) with �k by [e. and that F. q] r+ q.
in addition to mixed tensor fields. show that a nonzero element of F. If is another coordinate system and a · Clx = a t . instead. �) + GL( I .Tensors 133 10. For example. show that x' V a ' = det ( ax' )  ax'. In classical tensor analysis there are. show that the corre sponding equation is a ' = det (c) For this h. . (Xx '. . Conclude that the bundle of odd scalar densities is not trivial if M is not orientable. so every section on can be expressed as a . then are called odd scalar densities. Clx for a unique function Q . assignments for which (a) Let h : GL(n. �) take A into multiplication by det A. other "quantities" which are defined as sets of functions which transform according to yet other rules. a. and consider the I dimensional bundle Fh (TM) obtained by replacing each fibre Mp with F. x axl} are called (even) scalar densities. · a. If is a coordinate system. The Theorem in Problem 9 allows us to con struct a bundle whose sections correspond to these classical entities Oater we will have a more illuminating way): (x. These new rules are of the form A' = A operated on by h ( axa ) ax'P . (b) If.. I G:'� ) I . Let Fh be the functor given by the Theorem.(V) determines an orientation for V. assignments of a single function a to each coordinate system such that the function a ' assigned to satisfies x' a ' = det ( ax' ) · a . V) is nonzero. h takes A into multiplication by 1 det A I..(Mp).
in is an odd permutation of = ip i).k(�n) determines an element r. .k (A ) : r. . .k(A) of GL(IZk+I . . . (See Problem 712 for a geometric interpretation of these relative tensors. . . (�) (�) j (or the same formula with det(ax i lax. . Show that the transformation law for the components of sections of these bundles is = (det A)wr.k(A) w an integer.. . we define 'Ilk (f) : 'Ilk (V) + 'Ilk (V) by I .i" = 8. Also show that gi l . If (det A)W is replaced by I det A l w (w any real number). . . �) be defined by h(A) The bundle F(TM) is called the bundle of (even) relative tensors of type and weight w. ) " .134 Chapter 4 k+1 by taking (d) We can identify r/ (ll�n ) with �n Recall that if f : V + V. . �). is an even permutation of I . Al 0 f). . . n I. n .ill are the com ponents in every coordinate system of a certain contravariant relative tensor of weight 1 . . . Let iz : GL(II. . . AI 0 f. we can consider it as a map A : �n + �n.. . . . we obtain the bundle of odd relative tensors oftype and weight w. 1 . Vk .i n for some Ci oJ {J . �) + GL(n k +I . . .k(�n) + r.j ) replaced by I det(ax i lax. AI ) = T(/(vtl. For k = I = 0 we obtain the bundle of (even) relative scalars of weight w [the (even) scalar densities are the (even) relative scalars of weight 1]. ) I). . . Then r. f(Vk ). .) if ia iJ . (e) Define if if Show that there i s a covariant relative tensor o f weight .1 with these compo nents in every coordinate system. . 'Ilk (f)(T)(V Given A E GL(IZ. . . �).. · · · .
then we define It is not hard to check (Problem I) that a. 135 . U) around p and transfer the vector field X to x(U) C �n . . fn (q). p(O) = P (e. vector fields. In particular. if ex is a diffeomorphism. we have a vector field x. Cl* X does not make sense for Coo functions a : M + N. However. It = X(p(I» . we wish to introduce a coordinate system (x. in general.CHAPTER 5 VECTOR FIELDS AND DIFFERENTIAL EQ UATIONS e return to a more detailed study of the tangent bundle TM. Consider the curve The condition c = x op. e) P.X is Coo on a(M). that is.. and its Let X be a vector field defined in a neigh + M borhood of P E M.e. X)q has "components" P (q). a curve p with Wsections. Recall that. (�IJ = �. (� IJ dp = X(p(I» dl = X(p(I» .X on x (U) C �n . . .. Since this a local question. means that P. There is a function f : x(U) + � n with i. We would like to know if there is a curve p : through p whose tangent vectors coincide with X. (x.e. . i.
. = = (x. .136 hence Chapter 5 �It = x. will denote points of �n. If we use c'(I) to denote the ordinary derivative of the �nvalued function then this equation finally becomes simply This is a simple example of a diff erential equation for a function c : � + �n. A Curve p: + M with (. n . cn (l») We also want the "initial conditions" i = l. . Part of this terminology is still preserved. (� IJ = x. of course. Similar ter minology is applied. then a curve c : ) + M with V (e. . and for a while x. For quite sonle time. e V C'(I) = /(c(l» is called an integral curve for c(0) = x XEU / with initial condition c(O) = x. X)c(t) . . to the diff erential equations one obtains upon introducing a coordinate system.P. a form to which our particular equations never reduce). (X(p(l))) c'(I) = /(c(l» . c. . Solving a diff erential equation used to be described as "integrating" the equation (the process is integration when the equation has the special form C'(I) = /(1) for /: � + �. Ch(l) = f ( C' (I). e) p(O) = dl p <!£ = X(p(l» is called an integral curve for X with initial condition p(O) = p. which may also be considered as a system of II diff erential equations for the functions ci .e. solutions were consequently called "integrals" of the equation. . we will work entirely in Euclidean space.X) x(p(t)) (x. etc. . y. If c �n is open and / : + �n.. .
. the correct solution is for ali i C(I) = 0 .Fields and Diff erential Equations 137 Before stating the main theorem about the existence and uniqueness of such integral curves. In this case. then (F a y)' = I F(y(x» = + C for some C. (They really do make sense.=x+C Y Y Thus the curves are supposed to be solutions.[c(l)f. so C(I) = I + C = x + C· I I hence. we consider some special cases. we must take X I c(1) = This works in all cases except 1 + l /a ' a = O. if F' = I If.) To obtain the initial conditions c(O) = a.Vecto. This can be checked directly if you don't believe the above manipulations. the equation in question asserts that y' = J a y. The equation for a curve C with range R. which would be written classically in terms of a function . c'(I) = . is the special case J(a) = _a 2 is to write The standard method of solving this equation dy dx = y2 y: R + R as �2 = dx _y f �2 = f dX _y I .
or (J) for all I. Unique ness will always be insured when I : V + �n is C I . y Another phenomenon is illustrated by the equation c' (/ ) = C(I) 2/3 . since = = 0. . given by I(a) = a2/3 . can be defined f all I . . except c(/) = 0. This will continue to be true even if we modif the vector field near ° so that it is never 0. written classically as I . there is no K with I/(x) . but this has nothing to do with the case. namely .. and as 1 + 00 it approaches. In terms of vector fields. We say that the function I satisfies a Lipschitz condition on V if there is some K such that I/(x) . .Jla the curve escapes to infinity because the vector field gets big too f ast. but it can also be obtained with a rather Jess stringent condition. even or though X is defined on all of � . (2) C(I) = � 1 3 27 In this case.y l for all x. It might be thought that this somehow reflects the fact that X (0) = 0. but never reaches. as 1 + . the function I. For a > 0. the curves c are (which we missed by dividing by the integral curves of Notice that no integral curve.138 Chapter 5 y). in fact. 0.IU) I :5 Klx . Notice that I(a) a2/3 is not Lipschitz.1(0) 1 :5 Kl x l for x near 0. On the other hand. . There are two different solutions with the initial condition c(o) dy = y2/3 dx C(I) = ° f all I. the curve C (I) = J /(I + J la) is defined for all large I.y E U. i s not differentiable.
Xn+k ) � P(Xn . '"v' n times Thus P(Xn . f(x) = Ixl). A Lipschitz function is also clearly bounded on any bounded set. the sum L::'o e n converges. . � (C n + . that is. .. Let Xo E {xn } inductively by M such that f(x) = x (the function f has a unique M and define a sequence Xn+1 = f(xn ). � Cp (x. so C n + . THEOREM (THE CONTRACTION LEMMA). +C n+k I + Thus the sequence {xn } is Cauchy. so there is some x with Continuity of 0 as n + 00. Xn+k l +k. i. and Jet f : M + M be a "contraction". it satisfies a Lipschitz condition in a neighborhood of each pointthis follows from Lemma 25. . .e. P ROOF NOlice that f is clearly continuous. but not necessarily diff erentiable (for example. y E M. On the other hand. I . Xn+1 Then an easy induction argument shows that = fn (xo) = f a f a . lhat is. The basic existence and uniqueness theorem for differential equations de pends on a simple lemma about complete metric spaces.l )p(XO. f then shows that . Let (M.rector Fields and Dijferential Equations 139 A Lipschitz function is clearly continuous. a C l function is locally Lipschitz. XI). p) be a non empty complete metric space. j(y» Then there i s a unique x E "fixed point"). + Cn Since C < ] . . Xn+l) + . + p(Xn+k I .y) for all x. . suppose there is some C < I such that p(f(x). . . a f(xo).
(I) <>'(/) = I(<>(t)) <>(0) = X <>(/) = x + if it satisfies the integral equation (2) where the integral ofan � n valued function is defined by integrating each com ponent function separately. g) = sup p (f(x) .gil. then the set of all continuous functions I : X + M is complete with the metric (J(f. p) is a metric space and X is compact. hence for step functions. In particular. plus the fact that each ). then the new metric space is also complete. Moreover. I (x) exists since M is complete. . Conversely. Recall that if (M. b) + �n and I : defined on some interval around 0. clearly satisfies V V V. if <> satisfies (J). If a continuous function I: [a. this is basically just the theorem that the uniform limit of continuous functions is continuous.a). then a continuous function <> : (b. then the set of all continuous functions I : X + M is a metric space if we define the metric (J by If M is bounded. where 11 / 11 = sup I/(x) l · xeX Our basic strategy in solving diff erential equations will be to replace differen tiable functions and derivatives by continuous functions and integrals. <>(/) .x = <>(/) . hence continuous.<>(0) = For the proof of the basic theorem. bJ + �n satisfies 1 /1 :5 K. and thus f continuous functions.!.1 40 Chapter 5 We are going to apply the Contraction Lemma to certain spaces offunctions. if M is complete. It I(U) dUI :5 K(b . then we do not even need X to be compact. we note that it is true for constant functions. if M is a m n oo compact subset of �n . we need only one simple estimate. then <> is diff erentiable. If c + �n is continuous. then l' <>'(u) du = l' 1(<>('1» duo To prove this. b) of step functions. which are unif or orm limits on [a. g(x» . g) = II I . xeX (J (f. so l' 1(<> (u» du. thus af = f 0 ex is continuous.
Let Xo E and let a > 0 be a number such that the closed ball E2a(xo). Then for each x PROOF. Sa is clearly The curve I Sa(l) . where c �n is open. of radius 2a and center xo. b) we have l' I(a(u» du d bL by (I) E2a (xo)}. a E M. is contained in U. so Sa(l) E B2a (xo) c E2a (xo) for i E (b.(u) 1 la(u fJ . b). H Thus S: M M. < :s a Since Ix . Let I : + �n b e any function.y l for x.Vector Fields and Diff erential Equations 141 2 .I(Y)I :s Klx .xo l 2a. fJ E M. Moreover.xo l :s a. Let E Ea (xo) there is a unique ax : (b. b) V such that a/(I) = I(ax(l» ax(O) = x. for all I E (b. Then + li Sa . THEOREM.SfJ l = s�p < bK sup h<u<b If I(a(u») . for any I E (b.l(fJ(U» dU I by (2) = bKlIa . which will be fixed for the remainder ofthe proof + Then M i s a complete metric space. it follows that ISa(l) .b) by M = { continuous a : Sa(l) = x + (b. define a curve Sa o n . Now suppose a.fJ lI · . Y E E2a (xo). Choose x E Ea (xo).x l = < If I(a(u» u l by (3) .b) + (the integral exists since I is continuous on E2a (xo» continuous.b). Choose b > 0 so that (I) (2) (3) b :s aIL (4) b < 1 1K. For each (b. Suppose that V V V 1 / 1 :s L on E2a (xo) I/(x) .
This clearly implies that (J (a + E B2a(XO) f or sufficiently small > 0. in fact. we pay for it by finishing off with a finicky detail: The map a is the unique {J(I) = X + f f({J(u» du. the unique fixed point ax of the map S is the unique curve with the desired properties. So it must be that sup A = b.1 42 Chapter 5 Since we chose bK < I (by (4» . which contradicts the fact that a = sup A. a : (b. We can now use a simple least upper bound argument. (J(I) = X + provided that f f({J (u» du B2a(xo) [the open ball] (J(u) E B2a(xo) s o certainly if for all u with O ::s u I. Suppose a < b. •:. Let A < b and (J (u) E B2a (xo) for O ::s u I } . Having used the elegant Con traction Lemma. by (**).b) + B2a(xo) with a(l) = x + 10' f(a(u» du. in B2a(xo). this shows that Hence S has a unique fixed point: There is a unique S : M + M is a contraction. b) + This. {J : (b. A similar argument works for b < 1 ::s O. Consider first numbers t > O. < (J (u) E B2a (XO) = {I : O ::s 1 for all u with 0 ::s u ::s t . To sum up. We have already seen (statement H) that for each 1 with 0 :s 1 < b. We clearly have (J (u) E B2a(xo) for O ::s u < a. < Let a = sup A . So (J(a) E B2a(xo). is not quite what the theorem states. is in V satisfying Reason: We claim that any such (J actually lies in B2a(XO). alas. s s) .
(Jt'(I) = /((J. We will write ax (I) as a(l.x» x» . if iJ(l) then = a (10 + I). 3. We now revert to the situation in Theorem 2. that is. Hence {JI (I) = (J 2 (1) for sufficiently small I. then the functions {J i satisfy the same differential equation.(/) a. (J'(I) = a'(lo + I) = /(a(lo + I)) = /({J(I)). •:. so it equals I. by Theorem J. (/) . THEOREl'v!. and have the same initial condition (J. a/(I) = /(a.X) = /(a(l.x) = x d di a(I. (O) = a l (10) = a2 (10) E V. Suppose /: V + �n is locally Lipschitz. b) x Bp(xo ) + V [i.e. Let x E V and let a" a2 be two maps or on some open interval I with a l (/). 2. This remark allows us to extend the uniqueness part of Theorem 2. (Ji (I) = a.(O) = x i = 1 . It is clearly also closed and nonempty. Ifwe define PROOF. but we will frequently use /(a(l. alai or dIdl in this . that is. around each point there is a ball on which / satisfies condition (2) of Theorem 2 f some K or (and hence also condition (I) f some L). (lo + I). Thus the set is open. Suppose a l (10) = a2 (10) for some 10 E I.vecfor Fields and Dijferential Equations Notice that solutions of the differential equation 143 a' (I) = /(a(I» remain solutions under additive changes of parameter. so that we have a map {I E I : ad/) = a2 (1)) a: satisf ying (b. x) a(O.a2 (/) c V and Then a l = a2 on I. Dl a(l.x)..
then the integral curve Cix with the initial condition Cix(O) = x differs from the integral curve Ciy with initial condition Ciy(O) = Y only by a change of parameter. THEOREM. if we fix I. This map Ci is called a local flow for f in (b. To picture this map Ci. In fact. Then . the map 1 >+ Ci(I. the best we can do is to draw the images of the integral cUIVes Cix. the whole flow Ci is continuous (as a function of both I and xl: 4. is always continuous. PROOF. For each fixed x. then the flow Ci : (b. On the other hand. This map <p. Let us denote the map S defined in the proof of Theorem 2 by Sx. so the two images overlap. then the map x >+ Ci (I. we denote it by <p.b) x Ba(xo).: <Pdx) = Ci (I .1 44 Chapter 5 discussion] . to indicate explicitly the role of x. b) x Ba(xo) + V given by Theorem 2 is continuous. X) [ = Cix(/)].X) gives the result of pushing each x along the integral curve through it. If y = Cix (/O). If f : V + �n is locally Lipschitz. for a time inteIVal of I. To focus attention on this map.X) for b < 1 < b lies along part of the curve through x.
axll + . We will just accept this fact. this is a very hard theorem. . you must first learn the elements of Banach spaces. we have . besides.S.axll 1 ) I S (I + b K + .I'l ll · If S. then Recall also that in Theorem I the fixed point any Ct .b) x Ba (xo) + V satisfies a (O. b) x Ba(xo) + V is also C k Unfortunately. .axll S lIax .YI · I ay of Sy is the limit of S. . cxx. pp.: I 1 _ bK l x . pp. In order to read this highpowered proof. are consequently Coo if f is Coo. . If additional conditions are placed upon the map /..SI'l I s bKli a . then tJlej/ow a : (b .Y I · . Since the map a : (b. and then read about diff erential calculus in Banach spaces. then further smoothness conditions can be proved for a. + (b K n.ay ll S 1 Since Ilax ayll = sup la(l. x) a(l.). including the inverse and implicit function theorems (Real and Functional Ana!Jsis. . A clean exposition of the clas sical proof is given in Lang's Introduction to Diff erentiable Manifolds (2nd ed. so we obtain � Ilax . Real and Functional Ana!Jsis (3rd ed. and a recently discovered proof can be f ound in Lang.). In fact. + IIS.x) = x. Notice that the maps <P. y)l. when you're fin ished you'll also know about Banach spaces. this certainly proves continuity of a . 371379.S. but this is probably easier than reading the classical proof (and.a for lIax .Syax ll + I Syax . 360365).) Ix .T!ector Fields and Differential Equations Recall that 145 li Sa .I ax . and differential calculus in Banach spaces). including the HahnBanach theorem. Hence Cl)' = /� S. If f : V + �n is Ck .S.yl s l _ bK l x . denotes the nfold iterate of Sy.
<Ps = <Ps <p.X).X) E Ba(xo). . (J (I) = Ci(I. Roughly speaking. then Ci(I. In other words.X» = Ci(S + I. say: (J'(I) = f({J(I» (J (O) = Ci (S.. If we now let <p. then the point Ci(S. or if I sl. \I\�thout requiring any more proof. X). so we can also define y(l) = Ci(I. Everything we if a a have said. Consequently. Ci(S. that for ls i < e each <Ps is a diffeomorphism.1 46 Chapter 5 x Continuity of Ci and compactness of {OJ e > 0 such that lia/2 (Xo) imply that there is some Ci : (e. <p.(x) = Ci (I. We have also noted that satisfies (J(I) = Ci(S + I. II I. then <Ps(<P' (x)) = <ps+. <P'+s = <p. x) f x E Ba/2 (XO). (x). with inverse <Ps . and x E Ba/2 (xo). can be resaid. defined f or Is + I I < e.Ci(S. [If x E Ba/2 (XO).X» f 1 1 1 < e. in particular. then the integral curve with initial condition x stays in Ba (xo) for III < e. Is + I I < e and x. (x) E Ba/2 (xo). : Ba/2 (xo) + �n b e <p. since it is local.I = <ps. l s + l l < e.e) x Ba/2 (XO) + Ba(xo).] So if l s i < e. This satisfies y' (I) = f(Y(I» y(O) = Ci(S. III.Ci (S.X). on a manifold.X). we can or ls i. This shows. X» III < e.
THEOREl\![. 6. The support of a vector field X is just the closure of {p E M : Xp oJ OJ. q E V. and I I" I < ej2. such that there is a unique col lection of diffeomorphisms V + (V) c M for II I < e with the following properties: <PI: <PI (J) <p : (2) If (e./2 <Pr <Pr [[<P.Vector Fields and Dijferentiai Equations 1 47 5. defined by (I. . ls + 1 1 + < e. <PI: Clearly <p : (e. iterated k times] iterated k times] for k 2: 0 for k < O.e) x M + M is Coo. Notice that by uniqueness. . en ). or on all of M. PROOF. (3). then there are diffeomorphisms M + M for all I E � with properties (I). 11 + s l < e. . this can be attained. Let X be a Coo vector field on M./2/2 ' <P.0 0 0. If X has compact support (in particular./2 = 0. ls l . THEOREl\![. if M is compact). Then there is an open set V containing p and an e > 0. and M. e) x V Isl. Let It is easy to check that this is the desired {<PI }' ./2 <P. p) = . .:. <P <PI { <P.. . = is Coo. if '" support X.. = <Pi for E Vi n V .0 0 = k (ej2) + I" with k an integer. So j we can define if E Vi = . and and each is a diffeomorphism. write or <PI <PI 1 q <Pdq) {<Pi(q) q q <pi (q) (q) q if <PI+S <PI o <ps = 11 1 . . .. . en and diff eomorphisms <p. . then Xq is the tangent vector at I 0 of the curve I >+ <PI <Pdq)./2 <P. Let e = min(e l . III. d E V. To define f II I 2: e. and let p E M. (3) If The examples given previously show that we cannot expect to be defined for all I. Cover support X by a finite number of open sets VI . (2). " " V given by n Theorem 5 with con'esponding el .. In one case however. then q <P q) <P <Pdp)..
_ o � 0 of the curve t r+ <P. Then there is a coordinate system (x. and p = 0 E lffi. to say that Xq is the tangent vector at I = amounts to saying that '!!:.} (vector fields used to be called "infinitesimal transformations"). we obtain a "local I parameter group oflocal diff eomorphisms". ( ". M + �. (/) dl = df (c (I» dt = (/ o c)'(I). from � to the group of all diffeomorphisms of M. we can assume that The idea of the proof is that in a neighborhood of 0 there is a X(O) = unique integral cUl"e through each point (0. and which also has important theoretical uses.a'f+_ _ _ . if q lies on the integral a ax ' on U. the map <P. U) around p such that X = f (Xf) (q) = Xq / = hlim (<Ph (q) . This equation will be used very frequently.f(q) . The vector field X is sometimes called the "infinitesimal generator" of {<P. .n . alal ' lo ' . (q) PROOF It is easy to see that we can assume M = �n (with the standard coor dinate system t I " . . In the local case of Theorem 5. 7. and is said to be generated by X. Recall that I r+ Thus.} given by Theorem 6. a'.(O. . THEOREM. Condition (3) in Theorem 5 can be rephrased in terms of the action of Xq on a Coo function f. an ). say). The first use is to derive a corollary of Theorem 5 which allows us to simplif many calculations involving vector y fields. . l\1oreover.1 48 Chapter 5 The unique collection {<P. Let X be a Coo vector field on M with X(p) oJ O. is called a Iparameter group of diffeomorphisms. . or more precisely. .
for i > X. ( � IJ ] we can at least compute ( f) = = lim � [J ( X (O. curve through this point. ) a l o (f o X Moreover. an» . an). a . . . ./(x (a» ] (Xf)( x (a)). (�l ) a .uations q 1 49 We compute that for a = (a' .rector Fields alld Differential 1!. . Jet X generate <PI and consider the map X defined on a neighborhood of 0 in �n by x (a' . ./(x (a))J h l + h . . . . . (a/at ' ) = X o X . . . . The second use of the equation = �� I o' (Xf) (p) = lim .1+ 0 l is more comprehensive. is equivalent to X = a/ax' . O» . . h . . . X. [/(<PI> (p» . . a2. .I coordinates of q and the time interval it takes the curve to get to q as the first coordinate. . a2. a2. an as the last n . . = Jim _ [/( x (a' + 0 �l = Jim . a (f) = a ' a (f o X) /1 I . . . h./(0)] �l hO h Since X(O) = a/al ' l o by assumption. . . this shows that X. . an» . ./(O)] iI+O II = lim � [/(O. . This is the desh'ed coordinate system. f it is easy to see that the equation or X. . an) = <Pal (0. .[/(<PI> ( x (a» ) . . which we have just proved. an). . . . . . ./(x (a» ] I h /t70 = = /1+0 lim � [/(<Pa l +h (0. . To do this./(p)] . •:. . we will use a2.I may be used as a coordinate system in a neighborhood of O. . The fact that X/ can be defined totally in terms of the diffeomorphisms <PI> suggests that an action of X on other objects can be _ I . . .O = I is nonsingulaJ: Hence x = X. 0) .
. If Y is another vector field. it is another function. To emphasize the fundamental similarity of these notions. Y appearing here is a special case of the vector field a. . but we will Soon compute this vector field explicitly in a coordinate system. I (LxY)(p) = }.150 Chapter 5 Lx / obtained in a similar way.h (p). and these facts will then be obvious. for a : M + N a diffeomorphism and Y a vector field on M. (p) of X through p j p The definition of Lx Y can be made to look more closely analogous to Lx / and Lx w in the following way. we first introduce the notation for Xf. we define a new covariant vector field. whose value at p is denoted variously by (Lx f)( p) = Lx / ( p) = ( X/) ( p) = Xp (f). integral curve / . then A fairly easy direct argument (Problem 8) shows that this limit always exists. This is the limit of certain members of Mp' .w (p)]. the Lie derivative of w with respect to X. Thus (<Ph. and that the newly defined covariant vector field Lx w is Coo. Y)p].' (p) = <P..(<Ph .� [Yp . oh The vector field <Ph ... by (Lxw) (p) = . I l�o h [(<Ph' W) (P) . <P. We call Lx / the (Lie) derivative of / with respect to X . If a : M + N is a diff eomorphism and Y is a . Now if w is a COO covariant vector field. Recall that if Xp E Mp. Y)p = <Ph' (Y¢_h (P» is obtained by eValuating Y at <P.. Y defined at the beginning of the chapter. we can define the Lie derivative of Y with respect to X. and then moving it back to p by <p". .
(<Ph h h� h [ ( ¥h )p . then a vector field a * Y on M can b e defined by Now notice that Of course.!. Nevertheless.(<Ph* !Y)p]  = lim = h70 .!(<Ph (p» <Ph* Y¢_I.I ) * y. the one used in finding (/g)'(x). 2. [Yp = lim . then (3) Lx!Y = X! · Y + ! .<Ph* Y¢_h (pl] h !(<P (P» A. 8. The cal culation is made a lot easier by first observing (a .(<Pk. The remaining equations are all proved by the same trick. LxY.'!. (4) Lx ! · w = X! · w + ! · Lx w. If Lx Y and Lx w exist. + Y2) = LXYI + Lx h Lx (w. Finally.(A. (pl] lim = 1170 !(P). '¥ = k�O k [Yp . PROOF (I) and (2) are tri\�al. we will stick to the original (equivalent) definition. [f(p)Yp . If (I) (2) then (5) Lx (w(Y» = (Lx w)(y) + w(Lx Y). k7o k . LX Yi and Lx Wi exist for i = 1 ."ctor Fields and Dijferential Equations vector field on the range 151 N. We now wish to compute Lx w and Lx Y in a coordinate system.Il [f(p)Yp . a * (Y) is just * Y)p r I r I ' *y A.[. Y r 'f'!t* ¢I.<Ph* (/Y).(pl] [ ] . + W2) = Lx WI + Lx W2 . PROPOSITION.p = h�O !. [Yp . then Lx (Y.Y)p] = (LxY) (p). (P) · + h�O !(P) . We will do number (3) here.O l lim _ h I .k* y)p] Y ] r Yp .h h _ /t". if w(Y) denotes the function p r+ w(p)(Yp) and Lxw and Lx Y exist. lim (Lx !y)p = 1170 I [(/Y)p .
. aXl.(p) dx (p) .</Ih) (p) h ax' [ _ ] {this step will be justified in a moment} .) ...] a(x' 0 .dx' (p)] = lim Now the coefficient of dx j (p) is iJ"'O I � a(x' O </lh ) ' h L aXl· /t7 0 . but . n 8ai . = To justif *) we note that the map A (h . O · a I lim . where y is x. thus a2 A /ahax j = a2 A /ax j ah . .. Recall (Prob lem 41 ) that if f : M + N and y is a coordinate system on N. lim � a(x' O </Ih) h ax' .152 The first limit is clearly approaches Chapter 5 f(p) · Lx Y (p). Suppose X = L:7= . It now follows that ( a aa' x') . a . which is what the interchange of limits amounts to.f(</Idp»  k Xf(p). then f* (dy') = t j =1 .P X( = .We first compute Lx (dx').(p).o h [( 0 aXl.j=1 . In the second limit.  . • + : We are now ready to compute Lx in terms of a coordinate system (x. Lx dx' = I: axj dx ' . a'a/ax'.</10) (P) ax' (x' 0 </10)] . I a( �x�f) dxi We can apply this to </Ih * . Then Lx (dx')(p) = l�o h [(</Ih*)dx' (p) . [ _ 8' J (*) ] = lim = It. V) on M.x' </1. Yp . . . the term in brackets = while an easy argument shows that </IMY¢_h(P ) k7 0 lim f(p) . j =1 We could now use (2) and (4) of Proposition 8 to compute Lxw in general.dx' (p) [ . I Xl I � a(x' O .p /t". q) = X ' (</Ih (q» is Coo from � x M y to �.
· ax' axj axj ax' 0= .j .= .YX. . YJ. axj + b'Lx ( axj ) n .· axj "1 axj ax' . Clearly x = L a'. ax' ax' thus. because <P Lx oj = Lx [dX' ( a!j )] (Lx dx') ( a!j ) + dx' (Lx a!j ) so aa' a dx' (Lx. aaj ) axj ' Laxl Lx This somewhat complicated expression immediately leads to a much simpler coordinatefi·ee expression f or Y. If j . abj a n . aa' a = L a'. To compute Lx ( alax' ) we could imitate the calculations of Lx dx'. then Yj is a function.and Differential Equations 153 we are really interested in computing LxY. = . but there would be a complication. .. Summing over j and then interchanging and obtain '�I i j in the second double sum we '�I LXY = L j=1 .a The second partial derivatives which arise here cancel those in the expression for Y(Xj).aax. M + N is a Coo function.=1 axI n . a a Lx (b'.=1 a b � (� a'.b'.. also denoted by [X. T!ector Fields h* on vector fields involves one more composition than <Ph* on covariant vector fields.= Using (3) we obtain a .) . and (5) of Proposition 8. a n aa' a Lx . axj ) = Lxb'. and we find that LxY = XY ... so XYj = X(Yf) makes sense.L. and we can now use (5) to get the answer immediately..' . (4).L b' . = . The trick needed to deal with this complication has already been used to prove (3).
We are now in a very strange situation. If I : ( e. This sort of thing irks some people to no end.Yp (Xf). e) x M + � with = 0 for all p E M. Y]p(/) Often. 9. In Chapter 3 we proved a lemma which for the special case of � says that a Coo function I: (e. Y]p(g) + g(p)[X. •.e) x M + � is Coo and I(O. [X. p) ds r' a . p) = g(O. and can theref be considered as a member ore of Mp. : .YX. but they have Gcen proved equal using a coordinate system. p) = 10 as I(SI. Fortunately. A straightforward verification shows that [X. Y] have both been defined independently of any coordinate system. Y]p ( /g ) = l(p)[X.(sl) ds. Define g(l. Y]p is a derivation at p.p) then there is a Coo function g : (e. LEMMA. p) = Ig(l.e) + � with 1(0) = 0 can be written 1(1) = lg(l) for a Coo function g : (e. This has an immediate generalization. Y]p(/). (0). e) + � with g(O) = /. 1(1.154 Chapter 5 [X. in this case the coordinatefree proof is short. p) al & (0. Two vector fields Lx Y and [X. Y] (which is called the "bracket" of X and Y) is just defined as XY . note that this means = Xp (Yf) . though hardly obvious. so that [X. namely g(l) = l' /. pl . PROOF.
so I (p» ] Jim lim h /1 h70 . Let f: M � be Coo. we have L[x. there is a f amily of Coo functions g.(<Ph* Y)P](f) = h70 � [(Yf)(p) .Ly a Lx (which might be written as . Lx Z]. Lax. Ly D. it follows imme Since we obviously have Lx (aY. + bY2 ) = aLx Y."ctor Fields aud Differential Equations 155 Y are Coo vector fields. [Y.Yp(Xf). Lx X = O. [X. [Z. +bX. THEOREl\I[.YX reveals certain f acts about which are by no means obvious from the definition. [Lx. Y. Xll = o. go = Xf.[.[Yp . = f + lg. (P) (f a <Ph) = Y¢_I. + bLx Y2.l�o (YghH<Ph (p» = (LxYf) (p) . so sO [X.·ators on Coo functions.(YfH <Ph . Lx Y The equality Lx Y = [X. Y = aLx. diately that L is also linear with respect to X: LxY = LyX. + 1 0. Y] = [Y. Let X generate <p" II I < e. Z] = [Lx Y. then LxY = [X. Y]. Yll + [Y.(P)(f + hgh ). This equation is capable of two interpretations in terms of Lie derivatives: (a) Lx [Y. (b) as ope.YI = Lx a Ly . If X and By Lemma 9 Then (<Ph* Y)p(f) = <Ph* (Y¢_I. Finally. Y] = XY . Y + bLx. X]. PROOF.(P» (f) = Y¢_I. Zll + [Z. X] = o. Consequently. Z] + [Y. a straightforward calculation proves the 'Jacobi identity": [X. on M such that f a <P.(YgO) (p) = Xp(Yf) . *:. Clearly [X.
pops up in the definition of prac ticaJly all other tensors. at time I = 0. X)p = <Ph.156 Chapter 5 Finally. not over the Coo functions F. for Xf may have a nonzero derivative in the Yp direction even though (Xf) (p) = o. [X. h(p). gY] = fg[X. it does not necessarily follow that [X. ] is no t a tensorthat is.(p) is just the tangent vector at p <Ph (P) time t = h to the curve t >+ to the curve <P.. note that Lx Y is linear over constants only. and thus the tangent vector. Bef ore proceeding to examine its geometric interpretation. In f act. we will endeavor to become more at ease with the Lie derivative by taking time out to prove directly from the definition of Lx Y two f acts which are obvious from the defi nition of [X..* X¢_Ii (p) is the tangent vector. (I)  LxX = 0. Y]p = Oin the formula [X. to the curve But this tangent vector is just Xp. The bracket [X. even if Xp = 0. Y]. Thus. y(t) = <P. . Y]. If X generates <Pt. Y]p(f) = Xp(Yf) Yp(Xf) the first term Xp(Yf) is zero.(p) . it certainly suffices to show that «Pt.g(Yf)X. for reasons that \vill become more and more apparent.X)p = Xp for all h . at time I = 0.. (p). Y] + f(Xg)Y . but the second may not be. Thus </>. Recall that (<Ph.. Now X¢_. or a simple calculation using the definition of [X. shows that [fX. YJ. the bracket operation [ . Y]p does not de pend only on Xp and Yp (which is not surprisingwhat can one do to two vectors in a vector space except take linear combinations of them?). Proposition 8. but on the vector fields X and Y. although not a tensor.X¢_. In particular.
/(q)] : 1 2. LEMMA. consider the curve c : (e. of course.I }. If this by Corollary f all or Given q. X = X ifand only if <PI oa = a o<PI Then PROOF. LEMMA. Let X generate {<Pd and and only if <PI 0 "'S = "'s 0 <PI for all s. . or Conversely.X. If a : M for all I.I (q ) (/ o a) = lim _ [(/ 0 a)(<p" (a . Y generate {"'d.I (q))) . then LxY(p) = O.(/ 0 a) (a. e) + Mp given by .0 h I . X generates {a 0 <PI 0 a .l. so Lx Y(p) = O. COROLLARY.I (q» ] .I (q)](f) = X. Let a : M + N be a diffeomorphism and X a vector field on M which generates {<PI}. PROOF We have (a. I (q)) . p E M. •. . an integral curve starting at some other point can never get to p). suppose that [X. To develop an interpretation of I I . !. Y] we first prove two lemmas. Y [X. Then a. [X. [ f(a 0 <Ph 0 a 11". If <PI 0 "'S '" "'s 0 <PI for all s .X)q (/) = [a.T4!ctor Fields and Differential Equations 157 Since Xp = 0. + M. Then Yp = 0 and (2) If Xp and Yp are both 0. then a. so that 1 3. then <p Y '" " is true f all t. iI70 1 = lim . Y] = 0. then clearly Lx Y = O. Y] = 0 if 1 2. the unique integral curve c with c (O) = p and dcldl = X(C(I» is simply C(I) = P (an integral curve starting at p can never get away. conversely.
k. • . then there is a coordinate system (x. (O) = <p {l�o X [(<Ph.158 Chapter 5 I h c' (/) = lim .fJ � k. The remarkable f act is that the condition [X. 1 2. (P)] } using (*) with q = <Pd p) By Corollary = 0. As in the proof of Theorem 7.[c(/ For the derivative. : c(O). for the existence of the desired coordinate system. Xp] = ° for I � ex. so there is no hope offinding a coordinate system satisf ying (*) unless [X. (p)] " = <P. • . If X" .<p" Y¢_. . c' (/). [ a�' ' a�2 ] = a Y = ax2 • 14. Consequently c(/) = alJ s. 1jJ.Y¢. If Y is another vector field. Y] = 0. that = I . M = �n . <p. = 1jJ. V) around p such that a Xa = axa on V. •. . that ex ]. ex :::::: PROOF.. (p) . and. . . Y)¢. everywhere linearly independent of X. of this map into the vector space Mp we have h_O + h) . THEOREl\t[. a short calculation immediately gives the result 0. . . . Xk are linearly independent Coo vector fields in a neighborhood of p. I = l'!:'o h [<P. <P.C(I)] . as well as necessary. k. for a a We have already shown that if X(p) oJ 0. so <p" Y = Y. Y)¢_. Y] = ° is s'!!ficient. and [X. by a linear change ofcoordinates. (<Ph. we can assume that p = 0. l . then there is a coordinate system x with X = a/ax'. (p) . . . then we might expect to find a coordinate system with H a X = ax " However..
. . . As in the proof of Theorem 7. To make use of it. x ( a l . . . There is a more complicated. . n. . . . ( . . We thus see that the bracket [X. o al' 0 � l Ci = . .. . . generales {4>f } ' define X by I x ( a l . an) = 4>�. x . . . . . . then for sufficiently small h we can (I) follow the integral curve of through p for time h. (0) = a . ak +I . . an) .vector Fields alld Diff erential Equations If 159 X. X (2) starting from that point. » and our previous argument then shows that . . more difficult to Pl"Ove. . ax' •:. . (4)�. . which makes this assertion much more precise. it shows that f each between 1 and the map X can or also be written a ax l ' Ci k. . Xa = � . . al' Thus x = X I can be used as a coordinate system in a neighborhood of Moreover. . (0. . . the extent to which the integral curves of X and Y can be used to form the "coordinate lines" of a coordinate system. . an) = 4>� (4). p = O. Y] measures. . y X y X (4) then follow the integral curve of Y backwards for time h. . follow the integral curve of Y for time h. in some sense. Ci = k + I . ak + . and less important result.k I. If X and Y are two vector fields in a neighborhood of p. XI = Nothing said so f uses the hypothesis [X" Xp] = O. just as before we see that (� I ) � I 0 = 1 X.( . (0. we can compute that . .0. (4):. . . . . ( 3) then follow the integral curve of backwards for time h. . we ar appeal to Lemma 1 3. » . an» . . .0. .
.O. . . . (a) a dl. C (I) = a. . (1.0. O.I).160 Chapter 5 If there happens to be a coordinate system x with x (p) = 0 and a x = ax " (J) (2) a Y = ax2 ' then these steps take us to points with coordinates (3) (h.I) = a2 (1. 0) (h. which means that c' (O) = y'(O)] is the following.1) (b) . then Moreover. Let c(h) be the point which step (4) ends up at. . h) = <P. Even when X and Y are Oinearly independent) vector fields with [X. (I. . c'(O) = O.O. .1) = a. .(I. . If we define 1 5. a2 (0. PROPOSITION. c is the constant curve p up to first order.0.h) = 1fJd<Ph (P» a2 (I.O. h . Then the CUlve PROOF. the parallelogram is "closed up to first order". . .0) (O. h.d1fJh (<Ph (p» ) a. h) = 1fJd<Ph (1fJd<Ph (p» » . so that this "parallelogram" is always closed. Yj oJ 0. . The meaning of this phrase [an extension of the terminology "c = y up to first order at 0". . . that is.0). . O) (4) (0.(O.1) a.
(c). The proof. a(/ o aJ ) at = Yf o aJ a(/ a (2) = Xf o a 2 t . . that this operator e" (0) is a derivation. and a second addendum concerning linearly independent vector fields in dimension 2.O)] a Consequently. t .T4!ctor helds and Differential Equations and for any (c) (d) (e) while 161 Coo function f : M + �.aa (f) a(/.)(O. is an horrendous. : .h) = Xf(adO. O) + D2 (/ a aJ) (O. 0) = DJ (/ a a. 0) + DJ (/ a (2) (0. repeated use of the chain rule gives using (b) using (a). cal culation. Thus.)(O. the bracket [X.aa(/ a. 0) + D2(/ a (2) (0. 0) + [Ddf a (2)(0. e" ( 0) E Mp.al l aa. but clever. (e).h» .Xf(p) + Yf(p) + Xf(p) = o e" (O) (/) = (/ a e)" (O). which ends the chapter.(O. 0) + D2 (/ a a. (A more general construction is presented in Prob lem 1 7. (d). we can define a new vector e"(O) or d2e/dt 2 l o by A simple calculation shows. Y]p is related to this "second order" derivation. using the assumption e' (0) = 0. Until we get to Lie groups it will not be clear how anyone ever thought of the next theorem. Whenever we have a curve e: (e. •.) = Yf o a. and (f) give a (/ 0 (') ' (0) = Yf(p) . h (/ a c)' (0) = DJ (/ a a. but can easily be skipped. It is followed by an addendum containing some additional important points about diff erential equations which are used later.)(O.e) + M with e(O) = p and e'(O) = 0 E Mp. 0)] = DJ (/ a.)(O.) It turns out that for the curve e defined previously. 0) + [Ddf a J)(O.
. = (/0 CY.)(O.J)(O.2) (0.. Using the notation of the previous proof. Y]p .. I) PROOF.. ) (I.2XYf(p) by (e) by (b) and the chain rule by (d) by (a) and the chain rule by (c) and (f). Chapter 5 c"(O) = 2[X.df 0 CY.J (/ 0 CY. (2) 2D2.df o CY. 0) = DJ.)(O.)(O.2 (/ 0 CY.162 J 6..J (/ 0 CY.O) + D2( Xf 0 CY.2(/ 0 CY. 0) + 2D2. since (/ 0 C) (I) we have H (/ 0 cJ" (0) = DJ.2[DJ (Xf 0 CY. DJ. 0) by (d) XXf(p) .)(O.2YYf(p) . 0)] = 2XYf(p) .2)(0. 0)] by (d) and the chain rule + D2.J) (O.2[DJ (Yf 0 CY. 0) + 2D2. 0) + D2 (Yf 0 CY. J (/ 0 CY..2 )(0.0) + D2.) = 2[DJ (Yf 0 CY.) (O.2(/ 0 CY. 0)] = 2XYf(p) .0) = 2XYf(p) .O) = Y Yf(p) Now (J) by (e) by (e).0) by (c) and (f).2 D2 (Yf 0 CY..0) DJ(Xf 0 CY.2)(0.2) (0. O) = Dd.J)(O. = = = .)(O.J (/ 0 CY.2)(0.2XXf(p) + D2. Since (b) gives we have (3) D2.2 (/ 0 CY.2)(0.. 0) XXf(p) . 2 (/ 0 CY. O) + D2 (Yf 0 CY.O) = 2DJ (Yf o CY.2(/ 0 CY.)(O.Yf o CY. THEOREM.2)(0..2YXf(p) .O) + D2.2) (0.2) (0.2)(0.O).0) + 2D2 (Xf 0 CY.J)(O. 0) + D2.
. 0) = YYf(p) + 2XYf(p) + XXf(p) . . D1 J (f 0 CY(1)(0.2 (/ 0 CY(1)(0. 0) + 2D2. by (c) and (f).2 (/ 0 CY(2)(0.T4!ctor Fields and Diff erential Equalions Finally. from 163 we have (4) D2.1 (/ 0 CY(1)(0. 0) == Substituting (J)(4) in H yields the theorem.0) + D2.:.
b) x W + � X �n with (a' .s. x). There is also a clever trick.c) x V + �n .X) = X. Another omission in our treatment of diff erential equations is more glaring: the diff erential equations <>'(1) = /(<>(1» do not even include simple equations of the form <>'(1) = g(I).x» wherever it occurs in the proof. X) = /(a(l. at<> (I. or else just replace <> by 1 >+ <> (1 .<>(I. 2) = a : (b. /(S.x» a (O.lo.X» <>(O. Then there is a flow I(s. that <>(IO.c) x V + �n+l i<> (I. a _.X).x) = x. x) = (s. For the first component function & 1 this means that .x» . well have required.s. x) = ( I . a a ata(I. we would like to solve equations by where / : (c. Define I: (c.x) = /(I. let alone equations like <>' (1) = 1<>(1).x) = s.164 Chapter 5 ADDENDUM 1 DIFFERENTIAL EQUATIONS Although we have always solved diff erential equations with the initial condition we could just as a at<> (I. X) = /(<>(I. s. x» by /(I. S. To prove this. One way t o d o this i s t o replace /(<> (1.<>(I.x) = 1 a' (O. S.X» al <>(O.x) = x. In general. one can replace 0 by 10 everywhere in the proof of Theorem 2. for some to.
1 7.lo. S. even if / : (a. x) '" (s.X» (J (O. consider the special case of a linear diff erential equation a'(I) = g(I) .&2 (I.S. Then is the desired flow with a ii/{J (I. x) = x (by first finding & with & (lo. x).a(I)).X» = /(& ' (I. b). x) = /(I. then (c · a)' (l) = c · a'(I) = g(I) ' c · a (l) so c·a is also a solution of the same diff erential equation. b) x �n + �n is Coo. X» .X» '" /(s + 1.x) = s + 1.x).!' thus 165 For the second component & 2 w e have &' (I.S. alld Diff erential Equalion..X) = X. s. X) = g(I) ' x. na t by considering the curve 1 r+ (J (I .S. b). In this case If c is any n x n (constant) matrix.b).r�ctar Field. a 2 ii/& (I. which may have solutions defined only on a small time interval. If g is a continuous n x n matrixvalued function on (a. we could also have arranged for {J (10.& 2 (I.s. then the solutions of the equation a'(I ) = g(l) · a(l) can all be defined on (a. This remark allows us to prove an important property of linear diff erential equations. distinguishing them from general differential equations a'(I) = /(I. {J (I. . S. PROPOSITION. /(I. a(I). where g is an 11 x 11 matrixvalued function on (a.X» . x) = /(&(I. Finally. Of course.
(J'(I) = g (I ) . 1 hus Ci may be extended past I) as c . with any given initial condition. By uniqueness. Similarly.) "1 0.X) = g(l) . in a neighborhood of 10. a contradiction. let I. �. Ci must be defined for all I with a < I :s 10 . •: . c . x is locally Lipschitz. Hence there is c with (c · (J}(I *) = Ci(I*). Pick (J with Then {J (1 * ) "I 0 for 1 * < I. If the extended solution Ci is not defined for all I with 10 :s I < b.166 Chapter 5 PROOF Notice that continuity of g implies that /(I. . So for any 10 E ( a. (J(I) (J(I. or close enough to I. f I near I. (J coincides with Ci on the interval where they are defined. b) we can solve the equation. Extend it as far as possible. be the least upper bound of the set of I 's for which it is defined.
where V c �2 is open and p E f(U). . . . (O) = (e. . but have diff erent parameterizations.)o. we might hope to find an immersion f for which the j parameter cUIVes in the i th direction lie along the integral curves of the X. X 1 = o. and that X. PROOF.. . We can assume that p = 0 E �2.. On the other hand. .T4!ctor Fields and Differential Equations ADDENDUM 2 PARAMETER CUR VES IN TWO DIMENSIONS 167 If f : V + M is an immersion from an open set V C �n into an l1dimen sional manifold M. we know that there is usually no immersion f: V + M with p E f(U). A simple example (Problem 20) shows that even this modest hope cannot be fulfilled in dimension 3. an ) is called a parameter curve in the j th direction. l. aj+l . However. Similarly. q is on a unique integral curve of X through a point (X l (q). in the special case of dimension 2. . . 2 X l (q) The map q >+ (X l (q)... •:.X2 (q» we proved precisely this fact in Theorem 7. for we might not have [X.. 0).x2 (q» is Coo. . aj_ l . Its inverse. whose i th parameter lines lie along the integral curves of X. PROPOSITION. Every point q in a sufficiently small neighborhood of 0 is on a unique integral curve of XI thl"Ough a point (0. such an imbedding can be found: 18. . . Then there is an imbedding f : V + M. . X2 be linearly independent vector fields in a neighbol·hood of a point p in a 2dimensional manifold M. is the required diff eomorphism . whose parameter cUn"S in the i th direction are the integral curves of the X. with Jacobian equal to I at 0 (these facts also follow from the proof of Theorem 7). Given 11 vector fields Xl . in a sufficiently small neighborhood of 0. .. the curve I t+ / (a} . .. Xn defined in a neighborhood of p E M and linearly independent at p . Let XI . .
C C � 2 is the graph of a monotone function g. . t) = crt). Then there is an imbedding f : V + M. crt»�. g(y» takes the diagonal {(x. y) r+ (c.' (x). and let c be a curve in M with c(O) = p and c'(t) never a multiple of X. x)} to C. or X2 . Let X" X2 be linearly independent vector fields in a neighborhood of a point p in a 2dimensional manifold M. y) r+ (x. we can state 1 9 . and f which f(t. whose i th parameter lines lie along the integral curves of Xi . or c . Moreover. PROPOSITION. for example. we can further arrange that crt) maps to (c(t). f any particular or parameterization c = (c" (2 ) : � + � 2 of C. then the map (x.y) r+ (a(x). Consequently. y). tl(y» for diffeomorphisms a and tl of �. If. which gives us considerable flexibility. where V c �2 is open and p E f(V).168 Chapter 5 We can always compose f with a map of the form (x. by composing with (x.
y) E W there is a unique Ci = Ci (x. x) = f(l. b) + V for each initial condition x and "parameter" y. then Ci: (b. 4. Yo) E V x V. then there is a Coo vector field X on � such that Ci. : TM + TN is Coo. VYe sometimes have to solve equations "depending on parameters") a ii/Ci(I. V C �n are open. Ci' (I) = (J(I) (J ' (I) = f(I. y. y.Ci' (I» Ci(O) = X Ci'(O) = y. Ci(I. f(y» < p(x. 5. Prove that there is a neighborhood W of (xo. y E M. Ci(I). (b) If Ci : M + N is a diffeomorphism. Yo) and a num ber b > 0 such that for each (x. 3. y) for all X. where f: ( e.b) and Ci"(I) = f(l. if we write Ci (x . y). For example.[ector Fields and Differential Equalions 169 PROBLEMS 1. (c) If Ci : � + � is Ci(I) = 1 3 . 2. the equation Ci'(I) = yCi(I) Ci(O) = X. b) + V with Ci'(I) E V for I E (b. Find a nowhere 0 vector field on � such that all integral curves can be defined only on some interval around O. then Ci.x) : (b. { Hinl: Consider the system of equations Moreover.y) : (b. then Ci. and let (xo. . where V. Ci(I). {J (I» . and X is a Coo vector field on M. X» Ci(O. c) x V x V + �n . Find an example ofa complete metric space (M. x. y) (t) = Ci(I. for open V c �n and V c �m. X is a Coo vector field on N. but f has no fixed point. y.b) x W + V is Coo. x) = x. X is not a Coo vector field.c) x V x V + �n be Coo. y. Let f : (c. p) and a function f : M + M such that p(f(x). and we are solving for Ci (y. (a) If Ci : M + N is Coo.
X) (a result which follows directly from the original equation if ! is C 2 . j: (b.x» .x) = !(I.x» .x) = D2/(a(l. xo).a(l. D. by induction. and that if the derivative with respect to these arguments is denoted by D2a.! is C ' . i. (a) Define by If Chapter 5 a(l) = xeyt.x).e. y. = D2D..j(I.170 with solution is such a case. x) = !(a(l. it follows that D2 a is differentiable if D. x) = (y. x) = !(a(l. x» a(O.x) = x can be reduced to equations of the form (*) (and thus to equations a ai"a(l.a of the fOl"m (**). ultimately). then a is diff erentiable with respect to the arguments in W. x) = (y. x» j(t. y. x» a(O.x. and conclude that a satisfies (*). if ! is C 2 Diff erentiability of class C k is then proved similarly. D2a(l. y. a(l. x) = (O. a(l.x» . y. the hard part is to prove that if ! is C ' .a(l. x» · D2a(I. a2) = a : (Yo. y. c) x V x V + � m X �n show that we can write for some a. y.). (b) Show that equations of the form (**) a ai"a(l. since D.b) x (a' . x) = !(I.] D. [When one proves that a C k function ! : V + �n has a C k flow a : (b. then D. W + �m X �n is a flow for j in a neighborhood of ( c .b) x W + V. Since (***) is an equation for D. y. a (l. so that a ai"a(l.
p=O =O 2N (A + BJ P ( = (exp A) (exp B). where f g(l) dl denotes some function G with G'(I) = g (one can obtain all positive multiples simply by changing G). and let IA I denote the maximum of all l aij I. entry of the partial sums converge absolutely for each (i.A) = I . ill and uniformly in any bounded sel. we have I A I s i A l s n I A I. (a) Consider a linear differential equation 171 Ct' (I) = g(I)Ct(I). . The remainder of this problem inves tigates the extent to which similar results hold for a system of linear diff erential equations. is clearly differentiable (it is even analytic). (f) (exp A) (exp . p=O p. p. Show Ct(I) = ef g(r)dr . B).[eclor Fields and Dijjerenl. (b) Let A = (aij ) be an n x n matrix. N AP )( N BP ) + RN (Notice that for I A I.ons 6. (e) If AB = eA "" I + A + 2! + 3! + 4! + .al Equal. (d) Show that exp(TA T.= p i L i L p. A2 A' A4 j)th = BA. where g: � + �. �n2 . the usual norm of A E = B (= exp(O) . . 2 2 (g) The map exp. then exp(A + B) Hint: Write and show that I R N I + 0 as N + 00. considered as a map cxp : �n + �n . Show that IA + BI s i A l + I B I IA B I s n I A I · I B I · (c) Conclude that the infinite series of n x n matrices exp A converges absolutely [in the sense that the (i.' ) = T (exp A)T' . so exp A is always invertible. so that we are solving for a realvalued function that all solutions are multiples of Ct. Show that exp' (O) (B) L .) ..
show that the function al/axi. U) is a coordinate system on M.B) x M + M is a I parameter group of diff eomorphisms. t . (a) Let M and N be Coo manifolds. t. so every system of linear equations with constant co efficients can be solved explicitlythe exponential of J� g(s) ds = t A can be found by putting A in Jordan canonical form. and let (i) Let A � + �n b e diff : B(t) = exp(A (t» . Check that if the coordinate system x is x = XI . ( a/at ') = X 0 x. let 1( · . If B'(t) denotes the matrix whose enllies are the derivatives of the entries of B. �n + M. q) denote the function from M to � defined by 7. show that for every Coo function I : M + �. for x : X= a/ax' is equivalent 10 X. q » (P . 2 erentiable. defined by al a(f( · . the limit . (This is clealiy true if A(s)A(t) = A ( t ) A (s) for all S.172 Chapter 5 (h) Use the limit established in pan (g) to show that exp' (A)(B) = exp(A) · B if A B = BA . show that B'(t) = A'(t) · exp (A (t» . axi is a Coo function on M x N.) (j) Show that the linear differential equation et'(t) = g (t) · et (t) has the solution et(t) = exp prollided that g(s)g(t) = g(t)g (s) for all s. (b) If <p : (B. q). q) = � (p).) (fo' g(s) dS ) 8. For a Coo function I : M x and q E N. (This certainly happens when g(t) is a constant matrix A . If (x. then N + � p >+ I( p . provided that A(t)A'(t) = A'(t)A(t).
Y]. 9. and conclude that for every Coo vector field ant vector field w on M. i . and conclude that 4> * (df)(Y) = Y(f 0 4» . (v) . and f: M + IR a Coo function. M. }. X and covari exists and defines a Coo function on (d) Treat Lx Y similarly. unnecessarily clumsy way. In the next part we get a much simpler proof that Lx Y = [X. the limit 4>* (1.. e) show that 4>* is Coo. x TM + TM is defined by (c) If 4>* : 1 73 (e. using the technique which appeared in the proof of Proposi tion 15. (a) Prove that Lx(f · w) = Xf · w + f · Lxw Lx[w(Y)] = (Lx w)(Y) + w(LxY). isjust a special case derived in an The formula for Lx dx . If X generates {4>. (b) How would Proposition as 8 have to be changed ifwe had defined (Lx Y )(p) (b) Using (a). Lx that for Y E Mp. v) = 4>. show directly from the definition of II. (c) Let X and Y be vector fields on M. (a) Show that [Lx df(p)](Yp) = Yp(Lx f). Give the argument to show that tion 8. and defines a Coo function on M. derived in the text.Hxtor Fields and Diff erential Equations exists. 4>h* Y"'_h(P) + Yp in the proof of Proposi 10. define Lx df = d(Lx f).
If V I . . (</>. . and A is a tensor field of type (J) (�)J I (LxA)(p) = "_0 _ [(</>. . .. . . . b . .. . Vk E Mp. Y]p(f). ). Vk . Z be the vector fields a X = Zay a = Lx Y(p)(f) = [X. . . . 0) = Yp(Xf). Check the Jacobi identity. . . we define M generates {</>. (b) If the vector field X on (a) Check that under the identification of a vector field [or covariant vector field) with a tensor field of type (�) [or type this agrees with our old </>* Y.+ x ax a X + b Y + cZ is an isomorphism (from a certain set of vector fields to and that [U.x. If A is a tensor field of type on N and </> : M + N is a diffeomorphism. A[) = A (</>(P»(</>* V I . </>* Vk . 1 .c' (O) 1 2.Xp( Yf) D2a(0. r+ (a . (b) Show that the flow of aX + b Y + cZ is a rotation of about some axis through o. Y. c) E 1R3 1R3) 1R3 14. Yilz a az a a . A[ E Mp* . . on M. . . we define </> * A on M as follows.1 74 Show that Chapter 5 Dl a(O. lim . . . . .A(p)]. then (�) [4>*A(P)](VI . . O) = . Conclude that for c ( h ) = a(h. 13. Z=y ax ay (a) Show that the map a Y = z. h) we have . On 1R3 let X. (r l ) *A[). and A I . V] r+ the crossproduct of the images of U and V. A I > .1 ) * A I > .* A)(p) .
. Ap+ " .=1 + L A ( X1 . (c) Show that Lx (A + B) Lx{fA) = X{f)A + jLxA).lk .=1 show that the coordinates of Lx A are given by k n . .=1 . .!t _ ! ··. (�).. J ' a i + " " Al J. WI) + L A (XI . A + Lx. Xk . . Xk . j n .). . . W I . .. aai . . Xk .. ® . . Lx(A (XI . . A/ _. . . A � J .=1 n . Lx(CA) = C(Lx A). . . . . . . (CT)(v" Show that = contraction of . LxXi. . .' and Diflmlliiai Equations Show that 175 Lx (A ® B) = (LxA) ® B + A ® Lx B (so that in particular). .+ I .. V. . . W" .. . . Vk . (�). .. . . x . . WI ) can be obtained by applying contrac tions repeatedly to A ® X. . . . = LxA + LxB Lx. WI). . Vk_" A" . . .lleclor Field. W I . A . .. Xk . . . .J l}cr+J . . . (f) If A has components Ai ' in a coordinate system and X = L .. . ." A" .lk' = l .=1 I .. . k I .. V. . . • . . Ap_" A. . X cr LL 0'=1 . . use (d) to show that (LxA)(X" . . . .! ! !cr A' J ···. . wI) '':':: a " ai� " " LL L ax' cr=l j=1 . A) f+ T(V I . . . ® w[. A /_ ' ) (v . . . . Lx Wi . . 'J / . . . V. . . . A = Lx. ._ I . Xk . . Hint: We already know that i t is true for A o f type (d) Let be any contraction (�).•laJ11a+J ·'·'k . . +x. . ® Xk ® W I ® . . . . a axj n ai a/axi . . h � (LxA)IJJ . WI») = (e) Noting that A (X" . WI . W I . • .
0») Let A be a tensor field oft)l)e G).1 76 Chapter 5 15.h. Define g (t) = that the righthand derivative (�). then this unique extension is Lx . (0) . Y. Show . (d) Show that LJx = ILx  DXfj!!dJ .g(O) h  _ /. _ r g+ (0) . 2 (Use Taylor's Theorem . Show that if we define DAI = 0.. DY + DI . Show that the tangent vector e"(O) defined by e"(O)(/) = (1 0 e)"(O) can also be described by e"(O) = 2y'(0). (e) If T is of type (�).!� g + (h) . Hillt: Check this for functions and vector fields first. If we take DI = XI and D Y = Lx Y. and the Coo vector fields V to V. and (3). define y(t) = e ( . . DC = CD. so that we can consider A ( p ) E Elld(Mp). show that Generalize to tensors of type 16. then DA has a un�'1ue extension satisfying (I). Let D be an operator taking the Coo functions F to F. (c) Show that ( DA W) ( P) = A ( p) * (w ( p» . such that (J) (2) D ( A (I) D is linear over ® B) = DA ® B + A ® DB IR (3) for any contraction C. (a) Let I : IR + IR satisfy 1'(0) = o. DA X = A (X)./I ) for t ::: o. 1 (. (a) Show that D has a unique extemion to an operator taking tensor fields of type to themselves. (2)./1 ) for t ::: o. then A (X) is a vector field for each vector field X.) (b) Given c : IR + M with c'(O) = 0 E Mp. such that D : F + F and D : V + V are linear over IR and D U Y) = I .
Y with 51p = Xp and Yp = Yp. Let around pe x (p) 0. x is a coordinate system 0(/2) xi (e(/» ai/ 2 + 0(/ 2). Y]p = n a f. Given Xp.. 0 0 11 (0) . and b J if} a' b J . axi Ip ' a '. For Xp. xa IP) = i. show that where [X. be the curve of Theorems 15 and 16. (c) The rank of (d) Let I : M + define Show that (a2fjaxi axj (p» N p have •• a2I axi aXj (p). then there is a neighborhood U of E IR such that I. . 1 0) to the tangent vector 18. = denotes a function such that 19. : Mp Mp + NJ(p) x ( (( is a welldefined bilinear map. p Using the fact that [X.leclor Fields and DijjerClltiai Equations 1 77 1 7. show that 1 (Xp. IR is independent of the coordinate system. Y) g ) = 51p Y g 0 I»· I. (b) Show that I •• n n (LI axa i Ip ' j�1n . as a critical point. Y]p(f) conclude that it is welldefined. choose vector fields X. so that I. Yp E Mp. If with = and e"(O) defined by e"(O)(f) = (f 0 e)"(O).L1 L j= "= a' . = 0. show that e: 0 c•• (0) : 1R0 x 1R0 + Me(o) takes ( 1 0 . Yp) •• is symmetric. (e) If IR + M has as a critical point. Yp E M and g : N + I (X. (a) Let vectors Define I : M + IR have as a critical point..I (U) is diffeomorphic to x u.p = O. (a) If M is compact and is a regular value of I : M + IR.
20. as shown in the first part of the figure below. q2 sufficiently close. show that Proposition 18 does not hold in dimension 3 . then II (qI l and I. and let X will be a vector field one of whose integral curves is the xaxis. respectively. while certain other integral curves arc parabolas in the planes y = constant. (c) It follows f rom (b) that if all points of N are regular values. let Y and . then there is a neighborhood U of q and a diffeomorphism </> : II (q) x U + II (U) with I(</>(p.178 Cha pter 5 by a diff omorphism </> : 1 .I (q) x N? Z b e unit vector fields always pointing along the y . Using the second part of the figure. I)) = I. If I is onto N.q')) = q'.1 (0) x U + II (U) with 1(</>(p. if M is compact and q E N is a regular value of I: M + N. In IR ' .I (q2 ) are diffeomorphic for q l . 0») More generally. Hint: e Use Theorem 7 and a partition of unity to construct a vector field X on a neighborhood of II (0) such that I.and zaxes. does it follow that M is diff eomorphic to I.X = dldl.
