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
I am also extremely grateful to Richard Palais. f or otherwise the second half of these notes would not have been written.XlI Priface to the First Edition in manifold theory with a pace that few would describe as excessively I am leisurely. grateful to the class for keeping up wi th my accelerated pace. . whose expert knowledge saved me innumerable hours of labor.
. . Equivalence of Tangent Bundles Problems . . . . CHAPTER I. and on what pages. THE TANGENT BUNDLE The tangent space of �n The tangent space of an imbedded manifold . Immersion theorems Partitions of unity Problems CHAPTER 27 31 35 40 42 50 53 3. . . . . . MANIFOLDS Elementary properties of manifolds Examples of manifolds Problems CHAPTER . . . . Coo functions Partial derivatives Critical points . . DIFF ERENTIAL STRUCTURES COO structures . I . . . .. Xlll 63 67 71 75 77 82 84 89 95 . . . . . . . . . Vector bundles .TABLE OF CONTENTS Although the chapters are not divided into sections. and derivations Vector fields . 6 20 2. . Orientation . Addendum. . .. . . . the listing for each chapter gives some indication which topics are treated. . . The tangent bundle of a manifold Equivalence classes of curves. . . . .
Forms . . Addendum Addendum Problems I. . Problems 20 1 203 207 210 215 218 225 227 . .CHAPTER XlV Contents 4. TENSORS The dual bundle The differential of a function Classical Multilinear functions Covariant and contravariant tensors Mixed tensors. . . Differential Equations Parameter Curves in Two Dimensions . . Brackets . . . classical integrability theorems Local Theory. . 1 35 1 39 143 148 150 1 53 1 64 1 67 1 69 CHAPTER 6. . INTEGRAL MANIFOLDS Prologue. . Differential of a form Frobenius integrability theorem (second version) Closed and exact forms The Poincare Lemma . . The local flow Oneparameter groups of diffeomorphisms Lie derivatives .. 2. 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. 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. .
. . . LIE GRO UPS Lie groups . Left invariant vector fields Lie algebras . The equations of structure . . Oneparameter subgroups The exponential map Closed subgroups Left invariant forms Biinvariant metrics . . . . Subgroups and subalgebras Homomorphisms. 30 1 308 312 316 323 334 341 344 348 CHAPTER 1 0. ... . Tubular Neighborhoods Problems .. Stokes' Theorem . . Problems 37 1 374 376 379 380 382 384 39 1 394 400 402 406 . Riemannian metrics Length of curves The calculus of variations The First Variation Formula and geodesics The exponential map . .. . . . . 239 246 248 253 256 258 261 263 283 CHAPTER 9. .. .Contents CHAPTER xv 8. . . Geodesic completeness . . . de Rham cohomology Problems . . . . . . Addendum. Integrals over manifolds Volume elements . . . . RIEMANNIAN METRICS Inner products . INTE GRATION Classical line and surface integrals Integrals over singular kcubes The boundary of a chain Stokes' Theorem .
The Thorn class · · · · · · · · · . Index of a vector field . .CHAPTER XV! Contents I I. . . 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 . . . 477 · 481 . . . . .. . 2 6 7. . . . . M aye rVietoris sequence for compact supports The exact sequence of a pair Poincare Duality . . . . INDEX . 459 467 47 1 472 473 474 475 NOTATION INDEX . 9. . . · · To Chapters Problem . . The Euler characteristic. . . EXCURSION IN THE REALM OF ALGEBRAIC TOPOLOGY Complexes and exact sequences The MayerVietoris sequence Triangulations . . . . 10 . . .
Comprehensive Introduction to DIFFERENTIAL GEOMETRY A VOLUME ONE .
.
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. For n the single point A manifold is supposed to be "locally" like one of these exemplary metric spaces �n. . To be precise.n as a metric space.n. d(x. for each � n we can take U to be all of �n. [If you know anything about topological spaces. where � is the set of real x = (xl.y) = �)yi .xn) with each E �.CHAPTER 1 MANIFOLDS numbers. . a manifold is a metric space property: If 0 E �. =0 [=1 we will interpret �o as M with the following n=::O such that x E M. �n supplied with an equiva M is endowed plays almost no role. This example immediately suggests the next: any open . ] The second simplest example of a manifold is an open ball in �n. ric with which mention it. . just �n itself. is also a manifold. which should be consulted if one allows a manifold to be nonmetrizable.xi)2 . consisting of . in this case we can take Indeed. then there is some neighborhood U of U is homeomorphic to �n . Clearly. supplementing various chapters. and we shall almost never U to be the entire open ball since an open ball in �n is homeomorphic to �n. x and some integer lent metric (one which makes it homeomorphic to �n with the usual metric). 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. 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' unless another metric is explicitly suggested. Whenever we speak of ffi. . of course. you can replace "metric space" by "topological space" in our definition.
The open subsets of �n already provide many different examples of manifolds (just how many is the subject of Problem 24). some preliminary remarks need to be made. 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. and thus open in M. M. an open submanifold of the original manifold). then ¢(V) C �n is an open set containing ¢(x). though by no means all. Since ¢: V � �n is continuous. Before proceeding to examine other examples. quite naturally. there is an open ball W with ¢(x) E We ¢(V). the set ¢1 (W) is open in V. Exercising a mathematician's penchant for generalization. it is. of course. ConseIf is a point of a manifold some open set x quently.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. Thus x E ¢l(W) C V C U. 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. This compli cated little argument just shows that we can always choose the neighborhood U in our definition to be an open neighborhood. homeomorphic to W. . and thus to �n. which constitute mOst of this chapter.
by the way. it begins to appear that. if M C �3 is O} U {(x.z) : z = M. and d2 . example. with metrics d. for example.) 1.y E Mi. 3 V must be open.Manifolds With a little thought. U M2 . (It follows that f(V) is open for any open V C V.z) : x = M={(x.y. but pleasantly geometric. In general. Theorem 1 is called "Invariance of Domain" . we can define di or J di= i 1 + di M. given can first replace each all x. Topology ojPla ne Sets. . THEOREM If . Singular Homology Them].y. Notice V in our definition was not open). y) < 1 for = min(di. treatment may be found in Newman. 0 and z=I} then we can choose n=2 for points in M. is an unnecessarily complicated device for producing one manifold from two. and M2. then V C �n is open and f: V � �n is oneone and continu f(V) c �n is open. For example. in fact. = 1 were false. we di with an equivalent metric Ji such that Ji(x . 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. 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. we need the following theorem. * All proofs require some amount of machinery. But to prove this. and 11 = 1 for points in M2 • This M. An oldfashioned. then there would x. stated here without proof' ous. Homology Tlteor)1 and Massey. I ). for it implies that the property of being a "domain" (a connected open set) is invariant under oneone continuous maps into �n. and f is a homeomorphism. The quickest routes arc probably pro vided by Vick. sO fI is continuous.
we note that 11 �n. a nonopen subset of is unique at each Consider once more a discrete space. As a first step. In particular. If X is a connected.y) (we assume that which are). however. M is clearly a manifold also. for if n > 111. that the n of our definition x E M. We will often need to know that this is the only way in which acompactness can fail to hold. uncountable discrete space is not acompact (it cannot be written as a countable ::I 111.m into 2. then X is acompact. Although different Il'S may be required at different points of a manifold it would seem that only one M.". Y E M otherwise i to any number of spaceseven uncountably many. which is a Odimensional manifold. the manifold is not connected.4 Then we can define a metric Chapter 1 d(x. 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)= I = {�i(X'Y) d on M = M. THEOREM. if not. The same phenomenon occurs with higherdimen sional manifolds. If M.y):::r} . U M2 by if there is some i such that x. locally compact metric space. It is convenient to refer to the manifold M as Mn when we want to indicate that M has dimension n. sO is any discrete space {o ifx=y if x # y. A disjoint union of manifolds M. and M2 are open sets. A manifold has dimension n or is ndimensional or is an nmanifold if it has dimension 11 at each point. both M. This unique n is called the dimension of M at x. In these examples. an union of compact subsets). then there is a oneone continuous map from ffi. Consequently. PROOF For each x E X consider those numbers ball r > ° such that the closed {Y E X:d(x. n can work at a given point x E M. The only compact subsets of such a space are finite subsets. The further deduction. since a space with one point is a manifold. In the new space M. and M2 are disjoint. and M2 are manifolds. they can be replaced by new sets M. defined by the metric d(x. as we see by taking a disjoint union of uncountably many manifolds homeomorphic to ffi. the resulting metric space is called the is a manifold. This construction can be applied disjoint union of the metric spaces Mi. is left to the reader.
Y) S r + d(x"x2)}. > 0 is an interval. The so the function r: X consequence Suppose A C . since X is locally compact). . then X is acompact.Y) S r} C {y E X:d(x2. some subsequence of the Yi. proof is as follows. this set includes all r > 0.x 2)} C {y EX: d(x2 .Manifolds 5 is a compact set (there is at least one such r > 0. Now the closed ball B of radius �r(y) and Let that . If..Z2. which we might as well assume is the sequence itself. The set of all such.d(xj. Since A is compact. For each i there is a Yi E A such Zi is in the ball of radius r(Yi) with center Yi.. 00 The triangle inequality implies that {y E X: d(xj. so that which implies that S (I) Interchanging x. for all y E A.Z3.y) r}. since If not. Then A' is also compact. � Z1. for some x. . n=1 U {y E X:d(x. then for each all such x E X define rex) to be onehalf the least upper bound of X= r. {y E X: d(x"y) S r . and X2 gives (2) � is continuous. Let A' be the union of all closed balls of radius r(y) and center y. converges to some point Y E A.y) sn} . be a sequence in A'. This has the following important X is compact.
suppose that x is a point in the closure of A.  3 = 3 II which i s acompact. defined by Sl = = {x E �2: d(x.2d(x y) > rex) � .:. This shows that if y E A n ) then x E A n'. and consequently some subsequence converges. �r(x) �r(x) 2 > d (x.y). The only connected Imanifolds are the line � and the circle. Yi) r (Yi)} are contained in B. Since X is connected. and A # 0 is ope and closed.6 center y is compact. Then there is some Y E A with d(x. After this hassle with point·set topology. {y E X : dey. . S I. or I ·dimensional sphere. . y) �r(x). it must be that X A. To see this. � I = = < By (I). I rex) . It is also closed. so x E A. Since eventually the closed balls Chapter 1 ? Yi Y and since the function r is continuous. Their union A is clearly open.O) = I }. Moreover. Now let Xo E X and consider the compact sets A {xo} An+1 An'. r(y) =:: . the limit point is actually in the closed ball of radius r (y) and center Y (Problem 10). So the sequence Zi is eventually in the compact set B. we present the longpromised exam ples of manifolds. Thus A' is compact.
It is ob viously homeomorphic to a subset of �4. while Sl x Sl is commonly called "the torus". 2Jr) � Sl defined by J(e) = (cose. From these few manifolds we can already construct many others by noting that if Mi are manifolds of dimensionlli (i = 1.•   the above diagram..(Of course. The 2sphere S2. then MI x M2 is an (nl +n2) manifold... together with J it shows that Sl is indeed a manifold. though not oneone. .!2. and it is also homeomorphic to a certain subset of �3 which is what most people have in mind when they speak of . on [0. In particular . There is another way to prove this. The pro jection P from the point (0. I)} onto � x {I}: this is proved most simply by calculating P: Sl .Manifolds 7 The function J: (0.. is our first example of a compact 2manifold or surface.{CO. illustrated in (0.{CO. is also a homeomorphism. .) The function g: (Jr.Jr) � Sl. X Sl is called the ntorus. or it suf fices to note that Sf is "homogeneous"there is a homeomorphism taking point into any other (namely. The point (0. n times {x E �n+1 : d(x.sine) E Sl simply by e E [0. defined by the same formula. I) may be taken care of similar ly. by projecting onto � x {I}. . it is even continuous. is a homeomorphism of Sl . I)} � � x {I} explicitly. I) . We will often denote the point (cose. 2).2]'[].sine) is a home omorphism. better suited to generalization.. it is always necessary to check that use of this notation is valid. Considerations similar to these now show that the nsphere S" = is an nmanifold. an appropriate rotation of �2). I) onto the line � x {I} C � x �. O) = I} Sl x .!<. commonly known as "the sphere".2Jrl.
). we throw . it is easiest to begin with a "handle".:) E �3 : J' > OJ. The next simplest compact 2manifold is the 2holed torus. the region of the plane contained between two concentric circles. To provide a mOl'c explicit description of the 2holed torus. a spare homeomorphic to a torus with a hole ('ut out.z) E �3 : (y_I)2 + Z2 al"Ound = 1/4} contained in {CO. the latter of which is also homeomorphic to the annulus. The resulting surface.y.. called a surface of revolution. more precisely. has cOlnponents homeomorphic either to the torus or to the cylinder S' x �.8 Chapter 1 "the torus": This subset may be obtained by revolving the circle {(O.
_____ .h boundary piece of the sphere..h handle with corresponding points on the i.9 away all the points on one side of a certain circle. .. " I . :: . _____ . : : . It is homeomorphic to the space obtained by starting with the disjoint union of 11 handles and a sphere with n holes. and which will be referred to as the boundary of the handle. " . it is also described as the disjoint union of two handles with corresponding points on the boundaries "identified". and then identifying points on the boundary of the i. _____ . The 2holed torus may be obtained by piecing two of these together. C3EE1. : :. " I . which remains in OUf handle. " . . ' " :: '=' There is one 2manifold of which most budding mathematicians make the acquaintance when they still know more about paper and paste than about .. '83 . I ' . The nholed torus may be obtained by repeated applications of this proce dure. ..
the Mobius strip can also be described as [0. We wish to identify each point x E S 2 with . .10 � Chapter 1 metric spacesthe famous Mobius strip. .I. "identified". With our recently intro duced terminology. and ( I . we obtain the Mobius strip with a boundary. I) instead . If we define Ion [0. which you "make" by giving a strip of paper a half twist before pasting its ends together.l) = 2 cos e + 1 cos circle of radius 2 . I) with (0.�cose.2". This can be described by analytically as the image in �3 of the function I: [0. not to two disjoint circles.1) � �3 defined I(e.1. as investigation of the paper model will show. .( 2sine .+ 1 cos � sine.  o � . . but our next example will force the issue. I) I) ° 1  � • .) x ( . this boundary is homeomorphic to a circle.2". y ''''e have not yet had to make precise this notion of "identification". " . I ) x (. cos � sine .) x [I . 1 sin � ) ..:e) 2cosB � .
is a lot harder to visualize than previous examples. so that [p) = [p). indeed.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). We will postpone for a while the problem of defining the metric giving the distance between two points [p) and [q). We will denote this set by [p) E 11'2. there is no subset of �3 which represents it adequately. p} for p E S2. 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. 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 .Manifidds 11 its antipodal point x E S2. The points of 11'2 are defined to be the sets {p. The space which results. for which J(p) = J(q) implies p = ±q. the projective plane. . We thus have a map J: S2 � 11'2 given by J(p) = [p). 11'2.
I) (s. which is homeomorphic to the disc D2 = x E �2 : d(x.(I.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.l) (0. 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. )')plane. since they are identified with points above the (x. This leaves the upper hemisphere (including the bounding circle). s # 0.1)). 1 J«s.1) (1.12 Owpter 1 In exactly the same way. { 0) ::: I}.I)I)) ifs=OorI. J: [0.1) {(O. 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. 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) (s.I)) { res.1) (1. [(0. V (s. Squaring things off a bit.I).I)}. and we must identify each P E Sl with p E Sl .1)) [(1. JI (V) [0. we can make things easier for ourselves by first throwing away all points of S2 below the (x. The dotted lines in this picture are the key to understanding 1l'2 If we distort the  . y)plane anyway. 1) (s.
by identifying points on the boundary and points on the boundary of the disc... we obtain a Mobius strip with B l " " A " a boundary (namely. that is.Manifolds 13 region between them a bit w e see that the front part o f B followed by the back part of A. � " ' A B the disjoint union of a disc and a Mobius strip with a boundary. is to be identified with the same thing at the lower right. although J is not oneone. al though not homeomorphic to 11'2 .. a little experimentation wil1 convince you that this cannot be done (without having the two pieces of doth pass through each other).ffi. The projective plane is thus obtained from . Such a function J . Unfortunately. the dotted line. 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. which is a single circle). it is locally oneone. There is dearly a continuous function f: p2 ?. can still be described mathematically in terms ( B 1 1 I . every point P E 11'2 has a neighborhood U on which J is oneone. The subset of �3 obtained as the union of the Mobius strip and a disc. in reverse direction. : u of p2 . moreovel.. we are left with two pieces which can be rearranged to form something homeomorphic to a disc. If this Mobius strip is removed. 3 whose image is this subset. at the upper left. in other words. both of which are circles.
. . . although not topologically imbedded (there is no homeomorphism J from 1l'2 to a subset of �3).. The crosscap together with the disc at the bottom is a topologically immersed 1l'2 I��\ D'\jC � . . . . . . .. 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. .. explained in Chapter 2). . .rith... the disc can be added so as not to intersect the Mobius strip. In �4. with an extra dimension to play around v·... . (This is also obvious from the fact that the Mobius strip is the projective plane with a disc removed.. however. . . We can thus say that 1l'2 can be topologically immersed in �l... . this Can be done in the following way.. 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... . . . When the resulting figure is drawn up into 3space and the appropriate identifications are made we obtain the "crosscap".. ... .. ...
y. first of all (I) II' a.xy. Clearly J(p) = J( pl· We maintain that J(p) = J(q) implies that p = ±q.b.x2 +2y2+3z2). suppose that J(x..e yz= be xz=ac xy= abo #0. We have. 'Ie) prove this.xz. 15 'J'he one gap in the preceding discussion is the definition of a metric for p2 . are. b e . (fhose who know about topological spaces will recognize it as a disguised case of the Urysohn Metrization Theorem.e). it shows that things like ll'2 . whidI ought to be manifolds.Manifold. this gives a+ b+c= ±(x+y+z).:. The missing metric can be supplied by an appeal to Problem 31.b.:. Consider the function J: S2 � �4 defined by J(x.) For the present. which will Iatcr be used quite often. Roughly speaking.a+b+c=±x 1+. and which the reader should peruse sometime before J't·acling Chapter 3.z) = (yz.+. we wi)) obtain our metric by a trick that simultaneously provides an imbedding of ll'2 in �4.y. 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). =±x a+b+e ' a ( ) ( ) . so we also have hence (3) Using (2).z) = J(a. how('ver.
(x. so b (a. J('l))' = = = = . = 0. (4) also shows that the same sign holds in (7) and (8). we can define j: ll'2 � �4 b. so (5) gives = 3 .±I. 0. = This I11ap is oneone and we can LIse it to define the metric in p2 : ci( [p].O).O) or (O. Z = ±b.c) These equations clearly contradict = (O. Now suppose a = o. = 0 or e = 0 and Thus x = (4) 0 also. ['IJ) d (j([p)). In this case we have proved our contention without even using the fourth coordinate of f. which completes the proof."ith the same sign (which comes from (3)) holding for all three equations.y.O. _ b2 (8) z (this holds even if y = b = since then z. Similarly. Y Now (4) gives = = 3 ±b . b .O. j(['l))) d (j(p). =3 _ _ y2 ) = = I I.16 Chapter 1 so x = ±a. If x then (I) would immediately give y = z = so that #0. we obtain y = ±b. = ±c Since J(p) J('l) precisely when p ±'l.' j([p) ) J(p).y2 (7) 2y2 + 3 (1 y2 .±I). 0.z) But y = z = 0 implies (by (I) again) that be = (±I.c = ±I). 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 . Clearly. ".
y.xy) is a topological immersion of p2 in R 3 .x) [0. The image in R3 is Steiner's "Roman surface". or. g([x. (I. Consider the surface obtained from the square with identifications indicated below. it has a nicer.z)) (yz. Although this can be pictured as two crosscaps joined together. x' . = V"ith the new surface p2 at our disposal. which amounts to sewing two Mobius strips together along their boundary. For example. representation. By the way. 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. to a handle we can attach a projective space with a hole cut out. we can create other surfaces in t he same way as the nholed torus. and famous. the map ll'2 � �3 defined by the first 3 components of g: j. what amounts to the same thing.x').Manifolds 17 Then one can check that the open sets are indeed the ones described above. a !\16bius strip.xz. The closest we can come to picturing this is by drawing a cross cap sticking on a torus. We can also join together a pair of projective planes with holes cut out. (O. Notice that the [0.
so that the set {PI. P2. and P4 to be identified. which separates the surface into two parts. Xl). below. I) x Sl immediately suggests an immersion of the surface. The description ill terms of [0. P2. Turning one end of the cylinder around and pushing it through itself orients the lefthand boundary so that is directly opposite (I. form a single circle. P3. 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. one of which is shaded. P3. (O. to which il can then be joined.18 Chapter 1 identifications on the square force PI. P4} is a single point of our new space.x) �  . forming the "Klein bottle". dividing the sides into thirds. The dotted lines Pl P.
in addition to the family of nmanifolds S".Manifold.:: 0 such that U is homeomorphic to either�" or IHI". p} for p E S". if M is a manifoldwithboundary of dimension < 11. but \\. Points no these "boundaries" do not have neighborhoods homeomorphic to ffi. then aM = 0. then aM is not necessarily the same as the boundary of M in the old sense (defined for any subset of�"). then all points of M will be boundary points of M. then there is some neighborhood U of x and some integer . If manifoldswithboundary are studied as frequently as manifolds. We have already discussed some spaces which are not manifolds only because they have a "boundary".rill see later that the spaces pll for even n differ in a very important way from the same spaces for odd n.. The (closed) halfspace IHI" is defined by !\ manifoldwithboundary is a metric space M with the following property: IHI" = {(xt. we Can therefore distinguish those points x E M having a neighborhood homeomorphic to IHI".1I. 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 x E M. Projective nspace Il'" is defined as the rol1ection of all sets {p. .x") E�" :x" . a nonbounded compact manifold is calJed a "dosed manifold". The set of all such x is called the boundary of M and is denoted by aM.::O}. but they do have neighborhoods homeomorphic to an important subset of�". for example. If M is actually a manifold. One further definition is needed to complete this introduction to manifolds. Notice that if M is a subset of�".. indeed. 19 Examples of higherdimensional manifolds will not be treated in nearly such d"tail. but. We wi11 stick to the other terminology. we v. Often. we will mention the t'('bted family of "projective spaces". the Mobius strip and the disc. it becomes bothersome to use this long designation..j1J sometimes use "bounded manifold" instead of "manifoldwithboundary". A manifold in our sense is then called "non bounded". the word "manifold" is used lor "manifoldwithboundary". 11 A point in a manifoldwIthboundary cannot have a neighborhood homeo morphic to both�" and IHI" (Invariance of Domain again). . .
y) = 1 otherwise. The space (X. (c) X is locally connected if and only if components of open sets are open. (a) A subset of an nmanifold is an nmanifold if and only if it is open. X. and consequently homeomorphic to the di�joint union of its components. where X = U. and X. 6. then I is either increasing or decreasing.d) � (X. for i E I.. and a connected manifold is pathwise connected. y)== 3. '" 2.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. d. then (X.y E X. for each i. . (a) Every manifold is locally compact.. (a) The neighborhood U in our definition of a manifold is always open. is homeomorphic to X. are disjoint open sets and Y. n Xj for i # j.d. for some i. < I. (a) Connectedness does not imply local connectedness. which are open submanifolds. If (X. but a direct proof of arcwiseconnectedness can be given for manifolds.) are metric spaces. 0') An open subset of a locally connected space is locally connected. Each X.y) if x.(x. (a) If U c � is an inten'al and J: U � � is continuous and oneone. while d(x. then both d d!(I + d) and d min(I. A difficult theorem states that every path contains an arc between its end points. and d(x.e. 7. the identity map I: (X. d) are also metrics and that they are equivalent to d (i. and Y is homeomorphic to X if and only if Y = U..20 Chapter 1 PROBLEMS 1. 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.) 4.joint union of its com ponents.d) is a homeomorphism). 5.. (b) Every manifold is locally pathwise connected. where the Y. (A path is a continuous image of [0. then the dimension of M at x is the same for all x E M. is an open subset of X. Show that if d is a metric on X.d) is a metric space. (e) Every manifold is locally connected.11 but an arc is a oneone continuous image. Y. with metrics d. (b) If M is connected. (c) A connected manifold is arcwise connected. (d) A locally connected space is homeomorphic to the &. (b) The integer n in our definition is unique for each x.
(c) The map I is a homeomorphism.1I  B B is connected.{(O.1I for m ::I n.I)} � �nt. Every connected manifold (which is a metric space) has a countable base its topology.A has 2 components. then �n . 0') The analogous condition is necessary for the Mobius strip. (a) Compute the composition I = s' . c �n is homeomorphic to Dn (a) One component of �n . = {x E �n : d(x. then /(inside of A) inside of I( A) . � 13. Show that this last condition is actual1y unnecessary. (b) The image I( ) is open..{CO. . 0') Check that 1l'2 is a surface.A (the "outside of A") is unbounded. and A is the boundary of each.O. assume (I) (The GeneralizedJordan Cun'e Theorem) If A C �n is homeomorphic to snt. (2) If ffi. (b) Do the same for I: snt . where V c S2 is open and contains . .Manifold. For this problem. A C is homeomorphic to snI and I: � IR:II is oneone and continuous (so that J is a homeomorphism on A). 11.. (a) The text describes the open subsets of 1l'2 as sets of the form I(V). then V = V V 10.p whenever it contains p. 21 V 8. (h) I f c �n is open. Explain how the two cases differ. Ie)!' 9. and the other (til<' "inside of A") is bounded. and show that it is a homeomorphism. which is discussed immediately afterwards. (a) Check that the metric defined for 1l'2 gives the open sets described in the text. . In the proof of Tlleorem 2.O) S I}. 12.) (e) Pl'Ove Invariance of Domain. and a countable dense subset. (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. 14. show that the limit of a convergent subsequence or the Zi is actually in the closed ball of radius r(y) and center y. I)} �t x {I} � �t explicitly for the map P on page 7.. (First prove c.
Show that pn . C is homeomorphic . and that all compact surfaceswithboundary are obtained from these by cutting out a finite number of discs. we can consider pnI C pn in an obvious way. and since antipodal points in snI are still antipodal when considered as points in sn.22 Chapter 1 16.pnl is homeomorphic to interior Dn = {x E �n: d(x. To which of these "standard" surfaces are the following homeomorphic? 15. 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. Show that �2 to the surface shown at the top of the next page. (b) Since we can consider snl C S n. a hole in a hole in a hole 17. (a) Show that pI is homeomorphic to Sl .O) < J}. Let C c � C �2 be the Cantor set.
More generally. � . but not if n = I. c Cz implies e(Cz) C c(CIl. show that c(C) is either always the "left" component of� . the "right" one containing all numbers> some N. (b) Show that�n has only one end c for n > I. (a) �n has one end if n > I. Define a topology on Xu 8(X) by dlOosing as neighborhoods Nc(co) of an end cO the sets Nc(co) = 19. Show that Xu 8(X) is a compact Hausdorff space. (b) �n . locally compact Hausdorff space X. 18.1 23 (a) If C c� is compact.{OJ is not homeomorphic to�m. (A) The infiniteholed torus: for all compact C.C. This problem is a sequel to the previous one. Let 8(X) be the set of all ends of a connected. or always the "right" one.C has exactly 2 unbounded components. X has exactly one end c if and only if X "has one end" in the sense of Problem 18. Consider the following three surfaces. An end of X is a function c which assigns to each compact subset C c X a lionempty component c(C) of X . co(C) U {ends c: c(C) 20. it will be used in Problem 24. locally connected. the "left" component containing all numbers < some N. Thus� has 2 ends.K is connected.{OJ does not "have one end" so�n . A locally compact (but noncompact) space X "has one end" if for every rompact C C X there is a compact K such that C e K e X and X .Manifold. and�n U 8(�n) for n > I? = co(C)}. If c is an end of�. then� . in such a way that C. What is� U 8(�).C. (c) This part requires some knowledge of topological spaces.
which occurs at the left of (A). 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. while surface (B) does not. and consider the region between the two pairs. lH 1H 1\( 1\1 �1 ]l11'.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. Jll10l0' _ �JlllUlll �Il\ fiJllL � __ I _ e Ie .
. • • C') () () .J(x». (b) The points inside the three surfaces of Problem 20 are homeomorphic.A) is homeomorphic to A.e. j(x) . . � () () () () .j... For the purposes of this problem we will use a consequence of the Urysohn Metrization Theorem. J M M J: M j(M) M. that for any connected manifold there is a homeo morphism from to a subset of the countable product � x � x . which is eventually in the complement of every compact set. 23. (a) If is a connected noncompact manifold.A and �2 . define Show that is closed. (b) If A c �2 is closed and totally disconnected (the only components of A are points).) (b) Given a homeomorphism � � x � x . . . . then there is a continuous function � � such that "goes to co at co". . 25 ···· • • • C) C) () () . (b) There are only countably many nonhomeomorphic open subsets of �. then e(�2 . and a � � which = � � x (� x � x . C)··" 24. if { n} is a sequence � co. Hence �2 . (a) The three open subsets of �2 shown below are homeomorphic.Manifolds 21. .. (a) It is possible for �2 . . .A and �2 . ) by goes to co at co. . We define A(n) inductively by A(I) = A' and A(n+l) = (A(n».• • infinite swiss cheese .Bare nonhomeomorphic if A and B are nonhomeomorphic closed totally discon nected sets. (c) There are at most C nonhomeomorphic connected manifolds (where c = 2No is the cardinality of �). i. (a) Every open subset of � is homeomorphic to the disjoint union of inter vals. () () : i i. . (c) The derived set A' of A is the set of all nonisolated points. . J j: M J: M x J(x ) n g: M(g(x).. then (Compare with Problem 230...j C) 22. . For each n there is a subset An of � such that A/n ) consists of one point..Bto be homeomorphic even though A and B are nonhomeomorphic closed subsets.
26 Chapter 1 '(d) There are c nonhomeomorphic closed totally disconnected subsets of �'. . is a manifoldwithboundary. Hint: Let C be the Cantor set. e) E �3 : b'e'+ a'e'+alb' = abe} is the union of the Steiner surface and of the portions (co.co) of each axis.' . then the topological boundary of M.b. etc. < e3 < . abe.e) satisfies this equation and 0 oft D = Jb'e'+a'e'+a'b'.. then aM is a closed subset of M and aM and M . This is not necessarily true if M is not a closed subset. (a) Every point (a.b.b. one can add a set A i such that its n l1. b. (a.1/2) and (I /2. Hint: Let x = be/D. . (b) If M is a manifoldwithboundary. 27. (e) There are c nonhomeomorphic connected open subsets of �2 25. 26. then M . c) on Steiner's surface satisfies b'e2+a'e'+alb' (b) If = . a sequence of points in C.aM are manifolds. then (a. .U. and e) < e.'h derived set is {cil. and I' c J. (c) If C. as a subset of �n . For each sequence n) < nz < . . (c) The set {(a.eI' C. (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. i E J are the components of aM. If M C �n is a closed set and an ndimensional manifoldwithboundary.e) is on Steiner's surface. is aM.
.CHAPTER 2 DIFFERENTIABLE STRUCTURES freshly sharpened. if "' : V � �n is another homeomorphism. 229. so that I i = Xi . rather than by :n: (which will be used often enough in other contexts). including some problems.I : �n � � is also differentiable. The neces (x) we can expect J 0 1/1 1 to be differentiable for all J which make J 0 </>. we will denote the identity map from �n to �n by I. 225. Indeed.' differ entiable only if </> 0 ". and 335. 226. On a general manifold M the notion of a continuous function J : M � � makes sense. This is the case despite the fact that M is locally like �n. notably 29. where differentia bility of functions can be defined. 332. If U C M is an open set and we choose a homeomorphism </> : U � �n.. and U n V # 0. but the notion of a differentiable function J : M � � does not. it would seem reasonable to define J to be differentiable on U if J 0 </>1 : �n � � is differentiable. 215. Unfortunately. however. are contained in Chapters 2 and 3 of Calculus on Manifolds. then it is not necessarily true that J 0 ".I : �n � �n is differentiable. We will use freely the notation and results of these chapters. which the reader should bring along e are now ready to apply analysis to the study of manifolds. since Wsary tools of "advanced calculus". This is certainly not always 27 + .
V ) of an atlas A is called a chart (for the atlas A). for example. one need merely choose </> to be h o 1/!. Among all possible homeomorphisms from V C M onto �n. . namely. . it will suffice to consider homeomorphisms x: V � xCV) C �n onto open subsets of ffi. instead of considering homeomorphisms from open subsets U of M onto �n. we wish to select a certain collection with the property that </>01/!1 is diff erentiable whenever </>. the precise nature of which is suggested by the previous discussion. First of all. �n If we insist on defining differentiable functions on any manifold. each component function Ji possesses continuous partial derivatives of all orders). 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. The use of the letters X. where it is hard not to lapse back into the practice of denoting points by x and )'. where h : �n is a homeomorphism that is not diff erentiable. )" etc. if V n V = 0. and is automatically true. Moreover. Also./. instead of x alone. 1/! are in the collection. The only time this notation will be confusing (and it will be) is when we are referring to the manifold �n. V ). This make sense. but a few refinements will be introduced along the way. 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. sometimes we will use the words "differentiable': or "smooth" to mean Coo. for these homeomorphisms. just to provide a convenient name for the domain of x . . there is no way out of this impasse.28 Chapter 2 � the case. . We will often mention the pair (x. for the obvious reason that it provides a way of assigning "coordinates" to points on U. . since x( V n V) and y( V n V) are open subsets of �" . or a coordinate system on V. A particular member (x . VVe can even imagine a mesh of coordinate lines on U. xn (p ) to the point p E U.. . which has "coordinates" xl(p). . xn (p). This is preCisely what we shall do. we will be interested almost exclusively in functions J : �" � �n which are Coo (that is. . the coordinates xl(p). it makes sense. by considering the are Coo. It is necessary to adorn our manifolds with a little additional structure.
. We can easily make the atlas bigger. then .. LEMMA.4. one can easily construct 'U from . •:.4.. Another example is (�.\"3 .. (and one would be foolish not to do so once and for all).4.). denotes something which must be described in cumbersome language if one can refer only to A Aside from this. V) where V contains the homeomorphism x t? . Let .' 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 .4. where 'U (the "usual COOstructure for �n") is the maximal atlas containing {I}. if U and V are homeomorphic open subsets of �n...4. . PROOF. 'U) and (�..4. so . 'U differs only superficially from . Indeed. about the simplest example of a Coo manifold is (�n. 'U). is a maximal atlas for M. We now define a Coo manifold (or differentiable manifold.n parallel to one of the axes.. What has just been said for the atlas {I} applies to any atlas: I . V) are the sort of structures one would want to call "isomorphic".4. when applied to this atlas.. there is a oneone onto function J : � � � such that namely.4. we can adjoin any homeomorphism x : U � V with the property that x and XI are COO. is an atlas of COOrelated charts on M. or smooth manifold) to be a pair (M.' are Coorelated. where . 'U) and (�. If .' is an atlas..Differentiable 5lructureJ inverse images under x of lines in ffi. the identity J : �n � �n. and it is clearly the unique maximal atlas containing A . The term actually used is x E 'U if and only if x 0 J E V. The advantage of this bigger atlas 'U is that the single word "chart". we can adjoin as many such x's as we likeit is easy to check that they are all COOrelated to each other. of only one map.. Although (�.. is contained in a unique maximal atlas A' for M. V) are not the same. together with all charts Coorelated to it.4. the obvious map J(x ) = x 3 • Thus (�.4. Thus. whose inverse is not Coo. 29 The simplest example of a manifold together with an atlas consists of �n with an atlas .
: sn ..A) and (N. and II is clearly a diffeomorphism also.I ) of snI are easily seen to be Coorelated.(x) = g. 0. which are Coorelated to P and P2. It is easy to see that a diffeomorphism must be continuous. two homeomorphic manifolds are necessarily diff eomorphic. and the proof of the essential uniqueness of Coo structures on ffi. . . > O} � �n. . But there are 28 diffeomorphism classes of differentiable structures on S7. rather then differential geometry. xn). However. Later (Problem 924) we will be able to prove easily that � with any atlas is diffeomorphic to (�. 'Li). but we can already describe a diff crentiable structure . the definition of diffeo morphism would have had to be more complicated. There are. . we will suppress mention of the atlas for a differentiable manifold. Normally. of course. < O} � �nI .(x) = (Xl. This atlas may also be described in terms of the 2n homeomorphisms defined by f.30 Chapter 2 "diff eomorphic": two Coo manifolds (M.>4' on any open submanifold N of f. unique differentiable structures I on sn for 11 ::: 6. Consequently. . which are part of the field called "diff erential topology". we shall not come close to proving these assertions. the atlas for M is sometimes referred to as the diff erentiable structure for M. its inverse must also be continuous: so that a diff eomorphism is automatically a homeomorphism. . . conversely. . I) and (0. the projections PI and P2 from the points (0.. It will always be understood that �n refers to the pair (�n.I n {x E �n : x . x. 'Li). : snI n {x E �n : x . They therefore determine an atlac. . 2 A proof of the corresponding assertion for � is much harder. . . . .the "usual Coo structure for snb. . the proof for �3 would ccrtainly be too difficult for inclusion here.I g. up to diff eomorphism.I . and over 16 million on S 3 1 . Ifwe had not required our atlases to be maximal. .+ I . (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. X. . In the case of spheres.::: 5 requires very difficult techniques from topology. This raises the natural question whether. :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.n for n . . and speak elliptically of "the diff erentiable manifold M". 0.
V) in . as one would suspect. the atlas .. to the usual diff tiable structure on �.. it is easily seen to be true for any other pair. It is easy to see that (2) a function J : M � �m is diff erentiable ifand only ifeach Ji : M � �m is diff eren tiabl e.4. The diff erentiable structures on many manifolds are designed to make certain functions differentiable. 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. 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. We can thus define differentiability of J on any open subset M' e M.4. (3) a coordinate system (x. V ) is a diffeomorphism from V to xCV ). eren A differentiable function J : M � � refers. V) for N.). there are analogues of continuous maps.4. (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. with V C N.' 31 a differentiable manifold (M. . V ) and (y. V ) for M and (y. A function J : M � N is called differentiable if for every coordinate system (x . Consider first the product M. Just as diffeomorphisms are analogues for Coo manifolds of homeomor phisms. If this is true for one pair of coordinate systems. x M2 of two diff eren ..Differentiable Structure.' consists of all (x. the map y 0 J 0 XI : �n � �m is differentiable. Clearly. More particularly. a diff erentiable map is continuous. this coincides with differentiability of the restricted map JIM' : M' � N. of course. 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 . and hence J is differentiable if and only if J 0 XI is diff erentiable for each chart x.
pi) diff erentiable. A "handle" is then a Coo manifoldwithboundary. and any two such are COOrelated. To deal with Coo functions effectively. A differentiable structure on the 2holed torus can be obtained by "matching" the diff erentiable structure on two handles. 29). x M2 which makes each 7[1 diff erentiable.f. so that f l U is oneone. one needs to know that there are lots of I hem. UI) of coordinate systems on Mi) we construct the homeomorphism = defined by Then we extend this atlas to a maximal one. Similarly. It is easy to define a diff erentiable structure on M. Ca!. = = To obtain differentiable structures on other surfaces. there is a differentiable structure on ll'n which makes the map f: sn � ll'n (defined by f(p) [p) {p. We briefly recall here the necessary facts about such Coo functions (c. U) for sn. x M2 � Mi defined by 7[i(p" P2) Pi . The map X 0 (f I U)' is a homeomorphism on leU) C ll'n. The details involved in this process are reserved for Problem 1 4. . Consider any coordinate system (x. we cali f differentiable when it can be extended to a differentiable function on an open neighborhood of lHln. 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.culus on Malfif olds. The collection of these homeomorphisms can then be extended to a maximal atlas.32 Chapter 2 tiable manifolds MI. For each pair (XI. It is only necessary to know when a map f : lHln � �n is to be considered differentiable. where U does not contain . and the two "projections" 7[i : M. .p if it con tains p. we first note that a Coo manifoldwithboundary can be defined in an obvious way. pg.
j(xn ) = . there is a Coo function k : elsewhere. h: � � hex) { eI /x' x oft x h(n>(O) j: � � { e(xI)' . ]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. such that f = on C and support f e U. Let C c U e M with C compact and U open. The closure is called the support of f.8).1) I � � (I) The function = � defined by 0 o =0 n.:::: 0.. e) x . " x (e.) 1) I . LEMMA. il is 0 for x (4) The function g: �n � g(x) j(XI ) . it is positive on (e. 8) and 0 (3) The function I : � � � defined by is Coo. 8 > . and (2) The function i s Coo. e On a Coo manifold M we can now produce many nonconstant Coo func tions. J(x) x ¢ (1.Differentiable Structures 33 is Coo. e) and 0 elsewhere.I) . (Compare Case 2 of the proof of Theorem 15. increasing on (0. Then there is a Coo function f : M � [0. and denoted simply by support f (or sometimes supp f)· {x : /(x) oft O} 2. and � defined by /e je is Coo. Similarly.
ffi. then the charts of a maximal C� atlas are all COrelated.. This will not be proved here. J l + . Q' . . P p. for Q' = (J) the proof * For a proof see 1\. Of course. choose a coordinate system (x. Let J = I o (f + . Co manifolds for ".* In f act. # co will hardly ever be mentioned again. For each p E C. e ) for some e > o. V) with V c U and x(p ) = o. so a CO manifold is just a manifold in the sense of Chapter I. Since C is compact. of the corresponding functions has support C U. + J Pl p. the proof of Lemma 2 produces an appropriate Co filllction J OIl a Ccx lnanifold. for 0 :s. The symbol ew stands for analytic. a C� structure on M can always be extended to a Ccx structure in a unique way. + I . The function g o x (where g is defined in (4)) is Coo on V.n ?. ).1unkres. emellta1JI Diff ential Topolog)'_ 7 er . . e ) x . . it is not unique. (A function J : ffi. the smaller structure is the "stronger" one... . . and is positive on a neighborhood of p whose closure is contained in U. 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 . x (e. and the sum.:::: 00. where I is defined in (3). l:. Let J be the extended function. so on C it is 2: 8 for some 8 > o. but this atlas can always be extended to a bigger atlas of COrelated charts. The function J can be P P constructed for each p. we could have defined C' manifolds for each . not just for 00" . Clearly it remains Coo if we extend � U� it to be 0 outside of V. the CO structUl"e (consisting of all homeomorphisms x: U � �n) being the largest. 2: I. of course. finitely many such neighborhoods cover C. and it is convenient to agree that r < 00 < (J) for each integer .. but it is unique up to diffeomot·phism./) which converges in some neighborhood of a). . Then xC V) ::l (e. A "C o function" isjust a continuous function. One remark is in order now. is Cr ifit has continuous partial derivatives up to order r ) . On C it is positive. as in Lemma I. •: " r = By the way. Thus.34 Chapter 2 PROOF. If ". 2: o. every Ccx structure contains a CfJ structure for each f3 > 0' . < {3.. . The converse of this trivial remark is a h ard theorem: For Q' � 1.
. then is the classical symbol for this partial derivative.I ) define the curve Cj : (e.U) we define aJ aJ D.a' + h. . D. in fact it is just Notice that ifi = j ax' . then h. which . ax' (p) = ax' p c. .J(a) the number J(a' . . i= I X. h The Chain Rule states that if g : �m � �n and J: �n � �. e) � M by (or simply aI aXl = D.J(p) = aJlax' (p). then this partial derivative is just If x happens to be the identity map of �n. .U). ax} h� O r J(c. O lim Now.(f 0 x. 0» x 0 0 x. .. . so it measures the rate change of J along the curve c. .". Dd (a).(J J: M � � and a coordinate system (x. . an) . with respect to a coordinate system (x.J(a) .(h) = X.)(x(p» .(h» . so it is worth recalling a logical notation for the partial derivatives of a function J: �n � �. . .. At this point classical notation for partial derivatives is systematically introduced. We denote by D.Diff erentiable Structures 35 fails completely (and the result is falsean analytic function which is 0 on an open set is 0 everywhere). . If we .J(p) h ' {I (Joct!' (0).. . for a function Dj (f o g) (a) = L D. What we shall define are the partial derivatives of a dif ferentiable function J: M � �. h. . With differentiable functions now at our disposal. � (p) = 8}� = 0 if i # j.J(g (a» . ..I (x(p) + (0. I = ' as an equation between functions)... it is fitting that we begin differentiating them.
y) is the unique number in (0.. often not completely clarified. y) = we can introduce a "coordinate system" P : A ?..% (x .) We havc really defined P as an inverse function."raxis" 2" not restricted to the set A. (Of course.� .)..ffi. eo + 2".. y) sin e(x. r sin e). y).2 by p(x.) with x = r ex. e (x. On the subset A of �2 defined by A = �2 = �2 L _ {(x . �. and this sometimes justifies the sloppiness involved in the definition of the polar coordinate system. e) = (r cos e . One Can delete any ray other than L if e(x.. many results are essentially independent of which line is deleted. y) E �2 : y = 0 and x ::: O} where r ex. . the polar coordinate system is often + (r. Jx2 + y2 and e(x. 2rr).. yJ This really is a coordinate system on A in our sense.I (r. y) is restricted to lie in the appropriate interval (eo .2". y) cos e (x .. y) y = r ex. y) = (r(x.y� .I is defined simply by p . OJ "Oaxis" � +.36 Chapter 2 Another classical instance of this notation. y). with its image being the set {r : r > O} x (0. whose inverse p . is the use of the symbols alar and alae in connection with "polar coordinates". y» .
at the point (x. and thus the (x. y») + Dz f(x. y» is perpendicular to v. y). . The factor " (x. y) cos e(x. the inverse image under P (x' Y).). of the curve C2 v·. y). so aJ a. y). y) + Dzf(x. y) = DI /(r' (p(x.:. y» by the Chain Rule DI /(x. y) · cose(x. y) .(x.sin e(x. y).:. along a unit vector v = (cos e(x. [r')' (p(x..y)) .'hich is the inverse image under P of a cun.) /1.'n9(X. y). y) [r(x.DifJerentiable Structures From this formula we can compute aJlar explicitly: (f 0 p .y» · D.e along the "eaxis".y) /8(x.". .Y) = DI /(x.. . sin e(x. DI J 0 P . sin e(x.. A similar computation gives C l _ _ cos8(x. y)[r(x .I ) ( P(x. because CI .I )(r. e) = = 37 J(r cose.� / Xl' \ C2 . y» ( This formula just gives the value of the directional derivative of J at (x. This is to be expected.y) sin e(x. y) appears because this curve .y» + Dzf(P' (p(X. cos e(x. y) % direction. is just a line in this direction. r sin e). . y). . The vector w = ( .. "v' _ = ofa curve along the "raxis". y)) pointing outwards from the origin to (x. y»). DI [ rl )2 (p(x.
V) are coordinate systems on M. and suppressing the argument (x. of course. are. Using the notation aJlax for D1 !. Note that aJ/ae is independent of which line is deleted from the plane in order to define the function e unambiguously.38 Chapter 2 goes around a circle of that radius as e goes from 0 to 27[. the Chain Rule would always b e written i n this way. PROPOSITION. aJ ay sin e aJ cos e. y) denotes the identity coordinate system of �2 We have ax lar = COs e.. so Qur formulas can be put in the form  aJ aJ ax aJ ay = . if you just keep your cool: ar (p) = D.. these formulas also tell us what axlar etc. y) everyvvhere (thus writing an equation about functions). 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. I t is a pleasure to report that henceforth this may always be done: 3. etc.+ ar ax ar ay ar aJ aJ ax aJ ay = + ae ax ae ay ali ' I n classical notation. [x 0 yI )j (y(p» .. a ax ( rsm e) + / In particular. where (x. and J: M � � is differentiable. then on U n V we have (I) PROOF It's the Chain Rule. D.I )(y(p» a I I = D ([J 0 [ ) 0 [x 0 y. If (x. so it is going r(x. (J 0 y . U) and (y. we can write the above equations as �= ae ar aJ = aJ ax aJ cos e + .))(y(p)) i > = j =l I: Dj (f 0 xI )([x 0 yI )(y(p» ) .
the n matrix . n x PROOF.(pl· . which appears both "above" (in axj lay') and "below" (in aJlaxj ). No tice that the summation in this formula occurs for the index j.DifJerentiable Structures 39 = j= = :L Dj (J 0 [ I)(x(p» 1 1 j= aJ ax} :L � ( p ) . U) and (x'. . and felicitous choice of notation. U') are two coordinate systems on M. J= I n axj a here alay' is considered as an operator taking the function J to aJlay' . a)'i p ' I thus a ax j a = (p ) ax j P . PROPOSITION. one can avoid what Elie Cartan has called the "debauch of indices". I won't use this notation because whenever I do.:. There are scads of for mulas in which this happens. I soon forget I'm supposed to be summing. Left to the reader. and any coordinate system (x. often with hoards of indices being summed over. the operator f = alaxi lp satisfies f(Jg ) = J(p)f(g) + f(J)g(p ). The operator taking J to aJlay'(p) is denoted by a . We will often write formula (I) in the form a ay i = I: ayi axj . •:. ay' P J= l ayi n I I: For later use we record a property of f = alaxilp: it is a "pointderivation". ax) ayi n · Di [Xj 0 y. and because by do ing things "right". and the convention is to omit the L sign completelydouble indices (which by luck. If (x. For any differentiable J. almost always occur above and below) being summed over.I) (y(p» At this point we could introduce the "Einstein summation convention". I 4. g : M � �. U) with p E U. the nature of things.
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).a1 [a. In this case. If I : � � �. I = 0 everywhere. then D I = D. � � has all points as critical values. V) is a coordinate system around I(p). where v(Bn ) is the volume of Bn . it is called a regular point of f. is still a "small" subset of �'. . a function I: �. we will need some terminology. . We want to define the same concept for a subset of a manifold. b) to be critical points. U) or (y. of (closed Or open) rectangles with I and n=1 L v(Bn) < e. y) = x. 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). the rank of the m x 11 matrix (::. then x is a critical point of I if and only if j'(x) = O. however.1l has "measure zero" if for every e > 0 there is a sequence B" B" B3. Recall that a set A C ffi. The most important theorem about critica1 points generalizes this fact. This is true. may have all points as critical points without being constant. I(x. if q ¢ I(M) . It is called the rank of I at p. To state it. which we merely state. . . in fact. � �. although this can happen only if I is constant On [a. its Now if I : M n � N m is Coo and (y. Other points in N are regular values. the value I(p) is called a critical value of f. in particular. for example. the image I(�') = � x to} C �. If I : �.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. V). b). which in turn depends on a lemma from Calculus on Manif olds. if p is not a critical point of I. It is possible for all points of the inten.� (P») clearly does not depend on the coordinate system (x. It is nonsingular. so I is again constant. On the other hand. To do this we need a lemma.a nonvalue of I is still a "regular value".
If J : M � N is a C I function between two nmanifolds and A e M has measure 0. Thus J takes rectangles of diameter d into sets of diameter ::. (. V). n . .:0 n2 Klx . Lemma 5 implies that there is some K such that IJ(x) . Thus y(J(A) n V ) has measure 0.J(y ) 1 . This clearly implies that J(A) has measure ° if A does . PROOF. or has only countably many components. We can assume that A is contained in a compact set C (since �n is a countable union of compact sets). . then x(A n V) C �n has measure ° for any coordinate system (x. (But if M is the disjoint union of uncountably many copies of �. Since J (Ui Vi ) is contained in the Ullion of at most countably many components of N. y E C. Vi). y E A. such that each set xi (A n Vi) c �" has measure 0. then J(A) C N has measure 0. Conversely. it is easy to see that if A C M has measure 0. then A does not have measure 0).y l for all X. •:. if this condition is satisfied and M is connected. it follows that J(A) has measure 0. 6. n2 K d. There is a sequence of charts (Xi . then it follows easily from Theorem 12 that A has measure 0. PROOF.yl for all x. K on A for i. with A c Ui Vi. If J : �n � �n is C1 and A C �n has measure 0. Using Lemma 6. n 2 Klx . LEMMA. . A subset A of a Coo nmanifold M has measure zero if there is a sequence of charts (XI. and A consists of one point from each component. Then IJ(x) . If (y. Vi) with A c Ui Vi and each set xi (A n Vi) of measure 0. LEMMA. . by Lemma 6. = y 0 J o [l(x(A n Vi» . then J(A) has measure o.Differentiable Structures 41 5 . COROLLARY.J(y ) 1 ::. Each set y(J(A n Vi) n V ) has measure 0. V ) is a chart on N. Let A C �n b e a rectangle and let J : A � � n b e a function such that I DjJi l ::. Lemma 6 thus implies another result: 7. then J(A) n V = Ui J(A n Vi) n v. j = 1 .
. Topolog)' From the D{ff miiabtc Viewpoint or Sternberg. V) around J(p) with y 0 J 0 x t in the form yoJ0  I\1oreover. a n ) = (a t . If y is the identity of �m . 0).. and hence in a neighborhood of p. since J may not even be oneone on ffi. .1i1IIor. THEOREM (SARD 'S THEOREM). a n ) = (a t . contain only points with first coordinate o. . and its image could. . More exact information Can be given when J actually has rank k in a neighborhood of p. . . in terms of the rank k of J at p E M. SC'T 1\. O. and then permuting the coordi nates in �m. . . and M has at most countably many components. But this case is just Theorem 3.k + t (a). ak . If J : M � N is a C1 map between nmanifolds. It should be noted that J must have rank 2: k in some neighborhood of p. O) . . ". THEOREM. . then there are coordinate systems (x. . .14 of Calculus on Manif olds. then there is some coor dinate system (x. then the critical values of J form a set of measure 0 in N.k x to} c ffi. . XI k (a t . Remark: The special case M = � n . These diffcomorphisms on �n and �m are clearly necessary. and the case m > n is the trivial case (Problem 20). PROOF It clearly suffices to consider the case where M and N are �n . 9.n. Lectures 011 f'l" Diff erential Geomel1)'.42 Chapter 2 8. .. . part (I) says that by first performing a diffeomorphism on �n. U) and (y. which gives only local results.t (a t . . (I) If J: M n � N m has rank k at p. a . . 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. Theorem 8 is the easy case. * For a proof. for example. . . in Problem 824). because some k x k submatrix of (a(yi 0 J)/axi) has nonzero determinant at p . N = �m i s equivalent t o the general theo rem. . . (2) If J has rank k in a neighborhood of p. Although Theorem 8 will be very important later on. m (a» . . V) such that y o J o x. U) around p and some coordinate system (y. .In. The stronger version ofSard 's Theorem . . for the present we are more interested in knowing what the image of J : M � N looks like lo�ally. 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 . which we will never use (except once . . ".
an) = y 0 f(q) for q = x'(a'. k r = k + I.. . . � = I . . ..." o a. .. Condition (I) implies that (2) This shows that i s a coordinate system i n some neighbor hood of since (2) and the Inverse Function Theorem show that u ' is a diff eomorphism in a neighborhood of Now x = (x 0 u') 0 u p.. . 0 /» ) = ax} .. . . . since may not be contained in any kdimensional subspace of PROOF. fl (p») .. 43 fORn) JRn. . . . q = x'(a'. (I) det C au Define xa = ya 0 f xT = uT a = I.. .. . .ak. (2) Choose coordinate systems x and v so that v 0 f 0 x .(q) = a. . x0 u (p). hence x. { ya 0 f(q) = aar a = I.an) means x(q) = (a'.. . . .DifJerelltiabLe Structures In part (2) we must clearly allow more leeway in the choice of Y.. k . . (I) Choose some coordinate system u around p.an). k hence ur (q) = a r = k + I . n. has the form in (I). . . . __ ( a(v.n. ). . so y o f 0 x.an) = (a'. By a permutation of the coordinate functions ui and yi we can arrange that (ya . . . . Since rank f = k in a neighborhood of p. the lower square in the matrix . (a'....
. bm) the Jacobian matrix y is a coordinate system in a neighborhood f(p yofox'(a'. . THEOREM. vk). (1) If m S n and f: Mn . .an) = y (al... . .. then for any coordinate system (y.. ak» by (3) = (a 1 . .. . .. m . . ..44 Chapter 2 must vanish in a neighborhood of p. Thus we can write r = k + l. . m: V) U) . . . . .. around p with has nonzero determinant. . there is some coordinate system (x.. 1/rm(a» = (al. v(q) = (bl.. . " " 1/rm(a) _ 1fm(al. . . • . ak. . . . around f(p). . 1/rk+1 (a). 0.. . . so of ). . • . Define ya = va yr = vr _ 1fr for Since 0 (V i .an) = y o vI o v o j oxl(al.ak. .ak : Theorem 9 acquires a special form when the rank of f is n or 1 0. Moreover... Nm has rank m at p.. 0 V I . 1/rk+1 (a) _ 1fk+1 (al. .. . . . 0) . .. ak).
