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
. grateful to the class for keeping up wi th my accelerated pace. f or otherwise the second half of these notes would not have been written. I am also extremely grateful to Richard Palais.XlI Priface to the First Edition in manifold theory with a pace that few would describe as excessively I am leisurely. whose expert knowledge saved me innumerable hours of labor.
Vector bundles . . . . . . . Coo functions Partial derivatives Critical points . . .. . . . Addendum. . Orientation . . . . THE TANGENT BUNDLE The tangent space of �n The tangent space of an imbedded manifold .TABLE OF CONTENTS Although the chapters are not divided into sections. the listing for each chapter gives some indication which topics are treated. . . . . . . . . Xlll 63 67 71 75 77 82 84 89 95 . and on what pages. . . I . 6 20 2. The tangent bundle of a manifold Equivalence classes of curves. . . Immersion theorems Partitions of unity Problems CHAPTER 27 31 35 40 42 50 53 3. . . Equivalence of Tangent Bundles Problems . . . DIFF ERENTIAL STRUCTURES COO structures . MANIFOLDS Elementary properties of manifolds Examples of manifolds Problems CHAPTER . .. . . . . . CHAPTER I. and derivations Vector fields .
DIFF ERENTIAL FORMS Alternating functions The wedge product . . and contraction Problems versus 1 07 109 III modern terminology 1 15 1 17 121 127 CHAPTER 5. . 1 35 1 39 143 148 150 1 53 1 64 1 67 1 69 CHAPTER 6. Differential of a form Frobenius integrability theorem (second version) Closed and exact forms The Poincare Lemma . TENSORS The dual bundle The differential of a function Classical Multilinear functions Covariant and contravariant tensors Mixed tensors. 2. Brackets . . . classical integrability theorems Local Theory. Forms . Differential Equations Parameter Curves in Two Dimensions . . . V ECTOR FIELDS AND DIFF ERENTIAL EQUATIONS Integral curves Existence and uniqueness theorems . Frobenius integrability theorem Global Theory Problems 179 190 194 198 CHAPTER 7. Addendum Addendum Problems I. . .. INTEGRAL MANIFOLDS Prologue. . . The local flow Oneparameter groups of diffeomorphisms Lie derivatives . Problems 20 1 203 207 210 215 218 225 227 .CHAPTER XlV Contents 4. . . .
LIE GRO UPS Lie groups . 30 1 308 312 316 323 334 341 344 348 CHAPTER 1 0. . . Stokes' Theorem . . Geodesic completeness . . . . . . Addendum. de Rham cohomology Problems . . . . . . INTE GRATION Classical line and surface integrals Integrals over singular kcubes The boundary of a chain Stokes' Theorem . RIEMANNIAN METRICS Inner products . . ..Contents CHAPTER xv 8. . . . . Oneparameter subgroups The exponential map Closed subgroups Left invariant forms Biinvariant metrics . .. Tubular Neighborhoods Problems ... . Left invariant vector fields Lie algebras . . . Problems 37 1 374 376 379 380 382 384 39 1 394 400 402 406 . . Subgroups and subalgebras Homomorphisms. 239 246 248 253 256 258 261 263 283 CHAPTER 9. . .. Riemannian metrics Length of curves The calculus of variations The First Variation Formula and geodesics The exponential map . . The equations of structure . Integrals over manifolds Volume elements . . .
M aye rVietoris sequence for compact supports The exact sequence of a pair Poincare Duality . The Euler characteristic. INDEX . 459 467 47 1 472 473 474 475 NOTATION INDEX . 9. The Thorn class · · · · · · · · · . . 2 6 7. . . · · To Chapters Problem . . . . . 10 . . Index of a vector field . . . . . . . EXCURSION IN THE REALM OF ALGEBRAIC TOPOLOGY Complexes and exact sequences The MayerVietoris sequence Triangulations . . . .CHAPTER XV! Contents I I. . .. . . . . . 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 . .
Comprehensive Introduction to DIFFERENTIAL GEOMETRY A VOLUME ONE .
.
xi)2 . This example immediately suggests the next: any open . d(x. a manifold is a metric space property: If 0 E �. just �n itself. [If you know anything about topological spaces. To be precise. . =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. Whenever we speak of ffi.y) = �)yi . ric with which mention it. supplementing various chapters.n. . which should be consulted if one allows a manifold to be nonmetrizable.CHAPTER 1 MANIFOLDS numbers. 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. x and some integer lent metric (one which makes it homeomorphic to �n with the usual metric). where � is the set of real x = (xl.xn) with each E �. . you can replace "metric space" by "topological space" in our definition. x' unless another metric is explicitly suggested. then there is some neighborhood U of U is homeomorphic to �n . . for each � n we can take U to be all of �n. ] The second simplest example of a manifold is an open ball in �n. and we shall almost never U to be the entire open ball since an open ball in �n is homeomorphic to �n. �n supplied with an equiva M is endowed plays almost no role. 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. in this case we can take Indeed. 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. is also a manifold. consisting of . Clearly.n as a metric space. of course.
though by no means all. Exercising a mathematician's penchant for generalization. there is an open ball W with ¢(x) E We ¢(V). This compli cated little argument just shows that we can always choose the neighborhood U in our definition to be an open neighborhood. ConseIf is a point of a manifold some open set x quently. 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. homeomorphic to W. then ¢(V) C �n is an open set containing ¢(x). of course. M. quite naturally.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. some preliminary remarks need to be made. Thus x E ¢l(W) C V C U. which constitute mOst of this chapter. 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. . it is. and thus open in M. The open subsets of �n already provide many different examples of manifolds (just how many is the subject of Problem 24). Since ¢: V � �n is continuous. the set ¢1 (W) is open in V. and thus to �n. an open submanifold of the original manifold). Before proceeding to examine other examples.
Manifolds With a little thought. in fact. example. (It follows that f(V) is open for any open V C V. 0 and z=I} then we can choose n=2 for points in M.y. The quickest routes arc probably pro vided by Vick. it begins to appear that. Topology ojPla ne Sets. = 1 were false. sO fI is continuous. we can define di or J di= i 1 + di M. then V C �n is open and f: V � �n is oneone and continu f(V) c �n is open.z) : x = M={(x. and 11 = 1 for points in M2 • This M. 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. for example. . for it implies that the property of being a "domain" (a connected open set) is invariant under oneone continuous maps into �n. Singular Homology Them]. and M2.y. U M2 . y) < 1 for = min(di. Homology Tlteor)1 and Massey.) 1. is an unnecessarily complicated device for producing one manifold from two. THEOREM If . In general. Theorem 1 is called "Invariance of Domain" . we need the following theorem. treatment may be found in Newman. * All proofs require some amount of machinery. But to prove this. 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.y E Mi. I ). by the way. with metrics d. we di with an equivalent metric Ji such that Ji(x . and d2 . given can first replace each all x. if M C �3 is O} U {(x. then there would x. For example. stated here without proof' ous. but pleasantly geometric.z) : z = M. Notice V in our definition was not open). and f is a homeomorphism. An oldfashioned. 3 V must be open.
A disjoint union of manifolds M. M is clearly a manifold also. is left to the reader. sO is any discrete space {o ifx=y if x # y.". then there is a oneone continuous map from ffi.4 Then we can define a metric Chapter 1 d(x. uncountable discrete space is not acompact (it cannot be written as a countable ::I 111. The only compact subsets of such a space are finite subsets. If X is a connected. A manifold has dimension n or is ndimensional or is an nmanifold if it has dimension 11 at each point. In the new space M. This construction can be applied disjoint union of the metric spaces Mi. if not. In particular. both M. n can work at a given point x E M. a nonopen subset of is unique at each Consider once more a discrete space. they can be replaced by new sets M. Consequently. and M2 are manifolds.y)= I = {�i(X'Y) d on M = M. for if n > 111.m into 2. The further deduction. 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). the manifold is not connected. then X is acompact. as we see by taking a disjoint union of uncountably many manifolds homeomorphic to ffi. Y E M otherwise i to any number of spaceseven uncountably many. we note that 11 �n. We will often need to know that this is the only way in which acompactness can fail to hold. the resulting metric space is called the is a manifold. locally compact metric space. which is a Odimensional manifold. Although different Il'S may be required at different points of a manifold it would seem that only one M. U M2 by if there is some i such that x. and M2 are open sets. defined by the metric d(x. THEOREM. This unique n is called the dimension of M at x. It is convenient to refer to the manifold M as Mn when we want to indicate that M has dimension n. As a first step. since a space with one point is a manifold.y):::r} . an union of compact subsets). PROOF For each x E X consider those numbers ball r > ° such that the closed {Y E X:d(x. and M2 are disjoint. In these examples. that the n of our definition x E M. If M. The same phenomenon occurs with higherdimen sional manifolds. however.
then for each all such x E X define rex) to be onehalf the least upper bound of X= r. be a sequence in A'. > 0 is an interval. . Then A' is also compact. proof is as follows.y) sn} . For each i there is a Yi E A such Zi is in the ball of radius r(Yi) with center Yi. this set includes all r > 0.. then X is acompact..Manifolds 5 is a compact set (there is at least one such r > 0.d(xj. The set of all such. . for all y E A. converges to some point Y E A. 00 The triangle inequality implies that {y E X: d(xj. n=1 U {y E X:d(x.Y) S r + d(x"x2)}.Z3. � Z1. and X2 gives (2) � is continuous.Y) S r} C {y E X:d(x2. The so the function r: X consequence Suppose A C . Let A' be the union of all closed balls of radius r(y) and center y.x 2)} C {y EX: d(x2 . {y E X: d(x"y) S r . which we might as well assume is the sequence itself. since If not. If. so that which implies that S (I) Interchanging x. some subsequence of the Yi. Now the closed ball B of radius �r(y) and Let that .y) r}.Z2. for some x. Since A is compact. This has the following important X is compact. since X is locally compact).
Then there is some Y E A with d(x. defined by Sl = = {x E �2: d(x. I rex) . or I ·dimensional sphere. y) �r(x). so x E A. {y E X : dey. �r(x) �r(x) 2 > d (x. and A # 0 is ope and closed. Thus A' is compact.6 center y is compact. we present the longpromised exam ples of manifolds. It is also closed.:. � I = = < By (I). Now let Xo E X and consider the compact sets A {xo} An+1 An'. it must be that X A.O) = I }. So the sequence Zi is eventually in the compact set B. Moreover. Their union A is clearly open. suppose that x is a point in the closure of A. S I. the limit point is actually in the closed ball of radius r (y) and center Y (Problem 10). . . After this hassle with point·set topology.2d(x y) > rex) � . Since X is connected. This shows that if y E A n ) then x E A n'.  3 = 3 II which i s acompact. Since eventually the closed balls Chapter 1 ? Yi Y and since the function r is continuous. The only connected Imanifolds are the line � and the circle. r(y) =:: .y). Yi) r (Yi)} are contained in B. To see this. and consequently some subsequence converges.
then MI x M2 is an (nl +n2) manifold. is also a homeomorphism. .sine) E Sl simply by e E [0.) The function g: (Jr. Considerations similar to these now show that the nsphere S" = is an nmanifold. .sine) is a home omorphism. In particular . is our first example of a compact 2manifold or surface.Jr) � Sl.(Of course.Manifolds 7 The function J: (0. it is always necessary to check that use of this notation is valid. From these few manifolds we can already construct many others by noting that if Mi are manifolds of dimensionlli (i = 1.2]'[]. I) .!<. We will often denote the point (cose.•   the above diagram. I) onto the line � x {I} C � x �. better suited to generalization. 2). by projecting onto � x {I}. it is even continuous. illustrated in (0..2Jrl. 2Jr) � Sl defined by J(e) = (cose. . I)} onto � x {I}: this is proved most simply by calculating P: Sl . X Sl is called the ntorus. n times {x E �n+1 : d(x. or it suf fices to note that Sf is "homogeneous"there is a homeomorphism taking point into any other (namely. while Sl x Sl is commonly called "the torus". though not oneone. and it is also homeomorphic to a certain subset of �3 which is what most people have in mind when they speak of . The point (0. on [0. an appropriate rotation of �2). The 2sphere S2. I) may be taken care of similar ly. together with J it shows that Sl is indeed a manifold. There is another way to prove this.!2.{CO..{CO. The pro jection P from the point (0.. I)} � � x {I} explicitly. It is ob viously homeomorphic to a subset of �4. commonly known as "the sphere". is a homeomorphism of Sl . defined by the same formula. O) = I} Sl x ...
The next simplest compact 2manifold is the 2holed torus. The resulting surface.z) E �3 : (y_I)2 + Z2 al"Ound = 1/4} contained in {CO.). called a surface of revolution.y. more precisely. the latter of which is also homeomorphic to the annulus. To provide a mOl'c explicit description of the 2holed torus.8 Chapter 1 "the torus": This subset may be obtained by revolving the circle {(O.. a spare homeomorphic to a torus with a hole ('ut out. the region of the plane contained between two concentric circles. it is easiest to begin with a "handle". has cOlnponents homeomorphic either to the torus or to the cylinder S' x �. we throw .:) E �3 : J' > OJ.
.. : :. : : .. .h boundary piece of the sphere. I ' . . " . ' " :: '=' There is one 2manifold of which most budding mathematicians make the acquaintance when they still know more about paper and paste than about . It is homeomorphic to the space obtained by starting with the disjoint union of 11 handles and a sphere with n holes. " I . it is also described as the disjoint union of two handles with corresponding points on the boundaries "identified". which remains in OUf handle. and which will be referred to as the boundary of the handle. . _____ .h handle with corresponding points on the i. '83 . " I . :: .9 away all the points on one side of a certain circle.. _____ . _____ . 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. and then identifying points on the boundary of the i. " . C3EE1.
not to two disjoint circles. and ( I .1.  o � . but our next example will force the issue.2".+ 1 cos � sine. which you "make" by giving a strip of paper a half twist before pasting its ends together. as investigation of the paper model will show. " .( 2sine . I ) x (.l) = 2 cos e + 1 cos circle of radius 2 . I) I) ° 1  � • . . cos � sine . .2". the Mobius strip can also be described as [0. 1 sin � ) .) x [I . .�cose. We wish to identify each point x E S 2 with .) x ( . we obtain the Mobius strip with a boundary.10 � Chapter 1 metric spacesthe famous Mobius strip. this boundary is homeomorphic to a circle.I. . I) with (0. If we define Ion [0. y ''''e have not yet had to make precise this notion of "identification".1) � �3 defined I(e. This can be described by analytically as the image in �3 of the function I: [0.. I) instead .:e) 2cosB � . "identified". With our recently intro duced terminology.
p. 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). 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 . The points of 11'2 are defined to be the sets {p. is a lot harder to visualize than previous examples. for which J(p) = J(q) implies p = ±q. We will denote this set by [p) E 11'2.Manifidds 11 its antipodal point x E S2. The space which results. there is no subset of �3 which represents it adequately. 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. A subset V C 11'2 will be open if and only if JI (V) C S2 is open. so that [p) = [p). . We thus have a map J: S2 � 11'2 given by J(p) = [p). indeed. the projective plane. We will postpone for a while the problem of defining the metric giving the distance between two points [p) and [q). 11'2. p} for p E S2.
{ 0) ::: I}. J: [0. y)plane anyway. [(0. which is homeomorphic to the disc D2 = x E �2 : d(x.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.1)) [(1. JI (V) [0. The dotted lines in this picture are the key to understanding 1l'2 If we distort the  . V (s. Squaring things off a bit.I)I)) ifs=OorI. 1) (s.12 Owpter 1 In exactly the same way.I).I)}.(I. )')plane.1) (1.I)) { res. This leaves the upper hemisphere (including the bounding circle). 1 J«s. 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. we can make things easier for ourselves by first throwing away all points of S2 below the (x. I) (s.1) {(O. since they are identified with points above the (x. and we must identify each P E Sl with p E Sl .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.1) (1. 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.l) (0. s # 0.1) (s.
a little experimentation wil1 convince you that this cannot be done (without having the two pieces of doth pass through each other). is to be identified with the same thing at the lower right. which is a single circle). Unfortunately. in reverse direction. can still be described mathematically in terms ( B 1 1 I . The projective plane is thus obtained from . by identifying points on the boundary and points on the boundary of the disc.ffi. There is dearly a continuous function f: p2 ?. in other words. al though not homeomorphic to 11'2 .. : u of p2 . we obtain a Mobius strip with B l " " A " a boundary (namely. that is. the dotted line. it is locally oneone. Such a function J . we are left with two pieces which can be rearranged to form something homeomorphic to a disc. at the upper left.. every point P E 11'2 has a neighborhood U on which J is oneone. 3 whose image is this subset. although J is not oneone. 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. moreovel. The subset of �3 obtained as the union of the Mobius strip and a disc. � " ' A B the disjoint union of a disc and a Mobius strip with a boundary. both of which are circles...Manifolds 13 region between them a bit w e see that the front part o f B followed by the back part of A. If this Mobius strip is removed.
(This is also obvious from the fact that the Mobius strip is the projective plane with a disc removed. .. with an extra dimension to play around v·.. however. . . When the resulting figure is drawn up into 3space and the appropriate identifications are made we obtain the "crosscap".. . . .rith. ..... . the disc can be added so as not to intersect the Mobius strip. although not topologically imbedded (there is no homeomorphism J from 1l'2 to a subset of �3). . The figures below show that the Mobius strip may be ob tained from an annulus by identifyjng opposite points of the inner circle...) This inner circle can be replaced by a quadrilateral. .. .14 Chapter 1 is caned a topological immersion (the single word "immersion" has a more spe cialized meaning.... . The crosscap together with the disc at the bottom is a topologically immersed 1l'2 I��\ D'\jC � . explained in Chapter 2). . . this Can be done in the following way.. . ... . . . We can thus say that 1l'2 can be topologically immersed in �l. 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. . In �4.... . .
) For the present. whidI ought to be manifolds.xy. Roughly speaking.:. The missing metric can be supplied by an appeal to Problem 31. 15 'J'he one gap in the preceding discussion is the definition of a metric for p2 . (fhose who know about topological spaces will recognize it as a disguised case of the Urysohn Metrization Theorem. suppose that J(x. Clearly J(p) = J( pl· We maintain that J(p) = J(q) implies that p = ±q.e).b. =±x a+b+e ' a ( ) ( ) .. how('ver.y. so we also have hence (3) Using (2). 'Ie) prove this. we wi)) obtain our metric by a trick that simultaneously provides an imbedding of ll'2 in �4. Consider the function J: S2 � �4 defined by J(x.z) = (yz. 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). this gives a+ b+c= ±(x+y+z). which will Iatcr be used quite often. first of all (I) II' a. and which the reader should peruse sometime before J't·acling Chapter 3.:. are.y.Manifold. b e .a+b+c=±x 1+.xz.e yz= be xz=ac xy= abo #0. it shows that things like ll'2 .z) = J(a.+.b. We have.x2 +2y2+3z2).
z) But y = z = 0 implies (by (I) again) that be = (±I. so (5) gives = 3 . =3 _ _ y2 ) = = I I. ['IJ) d (j([p)). (x.O) or (O. Now suppose a = o.y. Clearly. _ b2 (8) z (this holds even if y = b = since then z. Y Now (4) gives = = 3 ±b . 0. j(['l))) d (j(p). ". If x then (I) would immediately give y = z = so that #0. = 0. = 0 or e = 0 and Thus x = (4) 0 also. Similarly. (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]. so b (a. which completes the proof.y2 (7) 2y2 + 3 (1 y2 .' j([p) ) J(p)."ith the same sign (which comes from (3)) holding for all three equations.c) These equations clearly contradict = (O.±I. In this case we have proved our contention without even using the fourth coordinate of f. = ±c Since J(p) J('l) precisely when p ±'l.16 Chapter 1 so x = ±a. 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 . Z = ±b.O). we obtain y = ±b.c = ±I).O. 0.O. b . J('l))' = = = = . we can define j: ll'2 � �4 b.±I).
x' .x'). or. and famous.z)) (yz. (O.xz. For example. representation. we can create other surfaces in t he same way as the nholed torus. We can also join together a pair of projective planes with holes cut out. By the way. Notice that the [0. to a handle we can attach a projective space with a hole cut out. it has a nicer. 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. = V"ith the new surface p2 at our disposal. a !\16bius strip. what amounts to the same thing. which amounts to sewing two Mobius strips together along their boundary. g([x. the map ll'2 � �3 defined by the first 3 components of g: j. (I. Consider the surface obtained from the square with identifications indicated below.x) [0. The image in R3 is Steiner's "Roman surface".xy) is a topological immersion of p2 in R 3 . Although this can be pictured as two crosscaps joined together.y. The closest we can come to picturing this is by drawing a cross cap sticking on a torus.Manifolds 17 Then one can check that the open sets are indeed the ones described above.
which separates the surface into two parts. one of which is shaded. dividing the sides into thirds. Xl). to which il can then be joined. forming the "Klein bottle". and P4 to be identified.x) �  . (O. P4} is a single point of our new space. I) x Sl immediately suggests an immersion of the surface. P3. The description ill terms of [0. P2. P2. below. The dotted lines Pl P. 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. so that the set {PI. P3. form a single circle. Turning one end of the cylinder around and pushing it through itself orients the lefthand boundary so that is directly opposite (I.18 Chapter 1 identifications on the square force PI.
then there is some neighborhood U of x and some integer .. If x E M. then all points of M will be boundary points of M. Points no these "boundaries" do not have neighborhoods homeomorphic to ffi. we will mention the t'('bted family of "projective spaces". 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. If manifoldswithboundary are studied as frequently as manifolds. If M is actually a manifold. . 11 A point in a manifoldwIthboundary cannot have a neighborhood homeo morphic to both�" and IHI" (Invariance of Domain again). in addition to the family of nmanifolds S".j1J sometimes use "bounded manifold" instead of "manifoldwithboundary". Notice that if M is a subset of�".x") E�" :x" . but they do have neighborhoods homeomorphic to an important subset of�". but. p} for p E S". but \\.. The (closed) halfspace IHI" is defined by !\ manifoldwithboundary is a metric space M with the following property: IHI" = {(xt. Often. .1I. We have already discussed some spaces which are not manifolds only because they have a "boundary". A manifold in our sense is then called "non bounded". we v. for example. 19 Examples of higherdimensional manifolds will not be treated in nearly such d"tail. the word "manifold" is used lor "manifoldwithboundary". The set of all such x is called the boundary of M and is denoted by aM. a nonbounded compact manifold is calJed a "dosed manifold". then aM = 0. indeed. then aM is not necessarily the same as the boundary of M in the old sense (defined for any subset of�"). . if M is a manifoldwithboundary of dimension < 11.::O}. we Can therefore distinguish those points x E M having a neighborhood homeomorphic to IHI". One further definition is needed to complete this introduction to manifolds.:: 0 such that U is homeomorphic to either�" or IHI". Projective nspace Il'" is defined as the rol1ection of all sets {p. the Mobius strip and the disc..rill see later that the spaces pll for even n differ in a very important way from the same spaces for odd n. We wi11 stick to the other terminology. it becomes bothersome to use this long designation.Manifold.
where the Y. Each X. while d(x.y) if x.) 4. y)== 3. and consequently homeomorphic to the di�joint union of its components. where X = U. (a) A subset of an nmanifold is an nmanifold if and only if it is open.20 Chapter 1 PROBLEMS 1.. but a direct proof of arcwiseconnectedness can be given for manifolds. (e) Every manifold is locally connected. n Xj for i # j.joint union of its com ponents. (b) Every manifold is locally pathwise connected. and a connected manifold is pathwise connected. 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. . < I.11 but an arc is a oneone continuous image. which are open submanifolds. (a) Connectedness does not imply local connectedness. is homeomorphic to X.d) (or any space homeomorphic to it) is called the disjoint union of the spaces X. so every neighborhood of a point in a locally connected space contains an open connected neighborhood. (a) The neighborhood U in our definition of a manifold is always open. for i E I.. (d) A locally connected space is homeomorphic to the &. (b) If M is connected. for some i. is an open subset of X.d) � (X.y E X. then I is either increasing or decreasing.d) is a metric space. (c) X is locally connected if and only if components of open sets are open.d) is a homeomorphism). Y. A difficult theorem states that every path contains an arc between its end points. the identity map I: (X. are disjoint open sets and Y. for each i. (b) The integer n in our definition is unique for each x. Show that if d is a metric on X. (a) If U c � is an inten'al and J: U � � is continuous and oneone. '" 2.) are metric spaces. then the dimension of M at x is the same for all x E M.(x. d) are also metrics and that they are equivalent to d (i. If (X. X. 7.. then both d d!(I + d) and d min(I. (a) Every manifold is locally compact. The space (X. then (X. 0') An open subset of a locally connected space is locally connected.e.d. (A path is a continuous image of [0. and X. 6. and d(x. with metrics d.y) = 1 otherwise. d. 5.. (c) A connected manifold is arcwise connected. and Y is homeomorphic to X if and only if Y = U.
. then �n . 21 V 8. = {x E �n : d(x. (b) The image I( ) is open.{(O. show that the limit of a convergent subsequence or the Zi is actually in the closed ball of radius r(y) and center y.p whenever it contains p. 12.1I for m ::I n. Explain how the two cases differ. . which is discussed immediately afterwards.A has 2 components. (c) The map I is a homeomorphism. For this problem. 0') Check that 1l'2 is a surface.. Ie)!' 9. c �n is homeomorphic to Dn (a) One component of �n . (2) If ffi.A (the "outside of A") is unbounded.I)} � �nt. (First prove c. assume (I) (The GeneralizedJordan Cun'e Theorem) If A C �n is homeomorphic to snt. then /(inside of A) inside of I( A) . (b) Do the same for I: snt .) (e) Pl'Ove Invariance of Domain. and the other (til<' "inside of A") is bounded. Every connected manifold (which is a metric space) has a countable base its topology.. then V = V V 10. (a) Compute the composition I = s' . and A is the boundary of each. (a) Check that the metric defined for 1l'2 gives the open sets described in the text. and show that it is a homeomorphism. Show that this last condition is actual1y unnecessary. � 13. 11.. In the proof of Tlleorem 2. 14. 0') The analogous condition is necessary for the Mobius strip. (h) I f c �n is open.O.1I  B B is connected. A C is homeomorphic to snI and I: � IR:II is oneone and continuous (so that J is a homeomorphism on A). where V c S2 is open and contains .O) S I}. (a) The text describes the open subsets of 1l'2 as sets of the form I(V).Manifold. (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. . I)} �t x {I} � �t explicitly for the map P on page 7. and a countable dense subset.{CO.
22 Chapter 1 16. and that all compact surfaceswithboundary are obtained from these by cutting out a finite number of discs. 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. C is homeomorphic . and since antipodal points in snI are still antipodal when considered as points in sn. (a) Show that pI is homeomorphic to Sl . Show that pn . Show that �2 to the surface shown at the top of the next page. Let C c � C �2 be the Cantor set. we can consider pnI C pn in an obvious way. To which of these "standard" surfaces are the following homeomorphic? 15. a hole in a hole in a hole 17.O) < J}. (b) Since we can consider snl C S n.pnl is homeomorphic to interior Dn = {x E �n: d(x.
(c) This part requires some knowledge of topological spaces.{OJ is not homeomorphic to�m. (b) �n .1 23 (a) If C c� is compact. (b) Show that�n has only one end c for n > I. show that c(C) is either always the "left" component of� .{OJ does not "have one end" so�n . More generally. Let 8(X) be the set of all ends of a connected. Consider the following three surfaces. the "right" one containing all numbers> some N. 18. Show that Xu 8(X) is a compact Hausdorff space. X has exactly one end c if and only if X "has one end" in the sense of Problem 18.C.C has exactly 2 unbounded components. Define a topology on Xu 8(X) by dlOosing as neighborhoods Nc(co) of an end cO the sets Nc(co) = 19. co(C) U {ends c: c(C) 20. or always the "right" one.C. in such a way that C. An end of X is a function c which assigns to each compact subset C c X a lionempty component c(C) of X . Thus� has 2 ends. it will be used in Problem 24. c Cz implies e(Cz) C c(CIl. then� . 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 . This problem is a sequel to the previous one. locally compact Hausdorff space X. the "left" component containing all numbers < some N.Manifold. � . but not if n = I. locally connected. (A) The infiniteholed torus: for all compact C. What is� U 8(�).K is connected. If c is an end of�. (a) �n has one end if n > I. and�n U 8(�n) for n > I? = co(C)}.
Jll10l0' _ �JlllUlll �Il\ fiJllL � __ I _ e Ie . Now draw in two more lines enclosing more holes. (b) Surfaces (A) and (C) are homeomorphic! Hint: The region cut out by the lines in the picture below is a cylinder. lH 1H 1\( 1\1 �1 ]l11'. which occurs at the left of (A).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. while surface (B) does not. and consider the region between the two pairs.
. (b) If A c �2 is closed and totally disconnected (the only components of A are points). .j. then e(�2 . For the purposes of this problem we will use a consequence of the Urysohn Metrization Theorem. .e. (c) The derived set A' of A is the set of all nonisolated points.Bare nonhomeomorphic if A and B are nonhomeomorphic closed totally discon nected sets. . ) by goes to co at co. (a) The three open subsets of �2 shown below are homeomorphic..A and �2 . For each n there is a subset An of � such that A/n ) consists of one point. that for any connected manifold there is a homeo morphism from to a subset of the countable product � x � x . J M M J: M j(M) M. (a) Every open subset of � is homeomorphic to the disjoint union of inter vals. if { n} is a sequence � co. Hence �2 . 23.• • infinite swiss cheese . ..Bto be homeomorphic even though A and B are nonhomeomorphic closed subsets. (a) It is possible for �2 . (a) If is a connected noncompact manifold...J(x». j(x) . (b) The points inside the three surfaces of Problem 20 are homeomorphic. . J j: M J: M x J(x ) n g: M(g(x). � () () () () .j C) 22.. () () : i i.Manifolds 21.. (c) There are at most C nonhomeomorphic connected manifolds (where c = 2No is the cardinality of �). then there is a continuous function � � such that "goes to co at co". (b) There are only countably many nonhomeomorphic open subsets of �.A) is homeomorphic to A. . C)··" 24. . . . • • C') () () . i.A and �2 . 25 ···· • • • C) C) () () . .. which is eventually in the complement of every compact set. .. then (Compare with Problem 230. We define A(n) inductively by A(I) = A' and A(n+l) = (A(n». and a � � which = � � x (� x � x .) (b) Given a homeomorphism � � x � x . define Show that is closed.
. is aM. . c) on Steiner's surface satisfies b'e2+a'e'+alb' (b) If = .co) of each axis.U. then aM is a closed subset of M and aM and M . then the topological boundary of M.b.26 Chapter 1 '(d) There are c nonhomeomorphic closed totally disconnected subsets of �'. e) E �3 : b'e'+ a'e'+alb' = abe} is the union of the Steiner surface and of the portions (co. (a) Every point (a.b. . then M .e) satisfies this equation and 0 oft D = Jb'e'+a'e'+a'b'.'h derived set is {cil. then (a. 26. (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. This is not necessarily true if M is not a closed subset. b.eI' C. For each sequence n) < nz < . abe. i E J are the components of aM.b. one can add a set A i such that its n l1. and e) < e. is a manifoldwithboundary. (e) There are c nonhomeomorphic connected open subsets of �2 25.' .1/2) and (I /2. < e3 < . (b) If M is a manifoldwithboundary. (c) If C. 27. Hint: Let x = be/D.e) is on Steiner's surface. etc. Hint: Let C be the Cantor set. as a subset of �n . (a. a sequence of points in C.. If M C �n is a closed set and an ndimensional manifoldwithboundary. (c) The set {(a. and I' c J. .aM are manifolds.
This is the case despite the fact that M is locally like �n. are contained in Chapters 2 and 3 of Calculus on Manifolds. 229. rather than by :n: (which will be used often enough in other contexts). we will denote the identity map from �n to �n by I. where differentia bility of functions can be defined.. The neces (x) we can expect J 0 1/1 1 to be differentiable for all J which make J 0 </>. . This is certainly not always 27 + .' differ entiable only if </> 0 ". since Wsary tools of "advanced calculus". 226.I : �n � � is also differentiable.CHAPTER 2 DIFFERENTIABLE STRUCTURES freshly sharpened. Unfortunately. which the reader should bring along e are now ready to apply analysis to the study of manifolds. however. and U n V # 0. If U C M is an open set and we choose a homeomorphism </> : U � �n. it would seem reasonable to define J to be differentiable on U if J 0 </>1 : �n � � is differentiable. On a general manifold M the notion of a continuous function J : M � � makes sense. 225. 332.I : �n � �n is differentiable. 215. then it is not necessarily true that J 0 ". We will use freely the notation and results of these chapters. including some problems. if "' : V � �n is another homeomorphism. and 335. so that I i = Xi . but the notion of a differentiable function J : M � � does not. notably 29. Indeed.
namely. V ) of an atlas A is called a chart (for the atlas A). we will be interested almost exclusively in functions J : �" � �n which are Coo (that is. instead of x alone. It is necessary to adorn our manifolds with a little additional structure. it will suffice to consider homeomorphisms x: V � xCV) C �n onto open subsets of ffi. VVe can even imagine a mesh of coordinate lines on U. one need merely choose </> to be h o 1/!. The use of the letters X. V ). which has "coordinates" xl(p). henceforth ad hered to almost religiousl)� is meant to encourage the casual confusion ofa point p E M with x(p) E �n. If V and V are open subsets of M. just to provide a convenient name for the domain of x . 1/! are in the collection. A particular member (x . �n If we insist on defining differentiable functions on any manifold. and is automatically true. where h : �n is a homeomorphism that is not diff erentiable. We will often mention the pair (x. Moreover. since x( V n V) and y( V n V) are open subsets of �" . where it is hard not to lapse back into the practice of denoting points by x and )'. )" etc. each component function Ji possesses continuous partial derivatives of all orders). . it makes sense.. . for these homeomorphisms. for example. . This make sense. xn (p). for the obvious reason that it provides a way of assigning "coordinates" to points on U. instead of considering homeomorphisms from open subsets U of M onto �n. xn (p ) to the point p E U. or a coordinate system on V. This is preCisely what we shall do. . . sometimes we will use the words "differentiable': or "smooth" to mean Coo. the coordinates xl(p). we wish to select a certain collection with the property that </>01/!1 is diff erentiable whenever </>. The only time this notation will be confusing (and it will be) is when we are referring to the manifold �n. Also. . there is no way out of this impasse.28 Chapter 2 � the case. the precise nature of which is suggested by the previous discussion. . but a few refinements will be introduced along the way. by considering the are Coo./. . Among all possible homeomorphisms from V C M onto �n. . 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. First of all. if V n V = 0.
. where . The advantage of this bigger atlas 'U is that the single word "chart".4.).4. is contained in a unique maximal atlas A' for M. 29 The simplest example of a manifold together with an atlas consists of �n with an atlas .. or smooth manifold) to be a pair (M. LEMMA.n parallel to one of the axes. Let . 'U). 'U differs only superficially from . (and one would be foolish not to do so once and for all). We can easily make the atlas bigger.4. PROOF... What has just been said for the atlas {I} applies to any atlas: I . 'U) and (�. the identity J : �n � �n.' 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 . The term actually used is x E 'U if and only if x 0 J E V.4. we can adjoin any homeomorphism x : U � V with the property that x and XI are COO. •:. 'U) and (�.. . then . V) where V contains the homeomorphism x t? . V) are not the same. the obvious map J(x ) = x 3 • Thus (�.. If . V) are the sort of structures one would want to call "isomorphic". Thus. Indeed.4. about the simplest example of a Coo manifold is (�n. Although (�. we can adjoin as many such x's as we likeit is easy to check that they are all COOrelated to each other. so .4. Another example is (�..4. is a maximal atlas for M. when applied to this atlas. . and it is clearly the unique maximal atlas containing A .. one can easily construct 'U from . denotes something which must be described in cumbersome language if one can refer only to A Aside from this.' is an atlas.\"3 . We now define a Coo manifold (or differentiable manifold. if U and V are homeomorphic open subsets of �n. whose inverse is not Coo.4.. there is a oneone onto function J : � � � such that namely.' are Coorelated. is an atlas of COOrelated charts on M..4. together with all charts Coorelated to it. of only one map.. where 'U (the "usual COOstructure for �n") is the maximal atlas containing {I}.Differentiable 5lructureJ inverse images under x of lines in ffi.4.
It is easy to see that a diffeomorphism must be continuous. . .. .(x) = g.I ) of snI are easily seen to be Coorelated. . : snI n {x E �n : x . up to diff eomorphism.A) and (N. xn).30 Chapter 2 "diff eomorphic": two Coo manifolds (M.the "usual Coo structure for snb. There are. which are part of the field called "diff erential topology". Consequently. I) and (0. . .I .I n {x E �n : x . This atlas may also be described in terms of the 2n homeomorphisms defined by f. 'Li). . (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. . They therefore determine an atlac. < O} � �nI . Later (Problem 924) we will be able to prove easily that � with any atlas is diffeomorphic to (�. Ifwe had not required our atlases to be maximal. . 0. . 'Li). we shall not come close to proving these assertions.n for n . and II is clearly a diffeomorphism also. .::: 5 requires very difficult techniques from topology. But there are 28 diffeomorphism classes of differentiable structures on S7.(x) = (Xl. and the proof of the essential uniqueness of Coo structures on ffi. the proof for �3 would ccrtainly be too difficult for inclusion here.+ I . and speak elliptically of "the diff erentiable manifold M". 2 A proof of the corresponding assertion for � is much harder. we will suppress mention of the atlas for a differentiable manifold. . X. Normally. .>4' on any open submanifold N of f. and over 16 million on S 3 1 . unique differentiable structures I on sn for 11 ::: 6. : sn . the definition of diffeo morphism would have had to be more complicated. conversely.. However. rather then differential geometry. of course. which are Coorelated to P and P2. . the projections PI and P2 from the points (0. x. This raises the natural question whether. 0. the atlas for M is sometimes referred to as the diff erentiable structure for M.I g. its inverse must also be continuous: so that a diff eomorphism is automatically a homeomorphism. > O} � �n. . but we can already describe a diff crentiable structure . two homeomorphic manifolds are necessarily diff eomorphic. It will always be understood that �n refers to the pair (�n. . . In the case of spheres. :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.
Just as diffeomorphisms are analogues for Coo manifolds of homeomor phisms. 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. to the usual diff tiable structure on �. with V C N. V) with p E V and J(p) E V. (4) a function J: M � N is differentiable if and only if each y i diff erentiable for each coordinate system y of N.4. 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. the atlas . V ) is a diffeomorphism from V to xCV ).4. it is easily seen to be true for any other pair.Differentiable Structure.. . V) in . (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. the map y 0 J 0 XI : �n � �m is differentiable. A function J : M � N is called differentiable if for every coordinate system (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 . More particularly. this coincides with differentiability of the restricted map JIM' : M' � N.. a diff erentiable map is continuous.' consists of all (x. Consider first the product M. there are analogues of continuous maps. V) for N.4. The diff erentiable structures on many manifolds are designed to make certain functions differentiable.. eren A differentiable function J : M � � refers.). V ) and (y.' 31 a differentiable manifold (M. Clearly. We can thus define differentiability of J on any open subset M' e M. (3) a coordinate system (x. as one would suspect. V ) for M and (y. of course. x M2 of two diff eren . and hence J is differentiable if and only if J 0 XI is diff erentiable for each chart x. If this is true for one pair of coordinate systems.
32 Chapter 2 tiable manifolds MI. The map X 0 (f I U)' is a homeomorphism on leU) C ll'n. there is a differentiable structure on ll'n which makes the map f: sn � ll'n (defined by f(p) [p) {p. where U does not contain . one needs to know that there are lots of I hem. pg. x M2 which makes each 7[1 diff erentiable. A "handle" is then a Coo manifoldwithboundary. . It is easy to define a diff erentiable structure on M. 29). To deal with Coo functions effectively. It is only necessary to know when a map f : lHln � �n is to be considered differentiable. = = To obtain differentiable structures on other surfaces. we cali f differentiable when it can be extended to a differentiable function on an open neighborhood of lHln. so that f l U is oneone. We briefly recall here the necessary facts about such Coo functions (c.p if it con tains p.pi) diff erentiable. A differentiable structure on the 2holed torus can be obtained by "matching" the diff erentiable structure on two handles.culus on Malfif olds. The collection of these homeomorphisms can then be extended to a maximal atlas. Consider any coordinate system (x. U) for sn. we first note that a Coo manifoldwithboundary can be defined in an obvious way. and the two "projections" 7[i : M. For each pair (XI. . UI) of coordinate systems on Mi) we construct the homeomorphism = defined by Then we extend this atlas to a maximal one. x M2 � Mi defined by 7[i(p" P2) Pi .f. 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. and any two such are COOrelated. Ca!. The details involved in this process are reserved for Problem 1 4. Similarly.
8 > . Similarly. il is 0 for x (4) The function g: �n � g(x) j(XI ) . e) x ..Differentiable Structures 33 is Coo. it is positive on (e. j(xn ) = ..I) .:::: 0. e) and 0 elsewhere. h: � � hex) { eI /x' x oft x h(n>(O) j: � � { e(xI)' . Let C c U e M with C compact and U open. there is a Coo function k : elsewhere. " x (e. and � defined by /e je is Coo.1) I � � (I) The function = � defined by 0 o =0 n. ]i[ +0 0= 1 x for E 2: 8. = 0 for all � defined by = o � which is positive on (0. e(x+I>' x E (I. 8) and 0 (3) The function I : � � � defined by is Coo.8). increasing on (0. Then there is a Coo function f : M � [0. e On a Coo manifold M we can now produce many nonconstant Coo func tions. LEMMA. J(x) x ¢ (1. (Compare Case 2 of the proof of Theorem 15. and denoted simply by support f (or sometimes supp f)· {x : /(x) oft O} 2. The closure is called the support of f. such that f = on C and support f e U.) 1) I . and (2) The function i s Coo.
This will not be proved here. and the sum. Of course. choose a coordinate system (x. The symbol ew stands for analytic. is Cr ifit has continuous partial derivatives up to order r ) . but this atlas can always be extended to a bigger atlas of COrelated charts. of the corresponding functions has support C U. Let J = I o (f + .1unkres. # co will hardly ever be mentioned again. l:. One remark is in order now. Since C is compact. Clearly it remains Coo if we extend � U� it to be 0 outside of V. V) with V c U and x(p ) = o. the smaller structure is the "stronger" one. every Ccx structure contains a CfJ structure for each f3 > 0' . 2: o. it is not unique. 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 . •: " r = By the way. the proof of Lemma 2 produces an appropriate Co filllction J OIl a Ccx lnanifold./) which converges in some neighborhood of a). e ) x .. ).34 Chapter 2 PROOF. . P p.ffi. Thus. A "C o function" isjust a continuous function. (A function J : ffi.. Co manifolds for ". emellta1JI Diff ential Topolog)'_ 7 er . as in Lemma I. so on C it is 2: 8 for some 8 > o. so a CO manifold is just a manifold in the sense of Chapter I. and is positive on a neighborhood of p whose closure is contained in U. Then xC V) ::l (e. 2: I. x (e. . If ". then the charts of a maximal C� atlas are all COrelated. + I . e ) for some e > o. ..n ?. . . the CO structUl"e (consisting of all homeomorphisms x: U � �n) being the largest. we could have defined C' manifolds for each . not just for 00" . and it is convenient to agree that r < 00 < (J) for each integer .* In f act. . Let J be the extended function. where I is defined in (3). Q' .. On C it is positive. The converse of this trivial remark is a h ard theorem: For Q' � 1. . + J Pl p. The function g o x (where g is defined in (4)) is Coo on V. For each p E C. for 0 :s. for Q' = (J) the proof * For a proof see 1\. J l + .. but it is unique up to diffeomot·phism. finitely many such neighborhoods cover C. of course. < {3. a C� structure on M can always be extended to a Ccx structure in a unique way.:::: 00. The function J can be P P constructed for each p.
J(p) = aJlax' (p). At this point classical notation for partial derivatives is systematically introduced. If we .Diff erentiable Structures 35 fails completely (and the result is falsean analytic function which is 0 on an open set is 0 everywhere). for a function Dj (f o g) (a) = L D.. then h. h.(f 0 x. With differentiable functions now at our disposal.I (x(p) + (0. O lim Now.)(x(p» ..I ) define the curve Cj : (e.U). so it is worth recalling a logical notation for the partial derivatives of a function J: �n � �. .". We denote by D..J(p) h ' {I (Joct!' (0). . . then is the classical symbol for this partial derivative. I = ' as an equation between functions). it is fitting that we begin differentiating them.J(a) the number J(a' . . . . in fact it is just Notice that ifi = j ax' . .U) we define aJ aJ D. . ..(h» .. an) . then this partial derivative is just If x happens to be the identity map of �n.J(a) . ax} h� O r J(c. What we shall define are the partial derivatives of a dif ferentiable function J: M � �. � (p) = 8}� = 0 if i # j. e) � M by (or simply aI aXl = D. so it measures the rate change of J along the curve c.a' + h. Dd (a). .J(g (a» . 0» x 0 0 x. . with respect to a coordinate system (x. h The Chain Rule states that if g : �m � �n and J: �n � �.(J J: M � � and a coordinate system (x. . .. i= I X. D. which . . . ax' (p) = ax' p c.(h) = X.
"raxis" 2" not restricted to the set A... y) is restricted to lie in the appropriate interval (eo . y). eo + 2". One Can delete any ray other than L if e(x..).36 Chapter 2 Another classical instance of this notation. y) y = r ex. (Of course. often not completely clarified.I is defined simply by p . y). OJ "Oaxis" � +.% (x . Jx2 + y2 and e(x. with its image being the set {r : r > O} x (0. �.I (r. y) = we can introduce a "coordinate system" P : A ?. . y) sin e(x. e) = (r cos e . y) = (r(x.� . and this sometimes justifies the sloppiness involved in the definition of the polar coordinate system.) with x = r ex.. On the subset A of �2 defined by A = �2 = �2 L _ {(x . r sin e). 2rr).2".. whose inverse p . e (x.2 by p(x. the polar coordinate system is often + (r.ffi.. yJ This really is a coordinate system on A in our sense. is the use of the symbols alar and alae in connection with "polar coordinates".y� .) We havc really defined P as an inverse function. many results are essentially independent of which line is deleted. y) E �2 : y = 0 and x ::: O} where r ex. y) cos e (x . y» .. y) is the unique number in (0.
e along the "eaxis". y» ( This formula just gives the value of the directional derivative of J at (x..y)) . "v' _ = ofa curve along the "raxis".y» · D. y) . y). DI J 0 P . because CI . y). is just a line in this direction. y) = DI /(r' (p(x. so aJ a.'n9(X. sin e(x. ..'hich is the inverse image under P of a cun.(x. .:. y) + Dzf(x.". y).sin e(x. [r')' (p(x.y) sin e(x.� / Xl' \ C2 .) /1.y) /8(x. y») + Dz f(x. sin e(x. The factor " (x. y) appears because this curve . cos e(x. y»). r sin e).. . . y» by the Chain Rule DI /(x. y). y) cos e(x.DifJerentiable Structures From this formula we can compute aJlar explicitly: (f 0 p . y). the inverse image under P (x' Y).I ) ( P(x. The vector w = ( . y). y) · cose(x. A similar computation gives C l _ _ cos8(x. at the point (x. e) = = 37 J(r cose. and thus the (x. y) [r(x.Y) = DI /(x. y) % direction.:. y)) pointing outwards from the origin to (x.y» + Dzf(P' (p(X. This is to be expected. y)[r(x . along a unit vector v = (cos e(x.). .. y» is perpendicular to v. DI [ rl )2 (p(x. of the curve C2 v·.I )(r.
aJ ay sin e aJ cos e. then on U n V we have (I) PROOF It's the Chain Rule. (J 0 y . the Chain Rule would always b e written i n this way. where (x. are. Using the notation aJlax for D1 !.+ ar ax ar ay ar aJ aJ ax aJ ay = + ae ax ae ay ali ' I n classical notation.. y) denotes the identity coordinate system of �2 We have ax lar = COs e.. V) are coordinate systems on M. Note that aJ/ae is independent of which line is deleted from the plane in order to define the function e unambiguously. D. PROPOSITION. so it is going r(x. we can write the above equations as �= ae ar aJ = aJ ax aJ cos e + . and J: M � � is differentiable. y) everyvvhere (thus writing an equation about functions). so Qur formulas can be put in the form  aJ aJ ax aJ ay = . [x 0 yI )j (y(p» . etc.38 Chapter 2 goes around a circle of that radius as e goes from 0 to 27[.. and suppressing the argument (x. I t is a pleasure to report that henceforth this may always be done: 3. If (x. U) and (y. a ax ( rsm e) + / In particular. etc.y) times as fast as it should go in order to be used to compute the directional derivative of J in the direction w. these formulas also tell us what axlar etc. if you just keep your cool: ar (p) = D.))(y(p)) i > = j =l I: Dj (f 0 xI )([x 0 yI )(y(p» ) .I )(y(p» a I I = D ([J 0 [ ) 0 [x 0 y. of course.
•:. I soon forget I'm supposed to be summing. No tice that the summation in this formula occurs for the index j.I) (y(p» At this point we could introduce the "Einstein summation convention".(pl· . J= I n axj a here alay' is considered as an operator taking the function J to aJlay' . the operator f = alaxi lp satisfies f(Jg ) = J(p)f(g) + f(J)g(p ).:. For any differentiable J. often with hoards of indices being summed over. ay' P J= l ayi n I I: For later use we record a property of f = alaxilp: it is a "pointderivation". The operator taking J to aJlay'(p) is denoted by a . the n matrix . ax) ayi n · Di [Xj 0 y. U') are two coordinate systems on M. U) with p E U. one can avoid what Elie Cartan has called the "debauch of indices". If (x. and because by do ing things "right". I 4. PROPOSITION. which appears both "above" (in axj lay') and "below" (in aJlaxj ). U) and (x'. I won't use this notation because whenever I do. a)'i p ' I thus a ax j a = (p ) ax j P . and the convention is to omit the L sign completelydouble indices (which by luck. and any coordinate system (x. and felicitous choice of notation.DifJerentiable Structures 39 = j= = :L Dj (J 0 [ I)(x(p» 1 1 j= aJ ax} :L � ( p ) . Left to the reader. almost always occur above and below) being summed over. g : M � �. We will often write formula (I) in the form a ay i = I: ayi axj . the nature of things. n x PROOF. There are scads of for mulas in which this happens. .
however. � �. b). we will need some terminology. .40 Chapter 2 is just the Jacobian matrix of x' 0 inverse is clearly XI at x(p). If p is a critical point of f. 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). then x is a critical point of I if and only if j'(x) = O. where v(Bn ) is the volume of Bn . if q ¢ I(M) . in fact. � � has all points as critical values. It is nonsingular. Other points in N are regular values. V). . In this case.1l has "measure zero" if for every e > 0 there is a sequence B" B" B3. It is possible for all points of the inten. . which we merely state.� (P») clearly does not depend on the coordinate system (x. then D I = D. I(x. If I : �. its Now if I : M n � N m is Coo and (y. It is called the rank of I at p. may have all points as critical points without being constant. U) or (y. which in turn depends on a lemma from Calculus on Manif olds. If I : � � �. I = 0 everywhere. in particular. This is true. is still a "small" subset of �'. y) = x.a nonvalue of I is still a "regular value". We want to define the same concept for a subset of a manifold. of (closed Or open) rectangles with I and n=1 L v(Bn) < e. The most important theorem about critica1 points generalizes this fact. Recall that a set A C ffi. V) is a coordinate system around I(p). the rank of the m x 11 matrix (::. To do this we need a lemma. On the other hand. the image I(�') = � x to} C �. 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). although this can happen only if I is constant On [a. b) to be critical points. .a1 [a. it is called a regular point of f. the value I(p) is called a critical value of f. a function I: �. To state it. so I is again constant. if p is not a critical point of I. for example.
such that each set xi (A n Vi) c �" has measure 0. Vi) with A c Ui Vi and each set xi (A n Vi) of measure 0. Lemma 5 implies that there is some K such that IJ(x) . •:. = y 0 J o [l(x(A n Vi» .y l for all X. 6. Vi). PROOF. K on A for i. There is a sequence of charts (Xi . LEMMA. n2 K d. .yl for all x. then J(A) n V = Ui J(A n Vi) n v. (But if M is the disjoint union of uncountably many copies of �. LEMMA. . n 2 Klx . V ) is a chart on N. Each set y(J(A n Vi) n V ) has measure 0. Thus J takes rectangles of diameter d into sets of diameter ::. then x(A n V) C �n has measure ° for any coordinate system (x. If (y. y E A. j = 1 . n . Since J (Ui Vi ) is contained in the Ullion of at most countably many components of N.:0 n2 Klx . or has only countably many components. it is easy to see that if A C M has measure 0. If J : M � N is a C I function between two nmanifolds and A e M has measure 0. Then IJ(x) . If J : �n � �n is C1 and A C �n has measure 0. with A c Ui Vi. Lemma 6 thus implies another result: 7. V). (. . if this condition is satisfied and M is connected. PROOF. Conversely.Differentiable Structures 41 5 . it follows that J(A) has measure 0. COROLLARY. Using Lemma 6. A subset A of a Coo nmanifold M has measure zero if 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 ::. and A consists of one point from each component. . then J(A) has measure o. y E C.J(y ) 1 . We can assume that A is contained in a compact set C (since �n is a countable union of compact sets).J(y ) 1 ::. then A does not have measure 0). then J(A) C N has measure 0. Thus y(J(A) n V ) has measure 0. then it follows easily from Theorem 12 that A has measure 0. This clearly implies that J(A) has measure ° if A does . by Lemma 6.
in Problem 824). O) . in terms of the rank k of J at p E M. . since J may not even be oneone on ffi. . 9.42 Chapter 2 8.k + t (a). SC'T 1\. Although Theorem 8 will be very important later on. and then permuting the coordi nates in �m. . . . Topolog)' From the D{ff miiabtc Viewpoint or Sternberg. . . * For a proof. and its image could. then the critical values of J form a set of measure 0 in N. 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 . m (a» . These diffcomorphisms on �n and �m are clearly necessary. THEOREM (SARD 'S THEOREM). contain only points with first coordinate o. . U) and (y. . THEOREM. Remark: The special case M = � n . . XI k (a t . ak . part (I) says that by first performing a diffeomorphism on �n. and the case m > n is the trivial case (Problem 20). which gives only local results. 0). . ". V) around J(p) with y 0 J 0 x t in the form yoJ0  I\1oreover. ". a n ) = (a t .t (a t . V) such that y o J o x. .1i1IIor. But this case is just Theorem 3. then there is some coor dinate system (x. given any coordinate system )" the appropriate coordinate system on N can be obtained merely by permuting the component functions of y. . we can insure that J keeps the first k components of a point fixed. The stronger version ofSard 's Theorem . and M has at most countably many components. More exact information Can be given when J actually has rank k in a neighborhood of p. . It should be noted that J must have rank 2: k in some neighborhood of p. .k x to} c ffi. . and hence in a neighborhood of p. then there are coordinate systems (x. . If J : M � N is a C1 map between nmanifolds.In.14 of Calculus on Manif olds. . . . U) around p and some coordinate system (y. . . N = �m i s equivalent t o the general theo rem. which we will never use (except once . for example. Lectures 011 f'l" Diff erential Geomel1)'. for the present we are more interested in knowing what the image of J : M � N looks like lo�ally...n. a n ) = (a t . (I) If J: M n � N m has rank k at p. Theorem 8 is the easy case. because some k x k submatrix of (a(yi 0 J)/axi) has nonzero determinant at p . If y is the identity of �m . PROOF It clearly suffices to consider the case where M and N are �n . . O. (2) If J has rank k in a neighborhood of p.
. Since rank f = k in a neighborhood of p. 0 /» ) = ax} .an).n.(q) = a..an) means x(q) = (a'. . . the lower square in the matrix .. . � = I . . has the form in (I).. . 43 fORn) JRn. since may not be contained in any kdimensional subspace of PROOF.ak. (a'. (I) det C au Define xa = ya 0 f xT = uT a = I.. . k r = k + I. . .DifJerelltiabLe Structures In part (2) we must clearly allow more leeway in the choice of Y. (I) Choose some coordinate system u around p.. . . 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. . . { ya 0 f(q) = aar a = I.. __ ( a(v.an) = y 0 f(q) for q = x'(a'. . k .. .. By a permutation of the coordinate functions ui and yi we can arrange that (ya . . .. hence x.an) = (a'. so y o f 0 x. q = x'(a'. fl (p») . . . . (2) Choose coordinate systems x and v so that v 0 f 0 x . n.. . .. . . ). k hence ur (q) = a r = k + I .. .." o a. . x0 u (p). . .
. m: V) U) . vk). . . Thus we can write r = k + l. Moreover. .. 0 V I . . • . ak). . 1/rk+1 (a). . . .. .44 Chapter 2 must vanish in a neighborhood of p. .an) = y o vI o v o j oxl(al. 1/rk+1 (a) _ 1fk+1 (al.. • . . m .. . 0) . ...an) = y (al. 1/rm(a» = (al. Define ya = va yr = vr _ 1fr for Since 0 (V i .. v(q) = (bl. there is some coordinate system (x.. . . THEOREM. . . .ak.. . ... then for any coordinate system (y. .. bm) the Jacobian matrix y is a coordinate system in a neighborhood f(p yofox'(a'.. . 0. . . ..ak : Theorem 9 acquires a special form when the rank of f is n or 1 0. around f(p). . . ak. Nm has rank m at p. around p with has nonzero determinant. ak» by (3) = (a 1 . . .. . so of ). .. " " 1/rm(a) _ 1fm(al. (1) If m S n and f: Mn .
. only the need be permuted.O.. 0) . .. = = .. . fl (p») .an) (a'. . 0) 0 y. . a ..bm) '" (X(¢I (bI .. . m. it is easily checked that A ljf is now the desired •:.bn+'.O.bn).. A(bI ... f M '" ]Rn ]Rn.. 0 Then A 1/1 0 0 ..bn) '" ¢(a) '" A(b'. V) = . (2) Since the rank of at any point must be :s n. . . 0).bn. i t i s only necessary to observe that when k = In. .an) = A 1/1 f o¢'(b'. . . .. Even if we clo not perform ¢' first.. it is clearly unnecessary.. in the proof of this case. 45 (I) This is practically a special case o f (I) i n Theorem 9. .O) (a I . bn)). . . . (CPI (bI . .. . 1/1 f o¢'(a'... .an. .an) (a' . � = l . . . Define A by x f 0 <� f x = ". .. . . y = 0 y.O. there is a coordinate system around f(p) with y o f ox'(a' . .O.60 f p. . 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. C �)�(1/r:(7nIT{(f »:12 0 f(a'.Differentiable Structures (2) If n :s m and system (x. .bn) for (b I . . . C(Yaau. . .. ... 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. .O.. This can be conected by another map on ]Rm. U) around PROOF. . bm).an.. 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.. to permute the yi in order to arrange that det f: Mn Nm has rank at p.... .O).. then for any coordinate p.'" ... . ..an. . > II (y. .
with /.0) . A more illuminating example is the function h : IR > 1R2 defined by h(x) = ( g (x). as defined in Chapter I). On the other hand. the curve itself manages to be differentiable by slowing down to velocity 0 at the point (0. Although the second and third immersions f31 and fh are oneone.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 ). Three immersions of in 1R 2 are shown below. Another example is IR IR x>O x < x=O o. a differentiable map f need not be an immersion even if it is globally oneone. I g (x)l). although its image is the graph of a nondiff erentiable function.(0) = o. and it is clear from Theorem 10(2) that an immersion is locally oneone (so it is a topological immersion. One can easily define a similar curve whose image looks like the picture below. at all points of M. A differentiable function f : M n > N m is called an immersion if the rank of f is 11. it is case (2) which most interests us. Of course. it is necessary that 111 ?: 11. The simplest example is the function f : > defined by f(x) = x3. the dimension of the domain M. their images are not homeomorphic IR .
. . such that nV OJ. . r i. indicating the image of an immersion fJ. . > �!r! u.+. Of course. In general. second time around shows that may even be a dense subset of Despite these complications. The following picture. g(q) = g(q'). = yn(q) = this consequence Theorem = g : MisI an immediate(considered as aoffunction on 10(2). l.k. even if the oneone immersion f : P 7 is not a homeo morphism onto its image. where { yr(q')) = 0ya (q) a = kl. with Mr) in aThus. MI M. 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' = = . 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. M VI = {q E M : yk+' (q) = . Coo �+ f Coo p V C p 0 i Vn .DifferenliabLe SlTucluTes 47 to 1R. MJ Mn VI MI E M" (y. .. : > S X S M MI M. . there is certainly some metric and some differentiable structure on f(P) which makes the inclusion map i : f(P) > an immer sion. I I.n. MI M MI M IR M M). 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.
= yn (q) = OJ . on U.' 0 f is Coo ifit is continuous. . V) for M around f(p) such that U. and (y ' I U" . . c M is an immersed manifold. Let i : M. We want to show that j. with its Coo structure. PROPOSITION. yk l U. this cannot be done if g is one of the functions rf. Given P E P. > M he the inclusion map.' : fJ. If M. .. = {q E V : yk+' (q) is a neighborhood of f(p) in M. as a map into M. = . .48 Chapter 2 On the other hand. The following figure shows that f might not even be continuous IR. If M. . we may not be able to define g on M. choose a coordinate system (y. Actually.. 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 . even if g is Coo on all of M. f : P > M is a Coo function with f(P) e M" and f is continuous considered as a map into M] .(M) > One other complication arises with immersed submanifolds. this is the only thing that can go wrong: 1 1 . C M is an immersed submanifold. J is a coordinate system of M. For example. then f is also Coo considered as a map into Mj • PROOF.
yk are a coordinate system on VI) the function f is Coo considered as a map into MI . = Left to the reader. . ) C P is open. .:. defined as {x : Ixl 2 I}. Thus j takes some neighborhood of P E P into U.' (y) is a closed submanifold of M of dimension n .k (or is empty). if y is a regular value of j : M" > N m . . . Now it turns out that ever (connected) Coo manifold can be imbedded in some )RN . f : P 7 M which are homeomorphisms onto their image. Since all y j 0 j are Coo. so that )' manifolds can be pictured as subsets of Euclidean space (though this picture is not always the most useful one).I 0 49 f is continuous. Since U. In particular. but we first develop some of the machine. and y ' . C M is called simply a (COO) submanifold of M if the inclusion map i : M. We will prove this fact only for compact manifolds. so 0 j' i (open set) is an open set. .' ( U. If j : M" > N has constant rank k on a neighborhood of j' (y). i. It is to be hoped that however abstract the notion of Coo manifolds may appear.m)dimensional submanifold of M (or is empty). PROOF. . then j . and gives the sphere S" . as a special case. An immersed subman ifold M.y which would be used in .' (y) is an (II .' 12. is open in M" this means that j. There is one way of getting submanifolds which is very important. Such an immersion f\1ost of these difficulties disappear when we consider oneone immersions is called an imbedding ("embedding" for the English). > M is an imbedding. •:. \� � • •• % submanifold . lS also a closed subset of M. . PROPOSITION. it is called a closed submanifold of M if M. then j .' C IR" . submanifolds of )R N will seem like fairly concrete objects.{OJ c )R".DifJerentiabLe Structures By assumption.
. compact and U.l . has an open neighborhood V. with U. . C Uj+1 .. <. Unfortunately. 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. . there are many definitions and theorems involved.Ui . . C3 . 1 3. we can choose all members of (9' to be open sets diffeomorphic to IRn. • • • with M C. we obtain open sets Vi. By Theorem 1. then there is a n open cover (9' of M which is locally finite and which refines (9. since we will need it later on anyway. Now M is the union for i > I of the "annular" regions Ai Vi .2. Then V. whose union contains all Ci . we can obviously cover Ai by a finite number of open sets. PROOF We can obviously assume that M is connected. Moreover. with compact closure. U C2 U C3 U . and each contained in Vi = Ui+ 1 . Continuing in this way.2 . each contained in some member of (9. If (9 is a cover of a space M. j = IRn. Since each Ai is compact. U C2 has an open = neighborhood U2 with compact closure. Clearly C. THEOREM. and hence is M. there are compact sets C" C2 . since a point in Vi is not in V for j 2: 2 + i. Let V_. We can also choose these open sets to be diffeomorphic to In this way we obtain a cover (9' v.Ui.50 Chapter 2 the general case. A cover (9 is called locally finite if every p E M has a neighborhood W which intersects only finitely many sets in (9 .lhich refines (9 and which is locally finite. I f (9 i s a n open cover o f a manifold M . = Vo = 0.
. }. THEOREM. . C] V{ UJV].(U2 U V3 U ... 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). C C 14. = 1 for all P E M (this sum is really a finite sum in some . . Un+i U�+j P E by cannot possibly eliminate p. V{ ...V2. I ] . V3 . . THEOREM (THE SHRINKING LEMMA). . U V3 U n V{ U V2 U·· . ' . one for each in (9. (2) "L <Pu (p) u neighborhood of p. <. Let (9 be an open locally finite cover of a manifold M.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. Now P E it follows that since replacing C2 V2 V2V2.V]. then intersects only finitely many members of (9. V' V' V PROOF. V2. Then there is a collection of functions <pu : M . For any p E M there is a largest n with p E because (9 is locally finite. open set with c c c Continue in this way. Let be an . Let (9 Then is a closed set contained in and M = open set with C C C Now C] V] . We can clearly assume that M is connected. 15. Then it is possible to choose. C] U V2 U V3 U = ) = V'V {V]. U V� U (Vn+] U V. Let W be an is a closed set contained in and M = i. . Let (9 be an open locally finite cover of a manifold M. V{ UVC2. V{UV2U " ' . . [0. an open set with is also an C in such a way that the collection of all open cover of M. such that Coo V (1) support <Pu C V for each V. ..+2 U··· ). by (I)). for each in (9.
If (9 is any open cover of a manifold M. : M > [0. (3) for each i there is a U E (9 such that support ¢. Choose the U' as in Theorem function 1/Iu : M > [0.52 PROOF. General case. where since it is a finite sum in a neighborhood of each point. Since (9' = { U E (9 : U C Up for some pl. 1 6. Cover Mwith open sets Va having compact closure and contained in M The open cover {Up. 1 ] 14. C. Ue(') L 1/Iu Case 2. Chapter 2 Case 1. clearly Coo Define Ue(') L 1/Iu > 0 ¢U = everywhere. (A collection {¢. : This sum is Coo. �. = Coo (2) L ¢. If M" is a compact cling ! : M > for some N. VeC}' L ¢U. C U. Let C C C C C. I]} satisfying (1) and (2) is called a partition of u nity. if it satisfies (3).(p) 1 for all p E M. it is called subordinate to (9. • . and ¢u(p) = 0 when U C Va . provided that Lemma 2 is true for c U e M with closed (but not necessarily compact) and U open. Apply Lemma 2 to U' c U e M to obtain a which i s I o n U ' and has support C U. THEOREM. it is easy to see that support f e u. != L u ¢u(P) = I for all p. 1 7 .(p) # O} is locally finite. I] such that (I) the collection of sets {p o ¢. Each U in (9 has compact closure. Since the U ' cover M .) It is now fairly easy to prove the last theorem of this chapter. V } has an open locally finite refinement (9 to which Case 1 a applies. This case can be proved in the same way. But this is a consequence of Case 1: For each p E choose an open set Up C U with compact closure. jRN Coo manifold. : M > [0. then there is an imbed . then there is a collection of functions ¢. clearly ! (p) = I for all p E Using the fact that (9 is locally finite. COROLLARY.
(Xk. . This shows that we must have q E Vi . by JRN . in fact. 1] which are 1 on Vj and have support C Vi. 211 + I . because any poiIlt p is in V! for some i.. and the composition of Coo functions is Coo. (a) All Coo functions are continuous. (b) In the proof of Lemma I. as claimed. jRn 3.1. 2. Problem 333 shows that. and on Vi.Ai are COOrelated. PROBLEMS 1. that is not the question being asked. the N x II Jacobian matrix It is also oneone. . N is Coo if and only if g 0 f is Coo for every Coo function g : N . .. Then Vri(p) = I. (b) Show similarly that aM is welldefined for a Coo manifoldwithboundary M. Define f : M . so p = q) since Xi is oneone on Vi . and functions Vri : M . . Vri ' Xi (P) = Vri ' Xi (q) . Moreover. where N = lIk + k. show that the 11 at each point is unique wi/houl using Illvariance of Domain. (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.. For suppose that f(p) = f(q).. There are a finite number of coordinate systems (Xl . so also Vri(q) = I . show that all charts in .. Vl ). [0.) . where VIi = I.Di/Ji:renliabLe S'lruclures 53 PROOF. 4. Vkl with M = Vl U · · · U Vk ..... we can always choose N = ar ( axf ) contains the 11 x 11 matrix (ax�) axf ' . •:. (b) A function f : M . (a) Show that being COOrelated is not an equivalence relation. How many distinct Coo structures are there on 1R? (There is only one up to diffeomorphism... There is some i such that p E Vj.. This is an immersion. Choose V: as in Theorem 14. JR.
. Check that the two projections PI and P2 on homeomorphisms Ji and g. ( PI . M is Coo.. 1 0 .. (d) If Ji : N . /1 x 12 : Nl x N2 . IRn1 1 2. is locally Coo. h) .....4. P2) of Ml . h) : N .. M. M with c (O) = and c ( I ) = q . / can be extended to a Coo function on a neighborhood of (Not so obvious.. 2).. M is Coo if and . p snl p are Coo related to the 2n p2 p sn snIP'n IRn lHIn IR IR lHIn.... Compare the rank of / and the rank of / 0 g . .' can also be described as the set of all (x I V n N.. be the map .. with U C N. (iit .. Show that / : only if / 0 g : .4... V) in A (c) Show that the inclusion i : N . 1 ] ... [pl. show that ... (b) Show that ... M2 are Coo. Let g : . are Coo (i = 1 .4. x M2) X M3 is diffeomorphic to MI x (M2 x M3) and that MI x M2 is diffeomorphic to M2 x MI ....) (b) If / : . 9.. If / : 1HI" . Moreover... h E M2... can one determine the rank of (fJ . then / is Coo... defined by Ji x h(Pl . is maximal for M... and that . (a) If M is a connected Coo manifold and .. (a) If N C M is open and .. V n N) for (x. has two extensions g. U) in .4.... Ml x M2 at in terms of the ranks of Ji at For Ji : N.. p? (pt}...e. then there is a Coo curve c: [0.. (a) If U C p IP'n i s open and / : U .. Ml x M2 differentiable for all fit E Ml ..' is the unique atlas with this property. (Obvious. If M is a Coo manifoldwith boundary.. q E M.54 Chapter 2 5. MI and 1f2 0 / : N .... Ml x M2 is Coo if and only if the compo sitions 1fl 0 / : N . .. Ml x M2.. then / is Coo. (b) The diffcrentiable structure on Ml x M2 makes the "slice" maps 8. M. 6.Ai consists of all (x. a map / : N . the Coo structure we have defined for Ml x M2 is the only one with this property.. M2 . (c) More generally. i s locally C oo (every point has a neighborhood on which / is COO). PI ... M is Coo. show thaI h( P2) ). 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).. P2 ) = (Ji is Coo and determine its rank in terms ofthe ranks of ji. (b) It is even possible to choose c to be oneone.' is maximal for N if .. 1 1 .. then there is a unique Coo structure on aM such that the inclusion map i : aM ...4..) IR lHIn.... M is an imbedding.... i. (a) Show that (MI 7.
then S boundary but U is not a 2dimensional manifoldwithboundary. The closure S of the final resulting figure is known as Alexander� Homed S plwe. (a) Let C be an open set such that boundary is an (n .I)dimen sional (differentiable) submanifold. then U is a manifoldwithboundary. < . O) < I or I < d(x. V = V. (It is well to bear in mind the following example: = {x E ]R" : d(x. O) (2) f(x. 0) for all x. (a) There is a map f : ]R2 . 14. ]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 .S. Show that U is an IIdimensional manifold withboundary.) V if V smaller copies of the two parts of S' into the regions indicated by arrows.. This figure may be extended by putting V Mn V.y) C ]R 2 _ 1HJ 2 for y = (x. y) c 1HJ 2 for y :0: 0. so part (a) is true only for differentiable submanifolds. and then repeating this construction indefinitely. (3) f(x...O) < 2). but au # boundary (b) Consider the figure shown below. 0.Differentiable Structures 55 13.
such that = for alI E and a similar diffeomorphism V x [0. .. (x. x x f define a Coo structure on P. fJ). Show that the resulting Coo manifolds are diffeomorphic. Show that this procedure does not f M N M UJ N f: aM . f(V aM) aN.I . . but that the diff eomorphism ca"not be chosen arbitrarily close to the identity map. I).O) Mp aM. I). aMfJ: . ..V U fllVJR" aM xi VV aM. aN M N p aM x aM f(x) aN. V U aM M a N fJ (a. aN aM a(p) V (p. . Show that there is a unique Coo structure a: V .56 (4) Chapter 2 restricted to the upper halfplane or the lower halfplane is Coo. on P such that the inclusions of and are Coo and such that the map from V to x ( . (b) Suppose and are Coo manifoldswithboundary and is a diff eomorphism. (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. but itself is not Coo. nn2 JR ann22. (We will be able to prove later that such diffeomorphisms always exist). 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. (c) Now suppose that there is a neighborhood of in and a diffeo morphism x [0. V) f(p). I) induced by and is a diff eomorphism. 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..
Consider the coordinate system (yl. axj ' axj axk ' J. . once you 11 (0) and know the trick. 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. I] be a function with ° on Vi and Functions like J.b) = a b.1 /x' is used just to get a function which is and e. we can even find with = 11 (1).C. I] such that C 11 (0) and C 11 (I). of order 1 . . C = and 11 (0).: M will be called mixed partials of ai = Show that sup of all mixed partials of Ji . . Can the inclusion be an imbedding? Are the projections on each factor maps? (b) If and are manifoldswithboundary.  a function with 11 (0). 2. . Actually. . . for any closed C (b) Let {V. I: JR JR I(x) = { �1/x x > 0o x > CooCoo. construct a structure on x sucl1 that all the "slice maps" (defined in Problem 8) are 16. defined by :s (a) It suffices to find. The proof turns out to be quite easy. y2) yl(a.} be a countable cover of V. where each Vi is of the form < E I}  onto ° on Let Vi .I /1xl could be used just as well). 2 Coo > x C1 I: M ° for < 0. Ji of all orders :::: i. > aJ. C M M. Coo I C = ({a JR": lal ) x MJ. I = �A I L a2i i=l 18.. . = X . (a) Find a structure on 1HI1 x 1HI1 which makes the inclusion into JR2 a map. I for some coordinate system taking an open subset of > [0. Cl = C1 C2 CM.I 17. + for JR2 defined by . =C M JR".b) = a y2 (a. a2J. Coo J. Let is Coo. .Differentiable Structures 57 15. Show that the function Coo Coo Coo M NM N Coo is (the formula e.
a2/ax2 f: Nm Cl aI. If M n . 1] .. f: [0. and not oneone.k for all 2""2'1I � I :S 2k2n · Show that f is continuous.. write each 1.: (rf. I . . in terms yl 19. the operator alai depends in terms of polar coordinates. 1 ]..l+ l . A n . Answer: then compute 20. A n •41+3) An •41+4 are contained in A n l . then f(M) has measure that M has only countably many components).2. (First compute from this). is and m > II. Notice that # on and i. 1] [0. (c) Squares A n. and 0 (provided [0 2 1 . 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 . /ayj .. x into 22n squares.k is labeled simply k. square A n . 1] (d) Squares An •41+ 1 . 2"'. (b) The upper left square is A n• 22n. a subdivision of ..b) yI I y af/ayl af/aI /. onto [0. for II = and 3. k +l have a common side. 1] by the condition kf(l) An. [0. (a) The lower left square is A n . Compute the "Laplacian" = 1 1 . . I .1] 1.:) + fe (Hii ) ]' a/ax in terms of alar alae. k and An . The numbering is determined by the following conditions: [0 . I] [0. .58 Chapter 2 (a) Compute from the definition. (b) Also compute it from Proposition 3 (to find of and y 2) . not just on even though af/ayl (a. �[f. The following pictures show. A n •41 +2 . .. A n .