• . a Coo distribution if such a vector field X can be chosen to be Coo in a neighborhood of each point. that is. if for every p E N we have i* ( N ) = /. a " . we have seen that the integral curves of a vector field where i : N + M is the inclusion map. Hardy. . Besicovitch. is spanned by a vector field locally. Littlewood.CHAPTER 6 INTEGRAL MANIFOLDS PROLOGUE A mathematician's repu tation rests on the number of bad proof" he has given . so that global results can be obtained.p C Mp . . quoted in]. . S. we can choose (in many possible ways) a vee to!' field X such that 0 # Xq E /. we just choose a vector field X with 0 # Xq E /. even n the previous chapter. For a given p E M. Then /.p p I on a manifold M may be definable only for some small time interval. This al'gument actually shows that for every p E M there is a coordinate system (x. G. For a l dimensional distribution the notion of an integral curve makes no sense. 1 79 .q f q in a neighborhood of p. with p E N. The function /. Instead of a vector field. but we define a (Idimensional) submanifold N of M to be an integral manifold of /. A Afatllcmaticia1l1 A1iscellan ). and then forget about the parameterization of e. A Mathematician's Apolog)' though the vector field is COO on all of M. Beauty is the first test: there is no permanent place in the world for ugly mathematics. find an integral curve c of X with initial condition or e(O) = p. . which include such things as the "afunction"). We call /. by defining N to be {c(I)}.q for all q in some open set around p. xn (q) an } = . sup pose that for each p E M we have a I dimensional subspace /. the set {q E U : x2 (q) = a 2. U) such that for each fixed set of numbers a 2 . E. we can always find all integral manifold N of a Coo distribution /. We will now vary our question a little. . H. [Pioneer work is clumsy] A. is called a I dimensional distribution (this kind of distribution has nothing whatsoever to do with the distributions of analysis.
It happens that the integral manif olds ofthesC' two distributions are also the integral curves for certain vector fields. This is still a local result: but because we are dealing with submanifolds. which locally look like  (rather than like or something even more complicated). on U.1 1 ���l il l l l l l 011 submanif old pictured in Chapter 2.180 Cha pter 6 is an integral manifold of !..hkh has olle compact connected integral manif old. and that these are the only integral manifolds in U. and all other integral man ifolds nOlicompact. there is a distribution the torus . . For example. On the other halld.. but on the . we can join overlapping integral submanifolcls together. rather than curves with a particular parameterization. The entire manif old M can be written as a disjoint union of connected integral submanifolds of !.. there is a distribution 011 lhe torus whose integral manifolds all look like the dense I dimensional = 1 11 .
a).p c= . b(x � . A (kdimensional) submanifold of is called an integral manifold of /. . the results will look very different. ( q).(a. . a Coo distribution if it is possible to choose Coo vector fields X" . s. integral manif olds do llot exist. Thus /. consider' the 2dimensional distribution D.br)p. Xdq) are a basis for /.P l where D. . As the simplest example. in a neighborhood of each point p . C For any P there is a neighborhood U is a k dimensional subspace of and k vector fields X" . We are leaving out the details involved in fitting together these local integral manif olds because we will eventually do this over again in the higher dimen sional casco For the.Integral Alanifolds 181 ]\1obiu5 strip there is a distribution which is spanned by a vector field only locally. .q. M E M N M i: N M + p Mp is the inclusion map. A kdimensionaI distribution on is a function p 1+ D. for each q U. . in �3 for which /. Thus /. . . S E } (r.c) is spanned by E pEN i*(Np) = /.moment we \'\'ill investigate higher dimensional cases only locally.p = If we identify T� 3 with � 3 x � J . . We call /. Xk such that X. . Xq with this property. then /.p where Mp.p consists of all may be pictured as the plane with the equation z {r �Ip + s �lp + br� lp : r. . In general. even locally. if for every we have Although the definitions given so f all look the same as the I dimensional ar casC'.p = /.b. .
with all tangent vectors having slope Yo. This will be a curve in the plane { (x.(a. z)} through (0. Suppose there were an integral manif old N of /. you can probably see that it has no integral manifolds. c) isjust parallel to the one through (a. yo. or the intersection N n { (x. b. :)plane through 0 whose tangent vectors would have to lie in the intersection of /. yo.182 Chapter 6 The figure below shows /. � � � � � � � � � � If you can picture this distribution. The plane /. . 0) contains vectors with third component nonzero. so Y lllust be the yaxis.z) and the (y. b. The only such vectors have third component 0.. f each fixed Yo. its tangent space at ( 1 . b. yox)}. yO. z)plane. 0). Yo.c) through (a.p for points p = (a.y. a proof can be given as follows. 0. y. 0).b. But this submanifold c10es not . with 0 E N.(O .. For example. s o i t must b e the linc { (x.'ork. The intel'section of N and {( O. Our intcgral manifold would have to look like the following picture. O). : ) } .z)} would be a curve y in the (y. Now consider.
the submanifold is given locally as the graph of a function: N = ((x. aa g(a. Since /.s E IR .. + a.y.(a.b. b. consider the somewhat more general case where D. b.b). y)}. has an integral manifold N through each point. 0). As in the first example. c) will be parallel to the one through (a.p = �{ + s � p + [rf(a.Ip . b)..z) z = a(x.b) . aay ay • = (a. ay = g.b)) is spanned by a aa a a.(a.b) = a.a(a.c = f(a.b) (y .b)] Ip : r.a) +g(a. S o we need t o find a function a: 1R2 > IR with aa � . y) plane.b. since f and g depend only on a and b. ax These tangent vectors are in /.f.Integral Manifolds 1 83 p /. (a.p is the plane with the equation 0 . /. (a. b) if ' p p .c) = D.b.. geometrically.b) = ay (a. : p Now the tangent space at aa f(a. To see in greater detail what is happening here. is r· l � } We now ask when the distribution /.b).(a. the plane /.p if and only if I I aa a a .p is never perpendicular to the (x.Ip + .c) through (a.b) + sg(a.b)(x .
O. defined in a PROOF We first define a(x. 0) = l(x. I(a . If I. = b. = 0' so this necessary condition is not satisfied. '!§. PROPOSITION. aa (x. then there is a function a(O.1 84 Clzapter 6 It is wellknown that this is not always possible. and ':0 E neighborhood of 0 E 1R2 . 0) = :0 (I) a.O) = :0 + namely. By using the equality of mixed partial derivatives. we find a necessary condition on f and g: In our previous example. O). It is also wellknown that the neces "JI'Y condition (**) is sl!lJirienl fOl' the existence of the function a satisfying (*) in a neighborhood of any point. 0) so that a(O.. such that �. we define foX 1(I. a(x. b ) g(a. and u._0' . g : 1R2 + IR satisfy in a neighborhood of 0. I 0'. al = ay ax I. O) dl. b) = 0.O) = :0 �=I ax aa ay = g.
aajay = g. I) dl 1 y This is 0 for ). 0) dl + 1(" g(x. ax y. y) = g(x. g : /'. y) so that (2) aa a:. = 0 by (I) . s �} . 0 0 laY g(x. consider.Integral Manifolds Then. y. = �3 + �. namely. To prove that it equals 0 for all shO\\' that its derivative is o. we define a(x. and always provide us with an a satisf y� ing (2). we define a(x. z) : z = a(x. We claim that if (**) holds. y). y)} . Suppose that N = : E { (x. y) . (x. I) dl. 0) + = Zo This construction does 1)ot use (**). for each 185 x. for each fixed x. y).y) = a(x. we just have to VVc are now ready to look at essentially the most general case of a 2dimen sional distribution in �3 : where I.p {r a�{ + s � Ip + [rl(p) + sg(p)] � Ip r. But its derivative at y is f+ aa (x. To prove this. the function + r 1(1. then also aajax = f../(x.
b» . y y • These tangent vectors are in /.a(a . since it still involves the unknown function a. b). I(a. b) a: p . b» . b» . a (a.b. In order to obtain necessary conditions for the existence of such a function a) we again use the equality of mixed partial derivatives. (a. Thus (*) and the chain' rule imply that al a2a al aa . b» + ilz (a. b. by aa a a a. b.. b. b» + . b) ay aY az ay a x II ag a 2a ag aa (a. Thus we obtain finally the necessary condition .b» + ilz (a. a . b» + ilz (a. b. once again.a(a . b. a(a . b. b) = . but we can substitute from (*) to obtain al al . b» · . b) = � (a. (a.(a.b) p aa a a a p + a (a. b . b. b.(a. b). b. through . b» = = aa � (a .ver ' point. The tangent space of N at p = (a.(a . b . b» g(a. a(a. which means that for each ) pair (a . a (a.a(a. b» is spanned. a(a. . ' + � p � (a. aa a. b» g(a. a(a. a(a. b) these equations must hold no matter what a(a. b» ag ag = � (a.(a. b). Now we are looking for conditions which will be satisfied by I and g when there is an integral manifold of /.a(a.186 Chapter 6 is an integral manifold of /.a(a. a x ay This condition is not very useful. b .a(a .p if and only if I I I I I(a. a (a . b) is.
In the following theorem. . J. . defined in a neighborhood W of ° in JRm. there is no need to restrict ourselves to equations f a single or function defined on JR2 . a partial diff erential equation for a function from �m to JR"). . there is at most one function a : W + V . where U is a neighborhood of ° E JRm.. O. .a(l. : U x V + JR" be Coo functions. we can treat a system of partial differential equations for n functions on �m (i. Ill. defined on WI and W2 . . 0. so for a function f : �m X �n + �k we use al aI' al ax' for for D. . . for i = 1. . . . rom the proof o f existence. m . 0) = 11 (1. . Necessity of PROOF Uniqueness will b e obvious f the conditions (**) is left to the reader as a simple exercise. the necessary condition again turns out to be suf ficient.e. Let U x V C JRm x JR" be open. satisfying a(O) = x aa (I. . In fact. j = 1. . .a(l» for ali I E W. . J. and we will concern ourselves with proving existence if these conditions do hold. such a function exists (and is automatically COO) in some neighborhood W if and only if there is a neighborhood of (0. . and let j. . Dm+. .l111egral Manifolds 187 In this more general case. 0». agree on the component of WI n W2 which contains 0. . 0) = x aa atT (l. al j (I) = Jj (More precisely. Then for every x E 11. . . . with a different twist at the end. x) E U V on which x i. We first want to define a(l. 0. The proofwill be like that of Proposition 0. 0. . . . 0) so that (I) a(O. .0. . 1.) Moreover. THEOREM. . any two such functions a l and a2 . . we will use I to denote points in JRm and x for points in �n . 0.
O. .al 2 . . . . . 0. I. 0. . . . . 0. f3dl» . defined for III III a(l.1.. . . ./. This has a unique solution f sufficiently small I. . (2) 0) = h(l ' . then for 11'1 < " the solutions of the equations f f32 with or the initial conditions f32 (0) = a(l ' . 0. f32 (1» .0. . 1. 0. O. axk a/2 ' k= ' (4) . We now derive an equation for g'(I).0. . 0» . (3) O = g(I) = iJiT (I . O) = x aa (1 . . 1. 0) f3/(I) = h (t ' .O. . . We then define fh (O) = x f3. 0». O. . 0) = f32 (1) Then a(O.0. . . " . . . . . . O. I. .. (t < < < " < '2. (I) = al 2 al' . . .1. . 0. . 1. 1. < a(l ' . I. .O. . O.I.a(1 . .. 0. . .L. . . . . . . 0.a(l . sufficiently small. . I. We have a2a a/. . ' (I) = jj (I. . 0. . . At this point the reader must or < < " . . . . . < " . . . . . all expressions involv ing a are to be evaluated at (t' . 0. for ali I with III aa . 0. . . and verif the following assertion: If we choose " ). .O. In the f ollowing.188 Cha pter 6 To do this. O) . 0) and all expressions involving Ii are to be evaluated at (I'. Define This equation has a unique solution. aak g. . 0) = f3dt) Then (I) holds f III or Now for each fixed I' with 11'1 '" consider the equation f32 (0) = a (I ' . Note first that g(O) = 0 by (I). . 0» a1 2 11 ' 1 ' " III '2· We claim that f each fixcd I' with 11 ' 1 or we also have. � a/. .0) will each be defined for III < '2 for some 82 > o. . . .a(I ' . . we consider the ordinary differential equation refer back to Theorem 52. . . . .
aft2 al by (2) by (2) again _ _ _ by definition. The remaining parts of our investigation. (3) Now equation is a diff erential equation with a unique solution for each initial condition. Theorem I essentially solves for us the problem of deciding which distributions have integral manifolds. Our investigation of the problem so far illustrates one basic fact about theorems in differential geometry: Many of the fundamental theorems of diff rential geometry fall into e one of two classes.(I) � aft k L. The second kind of theorem justifies this terminology.k=1. •:.. illustrates an even more important fact about the theorems of differential geometry: There al'e always incredibly concise and elegant ways to state the in tegrability conditions. a distribution with integral submanifolds through every point) then certain othel' conditions hold.g. and prove their sufficiency. without ever even mentioning partial derivatives. So (3) is true. given by (4).. by showing that the " integrability conditions" afe suffi cient for recovering the nice situation. al . en ) and satisfies t*) .Integral Manifolds and thus 1 89 (5) g . The first kind of theorem says that if one has a certain nice situation (e. It is a simple exercise to continue the definition of a until it is eventually defined on (e l .. e l ) x ' " x (en . these Con ditions are obtained by setting mixed partials equal. axk h ah + t ah aak aft t aft f: k = al l k=1 axk al l al 2 k=l axk 2 a a = h + t h [g k (l) + ft k J al l k= axk 1 aft _ t aft f: k al 2 k=1 axk 2 t ahk gk (l) = k=1 ax = ail a ( aa2 ) . in which we will essentially begin anew. is clearly g (t) = 0 for alI I . (5) . and are called "integrability conditions". The solution with initial condition g (O) = 0.
then there is a unique Coo vector field X on PROOF. . If g : N + JR is a Coo function. 2. . y X= n = 1 . IJ = a�. are Coo (problem 3) . we say that X and Y are Irelated if l p (Xp ) = Y p) for . and Y. . the latter condition is fulfilled: 2. then Y p) (g) = l Xp (g) . PROPOSITION. then X and Y are Irelated. .p f( = Xp (g o I). n . (Yg) 0 I = XU o g) . nor must a given vector field Y be Irelated to any vector field on M. . This is easily seen to imply that I p. V) around I(p) E N such that ' t J' 0 l o X. . O. fOI' i [Y" Y21 are Irelated.. PROPOSITION. axl Y= f( p) . Xp . . respectively. Clearly we must define Xp to be the unique element of Mp with Y p) = I . This ( a�. f(p) E I (Mp ). and X and Y are Coo vector fields on M and N. then where The most important property of Irelatedness for us is the following: 3. M which is Irelated to Y. .. •:. 0 I = I. a n ) = (a . ' L IY�. I a L a' a. . U) around P E M and (y.1 90 Chapter 6 LOCAL THEORY If I : M + N is a Coo function. To prove that X is Coo.= 1 implics that the functions I. Thus if where (i are Coo functions. . Let I : M + N be a Coo function such that I is an immersion. . . i f this i s true for all Coo functions g : N + JR. In One case. If Y is a Coo vector field on N with Y p. . . we use Theorem 210(2): there are p f( coordinate systems ( x . are Irelated. then [X[ ' X21 and . Of course. a n .= 1 . a given vector field X may not be Irelated to any vector field Y.I (a . 0). If X. so Conversely. f( each P E M. a.
(g f) i = {[Y" Y2]g} f = {Y.X2 (X.� + f� _ g af ) � . then [X. p In other words. (Y2g)} f .. Using the f ormula on page 1 56. Y]p E /. ( [Y2 g] f) . and Y and f are irelated. then for all p E N there are unique Xp . Y] are irelated. Thus i. and Proposition 3 then shows that [X.p = { I' f lp + s hip + [rf(p) +sg(p)] �Ip : I'. f]p E Np . in 1R 3 given by [X.Xp.{Y2 (Ytg)} f = X. Consequently. If X and Y are two vector fields which belong to /. ''''le will say that a vector field X belongs to /. Y] also belongs (I) So 0 0 0 0 0 0 If g: N + IR i s Coo. to have an integral manifold through every point.g) f = X. •.. Y]p.p for all p. through every point p.IntegraL JvlanifoLds PROOF.X2 ([Ytg] f) by (I). fp E Np such that Xp = i. respectively = XdX. if there is an integral manifold of /. ax ay az az az . we see that This belongs to D. 191 (Y. 0 0 0 .p .f . and i : N M is the inclusion map. 2. which is precisely the condition for /. S E IR } . Yp = i. For a moment look back at the distribution /. 0 + to /. Proposition 2 shows that X and f are Coo vector fields on N. (g f» by (I) = [X" X2](g f) : Now consider a k dimensional distribution Do. Suppose that N is an integral manifold of /. if Xp E /. /. f] and [X. this therefore shows that [X.(g f» . X and X are i related..[X..g and Ytg.Y] = (� . with g replaced by Y. The vector fields X = � + f� ax az a Y = a:ay +gazbelong to /. then 1 . only when the expression in parentheses is 0. f] p = [X. Here [X.
e). e) such that for each a k + I .. PROPOSITION.... [f. in a neighborhood V of p. [jX.fiXi i= f or Coo functions C. is called integrable if belongs to /'. x (x. x(p) = 0 ( . . ) a n with all fa i r < e. Y] f(Xg)Y g(Yf)X.. Any connected integral manifold o f /'. Theorem I can be derived f rom it (Problem 7). .Xi. •. /'. . it obviously suffices to treat each separately.. restricted t o V i s contained i n one of these sets. l Iich is spanned by the Conversely. the set {q E V : xk + l (q) = a k + l . THEOREM ([HE FROBENIUS INTEGRABILITY THEOREM. then /'. . It is equivalent to Theorem I.. . . FIRST VERSION).Xj] = L C/. If and belong to /'.Y] do : + . . Xj] is a linear combination k [Xi . . since E /'. x IV) = ( e. ..gY] = fg[X..J ' Such functions clearly exist if /'. we can clearly write PROOF Y To prove belongs to /'. be a Coo integrable kdimensional distribution on M.. Xj ]q q X k X = L1 . .. If is integrable on V if and only if each X" . x n (q) = a n } is an integral manifold of /'. . X. [Xi . . if X. [Xi. . Let /'. Xk span /'. Y] [jX.. Y and [X. 5. Y] X and Y 4. whenever belong to /'.1 92 Chapter 6 In general.e. X.gY] belongs to /'. in or fact. For every p E M there is a coordinate system V) with x .. Since we have clearly [X.gj Xj] We arc now ready f the main theorem.. But the proof is quite diff el·ent..(q).. suppose such f unctions exist. is integrable. This condition can be checked fairly easily: [X...
. I. there is a coordinate system x such that a X. . of Nq we h ave d(xm i)(Xq) = Xq(xm i) = i. . with p = O. k . So [X" X ] = o. .[X" Xj ]q = [ aaI" aaIj ] .Xq(xm) = 0. .. (on a neighborhood of 0 E IRn ) and a/aI (on IRk ) are 1[related. By continuity. We can clearly assume that we are in �n . . . ][* is oneone on f or near O. which is spanned by the a/axj l q f j = I . 1[.q q 1[. Moreover. ' Then the vector fields X. . . and 1[.X.0 C �no is spanned by Let IRn + IRk be projection onto the first k f actors. since i. k . . By Proposition 3. But. . since their tangent spaces are spanned by the a/ax' = X.6.. . is oneone on I'>q. . = ax' i = I. : 0. The sets {q E U : Xk + l (q) = ak + I . .Integral Manifolds 1 93 PROOF. . we can assume that . Then 1'>0 + IRko is an isomorphism. 1[. 0 . k .Xq E I'>q. which implies that xm i is constant on the connected mani f old N : so that + 0 0 0 m S. . j j By Theorem 514. •.(q) i = I. . 0 . . If N is a connected in tegral manifold of I'> restricted to U. So near we can choose unique 1[: D. Thus or d(xl1l i) = 0.(q) = 0. (q) = aaI. [X" X ]q E I'>q by assumption. . . for i = I. x" (q) = a n } are clearly integral man ifolds of 1'>. with inclusion map For any tangent vector Xq i: N U. consider d(xm i) for k + I s. 11. k.
Vi ) satisfying the conditions of Theorem 5. let us call each set (x. . 0) x . V). we see that we can cover M by a sequence of coordinate systems (Xi . and if around every point P E M there is a coordinate system (x.. we introduce the following terminology. with xIV) = (0. . .. . It is possible for a single slice S of Vi to intersect Uj in more than one slice of Vj' as shown below. x (0. . Let !. . 6. 0). {q E V : xk+l (q) ak +I . If M is a Coo manifold.1 94 Chapter 6 GLOBAL THEORY In order to express the global results succinctly. PROOF. Then M is foliated by an integral manifold of !.xn (q) an } = = NnV are the sets of the form Each component of N is called a folium or leaf of the foliation N. . THEOREM. ( () Using Theorem 12. For such a coordinate system V). a (usually disconnected) k dimensional submani fold N of M is called a foliation of M if every point of M is in (some com ponent of) N.). xn(q) an } = = a slice of U. (each component is called a maximal integral manifold of !. But S n Vj has at most countably many components. such that the components of {q E V : Xk+l (q) ak+I . . . Notice that two distinCI components of N nV might belong to the same leaf of the foliation. be a Coo kdimensional integrable distribution 011 M.
. [If we are allowing nonmetrizable manifolds. Appendix A describes a nonparacompact manifold which is foliated by a lowerdimensional connected submanifold. Vo) with P E Vo. . .. and only countably many such sequences. . Given P E M. sO S nVj is contained in at most countably many slices of Uj . i 1 . . the discussion to follow will not be validin fact.1 . since we do not have to find a countable number of coordinate systems for each leaf. 1 . In this case. . however. . •:. Since there are at most countably many such sequences of slices for each se quence io. A slice S of some V. . or totally disjoint from. For q #. and can merely describe the topology of the foliation as the smallest one which makes each slice an open set. . . the first union. . il = i and corresponding slices with a = 0. . choose a coordinate system (xo. . M is foliated by the disjoint union of all such submanifolds. this di'!ioint union is clearly an integral manifold of /'.p. there are at most countably many slices joined to p. we see that the union of all such slices is a submanifold of M.] . the corresponding union is either equal to. Consequently.InlegmL ManifoLds 1 95 and each component is contained in a single slice of Vj by Theorem 5. Using Problem 31. and let So be the slice of Vo containing p. i/. will be called joined to p if there is a sequence 0 = io . the proof is even easier.
. This allows us to apply a proposition from Chapter 2. Let be a Coo manifold.. at most cOUil/ablJ' man slices can belong to the same f y olium. choose a coordinate system f(p) such that the slices MI _ (x.1 96 Chapter 6 is a coordinate system of the sort considered in the proof Notice that if of the theorem. this folium would contain an uncountable disjoint family of open sets. . 7. Let be another Coo manifold and f: Then f is Coo considered as a a Coo function with C + map into as a map into Given P E P. U) U However. . it suffices to show that f is continuous M P MI .. (x. P M MI .. otherwise CiV ' . . so f takes f(P) MI . THEOREM. then infinitely many slices of may belong to the same f olium. . . and a f olium of the f olia tion determined by some distribution /'.xn(q) = an} are integral manifolds of /'.U) around {q E U : Xk+l (q) = ak +I . Now f is continuous as a map into M. According to Proposition 21 1. .. PROOF.
For k + I :s i :s n. •:. contradicting the fact that f( W) e M. (f(p'» = a. and x . This makes it clear that f is continuous as a map into MI . by continuity. Consequently. f( W) is contained in the single slice of U which contains p. x . if we had x . for all p' E W. . 0 f would take on all values between a . for any p' E W. .a. (f(p'» #.IntegraL ManifoLds 1 97 some neighborhood W of p into U. In other words. This would mean that f(W) contained points of un countably many slices. then x . we can choose W to be connected. (f(p'».
dt. If i : EI + E is the inclusion map. 3. . 1 . 0 f = f3.} actually are Coo .. M. and show that a kdimensional distribution is Coo if and only ifit is a Coo subbundle. (a) In the proof of Proposition 2. a Coo. Show that there is a coordinate system i s spanned by aj ax l . then the functions 4. show that the functions C. = 2. define a Coo subbundle. IB) is a bundle map.1 98 PROBLEMS Chapter 6 1 1[ : E' + B a 1. • . aj a xd" etc. show that 6'1 suffi (b) Complete the proof of Proposition 2 by showing that if so that with a. . 5. /'. . . . such that /'. 1 .. . we say that �' is a subbundle of � if (i. (a) In the proof of Theorem I. (x. • . and �I k plane bundle such that E' c E. x = L f3 ' ax. (a) Let � = 1[ : E + B be an nplane bundle. and IB : B + B the identity map. . be integrable distributions on Suppose that for each P E M. (b) Supply the proof of the uniqueness part of the theorem. .. . of dimensions dl . check the assertion about choosing ciently small. U) around each point. ' 1=1 f3. 1. Show that a kdimensional distribution on M is just a subbundle of TM.. Let /'. (b) For the case of Coo bundles � and �' over a Coo manifold M. In the proof of Proposition 4. are n .
by considering the distribution !. 7. al j ajj al i aft al j ' with the integrability conditions we nevertheless work in �m x �n ) rather than �m. This problem outlines another method of proving Theorem I. defined by Notice that even when the jj do not depend on x. x) . I)) = . VI) = f3 (UV. (Show that both functions satisfy the same differential equation as functions of u. This is connected with the classical technique of "introducing new independent variables". show af3 j iU (U. One has to check that one 0 can be picked which works for all l E W. VI). f3(U. f3(u. since the equation depends on the "parameter" I E JRm). 1) We know that we can solve such equations (we need Problem 55. jj (ul. by reducing x W + V. in JRm x JR" (with coordinates I. (b) Show that = x. l) al j = V· af3 ( I . (a) If we want D!(UI) = f3(u. = j m Ll 11 . with the same initial condition. I) for some function f3 : [0. al j .) By shrinking W. I). Prove Theorem I from Theorem 5. we can consequently assume that e = 1 . A similar technique will be very important in Chapter II.integraL ManifoLds 1 99 6. 0) that f3 must satisfy the equation 7. the partial differential equations to ordinary equations along lines through the origin. I) f3(0. (c) Conclude that af3 (v. so that the equations are of the form aD! (I) = jj (l).
y)) = . y). show that a satisfies the desired equation. y. and conclude that the (in which f denotes the complex derivative) has a solution in a neighborhood of =0. Noting that a(vl) = f3(V.Y �2 in a neighborhood of 0 E differential equation = = � v(x. ax aa ' a. f3(v. Let I: <C x <C + <C be complex analytic.200 Chapter 6 (d) Use the integrability condition on 1 to show that af3 (v. 2. I ) al j and v · h ( VI. y" X" y" then 1 = u + iv satisfies the CauchyRiemann equations Use Theorem I to prove that we can solve the equations au aXi au aYi av aYi av aXi i = 1 . If we denote the coordinate functions in <C x <C by ZI . (e) Define a(l) = f3 ( I . a' (x. I)) erential equation. I).ay <C aa ' u(x. 8. y ) . 2. y . Use (c) to conclude that satisfy the same diff the two functions are equal. I). y)) = ay (or of any point </>'(z) = I(z. a' (x. This problem is for those who know something about complex analysis. a ' (x. = XI . with any given initial condition 4>(zo) = Woo . a l (x. as functions of v.</>(z)) 20 E q.
The most familiar example of an alternating T is the determinant function det E T"(�n). so I1 I ( V) has dimension /1 . At the moment it is not clear what the dimension of I1 k (V) equals for k > I. ) Vj. . . Vi • • . . . Vi. but we will be concerned if Vi = Vj (i #. . [This is not true in the special case of a vector space over a field where 1 + 1 = 0. but one case is \vellknown. It is also convenient to set l1o(V) = :ro( V ) = JR. " Vj " . Uk ) = T(VI " .onsidered as a function of the n rows of a matrix we shall soon see that this function is. . • . " Vj" " . . Notice that I1 I ( V ) = :r1 (V) = V* . f\1osl discussions of the determinant begin by showing that of any two alternating nlinear functions on �n. Vi. Vj" " . one is a multiple of the 201 . Vi. . Vj" " . if f: V + W is a linear transformation. . Uk ) = . the discussion of which requires some e turn OUf attention once more to tensor fields. then T is also alternating. skewsymmetry is the same as symmetry. " . An element T E :r k (V) is called alternating if Wwith a special kind of tensor field. vk ! + 0. Vi" " . " Vi" " . " Vi" " . then f* : :r k (W) + :r k (V) preserves these subspaces f* : I1 k (W) + I1 k (V). Vi. • . . Vi + Vj • . ' " vk ! + T(vj" . Uk ). • . Vj .jl . the most general alternating function. . .CHAPTER 7 DIFFERENTIAL FORMS morc algebraic prclimi. in this case. Uk . • . in a certain sense. vk ! = 0 + T ( Vj " " . Vi + Vj • • • . Therefore. . if T is skewsymmetric.T(vJ • . vk ! = 0 If T is alternating� then for any VI . . . It is clear that I1 k (V) c :rk ( V) is a subspace of :r k (V). " vk ! + T(VI " " . . Vi" " . . .naries. we have 0 = T(v) • . . T(VI " " . Moreover. c. vk ! + T ( VI " . Of course. . " " Vj" . . and the condition of being alternating is the stronger one. . . . " Vj" . . . T is skewsymmetric: T(v) . Vj • . Let V be an /1dimensional vector space over JR. " vk ! + T ( V I " " .] We will denote by I1 k (V) the set of all alternating T E :r k ( V ) .
. of course. in other words. . then the first element on the right is 1lot necessarily that v whose index is a ( 1 ) . k}. . . ) is to rename things: V3 = WI ) V2 = Wz. which is a special case of the definition to follow. is the a ( l )" of the v's on the left side.e. . vk! = k! sgn a . . \'\. vd ) = a · (wl . vd = ( Va(I ). T 0 a. This definition has a builtin confusion. . k.1 if a is odd. O n the right side. . Then one proves dim 11"(IR") = I by actually constructing the nonzero function det (it follows. . Walk) ) = ( Vp(a(l)) ) . . . . . vd. . T(va(l )" ' " Va(k) ). ' . ' . the first element.202 Chapter 7 other. . we compute by setting so that Vp(l) = W I • . . (p . Vp(k) = Wk) a · (p . . . ueS" L sgn a . ' " Va(k) ) . The construction of det is usually by a messy: explicit formula. . Vp(a(k)) ) since Wet = vp(a)· Thus a . . if these v's have indices running in some order other than 1.here sgn a is + I if a is an even permutation and . . If (v" . . . VI = W3 • . Thus warned. vd) = ( pa ) . Vk ) is a k tuple (of any objects) we set a · (VI . . . . . . . .. . VI. The simplest way to figure out something like a ' (V3 ) V2 . (vJ . . . . dim 11" (IR") :S I. I Alt T(v" . . . . that dim 11" ( V ) = I if V is any n dimensional vector space). wd = (Wa(I) . Now for any T E :r k ( V ) we define the "alternation of T" I Alt T = k! aeSk L i. (VI . . . . . Let Sk denote the set of all permutations of { l. . for example. (v" . . an element a E Sk is a function i r+ a (i). . . . .
The funny coefficient is not essential. f\1oreover. associativity of /\ is proved in the following way. the wedge product of w and ry. (I) If S E :r k (V) and T E :rl ( V) and Alt(S) = 0. then Alt w = w. 7879 of Calculus on Manifolds) . In particular. PROOF. /\ ry + w. it is easy to see that (3) /\ is "anticommutative": w /\ ry = ( _ I) klry /\ w. ) /\ ry = w. e E I1m ( v ) . (3) If T E :rk ( V). 2. + w. ry E 111 ( V). by ( w /\ ry = k!/! Alt (w ® ry) . k ! l!m! . then Alt(T) E I1 k ( V ) . •:. + w /\ ry. aw /\ ry = w /\ ary = a(w /\ ry) k + /)! (2) I* (w /\ ry) = I*w /\ f* ry. (2) Alt(Alt (w ® ry) ® e ) = Alt(w ® ry ® e) = Alt(w ® Alt( ry ® e)). We now define. an element w/\ ry E I1 k +1 ( V ) . /\ ry w /\ (ry. then (w /\ 1]) /\ e = w /\ (ry /\ e) = (k + 1 + 111 ) ! Alt(w ® ry ® e). (2) I f w E 11k( V ) . (I) If T E :rk ( V). as we shall soon see.) = W /\ ry. then Alt(S ® T) = Alt(T ® S) = o. if k is odd then w /\ w = o. (3) If w E I1 k ( V). It is clear that (I) /\ is bilinear: (w. THEOREM. but it makes some things work out more nicely. 203 PROPOSITION. Finally. Left to the reader (or see pp. + ry. for w E 11 k (V) and ry E 111 ( V ) .Dijje"n liaI Forms I. then Alt(Alt(T)) = Alt(T).
(a . a contradiction. the sum over each being o. k + I fixed.ry ® e) = Alt( ry ® e) . (3) We have a'eG L sgn a' . \!\le can then continue in this .Alt(w ® ry ® e). ( Vb ."lay: breaking Sk+! up into disjoint subsets. S(Va(I ) . . . • . (vl . . Let aoG = {aoa' : a' E G). .I E G. (v l . .ry ® eJ ) = AIt(w ® Alt( ry ® e» . (2) Clearly Alt(Alt(1] ® e) . .204 Chapter 7 PROOF (k + /)! AIt(S ® T)(Vl . . Vk+I» ) . . . Notice that G n aoG = 0. . so (I) implies thaI 0 = Alt(w ® [Alt( ry ® e) . Vk+I » L sgn a . then a = aoa' for some a' E G. . Then L sgn a . . . ( S ® T) .Alt(1] ® e) = 0. . . Suppose now that ao <t G. the other equality is proved similarly. . Va'(k) ] . . . Va(k) . T(Va(k+l) . for if a E G n aoG. = ry ® (k + /)!m ! k!/ ! . • . . . . " " Va(k) ' T(Va(k+I). . T(Vk+l > " " Vk+I ) = sgn ao ' ''''le have just gjwwn that this is 0 (since ao . by (*) . so ao = a(a. Va(k+l) aeG = = o. . Vk +l) is just some other (k + I )tuple of vectors). (w /\ ry ) /\ e = The othel' equality is proved similarly. Vk+/» ueuuG [ a'eSk L sgn a ' · S(va'(1) . (S ® T)(a' .. .). . . The relation Alt(T ® S) = 0 is proved similarly. ( S ® T)(a . . S ( Va(I). . :. Vk+l) = = aeSk+l aeSk+l (1) We have L sgn a . (k + l + m ) ! Alt((w ) ® e) /\ ry (k + / ) ! m ! (k + 1 + m)! (k + /)! Alt(w ® e). . (ao ' (VI . . Then L sgn a . " " Va(k+l) ' Now let G C Sk+l consist of all a which leave k + I. .
then 1\ could be defined by w 1\ ry = 1 k!l! Alt(w @ ry).] The factor (k + I)!/ k ! / ! has been inserted into the definition of 1\ for the follO\ving rcason. Vk+l) occurs k ! / ! times (since w and ry are alternating). L ail . . ( I + . 4>il ® ' " . " . (So i f V I . @ </>n ) I!" . .iJ. This makes sense. and 4>1 . iJ. because each term in the sum Alt( w @ ry)(VI . . !. but without the factor 1/ k!. 1\ </>n = = In particular. ."2 k ( V) C :r k ( V) . . > 11. Vn ) = I . . .Dijjcrenliul j'lmlH 205 Notice that (2) just states that /\ is associative even if we had omitted the factor (k + /)!/ k! I! in the definition. .. . then </> 1 1\ . . The set of all 1 ::: i) < " is a basis for !. · + I)! Alt(</>I @ " . ."2 k ( V) can now be described. [If we had defined Alt just like Alt . · I\ </>n )(VI . . . . 1\ </>n = det. we would not have Alt(Al t T) = Alt T. THEOREM. . which therefore has dimension () 11 k : .) A basis for !."2k ( V) = {OJ for k PROOF. . . . the factor 1/ k! in the definition of Alt is essentialwithout it. . . Vn i s the standard basis for IRn . If w E !.. If V I . 4>n is the dual basis. we can write W = iJ . . < ik ::: n = 11! k! (I1 .) ® 4JiJ.. On the other hand.."2 k ( V). . 3. . . VII is a basis of V. then </>1 1\ " . and the first equation in the proof of (2) would fail. ellen over a field a ffinite characteristic. I! aeS" L sgn a · (</>I @ · · @ </>n ) o a · . and 1/ k! I! can be interpreted as meaning that these k ! / ! terms are replaced by just one.k) ! ' (In particular. . . . (</>I I\ .
. . . .ik 4>iI !\ · · · /\ <PiJ.ik Alt(4)i. ® 4>ik ) · Each Alt(4)i. . . vk! f+ determinant of a k One more simple theorem is in ordel� before we proceed to apply our con struction to manifolds. . /\ Wk is a basis element of !:) k ( V) . . < . . . . . /\ · · · /\ 4>ik . 4>n satisfy 4>i = Wi for I :s i :s k . /\ 4>h for some k }. . . .h L ai. .·. < }k span !:) ( V). . ® .. . and let 4>1 denote 4>i . Wk E !:) ' ( V). . so the elements 4>h /\ . . = o. ik!. If WI " 4. . ® . . . Vn • . . . there is a basis V I " of V such that the dual basis vectors 4>" . . so it is not o. it is convenient to let L al 4>l . then W' . . ® 4>ik ) is either 0 or = ± ( I / k!) 4>h /\ .206 Chapter 7 So W = Alt(w) = iI . 4>k. /\ . ) PROOF. If 0= i l < " '<ik L ai! . Uk • . . /\ . . if then uniquely expressible as k (i" . I j denote a typical "multiindex" Notice that Theorem 3 implies that every W E S'"2 k ( �n ) is a linear combination of the functions (v" . /\ 4>h for }. If W" . then applying both sides to (Vii " ' " Vik ) gives ail . . COROLLARY. ·. . . On the other hand.. x k minor of (:�). . . . .:. . . Then w. . .i}. . < . . . Wk are linearly inde pendent if and only if w. /\ Wk #. Wk are linearly independent. Then every element of !:) ( V ) is To abbreviate formulas.o. . . . . . . < }k. .
(p) w(p) !.talnVl). /w /\ ry = w /\ /ry = /(w /\ ry) w /\ ry = ( . . > O. Remember that covariant tensors actually map contravariantly: If /: M + N is Coo. we are ready to apply it to vector bundles. ann») = W(f>lIVl. . We can also define WI + w."2 k +/ (�) by (w /\ = /\ E In particular.1 A section W of !."2 k (�) by replacing each fibre 1[.ant!. If V is ndimensional and 0 #. . j=1 i PROOF. ."2 n ( JRn ). . If � = J[ : E ?. . The following properties of kforms are obvious from the corresponding properties for !."2 1 (�). ."2 n ( v) . . so ry = c · det for some c E JR. unique orientation [vt . . Define ry E :r n ( JR ) by n . . /\ ry W /\ (ryl + ry. A I form is just a covariant vector field. Vn) With our new algebraic construction at hand. v ] = /L IL for V such that n if and only if W(VI. Since !. THEOREM. and W is a kform on N. then there is a ry ( (u l I .en) = w(Vt. sections of n k ( TM)."2n ( v) . . " ."2 k (�) with is a function with E for each E B. . COROLLARY. . .ill E !. j=1 j=1 Then clearly ry E !. then we can define a section W /\ ry of !.) = w /\ ryl + W /\ ry. . . ."2k (1[1 (p» !. . . · .DifJemiliaL Forms 5. . . ."2 k (TM) can obviously be made into a Coo vector bundle. . we obtain a new bundle !. . and c = ry( et . •.B is a vector bundle. and let n Wi = L CXjiVj = l . then /* w is a kfol'ln on M. we can speak of Coo forms. If ry is a section of !. and w /\ ry.1 (p» .) /\ ry = WI /\ I/ + W. n. . " (Unl. . . ."2 k +1 (1[ 1 (p» . Vn be a basis for V. Let 207 Then VI . . all forms will be understood to be Coo forms unless the contrary is explicitly stated. let W E !. ."2 k (1[. . !. are called kforms on M."2 k ( V): (WI + w. p ry)(p) w(p) ry(p) . which are just alternating covariant tensor fields of order k. . vn ) : 6. .I/Iry /\ w /*(w /\ ry) = /*w /\ /*ry.
. . C(Y. .. . . w = = (il. If (x. then the dx. a. •:. f\. . L The problem of finding the relationship between the WI and the functions W'I when W = L WI dx I = L W'I dyl I I w Of. Now. 0 by Theorem 5 . ik). V) a coordinate system around E N.. . . /\ dyn (q) (L a(ix l f) (p) �l 1 . .. . f*(dy' /\ .. It suffices to show that f: (x..x�f» ) dx' /\ . U) is a coordinate system. . THEOREM.k(p) (i1 < . . then q = f(p) f*(gl/Y' /\ . PROOF. a�n IJ = dy' (q) /\ .1k(p). ay a � a(i o f) a 1 .. . . ay i= l  ( a(y. ... . /\ dyn (q) (f* a�' Ip . /\ dy")(p) ( a�' Ip . . /\ dxiJ. . and (y.' (p) /\ . 7. (p) are a basis for Mp '. J* a�n IJ n a = dy'(q) /\ .208 Chapter 7 is left to the reader (Problem 16). but we will do one special case here. /\ dxn .(p) . so the dx. /\ dyn ) = det .· by dx l for the multiindex I . . . . < ik) are a basis for !. If U) i s a coordinate system around P E M. o f» ) dx' dx". Thus every kform can be \\'Titten uniquely as W= if we denote dX i l /\ .. . . Wl dxl. f*(dy' /\ .. . . d et � A · · · f\. . by Problem 41. /\ dX. 1=1 q q ) . L. dy" ) = (g o f) . 1 M + N is a Coo function between l1manifolds.
V) are two coordinate systems on M and g dyl /\ . . Vn] = /Lp. If a Coo manifold M is orientable. /\ dyn = h dxl /\ . U) and (y. the bundle TM is orientable). .I (p » determines an orientation /Lp of . then there is an nform on M which is nowhere o. . so that /L = {/Lp} is an orientation of �. . [VI .o .I (p) by Corollary 6. or . . det a/ . .. if there is a nowhere zero l1�form w on an nmanifold M. zero section w of I1 n (�) has a special significance: For each p E B. ax) Apply the theorem with f = identity map . and a partition of unity {4>u } subordinate to e. and strict inequality holds f at least one U. . . .. U»).:. . . THEOREM.] If � = " : E + B is an nplane bundle. . W = L 4>uwu. w PROOF. Thus w(p) #. then a nowh". Vn E Mp satisfy (4)u Wu )(P)(VI . U) choose an nform Wu on U such that f VI > . It is easy to see that the collection of orientations {/Lp} satisf the "compatability condition" set forth y in Chapter 3. if VI . •:. If (x. . . .. Moreover. h ( ) [This corollary shows that nforms are the geometric objects corresponding to the "even scalar densities" defined in Problem 41 0. then M is orientable (i. Vn) :0:: 0. we can choose a cover e of M by a col lection ofcoordinate systems {(x. COROLLARY. By Theorem 213 and 21 5. P E U we have or WU ( VI > . Ue(9 for every Then w is a Coo nform. . . The converse also holds: 9. . .e. . For each ( x. . vn] = /Lp. Let /L be an orientation of M. .DijlerentiaL F orms 209 8. . the nonzero w(p) E I1 n (. /\ dxn. In particular. Vn E Mp . then eadl p.. = g . . .. Vn» Now let O if and only if [VI > . . then PROOF.
which implies that it is trivial. W = L W'l dx l I Notice that the bundle r:2"(TM) is Idimensional. I a= ax· f (x.) = dWI W.) = dWI + dw. then r:2"(TM) has a nowhere 0 section. We begin by finding some properties of (still defined with respect to this particular coordinate system). which in a coordinate system U) is given by dw with the The second method is a lot sneakier. df df = " af df = L1 � dxj . WI is a kform. [Generally. We have shown that if M is orientable. w). if the bundle r:2"(TM) is trivial. provided that the base space B is "paracompact" (every open cover has a locallyfinite refinement). and we can define a (k + I )form dw. f W'l WI. the differential of w. j= ax} If W is a kform W = L Wl dx l . Conversely. I then each dWI is a Iform. (I) (2) If d(wl + w. so M is orientable. if � is a kplane bundle..210 Chapter 7 It turns out that this definition does not depend on the coordinate system. d' = o. then r:2 k (n is trivial if and only if � is orientable. by dw = L dwl dx l I " aW = LL1 l dx· l\ dx I . of course. 1 0 . This can be proved in several ways.] Just as r:20( V) has been introduced as another name for 1R. + ( I /wl dw. (3) d(dw) = o. then it certainly has a nowhere 0 section. 1\ 1\  1\ . The first way is to use a bruteforce computation: comparing the coefficients in the expression f W f . then d(wl w. PROPOSITION.. Briefly. a Oform on M will just mean a function on M (and 1\ will just mean For every Oform we have the Iform (recall that ( X) XU»).
Then WI /\ W. + (llwI /\ dw.) = d'wl /\ w. == == == == == == (I) is clear.. = .. To prove (2) we fIrst note that because o f (I) it suffices to consider only In this sum. . =gdxJ. the terms a=1 /3=1 a=l Q  and cancel in pairs . (3) d'(d'f) = o.dx· /\ dxl L. Suppose for all k. ax. (I) d'(wl + W2 ) = d'w l + d'w. Then n af dw ". = fg dx l /\ dxJ and d(wl /\ w. d' takes kforms on U to (k + I )forms on U. WI = fdx l w.. PROPOSITION. (2) d'(wl /\ w. so n ( n a'f d(dw) L L aX"ax· dxP /\ dx· /\ dx l ) .Di/ ji<rential F orms 211 PROOF. + (I) k w I /\ d'w.) dUg) /\ dx l /\ dxJ g df /\ dx l /\ dxJ + f dg /\ dx l /\ dxJ dWI /\ w.. •:.. +(Ilfdxl /\ dg /\ dxJ dWI /\ W. (3) It clearly suffices to consider only kforms of the form w = fdxl . \!\le next note that these properties characterize I I . (4) d'f (the old) dJ: Then d' = d on U. and satisfies d on U.
We will uSe induction on k . + (. pick any U with P E U and define PROOF and agreeing with the old d on functions. /\ . by (2) by (3) and the inductive hypothesis . dw(p) = du(w [ U)(p) ) . d(wj + w. ) /\ d'xiz /\ . ' /\ d'Xik .I we have or d'(dxl) = d'(d'xi. where by (4) . . . . . = df /\ dxl + f /\ d'(dxl) dxl = dX il /\ .) = dw. . d(w. •:. by (2). Assuming it f k . f dx1 Now So it suffices to show that d'(dxl) d'(j dxl) = d'j /\ dxl + f /\ d'(dxl) by (4). U) we have a unique du defined. satisf ying = 0 . d' = 0. /\ dw. /\ w. + dw. Given the form w.0.d'X i. . /\ w. for all k. . /\ d'(d'xi. • The third way of proving that the definition of d does not depend on the coordinate gystem is to give an invariant definition. . COROLLARY.212 Chapter 7 It is clearly enough to show that d'w = dw when w = PROOF. ' /\ dXik = d 'xi. There is a unique operator d from the kforms on M to the (k + I)forms on M. and p E M. /\ d'Xik ) d'(d'xi. /\ . . /\ d'Xik ) = 1 2 . /\ . /\ d 'Xik = 0.I /w. For each coordinate system (x.) = dw.
. . It is easy to check that dw is alternating." ". .X" " " X. is clearly linear Coo Junctions y. k Theorem 4�2 shows that there is a unique covariant tensor field dw satisf y ing (*) . PROOF..i<j. . f\1oreover. In fact. . if Xio is and using the formulas [IX.. jY] = I[X. X" l::.) where over indicates that it is omitted. .Xk+'). + = L(li+'" (x. Y.. .. . Xk+') to }:. .. . .. .li+j w([x. Y] I[X.. .. . Y] + XI . . Moreove. ) Xk+1 . X" . .. by renumbering. . + }:. . . so that it is a (k + I)form. . becomes IX. it is actually linear over lhe replaced by then }:. .. I f i s a kform o n M. . This (k + I )form agrees with dw as defined previously. X. . .. becomes I}:. . . + I }:.L(lin+j (Xjl)w(X. The operator which takes over � . Xk+. .." say) . .. it is easily seen that 2:2 becomes I }:. . To compute dw in a coordinate system (x.) . . . Xk+. + }:.X". X}.J)w(X.::k+1 (= }:. Xj ]..<ill ..=1 + L (..Differential F orms 213 1 3." ' (X" ." .u . + }:. Y] .X. io<j a brief inspection then shows that }:. .. . we might as well assume w = I dx' /\ . /\ dx .. then there i s a unique (k + I)form dw on M such that for every set of vector fields we have w XI) . .. THEOREM. . . X}.. . X.YI · X [X. U) it clearly suffices to compute d(1 dx l ).
.214 Chapter 7 dw. although of except at it does depend on the values of at points other than this must enter into our f ormula somehow.a/ax·k+ ' ) = 0 unless some (0') . k ). dw(p . . . . . Vk+') Vi . . After all. But we can hardly hope f anything or does not depend on the values better. ax} dxj dx' L. . . as for any form. ..)(p) w p . Xi p. Q'k +d is a permutation of ( I . /\ dxk dxj dxk dxk . . •:. . /\ /\ . . One other feature of our definition is common to most invariant definitions of tensorsthe presence of a term involving brackets ofvarious vector fields. . For . . . . Xi dw(X" . directly. Since the a's are increasing. /\ /\ . . but first find where are vector fields extending and then evaluate this function at By some sort of magic. We do not find )(v" . . j>k " aj = L1 ax} dx}. this happens only if j > k. iii . . j= � /\ /\ /\ . . This term is what makes the operator dw(X" " " Xk+'). Xk+l . . . dx . X" . in which case so which is just the old definition . p. a� dw = L(I/ ax dx' j>k af = ". we have dw = It is clear from H that dw(a/ax· ' . . .. . /\ This is our first real example of an invariant definition of an important tensor. Xk+. . . . this turns out to be independent of the extensions This may not seem to be much of an improvement over using a coordinate system and checking that the definition is independent of the coordinate system. and our first use of Theorem 42. .
which by Corol lary 4 implies that dx1 (q). 1 4 . .lt(q)dxP (q) restricted to !. . . is a kdimensional distribution on M.). This enables us to state a second version of Theorem 65 (The Frobenius Inte grability Theorem) in terms of differential forms. for ali i. There f ore.q. dxl (p) dxk (p) is nonzero on !.W' J(!.k independent Iforms Wk+ 1. .) is an ideal in the ring \1(M)].w([X. there are Coo functions I such that t k dx"(q) = "L. Then d(J(!.Xl belong to !.n. Y]) I = E ry /\ E E n P E ..p. By continuity.). A distribution !. PROPOSITION (THE FROBENlUS INTEGRABILITY THEOREM. . but it disappears in computations in a coordinate system.X ) It is clear that WI + w.Di/Jemlliai F orms 215 linear over the Coo functions. the same is true for q sufficiently close t o p . .) w w whenever XI. J(!. Locally. . . on M is integrable if and only if .» c J(!. . . Theorem 13 gives the following formula. Y) X(w(Y» . Define the ring \1 (M) to be the direct sum of the rings of Iforms on M. /\ in . U) so that .. . dx k (q) are linearly independent !.It dxP. SECOND VERSION).q a = k + I . around any point M we can choose a coordinate system (x. In the particular case where is a Iform. W(XI . P= l /\ . then c \1 (M) will denote the subring generated by the set of all forms with the property that (if has degree I) I =O E w I dw(X. . J(!. If !. and that W J(!. In fact. J(!. the ideal J(!. wn .Y(w(X» .) if Wl .) if w J(!."L. . . • . . P=l We can therefore let k w" = dx" .) [thus.) is generated by . .
. k. write dw"(Xi . . P>k Once we have introduced a coordinate system (x. ) = L(ej /\wj )(Xi. :+ Now P=l = L OJ /\ wj If a > i<j for certain forms OJ ... . . is integrable ifand only ifthere are functions ct k [Xi . . • • k. . the first two terms on the right vanish.wn must be linear combinations of them. /\ wp . . dw"(Xi . . we can always For I :S i. Xn (q) = a n } are integral submanifolds of 1'>.Xj(w"(Xi» .w" ([X" Xj]).Xj..Xj) = 0 dw" Notice that since the w i /\w j (i < j) span fl2 (Mq) for each q. VVe therefore h ave the E . dxn are a basis for J(I'>.. Xj]) = 0 [Xi . But each if and only if each belongs to I'>. . is integrable). .). ... io ::. . j = I. if I'>. wn generate J (1'>. 0 = dw" (Xi. Xj.. so we can write the condition d(J(I'>... . .216 Chapter 7 such that PROOF. • • Then with Xl . Let Xl . . while each if and only if E J(I'>. Xj) = 0 w"([Xi . so wk + 1 .. So I'>. . .Xj) = Xi (w"(Xj» . .) = ej�/XiU). Xn be the vector fields with . wi (Xj) = oj.) as dw" = L e. (i. wn which span Mq for each q wk+ 1. j :S k and a > k. . . . Xk span 1'>.).» c J(I'>.Xj] = L ctxp i. we have following. k are distinct.).e. . . Xj]) = W"([Xi. and io. the forms dXk+1. Xj] dw"(Xi .. So if and only if o. U) such that the slices l l {q U : Xk+ (q) = J + . Locally we can choose Iforms w I .
wn e. /\ dXik' ») /\ f*dxik by the inductive hyposthesis f*(dg /\ dxi. /\ dXik' /\ dXik ) j*(dw) : = ) = = /\ We will use induction on k. If are linearly independent If orms in a neighborhood of E M. the cases k > I will hardly ever be used (a very significant exception occurs in the last chapter of Volume V).I. then j*(dw) d(f*w). /\ . 1 6. Assuming the formula for k . /\ dxik . w' = L ft dgP • P p > 1 5 . {3 k) with . . {3 > k) with Although Theorem 13 warms the heart of many an invariant lover. For k = 0 we have.. let (x. we have d(f*w) d((f*gdxi. We can assume w g dxiJ /\ . . . COROLLARY. . . . /\ dXik' j*dXik) d(J*(g dxi. /\ . For p M. dw.orms DifJerential F 217 wk+1 . If E = = = = j*(dg)(X) = dg(f* X) [f* X](g) = X(g 0 f) d(g o f)(X) (and. 0 0 =0 = = = = . d: f: M + N is Coo and w is a kform on N. tracing through some defini tions. . . .. /\ dXik' ») /\ j*dXik + since dj* dX ik d d(Xik f) j*(d(g dxi. /\ . PROOF. Problem 1 8 gives another invariant definition of using induction on the degree of which is much simpler. .g P (ex. p dw· = L e. /\ ' ... •. The reader may reflect on the difficulties which would be involved in using the definition of Theorem 13 to prove the f ollowing important property of w. f* g is to be interpreted as g o f). U) be a coordinate system around f(p). /\ . . of course. . then there are Iforms (ex. PROPOSITION. /\ dXik' ) /\ f*dxik f*(dg /\ dxi. /\w P if and only if there are functions f .
The term "dosed" is based on an analogy with chains. The relation = 0 is just an elegant way of stating that mixed partial derivatives are equal. (The terminology "exact" is classical diff erential f orms used to be called simply "differentials". af ax. dw 0. = 0 is a necessary condition for solving = If is a If orm d w dry w d' dw d2 dw = w W d(Pdx + Qdy + Rdz) aR _ aQ A ay az �: .=1 n One property of qualifies. a. ay C The necessary condition. dxi .e. a differential was then called "exact" ifit actually was the differential of something. by the criterion of the previous chapter.. If is a 2form on � 3 . as a basic theorem of differential geometry. however.. There is another set of terminology for stating the same thing. For 2forms the situation is more complicated. i. w W L w. which will be discussed in the next chapter. aWi aWj axj axi is necessary for solving w = dj.) Since = 0. = . then the condition dw = 0. w dry. is aA + aB + ac = 0.ap = .218 Chapter 7 Now we know from Theorem 61 that these conditions are also sufficient. A form is called closed if = 0 and exact if = for some form YJ. In other words. i. = Wj . ax ay az if and only if = then = =    .�� = B aQ .e. every exact form is dosed.
Consequently de and de. Like our results about solutions to differential equations. 21f) (a cos b. = e in the region A. with aj ay ag = ax We know how to find a function a on all of IR' with w = da.o. By deleting a different ray L. b) has determinant equal to a #. then it is exact. Then e. . e) : IR ' is the inverse of the map Coo. yo ) dt + On the other hand. = e+21f in the region A . Consider the case ora closed I form w on �2: w= j dx + g dy. this result is true only locally.b) r+ {L {I" : r > fY41Y g(x. It turns out that these necessary conditions are also sufficient: if w 1S dosed. t) dl. namely a(x.y) = r j(t. then Jxo defined in Chapter 2.DijJerential Forms 219 I n general.a sin b). . 00) x {O}. however.{O}. in fact. L (a. and e. + O} x (0. The reasons f restricting ourselves to local results are now or somewhat different. we can define a different function e. is (I". agree on their common . Recall that if L c IR ' is [0. whose derivative at (a. the situation is very different if w is defined only on IR' . we are dealing with a rather strange collection o f partial differential equations (carefully selected so that we can get integrability conditions).
This example shows that it is the shape of the region.) Clearly w is also not exact in any small region containing O. w is not dJ for any C' function J : lR'{O} Indeed. In fact.L d(de. U C lR " i s contractible t o Po E U if U h a s t h e property that H(p. . 0 [0. if w J. w de only on lR ' L. 1] + M x such that For example. (It is still exact i n a neighborhood of any point of lR' .L. �" is smoothly contractible to E �n . then = =d + so d(f .) = 0 on lR' . so that together they define a If orm w on lR' . I) = p H(p.O) = Po H: x for p E M. but this is an abuse of notation. The Iform w is usually denoted by de. but not exact. A manifold M is callcd (smoothly) contractible to a point Po E M if there is a Coo function H: M [0. which is impossible. rather than its size. that determines whether or not a closed form is necessarily exact. dw = 0 [the two relations d(de) = 0 on lR' . Nevertheless. clearly imply that this i s so] .e) = 0 on lR' . we can define lRn lRn H(p.L. A computation (Problem 20) shows that w x = x'yy' dx + x' + y' dy. S o w i s closed. since lR.L .{O}. + dJ = de on lR ' . which implies that afjax = aelax and afjay = ao/ay and hence J = e+ constant on lR' L.220 Chapter 7 domain. 1] + by l'vlore generally.I) = lp .{O}.
: M + M x [0. but we will not be in a position to prove this until the next chapter. many other regions are also contractible to a point.Dijje. I] as representing time. We will show that if M is smoothly contractible to a point. I] we define i.. If we think of [0.po) E U for ° :s I :s starshaped with respect to po). 221 1 (such a region U is called Of course. I] (for any manifold M). I] by We claim that if w is a form on M x [0. and at time ° it is the constant map. at time 1 this is just the identity map."tial F orms P E U implies po + I (p . then for each time I we have a map p r+ H (p. I) of M into itself. (By the way. then every closed form on M is exact.{o} is 1I0t contractible to a point. and pay hardly any attention at all to H.{o}. this result and our investigation of the form de prove the intuitively obvious fact that JR' . For I E [0. then . I] with dw = 0.) The trick in proving our result is to analyze M x [0. the same result holds for JR" .
t) is a coordinate system on U x [0. for Conve n n i. There is an obvious function I on M x [0. .a).L.l)dl f ' af (p. while J[M is the projection on M. dl... Now for W = 2::7= 1 Wi dx i + f dl we have � � af dw [terms not involving dl] . = i= x i=] i lfwe define g : M + IR by g(p) = l f(p. and if U) is a coordinate system on M. I] (Xl 0 1fM . l)dxi . We will denote nience..* (L Wi dXi + fdl ) = L Wi ( · . I) . i=1 a i=l So dw 0 implies that f+ =  where /\ =' /\ Consequently. the projection IT on the second coordinate).(p.. then W I]. .I)dl ) dxi .O)dxi = L] (In0 1 aafi (P. I) dl. (namely.Wi (p.O) = 1r0 ' aaw · (p. . It is easy to check (or should be) that xi 0 ITM by xi . . Consider first a I form on M x [0.222 Chapter 7 we will see later (and you may try to convince yourself right now) that the theorem follows trivially from this.l)dl.xn O J[M . . I]. I]. We will begin by working in a coordi nate system on M x [0.a)dxi. axl dx'. .. = r 10 axi n n n L Wi (p.. i==] i=] Wi t . (x. aWti dx'· dl + L. = so (I) Wi (p.a) denotes the function p wi (p.Ll w.
I] we have (M x [0. [0. and ry is a (k . t)dt. . (V ] + v. I] as w] + w. I] M w If a vector space V is a direct sum V = V] E !.) w. uniquely as where w] (v]. Vk) = w ° ifsome V = w] + i E ker HM. and hence also. = j In general.Diff erential Forms then (2) 223 ag t aj aXi (p) = 10 axi (p. Applying this to the decomposition (*). the function j. it is easy to see (Problem 22) that we can write w dt. . Notice that for the tangent space of M x [0.I)form with the (dt /\ ry) ...). EEl kerHM •. I]) (p.t) = kerH. . g Equations (I) and (2) show that Now although we seem to be using a coordinate system. we write the I form w on M x [0.11 ( V) can be written EEl V. there is then a unique j with w. of two subspaces. is really independent of the coordinate system. for a kform w. ) = = w(v] ) w(v. . then any where w] (v] + v.
io*w that holds even when dw #.�ik' f dl /\ d'X l . Now I(dw)(p) 1(. ) PROOF. at a!.*v" . analogous property.l)(i. Then af dw at dl /\ dx l . (2) w = / dl /\ (r�i. Since Iw is already invariantly defined. ..0. . i. Since Iw 0. Define a (k .io*w(p). (I) w f d'Xi. . O)]dx l (p) i. I) dl dx· /\ d . .n . I] we have i. *w io*w 0.. we can just as well work in a coordinate system (X l . * w io*w d(Iw) if dw 0.io*w d(Iw). . /\ .Vk_') = 10' ry(p. . For any kform w on M [0.a=] axa dl /\ d'X· /\ diel ) (P) � a! L = = = = = it is easy to see that = = =  + = = = d(Iw) d ([ /(p. j. Then i. I) . so we just have to consider two cases. (i{' / ) Clearly I(dw) d(Iw) 0 •:. . We claim that dw ° implies that i.*w .. (Consequently. . /\ . I ) dl) dxl (p) [f(p. = = and + = . *w(p) .*w . t ). this proves the result in this case. it is easier to find a f ormula for i. /\ d. . 1 7.i. /\ dieik f diel . THEOREM. The operator I is clearly linear.io*w d(Iw) I(dw). Actually.*Vk_.I)form = = x = = + = .)dl. .l) dl) dx l Xl =] o (p.224 Chapter 7 Iw on M as follows: Iw(p)(v" .*w . I(dw)(p) ([ � (P.f(p. .
is the identity is the constant map po. So But so w = (H 0 il )*(w) = i l * (H*w) 0 = (H 0 io)*(w) = io* (H*w). such that the following diagram w . X on Difjermlial F orms. 92 ( 1 959).DijJerential Forms 225 1 8. There are classical theorems about vector fields in 1R 3 which can be derived "'om the Poincare Lemma and its converse (Problem 27). where we have an explicit formula for H. it has always been something of a mystery to me just why d turns out to be so important.141 . Even though the Poincare Lemma and its converse fit very nicely into our pattern for basic theorems about differential geometry. If M is smoothly contractible to a point po E M. •:. I] + M with H(p. Corollary 18 is called the Poincare Lemma by most geometers. d(H*w) = H* (dw) = 0. and originally d was introduced in order to obtain a uniform generalization of all these results.0 = il * (H*w) . Soc. Amer. f evelY form w on U.O) = Po Thus H 0 il : M + M H 0 io : M + M for all P E M. COROLLARY. Trans. Math. while d' = 0 is called the Poincare Lemma by some (I don't even know whether Poincare had anything to do with it. or we can solve the system of partial differeIltial equations w = dry explicitly in terms of integrals.) In the case of a starshaped open subset U of IRn . we can find (Problem 23) an explicit formula for J(H*w). Since the new form is given by an integral. = d(l(H*w» . PROOF. then every closed f orm w on M is exact. Suppose we have any operator D from kforms to Iforms. I ) = p H(p.io* (H*w) by the Theorem . atwal Operations An answer to this question is provided by a theorem of Palais. We are given H : M x [0. 1 25.