. C(Yaau.bn) for (b I . it is easily checked that A ljf is now the desired •:. then for any coordinate p. .an) (a' .. V) = .an) = A 1/1 f o¢'(b'. This can be conected by another map on ]Rm. in the proof of this case... fl (p») ..... .. .. . 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. C �)�(1/r:(7nIT{(f »:12 0 f(a'. .. Even if we clo not perform ¢' first.bn). .. (2) Since the rank of at any point must be :s n.. y = 0 y.O. 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.. 0). only the need be permuted. . . Define A by x f 0 <� f x = "..an) (a'. . 0) 0 y.. . . . bn))... 1/1 f o¢'(a'. .. . ..bm) '" (X(¢I (bI . .. . .. a .'" . 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. .bn. . f M '" ]Rn ]Rn.Differentiable Structures (2) If n :s m and system (x. . . there is a coordinate system around f(p) with y o f ox'(a' . (CPI (bI . ..an.bn) '" ¢(a) '" A(b'. � = l . it is clearly unnecessary.. .an.. A(bI . . . ..O. ..O.. > II (y. = = . .O). to permute the yi in order to arrange that det f: Mn Nm has rank at p.an. 0 Then A 1/1 0 0 .O. .. . U) around PROOF. .bn+'. . i t i s only necessary to observe that when k = In. . . . 0) .O) (a I .. bm). . . m. 45 (I) This is practically a special case o f (I) i n Theorem 9..60 f p.O..
the dimension of the domain M. and it is clear from Theorem 10(2) that an immersion is locally oneone (so it is a topological immersion. Of course. A differentiable function f : M n > N m is called an immersion if the rank of f is 11. it is necessary that 111 ?: 11. Although the second and third immersions f31 and fh are oneone. their images are not homeomorphic IR .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 ). with /. at all points of M. Another example is IR IR x>O x < x=O o. A more illuminating example is the function h : IR > 1R2 defined by h(x) = ( g (x). The simplest example is the function f : > defined by f(x) = x3. I g (x)l). One can easily define a similar curve whose image looks like the picture below. it is case (2) which most interests us.(0) = o. although its image is the graph of a nondiff erentiable function. Three immersions of in 1R 2 are shown below. the curve itself manages to be differentiable by slowing down to velocity 0 at the point (0. a differentiable map f need not be an immersion even if it is globally oneone. as defined in Chapter I).0) . On the other hand.
In general. . 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. g(q) = g(q').k.+.. Coo �+ f Coo p V C p 0 i Vn . : > S X S M MI M.n. second time around shows that may even be a dense subset of Despite these complications. MJ Mn VI MI E M" (y. indicating the image of an immersion fJ. l. > �!r! u. where { yr(q')) = 0ya (q) a = kl. even if the oneone immersion f : P 7 is not a homeo morphism onto its image. . MI M. . such that nV OJ. . = yn(q) = this consequence Theorem = g : MisI an immediate(considered as aoffunction on 10(2). I I. with Mr) in aThus. r i. M VI = {q E M : yk+' (q) = .DifferenliabLe SlTucluTes 47 to 1R. MI M MI M IR M M). The following picture. if is a k dimensional immersed submanifold of and is a neighborhood in of a point P then there is a coordinate system V) of around p. Of course. there is certainly some metric and some differentiable structure on f(P) which makes the inclusion map i : f(P) > an immer sion. . .
48 Chapter 2 On the other hand. PROPOSITION. on U.' : fJ.' 0 f is Coo ifit is continuous. then f is also Coo considered as a map into Mj • PROOF. c M is an immersed manifold. .. yk l U. . Actually. . If M. = . J is a coordinate system of M. Given P E P.(M) > One other complication arises with immersed submanifolds. = {q E V : yk+' (q) is a neighborhood of f(p) in M. as a map into M. Let i : M. If M.. > M he the inclusion map. = yn (q) = OJ . C M is an immersed submanifold. 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 . with its Coo structure. We want to show that j. V) for M around f(p) such that U. choose a coordinate system (y. this cannot be done if g is one of the functions rf. even if g is Coo on all of M. . The following figure shows that f might not even be continuous IR. we may not be able to define g on M. this is the only thing that can go wrong: 1 1 . . f : P > M is a Coo function with f(P) e M" and f is continuous considered as a map into M] . For example. and (y ' I U" .
\� � • •• % submanifold . defined as {x : Ixl 2 I}. is open in M" this means that j. and y ' .I 0 49 f is continuous.' (y) is an (II . so 0 j' i (open set) is an open set. An immersed subman ifold M. it is called a closed submanifold of M if M. In particular. as a special case. If j : M" > N has constant rank k on a neighborhood of j' (y). . . PROOF. Now it turns out that ever (connected) Coo manifold can be imbedded in some )RN . C M is called simply a (COO) submanifold of M if the inclusion map i : M. . > M is an imbedding. f : P 7 M which are homeomorphisms onto their image. but we first develop some of the machine. We will prove this fact only for compact manifolds. = Left to the reader. Since U.k (or is empty). . . lS also a closed subset of M.m)dimensional submanifold of M (or is empty). so that )' manifolds can be pictured as subsets of Euclidean space (though this picture is not always the most useful one). PROPOSITION. if y is a regular value of j : M" > N m .:. then j . ) C P is open. •:.{OJ c )R". It is to be hoped that however abstract the notion of Coo manifolds may appear. submanifolds of )R N will seem like fairly concrete objects. i.DifJerentiabLe Structures By assumption. Such an immersion f\1ost of these difficulties disappear when we consider oneone immersions is called an imbedding ("embedding" for the English). and gives the sphere S" . There is one way of getting submanifolds which is very important. then j .' C IR" . yk are a coordinate system on VI) the function f is Coo considered as a map into MI . Thus j takes some neighborhood of P E P into U.y which would be used in . . .' (y) is a closed submanifold of M of dimension n . Since all y j 0 j are Coo.' ( U.' 12.
PROOF We can obviously assume that M is connected. <. THEOREM. j = IRn. there are many definitions and theorems involved.2 . we can obviously cover Ai by a finite number of open sets. since we will need it later on anyway.. whose union contains all Ci .lhich refines (9 and which is locally finite. . with compact closure. U C2 has an open = neighborhood U2 with compact closure. U C2 U C3 U . Then V. Since each Ai is compact.Ui . with U. If (9 is a cover of a space M. there are compact sets C" C2 . C3 .50 Chapter 2 the general case.2. and hence is M. I f (9 i s a n open cover o f a manifold M . Unfortunately. . By Theorem 1. Continuing in this way. Moreover. 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. and each contained in Vi = Ui+ 1 . compact and U. C Uj+1 . A cover (9 is called locally finite if every p E M has a neighborhood W which intersects only finitely many sets in (9 . since a point in Vi is not in V for j 2: 2 + i. .Ui. We can also choose these open sets to be diffeomorphic to In this way we obtain a cover (9' v. then there is a n open cover (9' of M which is locally finite and which refines (9. has an open neighborhood V. we obtain open sets Vi. • • • with M C.l . . each contained in some member of (9. Clearly C. Let V_. 1 3. we can choose all members of (9' to be open sets diffeomorphic to IRn. Now M is the union for i > I of the "annular" regions Ai Vi . = Vo = 0.
Now P E it follows that since replacing C2 V2 V2V2. Then there is a collection of functions <pu : M . V{UV2U " ' . Then it is possible to choose. C C 14.. an open set with is also an C in such a way that the collection of all open cover of M. Un+i U�+j P E by cannot possibly eliminate p. Let (9 be an open locally finite cover of a manifold M. . . . V3 . C] U V2 U V3 U = ) = V'V {V]. . by (I)). Let be an . one for each in (9.. THEOREM (THE SHRINKING LEMMA). For any p E M there is a largest n with p E because (9 is locally finite. }. Let (9 Then is a closed set contained in and M = open set with C C C Now C] V] .+2 U··· ). such that Coo V (1) support <Pu C V for each V. . I ] .(U2 U V3 U . (2) "L <Pu (p) u neighborhood of p.. [0. C] V{ UJV]. V{ . .. then intersects only finitely many members of (9. for each in (9. V2.. We can clearly assume that M is connected. open set with c c c Continue in this way. Let (9 be an open locally finite cover of a manifold M. U V3 U n V{ U V2 U·· . THEOREM.V]. V' V' V PROOF.DijjewztiabLe Structures 51 Notice that i f (9 i s an open locally finite cover o f a space M and C M is compact. .V2. 15. This shows that an open locally finite cover of a connected manifold must be countable Qike the cover constructed in the proof of Theorem 13). V{ UVC2. Let W be an is a closed set contained in and M = i. = 1 for all P E M (this sum is really a finite sum in some . ' .. <. U V� U (Vn+] U V.
But this is a consequence of Case 1: For each p E choose an open set Up C U with compact closure. Chapter 2 Case 1. (3) for each i there is a U E (9 such that support ¢. VeC}' L ¢U. 1 7 . != L u ¢u(P) = I for all p. If (9 is any open cover of a manifold M. it is called subordinate to (9. where since it is a finite sum in a neighborhood of each point. 1 6. Apply Lemma 2 to U' c U e M to obtain a which i s I o n U ' and has support C U. Let C C C C C. I] such that (I) the collection of sets {p o ¢.52 PROOF. C. then there is an imbed . Choose the U' as in Theorem function 1/Iu : M > [0.(p) 1 for all p E M. C U. Each U in (9 has compact closure. THEOREM. and ¢u(p) = 0 when U C Va . clearly ! (p) = I for all p E Using the fact that (9 is locally finite.) 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. �. if it satisfies (3). COROLLARY. Since (9' = { U E (9 : U C Up for some pl. provided that Lemma 2 is true for c U e M with closed (but not necessarily compact) and U open. (A collection {¢. If M" is a compact cling ! : M > for some N. Since the U ' cover M . : M > [0. clearly Coo Define Ue(') L 1/Iu > 0 ¢U = everywhere. Cover Mwith open sets Va having compact closure and contained in M The open cover {Up. General case. = Coo (2) L ¢. Ue(') L 1/Iu Case 2. I]} satisfying (1) and (2) is called a partition of u nity.(p) # O} is locally finite. : This sum is Coo. 1 ] 14. jRN Coo manifold. it is easy to see that support f e u. then there is a collection of functions ¢. • . : M > [0. This case can be proved in the same way.
Choose V: as in Theorem 14. JR. by JRN . (Xk. Vkl with M = Vl U · · · U Vk . (b) In the proof of Lemma I. 2... as claimed.. because any poiIlt p is in V! for some i. PROBLEMS 1. There are a finite number of coordinate systems (Xl . Then Vri(p) = I. where N = lIk + k. Moreover. that is not the question being asked.. N is Coo if and only if g 0 f is Coo for every Coo function g : N . This is an immersion. How many distinct Coo structures are there on 1R? (There is only one up to diffeomorphism. Problem 333 shows that. Define f : M . (a) All Coo functions are continuous.Di/Ji:renliabLe S'lruclures 53 PROOF. This shows that we must have q E Vi . show that the 11 at each point is unique wi/houl using Illvariance of Domain..Ai are COOrelated.. Vri ' Xi (P) = Vri ' Xi (q) . show that all charts in .. so also Vri(q) = I . we can always choose N = ar ( axf ) contains the 11 x 11 matrix (ax�) axf ' . .. (b) Show similarly that aM is welldefined for a Coo manifoldwithboundary M. and functions Vri : M . so p = q) since Xi is oneone on Vi . •:. the N x II Jacobian matrix It is also oneone. (b) A function f : M . [0. 1] which are 1 on Vj and have support C Vi. . where VIi = I. (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. and the composition of Coo functions is Coo. There is some i such that p E Vj.. . and on Vi. 4.) . ..1. Vl ). 211 + I .. (a) Show that being COOrelated is not an equivalence relation. jRn 3. in fact. For suppose that f(p) = f(q)..
) IR lHIn.Ai consists of all (x.. Moreover.. 2).... U) in .. M. (Obvious....... 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). Ml x M2 is Coo if and only if the compo sitions 1fl 0 / : N ..) (b) If / : .e. are Coo (i = 1 . show thaI h( P2) ).... then / is Coo.4.' is the unique atlas with this property. ( PI .4..... [pl. then / is Coo... can one determine the rank of (fJ .. PI .. . P2) of Ml . defined by Ji x h(Pl . 6. IRn1 1 2. M is Coo.. x M2) X M3 is diffeomorphic to MI x (M2 x M3) and that MI x M2 is diffeomorphic to M2 x MI ..... (c) More generally... then there is a unique Coo structure on aM such that the inclusion map i : aM .. (b) The diffcrentiable structure on Ml x M2 makes the "slice" maps 8. p snl p are Coo related to the 2n p2 p sn snIP'n IRn lHIn IR IR lHIn... (b) Show that .. (a) If N C M is open and . (a) If M is a connected Coo manifold and .. M with c (O) = and c ( I ) = q . 1 0 . p? (pt}. /1 x 12 : Nl x N2 ..54 Chapter 2 5. with U C N. (b) It is even possible to choose c to be oneone. M.' can also be described as the set of all (x I V n N.... i s locally C oo (every point has a neighborhood on which / is COO). (iit ..... i. (a) Show that (MI 7.. Ml x M2 at in terms of the ranks of Ji at For Ji : N.... M2 are Coo. h) : N . h) .. q E M. MI and 1f2 0 / : N .4. M is Coo. then there is a Coo curve c: [0..4.. (a) If U C p IP'n i s open and / : U .. Let g : . V n N) for (x... Show that / : only if / 0 g : . 1 1 . Check that the two projections PI and P2 on homeomorphisms Ji and g. and that ... (d) If Ji : N . 9.. M is an imbedding.. 1 ] .. V) in A (c) Show that the inclusion i : N .. be the map . P2 ) = (Ji is Coo and determine its rank in terms ofthe ranks of ji. Compare the rank of / and the rank of / 0 g ...' is maximal for N if . h E M2.. a map / : N . If M is a Coo manifoldwith boundary. the Coo structure we have defined for Ml x M2 is the only one with this property. Ml x M2 differentiable for all fit E Ml .. M is Coo if and .. Ml x M2. M2 . / can be extended to a Coo function on a neighborhood of (Not so obvious. is locally Coo. show that .4.. .. is maximal for M. has two extensions g. If / : 1HI" ..
Differentiable Structures 55 13.I)dimen sional (differentiable) submanifold. and then repeating this construction indefinitely. then S boundary but U is not a 2dimensional manifoldwithboundary. (a) There is a map f : ]R2 .. (It is well to bear in mind the following example: = {x E ]R" : d(x.. 0) for all x. V = V. (a) Let C be an open set such that boundary is an (n . ]R2 such that (I) f(x. 0. y) c 1HJ 2 for y :0: 0. but au # boundary (b) Consider the figure shown below. then U is a manifoldwithboundary.S. Show that U is an IIdimensional manifold withboundary. O) (2) f(x. (3) f(x.O) < 2). so part (a) is true only for differentiable submanifolds.y) C ]R 2 _ 1HJ 2 for y = (x. This figure may be extended by putting V Mn V. The closure S of the final resulting figure is known as Alexander� Homed S plwe. O) < I or I < d(x. Show that S is homeomorphic to S 2 (Hint: The additional points in the closure are homeomorphic to the Cantor seLl If is the unbounded component of ]R3 .) V if V smaller copies of the two parts of S' into the regions indicated by arrows. < .. 14.
O) Mp aM. on P such that the inclusions of and are Coo and such that the map from V to x ( . aMfJ: .I . (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. (b) Suppose and are Coo manifoldswithboundary and is a diff eomorphism. (c) Now suppose that there is a neighborhood of in and a diffeo morphism x [0. x x f define a Coo structure on P. such that = for alI E and a similar diffeomorphism V x [0. 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. I). fJ). I)..V U fllVJR" aM xi VV aM. I) induced by and is a diff eomorphism. . V U aM M a N fJ (a.56 (4) Chapter 2 restricted to the upper halfplane or the lower halfplane is Coo. Show that the resulting Coo manifolds are diffeomorphic. V) f(p).. (x. f(V aM) aN. but that the diff eomorphism ca"not be chosen arbitrarily close to the identity map. Show that this procedure does not f M N M UJ N f: aM .. (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. . Show that there is a unique Coo structure a: V . aN M N p aM x aM f(x) aN. nn2 JR ann22. aN aM a(p) V (p. . .
. y2) yl(a. . a2J. . . where each Vi is of the form < E I}  onto ° on Let Vi . I = �A I L a2i i=l 18. = X .Differentiable Structures 57 15. I] such that C 11 (0) and C 11 (I). 2.I 17. I: JR JR I(x) = { �1/x x > 0o x > CooCoo. > aJ. Consider the coordinate system (yl.I /1xl could be used just as well). Coo I C = ({a JR": lal ) x MJ.} be a countable cover of V.: M will be called mixed partials of ai = Show that sup of all mixed partials of Ji . Ji of all orders :::: i.b) = a y2 (a. .  a function with 11 (0). 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. Let is Coo. .b) = a b. 2 Coo > x C1 I: M ° for < 0. Show that the function Coo Coo Coo M NM N Coo is (the formula e. once you 11 (0) and know the trick. (a) Find a structure on 1HI1 x 1HI1 which makes the inclusion into JR2 a map. Actually.C. for any closed C (b) Let {V. C = and 11 (0). axj ' axj axk ' J. I for some coordinate system taking an open subset of > [0. + for JR2 defined by . . we can even find with = 11 (1).. The proof turns out to be quite easy. C M M. . of order 1 . defined by :s (a) It suffices to find. Coo J. Cl = C1 C2 CM. construct a structure on x sucl1 that all the "slice maps" (defined in Problem 8) are 16.1 /x' is used just to get a function which is and e. I] be a function with ° on Vi and Functions like J. =C M JR". Can the inclusion be an imbedding? Are the projections on each factor maps? (b) If and are manifoldswithboundary.
. a2/ax2 f: Nm Cl aI. . 1] [0. and 0 (provided [0 2 1 .. then f(M) has measure that M has only countably many components). in terms yl 19. I .1] 1. . a subdivision of . (b) Also compute it from Proposition 3 (to find of and y 2) . . The numbering is determined by the following conditions: [0 . (a) The lower left square is A n . A n .. square A n . is and m > II. 1] (d) Squares An •41+ 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 . x into 22n squares. I] [0. write each 1. /ayj .2. (b) The upper left square is A n• 22n. I .. Compute the "Laplacian" = 1 1 . A n •41+3) An •41+4 are contained in A n l .: (rf. Answer: then compute 20.l+ l . f: [0. �[f.:) + fe (Hii ) ]' a/ax in terms of alar alae.b) yI I y af/ayl af/aI /. The following pictures show. k +l have a common side. for II = and 3. A n . If M n .58 Chapter 2 (a) Compute from the definition. A n •41 +2 . 1] by the condition kf(l) An. k and An . the operator alai depends in terms of polar coordinates. . Notice that # on and i. (c) Squares A n.k is labeled simply k. onto [0.. 1] . 2"'. (First compute from this). not just on even though af/ayl (a. [0. and not oneone. 1 ].k for all 2""2'1I � I :S 2k2n · Show that f is continuous..
Coo if {(x. 1 ] '"' B = f(l) (a) Show that > measure 0 if the shaded triangles are chosen correctly.) . . V) on M such that M. Show that the image of a rectifiable curve has measure O. i=1 k C 24. Show that g is oneone. x E ]R) is not the image of any immersion of ]R into ]R2 ]Rk is open and f.C(li. dinates. P) = The curve is rectifiable if {i(c. "Ld(C(li). define > f is uniformly continuous. and that its image will not have ]Rn be continuous. Let c . P» )). a semicircle with diameter the line segment AB. a set M. . can be made into a closed submanif if and only such old coordinate systems exist around every point of M. C M can be made into a kdimen sional submanifold of M if and only if around each point in M. Id of i(c. (b) Every submanifold of ]Rn is locally of tllis form. 1]. You will need . (b) The subset M. below the image of g. [0. . For p/2n E [0. (Neither Theorem 9 nor l O is quite strong enough. P » ) is bounded above (with length equal to sup{i(c . [0. V > ]Rnk is then the graph of f = { (p. A = f(O)�"" g . . after renumbering coor 25. there is a coordinate system (x. define f(p/2n) E 59 ]R2 as shown below.Differentiable Structures 22. For each partition P "" {Io. The set 26. (a) If V C Coo. Ixl) . 1]. What does the inside of this curve look like? 23. (a) If M is a manifold. . (b) Consider the homeomorphic image of S ' obtained by adding. f(p» E ]R" . 1] [0. n V = {p : x k+ ' (p) = . p E V) is a submanifold of ]Rn. = xn (p) = 0). . so that it has a continuous extension ]R2.
29. 1R.. 1R 3 be the map g([x . then I is onto. A continuous function I: . Z]) = (yz. there is a proper map I : M . Geometry and the Imagination. Theorem l O is es sentially Theorem 213 of Calculus on Manifolds. has constant rank k on a neigh borhood of I'(Y).k (or is empty).. N (a) L(f) = 0 if and only if I is proper. X 3 ..) 27... Show that g fails to be an immersion at 6 points (the image points are the points at distance ± 1 /2 on each axis). (f) If M is a manifold. Prove Proposition 12: If I : Mn . Let I : IPz .. (c) Prove the Shrinking Lemma when (9 is a (not necessarily countable) point finite cover of any space... Y is proper if II is compact for every compact C Y.. The limit set L (f) of I is the set of all y E Y such that y = lim I(xn) for some sequence x" X . I] is in only finitely many members of the cover. but L(f) # 0. (a) An immersion from one nmanifold to another is an open map (the image of an open set is open). (e) A submanifold M.... There is a way of immersing IPz in IR 3 . Y is a homeomorphism (onto its image) if and only if L(f) n I(y) = 0... • • • E with no convergent z subsequence. consider collections e . See Hilbert and CohnVossen... 41). 30. .. the function I can be made if M is a manif old. whose image is the Steiner surface.. comparison with the implicit function theorem will show how some information has been allowed to escape. and I : M . C M is a closed submanifold if and only if the inclusion map i : M.. then II (y) is a (closed) submanifold of M of dimension n . N N N 28. (b) I( ) c Y is closed if and only if L(f) C (c) There is a continuous I: IR . C X (C) X X X f(X). M is proper. pp. 3 1 7321.. (a) Find a cover of [0. known as Boy's Surface. y.. is an immersion.. pg. (You will need Zorn's Lemma. I] which is not locally finite but which is "point finite": every point of [0.. (d) A oneone continuous function I: .xz. Coo Coo 31. (b) If M and are nmanifolds with M compact and connected. IRz with I(IR) closed.60 Chapter 2 the implicit function theorem (Calculus on Manifolds. xy) defined in Chapter I.. (b) Prove the Shrinking Lemma when the cover (9 is pointfinite and countable (notice that localfiniteness is not really used).
Coo ( a1/lU (A) ax kl ) has the same rank as ( a1/lij I axkl ( ) ). and that 1/1 0 RA = 1/1 for all A E O(n). (e) Using the formula 1/Iij (A) = L. 1R). 1R). (a) The set of all nonsingular n x /1 matrices with real entries is called GL(/1. together with all other U E 19 covers the space. U' C U. 33. = (0. IR). . and use partitions of unity). define RA : GL(II. pg. The special linear group SL(n . in fact. 1R) > (symmetric matrices) by 1/I(A) = A . Using the formula for D(det) in Calculus on Manifolds . > IR is then there is a function J: M > IR with J = f on M" and with support J c U. f has no continuous extension to a map from ]R2 to but the proof requires some topology. However. M. (Here xkl are the coordinate functions in IRn 2 .) Coo. 1R) by RA(B) = BA . (c) Show that O(n) is compact.IR) is called the orthogonal group 0(11). f can always be extended to a function in a neighborhood of M. aik ajk k (A = (a/j »). and f : M. A" where At is the transpose of A. The subgroup 1/1'(1) of GL(II. and 1/Iij the n(l1+ 1)/2 component functions of 1/1. (d) For any A E GL(/1. S I . 1R) > GL(n . the general linear group. (b) This is f alse if M = IR and M. show that for A E O(n ) the matrix 32. (extend locally. = = and f = identity.) Conclude from Proposition 12 that O(n) is a submanifold of GL(n. 1R) of dimension n 2 . Coo N. (a) If M. is any neighborhood. (c) This is false if IR is replaced by a disconnected manifold of pairs (U. Show that RA is a diffeomorphism. U ') where U E 19 . is the subgroup of all matrices with det = I. 24. Show that A E O(n ) ifand only if the rows [or columns] of A are orthonormal. in this case. Coo N S'. It is a manifold.I.Differentiable Structures 61 Remark: It is also false if M = 1R2 . since it is an open subset of IRn'. U :J M. By applying the chain rule. 1R) is a closed submanifold of GL(n . I). U') E e. IR). (b) The symmetric n x n matrices may be thought of as IRn (n + ' ) /2 Define 1/1 : GL(n . C M is a closed submanifold. or unimodular group. show that S L(n . and the union of all U' for (U .
Hint: If h denotes the k x k identity matrix. n .) Conclude that O(n) has dimension n (n .Ao are < e . x 1 aj l au k=i#J k=j#i k =i=j 2aj/ o otherwise. and M(m . n matrices. (f) Show that det A = ±I for all A E O(n).:s m.n) is a submanifold of dimension k(m + n . n .(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). .1)/2. k) C M(m. 1R) is called the special orthogonal group SO(n). where the entries of A . n) denote the set of all m of all m x n matrices of rank k .Ao are < e. l1. 34. then ( � �) (d) M(m .k) for all k . Let M(m . The group O(n)n SL(n. then X has rank k if and only i f D = CAJ B. or the rotation group R(n). n. k ) there are permutation matrices P and Q such that PXoQ = ( AO Co BO Do (b) There is some e > 0 such that A is nonsingular whenever all entries of A . k) the set (a) For every Xo E M(m . (c) If PXQ = ) ' where Ao is k x k and nonsingular.62 Chapter 2 show that a 1fr ij .
suppose c: JR > JR" is a differentiable curve. c"'(t» is just a point of JR". . crt) + c'(t ) 63 . 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. . but the line between crt) and c(t) + c'(t) is tangent to the curve. 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.
This picture gives rise to some TJRn JRn _ _ ]( '_ terminologywe call ](1 (p) the fibre over l  p p. the "tangent space of 1R"". it is more convenient to denote a member of TIR" by a single Jetter. we simply describe the "arrow" from p to p + V by the pair (p.. b) = a. For any tangent vector v.1 (p) may be pictured as all arrows starting at p. like v. v). Alternately. At times... which we will also denote by TIR". This fibre can be made into a . The set of all such pairs is just 1R" x 1R". we will denote the set of all (p. the point Jr ( v) is "where it's at". 1R" by ]( a . in conformity with this notation. To recover the first member of a pair v E TIR". the one visualizable case occurring when 11 = 1. elements of TIR" are called "tangent vectors" of 1R". we define the "projection" map 1r: 1R" x JR" .64 Chapter 3 To give this picture mathematical substance. v) E TIR" by vp ("the vector V at p"). it can be pictured more geometJ·ically as a particular subset of 1R" x 1R". We will often denote (p. v) for v E 1R" by 1R"p . The set ]( .
but. v) (j) (p.a · v). Since f (v) is defined to be a vector E JRmJ( ) . Usually we will just use ordinary + and · instead of If f : JR" > JRm is a differentiable map. 1( 0 f. W) = (p. whose apparently anomalous features will soon b e justified. D (g 0 f)(p) = Dg(f(p» 0 Df(p). has the map f. . Suppose that g : JRm > JRk is another differentiable function. denotes the map f : TJRn > TJRm which is the . is de noted by f . peJRfI x (j) and 0 (p. x respectively. the symbol f. and p E JRn. built into it. should really b e thought o fa s defined on and JR TJRn. g. it can be written . however.The Tangent Bundle 65 vector space in an obvious way: we define (The operations U 1( . v + w) a o (p. as well as all maps Df(p). using (I). J. .p union of all f. . Thus.' (p) 1( . = f 0 1(. the follow . This is not the only reason for defining f in this particular way. so that. ( [ Df(p)(v)lJ(p) } = (Dg(f(p)}(Df(p)(v» }g(f(p))' This looks horribly complicated. 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.p p p ing diagram "commutes" (the two possible compositions from TlRn to jRm are equal). (I) By our definition. v) = (p. and 0 . by the chain rule.' (p).
.(I. in terms of I. with our present definition of I. The tangent . this vector lies along the tangent line to the graph of 1 at (1. also. 0 I. Henceforth. it is merely an elegant restatement of the chain rule... I. The "tangent vector" of a curve c : JR .. 1(1 )) for I: JR . like rank or singularity.. JRm is differentiable. . . c'(I)c(I) E JR"C(I) ' [If c happens t o be of the form c(l ) = (1.. J ) is the "unit" tangent vector of 1R at t . . . rather than Df.] Notice that the tangent vector of c at 1 is the same as where I t = (t . JR : 1'(1) then C'(I)C(I) = (I . then g 0 c is a curve in JRm.) = [DC(I)(I)]c(l) = (c" (I). = (g 0 fl •. c"'(I» c(I) . The tangent vector of c at 1 may be defined as g.(vp) were not in JRmJ(p) .66 Chapter 3 thus we have This relation would clearly fall apart completely if I.. /'(I» c(I)... JR" can be defined in terms of this concept. If g: JR" . c... we will state almost all concepts about Jacobian matrices. /(1 » .
The Tangent Bundle 67 vector of g 0 C at t is (g 0 C).(I. This subspace doesn't depend on the coordinate system x) for if y is another coordinate system.I ). 0 (x 0 y ... and is an isomorphism (with inverse (y o x .(I. = (i o x .» v crt) = g. Consider now an ndimensional manifold M and an imbedding i : M .) = g.) ..I o x o y.. (lRnx(p) is an ndimen sional subspace of IR N . (i 0 x .I ) .I ) .. then (i o y I ). Then i 0 X .X(p) .(c.I ) . (x 0 y. IR N .(p). I = (i 0 x .I is a map from IRn to IR N with rank n. IR x( p) . (tangent vector of c at I) .I ) � �� � �. Suppose we take a coordinate system (x. n n 'y(p) : IR y(p) . U ) around p.. Consequently.
i)p. i)p con n(v) = p if V E (M. We can define a "projection" map n : T(M.(M). We will denote this ndimensional subspace by (M. 0 x . The Curve c(e) = (cos e. i)p' We now want to look at the (disjoint) union T(M. sin e) E S ' . i)p C i(M) x jR N C TjR N .i) = p EM U (M. let up = (. then a = . . Thus. Consider first the manifold M = S ' and the inclusion i : S ' > jR 2 . our ndimensional subspace is just the set of all tangent vectors at i(p) to differentiable curves in i(M).sine) passes through every point of S ' .I 0 c is a curve in jR N which lies in .i) > M by As in the case of TlRn .I (p) has a vector space structure also. each "fibre" 7f .68 Chapter 3 There is another way to see this. and c'(e) = (sine. which justifies the picture we have drawn. 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. Beyond this we have to look a little more carefully at some specific examples. For each p = (cos e. Now so the tangent vector of every a is in (i o x.sin e. e) > jRn is a curve with c(O) = x(p). cos e) # O. Then (S ' . If c : (e. cose)p (it clearly doesn't matter which of the infinitely many possible e's we choose). and every diff erentiable curve in i(M) is of this form (Proof?).J ) * (lRnX(p) ) ' Moreover.
b) a] = T(S2. would be a collection of nonzero tangent vectors. restricted to a fibre is a linear isomorphism on to the image.). . If we define the "fibres" of to be the sets (p). It is a wellknown (hard) theorem of topology that this is impossible (you can't comb the hair on a sphere). ).O in the set of vectors T(S'.The Tangent Bundle 69 sists of all multiples of the vector up . then. . . then each fibre has a vector space structure in a natural way. up = (p.S ' x IR' � S. 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. S2 fi. We Can therefore define a homeomorphism I. S2. S' x ( ) T(S ' . i) .. SS22 2 ]R2. which makes the following dia· [rr'(a. . ft rr' rr'' 1R ' by fi ). clearly I. i) fi .. for a fixed vector v #. Commutativity of the diagram means that fi takes fibres into fibres. one at each point of which varied continuously. : i) gram commute.
0) ).0) . This function would have to be nonzero everywhere and also satisfy A(21r) == A(O). 0): Ifwe could. shown * as dashed arrows in the picture.. 21r I r�� is a tangent vector. then each vector would be J (2. 1 )(8. 0 . The impossibility of choosing nonzero dashed vectors continuously clearly shows that there is no way to map T(M.  .0. The same is true for all multiples of 1 ( 0. f).1) = (2 cos8 + 1 cos � cos8. At each point p = (2cos 8. 1 )(2rr.0) (2.0. 0) This means that we can never pick nonzero dashed vectors continuouslY on the set of all points (2cos 8 . 1 ) (0. 2 sin8.M is the image of the map I: [0. The map f will just be the inclusion M > 1R 3 where M is a Mobius strip. 0) i .0) = ( .0) ) . for some continuous function A : [0. 0)(0. which it can't (by an easy theorem of topology). the particular subset of IR 3 defined in Chapter 1.0) . %2. 0) of M.70 Chapter 3 There is another example where we can prove that no appropriate homeo morphism T(M. i) > M X 1R 2 exists. 21r] > 1R. 2 sin 8 + 1 cos � sin 8. 2 sin 8. 1 ( 0. the vector (8. without appealing to a hard theorem of topology. to be precise. 0)(2. 1 sin n .0) ) = * (21r. 21r] x (I.0. = while = [� (O.0. 0 . fibre by fibre. 0) (2.O)J [ aal l [ DI(O.0) = ( 1 .0.1 . I) > 1R 3 defined by 1(8. Notice that 1 ( 0.
. B .. we can therefore define by f: f (mp(vx(p) ) n:I (V) . � = (E. In standard jargon. the part of T(M. the fibre where the abbreviation mp has been introduced temporarily. B is a continuous map onto B... i) is always simple locally: if (x. E. homeomorphically onto V x JR". .. (3) Ell and 0 are maps Ell : p EB U n:I (p) x n:I (p) . V = x JR" (p..The Tangent Bundle 71 homeomorphically onto M x JR2• We thus have another case where T(M. the structure of T(M. In fact. i) over V. however. 0: JR x E . Ell .. (2) n: : E ..... n:. for each p E V. This additional feature qualifies T(M. i) does not "look like" a product M x JR". JRN.. for all q E U. there is a neighborhood V of p and a homeomor phism I : n:I (V) . with Ell (n:I (p) x n:I (p») c n:I (p) and 0 (JR x n:I (p») c n:I (p). can always be mapped. For any imbedding i: M . 0) . fibre by fibre.. T(M. E. then n: . v) .. such that the following "local triviality" condition is satisfied: For each p E B. V) is a coordinate system on M....I (V). which make each fibre n:I (p) into an ndimensional vector space over lR.. 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).i) is "locally trivial". V x JR" which is a vector space isomorphism from each n:I (q) onto q x JR".
v or av. i) considered before is equivalent to . we will denote (j)(v. This is called the trivial Ilplane bundle over X and will be denoted by . (j). i») can be . 0) automatically gives rise to a bundle �IA over any subset A C B. and the obvious vector space structure on each fibre.n(B) is called trivial. for M can be obtained !i·om [0.. i) and T(M.n(x).. w) and 0(a. v) by v + w. i) are not trivial. but there is an even simpler example of a nontrivial bundle. > Ez which takes each fibre ". and we shall usually refer simply to a bundle . For vectors v. or even denote the bundle by E alone. The map h is called an equivalence. (The local triviality condition for a bundle � just says that �IU is trivial for some neighborhood U of p. Equivalence is here a technical term: Two vector bundles �.'(S')."nJ1sidered as a Idimensional vector bundle over S l .n(IRn). . respectively... B. a).' (p) and a E 1R. I] x IR by identifying (O. and a . Notation as cumbersome as all this invites abuse. while S' can be obtained from {O. W E . The bundle T(S ' .72 Chapter 3 Because this local triviality condition really is a local condition. . each bundle � = (E. : E. . to be precise. : X x 1R" > X the projection on the first factor. The "tangent bundle" TlRn is just . The Mobius strip itself (not T(M. = " . . > B and �z = "z : Ez > B are equivalent (�. The simplest example of an Ilplane bundle is juSt X x IRn with .a) with ( I .' (p) isomorphically onto "z' (p). I } .) The bundles T(S z .. :::: �z) if there is a homeomorphism h : E. : E > B. A bundle equivalent to .
An equivalence is obviously the analogue of an isomorphism. > If M" c are submanifolds. I] > with S(O) = S(I). fl is a bundle map fi'om . s(8) E 1(1 (8) must be the ° vector for some 8 E S'). f: jRk jRjRll . EI BI � Bz . I }. Since s must be ° somewhere. (I. fl.I U(p)) is a linear map. Thus f may be restricted to be an isomorphism on fibres and f to be the identity. the section s must be ° somewhere (that is. the map 1( is defined by 1( I. (2) j : 1(. Z jRk k l jRl TjR TjR jRk * There are actually several possible choices. The diagram above illustrates local triviality near the point {O. and the map f satisfies f(M) C N. a). then f . such that (I) the following diagram commutes jR 1(.The Tangent Bundle 73 [0. or a fixed base space with varying bundles. This surely shows that M is not a trivial bundle. 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). > Ez and f : B. ' (p) 1 � Ez l1(Z to for any differentiable The pair U . a)}) = {O.a) = I for 0 < I < I and 1( {(0. with j : E. depending on whether one is consider ing all bundles at once. > 1( . to �z is a pair of continuous maps (j. > Bz. fixed bundles over various spaces. The relations between some of these cases are considered in the problems. The analogue of a homomorphism is the following. . i : 'M > and Nm C and j : N > are the inclusions. I] by identifying ° with I .' A bundle map from �. or a homeomorphism. Such a map corresponds to a continuous function s : [0.
Elements of are of the form 0 ). In this way we obtain a bundle f o e in and consequently f. (v) E map from to Actually.i) TM.i) 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. just remember that v E is the tangent vector of a curve c in so f (v) is the tangent vector of the curve . (w) T(M.i(N) M N i T(M. M N.j). T(M.i)p T(M. this construction could be generalized much further. (M. to the case where and are merely imbeddings of two abstract manifolds and and f : > is Coo. then (M. T(M. rather than one bundle.j)p x. while and are two imbeddings of in and respectively. and obtain an equivalence from to In other words. while elements of If we map T(M. That is the T(M. In fact. The map > induced on fibres is independent of the coordinate system for if (y. i)p j (i xI M for WE WE IRnx( IRnx(p)' p). the dependence of on is al most illusory. N. i) T(N.i)I V T(M.I : i(M) N. V) is another coordinate system. IRk IRk i j M i. j)p are of the form (j 0 x 1). we could abbreviate to if we agreed that really denotes an equivalence class of bundles.j)IV. j).i) T(N. T(N. T(M. M. T(M..i)p k MIRIRI.74 Chapter 3 takes to to see this. (w) for we obtain a bundle map from to which is obviously an equivalence. j).j). it would have sufficed to begin with a Coo function f : > since f can be extended to locally. i) TMi . I I We can therefore put all these maps together.
: TM > TM is the identity. The identity map I of U to itself and the inclusion map i : U + M have to be considered as different. What we would like to do is to get a single bundle for each M. (2) There are equivalences In : TJRn > en (JRn) such that for every function f : JRn > JRm the following commutes.. (I) If 1 : M Coo TJRn � TJRm 1m In en (JRn) old f. sub terfuge. We will obtain a single bundle for TM. and it is undoubtedly ugly. 0 f .. _ _ _ . which has all the properties any one of these particular bundles T(M. f). en (JRn»). N) where f : M > N. but the elements of TM will each be large equivalence classes. in some natural way. It is possible to assign to each nmanifold M an nplane bun dle TM over M.. If • g : N > P. : TU > TN is just the restriction of f•. such that: Coo > M is the identity. . When we do. though quite natural. : TU > TM are certainly different (they map TU into two different sets).The T angent Bundle 75 sort of thing an algebraist might do. since the maps 1 * : TV + TV and i. TN PROOF The construction of TM is an ingenious. * vVhen using the notation h. THEOREM. and so will f for f : JRn > JRm. M. we can. and for f: M > N the map (flU). em (JRm) 1 1 (3) If U c M is an open submanifold. �TM/f.' TU � TM)IU/c ( _ _ _ '" i. More precisely. then TU is equivalent to (TM)IU. precisely we will write I'old J/' when necessary.. ea . I . then (g 0 f) . and to each map f : M > N a bundle map (f. it must be understood that the symbol "/" r lly refers to a triple (f. TM TU (flU). there is an equivalence TU :::: (TM)IU such that the following diagrams commute.. Can we do this? Yes. TJRn will be different from our old defini tion (namely.. = g. . where i : U > M is the inclusion. . i) has. then 1. so in stating our result .
It is the clue which enables us to now define TM. We will denote the fibre .. v]q � (q.... and (y 0 xI).I (p) isomorphicaIly onto {p) x If y is another coordinate system. w) for some w E We can easily figure out what the relationship between v and w would be. to some (x(p). 0 x. v) is taken to (y(p).I (p) by the formulas [x. These equivalence classes will be called tangent vectors at p. w) by y. v). v) p (y. W E we define if (a) is satisfied. This condition makes perfect sense without any mention of bundles. though TMp might be better. (x. in conformity with the notation p.. we would have if JRn. JRn . since x. v]p = [x. (a) It is easy to check (using the chain rule) that p is an equivalence relation. since (x(p).I (p) would be taken by x. is supposed to be the old (y 0 XI). namely [x. If x and y are coordinate systems whose domains contain p. Then (x. Here v is just an element of (and every v would occur.I (p) by Mp. JRn. JRn JRn). v) will be denoted by [x. We want this to be a homeomorphism. and this would be an equivalence (with inverse (xI). v]p. w) w = D(y o x.. Since TV should be essentially (TM)IV. I = (y 0 xI).). then y. : TM > M.I ) (x (p»)(v). There is a metric with exactly these sets as open sets. IRn. v).. : TV > T(x(V» . I (V) > V x JRn. we would have a map x. maps . V) is a coordinate system. w]p = [x. v]p. and TM is defined to be the set of all tangent vectors at all points p E M.76 Chapter 3 The construction is much easier to understand if we first imagine that we al rearfy had our bundles TM. We now have a bundle . We define a vector space structure on . this definition is independent of the particular coordinate system x or y. and T(x(U» should essentially be x(V) x a point e E . so we leave this one part of the proof to Problem l.. so we want I (A) to be open for every open A C V x and thus we want any union of such sets to be open. v + w]p a · [x. a . v]p + [x. but it is a little ticklish to produce.(e) would be (y(p). the map " takes p equivalence classes to p. and v. If JRn. the equivalence class of (x. because D(y o xI)(x(p» is an isomorphism from to Our definition of TM provides a oneone onto map (b) JRn JRn . Ix Ix : .
mapping to have the same derivative at O. i). with c (o) = p. we define (c) N. If (x . In fact. V) are coordinate systems around p and I(p) . we next ask how fortuitous this was. we consider curves c : (e. IRn Ix Henceforth. . and I is the identity coordinate system of then If . if (x. [y. In fact. i ). D (y 0 I 0 xJ )(x(p») (v)]J(p)' I. will be the elements of our new bun dle.. V) and respectively. ([x . Condition (I) of our theorem is obvious. the only difference is that each equivalence class for M contains some extra members. T'M. IRk i. Condition (3) is pI"actically obvious also. V]p) = [I. since in M there are more coordinate systems around p than there are in U C M. e) > M. it is trivial. : M IRk . •:. then TM is equivalent to T(M. the fibre of TV over p E V is almost exactly the same as the fibre of TM over p. In II. though perhaps confusing to the novice. > is an imbedding. To prove (2). For I : M > there is a map h taking the '1 equivalence class N x 0 CI and x a C2. Having succeeded in producing a bundle over each M. it must b e checked that this definition is independent o f x and y (the chain rule again). the bundle 1C : TM > M will be called the tangent bundle of M. we define to be where I is the identity map of and is defined in (b). But T(M. ( i ( p) . v]p) = O f course. and we proceed to define two different such bundles. V) is a coordinate system around p. D (i 0 x ')(x( p») (v)) E ( M . (y. i) will play no further role in this storythe abstract substitute TM will always be used instead. V) is a coordinate system around p. Can one find other bundles with the same properties? The answer is yes. In II I = by (c) CJ '1 C2 if and only if: The equivalence classes. we define In. i lp . For the first example. each defined on some interval around 0. IR IRn.([x . which is equivalent to T(M. to prove commutativity of the diagram. D(i 0 x J ) (x( p»)(V)] i (p) the composition is easily seen to be an equivalence. for all p E M.The Tangent Bundle 77 I : M > and (x.
f(O). then O = i(f . Coo i (fg) = J(p)i(g) + g (p)i(f).xn) for x E U. Then there are Coo functions gi Coo function in a convex open neighborhood U of 0 : U > with JR . y' (0) = V. . . vlp corresponds to: the 'J' equivalence class of where is a curve in with y JRn under this correspondence. . with support h C JI (0).. Then n J(x) = J(x) . xi dt. . There is a function h : M > with h (p) = 1 and support h C (0). • JRn . we may use this trick to define i (f): choose h to be 1 on a neighborhood of p. (The second condition actually follows from the first. 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. .) of c to the J equivalence class of J 0 c.xn) = L:7= l xigi (X I . Thus. let hx(t) J(tx). Without bothering to check details. we can already see that this example is "really the same" as TM [x. For. In the second example. clearly i (f) = i (g) if J = g in a neighborhood of p.g) = i (f) . suppose that J = 0 in a neighborhood of p. We have already seen that the operators i = have this property. things are not so simple. with J(O) = O. r i=1 Therefore we can let g (x) = J� Di J(tx) dt :PROOF. and define i(f) as i(fh). .. This comes out of the following. since U is convex. The set of all such operators is a vector space. if J = g in a neighborhood of O. If J is defined only in a neighborhood of p. this is defined for 0 :s t I. LEMMA. Then a/axi lp JI Coo JR 0 = i (O) = i (fh) = J(O)i(h) + h(O)i(f) = 0 + i(f). (2) gi (O) = D. 2. but it is not a priori clear what its dimension is.) For x E U. Let J be a in (I) J(x l .78 Chapter 3 (/.i(g). h corresponds to J•.J(O) = Jo I hx'(t)dt = l0 I L Di J(tX) . For these operators. This condition is actually true for any derivation i. = :s .
so i(l) = o. then Mn span this vector space. Notice that i(l) = i(l .. a]p. they are clearly linearly inde pendent. Given / on U. Then (Ii denotes the i(f) = i(f . The set of all linear derivations at p E is an ndimen sional vector space. and any derivation ./(0). for the function / . Assume U is convex./(0)) = e(£) igi ) coordinate function) 1= 1 = I )i (Ii )gi (0) + I i (O)i(gi)] 1= 1 ith PROOF. THEOREM. This shows that span the vector space. if (x. . From Theorem 3 we can see that. once again. It is a simple exercise to use the coordinate system to transfer this result from to . 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. Consider the case where M = JRn and p = o. [x. l) = l · i(I) + l · i(l). Hence i(c) = c . a�n Ip (so i is determined by the numbers i(xi »). U) is a coordinate system around p.e can be written a�I Ip . In fact. .The Tangent Bundle 79 3. •:.. choose gi as in Lemma 2. i(l) = 0 for any constant function c on U. .
.80 derived in Chapter 2.a]p most easily analyzed from the relation (f. 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. is often j . and the map f.a) 1=1 axi n a L1 a1. � . If e : JR M is a differentiable curve. shows that Chapter 3 if and only if bJ = L. the map which corresponds to can be defined as follows: (x. that is. In fact. . (i)](g) = i(g fl. we wilJ not hesitate to write v(f) for a differentiable function f defined in a neighborhood of p. and to is calJed the tangent vector to c at 10. en (JRn) . between E v Mp x ap [f.. We wilJ denote it by the suggestive symbol dt 10 de l · . ayj a' (p).IP 1= ax consequently. I [x. t i . b ) .v)(g) = v(g fl. It is easily checked that under this correspondence.. and this is precisely the equation which says that p (y . TJRn 0 f. JRn. 0 It is customary to denote the identity coordinate system on write JRI by this is a basis for JRlo.. then > dt 10 d I for I.=1 a a!. a tangent vector is often most easily described by telling what derivation it corresponds to.
.xn (Jr(v)). I TM. Since locally looks like x clearly is itself a manifold. an ) E ]R2n . L:7=1 ai Di(yl xI )(I). 0 = 0 0 0 then. I n .a) (y ox .x). 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 (v)) E ]R2n is a homeomorphism from (TM)IV to x(V) ]Rn .I (l. . .a l . . . as we have already seen. (x. 1=1 1=1 This shows that if (I. V ]Rn. and Let us denote x x p Jr( ).). = � a' ay}i (p) = � a'. . oxI )(x(p)). L.. but it is a little delicate to state. 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. Then the map v � (x l (Jr(v» . . If (y. . V) is another coordinate system. The tangent bundle of a manifold has a little more structure than an arbitrary nplane bundle. . Di (y1.I (I). moreover. .a) = (I I . TM Coo(TM) I V v a i by Xi (v) . ax L. This map. when we identifY TV with V ]Rn in the standard way.I is Coo. . so it has been quarantined in an Addendum to this chapter. L:7=1 ai Di (yn x. then y. . As you might well expect. This expression shows that y. TM TM. there is. . x: V ]Rn CooTM M.. is simply the map x. .The Tangent Bundle 81 This symbol will be subjected to the standard abuses one finds (unexplained) in calculus textbooks: the symbol de the subscript now denoting a particular number E as well as the identity coordinate system. . = V 0 b1. . (x.xl(v)..). . and quite a mess to prove. (x Jr.I )(t)).
we define the vector field fX by > (X + Y)(p) = X(p) + Y (p) . . s: M there is no vector field on S2 which is everywhere nonzero. the local trivializations x* are Coo. we define a new vector field TM is contin X+Y by Similarly. TM for all pE U. We have shown that there do not exist two vector fields on the Mobius strip which are everywhere linearly independent. It follows that 1C : E > B is Coo. or Z. then for any coordinate system (x. M TM M. if f: M JR. Y. With this Coo structure. we have Mp) .82 Chapter 3 We thus have a collection of COOrelated charts on which can be ex tended to a maximal atlas. In general. 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). 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. If X and Y are two vector fields. IRn. in some If we think of as the set of derivations. (fX ) (p) = f(p)X(p). Recall that a section of a bundle 1C : E > B is a continuous function B > E such that 1C o S = identity of B. The functions ai are continuous or Coo if and only if X: U > uous or Coo. 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. for Coo vector bundles we can also speak of Coo sections..
Let !F denote the set of all Coo functions on M. we clearly must define Xp(f) = A (f)(p) . X(f + g) = X (f) + X(g) X(fg) = IX(g) + gX(f). Y. and the operator Xp thus defined is a derivation at p . indeed. In fact. and I are Coo. thus X is a "derivation" of the ring !F. Clearly. then X (f) is Coo for every Coo function I. iflocally n a X(p) = L a'. a Coo vector field X is identified with the derivation X . then A = X for a unique Coo vector field X. if X (f) is Coo for every Coo function I. It is not hard to check that if X is a Coo vector field. Of ten. then we can define X (f): M IR by letting X operate on I at each point: X(f)(p) = Xp(f). On U we can write a/axi now denoting the vector field the symbol I: M IR is a Coo function. and X is a vector field. If > a newf unction > > > . (p) axi I ' P 1= 1 then . We have just seen that a Coo vector field X gives rise to a function X: !F !F.The Tangent Bundle 83 Clearly X + Y and IX are Coo if X. then X is a Coo vector field (since X (xi ) = a i l. al X(f) = � a ' axi' 1= ) which is a sum of products of Coo functions.L. Conversely. The reason for this is that if A : !F !F is any deriva tion.
. x n . . . Vn) E JL if and only if . Either of these two equivalence classes is called an orientation for V. There is no way to pass continuously between these two groups: if we identify linear maps IRn 7 IRn with n x n matrices. [VI .l . vn) belongs will be denoted by . V (VI. . . A simple example of the latter is the map I : IRn > IRn defined by I (x ) = (x l . . Vii L Qji Vj. 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. . . . the matrix A = (aij) of I is given by the equations V. For the present. . . .) = v'. . v'n) equally oriented if det A > 0 (i. The relation of being equally oriented is clearly an equivalence relation. The problem becomes more acute if we want to define orientation preserving isomorphisms between two diff erent (but isomorphic) vector spaces V and W.. by Robert A. vn] . . . this clearly makes no sense unless we supply V and W with more structure. . vn] = JL. we will discuss just one new concept about manifolds. . if I is orientation preserving) and oppositely oriented if det A < o. since we have not yet defined anything called "orientation". . vn) and (V'l . and those with det I < 0. . purportedly as a word from the Martian language. so that if JL is an orientation of V. in particular on the tangen t bundle of a manifold. we note that two ordered bases . . . 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). . . . Vn) and ( 'l . linear transformations in the first group are called orientation preserving and the others are called orientation reversing. _ xn ) (reflection in the hyperplane xn = 0).e. 'n ) for V determine an isomorphism I : V > V with I(v. then . .84 Chapter 3 The tangent bundle is the true beginning of the study of differentiable mani folds. If JL denotes one (VI. which is being preserved. j=J n = * A cult word of the sixties) "grok" was coined. . The terminology "orientation preserving" is a bit strange. We call . [VI . Its sense is nicely conveyed by the definition in The American Heritage DictiollaT ': } "To understand profoundly through intuition or empathy". . which arises in this very way from the notion of "orientation" in a vector space. those with det I > 0. The nonsingular linear maps I : V > V from a finite dimensional vector space to itselffall into two groups. (VI . and you should not read further until you grok it.' The next few chapters constitute a detailed study of this bundle. Heinlein in his pop science fiction novel Stranger in a Strange Land. . . . To provide this extra structure. The class to which . (VI.. divid ing the colJection of all ordered bases into just two equivalence classes.
(x. = For the trivial bundle we can put the "standard orientation" on each fibre If > is an equiva lence.. . en] then det (aij) : > is continuous and never o. . ed. then (' automatically satisfies the same condition. then ( is either orientation preserving or orientation reversing on all fibres. en (X) X x IRn [(x. an orientation of E is defined to be a collection of orientations (p) which satisfy the following "compatibility condition" for for any open connected set If t : > given the standard orientation. . ej ). since rr. . f(vn)] [VI . . [eI . ··.e. v) are two ndimensional vector spaces. If 1r: E > a non trivial nplane bundle. 1R2. for if we define the functions : > by f: [f(VI )' . .I (U) U x IRn is an equivalence. together with orientations. then is either orientation preserving or orientation reversing on each fibre. Vn ] fL. .. j=1 U C B: rr. . (x. and t' : > is another equivalence. and the fibres of U x IRn are fL B is Notice that if this condition is satisfied for a certain t. and (W. fL X IR fLp rr. .) = I:>ji (X) .The Tangent Bundle 85 VI WI Examples of equally orien ted ordered bases in 1R. and is connected. . . (VI . en )] {x} x IRn . . f: en (X) en (X) X f aij X IR n f(x. . an isomorphism V > W is called orientation preserving (with respect to and vI if = v whenever = if this holds for any one it clearly holds for all. and 1R' orientation of V. vn ). .I (U) U x lRn . the other will be denoted by and the orientation for will be called the "standard orientation".I IRn fL) fL. Now if (V .
. fLp fL fL (O.I IRn is M fL s: M [s(p)] 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. it has another orientation = { . in particular. although the Mobius strip has no nonzero section.I (p) with = A bundle is called orientable ifit has an orientation. we call itself orientable or nonorientable de pending on whether is orientable or nonorientable. the Mobius strip. For. and nonorientable oth erwise.1 U IRn U IRn U B. I } with identified with ( I .fL) . to the tangent bundle of a Coo manifold In this case. considered as a Idimensional bundle over SI .I . s(p) fL is M. an oriented bundle is just a pair (�. since on which we have the C also orientable..86 Chapter 3 > : x x is an equivalence. I]. TM. For example. This definition can be applied. TM IRn M. For example. To see this we standard orientation. we can let A be [0. I] x [. fLp . TM TM (M.fL) where is an orientation for �. TIRn :::: en(IRn). The sphere s(p) fLp .} but not every bundle has an orientation.I. has no 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 }. sn. we can pick two vectors from each fibre so that the totality A looks like two sections. we had compatible orientations we could define a section Sl > by choosing to be the unique vector E A n . I] x { . then A just looks like the boundary of the Mobius strip obtained from [0. If I'01.a) a).
this shows that T(S I S I ) is also trivial. Vn.: Vip > (Mi )p is an isomorphism and the subspaces Vip vary continuously (Problem 26).O) } c there are continuously varying vectors but that it is impossible to choose continuously from among the dashed vectors for = 1. 1)(0. A (p) p.: S2p > S2A p) (p. which also discusses the tangent bundle of a manifoldwithboundary. S2 JL is the standard orientation of S2 Jr : > are oppositely oriented. This shows that if and E E then [vt. " Vn. 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). ·· . x = . . w ] JL 11"2 p p p p vp .I .I E sn.. . The torus Sl S I is another example of an orientable manifold. «0..V2] JLA(p)' Thus the map A : S2 > S2 is . and consequently orientable.(snIp) E TIRnp (Problem 21).VI . A. . 0) i . .I the vector w = Pp E en (IRn ) :::: TIRn is not in i. Nonorientability of can be seen in another way. A(v» . x x x Any nholed torus is also orientablethe proof is presented in Problem 1 6.i. (vnI l) is in the standard orientation of IRnp .(vI l. JLp P E S. 2 sin e. Since TS I is trivial. V2] JLp .Ip we can define (VI . (A(p). the restriction �IB' Mobius strip (for any orientable bundle � = to any subset B' C B is also orientable). [A. by considering the "antipodal map" defined by = This map is just the restriction of a linear map defined by the same formula. if then we could simply choose The projective plane must be nonorientable also.0») and their negatives. so if = '' I t i 2" wp S M w [v . A subset = { (2 cos e. v) p I IR'A Vi (p.I } thus defined is called the "standard arien tarian" of Sn. since it contains the E B. The Mobius strip is the simplest example of a nonorientable 2manifold. The map ( is just � when is identified with a subspace of { x The map Uj) E Pl the bases is orientation reversing. The orientation JL {JLp : P E sn .The Tangent Bundle 87 note that for each p E Sn. For tlle imbedding of considered pre\�ously we have already seen that on the Ell M M (0. so for VI .I ) E JLp if and only if (w. S2p 11"2 A : S2 > S2 A: IR3 > IR3 A.i. If we had orientations = and wp otherwise.