. You will need . P) = The curve is rectifiable if {i(c.C(li. (b) The subset M. V) on M such that M. Show that g is oneone. n V = {p : x k+ ' (p) = . define > f is uniformly continuous. = xn (p) = 0). there is a coordinate system (x. (a) If V C Coo. p E V) is a submanifold of ]Rn.) . . dinates. 1] [0. (b) Every submanifold of ]Rn is locally of tllis form. and that its image will not have ]Rn be continuous. . a semicircle with diameter the line segment AB. a set M. "Ld(C(li). below the image of g. . Ixl) . For p/2n E [0. P » ) is bounded above (with length equal to sup{i(c . x E ]R) is not the image of any immersion of ]R into ]R2 ]Rk is open and f. A = f(O)�"" g . 1 ] '"' B = f(l) (a) Show that > measure 0 if the shaded triangles are chosen correctly. can be made into a closed submanif if and only such old coordinate systems exist around every point of M. Show that the image of a rectifiable curve has measure O. Id of i(c. f(p» E ]R" . (Neither Theorem 9 nor l O is quite strong enough. P» )). What does the inside of this curve look like? 23. .Differentiable Structures 22. [0. C M can be made into a kdimen sional submanifold of M if and only if around each point in M. define f(p/2n) E 59 ]R2 as shown below. Let c . Coo if {(x. (b) Consider the homeomorphic image of S ' obtained by adding. 1]. The set 26. i=1 k C 24. . 1]. after renumbering coor 25. For each partition P "" {Io. [0. V > ]Rnk is then the graph of f = { (p. (a) If M is a manifold. so that it has a continuous extension ]R2.
X 3 . (b) Prove the Shrinking Lemma when the cover (9 is pointfinite and countable (notice that localfiniteness is not really used). 30.. 41). has constant rank k on a neigh borhood of I'(Y).60 Chapter 2 the implicit function theorem (Calculus on Manifolds. (f) If M is a manifold. comparison with the implicit function theorem will show how some information has been allowed to escape. is an immersion. 29. 3 1 7321. 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 . There is a way of immersing IPz in IR 3 .. A continuous function I: . whose image is the Steiner surface. pp.. pg. I] which is not locally finite but which is "point finite": every point of [0. (b) I( ) c Y is closed if and only if L(f) C (c) There is a continuous I: IR . (e) A submanifold M... but L(f) # 0..) 27. (d) A oneone continuous function I: . (c) Prove the Shrinking Lemma when (9 is a (not necessarily countable) point finite cover of any space. M is proper.. (a) An immersion from one nmanifold to another is an open map (the image of an open set is open). Show that g fails to be an immersion at 6 points (the image points are the points at distance ± 1 /2 on each axis). Prove Proposition 12: If I : Mn . xy) defined in Chapter I. consider collections e . Let I : IPz ... I] is in only finitely many members of the cover. there is a proper map I : M .k (or is empty).. and I : M . See Hilbert and CohnVossen.. N (a) L(f) = 0 if and only if I is proper... y. 1R 3 be the map g([x . Coo Coo 31. then I is onto.... Theorem l O is es sentially Theorem 213 of Calculus on Manifolds... Z]) = (yz.. Geometry and the Imagination. 1R.. . N N N 28. • • • E with no convergent z subsequence. known as Boy's Surface. then II (y) is a (closed) submanifold of M of dimension n . C X (C) X X X f(X).. Y is a homeomorphism (onto its image) if and only if L(f) n I(y) = 0. IRz with I(IR) closed. C M is a closed submanifold if and only if the inclusion map i : M.. (a) Find a cover of [0. (b) If M and are nmanifolds with M compact and connected. (You will need Zorn's Lemma. Y is proper if II is compact for every compact C Y. the function I can be made if M is a manif old.xz...
= (0. A" where At is the transpose of A. (a) The set of all nonsingular n x /1 matrices with real entries is called GL(/1. 1R) > (symmetric matrices) by 1/I(A) = A . 33. = = and f = identity. It is a manifold. show that for A E O(n ) the matrix 32. and the union of all U' for (U . U' C U. (a) If M. is any neighborhood. 1R). (b) This is f alse if M = IR and M. in this case. or unimodular group. (e) Using the formula 1/Iij (A) = L. in fact. f can always be extended to a function in a neighborhood of M. By applying the chain rule. The special linear group SL(n .I. The subgroup 1/1'(1) of GL(II. IR). and 1/Iij the n(l1+ 1)/2 component functions of 1/1. pg. S I . Show that RA is a diffeomorphism. . U') E e.) Coo. 1R) by RA(B) = BA . (d) For any A E GL(/1. Coo ( a1/lU (A) ax kl ) has the same rank as ( a1/lij I axkl ( ) ). since it is an open subset of IRn'. IR). U ') where U E 19 . > IR is then there is a function J: M > IR with J = f on M" and with support J c U. However. 24. Coo N. I). Show that A E O(n ) ifand only if the rows [or columns] of A are orthonormal. 1R) > GL(n . together with all other U E 19 covers the space. C M is a closed submanifold.IR) is called the orthogonal group 0(11). show that S L(n .) Conclude from Proposition 12 that O(n) is a submanifold of GL(n. 1R) is a closed submanifold of GL(n . and use partitions of unity). (extend locally. aik ajk k (A = (a/j »).Differentiable Structures 61 Remark: It is also false if M = 1R2 . (c) Show that O(n) is compact. U :J M. 1R) of dimension n 2 . and f : M. is the subgroup of all matrices with det = I. 1R). Coo N S'. and that 1/1 0 RA = 1/1 for all A E O(n). Using the formula for D(det) in Calculus on Manifolds . (c) This is false if IR is replaced by a disconnected manifold of pairs (U. M. (Here xkl are the coordinate functions in IRn 2 . define RA : GL(II. f has no continuous extension to a map from ]R2 to but the proof requires some topology. (b) The symmetric n x n matrices may be thought of as IRn (n + ' ) /2 Define 1/1 : GL(n . the general linear group.
1R) is called the special orthogonal group SO(n). Hint: If h denotes the k x k identity matrix. (f) Show that det A = ±I for all A E O(n). n . n) denote the set of all m of all m x n matrices of rank k . then X has rank k if and only i f D = CAJ B. and M(m . n matrices. 34.Ao are < e.1)/2. x 1 aj l au k=i#J k=j#i k =i=j 2aj/ o otherwise. then ( � �) (d) M(m . 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 . The group O(n)n SL(n. (c) If PXQ = ) ' where Ao is k x k and nonsingular. k) the set (a) For every Xo E M(m . k) C M(m.n) is a submanifold of dimension k(m + n .) Conclude that O(n) has dimension n (n . n.k) for all k . or the rotation group R(n). .62 Chapter 2 show that a 1fr ij . Let M(m .Ao are < e .:s m. where the entries of A . l1.(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). n .
Then c'(t) = (cl'(t). . 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). .CHAPTER 3 THE TANGENT BUNDLE point v E JR" is frequently pictured as an arrow from 0 to v. crt) + c'(t ) 63 . 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. c"'(t» is just a point of JR". suppose c: JR > JR" is a differentiable curve. . but the line between crt) and c(t) + c'(t) is tangent to the curve. .
it can be pictured more geometJ·ically as a particular subset of 1R" x 1R". the point Jr ( v) is "where it's at". which we will also denote by TIR". v) for v E 1R" by 1R"p . Alternately. 1R" by ]( a . like v. To recover the first member of a pair v E TIR". v). the one visualizable case occurring when 11 = 1.64 Chapter 3 To give this picture mathematical substance. in conformity with this notation.. This picture gives rise to some TJRn JRn _ _ ]( '_ terminologywe call ](1 (p) the fibre over l  p p. it is more convenient to denote a member of TIR" by a single Jetter. b) = a.. elements of TIR" are called "tangent vectors" of 1R". we define the "projection" map 1r: 1R" x JR" . The set of all such pairs is just 1R" x 1R". For any tangent vector v. we simply describe the "arrow" from p to p + V by the pair (p. We will often denote (p.. v) E TIR" by vp ("the vector V at p").1 (p) may be pictured as all arrows starting at p. the "tangent space of 1R"". This fibre can be made into a . The set ]( . we will denote the set of all (p. At times.
built into it.' (p) 1( . but. Thus. however. Usually we will just use ordinary + and · instead of If f : JR" > JRm is a differentiable map. . This is not the only reason for defining f in this particular way. v + w) a o (p. it can be written . = f 0 1(.The Tangent Bundle 65 vector space in an obvious way: we define (The operations U 1( .' (p). . v) = (p. has the map f. so that. as well as all maps Df(p).a · v). (I) By our definition. D (g 0 f)(p) = Dg(f(p» 0 Df(p). ( [ Df(p)(v)lJ(p) } = (Dg(f(p)}(Df(p)(v» }g(f(p))' This looks horribly complicated.p union of all f.p p p ing diagram "commutes" (the two possible compositions from TlRn to jRm are equal). g. 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. Suppose that g : JRm > JRk is another differentiable function. W) = (p. and 0 . and p E JRn. v) (j) (p. Since f (v) is defined to be a vector E JRmJ( ) . denotes the map f : TJRn > TJRm which is the . the symbol f. peJRfI x (j) and 0 (p. whose apparently anomalous features will soon b e justified. x respectively. is de noted by f . 1( 0 f. should really b e thought o fa s defined on and JR TJRn. using (I). the follow . . by the chain rule. J.
c. c"'(I» c(I) .. JR" can be defined in terms of this concept.(vp) were not in JRmJ(p) . ..66 Chapter 3 thus we have This relation would clearly fall apart completely if I. The tangent vector of c at 1 may be defined as g. this vector lies along the tangent line to the graph of 1 at (1. . Henceforth.. then g 0 c is a curve in JRm. it is merely an elegant restatement of the chain rule. If g: JR" . J ) is the "unit" tangent vector of 1R at t . /(1 » . c'(I)c(I) E JR"C(I) ' [If c happens t o be of the form c(l ) = (1. I.. JR : 1'(1) then C'(I)C(I) = (I ... /'(I» c(I). The "tangent vector" of a curve c : JR .] Notice that the tangent vector of c at 1 is the same as where I t = (t . rather than Df. like rank or singularity.) = [DC(I)(I)]c(l) = (c" (I). The tangent . also. . in terms of I. . 0 I..... with our present definition of I.(I. 1(1 )) for I: JR . JRm is differentiable. we will state almost all concepts about Jacobian matrices.. = (g 0 fl •.
I ) ..) .I ) � �� � �. 0 (x 0 y . Suppose we take a coordinate system (x.I ) . (i 0 x .I is a map from IRn to IR N with rank n. (x 0 y.. n n 'y(p) : IR y(p) .. (tangent vector of c at I) .The Tangent Bundle 67 vector of g 0 C at t is (g 0 C). I = (i 0 x . Consider now an ndimensional manifold M and an imbedding i : M .. Then i 0 X .I ) .I o x o y.(I..(I. IR x( p) . This subspace doesn't depend on the coordinate system x) for if y is another coordinate system.» v crt) = g. IR N . Consequently. and is an isomorphism (with inverse (y o x . U ) around p.X(p) ..) = g.(p).(c. (lRnx(p) is an ndimen sional subspace of IR N . = (i o x . then (i o y I ).I ).
Thus.i) > M by As in the case of TlRn .i)p. and every diff erentiable curve in i(M) is of this form (Proof?).J ) * (lRnX(p) ) ' Moreover. We can define a "projection" map n : T(M. Then (S ' . which justifies the picture we have drawn.68 Chapter 3 There is another way to see this. each "fibre" 7f .I (p) has a vector space structure also. Now so the tangent vector of every a is in (i o x. cose)p (it clearly doesn't matter which of the infinitely many possible e's we choose). 0 x . cos e) # O. Beyond this we have to look a little more carefully at some specific examples. If c : (e. our ndimensional subspace is just the set of all tangent vectors at i(p) to differentiable curves in i(M). and c'(e) = (sine. i)p C i(M) x jR N C TjR N .sin e.(M). Consider first the manifold M = S ' and the inclusion i : S ' > jR 2 .sine) passes through every point of S ' .I 0 c is a curve in jR N which lies in . 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. then a = . i)p con n(v) = p if V E (M. i)p' We now want to look at the (disjoint) union T(M.i) = p EM U (M. The Curve c(e) = (cos e. We will denote this ndimensional subspace by (M. . sin e) E S ' . let up = (. e) > jRn is a curve with c(O) = x(p). For each p = (cos e.
. which makes the following dia· [rr'(a.b) a] = T(S2. i) fi . for a fixed vector v #. one at each point of which varied continuously. : i) gram commute. 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. SS22 2 ]R2. We Can therefore define a homeomorphism I. restricted to a fibre is a linear isomorphism on to the image. S' x ( ) T(S ' . . ft rr' rr'' 1R ' by fi ).The Tangent Bundle 69 sists of all multiples of the vector up . then. S2.S ' x IR' � S. then each fibre has a vector space structure in a natural way. Commutativity of the diagram means that fi takes fibres into fibres. would be a collection of nonzero tangent vectors.. . It is a wellknown (hard) theorem of topology that this is impossible (you can't comb the hair on a sphere)..). ). . S2 fi. i) .O in the set of vectors T(S'. up = (p. clearly I. If we define the "fibres" of to be the sets (p).
0)(0. This function would have to be nonzero everywhere and also satisfy A(21r) == A(O). for some continuous function A : [0. 0) of M. then each vector would be J (2.0. 1 ( 0.0. 2 sin8. %2. which it can't (by an easy theorem of topology).70 Chapter 3 There is another example where we can prove that no appropriate homeo morphism T(M. 0) (2.O)J [ aal l [ DI(O. 1 )(8.0) . 1 )(2rr. 2 sin 8. the vector (8. = while = [� (O. The map f will just be the inclusion M > 1R 3 where M is a Mobius strip.0) = ( . 1 sin n . I) > 1R 3 defined by 1(8. i) > M X 1R 2 exists.0) ). 1 ) (0. 21r] x (I. 0)(2.. 2 sin 8 + 1 cos � sin 8. 0): Ifwe could.0) = ( 1 .0. 21r] > 1R.  .1) = (2 cos8 + 1 cos � cos8. 0) This means that we can never pick nonzero dashed vectors continuouslY on the set of all points (2cos 8 .0) ) = * (21r. 21r I r�� is a tangent vector. 0 .0) ) . The same is true for all multiples of 1 ( 0. At each point p = (2cos 8.0.0) (2. to be precise. 0 . shown * as dashed arrows in the picture. Notice that 1 ( 0.1 . fibre by fibre. 0) i .0.M is the image of the map I: [0.0) . f). The impossibility of choosing nonzero dashed vectors continuously clearly shows that there is no way to map T(M. without appealing to a hard theorem of topology. the particular subset of IR 3 defined in Chapter 1.
Ell . which make each fibre n:I (p) into an ndimensional vector space over lR. fibre by fibre. In fact. T(M.. V x JR" which is a vector space isomorphism from each n:I (q) onto q x JR". such that the following "local triviality" condition is satisfied: For each p E B. (3) Ell and 0 are maps Ell : p EB U n:I (p) x n:I (p) . there is a neighborhood V of p and a homeomor phism I : n:I (V) . n:. In standard jargon. B .. the part of T(M.. B is a continuous map onto B. 0) . 0: JR x E . JRN. however. we can therefore define by f: f (mp(vx(p) ) n:I (V) . This additional feature qualifies T(M. (2) n: : E . For any imbedding i: M .. E. . � = (E.. i) over V. then n: . for each p E V. i) does not "look like" a product M x JR". i) is always simple locally: if (x... 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). V) is a coordinate system on M.... V = x JR" (p. the structure of T(M.....The Tangent Bundle 71 homeomorphically onto M x JR2• We thus have another case where T(M.i) is "locally trivial". can always be mapped. with Ell (n:I (p) x n:I (p») c n:I (p) and 0 (JR x n:I (p») c n:I (p). the fibre where the abbreviation mp has been introduced temporarily... homeomorphically onto V x JR". for all q E U... v) .I (V). E.
or even denote the bundle by E alone. each bundle � = (E.a) with ( I .n(B) is called trivial. for M can be obtained !i·om [0.n(x). i) are not trivial. . Equivalence is here a technical term: Two vector bundles �. W E . : X x 1R" > X the projection on the first factor. : E > B. a).'(S'). 0) automatically gives rise to a bundle �IA over any subset A C B.n(IRn). A bundle equivalent to . .) The bundles T(S z . w) and 0(a. . and the obvious vector space structure on each fibre. and a . The Mobius strip itself (not T(M.' (p) and a E 1R. > B and �z = "z : Ez > B are equivalent (�. For vectors v. I] x IR by identifying (O. v or av. The map h is called an equivalence. = " . This is called the trivial Ilplane bundle over X and will be denoted by . The bundle T(S ' .. > Ez which takes each fibre ".. (The local triviality condition for a bundle � just says that �IU is trivial for some neighborhood U of p. The "tangent bundle" TlRn is just . v) by v + w. :::: �z) if there is a homeomorphism h : E.. we will denote (j)(v. Notation as cumbersome as all this invites abuse. to be precise. : E. i) considered before is equivalent to . The simplest example of an Ilplane bundle is juSt X x IRn with ..72 Chapter 3 Because this local triviality condition really is a local condition."nJ1sidered as a Idimensional vector bundle over S l . (j). and we shall usually refer simply to a bundle .. . B.' (p) isomorphically onto "z' (p). but there is an even simpler example of a nontrivial bundle. while S' can be obtained from {O. respectively. i) and T(M. I } . i») can be .
I] > with S(O) = S(I).The Tangent Bundle 73 [0. > 1( . This surely shows that M is not a trivial bundle. EI BI � Bz . I] by identifying ° with I . and the map f satisfies f(M) C N. a)}) = {O. depending on whether one is consider ing all bundles at once.a) = I for 0 < I < I and 1( {(0. The relations between some of these cases are considered in the problems. The diagram above illustrates local triviality near the point {O. with j : E. (I. 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). the section s must be ° somewhere (that is. > Ez and f : B. > Bz. ' (p) 1 � Ez l1(Z to for any differentiable The pair U . fixed bundles over various spaces. (2) j : 1(. then f . s(8) E 1(1 (8) must be the ° vector for some 8 E S').I U(p)) is a linear map. I }. a). fl. or a fixed base space with varying bundles. Such a map corresponds to a continuous function s : [0. fl is a bundle map fi'om . such that (I) the following diagram commutes jR 1(. i : 'M > and Nm C and j : N > are the inclusions. An equivalence is obviously the analogue of an isomorphism. Thus f may be restricted to be an isomorphism on fibres and f to be the identity. to �z is a pair of continuous maps (j.' A bundle map from �. > If M" c are submanifolds. Z jRk k l jRl TjR TjR jRk * There are actually several possible choices. Since s must be ° somewhere. or a homeomorphism. . the map 1( is defined by 1( I. The analogue of a homomorphism is the following. f: jRk jRjRll .
N.i)p k MIRIRI. it would have sufficed to begin with a Coo function f : > since f can be extended to locally.I : i(M) N.i) TM.i) T(M. this construction could be generalized much further. M. I I We can therefore put all these maps together. i) TMi .j). Elements of are of the form 0 ).j)IV. (w) T(M. IRk IRk i j M i. and obtain an equivalence from to In other words.74 Chapter 3 takes to to see this. we could abbreviate to if we agreed that really denotes an equivalence class of bundles. T(N.i) T(N. (v) E map from to Actually.. to the case where and are merely imbeddings of two abstract manifolds and and f : > is Coo.j)p x. i) T(N.i)p T(M. rather than one bundle. j). That is the T(M. the dependence of on is al most illusory. j). 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. T(M. The map > induced on fibres is independent of the coordinate system for if (y.i)I V T(M. V) is another coordinate system.i(N) M N i T(M. j)p are of the form (j 0 x 1). i)p j (i xI M for WE WE IRnx( IRnx(p)' p). T(M. M N.j). then (M. (M. (w) for we obtain a bundle map from to which is obviously an equivalence. T(M. In fact. In this way we obtain a bundle f o e in and consequently f. while and are two imbeddings of in and respectively. just remember that v E is the tangent vector of a curve c in so f (v) is the tangent vector of the curve . while elements of If we map T(M.
(2) There are equivalences In : TJRn > en (JRn) such that for every function f : JRn > JRm the following commutes. N) where f : M > N. though quite natural. there is an equivalence TU :::: (TM)IU such that the following diagrams commute. Can we do this? Yes.. so in stating our result . : TU > TN is just the restriction of f•. The identity map I of U to itself and the inclusion map i : U + M have to be considered as different. then 1. precisely we will write I'old J/' when necessary. (I) If 1 : M Coo TJRn � TJRm 1m In en (JRn) old f. It is possible to assign to each nmanifold M an nplane bun dle TM over M. ea . More precisely.' TU � TM)IU/c ( _ _ _ '" i.. we can. What we would like to do is to get a single bundle for each M. and to each map f : M > N a bundle map (f. TJRn will be different from our old defini tion (namely. THEOREM. We will obtain a single bundle for TM. en (JRn»). but the elements of TM will each be large equivalence classes. and it is undoubtedly ugly. in some natural way. .. i) has.. f). and for f: M > N the map (flU). _ _ _ . : TU > TM are certainly different (they map TU into two different sets). I . such that: Coo > M is the identity. �TM/f.. sub terfuge. . where i : U > M is the inclusion. then (g 0 f) . it must be understood that the symbol "/" r lly refers to a triple (f. M. 0 f . and so will f for f : JRn > JRm. = g.The T angent Bundle 75 sort of thing an algebraist might do. TN PROOF The construction of TM is an ingenious. If • g : N > P. : TM > TM is the identity. then TU is equivalent to (TM)IU.. . * vVhen using the notation h. When we do. which has all the properties any one of these particular bundles T(M. TM TU (flU). since the maps 1 * : TV + TV and i. em (JRm) 1 1 (3) If U c M is an open submanifold.
v]p. : TM > M. namely [x. is supposed to be the old (y 0 XI).I (p) isomorphicaIly onto {p) x If y is another coordinate system. and (y 0 xI). We define a vector space structure on . w]p = [x. and v.. If JRn. These equivalence classes will be called tangent vectors at p. There is a metric with exactly these sets as open sets. v). v + w]p a · [x. Then (x. in conformity with the notation p.. JRn . since x. 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. so we leave this one part of the proof to Problem l. the map " takes p equivalence classes to p.. we would have if JRn. v]p. v) p (y. IRn. It is the clue which enables us to now define TM. I (V) > V x JRn. because D(y o xI)(x(p» is an isomorphism from to Our definition of TM provides a oneone onto map (b) JRn JRn . If x and y are coordinate systems whose domains contain p. We will denote the fibre . w) for some w E We can easily figure out what the relationship between v and w would be. and TM is defined to be the set of all tangent vectors at all points p E M. and this would be an equivalence (with inverse (xI). this definition is independent of the particular coordinate system x or y..). maps . This condition makes perfect sense without any mention of bundles. to some (x(p).I (p) by the formulas [x..76 Chapter 3 The construction is much easier to understand if we first imagine that we al rearfy had our bundles TM. and T(x(U» should essentially be x(V) x a point e E .. (x. the equivalence class of (x. since (x(p). V) is a coordinate system.. v) will be denoted by [x. w) by y. w) w = D(y o x. then y. we would have a map x. but it is a little ticklish to produce. : TV > T(x(V» . v]p + [x.I (p) would be taken by x. v) is taken to (y(p). v). JRn JRn).(e) would be (y(p). Ix Ix : .I (p) by Mp. v]q � (q. Since TV should be essentially (TM)IV. 0 x.. JRn.I ) (x (p»)(v). Here v is just an element of (and every v would occur. though TMp might be better. v]p = [x. I = (y 0 xI). (a) It is easy to check (using the chain rule) that p is an equivalence relation. We now have a bundle . W E we define if (a) is satisfied. We want this to be a homeomorphism. a .
i ). But T(M. •:. Having succeeded in producing a bundle over each M. V) is a coordinate system around p. For I : M > there is a map h taking the '1 equivalence class N x 0 CI and x a C2. In fact. For the first example. which is equivalent to T(M. we define (c) N. if (x. . i lp . and I is the identity coordinate system of then If . V]p) = [I. for all p E M. it must b e checked that this definition is independent o f x and y (the chain rule again). with c (o) = p. D(i 0 x J ) (x( p»)(V)] i (p) the composition is easily seen to be an equivalence. IR IRn. the bundle 1C : TM > M will be called the tangent bundle of M. v]p) = O f course. T'M. If (x . we consider curves c : (e. V) and respectively. In II. though perhaps confusing to the novice. i). IRk i. ( i ( p) . In fact. i) will play no further role in this storythe abstract substitute TM will always be used instead. the fibre of TV over p E V is almost exactly the same as the fibre of TM over p. each defined on some interval around 0. In II I = by (c) CJ '1 C2 if and only if: The equivalence classes. to prove commutativity of the diagram. Condition (I) of our theorem is obvious. we define In. ([x . e) > M.([x . since in M there are more coordinate systems around p than there are in U C M. Can one find other bundles with the same properties? The answer is yes. D (y 0 I 0 xJ )(x(p») (v)]J(p)' I. D (i 0 x ')(x( p») (v)) E ( M . IRn Ix Henceforth. (y.The Tangent Bundle 77 I : M > and (x. Condition (3) is pI"actically obvious also. we next ask how fortuitous this was. : M IRk . To prove (2). > is an imbedding. the only difference is that each equivalence class for M contains some extra members.. will be the elements of our new bun dle. then TM is equivalent to T(M. [y. and we proceed to define two different such bundles. it is trivial. V) is a coordinate system around p. V) are coordinate systems around p and I(p) . mapping to have the same derivative at O. we define to be where I is the identity map of and is defined in (b).
suppose that J = 0 in a neighborhood of p. we may use this trick to define i (f): choose h to be 1 on a neighborhood of p.) For x E U. and define i(f) as i(fh). then O = i(f . y' (0) = V.xn) for x E U. .. LEMMA. Then n J(x) = J(x) . The set of all such operators is a vector space. There is a function h : M > with h (p) = 1 and support h C (0). since U is convex. For these operators. with support h C JI (0).g) = i (f) . We have already seen that the operators i = have this property. (2) gi (O) = D. Thus. let hx(t) J(tx).J(O) = Jo I hx'(t)dt = l0 I L Di J(tX) .. . . . .i(g). (The second condition actually follows from the first. h corresponds to J•. things are not so simple. if J = g in a neighborhood of O. Then there are Coo functions gi Coo function in a convex open neighborhood U of 0 : U > with JR . Then a/axi lp JI Coo JR 0 = i (O) = i (fh) = J(O)i(h) + h(O)i(f) = 0 + i(f).) of c to the J equivalence class of J 0 c. Let J be a in (I) J(x l . 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. we can already see that this example is "really the same" as TM [x. . = :s . xi dt. clearly i (f) = i (g) if J = g in a neighborhood of p. but it is not a priori clear what its dimension is. If J is defined only in a neighborhood of p. r i=1 Therefore we can let g (x) = J� Di J(tx) dt :PROOF. this is defined for 0 :s t I.xn) = L:7= l xigi (X I . For. This comes out of the following. This condition is actually true for any derivation i. with J(O) = O. 2. • JRn .78 Chapter 3 (/. In the second example. vlp corresponds to: the 'J' equivalence class of where is a curve in with y JRn under this correspondence.f(O). Without bothering to check details. Coo i (fg) = J(p)i(g) + g (p)i(f).
. THEOREM. Consider the case where M = JRn and p = o.. •:. U) is a coordinate system around p./(0). they are clearly linearly inde pendent. l) = l · i(I) + l · i(l). It is a simple exercise to use the coordinate system to transfer this result from to .e can be written a�I Ip . so i(l) = o. Given / on U. a�n Ip (so i is determined by the numbers i(xi »). for the function / . This shows that span the vector space. . a]p. Notice that i(l) = i(l . The set of all linear derivations at p E is an ndimen sional vector space. and any derivation . From Theorem 3 we can see that. Then (Ii denotes the i(f) = i(f . if (x. once again. i(l) = 0 for any constant function c on U. then Mn span this vector space. [x. Hence i(c) = c .The Tangent Bundle 79 3./(0)) = e(£) igi ) coordinate function) 1= 1 = I )i (Ii )gi (0) + I i (O)i(gi)] 1= 1 ith PROOF. . choose gi as in Lemma 2. In fact. Assume U is convex. . 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.
. we wilJ not hesitate to write v(f) for a differentiable function f defined in a neighborhood of p. If e : JR M is a differentiable curve. b ) . 0 It is customary to denote the identity coordinate system on write JRI by this is a basis for JRlo.a]p most easily analyzed from the relation (f. the map which corresponds to can be defined as follows: (x..v)(g) = v(g fl. � . I [x. JRn. t i . and to is calJed the tangent vector to c at 10. TJRn 0 f. then > dt 10 d I for I. en (JRn) . We wilJ denote it by the suggestive symbol dt 10 de l · . that is. In fact. 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. ayj a' (p). and the map f.a) 1=1 axi n a L1 a1. . a tangent vector is often most easily described by telling what derivation it corresponds to.. is often j .=1 a a!. and this is precisely the equation which says that p (y . between E v Mp x ap [f.. It is easily checked that under this correspondence. shows that Chapter 3 if and only if bJ = L. (i)](g) = i(g fl.80 derived in Chapter 2.IP 1= ax consequently.
. I n .a) (y ox . .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. TM Coo(TM) I V v a i by Xi (v) . .I (I). I TM. . and quite a mess to prove.). moreover. Since locally looks like x clearly is itself a manifold. 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.xn (Jr(v)). ax L.x). L:7=1 ai Di (yn x. L. . . ..I )(t)). This expression shows that y. = V 0 b1. Then the map v � (x l (Jr(v» .a l . .I (l. there is. TM TM. . oxI )(x(p)). so it has been quarantined in an Addendum to this chapter. .. . The tangent bundle of a manifold has a little more structure than an arbitrary nplane bundle. 0 = 0 0 0 then. Di (y1. as we have already seen. and Let us denote x x p Jr( ). = � a' ay}i (p) = � a'. . . but it is a little delicate to state.a) = (I I .xl(v). x: V ]Rn CooTM M. 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. This map. . . (x Jr. L:7=1 ai Di(yl xI )(I). is simply the map x. 1=1 1=1 This shows that if (I. V) is another coordinate system. when we identifY TV with V ]Rn in the standard way. As you might well expect.I is Coo. xn (v)) E ]R2n is a homeomorphism from (TM)IV to x(V) ]Rn . . (x. . V ]Rn. . If (y. then y. (x.).. an ) E ]R2n .
In general. we define the vector field fX by > (X + Y)(p) = X(p) + Y (p) . 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. TM for all pE U. s: M there is no vector field on S2 which is everywhere nonzero. in some If we think of as the set of derivations. or Z. M TM M. we have Mp) . (fX ) (p) = f(p)X(p). for Coo vector bundles we can also speak of Coo sections. Y. With this Coo structure. the local trivializations x* are Coo. The functions ai are continuous or Coo if and only if X: U > uous or Coo. IRn. if f: M JR. we define a new vector field TM is contin X+Y by Similarly. then for any coordinate system (x. 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. Vector fields are customarily denoted by symbols like X. and the vector X(p) is of ten denoted by Xp (sometimes X may be used to denote a single vector. U). Recall that a section of a bundle 1C : E > B is a continuous function B > E such that 1C o S = identity of B. If X and Y are two vector fields. It follows that 1C : E > B is Coo. . We have shown that there do not exist two vector fields on the Mobius strip which are everywhere linearly independent..82 Chapter 3 We thus have a collection of COOrelated charts on which can be ex tended to a maximal atlas.
Clearly.The Tangent Bundle 83 Clearly X + Y and IX are Coo if X. Let !F denote the set of all Coo functions on M. Y. thus X is a "derivation" of the ring !F. Of ten. indeed. In fact. The reason for this is that if A : !F !F is any deriva tion. al X(f) = � a ' axi' 1= ) which is a sum of products of Coo functions. then A = X for a unique Coo vector field X. If > a newf unction > > > . and X is a vector field. On U we can write a/axi now denoting the vector field the symbol I: M IR is a Coo function. X(f + g) = X (f) + X(g) X(fg) = IX(g) + gX(f). then X (f) is Coo for every Coo function I. (p) axi I ' P 1= 1 then . 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. It is not hard to check that if X is a Coo vector field. we clearly must define Xp(f) = A (f)(p) . then we can define X (f): M IR by letting X operate on I at each point: X(f)(p) = Xp(f). Conversely. and I are Coo.L. a Coo vector field X is identified with the derivation X . iflocally n a X(p) = L a'. if X (f) is Coo for every Coo function I. then X is a Coo vector field (since X (xi ) = a i l.
linear transformations in the first group are called orientation preserving and the others are called orientation reversing. . if I is orientation preserving) and oppositely oriented if det A < o.' The next few chapters constitute a detailed study of this bundle. vn) and (V'l . . The nonsingular linear maps I : V > V from a finite dimensional vector space to itselffall into two groups. V (VI. . . Vn) and ( 'l . we will discuss just one new concept about manifolds. vn] . . . in particular on the tangen t bundle of a manifold. . We call . Vn) E JL if and only if . vn] = JL. . by Robert A. . Its sense is nicely conveyed by the definition in The American Heritage DictiollaT ': } "To understand profoundly through intuition or empathy". so that if JL is an orientation of V. . The class to which .84 Chapter 3 The tangent bundle is the true beginning of the study of differentiable mani folds. since we have not yet defined anything called "orientation". v'n) equally oriented if det A > 0 (i. . Heinlein in his pop science fiction novel Stranger in a Strange Land.e. . . 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. To provide this extra structure. . j=J n = * A cult word of the sixties) "grok" was coined. and those with det I < 0. The problem becomes more acute if we want to define orientation preserving isomorphisms between two diff erent (but isomorphic) vector spaces V and W. [VI . vn) belongs will be denoted by . . . the matrix A = (aij) of I is given by the equations V.. which is being preserved. and you should not read further until you grok it. (VI.) = v'. . divid ing the colJection of all ordered bases into just two equivalence classes. There is no way to pass continuously between these two groups: if we identify linear maps IRn 7 IRn with n x n matrices. which arises in this very way from the notion of "orientation" in a vector space.. . . . x n . . then . Either of these two equivalence classes is called an orientation for V. . . The terminology "orientation preserving" is a bit strange. [VI . purportedly as a word from the Martian language. 'n ) for V determine an isomorphism I : V > V with I(v. _ xn ) (reflection in the hyperplane xn = 0). 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). . . we note that two ordered bases . A simple example of the latter is the map I : IRn > IRn defined by I (x ) = (x l . . Vii L Qji Vj. . . this clearly makes no sense unless we supply V and W with more structure. The relation of being equally oriented is clearly an equivalence relation. If JL denotes one (VI. .l . those with det I > 0. . For the present. (VI .
. and 1R' orientation of V.I IRn fL) fL. 1R2. . Now if (V . = For the trivial bundle we can put the "standard orientation" on each fibre If > is an equiva lence. 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. . . en] then det (aij) : > is continuous and never o. . . since rr.. (x.The Tangent Bundle 85 VI WI Examples of equally orien ted ordered bases in 1R. . If 1r: E > a non trivial nplane bundle. . (VI . fL X IR fLp rr. j=1 U C B: rr.I (U) U x lRn . then is either orientation preserving or orientation reversing on each fibre. and (W. ed. en (X) X x IRn [(x. ··.e.I (U) U x IRn is an equivalence. for if we define the functions : > by f: [f(VI )' . and is connected. (x. and t' : > is another equivalence. together with orientations. then ( is either orientation preserving or orientation reversing on all fibres. and the fibres of U x IRn are fL B is Notice that if this condition is satisfied for a certain t. . the other will be denoted by and the orientation for will be called the "standard orientation". ej ).. Vn ] fL. 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:>ji (X) . en )] {x} x IRn . then (' automatically satisfies the same condition. f(vn)] [VI . . f: en (X) en (X) X f aij X IR n f(x. v) are two ndimensional vector spaces. . [eI . . . vn ). .
86 Chapter 3 > : x x is an equivalence. This definition can be applied. fLp . 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. we can let A be [0.I . fLp fL fL (O. TIRn :::: en(IRn).I (p) with = A bundle is called orientable ifit has an orientation. For example. If I'01. we had compatible orientations we could define a section Sl > by choosing to be the unique vector E A n .fL) ..1 U IRn U IRn U B.fL) where is an orientation for �. sn. TM.a) a). To see this we standard orientation. in particular. The sphere s(p) fLp .} but not every bundle has an orientation. 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 IRn is M fL s: M [s(p)] fLp. I]. we can pick two vectors from each fibre so that the totality A looks like two sections. although the Mobius strip has no nonzero section. TM IRn M. then A just looks like the boundary of the Mobius strip obtained from [0. considered as a Idimensional bundle over SI . the Mobius strip. and nonorientable oth erwise..I. TM TM (M. For. has no orientation. I] x [. we call itself orientable or nonorientable de pending on whether is orientable or nonorientable. to the tangent bundle of a Coo manifold In this case. I } with identified with ( I . since on which we have the C also orientable. it has another orientation = { . an oriented bundle is just a pair (�. I] x { . For example. s(p) fL is M.
0») and their negatives. (A(p).. . This shows that if and E E then [vt. [A. The Mobius strip is the simplest example of a nonorientable 2manifold. For tlle imbedding of considered pre\�ously we have already seen that on the Ell M M (0. V2] JLp . JLp P E S. since it contains the E B. Nonorientability of can be seen in another way. the restriction �IB' Mobius strip (for any orientable bundle � = to any subset B' C B is also orientable). 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). 1)(0. . ·· . The map ( is just � when is identified with a subspace of { x The map Uj) E Pl the bases is orientation reversing. . Since TS I is trivial. which also discusses the tangent bundle of a manifoldwithboundary. and consequently orientable.Ip we can define (VI . The orientation JL {JLp : P E sn . so if = '' I t i 2" wp S M w [v . Vn.I } thus defined is called the "standard arien tarian" of Sn. " Vn.VI . S2p 11"2 A : S2 > S2 A: IR3 > IR3 A. x x x Any nholed torus is also orientablethe proof is presented in Problem 1 6. by considering the "antipodal map" defined by = This map is just the restriction of a linear map defined by the same formula.O) } c there are continuously varying vectors but that it is impossible to choose continuously from among the dashed vectors for = 1. «0. so for VI . . v) p I IR'A Vi (p. this shows that T(S I S I ) is also trivial.i. w ] JL 11"2 p p p p vp .. A(v» . A (p) p. A subset = { (2 cos e.I the vector w = Pp E en (IRn ) :::: TIRn is not in i.(vI l.(snIp) E TIRnp (Problem 21).V2] JLA(p)' Thus the map A : S2 > S2 is .I . The torus Sl S I is another example of an orientable manifold. 2 sin e. A.I E sn. 0) i .: Vip > (Mi )p is an isomorphism and the subspaces Vip vary continuously (Problem 26).: S2p > S2A p) (p.I ) E JLp if and only if (w.i. S2 JL is the standard orientation of S2 Jr : > are oppositely oriented. if then we could simply choose The projective plane must be nonorientable also. If we had orientations = and wp otherwise.The Tangent Bundle 87 note that for each p E Sn. (vnI l) is in the standard orientation of IRnp . x = .
.' of the atlas . U) (y. because it is just the condition that 0 is orientation preserving. U) x(U) x.:IRnTMI U T(x(U)) :::: x(U) IRn (y xI ). since jj.4. the map would then Vip } on be orientation preserving with respect to jj. If g : map � we obviously can define orientations vp for by requiring g to be orientation preserving. > > (x. (x. which is impossible..4. for sueh that 11"2 S2n 11"2 s > f: M N > [p]. 11"2 f A p [pl. Condition (2) holds. the determinant condition will be very important later on. 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. these same arguments show that is orientable for n odd and nonorientable for n even. For projective 3space the situation is just the opposite. given A' we can orient the fibres of in such a way that is orientation preserving. axj > TM (x. and obtain an orientation of Although our original definition is easier to picture geometrically.: T(x(U)) T(x (U)) TMI U x* TM. is orient able if there is a subset . which does not use the tangent bundle of at all. E ( ayi ) 0 U n v.' cover M M det on . In general. (I) the domains of all (2) for all and V) .'. In this case. There is a more "elementary" definition of orientability.4. U) E M..4.4.. the antipodal map is the is orientation preserving. Conversely.' as the collection of all for whieh x is orientation preserving (when x is given the standard orientation).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). . 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... According to this definition.
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. and a bundle map (jil> for each Coo map M . Nevertheless. then there are equivalences such that the following diagram commutes for every Coo map eM ' TM . for certain equivalences 1m.. Those who have been through this sort of rigamarole before know (i... eMl PROOF. 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). (2) of Theorem I. 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.. T'M TM � TN f' M . N. N satisfying f) f' (I) of Theorem I.... have faith) that it's going to work out. occurs quite a ways on. is. If (x. for certain equivalences (3) of Theorem I.. * Functorites will notice that Theorems I and 4 say that there . T'U '" (T'M)IU. old f ) on * Euclidean spaces. up to natural equiv alence. the welcome symbol . 4.. 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. those for whom this sort of proof is a new experience should find it painful and instructive. JRn. U) is a chart on M. and to the restriction of the functor on open submanifolds.e. THEOREM'.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. Ifwe have a bundle T'M over M for each M..
Then fJx.1 0 ax : (TM)IU > (T'M)IU is an equivalence. we can define fJx . U) be a coordinate system on M. consider the following diagram. are the equivalences mentioned in condition (3).' for T'. Let ax denote the composition (/nlx (U» 0 '" 0 x. (TlR")ly(V) '" IY l. 0 (". U). (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) .TV � T(y(V)) =. Two of them.1 Similarly.TU � T(x(U)) � (TIR")lx(U) ' "lx(Ul. Then we have the following string of equivalences. (TM)IU ""'=". using equivalence ". Our main task is to show that this isomorphism between the fibres over p is independent of the coordinate system (x. To compare ax and ay. which are denoted by the same symbol '" .). so it takes the fibre of TM over p isomorphically to the fibre of T'M over p for each p E U.90 Chapter 3 ax = Let (x. e"(IR")lx(U) (1) (V (TM)I V =. ® Ik' ® Ie 0 Ie . e"(IR")ly(V) Ie <D Ij. This will be done in three stages.
(2). not the explicit construction of TM. 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). and (3). so that it is also true for T'M. i = (TM)IU (TM) I V 1e �Ti "Z ™� I .1 oay = fJx. fJy. Square @ obviously commutes. ax. Thus on the fibres over p . we enlarge it. (II) We now need a Lemma whieh applies to both TM and T'M.The Tangent Bundle 91 Each ofthe four squares in this diagram commutes. using only properties (I). i e� TV Square ® commutes because k ° y = x ° Square ® commutes for the same reason as square (D.1 oax ay fJx fJy. To see this for square (D. for every p E V. it will be proved for TM (where it is actually obvious). The two triangles on the left commute by con dition (3) for TM. Chasing through diagram (1) now shows that the following commutes. and (3). and the one on the right commutes because ° j I. Again. . the inclusions x(U) and y(V) come into play. JRn JRn This means that for P E V. (2). > > i. as shown below.
(TJRn) IA � en (JR)n IA TB =. j/' and JRn JRm > . If and following diagram commutes. A'. This implies that the two compositions TA =. p. There is a map j : JRn JRm with j f on A . General case." are equal on Case 2. (TJRn ) I A � en (JRn) I A � en (JRn ) old j. following diagram. p E A. where every :::: comes from the fact that some set is an open submanifold of another. Consider the JRn and j: B JRm are the inclusion maps. (TJRm) IB � em(JR)mI B PROOF Case 1. and f : A B is Coo.92 Chapter 3 LEMMA. em (JRm) and are equal and this proves the Lemma in Case I. then the > TA =. Everything in this diagram obviously commutes. A. en (JRn) laid j. 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. A C JRn f. p E A' C A . where i : A > > > 1 101d f• = . . B C JRm are open. since the maps "old "old f.
(T lRn) I A o A o i* = v O IL O K. � . TA +. for ease of reference. Thus it suffices to prove J* 0 i* = v 0 JL 0 K. which amounts to proving commutativity of the following diagram. . (TlRn)IA �8n(lR)nIA > i. :::: (T lRn) IA' Je (�) Since j 0 i is just the inclusion of Y T lRn A' in IRn . � . (J O i To prove that A 0 i. since v is oneone. @. in which j: is the inclusion map.The Tangent Bundle 93 and i: A' A is the inclusion map. it suffices to prove that v �l TA' TA y A 7 IRn T lRn :::: (A) (/C) � (V) . we imbed it in a larger diagram. . To see that square @ (which has a triangle within it) commutes. . this does commute. and ® obviously commute. ./ � / (TA)IA' '" @ e @ e Boxes (D. . and @ commutes by Case 1. and other maps have also been named. (2) / / . = � 0 /C .
1 0 ax follows immediately. V) and (y.94 Chapter 3 positions are equal in a neighborhood of any p E A. the two com Commutativity of diagram (2) shows that the composition (III) Now suppose (x. Diagram (3) thus shows that fJy = old (y 0 X. (TlRn)ly(U) / nly(U). and the rectangle commutes by part (II). e n (JRn)ly(u) The triangle obviously commutes. (TJRn) I A � en (JRn) I A old J. because part (I) applies to x and xiV n V. old (y o [I). l nlx(U). In other words. V) are any two coordinate systems with p E To prove that and induce the same isomorphism un the fibre of at p. we have the following diagram. .I). which proves the Lemma.TU � . Vn TA � TB =. 0 fJx. e n (JRn)lx (U) T(x(U)) =. on the subset (TA) I A'. em (JRm)IB. it is clearly an equivalence eM. TM fJy. as well as to y and ylV n v. •:. (T JRm)1 B � em (JRm)1 B coincides. T(y(U)) =. and on A' we can replace "old h" by "old J/'. (TJRn)lx(U) v.. with the composition TA =.I o ax)./ (y o [I). Now that we have a welldefined bundle map TM T'M (the union of all fJx. we can assume without loss of generality that V = V. and are thus equal. Assuming V = V. The proof that eN 0 f. = f� oeM is > Exactly the same result holds for T': left as a masochistic exercise for the reader.l o ay fJxl o ay (3) (TM)IU "":::. The desired result fJyI 0 ay = fJx .
..q) = iI: 2i d (Gi(P). It would seem that M ought to have a metric which makes each Vi open and each Xi a homeomorphism. Gi(q»). Vi)} a sequence ofoneone functions Xi : Vi JRn Vi C M and X (Vi) open in JRn. q E xj' ( Bj). =1 Show that p is the required pseudometric. 1].j : M [0. . (d) Let fi. then there is some Ji with Ji ( p ) ¢ fi(A nC) . j 1 . (e) Suppose that for every p . Hint: First arrange in a sequence all pairs (Ai. Show that p is actually a metric on M. Define gj. G3.* X2 (*) = 0. b e such a sequence for each open set Xi(Vi) .{O} U {*}. (P) . 1. . . this is not quite true: (a) Let M = JR U {*}. and define p on M by 00 1 p(p.j in a single sequence GJ. 1] by _ f g. 3. let d be a bounded metric on JR.Aj) of part (b) with A. such that each Xj Xi' : Xi(Vj n Vj) Xj(Vi n Vj) is continuous. (c) There is a sequence of continuous functions Ji : A [0. Never theless. a # 0. A3.. Let VI JR and XI : VI JR be the identity.{ ° i. and { (Xi. where * ¢ JR. 2. with support Ji c A. can find on M a pseudometric p (a function p: M M JR with all properties for a metric except that p(p. Let with > PROBLEMS 0 > = = > > we x > > > = • • • = . C Aj. with X 2 : V2 JR defined by x2 (a) = a.C. .j. Show that there is no metric on M of the required sort. G2 . A 2 . and let V2 JR .). . 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). . then there is a sequence A I .j (P) P E Vii· P¢V Arrange all gi. and Xi' ( Bi) nXjI (Bj) 0. of open subsets of A such that every open subset of A is a union of certain Ai 'S. by showing that every neighborhood of ° would have to intersect every neighborhood of * .The Tangent Bundle 95 M be any set. which "separates points and closed sets": if C is closed and p E A . Actually.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.
I ( V) .96 Chapter 3 V) and (y. the map 0 ..I (p) . (Ii) a figure eight c::><::::) . B.. Sn (P) (x. SII which are everywhere linearly independent. maps on TM. ... Iy : . Show that in the definition of a bundle map..... A weak equivalence between two bundles over the same base space B is a bundle map (]. then all Ix are homeomorphisms. (b) Show that an nplane bundle � is trivial if and only if there are II sections E SI .. v]q � (q.. giving rise to two Ix : . V IRn . w). f B2 fl . [y . .. on cach fibre.. Vi ) with M = Ui Vi . (c) Conclude from Problem I that there is a metric on TM which makes each Ix 2. and is a homeomorphism of B ontO itself.. (a) Suppose (x...I (p) s: s(p) p SJ (p). • 8.. V IRn). . . show that ] = g 0 h where g and h are contin uous maps such that h takes fibres linearly to fibres.. 6. : E .. note that locally it is just a map x x V IRn . f.e. v). . 3. (c) Show that locally every IIplane bundle has 11 linearly independent sections.. B . x x x x a homeomorphism . Show that in . E with the vector of is a section. [x. (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. i. fl 5. . (a) Show that for any bundle . Show that in the definition of an equivalence it suffices to assume that the map E I 7 E2 is continuous.... where ] is an isomorphism on each fibre. (To prove the inverse continuous.I (V) . f: BI . V) are two coordinate systems. continuity of follows automatically from continuity of ] : E I . E2. (c) Check the remaining details in Theorem l.I (A) for A C V IRn open are exactly the sets of the form Iy.I (B) for B C V IRn open.. but weakly equivalent. .I ( V n V) the sets of the form Ix. 7.. Given a bundle map (].. is a homeomorphism for a collection (Xi . w]q � (q... bundles over the following base spaces: (i) the disjoin t union of twO circles.. Find two inequivalent. are linearly independent for all E B. (iii) the torus. (b) Show that if there is a metric on TM such that Ix. V IRn .. · . while g is an isomorphism 4.. v) .
(a) Let :Fp be the set of all Coo functions f: M JR with f(p) 0. let { x +E O f.lp ' f* [f* (il](g) erg 0 fl. the same must be true p of :F / W. and let W be the subspace generated by all products. define a Coo structure on V and a homeomorphism from V V to TV which is independent of choice of bases. v] (0) v. = x 9. If x is a coordinate system with X(p) 0. Show that x f (x) /x 2 lim +o exists for all f E W. (c) Conclude that (V/W) * has dimension ce = 2 e• 14.:: o. p x"I + W is a basis for :Fp/ W (use Lemma 2). x ::: Show that all f. f* [x. 1 1. as the next problem shows. 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. 10. . The situation is quite different for C functions. for v.. p (c) Since (:F / W) * has dimension n dimension of M. and that they represent linearly independent elements of V/W. JR" [x. As in the case of JR".a. (b) For O < e < I . v). space (:F / W) * .g E :Fp.e has a unique extension to a derivation. 1 3. show that x l + W. (a) Show that the correspondence between and equivalence classes of curves under which p corresponds to the "j equivalence class of . (a) Let V be the vector space ofall C1 functions f: JR JR with frO) = 0.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. .The Tangent Bundle 97 xl . Show Jet e: :F that . g E :F Show that the vector space of all derivations at p is isomorphic to the dual p. (b) Show that under the correspondence the map can be defined by TM Vw > = > > = > = = = > > . L. W E V we will denote by E Vw the vector corresponding to (w.. and p JR be a linear operator with eUg) 0 for all f.. are in V. then f is constant on each component of M. If V is a finite dimensional vector space over JR. . If f: M N and f* is the 0 map on each fibre. (x) = 0 1 x .a/ax.a]p . .
the tangent bundle TM is defined exactly as for elen. Show that this subspace does not depend on the choice of X. in fact. v Mp p M U lHIn p. fog g f M M..: Mp NJ(p).l ] ( )p Mp. (Compare with Problem 233(d).(V) lHInx(p) M [v" . f o g f. .ents of are p equivalence classes of pairs (x. (e) Show that if has an orientation JL. (a) If is a manifoldwithboundary. so is a coordinate system around then directions". . . 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 . under the identification of with the vector is (a.) (e) Suppose we are in the setup of Problem 214. i. lHIn . A vector which is not in E is said to point "inward" if x. the rank of at E is the rank of the linear transformation (b) If at a = where is a diffeomorphism.) 16.98 Chapter 3 15. i* (8M)p > v) lHIn E points inward. TM U TN > .v" .] N x [0. TP g: aM x [0. Although x takes a neIghborhood of E a onto rather than IRn. (a) A map is an immersion if and only if is oneone on each fibre of TM. . [w. then the rank of equals the rank of at g(a). (b) Let a E IRn1 {OJ C A tangent vector in is said to point "in ward" if. it is where a is the inclusion.vn_.t) '" (/(p).. i.(aM)p . p M lHIn. n n g(p. More generally.t ). (The reason for this choice will become clear in Chapter 8. Vn. Define I) a Show that is obtained from I) by x. If E a and x : > > f. f: M N f p M f. Show that this definition does not depend on the coordinate system x. the still has tangent vectors "pointing in all vectors still run through IRn . where > O. TlHIn i: M M lHIna en(lHIn). i.\IRnIX(p» C Mp vn v Mp x lHIn. Mp v) . > is a subspace.
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. 18. b a (bl + b2) = a . Show that identifying  TIP'2 is homeomorphic to the space obtained from T(S2. and there are no zero divisors: . .. show that P has an orientation which agrees with J1.The Tangent Bundle 99 by identifYing v E (aM)p S2 with (.i)_p. b) = (Aa) . v) E (S2. What is the resulting manifold P? Sl [0. v) (S2. Show that such a vector field can be picked so that near (0.I ).av) is orientationreversing. and v on M C P and N C P. > (f) If M and N have orientations JL and v and f: (aM . . Although there is no everywhere nonzero vector field on S 2 .0. That is.a. I)}. b a (Ab) for A E ]R a · (1 .g. b) . a JL) (aN. I) the vector field looks like the foIlowing picture 17...B. . a · b (a l + a2) b a l b + a2 . (p.0) a = • = • = .(v) E (aN)J(p) .i)p with (p.. • • (a. (g) Suppose M is with two holes cut out. Sl .0. which is diffeomorphic to ]R2. Suppose we have a "multiplication" map from ]Rn x ]Rn to ]Rn that makes ]Rn into a (nonassociative) division algebra. (a "magnetic dipole"): 19. there is one on S2 0.i) by € {to. bl + a b2 A(a .
J " /l = . pg. the quaternions are not commutative and the Cayley numbers are not even associative.0. Suppose that (TM)IA is trivial whenever A C M is homeomorphic to S l .0).". [0. a = O). J. (f) TlP'n1 is trivial. F.) 22.(S 2p) by showing that the inner product (p . Adams has proved. 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'. the vector pp E ]R3p is not in i.100 Chapter 3 (For example. 4. e2 . (a. 0) = 2. (b) If a # then a . Multiplications with the required properties on ]R4 and ]Rs are provided by the "quaternions" and "Cayley numbers") respectively.0. so that we always have multiplicative inverses. el for a unique a E ]Rn. It is a classical theorem that the reals. for I. S.0. Conversely. (0. . (a) Consider the space obtained from I] x ]Rn by identifying with ( I . and quaternions are the only associative examples. . Hint: An arc e from Coo . = ( 1. g(t» ) + (j(t). . Arner. see Cai£ulus on Manifolds. a . (c) If p a · el E snI. (Recall that 1 (f. otherwise it suffices to assume the existence of a unique b with a . ' " . complexes. = /l = 2 = 0.. 20. Show that this can be made into the total space of a vector bundle over Sl (a generalized Mobius strip). . > (1. e'(O» ) = ° for all curves e with e ( O) = p and 1c (t)1 = for all t . 23. For a simple proof. v) 21. el . that n = I. then the projection of a . b b . Palais. Show that for p E S2 . (b) Show that the resulting bundle is orientabJe if and only if T is orientation preserving.ad + be). i)p are linearly independent. .d) (ae . this condition implies that there are no zero divisors jf the multiplication is associative. = (1. [Incidentally. . and for we can use "complex multiplication". (d) Multiplication by a is continuous. Let M be a manifold. 366368.bd. en be the standard basis of ]Rn. Tv). (a) Every point in snI is a . see R.) Let eJ . f Math.. a · en on (snl. where T : ]Rn ]Rn is a vector space isomorphism. . (e) TSn1 is trivial. where ' denotes the transpose. Show that M is orientabJe. The Classification o Real Division Algebras. we can use ordinary multiplication. The tangent bundles TS 3 and TS7 are both trivial. . b) · (e. Monthly 75 (1968).. using methods of algebraic topology.g '( t J'). .. or 8. . g)'(t) = (f'(t) '. en are linearly independent.
together with Prob lem 29. e') with 1C(e) = 1C'(e'). (g) Let � = 1C : E be a vector bundle. Show that �" '" �' = f*(�). Hint : Map e E E" to (1C"(e). This bundle is denoted by f*(�). then 1C*(�) is not orientable. e') = 1C(e) = 1C'(e'). f(e)) E E'. the fibre of � Ell ry over p is the direct sum 1C1 (p) Ell 1C'I (p). (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. Let E' c Y x E be the set of all (Y. f) from �" to � which is an isomorph�sm on each fibre. 23. Since 1C : E is a continuous map from a space to the base space of �. the symbol 1C*(�) makes sense. (b) Suppose we have another bundle f' = 1C" : E" Y and a bundle map > > X > > a X > B B > B 24. then f*(�) is also orientable. then i*(�) '" �IA. Thus one can "transport" the orientation of Mpo to Mp . Y. The next two problems deal with important constructions associated with vector bundles. It must be checked that this is independent of the choice of c.The Tangent Bundle 101 po E M to p E M is contained in some such A so (TM)lc is trivial. we can conclude that a neighborhood of some Sl C M is nonorientable if M is nonorientable. f) a bundle map which is an isomorphism on each fibre. Show that 1C': E' Y is a bundle.e) with /(y) = 1C(e). and is called the bundle induced (from �) by f. > B > B > B. and ( j. by using the vector space structure on 1C1 (f(y» . The general. Remark: Using results from the Addendum to Chapter 9. It is called the Whitney sum � Ell ry of � and ry. then (f g )*(�) '" g*(f* (m . (c) If g : Z (d) If A c and i : A is the inclusion map. Show that 1C" : E" is an (11 + m)plane bundle. (e) If � is orientable. possibly quite messy. e) = y. First consider pairs c. A vector space structure can be defined on > > > > X X (f.e) = e. define 1C' : E' Y by 1C' (y. Show that if � is not orientable. (f) Give an example where � is nonorientable. and define f : E' E by f(y. case can be treated by breaking up c into small pieces contained in coordinate neighborhoods. but f*(�) is orientable. (a) Suppose � = 1C : E is a bundle and f : Y is a continu ous map. Let 1C"(e. . show that f*(� Ell ry) '" f*(�) Ell f*( ry). c' which meet only at po and p. (b) If f: Y > B.
I shows that this orientation is independent of the choice (x. and use this to show that Ell is always orientable. show that � ry is also non = X � � � X Ell Ell Ell � 25. and 1CM [or 1CN] : M x N > M lor N] is the projection on M [or N].. .e.) A different proof that TM is orientable is given in Problem 29. (b) If M and N are orientable.x). then both M and N are orientable. then T(M x N) :::: 1CM*(TM) Ell 1CN TN . and ry is nonorientable. (a) Let be = [a()01C) /" .. (b) The Osection of E is a submanifold. and each fibre 1C 1 is a Coo submanifold of E. find two nonorientable Iplane bundles and ry over such that Ell ry is also nonorientable. carried diffeomorphically onto B by 1C. show that � ry is orientable. define 1C : E x E > BI X B2 by 1C(eJ. i. (f) If � is orientable. . (h) If is a "figure eight" (c. the bundle T(TM) is (lrientable.s orientable. (a) If 1C : E > M is a COO vector bundle. 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 . Problem 5). 27. (d) If Ll: B > B x B is the "diagonal map". (a) If M and N are Coo manifolds. I . (e) If � and ry are orientable. show that � ry :::: Ll* (� x ry).102 = Chapter 3 orientable. a ()o1C) /" a�I /" . (g) Define a "natural" orientation on V EEl V for any vector space V. then 1C* has maximal rank at . (c) Given bundles �i 1Ci : Ei > Bi. .f. Ll(X) = (x. (Here is a more conceptual formulation: for v E TM. . then M x N is orientable. the orientation for (TM)v can be defined as of x.1C2 (e2))· Show that this is a bundleI �I x �22 over BI x B2 . Show that the Jacobian matrix of *( ) y* 0 (x*) I is of the form This shows that the manifold TM is alw. (c) If M x N is orientable.e2) (1CI (eI l.!).(p) 26. 28. . a�n /J the form of Y* 0 (x*)..·arh point.
w) for a unique T*v). (v.. the map w . is defined by (T*v. then the manifold E is not orientable. Show that f is continuous and open. If A is the matrix of T with respect to the usual basis. (d) If n : E M is not orientable. E3 with f g = identity of B is exact if at each fibre it is exact as a sequence of vector space maps.. (a) Let Po E snI be the point (0. > > > an exact (e. w) = (v. I). then each bundle E.. (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.. This problem requires some familiarity with the notion of exact sequences (c. which determines a tangent vector of the fibre. (c) Show that 0 (11) has exactly two components. .. prove that SO(II ) is connected for all II 0::. and induction on II.( Chapter I I). I. . is orientable if the other two are. (c) T(TM) is always orientable. n) i v ' 29. > The next two Problems contain mOre information about the groups intro duced in Problem 233. Tw) (for each v.n* X).I). . 30. Show that I (po) is homeomorphic to SO (II .I) for all p E SnI .I (p) is home fomorphic to SO (II . Using part (a). the adjoint of T. so it is w . show that the matrix of T* is the transpose A t . and then show that f... For II define f: SO (n) > snI by f(A) = A (po) .. this informa tion will all be important in Chapter 10. > > 31.. . = a di (O) = ax. show that there is n*(�) TE n*(TB) O. (T*v.The Tangent Bundle 103 Show that (b) Find a curve whose tangent vector at 0 is a/a(X . 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. Tw) is linear. (a) If T : IRn IRn is a linear transformation. 0:: 2 (b) SO(l) is a point. In addition to being used in Problem 32. A sequence of bundle maps EI E2 l. 0. (2) Map X E (TE) e to E2 (b) If 0 EI E3 0 is exact. so it is connected. T* : IRn IRn. .) > > > > Coo vector bundle..
(d) Show that At . 1R) is homeomorphic to O(n) x IRn (n +' /2 and has exactly two com ponents. I] > Y such that Ji(x) = H (x . = Ji . Hint: Consider At . Two continuous functions f Ji : X > Y are called homotopic if there is a. v # o. show that H is continuous.) (f) Show that every A E GL(Il . Hint: Use the Schwarz inequality. A ) > (Y. The map H is called a homotopy between fa and Ji . Ji : (X.{O J . show that f : (IR" . {A : det A > OJ and {A : det A < OJ. IR) can be written uniquely as A = A. A 2 where A. ) 01) GL(Il.IR" . : (X. are homotopic as maps from (IRn. .IR" {OJ ) to (IRn.{OJ ) with det > 0 is homotopic to the identity map.) (b) A linear transformation IRn > IRn is selfadjoint if so that = = for all W E IRn. A) to (Y. At = A. E O(n) and A2 is positive definite. (g) The matrices A. (b) Suppose f : IRn > IRn is Coo and /(0) = 0. 122). We call /0. A. then some subsequence of {A (n) . A) > (Y. It is a standard theorem that a symmetric A can be written as CDC' for some diagonal matrix D (for an analytic proof. by showing that eigenvectors for distinct eigenvalues are orthogonal. . A ) > (y. pg. IRn . 1 . T: T T T* . 1 ] > IRn is defined by H(x. (x) = H(x. and positive definite if > 0 for all Show that a positive definite A is nonsingular. and use part (e). (c) A selfadjoint (or the corresponding symmetric A) is called positive semi definite if ?: 0 for all E IRn. 1R) is continuous and H: 1R" x [0. B) homotopic (as maps from (X. (e) Show that a positive semidefinite A can be written as A B2 for some B. Tw) T v. so that Ho and H. means that f : X > Y and f( A ) c B.104 Chapter 3 standard basis.{OJ) C 1R" . w) (v. and A2 are continuous functions of A. v) = 32. B)) if there is an H as above such that each H. (Remember that A is symmetric. while f(IRn . Show that C can be chosen orthogonal. see Calculus on ManijiJlds. (a) If A: [0. The notation f : (X. (Notice that this also gives us another way of finding the dimension of O(n). If Df(O) is nonsingular. i) i = 0. IRn . IRn . a continuous function H : X x [0.{OJ ) > (IRn.{OJ ) is T:   T . ) converges. B).{OJ ) > (IRn . B). Conclude that a nonsingular linear transformation (IR" . IRn {OJ ). t ) may be thought o fa s a path of functions from Ho = fa to H. : X > Y defined by H. If A is the matrix of with respect to the (Tv. (Tv. The functions H. t ) = A(I)(X ) . A (n)2 . for A C X and B e Y. A is always positive semidefinite. v) T v (Tv. then is selfadjoint if and only if A is symmetric. Hint: If A (n) > A and A (n ) = A (" ) . I] > GL(Il.
I . . such that e satisfies the condition in part (d). V) in e. Let Mil C ]RN be a Coo ndimensional submanifold. the map y o x . 33.The Tangent Bundle 105 homotopic to D/(O) : (]Rn . q E M. (d) For p E ]Rn . (c) Let U be a neighborhood of 0 E ]Rn and f: U > ]Rn a homeomorphism with frO) = O. where U. _x n ). if M is any (not necessarily differentiable) manifold.q for p. : B. we can define M to be orientab!e if there is a collection e of homeomorphisms x : U 7 whose domains cover M. (ii) no tangent plane Mp contains v. We will say that f is orientation preserving at 0 if f 0 h is homotopic to the identity map I : (]Rn . U) and (y. By a chord of M we mean a point of ]RN of the form p . ]Rn . . (e) Notice that the condition on y o x. f o /z : (]Rn . we will say that f is orientation preserving at p if T_ J(p ) 0 f 0 Tp is orientation preserving at O. . x n. . show that if M has such a collection e of homeomorphisms. Thus. then there is a collection e ofcharts whose domains cover M such that for every (x. h(x) = then (� arctan IXI) x.{OJ) > (]Rn .{O)) > (]Rn . use Lemma 2. Check that this does not depend on the choice of C V. t ) = f{lx) for 0 < t S 1 and H(x . 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 . V C ]Rn are open.I is not differentiable. To prove continuity at points (x. ]Rn . then there is a vector v E S N . (a) Prove that if N (i) no chord of M is parallel to v. ]Rn . ]Rn _ {O)) > (]Rn . Let C V be the open ball with center 0 and radius r. Hint: Define H(x. .l .I in part (d) makes sense even if y 0 x. . > 2n + I. .I such that M x M and .O). and let be the homeomorphism h : ]Rn > B. then precisely one of f and T 0 f is orientation preserving at O. xn ) = . Show that if M is orientable. then for any Coo structure on M the tangent bundle TM is orientable. B. If f: U > V is a p homeomorphism.{OJ).{O)).I is orientation preserving at X(p) for all p E U n V. ]Rn _ {Oj) . ]Rn . O) = Df(O) (x). Hint: Consider certain maps from appropriate open subsets of TM to S N . JRII (Xl . Assuming this result. let T ]Rn > ]Rn be Tp (q) = p + q.
(c) Every compact Coo Ildimensional manifold can be imbedded in jR2n + J . In particular. given by Problem ·230 (compare Guillemin . ff Ann. Proofs may be found in Auslander and MacKenzie. Di erentiable manifolds. Show that 1C I M is a oneone immersion. and 1C . :nd Pollack. Note: This is the easy case of "''bitney's classical theorem.J C jRN be the subspace perpendicular to V. In Munkres. together with the exis tence of a proper map f : M > jR. 220246). Whitney.106 Chapter 3 > (b) Let jRN. with N = (n + 1)2). 37 (1935). lntmduction IJJ Di erentiable Manifolds and Sternberg. then 1C I M is an imbedding.J the corresponding projection. if M is compact. Ann. 645680). 45 (1944). Elementary Diif lltial Topo/lJgy. Whitney. Lectures on Di ff f f erential Geometo'. jRN jRN. of Math. 17ze se/finl£rsections ofa smooth llmanifold in 2ns pace. there is a different m sort of argument to prove that a notnecessarilycompact nmanifold M can be imbedded in some jRN (in fact. A much harder J·esult of Whitney shows that f Mn can actually be imbedded in jR2n (H. Di ferential Topology). Then we may show that M imbeds in jR2n +J using essentially the argument above. which gives the same result even for noncompact manifolds (H. of Math. .
. not just on Vi alone. we can define v" E V" = (V')' unambigu ously by for every A E V'. . The linear function V'i depends on the entire set Vl . . if VI. The simplest case arises when we replace each fibre V by its dual space V'. then the elements V *i E V*. Vn is a basis for V. These simple remarks already show that f' is an isomorphism if f : V is. Recall that V' denotes the vector space of all linear functions A : V If f : V 7 is a linear transformation. for finite dimensional V. for (f' fl ' = I v' and > a a g: > g)' g' The dimension of V' is the same as that of V. Vn . On the other hand. 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. and then fit all these new vector spaces together to form a new vector bundle over the same base space. we replace each fibre (p) by some other vector space. . In fact. defined by (f a f. then I v· is the identity map of V' and if U V. = If V" (A) = 0 for every A E V'. if v E V. then (f = f ' . It is clear that if I v : V V is the identity. ll the constructions on vector bundles carried out in this chapter have a In each case.l W' > W > JR .l ) . > W a are easily checked to be a basis for V'. . . then A(V) 107 0 for all A E V'.CHAPTER 4 TENSORS Acommon feature. . . (j'A)(V) = A(jV). which implies that . = I w· . then there is a linear transformation f' : V· defined by rr.
108 v = O. it gives us an isomorphism At first it might appear that �. since each 1f1 (p) is isomorphic to However. (We first pick an isomorphism from (�n ) .) + 1f'1 (p). In contrast. we can prove that there is no natural isomorphism from V to V'. Actually. = . 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. '" �. is an isomorphism. to �n. the dual bundle �. and define the function 1f' : E ' and I : 1f1 (U) U X �n is a trivialization. the bundle �. It is called the natural isomorphism from V to V'*. (Problem 6 gives a precise meaning to the word "natural". formulated only after the term had long been in use. once and for all. If you B into a vector bundle. for the present. If U c B. * (�')' is alwa equivalent to �. and �. by We can make 1f': E' requiring that all such [ ' be local tnvializations. of �. we will see later that in "most" cases �* is equivalent to �. Chapter 4 Thus the map v >+ v " is an isomorphism from V (0 V" .) Now let � = 1f : E + B be any vector bundle. The lack of a natural isomorphism from V to V' prevents us from constructing an equivalence between �. readers may ponder this question for themselves. Once the meaning is made precise. Let E' = B to take each [1f1 (p)]* to p. in the obvious way: since the map [ restricted to a fibre. then we can define a function + + peB U [1fI (p)]'. this is true merely because the two vector spaces have the same dimension.
the fibre of T*M over p is (Mp)*.. to . * the same is clearly true for x. is called the cotangent bundle of M. Like TM. as well as continuous. sections of T OM. it will appear obvious that this map is indeed an equivalence. that dc/dI ll". denoted by T*M. the cotangent bundle T*M is actually a Coo vector bundle: since two trivializations x and y* of TM are Coorelated. If w is a Coo section of T*M and X is a Coo vector field. This function will be denoted simply by (J(s). Even if can be pictured geometrically (e. ' and y. in particular.Tensors 109 think about how is constructed. When this construction is applied to the tangent bundle TM of M. The seclion df is called the differential of f. �* . * * * We can thus define Coo. ' (in fact. If f: M � is a Coo function. there is seldom a operates on If s is a section of and (J geometric picture for Rather. Yo ' 0 (X . then a Coo section df of T*M can be + defined by df(p)(X) = X(f) for X E Mp. then w(X) is the Coo function p >+ w(p)(X(p» . �* �: � � P >+ (J(p)(s(p» s(p) E 1C 1 (p) (J(p) E 1CH (p) = 1C 1 (p)* . where e(lo) = p. is a section of then we can define a function from B to � by � �* �* .)I = y 0 (X )I ). Suppose. 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. the re sulting bundle. if is TM).g.
we see that = d(f(e(t» ) . . we obtain a classical formula: . this equation takes the nice form df dl (�) If (x.. . alaxnlp of Mp.=1 n W = L w. In fact. The section df must have some such expression. The section w is continuous or Coo if and only if the functions Wi are. .IrO Chapter 4 Adopting the elliptical notations de dl for dt r ' de l dg(l) dl for g' (I). U) is a coordinate system. This means that every section w can be expressed ulliquely on as V for certain functions Wi 011 U. Thus dx'(p). . We can also write W(p) = L Wi (p) dx i (p). dt T *M over U.=1 if c define sums of sections and products of functions and sections in the obvious way ("pointwise" addition and multiplication).. then the dxi are sections of Applying the definition. . .. . . . dxi .. . \. dxn(p) isjust the basis of M/ dual to the basis alax' lp.
No one wanted to admit that this was nonsense. df must be a function of this change. this point of view was in n af .TeIlS01J 111 I . which became tangent vectors. which means that df should be a function on tangent vectors.just as Leibnitz had. 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).(p) L a . except for quotients of infinitely small quantities. all classical notions involving infinitely small quantities became functions on tangent vectors. like dc/dt . and waiting for everybody to catch up. THEOREM. Looking back at the classical works fi·om OUf modern vantage point. df(p)(Xp) = Xp(f) = " a' . i. =L axi (p) dx 1=1 •:. 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. Classical differential geometers (and classical analysts) did not hesitate to talk about "infinitely small" changes dx i of the coordinates x i . no matter how obscurely expressed. which preserved the old notation. like df. The dx i themselves then metamorphosed into functions. In short. a tangent vector. . ' (p)(Xp) . and it became clear that they must be distinguished from the tangent vectors a/ax" Once this realization came. If (X. af .. U) is a coordinate system and f is a Coo function.=1 X. it was only a matter of making new definitions. Since df is supposed to be the infinitesimal change of f under an infinitesimal change of the point. one can usually see that.e. then on U we have P ROOF If Xp E Mp is then Thus n .
.](x(e(I» ) .=1 ... c: � + j lffi.) or = L: D. Then 1 0 C(I) = j(x l 0 e(I). dx i (e(I» dl dC · dx i dl dl gives 1=1 af dl = � ax' dx" L. . x"(I» . . . . say I I(XI. . CLASSICAL FORMULATION = Let I be a function on M.=1 d(f(c(I» ) dl Multiplying by .) = L: � (c(I» dl ax' . It is the closest approach in classical analysis to the realiza tion of as a function on tangent vectors.==1 " al = " �(c(I» . 1 (1) l(x l (I). we have C(I) is dl af dl = � axl dx" L . Then I becomes a function of I.=1 (f 0 C)' (I) ax' dl (The classical notation. . and x a coordinate system (so that I oX for some function on namely Let MODERN FORMULATION Let X i be functions of t.(This equation signifies that true results are obtained by dividing by dt again.) de al dl (. . where M be a curve. . . .1 12 Chapter 4 some sense the one always taken by classical geometers. x"). = j = l oxI). (Xi 0 e)' (I) . .. . In fact. as we shall point out once again in Chapter 7. (Xi 0 e)'(I) L 1=1 axl af = � ax' (C(I» L. " ( ) Since every tangent vector at of the form dc/dl.. which suppresses the curve C. .=1 Consequently. say X i X i (I). =j = I 0 c : � + �.. the diff erential d was usually introduced in the following way: l Let I be a function of the X l .x" 0 C(I» .x". no matter what tilef unctions X i (I) are. . is still used by physicists.1I . dl L. ... We now have We now have dl = � af dX i ..
we will continually examine the classical way of expressing all concepts which we introduce. then there is a map J : TM + TN. by J *w. . * for each P E M. Classical terminology used these same words.. we can * * do something with the cotangent bundle that we could not do with the tangent *N.) This section � is * denoted. the CCtranslation" of classical terminology becomes only a little more difficult than the translation of the German in which it was written. There is no corresponding way of trans ferring a vector field X on M over to a vector field on N. So it's very easy to remember which kind of vector field is covariant. we push it over to N by J . it gives rise to a map Strict notational propriety would dictate that this map be denoted by but everyone denotes it simply by (/*p)* . and which contravariantit's just the opposite of what it logically ought to be. the same q E N may be /(Pt) for more than one Pi E M. Suppose w is a section of T as follows: �(p) = w(/(p)) 0 J*p. should equal J P2 ' On the other hand. we can say. 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. Recall that if J : M + N is Coo. and then operate on it by w.Tensors 113 In preparation for our reading of Gauss and Riemann. roughly. 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". (The complex symbolism tends to hide the simple idea: to operate on a vector. Then we can define a section � of T *M bundle. And no one has had the gall or authority to reverse terminology so sanctified by years of usage. Since J p is a linear * * transformation between two vector spaces. Notice that we cannot put all J* together to obtain a bundle map from T *N p to T *M. and it just happens to have reversed this: a vector field is called a contravariant vector field. and there is no reason why J p.e. Despite these diff erences. After a while. in f act. we have a map J p : Mp + Nf(p) . naturally enough. i. while a section of T *M is called a covariant vector field.