If k = I.226 Chapter 7 commutes for every Coo map I: M + N [it actually suffices to assume that the diagram commutes only for diffeomorphisms 11 . In this case. (As a corollary. Roughly. since d ' makes the above diagram commute!) There is only one other case where a nonzero existswhen k is the dimension of M and I = o. kforms on M � kforms on N Iforms on M � If orms on M Palais' theorem says that. = o. with few exceptions. but nothing else. then can only be some multiple of d. If I = k + I. d ' = 0. which we discuss in the next chapter. then can be a multiple of the identity map. Dl D D DD D lD . can be a multiple of "integration". these excep tional cases are the following.
1 ) = w (v. E 11 1 ( V) and e E 11' ( V) .J w). For v E V and w E 11k (V).. + ( . Show that if we define then 3.. (b) Show that if V1. (Proving associativity is quite messy. . . < . This definition may be used even in a field of finite characteristic.Difforenlial F orms 227 PROBLEMS 1.J w E I1k. E 111 (V) we have (Use (b) and linearity of everything. ry 2. </>n. . k} and {k + I. < a rk + /) .. w ® ry(va(l) . (b) Show from this definition that w /\ ry is alternating. A cross seclion of S' is a subset K c Sk+1 containing exactly one element from each left coset of S' . (V1 . .J W)(V1.. . Vk +/) = aEK L sgn a ... .. . .. that 7\ is nol associative... j #.J w) = .. · . Vk).). Vk) = ap • (V1 . Vk_1). . V1 .. . .) . Vk .. 4. . (c) Show that for W1 V .w . . .J ( w .J w. and define W 7\ ry = Alt(w ® ry) . E 11 k (V) and w.. .) (c) A permutation a E Sk +1 is called a s"u permutalion if a ( l ) < a (2) < . . .J (W1 /\ w. .1 (V) by (v . . and w /\ ry = ( _ I ) klry /\ w. . Va(k+/) . then if j = i• . . . k +/} invariant. V . ..) = (v . . " . (Try w. · .J W 1 ) /\ w. Show a • p . Show that the set of all shuffle permutations is a cross section of S'..l lw1 /\ (v . . . . .any i. (a) Show that This is sometime also called the inner product and the notation ivw is also used. . Vn is a basis of V with dual basis </>1 . Let Alt be Alt without the factor 1 / k!.. Let S' C Sk+1 be the subgroup of all a which leave both sets { I .J ( v .flle a rk ) and a rk + I ) < a rk + 2) < " .) (a) Show that for any cross section K we have w /\ ry(v]. . we define the contraction v . . .. . . .
then W = (</>1 /\ </>.J Wl) /\ w. Show that with this definition WI /\ W. .228 Chapter 7 + W /\W.. .J(Wl /\ w. = ( _ I )k! w. /\ W l . . then X . n n. . then Ann(w l ) C Alln (w..J w. (a) Show that dim Ann(w ) ::. + (I/Wl /\ (X. . Show that if Wl is a kform. . Vk !) .. l (d) Formula (c) can be used to give a definition of by induction on k 1 (which works for vector spaces over any field): If is defined for forms of degree adding up to < k I.o. we define + + . . /\ Vk E !. Reformulate parts (a)(d) in terms of this /\ product.J w(p). Ann(wl ) + Ann(w. (e) If V has dimension then any W E !. i = 1 ..I)form X.1' (V) is decomposable. we define the annihilator of W to be Ann(w) = {</> E V* : </> /\ W = OJ. (a) If dim V 3. (b) Every subspace of V* is Ann(w) for some decomposable w.J Wl ) /\ w. (c) If W and are decomposable.J w by (X. /\ dl (p) #. Hin!: Look at W /\ w . (b) If </>i.. which is unique up to a multiplicative constant.(v" .). 7.J w.. n . n 6.1 k (V).J w)(p) = X(p). (f) If X is a vector field on M and W a kform on M we define a (k .4 = YJ W2 YJ.o. changes the sign of the right side).1 l (V) is decomposable. .) + (</>3 /\ </>4) is not decomposable. are independent. Vk !) + (. . and that equality holds if and only if W is decomposable.) = Ann(wl /\ w. . (f) Since Vi E V can be regarded as elements of V**. Show that functions fi. In this case. we can consider V /\ . For any W E !. /\ l + Wl /\ w. .](v" .I/[Wl /\ (Vl . then every W E !.)](v" . Vk !) = [(Vl . #. .) = (X. . . An element W !.). I : M IR form a coordinate system in a n neighborhood of p E M ifand only if dfi /\ . then Ann(wl ) Ann(w. /\ </>k for some </>i E V* = !..1 k ( V) is called decomposable if W = </>1 /\ .) = {OJ if and only if Wl /\ w. k. + n + E ::. (e) Prove by induction that /\ is bilinear and that W l /\ W.1k (V*). .) if and only if W2 Wj l/\ for some (d) I f Wi are decomposable. 5.1 1 (V) . is skewsymmetric (it is only necessary to check that interchanging V l and v.
. . where w ' does not involve 1/11 or 1/12. show that the rank of w is the rank of the ma trix ( aij). .p. For H = hl < . . .. . . " /\ Wn = L: C K VK . </>n of V* such that Hint: If W = L aijVti /\ Vtj . old it is called the rank of w. C K = det ( ap+ l 'k' : au. Let A = (aij) be an 11 X il matrix. . : al•hP ' . and let q = n . . < hp and K = k l < ' . let Vt. Vtn so that W = </>l /\ </>2 + w '. If 10.. Show that there is a basis </>" • • • . show that B H = det ( al. ' < kq.h] ap. .itp ) . Vn is a basis for V and Wj = 'LJ=l (ljiVj. i<j choose </>l involving Vth Vt3. and that the (r + I)f wedge product is O. ap. .1 2 (V). (c) If w = L i<j a ijVti /\ Vtj . K . (a) Let W E !. . . .k] (a) If v" . .. . . . Let I :s p :s n be fixed.Diff erential Forms 229 8. j= ! Wt /\ . . Vtn and </>2 involving Vt2. /\ Wp = L B H VH H Wp+l /\ . Thus r is welldetermined. 9. (b) Show that the rfold wedge product w /\ . . . /\ W is nonzero and decompos able. Vn is a basis of V and show that n Wi = L QjiVj.
.. where eH. the bundle whose fibre at Since we can regard as an element of ( Mp )** . kq = " . . ' ./\ . aIlY section of I1 n (T*M) can be written locally as h . hp . . (c) Prove "Laplace's expansion" det A 1/1" 1 1 . " } .H' K = H'. . Show that VH /\ Vx = . j= 1 k where bundle. For example. .. 1/Ik E y satisf . . eH. . . . . . . H E V* V* be independent and suppose that aji = aij ' Then =:: 1/Ii L Qjiq. a a (a) Show that if then .I (p)]*). L. . (I . . . we can consider 11 k (C). . h2. · /\ . /\ Vn " K #.H (arranged in increasing order). .H' VI /\ . . if � = :n: : E + B is a vector 12.' ax l axil TM using 11 and other operations. (Cartan's Lemma) Let <PI . • ".w B H C H' . . kl . H' is the sign of the permutation {o eH. (/>k .230 Chapter 7 (b) Let H' = { I . . .j . ) h l . we can consider sections of bundles constructed from p is 11k ([:n:. . In addition to forms. . .
(d) Show that the covariant relative tensor o f type e ) and weight 1 defined i n Problem 410. (b) Let :Tik [m] ( V) denote the vector space of all multilinear functions show that sections of :Tit. Interpret the relative tensor with com 'v' times V x . .] (TM) correspond to (even) relative tensors of type (1) and weight .w\ TM) correspond t o (even) relative tensors of type (J) and weight w. • . .I . Similarly for :Tit.Differential F orms 231 This shows that sections of r:2n ( T * M) are the geometric objects corresponding to the (even) relative scalars of weight I in Problem 410. We can identify :Tik ( V) with A w an integer Since r:2m(v) '" Am(v*) '" [Am(v)]*.. except that r:2m(v) is replaced by r:2m( v*).W ( V) . then elements of r:2n(v) can be represented by real numbers [times the element V*I /\ . these results can all be restated. (Notice that if VI > . </>n ) f+ ponents e'I .] · ··ill .w] (V) b e defined like :Tik [n] . x V + IR which are of for some W E r:2 n ( v) . . we can identify :Tik[m] (V) with :Tif ) ( V) with m I k Q9 V* ® Q9 V ® Am(v) I k Q9 V* ® Q9 V ® Am( V*) . k times I times .. W ( V) denotes all functions ry : the form V* x . Show that sections o f :Tik [n. . ] . Let :Tik[n . . x V* + r:2n ( V) n </>1 /\ /\ </>n . .i'1 similarly. . with components e. corresponds to the map Show that sections of :Tik [nl (TM) correspond to (even) relative tensors of type (1) and weight I . . Vn is a basis for V. . I times k limes m • • • given by (</>1> . /\ v*nJ .) (c) If :Tif ] ( V) is defined similarly. � x � + r:2m (v) . except that r:2n(v) i s replaced by r:2n .. W (f) For those who know about tensor products V ® W and exterior algebras k ( V). . (e) Suppose r:2n .
V + V is X(A) = det(A! .232 Consider. . more generally.) (b) Conclude that det AB = (det A) (det B). (a) If V has dimension n and A : V + V is a linear transformation. · · + ( .I + C2 An.. .  n trix (a! l (with respect to some basis). (c) Let 8/. (This may be used as a definition of det A. (b) Conclude that cdA B) = cdBA ).1k (V) + !. then the map A * : !. . then the function p r+ cdA (p» can be defined as a (2k)fold contraction of � A 0 · · · 0 A 0 8. Show that C = det A . 14. + (_ I)nc . Let P(Xij) be a polynomial in n 2 variables. show that .. .1 ) for all A and all invertible B .1n(v) must be multiplication by some constant c.jk V + V has a ma Thus. show that sections of 7ik[n. a!k ()Jl ··iN ' lk . j]. .'. k times 15. Cn defined in Problem 14.'h. if 8 is as defined on page 1 30.A) = An .wl (TM) and 7if . .(trace A)A n . This problem outlines a proof that any invariant P is a polynomial in the polynomials CI .1n(v) + !. Call P invariant if PtA) = P(BA B . For every n XI1 matrix A = (aij) we then have a number P(aij). and A is a tensor of type G).· k •l L i] . Chapter 7 and weight w and k 7it .::// be as defined in Problem 45 (xiii)..wl (V) = 0 V* 0 01 V 0 0 w A m ( V*) . I}. respectively. Recall that the characteristic polynomial of A : W.1 k (V) .I) " det A = An C l An.wl (TM) correspond to (even) relative tensors of type (1) n 7ik [m:wl ( V) = &/ V* 0 &/ V 0 0w A m ( V) 13. m Noting that An(V) 0 · · · 0 An(v) is always Idimensional. .2 + · .· . . We will . . . . (a) Show that Ck = trace of A * : !. If A : �l l l Ck (A) = � " a'1l a'2z .t + .
. it follows that ei(A) = a. . Recall that the at can be defined by the equation n i= 1 TI (y . . . . o .Yt) = yn . Yn can be written as a polynomial in at . The polynomial R has real coefficients if P does. . . they are the coefficients. by definition. Yn). . . . an(Yl . this is "the principal of irrelevance of algebraic inequalities". the roots of the polynomial X(A).. . . compare pg. that the equation holds for all matrices A over <C.Aj)'. Thus. Yn) in the variables Yl .) Now suppose that the coefficients of P are real and that PtA) = P(BA B . where ai is the ith elementary symmetric polynomial of Yb . (This last conclu sion follows even if <C is replaced by some other field. Conclude. . (b) PtA) = R(edA).0 ). an. Show that D(A) can be written as a polynomial in the entries of A . . \( 375. up to sign. .(A b " " An). .) = n' . . . (d) Show that P t A ) R(el (A). . . (Regard or the equation as polynomial equations in the aij and bij . We will first consider matrices A over the complex numbers <C (the coefficients of P may also be complex). (e) The same equation holds f complex A and complex invertible B. en(A» whenever D(A) # O. . . . . .DijJ rential FOWlS e 233 need the algebraic result that any symmetric polynomial Q( Yb ' . .I )"an.1 + . by continuity. . Yn. Since the eigenvalues A t . . Yn Then there is a polynomial R such that Q (Yb . en(A» for all diagonalizable A . of the polynomial with roots Y b . . . · · . . . Yn) = R (adyl o . . since the set where D # 0 is Zariskidense.1 ) for all real A and real invertible B.atyn. . " Yn) to be PtA) where A is the diagonal matrix Yl ( . . (c) The discriminant D ( A ) i s defined a s n ih (Ai . (a) Define Q(y l o . . + ( . An of a matrix A are. . . . Yn». where Ai are the eigenvalues of A. . . Yn .
Xk ) k + " (. show that X(w(X" . .. /\ j _ (b) Show that is (c) Using 514(e).·.. ik .Xi . E w 11k (V) show that W(WI. . (a) Show that i f W i s a kf orm.... (d) Deduce the following two expressions: dw(X" . . Show that d( Li<j aij dxi dx j ) = 0 if and only if aai aai aaj axk ax}k + ax'k = 0 for all i < j < k .XJ..I ) i+ LXiw(Xl.l)i+l w([X. 17. Xk» = LxW(X1 . . • • or Vn be a basis f V. . .. Vik )' i]. j=] Q'jiVj . . . .. . . Xk+l ) k+l l = " ( ... . . X".. 18. .Xi . .. (c) Check directly from (b) that the definition of d does not depend on the coordinate system. . then s o Lxw.» = Lx(w(XI. . . In Problem 514 we defined Lx A for any tensor field A . . .234 1 6. ..Xi]. Wk V be given by Wi = L n • • . L. .. . (a) Let Chapter 7 Vl . and let WI. . .Xk ). Xk+l ) L. (b) Generalize Theorem 7 and Corollary 8 to kforms.. . . . . . . For E . Wk ) = l=ilL a/ W(Vi" <···<lk where a{ is the determinant of the k x k submatrix of (aij ) obtained by selecting rows .
axj axi == x dy' + y dx x y' for all i. I aUi aU + j ai = '2 axj axi l == (LxI w)(X" . a'aij a'aik axk ax! .'2 I ) ..d(XI .A dx = dg for some function g with g (O) = g(I). == au) laxk "dO" 21. 19.J w)(X" . . . dw(Xt . . . 1 ] . 1R' . i. show that there is a unique number A such that W . Hin!: Integrate the equation w . show that d(Lxw) == Lx(dw). . . ..e. Un in a neighborhood of any point in IRn with ( ) it is necessary and sufficient that 20... . == (At most places 0 = arctan y Ix [ + a constant]. . . (b) Let i : S l . and let a' = i*(dO). . _ . . l. . ' functions on IRn with aU = aji . l . .{Oj be the inclusion. .d(X . Xk+ I ). . . I ] to find A. .I ) {Xi(w(Xt .axj ax! a'a/j a'a!k == axk axi . .J dw = Lxw . .. . If c : [0. Compute that Hin!: First set up partial differential equations for the functions fik aUk laxj .)  21f dx . . Xk+l ) k+l 1 i+1 Xi. (a) If w is a Iform / dx on [0. S l is c(x) = (cos 21fx. Xk+t ) .A dx = dg on [0.J w). Let aij be .) (f) Using (e). and use Theorem 61. . .erential Forms Diff dw(XI . . k . Show that in order for there to be functions U I .. . sin 21fx). Xk +t » i=l 235 _ (e) Show that X . .. Xk+l ) (This may be used to give an inductive definition of d.. show that cOra') (c) If w is a closed I form on S I show that there is a unique number A such that w Aa' is exact. I] with /(0) = /( 1 ).
is a (k . (The converse is proved in Problem 89. 1] + U by H(p. J(H*w) 1 (IX) dl) xi. . a= ' l < " '<'k l _ . show that 24.I Wi . .1 (1 I k . Let U c IRn be an open set starshaped with respect to 0. .(VI . . . (a) Let U c IR ' be a bounded open set such that IR ' .. but U is not contractible to a point.. vp) = 0 if some V. If W= WI 22.*.(V\.) Obtain U as an increasing union of sets. then W can be written uniquely as WI + (b) If dim V. ..l) Ip. Vk I ) = 0 ifsome Vi V. /\ dXik • Hint: (b) Find a bounded open set U C 1R 3 such that 1R 3 . . and hence smoothly contractible to a point. where W I is a kform and w.U i s connected. .236 Chapter 7 W 11k (VI V. Vk ) = 0 if some Vi V.. W. . . = I. . . 23. . and 0 #. ic).I)form such that WI (VI . va) = 0 if some Vi V2 W. . (a) Show that every E /\ W. and define H: U [0. . dxi. . Show that U is diffeomorphic to IR ' .Ik /\ . VI . . /\ d7a /\ . . . has degree f3 = k . = L. t( 1 ) .U is connected. 1/ k . where has degree 0' and WI EEl 0' E E E /\ E E x = on U.) can be written as a sum of forms w. the kth set being a finite union of squares containing the set of points in U whose distance from boundary U is ::. . .ic (w. V.and Wt(Vt.
. and express it in terms of the e*i.= " L aXl.t (JRn). but the "obvious" proof does not work. (a) If j : JRn + JR. 27.' . .l o w) n = x . the gradient of j. aXl. show that there is a unique vector Vt x . (a. (First find all e. using the expansion of a matrL by minors. . define a vector field grad j. x ej. . b) L>i bi. . .) 26.= 1 = (a) If Vt. . grad j = " . Let ( . Is U homeo morphic to JRn? (It would certainly appear so. ' x det (b) Show that x . . . Let U c JRn be an open set starshaped with respect to O. .Di erential Forms ff 237 25. .t for all W E JRn. since the length of rays from 0 to the boundary of the set could vary discontinuously. Vn _l with (Vt E �n. x E \1 n .) Intl"Oducing the formal symbolism i= 1 i=1 Dd . ) be the usual inner product on JRn. on JRn by a n a n aj . x (c) For JR3 show that () � Vn . L a X l. X Vn _l E �n V .
w ) . then X = curl Y for some vector field Y on U. X).238 Chapter 7 we can write grad I = VI If (grad J)(p) = wp . Define forms Wx = a l �x = a l Show that (c) Conclude that dl = Wgcad J d(wx) = �cudX d(�x) = (div X) dx /\ dy /\ dz. . dx + a2 dy + a3 dz dy /\ dz + a2 dz /\ dx + a3 dx /\ dy. 2::7=1 a ' a/ax' is a vector field on IR n . then X = grad I for some fU l Iction I : U > IR. Conclude that V/(p) is the direction in which I is changing fastest at p . curl grad I = 0 div curl X = o.) We also define. if we regard vp E IR np). (d) If X is a vector field on a starshaped open set U C IR n and curl X = 0. Similarly. if div X = 0. we define the divergence (b) If X of X as (Symbolically. show that Du/(p) = ( v. where Du/(p) denotes the directional derivative i n t h e direction v a t p (or simply vp (J). we can write div X = (V. for n = 3.
C(tl) e(�I) :_ C2(t1) . To be precise.lid. �n ). then the limit is denoted by l Jdx+gdy. as the maximum of Ii .c' (l1')] + g(C(�I))[C2{11) . i. that is. then we can divide the curve c into pieces.) . the piece going fi'om C(li. each such piece is approximately a straight segment.e' (til) horizontal projection C'{lI) . if IIPII = m!'x( ti . Suppose. 1]. .c2{1i_ l l.c2{11_l ll i=1 If these sums approach a limit as the "mesh" I PI I of P approaches 0. < In = 1 of [0.CHAPTER 8 INTEGRATION c: [0. y. (This is a complicated limit. g: JR2 JR. denote the coordinate functions on JR2). If we choose a partition 0 = 10 < .' ) to C(li). for example.\C.Ii_I are small.C'{lI_ ' ) and vertical projection c2 (11) .Ii_ I approadlCS 0.h x he basic concept of this chapter generalizes line and surface integrals. . 1] JR2 is a curve and w = J dx + g dy is a I form on JR2 (where J. and and. then the equation lim S(P. �) = L J(C(�I» [C'(li) . . �II) .�) = [ Jdx +gdy lc E IIPII� O 239 . We can choose points C(�I) on each piece by choosing points �I [IihIll For each partition P and each such choice � = (� .C2(t1_' ) e(O) c(I[_. with e(1) that > > n T which first arose fi'om very physical considerations. consider the sum " S(P. When the differences Ii .
�) IiI Ii J g J dx + g dy dx.Iid· i=1 . (In classical terminology. C(I) (I. while C2(li) .240 means: for all E we have > Chapter 8 0. Yo). If we consider a "force field" on described by the vector field � P. consider the special case where S(P. for all choices for The limit which we have just defined is called a "line integral". there is a Ii > 0 such that for all partitions P with II P II < Ii. the limit js the natural definition of the work done in the general case. the integral is the "sum" of these infinitely small displacements. a a J.+g ax ay then is the "work" involved in moving a unit mass along the curve c in the case where c is actually a straight line between and and and are constant along these straight line segments. so .) Before worrying about how to compute this limit. CI (li ) .IiI .) IR2. it has a natural physical interpretation. YO)(li .�) = LJ(�i. the differential would be described as the work done by the force field on an "in fInitely small" displacement with components ely.C2(ti_d S(P. = 0. = yo+� In this case.C l (tid = Ii .
e. In general.=1 it looks like they should approach. x c'.yo).lil l· i= = b These sums approach A somewhat messy argument (problem I) shows that these sums approach what (b .�) = (b . C' (li) . namely Physicists' notation (or abuse thereof) makes it easy to remember this result. L1 J(�ib + ( 1 . Yo)(li . if Yo C(I) = (Ib + (1 . so S(P. .. YO)dX.C' (li' ) = C" (Cii)(li liIl Cii E [Ii. for any curve c.c' (Ii' ) = (b . of are denoted simply by and The components [i.integration These sums approach 241 lJdx + gdy = l ' J(X.a) l ' J(xb + ( 1 x)a. On the other hand. So S(P. we have.I)a. then a c' (Ii) .yo)dx. by the mean value theorem.a) .�i)a.yo)dx lb J(x.�) = L { J(C(�i» C" (Cii) + g(C(�i » C" (f3i)} (Ii liI).Ii] C' (li) . denotes l' [J(C(I» C" (I) + g(C(I»)c" (I)] dl. c' C y x .a)(ti .li' ).C' (ti1l = C" (f3i)(li .li1l f3i E [Ii"" Ii].
this is indicated c1assicaIJy by saying "let x = X(I). .242 0 y y o Cha pter 8 y =y x e and denotes e. The above in tegral is then written 1 / dx +gdy [o I [j(x'Y) dx +g(x'Y) di] d/ . dY d1 J " In preference to this physical interpretation of "line integrals". e goes with constant velocity on each (IiI . Ii). we Can in troduce a more geometrical interpretation. (I) ] .li1l [=1 clearly also approach Consider the special case where 1 1 [j(e(l) e ll(l) + g(e(I» e21(1)] d/ . RecaIJ that denotes the = de/dl(�i) e(ti ) tangent vector of c at lime de dl (�i) �i . Then the sums = L [j(e(�i» e ll(�i) + g(e(�i»)c21(�i)] ' (Ii .
�i E (li' . and then extend W to IR'. the infinitely small piece at with components having length 1W c. dc/dl C C.integration 243 In this case. t' [ i= length of dx. � (�i)] . . then dc dt (�i) = the constant speed on (Ii). we choose W(C(I» E !1 ' (IR'c (I» so that W(c(t)) (� ) = 1 . (Ii . whereby the curve is divided into infinitely small parts. /(C(I)) dx + g(c(I » dy. Before pushing this geometrical interpretation too far. If we choose any length of  is the length of and the limit of such sums. we should note that there is no Iform w on IR' such that 1 w = length of C for all curves c. dy. kernel w (c(t» But if W arbitrarily). (choosing the kernel of can be thought of as the "length" of when our ruler is changing contin uously in a way specified by Notice that the restriction of to the I dimensional subspace of IR'c(l) spanned by is a constant times "signed length". can be used as a w(C(I» w: = limit of the sums (*) C crt).IiI) = length of segment from C(li1l to C(li). (Ii .li1l dl c. C is . for a general definition of the length of The line integral C. Ii).. It is true that for a given oneone curve c we can produce a form w which works for c. The natural way to specify a continuously changing length along is to specif a length on its tangent vectors.Ii) length of the segment from C(li 1l t o C(li) Ii {I I so [length of � (�i)] . this is the modern counterpart of y the classical conception.
ffi. J] 7 ffi. l l c c: [0. so we can define Ie W as the limit of these sums. c: [0.� there is no element of !1 ' (IR 2c(d) which has the value on all three vectors. 1] 7 [ [/(c(l» C" (I) g(c(t))c2l(t)] cit. This is no longer 50 dear . namely c: [0. any curve contained in an integral submanifold of /'. In general.�) for c is dearly equal to a sum S ( P'.even for a curve but in this case we can proceed right to the integral these sums approach. the subspaces /'. because it is ob vious with our original definition and merely true for our new definition. 1] 7 ffi."hen we consider the sums (*) f a curve or M. in the situation shown belov.2 wc op w ("the integral of w over (' is independent of the parameterization"). Every sum S(P. 1 2 c One property of line integrals should be mentioned now. If is a oneone increasing function from onto then the p: I > > M is called a reparameterization of cit has exactly the curve c o p : same image as but transverses it at a different rate. + . I] [0.244 Chapter 8 not oneone this may be impossible. used to define this generalized "length". and conversely. will have "length" Later we will see a way of circumventing this difficulty. make sense even if is a curve in a manifold M (where there is no notion of "length").2. if we are interested in obtaining the ordinary length of a curve. For the present. for example. so it is clear from our first definition thal for a curve we have [0. I]. we note that the sums t*j. I] [0.2 0. I] c. ] [0. [0. given any w on which is everywhere nonzero. and w is a I form on M. �') for c o p. nor is it deal.p = ker w(p) form a I dimensional distribution on IR .
Consider once again the case of a Iform on where [ [/(e(I»)c" (I) g(e(I»)c2l(1)] dl = lP' ( I) [/(e(p(u» )e" (p(u)) g(e(p(u)))e2l(p(u))]p'(u) du = [ [/(e p(u))(e p)" (u) g(e p(u))(e p)21(u)] duo p ' (O) + 0 I = p(u) gives 0 + 0 0 e([O. 1]2 JRn . Everything we have said for curves e: [0.. let � = e* (�) a ae ay = e* ( ay ) . For a pair of partitions and 10 < .n . which will eventually become our formal choice. 1]. and take the integral of h on [0. We are being a bit sloppy about all this because we are about to introduce yet a third definition. and a l form there is a similar calcula 2:::7=1 tion. we can introduce a coordinate system for our calculations if lies in one coordinate system. e I is the standard coordinate system on JR. . 0 0 0 0 0 e: [0. for a general manifold M.integration 245 The result then follows from a calculation: the substitution + For a curve in ffi. < In of [0. or break up into several pieces otherwise.2. 1]. to be precise. we write c*(J dx + g dy) = h cll (in the unique possible way). . so that formally we just integrate e*(/ dx + g dy). 1] JRn could be generalized to functions e: [0. 1 w = [ [/(e(I» e"(I) + g(e(l))e2l(1)] dl. If x and y m·e the coordinate functions on JR 2.. then for the map 2 e*(J dx + g dy) = (J e)e*(dx) + (g e)e*(dy) = (J c) d(x c) + (g c) dry c) = (J e)e" dl + (g e)e 21 dl. I]) w= Wi dxi . < Notice that if ffi. I] > JR we have 0 0 0 > > So s'" �ffQS . if we choose < .
The inclusion map of It in will be denoted by It > it is called the standard kcube. . are the coordinate functions. we have a special definition: a Oform is a function !. .Sj_JJ(lj .. We will Jet I Jo = = so that a singular Ocube is determined by the one point M. c [0. . * (�ij» ) (Sf . and for a singular Ocube c we define 1 ! !(c(0)). ° E IR. For k 0. Xk w w = ! dx l /\ .lj_l ) is a "generalized area" of the parallelogram spanned by ac ac a. l lk where the right hand side has just been defined. . llk in c1assical notation. and . and c is a singular k cube in M. /\ dxk We define = J ! (X I . ( = ) = . dXk [O. IRo ck c(O) E IR Jk : [0. ..f ! notatIOn attempts to w�ic? modern . . A Coo function It > M is called a singular kcube in M (the word "singular" indicates that is not necessarily oneone). [0. then Cha pter 8 [0. . we define 1 w = f c'w. . IRk. The limit of sums of these terms can be thought of as a "generalized area" of c. .Xk ) dX I . To make a long slOry short. If is a kform on It. f W lO be O.246 X �jj E [Sj_l .. we now proceed with the formal definitions. Si_1 c: [0. Sj1 [ljI . then can be written uniquely as Sj w(C(�ij)) (�: (�jj). w X l . ay(�ij ). lj1 and w is a 2form on IRn. [O.(�fj ). mImIC as far [O. . l lk [ l lk as logic permits If w is a k form on M.
. Let be oneone onto with det let be a singular k cube in M and let w be a kform on M. I}" 0 det e([O. We have L w = f (e plow = f p*(e*w) p = f e*(w) 0 � I� � It [0. By definition. since p is onto . • . C{[O.1]" p: [0. by the Proposition.illlegmiion 247 > JRn be a oneone singular ncube with PROPOSITION. 1]n e' 0 [0. l w = f f. II") by the change of variable formula . Then e: [0. /\ /\ . It [0. 1]". /\ .11" = [O. . Then e PROOF.1)II) PROOF. It [ w = [ w. Let det 2: on Let be the nform 1. /\ dx" by Theorem 7 7 dx" by assumption > 2. w w = J dxl /\ • • . Jc leo p p' 2: 0. /\ dx". •:. = 1 w f e*(w) f (J e') dxl f (J e)1 e'l dxl f J [0.11" = 0 e) (det [0. COROLLARY.
=I 2(c.. so that the change of variable formula will pop up). . in the obvious The reason for introducing k chains is that to every kchain c (which may be just a singular k cube) we wish to associate a (k . Our definition of the integral of a k form w aver a singular k cube < can immediately be generalized.248 > Chapter 8 > independence of parameterization. <. purely formally. it is called orientation preserving or orientation reversing depending on whether det p' 0 or det p' < 0 everywhere. j = Li ajC. diff rential forms are the things we integrate be e cause they transform correctly (i. in accordance with Theorem 77. I t is a Coo oneone onto map with det p' i' 0 everywhere (so that p' is also COO). Notice that there would be no such result if we tried to define the integral over of a Coo function M > )R by the formula The map c o p: [0. + 3<4 ) + (2)« . + <3 + C2) = 2C2 . e.g. /\ dx k ).I )chain a c. which is called the boundary of c. and multiply them by integers. The corollary thus shows > I: c [O. will also be denoted simply by Ct . I I . c: [0. e.2<3 + 6c4.e. 1\1orcovcr. and which is supposed to be the sum of the various singular . if From a formal point of view. provided it is orientation preserving. I] M then l' I(c(t)) dt is generally i' l' I(c(p(t»)) cll. The k chain Ie. > f l o c. an ori entation reversing reparameterization clearly changes the sign of the integral. It M is called a reparameterization of c if p: [0. It [0..' lk For example. .g. . functions On a manifold cannot be integrated (we can integrate a function on the manifold )Rk only because it gives us a form dx ' /\ . We add kchains. w(' define the integral of w over a kchain way: c J2::: uiCf [ W = I>lCj W. A kehain is simply a formal (finite) sum of singular k cubes multiplied by integers.
.xn . . 1 .l) /11 . .l) = (x l . o f the four singular I cubes shown o n the right. . it for example.xnI ) 1. .xil . ..Integration 249 (k .I ). i . . .1) . but tbe sum. . . Xi . . /11 . . then 1i). .I . . . The boundary of y the sum of the four singular I cubes indicated below on the left. (x) = I" (xl . I) = (X l . . . . xiI .I)cubes around the boundary of each singular k cube in c.xn.:. l . will not be is convenient to modif this idea. . .I . I )face of as follows: If X E [0.O) (x) = I" (x l . 1._n . O. .xi . .0) . ). In practice. .I . . P I I 12.:. +1 +1 with the indicated coefficients.O.O) In) I]n. . . (Notice that this will not change the integral of a I form over a/ 2 ) For each i with 1 :0 i :0 n we first define two singular (n .O)face and (i. . . .l ).x i . .. .I)cubes and (the (i. xi.x i . X Ii).
O) s  ..1 = O.1) a(aC) = (R . 1]2 + C(2 .(R ..).) ' C(I . 1]0 Notice that for a I cube c : [0. which we will usually simply identif with the point P = c(O).P) p C(2.S) . > M. C(1. 1 .I) C(1. I] > M we have so a(aC) = ac = C(2.250 Chapter 8 The (i. i=1 a=O. for a singular 2cube c: [0.(S PH (Q = 0.) = C (Iv. C(l..Q) .l These definitions all make sense only for n ::: I ." )face of a singular ncube Now we define C(O) � C is defined by C(i. y we define a c to be the number 1 E JR.O) We also have. For the case of a Ocube > M.O) / C(I) C 0 Finally. 0).1) R . C(i. the boundary of all nchain Li aici is defined by aC = L L ( _ I )i+. and for a Ochain Li aiCi we define c: [0.
this is the case if c = c) e2 .P) (i. Xn2). from the definition ( /(�.. we have: 3.) occur with opposite signs. a2 = o. xn2) = In (x l .. " . PROPOSITION. xj . ..) ( /.xj. Now i I a(ac) = a = L a=O.xj. C{ . . xi. .Pl = ( /'( + . . . Pl)(" . It follows easily for any j I singular ncube c that ( c( . ( /U+I. . Therefore ' all terms cancel in pairs.) = J. .) U. xiI . . Xj .) for i :0 j :0 n I.=1  In this sum. P) ( . .1 nLI P=O. a . we have. .1 .integration 251 From a picture it can be checked that this also happens for a singular 3cube.. . . Briefly. Thus ( /. a good exercise because this involves figuring out just what the boundary of a 3cube looks like.) U.) (X I . p. xj. but even ac = O . .( P) (x » / = J(�.1 (I )'+. /3. and consider ( /D.)) L .P) = (CU + . Notice that for some nchains c we have not only a(ac) = 0. 1]"2 . and a(ac) = O. Pl)(" . it is clearly also true for singular nchains.) u .P) and ( CU+I . . where CT and C2 are two I cubes  . then a(ac) = o.j+I ..P) (XI. X i .l (I)'+·+j+p(C(" . xn2). xj .. .I . C(" . . . xn2) j I i = In (x l . PROOF Let i :0 j :0 n . a. .) u. . xiI ..: ..1 .) for i :0 j ::S n . .) U. Similarly.P) (x ) = J'J . fj.P) · L .1 . .. ( C(i. . xi. . . .P) ' For x E [0. . For example.\ (x » = J{ + . . If c is any nchain in M.•) U . •:. .I. x . . . Since the theorem is true for singular ncubes.L 1=1 j=1 (t a=O. In general. . . .I . . . . .Pp(�:.
The connection between forms and chains goes much deeper than that. then ac = 0 precisely when c(O) = c(I).{OJ there is a I form "de" which is closed but not exact. If c is just a singular I cube itself. by the classical term of "exact". Recall that a differential form w with dw = 0 is also called "closed".e. this terminology has been purposely chosen to parallel the terminology for chains (on the other hand. In general.. namely. reciprocally. There is also a I chain c which is closed but not a boundary.252 with c. (0) = Chapter 8 C2 (0) and CI (1) = c2 (1). expressed by the relations d' = 0 and a ' = O. we have seen that on IR ' . This parallel terminology \'\'as not chosen merely because of the formal similarities between d and a. For example. any k chain c is called closed if ac = O. i. but is simply called "a boundary"). a chain of the form ac is not desCllbed. when c is a "closed" curve. a closed curve encircling .
. .)* U dx l j. .1/ f [O. Therefore = ( . the simplest proof uses the theorem which establishes the connection between forms.integra/lOll 253 the point ° once. 1 . chains. . /\ = jof dxi /\ . . /\ . then Jc [ dw = [ w. . .k ) dx. . O. a. dx. .l)k f l(xl .I )form on M and c is a k chain in M.l)k1 [ J(k .I )form on IR k and c = J k In this case. We now compute.k ) d.x. .I)k + (.x·1 . . d. . .{OJ. Jae PROOF Most of the proof involves the special case where w is a (k . . . dxk /(. . • . l . • • . . I . . 4. . a little notation translation shows that J[o. . . and a. x. THEOREM (STOKES' THEOREM).1 ) 1+ 1 [O. First. w is a sum of (k . . . . .x·1 .k if j i. . . /\ dxk ) if j i' i = [0. xk ) dxl . 1)" I(x. Although it is intuitively clear that (" is not the boundary of a 2chain in IR 2 . dx·k � .I)forms of the type and it suffices to prove the theorem for each of these. If w is a (k . . .
dxk [O. . . For an arbitrary singular k cube. . xk ) dx' . O. . /\ dxi /\ " . xk ) dx ' . . 1tk Jatk c'w = [ w . ) . Xk ) . . . /\ dxi /\ . . . . . dxk = (_) 1. . dxi . )k + ( )' f l(x ' . /\ dxk ) J . .1 f . )k Thus .. . . . . . . . [O. . 0. . . . (1' D (x' .. . . dxk ' ' = () 1 . . chasing through the definitions shows that [ dw = [ W. . . . . . . Chapter 8 [ d(J dx ' }lk /\ . . . . /\ = = f Di 1 dx' dx. I ' I(x . dxk . . )k By Fubini's theorem and the fundamental theorem of calculus we have [ d(J dx ' 1tk = (_I)i. . .' l' . [O. . latH 1tk [ w = [ c·w. . . Jac Jatk Theref ore The theorem clearly follows for k chains also . . •:. dxi . . . .1 1 · · · 1 [/(x'. . .. xk ) ] dx ' . lc dw = [ c'(dw) = [ d(c'w) = [ Jtl.' f DJ. ) . . lac . /\ dxi /\ . . . /\ .(x ' . xk) dXi) dx' . )k (_I )i . /\ dxk ) /\ .254 On the other hand. . . . . . . /\ dx k [O. .
{OJ > jR. For example. For. As an application of Stokes' Theorem. but actually becomes that theorem when c = / 1 and w = f. then we would have d(de) = 0 = o.£) .x2 + y 1c de = 1c y2 dx + 2 dy x2 + y 1c Jac2 1 c 1 c2 "de" really c l (£.e(£) > 21f as £ > 0. although closed. then we would have 21f = \Ale were previously able to give a simpler argument to show that "de" is not exact.([0. I £) f de = e(l . we show that the curve c : [0. Tfwe did have c = ac 2 . but Stokes' Theorem is the tool which will enable us to deal with forms on jRn {OJ. sin 21f1).e(£) . de = [ de = But a straightforward computation (which will be good f the soul) shows that or = 21f.] Although we used this calculation to show that c is not a boundary.  1c 1c dl w= = [ I= Jac 1 0 I = o.£) . using the fact that is de f e : jR 2 . if we had w = dl for some Coo function I : jR 2 . 00) x {OJ) > jR: We have or x . we wiII eventually obtain a 2form w on jR3 . and e(1 . we could just as well have used it to show that w = "de" is not exact.integration 255 Notice that Stokes' 1l1eorem not only uses the fundamental theorem of cal culus. [There is also a noncomputational argument. 1] > jR 2 _ {OJ defined by C(I) = (cos 21f1.{OJ. w= x dy /\ dl Y dx /\ dz + z dx /\ dy (x 2 + y2 + z2) 3/2  . is not ac 2 for any 2chain c 2 .
THEOREM. I]") . If w is an nf orm 011 M such that support w c [0.l a c) )' 2: 0. the reason will be clear from the next result. where c is a singular ncube of this sort. There are lots of ways of doing this.l W• CI C20{c2.. for singular ncubes c : [0. In fact.([O. because of the fact that support w C (' . To prove that w is not exact we will \'vant to integrate it over a 2chain which "fills up" the 2sphere S' C JR 3 . since ('1 and ("2 are both orientation preserving). I]") n c. C2 then PROOF We want to use Corollary 2. 11" (it does satisfy det{c2 . and ('.1 OC] ) w= . I ]" 5. Assume that c.lion jm!servillg (with respect to the orientation jJ. : M be two singular ncubes which can be extended to be diffeomor > c. we first want to describe a way of integrating nforms over nmanifolds. I]" ) .([O.0 (' .256 Chapter 8 which is closed but not exact. ([0. phisms in a neighborhood of [0. and the usual orientation on JR"). w. I]" ) n c. will be denoted by > 1 with sup If w is an arbitrary nform 011 M. which is basic for our definition. Let M be an nmanifold with an orientation {/" and let c" c. However. is not defmed on all of [0. I]" M ('([0. but they all turn out to give the same result. .{OJ. a glance at the proof of Corollary 2 will show that the result still follows. For the moment we are keeping the origin of w a secret. are both orienlo. ([0. l C] w. I]"). but a straightforward calculation shows that dw = 0. •:. The common number 1 w=l C2 1 w . This is possible only when M is orientable. each contained in some ('([0. and write port w C The only problem is that c. on M. I]") and c orientation preserving. then there is a cover I!J of M by open sets U. IJ". then . if <I> is a partition of unity subordinate to this cover.
of M we clearly have IM. /\ dX" = f f.!') W .. If M C IR n is an ndimensional manifoldwithboundary and J : M > IR has compact support. M . and w is an nform with compact support on N.!') W M where the right hand side denotes the ordinary integral. then if J is orientation preserving if J is orientation reversing. Likewise. . We wish to define 257 We will adopt this definition only when w has compact support. fM Jdx' /\ . . which form a locally finite collection. This is a simple conse quence of Proposition 1.) \t\'ith minor modifications we can define J w even if M is an nmanifold withboundary. (We really should denote this sum by for the OIientation  {/.1M.1nlegralion is defined for each ¢ E <1>. E [ so that our definition does not depend on the partition. then these sums are all finite. in which case the sum is actually finite. Ifwe have another partition of unity \jJ (subordinate to a cover <9'). and the last sum can clearly also be written as [ 1M W = ¢Lc'pJM ¢ · w. then IM. since support w can intersect only finitely many of the sets {p : ¢(p) i' OJ..!')W' However) we usually omit explicit mention of p. if J: M n > N n is a diffeomorphism onto.
i. 10calIy. /\ dyn J) = g: (g I ( a(y'o J) ) 1 0 f) . Howeve. 77'. then for nonnegative V > ]R we have J*{g Idy ' J(p) PROOF Go through the proof o f Theorem 77. pUlling i n absolute value signs in the right place . we can define w = Idx ' /\ . n ( Mp) . consider the Mobius strip M. One way of obtaining a volume element is to begin with an IIform � and then define w(p) = 1 � p) l . there is a way of discussing integration on nonorientable manifolds. U) is a coordinate system around p E M. But this cannot be true on all of M. w is of the form w = I�I for an l1form � . . Theorem 77 has an obvious modification for volume elements: w(p)(vp. Suppose that w is a function on M such that for each p E M we have for some �p E g.e. . wp) = area of paralIelogram spanned by v and w. . ( It is not hard to see that w is a volume element. If J : M > N is a Coo function between l1manifolds. not every volume element arises in this waythe form 1}p may not vary con tinuously with p. Vn E Mp we have Sudl a function w is called a volume element on each vector space it deter mines a way of measuring IIdimensional volume (not signed volume). .. /\ dx"l· . . then on U we can write we calI w a Coo volume element if is COO. THEOREM. •:. /\ . /\ dxnl J J for J ?: 0. If (x. For example. .r · Idx ' /\ . V) a coordinate system around q = E N. det a. imbedded in ]R 3 Since Mp can be considered as a subspace of ]R3p. for any n vectors VI . . . U) is a coordinate system. (x. ..258 Chapter 8 Although IIforms can be integrated only over orientable manifolds. and (y. since there is no IIform � on M which is everywhere nonzero. .
integration
259
then
78'. COROLLARY. If (x, U) and (y, V) are two coordinate systems on M and g Idy l /\ . . . /\ dY"1 = h Idx l /\ . . . /\ dx" l g, h ? 0
[This corollary shows that volume elements are the geometric objects corre sponding to the "odd scalar densities" defined in Problem 41 0.]
It is now an easy matter to integrate a volume element w over any manifold. First we define
1[0,1}11
[
w=
[O,I}"
f
J for w = J Idxl
>
Then for an nchain c : [0, I]"
Theorem 77' shows that Proposition I holds for a volume element w = J Idxl /\ . . . /\ dx"l even if det c ' is not � O. Thus Corollary 2 holds f volume elements or even jf det p' is not � O. From this we conclude that Theorem 5 holds for volume elements w on any manifold M, without assuming Cl , C2 orientation preserving (or even Ulat M is orientable). Consequenuy we can define JM w for any volume element w with compact support. Of course, when M is orientable these considerations are unnecessary. For, there is a nowhere zero nform 1] on M, and consequently any volume element w can be vvritten w = JI� J, J ? O. If we choose an orientation jJ., for M such that W(VI , " " v,J positively oriented, then we can define
>
l
e
M we define
/\ . . . /\
dx"J, J ? O.
w=
J[O,1}rI
[
c'w .
0 for V I , . . . , V
n
1M
[ w= [
1(M,,,)
h
Volume clements will be important later, but for the remainder of this chapter we are concerned only with integrating forms over oriented manifolds. In fact) ) our main result about integrals of forms over manifolds ) an analogue of Stokes Theorem about the integral of forms over chains) does not work for volume elements.
260
Chapter 8
Recall from Problem 316 that if M is a manifoldwithboundary, and p E then certain vectors v E Mp can be distinguished by tbe fact that for any coordinate system x : U > IHJ n around p, the vector x.(v) E IHJ nf(p) points "outwards') . Vile call such vectors v E Mp "outward pointing". If M has an
aM,
orientation {/" we define the induced orientation a{/, f aM by the condition that or [vl , . . . , vn _d E (a{/,)p if and only if [W, V I , . . . , Vn _ l ] E {/,p for every outward pointing W E Mp. If {/, is the usual orientation of IHJ n , then for p = (a , O) E IHJ n we have
{/,p = [(el lp, . . . , (en )p]
=
=
(_1)"  1 [(en )p, (el lp, . . . , (en llp] (  I ) n [( en )p, (el lp, . . , , (en l lp].
Since (  en )p is an outward pointing vector, this shows that the induced orien tation on IRn 1 x {OJ = alHJ n is (_ I) n times the usual one. The reason for this choice is the following. Let c be an orientation preserving singular ncube in (M, {/,) such that aM n c([O, 1]") = C(�,o) ([O, 1]"1). Then c(n ,O) : [O, l]n1 >
(aM, ap.,) is orientation preserving for even n, and orientation reversing for odd n. If w is an (n  I )fol'm on M whose support is contained in the interior
llllegratio11
261
of the image of c (this interior contains points in the image of C(n ,O)), it follows that
Jell/.m
But
[
w = (l)" [ w .
JaM
n c(n,O) appears with coefficient (_1) in ac. So
Jae
[ w= [
J<I)II C(II ,1ll
n w = (_1) [
Jell/.Ill
w = [ w.
JaM
Ifit were not for this choice of in the following theorem.
aft we would have some unpleasant minus signs
6. THEOREM (STOKES' THEOREM). If M is an oriented ndimensional manifoldwithboundaIY, and aM is given the induced orientation, and w is an (/l  I )form on M with compact support, then
JM
[ dw = [ w . JaM
PROOF Suppose first that there is an orientation preserving singular ncube c in M  aM such that support w C interior of image c . Then
[ dw = [ dw = [ w 1M Jc Jae =0
while we clearly have
by Theorem 4 since support w C interior of image c ,
Suppose next that there is an orientation preserving singular ncube c in M such that aMn c ([O, I J ") C(n, O)([O, 1]" 1 ), and support w C interior ofimage c. = Then once again
lM w = O.
[ dw = [ dw = [ w = [ w by (*). 1M Jc Jae JaM
262
ChapI,," 8
In general, there is an open cover I!J of M and a partition of unity <I> sub ordinate to I!J such that for each ¢ E <I> the form ¢ " w is one of the two sorts already considered. We h ave
0 = d(l) = d I »
¢E 4>
so
(
=
¢E 4>
L d¢,
Since w has compact support, this is really a finite sum, and we conclude that
¢E4>
L d¢ /\ w = O.
Therefore
1M dw = L 1M ¢ . dw = L 1M d¢ /\ w
¢E 4> ¢E 4> ¢E 4> 1M ¢E 4> laM
+¢
. dw
= L [ d(¢ ' W) = L [ ¢ · w = [ w. ·:·
laM
One of the simplest applications of Stokes' Theorem occurs when the oriented Ilmanifold (M,!,) is compact (so that every form has compact support) and aM = 0. In this case, if � is any (/1 I )form, then
[ d� = [ � = O. 1M laM
Therefore we can fwd an nform w on M v.lhich is not exact (even though it must be closed, because all (Il + I )fOrms on M are 0), simply by finding an w with Such a form w always exists. Indeed we have seen that there is a form that for VI , . . , VII E Mp we have

.
L w i' O.
w such
w(v] , . . . , v,,» O if [v] , . . . , v,,] = !'p.
If c; [0. 1]" > ( M . ll ) is orientation preserving. then the form c'w on clearly for some g > 0 on [0, I] ".
[0, I]" is
inl£gralion
263
so Ie W > o. It follows that 1M W > o. There is, moreover, no need to choose a form W with H holding evelywherewe can allow the > sign to be replaced by 2: . Thus we can even obtain a nonexact nform on M which has support contained in a coordinate neighborhood. This seemingly minor result already proves a theorem: a compact oriented manifold is not smoothly contractible to a point. As we have already empha sized, it is the "shape" of M, rather than its "size", which determines whether or not every closed form on M is exact. Roughly speaking, we can obtain more information about the shape of M by analyzing more closely the extent to which closed forms are not necessarily exact. In particular, we would now like to ask just how many nonexact nforms there are on a compact oriented Ilmanifold M. Naturally, if w is not exact, then the same is true for w + d� for any (n  I )form �, so we really want to consider w and w + d� as equivalent. There is, of course, a standard way of doing this, by considering quotient spaces. We will apply this construction not only to Ilforms, but to forms of any degree. For each k, the collection Z k (M) of all closed kforms on M is a vector space. The space B k (M) of all exact kforms is a subspace (since d 2 = 0), so we can form the quotient vector space H k (M) = Z k (M)/B k (M); this vector space Hk (M) is called the kdimensional de Rham cohomology vector space of M. rde RJzam � T heorem states that this veClor space is isomorphic to a ccrtain vector space defined purely in terms of the topology of M (for any space M), called the "kdimensional cohomology group of M with real coef ficients"; the notation Z k , Bk is chosen to correspond to the notation used in algebraic topology, where these groups are defined.] An element of H k (M) is an equivalence class [w] of a closed kform w, two closed kforms w, and W2 being equivalent if and only if their difference is exact. In terms of these vector spaces, the Poincare Lemma says that H k (JRn ) = 0 (the vector space containing only 0) if k > 0, or more generally, Hk (M) = 0 if M is contractible and k > o. To compute HO (M) we 110te first that B O (M) = 0 (there are no nonzero exact Oforms, since there are no nonzero (  I )forms for them to be the dif ferential of). So H O (M) is the same as the vector space of all Coo functions I : M > JR with dl = o. If M is connected, the condition dl = 0 implies that I is constant, sO HO (M) '" K (In general, the dimension of H O (M) is the number of components of M.) Aside from these trivial remarks, we presently knO\"/ only one other fact about Hk (M)if M is compact and oriented, then Hn (M) has dimension 2: I. The further study of Hk (M) requires a careful look at spheres and Euclidean space.
264
On S,,l C JR"  {OJ ISIII (1' > 0: for (vJ)P ) "
Chapter 8
there is a natural choice of an (n  I ) form (J' with " (Vnl )p E snlp , we define
Clearly this is > 0 if (V l )p , . . . , (V )p is a positively oriented basis. In fact, ,,l we defined the orientation of sn l in precisely this waythis orientation is just the induced orientation \'vhen sn l is considered as the boundary of the unit ball {p E JR" : ipi :0 Ij with the usual orientation. Using the expansion of a determinant by minors along the top fOW we see that (1' is the restriction to s n l of the form (J on JR" defined by
The form (J' on S,,l will now be used to find an (n  I )form on JR"  {OJ which is closed but not exact (thus showing that H,,l (JR"  {OJ) i' 0). Consider the map r : JR"  {OJ > S,,l defined by
n (J = L (  I ) i l X i dx 1
i=]
/\
. • .
/\
dxi /\ . . . /\ dx".
r(p)
Clearly r(p) = inclusion, then
=
TPi
p
=
p v(p) '
> JR" 
p if p
E
S"l; otherwise said, if i : S,,l
r 0
{OJ
is the
(In general, if A C X and r : X > A satisfies r(a ) = a for a E A, then r is called a retraction of X onto A.) Clearly, r* (1' is closed:
i = identity of S"l .
d(r*(J') = r*d(J' = O.
However, it is not exact, for if "*(1' = d1], then
(1' = i*r*a' = di*1J;
but we know that (1' is not exact.
inlegmlion
It is a worthwhile exercise to compute by brute force that for n = 2, for n = 3,
265
 dx x dy  y dx r *(J' = x dy + y = = de x 2 y2 v2 r*(J' =
1
Since we will actually need to know r*(11 in general, we evaluate it in another way:
= ;;; [x dy A dz  y dx A dz + z dx A dy].
x dy A dz  y dx A dz + z dx A dy (x 2 + y 2 + z 2 ) 3 /2
7. LEMMA. If (J is the form on IF!.n defined by
(J =
L (I);I x; dx l A · · · A d;! A · · · A dxn,
i= ]
n
and (1' is the restriction i*a of (J to snl) then
r*(J'(p) =
L i=]
(J(p) . I pln
So
r*(J' = �
VII
"(I);IX; dxl A · · · A d;! A · · · A dxn. 
n
PROOF. At any point p E IF!.n {OJ, the tangent space IF!.np is spanned by Pp and the vectors up in the tangent space of the sphere sn l (l pl) of radius I pl . So it suffices to check that both sides of (*) give the same result when applied to 11  1 vectors each of which is one of these two sorts. Now Pp is the tangent vector of a curve y lying along the straight line through 0 and p; this curve is taken to the single point r(p) by r, so r (pp ) = o. On the other hand,
*
0(,> (" , ( ."" . . . , (�,),) = de.
(,fJ
= o.
So it suffices to apply both sides of (*) to vectors in the tangent space of snl (Ipl). Thus (problem 15), it suffices to show that for such vectors up we
266
Chapter 8
But this is almost obvious: since the vector vp is the tangent vector of a circle y lying in sn l (Ipl), and the curve .. 0 y lies in snl and goes I II pi as far in the Same time . •:.
J: B
and define
8. COROLLARY (INTEGRATION IN "POLAR COORDINATES"). Let > JR, where B = {p E JRn : Ipl :0: I),
g:
snl
>
JR b,'
g
(p) =
lo' unl J(u · p) duo
I]
>
[ J = [ J dx ' /\ . . . /\ dxn = [ g(J' . }SIII JB JB PROOF. Consider snl X [0, I] and the two projections
1f l : S,, l 1f2 : snl
x [0,
X
Then
[0, J]
>
sn l [0, I].
[0, I]
Let us use the abbreviation
I
....!2
�
1'" c=::>
snl
From this it is easy to see that if we define h : snl
If (y, U ) is a coordinate system on sn l ) \"ith a corresponding coordinate sys tem ()' , I) = (y 0 1f , 1f2) on snl X [0, I], and (J' = Ci dy ' /\ . . . /\ dyn ' , then l clearly , ' (J A dl = Ix 0 1f1 dy ' /\ . . . /\ dyn /\ dl.
h(p, u) = unl J(u . p),
then
X
[0, J]
>
JR by
Now we can denne a diffeomorphism ¢ : B  {OJ ¢ (p) =
}SIII g(J
[
'
=
/w' A dl . ( _ I) nI [ 1s"1 x[O,I J
>
snl
(,, (p), v(p»
=
(pll pl, Ipl).
X
(0, I] by
bzle.gralian
267
Then
¢*«J' A dt) = ¢*(1C , *(J' /\ 1Cz*dt) = ¢*1f ] *a' 1\ ¢*1f2*dt = (1CI O ¢)*(J' /\ (1C2 0 ¢)*dt = r *a' 1\ v*dt n Xi i I n i = ;;n L (  I ) i  J )C i dx I /\ · · · /\ dx /\ · · · /\ dxn /\ L ; dx ,= 1 1=] ( _ I )n ' n . 2 ' . . . = '"'(X') dx /\ /\ dx n Vn + 1 � i=]

(
)
Hence
¢*(lw ' A dt) = (h ¢)¢*«J ' A dt)
( 1 )"' = vn 1 J . ;,;::r dx 1
= °
( _ I ) n ' J dx ' /\ . . . /\ dxn.
1\ . . . 1\
dxn
So,
iB
[ J dx ' /\ . . . /\ dxn = ( _ 1 )"1 [
= ( _ I ) n' [
iB{a)
¢*(h(J' A dt) }l(J' A dt
(This last step requires some justification, which should be supplied by the reader, since the forms involved do not have compact support on the mani folds B  {OJ and sn ' x (0, I] where they are defined.) .:. We are about ready to compute H k (M) in a few more cases. We are going to reduce our calculations to calculations within coordinate neighborhoods, which are submanifolds of M, but not compact. It is therefore necessary to introduce another collection of vector spaces, which are interesting in their own right.
= L,,, g(J'.
)snIx{o,l}
268
Chapter 8
Z� (M) is the vector space of closed kforms with compact support, and B�(M) is the vector space of all kforms d� where � is a (k  I )form with compact support. Of course, if M is compact, then H;(M) H k (M). Notice that B� (M) is not the same as the set of all exact kforms with compact support. For example, on IRn, if J 2: 0 is a function with compact support, and J > 0
where
=
The de Rham cohomology vector spaces with compact supports defined as H;(M) = Z;(M)/B�(M),
Ii; (M) are
at some point, then
w = J dx'
/\ . . . /\ dxn
is exact (every closed form on IRn is) and has compact support, but w is not d� for any form � with compact support. Indeed, if w = d� where � has compact support, then by Stokes' Theorem
This example shows that H;(IRn) i' 0, and a similar argument shows that if M is any orientable manifold, then H;(M) i' o . We are now going to show that for any connected orientable manifold M we actually have
L w L d� faR" �
= =
= o.
H;(M) "" R
This means that if we choose a fixed w with f w i' 0, then for any nform w' with compact support there is a real number a such that w' aw is exact. The number a can be described easily: if
M
w'
then so


[ [ ' [ 1M w' 1M aw 1M d�
=
GW
=
d1].
=
0,
H;(M) '" IR is equivalent to the assertion that
[w] >>
the problem, of course, is showing that 1] exists. Notice that the assertion that
is an isomorphism of H;(M) with �, i.e.) to the assertion that a dosed form w with compact support is the differential of another form with compact support if f w = o.
Lw
M
then H.(_ J)i' (10 . (3) If the theorem is true for JR". dl 0 outside some interval [N.J. Moreover.n. N) and a constant Cz on (N. we have an explicit formula for �.I)manifolds.c has compact support.I )form � o n JR " such that w = d�. . p) dl xi dx' /\ .c) Step 2. ) . Let w = 1 dx' /\ . If M is a connected orientabIe nmanifoId. . . /\ dx". In fact. Therefore c] = Cz = c and we have w = d(J where 1 .(M) '" IF!.i /\ . /\ . PROOF. Step I . N]. Since support w is compact. .269 9. . then it is true for any connected oriented nmanifoId. then it is true for ffi. There is some function 1 (not necessarily with compact support) such that w = dJ. /\ dx" be an nform with compact support on JR" such that III" w = o. �(p) " = {. in particular for S". (2) If the theorem is true for (n . 1"' 1(1 . from Problem 723. oo).. THEOREM. We know that there is an ( n .. Let w be a I form on JR with compact support such that III w = o. For simplicity assume that support w e {p E JR" : ipi < I ) . We wiII establish the theorem in three steps: = f N constant c] N on ( 00.' . . . so 1 is a (I) The theorem is true for M = IF!.
1 f ( u . Let h : [0.is for Stefl 2. . II this becomes . g(J' = dA for some Hence (11 . /\ dxn L i=] = (t i r l f (u · � ) dU) . Then hr* A is a Coo form on and � = r*(dA) = d(r*A). I] be any Coo function with h = I on A and h = 0 in a neighborhood of o. . by the hypothe. g(J' = snl. r f dx1 /\ .2)form A on snl . . .. .d(hr*A» : .270 Using the substitution Chapter 8 Define On the set or u = I pl l �(p) = ( ripi un1 f (u I pl ) dU) �n I pl Jo n (_J )il dx1 /\ . r*(J'(p). . r*(J' = r*(g(J').. Thus.II JB w o. X Xi by Lemma 7. IR" > IRn w = d� = d(� . by Corollary 8 we have for the (11 . > E > I I we have so on we have Moreover. � ) dU ) .I )form g(J' on 1s"1 r � = (g 0 r ) . r*(J'( p) II g: snl IR by g(p) = 10' u n1 f(u· p) duo A = {p IRn : I pi f = 0. A �(p) = (10' u n . /\ dxn = = r JR. !!. /\ dxi /\ .