Condition (2) holds. for sueh that 11"2 S2n 11"2 s > f: M N > [p].. For projective 3space the situation is just the opposite. which is impossible. According to this definition. which does not use the tangent bundle of at all. U) (y.. since jj.. these same arguments show that is orientable for n odd and nonorientable for n even. U) x(U) x. 11"2 f A p [pl.' cover M M det on . E ( ayi ) 0 U n v..4.. (x.4. is orient able if there is a subset .' 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). In this case. U) E M. 11"3 A : S3 S 3 > S3 11"3 11"3 > pn M An orientation JL of allows us to distinguish the subset .:IRnTMI U T(x(U)) :::: x(U) IRn (y xI ). If g : map � we obviously can define orientations vp for by requiring g to be orientation preserving.4. In general. Conversely.: T(x(U)) T(x (U)) TMI U x* TM. > > (x. .' of the atlas .4. the determinant condition will be very important later on. the map would then Vip } on be orientation preserving with respect to jj. axj > TM (x. = JL or JL. and obtain an orientation of Although our original definition is easier to picture geometrically. given A' we can orient the fibres of in such a way that is orientation preserving.'. There is a more "elementary" definition of orientability. the antipodal map is the is orientation preserving.4. 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. (I) the domains of all (2) for all and V) . because it is just the condition that 0 is orientation preserving..
old f ) on * Euclidean spaces.:. 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). for certain equivalences 1m. N satisfying f) f' (I) of Theorem I. T'U '" (T'M)IU. then there are equivalences such that the following diagram commutes for every Coo map eM ' TM .. 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. eMl PROOF.. the welcome symbol . 4... * Functorites will notice that Theorems I and 4 say that there . those for whom this sort of proof is a new experience should find it painful and instructive. JRn. (2) of Theorem I..... for certain equivalences (3) of Theorem I. 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. up to natural equiv alence. Nevertheless.. occurs quite a ways on. If (x. THEOREM'.. Those who have been through this sort of rigamarole before know (i. T'M TM � TN f' M . and a bundle map (jil> for each Coo map M . have faith) that it's going to work out. and to the restriction of the functor on open submanifolds. N. Ifwe have a bundle T'M over M for each M. the idea behind the proof is simple enough. 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. is. U) is a chart on M.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.e.
which are denoted by the same symbol '" . This will be done in three stages.). we can define fJx . To compare ax and ay. ® Ik' ® Ie 0 Ie . Let ax denote the composition (/nlx (U» 0 '" 0 x. are the equivalences mentioned in condition (3).' for T'. so it takes the fibre of TM over p isomorphically to the fibre of T'M over p for each p E U. using equivalence ". e"(IR")lx(U) (1) (V (TM)I V =.90 Chapter 3 ax = Let (x. consider the following diagram. U). Two of them.TU � T(x(U)) � (TIR")lx(U) ' "lx(Ul. (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) . (TM)IU ""'=". Then we have the following string of equivalences.1 Similarly.1 0 ax : (TM)IU > (T'M)IU is an equivalence. (TlR")ly(V) '" IY l. e"(IR")ly(V) Ie <D Ij. U) be a coordinate system on M.TV � T(y(V)) =. 0 (". Then fJx. Our main task is to show that this isomorphism between the fibres over p is independent of the coordinate system (x.
it will be proved for TM (where it is actually obvious). ax. (II) We now need a Lemma whieh applies to both TM and T'M. Thus on the fibres over p . > > i. and the one on the right commutes because ° j I. the inclusions x(U) and y(V) come into play. and (3). for every p E V. To see this for square (D.1 oay = fJx. fJy. we enlarge it. Square @ obviously commutes. i = (TM)IU (TM) I V 1e �Ti "Z ™� I . not the explicit construction of TM. (2). and (3). Chasing through diagram (1) now shows that the following commutes.1 oax ay fJx fJy. 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). JRn JRn This means that for P E V. using only properties (I). The two triangles on the left commute by con dition (3) for TM. as shown below. i e� TV Square ® commutes because k ° y = x ° Square ® commutes for the same reason as square (D. so that it is also true for T'M.The Tangent Bundle 91 Each ofthe four squares in this diagram commutes. . Again. (2).
en (JRn) laid j. Everything in this diagram obviously commutes.92 Chapter 3 LEMMA. (TJRn ) I A � en (JRn) I A � en (JRn ) old j. . A C JRn f. B C JRm are open. If and following diagram commutes." are equal on Case 2. following diagram. 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. j/' and JRn JRm > . A'. p. em (JRm) and are equal and this proves the Lemma in Case I. then the > TA =. There is a map j : JRn JRm with j f on A . This implies that the two compositions TA =. (TJRm) IB � em(JR)mI B PROOF Case 1. where i : A > > > 1 101d f• = . 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. A. General case. (TJRn) IA � en (JR)n IA TB =. and f : A B is Coo. since the maps "old "old f. p E A' C A . p E A.
(TlRn)IA �8n(lR)nIA > i. for ease of reference. (2) / / . . :::: (T lRn) IA' Je (�) Since j 0 i is just the inclusion of Y T lRn A' in IRn . since v is oneone. � ./ � / (TA)IA' '" @ e @ e Boxes (D. and ® obviously commute. which amounts to proving commutativity of the following diagram. Thus it suffices to prove J* 0 i* = v 0 JL 0 K. in which j: is the inclusion map. @. and @ commutes by Case 1. � . we imbed it in a larger diagram. . To see that square @ (which has a triangle within it) commutes. = � 0 /C . (T lRn) I A o A o i* = v O IL O K. it suffices to prove that v �l TA' TA y A 7 IRn T lRn :::: (A) (/C) � (V) . TA +. . and other maps have also been named.The Tangent Bundle 93 and i: A' A is the inclusion map. (J O i To prove that A 0 i. . this does commute. .
which proves the Lemma. Assuming V = V. and on A' we can replace "old h" by "old J/'. and are thus equal. and the rectangle commutes by part (II). the two com Commutativity of diagram (2) shows that the composition (III) Now suppose (x. The proof that eN 0 f. on the subset (TA) I A'.TU � .l o ay fJxl o ay (3) (TM)IU "":::.I). In other words. 0 fJx. (T JRm)1 B � em (JRm)1 B coincides. we have the following diagram. because part (I) applies to x and xiV n V. T(y(U)) =.. with the composition TA =./ (y o [I). •:. em (JRm)IB. TM fJy. V) are any two coordinate systems with p E To prove that and induce the same isomorphism un the fibre of at p. it is clearly an equivalence eM. Now that we have a welldefined bundle map TM T'M (the union of all fJx. e n (JRn)lx (U) T(x(U)) =. V) and (y. The desired result fJyI 0 ay = fJx . (TJRn) I A � en (JRn) I A old J.1 0 ax follows immediately.94 Chapter 3 positions are equal in a neighborhood of any p E A.I o ax). Diagram (3) thus shows that fJy = old (y 0 X. old (y o [I). we can assume without loss of generality that V = V. as well as to y and ylV n v. (TlRn)ly(U) / nly(U). l nlx(U). (TJRn)lx(U) v. e n (JRn)ly(u) The triangle obviously commutes. = f� oeM is > Exactly the same result holds for T': left as a masochistic exercise for the reader. . Vn TA � TB =.
b e such a sequence for each open set Xi(Vi) . can find on M a pseudometric p (a function p: M M JR with all properties for a metric except that p(p.. G2 .q) = iI: 2i d (Gi(P).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. (P) . then there is some Ji with Ji ( p ) ¢ fi(A nC) . Let VI JR and XI : VI JR be the identity. It would seem that M ought to have a metric which makes each Vi open and each Xi a homeomorphism. .j (P) P E Vii· P¢V Arrange all gi. Gi(q»). such that each Xj Xi' : Xi(Vj n Vj) Xj(Vi n Vj) is continuous. =1 Show that p is the required pseudometric. Define gj.j. Let with > PROBLEMS 0 > = = > > we x > > > = • • • = . which "separates points and closed sets": if C is closed and p E A . . this is not quite true: (a) Let M = JR U {*}. A3. 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. a # 0. Vi)} a sequence ofoneone functions Xi : Vi JRn Vi C M and X (Vi) open in JRn.{O} U {*}.The Tangent Bundle 95 M be any set. by showing that every neighborhood of ° would have to intersect every neighborhood of * .. 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). q E xj' ( Bj). then there is a sequence A I .j : M [0. (d) Let fi. . Show that p is actually a metric on M. G3. . .C. Hint: First arrange in a sequence all pairs (Ai.Aj) of part (b) with A. 3.* X2 (*) = 0. . 1] by _ f g. 1]. Never theless. with support Ji c A. 1. where * ¢ JR. Actually. with X 2 : V2 JR defined by x2 (a) = a. (c) There is a sequence of continuous functions Ji : A [0. (e) Suppose that for every p . and let V2 JR . and { (Xi. and define p on M by 00 1 p(p.). let d be a bounded metric on JR. C Aj.{ ° i.j in a single sequence GJ. . 2. j 1 . A 2 . Show that there is no metric on M of the required sort.
SII which are everywhere linearly independent.... v) . V) are two coordinate systems.. continuity of follows automatically from continuity of ] : E I . x x x x a homeomorphism . Find two inequivalent.e.. while g is an isomorphism 4. fl 5. v). note that locally it is just a map x x V IRn .. B. · .. .I (B) for B C V IRn open.. then all Ix are homeomorphisms. V IRn . (c) Check the remaining details in Theorem l. where ] is an isomorphism on each fibre. . the map 0 .. (To prove the inverse continuous. (c) Conclude from Problem I that there is a metric on TM which makes each Ix 2. [y . show that ] = g 0 h where g and h are contin uous maps such that h takes fibres linearly to fibres. A weak equivalence between two bundles over the same base space B is a bundle map (].. Given a bundle map (].. (b) Show that an nplane bundle � is trivial if and only if there are II sections E SI . and is a homeomorphism of B ontO itself.I ( V) . are linearly independent for all E B... Show that in .I ( V n V) the sets of the form Ix.. ... is a homeomorphism for a collection (Xi . E2.. but weakly equivalent. (Ii) a figure eight c::><::::) . [x. 6. giving rise to two Ix : . Vi ) with M = Ui Vi . i.. f. f B2 fl .. . (iii) the torus..I (A) for A C V IRn open are exactly the sets of the form Iy. v]q � (q. Sn (P) (x. • 8. E with the vector of is a section. ... V IRn). bundles over the following base spaces: (i) the disjoin t union of twO circles.. Iy : . .I (V) .96 Chapter 3 V) and (y.. B . .. (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. Show that in the definition of an equivalence it suffices to assume that the map E I 7 E2 is continuous. 7.I (p) .. on cach fibre. 3. (c) Show that locally every IIplane bundle has 11 linearly independent sections..I (p) s: s(p) p SJ (p). maps on TM. (a) Suppose (x. V IRn . (b) Show that if there is a metric on TM such that Ix. w]q � (q. w). : E . (a) Show that for any bundle . f: BI .... Show that in the definition of a bundle map.
(a) Show that the correspondence between and equivalence classes of curves under which p corresponds to the "j equivalence class of . JR" [x. (b) Let W be the vector subspace of :Fp generated by all products fg for f.a/ax. then f is constant on each component of M. 10. f* [x. (b) Show that under the correspondence the map can be defined by TM Vw > = > > = > = = = > > . g E :F Show that the vector space of all derivations at p is isomorphic to the dual p. L.a]p . (x) = 0 1 x . If V is a finite dimensional vector space over JR. (b) For O < e < I .a.. the same must be true p of :F / W.The Tangent Bundle 97 xl . are in V. v). p x"I + W is a basis for :Fp/ W (use Lemma 2). and p JR be a linear operator with eUg) 0 for all f. as the next problem shows. . (a) Let :Fp be the set of all Coo functions f: M JR with f(p) 0.. . show that x l + W. for v.:: o. 1 3. The situation is quite different for C functions.g E :Fp. (a) Let V be the vector space ofall C1 functions f: JR JR with frO) = 0. W E V we will denote by E Vw the vector corresponding to (w. v] (0) v.o y for y a curve in with y ' = makes correspond to In. define a Coo structure on V and a homeomorphism from V V to TV which is independent of choice of bases. and that they represent linearly independent elements of V/W. 1 1. Show Jet e: :F that . If f: M N and f* is the 0 map on each fibre. Show that x f (x) /x 2 lim +o exists for all f E W. . space (:F / W) * . let { x +E O f. . If x is a coordinate system with X(p) 0. and let W be the subspace generated by all products.e has a unique extension to a derivation. p (c) Since (:F / W) * has dimension n dimension of M. = x 9. As in the case of JR". (c) Conclude that (V/W) * has dimension ce = 2 e• 14. 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.lp ' f* [f* (il](g) erg 0 fl.. x ::: Show that all f.
n n g(p. under the identification of with the vector is (a. TP g: aM x [0. lHIn . Show that this subspace does not depend on the choice of X. i* (8M)p > v) lHIn E points inward. i. Define I) a Show that is obtained from I) by x.t ). If E a and x : > > f. (a) A map is an immersion if and only if is oneone on each fibre of TM. (a) If is a manifoldwithboundary.. in fact. v Mp p M U lHIn p. fog g f M M. p M lHIn. then the rank of equals the rank of at g(a). (The reason for this choice will become clear in Chapter 8. (b) Let a E IRn1 {OJ C A tangent vector in is said to point "in ward" if.(aM)p . (e) Show that if has an orientation JL. the tangent bundle TM is defined exactly as for elen. f o g f. . f: M N f p M f.vn_. the still has tangent vectors "pointing in all vectors still run through IRn . Show that this definition does not depend on the coordinate system x.v" . > is a subspace. More generally.ents of are p equivalence classes of pairs (x.l ] ( )p Mp. . then aM has a unique orientation aJL such that = JLp for = aJL if and only if every outward pointing W E (d) If JL is the usual orientation of show that aJL is (I)" times the usual orientation of IR 1 = a lHl . Mp v) . [w. . it is where a is the inclusion. i. . where > O. i.t) '" (/(p). TM U TN > . TlHIn i: M M lHIna en(lHIn).98 Chapter 3 15. (Compare with Problem 233(d).(V) lHInx(p) M [v" . so is a coordinate system around then directions".: Mp NJ(p).) (e) Suppose we are in the setup of Problem 214.] N x [0. Vn. Although x takes a neIghborhood of E a onto rather than IRn.\IRnIX(p» C Mp vn v Mp x 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..
> (f) If M and N have orientations JL and v and f: (aM . Suppose we have a "multiplication" map from ]Rn x ]Rn to ]Rn that makes ]Rn into a (nonassociative) division algebra. .The Tangent Bundle 99 by identifYing v E (aM)p S2 with (.a.i)_p..g. Show that such a vector field can be picked so that near (0. (a "magnetic dipole"): 19.i)p with (p. and v on M C P and N C P. b) . Although there is no everywhere nonzero vector field on S 2 . and there are no zero divisors: . a · b (a l + a2) b a l b + a2 . What is the resulting manifold P? Sl [0. 18. • • (a. I) the vector field looks like the foIlowing picture 17. a JL) (aN.0.. . Sl .0) a = • = • = . .(v) E (aN)J(p) . bl + a b2 A(a . b a (Ab) for A E ]R a · (1 .0. which is diffeomorphic to ]R2. (p. I)}.. b) = (Aa) . v) E (S2. b a (bl + b2) = a .av) is orientationreversing. 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. Show that identifying  TIP'2 is homeomorphic to the space obtained from T(S2.B. show that P has an orientation which agrees with J1.i) by € {to.I ).. (g) Suppose M is with two holes cut out. That is. v) (S2. there is one on S2 0.
. or 8. 0) = 2. (d) Multiplication by a is continuous. f Math.) Let eJ .0. we can use ordinary multiplication. (f) TlP'n1 is trivial. a · en on (snl. ' " . so that we always have multiplicative inverses. . > (1. Adams has proved. Monthly 75 (1968). b) · (e. (a) Every point in snI is a .g '( t J'). see R. the quaternions are not commutative and the Cayley numbers are not even associative. 23. 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'. 366368. S. Arner. = /l = 2 = 0. e'(O» ) = ° for all curves e with e ( O) = p and 1c (t)1 = for all t . [Incidentally.. Show that M is orientabJe. (e) TSn1 is trivial.0.J " /l = . Palais. F.0. a = O). for I. el ..0). g(t» ) + (j(t). that n = I. and quaternions are the only associative examples. . (a. e2 . Multiplications with the required properties on ]R4 and ]Rs are provided by the "quaternions" and "Cayley numbers") respectively. Show that for p E S2 . . and for we can use "complex multiplication". g)'(t) = (f'(t) '.. 4. Tv). this condition implies that there are no zero divisors jf the multiplication is associative. .(S 2p) by showing that the inner product (p . . . pg. The Classification o Real Division Algebras.100 Chapter 3 (For example.bd. [0. Conversely. Let M be a manifold. It is a classical theorem that the reals. a . (Recall that 1 (f. en be the standard basis of ]Rn. en are linearly independent. complexes. the vector pp E ]R3p is not in i. . (b) Show that the resulting bundle is orientabJe if and only if T is orientation preserving. i)p are linearly independent.". = ( 1..d) (ae . (b) If a # then a . where T : ]Rn ]Rn is a vector space isomorphism. The tangent bundles TS 3 and TS7 are both trivial. (c) If p a · el E snI. el for a unique a E ]Rn. otherwise it suffices to assume the existence of a unique b with a . v) 21. Suppose that (TM)IA is trivial whenever A C M is homeomorphic to S l .ad + be). = (1. where ' denotes the transpose. . Hint: An arc e from Coo . . see Cai£ulus on Manifolds. (a) Consider the space obtained from I] x ]Rn by identifying with ( I . Show that this can be made into the total space of a vector bundle over Sl (a generalized Mobius strip). 20. using methods of algebraic topology. b b . For a simple proof.) 22. then the projection of a . J. (0.
c' which meet only at po and p. Show that �" '" �' = f*(�). Show that if � is not orientable. the symbol 1C*(�) makes sense. Since 1C : E is a continuous map from a space to the base space of �. (g) Let � = 1C : E be a vector bundle. e) = y. define 1C' : E' Y by 1C' (y.The Tangent Bundle 101 po E M to p E M is contained in some such A so (TM)lc is trivial. by using the vector space structure on 1C1 (f(y» . It is called the Whitney sum � Ell ry of � and ry. e') with 1C(e) = 1C'(e'). f(e)) E E'. (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. (b) If f: Y > B.e) = e. A vector space structure can be defined on > > > > X X (f. Show that 1C': E' Y is a bundle. and define f : E' E by f(y. Y. case can be treated by breaking up c into small pieces contained in coordinate neighborhoods. then 1C*(�) is not orientable. (c) If g : Z (d) If A c and i : A is the inclusion map. Hint : Map e E E" to (1C"(e). (a) Suppose � = 1C : E is a bundle and f : Y is a continu ous map. The next two problems deal with important constructions associated with vector bundles. (b) Suppose we have another bundle f' = 1C" : E" Y and a bundle map > > X > > a X > B B > B 24.e) with /(y) = 1C(e). then (f g )*(�) '" g*(f* (m . and is called the bundle induced (from �) by f. This bundle is denoted by f*(�). Let E' c Y x E be the set of all (Y. the fibre of � Ell ry over p is the direct sum 1C1 (p) Ell 1C'I (p). e') = 1C(e) = 1C'(e'). Let 1C"(e. The general. (f) Give an example where � is nonorientable. together with Prob lem 29. we can conclude that a neighborhood of some Sl C M is nonorientable if M is nonorientable. > B > B > B. Remark: Using results from the Addendum to Chapter 9. then i*(�) '" �IA. 23. show that f*(� Ell ry) '" f*(�) Ell f*( ry). but f*(�) is orientable. It must be checked that this is independent of the choice of c. then f*(�) is also orientable. f) from �" to � which is an isomorph�sm on each fibre. (e) If � is orientable. f) a bundle map which is an isomorphism on each fibre. Thus one can "transport" the orientation of Mpo to Mp . Show that 1C" : E" is an (11 + m)plane bundle. and ( j. First consider pairs c. possibly quite messy. .
x). (g) Define a "natural" orientation on V EEl V for any vector space V. find two nonorientable Iplane bundles and ry over such that Ell ry is also nonorientable. then 1C* has maximal rank at . (b) The Osection of E is a submanifold. Ll(X) = (x. a ()o1C) /" a�I /" . the bundle T(TM) is (lrientable. (d) If Ll: B > B x B is the "diagonal map". 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 . I . show that � ry is orientable. and 1CM [or 1CN] : M x N > M lor N] is the projection on M [or N].102 = Chapter 3 orientable.e2) (1CI (eI l. and use this to show that Ell is always orientable. the orientation for (TM)v can be defined as of x. then T(M x N) :::: 1CM*(TM) Ell 1CN TN .I shows that this orientation is independent of the choice (x. and ry is nonorientable. then both M and N are orientable. (a) If 1C : E > M is a COO vector bundle. .. i.f. show that � ry :::: Ll* (� x ry). 28.s orientable. Show that the Jacobian matrix of *( ) y* 0 (x*) I is of the form This shows that the manifold TM is alw. carried diffeomorphically onto B by 1C. show that � ry is also non = X � � � X Ell Ell Ell � 25. a�n /J the form of Y* 0 (x*). (c) Given bundles �i 1Ci : Ei > Bi.1C2 (e2))· Show that this is a bundleI �I x �22 over BI x B2 .·arh point. define 1C : E x E > BI X B2 by 1C(eJ.!).e.(p) 26. then M x N is orientable. . (c) If M x N is orientable. .. (h) If is a "figure eight" (c. (a) If M and N are Coo manifolds. (Here is a more conceptual formulation: for v E TM. (b) If M and N are orientable. 27. (e) If � and ry are orientable. . Problem 5).) A different proof that TM is orientable is given in Problem 29.. and each fibre 1C 1 is a Coo submanifold of E. (a) Let be = [a()01C) /" . (f) If � is orientable. .
(d) If n : E M is not orientable.. If A is the matrix of T with respect to the usual basis. Show that f is continuous and open. this informa tion will all be important in Chapter 10. = a di (O) = ax. 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. then the manifold E is not orientable.I) for all p E SnI . so it is w . (2) Map X E (TE) e to E2 (b) If 0 EI E3 0 is exact. and then show that f. the map w .. prove that SO(II ) is connected for all II 0::. w) for a unique T*v). (v... I. . Tw) (for each v. 0.I (p) is home fomorphic to SO (II . . (c) Show that 0 (11) has exactly two components. > > > an exact (e. show that the matrix of T* is the transpose A t . (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. (a) If T : IRn IRn is a linear transformation. 30.. n) i v ' 29. and induction on II. 0:: 2 (b) SO(l) is a point. Tw) is linear. For II define f: SO (n) > snI by f(A) = A (po) . A sequence of bundle maps EI E2 l. (T*v. is orientable if the other two are. .I). .. the adjoint of T. w) = (v... so it is connected.( Chapter I I). show that there is n*(�) TE n*(TB) O.. This problem requires some familiarity with the notion of exact sequences (c. E3 with f g = identity of B is exact if at each fibre it is exact as a sequence of vector space maps. . In addition to being used in Problem 32. Show that I (po) is homeomorphic to SO (II . Using part (a).The Tangent Bundle 103 Show that (b) Find a curve whose tangent vector at 0 is a/a(X . > > 31.. then each bundle E. which determines a tangent vector of the fibre.) > > > > Coo vector bundle. T* : IRn IRn. is defined by (T*v. > The next two Problems contain mOre information about the groups intro duced in Problem 233. (a) Let Po E snI be the point (0. (c) T(TM) is always orientable.n* X). I).
{OJ ) with det > 0 is homotopic to the identity map. and A2 are continuous functions of A. 122). Two continuous functions f Ji : X > Y are called homotopic if there is a. : (X. by showing that eigenvectors for distinct eigenvalues are orthogonal. (a) If A: [0. then some subsequence of {A (n) .) (f) Show that every A E GL(Il . Show that C can be chosen orthogonal. IR) can be written uniquely as A = A. (g) The matrices A.) (b) A linear transformation IRn > IRn is selfadjoint if so that = = for all W E IRn. B). B)) if there is an H as above such that each H. v) T v (Tv. A) to (Y. Hint: Consider At . 1R) is homeomorphic to O(n) x IRn (n +' /2 and has exactly two com ponents. IRn . B) homotopic (as maps from (X. A (n)2 . Ji : (X. E O(n) and A2 is positive definite. 1R) is continuous and H: 1R" x [0. The notation f : (X. (b) Suppose f : IRn > IRn is Coo and /(0) = 0. and positive definite if > 0 for all Show that a positive definite A is nonsingular.{OJ) C 1R" . {A : det A > OJ and {A : det A < OJ. IRn {OJ ). ) 01) GL(Il. IRn . for A C X and B e Y. At = A. Tw) T v. (Remember that A is symmetric.IR" {OJ ) to (IRn.{O J . We call /0.{OJ ) > (IRn. : X > Y defined by H. i) i = 0. IRn . show that f : (IR" . see Calculus on ManijiJlds. T: T T T* . A ) > (y.{OJ ) > (IRn . are homotopic as maps from (IRn. A ) > (Y. If Df(O) is nonsingular. pg. a continuous function H : X x [0. Hint: If A (n) > A and A (n ) = A (" ) . ) converges. so that Ho and H. A 2 where A. (d) Show that At . B). while f(IRn . (Tv. Conclude that a nonsingular linear transformation (IR" .IR" . The functions H. If A is the matrix of with respect to the (Tv.{OJ ) is T:   T . and use part (e). then is selfadjoint if and only if A is symmetric. = Ji . t ) = A(I)(X ) . t ) may be thought o fa s a path of functions from Ho = fa to H. A. (e) Show that a positive semidefinite A can be written as A B2 for some B. . A is always positive semidefinite.104 Chapter 3 standard basis. 1 . . (Notice that this also gives us another way of finding the dimension of O(n). v) = 32. 1 ] > IRn is defined by H(x. I] > Y such that Ji(x) = H (x . It is a standard theorem that a symmetric A can be written as CDC' for some diagonal matrix D (for an analytic proof. v # o. The map H is called a homotopy between fa and Ji . show that H is continuous. means that f : X > Y and f( A ) c B. w) (v. I] > GL(Il. (x) = H(x. A) > (Y. (c) A selfadjoint (or the corresponding symmetric A) is called positive semi definite if ?: 0 for all E IRn. Hint: Use the Schwarz inequality.
(d) For p E ]Rn . ]Rn . (c) Let U be a neighborhood of 0 E ]Rn and f: U > ]Rn a homeomorphism with frO) = O. Let Mil C ]RN be a Coo ndimensional submanifold. xn ) = . where U. > 2n + I. ]Rn . Show that if M is orientable.{O)). U) and (y. Hint: Consider certain maps from appropriate open subsets of TM to S N . Assuming this result. V C ]Rn are open. : B. _x n ).O). By a chord of M we mean a point of ]RN of the form p . the map y o x . then there is a vector v E S N . JRII (Xl .I is not differentiable. Check that this does not depend on the choice of C V. Let C V be the open ball with center 0 and radius r. then for any Coo structure on M the tangent bundle TM is orientable. To prove continuity at points (x. . (a) Prove that if N (i) no chord of M is parallel to v. Hint: Define H(x. let T ]Rn > ]Rn be Tp (q) = p + q.The Tangent Bundle 105 homotopic to D/(O) : (]Rn . Thus. x n.l .{OJ) > (]Rn . t ) = f{lx) for 0 < t S 1 and H(x . h(x) = then (� arctan IXI) x. q E M. ]Rn _ {Oj) . 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 .I . B. ]Rn . and let be the homeomorphism h : ]Rn > B. ]Rn . show that if M has such a collection e of homeomorphisms. ]Rn _ {O)) > (]Rn . V) in e. we will say that f is orientation preserving at p if T_ J(p ) 0 f 0 Tp is orientation preserving at O. . 33. O) = Df(O) (x). .q for p. . use Lemma 2. if M is any (not necessarily differentiable) manifold.I is orientation preserving at X(p) for all p E U n V. We will say that f is orientation preserving at 0 if f 0 h is homotopic to the identity map I : (]Rn .I in part (d) makes sense even if y 0 x.{O)) > (]Rn . .I such that M x M and . (ii) no tangent plane Mp contains v. . f o /z : (]Rn . (e) Notice that the condition on y o x.{OJ). such that e satisfies the condition in part (d). If f: U > V is a p homeomorphism. . then there is a collection e ofcharts whose domains cover M such that for every (x. then precisely one of f and T 0 f is orientation preserving at O. we can define M to be orientab!e if there is a collection e of homeomorphisms x : U 7 whose domains cover M.
J C jRN be the subspace perpendicular to V. with N = (n + 1)2). . together with the exis tence of a proper map f : M > jR. lntmduction IJJ Di erentiable Manifolds and Sternberg. Whitney. Note: This is the easy case of "''bitney's classical theorem. Show that 1C I M is a oneone immersion. A much harder J·esult of Whitney shows that f Mn can actually be imbedded in jR2n (H. Proofs may be found in Auslander and MacKenzie. ff Ann. 220246). 17ze se/finl£rsections ofa smooth llmanifold in 2ns pace. In Munkres. if M is compact. Di erentiable manifolds. jRN jRN. given by Problem ·230 (compare Guillemin . 645680). (c) Every compact Coo Ildimensional manifold can be imbedded in jR2n + J . which gives the same result even for noncompact manifolds (H. 45 (1944). Ann. Elementary Diif lltial Topo/lJgy. In particular. :nd Pollack. Then we may show that M imbeds in jR2n +J using essentially the argument above.J the corresponding projection. of Math.106 Chapter 3 > (b) Let jRN. Whitney. then 1C I M is an imbedding. Di ferential Topology). and 1C . there is a different m sort of argument to prove that a notnecessarilycompact nmanifold M can be imbedded in some jRN (in fact. of Math. 37 (1935). Lectures on Di ff f f erential Geometo'.
. for (f' fl ' = I v' and > a a g: > g)' g' The dimension of V' is the same as that of V. ll the constructions on vector bundles carried out in this chapter have a In each case. then the elements V *i E V*. Recall that V' denotes the vector space of all linear functions A : V If f : V 7 is a linear transformation. The linear function V'i depends on the entire set Vl . On the other hand. defined by (f a f. we can define v" E V" = (V')' unambigu ously by for every A E V'. then I v· is the identity map of V' and if U V. we replace each fibre (p) by some other vector space. Vn . then (f = f ' . . . . . > W a are easily checked to be a basis for V'. for finite dimensional V. not just on Vi alone. Vn is a basis for V. It is clear that if I v : V V is the identity. In fact. . 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. = I w· . .l W' > W > JR . (j'A)(V) = A(jV). if VI. then A(V) 107 0 for all A E V'. which implies that . These simple remarks already show that f' is an isomorphism if f : V is.CHAPTER 4 TENSORS Acommon feature. and then fit all these new vector spaces together to form a new vector bundle over the same base space. then there is a linear transformation f' : V· defined by rr. if v E V. The simplest case arises when we replace each fibre V by its dual space V'. = If V" (A) = 0 for every A E V'.l ) . .
) Now let � = 1f : E + B be any vector bundle.108 v = O. and define the function 1f' : E ' and I : 1f1 (U) U X �n is a trivialization. = . Once the meaning is made precise. for the present. and �. we will see later that in "most" cases �* is equivalent to �. to �n. formulated only after the term had long been in use. (We first pick an isomorphism from (�n ) . '" �.) + 1f'1 (p). then we can define a function + + peB U [1fI (p)]'. If you B into a vector bundle. the bundle �. we can prove that there is no natural isomorphism from V to V'. * (�')' is alwa equivalent to �. is an isomorphism. readers may ponder this question for themselves. Chapter 4 Thus the map v >+ v " is an isomorphism from V (0 V" . 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. in the obvious way: since the map [ restricted to a fibre. Let E' = B to take each [1f1 (p)]* to p. of �. In contrast. the dual bundle �. The lack of a natural isomorphism from V to V' prevents us from constructing an equivalence between �. by We can make 1f': E' requiring that all such [ ' be local tnvializations. (Problem 6 gives a precise meaning to the word "natural". once and for all. since each 1f1 (p) is isomorphic to However. Actually. If U c B. It is called the natural isomorphism from V to V'*. this is true merely because the two vector spaces have the same dimension. it gives us an isomorphism At first it might appear that �.
The seclion df is called the differential of f. then a Coo section df of T*M can be + defined by df(p)(X) = X(f) for X E Mp. there is seldom a operates on If s is a section of and (J geometric picture for Rather. that dc/dI ll". is a section of then we can define a function from B to � by � �* �* . to . If w is a Coo section of T*M and X is a Coo vector field. if is TM). ' (in fact. then w(X) is the Coo function p >+ w(p)(X(p» . is called the cotangent bundle of M. the re sulting bundle. This function will be denoted simply by (J(s). Like TM. * * * We can thus define Coo. 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 same is clearly true for x. as well as continuous.. Yo ' 0 (X . where e(lo) = p. denoted by T*M. in particular. Even if can be pictured geometrically (e. When this construction is applied to the tangent bundle TM of M. Suppose. the fibre of T*M over p is (Mp)*. the cotangent bundle T*M is actually a Coo vector bundle: since two trivializations x and y* of TM are Coorelated. ' and y.)I = y 0 (X )I ).Tensors 109 think about how is constructed. If f: M � is a Coo function. �* �: � � P >+ (J(p)(s(p» s(p) E 1C 1 (p) (J(p) E 1CH (p) = 1C 1 (p)* .g. sections of T OM. �* . it will appear obvious that this map is indeed an equivalence.
Thus dx'(p). then the dxi are sections of Applying the definition... . . The section df must have some such expression. dt T *M over U. . dxi .. alaxnlp of Mp. we obtain a classical formula: . . This means that every section w can be expressed ulliquely on as V for certain functions Wi 011 U. The section w is continuous or Coo if and only if the functions Wi are. We can also write W(p) = L Wi (p) dx i (p). In fact. dxn(p) isjust the basis of M/ dual to the basis alax' lp. \.. .IrO Chapter 4 Adopting the elliptical notations de dl for dt r ' de l dg(l) dl for g' (I). we see that = d(f(e(t» ) .=1 n W = L w. .=1 if c define sums of sections and products of functions and sections in the obvious way ("pointwise" addition and multiplication). . U) is a coordinate system. . . this equation takes the nice form df dl (�) If (x. .
Looking back at the classical works fi·om OUf modern vantage point. this point of view was in n af . like df. except for quotients of infinitely small quantities.just as Leibnitz had. one can usually see that. and waiting for everybody to catch up. df(p)(Xp) = Xp(f) = " a' . The dx i themselves then metamorphosed into functions. THEOREM. =L axi (p) dx 1=1 •:.. which preserved the old notation. which became tangent vectors.TeIlS01J 111 I . a tangent vector. like dc/dt . 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 . which means that df should be a function on tangent vectors. it was only a matter of making new definitions. U) is a coordinate system and f is a Coo function. no matter how obscurely expressed. If (X.e. all classical notions involving infinitely small quantities became functions on tangent vectors. df must be a function of this change. 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). i.=1 X. . af . No one wanted to admit that this was nonsense. then on U we have P ROOF If Xp E Mp is then Thus n . and it became clear that they must be distinguished from the tangent vectors a/ax" Once this realization came. In short.(p) L a . Since df is supposed to be the infinitesimal change of f under an infinitesimal change of the point. ' (p)(Xp) .
1 (1) l(x l (I).. x"(I» .=1 (f 0 C)' (I) ax' dl (The classical notation. " ( ) Since every tangent vector at of the form dc/dl. CLASSICAL FORMULATION = Let I be a function on M. no matter what tilef unctions X i (I) are. Then 1 0 C(I) = j(x l 0 e(I). . say I I(XI... . (Xi 0 e)'(I) L 1=1 axl af = � ax' (C(I» L.. =j = I 0 c : � + �. . . x").) de al dl (. .. = j = l oxI). . which suppresses the curve C. .1 12 Chapter 4 some sense the one always taken by classical geometers. .==1 " al = " �(c(I» . where M be a curve. . . and x a coordinate system (so that I oX for some function on namely Let MODERN FORMULATION Let X i be functions of t.) or = L: D. (Xi 0 e)' (I) .=1 Consequently. say X i X i (I).1I . It is the closest approach in classical analysis to the realiza tion of as a function on tangent vectors. Then I becomes a function of I. . is still used by physicists..x" 0 C(I» . as we shall point out once again in Chapter 7.=1 . dl L.=1 d(f(c(I» ) dl Multiplying by . .(This equation signifies that true results are obtained by dividing by dt again. . We now have We now have dl = � af dX i .. c: � + j lffi. In fact.. . .](x(e(I» ) .x". dx i (e(I» dl dC · dx i dl dl gives 1=1 af dl = � ax' dx" L. we have C(I) is dl af dl = � axl dx" L . . the diff erential d was usually introduced in the following way: l Let I be a function of the X l .) = L: � (c(I» dl ax' .
So it's very easy to remember which kind of vector field is covariant. roughly. should equal J P2 ' On the other hand. There is no corresponding way of trans ferring a vector field X on M over to a vector field on N. in f act. Since J p is a linear * * transformation between two vector spaces.. * for each P E M. After a while. Then we can define a section � of T *M bundle. Classical terminology used these same words. Suppose w is a section of T as follows: �(p) = w(/(p)) 0 J*p. and then operate on it by w. and it just happens to have reversed this: a vector field is called a contravariant vector field. And no one has had the gall or authority to reverse terminology so sanctified by years of usage.Tensors 113 In preparation for our reading of Gauss and Riemann. then there is a map J : TM + TN.e. the same q E N may be /(Pt) for more than one Pi E M. we can * * do something with the cotangent bundle that we could not do with the tangent *N. Notice that we cannot put all J* together to obtain a bundle map from T *N p to T *M. i. we will continually examine the classical way of expressing all concepts which we introduce. Despite these diff erences. while a section of T *M is called a covariant vector field. . 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 there is no reason why J p. and which contravariantit's just the opposite of what it logically ought to be. we can say.) This section � is * denoted. Recall that if J : M + N is Coo. naturally enough. 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. we have a map J p : Mp + Nf(p) . the CCtranslation" of classical terminology becomes only a little more difficult than the translation of the German in which it was written. we push it over to N by J . (The complex symbolism tends to hide the simple idea: to operate on a vector. by J *w. 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".
n W = L Wi dxi .n x(v. any combination xi. .1 14 Chapter 4 The rationale behind the classical terminology can be seen by considering coordinate systems on which are linear transformations. so the x coordinate system is just an "oblique Cartesian coordinate system". . x(a l vl + . In this case. _  1' P I x(p) = (a.=1 is also called "covariant". . Comparing (*) with D (x' = x' n I dxIj = L � dXI . .=1 n " W = L WIi dxII . we see that the diff the coordinates hence they are cccovariant". COllsequently. if then x lffi. . (ax'j laxl) is the constant I h is can be seen directly from the fact that the matrix matrix 0 X.I ) o XI . . then Clearly so x' aij = ax'j laxl . + anvn) = (a l . an).) = el.b) If is another such coordinate system. x'j = L7=1 aijx" for certain au .j 1=1 dxi erentials "change in the same way" as from Theorem I. Notice that if we also have . " aax. . bv.
i. . a function is multilinear if is linear for each choice of is clearly a vector space. .) Covariant and contravariant vector fields. . . Vk _l . .l lj = '"" i aaxlj . . If Vb " " Vm are vector spaces.. . this vector space will be denoted . We begin with some worse algebra. . respectively. . . . Substituting 115 into the first expression for w and comparing coefficients with the second. .=1 L. Vk+I . . If V t+ T(VI . Wi ax" j L ax'j 1=1 La On the other hand.Tensors then we can express the W'. XII used to be denoted by XI .. . . we find that n I W = . Xn . = � 'I axIl . . i n terms o f the Wi . .. sections of T *M and TM. .a X for a vector field. Af ter this was firmly established.. are also called covariant and contravariant tensors (or tensor fields) of order I .e. . This suggested subscripts Wi for covariant vector fields and superscripts at for contravariant vector fields. which is a warning that worse things are to COme. . . the functions a'j must satisf y a n . The set of all such T VI . . Vm = V. Vrn. given two expressions � a1 L 1=1 a a ax. . (Coordinate func tions X l . Vk+ h " " Vrn) VI . 1=1 These expressions can always be remembered by noting that indices which are summed Over always appear once "above" and once "below" . V. the indices on the x's were shifted upstairs again to make the summation convention work out. Vk_l . .
and S E r i (V) we can define the "tensor product" T ® S E rk+i (V) by T ® S(v" . . . S. Vn is a basis for V and v *I . then the elements For T E r k (V). . . . (S. defined completely analogously to the case m = I: are easily scen to be a basis for r k (V). if VI . (S ® T) ® U = S ® (T ® U). T ® S is not S ® T. . Vk+i ). so we can define nfold tensor products unambiguously. . S(Vk+ " . . ® T S. . In particular. which thus has dimension nk We can use this new algebraic construction to obtain a new bundle from any vector bundle � = 1C : E + B. .116 Chapter 4 by rm(v). + + and let If = peB U rk (1C' (p» . . Notice that r ' ( V) = V*. . Vk +i ) = T(v" . ® T. We let E' O fcourse. Vk . Vk ) . v *n is the dual basis for V* = r'(V). . . . · · . this tensor product operation is itself multilinear. O n the other hand. etc. . 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. If f: V + W is a linear transfor mation. . .) ® T = S. these maps can 1C'' (U) + U X �n k . U c B and is a trivialization. Vk +' . . then there is a linear transformation f* : rm(w) + rm (V).
. . so that TM. . Xi + X'I . . . . . X'" . . If (x. . . and a section is called a covariant tensor field of order k.. which yields a function: . . To derive this equation. ik are. • • . n are a basis for Tk (Mp).k or simply Ifwe also have then (the products are just ordinary products of functions). . ® dX lk (p) E Tk (Mp) I ::. . Xk ). Thus.ik ::. . t" . ® dXlk . . . then the kfold tensor products dXil (p) ® . . . .. on U every covariant tensor field A of order k is a basis for can be written dx l (p).. .Tensors 117 We make H ' : E ' + B into a vector bundle by requiring that all such / ' be local trivializations. . . . ® dxlk (p). .. A covariant tensor field A of order k can just be thought of as an operation A on k vector fields XI . . . U) is a coordinate system.i/. . . .. A(Xh . The section A is continuous or Coo if and only if the functions A i . . . .. aXi. . . Xk Notice that A is multilinear on the set V of Coo vector fields: A(XI . . .: A = L A II . and multilinearity of ®.··. . . we just use equation (**) on page 1 1 4.lk (P) dxi' (p) ® . . il.. . Tk (�) �*k T (TM) (Mp) *.···..: where dX il ® .. . . .. . For the case of the bundle is called the bundle of covariant tensors of order k. Xk) = A(XI . . . ./. Xk ) Xk ) = aA(XI . .ik dX11 ® . il . . ® dX ik now denotes a section of Tk (TM) . The bundle is the special case k = I . il. dxn (p) A (p) = L AI) .Xk ) + A(XI . .XI . .
. . . Xk)(P)· = l (p)A(p)(Xdp). . If P ROOF Note first that if v E Mp is any tangent vector.e. XI . . (p). (p) for each we claim that i. then there is a vector Geld X E V with X(p) = v. 1X" . . . . . Xt (p). then there is a unique tensor field A with A = A. . then we can define v = L al axl p ' . . · · · . Xk)(P) = A (p)(Xd p). . . . Vk E Mp are extended to vector fields X" . if I is Coo. . because A is defined "pointwise". Xd p» We are finally ready for another theorem. . if U) is a coordinate system and is linear over F. . In fact.1 18 Chapter 4 rvforeover. .. . . . \vhere each at now denotes a constant function and f is a Coo function with I(p) = I and support I C V. then for we have A(X" . . . f(p)X. . Xk E V we clearly must define The problem is to prove that this is welldefined: If XI (p) = Y. . . . THEOREM. . . Now if v" . Xd p» = I(p) · A(X" . . one that is used over and over. . . A : V x · · · x V + F "..=1 n a I on U outside V. it is actually linear over the Coo I Ullclions F. i. 2.' k limes (x.
so f A(X) evaluating at p gives X(p) = o.) For simplicity. can write = (The map A "lives at points". where T k (TM) = T*M.b' a x ' V) be a coordinate system around p.. = 0 if g C V. A since bl (p) (g � ) a 1 = (p) o. The proof that A(X)(p) = A(Y)(p) when X(p) = Y(p) is in two steps. and support 1=1 � a . Note that Theorem 2 applies. AUX) = AUY) A(X)(p) (2) To prove the result.Tensors 1 19 (I) Suppose first that X = Y in a neighborhood V of p . � bl (p) . then is a welldefined Coo vector field on all of A(X)(p) = M which equals X on V. 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: . to use the in terminology.. in particular. where all b' (p) = o. the cotangent bundle: a function fi·om V + 'F which is linear over 'F comes fiom a covariant vector field w . Because of Theorem 2.. A(Y) (p). to the case k = 1. = fA(Y). •:. Then fX = fY. Let f be a Coo function with f(p) = 1 and support f C V. take the case k = 1 (the general case is exactly analogous). nor will we use the symbol A any longer. it obviously suffices to show that A(X)(p) X = L. Now A(Y)(p) = = 0. so that by (I). If g is 1 in a neighborhood V of p. Let ( x . so that on V we = A(Y)(p). we will never distinguish between the tensor field A and the operation A.
Now an element v ofa vector space V can be thought of as a linear function v : V* + �. . . We could also use the notation '1k(TM). proved exactly the same way. In local coordinates we can write k + 1 times).j. .: Ah . 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 . ·· h __ <8> . .h� <8> . x Mp Although covariant tensor fields will be our main concern. 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 . if and are covariant tensor fields of orders k and I._ ' aX'Jk aX'}1. A A(p) a. h.. j(. " <8> . respectively. there is an analogue of Theorem 2. then we easily compute that A. a A(p) = L A}. <8> __ . . that allows us to dispense with the notation A.. we just define V(A) to be A(V). If we have another such expression.IP aX}k ax}'. .k (p) . .120 Chapter 4 A <8> B (operating on Moreover.jk (remember that each alax j Ip operates on Mp*). each on Mp* . a A = L A. or simply a A = Lh.. if we use TdV) to denote all klinear functions on V*. .. if only for the sake of completeness we should define contravariant tensor fields.P" " Pk = L jl.. So each Xp E Mp. " <8> _a_. · ·}. Wk into a function: A Naturally. . Recall that a contravariant vector field is a section X of TM. . j(.IP <8> .···.. .··. aX}k ax}' h. thus. A contravariant tensor field of order k is just is a klinear function a section of the bundle T k (T*M).jk A contravariant tensor field A of order k can be considered as an operator taking k covariant vector fields WI .···.
we can take the trace of the corresponding S : V + V. Given S. .' (�). obtained by replacing each fibre 1C . In particular. Moreover. If v" . however. our isomorphism can be used to transfer any operation from End( V) to 7/ (V). this map is an isomorphism.' (1C .j . although they should be kept to a minimum. covariant of order I and contravariant of order I . for 5 = implies that A (S(V» = for all A. 5(V. certain ones are quite important. i. 5: by the formula Vx v· + �.' (p) by 7. If V is a vector space.Tensors 121 Finally. we are ready to introduce "mixed" tensor fields.' ( V) is linear and one one. To make the intro duction less painful.' ( 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. V is a basis of V and n T = L T/v·i <8> Vj . sections of 7.' ( V) have dimension n'.A) = A (S(V» . ' (V).' (p» . we consider a special case first.' (TM) are called tensor fields. . for each v the vector S(v) E V is merely determined by describing the action ofa A on it according to (*).' ( V) denote all bilinear functions T: V x V· + R A vector bundle � = 1C : E + B gives rise to a vector bundle 7.' ( V). is not so easy to describe. given by Generally speaking. given a bilinear e(v. Let End( V) denote the vector space of all linear transformations T : V + V ("endomor phisms" of V). In particular. Since both End(V) and 7. which implies that S(v) = for all v. The inverse. let 7. It is not hard to check that this isomorphism of End(V) and 7. . Notice that each S E End( V) gives rise to a bilinear 5 E 7.A) = A(V). There are all sorts of algebraic tricks one can play with 7. T: V x V · + �. the correspondence S >+ 5 from End( V) to 7. this number is called the contraction of T.
defined by S(vt) L aji Vj . in any fibre bundle 7i (�). I L a axj ' II (contraction of A) = The general notion of a mixed tensor field is a straightforward generalization. in fact. More over. In particular. each A (p) is an endomorphism of rr1 (p). each section A gives rise to af unction (contraction of A) : B defined by p r+ contraction of A (p) = + 2 T L T/. i=1 = II � trace A (p) if we consider A (p) E End(rrl (p» . In this case.122 then we can find the matrix A = in terms of the T/. obtained by replacing each fibre rrI (p) by End(rrI (p» . Thus. <8> i i. which is a section of we can consider each A (p) as an endomorphism of Mp. given a tensor field A.' i limes <8> • • • <8> V* + R . A= A J dX '. J. Define to be the set of all (k + I)linear L A. covariant of order and contravariant of order 1 . . If in a coordinate system . aj/ v*j (S(v/» T(v/. 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. 1=1 1/ ( V) T: V 'v' k times <8> • • • <8> V x V* '.) These identifications and operations can be carried out. then 7. fibre by fibre. = Chapter 4 (aij) of S. a section A of Ti (�) can just as well be considered as a section of the bundle Elld(�). and we obtain a function "contraction of A". v*j) T/ .1 (TM).
nk+l functions in terms of the coordinate system and then checking that (*) holds. Consequently.Tensors 123 Every bundle gives rise to a bundle Sections of are called tensor fJelds.'.j ax. all important tensors are actually defined by defining the nHI "HI x x'..···..a axj et L this is certainly true. . h. aX. ·.j � ax'p . in classical diff erential geometry.ijr . i..: � ..e.. and if A' = L il. i. J. Locally.k (TM) (�).ijr • .lh/. . x.. n' of of oP ." In other words.·.'\'o. In fact.: jl. or simply of type an abbreviation that also saves everybody embarrassment about the use of the words "covariant" and "contravariant". Ai. ·. . axh i( /. Here is an important example.al ' " ax'ctkJ.. covariant of order k and contravariant of order I.. then (*) holds.k (�) . for . Classical differential geometry books are filled with monstrosities like this equation. . a tensor field of type can be expressed as � J... cca set of functions which changes according to (*)".. one will find the following assertion: "The Kronecker delta is a tensor. A (�) A= L il k . it is asserted that if one chooses the same functions for each coordinate system.: then axil axl ax'PI A'g::: g� = L Af(.. In every classical diff erential geometry book. ax'P. the classical definition of a tensor field is: an assignment of functions to every coordinate system so that (*) holds between the functions assigned to any two coordinate systems and (!) Or even. jl.·.i��.
L bp dXP (p).J is a certain tensor field. classically. sat isf ying A(p) AI. a I " n Thus isjuSI the evaluation map Mp x Mp' + �. what this equation shows is that '.) LO! dx'(p) <8> a!j IP (v. aib..124 Chapter 4 a " A Le 0f..j j b i. in a similar manner. P=I = + �. i=1 = J..) a collection of functions one for each coordinate system.. dxl <8> axj = From our point of view.J..e.. The contraction of a tensor is defined. . Given a tensor. it isjust the identity map. then A (p)(v. Mp x Mp' we consider = v E Mp and J. _ L..j j = L. considered as an endo morphism of Mp.0.J.. independent of the choice of the coordinate system x To identify the mysterious map A(p). E Mp' with the expressions J. i.(v).) i.
Tensors we note that 125 a=1 � A'aa = � (" Aj' � aX. so that vector fields can be defined as sections.A) . While a COlltravarianl vector field is classically a set of n functions which "transforms in a certain way". V2. This calculation tends to obscure the one part which is really necessaryverification of the fact that the trace of a linear transformation..3 (Mp ). A) >+ A (p)(v. a vector at a single point p is classically just an such that the assignment of n TZumbers a l . (Mp) to be B(p)(v) . . a n to each coordinate system numbers a l ) . ax/ja .(p).. . defined as the sum of the diagonal entries of its matrix. .j a=1 ax = L A{a) i. a 'n assigned to satisfy x. a=1 "transform correctly" if the A�r do. A). then 2 E T.yv = L AaPvya P " J1. A2. A2) = contraction of: .ax i . the functions (�) we are laking B(p) n BJ1.ja ) L L L: ax">: ax a=1 i. . n . x' ' aJ (v. For example.. VI . . A) .= This is precisel the definition we adopted when we defined tangent vectors as y equivalence classes [x. . Incidentally. and that all other alp.j = L AL . Ifwe consider each A(p) E T.} n ax = L A !. ax'j = ") a'. rather than as equivalence classes of sets offunctions. . L. .=1 n so that this sum is a welldefined function.. . is independent of the basis with respect to which the matrix is written. a tensor of type can be contracted with respect to any pair of upper and lower indices. The revolution in the modern approach is tllat the set of all vectors is made into a bundle. V2. L ax'a.
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. such invariant definitions are usually not so easy to come by. instead we define a function A on vector fields. We seldom define A (p) directly. and hence must come from some A . We won't go quite that far. Unlike the "Kronecker delta)' and contractions. The tangent bundle itselfwas almost a victim of the excesses of revolutionary zeal.126 Chapter 4 types of tensors are constructed from this bundle. or as equivalence classes of curves. which miraculously turns out to be linear over the Coo functions F. the return to the old definition was influenced by the "functorial" point of view of Theorems 31 and 34. . the party line held thaI TM must be defined either as derivations. invariant definitions of all the important tensors in diff erential geometry are made by means of Theorem 2. For a long time. At the appropriate time we will discuss whether or not this is all a big cheat. but we will give an "invariant" definition (one that does not use a coordinate system) of any tensors that are defined. As we shall see.
.x� f) (p) . w*) . dxl (p). . L. Let + dUg) = fdg+g dj..g: M f: M dx" f* ( . then the same is true of the bases n and n for V*. Y j=1 and. is orientable. . V) are coordinate p and f(p).. U) and (y.···. . . show that f.(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. For v E Mp. (c) Show that U*dyj)(p) = t a(y. (a) If g: N �. v* N are Coo. If f. 1=1 0 f = identity. . . Vn and WI . . . Jn • • • + 3. . ) w* . . (b) Show that a bundle � is orientable if and only if �. respectively. more generally. .j" dyh <8> <8>dyj. express j�(L7=1 a i alax i l p ) in terms of the alay j lp .) (d) Express in terms of the 2. � be Coo. ah . • • Wn for V are equally oriented. and suppose that (x. .) } 1 . show that 4 . (a) Show that i fthe ordered bases VI .(v) = df(v)f(p) E �f(p)· v*). then 1 .Tensors PROBLEMS 127 f: Nm.
if Vi is defined [cJ (iii)] by a ij Aali v j = 0. . Ci = 1.(Jj )�i� j = 0 is satisfied by every vector field �i. (iv) If aij = a ji for the components of a tensor field in one coordinate system. . If aij and bkl are components of a tensor field. where fJi is a given covariant vector. Il indicates the component and ex for ex = I.= 0 is expressible linearly in terms of II . then a ij + aj i are the i components of a tensor. and then translate the problem and solution into modern terms. (v) If aij and bij are components of a tensor field. then a. Remember that the summation convention is always used. then (alj . 1.128 Chapter 4 (i) If the quantity ). the other sets are com ponents of a vector field.are the components of a given vectorfield. n . or b = c and ars is skewsymmetric. then aij + Qj = O. Hints and answers are given at the end. .1 and fJi Vi oJ 0. Show that the rank is invariant under all transformations of coordinates. check them. . and by definition 5. that is. . . . i (vii) If a ij AiJ. In each case. n . where i for i = I.. . so are aiJ bki . .is an invariant and either Ai or fJi are the components i of an arbitrary [covariant or contravariant] vector field. using the classical methods.) = 0.j = 0 for all vectors Ai such that Ai fJi = 0. so Ai jLi means L7= 1 Ai JLi . .. An "invariant" is just a (welldefined) function. . (vi) I f a j AiA j i s a n invariant for A i a n arbitrary vector. det(AaI') oJ 0. .ij = a'ji for the coordinates in any other coordinate system. . if Qij)J). and consequently i 1 a ij + aj = (fJj!Jj + fJj (Ji ). . i . and these vectors are independent. (ii) If Aali are the components of n vector fields [in an nmanifold] .1 independent vector fields i Aali for Ci = . after (xiii). (ix) By definition the rallk of a tensor of the second order aij is the rank of the matrix (aU). so are aij + b lj .} fJ. in particular. . . The following statements and problems are all taken from Eisenhart's clas sical work Riemallnian Geomet�). then either b = c and ars is symmetric. n the vector. show that if bars + casr = 0.1 which satisfy the equation. then any vectorfield Ai is expressible in the f orm where the a's are invariants. any vectorfield A satisfy i ing Ai fJ.�fJ. i (iii) If fJ.: (viii) If ars are the components of a tensor and b and c are invariants.
where fJ. w) = T(w. . For. v) + T(v. . (ix) T: V x V + � corresponds to T: V + V' [where T(v)(w) = T(v. = ° 1 I 1 if ja = jp for some a oJ fJ Or ia = ip for SOme a oJ fJ or if U" . . and vice versa. . .E V * and T(v) = av for all v E kerfJ. . (iii) Given w [with w (p) oJ for all pj . 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.) (vi) For T: V x V + �. T(v. ip (i) w is determined if w(X) is known for all X. there are everywhere linearly indepen dent vector fields X" . (xii) Let V = p' . then (xiii) If j 8 h .is a given vector. Then T + T' is determined by S(v) = T(v.a8lj)Ai = 0. . jp) oJ {i" . (vii) Given w [with w(p) oJ for all pj . . Similarly.and writing v uniquely as vo+cYo for Vo E ker fJ. . . v + w) = T(v. choose Y complementary to kerw at all points. . . . . . . ° ° ° M T(v) = av + fJ(v)y for all v . where ai and bj are the components of two vectors. . 1f T : V + V and fJ. . w)] .l. I) consists of vectors tangent to S' x {I}. ip if j) . show that for the symmetric tensor aibj + ajbi the rank is two. Z). let T' (v. If O"(Z) = T( Y. .. w) + T (w.Tensors 129 (x) Show that the rank of the tensor of components aibj. . . . can i be written in the form (a j . v) + T (w. . (Begin by choosing Yo complementary to ker fJ. . then T(Z. w). jp is an odd permutation of iI. . (xi) Show that the tensor equation aij Ai = aAj. . is one. . T(v + w. v) . . .) . jp is a n even permutation o f i) .such that HINTS AND ANSWERS. v) = for all v implies that T + T' = 0. if the equation is to hold for an arbitrary vector Ai . i (xii) I f a ij Ai = aAj holds for all vectors A i such that p iAi = 0 .:·. Xn _. . . For example. which span ker w at each point. .i�p are the components of a tensor in all coordinate systems. v* n) . there is a y complementary to ker fJ.. on S' x � there is an w such that ker w (p. . v) . Show also that a ij = 8 ija. Vn and v* ) . ip} i f j) . 'I ···'p then Of. where a is an invariant. (This is true only locally.