=1 is also called "covariant". .n x(v. Comparing (*) with D (x' = x' n I dxIj = L � dXI . .1 14 Chapter 4 The rationale behind the classical terminology can be seen by considering coordinate systems on which are linear transformations.j 1=1 dxi erentials "change in the same way" as from Theorem I. we see that the diff the coordinates hence they are cccovariant".I ) o XI . . x(a l vl + . + anvn) = (a l . " aax. n W = L Wi dxi . _  1' P I x(p) = (a. bv. any combination xi. COllsequently. then Clearly so x' aij = ax'j laxl . . an). Notice that if we also have . if then x lffi. .b) If is another such coordinate system. (ax'j laxl) is the constant I h is can be seen directly from the fact that the matrix matrix 0 X. x'j = L7=1 aijx" for certain au . .) = el.=1 n " W = L WIi dxII . so the x coordinate system is just an "oblique Cartesian coordinate system". In this case. .
. . Vk+I . Substituting 115 into the first expression for w and comparing coefficients with the second. . If V t+ T(VI . Wi ax" j L ax'j 1=1 La On the other hand.. . .e. XII used to be denoted by XI . . the indices on the x's were shifted upstairs again to make the summation convention work out. Vk+ h " " Vrn) VI . .. . . .l lj = '"" i aaxlj . This suggested subscripts Wi for covariant vector fields and superscripts at for contravariant vector fields. the functions a'j must satisf y a n . (Coordinate func tions X l . which is a warning that worse things are to COme. = � 'I axIl . The set of all such T VI . . Vm = V. Af ter this was firmly established.a X for a vector field. .Tensors then we can express the W'.. . i n terms o f the Wi . Vk _l . Xn . Vrn. . . are also called covariant and contravariant tensors (or tensor fields) of order I . given two expressions � a1 L 1=1 a a ax. . .=1 L. . sections of T *M and TM. a function is multilinear if is linear for each choice of is clearly a vector space. . V. we find that n I W = . 1=1 These expressions can always be remembered by noting that indices which are summed Over always appear once "above" and once "below" . Vk_l .) Covariant and contravariant vector fields. . this vector space will be denoted . respectively. If Vb " " Vm are vector spaces. . . . . i. We begin with some worse algebra. .
® T. . (S. and S E r i (V) we can define the "tensor product" T ® S E rk+i (V) by T ® S(v" . S. . Vk ) . . if VI . U c B and is a trivialization. . . then the elements For T E r k (V). so we can define nfold tensor products unambiguously. Vn is a basis for V and v *I . 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. . which thus has dimension nk We can use this new algebraic construction to obtain a new bundle from any vector bundle � = 1C : E + B. .) ® T = S. In particular. . Vk +' . S(Vk+ " . · · . . . + + and let If = peB U rk (1C' (p» . Vk+i ). these maps can 1C'' (U) + U X �n k .116 Chapter 4 by rm(v). Vk . . . Vk +i ) = T(v" . . . . . this tensor product operation is itself multilinear. . v *n is the dual basis for V* = r'(V). T ® S is not S ® T. defined completely analogously to the case m = I: are easily scen to be a basis for r k (V). If f: V + W is a linear transfor mation. etc. Notice that r ' ( V) = V*. . then there is a linear transformation f* : rm(w) + rm (V). O n the other hand. We let E' O fcourse. (S ® T) ® U = S ® (T ® U). ® T S. .
. The section A is continuous or Coo if and only if the functions A i . . Tk (�) �*k T (TM) (Mp) *. . ® dX lk (p) E Tk (Mp) I ::. il . . ® dX ik now denotes a section of Tk (TM) .. . ../. .: where dX il ® . ® dXlk . Xk Notice that A is multilinear on the set V of Coo vector fields: A(XI . Xk) = A(XI . . on U every covariant tensor field A of order k is a basis for can be written dx l (p). .lk (P) dxi' (p) ® . . . t" . . . .Tensors 117 We make H ' : E ' + B into a vector bundle by requiring that all such / ' be local trivializations.i/. . . . we just use equation (**) on page 1 1 4. . Xk ). . For the case of the bundle is called the bundle of covariant tensors of order k. then the kfold tensor products dXil (p) ® . U) is a coordinate system. . . . A covariant tensor field A of order k can just be thought of as an operation A on k vector fields XI . . . Xi + X'I .. The bundle is the special case k = I .ik dX11 ® . . so that TM.. .. If (x. A(Xh .k or simply Ifwe also have then (the products are just ordinary products of functions). ... . il.···. . • • .ik ::. aXi. ® dxlk (p). dxn (p) A (p) = L AI) . il.. ..··. . . . . Xk ) Xk ) = aA(XI . . . . . ik are. . . To derive this equation. .: A = L A II . . . . n are a basis for Tk (Mp). . X'" . Thus..Xk ) + A(XI .XI . . . which yields a function: . . and multilinearity of ®. . and a section is called a covariant tensor field of order k. . .
. THEOREM. Xd p» We are finally ready for another theorem. If P ROOF Note first that if v E Mp is any tangent vector. . 1X" . (p). In fact. Now if v" . (p) for each we claim that i. . Xd p» = I(p) · A(X" . . then we can define v = L al axl p ' .. . . A : V x · · · x V + F ". . . . . . Xk)(P)· = l (p)A(p)(Xdp). . . \vhere each at now denotes a constant function and f is a Coo function with I(p) = I and support I C V. . . . . Xk)(P) = A (p)(Xd p). . . . 2. it is actually linear over the Coo I Ullclions F. .e. then for we have A(X" . . .=1 n a I on U outside V. because A is defined "pointwise". Vk E Mp are extended to vector fields X" . then there is a unique tensor field A with A = A. if I is Coo. . . if U) is a coordinate system and is linear over F.' k limes (x. . . then there is a vector Geld X E V with X(p) = v. XI . · · · . f(p)X..1 18 Chapter 4 rvforeover. one that is used over and over. . i. . . Xk E V we clearly must define The problem is to prove that this is welldefined: If XI (p) = Y. . . Xt (p).
Note that Theorem 2 applies. the cotangent bundle: a function fi·om V + 'F which is linear over 'F comes fiom a covariant vector field w .) For simplicity. Let f be a Coo function with f(p) = 1 and support f C V. to the case k = 1. = 0 if g C V. A since bl (p) (g � ) a 1 = (p) o. � bl (p) .Tensors 1 19 (I) Suppose first that X = Y in a neighborhood V of p . take the case k = 1 (the general case is exactly analogous). then is a welldefined Coo vector field on all of A(X)(p) = M which equals X on V. The proof that A(X)(p) = A(Y)(p) when X(p) = Y(p) is in two steps. where T k (TM) = T*M. so that by (I). in particular. to use the in terminology.b' a x ' V) be a coordinate system around p. Now A(Y)(p) = = 0. we will never distinguish between the tensor field A and the operation A.. so f A(X) evaluating at p gives X(p) = o. If g is 1 in a neighborhood V of p. and support 1=1 � a .. Because of Theorem 2. 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: . A(Y) (p). it obviously suffices to show that A(X)(p) X = L. where all b' (p) = o. so that on V we = A(Y)(p). = fA(Y).. Then fX = fY. can write = (The map A "lives at points". Let ( x . •:. AUX) = AUY) A(X)(p) (2) To prove the result. nor will we use the symbol A any longer.
a A(p) = L A}. x Mp Although covariant tensor fields will be our main concern.jk A contravariant tensor field A of order k can be considered as an operator taking k covariant vector fields WI . . h. j(. . if and are covariant tensor fields of orders k and I. . If we have another such expression. <8> __ . . · ·}. .. there is an analogue of Theorem 2.j. Recall that a contravariant vector field is a section X of TM._ ' aX'Jk aX'}1.. 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 . .P" " Pk = L jl. or simply a A = Lh..k (p) .jk (remember that each alax j Ip operates on Mp*). . proved exactly the same way.. each on Mp* . that allows us to dispense with the notation A. . we just define V(A) to be A(V). " <8> _a_. So each Xp E Mp. A A(p) a. thus.··. aX}k ax}' h.. In local coordinates we can write k + 1 times).h� <8> .···. ·· h __ <8> . if only for the sake of completeness we should define contravariant tensor fields. respectively. Wk into a function: A Naturally. Now an element v ofa vector space V can be thought of as a linear function v : V* + �. " <8> . A contravariant tensor field of order k is just is a klinear function a section of the bundle T k (T*M). if we use TdV) to denote all klinear functions on V*.: Ah .···.120 Chapter 4 A <8> B (operating on Moreover.IP <8> . then we easily compute that A. .. We could also use the notation '1k(TM). . a A = L A. . 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 . j(.IP aX}k ax}'.
Moreover.' ( V) have dimension n'.j . If V is a vector space. ' (V). however.' (p) by 7.A) = A(V). covariant of order I and contravariant of order I .' (p» . In particular. which implies that S(v) = for all v.' ( V) denote all bilinear functions T: V x V· + R A vector bundle � = 1C : E + B gives rise to a vector bundle 7. Given S. 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. T: V x V · + �. sections of 7.' (�). certain ones are quite important. Notice that each S E End( V) gives rise to a bilinear 5 E 7. 5: by the formula Vx v· + �. i. . The inverse.A) = A (S(V» . Let End( V) denote the vector space of all linear transformations T : V + V ("endomor phisms" of V).' (1C . this number is called the contraction of T. although they should be kept to a minimum. given by Generally speaking.Tensors 121 Finally. this map is an isomorphism. we can take the trace of the corresponding S : V + V. for each v the vector S(v) E V is merely determined by describing the action ofa A on it according to (*). 5(V. If v" . obtained by replacing each fibre 1C . our isomorphism can be used to transfer any operation from End( V) to 7/ (V).' (TM) are called tensor fields. given a bilinear e(v. V is a basis of V and n T = L T/v·i <8> Vj . we are ready to introduce "mixed" tensor fields. . Since both End(V) and 7. is not so easy to describe. It is not hard to check that this isomorphism of End(V) and 7. There are all sorts of algebraic tricks one can play with 7. let 7. .' ( V) is linear and one one.' ( V). for 5 = implies that A (S(V» = for all A. To make the intro duction less painful. In particular. the correspondence S >+ 5 from End( V) to 7.
i=1 = II � trace A (p) if we consider A (p) E End(rrl (p» . Thus. in any fibre bundle 7i (�). in fact. each A (p) is an endomorphism of rr1 (p). More over. fibre by fibre. I L a axj ' II (contraction of A) = The general notion of a mixed tensor field is a straightforward generalization. obtained by replacing each fibre rrI (p) by End(rrI (p» . If in a coordinate system .) These identifications and operations can be carried out.' i limes <8> • • • <8> V* + R . covariant of order and contravariant of order 1 . aj/ v*j (S(v/» T(v/. J. a section A of Ti (�) can just as well be considered as a section of the bundle Elld(�). which is a section of we can consider each A (p) as an endomorphism of Mp. <8> i i.1 (TM). 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.122 then we can find the matrix A = in terms of the T/. . In particular. each section A gives rise to af unction (contraction of A) : B defined by p r+ contraction of A (p) = + 2 T L T/. 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. defined by S(vt) L aji Vj . v*j) T/ . A= A J dX '. In this case. = Chapter 4 (aij) of S. then 7. given a tensor field A. and we obtain a function "contraction of A".
Consequently.k (TM) (�).. A (�) A= L il k . Here is an important example. J... ·. ·.. one will find the following assertion: "The Kronecker delta is a tensor. for ... all important tensors are actually defined by defining the nHI "HI x x'..e.. a tensor field of type can be expressed as � J. h.j � ax'p .: jl.Tensors 123 Every bundle gives rise to a bundle Sections of are called tensor fJelds.'\'o.···. . . ax'P. x. i. cca set of functions which changes according to (*)". .. and if A' = L il. nk+l functions in terms of the coordinate system and then checking that (*) holds. aX. jl.'.k (�) . . i.. Classical differential geometry books are filled with monstrosities like this equation.: � .ijr .·. 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. in classical diff erential geometry.j ax. axh i( /. Locally.·.ijr • .i��. In every classical diff erential geometry book." In other words. Ai.: then axil axl ax'PI A'g::: g� = L Af(..a axj et L this is certainly true. In fact.lh/. or simply of type an abbreviation that also saves everybody embarrassment about the use of the words "covariant" and "contravariant". then (*) holds. covariant of order k and contravariant of order I.al ' " ax'ctkJ.. it is asserted that if one chooses the same functions for each coordinate system. n' of of oP .
. independent of the choice of the coordinate system x To identify the mysterious map A(p). in a similar manner. it isjust the identity map..124 Chapter 4 a " A Le 0f.) a collection of functions one for each coordinate system.) LO! dx'(p) <8> a!j IP (v.j j = L.J.. a I " n Thus isjuSI the evaluation map Mp x Mp' + �.J is a certain tensor field. P=I = + �.) i. i. Mp x Mp' we consider = v E Mp and J. considered as an endo morphism of Mp... L bp dXP (p). i=1 = J. Given a tensor.. what this equation shows is that '. aib.J. E Mp' with the expressions J.e. dxl <8> axj = From our point of view.j j b i. _ L.. classically.0.. The contraction of a tensor is defined..(v). sat isf ying A(p) AI. then A (p)(v.
a=1 "transform correctly" if the A�r do. a n to each coordinate system numbers a l ) . ax'j = ") a'.j = L AL . A) . so that vector fields can be defined as sections. .. (Mp) to be B(p)(v) . While a COlltravarianl vector field is classically a set of n functions which "transforms in a certain way".. . is independent of the basis with respect to which the matrix is written. VI . the functions (�) we are laking B(p) n BJ1. .3 (Mp ).} n ax = L A !. A2) = contraction of: . ax/ja . L.j a=1 ax = L A{a) i.=1 n so that this sum is a welldefined function..(p).ja ) L L L: ax">: ax a=1 i. A). defined as the sum of the diagonal entries of its matrix. V2.yv = L AaPvya P " J1. x' ' aJ (v. then 2 E T. This calculation tends to obscure the one part which is really necessaryverification of the fact that the trace of a linear transformation. For example. . Ifwe consider each A(p) E T. A2. L ax'a. . . . a vector at a single point p is classically just an such that the assignment of n TZumbers a l . . a tensor of type can be contracted with respect to any pair of upper and lower indices. V2.A) . . n .ax i . Incidentally. rather than as equivalence classes of sets offunctions. The revolution in the modern approach is tllat the set of all vectors is made into a bundle..= This is precisel the definition we adopted when we defined tangent vectors as y equivalence classes [x. and that all other alp. a 'n assigned to satisfy x.Tensors we note that 125 a=1 � A'aa = � (" Aj' � aX. A) >+ A (p)(v.
such invariant definitions are usually not so easy to come by. 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. but we will give an "invariant" definition (one that does not use a coordinate system) of any tensors that are defined. . and hence must come from some A . We won't go quite that far. Unlike the "Kronecker delta)' and contractions. We seldom define A (p) directly. For a long time. instead we define a function A on vector fields. The tangent bundle itselfwas almost a victim of the excesses of revolutionary zeal. As we shall see. invariant definitions of all the important tensors in diff erential geometry are made by means of Theorem 2. the party line held thaI TM must be defined either as derivations. or as equivalence classes of curves.126 Chapter 4 types of tensors are constructed from this bundle. At the appropriate time we will discuss whether or not this is all a big cheat. which miraculously turns out to be linear over the Coo functions F.
) } 1 . and suppose that (x. U) and (y. . . V) are coordinate p and f(p). (c) Show that U*dyj)(p) = t a(y. • • Wn for V are equally oriented. . L. respectively..(v) = df(v)f(p) E �f(p)· v*).···. show that 4 .Tensors PROBLEMS 127 f: Nm. then 1 . (a) If g: N �. Jn • • • + 3. (b) Show that a bundle � is orientable if and only if �. (a) Show that i fthe ordered bases VI . Y j=1 and. . ) w* .(p) · ifJ If(p) ' L. Let Mn systems around + + (Proposition 23 is the special case (b) Show that a a(y f) a f* ( axi Ip) = � ajXI. � be Coo. . show that f. is orientable.j" dyh <8> <8>dyj.g: M f: M dx" f* ( . v* N are Coo. . dxl (p). . w*) . . . then the same is true of the bases n and n for V*. .. Vn and WI . If f. ah . Let + dUg) = fdg+g dj. express j�(L7=1 a i alax i l p ) in terms of the alay j lp . .x� f) (p) . . 1=1 0 f = identity.) (d) Express in terms of the 2. . more generally. For v E Mp.
or b = c and ars is skewsymmetric..1 which satisfy the equation.ij = a'ji for the coordinates in any other coordinate system. (vi) I f a j AiA j i s a n invariant for A i a n arbitrary vector.is an invariant and either Ai or fJi are the components i of an arbitrary [covariant or contravariant] vector field. and consequently i 1 a ij + aj = (fJj!Jj + fJj (Ji ). then a. i (vii) If a ij AiJ. (ix) By definition the rallk of a tensor of the second order aij is the rank of the matrix (aU). i (iii) If fJ. Show that the rank is invariant under all transformations of coordinates. n . check them. The following statements and problems are all taken from Eisenhart's clas sical work Riemallnian Geomet�). then any vectorfield Ai is expressible in the f orm where the a's are invariants. where i for i = I. . . (iv) If aij = a ji for the components of a tensor field in one coordinate system. then either b = c and ars is symmetric. (v) If aij and bij are components of a tensor field. if Qij)J). so are aiJ bki . using the classical methods. 1. . (ii) If Aali are the components of n vector fields [in an nmanifold] . that is. . so Ai jLi means L7= 1 Ai JLi .�fJ.1 and fJi Vi oJ 0.j = 0 for all vectors Ai such that Ai fJi = 0.1 independent vector fields i Aali for Ci = . . Ci = 1. if Vi is defined [cJ (iii)] by a ij Aali v j = 0. . . det(AaI') oJ 0.. n .= 0 is expressible linearly in terms of II . in particular. Remember that the summation convention is always used.: (viii) If ars are the components of a tensor and b and c are invariants. . and these vectors are independent.128 Chapter 4 (i) If the quantity ). . so are aij + b lj .are the components of a given vectorfield. Hints and answers are given at the end. n the vector. show that if bars + casr = 0. . then a ij + aj i are the i components of a tensor.} fJ. after (xiii). . the other sets are com ponents of a vector field. . . . i . . where fJi is a given covariant vector. then (alj . and then translate the problem and solution into modern terms. An "invariant" is just a (welldefined) function. and by definition 5.) = 0. any vectorfield A satisfy i ing Ai fJ. Il indicates the component and ex for ex = I. . In each case.(Jj )�i� j = 0 is satisfied by every vector field �i. If aij and bkl are components of a tensor field. then aij + Qj = O.
. T(v. then T(Z. v + w) = T(v. . choose Y complementary to kerw at all points. T(v + w. .) . I) consists of vectors tangent to S' x {I}. . Z). If O"(Z) = T( Y. . For example. 1f T : V + V and fJ. v* n) . i (xii) I f a ij Ai = aAj holds for all vectors A i such that p iAi = 0 . . . is one. . ip (i) w is determined if w(X) is known for all X. ip} i f j) . . . which span ker w at each point. can i be written in the form (a j . let T' (v. (iii) Given w [with w (p) oJ for all pj . For. Xn _. (This is true only locally. jp is a n even permutation o f i) .E V * and T(v) = av for all v E kerfJ. (vii) Given w [with w(p) oJ for all pj .. and vice versa.. .l. (xii) Let V = p' . w) = T(w. then (xiii) If j 8 h . Similarly. where fJ. w)] . . . . (ix) T: V x V + � corresponds to T: V + V' [where T(v)(w) = T(v. . where a is an invariant.such that HINTS AND ANSWERS. . show that for the symmetric tensor aibj + ajbi the rank is two. . . jp) oJ {i" . . . Show also that a ij = 8 ija. there are everywhere linearly indepen dent vector fields X" . v) + T (w. 'I ···'p then Of. jp is an odd permutation of iI. . Then T + T' is determined by S(v) = T(v. there is a y complementary to ker fJ. . = ° 1 I 1 if ja = jp for some a oJ fJ Or ia = ip for SOme a oJ fJ or if U" . . v) . . . . . . The rank of T may be defined as the rank of T (consider the matrix of t with respect to bases vI . . Z) = w(Z)O"(Z)!w(y) for all vector fields Z. Vn and v* ) . w) + T (w. .i�p are the components of a tensor in all coordinate systems. (xi) Show that the tensor equation aij Ai = aAj.Tensors 129 (x) Show that the rank of the tensor of components aibj.:·. v) = for all v implies that T + T' = 0. w). v) + T(v. ip if j) . . v) . . on S' x � there is an w such that ker w (p. ° ° ° M T(v) = av + fJ(v)y for all v .and writing v uniquely as vo+cYo for Vo E ker fJ. (Begin by choosing Yo complementary to ker fJ.a8lj)Ai = 0. .is a given vector. . where ai and bj are the components of two vectors.) (vi) For T: V x V + �. if the equation is to hold for an arbitrary vector Ai .
k F U) ( T)(AI> . . . F(V) = V*'. F U) = f * is a contravariant functor. . such that F( I v) = I F(V) and F(g 0 f) = F(g) 0 F(v). (a) Let iv : V + V** be the "natural isomorphism" iV (V)(A) = A(V). 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).130 Chapter 4 0 : V x . + W is an isomorphism. . except that FU) : F(W) + F(V) and F(g 0 f) = FU) 0 F(g). Ak 0 f) is a functor. FU) = f is a functor. FU) = f'* is a functor. . Ap) = det(Ai (Vj» . . (c) The "T functor". f l fr + �* which makes the (a) The "identity functor". . . W � W* Hinl: There does not even exist an isomorphism i : � diagram commute for all linear f : � + �. FU) = f * is a contravariant functor. . x V x V* p times x (xiii) Define p times . . then FU) is an A conl>ava/ian/f ullclor is defined similarly. . F(V) = T k ( V). (f) The "T k functor". ' Ak ) = T(AI 0 f. . . 6. . x V* + � by O(VI . !" . (b) The "double dual functor". may also be defined . (e) The "dual functor". F(V) = V*. . . Functors of morc than one argument. covariant in some and contravariant in others. Vp . F(V) = 1k ( V) = Tk ( V*). V � V* 1** . Show that for any linear transformation f : V + W. (d) If F is any functor and f : V isomorphism. W � W** l 1 + V* such that the fol 7.
(e) The functor F(V) = V* is continuous. once and for all. Call F continuous if this map is always continuous [using the metric in part (a)] . W) with the m x n matrices. the same construction can be used when F is a functor of several arguments. and consequently give it the metric of �nm . �) is a homomorphism. wn ) be ordered bases of V and let e = (e l . . Show that a different choice of bases leads to a homeomorphic metric on Hom(V. we can identify Hom( V. vn ) and w = (w" . If e + v denotes the isomorphism n taking ei to Vi. . and show how to construct a bundle F(�). Then F(A) : F(�n) + F(�"). �) we can consider it as a map A : �n + �n.) Generally. W). (b) A functor F gives a map from Hom (V. The bundles 'Jjk (M) are all special cases. the class of ndimensional vector spaces. . . (a) Let Hom(V. (The bundle Ti (TM) is just a case of the construction in (b). . to yk Given A E GL(n. show that the following diagram commutes . W) to Hom(F(V). . (b) How does the homomorphism h depend on the initial choice of the isomor phism F(�") + �k ? (c) Let v = (V I . See the next two problems for other examples.) (d) DefIne a continuous contravariant functor F. then there is a bundle F(�) = 1C' : E' + B for which 1C'1 (p) = F(1C. W ) denote all linear transformations f rom V t o W. an isomorphism F(�n ) + �k Then F(A) can be considered as a map h(A) : �k + �k Show that Iz : GL(n. . F(W» .1 (p» . and F is continuous. . �) + GL(k . Show that if � = 1C : E + B is any vector bundle. . as well as an example of a functor which is not continuous. . . (a) Let F be a functor from Y".Tensors 131 corresponds a trivialization 8. Choose. e ) be the standard basis of �n . Choos ing a basis for V and W. 9. (The bundle T*M is a special case of the construction in (d). and such that to every trivialization I : 1C1(U) + U X �n (c) The functor 7J (V) = 7 1 ( V*) = V** is continuous.
and let h : GL(n. define A by (f) If V.q] = [v. that Fh is a functor.Vj . W. This suggests a way of proving the following. j =1 . Show that this is a welldefined linear transformation. and f: V + W.�) + GL(k . + q2] a · [v. W E f(v. (A) = h(A) when we identify Fh (�n) with �k by [e..vj .. aq] � (w . We will denote the equivalence class of (v. choose ordered bases V . q) for a given v. define ( V .j) is defined by Wi = L aj. aj. �) + GL( I . and that everyequivalence class contains exactly one element (v. aj. and that F. Show that � is an equivalence relation. + such that the homomorphism defined there is a functor Fh in part (a) is equal to h. q] r+ q. �) is any homomorphism.wj.q) by [v. q2] = [v. q]. this means that �k h(A) �k [ // F(e + v) F(e + w) also commutes. Then Fh : + is a noncontinuous functor. q') are welldefined operations making the set of all equivalence classes into a kdimensional vector space Fh ( V).q' E �k . THEOREM. (f)[v . yn y ' . = L'j�. q] = [w . q) if q = h(A)q' where W. qd + [V . (g) Let Ci: � + � be a noncontinuous homomorphism (compare page 380).) = L'j�. F(V) n After identifying F(�n) with �k . �) = � be h (A) = Ci (det A). and define yn F. (e) Show that the operations [v.132 where A = Chapter 4 (a. h (A)(q)].q . If fl : GL(n. : yn yk (d) For q.
. then are called odd scalar densities. instead. x axl} are called (even) scalar densities. show that x' V a ' = det ( ax' )  ax'. �) take A into multiplication by det A. Let Fh be the functor given by the Theorem. a. in addition to mixed tensor fields. I G:'� ) I . assignments for which (a) Let h : GL(n. and consider the I dimensional bundle Fh (TM) obtained by replacing each fibre Mp with F. assignments of a single function a to each coordinate system such that the function a ' assigned to satisfies x' a ' = det ( ax' ) · a .(V) determines an orientation for V. show that a nonzero element of F. (Xx '. These new rules are of the form A' = A operated on by h ( axa ) ax'P . V) is nonzero. show that the corre sponding equation is a ' = det (c) For this h. other "quantities" which are defined as sets of functions which transform according to yet other rules. so every section on can be expressed as a .(Mp). For example. · a. If is another coordinate system and a · Clx = a t . If is a coordinate system.Tensors 133 10.. Clx for a unique function Q . 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. . . In classical tensor analysis there are. (b) If. �) + GL( I . Conclude that the bundle of odd scalar densities is not trivial if M is not orientable. h takes A into multiplication by 1 det A I.
is an even permutation of I . . (�) (�) j (or the same formula with det(ax i lax. . . ) I). �) + GL(n k +I . . . �) be defined by h(A) The bundle F(TM) is called the bundle of (even) relative tensors of type and weight w. .k(�n) + r. Let iz : GL(II. .k (A ) : r. Vk . 'Ilk (f)(T)(V Given A E GL(IZ. . AI 0 f. we can consider it as a map A : �n + �n.) if ia iJ . . .i" = 8. 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].ill are the com ponents in every coordinate system of a certain contravariant relative tensor of weight 1 . . .j ) replaced by I det(ax i lax. Al 0 f). . · · · . Show that the transformation law for the components of sections of these bundles is = (det A)wr. f(Vk ). AI ) = T(/(vtl. n I. . we define 'Ilk (f) : 'Ilk (V) + 'Ilk (V) by I . . 1 .1 with these compo nents in every coordinate system. (See Problem 712 for a geometric interpretation of these relative tensors.k(A) of GL(IZk+I . �). . . .i n for some Ci oJ {J . �). . Then r. n .134 Chapter 4 k+1 by taking (d) We can identify r/ (ll�n ) with �n Recall that if f : V + V.k(A) w an integer. .. . Also show that gi l . (e) Define if if Show that there i s a covariant relative tensor o f weight . If (det A)W is replaced by I det A l w (w any real number).. . . . . in is an odd permutation of = ip i). .. . . . we obtain the bundle of odd relative tensors oftype and weight w.k(�n) determines an element r. ) " .
CHAPTER 5 VECTOR FIELDS AND DIFFERENTIAL EQ UATIONS e return to a more detailed study of the tangent bundle TM. we wish to introduce a coordinate system (x. we have a vector field x. in general. Recall that. (x.e. Consider the curve The condition c = x op. means that P. (� IJ dp = X(p(I» dl = X(p(I» . 135 . p(O) = P (e. e) P. Cl* X does not make sense for Coo functions a : M + N. It = X(p(I» . However. a curve p with Wsections. . There is a function f : x(U) + � n with i. . if ex is a diffeomorphism. then we define It is not hard to check (Problem I) that a. U) around p and transfer the vector field X to x(U) C �n . i. We would like to know if there is a curve p : through p whose tangent vectors coincide with X.. . X)q has "components" P (q). vector fields. .e. (�IJ = �. fn (q).X is Coo on a(M). and its Let X be a vector field defined in a neigh + M borhood of P E M. Since this a local question.X on x (U) C �n . that is.. In particular.
of course. (� IJ = x. solutions were consequently called "integrals" of the equation. . (X(p(l))) c'(I) = /(c(l» . X)c(t) . cn (l») We also want the "initial conditions" i = l. . Ch(l) = f ( C' (I).P. e) p(O) = dl p <!£ = X(p(l» is called an integral curve for X with initial condition p(O) = p. we will work entirely in Euclidean space. 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. then a curve c : ) + M with V (e. = = (x. 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 /: � + �.e. y. a form to which our particular equations never reduce). Part of this terminology is still preserved. . n . If c �n is open and / : + �n. which may also be considered as a system of II diff erential equations for the functions ci . . etc. e V C'(I) = /(c(l» is called an integral curve for c(0) = x XEU / with initial condition c(O) = x. and for a while x.136 hence Chapter 5 �It = x. to the diff erential equations one obtains upon introducing a coordinate system.X) x(p(t)) (x. A Curve p: + M with (.. . . . Similar ter minology is applied. . For quite sonle time. will denote points of �n. c. .
then (F a y)' = I F(y(x» = + C for some C. the correct solution is for ali i C(I) = 0 .) To obtain the initial conditions c(O) = a.=x+C Y Y Thus the curves are supposed to be solutions. In this case. which would be written classically in terms of a function .Fields and Diff erential Equations 137 Before stating the main theorem about the existence and uniqueness of such integral curves.. 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 . This can be checked directly if you don't believe the above manipulations.Vecto. we consider some special cases. c'(I) = . we must take X I c(1) = This works in all cases except 1 + l /a ' a = O. The equation for a curve C with range R. (They really do make sense. so C(I) = I + C = x + C· I I hence. if F' = I If.[c(l)f. the equation in question asserts that y' = J a y.
For a > 0. i s not differentiable. since = = 0. in fact. written classically as I . can be defined f all I . even or though X is defined on all of � . there is no K with I/(x) . . . and as 1 + 00 it approaches. This will continue to be true even if we modif the vector field near ° so that it is never 0. the curve C (I) = J /(I + J la) is defined for all large I. as 1 + .IU) I :5 Klx . (2) C(I) = � 1 3 27 In this case. y Another phenomenon is illustrated by the equation c' (/ ) = C(I) 2/3 .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. namely . 0. the curves c are (which we missed by dividing by the integral curves of Notice that no integral curve. given by I(a) = a2/3 . Notice that I(a) a2/3 is not Lipschitz.y E U.. except c(/) = 0. In terms of vector fields.y l for all x. . There are two different solutions with the initial condition c(o) dy = y2/3 dx C(I) = ° f all I. We say that the function I satisfies a Lipschitz condition on V if there is some K such that I/(x) . the function I. On the other hand. but never reaches. or (J) for all I. Unique ness will always be insured when I : V + �n is C I .1(0) 1 :5 Kl x l for x near 0.138 Chapter 5 y). but this has nothing to do with the case. . It might be thought that this somehow reflects the fact that X (0) = 0.
so there is some x with Continuity of 0 as n + 00. � (C n + . . a C l function is locally Lipschitz. f then shows that . . +C n+k I + Thus the sequence {xn } is Cauchy. I .y) for all x. � Cp (x. and Jet f : M + M be a "contraction". i.e. p) be a non empty complete metric space. Let (M. y E M. Xn+k ) � P(Xn . On the other hand.rector Fields and Dijferential Equations 139 A Lipschitz function is clearly continuous. lhat is. f(x) = Ixl). + p(Xn+k I . .l )p(XO. it satisfies a Lipschitz condition in a neighborhood of each pointthis follows from Lemma 25. Xn+l) + . A Lipschitz function is also clearly bounded on any bounded set. . . THEOREM (THE CONTRACTION LEMMA). + Cn Since C < ] .. . XI). the sum L::'o e n converges. Xn+k l +k. The basic existence and uniqueness theorem for differential equations de pends on a simple lemma about complete metric spaces. a f(xo). . but not necessarily diff erentiable (for example. so C n + . j(y» Then there i s a unique x E "fixed point"). '"v' n times Thus P(Xn . P ROOF NOlice that f is clearly continuous. 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 ). . Xn+1 Then an easy induction argument shows that = fn (xo) = f a f a . suppose there is some C < I such that p(f(x). that is.
this is basically just the theorem that the uniform limit of continuous functions is continuous.gil. p) is a metric space and X is compact. hence continuous. which are unif or orm limits on [a. I (x) exists since M is complete. 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. . if M is a m n oo compact subset of �n . and thus f continuous functions. thus af = f 0 ex is continuous. (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. then l' <>'(u) du = l' 1(<>('1» duo To prove this. b) + �n and I : defined on some interval around 0. 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. g) = sup p (f(x) . we need only one simple estimate. It I(U) dUI :5 K(b . clearly satisfies V V V.<>(0) = For the proof of the basic theorem.a).!. If a continuous function I: [a. If c + �n is continuous. if <> satisfies (J). then the new metric space is also complete. then we do not even need X to be compact. <>(/) . plus the fact that each ). bJ + �n satisfies 1 /1 :5 K. then the set of all continuous functions I : X + M is complete with the metric (J(f. Conversely. xeX (J (f. hence for step functions. In particular. g(x» . if M is complete. then <> is diff erentiable. then a continuous function <> : (b. so l' 1(<> (u» du. Moreover.1 40 Chapter 5 We are going to apply the Contraction Lemma to certain spaces offunctions. we note that it is true for constant functions. b) of step functions.x = <>(/) . g) = II I . Recall that if (M.
of radius 2a and center xo.x l = < If I(a(u» u l by (3) .xo l :s a. Choose b > 0 so that (I) (2) (3) b :s aIL (4) b < 1 1K. Then + li Sa . Moreover.(u) 1 la(u fJ . is contained in U. it follows that ISa(l) . H Thus S: M M. For each (b. define a curve Sa o n .fJ lI · . a E M. which will be fixed for the remainder ofthe proof + Then M i s a complete metric space.b) + (the integral exists since I is continuous on E2a (xo» continuous.l(fJ(U» dU I by (2) = bKlIa . < :s a Since Ix . Then for each x PROOF.I(Y)I :s Klx . for all I E (b. where c �n is open.y l for x. Now suppose a. Let Xo E and let a > 0 be a number such that the closed ball E2a(xo). Let E Ea (xo) there is a unique ax : (b. b) we have l' I(a(u» du d bL by (I) E2a (xo)}. THEOREM. fJ E M. so Sa(l) E B2a (xo) c E2a (xo) for i E (b. b) V such that a/(I) = I(ax(l» ax(O) = x. Y E E2a (xo). Sa is clearly The curve I Sa(l) .xo l 2a.Vector Fields and Diff erential Equations 141 2 . Choose x E Ea (xo). for any I E (b.b). Let I : + �n b e any function. Suppose that V V V 1 / 1 :s L on E2a (xo) I/(x) .b) by M = { continuous a : Sa(l) = x + (b. b).SfJ l = s�p < bK sup h<u<b If I(a(u») .
Suppose a < b. in B2a(xo). 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. So (J(a) E B2a(xo). is not quite what the theorem states. a : (b. To sum up. is in V satisfying Reason: We claim that any such (J actually lies in B2a(XO). We can now use a simple least upper bound argument. {J : (b. alas. A similar argument works for b < 1 ::s O. This clearly implies that (J (a + E B2a(XO) f or sufficiently small > 0. by (**). So it must be that sup A = b. s s) . the unique fixed point ax of the map S is the unique curve with the desired properties. this shows that Hence S has a unique fixed point: There is a unique S : M + M is a contraction. in fact. b) + This. We have already seen (statement H) that for each 1 with 0 :s 1 < b. < (J (u) E B2a (XO) = {I : O ::s 1 for all u with 0 ::s u ::s t .b) + B2a(xo) with a(l) = x + 10' f(a(u» du. Let A < b and (J (u) E B2a (xo) for O ::s u I } . Consider first numbers t > O. (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. which contradicts the fact that a = sup A. < Let a = sup A .1 42 Chapter 5 Since we chose bK < I (by (4» . We clearly have (J (u) E B2a(xo) for O ::s u < a. Having used the elegant Con traction Lemma. •:.
Let x E V and let a" a2 be two maps or on some open interval I with a l (/).a2 (/) c V and Then a l = a2 on I. •:. Suppose a l (10) = a2 (10) for some 10 E I. that is. (/) . We now revert to the situation in Theorem 2.x» x» . alai or dIdl in this . by Theorem J.x) = x d di a(I. b) x Bp(xo ) + V [i. a/(I) = /(a.(O) = x i = 1 . Dl a(l. 3. Thus the set is open. This remark allows us to extend the uniqueness part of Theorem 2. and have the same initial condition (J.(/) a. x) a(O. (O) = a l (10) = a2 (10) E V. so it equals I.e. Hence {JI (I) = (J 2 (1) for sufficiently small I. We will write ax (I) as a(l. so that we have a map {I E I : ad/) = a2 (1)) a: satisf ying (b. THEOREl'v!. (J'(I) = a'(lo + I) = /(a(lo + I)) = /({J(I)). Ifwe define PROOF. 2. (Jt'(I) = /((J.. 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).x). then the functions {J i satisfy the same differential equation. but we will frequently use /(a(l. if iJ(l) then = a (10 + I). Suppose /: V + �n is locally Lipschitz.vecfor Fields and Dijferential Equations Notice that solutions of the differential equation 143 a' (I) = /(a(I» remain solutions under additive changes of parameter. (lo + I). (Ji (I) = a. It is clearly also closed and nonempty.X) = /(a(l. that is.
This map Ci is called a local flow for f in (b. This map <p. Then .1 44 Chapter 5 discussion] . is always continuous. For each fixed x. we denote it by <p. Let us denote the map S defined in the proof of Theorem 2 by Sx. for a time inteIVal of I. if we fix I. In fact. the whole flow Ci is continuous (as a function of both I and xl: 4. then the map x >+ Ci (I. 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. If y = Cix (/O). the map 1 >+ Ci(I. to indicate explicitly the role of x. b) x Ba(xo) + V given by Theorem 2 is continuous. so the two images overlap. If f : V + �n is locally Lipschitz. then the flow Ci : (b. To picture this map Ci. To focus attention on this map.X) gives the result of pushing each x along the integral curve through it. On the other hand.: <Pdx) = Ci (I .b) x Ba(xo). X) [ = Cix(/)]. PROOF. the best we can do is to draw the images of the integral cUIVes Cix. THEOREM.X) for b < 1 < b lies along part of the curve through x.
you must first learn the elements of Banach spaces. Real and Functional Ana!Jsis (3rd ed. then tJlej/ow a : (b .S. we have . 371379. denotes the nfold iterate of Sy. and then read about diff erential calculus in Banach spaces. then Recall also that in Theorem I the fixed point any Ct . . .SI'l I s bKli a .. .ay ll S 1 Since Ilax ayll = sup la(l. are consequently Coo if f is Coo. . pp.T!ector Fields and Differential Equations Recall that 145 li Sa . besides.Y I · .) Ix . + (b K n. but this is probably easier than reading the classical proof (and. and differential calculus in Banach spaces). If f : V + �n is Ck .axll + . including the inverse and implicit function theorems (Real and Functional Ana!Jsis. In order to read this highpowered proof. so we obtain � Ilax . including the HahnBanach theorem. A clean exposition of the clas sical proof is given in Lang's Introduction to Diff erentiable Manifolds (2nd ed. . when you're fin ished you'll also know about Banach spaces. cxx.a for lIax . this is a very hard theorem.S. 360365).axll 1 ) I S (I + b K + .I'l ll · If S.yl s l _ bK l x .Syax ll + I Syax . this certainly proves continuity of a .axll S lIax . pp. y)l.).: I 1 _ bK l x . and a recently discovered proof can be f ound in Lang. then further smoothness conditions can be proved for a. Hence Cl)' = /� S.). b) x Ba(xo) + V is also C k Unfortunately.b) x Ba (xo) + V satisfies a (O.I ax .x) = x.YI · I ay of Sy is the limit of S. x) a(l. In fact. If additional conditions are placed upon the map /. Since the map a : (b.S. We will just accept this fact. + IIS. Notice that the maps <P.
<p. [If x E Ba/2 (XO). (J (I) = Ci(I. . If we now let <p.(x) = Ci (I. so we can also define y(l) = Ci(I. (x).] So if l s i < e. In other words. on a manifold. in particular.Ci(S.Ci (S. Roughly speaking.X» = Ci(S + I. since it is local. then Ci(I. and x E Ba/2 (xo). (x) E Ba/2 (xo). X).X) E Ba(xo).. We have also noted that satisfies (J(I) = Ci(S + I. III. x) f x E Ba/2 (XO). then the point Ci(S.X). or if I sl. <P'+s = <p. defined f or Is + I I < e.I = <ps.X). we can or ls i. that for ls i < e each <Ps is a diffeomorphism. : Ba/2 (xo) + �n b e <p.X» f 1 1 1 < e. say: (J'(I) = f({J(I» (J (O) = Ci (S. <Ps = <Ps <p. \I\�thout requiring any more proof.e) x Ba/2 (XO) + Ba(xo). then <Ps(<P' (x)) = <ps+. II I. Consequently. can be resaid. Everything we if a a have said.X). X» III < e. l s + l l < e. This shows.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. This satisfies y' (I) = f(Y(I» y(O) = Ci(S. with inverse <Ps . then the integral curve with initial condition x stays in Ba (xo) for III < e. Ci(S. Is + I I < e and x.
= <Pi for E Vi n V . and let p E M. . THEOREl\![. e) x V Isl. THEOREl\![. write or <PI <PI 1 q <Pdq) {<Pi(q) q q <pi (q) (q) q if <PI+S <PI o <ps = 11 1 . if '" support X. Let X be a Coo vector field on M. en and diff eomorphisms <p./2/2 ' <P. .. Notice that by uniqueness.. So j we can define if E Vi = . PROOF. (2). (3) If The examples given previously show that we cannot expect to be defined for all I. " " V given by n Theorem 5 with con'esponding el . <PI: Clearly <p : (e. q E V.0 0 0. In one case however. . = is Coo. III.. Cover support X by a finite number of open sets VI ./2 <P. The support of a vector field X is just the closure of {p E M : Xp oJ OJ.e) x M + M is Coo. and and each is a diffeomorphism. Then there is an open set V containing p and an e > 0. If X has compact support (in particular. 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. To define f II I 2: e. 11 + s l < e. then Xq is the tangent vector at I 0 of the curve I >+ <PI <Pdq). Let e = min(e l . then there are diffeomorphisms M + M for all I E � with properties (I).0 0 = k (ej2) + I" with k an integer. .. . ls + 1 1 + < e./2 <P. en )./2 = 0. (3). . and I I" I < ej2./2 <Pr <Pr [[<P. then q <P q) <P <Pdp). defined by (I. <P <PI { <P. . or on all of M. d E V. p) = .Vector Fields and Dijferentiai Equations 1 47 5.. .:. and M. . Let It is easy to check that this is the desired {<PI }' . if M is compact). iterated k times] iterated k times] for k 2: 0 for k < O. ls l . 6. this can be attained.
alal ' lo ' . . . l\1oreover.(O. THEOREM. ( ".} given by Theorem 6. and is said to be generated by X. say). a'. the map <P.} (vector fields used to be called "infinitesimal transformations"). is called a Iparameter group of diffeomorphisms. and p = 0 E lffi. (/) dl = df (c (I» dt = (/ o c)'(I). or more precisely. Condition (3) in Theorem 5 can be rephrased in terms of the action of Xq on a Coo function f. The vector field X is sometimes called the "infinitesimal generator" of {<P. to say that Xq is the tangent vector at I = amounts to saying that '!!:. . from � to the group of all diffeomorphisms of M. _ o � 0 of the curve t r+ <P. an ). . This equation will be used very frequently.f(q) . Recall that I r+ Thus.n . if q lies on the integral a ax ' on U. M + �. 7. Let X be a Coo vector field on M with X(p) oJ O. Then there is a coordinate system (x.1 48 Chapter 5 The unique collection {<P. 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+_ _ _ . In the local case of Theorem 5. . we obtain a "local I parameter group oflocal diff eomorphisms". U) around p such that X = f (Xf) (q) = Xq / = hlim (<Ph (q) . (q) PROOF It is easy to see that we can assume M = �n (with the standard coor dinate system t I " . . The first use is to derive a corollary of Theorem 5 which allows us to simplif many calculations involving vector y fields.
./(x (a» ] I h /t70 = = /1+0 lim � [/(<Pa l +h (0. . curve through this point. this shows that X. . •:. . 0) . (�l ) a ./(0)] �l hO h Since X(O) = a/al ' l o by assumption. X.I coordinates of q and the time interval it takes the curve to get to q as the first coordinate. an as the last n . an). h . . a2. . . . . a (f) = a ' a (f o X) /1 I . ./(x (a» ] (Xf)( x (a)). O» . . . ./(O)] iI+O II = lim � [/(O. for i > X. .uations q 1 49 We compute that for a = (a' . The second use of the equation = �� I o' (Xf) (p) = lim . . we will use a2. which we have just proved. . ) a l o (f o X Moreover. . . . . .[/(<PI> ( x (a» ) . a2. . . an) = <Pal (0. ./(p)] . . . . To do this. This is the desh'ed coordinate system. an» . .1+ 0 l is more comprehensive. . Jet X generate <PI and consider the map X defined on a neighborhood of 0 in �n by x (a' . . an). a2. . (a/at ' ) = X o X ./(x (a))J h l + h . = Jim _ [/( x (a' + 0 �l = Jim .rector Fields alld Differential 1!. . . an» . 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 . . .I may be used as a coordinate system in a neighborhood of O.O = I is nonsingulaJ: Hence x = X. [/(<PI> (p» . a . h. f it is easy to see that the equation or X. . . . ( � IJ ] we can at least compute ( f) = = lim � [J ( X (O. . . is equivalent to X = a/ax' .
oh The vector field <Ph . . for a : M + N a diffeomorphism and Y a vector field on M. we define a new covariant vector field. We call Lx / the (Lie) derivative of / with respect to X . If a : M + N is a diff eomorphism and Y is a . then A fairly easy direct argument (Problem 8) shows that this limit always exists. Y)p = <Ph' (Y¢_h (P» is obtained by eValuating Y at <P. by (Lxw) (p) = .. To emphasize the fundamental similarity of these notions.. <P. Now if w is a COO covariant vector field. it is another function.' (p) = <P.150 Chapter 5 Lx / obtained in a similar way.w (p)]. integral curve / . but we will Soon compute this vector field explicitly in a coordinate system. I (LxY)(p) = }. I l�o h [(<Ph' W) (P) . we first introduce the notation for Xf. Y)p]..(<Ph . ... and these facts will then be obvious. and then moving it back to p by <p". Recall that if Xp E Mp. we can define the Lie derivative of Y with respect to X. and that the newly defined covariant vector field Lx w is Coo. Thus (<Ph. (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. Y defined at the beginning of the chapter. If Y is another vector field.� [Yp .h (p). Y appearing here is a special case of the vector field a. whose value at p is denoted variously by (Lx f)( p) = Lx / ( p) = ( X/) ( p) = Xp (f). This is the limit of certain members of Mp' . the Lie derivative of w with respect to X.
then a vector field a * Y on M can b e defined by Now notice that Of course. lim (Lx !y)p = 1170 I [(/Y)p .(<Ph* !Y)p]  = lim = h70 . [Yp = lim . Finally. The remaining equations are all proved by the same trick.h h _ /t".Il [f(p)Yp . Nevertheless. If (I) (2) then (5) Lx (w(Y» = (Lx w)(y) + w(Lx Y). 8.'!. then Lx (Y. If Lx Y and Lx w exist. if w(Y) denotes the function p r+ w(p)(Yp) and Lxw and Lx Y exist. LX Yi and Lx Wi exist for i = 1 .k* y)p] Y ] r Yp . We will do number (3) here. (4) Lx ! · w = X! · w + ! · Lx w. (pl] lim = 1170 !(P). we will stick to the original (equivalent) definition. the one used in finding (/g)'(x).(A."ctor Fields and Dijferential Equations vector field on the range 151 N.(<Ph h h� h [ ( ¥h )p .[. + Y2) = LXYI + Lx h Lx (w. a * (Y) is just * Y)p r I r I ' *y A. 2. LxY. (P) · + h�O !(P) . [f(p)Yp .!. The cal culation is made a lot easier by first observing (a .I ) * y. + W2) = Lx WI + Lx W2 .<Ph* (/Y).(<Pk.!(<Ph (p» <Ph* Y¢_I. '¥ = k�O k [Yp . Y r 'f'!t* ¢I. [Yp . We now wish to compute Lx w and Lx Y in a coordinate system. PROOF (I) and (2) are tri\�al.p = h�O !.O l lim _ h I .Y)p] = (LxY) (p).<Ph* Y¢_h (pl] h !(<P (P» A.(pl] [ ] . k7o k . PROPOSITION. then (3) Lx!Y = X! · Y + ! .
the term in brackets = while an easy argument shows that </IMY¢_h(P ) k7 0 lim f(p) .</10) (P) ax' (x' 0 </10)] .152 The first limit is clearly approaches Chapter 5 f(p) · Lx Y (p). .(p) dx (p) . lim � a(x' O </Ih) h ax' . . V) on M. where y is x.p /t". O · a I lim . I a( �x�f) dxi We can apply this to </Ih * .dx' (p) [ . thus a2 A /ahax j = a2 A /ax j ah .P X( = . [ _ 8' J (*) ] = lim = It.) . • + : We are now ready to compute Lx in terms of a coordinate system (x.  .j=1 . q) = X ' (</Ih (q» is Coo from � x M y to �. a .] a(x' 0 .We first compute Lx (dx').f(</Idp»  k Xf(p).o h [( 0 aXl.dx' (p)] = lim Now the coefficient of dx j (p) is iJ"'O I � a(x' O </lh ) ' h L aXl· /t7 0 . n 8ai . Suppose X = L:7= . a'a/ax'. . = To justif *) we note that the map A (h . Then Lx (dx')(p) = l�o h [(</Ih*)dx' (p) . then f* (dy') = t j =1 . which is what the interchange of limits amounts to. aXl.</Ih) (p) h ax' [ _ ] {this step will be justified in a moment} . In the second limit. but .x' </1.. Recall (Prob lem 41 ) that if f : M + N and y is a coordinate system on N. It now follows that ( a aa' x') . . I Xl I � a(x' O ... Lx dx' = I: axj dx ' . j =1 We could now use (2) and (4) of Proposition 8 to compute Lxw in general.(p). Yp .
. aaj ) axj ' Laxl Lx This somewhat complicated expression immediately leads to a much simpler coordinatefi·ee expression f or Y.. YJ. Clearly x = L a'. .) .' . axj ) = Lxb'.L. and we can now use (5) to get the answer immediately. T!ector Fields h* on vector fields involves one more composition than <Ph* on covariant vector fields.=1 a b � (� a'.aax.· axj "1 axj ax' .L b' . aa' a = L a'. also denoted by [X. and (5) of Proposition 8. = . a a Lx (b'. If j . because <P Lx oj = Lx [dX' ( a!j )] (Lx dx') ( a!j ) + dx' (Lx a!j ) so aa' a dx' (Lx.= . The trick needed to deal with this complication has already been used to prove (3).· ax' axj axj ax' 0= . and we find that LxY = XY .YX. so XYj = X(Yf) makes sense..j .a The second partial derivatives which arise here cancel those in the expression for Y(Xj). axj + b'Lx ( axj ) n .=1 axI n . .. then Yj is a function. = . To compute Lx ( alax' ) we could imitate the calculations of Lx dx'. ax' ax' thus. abj a n . M + N is a Coo function. a n aa' a Lx . but there would be a complication.and Differential Equations 153 we are really interested in computing LxY.b'.. Summing over j and then interchanging and obtain '�I i j in the second double sum we '�I LXY = L j=1 .= Using (3) we obtain a . (4).
p) al & (0. so that [X. A straightforward verification shows that [X. p) = Ig(l. Y] have both been defined independently of any coordinate system. : .154 Chapter 5 [X. in this case the coordinatefree proof is short.YX.e) + � with 1(0) = 0 can be written 1(1) = lg(l) for a Coo function g : (e. p) = 10 as I(SI. (0). Two vector fields Lx Y and [X. Y]p is a derivation at p. Y]p(g) + g(p)[X. namely g(l) = l' /.e) x M + � is Coo and I(O. though hardly obvious. PROOF. 9. Y]p(/) Often. Define g(l. pl . note that this means = Xp (Yf) . e) x M + � with = 0 for all p E M. We are now in a very strange situation. e) + � with g(O) = /. Y]p(/). p) ds r' a .(sl) ds. Fortunately. This sort of thing irks some people to no end. Y] (which is called the "bracket" of X and Y) is just defined as XY . but they have Gcen proved equal using a coordinate system. •. This has an immediate generalization. p) = g(O.p) then there is a Coo function g : (e. 1(1. Y]p ( /g ) = l(p)[X. and can theref be considered as a member ore of Mp. If I : ( e. In Chapter 3 we proved a lemma which for the special case of � says that a Coo function I: (e. [X.Yp (Xf). LEMMA.
[.Yp(Xf). X]. diately that L is also linear with respect to X: LxY = LyX. Y]. Z] = [Lx Y. so sO [X. Y] = XY ."ctor Fields aud Differential Equations 155 Y are Coo vector fields. This equation is capable of two interpretations in terms of Lie derivatives: (a) Lx [Y. (P) (f a <Ph) = Y¢_I. Clearly [X. Lax. [Lx. + 1 0. Let X generate <p" II I < e. Yll + [Y. *:. Consequently.(<Ph* Y)P](f) = h70 � [(Yf)(p) . Z] + [Y. Y. = f + lg. there is a f amily of Coo functions g. [Y. then LxY = [X.Ly a Lx (which might be written as . it follows imme Since we obviously have Lx (aY. go = Xf. [Z. PROOF.[Yp . [X. Ly D. Lx X = O. THEOREl\I[. Y = aLx.(P» (f) = Y¢_I. Lx Y The equality Lx Y = [X. Y] = [Y. Xll = o. we have L[x. +bX. Y + bLx. on M such that f a <P. Lx Z].YX reveals certain f acts about which are by no means obvious from the definition. + bY2 ) = aLx Y.·ators on Coo functions. a straightforward calculation proves the 'Jacobi identity": [X.YI = Lx a Ly . + bLx Y2. If X and By Lemma 9 Then (<Ph* Y)p(f) = <Ph* (Y¢_I.l�o (YghH<Ph (p» = (LxYf) (p) .(P)(f + hgh ). (b) as ope. Finally.(YgO) (p) = Xp(Yf) .(YfH <Ph . X] = o. Let f: M � be Coo. Zll + [Z. so I (p» ] Jim lim h /1 h70 .
[X. Y]. Recall that (<Ph. but on the vector fields X and Y. ] is no t a tensorthat is. If X generates <Pt. 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¢_. not over the Coo functions F. or a simple calculation using the definition of [X. note that Lx Y is linear over constants only. In f act. to the curve But this tangent vector is just Xp. (p). even if Xp = 0. Proposition 8.156 Chapter 5 Finally. Y]. (I)  LxX = 0. . it certainly suffices to show that «Pt. for reasons that \vill become more and more apparent. although not a tensor.. The bracket [X. y(t) = <P.. Thus </>.* X¢_Ii (p) is the tangent vector. h(p). for Xf may have a nonzero derivative in the Yp direction even though (Xf) (p) = o. at time I = 0. Thus.X)p = Xp for all h . X)p = <Ph. pops up in the definition of prac ticaJly all other tensors. Y] + f(Xg)Y .(p) . Now X¢_. at time I = 0. it does not necessarily follow that [X... gY] = fg[X. and thus the tangent vector. Y]p(f) = Xp(Yf) Yp(Xf) the first term Xp(Yf) is zero. but the second may not be. Bef ore proceeding to examine its geometric interpretation. YJ.g(Yf)X. Y]p = Oin the formula [X. the bracket operation [ . shows that [fX. 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?). In particular.(p) is just the tangent vector at p <Ph (P) time t = h to the curve t >+ to the curve <P.
Y [X. X = X ifand only if <PI oa = a o<PI Then PROOF. p E M. [ f(a 0 <Ph 0 a 11". conversely.X)q (/) = [a. Y] = 0. Y generate {"'d. LEMMA. PROOF We have (a.(/ 0 a) (a. LEMMA. then a. e) + Mp given by . then LxY(p) = O. an integral curve starting at some other point can never get to p). !. suppose that [X.l. Y] = 0 if 1 2. Then a.I (q» ] . of course. + M. 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. or Conversely. [X. consider the curve c : (e. If this by Corollary f all or Given q. Y] we first prove two lemmas. X generates {a 0 <PI 0 a . I (q)) . If <PI 0 "'S '" "'s 0 <PI for all s . iI70 1 = lim ./(q)] : 1 2.T4!ctor Fields and Differential Equations 157 Since Xp = 0. To develop an interpretation of I I .I }. Then Yp = 0 and (2) If Xp and Yp are both 0.0 h I . If a : M for all I.I (q))) . so that 1 3. so Lx Y(p) = O. . Let a : M + N be a diffeomorphism and X a vector field on M which generates {<PI}. •. then <p Y '" " is true f all t. COROLLARY.I (q ) (/ o a) = lim _ [(/ 0 a)(<p" (a .X.I (q)](f) = X. . Let X generate {<Pd and and only if <PI 0 "'S = "'s 0 <PI for all s. then clearly Lx Y = O.
• . . k. l . Consequently c(/) = alJ s. . so <p" Y = Y. . Y)¢. ex :::::: PROOF. = 1jJ. c' (/). Xk are linearly independent Coo vector fields in a neighborhood of p. 1jJ. then there is a coordinate system (x.C(I)] . •. . so there is no hope offinding a coordinate system satisf ying (*) unless [X. Xp] = ° for I � ex. k.. I = l'!:'o h [<P. If Y is another vector field. everywhere linearly independent of X. for the existence of the desired coordinate system. of this map into the vector space Mp we have h_O + h) . then there is a coordinate system x with X = a/ax'. : c(O). <p. . . then we might expect to find a coordinate system with H a X = ax " However. V) around p such that a Xa = axa on V. The remarkable f act is that the condition [X. Y] = 0. that = I . that ex ]. by a linear change ofcoordinates. as well as necessary.<p" Y¢_. and. (p)] " = <P.. 1 2. THEOREl\t[. [ a�' ' a�2 ] = a Y = ax2 • 14. As in the proof of Theorem 7. for a a We have already shown that if X(p) oJ 0. (<Ph. a short calculation immediately gives the result 0.fJ � k. (p) . If X" . Y] = ° is s'!!ficient. and [X. . (O) = <p {l�o X [(<Ph. <P.Y¢. we can assume that p = 0. (p) . Y)¢_. . • . (P)] } using (*) with q = <Pd p) By Corollary = 0.158 Chapter 5 I h c' (/) = lim . M = �n . .[c(/ For the derivative.
. just as before we see that (� I ) � I 0 = 1 X. . . . . . it shows that f each between 1 and the map X can or also be written a ax l ' Ci k. . XI = Nothing said so f uses the hypothesis [X" Xp] = O. » . If X and Y are two vector fields in a neighborhood of p. ( 3) then follow the integral curve of backwards for time h. .( . . . . . . . follow the integral curve of Y for time h. n. Xa = � . . x ( a l . . an) = 4>� (4). p = O. . we ar appeal to Lemma 1 3. in some sense. . we can compute that .. There is a more complicated. . . To make use of it. X (2) starting from that point. . (4):. (4)�. generales {4>f } ' define X by I x ( a l . ak +I . . Ci = k + I . . and less important result. . (0) = a . . . ax' •:. more difficult to Pl"Ove. al' Thus x = X I can be used as a coordinate system in a neighborhood of Moreover. . o al' 0 � l Ci = . . an» .0. ( . an) = 4>�. (0.vector Fields alld Diff erential Equations If 159 X. As in the proof of Theorem 7. Y] measures. . . » and our previous argument then shows that . .0. an) . (0. . ak + . y X y X (4) then follow the integral curve of Y backwards for time h. . . x . which makes this assertion much more precise. the extent to which the integral curves of X and Y can be used to form the "coordinate lines" of a coordinate system. . . then for sufficiently small h we can (I) follow the integral curve of through p for time h. . . .k I. We thus see that the bracket [X. .
Yj oJ 0. h) = 1fJd<Ph (1fJd<Ph (p» » . . . O.1) = a. (I. PROPOSITION.0. . . h. that is.O. If we define 1 5. so that this "parallelogram" is always closed. .O.(I. . c is the constant curve p up to first order. .0).0.h) = 1fJd<Ph (P» a2 (I.0) (O. a2 (0. O) (4) (0. the parallelogram is "closed up to first order".O. 0) (h. Then the CUlve PROOF. .d1fJh (<Ph (p» ) a. c'(O) = O.I) = a2 (1. h) = <P. Even when X and Y are Oinearly independent) vector fields with [X. .1) (b) . which means that c' (O) = y'(O)] is the following.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. .1) a. then Moreover. The meaning of this phrase [an extension of the terminology "c = y up to first order at 0". . . Let c(h) be the point which step (4) ends up at. .I). h . (1.(O. . C (I) = a. (a) a dl. . .
but clever.Xf(p) + Yf(p) + Xf(p) = o e" (O) (/) = (/ a e)" (O). and a second addendum concerning linearly independent vector fields in dimension 2. and (f) give a (/ 0 (') ' (0) = Yf(p) . 0) + [Ddf a (2)(0. e" ( 0) E Mp. but can easily be skipped. is an horrendous. cal culation. repeated use of the chain rule gives using (b) using (a). (c). : . 0)] = DJ (/ a. Y]p is related to this "second order" derivation.h» .)(O. 0) + DJ (/ a (2) (0.)(O. that this operator e" (0) is a derivation. 0) + D2(/ a (2) (0. t . It is followed by an addendum containing some additional important points about diff erential equations which are used later. a(/ o aJ ) at = Yf o aJ a(/ a (2) = Xf o a 2 t . (A more general construction is presented in Prob lem 1 7. . 0) + D2 (/ a a.O)] a Consequently. O) + D2 (/ a aJ) (O.al l aa.h) = Xf(adO.) It turns out that for the curve e defined previously.aa(/ a. (d). which ends the chapter.)(O. 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.T4!ctor helds and Differential Equations and for any (c) (d) (e) while 161 Coo function f : M + �. the bracket [X. (e). •. Thus. The proof. using the assumption e' (0) = 0.) = Yf o a.)(O. h (/ a c)' (0) = DJ (/ a a.aa (f) a(/.(O. 0) + [Ddf a J)(O. we can define a new vector e"(O) or d2e/dt 2 l o by A simple calculation shows. 0) = DJ (/ a a.e) + M with e(O) = p and e'(O) = 0 E Mp.
2(/ 0 CY..J (/ 0 CY.)(O.2)(0.O) = Y Yf(p) Now (J) by (e) by (e). Chapter 5 c"(O) = 2[X.O) + D2( Xf 0 CY. since (/ 0 C) (I) we have H (/ 0 cJ" (0) = DJ.)(O. THEOREM.2)(0.2 (/ 0 CY.O) + D2.2 D2 (Yf 0 CY..2[DJ (Xf 0 CY.2)(0. 0)] by (d) and the chain rule + D2.O)..0) + 2D2 (Xf 0 CY. 0) + D2. 0) = DJ.2 )(0.2) (0. Y]p .2) (0.2XXf(p) + D2.. DJ. 0)] = 2XYf(p) .2[DJ (Yf 0 CY.)(O.)(O. = = = .O) = 2DJ (Yf o CY.0) + D2. J (/ 0 CY. 0) + 2D2.2YXf(p) ..162 J 6.0) DJ(Xf 0 CY.2(/ 0 CY.2)(0.)(O.. Since (b) gives we have (3) D2.)(O.0) = 2XYf(p) . O) + D2 (Yf 0 CY. ) (I.J)(O..df o CY.2 (/ 0 CY.2(/ 0 CY.0) by (c) and (f). I) PROOF.) = 2[DJ (Yf 0 CY. 2 (/ 0 CY. (2) 2D2. 0) + 2D2. 0) XXf(p) .2) (0.2)(0.J)(O. 0)] = 2XYf(p) . 0) by (d) XXf(p) .) (O.J (/ 0 CY.J) (O.J)(O. O) = Dd. 0) + D2 (Yf 0 CY. Using the notation of the previous proof.2YYf(p) .2) (0.df 0 CY.J (/ 0 CY.2XYf(p) by (e) by (b) and the chain rule by (d) by (a) and the chain rule by (c) and (f). = (/0 CY.2)(0...Yf o CY.
T4!ctor Fields and Diff erential Equalions Finally. . .1 (/ 0 CY(1)(0. 0) + 2D2. 0) = YYf(p) + 2XYf(p) + XXf(p) . by (c) and (f). 0) == Substituting (J)(4) in H yields the theorem.:. D1 J (f 0 CY(1)(0. from 163 we have (4) D2.2 (/ 0 CY(1)(0.0) + D2.2 (/ 0 CY(2)(0.
for some to. x) = ( I . x) = (s. a a ata(I. x).X) = X.x» a (O. 2) = a : (b. s. a _. /(S. One way t o d o this i s t o replace /(<> (1. we would like to solve equations by where / : (c.x» wherever it occurs in the proof. X) = /(<>(I. Then there is a flow I(s. well have required.c) x V + �n+l i<> (I.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) = x. For the first component function & 1 this means that .x) = s.X» <>(O.<>(I.x) = /(I. S.s.X» al <>(O. Define I: (c. at<> (I.lo. There is also a clever trick. X) = /(a(l. b) x W + � X �n with (a' .<>(I. 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). To prove this. In general.x» .x) = x. one can replace 0 by 10 everywhere in the proof of Theorem 2.X).x) = 1 a' (O.s. S.c) x V + �n . let alone equations like <>' (1) = 1<>(1). or else just replace <> by 1 >+ <> (1 . that <>(IO. x» by /(I.
& 2 (I. PROPOSITION. x) = /(&(I.x) = s + 1. Finally.!' thus 165 For the second component & 2 w e have &' (I. .X» = /(& ' (I. X» . x) '" (s. S. consider the special case of a linear diff erential equation a'(I) = g(I) . then the solutions of the equation a'(I ) = g(l) · a(l) can all be defined on (a. Then is the desired flow with a ii/{J (I. na t by considering the curve 1 r+ (J (I . X) = g(I) ' x. a 2 ii/& (I.r�ctar Field.&2 (I.X) = X. x) = x (by first finding & with & (lo..S. a(I). b). s. 1 7. distinguishing them from general differential equations a'(I) = /(I.S.x). If g is a continuous n x n matrixvalued function on (a. This remark allows us to prove an important property of linear diff erential equations. {J (I. where g is an 11 x 11 matrixvalued function on (a.X» . b) x �n + �n is Coo. Of course.X» (J (O. In this case If c is any n x n (constant) matrix. x).lo. x) = /(I. b). we could also have arranged for {J (10. alld Diff erential Equalion. even if / : (a.a(I)). /(I.S. 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. which may have solutions defined only on a small time interval. S.X» '" /(s + 1.s.b).
Similarly. b) we can solve the equation. (J'(I) = g (I ) . Extend it as far as possible. Ci must be defined for all I with a < I :s 10 . a contradiction. 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. By uniqueness.166 Chapter 5 PROOF Notice that continuity of g implies that /(I. (J coincides with Ci on the interval where they are defined.) "1 0. or close enough to I. be the least upper bound of the set of I 's for which it is defined. 1 hus Ci may be extended past I) as c . •: . with any given initial condition. f I near I. . �. let I.X) = g(l) . c . (J(I) (J(I. in a neighborhood of 10. x is locally Lipschitz. So for any 10 E ( a.
for we might not have [X. X2 be linearly independent vector fields in a neighbol·hood of a point p in a 2dimensional manifold M.. . 0). . whose parameter cUn"S in the i th direction are the integral curves of the X. whose i th parameter lines lie along the integral curves of X. Let XI . l. A simple example (Problem 20) shows that even this modest hope cannot be fulfilled in dimension 3. We can assume that p = 0 E �2. we know that there is usually no immersion f: V + M with p E f(U). . but have diff erent parameterizations. Similarly.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. .. . 2 X l (q) The map q >+ (X l (q). Every point q in a sufficiently small neighborhood of 0 is on a unique integral curve of XI thl"Ough a point (0. . PROOF. where V c �2 is open and p E f(U).X2 (q» we proved precisely this fact in Theorem 7.. the curve I t+ / (a} . . .)o. X 1 = o. . .. in the special case of dimension 2. and that X. However... (O) = (e. aj_ l . Xn defined in a neighborhood of p E M and linearly independent at p . . q is on a unique integral curve of X through a point (X l (q). On the other hand. . . in a sufficiently small neighborhood of 0. with Jacobian equal to I at 0 (these facts also follow from the proof of Theorem 7). Then there is an imbedding f : V + M. . Given 11 vector fields Xl . •:. an ) is called a parameter curve in the j th direction. PROPOSITION. such an imbedding can be found: 18. aj+l .x2 (q» is Coo.. 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. . Its inverse. is the required diff eomorphism .
we can further arrange that crt) maps to (c(t). we can state 1 9 .168 Chapter 5 We can always compose f with a map of the form (x. or c . . and f which f(t. then the map (x. whose i th parameter lines lie along the integral curves of Xi . C C � 2 is the graph of a monotone function g. y) r+ (c. Then there is an imbedding f : V + M. crt»�. or X2 . where V c �2 is open and p E f(V). g(y» takes the diagonal {(x. and let c be a curve in M with c(O) = p and c'(t) never a multiple of X. Consequently. x)} to C. which gives us considerable flexibility. t) = crt). PROPOSITION. y) r+ (x. If. y). Moreover. f any particular or parameterization c = (c" (2 ) : � + � 2 of C. Let X" X2 be linearly independent vector fields in a neighborhood of a point p in a 2dimensional manifold M.y) r+ (a(x). by composing with (x. for example.' (x). tl(y» for diffeomorphisms a and tl of �.
y) E W there is a unique Ci = Ci (x. X is a Coo vector field on N. x. For example. y) (t) = Ci(I. but f has no fixed point. X is not a Coo vector field. Yo) E V x V. X» Ci(O. then Ci. 4. the equation Ci'(I) = yCi(I) Ci(O) = X. p) and a function f : M + M such that p(f(x).Ci' (I» Ci(O) = X Ci'(O) = y. { Hinl: Consider the system of equations Moreover.x) : (b. then Ci: (b. Find an example ofa complete metric space (M. x) = x. b) + V with Ci'(I) E V for I E (b. 2. then there is a Coo vector field X on � such that Ci. 3. where f: ( e. y. y E M. Ci(I). Let f : (c. Find a nowhere 0 vector field on � such that all integral curves can be defined only on some interval around O. y. y. x) = f(l.b) x W + V is Coo. Prove that there is a neighborhood W of (xo. b) + V for each initial condition x and "parameter" y. . and we are solving for Ci (y. and X is a Coo vector field on M. (b) If Ci : M + N is a diffeomorphism. 5.c) x V x V + �n be Coo. y) for all X. for open V c �n and V c �m. c) x V x V + �n . V C �n are open. y. (a) If Ci : M + N is Coo. {J (I» .[ector Fields and Differential Equalions 169 PROBLEMS 1. Yo) and a num ber b > 0 such that for each (x. and let (xo. f(y» < p(x. (c) If Ci : � + � is Ci(I) = 1 3 . y).y) : (b. if we write Ci (x . VYe sometimes have to solve equations "depending on parameters") a ii/Ci(I. where V. Ci' (I) = (J(I) (J ' (I) = f(I. Ci(I. : TM + TN is Coo. then Ci. Ci(I).b) and Ci"(I) = f(l.
! is C ' . x) = !(a(l. y. then D. y. x) = (y. [When one proves that a C k function ! : V + �n has a C k flow a : (b. (b) Show that equations of the form (**) a ai"a(l. a(l.. a(l. x» · D2a(I. x) = (O. it follows that D2 a is differentiable if D.] D.x) = !(I.a(l. x) = (y. x» a(O.a of the fOl"m (**).x» .x) = D2/(a(l. D. the hard part is to prove that if ! is C ' . Since (***) is an equation for D. so that a ai"a(l. xo).170 with solution is such a case.j(I. c) x V x V + � m X �n show that we can write for some a. y. and conclude that a satisfies (*).x» . and that if the derivative with respect to these arguments is denoted by D2a. y. x) = !(I.x» . a (l. x» a(O. if ! is C 2 Diff erentiability of class C k is then proved similarly. y. by induction. D2a(l. X) (a result which follows directly from the original equation if ! is C 2 .).e. ultimately).a(l.x). W + �m X �n is a flow for j in a neighborhood of ( c . x» j(t.x) = x can be reduced to equations of the form (*) (and thus to equations a ai"a(l. y.b) x W + V. then a is diff erentiable with respect to the arguments in W.b) x (a' . y. since D. = D2D. x) = !(a(l. (a) Define by If Chapter 5 a(l) = xeyt.x. i. a2) = a : (Yo. j: (b.