) = o.C1 W = dl}l Wz . with ¢. .:.CrWr. it is easy to see that there is a sequence of open sets U = V" . we want to show that there is a number c and a form � with compact support such that w' = cw + d�. have compact support (C V. with Vi n Vi+ 1 =j:. If w' is any other nform with compact support. we can assume w' has support contained in some open V C M which is diffeomorphic to IRn. .). since on A we have � . o.) = � .n we h ave WI . .d(hr*). It obviously suffices to find c. In other words. Step 3.l1. Choose an IIform w such that JM w i' 0 and w has compact support contained in an open set U C M.d(hr*).gral. Vr = V diffeomorphic to ffi.CZWI = dI}2 w' . Using the connectedness of M.lille. . we can write where each ¢iW' has compact support contained in some open set Vi C M with U. with U diffeomorphic to IRn .l = d1Jr : where all �. From this we clearly obtain the desired result. 0.d(r*).w' = ("jW + d1Ji' for each i.) has compact support. Since we are assuming the theorem for ffi.oll 27 1 the form � . and �. diff eomorphic to IRn . Choose forms Wi with support Wi C Vi n Vi+ l and JVi Wi =j:. . Using a partition of unity.
n orientation preserving. U4.(M) = o. we have Iv. JVI [ Wi /J� [ Wi . Xi) where each Xi 0 Xi+ l . UJ . there is an infinite sequence U = UJ . such that fu w i' 0 (this integral makes sense. • . THEOREM. Choose the forms Wi in Step 3 so that. It obviously suffices to show that w = d� for some form � with compact support. THEOREM. U2. Consider a sequence of coordinate systems ( Vi. : PROOF Consider first an Ilform w with support contained in a coordinate neighborhood U which is diffeomorphic to JRn .1 is orientation rel1ersillg. then Hn(M) = O. •. I I . I t follows that Wi = cw + d� where c > O. Wi > 0. there is such a sequence where Vr = VI but Xr OX1 . then also !Vi+1 Wi > O. . Choose an Ilform W with compact support contained in an open set U diff eomorphic to JRn . O . then PROOF. Consequently.l are positive. If M is any connected nonorientable nmanifold. If M is a connected noncompact nmanifold (orientable or not). . Since M is not compact. since U is orientable). we have W = cw + d� for for c  >0 J so (c  We can also compute Hn(M) for noncompact M.1 is orientation preserving. the numbers Ci = H. Taking w ' = W.272 Chapter 8 The method used in the last step can be used to derive another result. Now if M is unorientable. using the orientation of Vi which makes Xi : Vi 7 ffi. 1 0. i)w = d� c =j:.
d�2 + C. but the sequence is still eventually outside of any compact set).+. V3 . C2C3W3 = where the right side makes sense since the V. has support contained in V. + C. Let {¢u) be a partition of unity subordinate to !9. such that w Wi Then = = c. = w = d�.C2C3d�4 + .. V. Now it can be shown (Problem 20) that there is actually such a sequence V.c2d�3 + C. . . where �. whose union is all of M (repetitions are allowed. + c. i' o. c2d�3 + c. then for each V. we have W d�. V2 . If W is an nform 011 M. may intersect several Vj for j < i. .w.'s. . d�. . and V. and forms �. are eventually outside of any compact set. d�2 + c. d�2 + c. Since any point P E M is eventually in the complement of the V. we have seen that ¢u. V. + d�.+. n V. . C2W2 = d�. with compact support . . C V. i' 0. + C. and such that the sequence is eventually in the complement of any compact set. Cj+lWi+l + d1]i+l i 2: J .. + C.+2 Hence . There are constants C.Integra/ion 273 of such coordinate neighborhoods such that V. Now choose nforms Wi with compact support contained in Vj n Uj+h such that Ju W. U U U .w = d�. The cover !9 = {V) is then locally finite. . w.
with Cith component equal to c. then I*w is also closed (dl*w I*dw = 0). On the other hand. so 1* takes Z k (N) to Z k (M). consider the case k o. Then I*(c) = c o l is also a constant function. Ci E A .(M) '" {� Hn Hn(M) '" { c 0 HO(M) '" IR if M is orientable if M is nonorientable if M is compact if M is not compact. I f M i s disconnected. where each IR. then 1* : HO(N) > HO(M) is just the identity map under the natural identification of HO(N) and HO(M) with IR. For example.1 (IRn .274 Chapter 8 SUMMAR OF RESULTS Y (I) For IRn we have (2) If M is a connected nmanifold. We also know that Hn. If M is connected. fJ E B. This shows that 1* induces a map = also denoted by 1*: /* : Hk (N) > H k ( M ) . since we will eventually improve it. = . ctEA PEB takes the element {cp) of EBpEB IRp to {c�}.. In order to proceed further with our computations we need to examine the behavior of the de Rham cohomology vector spaces under Coo maps I: M > N. If N is also disconnected. then H. since I*(d�) = d(J*�). then HO( M ) i s isomorphic to the direct sum E9 IR. If w is a closed kform on N. If N is connected. with components M. with components Np. .EA E9 IRp > E9 IR. then HO(N) is just the collection of constant functions c : N > IR.{OJ) i' 0. but we have not listed this result. 1* also takes B k (N) to B k (M).. then = cp when I( M. where c� /* : the map 1* takes C E IR into the element of EB IR.) C Np. '" IR.
There is when no natural way to make H"(M) isomorphic to JR.on 275 A more interesting case. L J*w = a .l. J. it is almost unbelievable that this number is alwqys an integer. so we really want to compare M J*: H" (N) L J*w > H"(M) and Wo for an nform on N. but is proper (the inverse image of any compact If set is compact). then we have a map M a = J.Integral. Until one sees the proof of the next theorem. let For each P E sign J: M. and a number deg such that for all forms on N with compact support. J M) and JR (and similarly for N) it J.p: Mp J pJ= J !*p w L J*w = (deg /) Lw { if + Nq is orientation preserving (using the orientations J. form we have w fM w w 1M J*wo = a · L woo with fN Wo Lw i' o. Let q E N be a regular value of I (q). ) and (N. M J1 !. Let oriented nmanifolds ( J. Choose one number such that w a is an isomorphism of H " ( Since r> follows that for eVeI). THEOREM. L w. is the map and N are both compact connected oriented nmanifolds. v). if Mp . > N be a proper map between two connected 1 2. Then there is some deg which depends only o n i s called the degree o f I The number and N are not compact.P for and Vq for Nq) is orientation reversing.l. and the only one we are presently in a position to look at.
I .I (q ) is finite. We want to choose a coordinate system (V.(V. and the Vi are disjoint. . .J(W'). u . Then support J* w C V.I] tIllS proves the theorem . first choose a compact neighborhood W of q. Moreover. . So Since J is a diff eomorphism from each Vi to V we have L J* w = {. . Vk . since it is compact and consists of isolated points. Let J' (q ) = { p " . Then J(W') is a closed set which does not contain q. xi i around Pi such that all points in Vi are regular values of J.' (q) PROOF Notice first that regular values exist. . To do this. . We can therefore choose V C W . · · · Vk . pd· Choose coordinate systems (Vi . Now choose w on N to be w = g dy ' /\ . U U U [ J* w = [ w if J is orientation preserving 11' lUi = Since J is orientation preserving [or reversing] precisely when signp J = I lor . . . This ensures that J' (V) c V. so the sum above is a finite sum.276 Then Chapter 8 L signp J (= 0 if r ' (p) = 0). ii J*w k U · · · U Vk !. •:. Iv w if J is orientation reversing. /\ dyn where g 2: 0 has compact support contained in V. · · ·UVk • Finally. by Sard's Theorem. PEj. redefine Vi to be Vi n J' ( V). y) around q such that J' (V ) = V. and let deg J = U W' C M be the compact set w' = J' ( W) .
We have already seen that A is orientation preserving or reversing at all points. I ] > N the map H is called a (smooth) homotopy between J and g.inle. there is a smooth map H : M x [0. H: M with x [0. Then J = H o io g = H o i) . Recall that for every kform w on M x [0. O) = J(p) H(p. Two functions J. We used this fact to show that all closed forms on a smoothly contractible man ifold are exact. J] = g* .I )form /w on M such that i l*W . we compute the degree of the "antipodal map" A : S" > S" defined by A (p) = po . depending on whether II is odd or even.io*w = d(Iw) + /(dw). J* PROOF By assumption.*(H*w) . Any element of H k (N) i s the equivalence class [w] of some closed kform w on N. N with g*w . then the maps /* : H k (N) H k (M) g* : Hk (N) H k (M) > > > > are equal.io*(H*w) = d(IH*w) + /(dH*w) = d(IH*w) + 0. Since A .l (p) consists of just one point. •. 1 3.1.1 ) " . N are smoothly homotopic. We can HOW prove a more general result. g: M H(p. THEOREM. but we need to intro duce another important concept first. we conclude that deg A = (. But this means that g*([w]) = J*([w]) : .gmlioll 277 We can draw an interesting conclusion from this result. g : M > N between two Coo manifolds are called (smoothly) homotopic if there is a smooth function As an immediate application of the theorem. I ) = g(p) for all p E M. J] we defined a (k .(H 0 io)*w = i. Notice that M is smoothly contractible to a point po E M if and only if the identity map of M is homotopic to the constant map po . If f.J*w = (H o i) *w . .
1 4. then we can construct a homotopy between A and the identity map as follows. •. For another application of TIleorem 13. . COROLLARY. If vector field on S" . COROLLARY. X"+ l ). . For H (p. X"+l ) E S" we define = For n odd we can explicitly construct a nowhere zero vector field on S" . If M and N are compact oriented nmanifolds and the maps j.J)".) The vector field on S" can then be used to give a homotopy between A and the identity map. then r 0 i : sn. Define S" is ( . (On S l this gives the standard picture. . then deg j = deg g. consider the retraction r: JR" . r (p) = p/ ipi.l .l > JR" . .l 7 sn. . and therefore in S"p .{OJ is the inclusion. . Il is even. For each p.278 Chapter 8 1 5 .l is the identity of S". there is a unique great semicircle Yp from p to A (p) = p whose tangent vector at p is a multiple of X(p). I ) = Yp (I ) : . . p (Xl . Since the identity map has degree ) . g : M > N are homotopic. .{OJ > S". X2 . I . this is perpendicular to p = (X l .l . then there does not exist a nowhere zero PROOF We have already seen that the degree of the antipodal map A : S" > A is not homotopic to the identity for n even.l If i: S". But if there is a nowhere zero vector field on S".
We need one further observation. = "r( p) (t=O) p(/=1 1 A retraction with this property is called a deformation retraction.I)r(p) E jRn .{OJ) for all k. Thus In particular. We are now going to compute Hk (jRn . A generator of Hn l (jRn .{OJ i 0 I" ( p) = p/ i p i H(p. l ) Ip + ( J . The manifold for all I . we can define the homotopy H by . we have and 1" ' 0 i' = (i 0 1" ) ' i ' 0 1" ' = ( I" 0 i)' = = identity of Hk( sn.{OJ) is the closed form r*(1'.I ) . the maps (r o il' and (i 0 1" ) ' are the identity. M x {OJ c M x jR' is clearly a deformation retraction of M x jRt So Hk(M) "" Hk(M x jRl ) . but it is homotopic to the identity.{OJ) "" jR .{OJ. Thus.{OJ > jR" . we have Hnl (jRn . identity of Hk(jR"  {OJ) So i* and r* are inverses of each other.integration 279 The map is. of course. for the case of s nI C jRn .// ' i 0 1" : jRn . Whenever I" is a deformation retraction.{OJ. not the identity.
O) x [O.I we have H k (w.0.1) Since A and B are both starshaped (with respect to the points (0. using A = w.4. there is a similar argument.{OJ) = H k (snl) = 0.{(O. Th us w is exact. . 0.4 .{all r B so clearly JA . there are Oforms JA and Is on A and B with w = dJA l w=dB Now d(JA on A on B.3 . 0]) B = w. Let w be a closed I form on w. oo)}. THEOREM. 0.{(O. for on A on B A w = d(J . If w is a closed I form on ffi.3 . We claim H I (w. 0) x (00 .c = Is on A n B. 0]) [0. For ° < k < /1 .4 .{(O.2 .3 Let A and B be the open sets (0. . .n .O) B w. (0. 00») .4 .280 pter 8 Cha 1 6.O.0. n= PROOF. If w is a closed 2form on w.3 .c) B) w = d(J and JA . I) A = w. on A n B.{ (O.0. I) and (0.I). 0) = x x (00.I is a constant on A n B. respectively). Induction on n.{O) = 0. then we obtain Iforms �A and �B with w = d�A w on A = d�B on B. x w" B) I =° and A n B = [ w. The first case where there is anything to prove is 3 .
The general inductive step is similar.d(<PBA) = �A .d(<PBA) on A on B. To circumvent this difficulty. <PBj for the cover { A . if <PBA denotes { <PBA O and similarly for <PAA.integration 281 Now d(�A . Clearly.{OJ = A on A.�B = dA for some Oform A on A n B. So we can define a Coo form on IR n . since this is not defined on A . = d�B /\ A U B by letting it be �A .{OJ: <PA + <PB d<PA + d<PB support <PA support<PB CA C B.(A n B). and �B + d(<PAA) on B. on A n B =) =0 Now. •:. on A n B on A . Bj of IR J .dA + d(<PAA) is a COO form on A is a Coo form on B. we cannot simply consider �A . = �A + (<PA .d(<PBA» = d(�B + d(<PAA» so w is exact. .d A.J) dA + d<PA = �B + d(<PAA).<PB dA . then <PBA <PA A On A n B we have �A . w = d�A = d(�A .d<PB /\ A = �A . note that there is a partition of unity {<PA .�B) = 0 and SO �A . Unlike the previous case.
.. 1 7 .n > [0. where 2B denotes the ball of twice the radius of B.2B.282 Chapter 8 We end this chapter with one more calculation.n with compact support. Let w be a kform on w.{OJ and k  I < 11 . the form � ..n.I we have from Theorem 16 that � = dA for SOme (k . which we will need in Chap ter 1 1. I] be a Coo function with / = 0 in a neighborhood of B and / = I on w.. 0 < k < n.2)form A on A .n and w = d� = d(� .n .. THEOREM.n . Let B be a dosed ball containing support w.n) = 0 is left to the reader.. Since A is diffeomorphic to O = w = dry w.. We know that w = d� for some (k . Then on A = w. •: . Let / : w.n) = o.d(JA» ..B we have d� = O.. PROOF The proof that H�(w.n .I )form � on w. Then d(JA) makes sense on all of w. For 0 � k < 11 we have H: (w.d (/A) dearly has compact support contained in 2B .
We call J Darboux integrable if the sup of all L (J. . Show however. . . Pl.) [a. s. and define is an ntuple = similarly. Let [a. that cannot exist if is unbounded. . [a. . b].b]. < be the inf of of [a. this sup or inf is called the Darboux integral of J on We call J Riemann integrable if . and "Riemann sum" for a partition and choice by J: P I 111i mi(J) In} P � (�I" I"i] .] which are contained in [li_] . b].lil ) . The Riema1l1l integral UeJJUS the Darboux integral.I . For a partition = { o < . [a. P) S(j. I � I I I. Mi(f) Ii]. b] ei 111 S s [a. b].�) U (J. P) = L Mi(J) ' (Ii . < s } and Q = Io < .add up to e. . I ti_l .= ] S [a.�) (b) If J is continuous on then J is Riemann and Darboux integrable on and the two integrals are equal. S J on J J 1I). P. letm Ii] = length of the limit is called the Riemann integral of (a) We can define even if is not bounded.�) = L J(�i)(I' IiI)· Clearly L(J. [S. let = on [Ii_I. b]. .�n) �i M. b]. (d) Let J M on Let P = < . b]. b]. P) = Ll11i(J) ' (Ii IiI) i=l J U(J.. I .i� o S(J. [a. .lnlegralion 283 PROBLEMS I._I . P. P). . P. P � n L(J. P. .=1 S(J. "upper sum".Ii.�) [a. t] shaded lengths .sum of lengths of all [s. [a.]'S cmltained in [Ii. A choice for = with E [ I We define the "lower sum". < In} be two partitions of For each i = I. S.I . (Use uniform continuity of J on (c) If J is Riemann integrable on then J is Darboux integrable on and the two integrals are equal. b] > IR be bounded. {so { [Ii. P) equals the inf of all U (J. n. S(J. li]. ._] . b]. [a.
then U(J. Q) : Q a partition of [a. (a) Show that there is a singular 2cube c : [0.il� o L(J.n : [0. i=l There is a similar result for lower sums.)f(�i)g(�'. (e) Show that .M.n (t ) = () = (cos 21f1. b]. b]). equals > > H O n UPU+O 2. Show that fOJ choices �. Let f and g be integrable on [a. defined as a limit of sums.b]) H ). P) :s U(J. (R cos 2 1fl. P) = sup{ L (J. if Mi denotes the sup of f on [If_I . then f is Riemann integrable on [a. Compute fe de = f[o. (g) (Osgood's Theorem).1 ) = [ fg· Ja i=l Hint: If l g l :s M on [a . Q) + iL(M . For 11 an integer. I] JR2 . b].11 . It is called the winding number of C around 0. )1 :s MIJ(�'i )f(�.m) L ei. (h) Show that fe f dx + g dy. Q) + (M . P) = inf{ U (J.£7= 1 e. > R > 0. I I c· de.284 Chapter 8 Show that. b lim L f(�i)g(�'i)(li . (f) If f is Darboux integrable on [a.{OJ be defined by (c) Show that n is unique. CR2. let CR.11 = > CR"n  n > = C . li]. b].�' for P. then 1J(�'. ° as I PI I 0. IF JR2 . show that there is some such that is a boundary in JR2 .)g(�'. I].{OJ is any cUlve with c(O) c(I). sin 2Jr/ ) on [0.CI . R sin2n1fI). where C I 3. Q) : Q a partition of [a.n . (b) If c : [0. )I. I] n JR2 .)e. b]. =l :s U(J. and deduce Darboux's Theorem: ). and CR.{OJ such that ac.{OJ.� U(J.
+an .(" n . Hint: If /(z) i' ° for all z with Izl :S R. Let /: <C > <C be a polynomial. Define a. x l . .J)'a. there are some advantages in using singular simplexes. : " n I > " n by 7 ao(x) = ([I . I ' = 5. i=O 0 < i ::: 11. ! is a boundary.c. I (I)] n .+ . xn . n (t) = [CR.(x) = (x l .I ) a. Then we define and for singular nsimplexes c. Some approaches to integration use singular simplexes instead of singular cubes. .co. ! = / O CR . . then cR. . /(2) zn +a l znI + . Hint: Note that CR".{OJ..c = C o a. xn . where n :0: I .Integration 285 n l /(2) = Zn ( 1 + . x'.I ) a. let r : " n > JRn be the inclusion map. and write 4. · + . and an nchain is a formal sum of singular nsimplexes. define (a) Describe geometrically the images (b) Show that a2 = 0. . . n i s the boundary of a chain in <C . zn z (b) Show that /(z) = ° for some z E <C ("Fundamental Theorem of Algebra"). as indicated in the next Problem. Define cR. Although Stokes' Theorem becomes more complicated. . . . .) . 1] > <C by cR. . a. ! : [0. . Let !J. n C JRn be the set of all x E JRn such that A singular nsimplex in M is a Coo function c : 6.CR. n M. a a (a) Show that if R is large enough. ! . then CR .I ) in " n . ! . . . As before.. x' ] . aC = L (. . O.'£7.I ..x' . .
C2 is a boundary.c.286 (c) Show that if w = J dxl then /\ . L. (d) If C I (O) = C2 ( 0) and CI (l) = C2 (1). and prove that 1 /" dw = [ w. ? .J )form on IRn . k 7 c(x) = I . and x ' E ao("' kl .a. We then define (c) Show that we do have c = au + c' where c' is degenerate. I] > IR 2 be a closed curve. IRn . 6. x' is unique except when t = O. I]) is a point.. Hint: If au = L. • . that is. .I )form w.) (d) Define Ie w for any kchain C in M and kform w on M. Every x E "' k + 1 can be written as ' IX . a. (a) Show that a c = 0 implies that c = at. define c: "' k +1 > IRn by IV!orcover. c(x ' ). show that CJ . what can be said about c for chains c in the obvious way. (b) Let c : [0. . Ozapter 8 /\ (Imitate the proof for cubes. c'([O. For any singular k simplex c : 6. for 0 :0 :0 I . /\ dxn Jalll is an (n . dxi /\ . [ dw = [ w lc Jac for any (k . Show that c is not the boundary of any sum u of singular 2cubes. using either simplexes or cubes.
A manifold M is called simplyconnected if M is connected and if every smooth map I : S l M is smoothly contractible to a point. It is > 1 w = l(c(1» . partition S l into s a finite number of intervals each of which is taken into either U or V.U. Show that U is not smoothly contractible to a point. (c) sn is simplyconnected for n > J . Hint: If we do have w = curve c we have 7. (c) This is false for open subsets of IR 3 . (d) If M is simplyconnected and P E M. (See Problem 7. (b) A bounded connected open set U C IR 2 is smoothly contractible to a point if and only if it is simplyconnected. then M is simplyconnected. Let w be a J form on a manifold M. then M is simplyconnected. Show that w is exact. then any smooth map I : Sl > M is smoothly contractible to p. (f) If M is simplyconnected. then H I (M) = O.] (a) If M is smoothly contractible to a point. > dl. Suppose that Ie W '" 0 for every closed 8.) Hint: If p is in a bounded component of IR 2 . [Actually. (b) Sl is not simplyconnected. show that there is a curve in U which " surrounds" p. Hint: Show that a smooth I : Sl > s n is not onto. (Converse of nV UV V Problem 724.) Hint: Given I : Sl > M. any space M (not necessarily a manifold) is called simplyconnected ifit is connected and any continuous I : S l M is (continuously) contractible to a point. (e) If M = U where U and are simplyconnected open subsets with U connected.U is not con nected. then for any not hard to show that for a manifold we may insert "smooth" at both places.l(c(O» .) 9. (This gives another proof that n is simplyconnected for n > J .Integration 287 curve c in M. (a) Let U C IR 2 be a bounded open set such that IR 2 .
and show the same result with w replaced by Iwl. define geometric objects corresponding to odd relative tensors of type (�) and weight w (w any real number). n + J) be closed sets.288 Chapter 8 L 1M ¢ . (d) Let A n C (n. (Note that for each ¢ . M (b) Show that L 1M ¢ .) 0) Similarly. L dw . and suppose that w I I . 1 2. w = L L 1M " + w. Iwl < 00. Let I : IR IR be a Coo function with fAn 1 = ( . and that M L lM ¢ .I )"/n and support I C Un A n . I dx converge absolutely Il. Show that w. · Iwl < 00. w converges absolutely. w = L lM .) (c) Show that L"' E 'l' f . find a counterexample to Stokes' Theorem when M = (0. Find two partitions of unity <I> and such that L¢E ¢> f ¢ .. w = x dy. there are only finitely many " which are nonzero on support ¢ .y) l(x. Following Problem 712. and let o w . " . We define this common sum to be fM w.. Let <I> and lj! be two partitions of unity by functions with compact support." laM Problem 1 0. "' to different values. (e) Examine a partition of unity for (0.. J ) and i s a Oform whose support i s not compact. Let be an nform on an oriented manifold M n .y)1 (No computations needednote that equality would hold if we had the entire boundary. even though both sides make sense. J) by functions with compact support to see just why the proof of Stokes' Theorem breaks down in this case. ¢E ¢> cpe¢ cpe¢ l/re\ll cpe¢ l/fe\ll > 1 0. together with a proper portion lj! « x. (a) This implies that L¢ E¢> f ¢ . I dx and L E 'l' fIl. using of its boundary. (a) Let M be E IR 2 : < J }. · w .
Wm are positively oriented in Mp and Nq. Use the formula for r*dxl (problem 41) to compute r*(J'. . . q). Wm is positively oriented in (M x N )(p.J )form w on MI.d· 1 5. . . respectively. considered as a map into JR" . .J )form on M. (Note tbat the map i o r : JR" . . WI . )�.1 vectors 1 6.) r * (1' = r * i*(1 = (i 0 r ) *O".I . show that (b) If h : M x N > [ 1f1*W /\ 1fz*� = [ W · [ �. 1M IN JMxN JR is Coo. . ) = q 1+ h (p. We will orient M x N by agreeing that V I . " (p. VI .aMI. . on M and N. vn. .Integration 289 13. . Account for the factor I/Ipl" in Lemma 7 (we have r* (vp) = (J /lpl)vr (p) . respectively. Show that for any closed (n . If 1f1 : M x N > M or N is projection on the ith factor. · · · . and let w and � be an nform and an mform with compact support. . . V" and W I .{OJ. Let MI . (a) Let M" and Nm be oriented manifolds. but this only accounts for a factor of 1 /Ipl" . and e is an (n . N is h 1fI*W /\ 1fz*� for some w and �. M2 C JR" be compact ndimensional manifoldswithboundary with M2 c MI . 1 7. 1 4. . since there are n .{OJ is just r.{OJ > JR" . Show that de is 0 at some point. . then where g(p) = (c) Every (m + n)form on M x IN h (p. . Suppose M is a compact orientable nmanifold (witb no boundary). " " Un.q) '" Mp Ell Nq if VI .
0]. C (Ml C(Ml n S' D·. and let C(M) = {Ap : p E M .290 Chapter 8 1 8.{OJ be a compact (n . The latter integral is the measure of the solid angle sub tended by reason we often denote r * (J' by de . 1M 1c(M)ns2 M.�l E JR3 except those with x = 0. we define ¢(x. . r *a' (v. . Z E (00. A ::c OJ . y. " " Wn2 some A E IR. Show that E JR"p and let v E JR"p be (Ap)p for (b) Let M C JR" . �) to be the angle between the positive zaxis and the ray from 0 through (x. For aJJ (x. Wn2) = o.J )manifoldwithboundary which inter sects every ray through 0 at most once.. y. For this n 1 9. zl. . . Show that JM r *(J' = O. . (a) Let p E JR" .I )manifoldwithboundary which is the union of segments of rays through O . . (c) Let M C JRn{O) be a compact (11 . Let WI . W I . y = 0. M Show that [ r *(J' = [ r *(J' . y .{OJ .
00) or with x = 0. y . z) E S 2 (r). z) in ]R3 except those with y = 0. (c) If v is a longitudinal unit tangent vector on the sphere S 2 (r) of radius r. z) = vix2 + y2 (J' = h de /\ d¢. (e) Conclude that (J' = d( . y. ¢) is a coordinate system on the set of all points (x. .z) (x. 0]. y. e. then (v. then where . then d¢(v) = 1 .Integration 291 (x.. y.cos ¢ de ) . y = 0. (the minus sign comes from the orientation). Z E (00.. y. and e is considered as a function on ]R 3 .z) arctan y/x. y) (a) ¢(x. then de(wp) = Y X2 + / (d) If e and ¢ are taken to mean the restrictions of e and of] S 2. If w points along a meridian through p = (x. ) . X E [0. ¢ to [certain portions h: S 2 > ]R is h(x.z) = arctan( vlx 2 + y 2 /z) (with appropriate conventions). (b) If v (p ) = Ipl.. e(x.y..
where i' 0. .292 (f) Let {OJ on Show that If > for the form (Jzapter 8 > IR2 1"2 : IR2 1C : IR3 IRde2 Sl be the retraction. l w I(x+i ) u(x. for the form (J de IR3 is just 1C*de. and the sequence is 2 1 . . I/'(x iy )12 DI(x.y) x aav (x.. ·+a _ . 22. de = r2*i*(J. 0 on {OJ in terms of on IR2 r2*de = de . Let > Nn be a proper map between oriented IImanifolds such that p 7 Nf(p} is orientation preserving whenever p is a regular point. and the fact that p vr(p)/lpl for tangent to (h) Conclude that = = * (v ) = (i) Similarly.y) == for realvalued functions I(z) zn+aI Zn1 + I'(z) u v. y au h.To[/(" + w) yl(z)]lw. U U U .y)+iv(x. and then to be == + det . dBn_1 IRn1 dBn IRn Vi n VjVI V2 V3 .) = IIZn111 +(11 . ( + nl l.)a "n2I: ..y). then the form on [part of] on [part of] Use this to show that = r (g) Also prove this directly by using the result in part (c).. {OJ. given by = is proper � J). so that is the projection. or else all points are critical points of J. express 20.. . v dB3 r*(J' d(cos(¢ r) de) = d(cos¢de).y). S 2(1 p l). ih. with eventually outside of any compact set. (a) Show that a polynomial map ce > ce. Vi r*de de . +an/'(. Show that if N is connected. = and Show Hint: Choose w to be a real (d) Conclude that av /'(x+iy) = aau (x. then either 1 is onto N. (b) Let J Show that we have i where varies over complex numbers.y) i ay (x.y)+i ax (x. = J* : I: Mn M (e) Write that . Prove that a connected manifold is the union the are coordinate neighborhoods.
Integmtion 293 where J' is defined in part (b).M.I (a) is a locally constant function on the set of regular values of f Show that this set is connected for a polynomial I: ce > ce.{p) > L: be the obvious retraction.I )sphere L: around p such that all points inside L: are in JR" . (b) Show that this definition does not depend on the choice of L:. Let I"p : JR" . Show directly that if I : M > N is proper. 23. For p E JR" . Let M" . and conclude that I takes on all values. while DI is the linear transformation defined for any differentiable I: JR 2 > JR2 (e) Using Problem 21.I c JR" be a compact oriented manifold. then the number of points in I. (a) Show that this definition agrees with that in Problem 3. (c) Show that w is constant in a neighborhood of p. not using Problem 2 1 (which relies on many theorems of this chapter). (d) Suppose M contains a portion A of an (n . give another proof of the Fundamental Theorem of Al gebra. (f) There is a still simpler argument.M. choose an (11 .M. Define the winding number w (p) of M around p to be the degree of rp lM.I )plane. Let p and q be points . Conclude that w is con stant on each component of JR" .
H'(p. and 0 elsewhere. (b) Let IH:" > IH: be a Coo function which is positive on the interior of the unit ball. #11('1) I#rl (q) '" #gI (q) Hil l: H1(q) 24. = H': M x [0. Show that = ± 1. For E s n. q E N I (q). close to this plane. H.t) g(p) t g t. x w(q) w(p) M > N be smoothly denote the (finite) number 1 g. H(p.) Show that each 1+ is 1I diffeomorphism. 1] > be a regular value of Show that Let (mod 2).g:g. that this result holds so long as 'I is a regular value of both and I.A and which does not take any point of A onto the point in the figure. = == x H(t. N we will write / :::e g to indicate that 1 is (a) If l :::e then there is a smooth homotopy Q)) 26. The next few Problems show how to prove the same result even if M is not orientable. more gencrally. g: H: N.I .x» · p x at H(O. let IR x IRn > IRn satisfy ¢: (Each solution is defined for all by TheOI·em 56. I] > N such that for in a neighborhood of 0. Let M and N be compact nmanifolds. More precise conclusions are drawn in Chapter I I. 1 H p (a) Being smoothly isotopic is an equivalence relation.x) x. for in a neighborhood of I . which is smoothly isotopic to the identity.) ¢(H(t. g. 25. but on opposite sides. t. (This is one place \·vhere we use the stronger form of Sard's Theorem. H: p aH(t. and let homotopic. x) .t) /(p) t H'( p. The number of points in its boundary is clearly even. then IRn . by a smooth homotopy M x [0.) (e) Show that.t) �is an equivalence relatiOll.) (b) Show. if M is orientable. (Show that I"q l M is homotopic to a map which equals 'p l M on M . in general.294 Chapter 8 (a) Let of points in is a compact I manifoldwithboundary. For two maps M > smoothly homotopic to = I.M has at least 2 com ponents. and leaves all points outside the unit ball fixed. If / is smoothly homotopic to by a smooth homotopy such that 1+ is a diffeomorphism for each we say that is smoothly isotopic to g.
q E M. suppose X is a Coo vector field on M x [0. M we define the kform Wt +h . we have �. } of diffeomorphisms of M [not necessarily a J parameter group] .. then XI (p) will denote 1CM * X(p . of kforms on (where #1..e.. Let {X' ) be a Coo family of Coo vector fields on a compact manifold M.. and define w. O For a family w.0) be any point in the interior of the unit ball. and the argument which was used in the proof of Theorem 56.1 )Wo + IWI · Show that the family cp. (f) If M and N are compact nmanifolds. 1 ] . show that if M c JR" is a compact (n . This number is called the mod 2 degree of f. (To be more precise. (g) By replacing "degree" with "mod 2 degree" in Problem 23. I1 h.Wt .* (Lx' w. of diffeomorphisms generated by {X'} satisfies for ali I if and only if LXI Wt = Wo .WI · . + w.M has at least 2 components.). (d) If M is connected and p . which is generated by {X ' } . it follows that there is a Coo family {cp. then for regular values ql .. Wt = I " . with <Po = identity.. (p» (Xlf)(p) = lim CP'+h P» . O.:. then there is a diffeomorphism I: M > M such that I(p) = q and I is smoothly isotopic to the identity. for any Coo function I : M > JR we have I( ( I(cp. (e) Use part (d) to give an alternate proof of Sl£p 3 of Theorem 9.J )manifold.lnl£gmlion 295 (c) Show that by choosing suitable p and I we can make H (1. then JR" . q2 E N we have 27. o h� h (a) Show that for �(t) = cp. and I : M > N. = (l .. (b) Let Wo and W I b e nowhere zero nforms o n a compact oriented nmani fold M. = cp. i.1 (q) is defined in Problem 24).) From the addendum to Chapter 5. *w.
J) such that = = I(p) K(q . Let I : M k > JRn and g : Nt > JRn be Coo maps.W I = d A for some A. q ) = We define the linking number of I and g to be '" g(q) I(p) Ig (q) . t ) : q Show that n E N } = 0 for every I .I(p» C JRn  (OJ  Cif. t ) : p E M} (K(q . J) = g(q) j(p) (H(p. g) (c) For I.g . show that this holds if and only if 28. J ). i(f.  1 1 10 10 . (b) Let H: M x [0. J A (u v) 1C o o [r (u. and I( M ) g( N) = 0. g) = (Jlt + li(g. n Cif.O) H(p. g) = deg Cif.l(p)I ' i(f. Show that there is a diffeomorphism II : M > M such that Wo = J. g) = 4 = i(j. where M x N is oriented as in Problem 18. g).I r (g(q) . n = k + 1 + J. J ] > JRn and K: with JRn be smooth homotopies H(p. 'v)]3 du dv. g : S l > JR3 show that i ( f.g : by M x N > s n. Define (d) Suppose that JM Wo = JM W I . where M and N are compact oriented manifolds. O) = g(q) K(q. so that Wo .g (P. *W I ' . J ] > (a) i(f. N x [0.296 Chapter 8 (c) Using Problem 718.
c') be points close to p E M. M C ]R3 and (a. b. b. (d) Show that C(/. y)plane). b.b.c')'J> P Hint: First show that if and M = aN.b.(a. v) = det (g l )' (V) g l (V) . c) rt N. b. b . (V2 )p is positively oriented in ]R 3p when oriented in Mp.JI (u) g2 (V) . (a.J'(u) ( ) (the factor ) /41C comes from the fact that Is' (J ' = 41C [Problem 914]).b) dx /\ dz + (z c) dx /\ dy [(x _ a) 2 + (y _ b) 2 + (z _ C) 2]'/2  g. (a.c) . (a' .c) and (a'.p (u) g3 (V) . The next problem shows how to determine C(J. g) = 0 if J and g both lie in the same plane (first do it for (x.b'.a) dy /\ dz (y . {a.c') p triple wp. c) E N . (a. c) . (V2)P is positively lim g.c) /(a'.c) _ For a compact oriented 2manifoldwithboundary let  (x .g.Mp for which the Wp 1M dEl(a. then g. c') = 41C. b.b'. (U» (/1 )'(U) (g3 )'(V) (g 2)'(V) A (u. (V I )p. b' . c) is on the same side as a vector wp E ]R 3p . b.b. c) rt M. on opposite sides of M.  .c) E ]R 3 define dEl(a.J(u) 1 (P)'(U) (/3 ).c) = 0 for (a.M g. b.b'.hltegmtion where 297 1" (u.b. 29. v) = Ig(v) .b. . c) = Let (a.c)'I> p (a'. b. g) without calculating. (a) For (a. Show that (VI )p. (a. c) = 41C for (a. Suppose (a.
lSI J l(x.lSI J l(x .(a.(Z . b . ac I(x.the number of other intersections. z) I' I _ 1 ( ( ( l g*(dg. z) 1 3 X .+ be the number of inter sections where dgjdl points in the same direction as the vector wp of part (a). [ * Z . and .C) dy aa (a.298 Chapter 8 (b) Let J : Sl > IR 3 be an imbedding such that f( S I ) = a M for some com pact oriented 2manifoldwithboundary M. n = .b) dX ag.b) d.(X ..).(Y . .a) dy ..g) for the pairs shown below.[ * y . .C) dx . z) 13 ag.=  J 41T (c) Show that ag. b. Let . (An M with this property always exists. .. + . 1 38. See Fort. y. c) = S J* . y.a) dZ ab (a.g). Topowgy qf 3Manijolds. c) . .) Let g : S I > IR 3 and suppose The figure on the Jeft shows a nonorientable surface whose boundary is the "trefoil" knot. s' ) ) ) (d) Show that n = i(f. pg. y. that when g(l) = P E M we have dgjdl rt Mp. c) . Show that hut the surface on the rightincJuding the hemisphere behind the pJane of the paper is orientabJe.. Compute i(f. b .n.
q E JRn be distinct. If J: M . � is closed. q }) has dimension 2.1 (JRn . B C JRn .Integration 299 30. (b) Find the de Rham cohomology vector spaces of JRn .(p.(p . We define the cup product v : Hk (M) x HI (M) . and that Hn. (a) Let p .. and Ci E H k ( N ) .. i. fJ E HI ( N ) . q} so that A and B are diffeomorphic to JRn . Choose open sets A ... Hk+1 (M x N) is defined by Show that x is welldefined. and A n B is diff eomorphic to JRn. q } ) = 0 for 0 < k < n .{p. then [w] v [�] = [w 1\ �].. 31.J .. w 1\ � is exact if w is exact and Show that v is bilinear. N..F where F C JRn is a finite set...(OJ.. J* (Ci V fJ ) = J*Ci V J*fJ· (e) The crossproduct x : H k (M) x HI (N) . Hk+l (M) by (a) (b) (c) (d) Show that v is welldefined.e. then Ci V fJ = (J llfJ v Ci . show that A B po oq Hk (JRn . If Ci E H k (M) and fJ E HI (M). Using an argument similar to that in the proof of Theorem 16. and that .
> X n times . show f]).. where 1fj : Tn 7 S l is projection on the jth factor. ) Equality is proved in the Problems for Chapter I I . /\ de ik represent diff erent elements of H k (Tn ). (b) Show that every map J: sn let . Hence dim H k (Tn ) � (. . Sl . On the ndimensional torus Tn = S l de . • . : Chapter 8 M > M x M is the "diagonal map". p). (a) Show Ihat all de i l /\ . Tn has degree O. denote 1Cj*de.. 32. . Hint: Use Problem 25. by finding submanifolds of T n over which they have diff erent integrals. given by "(p) Ci V f] = "* (Ci X X = (p.300 (f) If that .
but there has been one notable w) = (w. (v.=1 . = "L. and conse 30) . . ) is called positive definite if (v. ( . we can define an inner product ( . which is symmetric. then n ( a 1 .CHAPTER 9 RIEMANNIAN METRICS exceptionwe have never mentioned inner products. v) i' o. w). . for r = 11 we obtain the "usual inner product".n . b). . ."L. then there is some w i' 0 such that (w. . a. = "L. albl . v) > 0 A positive definite bilinear function ( quently an inner product. w) 1+ (v. 11. for all v i' o. An inner product on a vector space V over a field F is a bilinear function from V x V to F. b) . I with vector spaces. ) is clearly nondegenerate. . a) > 0 for any a i' o. ) on (a.=1 > 0. . The time has now come to make use of this neglected tool. For us. . a ») = " (a ')2 L n r In particular. a'b'. For each r with 0 :0. . an ) . on JRn by (a. a'b'. a symmetric bilinear function ( . ). r :0.=1 For this inner product we have (a. n previous chapters we have exploited nearly every construction associated and nondegenerate: if v i' 0. . In general. v). the field F will always be JR.+1 . (a 1 . ffi. and thus with bundles. . a. i=r+ l n this is nondegenerate because if a i' 0. denoted by (v. . . .
a 2 b2 . J. then J* ( . n I n this expression.. Clearly. ) corresponds to the more complicated condition that the matrix {gij } be "positive definite". a n with a t least one a i =j:. )= i.= 1 . . ) is cquivalenl to the condition that (gij ) is nonsingular. Vn of V. o. Thus. wI· Since ( . . ) is defined on IR 2 by (a. e. Positive definiteness of ( . .g. W W.9 Notice that an inner product ( . we can write ( . for each V E V.302 ChajJle.j=i L gijv*i 0 v*j . . ) implies that the matrix The matrix {gij } has another important illl crpretation. w) = a(v)(w). a). ) on V is an element of 7 2 (V). then J* ( . with (v. an inner product ( . . so symmetry of ( . . meaning that . nondegeneracy of ( . and J: IR > IR2 is J(a) = (a. Since an inner product { . gij = (Vi . Vj ) ' gij = gji. Thus. . However. if ( . (gij) is symmetric. . Also. if ( . . we can define a linear functional ¢v E V*. ) is positive definite if and only if J is oneone. if V is finil<. ) i!i linear in the first argument. bv <pv(w) = (v . det(gij) i' 0. For allY basis V I . so if > V is a linear transformation. for all a l . gij a. . .' dimensional. ) gives lIS an iSOITIOI"phism a : V + V*. the map v !)o ¢v is a linear transfor mation fi'om V 10 V*. ) is clearly nondegenerate if J is an isomorphism OllW V. v*u. b ) = a I b l .a j > 0 L . with corresponding dual basis v*} . .. the matrix (gij) is just the matrix of a: V > V* with I 'espect to the bases {Vi } lor V and {V*i} for V*. . J* ( . Nondegeneracy of ( . ) implies that <pv i' 0 if v i' O. ) is a symmetric bilinear function on This symmetric bilinear function may be degenerate even if J is oneone. } is linear in the second argument. ) is positive definite.
w PROOF. ) simply by J(a. (3) II v + w ll :5 II v ll + II w ll (Triangle inequality).2A (v.V) (the positive square root is to be taken). and its discriminant must be negative. the func tion I I on JRII is not differentiable at 0 E JRIIwhich do not arise for the function II 11 2 . If not. i=l The principal properties of II II are the following.w ll 2 = (AV .ja. W) 2 . equality clearly holds. a ) = L )a.l 303 norm In Given any positive difmite inner product ( . This latter function is a "quadratic function" on V in terms of a basis {Vi } for V it can be written as a "homogeneous polynomial of degree 2" in the components: = g. then 0 # AV . AV . Il t jvf t 1=1 i. so 0 < II Av . v + w) = II v ll 2 + II w ll 2 + 2 (v . Il w ll = (II v ll + II w ll) 2 . Thus 2 2 4(v.:5 o. w ) II v ll 2 + II w ll 2 + 2 11 v ll . (2) If v and w are linearly dependent. •:. w) + II w ll 2 So the right side is a quadratic equation in A with no real solution. ) on V we define the associated II II by II v ll = j(V.Riemannian 1I1etri(. II w lL with equality if and only if v and w are linearly dependent (Schwarz inequality). . THEOREM.4 11 v ll 11 w ll < (3) II v + w ll 2 = ( v + w. For all v .w for all A (I) lIav ll = JaJ · II v ll .)2 . (2) J (v . E V we have E IR.j=1 . by (2) The function II II has certain unpleasant propeJ1iesfor example.W .w ) = A 2 11 v ll 2 . a j . IRn we denote the norm corresponding to ( laJ = . w) J :5 II v ll . (I) is trivial. I .
Vn for V such that (Vi . Chapter 9 11 11 2 = L gijV*.IIvll2 . . \Ve will now see that. ) . PROOF Let WI . II/(v)1I = IIvll for all v E V. W) = Wv + w ll2 . that is. 3. • • . If ( . . (f(v). 2. then I is also inner product preserving.w 1l2 ].) Consequentl. I(w» = (v.304 More succinctly. Suppose that we have constructed 1 :'0 i. . •:. there is only one positive definite inner product. ) is a positive definite inner product on an Ildimen sional vector spare V. Theorem 2 shows that tv.IIw1l 2] (v. that is. We obtain the desired basis by applying the "GramSchmidt orthonormalization process" to this basis: Since W I # 0. Similarly. Vi) = DU' (Such a basis is called orthonormal with respect to ( . ) on V. a. we can define /* ( . "up to isomorphism". w) = Wv + wll 2 .: there is an isomorphism f : JRII T V such that (a. ancJ clearly IIvl li = 1. /(b» . then (I) " i.IIv . Vk so that . .'O inner products which induce the same norm are themselves equal. THEOREM. If ated to an inner product ( . ) = ( . I n other words. An invariant definition of a quadratic function can be obtained (Problem I) from the following observation. } � k VI . . . ) . . THEOREM (POLARIZATION IDENTITY).b E IR". w" be any basis for V. if I: V + V is norm preserving. b) = (f(a). w E V. . w) for all v.j=i II II is the norm associ (2) (v. . ' V'j. PROOF Compute. then there is a basis VI .
A(V). Recall that an inner product ( .ic. Vk . • .wli. ) on V provides an isomorphism Ci : V + V* with Ci(V)(W) = (v. This is because we obtain a metric p on V by defining p(V. ) on V is sometimes called a Euclidean V. • • . . . . k . w). Now. ) * on V* by + ( V*)*. . defined by i(V)(A) we obtain an isomorphism fJ : V* + = a. i Using the natural isomorphism i: V + V**. . = I.I V We can now use fJ to define a bilinear function ( . . The "triangle inequality" (Theorem I (3)) shows that this is indeed a metric. . . Then W k +1 is linearly independent of V I . We also call IIvll the length of v. the symmetry of ( . w) = IIv . .I 305 and span V I . .. We have only one more algebraic trick to play. . : melric on A positive definite inner product ( .Riemannian 1\1el. Vk . W k . . ) can be expressed by the equation Ci(V)(W) = Ci(W)(V). .:::::: span Wj . Let It is easy to see that . . So we can define and continue inductively.
and let ( Then . ® v*j . without the invariant definition.)=1 n n = . the one which produces fJ).} for V..} and {v·*. )= . let {v*.1 So aI : V* fJ : + {v*.} be the dual basis for V*. Consequently ( . (gU).1 V· + Thus. ) * = L gij V**i ® V**j i. )* is an inner product on the dual space V* (in fact. ( .306 Letting this can be written which shows that Clzapw 9 a(v) = ft.) (gU ). a(w) = /1.j v*. (g U ) = (gU )I . To see what this all means.. if we let g U b e the entries of the inverse matrix. that this equation defines ( .) .j= 1 L g ij Vi ® vj . One can check directly (Problem 9). choose a basis {v. so that then ( .} and {v*. ) * independently of the choice of basis. n (gu) so is the matrix of is the matrix of is the matrix of a: V V* V V** with respect to with respect to with respect to {v. if we consider Vj E V**.} and {v. )* is also symmetric.j=l + L g. .) {v*..
letting a(v) = A. we have a(v)(v) > ° for v # 0. we need only describe which two elements. } an orthonormal basis of V. . . TdV) = T k (V*). . which implies that det A = ± l . we are interested in only one case. . It follows Iiom Theorem 75 that for any W E [I"(V) we have = L Cik i Cikj . Let V I . . . the simplest way to describe ( ) )* is as follows: The basis V*l. vn ) = ±w(w" . Similar tricks can b e used (Problem 4) t o produce a n inner product on all the vector spaces T k (V). for A # 0. . The vector space [I n (V) is I dimensional. /\ v*n for {v. . So to produce an inner product on it. However. . If we write n Wj = L CijjVj. v*n of V* is orthonormal with respect to ( . which we will not describe in a completely invariant way. Wn b e two bases o f V which are orthonormal with respect to ( . Vn is orthonormahvith respect to ( . • It clearly follows that v *1 /\ We have thus distinguished two elements of [I n (V). . Vn and WI" . so that then. .' 307 Notice that if ( . and [l k (V). /\ W *1/. ) is positive definite. . . /\ . . k�' " w(v" . This can also be checked directly from the definition in terms of a basis. )* is also positive definite. ). . so ( . w and w.Riemannian 1I1etric. In the positive definite case. . . they are both of the form v * . . . )* if and only if v" . . . ) . will have length 1 . . 1'=1 then So the transpose matrix A t of A = (aU) satisfies A · A t = 1. . . . We will call these two elements /\ V*n = ± W * 1 /\ • . wn ) .
. To express the elements of norm 1 in terms of an arbitrary basis W I . .s /Jositivf.. . \IVe now apply our new tool t o vector bundles.j ). .j=l L gU w*. Ifwe also have an orientation J1. Consequently. ) which assigns to each p E B a positive definite inner product ( .. )p on . . . If � = 1f : E T B i s a vector bundle. . . S2 : B T E.' (p). det(gU) is a/wll). " so if A = (Ci. " . /\ · · · /\ w * " gu = (w. . Will we choose an orthonormal basis v} . . .308 Chapw 9 the elements of norm 1 in n n (v). . Wj ) . n then = k� ' L CikiCikj . vn ] = /1. If � is a Coo vector bundle over a Coo manifold we can also speak of Coo Riemannian metrics. we define a Riemannian metric on � to be a function ( . and which is continuous in the sense that for any two continuous sections S1 . )= i. then we can further distinguish the one which is positive when applied to any (VI . the function is also continuous. vn ) with [v" . • . Vn and write Wi = j=l L Cijj Vj o n Problem 79 implies that Ifwe write ( . 0 w*j . we will call it the positive element of norm 1 in n n (v). then In particular.. the elements of norm in n"(V) arc * ± Jdet(gu) w .
Let � = . ) u In fact (Problem 7). + « p . there is no ( . To show that it is positive definite. and for some U strict inequality holds. UeC9 each ¢u(p)(v. Then ( . THEOREM. and let Eue(�) = U ElIe(.v) � . The only problem is that Eue(V) is not a vector space. define Then ( . note that (v. W E . We define ( . For v. W E . p eB then a Riemannian metric on � can be defined to be a section of EliC(�). )p is a symmetric bilinear function on . ) on TS 2 which gives a symmetric bilinear function on each S2p which is not positive definite or negative definite but is still nondegenerate.. ) is continuous [COO] and each ( . there is an obvious Riemannian metric. ) by (v.. the new object Elie(�) that we obtain is not a vector bundle at all.I (p). ) u i s a [COO] Riemannian metric for � I U. .b» p = (a.1 (p» . a ) .Riemannian 1I1etricJ 309 [Another approach to the definition can be given. v)� [The same argument shows that any vector bundle over a paracompact space has a Riemannian metric. : E + M be a [COO] k plane bundle over a Coo manifold M. I(U) U x IRk IRk .W)p = L ¢u (p) (v. •:.I (p) by ElIc(.I (p). but an instance of a more general structure.I (p» .. 2: 0.I (p). w)� = (lu (V) . Then there is a [COO] Riemannian metric on �. PROOF There is an open locally finite cover (9 of M by sets U for which there exists [COO] trivializations On V x IU : .. Let {¢u} b e a partition of unity subordinate to (9. b ) .] Notice that the argument in the final step would not work if we had merely picked liondegenerate inner products ( ..] 4. v.. (p.. lu (w» p . w)� UeC9 (v. a fibre bundle.. Let ElIe(V) be the set of all positive definite inner products on V. v)p = L ¢u (p) (v. Ifwe replace each .
) is a with (S(p) .310 Chapter 9 As an application of Theorem 4.p . All these considerations take on special significance when our bundle is the tangent bundle TM of a Coo manifold M. ) p on . M x IR by ALTERN ATIVE PROOF We know (see the discussion after Theorem 79) that if � is orientable. If � = 1[ : E + PROOF.1 (p)] * . so that �* is trivial. COROLLARY. ) implies that the union ofall "p is a homeomorphism from E to E' = UPEM[1[ . W E 1[ . which gives a positive definite inner product ( . then � is trivial if /J. Then for each I' E have an isomorphism defined by M is a k plane bundle. We "p (v)(w) = (v. •:. S(p» p = I . we settle some questions which have till now remained unanswered. + M is a I plane bundle. But � :::: �* . ) for TM. If � = 1[ : E and only if � is orientable. In this case. COROLLARY. Let ( . then there is a nowhere 0 section of [1 1 m = �*. PROOF The "only if" part is trivial. •:. [S(p)] = /J. Continuiry of ( . then � "" �*. If � has an orientation Riemannian metric on M then there is a unique and ( . J. w) p v. we then define an equivalence ! : E + !(A S(p» = (p. 6.1 (1'). A ) . ) be a Riemannian metric for �. Clearly s is a section. M. a Coo Riemannian metric ( .
j ) Idx 1 /\ This volume element is denoted by dV.j) > 0 since ( . U) is a coordinate system on M. as defined Even if M is not oricntable. dx 1 /\ . for each p E M the Riemannian metric ( .Ib·emalUzian A1etric. A Riemannian metric ( .I 31 1 each Mp. thus p. of course. . the elements of norm I. ) on M is. which ( . .Jdet(gu(p» dx 1 (p) /\ .j ) Similarly.. i i ) = L g. We have seen that they can be written ± . . then on V this form can be written . is called a R emann an metric on M. even though it is usually not d of anything (even when M is orientable and it can be considered to be an nform). ) as ( . M. the k=l L g. and det(g. and we can write it as Our discussion of inner products induced on V'" shows that for each matrix (gU(p» is the inverse of the matrix (gU(p» . ) induces on the dual bundle T*M. . /\ dx" . /\ dx n (p). j= 1 \vherc the Coo functions gij satisfy gjj = gji. If (x. in a coordinate system (x. then /J.Jdet(g. if x : V T JRn is orientation preserving. then on V we can write our Riemannian metric ( . n If M has an orientation /J. since ( . V) it can be written as . and we obtain an nform on M. ) is positive definite. ) is symmetric. it is a Riemannian metric on N if and only if f is an immersion (f p is oneone for * all p E N). is a contravariant tensor of order 2.j dx. i. • • /\ dx" l · .k g kj = 8( . ® dxj . a covariant tensor of order 2. ) on M determines two clements of n n (Mp). So for every Coo map f: N T M there is a covariant tensor f* ( .Jdet(g. ) * . we obtain a "volume element" on in Chapter 8. The Riemannian metric ( . which is dearly symmetric.p allows us to pick out the positive element of norm 1. ) on N.
with the "usual Riemannian metric" 1M dV. or becomes arbitrarily large over compact subsets of M.=1 ' /\ . L� (y ) I" If y is merely jJuuwise smootli._. /\ and ccvolume" becomes ordinary volume. which will occupy us for the rest of the chapter. There is an even more important construction associated with a Riemannian metric on M. We can then define the This certainly makes sense if M is compact. = ) (t lI y'(t) 1I d ) = . in which case we say that M has "infinite volume".. . so .. dV = I dx . For every Coo curve y : [a .] (with possibly different = = t I dr I dl Y from a to b. ® dx. and can therefore use ( . dxn l . If M is an IIdimensional manifold (withboundary) in IRn . . we have tangent vectors y' (t ) = dt dy E My (t). ). b] such that y is smooth on each [/. b] + M. ) = L dx. then gij = 8U.312 Clzapw 9 and is called the volume element determined by volume of M as ( . and in the noncompact case (see Problem 810) it either converges to a definite number. ) to define their length Wc can then define the length of dl dl Il drll ( dY ' dY ) ( = ( dY dY ) = dt' dt y(t ) ' to be precise . meaning that there is a partition a to < ' " < b of [a. 1.
I dl.Riemannian Metnes 313 left. A little argument shows (Problem 1 5) that f piecewise smooth curves in or JRn . s(t) = L�(y) = L�+a(y) = old s(t + a) . bJ + JR. ® dx. b .t}. y(1 .a.II .aJ by defining y(t) = For the new curve new y we have = 1. Then b a = I �. .a). Whenever there is no possibility of misunderstanding we will denote L� simply by L. . 1. We can reparameterize y to be a curve on [0. [ I �. (*) s'(t) = Consequently dyjdl has constant length I precisely when S(I) = I thus precisely when s(t) = I . We can also define a function s : [a.l (y I [IH .old s(a) If y satisfies s(t) = I we say that y is parameterized by arclength (and then often use s instead of I to denote the argument in the domain of V).and righthand derivatives at 11 . s(b) = L�(y).=1 this definition agrees with the definition of length as the least upper bound of the lengths of inscribed polygonal curves. + constant. with the usual Riemannian metric L dx. In .. . the "arclength function of y" by s(t) = L�(y) = Naturally.]) . . i=1 . . we can define the length of y by L�(Y) = I >:.
q) < t q to r with L(Y2) d(q. If p. q) = inf { L ( y) : y a piecewise smooth curve from p to q}. this is sometimes interpreted as being the equivalent of the modern equation ( . : [a. but what it always actually meant was i i. ) . q) 2: 0 and d ( p.314 Chapw 9 Classically. ® dxi . the Riemannian metric was not a function on tangent vectors.dxi at p should not be interpreted as a bilinear function at ail.8xi n where dx and 8x are independent infinitesimals. then for any £ > 0.d(p.dxi)(p) of dx.b] + [b. ) = L7.) Consider now a Riemannian metric ( .dxi n II 11 2 = The symbol dx. but the inner product of two "infinitely small displacements" dx and 8x.dxi dxi ® dxj was very strange: i. Define d(p. ) on a cOTlnected manifold M. but as the quadratic function i.j=l L gu dx.j=1 and we would use the same symbol today. we can choose piecewise smooth curves y. Moreover. (Classically. q E M are any two points. It i s clear that d(p.c] + Y2 : M M from from p to q with L(y. then there is at least one piecewise smooth curve y : [a.j=1 L gu dx.r) < t. the norm II II on M was denoted by ds. .) Consequently. ® dxi the value (dx. in classical books one usually sees the equation ds2 = Nowadays. b] . (This makes some Sort of sense even in modern notation. i f r E M i s a third point. equation (*) says that for each curve y and corresponding s : [a. b] + IR we have Ids l = Y* ( II II ) on [a. b] + M from p to q (there is even a smooth curve from p to q). �' gU dx.dxi appearing here is Tlol a classical substitute for dx.i dx. The classical way of indicating one wrote L g. p) = O.
it follows that :::: d(p. r) + 2e.c] . PROOF Define G: B " S"I + lR by Then G is continuous and positive. M > 0 such that on B " S". and if p : M " M + lR is the original metric on M (which makes M a manifold). . This is made clear in the following. p). and II II be the corresponding norms. a) < M lb l. Let U be an open neighborhood of the closed ball B = {p E lR" : I p l :::: I } .b] and Y2 on [b. M > 0 such that m . I I on B. Then mlbl < Ib IG(p . r) 0. q) + d(q. lI ap lip = lI<l b l a)p llp = II b llp.I .(y) :::: L(y) :::: M L. this gives the desi�ed inequality (which clearly also holds for b = 0) . and consequently for any curve y : [a. let ( . m < G <M Now if P E B and 0 # bp E lR"p .I be a = b/lb l . but d would still turn out to be the same (Problem 1 7). THEOREM. since Ib I G(p.1 315 If we define y: [a.c] + M to be YI on [a. q) + d(q. The function d : M " M + lR is a metric on M. 7.a) = lIap lip. G(p. let a E S". I I :::: II II :::: M . ).Riemannian Metric. Let I I = II II. a) = Ib l . except that it is not so clear that d(p . ( . then Y is a piecewise smooth curve from p to r and L(y) = L(YI ) + L (Y2 ) < d(p.(y). q ) > 0 for p # q . PROOF Both parts of the theorem are obviously consequences of the following 7'. i=1 " and let ( . d) is homeomorphic to (M. L dxi ® dxi. be the "Euclidean" or usual Riemannian metric on U. r). there would be difficulties in fitting together YI and Y2 . Since B x SIl . •:. ). ) be any other Riemannian metric. Since this is true for all £ > d(p. LEMMA. Then there are numbers 111 .l is compact there are num bers m.] The function d : M " M + lR has all properties for a metric. then (M. = [Ifwe did not allow piecewise smooth curves.b] + B we have mL.
I n order to investigate the question of shortest curves more thoroughly. Suppose we are given a (suitably differentiable) function F. For example.g.316 Chapter 9 Notice that the distance d(p . y) = �. e. we consider first a simple problem of this sort.. lR x lR x lR + R We seek. q ) defined by our metric need not be L(y) for any piecewise smooth curve from p to q. the manifold M might be 1R2 {OJ. the two semicircles between the points p and p on S l ).b] + IR with f(a) = a' and f(b) = b' one a' which will maximize (or minimize) the quantity t F(t. As an introduction to such techniques. we have to employ techniques fi'om the "calculus of variations". q) = L(y) for some y.f'(t» a b dl . if d(p. . Of course. and q might be p. f(t). among all functions b' f: [a. For example. then y is clearly a shortest piecewise smooth curve from p to q (there might be morc than one shortest curve. x . if" F(t.
b') of shortest length t )1 + [f'(t)F dl.y + l) . a') and (b. which is given (Problem 1 2) by To approach this sort of problem we recall first the methods used for solv ing the much simpler problem of determining the maximum or minimum of a function I: ]R + IR.y) = Dd(x. x. but critical points arc the only candidates for maxima or minima if I is everywhere differ entiable. Similarly.y) = 21f X JI+Y2 . /(x. then we are trying to minimize the area of the surface obtained by revolving the graph of I around the xaxis. y) t r> l(x. y) E ]R2 for which (*) D. those points x for which I'(x) = O. To solve this problem. for a function I : ]R2 + ]R we consider points (x.e. if F(I. l(t» between (a. or even a local maximum or minimum. As a second example. we examine the critical points of I.y) = o..Riemannian 1I1etrics 317 I r> then w e are looking for a function I on [a. i. A critical point is not necessarily a maximum or minimum. This is the same as saying that the curves I r> l(x + l.b] which makes the curve (t.