(f) The "T k functor". k F U) ( T)(AI> . + W is an isomorphism. except that FU) : F(W) + F(V) and F(g 0 f) = FU) 0 F(g). . . W � W* Hinl: There does not even exist an isomorphism i : � diagram commute for all linear f : � + �. . . Ap) = det(Ai (Vj» . . then FU) is an A conl>ava/ian/f ullclor is defined similarly. F(V) = T k ( V). F(V) = V*'. . may also be defined . (c) The "T functor". FU) = f is a functor. such that F( I v) = I F(V) and F(g 0 f) = F(g) 0 F(v). covariant in some and contravariant in others. . . . . f l fr + �* which makes the (a) The "identity functor". . (e) The "dual functor". F(V) = V*. (d) If F is any functor and f : V isomorphism. Show that for any linear transformation f : V + W. Functors of morc than one argument. . 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). F U) = f * is a contravariant functor. (b) The "double dual functor". (a) Let iv : V + V** be the "natural isomorphism" iV (V)(A) = A(V). x V x V* p times x (xiii) Define p times . FU) = f * is a contravariant functor. 6. the following diagram commutes: V � V** f 1) Show that there do nOI exist isomorphisms i v : V lowing diagram always commutes. . x V* + � by O(VI . ' Ak ) = T(AI 0 f. FU) = f'* is a functor. W � W** l 1 + V* such that the fol 7. . .130 Chapter 4 0 : V x . . Vp . · · . V � V* 1** . Ak 0 f) is a functor. . F(V) = 1k ( V) = Tk ( V*). !" .
Call F continuous if this map is always continuous [using the metric in part (a)] . See the next two problems for other examples. (The bundle Ti (TM) is just a case of the construction in (b). once and for all. . and consequently give it the metric of �nm . W) to Hom(F(V). (e) The functor F(V) = V* is continuous. the same construction can be used when F is a functor of several arguments. �) we can consider it as a map A : �n + �n.) (d) DefIne a continuous contravariant functor F. (a) Let Hom(V. . we can identify Hom( V. and show how to construct a bundle F(�). . W ) denote all linear transformations f rom V t o W. . . the class of ndimensional vector spaces.1 (p» . (b) A functor F gives a map from Hom (V. (b) How does the homomorphism h depend on the initial choice of the isomor phism F(�") + �k ? (c) Let v = (V I . W) with the m x n matrices. Choose. . (a) Let F be a functor from Y". Show that if � = 1C : E + B is any vector bundle. (The bundle T*M is a special case of the construction in (d).Tensors 131 corresponds a trivialization 8. and F is continuous. Choos ing a basis for V and W. �) is a homomorphism. Show that a different choice of bases leads to a homeomorphic metric on Hom(V. If e + v denotes the isomorphism n taking ei to Vi. 9. . W). and such that to every trivialization I : 1C1(U) + U X �n (c) The functor 7J (V) = 7 1 ( V*) = V** is continuous. to yk Given A E GL(n.) Generally. e ) be the standard basis of �n . then there is a bundle F(�) = 1C' : E' + B for which 1C'1 (p) = F(1C. Then F(A) : F(�n) + F(�"). wn ) be ordered bases of V and let e = (e l . �) + GL(k . show that the following diagram commutes . as well as an example of a functor which is not continuous. . . an isomorphism F(�n ) + �k Then F(A) can be considered as a map h(A) : �k + �k Show that Iz : GL(n. . . The bundles 'Jjk (M) are all special cases. F(W» . vn ) and w = (w" .
aq] � (w . (g) Let Ci: � + � be a noncontinuous homomorphism (compare page 380). THEOREM. define ( V . that Fh is a functor.vj .q) by [v. q] = [w .j) is defined by Wi = L aj. (f)[v .) = L'j�. choose ordered bases V . aj.Vj . This suggests a way of proving the following.132 where A = Chapter 4 (a. aj. : yn yk (d) For q. = L'j�. define A by (f) If V. If fl : GL(n. �) = � be h (A) = Ci (det A). �) is any homomorphism. and that everyequivalence class contains exactly one element (v. q) for a given v.q] = [v. yn y ' . + such that the homomorphism defined there is a functor Fh in part (a) is equal to h. q]. W E f(v. Then Fh : + is a noncontinuous functor. h (A)(q)]. (A) = h(A) when we identify Fh (�n) with �k by [e. F(V) n After identifying F(�n) with �k . q2] = [v. Show that � is an equivalence relation. q) if q = h(A)q' where W. q] r+ q.q .. and let h : GL(n. We will denote the equivalence class of (v. W. and define yn F. q') are welldefined operations making the set of all equivalence classes into a kdimensional vector space Fh ( V). + q2] a · [v. (e) Show that the operations [v. this means that �k h(A) �k [ // F(e + v) F(e + w) also commutes. Show that this is a welldefined linear transformation.q' E �k . �) + GL( I .�) + GL(k . and f: V + W. qd + [V . and that F.. j =1 .wj.
Conclude that the bundle of odd scalar densities is not trivial if M is not orientable. so every section on can be expressed as a . (Xx '. Let Fh be the functor given by the Theorem. . �) take A into multiplication by det A. assignments of a single function a to each coordinate system such that the function a ' assigned to satisfies x' a ' = det ( ax' ) · a . instead.(Mp). Clx for a unique function Q . show that the corre sponding equation is a ' = det (c) For this h. x axl} are called (even) scalar densities. · a.. other "quantities" which are defined as sets of functions which transform according to yet other rules. I G:'� ) I . h takes A into multiplication by 1 det A I.(V) determines an orientation for V. a. If is another coordinate system and a · Clx = a t . . V) is nonzero. If is a coordinate system. and consider the I dimensional bundle Fh (TM) obtained by replacing each fibre Mp with F.. For example. 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. �) + GL( I . In classical tensor analysis there are. assignments for which (a) Let h : GL(n.Tensors 133 10. in addition to mixed tensor fields. show that x' V a ' = det ( ax' )  ax'. show that a nonzero element of F. then are called odd scalar densities. (b) If. These new rules are of the form A' = A operated on by h ( axa ) ax'P .
. ) I). we obtain the bundle of odd relative tensors oftype and weight w. 'Ilk (f)(T)(V Given A E GL(IZ. (�) (�) j (or the same formula with det(ax i lax. ) " .) if ia iJ . . . f(Vk ).k(A) of GL(IZk+I . .1 with these compo nents in every coordinate system. n . �) + GL(n k +I . in is an odd permutation of = ip i). Vk .134 Chapter 4 k+1 by taking (d) We can identify r/ (ll�n ) with �n Recall that if f : V + V. �). . . . is an even permutation of I .k(�n) + r. . . .k(A) w an integer. . . .. AI ) = T(/(vtl. . . . . . . �) be defined by h(A) The bundle F(TM) is called the bundle of (even) relative tensors of type and weight w. we define 'Ilk (f) : 'Ilk (V) + 'Ilk (V) by I . . n I. . .ill are the com ponents in every coordinate system of a certain contravariant relative tensor of weight 1 . . �). Show that the transformation law for the components of sections of these bundles is = (det A)wr... (e) Define if if Show that there i s a covariant relative tensor o f weight . 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]. · · · . . AI 0 f.k (A ) : r. .i n for some Ci oJ {J .i" = 8. Then r. Also show that gi l . (See Problem 712 for a geometric interpretation of these relative tensors.k(�n) determines an element r. . . . If (det A)W is replaced by I det A l w (w any real number). we can consider it as a map A : �n + �n. Al 0 f). Let iz : GL(II. .j ) replaced by I det(ax i lax. 1 .
in general. p(O) = P (e. . i. that is. 135 .X is Coo on a(M). Since this a local question. means that P. Consider the curve The condition c = x op.X on x (U) C �n . (�IJ = �. It = X(p(I» . we wish to introduce a coordinate system (x. Recall that. Cl* X does not make sense for Coo functions a : M + N.. a curve p with Wsections. However. fn (q). There is a function f : x(U) + � n with i. We would like to know if there is a curve p : through p whose tangent vectors coincide with X.e. vector fields.. and its Let X be a vector field defined in a neigh + M borhood of P E M. X)q has "components" P (q).e. U) around p and transfer the vector field X to x(U) C �n . (� IJ dp = X(p(I» dl = X(p(I» . if ex is a diffeomorphism. . e) P. In particular. we have a vector field x.CHAPTER 5 VECTOR FIELDS AND DIFFERENTIAL EQ UATIONS e return to a more detailed study of the tangent bundle TM. . . then we define It is not hard to check (Problem I) that a. (x.
. . of course. (� IJ = x. e V C'(I) = /(c(l» is called an integral curve for c(0) = x XEU / with initial condition c(O) = x. we will work entirely in Euclidean space. c. = = (x. If c �n is open and / : + �n.136 hence Chapter 5 �It = x. to the diff erential equations one obtains upon introducing a coordinate system.X) x(p(t)) (x.e. . cn (l») We also want the "initial conditions" i = l. Ch(l) = f ( C' (I). a form to which our particular equations never reduce). e) p(O) = dl p <!£ = X(p(l» is called an integral curve for X with initial condition p(O) = p. For quite sonle time. which may also be considered as a system of II diff erential equations for the functions ci .. and for a while x. A Curve p: + M with (. . . solutions were consequently called "integrals" of the equation. (X(p(l))) c'(I) = /(c(l» . n . . . Part of this terminology is still preserved. Similar ter minology is applied. 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. . . y. 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 /: � + �. will denote points of �n. X)c(t) . then a curve c : ) + M with V (e.P. etc.
In this case. c'(I) = . (They really do make sense. the equation in question asserts that y' = J a y.=x+C Y Y Thus the curves are supposed to be solutions.) To obtain the initial conditions c(O) = a.Fields and Diff erential Equations 137 Before stating the main theorem about the existence and uniqueness of such integral curves.[c(l)f. then (F a y)' = I F(y(x» = + C for some C. if F' = I If. so C(I) = I + C = x + C· I I hence. 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 . The equation for a curve C with range R. which would be written classically in terms of a function . the correct solution is for ali i C(I) = 0 .. we must take X I c(1) = This works in all cases except 1 + l /a ' a = O. we consider some special cases.Vecto. This can be checked directly if you don't believe the above manipulations.
the curve C (I) = J /(I + J la) is defined for all large I. but it can also be obtained with a rather Jess stringent condition. except c(/) = 0. It might be thought that this somehow reflects the fact that X (0) = 0. In terms of vector fields. For a > 0. even or though X is defined on all of � . namely . We say that the function I satisfies a Lipschitz condition on V if there is some K such that I/(x) .138 Chapter 5 y). (2) C(I) = � 1 3 27 In this case.IU) I :5 Klx . This will continue to be true even if we modif the vector field near ° so that it is never 0. On the other hand. as 1 + . but never reaches. given by I(a) = a2/3 . Notice that I(a) a2/3 is not Lipschitz. or (J) for all I. i s not differentiable. can be defined f all I . 0. since = = 0.1(0) 1 :5 Kl x l for x near 0.y E U. but this has nothing to do with the case. the function I. . . 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. written classically as I . There are two different solutions with the initial condition c(o) dy = y2/3 dx C(I) = ° f all I. Unique ness will always be insured when I : V + �n is C I .. in fact. there is no K with I/(x) . and as 1 + 00 it approaches. . . the curves c are (which we missed by dividing by the integral curves of Notice that no integral curve.y l for all x.
lhat is. . XI). a f(xo). . and Jet f : M + M be a "contraction". suppose there is some C < I such that p(f(x). '"v' n times Thus P(Xn . + Cn Since C < ] . . i. f then shows that . y E M. Xn+l) + . so C n + . + p(Xn+k I . j(y» Then there i s a unique x E "fixed point"). .l )p(XO. +C n+k I + Thus the sequence {xn } is Cauchy. Xn+1 Then an easy induction argument shows that = fn (xo) = f a f a .y) for all x. A Lipschitz function is also clearly bounded on any bounded set. that is. On the other hand. THEOREM (THE CONTRACTION LEMMA). Xn+k l +k. . the sum L::'o e n converges. it satisfies a Lipschitz condition in a neighborhood of each pointthis follows from Lemma 25. � (C n + . I . .. Xn+k ) � P(Xn . Let (M. f(x) = Ixl). p) be a non empty complete metric space. .e.rector Fields and Dijferential Equations 139 A Lipschitz function is clearly continuous. . but not necessarily diff erentiable (for example. 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 ). The basic existence and uniqueness theorem for differential equations de pends on a simple lemma about complete metric spaces. so there is some x with Continuity of 0 as n + 00. P ROOF NOlice that f is clearly continuous. a C l function is locally Lipschitz. � Cp (x.
clearly satisfies V V V. then a continuous function <> : (b. plus the fact that each ). If c + �n is continuous. xeX (J (f. if M is a m n oo compact subset of �n . if M is complete. Recall that if (M. then we do not even need X to be compact. . Moreover. b) + �n and I : defined on some interval around 0. and thus f continuous functions. if <> satisfies (J). so l' 1(<> (u» du. then the set of all continuous functions I : X + M is complete with the metric (J(f. then <> is diff erentiable. we need only one simple estimate.gil.1 40 Chapter 5 We are going to apply the Contraction Lemma to certain spaces offunctions. thus af = f 0 ex is continuous. 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. hence for step functions. g) = sup p (f(x) . 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. (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. I (x) exists since M is complete.x = <>(/) . this is basically just the theorem that the uniform limit of continuous functions is continuous.<>(0) = For the proof of the basic theorem. If a continuous function I: [a. we note that it is true for constant functions. g(x» . hence continuous. then the new metric space is also complete. Conversely. It I(U) dUI :5 K(b . In particular. bJ + �n satisfies 1 /1 :5 K. <>(/) . which are unif or orm limits on [a.a). then l' <>'(u) du = l' 1(<>('1» duo To prove this. b) of step functions. p) is a metric space and X is compact. g) = II I .!.
for any I E (b. Now suppose a. Let I : + �n b e any function. Moreover. Sa is clearly The curve I Sa(l) .b) by M = { continuous a : Sa(l) = x + (b. b) we have l' I(a(u» du d bL by (I) E2a (xo)}. for all I E (b. Then for each x PROOF. a E M.y l for x. < :s a Since Ix . is contained in U. H Thus S: M M. fJ E M. For each (b.(u) 1 la(u fJ .l(fJ(U» dU I by (2) = bKlIa . Let E Ea (xo) there is a unique ax : (b. b) V such that a/(I) = I(ax(l» ax(O) = x. Choose b > 0 so that (I) (2) (3) b :s aIL (4) b < 1 1K.xo l 2a. define a curve Sa o n . b).b) + (the integral exists since I is continuous on E2a (xo» continuous.x l = < If I(a(u» u l by (3) .fJ lI · . of radius 2a and center xo. Let Xo E and let a > 0 be a number such that the closed ball E2a(xo). which will be fixed for the remainder ofthe proof + Then M i s a complete metric space. THEOREM. Then + li Sa .Vector Fields and Diff erential Equations 141 2 .b). it follows that ISa(l) . so Sa(l) E B2a (xo) c E2a (xo) for i E (b.I(Y)I :s Klx . Suppose that V V V 1 / 1 :s L on E2a (xo) I/(x) . where c �n is open.SfJ l = s�p < bK sup h<u<b If I(a(u») .xo l :s a. Choose x E Ea (xo). Y E E2a (xo).
1 42 Chapter 5 Since we chose bK < I (by (4» . •:. (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. We can now use a simple least upper bound argument. < (J (u) E B2a (XO) = {I : O ::s 1 for all u with 0 ::s u ::s t . This clearly implies that (J (a + E B2a(XO) f or sufficiently small > 0. We have already seen (statement H) that for each 1 with 0 :s 1 < b. So it must be that sup A = b. So (J(a) E B2a(xo). the unique fixed point ax of the map S is the unique curve with the desired properties. < Let a = sup A . in B2a(xo). this shows that Hence S has a unique fixed point: There is a unique S : M + M is a contraction. {J : (b. To sum up. Suppose a < b. which contradicts the fact that a = sup A. We clearly have (J (u) E B2a(xo) for O ::s u < a. in fact. b) + This. 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.b) + B2a(xo) with a(l) = x + 10' f(a(u» du. a : (b. Having used the elegant Con traction Lemma. alas. by (**). Let A < b and (J (u) E B2a (xo) for O ::s u I } . is in V satisfying Reason: We claim that any such (J actually lies in B2a(XO). is not quite what the theorem states. A similar argument works for b < 1 ::s O. Consider first numbers t > O. s s) .
. We will write ax (I) as a(l. (Jt'(I) = /((J.X) = /(a(l. (/) .vecfor Fields and Dijferential Equations Notice that solutions of the differential equation 143 a' (I) = /(a(I» remain solutions under additive changes of parameter. This remark allows us to extend the uniqueness part of Theorem 2. (lo + I).a2 (/) c V and Then a l = a2 on I. Suppose /: V + �n is locally Lipschitz.(/) a. by Theorem J. •:.e. 2. that is. x) a(O.x) = x d di a(I. (Ji (I) = a. so that we have a map {I E I : ad/) = a2 (1)) a: satisf ying (b. alai or dIdl in this . Let x E V and let a" a2 be two maps or on some open interval I with a l (/). Thus the set is open. a/(I) = /(a. Dl a(l. if iJ(l) then = a (10 + I).x). It is clearly also closed and nonempty. so it equals I.(O) = x i = 1 . b) x Bp(xo ) + V [i. Ifwe define PROOF. 3. Suppose a l (10) = a2 (10) for some 10 E I. 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). THEOREl'v!. (O) = a l (10) = a2 (10) E V. (J'(I) = a'(lo + I) = /(a(lo + I)) = /({J(I)). but we will frequently use /(a(l. We now revert to the situation in Theorem 2. then the functions {J i satisfy the same differential equation. that is. Hence {JI (I) = (J 2 (1) for sufficiently small I.x» x» . and have the same initial condition (J.
1 44 Chapter 5 discussion] . If f : V + �n is locally Lipschitz.: <Pdx) = Ci (I . Let us denote the map S defined in the proof of Theorem 2 by Sx. the best we can do is to draw the images of the integral cUIVes Cix. to indicate explicitly the role of x. the whole flow Ci is continuous (as a function of both I and xl: 4. 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. so the two images overlap. If y = Cix (/O). This map Ci is called a local flow for f in (b.X) gives the result of pushing each x along the integral curve through it. then the flow Ci : (b. In fact. for a time inteIVal of I. b) x Ba(xo) + V given by Theorem 2 is continuous. This map <p. To picture this map Ci.b) x Ba(xo). then the map x >+ Ci (I. is always continuous. Then .X) for b < 1 < b lies along part of the curve through x. THEOREM. On the other hand. X) [ = Cix(/)]. PROOF. To focus attention on this map. we denote it by <p. if we fix I. For each fixed x. the map 1 >+ Ci(I.
S.: I 1 _ bK l x .). 360365).Y I · .SI'l I s bKli a .YI · I ay of Sy is the limit of S. pp.S. denotes the nfold iterate of Sy. + (b K n. but this is probably easier than reading the classical proof (and.ay ll S 1 Since Ilax ayll = sup la(l. 371379.T!ector Fields and Differential Equations Recall that 145 li Sa .S.) Ix . If additional conditions are placed upon the map /. x) a(l.I'l ll · If S.b) x Ba (xo) + V satisfies a (O. In order to read this highpowered proof. Notice that the maps <P.). Real and Functional Ana!Jsis (3rd ed. .x) = x. we have . including the inverse and implicit function theorems (Real and Functional Ana!Jsis. . and then read about diff erential calculus in Banach spaces. A clean exposition of the clas sical proof is given in Lang's Introduction to Diff erentiable Manifolds (2nd ed. this is a very hard theorem. so we obtain � Ilax . then further smoothness conditions can be proved for a.I ax .axll + . b) x Ba(xo) + V is also C k Unfortunately.axll 1 ) I S (I + b K + . besides.axll S lIax . + IIS. and a recently discovered proof can be f ound in Lang. then Recall also that in Theorem I the fixed point any Ct . .Syax ll + I Syax . . then tJlej/ow a : (b . and differential calculus in Banach spaces).a for lIax . when you're fin ished you'll also know about Banach spaces. If f : V + �n is Ck . are consequently Coo if f is Coo. Hence Cl)' = /� S.yl s l _ bK l x . . pp.. cxx. y)l. We will just accept this fact. this certainly proves continuity of a . you must first learn the elements of Banach spaces. In fact. including the HahnBanach theorem. Since the map a : (b.
since it is local. Consequently. <p. say: (J'(I) = f({J(I» (J (O) = Ci (S. (x) E Ba/2 (xo). Ci(S.X). then <Ps(<P' (x)) = <ps+.e) x Ba/2 (XO) + Ba(xo). then the integral curve with initial condition x stays in Ba (xo) for III < e.X» f 1 1 1 < e. X). defined f or Is + I I < e. that for ls i < e each <Ps is a diffeomorphism.X) E Ba(xo). or if I sl.I = <ps. (J (I) = Ci(I. .X» = Ci(S + I. <Ps = <Ps <p.Ci (S. X» III < e. II I. with inverse <Ps .. Roughly speaking. and x E Ba/2 (xo). III. If we now let <p. \I\�thout requiring any more proof.] So if l s i < e. l s + l l < e. x) f x E Ba/2 (XO). then Ci(I. In other words.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.X). (x). Is + I I < e and x. We have also noted that satisfies (J(I) = Ci(S + I. so we can also define y(l) = Ci(I. then the point Ci(S. on a manifold. <P'+s = <p. in particular. can be resaid. This shows. : Ba/2 (xo) + �n b e <p. we can or ls i. Everything we if a a have said. This satisfies y' (I) = f(Y(I» y(O) = Ci(S.X).Ci(S.(x) = Ci (I. [If x E Ba/2 (XO).
= is Coo.0 0 0.. write or <PI <PI 1 q <Pdq) {<Pi(q) q q <pi (q) (q) q if <PI+S <PI o <ps = 11 1 . III. and M. then there are diffeomorphisms M + M for all I E � with properties (I). and I I" I < ej2. and let p E M. (2). If X has compact support (in particular. or on all of M. en and diff eomorphisms <p. iterated k times] iterated k times] for k 2: 0 for k < O. 6. ls l . Notice that by uniqueness. 11 + s l < e. PROOF... e) x V Isl. (3).e) x M + M is Coo. p) = . Then there is an open set V containing p and an e > 0. " " V given by n Theorem 5 with con'esponding el . this can be attained. (3) If The examples given previously show that we cannot expect to be defined for all I. en ). ./2 <Pr <Pr [[<P. . THEOREl\![. defined by (I./2 <P. The support of a vector field X is just the closure of {p E M : Xp oJ OJ. .. . Cover support X by a finite number of open sets VI . So j we can define if E Vi = . then Xq is the tangent vector at I 0 of the curve I >+ <PI <Pdq).Vector Fields and Dijferentiai Equations 1 47 5. . if '" support X./2 = 0. To define f II I 2: e. Let X be a Coo vector field on M. d E V.:.. = <Pi for E Vi n V . Let It is easy to check that this is the desired {<PI }' . <PI: Clearly <p : (e. . . <P <PI { <P. and and each is a diffeomorphism. if M is compact). In one case however. q E V. ls + 1 1 + < e. ./2/2 ' <P. . 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. then q <P q) <P <Pdp). Let e = min(e l . THEOREl\![./2 <P.0 0 = k (ej2) + I" with k an integer.
(/) dl = df (c (I» dt = (/ o c)'(I). . In the local case of Theorem 5. we obtain a "local I parameter group oflocal diff eomorphisms". The vector field X is sometimes called the "infinitesimal generator" of {<P. l\1oreover. if q lies on the integral a ax ' on U.1 48 Chapter 5 The unique collection {<P. or more precisely.(O. . an ). The first use is to derive a corollary of Theorem 5 which allows us to simplif many calculations involving vector y fields. to say that Xq is the tangent vector at I = amounts to saying that '!!:.n . . U) around p such that X = f (Xf) (q) = Xq / = hlim (<Ph (q) .f(q) . is called a Iparameter group of diffeomorphisms. the map <P. alal ' lo ' . 7.} given by Theorem 6.a'f+_ _ _ . Let X be a Coo vector field on M with X(p) oJ O. ( ". and which also has important theoretical uses. a'. . _ o � 0 of the curve t r+ <P. . This equation will be used very frequently. (q) PROOF It is easy to see that we can assume M = �n (with the standard coor dinate system t I " . M + �. say).} (vector fields used to be called "infinitesimal transformations"). from � to the group of all diffeomorphisms of M. Condition (3) in Theorem 5 can be rephrased in terms of the action of Xq on a Coo function f. 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. THEOREM. Recall that I r+ Thus. . and is said to be generated by X. and p = 0 E lffi. Then there is a coordinate system (x.
h. . an» .uations q 1 49 We compute that for a = (a' . we will use a2. ./(p)] . . . ./(O)] iI+O II = lim � [/(O. . . which we have just proved. . X. .1+ 0 l is more comprehensive. . . 0) . . h . .O = I is nonsingulaJ: Hence x = X. . a2. . . . a2. . This is the desh'ed coordinate system. a (f) = a ' a (f o X) /1 I . .rector Fields alld Differential 1!./(x (a» ] I h /t70 = = /1+0 lim � [/(<Pa l +h (0. . an) = <Pal (0. . . •:. . .I may be used as a coordinate system in a neighborhood of O. O» . for i > X. . = Jim _ [/( x (a' + 0 �l = Jim . ( � IJ ] we can at least compute ( f) = = lim � [J ( X (O. is equivalent to X = a/ax' . . an» . . an).I coordinates of q and the time interval it takes the curve to get to q as the first coordinate./(x (a» ] (Xf)( x (a)). . . curve through this point. To do this. (�l ) a . an). . . a2. (a/at ' ) = X o X . .[/(<PI> ( x (a» ) . . . ) a l o (f o X Moreover. . f it is easy to see that the equation or X. . ./(0)] �l hO h Since X(O) = a/al ' l o by assumption. The second use of the equation = �� I o' (Xf) (p) = lim . Jet X generate <PI and consider the map X defined on a neighborhood of 0 in �n by x (a' . an as the last n . [/(<PI> (p» . . 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 ./(x (a))J h l + h . a . . this shows that X. .
I (LxY)(p) = }. . oh The vector field <Ph . we can define the Lie derivative of Y with respect to X. This is the limit of certain members of Mp' .(<Ph . If Y is another vector field. whose value at p is denoted variously by (Lx f)( p) = Lx / ( p) = ( X/) ( p) = Xp (f).. we first introduce the notation for Xf. the Lie derivative of w with respect to X. We call Lx / the (Lie) derivative of / with respect to X . (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 appearing here is a special case of the vector field a.. by (Lxw) (p) = . .' (p) = <P. Recall that if Xp E Mp. integral curve / . then A fairly easy direct argument (Problem 8) shows that this limit always exists.150 Chapter 5 Lx / obtained in a similar way. we define a new covariant vector field. I l�o h [(<Ph' W) (P) . Now if w is a COO covariant vector field. Y)p]... and these facts will then be obvious. Y defined at the beginning of the chapter. but we will Soon compute this vector field explicitly in a coordinate system. If a : M + N is a diff eomorphism and Y is a . it is another function. Y)p = <Ph' (Y¢_h (P» is obtained by eValuating Y at <P. To emphasize the fundamental similarity of these notions. and then moving it back to p by <p". Thus (<Ph.w (p)]. <P. and that the newly defined covariant vector field Lx w is Coo.h (p).� [Yp . for a : M + N a diffeomorphism and Y a vector field on M.
(P) · + h�O !(P) .(pl] [ ] .<Ph* Y¢_h (pl] h !(<P (P» A. the one used in finding (/g)'(x).(<Ph* !Y)p]  = lim = h70 ."ctor Fields and Dijferential Equations vector field on the range 151 N.[. 8.<Ph* (/Y).Il [f(p)Yp . The remaining equations are all proved by the same trick. 2. Nevertheless. + W2) = Lx WI + Lx W2 .h h _ /t". [f(p)Yp . we will stick to the original (equivalent) definition. if w(Y) denotes the function p r+ w(p)(Yp) and Lxw and Lx Y exist. [Yp . If Lx Y and Lx w exist. The cal culation is made a lot easier by first observing (a .p = h�O !. We will do number (3) here.(<Pk. (4) Lx ! · w = X! · w + ! · Lx w.!(<Ph (p» <Ph* Y¢_I.(<Ph h h� h [ ( ¥h )p .'!. [Yp = lim . Y r 'f'!t* ¢I. PROPOSITION. a * (Y) is just * Y)p r I r I ' *y A. (pl] lim = 1170 !(P). then (3) Lx!Y = X! · Y + ! .O l lim _ h I . If (I) (2) then (5) Lx (w(Y» = (Lx w)(y) + w(Lx Y).(A. '¥ = k�O k [Yp .!. + Y2) = LXYI + Lx h Lx (w.I ) * y.k* y)p] Y ] r Yp . then a vector field a * Y on M can b e defined by Now notice that Of course. k7o k .Y)p] = (LxY) (p). We now wish to compute Lx w and Lx Y in a coordinate system. LxY. LX Yi and Lx Wi exist for i = 1 . Finally. then Lx (Y. lim (Lx !y)p = 1170 I [(/Y)p . PROOF (I) and (2) are tri\�al.
(p). I Xl I � a(x' O . V) on M. thus a2 A /ahax j = a2 A /ax j ah . Then Lx (dx')(p) = l�o h [(</Ih*)dx' (p) .  . Recall (Prob lem 41 ) that if f : M + N and y is a coordinate system on N.j=1 .] a(x' 0 . Suppose X = L:7= .P X( = . a'a/ax'.o h [( 0 aXl.p /t". n 8ai .. .. the term in brackets = while an easy argument shows that </IMY¢_h(P ) k7 0 lim f(p) . where y is x. Yp .dx' (p)] = lim Now the coefficient of dx j (p) is iJ"'O I � a(x' O </lh ) ' h L aXl· /t7 0 . . It now follows that ( a aa' x') .We first compute Lx (dx'). q) = X ' (</Ih (q» is Coo from � x M y to �. I a( �x�f) dxi We can apply this to </Ih * . but . = To justif *) we note that the map A (h . .(p) dx (p) . aXl. lim � a(x' O </Ih) h ax' .f(</Idp»  k Xf(p). then f* (dy') = t j =1 . [ _ 8' J (*) ] = lim = It. Lx dx' = I: axj dx ' .</Ih) (p) h ax' [ _ ] {this step will be justified in a moment} . • + : We are now ready to compute Lx in terms of a coordinate system (x.. which is what the interchange of limits amounts to. a . j =1 We could now use (2) and (4) of Proposition 8 to compute Lxw in general.) . In the second limit.dx' (p) [ . .x' </1. O · a I lim .</10) (P) ax' (x' 0 </10)] .152 The first limit is clearly approaches Chapter 5 f(p) · Lx Y (p).
ax' ax' thus. a n aa' a Lx . If j . because <P Lx oj = Lx [dX' ( a!j )] (Lx dx') ( a!j ) + dx' (Lx a!j ) so aa' a dx' (Lx. then Yj is a function.= . and we find that LxY = XY .L. .= Using (3) we obtain a . T!ector Fields h* on vector fields involves one more composition than <Ph* on covariant vector fields.' . The trick needed to deal with this complication has already been used to prove (3).=1 axI n .b'... YJ... aaj ) axj ' Laxl Lx This somewhat complicated expression immediately leads to a much simpler coordinatefi·ee expression f or Y. a a Lx (b'. Summing over j and then interchanging and obtain '�I i j in the second double sum we '�I LXY = L j=1 .j .· ax' axj axj ax' 0= . axj + b'Lx ( axj ) n . so XYj = X(Yf) makes sense. abj a n .) .=1 a b � (� a'. but there would be a complication. M + N is a Coo function.YX.and Differential Equations 153 we are really interested in computing LxY.a The second partial derivatives which arise here cancel those in the expression for Y(Xj). and (5) of Proposition 8. aa' a = L a'. = . (4).aax. . To compute Lx ( alax' ) we could imitate the calculations of Lx dx'. = . and we can now use (5) to get the answer immediately. also denoted by [X. Clearly x = L a'.L b' ..· axj "1 axj ax' . axj ) = Lxb'.
: . but they have Gcen proved equal using a coordinate system. Y]p(g) + g(p)[X. PROOF. 9.154 Chapter 5 [X. Y] have both been defined independently of any coordinate system.(sl) ds.e) x M + � is Coo and I(O. so that [X. Two vector fields Lx Y and [X. Y]p(/). note that this means = Xp (Yf) . p) ds r' a . Y] (which is called the "bracket" of X and Y) is just defined as XY . A straightforward verification shows that [X. LEMMA. Y]p ( /g ) = l(p)[X. If I : ( e. Fortunately. We are now in a very strange situation. e) + � with g(O) = /. Define g(l.e) + � with 1(0) = 0 can be written 1(1) = lg(l) for a Coo function g : (e. and can theref be considered as a member ore of Mp. (0). In Chapter 3 we proved a lemma which for the special case of � says that a Coo function I: (e.p) then there is a Coo function g : (e. namely g(l) = l' /. •. Y]p(/) Often. [X. though hardly obvious.Yp (Xf). p) al & (0. p) = g(O. e) x M + � with = 0 for all p E M. p) = Ig(l. 1(1. pl . This sort of thing irks some people to no end. This has an immediate generalization. Y]p is a derivation at p.YX. p) = 10 as I(SI. in this case the coordinatefree proof is short.
Lx X = O. Ly D. so sO [X.l�o (YghH<Ph (p» = (LxYf) (p) .(P)(f + hgh ).YX reveals certain f acts about which are by no means obvious from the definition. PROOF. *:. Y] = [Y. [Z. This equation is capable of two interpretations in terms of Lie derivatives: (a) Lx [Y. [Y. there is a f amily of Coo functions g. + 1 0. [Lx. go = Xf. If X and By Lemma 9 Then (<Ph* Y)p(f) = <Ph* (Y¢_I. + bY2 ) = aLx Y. X] = o. a straightforward calculation proves the 'Jacobi identity": [X. Yll + [Y. Consequently. diately that L is also linear with respect to X: LxY = LyX. Z] = [Lx Y. then LxY = [X.(<Ph* Y)P](f) = h70 � [(Yf)(p) . Lx Z].·ators on Coo functions.(P» (f) = Y¢_I. Y.[Yp . Y = aLx.(YfH <Ph ."ctor Fields aud Differential Equations 155 Y are Coo vector fields. = f + lg. Let f: M � be Coo.YI = Lx a Ly . Finally.Yp(Xf). (P) (f a <Ph) = Y¢_I. [X. Xll = o. so I (p» ] Jim lim h /1 h70 . +bX. Clearly [X. X]. Z] + [Y. on M such that f a <P. + bLx Y2. Y + bLx. Y] = XY .Ly a Lx (which might be written as . Y].[. Lax. we have L[x. THEOREl\I[. (b) as ope. Lx Y The equality Lx Y = [X.(YgO) (p) = Xp(Yf) . Zll + [Z. it follows imme Since we obviously have Lx (aY. Let X generate <p" II I < e.
y(t) = <P. Y]p(f) = Xp(Yf) Yp(Xf) the first term Xp(Yf) is zero. If X generates <Pt. Thus </>.(p) . YJ. gY] = fg[X. [X. even if Xp = 0.. but the second may not be.. although not a tensor. (p).g(Yf)X. Y]p = Oin the formula [X. it does not necessarily follow that [X.. the bracket operation [ . at time I = 0.X)p = Xp for all h . and thus the tangent vector. Proposition 8.(p) is just the tangent vector at p <Ph (P) time t = h to the curve t >+ to the curve <P. 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?). Y]. note that Lx Y is linear over constants only. Y] + f(Xg)Y . ] is no t a tensorthat is. Y].X¢_.* X¢_Ii (p) is the tangent vector. 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. pops up in the definition of prac ticaJly all other tensors. but on the vector fields X and Y. to the curve But this tangent vector is just Xp. Thus. for reasons that \vill become more and more apparent. h(p).. The bracket [X. In f act. for Xf may have a nonzero derivative in the Yp direction even though (Xf) (p) = o. shows that [fX. or a simple calculation using the definition of [X. . it certainly suffices to show that «Pt.156 Chapter 5 Finally. X)p = <Ph. Now X¢_. (I)  LxX = 0. not over the Coo functions F. at time I = 0. Bef ore proceeding to examine its geometric interpretation. Recall that (<Ph. In particular.
Y generate {"'d. consider the curve c : (e.0 h I . . suppose that [X.I (q)](f) = X. of course. Let a : M + N be a diffeomorphism and X a vector field on M which generates {<PI}. or Conversely. conversely. [ f(a 0 <Ph 0 a 11".X)q (/) = [a. so Lx Y(p) = O. LEMMA. PROOF We have (a. Y [X.I }.l.T4!ctor Fields and Differential Equations 157 Since Xp = 0. Then a.X. an integral curve starting at some other point can never get to p). Y] = 0. Then Yp = 0 and (2) If Xp and Yp are both 0. If this by Corollary f all or Given q. then <p Y '" " is true f all t. !. X = X ifand only if <PI oa = a o<PI Then PROOF. 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. X generates {a 0 <PI 0 a . Let X generate {<Pd and and only if <PI 0 "'S = "'s 0 <PI for all s.I (q» ] ./(q)] : 1 2. Y] = 0 if 1 2. To develop an interpretation of I I . LEMMA. [X. p E M. then LxY(p) = O. If <PI 0 "'S '" "'s 0 <PI for all s . so that 1 3. then a.(/ 0 a) (a. If a : M for all I. then clearly Lx Y = O. . •. iI70 1 = lim . COROLLARY.I (q ) (/ o a) = lim _ [(/ 0 a)(<p" (a .I (q))) . + M. I (q)) . e) + Mp given by . Y] we first prove two lemmas.
. As in the proof of Theorem 7. for a a We have already shown that if X(p) oJ 0. ex :::::: PROOF. Consequently c(/) = alJ s.158 Chapter 5 I h c' (/) = lim . • . •. <P. and. : c(O). = 1jJ. and [X.C(I)] .. . (p) . by a linear change ofcoordinates. k. (<Ph. The remarkable f act is that the condition [X. for the existence of the desired coordinate system. then there is a coordinate system x with X = a/ax'. • . c' (/). . Y] = 0. as well as necessary. then there is a coordinate system (x. <p. . V) around p such that a Xa = axa on V. Y] = ° is s'!!ficient. that = I . l . so <p" Y = Y. [ a�' ' a�2 ] = a Y = ax2 • 14. k. (p) . so there is no hope offinding a coordinate system satisf ying (*) unless [X. M = �n . . 1jJ.fJ � k. Xp] = ° for I � ex.[c(/ For the derivative. If Y is another vector field. that ex ]. (P)] } using (*) with q = <Pd p) By Corollary = 0. everywhere linearly independent of X. of this map into the vector space Mp we have h_O + h) . If X" . . Xk are linearly independent Coo vector fields in a neighborhood of p. (O) = <p {l�o X [(<Ph. Y)¢. THEOREl\t[. we can assume that p = 0.<p" Y¢_. . Y)¢_.Y¢. . . I = l'!:'o h [<P. (p)] " = <P. a short calculation immediately gives the result 0. then we might expect to find a coordinate system with H a X = ax " However. 1 2. .
. just as before we see that (� I ) � I 0 = 1 X. an) . 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. (0) = a . Xa = � . » . an) = 4>� (4). . an» . then for sufficiently small h we can (I) follow the integral curve of through p for time h. (0. and less important result. (4)�. . . . . ak + . . (0. . . » and our previous argument then shows that . . an) = 4>�. . . . Y] measures. . . . .k I. . . y X y X (4) then follow the integral curve of Y backwards for time h. . . To make use of it.vector Fields alld Diff erential Equations If 159 X. al' Thus x = X I can be used as a coordinate system in a neighborhood of Moreover. Ci = k + I . in some sense.0. . As in the proof of Theorem 7. generales {4>f } ' define X by I x ( a l . (4):. . . ax' •:. . There is a more complicated. . . we ar appeal to Lemma 1 3. ak +I . . . . x .( . which makes this assertion much more precise. . . o al' 0 � l Ci = . . . p = O. . . x ( a l . . ( . ( 3) then follow the integral curve of backwards for time h. n. follow the integral curve of Y for time h.. we can compute that . the extent to which the integral curves of X and Y can be used to form the "coordinate lines" of a coordinate system.0. We thus see that the bracket [X. . . If X and Y are two vector fields in a neighborhood of p. more difficult to Pl"Ove. . X (2) starting from that point.
0) (O.0.1) (b) . 0) (h.(I. Yj oJ 0.0).h) = 1fJd<Ph (P» a2 (I.160 Chapter 5 If there happens to be a coordinate system x with x (p) = 0 and a x = ax " (J) (2) a Y = ax2 ' then these steps take us to points with coordinates (3) (h. O. so that this "parallelogram" is always closed.0. .I). .O. . .O. Even when X and Y are Oinearly independent) vector fields with [X. which means that c' (O) = y'(O)] is the following. If we define 1 5. Then the CUlve PROOF. .1) a. that is. O) (4) (0. h. . . (1. the parallelogram is "closed up to first order". . The meaning of this phrase [an extension of the terminology "c = y up to first order at 0". c is the constant curve p up to first order.I) = a2 (1. .(O. then Moreover. . . c'(O) = O. Let c(h) be the point which step (4) ends up at.1) = a.d1fJh (<Ph (p» ) a. (a) a dl. . C (I) = a.O. h) = 1fJd<Ph (1fJd<Ph (p» » . a2 (0. (I. . PROPOSITION. h . . . h) = <P. .
Thus.al l aa.(O.h» . 0) + DJ (/ a (2) (0.)(O. cal culation. t . the bracket [X.) = Yf o a. Y]p is related to this "second order" derivation. (A more general construction is presented in Prob lem 1 7. but clever. 0) + [Ddf a J)(O. 0)] = DJ (/ a.Xf(p) + Yf(p) + Xf(p) = o e" (O) (/) = (/ a e)" (O). h (/ a c)' (0) = DJ (/ a a. using the assumption e' (0) = 0. is an horrendous. repeated use of the chain rule gives using (b) using (a). 0) + D2(/ a (2) (0. 0) = DJ (/ a a. and (f) give a (/ 0 (') ' (0) = Yf(p) .) It turns out that for the curve e defined previously. which ends the chapter.O)] a Consequently. (d). we can define a new vector e"(O) or d2e/dt 2 l o by A simple calculation shows. The proof.)(O. 0) + D2 (/ a a.aa (f) a(/.aa(/ a.T4!ctor helds and Differential Equations and for any (c) (d) (e) while 161 Coo function f : M + �. and a second addendum concerning linearly independent vector fields in dimension 2. but can easily be skipped. It is followed by an addendum containing some additional important points about diff erential equations which are used later.)(O. e" ( 0) E Mp. O) + D2 (/ a aJ) (O. Until we get to Lie groups it will not be clear how anyone ever thought of the next theorem. (c).)(O. that this operator e" (0) is a derivation. 0) + [Ddf a (2)(0.e) + M with e(O) = p and e'(O) = 0 E Mp. •. : .h) = Xf(adO. . Whenever we have a curve e: (e. a(/ o aJ ) at = Yf o aJ a(/ a (2) = Xf o a 2 t . (e).
) (I.2) (0.0) + D2. Chapter 5 c"(O) = 2[X. THEOREM.)(O.0) + 2D2 (Xf 0 CY.df o CY. J (/ 0 CY.J (/ 0 CY.df 0 CY.2 (/ 0 CY. 2 (/ 0 CY.J (/ 0 CY. O) = Dd.2)(0.2 (/ 0 CY.2XXf(p) + D2.2YYf(p) .)(O.2 )(0. 0)] = 2XYf(p) .2XYf(p) by (e) by (b) and the chain rule by (d) by (a) and the chain rule by (c) and (f)... 0) + 2D2. O) + D2 (Yf 0 CY.O) + D2( Xf 0 CY.2(/ 0 CY.2)(0.. = (/0 CY.)(O.)(O. Y]p .2 D2 (Yf 0 CY.2(/ 0 CY. since (/ 0 C) (I) we have H (/ 0 cJ" (0) = DJ.2[DJ (Yf 0 CY. 0) by (d) XXf(p) ..J) (O. (2) 2D2. 0)] by (d) and the chain rule + D2. 0) + D2 (Yf 0 CY.2[DJ (Xf 0 CY.O) = Y Yf(p) Now (J) by (e) by (e).J)(O..J)(O.) = 2[DJ (Yf 0 CY.) (O. = = = .2)(0. Using the notation of the previous proof.)(O.J)(O. 0)] = 2XYf(p) .2) (0.2)(0.2) (0.Yf o CY.0) = 2XYf(p) .O). 0) = DJ.2(/ 0 CY.0) DJ(Xf 0 CY. DJ. Since (b) gives we have (3) D2.)(O.2YXf(p) .O) + D2.O) = 2DJ (Yf o CY. 0) XXf(p) .. I) PROOF.2) (0.162 J 6.. 0) + D2.J (/ 0 CY..2)(0.. 0) + 2D2.0) by (c) and (f).2)(0.
:. . from 163 we have (4) D2.2 (/ 0 CY(1)(0. D1 J (f 0 CY(1)(0.2 (/ 0 CY(2)(0. 0) == Substituting (J)(4) in H yields the theorem. . 0) = YYf(p) + 2XYf(p) + XXf(p) . 0) + 2D2. by (c) and (f).T4!ctor Fields and Diff erential Equalions Finally.1 (/ 0 CY(1)(0.0) + D2.
x) = ( I .<>(I. 2) = a : (b. a a ata(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). x). X) = /(a(l.X). for some to.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.<>(I.x) = 1 a' (O. s. Define I: (c.X» <>(O. S. One way t o d o this i s t o replace /(<> (1.x» . X) = /(<>(I. a _.c) x V + �n+l i<> (I. x» by /(I. Then there is a flow I(s.c) x V + �n . let alone equations like <>' (1) = 1<>(1). b) x W + � X �n with (a' . at<> (I. To prove this.x) = /(I.x» wherever it occurs in the proof. For the first component function & 1 this means that .s.X» al <>(O. well have required. x) = (s. In general. that <>(IO. /(S. S. There is also a clever trick.x) = s. or else just replace <> by 1 >+ <> (1 .s.x) = x.x» a (O.lo.X) = X.x) = x. we would like to solve equations by where / : (c. one can replace 0 by 10 everywhere in the proof of Theorem 2.
& 2 (I. where g is an 11 x 11 matrixvalued function on (a. In this case If c is any n x n (constant) matrix. consider the special case of a linear diff erential equation a'(I) = g(I) .X» . a 2 ii/& (I. . 1 7. If g is a continuous n x n matrixvalued function on (a. which may have solutions defined only on a small time interval. b) x �n + �n is Coo.X» '" /(s + 1. then the solutions of the equation a'(I ) = g(l) · a(l) can all be defined on (a. b).b). x) = x (by first finding & with & (lo.&2 (I. /(I. we could also have arranged for {J (10.X) = X.S.S. {J (I.!' thus 165 For the second component & 2 w e have &' (I.X» (J (O. alld Diff erential Equalion. s. PROPOSITION. na t by considering the curve 1 r+ (J (I .X» = /(& ' (I. distinguishing them from general differential equations a'(I) = /(I. S.r�ctar Field.s.. x) = /(I. x) '" (s. X) = g(I) ' x. X» .lo. x) = /(&(I. Finally. even if / : (a. x). This remark allows us to prove an important property of linear diff erential equations.x) = s + 1. Then is the desired flow with a ii/{J (I. S. a(I). Of course.a(I)). then (c · a)' (l) = c · a'(I) = g(I) ' c · a (l) so c·a is also a solution of the same diff erential equation.S. b).x).
Pick (J with Then {J (1 * ) "I 0 for 1 * < I. Hence there is c with (c · (J}(I *) = Ci(I*). (J(I) (J(I. Extend it as far as possible. be the least upper bound of the set of I 's for which it is defined.X) = g(l) . c . 1 hus Ci may be extended past I) as c . •: . (J coincides with Ci on the interval where they are defined. x is locally Lipschitz. Ci must be defined for all I with a < I :s 10 . with any given initial condition. b) we can solve the equation. a contradiction.166 Chapter 5 PROOF Notice that continuity of g implies that /(I. in a neighborhood of 10. or close enough to I. If the extended solution Ci is not defined for all I with 10 :s I < b. f I near I. By uniqueness. let I.) "1 0. �. So for any 10 E ( a. (J'(I) = g (I ) . Similarly. .
Every point q in a sufficiently small neighborhood of 0 is on a unique integral curve of XI thl"Ough a point (0. Let XI . aj_ l . where V c �2 is open and p E f(U).. X 1 = o..)o.. 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. is the required diff eomorphism . but have diff erent parameterizations. Given 11 vector fields Xl . aj+l . . in a sufficiently small neighborhood of 0.x2 (q» is Coo. However. . and that X. whose i th parameter lines lie along the integral curves of X. 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..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.. such an imbedding can be found: 18. . whose parameter cUn"S in the i th direction are the integral curves of the X. Its inverse. (O) = (e. . with Jacobian equal to I at 0 (these facts also follow from the proof of Theorem 7). . l. . . Xn defined in a neighborhood of p E M and linearly independent at p . 2 X l (q) The map q >+ (X l (q). 0). . . •:. .X2 (q» we proved precisely this fact in Theorem 7. we know that there is usually no immersion f: V + M with p E f(U). PROOF. for we might not have [X. Then there is an imbedding f : V + M. the curve I t+ / (a} . q is on a unique integral curve of X through a point (X l (q). Similarly. . . On the other hand.. X2 be linearly independent vector fields in a neighbol·hood of a point p in a 2dimensional manifold M. . in the special case of dimension 2. . PROPOSITION. an ) is called a parameter curve in the j th direction. .
or c . then the map (x. and let c be a curve in M with c(O) = p and c'(t) never a multiple of X. tl(y» for diffeomorphisms a and tl of �. we can further arrange that crt) maps to (c(t). f any particular or parameterization c = (c" (2 ) : � + � 2 of C. or X2 . crt»�. . y). Consequently. which gives us considerable flexibility.' (x). PROPOSITION. g(y» takes the diagonal {(x. whose i th parameter lines lie along the integral curves of Xi . y) r+ (c. we can state 1 9 . Then there is an imbedding f : V + M. x)} to C. If. where V c �2 is open and p E f(V). 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. t) = crt). and f which f(t.168 Chapter 5 We can always compose f with a map of the form (x. Moreover. y) r+ (x. for example. C C � 2 is the graph of a monotone function g.
b) + V with Ci'(I) E V for I E (b. (a) If Ci : M + N is Coo. For example. y E M. where V.c) x V x V + �n be Coo. Find a nowhere 0 vector field on � such that all integral curves can be defined only on some interval around O. then Ci. y. but f has no fixed point. x) = x. and let (xo.b) x W + V is Coo.x) : (b. and we are solving for Ci (y. X» Ci(O. x. Ci' (I) = (J(I) (J ' (I) = f(I.y) : (b. (c) If Ci : � + � is Ci(I) = 1 3 . Let f : (c. y) (t) = Ci(I. p) and a function f : M + M such that p(f(x). V C �n are open. Yo) and a num ber b > 0 such that for each (x. 5. then Ci. 3. Ci(I). X is a Coo vector field on N. y). Ci(I). VYe sometimes have to solve equations "depending on parameters") a ii/Ci(I. Yo) E V x V. { Hinl: Consider the system of equations Moreover.[ector Fields and Differential Equalions 169 PROBLEMS 1. and X is a Coo vector field on M. 4. f(y» < p(x. y. x) = f(l. for open V c �n and V c �m. .b) and Ci"(I) = f(l. Find an example ofa complete metric space (M. Ci(I. where f: ( e. y. then there is a Coo vector field X on � such that Ci. y) E W there is a unique Ci = Ci (x. c) x V x V + �n . the equation Ci'(I) = yCi(I) Ci(O) = X. : TM + TN is Coo. b) + V for each initial condition x and "parameter" y. Prove that there is a neighborhood W of (xo. X is not a Coo vector field. y) for all X.Ci' (I» Ci(O) = X Ci'(O) = y. then Ci: (b. (b) If Ci : M + N is a diffeomorphism. {J (I» . y. if we write Ci (x . 2.
x) = x can be reduced to equations of the form (*) (and thus to equations a ai"a(l.a(l.e. x» a(O. W + �m X �n is a flow for j in a neighborhood of ( c . x) = (O. X) (a result which follows directly from the original equation if ! is C 2 . the hard part is to prove that if ! is C ' . D2a(l.x» . then D.! is C ' .x» .x) = !(I.b) x (a' . = D2D.x) = D2/(a(l.. [When one proves that a C k function ! : V + �n has a C k flow a : (b. (a) Define by If Chapter 5 a(l) = xeyt. y. y. then a is diff erentiable with respect to the arguments in W. x) = !(a(l. D. x» · D2a(I. ultimately). y. and that if the derivative with respect to these arguments is denoted by D2a. y. xo).). since D.j(I. if ! is C 2 Diff erentiability of class C k is then proved similarly. x) = (y.a of the fOl"m (**). by induction. y. a(l. a2) = a : (Yo.b) x W + V. x» a(O.170 with solution is such a case. and conclude that a satisfies (*). a (l.x). it follows that D2 a is differentiable if D. c) x V x V + � m X �n show that we can write for some a. Since (***) is an equation for D. y. j: (b. x) = !(I.x» .a(l.x. y.] D. so that a ai"a(l. x) = (y. x) = !(a(l. i. (b) Show that equations of the form (**) a ai"a(l. a(l. x» j(t.
and let IA I denote the maximum of all l aij I.al Equal. (a) Consider a linear differential equation 171 Ct' (I) = g(I)Ct(I). Show Ct(I) = ef g(r)dr . (e) If AB = eA "" I + A + 2! + 3! + 4! + . (d) Show that exp(TA T.[eclor Fields and Dijjerenl. where f g(l) dl denotes some function G with G'(I) = g (one can obtain all positive multiples simply by changing G). where g: � + �. is clearly differentiable (it is even analytic). so exp A is always invertible. p=O =O 2N (A + BJ P ( = (exp A) (exp B). The remainder of this problem inves tigates the extent to which similar results hold for a system of linear diff erential equations. 2 2 (g) The map exp. B).' ) = T (exp A)T' . N AP )( N BP ) + RN (Notice that for I A I. the usual norm of A E = B (= exp(O) . then exp(A + B) Hint: Write and show that I R N I + 0 as N + 00.. 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. (f) (exp A) (exp . �n2 . we have I A I s i A l s n I A I. p=O p.ons 6.= p i L i L p. Show that exp' (O) (B) L . (b) Let A = (aij ) be an n x n matrix. A2 A' A4 j)th = BA. so that we are solving for a realvalued function that all solutions are multiples of Ct. .A) = I .) . entry of the partial sums converge absolutely for each (i. p. considered as a map cxp : �n + �n . ill and uniformly in any bounded sel. .
) (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. 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. defined by al a(f( · . t. ( a/at ') = X 0 x. show that for every Coo function I : M + �. q) = � (p). show that B'(t) = A'(t) · exp (A (t» . Check that if the coordinate system x is x = XI . �n + M. q). (b) If <p : (B. If (x.172 Chapter 5 (h) Use the limit established in pan (g) to show that exp' (A)(B) = exp(A) · B if A B = BA . provided that A(t)A'(t) = A'(t)A(t). let 1( · .B) x M + M is a I parameter group of diff eomorphisms. (This is clealiy true if A(s)A(t) = A ( t ) A (s) for all S. 2 erentiable. for x : X= a/ax' is equivalent 10 X. For a Coo function I : M x and q E N.) (fo' g(s) dS ) 8. q » (P . axi is a Coo function on M x N. show that the function al/axi. U) is a coordinate system on M. If B'(t) denotes the matrix whose enllies are the derivatives of the entries of B. then N + � p >+ I( p . (This certainly happens when g(t) is a constant matrix A . and let (i) Let A � + �n b e diff : B(t) = exp(A (t» . the limit . t . q) denote the function from M to � defined by 7. (a) Let M and N be Coo manifolds.