(e) If AB = eA "" I + A + 2! + 3! + 4! + ..ons 6. (b) Let A = (aij ) be an n x n matrix.A) = I . p=O =O 2N (A + BJ P ( = (exp A) (exp B).) . (d) Show that exp(TA T. then exp(A + B) Hint: Write and show that I R N I + 0 as N + 00. and let IA I denote the maximum of all l aij I. (f) (exp A) (exp . the usual norm of A E = B (= exp(O) . p. 2 2 (g) The map exp. ill and uniformly in any bounded sel. so exp A is always invertible.[eclor Fields and Dijjerenl. 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. A2 A' A4 j)th = BA. . where g: � + �.al Equal. . 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). entry of the partial sums converge absolutely for each (i. considered as a map cxp : �n + �n . Show Ct(I) = ef g(r)dr . where f g(l) dl denotes some function G with G'(I) = g (one can obtain all positive multiples simply by changing G). B).= p i L i L p. p=O p. The remainder of this problem inves tigates the extent to which similar results hold for a system of linear diff erential equations. so that we are solving for a realvalued function that all solutions are multiples of Ct.' ) = T (exp A)T' . �n2 . is clearly differentiable (it is even analytic). Show that exp' (O) (B) L . N AP )( N BP ) + RN (Notice that for I A I.
axi is a Coo function on M x N. �n + M. let 1( · . show that B'(t) = A'(t) · exp (A (t» . q) denote the function from M to � defined by 7. t. provided that A(t)A'(t) = A'(t)A(t). t . defined by al a(f( · . U) is a coordinate system on M. If (x. For a Coo function I : M x and q E N. and let (i) Let A � + �n b e diff : B(t) = exp(A (t» . the limit . ( a/at ') = X 0 x. show that for every Coo function I : M + �.) (fo' g(s) dS ) 8. q) = � (p). q).) (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. (a) Let M and N be Coo manifolds. Check that if the coordinate system x is x = XI . (This is clealiy true if A(s)A(t) = A ( t ) A (s) for all S. then N + � p >+ I( p . q » (P . show that the function al/axi. for x : X= a/ax' is equivalent 10 X. 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.172 Chapter 5 (h) Use the limit established in pan (g) to show that exp' (A)(B) = exp(A) · B if A B = BA . (b) If <p : (B. 2 erentiable. If B'(t) denotes the matrix whose enllies are the derivatives of the entries of B.B) x M + M is a I parameter group of diff eomorphisms. (This certainly happens when g(t) is a constant matrix A .
Give the argument to show that tion 8. isjust a special case derived in an The formula for Lx dx .Hxtor Fields and Diff erential Equations exists. (a) Show that [Lx df(p)](Yp) = Yp(Lx f). In the next part we get a much simpler proof that Lx Y = [X. the limit 4>* (1. derived in the text. x TM + TM is defined by (c) If 4>* : 1 73 (e. Lx that for Y E Mp.. and f: M + IR a Coo function. If X generates {4>. e) show that 4>* is Coo. }. define Lx df = d(Lx f). 9. X and covari exists and defines a Coo function on (d) Treat Lx Y similarly. (a) Prove that Lx(f · w) = Xf · w + f · Lxw Lx[w(Y)] = (Lx w)(Y) + w(LxY). v) = 4>. unnecessarily clumsy way. i . and conclude that 4> * (df)(Y) = Y(f 0 4» . 4>h* Y"'_h(P) + Yp in the proof of Proposi 10. and conclude that for every Coo vector field ant vector field w on M. using the technique which appeared in the proof of Proposi tion 15. (b) How would Proposition as 8 have to be changed ifwe had defined (Lx Y )(p) (b) Using (a). and defines a Coo function on M. show directly from the definition of II. (v) . (c) Let X and Y be vector fields on M. M. Y].
. Z be the vector fields a X = Zay a = Lx Y(p)(f) = [X. h) we have . then (�) [4>*A(P)](VI . A[) = A (</>(P»(</>* V I . we define M generates {</>.1 ) * A I > .c' (O) 1 2. (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. .A(p)]. . Vk . . . . 13.* A)(p) .. . . (</>. A[ E Mp* . . . .Xp( Yf) D2a(0. (r l ) *A[). If V I . . r+ (a . Check the Jacobi identity. . . A I > . . Y. . on M. b . 0) = Yp(Xf). and A I . On 1R3 let X. 1 . Y]p(f). </>* Vk . . we define </> * A on M as follows. ). (b) Show that the flow of aX + b Y + cZ is a rotation of about some axis through o. . . Z=y ax ay (a) Show that the map a Y = z. and A is a tensor field of type (J) (�)J I (LxA)(p) = "_0 _ [(</>. lim . c) E 1R3 1R3) 1R3 14. .1 74 Show that Chapter 5 Dl a(O. . Yilz a az a a . . V] r+ the crossproduct of the images of U and V. O) = . Vk E Mp..+ x ax a X + b Y + cZ is an isomorphism (from a certain set of vector fields to and that [U. If A is a tensor field of type on N and </> : M + N is a diffeomorphism.x. Conclude that for c ( h ) = a(h.
. ." A" . LxXi.•laJ11a+J ·'·'k . . . .. WI) + L A (XI . ._ I . V. A + Lx. . . . . . .. .lleclor Field. . . . Ap_" A. . Lx(CA) = C(Lx A). . X cr LL 0'=1 . W I . aai . .! ! !cr A' J ···.. . ® .=1 I . (c) Show that Lx (A + B) Lx{fA) = X{f)A + jLxA). .). . . . WI ) can be obtained by applying contrac tions repeatedly to A ® X. • .. WI») = (e) Noting that A (X" . Xk .. A/ _. . x . = LxA + LxB Lx.=1 n . . A � J .lk' = l . use (d) to show that (LxA)(X" . . . k I . . . (�).+ I . Lx Wi . . . V. Xk . . A /_ ' ) (v . . . (CT)(v" Show that = contraction of . ® Xk ® W I ® . . ..lk . . WI). (f) If A has components Ai ' in a coordinate system and X = L . . a axj n ai a/axi . .J l}cr+J . . . Vk_" A" .=1 . WI . . . Xk . W I . Hint: We already know that i t is true for A o f type (d) Let be any contraction (�). V. . . A . A) f+ T(V I . . wI) '':':: a " ai� " " LL L ax' cr=l j=1 . . . h � (LxA)IJJ . j n .. . • . . 'J / .=1 + L A ( X1 . .. . A = Lx. . W I .. . . . . Xk .' and Diflmlliiai Equations Show that 175 Lx (A ® B) = (LxA) ® B + A ® Lx B (so that in particular). Vk .=1 show that the coordinates of Lx A are given by k n . W" . J ' a i + " " Al J. . . Xk . .. . . Lx(A (XI . +x. . (�). . .!t _ ! ··. . . Ap+ " . . . ® w[. . .
DC = CD. Show that if we define DAI = 0. then this unique extension is Lx . Hillt: Check this for functions and vector fields first. (c) Show that ( DA W) ( P) = A ( p) * (w ( p» . Show ./I ) for t ::: o.) (b) Given c : IR + M with c'(O) = 0 E Mp.h. DA X = A (X).!� g + (h) . define y(t) = e ( . such that (J) (2) D ( A (I) D is linear over ® B) = DA ® B + A ® DB IR (3) for any contraction C. then DA has a un�'1ue extension satisfying (I). Show that the tangent vector e"(O) defined by e"(O)(/) = (1 0 e)"(O) can also be described by e"(O) = 2y'(0). Let D be an operator taking the Coo functions F to F. so that we can consider A ( p ) E Elld(Mp).1 76 Chapter 5 15. (a) Let I : IR + IR satisfy 1'(0) = o. (a) Show that D has a unique extemion to an operator taking tensor fields of type to themselves. Define g (t) = that the righthand derivative (�). show that Generalize to tensors of type 16.g(O) h  _ /. (d) Show that LJx = ILx  DXfj!!dJ . then A (X) is a vector field for each vector field X. (e) If T is of type (�). 0») Let A be a tensor field oft)l)e G).. 2 (Use Taylor's Theorem . Y. 1 (./1 ) for t ::: o. _ r g+ (0) . such that D : F + F and D : V + V are linear over IR and D U Y) = I . If we take DI = XI and D Y = Lx Y. (2). DY + DI . and (3). . (0) . and the Coo vector fields V to V.
show that where [X. If with = and e"(O) defined by e"(O)(f) = (f 0 e)"(O). (a) If M is compact and is a regular value of I : M + IR. Y]p(f) conclude that it is welldefined. : Mp Mp + NJ(p) x ( (( is a welldefined bilinear map. 1 0) to the tangent vector 18. 0 0 11 (0) . choose vector fields X. . so that I. x is a coordinate system 0(/2) xi (e(/» ai/ 2 + 0(/ 2). (b) Show that I •• n n (LI axa i Ip ' j�1n . show that 1 (Xp. Yp E Mp. Y]p = n a f.L1 L j= "= a' . axi Ip ' a '. and b J if} a' b J . IR is independent of the coordinate system.I (U) is diffeomorphic to x u. Given Xp. For Xp.leclor Fields and DijjerClltiai Equations 1 77 1 7.p = O. then there is a neighborhood U of E IR such that I. (a) Let vectors Define I : M + IR have as a critical point. Yp) •• is symmetric. be the curve of Theorems 15 and 16. Yp E M and g : N + I (X. Let around pe x (p) 0. = denotes a function such that 19. (e) If IR + M has as a critical point. = 0. show that e: 0 c•• (0) : 1R0 x 1R0 + Me(o) takes ( 1 0 . Y) g ) = 51p Y g 0 I»· I... xa IP) = i. p Using the fact that [X. Y with 51p = Xp and Yp = Yp. (c) The rank of (d) Let I : M + define Show that (a2fjaxi axj (p» N p have •• a2I axi aXj (p). as a critical point.
20. If I is onto N. then there is a neighborhood U of q and a diffeomorphism </> : II (q) x U + II (U) with I(</>(p.and zaxes. 0») More generally. as shown in the first part of the figure below. respectively. Using the second part of the figure. In IR ' .q')) = q'. while certain other integral curves arc parabolas in the planes y = constant.178 Cha pter 5 by a diff omorphism </> : 1 .I (q) x N? Z b e unit vector fields always pointing along the y . 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. I)) = I. then II (qI l and I. show that Proposition 18 does not hold in dimension 3 . (c) It follows f rom (b) that if all points of N are regular values. q2 sufficiently close. does it follow that M is diff eomorphic to I. if M is compact and q E N is a regular value of I: M + N.X = dldl. and let X will be a vector field one of whose integral curves is the xaxis.I (q2 ) are diffeomorphic for q l .1 (0) x U + II (U) with 1(</>(p. let Y and .
The function /. even n the previous chapter. with p E N. is called a I dimensional distribution (this kind of distribution has nothing whatsoever to do with the distributions of analysis.p p I on a manifold M may be definable only for some small time interval. Beauty is the first test: there is no permanent place in the world for ugly mathematics. is spanned by a vector field locally.p C Mp . H. A Mathematician's Apolog)' though the vector field is COO on all of M. This al'gument actually shows that for every p E M there is a coordinate system (x. the set {q E U : x2 (q) = a 2. . by defining N to be {c(I)}. we can choose (in many possible ways) a vee to!' field X such that 0 # Xq E /. . Hardy. quoted in]. S. a " . A Afatllcmaticia1l1 A1iscellan ). we can always find all integral manifold N of a Coo distribution /. so that global results can be obtained. We will now vary our question a little. xn (q) an } = .CHAPTER 6 INTEGRAL MANIFOLDS PROLOGUE A mathematician's repu tation rests on the number of bad proof" he has given . . For a l dimensional distribution the notion of an integral curve makes no sense. we have seen that the integral curves of a vector field where i : N + M is the inclusion map. find an integral curve c of X with initial condition or e(O) = p. We call /.q for all q in some open set around p. Besicovitch. G. Instead of a vector field. U) such that for each fixed set of numbers a 2 . Then /. we just choose a vector field X with 0 # Xq E /. 1 79 . which include such things as the "afunction"). sup pose that for each p E M we have a I dimensional subspace /. Littlewood. and then forget about the parameterization of e. that is.q f q in a neighborhood of p. if for every p E N we have i* ( N ) = /. . • . [Pioneer work is clumsy] A. a Coo distribution if such a vector field X can be chosen to be Coo in a neighborhood of each point. E. For a given p E M. . but we define a (Idimensional) submanifold N of M to be an integral manifold of /.
. The entire manif old M can be written as a disjoint union of connected integral submanifolds of !. On the other halld.hkh has olle compact connected integral manif old. there is a distribution 011 lhe torus whose integral manifolds all look like the dense I dimensional = 1 11 . rather than curves with a particular parameterization. This is still a local result: but because we are dealing with submanifolds. we can join overlapping integral submanifolcls together..1 1 ���l il l l l l l 011 submanif old pictured in Chapter 2. It happens that the integral manif olds ofthesC' two distributions are also the integral curves for certain vector fields. For example.. and that these are the only integral manifolds in U. there is a distribution the torus .. and all other integral man ifolds nOlicompact. on U. but on the . which locally look like  (rather than like or something even more complicated).180 Cha pter 6 is an integral manifold of !.
p consists of all may be pictured as the plane with the equation z {r �Ip + s �lp + br� lp : r. integral manif olds do llot exist. for each q U. ( q). if for every we have Although the definitions given so f all look the same as the I dimensional ar casC'.b. .p = If we identify T� 3 with � 3 x � J .P l where D. in a neighborhood of each point p . .c) is spanned by E pEN i*(Np) = /. . As the simplest example. Xq with this property. A (kdimensional) submanifold of is called an integral manifold of /.p where Mp. 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. .(a. .br)p. In general. . in �3 for which /. A kdimensionaI distribution on is a function p 1+ D. b(x � . C For any P there is a neighborhood U is a k dimensional subspace of and k vector fields X" .moment we \'\'ill investigate higher dimensional cases only locally. Xdq) are a basis for /. Thus /. Thus /. . Xk such that X.Integral Alanifolds 181 ]\1obiu5 strip there is a distribution which is spanned by a vector field only locally. even locally. . a Coo distribution if it is possible to choose Coo vector fields X" . then /. the results will look very different. s.p c= .a).q. M E M N M i: N M + p Mp is the inclusion map. consider' the 2dimensional distribution D. . . .p = /. S E } (r. We call /.
(O .182 Chapter 6 The figure below shows /..c) through (a. For example. :)plane through 0 whose tangent vectors would have to lie in the intersection of /. its tangent space at ( 1 . 0) contains vectors with third component nonzero. � � � � � � � � � � If you can picture this distribution. y. The only such vectors have third component 0. c) isjust parallel to the one through (a. Now consider. yO.y. z)} through (0. Yo. z)plane. b. b. a proof can be given as follows. : ) } .b. The plane /. But this submanifold c10es not . f each fixed Yo. with 0 E N. s o i t must b e the linc { (x. 0). or the intersection N n { (x. The intel'section of N and {( O. . yo. Suppose there were an integral manif old N of /. so Y lllust be the yaxis.z) and the (y.(a.'ork.. you can probably see that it has no integral manifolds. with all tangent vectors having slope Yo.z)} would be a curve y in the (y. Our intcgral manifold would have to look like the following picture. This will be a curve in the plane { (x. O). yox)}. 0). yo.p for points p = (a. 0. b.
b)(x . ax These tangent vectors are in /. b. y) plane. c) will be parallel to the one through (a..a) +g(a. consider the somewhat more general case where D. since f and g depend only on a and b. Since /. has an integral manifold N through each point.b) + sg(a. + a. the submanifold is given locally as the graph of a function: N = ((x. aa g(a. To see in greater detail what is happening here.Integral Manifolds 1 83 p /.c = f(a.b)] Ip : r.Ip .Ip + . b) if ' p p .p = �{ + s � p + [rf(a.(a. the plane /.b.c) = D.p is the plane with the equation 0 .b) = ay (a.b).s E IR . ay = g.(a. /.p if and only if I I aa a a . S o we need t o find a function a: 1R2 > IR with aa � . (a. aay ay • = (a.b) (y .. geometrically. b).b.b) = a. : p Now the tangent space at aa f(a.c) through (a.(a.b) .z) z = a(x.b. is r· l � } We now ask when the distribution /. (a.y.p is never perpendicular to the (x.b)..b)) is spanned by a aa a a. y)}. 0). b. As in the first example.f.a(a.
O) = :0 �=I ax aa ay = g. 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. If I. O) dl. 0) = l(x. g : 1R2 + IR satisfy in a neighborhood of 0.O) = :0 + namely. '!§. 0) so that a(O. we find a necessary condition on f and g: In our previous example. aa (x. O. PROPOSITION. defined in a PROOF We first define a(x. we define foX 1(I. I 0'. O). I(a . = 0' so this necessary condition is not satisfied. then there is a function a(O. such that �. b ) g(a._0' . and ':0 E neighborhood of 0 E 1R2 . By using the equality of mixed partial derivatives. b) = 0. = b. and u. a(x. 0) = :0 (I) a.. al = ay ax I.1 84 Clzapter 6 It is wellknown that this is not always possible.
then also aajax = f.y) = a(x. aajay = g.Integral Manifolds Then. y) so that (2) aa a:. y) = g(x. I) dl. 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. y) . for each fixed x. 0) dl + 1(" g(x. I) dl 1 y This is 0 for ). and always provide us with an a satisf y� ing (2). y). To prove that it equals 0 for all shO\\' that its derivative is o. s �} . g : /'. z) : z = a(x. y)} . Suppose that N = : E { (x. = 0 by (I) . 0) + = Zo This construction does 1)ot use (**). for each 185 x. namely. ax y. the function + r 1(1./(x. y. (x. we define a(x. 0 0 laY g(x. To prove this.. y). consider.p {r a�{ + s � Ip + [rl(p) + sg(p)] � Ip r. We claim that if (**) holds. = �3 + �. But its derivative at y is f+ aa (x. we define a(x.
ver ' point. a(a . b» g(a. since it still involves the unknown function a. a (a.a(a. In order to obtain necessary conditions for the existence of such a function a) we again use the equality of mixed partial derivatives.b. a x ay This condition is not very useful. b» · . a(a. b) a: p . b . a (a. b» = = aa � (a . ' + � p � (a. a(a.a(a . (a. which means that for each ) pair (a .(a . b.a(a . Now we are looking for conditions which will be satisfied by I and g when there is an integral manifold of /. b) = . b).p if and only if I I I I I(a. b» . b. b. b.(a.. but we can substitute from (*) to obtain al al . b» . a (a . Thus (*) and the chain' rule imply that al a2a al aa . b) these equations must hold no matter what a(a. by aa a a a. b). b» + ilz (a. b.a(a . I(a. b) = � (a. (a. b). a(a.(a. through . b. a(a. b .a(a.b) p aa a a a p + a (a. b. y y • These tangent vectors are in /. aa a. b» + ilz (a. b . b» g(a. b» is spanned. Thus we obtain finally the necessary condition .a(a.b» + ilz (a. b» . b) is. b» ag ag = � (a. b) ay aY az ay a x II ag a 2a ag aa (a. a . .186 Chapter 6 is an integral manifold of /. b. b. once again. The tangent space of N at p = (a.(a. b» + .
We first want to define a(l. 0. . THEOREM. and we will concern ourselves with proving existence if these conditions do hold. . . . rom the proof o f existence. . . J. a partial diff erential equation for a function from �m to JR").a(l. defined in a neighborhood W of ° in JRm.. for i = 1. agree on the component of WI n W2 which contains 0. . . . . 0) = x aa atT (l. . with a different twist at the end. we can treat a system of partial differential equations for n functions on �m (i. 1. . . there is at most one function a : W + V . any two such functions a l and a2 . : U x V + JR" be Coo functions. al j (I) = Jj (More precisely. . . m . 0. j = 1. there is no need to restrict ourselves to equations f a single or function defined on JR2 .l111egral Manifolds 187 In this more general case. . Ill. In the following theorem. J. so for a function f : �m X �n + �k we use al aI' al ax' for for D. . satisfying a(O) = x aa (I. 0. . and let j. Let U x V C JRm x JR" be open. . .e. . .a(l» for ali I E W. such a function exists (and is automatically COO) in some neighborhood W if and only if there is a neighborhood of (0. 0) = 11 (1. we will use I to denote points in JRm and x for points in �n . x) E U V on which x i. Dm+. The proofwill be like that of Proposition 0. 0». 0) so that (I) a(O. .0. where U is a neighborhood of ° E JRm. . In fact. O. Then for every x E 11. defined on WI and W2 . 0.) Moreover. . 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. . . . .
0» a1 2 11 ' 1 ' " III '2· We claim that f each fixcd I' with 11 ' 1 or we also have. . all expressions involv ing a are to be evaluated at (t' . Define This equation has a unique solution. 0. . 0. . ' (I) = jj (I.al 2 . In the f ollowing. .0. . I. O) = x aa (1 . This has a unique solution f sufficiently small I. . (t < < < " < '2. . . . . . . . . . . < " . . (I) = al 2 al' . 0. 0. and verif the following assertion: If we choose " ).O. . � a/.I. (2) 0) = h(l ' . 0) = f3dt) Then (I) holds f III or Now for each fixed I' with 11'1 '" consider the equation f32 (0) = a (I ' . O. O. . ./. .1. 0. . . 1. O) .. 0) and all expressions involving Ii are to be evaluated at (I'. O. . 0». . f32 (1» . (3) O = g(I) = iJiT (I . O. 0. 1. At this point the reader must or < < " . . . O. axk a/2 ' k= ' (4) . . We now derive an equation for g'(I). sufficiently small. . . . . . . . " . . aak g. Note first that g(O) = 0 by (I). 0. for ali I with III aa . . 0. . 0. .. . .a(l .O. 1. . < a(l ' . . I. . 0» . . . . I. . 0. .0) will each be defined for III < '2 for some 82 > o. . f3dl» .1. then for 11'1 < " the solutions of the equations f f32 with or the initial conditions f32 (0) = a(l ' . . . . . defined for III III a(l.O. . .0. I. . . 1. 0) f3/(I) = h (t ' . .188 Cha pter 6 To do this. We then define fh (O) = x f3. We have a2a a/. . . . . ..0. . 0) = f32 (1) Then a(O.L. . we consider the ordinary differential equation refer back to Theorem 52. .a(1 . 0. . .a(I ' .
It is a simple exercise to continue the definition of a until it is eventually defined on (e l . and are called "integrability conditions". The second kind of theorem justifies this terminology..Integral Manifolds and thus 1 89 (5) g . al . So (3) is true. in which we will essentially begin anew. (5) . 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. 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. is clearly g (t) = 0 for alI I .. The remaining parts of our investigation. without ever even mentioning partial derivatives. a distribution with integral submanifolds through every point) then certain othel' conditions hold. by showing that the " integrability conditions" afe suffi cient for recovering the nice situation. The solution with initial condition g (O) = 0.. given by (4). en ) and satisfies t*) . (3) Now equation is a diff erential equation with a unique solution for each initial condition. 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 ) .(I) � aft k L.g. Theorem I essentially solves for us the problem of deciding which distributions have integral manifolds. •:. e l ) x ' " x (en . and prove their sufficiency.k=1.aft2 al by (2) by (2) again _ _ _ by definition. these Con ditions are obtained by setting mixed partials equal. The first kind of theorem says that if one has a certain nice situation (e.
O. fOI' i [Y" Y21 are Irelated. .I (a . f( each P E M. . we say that X and Y are Irelated if l p (Xp ) = Y p) for . Of course. ' L IY�. U) around P E M and (y.. M which is Irelated to Y.1 90 Chapter 6 LOCAL THEORY If I : M + N is a Coo function. . . a given vector field X may not be Irelated to any vector field Y. the latter condition is fulfilled: 2. axl Y= f( p) . a n . then where The most important property of Irelatedness for us is the following: 3. PROPOSITION. respectively. Xp . Let I : M + N be a Coo function such that I is an immersion. Clearly we must define Xp to be the unique element of Mp with Y p) = I . If X. I a L a' a. and Y. In One case. f(p) E I (Mp ). a. a n ) = (a . . 2. . then there is a unique Coo vector field X on PROOF. . y X= n = 1 .= 1 implics that the functions I. . 0 I = I. .. i f this i s true for all Coo functions g : N + JR. n . then Y p) (g) = l Xp (g) . V) around I(p) E N such that ' t J' 0 l o X. are Irelated. If g : N + JR is a Coo function. 0). . . If Y is a Coo vector field on N with Y p. then [X[ ' X21 and . . IJ = a�. (Yg) 0 I = XU o g) . . then X and Y are Irelated.p f( = Xp (g o I). PROPOSITION. so Conversely. and X and Y are Coo vector fields on M and N.= 1 . nor must a given vector field Y be Irelated to any vector field on M. This ( a�. we use Theorem 210(2): there are p f( coordinate systems ( x . are Coo (problem 3) . •:. Thus if where (i are Coo functions. . This is easily seen to imply that I p. To prove that X is Coo.
0 0 0 . Suppose that N is an integral manifold of /. Y]p E /. ''''le will say that a vector field X belongs to /. in 1R 3 given by [X. f] p = [X.g) f = X. f] and [X. we see that This belongs to D.g and Ytg.f .(g f) i = {[Y" Y2]g} f = {Y. then for all p E N there are unique Xp .� + f� _ g af ) � . S E IR } . Y] are irelated. f]p E Np . •. to have an integral manifold through every point. if there is an integral manifold of /. and Proposition 3 then shows that [X. then 1 .(g f» . Thus i. Using the f ormula on page 1 56.[X. Yp = i. and Y and f are irelated.Y] = (� . Proposition 2 shows that X and f are Coo vector fields on N. Y] also belongs (I) So 0 0 0 0 0 0 If g: N + IR i s Coo. this therefore shows that [X. If X and Y are two vector fields which belong to /. only when the expression in parentheses is 0. fp E Np such that Xp = i. if Xp E /. (Y2g)} f . /. The vector fields X = � + f� ax az a Y = a:ay +gazbelong to /.X2 (X. then [X. (g f» by (I) = [X" X2](g f) : Now consider a k dimensional distribution Do. 0 + to /. and i : N M is the inclusion map. with g replaced by Y. X and X are i related. Here [X. p In other words.{Y2 (Ytg)} f = X. ( [Y2 g] f) ..IntegraL JvlanifoLds PROOF. ax ay az az az ... Consequently.X2 ([Ytg] f) by (I). 191 (Y.p = { I' f lp + s hip + [rf(p) +sg(p)] �Ip : I'. For a moment look back at the distribution /.p . 2.Xp. Y]p. through every point p.. which is precisely the condition for /.p for all p. respectively = XdX.
x (x. . Y and [X.e. if X. . THEOREM ([HE FROBENIUS INTEGRABILITY THEOREM. [Xi.. PROPOSITION.. Y] [jX. •.J ' Such functions clearly exist if /'. it obviously suffices to treat each separately. Since we have clearly [X. Xj] is a linear combination k [Xi . It is equivalent to Theorem I. Xk span /'.(q). ) a n with all fa i r < e. FIRST VERSION).. 5... . is called integrable if belongs to /'.fiXi i= f or Coo functions C. . If and belong to /'. Any connected integral manifold o f /'. .. Y] X and Y 4. [jX.Xi.gj Xj] We arc now ready f the main theorem.... . . then /'. Theorem I can be derived f rom it (Problem 7).Y] do : + . . For every p E M there is a coordinate system V) with x .. But the proof is quite diff el·ent. x n (q) = a n } is an integral manifold of /'. /'. e). suppose such f unctions exist. Let /'.... . is integrable. be a Coo integrable kdimensional distribution on M.gY] = fg[X. the set {q E V : xk + l (q) = a k + l . since E /'. . x IV) = ( e. x(p) = 0 ( . restricted t o V i s contained i n one of these sets. Y] f(Xg)Y g(Yf)X. X. X. . [Xi ..Xj] = L C/..1 92 Chapter 6 In general. l Iich is spanned by the Conversely. . [f.. whenever belong to /'..gY] belongs to /'. If is integrable on V if and only if each X" . This condition can be checked fairly easily: [X. Xj ]q q X k X = L1 . in or fact. we can clearly write PROOF Y To prove belongs to /'.. . in a neighborhood V of p. e) such that for each a k + I .
. which implies that xm i is constant on the connected mani f old N : so that + 0 0 0 m S. . k .X. Moreover. But. (on a neighborhood of 0 E IRn ) and a/aI (on IRk ) are 1[related. If N is a connected in tegral manifold of I'> restricted to U. . . . 1[. . .6. = ax' i = I. . .Integral Manifolds 1 93 PROOF. The sets {q E U : Xk + l (q) = ak + I . 11. : 0. for i = I. k . So near we can choose unique 1[: D. ][* is oneone on f or near O. Then 1'>0 + IRko is an isomorphism. (q) = aaI. Thus or d(xl1l i) = 0. consider d(xm i) for k + I s. ' Then the vector fields X. since their tangent spaces are spanned by the a/ax' = X. .[X" Xj ]q = [ aaI" aaIj ] . of Nq we h ave d(xm i)(Xq) = Xq(xm i) = i.. k. j j By Theorem 514. is oneone on I'>q. . with p = O.Xq E I'>q. and 1[. I. which is spanned by the a/axj l q f j = I . x" (q) = a n } are clearly integral man ifolds of 1'>. k . By continuity. we can assume that . We can clearly assume that we are in �n . 0 . 1[. By Proposition 3. . with inclusion map For any tangent vector Xq i: N U. So [X" X ] = o. since i. . 0 .(q) i = I. there is a coordinate system x such that a X. •. . . .(q) = 0. . . [X" X ]q E I'>q by assumption.0 C �no is spanned by Let IRn + IRk be projection onto the first k f actors. .q q 1[.. .Xq(xm) = 0.
with xIV) = (0.. . and if around every point P E M there is a coordinate system (x. x (0. . . be a Coo kdimensional integrable distribution 011 M.. PROOF. . {q E V : xk+l (q) ak +I .xn (q) an } = = NnV are the sets of the form Each component of N is called a folium or leaf of the foliation N. (each component is called a maximal integral manifold of !. . . Notice that two distinCI components of N nV might belong to the same leaf of the foliation. But S n Vj has at most countably many components. xn(q) an } = = a slice of U. V). 0). 6. For such a coordinate system V). we introduce the following terminology. we see that we can cover M by a sequence of coordinate systems (Xi . 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. Then M is foliated by an integral manifold of !. let us call each set (x. ( () Using Theorem 12. THEOREM. . such that the components of {q E V : Xk+l (q) ak+I . . It is possible for a single slice S of Vi to intersect Uj in more than one slice of Vj' as shown below. If M is a Coo manifold. 0) x .1 94 Chapter 6 GLOBAL THEORY In order to express the global results succinctly. Let !. Vi ) satisfying the conditions of Theorem 5. .).
and only countably many such sequences. sO S nVj is contained in at most countably many slices of Uj . since we do not have to find a countable number of coordinate systems for each leaf. Given P E M. choose a coordinate system (xo. or totally disjoint from. il = i and corresponding slices with a = 0. . . and let So be the slice of Vo containing p.] . In this case. Using Problem 31.1 . . Since there are at most countably many such sequences of slices for each se quence io. . Appendix A describes a nonparacompact manifold which is foliated by a lowerdimensional connected submanifold. A slice S of some V. the corresponding union is either equal to. . M is foliated by the disjoint union of all such submanifolds. . there are at most countably many slices joined to p. . . [If we are allowing nonmetrizable manifolds. i/. the first union. will be called joined to p if there is a sequence 0 = io . however. and can merely describe the topology of the foliation as the smallest one which makes each slice an open set. this di'!ioint union is clearly an integral manifold of /'. we see that the union of all such slices is a submanifold of M. . the discussion to follow will not be validin fact. . Vo) with P E Vo. •:. . . the proof is even easier. 1 .p. For q #.InlegmL ManifoLds 1 95 and each component is contained in a single slice of Vj by Theorem 5. . Consequently. i 1 ..
choose a coordinate system f(p) such that the slices MI _ (x. .U) around {q E U : Xk+l (q) = ak +I . This allows us to apply a proposition from Chapter 2.. U) U However. . (x. .xn(q) = an} are integral manifolds of /'. it suffices to show that f is continuous M P MI . . so f takes f(P) MI . Let be a Coo manifold.. . . . THEOREM. 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.1 96 Chapter 6 is a coordinate system of the sort considered in the proof Notice that if of the theorem. Now f is continuous as a map into M. then infinitely many slices of may belong to the same f olium. According to Proposition 21 1. and a f olium of the f olia tion determined by some distribution /'. P M MI .. PROOF. this folium would contain an uncountable disjoint family of open sets. 7.. otherwise CiV ' . at most cOUil/ablJ' man slices can belong to the same f y olium.
for all p' E W. . if we had x . by continuity. then x . we can choose W to be connected. (f(p'» = a. (f(p'» #. •:. and x . For k + I :s i :s n. Consequently. (f(p'». This makes it clear that f is continuous as a map into MI . 0 f would take on all values between a . f( W) is contained in the single slice of U which contains p.a.IntegraL ManifoLds 1 97 some neighborhood W of p into U. In other words. contradicting the fact that f( W) e M. This would mean that f(W) contained points of un countably many slices. . for any p' E W. x .
In the proof of Proposition 4. x = L f3 ' ax. are n . IB) is a bundle map. (a) In the proof of Proposition 2. (a) Let � = 1[ : E + B be an nplane bundle. check the assertion about choosing ciently small. /'. we say that �' is a subbundle of � if (i. be integrable distributions on Suppose that for each P E M. 0 f = f3. and show that a kdimensional distribution is Coo if and only ifit is a Coo subbundle. aj a xd" etc. show that the functions C. . and �I k plane bundle such that E' c E. . ' 1=1 f3. . (b) Supply the proof of the uniqueness part of the theorem.. 5. a Coo. Let /'. (b) For the case of Coo bundles � and �' over a Coo manifold M. • . . Show that a kdimensional distribution on M is just a subbundle of TM.} actually are Coo . . then the functions 4. . such that /'. dt. . and IB : B + B the identity map.1 98 PROBLEMS Chapter 6 1 1[ : E' + B a 1. 1 . show that 6'1 suffi (b) Complete the proof of Proposition 2 by showing that if so that with a. 1 . 1. .. (x. (a) In the proof of Theorem I. Show that there is a coordinate system i s spanned by aj ax l .. .. of dimensions dl . M. = 2. U) around each point. If i : EI + E is the inclusion map. • . 3. define a Coo subbundle.
7. we can consequently assume that e = 1 . f3(U. I) for some function f3 : [0. jj (ul. Prove Theorem I from Theorem 5. x) . One has to check that one 0 can be picked which works for all l E W. VI). show af3 j iU (U. by considering the distribution !. = j m Ll 11 . This problem outlines another method of proving Theorem I.integraL ManifoLds 1 99 6. defined by Notice that even when the jj do not depend on x. the partial differential equations to ordinary equations along lines through the origin. in JRm x JR" (with coordinates I. (c) Conclude that af3 (v. VI) = f3 (UV. (b) Show that = x.) By shrinking W. by reducing x W + V. 1) We know that we can solve such equations (we need Problem 55. 0) that f3 must satisfy the equation 7. (Show that both functions satisfy the same differential equation as functions of u. al j . with the same initial condition. This is connected with the classical technique of "introducing new independent variables". since the equation depends on the "parameter" I E JRm). I)) = . al j ajj al i aft al j ' with the integrability conditions we nevertheless work in �m x �n ) rather than �m. I) f3(0. f3(u. A similar technique will be very important in Chapter II. l) al j = V· af3 ( I . I). so that the equations are of the form aD! (I) = jj (l). (a) If we want D!(UI) = f3(u.
This problem is for those who know something about complex analysis. (e) Define a(l) = f3 ( I . Noting that a(vl) = f3(V. Use (c) to conclude that satisfy the same diff the two functions are equal. Let I: <C x <C + <C be complex analytic. I). I)) erential equation. y)) = ay (or of any point </>'(z) = I(z. a l (x. ax aa ' a.200 Chapter 6 (d) Use the integrability condition on 1 to show that af3 (v. = XI . as functions of v. a' (x. show that a satisfies the desired equation. I). y). 8. 2.ay <C aa ' u(x. and conclude that the (in which f denotes the complex derivative) has a solution in a neighborhood of =0.Y �2 in a neighborhood of 0 E differential equation = = � v(x. y)) = . a' (x. a ' (x. 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 . I ) al j and v · h ( VI. If we denote the coordinate functions in <C x <C by ZI . y. 2.</>(z)) 20 E q. with any given initial condition 4>(zo) = Woo . f3(v. y ) . y .
• . It is clear that I1 k (V) c :rk ( V) is a subspace of :r k (V). " Vi" " . Vi. . " vk ! + T(VI " " . . the discussion of which requires some e turn OUf attention once more to tensor fields. Vj . if f: V + W is a linear transformation. " Vj" " . • .CHAPTER 7 DIFFERENTIAL FORMS morc algebraic prclimi. Uk . but one case is \vellknown. Vi. . Let V be an /1dimensional vector space over JR. . . It is also convenient to set l1o(V) = :ro( V ) = JR. • . vk ! + 0. Uk ). . [This is not true in the special case of a vector space over a field where 1 + 1 = 0. we have 0 = T(v) • . An element T E :r k (V) is called alternating if Wwith a special kind of tensor field. Therefore. Vi" " . The most familiar example of an alternating T is the determinant function det E T"(�n). Moreover. vk ! + T ( VI " . if T is skewsymmetric. Vi + Vj • . . Vi" " . and the condition of being alternating is the stronger one.jl . then T is also alternating. " Vi" " . ) Vj. . vk ! = 0 + T ( Vj " " . but we will be concerned if Vi = Vj (i #. " vk ! + T ( V I " " . in this case. Notice that I1 I ( V ) = :r1 (V) = V* .onsidered as a function of the n rows of a matrix we shall soon see that this function is. then f* : :r k (W) + :r k (V) preserves these subspaces f* : I1 k (W) + I1 k (V). . Vi. T(VI " " . so I1 I ( V) has dimension /1 . .naries. " . Uk ) = T(VI " . the most general alternating function. c. Uk ) = . . . At the moment it is not clear what the dimension of I1 k (V) equals for k > I. . . " " Vj" . Vi. . .T(vJ • . . " Vj" . Vi • • . skewsymmetry is the same as symmetry. Vj • . . one is a multiple of the 201 . T is skewsymmetric: T(v) . vk ! = 0 If T is alternating� then for any VI . . . in a certain sense. . • . . ' " vk ! + T(vj" . .] We will denote by I1 k (V) the set of all alternating T E :r k ( V ) . Vj" " . Vi + Vj • • • . " Vj " . Of course. f\1osl discussions of the determinant begin by showing that of any two alternating nlinear functions on �n. Vj" " . . . .
VI. vd) = ( pa ) . wd = (Wa(I) . . Thus warned. (VI . an element a E Sk is a function i r+ a (i). k. . . . . . (vJ . . . for example. k}. dim 11" (IR") :S I. . . . .202 Chapter 7 other. Walk) ) = ( Vp(a(l)) ) . . . ' . VI = W3 • . (v" . . . . . of course. then the first element on the right is 1lot necessarily that v whose index is a ( 1 ) . . . . . . I Alt T(v" . Let Sk denote the set of all permutations of { l.e. ' " Va(k) ) . . . T(va(l )" ' " Va(k) ). which is a special case of the definition to follow. .. Now for any T E :r k ( V ) we define the "alternation of T" I Alt T = k! aeSk L i. vd = ( Va(I ). . T 0 a. if these v's have indices running in some order other than 1. ' . we compute by setting so that Vp(l) = W I • . This definition has a builtin confusion. . ) is to rename things: V3 = WI ) V2 = Wz. . Vk ) is a k tuple (of any objects) we set a · (VI . If (v" . vk! = k! sgn a . . . O n the right side. . vd. The construction of det is usually by a messy: explicit formula. The simplest way to figure out something like a ' (V3 ) V2 . . \'\. . . . .here sgn a is + I if a is an even permutation and . Then one proves dim 11"(IR") = I by actually constructing the nonzero function det (it follows. . is the a ( l )" of the v's on the left side. . . . Vp(k) = Wk) a · (p . ueS" L sgn a . . vd ) = a · (wl .1 if a is odd. that dim 11" ( V ) = I if V is any n dimensional vector space). . in other words. . Vp(a(k)) ) since Wet = vp(a)· Thus a . . . . the first element. (p .
then Alt w = w. Finally. if k is odd then w /\ w = o. THEOREM. k ! l!m! . + w /\ ry. then Alt(T) E I1 k ( V ) . (I) If S E :r k (V) and T E :rl ( V) and Alt(S) = 0. f\1oreover. We now define. 7879 of Calculus on Manifolds) . 2. it is easy to see that (3) /\ is "anticommutative": w /\ ry = ( _ I) klry /\ w. 203 PROPOSITION. ) /\ ry = w. + w. an element w/\ ry E I1 k +1 ( V ) . /\ ry + w. ry E 111 ( V). then Alt(S ® T) = Alt(T ® S) = o.) = W /\ ry. /\ ry w /\ (ry. but it makes some things work out more nicely. Left to the reader (or see pp. •:. The funny coefficient is not essential. as we shall soon see. e E I1m ( v ) . for w E 11 k (V) and ry E 111 ( V ) . In particular. PROOF. 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)). It is clear that (I) /\ is bilinear: (w. associativity of /\ is proved in the following way. the wedge product of w and ry. then Alt(Alt(T)) = Alt(T). (2) I f w E 11k( V ) . (I) If T E :rk ( V).Dijje"n liaI Forms I. then (w /\ 1]) /\ e = w /\ (ry /\ e) = (k + 1 + 111 ) ! Alt(w ® ry ® e). (3) If w E I1 k ( V). (3) If T E :rk ( V). + ry. by ( w /\ ry = k!/! Alt (w ® ry) .
so (I) implies thaI 0 = Alt(w ® [Alt( ry ® e) . . Vk+/» ueuuG [ a'eSk L sgn a ' · S(va'(1) . the other equality is proved similarly. the sum over each being o. Vk +l) is just some other (k + I )tuple of vectors). Let aoG = {aoa' : a' E G). a contradiction. (v l .Alt(w ® ry ® e). (k + l + m ) ! Alt((w ) ® e) /\ ry (k + / ) ! m ! (k + 1 + m)! (k + /)! Alt(w ® e). .Alt(1] ® e) = 0. . Then L sgn a . Suppose now that ao <t G. :. Notice that G n aoG = 0. . Va'(k) ] . . . \!\le can then continue in this . • . Vk+l) = = aeSk+l aeSk+l (1) We have L sgn a . . . .). (S ® T)(a' . . . = ry ® (k + /)!m ! k!/ ! . . . " " Va(k+l) ' Now let G C Sk+l consist of all a which leave k + I. (w /\ ry ) /\ e = The othel' equality is proved similarly. • .ry ® e) = Alt( ry ® e) . . (3) We have a'eG L sgn a' . then a = aoa' for some a' E G. ( S ® T) . Va(k+l) aeG = = o. . T(Va(k+l) . (2) Clearly Alt(Alt(1] ® e) . . k + I fixed. . . . (a . . . . Va(k) . Then L sgn a . for if a E G n aoG. . S ( Va(I). (vl . S(Va(I ) . by (*) . . .204 Chapter 7 PROOF (k + /)! AIt(S ® T)(Vl . . . Vk+I» ) . . ."lay: breaking Sk+! up into disjoint subsets. (ao ' (VI . The relation Alt(T ® S) = 0 is proved similarly. . so ao = a(a.I E G.. T(Vk+l > " " Vk+I ) = sgn ao ' ''''le have just gjwwn that this is 0 (since ao . ( S ® T)(a . ( Vb .ry ® eJ ) = AIt(w ® Alt( ry ® e» . . " " Va(k) ' T(Va(k+I). Vk+I » L sgn a .
(So i f V I . the factor 1/ k! in the definition of Alt is essentialwithout it. ellen over a field a ffinite characteristic. .. . !. . . because each term in the sum Alt( w @ ry)(VI . · I\ </>n )(VI . . ."2 k ( V) can now be described."2 k ( V) C :r k ( V) .. .. and 1/ k! I! can be interpreted as meaning that these k ! / ! terms are replaced by just one. < ik ::: n = 11! k! (I1 . [If we had defined Alt just like Alt . . If w E !. 4>n is the dual basis.) ® 4JiJ. . Vn i s the standard basis for IRn . we can write W = iJ . THEOREM. . which therefore has dimension () 11 k : . VII is a basis of V. This makes sense. . 1\ </>n = det. · + I)! Alt(</>I @ " .iJ. . . L ail . then </> 1 1\ . but without the factor 1/ k!. . then </>1 1\ " .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.) A basis for !. iJ. @ </>n ) I!" . then 1\ could be defined by w 1\ ry = 1 k!l! Alt(w @ ry). If V I . . . we would not have Alt(Al t T) = Alt T.. .] 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).k) ! ' (In particular. . I! aeS" L sgn a · (</>I @ · · @ </>n ) o a · . > 11."2k ( V) = {OJ for k PROOF. " . and the first equation in the proof of (2) would fail. Vn ) = I . ( I + . and 4>1 . (</>I I\ . 1\ </>n = = In particular. 4>il ® ' " . . On the other hand. 3. . ."2 k ( V). . . . . . . The set of all 1 ::: i) < " is a basis for !.
. . On the other hand. < .·. < }k. vk! f+ determinant of a k One more simple theorem is in ordel� before we proceed to apply our con struction to manifolds.. Then w. then W' . ® . If WI " 4. . . Wk E !:) ' ( V). . Vn • . . . ® . If 0= i l < " '<ik L ai! .ik 4>iI !\ · · · /\ <PiJ. . < . /\ · · · /\ 4>ik . . . . .:. ® 4>ik ) is either 0 or = ± ( I / k!) 4>h /\ . . /\ . . . . . . so it is not o. . Uk • . . if then uniquely expressible as k (i" . Then every element of !:) ( V ) is To abbreviate formulas.206 Chapter 7 So W = Alt(w) = iI . . ik!. If W" . . so the elements 4>h /\ . . then applying both sides to (Vii " ' " Vik ) gives ail . . . . . Wk are linearly inde pendent if and only if w. . . . .o. . .. . 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" . . . ·.h L ai. x k minor of (:�). /\ 4>h for }. 4>k. it is convenient to let L al 4>l . . .ik Alt(4)i.i}. /\ 4>h for some k }. /\ Wk is a basis element of !:) k ( V) . . . = o. . /\ Wk #. 4>n satisfy 4>i = Wi for I :s i :s k . . Wk are linearly independent. . /\ . . there is a basis V I " of V such that the dual basis vectors 4>" . . . < }k span !:) ( V). COROLLARY. . . ® 4>ik ) · Each Alt(4)i. ) PROOF. and let 4>1 denote 4>i . . .
.B is a vector bundle. . .) = w /\ ryl + W /\ ry. we obtain a new bundle !.) /\ ry = WI /\ I/ + W. . (p) w(p) !. and let n Wi = L CXjiVj = l . /\ ry W /\ (ryl + ry. . /w /\ ry = w /\ /ry = /(w /\ ry) w /\ ry = ( . sections of n k ( TM). . . " (Unl. . vn ) : 6. then there is a ry ( (u l I . We can also define WI + w. . . ."2n ( v) . . . . . . ."2k (1[1 (p» !. . > O. A I form is just a covariant vector field. •. j=1 i PROOF. . all forms will be understood to be Coo forms unless the contrary is explicitly stated."2 n ( v) . then we can define a section W /\ ry of !. . If � = J[ : E ?. ann») = W(f>lIVl. . . · . . and c = ry( et .ant!. Vn) With our new algebraic construction at hand. .I/Iry /\ w /*(w /\ ry) = /*w /\ /*ry."2 k (1[. COROLLARY. ."2 k ( V): (WI + w. Define ry E :r n ( JR ) by n . Since !. are called kforms on M.en) = w(Vt."2 k (�) by replacing each fibre 1[."2 k +1 (1[ 1 (p» . Remember that covariant tensors actually map contravariantly: If /: M + N is Coo.ill E !. . If V is ndimensional and 0 #. unique orientation [vt . we are ready to apply it to vector bundles. and w /\ ry. . THEOREM. !. Let 207 Then VI .DifJemiliaL Forms 5."2 1 (�). p ry)(p) w(p) ry(p) . . j=1 j=1 Then clearly ry E !."2 k (�) with is a function with E for each E B. Vn be a basis for V. then /* w is a kfol'ln on M. The following properties of kforms are obvious from the corresponding properties for !. and W is a kform on N. let W E !. If ry is a section of !. .talnVl). we can speak of Coo forms. . so ry = c · det for some c E JR.1 A section W of !. . ."2 n ( JRn ). ."2 k +/ (�) by (w /\ = /\ E In particular."2 k (TM) can obviously be made into a Coo vector bundle. which are just alternating covariant tensor fields of order k. v ] = /L IL for V such that n if and only if W(VI.1 (p» . n. " .
It suffices to show that f: (x. . /\ dxiJ. . then the dx. U) is a coordinate system. . . . . .. ay i= l  ( a(y. THEOREM. . V) a coordinate system around E N. 0 by Theorem 5 . by Problem 41.1k(p). .. dy" ) = (g o 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. < ik) are a basis for !. . . . f\. /\ dyn (q) (f* a�' Ip .208 Chapter 7 is left to the reader (Problem 16).. f*(dy' /\ . f*(dy' /\ . . /\ dX.. . . Now. 7. ik). Wl dxl. 1 M + N is a Coo function between l1manifolds. .· by dx l for the multiindex I . then q = f(p) f*(gl/Y' /\ .. Thus every kform can be \\'Titten uniquely as W= if we denote dX i l /\ . /\ dy")(p) ( a�' Ip . .. and (y. so the dx. /\ dyn ) = det ..x�f» ) dx' /\ . 1=1 q q ) .. . L. J* a�n IJ n a = dy'(q) /\ . . ay a � a(i o f) a 1 .(p) . /\ dyn (q) (L a(ix l f) (p) �l 1 . but we will do one special case here. •:. C(Y.. . .. . /\ dxn .' (p) /\ . a. PROOF. (p) are a basis for Mp '.. o f» ) dx' dx". . d et � A · · · f\. If (x. If U) i s a coordinate system around P E M. . w = = (il.k(p) (i1 < . a�n IJ = dy' (q) /\ .
Vn) :0:: 0.o . Vn» Now let O if and only if [VI > . if there is a nowhere zero l1�form w on an nmanifold M. . . then a nowh". . then M is orientable (i. h ( ) [This corollary shows that nforms are the geometric objects corresponding to the "even scalar densities" defined in Problem 41 0. In particular. P E U we have or WU ( VI > . . . .I (p » determines an orientation /Lp of . if VI ..:. or .I (p) by Corollary 6. then PROOF. . . so that /L = {/Lp} is an orientation of �. . .. the nonzero w(p) E I1 n (. . zero section w of I1 n (�) has a special significance: For each p E B. W = L 4>uwu. . . . U) choose an nform Wu on U such that f VI > . If a Coo manifold M is orientable. and a partition of unity {4>u } subordinate to e. /\ dxn. By Theorem 213 and 21 5. THEOREM.] If � = " : E + B is an nplane bundle. /\ dyn = h dxl /\ . . . and strict inequality holds f at least one U. •:..DijlerentiaL F orms 209 8. Vn E Mp satisfy (4)u Wu )(P)(VI . If (x. U»).e.. det a/ . U) and (y. It is easy to see that the collection of orientations {/Lp} satisf the "compatability condition" set forth y in Chapter 3. For each ( x. COROLLARY. Vn] = /Lp.. we can choose a cover e of M by a col lection ofcoordinate systems {(x. . . Let /L be an orientation of M. Vn E Mp . . . then there is an nform on M which is nowhere o. Ue(9 for every Then w is a Coo nform. . vn] = /Lp. ax) Apply the theorem with f = identity map . . . the bundle TM is orientable). w PROOF. . = g . The converse also holds: 9. V) are two coordinate systems on M and g dyl /\ . . [VI . then eadl p. Moreover. Thus w(p) #. .
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»).] Just as r:20( V) has been introduced as another name for 1R. then it certainly has a nowhere 0 section. 1\ 1\  1\ . We have shown that if M is orientable.) = dWI W.. then r:2 k (n is trivial if and only if � is orientable. The first way is to use a bruteforce computation: comparing the coefficients in the expression f W f . Conversely. PROPOSITION. which in a coordinate system U) is given by dw with the The second method is a lot sneakier. d' = o. I then each dWI is a Iform. w). I a= ax· f (x. if the bundle r:2"(TM) is trivial. and we can define a (k + I )form dw.210 Chapter 7 It turns out that this definition does not depend on the coordinate system. j= ax} If W is a kform W = L Wl dx l . df df = " af df = L1 � dxj . This can be proved in several ways. the differential of w. [Generally. + ( I /wl dw. which implies that it is trivial. f W'l WI. so M is orientable. We begin by finding some properties of (still defined with respect to this particular coordinate system). of course. (3) d(dw) = o.. W = L W'l dx l I Notice that the bundle r:2"(TM) is Idimensional. provided that the base space B is "paracompact" (every open cover has a locallyfinite refinement). 1 0 . then d(wl w. (I) (2) If d(wl + w. if � is a kplane bundle. Briefly. WI is a kform. then r:2"(TM) has a nowhere 0 section.) = dWI + dw. by dw = L dwl dx l I " aW = LL1 l dx· l\ dx I .
.) dUg) /\ dx l /\ dxJ g df /\ dx l /\ dxJ + f dg /\ dx l /\ dxJ dWI /\ w. =gdxJ. and satisfies d on U. = fg dx l /\ dxJ and d(wl /\ w. = . d' takes kforms on U to (k + I )forms on U.. (3) d'(d'f) = o. (2) d'(wl /\ w. PROPOSITION. ax.. WI = fdx l w. (I) d'(wl + W2 ) = d'w l + d'w. + (I) k w I /\ d'w. •:. Then WI /\ W. +(Ilfdxl /\ dg /\ dxJ dWI /\ W. (4) d'f (the old) dJ: Then d' = d on U. == == == == == == (I) is clear. . so n ( n a'f d(dw) L L aX"ax· dxP /\ dx· /\ dx l ) . \!\le next note that these properties characterize I I ..) = d'wl /\ w. Suppose for all k. Then n af dw ". the terms a=1 /3=1 a=l Q  and cancel in pairs .dx· /\ dxl L. (3) It clearly suffices to consider only kforms of the form w = fdxl . To prove (2) we fIrst note that because o f (I) it suffices to consider only In this sum..Di/ ji<rential F orms 211 PROOF. + (llwI /\ dw.
by (2).212 Chapter 7 It is clearly enough to show that d'w = dw when w = PROOF.I /w. There is a unique operator d from the kforms on M to the (k + I)forms on M. . /\ dw. /\ w. . ) /\ d'xiz /\ . •:. . /\ d'Xik ) = 1 2 . . = df /\ dxl + f /\ d'(dxl) dxl = dX il /\ . ' /\ dXik = d 'xi. We will uSe induction on k . + dw. . Assuming it f k . satisf ying = 0 . d' = 0. f dx1 Now So it suffices to show that d'(dxl) d'(j dxl) = d'j /\ dxl + f /\ d'(dxl) by (4). /\ .) = dw. dw(p) = du(w [ U)(p) ) . U) we have a unique du defined. /\ d'(d'xi. Given the form w. . For each coordinate system (x.0. and p E M. /\ w. /\ . pick any U with P E U and define PROOF and agreeing with the old d on functions. COROLLARY. d(w. by (2) by (3) and the inductive hypothesis .) = dw. . /\ d'Xik ) d'(d'xi. .I we have or d'(dxl) = d'(d'xi.d'X i. • The third way of proving that the definition of d does not depend on the coordinate gystem is to give an invariant definition. where by (4) . /\ d 'Xik = 0. . d(wj + w. /\ . + (. ' /\ d'Xik . for all k.
." say) .. /\ dx .::k+1 (= }:.L(lin+j (Xjl)w(X. . . . It is easy to check that dw is alternating. becomes I}:..YI · X [X. ..Xk+'). . X}. . THEOREM. .i<j. Y] . Xk+. X. Y] + XI . it is easily seen that 2:2 becomes I }:. f\1oreover. . Y] I[X. if Xio is and using the formulas [IX. + }:. . Y." ' (X" . + }:.. we might as well assume w = I dx' /\ . ) Xk+1 . ...<ill .. In fact.li+j w([x. by renumbering. becomes IX. X}. + }:. k Theorem 4�2 shows that there is a unique covariant tensor field dw satisf y ing (*) ." . then there i s a unique (k + I)form dw on M such that for every set of vector fields we have w XI) .X.) where over indicates that it is omitted. X" . . . . jY] = I[X. . . . .. it is actually linear over lhe replaced by then }:. . . Xk+') to }:.. . .. . X" l::. . PROOF. . . io<j a brief inspection then shows that }:. . To compute dw in a coordinate system (x. The operator which takes over � .Differential F orms 213 1 3.. .. Moreove. .. . This (k + I )form agrees with dw as defined previously. .J)w(X. so that it is a (k + I)form.X".) . .. Xk+. . + I }:.. is clearly linear Coo Junctions y.X" " " X. .. U) it clearly suffices to compute d(1 dx l ).=1 + L (.. + = L(li+'" (x. . Xj ]. I f i s a kform o n M.. .u . X.. ." ". .
dw(p . . k ). in which case so which is just the old definition . ax} dxj dx' L. . .a/ax·k+ ' ) = 0 unless some (0') . j>k " aj = L1 ax} dx}.214 Chapter 7 dw. Xk+l . . 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. but first find where are vector fields extending and then evaluate this function at By some sort of magic. . . directly. this happens only if j > k. . . /\ /\ . This term is what makes the operator dw(X" " " Xk+'). X" . . iii . as for any form. For . . One other feature of our definition is common to most invariant definitions of tensorsthe presence of a term involving brackets ofvarious vector fields. Since the a's are increasing. . . . Xi p. . /\ This is our first real example of an invariant definition of an important tensor. . j= � /\ /\ /\ . •:. After all. we have dw = It is clear from H that dw(a/ax· ' . . and our first use of Theorem 42. Xi dw(X" . /\ dxk dxj dxk dxk . . dx . . although of except at it does depend on the values of at points other than this must enter into our f ormula somehow. . .)(p) w p .. . Q'k +d is a permutation of ( I . . p. . But we can hardly hope f anything or does not depend on the values better. We do not find )(v" . . .. /\ /\ . Vk+') Vi . . a� dw = L(I/ ax dx' j>k af = ". . . Xk+. .
• . for ali i.). . . . Y]) I = E ry /\ E E n P E . the ideal J(!. . A distribution !.q.) w w whenever XI. around any point M we can choose a coordinate system (x. J(!. SECOND VERSION). In fact. wn . is a kdimensional distribution on M. .) is an ideal in the ring \1(M)]. Locally. .n. . If !. There f ore. . J(!.Y(w(X» .Xl belong to !. Y) X(w(Y» . but it disappears in computations in a coordinate system. P=l We can therefore let k w" = dx" . dxl (p) dxk (p) is nonzero on !. 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. U) so that ."L. 1 4 .Di/Jemlliai F orms 215 linear over the Coo functions.X ) It is clear that WI + w. on M is integrable if and only if . dx k (q) are linearly independent !. J(!.p. .w([X. there are Coo functions I such that t k dx"(q) = "L. . /\ in . Theorem 13 gives the following formula.) if Wl . By continuity.).. PROPOSITION (THE FROBENlUS INTEGRABILITY THEOREM. W(XI . the same is true for q sufficiently close t o p . In the particular case where is a Iform.) if w J(!. . This enables us to state a second version of Theorem 65 (The Frobenius Inte grability Theorem) in terms of differential forms. .W' J(!.» c J(!. Then d(J(!.It dxP. P= l /\ . which by Corol lary 4 implies that dx1 (q). and that W J(!.) [thus.lt(q)dxP (q) restricted to !. . .) is generated by . .k independent Iforms Wk+ 1.q a = k + I . Define the ring \1 (M) to be the direct sum of the rings of Iforms on M. ..
if I'>. .Xj. Xn (q) = a n } are integral submanifolds of 1'>.. 0 = dw" (Xi. . .. .e. Xk span 1'>. . Xj]) = 0 [Xi . . so we can write the condition d(J(I'>. /\ wp . .. Let Xl . ) = L(ej /\wj )(Xi. Locally we can choose Iforms w I . But each if and only if each belongs to I'>. and io. .w" ([X" Xj]). . .). :+ Now P=l = L OJ /\ wj If a > i<j for certain forms OJ ..wn must be linear combinations of them. the first two terms on the right vanish. Xj] dw"(Xi . . j = I. we can always For I :S i. .Xj(w"(Xi» . wn which span Mq for each q wk+ 1. is integrable ifand only ifthere are functions ct k [Xi .. j :S k and a > k. Xj) = 0 w"([Xi .216 Chapter 7 such that PROOF. dw"(Xi .Xj) = 0 dw" Notice that since the w i /\w j (i < j) span fl2 (Mq) for each q. io ::.. VVe therefore h ave the E . • • Then with Xl . wi (Xj) = oj. . we have following. .).» c J(I'>. • • k.. dxn are a basis for J(I'>. P>k Once we have introduced a coordinate system (x. .). Xn be the vector fields with ..) = ej�/XiU). So if and only if o. U) such that the slices l l {q U : Xk+ (q) = J + . . Xj.Xj] = L ctxp i...Xj) = Xi (w"(Xj» .. so wk + 1 . . Xj]) = W"([Xi. So I'>. . . . the forms dXk+1. is integrable). . write dw"(Xi . . . wn generate J (1'>. (i. k.) as dw" = L e. while each if and only if E J(I'>. k are distinct.
let (x. . .. •.. We can assume w g dxiJ /\ . . /\ .orms DifJerential F 217 wk+1 . then j*(dw) d(f*w). /\ dXik' /\ dXik ) j*(dw) : = ) = = /\ We will use induction on k. . 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. PROOF. PROPOSITION. 0 0 =0 = = = = . f* g is to be interpreted as g o f). then there are Iforms (ex. /\ dXik' ) /\ f*dxik f*(dg /\ dxi. U) be a coordinate system around f(p). For k = 0 we have. If are linearly independent If orms in a neighborhood of E M. . tracing through some defini tions.. Problem 1 8 gives another invariant definition of using induction on the degree of which is much simpler. /\ dXik' ») /\ j*dXik + since dj* dX ik d d(Xik f) j*(d(g dxi. /\ dXik' ») /\ f*dxik by the inductive hyposthesis f*(dg /\ dxi. d: f: M + N is Coo and w is a kform on N. /\ dxik . . COROLLARY. p dw· = L e. the cases k > I will hardly ever be used (a very significant exception occurs in the last chapter of Volume V). 1 6. If E = = = = j*(dg)(X) = dg(f* X) [f* X](g) = X(g 0 f) d(g o f)(X) (and.I. . .g P (ex. w' = L ft dgP • P p > 1 5 . /\w P if and only if there are functions f . of course. dw.wn e. . For p M. /\ . /\ . {3 k) with . . /\ ' . {3 > k) with Although Theorem 13 warms the heart of many an invariant lover. /\ dXik' j*dXik) d(J*(g dxi. . Assuming the formula for k . . /\ . we have d(f*w) d((f*gdxi..
�� = B aQ .. a differential was then called "exact" ifit actually was the differential of something. dw 0. af ax. w W L w. is aA + aB + ac = 0.e. The term "dosed" is based on an analogy with chains.. = Wj .=1 n One property of qualifies. however. i. dxi .) Since = 0. every exact form is dosed. which will be discussed in the next chapter. = 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 �: . ax ay az if and only if = then = =    . For 2forms the situation is more complicated. ay C The necessary condition. (The terminology "exact" is classical diff erential f orms used to be called simply "differentials". a.ap = . i. A form is called closed if = 0 and exact if = for some form YJ. The relation = 0 is just an elegant way of stating that mixed partial derivatives are equal.e. If is a 2form on � 3 . aWi aWj axj axi is necessary for solving w = dj.218 Chapter 7 Now we know from Theorem 61 that these conditions are also sufficient. w dry. then the condition dw = 0. In other words. = . There is another set of terminology for stating the same thing. as a basic theorem of differential geometry. by the criterion of the previous chapter.
with aj ay ag = ax We know how to find a function a on all of IR' with w = da.y) = r j(t. then Jxo defined in Chapter 2. in fact. . Consider the case ora closed I form w on �2: w= j dx + g dy. then it is exact. e) : IR ' is the inverse of the map Coo. 00) x {O}.o. we can define a different function e.{O}. The reasons f restricting ourselves to local results are now or somewhat different. we are dealing with a rather strange collection o f partial differential equations (carefully selected so that we can get integrability conditions). L (a. = e+21f in the region A . whose derivative at (a. agree on their common . It turns out that these necessary conditions are also sufficient: if w 1S dosed. t) dl. + O} x (0. is (I". = e in the region A. yo ) dt + On the other hand. b) has determinant equal to a #. Like our results about solutions to differential equations.DijJerential Forms 219 I n general. Consequently de and de. and e. . Recall that if L c IR ' is [0. the situation is very different if w is defined only on IR' . Then e. however. namely a(x.b) r+ {L {I" : r > fY41Y g(x. 21f) (a cos b. this result is true only locally. By deleting a different ray L.a sin b).
e) = 0 on lR' .L . . U C lR " i s contractible t o Po E U if U h a s t h e property that H(p. 1] + by l'vlore generally. �" is smoothly contractible to E �n . 1] + M x such that For example. w is not dJ for any C' function J : lR'{O} Indeed. so that together they define a If orm w on lR' . which implies that afjax = aelax and afjay = ao/ay and hence J = e+ constant on lR' L.L. clearly imply that this i s so] .) = 0 on lR' . This example shows that it is the shape of the region.L. but this is an abuse of notation.O) = Po H: x for p E M. 0 [0. if w J. since lR.) Clearly w is also not exact in any small region containing O. rather than its size. A computation (Problem 20) shows that w x = x'yy' dx + x' + y' dy. which is impossible. + dJ = de on lR ' . In fact. 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. then = =d + so d(f . (It is still exact i n a neighborhood of any point of lR' . that determines whether or not a closed form is necessarily exact. we can define lRn lRn H(p.{O}.I) = lp . dw = 0 [the two relations d(de) = 0 on lR' . I) = p H(p.L d(de.220 Chapter 7 domain. Nevertheless. S o w i s closed. The Iform w is usually denoted by de. w de only on lR ' L.{O}.
We will show that if M is smoothly contractible to a point. and at time ° it is the constant map. I] (for any manifold M). many other regions are also contractible to a point. I] by We claim that if w is a form on M x [0. then for each time I we have a map p r+ H (p. 221 1 (such a region U is called Of course.{o} is 1I0t contractible to a point.Dijje. the same result holds for JR" .{o}."tial F orms P E U implies po + I (p . I] with dw = 0. this result and our investigation of the form de prove the intuitively obvious fact that JR' .po) E U for ° :s I :s starshaped with respect to po). If we think of [0.. I] we define i. then every closed form on M is exact. at time 1 this is just the identity map. I) of M into itself. and pay hardly any attention at all to H. For I E [0.) The trick in proving our result is to analyze M x [0. but we will not be in a position to prove this until the next chapter. then . I] as representing time. : M + M x [0. (By the way.
Consider first a I form on M x [0.l)dl.xn O J[M . I] (Xl 0 1fM . . l)dxi . = i= x i=] i lfwe define g : M + IR by g(p) = l f(p. I) dl. ..L.. for Conve n n i. .. I]. = so (I) Wi (p. I) .* (L Wi dXi + fdl ) = L Wi ( · . .I)dl ) dxi .O) = 1r0 ' aaw · (p. the projection IT on the second coordinate).. dl. i=1 a i=l So dw 0 implies that f+ =  where /\ =' /\ Consequently. while J[M is the projection on M.a). .a)dxi. We will begin by working in a coordi nate system on M x [0. = r 10 axi n n n L Wi (p. I]. . i==] i=] Wi t .Wi (p. (x. axl dx'.Ll w. aWti dx'· dl + L. and if U) is a coordinate system on M.222 Chapter 7 we will see later (and you may try to convince yourself right now) that the theorem follows trivially from this..(p. Now for W = 2::7= 1 Wi dx i + f dl we have � � af dw [terms not involving dl] . (namely. It is easy to check (or should be) that xi 0 ITM by xi . We will denote nience.l)dl f ' af (p.. There is an obvious function I on M x [0. t) is a coordinate system on U x [0. then W I].O)dxi = L] (In0 1 aafi (P.a) denotes the function p wi (p.
for a kform w.Diff erential Forms then (2) 223 ag t aj aXi (p) = 10 axi (p. I] M w If a vector space V is a direct sum V = V] E !. I] we have (M x [0. I]) (p. t)dt. EEl kerHM •. g Equations (I) and (2) show that Now although we seem to be using a coordinate system. and hence also. (V ] + v. . ) = = w(v] ) w(v. of two subspaces. then any where w] (v] + v. it is easy to see (Problem 22) that we can write w dt.) w... I] as w] + w. we write the I form w on M x [0. Applying this to the decomposition (*). . Vk) = w ° ifsome V = w] + i E ker HM. . [0. . and ry is a (k . there is then a unique j with w. is really independent of the coordinate system. = j In general. uniquely as where w] (v].t) = kerH.11 ( V) can be written EEl V. Notice that for the tangent space of M x [0.). the function j.I)form with the (dt /\ ry) .
/\ . so we just have to consider two cases. (i{' / ) Clearly I(dw) d(Iw) 0 •:.. /\ d.)dl. it is easier to find a f ormula for i.224 Chapter 7 Iw on M as follows: Iw(p)(v" .*w . 1 7. ) PROOF. *w io*w 0. Since Iw is already invariantly defined. Actually. I ) dl) dxl (p) [f(p. * w io*w d(Iw) if dw 0.a=] axa dl /\ d'X· /\ diel ) (P) � a! L = = = = = it is easy to see that = = =  + = = = d(Iw) d ([ /(p. THEOREM. .n .0.io*w d(Iw) I(dw).*w . Then i. O)]dx l (p) i. For any kform w on M [0.. Now I(dw)(p) 1(. (Consequently. Then af dw at dl /\ dx l . . I) dl dx· /\ d .*Vk_.*w . i. . . .l)(i. *w(p) . (2) w = / dl /\ (r�i. We claim that dw ° implies that i. analogous property. we can just as well work in a coordinate system (X l .io*w d(Iw). I] we have i. j. . this proves the result in this case. .l) dl) dx l Xl =] o (p. at a!. The operator I is clearly linear. /\ dieik f diel . = = and + = .I)form = = x = = + = .Vk_') = 10' ry(p. . Define a (k . (I) w f d'Xi.f(p. I) .. /\ .*v" .�ik' f dl /\ d'X l .io*w(p).i. Since Iw 0. t ). . .io*w that holds even when dw #. . I(dw)(p) ([ � (P.
O) = Po Thus H 0 il : M + M H 0 io : M + M for all P E M. Amer. f evelY form w on U.0 = il * (H*w) . So But so w = (H 0 il )*(w) = i l * (H*w) 0 = (H 0 io)*(w) = io* (H*w). atwal Operations An answer to this question is provided by a theorem of Palais. we can find (Problem 23) an explicit formula for J(H*w). We are given H : M x [0. There are classical theorems about vector fields in 1R 3 which can be derived "'om the Poincare Lemma and its converse (Problem 27). Trans. Even though the Poincare Lemma and its converse fit very nicely into our pattern for basic theorems about differential geometry. it has always been something of a mystery to me just why d turns out to be so important.DijJerential Forms 225 1 8.141 . PROOF. Math. and originally d was introduced in order to obtain a uniform generalization of all these results. COROLLARY. Suppose we have any operator D from kforms to Iforms. = d(l(H*w» . 1 25. 92 ( 1 959). then every closed f orm w on M is exact. or we can solve the system of partial differeIltial equations w = dry explicitly in terms of integrals.io* (H*w) by the Theorem . such that the following diagram w . I ) = p H(p. Soc. If M is smoothly contractible to a point po E M. Corollary 18 is called the Poincare Lemma by most geometers. d(H*w) = H* (dw) = 0. while d' = 0 is called the Poincare Lemma by some (I don't even know whether Poincare had anything to do with it. X on Difjermlial F orms. is the identity is the constant map po. where we have an explicit formula for H. I] + M with H(p. •:.) In the case of a starshaped open subset U of IRn . Since the new form is given by an integral.
with few exceptions. then can only be some multiple of d. which we discuss in the next chapter. but nothing else. 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. can be a multiple of "integration". If k = I. then can be a multiple of the identity map. (As a corollary. Dl D D DD D lD . If I = k + I. d ' = 0. kforms on M � kforms on N Iforms on M � If orms on M Palais' theorem says that. Roughly. = o.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. these excep tional cases are the following.
. . </>n. . E 11 k (V) and w. This definition may be used even in a field of finite characteristic. . < a rk + /) ..1 ) = w (v. . . + ( . (c) Show that for W1 V . . Vk) = ap • (V1 .. Let Alt be Alt without the factor 1 / k!. Show a • p . (a) Show that This is sometime also called the inner product and the notation ivw is also used. (b) Show from this definition that w /\ ry is alternating. ry 2. then if j = i• . Vk .J (W1 /\ w. .).w . (b) Show that if V1.. Vk_1).. Vn is a basis of V with dual basis </>1 . Show that the set of all shuffle permutations is a cross section of S'.J w) = ..) (c) A permutation a E Sk +1 is called a s"u permutalion if a ( l ) < a (2) < . (Proving associativity is quite messy. Vk +/) = aEK L sgn a . . 4. .. Let S' C Sk+1 be the subgroup of all a which leave both sets { I . .. . . Vk).J ( v . ..flle a rk ) and a rk + I ) < a rk + 2) < " . k} and {k + I.any i. . .. and define W 7\ ry = Alt(w ® ry) . V1 . V . · .Difforenlial F orms 227 PROBLEMS 1. . . For v E V and w E 11k (V). .J w E I1k. . that 7\ is nol associative..) (a) Show that for any cross section K we have w /\ ry(v]. and w /\ ry = ( _ I ) klry /\ w. A cross seclion of S' is a subset K c Sk+1 containing exactly one element from each left coset of S' .J w. .. (V1 . .) .) = (v . E 111 (V) we have (Use (b) and linearity of everything.J W)(V1.J w). . < . Va(k+/) . . k +/} invariant.J W 1 ) /\ w..1 (V) by (v .J ( w . . . j #. .l lw1 /\ (v .. . w ® ry(va(l) . . E 11 1 ( V) and e E 11' ( V) . · .. Show that if we define then 3. . . .. . . we define the contraction v .... " . . (Try w.
is skewsymmetric (it is only necessary to check that interchanging V l and v. I : M IR form a coordinate system in a n neighborhood of p E M ifand only if dfi /\ . then every W E !..J w by (X. . . .. Ann(wl ) + Ann(w.4 = YJ W2 YJ. /\ W l .. #. 5. For any W E !.I)form X.1' (V) is decomposable. (a) Show that dim Ann(w ) ::. Show that if Wl is a kform.). (b) Every subspace of V* is Ann(w) for some decomposable w.1 l (V) is decomposable.) = {OJ if and only if Wl /\ w. we define + + . we can consider V /\ . then Ann(wl ) Ann(w. . 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. Vk !) . Show that functions fi.J w. (c) If W and are decomposable. In this case. Reformulate parts (a)(d) in terms of this /\ product. Vk !) = [(Vl . . 7.1k (V*).) if and only if W2 Wj l/\ for some (d) I f Wi are decomposable. we define the annihilator of W to be Ann(w) = {</> E V* : </> /\ W = OJ.1 k (V). (e) If V has dimension then any W E !. changes the sign of the right side). An element W !. (b) If </>i. then Ann(w l ) C Alln (w.J w.1 1 (V) .) + (</>3 /\ </>4) is not decomposable.)](v" . k.J Wl) /\ w. and that equality holds if and only if W is decomposable. then W = (</>1 /\ </>.. Vk !) + (. then X . (f) Since Vi E V can be regarded as elements of V**.J Wl ) /\ w.).J w)(p) = X(p). /\ Vk E !. Show that with this definition WI /\ W.](v" . (a) If dim V 3. = ( _ I )k! w. n 6.) = Ann(wl /\ w. i = 1 .. are independent. /\ l + Wl /\ w. + (I/Wl /\ (X. which is unique up to a multiplicative constant. . . . n . .(v" .) = (X. /\ dl (p) #..o.o. .228 Chapter 7 + W /\W. (e) Prove by induction that /\ is bilinear and that W l /\ W. + n + E ::. .. .J(Wl /\ w.1 k ( V) is called decomposable if W = </>1 /\ . /\ </>k for some </>i E V* = !. n n. Hin!: Look at W /\ w .I/[Wl /\ (Vl . . (f) If X is a vector field on M and W a kform on M we define a (k . . .J w(p).