This can be done by considering a "variation" of f. & (u)(b) = b'. To find maxima and minima for funclions f: [a.a(u. that is) a function J (f) = t F(I. such that The functions I r> a(u. e) to the set offunctions f : [a.= du u=o du I aa at I Ib F (l.y). We might try to get more information by considering 0 = (f 0 e)' (O) for every curve e : (. For a variation Ci we now compute dJ(&(u» d . l) are then a family of functions on ( e.b) = b' b' a ' a b for all u E (e. f(t).(U. then we call a a variation of I keeping endpoints fixed. f'(I))dl a : (e. but it turns out that these conditions follow from (*). in other words if a(u. e) x [a.b] + IR a(O. Thus & is a function from (e. I).318 Chapw 9 have derivative 0 at the condition O. e). I)) dl u=o a . . If each &(u) satisfies &(u)(a) = a' . We will denote this function by &(u).e ) which pass through f for u = O. I) = f(I).e.b] + we wish to proceed in an analogous way.a) = a' a(u. b] + IR. because of the chain rule. by considering curves in the sel o all f IR.e) + 1R2 with e(O) = (x.
d1(1). . l) = 1(I) + u�(t). f'(t)))] dl aa aF ' I + au (0. f(I). 1'(1» d ( aF I.f (� (t. b] + IR with �(a) = �(b) = O.Riemannian 1I1etl1cS 319 = f [ :u lu�o F (I. (I» ( ay (I.aF1.] dl. the second term is 0.(1. 1(1).. f(l). ofcourse. flu l u�o J(&(u» J aF d aF 8J ax dl ay Ib � [. thus obtaining (*)  dJ(&(u» du For variations Ci keeping endpoints fixed. 1. I) [aF (1. = a The final result is. and we obtain (**) � dJ(&(u» ax au I u�o = Ib aa (0. f'(t» )] dl. a(U. for some � dJ(&(u» � : [a. we can apply integration by parts to the second term in the integrand. the variations Ci were taken to be of the special form . �7 (U. I). I) ay (1. (I» a aF aa a a I Ib U (0. Then we obtain a (u. 1(1). t)) ] dl Since a 2 ajaual = a 2 ajalau.dt = a . The derivative is called the "first variation" of and is denoted classically by a/ ax Iu�o Ib �(I) [ aF (1..1) [.dt a ' In classical treatments ofthe calculus ofvariations. f'(I))) ] dl. essentially the same. /(1) . 1 (I» b 11 =0 = a X . 1(1).
f. J'(t» ) = ax di ay 1)(t) = aCi (0. So the theorem is a consequence of the following simple 8'. 1 ) au o. •:' 0= di d ( '/1 /'(1)' 2 ) ' + F(t. As in the case of I dimensional calculus. J(t). which vanishes at a and b.(t» . LEMMA. J Ci J) 8J(Ci) 8J 8. If a continuous function g : [a. consider the case where equation is = ¢(b) = 0.(t). The point of if and only if I satisfies J C2 function I is a critical PROOF Clearly I must make the integral in (**) vanish for every aF (I. THEOREM (EULER'S EQUATION). Ci 8J(Ci) 8J(Ci) J.320 Chapw 9 As is usual in classical notation. The particular form (**) into which we have put now allows us to deduce an important condition. . We call I a critical point of (or an extremal for if =0 for all variations of I keeping endpoints fixed. y) = !i+7. then g = O. b] + IR satisfies l b 1)(I)g(t) dl = 0 for every Coo function 1) on [a.b] with II(a) = 1)(b) = 0. not only are the ar guments I and (t. there is no reason to expect that the condition = 0 for all will imply that I is even a local maximum or minimum for and we emphasize this by introducing a definition. . b) and ¢(a) As an example. 8J Ci J.d ( aF (t. l(t). but the dependence of on is not indicated (which can make things pretty confusing). /'(1» omitted (resulting in the disappearance of the function I for which we are solving). If I is to maximize or minimize then must be 0 for every variation Ci of I keeping endpoints fixed. x PROOF Choos C 1) to be ¢g where ¢ ispositive on (a. the arguments of functions are either put in indiscriminately or left out indiscriminatelyin this case. The Euler [/ (1)] . J.
�"��� ) . dl (t) f (t) 1 + . dy dx 1 + ()2 y=O.Riemannian Metrics so 321 hence which implies that So is linear.y) = xJI+Y2..11 +1' 0 = + 1'2)/" I'!" = I' + 1'2)/". We let Then .( )1l+ [f''(t) F ) . Notice that we would have obtained the same result if we had considered the case for then the Euler equation is simply This is analogous to the situation in I dimensional calculus. I F(I.x. J7 I.11+1'2 ' / . where the critical points of are the same as those of since For the case of the surface of revolution. dx d2y2 dy P = y..I �2 o= �(. (I (I this leads to the equation which we will also write in the classical form 0= )1 + [f'(t)F :!.y) = 1 + y2. 2J7 F(I. ( J7)' = L . o = �(2/'(1» . = dx ' the To solve this.x. I" = 0. we use one of the �o standard tricks Ocaving justification of the details to the reader). where Euler equation is _ .f'2 I/" = 0.
The graph of X +k = c cosh ( . and cosh . we write this as Y = x +k.) . � dy p = dx = Jey2 . dy . e eX + ex x = 2 cosh is shown below.dp = � d)'. x x � 0.322 Chapter 9 so our equation becomes 1 + p 2 . Pp2 1+ � log(l + p2) = log y + constant 2 y = constant .l ey Replacing c e by 1/ e.1 dy 1 = dx Jey2 Y and thus (see Problem 20 for the definition and properties of the "hyperbolic cosine" function cosh and its inverse) cosh. it is symmetric about the yaxis.yp dp = 0. decreasing for increasing for � o.
we define a variation of y to be a function aF d (a axl (/. J(I). 1' (1) ) ] dl 1 X o. b] + M J of J must satisfy the n equations a. : (t. J(I). /'(1» dl . t ) x for F: IR x IRn x IRn + IR. any critical point We are now going to apply these results to the problem of finding shortest paths in a manifold M.. Problem 2 1 investigates the special case where a' = b' ...f G.di ayFl (I. a 1= 1 U + Thus. I) aayFl (1.t) x [a .b] + IRn J(f) = In this case we consider that t F(I. not trivial to decide whether there a7" constants k and c which will make the graph of (*) pass through (a.. 1(/ ).1) [D (I. b'). and compute = L. with y (a ) = p and y (b) = q. b] + M is a piecewise smooth curve. 1'(/ » Ib n . f.a') and (b. (1. dJ(&(u)) dU b � a1 (0.b] + IRn with &(0) = J.Riemannian 1\1etnes 323 So our surface must look like the one drawn below..(1» ) = ..(I» I l a. If y : [a . It is. J'(I)) . by the way.. J(I). � a. (0. It is easy to generalize these considerations to the case where and J : [a. a . aF u=o [a.. : ( t . f(l). /. 1(1).
b) = q for all (I) . . (2) there is a partition on each strip We call " a variation of y keeping endpoints fixed if (3) u E (t.r) =" (u.(u. U) (othelwise wejust refine the partition). (0.t) x [a. (. aftenvards we will consider the relation between the two integrals.. < IN = b of [a.t) x [1. . we let paths y satisfy & (u) be the path 1 dL(&(u» du r> . . (u. I)..1) = y (t). However. b] we write lb I dy 1 2 dl = Zl ib ( dY dY ) dl. . such that a = 10 < II < .t). dt a dt ' dt aa" .1. I) is the standard coordinate system in (t..._ 1. l) . _ 1 .324 Chapter 9 > for some t 0.a) = p .r) .b] so that " is Coo (t.(u. We would like to find which for all variations Ci keeping endpoints fixed. As before.I ) aa aa"I (u. If (u. we will take a hint from our first example and first find the critical points for the CCenergy" I u=o =° E ( y) = z a I which has a much nicer integrand.] lies in some coordinate system (x. (Iaa I ) U ( U.]. 1. I) = " U (u. We can assume that each y l [1.
ti _ l i. I» .dt.j = 1 1 tj _ l " Then and So .c.\ 325 Then aet/al(U.j =l dl dl If we usc the coordinate system x to identify U with as functions on JRn . then we are considering IRn . I) is the tangent vector at time I to the curve &(u).I) So = L ii/(U. F(y(l). i yi i. then aa iii (U. Ifwe adopt the abbreviations eti (u.=1 1 l1 [}e/ 0 1 . a(u.t) ti_ l E(y! [li l .l) . (dY dY ) dl ' 1.y) = '2 L gij (x) .Riemannian 1\1elr. ndI dI dyi dyj 1 L gij (y(l» .. y'(t)) dt where F(x.. and consider the gij [. i ih. li J) = 2 1 = 2 11. t) = xi (et(U.
dyj dy' = �..= left hand tangent vector of y at I (l i· dl i ) �(tn dl To abbreviate the integral somewhat we introduce the symbols [ij.. j= l ax' dt dt So �.326 Chapw 9 In order to obtain a symmetrical looking result. ax' ) [ . agl.. we note that a little index juggling gives r.1.L. L.. ax' dt dt j= From (***) we now obtain Notice that the final sum in the above formula is simpl\" Remember that y is only piecewise Coo. j= 1 ax' dl dl i. i d i dy . agi/ dyi dyj = � agjl dyi dyj L.. /] a = � gi/ 2 ( ax' + agjl ax' _ agij . Let dyI(l + ) = nght hand tangent vector 0f y at I . i..
the sum can be written [ij. I] . . clearly cannot.� (aCi (O. but the integral  327 iI. but there it is. For the time being we present: with apologies. which appears in our result.Riemannian 1I1etrics These depend on the coordinate system. Consequently.fi).) This result is not very pretty. we will use the exact same expression for each [IiI .. Ii]. Il" d/) a. of course.1 . Now we just have to add up these results. Nevertheless.I](y(t) d� d�j n I [ n r n i ] dl. (In the case of a variation from 1 to N . 1. THEOREM (FIRST VA RIATION FORMULA). 9. simple to obtain conditions for critical points of E. this coordinate dependent approach. even though different coordinate systems may actually be involved (and hence different gij and V i ). t) !. i = 1. N . For any variation w e have dE(Et(u» du I Ci. Let � aa: (O. leaving endpoints fixed. . . From the first variation formula it is. later on we will have an invariant interpretation of the first variation formula._. . It should be noted that are not the components of a tensor.' \�I [ij. glr (Y(t) dd2. Ci .l Then we obtain the following formula (where there is a convention being used in the integral). 11=0 N dy .
then in the first variation formula we can assume that the part of the integral from Ii.. such that satisfies y: E y E y(O) dy (0) = v (ji d2yk � j dy' dyjt = .L [ij. ) for some i. COROLLARY. •.I. but a positive function times the term in brackets on (IiI. Suppose y is a critical point.j=1 =P o. choosing all aCi' jau(O. I!) : n r I. d j =l r=1 PROOF. . r/k ( y (I » dt d l. Given / with y(l) E U. which is 0 outside of (1'1 . I) to be 0.[i2 + .b] with I E (1.. and such that y l [I'_I. except one.328 Chapw 9 1 0 . then is a critical point of E� if and only if for every coordinate system (x. I](y(l)) d� d�j = 0 for y (t) E U.I to Ii is written in terms of (x. that for each P M and each v Mp. there is a unique (t.1 . (y(l)) d2. The final term in the formula vanishes since y is Coo. for some t > 0.' + i.] is in U. U). choose a partition of [a. L.). In order to put the equations of Corollary 10 in a standard form we introduce another set of symbols OUf equations Can now be written We know from the standard theorem about systems of second order differential equations (Problem 54). + M is a Coo path.t) + M. . If . Now apply the method of proof in Lemma 8'.. 1. U) we have y: [a b] 11 y i Lg/. If Ci is a variation of y keeping endpoints fixed.
Naturally..] satisfies the equation. Let y be a critical point. we see that yl [I' _h l. 1. We now have the more precise result. (0) = Y2 (0) then y.. . this means that is actually Coo on all of [a. ) = L dx ' ® dx'i=l Here g'j = D. This last fact shows that if y. so all ag'j /ax k = energy function satisfy 0. (0.Riemannian 1I1et. this y is Coo on (e. N . b] + M is a critical point Eg if and only if Y is actually Coo on [a. consider the Euclidean metric on * y JRn ) ( . As the simplest possible case. We already know that the integral in the first variation formula vanishes. we could replace 0 by any other I. A piecewise Coo path for (x. Now choose Ci so that c/ c/ a. Choosing the same as before (all are 0 outside of (1.e). So we obtain which implies that all !:1/. because the final term in the first variation formula still vanishes. By our previous remarks. The critical points y for the . .e) + M and Y2 ' (e./' dt ' aCt dy i = 1 . + d Y2 dl (0 ) = dl (0 ) y.cs 329 Moreover.e).1. b] and for every coordinate system for y(I) E U. U) satisfies Y ' [a. O] + M are Coo functions satisf ying this equation. . I.:. PROOF.. and Y2 together give a Coo function on (e._. . and ifmoreover d y. . ' [O.» . ' I I . and rt = O.) = !J. .b]. are O.j. COROLLARY.
when we considered only curves of the form I r> Any reparameterization of is also a critical point for length. both terms in parentheses are 0 (CorollalY 10).330 Chapter 9 Thus lies along a straight line. (y(t » dt df2 ' =l i. If eterized proportionally to arclength. THEOREM. y y y (t. since length is independent ofparameterization (Problem 16). since we have just seen that for to be a critical point for energy. This situation always prevails. TllllS the length is constant. PROOF Observe first. i]. this can be written as di dt r n Since is a critical point for E. Replacing agij lax' by the value given above.L ax' l.b] + M is a critical point for 12.j=l /=l r. The situation is now quite different from the first variational problem we considered.�. then y is param Now w e have ��:{ = [i/ . that y: E. }] + [}/. so is a critical point for the length function as well. from the definitions. y y y [a. the component functions of must be linear. d dy 2 d " gij (y(t» dit dyjt) I 1 = di (i. and hence must be parameterized proportionally to arc length. f(t» . dy d i dy' dy dyj " d2y dyj " " L agij (y(t» dt dtdt + L g'j (y(t) df'2dt = . +!+ y I dyldlll . This shows that there are critical points for length which definitely aren't critical points for energy.j=l dyi d2y' + L gi.
...j=l .m _ '"'" if _  _ '"'" if =  or We can find the equations for critical points of the length function L in ex actly the same way as we treated the energy function.j] + [jk. Ii] y [a dy/dl (x. For the por tion I of contained in a coordinate system U). For the moment we consider only paths + M with # 0 everywhere. we have y: .Riemannian 111etries The formula 331 ·· ag't [ik. g glm ogmj L. It will also be useful to know a f ormula for ag ij / ax k To derive one.. g gmj oglkm aY I.. Considering our coordinate system as JRn) we are now dealing with the case F(x.m aY il gmj ([lk.y) = \ iL gij (x)yiyj .b] y [li1. we first differentiate = m=l to obtain Thus we have ogikj L.. l]) by (*) Lg I.i] ax occurring in this proof will be used on several occasions later on.m] + [mk..m aYk I.
but be Coo because it has a zero tangent vector y oS : [0. L� (y)] 1 is a diff omorphism. Conversely. the third term vanishes. so y o S 1 is a critical point for E. and we can consider the reparameterized curve e This reparameterized curve is automatically also a critical point for L . so it must satisf the same diff rential equation. given a critical point y for L with y I # ° everywhere. the function d /d d2s/dl2 s: [a. dl .L ax dl dl aF ( I dy ) _ � i.2 ax dl After a little more calculation we finally obtain the equations for a critical point of L : dt ds d2s dyk dl2 = O' di It is clear from this that critical points of E are also critical points of L (since they satisf y = 0).j �1 ds ' y ( ) . Conceivably a critical point for L might have a kink. b] + [0. There is only one detail which remains unsettled.332 Clzapw 9 s (t) = L� (y) . . We introduce the arclength function Then So we have " a · d id j L 4· (y (t» � . L� (y)] + M. Since it is now parameterized by y e arclength.
2) + M satisf ying Yv(O) = q.Riemannian 111el1ics 333 there. This feature of the equation allows us to improve the result given by the basic existence and uniqueness theorems. The most elementary properties of geodesics depend only on facts about differential equations. Problem 37 shows that this situation cannot arise. Henceforth we will call a critical point of E a geodesic on M (for the Rie mannian metric ( . This name comes f rom the science of geodesy. A geodesic on the earth 's surface is a segment of a great circle. Let p E M. which is the shortest path between two points. as in the figure below. <2 > 0 so that f q E U and v E Mq with IIvll < <1 there is a unique geodesic . we must initiate or a local study of geodesics. In this case it would not be possible to reparame terize y by arclength. »). which is concerned with the measurement of the earth's surface. � PROOF. THEOREM. Observe that the equations f a geodesic. Then there is a neighborhood U of p and a number < > 0 such that for every q E U and every tangent vector v E Mq with II v ll < < there is a unique geodesic yv : (2. then I r> y(cI) is also clearly a geodesic. or have an important homogeneity property: if y is a geodesic. d v (0) = v. Before we can say whether this is true for geodesics in general. 13. which are so far merely known to be critical points f length. The fundamental existence and uniqueness theorem says that there or is a neighborhood U of p and <] . including surveying and the measurement of degrees of latitude and longitude.
y(l).« 21) . In fact. If (9 c Mq is the set of all vectors v E Mq for which expq (v) is defined. we claim that the map (expq)o. exp(v) = expq(v) = (The reason for this terminology will b e explained in the next chapter.. . d"i (0) = v. Identif ying the tangent space (Mq). •:.(v) = :!.I dl 1=0 eXpq (c(I» = v. : Mq + Mexp.) The geodesic y can thus be described as y(l) = expq (lv).334 Chapter 9 with the required initial conditions. 1] + M If v E Mq is a vector for which there is a geodesic satisf ying then we define the exponential of v to be dy y(O) = q. we have an induced map (expq). then the map expq : (9 + M is Coo) since the solutions of the diff rential equations for geodesics have a Coo e flow. we can let crt) = IV. : Mq + Mq is the identity. to obtain a curve c in the manifold Mq with dc/dl(O) = v E Mq = (Mq)o.. there is a natural way to give it a Coo structure. Since Mq is an ndimensional vector space. Choose < < < 1 <2 : Then if Ivl < < and III < 2 we have So we can define yv(l) to be yv/. y : [0. at v E Mq with Mq itself. (v)· In particular... the geodesic with tangent vector v at time 0) So (expq)o. Then expq 0 c(l) = expq (lv).
.=1 a L X'. . x l . (v) axi n We will denote ai by Xi (v). . For every P E M there is a neighborhood W and a number e > 0 such that (1) Any two points of W are joined by a unique geodesic in M of length < e. we recall some facts about the manifold TM. Then is a coordinate system on 1[ . . I I ' rr(v) q E U we therefore have the vectors while the vectors a/axt are all in the tangent space of the submanifold Mq a/axi l v span a complimentary subspace. xn ) v E Mq. .=1 where 1[ : TM + M is the projection. . q') r> v(q.. Then (q. . so that v= . xn . q . For tangent vectors l (X 0 1f) O .l (U). 1 4. v(q. . U) is a coordinate system on M. (3) For each q E W. . . . a v = L a' ox. If (x. the map expq maps the open eball in Mq diffeomor phically onto an open set Uq :::I W. . q') is a Coo function from W x W + . C TM. xn o n. x l .cJ 335 Before proving the next result. .Riemannian 1I1el. . xn ) = (xl. THEOREM. .. (2) Let expq(v) TM. q') denote the unique vector v E Mq of length < e such that = q'. then for q E U we can express every vector v E Mq uniquely as " .
Now. then + M is projection on the i th is a coordinate system on V x U. p) = . We may assume that V' consists of all vectors v E Mq with q in some neighborhood V' of p and IIvll < e. . p) E M x M. F. Define the Coo function F: V + M x M by F(v) = (. so maps some neighborhood V' of 0 diffeomorphically onto some neighborhood of (p. I + a!i I F.(v). = 0 E Mp we have (P' p) + Mp (p . •:. consider the geodesics through q of the form I r> eXpq (lv) f IIvll < e. ( a�i IJ a 1 . : M x M factor.336 C/zapw 9 PROOF. using the fact that (expp)o. Theorem 13 says that the vector 0 E Mp has a neighborhood V in the manifold TM such that exp is defined on V. Let (x. for . and q E W. We will use the coordinate system described above. V) be a coordinate system around p.p) F 0E F... Choose W to be a smaller neighborhood of p for which F(V') :J W x W.1 (V). a . it is not hard to see that at Given a W as in the theorem. : Mp is the identity. These fill out Vq . If ". ( � IJ a!. exp(v» . The close analysis of or geodesics depends on the f ollowing.. !2 . Consequently. is oneone at Mp .
which is just twice the second term on the right of (1). I) is just the tangent vector at time u to the geodesiC u r> expq ( u . v: lR + Mq be a smooth curve with IIvU) II k < e for all I. I) are omitted. So But Ct(O. In Uq . the geodesics through q are perpen dicular to the hypersurfaces { expq(v) .I) = ( � ' � ) is independent of u. .Riemannian Metrics 337 1 5. vU» I <u< 1. and define Ct(U. au al . vU» . for convenience: The first term on the right is Similarly. I) )=0 f all (u . LEMMA (GAUSS ' LEMMA).I) ' at (U. we obtain 0 since each curve u r> Ct( u l) is a geodesic. aCt aCt � (U. so lIaCt/aull = k . au al expq(O) = q. Thus the second term on the right of(2) is also O. I). It follows that (� ' � ) = 0 for all (u. I). But aCt/ au(u. IIvll FIRST PROOF. or A calculation precisely like that in the proofofTheorem 1 2 proves the following equation. Let = constant < e}. I) = a constant We are claiming that for every such Ct we have ( = expq (u . where IIvU)1I = k. so aCt/al(O. I) = o. in which the arguments (u) I) and Ci(U.
COROLLARY. v(u» fJ(u. ..{q} be a piecewise smooth curve. the first variation formula. •:. = (t) r dl =  (� O ( 1\ 2 dl = k2 . defined on [0. I].338 SECO}fD PROOF Chapter 9 Let v: IR + Mq be any smooth curve with IIv(t) II = a constant (note carefully the roles played by I and k < e for ali I. But each curve ('!t (0. we have = dE(du(u» 1u=o � E(  = the integral vanishing since y is a geodesic.bj + Uq . I). By u). c(t) = expq(u(t) . v(O» .0). v(t» . (0. Let c : [a.( a. Tt(I) ) . � (u) has energy . ddly ( I ») ('!t (0. and define expq(t . ddly (0») au au afJ dy .. � u» 10' 1 0 = dE�(U» lu� o ( = dg �.l) = Then fJ is a variation of the geodesic y(t) = expq(t . I). _ so 1 6. I ).(1») . .
If Ct (u. I) and u au (I) aCt . v(l» .' iii = ) = e(l) = Ct(u(l). the path c must contain a segment which joins the spherical shell of radius 8 to the spherical shell of radius r. PROOF with equality if and only if u is monotonic and v is constant.u(a)l. a. aCt + iii ' we have with equality if and only if aCt/al with equality if and only if I �r = + 1 �1 t I�I t lu'(M dl 2: lu'(I) l dl I �II = I . and lies between them. So the length of e is 2: r. For 8 > 0.{q} (otherwise break e up into smaller pieces). Then I L(y) ::: L(e). •:. and hence 2: lu(b) . ] + M be any piecewise Coo path from q to q'. 1 7 . 2 2: lu'(I)12 . Thus 0. q' E W. the length of this segment has length 2: r . and clearly e must be a reparameterization of y f equality or to hold. COROLLARY. v'(I) = 0. let y : [0. then di de Since ( aCt aCt = 0..u(a)l. with equality holding if and only if c is a reparameterization of PROOF y. We can assume that q' = expq(rv) E Uq . Let W and e be as in Theorem 15. •:. I ] + M be the geodesic of length < e joining q.for ° < u(l) < e and II v (l ) 1 I = Riemannian 111ellics 1. and let e : [0. Then 339 L�e 2: lu(b) . so that c is a radial geodesic joining two concentric spherical shells around q . u is monotonic and v is constant. . By Corollary 16.8. I) = expq(u .
If (M. q E C be two points with a unique geodesic C' of minimal length between them. ( . )'. and if c is a geodesic. n Notice that a portion of a great circle which is larger than a semicircle is defll1itely not of minimal length. the sphere have a continuum of geodesics of minimal length between them.340 Chapw 9 We thus see that su fficient6' small pieces of geodesics are minimal paths for arc length. Since there is a great circle through any point of sn in any given direction) these are all the geodesics. It is clear that if c : [0. then a oneone Coo function f : M T M' is called an isometry of M into M' if f*( . depending on how many time� the geodesic goes around the sphere and in which direction it starts. For the isometry I : sn + sn mentioned above. . So C' = I(C'). For example. ( . reflection through a plane £2 C )Rn+l is an isometry I : sn + sn. Then I(C') is a geodesic of the same length as C' between I(p) = p and I(q) = q. )) and (M'. ) ' = ( . the fixed point set is the great circle C = s n £2 Let p. We can use Corollary 17 to determine the geodesics on a few simple surfaces. ). 1] + M is a Coo curve. ) is the length of f o e with respect to ( . ) ' ) are Coo manifolds with Riemannian metrics. then f o e is likewise a geodesic. All other pairs of points have a unique geodesic of minimal length between them but an infinite family of nonminimal geodesics. even among nearb paths. so that C is a geodesic. Antipodal points on ). which implies that C' c C. then the length of c with respect to ( . if we first introduce a notion which will play a crucial role later. without any computations.
. This is false on a cylinder of bounded height. The geodesics on Z are just the images under I. b] can be extended to a geodesic defined on all of lR. d(p.L + JR2 by rolling Z onto JR 2 .Riemannian 111etlicJ 341 The geodesics on a right circular cylinder Z are the generating lines. the . In fact) if L is a generating line of Z) then we can set up an isometry I : Z . ) and the metric d : M x M + JR it determines.. b] + M can be extended to a geodesic from JR to M.. or JR" .. and the helices on Z . Notice that on both the sphere and the infinite cylinder every geodesic y defined on an interval [a. .{OJ. circles cut by planes perpendicular to the generating lines. ) is called geodesically complete if evelY geodesic y : [a. q) = inf{L(y) : y a piecewise smooth curve from p to q}. In general. . We are now in a position to wind up our discussion of Riemannian me\ rics on M by establishing an important connection between the Riemannian metric ( .. a bounded portion of JR".. a manifold M with a Riemannian metric ( .I of the straight lines in JR2 Two points on Z have infinitely many geodesics between them.
q) for all s E S. choose Up as in Theorem 14. There is a point on PROOF. q) :. we clearly have d(p.. then M is geodesically complete if and only if M is complete in the metric d determined by ( . old r > 0. any two points in a geodesi cally complete manif can be joined by a geodesic of minimal length. s) + d(s .q) = �ir (d(p s) + d(s. First of all. q). Moreover. Suppose 10 < r. Then d(y(to). r] be the least upper bound of all I f which (**) holds. q») = 8 + d( po. To prove this result.q). q E M with d(p. ). Given p. q») = 8' + d(po' . q) = r .I IE [8. Suppose M is geodesically complete. ) is a Riemannian metric on M. Then or (**) holds for 10 also. THEOREM (HOPFRINOWDE RHAM). q) = r . 8 < e. we will prove that S such that d(po. since every curve from . by continuity. If ( . . II v ll = 1 this will show that the geodesic y(t) = expp (t v) is a geodesic of minimal length between p and q. Let S' be a spherical shell p to q must intersect S. drs. This proves that (**) holds for I = 8. We claim that expp (r v ) = q. of radius 8' around y(to) and let po' E S' be a point closest to q. q) = Let S C Up be the spherical shell of radius po = expp 8v. d(y(t).(d(y(to).8.342 Chapw 9 18.r]. Now let 10 E [8. q) = ��. So d(po.
showing that (**) holds for 10 + 8'. But the path e obtained by following y from p to y(lo) and then the minimal geodesic from y(lo) to po' has length precisely 10 +8'. Using Theorem 14. •:. if A C M has diameter D. suppose M is complete as a metric space. and must therefore be a geodesic.10) . q) = r . Hence d(p. So e is a path of minimal length. it is not difficult to show that y can be extended past b. (**) holds for I = r. fact. and p E A . Given any geodesic y : (a.8'.Riemannian Metrics so 343 d( po'. q) .(10 + 8'). III other words. by a least upper bound argument. choose In + b. I n other words. po') 2: d( p. Conversely. any geodesic can be extended to ffi. As a particular consequence of Theorem 18. then the map expp : Mp + M maps the closed disc of radius D in Mp onto a compact set containing A.d(po'. bounded subsets of M have compact closure. note that there is always a min imal geodesic joining any two points of a compact manifold. Hence y(lo + 8') = po'. Clearly y(ln ) is a Cauchy sequence in M. From this it is clear that Cauchy sequences converge. . so it converges to Some point p E M. which proves (*). q) = 10 + 8'. which means that it coincides with y. q) = (r . so it must be that 10 = r. d(y(lo + 8'). From this result. In Hence (***) gives . it follows easily that M is complete with the metric d. b) + M. This contradicts the choice of 10. Consequently.
. with i : M + N the inclusion map. It is not hard to see that v = w : E + M is a kplane bundle over M.w) = 0 f all w E Mp} . On the other hand if M is the Mobius strip and S I C M is a circle around the center. then we can define M/.344 Chapw 9 ADDENDUM TUBULAR NEIGHBORHOODS Let M" C Nn+ k be a submanifold of N. then it is not hard to see that the normal bundle v will be isomorphic to the (nontlivial) bundle M + S If we consider S I C M C J1> 2 . then the normal bundle 1 of S I in J1> 2 is exactly the Same as the normal bundle of S I in M. so that for every p E M we have i. the normal bundle v of sn 1 C IRn is the trivial I plane bundle. i. for v has a section consistiIlg of unit outward normal vectors. E + M take MpL to p. so it too is nontrivial. For example. the normal bundle of M in N. ) is a Riemannian metric for N.C Np as MpL = {v E Np : (v. or p eM Let E= U Mp L and w .(Mp) C Np ' If ( .
and the map 1f I D : D + M is also a def ormation retraction. Then there is a neighborhood U of Xo such that IIU is oneone. For example. LEMMA. this open neighborhood can be taken to be the whole total space. But in general the neighborhood cannot be all of N. and H k (U) "" Hk (M). s 0 1f . for which the Osection s : M + U is the inclusion of M in U. ) f 1f : U + M and define D = {e E U : (e. A bundle 1f : U + M with U an open neighborhood of M in N. we add some remarks and a Lemma. 345 Our aim is to prove that for compact M the normal bundle of M in N is always equivalent to a bundle 1f : U + M for which U is an open neighborhood of M in N.Riemannian Metric. Before proving the existence of tubular neighborhoods. Moreover. Let I : X + Y be a local homeomorphism such that IIXo is oneone. In the case where N is the total space of a bundle over M.{OJ. so 1f is a deformation retraction. If 1f : U + M is a tubular neighborhood. or then D is a submanif oldwithboundary o f U. is called a tubular neighborhood of M in N. Let X be a compact metric space and Xo C X a closed subset.e) ::: I } . if we choose a Riemannian metric ( . and for which the Osection s : M + U is just the inclusion of M into U. then clearly 1f 0 S = identity of M. 1 9. as an appropriate neighborhood of S l C )R2 we can choose )R2 . So M also has the same de Rham cohomology as a closed neighborhood. thus M has the same de Rham cohomology as an open neighborhood. is smoothly homotopic to the identity of U.
. the map or or exp is a diff omorphism f e rom E. e It is clear also that exp(E. with the corresponding It f ollows easily from Theorem 13. which is equivalent to the normal bundle of M in N. = {v E E : I v l < e ) U. x + Choose a Riemannian metric ( . f if (Xn l Yn) is a sequence in C with Xn T x and or T y. there is e 0 such that g 2: 2< on C. for some p E M } E . and also x # y since f is locally oneone. Xo). •:. onto U. n q q y (0) y ( 1 ) q. We claim that f sufficiently small e. norm I I .) C U. and metric d: N N JR. f sufficiently small < > O. PROOF. •:. . 1] y One of the illteresting features of Theorem 20 is that all the paraphernalia of Riemannian metrics and geodesics are used in its proof. THEOREM. Chapter 9 Let C C X x X be x <neighborhood of Xo. = V E.346 PROOF. Let M e N be a compact submanif of N. This means that = expp / O) where E E. . that exp is defined on E. is compact. = {q E N : d (q. since exp is oneone on M e V" it f ollows from Lemma 19 that for sufficiently small < the map exp is a diff omorphism on E. Theorem 20 will be needed only in Chapter I I . dyjd ( dyjdl(O) q. then f(x ) = lim f(xn) lim f( Yn ) = fry). To prove that exp is onto U" choose any E U" and a point p E M closest to If + N is the geodesic of length < < with = p and = it is easy to see that is perpendicular to M at p (compare the second proof of Gauss' Lemma). If g: C JR is g(x. Since C is compact. y) = d(x. M) < e). Xo) + dry.. where we will also need the following modification. while they do not even appear in the statement. x # y and f(x) = f(y») . Then M has a tubular neighborhood 1f : U + M in N. Then f is oneone on the {(x. y: [0. . ) for N.y) E X > = X: Yn + > old 20. Let E = {v : v E Np and v E M/". Then C is closed. Let V e E be the set of a noncritical points f expo Then V :J M (consid or ered as a subset of E via the Osection). This will clearly prove the theorem. and compactness of M. and V. then g 0 on C.
Riemannian 111etrics 347 2 1 . using only inward point ing normal vectors . •:. PROOF Exactly the same as the proof ofTheorem 20. Let N be a manif oldwithboundary. THEOREM. Then aN has (arbitrarily small) open [and closed] neighborhoods f or which there are deformation retractions onto aN. . with compact bound ary aN.
" " ¢n is a basis q= i. . v) (c) Suppose q : V + F satisfies q( . . .q (u) 0. Suppose ¢i. + (a) Define q : V for V*.othe:rvvise.348 Chapw 9 PROBLEMS V I.v) = q(v) h ( u . • /\ Vk ofvectors Vi E V.) (c) Using Problem 79. /\ ¢k ¢ O. show that the signed kdimensional volume of the parallelepiped spanned by ¢1 . Show that if ¢1 . .v) = v. ) an inner product on V which is not necessarily positive definite.q( u ) . = q(u+v) . . ¢ (b) Let U1 . (a) Show that W E W if and only if W /\ ¢1 /\ . (d) Conversely. .q ( u ) . Conclude that W = W". . = = q(u + v) . ' . . v) q(v) is bilinear. . Let ( . and ( . " " ¢k has the same orientation as al l ' . .q(u) . 3. If ¢i = Lj aj. he deduces geometrically the COlTcsponding conditions on the coordinates of Vi.q(v) . then we have a geometl"ic condition for equality with WI /\ . . (The sign is + if ¢1 . Ifwe identify V with V**. .q(u ) . . /\ ¢k = 1/11 /\ . and let W and W". v). . or (b) Show that F by q(v) = h (v. 2. and that iJ(u. Uk be an orthonormal basis of W¢ = W". . Show that q (u+ v + w ) . 1/Ii E V* satisfy ¢1 /\ . = .q( w ). Vn for V is called orthonormal if ( Vi. . . v). ' " ¢k is det(aij ). . Let V be an ndimensional vector space. E. /\ Wk In Lcfons sur La Ceometrie des Espaces de Rieman1l.uj . + Let V be a vector space over a field F of characteristic # 2.q( u + w ) .q (v + w) Conclude that '1(0) =H = q( u + v) . if W = W"' . show that ¢1 /\ . and q(2u) 4q(u). . then f some aij . A basis VI . . Then show that q(v) = h ( v. Wi. . . Cartan uses this condition to define n k (V*) as formal sums of equivalence classes of k vectors. and let h : V x F be symmetric and bilinear. /\ 1/Ik . 1/Ik. ) be a Euclidean metric for V*. . .q(v) ] . be the subspaces of V* ¢ spanned by the ¢i and 1/Ii. j =l L Qij v*. • /\ ¢k = 1/11 /\ . so that we have a wedge product VI /\ . Vj ) = ±8ij. • . . ab and . and the signed volumes of the parallelepipeds are ¢ the same. /\ 1/Ik # 0. · v*j n q ( . show that this volume is the same for 1/11 .
) has index i . w) = 0 for all W E W } . ) k is independent of the basis v" (b) Show that (¢' /\ . where gij = (Vi . 5. ). (c) I f ( . v" 4. V n be an orthonormal basis (see Problem 3). Prove that dim WL 2: n . (b) Show that lv. for some r (the inner product ( . Hint: If {w. Hint: Apply the result on page 308 to a certain (II I )dimensional subspace of IRn . (Use Problem 716 . ) = ( . . . (e) The index of ( .dim W. then there is a vector v E V with (v. then = . . . Vj ) . let WL = {v E V : (v.. . Vk. X Vn _ ' . is defined on page 301)._. /\ ¢k o 1/1.  . I = . . . ) i s also nondegenerate on WL . • . . ) i s nondegenerate o n W. define inner products on 0k V and A k V by using the isomorphism V + V* given by the inner product on V.} is a basis for W. and let .Riemannian 1\1etries 349 (a) If V # {OJ. 1/Ji}'). ) I W is negative definite. x . then V = W Ell WL . V i ) = O. < ik � n be an orthonormal basis. . consider the linear functionals Ai : V + IR defined by Ai(v) = (v. and ( . .r. ) be a (possibly nonpositive definite) inner product on V. ) k on n k (V) by requiring that 1 � i1 < . (b) For W C V. in Problem 726. ) is the largest dimension of a subspace W C V such that ( . X Vn _. . x . v) # O. /\ 1/Id (c) I f ( . Show that these inner products agree with the ones defined above. x V. . there is an isomorphism f: IRn + V with f*( . . . with (a) Show that ( . (d) V has an orthonormal basis. Wi ) . Let ( . . x . S how that the index is n . Recall the definition of v. (d) For those who know about 0 and A k Using the isomorphisms 0 k V* "" (0k V) * and A k (V*) "" (A k V)*. /\ . (a) Show that (v. thus showing that r is unique ("Sylvester's Law of Inertia"). Define an inner product . ). Thus. 1/Ij )* } = det( ¢i.) det( ¢i. .Jdet(gij ).
7..) (b) There is no Riemannian metric of index 1 on S 2 gij and g'ij are related by axi axj g'r:t� = Lgjj a '.1 (p). )p is constant on each component of B . (a) Let ( . and define S' similarly. An indefinite metric on � is a continuous choice of a nonpositive definite inner product ( . Show that the index of ( . '. This problem requires a little knowledge of simpleconnectedness and cov ering spaces. (a) There is no way of continuously choosing a I dimensional subspace of S 2p. is the classical way of defining the tensor [having the compo nents] 10.350 Chapter 9 6. i}�� ' i. g C). Show that S is homeomorphic to S'. ) be a Riemannian metric on M. Let S be the set of e E E with (e. a . If � is a smooth bundle over a manifold M..j x '  8. . Let � '= 1[ : E + B be a vector bundle. + = ' X /J then g This. Show by a computation that if the functions .� 'r:t� _ . ) be two Riemannian metrics on a vector bundle � 1[ : E B. and the functions gij ."". show that S is diffeomorphic to S'. ) and ( . Define a tensor B of type @ by gil a '· a . of course. for each p E S 2 (Consider the space consisting of the two unit vectors in each subspace.j axi axj  L. and A a tensor of type that A (p) : Mp + Mp .. )p on each 1[ ." with det(gij) # 0. Let ( . g ij are defined by 9.e) 1. So If the expression for A in a coordinate system is .
Ai gjk · j=1 (b) Similarly. 12. S how that if we consider v(p) as an element of IRn . show that the area of the surface obtained by revolving the graph of 1 around the xaxis is t 21f1JI + (/')2 . . Vnl = = E /1.. Y ) j ±81j in a neighborhood of any point. . . _I)p) = det �I . (Vn I)p is positively oriented in IRnp when (VI)p. (a) If M = aN for an ndimensional manifoldwithboundary N C IRn . then v(p) is outward pointing in the sense of Chapter 8. (VI)p. Xn be linearly independent vector fields on a manifold M with a Riemannian metric ( . .bj + IR is positive. then V (p) dVn_l(p)( VI)p. . find Yh ...A2)' Show that if C has components C kJ . . i=1 The tensors B and C are said to be obtained from A by "raising and lowering indices" . . ). (a) Let XI. so that we obtain n everywhere orthonor mal vector fields Yh . Yn . (b) Let d VnI be the volume element of M determined by the Riemannian metric it acquires a submanifold of IRn . (vnIlp is positively oriented in Mp. 1 1 . . where n Blk = 'L.Riemannian 111etr. define a tensor C of type (�) by C (p)(AI . . (b) For the case of a nonpositive definite metric.k Blk dxl ® dxk . Yn with (YI. A2) = (A(p) * (AIl.I)dimensional submanifold with orientation The outward unit normal v (p) at p M is defined to be that vector in IRnp oflength I such that v (p). . then n ckj 'L. (V. Let M C IRn be an (n . .gkI A{ . . . Show that the GramSchmidt process can be applied to the vector fields all at once.. . as () . (b) Compute the area of S2 13.cs show that = 351 B LI. (a) If I: [a. . . . .
' . . x . /\dxn (p). W n w 14. i=1 (e) Let M 1R3 be a compact 2dimensional manifoldwithboundary. Prove (the original) W E (w . Denote the volume element of M by dVn . and that of aM by dVn _ l . Let T be the vector field on aM consisting of positively oriented unit vectors. /\ dxn . Show that for IRn we have n v(p)) . . T) ds JaM JM x X is defined in Problem 727). x vnd.•.l )iI ai dx' /\ . (a) Let v" be the volume of the unit ball in (V [ Cv x X.i /\ . /\(fXi(J.  . /\ dxn (p). /\ (fxi(J. .. and that of aM by ds. /\ d.352 Conclude that Chapter 9 dVn_1 (p) is the restriction to Mp of t(l)iI vi (p) dx' (p) /\ . dVn_1 (p) = restriction to Mp of (l)i'dx' (p) /\ . (c) Note that VI x . IRn . . .x2)(nI )/2 Vn_ 1 dx. .. (d) Let M c IRn be a compact ndimensional manifoldwithboundary. X (w. Let X be a vector field on M. with odentation and outward unit normal v. Conclude that vi (p) . .. V _1 = v(p) for some IR (by Problem 5).) /\ . X Ci Ci E X . Let X = Li a i a/ax i be a vector field on M. Denote the volume element of M by dA. C Slfikes' Theorem: J1. . v) dVn1 JM JaM (the function div X is defined in Problem 727). Hint: Consider the form on M defined by = L(.. (VI VnI> v(p)) = V.) /\ . . v ) dA = [ (X. Show that Vn = /'I (1 . Prove the Divergence Theorem: [ div X dVn = [ (X. . with v the outward unit normal on aM..
c(lll) for . l' jl + JR2 given by C(I) = (l. show that Vn = (In terms of the r function. ) = Li dx i ® dx i . this can be written (d) Let A n. f(I» . n n even ' 2. (1 . Using the method of proof in Corollary 88.I )volume of S nI . In = .Riemannian Metlies 353 (b) If In = ['. n JRn has the usual L(c) = (b) For the special case length. V2 = IT.In2 . show that nI (c) Using V.x2)(nI)/2 dx = . (a) Let c: [0. 1fnf2 r(l + n/2) . Show that I l ' rnI An1 dr = AnI . 15.) Vn = (e) Obtain this same result by applying the Divergence 111eorem (Problem 13). with X(p) = pp . where Riemannian metric ( . I] l' i=1 L [(ci )'(I)j2 dl. is the least upper bound of the lengths of inscribed polygonal curves. show that this + [/'(1)]2 dl . but reversing the order of integration. show that I (:��:! o 2 (n +' )/2. c: [0.. Hint: If the inscribed polygonal curve is determined by the points (Ii .(nI)/2 I · 3·5" 11 n odd.I be the (11 . I] + JRn be a differentiable curve.
(c) Prove the same result in the general case. we maintain the hexagonal arrangement. so that the triangles are more nearly in a plane parallel to the bases of the cylinder.354 a partition Chapter 9 1 [0. It is natural to suppose that the area of a surface is.c(liI ll = J(li . we can increase the number of triangles indefinitely. The topic of surface area for nondifferentiable sUlfaces is a complex one. but as H.li_I l2 + (J(lt) . CmmaK. .planes of the hexagons closer together. .J(ti_Il)2 J(li .li_I l 2 for some �i E [Ii _ I . on a compact set.li _I l 2 + f'(�i )(li . 1]. while the area of each approaches hi /2. and uniform continuity of "'. In this way. similarly. then we have IC(li ) . Ii ] . the least upper bound of the areas of inscribed polygonal surfaces. Top view To in(T("asc the number of triangles. which we will not go into here. Hillt: Use the results of Prob = 10 0 < . written by someone called M. . Schwarz first observed. bUl mow the. < In = of = lem 81. this least upper bound is infinite for a bounded portion of a cylinderl To illustrate Schwarz's example I have plagiarized the following picture from a book called Maml�Mamu't{eCKu.u A1fClJlUJ 'Ua MrIOZOo6jJa3URX.
f M .B and B . lb �' (I)[g(l) . I) [a au a a Y this result makes sense even if 1 is only C'. If p : [0. d d 19. (b) If p E B . q').e. g o n [a.) 18. If a continuous function . (b) D u Bois Re )'mond� Lemma.) d L (e) = L(e 0 pl. (a) By applying integration by parts to the equation on pages 3 1 8319. show that (d& (u» I dJ U u=o = a2Ci aF lb t (0.S. these details become quite important. q) ? min (p . (Show how to round offcorners of a piecewise Coo path so that the length increases by less than any given £ > 0. instead of piece wise Coo curves.S is disconnected. (In the theory of infinite dimensional manifolds. show that 17.l' �� (1'/(1). 1 (I» .B.e] dl = 0. ). Show that M . . Use q!eS this fact and Lemma 7' to complete the proof of Theorem 7. I) + M be a curve in a manifold M with a Riemannian metric ( . /'(1» dl] dl.b] with Hillt: TIle constant c must be lb �'(I)g(l) dl = ° __ l b a � (a) = � (b) g is a constant. then = (1. show that (p. (c) Conclude that if the C' function 1 is a critical point of J. and Theorem 7 is false. (a) Let B C M be homeomorphic to the ball {p E JR" : Ipl :s: I } and let S c M be the subset corresponding to {p E JR" : I p l = I}. 1] + [0. for all Coo functions � on [a.b] satisfies 0. then 1 still satisfies the Euler equations (which are not a priori meaningful if 1 is not C2). remember that the f ormula for length involves only first derivatives.S and q E M . a 1rbg(u)du. /(1). Show that the metric on M may be defined using Coo. Let e: [0.S are disjoint open subsets of M . 1] is a diffeomorphism. by showing that M .S or does no/ have to be disconnected. e= We clearly have so we need to find a suitable � with �'(I) = g(l) .Riemannian 1I1ettics 355 16.
at a and a. (b) Show that 2 2 cosh .and tanh. situated. Show that sinh(coshI x) I I = cosh(sinh.. (c) For those who know about complex power series: sinh x = sin ix . and hyperbolic tangent functions sinh..x) = J 1 + x2 1 v1x R=I r:. and tanh are defined by sinhx = . for convenience. and tanh. I respectively.' eX _ 2 eX cosh x = .sinh = 1 2 2 tanh + 1 / cosh = 1 sinh(x + y) = sinh x cosh y + cosh x sinh y cosh (x + y) = cosh x cosh y + sinh x sinh y sinh' = cosh cosh' = sinh . cosh. Consider the problem of finding a surface of revolution joining two circles of radius I . hyperbolic cosine. 2 _1 1 . ) I cosh x = cos ix.x) = cosh(tanh. We are looking for a function Vx �.. cosh. while cosh. 2 (cosh _ ' )'(x) = 21. The hyperbolic sine.. coshx (a) Graph sinh. 00).' eX + ex 2 sinh x tan h X = .denotes the inverse of cosh I [0. I I (d) The inverse functions of sinh and tanh are denoted by sinh.356 Chapter 9 20.
If a < yo/ cosh yo. 67 .Riemannian Metrics of the 357 form where e is supposed to satisfY x /(x) = c cosh c (c > e cosh :!. and sketch the graph. c Show that Aa has a minimum at a/yo. then there are two such c. If a = yo/ cosh Yo. find the value of Aa there. then there is a unique such c.y sinh y for y > O.1 . Examine the sign of 1/ y tanh y f y > O. 2 c. yo/ cosh Yo 0 a Yo/ cosh yo (e) Using Problem 20 (d). It turns out that the surface for C has smaller area.. namely c = a/yo = I / cosh yo.2. Aa (e) = c cosh :!.I � . or (b) Examine the sign of cosh y . (a) There is a unique Yo > 0 with tanh Yo = 1/ Yo. so JY02 . with CI < a/yo < c'}. (c) Let c > O. show that � = JY02 . cosh Yo [YO � 1 . (d) There exists c with c cosha/c = I if and only if a :s: yo/ cosh Yo.] . = I e 0).
is tangent to the envelope at points outside [a. so the envelope consists of the straight lines c cosh  x c = c cosh Yo. ftr a < )'0/ cosh )'0.358 Chapter 9 [These phenomena can be pictured more easily if we use the notion of an envelopec.a]. a). 176ff. '" a Yo/ ('osh Yo the envelope at points P.x . the graph of f is tangent to q cosh Yo :' y = � x/ .[ Volume III. cosh Yo y = ± . Yo The unique member of the family through (yo / cosh Yo . I) does no/ . y= = cosh :: _ c c :: sinh C :: (' . but the graph of /c.= ±Yo. Q E (a. The envelope of the I parameter family of curves x /c (x ) = c cosh is determined by solving the equations 0= We obtain x c a/c (x) oc . I) is tangent to the envelope at that point. pp. TIle point Q is called conjugate to P along t he extremal hi ) and it is shown in the calculus of variations that the existence of this conjugate point implies that the portion of /cl from P to (a. . .
j. (Compare with the discussion of aF d aF (/(1).) ax. ag'.� l. so that the Euler equations are actually f f.396. ax) ox'Y " [}2xl ox'Y " rk. k] and [a!'l.k=1 11 n ax' a2xJ ax' ax} axk L [ij. ax'· ax'� ax' . + L. y] ' = so that [ij.. ture. Show that � which we eventually obtained in our solution to the problem. 'l ax'. ax'� ax k L. F. k] ax'. y) = . = . 11 n [ij. /' (1» ax (a) Show that for any and conclude that the extremals for our problem satisfy � (F f aF ) f [ aF �aF] dl ay ax dl ay _ .)] 22.k] are not the components of a tensor.. and note the remark on pg. All of our illustratiOlls of calculus of variations problems involved an F which does not involve t. i. ax'� ax' ' )=1 .Riemannian 1I1ebic.. _ f and F: 1R2 + IR we have ( ) = O. show that 24.1J + (/') 2 .. ay 23. . /' (1» . (a) Let x and x' be two coordinate systems. with corresponding gil and g 'i) for the expression of a Riemannian metric.+ ax'/3 (Jx'l3ax'Y ox.ct(Jx'Y . Chapter 8.) For the corresponding [a!'l. Q.1 359 give a local minimum for conjugate points of a geodesic in Volume IV.dt ay (/(1). Y]'. V. i.k=1 ax ax'Y ax'· ax'� n ax} a2x' ax' a2x) + " g'l' ox'ct .}.. k=1 1=1 Show that any Coo structure on IR is diffeomorphic to the usual Coo struc Y r '. (Consider the arclength function on a geodesic for Some Riemannian metric on IR.j.f (b) Apply this to F(x.� ax'Y  x . ax'� ax'Y + L ax'Y ..=1 (   ) . aF = 0.Jl+7 to obtain directly the equation dy/dx = _ � ag'J �� ax} k L. (c) Also show that i. � = __ _ __ _ .
) = Li dx i ® dX i be the usual Riemannian metric on JR".) + Bk). . How can we .360 Chapter 9 (a) Let al/au i = ei : JR" + JRn . (b) Show that gil = (ei. and let L.r A k) r=1 Bik . sufficient t o solve for the ei . Suppose we are told that there is a diffeo morphism I : JRn + JRn such that Li ) gij dui ® du) = I' ( . (c) To solve for I i t is. (Note that the curveS obtained by setting all but one Ii constant are geodesics. E. Ifwe consider ei as a vector field on JR".k = [ij. Cartan uses j this approach to motivate the introduction of the ri� . show that I (a/au i ) = ei . and to solve for these we want to find differential equations 25.) Bi). . where u l . Let ( . j. • • • ) un again denotes the .i· (e) By cyclically permuting i . since they correspond to lines parallel to the x i axis.J standard coordinate system on JR" . In Lcfons sur La Ciomitrie des Espaces de Riemann. k. e) ) . deduce that so that A� = rD.) = r[. k]. ill theory a t least. ). go about finding I? satisfied by the ei 's. Show that we must have (d) Show that = = " L g)r A �k + g.i gij du i ® du J be another metric. directly from our equations for a geodesic. (f) Deduce the result A.
w")". then y is a geodesic. ( . Show that for v E Mc (r ) with (v. (The equation p"(I) = can be solved explicitly: p(t) = J' e M (. ) is an inner product.) . de /dl) 0. we define ( .) p'(I)/1(I) 28.Riemannian Metlies 361 26. (b) Given Riemannian metrics on M and N. if e satisfies this equation. ) ") are two vector spaces with inner products. Describe the geodesics on M x N for this metric.b) + M be a geodesic. 27.b) be a (b) Conversely. (Gauss' Lemma is the '�special case" where c is constant. and let diffeomorphism. If (V'. Show that e = y o p satisfies p : [". I'l) + [a. ( . (c) If e satisfies for Il' JR + JR. then e is a reparameterization of a geodesic. where M'(s) = /1(5). ) on V = V' Ell V" by (V' E9 v". where v E Mc (r ) with = OJ. the geodesic U r> eXPc(r) u . (a) Let y : [a. n and (V". v is perpendicular to these hypersurfaces. w' E9 w") = (v'. and consider the hy (v. it follows that there is a natural way to put a Riemannian metric on M x N.) ds. (a) Show that ( . Let c be a curve in persurfaces {exPC (r ) v : Il v ll = M with dc/dl # ° everywhere. w')' + (v". de/dl) = constant.
) 30. . (Gauss' Lemma still works on exp«('J). 31. b) + M be a geodesic with L � (y) = d(y(a). ba ba L� (e) = d(e(a). . If era) = y(a) and e(b) = y(b). . so that each y k is linear. . Choose a basis VI . Let y: [a. and note that a geodesic y through p is just exp composed with a straight line through 0 in Mp. unless e is also a geodesic with (a) Show that in this coordinate system we have r6(p) = O. . . Schwarz's inequality for continuous functions states that with equality if and only if f and g are linearly dependent (over IR). . let x be the coordinate system xo exp. 32.5 E(e). y(b» = L(y) is a geodesic. e(b» . ) is a Riemannian metric on M and d : M x M + IR is the corre sponding metric. Let p be a point of a manifold M with a Riemannian metric ( . defined in a neighborhood U of p. b] + M with d(y(a). If ( . Show e that y is a curve of minimal 1ength between p and q = y(b). among all curves in exp«('J) . a n ) . ) . .t . y(b» . and suppose that expp is a diff omorphism on a neighborhood ('J C Mp of {ly ' (O) : 0 ::s I ::s I ).5 (b . . Vn of MP l so that we have a �Crectangular" coordinate system X on Mp given by Li a i vi r> (a t ..a)E� (y).5 . b] + M be a geodesic with y(a) = p. Hint: Recall the equations for a geodesic.. sufficiently small pieces of a geodesic mil limize energy. In particular.362 Chapter 9 29. (a) Prove Schwarz's inequality by imitating the proof of Theorem 1 (2). (c) Let y: [a . then a curve y : [a. (b) For any curve y show that with equality if and only if y is parameterized proportionally to arclength. Conclude that E(y) < E(e) L(y)2 L(e) 2 E(y) = . show that [L�(Y)f .
y . u (O) . then there is Ii > 0 (depending on such that >t exp(B. (g) Let f: U + IRn be a diffeomorphism of a neighborhood U of 0 E IRn into IRn.) and such that y'(O) is tangent to exp(S. ( dyk )2 _ " Y2rk dyi dyJ ] . q ' E U has a unique geodesic of minimum length between them. 8).q). '.Riemannian 111etricJ 363 = r(q) d(p. r(q') < . the maximum of occurs at either q or q ' . Show that for sufficiently small " the image of the open . dl 2 L.. 0 so that (d) Let B. Show that d2 ( r o d) = 2 [ . and let be the unique geodesic of length < joining them. Show that for sufficiently small .p B. dl .. y y o = tangent to exp(S.. Hint: If ' ( ) is y(O) d(r y)2/dl = y) y(l) s. d(r y)/dl o. (f) A set U C M is geodesically convex if every pair q. > O. i.) for 0 # I E (8. then 0 (e) Let q and '1' be two points with r (q) . conclude that if Il y '(O) I (b) Let r: U + IR be = 'i I :s: is strictly increasing in a neighborhood of O.). roy 2. and this geodesic lies completely in U. Show that exp ( { v E Mp : IIvll < d) is geodesically convex for sufficiently small . {v E Mp : IIvll :s: 'I and S.J.. dl dl k (c) Note that l L ddyi dyl n2 L( dylk )2. then Using part (a).). So that r o y Lk (yk)2.. = {v E Mp II v ll = d.ball is convex.k iL. Show that the following is true for all sufficiently small ' > 0: if is a geodesic such that E exp(S.J d k d I is sufficiently small.
I) 1 2 dl a u(h) q..c (O)1 1. (a) There is an everywhere differentiable curve that length of c 1 [0 . Show that if is some K > 0 such that for all near 0 we have I I u'(t) I 1 5.t) I · K2 such that V at points near lim h�o lim h�o d(p. c (h» L (c I [0.:.'=__"::::__"10 _ u'(t)2 + I '!. 1. f(I» in ]R2 such = Hint: Make c look something like the following picture.. 1 d�.ICi (U (I). I C(I) V(I) if and only if approaches a c is C'.. and suppose limit as + 0 (even though v(O) is undefined). h�o Ic(h) . ..364 Chapter 9 33. then there = q Hint: In Mq we clearly have Since expq is locally a diffeomorphism there are 0 < K.0) is I length from ����+ (�. h]) .O) to (!. < for all tangent vectors (c) Conclude that I d(u��(t» I [ h = = lu i .O) to (1. then = a. length from (!. h] # lim c(t) (I.O) is ! (b) Consider the situation in Corollary except that > 0 for near o. If c is C'.
(a) If y is a geodesic. show that L(c) is the least upper bound of inscribed piecewise geodesic curves. Let f : M + M be a map of M onto itself which preserves the metric d.w�O L (cv. w ) = 1.Riemannian Metrics 365 (d) If c is C'. lim v. l�� II v _ w ll v. let . 34. exp w) .w ) (c) Conclude that d(exp v. if Yv.(cX) = c/'(X). o Let f : M + N be an isometry. Y ) II X II · II Y II = il X II · II Y II II IX II · II I Y II ' II X II 2 + II Y II 2 [d(exp IX. then L(yv.w.(y'(O» = d/(y(I» dl Show that II /'(X) II = II X II . v.I Y II 2 II X II 2 + II Y II 2 2(X. Y E Mp. exp l y)] = hm II IX II · II I Y II II X II · II Y II _ HO I1=0 .w is the unique geodesic joining exp(v) and exp(w). then f 0 y is a geodesic.w�O L(c) (b) Similarly. 36. ) and corresponding metric d. (a) Using the methods of Problem 33. (c) Given X. = l . (b) Define /' : Mp + M f(p) as follows: For y a geodesic with y(O) /. and that /. uSe Problem 34 to show that II IX . W E MP l then L (exp o c) = l lim . = p. ric space structures determined on M and N by their respective Riemannian metrics. and Yv.w = exp 0 cv. show that if c is the straight line joining v. Show that f is an isometry of the met 35. Let M be a manifold with Riemannian metric ( .
show that hO lim II v + l w ll . and finally the rest of y.lIv ll . ) to y(to + u). one can use the equation (u. ). Show that if II is ti'O hm . Choose II < 10 and consider the variation Ci for which & (u) is obtained by following y up to I" then the unique geodesic from y(t. and hence an isometlY. and suppose that y'(lo+ ) # y'(to) for some 10 E (0. Alternately. (b) Conclude that if w is linearly independent of v. (/'(X). and then that f'(X + Y) = IRn with w # 0. The same result then holds in allY vector space with a Euclidean metric ( . critical . then Hint: (c) Let y : [0. /'(Y» f(p) .lIl w ll t # O. II v ll · cos Ii where Ii is the angle between u and v. then dL(&(u» /dulu�o paths for length cannot have kinks.lI v ll I = (v. If v : IRn + IR is the norm. # 0. e 37. lI v + l w ll . v) = lI u l i . y(l) y(t. w ) II v ll . TIlus. then the limit is Dv (v)(w). 1 ). a contradiction. (a) For v. w E f'(X) + f'(Y). 1] + M be a piecewise C' critical point for length. Y)p = (d) Part (c) shows that 1' : Mp + Mf(p ) is a diffeomorphism. Use this to show that f is itself a diff omorphism.366 Conclude that Chapter 9 (X.) y(O) close enough to 10.
41. TIle Poincare upper half plane with the Riemannian metric Jf' is the manif old {(x. . 39. Y2 (a) Compute that : y > O} ! ri. = . Investigate the geodesics on a cone (the number of geodesics between two points depends on the angle of the cone. ) = dx ® dx + dy ® dy . Consider a cone C (without the vertex). (c) Show that there are isometries of P" onto itself taking any tangent vector at one point to any tangent vector at any other point. p }. The sum of the angles in any triangle is > 1f.3. and let L be a generating line. = r"2 = rl.�. = . r�. TIlese results show that pH provides a model for "elliptical" nonEuclidean geometry. Q» Show that evcry geodesic y : jR + pn is closed (that is. Unfolding C ..L onto jR 2 produces a map f: C . Let g : S" + pn be the map g(p) = [p] = {p.\' 367 38. )) on pn such that g * « . y all other ri� = O. Consider a cylinder Z C jR3 of radius t.Ri£1nannian 1\1etfie. )) is the usual Riemannian metric on sn (thc one that makes the inclusion of Sn into jRn+' an isometry). there is a number a such that y(t + a) = y(l) f aU I). Find the metric d induced by the Riemannian metric it acquires as a subset of ffi. and some geodesics may come back to their initial point). (a) Show that there i s a unique Riemannian metric « . 40. and that every two geodesiCS intersect exactly or once. y) E jR' ( .L + jR2 which is a local � P(� isometl)� but which is usually not oneone.y.
(d) Show that these geodesics have infinite length in either direction. together with the straight lines parallel to the yaxis. y(t)).368 Chapter 9 (0. B' . Consider the upper halfplane as a subset of the complex num bers <C. upper half plane is a model for Lobachevskian nonEuclidean geometry. The sum of the angles in any triangle is < 7[.. c.: J 3' YN. Show that the maps z J( ) = az + b cz + d a. c) and radius R. b .bc > O are isometries. These results show that the Poincare = = A ti c : c . d E 'llI. show that all the geodesics in Jf2 are the (suitably pa rameterized) semicircles with center on the xaxis. show that y'(t) 2 y(t) (c) Using Problem 27.. so that the upper halfplane is complete. (f) For those who know a little about conformal mapping (compare with Prob lem IV 76). and that we can take any tangent vector at one point to any tangent vector at any other point by some J. ad .. . (e) Show that if y is a geodesic and p ¢ y. . Conclude that if length A B length A ' B' and length A C length A ' C' and the angle between the tangent vectors of f3 and y at A equals the angle between the tangent vectors of f3' and y' at A'. A. then length B C = length B'C' and the angles at B and B' and at C and C ' are equal ("sideangleside"). then there are infinitely many geodesics through p which do not intersect y. Considering (b) Let C be a semicircle in Jf2 with center at it as a curve t 1+ (t.