Give the argument to show that tion 8. unnecessarily clumsy way.. 4>h* Y"'_h(P) + Yp in the proof of Proposi 10. If X generates {4>. Lx that for Y E Mp. e) show that 4>* is Coo. (a) Prove that Lx(f · w) = Xf · w + f · Lxw Lx[w(Y)] = (Lx w)(Y) + w(LxY). x TM + TM is defined by (c) If 4>* : 1 73 (e. Y]. v) = 4>. In the next part we get a much simpler proof that Lx Y = [X. and conclude that for every Coo vector field ant vector field w on M. and conclude that 4> * (df)(Y) = Y(f 0 4» . }. using the technique which appeared in the proof of Proposi tion 15. show directly from the definition of II. and defines a Coo function on M. define Lx df = d(Lx f). (b) How would Proposition as 8 have to be changed ifwe had defined (Lx Y )(p) (b) Using (a). (a) Show that [Lx df(p)](Yp) = Yp(Lx f). X and covari exists and defines a Coo function on (d) Treat Lx Y similarly. i . (c) Let X and Y be vector fields on M. derived in the text. isjust a special case derived in an The formula for Lx dx . (v) . and f: M + IR a Coo function. 9. the limit 4>* (1. M.Hxtor Fields and Diff erential Equations exists.
h) we have . 1 . . c) E 1R3 1R3) 1R3 14. on M. (b) Show that the flow of aX + b Y + cZ is a rotation of about some axis through o. and A I .Xp( Yf) D2a(0. </>* Vk . . . .c' (O) 1 2. we define M generates {</>..A(p)]. A I > . O) = . ).1 74 Show that Chapter 5 Dl a(O. On 1R3 let X.. . Yilz a az a a . . . V] r+ the crossproduct of the images of U and V. . .+ x ax a X + b Y + cZ is an isomorphism (from a certain set of vector fields to and that [U. 13. (</>. Vk E Mp. (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. 0) = Yp(Xf). . and A is a tensor field of type (J) (�)J I (LxA)(p) = "_0 _ [(</>. Conclude that for c ( h ) = a(h. If A is a tensor field of type on N and </> : M + N is a diffeomorphism. . . If V I . . . Y]p(f). b . . Z be the vector fields a X = Zay a = Lx Y(p)(f) = [X. A[ E Mp* . lim . .1 ) * A I > . Vk . . . . Z=y ax ay (a) Show that the map a Y = z. we define </> * A on M as follows. Y. A[) = A (</>(P»(</>* V I . (r l ) *A[). r+ (a . . . then (�) [4>*A(P)](VI . Check the Jacobi identity.x. .* A)(p) .
. . . Xk . . . Lx Wi . Xk . . WI ) can be obtained by applying contrac tions repeatedly to A ® X. A . J ' a i + " " Al J. . +x.. A/ _. Hint: We already know that i t is true for A o f type (d) Let be any contraction (�). Lx(A (XI . X cr LL 0'=1 .. .=1 I . .lk .! ! !cr A' J ···.•laJ11a+J ·'·'k . (c) Show that Lx (A + B) Lx{fA) = X{f)A + jLxA). . a axj n ai a/axi . .+ I . • . . . . .. . WI . WI). .lk' = l . (CT)(v" Show that = contraction of . . WI») = (e) Noting that A (X" . W I . . .=1 + L A ( X1 . Xk . . . . LxXi. Ap+ " . . . A � J . ..' and Diflmlliiai Equations Show that 175 Lx (A ® B) = (LxA) ® B + A ® Lx B (so that in particular). Vk_" A" . Ap_" A. . Xk .!t _ ! ··. .. . . use (d) to show that (LxA)(X" . j n . Vk . .=1 show that the coordinates of Lx A are given by k n .. . .. . . . (f) If A has components Ai ' in a coordinate system and X = L . wI) '':':: a " ai� " " LL L ax' cr=l j=1 ." A" . . W I . ® Xk ® W I ® . . . Lx(CA) = C(Lx A). Xk . = LxA + LxB Lx.). . A + Lx. . x . • . ® w[. k I . ® . . V. WI) + L A (XI . . (�)._ I . . 'J / . . . . . . .. V. . W" . . .=1 n . . . . h � (LxA)IJJ . A /_ ' ) (v . . . . . (�). . aai .. A = Lx.J l}cr+J . . . . ..=1 . V. . . . . .lleclor Field. A) f+ T(V I . W I . . .
(c) Show that ( DA W) ( P) = A ( p) * (w ( p» . (e) If T is of type (�). and (3). such that D : F + F and D : V + V are linear over IR and D U Y) = I . Show that if we define DAI = 0./1 ) for t ::: o. then this unique extension is Lx . (2). Define g (t) = that the righthand derivative (�).) (b) Given c : IR + M with c'(O) = 0 E Mp. (d) Show that LJx = ILx  DXfj!!dJ .g(O) h  _ /. 2 (Use Taylor's Theorem . Let D be an operator taking the Coo functions F to F. 1 (. Show .1 76 Chapter 5 15. then A (X) is a vector field for each vector field X. Hillt: Check this for functions and vector fields first. show that Generalize to tensors of type 16. 0») Let A be a tensor field oft)l)e G). (a) Show that D has a unique extemion to an operator taking tensor fields of type to themselves. _ r g+ (0) . Show that the tangent vector e"(O) defined by e"(O)(/) = (1 0 e)"(O) can also be described by e"(O) = 2y'(0). DC = CD.h. DY + DI . DA X = A (X). define y(t) = e ( ./I ) for t ::: o. If we take DI = XI and D Y = Lx Y. Y. (a) Let I : IR + IR satisfy 1'(0) = o. then DA has a un�'1ue extension satisfying (I). such that (J) (2) D ( A (I) D is linear over ® B) = DA ® B + A ® DB IR (3) for any contraction C. so that we can consider A ( p ) E Elld(Mp).!� g + (h) . and the Coo vector fields V to V.. . (0) .
show that e: 0 c•• (0) : 1R0 x 1R0 + Me(o) takes ( 1 0 . = denotes a function such that 19.L1 L j= "= a' . be the curve of Theorems 15 and 16. (a) If M is compact and is a regular value of I : M + IR. as a critical point. For Xp. x is a coordinate system 0(/2) xi (e(/» ai/ 2 + 0(/ 2). . axi Ip ' a '. p Using the fact that [X. and b J if} a' b J . (a) Let vectors Define I : M + IR have as a critical point. 1 0) to the tangent vector 18.I (U) is diffeomorphic to x u. Y]p = n a f. Y with 51p = Xp and Yp = Yp. (b) Show that I •• n n (LI axa i Ip ' j�1n . 0 0 11 (0) . xa IP) = i.leclor Fields and DijjerClltiai Equations 1 77 1 7. Yp E M and g : N + I (X. Y) g ) = 51p Y g 0 I»· I. Let around pe x (p) 0. choose vector fields X. so that I.. Yp) •• is symmetric.p = O.. (e) If IR + M has as a critical point. = 0. Yp E Mp. If with = and e"(O) defined by e"(O)(f) = (f 0 e)"(O). Y]p(f) conclude that it is welldefined. show that 1 (Xp. : Mp Mp + NJ(p) x ( (( is a welldefined bilinear map. Given Xp. then there is a neighborhood U of E IR such that I. (c) The rank of (d) Let I : M + define Show that (a2fjaxi axj (p» N p have •• a2I axi aXj (p). show that where [X. IR is independent of the coordinate system.
If I is onto N. I)) = I.1 (0) x U + II (U) with 1(</>(p.I (q2 ) are diffeomorphic for q l . show that Proposition 18 does not hold in dimension 3 .q')) = q'. if M is compact and q E N is a regular value of I: M + N. while certain other integral curves arc parabolas in the planes y = constant.I (q) x N? Z b e unit vector fields always pointing along the y . then II (qI l and I. does it follow that M is diff eomorphic to I. In IR ' . let Y and .and zaxes.178 Cha pter 5 by a diff omorphism </> : 1 . and let X will be a vector field one of whose integral curves is the xaxis. 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. 0») More generally. then there is a neighborhood U of q and a diffeomorphism </> : II (q) x U + II (U) with I(</>(p. Using the second part of the figure. (c) It follows f rom (b) that if all points of N are regular values.X = dldl. q2 sufficiently close. 20. respectively. as shown in the first part of the figure below.
[Pioneer work is clumsy] A. if for every p E N we have i* ( N ) = /. so that global results can be obtained. we have seen that the integral curves of a vector field where i : N + M is the inclusion map. . For a l dimensional distribution the notion of an integral curve makes no sense. that is. and then forget about the parameterization of e. Hardy. we can choose (in many possible ways) a vee to!' field X such that 0 # Xq E /. 1 79 . we can always find all integral manifold N of a Coo distribution /. We call /. For a given p E M. the set {q E U : x2 (q) = a 2. This al'gument actually shows that for every p E M there is a coordinate system (x. find an integral curve c of X with initial condition or e(O) = p. Beauty is the first test: there is no permanent place in the world for ugly mathematics. sup pose that for each p E M we have a I dimensional subspace /. by defining N to be {c(I)}. a " .q for all q in some open set around p.q f q in a neighborhood of p. is spanned by a vector field locally.CHAPTER 6 INTEGRAL MANIFOLDS PROLOGUE A mathematician's repu tation rests on the number of bad proof" he has given . . is called a I dimensional distribution (this kind of distribution has nothing whatsoever to do with the distributions of analysis. Littlewood. G. quoted in]. we just choose a vector field X with 0 # Xq E /. A Mathematician's Apolog)' though the vector field is COO on all of M. . .p C Mp . We will now vary our question a little. Besicovitch. A Afatllcmaticia1l1 A1iscellan ). which include such things as the "afunction"). xn (q) an } = .p p I on a manifold M may be definable only for some small time interval. Then /. S. with p E N. even n the previous chapter. • . H. a Coo distribution if such a vector field X can be chosen to be Coo in a neighborhood of each point. Instead of a vector field. U) such that for each fixed set of numbers a 2 . but we define a (Idimensional) submanifold N of M to be an integral manifold of /. The function /. E. .
.180 Cha pter 6 is an integral manifold of !.hkh has olle compact connected integral manif old.1 1 ���l il l l l l l 011 submanif old pictured in Chapter 2.. on U. On the other halld. rather than curves with a particular parameterization. and that these are the only integral manifolds in U. The entire manif old M can be written as a disjoint union of connected integral submanifolds of !. This is still a local result: but because we are dealing with submanifolds. and all other integral man ifolds nOlicompact. For example. which locally look like  (rather than like or something even more complicated). there is a distribution 011 lhe torus whose integral manifolds all look like the dense I dimensional = 1 11 . we can join overlapping integral submanifolcls together. there is a distribution the torus . but on the . . It happens that the integral manif olds ofthesC' two distributions are also the integral curves for certain vector fields..
In general. . . Thus /.P l where D. even locally. 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. if for every we have Although the definitions given so f all look the same as the I dimensional ar casC'.(a. integral manif olds do llot exist. As the simplest example. consider' the 2dimensional distribution D. . .br)p. Xdq) are a basis for /.p where Mp.q.p = /. . . b(x � . Xk such that X.p c= . Xq with this property.moment we \'\'ill investigate higher dimensional cases only locally. for each q U.c) is spanned by E pEN i*(Np) = /. C For any P there is a neighborhood U is a k dimensional subspace of and k vector fields X" . We call /. in �3 for which /. then /. S E } (r. . .p = If we identify T� 3 with � 3 x � J . A kdimensionaI distribution on is a function p 1+ D. . A (kdimensional) submanifold of is called an integral manifold of /. the results will look very different. s. M E M N M i: N M + p Mp is the inclusion map. a Coo distribution if it is possible to choose Coo vector fields X" . .a). Thus /. . in a neighborhood of each point p . ( q).Integral Alanifolds 181 ]\1obiu5 strip there is a distribution which is spanned by a vector field only locally.b.p consists of all may be pictured as the plane with the equation z {r �Ip + s �lp + br� lp : r.
For example. b. This will be a curve in the plane { (x. y.p for points p = (a. b. The plane /. f each fixed Yo. b.(a. The only such vectors have third component 0.(O .. yo. Our intcgral manifold would have to look like the following picture.c) through (a. z)plane. yox)}. its tangent space at ( 1 .z) and the (y.'ork. But this submanifold c10es not . 0). : ) } .z)} would be a curve y in the (y. with 0 E N. you can probably see that it has no integral manifolds.. z)} through (0. Suppose there were an integral manif old N of /. 0) contains vectors with third component nonzero. 0). Now consider. O).b. 0.182 Chapter 6 The figure below shows /.y. The intel'section of N and {( O. so Y lllust be the yaxis. . � � � � � � � � � � If you can picture this distribution. or the intersection N n { (x. Yo. with all tangent vectors having slope Yo. a proof can be given as follows. yo. :)plane through 0 whose tangent vectors would have to lie in the intersection of /. s o i t must b e the linc { (x. c) isjust parallel to the one through (a. yO.
z) z = a(x. b).c) through (a.p = �{ + s � p + [rf(a.b)(x . b.p is the plane with the equation 0 . As in the first example.(a. y)}. ay = g.b)] Ip : r. has an integral manifold N through each point.p is never perpendicular to the (x.a) +g(a.b. the plane /.b) .b).Integral Manifolds 1 83 p /.b.b) (y .c) = D. (a.b.p if and only if I I aa a a . y) plane. + a. b) if ' p p . S o we need t o find a function a: 1R2 > IR with aa � . To see in greater detail what is happening here. aay ay • = (a.Ip + . Since /.. c) will be parallel to the one through (a. since f and g depend only on a and b..b) = a.(a.f.s E IR . 0).b)) is spanned by a aa a a. ax These tangent vectors are in /. : p Now the tangent space at aa f(a.b).(a.b) + sg(a. (a.. geometrically. aa g(a.Ip .b) = ay (a. is r· l � } We now ask when the distribution /. consider the somewhat more general case where D. b.c = f(a.y. the submanifold is given locally as the graph of a function: N = ((x.a(a. /.
= b. aa (x. and ':0 E neighborhood of 0 E 1R2 . b) = 0. we find a necessary condition on f and g: In our previous example. then there is a function a(O. It is also wellknown that the neces "JI'Y condition (**) is sl!lJirienl fOl' the existence of the function a satisfying (*) in a neighborhood of any point. O).O) = :0 �=I ax aa ay = g. 0) so that a(O.. 0) = l(x. al = ay ax I. such that �. I 0'. O. I(a . we define foX 1(I.1 84 Clzapter 6 It is wellknown that this is not always possible. = 0' so this necessary condition is not satisfied._0' . 0) = :0 (I) a. a(x. '!§.O) = :0 + namely. By using the equality of mixed partial derivatives. and u. O) dl. If I. PROPOSITION. defined in a PROOF We first define a(x. g : 1R2 + IR satisfy in a neighborhood of 0. b ) g(a.
I) dl 1 y This is 0 for ). = 0 by (I) . y) so that (2) aa a:. y).p {r a�{ + s � Ip + [rl(p) + sg(p)] � Ip r./(x. z) : z = a(x. namely. y. = �3 + �. then also aajax = f. Suppose that N = : E { (x. consider. To prove that it equals 0 for all shO\\' that its derivative is o. and always provide us with an a satisf y� ing (2). y). g : /'. 0) + = Zo This construction does 1)ot use (**). 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) = g(x.y) = a(x. s �} . I) dl.Integral Manifolds Then. 0) dl + 1(" g(x. ax y.. 0 0 laY g(x. (x. To prove this. y) . the function + r 1(1. we define a(x. for each 185 x. aajay = g. But its derivative at y is f+ aa (x. we define a(x. We claim that if (**) holds. for each fixed x. y)} .
b» ag ag = � (a. b» is spanned. by aa a a a. a (a . b» . a(a. The tangent space of N at p = (a. b» = = aa � (a . b.a(a. b. b) = � (a. ' + � p � (a. b» . once again.(a. b. which means that for each ) pair (a . y y • These tangent vectors are in /. b . In order to obtain necessary conditions for the existence of such a function a) we again use the equality of mixed partial derivatives.. Now we are looking for conditions which will be satisfied by I and g when there is an integral manifold of /. b) is. b. since it still involves the unknown function a. a (a. b» g(a. b» + . a(a . b). I(a. b .b» + ilz (a. b) ay aY az ay a x II ag a 2a ag aa (a.186 Chapter 6 is an integral manifold of /. a(a. b . but we can substitute from (*) to obtain al al . .b) p aa a a a p + a (a. a . through . b» g(a. b» + ilz (a.a(a . b.p if and only if I I I I I(a.a(a.a(a . b) these equations must hold no matter what a(a.b.a(a.a(a . b» · . b).(a.ver ' point. b. b. b).(a. (a. a x ay This condition is not very useful.(a . b» + ilz (a. a(a. b. b) = . aa a. b) a: p . (a. a(a. b» . Thus we obtain finally the necessary condition . Thus (*) and the chain' rule imply that al a2a al aa . b. a (a.
In the following theorem. . there is at most one function a : W + V . 0. O. J. . . .0. .. . satisfying a(O) = x aa (I. there is no need to restrict ourselves to equations f a single or function defined on JR2 . . . we can treat a system of partial differential equations for n functions on �m (i. : U x V + JR" be Coo functions. . the necessary condition again turns out to be suf ficient. In fact. agree on the component of WI n W2 which contains 0. defined in a neighborhood W of ° in JRm. . a partial diff erential equation for a function from �m to JR"). .a(l. rom the proof o f existence. . 1. 0».e. Necessity of PROOF Uniqueness will b e obvious f the conditions (**) is left to the reader as a simple exercise. . Then for every x E 11. defined on WI and W2 . . . x) E U V on which x i. THEOREM. . J.l111egral Manifolds 187 In this more general case. . . for i = 1. so for a function f : �m X �n + �k we use al aI' al ax' for for D. any two such functions a l and a2 . . 0) so that (I) a(O. . m . 0) = 11 (1. . . . 0. Let U x V C JRm x JR" be open. with a different twist at the end. . We first want to define a(l. where U is a neighborhood of ° E JRm. . Dm+. . . The proofwill be like that of Proposition 0. 0. we will use I to denote points in JRm and x for points in �n . 0. .) Moreover. . j = 1. 0) = x aa atT (l.a(l» for ali I E W. and we will concern ourselves with proving existence if these conditions do hold. Ill. such a function exists (and is automatically COO) in some neighborhood W if and only if there is a neighborhood of (0. al j (I) = Jj (More precisely. and let j.
. . 0. .I. f32 (1» . In the f ollowing. . f3dl» . . We now derive an equation for g'(I). 0.a(l . . . . 0. (3) O = g(I) = iJiT (I . . . 0» a1 2 11 ' 1 ' " III '2· We claim that f each fixcd I' with 11 ' 1 or we also have. . . defined for III III a(l. . . 0.O. .a(I ' . . then for 11'1 < " the solutions of the equations f f32 with or the initial conditions f32 (0) = a(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.0. � a/. . . . 0) f3/(I) = h (t ' . 0. I.1. for ali I with III aa .188 Cha pter 6 To do this.0) will each be defined for III < '2 for some 82 > o. We have a2a a/. .0. . . O. ' (I) = jj (I. . . . . . < a(l ' . all expressions involv ing a are to be evaluated at (t' . . . . . 0» .O. . . I. . < " . 1. 1. 0) and all expressions involving Ii are to be evaluated at (I'. 1. . At this point the reader must or < < " . O. 0. O) . aak g. and verif the following assertion: If we choose " ). . " . . . .L. Note first that g(O) = 0 by (I). axk a/2 ' k= ' (4) . . . 0. . . (t < < < " < '2. 0. (2) 0) = h(l ' . . I. .a(1 . . .. This has a unique solution f sufficiently small I. . 0) = f32 (1) Then a(O. . . . .0. . . . we consider the ordinary differential equation refer back to Theorem 52. .al 2 . O. (I) = al 2 al' . sufficiently small. 0. . . 0. . . I. Define This equation has a unique solution. . 0». . . .1. . O. We then define fh (O) = x f3. ../. . . . O. 1. 0. O) = x aa (1 .
(5) . The first kind of theorem says that if one has a certain nice situation (e. these Con ditions are obtained by setting mixed partials equal.. a distribution with integral submanifolds through every point) then certain othel' conditions hold.aft2 al by (2) by (2) again _ _ _ by definition. So (3) is true. by showing that the " integrability conditions" afe suffi cient for recovering the nice situation.(I) � aft k L.g. •:.k=1. It is a simple exercise to continue the definition of a until it is eventually defined on (e l . and are called "integrability conditions". in which we will essentially begin anew. The remaining parts of our investigation. Theorem I essentially solves for us the problem of deciding which distributions have integral manifolds. Our investigation of the problem so far illustrates one basic fact about theorems in differential geometry: Many of the fundamental theorems of diff rential geometry fall into e one of two classes. The second kind of theorem justifies this terminology.. without ever even mentioning partial derivatives. en ) and satisfies t*) . given by (4).. is clearly g (t) = 0 for alI I . al . e l ) x ' " x (en . (3) Now equation is a diff erential equation with a unique solution for each initial condition. The solution with initial condition g (O) = 0. and prove their sufficiency. 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 ) . 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.Integral Manifolds and thus 1 89 (5) g .
so Conversely. then Y p) (g) = l Xp (g) . the latter condition is fulfilled: 2.p f( = Xp (g o I). O. Thus if where (i are Coo functions. i f this i s true for all Coo functions g : N + JR.= 1 implics that the functions I. Let I : M + N be a Coo function such that I is an immersion. .I (a .1 90 Chapter 6 LOCAL THEORY If I : M + N is a Coo function. •:. then where The most important property of Irelatedness for us is the following: 3. then X and Y are Irelated. then [X[ ' X21 and . PROPOSITION. f(p) E I (Mp ). . and Y. . (Yg) 0 I = XU o g) . PROPOSITION. . If Y is a Coo vector field on N with Y p. f( each P E M. V) around I(p) E N such that ' t J' 0 l o X. . If g : N + JR is a Coo function. a n ) = (a . we say that X and Y are Irelated if l p (Xp ) = Y p) for . . In One case.. y X= n = 1 . This is easily seen to imply that I p. 0). a. and X and Y are Coo vector fields on M and N. then there is a unique Coo vector field X on PROOF.= 1 . . Xp . Clearly we must define Xp to be the unique element of Mp with Y p) = I . a n . fOI' i [Y" Y21 are Irelated. 0 I = I. . . ' L IY�. I a L a' a. Of course. . M which is Irelated to Y. To prove that X is Coo. are Irelated. a given vector field X may not be Irelated to any vector field Y. IJ = a�. If X. . are Coo (problem 3) . This ( a�. U) around P E M and (y.. . n . . 2. respectively. axl Y= f( p) . we use Theorem 210(2): there are p f( coordinate systems ( x . . nor must a given vector field Y be Irelated to any vector field on M.
then [X.p = { I' f lp + s hip + [rf(p) +sg(p)] �Ip : I'. X and X are i related. 0 + to /. only when the expression in parentheses is 0. then for all p E N there are unique Xp . Thus i. f] p = [X. which is precisely the condition for /. p In other words. Using the f ormula on page 1 56. and Y and f are irelated. and Proposition 3 then shows that [X.{Y2 (Ytg)} f = X. 0 0 0 . (Y2g)} f . this therefore shows that [X.(g f» .f . respectively = XdX. then 1 . if there is an integral manifold of /.X2 ([Ytg] f) by (I).. Y] are irelated. if Xp E /. in 1R 3 given by [X. with g replaced by Y. Proposition 2 shows that X and f are Coo vector fields on N. Y]p. If X and Y are two vector fields which belong to /.Y] = (� .g and Ytg. S E IR } . Here [X.IntegraL JvlanifoLds PROOF. and i : N M is the inclusion map. Y]p E /.. ''''le will say that a vector field X belongs to /.p for all p. ( [Y2 g] f) .. The vector fields X = � + f� ax az a Y = a:ay +gazbelong to /. we see that This belongs to D.(g f) i = {[Y" Y2]g} f = {Y.X2 (X. 191 (Y. •. /. ax ay az az az . 2..Xp. Consequently. Y] also belongs (I) So 0 0 0 0 0 0 If g: N + IR i s Coo.g) f = X. For a moment look back at the distribution /. f]p E Np . Yp = i.p . f] and [X. (g f» by (I) = [X" X2](g f) : Now consider a k dimensional distribution Do. fp E Np such that Xp = i.� + f� _ g af ) � . to have an integral manifold through every point. Suppose that N is an integral manifold of /.[X. through every point p.
. It is equivalent to Theorem I.. ..fiXi i= f or Coo functions C. 5. •. . Theorem I can be derived f rom it (Problem 7). [f..Xi. e) such that for each a k + I .. [jX... X.Y] do : + . . whenever belong to /'. . X. in or fact. FIRST VERSION).gY] = fg[X.J ' Such functions clearly exist if /'.. If is integrable on V if and only if each X" . Since we have clearly [X.. PROPOSITION.Xj] = L C/.. .. be a Coo integrable kdimensional distribution on M. x IV) = ( e. For every p E M there is a coordinate system V) with x . we can clearly write PROOF Y To prove belongs to /'. ) a n with all fa i r < e. [Xi. Xj ]q q X k X = L1 .gY] belongs to /'. if X. . then /'. is called integrable if belongs to /'. This condition can be checked fairly easily: [X. Y] [jX. e). Y] f(Xg)Y g(Yf)X. suppose such f unctions exist...1 92 Chapter 6 In general. . If and belong to /'.. But the proof is quite diff el·ent.. x(p) = 0 ( .. .gj Xj] We arc now ready f the main theorem. since E /'. x n (q) = a n } is an integral manifold of /'.e. Y and [X. l Iich is spanned by the Conversely.. [Xi . Let /'. restricted t o V i s contained i n one of these sets. THEOREM ([HE FROBENIUS INTEGRABILITY THEOREM. . Xj] is a linear combination k [Xi . x (x. . . Xk span /'. . /'. .(q).. is integrable. Any connected integral manifold o f /'. in a neighborhood V of p. the set {q E V : xk + l (q) = a k + l . it obviously suffices to treat each separately. Y] X and Y 4.
11. . Then 1'>0 + IRko is an isomorphism. consider d(xm i) for k + I s.(q) = 0. . k . of Nq we h ave d(xm i)(Xq) = Xq(xm i) = i.Integral Manifolds 1 93 PROOF. which is spanned by the a/axj l q f j = I . we can assume that . Thus or d(xl1l i) = 0.Xq(xm) = 0. with p = O. . By Proposition 3.6. and 1[. 0 . ][* is oneone on f or near O.0 C �no is spanned by Let IRn + IRk be projection onto the first k f actors. (on a neighborhood of 0 E IRn ) and a/aI (on IRk ) are 1[related. k . But.X. .q q 1[. = ax' i = I. So near we can choose unique 1[: D. . k. j j By Theorem 514. with inclusion map For any tangent vector Xq i: N U.(q) i = I. . .. (q) = aaI. The sets {q E U : Xk + l (q) = ak + I . 1[. . . I. [X" X ]q E I'>q by assumption. . ..[X" Xj ]q = [ aaI" aaIj ] . If N is a connected in tegral manifold of I'> restricted to U. •. . x" (q) = a n } are clearly integral man ifolds of 1'>. Moreover. is oneone on I'>q. . . 0 . . We can clearly assume that we are in �n . So [X" X ] = o. there is a coordinate system x such that a X.Xq E I'>q. By continuity. . since their tangent spaces are spanned by the a/ax' = X. which implies that xm i is constant on the connected mani f old N : so that + 0 0 0 m S. 1[. . . : 0. . for i = I. . ' Then the vector fields X. since i. k .
Notice that two distinCI components of N nV might belong to the same leaf of the foliation. Vi ) satisfying the conditions of Theorem 5. 6.). 0) x . PROOF.. we see that we can cover M by a sequence of coordinate systems (Xi . ( () Using Theorem 12. It is possible for a single slice S of Vi to intersect Uj in more than one slice of Vj' as shown below. we introduce the following terminology. . such that the components of {q E V : Xk+l (q) ak+I . {q E V : xk+l (q) ak +I . But S n Vj has at most countably many components. . let us call each set (x. be a Coo kdimensional integrable distribution 011 M. and if around every point P E M there is a coordinate system (x. . 0). . xn(q) an } = = a slice of U. 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. .1 94 Chapter 6 GLOBAL THEORY In order to express the global results succinctly. . x (0. Then M is foliated by an integral manifold of !. V). . If M is a Coo manifold. Let !. . For such a coordinate system V).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 !. THEOREM. . with xIV) = (0..
choose a coordinate system (xo.InlegmL ManifoLds 1 95 and each component is contained in a single slice of Vj by Theorem 5. sO S nVj is contained in at most countably many slices of Uj . i 1 . . Appendix A describes a nonparacompact manifold which is foliated by a lowerdimensional connected submanifold. . since we do not have to find a countable number of coordinate systems for each leaf. i/. will be called joined to p if there is a sequence 0 = io . Since there are at most countably many such sequences of slices for each se quence io. . In this case. we see that the union of all such slices is a submanifold of M. . . . the first union.p. there are at most countably many slices joined to p. . . this di'!ioint union is clearly an integral manifold of /'. A slice S of some V. and only countably many such sequences. Vo) with P E Vo. . Consequently. Given P E M.. il = i and corresponding slices with a = 0. the discussion to follow will not be validin fact. or totally disjoint from. •:. 1 . . For q #.] . Using Problem 31. and can merely describe the topology of the foliation as the smallest one which makes each slice an open set. however. . [If we are allowing nonmetrizable manifolds.1 . the corresponding union is either equal to. and let So be the slice of Vo containing p. . . the proof is even easier. M is foliated by the disjoint union of all such submanifolds.
.. and a f olium of the f olia tion determined by some distribution /'. PROOF. . 7. U) U However.. at most cOUil/ablJ' man slices can belong to the same f y olium. this folium would contain an uncountable disjoint family of open sets. choose a coordinate system f(p) such that the slices MI _ (x. so f takes f(P) MI . 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. According to Proposition 21 1.. then infinitely many slices of may belong to the same f olium. (x. . . . . . Let be a Coo manifold.U) around {q E U : Xk+l (q) = ak +I . otherwise CiV ' . Now f is continuous as a map into M. . THEOREM.xn(q) = an} are integral manifolds of /'. This allows us to apply a proposition from Chapter 2.1 96 Chapter 6 is a coordinate system of the sort considered in the proof Notice that if of the theorem. P M MI . it suffices to show that f is continuous M P MI .
Consequently. we can choose W to be connected. for all p' E W. for any p' E W. This would mean that f(W) contained points of un countably many slices. •:. 0 f would take on all values between a .a. (f(p'». (f(p'» = a. contradicting the fact that f( W) e M. This makes it clear that f is continuous as a map into MI . by continuity. In other words. if we had x . . (f(p'» #. .IntegraL ManifoLds 1 97 some neighborhood W of p into U. f( W) is contained in the single slice of U which contains p. For k + I :s i :s n. x . and x . then x .
' 1=1 f3.. (a) Let � = 1[ : E + B be an nplane bundle. . and show that a kdimensional distribution is Coo if and only ifit is a Coo subbundle. IB) is a bundle map. we say that �' is a subbundle of � if (i. show that 6'1 suffi (b) Complete the proof of Proposition 2 by showing that if so that with a. x = L f3 ' ax. (b) Supply the proof of the uniqueness part of the theorem. U) around each point. 3. . 5. Let /'. 1 . • . 1. . • . (a) In the proof of Theorem I. .. Show that a kdimensional distribution on M is just a subbundle of TM.. 0 f = f3. check the assertion about choosing ciently small. . /'. . such that /'. (a) In the proof of Proposition 2.1 98 PROBLEMS Chapter 6 1 1[ : E' + B a 1. be integrable distributions on Suppose that for each P E M. and IB : B + B the identity map. a Coo. then the functions 4. of dimensions dl .} actually are Coo . define a Coo subbundle. dt. 1 . are n . M. In the proof of Proposition 4. = 2.. (x. show that the functions C. Show that there is a coordinate system i s spanned by aj ax l . . (b) For the case of Coo bundles � and �' over a Coo manifold M. aj a xd" etc. . and �I k plane bundle such that E' c E. . If i : EI + E is the inclusion map.
I) for some function f3 : [0. = j m Ll 11 . show af3 j iU (U. defined by Notice that even when the jj do not depend on x. I) f3(0. al j . in JRm x JR" (with coordinates I. by considering the distribution !. VI) = f3 (UV. with the same initial condition. 7. we can consequently assume that e = 1 . jj (ul. (Show that both functions satisfy the same differential equation as functions of u. x) . f3(u.) By shrinking W. the partial differential equations to ordinary equations along lines through the origin. VI). I).integraL ManifoLds 1 99 6. (b) Show that = x. l) al j = V· af3 ( I . Prove Theorem I from Theorem 5. f3(U. (a) If we want D!(UI) = f3(u. 1) We know that we can solve such equations (we need Problem 55. I)) = . This problem outlines another method of proving Theorem I. A similar technique will be very important in Chapter II. al j ajj al i aft al j ' with the integrability conditions we nevertheless work in �m x �n ) rather than �m. (c) Conclude that af3 (v. since the equation depends on the "parameter" I E JRm). One has to check that one 0 can be picked which works for all l E W. 0) that f3 must satisfy the equation 7. This is connected with the classical technique of "introducing new independent variables". by reducing x W + V. so that the equations are of the form aD! (I) = jj (l).
a' (x. f3(v. y)) = . y)) = ay (or of any point </>'(z) = I(z. = XI . I). a ' (x. a l (x. as functions of v. with any given initial condition 4>(zo) = Woo . and conclude that the (in which f denotes the complex derivative) has a solution in a neighborhood of =0. y. y). (e) Define a(l) = f3 ( I .</>(z)) 20 E q. y . 2. 2.Y �2 in a neighborhood of 0 E differential equation = = � v(x. I). Let I: <C x <C + <C be complex analytic. I)) erential equation. Use (c) to conclude that satisfy the same diff the two functions are equal. y ) . 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 .ay <C aa ' u(x. a' (x. ax aa ' a. Noting that a(vl) = f3(V. If we denote the coordinate functions in <C x <C by ZI . This problem is for those who know something about complex analysis. 8. show that a satisfies the desired equation.200 Chapter 6 (d) Use the integrability condition on 1 to show that af3 (v. I ) al j and v · h ( VI.
CHAPTER 7 DIFFERENTIAL FORMS morc algebraic prclimi. • . if f: V + W is a linear transformation. " Vj" . Vj" " . . [This is not true in the special case of a vector space over a field where 1 + 1 = 0. . . . . It is clear that I1 k (V) c :rk ( V) is a subspace of :r k (V). Vj • . the discussion of which requires some e turn OUf attention once more to tensor fields. . Vi. Vi.naries. Of course. . we have 0 = T(v) • . Therefore. " vk ! + T ( V I " " . Vi + Vj • . Vi" " . . Notice that I1 I ( V ) = :r1 (V) = V* . . vk ! + 0. • . but we will be concerned if Vi = Vj (i #. Let V be an /1dimensional vector space over JR. then T is also alternating. ) Vj. " vk ! + T(VI " " . . . . then f* : :r k (W) + :r k (V) preserves these subspaces f* : I1 k (W) + I1 k (V). skewsymmetry is the same as symmetry. . " Vj " . " Vi" " . f\1osl discussions of the determinant begin by showing that of any two alternating nlinear functions on �n. " Vj" " . It is also convenient to set l1o(V) = :ro( V ) = JR. Vi" " . . An element T E :r k (V) is called alternating if Wwith a special kind of tensor field. vk ! = 0 + T ( Vj " " . T(VI " " . Vi. Vi • • . . . . Vj .T(vJ • . c. . . The most familiar example of an alternating T is the determinant function det E T"(�n). . Vj" " .jl . in a certain sense. and the condition of being alternating is the stronger one. Vi + Vj • • • . . so I1 I ( V) has dimension /1 . vk ! = 0 If T is alternating� then for any VI . if T is skewsymmetric. Uk ) = T(VI " . one is a multiple of the 201 . " Vi" " . in this case. At the moment it is not clear what the dimension of I1 k (V) equals for k > I. " " Vj" . Uk ). • . the most general alternating function. . " . . . Uk ) = . . ' " vk ! + T(vj" . vk ! + T ( VI " . . T is skewsymmetric: T(v) . but one case is \vellknown.onsidered as a function of the n rows of a matrix we shall soon see that this function is. • .] We will denote by I1 k (V) the set of all alternating T E :r k ( V ) . Uk . Moreover. Vi.
the first element. ) is to rename things: V3 = WI ) V2 = Wz. wd = (Wa(I) . . Let Sk denote the set of all permutations of { l. If (v" . . . T 0 a. . . ' " Va(k) ) . Now for any T E :r k ( V ) we define the "alternation of T" I Alt T = k! aeSk L i. Then one proves dim 11"(IR") = I by actually constructing the nonzero function det (it follows. of course. vd = ( Va(I ). ueS" L sgn a . for example. . T(va(l )" ' " Va(k) ). . This definition has a builtin confusion. . . ' . vk! = k! sgn a . . . . that dim 11" ( V ) = I if V is any n dimensional vector space). The construction of det is usually by a messy: explicit formula. is the a ( l )" of the v's on the left side. . .. . . . . . . . .202 Chapter 7 other. I Alt T(v" . vd. . Vp(a(k)) ) since Wet = vp(a)· Thus a . . an element a E Sk is a function i r+ a (i). O n the right side. . . . . k. . (p . we compute by setting so that Vp(l) = W I • . . if these v's have indices running in some order other than 1. . then the first element on the right is 1lot necessarily that v whose index is a ( 1 ) . .1 if a is odd. k}. . dim 11" (IR") :S I.e. . (v" . (VI . . . . . . . . ' . VI. in other words. . Walk) ) = ( Vp(a(l)) ) . vd) = ( pa ) . Thus warned. .here sgn a is + I if a is an even permutation and . VI = W3 • . . . . The simplest way to figure out something like a ' (V3 ) V2 . . \'\. vd ) = a · (wl . (vJ . which is a special case of the definition to follow. Vk ) is a k tuple (of any objects) we set a · (VI . Vp(k) = Wk) a · (p .
f\1oreover. /\ ry w /\ (ry. as we shall soon see. + w. ) /\ ry = w. then Alt w = w. 7879 of Calculus on Manifolds) . Finally. (I) If S E :r k (V) and T E :rl ( V) and Alt(S) = 0. PROOF. associativity of /\ is proved in the following way. The funny coefficient is not essential. (3) If w E I1 k ( V). then Alt(S ® T) = Alt(T ® S) = o. the wedge product of w and ry. /\ ry + w. (3) If T E :rk ( V). Left to the reader (or see pp. ry E 111 ( V). but it makes some things work out more nicely. an element w/\ ry E I1 k +1 ( V ) . then Alt(Alt(T)) = Alt(T). (I) If T E :rk ( V). aw /\ ry = w /\ ary = a(w /\ ry) k + /)! (2) I* (w /\ ry) = I*w /\ f* ry. 203 PROPOSITION.Dijje"n liaI Forms I. then (w /\ 1]) /\ e = w /\ (ry /\ e) = (k + 1 + 111 ) ! Alt(w ® ry ® e). for w E 11 k (V) and ry E 111 ( V ) . e E I1m ( v ) . + ry. it is easy to see that (3) /\ is "anticommutative": w /\ ry = ( _ I) klry /\ w. In particular. We now define. k ! l!m! . THEOREM. It is clear that (I) /\ is bilinear: (w. 2. + w /\ ry. if k is odd then w /\ w = o. by ( w /\ ry = k!/! Alt (w ® ry) . •:. then Alt(T) E I1 k ( V ) . (2) Alt(Alt (w ® ry) ® e ) = Alt(w ® ry ® e) = Alt(w ® Alt( ry ® e)).) = W /\ ry. (2) I f w E 11k( V ) .
. the sum over each being o. . a contradiction.ry ® eJ ) = AIt(w ® Alt( ry ® e» . so ao = a(a. . T(Vk+l > " " Vk+I ) = sgn ao ' ''''le have just gjwwn that this is 0 (since ao . . S(Va(I ) . . ( S ® T)(a . . (k + l + m ) ! Alt((w ) ® e) /\ ry (k + / ) ! m ! (k + 1 + m)! (k + /)! Alt(w ® e).204 Chapter 7 PROOF (k + /)! AIt(S ® T)(Vl . . S ( Va(I).Alt(w ® ry ® e). T(Va(k+l) . Vk+I» ) . Va(k+l) aeG = = o. . Then L sgn a . (3) We have a'eG L sgn a' . . Vk+/» ueuuG [ a'eSk L sgn a ' · S(va'(1) . . . . Va'(k) ] . Vk +l) is just some other (k + I )tuple of vectors). . (a .). (w /\ ry ) /\ e = The othel' equality is proved similarly. . • . . The relation Alt(T ® S) = 0 is proved similarly. Vk+l) = = aeSk+l aeSk+l (1) We have L sgn a . . . Let aoG = {aoa' : a' E G). Suppose now that ao <t G.. then a = aoa' for some a' E G. . " " Va(k) ' T(Va(k+I). (2) Clearly Alt(Alt(1] ® e) . .ry ® e) = Alt( ry ® e) . . :. . for if a E G n aoG. " " Va(k+l) ' Now let G C Sk+l consist of all a which leave k + I. Va(k) . so (I) implies thaI 0 = Alt(w ® [Alt( ry ® e) .I E G. ( S ® T) . . . the other equality is proved similarly. (vl . . (S ® T)(a' . • . .Alt(1] ® e) = 0. = ry ® (k + /)!m ! k!/ ! . . Notice that G n aoG = 0. . Then L sgn a . ( Vb . . by (*) . \!\le can then continue in this . Vk+I » L sgn a . k + I fixed. (ao ' (VI . (v l ."lay: breaking Sk+! up into disjoint subsets. . . .
. . .iJ. . . . !. . 3. . . . because each term in the sum Alt( w @ ry)(VI . . . we would not have Alt(Al t T) = Alt T. and the first equation in the proof of (2) would fail. [If we had defined Alt just like Alt . . . If V I . . ellen over a field a ffinite characteristic.."2k ( V) = {OJ for k PROOF. VII is a basis of V. . I! aeS" L sgn a · (</>I @ · · @ </>n ) o a · . < ik ::: n = 11! k! (I1 .. · I\ </>n )(VI . 4>n is the dual basis. and 1/ k! I! can be interpreted as meaning that these k ! / ! terms are replaced by just one... 4>il ® ' " .] The factor (k + I)!/ k ! / ! has been inserted into the definition of 1\ for the follO\ving rcason. which therefore has dimension () 11 k : . . Vk+l) occurs k ! / ! times (since w and ry are alternating). but without the factor 1/ k!. . If w E !. . THEOREM. . " .) ® 4JiJ. ."2 k ( V) can now be described. iJ. . .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. (</>I I\ . then </>1 1\ " . 1\ </>n = det. Vn i s the standard basis for IRn .k) ! ' (In particular. 1\ </>n = = In particular. The set of all 1 ::: i) < " is a basis for !. L ail . On the other hand. ."2 k ( V). > 11. This makes sense. (So i f V I . . ."2 k ( V) C :r k ( V) . ( I + . then </> 1 1\ . . @ </>n ) I!" . · + I)! Alt(</>I @ " . Vn ) = I . then 1\ could be defined by w 1\ ry = 1 k!l! Alt(w @ ry). we can write W = iJ . the factor 1/ k! in the definition of Alt is essentialwithout it. and 4>1 .) A basis for !.
. . . so it is not o. . . Wk are linearly inde pendent if and only if w. . . . . then W' . . If W" . ik!. < . < }k. On the other hand. I j denote a typical "multiindex" Notice that Theorem 3 implies that every W E S'"2 k ( �n ) is a linear combination of the functions (v" . ® . ® 4>ik ) is either 0 or = ± ( I / k!) 4>h /\ . . so the elements 4>h /\ . If WI " 4. . . . . . .. . Wk E !:) ' ( V). . .i}. . .o. Vn • . . . . /\ Wk #. Then w.ik 4>iI !\ · · · /\ <PiJ. . . COROLLARY. . Then every element of !:) ( V ) is To abbreviate formulas. < }k span !:) ( V). . . Wk are linearly independent. ® . Uk • . /\ 4>h for }. . . . then applying both sides to (Vii " ' " Vik ) gives ail .ik Alt(4)i. /\ . . ·. /\ Wk is a basis element of !:) k ( V) . if then uniquely expressible as k (i" .206 Chapter 7 So W = Alt(w) = iI . x k minor of (:�). . . . . vk! f+ determinant of a k One more simple theorem is in ordel� before we proceed to apply our con struction to manifolds. . ..h L ai. < . 4>k. there is a basis V I " of V such that the dual basis vectors 4>" . ) PROOF. .:. . /\ · · · /\ 4>ik . . 4>n satisfy 4>i = Wi for I :s i :s k . = o. . . .·. . . If 0= i l < " '<ik L ai! . and let 4>1 denote 4>i . it is convenient to let L al 4>l . ® 4>ik ) · Each Alt(4)i. /\ 4>h for some k }. . /\ . .
Vn be a basis for V. . j=1 i PROOF. Define ry E :r n ( JR ) by n . ann») = W(f>lIVl. . . The following properties of kforms are obvious from the corresponding properties for !."2 k ( V): (WI + w. · . and c = ry( et . If ry is a section of !.B is a vector bundle. . unique orientation [vt . ."2 k (1[. which are just alternating covariant tensor fields of order k. . Vn) With our new algebraic construction at hand. . . and w /\ ry.1 (p» . so ry = c · det for some c E JR. If V is ndimensional and 0 #."2 k (TM) can obviously be made into a Coo vector bundle. . let W E !. .) = w /\ ryl + W /\ ry. " . . we are ready to apply it to vector bundles. !."2k (1[1 (p» !. . ."2 k (�) with is a function with E for each E B. we obtain a new bundle !. .ill E !. vn ) : 6. /w /\ ry = w /\ /ry = /(w /\ ry) w /\ ry = ( . (p) w(p) !."2n ( v) . and W is a kform on N. .DifJemiliaL Forms 5. . are called kforms on M."2 n ( v) . . . " (Unl. A I form is just a covariant vector field. Remember that covariant tensors actually map contravariantly: If /: M + N is Coo. j=1 j=1 Then clearly ry E !. and let n Wi = L CXjiVj = l . COROLLARY. . . Let 207 Then VI . . We can also define WI + w."2 n ( JRn ). then there is a ry ( (u l I ."2 k (�) by replacing each fibre 1[. THEOREM. . then /* w is a kfol'ln on M.talnVl). p ry)(p) w(p) ry(p) . ."2 k +/ (�) by (w /\ = /\ E In particular.en) = w(Vt."2 1 (�).I/Iry /\ w /*(w /\ ry) = /*w /\ /*ry.ant!. then we can define a section W /\ ry of !.) /\ ry = WI /\ I/ + W. . . v ] = /L IL for V such that n if and only if W(VI. . Since !. n. . all forms will be understood to be Coo forms unless the contrary is explicitly stated. . sections of n k ( TM). . > O. we can speak of Coo forms."2 k +1 (1[ 1 (p» .1 A section W of !. . •. /\ ry W /\ (ryl + ry. . If � = J[ : E ?. . .
/\ dX. . f*(dy' /\ .k(p) (i1 < . THEOREM. /\ dy")(p) ( a�' Ip . 7. then q = f(p) f*(gl/Y' /\ . 0 by Theorem 5 . o f» ) dx' dx". a. /\ dyn ) = det . .. . . and (y. d et � A · · · f\. J* a�n IJ n a = dy'(q) /\ . . w = = (il. but we will do one special case here.(p) . . by Problem 41. Now. If U) i s a coordinate system around P E M. 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.1k(p).208 Chapter 7 is left to the reader (Problem 16).x�f» ) dx' /\ . < ik) are a basis for !. . .. . then the dx.. . C(Y. . PROOF.' (p) /\ . 1 M + N is a Coo function between l1manifolds. It suffices to show that f: (x. . dy" ) = (g o f) . U) is a coordinate system.. ..· by dx l for the multiindex I . . . ay a � a(i o f) a 1 . 1=1 q q ) . . (p) are a basis for Mp '. /\ dyn (q) (f* a�' Ip . f*(dy' /\ . f\.. /\ dyn (q) (L a(ix l f) (p) �l 1 . .. .. L. /\ dxiJ. ay i= l  ( a(y. . .. so the dx. If (x. . Thus every kform can be \\'Titten uniquely as W= if we denote dX i l /\ . •:. a�n IJ = dy' (q) /\ . . Wl dxl.. . .. /\ dxn . V) a coordinate system around E N. . ik).
o . then M is orientable (i. .. P E U we have or WU ( VI > . if there is a nowhere zero l1�form w on an nmanifold M. . .DijlerentiaL F orms 209 8. . Moreover. . . . zero section w of I1 n (�) has a special significance: For each p E B. = g . h ( ) [This corollary shows that nforms are the geometric objects corresponding to the "even scalar densities" defined in Problem 41 0. For each ( x. Vn» Now let O if and only if [VI > . .e.. Thus w(p) #. w PROOF. we can choose a cover e of M by a col lection ofcoordinate systems {(x. . then eadl p. ax) Apply the theorem with f = identity map .:. then there is an nform on M which is nowhere o. if VI . and a partition of unity {4>u } subordinate to e. .. . . . . vn] = /Lp. . . W = L 4>uwu. det a/ . then PROOF.] If � = " : E + B is an nplane bundle. . U) choose an nform Wu on U such that f VI > . U) and (y. . the nonzero w(p) E I1 n (. THEOREM. Ue(9 for every Then w is a Coo nform. the bundle TM is orientable). . V) are two coordinate systems on M and g dyl /\ . so that /L = {/Lp} is an orientation of �. . and strict inequality holds f at least one U. In particular. .I (p » determines an orientation /Lp of . U»). or . Vn E Mp . The converse also holds: 9. Vn] = /Lp. COROLLARY. then a nowh".I (p) by Corollary 6. If a Coo manifold M is orientable. . If (x. /\ dxn. By Theorem 213 and 21 5. /\ dyn = h dxl /\ . . It is easy to see that the collection of orientations {/Lp} satisf the "compatability condition" set forth y in Chapter 3. •:.. Vn) :0:: 0. Let /L be an orientation of M. [VI . . . Vn E Mp satisfy (4)u Wu )(P)(VI . ..
WI is a kform. + ( I /wl dw. Conversely. W = L W'l dx l I Notice that the bundle r:2"(TM) is Idimensional. I a= ax· f (x. [Generally. (3) d(dw) = o. so M is orientable.] Just as r:20( V) has been introduced as another name for 1R.) = dWI + dw. f W'l WI. then it certainly has a nowhere 0 section. of course. 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»). We begin by finding some properties of (still defined with respect to this particular coordinate system). The first way is to use a bruteforce computation: comparing the coefficients in the expression f W f . which implies that it is trivial.. 1 0 . 1\ 1\  1\ . w). d' = o. then r:2"(TM) has a nowhere 0 section. This can be proved in several ways. if the bundle r:2"(TM) is trivial. (I) (2) If d(wl + w. I then each dWI is a Iform.. which in a coordinate system U) is given by dw with the The second method is a lot sneakier. We have shown that if M is orientable.) = dWI W. if � is a kplane bundle. j= ax} If W is a kform W = L Wl dx l .210 Chapter 7 It turns out that this definition does not depend on the coordinate system. then r:2 k (n is trivial if and only if � is orientable. then d(wl w. by dw = L dwl dx l I " aW = LL1 l dx· l\ dx I . the differential of w. and we can define a (k + I )form dw. PROPOSITION. Briefly. provided that the base space B is "paracompact" (every open cover has a locallyfinite refinement). df df = " af df = L1 � dxj .
so n ( n a'f d(dw) L L aX"ax· dxP /\ dx· /\ dx l ) . (4) d'f (the old) dJ: Then d' = d on U. and satisfies d on U.) dUg) /\ dx l /\ dxJ g df /\ dx l /\ dxJ + f dg /\ dx l /\ dxJ dWI /\ w. (3) It clearly suffices to consider only kforms of the form w = fdxl . + (I) k w I /\ d'w. •:. (I) d'(wl + W2 ) = d'w l + d'w. (3) d'(d'f) = o. PROPOSITION. =gdxJ. To prove (2) we fIrst note that because o f (I) it suffices to consider only In this sum. Suppose for all k. the terms a=1 /3=1 a=l Q  and cancel in pairs . d' takes kforms on U to (k + I )forms on U.. .Di/ ji<rential F orms 211 PROOF. = fg dx l /\ dxJ and d(wl /\ w..) = d'wl /\ w. Then n af dw ". WI = fdx l w. ax.. Then WI /\ W. +(Ilfdxl /\ dg /\ dxJ dWI /\ W. \!\le next note that these properties characterize I I . (2) d'(wl /\ w.dx· /\ dxl L.. + (llwI /\ dw. == == == == == == (I) is clear. = ..
pick any U with P E U and define PROOF and agreeing with the old d on functions. + (.I /w. For each coordinate system (x. /\ d'(d'xi. d(w. = df /\ dxl + f /\ d'(dxl) dxl = dX il /\ .d'X i. . for all k. . /\ w.) = dw. + dw. Assuming it f k . U) we have a unique du defined. . •:. /\ . . There is a unique operator d from the kforms on M to the (k + I)forms on M. • The third way of proving that the definition of d does not depend on the coordinate gystem is to give an invariant definition. d' = 0.0. and p E M. ' /\ dXik = d 'xi.212 Chapter 7 It is clearly enough to show that d'w = dw when w = PROOF. /\ w. We will uSe induction on k . by (2). /\ d 'Xik = 0. d(wj + w. by (2) by (3) and the inductive hypothesis . satisf ying = 0 . . dw(p) = du(w [ U)(p) ) . ' /\ d'Xik . . ) /\ d'xiz /\ . .I we have or d'(dxl) = d'(d'xi. COROLLARY. /\ . . /\ d'Xik ) d'(d'xi. . where by (4) . f dx1 Now So it suffices to show that d'(dxl) d'(j dxl) = d'j /\ dxl + f /\ d'(dxl) by (4). Given the form w.) = dw. /\ d'Xik ) = 1 2 . /\ . /\ dw.
." .::k+1 (= }:. .. .X". + }:. . + }:.. . X" l::. . . .u ..L(lin+j (Xjl)w(X. Y] .. . . . . . it is easily seen that 2:2 becomes I }:. X. . . + = L(li+'" (x. I f i s a kform o n M. it is actually linear over lhe replaced by then }:. Xk+. .li+j w([x. .. PROOF. becomes I}:." say) .... This (k + I )form agrees with dw as defined previously. . Xk+. by renumbering." ' (X" .. .J)w(X. X" ... + I }:..=1 + L (. k Theorem 4�2 shows that there is a unique covariant tensor field dw satisf y ing (*) . . . . U) it clearly suffices to compute d(1 dx l ). + }:.X" " " X.. Moreove. In fact. f\1oreover. Xk+') to }:.. . .) where over indicates that it is omitted." "...i<j.) . Y.Differential F orms 213 1 3. It is easy to check that dw is alternating. X. .YI · X [X. . . X}.. To compute dw in a coordinate system (x. so that it is a (k + I)form.. . Xj ]. . THEOREM. io<j a brief inspection then shows that }:. . we might as well assume w = I dx' /\ . . is clearly linear Coo Junctions y.. . becomes IX. /\ dx . . ) Xk+1 . . Y] I[X. Y] + XI . . jY] = I[X.Xk+').<ill . then there i s a unique (k + I)form dw on M such that for every set of vector fields we have w XI) . if Xio is and using the formulas [IX. . . .X. The operator which takes over � . X}.