' < kq. Show that there is a basis </>" • • • . . where w ' does not involve 1/11 or 1/12. . . (c) If w = L i<j a ijVti /\ Vtj . . Vn is a basis for V and Wj = 'LJ=l (ljiVj. Let I :s p :s n be fixed. . For H = hl < . " /\ Wn = L: C K VK . < hp and K = k l < ' . Let A = (aij) be an 11 X il matrix. If 10. j= ! Wt /\ . (a) Let W E !.. K . Vtn and </>2 involving Vt2. . i<j choose </>l involving Vth Vt3. </>n of V* such that Hint: If W = L aijVti /\ Vtj . . . (b) Show that the rfold wedge product w /\ . C K = det ( ap+ l 'k' : au. . .. /\ W is nonzero and decompos able.k] (a) If v" . show that the rank of w is the rank of the ma trix ( aij). 9. . ap. : al•hP ' . . Vtn so that W = </>l /\ </>2 + w '.itp ) .Diff erential Forms 229 8.. . Vn is a basis of V and show that n Wi = L QjiVj. . . show that B H = det ( al.1 2 (V). /\ Wp = L B H VH H Wp+l /\ . Thus r is welldetermined. and that the (r + I)f wedge product is O. .p. . and let q = n . old it is called the rank of w.h] ap. . . let Vt. . .
.230 Chapter 7 (b) Let H' = { I . we can consider sections of bundles constructed from p is 11k ([:n:.. (c) Prove "Laplace's expansion" det A 1/1" 1 1 . · /\ . h2. . a a (a) Show that if then . . (Cartan's Lemma) Let <PI . . hp . . ) h l . the bundle whose fibre at Since we can regard as an element of ( Mp )** . For example. ' . . .j . . . ./\ .H' VI /\ . .. L. eH.' ax l axil TM using 11 and other operations. . . . . .H (arranged in increasing order).H' K = H'. aIlY section of I1 n (T*M) can be written locally as h . (I . . 1/Ik E y satisf . . • ".w B H C H' . . /\ Vn " K #. . . . j= 1 k where bundle. (/>k . kl . kq = " . H' is the sign of the permutation {o eH. Show that VH /\ Vx = . where eH. In addition to forms. we can consider 11 k (C). if � = :n: : E + B is a vector 12. . " } . H E V* V* be independent and suppose that aji = aij ' Then =:: 1/Ii L Qjiq. .I (p)]*).
Vn is a basis for V. (b) Let :Tik [m] ( V) denote the vector space of all multilinear functions show that sections of :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.) (c) If :Tif ] ( V) is defined similarly. W ( V) denotes all functions ry : the form V* x . .W ( V) . k times I times . W (f) For those who know about tensor products V ® W and exterior algebras k ( V). Let :Tik[n .] · ··ill . (e) Suppose r:2n .] (TM) correspond to (even) relative tensors of type (1) and weight . x V* + r:2n ( V) n </>1 /\ /\ </>n . Show that sections o f :Tik [n.I . � x � + r:2m (v) . . then elements of r:2n(v) can be represented by real numbers [times the element V*I /\ . . with components e. . except that r:2m(v) is replaced by r:2m( v*). (Notice that if VI > . /\ v*nJ .w\ TM) correspond t o (even) relative tensors of type (J) and weight w. .. Interpret the relative tensor with com 'v' times V x . Similarly for :Tit. except that r:2n(v) i s replaced by r:2n . x V + IR which are of for some W E r:2 n ( v) . corresponds to the map Show that sections of :Tik [nl (TM) correspond to (even) relative tensors of type (1) and weight I . (d) Show that the covariant relative tensor o f type e ) and weight 1 defined i n Problem 410. I times k limes m • • • given by (</>1> . . these results can all be restated. . We can identify :Tik ( V) with A w an integer Since r:2m(v) '" Am(v*) '" [Am(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*) .w] (V) b e defined like :Tik [n] .i'1 similarly. </>n ) f+ ponents e'I .. ] . . • ..
I + C2 An. V + V is X(A) = det(A! .1k (V) + !.'. .1n(v) + !..1 ) for all A and all invertible B . a!k ()Jl ··iN ' lk .(trace A)A n . j]. (c) Let 8/.::// be as defined in Problem 45 (xiii).2 + · .1 k (V) . Show that C = det A .  n trix (a! l (with respect to some basis).'h. Chapter 7 and weight w and k 7it . then the function p r+ cdA (p» can be defined as a (2k)fold contraction of � A 0 · · · 0 A 0 8. (a) If V has dimension n and A : V + V is a linear transformation. m Noting that An(V) 0 · · · 0 An(v) is always Idimensional.wl (V) = 0 V* 0 01 V 0 0 w A m ( V*) .wl (TM) correspond to (even) relative tensors of type (1) n 7ik [m:wl ( V) = &/ V* 0 &/ V 0 0w A m ( V) 13. This problem outlines a proof that any invariant P is a polynomial in the polynomials CI . (This may be used as a definition of det A. respectively. show that . . . Call P invariant if PtA) = P(BA B . I}. more generally. 14. (b) Conclude that cdA B) = cdBA ).. .· . .A) = An .wl (TM) and 7if .232 Consider. then the map A * : !. . (a) Show that Ck = trace of A * : !. and A is a tensor of type G). . . .1n(v) must be multiplication by some constant c. If A : �l l l Ck (A) = � " a'1l a'2z . Recall that the characteristic polynomial of A : W. k times 15. We will .· k •l L i] . . · · + ( . Let P(Xij) be a polynomial in n 2 variables.jk V + V has a ma Thus. + (_ I)nc .. .I) " det A = An C l An.t + . For every n XI1 matrix A = (aij) we then have a number P(aij).) (b) Conclude that det AB = (det A) (det B). Cn defined in Problem 14. if 8 is as defined on page 1 30. show that sections of 7ik[n.
. (c) The discriminant D ( A ) i s defined a s n ih (Ai . . . . . . o . compare pg. Yn can be written as a polynomial in at .1 ) for all real A and real invertible B. \( 375. We will first consider matrices A over the complex numbers <C (the coefficients of P may also be complex). . . . up to sign. .Yt) = yn . . . . . . of the polynomial with roots Y b .atyn. . Yn) = R (adyl o . by continuity. Yn) in the variables Yl . . . (d) Show that P t A ) R(el (A). . Yn . that the equation holds for all matrices A over <C. . . they are the coefficients. Yn. . . (Regard or the equation as polynomial equations in the aij and bij . (This last conclu sion follows even if <C is replaced by some other field.) Now suppose that the coefficients of P are real and that PtA) = P(BA B . . . An of a matrix A are. . . by definition. . .(A b " " An). Yn Then there is a polynomial R such that Q (Yb . the roots of the polynomial X(A).DijJ rential FOWlS e 233 need the algebraic result that any symmetric polynomial Q( Yb ' .) = n' . Show that D(A) can be written as a polynomial in the entries of A . .I )"an. The polynomial R has real coefficients if P does. (a) Define Q(y l o . .0 ). an(Yl . . Recall that the at can be defined by the equation n i= 1 TI (y .1 + . . (b) PtA) = R(edA). (e) The same equation holds f complex A and complex invertible B. since the set where D # 0 is Zariskidense. it follows that ei(A) = a. " Yn) to be PtA) where A is the diagonal matrix Yl ( . en(A» whenever D(A) # O. . Since the eigenvalues A t . Conclude. this is "the principal of irrelevance of algebraic inequalities". . . an. Yn). . · · . Thus. Yn». + ( .Aj)'. . where Ai are the eigenvalues of A. where ai is the ith elementary symmetric polynomial of Yb . . . . . en(A» for all diagonalizable A .
. Wk V be given by Wi = L n • • . (b) Generalize Theorem 7 and Corollary 8 to kforms. . then s o Lxw. . . Vik )' i]. . . .Xi . .l)i+l w([X. Wk ) = l=ilL a/ W(Vi" <···<lk where a{ is the determinant of the k x k submatrix of (aij ) obtained by selecting rows .. E w 11k (V) show that W(WI. Xk+l ) L. Xk» = LxW(X1 . ..XJ. 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 . . 18.·. (a) Show that i f W i s a kf orm.. . .Xi .. Xk ) k + " (.Xk ). .234 1 6. show that X(w(X" .. Xk+l ) k+l l = " ( ..I ) i+ LXiw(Xl.. 17.» = Lx(w(XI. In Problem 514 we defined Lx A for any tensor field A . /\ j _ (b) Show that is (c) Using 514(e)..Xi]. . . . ik .. . . • • or Vn be a basis f V. .. . ... .. . . . (c) Check directly from (b) that the definition of d does not depend on the coordinate system.. (d) Deduce the following two expressions: dw(X" . and let WI. . L. j=] Q'jiVj . . . (a) Let Chapter 7 Vl . . For E .
Show that in order for there to be functions U I . If c : [0. . I aUi aU + j ai = '2 axj axi l == (LxI w)(X" . . (a) If w is a Iform / dx on [0. Xk+l ) (This may be used to give an inductive definition of d..'2 I ) . .J w). . . Hin!: Integrate the equation w . . . . l.d(X . 19.d(XI . ' functions on IRn with aU = aji . Let aij be . l .I ) {Xi(w(Xt .) (f) Using (e).. . .J dw = Lxw .A dx = dg on [0. Xk+ I ). a'aij a'aik axk ax! . 1 ] . i. sin 21fx).)  21f dx . . 1R' . .axj ax! a'a/j a'a!k == axk axi .{Oj be the inclusion. Un in a neighborhood of any point in IRn with ( ) it is necessary and sufficient that 20. Compute that Hin!: First set up partial differential equations for the functions fik aUk laxj . . . .. I ] to find A. . .. and use Theorem 61. Xk+l ) k+l 1 i+1 Xi.. (b) Let i : S l . == au) laxk "dO" 21. ...erential Forms Diff dw(XI . . 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. .A dx = dg for some function g with g (O) = g(I).. .J w)(X" .axj axi == x dy' + y dx x y' for all i.e. I] with /(0) = /( 1 ). and let a' = i*(dO). dw(Xt . . . . _ . == (At most places 0 = arctan y Ix [ + a constant]. Xk+t ) . show that d(Lxw) == Lx(dw). Xk +t » i=l 235 _ (e) Show that X . . S l is c(x) = (cos 21fx. . k . . show that there is a unique number A such that W .
(V\. . (a) Let U c IR ' be a bounded open set such that IR ' ... If W= WI 22. show that 24. . but U is not contractible to a point.and Wt(Vt. has degree f3 = k . Vk I ) = 0 ifsome Vi V.I Wi . (a) Show that every E /\ W.) can be written as a sum of forms w. .U is connected.1 (1 I k . where has degree 0' and WI EEl 0' E E E /\ E E x = on U.(VI . /\ d7a /\ . ic). J(H*w) 1 (IX) dl) xi. = L. . . 1] + U by H(p.ic (w. . the kth set being a finite union of squares containing the set of points in U whose distance from boundary U is ::.I)form such that WI (VI . a= ' l < " '<'k l _ .*. .Ik /\ .U i s connected. .236 Chapter 7 W 11k (VI V. va) = 0 if some Vi V2 W. and 0 #. . Vk ) = 0 if some Vi V. . (The converse is proved in Problem 89. Show that U is diffeomorphic to IR ' . then W can be written uniquely as WI + (b) If dim V. 23. 1/ k . . t( 1 ) . Let U c IRn be an open set starshaped with respect to 0. /\ dXik • Hint: (b) Find a bounded open set U C 1R 3 such that 1R 3 . and define H: U [0..l) Ip. vp) = 0 if some V. . W. . . . . dxi. VI .. . = I. .) Obtain U as an increasing union of sets. where W I is a kform and w. . . V. . and hence smoothly contractible to a point. is a (k .
. ' x det (b) Show that x . and express it in terms of the e*i. (a.t (JRn). .l o w) n = x .Di erential Forms ff 237 25. since the length of rays from 0 to the boundary of the set could vary discontinuously. the gradient of j. Vn _l with (Vt E �n.= " L aXl.t for all W E JRn. define a vector field grad j. L a X l. X Vn _l E �n V . Is U homeo morphic to JRn? (It would certainly appear so. . (First find all e. (a) If j : JRn + JR. using the expansion of a matrL by minors. b) L>i bi. . . on JRn by a n a n aj . . . .' . . x E \1 n . Let U c JRn be an open set starshaped with respect to O.) 26.) Intl"Oducing the formal symbolism i= 1 i=1 Dd . Let ( . . grad j = " . 27. ) be the usual inner product on JRn. x ej. .= 1 = (a) If Vt. x (c) For JR3 show that () � Vn . show that there is a unique vector Vt x . but the "obvious" proof does not work. aXl.
if we regard vp E IR np). 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. Conclude that V/(p) is the direction in which I is changing fastest at p . then X = grad I for some fU l Iction I : U > IR. where Du/(p) denotes the directional derivative i n t h e direction v a t p (or simply vp (J). Similarly.238 Chapter 7 we can write grad I = VI If (grad J)(p) = wp . for n = 3. we can write div X = (V. show that Du/(p) = ( v. curl grad I = 0 div curl X = o. 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 = curl Y for some vector field Y on U. w ) . we define the divergence (b) If X of X as (Symbolically. X). if div X = 0.) We also define. . (d) If X is a vector field on a starshaped open set U C IR n and curl X = 0.
Ii_ I approadlCS 0. �n ). < In = 1 of [0.C'{lI_ ' ) and vertical projection c2 (11) . To be precise.�) = [ Jdx +gdy lc E IIPII� O 239 .C2(t1_' ) e(O) c(I[_.) . with e(1) that > > n T which first arose fi'om very physical considerations. 1]. for example. C(tl) e(�I) :_ C2(t1) . We can choose points C(�I) on each piece by choosing points �I [IihIll For each partition P and each such choice � = (� . . . If we choose a partition 0 = 10 < . that is. (This is a complicated limit. When the differences Ii . if IIPII = m!'x( ti . consider the sum " S(P.Ii_I are small. and and.\C.e' (til) horizontal projection C'{lI) . as the maximum of Ii . .c2{1i_ l l. then the limit is denoted by l Jdx+gdy.' ) to C(li). y.c2{11_l ll i=1 If these sums approach a limit as the "mesh" I PI I of P approaches 0.c' (l1')] + g(C(�I))[C2{11) . denote the coordinate functions on JR2). Suppose.CHAPTER 8 INTEGRATION c: [0.lid.h x he basic concept of this chapter generalizes line and surface integrals. then we can divide the curve c into pieces. g: JR2 JR. �) = L J(C(�I» [C'(li) . 1] JR2 is a curve and w = J dx + g dy is a I form on JR2 (where J. then the equation lim S(P. �II) . the piece going fi'om C(li. i. each such piece is approximately a straight segment.
+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 .240 means: for all E we have > Chapter 8 0. it has a natural physical interpretation. CI (li ) . a a J.�) IiI Ii J g J dx + g dy dx. for all choices for The limit which we have just defined is called a "line integral".�) = LJ(�i.) IR2. consider the special case where S(P. the differential would be described as the work done by the force field on an "in fInitely small" displacement with components ely. while C2(li) . (In classical terminology. the integral is the "sum" of these infinitely small displacements.IiI . = yo+� In this case. = 0. YO)(li . C(I) (I. Yo). there is a Ii > 0 such that for all partitions P with II P II < Ii.C l (tid = Ii . the limit js the natural definition of the work done in the general case.) Before worrying about how to compute this limit.C2(ti_d S(P. If we consider a "force field" on described by the vector field � P.Iid· i=1 .
li1l f3i E [Ii"" Ii]. .Ii] C' (li) .integration These sums approach 241 lJdx + gdy = l ' J(X. then a c' (Ii) . so S(P.I)a.e.a) l ' J(xb + ( 1 x)a.yo)dx.C' (ti1l = C" (f3i)(li . denotes l' [J(C(I» C" (I) + g(C(I»)c" (I)] dl. In general.�i)a.C' (li' ) = C" (Cii)(li liIl Cii E [Ii. C' (li) . namely Physicists' notation (or abuse thereof) makes it easy to remember this result. we have. c' C y x . So S(P.=1 it looks like they should approach.yo)dx lb J(x. if Yo C(I) = (Ib + (1 .c' (Ii' ) = (b . for any curve c. YO)dX. L1 J(�ib + ( 1 .�) = L { J(C(�i» C" (Cii) + g(C(�i » C" (f3i)} (Ii liI). x c'.. by the mean value theorem. On the other hand.yo). of are denoted simply by and The components [i.lil l· i= = b These sums approach A somewhat messy argument (problem I) shows that these sums approach what (b .a) . Yo)(li .a)(ti .�) = (b .li' ).
we Can in troduce a more geometrical interpretation. Then the sums = L [j(e(�i» e ll(�i) + g(e(�i»)c21(�i)] ' (Ii . dY d1 J " In preference to this physical interpretation of "line integrals". 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. . RecaIJ that denotes the = de/dl(�i) e(ti ) tangent vector of c at lime de dl (�i) �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/ . (I) ] . Ii). e goes with constant velocity on each (IiI . The above in tegral is then written 1 / dx +gdy [o I [j(x'Y) dx +g(x'Y) di] d/ .
whereby the curve is divided into infinitely small parts. then dc dt (�i) = the constant speed on (Ii). kernel w (c(t» But if W arbitrarily). Ii).Ii) length of the segment from C(li 1l t o C(li) Ii {I I so [length of � (�i)] . dc/dl C C. dy. (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". and then extend W to IR'. /(C(I)) dx + g(c(I » dy. . The natural way to specify a continuously changing length along is to specif a length on its tangent vectors. t' [ i= length of dx. can be used as a w(C(I» w: = limit of the sums (*) C crt). (Ii . � (�i)] .integration 243 In this case. we should note that there is no Iform w on IR' such that 1 w = length of C for all curves c. (Ii .IiI) = length of segment from C(li1l to C(li). C is . we choose W(C(I» E !1 ' (IR'c (I» so that W(c(t)) (� ) = 1 . Before pushing this geometrical interpretation too far. the infinitely small piece at with components having length 1W c. It is true that for a given oneone curve c we can produce a form w which works for c.. for a general definition of the length of The line integral C. this is the modern counterpart of y the classical conception. If we choose any length of  is the length of and the limit of such sums.li1l dl c. �i E (li' .
given any w on which is everywhere nonzero.� there is no element of !1 ' (IR 2c(d) which has the value on all three vectors. + . used to define this generalized "length".244 Chapter 8 not oneone this may be impossible. Every sum S(P. any curve contained in an integral submanifold of /'. and w is a I form on M."hen we consider the sums (*) f a curve or M. namely c: [0. for example. This is no longer 50 dear . 1 2 c One property of line integrals should be mentioned now. ffi. For the present. in the situation shown belov. I] c. I] [0.2 wc op w ("the integral of w over (' is independent of the parameterization"). make sense even if is a curve in a manifold M (where there is no notion of "length"). will have "length" Later we will see a way of circumventing this difficulty. �') for c o p. 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. so it is clear from our first definition thal for a curve we have [0. I].2 0. and conversely.p = ker w(p) form a I dimensional distribution on IR . we note that the sums t*j. 1] 7 ffi. 1] 7 [ [/(c(l» C" (I) g(c(t))c2l(t)] cit.�) for c is dearly equal to a sum S ( P'. c: [0.even for a curve but in this case we can proceed right to the integral these sums approach. I] [0. l l c c: [0. because it is ob vious with our original definition and merely true for our new definition. ] [0. the subspaces /'. J] 7 ffi. In general. so we can define Ie W as the limit of these sums. [0. nor is it deal.2. if we are interested in obtaining the ordinary length of a curve.
1]. 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. 1 w = [ [/(e(I» e"(I) + g(e(l))e2l(1)] dl. for a general manifold M. so that formally we just integrate e*(/ dx + g dy). we can introduce a coordinate system for our calculations if lies in one coordinate system. and take the integral of h on [0. and a l form there is a similar calcula 2:::7=1 tion.integration 245 The result then follows from a calculation: the substitution + For a curve in ffi. . 1]2 JRn . If x and y m·e the coordinate functions on JR 2.2. 1]. e I is the standard coordinate system on JR.n . or break up into several pieces otherwise. we write c*(J dx + g dy) = h cll (in the unique possible way). which will eventually become our formal choice. For a pair of partitions and 10 < . 0 0 0 0 0 e: [0. < In of [0.. We are being a bit sloppy about all this because we are about to introduce yet a third definition. 1] JRn could be generalized to functions e: [0. . I] > JR we have 0 0 0 > > So s'" �ffQS . if we choose < . 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. I]) w= Wi dxi . < Notice that if ffi. to be precise. Everything we have said for curves e: [0. let � = e* (�) a ae ay = e* ( ay ) ..
.246 X �jj E [Sj_l . dXk [O. The inclusion map of It in will be denoted by It > it is called the standard kcube. Xk w w = ! dx l /\ . lj1 and w is a 2form on IRn. . . IRk. f W lO be O. are the coordinate functions. w X l . .(�fj ). ay(�ij ). ( = ) = .. To make a long slOry short. we have a special definition: a Oform is a function !. and . . We will Jet I Jo = = so that a singular Ocube is determined by the one point M. A Coo function It > M is called a singular kcube in M (the word "singular" indicates that is not necessarily oneone). For k 0.Sj_JJ(lj .Xk ) dX I . [O. If is a kform on It. [0. and c is a singular k cube in M. l lk where the right hand side has just been defined. . we now proceed with the formal definitions. . The limit of sums of these terms can be thought of as a "generalized area" of c. ° E IR. we define 1 w = f c'w.. then can be written uniquely as Sj w(C(�ij)) (�: (�jj).lj_l ) is a "generalized area" of the parallelogram spanned by ac ac a.. . Si_1 c: [0. . . * (�ij» ) (Sf . mImIC as far [O. c [0. then Cha pter 8 [0. llk in c1assical notation.f ! notatIOn attempts to w�ic? modern . Sj1 [ljI . IRo ck c(O) E IR Jk : [0. l lk [ l lk as logic permits If w is a k form on M. /\ dxk We define = J ! (X I . and for a singular Ocube c we define 1 ! !(c(0)).
II") by the change of variable formula . by the Proposition. I}" 0 det e([O. Then e PROOF.illlegmiion 247 > JRn be a oneone singular ncube with PROPOSITION. since p is onto . /\ dx" by Theorem 7 7 dx" by assumption > 2. Then e: [0. w w = J dxl /\ • • .11" = 0 e) (det [0. Jc leo p p' 2: 0. . 1]".1)II) PROOF. = 1 w f e*(w) f (J e') dxl f (J e)1 e'l dxl f J [0. •:. l w = f f. It [0. We have L w = f (e plow = f p*(e*w) p = f e*(w) 0 � I� � It [0. COROLLARY. /\ /\ .1]" p: [0. /\ . Let be oneone onto with det let be a singular k cube in M and let w be a kform on M. It [ w = [ w. Let det 2: on Let be the nform 1. • . /\ dx". 1]n e' 0 [0.11" = [O. By definition. C{[O. .
+ 3<4 ) + (2)« . if From a formal point of view.I )chain a c. provided it is orientation preserving.g. will also be denoted simply by Ct . and multiply them by integers. . > f l o c. I I . in accordance with Theorem 77. /\ dx k ). so that the change of variable formula will pop up). j = Li ajC.. I] M then l' I(c(t)) dt is generally i' l' I(c(p(t»)) cll. and which is supposed to be the sum of the various singular . . 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 .2<3 + 6c4. w(' define the integral of w over a kchain way: c J2::: uiCf [ W = I>lCj W. The k chain Ie. The corollary thus shows > I: c [O. I t is a Coo oneone onto map with det p' i' 0 everywhere (so that p' is also COO). <. it is called orientation preserving or orientation reversing depending on whether det p' 0 or det p' < 0 everywhere.e. 1\1orcovcr. functions On a manifold cannot be integrated (we can integrate a function on the manifold )Rk only because it gives us a form dx ' /\ . c: [0. A kehain is simply a formal (finite) sum of singular k cubes multiplied by integers. It M is called a reparameterization of c if p: [0. Our definition of the integral of a k form w aver a singular k cube < can immediately be generalized. purely formally. It [0. We add kchains.' lk For example.. diff rential forms are the things we integrate be e cause they transform correctly (i. . an ori entation reversing reparameterization clearly changes the sign of the integral. which is called the boundary of c.g. =I 2(c.248 > Chapter 8 > independence of parameterization. e. + <3 + C2) = 2C2 . 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. e.
._n . /11 . . . . .I)cubes and (the (i. . (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 . . . . .l) /11 . .:. .:. . +1 +1 with the indicated coefficients. .I . O.xil .1) .O) (x) = I" (x l . . . will not be is convenient to modif this idea. .Integration 249 (k . .l ). (x) = I" (xl . . o f the four singular I cubes shown o n the right. . i . but tbe sum.O) In) I]n.xn . then 1i). The boundary of y the sum of the four singular I cubes indicated below on the left. xiI .. . . I) = (X l . .. 1.O. . . . 1 .O)face and (i. . X Ii).I . In practice.I .xn.0) . . Xi . P I I 12. it for example.l) = (x l .I)cubes around the boundary of each singular k cube in c. ). l .x i . xi.xi . . .I ). I )face of as follows: If X E [0.xnI ) 1. .x i .
) ' C(I .) = C (Iv." )face of a singular ncube Now we define C(O) � C is defined by C(i. 1]0 Notice that for a I cube c : [0.). and for a Ochain Li aiCi we define c: [0.O) s  . C(i. C(l. for a singular 2cube c: [0.. C(1. which we will usually simply identif with the point P = c(O).S) .(R .(S PH (Q = 0.1 = O.O) We also have. the boundary of all nchain Li aici is defined by aC = L L ( _ I )i+.O) / C(I) C 0 Finally.P) p C(2. i=1 a=O.1) R . For the case of a Ocube > M.. 1 . 0).250 Chapter 8 The (i. 1]2 + C(2 .l These definitions all make sense only for n ::: I .Q) .I) C(1.. > M.1) a(aC) = (R . y we define a c to be the number 1 E JR. I] > M we have so a(aC) = ac = C(2.
Xn2).P) (i. /3.integration 251 From a picture it can be checked that this also happens for a singular 3cube.P) ' For x E [0.1 . Xj . and consider ( /D..) u . .) u. xj . xn2) j I i = In (x l . . . " . a2 = o. . PROOF Let i :0 j :0 n .) (X I . Therefore ' all terms cancel in pairs.j+I . .. Similarly.P) (x ) = J'J .1 . . Now i I a(ac) = a = L a=O. . it is clearly also true for singular nchains. we have: 3. C{ .) ( /. . . . . we have.P) · L . .P) = (CU + . .I. ( /U+I. . xiI . but even ac = O . . .1 (I )'+. Pl)(" . In general. a good exercise because this involves figuring out just what the boundary of a 3cube looks like. P) ( .)) L .1 . . a. . fj.. xiI .) U. Pl)(" . . For example. xj. xj .P) and ( CU+I . xn2). C(" ..) occur with opposite signs. . . a .. . .) for i :0 j ::S n . ( C(i.) for i :0 j :0 n I. . ... .I .: . from the definition ( /(�. and a(ac) = O. where CT and C2 are two I cubes  . Thus ( /.( P) (x » / = J(�. X i .Pp(�:. Since the theorem is true for singular ncubes. •:. .) U. 1]"2 . xi.Pl = ( /'( + . . . xi. .. x . . It follows easily for any j I singular ncube c that ( c( . If c is any nchain in M.\ (x » = J{ + .xj. . then a(ac) = o.) = J.xj. ..) U.P) (XI.•) U . . PROPOSITION. .I . .L 1=1 j=1 (t a=O. xn2) = In (x l .=1  In this sum. Briefly. . Notice that for some nchains c we have not only a(ac) = 0. this is the case if c = c) e2 .l (I)'+·+j+p(C(" . p.1 nLI P=O.
when c is a "closed" curve. a chain of the form ac is not desCllbed.. a closed curve encircling .252 with c. (0) = Chapter 8 C2 (0) and CI (1) = c2 (1). this terminology has been purposely chosen to parallel the terminology for chains (on the other hand. This parallel terminology \'\'as not chosen merely because of the formal similarities between d and a. by the classical term of "exact". we have seen that on IR ' . i. There is also a I chain c which is closed but not a boundary. If c is just a singular I cube itself. For example.{OJ there is a I form "de" which is closed but not exact. any k chain c is called closed if ac = O.e. The connection between forms and chains goes much deeper than that. Recall that a differential form w with dw = 0 is also called "closed". expressed by the relations d' = 0 and a ' = O. In general. namely. then ac = 0 precisely when c(O) = c(I). reciprocally. but is simply called "a boundary").
l)k f l(xl .x. First. 4. . dxk /(. . . /\ = jof dxi /\ . . dx·k � . Although it is intuitively clear that (" is not the boundary of a 2chain in IR 2 . . If w is a (k .1/ f [O. chains. dx. .{OJ. . .I )form on IR k and c = J k In this case. . 1 .l)k1 [ J(k . then Jc [ dw = [ w. 1)" I(x. Therefore = ( .)* U dx l j. xk ) dxl . /\ dxk ) if j i' i = [0. . w is a sum of (k . THEOREM (STOKES' THEOREM). . I . . l . . . O. /\ . . . . .x·1 . a. x.1 ) 1+ 1 [O. • • . Jae PROOF Most of the proof involves the special case where w is a (k .I)forms of the type and it suffices to prove the theorem for each of these. • . . . . .k ) d.integra/lOll 253 the point ° once. . . . . . and a.I)k + (. d. . .k ) dx.x·1 . We now compute.k if j i. the simplest proof uses the theorem which establishes the connection between forms.I )form on M and c is a k chain in M. a little notation translation shows that J[o.
. ) . /\ dxi /\ " . xk ) dx ' . . . .. /\ dxk ) /\ . chasing through the definitions shows that [ dw = [ W. . For an arbitrary singular k cube. . . . /\ = = f Di 1 dx' dx. . 1tk Jatk c'w = [ w . •:. . lc dw = [ c'(dw) = [ d(c'w) = [ Jtl. . )k + ( )' f l(x ' . . xk ) dx' . . (1' D (x' . . dxi . . xk ) ] dx ' . . lac .' l' . . . dxk = (_) 1. . Chapter 8 [ d(J dx ' }lk /\ . 0. . . Xk ) . . . . /\ dxi /\ . I ' I(x . xk) dXi) dx' . . . dxk ' ' = () 1 . . . .' f DJ. . dxk . dxk [O. . . . . . . . . /\ dx k [O. [O. . . . . /\ . . . . ) . .. )k Thus . /\ dxk ) J . /\ dxi /\ . .(x ' . . .254 On the other hand. . Jac Jatk Theref ore The theorem clearly follows for k chains also . latH 1tk [ w = [ c·w. dxi . . . . . [O. . )k By Fubini's theorem and the fundamental theorem of calculus we have [ d(J dx ' 1tk = (_I)i. . . .1 1 · · · 1 [/(x'.1 f . O.. )k (_I )i .
is not ac 2 for any 2chain c 2 . As an application of Stokes' Theorem. we wiII eventually obtain a 2form w on jR3 . For example. although closed. we could just as well have used it to show that w = "de" is not exact. 1] > jR 2 _ {OJ defined by C(I) = (cos 21f1.{OJ.{OJ > jR. de = [ de = But a straightforward computation (which will be good f the soul) shows that or = 21f.e(£) .  1c 1c dl w= = [ I= Jac 1 0 I = o.] Although we used this calculation to show that c is not a boundary.integration 255 Notice that Stokes' 1l1eorem not only uses the fundamental theorem of cal culus. if we had w = dl for some Coo function I : jR 2 .£) . For. 00) x {OJ) > jR: We have or x . then we would have 21f = \Ale were previously able to give a simpler argument to show that "de" is not exact.£) . [There is also a noncomputational argument.([0. sin 21f1). using the fact that is de f e : jR 2 . but actually becomes that theorem when c = / 1 and w = f. Tfwe did have c = ac 2 .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. I £) f de = e(l . but Stokes' Theorem is the tool which will enable us to deal with forms on jRn {OJ. we show that the curve c : [0. then we would have d(de) = 0 = o. and e(1 . w= x dy /\ dl Y dx /\ dz + z dx /\ dy (x 2 + y2 + z2) 3/2  .
The common number 1 w=l C2 1 w . and write port w C The only problem is that c.0 (' . 11" (it does satisfy det{c2 . the reason will be clear from the next result. ([0.lion jm!servillg (with respect to the orientation jJ. I]") . There are lots of ways of doing this. we first want to describe a way of integrating nforms over nmanifolds. . I]" ) .l W• CI C20{c2. In fact. for singular ncubes c : [0. However. each contained in some ('([0. : M be two singular ncubes which can be extended to be diffeomor > c. but they all turn out to give the same result. I]" M ('([0.([O. and ('. and the usual orientation on JR"). is not defmed on all of [0. For the moment we are keeping the origin of w a secret. on M. I]" ) n c.l a c) )' 2: 0. which is basic for our definition. I ]" 5. If w is an nf orm 011 M such that support w c [0.256 Chapter 8 which is closed but not exact. a glance at the proof of Corollary 2 will show that the result still follows. Let M be an nmanifold with an orientation {/" and let c" c.1 OC] ) w= . Assume that c. I]") n c. but a straightforward calculation shows that dw = 0. This is possible only when M is orientable. l C] w. because of the fact that support w C (' .{OJ.([O. will be denoted by > 1 with sup If w is an arbitrary nform 011 M. if <I> is a partition of unity subordinate to this cover. since ('1 and ("2 are both orientation preserving). IJ". then there is a cover I!J of M by open sets U. C2 then PROOF We want to use Corollary 2. THEOREM. w. I]"). phisms in a neighborhood of [0. 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 . ([0. then . where c is a singular ncube of this sort. I]") and c orientation preserving. •:. are both orienlo..
1nlegralion is defined for each ¢ E <1>. Likewise. if J: M n > N n is a diffeomorphism onto. then IM. and w is an nform with compact support on N.. We wish to define 257 We will adopt this definition only when w has compact support. If M C IR n is an ndimensional manifoldwithboundary and J : M > IR has compact support. of M we clearly have IM.!') W . /\ dX" = f f.!')W' However) we usually omit explicit mention of p. This is a simple conse quence of Proposition 1. (We really should denote this sum by for the OIientation  {/. fM Jdx' /\ .!') W M where the right hand side denotes the ordinary integral.. M . since support w can intersect only finitely many of the sets {p : ¢(p) i' OJ. Ifwe have another partition of unity \jJ (subordinate to a cover <9').1M. . E [ so that our definition does not depend on the partition. . which form a locally finite collection. then these sums are all finite. in which case the sum is actually finite.) \t\'ith minor modifications we can define J w even if M is an nmanifold withboundary. and the last sum can clearly also be written as [ 1M W = ¢Lc'pJM ¢ · w. then if J is orientation preserving if J is orientation reversing.
(x. and (y. If (x. .. /\ dx"l· . . If J : M > N is a Coo function between l1manifolds. imbedded in ]R 3 Since Mp can be considered as a subspace of ]R3p. i. . ( It is not hard to see that w is a volume element. . consider the Mobius strip M. U) is a coordinate system around p E M. One way of obtaining a volume element is to begin with an IIform � and then define w(p) = 1 � p) l . . 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). Suppose that w is a function on M such that for each p E M we have for some �p E g. . For example. V) a coordinate system around q = E N. then on U we can write we calI w a Coo volume element if is COO. for any n vectors VI . we can define w = Idx ' /\ . THEOREM. . /\ dyn J) = g: (g I ( a(y'o J) ) 1 0 f) .e. U) is a coordinate system. since there is no IIform � on M which is everywhere nonzero. det a. then for nonnegative V > ]R we have J*{g Idy ' J(p) PROOF Go through the proof o f Theorem 77. 77'.258 Chapter 8 Although IIforms can be integrated only over orientable manifolds. wp) = area of paralIelogram spanned by v and w. n ( Mp) . Theorem 77 has an obvious modification for volume elements: w(p)(vp. there is a way of discussing integration on nonorientable manifolds. not every volume element arises in this waythe form 1}p may not vary con tinuously with p.. pUlling i n absolute value signs in the right place . •:. Howeve. 10calIy.r · Idx ' /\ . w is of the form w = I�I for an l1form � . But this cannot be true on all of M. . /\ . . /\ dxnl J J for J ?: 0. .
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
' .c has compact support. Therefore c] = Cz = c and we have w = d(J where 1 . In fact. Let w be a I form on JR with compact support such that III w = o. PROOF. We know that there is an ( n . in particular for S". If M is a connected orientabIe nmanifoId. 1"' 1(1 .i /\ . then it is true for ffi. p) dl xi dx' /\ . dl 0 outside some interval [N. �(p) " = {. Since support w is compact. . . We wiII establish the theorem in three steps: = f N constant c] N on ( 00. There is some function 1 (not necessarily with compact support) such that w = dJ. For simplicity assume that support w e {p E JR" : ipi < I ) . . we have an explicit formula for �. from Problem 723. Moreover.n.(M) '" IF!. . ) . /\ . Let w = 1 dx' /\ .. /\ dx". N) and a constant Cz on (N. .269 9.c) Step 2.I )form � o n JR " such that w = d�. THEOREM. Step I .(_ J)i' (10 . (2) If the theorem is true for (n .. so 1 is a (I) The theorem is true for M = IF!. N]. then H. then it is true for any connected oriented nmanifoId.J. oo). /\ dx" be an nform with compact support on JR" such that III" w = o. . .I)manifolds. (3) If the theorem is true for JR".
. II this becomes . X Xi by Lemma 7. IR" > IRn w = d� = d(� .d(hr*A» : . I] be any Coo function with h = I on A and h = 0 in a neighborhood of o. g(J' = dA for some Hence (11 . r*(J' = r*(g(J'). Thus. r*(J'(p). !!. /\ dxi /\ . . r*(J'( p) II g: snl IR by g(p) = 10' u n1 f(u· p) duo A = {p IRn : I pi f = 0. Then hr* A is a Coo form on and � = r*(dA) = d(r*A). by the hypothe. . /\ dxn = = r JR. by Corollary 8 we have for the (11 .. Let h : [0. .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 /\ . > E > I I we have so on we have Moreover.. A �(p) = (10' u n . r f dx1 /\ .II JB w o.2)form A on snl . . g(J' = snl. � ) dU ) . .is for Stefl 2. /\ dxn L i=] = (t i r l f (u · � ) dU) .I )form g(J' on 1s"1 r � = (g 0 r ) .1 f ( u .
Choose an IIform w such that JM w i' 0 and w has compact support contained in an open set U C M. Choose forms Wi with support Wi C Vi n Vi+ l and JVi Wi =j:. Vr = V diffeomorphic to ffi.d(r*). since on A we have � .).n we h ave WI . In other words.d(hr*). . From this we clearly obtain the desired result.C1 W = dl}l Wz . with Vi n Vi+ 1 =j:.CZWI = dI}2 w' . we want to show that there is a number c and a form � with compact support such that w' = cw + d�. 0. .d(hr*). with U diffeomorphic to IRn . Step 3. diff eomorphic to IRn . . It obviously suffices to find c.l1. Since we are assuming the theorem for ffi. Using a partition of unity.:. with ¢.lille.CrWr. If w' is any other nform with compact support. Using the connectedness of M. we can write where each ¢iW' has compact support contained in some open set Vi C M with U. . we can assume w' has support contained in some open V C M which is diffeomorphic to IRn.) = o. . have compact support (C V.) = � .l = d1Jr : where all �.gral. and �.oll 27 1 the form � .w' = ("jW + d1Ji' for each i. it is easy to see that there is a sequence of open sets U = V" .) has compact support. o.
O . JVI [ Wi /J� [ Wi . . I I . UJ . Wi > 0.n orientation preserving. Since M is not compact. we have Iv. then PROOF. If M is a connected noncompact nmanifold (orientable or not). the numbers Ci = H. THEOREM. such that fu w i' 0 (this integral makes sense.272 Chapter 8 The method used in the last step can be used to derive another result.1 is orientation preserving. U4. then also !Vi+1 Wi > O. : PROOF Consider first an Ilform w with support contained in a coordinate neighborhood U which is diffeomorphic to JRn . U2. •. there is an infinite sequence U = UJ .(M) = o. If M is any connected nonorientable nmanifold. It obviously suffices to show that w = d� for some form � with compact support. Consider a sequence of coordinate systems ( Vi. . I t follows that Wi = cw + d� where c > O. using the orientation of Vi which makes Xi : Vi 7 ffi.1 is orientation rel1ersillg. we have W = cw + d� for for c  >0 J so (c  We can also compute Hn(M) for noncompact M. THEOREM. i)w = d� c =j:. Choose the forms Wi in Step 3 so that. Choose an Ilform W with compact support contained in an open set U diff eomorphic to JRn . Taking w ' = W. then Hn(M) = O.l are positive. there is such a sequence where Vr = VI but Xr OX1 . Xi) where each Xi 0 Xi+ l . Consequently. • . since U is orientable). Now if M is unorientable. 1 0.
we have seen that ¢u. i' 0. V. C V. V3 . .w. has support contained in V. C2C3W3 = where the right side makes sense since the V. Now choose nforms Wi with compact support contained in Vj n Uj+h such that Ju W. Let {¢u) be a partition of unity subordinate to !9.'s.c2d�3 + C. w. .+. we have W d�.w = d�. d�2 + C. Since any point P E M is eventually in the complement of the V. . V2 . where �.+2 Hence . . and V. d�2 + c. The cover !9 = {V) is then locally finite. . but the sequence is still eventually outside of any compact set). V. c2d�3 + c. + c. then for each V. Now it can be shown (Problem 20) that there is actually such a sequence V. + d�. are eventually outside of any compact set. i' o. C2W2 = d�. Cj+lWi+l + d1]i+l i 2: J .. If W is an nform 011 M. and forms �.C2C3d�4 + .+. whose union is all of M (repetitions are allowed. with compact support . d�. such that w Wi Then = = c. may intersect several Vj for j < i. + C.Integra/ion 273 of such coordinate neighborhoods such that V. . . There are constants C. + C. d�2 + c.. and such that the sequence is eventually in the complement of any compact set. = w = d�. + C. n V. U U U . .
{OJ) i' 0.. so 1* takes Z k (N) to Z k (M). where c� /* : the map 1* takes C E IR into the element of EB IR. ctEA PEB takes the element {cp) of EBpEB IRp to {c�}. I f M i s disconnected. since I*(d�) = d(J*�). with components Np. For example. If N is also disconnected.(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. then H. . If w is a closed kform on N.) C Np. consider the case k o. where each IR. If M is connected. This shows that 1* induces a map = also denoted by 1*: /* : Hk (N) > H k ( M ) . Then I*(c) = c o l is also a constant function. fJ E B. then I*w is also closed (dl*w I*dw = 0). '" IR. 1* also takes B k (N) to B k (M). with Cith component equal to c.EA E9 IRp > E9 IR. then HO( M ) i s isomorphic to the direct sum E9 IR.274 Chapter 8 SUMMAR OF RESULTS Y (I) For IRn we have (2) If M is a connected nmanifold. then HO(N) is just the collection of constant functions c : N > IR.1 (IRn . Ci E A . with components M. If N is connected. On the other hand. We also know that Hn. but we have not listed this result. then = cp when I( M. since we will eventually improve it. 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. = .. then 1* : HO(N) > HO(M) is just the identity map under the natural identification of HO(N) and HO(M) with IR.
p: Mp J pJ= J !*p w L J*w = (deg /) Lw { if + Nq is orientation preserving (using the orientations J.l. ) and (N. then we have a map M a = J. J. J M) and JR (and similarly for N) it J. M J1 !. THEOREM. so we really want to compare M J*: H" (N) L J*w > H"(M) and Wo for an nform on N.Integral. if Mp .on 275 A more interesting case. 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. L w. let For each P E sign J: M. and a number deg such that for all forms on N with compact support. L J*w = a . v). and the only one we are presently in a position to look at. is the map and N are both compact connected oriented nmanifolds. Until one sees the proof of the next theorem. but is proper (the inverse image of any compact If set is compact). There is when no natural way to make H"(M) isomorphic to JR. > N be a proper map between two connected 1 2. Let oriented nmanifolds ( J. it is almost unbelievable that this number is alwqys an integer.P for and Vq for Nq) is orientation reversing. Let q E N be a regular value of I (q). Choose one number such that w a is an isomorphism of H " ( Since r> follows that for eVeI). form we have w fM w w 1M J*wo = a · L woo with fN Wo Lw i' o.l.
since it is compact and consists of isolated points. •:. . To do this. This ensures that J' (V) c V. . /\ dyn where g 2: 0 has compact support contained in V.I (q ) is finite. xi i around Pi such that all points in Vi are regular values of J. Let J' (q ) = { p " . We want to choose a coordinate system (V. 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 . 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 !. y) around q such that J' (V ) = V. u . Iv w if J is orientation reversing.J(W'). .(V. Vk .' (q) PROOF Notice first that regular values exist. first choose a compact neighborhood W of q. . by Sard's Theorem. So Since J is a diff eomorphism from each Vi to V we have L J* w = {. Then J(W') is a closed set which does not contain q. and the Vi are disjoint. PEj. redefine Vi to be Vi n J' ( V). . · · · Vk . · · ·UVk • Finally. . Moreover. I . We can therefore choose V C W . Now choose w on N to be w = g dy ' /\ . . pd· Choose coordinate systems (Vi . . Then support J* w C V.I] tIllS proves the theorem . and let deg J = U W' C M be the compact set w' = J' ( W) .
1 3.J*w = (H o i) *w . O) = J(p) H(p. N are smoothly homotopic. J] we defined a (k .1.(H 0 io)*w = i. but we need to intro duce another important concept first. Since A . Two functions J. •. Then J = H o io g = H o i) . If f. we conclude that deg A = (. g: M H(p.l (p) consists of just one point.1 ) " . then the maps /* : H k (N) H k (M) g* : Hk (N) H k (M) > > > > are equal. We have already seen that A is orientation preserving or reversing at all points. THEOREM.io*w = d(Iw) + /(dw). H: M with x [0.gmlioll 277 We can draw an interesting conclusion from this result.*(H*w) . J] = g* . N with g*w . Recall that for every kform w on M x [0. We used this fact to show that all closed forms on a smoothly contractible man ifold are exact. Any element of H k (N) i s the equivalence class [w] of some closed kform w on N. g : M > N between two Coo manifolds are called (smoothly) homotopic if there is a smooth function As an immediate application of the theorem. We can HOW prove a more general result.inle.io*(H*w) = d(IH*w) + /(dH*w) = d(IH*w) + 0. we compute the degree of the "antipodal map" A : S" > S" defined by A (p) = po . But this means that g*([w]) = J*([w]) : . there is a smooth map H : M x [0.I )form /w on M such that i l*W . I ) = g(p) for all p E M. . 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 . I ] > N the map H is called a (smooth) homotopy between J and g. depending on whether II is odd or even. J* PROOF By assumption.
COROLLARY. there is a unique great semicircle Yp from p to A (p) = p whose tangent vector at p is a multiple of X(p). If vector field on S" . then deg j = deg g. I ) = Yp (I ) : . then we can construct a homotopy between A and the identity map as follows. If M and N are compact oriented nmanifolds and the maps j. then r 0 i : sn. . 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. this is perpendicular to p = (X l .{OJ is the inclusion. X"+ l ). But if there is a nowhere zero vector field on S".l .l 7 sn. (On S l this gives the standard picture. consider the retraction r: JR" . For another application of TIleorem 13. r (p) = p/ ipi. g : M > N are homotopic. and therefore in S"p . I . X"+l ) E S" we define = For n odd we can explicitly construct a nowhere zero vector field on S" . For H (p.{OJ > S". Il is even.) The vector field on S" can then be used to give a homotopy between A and the identity map. COROLLARY. •.l is the identity of S". .J)".l If i: S". . Define S" is ( . .l > JR" . .l . 1 4. . . p (Xl . .278 Chapter 8 1 5 . For each p. Since the identity map has degree ) . X2 .
M x {OJ c M x jR' is clearly a deformation retraction of M x jRt So Hk(M) "" Hk(M x jRl ) .integration 279 The map is. but it is homotopic to the identity.I)r(p) E jRn .{OJ) for all k.{OJ i 0 I" ( p) = p/ i p i H(p. we have and 1" ' 0 i' = (i 0 1" ) ' i ' 0 1" ' = ( I" 0 i)' = = identity of Hk( sn.{OJ > jR" . we can define the homotopy H by .// ' i 0 1" : jRn . = "r( p) (t=O) p(/=1 1 A retraction with this property is called a deformation retraction.{OJ. not the identity. The manifold for all I . We are now going to compute Hk (jRn . Whenever I" is a deformation retraction. A generator of Hn l (jRn . Thus. for the case of s nI C jRn .{OJ.{OJ) is the closed form r*(1'.{OJ) "" jR . identity of Hk(jR"  {OJ) So i* and r* are inverses of each other.I ) . l ) Ip + ( J . Thus In particular. We need one further observation. of course. we have Hnl (jRn . the maps (r o il' and (i 0 1" ) ' are the identity.
3 . .3 . 0]) B = w.4.O) B w.I is a constant on A n B. .4 .{(O. on A n B.{ (O. 0) = x x (00.0.4 . respectively).O) x [O. x w" B) I =° and A n B = [ w.0. 00») . there is a similar argument.3 .4 . If w is a closed 2form on w. I) A = w. If w is a closed I form on ffi. n= PROOF. I) and (0. then we obtain Iforms �A and �B with w = d�A w on A = d�B on B.{(O. 0]) [0. Let w be a closed I form on w. Induction on n.{O) = 0. (0. THEOREM. .n .3 Let A and B be the open sets (0. using A = w. 0.O.2 .{all r B so clearly JA . For ° < k < /1 . Th us w is exact.{(O.I). 0) x (00 . there are Oforms JA and Is on A and B with w = dJA l w=dB Now d(JA on A on B. 0. for on A on B A w = d(J .0.{OJ) = H k (snl) = 0. oo)}. We claim H I (w.c = Is on A n B.I we have H k (w. The first case where there is anything to prove is 3 .280 pter 8 Cha 1 6.c) B) w = d(J and JA .1) Since A and B are both starshaped (with respect to the points (0.
�B) = 0 and SO �A . So we can define a Coo form on IR n .d A. if <PBA denotes { <PBA O and similarly for <PAA.�B = dA for some Oform A on A n B. on A n B on A .dA + d(<PAA) is a COO form on A is a Coo form on B. then <PBA <PA A On A n B we have �A .J) dA + d<PA = �B + d(<PAA).integration 281 Now d(�A .d<PB /\ A = �A . To circumvent this difficulty. Unlike the previous case. •:. = �A + (<PA .{OJ = A on A.(A n B). . note that there is a partition of unity {<PA . The general inductive step is similar. since this is not defined on A . and �B + d(<PAA) on B. we cannot simply consider �A . <PBj for the cover { A . = d�B /\ A U B by letting it be �A .d(<PBA) on A on B. Bj of IR J . Clearly.<PB dA . on A n B =) =0 Now. w = d�A = d(�A .d(<PBA» = d(�B + d(<PAA» so w is exact.d(<PBA) = �A .{OJ: <PA + <PB d<PA + d<PB support <PA support<PB CA C B.
We know that w = d� for some (k . PROOF The proof that H�(w..2)form A on A . Let B be a dosed ball containing support w.n .. which we will need in Chap ter 1 1. 0 < k < n.282 Chapter 8 We end this chapter with one more calculation. Then d(JA) makes sense on all of w. Let / : w.n. Let w be a kform on w..n with compact support.n and w = d� = d(� .n . the form � .n . Since A is diffeomorphic to O = w = dry w..d (/A) dearly has compact support contained in 2B .2B.I )form � on w. THEOREM..n > [0. For 0 � k < 11 we have H: (w..B we have d� = O..d(JA» . I] be a Coo function with / = 0 in a neighborhood of B and / = I on w.n) = o. 1 7 . where 2B denotes the ball of twice the radius of B..n) = 0 is left to the reader. Then on A = w.I we have from Theorem 16 that � = dA for SOme (k . •: ..{OJ and k  I < 11 .
that cannot exist if is unbounded.�n) �i M. . [a. P) = Ll11i(J) ' (Ii IiI) i=l J U(J. . [S.Ii.. 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] . li]. Let [a. . letm Ii] = length of the limit is called the Riemann integral of (a) We can define even if is not bounded. b] > IR be bounded. b]. (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. < s } and Q = Io < . b]. [a. S. P. and define is an ntuple = similarly. {so { [Ii. b]. For a partition = { o < . b]. P.lil ) . I . [a. P � n L(J.I . P) = L Mi(J) ' (Ii . S J on J J 1I).= ] S [a. We call J Darboux integrable if the sup of all L (J. A choice for = with E [ I We define the "lower sum". . P). (d) Let J M on Let P = < . "upper sum".�) (b) If J is continuous on then J is Riemann and Darboux integrable on and the two integrals are equal._] . b] ei 111 S s [a. [a. Show however. P. < In} be two partitions of For each i = I. P) equals the inf of all U (J. t] shaded lengths .]'S cmltained in [Ii.�) = L J(�i)(I' IiI)· Clearly L(J._I . let = on [Ii_I. The Riema1l1l integral UeJJUS the Darboux integral.b]. I ti_l .I .add up to e. . b]. b].�) U (J. .=1 S(J. P. Mi(f) Ii]. . .lnlegralion 283 PROBLEMS I. [a. S(J. < be the inf of of [a.) [a. b]. . .sum of lengths of all [s. n. s.�) [a. P) S(j.i� o S(J. I � I I I. . b]. Pl.] which are contained in [li_] .
(b) If c : [0. b]).{OJ be defined by (c) Show that n is unique. P) :s U(J.M.{OJ is any cUlve with c(O) c(I).{OJ such that ac. sin 2Jr/ ) on [0. I] JR2 . Q) : Q a partition of [a.£7= 1 e. (f) If f is Darboux integrable on [a.CI . I] n JR2 . ° as I PI I 0. P) = inf{ U (J. i=l There is a similar result for lower sums.n .1 ) = [ fg· Ja i=l Hint: If l g l :s M on [a . P) = sup{ L (J.n (t ) = () = (cos 21f1. equals > > H O n UPU+O 2.b]) H ). b]. =l :s U(J. b].11 . b lim L f(�i)g(�'i)(li . I]. defined as a limit of sums. (a) Show that there is a singular 2cube c : [0.284 Chapter 8 Show that.{OJ. b]. For 11 an integer. then 1J(�'. where C I 3. Q) : Q a partition of [a. )1 :s MIJ(�'i )f(�.m) L ei. > R > 0. Q) + (M . then f is Riemann integrable on [a. show that there is some such that is a boundary in JR2 .)g(�'. It is called the winding number of C around 0.)e. b]. R sin2n1fI). Let f and g be integrable on [a. let CR.il� o L(J. then U(J.n : [0. and deduce Darboux's Theorem: ). Compute fe de = f[o. li].11 = > CR"n  n > = C . I I c· de. if Mi denotes the sup of f on [If_I .� U(J.)f(�i)g(�'.�' for P. IF JR2 . CR2. Q) + iL(M . Show that fOJ choices �. )I. (h) Show that fe f dx + g dy. (e) Show that . (g) (Osgood's Theorem). and CR. (R cos 2 1fl.
c = C o a. . . . Define cR. : " n I > " n by 7 ao(x) = ([I . . · + . .. . Then we define and for singular nsimplexes c. xn . a a (a) Show that if R is large enough. xn . a. x'. ! .I ) a.Integration 285 n l /(2) = Zn ( 1 + . zn z (b) Show that /(z) = ° for some z E <C ("Fundamental Theorem of Algebra").x' . +an . as indicated in the next Problem. /(2) zn +a l znI + .(" n . As before. . n i s the boundary of a chain in <C .I ) in " n .. Define a. 1] > <C by cR. Although Stokes' Theorem becomes more complicated. .CR. I (I)] n . x l . . Let /: <C > <C be a polynomial. define (a) Describe geometrically the images (b) Show that a2 = 0. ! : [0. . ! .+ . ! is a boundary. I ' = 5.{OJ. n C JRn be the set of all x E JRn such that A singular nsimplex in M is a Coo function c : 6.'£7. where n :0: I . . n (t) = [CR.(x) = (x l . n M. . let r : " n > JRn be the inclusion map.) . Some approaches to integration use singular simplexes instead of singular cubes.co. Hint: Note that CR".J)'a. . . ! = / O CR . and an nchain is a formal sum of singular nsimplexes.I .I ) a. then cR. Hint: If /(z) i' ° for all z with Izl :S R. Let !J. i=O 0 < i ::: 11. aC = L (. x' ] . and write 4. ..c. there are some advantages in using singular simplexes. O. then CR . .
C2 is a boundary. Hint: If au = L. show that CJ . define c: "' k +1 > IRn by IV!orcover. • . that is. I] > IR 2 be a closed curve. what can be said about c for chains c in the obvious way. Every x E "' k + 1 can be written as ' IX . .a. and x ' E ao("' kl . I]) is a point. For any singular k simplex c : 6. and prove that 1 /" dw = [ w. We then define (c) Show that we do have c = au + c' where c' is degenerate. ? . [ dw = [ w lc Jac for any (k .286 (c) Show that if w = J dxl then /\ .c. IRn . (b) Let c : [0. 6. c'([O.I )form w. k 7 c(x) = I .J )form on IRn . dxi /\ . x' is unique except when t = O. . for 0 :0 :0 I . using either simplexes or cubes. a. c(x ' ). Show that c is not the boundary of any sum u of singular 2cubes. Ozapter 8 /\ (Imitate the proof for cubes..) (d) Define Ie w for any kchain C in M and kform w on M. (d) If C I (O) = C2 ( 0) and CI (l) = C2 (1). L. (a) Show that a c = 0 implies that c = at. /\ dxn Jalll is an (n .
(See Problem 7. 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. > dl. It is > 1 w = l(c(1» . (e) If M = U where U and are simplyconnected open subsets with U connected. (b) Sl is not simplyconnected. then any smooth map I : Sl > M is smoothly contractible to p. (a) Let U C IR 2 be a bounded open set such that IR 2 . then H I (M) = O.U. Let w be a J form on a manifold M.) Hint: If p is in a bounded component of IR 2 . (f) If M is simplyconnected. [Actually. 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. Show that U is not smoothly contractible to a point. (This gives another proof that n is simplyconnected for n > J .U is not con nected.] (a) If M is smoothly contractible to a point.) Hint: Given I : Sl > M. (d) If M is simplyconnected and P E M. then for any not hard to show that for a manifold we may insert "smooth" at both places. (c) This is false for open subsets of IR 3 .) 9. Hint: If we do have w = curve c we have 7. show that there is a curve in U which " surrounds" p. partition S l into s a finite number of intervals each of which is taken into either U or V. Show that w is exact. then M is simplyconnected. then M is simplyconnected. (c) sn is simplyconnected for n > J . (Converse of nV UV V Problem 724. Suppose that Ie W '" 0 for every closed 8. (b) A bounded connected open set U C IR 2 is smoothly contractible to a point if and only if it is simplyconnected.l(c(O» .Integration 287 curve c in M. Hint: Show that a smooth I : Sl > s n is not onto.
Iwl < 00. L dw . w converges absolutely. J ) and i s a Oform whose support i s not compact. M (b) Show that L 1M ¢ . Find two partitions of unity <I> and such that L¢E ¢> f ¢ . using of its boundary. n + J) be closed sets. there are only finitely many " which are nonzero on support ¢ . and let o w . (d) Let A n C (n.) 0) Similarly. 1 2. J) by functions with compact support to see just why the proof of Stokes' Theorem breaks down in this case. w = L L 1M " + w. (Note that for each ¢ . We define this common sum to be fM w. Let I : IR IR be a Coo function with fAn 1 = ( . Let be an nform on an oriented manifold M n . I dx and L E 'l' fIl. w = x dy. and show the same result with w replaced by Iwl.) (c) Show that L"' E 'l' f . ¢E ¢> cpe¢ cpe¢ l/re\ll cpe¢ l/fe\ll > 1 0. Let <I> and lj! be two partitions of unity by functions with compact support. w = L lM .. find a counterexample to Stokes' Theorem when M = (0.y)1 (No computations needednote that equality would hold if we had the entire boundary. Show that w.. (e) Examine a partition of unity for (0. Following Problem 712." laM Problem 1 0. together with a proper portion lj! « x. and suppose that w I I . define geometric objects corresponding to odd relative tensors of type (�) and weight w (w any real number).. · Iwl < 00. " . even though both sides make sense. and that M L lM ¢ . "' to different values.I )"/n and support I C Un A n .288 Chapter 8 L 1M ¢ .y) l(x. · w . (a) Let M be E IR 2 : < J }. (a) This implies that L¢ E¢> f ¢ . I dx converge absolutely Il.
.{OJ. Show that de is 0 at some point. (a) Let M" and Nm be oriented manifolds. . and e is an (n .{OJ is just r.aMI. respectively.{OJ > JR" . on M and N. . WI . Account for the factor I/Ipl" in Lemma 7 (we have r* (vp) = (J /lpl)vr (p) . We will orient M x N by agreeing that V I . VI . but this only accounts for a factor of 1 /Ipl" . )�.I . Wm are positively oriented in Mp and Nq. V" and W I . and let w and � be an nform and an mform with compact support.J )form on M. " " Un. 1 7. vn. M2 C JR" be compact ndimensional manifoldswithboundary with M2 c MI . then where g(p) = (c) Every (m + n)form on M x IN h (p. · · · . since there are n .Integration 289 13. (Note tbat the map i o r : JR" . . 1M IN JMxN JR is Coo.q) '" Mp Ell Nq if VI . Suppose M is a compact orientable nmanifold (witb no boundary). . show that (b) If h : M x N > [ 1f1*W /\ 1fz*� = [ W · [ �. . q).1 vectors 1 6. respectively. . If 1f1 : M x N > M or N is projection on the ith factor. Let MI . . .d· 1 5. " (p. . . . Show that for any closed (n .J )form w on MI.) r * (1' = r * i*(1 = (i 0 r ) *O". . Use the formula for r*dxl (problem 41) to compute r*(J'. Wm is positively oriented in (M x N )(p. N is h 1fI*W /\ 1fz*� for some w and �. 1 4. considered as a map into JR" . . ) = q 1+ h (p. .
For this n 1 9. Show that E JR"p and let v E JR"p be (Ap)p for (b) Let M C JR" . we define ¢(x. 1M 1c(M)ns2 M. Z E (00. .�l E JR3 except those with x = 0. For aJJ (x.. Wn2) = o. " " Wn2 some A E IR. and let C(M) = {Ap : p E M . .J )manifoldwithboundary which inter sects every ray through 0 at most once. (c) Let M C JRn{O) be a compact (11 . Let WI . (a) Let p E JR" . y. W I . . r *a' (v.{OJ . . A ::c OJ . 0]. y . y. The latter integral is the measure of the solid angle sub tended by reason we often denote r * (J' by de .{OJ be a compact (n .I )manifoldwithboundary which is the union of segments of rays through O . �) to be the angle between the positive zaxis and the ray from 0 through (x. y = 0. zl. Show that JM r *(J' = O. M Show that [ r *(J' = [ r *(J' . . . C (Ml C(Ml n S' D·.290 Chapter 8 1 8.
Integration 291 (x. 0]. ¢) is a coordinate system on the set of all points (x. y ..cos ¢ de ) . then d¢(v) = 1 . (b) If v (p ) = Ipl. (the minus sign comes from the orientation). z) in ]R3 except those with y = 0. ¢ to [certain portions h: S 2 > ]R is h(x.. and e is considered as a function on ]R 3 .z) arctan y/x. then de(wp) = Y X2 + / (d) If e and ¢ are taken to mean the restrictions of e and of] S 2. y. then (v. Z E (00. y.. ) . . then where . z) = vix2 + y2 (J' = h de /\ d¢. y.. y) (a) ¢(x. z) E S 2 (r). If w points along a meridian through p = (x. 00) or with x = 0. X E [0. e(x. (c) If v is a longitudinal unit tangent vector on the sphere S 2 (r) of radius r.z) = arctan( vlx 2 + y 2 /z) (with appropriate conventions).y. e. (e) Conclude that (J' = d( . y = 0.z) (x. y.
for the form (J de IR3 is just 1C*de. and the sequence is 2 1 . Let > Nn be a proper map between oriented IImanifolds such that p 7 Nf(p} is orientation preserving whenever p is a regular point. (a) Show that a polynomial map ce > ce. S 2(1 p l). U U U . then either 1 is onto N. . ( + nl l.y)+i ax (x. = and Show Hint: Choose w to be a real (d) Conclude that av /'(x+iy) = aau (x. Show that if N is connected.y)+iv(x.y) i ay (x..y).) = IIZn111 +(11 . l w I(x+i ) u(x. dBn_1 IRn1 dBn IRn Vi n VjVI V2 V3 . or else all points are critical points of J. I/'(x iy )12 DI(x. ·+a _ . y au h.y). .. express 20.To[/(" + w) yl(z)]lw. so that is the projection. de = r2*i*(J. ih. 0 on {OJ in terms of on IR2 r2*de = de ..y) == for realvalued functions I(z) zn+aI Zn1 + I'(z) u v.292 (f) Let {OJ on Show that If > for the form (Jzapter 8 > IR2 1"2 : IR2 1C : IR3 IRde2 Sl be the retraction. . Prove that a connected manifold is the union the are coordinate neighborhoods.)a "n2I: . 22. given by = is proper � J). where i' 0. 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).. (b) Let J Show that we have i where varies over complex numbers. Vi r*de de . with eventually outside of any compact set. and then to be == + det . {OJ.y) x aav (x. = J* : I: Mn M (e) Write that . +an/'(. v dB3 r*(J' d(cos(¢ r) de) = d(cos¢de). and the fact that p vr(p)/lpl for tangent to (h) Conclude that = = * (v ) = (i) Similarly.
(c) Show that w is constant in a neighborhood of p. (f) There is a still simpler argument. (a) Show that this definition agrees with that in Problem 3.I )sphere L: around p such that all points inside L: are in JR" . not using Problem 2 1 (which relies on many theorems of this chapter). For p E JR" .{p) > L: be the obvious retraction. and conclude that I takes on all values.M.I )plane. then the number of points in I.M. (d) Suppose M contains a portion A of an (n . choose an (11 . Define the winding number w (p) of M around p to be the degree of rp lM. Show directly that if I : M > N is proper. give another proof of the Fundamental Theorem of Al gebra.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. Conclude that w is con stant on each component of JR" . (b) Show that this definition does not depend on the choice of L:. Let p and q be points .M.I c JR" be a compact oriented manifold. while DI is the linear transformation defined for any differentiable I: JR 2 > JR2 (e) Using Problem 21.Integmtion 293 where J' is defined in part (b). Let M" . Let I"p : JR" . 23.
1] > be a regular value of Show that Let (mod 2). H'(p.x» · p x at H(O. The number of points in its boundary is clearly even. g: H: N.x) x.t) /(p) t H'( p.t) �is an equivalence relatiOll. #11('1) I#rl (q) '" #gI (q) Hil l: H1(q) 24. then IRn . if M is orientable.294 Chapter 8 (a) Let of points in is a compact I manifoldwithboundary. for in a neighborhood of I . and leaves all points outside the unit ball fixed. more gencrally. but on opposite sides. = H': M x [0. Let M and N be compact nmanifolds. and let homotopic. and 0 elsewhere.g:g. close to this plane. For two maps M > smoothly homotopic to = I. x) . 25. H(p.) ¢(H(t.) (b) Show. that this result holds so long as 'I is a regular value of both and I. I] > N such that for in a neighborhood of 0. = == x H(t. H.t) g(p) t g t. let IR x IRn > IRn satisfy ¢: (Each solution is defined for all by TheOI·em 56.M has at least 2 com ponents. q E N I (q). 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. x w(q) w(p) M > N be smoothly denote the (finite) number 1 g.) Show that each 1+ is 1I diffeomorphism. (b) Let IH:" > IH: be a Coo function which is positive on the interior of the unit ball. N we will write / :::e g to indicate that 1 is (a) If l :::e then there is a smooth homotopy Q)) 26. For E s n. (Show that I"q l M is homotopic to a map which equals 'p l M on M . More precise conclusions are drawn in Chapter I I. Show that = ± 1.) (e) Show that. which is smoothly isotopic to the identity. (This is one place \·vhere we use the stronger form of Sard's Theorem. by a smooth homotopy M x [0. g. t. The next few Problems show how to prove the same result even if M is not orientable. 1 H p (a) Being smoothly isotopic is an equivalence relation. in general. H: p aH(t.I .A and which does not take any point of A onto the point in the figure.
then there is a diffeomorphism I: M > M such that I(p) = q and I is smoothly isotopic to the identity.. it follows that there is a Coo family {cp. which is generated by {X ' } . 1 ] ..WI · . + w. and the argument which was used in the proof of Theorem 56. (p» (Xlf)(p) = lim CP'+h P» .J )manifold. (d) If M is connected and p . q2 E N we have 27.. i. O For a family w.1 (q) is defined in Problem 24). Let {X' ) be a Coo family of Coo vector fields on a compact manifold M. (e) Use part (d) to give an alternate proof of Sl£p 3 of Theorem 9. = (l . and I : M > N. I1 h.* (Lx' w. of diffeomorphisms generated by {X'} satisfies for ali I if and only if LXI Wt = Wo .e. then JR" .Wt .lnl£gmlion 295 (c) Show that by choosing suitable p and I we can make H (1. = cp. (b) Let Wo and W I b e nowhere zero nforms o n a compact oriented nmani fold M. then for regular values ql . q E M. of kforms on (where #1. then XI (p) will denote 1CM * X(p .1 )Wo + IWI · Show that the family cp.. for any Coo function I : M > JR we have I( ( I(cp.. (To be more precise. show that if M c JR" is a compact (n . *w.M has at least 2 components. o h� h (a) Show that for �(t) = cp. M we define the kform Wt +h . and define w. (f) If M and N are compact nmanifolds. This number is called the mod 2 degree of f.). (g) By replacing "degree" with "mod 2 degree" in Problem 23.:.0) be any point in the interior of the unit ball. we have �. suppose X is a Coo vector field on M x [0..) From the addendum to Chapter 5. with <Po = identity.. Wt = I " . O. } of diffeomorphisms of M [not necessarily a J parameter group] .
'v)]3 du dv. and I( M ) g( N) = 0. g) (c) For I. g) = 4 = i(j. so that Wo . where M and N are compact oriented manifolds.I r (g(q) .g (P. g) = deg Cif.l(p)I ' i(f. n = k + 1 + J. O) = g(q) K(q. J ] > JRn and K: with JRn be smooth homotopies H(p. g). Let I : M k > JRn and g : Nt > JRn be Coo maps. where M x N is oriented as in Problem 18.O) H(p.g . J A (u v) 1C o o [r (u. J) = g(q) j(p) (H(p. n Cif. show that this holds if and only if 28.W I = d A for some A. t ) : p E M} (K(q . g) = (Jlt + li(g. J) such that = = I(p) K(q . Show that there is a diffeomorphism II : M > M such that Wo = J.  1 1 10 10 . g : S l > JR3 show that i ( f. q ) = We define the linking number of I and g to be '" g(q) I(p) Ig (q) .g : by M x N > s n. Define (d) Suppose that JM Wo = JM W I . (b) Let H: M x [0. J ] > (a) i(f.I(p» C JRn  (OJ  Cif. i(f. t ) : q Show that n E N } = 0 for every I .296 Chapter 8 (c) Using Problem 718. J ). N x [0. *W I ' .
(U» (/1 )'(U) (g3 )'(V) (g 2)'(V) A (u.(a.c')'J> P Hint: First show that if and M = aN. (V I )p. Show that (VI )p. g) = 0 if J and g both lie in the same plane (first do it for (x.b. (a. (a.c') p triple wp. . c) = 41C for (a. b' .c) = 0 for (a. c) is on the same side as a vector wp E ]R 3p . (a) For (a.hltegmtion where 297 1" (u.b.b'. v) = Ig(v) . b. (a' . (d) Show that C(/.c)'I> p (a'. b. The next problem shows how to determine C(J. M C ]R3 and (a. g) without calculating.J'(u) ( ) (the factor ) /41C comes from the fact that Is' (J ' = 41C [Problem 914]).c) .J(u) 1 (P)'(U) (/3 ). c') be points close to p E M. v) = det (g l )' (V) g l (V) . c) rt M. (V2)P is positively lim g.M g.JI (u) g2 (V) .c) and (a'. 29.  .b.Mp for which the Wp 1M dEl(a. (V2 )p is positively oriented in ]R 3p when oriented in Mp. b. y)plane).b'. Suppose (a. c) rt N. b.b) dx /\ dz + (z c) dx /\ dy [(x _ a) 2 + (y _ b) 2 + (z _ C) 2]'/2  g. then g. c) . c) = Let (a. on opposite sides of M. b.g. (a.c) /(a'.b. c') = 41C.a) dy /\ dz (y .c) _ For a compact oriented 2manifoldwithboundary let  (x . b. (a. {a. b.c) E ]R 3 define dEl(a.b'.b.p (u) g3 (V) . b . c) E N . b.
a) dZ ab (a.the number of other intersections. pg. n = .(Y . y.+ be the number of inter sections where dgjdl points in the same direction as the vector wp of part (a). (An M with this property always exists. . ac I(x.=  J 41T (c) Show that ag.lSI J l(x. c) .(Z . . . Topowgy qf 3Manijolds.. Show that hut the surface on the rightincJuding the hemisphere behind the pJane of the paper is orientabJe. z) 1 3 X . .).C) dy aa (a. b .b) d. z) I' I _ 1 ( ( ( l g*(dg. z) 13 ag. c) .lSI J l(x .n. c) = S J* . y.(a.. See Fort. b . s' ) ) ) (d) Show that n = i(f.b) dX ag.[ * y .. that when g(l) = P E M we have dgjdl rt Mp..C) dx . Compute i(f.(X .g) for the pairs shown below. y.) Let g : S I > IR 3 and suppose The figure on the Jeft shows a nonorientable surface whose boundary is the "trefoil" knot. 1 38.a) dy . [ * Z . + .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.g). Let . and . b.
(p .(OJ. B C JRn . q} so that A and B are diffeomorphic to JRn .F where F C JRn is a finite set.. If Ci E H k (M) and fJ E HI (M). If J: M . fJ E HI ( N ) . Choose open sets A . Hk+1 (M x N) is defined by Show that x is welldefined.. then Ci V fJ = (J llfJ v Ci ... We define the cup product v : Hk (M) x HI (M) . show that A B po oq Hk (JRn . � is closed. q E JRn be distinct.. (b) Find the de Rham cohomology vector spaces of JRn . w 1\ � is exact if w is exact and Show that v is bilinear.. (a) Let p .e.{p.. i.Integration 299 30. and that Hn. Hk+l (M) by (a) (b) (c) (d) Show that v is welldefined. 31. Using an argument similar to that in the proof of Theorem 16. J* (Ci V fJ ) = J*Ci V J*fJ· (e) The crossproduct x : H k (M) x HI (N) ..1 (JRn .. q } ) = 0 for 0 < k < n .(p.J . and that . and A n B is diff eomorphic to JRn.. and Ci E H k ( N ) . then [w] v [�] = [w 1\ �]. N. q }) has dimension 2.
: Chapter 8 M > M x M is the "diagonal map". Sl . (a) Show Ihat all de i l /\ . /\ de ik represent diff erent elements of H k (Tn ). On the ndimensional torus Tn = S l de .. (b) Show that every map J: sn let . denote 1Cj*de. . show f]).300 (f) If that . > X n times . by finding submanifolds of T n over which they have diff erent integrals. . given by "(p) Ci V f] = "* (Ci X X = (p. where 1fj : Tn 7 S l is projection on the jth factor.. Hence dim H k (Tn ) � (. Hint: Use Problem 25. • . ) Equality is proved in the Problems for Chapter I I . p). 32. Tn has degree O.