45. where N is a Coo manifold. (b) If � = ]f : E > M is a smooth bundle. and choose a fixed point P E M. 48. more generally. 46. E. Show that the Riemannian manifold M is geodesically complete. Show that the Rie mannian manifold M x N is also complete. b 1 > M with initial value y(a) = p can be extended to all of JR. show that the normal bundle of M C E is equivalent to �.Riemanni(J}l Metric. Prove that there is a geodesic y : [0. . where the bundles are over a smooth manifold M [or. conclude that TE '" ]f*(�) Ell ]f*(TM) . (a) Show that the normal bundles v" V2 of M n C N n +k defined for two diff erent Riemannian metrics are equivalent. If ( . Ell E3 . Let is indeed a kplane bundle. )) is complete. (a) If M n C N n +k is a submanifold of N. then g'( . 49. (b) Using the notion of Whitney sum Ell introduced in Problem 352. Let M be a manifold with a Riemannian metric ( . and give N the Riemannian metric described in Problem 26.g'( . show that v M and N be geodesically complete Riemannian manifolds.00) > M with the initial value y(O) = p. ). ( . )) is complete if and only if (N. 369 42. over a paracompact space]. 44. (a) Given an exact sequence of bundle maps as in Problem 328. having the property that y is a minimal geodesic between any two of its points. Let p be a point in a complete noncompact Riemannian manifold M. (b) If � = ]f : E > M is a smooth kplane bundle over M". Does it necessarily follow that the Riemannian manifold M is complete? 43. show that E2 ". show that the normal bundle v M x Ell TM '" (TN)IM. ) is a Riemannian metric on N. This problem presupposes knowledge of covering spaces. ) is a Riemannian metric on M. Let g : M > N be a covering space. and (M. Then there is a unique Coo structure on M which makes g an immersion. Suppose that every geodesic y : [a. 47.. Let M be a Riemannian manifold such that every two points of M can be joined by a unique geodesic of minimal length.
According to Problem 322 there is S' c M so that (TM)IS' is not orientable (the Problem deals with the case where (TM)IS' is always trivial. in fact.) . (Thus. but the same conclusions will hold if each (TM) IS' is orientable. it is not hard to show that a bundle over S' is trivial if and only if it is orientable). Using Problem 47. show that the normal bundle v of S' C M is not orientable. (b) Use Problem 329 to conclude that there is a neighborhood of some S' C M which is not orientable. (a) Let M be a nonorient able manifold.370 Chapter 9 50. any nonorientable manifold contains a "fairly small" nonorientable open submanifold.
with the operation + . y) is Coo. ab aI from G x G to G from G to G are continuous. A Lie group is a group G which is also a manifold with a Coo structure such that (x. The circle S' is also a Lie group. y) 1+ xy X 1+ X . and at each point at least one of them is a coordinate system. where Z C JR denotes the subgroup of integers. A topological group is a space G which also has a group structure (the product of a. b) a 1+ 1+ T preceding chapters.I are Coo functions. The functions x 1+ cos 2n: X and x 1+ sin 2n: x are Coo functions on JR/Z. and these groups can be studied by the methods now at our disposal.y) sin2Jf(x . Thus the map (x.CHAPTER 1 0 LIE GROUPS his chapter uses.I In In In JR x JR > JR > S' = JR/Z. One way to put a group structure on S' is to consider it as the quotient group JR/Z. because in the study of these problems the groups of autom orph isms of various structures play a central role. It clearly suffices to aSSume instead that the single map (a. y) 1+ xy' is Coo. y) (x. It clearly suffices to assume that the map (x. y) 1+ xy is Coo.' from S' x S' to S' is also = 1+ 1+ cos2Jf(x . and illuminates. consequently the map (x. y) 1+ xy. where we are concerned 'V'lith geometric problems. As a matter offact (Problem I).Y 1+ xy.cos2JfX sin 2Jfy.y) = Coo . b) 1+ ab' is continuous. 371 . We will mainly be interested in a very special kind of topological group. b E G being denoted by ab) such that the maps (a. many of the results and concepts of the It will also play an important role in later Volumes. The simplest example of a Lie group is JRn . it even suffices to assume that the map (x. y) 1+ X . which can be expressed in coordinates as one of the two maps (x. cos2Jfx cos2JfY + sin 2JfX sin 2JfY sin 2Jfx cos2JfY .
since the entries of A B are polynomials in the entries of A and B. By Proposition 21 I . it suffices to show that it is continuous. 1R). is norm presen1ing. To show that 0(11) is a Lie group we must show that the map (x.a E Z and b' . 1R) is a homeomorphism (since O(n) is a submanifold of GL(II. 1R) is the group of all nonsingular real II x II matrices. the set GL(II. the pairs (a.. The argument in the previous paragraph shows. (This gives another proof that S' is a Lie group. a manifold. the torus S' x S' is a Lie group. One of the most important examples of a Lie group is the orthogonal group 0(11). generally. The general linear group GL(II. 1R)). IR) = det. and hence can be given the Coo structure which makes it an open submanifold of IRn' .{OJ) is open.1)/2. 1R) with A . The torus may also be described as the quotient group IR x 1R/(Z x Z).b') represent the same element of S' x S' if and only if a ' .' (1R .' )F = det A ij / det A. that if H e G is a subgroup of G and also a submanifold of G. Smoothness of the inverse map follows similarly fi'om Cramer's Rule: IIIIIIIII!! (A. considered as a 2 subset of IRn Since the function det : IRn' > IR is continuous (it is a polynomial map). of dimension 11(11 . IR). but this is true because the inclusion of 0(11) > GL(n . In particular.e. then G x H.372 C7wjJler 10 If G and H are Lie groups. where A ij i s t h e matrix obtained from A by deleting row i and column j. This condition is equivalent to the condition that the rOws [and columns] of A are orthonormal. and thus inner product presen1ing. f S' C 1R 2 can be considered as the group or . which is equivalent to the condition that. then H is a Lie group. the matrix A represents a linear transformation which is an "isometry'" i. Problem 233 presents a proof that 0(11) is a closed submanifold of GL(II. Many imporlant Lie groups are matrix groups. Later in the chapter we will havc another way of proving that 0(11) is a Lie group. is easily seen to be a Lie group. b) and (a'. with respect to the usual basis of jRn. A t = I.b E Z. Multiplication of matrices is Coo. with the product Coo structure. y) 1+ xy . and the direct product group structure. cOJ"isting of all A E GL(n. where A t is the transpose of A. and in particular. is also Coo as a map from 0(11) x 0(11) to 0(11).' which is Coo on GL(II.
K is dosed since components are always closed. then the component K con taining the identity e E G is a closed normal subgroup of G. It is know that these are the only spheres which admit a Lie group structure.) It is possible for a subgroup H of G to be Lie group with respect to a COO structure that makes it merely an immersed submanifold. If G is a topological group. The group 0(11) is disconnected. Clearly SO(I1) = {A E 0(11) : det A = I } . PROPOSITION. As we have seen.I . •:. image of L in S l x S l = lR x lR/(Z x Z) is a dense subgroup. but we will not need this fact. Since e = aIa E aI K. Since e E bK bI . PROOF If a E K. then the immersion. The group SO(2) is just S l . after some work (Problem 18).Lie Groups 373 of complex numbers of norm I. so K is normal. It even turns out. If G is a Lie group. we mention £(11). If G is a Lie group. and also a Lie group for some Coo structure which makes the inclusion map i : H > G an I l lnl l I . respectively.1 K C K. we have aI K C K. it follows similarly that bK b I is connected. then aI K is connected. Similarly. the group of all Euclidean . S 3 is the Lie group of quaternions of norm I. we have K. the two components consist of all A E 0(11) with det A = +1 and det A = . so K is a submanifold and a subgroup of G. Since this is true for all a E K. For example. We define a Lie subgroup H of G to be a subset H of G which is a subgroup of G. which proves that K is a subgroup. then K is an open Lie subgroup. For any b E G. is a subgroup. since G is locally connected. if L C lR x lR is a subgroup consisting of all (x. we have bKb1 C K. a subgroup which is an (imbedded) submanif old is always a Lie subgl"Oup.I b is a homeomor phism of K to itself. the component containing the identity I. As a final example of a Lie group. since b 1+ a . ex) for e irrational. which we have already seen is a Lie group. I\1oreover. that a subgroup which is an immersed submanifold is always a Lie subgroup. Hence K is a Lie subgroup. This is not accidental. then K is also open.
the maps La* : Gb Ra * : Gb 7 Gab 7 Gba for all are isomorphisms.l . and T is a translation: T(X) = Ta (X ) = X + a . b It is easy to see that this is t�ue if we merely hav(' for all a E II E G. In fact.e.1 = A B. La : G > G and Ra : G > G. T where A C 0(11). COJlsequently. A Ta(BTb). which shows that E(I1) is a Lie group.1 TB(a_b). given Xe E Geo there is a unique left invariant vector field X On G which has the value Xe at e .1 X + a) = x + A(a) = TA(a) (X) .374 Chapter 10 motions.. Vie can give E(I1) the C"" structure which makes it diff eomorphic to 0(11) x IR" . E G. by La(b) = ab Ra(b) = ba . isometries of IR". since translations and orthogonal transformations do not generally commute. A vector field X on G is called left invariant if La.l = A Ta_bB. A little argument shows (Problem 5) that every element of E(I1) can be written uniquely as A . G. Consequently. so A TaA. Clearly the component of e E E(I1) is the subgroup of all A T with A E SO(I1). Consequently. . respectively. with inverses La . Now E(I1) is not the direct product 0(11) x IRn as a group. if a E G we define the left and right translations.1 = A TaTbB. Notice that La and Ra are both diff eomorphisms.l and Ra.1 (x) = A(A.X = X Recall this means that for all a . i. For any Lie group G.
so we can define an equivalence G x lR" J (t J=I ciX (a) i ) = " (a. c ) . •:. Since the map (a. X" be the left invari ant vector fields v. . . ' ./f.. b) 1+ a b i s C oo o n V x V we can write for some Coo function Ji on x(V) x x(V).I 375 PROOF. Then XI . A Lie group G always has a trivial tangent bundle (and is consequently orientable). . • : 3. Coo .Xe(xi) 0 = Xe (xi La) . . COROLLARY.rith these values at e. It suffices to prove that X is Coo in a neighborhood of e. . XXi (a) = Then " Xe (xi = j= 1 = . Let X" . Choose a basis X' e . Then for a E V we have XXi (a) = 2. . . Cj a(xi 0 La) � ax} 0 La) j= 1 L c j D"+j Ji (x(a). PROPOSITION. . . U) be a coordinate system around e.Lie Grau/}. This implies that X is I e where Xe = j =l L cj D X a I e which shows that XXi is Coo . Xn are clearly eveI)"Alhere J : TG > linearly independent. b E V implies ab ' E U. La. . PROOF. x(e)). " " X"e for Ge. Every left invariant vector field X on a Lie group G is COO.X around a (Prob lem 51). Choose a neighbor hood V of e so that a. since the diffeomorphism La then takes X to the C"" vector field La. Let (x. c ' .
X].376 Chapter 10 A left invariant vector field X is just one that is Larelated to itself for all a . Consequently. ] satisfying [X. etc. Y. and will be denoted by £(G). 1R) or f the Lie algebra of 0(11). Since GL(n. If If. Y. x V as 2 a set) by It is easy to check that this makes V into a Lie algebra. etc.. Henceforth we will use X. 1R) f the Lie algebra of GL(Il. We can then define an operation [ . . Z E V. then or we can define an operation [ . YJ = [Y. the tangent space of GL(n . YJ.. ] on the direct sum V = V. a Lie algebra is a finite dimensional vector space V. Z] + [[Y. to denote the left invariant vector fields with X(e) = X. 1R) is an open submanifold of IRn ' . or Since the [ . Consequently. X S') is also abelian. Y. Proposition 63 shows that [ X . Clearly £(IRn) is abelian. we write or gl (ll. The vector space Ge . together with this [ . Ell V2 (= V. or 0(11) In general. we call a Lie algebra abelian or commutative if [X. X] + [[Z . since the vector fields a/ax . 2. Z] . ] i f i = 1 . [X. or The Lie algebra of IRn is isomorphic as a vector space to IRn. YJ i s left invariant i f X and Y are. This notation requires o some conventions f particular groups. the Lie algebra £(S' x . and X. IR) at the identity I can be . 1R) is more complicated. with a bilinear operation [ . ] operation. X] = o [[X. are left invariant and [a/ax . YJ '= 0 f all X. (Sometimes the Lie algebra of G is defined instead to be the set of left invariant vector fields. . etc. Y(e) = Y. Y](e). YJ = [X . YJ = 0 'jacobi identity" f all X. and consequently must be abelian. are Lie algebras with bracket operations [ . ] operation is assumed alternating. Y. a/axj ] = O. The Lie algebra £(S') of s' is I dimensional. it is also skewsymmetric. .) We will also use the more customary notation 9 (a German Fraktur g) f r £(G). ] on Ge by [X. Xl Consequently. The structure of gf (11. is called the Lie algebra of G. and that £(G x H) is isomorphic to £(G) x £(H) with this bracket operation. to denote elements of Ge.
2 . So if N is another 11 x a __ .NM lkl rJ . 1R) corresponding to M. we have i. 1R) i s the linear function LA) (B) 0 = xkl (AB) = L A k . Now the function xkl Mxkl (A) = L Mij :ij a i.[M.Lie Groups 377 identified ".IA k Thus.(Mx kl ) . 1R). k=i k # i.. 0 1 (x kl 0 LA : GL(n . then an 11 x 11 (possibly singular) matrix M = (Mij ) can be identified with M/ Let M be the left invariant vector field on GL(n. ax l / a k. NlI = � (MN . N/ (Mx. M/ (xkl ) = M/ (xkl LA) .kl ) matrix. 1R) as f ollows. 1R) > GL(n.lith jRn 2 .l ..kl _ Mjl Mx {0 ax ij 11 . For every A E GL(n. a Xl} a = L Nkj Mjl = (NM lkl· j=l From this we see that " .I.j I = MA (xkl ) = LA. 0'=1 n with (constant) partial derivatives � (xkl ax" 0 LA) = A ki {0 j =I J # I. So = L Mi/ A ki = L M.B. If \ve use the standard cOOldinates xij on jRn.j = L Nij . Wc compute the function M xk l on GL(n.=) n 0'=1 n •.
378 thus. we can define a cun'e A : IR > 0(11) by A (') � ( j ' cos t . then [ .ba . This subspace may be determined as follows. . IR) we can consider 0 (11) / as a subspace ' of GL(I1.Aj/(O) . . which is exactly the dimension of 0(11). :2 Thus 0(11)1 C jRn can contain only matrices M= ( M which are skewsymmetric. If we did not know the dimension of 0(11). then k�i differentiating give.sin ! sin ! cos t J (rotation in the (i. j)plane) j .b] = ab . j with i < j. 0 M1 2 M1 3 .NM.8) > 0(11) is a curve with A(O) = I. and thus identify 0(11) with a certain subspace of IRn . IR) with IRn' . If A : (8.1 )/2. Mi n M1 2 0 0 M1 3 . the bracket operation is just [M. if we define [a. For each i. so 0(111I must consist exactly of skewsymmetric matrices. which shows that L A ik (t)Ajk (t) = �ih A i/ (O) = . Mi n 0 ) This subspace has dimension 11(11 . if we identify Chapter 10 gl(I1. ] satisfies the Jacobi identi�: Since 0(11) is a submanifold of GL(I1. IR)/. and we denote (A(t))ij by Aij (t). Notice that i n any ring. we could use the following line of reasoning. N] = MN .
[ G be a Lie group. Since i* : He Ge is an isomorphism into. 1 operation is juSt I) = £(H).sin t at (i. then Lb. takes H to the maximal integral manifold containing Lb. Any X E He can be extended to a left invariant vector field on H and a left invariant vector field on G. (H) = H. Let H be the maximal integral manifold of /j. and let i : H > G be the inclusion. that is. let /j. For E G. f] X [ i. i. = = La.1 )/2. is an integrable distribution. Then there is a unique connected Lie subgroup H of G whose Lie algebra is I). (i. He is a subspace of � which is closed under the . if E H.ba. Consequently. j) and (j. b) ab. among themselves. In particular. He. For each E H C G. Y](e)). = so Lb.Lie Graul" 379 with sin t and . ([X.(/j. if Y E and are irelated. i). La. i) and (j. THEOREM. He with this induced . other words. then [X. Using Theorem 67. containing If b E G. Now this map is clearly Coo as a map into G. It follows immediately that Lb permutes the various maximal integral manifolds of /j. moreover.. invariant. consider a Lie subgroup H of any Lie group G. We do not need any new calculations to determine the bracket operation in o(n). •: a X(a) e.a be the subspace of Ga consisting of all for X E I). then clearly Lb. 4. To prove that it is a Lie subgroup we just need to show that 1+ is Coo . The proof of uniqueness is left to the reader. it follows that it is Coo as a map into H. 1 operation. Then the set of all A'(O) span the skewsymmetric ma trices.a) = /j.1 . which means that [X. X) = X(a). In fact. X La o f i 0 La. The fact that I) is a subalgebra of � implies that /j. = La : G > G In. j). and I) a subalgebra of �. I 's on the diagonal except at (i. so Lb . Let PROOF. b (b) e. He C Ge = � is a subalgebra of �. This implies that H is a subgroup of G. and O's elsewhere. and are irelated. Yl [X. This correspondence between Lie subgroups of G and subalgebras of � turns out to work in the other direction also. and O(n) must have dimension n(n . Hence O(n)/ must consist exactly of skewsymmetric matrices. (a. leaves the distribution /j. we can identify He with a subspace of Ge. X(a) X = i. we have left translations 7 X X a La : H and So > H. Y](e) Thus..
Usually. Having considered subgroups of Lie groups (and subalgebras of their Lie algebras). Lie algebra is isomorphic to Ihe lie algebra ifsome Lie group. then the condition 4>(s + I) = 4>(s) + 4>(1) implies that d4>(1 + s) ds = 0 gives d4>(s) . and every linear transformation is a group homomorphism. For example. homomorphisms between Lie groups.. 4>.. and % value 4>*eX at e. Consequently. ds which means that 4>(1) = cl for some c (= 4>'(0)). Later ) on we will be able to obtain a "local" version of this result.380 Chapter 10 There is a very difficult theorem of Ado which states that every Lie algebra is isomorphic to a subalgebra of G L( N . It is not hard to see that even a continuous 4> must be of this form (one first shows that 4> is of this form on the evaluating at s 4>' (1) = 4>' (0).P.X = L¢(aj.. we next consider. X + b4>. It then follows from Theorem 4 that ever . the map algebra homomorphism. suppose that G = H = JR.. . : � > Ij. (aX + bY) = a4>.. We will soon see to what extent the Lie algebra of G determines G... For any a E G we clearly have 4> 0 La = L¢(aj 0 4>.. then 4>. 1R) for some N. Y]. Thus X and % are 4>related. X = %(4)(a)). Y 4>. because IR is a vector space of uncountable dimension over Q. Yj = [4>. simply by 4>. that is.. But if 4> is Coo. we will denote 4>. : � > Ij is a Lie 4>. so if X E Ge. If 4> : G > H is a Coo homomorphism. then ..x is the left invariant vector field on H with 4>*a%(a) = 4>.4>. There are an enormous number of homomorphisms 4> : IR > 1R. X. 4>. We continue the study of Lie groups along the same route used in the study of groups. more generally. : Ge > He ..aLa. [X.
4>. P ROOF.(X. Notice that the only continuous homomorphism 4> : S' > IR is the 0 map (since {OJ is the only compact subgroup of 1R). there is a unique connected Lie subgroup K of G x H whose Lie algebra is f. We can identify £(IR) with 1R . Since <I> is a homomorphism. Now suppose that G = 1R. there is an open neigh borhood V of (e. but H = S' = IR/Z. then w : K > G is a Coo homomorphism. Let G and H be Lie groups. The continuous homomorphisms 4> : IR > S' are clearly of the form once again. we Can define Moreover.b. <I> (X)) = X .. Consequently. = <1>. 4>.e) E K such that w takes V diff eomorphically onto an open neighborhood U of e E G. X E g we have 4>. and G is connected.e} Ge is an isomorphism. For X E g we have SO w* : K(e. if there are two Coo homomorphisms Vr. X = <I>(X). : IR > IR is multiplication by c. for X E g. and <1> : g > l) a Lie algebra homomorphism. I K. giving rise to an identification of £(S') with 1R. . THEOREM. . If H2 : G x H > H is projection on the second factor.Lie Groups 381 rational numbers). . a Lie algebra ho momorphism g > I) may not come from any Coo homomorphism 4> : G > H. : IR > IR is just multiplication by c. Then there is a neighborhood U of e E G and a Coo map 4> : U > H such that 4>(ab) = 4>(a)4>(b) and such that for every when a. A neighborhood of the identity e E S' can be identified with a neighborhood of 0 E 1R. 7 w. then 4> = Vr. <I>(X)). Vr: G > H with 4>. Let I (German Fraktur k) be the subset f C g x l) of all (X. Clearly the map 4>.ab E U. If H. 5. However. Consequently. By Theorem 4. I is a subalgebra of g x l) = £(G x H). we do have a local result. and w = H. : G x H > G is projection on the first factor.
COROLLARY. then they are locally isomorphic. (X. (X. x so £(G') = f. b in a neighborhood of e. let 4> be the map given by Theo rem 5. As for the second. = (X. <I> (X)) = <I>(X). It follows that G is abelian. Since 4>*e = <P is an isomorphism. 6. By Corollary 6. there is a unique d4> dl Coo homomorphism 1 . PROOF. neighborhood of e generates G . The image G' of e is a Lie subgroup of G x H and for X E A we clearly have e. so ab = ba for a. if X E g. 1jr (a)). 1jr: G > H. define the oneone map = ][2. If two Lie groups G and H have isomorphic Lie algebras. <I> (X)). since (Problem 4) mH' . A connected Lie group G with an abelian Lie algebra is itself abelian . COROLLARY. PROOF. and that all connected Lie groups with a given Lie algebra are covered by the same simplyconnected Lie group.X Given 4>.�O . which implies that 1jr (a) = 4>(a) for all a E G . For every X E Ge. •:. •:Remark: For those who know about simplyconnected spaces it is fairly easy (Problem 8) to conclude that two simplyconnected Lie groups with isomorphic Lie algebras are actually isomorphic. G is locally isomorphic to IR". <I> (X)) = X. Thus G' = K. •:. then w. e: G > G x H by e(a) = (a. 4> : IR > G such that 8. 4> is a diffeomorphism in a neighbor hood of e E G . 7.X•  .382 Chapter 10 TIle first condition on 4> is obvious. so 4>. COROLLARY. Given an isomorphism <1> : A > I).
Uniqueness also follows from Theorem 5. 4>(s + 1) = 4> (s)4> (I ) and Clearly <I> is a Lie algebra homomorphism. Irl e/2 r �O ddl4> I Isl. they can be extended to all of IR using the method of the first proof.· 4>(e/2) · 4>(r) . .X. < o. then 1 1 is an integral curve of % which passes through 4>(s) at time clearly true for 1+ 1+ 4>(s) . By 111eorem 5. ls + 1 1 (:!. which proves uniqueness. then [4>(e/2) appears k times] k '" 0 4>(e/2) · 4> (e/2) · 4>(r) [4>( e/2) appears k times] k /: G > IR is Coo. If homomorphism.0 dl ' /(4)(1)4>('')) .( ) du u=o j' = L¢(r»XU) = %(4)(I))U). SECON ( D DIRECT) PROOF.:. 4>(1) 4>(s + I). The same is so 4> is a homomorphism. I 1 Thus 4> must be an integral cun'e of X. e) of 0 E IR there is a map 4> : (e. . . on some neighbor hood (e./(4)(1)) lim = h'l>O h = � / 0 L¢(l) 0 4> du u=o d4> = L ¢(l)' . 6") > G with . III. We know that integral cun'es of % exist locally. I = O.I ) and define 4>(1) = { 4>(e/2) · · . Conversely) if 4> : IR > G is an integral cun'e of %. and < < e 4>: IR > G is a Coo d4> ( f) = lim /(4)(1 + ")) .Lie Groups F IRST PROOF.4>* dl r�O To extend 4> to IR we write every I with II I '" e uniquely as 1 = k(e/2) + r k an integer. Define <1> : IR > £(G) by 383 <I>(a) = aX./(4)(1)) " 1.
Then all Coo homomorphisms 4> : IR > IR . with multiplication as the group operation.. Consequently: IA + BI IA B I :s: :s: I A I + IB I n I A I ' IBI: I .+ +.{OJ. . IR) where · now denotes matrix multiplication. must satisfy 4>' (1 ) = 4> ' (0)4>(1 ) 4> (0) The solutions of this equation are 4>(1 ) = = I. with 4> (S + I) = 4>(s)4>(I). where exponentiation of matrices is defined by exp(A) = I+ This follows from the facts in Problem 56. some of which will be briefly reca pitulated here. More interesting things happen when we take G to be IR ..{OJ. We thus see that there is a unique I parameter subgroup 4> of G with given tangent vertnr d4>/dl(O) E Ge.' I A lk :s: nk l A l k . GL(n. 4> (0) = I. Ii 2! 3! hence I A lk I A N+K AN (n I A I) N (nI A I )N+K +". (N + K ) ! :s: y.384 Chapter 1 0 A homomorphism 4> : IR > G i s called a Iparameter sub group o f G . 4>(1 ).nN! (N + K ) ! :s: nk. We have already examined the I parameter subgroups vi Iii . The solutions of these equations can be written formally in the same way 4> (1 ) = exp(I4> ' (O))."+ > 0 as N > 00. IR). If A = (a'j ) and I A I = max laij I. All Coo homomorphisms 4> : IR > must satisfy the analogous diff erential equation 4>' (1 ) = 4>'(0) .{OJ is just GL( I . Notice that IR . then clearly A A2 A3 + + +. e¢'(O)r.
Then exp(X) = </> ( 1 ) . so </> does satisfy (*) . (exp A ) (exp B). let </> : lR > G be the unique with d</>/dl(O) = X. If 1fr : G > H is any Coo homomorphism... h_O lim exp(I</>'(O) + h</>'(O)) .I . Given X E �.exp(I</>'(O)) h [exp(h</> (O)) 1 = lim h'l>O h</>'(O) = . and 0 is a regular point..Lie Grau/}. PROPOSITION. if </>(1) is defined by (* *).. 9.I 385 so the series for exp(A) converges (the J) th entry of the partial sums converge). and convergence is absolute and uniform in any bounded set. X) (exp 12 X) exp(IX) = (exp IX). then Coo homomorphism We clearly have Ge exp 0 1fr. Moreover (see Problem 56). if A B = BA. then </>' (1) = = (i. exp(11 + 12 ) X = (exp I . I] exp(I</>'(O)) +. we now define the "exponential map" exp : � > G as follows. The map exp : Ge > G is Coo (note that Ge '" lRn has a natural Coo structure). so that exp takes a neighborhood of 0 E Ge diff eomorphically onto a neighborhood of e E G. then exp(A + B) Hence. For any Lie group G.+O lim Ii + h2 </>'(0) 2 h 2! exp(I</>'(O)) = </>'(0)</>(1). h. = 1fr 0 expo exp G�H l � He l exp .
a) can be identified with Ge Ell Ga. exp(Vr. 0 Ell X). then also 4>.X) = Vr 0 4>(1) = Vr (exp X) .386 Chapter 1 0 PROOF. If 4>. dl r�O Consequently. (tX) = 4>(exp(IX)).. If we identify a vector V E (Ge) o with Ge. (X) = But then e = exp 4>.o is the identity. The tangent space (Ge x G)(X. Given Vr : G > H. and X E Ge. •: 1 1 1 0. which we know is exp X = projection on G of a ( l . Coo. PROOF.X. We define a vector field Y on Ge x G by Y(X.p( X (p)) = 0 for some nOllzero X E �.a) o f the Coo manifold G e x G a t the point (X. contradicting the fact that 4> is oneone. dl r �O Then Vr 0 4> : IR > H is a homomorphism with d(Vr 0 4» = Vr. . Therefore exp is a diffeomorphism in a neighborhood of O. and hence oneone. So exp. COROLLARY. let 4> : IR > G be a homomorphism with d4> = X.o(v) = = d exp(c(I)) dt v. .. Since it follows that exp is Coo. Then Y has a flow a: IR x (Ge x G) > Ge x G.a) = 0 Ell X (a).:. I 1=0 = d dt I 1=0 exp(lv) So exp. Every oneone Coo homomorphism 4>: G > H is an immersion (so 4>(G) is a Lie subgroup of H). o. then the curve C(I) = IV in Ge has tangent vector v at o.
4> ( expI. Yt l · · . exp X /4 E exp( t U ) .I . Every continuous homomorphism 4> : G > H is Hence Coo. Since 2 [4>( e/2)] = 4>(e) = [exp X/2f . Then on V. so X/4 = Y/2.. X E exp( ! U). then 2 a = exp(X/2) = [exp(X/4)] . so 2 2 exp(X/2) = a = b = [exp(Y/2)] = exp Y.. Let U be a starshaped open neighborhood of 0 E G. . . . COROLLARY. Choose a basis X" . PROOF Since X/2. Thus. Every continuous homomorphism 4>: JR > G is Coo. COROLLARY. X. 4> (se) = expsX 1 2. Now choose e > 0 so that 4>(1) E exp( ! U) for I I I � e. on which exp is oneone. then S o a has a square root i n exp( b = exp(Y/2) for Y E U.) = exp I Y. 4> (m/2n . Vr* a i = X"� so Vr is a diffeomorphism of a neighborhood U of 0 E JR" onto a neighbor hood V of e E G. For any a E exp( ! U). )m = [exp(X/2 n )]m = exp(m/2n . .. . PROOF. t U). ) ) = (exp I. i f a = b2 for b E exp( ! U). and thus everywhere. •:.Lie Groups 387 I I . Moreover. ( � IJ . Y E U it follows that X/2 = Y. X" for G. Xtl · · . The map 1 1+ 4>(exp IXi) is a continuous homomorphism of IR to H. By continuity. •:. I] . (exp In Yn ) . if a = exp(X/2) for X E U. This shows that every a E exp( t U) has a unique square root in the set exp( t U). . X) . for all s E [ . (exp In X. 4> = (4) o Vr ) o r ' . In ) = (exp l. it follows from the above that 4>(e/2) = exp(X /2). .) · · · (exp ln Xn ) is Coo and clearly (*) and (*) shows that 4> 0 Vr is Coo. e) = 4>(e/2. Let 4>(e) = exp X. By induction we have 4> (e/2n ) = exp(X/2n ) . Now the map Vr : JRn > G given by Vr (l" . So 4> is Coo at e. so there is Yj E He such that 4> (exp IX.
we have = I. y] + O(l3)} (3) exp lX exp IY exp(IX) = exp{IY + 1 2 [X.. [X. If G is a Lie group and 1 exp(IX) exp(IY) exp lX exp l Y = exp { 1 2 [X. Thus 0(13) will denote differenl functions at different limes. 2 The properties of the particular exponential map exp : IRn may !lOW be used to show that O(n) is a Lie group. then 12 . that is. the x ver (exp M)(exp M)t = I . (2) (I) exp lX exp l Y = exr l (X + Y) + '2 [X. IR ») > GL(n. THEOREM. we will denote it by 0(13). bounded for small I. Then I = A . to indicate that a function c : IR > Ge has the property that C(I)/13 i. IR). M = _Mt. It is easy to see that (= �l(n . •:. In the following Theorem. If G and G' are Lie groups which are isomorphic as topo logical groups. exp M E O(n). IR) exp(Mt) = ( exp M)t . Moreover. any A E O(n) sufficiently close t o I can be written A = exp M for some M. Y] measures. Y E Ge.388 Chapter 10 1 3 . It follows that a neighborhood of I in O(n) is an n(n . y] + 0(t 3 )}. In general. then they are isomorphic as Lie groups. Let At = exp N. up to first order. and in its proof. O(n) is itself a submanifold of GL(n . Since O(n) is a subgroup. 14.1)/2 dimensional submanifold of GL(n . At = (cxp M)(exp N). the equation exp(X + Y) = exp X exp Y holds whenever [X. hencc Mt = M. IR). For sufficiently small M and N this implies that N = M.1 = exp(M). then = (exp M) ( exp N) when MN = NM. so e p N = (exp M) . i. PROOF Apply Corollary 1 1 to the continuous isomorphism and its inverse . there is a diffeo morphism between them which is also a group isomorphism. Just as in GL(n.e. y] = 0 (Problem 1 3). since exp(M + N) (exp M)(exp M) So if M is skewsymmetric. COROLLARY. So exp Mt = At = exp(M). Y] + 0(1 3 ) \ j X. IR). C on sely. extent (0 which this equation fails to hold.
X(J) = X(J 0 La) = J(a . 2 .Lie Graul)j 389 PROOF.J(exp sX exp l Y ) = dt d = ( YJ) (exp sX exp l Y ) . exp uX). du u=o d YJ(a) = J(a · exp uY). d d (iii) </> (I) = .o 12 _ _ + "2 [Y ( YJ)](exp sX) + 0(1 3 ) . let Then l </>(1) = J(exp sX exp l y). Then we have '�. Now Taylor's Theorem says that Suppose that J(e) = (v) J(expsX exp l Y ) = J(exp sX) + I ( YJ)(expsX) _ _ o. (ii) d XJ(a) = Xa (J) = La. d d. (i) We have Similarly. </>(1) = </>(0) + </>' (0)1 + </> ) 1 2 + 0(1 3 ). du u=o _ I For fixed s. for any F. I J(exp sX exp l Y exp uy) Applying (iii) to YJ instead of J gives (iv) </>" (1) = [Y ( YJ)](exp sX exp I Y). Similarly. F(expsX) = (XF)(expsX) d2 F(exp sX) = [X(XF)] (exp sX) ds 2 F(expsX) = F(e) + s( XF)(e) + � [X (XF)](e) + 0(S3 ) . u u=o by (ii).
+ Z2 = . Applying Taylor's formula to t Z . (vii) J(exp tX exp t Y) = t [(X + YJ Jl(e) No\v for small t we can write +t2 [( t x = + XY + YY 2 ) J (e) + 0(1 3 ). and F = Y (YJ) gives (vi) J<exp sX exp t Y) + s2 _ = s ( XJ) (e) + t ( YJ)(e) t2 _ _ _ 2 [X (XJ)](e) + 2 [Y (YJ)] (e) + stX (YJ)(e) _ _ 2 2 + 0(S 3 ) + 0(1 3 ) + 0 (s t ) + 0(st ) . F = YJ. + t 2 Z2 + 0(1 3 ). = Z for some Z" Z 2 E Ge.  X + Y = 2. Z{I) = J(A(t)) + Since w e can take the . xx YY .+ X Y + 2 · 2 2 which gives thus proving (I). . = X + Y.  Z. so by (vi) we have (viii) /(exp Z (t)) = J (exp(t Z . (Zt /)](e) + 0(1 3 ) .390 Chapter 10 Substituting in (v) for F = J.f's t o be coordinate functions. If J(e) = 0. In particular. J)(e) + t 2(22 1)(e) t2 _ _ + 2 [Z. 2. ] exp tX exp t Y for some gives exp Z (I ) Coo function Z with values in Ge. comparison o f (vii) and (viii) gives 2. + t 2 Z2 )) + 0(1 3 ) = t ( 2. Equation (2) follows immediately from (I). then clearly J(A(t) + 0(1 3 )) 3 O{l ).
X. again choose J with J(e) = O. The work involved in proving Theorem 14 is justified by its role in the f ol lowing beautiful theorem. Let X.. then we also have (x) Comparing (ix) and (x) gives the desired result. 1 5. 3 + [ . We attempt to reconstruct the Lie algebra of H as f ollows.X.X. E H f all i.). (I) = largest integer � Then �I < I.X)J](e) + 1 + .)) + 0(1 ) 5. or PROOF.. with llle relative topologJ') that makes it a Lie subgroup of G. Ifwe write 0 ' (1 ) .) = (expl. > X and let I. 3 [(XX 2  yy + ""2 + 2 + X Y  XX  xx .51 ( . . ' 3 J(exp lX exp l Y exp(IX)) = J ( exp(lS. Let 1) c Ge be the set of all X E Ge such that exp lX E H f all I .:. J)(e) + 1 ' ( .)]  (e) ' 3 exp lX exp I Y exp(IX) = exp(lS. More precisely.YX .(t) ::: ::: I. or Proqf We can assume I.5. then H is a Lie subgroup of G. since exp(I. I Notice that f ormula (2) is a special case of Theorem 516 (compare also with Problems 516 and 518). Then X E lj.51 J)](e) + 0(1 ). # O . + 0(1 )). THEOREI'vl.Lie Groups 391 To prove (3). there is a Coo structure on H. Suppose expl. > 0 with each I. . > 0. ' 1. Then similar calculations give (ix) J(exp lX exp l Y exp(IX)) = I [(X + Y . .1 E H. E Ge with X. For I > 0. . Assertion J . + I S. let k. k. If G is a Lie group and H e G is a closed subset which is also a subgroup (algebraically) . + I S.J)(e) = 1 (.
X exp . . Choose positive I. . We clearly also have exp lX E H for 1 < 0. since H is closed and exp is continuous. Proq{ Choose a basis Then t/> is given by of G.X. exp(k. We now claim that I) c Ge is a vector subspace. we can write. Assertion 2. > G defined by t/>(X + X') = exp X exp X' Xl . .Y 11 11 1 I )n = exp {I(X + Y) + [X. Choose a subspace Il' C Ge complementary to I). but we will not even use this fact.(t)I. . . using (2) of Theorem 1 4 we see that [X. . QE. is a dilleomorphism in some neighborhood of O. Y E I). Thus exp IX E H.) k.X. Alternatively. so that G. If X. ( exp . Xk . The map t/> : G. sX E I) for all s E IR. > IX. . ). (t) > I. so X E I). . Xn X E 1). for fixed I. Y] E 1). 12 211 f taking limits as n > 00 gives exp 1 (X + Y) E H. = X + Y + Z(t. with . Clearly X E I) implie.k. = I) Ell Ij'.(t)I. Y] + O( l /n2) l . It dearly suffices to show thaI old if U is small enough. so that I) is a subalgebra. . t (l) E H. then H n exp(U) = exp(I) n U).X. Then exp(1) n U) is a submanif of G.392 so Now Olapter 1 0 I. [Similarly. X' E I)' X" . > 0 and let X.D. . we can write by (I) of Theorem 1 4 exp lX exp l Y = exp{I(X + y) + IZ(t)} where Z(t) > 0 as 1 > O. . Xk a basis for I). ) = [ exp (t.] Now let U be an open neighborhood of 0 E Ge on which exp is a diffeomor phism. Then Assertion J implies that X + Y E 1).
This is clear. Choose an inner product on �' and let K C I)' be the compact set of all X' E IJ' with I ::: IX'I ::: 2.Lie Grau!}. 1 it suffices to show thaI 1+ (a) . Choose integers JI j 'V'lith Assertion 3. let a E H n exp( U).D. .\ 393 Since the map . there would be X. X' E W'. since QE. If the assertion were false. ax 1jr* (�' I ) 0 = X. a contradiction. is a diffeomorphism in a neighborhood of 0 E ]R" . Choose a neighborhood = W x W' of Ge on which exp is a diffeomorphism. we can assume Xi' I/n. with U W a neighborhood of 0 E I) W' a neighborhood of 0 E such that W' is contained in V' of Assertion 3. There is a neighborhood V' of 0 in I)' such that exp X' ¢ H if 0 # X' E V'. Then a = exp X exp X' X E W. . and a E exp(1) n U) . .X. Since exp(I /I1. We can now complete the proof of the theorem. > 0.X. so 0 = X'. To prove the reverse inclusion. . and 4> of Assertion 2 is a diffeomor phism on W x W'. Clearly exp(1) 11' n U) CH n exp(U). •:. . it follows fi'om Assertion 1 that X' E I).D.') E H. Prorif. an) is a diffeomorphism of Ge onto jRn . . exp X E H we obtain exp X' E H. 7 Choosing a subsequence jf necessary.L7= a. )(11. QE. .)0 X' E K.' E 1)' with X/ 0 and exp X/ E H. Since a.
of Lie groups are better expressed in terms of f orms. iJ. Y) = w(e)([X. . cvn (e) span G/'. . Clearly. also left invariant. wn(e) is the dual basis to or then any Coo vector field X can be written Xl . fj Xj j=1 Then f or Coo functions /i so 0/ is Coo. .Y(w(X)) . . . 9]) = w([xj ']). /\ o/k = L Aj o/ J f certain COI/SWI/Is aj. the bracket being the operation in A. but man� propertie. H ence dw(e)(X. . If wi (e).··· <iJ.. The interplay between left im"ariant and right invariant vector fields is the subject of Problem I I.394 Olapte. L ai1 . . . The f ormula on page 215 implies that for a ieI'. A f orm w is called left invariant if La'w = w f all a E G. It f oHows that any left invariant f orm is If (J) is left invariant. then every left invariant kform i� i]<. H ere we consider the case of f orms. " " Xn E Ge. then f a E G we have or CO:: . Y) = X(w(l7)) .w([X. . (J)n are left invariant I f orms such that cv I (e). Y]). . This means that or web) = La * [w(ab)]. X = 'L. .o} I /\ . 1 0 U p t o Ilo\V: w e have concentrated on the left invariant vector fields. = so dw i. La* dw = d(La*w) dw. a left invariant kform w is determined by its value wee) E Qk(Ge) ' Hence� i f cv l � .. . . . . invariant Iform (J) and left invariant vector fields X and Y we have dw(X.
(X) = X for X E Ge. then A is abelian (converse of Corollary 7). Now X is the tangent vector at t = 0 of the curve 1 1+ explX. If W is left and right invariant. PROPOSITION. this tangent vector is just x. So 1jr"X is the tangent vector at 1 = 0 of 1 1+ (exptX).Lie Group. then 1jr*w is right invariant. then dw = o. Let 1jr : G > G be 1jr(a) = a. (3) If W is a left and right invariant kform. 1jr* (we) = ( _ J )k We.I . (2) It clearly suffices to prove this for k = J. .! 395 16. Since o/*w and (J) are both left invariant. So dw = 0 for all left invariant Iforms. (I) Clearly so If (J) is left invariant. then so o/*w is right invariant.. we have dw is also left and right invariant. P ROOF. so The f orm But So dw = o. Yj = 0 for all X. (4) If G is abelian. (2) If We E Qk(Gel. The converse is similar. then all left invariant If orms w are also right invariant. It follows fi·om (*) that [X.1 = exp( I X). Y E A. So it is enough to show that 1jr. then 1jr * we = (3) (I) A f orm W is left invariant if and only if (4) If G is abelian. (_J)k we.
It turns out that (2) is exactly what we obtain fi'om the relation d 2 w k = O. From skewsymmetry of [ . First choose XI . Y E Ge we Allernale proqf qf (4). Y] [Y.:. � j k=l n The numbers Ci) are called the constants of structure of G (with respect to the basis XI . Con dition (2) is thus an integrability condition. wn(e). . ' " X E Ge dual to Wi (e).j � . . By Theorem 1 4. . . it foHows that for a basis w 1 . X)] = " Cik Xk . . dwk in terms of the 0/ /\ wi . . . . ] and the Jacobi identity we obtain (I) Ci) = C)� L( C! C/'k + cJk cL + Cl'iC/') ) = . of of invariant Iforms we can express each such that [Xi . we can prove (Problem 30) that if Ci� are constants satisf ying (I) and (2). . There are constants Ci� n Since . if G is abelian. . we obtain [X. [Xi. . ll=l dwk = . X)] = L Ci�Xk : k=1 dearly we also have . . Letting 1 > 0. (J)1I in a neighborhood of 0 E jRn such that .L Ci� wi /\ wi i. In fact.396 Ozapter /0 X. then for have Hence HX. Y] + 0(1 3)/1 2 = HY. = dw is left invariant for any left invariant w. then we can find evelywherc linearly independent I forms cv 1 .L Ci� Wi /\ w) i<j n (2) From (*) on page 394 we obtain = o. " " Xn of � ). Xl . X] + 0(1 3)/1 2 .
iiJn are each linearly inde pendent. . so on J(p. the existence of such (J/ implies (Problem 29) that we can define a multiplication (a. . a). M x G is foliated by IIdimensional manifolds whose tangent spaces at each point are annihilated by all iJk .wk Choose G and let r be the folium through (p. [i} M "2 : > a.iii ) . 397 Moreover. THEOREM. • . Let G be a Lie group with a basis of left invariant I forms w'. This means that r contains the graph of a diff eomorphism J fi·om a neighborhood U of p to a n eighborhood of = = Then "2 : > a E a). . ei /\ ej de L i<j Then f every p or M there is a neighborhood U and a diffeomorphism J: U G such that ei J*wi PROOF. . wIn and constants of structure Ci� . w'. . en and wi. = E > IJ . . far 1 7 . . From this latter fact and (a suitable local version of) Theorem 5 we could immediately deduce the following Theorem. . b) 1+ ab in a neighborhood of 0 which is a group as as it can be and which has the (J/ as left invariant I forms. . . wn are linearly independent everyvvhere.L Ci� [iJi /\ (iJj _ w j )+ (iJ i _ wi ) /\ wj ]. Let . a} .Lie Group. . Let '" : M x G M and M x G G be the projections. = > cI(iJk . . Let Mn be a differentiable manifold and let e . . for which we supply an independent proof..a} where {jk iiJk = 0. which is the set of vectors in (M x G)(p.i< Ci� ([iJi /\ iJj ] . e n be everywhere linearly independent I forms on M satisf ying k . .iJ n .. i<j By Proposition 714.. Now iJ'. . /\ wj J) L j . . _ " > . the sets 8 1 . . ." Ck.[a'. Hence . : r M and r G are each diffeomorphisms in some neighborhood of (p.
Then fi and .(I)y.Ii : M > G be two Coo maps such that . (�I ) [Ly. .l ) w (Y2(1)) ( [ Ly. [ (L y. M C�2) == Y2* W == = == It follows thai U we regard Yl a� b riven.j* H2* a/ (H. then it becomes an ordinary dificrclltial equation for dY dt = 2 (d. (I) _ ' J. there is a (unique) a E G such that j.iii ) == (q.(I).:. J(q)) c r. )* W (Yz (l )) ] . THEOREM.iiJk = 0 On J. and let fi . (I)y. (I)) (�IJ IR and the fwo maps YI ' Y2 : IR > G satisji Y2· For every left invariant Iform w == (d. j* H tek .Ii * (w) for all lefi im·ariant I forms w.(I)y. we have 0 = j*( {i .1 ) YI* w (�IJ * dI · Y2 (of the type considered . let G be a Lie group. that is.Ii differ by a left translation.(H2 ° i)* a/ ek J*w k . we have YI (O) == Y2(O) .398 Let OUipter 10 j: U > M x G be t h e map j( q ) Since (jk . (I). 0 j)* e k . == PEDESTRIAN PROOF.li . Let M be a connected manifold. w (y. == == == j. _ == == == It is also po"ible to say by how much any two such maps differ: 1 8 . and write this equation out in a coordinate system.' ] d y.*(w) La o . We must show that YI W (Y2(1)) Case 1.
is an integrable distribution on G x G. are arbitrary'. Let 7[... it follows that heM) is contained in some left coset of t. b) E G x G. Choose a E G so that y.(G) c G x G is the diagonal subgroup {(a. (t) = a . Gelleral case. Since M is connected.(O). ELEGANTPROOF. = y. = y. By Case 2. y. If (J) is a left invariant If orm.lr2*(J/ ). Now define h : M > G x G bv h(p) = Uj (p)./i (p). if t. M = IR. 399 in the Addendum to Chapter 5).Lie Group. theu the maximal integral manif olds of t. . .. But this solution is clearly y.(G). Let po E M. there are a .:* . (0). For (a..*(w).(G). /2 (p)). (0) = y. . s o /2(1') = a · .. let Ll(a../i *(w) = y/(w) . arc the left cosets of t. a) : a E G}. In fact. but the maps y" y. i n particular for t = I . w" for the left invariant Iforms. y. . . = f. (t) for ali I . 0 c.(O) = a · ydO). " Then t. then Since La 0 y. Case 3.(O) = y. Choose a E G so that (La 0 y. . b E G with . so it has a unique solution with the initial condition y. we have For any p E M there is a Le t y. aJd p ) = bJ ( p) for all p E M .. By assumption.*(w) = c* J 2(W) = c* .. Choose a basis w ' . Case 2. )*(w) = yt ( L:w) = y/(w) = y. In other words.b} = i=l n ker(Jr) *a/ . . y. Then Coo curve c: IR > M with dO) = Po and c ( I ) = p. it follows fi'om Case 1 that La 0 y.. : G x G > G be projection on th e ith factor...
COROLLARY. For example. PROOF. then J for a unique a E G. the form Ra*w is left invariant. 20. we can define Coo an an 7 Since L Ja '" a" is llsLlal1y kept fixed in any discussion. . and if J : G IR is a Coo function v'lith compact support. th e left invariant nforms are also very important. ''\'e can. left invariancr of implies that a" in other words. But in one case they coincide. of c. If G is a connected Lie group and J: G > G is a map preserving left invariant forms.400 Ozapter 10 1 9. also consider right invariant l1forms. so L Jan = L Lb*uan) = LU o Lb )Lb*a" = LU o Lb )aN• g(a) J(ba). this is often abbreviated to L J or fG J(a)dCl. If is a left invarjant nform. So { G c IR is a compact connected subgroup of IR{O}. == where I note that Lb is an orientation presenTing diffeomorphism. so there is a unique real number with J(a) = Ra'w = l(a)w. For each a E G. •:. PROPOSITION. an left invariant nform� arc a constant multiple of any nonzero one. ( ) = {I} . These generally turn out to be quite diff erent from the left invariant IZforms (see the example in Problem 25). iG /(a) da L /(ba) da. If G is compact and connected and w is a left invariant Jlform� then (J) is also right invariant. Clearly.ourse. Suppose w # O. = La While left im'ariant Iforms play a fundamental role in the study of G. The latter notation has advantages in certain cases. = \\'hich proves the formula]. Hence J G () l(ab) J(ba) = J(a) · J(b). then determines an orientation on G.
(I) For any a E G.. But also .) (Y(u .Lie Graul). So = fy(. Clearly Ie reverses geodesics through e.. so it is an isometry on Ge.e. the maps I 1+ exp(tX) for some X E A. which reverses geodesics through fa (y(t)) = y (I). : Ge > Ge is just multiplication by . In the case of a compact group G there is always a Riemannian metric on G which is both left and right invariant.I (see the proof of Proposi tion 16 (2)). let or y (1 + u ). Since fa = Ra1e Ra . then (2) The geodesics y with y(O) = e are precisely the I parameter subgroups of G. G. a. )) on G by We are finally ready to account for some terminology from Chapter 9. the map le* : Ga 7 Ga.l is also an isometry.I)) = y(t + u) = y(u + 21). the map fa : G 2 1 . 401 Vife can also consider Riemannian metrics on G. PROOF. i.e. ) is any Riemannian metric we can choose a biinvariant IIf orm an and define a biinvariant (( . In fact: if ( . F fixed I. Since for any a E G.) : it is dear that fa is an isometry reversing geodesic through (2) Let y : IR > G be a geodesic with y(O) Y(U) = = e. (I) Since the map 1. PROPOSITION. Then y is a geodesic and y(O) fy(ljfe( Y (u)) = = Y(I) .)(y (u)) fy(. Let G be a Lie group with a biinvariant m etric.. if y > G given by 1a ( b) = ab' a is an isometry is a geodesic and y (O) = a. i.
y(l1I) = y(t)n for any integer n. Vi . x Mp > V.. . If V is a ddimensional vector space. . '. . Xk )v. . Vd is a basis for V. By continuity. . . . . we define a Vvalued kform On M to be a function w such that each w(p) is an alternating map w(p) : Mp x . . It follows by induction that If t f = n'! and t ff = n"t for integers n' and n". . . .: i=1 we will write simply ror anv Vvalued kform w we define a Vvalued (k + I)form dw by d dw = L do/ . since there are I parameter subgroups with any tange nt vector at I = 0. . Xk E Mp we have or k times d w(p)(X" " " Xk ) = L w.402 so Chapter 10 y(t)y(u)y(t) = y(u + 21).. .. so y is a homomorphism on Q. then there are ordinary kforms cv l . . then y(t' + I") = y(l)n'+n" = y(t')y(t "). •:. .' If VI . . . . These are the only geodesics. We conclude this chapter by introducing some neat f ormalism which allows us to write the expression f dwk in an invariant way that does not use the or constants of structure of G. (p)(X" . y is a I parameter subgroup. Vd for V. and geodesics through e are determined by their tangent vectors at / = o. i=l d w = LO/ ' Vi o i=l a simple calculation shows that this definition does not depend on the choice of basis VI . . (J)d such that f X" .
/\ ryj · p(u. where V and V have . Using the bilinear map [ . . for any Avalued kform ry alld any Avalued Iform A on G. namely the form w defined by w(a)(X(a)) = X E A. These concepts have a natural place in the study of a Lie group G. Uc and 403 p: V" . • • V x V > W is a bilinear map. Vd. Uc or VJ . . respectively. a calculation shows that this does not depend on the choice of bases u" • .Lie Groups Similarly. Although there is no natural way to choose a basis ofleft invariant Iforms on G. Now suppose that Xl " ' " Xn E Ge = A is a basis. we have. The form w defined by (*) can clearly be written Then (I) . Vj ) i=l j=l is a Wvalued (k+I)form. and that cvI . there iI a natural nvalued If orm on G . . suppose bases UJ . Vd . • . . vf . . ]: A x A > A. a new Avalued (k + I)form fry /\ A] on G. If w is a Vvalued kform c w = L {J/ ' Ui i= 1 ry = I >j . We will denote this Wvalued (k + 1) form by p ( w /\ ry).. . j=l d and ry is a Vvalued Iform then c d L L w. • . (J)n is a dual basis of lefi invariant I forms. • • • .
Y] is left invariant [ . Y E G" then dw(a)(Xa. V = Y for some X. Yo) = 0 . ] in G. . then (Problem 20) we can define a �valued function V(f) on G.404 On the other hand. w(V) is a �valued function on G. For the term dw we just modif the formula in Theorem 71 3: If V is a vector field y on G and J is a �valued function on G. w(a)(Ya)] It f ollows that for any vector fields since by definition of Problem 20 gives an invariant definition of p(w/\ry) and shows that this equation is equivalent to the equations of structure.w(a)([X. For vector fields V and V we can then define dw(U. V) = .[W(V).. I f we choose V = X. V]).W( V)] · I V and V we have } [X. On the other hand. V) = V(w( V)) . I dw(U. Recall that the value at a E G of the right side depends only on the values Va and Va of V and V at a. Chapter 10 [X" Xj] = L C.: Xk . Y]a) = w(e)([X.0 .w([V. by definition of w. For the present we merely wish to point out that the terms dw and [w /\ w] appearing in this equation can also be defined in an invariant way.V(w(V)) . we obtain the equations of structure of G : The equations of structure of a Lie group will play a n important role in Volume III. Y]) = [X. Y] = [W(a)(Xa). h) = w(e)([X. k=1 n so (2) Comparing (I) and (2).
HOJ(U). The appearance of the factor 4 bere has nothing to do with the 4 in the other f orm of the structure equations. makes their dOJ equal to � of ours for Iforms OJ. Tbis makes their A /\ ry equal to 4 of ours for Iforms A and ry. ) as L dOJ.\ 405 WARNING: In some books the equation which we have just deduced appears as dOJ(U. /\ dx . V) = . Then the definition of deL OJ. dx .Lie Gmu/i. OJ(V)]. . It comes about because some books do not use the factor ( k + /)!/ k! I! in the definition of /\ .
1 n M xkl (A) = " MaI A ka . 5. . and suppose that (x. � a= 1 (a) Show that this means that (It is actually clear a priori that M defined in this way is left invariant.) (b) Find the right invariant vector field with value M at 1 . Show that the closure ii of H is also a subgroup.L Mxkl . We have computed that f or M E �l(n. then H is closed. E U. xy) . and hence an element of O(n). an for a. then G is acompact. e) is a regular point of . U a neighborhood of e E G. 4. (a) Show that I takes straight lines to straight lines. Let G be a topological group. (a) If H is open. Let I : IRn > IRn be distance presen'ing. 6. and H e G a subgroup.I when I : G x G > G x G is I(x.ak ax k. (c) Show that I is a linear transformation.f(c) Conclude that G is a Lie group. with I(O) = o. IR) we have where M= . (b) Conclude that Un u n = G. Let G be a topological group and H e G a subgroup. Show that the tangent bWldle TG ofa Lie group G can always be made into a Lie group. y) 1+ (a) Find I. f LA. (a) Show that U n +1 is a neighborhood of u n . . 7. (b) Show that I takes planes to planes. then so is every coset g H.y) (b) Show that (e. 3. r for A E O(n) and r a translation. xy 2. Let G be a connected topological group. or LA since LA is linear. = (x. and Let u n denote all products a l . (d) Show that any element of £(n) can be written A . (Use Problem 3. .) (c) If G is locally compact and connected. lRlA = . l .406 Chapter 10 PROBLEMS is I. M(A) = A · M E IRn ' = GL(n. (b) If H is open. Coco Let G be a group which is also a Coo manifold.
Then (A. (a) Use Lemma 513 to show that (exp sX)(exp t Y) = (exp t y)(expsX). 1 2. then 4> can be extended unique!. where V e G is an open neighbor hood of c with V · V. show that 4> and ljr are equal even if they are defined only on a neighborhood U of e E G.sin a sin a cos a 13. (c) £(GO) is isomorphic to [£(G)]O = AO. (e) Use this to give another proof that A is abelian when G is abelian. In particular. can be made into a covering space of G. then ljr is an isomorphism from G to Go. (a) Show that exp QJ) Use the matrices A and B below to show that exp(A + B) is not generally equal to (exp A) (exp B). exp(X + Y) = (exp X)(exp Y). Show that Corollary 7 is false if G is not assumed connected. In Theorem 5. on some smaller neighborhood of c. ]) is isomorphic to £( GO). and hence to �o. Y]. 10. ]. 8. ( �) O I ) . Show that the set of all ? equivalence classes. Let X. b E V. use Theorem 5 to show that exp is a homomorphism on the subspace of Ge spanned by X and Y.) if ljrl = ljr. (b) AO is a Lie algebra. [ . consider pairs (V. If G is a group. for all c E G. we define the opposite group GO to be the same set with the multiplication . I I. ] be the operation on Ge obtained by using right invariant vector fields instead oneft invariant ones. and where ljr : V > H satisfies ljr(a) · ljr (b)' = 4>(ab . (d) Let [ . with operation [ . and X 1+ X is an isomorphism of A onto AO. ljr). b = b · a. Define ( VI . Y]0 = [X. If A is a Lie algebra. we define the opposite Lie algebra AO to be the same set with the operation [X. provided that U is connected.I . and if ljr : G > G is a 1+ a. b.Lie Groups 407 (a) For each C E G. (b) Conclude that if G is simplyconnected. A= ( 0 a a 0 ) ( = cos a . (a) GO is a group. Y] (� �) = B= O. ljrd ? ( V" ljr. . (b) More generally.I C U. defined by a. ab E U 9. Let G and H be topological groups and 4> : U > H a map on a connected open neighborhood U of e E G such that 4>(ab) = 4>(a)4>(b) when a. Y E Ge with [X.' ) for a.· to a homomorphism of G into H .
etracc M . (e) Usc this fact to give a fancy proof that trace MN = trace NM. : AI(n..408 Chapter 10 (a) Show that 14. 0 0/. and let ry. � dry. Y] = 0. V) be a coordinate system around e E G with x. (c) For G = S' . the map det. IR) > IR . (b) Using Corollary 51 2. show that for the homomorphism det : GL(n. compare Problem 715. (a) Let (x. Show tbat no matter how small U is. (The diagonalizable matrices are dense.. (Since det. 1 5. N]. Suppose that [X. ( Y(o/. (a) If M is a diagonal matrix with complex entries. {ryrl is generated by X + Y. = <p. IR) > £(IR . (e) Conclude that it holds for all M with complex entries.NM) = trace[M. 1 6. 0 <p.) (g) Now usc this result and Proposition 9 to give a fancy proof of (c). there is a neighborhood U of e E G such that every element in U has a unique nth foot in U. show tbat there is no neighborbood U which has this property for all n. (a) Let U be a neighborhood of the identity ( 1 .{Oll = IR is just M 1+ trace M. and trace are homomorphisms. 0) of S' (considered as a subsct of IR'). Let . show that det exp M = (b) Show that the same equation holds for all diagonalizable M with complex entries. Problem 1 3 implies that exp t (X + Y) = (exp tX)(exp t Y ) i f [X.}. show that In other words.(p)) + <p. Y j = O.) (d) Using Proposition 9. (b) Show tbat for each n ::: I . it suffices to look at matrices with only one nonzero entry.{O}. 1 7. {o/s}. (p)). = 0/. A more general result holds.) (0 Prove the result in part (d) directly. (p) = X(ry. (Look at trace(M N . witbout using (c). there are elements a E U which have square roots outside U in addition to their square root in U. ( e ) = O. Let X and Y be vector fields on a Coo manif old M with corresponding local I parameter families of local diffeomorphisms {<p.
Hint: This f ollows immediately from one of our propositions. the automorphism group of A. Hint: Join any a E H to e by a Coo curve c . and show that the tangent vectors of c lie in the distribution constructed in the proof of Theorem 4. usually Ad(a)(X) is denoted simply by Ad(a)X. (a) Ad(ab) = Ad(a)oAd(b). . see Yamabe. . Thus. 19. 1314. 2 (1 950). The argument of part (a) is essentially equivalent to the initial part of the deduction of (I). . H is a Lie subgroup of G. . isomorphic to G L(n .) (b) Let K e G be the connected Lie subgroup of G with Lie algebra �. (c) Let CI ' ' ' ' . Ck be curves in H with {c. show that (b) If Ol. where AUI (A). . . such that every a E H can be joined to e by a Coo path lying in H. Let � c Ge be the set of tangent vectors to all Coo paths lying in H. Ck (tk ) . IR) if A has dimension n). . LaRa . (b) Show thai exp(Ad(a)X) = a ( exp X )a . For a E G. The map is denoted by Ad(a).' . is the set of all nonsingular linear transformations of the vector space A onto itself (thus.l (b).].) 18. (Not even the full strength of (I) is needed. . it suffices to know that exptX exp t Y = exp{t(X + y) + O(t)}. Thus we have a homomorphism Ad : G > Aut(�). The map Ad is called the adjoint representation. then H is a Lie subgroup. By considering or k the map J(t l . show that K C H. Show that (e. Osaka Math.e) > G are differentiable. consider the map b 1+ aba . It is even true that H e G is a Lie subgroup if H is path connected (by not necessarily Coo paths). (a) Show that I) is a subalgebra of Ge• (Use Theorem 14. On an arcwise connected subgroup ifa lie group.'(O)) a basis f �. t ) = cd t l ) . and let H C G be a subgroup of G (algebraically). fJ : (01 ' fJ)'(O) (c) Also deduce this result from Theorem 1 4 (1). Show that H C K.l = = 01 ' (0) + fJ'(O) . (d) If H e G is a subgroup and an immersed submanifold.Lie Groups for 409 Coo functions J. Let G be a Lie group.
nt space of Aut(�) . JR). This tangent space is isomorphic to End(A). ifand only if [X. (It suffices to show this for M in a neighborhood of 0. w e have the map Ad*. with basis X(J) = L Xp (Jl) i=] . . Vd. (e) Since Ad : G > �. . V J . Yj E � for all X E �.) The map (ad X) ' Ig + ad X + . Yj is denoted by ad X E End(�). LxY. (a) Let J : M > V.) (d) Show that Ad(exp tX)Y = Y + t [X. where J = "L1=J Jl . For Xp E Mp. Vl for Jl : M > JR. . Show that H is a normal subgroup of G if and only if I) = £ ( H) is an ideal of � = £(G). (f) Conclude that Ad(exp X) = exp(ad X) = c . (O)(Y) = dt dI 1=0 c(Y). Y E I). (A proof may also be given using the fact that [X . define Xp (J) E V by d 20. JR) show that J . we let it operate on Y E A by (Compare with the case A = JR".+ . AlIt(�) = GL(n. (X)(y) = [X. � (= Ge ) > . . . where End(A) is the vector space of all linear transformations of A into itself: If c is a curve in Aut(�) with c(O) = 10. that is. End(�) = n x n matrices. Y] + 0 (1 ' ) . .. . �ange. where V is a finite dimensional vector space.) Use (d) to show that Ad. . . Yj. 2! (g) Let G be a connected Lie group and H e G a Lie subgroup. Vd for V. Y j Y 1+ [X.JR) and M E � 1 (n . .410 Chapter 1 0 Ad(A)M = A Mr (c) For A E GL(n. . Show that this definition is indepen dent of the choice of basis VJ . then to regard c'(O) as an element of Aut(A).at the Jdenuty map 10 of � to Jtself. • . Vl.