214 Chapter 7 dw. .)(p) w p . Xi dw(X" . . Xk+l . . this happens only if j > k. . This term is what makes the operator dw(X" " " Xk+'). p. . dx . . X" . . . . a� dw = L(I/ ax dx' j>k af = ". . . . For . . •:. 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. . . . . j>k " aj = L1 ax} dx}. . /\ dxk dxj dxk dxk . We do not find )(v" . as for any form. . /\ /\ . and our first use of Theorem 42. Vk+') Vi . But we can hardly hope f anything or does not depend on the values better. Xk+. . we have dw = It is clear from H that dw(a/ax· ' . After all. . One other feature of our definition is common to most invariant definitions of tensorsthe presence of a term involving brackets ofvarious vector fields. j= � /\ /\ /\ . /\ This is our first real example of an invariant definition of an important tensor. iii . . but first find where are vector fields extending and then evaluate this function at By some sort of magic.. in which case so which is just the old definition . Since the a's are increasing. . . ax} dxj dx' L. . . /\ /\ . directly. dw(p .. . .a/ax·k+ ' ) = 0 unless some (0') . k ). Q'k +d is a permutation of ( I . although of except at it does depend on the values of at points other than this must enter into our f ormula somehow. . Xi p.
on M is integrable if and only if .lt(q)dxP (q) restricted to !. the same is true for q sufficiently close t o p . This enables us to state a second version of Theorem 65 (The Frobenius Inte grability Theorem) in terms of differential forms. .Di/Jemlliai F orms 215 linear over the Coo functions.w([X.k independent Iforms Wk+ 1.) is an ideal in the ring \1(M)].). . Theorem 13 gives the following formula. J(!. • . . By continuity. there are Coo functions I such that t k dx"(q) = "L. . P= l /\ . . A distribution !. .q a = k + I . W(XI . Locally.) [thus. Y]) I = E ry /\ E E n P E . .). There f ore.Xl belong to !. but it disappears in computations in a coordinate system. J(!. Then d(J(!. U) so that . Y) X(w(Y» . .n. SECOND VERSION). /\ in . .. . In fact. is a kdimensional distribution on M.Y(w(X» .q.) is generated by . . PROPOSITION (THE FROBENlUS INTEGRABILITY THEOREM. . . and that W J(!. dxl (p) dxk (p) is nonzero on !..» c J(!.X ) It is clear that WI + w. dx k (q) are linearly independent !. J(!.) if Wl . . for ali i. P=l We can therefore let k w" = dx" ."L. the ideal J(!. 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. wn . which by Corol lary 4 implies that dx1 (q).It dxP. In the particular case where is a Iform. . If !. .W' J(!.) if w J(!. 1 4 .p. Define the ring \1 (M) to be the direct sum of the rings of Iforms on M.) w w whenever XI. around any point M we can choose a coordinate system (x.
j :S k and a > k. But each if and only if each belongs to I'>. . Xj]) = W"([Xi..wn must be linear combinations of them. while each if and only if E J(I'>. Locally we can choose Iforms w I . . io ::. . . Let Xl .e. wi (Xj) = oj. we have following.. . . . wn generate J (1'>. . the forms dXk+1. Xj] dw"(Xi . P>k Once we have introduced a coordinate system (x. Xn be the vector fields with . (i.Xj. Xj.Xj) = Xi (w"(Xj» .Xj) = 0 dw" Notice that since the w i /\w j (i < j) span fl2 (Mq) for each q. wn which span Mq for each q wk+ 1. U) such that the slices l l {q U : Xk+ (q) = J + . ..Xj(w"(Xi» . .. dxn are a basis for J(I'>. so we can write the condition d(J(I'>. So if and only if o. Xj) = 0 w"([Xi .) as dw" = L e. . • • Then with Xl .). k are distinct. write dw"(Xi . the first two terms on the right vanish. So I'>. so wk + 1 .). .. /\ wp .. Xj]) = 0 [Xi . . . • • k. is integrable). . .w" ([X" Xj]). :+ Now P=l = L OJ /\ wj If a > i<j for certain forms OJ .. VVe therefore h ave the E . . . we can always For I :S i. if I'>. 0 = dw" (Xi.).. k.) = ej�/XiU). j = I. .Xj] = L ctxp i. . . Xn (q) = a n } are integral submanifolds of 1'>. dw"(Xi . Xk span 1'>.216 Chapter 7 such that PROOF. .» c J(I'>.. . ) = L(ej /\wj )(Xi.. and io.. is integrable ifand only ifthere are functions ct k [Xi .
d: f: M + N is Coo and w is a kform on N. /\ dXik' ») /\ j*dXik + since dj* dX ik d d(Xik f) j*(d(g dxi. . . . tracing through some defini tions. .. PROOF. /\ ' . {3 > k) with Although Theorem 13 warms the heart of many an invariant lover. 1 6. For p M. If are linearly independent If orms in a neighborhood of E M. then j*(dw) d(f*w). . Problem 1 8 gives another invariant definition of using induction on the degree of which is much simpler. COROLLARY.. {3 k) with . . . w' = L ft dgP • P p > 1 5 . /\ dXik' ) /\ f*dxik f*(dg /\ dxi. p dw· = L e.. PROPOSITION. For k = 0 we have. 0 0 =0 = = = = . then there are Iforms (ex.I. /\ dXik' ») /\ f*dxik by the inductive hyposthesis f*(dg /\ dxi. /\ . 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. dw. the cases k > I will hardly ever be used (a very significant exception occurs in the last chapter of Volume V). . /\ . /\ dXik' /\ dXik ) j*(dw) : = ) = = /\ We will use induction on k. of course. let (x. /\ . •. /\ dXik' j*dXik) d(J*(g dxi. /\ . ..wn e. we have d(f*w) d((f*gdxi. We can assume w g dxiJ /\ . /\w P if and only if there are functions f . f* g is to be interpreted as g o f). /\ dxik . . If E = = = = j*(dg)(X) = dg(f* X) [f* X](g) = X(g 0 f) d(g o f)(X) (and.orms DifJerential F 217 wk+1 .g P (ex. . . Assuming the formula for k . U) be a coordinate system around f(p).
A form is called closed if = 0 and exact if = for some form YJ. = 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 �: . = .) Since = 0. i. ay C The necessary condition. every exact form is dosed. w dry. dw 0. is aA + aB + ac = 0. = Wj .e. If is a 2form on � 3 . then the condition dw = 0. (The terminology "exact" is classical diff erential f orms used to be called simply "differentials". The relation = 0 is just an elegant way of stating that mixed partial derivatives are equal. however. ax ay az if and only if = then = =    . by the criterion of the previous chapter... a differential was then called "exact" ifit actually was the differential of something.=1 n One property of qualifies. In other words. af ax.ap = . w W L w. as a basic theorem of differential geometry. For 2forms the situation is more complicated.e. dxi .218 Chapter 7 Now we know from Theorem 61 that these conditions are also sufficient. i.�� = B aQ . aWi aWj axj axi is necessary for solving w = dj. There is another set of terminology for stating the same thing. which will be discussed in the next chapter. The term "dosed" is based on an analogy with chains. a.
whose derivative at (a. however. Recall that if L c IR ' is [0. and e. The reasons f restricting ourselves to local results are now or somewhat different. Like our results about solutions to differential equations.b) r+ {L {I" : r > fY41Y g(x.DijJerential Forms 219 I n general. e) : IR ' is the inverse of the map Coo. 00) x {O}. with aj ay ag = ax We know how to find a function a on all of IR' with w = da. namely a(x. in fact. L (a. b) has determinant equal to a #.o. then it is exact. = e in the region A. then Jxo defined in Chapter 2. It turns out that these necessary conditions are also sufficient: if w 1S dosed. Then e. yo ) dt + On the other hand. agree on their common . Consequently de and de. Consider the case ora closed I form w on �2: w= j dx + g dy.y) = r j(t. + O} x (0. . By deleting a different ray L. we are dealing with a rather strange collection o f partial differential equations (carefully selected so that we can get integrability conditions).a sin b). we can define a different function e. . 21f) (a cos b. t) dl. the situation is very different if w is defined only on IR' . is (I".{O}. this result is true only locally. = e+21f in the region A .
) Clearly w is also not exact in any small region containing O.O) = Po H: x for p E M.I) = lp . which implies that afjax = aelax and afjay = ao/ay and hence J = e+ constant on lR' L. A computation (Problem 20) shows that w x = x'yy' dx + x' + y' dy. w de only on lR ' L. �" is smoothly contractible to E �n .L d(de. but not exact. that determines whether or not a closed form is necessarily exact. 1] + M x such that For example. so that together they define a If orm w on lR' . (It is still exact i n a neighborhood of any point of lR' .{O}. This example shows that it is the shape of the region. 1] + by l'vlore generally. A manifold M is callcd (smoothly) contractible to a point Po E M if there is a Coo function H: M [0. S o w i s closed. but this is an abuse of notation.L. which is impossible. 0 [0.) = 0 on lR' .{O}. I) = p H(p.e) = 0 on lR' . . The Iform w is usually denoted by de. clearly imply that this i s so] . if w J. since lR. rather than its size. then = =d + so d(f .220 Chapter 7 domain. w is not dJ for any C' function J : lR'{O} Indeed.L. In fact. we can define lRn lRn H(p. + dJ = de on lR ' . Nevertheless.L . dw = 0 [the two relations d(de) = 0 on lR' . U C lR " i s contractible t o Po E U if U h a s t h e property that H(p.
I] with dw = 0. If we think of [0.po) E U for ° :s I :s starshaped with respect to po). but we will not be in a position to prove this until the next chapter. and pay hardly any attention at all to H. then every closed form on M is exact. this result and our investigation of the form de prove the intuitively obvious fact that JR' .{o}. at time 1 this is just the identity map. then . I] (for any manifold M).) The trick in proving our result is to analyze M x [0.{o} is 1I0t contractible to a point. I] by We claim that if w is a form on M x [0."tial F orms P E U implies po + I (p . then for each time I we have a map p r+ H (p. the same result holds for JR" . For I E [0. (By the way. 221 1 (such a region U is called Of course. I] as representing time. : M + M x [0..Dijje. I] we define i. We will show that if M is smoothly contractible to a point. I) of M into itself. and at time ° it is the constant map. many other regions are also contractible to a point.
Now for W = 2::7= 1 Wi dx i + f dl we have � � af dw [terms not involving dl] . the projection IT on the second coordinate). axl dx'. Consider first a I form on M x [0..a).(p. i==] i=] Wi t . We will begin by working in a coordi nate system on M x [0.. .Ll w. = r 10 axi n n n L Wi (p. ... I]. I] (Xl 0 1fM . for Conve n n i.l)dl. There is an obvious function I on M x [0.* (L Wi dXi + fdl ) = L Wi ( · .xn O J[M .. It is easy to check (or should be) that xi 0 ITM by xi .Wi (p. l)dxi .a)dxi. aWti dx'· dl + L. then W I]. while J[M is the projection on M. (namely. and if U) is a coordinate system on M.O) = 1r0 ' aaw · (p. .222 Chapter 7 we will see later (and you may try to convince yourself right now) that the theorem follows trivially from this. I].I)dl ) dxi . . I) dl. I) .l)dl f ' af (p. t) is a coordinate system on U x [0. dl. (x. i=1 a i=l So dw 0 implies that f+ =  where /\ =' /\ Consequently.a) denotes the function p wi (p.L. . = i= x i=] i lfwe define g : M + IR by g(p) = l f(p. . We will denote nience..O)dxi = L] (In0 1 aafi (P. = so (I) Wi (p.
I]) (p. = j In general. the function j. it is easy to see (Problem 22) that we can write w dt.I)form with the (dt /\ ry) . we write the I form w on M x [0. ) = = w(v] ) w(v. for a kform w. Applying this to the decomposition (*).. t)dt. I] as w] + w. EEl kerHM •.. . there is then a unique j with w. of two subspaces.). uniquely as where w] (v]. and hence also. I] M w If a vector space V is a direct sum V = V] E !. then any where w] (v] + v. and ry is a (k . Notice that for the tangent space of M x [0. (V ] + v.11 ( V) can be written EEl V.) w.t) = kerH. . I] we have (M x [0. [0.Diff erential Forms then (2) 223 ag t aj aXi (p) = 10 axi (p. g Equations (I) and (2) show that Now although we seem to be using a coordinate system. Vk) = w ° ifsome V = w] + i E ker HM. . . is really independent of the coordinate system.
(2) w = / dl /\ (r�i.io*w that holds even when dw #.)dl. i. The operator I is clearly linear. so we just have to consider two cases. . = = and + = . /\ dieik f diel . 1 7. Now I(dw)(p) 1(. . O)]dx l (p) i.n . .f(p. Define a (k . . . For any kform w on M [0. Then af dw at dl /\ dx l . (i{' / ) Clearly I(dw) d(Iw) 0 •:.a=] axa dl /\ d'X· /\ diel ) (P) � a! L = = = = = it is easy to see that = = =  + = = = d(Iw) d ([ /(p. . I) dl dx· /\ d . * w io*w d(Iw) if dw 0. analogous property.l)(i.. . THEOREM. I] we have i. j. (I) w f d'Xi. I ) dl) dxl (p) [f(p.i. /\ d.. t ). ) PROOF. I) .I)form = = x = = + = . Since Iw is already invariantly defined.224 Chapter 7 Iw on M as follows: Iw(p)(v" .*w . Actually.�ik' f dl /\ d'X l .io*w d(Iw) I(dw). at a!. this proves the result in this case.io*w(p). I(dw)(p) ([ � (P.io*w d(Iw).*v" . . we can just as well work in a coordinate system (X l .0.*w .l) dl) dx l Xl =] o (p. (Consequently. *w io*w 0.*w . Then i.Vk_') = 10' ry(p. . .. Since Iw 0. . We claim that dw ° implies that i. *w(p) . /\ . it is easier to find a f ormula for i.*Vk_. /\ .
We are given H : M x [0. COROLLARY. Since the new form is given by an integral. Soc. Trans.O) = Po Thus H 0 il : M + M H 0 io : M + M for all P E M. PROOF. is the identity is the constant map po. •:.) In the case of a starshaped open subset U of IRn . Suppose we have any operator D from kforms to Iforms. There are classical theorems about vector fields in 1R 3 which can be derived "'om the Poincare Lemma and its converse (Problem 27). we can find (Problem 23) an explicit formula for J(H*w). it has always been something of a mystery to me just why d turns out to be so important. I] + M with H(p.0 = il * (H*w) . such that the following diagram w . 1 25. d(H*w) = H* (dw) = 0. or we can solve the system of partial differeIltial equations w = dry explicitly in terms of integrals. and originally d was introduced in order to obtain a uniform generalization of all these results. Amer.141 . f evelY form w on U. X on Difjermlial F orms. Corollary 18 is called the Poincare Lemma by most geometers.io* (H*w) by the Theorem . 92 ( 1 959). while d' = 0 is called the Poincare Lemma by some (I don't even know whether Poincare had anything to do with it. = d(l(H*w» . If M is smoothly contractible to a point po E M. Even though the Poincare Lemma and its converse fit very nicely into our pattern for basic theorems about differential geometry. then every closed f orm w on M is exact. atwal Operations An answer to this question is provided by a theorem of Palais. I ) = p H(p.DijJerential Forms 225 1 8. Math. So But so w = (H 0 il )*(w) = i l * (H*w) 0 = (H 0 io)*(w) = io* (H*w). where we have an explicit formula for H.
If k = I. then can be a multiple of the identity map. Dl D D DD D lD . kforms on M � kforms on N Iforms on M � If orms on M Palais' theorem says that.226 Chapter 7 commutes for every Coo map I: M + N [it actually suffices to assume that the diagram commutes only for diffeomorphisms 11 . which we discuss in the next chapter. can be a multiple of "integration". then can only be some multiple of d. with few exceptions. If I = k + I. (As a corollary. these excep tional cases are the following. In this case. = o. Roughly. d ' = 0. 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. but nothing else.
V1 . and w /\ ry = ( _ I ) klry /\ w. (Try w. Show that the set of all shuffle permutations is a cross section of S'. . k} and {k + I.J w).w . 4.. .J w) = . . Let Alt be Alt without the factor 1 / k!. that 7\ is nol associative. (b) Show that if V1. . ry 2.l lw1 /\ (v .) = (v . E 11 1 ( V) and e E 11' ( V) . Vk) = ap • (V1 . Vk +/) = aEK L sgn a . E 11 k (V) and w.J W)(V1. Va(k+/) . . . .) (c) A permutation a E Sk +1 is called a s"u permutalion if a ( l ) < a (2) < . ..) . Vk_1). then if j = i• .J w. . (b) Show from this definition that w /\ ry is alternating. . · . V . . . . k +/} invariant.. . + ( . . E 111 (V) we have (Use (b) and linearity of everything.J (W1 /\ w. (c) Show that for W1 V . . and define W 7\ ry = Alt(w ® ry) . . < .1 ) = w (v. " . . . . .flle a rk ) and a rk + I ) < a rk + 2) < " . we define the contraction v . Vk . </>n.. (Proving associativity is quite messy. This definition may be used even in a field of finite characteristic.. .. j #. . w ® ry(va(l) . Let S' C Sk+1 be the subgroup of all a which leave both sets { I . Show that if we define then 3.. .J ( v . Vk). ..). . ... (a) Show that This is sometime also called the inner product and the notation ivw is also used...) (a) Show that for any cross section K we have w /\ ry(v]. . . .. Vn is a basis of V with dual basis </>1 . A cross seclion of S' is a subset K c Sk+1 containing exactly one element from each left coset of S' .Difforenlial F orms 227 PROBLEMS 1. ..J ( w . For v E V and w E 11k (V). . . · . (V1 . ..J w E I1k..1 (V) by (v . < a rk + /) . ..J W 1 ) /\ w. Show a • p ..any i.
1 1 (V) .1' (V) is decomposable.1 k (V).J w. changes the sign of the right side).1 l (V) is decomposable. .J w)(p) = X(p). then every W E !.1k (V*).4 = YJ W2 YJ. Vk !) .. + n + E ::. (e) Prove by induction that /\ is bilinear and that W l /\ W. . (f) Since Vi E V can be regarded as elements of V**.J w(p). . Show that functions fi.228 Chapter 7 + W /\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. are independent..](v" . + (I/Wl /\ (X. Show that if Wl is a kform.. (b) If </>i. is skewsymmetric (it is only necessary to check that interchanging V l and v.) + (</>3 /\ </>4) is not decomposable.) if and only if W2 Wj l/\ for some (d) I f Wi are decomposable.I)form X. = ( _ I )k! w.J w. In this case. we define the annihilator of W to be Ann(w) = {</> E V* : </> /\ W = OJ. .. /\ W l . (a) Show that dim Ann(w ) ::. then Ann(wl ) Ann(w. i = 1 . (c) If W and are decomposable. . we can consider V /\ . .) = Ann(wl /\ w. Show that with this definition WI /\ W. which is unique up to a multiplicative constant. /\ Vk E !. . then X .. n 6. I : M IR form a coordinate system in a n neighborhood of p E M ifand only if dfi /\ . (a) If dim V 3. (f) If X is a vector field on M and W a kform on M we define a (k .o. .J w by (X. n n. #. then W = (</>1 /\ </>.) = (X. .1 k ( V) is called decomposable if W = </>1 /\ ..). An element W !. . /\ </>k for some </>i E V* = !. Ann(wl ) + Ann(w.. 5. k.) = {OJ if and only if Wl /\ w.J Wl) /\ w. Reformulate parts (a)(d) in terms of this /\ product. (e) If V has dimension then any W E !. For any W E !.J(Wl /\ w. (b) Every subspace of V* is Ann(w) for some decomposable w. . /\ dl (p) #. Vk !) = [(Vl . n . and that equality holds if and only if W is decomposable. .(v" . /\ l + Wl /\ w. we define + + . then Ann(w l ) C Alln (w.o. .). . Hin!: Look at W /\ w .J Wl ) /\ w. 7. Vk !) + (.I/[Wl /\ (Vl .)](v" . .
. where w ' does not involve 1/11 or 1/12. K . .. . j= ! Wt /\ . . i<j choose </>l involving Vth Vt3. and that the (r + I)f wedge product is O. Show that there is a basis </>" • • • .1 2 (V). (c) If w = L i<j a ijVti /\ Vtj . /\ W is nonzero and decompos able. /\ Wp = L B H VH H Wp+l /\ . . . : al•hP ' . show that B H = det ( al. .Diff erential Forms 229 8. . . .k] (a) If v" . Thus r is welldetermined.itp ) . old it is called the rank of w. (b) Show that the rfold wedge product w /\ . Vtn so that W = </>l /\ </>2 + w '.p. C K = det ( ap+ l 'k' : au. ap. . and let q = n . .h] ap. Let A = (aij) be an 11 X il matrix. . show that the rank of w is the rank of the ma trix ( aij). " /\ Wn = L: C K VK . let Vt. Let I :s p :s n be fixed. . . ' < kq.. Vn is a basis of V and show that n Wi = L QjiVj. < hp and K = k l < ' . </>n of V* such that Hint: If W = L aijVti /\ Vtj . Vtn and </>2 involving Vt2. (a) Let W E !. Vn is a basis for V and Wj = 'LJ=l (ljiVj. . If 10. 9. . . For H = hl < .. . . .
. (Cartan's Lemma) Let <PI . H' is the sign of the permutation {o eH. ) h l . the bundle whose fibre at Since we can regard as an element of ( Mp )** . /\ Vn " K #. . (c) Prove "Laplace's expansion" det A 1/1" 1 1 . .. · /\ .j . we can consider sections of bundles constructed from p is 11k ([:n:.. . . . ./\ . In addition to forms. . we can consider 11 k (C). . eH. Show that VH /\ Vx = .H' VI /\ . kl . 1/Ik E y satisf . ..I (p)]*).H (arranged in increasing order). . . For example. . where eH. . . . . . j= 1 k where bundle. H E V* V* be independent and suppose that aji = aij ' Then =:: 1/Ii L Qjiq. hp .230 Chapter 7 (b) Let H' = { I . L. if � = :n: : E + B is a vector 12. a a (a) Show that if then .w B H C H' .H' K = H'. . .' ax l axil TM using 11 and other operations. (/>k . " } . . kq = " . . . h2. • ". (I . ' . aIlY section of I1 n (T*M) can be written locally as h .
W ( V) denotes all functions ry : the form V* x . . .] · ··ill . . (e) Suppose r:2n .) (c) If :Tif ] ( V) is defined similarly. I times k limes m • • • given by (</>1> . except that r:2m(v) is replaced by r:2m( v*). . . Let :Tik[n . Show that sections o f :Tik [n. . ] .. . (b) Let :Tik [m] ( V) denote the vector space of all multilinear functions show that sections of :Tit. We can identify :Tik ( V) with A w an integer Since r:2m(v) '" Am(v*) '" [Am(v)]*. • . � x � + r:2m (v) . . x V* + r:2n ( V) n </>1 /\ /\ </>n .i'1 similarly. we can identify :Tik[m] (V) with :Tif ) ( V) with m I k Q9 V* ® Q9 V ® Am(v) I k Q9 V* ® Q9 V ® Am( V*) . k times I times .] (TM) correspond to (even) relative tensors of type (1) and weight . /\ v*nJ .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. corresponds to the map Show that sections of :Tik [nl (TM) correspond to (even) relative tensors of type (1) and weight I . (Notice that if VI > . with components e. . Interpret the relative tensor with com 'v' times V x . then elements of r:2n(v) can be represented by real numbers [times the element V*I /\ . Vn is a basis for V. .w] (V) b e defined like :Tik [n] . except that r:2n(v) i s replaced by r:2n .. (d) Show that the covariant relative tensor o f type e ) and weight 1 defined i n Problem 410. x V + IR which are of for some W E r:2 n ( v) .I .w\ TM) correspond t o (even) relative tensors of type (J) and weight w.W ( V) .. </>n ) f+ ponents e'I . these results can all be restated. W (f) For those who know about tensor products V ® W and exterior algebras k ( V). Similarly for :Tit.
.1 ) for all A and all invertible B . and A is a tensor of type G).I + C2 An. Chapter 7 and weight w and k 7it .t + . For every n XI1 matrix A = (aij) we then have a number P(aij). a!k ()Jl ··iN ' lk . . + (_ I)nc . I}. then the map A * : !. Show that C = det A .) (b) Conclude that det AB = (det A) (det B). then the function p r+ cdA (p» can be defined as a (2k)fold contraction of � A 0 · · · 0 A 0 8.wl (V) = 0 V* 0 01 V 0 0 w A m ( V*) .· .232 Consider. j]. . . Let P(Xij) be a polynomial in n 2 variables.. if 8 is as defined on page 1 30. Recall that the characteristic polynomial of A : W. more generally. respectively.2 + · . .1 k (V) . . . 14. m Noting that An(V) 0 · · · 0 An(v) is always Idimensional. This problem outlines a proof that any invariant P is a polynomial in the polynomials CI . . V + V is X(A) = det(A! . (a) Show that Ck = trace of A * : !. . . · · + ( . Call P invariant if PtA) = P(BA B . We will .'.I) " det A = An C l An.. (This may be used as a definition of det A.1k (V) + !.A) = An . .'h. If A : �l l l Ck (A) = � " a'1l a'2z . (a) If V has dimension n and A : V + V is a linear transformation. show that .(trace A)A n .1n(v) must be multiplication by some constant c.wl (TM) correspond to (even) relative tensors of type (1) n 7ik [m:wl ( V) = &/ V* 0 &/ V 0 0w A m ( V) 13.  n trix (a! l (with respect to some basis).jk V + V has a ma Thus. k times 15. show that sections of 7ik[n. (c) Let 8/.. Cn defined in Problem 14.1n(v) + !.wl (TM) and 7if . (b) Conclude that cdA B) = cdBA ).· k •l L i] .::// be as defined in Problem 45 (xiii).
Yn) in the variables Yl . \( 375. .1 ) for all real A and real invertible B. . + ( .atyn. . . . . . . (e) The same equation holds f complex A and complex invertible B. . .DijJ rential FOWlS e 233 need the algebraic result that any symmetric polynomial Q( Yb ' ..1 + . . . where ai is the ith elementary symmetric polynomial of Yb . . since the set where D # 0 is Zariskidense. . Since the eigenvalues A t . . (d) Show that P t A ) R(el (A). up to sign. . . an(Yl . (b) PtA) = R(edA). . of the polynomial with roots Y b . .0 ). . Yn) = R (adyl o . Recall that the at can be defined by the equation n i= 1 TI (y . . en(A» for all diagonalizable A . Thus. . Yn . . en(A» whenever D(A) # O.Aj)'. . o . Conclude. . . Yn Then there is a polynomial R such that Q (Yb . An of a matrix A are. by continuity. The polynomial R has real coefficients if P does. . We will first consider matrices A over the complex numbers <C (the coefficients of P may also be complex). (c) The discriminant D ( A ) i s defined a s n ih (Ai .I )"an. . Show that D(A) can be written as a polynomial in the entries of A .) = n' . .) Now suppose that the coefficients of P are real and that PtA) = P(BA B . " Yn) to be PtA) where A is the diagonal matrix Yl ( . . this is "the principal of irrelevance of algebraic inequalities". Yn. the roots of the polynomial X(A). that the equation holds for all matrices A over <C. Yn can be written as a polynomial in at . . by definition. (a) Define Q(y l o . where Ai are the eigenvalues of A. . . (Regard or the equation as polynomial equations in the aij and bij .(A b " " An). . an. . Yn». . . (This last conclu sion follows even if <C is replaced by some other field. . Yn). . they are the coefficients. · · . . .Yt) = yn . it follows that ei(A) = a. compare pg.
. In Problem 514 we defined Lx A for any tensor field A . L. 18. .. • • or Vn be a basis f V. and let WI... /\ j _ (b) Show that is (c) Using 514(e). ik .·..234 1 6. Wk V be given by Wi = L n • • . Xk+l ) L. . 17. Xk+l ) k+l l = " ( .l)i+l w([X. . . .Xk ).. (b) Generalize Theorem 7 and Corollary 8 to kforms. show that X(w(X" ..I ) i+ LXiw(Xl. Wk ) = l=ilL a/ W(Vi" <···<lk where a{ is the determinant of the k x k submatrix of (aij ) obtained by selecting rows . . Xk» = LxW(X1 . . . .... . . . . (a) Let Chapter 7 Vl .. . then s o Lxw.Xi . . E w 11k (V) show that W(WI..» = Lx(w(XI. . . ..XJ. . For E . . . X". Vik )' i].. j=] Q'jiVj . . ..Xi]. . Xk ) k + " (. 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 . . .Xi .. (d) Deduce the following two expressions: dw(X" . . . . . (c) Check directly from (b) that the definition of d does not depend on the coordinate system. (a) Show that i f W i s a kf orm.
.A dx = dg on [0.. and let a' = i*(dO). . .{Oj be the inclusion. .)  21f dx . Un in a neighborhood of any point in IRn with ( ) it is necessary and sufficient that 20. I ] to find A. . 1 ] . . . a'aij a'aik axk ax! . . l. If c : [0. Compute that Hin!: First set up partial differential equations for the functions fik aUk laxj ..J w)(X" .d(X .. Let aij be . Xk+t ) . . I aUi aU + j ai = '2 axj axi l == (LxI w)(X" .J w).A dx = dg for some function g with g (O) = g(I).axj ax! a'a/j a'a!k == axk axi . . . 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.. Xk +t » i=l 235 _ (e) Show that X . k . . Xk+l ) k+l 1 i+1 Xi. S l is c(x) = (cos 21fx. I] with /(0) = /( 1 ). dw(Xt ..axj axi == x dy' + y dx x y' for all i. show that d(Lxw) == Lx(dw). . . l . Hin!: Integrate the equation w . .. i. .erential Forms Diff dw(XI . .) (f) Using (e). ' functions on IRn with aU = aji . == (At most places 0 = arctan y Ix [ + a constant]. _ . show that there is a unique number A such that W .'2 I ) . 19. . Xk+ I ). Xk+l ) (This may be used to give an inductive definition of d. and use Theorem 61. (b) Let i : S l .I ) {Xi(w(Xt . (a) If w is a Iform / dx on [0. . . . . . . . Show that in order for there to be functions U I .J dw = Lxw . sin 21fx).e. == au) laxk "dO" 21. .. . . 1R' .d(XI .
va) = 0 if some Vi V2 W.I Wi . . If W= WI 22..U is connected.and Wt(Vt.) can be written as a sum of forms w. and 0 #. . W. has degree f3 = k .1 (1 I k .I)form such that WI (VI . .. Vk I ) = 0 ifsome Vi V. Show that U is diffeomorphic to IR ' . a= ' l < " '<'k l _ . . t( 1 ) . where has degree 0' and WI EEl 0' E E E /\ E E x = on U. Let U c IRn be an open set starshaped with respect to 0. = I. . and hence smoothly contractible to a point. vp) = 0 if some V. the kth set being a finite union of squares containing the set of points in U whose distance from boundary U is ::. and define H: U [0. VI . /\ d7a /\ . 1] + U by H(p.(VI .ic (w.236 Chapter 7 W 11k (VI V. . . (The converse is proved in Problem 89. 23.Ik /\ . . /\ dXik • Hint: (b) Find a bounded open set U C 1R 3 such that 1R 3 .. . then W can be written uniquely as WI + (b) If dim V. dxi. . = L. . J(H*w) 1 (IX) dl) xi. is a (k .l) Ip.. Vk ) = 0 if some Vi V. . (a) Show that every E /\ W. .U i s connected. .*.(V\. 1/ k . ic). .) Obtain U as an increasing union of sets. V. (a) Let U c IR ' be a bounded open set such that IR ' . but U is not contractible to a point. . show that 24. where W I is a kform and w. . . . . .
L a X l. . on JRn by a n a n aj . .t for all W E JRn. (a. aXl. show that there is a unique vector Vt x . ' x det (b) Show that x . . define a vector field grad j. . x E \1 n . . x ej. X Vn _l E �n V .) Intl"Oducing the formal symbolism i= 1 i=1 Dd . Let ( . . but the "obvious" proof does not work. . (First find all e. .' . grad j = " .) 26. x (c) For JR3 show that () � Vn . Is U homeo morphic to JRn? (It would certainly appear so. .= " L aXl. and express it in terms of the e*i.t (JRn). Vn _l with (Vt E �n. since the length of rays from 0 to the boundary of the set could vary discontinuously. 27.Di erential Forms ff 237 25.l o w) n = x . Let U c JRn be an open set starshaped with respect to O. . using the expansion of a matrL by minors. the gradient of j. ) be the usual inner product on JRn. .= 1 = (a) If Vt. b) L>i bi. (a) If j : JRn + JR.
if div X = 0. Conclude that V/(p) is the direction in which I is changing fastest at p . 2::7=1 a ' a/ax' is a vector field on IR n . .238 Chapter 7 we can write grad I = VI If (grad J)(p) = wp . (d) If X is a vector field on a starshaped open set U C IR n and curl X = 0. Similarly. w ) . for n = 3. then X = curl Y for some vector field Y on U. dx + a2 dy + a3 dz dy /\ dz + a2 dz /\ dx + a3 dx /\ dy. we can write div X = (V.) We also define. we define the divergence (b) If X of X as (Symbolically. then X = grad I for some fU l Iction I : U > IR. 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. X). where Du/(p) denotes the directional derivative i n t h e direction v a t p (or simply vp (J). curl grad I = 0 div curl X = o. show that Du/(p) = ( v. if we regard vp E IR np).
We can choose points C(�I) on each piece by choosing points �I [IihIll For each partition P and each such choice � = (� . if IIPII = m!'x( ti .\C. . then the equation lim S(P.�) = [ Jdx +gdy lc E IIPII� O 239 .c' (l1')] + g(C(�I))[C2{11) .C'{lI_ ' ) and vertical projection c2 (11) .c2{1i_ l l. C(tl) e(�I) :_ C2(t1) .' ) to C(li). as the maximum of Ii . 1] JR2 is a curve and w = J dx + g dy is a I form on JR2 (where J.lid. i. the piece going fi'om C(li. with e(1) that > > n T which first arose fi'om very physical considerations.C2(t1_' ) e(O) c(I[_. consider the sum " S(P. . then the limit is denoted by l Jdx+gdy. then we can divide the curve c into pieces. �II) .) . (This is a complicated limit. < In = 1 of [0. When the differences Ii . To be precise. and and. for example.Ii_I are small. �n ). that is.c2{11_l ll i=1 If these sums approach a limit as the "mesh" I PI I of P approaches 0. y. . 1]. g: JR2 JR.e' (til) horizontal projection C'{lI) .h x he basic concept of this chapter generalizes line and surface integrals. �) = L J(C(�I» [C'(li) . If we choose a partition 0 = 10 < .Ii_ I approadlCS 0. denote the coordinate functions on JR2). Suppose. each such piece is approximately a straight segment.CHAPTER 8 INTEGRATION c: [0.
Yo). (In classical terminology. so . it has a natural physical interpretation. = yo+� In this case. the differential would be described as the work done by the force field on an "in fInitely small" displacement with components ely.C2(ti_d S(P.Iid· i=1 . while C2(li) . consider the special case where S(P.) IR2. C(I) (I.240 means: for all E we have > Chapter 8 0. a a J.C l (tid = Ii . = 0. If we consider a "force field" on described by the vector field � P. the integral is the "sum" of these infinitely small displacements.�) IiI Ii J g J dx + g dy dx. there is a Ii > 0 such that for all partitions P with II P II < Ii.) Before worrying about how to compute this limit. CI (li ) . for all choices for The limit which we have just defined is called a "line integral". the limit js the natural definition of the work done in the general case.IiI .+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.�) = LJ(�i. YO)(li .
e.li' ). we have.c' (Ii' ) = (b .a) l ' J(xb + ( 1 x)a.Ii] C' (li) . In general.a)(ti . c' C y x . of are denoted simply by and The components [i.=1 it looks like they should approach.yo)dx lb J(x. denotes l' [J(C(I» C" (I) + g(C(I»)c" (I)] dl. for any curve c.I)a. namely Physicists' notation (or abuse thereof) makes it easy to remember this result.lil l· i= = b These sums approach A somewhat messy argument (problem I) shows that these sums approach what (b .C' (li' ) = C" (Cii)(li liIl Cii E [Ii. L1 J(�ib + ( 1 . C' (li) . by the mean value theorem. So S(P.li1l f3i E [Ii"" Ii].yo)dx. YO)dX. so S(P.. On the other hand.yo).�) = L { J(C(�i» C" (Cii) + g(C(�i » C" (f3i)} (Ii liI).�i)a.�) = (b . Yo)(li .integration These sums approach 241 lJdx + gdy = l ' J(X.a) . if Yo C(I) = (Ib + (1 . then a c' (Ii) .C' (ti1l = C" (f3i)(li . x c'. .
dY d1 J " In preference to this physical interpretation of "line integrals". RecaIJ that denotes the = de/dl(�i) e(ti ) tangent vector of c at lime de dl (�i) �i . (I) ] . . Ii). Then the sums = L [j(e(�i» e ll(�i) + g(e(�i»)c21(�i)] ' (Ii . e goes with constant velocity on each (IiI .li1l [=1 clearly also approach Consider the special case where 1 1 [j(e(l) e ll(l) + g(e(I» e21(1)] d/ . The above in tegral is then written 1 / dx +gdy [o I [j(x'Y) dx +g(x'Y) di] d/ .242 0 y y o Cha pter 8 y =y x e and denotes e. this is indicated c1assicaIJy by saying "let x = X(I). we Can in troduce a more geometrical interpretation.
can be used as a w(C(I» w: = limit of the sums (*) C crt).Ii) length of the segment from C(li 1l t o C(li) Ii {I I so [length of � (�i)] . � (�i)] .. (Ii . (Ii . (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".integration 243 In this case. C is .li1l dl c. Before pushing this geometrical interpretation too far.IiI) = length of segment from C(li1l to C(li). It is true that for a given oneone curve c we can produce a form w which works for c. . �i E (li' . Ii). whereby the curve is divided into infinitely small parts. this is the modern counterpart of y the classical conception. t' [ i= length of dx. and then extend W to IR'. the infinitely small piece at with components having length 1W c. dc/dl C C. The natural way to specify a continuously changing length along is to specif a length on its tangent vectors. we choose W(C(I» E !1 ' (IR'c (I» so that W(c(t)) (� ) = 1 . /(C(I)) dx + g(c(I » dy. kernel w (c(t» But if W arbitrarily). we should note that there is no Iform w on IR' such that 1 w = length of C for all curves c. dy. then dc dt (�i) = the constant speed on (Ii). for a general definition of the length of The line integral C. If we choose any length of  is the length of and the limit of such sums.
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. J] 7 ffi. used to define this generalized "length".�) for c is dearly equal to a sum S ( P'.244 Chapter 8 not oneone this may be impossible.2.2 wc op w ("the integral of w over (' is independent of the parameterization"). [0. nor is it deal. so it is clear from our first definition thal for a curve we have [0. I]. in the situation shown belov. if we are interested in obtaining the ordinary length of a curve. I] [0. c: [0. l l c c: [0. because it is ob vious with our original definition and merely true for our new definition. ] [0. This is no longer 50 dear . make sense even if is a curve in a manifold M (where there is no notion of "length").� there is no element of !1 ' (IR 2c(d) which has the value on all three vectors. I] c. Every sum S(P.2 0. 1] 7 [ [/(c(l» C" (I) g(c(t))c2l(t)] cit.p = ker w(p) form a I dimensional distribution on IR . + . and w is a I form on M. for example. For the present.even for a curve but in this case we can proceed right to the integral these sums approach. 1] 7 ffi. In general. the subspaces /'. 1 2 c One property of line integrals should be mentioned now. given any w on which is everywhere nonzero. will have "length" Later we will see a way of circumventing this difficulty. ffi. �') for c o p. so we can define Ie W as the limit of these sums. and conversely. we note that the sums t*j. any curve contained in an integral submanifold of /'. I] [0. namely c: [0."hen we consider the sums (*) f a curve or M.
For a pair of partitions and 10 < . 1]2 JRn .. let � = e* (�) a ae ay = e* ( ay ) . < In of [0. which will eventually become our formal choice..2. I] > JR we have 0 0 0 > > So s'" �ffQS . .integration 245 The result then follows from a calculation: the substitution + For a curve in ffi. We are being a bit sloppy about all this because we are about to introduce yet a third definition. < Notice that if ffi. so that formally we just integrate e*(/ dx + g dy). and take the integral of h on [0. If x and y m·e the coordinate functions on JR 2. we can introduce a coordinate system for our calculations if lies in one coordinate system. and a l form there is a similar calcula 2:::7=1 tion. 1].n . 1]. 1 w = [ [/(e(I» e"(I) + g(e(l))e2l(1)] dl. 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. e I is the standard coordinate system on JR. or break up into several pieces otherwise. 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. 0 0 0 0 0 e: [0. if we choose < . Everything we have said for curves e: [0. . 1] JRn could be generalized to functions e: [0. for a general manifold M. to be precise. I]) w= Wi dxi . we write c*(J dx + g dy) = h cll (in the unique possible way).
c [0. . . and c is a singular k cube in M. Sj1 [ljI .Sj_JJ(lj . IRo ck c(O) E IR Jk : [0. The limit of sums of these terms can be thought of as a "generalized area" of c. dXk [O. . For k 0.lj_l ) is a "generalized area" of the parallelogram spanned by ac ac a. Si_1 c: [0. The inclusion map of It in will be denoted by It > it is called the standard kcube. we now proceed with the formal definitions. [O. llk in c1assical notation.Xk ) dX I . we have a special definition: a Oform is a function !.. lj1 and w is a 2form on IRn. . . w X l . are the coordinate functions. f W lO be O. ° E IR. Xk w w = ! dx l /\ . ( = ) = . IRk. and . To make a long slOry short. mImIC as far [O. We will Jet I Jo = = so that a singular Ocube is determined by the one point M. [0..246 X �jj E [Sj_l . . /\ dxk We define = J ! (X I .f ! notatIOn attempts to w�ic? modern . ay(�ij ). . we define 1 w = f c'w.(�fj ). If is a kform on It. l lk [ l lk as logic permits If w is a k form on M. A Coo function It > M is called a singular kcube in M (the word "singular" indicates that is not necessarily oneone). . * (�ij» ) (Sf . and for a singular Ocube c we define 1 ! !(c(0)). . .. then Cha pter 8 [0. then can be written uniquely as Sj w(C(�ij)) (�: (�jj). l lk where the right hand side has just been defined.
I}" 0 det e([O. Then e PROOF. w w = J dxl /\ • • . •:.11" = 0 e) (det [0. It [0.1]" p: [0.1)II) PROOF. Jc leo p p' 2: 0. COROLLARY. by the Proposition. l w = f f. = 1 w f e*(w) f (J e') dxl f (J e)1 e'l dxl f J [0. /\ dx" by Theorem 7 7 dx" by assumption > 2. . It [ w = [ w.11" = [O. • . 1]n e' 0 [0. We have L w = f (e plow = f p*(e*w) p = f e*(w) 0 � I� � It [0. . since p is onto . /\ /\ . By definition. Then e: [0.illlegmiion 247 > JRn be a oneone singular ncube with PROPOSITION. C{[O. II") by the change of variable formula . 1]". /\ . /\ dx". Let det 2: on Let be the nform 1. Let be oneone onto with det let be a singular k cube in M and let w be a kform on M.
2<3 + 6c4. e. The corollary thus shows > I: c [O. It M is called a reparameterization of c if p: [0. 1\1orcovcr. which is called the boundary of c. I] M then l' I(c(t)) dt is generally i' l' I(c(p(t»)) cll.e. j = Li ajC. I t is a Coo oneone onto map with det p' i' 0 everywhere (so that p' is also COO). so that the change of variable formula will pop up). in accordance with Theorem 77. 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 .. 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. We add kchains.I )chain a c. will also be denoted simply by Ct . e. > f l o c. an ori entation reversing reparameterization clearly changes the sign of the integral. provided it is orientation preserving. and which is supposed to be the sum of the various singular . functions On a manifold cannot be integrated (we can integrate a function on the manifold )Rk only because it gives us a form dx ' /\ . A kehain is simply a formal (finite) sum of singular k cubes multiplied by integers.. <. Our definition of the integral of a k form w aver a singular k cube < can immediately be generalized. if From a formal point of view. It [0. . . diff rential forms are the things we integrate be e cause they transform correctly (i. I I . w(' define the integral of w over a kchain way: c J2::: uiCf [ W = I>lCj W. + 3<4 ) + (2)« . it is called orientation preserving or orientation reversing depending on whether det p' 0 or det p' < 0 everywhere. c: [0. purely formally. + <3 + C2) = 2C2 . and multiply them by integers. /\ dx k ). =I 2(c.' lk For example. The k chain Ie. .g.g.248 > Chapter 8 > independence of parameterization.
:. /11 . (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 . . then 1i). 1..xi . . .l ). I )face of as follows: If X E [0. P I I 12. .O) (x) = I" (x l . . . . . . .O) In) I]n.xn.. . . .l) = (x l . . it for example. . i .0) . xi. . . . .xil . . will not be is convenient to modif this idea.I .Integration 249 (k .l) /11 . . . X Ii). +1 +1 with the indicated coefficients. . Xi . In practice.xn . .x i .O. o f the four singular I cubes shown o n the right. O.1) . The boundary of y the sum of the four singular I cubes indicated below on the left. ).O)face and (i. (x) = I" (xl .:. but tbe sum.I ._n .I ). .xnI ) 1. .I . I) = (X l . .x i . . xiI . 1 . l .I)cubes around the boundary of each singular k cube in c. . . .I)cubes and (the (i.
1 .1) R .). 1]2 + C(2 .) ' C(I . and for a Ochain Li aiCi we define c: [0.O) / C(I) C 0 Finally.O) We also have. 1]0 Notice that for a I cube c : [0. i=1 a=O. 0). y we define a c to be the number 1 E JR. C(1.1) a(aC) = (R .250 Chapter 8 The (i.l These definitions all make sense only for n ::: I .. For the case of a Ocube > M.. C(i.Q) . the boundary of all nchain Li aici is defined by aC = L L ( _ I )i+.(S PH (Q = 0.. > M.O) s  .P) p C(2." )face of a singular ncube Now we define C(O) � C is defined by C(i.I) C(1.S) .) = C (Iv. for a singular 2cube c: [0. which we will usually simply identif with the point P = c(O).(R .1 = O. I] > M we have so a(aC) = ac = C(2. C(l.
. . . Pl)(" . and a(ac) = O..xj. PROOF Let i :0 j :0 n . . If c is any nchain in M.: . . xi.P) and ( CU+I . xn2) j I i = In (x l . . this is the case if c = c) e2 . Xn2).L 1=1 j=1 (t a=O. .) u . Notice that for some nchains c we have not only a(ac) = 0. •:. . .1 . where CT and C2 are two I cubes  . .) for i :0 j ::S n . . . Thus ( /. . Now i I a(ac) = a = L a=O.P) (XI.)) L . It follows easily for any j I singular ncube c that ( c( . . a2 = o. . ( C(i. we have.1 .l (I)'+·+j+p(C(" . . 1]"2 . . .xj..Pp(�:. Similarly. X i .=1  In this sum. .) occur with opposite signs.Pl = ( /'( + . . . xj .) = J. . we have: 3. ..) u. Therefore ' all terms cancel in pairs. . xiI . xiI .1 nLI P=O. .. . then a(ac) = o.) (X I . and consider ( /D. C{ . P) ( . xj . .1 .( P) (x » / = J(�...) U. Briefly. xn2) = In (x l . . /3.I .. Since the theorem is true for singular ncubes. .1 (I )'+. a . For example. . . . Pl)(" . PROPOSITION.P) (x ) = J'J . Xj . In general. xn2)..) U. a good exercise because this involves figuring out just what the boundary of a 3cube looks like.P) = (CU + .I. it is clearly also true for singular nchains. xj.) ( /.) U.•) U . fj. . .P) · L . ( /U+I.P) ' For x E [0.\ (x » = J{ + . x . .) for i :0 j :0 n I. " . .. C(" .integration 251 From a picture it can be checked that this also happens for a singular 3cube.I . xi. from the definition ( /(�.j+I . . p. a.P) (i. but even ac = O .
any k chain c is called closed if ac = O. when c is a "closed" curve. but is simply called "a boundary"). In general. There is also a I chain c which is closed but not a boundary. then ac = 0 precisely when c(O) = c(I). we have seen that on IR ' . i. (0) = Chapter 8 C2 (0) and CI (1) = c2 (1). Recall that a differential form w with dw = 0 is also called "closed". namely. For example. reciprocally.e. The connection between forms and chains goes much deeper than that.. this terminology has been purposely chosen to parallel the terminology for chains (on the other hand. by the classical term of "exact".{OJ there is a I form "de" which is closed but not exact. expressed by the relations d' = 0 and a ' = O. This parallel terminology \'\'as not chosen merely because of the formal similarities between d and a.252 with c. If c is just a singular I cube itself. a closed curve encircling . a chain of the form ac is not desCllbed.
xk ) dxl . .x. dx·k � . .l)k f l(xl . We now compute. . l .I )form on M and c is a k chain in M. • • . the simplest proof uses the theorem which establishes the connection between forms.I )form on IR k and c = J k In this case.x·1 . Therefore = ( . chains. dx. . . /\ = jof dxi /\ . . . . . .{OJ. I . First. 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. THEOREM (STOKES' THEOREM). dxk /(.)* U dx l j. d. . . . . • . then Jc [ dw = [ w. /\ dxk ) if j i' i = [0.k ) d. w is a sum of (k . Although it is intuitively clear that (" is not the boundary of a 2chain in IR 2 . . a. . x. . If w is a (k . .integra/lOll 253 the point ° once. . 1 . . .1 ) 1+ 1 [O.k if j i. . a little notation translation shows that J[o. /\ .1/ f [O. . 1)" I(x. 4. .I)k + (.l)k1 [ J(k . .x·1 . and a.k ) dx. O. . . . .
. . . Xk ) . 0. dxk = (_) 1. . /\ dxi /\ . . )k By Fubini's theorem and the fundamental theorem of calculus we have [ d(J dx ' 1tk = (_I)i. chasing through the definitions shows that [ dw = [ W. [O. ) . xk ) dx ' . Jac Jatk Theref ore The theorem clearly follows for k chains also . .. . xk) dXi) dx' . .. . . . . . dxk . •:. /\ dxk ) J . Chapter 8 [ d(J dx ' }lk /\ . . latH 1tk [ w = [ c·w. . .(x ' . dxi . . . . dxk [O. . . . .1 1 · · · 1 [/(x'. . . . . . For an arbitrary singular k cube. . 1tk Jatk c'w = [ w . . . O. /\ = = f Di 1 dx' dx. /\ . ) . . xk ) dx' . . lac . I ' I(x . . . . . . . )k (_I )i . . . /\ dxk ) /\ . . )k Thus .. . . )k + ( )' f l(x ' . . (1' D (x' . . dxk ' ' = () 1 . /\ dxi /\ . . . . lc dw = [ c'(dw) = [ d(c'w) = [ Jtl. [O. . . /\ dxi /\ " . . dxi .' l' . . /\ dx k [O.1 f .' f DJ. . xk ) ] dx ' . . .254 On the other hand. . .
we could just as well have used it to show that w = "de" is not exact. but Stokes' Theorem is the tool which will enable us to deal with forms on jRn {OJ. we show that the curve c : [0. I £) f de = e(l .e(£) . we wiII eventually obtain a 2form w on jR3 . but actually becomes that theorem when c = / 1 and w = f. although closed. then we would have d(de) = 0 = o.  1c 1c dl w= = [ I= Jac 1 0 I = o.x2 + y 1c de = 1c y2 dx + 2 dy x2 + y 1c Jac2 1 c 1 c2 "de" really c l (£. then we would have 21f = \Ale were previously able to give a simpler argument to show that "de" is not exact.([0.] Although we used this calculation to show that c is not a boundary.£) .{OJ > jR. sin 21f1). w= x dy /\ dl Y dx /\ dz + z dx /\ dy (x 2 + y2 + z2) 3/2  . is not ac 2 for any 2chain c 2 . 00) x {OJ) > jR: We have or x . For.{OJ. 1] > jR 2 _ {OJ defined by C(I) = (cos 21f1.integration 255 Notice that Stokes' 1l1eorem not only uses the fundamental theorem of cal culus. [There is also a noncomputational argument. For example. Tfwe did have c = ac 2 . if we had w = dl for some Coo function I : jR 2 . de = [ de = But a straightforward computation (which will be good f the soul) shows that or = 21f. using the fact that is de f e : jR 2 . and e(1 .£) .e(£) > 21f as £ > 0. As an application of Stokes' Theorem.
and the usual orientation on JR"). will be denoted by > 1 with sup If w is an arbitrary nform 011 M.256 Chapter 8 which is closed but not exact. phisms in a neighborhood of [0.. each contained in some ('([0. where c is a singular ncube of this sort. Assume that c. are both orienlo. but a straightforward calculation shows that dw = 0.([O. then there is a cover I!J of M by open sets U. and write port w C The only problem is that c. l C] w. C2 then PROOF We want to use Corollary 2. However. IJ". I]") . There are lots of ways of doing this. •:.l W• CI C20{c2. and ('. a glance at the proof of Corollary 2 will show that the result still follows. if <I> is a partition of unity subordinate to this cover.lion jm!servillg (with respect to the orientation jJ. I]"). I]") and c orientation preserving. Let M be an nmanifold with an orientation {/" and let c" c.{OJ. 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 . on M. 11" (it does satisfy det{c2 . : M be two singular ncubes which can be extended to be diffeomor > c. but they all turn out to give the same result. the reason will be clear from the next result. because of the fact that support w C (' .([O. If w is an nf orm 011 M such that support w c [0. The common number 1 w=l C2 1 w . for singular ncubes c : [0.l a c) )' 2: 0. since ('1 and ("2 are both orientation preserving). . which is basic for our definition. ([0.1 OC] ) w= . THEOREM. ([0. I ]" 5. This is possible only when M is orientable. I]") n c. we first want to describe a way of integrating nforms over nmanifolds. is not defmed on all of [0. I]" ) n c. I]" M ('([0. I]" ) . For the moment we are keeping the origin of w a secret. w. In fact.
. then these sums are all finite. M .1nlegralion is defined for each ¢ E <1>. . This is a simple conse quence of Proposition 1. We wish to define 257 We will adopt this definition only when w has compact support. of M we clearly have IM. and the last sum can clearly also be written as [ 1M W = ¢Lc'pJM ¢ · w. then IM..!') W . Ifwe have another partition of unity \jJ (subordinate to a cover <9'). then if J is orientation preserving if J is orientation reversing.) \t\'ith minor modifications we can define J w even if M is an nmanifold withboundary. since support w can intersect only finitely many of the sets {p : ¢(p) i' OJ. E [ so that our definition does not depend on the partition. which form a locally finite collection. (We really should denote this sum by for the OIientation  {/. and w is an nform with compact support on N. /\ dX" = f f. fM Jdx' /\ .!')W' However) we usually omit explicit mention of p. If M C IR n is an ndimensional manifoldwithboundary and J : M > IR has compact support. if J: M n > N n is a diffeomorphism onto.!') W M where the right hand side denotes the ordinary integral. Likewise..1M. in which case the sum is actually finite.
we can define w = Idx ' /\ . But this cannot be true on all of M. THEOREM. (x.r · Idx ' /\ . not every volume element arises in this waythe form 1}p may not vary con tinuously with p. imbedded in ]R 3 Since Mp can be considered as a subspace of ]R3p. since there is no IIform � on M which is everywhere nonzero. then on U we can write we calI w a Coo volume element if is COO. . U) is a coordinate system. . for any n vectors VI . Theorem 77 has an obvious modification for volume elements: w(p)(vp. Howeve.. /\ dx"l· . . . V) a coordinate system around q = E N. /\ dxnl J J for J ?: 0. then for nonnegative V > ]R we have J*{g Idy ' J(p) PROOF Go through the proof o f Theorem 77.e. and (y. 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). ( It is not hard to see that w is a volume element. •:.. n ( Mp) . /\ dyn J) = g: (g I ( a(y'o J) ) 1 0 f) . . . det a. w is of the form w = I�I for an l1form � . One way of obtaining a volume element is to begin with an IIform � and then define w(p) = 1 � p) l . U) is a coordinate system around p E M. For example. there is a way of discussing integration on nonorientable manifolds. . 10calIy. consider the Mobius strip M. . . . Suppose that w is a function on M such that for each p E M we have for some �p E g. If J : M > N is a Coo function between l1manifolds.258 Chapter 8 Although IIforms can be integrated only over orientable manifolds. pUlling i n absolute value signs in the right place . i. If (x. /\ . wp) = area of paralIelogram spanned by v and w. 77'.
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
(2) If the theorem is true for (n .J.269 9.I )form � o n JR " such that w = d�.n. /\ dx". Let w = 1 dx' /\ . Step I . N]. then it is true for ffi.I)manifolds. N) and a constant Cz on (N. Since support w is compact.(_ J)i' (10 . we have an explicit formula for �. 1"' 1(1 . . In fact.(M) '" IF!. . There is some function 1 (not necessarily with compact support) such that w = dJ. .. THEOREM. ) . We know that there is an ( n .' . oo). p) dl xi dx' /\ . .c has compact support. /\ . then it is true for any connected oriented nmanifoId. (3) If the theorem is true for JR". . dl 0 outside some interval [N. Let w be a I form on JR with compact support such that III w = o. Moreover. . If M is a connected orientabIe nmanifoId. �(p) " = {. We wiII establish the theorem in three steps: = f N constant c] N on ( 00. PROOF. . so 1 is a (I) The theorem is true for M = IF!.c) Step 2. then H. Therefore c] = Cz = c and we have w = d(J where 1 . For simplicity assume that support w e {p E JR" : ipi < I ) ..i /\ . in particular for S". /\ dx" be an nform with compact support on JR" such that III" w = o. from Problem 723.
X Xi by Lemma 7.d(hr*A» : . Thus. IR" > IRn w = d� = d(� . . g(J' = snl. r*(J'( p) II g: snl IR by g(p) = 10' u n1 f(u· p) duo A = {p IRn : I pi f = 0. r*(J'(p). /\ dxn = = r JR. .. > E > I I we have so on we have Moreover. Let h : [0. r*(J' = r*(g(J').1 f ( u . I] be any Coo function with h = I on A and h = 0 in a neighborhood of o. II this becomes . Then hr* A is a Coo form on and � = r*(dA) = d(r*A). . A �(p) = (10' u n . .is for Stefl 2. /\ dxi /\ . /\ dxn L i=] = (t i r l f (u · � ) dU) .II JB w o. g(J' = dA for some Hence (11 . . by the hypothe.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 /\ . .I )form g(J' on 1s"1 r � = (g 0 r ) . r f dx1 /\ . !!. by Corollary 8 we have for the (11 . � ) dU ) .2)form A on snl ..
with U diffeomorphic to IRn . Choose forms Wi with support Wi C Vi n Vi+ l and JVi Wi =j:. we can write where each ¢iW' has compact support contained in some open set Vi C M with U. Using the connectedness of M. In other words. o. Step 3. Since we are assuming the theorem for ffi. we want to show that there is a number c and a form � with compact support such that w' = cw + d�. From this we clearly obtain the desired result. Choose an IIform w such that JM w i' 0 and w has compact support contained in an open set U C M. it is easy to see that there is a sequence of open sets U = V" .n we h ave WI . diff eomorphic to IRn . . . have compact support (C V. with Vi n Vi+ 1 =j:. .C1 W = dl}l Wz .w' = ("jW + d1Ji' for each i.). .CZWI = dI}2 w' . and �. Using a partition of unity.gral. with ¢. Vr = V diffeomorphic to ffi.l1. 0.d(hr*).lille. It obviously suffices to find c.CrWr.l = d1Jr : where all �. . we can assume w' has support contained in some open V C M which is diffeomorphic to IRn.) has compact support.:. If w' is any other nform with compact support. since on A we have � .) = � .oll 27 1 the form � .d(r*).d(hr*).) = o.
using the orientation of Vi which makes Xi : Vi 7 ffi. Xi) where each Xi 0 Xi+ l . Taking w ' = W. THEOREM. such that fu w i' 0 (this integral makes sense. .l are positive.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. Since M is not compact. Now if M is unorientable. i)w = d� c =j:. If M is any connected nonorientable nmanifold. there is an infinite sequence U = UJ . there is such a sequence where Vr = VI but Xr OX1 . then Hn(M) = O. U4. I I . Wi > 0. the numbers Ci = H. 1 0. we have Iv. since U is orientable). U2. Choose an Ilform W with compact support contained in an open set U diff eomorphic to JRn . Consider a sequence of coordinate systems ( Vi. then PROOF. . Choose the forms Wi in Step 3 so that. It obviously suffices to show that w = d� for some form � with compact support.1 is orientation preserving. : PROOF Consider first an Ilform w with support contained in a coordinate neighborhood U which is diffeomorphic to JRn . O . I t follows that Wi = cw + d� where c > O. then also !Vi+1 Wi > O.n orientation preserving. JVI [ Wi /J� [ Wi . • .272 Chapter 8 The method used in the last step can be used to derive another result. •. If M is a connected noncompact nmanifold (orientable or not). THEOREM. UJ . Consequently.(M) = o.