+1 . but there has been one notable w) = (w. ). . = "L. ffi. n previous chapters we have exploited nearly every construction associated and nondegenerate: if v i' 0. (v. v) > 0 A positive definite bilinear function ( quently an inner product. ) is called positive definite if (v. w) 1+ (v.=1 > 0. . . In general.=1 . a) > 0 for any a i' o. (a 1 . . a. ( . An inner product on a vector space V over a field F is a bilinear function from V x V to F. for r = 11 we obtain the "usual inner product". . a ») = " (a ')2 L n r In particular. an ) . ) on (a. w). the field F will always be JR. b) . . ) is clearly nondegenerate. for all v i' o. . . b). v) i' o. a'b'. r :0. . For each r with 0 :0. then there is some w i' 0 such that (w. . denoted by (v. . and conse 30) . 11.n . . I with vector spaces. albl . then n ( a 1 .=1 For this inner product we have (a.CHAPTER 9 RIEMANNIAN METRICS exceptionwe have never mentioned inner products. a symmetric bilinear function ( . a'b'. The time has now come to make use of this neglected tool. = "L. on JRn by (a. and thus with bundles."L. . we can define an inner product ( . For us. . which is symmetric. a. v). i=r+ l n this is nondegenerate because if a i' 0.
so if > V is a linear transformation. ) gives lIS an iSOITIOI"phism a : V + V*. if ( . Thus. n I n this expression.a j > 0 L . w) = a(v)(w). Positive definiteness of ( . gij = (Vi .g. then J* ( . a n with a t least one a i =j:. Vn of V. ) corresponds to the more complicated condition that the matrix {gij } be "positive definite". . ) is positive definite if and only if J is oneone. 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*.9 Notice that an inner product ( . a). . Clearly. with (v. ) i!i linear in the first argument. )= i.j=i L gijv*i 0 v*j . meaning that . } is linear in the second argument. .' dimensional. Nondegeneracy of ( . Thus. . (gij) is symmetric. v*u. ) on V is an element of 7 2 (V). b ) = a I b l . we can write ( . an inner product ( . nondegeneracy of ( . For allY basis V I . wI· Since ( . . Vj ) ' gij = gji. . . so symmetry of ( . bv <pv(w) = (v . W W. if ( . Also. and J: IR > IR2 is J(a) = (a. det(gij) i' 0. . then J* ( . ) is cquivalenl to the condition that (gij ) is nonsingular. Since an inner product { . we can define a linear functional ¢v E V*.. ) is a symmetric bilinear function on This symmetric bilinear function may be degenerate even if J is oneone. .a 2 b2 . if V is finil<. ) is positive definite. . 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. J. gij a. . with corresponding dual basis v*} . J* ( . for each V E V. ) implies that the matrix The matrix {gij } has another important illl crpretation. .302 ChajJle. o. e. .= 1 . ) implies that <pv i' 0 if v i' O. However. for all a l . ) is defined on IR 2 by (a.
l 303 norm In Given any positive difmite inner product ( . (3) II v + w ll :5 II v ll + II w ll (Triangle inequality). (I) is trivial.ja. a j .:5 o. by (2) The function II II has certain unpleasant propeJ1iesfor example. E V we have E IR. II w lL with equality if and only if v and w are linearly dependent (Schwarz inequality). Il t jvf t 1=1 i. AV .w for all A (I) lIav ll = JaJ · II v ll . equality clearly holds. a ) = L )a. W) 2 . THEOREM. w ) II v ll 2 + II w ll 2 + 2 11 v ll . For all v .j=1 .w ) = A 2 11 v ll 2 . . IRn we denote the norm corresponding to ( laJ = . I . ) simply by J(a. so 0 < II Av .2A (v. Thus 2 2 4(v.w ll 2 = (AV . w) + II w ll 2 So the right side is a quadratic equation in A with no real solution.V) (the positive square root is to be taken). then 0 # AV . (2) If v and w are linearly dependent. and its discriminant must be negative. w) J :5 II v ll .4 11 v ll 11 w ll < (3) II v + w ll 2 = ( v + w. the func tion I I on JRII is not differentiable at 0 E JRIIwhich do not arise for the function II 11 2 . i=l The principal properties of II II are the following. •:.Riemannian 1I1etri(.)2 . w PROOF. Il w ll = (II v ll + II w ll) 2 . If not. 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. ) on V we define the associated II II by II v ll = j(V.W . (2) J (v . v + w) = II v ll 2 + II w ll 2 + 2 (v .
} � k VI . II/(v)1I = IIvll for all v E V. that is. w) for all v.IIvll2 . /(b» .IIv . (f(v). . ancJ clearly IIvl li = 1. ) is a positive definite inner product on an Ildimen sional vector spare V. PROOF Let WI . Similarly. . If ated to an inner product ( .) Consequentl. ) . then there is a basis VI . then I is also inner product preserving. I n other words. THEOREM. a.304 More succinctly. . w) = Wv + wll 2 . we can define /* ( .IIw1l 2] (v. PROOF Compute. there is only one positive definite inner product. . b) = (f(a). 3. Vk so that . \Ve will now see that. THEOREM (POLARIZATION IDENTITY). w" be any basis for V. ) on V. Suppose that we have constructed 1 :'0 i. . . . w E V. ' V'j.b E IR". W) = Wv + w ll2 . .w 1l2 ]. Chapter 9 11 11 2 = L gijV*. An invariant definition of a quadratic function can be obtained (Problem I) from the following observation. that is. Theorem 2 shows that tv. If ( . . • • .: there is an isomorphism f : JRII T V such that (a. 2. "up to isomorphism". ) .j=i II II is the norm associ (2) (v. We obtain the desired basis by applying the "GramSchmidt orthonormalization process" to this basis: Since W I # 0. Vi) = DU' (Such a basis is called orthonormal with respect to ( . I(w» = (v.'O inner products which induce the same norm are themselves equal. •:. ) = ( . then (I) " i. if I: V + V is norm preserving. Vn for V such that (Vi .
.I V We can now use fJ to define a bilinear function ( . . This is because we obtain a metric p on V by defining p(V. = I.Riemannian 1\1el. defined by i(V)(A) we obtain an isomorphism fJ : V* + = a. Recall that an inner product ( . . : melric on A positive definite inner product ( . . A(V). The "triangle inequality" (Theorem I (3)) shows that this is indeed a metric. . k . .wli. We also call IIvll the length of v. ) on V provides an isomorphism Ci : V + V* with Ci(V)(W) = (v. So we can define and continue inductively.I 305 and span V I . • . Vk . w). W k . Vk . Then W k +1 is linearly independent of V I . .:::::: span Wj . the symmetry of ( . . w) = IIv . . Now. . We have only one more algebraic trick to play. . ) on V is sometimes called a Euclidean V. ) * on V* by + ( V*)*.. • • .ic. i Using the natural isomorphism i: V + V**. . ) can be expressed by the equation Ci(V)(W) = Ci(W)(V). Let It is easy to see that . . .
) (gU ).} and {v*.1 So aI : V* fJ : + {v*. )* is also symmetric.)=1 n n = .j v*. let {v*. a(w) = /1. 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. . (gU). if we let g U b e the entries of the inverse matrix. ) * = L gij V**i ® V**j i.j= 1 L g ij Vi ® vj .) {v*.1 V· + Thus.. and let ( Then . ( .} and {v·*.} and {v.. choose a basis {v.} for V. that this equation defines ( . Consequently ( . so that then ( . if we consider Vj E V**. without the invariant definition.} be the dual basis for V*. One can check directly (Problem 9).j=l + L g.306 Letting this can be written which shows that Clzapw 9 a(v) = ft.) .. ® v*j . ) * independently of the choice of basis. )* is an inner product on the dual space V* (in fact. To see what this all means. (g U ) = (gU )I . the one which produces fJ). )= .
. . Wn b e two bases o f V which are orthonormal with respect to ( . we are interested in only one case. • It clearly follows that v *1 /\ We have thus distinguished two elements of [I n (V). If we write n Wj = L CijjVj. . ) is positive definite. . . . Vn is orthonormahvith respect to ( . vn ) = ±w(w" . /\ . . )* is also positive definite. v*n of V* is orthonormal with respect to ( . . . . which we will not describe in a completely invariant way. } an orthonormal basis of V.Riemannian 1I1etric. . However. k�' " w(v" . Vn and WI" . .' 307 Notice that if ( . 1'=1 then So the transpose matrix A t of A = (aU) satisfies A · A t = 1. . . the simplest way to describe ( ) )* is as follows: The basis V*l. ) . . which implies that det A = ± l . Similar tricks can b e used (Problem 4) t o produce a n inner product on all the vector spaces T k (V). . we have a(v)(v) > ° for v # 0. The vector space [I n (V) is I dimensional. . In the positive definite case. ). will have length 1 . /\ v*n for {v. . . so that then. . wn ) . /\ W *1/. So to produce an inner product on it. We will call these two elements /\ V*n = ± W * 1 /\ • . for A # 0. we need only describe which two elements. . so ( . . This can also be checked directly from the definition in terms of a basis. TdV) = T k (V*). they are both of the form v * . . )* if and only if v" . letting a(v) = A. and [l k (V). It follows Iiom Theorem 75 that for any W E [I"(V) we have = L Cik i Cikj . . . w and w. Let V I .
Vn and write Wi = j=l L Cijj Vj o n Problem 79 implies that Ifwe write ( . . . To express the elements of norm 1 in terms of an arbitrary basis W I .. . If � is a Coo vector bundle over a Coo manifold we can also speak of Coo Riemannian metrics. S2 : B T E. " . Wj ) .' (p).j ). • . ) which assigns to each p E B a positive definite inner product ( . Ifwe also have an orientation J1. . . we will call it the positive element of norm 1 in n n (v). then In particular. . If � = 1f : E T B i s a vector bundle. det(gU) is a/wll). n then = k� ' L CikiCikj .308 Chapw 9 the elements of norm 1 in n n (v).. /\ · · · /\ w * " gu = (w. we define a Riemannian metric on � to be a function ( . Will we choose an orthonormal basis v} . 0 w*j . vn ) with [v" . . the function is also continuous.s /Jositivf. )= i. . )p on . and which is continuous in the sense that for any two continuous sections S1 . then we can further distinguish the one which is positive when applied to any (VI . " so if A = (Ci. .j=l L gU w*. the elements of norm in n"(V) arc * ± Jdet(gu) w .. . \IVe now apply our new tool t o vector bundles. . vn ] = /1. . Consequently.
w)� UeC9 (v. v)� [The same argument shows that any vector bundle over a paracompact space has a Riemannian metric. THEOREM. We define ( . lu (w» p .] Notice that the argument in the final step would not work if we had merely picked liondegenerate inner products ( .. .I (p). )p is a symmetric bilinear function on . a fibre bundle.b» p = (a. v. The only problem is that Eue(V) is not a vector space.I (p» . PROOF There is an open locally finite cover (9 of M by sets U for which there exists [COO] trivializations On V x IU : . and let Eue(�) = U ElIe(.I (p) by ElIc(. Let ElIe(V) be the set of all positive definite inner products on V. ) u In fact (Problem 7). Then ( . ) is continuous [COO] and each ( . ) on TS 2 which gives a symmetric bilinear function on each S2p which is not positive definite or negative definite but is still nondegenerate. UeC9 each ¢u(p)(v..] 4. ) u i s a [COO] Riemannian metric for � I U. Let {¢u} b e a partition of unity subordinate to (9. 2: 0. + « p . there is no ( . define Then ( . Ifwe replace each . v)p = L ¢u (p) (v. note that (v. ) by (v. For v. a ) ..I (p). p eB then a Riemannian metric on � can be defined to be a section of EliC(�). the new object Elie(�) that we obtain is not a vector bundle at all. w)� = (lu (V) . : E + M be a [COO] k plane bundle over a Coo manifold M. Let � = . W E . W E .I (p).Riemannian 1I1etricJ 309 [Another approach to the definition can be given..1 (p» . To show that it is positive definite. Then there is a [COO] Riemannian metric on �..v) � . and for some U strict inequality holds. b ) . •:. there is an obvious Riemannian metric. but an instance of a more general structure.. I(U) U x IRk IRk ... (p.W)p = L ¢u (p) (v.
a Coo Riemannian metric ( . But � :::: �* . + M is a I plane bundle. ) for TM. A ) .310 Chapter 9 As an application of Theorem 4. In this case. COROLLARY. Clearly s is a section. COROLLARY. ) be a Riemannian metric for �. 6. All these considerations take on special significance when our bundle is the tangent bundle TM of a Coo manifold M. •:. J. then there is a nowhere 0 section of [1 1 m = �*. Continuiry of ( . then � is trivial if /J. If � = 1[ : E + PROOF. W E 1[ . M. Then for each I' E have an isomorphism defined by M is a k plane bundle.p . w) p v. ) implies that the union ofall "p is a homeomorphism from E to E' = UPEM[1[ . [S(p)] = /J. M x IR by ALTERN ATIVE PROOF We know (see the discussion after Theorem 79) that if � is orientable. then � "" �*. If � = 1[ : E and only if � is orientable.1 (p)] * . so that �* is trivial. ) is a with (S(p) . we settle some questions which have till now remained unanswered. we then define an equivalence ! : E + !(A S(p» = (p. •:. S(p» p = I . Let ( .1 (1'). ) p on . We "p (v)(w) = (v. PROOF The "only if" part is trivial. which gives a positive definite inner product ( . If � has an orientation Riemannian metric on M then there is a unique and ( .
Ib·emalUzian A1etric. then /J. which ( . we obtain a "volume element" on in Chapter 8. . even though it is usually not d of anything (even when M is orientable and it can be considered to be an nform). .j) > 0 since ( . n If M has an orientation /J. V) it can be written as .j dx. j= 1 \vherc the Coo functions gij satisfy gjj = gji. if x : V T JRn is orientation preserving. then on V this form can be written . i. . of course. since ( .Jdet(g. If (x. /\ dx n (p). ) * . /\ dx" . We have seen that they can be written ± . ) is symmetric. • • /\ dx" l · . dx 1 /\ . thus p. is a contravariant tensor of order 2.k g kj = 8( . A Riemannian metric ( . the k=l L g.Jdet(gu(p» dx 1 (p) /\ .j ) Similarly.I 31 1 each Mp. a covariant tensor of order 2. ) induces on the dual bundle T*M. as defined Even if M is not oricntable. it is a Riemannian metric on N if and only if f is an immersion (f p is oneone for * all p E N). ) on M is. and det(g. in a coordinate system (x. for each p E M the Riemannian metric ( .j ) Idx 1 /\ This volume element is denoted by dV. So for every Coo map f: N T M there is a covariant tensor f* ( .p allows us to pick out the positive element of norm 1.. then on V we can write our Riemannian metric ( .Jdet(g. The Riemannian metric ( . ) on N. ) is positive definite. M. which is dearly symmetric. i i ) = L 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» . and we obtain an nform on M. ® dxj . U) is a coordinate system on M. ) on M determines two clements of n n (Mp). . is called a R emann an metric on M. the elements of norm I. ) as ( .
312 Clzapw 9 and is called the volume element determined by volume of M as ( . with the "usual Riemannian metric" 1M dV. . 1._. We can then define the This certainly makes sense if M is compact. then gij = 8U. ) 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 . ). b] such that y is smooth on each [/. dV = I dx . meaning that there is a partition a to < ' " < b of [a. and in the noncompact case (see Problem 810) it either converges to a definite number.] (with possibly different = = t I dr I dl Y from a to b. b] + M. which will occupy us for the rest of the chapter. There is an even more important construction associated with a Riemannian metric on M. .. L� (y ) I" If y is merely jJuuwise smootli. For every Coo curve y : [a . dxn l . /\ and ccvolume" becomes ordinary volume. we have tangent vectors y' (t ) = dt dy E My (t). so . in which case we say that M has "infinite volume".. If M is an IIdimensional manifold (withboundary) in IRn .=1 ' /\ . = ) (t lI y'(t) 1I d ) = . or becomes arbitrarily large over compact subsets of M. ® dx. and can therefore use ( . ) = L dx.
i=1 . In .=1 this definition agrees with the definition of length as the least upper bound of the lengths of inscribed polygonal curves. . . s(t) = L�(y) = L�+a(y) = old s(t + a) . the "arclength function of y" by s(t) = L�(y) = Naturally. s(b) = L�(y). with the usual Riemannian metric L dx. + constant.a).II . ® dx. I dl. we can define the length of y by L�(Y) = I >:. (*) s'(t) = Consequently dyjdl has constant length I precisely when S(I) = I thus precisely when s(t) = I .l (y I [IH . [ I �. b .t}. We can also define a function s : [a.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).Riemannian Metnes 313 left.]) . . .a. . Then b a = I �. We can reparameterize y to be a curve on [0. y(1 . 1. A little argument shows (Problem 1 5) that f piecewise smooth curves in or JRn ..aJ by defining y(t) = For the new curve new y we have = 1.and righthand derivatives at 11 . Whenever there is no possibility of misunderstanding we will denote L� simply by L. bJ + JR.
If p. b] . p) = O. i f r E M i s a third point. ® dxi . we can choose piecewise smooth curves y. The classical way of indicating one wrote L g.dxi at p should not be interpreted as a bilinear function at ail. : [a. this is sometimes interpreted as being the equivalent of the modern equation ( .j=l L gu dx. equation (*) says that for each curve y and corresponding s : [a. but what it always actually meant was i i.j=1 and we would use the same symbol today.b] + [b. ) = L7. ® dxi the value (dx. the Riemannian metric was not a function on tangent vectors. It i s clear that d(p. but as the quadratic function i. then for any £ > 0. b] + M from p to q (there is even a smooth curve from p to q).dxi)(p) of dx. q) 2: 0 and d ( p. q) = inf { L ( y) : y a piecewise smooth curve from p to q}. Moreover.) Consequently. in classical books one usually sees the equation ds2 = Nowadays.dxi appearing here is Tlol a classical substitute for dx.i dx. q) < t q to r with L(Y2) d(q. �' gU dx. but the inner product of two "infinitely small displacements" dx and 8x.r) < t. the norm II II on M was denoted by ds. (This makes some Sort of sense even in modern notation. then there is at least one piecewise smooth curve y : [a.8xi n where dx and 8x are independent infinitesimals.) Consider now a Riemannian metric ( .j=1 L gu dx.dxi n II 11 2 = The symbol dx. q E M are any two points. b] + IR we have Ids l = Y* ( II II ) on [a. Define d(p.c] + Y2 : M M from from p to q with L(y. ) on a cOTlnected manifold M. ) .dxi dxi ® dxj was very strange: i. (Classically. .d(p.314 Chapw 9 Classically.
PROOF Both parts of the theorem are obviously consequences of the following 7'.c] . a) < M lb l. ). . i=1 " and let ( . I I on B. The function d : M " M + lR is a metric on M. •:. M > 0 such that m . since Ib I G(p. M > 0 such that on B " S". then Y is a piecewise smooth curve from p to r and L(y) = L(YI ) + L (Y2 ) < d(p. LEMMA. ( .b] + B we have mL. lI ap lip = lI<l b l a)p llp = II b llp.b] and Y2 on [b. 7.(y).c] + M to be YI on [a. ). let ( .a) = lIap lip. and if p : M " M + lR is the original metric on M (which makes M a manifold). let a E S". Let I I = II II.] The function d : M " M + lR has all properties for a metric. ) be any other Riemannian metric. Since B x SIl . a) = Ib l . there would be difficulties in fitting together YI and Y2 . and consequently for any curve y : [a. Let U be an open neighborhood of the closed ball B = {p E lR" : I p l :::: I } .Riemannian Metric. q) + d(q. p). Then there are numbers 111 . m < G <M Now if P E B and 0 # bp E lR"p .(y) :::: L(y) :::: M L. G(p. q ) > 0 for p # q . and II II be the corresponding norms.l is compact there are num bers m. d) is homeomorphic to (M. but d would still turn out to be the same (Problem 1 7). PROOF Define G: B " S"I + lR by Then G is continuous and positive. THEOREM. Then mlbl < Ib IG(p . r) 0. = [Ifwe did not allow piecewise smooth curves. r) + 2e. Since this is true for all £ > d(p.1 315 If we define y: [a. r). this gives the desi�ed inequality (which clearly also holds for b = 0) . be the "Euclidean" or usual Riemannian metric on U. it follows that :::: d(p.I . then (M. This is made clear in the following. L dxi ® dxi. except that it is not so clear that d(p . q) + d(q. I I :::: II II :::: M .I be a = b/lb l .
b] + IR with f(a) = a' and f(b) = b' one a' which will maximize (or minimize) the quantity t F(t.f'(t» a b dl . Of course.. we have to employ techniques fi'om the "calculus of variations". . f(t). x . the two semicircles between the points p and p on S l ).316 Chapter 9 Notice that the distance d(p . I n order to investigate the question of shortest curves more thoroughly. and q might be p. we consider first a simple problem of this sort. if" F(t. if d(p. For example. As an introduction to such techniques. Suppose we are given a (suitably differentiable) function F. y) = �. lR x lR x lR + R We seek. among all functions b' f: [a. the manifold M might be 1R2 {OJ. q) = L(y) for some y. q ) defined by our metric need not be L(y) for any piecewise smooth curve from p to q. then y is clearly a shortest piecewise smooth curve from p to q (there might be morc than one shortest curve. For example. e.g.
we examine the critical points of I. A critical point is not necessarily a maximum or minimum. /(x.y) = Dd(x. i. for a function I : ]R2 + ]R we consider points (x.b] which makes the curve (t. b') of shortest length t )1 + [f'(t)F dl. As a second example. but critical points arc the only candidates for maxima or minima if I is everywhere differ entiable. l(t» between (a. or even a local maximum or minimum.Riemannian 1I1etrics 317 I r> then w e are looking for a function I on [a. those points x for which I'(x) = O. if F(I.y) = o.y) = 21f X JI+Y2 . This is the same as saying that the curves I r> l(x + l. a') and (b. x. 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) t r> l(x. To solve this problem. y) E ]R2 for which (*) D. Similarly.e.y + l) .. then we are trying to minimize the area of the surface obtained by revolving the graph of I around the xaxis.
y). This can be done by considering a "variation" of f.a(u. For a variation Ci we now compute dJ(&(u» d . I). e). but it turns out that these conditions follow from (*). e) x [a.e ) which pass through f for u = O. e) to the set offunctions f : [a. f(t). . I) = f(I).b] + we wish to proceed in an analogous way.b) = b' b' a ' a b for all u E (e. b] + IR. in other words if a(u. Thus & is a function from (e. We will denote this function by &(u). by considering curves in the sel o all f IR. then we call a a variation of I keeping endpoints fixed.b] + IR a(O.(U. because of the chain rule.= du u=o du I aa at I Ib F (l.a) = a' a(u.e) + 1R2 with e(O) = (x. & (u)(b) = b'. f'(I))dl a : (e.e. We might try to get more information by considering 0 = (f 0 e)' (O) for every curve e : (. I)) dl u=o a . such that The functions I r> a(u.318 Chapw 9 have derivative 0 at the condition O. If each &(u) satisfies &(u)(a) = a' . l) are then a family of functions on ( e. To find maxima and minima for funclions f: [a. that is) a function J (f) = t F(I.
(I» a aF aa a a I Ib U (0. I) ay (1. b] + IR with �(a) = �(b) = O. 1(1). essentially the same.] dl. a(U. f(l). f'(t)))] dl aa aF ' I + au (0. 1(1). 1(1). I) [aF (1.1) [.Riemannian 1I1etl1cS 319 = f [ :u lu�o F (I. f(I). (I» ( ay (I. 1 (I» b 11 =0 = a X . The derivative is called the "first variation" of and is denoted classically by a/ ax Iu�o Ib �(I) [ aF (1. = a The final result is. f'(t» )] dl. 1.aF1. l) = 1(I) + u�(t). t)) ] dl Since a 2 ajaual = a 2 ajalau. ofcourse. Then we obtain a (u. �7 (U. flu l u�o J(&(u» J aF d aF 8J ax dl ay Ib � [.dt a ' In classical treatments ofthe calculus ofvariations.d1(1). the second term is 0. the variations Ci were taken to be of the special form . /(1) . 1'(1» d ( aF I.(1.f (� (t.. I). f'(I))) ] dl..dt = a . . thus obtaining (*)  dJ(&(u» du For variations Ci keeping endpoints fixed. we can apply integration by parts to the second term in the integrand. for some � dJ(&(u» � : [a. and we obtain (**) � dJ(&(u» ax au I u�o = Ib aa (0.
. •:' 0= di d ( '/1 /'(1)' 2 ) ' + F(t. The particular form (**) into which we have put now allows us to deduce an important condition. So the theorem is a consequence of the following simple 8'. which vanishes at a and b. consider the case where equation is = ¢(b) = 0. Ci 8J(Ci) 8J(Ci) J. J(t). If a continuous function g : [a. 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. not only are the ar guments I and (t.b] with II(a) = 1)(b) = 0. x PROOF Choos C 1) to be ¢g where ¢ ispositive on (a.(t» . THEOREM (EULER'S EQUATION). We call I a critical point of (or an extremal for if =0 for all variations of I keeping endpoints fixed.f. b] + IR satisfies l b 1)(I)g(t) dl = 0 for every Coo function 1) on [a. J'(t» ) = ax di ay 1)(t) = aCi (0. but the dependence of on is not indicated (which can make things pretty confusing). y) = !i+7. LEMMA. . /'(1» omitted (resulting in the disappearance of the function I for which we are solving). b) and ¢(a) As an example. 1 ) au o. J. 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. the arguments of functions are either put in indiscriminately or left out indiscriminatelyin this case.d ( aF (t. As in the case of I dimensional calculus. J Ci J) 8J(Ci) 8J 8. If I is to maximize or minimize then must be 0 for every variation Ci of I keeping endpoints fixed. 8J Ci J. l(t).320 Chapw 9 As is usual in classical notation. then g = O.(t). The Euler [/ (1)] .
( J7)' = L . 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.( )1l+ [f''(t) F ) . dy dx 1 + ()2 y=O. (I (I this leads to the equation which we will also write in the classical form 0= )1 + [f'(t)F :!.f'2 I/" = 0.x.y) = xJI+Y2.�"��� ) . = dx ' the To solve this.I �2 o= �(..y) = 1 + y2. We let Then .11 +1' 0 = + 1'2)/" I'!" = I' + 1'2)/".11+1'2 ' / .x. where the critical points of are the same as those of since For the case of the surface of revolution. 2J7 F(I. o = �(2/'(1» . we use one of the �o standard tricks Ocaving justification of the details to the reader).Riemannian Metrics so 321 hence which implies that So is linear. dx d2y2 dy P = y. where Euler equation is _ . I" = 0. dl (t) f (t) 1 + . J7 I. I F(I..
) .yp dp = 0. e eX + ex x = 2 cosh is shown below. and cosh . The graph of X +k = c cosh ( . we write this as Y = x +k. � dy p = dx = Jey2 . dy . decreasing for increasing for � o.l ey Replacing c e by 1/ e. Pp2 1+ � log(l + p2) = log y + constant 2 y = constant . x x � 0.322 Chapter 9 so our equation becomes 1 + p 2 . it is symmetric about the yaxis.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.dp = � d)'.
I) aayFl (1.1) [D (I.a') and (b. J(I).(I» I l a.. /. If y : [a .b] + IRn with &(0) = J. t ) x for F: IR x IRn x IRn + IR. a . with y (a ) = p and y (b) = q. dJ(&(u)) dU b � a1 (0. J(I).. b] + M is a piecewise smooth curve. b] + M J of J must satisfy the n equations a. f(l).Riemannian 1\1etnes 323 So our surface must look like the one drawn below. b'). (0. by the way. J'(I)) . (1.. J(I). 1' (1) ) ] dl 1 X o.. 1(1)..(1» ) = . a 1= 1 U + Thus. any critical point We are now going to apply these results to the problem of finding shortest paths in a manifold M. � a. not trivial to decide whether there a7" constants k and c which will make the graph of (*) pass through (a. we define a variation of y to be a function aF d (a axl (/.t) x [a . : ( t .f G. aF u=o [a.di ayFl (I. and compute = L.b] + IRn J(f) = In this case we consider that t F(I.. f. : (t. Problem 2 1 investigates the special case where a' = b' . 1'(/ » Ib n . It is easy to generalize these considerations to the case where and J : [a.. 1(/ ). /'(1» dl . It is..
We would like to find which for all variations Ci keeping endpoints fixed. (. l) . As before. I) = " U (u.r) =" (u. (u. However. . . 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.1...r) . I). such that a = 10 < II < .I ) aa aa"I (u.t) x [a.t) x [1.a) = p .(u.].. I) is the standard coordinate system in (t. we let paths y satisfy & (u) be the path 1 dL(&(u» du r> .(u. .324 Chapter 9 > for some t 0. (0. dt a dt ' dt aa" .b] so that " is Coo (t. b] we write lb I dy 1 2 dl = Zl ib ( dY dY ) dl..t). We can assume that each y l [1.1) = y (t)._ 1. < IN = b of [a. . b) = q for all (I) . If (u. _ 1 . (2) there is a partition on each strip We call " a variation of y keeping endpoints fixed if (3) u E (t.] lies in some coordinate system (x. (Iaa I ) U ( U. U) (othelwise wejust refine the partition). aftenvards we will consider the relation between the two integrals. 1.
ndI dI dyi dyj 1 L gij (y(l» . (dY dY ) dl ' 1. I) is the tangent vector at time I to the curve &(u).c. t) = xi (et(U..y) = '2 L gij (x) .t) ti_ l E(y! [li l . F(y(l). i ih. ti _ l i.j = 1 1 tj _ l " Then and So . li J) = 2 1 = 2 11. then we are considering IRn .Riemannian 1\1elr.I) So = L ii/(U.\ 325 Then aet/al(U. I» . and consider the gij [. a(u.dt.. Ifwe adopt the abbreviations eti (u. then aa iii (U.=1 1 l1 [}e/ 0 1 . i yi i.l) . y'(t)) dt where F(x.j =l dl dl If we usc the coordinate system x to identify U with as functions on JRn .
agl. /] a = � gi/ 2 ( ax' + agjl ax' _ agij .. ax' ) [ ... 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= l ax' dt dt So �. dyj dy' = �. we note that a little index juggling gives r. i d i dy .L.= left hand tangent vector of y at I (l i· dl i ) �(tn dl To abbreviate the integral somewhat we introduce the symbols [ij. L. i.326 Chapw 9 In order to obtain a symmetrical looking result. agi/ dyi dyj = � agjl dyi dyj L. j= 1 ax' dl dl i.1. Let dyI(l + ) = nght hand tangent vector 0f y at I ...
i = 1. glr (Y(t) dd2.fi)._.. but the integral  327 iI. but there it is.I](y(t) d� d�j n I [ n r n i ] dl. simple to obtain conditions for critical points of E. Il" d/) a. Now we just have to add up these results. For the time being we present: with apologies.' \�I [ij. I] . the sum can be written [ij. .) This result is not very pretty. . THEOREM (FIRST VA RIATION FORMULA). Ci . this coordinate dependent approach. leaving endpoints fixed. later on we will have an invariant interpretation of the first variation formula. (In the case of a variation from 1 to N .1 . Let � aa: (O. 11=0 N dy . . For any variation w e have dE(Et(u» du I Ci. which appears in our result. we will use the exact same expression for each [IiI . From the first variation formula it is. clearly cannot. N . Nevertheless. 9. even though different coordinate systems may actually be involved (and hence different gij and V i ). of course.Riemannian 1I1etrics These depend on the coordinate system. Ii]. . t) !. Consequently.l Then we obtain the following formula (where there is a convention being used in the integral). It should be noted that are not the components of a tensor. 1.� (aCi (O.
. .. L. + M is a Coo path. U) we have y: [a b] 11 y i Lg/. If .L [ij.' + i. 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). (y(l)) d2. U).1 . I) to be 0. except one. that for each P M and each v Mp.328 Chapw 9 1 0 . . The final term in the formula vanishes since y is Coo.[i2 + . If Ci is a variation of y keeping endpoints fixed.I to Ii is written in terms of (x. Now apply the method of proof in Lemma 8'. d j =l r=1 PROOF. which is 0 outside of (1'1 . COROLLARY. there is a unique (t. r/k ( y (I » dt d l. then is a critical point of E� if and only if for every coordinate system (x. I!) : n r I. choosing all aCi' jau(O.b] with I E (1.). 1. Suppose y is a critical point. then in the first variation formula we can assume that the part of the integral from Ii. for some t > 0.t) + M..j=1 =P o. but a positive function times the term in brackets on (IiI. such that satisfies y: E y E y(O) dy (0) = v (ji d2yk � j dy' dyjt = . Given / with y(l) E U.] is in U. •. choose a partition of [a. ) for some i. and such that y l [I'_I.I. I](y(l)) d� d�j = 0 for y (t) E U.
_. (0) = Y2 (0) then y. As the simplest possible case. So we obtain which implies that all !:1/.. this y is Coo on (e.e) + M and Y2 ' (e.e). PROOF. ..b]. U) satisfies Y ' [a. . b] and for every coordinate system for y(I) E U. Naturally. Now choose Ci so that c/ c/ a. COROLLARY. (0.j. and rt = O. so all ag'j /ax k = energy function satisfy 0.:. + d Y2 dl (0 ) = dl (0 ) y. ' [O. . N . we see that yl [I' _h l. consider the Euclidean metric on * y JRn ) ( .. We already know that the integral in the first variation formula vanishes. This last fact shows that if y.Riemannian 1I1et. A piecewise Coo path for (x. ) = L dx ' ® dx'i=l Here g'j = D. we could replace 0 by any other I.e). The critical points y for the . and Y2 together give a Coo function on (e. because the final term in the first variation formula still vanishes.cs 329 Moreover. and ifmoreover d y. Choosing the same as before (all are 0 outside of (1.» . . We now have the more precise result. are O. . b] + M is a critical point Eg if and only if Y is actually Coo on [a. Let y be a critical point. . I.1.) = !J. By our previous remarks./' dt ' aCt dy i = 1 . O] + M are Coo functions satisf ying this equation.] satisfies the equation. this means that is actually Coo on all of [a. ' I I . 1.
PROOF Observe first. y y y [a. }] + [}/. TllllS the length is constant. If eterized proportionally to arclength. from the definitions.330 Chapter 9 Thus lies along a straight line. (y(t » dt df2 ' =l i. since length is independent ofparameterization (Problem 16). that y: E. Replacing agij lax' by the value given above.j=l dyi d2y' + L gi. y y y (t. This situation always prevails. when we considered only curves of the form I r> Any reparameterization of is also a critical point for length. f(t» . d dy 2 d " gij (y(t» dit dyjt) I 1 = di (i. and hence must be parameterized proportionally to arc length. +!+ y I dyldlll . i].L ax' l. The situation is now quite different from the first variational problem we considered.�. this can be written as di dt r n Since is a critical point for E.b] + M is a critical point for 12. THEOREM. both terms in parentheses are 0 (CorollalY 10). the component functions of must be linear. since we have just seen that for to be a critical point for energy. so is a critical point for the length function as well. dy d i dy' dy dyj " d2y dyj " " L agij (y(t» dt dtdt + L g'j (y(t) df'2dt = . then y is param Now w e have ��:{ = [i/ . This shows that there are critical points for length which definitely aren't critical points for energy.j=l /=l r.
It will also be useful to know a f ormula for ag ij / ax k To derive one.b] y [li1. l]) by (*) Lg I.j=l . For the por tion I of contained in a coordinate system U).y) = \ iL gij (x)yiyj .. g gmj oglkm aY I. Ii] y [a dy/dl (x... For the moment we consider only paths + M with # 0 everywhere.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..m aY il gmj ([lk. we first differentiate = m=l to obtain Thus we have ogikj L.Riemannian 111etries The formula 331 ·· ag't [ik.i] ax occurring in this proof will be used on several occasions later on. Considering our coordinate system as JRn) we are now dealing with the case F(x.m] + [mk. g glm ogmj L. .m aYk I.. we have y: .j] + [jk..
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). dl . L� (y)] + M.j �1 ds ' y ( ) . and we can consider the reparameterized curve e This reparameterized curve is automatically also a critical point for L . given a critical point y for L with y I # ° everywhere. L� (y)] 1 is a diff omorphism.332 Clzapw 9 s (t) = L� (y) . There is only one detail which remains unsettled. Conceivably a critical point for L might have a kink. so it must satisf the same diff rential equation. Conversely. so y o S 1 is a critical point for E. b] + [0. the third term vanishes. Since it is now parameterized by y e arclength. We introduce the arclength function Then So we have " a · d id j L 4· (y (t» � . the function d /d d2s/dl2 s: [a. . but be Coo because it has a zero tangent vector y oS : [0.
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. In this case it would not be possible to reparame terize y by arclength. which are so far merely known to be critical points f length. Henceforth we will call a critical point of E a geodesic on M (for the Rie mannian metric ( . A geodesic on the earth 's surface is a segment of a great circle. including surveying and the measurement of degrees of latitude and longitude. Observe that the equations f a geodesic. as in the figure below. � PROOF. Problem 37 shows that this situation cannot arise. we must initiate or a local study of geodesics. then I r> y(cI) is also clearly a geodesic. which is concerned with the measurement of the earth's surface. Let p E M. 2) + M satisf ying Yv(O) = q. <2 > 0 so that f q E U and v E Mq with IIvll < <1 there is a unique geodesic . which is the shortest path between two points. THEOREM. This feature of the equation allows us to improve the result given by the basic existence and uniqueness theorems. or have an important homogeneity property: if y is a geodesic.Riemannian 111el1ics 333 there. Before we can say whether this is true for geodesics in general. »). The most elementary properties of geodesics depend only on facts about differential equations. d v (0) = v. The fundamental existence and uniqueness theorem says that there or is a neighborhood U of p and <] . 13. This name comes f rom the science of geodesy.
Choose < < < 1 <2 : Then if Ivl < < and III < 2 we have So we can define yv(l) to be yv/. : Mq + Mq is the identity.) The geodesic y can thus be described as y(l) = expq (lv). to obtain a curve c in the manifold Mq with dc/dl(O) = v E Mq = (Mq)o. In fact. y(l). we can let crt) = IV...I dl 1=0 eXpq (c(I» = v. exp(v) = expq(v) = (The reason for this terminology will b e explained in the next chapter. 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. Since Mq is an ndimensional vector space. (v)· In particular.(v) = :!. there is a natural way to give it a Coo structure..334 Chapter 9 with the required initial conditions. . •:. Then expq 0 c(l) = expq (lv).. we have an induced map (expq). d"i (0) = v. we claim that the map (expq)o. Identif ying the tangent space (Mq).« 21) . then the map expq : (9 + M is Coo) since the solutions of the diff rential equations for geodesics have a Coo e flow. y : [0. If (9 c Mq is the set of all vectors v E Mq for which expq (v) is defined. : Mq + Mexp. at v E Mq with Mq itself. the geodesic with tangent vector v at time 0) So (expq)o.
Riemannian 1I1el. 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. . . (3) For each q E W. 1 4. For tangent vectors l (X 0 1f) O . . x l . . . q') r> v(q. U) is a coordinate system on M. xn .l (U). q') is a Coo function from W x W + . . x l .. . If (x. .=1 where 1[ : TM + M is the projection. . xn o n. 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. (v) axi n We will denote ai by Xi (v). so that v= . then for q E U we can express every vector v E Mq uniquely as " . Then is a coordinate system on 1[ . v(q. . . . Then (q. . THEOREM. (2) Let expq(v) TM. C TM. we recall some facts about the manifold TM. xn ) = (xl. a v = L a' ox. the map expq maps the open eball in Mq diffeomor phically onto an open set Uq :::I W. q') denote the unique vector v E Mq of length < e such that = q'. . . q .=1 a L X'.. xn ) v E Mq.cJ 335 Before proving the next result.
using the fact that (expp)o. The close analysis of or geodesics depends on the f ollowing. Choose W to be a smaller neighborhood of p for which F(V') :J W x W.(v).1 (V). We may assume that V' consists of all vectors v E Mq with q in some neighborhood V' of p and IIvll < e. consider the geodesics through q of the form I r> eXpq (lv) f IIvll < e. !2 . We will use the coordinate system described above. = 0 E Mp we have (P' p) + Mp (p . then + M is projection on the i th is a coordinate system on V x U. p) E M x M. for . If ".336 C/zapw 9 PROOF. it is not hard to see that at Given a W as in the theorem. exp(v» .p) F 0E F. These fill out Vq . Consequently. Define the Coo function F: V + M x M by F(v) = (... ( a�i IJ a 1 . and q E W.. Theorem 13 says that the vector 0 E Mp has a neighborhood V in the manifold TM such that exp is defined on V. so maps some neighborhood V' of 0 diffeomorphically onto some neighborhood of (p. a . p) = . V) be a coordinate system around p. •:. I + a!i I F. : Mp is the identity. : M x M factor. . is oneone at Mp . F. Now. Let (x. ( � IJ a!.
I) is just the tangent vector at time u to the geodesiC u r> expq ( u .Riemannian Metrics 337 1 5. we obtain 0 since each curve u r> Ct( u l) is a geodesic. In Uq . I). the geodesics through q are perpen dicular to the hypersurfaces { expq(v) . Let = constant < e}. au al . IIvll FIRST PROOF. . aCt aCt � (U.I) = ( � ' � ) is independent of u. But aCt/ au(u. vU» I <u< 1. where IIvU)1I = k. I) = a constant We are claiming that for every such Ct we have ( = expq (u . So But Ct(O.I) ' at (U. for convenience: The first term on the right is Similarly. I) )=0 f all (u . I). I) = o. au al expq(O) = q. or A calculation precisely like that in the proofofTheorem 1 2 proves the following equation. and define Ct(U. v: lR + Mq be a smooth curve with IIvU) II k < e for all I. which is just twice the second term on the right of (1). so aCt/al(O. vU» . so lIaCt/aull = k . LEMMA (GAUSS ' LEMMA). I) are omitted. It follows that (� ' � ) = 0 for all (u. Thus the second term on the right of(2) is also O. in which the arguments (u) I) and Ci(U.
. COROLLARY. the first variation formula. ddly (0») au au afJ dy . •:. we have = dE(du(u» 1u=o � E(  = the integral vanishing since y is a geodesic. . and define expq(t . . By u). c(t) = expq(u(t) . I). I].{q} be a piecewise smooth curve. v(O» .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. v(t» . Let c : [a.l) = Then fJ is a variation of the geodesic y(t) = expq(t . = (t) r dl =  (� O ( 1\ 2 dl = k2 . But each curve ('!t (0. � u» 10' 1 0 = dE�(U» lu� o ( = dg �. I ).( a. v(u» fJ(u.(1») .. � (u) has energy . (0. _ so 1 6. Tt(I) ) .0). defined on [0.bj + Uq . I). ddly ( I ») ('!t (0.
I) and u au (I) aCt . and let e : [0. COROLLARY. So the length of e is 2: r. We can assume that q' = expq(rv) E Uq . and clearly e must be a reparameterization of y f equality or to hold.u(a)l.' iii = ) = e(l) = Ct(u(l). By Corollary 16. v'(I) = 0. I) = expq(u . Let W and e be as in Theorem 15. with equality holding if and only if c is a reparameterization of PROOF y. If Ct (u. 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 . u is monotonic and v is constant.8. •:. then di de Since ( aCt aCt = 0.{q} (otherwise break e up into smaller pieces). Then I L(y) ::: L(e). a. I ] + M be the geodesic of length < e joining q. For 8 > 0. 1 7 . •:. Thus 0. Then 339 L�e 2: lu(b) .for ° < u(l) < e and II v (l ) 1 I = Riemannian 111ellics 1. and lies between them. v(l» . q' E W.u(a)l. the path c must contain a segment which joins the spherical shell of radius 8 to the spherical shell of radius r. and hence 2: lu(b) . ] + M be any piecewise Coo path from q to q'. so that c is a radial geodesic joining two concentric spherical shells around q . PROOF with equality if and only if u is monotonic and v is constant. let y : [0. the length of this segment has length 2: r . .. 2 2: lu'(I)12 .
even among nearb paths.340 Chapw 9 We thus see that su fficient6' small pieces of geodesics are minimal paths for arc length. For example. depending on how many time� the geodesic goes around the sphere and in which direction it starts. the sphere have a continuum of geodesics of minimal length between them. the fixed point set is the great circle C = s n £2 Let p. n Notice that a portion of a great circle which is larger than a semicircle is defll1itely not of minimal length. )) and (M'. then the length of c with respect to ( . So C' = I(C'). ). We can use Corollary 17 to determine the geodesics on a few simple surfaces. which implies that C' c C. )'. If (M. Antipodal points on ). 1] + M is a Coo curve. ) is the length of f o e with respect to ( . so that C is a geodesic. ) ' ) are Coo manifolds with Riemannian metrics. then f o e is likewise a geodesic. Then I(C') is a geodesic of the same length as C' between I(p) = p and I(q) = q. It is clear that if c : [0. if we first introduce a notion which will play a crucial role later. then a oneone Coo function f : M T M' is called an isometry of M into M' if f*( . ( . ( . . q E C be two points with a unique geodesic C' of minimal length between them. and if c is a geodesic. For the isometry I : sn + sn mentioned above. without any computations. All other pairs of points have a unique geodesic of minimal length between them but an infinite family of nonminimal geodesics. reflection through a plane £2 C )Rn+l is an isometry I : sn + sn. Since there is a great circle through any point of sn in any given direction) these are all the geodesics. ) ' = ( .
) is called geodesically complete if evelY geodesic y : [a... 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 ( . b] + M can be extended to a geodesic from JR to M. In fact) if L is a generating line of Z) then we can set up an isometry I : Z . This is false on a cylinder of bounded height. . the . or JR" . The geodesics on Z are just the images under I. ) and the metric d : M x M + JR it determines. Notice that on both the sphere and the infinite cylinder every geodesic y defined on an interval [a. and the helices on Z . . . In general.. 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.I of the straight lines in JR2 Two points on Z have infinitely many geodesics between them.. a bounded portion of JR".{OJ. q) = inf{L(y) : y a piecewise smooth curve from p to q}. circles cut by planes perpendicular to the generating lines. a manifold M with a Riemannian metric ( . b] can be extended to a geodesic defined on all of lR.
q) = r . choose Up as in Theorem 14.(d(y(to). we will prove that S such that d(po. We claim that expp (r v ) = q. Then d(y(to). drs. If ( . Given p. then M is geodesically complete if and only if M is complete in the metric d determined by ( . First of all. since every curve from . q E M with d(p. of radius 8' around y(to) and let po' E S' be a point closest to q. q) for all s E S. This proves that (**) holds for I = 8. q). Then or (**) holds for 10 also. Moreover. q) :. any two points in a geodesi cally complete manif can be joined by a geodesic of minimal length. Now let 10 E [8. THEOREM (HOPFRINOWDE RHAM). we clearly have d(p. Let S' be a spherical shell p to q must intersect S. Suppose M is geodesically complete. old r > 0. ) is a Riemannian metric on M. To prove this result. r] be the least upper bound of all I f which (**) holds. d(y(t). s) + d(s . by continuity. 8 < e. q) = ��.I IE [8. Suppose 10 < r.q) = �ir (d(p s) + d(s.r].q). ).342 Chapw 9 18. q) = Let S C Up be the spherical shell of radius po = expp 8v. There is a point on PROOF.. So d(po. 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. q») = 8' + d(po' . .8. q») = 8 + d( po. q) = r .
III other words. so it must be that 10 = r. (**) holds for I = r. q) . po') 2: d( p. fact. it is not difficult to show that y can be extended past b. showing that (**) holds for 10 + 8'. note that there is always a min imal geodesic joining any two points of a compact manifold. As a particular consequence of Theorem 18. and p E A . so it converges to Some point p E M. by a least upper bound argument. which proves (*). b) + M. suppose M is complete as a metric space. Given any geodesic y : (a. and must therefore be a geodesic. I n other words.10) . Using Theorem 14. if A C M has diameter D. Hence y(lo + 8') = po'. . In Hence (***) gives . Conversely. Clearly y(ln ) is a Cauchy sequence in M. then the map expp : Mp + M maps the closed disc of radius D in Mp onto a compact set containing A.d(po'. it follows easily that M is complete with the metric d.(10 + 8'). From this result. 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'. •:. So e is a path of minimal length. d(y(lo + 8'). q) = (r . q) = 10 + 8'. From this it is clear that Cauchy sequences converge. choose In + b. q) = r . Consequently. bounded subsets of M have compact closure.8'. This contradicts the choice of 10.Riemannian Metrics so 343 d( po'. Hence d(p. which means that it coincides with y. any geodesic can be extended to ffi.
so it too is nontrivial. On the other hand if M is the Mobius strip and S I C M is a circle around the center. or p eM Let E= U Mp L and w . E + M take MpL to p. the normal bundle v of sn 1 C IRn is the trivial I plane bundle. so that for every p E M we have i. then we can define M/. then the normal bundle 1 of S I in J1> 2 is exactly the Same as the normal bundle of S I in M.C Np as MpL = {v E Np : (v. ) is a Riemannian metric for N.(Mp) C Np ' If ( . the normal bundle of M in N. for v has a section consistiIlg of unit outward normal vectors.w) = 0 f all w E Mp} . 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 . with i : M + N the inclusion map. i. It is not hard to see that v = w : E + M is a kplane bundle over M.344 Chapw 9 ADDENDUM TUBULAR NEIGHBORHOODS Let M" C Nn+ k be a submanifold of N. . For example.
if we choose a Riemannian metric ( . then clearly 1f 0 S = identity of M. and for which the Osection s : M + U is just the inclusion of M into U. 1 9.Riemannian Metric. Let I : X + Y be a local homeomorphism such that IIXo is oneone. Before proving the existence of tubular neighborhoods. is smoothly homotopic to the identity of U. But in general the neighborhood cannot be all of N. In the case where N is the total space of a bundle over M. and the map 1f I D : D + M is also a def ormation retraction. thus M has the same de Rham cohomology as an open neighborhood. so 1f is a deformation retraction.{OJ. So M also has the same de Rham cohomology as a closed neighborhood. LEMMA. Let X be a compact metric space and Xo C X a closed subset.e) ::: I } . or then D is a submanif oldwithboundary o f U. 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. and H k (U) "" Hk (M). Then there is a neighborhood U of Xo such that IIU is oneone. Moreover. we add some remarks and a Lemma. ) f 1f : U + M and define D = {e E U : (e. If 1f : U + M is a tubular neighborhood. for which the Osection s : M + U is the inclusion of M in U. For example. as an appropriate neighborhood of S l C )R2 we can choose )R2 . this open neighborhood can be taken to be the whole total space. s 0 1f . A bundle 1f : U + M with U an open neighborhood of M in N. is called a tubular neighborhood of M in N.
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). onto U. PROOF. where we will also need the following modification. then g 0 on C. . dyjd ( dyjdl(O) q.. is compact. This will clearly prove the theorem. f if (Xn l Yn) is a sequence in C with Xn T x and or T y. norm I I . Then f is oneone on the {(x. •:. Xo). for some p E M } E . Chapter 9 Let C C X x X be x <neighborhood of Xo. x # y and f(x) = f(y») . and compactness of M. n q q y (0) y ( 1 ) q. Let E = {v : v E Np and v E M/". Then C is closed. Theorem 20 will be needed only in Chapter I I . with the corresponding It f ollows easily from Theorem 13. If g: C JR is g(x. = {v E E : I v l < e ) U. y) = d(x. and metric d: N N JR. while they do not even appear in the statement. e It is clear also that exp(E. This means that = expp / O) where E E.. which is equivalent to the normal bundle of M in N. •:. then f(x ) = lim f(xn) lim f( Yn ) = fry). ) for N.y) E X > = X: Yn + > old 20.) C U. 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. . 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). THEOREM. Since C is compact. that exp is defined on E. Let M e N be a compact submanif of N.346 PROOF. the map or or exp is a diff omorphism f e rom E. 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. M) < e). . f sufficiently small < > O. We claim that f sufficiently small e. = {q E N : d (q. = V E. Then M has a tubular neighborhood 1f : U + M in N. y: [0. and V. there is e 0 such that g 2: 2< on C. Xo) + dry. and also x # y since f is locally oneone. x + Choose a Riemannian metric ( .
.Riemannian 111etrics 347 2 1 . using only inward point ing normal vectors . Then aN has (arbitrarily small) open [and closed] neighborhoods f or which there are deformation retractions onto aN. Let N be a manif oldwithboundary. •:. with compact bound ary aN. THEOREM. PROOF Exactly the same as the proof ofTheorem 20.
A basis VI . Wi.q( w ). (d) Conversely. ) an inner product on V which is not necessarily positive definite. E. /\ ¢k = 1/11 /\ . . so that we have a wedge product VI /\ . 1/Ii E V* satisfy ¢1 /\ . (a) Show that W E W if and only if W /\ ¢1 /\ . show that ¢1 /\ . v) (c) Suppose q : V + F satisfies q( . then f some aij . and let h : V x F be symmetric and bilinear. If ¢i = Lj aj. Let V be an ndimensional vector space. then we have a geometl"ic condition for equality with WI /\ . . . or (b) Show that F by q(v) = h (v. (The sign is + if ¢1 . Ifwe identify V with V**.q ( u ) . Cartan uses this condition to define n k (V*) as formal sums of equivalence classes of k vectors. . · v*j n q ( . . . ' . and ( . + Let V be a vector space over a field F of characteristic # 2. 1/Ik.q (u) 0. Let ( . ¢ (b) Let U1 . . if W = W"' .v) = v. . + (a) Define q : V for V*. Show that if ¢1 .348 Chapw 9 PROBLEMS V I. Then show that q(v) = h ( v. . Show that q (u+ v + w ) .) (c) Using Problem 79. . . /\ 1/Ik . he deduces geometrically the COlTcsponding conditions on the coordinates of Vi. . and q(2u) 4q(u). /\ Wk In Lcfons sur La Ceometrie des Espaces de Rieman1l.q (v + w) Conclude that '1(0) =H = q( u + v) .q( u ) . " " ¢k has the same orientation as al l ' . .othe:rvvise. 2. 3.q(u) . ab and . Vj ) = ±8ij. .q(u ) .v) = q(v) h ( u .uj . = q(u+v) . ) be a Euclidean metric for V*. show that this volume is the same for 1/11 . . . v) q(v) is bilinear. and the signed volumes of the parallelepipeds are ¢ the same. . = . . j =l L Qij v*.q(v) . v). " " ¢n is a basis q= i. . ' " ¢k is det(aij ). . and let W and W".q( u + w ) . • /\ Vk ofvectors Vi E V. . = = q(u + v) . /\ ¢k ¢ O. Uk be an orthonormal basis of W¢ = W". Conclude that W = W". /\ 1/Ik # 0. show that the signed kdimensional volume of the parallelepiped spanned by ¢1 .q(v) ] . . • . . • /\ ¢k = 1/11 /\ . v). Vn for V is called orthonormal if ( Vi. be the subspaces of V* ¢ spanned by the ¢i and 1/Ii. Suppose ¢i. and that iJ(u. . .
x . V n be an orthonormal basis (see Problem 3). . then = . ) = ( . w) = 0 for all W E W } . Vj ) . . then there is a vector v E V with (v. in Problem 726. /\ .  . < ik � n be an orthonormal basis. ) k is independent of the basis v" (b) Show that (¢' /\ .Jdet(gij ). . . then V = W Ell WL . 1/Ji}'). . . v) # O. Let ( . Show that these inner products agree with the ones defined above. V i ) = O. (b) For W C V. Vk. for some r (the inner product ( . x . (c) I f ( . x V. and let . where gij = (Vi . (a) Show that (v. . . ) i s also nondegenerate on WL .. . S how that the index is n . ) is the largest dimension of a subspace W C V such that ( . (d) V has an orthonormal basis. ). (e) The index of ( . /\ ¢k o 1/1. consider the linear functionals Ai : V + IR defined by Ai(v) = (v. Thus. define inner products on 0k V and A k V by using the isomorphism V + V* given by the inner product on V. ) I W is negative definite. I = . . 5. Define an inner product . . X Vn _ ' . and ( . Wi ) . ). • . .) det( ¢i. is defined on page 301). . ) be a (possibly nonpositive definite) inner product on V._. there is an isomorphism f: IRn + V with f*( . . thus showing that r is unique ("Sylvester's Law of Inertia").Riemannian 1\1etries 349 (a) If V # {OJ.} is a basis for W. . (Use Problem 716 . X Vn _. (b) Show that lv. ) k on n k (V) by requiring that 1 � i1 < . 1/Ij )* } = det( ¢i. ) has index i . Prove that dim WL 2: n . Hint: If {w. with (a) Show that ( . ) i s nondegenerate o n W. .dim W. Recall the definition of v. let WL = {v E V : (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)*. /\ 1/Id (c) I f ( . Hint: Apply the result on page 308 to a certain (II I )dimensional subspace of IRn . v" 4.r. x .
)p on each 1[ . '. Let S be the set of e E E with (e.j x '  8. a . g C). is the classical way of defining the tensor [having the compo nents] 10. + = ' X /J then g This. for each p E S 2 (Consider the space consisting of the two unit vectors in each subspace. Define a tensor B of type @ by gil a '· a . Let � '= 1[ : E + B be a vector bundle. )p is constant on each component of B .. (a) Let ( . Let ( . Show that the index of ( ..� 'r:t� _ . i}�� ' i. So If the expression for A in a coordinate system is . An indefinite metric on � is a continuous choice of a nonpositive definite inner product ( .1 (p). and the functions gij . and define S' similarly. . Show that S is homeomorphic to S'. This problem requires a little knowledge of simpleconnectedness and cov ering spaces..e) 1. (a) There is no way of continuously choosing a I dimensional subspace of S 2p. of course. Show by a computation that if the functions . If � is a smooth bundle over a manifold M.j axi axj  L. ) and ( . ) be two Riemannian metrics on a vector bundle � 1[ : E B. ) be a Riemannian metric on M."". g ij are defined by 9.350 Chapter 9 6.) (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 '." with det(gij) # 0. 7. show that S is diffeomorphic to S'. and A a tensor of type that A (p) : Mp + Mp .
(b) For the case of a nonpositive definite metric.cs show that = 351 B LI. . Xn be linearly independent vector fields on a manifold M with a Riemannian metric ( .gkI A{ . Y ) j ±81j in a neighborhood of any point. then v(p) is outward pointing in the sense of Chapter 8.k Blk dxl ® dxk . Yn . A2) = (A(p) * (AIl. Let M C IRn be an (n . Ai gjk · j=1 (b) Similarly. (VI)p. (a) Let XI. 1 1 . . .A2)' Show that if C has components C kJ . (V. show that the area of the surface obtained by revolving the graph of 1 around the xaxis is t 21f1JI + (/')2 . (vnIlp is positively oriented in Mp. (b) Let d VnI be the volume element of M determined by the Riemannian metric it acquires a submanifold of IRn . i=1 The tensors B and C are said to be obtained from A by "raising and lowering indices" .. . (b) Compute the area of S2 13. define a tensor C of type (�) by C (p)(AI .. . . . (Vn I)p is positively oriented in IRnp when (VI)p. then n ckj 'L.Riemannian 111etr. Show that the GramSchmidt process can be applied to the vector fields all at once. then V (p) dVn_l(p)( VI)p. . . . _I)p) = det �I . . as () . ). . Vnl = = E /1.bj + IR is positive. where n Blk = 'L.. . . (a) If I: [a. . 12. so that we obtain n everywhere orthonor mal vector fields Yh . . (a) If M = aN for an ndimensional manifoldwithboundary N C IRn . . Yn with (YI. S how that if we consider v(p) as an element of IRn .. find Yh . .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).
X (w.) /\ . with v the outward unit normal on aM. .l )iI ai dx' /\ . (a) Let v" be the volume of the unit ball in (V [ Cv x X. Prove the Divergence Theorem: [ div X dVn = [ (X. /\ dxn . x vnd. W n w 14. Conclude that vi (p) ..•. /\(fXi(J. Show that for IRn we have n v(p)) . Show that Vn = /'I (1 . /\dxn (p). /\ (fxi(J. x . Let X = Li a i a/ax i be a vector field on M. Let X be a vector field on M.. . T) ds JaM JM x X is defined in Problem 727).x2)(nI )/2 Vn_ 1 dx. Let T be the vector field on aM consisting of positively oriented unit vectors. . Denote the volume element of M by dVn . C Slfikes' Theorem: J1. v ) dA = [ (X. . (d) Let M c IRn be a compact ndimensional manifoldwithboundary. .352 Conclude that Chapter 9 dVn_1 (p) is the restriction to Mp of t(l)iI vi (p) dx' (p) /\ . Hint: Consider the form on M defined by = L(. Prove (the original) W E (w .  . and that of aM by dVn _ l . ' . . X Ci Ci E X .. (VI VnI> v(p)) = V. . . dVn_1 (p) = restriction to Mp of (l)i'dx' (p) /\ . V _1 = v(p) for some IR (by Problem 5).i /\ . v) dVn1 JM JaM (the function div X is defined in Problem 727). /\ dxn (p).. /\ d.) /\ . .. Denote the volume element of M by dA. and that of aM by ds. IRn . with odentation and outward unit normal v. (c) Note that VI x . i=1 (e) Let M 1R3 be a compact 2dimensional manifoldwithboundary. .
is the least upper bound of the lengths of inscribed polygonal curves. I] l' i=1 L [(ci )'(I)j2 dl. Using the method of proof in Corollary 88. V2 = IT. where Riemannian metric ( . show that Vn = (In terms of the r function.Riemannian Metlies 353 (b) If In = ['. n JRn has the usual L(c) = (b) For the special case length.) Vn = (e) Obtain this same result by applying the Divergence 111eorem (Problem 13).x2)(nI)/2 dx = . show that I (:��:! o 2 (n +' )/2. In = .(nI)/2 I · 3·5" 11 n odd. l' jl + JR2 given by C(I) = (l. ) = Li dx i ® dx i .I )volume of S nI . but reversing the order of integration.In2 .I be the (11 . I] + JRn be a differentiable curve. c(lll) for .. (1 . with X(p) = pp . 15. Hint: If the inscribed polygonal curve is determined by the points (Ii . n n even ' 2. Show that I l ' rnI An1 dr = AnI . f(I» . c: [0. show that this + [/'(1)]2 dl . this can be written (d) Let A n. 1fnf2 r(l + n/2) . show that nI (c) Using V. (a) Let c: [0.
u A1fClJlUJ 'Ua MrIOZOo6jJa3URX. but as H.J(ti_Il)2 J(li . The topic of surface area for nondifferentiable sUlfaces is a complex one. on a compact set. written by someone called M. . CmmaK. Ii ] . and uniform continuity of "'. 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.c(liI ll = J(li . Schwarz first observed. which we will not go into here. then we have IC(li ) . similarly.li _I l 2 + f'(�i )(li . bUl mow the. In this way. while the area of each approaches hi /2. the least upper bound of the areas of inscribed polygonal surfaces. Top view To in(T("asc the number of triangles. we can increase the number of triangles indefinitely. It is natural to suppose that the area of a surface is.354 a partition Chapter 9 1 [0. (c) Prove the same result in the general case. we maintain the hexagonal arrangement.li_I l2 + (J(lt) .li_I l 2 for some �i E [Ii _ I . . 1]. Hillt: Use the results of Prob = 10 0 < . .planes of the hexagons closer together. < In = of = lem 81. so that the triangles are more nearly in a plane parallel to the bases of the cylinder.
S and q E M . 1] + [0. (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}.Riemannian 1I1ettics 355 16. f M . If p : [0. q').b] satisfies 0. (In the theory of infinite dimensional manifolds. (a) By applying integration by parts to the equation on pages 3 1 8319.e] dl = 0.S or does no/ have to be disconnected. instead of piece wise Coo curves. show that (p. remember that the f ormula for length involves only first derivatives.e. (b) If p E B . I) [a au a a Y this result makes sense even if 1 is only C'. then 1 still satisfies the Euler equations (which are not a priori meaningful if 1 is not C2).S. show that (d& (u» I dJ U u=o = a2Ci aF lb t (0. ). Show that M . Use q!eS this fact and Lemma 7' to complete the proof of Theorem 7. g o n [a. (c) Conclude that if the C' function 1 is a critical point of J.) d L (e) = L(e 0 pl.) 18.B and B . (Show how to round offcorners of a piecewise Coo path so that the length increases by less than any given £ > 0. Let e: [0. I) + M be a curve in a manifold M with a Riemannian metric ( . lb �' (I)[g(l) . 1 (I» . (b) D u Bois Re )'mond� Lemma.S is disconnected. show that 17. . 1] is a diffeomorphism. these details become quite important. for all Coo functions � on [a. then = (1. a 1rbg(u)du.b] with Hillt: TIle constant c must be lb �'(I)g(l) dl = ° __ l b a � (a) = � (b) g is a constant. If a continuous function . /(1).l' �� (1'/(1). d d 19.B. by showing that M . q) ? min (p .S are disjoint open subsets of M . Show that the metric on M may be defined using Coo. /'(1» dl] dl. and Theorem 7 is false. e= We clearly have so we need to find a suitable � with �'(I) = g(l) .
(c) For those who know about complex power series: sinh x = sin ix . for convenience.. Show that sinh(coshI x) I I = cosh(sinh.and tanh.356 Chapter 9 20. 2 (cosh _ ' )'(x) = 21.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 ..' eX _ 2 eX cosh x = .. coshx (a) Graph sinh. 2 _1 1 .' eX + ex 2 sinh x tan h X = . Consider the problem of finding a surface of revolution joining two circles of radius I . I respectively.. and tanh are defined by sinhx = . and tanh. while cosh. The hyperbolic sine. I I (d) The inverse functions of sinh and tanh are denoted by sinh. 00). We are looking for a function Vx �. cosh. (b) Show that 2 2 cosh .x) = cosh(tanh. at a and a. hyperbolic cosine. and hyperbolic tangent functions sinh.x) = J 1 + x2 1 v1x R=I r:. situated. ) I cosh x = cos ix.denotes the inverse of cosh I [0. cosh.
then there is a unique such c. (d) There exists c with c cosha/c = I if and only if a :s: yo/ cosh Yo. so JY02 .2. = I e 0). cosh Yo [YO � 1 . Aa (e) = c cosh :!. It turns out that the surface for C has smaller area.Riemannian Metrics of the 357 form where e is supposed to satisfY x /(x) = c cosh c (c > e cosh :!. c Show that Aa has a minimum at a/yo.1 . with CI < a/yo < c'}. Examine the sign of 1/ y tanh y f y > O. 2 c. (c) Let c > O. or (b) Examine the sign of cosh y . 67 . If a < yo/ cosh yo. yo/ cosh Yo 0 a Yo/ cosh yo (e) Using Problem 20 (d). then there are two such c. If a = yo/ cosh Yo. and sketch the graph. namely c = a/yo = I / cosh yo. (a) There is a unique Yo > 0 with tanh Yo = 1/ Yo.] . show that � = JY02 . find the value of Aa there..y sinh y for y > O.I � .
= ±Yo. but the graph of /c. Yo The unique member of the family through (yo / cosh Yo . . pp. I) does no/ .[ Volume III. Q E (a.358 Chapter 9 [These phenomena can be pictured more easily if we use the notion of an envelopec. .x .a]. 176ff. a). the graph of f is tangent to q cosh Yo :' y = � x/ . ftr a < )'0/ cosh )'0. '" a Yo/ ('osh Yo the envelope at points P. 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 . cosh Yo y = ± . y= = cosh :: _ c c :: sinh C :: (' . is tangent to the envelope at points outside [a. so the envelope consists of the straight lines c cosh  x c = c cosh Yo. 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. I) is tangent to the envelope at that point.
k] and [a!'l.f (b) Apply this to F(x. (Consider the arclength function on a geodesic for Some Riemannian metric on IR. ag'. aF = 0. � = __ _ __ _ . i.k] are not the components of a tensor.� l. ax) ox'Y " [}2xl ox'Y " rk. ax'� ax'Y + L ax'Y .� ax'Y  x . ax'� ax' ' )=1 .Riemannian 1I1ebic.396..}. /' (1» . (c) Also show that i.1 359 give a local minimum for conjugate points of a geodesic in Volume IV.=1 (   ) . 'l ax'. = .k=1 11 n ax' a2xJ ax' ax} axk L [ij.. show that 24.j.. /' (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 _ . and note the remark on pg. with corresponding gil and g 'i) for the expression of a Riemannian metric. ax'� ax k L. + L. ay 23.)] 22. (a) Let x and x' be two coordinate systems. All of our illustratiOlls of calculus of variations problems involved an F which does not involve t.ct(Jx'Y . Chapter 8.Jl+7 to obtain directly the equation dy/dx = _ � ag'J �� ax} k L. .. Y]'.k=1 ax ax'Y ax'· ax'� n ax} a2x' ax' a2x) + " g'l' ox'ct .) ax. V.) For the corresponding [a!'l. 11 n [ij. y] ' = so that [ij.+ ax'/3 (Jx'l3ax'Y ox. Q. i. k] ax'.1J + (/') 2 . (Compare with the discussion of aF d aF (/(1). k=1 1=1 Show that any Coo structure on IR is diffeomorphic to the usual Coo struc Y r '. F.j. so that the Euler equations are actually f f.. Show that � which we eventually obtained in our solution to the problem.. ture. ax'· ax'� ax' . y) = . _ f and F: 1R2 + IR we have ( ) = O.dt ay (/(1).
). In Lcfons sur La Ciomitrie des Espaces de Riemann. ill theory a t least. go about finding I? satisfied by the ei 's.) = r[. How can we . and let L. e) ) . sufficient t o solve for the ei . deduce that so that A� = rD.i gij du i ® du J be another metric. Show that we must have (d) Show that = = " L g)r A �k + g. (c) To solve for I i t is. k]. ) = Li dx i ® dX i be the usual Riemannian metric on JR". j. (f) Deduce the result A. (Note that the curveS obtained by setting all but one Ii constant are geodesics. Ifwe consider ei as a vector field on JR". show that I (a/au i ) = ei . Let ( . k. directly from our equations for a geodesic. since they correspond to lines parallel to the x i axis.r A k) r=1 Bik . Suppose we are told that there is a diffeo morphism I : JRn + JRn such that Li ) gij dui ® du) = I' ( . and to solve for these we want to find differential equations 25. where u l .) Bi).J standard coordinate system on JR" . .) + Bk).360 Chapter 9 (a) Let al/au i = ei : JR" + JRn . Cartan uses j this approach to motivate the introduction of the ri� .i· (e) By cyclically permuting i .k = [ij. E. (b) Show that gil = (ei. . • • • ) un again denotes the .
I'l) + [a. ( . Let c be a curve in persurfaces {exPC (r ) v : Il v ll = M with dc/dl # ° everywhere. v is perpendicular to these hypersurfaces. w' E9 w") = (v'.b) be a (b) Conversely. Show that e = y o p satisfies p : [". (a) Show that ( . (Gauss' Lemma is the '�special case" where c is constant.) ds. then y is a geodesic. (The equation p"(I) = can be solved explicitly: p(t) = J' e M (. If (V'. 27. w')' + (v". n and (V". de /dl) 0. w")".) . Show that for v E Mc (r ) with (v. ( . the geodesic U r> eXPc(r) u .b) + M be a geodesic. where v E Mc (r ) with = OJ. (a) Let y : [a. de/dl) = constant. if e satisfies this equation.) p'(I)/1(I) 28. Describe the geodesics on M x N for this metric. and let diffeomorphism. we define ( . ) on V = V' Ell V" by (V' E9 v". (c) If e satisfies for Il' JR + JR. ) is an inner product.Riemannian Metlies 361 26. then e is a reparameterization of a geodesic. (b) Given Riemannian metrics on M and N. ) ") are two vector spaces with inner products. it follows that there is a natural way to put a Riemannian metric on M x N. where M'(s) = /1(5). and consider the hy (v.
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). . 32. (Gauss' Lemma still works on exp«('J). Conclude that E(y) < E(e) L(y)2 L(e) 2 E(y) = . y(b» = L(y) is a geodesic. (b) For any curve y show that with equality if and only if y is parameterized proportionally to arclength.362 Chapter 9 29. and suppose that expp is a diff omorphism on a neighborhood ('J C Mp of {ly ' (O) : 0 ::s I ::s I ).5 E(e). unless e is also a geodesic with (a) Show that in this coordinate system we have r6(p) = O. sufficiently small pieces of a geodesic mil limize energy. . . . a n ) . Hint: Recall the equations for a geodesic.) 30. . b) + M be a geodesic with L � (y) = d(y(a). b] + M be a geodesic with y(a) = p. then a curve y : [a. defined in a neighborhood U of p. let x be the coordinate system xo exp.. e(b» .5 .. Schwarz's inequality for continuous functions states that with equality if and only if f and g are linearly dependent (over IR). among all curves in exp«('J) . In particular. Let p be a point of a manifold M with a Riemannian metric ( . (a) Prove Schwarz's inequality by imitating the proof of Theorem 1 (2). ) . . y(b» .t . ) is a Riemannian metric on M and d : M x M + IR is the corre sponding metric. . If era) = y(a) and e(b) = y(b). Choose a basis VI . and note that a geodesic y through p is just exp composed with a straight line through 0 in Mp. 31. .5 (b . Let y: [a. show that [L�(Y)f . . Show e that y is a curve of minimal 1ength between p and q = y(b). ba ba L� (e) = d(e(a). If ( . so that each y k is linear. (c) Let y: [a . b] + M with d(y(a).
). roy 2. q ' E U has a unique geodesic of minimum length between them. y . and this geodesic lies completely in U. u (O) .ball is convex. Show that for sufficiently small .) for 0 # I E (8. the maximum of occurs at either q or q ' . ( dyk )2 _ " Y2rk dyi dyJ ] .q). (g) Let f: U + IRn be a diffeomorphism of a neighborhood U of 0 E IRn into IRn. Hint: If ' ( ) is y(O) d(r y)2/dl = y) y(l) s. > O.. y y o = tangent to exp(S. '. 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..Riemannian 111etricJ 363 = r(q) d(p. Show that d2 ( r o d) = 2 [ . So that r o y Lk (yk)2.J d k d I is sufficiently small. 8). and let be the unique geodesic of length < joining them. d(r y)/dl o. Show that the following is true for all sufficiently small ' > 0: if is a geodesic such that E exp(S.p B. dl 2 L. i.J. r(q') < . {v E Mp : IIvll :s: 'I and S. = {v E Mp II v ll = d. then Using part (a)..) and such that y'(O) is tangent to exp(S. then there is Ii > 0 (depending on such that >t exp(B. Show that for sufficiently small " the image of the open .). dl .k iL. Show that exp ( { v E Mp : IIvll < d) is geodesically convex for sufficiently small .. (f) A set U C M is geodesically convex if every pair q.. 0 so that (d) Let B. dl dl k (c) Note that l L ddyi dyl n2 L( dylk )2.
If c is C'. (a) There is an everywhere differentiable curve that length of c 1 [0 .. then there = q Hint: In Mq we clearly have Since expq is locally a diffeomorphism there are 0 < K.O) is ! (b) Consider the situation in Corollary except that > 0 for near o. c (h» L (c I [0.c (O)1 1. ..0) is I length from ����+ (�. I C(I) V(I) if and only if approaches a c is C'. 1.364 Chapter 9 33.O) to (!.ICi (U (I).. h] # lim c(t) (I. f(I» in ]R2 such = Hint: Make c look something like the following picture.I) 1 2 dl a u(h) q. < for all tangent vectors (c) Conclude that I d(u��(t» I [ h = = lu i . then = a.:. length from (!. h�o Ic(h) .'=__"::::__"10 _ u'(t)2 + I '!. and suppose limit as + 0 (even though v(O) is undefined).O) to (1. h]) .t) I · K2 such that V at points near lim h�o lim h�o d(p.. Show that if is some K > 0 such that for all near 0 we have I I u'(t) I 1 5. 1 d�.
W E MP l then L (exp o c) = l lim . lim v. uSe Problem 34 to show that II IX . Let f : M + M be a map of M onto itself which preserves the metric d. 36. show that if c is the straight line joining v.w�O L(c) (b) Similarly. 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. show that L(c) is the least upper bound of inscribed piecewise geodesic curves.w ) (c) Conclude that d(exp v.(y'(O» = d/(y(I» dl Show that II /'(X) II = II X II .w�O L (cv.w.Riemannian Metrics 365 (d) If c is C'. then L(yv. exp l y)] = hm II IX II · II I Y II II X II · II Y II _ HO I1=0 . = p. (c) Given X. (a) Using the methods of Problem 33. Y E Mp. (b) Define /' : Mp + M f(p) as follows: For y a geodesic with y(O) /. exp w) . and that /. 34. ) and corresponding metric d.w is the unique geodesic joining exp(v) and exp(w). ric space structures determined on M and N by their respective Riemannian metrics. o Let f : M + N be an isometry.I Y II 2 II X II 2 + II Y II 2 2(X. Show that f is an isometry of the met 35. v. Let M be a manifold with Riemannian metric ( . and Yv. let . = l . w ) = 1.(cX) = c/'(X). then f 0 y is a geodesic. l�� II v _ w ll v. if Yv. (a) If y is a geodesic.w = exp 0 cv.