Lie GmU/>5
411
(b) If w is a Vvalued k form, show that dw may be defined invariantly by the formula in Theorem 713 (using the definition in part (a)). (c) For p : U x V > W show that p(w /\ ry) may be defined invariantly by
p(w /\ ry)( Xl " ' " Xk , Xk+ , . . . , Xk+l ) l I = L sgn a · p (w( Xq(l j , . . . , Xq(kj ) , ry( Xq(k+lj , ' " , Xq(k+/) )) . ! l! k q eSk+1
Conclude, in particular, that
[w /\ w](X, Y) = 2[w( X ), w( Y )] .
(d) Deduce the structure equations f rom (b) and (c).
21. (a) If w is a U valued kform and ry is a V valued Iform, and p : U x V
W, then
d(p(w /\ ry)) = p(dw /\ ry) + ( _ I ) k p(w /\ dry) . + [w /\ ry] = ( _ I ) k l l [ry /\ w] .
>
(b) For a �valued kform w and Iform ry we have
(c) Moreover, if A is a �valued m form, then
22. Let G c GL(n, IR) be a Lie subgroup. The inclusion map G > G L(n , IR) > ' ' IRn will be denoted by P (for "point"). Then dP is an IRn valued Iform (it corresponds to the identity map of the tangent space of G into itself). We can also consider dP as a matrix of I forms; it is just the matrix (dx;j ), where each 2 dxU is restricted to the tangent bundle of G. We also have the IRn valued I form (or matrix of I forms) pl . dP, where · denotes matrix multiplication, and pl denotes the map A l+ A1 on G. ' ' ' (a) pl ·dP = p(pl /\ dP), where p: IRn x IRn > IRn is matrix multiplication. (b) LA*dP = A . dP. (Use J*d = d .) J* (c) pl . dP is left invariant; and (dP) . pl is right invariant. (d) pl . dP is the natural �valued Iform w on G. (It suffices to check that pl . dP = w at I.) (e) Using dP = P . w , show that 0 = dP . w + P . dw, where the matrix of 2forms p . dw is computed by formally multiplying the matrices of I forms dP and w . Deduce that
dw + w · w = O .
412
If w i s the matrix o f Iforms w
=
ChapteT 10
(w u ), this says that
/\
d
wij = _ L Wik k
(J)kj .
nience, denote the coordinates x
23. Let G C GL(2, IR) consist of all matrices with a # ll and on GL(2, 1R) by and
Check that these equations are equivalent to the equations of structure (use the form dw(X, Yl = [w(X),w( Yl].)
(a) Show that for the natural �valued form w on G we have
w
so that and are left invariant I forms on G, and a len invariant 2form is /\ (b) Find the structure constants for these forms. (c) Show that
dxjx dy)jx2 (dx dyjx
=
xl' � ( dX dy ) x o '
0
(� � )
x
o.
For conve
y.
(dP) .
p l
= ; ( d;
y dxo+ Xdy )
and find the right invariant 2forms.
24. (a) Show that the natural �((n, IR)valued I form w on GL(n, IR) is given by
where
(b) Show that botb tbe left and right invariant n ' forms are multiples of
(det(xap) ) I
n (dxJ l
25. Tbe special linear group SL(n, IR) C GL(n, IR) is tbe set of all matrices of determinant I .
(a) Using Problem 1 5 . sbow tbat its Lie algebra ,:;((n, lR) consists o f all matrices v\·ith trace =
/\
• . •
/\
dxn l)
/\ . . . /\
(dx1n
/\
. . •
/\
dxnn) .
o.
Lie Grouj>.,
(b) For the case of SL(2, IR), show that
413
where we use for differentiating the equation = I. (c) Show that a left invariant 3form is
/\ /\
vdx ydu vdy J'dV PJ . dP= ( udx+xdu udy +xdv ) ' x, y, u, v xvxll,yXJ2, x2J, x22. u v dx du dy  y dx du dv.
=
Check that the trace is 0 by
26. For M, N E 0(11)
£ (0(11))
=
{M : M
=
/\
_ Mt}, define
/\
(N, M) =  trace M . Nt. (a) ( , ) is a positive definite inner product on 0(11). (b) If A E 0(11), then (Ad(A)M,Ad(A)N)
=
(M, N) . , ) at 0(11)/ is also right
(Ad(A) is defined in Problem 19.) (c) The left invariant metric on 0(11) with value invariant.
27. (a) If G is a compact Lie group, then exp : A > G is onto. Hint: Use Proposition 2 1 . (b) Let A E SL(2, IR ) . Recall that A satisfies its characteristic polynomial, so A'  (trace A)A I = 0. Conclude that trace A' � 2. (c) Show that the following element of SL(2, IR) is not A ' for any A . Conclude that it is not in the image of expo
+
(
_0 o
0  1 /2
)
(d) SL(2, IR) does not have a biinvariant metric.
28. Let be a coordinate system around e in a Lie group G, let Kj : G x G > G be the projections, and let be the coordinate system around (e, e) given by y i = Xi =i = 0 Define 4/ : G x G 7 IR by
x
O]l"J , xi ][(2y. ,:)
414
Chapter 1 0 X, b e the left invariant vector field o n G with X; (e) = a ;
and let
(a) Show that
n a X; = " 1frf aXl j= l
� I, .
.
L
. ,
where (b) Using
LaLb = Lab, show that l La.X; (b)](xl ) = [X,(ab)](xl ).
n / L Vr/ (b) . aqj (a, b) = 1fr!(ab). az j= ] (V,}) be the inverse matrix of 1fr (1fr/), we can write n aq/ _ j (a, b) L 1fr,I (ab) . 1frj; (b). az I=J
= =
. aq/ 1frf (a) = a ; (a, e). o
Deduce that and then that
Letting
V,
=
This equation (or any of numerous things equivalent to it) is known as lie's.firsl J undamental theorem . The as,ociativity of G is implicitly contained in it, since we used the fact that LaLb = Lab. (c) Prove the convene o lie's .first Jundamenwl theorem, which states the following. j ' Let </> == (</> ] , . • • ,</>n ) be a differentiable function in a neighborhood of 0 E lR n z [with standard coordinate system y l , . . . , yn, : I , . . . , n] such that
</>(a,O) = a 1frJ (O) 8j I _; a</> I j a.. (a, b) L 1fr; (</>(a, b)) · 1frj (b)
= _ 
for
a E }R1I .
Suppose there arc differentiable functions 1frJ in a neighborhood of 0 E IR" [with standard coordinate system x I , . . . , xn ] such that
II
i=l
for (a, b) in a neighborhood of O E IR'''.
Lie Grauj).,
415
Then (a, b) 1+ 4>(a, b) i s a local Lie group structure o n a neighborhood o f0 E JRn (it is associative and has inverses for points close enough to 0, which serves as the identity); the corresponding left invariant vector fields are
[To prove associativity, note that
=
n a Xi = L,Vrf axj ' j=1
.
and then show that 4>(a, 4>(b, z )) satisfies the same equation.]
a4>1 (4)�:; b), z) t Vr! (4)(4>(a, b), z)) . v,,} (z) 1=1 [Xi , Xj] L, Ci)Xk k=1
=
29 . Lie's secondJ undamental theorem states that the left invariant vector fields a Lie group G satisfy
Xi of
for certain constants  in other words, the bracket of two left invariant vector fields is left invariant. The aim of this problem is to prove the converse if Lie� second undamental theorem, which states the following: A Lie algebra of vector fields f on a neighborhood of 0 E jRn) which is of dimension 12 over IR and contains a basis for JR"o, is the set of left invariant vector fields for some local Lie group structure on a neigh borhood of 0 E JRn. (a) Choose
Ci�
n " ,'f'/,Ji dxj ) L j=l then the 0/ are the dual forms, and consequently dwk   " C�· wi Wj Ci� constants. L lJ i<j (b) Let Hj : JR" x JRn JRn be the projections. Then H2'Wj HI'Wj t(v,,/ OH2) [d(xl OH2)  t(Vr! OH2) ' HI'Wi] . ;=1 1=1
0/
=
If
X" . . . , Xn in the Lie algebra so that Xi (O) a/axi lo and set n a Xi L,Vrf. ax} ' j=1
= =
>
=
/\
416
Chapter 1 0
ndimensional manifolds on which the forms d ( xl 0 ][2)  ,£7=1 (1frf 0 ][2) . ][ 1 *Wi all vanish. (c) Conclude, as in the proof of Theorem 17, that for fixed G, there is a function <Po : JRn > JR" satisfying <Po(O) = G and or equivalentl:;
Consequently, the ideal generated by the forms d (xl 0][2) ,£7= 1 (1frf 0][2)'][I*W ' is the same as the ideal J generated by the forms ][/w j ][ I *wj . Using the faci that the ej are COllstants, show that d ( J ) C J. Hence JRn x JRn is foliated by
k
d<P� (b) = L 1frf (<Po(b)) . w' (b), ;=1
=
30. lie's thirdfimdamenlal Ilzearrm states that the ej satisfy equations (I) and (2) k on page 396, i.e., that the left invariant vector fields form a Lie algebra under [ , ]. The aim of this problem is to prove the callverse if LU's thirdf undamental :/;mrem: which states that any ndimensional Lie algebra is the Lie algebra for SOlDe local Lie group in a neighborhood of 0 E JR". Let ei) be constants satisfying equations (1) and (2) on page 396. We would like to find vector fields Xh . . . , Xn on a neighborhood of 0 E JR" such that [X,. Xj ] = '£;;=1 e,) Xk · Equivalently, we want to find forms w' with
Now set </>(a, b ) rem.
=
<Pa ( b), and use the converse ofLie's first fundamental theo
a cp� ax j (b )
_ L 1fr,I (<Po (b)) . 1frjj (b). ;=1
n
do}
=

Then the result will follow from the converse of Lie's second fundamental the orem. (a) Let h� be functions on
L C;1 o/ /\ wj . i< j
JR x JRn such that
a h� at
=
ok
r
_
'\' ck
i, j
X L IJ
h;(O,x)
=
O.
j 'IIr
These al'e equations "depending on the parameters x" (see Problem 55 (b)). Note that h;( t , O) = O�I, so that h� ( l , O) = Let ak be the I form on JR x JR" defined by
o�.
Lie
Grauj!.'
417
and write
where ),.k and C{k do not involve ),. k =
j (ik = dxk  L Ci� x ia j . i,)
(b) Show that
ahk L ( ah!.k  ', ) dx'. /\ dx!.. ax!. ax' l<
�
dak = ),.k + (dl
/\ C{k). dl. Show that
(c) Let
Show that
dek
=
dl /\  " eA x'),.j  " " cAe' x' a' /\ a j L lJ L L IJ rs i,} i,} r. � + terms not involving dt.
.  � " " e  2 L L ( ik rs j e' _
(
)
Using
" L ek er.� a /\ aj " Ij " L i,} T,S
i,} T,S i,) T,S
k r eis e'j ) a' /\ a j j
'  � " " (ek ers + ek e'T ) a' /\ aj , is  LL I
2
j
and equation (2) on page 396, show that
dek
=
" I 5J dl /\  " eA x'),.j + � " L dT e' x' a' /\ aj L IJ 2L i,} i,} T,S + terms not involving dl.
=
(
)
Finally deduce that
dek
dl /\  L ej� xj e/ + terms not involving dl. j,/
418
(d) We can write
Cizaj)ter J ()
where gt(O.X) = 0 (Why?). Using (c), show that
k ek = '" gIJ. dxi /\ dxi . L i<j
Conclude that (e) We now haw
ek
=
O.
), k =  I L. c.IJk·ai /\a . . dak
=
L Ci� a i /\ a) + (dl /\ Cik ) . i,j Show that the (arms ,,/ (x) = a k ( l,x ) satis(,' do}  � L C;} Wi /\wl i,j
=
�
2
I,)
j
CHAPTER 1 1 EXCURSION IN THE REALM OF ALGEBRAIC TOPOLOGY
de Rham cohomology, of fundamental properties of the ordinary cohomology which is studied in algebraic topology, Because we deal only with manifolds, many of the proofs become significantly easier, On the other hand, we will be using some of the main tools of algebraic topology, thus retaining much of the flavor of that sul�lect. Along the way we will deduce all sorts of interesting con sequences, including a theorem about the po"ibilitv of imbedding nmanifolds in JRlI + l . Let M be a manifold with M = U ]I for open sets U, ]I C M, Before examining the cohomology of M we will simply look at the vector space Ck(M) of kforms 01 1 M, Let
Tspaces of a manifold. Our main results will be restatements, in terms of the
his chapter explores further properties of the de Rham cohomologv vector
U
iu : U > M iu : U n ]l > U
iv : ]1 > M iv : U n ]l > ]I
Q'
be the inclusion!o). Then we have two linear maps
and {3.
defined by
O'(ev) = (iu"(ev), iv'(ev))
Here iu'(ev) is just the restriction of ev to U. etc. Clearly fJ 0 0' = (): In other wOI'ds, imagc O' C ker fJ, 1\101eovel, the converse holds: ker fJ C image O'_ For, if fJ O" , A 2 ) = 0, then A I = A 2 on U n ]l. so we can define ev on M to be A I o n U and A 2 o n jl and then O'(ev) = O" , A2) The equation image 0' = k e r fJ i s expressed by saying that the above diagram is exact a t the middle vector space. '>\Ie can extend this diagram by putting the vector space containing only 0 at the 419
420
C1wpter J J
ends: the al·ro"� at either cnd of the following sequence are the only possibk linear maps .
I . LEMMA. The sequence
0 +
is exact at all places.
Ck (M) � Ck (V) Ell Ck (V) � Ck (V n V) + 0
PROOF. It is clear that Cl is oneone. This is equivalent to exactness at Ck (M). since the image ofthe first map is {OJ C Ck (M). Similarly, exactness at Ck (V n V) is equivalent to {J bein!! onto. To prove that {J is onto, let {</>u , </>v } be a partition 0(" unity subordinate to {V, VJ. Then W E Ck(V n V ) is w = (J(</>vw, </>uw).
where
</>vw denotes the (orm equal to </>vw on V n V, and equal to 0 on V  (V n V ) :
. •.
By putting in the maps d.
\\"(:'
can expand our diagram as follows,
0  C k (M) � Ck (V) Ell Ck ( V) � C k(V
()
�
Ck + I (M) � Ck +I (V) Ell Ck+ I ( V ) L Ck + I (U n V) � 0
jd
j
j d Ell d j
j
jn
jd j
V) ..., 0
.
j
so that the roWS arc all exact. It is easy to check that this diagram commute� . that is, any tvo/O compositions {rom one vector spaee to another are equal:
_ Cl_ (d Ell d) O Cl = Cl o d fJ d o {J = fJ o (d Ell d)
J d Ell d jd
= d
l _Cl_ l _fJ_
d Ell d
b:cunioll ill the Realm 'IfA(�ehmi( 7ojJolog}
421
OUf first main theorem depends only on the simple algebraic structure in� h erent in this diagram. To isolate this purely algebraic structure, we make the following definitions. A complex C is a sequence of vector spaces C' . k = 0, 1 , 2, . . . , together with a sequence of linear maps
satisf ying dk+' 0 dA = 0, or briefly, d2 = O. A map "' ; C, > C2 b e tween complexes is a sequence of linear maps
such that the following diagram commutes for all k . C,k
d/
l Clk + l � cl+
j
. "''_. C:'
j d! r r
The most important examples of complexes are obtained by choosing C" = C" (M) for some manifold M, with dk the ope ato d on kforms. Another example, implicit in our di cu ssion, is the direct Sum C = Cl EB C2 of two complexes, defined by
s
For any complex
C' we
can define the cohomology vector spaces of C by H " (C)
=
ker d' image dk  '
·
Naturally, if C = l C' (M)), then Hk(C) is just Hk(M). If "' ; C, > C2 is a map between compl exes. then we have a map. also denoted by "'.
x E C,k with
that every dement of H" (Cil is determined by some = o. Commutativitv of the above diagram shows that d/(",k(x)) = ",k +' d, " (x) = 0, so ",k(x) determines an element of Hk (C2), which we define to be ",(the class determined bv x). This map is welldefined.
To
d e fi
ne ", we not
d,"CY)
e
if V and V arc overlapping portions of S 2 fOJ which there is a deformation retraction of V n V into S l . .J J for if we change x to x + dl k .I (y)). Whal does this imply about the maps Ci: Hk(C. This is nol true .f* : Hk (M) > Hk(N). For example. then Cik (x) is chanl!ed to Cik (X + dl k . Cik (x) + d/I (Cik . ) > Ci u v . and Ci : Ck (M) > Ck (N) is J* for J : N > M. then this map is jus. �ow suppose that \·ve have an exact sequence of complexes fJ Ci o > C > C2 > C3 > O. ) > Hk(C2 ) and fJ: Hk(C2 ) > Hk (C3)? The nicest thing that could h appen would be for the following dia gram to be exact: fJ Hk(C2 ) > Hk (C3) > o.I (y)) = Cik (x) + Cik(dl k .I (y) for some y E C/ . C/ = Ck (N).422 GiW/)le..I . I \·vhich re ally 1l1eans a vast c om mutative diagram in \·vhich all rows are exact. then we have an exac i sequen lT 0 > Hk(C. When Clk = Ck(M).I (y)) = which determines the same element of Hk ( C2 ).
1 (y') . something very nice is true: o II o II IR: 2. If 0 > CI > C2 > complexes) then there are linear maps '" fJ C. > Hk (Cr) > Hk (C2) > Hk(C3) > Hk+l (Cr ) > . Moreover. and hence z = o. So we have to show that V. THEOREM. so z determines an ele ment of Hk+I (C ): this element is defined to b e ok of the element of Hk (C3) I determined by x. · E d/ . I ci with = PROOF. this implies that dl k+1 (0) = 0.k with fJk(y) = x. Throughout the proof. Bv exactness of the middle row of (*) . This means that dl (y) = sO we choose d/ . Ho (Cr) > H0 (C2) > H0 (C3) > H I (Cr) > . .k+I (z) for some (unique) Z E C k+I .I (y') = fJkdl . . there is y E C. thus dle)') ". Then x = tli (x) = O.)' 423 an exact sequence II 1'\ evertheless.le obtain 0 E Hk+I(Cl ) if we start with an element of the form di .I (x') for x' E C. 0. we must check that the result does not depend on the choice of x E cl representing the element of Hk (C3). fJ '" 8 .J". Theil o = dl (x) = dl fJk (y) = fJk+1 dl (y) .k+ l . . In this case.k . let x' = fJk . In order to prove that ok is welldefined. diagram (*) should be kept at hand.I (x') = dll fJk . . .I Cr') as .I (y').k+1 is oneone.on in the Realm qfAlgebraic but not 70/l0/0t. So dl(y) E ker fJk+1 = i lnage". > 0 is a short exact sequence of so that the following infinitelv long sequence is exact (everywhere): 0 > Since ". Let x fJ '" /.EtCU/:r.I .
and a deformation retraction or V Viet oris sequence i� n 11 onto 2 circles. Let x E q k satisfy dl k (X) = 0. All of Theorem 816 can be derived directly from the followin. and suppose that O'k(x) E cl represents 0 E Hk(C2 )' This means that O'k(x) = dlI (y) lor some y E C/. the :Maver. ".. then we have an exact sequence (eventually ending in O's): o� U HO l M ) .. If M = U ]I. this is left to the reader. � As several of the Problems shO\.. 3.. the cohomology of nearly everything can be computed by a suitable application of the MayerVietoris sequence.� corollarv of Lemma I and Theorem 2. As a simple example.·. It is a worthwhile exercise to check that the main step in the proof of Theo rem 81 6 is preei se lv the proof that kerO' C image 0. the definition of 8 i mmed i ately shows that the image of this element under 0 is precisely the cla" represented by x . together with the first part of the proof that 0 is welldefined. THEOREM (THE I\1AYERVIETORIS SEQ UENCE). and the open sets 1) and V illustrated below. . The proof that the sequence is exact consists of 6 similar diagram chases. onto circles. Moreover.424 Glza/Jler J J 1 t is also necessary to check that our definition is independent of the choice of y with fik(y) = x . HA +1 (M) > .I .I (y) represents alI element of HkI (C3). '''Ie will supply the proof that kerO' C image o. ''\'(' consider the torus T = S l X S ' . Now So fik. •:. Since there is a deformation retraction of V and r H A (M) � HA (U) Ell HA ( V) > HA (U n )1) 2. .'h ere V and V are open. .
H ' (U n VI � H' (T) � o. . +:+ 0 = dim V2 /a( Vd . then 112 7 . So the kernel of this map is I dimensional.dim VI + dim V2 . 7 Vk l 7 Vk 7 0 0 = dim h . By induction on k. For k I We have the sequenu Exactness means that {OJ C VI is the kernel of the map VI > 0. = . . PROPOSITION. Thus the image of the map H (U) Ell H I (Y) > H I (U n V) is Idimensional. it induces a map V'.J vector spac("� hence which proves the theorem for k. which implie> that VI = O .I dim Vk . The reasoning used here can fortunately be systematized.J .. . So we have an exact sequence of k .Excursion in the Realm 'IfAlgelJ1'aic T j O JOlogT 425 � HO (TI � HO (U) Ell HO (VI � HO W n VI � H ' (TI . so its kernel is I I dimensional. .dim V2 + dim V3 . If the sequence o 7 VI + a is exact. " III E1l IR 11 IR � H ' W) Ell H ' (V) � IR /I IR The In ap H I (U n V) > H 2 (T) is not 0 (it is onto H 2 (T)). Moreovel. this map is oneone. = PROOF.dim V3 + . Aswme the theorem for k . and consequencly the map H I (T) > H I (U) Ell H I (V) has a Idimensional image.a( VI ) > V3 .dim V3 + . " + ( _ I ) k . It follows that dim H I (T) = 2. Similar rea soning shows that this map also has a I dimensional kernel. Since the map V2 > V3 has kernel a(Yd. 4.
. 1 t is a difficult I heorem that every ('OC manifold has a triangulation. we will use the MaverVietoris sequence to relate the dimensions of H k (M) to an entirely difT�rent set of numbers. Elcmc77la�J' Dijjerf1lial Topology. Now by a triangulation of a compact nmanifold M we mean a [mite coHection 1 allj} of diffeomorphic ima. /j. m . arising from a «triangulation" of M.. 1 .".n = Jx ) E < < jR'"+1 . m is called a k face of !:J.426 CizajJter J J Rather than compute the cohomologv of other manifolds.X i .vard 3simplcxc . ( 0 0 1 f S' (afiter SI ' �... . l>o \ (In Problem 8S . then for some k the intersection ani n an} is a k ·lacr of both alii and an}.. .. n allj #. standard tflangulaooll b' In al id \Z!J"lria.���:oI1S v � A !1\ � � � � � or 2 4 \f mters(.0. ��_< 0V ��':he .'cUom of � . The standard nsim pl ex /j. . �. 0 .=l xi = Ji J.j The subsel of /j. C M is a dif leomorphic image 0(' some /j.�+ I L. .. k . Geometric Integration 1 . . a ne\\' structur� which we will now define. ''''hitney. n which cover M and which satisf the following condition: y If a n.I and . J " " . although homeomorphic." obtained by setting n .. If /j. .!!cs of /:). ..::." is defined as the set /j. and is called a kface of /j. set." is defined to be a dillcrent.k of the coordinates xi equal to 0 is homeomorphic to /j.ft ' €SO . then the image of a kface of /j. for a proof see !\1unkres.
Now let U be the disjoint union of open balls. each ani an nsimplex of the triangulation.I. will be called a ksimplex of the triangulation .1 where V n Vn _ 1 has the same cohomology as a di�oint union of Gi copies of sn. We let ctk be the number of these ksimplexes.> Hk (M) .I i � H n I (M ) � H . any kface of any an.> Hk (U) Ell Hk (Vn_ I i II n II o 11 .I )simplexes of M. and let Vn_ 1 be the complement ofthe set consisting of the centers of these balls.I (U) Ell H I ( Vn_d � � H� I (U n Vn_d n � Hn(M) � H ( u ) Ell Hn ( Vn_ J l II . n II 0 VnJ l > Hk ( U n Vn.l . The :MayerVietoris n sequence breaks into pieces: (2) For I < k � H I ( U ) Ell H I ( V J l � H I (U n II o < Hkleu II (I n Vnd .Excunioll ill the Realm rljAlgebraic TO/Jolog} 427 Theor Assuming that our manifold M has a triangulation {an.I i (3 ) H n'(U II nn Vn. Then M = V U 1111.} we will call ). one within each nsimplex ani. Consider first the case where 11 > 2. so that Vn_ 1 is a neighborhood of the union of all (n ..
The above equations then imply that X ( Vn . defined bv n X (M) = dim Ho(M) .al + a2 .2 L ( . For any triangulation of a compact manifold M we have PROOF..] X ( M) = X ( Vnd + ( . one for each (n .(. X ( M) = aO . .2 This makes sense for any manifold in which all Hk (M) are finite dimensional: we anticipate here a later result that Hk (M) is finite dimensional whenever M is compact. . THEOREM.i l = dim Hk ( M) O ::: k � n .I (M) .'.dim H (M) + a.I ) "a . joining the balls of the old U. 111 the manifold Vn.t l = n.I )face. . � comp ( "eWUonents of (11 = 2) . dim Hk ( Vn .l ( Vn _ l ) = dim H n.I )k dim H k ( V.I For the case 11 = 2 we easily obtain the same result without splitting up the sequence._ t l = X(M) .l [dim H .. ..428 Chapter J J Applying Proposition 4 to these pieces yield.dim H n (M) + an .I ) " an · L ( . dim H n. .I we define a new open set U which consists of a disjoint union of sets diff eomorphic to IR".dim H I (M) + dim H'(M) .. We now introduce the Euler characteristic X(M) of M. = k �O n. + (I)" dim H (M).I )k dim Hk(M ) or 5. " k =o n n + ( _ I )n.l (M) . + (I) n an .
.. .. . . . each of which is smoothly co Iltractil)le to a point. joining the centers of the balls in the old U. . � .I ) Q'n . we have X(M) = X(Vn!) + (I )"O'n = X(Vn .3.' . + ( .1 ) 1 0'1 + .11 the Realm q[Algebraic 7opolog1 429 vVe wil1 let 1111 2 be the complement of arcs� in the new V.2 ) + [(. . /<� .Lxcuniol1 .2 IS . and F faces..I)" IO'n _1 + ( . while in an other cases we have Combining these equations. E edges. . Hence X( Vol = 0'0.. If a convex polyhedron has vertices. ' (11 = 3) An argument precisely like that which proves the equation X(M) = X( Vnl l + ( .I )"O'n] = x(Vo) + [(. . • . the. COROLLARY (DESCARTES. .I )"O'n also shows thai Similal'I)� we introduce VIl.EULER). the last of these is a disjoint union of 0'0 .E+F=2 . + ( .I )"O'n] n = 0'0 .rt. Vo. ('omplf'ment of � I/n. : V 6. .0'1 + . thell V . . . . . . . .
. . and let {</>u .hat O" .. To prove . If V c M is open. C V n V. </>v} be a partition of unity for 'he cover { V ..(V) satisfies iU'(A t ) + iv'(A2) = 0. If C: (M) den otes the vector space of kforms with compact suppOrt on M.</>vw) E C. thi' shows that support i" C V n V and support A. we encounter a very different situation. a form (J) with compact suppOrt C M may not restrict to a fOfm with compact suppOrt C V: the inclusion map of V into M is not proper. we can define a new sequence. (U) EIl C.he converse. This means that AI = i. I is clear h ac is exact. . V}.. suppose . Then w t t = </>uw + ¢vw E C. It is clear that image Uu' Ell iv') C ker ( i u' + iv'). Since support At C V and support A. C V. So (A. we will denme this extended form by i u'(w). (V) EIl C. A. then w can be extended M by letting it be 0 am side V. in fact.) iu' Ell iv' is oneone. ) is the i mage of At E C:(V n V ) .430 ChapteT Jl If we turn from Hk to H. if w is a form with compact support C U. let (J) b e a kf())'ffi with compact support on M. On tho' �'"::. support w U . ( V ). To prove that iuT +iv' is OntO. @support w LEMMA." w other hand. •:. is clearly 'he image of (</>uw. The sequence PROOF. each map iu' and i v' is oneone. A. 7.
whel'e (U x IR) n (V x IR) is diffeomorphic to snI x 1R2 The only way to use induction is to find H. 0 a= O. (M) This sequence is much harder to work with Ihan the M ayerVietoris sequence.. so Ihat unv is diffeomorphic 1 0 S. . V open in M. > IR wilh A = fJ *( X). Apply Theorem 2 10 the short exaci sequence of complexes given by the Lemma . = O. The details will be left to the reader. Then A WI � w2 L wJ A there is �: W. •:. > H. If M = long exact sequence .. U U V fol' U. .I x IR. .. which is diffeomorphic to S. ( M)' > [H:(U)EIlH:(V)]* > H. . A IR j . i..on in the Realm qfAlgelnaic 1O/!% g) 431 8. Now suppose A E satisfies a*(A) fJ ) = A 0 (fJ 0 a) = A 0 0 = o. (V) > H. and then proceed to yet anO! her application of Theorem 2. . + 8 H.e. then there is a dual .I x lR . PROOF. . starting with S l X IRm. . So a* 0 fJ * = O. suppose we want to find H.:' I. THEOREM (MAYERVIETORIS FOR COMPACT SUPPORTS). (UnV)' > WI a W2 fJ PROOF. Ifwe write S" = UuV in the usual way. Ihen sn x IR = (U x IR) U ( V x IR).EtcU1..+I (UnV)' > H.+I (U n V) > . for all sn x IR"'. then there is a long exact sequence . > H: (U n V) > H. For example. We claim that = fJ 0 t Given a w E W3 which is . COROLLARY. then so is the sequence of dual maps and space� For any A E W3* we have a* fJ* (A) = a*(A W. we "vi}} merely record one further resu1t� for later use.'" /. (U) Ell H.* W3* � W]* 0 � WJ* . 9. If the manifold M = U U V for U.. for IR" {OJ. We just have to show that if the sequence of linear map' )0 )0 W3 is cxaCi at W2.. V open in M.
we define a map of complex". '>\'e now construct l1 sequence of such neighborhoods V = VI :::J V2 :::J V3 :::J . Moreover. /I' + e C: (M) + i* Ck (N). We will call Wi and Wj wil Vl = Wj I VI · l t is clear that we can make the set o f all equivalence. This defines X on fi( W2 ) C W3.w" = 0'(=) for some z. and define X to be 0 on W. vVc appeal first to a result fJ"Om the Addendum to Chapter 9.A(W") = AO'(=) = O.k ( N ) . Finally. • • with ni Vi = 1\'. sO that we obtain a complex g. •:.k (N) > gk+ I (N). j such that Wj E Ck(Vj). so A(W) . while j 0 lf is smoothly homotopic to the identity of V. vVe now consider a rather different situation. we define �(W) = AW). There is a compact neighborhood V of N and a map l[ : V > N such that V is a manifoldwith· boundary.. We therefore have the sequencL C:(M . then w . the "germs of kfonns in a neighborhood of M". we will have to use a technical device. To circumvent this difficulty. while the image of e contains all W E e: (M) which are 0 in a neighborhood of N. Now chooSt W e W) with W3 = fi( W2) Ell W." classes into a Vector space f'/ (N).432 of the form Chapter Jl fi (w'). Now consider two forms Wi E Ck(Vi ). then l[ 0 j is the identity of 1\'. C: (M) in the obvious way: (J) � the equivalence class o r any w i lli . equivalenl if there is f > i. it is eas\" to define d : g.N) where e is "extension". This makes sense. Then M N is also a manifold. This sequence is 1lot exact at C! (M): the kernel of i" contains all W E e: (M) which are 0 on N. Let N C M be a compact sub manifold of M. for if fi (W' ) = fi(w"). and if j : N > V is the inclusion. + j" g.
(M). Clearly e is oneone.N. LEMMA. LEMMA. PROOF. and A = e(AIM . and l� represents the same element of fJ. which we will need later. If W E C. Let M be a manifoldwithboundary.k (N)) PROOF.N). > H. Finally. Vi = N. PROOF. then there is an exact sequence N c . 1] be a Coo function which is 1 on v.k (N). and then use Lemma I I . having support I C + interior Vi .k (N) > 0 I I . and consequently W = 0 on Vi . (M) > Hk (N) 10. (M . Details are left to the reader. (M . + e C. Then there is an exact sequence + o H. This follows easily from the fact that j* : Hk ( Vi) isomorphism for each Vi . .k (N) as 1 ] . then W '= 0 in some neighborhood U of N. with compa�t bound ary aM. The sequence 433 0 > C. Since N i.N) > H. Then l� E C. If compact submanifold of M. THEOREM (THE EXACT SEQ UENCE OF A PAIR). Apply Theorem 2 to the exact sequence of complexes given by Lemma In the proof of this theorem. B y definition o f fJ. THEOREM. ! .) of the complex {fJ.f�). > Hk (N) is an M is a 1 2 .k (N) is represented by a form � on some V. suppose A E C/ (M) satisfies i*(A) = O.ExcU):rion in the Realm ifAlgebraic Topolog). entered only as an intermediary (and we could have replaced the Vi by their interiors). •:.N) > . consequently this element is i*(. this means that A Wi = 0 for some i . This means that i*e(w) = O. .N) is exact. Let I : M > [0. •:. 1 3 . Hk (fJ. the de Rham cohomology of the manifoldwith boundary V.N has compact support C M . Conversely. Hence AIM . •:. +I (M . there is some i such that Vi C U. But in the next theorem. it is the object of primary interest .N). (M) + i* fJ. . compact and n. (M . The cohomology vector spaces are isomorphic to Hk (N) for all k. any element of fJ. 1 0.
of aM in M. •:.434 C hapter J J PROOF.n+1 . Just like the proof of Theorem 1 2. dim H"(M) + 1 ? 2 . M) gives H. {p) X �m). (S" x �m).'(�"+ I ) � H"(M) � H. . M is the boundary of each component. with H. (�") from a knowledge of Hk (S" .'+I (�"+I . ( B) "" H k ( B) = 0 for k '" o. PROOF. THEOREM. using tubular neighborhoods V. The reader may use Theorem 12 to compute O R O It follows thaI (*) number of components of �"+I . we can rederive H. But we also know (Problem 825) thaI (**) number of components of�"+ 1 . Then Theo rem 13 may be used to compute the cohomology of S" X sm .I ) . by considering the pair (S" x �m. Let M C �"+ I be a compact ndimensional submanifold of �n+ 1 (a com As a simple application of Theorem 1 3 . then M is ori entable.M has exactly 2 components. the sequence of the pair (�"+ I .M ? 2 . pact "hypersurface" of �"+I). From (*) and (**) we obtain 14. Using Theorem 817. U R n H. by choosing M to be the closed ball B in �". and ffi.M = dim H" (M) + 1 . If M c �"+I is a compact hypersurface.M) � H:+I (�"+I ) � H"+I (M).I = a(S" x closed ball in �m). I\1oreover. For our next application we wiIJ seek bigger game.
then jRn+1 . so every point of M is in the boundal"\' of each of the h¥O components. 1 7. Then U can be described as U = 11x : x E f a certain function p: or Sn I  and 0 5 I < p(x)) S. If p is Coo. If M C jRn+1 is a submanifold homeomorphic to sn. there are a number of technicalities involved. 1 5.Excunillll ill the Realm ifAlgelna. x E U. LEMMA. However. we conclude that dim Hn (M) = 1. . then we can prove that U is diff eomorphic to the open ball B of radius 1 in jRn. and M is the boundary of each.. If the radial function p of a starshaped open set U C jRn is Coo..M has exactly two components. COROLLARY. then H shows that jRn+1 . so M is orientable. which we will have to dispose of first.M. We will call p the radial function of U. •:. so a modification is necessary.M has two components. The basic idea of the proof is to take Ix E B to p(X)1 . 16. Neither the projective plane nor the Klein bottle can be imbedded in jR3 OUf next main result will combine some of the theorems we already have. This produces difficulties at 0. Consider a bounded open set U C jRn which is starshaped with respect to o. COROLLARY (GENERALIZED [COOl JORDAN CURVE THEOREM). The proof in Problem 825 shows that every point of M is arbitrarily close to points in different components of jRn+1 . then U is diffeomorphic to the open baJJ B of radius 1 in jRn.I > JR .t 7ojJO/ogy 435 Since dim Hn(M) is either 0 or 1.
and hence in U. This means that for .I . ° /(J) I . Let I < 1. .I .436 eha/ller J J PROOF. 1 8 . 1] be a Coo function with = Define h : B / ° in a neighborhood of ° /' 2':. 1] > We can assume. Since U is open. so it is Coo. that p [0. I t is the identity i n a neighborhood of 0. At each point x E sn. I > p(x) .J' E W we have p (y) 2':. / : [0. it might not even be continuous.e for all Y E W.e . with a nonzero Jacobian. at 0.11 is "lower semicontinuous": for every E: > 0 there is a neigh borhood W of x in sn. •:. •:. I Coo. At any other point the same conclusion follows from the fact that I r+ 1 + (p(x) . There is clearly a neighborhood W of x with the property that for . However� the discontinuities of p can be of a certain form only. > U by x E sn.1' E W the point IJ' is in B. without loss of generality. (he function p need 110t be h(lx) = [I + (p(x) . In general.I such that p()') > p(x) . 1 on sn. I < e. 1L 0 ::: 2':. there is an open ball B with x E B C U.1)/(1) is a Coo function with strictly positive derivative.1 )/(I)]X.I . Choose IX E U with p(x ) PROOF. = Clearly h is a oneone map of B onto U. the radial function p of a starshaped open set V C ffi. LEMMA.
(U ) "" IR "" H. IVx. Since 1>dx) + .I such that Ix may also be used as 1. for then V is diffeomorphic to IRn . and consequently w = d� where � has compact support contained in V. 19.EXCU1J. (IRn) for all k. . Similarly. and hence in U. and define Any point X E snI is in a ceI"tain sllbcollection of the J. . and we will be content with proving the following. Let IVx" . . •:. PROOF. 437 Even when p is discontinuous. .11 the Realm ifAlgebmic ToJ)olog).1' for all Y E W.. lxl is < p(x). Since K is closed and p is lower semi K C V = {IX : x E SnI and 0 :s I < p(x)}. choose Ix < p(x) such that all points in K of the form ux for 0 :s u :s p(x) actually have u < Ix. Each Ix" . For each x E sn. (x) are o. . it looks as if U should be diffeomorphic to IR". .1lx. .I . then Hk (U) "" H k (IR") and H. continuous.Oll . p. If U is an open starshaped set in IRn. cover S. . . . + 1>1 (x) = 1. . . . 1>. be a partition of unity subordinate to this cover. it follows that p(x) < p(x). K C V. (lRn). say WX1 " ' " Wx/ for convenience. . We claim that there is a Coo function p : snI > IR such that p < p and This will prove the Lemma. (U) "" H. . .I . By Theorem 817. We also know that H. let 'PI . since U is smoothly contractible to a point. there is a neighborhood Wx of x in sn. Let w be a closed kform with compact support K C U. LEMMA. Then PI+dx). Proving this turns out to be quite a f eat. The prooffor Hk is clear. we just have to show that H. .. . (U) = 0 for O :S k < Il.
then expp establishes a diffeomorphism of V with an open starshaped set in Mp . . . and . every point has a neighborhood V which is geodesically convex. 3. we can also choose V so that for any p E V the map expp takes an open subset of Mp diffeomorphic ally onto V. where V is a disjoint union of balls or around 1 .:. Ii follows from Lemma 19 that V has the same Hk and H: as IRn. and choose a Riemannian metric for M. such a cover will be called nice. clear for PROOF B v induction o n the number o f open sets r in a nicc cover. V" V) of M.438 Chapter II We can applv this last Lemma in the following way. . It is fairlv clear that if we consider /II = { I . Then the theorem is true for V = V. . . . is nonempty.} such that each nonempty intersection has the same Hk and H: as IRn.. } as a subset of 1R2. Suppose i t i s true for a certain 1'. . Let {V" . this shows Ihat H' (M) is infinite dimensional (see Problem 7 for more informalion abOUI the cohomology of M). II 0 where M n V has the same H' as a disjoint union of infinitely many copies of S ' . In general. PROPOSITION. U · · · U V. 2. If any V = Vi. n· . According to Problem 932. . . } be a finite cover by such open sets. (M) are finite dimensional for all k . Ii i s r = 1. a manifold M will be called of finite type if there is a finite cover {V" . If p E V. To prove this rigorously. On the other hand. 2. then Hk (M) and H. . 20. . 3. and consider a nice cover {V" . If M has finite type. . n Vi. then M = 1R2 . then V is clearly geodesically convex. . V. We obtain H' (1R2) II 0 � . H' (M) (fJ H' ( V) II 0 � H ' (M n V) � H2(1R2). V. Let M be a compaci manifold. . /II is not of finite type. we first use the MayerVielDris sequence f 1R2 = M U V. . . .
by the same f ormula. So Hk (M) must be finite dimensional. the main part of the proof is called a Lemma. There is then a unique element of H. . (a v f3). in order not to plague ourselves with additional technical details.(M) represented by any � E C. •:.ExcuIJion in the Realm ifAlgebraic 7ojJolog} 439 for U. V n V. but we shall restrict the theorem to manif olds of finite type. Now consider the MayerVietoris sequence . /J. (M) and the isomorphism H�'(M) > IR which takes this element to I E 1R. with orientation IJ. or For any manif old + a Hk (V) (fJ Hk ( V) > ' " . and the theorem itself is a simple corollar\'. .. M we can define (see Problem 831) the cup product map Hk (M) x HI (M) "::' Hk+/ (M) r+ by ( [a!]. (M) + /J. . the "Poincare dual" of a. > Hk. . and the kernel of a is also finite dimensional. so P D(a) (f3) = v f3 E H. As with most big theorems of algebraic topology.k (M)* by f3 r+ 1 (M. It is also true for V n V.) � = J.p. a We denote this element of that we have a map H. . Now every a E Hk (M) determines an element of the dual space H. . since this has the nice cover {V n V. Now suppose that Mn is connected and oriented. to denote both this element of H.l (V n V) Ii H k (M) + The map a maps Hk (M) onto a finite dimensional vector space. We are all set up to prove this fact. [�]) We can also define [a! I\ �]. (M) is similar. ] . 1R.k (M)* by PD(a) . (M) with It is convenient to also use /J. The proof f H. . One of the fundamental theorems of manifold theory states that P is always D an isomorphism. since a! 1\ � has compact support if � does. .
Suppose that the rows are exact. ffi n � ffi n 1>1 .(M)· [H. Hence "THE FIVE LEMMA". This proves the ."2(Y) . in which the top row is the MayerVietoris sequence. and is left original Lemma . 1>3 the reader. PD V PD � � Ell � � n By as�umption.+I(U) H. 1>4 0 = 1>3(X) = 1>3"2(Y) = f321>2 V3. By exactness at there i.k. Consider the following commutative diagram of vec tor spaces and linear maps.(U) H. and U n V. 1>4 3 = "3 (X) = 0.+I(U V)+ H. 1>3(X) = 0 f331>3 (x) = 0. for all k on U. 1>3 x= "2(Y) = "2 "dw)) = ( (Q o. aU vertical maps. Vve now forget all about our manifold and use a purely algebraic resull. LEMMA. except possibly the middle one. = = 1>dw) w E VI . Then is also an isomorphism. 1>4. n � is an isomorphism 2 1 . •:." y E 1>2(Y)V2 f3d:) x = E WI . V. and the bottom row is the dual of the Mayer Vietoris sequence for compact supports. and that are isomorphisms. Consider the following diagram. Hence since Thus (Y) · Hence with for some Z Moreover. 1>3 VI � V2 � V3 � V4 � V5 j 1>1 f3I 11>2 11>3 11>4 j 1>5 WI __� W2 � W3 � W4 � W5 PROOF.H. : for some Then which implies that y = "dw). 1>5 Supposc f some or Then (x) so is an isomorphism. x E V3.+I(Vn. then V PROOF. Thc proof that is onto is similar. are isomor phisms. So is oneonc. o. 1>2. If M = U U for open sets U and and is also an isomorphism for all k on M.440 C1zapter J J H'IW) H'I(V) H'IW V) H'(M) H'(U) H'(V) HkW V) j PD j PDEIlPD 1PD 1PDEIlPD j PD [H. It is not hard to check (Problem 8) that every square in this diagram commutes up to sign� so that by changing �ome of the vertical isomorphism� to their negatives: we obtain a commutative diagram.<VW H.<U v)· Ell � � Let I = 11 .
Even though the Poincare Duality Theorem holds for manifolds which are not of finite type. and consider a nice cover {V" . •:. Let V = V. 23. then a slight modification of the proof for Lemma 1 9 ./II) have different (infinite) dimensions. or PROOF B y induction on the number r o f open sets in a nice cover o f M .ExeulJion in Ilze Realm ifAlgebraic Topolog). We begin by considering a smooth kdimensional orientabie vector bundle S = 1f : E > M over M. then Hk (M) and Hnk (M) have the same dimension. then X(M) = O. If M is a compact orientable odddimensional manifold. . If { V" . since E is locally a product. . Ell v for the (n + k )manif old E. = ( _ I ) k+ ' dim Hnk (M) A more involved use of Poincare duality will eventually allow us to say much more about the Euler characteristic of any compact connected oriented man if old M"./II) and H� (IR2 . . COROLLARY. THEOREM (THE POINCARE DUALITY THEOREM).k ( M) ' is an isomorphism f all k . The theorem is clearly true for r = 1 . then the map PD : Hk (M) > H. This completes the induction step. . and for V n V (as in the proof of Proposition 19). COROLLARY. f M and v f S give an or or orientation /J. V} of M. 441 22. Vr . Corollary 23 does not. If M is a compact connected orientable nmanifold. 25. Orientations /J. COROLLARY. The theorem is true for V. If M is a connected oriented nmanif old of finite type. PROOF In the expression for X(M). noting that V' is isomorphic to V if V is finite dimensional . . Vr } is a nice cover of M by geodesically convex sets so small that each bundle S I Vi is trivial. Problem 7 shows that H' (IR2 . then H k (M) and H. V. In fact. . 24.k (M) have the same dimension. PROOF Use the Theorem and Proposition 19. it is true for M. U · · · U Vr . By the Lemma. •: . the terms (l)k dim Hk(M) and ( _ I )"k dim Hnk (M) cancel in pairs. If M is a connected oriented nmanifold of finite type. •:. Suppose it is true for a certain r.
26. This class U is called the Thorn class of $. . so E is a manifold of M 1f 'E we haw' 1f s O 1f so 1f' : H I (M) > H I (E) is an isomorphism for all I.vp) jp'w = l .). Also choose A so that there is an equivalence (Fp.) PROOF. the orientation v for $ determines an orientation vp for Fp. L 1f'� l\ w = 1. Notice also that for the maps s = Osection E. Let (M. . so that A i s smoothly contractible to any point p E A . so that J( M. (This condition means that f where U is the class of the closed form w. /L) be a compact connected oriented manifold. Pick some closed form w E c� (E) representing U. . Our definition of U states that ( I) Let A C M be an open set which i s diffeomorphic to IRn. Then the Thorn class U is the unique element of H�(E) with the property that f or all P E M we have jp 'U = vp.442 Chapter J J shows that {1f ' (U. 1f . and hence an element vp E H�(Fp). ) � = 1. On the other hand. and let jp : Fp > E be the inclusion map.' (Ur)} is a nice cover of finite type. Let Fp = 1f' (p) be the fibre of $ over any point p E M. The Poincare duality theorem shows that there is a unique class U E H� (E) such that 0S = identity of M is smoothly homotopic to identity of E. . THEOREM. Our first goal will be to find a simpler property to characterize U. there is an element jp' U E H� (Fp). and $ = 1f : E > M an oriented kplane bundle over M with orientation v. p. and let � E Cn(M) be a form representing /L. . Since jp is proper.
I) 'f. it is easy to see that there is a smooth homotopy H : (A x Fp) x [0. the form H ' w o n (A x Fp) x [0. 1] > A x Fp such that H(e. 1f2 (e)) = }i> (1f2 (e)).I (A) with A x Fp. under the identification of 1f . By choosing a smaller A if necessary. e. O) = e me. the map jp : Fp > 1f1 (A) corresponds to the map e r+ (p. Consequently. the support of wl1f1 (A) is contained in {(q. lleII < K ) . We will also use 1f2 : A x Fp > Fp to denote projection on the second factor. A glance a t the definition o f J (page 224) shows that the . support w if lie II ? K . Under this identifica y tion.EtC/miDlI in lize Realm ifAl gebraic ToJ)ology 443 This equivalence allows us to identif 1f1 (A) with A x Fp. 1) = (p. or which we will continue to denote by jp. we can assume that there is some K > 0 such that. wejust pull the fibres along the smooth homotopy which makes A contractible to e. Let II II be a norm on Fp. For the H constructed in this way it follows that m e. 1] has support contained i n {(q. I) : lIell < K}. support w Using the fact that A is smoothly contractible to p. e) f e E Fp.e) : q E A .
To prove this. " learly suffices to prove that the integral is 0 over A' x Fp for any closed ball A' C A. Combining (3).444 Chapter II lIell < K}. e) p rem 714 shows that (jp Jr2 ) 'W . we claim that the last integral in (3) is O. Thm (2 ) So support A c {(q. Since we have where Jr* /L 1\ A has compact support on A' x Fp by (2) = = ± 0 . e) : lIell < K J .io*(H*w) ° = = d(lH*w) + I(dH*w) d(lH*w). Now. Theo form IH*w on A x F has support contained in {(q. JaA1XF/. (5) we See tha. i.w = i.*(H*w) . (4). on the one hand we have (Problem 817) (4) On the other hand.  I ) dimen . [ Jr*/L I\ A bv Stokes' Theorem because the form Jr*/L I\ A is clearly 0 on aA' x Fp (since aA' is (n sional).
ExcUlJion in the Realm ifAlgebraic ToJ10Iogy
445
This shows that iFp is independent of p, for p E A . Using connectedness. it is easy to see that it is independent of p for all p E M, so we will denote il simply by iF Thus
ip*a!
i*a!.
11fI(A) ,,*� a!
1\
=
JA
[ ,,*� . [ i*a!.
Comparing with equation (I), and utilizing partitions of unity, we conclude thaI
which proves the first part of the theorem. Now suppose we have another class V' E H; (E). Since
it f ollows that V' = cV for some C E
1R.
Consequently.
Hence V' has the same property as V only if c =
1. .:.
The Thorn class V of S = : E > M can now be used to determine an element of Hk (M). Let s : M > E be any section; there always is one (namely, Ihe Osection) and anv two are clearly smoothly homotopic. We define the Euler class X(S ) E Hk (M) of s by
,,
X(s) = s V .
Notice that if S has a nonzero section s : M > E, and E C;(E) rep resents V, then a suitable multiple c . s of s takes M to the complement of suppOrt w. Hence, in this case
*
a!
X(s) = (c · s)*V = o.
The terminology "Euler class" is connected with the special case of the bundle
TM, whose sections are, of course, vector fields on M. If X is a vector field on M which has an isolaled 0 at some point p (that is, X (p) = 0, but X(q) '" 0 (or q '" p in a neighborhood of p), then, quite independently of our previous considerations, we can define an "index" of X at p. Consider first a vector
446
field X O n a n open set U C JR" with a n isolated zero a t 0 E U . We can define x x a function I : U {O} > S"I by I (p) = X(p)/IX(p) l. If i : S"I > U is i(p) = <p, mapping S"I into U, then the map I 0 i : S"I > S"I has a x certain degree; it is independent of FJ, for small e, since the maps i , i2 : snI I U corresponding to < I and <2 will be smoothly homotopic. This degree is called the index of X at O.

Chapter J J
7
,, ~ ~
index 0
index
0
dI/� )V
index
I
~ ~ 2
)�
index index 2
index 1
index
I
index
I
in
�n
index
(_ I )"
h.X is the vector field On V with
Now consider a diffeomorphism h : U
>
V C JR" with h(O)
in W
=
O. Recall thaI
Clearly 0 is also an isolated zero of h.%. 27. LEMMA. If h : U > V C JR" is a diff eomorphism with h(O) = 0, and X has an isolated 0 at 0, then the index of fl.X at 0 equals the index of X at O.
Excursion in the Realm ifAlgebraic Topology
447
PROOF.
by
Suppose first that h is orientation preserving. Define
H : IRn x [0, I J H (x , l ) 
{ h(IX)
>
Dh (O)(x)
O < I ::S I 1 = 0.
JR:"
This is a smooth homotopy; to prove that it is smooth at 0 we use Lemma 32 (compare Problem 332). Each map H, = x r+ H(x, I) is clearly a diffeomor phism, 0 ::s 1 ::s 1 . Note that H, E SO(n), since h is orientation preserving. There is also a smooth homotopy { H, } , I ::s 1 ::s 2 with each H, E SO(n) and H2 = identity, since SO(n) is connected. So (see Problem 825), the map h is smoothly homotopic to the identity, \�a maps which are diffeomorphisms. This shows that f . X is smoothly homotopic to f on a sufficiently small region of x h x IRn  {OJ. Hence the degree of f h.X 0 i is the same as the degree of f 0 i. To deal with nonorientation preserving h, it ob�ously suffices to check the theorem for h(x) = (x ' , . . . , xn' , _xn). In this case which shows that degree f h.X 0 i
As a consequence of Lemma 27, we can now define the index of a vector field on a manifold. If X is a vector field on a manif old M, with an isolated zero at p E M, we choose a coordinate system (x, U) with x(p) = 0, and define the index of X at p to be the index of x,X at O. 28. THEOREM. Let M be a compact connected manifold with an orien tation /J. , which is, by definition, also an orientation for the tangent bundle S = 1f : TM > M. Let X : M > TM be a vector field with only a finite number of zeros, and let a be the sum of the indices of X at these zeros. Then
fh.X = h 0 f 0 h' , x = degree f 0 i . •:. x
X( S) = a . /J. E H" (M).
PROOF.
Let Pl, . . . , p, be the zeros of X. Choose disjoint coordinate systems (U" x, ) , . . . , (U" x,) with x/ (P/) = 0, and let
B, = X, l ({p E IRn : Ipl
I f W E C; (E) i s a closed form representing the Thorn class U o f S, then we are trying to prove tha,
::s
I ll .
f(M,p.) X'(w) = a.
448
ChajJter J J
We can clearly suppose that X(q) " support a! for q " Ui B,. So
thus it suffices to prove that
L X*(a!) = t.l, X*(a!);
X*(a!)
lBi
[
=
index of X at Pi .
It will be convenient to drop the subscript i f rom now on. We can assume that TM is tri\�aJ over B, so that 1f1 (B) can be identified with B x Mp. Let ip and 1f2 have the same meaning as in the proof of 111e orem 26. Also choose a norm II II on Mp. We can assume that under the identification of 1f I ( B ) with B x Mp, the support of a! 11f' (B) is contained in {(q, v) : q E A, IIvll � I }. Recall f rom the proof of Theorem 26 that support J.. c {(q, v) : IIvll Since we can assume that X(q) " support J.. for q
( I)
l
X*(a!)
= =
= [ X*1f,*(j;a!) [
JB
l
X*1f,*(jp *a!)
l
E aB, we have
:::
I ).
X* (dJ..) X* (J..)
by Stokes' Theorem
On the manifold Mp we have
l

JaB
X*1f,* (i/a! )·
ip' a!
= dp
p an (11  1 )form on Mp (with noncompact support).
If D c Mp is the unil disc (with respect to the norm II III and aD c Mp , then
snI
denotes
(2
}SIII
[
p
d [ = laD p = 1 p D
= 1D jp*a! by 111eorem 26, the = 1. that support jp*a!andD. facI C
.
1,�'Cl/nioll ill the Realm ifAlgehm;( 7ojlOlol{1'
Now, for q and X: (3)
449
E B  {pI, we can define
X(q) = X(q)/IX(q)l,
aB
l
>
TM is smoothly homotopic to X: aB
X * 1f2 (J/w)
= =
=
l
[ [ [
>
TM. So
X * 1f2 dp X * 1f2 P X*1f2 P (1f2
0
JaB JaB JaB
by Stokes' 1l1eorem
=
X) *p.
From the definition of the index of a vector field, together with equation (2), it follows that (4)
JaB
[
(1f2
Equations (I), (3), (4) together imply (*) .
0
X*)p = index of X at p.
•:.
At the momcnt, we do not even know that there is a vector field on M with finitdy many zeros, nor do we know what this constant sum of the indices is (although our terminology certainly suggests a good guess). To resolve these questions: we consider once again a triangulation of M. \!I./e can then find a vector field X with just one zero in each ksimplex of the triangulation. We begin by drawing the integral curve. of X along the I simplexes, with a zero at cach Osimplex and at one point in each I simplex. We then extend this picture
29. COROLLARY. If X and Y are two vector fields with only finitely many zero. on a compact orientable manifold, then the sum of the indices of X equal. the sum of the indices of Y.
450
Chapter ]]
to include the integral curves of X on the 2simplexes, producing a zero at onf
point in each of them. We then continue similarly until the nsimplexes are filled.
30. THEOREM (POINCARE HOPF). The sum of the indices of this vector field (and hence of any vector field) on M is the Euler characteristic X(M). Thus, for S = 1f : TM > M we have X(s) = X(M) ' /J..
PROOF.
At each Osimplex of the triangulation, the vector field looks like
with index 1 . Now consider the vector field in a neighborhood of the place where it is zero on a I simplex. The vector field looks like a vector field on IRn = IR I X IRn 1 which points direc inwards on IR I x {O} and directly outwards on {O} x IR,, I . lly
(b) II = 3
E, xcuTsion in Ihe Realm ifAlgebraic 7opology
451
For 1 1 = 2, the index is clearly  I . To compute the index in general, we note that I takes the "north pole" N = (0, . . . , 0, 1 ) to itself and no other point x goes to N. By Theorem 812 we just have to compute signN I . Now at N we x can pick projection on jRn  1 X {O} as the coordinate system. Along the inverse l axis the vector field looks exactly like figure (a) above, where we image of the x already know the degree is  I , so I takes the subspace of sn I N consisting of x. tangent vectors to this curve into the same subspace, in an orientation reversing way. Along the inverse image of the x 2 _, . • • , xn I axes the vector field looks like
so ! takes the corresponding subspaces of snl N into themselves in an ori x* entation preserving way. Thus signN I =  I , which is theref x ore the index of the vector field. In general, near a zero within a ksimplex, X looks like a vector field on jRn = jRk x jRn k which points directly inwards on jRk x {O} and directly outwards on {O} x jRnk The same argument shows that the index is ( _ I ) k Consequently, the sum of the indices is
ao  a t + C¥2

. . •
=
X(M) . •:.
We end this chapter with one more observation, which we will need in the last chapter of Volume V ! Let � = 1f : E > M be a smooth oriented k plane bundle over a compact connected oriented l1manifold M, and let ( , ) be a Riemannian metric f �. Then we can form the "associated disc bundle" and or "associated sphere bundle"
D = {e :
(e, e)
S = {e : ( e , e) = I ) .
5
I)
I t i s easy t o see that D is a compact oriented (11 + k)manifold, with a D = S; moreover, the D constructed f any other Riemannian metric is diff or eomorphic to this one. We let 1fo : S > M be 1f I S.
452
3 1 . THEOREM. A class multiple of X($ ).
Chapter II
PROOF.
Consider the following picture. The top row is the exact sequence
H! (D  SI
�':'ir: D�
' Hk (M)
since HO
a E Hk (M) satisfies Ho'(a) = 0 if and only if a is a
H'(�
for (D, S) given by Theorem 13. The map s: M > D  S is the Osection, \vhile s: M 7 D is tlle same Osection. Note that everything commutes.
HO' = i' 0 (HI D)'
and that
since (H I D) 0 S is smoothly homotopic to the identity. Now let a E H ( M) satisfy Ho'(a) = O. Then i'(HI D)'a = 0, so (H I D)'a E image e. Since D  S is diffeomorphic to E, and every element of H; (D  S) is a multiple of the Thorn class U of � , we conclude that
k
s* = s* 0 e
.,' 0
since extending a form to D does not affect its value on s(M),
= (H I D) 0 i,
(HI D)' = identity of Hk (M),
(H I D)'a = c · e(U)
for some C
ER
Hence
a = s'(HI D)'a = c · s' (e(U)) = c · s'U = c · X(�).
The proof of the converse is similar. •:.
i'.'xclIniou iu flu' Reahn ofAlgebraic 7opolog)
PROBLEMS
L Find Hk(S' x . , . dim Hk = G; )'] X
453
S ' ) by induction on the number 11 offactors. [Answer:
2. (a) Use the MayerVietoris sequence to determine Hk (M  {pI) in terms of Hk (M), for a connected manifold M. (b) If M and N are two connected l1manifolds, let M # N be obtained by joining M and N as shown below. Find the cohomology of M # N in terms of that of M and N.
(c) Find X f the l1holed torus. [Answer: 2  211.] or 3. (a) Find Hk(Mobius strip). (b) Find Hk(JP'2 ) . (c) Find Hk (p"'). (Use Problem 115 (b); it is necessary to consider whether a neighbol'ilOod of 1P'''' in 1P''' is orientable or not.) [Answer: dim Hk (IP'n) = 1 if k even and ::: 11, = 0 otherwise.] (d) Find Hk (Klein bottle). (e) Find the cohomology of M # (Mobius strip) and M # (Klein bottle) if M is the nholed torus. indicated identifications of edges we do 1101 obtain a triangulation of the torus. Why not>
4. (a) The figure below is a triangulation of a rectangle. If we perform the
A
A
454
Chapter II
(b) The figure below does give a triangulation of the torus when sides are iden tified. Find "0, " " " for this triangulation; compare with Theorem 5 and 2 Problem I .
5. (a) For any triangulation o f a compact 2manifold M, show that
3"2 = 2" , '" = 3("0  X(M))  1) "0("0 2: C¥) 2
1
"0 '" 2 (7 + V49  24X ( M ) ) .