C2W2 = d�.. + c. we have W d�. are eventually outside of any compact set. we have seen that ¢u. but the sequence is still eventually outside of any compact set). . such that w Wi Then = = c. d�2 + C. i' 0. + C. . Now choose nforms Wi with compact support contained in Vj n Uj+h such that Ju W. + C. U U U . . d�2 + c. whose union is all of M (repetitions are allowed. Cj+lWi+l + d1]i+l i 2: J . . Let {¢u) be a partition of unity subordinate to !9.+. + C.C2C3d�4 + . There are constants C. C2C3W3 = where the right side makes sense since the V. n V. then for each V. d�. may intersect several Vj for j < i. V2 . w.w. . has support contained in V. . V. i' o. d�2 + c.+2 Hence . and V. Now it can be shown (Problem 20) that there is actually such a sequence V. If W is an nform 011 M. + d�. and forms �. c2d�3 + c. C V.Integra/ion 273 of such coordinate neighborhoods such that V. V3 . V. = w = d�. .w = d�.c2d�3 + C. . Since any point P E M is eventually in the complement of the V.+.. The cover !9 = {V) is then locally finite. where �.'s. with compact support . and such that the sequence is eventually in the complement of any compact set.
then H. consider the case k o.(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. where each IR.{OJ) i' 0. so 1* takes Z k (N) to Z k (M). ctEA PEB takes the element {cp) of EBpEB IRp to {c�}. On the other hand. If N is also disconnected. with components M.EA E9 IRp > E9 IR. 1* also takes B k (N) to B k (M). We also know that Hn.274 Chapter 8 SUMMAR OF RESULTS Y (I) For IRn we have (2) If M is a connected nmanifold. I f M i s disconnected. '" IR. If w is a closed kform on N..) C Np.1 (IRn . since we will eventually improve it. Ci E A . This shows that 1* induces a map = also denoted by 1*: /* : Hk (N) > H k ( M ) . fJ E B. with Cith component equal to c. For example. 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. with components Np. . then HO( M ) i s isomorphic to the direct sum E9 IR. where c� /* : the map 1* takes C E IR into the element of EB IR. but we have not listed this result. If N is connected. then 1* : HO(N) > HO(M) is just the identity map under the natural identification of HO(N) and HO(M) with IR. then HO(N) is just the collection of constant functions c : N > IR. = . then I*w is also closed (dl*w I*dw = 0). Then I*(c) = c o l is also a constant function. If M is connected. then = cp when I( M. since I*(d�) = d(J*�)..
L w. J.l.P for and Vq for Nq) is orientation reversing. ) and (N. J M) and JR (and similarly for N) it J. and the only one we are presently in a position to look at. 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. THEOREM. if Mp . it is almost unbelievable that this number is alwqys an integer. Let oriented nmanifolds ( J. Let q E N be a regular value of I (q). > N be a proper map between two connected 1 2. let For each P E sign J: M. then we have a map M a = J.p: Mp J pJ= J !*p w L J*w = (deg /) Lw { if + Nq is orientation preserving (using the orientations J. so we really want to compare M J*: H" (N) L J*w > H"(M) and Wo for an nform on N. Until one sees the proof of the next theorem. Choose one number such that w a is an isomorphism of H " ( Since r> follows that for eVeI). and a number deg such that for all forms on N with compact support. L J*w = a .l.Integral. M J1 !. is the map and N are both compact connected oriented nmanifolds.on 275 A more interesting case. but is proper (the inverse image of any compact If set is compact). form we have w fM w w 1M J*wo = a · L woo with fN Wo Lw i' o. There is when no natural way to make H"(M) isomorphic to JR. v).
J(W'). Then J(W') is a closed set which does not contain q.(V. So Since J is a diff eomorphism from each Vi to V we have L J* w = {. . Let J' (q ) = { p " .276 Then Chapter 8 L signp J (= 0 if r ' (p) = 0). · · ·UVk • Finally. PEj. •:. We want to choose a coordinate system (V. and let deg J = U W' C M be the compact set w' = J' ( W) . and the Vi are disjoint.I] tIllS proves the theorem . .' (q) PROOF Notice first that regular values exist. first choose a compact neighborhood W of q. xi i around Pi such that all points in Vi are regular values of J.I (q ) is finite. y) around q such that J' (V ) = V. Iv w if J is orientation reversing. To do this. . 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 . /\ dyn where g 2: 0 has compact support contained in V. redefine Vi to be Vi n J' ( V). ii J*w k U · · · U Vk !. . u . by Sard's Theorem. pd· Choose coordinate systems (Vi . I . . We can therefore choose V C W . This ensures that J' (V) c V. Now choose w on N to be w = g dy ' /\ . Then support J* w C V. . · · · Vk . so the sum above is a finite sum. . . since it is compact and consists of isolated points. Moreover. Vk .
If f. then the maps /* : H k (N) H k (M) g* : Hk (N) H k (M) > > > > are equal.1.l (p) consists of just one point. . 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 . depending on whether II is odd or even. J] we defined a (k .(H 0 io)*w = i. Two functions J.io*(H*w) = d(IH*w) + /(dH*w) = d(IH*w) + 0. but we need to intro duce another important concept first. I ) = g(p) for all p E M.inle. H: M with x [0. •. Since A . I ] > N the map H is called a (smooth) homotopy between J and g. We used this fact to show that all closed forms on a smoothly contractible man ifold are exact. we conclude that deg A = (. Any element of H k (N) i s the equivalence class [w] of some closed kform w on N. g: M H(p. We can HOW prove a more general result. J] = g* . N are smoothly homotopic. g : M > N between two Coo manifolds are called (smoothly) homotopic if there is a smooth function As an immediate application of the theorem. N with g*w . 1 3.*(H*w) . J* PROOF By assumption.1 ) " . we compute the degree of the "antipodal map" A : S" > S" defined by A (p) = po .I )form /w on M such that i l*W .J*w = (H o i) *w . Recall that for every kform w on M x [0. THEOREM.gmlioll 277 We can draw an interesting conclusion from this result. O) = J(p) H(p. But this means that g*([w]) = J*([w]) : .io*w = d(Iw) + /(dw). We have already seen that A is orientation preserving or reversing at all points. Then J = H o io g = H o i) . there is a smooth map H : M x [0.
g : M > N are homotopic. 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. . and therefore in S"p . X2 . •.l . Since the identity map has degree ) .) The vector field on S" can then be used to give a homotopy between A and the identity map. consider the retraction r: JR" . COROLLARY. .{OJ is the inclusion. Il is even. I ) = Yp (I ) : .l > JR" .278 Chapter 8 1 5 . Define S" is ( .l is the identity of S". 1 4. . . X"+ l ). then we can construct a homotopy between A and the identity map as follows. For each p. . If M and N are compact oriented nmanifolds and the maps j.J)".l . But if there is a nowhere zero vector field on S". then deg j = deg g. For another application of TIleorem 13. X"+l ) E S" we define = For n odd we can explicitly construct a nowhere zero vector field on S" .l 7 sn.l If i: S". . If vector field on S" . . this is perpendicular to p = (X l . p (Xl . r (p) = p/ ipi. there is a unique great semicircle Yp from p to A (p) = p whose tangent vector at p is a multiple of X(p). then r 0 i : sn.{OJ > S". (On S l this gives the standard picture. I . COROLLARY. For H (p. .
= "r( p) (t=O) p(/=1 1 A retraction with this property is called a deformation retraction. we have Hnl (jRn .{OJ > jR" .{OJ) is the closed form r*(1'.{OJ i 0 I" ( p) = p/ i p i H(p. The manifold for all I . A generator of Hn l (jRn . Thus In particular. the maps (r o il' and (i 0 1" ) ' are the identity. We need one further observation.I ) . but it is homotopic to the identity.// ' i 0 1" : jRn .I)r(p) E jRn . of course.integration 279 The map is. Thus. we can define the homotopy H by .{OJ. M x {OJ c M x jR' is clearly a deformation retraction of M x jRt So Hk(M) "" Hk(M x jRl ) . not the identity. we have and 1" ' 0 i' = (i 0 1" ) ' i ' 0 1" ' = ( I" 0 i)' = = identity of Hk( sn. identity of Hk(jR"  {OJ) So i* and r* are inverses of each other.{OJ. l ) Ip + ( J . Whenever I" is a deformation retraction. for the case of s nI C jRn . We are now going to compute Hk (jRn .{OJ) "" jR .{OJ) for all k.
280 pter 8 Cha 1 6. 0]) [0.0.O) x [O.{(O.4 .{ (O. . there is a similar argument. I) A = w.{(O. .1) Since A and B are both starshaped (with respect to the points (0. Induction on n.{all r B so clearly JA .3 . using A = w. I) and (0. oo)}.0.O) B w. We claim H I (w. The first case where there is anything to prove is 3 .c) B) w = d(J and JA . n= PROOF.n . THEOREM. Let w be a closed I form on w.0.{O) = 0.c = Is on A n B.4 . (0.I we have H k (w.{OJ) = H k (snl) = 0.O. respectively). If w is a closed I form on ffi. 00») . x w" B) I =° and A n B = [ w.3 Let A and B be the open sets (0. Th us w is exact. . then we obtain Iforms �A and �B with w = d�A w on A = d�B on B. on A n B.I). For ° < k < /1 .{(O. 0. 0) x (00 . 0. there are Oforms JA and Is on A and B with w = dJA l w=dB Now d(JA on A on B. 0]) B = w. If w is a closed 2form on w.I is a constant on A n B.4 .2 .4.3 .3 . for on A on B A w = d(J . 0) = x x (00.
since this is not defined on A . we cannot simply consider �A .�B = dA for some Oform A on A n B. .d(<PBA) = �A . = �A + (<PA .<PB dA .(A n B).integration 281 Now d(�A . note that there is a partition of unity {<PA . So we can define a Coo form on IR n .d(<PBA) on A on B. on A n B =) =0 Now. on A n B on A . and �B + d(<PAA) on B.{OJ = A on A. The general inductive step is similar.d(<PBA» = d(�B + d(<PAA» so w is exact.dA + d(<PAA) is a COO form on A is a Coo form on B.{OJ: <PA + <PB d<PA + d<PB support <PA support<PB CA C B. = d�B /\ A U B by letting it be �A .d A. Clearly. Bj of IR J . then <PBA <PA A On A n B we have �A . Unlike the previous case. if <PBA denotes { <PBA O and similarly for <PAA.d<PB /\ A = �A .J) dA + d<PA = �B + d(<PAA). •:.�B) = 0 and SO �A . To circumvent this difficulty. w = d�A = d(�A . <PBj for the cover { A .
B we have d� = O. where 2B denotes the ball of twice the radius of B.. 0 < k < n. Let / : w..n.n) = 0 is left to the reader.n with compact support. For 0 � k < 11 we have H: (w.. 1 7 . which we will need in Chap ter 1 1.n . PROOF The proof that H�(w.. Then on A = w.n) = o. Since A is diffeomorphic to O = w = dry w..n > [0. THEOREM.I )form � on w.2B.n ..{OJ and k  I < 11 .d (/A) dearly has compact support contained in 2B . Let B be a dosed ball containing support w..I we have from Theorem 16 that � = dA for SOme (k .d(JA» . Then d(JA) makes sense on all of w..n and w = d� = d(� .n .. Let w be a kform on w.2)form A on A . the form � . We know that w = d� for some (k . I] be a Coo function with / = 0 in a neighborhood of B and / = I on w. •: .282 Chapter 8 We end this chapter with one more calculation.
t] shaded lengths . P. P � n L(J. Mi(f) Ii].�) = L J(�i)(I' IiI)· Clearly L(J. b]. We call J Darboux integrable if the sup of all L (J. S. . . P. (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. . .I . (d) Let J M on Let P = < . P) equals the inf of all U (J._] .i� o S(J. P. b]. P) S(j. Let [a. b]. P. b].) [a.lnlegralion 283 PROBLEMS I. < s } and Q = Io < . . < be the inf of of [a. . P) = L Mi(J) ' (Ii . li]. Pl. A choice for = with E [ I We define the "lower sum". {so { [Ii. . [a.add up to e. The Riema1l1l integral UeJJUS the Darboux integral. this sup or inf is called the Darboux integral of J on We call J Riemann integrable if .sum of lengths of all [s.�n) �i M. . . I � I I I. letm Ii] = length of the limit is called the Riemann integral of (a) We can define even if is not bounded. b]. P). s. Show however._I . b] > IR be bounded. [a.�) U (J. I .�) [a.]'S cmltained in [Ii. S(J.= ] S [a. b].I . S J on J J 1I). that cannot exist if is unbounded. .�) (b) If J is continuous on then J is Riemann and Darboux integrable on and the two integrals are equal. and define is an ntuple = similarly. [a.b]. let = on [Ii_I.. "upper sum". I ti_l . [a. b]. < In} be two partitions of For each i = I. n. [a. and "Riemann sum" for a partition and choice by J: P I 111i mi(J) In} P � (�I" I"i] . For a partition = { o < .Ii. P) = Ll11i(J) ' (Ii IiI) i=l J U(J.] which are contained in [li_] . b] ei 111 S s [a. [S.=1 S(J. .lil ) . b].
i=l There is a similar result for lower sums. and CR. (h) Show that fe f dx + g dy. I] JR2 .n : [0. let CR.n .il� o L(J. b]. then f is Riemann integrable on [a. sin 2Jr/ ) on [0. CR2.)f(�i)g(�'. It is called the winding number of C around 0. then U(J.{OJ such that ac.£7= 1 e. I I c· de.M.� U(J.b]) H ). (e) Show that . (a) Show that there is a singular 2cube c : [0.{OJ is any cUlve with c(O) c(I). and deduce Darboux's Theorem: ). Compute fe de = f[o. IF JR2 .1 ) = [ fg· Ja i=l Hint: If l g l :s M on [a .CI . show that there is some such that is a boundary in JR2 . b]. equals > > H O n UPU+O 2.284 Chapter 8 Show that.n (t ) = () = (cos 21f1. (R cos 2 1fl.)e. then 1J(�'. Q) : Q a partition of [a. li]. (b) If c : [0. (f) If f is Darboux integrable on [a. b]). Q) : Q a partition of [a. Q) + iL(M .{OJ. > R > 0. )I. =l :s U(J. P) :s U(J. For 11 an integer.m) L ei. Show that fOJ choices �. R sin2n1fI). (g) (Osgood's Theorem). ° as I PI I 0. b]. I]. Let f and g be integrable on [a. )1 :s MIJ(�'i )f(�.11 . Q) + (M . if Mi denotes the sup of f on [If_I . defined as a limit of sums. P) = sup{ L (J. where C I 3. I] n JR2 . b]. P) = inf{ U (J.{OJ be defined by (c) Show that n is unique.)g(�'. b lim L f(�i)g(�'i)(li .11 = > CR"n  n > = C .�' for P.
x'. .co. . Some approaches to integration use singular simplexes instead of singular cubes.I . ! is a boundary.c. a.CR. Although Stokes' Theorem becomes more complicated.I ) a. . Hint: Note that CR". Let /: <C > <C be a polynomial. . .. n M. as indicated in the next Problem. .. aC = L (. n (t) = [CR. Define cR. . i=O 0 < i ::: 11.{OJ. x' ] . .(" n . · + . . .+ . Define a.. . xn .I ) a. ! .) . Then we define and for singular nsimplexes c. ! = / O CR . +an . let r : " n > JRn be the inclusion map. /(2) zn +a l znI + . 1] > <C by cR. where n :0: I . I ' = 5. then CR .'£7. .c = C o a. xn .(x) = (x l . As before.I ) in " n . . Let !J. n i s the boundary of a chain in <C . zn z (b) Show that /(z) = ° for some z E <C ("Fundamental Theorem of Algebra").x' . and an nchain is a formal sum of singular nsimplexes. ! : [0.Integration 285 n l /(2) = Zn ( 1 + . O. then cR. ! . and write 4. . n C JRn be the set of all x E JRn such that A singular nsimplex in M is a Coo function c : 6. I (I)] n . x l . : " n I > " n by 7 ao(x) = ([I . .J)'a. there are some advantages in using singular simplexes. define (a) Describe geometrically the images (b) Show that a2 = 0. Hint: If /(z) i' ° for all z with Izl :S R. . a a (a) Show that if R is large enough.
for 0 :0 :0 I . IRn . We then define (c) Show that we do have c = au + c' where c' is degenerate. a. [ dw = [ w lc Jac for any (k . ? . that is. and x ' E ao("' kl . Hint: If au = L. dxi /\ .) (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). (b) Let c : [0. For any singular k simplex c : 6. using either simplexes or cubes.C2 is a boundary. c(x ' ). c'([O. .J )form on IRn . Show that c is not the boundary of any sum u of singular 2cubes.c. and prove that 1 /" dw = [ w. k 7 c(x) = I .286 (c) Show that if w = J dxl then /\ . Every x E "' k + 1 can be written as ' IX . what can be said about c for chains c in the obvious way.a.I )form w. show that CJ . L. I] > IR 2 be a closed curve. define c: "' k +1 > IRn by IV!orcover. x' is unique except when t = O. I]) is a point. • . /\ dxn Jalll is an (n . . Ozapter 8 /\ (Imitate the proof for cubes. (a) Show that a c = 0 implies that c = at. 6.
) Hint: Given I : Sl > M. any space M (not necessarily a manifold) is called simplyconnected ifit is connected and any continuous I : S l M is (continuously) contractible to a point.) 9. Hint: If we do have w = curve c we have 7. (Converse of nV UV V Problem 724. Show that U is not smoothly contractible to a point. Let w be a J form on a manifold M.Integration 287 curve c in M. then any smooth map I : Sl > M is smoothly contractible to p. It is > 1 w = l(c(1» . then for any not hard to show that for a manifold we may insert "smooth" at both places. then M is simplyconnected. (c) sn is simplyconnected for n > J . (a) Let U C IR 2 be a bounded open set such that IR 2 . 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.U. (d) If M is simplyconnected and P E M. Hint: Show that a smooth I : Sl > s n is not onto. partition S l into s a finite number of intervals each of which is taken into either U or V. > dl. (See Problem 7. [Actually. Suppose that Ie W '" 0 for every closed 8. (f) If M is simplyconnected.U is not con nected. then M is simplyconnected. (c) This is false for open subsets of IR 3 . (b) A bounded connected open set U C IR 2 is smoothly contractible to a point if and only if it is simplyconnected. (b) Sl is not simplyconnected.) Hint: If p is in a bounded component of IR 2 . then H I (M) = O. (This gives another proof that n is simplyconnected for n > J .l(c(O» . Show that w is exact.] (a) If M is smoothly contractible to a point. (e) If M = U where U and are simplyconnected open subsets with U connected. show that there is a curve in U which " surrounds" p.
We define this common sum to be fM w. Let be an nform on an oriented manifold M n .) (c) Show that L"' E 'l' f . M (b) Show that L 1M ¢ .I )"/n and support I C Un A n . (a) Let M be E IR 2 : < J }. n + J) be closed sets. together with a proper portion lj! « x. (a) This implies that L¢ E¢> f ¢ . Show that w. " . w = L L 1M " + w. w = x dy. J ) and i s a Oform whose support i s not compact. Let I : IR IR be a Coo function with fAn 1 = ( . I dx converge absolutely Il. there are only finitely many " which are nonzero on support ¢ . even though both sides make sense. ¢E ¢> cpe¢ cpe¢ l/re\ll cpe¢ l/fe\ll > 1 0. 1 2.y) l(x. using of its boundary.) 0) Similarly. · Iwl < 00. Find two partitions of unity <I> and such that L¢E ¢> f ¢ . (d) Let A n C (n. find a counterexample to Stokes' Theorem when M = (0.. and suppose that w I I ." laM Problem 1 0. w = L lM .. w converges absolutely. Let <I> and lj! be two partitions of unity by functions with compact support. and show the same result with w replaced by Iwl. Iwl < 00. (Note that for each ¢ . (e) Examine a partition of unity for (0. J) by functions with compact support to see just why the proof of Stokes' Theorem breaks down in this case. and that M L lM ¢ . Following Problem 712. L dw . define geometric objects corresponding to odd relative tensors of type (�) and weight w (w any real number). · w .y)1 (No computations needednote that equality would hold if we had the entire boundary. "' to different values. and let o w . I dx and L E 'l' fIl.288 Chapter 8 L 1M ¢ ..
Integration 289 13. · · · . . . (Note tbat the map i o r : JR" . 1 7. and let w and � be an nform and an mform with compact support. vn. . We will orient M x N by agreeing that V I . Show that de is 0 at some point. and e is an (n . " " Un. . Let MI . Account for the factor I/Ipl" in Lemma 7 (we have r* (vp) = (J /lpl)vr (p) . . N is h 1fI*W /\ 1fz*� for some w and �.{OJ is just r. Wm are positively oriented in Mp and Nq. since there are n . respectively. . . (a) Let M" and Nm be oriented manifolds. M2 C JR" be compact ndimensional manifoldswithboundary with M2 c MI . ) = q 1+ h (p. 1 4. Show that for any closed (n . WI . V" and W I .d· 1 5.J )form on M. Use the formula for r*dxl (problem 41) to compute r*(J'. respectively. . )�.{OJ. If 1f1 : M x N > M or N is projection on the ith factor.J )form w on MI. " (p. . . q). Wm is positively oriented in (M x N )(p. on M and N. . show that (b) If h : M x N > [ 1f1*W /\ 1fz*� = [ W · [ �. then where g(p) = (c) Every (m + n)form on M x IN h (p. 1M IN JMxN JR is Coo. .I .aMI. VI . considered as a map into JR" .{OJ > JR" . . .1 vectors 1 6. but this only accounts for a factor of 1 /Ipl" .q) '" Mp Ell Nq if VI .) r * (1' = r * i*(1 = (i 0 r ) *O". . Suppose M is a compact orientable nmanifold (witb no boundary).
Z E (00. zl. Show that E JR"p and let v E JR"p be (Ap)p for (b) Let M C JR" . we define ¢(x. (c) Let M C JRn{O) be a compact (11 .{OJ . y .290 Chapter 8 1 8.J )manifoldwithboundary which inter sects every ray through 0 at most once. y = 0.�l E JR3 except those with x = 0. (a) Let p E JR" . . For this n 1 9. . The latter integral is the measure of the solid angle sub tended by reason we often denote r * (J' by de . y. M Show that [ r *(J' = [ r *(J' .. " " Wn2 some A E IR. 1M 1c(M)ns2 M.{OJ be a compact (n . A ::c OJ . �) to be the angle between the positive zaxis and the ray from 0 through (x. . y. and let C(M) = {Ap : p E M . .I )manifoldwithboundary which is the union of segments of rays through O . For aJJ (x. Let WI . Show that JM r *(J' = O. 0]. . . Wn2) = o. C (Ml C(Ml n S' D·. r *a' (v. W I .
y. (e) Conclude that (J' = d( . z) = vix2 + y2 (J' = h de /\ d¢. y . z) E S 2 (r).z) (x. If w points along a meridian through p = (x. ) . and e is considered as a function on ]R 3 .. (c) If v is a longitudinal unit tangent vector on the sphere S 2 (r) of radius r.Integration 291 (x. y) (a) ¢(x. ¢) is a coordinate system on the set of all points (x. e. X E [0... (the minus sign comes from the orientation). 0].z) = arctan( vlx 2 + y 2 /z) (with appropriate conventions). ¢ to [certain portions h: S 2 > ]R is h(x. then de(wp) = Y X2 + / (d) If e and ¢ are taken to mean the restrictions of e and of] S 2. e(x. then (v. y = 0. Z E (00. y..y. y. . y.cos ¢ de ) . (b) If v (p ) = Ipl. then where .z) arctan y/x. then d¢(v) = 1 . 00) or with x = 0. z) in ]R3 except those with y = 0.
+an/'(. S 2(1 p l). Let > Nn be a proper map between oriented IImanifolds such that p 7 Nf(p} is orientation preserving whenever p is a regular point. Vi r*de de .) = IIZn111 +(11 . ih. = J* : I: Mn M (e) Write that . 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). 22. (a) Show that a polynomial map ce > ce. dBn_1 IRn1 dBn IRn Vi n VjVI V2 V3 . Prove that a connected manifold is the union the are coordinate neighborhoods. (b) Let J Show that we have i where varies over complex numbers.y)... de = r2*i*(J.y)+i ax (x. I/'(x iy )12 DI(x.y) i ay (x. and the fact that p vr(p)/lpl for tangent to (h) Conclude that = = * (v ) = (i) Similarly. so that is the projection.y). v dB3 r*(J' d(cos(¢ r) de) = d(cos¢de). Show that if N is connected. express 20.)a "n2I: . for the form (J de IR3 is just 1C*de. or else all points are critical points of J. ( + nl l. l w I(x+i ) u(x. given by = is proper � J). 0 on {OJ in terms of on IR2 r2*de = de . .292 (f) Let {OJ on Show that If > for the form (Jzapter 8 > IR2 1"2 : IR2 1C : IR3 IRde2 Sl be the retraction. = and Show Hint: Choose w to be a real (d) Conclude that av /'(x+iy) = aau (x. then either 1 is onto N.To[/(" + w) yl(z)]lw. with eventually outside of any compact set. .y) x aav (x. and the sequence is 2 1 . . {OJ. where i' 0. ·+a _ . U U U .y)+iv(x. y au h..y) == for realvalued functions I(z) zn+aI Zn1 + I'(z) u v. and then to be == + det ..
while DI is the linear transformation defined for any differentiable I: JR 2 > JR2 (e) Using Problem 21. Let M" . (a) Show that this definition agrees with that in Problem 3. and conclude that I takes on all values.I c JR" be a compact oriented manifold. then the number of points in I.M.I (a) is a locally constant function on the set of regular values of f Show that this set is connected for a polynomial I: ce > ce. (b) Show that this definition does not depend on the choice of L:. (c) Show that w is constant in a neighborhood of p.{p) > L: be the obvious retraction.Integmtion 293 where J' is defined in part (b).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). Define the winding number w (p) of M around p to be the degree of rp lM. (f) There is a still simpler argument. Conclude that w is con stant on each component of JR" . Let p and q be points . (d) Suppose M contains a portion A of an (n . choose an (11 . Let I"p : JR" . Show directly that if I : M > N is proper.M.I )plane. For p E JR" . 23. give another proof of the Fundamental Theorem of Al gebra.M.
t) g(p) t g t. by a smooth homotopy M x [0.M has at least 2 com ponents. then IRn . 1 H p (a) Being smoothly isotopic is an equivalence relation. 1] > be a regular value of Show that Let (mod 2).294 Chapter 8 (a) Let of points in is a compact I manifoldwithboundary. (Show that I"q l M is homotopic to a map which equals 'p l M on M . g: H: N. H(p. and 0 elsewhere. More precise conclusions are drawn in Chapter I I. and leaves all points outside the unit ball fixed. q E N I (q). let IR x IRn > IRn satisfy ¢: (Each solution is defined for all by TheOI·em 56. which is smoothly isotopic to the identity. in general. #11('1) I#rl (q) '" #gI (q) Hil l: H1(q) 24. and let homotopic. g. t. For E s n.I . 25. N we will write / :::e g to indicate that 1 is (a) If l :::e then there is a smooth homotopy Q)) 26. Show that = ± 1. The next few Problems show how to prove the same result even if M is not orientable. close to this plane. For two maps M > smoothly homotopic to = I.) (e) Show that. I] > N such that for in a neighborhood of 0. if M is orientable.x» · p x at H(O. more gencrally. but on opposite sides.x) x. that this result holds so long as 'I is a regular value of both and I. (b) Let IH:" > IH: be a Coo function which is positive on the interior of the unit ball.g:g. = == x H(t. The number of points in its boundary is clearly even.t) /(p) t H'( p. Let M and N be compact nmanifolds. H'(p.) ¢(H(t. x) . H.) (b) Show. (This is one place \·vhere we use the stronger form of Sard's Theorem. for in a neighborhood of I .A and which does not take any point of A onto the point in the figure. x w(q) w(p) M > N be smoothly denote the (finite) number 1 g.) Show that each 1+ is 1I diffeomorphism. H: p aH(t. 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.t) �is an equivalence relatiOll. = H': M x [0.
This number is called the mod 2 degree of f.) From the addendum to Chapter 5.* (Lx' w. which is generated by {X ' } .lnl£gmlion 295 (c) Show that by choosing suitable p and I we can make H (1. then XI (p) will denote 1CM * X(p . (e) Use part (d) to give an alternate proof of Sl£p 3 of Theorem 9. q E M.1 )Wo + IWI · Show that the family cp.). it follows that there is a Coo family {cp. for any Coo function I : M > JR we have I( ( I(cp. = (l . (p» (Xlf)(p) = lim CP'+h P» . *w. I1 h.M has at least 2 components. = cp. show that if M c JR" is a compact (n .. + w. of kforms on (where #1.:.Wt .0) be any point in the interior of the unit ball.1 (q) is defined in Problem 24). (d) If M is connected and p . O. Let {X' ) be a Coo family of Coo vector fields on a compact manifold M. we have �.e. (g) By replacing "degree" with "mod 2 degree" in Problem 23. and the argument which was used in the proof of Theorem 56.WI · .. and I : M > N. O For a family w. q2 E N we have 27. (To be more precise. 1 ] . of diffeomorphisms generated by {X'} satisfies for ali I if and only if LXI Wt = Wo . then there is a diffeomorphism I: M > M such that I(p) = q and I is smoothly isotopic to the identity. i.J )manifold. then for regular values ql ... o h� h (a) Show that for �(t) = cp.. with <Po = identity.. (f) If M and N are compact nmanifolds. and define w. suppose X is a Coo vector field on M x [0. Wt = I " .. (b) Let Wo and W I b e nowhere zero nforms o n a compact oriented nmani fold M. then JR" . } of diffeomorphisms of M [not necessarily a J parameter group] . M we define the kform Wt +h .
J ] > JRn and K: with JRn be smooth homotopies H(p. t ) : q Show that n E N } = 0 for every I . Define (d) Suppose that JM Wo = JM W I . J A (u v) 1C o o [r (u. and I( M ) g( N) = 0.l(p)I ' i(f. q ) = We define the linking number of I and g to be '" g(q) I(p) Ig (q) . n Cif. N x [0.I r (g(q) . Show that there is a diffeomorphism II : M > M such that Wo = J. J ). O) = g(q) K(q.296 Chapter 8 (c) Using Problem 718. Let I : M k > JRn and g : Nt > JRn be Coo maps.  1 1 10 10 . n = k + 1 + J.O) H(p. where M and N are compact oriented manifolds. (b) Let H: M x [0. J) such that = = I(p) K(q . 'v)]3 du dv. so that Wo . where M x N is oriented as in Problem 18. g) (c) For I. J ] > (a) i(f. g : S l > JR3 show that i ( f. *W I ' . J) = g(q) j(p) (H(p.g .g : by M x N > s n. t ) : p E M} (K(q . g) = deg Cif. g) = (Jlt + li(g.g (P. i(f. show that this holds if and only if 28. g) = 4 = i(j.I(p» C JRn  (OJ  Cif.W I = d A for some A. g).
b'. c) = 41C for (a. b.J'(u) ( ) (the factor ) /41C comes from the fact that Is' (J ' = 41C [Problem 914]).b'. Show that (VI )p.b'. (a. c) is on the same side as a vector wp E ]R 3p . (U» (/1 )'(U) (g3 )'(V) (g 2)'(V) A (u. Suppose (a.c) . c') = 41C.b.a) dy /\ dz (y .b.b. b. y)plane). c) . b.c') p triple wp. (a) For (a.c) = 0 for (a.c) /(a'.c')'J> P Hint: First show that if and M = aN.(a. (a. (V2 )p is positively oriented in ]R 3p when oriented in Mp.c) and (a'. b.c)'I> p (a'. c') be points close to p E M. g) = 0 if J and g both lie in the same plane (first do it for (x. b' . v) = Ig(v) . (a. b. . c) = Let (a. c) rt N. (a.JI (u) g2 (V) . g) without calculating.p (u) g3 (V) .b. b. on opposite sides of M.b.c) _ For a compact oriented 2manifoldwithboundary let  (x . (d) Show that C(/. v) = det (g l )' (V) g l (V) .J(u) 1 (P)'(U) (/3 ).M g.hltegmtion where 297 1" (u.c) E ]R 3 define dEl(a. b .  . b. M C ]R3 and (a. c) E N .b) dx /\ dz + (z c) dx /\ dy [(x _ a) 2 + (y _ b) 2 + (z _ C) 2]'/2  g. 29. c) rt M.Mp for which the Wp 1M dEl(a. (V I )p. b. (a' . (V2)P is positively lim g. The next problem shows how to determine C(J. {a. then g.g.
. Show that hut the surface on the rightincJuding the hemisphere behind the pJane of the paper is orientabJe. . . c) = S J* .(X .(Y . c) . z) I' I _ 1 ( ( ( l g*(dg.. z) 1 3 X . 1 38.(Z . y. See Fort.C) dx . y.lSI J l(x.. Let ..a) dy . Topowgy qf 3Manijolds. (An M with this property always exists.=  J 41T (c) Show that ag.g) for the pairs shown below. + . .g).b) d. z) 13 ag. [ * Z . n = . Compute i(f.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. and .+ be the number of inter sections where dgjdl points in the same direction as the vector wp of part (a). that when g(l) = P E M we have dgjdl rt Mp.[ * y . b .lSI J l(x . y.the number of other intersections.n.). s' ) ) ) (d) Show that n = i(f. ac I(x.(a. c) .C) dy aa (a.a) dZ ab (a. pg.b) dX ag.) Let g : S I > IR 3 and suppose The figure on the Jeft shows a nonorientable surface whose boundary is the "trefoil" knot. b. b ..
31..J . show that A B po oq Hk (JRn .. J* (Ci V fJ ) = J*Ci V J*fJ· (e) The crossproduct x : H k (M) x HI (N) . q} so that A and B are diffeomorphic to JRn ...{p.. If Ci E H k (M) and fJ E HI (M). � is closed.(OJ. We define the cup product v : Hk (M) x HI (M) .1 (JRn ..Integration 299 30. Hk+l (M) by (a) (b) (c) (d) Show that v is welldefined. and Ci E H k ( N ) . w 1\ � is exact if w is exact and Show that v is bilinear. i. and that Hn. Hk+1 (M x N) is defined by Show that x is welldefined.F where F C JRn is a finite set. q E JRn be distinct. If J: M . fJ E HI ( N ) . then Ci V fJ = (J llfJ v Ci . Choose open sets A . (b) Find the de Rham cohomology vector spaces of JRn .e. q } ) = 0 for 0 < k < n . then [w] v [�] = [w 1\ �].(p . N. and that .....(p. q }) has dimension 2. (a) Let p . B C JRn . Using an argument similar to that in the proof of Theorem 16. and A n B is diff eomorphic to JRn.
(b) Show that every map J: sn let . given by "(p) Ci V f] = "* (Ci X X = (p..300 (f) If that . denote 1Cj*de. ) Equality is proved in the Problems for Chapter I I . Tn has degree O. > X n times . p). Sl . (a) Show Ihat all de i l /\ . show f]). where 1fj : Tn 7 S l is projection on the jth factor. Hint: Use Problem 25. On the ndimensional torus Tn = S l de .. : Chapter 8 M > M x M is the "diagonal map". . /\ de ik represent diff erent elements of H k (Tn ). • . Hence dim H k (Tn ) � (. by finding submanifolds of T n over which they have diff erent integrals. 32. .
An inner product on a vector space V over a field F is a bilinear function from V x V to F. and conse 30) . v) i' o. b). ffi. a. .=1 > 0.CHAPTER 9 RIEMANNIAN METRICS exceptionwe have never mentioned inner products. a ») = " (a ')2 L n r In particular. ) is clearly nondegenerate. The time has now come to make use of this neglected tool. . . . .=1 For this inner product we have (a. the field F will always be JR. (a 1 . a."L. b) .=1 . a symmetric bilinear function ( . v). albl . then there is some w i' 0 such that (w. an ) . n previous chapters we have exploited nearly every construction associated and nondegenerate: if v i' 0. but there has been one notable w) = (w. which is symmetric. and thus with bundles. r :0. ) on (a. 11. a'b'. = "L. . . for r = 11 we obtain the "usual inner product". . . I with vector spaces. (v. a'b'. i=r+ l n this is nondegenerate because if a i' 0. v) > 0 A positive definite bilinear function ( quently an inner product. . For us. In general. ( . . for all v i' o. . ) is called positive definite if (v. a) > 0 for any a i' o. denoted by (v. . we can define an inner product ( . = "L. then n ( a 1 . w). For each r with 0 :0. . w) 1+ (v. ).n . on JRn by (a.+1 .
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*. ) corresponds to the more complicated condition that the matrix {gij } be "positive definite". with (v. J. ) gives lIS an iSOITIOI"phism a : V + V*. Thus. and J: IR > IR2 is J(a) = (a. ) implies that <pv i' 0 if v i' O. for all a l . ) is positive definite. . . Thus. with corresponding dual basis v*} . ) on V is an element of 7 2 (V). .9 Notice that an inner product ( . wI· Since ( . e. . However. an inner product ( .a j > 0 L . ) is positive definite if and only if J is oneone. det(gij) i' 0. then J* ( . so if > V is a linear transformation. Vn of V. if ( . v*u. . for each V E V.a 2 b2 . } is linear in the second argument. w) = a(v)(w).' dimensional. gij a. nondegeneracy of ( . ) is a symmetric bilinear function on This symmetric bilinear function may be degenerate even if J is oneone. (gij) is symmetric. a n with a t least one a i =j:. ) is cquivalenl to the condition that (gij ) is nonsingular. the map v !)o ¢v is a linear transfor mation fi'om V 10 V*. . . o. b ) = a I b l . we can define a linear functional ¢v E V*. W W.. . meaning that . For allY basis V I .g. bv <pv(w) = (v . ) i!i linear in the first argument. ) is clearly nondegenerate if J is an isomorphism OllW V. ) is defined on IR 2 by (a. )= i. Nondegeneracy of ( .= 1 . we can write ( . . . a).j=i L gijv*i 0 v*j .302 ChajJle. . Also. ) implies that the matrix The matrix {gij } has another important illl crpretation. . n I n this expression. so symmetry of ( . if V is finil<. Since an inner product { . Vj ) ' gij = gji. gij = (Vi . J* ( . . Clearly.. then J* ( . if ( . Positive definiteness of ( .
so 0 < II Av . E V we have E IR.w ) = A 2 11 v ll 2 .V) (the positive square root is to be taken). •:. THEOREM. equality clearly holds.j=1 . by (2) The function II II has certain unpleasant propeJ1iesfor example. v + w) = II v ll 2 + II w ll 2 + 2 (v . the func tion I I on JRII is not differentiable at 0 E JRIIwhich do not arise for the function II 11 2 .w for all A (I) lIav ll = JaJ · II v ll . For all v . w ) II v ll 2 + II w ll 2 + 2 11 v ll . AV . then 0 # AV .:5 o. . w) J :5 II v ll . (3) II v + w ll :5 II v ll + II w ll (Triangle inequality). w PROOF.ja. (I) is trivial.w ll 2 = (AV . Il t jvf t 1=1 i. (2) J (v . If not. a ) = L )a. i=l The principal properties of II II are the following.l 303 norm In Given any positive difmite inner product ( . ) on V we define the associated II II by II v ll = j(V. Il w ll = (II v ll + II w ll) 2 . ) simply by J(a. I . IRn we denote the norm corresponding to ( laJ = .)2 . w) + II w ll 2 So the right side is a quadratic equation in A with no real solution. a j . and its discriminant must be negative.2A (v.4 11 v ll 11 w ll < (3) II v + w ll 2 = ( v + w.Riemannian 1I1etri(.W . (2) If v and w are linearly dependent. 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. II w lL with equality if and only if v and w are linearly dependent (Schwarz inequality). Thus 2 2 4(v. W) 2 .
ancJ clearly IIvl li = 1. ) . then (I) " i. . /(b» .w 1l2 ].IIw1l 2] (v. Similarly. 3. Chapter 9 11 11 2 = L gijV*.j=i II II is the norm associ (2) (v. . Vi) = DU' (Such a basis is called orthonormal with respect to ( . there is only one positive definite inner product. Vk so that .'O inner products which induce the same norm are themselves equal. . a. ) . w) for all v. w E V. ' V'j. Theorem 2 shows that tv. "up to isomorphism". . (f(v). ) is a positive definite inner product on an Ildimen sional vector spare V.IIvll2 . . then there is a basis VI . .IIv . THEOREM.: there is an isomorphism f : JRII T V such that (a. If ated to an inner product ( . Vn for V such that (Vi . 2. I n other words. ) on V. \Ve will now see that. II/(v)1I = IIvll for all v E V. that is. ) = ( . I(w» = (v. PROOF Let WI . . •:.b E IR". W) = Wv + w ll2 . We obtain the desired basis by applying the "GramSchmidt orthonormalization process" to this basis: Since W I # 0. then I is also inner product preserving. PROOF Compute. } � k VI . w) = Wv + wll 2 .304 More succinctly. that is. • • . . . b) = (f(a). if I: V + V is norm preserving. w" be any basis for V. THEOREM (POLARIZATION IDENTITY). If ( .) Consequentl. An invariant definition of a quadratic function can be obtained (Problem I) from the following observation. Suppose that we have constructed 1 :'0 i. we can define /* ( .
) on V is sometimes called a Euclidean V. Let It is easy to see that . ) on V provides an isomorphism Ci : V + V* with Ci(V)(W) = (v. w). . . This is because we obtain a metric p on V by defining p(V. . . . . . • • . Vk . . We have only one more algebraic trick to play. the symmetry of ( . We also call IIvll the length of v. So we can define and continue inductively. The "triangle inequality" (Theorem I (3)) shows that this is indeed a metric. Now. Then W k +1 is linearly independent of V I . k . Recall that an inner product ( . w) = IIv .. A(V). W k . defined by i(V)(A) we obtain an isomorphism fJ : V* + = a. : melric on A positive definite inner product ( . . .ic.wli. = I. .:::::: span Wj .I V We can now use fJ to define a bilinear function ( . .I 305 and span V I . . • . Vk . . ) can be expressed by the equation Ci(V)(W) = Ci(W)(V). i Using the natural isomorphism i: V + V**. ) * on V* by + ( V*)*.Riemannian 1\1el.
choose a basis {v.j=l + L g. if we let g U b e the entries of the inverse matrix. so that then ( . and let ( Then . To see what this all means. 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. ) * independently of the choice of basis. without the invariant definition. ) * = L gij V**i ® V**j i.) {v*.1 V· + Thus.} and {v·*.j v*. ® v*j . . ( . )= .} and {v*.} and {v.) .. Consequently ( . let {v*. a(w) = /1.306 Letting this can be written which shows that Clzapw 9 a(v) = ft.) (gU ). One can check directly (Problem 9)...} be the dual basis for V*.)=1 n n = . (g U ) = (gU )I .1 So aI : V* fJ : + {v*. the one which produces fJ). (gU). if we consider Vj E V**.} for V.j= 1 L g ij Vi ® vj . )* is an inner product on the dual space V* (in fact. that this equation defines ( . )* is also symmetric.
• It clearly follows that v *1 /\ We have thus distinguished two elements of [I n (V). will have length 1 . . we are interested in only one case. .Riemannian 1I1etric. . It follows Iiom Theorem 75 that for any W E [I"(V) we have = L Cik i Cikj . . /\ . ) . we need only describe which two elements. so ( . ). . . . Let V I . . /\ v*n for {v. letting a(v) = A. . . )* if and only if v" . . v*n of V* is orthonormal with respect to ( . This can also be checked directly from the definition in terms of a basis. which implies that det A = ± l . The vector space [I n (V) is I dimensional. wn ) . We will call these two elements /\ V*n = ± W * 1 /\ • . . If we write n Wj = L CijjVj. Vn is orthonormahvith respect to ( . and [l k (V). Similar tricks can b e used (Problem 4) t o produce a n inner product on all the vector spaces T k (V). . . w and w. . . they are both of the form v * . which we will not describe in a completely invariant way. vn ) = ±w(w" . . for A # 0. . In the positive definite case. TdV) = T k (V*). Vn and WI" . So to produce an inner product on it. . k�' " w(v" . we have a(v)(v) > ° for v # 0. )* is also positive definite. so that then. . /\ W *1/. . . the simplest way to describe ( ) )* is as follows: The basis V*l. However. Wn b e two bases o f V which are orthonormal with respect to ( . . 1'=1 then So the transpose matrix A t of A = (aU) satisfies A · A t = 1. .' 307 Notice that if ( . . } an orthonormal basis of V. ) is positive definite.
. )p on .' (p). vn ] = /1.. we will call it the positive element of norm 1 in n n (v). \IVe now apply our new tool t o vector bundles.308 Chapw 9 the elements of norm 1 in n n (v). . If � is a Coo vector bundle over a Coo manifold we can also speak of Coo Riemannian metrics. . . . we define a Riemannian metric on � to be a function ( . . . • .s /Jositivf. then we can further distinguish the one which is positive when applied to any (VI . then In particular. . Consequently. " so if A = (Ci. ) which assigns to each p E B a positive definite inner product ( . . " . det(gU) is a/wll). . Will we choose an orthonormal basis v} . Ifwe also have an orientation J1. Wj ) . 0 w*j .j ). . /\ · · · /\ w * " gu = (w. n then = k� ' L CikiCikj . Vn and write Wi = j=l L Cijj Vj o n Problem 79 implies that Ifwe write ( .j=l L gU w*. If � = 1f : E T B i s a vector bundle. vn ) with [v" . the elements of norm in n"(V) arc * ± Jdet(gu) w . .. )= i. the function is also continuous. To express the elements of norm 1 in terms of an arbitrary basis W I .. and which is continuous in the sense that for any two continuous sections S1 . S2 : B T E.
) u i s a [COO] Riemannian metric for � I U. v. PROOF There is an open locally finite cover (9 of M by sets U for which there exists [COO] trivializations On V x IU : ..I (p).Riemannian 1I1etricJ 309 [Another approach to the definition can be given. Then there is a [COO] Riemannian metric on �. THEOREM. a fibre bundle. and let Eue(�) = U ElIe(.1 (p» . 2: 0. ) is continuous [COO] and each ( . For v. )p is a symmetric bilinear function on .I (p)... there is an obvious Riemannian metric. W E . lu (w» p . Let � = . : E + M be a [COO] k plane bundle over a Coo manifold M.] 4. ) on TS 2 which gives a symmetric bilinear function on each S2p which is not positive definite or negative definite but is still nondegenerate. Let ElIe(V) be the set of all positive definite inner products on V.] Notice that the argument in the final step would not work if we had merely picked liondegenerate inner products ( . The only problem is that Eue(V) is not a vector space.I (p) by ElIc(.W)p = L ¢u (p) (v. ) by (v. + « p .. define Then ( .I (p). Let {¢u} b e a partition of unity subordinate to (9.. w)� UeC9 (v. and for some U strict inequality holds. there is no ( . ) u In fact (Problem 7). W E . .. a ) .b» p = (a. I(U) U x IRk IRk . •:. v)� [The same argument shows that any vector bundle over a paracompact space has a Riemannian metric. but an instance of a more general structure..I (p» . w)� = (lu (V) . Ifwe replace each . UeC9 each ¢u(p)(v. the new object Elie(�) that we obtain is not a vector bundle at all. (p.v) � . p eB then a Riemannian metric on � can be defined to be a section of EliC(�). To show that it is positive definite.. v)p = L ¢u (p) (v. b ) . Then ( . note that (v. We define ( .
S(p» p = I . ) p on . a Coo Riemannian metric ( . [S(p)] = /J.1 (p)] * . W E 1[ . All these considerations take on special significance when our bundle is the tangent bundle TM of a Coo manifold M. •:. M x IR by ALTERN ATIVE PROOF We know (see the discussion after Theorem 79) that if � is orientable. We "p (v)(w) = (v. ) for TM. •:. If � = 1[ : E + PROOF. Clearly s is a section. If � has an orientation Riemannian metric on M then there is a unique and ( . Continuiry of ( . If � = 1[ : E and only if � is orientable. which gives a positive definite inner product ( . COROLLARY. Let ( . ) is a with (S(p) .310 Chapter 9 As an application of Theorem 4. ) be a Riemannian metric for �. we settle some questions which have till now remained unanswered. COROLLARY. A ) . ) implies that the union ofall "p is a homeomorphism from E to E' = UPEM[1[ . 6. Then for each I' E have an isomorphism defined by M is a k plane bundle. PROOF The "only if" part is trivial. w) p v. But � :::: �* . then there is a nowhere 0 section of [1 1 m = �*. then � "" �*.1 (1'). then � is trivial if /J. so that �* is trivial. J. we then define an equivalence ! : E + !(A S(p» = (p. M.p . + M is a I plane bundle. In this case.
) on N. So for every Coo map f: N T M there is a covariant tensor f* ( .j dx. then /J. even though it is usually not d of anything (even when M is orientable and it can be considered to be an nform). . for each p E M the Riemannian metric ( . 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» . of course. and det(g. ) is positive definite.Jdet(g. then on V this form can be written .Jdet(gu(p» dx 1 (p) /\ . We have seen that they can be written ± . is called a R emann an metric on M.j) > 0 since ( . /\ dx n (p). The Riemannian metric ( .j ) Similarly. ) on M determines two clements of n n (Mp).I 31 1 each Mp.j ) Idx 1 /\ This volume element is denoted by dV. ® dxj . ) * .Ib·emalUzian A1etric. a covariant tensor of order 2. is a contravariant tensor of order 2. and we obtain an nform on M. it is a Riemannian metric on N if and only if f is an immersion (f p is oneone for * all p E N). in a coordinate system (x. dx 1 /\ .Jdet(g. U) is a coordinate system on M. n If M has an orientation /J. the elements of norm I. V) it can be written as . If (x. ) is symmetric. we obtain a "volume element" on in Chapter 8. i. ) on M is. A Riemannian metric ( . . i i ) = L g.. as defined Even if M is not oricntable. which ( . the k=l L g. which is dearly symmetric. .p allows us to pick out the positive element of norm 1. ) induces on the dual bundle T*M. if x : V T JRn is orientation preserving. • • /\ dx" l · . M. j= 1 \vherc the Coo functions gij satisfy gjj = gji. thus p. then on V we can write our Riemannian metric ( . ) as ( .k g kj = 8( . . since ( . /\ dx" .
) 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 . and in the noncompact case (see Problem 810) it either converges to a definite number. ® dx. L� (y ) I" If y is merely jJuuwise smootli. we have tangent vectors y' (t ) = dt dy E My (t). b] + M.] (with possibly different = = t I dr I dl Y from a to b. There is an even more important construction associated with a Riemannian metric on M. For every Coo curve y : [a . and can therefore use ( . 1. If M is an IIdimensional manifold (withboundary) in IRn . We can then define the This certainly makes sense if M is compact. which will occupy us for the rest of the chapter. with the "usual Riemannian metric" 1M dV. ).312 Clzapw 9 and is called the volume element determined by volume of M as ( . /\ and ccvolume" becomes ordinary volume. = ) (t lI y'(t) 1I d ) = ._. ) = L dx. . meaning that there is a partition a to < ' " < b of [a. then gij = 8U. .. dV = I dx . b] such that y is smooth on each [/. in which case we say that M has "infinite volume".. so . or becomes arbitrarily large over compact subsets of M. dxn l .=1 ' /\ .
=1 this definition agrees with the definition of length as the least upper bound of the lengths of inscribed polygonal curves. Whenever there is no possibility of misunderstanding we will denote L� simply by L.a.]) . A little argument shows (Problem 1 5) that f piecewise smooth curves in or JRn . we can define the length of y by L�(Y) = I >:. ® dx. 1. b . bJ + JR. i=1 .and righthand derivatives at 11 ..a). .t}. [ I �. with the usual Riemannian metric L dx. We can also define a function s : [a.II .aJ by defining y(t) = For the new curve new y we have = 1.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. + constant. In . s(b) = L�(y). the "arclength function of y" by s(t) = L�(y) = Naturally. . We can reparameterize y to be a curve on [0. I dl. . .l (y I [IH . Then b a = I �. s(t) = L�(y) = L�+a(y) = old s(t + a) . y(1 . . (*) s'(t) = Consequently dyjdl has constant length I precisely when S(I) = I thus precisely when s(t) = I .
It i s clear that d(p. the Riemannian metric was not a function on tangent vectors. Moreover.) Consequently. i f r E M i s a third point. If p. then there is at least one piecewise smooth curve y : [a. �' gU dx. q) 2: 0 and d ( p. we can choose piecewise smooth curves y. (Classically. this is sometimes interpreted as being the equivalent of the modern equation ( . .j=1 and we would use the same symbol today. q E M are any two points. ® dxi . q) = inf { L ( y) : y a piecewise smooth curve from p to q}. equation (*) says that for each curve y and corresponding s : [a.dxi dxi ® dxj was very strange: i. ) = L7.8xi n where dx and 8x are independent infinitesimals.dxi appearing here is Tlol a classical substitute for dx.j=1 L gu dx. but what it always actually meant was i i. ® dxi the value (dx. ) . but the inner product of two "infinitely small displacements" dx and 8x. Define d(p.) Consider now a Riemannian metric ( . The classical way of indicating one wrote L g.r) < t. : [a.314 Chapw 9 Classically.dxi)(p) of dx. ) on a cOTlnected manifold M.b] + [b.d(p.c] + Y2 : M M from from p to q with L(y.j=l L gu dx. the norm II II on M was denoted by ds.i dx.dxi at p should not be interpreted as a bilinear function at ail. but as the quadratic function i. q) < t q to r with L(Y2) d(q. then for any £ > 0. b] . b] + M from p to q (there is even a smooth curve from p to q). (This makes some Sort of sense even in modern notation. b] + IR we have Ids l = Y* ( II II ) on [a. in classical books one usually sees the equation ds2 = Nowadays.dxi n II 11 2 = The symbol dx. p) = O.
The function d : M " M + lR is a metric on M.c] + M to be YI on [a. and consequently for any curve y : [a.I .a) = lIap lip. G(p. ). except that it is not so clear that d(p .I be a = b/lb l . and if p : M " M + lR is the original metric on M (which makes M a manifold). be the "Euclidean" or usual Riemannian metric on U. . Since B x SIl . since Ib I G(p. m < G <M Now if P E B and 0 # bp E lR"p . q) + d(q. this gives the desi�ed inequality (which clearly also holds for b = 0) . r) 0. M > 0 such that m . ) be any other Riemannian metric.Riemannian Metric. it follows that :::: d(p. PROOF Both parts of the theorem are obviously consequences of the following 7'. LEMMA.b] and Y2 on [b. Let U be an open neighborhood of the closed ball B = {p E lR" : I p l :::: I } . then Y is a piecewise smooth curve from p to r and L(y) = L(YI ) + L (Y2 ) < d(p. let a E S". p).l is compact there are num bers m. then (M. d) is homeomorphic to (M. L dxi ® dxi.c] . i=1 " and let ( .b] + B we have mL. a) < M lb l. there would be difficulties in fitting together YI and Y2 . lI ap lip = lI<l b l a)p llp = II b llp. I I on B. a) = Ib l . PROOF Define G: B " S"I + lR by Then G is continuous and positive. Then mlbl < Ib IG(p . 7. but d would still turn out to be the same (Problem 1 7). Then there are numbers 111 . ). I I :::: II II :::: M . •:.(y) :::: L(y) :::: M L. Since this is true for all £ > d(p. and II II be the corresponding norms. q ) > 0 for p # q . r).(y). Let I I = II II. = [Ifwe did not allow piecewise smooth curves.] The function d : M " M + lR has all properties for a metric. This is made clear in the following. THEOREM. r) + 2e. M > 0 such that on B " S". let ( . q) + d(q. ( .1 315 If we define y: [a.
g.316 Chapter 9 Notice that the distance d(p . x .f'(t» a b dl . the two semicircles between the points p and p on S l ). e. For example. I n order to investigate the question of shortest curves more thoroughly. and q might be p. we have to employ techniques fi'om the "calculus of variations". if" F(t. As an introduction to such techniques.b] + IR with f(a) = a' and f(b) = b' one a' which will maximize (or minimize) the quantity t F(t. q) = L(y) for some y. if d(p. among all functions b' f: [a. f(t). For example. we consider first a simple problem of this sort. . Of course.. then y is clearly a shortest piecewise smooth curve from p to q (there might be morc than one shortest curve. q ) defined by our metric need not be L(y) for any piecewise smooth curve from p to q. Suppose we are given a (suitably differentiable) function F. y) = �. the manifold M might be 1R2 {OJ. lR x lR x lR + R We seek.