Show that if II is ti'O hm . The same result then holds in allY vector space with a Euclidean metric ( .lI v ll I = (v. /'(Y» f(p) . TIlus. critical . then Hint: (c) Let y : [0. then the limit is Dv (v)(w). (b) Conclude that if w is linearly independent of v. one can use the equation (u. Y)p = (d) Part (c) shows that 1' : Mp + Mf(p ) is a diffeomorphism. ) to y(to + u).) y(O) close enough to 10. If v : IRn + IR is the norm.lIv ll . II v ll · cos Ii where Ii is the angle between u and v.366 Conclude that Chapter 9 (X. lI v + l w ll . w ) II v ll . a contradiction. v) = lI u l i . Alternately. then dL(&(u» /dulu�o paths for length cannot have kinks. and then that f'(X + Y) = IRn with w # 0. and finally the rest of y. # 0. and hence an isometlY. and suppose that y'(lo+ ) # y'(to) for some 10 E (0. 1] + M be a piecewise C' critical point for length. ). Use this to show that f is itself a diff omorphism.lIl w ll t # O. e 37. (/'(X). 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. w E f'(X) + f'(Y). show that hO lim II v + l w ll . (a) For v. y(l) y(t. 1 ).
39. and some geodesics may come back to their initial point). 41. = r"2 = rl. Find the metric d induced by the Riemannian metric it acquires as a subset of ffi. . there is a number a such that y(t + a) = y(l) f aU I). )) is the usual Riemannian metric on sn (thc one that makes the inclusion of Sn into jRn+' an isometry).L + jR2 which is a local � P(� isometl)� but which is usually not oneone. p }. and that every two geodesiCS intersect exactly or once.y.Ri£1nannian 1\1etfie. (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.3. Consider a cone C (without the vertex).. TIlese results show that pH provides a model for "elliptical" nonEuclidean geometry. and let L be a generating line. y) E jR' ( .L onto jR 2 produces a map f: C . Q» Show that evcry geodesic y : jR + pn is closed (that is. The sum of the angles in any triangle is > 1f. = . 40.�. r�. Unfolding C . y all other ri� = O. Y2 (a) Compute that : y > O} ! ri. )) on pn such that g * « . = . Investigate the geodesics on a cone (the number of geodesics between two points depends on the angle of the cone. Consider a cylinder Z C jR3 of radius t. TIle Poincare upper half plane with the Riemannian metric Jf' is the manif old {(x. (a) Show that there i s a unique Riemannian metric « . Let g : S" + pn be the map g(p) = [p] = {p.\' 367 38. ) = dx ® dx + dy ® dy .
A. (e) Show that if y is a geodesic and p ¢ y. The sum of the angles in any triangle is < 7[. together with the straight lines parallel to the yaxis. show that all the geodesics in Jf2 are the (suitably pa rameterized) semicircles with center on the xaxis.368 Chapter 9 (0..bc > O are isometries. and that we can take any tangent vector at one point to any tangent vector at any other point by some J. B' . . 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'. b . ad .. so that the upper halfplane is complete. . Considering (b) Let C be a semicircle in Jf2 with center at it as a curve t 1+ (t. then there are infinitely many geodesics through p which do not intersect y. y(t)).: J 3' YN. upper half plane is a model for Lobachevskian nonEuclidean geometry. Consider the upper halfplane as a subset of the complex num bers <C. (d) Show that these geodesics have infinite length in either direction. (f) For those who know a little about conformal mapping (compare with Prob lem IV 76). c) and radius R. These results show that the Poincare = = A ti c : c . then length B C = length B'C' and the angles at B and B' and at C and C ' are equal ("sideangleside"). Show that the maps z J( ) = az + b cz + d a. c. show that y'(t) 2 y(t) (c) Using Problem 27. d E 'llI..
(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 ( . show that the normal bundle v M x Ell TM '" (TN)IM. Prove that there is a geodesic y : [0. (b) Using the notion of Whitney sum Ell introduced in Problem 352. conclude that TE '" ]f*(�) Ell ]f*(TM) . where N is a Coo manifold. . show that v M and N be geodesically complete Riemannian manifolds. having the property that y is a minimal geodesic between any two of its points. ) is a Riemannian metric on N. )) is complete. ( . E. and give N the Riemannian metric described in Problem 26. This problem presupposes knowledge of covering spaces. 47. Suppose that every geodesic y : [a.g'( . ).00) > M with the initial value y(O) = p. 48. (a) Given an exact sequence of bundle maps as in Problem 328. (b) If � = ]f : E > M is a smooth kplane bundle over M". (a) If M n C N n +k is a submanifold of N. 44. Let M be a Riemannian manifold such that every two points of M can be joined by a unique geodesic of minimal length. Let is indeed a kplane bundle. 46. b 1 > M with initial value y(a) = p can be extended to all of JR. and choose a fixed point P E M. Let g : M > N be a covering space. Show that the Riemannian manifold M is geodesically complete.. more generally. then g'( . ) is a Riemannian metric on M. show that E2 ". Show that the Rie mannian manifold M x N is also complete. and (M. where the bundles are over a smooth manifold M [or.Riemanni(J}l Metric. Ell E3 . Does it necessarily follow that the Riemannian manifold M is complete? 43. (b) If � = ]f : E > M is a smooth bundle. Then there is a unique Coo structure on M which makes g an immersion. over a paracompact space]. )) is complete if and only if (N. Let p be a point in a complete noncompact Riemannian manifold M. show that the normal bundle of M C E is equivalent to �. Let M be a manifold with a Riemannian metric ( . 49. 45. 369 42.
(b) Use Problem 329 to conclude that there is a neighborhood of some S' C M which is not orientable. it is not hard to show that a bundle over S' is trivial if and only if it is orientable). 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. show that the normal bundle v of S' C M is not orientable.) . (Thus. (a) Let M be a nonorient able manifold. in fact. any nonorientable manifold contains a "fairly small" nonorientable open submanifold. Using Problem 47. but the same conclusions will hold if each (TM) IS' is orientable.370 Chapter 9 50.
ab aI from G x G to G from G to G are continuous. The simplest example of a Lie group is JRn . It clearly suffices to aSSume instead that the single map (a. We will mainly be interested in a very special kind of topological group. many of the results and concepts of the It will also play an important role in later Volumes. b) a 1+ 1+ T preceding chapters. consequently the map (x. which can be expressed in coordinates as one of the two maps (x. The functions x 1+ cos 2n: X and x 1+ sin 2n: x are Coo functions on JR/Z. y) is Coo. y) 1+ xy' is Coo. and illuminates. and at each point at least one of them is a coordinate system. One way to put a group structure on S' is to consider it as the quotient group JR/Z. The circle S' is also a Lie group. y) 1+ X . and these groups can be studied by the methods now at our disposal.cos2JfX sin 2Jfy. b) 1+ ab' is continuous. it even suffices to assume that the map (x. As a matter offact (Problem I). y) 1+ xy X 1+ X . It clearly suffices to assume that the map (x. b E G being denoted by ab) such that the maps (a. y) 1+ xy.y) sin2Jf(x .' from S' x S' to S' is also = 1+ 1+ cos2Jf(x . cos2Jfx cos2JfY + sin 2JfX sin 2JfY sin 2Jfx cos2JfY . where Z C JR denotes the subgroup of integers.Y 1+ xy.I are Coo functions. Thus the map (x. A Lie group is a group G which is also a manifold with a Coo structure such that (x. because in the study of these problems the groups of autom orph isms of various structures play a central role.CHAPTER 1 0 LIE GROUPS his chapter uses.I In In In JR x JR > JR > S' = JR/Z. 371 . where we are concerned 'V'lith geometric problems. y) (x. y) 1+ xy is Coo. with the operation + . A topological group is a space G which also has a group structure (the product of a.y) = Coo .
372 C7wjJler 10 If G and H are Lie groups. generally. which is equivalent to the condition that. is norm presen1ing. and in particular. The argument in the previous paragraph shows. Multiplication of matrices is Coo. To show that 0(11) is a Lie group we must show that the map (x. considered as a 2 subset of IRn Since the function det : IRn' > IR is continuous (it is a polynomial map). By Proposition 21 I . 1R)). where A t is the transpose of A. cOJ"isting of all A E GL(n. A t = I. (This gives another proof that S' is a Lie group. 1R) is a homeomorphism (since O(n) is a submanifold of GL(II. 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. f S' C 1R 2 can be considered as the group or . is easily seen to be a Lie group. that if H e G is a subgroup of G and also a submanifold of G. then G x H. The general linear group GL(II.b') represent the same element of S' x S' if and only if a ' . Many imporlant Lie groups are matrix groups. but this is true because the inclusion of 0(11) > GL(n . a manifold.' )F = det A ij / det A. The torus may also be described as the quotient group IR x 1R/(Z x Z). since the entries of A B are polynomials in the entries of A and B.{OJ) is open. IR) = det. and the direct product group structure. the torus S' x S' is a Lie group. of dimension 11(11 . 1R).. This condition is equivalent to the condition that the rOws [and columns] of A are orthonormal. where A ij i s t h e matrix obtained from A by deleting row i and column j. IR). the pairs (a. and thus inner product presen1ing. 1R) with A .1)/2. 1R) is the group of all nonsingular real II x II matrices. b) and (a'.' (1R . One of the most important examples of a Lie group is the orthogonal group 0(11). it suffices to show that it is continuous. with the product Coo structure. and hence can be given the Coo structure which makes it an open submanifold of IRn' . In particular. the set GL(II. is also Coo as a map from 0(11) x 0(11) to 0(11). Later in the chapter we will havc another way of proving that 0(11) is a Lie group.' which is Coo on GL(II. y) 1+ xy .a E Z and b' . Smoothness of the inverse map follows similarly fi'om Cramer's Rule: IIIIIIIII!! (A.b E Z. with respect to the usual basis of jRn.e.
the component containing the identity I. If G is a Lie group. is a subgroup.1 K C K. For example. a subgroup which is an (imbedded) submanif old is always a Lie subgl"Oup. then K is also open. •:. As we have seen. so K is normal. then the immersion.) 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. but we will not need this fact. respectively. Since e = aIa E aI K. which proves that K is a subgroup. PROOF If a E K. the two components consist of all A E 0(11) with det A = +1 and det A = . It even turns out. we have bKb1 C K. since G is locally connected. S 3 is the Lie group of quaternions of norm I. It is know that these are the only spheres which admit a Lie group structure. If G is a Lie group. We define a Lie subgroup H of G to be a subset H of G which is a subgroup of G. Hence K is a Lie subgroup. the group of all Euclidean . since b 1+ a . then aI K is connected. PROPOSITION. image of L in S l x S l = lR x lR/(Z x Z) is a dense subgroup.I . I\1oreover. then the component K con taining the identity e E G is a closed normal subgroup of G. Clearly SO(I1) = {A E 0(11) : det A = I } . The group SO(2) is just S l . then K is an open Lie subgroup. Since this is true for all a E K. which we have already seen is a Lie group. This is not accidental. we have aI K C K. K is dosed since components are always closed. after some work (Problem 18).I b is a homeomor phism of K to itself. The group 0(11) is disconnected. we have K. ex) for e irrational. For any b E G. it follows similarly that bK b I is connected. so K is a submanifold and a subgroup of G. if L C lR x lR is a subgroup consisting of all (x. and also a Lie group for some Coo structure which makes the inclusion map i : H > G an I l lnl l I . that a subgroup which is an immersed submanifold is always a Lie subgroup. As a final example of a Lie group.Lie Groups 373 of complex numbers of norm I. we mention £(11). Since e E bK bI . Similarly. If G is a topological group.
which shows that E(I1) is a Lie group. b It is easy to see that this is t�ue if we merely hav(' for all a E II E G. since translations and orthogonal transformations do not generally commute. with inverses La . 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) . COJlsequently.1 = A B.1 = A TaTbB. In fact. . T where A C 0(11). A little argument shows (Problem 5) that every element of E(I1) can be written uniquely as A .. G. the maps La* : Gb Ra * : Gb 7 Gab 7 Gba for all are isomorphisms.1 (x) = A(A. E G. and T is a translation: T(X) = Ta (X ) = X + a . For any Lie group G. isometries of IR". Consequently. Clearly the component of e E E(I1) is the subgroup of all A T with A E SO(I1). Now E(I1) is not the direct product 0(11) x IRn as a group. La : G > G and Ra : G > G. Notice that La and Ra are both diff eomorphisms.e.l and Ra. A vector field X on G is called left invariant if La.l = A Ta_bB. by La(b) = ab Ra(b) = ba .374 Chapter 10 motions. A Ta(BTb). if a E G we define the left and right translations. so A TaA.X = X Recall this means that for all a .1 TB(a_b). respectively. Vie can give E(I1) the C"" structure which makes it diff eomorphic to 0(11) x IR" . Consequently. i.l .
. c ) . COROLLARY.rith these values at e. Then XI .Xe(xi) 0 = Xe (xi La) . Xn are clearly eveI)"Alhere J : TG > linearly independent. . PROOF. Coo . A Lie group G always has a trivial tangent bundle (and is consequently orientable)./f. Cj a(xi 0 La) � ax} 0 La) j= 1 L c j D"+j Ji (x(a). It suffices to prove that X is Coo in a neighborhood of e. Choose a neighbor hood V of e so that a. •:. Let X" . x(e)). This implies that X is I e where Xe = j =l L cj D X a I e which shows that XXi is Coo . • : 3. .Lie Grau/}. XXi (a) = Then " Xe (xi = j= 1 = . . . .I 375 PROOF. . c ' . Then for a E V we have XXi (a) = 2. Let (x. since the diffeomorphism La then takes X to the C"" vector field La. . ' . . La. so we can define an equivalence G x lR" J (t J=I ciX (a) i ) = " (a. PROPOSITION. 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). U) be a coordinate system around e. Every left invariant vector field X on a Lie group G is COO.. Choose a basis X' e . .X around a (Prob lem 51). " " X"e for Ge. Since the map (a. X" be the left invari ant vector fields v. b E V implies ab ' E U.
YJ '= 0 f all X. Consequently. then or we can define an operation [ . are Lie algebras with bracket operations [ . X S') is also abelian. 2. and consequently must be abelian. ] on Ge by [X.) We will also use the more customary notation 9 (a German Fraktur g) f r £(G). 1R) or f the Lie algebra of 0(11). Y. or The Lie algebra of IRn is isomorphic as a vector space to IRn. Henceforth we will use X.376 Chapter 10 A left invariant vector field X is just one that is Larelated to itself for all a . or 0(11) In general. to denote the left invariant vector fields with X(e) = X. Z] . Xl Consequently. YJ i s left invariant i f X and Y are. ] satisfying [X. Clearly £(IRn) is abelian. If If. . etc. x V as 2 a set) by It is easy to check that this makes V into a Lie algebra. ] i f i = 1 . we write or gl (ll. Y](e). the Lie algebra £(S' x . ] operation is assumed alternating. the tangent space of GL(n . . Ell V2 (= V. we call a Lie algebra abelian or commutative if [X. YJ = 0 'jacobi identity" f all X. ] on the direct sum V = V.. to denote elements of Ge. since the vector fields a/ax . X] = o [[X. X]. 1R) is an open submanifold of IRn ' . Y. or Since the [ . Y. 1R) is more complicated. etc. Z] + [[Y. We can then define an operation [ . together with this [ . etc. Y(e) = Y. IR) at the identity I can be . and X. Y. Proposition 63 shows that [ X . it is also skewsymmetric. (Sometimes the Lie algebra of G is defined instead to be the set of left invariant vector fields. YJ = [X . with a bilinear operation [ . are left invariant and [a/ax . Consequently. The vector space Ge . This notation requires o some conventions f particular groups. Z E V. The Lie algebra £(S') of s' is I dimensional. is called the Lie algebra of G. and that £(G x H) is isomorphic to £(G) x £(H) with this bracket operation. YJ = [Y. ] operation. [X. 1R) f the Lie algebra of GL(Il. a/axj ] = O. . a Lie algebra is a finite dimensional vector space V. Since GL(n. YJ. and will be denoted by £(G). The structure of gf (11. X] + [[Z ..
0'=1 n with (constant) partial derivatives � (xkl ax" 0 LA) = A ki {0 j =I J # I. Now the function xkl Mxkl (A) = L Mij :ij a i. NlI = � (MN ..kl _ Mjl Mx {0 ax ij 11 .IA k Thus.j = L Nij . 1R) i s the linear function LA) (B) 0 = xkl (AB) = L A k .l . M/ (xkl ) = M/ (xkl LA) .B. we have i. N/ (Mx. If \ve use the standard cOOldinates xij on jRn.NM lkl rJ . k=i k # i. 1R) as f ollows.. So = L Mi/ A ki = L M. So if N is another 11 x a __ . 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. Wc compute the function M xk l on GL(n. 1R) corresponding to M.j I = MA (xkl ) = LA.lith jRn 2 .kl ) matrix. 1R) > GL(n.I. 1R).=) n 0'=1 n •.[M. 0 1 (x kl 0 LA : GL(n . For every A E GL(n. a Xl} a = L Nkj Mjl = (NM lkl· j=l From this we see that " .Lie Groups 377 identified ".(Mx kl ) .2 . ax l / a k.
. If we did not know the dimension of 0(11).ba . if we identify Chapter 10 gl(I1. IR) we can consider 0 (11) / as a subspace ' of GL(I1. j)plane) j . then k�i differentiating give.NM.Aj/(O) . which is exactly the dimension of 0(11). and we denote (A(t))ij by Aij (t).b] = ab . .378 thus. ] satisfies the Jacobi identi�: Since 0(11) is a submanifold of GL(I1. j with i < j. then [ . :2 Thus 0(11)1 C jRn can contain only matrices M= ( M which are skewsymmetric. IR) with IRn' . we could use the following line of reasoning. Notice that i n any ring. the bracket operation is just [M. if we define [a. 0 M1 2 M1 3 . and thus identify 0(11) with a certain subspace of IRn . If A : (8. IR)/. we can define a cun'e A : IR > 0(11) by A (') � ( j ' cos t . N] = MN . For each i.8) > 0(11) is a curve with A(O) = I. so 0(111I must consist exactly of skewsymmetric matrices. which shows that L A ik (t)Ajk (t) = �ih A i/ (O) = .1 )/2. This subspace may be determined as follows.sin ! sin ! cos t J (rotation in the (i. Mi n 0 ) This subspace has dimension 11(11 . Mi n M1 2 0 0 M1 3 .
This correspondence between Lie subgroups of G and subalgebras of � turns out to work in the other direction also. and O's elsewhere. leaves the distribution /j. He. The proof of uniqueness is left to the reader. In particular. i. so Lb . For each E H C G. Let PROOF. ([X.1 )/2. and are irelated. if E H. Y](e) Thus. i). and O(n) must have dimension n(n . = = La. I 's on the diagonal except at (i. and let i : H > G be the inclusion. f] X [ i. then [X. that is. Then the set of all A'(O) span the skewsymmetric ma trices.. He with this induced . (a. among themselves. X La o f i 0 La. j) and (j. Consequently. let /j. •: a X(a) e. 4. Let H be the maximal integral manifold of /j. 1 operation. X) = X(a). Using Theorem 67. b (b) e. THEOREM. invariant. La. Since i* : He Ge is an isomorphism into.(/j. X(a) X = i. It follows immediately that Lb permutes the various maximal integral manifolds of /j. i) and (j. we have left translations 7 X X a La : H and So > H. moreover. We do not need any new calculations to determine the bracket operation in o(n).1 .ba. which means that [X. 1 operation is juSt I) = £(H). b) ab. This implies that H is a subgroup of G. if Y E and are irelated. Now this map is clearly Coo as a map into G. and I) a subalgebra of �. other words. He C Ge = � is a subalgebra of �. The fact that I) is a subalgebra of � implies that /j. consider a Lie subgroup H of any Lie group G. Any X E He can be extended to a left invariant vector field on H and a left invariant vector field on G. Then there is a unique connected Lie subgroup H of G whose Lie algebra is I). To prove that it is a Lie subgroup we just need to show that 1+ is Coo . j). Y](e)). it follows that it is Coo as a map into H. takes H to the maximal integral manifold containing Lb.Lie Graul" 379 with sin t and . In fact. we can identify He with a subspace of Ge. He is a subspace of � which is closed under the . = so Lb.a be the subspace of Ga consisting of all for X E I).. For E G. Yl [X. is an integrable distribution. [ G be a Lie group. then clearly Lb. containing If b E G. (H) = H. then Lb. (i.a) = /j.sin t at (i. = La : G > G In. Hence O(n)/ must consist exactly of skewsymmetric matrices.
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 . . For any a E G we clearly have 4> 0 La = L¢(aj 0 4>.X = L¢(aj.. homomorphisms between Lie groups. Lie algebra is isomorphic to Ihe lie algebra ifsome Lie group. because IR is a vector space of uncountable dimension over Q. ds which means that 4>(1) = cl for some c (= 4>'(0)). But if 4> is Coo. We will soon see to what extent the Lie algebra of G determines G. then 4>.. we next consider. the map algebra homomorphism. then .. simply by 4>. so if X E Ge. We continue the study of Lie groups along the same route used in the study of groups. There are an enormous number of homomorphisms 4> : IR > 1R.. we will denote 4>.x is the left invariant vector field on H with 4>*a%(a) = 4>.P. : � > Ij is a Lie 4>. Y 4>.aLa. Consequently. and every linear transformation is a group homomorphism. If 4> : G > H is a Coo homomorphism. and % value 4>*eX at e. It then follows from Theorem 4 that ever . Yj = [4>.. then the condition 4>(s + I) = 4>(s) + 4>(1) implies that d4>(1 + s) ds = 0 gives d4>(s) . X.. X = %(4)(a)). Thus X and % are 4>related. 1R) for some N.. Later ) on we will be able to obtain a "local" version of this result. Having considered subgroups of Lie groups (and subalgebras of their Lie algebras). Usually.4>. more generally. Y]. [X. : � > Ij. 4>. For example. suppose that G = H = JR. 4>. 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)... that is. : Ge > He . X + b4>.. (aX + bY) = a4>..
Consequently. Now suppose that G = 1R. Let I (German Fraktur k) be the subset f C g x l) of all (X. and G is connected. A neighborhood of the identity e E S' can be identified with a neighborhood of 0 E 1R. if there are two Coo homomorphisms Vr. 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. 7 w. . then 4> = Vr. there is an open neigh borhood V of (e. : IR > IR is multiplication by c. but H = S' = IR/Z. Let G and H be Lie groups. . : IR > IR is just multiplication by c. a Lie algebra ho momorphism g > I) may not come from any Coo homomorphism 4> : G > H. we do have a local result. I K. 5.(X. If H. By Theorem 4. I is a subalgebra of g x l) = £(G x H). for X E g. giving rise to an identification of £(S') with 1R. Vr: G > H with 4>. Since <I> is a homomorphism. Notice that the only continuous homomorphism 4> : S' > IR is the 0 map (since {OJ is the only compact subgroup of 1R). THEOREM. = <1>. then w : K > G is a Coo homomorphism. Consequently. and <1> : g > l) a Lie algebra homomorphism.e} Ge is an isomorphism. : G x H > G is projection on the first factor. For X E g we have SO w* : K(e. However.b.Lie Groups 381 rational numbers). and w = H. <I> (X)) = X . . there is a unique connected Lie subgroup K of G x H whose Lie algebra is f. X = <I>(X). If H2 : G x H > H is projection on the second factor.. 4>. We can identify £(IR) with 1R .e) E K such that w takes V diff eomorphically onto an open neighborhood U of e E G. we Can define Moreover.ab E U. The continuous homomorphisms 4> : IR > S' are clearly of the form once again. <I>(X)). Clearly the map 4>. 4>. X E g we have 4>. P ROOF.
neighborhood of e generates G . COROLLARY.X Given 4>. A connected Lie group G with an abelian Lie algebra is itself abelian . <I> (X)).X•  . If two Lie groups G and H have isomorphic Lie algebras. It follows that G is abelian. For every X E Ge.382 Chapter 10 TIle first condition on 4> is obvious. G is locally isomorphic to IR". define the oneone map = ][2. PROOF. PROOF. x so £(G') = f. (X. 1jr: G > H.�O . let 4> be the map given by Theo rem 5. e: G > G x H by e(a) = (a. •: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. The image G' of e is a Lie subgroup of G x H and for X E A we clearly have e. 4> is a diffeomorphism in a neighbor hood of e E G . 1jr (a)). 4> : IR > G such that 8. •:. so 4>. •:. = (X. Since 4>*e = <P is an isomorphism. and that all connected Lie groups with a given Lie algebra are covered by the same simplyconnected Lie group. then they are locally isomorphic. By Corollary 6. COROLLARY. 7. As for the second. so ab = ba for a. COROLLARY. (X. then w. 6. <I> (X)) = X. if X E g. Given an isomorphism <1> : A > I). since (Problem 4) mH' . which implies that 1jr (a) = 4>(a) for all a E G . there is a unique d4> dl Coo homomorphism 1 . Thus G' = K. b in a neighborhood of e. <I> (X)) = <I>(X).
We know that integral cun'es of % exist locally. .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. e) of 0 E IR there is a map 4> : (e.X. 6") > G with . which proves uniqueness. Irl e/2 r �O ddl4> I Isl. < o.( ) du u=o j' = L¢(r»XU) = %(4)(I))U)./(4)(1)) " 1. 4>(s + 1) = 4> (s)4> (I ) and Clearly <I> is a Lie algebra homomorphism. III.I ) and define 4>(1) = { 4>(e/2) · · .0 dl ' /(4)(1)4>('')) . SECON ( D DIRECT) PROOF. By 111eorem 5. I = O. then 1 1 is an integral curve of % which passes through 4>(s) at time clearly true for 1+ 1+ 4>(s) . The same is so 4> is a homomorphism. 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.Lie Groups F IRST PROOF. Uniqueness also follows from Theorem 5. ls + 1 1 (:!. If homomorphism.· 4>(e/2) · 4>(r) ./(4)(1)) lim = h'l>O h = � / 0 L¢(l) 0 4> du u=o d4> = L ¢(l)' . Define <1> : IR > £(G) by 383 <I>(a) = aX. . . 4>(1) 4>(s + I). Conversely) if 4> : IR > G is an integral cun'e of %.:. I 1 Thus 4> must be an integral cun'e of X. on some neighbor hood (e. and < < e 4>: IR > G is a Coo d4> ( f) = lim /(4)(1 + ")) . they can be extended to all of IR using the method of the first proof.
4>(1 ). then clearly A A2 A3 + + +. More interesting things happen when we take G to be IR . . We thus see that there is a unique I parameter subgroup 4> of G with given tangent vertnr d4>/dl(O) E Ge. Notice that IR .' I A lk :s: nk l A l k . (N + K ) ! :s: y. some of which will be briefly reca pitulated here.{OJ. Consequently: IA + BI IA B I :s: :s: I A I + IB I n I A I ' IBI: I . GL(n.+ +. If A = (a'j ) and I A I = max laij I.. Ii 2! 3! hence I A lk I A N+K AN (n I A I) N (nI A I )N+K +". IR). IR) where · now denotes matrix multiplication. The solutions of these equations can be written formally in the same way 4> (1 ) = exp(I4> ' (O)).."+ > 0 as N > 00.{OJ is just GL( I . All Coo homomorphisms 4> : IR > must satisfy the analogous diff erential equation 4>' (1 ) = 4>'(0) . must satisfy 4>' (1 ) = 4> ' (0)4>(1 ) 4> (0) The solutions of this equation are 4>(1 ) = = I.nN! (N + K ) ! :s: nk. e¢'(O)r. where exponentiation of matrices is defined by exp(A) = I+ This follows from the facts in Problem 56.384 Chapter 1 0 A homomorphism 4> : IR > G i s called a Iparameter sub group o f G . 4> (0) = I. with 4> (S + I) = 4>(s)4>(I).{OJ. with multiplication as the group operation. Then all Coo homomorphisms 4> : IR > IR . We have already examined the I parameter subgroups vi Iii .
h_O lim exp(I</>'(O) + h</>'(O)) .. If 1fr : G > H is any Coo homomorphism. then Coo homomorphism We clearly have Ge exp 0 1fr. = 1fr 0 expo exp G�H l � He l exp . exp(11 + 12 ) X = (exp I .I .Lie Grau/}. if </>(1) is defined by (* *).. h. Given X E �. PROPOSITION. I] exp(I</>'(O)) +.+O lim Ii + h2 </>'(0) 2 h 2! exp(I</>'(O)) = </>'(0)</>(1). X) (exp 12 X) exp(IX) = (exp IX).. let </> : lR > G be the unique with d</>/dl(O) = X. then </>' (1) = = (i. so that exp takes a neighborhood of 0 E Ge diff eomorphically onto a neighborhood of e E G. so </> does satisfy (*) . The map exp : Ge > G is Coo (note that Ge '" lRn has a natural Coo structure). 9. and 0 is a regular point..exp(I</>'(O)) h [exp(h</> (O)) 1 = lim h'l>O h</>'(O) = . (exp A ) (exp B). Then exp(X) = </> ( 1 ) . if A B = BA. Moreover (see Problem 56). we now define the "exponential map" exp : � > G as follows.I 385 so the series for exp(A) converges (the J) th entry of the partial sums converge). then exp(A + B) Hence. and convergence is absolute and uniform in any bounded set. For any Lie group G.
I 1=0 = d dt I 1=0 exp(lv) So exp. The tangent space (Ge x G)(X. So exp. COROLLARY. then also 4>. If we identify a vector V E (Ge) o with Ge.X. (tX) = 4>(exp(IX)). a) can be identified with Ge Ell Ga. Coo. exp(Vr. dl r �O Then Vr 0 4> : IR > H is a homomorphism with d(Vr 0 4» = Vr.a) o f the Coo manifold G e x G a t the point (X. and hence oneone.o is the identity.o(v) = = d exp(c(I)) dt v.:. •: 1 1 1 0. .X) = Vr 0 4>(1) = Vr (exp X) .a) = 0 Ell X (a).. Then Y has a flow a: IR x (Ge x G) > Ge x G. We define a vector field Y on Ge x G by Y(X. o. Since it follows that exp is Coo. let 4> : IR > G be a homomorphism with d4> = X. Therefore exp is a diffeomorphism in a neighborhood of O. dl r�O Consequently. . (X) = But then e = exp 4>. PROOF.. contradicting the fact that 4> is oneone. Given Vr : G > H. and X E Ge.386 Chapter 1 0 PROOF.p( X (p)) = 0 for some nOllzero X E �. which we know is exp X = projection on G of a ( l . then the curve C(I) = IV in Ge has tangent vector v at o. If 4>. Every oneone Coo homomorphism 4>: G > H is an immersion (so 4>(G) is a Lie subgroup of H). 0 Ell X).
) = exp I Y. it follows from the above that 4>(e/2) = exp(X /2). . exp X /4 E exp( t U ) . . 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. e) = 4>(e/2.I . So 4> is Coo at e. Yt l · · . so X/4 = Y/2. X. . if a = exp(X/2) for X E U.. .. I] . Every continuous homomorphism 4>: JR > G is Coo. 4> (se) = expsX 1 2. and thus everywhere. COROLLARY. . X E exp( ! U). This shows that every a E exp( t U) has a unique square root in the set exp( t U). Now the map Vr : JRn > G given by Vr (l" . By continuity. •:.Lie Groups 387 I I . t U). •:. The map 1 1+ 4>(exp IXi) is a continuous homomorphism of IR to H. then 2 a = exp(X/2) = [exp(X/4)] . ( � IJ . Let U be a starshaped open neighborhood of 0 E G. Thus. COROLLARY. PROOF. i f a = b2 for b E exp( ! U). Now choose e > 0 so that 4>(1) E exp( ! U) for I I I � e. Let 4>(e) = exp X.. so 2 2 exp(X/2) = a = b = [exp(Y/2)] = exp Y. Then on V. (exp In Yn ) . For any a E exp( ! U). By induction we have 4> (e/2n ) = exp(X/2n ) . then S o a has a square root i n exp( b = exp(Y/2) for Y E U. for all s E [ . Moreover. Choose a basis X" . (exp In X. 4> (m/2n . ) ) = (exp I. X) . so there is Yj E He such that 4> (exp IX. In ) = (exp l. Xtl · · . 4> = (4) o Vr ) o r ' . Y E U it follows that X/2 = Y. 4> ( expI. Since 2 [4>( e/2)] = 4>(e) = [exp X/2f . Every continuous homomorphism 4> : G > H is Hence Coo. on which exp is oneone.) · · · (exp ln Xn ) is Coo and clearly (*) and (*) shows that 4> 0 Vr is Coo. . )m = [exp(X/2 n )]m = exp(m/2n . X" for G. . PROOF Since X/2.
COROLLARY. exp M E O(n). the equation exp(X + Y) = exp X exp Y holds whenever [X. Then I = A .1 = exp(M). so e p N = (exp M) . to indicate that a function c : IR > Ge has the property that C(I)/13 i. Y] measures. we have = I. bounded for small I. •:. M = _Mt. [X.e. Just as in GL(n. 2 The properties of the particular exponential map exp : IRn may !lOW be used to show that O(n) is a Lie group. At = (cxp M)(exp N). O(n) is itself a submanifold of GL(n . If G and G' are Lie groups which are isomorphic as topo logical groups. Y E Ge. i. IR ») > GL(n. (2) (I) exp lX exp l Y = exr l (X + Y) + '2 [X. since exp(M + N) (exp M)(exp M) So if M is skewsymmetric. THEOREM. Let At = exp N. For sufficiently small M and N this implies that N = M. PROOF Apply Corollary 1 1 to the continuous isomorphism and its inverse . Thus 0(13) will denote differenl functions at different limes. up to first order. C on sely. we will denote it by 0(13). then 12 . IR). y] + 0(t 3 )}. y] = 0 (Problem 1 3). IR). It follows that a neighborhood of I in O(n) is an n(n . hencc Mt = M. Moreover.. If G is a Lie group and 1 exp(IX) exp(IY) exp lX exp l Y = exp { 1 2 [X. there is a diffeo morphism between them which is also a group isomorphism. So exp Mt = At = exp(M). y] + O(l3)} (3) exp lX exp IY exp(IX) = exp{IY + 1 2 [X.388 Chapter 10 1 3 . and in its proof. In the following Theorem. then = (exp M) ( exp N) when MN = NM. that is. any A E O(n) sufficiently close t o I can be written A = exp M for some M. IR). the x ver (exp M)(exp M)t = I . It is easy to see that (= �l(n .1)/2 dimensional submanifold of GL(n . 14. Y] + 0(1 3 ) \ j X. Since O(n) is a subgroup. IR) exp(Mt) = ( exp M)t . extent (0 which this equation fails to hold. In general. then they are isomorphic as Lie groups.
d d. d d (iii) </> (I) = . du u=o _ I For fixed s. Similarly. Then we have '�. du u=o d YJ(a) = J(a · exp uY). (ii) d XJ(a) = Xa (J) = La.J(exp sX exp l Y ) = dt d = ( YJ) (exp sX exp l Y ) .Lie Graul)j 389 PROOF. u u=o by (ii). </>(1) = </>(0) + </>' (0)1 + </> ) 1 2 + 0(1 3 ). Now Taylor's Theorem says that Suppose that J(e) = (v) J(expsX exp l Y ) = J(exp sX) + I ( YJ)(expsX) _ _ o.X(J) = X(J 0 La) = J(a . exp uX). 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). (i) We have Similarly. 2 .o 12 _ _ + "2 [Y ( YJ)](exp sX) + 0(1 3 ) . let Then l </>(1) = J(exp sX exp l y). 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 ) . for any F.
+ X Y + 2 · 2 2 which gives thus proving (I). (Zt /)](e) + 0(1 3 ) .+ Z2 = . (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 ). comparison o f (vii) and (viii) gives 2. so by (vi) we have (viii) /(exp Z (t)) = J (exp(t Z .390 Chapter 10 Substituting in (v) for F = J. If J(e) = 0. . Equation (2) follows immediately from (I). 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 ) . Z{I) = J(A(t)) + Since w e can take the . + t 2 Z2 + 0(1 3 ).  Z. F = YJ. xx YY . = Z for some Z" Z 2 E Ge. In particular. ] exp tX exp t Y for some gives exp Z (I ) Coo function Z with values in Ge. J)(e) + t 2(22 1)(e) t2 _ _ + 2 [Z. then clearly J(A(t) + 0(1 3 )) 3 O{l ).  X + Y = 2. + t 2 Z2 )) + 0(1 3 ) = t ( 2. = X + Y.f's t o be coordinate functions. Applying Taylor's formula to t Z . 2.
X. with llle relative topologJ') that makes it a Lie subgroup of G. . or PROOF. 1 5.) = (expl.J)(e) = 1 (. . > X and let I. since exp(I. let k. . k.51 ( . 3 [(XX 2  yy + ""2 + 2 + X Y  XX  xx .5. E Ge with X. Then X E lj. + I S. again choose J with J(e) = O. then H is a Lie subgroup of G. Ifwe write 0 ' (1 ) . ' 1. Suppose expl.)) + 0(1 ) 5. Assertion J . E H f all i. # O . then we also have (x) Comparing (ix) and (x) gives the desired result. + 0(1 )). We attempt to reconstruct the Lie algebra of H as f ollows. I Notice that f ormula (2) is a special case of Theorem 516 (compare also with Problems 516 and 518).X. + I S. THEOREI'vl. J)(e) + 1 ' ( . > 0. Let X. > 0 with each I. there is a Coo structure on H. . ' 3 J(exp lX exp l Y exp(IX)) = J ( exp(lS.. More precisely.:. If G is a Lie group and H e G is a closed subset which is also a subgroup (algebraically) . or Proqf We can assume I.)]  (e) ' 3 exp lX exp I Y exp(IX) = exp(lS.X)J](e) + 1 + .Lie Groups 391 To prove (3).1 E H. For I > 0.X. 3 + [ ..51 J)](e) + 0(1 ). Then similar calculations give (ix) J(exp lX exp l Y exp(IX)) = I [(X + Y .).(t) ::: ::: I.YX . Let 1) c Ge be the set of all X E Ge such that exp lX E H f all I . The work involved in proving Theorem 14 is justified by its role in the f ol lowing beautiful theorem. (I) = largest integer � Then �I < I.
= I) Ell Ij'. Y E I). We clearly also have exp lX E H for 1 < 0. . so that G.k. We now claim that I) c Ge is a vector subspace. with . exp(k. sX E I) for all s E IR. X' E I)' X" . > 0 and let X. . . ( exp .X. . Assertion 2. Alternatively. is a dilleomorphism in some neighborhood of O.D. ). then H n exp(U) = exp(I) n U). ) = [ exp (t. . (t) > I. Thus exp IX E H. [Similarly. If X. Y] E 1). for fixed I. Clearly X E I) implie. we can write.] Now let U be an open neighborhood of 0 E Ge on which exp is a diffeomor phism. . but we will not even use this fact. Xn X E 1).X. Proq{ Choose a basis Then t/> is given by of G. using (2) of Theorem 1 4 we see that [X.) k. Xk . .X exp . > G defined by t/>(X + X') = exp X exp X' Xl . since H is closed and exp is continuous. Choose positive I. It dearly suffices to show thaI old if U is small enough. = X + Y + Z(t. so X E I). .(t)I. 12 211 f taking limits as n > 00 gives exp 1 (X + Y) E H. QE.X.392 so Now Olapter 1 0 I. Xk a basis for I). The map t/> : G. . Then exp(1) n U) is a submanif of G. Y] + O( l /n2) l . .(t)I. Then Assertion J implies that X + Y E 1).Y 11 11 1 I )n = exp {I(X + Y) + [X. > IX. so that I) is a subalgebra. t (l) E H. Choose a subspace Il' C Ge complementary to I). 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. .
.D. There is a neighborhood V' of 0 in I)' such that exp X' ¢ H if 0 # X' E V'. . a contradiction. it follows fi'om Assertion 1 that X' E I). Choose a neighborhood = W x W' of Ge on which exp is a diffeomorphism. let a E H n exp( U). X' E W'. Choose an inner product on �' and let K C I)' be the compact set of all X' E IJ' with I ::: IX'I ::: 2. 7 Choosing a subsequence jf necessary.)0 X' E K. If the assertion were false. we can assume Xi' I/n. and 4> of Assertion 2 is a diffeomor phism on W x W'. 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. exp X E H we obtain exp X' E H. and a E exp(1) n U) . since QE.') E H. > 0. . an) is a diffeomorphism of Ge onto jRn . Choose integers JI j 'V'lith Assertion 3. To prove the reverse inclusion. This is clear. . there would be X.' E 1)' with X/ 0 and exp X/ E H. Since exp(I /I1. •:. QE. Prorif. Clearly exp(1) 11' n U) CH n exp(U).X. Since a.X. is a diffeomorphism in a neighborhood of 0 E ]R" .D. . 1 it suffices to show thaI 1+ (a) . so 0 = X'. )(11.L7= a. . ax 1jr* (�' I ) 0 = X. Then a = exp X exp X' X E W. We can now complete the proof of the theorem. .\ 393 Since the map .Lie Grau!}.
A f orm w is called left invariant if La'w = w f all a E G.··· <iJ. La* dw = d(La*w) dw. " " Xn E Ge. .. invariant Iform (J) and left invariant vector fields X and Y we have dw(X. 1 0 U p t o Ilo\V: w e have concentrated on the left invariant vector fields..w([X.o} I /\ . . . . wn(e) is the dual basis to or then any Coo vector field X can be written Xl . . . then f a E G we have or CO:: . Y) = w(e)([X. H ence dw(e)(X. The f ormula on page 215 implies that for a ieI'. a left invariant kform w is determined by its value wee) E Qk(Ge) ' Hence� i f cv l � . fj Xj j=1 Then f or Coo functions /i so 0/ is Coo. This means that or web) = La * [w(ab)]. H ere we consider the case of f orms. Clearly. then every left invariant kform i� i]<. Y) = X(w(l7)) . L ai1 . . iJ.Y(w(X)) . . . . . of Lie groups are better expressed in terms of f orms. If wi (e). = so dw i. Y]). It f oHows that any left invariant f orm is If (J) is left invariant. the bracket being the operation in A. also left invariant. but man� propertie. cvn (e) span G/'. . The interplay between left im"ariant and right invariant vector fields is the subject of Problem I I. (J)n are left invariant I f orms such that cv I (e).394 Olapte. X = 'L. /\ o/k = L Aj o/ J f certain COI/SWI/Is aj. . . 9]) = w([xj ']). .
Now X is the tangent vector at t = 0 of the curve 1 1+ explX. It follows fi·om (*) that [X. we have dw is also left and right invariant. then 1jr*w is right invariant. The converse is similar. then A is abelian (converse of Corollary 7). (4) If G is abelian. So 1jr"X is the tangent vector at 1 = 0 of 1 1+ (exptX).1 = exp( I X). (2) If We E Qk(Gel.! 395 16. Since o/*w and (J) are both left invariant. 1jr* (we) = ( _ J )k We. Let 1jr : G > G be 1jr(a) = a. then dw = o. then so o/*w is right invariant. PROPOSITION. this tangent vector is just x. (3) If W is a left and right invariant kform. then 1jr * we = (3) (I) A f orm W is left invariant if and only if (4) If G is abelian.Lie Group. so The f orm But So dw = o. P ROOF. (I) Clearly so If (J) is left invariant. Y E A. . So dw = 0 for all left invariant Iforms. If W is left and right invariant.(X) = X for X E Ge. So it is enough to show that 1jr. (2) It clearly suffices to prove this for k = J. then all left invariant If orms w are also right invariant. (_J)k we. Yj = 0 for all X..I .
. There are constants Ci� n Since . ' " X E Ge dual to Wi (e). . Con dition (2) is thus an integrability condition. X)] = L Ci�Xk : k=1 dearly we also have . wn(e). . . . [Xi. � j k=l n The numbers Ci) are called the constants of structure of G (with respect to the basis XI . then for have Hence HX. Xl . Y] + 0(1 3)/1 2 = HY. First choose XI . .:. . " " Xn of � ). X)] = " Cik Xk . In fact. Y E Ge we Allernale proqf qf (4). From skewsymmetry of [ . it foHows that for a basis w 1 .L Ci� Wi /\ w) i<j n (2) From (*) on page 394 we obtain = o. . dwk in terms of the 0/ /\ wi .396 Ozapter /0 X. .L Ci� wi /\ wi i. we obtain [X. . By Theorem 1 4.j � . . we can prove (Problem 30) that if Ci� are constants satisf ying (I) and (2). It turns out that (2) is exactly what we obtain fi'om the relation d 2 w k = O. X] + 0(1 3)/1 2 . = dw is left invariant for any left invariant w. ] and the Jacobi identity we obtain (I) Ci) = C)� L( C! C/'k + cJk cL + Cl'iC/') ) = . then we can find evelywherc linearly independent I forms cv 1 . Y] [Y. Letting 1 > 0. . ll=l dwk = . if G is abelian. (J)1I in a neighborhood of 0 E jRn such that . of of invariant Iforms we can express each such that [Xi .
= > cI(iJk . 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. i<j By Proposition 714. 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. which is the set of vectors in (M x G)(p. . the sets 8 1 . . a} . • . . .iii ) . = E > IJ . wIn and constants of structure Ci� ." Ck.. Hence . Let G be a Lie group with a basis of left invariant I forms w'. : r M and r G are each diffeomorphisms in some neighborhood of (p. so on J(p. .wk Choose G and let r be the folium through (p. . .i< Ci� ([iJi /\ iJj ] .[a'.a} where {jk iiJk = 0. a). . . . 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). . Let .iJ n . iiJn are each linearly inde pendent. From this latter fact and (a suitable local version of) Theorem 5 we could immediately deduce the following Theorem. w'. THEOREM. .L Ci� [iJi /\ (iJj _ w j )+ (iJ i _ wi ) /\ wj ]. [i} M "2 : > a. . Let Mn be a differentiable manifold and let e . for which we supply an independent proof. Let '" : M x G M and M x G G be the projections. ... . the existence of such (J/ implies (Problem 29) that we can define a multiplication (a. far 1 7 . en and wi.Lie Group. M x G is foliated by IIdimensional manifolds whose tangent spaces at each point are annihilated by all iJk . _ " > . /\ wj J) L j . . Now iJ'. e n be everywhere linearly independent I forms on M satisf ying k . . wn are linearly independent everyvvhere. . 397 Moreover.
let G be a Lie group. and write this equation out in a coordinate system.l ) w (Y2(1)) ( [ Ly.Ii : M > G be two Coo maps such that . == PEDESTRIAN PROOF.(I)y.Ii differ by a left translation. (�I ) [Ly. j* H tek .(I). .li .Ii * (w) for all lefi im·ariant I forms w. then it becomes an ordinary dificrclltial equation for dY dt = 2 (d. == == == j. (I).iii ) == (q.(H2 ° i)* a/ ek J*w k .j* H2* a/ (H. J(q)) c r. _ == == == It is also po"ible to say by how much any two such maps differ: 1 8 .:. (I)) (�IJ IR and the fwo maps YI ' Y2 : IR > G satisji Y2· For every left invariant Iform w == (d.*(w) La o . )* W (Yz (l )) ] . (I) _ ' J. [ (L y.1 ) YI* w (�IJ * dI · Y2 (of the type considered . 0 j)* e k .398 Let OUipter 10 j: U > M x G be t h e map j( q ) Since (jk . M C�2) == Y2* W == = == It follows thai U we regard Yl a� b riven. We must show that YI W (Y2(1)) Case 1.iiJk = 0 On J. w (y. (I)y.' ] d y. Let M be a connected manifold. and let fi . Then fi and .(I)y. we have 0 = j*( {i . that is. THEOREM. we have YI (O) == Y2(O) . there is a (unique) a E G such that j.
aJd p ) = bJ ( p) for all p E M . is an integrable distribution on G x G. i n particular for t = I . /2 (p))...b} = i=l n ker(Jr) *a/ .(G). : G x G > G be projection on th e ith factor. y. Choose a basis w ' .. w" for the left invariant Iforms. Then Coo curve c: IR > M with dO) = Po and c ( I ) = p. Gelleral case..(O) = y. (t) for ali I . 399 in the Addendum to Chapter 5). Let 7[. it follows fi'om Case 1 that La 0 y. there are a . . Choose a E G so that (La 0 y. Case 3..(G) c G x G is the diagonal subgroup {(a.. b) E G x G. )*(w) = yt ( L:w) = y/(w) = y. For (a. Since M is connected. if t. we have For any p E M there is a Le t y. theu the maximal integral manif olds of t. s o /2(1') = a · .lr2*(J/ ).:* . so it has a unique solution with the initial condition y.. Now define h : M > G x G bv h(p) = Uj (p). it follows that heM) is contained in some left coset of t. .Lie Group. = y. 0 c. = f. M = IR.(O). y. (t) = a . ELEGANTPROOF. In fact.. If (J) is a left invariant If orm. By Case 2. .*(w) = c* J 2(W) = c* . But this solution is clearly y. b E G with . Case 2. (0)./i (p). . By assumption. y. . then Since La 0 y. " Then t.. but the maps y" y. (0) = y..(O) = a · ydO). a) : a E G}. In other words.(G). arc the left cosets of t.. . . Choose a E G so that y. = y. let Ll(a./i *(w) = y/(w) .*(w). Let po E M. are arbitrary'.
th e left invariant nforms are also very important. Hence J G () l(ab) J(ba) = J(a) · J(b). 20. So { G c IR is a compact connected subgroup of IR{O}. . PROPOSITION. For example. But in one case they coincide. Suppose w # O. then J for a unique a E G.ourse. we can define Coo an an 7 Since L Ja '" a" is llsLlal1y kept fixed in any discussion. COROLLARY. = \\'hich proves the formula]. If is a left invarjant nform. If G is a connected Lie group and J: G > G is a map preserving left invariant forms. These generally turn out to be quite diff erent from the left invariant IZforms (see the example in Problem 25). left invariancr of implies that a" in other words. so L Jan = L Lb*uan) = LU o Lb )Lb*a" = LU o Lb )aN• g(a) J(ba). the form Ra*w is left invariant. •:. this is often abbreviated to L J or fG J(a)dCl. and if J : G IR is a Coo function v'lith compact support. ( ) = {I} . ''\'e can. so there is a unique real number with J(a) = Ra'w = l(a)w. of c. = La While left im'ariant Iforms play a fundamental role in the study of G. == where I note that Lb is an orientation presenTing diffeomorphism. If G is compact and connected and w is a left invariant Jlform� then (J) is also right invariant. PROOF. iG /(a) da L /(ba) da. For each a E G. Clearly. then determines an orientation on G. The latter notation has advantages in certain cases. an left invariant nform� arc a constant multiple of any nonzero one. also consider right invariant l1forms.400 Ozapter 10 1 9.
I)) = y(t + u) = y(u + 21). the map le* : Ga 7 Ga.e.)(y (u)) fy(. Since fa = Ra1e Ra . But also . 401 Vife can also consider Riemannian metrics on G. Then y is a geodesic and y(O) fy(ljfe( Y (u)) = = Y(I) . if y > G given by 1a ( b) = ab' a is an isometry is a geodesic and y (O) = a. )) on G by We are finally ready to account for some terminology from Chapter 9. In fact: if ( . i. PROOF. F fixed I.. PROPOSITION. So = fy(.. (I) Since the map 1.e.I (see the proof of Proposi tion 16 (2)). the map fa : G 2 1 . a. let or y (1 + u ). which reverses geodesics through fa (y(t)) = y (I). i. In the case of a compact group G there is always a Riemannian metric on G which is both left and right invariant. Let G be a Lie group with a biinvariant m etric. Since for any a E G.l is also an isometry. then (2) The geodesics y with y(O) = e are precisely the I parameter subgroups of G. the maps I 1+ exp(tX) for some X E A. G.) (Y(u . (I) For any a E G.Lie Graul). Clearly Ie reverses geodesics through e. : Ge > Ge is just multiplication by .. ) is any Riemannian metric we can choose a biinvariant IIf orm an and define a biinvariant (( . so it is an isometry on Ge.) : 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.
. . . . . and geodesics through e are determined by their tangent vectors at / = o. so y is a homomorphism on Q. y is a I parameter subgroup. Xk )v. . then y(t' + I") = y(l)n'+n" = y(t')y(t "). It follows by induction that If t f = n'! and t ff = n"t for integers n' and n". Vd for V. . . .' If VI . . 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 .. x Mp > V. •:. then there are ordinary kforms cv l . Vi . .: i=1 we will write simply ror anv Vvalued kform w we define a Vvalued (k + I)form dw by d dw = L do/ . . (J)d such that f X" . '.. . Vd is a basis for V. . (p)(X" . since there are I parameter subgroups with any tange nt vector at I = 0. . 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 . y(l1I) = y(t)n for any integer n. These are the only geodesics. . . . If V is a ddimensional vector space. Xk E Mp we have or k times d w(p)(X" " " Xk ) = L w. By continuity. .. 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. .402 so Chapter 10 y(t)y(u)y(t) = y(u + 21).
(J)n is a dual basis of lefi invariant I forms. /\ ryj · p(u. • . . we have. • • V x V > W is a bilinear map. If w is a Vvalued kform c w = L {J/ ' Ui i= 1 ry = I >j . .Lie Groups Similarly. We will denote this Wvalued (k + 1) form by p ( w /\ ry). These concepts have a natural place in the study of a Lie group G. respectively. ]: A x A > A. there iI a natural nvalued If orm on G . The form w defined by (*) can clearly be written Then (I) . • . Now suppose that Xl " ' " Xn E Ge = A is a basis. Using the bilinear map [ . • • • . suppose bases UJ . where V and V have . Uc or VJ . Vj ) i=l j=l is a Wvalued (k+I)form. . . . a new Avalued (k + I)form fry /\ A] on G. vf . Vd . a calculation shows that this does not depend on the choice of bases u" • . Although there is no natural way to choose a basis ofleft invariant Iforms on G. and that cvI . for any Avalued kform ry alld any Avalued Iform A on G. Uc and 403 p: V" . . Vd. namely the form w defined by w(a)(X(a)) = X E A. j=l d and ry is a Vvalued Iform then c d L L w..
I dw(U. 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. Yo) = 0 . h) = w(e)([X.. ] in G.0 . V) = . by definition of w. On the other hand. 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.W( V)] · I V and V we have } [X. V = Y for some X. . Y]a) = w(e)([X. Y E G" then dw(a)(Xa.[W(V). 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. k=1 n so (2) Comparing (I) and (2). Y] is left invariant [ . Chapter 10 [X" Xj] = L C. V]).: Xk . 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. For vector fields V and V we can then define dw(U.404 On the other hand. I f we choose V = X. then (Problem 20) we can define a �valued function V(f) on G. Y] = [W(a)(Xa). Y]) = [X.V(w(V)) . 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.w([V.w(a)([X. V) = V(w( V)) . w(V) is a �valued function on G.
It comes about because some books do not use the factor ( k + /)!/ k! I! in the definition of /\ . makes their dOJ equal to � of ours for Iforms OJ. Then the definition of deL OJ.\ 405 WARNING: In some books the equation which we have just deduced appears as dOJ(U. dx . OJ(V)]. The appearance of the factor 4 bere has nothing to do with the 4 in the other f orm of the structure equations. /\ dx . V) = . ) as L dOJ.HOJ(U). .Lie Gmu/i. Tbis makes their A /\ ry equal to 4 of ours for Iforms A and ry.
then so is every coset g H. an for a. (c) Show that I is a linear transformation.f(c) Conclude that G is a Lie group. . or LA since LA is linear. and suppose that (x. 3. Let I : IRn > IRn be distance presen'ing. E U. M(A) = A · M E IRn ' = GL(n.) (b) Find the right invariant vector field with value M at 1 .L Mxkl . We have computed that f or M E �l(n. Show that the tangent bWldle TG ofa Lie group G can always be made into a Lie group. (a) Show that U n +1 is a neighborhood of u n . Let G be a connected topological group. 6. and hence an element of O(n). and Let u n denote all products a l . r for A E O(n) and r a translation. xy) . Let G be a topological group. lRlA = . f LA. . e) is a regular point of . xy 2.) (c) If G is locally compact and connected.406 Chapter 10 PROBLEMS is I. l . (Use Problem 3. (b) Conclude that Un u n = G. with I(O) = o. Coco Let G be a group which is also a Coo manifold. Let G be a topological group and H e G a subgroup.1 n M xkl (A) = " MaI A ka . (a) Show that I takes straight lines to straight lines. and H e G a subgroup. then H is closed.I when I : G x G > G x G is I(x. then G is acompact. (a) If H is open. U a neighborhood of e E G. 5. (d) Show that any element of £(n) can be written A . � a= 1 (a) Show that this means that (It is actually clear a priori that M defined in this way is left invariant. (b) If H is open. . IR) we have where M= . (b) Show that I takes planes to planes. = (x.ak ax k. 4. Show that the closure ii of H is also a subgroup. y) 1+ (a) Find I.y) (b) Show that (e. 7.
[ . 8. Show that Corollary 7 is false if G is not assumed connected. use Theorem 5 to show that exp is a homomorphism on the subspace of Ge spanned by X and Y. . (c) £(GO) is isomorphic to [£(G)]O = AO. b. Show that the set of all ? equivalence classes. 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.Lie Groups 407 (a) For each C E G. (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). (b) AO is a Lie algebra. can be made into a covering space of G. consider pairs (V. Y]0 = [X. I I. In particular. ]. b E V. 1 2. Define ( VI . ljr). If A is a Lie algebra. In Theorem 5.I C U. A= ( 0 a a 0 ) ( = cos a . for all c E G. show that 4> and ljr are equal even if they are defined only on a neighborhood U of e E G.) if ljrl = ljr. (a) GO is a group. ljrd ? ( V" ljr. exp(X + Y) = (exp X)(exp Y). Then (A. provided that U is connected. (a) Use Lemma 513 to show that (exp sX)(exp t Y) = (exp t y)(expsX). ] be the operation on Ge obtained by using right invariant vector fields instead oneft invariant ones.· to a homomorphism of G into H . then 4> can be extended unique!. Y] (� �) = B= O. we define the opposite Lie algebra AO to be the same set with the operation [X. (b) Conclude that if G is simplyconnected. (d) Let [ . with operation [ . defined by a. (e) Use this to give another proof that A is abelian when G is abelian. and where ljr : V > H satisfies ljr(a) · ljr (b)' = 4>(ab . If G is a group. Let X. and hence to �o. ( �) O I ) . on some smaller neighborhood of c. ]) is isomorphic to £( GO). then ljr is an isomorphism from G to Go. ab E U 9.I . and if ljr : G > G is a 1+ a. (b) More generally.sin a sin a cos a 13. Y]. 10. Y E Ge with [X. where V e G is an open neighbor hood of c with V · V. we define the opposite group GO to be the same set with the multiplication . b = b · a.' ) for a. and X 1+ X is an isomorphism of A onto AO.
( Y(o/. Suppose that [X. and let ry. 0) of S' (considered as a subsct of IR'). (p) = X(ry. (e) Usc this fact to give a fancy proof that trace MN = trace NM. Y] = 0. � dry. 1 6..}. and trace are homomorphisms. (b) Using Corollary 51 2.NM) = trace[M. the map det. (Since det. Let . V) be a coordinate system around e E G with x. A more general result holds. (p)). (a) If M is a diagonal matrix with complex entries. it suffices to look at matrices with only one nonzero entry. {ryrl is generated by X + Y. show that for the homomorphism det : GL(n. (a) Let U be a neighborhood of the identity ( 1 .) (g) Now usc this result and Proposition 9 to give a fancy proof of (c). : AI(n.{O}.408 Chapter 10 (a) Show that 14. IR) > IR . 0 <p. N].{Oll = IR is just M 1+ trace M.(p)) + <p.. (a) Let (x. show that In other words. 1 5. Let X and Y be vector fields on a Coo manif old M with corresponding local I parameter families of local diffeomorphisms {<p. {o/s}. witbout using (c). 1 7. 0 0/. there is a neighborhood U of e E G such that every element in U has a unique nth foot in U. (e) Conclude that it holds for all M with complex entries. = <p. show that det exp M = (b) Show that the same equation holds for all diagonalizable M with complex entries. (Look at trace(M N . (b) Show tbat for each n ::: I .) (0 Prove the result in part (d) directly. (c) For G = S' . there are elements a E U which have square roots outside U in addition to their square root in U. Problem 1 3 implies that exp t (X + Y) = (exp tX)(exp t Y ) i f [X. = 0/. IR) > £(IR . (The diagonalizable matrices are dense. etracc M . compare Problem 715.) (d) Using Proposition 9. show tbat there is no neighborbood U which has this property for all n. Show tbat no matter how small U is. Y j = O. ( e ) = O.
then H is a Lie subgroup. t ) = cd t l ) . Ck (tk ) . consider the map b 1+ aba . .l (b). The argument of part (a) is essentially equivalent to the initial part of the deduction of (I). see Yamabe. Show that H C K. show that (b) If Ol. usually Ad(a)(X) is denoted simply by Ad(a)X. By considering or k the map J(t l . . (d) If H e G is a subgroup and an immersed submanifold. The map is denoted by Ad(a). . It is even true that H e G is a Lie subgroup if H is path connected (by not necessarily Coo paths). . Thus we have a homomorphism Ad : G > Aut(�). (a) Show that I) is a subalgebra of Ge• (Use Theorem 14. (c) Let CI ' ' ' ' . .l = = 01 ' (0) + fJ'(O) . On an arcwise connected subgroup ifa lie group. Hint: This f ollows immediately from one of our propositions. Let G be a Lie group. . 19. such that every a E H can be joined to e by a Coo path lying in H. it suffices to know that exptX exp t Y = exp{t(X + y) + O(t)}. (a) Ad(ab) = Ad(a)oAd(b). (Not even the full strength of (I) is needed.'(O)) a basis f �. H is a Lie subgroup of G. Osaka Math. Ck be curves in H with {c.Lie Groups for 409 Coo functions J. Show that (e.) (b) Let K e G be the connected Lie subgroup of G with Lie algebra �. where AUI (A). and let H C G be a subgroup of G (algebraically). The map Ad is called the adjoint representation. 2 (1 950). For a E G. Let � c Ge be the set of tangent vectors to all Coo paths lying in H. .]. isomorphic to G L(n .e) > G are differentiable. LaRa . show that K C H. IR) if A has dimension n).) 18. fJ : (01 ' fJ)'(O) (c) Also deduce this result from Theorem 1 4 (1). .' . is the set of all nonsingular linear transformations of the vector space A onto itself (thus. (b) Show thai exp(Ad(a)X) = a ( exp X )a . and show that the tangent vectors of c lie in the distribution constructed in the proof of Theorem 4. 1314. Hint: Join any a E H to e by a Coo curve c . Thus. the automorphism group of A.
(a) Let J : M > V. (f) Conclude that Ad(exp X) = exp(ad X) = c . . V J . Show that H is a normal subgroup of G if and only if I) = £ ( H) is an ideal of � = £(G). . For Xp E Mp. (It suffices to show this for M in a neighborhood of 0. define Xp (J) E V by d 20.nt space of Aut(�) . . This tangent space is isomorphic to End(A). that is. � (= Ge ) > . (e) Since Ad : G > �. . Y E I).) Use (d) to show that Ad.at the Jdenuty map 10 of � to Jtself. . (O)(Y) = dt dI 1=0 c(Y).+ . 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. Yj. Vl for Jl : M > JR. . Show that this definition is indepen dent of the choice of basis VJ . with basis X(J) = L Xp (Jl) i=] . Y j Y 1+ [X.) (d) Show that Ad(exp tX)Y = Y + t [X. End(�) = n x n matrices. Vl. . . w e have the map Ad*. Yj is denoted by ad X E End(�). Vd. �ange. (X)(y) = [X. .) The map (ad X) ' Ig + ad X + . Vd for V. JR). . JR) show that J . • . where V is a finite dimensional vector space.410 Chapter 1 0 Ad(A)M = A Mr (c) For A E GL(n.JR) and M E � 1 (n . where J = "L1=J Jl . 2! (g) Let G be a connected Lie group and H e G a Lie subgroup. . Yj E � for all X E �. AlIt(�) = GL(n.. LxY. then to regard c'(O) as an element of Aut(A). we let it operate on Y E A by (Compare with the case A = JR". (A proof may also be given using the fact that [X . Y] + 0 (1 ' ) . ifand only if [X.
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
J J for if we change x to x + dl k . Whal does this imply about the maps Ci: Hk(C. ) > 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. ) > Ci u v . �ow suppose that \·ve have an exact sequence of complexes fJ Ci o > C > C2 > C3 > O.I (y) for some y E C/ . I \·vhich re ally 1l1eans a vast c om mutative diagram in \·vhich all rows are exact.I (y)) = which determines the same element of Hk ( C2 ). For example.422 GiW/)le.f* : Hk (M) > Hk(N). then Cik (x) is chanl!ed to Cik (X + dl k . Cik (x) + d/I (Cik . and Ci : Ck (M) > Ck (N) is J* for J : N > M. .I . When Clk = Ck(M). This is nol true . then we have an exac i sequen lT 0 > Hk(C. if V and V arc overlapping portions of S 2 fOJ which there is a deformation retraction of V n V into S l ..I (y)) = Cik (x) + Cik(dl k . C/ = Ck (N).I (y)). then this map is jus.
k+1 is oneone. . > Hk (Cr) > Hk (C2) > Hk(C3) > Hk+l (Cr ) > .J". In this case.I (y') = fJkdl . thus dle)') ". . > 0 is a short exact sequence of so that the following infinitelv long sequence is exact (everywhere): 0 > Since ".I (x') for x' E C.I . diagram (*) should be kept at hand.I Cr') as . this implies that dl k+1 (0) = 0. If 0 > CI > C2 > complexes) then there are linear maps '" fJ C. Moreover.on in the Realm qfAlgebraic but not 70/l0/0t. So we have to show that V. Bv exactness of the middle row of (*) .)' 423 an exact sequence II 1'\ evertheless. THEOREM. In order to prove that ok is welldefined.1 (y') . fJ '" 8 .I (y'). and hence z = o. Then x = tli (x) = O. there is y E C. 0. . . This means that dl (y) = sO we choose d/ . Throughout the proof.EtCU/:r. let x' = fJk . So dl(y) E ker fJk+1 = i lnage". something very nice is true: o II o II IR: 2. 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.k . · E d/ .k+I (z) for some (unique) Z E C k+I . we must check that the result does not depend on the choice of x E cl representing the element of Hk (C3). Ho (Cr) > H0 (C2) > H0 (C3) > H I (Cr) > . I ci with = PROOF.k+ l . Theil o = dl (x) = dl fJk (y) = fJk+1 dl (y) .le obtain 0 E Hk+I(Cl ) if we start with an element of the form di .k with fJk(y) = x.I (x') = dll fJk . Let x fJ '" /. .
. THEOREM (THE I\1AYERVIETORIS SEQ UENCE). Now So fik. this is left to the reader.. onto circles. and the open sets 1) and V illustrated below.� corollarv of Lemma I and Theorem 2.I . HA +1 (M) > . 3. then we have an exact sequence (eventually ending in O's): o� U HO l M ) . together with the first part of the proof that 0 is welldefined. All of Theorem 816 can be derived directly from the followin. . The proof that the sequence is exact consists of 6 similar diagram chases. and a deformation retraction or V Viet oris sequence i� n 11 onto 2 circles. .. � As several of the Problems shO\. the definition of 8 i mmed i ately shows that the image of this element under 0 is precisely the cla" represented by x . the :Maver. 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/.I (y) represents alI element of HkI (C3). ". Let x E q k satisfy dl k (X) = 0. Since there is a deformation retraction of V and r H A (M) � HA (U) Ell HA ( V) > HA (U n )1) 2. •:. Moreover. '''Ie will supply the proof that kerO' C image o. 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.·.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 . As a simple example... ''\'(' consider the torus T = S l X S ' .'h ere V and V are open. If M = U ]I. the cohomology of nearly everything can be computed by a suitable application of the MayerVietoris sequence.
By induction on k. " 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)).a( VI ) > V3 . then 112 7 . it induces a map V'.. Since the map V2 > V3 has kernel a(Yd. so its kernel is I I dimensional.dim V3 + .H ' (U n VI � H' (T) � o. " + ( _ I ) k . . which implie> that VI = O . PROPOSITION.J . 7 Vk l 7 Vk 7 0 0 = dim h . So we have an exact sequence of k .dim V3 + .Excursion in the Realm 'IfAlgelJ1'aic T j O JOlogT 425 � HO (TI � HO (U) Ell HO (VI � HO W n VI � H ' (TI . It follows that dim H I (T) = 2. For k I We have the sequenu Exactness means that {OJ C VI is the kernel of the map VI > 0.J vector spac("� hence which proves the theorem for k. Thus the image of the map H (U) Ell H I (Y) > H I (U n V) is Idimensional. .dim VI + dim V2 .I dim Vk . = PROOF. Aswme the theorem for k . If the sequence o 7 VI + a is exact. . +:+ 0 = dim V2 /a( Vd . this map is oneone. 4. So the kernel of this map is I dimensional. The reasoning used here can fortunately be systematized.dim V2 + dim V3 . and consequencly the map H I (T) > H I (U) Ell H I (V) has a Idimensional image. . Similar rea soning shows that this map also has a I dimensional kernel. = . Moreovel.
( 0 0 1 f S' (afiter SI ' �. .. and is called a kface of /j.. ''''hitney. k . .�+ I L. C M is a dif leomorphic image 0(' some /j.���:oI1S v � A !1\ � � � � � or 2 4 \f mters(. The standard nsim pl ex /j.k of the coordinates xi equal to 0 is homeomorphic to /j.. for a proof see !\1unkres. .'cUom of � ." is defined to be a dillcrent.=l xi = Ji J.. then the image of a kface of /j. 1 t is a difficult I heorem that every ('OC manifold has a triangulation. ." is defined as the set /j. a ne\\' structur� which we will now define. although homeomorphic. 1 . we will use the MaverVietoris sequence to relate the dimensions of H k (M) to an entirely difT�rent set of numbers. . n which cover M and which satisf the following condition: y If a n. J " " . Geometric Integration 1 ..ft ' €SO .X i .vard 3simplcxc .426 CizajJter J J Rather than compute the cohomologv of other manifolds.0.".. m is called a k face of !:J. l>o \ (In Problem 8S . n allj #. Elcmc77la�J' Dijjerf1lial Topology. ..I and .." obtained by setting n . /j. then for some k the intersection ani n an} is a k ·lacr of both alii and an}. standard tflangulaooll b' In al id \Z!J"lria. m .::. �.j The subsel of /j. set. ��_< 0V ��':he .!!cs of /:). If /j. .n = Jx ) E < < jR'"+1 . Now by a triangulation of a compact nmanifold M we mean a [mite coHection 1 allj} of diffeomorphic ima. 0 . arising from a «triangulation" of M. .
any kface of any an.l .I i (3 ) H n'(U II nn Vn. so that Vn_ 1 is a neighborhood of the union of all (n .1 where V n Vn _ 1 has the same cohomology as a di�oint union of Gi copies of sn. will be called a ksimplex of the triangulation .> Hk (U) Ell Hk (Vn_ I i II n II o 11 . n II 0 VnJ l > Hk ( U n Vn. each ani an nsimplex of the triangulation.I i � H n I (M ) � H .} we will call ).Excunioll ill the Realm rljAlgebraic TO/Jolog} 427 Theor Assuming that our manifold M has a triangulation {an. We let ctk be the number of these ksimplexes..I.> Hk (M) .I )simplexes of M. Consider first the case where 11 > 2. 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 .I (U) Ell H I ( Vn_d � � H� I (U n Vn_d n � Hn(M) � H ( u ) Ell Hn ( Vn_ J l II . one within each nsimplex ani. Now let U be the disjoint union of open balls. and let Vn_ 1 be the complement ofthe set consisting of the centers of these balls. Then M = V U 1111.
t l = n.I (M) . For any triangulation of a compact manifold M we have PROOF.I )k dim H k ( V.428 Chapter J J Applying Proposition 4 to these pieces yield. THEOREM.I ) " an · L ( . . defined bv n X (M) = dim Ho(M) . dim H n. joining the balls of the old U. ..dim H I (M) + dim H'(M) .dim H n (M) + an .._ t l = X(M) .] X ( M) = X ( Vnd + ( .l (M) .I )k dim Hk(M ) or 5. dim Hk ( Vn .I we define a new open set U which consists of a disjoint union of sets diff eomorphic to IR".l [dim H .i l = dim Hk ( M) O ::: k � n . 111 the manifold Vn. . one for each (n . = k �O n.dim H (M) + a. " k =o n n + ( _ I )n. We now introduce the Euler characteristic X(M) of M.l ( Vn _ l ) = dim H n.'. .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...I ) "a . � comp ( "eWUonents of (11 = 2) . X ( M) = aO . + (I) n an .I For the case 11 = 2 we easily obtain the same result without splitting up the sequence. + (I)" dim H (M).(.2 L ( .al + a2 . The above equations then imply that X ( Vn .I )face. .
I)" IO'n _1 + ( . we have X(M) = X(Vn!) + (I )"O'n = X(Vn .' . . . . thell V . COROLLARY (DESCARTES. . ' (11 = 3) An argument precisely like that which proves the equation X(M) = X( Vnl l + ( . . while in an other cases we have Combining these equations. the. . joining the centers of the balls in the old U. ..11 the Realm q[Algebraic 7opolog1 429 vVe wil1 let 1111 2 be the complement of arcs� in the new V. • . . ('omplf'ment of � I/n.I )"O'n] = x(Vo) + [(. Vo. each of which is smoothly co Iltractil)le to a point.0'1 + .2 ) + [(. + ( . Hence X( Vol = 0'0. . + ( . and F faces.. /<� . E edges.I )"O'n] n = 0'0 . .I )"O'n also shows thai Similal'I)� we introduce VIl. . .1 ) 1 0'1 + . If a convex polyhedron has vertices..Lxcuniol1 . : V 6. .3. .rt.2 IS . .I ) Q'n .. . . the last of these is a disjoint union of 0'0 . .EULER).E+F=2 . � .
. let (J) b e a kf())'ffi with compact support on M. (V) EIl C. On tho' �'"::. in fact. (U) EIl C. A. ) is the i mage of At E C:(V n V ) . we will denme this extended form by i u'(w). A. each map iu' and i v' is oneone. support w U .. It is clear that image Uu' Ell iv') C ker ( i u' + iv'). C V. The sequence PROOF. suppose .. To prove . @support w LEMMA. . 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.he converse. To prove that iuT +iv' is OntO.) iu' Ell iv' is oneone. 7. This means that AI = i. V}. C V n V. If V c M is open.(V) satisfies iU'(A t ) + iv'(A2) = 0. then w can be extended M by letting it be 0 am side V. </>v} be a partition of unity for 'he cover { V .. thi' shows that support i" C V n V and support A. Since support At C V and support A.." w other hand. if w is a form with compact support C U. So (A.430 ChapteT Jl If we turn from Hk to H. •:. ( V ).</>vw) E C. Then w t t = </>uw + ¢vw E C. I is clear h ac is exact. is clearly 'he image of (</>uw. If C: (M) den otes the vector space of kforms with compact suppOrt on M. we encounter a very different situation. we can define a new sequence.hat O" . and let {</>u .
9. PROOF. If M = long exact sequence . Now suppose A E satisfies a*(A) fJ ) = A 0 (fJ 0 a) = A 0 0 = o.on in the Realm qfAlgelnaic 1O/!% g) 431 8. The details will be left to the reader. .. Ihen sn x IR = (U x IR) U ( V x IR). V open in M. •:. U U V fol' U.'" /. V open in M. (M) This sequence is much harder to work with Ihan the M ayerVietoris sequence.+I (U n V) > . + 8 H. i.. 0 a= O. = O.EtcU1.. COROLLARY. . (V) > H. We just have to show that if the sequence of linear map' )0 )0 W3 is cxaCi at W2. .. suppose we want to find H. . Then A WI � w2 L wJ A there is �: W. then there is a dual . THEOREM (MAYERVIETORIS FOR COMPACT SUPPORTS).e. (U) Ell H. for all sn x IR"'. . . > IR wilh A = fJ *( X). Ifwe write S" = UuV in the usual way. So a* 0 fJ * = O. We claim that = fJ 0 t Given a w E W3 which is . 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.+I (UnV)' > H.. then so is the sequence of dual maps and space� For any A E W3* we have a* fJ* (A) = a*(A W. starting with S l X IRm.. (UnV)' > WI a W2 fJ PROOF.:' I. For example. and then proceed to yet anO! her application of Theorem 2.* W3* � W]* 0 � WJ* . which is diffeomorphic to S.I x lR . . for IR" {OJ. Apply Theorem 2 10 the short exaci sequence of complexes given by the Lemma . If the manifold M = U U V for U. so Ihat unv is diffeomorphic 1 0 S. then there is a long exact sequence .. A IR j . ( M)' > [H:(U)EIlH:(V)]* > H.I x IR. we "vi}} merely record one further resu1t� for later use. > H: (U n V) > H. > H.