=
(b) Show that for triangulations of S2 and the torus r2
S'
x
S'
we have
'" 14. 2 Find triangulations for which these inequalities are all equalities.
r2 :
S2 :
0'0 2: 4 "0 :0: 7
Q' l :::: 6
'" '" 21
"
0' 2
2:
4
7. (a) The vector space H' (1R2  /II) may be described as the set of all se quences of real numbers. Using the exact sequence of the pair (1R2 , /II) , show that H� (1R2  /II) may be considered as the set of all real sequences {a,,} such that an = 0 for all but finitely many II. (b) Describe the map PD: H ' ( 1R2  /II) .... H) (1R2  /11) * in terms of these descriptions of H' (1R2  /II) and H� (1R2  /II) , and show that it is an isomorphism. (c) Clearly H) (1R2  /II) has a countable basis. Show that H' (1R2  /II) does no\. Hint: If Vi = {aij} E H' (1R2  /II) , choose (b" b2) E 1R2 linearly indepen dent of (a , ' , a , 2): then choose (b" b4,b,) E IR' linearly independent of both (a ,' , a ,4, a ,') and (a ' , a 4, a '); etc. 2 2 2
6. (a) Find H; (S" x IR"') by induction on 11, using the MayerVietoris sequence for compact suPPOrts. (b) Use the exact sequence of the pair (s n x IRm , {pI x IRm ) to compute the same vector spaces. (c) Compute H k (S" x sm'), using Theorem 13.
"2.) 9. is union ofshaded & . "3 . (It will be necessary to recall how various maps are defined. VI. . V u. "3. the only slightly difficult maps are the ones involved in the above diagram.E. this "direct sum" consisting of all sequences (" I . this "direct product" consisting of ali sequences ("1 . then it holds for M. 455 8. Use an analogous decomposition in n dimen sions to prove that Poincare duality holds for any triangulated manifold. Show that H. " ) with "i E H. (M) "" EEli H. Show that the squares in the diagram in the proof of Lemma 21 commute.l l . (d) The figure below shows a decomposition of a triangulated 2manifold into three open sets Vo. ) with "i E Hk (Mi). and U2 . (a) Let M = MI U M2 U M3 U · · be a disjoint union of oriented IImanif olds. ( Mi). xcwsioll ill lize Realm ifA lgebraic Tapolog).I (u n V) H�+I (U n V)' lPD > Hk (M) > H� (M)' lPD which commutes up to the sign ( . (c) Show that if the Poincare duality theorem holds for each Mi. "2 . . (b) Show that Hk (M) "" n i Hk(Mi). (Mi). (Mi) and all but finitely many "i = 0 E H. . which is a good exercise. except for the square Hk. Vo is u nion of shaded VI is union of un shaded C�·. .
show that this is 0 for k odd. : E > M and r = . . and (j. (Begin with a triangulation to obtain a vector field with a discrete set of zeros. such that all Pi E interior D. : E > M be an oriented kplane bundle over an oriented manifold M. (Using the notation of Problem 323. .here zero vector field on M. and push everything off to infinity. we have f*(x(�)) = 12. enclose this ray in a cone. Using this. (d) If M is connected and not compact. X(j*(�))·) (e) If M is a connected manifoldwithboundary. over a compact oriented manifold M. thaI X(�) = 0 when � = . prove the Thorn Isomorphism Theorem: The map HI (E) > H:+k (E) given by " r+ " V U is an isomorphism for all I. (a) Let P I .456 Chapter JJ 10. Using Problem 826. If a vector field X has an isolated singulaI"ity at p E Mn . (E') are the Thorn classes. : TM > M fOl' M of odd dimension.X at p is ( _ I )n times the index of X at p. in particular. .. f) a bundle map from �' to I: which is an isomorphism on each fibre. This p rovides another proof that X(M) = 0 for odd 11. Let � = . (b) . ' : E' > M be oriented kplane bundles. 1 3. p. show that there is a subsel D C M difTeomorphic to the closed ball. show that if X(M) = 0. then j* ( U) = U' . It follows. (a) If U E H.k (E).I > snI of degree 0 is smoothly ho motopic to a constant map. with Thorn class U. Conclude that U represents 0 E Hk (E).. . we can form U v U E H. then there is a . then there is a nowhere 0 vector field on M. aM '" 0.. Q») Since we can also consider U as being in Hk (E).f*(X(n) = xW). show that the index of . (E) and U' E H. then there is a nowhere 0 vector field on M. Join these by a ray going to infinity. (c) it is a fact that a Coo map f : sn. then there is a vector field X on M with only one singu larity.. providing anotllCl proof that X(M) = 0 in this case.) 1 I . so that X(�) = O. (a) Let � = . Using anticommutativity of /\. with nov\. E M. Using Poincare duality. 0» If M is compact.
(d) Conclude that Rh is an isomorphism if M is of finite type. and let i : S'k(M) > Sd M) be the inclusion.) It follows that there is an induced map Rh from the de Rham cohomology of M to the singular cohomology of M. This Problem proves de Rham's Theorem. while i 0 r is chain homotopic to the identity of Sd M) [basically. Basic knowledge of singular cohomology is required. 1 w. .) (e) Check that the cup product defined using 1\ corresponds to the cup product defined in singular cohomology. IR) be Rh(w)(c) = Show that Rh i s a chain map from { Ck (M)} t o {Hom(S'k ( M). (Hillt: Stokes' Theorem. (a) If w is a closed kform on M. IR)}. For a manifold M. r is approximation by a Coo chain]. let Rh(w) E Hom(S'k ( M). V. (b) Show that Rh is an isomorphism on a smoothly contractible manifold (Lem mas 17.EXCU1JiOl1 in tlze Realm ifA( �ebmic TopologJ' 457 1 4. we let S'k( M) denote the Coo singular kchains. We will denote the group of singular kchains of X by SdX). It is not hard to show that there is a chain map r : SdM) > S'k(M) so that r 0 i = identity of S'k (M). it follows that Rll is an isomorphism f any triangulated or manifold. This means that we obtain the correct singular cohomology of M if we consider the complex Hom(S'k (M). then it is an isomorphism for U U V. using the MayerVietoris sequence for singular cohomology. and U n V. and 19 will not be necessary for this. to show that if Rh is an isomorphism for U. 1R). (Using the method of Problem 9. 1 8.) (c) Imitate the proof of Theorem 2 1.
.
We bave just seen that the Hausdorff property is not a "local property". there are manifolds which are not normal (Pt·oblem 6). that almost all "nice': properties of a � anifold are equivalent. but local compactness is. we first note . are important in certain cases. :Moreover. . Before exhibiting nonmetrizable manifolds. we will now define a manifold to be a topological space M such that (I) M is Hausdorff. these need not be regular even if they are Hausdorff) On the other hand. (b) Each component of M is second countable (has a countable base for the topology). so every component is open. this argument does not work f "infinite dimensional" manifolds. which are lo or cally like Banach spaces.iust like neighborhoods of o. (c) M is metrizable. so every manifold is locally compact. a compact manifold is metrizable. a Haus dorfflocally compact space is regular. (2) For each x E M there is a neighborhood such that U is homeomorphic to lR:". then (U n JR) U {OJ is a neighborhood of 0 (in JR). Although nonHausdorff manifold. The following properties are equivalent for any manifold M: (a) Each component of M is a compact.APPENDIX A CHAPTER I Following the suggestions in this chapter. we will not consider them.) 459 U n JR is open. This space may also be obtained by identif ying all points except 0 in one copy of JR with the corresponding point in another copy of JR. I t consists of JR U {*} where * 'f. so every manifold is regular. (In particular. THEOREM. (I) U of x and an integer n Condition (I) is necessary. with the following topology: A set U is open if and only if (2) If :>: 0 Thus the neighborhoods of * look . Ev ery manifold is also clearly locally connected. JR. (d) M is paracompact. and thus a manifold itself. * E U. for there is even a I dimensional "manifold" which is not Hausdorff. (By the way.
s countable union of C0111pact set�. It lurWi out thaI there are even l manifolds "dlich are not paracompact. = Uo U . . Similarly Do U VI U . U C. is compact. we obta1n open sets Vi with Vi compact and Vi C Vi+ l . Since the space 1S connected. . is clearly open. (Ord1nal numbers will nol be needed for a 2d1111ensional ex ample 10 come later. Let M = C.460 Appendix A FIRST PROOF. This pro\'es the Lemma and the Theorem. where each C./. paracompact space is a compact. pg. U . U . ' . M. U [in . U · .. U V". . Then VI U C2 has an open nl'i�hborhood U:!.) oflhe others. locally compact. There is a locall\' finite COver of the space by open sets with compact clo sure. \". The cOllSl ruct10n of J11Cse eXalnples requires the ordinal numbers. U . . U . . 160). The unioll Va U " . then x must be in the closure of a finite union of these Vi. \l\'1th compact closure_ Continu1ng in th1s way. . . is (a) is Theorem 12. (cJ =} (d) because anv metric space is paracompact (Kelley. LEMMA. Gellem/ 7ojJo/agT. . If Vo is one ofthe�e.. . an open neighborhood VI "·. (a I =} (b) follows immediateh' fi·om the simple proposition that a acompact locally second countable space is second countable. . Un':!. . ORDINAL NUMBERS Recall thaI an ordering < on a set A is a ]e1ation such thai (I) a < b and b < C implies a < c f all a . it equals thi. . (b) =} (c) follows (rom the Urvsohn metrization theorem. U Un . A connected. c E A (tl"ansitivit\·. or . U V'I I imersects only Un l+1 • . which are briefly explained here. and hence =} SECOND PROOF (0) (a) =} (b) =} (c) and (d) =} (a) as before.:. b . then Vo can intersect onlya finite number lJ1 • . It is also dosed. PlOOf. Clearly C. ha. (a) =} (d). The serond proof does not rei v on this difficult theorem. .ith compact closure. · .'hos('" union contains all Cj. ' U U". II is easy to show iiom th1s that M is paracompact. : and so On. for if x is in the closure. . (d) =} (a) is a consequence of the following. because x has a neighborhood \vhich intersects only finitely man}: Thus x is 1n the union. l..
. O < 1 < 'J < · · .· <w<w+l <w+2< . . < w < . . 2. for the prese]]!. An orderinf! < On A lS a wellordering jf every nonelnpty subset B C A has a. 3 < . .· < · '] < . an element b such that b ::5 b' for all b' E B.··<w·2 <w·2+ 1 < . O < l < '2 < · · · < w 0 < 1 <2 < .. o< 1 < ' " < w 1. I ) ) CO.I' 0. 2)\ 0 < 1 <2<3< . .Appendi. just a set distinct from those already mentionedl 0 < 1 < 2 < . 0<1 <2< 0 < 1 < 2 < · · · < w < w + l <w+2 < ·· · < w · 2 < w · 2+ 1 < ··· < w O< 1 <2 < ··· < w < . < w .. one and onlv One of the following holds: (trichotomy).. . Two ordered sets (A. b 461 E A. that is. < w .firs/ clement.· <w·2 « ··<w·3 < . = (O.J (2) For all a. 1 . ) (w + I is. <) and (B. . <) where < is an ordering on A .· · < w <w+1 (w is some set . <) are order isomorphic jf there is a oneOne onto function / : A > B such that a < b implies lea) < feb): the map f itself is called an order isomorphism. (i) a = h (ii) a < h (iii) b < a (also written a h' o {OJ 0<1 0 < 1 < :' 0<1 <2<3 (A (A = etc. < w < w + l < w + 2 < · · · < w · ] < w <w+1 <w+2 < . 1 . . . Some wellordered sets are illustrated below: in this scheme we do not list any of the < relations which are consequences of the ones ab'eady 1isted. 'J < . and II is easily seen to be an order isomorphism also. < .. . . > An ordered set is just a pair (A .. .. 3 .
while the first does not. the "wth H .c.. <) and (B.hen' PROOF.rith the "w th". ...462 AfJPclldix A Any �ub�et of a wellordered set is. We match the first element of A with the first of B. Notice that each set on OUf list 1S <1 .. It is easy to see that if B is " seglncm of A . TI. ('ven for the "wth" eienlellt (the first one after the first: second: third. a subset B of a wellordered set A is called al l (initial) segment if b E B and a < b imply a E B.c?ment of the succeeding ones. If B '" A is a segment of A. If (A. PROOF. .. until we run out of one . . the second with the second. < ) are wellordered sets. <w and 0 < I < . To do Ihls rigorously. PROPOSITION. .en feb') must be b'. In particular. of course. also a wenordered set with tbt same ordering.)! The way we prove this ligorouslv is amazingly simple: If feb) '" b for some b E B. then B is not order iSOlnol· phl( to A. the only (wder isomorphism f)'om B to a segmeni of A is 1111 ldentit:: > f(h) = b. . on the smaller of their two domains (just cOllsider the smaJJesl element \. If f : B B' C A is all order isomorphism and B' is a segment of A . <w <w+ I arc nOl order isomorphic because the second has both a last and a next to la�l eien1cnt. etc. PROPOSI TION. It 1S not hard to sec that no two sets on our lis1 are order isomorphic.. For example. etc.B. B = {a' E A : a' < a J : o < I < . just consider the first element of {b E B : feb) '" b): an outright contradiction appears almost immediately. And so Oll . •:. then for the firs! clement b of B (and hence of A) we clearly mus! haw 2 . a i s the first element of A .. then one i. order isomorphic to a segment of the other. In fact. It is easy to show that any tWo such order isomorphisms agrC'(. which makes the study of wellordered set' simply delightful.. But there lS a much more general propOSltlOll which will settle all cases at once: J..srI. where b' is the second element.. then either B = A or else there 1S some a E A such that In fact. consider order isomorphisms from segments of A onto segments of B. Proposition I has a com pall ion.
If it is defined on all of A we are done.A.+ I the smallest ordinal after " (if " is represented by the wellordered set ( A .rl from each ordrr l:. <). <)". it would be much !llcer to choose one specific well o)'dered :. <) < (B.. <) = (B. < is a wellordering of the ordinal numbers. Suppose wc define a relation < between wellordered sets by stipulating thaI (A . <r! must be replaced by "(A . Notice that some ordinals are not of the form " + I for any ".'enordered sets. let ( A . < ) ! Roughly speaking: An ordinal number is order isomorphic to the set of an ordinals less than 1t.ularly elegant way to do this. This fJ is the smallest element of . due to von Neu mann . Given a nonemptv set .gnore elements 2:. c¥. These older lsomorphism classes arc called ordinal numbers. There is a panic. which is c1earlv the largest of all. <) be a well ordered set representing one of its elements ex. then " + I is represented by a \". instead of '''lith the wellordered sets themselves.A of ordinal numbers. <) order isomorphic to (B. It determines a segmen t wh1ch represents somt' fJ E . I f il ' is noL then its range m� st be all of B (or we could easily extend it) and we arc stl1l done.. To obviate this difficulty V'!e' need only '''lark with order isomorphism classes of . and dcfine thesr speclfic sets to be thr PROOF. E\'ery element < ex is represented by an ordered set which 1S order iS01TI01phic to some pl"Oper segment of A : earh o f these is the segment consisting o f elements o f A less that some a E A . we will denote by . . <) when (A . . then the wellordered set of all ordinals fJ < " has a particularly simpk representation: it is order isomorphic to the set (A .A. <) is order isomorphic to a proper segment 01 (B. v·:hkh can be f ound in (he Appendix to Kelley. these are called limit ordinals. •:. Each ordinal number ls a horribly large set.Appendix A 463 thev don't). <). They are beautiful:' 3..A \'\'C can obviously i.. General Topolog). larger than all member!' of A). ''Almost'': because the condition H(A.oll1orphism class.:. To produce a smallest element of . <). represented by a wellordered set (A . Transitivity of < is obvious.'ellordered set with just one more element. So all such order isomorphisms can be put together to give another.. Consider the least of these a's. PROPOSITION. while those of the form " + I are called Successor * Only one feature mars the beauty of the ordinal numbers as presented here. If " is an ordinal numbe. ordi nal !lumbers. I\otice that if ex is an o)dinal number. and Propositions I and 2 show that we al most have trichotomy.
Then SI is uncountable. 4. this ". w' . 1 . and ell counter �rts with an ullcountable nmnber of degree� of complexity. PROPOSITION."ts llWlll loned so fal" are countable. one can see how the symbol> w3. to a proper segment o f itsdL contradicting* Prop osition J . . . il would represent a countable ordinal " E SI. any member of SI is countable. If it were countable. For . fO I· 011e would have to go uncountably far.. ( SI . By the remark after Proposition 3. OUl· list of wellordered sets onlv begins to suggest the complexity which well ordered Sets can achieve. a countable wellordered set). . . w . We will also denote some ordinals by the symbols appearing before: 0.tI. i. A leap of faith is required. .wuld mean that n is order i�omorphi(' to the C'oJknion of ordinals < Q .I resohllioll of this paradox. etc.464 Appendix A ordinals. After one comes to and this 1S only the beginnln�� A ll the wellordered Sf. With a little thought. (h is hopeless to tl'Y 10 "reach" n by C'o 'ltinuing the llsting of well ordered �ets begun above. There are indeed an enormous nUlnber of countable v·. 2. We have thus established the existence of an uncountable ordinal. 3 + w · 4 + 6 would be used sOlnewhere between w3 and w4 ) : after an these one would need and after all these the symbol <0 pops up. . . Let SI bc the collection of aJ! countable ordinals (ordinal. •:. see Kelley's Appendix. . is clearly the first uncountable ordinal. w + 1. so it represents an ordinal ex E (htl. Our sv example. . . and henct' is order isomorphic to an initial s<'glllt'nt of (). re presented by n. 3 .) Although the countable ordinals exhibit ul1countably many degrees of COIll pkxit�: they are each simple in one way: rihe PROOF By Proposition 3.e. <) is a wellordered set. and consequentlv has onlv countably many pre decessors. would appear (svmbols like w3 + w2 .reDordered sets: represented by * By ddetinll the words cOlllllable ami uncolilltable in this proofolle obtains the "Burali Forti Parmlox'·: the set Ord of all ordinal nurubers is wellordered.
thus also by a subset of (00.AJ)pelldix A 465 5. 6. aside f)'om the line and the circ1e� there are 110 other connected Imanifolds. then there would be uncoul1\ablv mallY disjoint intervals in � . 1) with the order topology (a subbase consists of sets of the form {s : X < xo l and (x : x > . . I ) if a < f3 or if a = f3 and s < I . The union of all these sets would then represent ex. The Corollarv to Proposition 5 implies easilv' that the long ray and the long line are I dimensional manifolds. COROLLARY. .0) .0) )' (2. for then f3 would be represented by a subset of �. (1. To disllnguish L + and L.. then there is a sequence f3 t < f3. there is a sequence f3 . a contradiction. The first example of a nonmtU'izable manifold is defined in terms of n. Then f3. This is impossible . 11S1 \vhich comes after fill ' (. namely those bel\'I'eCll the points representln. PROOF. < f32 < f33 < .O) (w+2. with the order < defined a s follows: (a. •:. ). i). Suppose there were one. < a cofinal in a . ). If a E n is a limit ordinal. such that every f3 < a satisfies f3 < f3n for some 11 (we sa�' that (f3nl is "cofinal" in a). then a is represented bv some wellordered subset of �. and we can easily arrange that the subset representing f3i is a segment of the subset representing f3j for i < j. and hence a smallest.O) (W+I. no subset of � is order isomorphic to n .0) . . So by Proposition 5. ) tw·2. . s ) < (f3 . is represented by a subset of ( 00. 1 ) . If a E n. This can be pictured as follows: (0.\'0 ) ) is called the closed long ray (with "origin" (0. 0)). O)} is the (open) long ray. 1) . < a. If a subset of � were order isomorphic to n. . 0) and a could bv represented by a subset of � . . .o)  The set n x [0. Howevel.Q and a + 1 for all a E n. The disjoint union of two copies of the closed long ray with their origins identified is the long line L.0) (w. PROOF. and L + = n x [0.g. a E n not represented bv some subset of � . the names "halflong linel ' alld "long line" may also be used. Since a is countable. all its members can be listed (in notnecessarilv Let f3 t = Yt and let f3n + t be the first Y in the increasing order) y" Y2. YJ .{(O. It cannot happen that a = f3 + 1. Consider n x [0. PROPOSITION. < f3) < .
We begin with the open upper halfplane �� = {(x. This is a Hausdorff manifold. with P E (��) and f (p) E JR � o + identified. 0) by (x. The manifold itself is. .466 Appendix A Quite a few new 2manifolds can now be constructed: L+ X S ' L + x IR LxL L+ x L+ (halflong cvlinder).:. Define a map /0 : (��) > �� b." � /_. the following diagram shows two open sets homeomorphic to �2. y). y)." '" + q�. we will denote this set by �6. we could have thro\vn away �� to begin with slnce it is ldcntified by a homeomorphism with (��) . + to JR2 . + But consider now.\' .' f .) > �� b�' + j� «X. homeomorphic �F='. 8 ) in the product of the closed long ray and S' produces another 2manifold. Identifying all points « 0. There is another way of producing a nonmcuizable 2manifold which doe� not use n at all. L x L+ (big halfplane).)')a) = (a + y. another copy of �2. in fact. (halflong strip. (big quadrant). for each a E �. and denote the point (x. Y ) o ) = o« (xy. Deline j� : (�.\".. L x S ' (long cylinder).:. y) E �2 : r > OJ and another copy �2 x {OJ of the plane. which might be called the "big disc". Consider the disjoint union of �� and ��. which we will denote by �.. y)o. (big plancj. y. 0). say �2 x {a}. L x IR (long strip).
O)a ] . and Problem I. then any a equals f3 + II for a unique limit ordinal f3 and integer II � O. show that if X and Y are disjoint equivalent infinite sets.y')b for which y = y' > 0 and .l'+b. (a) L + and L are not metrizable. c(A ) is defined for each set A '" 0. 3. show that one of them is equivalent to (can be put in oneone correspondence with) a subset of the other.ApjJendix A 467 In the disjoint union of �� and all ��. and in the disjoint union of all �� identify each (x. one is an extension of the oth. (Zorn's Lemma may be deduced from this fact fairly easily.. This space is still a Hausdorff manifold. are a space homeomorphic to ��� so we will consider �� a subset of the resulting space. (a + 2) + 1 by a + 3. the Priifer manifold.) (c) Given two sets. etc. developed in the problems. a wellordering < on a subset Y of X will be called "distinguished" if for all Y E Y. Given a set X.\3 :s n polnt. and c(A ) E A for all A. . We mav dispense with �� completely. (b) ]( we denote (a + 1 ) + 1 by a + 2. is a sequence in L +. for it has an uncountable discrete subset. (e) From (d). (d) Show that on any infinite set th"'e is a wellordering which represents a limit ordinal. The equivalence c1asses� of course. then X U Y 1S equivalent to 1'..) L X2 > 2. PROBLEMS (a) A wenordered set cannot contain a decreasing infinite sequence X l > X3 > . any sequence has a convergent subsequence (but L + i� not compact!). Consequently. (Thus one can define even and odd ordinals.. y = e(Y . i. but it cannot be second countable.\. then {x } converges to some (b) If Xl � X2 :s: . (a) Show that of any two distingulshed wellorderings. Let c be a "choice function". a E IR we wish to identify each p E (��) + with f a(P) E ��.y+a = x'. This manifold.{y' E Y : / < y}).: (b) Show that there is a wellordering on X. namely the set { (O. and related man]folds� have some very strange properties.e.Y)a and (X'.
(d) The StoneCech compactif. L + x L. Show that U include. (Using Problem 3(g). one can explicitly construct these Slone.r. L + x L" . I] L + with H (x.s]) and . (d) L (e) More generally. Hint: Let A n = {a : the wedge centered at a has width � I /n } . L+ x S l and L x S' are nOl homeomorphic. 6. (c) Let U be an open set in �� \vhich 1S the union of "wedges" centered a t (a. (Use (c)). Hint: Given H : L + x [0. (gJ If f : L + > � is continuous. L x L +. and the big disc arc all distinct.. L x �. + (and also L) are normal. (e) Of the 2manifolds constructed from L + or L with HI = 0 and one para compact end. (f) If f : L + > � is continuous. defin. X Sl ) = z. L + x L +. 00)) is countablt. a whole rectangle of the form J (a.cations of L x L. 4. (a) L + and L are not homeomorphic.t ech compactificatlon�. " paracompact ends'·. Similarly for L + x �.O) for every irrational a. (a) Show that the Proler manifold P is Hausdorff (b) P does not have a countable dense subset.I)} L +. and r > s. (b) HI(L +) Ht {L) O . Sillee IR: = Q U Un A . show that for every I we have ( H (x. some A n 1S not nowhere denst .' ([r. &). any order topology is normal (completely different proof!. 0) x for all x. n . L x L. then one of the sets f' « oo. only L + x � has the homotopy type of L +. then f is eventually constant. then n n both sequences converge to the same point. b) x (0. (a) L + is not contractible.468 Appendix A (c) If {X } and {Y } are sequences in L + with Xn S Yn :5 Xn +! for all 11. (b) L + x III and L x � are not homeomorphic. Hint: Imitating Problem 119. (c) HI(L + x S') = = Ht {L = = > = 5.
then the closure A ha. but which has a countable dense subset. s ) = Show that H is welldefined and that H(p. but that thev are not contained in dis open sets. (f) P . { (O. (e) Define H : P x [0.Appendix A (d) Let C" C2 C P be C. and such thai f lO) = { p ) (Consider functions defined on inidal segments of n vvith these same propertle�! and apply Zorn's Lemma. = 469 C2 = { (0. O)a a rational ) . Let M be a connected Hausdorff manifold which is not a point.s + sy a ) if y if y > 0 : � o. then C has an open neighborhood with cardinality (. Alternatively. whose boundary is a disjoint union of uncountably many copies of JR . 1( x 1 s + sy �' y  (x �. ( g) The dis joint union o f two copies o f P' . one can require f(rx) to be the result of applying the choice function to the set of all open neighborhoods of the c10sun f(rx) is an open neighborhood of the closure of Up <a f( f3 ) . There is a function f: n > (set of subsets of M ) such that f(rx) has cardinality c for all rx E n. 1 ] > P by H « x .{ (x . is a manifold which is not metrizable. Describe a homeomorphism of this manifold onto �2 8. (c) Let p E M. and C2 are closed. 7. 1) E �i U {(O. (a) If A C M has cardinality c (the cardinality of �). Conclude that P is contractible. � )a J'\ I 1 + sy l . It is known that everY second countable contractible 2manifold is S 2 or � 2 Hence the result of ro �structing the Priifer manifold using only copies �� for ratlonal a must be hOlneomorphic to � 2 . Its fundamental group is uncountable. cardinality r. O)a a irrational) : : joint Show that C. (b) If C e M is closed and has cardinality c.O)aJ for all p E P. with con·esponding points o n tht boundary identified. . Y)a. y)a y < 0) is a manifoldwithboundary P'.
1 ] ) with the properties of tbe func tion in part (c) is eventually constant. the long line.) (d) A function f : n > (set of subsets of [0. or the halflong line. Then there is a unique f with the required properties. (b) Every I manifold M contains a maximal open submanifold N whose topol ogy is tbe order topology for some ordelC (c) If M is connected and N '" M. consider an arc from p to p'. is homeomorphic to either the real line. (e) M has cardinality r. This is an example of defining a function by "transfinite induction" . of UP<a f(f3) with cardinality C. (a) A connected I manifold whose topology is the order topology for some . (Given p' E M.470 Appendix A order.) 9. then M is homeomorphic to S' .
are all CW. then it yields a CW structure for L + . . Continuing i n this wa)� we obtain Xo < Xl < X2 < . in other words. f(/(/(q ))) > f(/(q)). If (!) is a CW structure for L +. we must have f(q) > q.L +p. and then it is easy to see P I that we must also have f(/(q)) > f(q). has a limit point Xo E L + . f(/(q)). with f(xn } = Xn .Appendix A 471 CHAPTER 2 The long ray L + can be given a Coo structure. etc.L+p is clearly homeomor phic to L +. and even a CW structure. This sequence has a limit in L + . we obtain Xl > Xo with f(x} ) = X l . In fact. .L +p. So for some ql > Xo we have f(ql ) '" ql. the problem is that a complex analytic structure on �2 mav bc conformally equivalent to the disc. An easy appllcation of Zorn's Lemma then shows that Coo and CW structures eXlst on L+' I do not know whether all Coo structures on L + are difleomorphic.L +p. A CW stmcture exists on the Prufer manifold. since it is CW). then L + . Reasoning as before. and I(xo) = "0 . I do not know whether every 2manifold has a Coo struCtUlT. . . but this implies that f(x) = x for all x. If p E L +. and hence extendable. a contradiction. This implies that it is also diffeomorphic to (0.e can get a CW structure on L + x L +. f(q). and consequently that the structure on M can be extended if M is a proper subset of L +. To see this we need the result of Problem 924any Coo [or CW] structuff on a manifold M homeomorphic to � is diffeomorphic to � with the usual structure. points in ��. we can assume f(q J ) > q l . . since we can consider fI in the cOlltrary case. These are all distinct. this follows immediately from the fact that the maps f used for identifying points in various (��) + with a. Now f cannot be the identity on all points > Xo (for then it would be the identity ever)'\'Vhere. Using the CW structure on L + . \". there is no CW map q > p with a CW inverse. How ever. and L+p denotes all points S p. 1 ). The increasing sequence q. but it . and hence for L +. the method used fOI" obtaining a CW structUl'e on L + will not yield a complex ana!wic Jlruc/ure on L + x L +. It is knOvvll that there are uncountably many inequivalent CW structures on L + .
Ann. Sic. I 25115 (1958). PROBLEMS L+ . AnaiJ'lische Slrucklur und A b<. Soc.472 Appendix A may also be equivalent to the complex plane. On the other hand. locally connected. Of the Yalious manifolds mentioned in the previous section. it is a classical theorem of Rado that every Riemannian surface (2manifold with a complex analytic structure) is second countable. Math. a mod ification of the Priifer manifold yields a nonmetrizable manifold of complex dimension 2.L+q . . pp. Fennicae Se ries A. Show that d is a metric on X which gives the same topology for X q > p there is no nonconstant CW map f : L + . In fact. Let (Y. H. (b) Let d(x. 18. where X is H ausdorfl connected. Prove that lor I I .iihlbarkeil. pp. Complex AnaD'lic Manifolds wilhoul Counlable Base. Kneser. and not extendable. E X are contained In a compact connected C C X. 335340. (a) Every twO points X. 10. References to these matters are to be found 111 Calabi and Rosenlicht.hich can be immersed in which. try to deter mine v·. y) be the greatest lower bound of the diameters of f(C) (in the pmetric) lor all compact connected C containing x and y. Arner. Acad.L +p > 1 2. p) be a metric space and let f : X > Y be a continuous locall" oneone map. Proc. 4 (1953). ). and locall" compact.
Appendix A 473 CHAPTER 6 Problem A6(g) describes a nonparacompact 2manifold in which two open halfplanes are a dense set. y. On thi. the set of points (x. y. This example is due to M. y . Z )a (x . Since we can choose a = c/2. and the sets z = constant form a foliation of M by a 2dimensional manifold N.a). and for each a E � let �� be a copy of �J. Archiv. y. Kneser. z .· chell A bbildung "wischen iiberiib"ahlbaren Mannigfaltigkeilen. a E IR we identify (X. y. z + a ) (a + yx. y. II (1960). c . c) E A with y > 0 is identified with the set of points (x. O) : y < O} C A . y. Math.ala E �� with Y > O. o) a . The remarkable fact about this 2dimensional manifold N is that it is connected. y .2a) E A with y < O. For.z) E �3 : y '" OJ. we see that all leaves of the foliation are the same as the ieaf containing { (x. points in �� being denoted by (x. and these are identified with the set of points (x. y. . Let A = {(x.. y. manifold there is an obvious function ". The equivalence classes form a 3dimensional H ausdorff manifold M. Beispiel eiller dimensio'nserhohenden anab'li. We will now describe a 3dimensional version with a twist. y . c . 280281.al a with y < O. pp.ala) contains the points (x. Now the folium containing { (x. z) a for y > O for y < 0 with with (a + yx. c .z". y. c . In the di�oint union of A and all ��.
even though the Prufer manifold is contractible. Although the results in the Addendum to Chapter 9 can be extended to closed. Notice also that if M is nonparacompact. also metrizable. a Riemannlan metric on M. but a much easier proof is now available: Let ( . Problem AI I implies that a manifold N immersed in a paracompact mani fold M is paracompact. (On the other hand.474 CH APTERS 7. not necessarily compact. for the folium is a submanifold.. Appendix A 9. So there are at least two inequivalent nontrivial bundles over M. The tangent bundle of the long line L is clearly orientable. a locally compact connected topological group is acompacl (Problem 104. The converSe also holds: slnce a Riemannlan metric determines an ordinary metric. since an equivalence would determine a Riemannian metric.) 4. so there can· no/ be a nowhere zero Iform w on L! for w and the orientation would de termine a nowhere zero "ector field. We have seen that any paracompact Coo manifold has a Riemannian metric. Y. 6. 7 V'ie can now dispense with the argument in the proof of Theorem 66 which was used to show that each folium of a distribution on a metrizable manifold i. submanifolds. O)a) of the Prufer manifold. and hence paracompact. the tangent bundle TM cannot be trivial. they cannot be extended to non· paracompact manifolds. 5. contradicting the fact that the tangem bundle is not trivial. ). then TM lS definitely not equivalent to T* M. . then N has the Riemannian metric f* ( . if f: N M lS an immerslon. Theorem 79 fails for L. Compare pg. 3. 2. 2 7 2. A Lie group is automatically paracompact. a basic result about bundles says that a bundle over a paracompact contractible space is trivial. Thus. 10 I. Thus the tangent bundle of the long line is not trivial. Since there is no Riemannian metric on a nonparacompact manifold M. since its tangent bundle is trivial. More generall\'. as can be seen by considering the Odimensional sub· manifold {(O. nor is the tangent bundle of the Pnjfer manifold. ) b.
TIlis will b. proved in Volume II (Chapter 8. It i s not clear that a nonparacompact manifold cannot have an lndefllll](' metric (a nondegenerate inner product on each tangent space). Addendum 1). Is there a nowhere zerO 2form on the various nonparacompact 2mani folds which have been described? . PROBLEM 13.Ajijlclldix A 475 7.
.
75 65 101 T'M T®S 109 1 10 1 21 107. i. 61 34 34 28 34 35 61 6] v" v { f) Iv) .i) Till" T' Xp J . �) O (n) R(n) (�n . 1 1 9 1 13 131 109 1 16 1 16 1 20 121 121 1 22 I ii .· ' CHAPTER 3 E"(X) J. 1 1 6. p I'm F 65 72 83 65. U ) aJ ar I ax. �) SO(n ) (x. VnJ IA � Ix. �.'(V) Tt' (� ) 7ik ( V ) I' J p' End(V) df d.. v p a . (p). . ax . 'U) SL(n.NOTATION I NDEX CHAPTER ] CHAPTER 2 M p2n pn sn aM JRn 8(X) lHln 23 19 4 S' 11 19 1 6 19 .· dl de di t(l ax l a 76 68 64 75 68 64 103 82 83 81 76 64 80 84 72 101 102 81 80 83 l CHAPTER 4 Hom( V. .i)p np TM T(M. W ) Tk < v ) Tk(i' ) T V) d T. x �2 � Mp (M. a a ar' a8 DJ � Ip ip 62 29 61 62 28 35 39 36 d. .' X At C' CO COC CW DJ(a ) GL(n. }. � Ell 1.
2 1 5. g ) mj '[La) wa d f l(LI) i1·w Iw dw M. x) (01. Vk ) (v" . 224 215 22.«) ('R. II P II Lx w S. 8fi.I/.) / ')" r" . 1 0... 1'·* V' 1 23 10. II. CHAPTER q" 7 202 205 238 238 2 19. . Vk ) IA I (/(1 2 ) exp A Lx A Lx f Lx ) " Lx w [X.J w <PI a· /\ a· CHAPTER ("" Q(M) Q k (V) QO(l ') (v. /\ dxn l de den d8ta .c) f' ' .wl (V) n. 1 29 1 34 1 34 1 08 1 09 1 0. . . 234 23.478 N olation Inde> T (I.. " . (. signp J w(p) Zk ( M ) . . X I. 235 2 10.n OIx(l) = 01 (1 .�:��/ . . 234 202 Ali curl X All I dx ' /\ . w(X) Iv t' �j l . 5 171 161 171 1 74 1 5 (1 1 50 1 50 1 ii 1 53 1 43 1 35 1 44 7ik1ml (V) 7i�ml ( V ) 7ik1n...f Hk (M ) H� ( M ) I' deg f div A" de t(f.wI 7i� ( V ) T°(V) 231 231 231 231 20 1 227 206 202 227 215 20 1 201 203 v.h. 2 1 3. 10. Y] CHAPTER 8 263 268 250 284 275 258 252 290 297 292 274 263 268 246 249 296 283 283 239 264 275 293 263 Bk (M ) B� ( M ) cu.i l .il.
) (g. I Jo CHAPTER A ( ( ( ( ( ( ( « I II II rti .u E(II) Elld(nl exp O C O B C' 'J [ij.. 245. �) g[(". HI ii(u) oj 0. 257. 259. 356 356 314 314 31 I 309 309 324 334 302 306 32G 312 344 sinb tanh x'( v) xi r . 404 373 410 385 385 372 407 407 376 CHAPTER d(p.\ 479 Z. 291 264 264 263 248. }/ . 285 285 260 246 239 243. . ). )' .1 IAI A i) Ad( ) ad X Aul( g) a Euc(V) Euc(�) E(y ) J' ( ..N olalion il1dc.!.J ) exp d. 313 356 356 349 335 335 318 319 314 353 328 30 1 315 308 301 349 305 301 367 303 303 315 10 384 372 409 410 409 396 402. �) .' // / J dx + g dy /w Lw # (' (q) u [O. 246./J L� Mp" exp(A) GL(II. )r . 248.) . )' . I II II.. q) ds d l' ("osh ("osh .(M) 6n a' 0' [wJ al' di ( 268 285 29 1 264. ). } . 288 294 246 299 299 266 9 322.
Ix 6" [' M # !I' p. 0(13 ) .480 i. L. APPENDIX A k=d L W' U! to Q Old <> + 1 464 463 464 464 461 46 1 463 . £(G) o(n) P Ck (M) 9!(N) PD RI.I [ f(o)da " . 'II p(W A ry) [ .] [ry A A] p x (M) x (�) 419 446 4 32 453 439 457 442 426 423 435 428 445 439 w (natural nvaluecl J f ormJ [ fa" [ . N otation index 401 374 379 376 388 376 41 1 41 1 374 373 376 379 380 395 403. 410 403 376 403 400 400 400 403 CHAPTER I I L.: <jJ. u .1 R" SO(n) X .
466 plane. J O� dual. 55 AJgebra. 451 Atlas. 293 AJgebraic inequaJities. �). 31 2 function.. 465 manifold. �).INDEX Abelian Lie algebra. 46� Circle. 233 Alternatinc ("ovaria� t tensor field. 7 i trivial. 1 9 long ray. 35. 1 03 Ado. 45 I sphere bundle. 36 i halfspace. F. 108 fibre. 20J Alternation.. 409 Area. 46b half plane. 7 ) Basi� dual. 395 Adams. 4!=l . A S" 1 7 9 Big disc. 27�: point. ] ] I tangerlt. 466 Biinvariant metric. 1 07 for Mp'. 379 Bundle cotangent. 1 9 subgroup of a Lie group. 2 1 0 vector. 1 54 in gl(n. 23(l CauchyRiemann equations. 2� Auslander. 380 Alexander's Horned Sphere. 3 3 2 Arcwise connected. 40J Boundary. R Antipodal map. 202 Analytic manifold. 391 submanifold. 7 3 nplane. J 45 Base space. J. 7 J BuraliForti Paradox. 34 Annihilator. 309 induced. 382. 360 Cartan's Lemma. I CJ Boy's Surface. Elie. 248. 208 or Belongs to a distribution. 46() Calabi. 1 1 1 Chart. 248. 285. 283 Choice function. 20S f g. 200 Cayley numbers) 100 Chain. 348. I I Arclength. 252 Bounded manifold. 38 Change) infinitely small. 106 quadrant. 19. 60 Bracket. 7 1 normal. generalized. ] 20 of covariant tensors. 285 Chain Rule. 20i multilinear function. 246 Associated disc bundle. 3 9 . 378 in o(n. 344 of contra\'ariant tensors. 218. L. 376. l Oll Adjoint T* of a linear transf ormation T. 19J Besicovitch. 252 geodesic. principle of irrelevance of. 1 0 1 map. 464 Banach space. 228 Annulus. 28 maximal. £" 472 Calculus of variations. L n. 6 Closed form. 7 2 . 28 Choice. Fundamentru Theorem of. 3 1 6 Cartan.k(p). 20 subgroup of a Lie group. 3 1 3.
465 Cohomology. 420 Commutative Lie algebra. 421 analytic structure. 34 C� distribution. ] 2] . 1 20 vector field. 286 Degree. Uy. 227 Crosscap. 4 i ) numhers of norm 1 . 78 of a ring. 295 Densin· eve. ]7 9 form. 133 Derh<ation. 148 DiAerentiable. 279 Degenerate. 25 DescartesEuler Theorem. 28 Darbom: integrable. geodesic'. 284 Debauch of indices. 214 Deformation retraction.10 pointfinite. 29 structure. 39. J O �l Covariant functor. 228 Definition. 123 Decomposable. 28. 38i Contractible. 28. 14 Crossproduct. 283 Darboux's Theorem. 207 function. 372 Critical poim. invariant. manifold. 32 at a point. 232 DiAeomorphk.. 429 Coordinate lines. 3 1 manifold. 429 Determinant. 39. 220. 238 Cvlinder. 341 Complex. 1 1 3 COWl locally finite. 30 on the long line. 3ib Complete.482 Closed (contiuued. 259 relative scalar. 439 Curl. 299 Cube. 5(1 Cramer's Rule. 4 19 d e Rham. 28 Cotangent bundle. 24b index Cup product. 358 Constants of structure. 133. 263 of a complex. ] 30 tensor field. singular. 225. 15H Coordinates. 139. scalar. I 1 3 Convex geodesicaU)" 363 polyhedron. 40 i ll the calculus of \·ariations. 15� Coordinate system. 42 [ Commutative diagram. 30 oneparameter group of. 47 1 on pn. 22i Lemma. 130 tensor field. 30 Dilleomorphism. 83 Derived set. 263 group of M with real coefficient:. 8 C. . vector field. 236 Contraction. I I . 209 odd scalar. 32 Riemannian metric. 23 J scalar. 1 33. 32 manifold. 308 structure on TM. 27. 34 CO manifold. up to first order. 1 39 Contravariant functor. 29 manifo]dwithboundary. 60 refinement of. 3 1. 32 . 275 mod 2. 39b Continuous homomorphism. 65. 299. 3i3 Conjugate. 82 Coorelated. 40 Cross section. 283 integral. J 6{1 Cofinal. 320 Critical value.
space. 139 Foliation. 207 differential. 433 of vector bundles. 1 69 Equations of structure. l OB Euler. 16S forms. 201 of a function. 254 . 72 weak. 194 Folium. 3 19. 254 Functor. 374 right invariant. 384 Extension. 404 Equivalence (or vector bundles). 233 Disjoint union. 3 1 S motion. 215 Fubini's theorem. 109 Dimension. 96 Euclidean metric. 438 First element. 421 Disc bundJe. 1 0 7 vector bundle. 334. 169 linear. 355 . 29 on s n . 205 type. 1 03 E. 4� End. 238 Theorem. 358 Equations depending on parameter�. ponential map. 240 Form. 68.. 194 Force field. 30 Differential. 36i Embedding. 164 depending on parameters. 1 8 1 ideal of. Five Lemma. 23 1 relative tensor. 2 1 5 o n torus. 432 Extremal. 2 113 sequence. 23 1 scalar density. 20 Distribution. 179. 422 of a pair. 1 36. 4 Direct sum. 209 Exact form. 23 paracompact. 429 characteristic. 39 Elements of norm 1 . 32(1 Even ordinal. 468 Endomorphism. 7 1 Finitt> characteristic.' Differentiable (cOTlliTlurd (structure con1iTllled. 10. 3 Du Bois Revmond's Lemma. 320 483 Einstein summation conv(. on �n . 352 Domain. 89 Fundamental Theorem of Algebra. 64. leap of. associated.. 464 Fibre. I Faith. 1 30 Functorites. 461 First variation. 440 Fixed point. 324 Envelope.ntion. 374 nspace. 285. 293 Fundamental Theorem of Calculus. 45 1 Discriminant. 467 relative scalar.Inde. 1 92. 4. J 2 1 Energy.. 32. 428 class. 133. 134. 419. Exponential of matrices. 1 8ll Divergence. 201 left invariant. 385 . 445 Euler's Equation. 305. Dual basis. 190 Frobenius Integrability Theorem. 308 Elliptical nonEuclidean geometry. 400 Irelated. 2 1 0 equation.
23i GramSclllnidt orthonormalization process. 277 Hopf. 23:. 7 Homomorphism continuous. 349 Infinite volume. 84 Hausdorff. 432 Global theory of integral manifold�.. 104. 283 . 350 Independent infiniteBimals. 69 HalfIonp cylinder. 192 Integrable functioll Darboux. 277 Homotopy. 342. 46 Implicit function theorem. 36. 47 Immersion. 3 1 4 Initial conditions. 44 6 Indite:debauch of. 3 7 J Guillemin. 2 3 Heiukin. I I I Infinitely small displacements. 194 Gradient. principle of irrelevance of algebraic. reversing map.484 Gauss's Lemma. H. 1 9 Handle. 14. 2 1 Integrability conditions. 1 7 9 Ha s one end. 333 closed. 387 of Lie algebras. 145 Hair. 104. 40i orthogonaJ. 356 tangent. 462 Inner product. 372 topological. H. Sylvester's Law of. I nertia.. 136 of integral ClJrve. 46S strip. 30l preserving. 363 ' Geodesy. 246 Geodesic. 349 Index of vector field on a manifold. 44i on �n . 14 Immersed submanif old.. 337 General linear group. 49 topological. 84 Group Lie. V W. 356 Ideal of a Lie algebra. 356 sine. 227. 342 Inde> Hyperboli( cosine. 410 Identification. 466 Halfspace. 3 7 1 matrix. 283 Riemann. 459 Homogeneous. 148 Infinitesimals. 372 usual. 314 Index of inner product. 1 0 Imbedding. Robert A. 304 Gmk. 101 orientation. 106 HahnBanach theorem. 39. 61. 3 i 2 opposite. 351 Induced bundle. 123 raising and lowering. 46 lOpologi �al. l 36 Initial segment. 401 Geodesically complete. 341 Geodesicallv convex. 301 Inside. 37� Generalized area. 450 Hopf Ril1owde Rham Theorem. 380 Homotopic. 3 33 Germs of kforms. independent. 260 Inequalities. 466 line. 189 Integrable distribution. H Hardy. 3 1 4 Infinitesimal generator. G. 60 Indefinite metric. 304. 3 1 2 Infinitely small cbange.
239 In\'ariance of Domain. 464 Kink. 340 Isomorphic Lie groups.. 3 1 2 o f a curve) 59 Lie algebra. E.. principle of.]. 243 manif old. 1 8 1 maximal. 394 n form. H.. 382. 473 Lang. 194 of a differential equation.ket in any ring. 43S 485 vector field. 472 Kneser. 245 Integration. 233 Isometry. 395 commutative. 435 Kneser. 1 7 J Lin� ar transformation ac\joint of. 3 76 abelian. 368 Local flow. ]. 136 Darboux. 150 Lie group. 283 surface. 4 1 6 Limit ordinal.. 400 . 34 9 Leaf. 407 Lie derivative. 3 7 1 arcwise connected subgroup of. 373 Lie's fundamental theorem first. Invariant. 388 Lie $ubgroup. 239. 103 contraction of. 296 Lipschitz condition. 138 Littlewood. 18. 463. 243. J44 Lie group.inde. 2 J 4 Irrelevance of algebraic inegualitie�. 374 Length. 104 Linking number. . 58 Lav�' of ] nertia... 148 spanned locally. l O P Isotopic. 165 svstems of. S. 155. 121 positive definite. 194 Leap of f aith. natural. 21. 460. 464 Left iJwariant form. 366 Klein bottle. 226. 1 36 Riemann. 376. 230 Laplacian. 376 homomorphism of. :M. 243 Linear differential equations. locally: 382 lsomorphism. 410 topologically isomorphic. 294 Jacobi identity. 4 1 5 oneparameter group of local diffeo morphism. 7 1 Local theory o f integral manifoIds� 190 Kelley. 378 Jacobian matrix. 232 definition. 60 Line integral. 128. 415 third. Integral curve. 136. 1 1 9 Lobachevskian nonEuclidean geome try. 104 positive semidefinite.. 179. 179 triviality. 305. 410 opposite. 40 J ordan Curve Theorem. 179 Lives at points. 283 line. 239. 374 Left translation. 409 closed subgroup of. 463 ' set. Sylvester's. 376 for the blac. 414 second. 391 local. 415 normal su bgloup of. 380 ideal of. 145 Laplace's expansion.
304. 350 Riemannian. 3 4 ' atlas for. 20 finite cover. 2 1 4 Manifold. 30 I NonEuclidean geometr>" elliptical. Ion Multiindex. 2�1 diflerentiable. H . 1�1 C' 34 Co: 34 Coo.m. 465. I . 20 connected. R. 9 nmanifold. 208 Multilinear function. tubular. 401 Euclidean.vitllbnundarv. S. 20 Milnor. 3 Nice cover.486 Locall) compact. V. 2�1 Manifold. Ht) smooth. 4 . 438 Nonbounded. 4 imbedding in �N. 403 Natural isomorphi. 308. 465. 344 . 305.. 20 Long cylinder.<. 466 N onorientablc bundle. 460 closed. 7 1 nsphere. 46:} Lower sum. i ntorus. 86 manifold. 3 1 1 usual. 4 J Mesh. 2 8 boundary o C 1 �l ' bounded. 73 rank of. 179. J 15 Munkres. 424 for compact supports. 367 Lobachevskian. left invariant.]. 345 Newman. 466 line. 4 5 9 analvtic.'.1 Mobius strip. :)7�' Maximal intcgral manifold. J 9 closed. 37:2 Normal bundle. 43 J :Measure zero. 36H Nonmctrizable manifold. 2�1 dimension of. 4 ( 1 Massey. 1 9 4 nOllmctrizable. 4 nplane bundle. 239 Metric biinvariant. l OCi Magic. Coo. 19 Nondegenerate. 42 1 bundle. E. 50 isomorphic Lie groups. R. 303 preserving.]. 1 39 oneone... 283 index MayerVietoris Sequence. disjoint union of. ](l generalized. 3� Map betweell complexes. I�l . 46] open. 52 integral. W. 34. 3 1 5 indefinite. 400 nholed torus. 46" ray. 1 8 1 maximal. A . M . 40. 4 nforms. 1 3 path wise connected. 3 1 2 spaces.. 108 Neighborhood. 3 I\'latrix grou ps. 8(1 Norm. 194 ndimensional. 4 6G orielltation of. 382 Lipschitz. 42 Mod 2 degree. i Natural gvalued Iform. J06 MacKenzie. 29.
l l6 Projection. 468 Parameter curves.. 367 Point inward. 384 Opm long ray. 407 Lie algebra.index 487 Normal (cantinued) space. S. 248 reversing. 148 Oneparameter subgroup. 39 Pointfinite cover. of a bundle. 52 Path wise connected. 88 Proper map. 60. 225 Poincare upper half plane. 86 Oriel1latim. 351 Nowhere zero section. 266 Polarization. 35 Partition. 100. 85 of a manif old. yellow. 450 Poincare dual. 465 map. 450 Poincare Lemma. 435 space. 304. 86 of a vector space. 3 1 3 Partial derivatives. 461 Ordering. 461 isomorphism. 133. 95 . 460 Ordinal numbers. 407 Orde. 461 topology. 248 Orthogonal group. 86 manif old.. 441 PoincareHopf Theorem. 148 of local diff eomorphisms. 259 Onedimensional distribution. 167 Parameterized by arclength . 32 Projective plane. 439 Poincare Duality Theorem. ] 34. 84 preserving. 7. 60 submanifold. 465 Ordered set. GramSchmidt. 301 Positive element of norm I . 467 relative tensor. 88. I I. 98. 20 Piecewise smooth. 304 Osgood's Theorem. 304 Pollack. 225 Paracompact. H. 308 Positive semidefinite. 351 Odd ordinal. 209 Outside. 19. 275 Prufer manif old� 467 Pseudometric. 105. 260 Outward unit normal. special. 410 outward unit. isomorphic. 36 integration in. 21 Outward pointing. 463 Orientablc:bundle. Opposite group. 30. 61.. 104. 245 of unity. A. 84. 434 Poincare. 98 outward. 3 1 2 Pig. 17!=l Onedimensional sphere: 6 Oneparameter group of diff eomorphisms. 84. 288 scalar density. 372 Orthonormal. J06 Positive definite. 459 subgroup of a Lie group. R. 102 tensor. 88. 85. 210. 459 end. 284 Palais. local. 60 Polar coordinates. 104 Product of vector bundles. 239. 260 Pointderivation. 348 Orthonormalization process.
7 3 zero. 263. positive. 3 1 2 Smoothly contractible. 62 Sphere. 459 value. 3 J J usual. 268 de Rham's Theorem. 58 Spanned locally. 134. 290 Space filling curve. 42... 2 3 1 tensor.. 400 Right translatioll. 246 simplex.. 100 of norm 1. 354 Schwarz inequality. 264 def ormation. 22i Simplex of a triangulation. 424 f compact supports . 104 Separate points and dosed sets. 321 de Rham. associated. 4 7'2 Rotation group. 220 homotopic. 28 homotopy. G. 134. 277 manif old. relative. 59 Rdlnement of a cover. 28:. 54 Smooth. 472 Rank of a f orm. 7 Sphele bundle. 308.. 248 Retraction. 134. 433 Shrinking Lemma. sum. 5 I Shrinking Lemma. 294 Solid angle. 20J. 283 integral. 133 Schwarz. 40 space. 103 MayerVietoris. 104 Semidefinite. 231.. 263 with compact supports. surface of. 95 Sequence exact. 382 Singular cube. 50 Regular point. 45!=l Section of a vector bundle. initial. 9 6 Segment. 6 J Special orthogonal group . T. \fl. 22� of a map. 342 Roman surf ac{'. 17'1 Special linear group.connected. 451 RadiaJ f unction. 435 Rado. 427 singular. 40. 457 RicmaJ1Jl integrable. 244. 288 Reparameterizatiol1. 3 1 2 Right invariant nform. 287 Lie group. 303. M. 1 90 Relative scalar. 285 Simply. 422 of vector bundles. 362 Second countable. 279 Revolution.'.488 Quaternions. 6(J Shuffle permutation. 40 Related vector fields. 231 Scalar density. 3iB Slice. 462 Selfadjoint linear transf ormation. H. 43J or of a pair. 194 Slice maps. 419. 29 piecewise. J7. 62 . 285 Skewsymmetric. 277 isotopic. 98 Rectifiable. 26 Roscnlichl. 3i 3 Sard's Theorem. 283 Riemannian metric. 374 Rinow. 8. 294 Scalar. 342 de Rham cohomology v{'ctor spaces.
14. 123 contravariant. 8. Whitney. 49 closed. 64 Tangent vector inward pointing. 461 Trivial vector bundle. 33. 384 Submanif old. 426 simpJex of. 203 Whitney. 98. 349 Symmetric bilinear f orm. 77 Tangent space of �Ii. 301 System of linear diff erential equations. 231 odd relative. 4 7 open. 134. 374 right. 63. 64 outward pointing. 8 Tangent bundle. 106 '. 460 Translation left.. 4 .Inrie. 7 1 Totally disconnected.. 42. 147 Surface. J W. 456 Topological group. 25 Transitivity. 426 singular cube. 7 area. :2 Su ccessor ordinal. 468 Structure constants. 106 Stokes' Theorem. 17. 352 Stonee ech compactification. 442 Thorn lsomorphism Theorem. 3 Wedge product. 120 covariant. 1 1 3 even relative. 3 7 1 imbedding. 4�) immersed. 66 Vick. 464 Sum of vector bundles. 396 Subalgebra of a Lie algebra. 116 Thorn class. 72 Tubular neighborhood. 303 Triangulation. 459 489 Tensor contravariant. 9 Total space. S. 261. 50 Subgroup Lie. 321 Sylvester's Law of Inertia. 285. 76 of �". 3 7� Subbundle. H. 8 nholed. 26 Sternberg. 198 Subcover. 1 2 1 . 1 1 3 mixed. 98 of a manif old. 120 covariant. 373 oneparameter. 101 Support. 7. 134. 101 . 46 Topologically isomorphic Lie groups. 288 Tensor field classical definition of. 221 Steiner's surf ace. 345 Twoholed torus. 171 acompact. 374 Triangle inequ ality. 354 integral. 246 Starshaped. 49 Coo.\'hitney sum. Standard nsimplex. 239 of revolution.. 122 Tensor product. 42 7 Trichotomy. 388 Torus. 1 4 immersion. 260 to a curve. 7. 253.
.
and additional special symbols. including bold symbols. . which was created in three essential sizes: 9. with su itably modified versions used for su perscripts.hese books were typeset using Donald E. Although a Baskerville bold face is unhistorical.e Rime qfthe Ancient Ma. Knuth's 'lEX typesetting system. script letters. The bold face supplied by Monotype� even the "semibold": is obtru sively extended.. Numerous other modifications of this sort. DVIPSONE PostScript driver. bold type was useful in special circumstancesmainly for indicating defined terms. The text font is I I point Monotype Baskervillethough the emdash has been modified together with its italic. as well as the fact that the upper case Baskerville letters are of somewhat heavier weight than the lower case. The figures were produced with Adobe Illustrator. are loosely based on Samuel Tavlor Coleridge's poem TI. The bold letters for mathematics come directly from the bold fonts of the Times families. together with the Monotype Times N R Seven and Times Small Text familiespresumably these three families are based on the original designs for Times New Roman. . On the other hand. since simple scaling would have made them u npleasantly heavy. including additional kerns and alterations of set widths. were made for various pu rposes. and new or modified fonts were created using FontographeI. and J cause problems in words like topologl' and apology. in superscripts. The somewhat bold appearance of chapter headings. A thicker set ofnumerals was constructed for use with the upper case lettering in chapter headings and statements of theorems. 7 and 5� point. while blackboard bold letters were made by hollowing out these bold letters. The elegant swashes of the italic ). and special parentheses and other punctuation symbols were also required.. results from the linear scaling. T together with Berthold H orn's The mathematics fonts are a variation of the Math TlIne fonts: now based on the Monotype Times New Roman family. painted by the author in his spare moments. special gg and gf ligatures were also required. so a special gy ligature was added. in 16 point type. so a nonextended version was created. The Adobe Mathematical Pi 2 font was used for the ordinary sized German Fraktur letters. The italic fonts of these three families were used as the basis for creating the three separate " math italic" fontsfor use at ordinary size. The covers. the tall initial letters beginning each chapter were designed specially. as well as for the extension font and its bold version. and in second order superscripts. The proportions and weights for these were then used for the three sizes of the symbol font and the other mathematics fonts.in.
JR2 . Xk) is a Coo f unction for all COO vector fields XI . 103. and aM itself is always a manifold JR" (but only one homeomorphic V pg. homeomorphically onto 2 T(M. 6 1 . . 70: Replace the last two lines of page 70 and the first two lines of page 7 1 with the f ollowing: correspond to ( p. pg. �3 . . 14: relabel the lower left part o f the central figure as di(x. . .� �\ B: :B A. 43: Replace the last line and displayed equation with the f ollowing: Since rank f =k in a neighborhood of p. . M. line 3: change pg. Add the hypothesis that = e(fh) = J(p)e(h) + h(p)e(f) = 0 + e(f). corresponding to a M x JR2 .y) ::: 1. 1 1 7: After the next to last display. . 1 9 : replace the nexttolast paragraph with the f ollowing: to The set of points in a manifoldwithboundary that do not have a neighborhood homeomorphic to a homeomorphism (without boundary). then is Coo. . Xk)(p) If A pg. . Problem 30: Change part (f) and add part (g): pg. A(X" . 78: the third display should read: 0 = e(O) pg. pg. the function J can be made Coo if (g) The same is true if M has at most countably many components. theorem of topology). A is also. restate part (c) as f ollows: (c) This is f alse if J : (f) If M is a connected manifold. JR is replaced with J : M. . and we could continuously pick w (p) E criterion that w(p) should make a positive angle with v(p).y) < 1 to di (x. x E aM if and only if there is a neighborhood of x and O. then A is Coo. Problem 32: For clarity. 22: Replace the top left figure with IHI" such that M and is denoted by 4>(x) = aM. . the lower rectangle in the matrix ( a(Ui ) o f) ax} = pg. then aM = 0. If there were a way to map + N for a disconnected manifold N.. 1 1 8: Add the following to the statement of the theorem : If . by using the up would pg. Problem 29(d). 3. If M is actually a manif old. fibre by fibre. then each dashed vector. .i). in the sense that A(X) . . A (p) (X. A. Equivalently. liP) is called the boundary of 4> : V + pg. + M is a Coo manif old. . . v(p )) for some v(p) E JR . Xk . = pg. . Xdp)). there is a proper map J : M + JR. ( p). . 60.A is Coo. . '� .CORRECTIONS FOR VOLUME I pg. add: M is orientable.
when the transformation rule involves (det (7) A)W. . End with End . the lower right entry should be "Iforms on pg.. show that . 2 C V. ajaXI. 233. pg. V375" refers to pg. 237.] is integrable. 0<2 be two maps on some open interval 1 such that 0<1(1). and then replace Problem 19.i. 226.1) x (n . . what are called "even scalar densities" might best be called "signed scalar densities". . In Problem 26. Aut should be replaced with Aut . must also assume that each /'. 143. Extending the changed terminology f rom pg. part (c) should read: . 1 1 9 : Add the f ollowing at the end of the proof: Smoothness of pg. 133. and suppose that f 5. For consistency with standard usage. The display in Problem 21 . . p. pg. 292. 134. h component of VI x ··· X VnI in terms of the (n . the condition V. For X Y E Mp and .. replace parts (b) and (c) with: I (b) Determine the . # 0. .p N..+I pp. pg. part (d) should begin: = for some to E I. pg. The hypothesis o f Theorem Let pg. 4 1 1 . . 0<2(/) 0<1 (10) = 0<2(10 ) 0<1 ' (1 ) = f(O<I(I)) i = And the first sentence of the proof should be deleted. 1.pg. is . we should probably speak of the bundle of "signed tensor densities of type and weight w " (though sometimes the term relative tensor is used instead. . Though there is considerable variation in terminology. A follows from the fact that the function Ail . . Problem 16 (b) should read: "For any Lie group G. for 1R3 . scalar densities In part (c) of Problem 10. " . pg. 1 3 1 . we should be considering the h of part (a). omitting the modifier «signed" when it involves I det A l w. ). . Problem (d) Let f: M pg. 408. what are here called "odd scalar densities" should probably simply be called uscalar densities". The reference "pg. In Problem + . p N". In Problem 20. . restricting densities to those of weight 1). 177. 198. 375 of V olume V pg. 133. Ell /'. show that n Vj # 0 should be CJ n V. O. 408410. Problem 9: Let F be a covariant functor from V. . we 17. . In the comutative diagram. V. 3 should b e changed s o that i t reads: x and E V and let 0<1 . add the hypothesis that pg. In part (g) of H is a connected Lie subgroup. not the h of part (b)! Thus conclude that the bundle of signed (not the scalar densities) is not trivial if M is not orientable.A(ajaxl" .. pg.1) submatrices of the matrix In particular.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.