Finally. Now chooSt W e W) with W3 = fi( W2) Ell W. Let N C M be a compact sub manifold of M.A(W") = AO'(=) = O." classes into a Vector space f'/ (N). we define a map of complex". the "germs of kfonns in a neighborhood of M". and if j : N > V is the inclusion. vVc appeal first to a result fJ"Om the Addendum to Chapter 9. This sequence is 1lot exact at C! (M): the kernel of i" contains all W E e: (M) which are 0 on N. while the image of e contains all W E e: (M) which are 0 in a neighborhood of N. we will have to use a technical device. There is a compact neighborhood V of N and a map l[ : V > N such that V is a manifoldwith· boundary. Then M N is also a manifold. sO that we obtain a complex g.432 of the form Chapter Jl fi (w'). + j" g. We therefore have the sequencL C:(M . '>\'e now construct l1 sequence of such neighborhoods V = VI :::J V2 :::J V3 :::J . and define X to be 0 on W. j such that Wj E Ck(Vj). vVe now consider a rather different situation. for if fi (W' ) = fi(w"). then l[ 0 j is the identity of 1\'. •:. Moreover.w" = 0'(=) for some z. To circumvent this difficulty. This defines X on fi( W2 ) C W3. so A(W) . while j 0 lf is smoothly homotopic to the identity of V.N) where e is "extension". it is eas\" to define d : g. C: (M) in the obvious way: (J) � the equivalence class o r any w i lli . 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. Now consider two forms Wi E Ck(Vi ). /I' + e C: (M) + i* Ck (N).k (N) > gk+ I (N).k ( N ) . This makes sense. equivalenl if there is f > i. then w . • • with ni Vi = 1\'. we define �(W) = AW)..
suppose A E C/ (M) satisfies i*(A) = O. then there is an exact sequence N c . and then use Lemma I I . +I (M .N has compact support C M . The sequence 433 0 > C. (M . (M) > Hk (N) 10. The cohomology vector spaces are isomorphic to Hk (N) for all k. 1] be a Coo function which is 1 on v. •:. Clearly e is oneone. Let M be a manifoldwithboundary. If W E C. the de Rham cohomology of the manifoldwith boundary V. then W '= 0 in some neighborhood U of N. If compact submanifold of M.N) > .k (N) as 1 ] . Vi = N.k (N) > 0 I I . (M). 1 0. . Let I : M > [0.k (N)) PROOF.k (N) is represented by a form � on some V. Finally. 1 3 . entered only as an intermediary (and we could have replaced the Vi by their interiors). LEMMA. > H. and A = e(AIM . . it is the object of primary interest . this means that A Wi = 0 for some i . ! .N. B y definition o f fJ. THEOREM (THE EXACT SEQ UENCE OF A PAIR). •:. which we will need later. •:.) of the complex {fJ. .N) is exact. THEOREM.N). Conversely.f�). Hence AIM . But in the next theorem. consequently this element is i*(. Details are left to the reader. LEMMA. Then there is an exact sequence + o H. This means that i*e(w) = O. (M . > Hk (N) is an M is a 1 2 . Since N i.N) > H. Then l� E C. + e C. having support I C + interior Vi . with compa�t bound ary aM. This follows easily from the fact that j* : Hk ( Vi) isomorphism for each Vi . PROOF.k (N). any element of fJ. and consequently W = 0 on Vi . (M) + i* fJ. Hk (fJ.ExcU):rion in the Realm ifAlgebraic Topolog). (M . Apply Theorem 2 to the exact sequence of complexes given by Lemma In the proof of this theorem. compact and n. and l� represents the same element of fJ. PROOF.N). there is some i such that Vi C U.
From (*) and (**) we obtain 14. then M is ori entable. M is the boundary of each component. and ffi. I\1oreover. THEOREM. PROOF. {p) X �m). by choosing M to be the closed ball B in �".'(�"+ I ) � H"(M) � H.M ? 2 .n+1 . But we also know (Problem 825) thaI (**) number of components of�"+ 1 . the sequence of the pair (�"+ I . •:. using tubular neighborhoods V.'+I (�"+I . If M c �"+I is a compact hypersurface. M) gives H.M has exactly 2 components. Using Theorem 817. Let M C �"+ I be a compact ndimensional submanifold of �n+ 1 (a com As a simple application of Theorem 1 3 . (S" x �m). U R n H. ( B) "" H k ( B) = 0 for k '" o. For our next application we wiIJ seek bigger game.M = dim H" (M) + 1 .M) � H:+I (�"+I ) � H"+I (M). (�") from a knowledge of Hk (S" . Just like the proof of Theorem 1 2. Then Theo rem 13 may be used to compute the cohomology of S" X sm .I = a(S" x closed ball in �m).434 C hapter J J PROOF. The reader may use Theorem 12 to compute O R O It follows thaI (*) number of components of �"+I . pact "hypersurface" of �"+I). by considering the pair (S" x �m. dim H"(M) + 1 ? 2 . with H.I ) . . of aM in M. we can rederive H.
so M is orientable.M has two components. so every point of M is in the boundal"\' of each of the h¥O components. The proof in Problem 825 shows that every point of M is arbitrarily close to points in different components of jRn+1 . COROLLARY (GENERALIZED [COOl JORDAN CURVE THEOREM). Consider a bounded open set U C jRn which is starshaped with respect to o. then U is diffeomorphic to the open baJJ B of radius 1 in jRn. we conclude that dim Hn (M) = 1. then jRn+1 . The basic idea of the proof is to take Ix E B to p(X)1 .M.I > JR .. then we can prove that U is diff eomorphic to the open ball B of radius 1 in jRn. then H shows that jRn+1 . If M C jRn+1 is a submanifold homeomorphic to sn.Excunillll ill the Realm ifAlgelna. However.. LEMMA. and M is the boundary of each.M has exactly two components. 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. so a modification is necessary. there are a number of technicalities involved. 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. 1 5. which we will have to dispose of first. This produces difficulties at 0. x E U. COROLLARY. .t 7ojJO/ogy 435 Since dim Hn(M) is either 0 or 1. If p is Coo. If the radial function p of a starshaped open set U C jRn is Coo. 1 7. We will call p the radial function of U. 16. •:.
1 8 .1)/(1) is a Coo function with strictly positive derivative. there is an open ball B with x E B C U. / : [0. the radial function p of a starshaped open set V C ffi. At any other point the same conclusion follows from the fact that I r+ 1 + (p(x) .436 eha/ller J J PROOF. There is clearly a neighborhood W of x with the property that for . = Clearly h is a oneone map of B onto U. ° /(J) I . 1L 0 ::: 2':. Choose IX E U with p(x ) PROOF. it might not even be continuous. at 0. so it is Coo.e for all Y E W. 1] > We can assume. and hence in U. Let I < 1. •:. LEMMA.I . I > p(x) . 1 on sn. •:.e .I such that p()') > p(x) .I . I t is the identity i n a neighborhood of 0.11 is "lower semicontinuous": for every E: > 0 there is a neigh borhood W of x in sn. > U by x E sn.1' E W the point IJ' is in B. without loss of generality.1 )/(I)]X.J' E W we have p (y) 2':. At each point x E sn. (he function p need 110t be h(lx) = [I + (p(x) . with a nonzero Jacobian. I < e. This means that for . that p [0.I . . In general. 1] be a Coo function with = Define h : B / ° in a neighborhood of ° /' 2':. I Coo. However� the discontinuities of p can be of a certain form only. Since U is open.
We claim that there is a Coo function p : snI > IR such that p < p and This will prove the Lemma. . and we will be content with proving the following. cover S. . •:.1lx. By Theorem 817. lxl is < p(x). .. say WX1 " ' " Wx/ for convenience. . . For each x E sn. and hence in U.1' for all Y E W. Similarly. . IVx. . . it follows that p(x) < p(x). .11 the Realm ifAlgebmic ToJ)olog). and consequently w = d� where � has compact support contained in V. The prooffor Hk is clear. . (U) "" H. then Hk (U) "" H k (IR") and H. . .I . 437 Even when p is discontinuous. Since K is closed and p is lower semi K C V = {IX : x E SnI and 0 :s I < p(x)}.I such that Ix may also be used as 1. LEMMA. (IRn) for all k. p. since U is smoothly contractible to a point. + 1>1 (x) = 1. PROOF. . . (U ) "" IR "" H. it looks as if U should be diffeomorphic to IR".. continuous. Let IVx" . Proving this turns out to be quite a f eat. (x) are o. let 'PI . 19. Let w be a closed kform with compact support K C U. . Each Ix" . be a partition of unity subordinate to this cover. Then PI+dx). Since 1>dx) + . for then V is diffeomorphic to IRn .Oll . we just have to show that H. . 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. .EXCU1J. (lRn). We also know that H.I . (U) = 0 for O :S k < Il. and define Any point X E snI is in a ceI"tain sllbcollection of the J. . there is a neighborhood Wx of x in sn. If U is an open starshaped set in IRn. K C V. 1>.
n· . } be a finite cover by such open sets. U · · · U V. every point has a neighborhood V which is geodesically convex. . Suppose i t i s true for a certain 1'. In general.438 Chapter II We can applv this last Lemma in the following way. . is nonempty. V" V) of M. and consider a nice cover {V" . we first use the MayerVielDris sequence f 1R2 = M U V. 3. } as a subset of 1R2. It is fairlv clear that if we consider /II = { I . and . . then expp establishes a diffeomorphism of V with an open starshaped set in Mp . . 2. a manifold M will be called of finite type if there is a finite cover {V" . . this shows Ihat H' (M) is infinite dimensional (see Problem 7 for more informalion abOUI the cohomology of M). If any V = Vi. then V is clearly geodesically convex. 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. 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. Then the theorem is true for V = V. then M = 1R2 .:. V. If M has finite type. . . n Vi. . On the other hand. . . .} such that each nonempty intersection has the same Hk and H: as IRn. . To prove this rigorously. 20. Let {V" . 3. . H' (M) (fJ H' ( V) II 0 � H ' (M n V) � H2(1R2). We obtain H' (1R2) II 0 � . .. V. /II is not of finite type. Let M be a compaci manifold. If p E V. Ii i s r = 1. PROPOSITION. . . Ii follows from Lemma 19 that V has the same Hk and H: as IRn. and choose a Riemannian metric for M. II 0 where M n V has the same H' as a disjoint union of infinitely many copies of S ' . According to Problem 932. 2. (M) are finite dimensional for all k . where V is a disjoint union of balls or around 1 . then Hk (M) and H. .
p. (M) is similar. (a v f3). so P D(a) (f3) = v f3 E H. the "Poincare dual" of a. ] . (M) with It is convenient to also use /J. /J. Now every a E Hk (M) determines an element of the dual space H. . V n V. So Hk (M) must be finite dimensional. the main part of the proof is called a Lemma.(M) represented by any � E C. (M) + /J. > Hk. As with most big theorems of algebraic topology. to denote both this element of H. but we shall restrict the theorem to manif olds of finite type. •:. since this has the nice cover {V n V. and the theorem itself is a simple corollar\'. Now consider the MayerVietoris sequence .l (V n V) Ii H k (M) + The map a maps Hk (M) onto a finite dimensional vector space. It is also true for V n V. One of the fundamental theorems of manifold theory states that P is always D an isomorphism. 1R. with orientation IJ. There is then a unique element of H..k (M)* by PD(a) . . or For any manif old + a Hk (V) (fJ Hk ( V) > ' " . The proof f H.ExcuIJion in the Realm ifAlgebraic 7ojJolog} 439 for U.) � = J. by the same f ormula. . We are all set up to prove this fact. M we can define (see Problem 831) the cup product map Hk (M) x HI (M) "::' Hk+/ (M) r+ by ( [a!]. a We denote this element of that we have a map H. . Now suppose that Mn is connected and oriented.k (M)* by f3 r+ 1 (M. (M) and the isomorphism H�'(M) > IR which takes this element to I E 1R. and the kernel of a is also finite dimensional. in order not to plague ourselves with additional technical details. [�]) We can also define [a! I\ �]. . since a! 1\ � has compact support if � does. . . .
•:. If M = U U for open sets U and and is also an isomorphism for all k on M. and is left original Lemma . = = 1>dw) w E VI . o."2(Y) ." y E 1>2(Y)V2 f3d:) x = E WI . 1>5 Supposc f some or Then (x) so is an isomorphism.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. 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. 1>3 the reader. in which the top row is the MayerVietoris sequence. V. Consider the following commutative diagram of vec tor spaces and linear maps. 1>2. So is oneonc. Then is also an isomorphism. Vve now forget all about our manifold and use a purely algebraic resull. Suppose that the rows are exact. and that are isomorphisms. LEMMA. PD V PD � � Ell � � n By as�umption. 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. for all k on U. and U n V.+I(U V)+ H. then V PROOF.+I(Vn. ffi n � ffi n 1>1 .(U) H.k. By exactness at there i. x E V3. Thc proof that is onto is similar. 1>4. This proves the . except possibly the middle one. Hence "THE FIVE LEMMA".(M)· [H. 1>4 0 = 1>3(X) = 1>3"2(Y) = f321>2 V3.<U v)· Ell � � Let I = 11 . 1>3 x= "2(Y) = "2 "dw)) = ( (Q o. aU vertical maps.H. 1>4 3 = "3 (X) = 0. 1>3(X) = 0 f331>3 (x) = 0. : for some Then which implies that y = "dw). and the bottom row is the dual of the Mayer Vietoris sequence for compact supports.+I(U) H. Consider the following diagram. n � is an isomorphism 2 1 . are isomor phisms.
. . PROOF Use the Theorem and Proposition 19. In fact. COROLLARY. since E is locally a product.ExeulJion in Ilze Realm ifAlgebraic Topolog). and consider a nice cover {V" . . U · · · U Vr . Even though the Poincare Duality Theorem holds for manifolds which are not of finite type. V. The theorem is true for V. 23. Ell v for the (n + k )manif old E. V} of M. Problem 7 shows that H' (IR2 . •:. •:. THEOREM (THE POINCARE DUALITY THEOREM). 25. it is true for M. If M is a connected oriented nmanifold of finite type. noting that V' is isomorphic to V if V is finite dimensional . If M is a compact connected orientable nmanifold. . COROLLARY. = ( _ 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". 441 22. and for V n V (as in the proof of Proposition 19). •: .k (M) have the same dimension. The theorem is clearly true for r = 1 . then X(M) = O. Orientations /J. Let V = V. Suppose it is true for a certain r. then a slight modification of the proof for Lemma 1 9 . This completes the induction step. If M is a compact orientable odddimensional manifold. f M and v f S give an or or orientation /J. PROOF In the expression for X(M). If { V" . We begin by considering a smooth kdimensional orientabie vector bundle S = 1f : E > M over M. If M is a connected oriented nmanif old of finite type. then H k (M) and H. . then the map PD : Hk (M) > H. Corollary 23 does not. the terms (l)k dim Hk(M) and ( _ I )"k dim Hnk (M) cancel in pairs. By the Lemma. Vr . Vr } is a nice cover of M by geodesically convex sets so small that each bundle S I Vi is trivial./II) and H� (IR2 . then Hk (M) and Hnk (M) have the same dimension. COROLLARY. 24.k ( M) ' is an isomorphism f all k . or PROOF B y induction on the number r o f open sets in a nice cover o f M ./II) have different (infinite) dimensions. .
. Let Fp = 1f' (p) be the fibre of $ over any point p E M. Notice also that for the maps s = Osection E. /L) be a compact connected oriented manifold. . 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. and $ = 1f : E > M an oriented kplane bundle over M with orientation v. . so that J( M. . On the other hand. and let � E Cn(M) be a form representing /L.). This class U is called the Thorn class of $.' (Ur)} is a nice cover of finite type. Our first goal will be to find a simpler property to characterize U. p. 26.vp) jp'w = l . and hence an element vp E H�(Fp). the orientation v for $ determines an orientation vp for Fp. Also choose A so that there is an equivalence (Fp.442 Chapter J J shows that {1f ' (U. Our definition of U states that ( I) Let A C M be an open set which i s diffeomorphic to IRn. THEOREM. . 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. Let (M. (This condition means that f where U is the class of the closed form w. ) � = 1. there is an element jp' U E H� (Fp). so that A i s smoothly contractible to any point p E A . L 1f'� l\ w = 1.) PROOF. and let jp : Fp > E be the inclusion map. Since jp is proper. Pick some closed form w E c� (E) representing U. 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. 1f .
the form H ' w o n (A x Fp) x [0. 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 . it is easy to see that there is a smooth homotopy H : (A x Fp) x [0. under the identification of 1f . support w Using the fact that A is smoothly contractible to p. Let II II be a norm on Fp. By choosing a smaller A if necessary. or which we will continue to denote by jp. the map jp : Fp > 1f1 (A) corresponds to the map e r+ (p.I (A) with A x Fp. 1) = (p. I) : lIell < K}. For the H constructed in this way it follows that m e. 1] > A x Fp such that H(e.e) : q E A . e. we can assume that there is some K > 0 such that. Under this identifica y tion. I) 'f. support w if lie II ? K . lleII < K ) . 1f2 (e)) = }i> (1f2 (e)).EtC/miDlI in lize Realm ifAl gebraic ToJ)ology 443 This equivalence allows us to identif 1f1 (A) with A x Fp. 1] has support contained i n {(q. Consequently. wejust pull the fibres along the smooth homotopy which makes A contractible to e. O) = e me. the support of wl1f1 (A) is contained in {(q. e) f e E Fp.
*(H*w) .  I ) dimen . Now. Theo form IH*w on A x F has support contained in {(q.444 Chapter II lIell < K}. on the one hand we have (Problem 817) (4) On the other hand. Combining (3). e) : lIell < K J . (4). " learly suffices to prove that the integral is 0 over A' x Fp for any closed ball A' C A. (5) we See tha. Since we have where Jr* /L 1\ A has compact support on A' x Fp by (2) = = ± 0 . Thm (2 ) So support A c {(q. To prove this. JaA1XF/. e) p rem 714 shows that (jp Jr2 ) 'W . [ 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).io*(H*w) ° = = d(lH*w) + I(dH*w) d(lH*w). i. we claim that the last integral in (3) is O.w = i.
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.
(Mi). xcwsioll ill lize Realm ifA lgebraic Tapolog). . V u. 455 8.) 9. except for the square Hk.l l .I (u n V) H�+I (U n V)' lPD > Hk (M) > H� (M)' lPD which commutes up to the sign ( . the only slightly difficult maps are the ones involved in the above diagram. Vo is u nion of shaded VI is union of un shaded C�·. (M) "" EEli H. and U2 . (d) The figure below shows a decomposition of a triangulated 2manifold into three open sets Vo. "3 . "2 . Show that the squares in the diagram in the proof of Lemma 21 commute. (Mi) and all but finitely many "i = 0 E H. .E. "3. " ) with "i E H. VI. this "direct sum" consisting of all sequences (" I . (a) Let M = MI U M2 U M3 U · · be a disjoint union of oriented IImanif olds. (b) Show that Hk (M) "" n i Hk(Mi). Show that H. (It will be necessary to recall how various maps are defined. "2. is union ofshaded & . . (c) Show that if the Poincare duality theorem holds for each Mi. then it holds for M. ( Mi). ) with "i E Hk (Mi). . this "direct product" consisting of ali sequences ("1 . Use an analogous decomposition in n dimen sions to prove that Poincare duality holds for any triangulated manifold. which is a good exercise.
(a) Let P I .) 1 I . enclose this ray in a cone. Using Problem 826. Let � = .. (a) If U E H. then there is a nowhere 0 vector field on M.here zero vector field on M. : E > M and r = . Conclude that U represents 0 E Hk (E). show that the index of . Join these by a ray going to infinity..X at p is ( _ I )n times the index of X at p.f*(X(n) = xW). ' : E' > M be oriented kplane bundles. thaI X(�) = 0 when � = . so that X(�) = O. prove the Thorn Isomorphism Theorem: The map HI (E) > H:+k (E) given by " r+ " V U is an isomorphism for all I. then there is a vector field X on M with only one singu larity. . over a compact oriented manifold M. (Begin with a triangulation to obtain a vector field with a discrete set of zeros. Using this. If a vector field X has an isolated singulaI"ity at p E Mn . . X(j*(�))·) (e) If M is a connected manifoldwithboundary. Using Poincare duality. (E) and U' E H. aM '" 0. f) a bundle map from �' to I: which is an isomorphism on each fibre. show that if X(M) = 0. then there is a . with nov\. E M.456 Chapter JJ 10. It follows. 1 3. (a) Let � = . 0» If M is compact.. we have f*(x(�)) = 12. (d) If M is connected and not compact. then there is a nowhere 0 vector field on M. such that all Pi E interior D. (c) it is a fact that a Coo map f : sn. providing anotllCl proof that X(M) = 0 in this case. p. : E > M be an oriented kplane bundle over an oriented manifold M. Q») Since we can also consider U as being in Hk (E). . (E') are the Thorn classes. and push everything off to infinity. and (j. show that there is a subsel D C M difTeomorphic to the closed ball. we can form U v U E H. .. (b) . show that this is 0 for k odd. in particular.I > snI of degree 0 is smoothly ho motopic to a constant map. This p rovides another proof that X(M) = 0 for odd 11. (Using the notation of Problem 323.k (E). with Thorn class U. Using anticommutativity of /\. : TM > M fOl' M of odd dimension. then j* ( U) = U' .
(b) Show that Rh is an isomorphism on a smoothly contractible manifold (Lem mas 17. This means that we obtain the correct singular cohomology of M if we consider the complex Hom(S'k (M). (Hillt: Stokes' Theorem. (d) Conclude that Rh is an isomorphism if M is of finite type. and 19 will not be necessary for this. Basic knowledge of singular cohomology is required. . and U n V. to show that if Rh is an isomorphism for U. For a manifold M. 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.EXCU1JiOl1 in tlze Realm ifA( �ebmic TopologJ' 457 1 4. IR) be Rh(w)(c) = Show that Rh i s a chain map from { Ck (M)} t o {Hom(S'k ( M). This Problem proves de Rham's Theorem. 1R). 1 8. IR)}. 1 w. while i 0 r is chain homotopic to the identity of Sd M) [basically.) (e) Check that the cup product defined using 1\ corresponds to the cup product defined in singular cohomology. V. then it is an isomorphism for U U V. using the MayerVietoris sequence for singular cohomology. r is approximation by a Coo chain]. (Using the method of Problem 9. let Rh(w) E Hom(S'k ( M). 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.) (c) Imitate the proof of Theorem 2 1. We will denote the group of singular kchains of X by SdX). (a) If w is a closed kform on M. we let S'k( M) denote the Coo singular kchains.
.
iust like neighborhoods of o. we will not consider them. . 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. (c) M is metrizable. (I) U of x and an integer n Condition (I) is necessary. (b) Each component of M is second countable (has a countable base for the topology). (By the way.) 459 U n JR is open. so every component is open. JR. :Moreover. but local compactness is. (2) For each x E M there is a neighborhood such that U is homeomorphic to lR:". We bave just seen that the Hausdorff property is not a "local property". Although nonHausdorff manifold. (d) M is paracompact. so every manifold is regular. we will now define a manifold to be a topological space M such that (I) M is Hausdorff. Before exhibiting nonmetrizable manifolds. so every manifold is locally compact. and thus a manifold itself. * E U. this argument does not work f "infinite dimensional" manifolds. which are lo or cally like Banach spaces. that almost all "nice': properties of a � anifold are equivalent. a Haus dorfflocally compact space is regular. (In particular. I t consists of JR U {*} where * 'f. The following properties are equivalent for any manifold M: (a) Each component of M is a compact. we first note . for there is even a I dimensional "manifold" which is not Hausdorff.APPENDIX A CHAPTER I Following the suggestions in this chapter. Ev ery manifold is also clearly locally connected. are important in certain cases. THEOREM. these need not be regular even if they are Hausdorff) On the other hand. a compact manifold is metrizable. then (U n JR) U {OJ is a neighborhood of 0 (in JR). with the following topology: A set U is open if and only if (2) If :>: 0 Thus the neighborhoods of * look . there are manifolds which are not normal (Pt·oblem 6).
/. M. The cOllSl ruct10n of J11Cse eXalnples requires the ordinal numbers. U . then x must be in the closure of a finite union of these Vi. U . U V'I I imersects only Un l+1 • . Then VI U C2 has an open nl'i�hborhood U:!. II is easy to show iiom th1s that M is paracompact.s countable union of C0111pact set�. . and hence =} SECOND PROOF (0) (a) =} (b) =} (c) and (d) =} (a) as before. : and so On. is (a) is Theorem 12. which are briefly explained here. The serond proof does not rei v on this difficult theorem. is clearly open. ' U U". . (b) =} (c) follows (rom the Urvsohn metrization theorem. U C. = Uo U . . A connected.. U · . paracompact space is a compact. . 160). PlOOf. Let M = C. (Ord1nal numbers will nol be needed for a 2d1111ensional ex ample 10 come later. Since the space 1S connected. (d) =} (a) is a consequence of the following. Clearly C. U V". .ith compact closure. then Vo can intersect onlya finite number lJ1 • . 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 . . for if x is in the closure. It lurWi out thaI there are even l manifolds "dlich are not paracompact. LEMMA. . \l\'1th compact closure_ Continu1ng in th1s way. . is compact. . U Un .460 Appendix A FIRST PROOF. (cJ =} (d) because anv metric space is paracompact (Kelley. Similarly Do U VI U . or .) oflhe others. an open neighborhood VI "·. pg. ' . It is also dosed. . If Vo is one ofthe�e. l. The unioll Va U " . · .'hos('" union contains all Cj. This pro\'es the Lemma and the Theorem. . we obta1n open sets Vi with Vi compact and Vi C Vi+ l . Un':!. b . (a I =} (b) follows immediateh' fi·om the simple proposition that a acompact locally second countable space is second countable.. it equals thi. c E A (tl"ansitivit\·. U [in . U . .. where each C. locally compact. ha.:. . because x has a neighborhood \vhich intersects only finitely man}: Thus x is 1n the union. . Gellem/ 7ojJo/agT. U . There is a locall\' finite COver of the space by open sets with compact clo sure. (a) =} (d). \".
2)\ 0 < 1 <2<3< . o< 1 < ' " < w 1. I ) ) CO. ) (w + I is. An orderinf! < On A lS a wellordering jf every nonelnpty subset B C A has a. just a set distinct from those already mentionedl 0 < 1 < 2 < . < w .. an element b such that b ::5 b' for all b' E B. < w < . one and onlv One of the following holds: (trichotomy). .. Two ordered sets (A. 2. . O < l < '2 < · · · < w 0 < 1 <2 < .··<w·2 <w·2+ 1 < . for the prese]]!. = (O.· < · '] < . < w . <) 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. 3 . . . 0<1 <2< 0 < 1 < 2 < · · · < w < w + l <w+2 < ·· · < w · 2 < w · 2+ 1 < ··· < w O< 1 <2 < ··· < w < . 'J < . .I' 0. <) where < is an ordering on A . .· <w·2 « ··<w·3 < . b 461 E A.. that is. O < 1 < 'J < · · .. . . and II is easily seen to be an order isomorphism also. <) and (B. 1 . . 3 < . > An ordered set is just a pair (A . 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.Appendi.J (2) For all a.firs/ clement. < w < w + l < w + 2 < · · · < w · ] < w <w+1 <w+2 < .. 1 . < .· <w<w+l <w+2< . (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+1 (w is some set .. .. .
of course.. also a wenordered set with tbt same ordering.c?ment of the succeeding ones. then for the firs! clement b of B (and hence of A) we clearly mus! haw 2 .. . which makes the study of wellordered set' simply delightful.. It is easy to see that if B is " seglncm of A . . If (A. If f : B B' C A is all order isomorphism and B' is a segment of A . <w and 0 < I < . If B '" A is a segment of A.c. . etc. the "wth H .)! The way we prove this ligorouslv is amazingly simple: If feb) '" b for some b E B.B. the second with the second. just consider the first element of {b E B : feb) '" b): an outright contradiction appears almost immediately. Proposition I has a com pall ion. B = {a' E A : a' < a J : o < I < . < ) are wellordered sets. We match the first element of A with the first of B. Notice that each set on OUf list 1S <1 .en feb') must be b'. a i s the first element of A . where b' is the second element. But there lS a much more general propOSltlOll which will settle all cases at once: J. In particular..rith the "w th". It 1S not hard to sec that no two sets on our lis1 are order isomorphic.srI. order isomorphic to a segment of the other. PROPOSI TION. . PROPOSITION. until we run out of one . To do Ihls rigorously.. consider order isomorphisms from segments of A onto segments of B... then B is not order iSOlnol· phl( to A. etc. For example.. In fact. And so Oll ... •:. PROOF. the only (wder isomorphism f)'om B to a segmeni of A is 1111 ldentit:: > f(h) = b.462 AfJPclldix A Any �ub�et of a wellordered set is. ('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. It is easy to show that any tWo such order isomorphisms agrC'(.hen' PROOF. <w <w+ I arc nOl order isomorphic because the second has both a last and a next to la�l eien1cnt. <) and (B. TI. while the first does not. . then one i. on the smaller of their two domains (just cOllsider the smaJJesl element \. then either B = A or else there 1S some a E A such that In fact.
So all such order isomorphisms can be put together to give another. < is a wellordering of the ordinal numbers. To obviate this difficulty V'!e' need only '''lark with order isomorphism classes of .A. ordi nal !lumbers. <) be a well ordered set representing one of its elements ex. <r! must be replaced by "(A . <) < (B. .. . If " is an ordinal numbe. <)". ''Almost'': because the condition H(A.Appendix A 463 thev don't). To produce a smallest element of . < ) ! Roughly speaking: An ordinal number is order isomorphic to the set of an ordinals less than 1t.rl from each ordrr l:. Given a nonemptv set . •:. These older lsomorphism classes arc called ordinal numbers.:. we will denote by .. while those of the form " + I are called Successor * Only one feature mars the beauty of the ordinal numbers as presented here. <). it would be much !llcer to choose one specific well o)'dered :.+ I the smallest ordinal after " (if " is represented by the wellordered set ( A . <) order isomorphic to (B. due to von Neu mann . Transitivity of < is obvious. PROPOSITION. There is a panic. <) = (B. I\otice that if ex is an o)dinal number. Each ordinal number ls a horribly large set.'ellordered set with just one more element.oll1orphism class. <). Suppose wc define a relation < between wellordered sets by stipulating thaI (A .. Notice that some ordinals are not of the form " + I for any ". instead of '''lith the wellordered sets themselves.A of ordinal numbers.. <) when (A . larger than all member!' of A). General Topolog). It determines a segmen t wh1ch represents somt' fJ E . represented by a wellordered set (A . and Propositions I and 2 show that we al most have trichotomy.gnore elements 2:. They are beautiful:' 3.ularly elegant way to do this. <). Consider the least of these a's. 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.A \'\'C can obviously i. then " + I is represented by a \". This fJ is the smallest element of . 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 ..'enordered sets. which is c1earlv the largest of all. If it is defined on all of A we are done.A. these are called limit ordinals. <) is order isomorphic to a proper segment 01 (B. c¥. and dcfine thesr speclfic sets to be thr PROOF. let ( A . then the wellordered set of all ordinals fJ < " has a particularly simpk representation: it is order isomorphic to the set (A . v·:hkh can be f ound in (he Appendix to Kelley.
and henct' is order isomorphic to an initial s<'glllt'nt of (). il would represent a countable ordinal " E SI.e.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. 2. <) is a wellordered set. and consequentlv has onlv countably many pre decessors. etc.) Although the countable ordinals exhibit ul1countably many degrees of COIll pkxit�: they are each simple in one way: rihe PROOF By Proposition 3. . 3 . . . w' . PROPOSITION. A leap of faith is required. a countable wellordered set). After one comes to and this 1S only the beginnln�� A ll the wellordered Sf. i. . any member of SI is countable. this ". one can see how the symbol> w3. see Kelley's Appendix. . By the remark after Proposition 3. would appear (svmbols like w3 + w2 .. w . We have thus established the existence of an uncountable ordinal. .I resohllioll of this paradox. w + 1. There are indeed an enormous nUlnber of countable v·. Our sv example. (h is hopeless to tl'Y 10 "reach" n by C'o 'ltinuing the llsting of well ordered �ets begun above. With a little thought.464 Appendix A ordinals. . to a proper segment o f itsdL contradicting* Prop osition J . We will also denote some ordinals by the symbols appearing before: 0.tI. . fO I· 011e would have to go uncountably far. re presented by n. For ."ts llWlll loned so fal" are countable. and ell counter �rts with an ullcountable nmnber of degree� of complexity. . 4. OUl· list of wellordered sets onlv begins to suggest the complexity which well ordered Sets can achieve.wuld mean that n is order i�omorphi(' to the C'oJknion of ordinals < Q . . Let SI bc the collection of aJ! countable ordinals (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. 1 . so it represents an ordinal ex E (htl. If it were countable. ( SI . is clearly the first uncountable ordinal. Then SI is uncountable.
s ) < (f3 . for then f3 would be represented by a subset of �.{(O. . The union of all these sets would then represent ex. PROPOSITION. Then f3.o)  The set n x [0.0) (w. no subset of � is order isomorphic to n . < a cofinal in a .g.O) (w+2. . < f3) < . The Corollarv to Proposition 5 implies easilv' that the long ray and the long line are I dimensional manifolds. This can be pictured as follows: (0.0) . If a subset of � were order isomorphic to n. . and we can easily arrange that the subset representing f3i is a segment of the subset representing f3j for i < j. 1 ) . So by Proposition 5. Suppose there were one.\'0 ) ) is called the closed long ray (with "origin" (0. •:. YJ . ) tw·2. 0) and a could bv represented by a subset of � . The disjoint union of two copies of the closed long ray with their origins identified is the long line L. If a E n. This is impossible . If a E n is a limit ordinal.0) . thus also by a subset of (00. and hence a smallest. a E n not represented bv some subset of � .O) (W+I. O)} is the (open) long ray. < a. and L + = n x [0. aside f)'om the line and the circ1e� there are 110 other connected Imanifolds. namely those bel\'I'eCll the points representln. Howevel. .Q and a + 1 for all a E n. such that every f3 < a satisfies f3 < f3n for some 11 (we sa�' that (f3nl is "cofinal" in a). Since a is countable. 6.. is represented by a subset of ( 00. a contradiction. PROOF. then there is a sequence f3 t < f3.0) )' (2. 0)). It cannot happen that a = f3 + 1. PROOF. 11S1 \vhich comes after fill ' (. < f32 < f33 < . COROLLARY. i). Consider n x [0. then there would be uncoul1\ablv mallY disjoint intervals in � . . 1) . there is a sequence f3 . 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. The first example of a nonmtU'izable manifold is defined in terms of n. . with the order < defined a s follows: (a. To disllnguish L + and L. (1. ). then a is represented bv some wellordered subset of �. . the names "halflong linel ' alld "long line" may also be used. I ) if a < f3 or if a = f3 and s < I .AJ)pelldix A 465 5. 1) with the order topology (a subbase consists of sets of the form {s : X < xo l and (x : x > . ).
0). The manifold itself is. + But consider now. Y ) o ) = o« (xy. Consider the disjoint union of �� and ��. which might be called the "big disc". with P E (��) and f (p) E JR � o + identified. y) E �2 : r > OJ and another copy �2 x {OJ of the plane. the following diagram shows two open sets homeomorphic to �2. homeomorphic �F='. Define a map /0 : (��) > �� b." '" + q�. We begin with the open upper halfplane �� = {(x. and denote the point (x. There is another way of producing a nonmcuizable 2manifold which doe� not use n at all. y). 0) by (x. in fact. Identifying all points « 0." � /_.\' .. 8 ) in the product of the closed long ray and S' produces another 2manifold. y. L x IR (long strip). + to JR2 . L x L+ (big halfplane).) > �� b�' + j� «X.466 Appendix A Quite a few new 2manifolds can now be constructed: L+ X S ' L + x IR LxL L+ x L+ (halflong cvlinder). say �2 x {a}. (halflong strip.:. Deline j� : (�. which we will denote by �. we could have thro\vn away �� to begin with slnce it is ldcntified by a homeomorphism with (��) .)')a) = (a + y. for each a E �. we will denote this set by �6.' f . (big plancj. L x S ' (long cylinder).. .:. y). This is a Hausdorff manifold. (big quadrant).\". y)o. another copy of �2.
(d) Show that on any infinite set th"'e is a wellordering which represents a limit ordinal. a E IR we wish to identify each p E (��) + with f a(P) E ��.\. developed in the problems..ApjJendix A 467 In the disjoint union of �� and all ��. any sequence has a convergent subsequence (but L + i� not compact!). and c(A ) E A for all A. . (Zorn's Lemma may be deduced from this fact fairly easily. O)a ] . then X U Y 1S equivalent to 1'.) (c) Given two sets. and related man]folds� have some very strange properties. (e) From (d). This manifold. Let c be a "choice function". PROBLEMS (a) A wenordered set cannot contain a decreasing infinite sequence X l > X3 > . 3. namely the set { (O. show that one of them is equivalent to (can be put in oneone correspondence with) a subset of the other. one is an extension of the oth. the Priifer manifold. (Thus one can define even and odd ordinals. (b) ]( we denote (a + 1 ) + 1 by a + 2.l'+b. for it has an uncountable discrete subset.\3 :s n polnt. Given a set X. (a + 2) + 1 by a + 3. but it cannot be second countable. then {x } converges to some (b) If Xl � X2 :s: . then any a equals f3 + II for a unique limit ordinal f3 and integer II � O. and in the disjoint union of all �� identify each (x.) L X2 > 2. c(A ) is defined for each set A '" 0... a wellordering < on a subset Y of X will be called "distinguished" if for all Y E Y. Consequently. are a space homeomorphic to ��� so we will consider �� a subset of the resulting space. i. (a) L + and L are not metrizable. is a sequence in L +. show that if X and Y are disjoint equivalent infinite sets.: (b) Show that there is a wellordering on X. The equivalence c1asses� of course.y+a = x'. We mav dispense with �� completely. (a) Show that of any two distingulshed wellorderings.{y' E Y : / < y}). y = e(Y . This space is still a Hausdorff manifold. and Problem I. etc.Y)a and (X'.y')b for which y = y' > 0 and .e.
(b) HI(L +) Ht {L) O . 00)) is countablt.cations of L x L. Show that U include. L + x L +. and r > s. defin. one can explicitly construct these Slone. (e) Of the 2manifolds constructed from L + or L with HI = 0 and one para compact end. and the big disc arc all distinct. some A n 1S not nowhere denst . &). then n n both sequences converge to the same point. Hint: Given H : L + x [0. (b) L + x III and L x � are not homeomorphic. L + x L. only L + x � has the homotopy type of L +. Sillee IR: = Q U Un A . (a) L + is not contractible. L x �.t ech compactificatlon�. (Use (c)). L + x L" . show that for every I we have ( H (x. X Sl ) = z.O) for every irrational a.468 Appendix A (c) If {X } and {Y } are sequences in L + with Xn S Yn :5 Xn +! for all 11. 4. " paracompact ends'·. (f) If f : L + > � is continuous. I] L + with H (x. Hint: Imitating Problem 119. L x L. 0) x for all x. (d) The StoneCech compactif. then f is eventually constant. 6. L x L +. any order topology is normal (completely different proof!. (Using Problem 3(g). + (and also L) are normal. (c) HI(L + x S') = = Ht {L = = > = 5. (a) Show that the Proler manifold P is Hausdorff (b) P does not have a countable dense subset.s]) and .' ([r. Similarly for L + x �..I)} L +. (c) Let U be an open set in �� \vhich 1S the union of "wedges" centered a t (a. n . Hint: Let A n = {a : the wedge centered at a has width � I /n } .r. L+ x S l and L x S' are nOl homeomorphic. b) x (0. a whole rectangle of the form J (a. then one of the sets f' « oo. (d) L (e) More generally. (a) L + and L are not homeomorphic. (gJ If f : L + > � is continuous.
(a) If A C M has cardinality c (the cardinality of �). (b) If C e M is closed and has cardinality c. whose boundary is a disjoint union of uncountably many copies of JR . but which has a countable dense subset. s ) = Show that H is welldefined and that H(p. (e) Define H : P x [0. � )a J'\ I 1 + sy l . (c) Let p E M. = 469 C2 = { (0. { (O. Conclude that P is contractible.{ (x . 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 ) . .s + sy a ) if y if y > 0 : � o. O)a a rational ) . then the closure A ha. with con·esponding points o n tht boundary identified. (f) P . 1 ] > P by H « x . 1) E �i U {(O.Appendix A (d) Let C" C2 C P be C. ( g) The dis joint union o f two copies o f P' . 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 . There is a function f: n > (set of subsets of M ) such that f(rx) has cardinality c for all rx E n. Alternatively.O)aJ for all p E P. cardinality r. then C has an open neighborhood with cardinality (. but that thev are not contained in dis open sets. Describe a homeomorphism of this manifold onto �2 8. Let M be a connected Hausdorff manifold which is not a point. is a manifold which is not metrizable. 7. Its fundamental group is uncountable. O)a a irrational) : : joint Show that C. 1( x 1 s + sy �' y  (x �. and such thai f lO) = { p ) (Consider functions defined on inidal segments of n vvith these same propertle�! and apply Zorn's Lemma. Y)a. y)a y < 0) is a manifoldwithboundary P'. and C2 are closed.
(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. is homeomorphic to either the real line. Then there is a unique f with the required properties. the long line. (a) A connected I manifold whose topology is the order topology for some .) 9. then M is homeomorphic to S' . (Given p' E M. of UP<a f(f3) with cardinality C. (e) M has cardinality r. consider an arc from p to p'. 1 ] ) with the properties of tbe func tion in part (c) is eventually constant. or the halflong line.) (d) A function f : n > (set of subsets of [0.470 Appendix A order. This is an example of defining a function by "transfinite induction" .
are all CW. and I(xo) = "0 . there is no CW map q > p with a CW inverse. Reasoning as before. we must have f(q) > q. etc. but this implies that f(x) = x for all x.L +p. but it . since it is CW). since we can consider fI in the cOlltrary case. I do not know whether every 2manifold has a Coo struCtUlT. the problem is that a complex analytic structure on �2 mav bc conformally equivalent to the disc. the method used fOI" obtaining a CW structUl'e on L + will not yield a complex ana!wic Jlruc/ure on L + x L +. \".L +p. A CW stmcture exists on the Prufer manifold. This implies that it is also diffeomorphic to (0. has a limit point Xo E L + . This sequence has a limit in L + . How ever. . then it yields a CW structure for L + . Using the CW structure on L + . and even a CW structure. points in ��. Continuing i n this wa)� we obtain Xo < Xl < X2 < . In fact. If (!) is a CW structure for L +. So for some ql > Xo we have f(ql ) '" ql. and consequently that the structure on M can be extended if M is a proper subset of L +. and hence extendable. f(/(q)). this follows immediately from the fact that the maps f used for identifying points in various (��) + with a. we can assume f(q J ) > q l . .L +p. with f(xn } = Xn . 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. If p E L +. and hence for L +. f(/(/(q ))) > f(/(q)). . . Now f cannot be the identity on all points > Xo (for then it would be the identity ever)'\'Vhere. 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. 1 ). we obtain Xl > Xo with f(x} ) = X l . f(q).Appendix A 471 CHAPTER 2 The long ray L + can be given a Coo structure. then L + .e can get a CW structure on L + x L +. in other words. a contradiction.L+p is clearly homeomor phic to L +. It is knOvvll that there are uncountably many inequivalent CW structures on L + . These are all distinct. The increasing sequence q. and L+p denotes all points S p. and then it is easy to see P I that we must also have f(/(q)) > f(q). .
Complex AnaD'lic Manifolds wilhoul Counlable Base. . 4 (1953). 335340. (a) Every twO points X. it is a classical theorem of Rado that every Riemannian surface (2manifold with a complex analytic structure) is second countable. a mod ification of the Priifer manifold yields a nonmetrizable manifold of complex dimension 2. Soc. Sic. and locall" compact.L+q . try to deter mine v·. p) be a metric space and let f : X > Y be a continuous locall" oneone map. Let (Y. Ann. Arner. (b) Let d(x. Math.L +p > 1 2. Prove that lor I I . 10.hich can be immersed in which. pp.iihlbarkeil. y) be the greatest lower bound of the diameters of f(C) (in the pmetric) lor all compact connected C containing x and y. I 25115 (1958). E X are contained In a compact connected C C X. AnaiJ'lische Slrucklur und A b<. References to these matters are to be found 111 Calabi and Rosenlicht. Kneser. pp. Of the Yalious manifolds mentioned in the previous section. Acad. On the other hand. where X is H ausdorfl connected. In fact. and not extendable. H. PROBLEMS L+ . locally connected. ).472 Appendix A may also be equivalent to the complex plane. Proc. 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 + . Fennicae Se ries A. 18.
Z )a (x . The remarkable fact about this 2dimensional manifold N is that it is connected. y. points in �� being denoted by (x. y . y.z) E �3 : y '" OJ.al a with y < O. The equivalence classes form a 3dimensional H ausdorff manifold M. Beispiel eiller dimensio'nserhohenden anab'li. c .2a) E A with y < O. and for each a E � let �� be a copy of �J. Archiv. Math. a E IR we identify (X. Kneser. This example is due to M. y.. .Appendix A 473 CHAPTER 6 Problem A6(g) describes a nonparacompact 2manifold in which two open halfplanes are a dense set. z) a for y > O for y < 0 with with (a + yx. y. we see that all leaves of the foliation are the same as the ieaf containing { (x. y. II (1960). In the di�oint union of A and all ��. c . Let A = {(x. z .ala E �� with Y > O. O) : y < O} C A . z + a ) (a + yx. c . We will now describe a 3dimensional version with a twist. and these are identified with the set of points (x.· chell A bbildung "wischen iiberiib"ahlbaren Mannigfaltigkeilen. y. y .a). the set of points (x. o) a . and the sets z = constant form a foliation of M by a 2dimensional manifold N. Since we can choose a = c/2. pp. c) E A with y > 0 is identified with the set of points (x. manifold there is an obvious function ". y . 280281. Now the folium containing { (x. y. For. c . y. On thi.z".ala) contains the points (x. y.
contradicting the fact that the tangem bundle is not trivial. . Thus. and hence paracompact. if f: N M lS an immerslon. Although the results in the Addendum to Chapter 9 can be extended to closed. 2 7 2. Compare pg. not necessarily compact. Theorem 79 fails for L. More generall\'. the tangent bundle TM cannot be trivial. Problem AI I implies that a manifold N immersed in a paracompact mani fold M is paracompact. 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.474 CH APTERS 7. then TM lS definitely not equivalent to T* M. they cannot be extended to non· paracompact manifolds. then N has the Riemannian metric f* ( . Appendix A 9. ). for the folium is a submanifold. O)a) of the Prufer manifold. We have seen that any paracompact Coo manifold has a Riemannian metric. 3. (On the other hand. as can be seen by considering the Odimensional sub· manifold {(O. a basic result about bundles says that a bundle over a paracompact contractible space is trivial. Thus the tangent bundle of the long line is not trivial. but a much easier proof is now available: Let ( . 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. The tangent bundle of the long line L is clearly orientable. since an equivalence would determine a Riemannian metric. 5. also metrizable. a locally compact connected topological group is acompacl (Problem 104. 10 I. nor is the tangent bundle of the Pnjfer manifold. Y. even though the Prufer manifold is contractible.. So there are at least two inequivalent nontrivial bundles over M. The converSe also holds: slnce a Riemannlan metric determines an ordinary metric. Notice also that if M is nonparacompact. A Lie group is automatically paracompact. a Riemannlan metric on M. submanifolds. since its tangent bundle is trivial. 2.) 4. Since there is no Riemannian metric on a nonparacompact manifold M. ) b.
Ajijlclldix A 475 7. TIlis will b. Is there a nowhere zerO 2form on the various nonparacompact 2mani folds which have been described? . 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). PROBLEM 13. Addendum 1).
.
75 65 101 T'M T®S 109 1 10 1 21 107. �) SO(n ) (x.NOTATION I NDEX CHAPTER ] CHAPTER 2 M p2n pn sn aM JRn 8(X) lHln 23 19 4 S' 11 19 1 6 19 . }. v p a . 1 1 9 1 13 131 109 1 16 1 16 1 20 121 121 1 22 I ii .. 61 34 34 28 34 35 61 6] v" v { f) Iv) . � Ell 1. VnJ IA � Ix. . p I'm F 65 72 83 65.· ' CHAPTER 3 E"(X) J. . 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. i. �. 1 1 6. 'U) SL(n. ax . U ) aJ ar I ax.i) Till" T' Xp J . �) O (n) R(n) (�n . .'(V) Tt' (� ) 7ik ( V ) I' J p' End(V) df d. W ) Tk < v ) Tk(i' ) T V) d T. (p).· 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. x �2 � Mp (M.
478 N olation Inde> T (I.�:��/ . 8fi. CHAPTER q" 7 202 205 238 238 2 19. X I. 234 202 Ali curl X All I dx ' /\ . 234 23.) / ')" r" . 10. Vk ) IA I (/(1 2 ) exp A Lx A Lx f Lx ) " Lx w [X. 2 1 5. x) (01.I/. 1 29 1 34 1 34 1 08 1 09 1 0. II P II Lx w S.. 235 2 10... .J w <PI a· /\ a· CHAPTER ("" Q(M) Q k (V) QO(l ') (v..c) f' ' . II.wI 7i� ( V ) T°(V) 231 231 231 231 20 1 227 206 202 227 215 20 1 201 203 v.h. w(X) Iv t' �j l .n OIx(l) = 01 (1 . " . 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. 2 1 3. 1 0.il. (. /\ dxn l de den d8ta .f Hk (M ) H� ( M ) I' deg f div A" de t(f. . Vk ) (v" .i l . g ) mj '[La) wa d f l(LI) i1·w Iw dw M. 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.«) ('R. signp J w(p) Zk ( M ) . . . 224 215 22. 1'·* V' 1 23 10. .wl (V) n..
(M) 6n a' 0' [wJ al' di ( 268 285 29 1 264. ) (g. I Jo CHAPTER A ( ( ( ( ( ( ( « I II II rti . }/ . 259. I II II. 356 356 314 314 31 I 309 309 324 334 302 306 32G 312 344 sinb tanh x'( v) xi r .J ) exp d. �) g[(".. 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. . 291 264 264 263 248.N olalion il1dc..' // / J dx + g dy /w Lw # (' (q) u [O. ). 245./J L� Mp" exp(A) GL(II. 257. HI ii(u) oj 0. )r . 248. )' .) . �) . } . 285 285 260 246 239 243. 404 373 410 385 385 372 407 407 376 CHAPTER d(p. 246.1 IAI A i) Ad( ) ad X Aul( g) a Euc(V) Euc(�) E(y ) J' ( .u E(II) Elld(nl exp O C O B C' 'J [ij. ).!..\ 479 Z. q) ds d l' ("osh ("osh . 288 294 246 299 299 266 9 322. )' .
] [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" [ .I [ f(o)da " . APPENDIX A k=d L W' U! to Q Old <> + 1 464 463 464 464 461 46 1 463 .480 i.1 R" SO(n) X . L. 0(13 ) . 410 403 376 403 400 400 400 403 CHAPTER I I L. £(G) o(n) P Ck (M) 9!(N) PD RI. 'II p(W A ry) [ .Ix 6" [' M # !I' p. N otation index 401 374 379 376 388 376 41 1 41 1 374 373 376 379 380 395 403.: <jJ. u .
27�: point. 20J Alternation. 228 Annulus. generalized. L. 36 i halfspace. Elie. 31 2 function. 7 2 . 202 Analytic manifold. 1 03 Ado. A S" 1 7 9 Big disc. 248. 283 Choice function. J 45 Base space. 7 ) Basi� dual. 28 maximal. 248. 108 fibre. 309 induced.. I CJ Boy's Surface. 46() Calabi. 28 Choice. 40J Boundary. 46� Circle. 7 i trivial. 466 Biinvariant metric.INDEX Abelian Lie algebra. 464 Banach space. 6 Closed form. 38 Change) infinitely small. 4!=l . Fundamentru Theorem of. 106 quadrant. 3 9 . 382. 20i multilinear function. 3 1 3. 380 Alexander's Horned Sphere. ] 20 of covariant tensors. 1 0 1 map. 208 or Belongs to a distribution. �). 200 Cayley numbers) 100 Chain. 465 manifold. 3 3 2 Arcwise connected. 23(l CauchyRiemann equations. 7 J BuraliForti Paradox. 34 Annihilator. 19J Besicovitch. 379 Bundle cotangent. 7 1 normal. 376. 45 I sphere bundle. F. 466 plane. £" 472 Calculus of variations. 293 AJgebraic inequaJities. 252 geodesic. 19. ] ] I tangerlt. J. L n. 35. 46b half plane. 20 subgroup of a Lie group. 285 Chain Rule. 246 Associated disc bundle. 391 submanifold. l Oll Adjoint T* of a linear transf ormation T. 20S f g. 378 in o(n. 1 07 for Mp'. 285.. �). 55 AJgebra. 348. I I Arclength. 3 1 6 Cartan. 344 of contra\'ariant tensors. 409 Area. 252 Bounded manifold. 7 3 nplane. 1 9 subgroup of a Lie group.k(p). 2� Auslander. 1 9 long ray. 451 Atlas. principle of irrelevance of. 1 54 in gl(n. 1 1 1 Chart. 395 Adams. R Antipodal map. 360 Cartan's Lemma. 2 1 0 vector. 218. J O� dual. 233 Alternatinc ("ovaria� t tensor field. 60 Bracket.
341 Complex. 83 Derived set. 4 19 d e Rham. 29 manifo]dwithboundary. 42 [ Commutative diagram. 299. 23 J scalar. 39b Continuous homomorphism. up to first order. 82 Coorelated. 295 Densin· eve. 27. 60 refinement of. invariant. 3 1. 3ib Complete. 34 C� distribution. J O �l Covariant functor. 30 on the long line. 372 Critical poim. 279 Degenerate. 34 CO manifold. 238 Cvlinder. 421 analytic structure. 225. 40 Cross section. 263 group of M with real coefficient:. 8 C. . 220. 40 i ll the calculus of \·ariations.482 Closed (contiuued. 209 odd scalar. 28 Darbom: integrable. 29 structure. 283 integral. 38i Contractible. 214 Deformation retraction. 130 tensor field. manifold. 284 Debauch of indices. 32 at a point. 1 20 vector field. 465 Cohomology. 14 Crossproduct. 227 Crosscap. 263 of a complex. 30 oneparameter group of. 228 Definition. 28. 3 1 manifold. 123 Decomposable. 4 i ) numhers of norm 1 . ] 30 tensor field. 65. 25 DescartesEuler Theorem. 1 39 Contravariant functor. 429 Coordinate lines. 439 Curl. 133. 148 DiAerentiable. 32 Riemannian metric. 32 manifold. geodesic'. 39. 133 Derh<ation. ]7 9 form. 259 relative scalar. scalar. 308 structure on TM. 429 Determinant. 139. 286 Degree. 39. 283 Darboux's Theorem. 232 DiAeomorphk. 28 Cotangent bundle. 47 1 on pn. 24b index Cup product. I I . 320 Critical value. 1 1 3 COWl locally finite. 1 33. 420 Commutative Lie algebra. I 1 3 Convex geodesicaU)" 363 polyhedron. 30 Dilleomorphism. ] 2] . 207 function.. 15� Coordinate system. 299 Cube. 22i Lemma. 236 Contraction. 15H Coordinates. 3i3 Conjugate. Uy.10 pointfinite. 32 . 5(1 Cramer's Rule. J 6{1 Cofinal. 28. 275 mod 2. 78 of a ring. singular. 358 Constants of structure. vector field.
355 . 1 92. 464 Fibre. 68. 194 Folium.. 422 of a pair. 89 Fundamental Theorem of Algebra. associated. 134. 468 Endomorphism. 1 03 E. 461 First variation. 445 Euler's Equation. 320 483 Einstein summation conv(. 179. 2 1 0 equation. 10. 72 weak. 384 Extension. 400 Irelated. 2 113 sequence. 438 First element. 207 differential. 3 Du Bois Revmond's Lemma.ntion. 1 8 1 ideal of. Five Lemma. 16S forms. 1 69 Equations of structure. 419. 201 of a function. 30 Differential. 23 paracompact.. 96 Euclidean metric. 358 Equations depending on parameter�. 133. 39 Elements of norm 1 . on �n . 308 Elliptical nonEuclidean geometry. J 2 1 Energy. 190 Frobenius Integrability Theorem. 3 1 S motion.. 139 Foliation. Exponential of matrices. 201 left invariant. 467 relative scalar. 64. 23 1 relative tensor.Inde. 254 Functor. 440 Fixed point. 374 right invariant. 7 1 Finitt> characteristic. 433 of vector bundles. 285. I Faith. 305. 324 Envelope. 254 . 4 Direct sum. 32. 169 linear. 1 0 7 vector bundle. 1 30 Functorites. 429 characteristic. 36i Embedding. 352 Domain. 20 Distribution. 238 Theorem. 215 Fubini's theorem. 205 type. 194 Force field. 233 Disjoint union. 240 Form. 404 Equivalence (or vector bundles). 2 1 5 o n torus. 3 19. 164 depending on parameters. space. 374 nspace. 1 8ll Divergence. Dual basis. 4� End. 109 Dimension. 32(1 Even ordinal. 428 class. leap of. l OB Euler. 421 Disc bundJe. 334. 1 36. 385 .' Differentiable (cOTlliTlurd (structure con1iTllled. 432 Extremal. 4. 209 Exact form. ponential map. 23 1 scalar density. 293 Fundamental Theorem of Calculus. 45 1 Discriminant. 29 on s n .
3 7 J Guillemin. 39. V W. 227. 1 9 Handle. 69 HalfIonp cylinder. 349 Index of vector field on a manifold. 40i orthogonaJ. 349 Infinite volume. 104. 341 Geodesicallv convex. 14 Immersed submanif old. 380 Homotopic. 363 ' Geodesy. 350 Independent infiniteBimals. principle of irrelevance of algebraic. 61. 47 Immersion. H Hardy. 337 General linear group. 356 sine. 46 Implicit function theorem.. 283 . 432 Global theory of integral manifold�. I I I Infinitely small displacements. 44i on �n . 459 Homogeneous. 304 Gmk. 46S strip. 372 usual. 314 Index of inner product. 23i GramSclllnidt orthonormalization process. 342 Inde> Hyperboli( cosine. 148 Infinitesimals. 36. 101 orientation. 304. 7 Homomorphism continuous. 277 Hopf. 342. 3 1 4 Initial conditions. reversing map. 3 i 2 opposite. 145 Hair. 192 Integrable functioll Darboux. 466 line. 44 6 Indite:debauch of. H. Robert A. 3 1 2 Infinitely small cbange. independent. Sylvester's Law of. 123 raising and lowering. 246 Geodesic. 260 Inequalities. 333 closed. 351 Induced bundle. 3 1 4 Infinitesimal generator. 3 33 Germs of kforms.. 410 Identification. 466 Halfspace. 30l preserving. 60 Indefinite metric. G. 356 Ideal of a Lie algebra. 106 HahnBanach theorem. 450 Hopf Ril1owde Rham Theorem. 46 lOpologi �al.. 84 Group Lie. 2 3 Heiukin. 189 Integrable distribution. 37� Generalized area. 136 of integral ClJrve. 49 topological. 1 7 9 Ha s one end. 3 7 1 matrix. 2 1 Integrability conditions. 301 Inside. 401 Geodesically complete. 356 tangent.484 Gauss's Lemma. 23:. 372 topological. l 36 Initial segment. 14. 283 Riemann. 387 of Lie algebras. 104. I nertia. 462 Inner product. 194 Gradient. 84 Hausdorff. H. 1 0 Imbedding. 277 Homotopy.
194 of a differential equation. :M. 243 Linear differential equations. 382. 294 Jacobi identity. 409 closed subgroup of. 374 Left translation. 232 definition. 165 svstems of. 7 1 Local theory o f integral manifoIds� 190 Kelley.]. 34 9 Leaf. J44 Lie group. 43S 485 vector field.. 179 triviality. 243 manif old. 230 Laplacian. 58 Lav�' of ] nertia. 435 Kneser. 460. 414 second. 464 Left iJwariant form. 378 Jacobian matrix. 472 Kneser.. 243. 239 In\'ariance of Domain. 2 J 4 Irrelevance of algebraic inegualitie�. 410 topologically isomorphic. 179 Lives at points. natural. 60 Line integral. 21. 239. Invariant. 380 ideal of. 394 n form. 1 1 9 Lobachevskian nonEuclidean geome try.. 376. 104 positive semidefinite.. 340 Isomorphic Lie groups. 3 76 abelian. 1 7 J Lin� ar transformation ac\joint of. 194 Leap of f aith.. 233 Isometry. 400 . l O P Isotopic. E. 1 8 1 maximal. 226. locally: 382 lsomorphism. 463. 407 Lie derivative. 1 36 Riemann. 128. . 179. 245 Integration. 395 commutative.ket in any ring. 410 opposite. 373 Lie's fundamental theorem first. 103 contraction of. 18. 3 7 1 arcwise connected subgroup of. 388 Lie $ubgroup. 415 normal su bgloup of. 148 spanned locally. 155.inde. 305. 4 1 6 Limit ordinal. 40 J ordan Curve Theorem. 366 Klein bottle. 283 line. 376 for the blac. H. 104 Linking number. 473 Lang.. 138 Littlewood. 239. 374 Length. 150 Lie group. 3 1 2 o f a curve) 59 Lie algebra. ]. 368 Local flow. 145 Laplace's expansion. 415 third. Integral curve. 296 Lipschitz condition. 376 homomorphism of. principle of. 4 1 5 oneparameter group of local diffeo morphism. S. 464 Kink. 136.. Sylvester's. 136 Darboux. 391 local. 283 surface. 121 positive definite. 463 ' set.
E.]. 2�1 Manifold. disjoint union of. 3 4 ' atlas for. I�l . W. 303 preserving. 194 ndimensional. 46] open. 1 9 4 nOllmctrizable. Ht) smooth. 283 index MayerVietoris Sequence. 9 nmanifold. I . 29. 4 6G orielltation of. Coo. Ion Multiindex. 4 J Mesh.. 424 for compact supports. 460 closed. 46" ray. 401 Euclidean. ](l generalized. 73 rank of. S. 3 I\'latrix grou ps. 345 Newman.m. 465. 2�1 diflerentiable. 2 1 4 Manifold. 108 Neighborhood. 8(1 Norm. 1 3 path wise connected. :)7�' Maximal intcgral manifold.. M . l OCi Magic. 1�1 C' 34 Co: 34 Coo.<. 34. 20 Long cylinder. V. 19 Nondegenerate. H . 3 Nice cover. 2 8 boundary o C 1 �l ' bounded. 4 . 3 1 2 spaces. R. 43 J :Measure zero. 40. J 9 closed. 4 nplane bundle. 465. left invariant. tubular. 2�1 dimension of.. 350 Riemannian. 400 nholed torus. 308. 42 1 bundle. 86 manifold. 3 1 1 usual. 46:} Lower sum. 466 N onorientablc bundle. J 15 Munkres.486 Locall) compact. 42 Mod 2 degree. 30 I NonEuclidean geometr>" elliptical. 20 connected. 52 integral. 344 . 208 Multilinear function. 4 nforms.].vitllbnundarv. 4 imbedding in �N. 1 39 oneone. J06 MacKenzie.'. 403 Natural isomorphi. 367 Lobachevskian.. i Natural gvalued Iform. 1 8 1 maximal. 36H Nonmctrizable manifold. 20 finite cover. 466 line. 3 1 5 indefinite. 3� Map betweell complexes. R. 50 isomorphic Lie groups. 382 Lipschitz. 7 1 nsphere. 179. 4 5 9 analvtic. A . 305. 37:2 Normal bundle. 239 Metric biinvariant.1 Mobius strip. 438 Nonbounded. 4 ( 1 Massey. 304. 20 Milnor. i ntorus.
36 integration in. 32 Projective plane. 439 Poincare Duality Theorem. 210. 20 Piecewise smooth. special. 463 Orientablc:bundle. 39 Pointfinite cover. 465 Ordered set. 98. 225 Poincare upper half plane. S. I I. 61. 348 Orthonormalization process. 86 manif old. 351 Odd ordinal. 467 relative tensor. 60 submanifold. J06 Positive definite. 102 tensor. 104. 30. 275 Prufer manif old� 467 Pseudometric. 441 PoincareHopf Theorem. 148 of local diff eomorphisms. 450 Poincare Lemma. 260 Outward unit normal. 248 Orthogonal group. 225 Paracompact.index 487 Normal (cantinued) space. 17!=l Onedimensional sphere: 6 Oneparameter group of diff eomorphisms. isomorphic. 88. 266 Polarization. 284 Palais. 351 Nowhere zero section. 367 Point inward. 60 Polar coordinates. A. 308 Positive semidefinite. 407 Orde. 461 isomorphism. l l6 Projection. 461 topology. R. 85 of a manif old.. 209 Outside.. 98 outward. Opposite group. 86 of a vector space. 410 outward unit. 260 Pointderivation. local. 372 Orthonormal. 3 1 2 Pig. 60. 88 Proper map. 100. 52 Path wise connected. 304 Pollack. GramSchmidt. 435 space. 84. 7. 407 Lie algebra. 304. of a bundle. 460 Ordinal numbers. ] 34. 21 Outward pointing. 239. 465 map. 3 1 3 Partial derivatives. H. 104 Product of vector bundles. 248 reversing. yellow. 95 . 459 subgroup of a Lie group. 84. 105. 148 Oneparameter subgroup. 85. 459 end. 434 Poincare. 19. 88. 468 Parameter curves. 35 Partition. 461 Ordering. 384 Opm long ray. 288 scalar density. 84 preserving. 86 Oriel1latim.. 259 Onedimensional distribution. 245 of unity. 450 Poincare dual. 301 Positive element of norm I . 304 Osgood's Theorem. 133. 167 Parameterized by arclength .
. 194 Slice maps. J7. 283 integral. 472 Rank of a f orm.488 Quaternions. 231 Scalar density. 277 isotopic. 424 f compact supports . 7 Sphele bundle. 134. 342 de Rham cohomology v{'ctor spaces. 134. 22� of a map. 287 Lie group. 59 Rdlnement of a cover. 285 Skewsymmetric. 246 simplex. 308. 28:. 419. 98 Rectifiable. surface of. 104 Separate points and dosed sets. relative. 26 Roscnlichl. 288 Reparameterizatiol1. 7 3 zero. 303. 103 MayerVietoris. 382 Singular cube. 100 of norm 1. 342 Roman surf ac{'. 290 Space filling curve. 3iB Slice. 422 of vector bundles. 321 de Rham. 263 with compact supports. 220 homotopic. 264 def ormation. 462 Selfadjoint linear transf ormation. 9 6 Segment. 231. 427 singular.'. 40 Related vector fields. 40. 17'1 Special linear group. 20J. H... 40 space. 244. M. 133 Schwarz. 8.. initial. 285 Simply. 50 Regular point. 4 7'2 Rotation group. 435 Rado. 362 Second countable. 277 manif old. 294 Solid angle. 6(J Shuffle permutation.. 283 Riemannian metric. associated. 6 J Special orthogonal group . 134. 279 Revolution. 29 piecewise. 43J or of a pair. 263. 104 Semidefinite. 28 homotopy. 45!=l Section of a vector bundle.. 3 1 2 Smoothly contractible. 3 1 2 Right invariant nform. 1 90 Relative scalar. T. 42. 459 value. 374 Rinow. 2 3 1 tensor.connected. 268 de Rham's Theorem. \fl. 3i 3 Sard's Theorem. 294 Scalar. 354 Schwarz inequality. 62 . 433 Shrinking Lemma. 3 J J usual. G. 457 RicmaJ1Jl integrable. positive. 58 Spanned locally. 5 I Shrinking Lemma. 400 Right translatioll. 22i Simplex of a triangulation. 451 RadiaJ f unction. sum. 248 Retraction. 54 Smooth. 95 Sequence exact. 62 Sphere.
H. 464 Sum of vector bundles. 49 Coo. 1 4 immersion.Inrie. 1 2 1 . 253. 106 '. 3 7 1 imbedding. 384 Submanif old. Standard nsimplex. 120 covariant. 373 oneparameter. 98. 301 System of linear diff erential equations. 396 Subalgebra of a Lie algebra. 98 of a manif old. 246 Starshaped. 25 Transitivity. 14. 3 Wedge product. 116 Thorn class. 198 Subcover. 120 covariant. 239 of revolution. 63. 7 1 Totally disconnected. 203 Whitney. Whitney. 171 acompact. 9 Total space. 7. 7. 123 contravariant. 42. 42 7 Trichotomy. 8 nholed. 4�) immersed.. 426 simpJex of. 261. 388 Torus. 76 of �". 8 Tangent bundle. 134. 17. 106 Stokes' Theorem. 4 . 374 Triangle inequ ality.. 134. 345 Twoholed torus. 456 Topological group. 77 Tangent space of �Ii.\'hitney sum. 4 7 open. :2 Su ccessor ordinal. 8. 26 Sternberg. 122 Tensor product. 101 . 260 to a curve. 426 singular cube. 101 Support. 33.. 46 Topologically isomorphic Lie groups. 72 Tubular neighborhood. 468 Structure constants. 3 7� Subbundle. 7 area. 66 Vick. 49 closed. 374 right. 354 integral. 442 Thorn lsomorphism Theorem. 460 Translation left. J W. 1 1 3 even relative. 303 Triangulation. 1 1 3 mixed. 321 Sylvester's Law of Inertia. 64 outward pointing. 285. 50 Subgroup Lie. 352 Stonee ech compactification. 221 Steiner's surf ace. S. 349 Symmetric bilinear f orm. 147 Surface. 64 Tangent vector inward pointing. 459 489 Tensor contravariant. 288 Tensor field classical definition of. 231 odd relative. 461 Trivial vector bundle.
.
The figures were produced with Adobe Illustrator. A thicker set ofnumerals was constructed for use with the upper case lettering in chapter headings and statements of theorems.. were made for various pu rposes. Numerous other modifications of this sort. so a nonextended version was created. results from the linear scaling. the tall initial letters beginning each chapter were designed specially. The bold face supplied by Monotype� even the "semibold": is obtru sively extended. 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.in. and special parentheses and other punctuation symbols were also required. The covers. as well as the fact that the upper case Baskerville letters are of somewhat heavier weight than the lower case. Although a Baskerville bold face is unhistorical. bold type was useful in special circumstancesmainly for indicating defined terms. in superscripts.e Rime qfthe Ancient Ma. special gg and gf ligatures were also required. and in second order superscripts. as well as for the extension font and its bold version. Knuth's 'lEX typesetting system. painted by the author in his spare moments. The text font is I I point Monotype Baskervillethough the emdash has been modified together with its italic. are loosely based on Samuel Tavlor Coleridge's poem TI. in 16 point type. and additional special symbols.. The bold letters for mathematics come directly from the bold fonts of the Times families. and new or modified fonts were created using FontographeI. since simple scaling would have made them u npleasantly heavy. so a special gy ligature was added. 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. and J cause problems in words like topologl' and apology. 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 proportions and weights for these were then used for the three sizes of the symbol font and the other mathematics fonts. .hese books were typeset using Donald E. On the other hand. including bold symbols. which was created in three essential sizes: 9. DVIPSONE PostScript driver. including additional kerns and alterations of set widths. . The elegant swashes of the italic ). 7 and 5� point. The Adobe Mathematical Pi 2 font was used for the ordinary sized German Fraktur letters. The somewhat bold appearance of chapter headings. while blackboard bold letters were made by hollowing out these bold letters. with su itably modified versions used for su perscripts. script letters.
. corresponding to a M x JR2 . . . 43: Replace the last line and displayed equation with the f ollowing: Since rank f =k in a neighborhood of p. . liP) is called the boundary of 4> : V + pg. the function J can be made Coo if (g) The same is true if M has at most countably many components. v(p )) for some v(p) E JR . Problem 32: For clarity. .A is Coo. and aM itself is always a manifold JR" (but only one homeomorphic V pg. line 3: change pg. the lower rectangle in the matrix ( a(Ui ) o f) ax} = pg. there is a proper map J : M + JR. homeomorphically onto 2 T(M. A. 1 1 8: Add the following to the statement of the theorem : If . A is also. Add the hypothesis that = e(fh) = J(p)e(h) + h(p)e(f) = 0 + e(f).y) ::: 1. . then aM = 0. pg. . If there were a way to map + N for a disconnected manifold N. restate part (c) as f ollows: (c) This is f alse if J : (f) If M is a connected manifold. Xk . 60. �3 . . 1 1 7: After the next to last display. . . A (p) (X. . JR is replaced with J : M. Problem 29(d). Xdp)). then each dashed vector. . JR2 . fibre by fibre. Xk)(p) If A pg. + M is a Coo manif old. and we could continuously pick w (p) E criterion that w(p) should make a positive angle with v(p). ( p). 78: the third display should read: 0 = e(O) pg. Xk) is a Coo f unction for all COO vector fields XI . then A is Coo. . Equivalently. x E aM if and only if there is a neighborhood of x and O. 3. Problem 30: Change part (f) and add part (g): pg.CORRECTIONS FOR VOLUME I pg. 14: relabel the lower left part o f the central figure as di(x. '� . . 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. 103. . then is Coo. . add: M is orientable. pg. 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). theorem of topology). A(X" . in the sense that A(X) . = pg. 22: Replace the top left figure with IHI" such that M and is denoted by 4>(x) = aM..i). 6 1 .y) < 1 to di (x. by using the up would pg. If M is actually a manif old. M. .� �\ B: :B A.
.p N. 143.i. 226. # 0. Though there is considerable variation in terminology. pg. 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. 292. 408. for 1R3 . . " . is . what are called "even scalar densities" might best be called "signed scalar densities". replace parts (b) and (c) with: I (b) Determine the . we should be considering the h of part (a). Ell /'. O. pg. 233. ajaXI. In the comutative diagram. In Problem 26. 134.A(ajaxl" . V375" refers to pg. show that . 408410. In Problem + . 375 of V olume V pg.1) x (n . 133. ). part (c) should read: . . the lower right entry should be "Iforms on pg. Aut should be replaced with Aut . . scalar densities In part (c) of Problem 10. In Problem 20.+I pp. pg. show that n Vj # 0 should be CJ n V. End with End . we 17. p. 177. Extending the changed terminology f rom pg. omitting the modifier «signed" when it involves I det A l w. A follows from the fact that the function Ail . and suppose that f 5. . 1 3 1 . . what are here called "odd scalar densities" should probably simply be called uscalar densities". .] is integrable. 237.. pg. 4 1 1 . 3 should b e changed s o that i t reads: x and E V and let 0<1 . . when the transformation rule involves (det (7) A)W. restricting densities to those of weight 1).. The hypothesis o f Theorem Let pg. the condition V. For consistency with standard usage. 198.pg. 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. The display in Problem 21 .1) submatrices of the matrix In particular. The reference "pg. . 1 1 9 : Add the f ollowing at the end of the proof: Smoothness of pg.. Problem (d) Let f: M pg. . h component of VI x ··· X VnI in terms of the (n . V. 1. must also assume that each /'. and then replace Problem 19. Problem 16 (b) should read: "For any Lie group G. part (d) should begin: = for some to E I. 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.. pg. . p N". In part (g) of H is a connected Lie subgroup. add the hypothesis that pg. pg. 2 C V. Problem 9: Let F be a covariant functor from V. 133. For X Y E Mp and . 0<2 be two maps on some open interval 1 such that 0<1(1). .
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.