To solve this problem. Similarly..y + l) .Riemannian 1I1etrics 317 I r> then w e are looking for a function I on [a. then we are trying to minimize the area of the surface obtained by revolving the graph of I around the xaxis. a') and (b.b] which makes the curve (t. but critical points arc the only candidates for maxima or minima if I is everywhere differ entiable. This is the same as saying that the curves I r> l(x + l. we examine the critical points of I. or even a local maximum or minimum. y) t r> l(x.y) = o.e. i. l(t» between (a. y) E ]R2 for which (*) D.y) = 21f X JI+Y2 . b') of shortest length t )1 + [f'(t)F dl. if F(I. those points x for which I'(x) = O. for a function I : ]R2 + ]R we consider points (x.y) = Dd(x. x. /(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. As a second example. A critical point is not necessarily a maximum or minimum.
This can be done by considering a "variation" of f.= du u=o du I aa at I Ib F (l. because of the chain rule. I) = f(I). I).(U. e) to the set offunctions f : [a. f'(I))dl a : (e. by considering curves in the sel o all f IR.b) = b' b' a ' a b for all u E (e.b] + we wish to proceed in an analogous way.e.b] + IR a(O.y). We might try to get more information by considering 0 = (f 0 e)' (O) for every curve e : (. e). & (u)(b) = b'. To find maxima and minima for funclions f: [a. e) x [a.e) + 1R2 with e(O) = (x.e ) which pass through f for u = O. that is) a function J (f) = t F(I. then we call a a variation of I keeping endpoints fixed. l) are then a family of functions on ( e. If each &(u) satisfies &(u)(a) = a' . b] + IR.a(u. We will denote this function by &(u). Thus & is a function from (e. such that The functions I r> a(u. f(t). but it turns out that these conditions follow from (*). .318 Chapw 9 have derivative 0 at the condition O. I)) dl u=o a .a) = a' a(u. in other words if a(u. For a variation Ci we now compute dJ(&(u» d .
d1(1). �7 (U. 1(1). (I» a aF aa a a I Ib U (0. 1 (I» b 11 =0 = a X .dt = a . t)) ] dl Since a 2 ajaual = a 2 ajalau. ofcourse. the variations Ci were taken to be of the special form . thus obtaining (*)  dJ(&(u» du For variations Ci keeping endpoints fixed. The derivative is called the "first variation" of and is denoted classically by a/ ax Iu�o Ib �(I) [ aF (1.(1. I). flu l u�o J(&(u» J aF d aF 8J ax dl ay Ib � [. = a The final result is. 1(1).1) [.. 1'(1» d ( aF I. /(1) .] dl. l) = 1(I) + u�(t). . I) ay (1.Riemannian 1I1etl1cS 319 = f [ :u lu�o F (I. and we obtain (**) � dJ(&(u» ax au I u�o = Ib aa (0. f'(t» )] dl. f(l). 1(1). f(I). essentially the same.f (� (t. Then we obtain a (u. f'(t)))] dl aa aF ' I + au (0. for some � dJ(&(u» � : [a. (I» ( ay (I. b] + IR with �(a) = �(b) = O.aF1. f'(I))) ] dl. we can apply integration by parts to the second term in the integrand. a(U. 1.dt a ' In classical treatments ofthe calculus ofvariations. the second term is 0.. I) [aF (1.
1 ) au o. Ci 8J(Ci) 8J(Ci) J. So the theorem is a consequence of the following simple 8'. J. .d ( aF (t. b] + IR satisfies l b 1)(I)g(t) dl = 0 for every Coo function 1) on [a.(t» . then g = O. J'(t» ) = ax di ay 1)(t) = aCi (0. . 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. •:' 0= di d ( '/1 /'(1)' 2 ) ' + F(t. y) = !i+7. but the dependence of on is not indicated (which can make things pretty confusing). the arguments of functions are either put in indiscriminately or left out indiscriminatelyin this case. J(t). The particular form (**) into which we have put now allows us to deduce an important condition. 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.320 Chapw 9 As is usual in classical notation. If a continuous function g : [a. THEOREM (EULER'S EQUATION). J Ci J) 8J(Ci) 8J 8. l(t). which vanishes at a and b. /'(1» omitted (resulting in the disappearance of the function I for which we are solving). As in the case of I dimensional calculus. not only are the ar guments I and (t. The Euler [/ (1)] .(t). If I is to maximize or minimize then must be 0 for every variation Ci of I keeping endpoints fixed. x PROOF Choos C 1) to be ¢g where ¢ ispositive on (a. 8J Ci J. consider the case where equation is = ¢(b) = 0.b] with II(a) = 1)(b) = 0. We call I a critical point of (or an extremal for if =0 for all variations of I keeping endpoints fixed. b) and ¢(a) As an example.f. LEMMA.
11+1'2 ' / . where the critical points of are the same as those of since For the case of the surface of revolution. I F(I. (I (I this leads to the equation which we will also write in the classical form 0= )1 + [f'(t)F :!. Notice that we would have obtained the same result if we had considered the case for then the Euler equation is simply This is analogous to the situation in I dimensional calculus.. = dx ' the To solve this. dx d2y2 dy P = y.11 +1' 0 = + 1'2)/" I'!" = I' + 1'2)/". I" = 0. We let Then . ( J7)' = L . 2J7 F(I.y) = 1 + y2. where Euler equation is _ . J7 I.f'2 I/" = 0.( )1l+ [f''(t) F ) . we use one of the �o standard tricks Ocaving justification of the details to the reader). dy dx 1 + ()2 y=O..y) = xJI+Y2. dl (t) f (t) 1 + .�"��� ) . o = �(2/'(1» .x.Riemannian Metrics so 321 hence which implies that So is linear.I �2 o= �(.x.
The graph of X +k = c cosh ( .322 Chapter 9 so our equation becomes 1 + p 2 . e eX + ex x = 2 cosh is shown below. � dy p = dx = Jey2 .l ey Replacing c e by 1/ e. x x � 0.yp dp = 0. it is symmetric about the yaxis. we write this as Y = x +k.) . and cosh . Pp2 1+ � log(l + p2) = log y + constant 2 y = constant . dy .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)'. decreasing for increasing for � o.
(1» ) = . we define a variation of y to be a function aF d (a axl (/.a') and (b..b] + IRn with &(0) = J. 1' (1) ) ] dl 1 X o. dJ(&(u)) dU b � a1 (0. t ) x for F: IR x IRn x IRn + IR. If y : [a . f(l).t) x [a .Riemannian 1\1etnes 323 So our surface must look like the one drawn below.. not trivial to decide whether there a7" constants k and c which will make the graph of (*) pass through (a.(I» I l a. /'(1» dl . with y (a ) = p and y (b) = q. It is easy to generalize these considerations to the case where and J : [a. b] + M is a piecewise smooth curve. (0. b] + M J of J must satisfy the n equations a.di ayFl (I. a . by the way.b] + IRn J(f) = In this case we consider that t F(I. 1(/ ).. Problem 2 1 investigates the special case where a' = b' .. aF u=o [a. 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.1) [D (I. J(I). f.f G. : ( t . and compute = L. : (t.... (1. J'(I)) . J(I). I) aayFl (1. /. � a. b'). 1'(/ » Ib n .. 1(1). It is. J(I).
I ) aa aa"I (u. such that a = 10 < II < .t). As before. (2) there is a partition on each strip We call " a variation of y keeping endpoints fixed if (3) u E (t. I) is the standard coordinate system in (t. we let paths y satisfy & (u) be the path 1 dL(&(u» du r> .(u.] lies in some coordinate system (x._ 1. b) = q for all (I) .. . (0. . _ 1 .r) =" (u..b] so that " is Coo (t.r) . U) (othelwise wejust refine the partition). (. If (u. we will take a hint from our first example and first find the critical points for the CCenergy" I u=o =° E ( y) = z a I which has a much nicer integrand. We can assume that each y l [1.1.t) x [a. 1. < IN = b of [a. dt a dt ' dt aa" .324 Chapter 9 > for some t 0. (u. . I) = " U (u. (Iaa I ) U ( U. . b] we write lb I dy 1 2 dl = Zl ib ( dY dY ) dl.. l) .1) = y (t). I). We would like to find which for all variations Ci keeping endpoints fixed.].a) = p .(u..t) x [1. However. aftenvards we will consider the relation between the two integrals.
i ih. then aa iii (U.I) So = L ii/(U.Riemannian 1\1elr. I) is the tangent vector at time I to the curve &(u).j =l dl dl If we usc the coordinate system x to identify U with as functions on JRn . ti _ l i.dt.y) = '2 L gij (x) .=1 1 l1 [}e/ 0 1 .l) . F(y(l).. i yi i.. I» . Ifwe adopt the abbreviations eti (u.t) ti_ l E(y! [li l .\ 325 Then aet/al(U.c. y'(t)) dt where F(x. ndI dI dyi dyj 1 L gij (y(l» . t) = xi (et(U.j = 1 1 tj _ l " Then and So . a(u. li J) = 2 1 = 2 11. and consider the gij [. (dY dY ) dl ' 1. then we are considering IRn .
.326 Chapw 9 In order to obtain a symmetrical looking result.. ax' ) [ .. i. we note that a little index juggling gives r. j= l ax' dt dt So �. agi/ dyi dyj = � agjl dyi dyj L.1.. i d i dy . dyj dy' = �. Let dyI(l + ) = nght hand tangent vector 0f y at I .L.. j= 1 ax' dl dl i. 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. /] a = � gi/ 2 ( ax' + agjl ax' _ agij . 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.. agl.
(In the case of a variation from 1 to N . but the integral  327 iI. glr (Y(t) dd2. For the time being we present: with apologies. which appears in our result. Il" d/) a. .� (aCi (O. Now we just have to add up these results. even though different coordinate systems may actually be involved (and hence different gij and V i ).1 . the sum can be written [ij. leaving endpoints fixed. From the first variation formula it is.) This result is not very pretty. this coordinate dependent approach. Let � aa: (O._. but there it is. simple to obtain conditions for critical points of E. I] . clearly cannot. Ii]. 1.. THEOREM (FIRST VA RIATION FORMULA).' \�I [ij. . . . later on we will have an invariant interpretation of the first variation formula. Ci .I](y(t) d� d�j n I [ n r n i ] dl. Consequently. 11=0 N dy . 9.fi). t) !. Nevertheless. i = 1. It should be noted that are not the components of a tensor. of course.Riemannian 1I1etrics These depend on the coordinate system.l Then we obtain the following formula (where there is a convention being used in the integral). we will use the exact same expression for each [IiI . For any variation w e have dE(Et(u» du I Ci. N .
. U) we have y: [a b] 11 y i Lg/.] is in U. L. In order to put the equations of Corollary 10 in a standard form we introduce another set of symbols OUf equations Can now be written We know from the standard theorem about systems of second order differential equations (Problem 54). that for each P M and each v Mp. Suppose y is a critical point. I](y(l)) d� d�j = 0 for y (t) E U. U). such that satisfies y: E y E y(O) dy (0) = v (ji d2yk � j dy' dyjt = . choose a partition of [a. which is 0 outside of (1'1 . .I to Ii is written in terms of (x. + M is a Coo path..[i2 + . for some t > 0. but a positive function times the term in brackets on (IiI. then in the first variation formula we can assume that the part of the integral from Ii.b] with I E (1. I!) : n r I. then is a critical point of E� if and only if for every coordinate system (x.328 Chapw 9 1 0 . If . choosing all aCi' jau(O.t) + M. I) to be 0.. ) for some i.' + i. r/k ( y (I » dt d l. (y(l)) d2. and such that y l [I'_I. 1. The final term in the formula vanishes since y is Coo. If Ci is a variation of y keeping endpoints fixed. Given / with y(l) E U. d j =l r=1 PROOF.I.1 . •. there is a unique (t.. except one.L [ij.j=1 =P o. COROLLARY.). Now apply the method of proof in Lemma 8'.
We now have the more precise result. By our previous remarks.cs 329 Moreover. ' [O. + d Y2 dl (0 ) = dl (0 ) y. because the final term in the first variation formula still vanishes. ' I I .. O] + M are Coo functions satisf ying this equation. . Choosing the same as before (all are 0 outside of (1.e) + M and Y2 ' (e. ) = L dx ' ® dx'i=l Here g'j = D. b] and for every coordinate system for y(I) E U.] satisfies the equation. and ifmoreover d y.» . As the simplest possible case. .. this means that is actually Coo on all of [a._. This last fact shows that if y. . (0) = Y2 (0) then y. so all ag'j /ax k = energy function satisfy 0. We already know that the integral in the first variation formula vanishes. (0. N . Now choose Ci so that c/ c/ a. and rt = O. this y is Coo on (e.e). Naturally. A piecewise Coo path for (x.:. So we obtain which implies that all !:1/. PROOF. 1. we see that yl [I' _h l. The critical points y for the .j.1. ./' dt ' aCt dy i = 1 . . consider the Euclidean metric on * y JRn ) ( . are O. U) satisfies Y ' [a. we could replace 0 by any other I.Riemannian 1I1et. COROLLARY. .e).b]. Let y be a critical point.. b] + M is a critical point Eg if and only if Y is actually Coo on [a. I. and Y2 together give a Coo function on (e.) = !J.
PROOF Observe first. that y: E. dy d i dy' dy dyj " d2y dyj " " L agij (y(t» dt dtdt + L g'j (y(t) df'2dt = . when we considered only curves of the form I r> Any reparameterization of is also a critical point for length. +!+ y I dyldlll . this can be written as di dt r n Since is a critical point for E. both terms in parentheses are 0 (CorollalY 10). y y y (t. the component functions of must be linear. }] + [}/.L ax' l. If eterized proportionally to arclength. TllllS the length is constant. THEOREM. The situation is now quite different from the first variational problem we considered. d dy 2 d " gij (y(t» dit dyjt) I 1 = di (i. so is a critical point for the length function as well. y y y [a.�. Replacing agij lax' by the value given above. This situation always prevails.j=l dyi d2y' + L gi. since we have just seen that for to be a critical point for energy. This shows that there are critical points for length which definitely aren't critical points for energy. from the definitions. f(t» . (y(t » dt df2 ' =l i. since length is independent ofparameterization (Problem 16).330 Chapter 9 Thus lies along a straight line. and hence must be parameterized proportionally to arc length. i].b] + M is a critical point for 12.j=l /=l r. then y is param Now w e have ��:{ = [i/ .
m aY il gmj ([lk. Ii] y [a dy/dl (x. For the por tion I of contained in a coordinate system U). g glm ogmj L.m aYk I.m] + [mk. 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. . It will also be useful to know a f ormula for ag ij / ax k To derive one..j] + [jk.. we have y: ..y) = \ iL gij (x)yiyj ..i] ax occurring in this proof will be used on several occasions later on. g gmj oglkm aY I.Riemannian 111etries The formula 331 ·· ag't [ik.b] y [li1. Considering our coordinate system as JRn) we are now dealing with the case F(x... l]) by (*) Lg I. we first differentiate = m=l to obtain Thus we have ogikj L.j=l .
dl . Conceivably a critical point for L might have a kink. and we can consider the reparameterized curve e This reparameterized curve is automatically also a critical point for L . L� (y)] 1 is a diff omorphism. Conversely. the function d /d d2s/dl2 s: [a. the third term vanishes.L ax dl dl aF ( I dy ) _ � i. We introduce the arclength function Then So we have " a · d id j L 4· (y (t» � . given a critical point y for L with y I # ° everywhere.j �1 ds ' y ( ) .332 Clzapw 9 s (t) = L� (y) . There is only one detail which remains unsettled. Since it is now parameterized by y e arclength. . L� (y)] + M. but be Coo because it has a zero tangent vector y oS : [0. so y o S 1 is a critical point for E.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). b] + [0. so it must satisf the same diff rential equation.
d v (0) = v. Observe that the equations f a geodesic. including surveying and the measurement of degrees of latitude and longitude. A geodesic on the earth 's surface is a segment of a great circle. The fundamental existence and uniqueness theorem says that there or is a neighborhood U of p and <] . 2) + M satisf ying Yv(O) = q. 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. which is concerned with the measurement of the earth's surface. 13. This name comes f rom the science of geodesy. which are so far merely known to be critical points f length. as in the figure below. or have an important homogeneity property: if y is a geodesic. Henceforth we will call a critical point of E a geodesic on M (for the Rie mannian metric ( . In this case it would not be possible to reparame terize y by arclength. Before we can say whether this is true for geodesics in general. which is the shortest path between two points. Let p E M.Riemannian 111el1ics 333 there. <2 > 0 so that f q E U and v E Mq with IIvll < <1 there is a unique geodesic . we must initiate or a local study of geodesics. Problem 37 shows that this situation cannot arise. This feature of the equation allows us to improve the result given by the basic existence and uniqueness theorems. »). THEOREM. then I r> y(cI) is also clearly a geodesic. � PROOF. The most elementary properties of geodesics depend only on facts about differential equations.
I dl 1=0 eXpq (c(I» = v. Then expq 0 c(l) = expq (lv).. d"i (0) = v. there is a natural way to give it a Coo structure. y(l).. at v E Mq with Mq itself.« 21) . then the map expq : (9 + M is Coo) since the solutions of the diff rential equations for geodesics have a Coo e flow. exp(v) = expq(v) = (The reason for this terminology will b e explained in the next chapter. If (9 c Mq is the set of all vectors v E Mq for which expq (v) is defined. Choose < < < 1 <2 : Then if Ivl < < and III < 2 we have So we can define yv(l) to be yv/. Identif ying the tangent space (Mq).(v) = :!. . y : [0. •:. we claim that the map (expq)o. : Mq + Mexp. In fact. the geodesic with tangent vector v at time 0) So (expq)o. we can let crt) = IV.334 Chapter 9 with the required initial conditions.. 1] + M If v E Mq is a vector for which there is a geodesic satisf ying then we define the exponential of v to be dy y(O) = q. we have an induced map (expq). to obtain a curve c in the manifold Mq with dc/dl(O) = v E Mq = (Mq)o.) The geodesic y can thus be described as y(l) = expq (lv).. : Mq + Mq is the identity. Since Mq is an ndimensional vector space. (v)· In particular.
. . x l . . C TM. the map expq maps the open eball in Mq diffeomor phically onto an open set Uq :::I W. . 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. . xn ) = (xl. xn .. . . . v(q. q') is a Coo function from W x W + . . 1 4. 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.. . x l . Then is a coordinate system on 1[ . . THEOREM. If (x. . a v = L a' ox.cJ 335 Before proving the next result. so that v= . q') r> v(q. U) is a coordinate system on M. (3) For each q E W. then for q E U we can express every vector v E Mq uniquely as " . xn ) v E Mq. .=1 where 1[ : TM + M is the projection. .l (U).=1 a L X'. q . (v) axi n We will denote ai by Xi (v). Then (q. we recall some facts about the manifold TM. For tangent vectors l (X 0 1f) O .Riemannian 1I1el. (2) Let expq(v) TM. q') denote the unique vector v E Mq of length < e such that = q'. . xn o n.
Choose W to be a smaller neighborhood of p for which F(V') :J W x W. p) E M x M. These fill out Vq . Let (x. Consequently. Define the Coo function F: V + M x M by F(v) = (. I + a!i I F. We may assume that V' consists of all vectors v E Mq with q in some neighborhood V' of p and IIvll < e. F. consider the geodesics through q of the form I r> eXpq (lv) f IIvll < e. If ". : Mp is the identity.. using the fact that (expp)o.(v).. 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. it is not hard to see that at Given a W as in the theorem. We will use the coordinate system described above. p) = . then + M is projection on the i th is a coordinate system on V x U. ( a�i IJ a 1 . •:. !2 .. V) be a coordinate system around p. : M x M factor. exp(v» .336 C/zapw 9 PROOF.p) F 0E F. ( � IJ a!. . a . Now. so maps some neighborhood V' of 0 diffeomorphically onto some neighborhood of (p. for . is oneone at Mp . = 0 E Mp we have (P' p) + Mp (p . The close analysis of or geodesics depends on the f ollowing.1 (V).
vU» . au al expq(O) = q. Thus the second term on the right of(2) is also O. LEMMA (GAUSS ' LEMMA). But aCt/ au(u. or A calculation precisely like that in the proofofTheorem 1 2 proves the following equation. . Let = constant < e}. I).Riemannian Metrics 337 1 5. in which the arguments (u) I) and Ci(U. where IIvU)1I = k.I) ' at (U. and define Ct(U. we obtain 0 since each curve u r> Ct( u l) is a geodesic. I). It follows that (� ' � ) = 0 for all (u. v: lR + Mq be a smooth curve with IIvU) II k < e for all I. I) = a constant We are claiming that for every such Ct we have ( = expq (u . for convenience: The first term on the right is Similarly. vU» I <u< 1. au al .I) = ( � ' � ) is independent of u. aCt aCt � (U. I) = o. so aCt/al(O. I) are omitted. I) )=0 f all (u . I) is just the tangent vector at time u to the geodesiC u r> expq ( u . the geodesics through q are perpen dicular to the hypersurfaces { expq(v) . so lIaCt/aull = k . which is just twice the second term on the right of (1). So But Ct(O. In Uq . IIvll FIRST PROOF.
v(O» . By u).bj + Uq . c(t) = expq(u(t) . I). (0. I). •:.(1») . . defined on [0. ddly (0») au au afJ dy .{q} be a piecewise smooth curve. . COROLLARY. But each curve ('!t (0. � (u) has energy . = (t) r dl =  (� O ( 1\ 2 dl = k2 .0). _ so 1 6. the first variation formula. and define expq(t . � u» 10' 1 0 = dE�(U» lu� o ( = dg �. I ). we have = dE(du(u» 1u=o � E(  = the integral vanishing since y is a geodesic. ddly ( I ») ('!t (0. v(u» fJ(u.( a.. Tt(I) ) ..338 SECO}fD PROOF Chapter 9 Let v: IR + Mq be any smooth curve with IIv(t) II = a constant (note carefully the roles played by I and k < e for ali I. Let c : [a. I]. v(t» .l) = Then fJ is a variation of the geodesic y(t) = expq(t .
] + M be any piecewise Coo path from q to q'. . and lies between them.{q} (otherwise break e up into smaller pieces). Let W and e be as in Theorem 15. let y : [0. I) = expq(u . the path c must contain a segment which joins the spherical shell of radius 8 to the spherical shell of radius r.for ° < u(l) < e and II v (l ) 1 I = Riemannian 111ellics 1. •:.u(a)l.8. and hence 2: lu(b) . We can assume that q' = expq(rv) E Uq .' iii = ) = e(l) = Ct(u(l). so that c is a radial geodesic joining two concentric spherical shells around q . 2 2: lu'(I)12 . By Corollary 16. u is monotonic and v is constant. Then I L(y) ::: L(e). then di de Since ( aCt aCt = 0. q' E W. v(l» . PROOF with equality if and only if u is monotonic and v is constant. and clearly e must be a reparameterization of y f equality or to hold. the length of this segment has length 2: r . COROLLARY. •:. I) and u au (I) aCt . For 8 > 0. So the length of e is 2: r.u(a)l. If Ct (u. I ] + M be the geodesic of length < e joining q. v'(I) = 0. Thus 0. with equality holding if and only if c is a reparameterization of PROOF y. 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 . Then 339 L�e 2: lu(b) .. and let e : [0. 1 7 . a.
Antipodal points on ). the sphere have a continuum of geodesics of minimal length between them. ) ' ) are Coo manifolds with Riemannian metrics. )) and (M'. q E C be two points with a unique geodesic C' of minimal length between them. Then I(C') is a geodesic of the same length as C' between I(p) = p and I(q) = q. depending on how many time� the geodesic goes around the sphere and in which direction it starts. For the isometry I : sn + sn mentioned above. then the length of c with respect to ( . reflection through a plane £2 C )Rn+l is an isometry I : sn + sn. So C' = I(C'). ( . All other pairs of points have a unique geodesic of minimal length between them but an infinite family of nonminimal geodesics.340 Chapw 9 We thus see that su fficient6' small pieces of geodesics are minimal paths for arc length. For example. n Notice that a portion of a great circle which is larger than a semicircle is defll1itely not of minimal length. if we first introduce a notion which will play a crucial role later. )'. which implies that C' c C. without any computations. and if c is a geodesic. We can use Corollary 17 to determine the geodesics on a few simple surfaces. It is clear that if c : [0. 1] + M is a Coo curve. so that C is a geodesic. even among nearb paths. Since there is a great circle through any point of sn in any given direction) these are all the geodesics. . then a oneone Coo function f : M T M' is called an isometry of M into M' if f*( . ( . then f o e is likewise a geodesic. ) is the length of f o e with respect to ( . ) ' = ( . If (M. the fixed point set is the great circle C = s n £2 Let p. ).
In fact) if L is a generating line of Z) then we can set up an isometry I : Z . ... the .L + JR2 by rolling Z onto JR 2 .{OJ. . . In general. and the helices on Z . Notice that on both the sphere and the infinite cylinder every geodesic y defined on an interval [a. b] + M can be extended to a geodesic from JR to M. a manifold M with a Riemannian metric ( . circles cut by planes perpendicular to the generating lines. The geodesics on Z are just the images under I. q) = inf{L(y) : y a piecewise smooth curve from p to q}. This is false on a cylinder of bounded height. 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 ( .I of the straight lines in JR2 Two points on Z have infinitely many geodesics between them..Riemannian 111etlicJ 341 The geodesics on a right circular cylinder Z are the generating lines.. ) is called geodesically complete if evelY geodesic y : [a. a bounded portion of JR".. b] can be extended to a geodesic defined on all of lR. ) and the metric d : M x M + JR it determines. d(p. or JR" .
we clearly have d(p.(d(y(to). q) = r . q) = ��. Given p. s) + d(s .q). 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) = r . To prove this result. drs. r] be the least upper bound of all I f which (**) holds. Moreover. Let S' be a spherical shell p to q must intersect S.r]. There is a point on PROOF. This proves that (**) holds for I = 8.. q») = 8 + d( po. Then or (**) holds for 10 also. q) = Let S C Up be the spherical shell of radius po = expp 8v. First of all. If ( .q) = �ir (d(p s) + d(s.I IE [8. we will prove that S such that d(po. q) for all s E S. Suppose 10 < r. of radius 8' around y(to) and let po' E S' be a point closest to q.8. q) :. by continuity. Now let 10 E [8.342 Chapw 9 18. q») = 8' + d(po' . Then d(y(to). ) is a Riemannian metric on M. So d(po. THEOREM (HOPFRINOWDE RHAM). any two points in a geodesi cally complete manif can be joined by a geodesic of minimal length. since every curve from . q E M with d(p. choose Up as in Theorem 14. We claim that expp (r v ) = q. ). old r > 0. then M is geodesically complete if and only if M is complete in the metric d determined by ( . d(y(t). Suppose M is geodesically complete. q). . 8 < e.
Conversely.Riemannian Metrics so 343 d( po'. So e is a path of minimal length.(10 + 8').10) . bounded subsets of M have compact closure. any geodesic can be extended to ffi. Hence y(lo + 8') = po'. it is not difficult to show that y can be extended past b. so it converges to Some point p E M. if A C M has diameter D. and p E A . q) = 10 + 8'. . Consequently. so it must be that 10 = r. fact. which proves (*). b) + M. then the map expp : Mp + M maps the closed disc of radius D in Mp onto a compact set containing A. showing that (**) holds for 10 + 8'. Given any geodesic y : (a. po') 2: d( p. note that there is always a min imal geodesic joining any two points of a compact manifold. suppose M is complete as a metric space. •:. by a least upper bound argument. From this result. I n other words. Hence d(p.8'.d(po'. In Hence (***) gives . which means that it coincides with y. and must therefore be a geodesic. (**) holds for I = r. q) = r . q) . d(y(lo + 8'). But the path e obtained by following y from p to y(lo) and then the minimal geodesic from y(lo) to po' has length precisely 10 +8'. From this it is clear that Cauchy sequences converge. This contradicts the choice of 10. q) = (r . III other words. Clearly y(ln ) is a Cauchy sequence in M. choose In + b. it follows easily that M is complete with the metric d. Using Theorem 14. As a particular consequence of Theorem 18.
or p eM Let E= U Mp L and w .C Np as MpL = {v E Np : (v. 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 . E + M take MpL to p. the normal bundle of M in N. then the normal bundle 1 of S I in J1> 2 is exactly the Same as the normal bundle of S I in M. with i : M + N the inclusion map. then we can define M/. i. so that for every p E M we have i.(Mp) C Np ' If ( . . On the other hand if M is the Mobius strip and S I C M is a circle around the center. so it too is nontrivial. For example. It is not hard to see that v = w : E + M is a kplane bundle over M. for v has a section consistiIlg of unit outward normal vectors. the normal bundle v of sn 1 C IRn is the trivial I plane bundle. ) is a Riemannian metric for N.344 Chapw 9 ADDENDUM TUBULAR NEIGHBORHOODS Let M" C Nn+ k be a submanifold of N.w) = 0 f all w E Mp} .
for which the Osection s : M + U is the inclusion of M in U. this open neighborhood can be taken to be the whole total space. is smoothly homotopic to the identity of U. or then D is a submanif oldwithboundary o f U. Then there is a neighborhood U of Xo such that IIU is oneone. Before proving the existence of tubular neighborhoods. If 1f : U + M is a tubular neighborhood. So M also has the same de Rham cohomology as a closed neighborhood. Moreover.e) ::: I } . ) f 1f : U + M and define D = {e E U : (e. s 0 1f . LEMMA. For example. so 1f is a deformation retraction. Let I : X + Y be a local homeomorphism such that IIXo is oneone. 345 Our aim is to prove that for compact M the normal bundle of M in N is always equivalent to a bundle 1f : U + M for which U is an open neighborhood of M in N.Riemannian Metric. is called a tubular neighborhood of M in N. 1 9. as an appropriate neighborhood of S l C )R2 we can choose )R2 . A bundle 1f : U + M with U an open neighborhood of M in N. if we choose a Riemannian metric ( . and for which the Osection s : M + U is just the inclusion of M into U. and H k (U) "" Hk (M). thus M has the same de Rham cohomology as an open neighborhood. then clearly 1f 0 S = identity of M. 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. we add some remarks and a Lemma. But in general the neighborhood cannot be all of N. Let X be a compact metric space and Xo C X a closed subset.{OJ.
M) < e). norm I I . . for some p E M } E . x # y and f(x) = f(y») . and compactness of M. Xo) + dry. = {q E N : d (q. e It is clear also that exp(E. Xo). 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). ) for N... Let E = {v : v E Np and v E M/".) C U. that exp is defined on E. dyjd ( dyjdl(O) q. PROOF. then f(x ) = lim f(xn) lim f( Yn ) = fry). To prove that exp is onto U" choose any E U" and a point p E M closest to If + N is the geodesic of length < < with = p and = it is easy to see that is perpendicular to M at p (compare the second proof of Gauss' Lemma). while they do not even appear in the statement. with the corresponding It f ollows easily from Theorem 13. y: [0. which is equivalent to the normal bundle of M in N. f if (Xn l Yn) is a sequence in C with Xn T x and or T y. is compact. Then C is closed. . f sufficiently small < > O. Since C is compact. where we will also need the following modification.346 PROOF. there is e 0 such that g 2: 2< on C. Then f is oneone on the {(x. = V E. . Chapter 9 Let C C X x X be x <neighborhood of Xo. We claim that f sufficiently small e. THEOREM. onto U. 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. Then M has a tubular neighborhood 1f : U + M in N. then g 0 on C. and V. •:. 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. n q q y (0) y ( 1 ) q. Theorem 20 will be needed only in Chapter I I . Let M e N be a compact submanif of N. This will clearly prove the theorem. This means that = expp / O) where E E. x + Choose a Riemannian metric ( . the map or or exp is a diff omorphism f e rom E. and also x # y since f is locally oneone. y) = d(x. If g: C JR is g(x.y) E X > = X: Yn + > old 20. •:. and metric d: N N JR. = {v E E : I v l < e ) U.
Riemannian 111etrics 347 2 1 . •:. with compact bound ary aN. . Let N be a manif oldwithboundary. using only inward point ing normal vectors . PROOF Exactly the same as the proof ofTheorem 20. Then aN has (arbitrarily small) open [and closed] neighborhoods f or which there are deformation retractions onto aN. THEOREM.
q(u) . Let V be an ndimensional vector space. . /\ Wk In Lcfons sur La Ceometrie des Espaces de Rieman1l. . if W = W"' . Suppose ¢i. then f some aij .q(v) .q( u ) . If ¢i = Lj aj.v) = v. . /\ ¢k = 1/11 /\ . and let h : V x F be symmetric and bilinear.) (c) Using Problem 79. /\ 1/Ik . . . ' .q( u + w ) . . v). . Let ( . . . show that ¢1 /\ . j =l L Qij v*. and the signed volumes of the parallelepipeds are ¢ the same. . . be the subspaces of V* ¢ spanned by the ¢i and 1/Ii. . 1/Ik. and let W and W". 2.q(v) ] . and that iJ(u. . Uk be an orthonormal basis of W¢ = W". A basis VI . . · v*j n q ( . ) an inner product on V which is not necessarily positive definite. Show that q (u+ v + w ) . + Let V be a vector space over a field F of characteristic # 2. Show that if ¢1 . " " ¢k has the same orientation as al l ' . he deduces geometrically the COlTcsponding conditions on the coordinates of Vi. . ) be a Euclidean metric for V*. ' " ¢k is det(aij ). show that the signed kdimensional volume of the parallelepiped spanned by ¢1 . then we have a geometl"ic condition for equality with WI /\ .348 Chapw 9 PROBLEMS V I. and ( . Vj ) = ±8ij. Wi. Conclude that W = W". Cartan uses this condition to define n k (V*) as formal sums of equivalence classes of k vectors. or (b) Show that F by q(v) = h (v. Vn for V is called orthonormal if ( Vi. .uj . = = q(u + v) .q(u ) . Then show that q(v) = h ( v.q (v + w) Conclude that '1(0) =H = q( u + v) . E. so that we have a wedge product VI /\ . • . ab and . . . and q(2u) 4q(u). show that this volume is the same for 1/11 . /\ ¢k ¢ O. " " ¢n is a basis q= i. • /\ Vk ofvectors Vi E V. . (The sign is + if ¢1 . /\ 1/Ik # 0. (a) Show that W E W if and only if W /\ ¢1 /\ .q (u) 0. ¢ (b) Let U1 . . v) (c) Suppose q : V + F satisfies q( . . . Ifwe identify V with V**. 3. . . • /\ ¢k = 1/11 /\ . 1/Ii E V* satisfy ¢1 /\ . v) q(v) is bilinear.q( w ). = . .q ( u ) . (d) Conversely.v) = q(v) h ( u .othe:rvvise. v). = q(u+v) . + (a) Define q : V for V*.
} is a basis for W. . ) I W is negative definite. and let . Thus. then = . (Use Problem 716 . 5. x . Wi ) . (d) V has an orthonormal basis. ) be a (possibly nonpositive definite) inner product on V. and ( .) det( ¢i. x . (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)*.. (b) Show that lv. define inner products on 0k V and A k V by using the isomorphism V + V* given by the inner product on V. thus showing that r is unique ("Sylvester's Law of Inertia"). I = . (a) Show that (v. ) i s also nondegenerate on WL . Recall the definition of v. . ). . v) # O. /\ . (e) The index of ( . with (a) Show that ( . 1/Ji}'). (b) For W C V.dim W. where gij = (Vi . is defined on page 301). S how that the index is n . . there is an isomorphism f: IRn + V with f*( . . Prove that dim WL 2: n . (c) I f ( . . . .Riemannian 1\1etries 349 (a) If V # {OJ. Show that these inner products agree with the ones defined above. X Vn _ ' . .r. for some r (the inner product ( .  . . ) k is independent of the basis v" (b) Show that (¢' /\ . v" 4. Let ( . ) = ( . ). Hint: If {w. . /\ 1/Id (c) I f ( . . Hint: Apply the result on page 308 to a certain (II I )dimensional subspace of IRn . ) i s nondegenerate o n W.Jdet(gij ). Define an inner product . /\ ¢k o 1/1. w) = 0 for all W E W } . let WL = {v E V : (v. then V = W Ell WL . Vj ) . x V. 1/Ij )* } = det( ¢i. x . V i ) = O. • . consider the linear functionals Ai : V + IR defined by Ai(v) = (v. in Problem 726. ) is the largest dimension of a subspace W C V such that ( . . then there is a vector v E V with (v. ) has index i . . Vk. ) k on n k (V) by requiring that 1 � i1 < . X Vn _. V n be an orthonormal basis (see Problem 3). . < ik � n be an orthonormal basis._. .
and the functions gij . g C)."".j axi axj  L.350 Chapter 9 6. ..j x '  8. is the classical way of defining the tensor [having the compo nents] 10. ) be a Riemannian metric on M. An indefinite metric on � is a continuous choice of a nonpositive definite inner product ( . This problem requires a little knowledge of simpleconnectedness and cov ering spaces. Show by a computation that if the functions .) (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 '. Show that the index of ( ." with det(gij) # 0. (a) There is no way of continuously choosing a I dimensional subspace of S 2p. So If the expression for A in a coordinate system is . i}�� ' i. for each p E S 2 (Consider the space consisting of the two unit vectors in each subspace. ) be two Riemannian metrics on a vector bundle � 1[ : E B. a . Let � '= 1[ : E + B be a vector bundle. 7. ) and ( . Let S be the set of e E E with (e. and define S' similarly. Let ( . Define a tensor B of type @ by gil a '· a . and A a tensor of type that A (p) : Mp + Mp .1 (p). )p on each 1[ . If � is a smooth bundle over a manifold M. show that S is diffeomorphic to S'.� 'r:t� _ . (a) Let ( .e) 1. + = ' X /J then g This. of course.. g ij are defined by 9. '. )p is constant on each component of B . Show that S is homeomorphic to S'..
Ai gjk · j=1 (b) Similarly. i=1 The tensors B and C are said to be obtained from A by "raising and lowering indices" .I)dimensional submanifold with orientation The outward unit normal v (p) at p M is defined to be that vector in IRnp oflength I such that v (p). . . then n ckj 'L.cs show that = 351 B LI. find Yh . then V (p) dVn_l(p)( VI)p. . (Vn I)p is positively oriented in IRnp when (VI)p. (a) If I: [a. so that we obtain n everywhere orthonor mal vector fields Yh . Vnl = = E /1. (b) Compute the area of S2 13. . (b) Let d VnI be the volume element of M determined by the Riemannian metric it acquires a submanifold of IRn . . (VI)p. .. Yn .bj + IR is positive. where n Blk = 'L. Yn with (YI. S how that if we consider v(p) as an element of IRn .Riemannian 111etr. (a) If M = aN for an ndimensional manifoldwithboundary N C IRn . . . . Y ) j ±81j in a neighborhood of any point. ). . (V. as () . .k Blk dxl ® dxk . define a tensor C of type (�) by C (p)(AI . Show that the GramSchmidt process can be applied to the vector fields all at once. (b) For the case of a nonpositive definite metric. show that the area of the surface obtained by revolving the graph of 1 around the xaxis is t 21f1JI + (/')2 .A2)' Show that if C has components C kJ .gkI A{ . . (vnIlp is positively oriented in Mp. (a) Let XI. .. 12. Let M C IRn be an (n . then v(p) is outward pointing in the sense of Chapter 8. _I)p) = det �I . . 1 1 . Xn be linearly independent vector fields on a manifold M with a Riemannian metric ( . A2) = (A(p) * (AIl.. .. . . .
/\ d. /\(fXi(J. /\ dxn (p). . (d) Let M c IRn be a compact ndimensional manifoldwithboundary. . X Ci Ci E X . x . x vnd. . Let T be the vector field on aM consisting of positively oriented unit vectors. v ) dA = [ (X.x2)(nI )/2 Vn_ 1 dx. and that of aM by ds. (VI VnI> v(p)) = V. Conclude that vi (p) . (c) Note that VI x .. . ' . /\dxn (p).. /\ (fxi(J. Prove the Divergence Theorem: [ div X dVn = [ (X. Let X be a vector field on M. Prove (the original) W E (w . X (w. IRn . Show that Vn = /'I (1 .l )iI ai dx' /\ . dVn_1 (p) = restriction to Mp of (l)i'dx' (p) /\ .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(.) /\ .  .) /\ .. (a) Let v" be the volume of the unit ball in (V [ Cv x X. C Slfikes' Theorem: J1. i=1 (e) Let M 1R3 be a compact 2dimensional manifoldwithboundary. Denote the volume element of M by dA. . v) dVn1 JM JaM (the function div X is defined in Problem 727). Let X = Li a i a/ax i be a vector field on M. Show that for IRn we have n v(p)) . with v the outward unit normal on aM.. . .i /\ . /\ dxn . T) ds JaM JM x X is defined in Problem 727). with odentation and outward unit normal v. Denote the volume element of M by dVn . and that of aM by dVn _ l . . .•. V _1 = v(p) for some IR (by Problem 5). W n w 14.
I be the (11 . where Riemannian metric ( . Show that I l ' rnI An1 dr = AnI . c: [0. f(I» . (a) Let c: [0. l' jl + JR2 given by C(I) = (l. this can be written (d) Let A n. show that this + [/'(1)]2 dl . n n even ' 2. with X(p) = pp .In2 . is the least upper bound of the lengths of inscribed polygonal curves.. V2 = IT. but reversing the order of integration. In = . n JRn has the usual L(c) = (b) For the special case length. I] l' i=1 L [(ci )'(I)j2 dl. show that Vn = (In terms of the r function.) Vn = (e) Obtain this same result by applying the Divergence 111eorem (Problem 13).x2)(nI)/2 dx = . Hint: If the inscribed polygonal curve is determined by the points (Ii .(nI)/2 I · 3·5" 11 n odd. I] + JRn be a differentiable curve. 15.I )volume of S nI . ) = Li dx i ® dx i . (1 . show that I (:��:! o 2 (n +' )/2. c(lll) for .Riemannian Metlies 353 (b) If In = ['. show that nI (c) Using V. 1fnf2 r(l + n/2) . Using the method of proof in Corollary 88.
(c) Prove the same result in the general case. we maintain the hexagonal arrangement. CmmaK. 1]. Top view To in(T("asc the number of triangles. written by someone called M.354 a partition Chapter 9 1 [0. 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. Schwarz first observed.li_I l2 + (J(lt) . In this way. on a compact set. we can increase the number of triangles indefinitely.planes of the hexagons closer together. then we have IC(li ) . similarly.li_I l 2 for some �i E [Ii _ I .J(ti_Il)2 J(li .u A1fClJlUJ 'Ua MrIOZOo6jJa3URX. The topic of surface area for nondifferentiable sUlfaces is a complex one. while the area of each approaches hi /2. the least upper bound of the areas of inscribed polygonal surfaces. which we will not go into here. < In = of = lem 81. Ii ] . It is natural to suppose that the area of a surface is. Hillt: Use the results of Prob = 10 0 < . bUl mow the. .li _I l 2 + f'(�i )(li .c(liI ll = J(li . and uniform continuity of "'. but as H. so that the triangles are more nearly in a plane parallel to the bases of the cylinder. . .
I) [a au a a Y this result makes sense even if 1 is only C'.l' �� (1'/(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. and Theorem 7 is false. lb �' (I)[g(l) . .S or does no/ have to be disconnected.B and B . q').B. Show that the metric on M may be defined using Coo. (a) By applying integration by parts to the equation on pages 3 1 8319. q) ? min (p . e= We clearly have so we need to find a suitable � with �'(I) = g(l) . If a continuous function . (Show how to round offcorners of a piecewise Coo path so that the length increases by less than any given £ > 0.S and q E M .) d L (e) = L(e 0 pl. then 1 still satisfies the Euler equations (which are not a priori meaningful if 1 is not C2). by showing that M . 1] is a diffeomorphism. Use q!eS this fact and Lemma 7' to complete the proof of Theorem 7. these details become quite important. instead of piece wise Coo curves.S is disconnected.Riemannian 1I1ettics 355 16. show that 17. d d 19. (In the theory of infinite dimensional manifolds. (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}. Let e: [0. for all Coo functions � on [a.S are disjoint open subsets of M .e.S. then = (1. I) + M be a curve in a manifold M with a Riemannian metric ( .b] satisfies 0. Show that M .) 18.e] dl = 0. (b) If p E B . ). If p : [0. (b) D u Bois Re )'mond� Lemma. f M . remember that the f ormula for length involves only first derivatives. g o n [a. 1] + [0. show that (p. show that (d& (u» I dJ U u=o = a2Ci aF lb t (0. /'(1» dl] dl. (c) Conclude that if the C' function 1 is a critical point of J. /(1). 1 (I» .
for convenience..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 . ) I cosh x = cos ix.. hyperbolic cosine. I respectively. The hyperbolic sine. coshx (a) Graph sinh. 2 (cosh _ ' )'(x) = 21. while cosh.. and tanh are defined by sinhx = . at a and a.denotes the inverse of cosh I [0.' eX + ex 2 sinh x tan h X = .and tanh. 00). (c) For those who know about complex power series: sinh x = sin ix . and hyperbolic tangent functions sinh.356 Chapter 9 20.x) = J 1 + x2 1 v1x R=I r:. We are looking for a function Vx �. situated.. Consider the problem of finding a surface of revolution joining two circles of radius I . Show that sinh(coshI x) I I = cosh(sinh. (b) Show that 2 2 cosh .x) = cosh(tanh. 2 _1 1 . cosh.' eX _ 2 eX cosh x = . and tanh. I I (d) The inverse functions of sinh and tanh are denoted by sinh. cosh.
Examine the sign of 1/ y tanh y f y > O. find the value of Aa there. or (b) Examine the sign of cosh y .I � .Riemannian Metrics of the 357 form where e is supposed to satisfY x /(x) = c cosh c (c > e cosh :!. = I e 0). so JY02 .] . with CI < a/yo < c'}. yo/ cosh Yo 0 a Yo/ cosh yo (e) Using Problem 20 (d). If a < yo/ cosh yo. cosh Yo [YO � 1 . If a = yo/ cosh Yo. namely c = a/yo = I / cosh yo.1 .. Aa (e) = c cosh :!.y sinh y for y > O. It turns out that the surface for C has smaller area. and sketch the graph. (c) Let c > O. (d) There exists c with c cosha/c = I if and only if a :s: yo/ cosh Yo. then there is a unique such c. show that � = JY02 . 2 c. then there are two such c.2. 67 . (a) There is a unique Yo > 0 with tanh Yo = 1/ Yo. c Show that Aa has a minimum at a/yo.
ftr a < )'0/ cosh )'0. y= = cosh :: _ c c :: sinh C :: (' . 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 .[ Volume III. I) does no/ . the graph of f is tangent to q cosh Yo :' y = � x/ . Q E (a. a). so the envelope consists of the straight lines c cosh  x c = c cosh Yo.= ±Yo.x . I) is tangent to the envelope at that point.358 Chapter 9 [These phenomena can be pictured more easily if we use the notion of an envelopec. pp. TIle point Q is called conjugate to P along t he extremal hi ) and it is shown in the calculus of variations that the existence of this conjugate point implies that the portion of /cl from P to (a. . . but the graph of /c. Yo The unique member of the family through (yo / cosh Yo . is tangent to the envelope at points outside [a. '" a Yo/ ('osh Yo the envelope at points P. 176ff.a]. cosh Yo y = ± .
ax'· ax'� ax' . � = __ _ __ _ .) ax.1J + (/') 2 . Chapter 8. Y]'.. show that 24.� l. All of our illustratiOlls of calculus of variations problems involved an F which does not involve t. ture. Q. ay 23.f (b) Apply this to F(x. ag'.j.=1 (   ) . 11 n [ij. k] and [a!'l. with corresponding gil and g 'i) for the expression of a Riemannian metric. ax'� ax' ' )=1 .. i. ax) ox'Y " [}2xl ox'Y " rk. (a) Let x and x' be two coordinate systems. k=1 1=1 Show that any Coo structure on IR is diffeomorphic to the usual Coo struc Y r '. ax'� ax k L. i.}.+ ax'/3 (Jx'l3ax'Y ox.396.k=1 ax ax'Y ax'· ax'� n ax} a2x' ax' a2x) + " g'l' ox'ct . y] ' = so that [ij.dt ay (/(1). F.ct(Jx'Y . (Compare with the discussion of aF d aF (/(1). + L.� ax'Y  x .1 359 give a local minimum for conjugate points of a geodesic in Volume IV. k] ax'. _ f and F: 1R2 + IR we have ( ) = O. so that the Euler equations are actually f f. = . 'l ax'. aF = 0.j.. (Consider the arclength function on a geodesic for Some Riemannian metric on IR. and note the remark on pg. ax'� ax'Y + L ax'Y .) For the corresponding [a!'l..)] 22. .Riemannian 1I1ebic. (c) Also show that i. /' (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 _ .k=1 11 n ax' a2xJ ax' ax} axk L [ij. /' (1» .. Show that � which we eventually obtained in our solution to the problem. y) = .Jl+7 to obtain directly the equation dy/dx = _ � ag'J �� ax} k L.k] are not the components of a tensor. V..
(f) Deduce the result A. ). deduce that so that A� = rD. Ifwe consider ei as a vector field on JR". go about finding I? satisfied by the ei 's.) Bi). How can we . Suppose we are told that there is a diffeo morphism I : JRn + JRn such that Li ) gij dui ® du) = I' ( . since they correspond to lines parallel to the x i axis. and to solve for these we want to find differential equations 25. k]. directly from our equations for a geodesic. e) ) .) = r[.) + Bk).r A k) r=1 Bik . and let L. (c) To solve for I i t is.k = [ij. show that I (a/au i ) = ei . . Cartan uses j this approach to motivate the introduction of the ri� .360 Chapter 9 (a) Let al/au i = ei : JR" + JRn . E.i gij du i ® du J be another metric. sufficient t o solve for the ei . k. In Lcfons sur La Ciomitrie des Espaces de Riemann. • • • ) un again denotes the .J standard coordinate system on JR" . (b) Show that gil = (ei. ) = Li dx i ® dX i be the usual Riemannian metric on JR". j. Show that we must have (d) Show that = = " L g)r A �k + g. where u l . . (Note that the curveS obtained by setting all but one Ii constant are geodesics. Let ( . ill theory a t least.i· (e) By cyclically permuting i .
(Gauss' Lemma is the '�special case" where c is constant. de /dl) 0. the geodesic U r> eXPc(r) u .b) be a (b) Conversely. Show that e = y o p satisfies p : [". (a) Show that ( . ) is an inner product. (b) Given Riemannian metrics on M and N. (c) If e satisfies for Il' JR + JR. v is perpendicular to these hypersurfaces. if e satisfies this equation. (The equation p"(I) = can be solved explicitly: p(t) = J' e M (. we define ( .Riemannian Metlies 361 26. w')' + (v". w")". ) ") are two vector spaces with inner products. n and (V". Show that for v E Mc (r ) with (v. and let diffeomorphism. where v E Mc (r ) with = OJ.) p'(I)/1(I) 28. w' E9 w") = (v'. it follows that there is a natural way to put a Riemannian metric on M x N. then e is a reparameterization of a geodesic. Let c be a curve in persurfaces {exPC (r ) v : Il v ll = M with dc/dl # ° everywhere. ( . ) on V = V' Ell V" by (V' E9 v". 27. If (V'. de/dl) = constant. where M'(s) = /1(5).b) + M be a geodesic. ( . Describe the geodesics on M x N for this metric.) ds. (a) Let y : [a.) . and consider the hy (v. I'l) + [a. then y is a geodesic.
a n ) . Vn of MP l so that we have a �Crectangular" coordinate system X on Mp given by Li a i vi r> (a t . defined in a neighborhood U of p. (a) Prove Schwarz's inequality by imitating the proof of Theorem 1 (2). Show e that y is a curve of minimal 1ength between p and q = y(b).5 E(e). . (Gauss' Lemma still works on exp«('J). . 31. In particular.5 (b .) 30. . sufficiently small pieces of a geodesic mil limize energy. e(b» . . ba ba L� (e) = d(e(a). and suppose that expp is a diff omorphism on a neighborhood ('J C Mp of {ly ' (O) : 0 ::s I ::s I ). . Let y: [a. (c) Let y: [a . among all curves in exp«('J) . b] + M with d(y(a).a)E� (y). . so that each y k is linear. b] + M be a geodesic with y(a) = p. y(b» = L(y) is a geodesic. b) + M be a geodesic with L � (y) = d(y(a). show that [L�(Y)f .t . Let p be a point of a manifold M with a Riemannian metric ( . . If ( . let x be the coordinate system xo exp. unless e is also a geodesic with (a) Show that in this coordinate system we have r6(p) = O. Hint: Recall the equations for a geodesic.. Conclude that E(y) < E(e) L(y)2 L(e) 2 E(y) = .5 . . and note that a geodesic y through p is just exp composed with a straight line through 0 in Mp.362 Chapter 9 29.. ) is a Riemannian metric on M and d : M x M + IR is the corre sponding metric. . 32. y(b» . Schwarz's inequality for continuous functions states that with equality if and only if f and g are linearly dependent (over IR). If era) = y(a) and e(b) = y(b). ) . (b) For any curve y show that with equality if and only if y is parameterized proportionally to arclength. then a curve y : [a. Choose a basis VI .
. and this geodesic lies completely in U. > O.ball is convex.p B.q). So that r o y Lk (yk)2. y y o = tangent to exp(S.). {v E Mp : IIvll :s: 'I and S. and let be the unique geodesic of length < joining them. '. dl . 8).J d k d I is sufficiently small. Show that the following is true for all sufficiently small ' > 0: if is a geodesic such that E exp(S.) for 0 # I E (8. then there is Ii > 0 (depending on such that >t exp(B.k iL. then 0 (e) Let q and '1' be two points with r (q) .) and such that y'(O) is tangent to exp(S. 0 so that (d) Let B. conclude that if Il y '(O) I (b) Let r: U + IR be = 'i I :s: is strictly increasing in a neighborhood of O. the maximum of occurs at either q or q ' . (f) A set U C M is geodesically convex if every pair q.. i. dl dl k (c) Note that l L ddyi dyl n2 L( dylk )2. = {v E Mp II v ll = d..). Show that for sufficiently small .J. d(r y)/dl o. r(q') < . roy 2.. Show that for sufficiently small " the image of the open . dl 2 L. Hint: If ' ( ) is y(O) d(r y)2/dl = y) y(l) s. then Using part (a).Riemannian 111etricJ 363 = r(q) d(p. Show that d2 ( r o d) = 2 [ .. q ' E U has a unique geodesic of minimum length between them. y . u (O) . Show that exp ( { v E Mp : IIvll < d) is geodesically convex for sufficiently small . (g) Let f: U + IRn be a diffeomorphism of a neighborhood U of 0 E IRn into IRn. ( dyk )2 _ " Y2rk dyi dyJ ] .
.. c (h» L (c I [0. < for all tangent vectors (c) Conclude that I d(u��(t» I [ h = = lu i . then = a. Show that if is some K > 0 such that for all near 0 we have I I u'(t) I 1 5. 1. I C(I) V(I) if and only if approaches a c is C'. h]) .:.O) to (1. then there = q Hint: In Mq we clearly have Since expq is locally a diffeomorphism there are 0 < K. h�o Ic(h) .c (O)1 1.'=__"::::__"10 _ u'(t)2 + I '!. If c is C'.O) is ! (b) Consider the situation in Corollary except that > 0 for near o.O) to (!. h] # lim c(t) (I. (a) There is an everywhere differentiable curve that length of c 1 [0 ... 1 d�..0) is I length from ����+ (�. length from (!. f(I» in ]R2 such = Hint: Make c look something like the following picture.364 Chapter 9 33.I) 1 2 dl a u(h) q.t) I · K2 such that V at points near lim h�o lim h�o d(p. and suppose limit as + 0 (even though v(O) is undefined).ICi (U (I).
show that L(c) is the least upper bound of inscribed piecewise geodesic curves.Riemannian Metrics 365 (d) If c is C'.I Y II 2 II X II 2 + II Y II 2 2(X. let . Let f : M + M be a map of M onto itself which preserves the metric d. Y E Mp. then L(yv. 36. if Yv. ric space structures determined on M and N by their respective Riemannian metrics. show that if c is the straight line joining v. and that /. (a) If y is a geodesic. lim v. (a) Using the methods of Problem 33. ) and corresponding metric d. exp l y)] = hm II IX II · II I Y II II X II · II Y II _ HO I1=0 . Let M be a manifold with Riemannian metric ( .w is the unique geodesic joining exp(v) and exp(w). = l . exp w) .w�O L (cv. v.w ) (c) Conclude that d(exp v. (c) Given X. W E MP l then L (exp o c) = l lim . uSe Problem 34 to show that II IX . and Yv.w = exp 0 cv.w.(cX) = c/'(X). 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. (b) Define /' : Mp + M f(p) as follows: For y a geodesic with y(O) /. then f 0 y is a geodesic. Show that f is an isometry of the met 35. = p. o Let f : M + N be an isometry.(y'(O» = d/(y(I» dl Show that II /'(X) II = II X II . l�� II v _ w ll v. 34. w ) = 1.w�O L(c) (b) Similarly.
then Hint: (c) Let y : [0. II v ll · cos Ii where Ii is the angle between u and v. w E f'(X) + f'(Y). ) to y(to + u). Show that if II is ti'O hm .) y(O) close enough to 10. 1 ). Choose II < 10 and consider the variation Ci for which & (u) is obtained by following y up to I" then the unique geodesic from y(t. and finally the rest of y. critical .lIl w ll t # O. and then that f'(X + Y) = IRn with w # 0. (/'(X). w ) II v ll . ). TIlus. Use this to show that f is itself a diff omorphism.366 Conclude that Chapter 9 (X. (b) Conclude that if w is linearly independent of v. /'(Y» f(p) . one can use the equation (u. lI v + l w ll . 1] + M be a piecewise C' critical point for length. and suppose that y'(lo+ ) # y'(to) for some 10 E (0. v) = lI u l i . e 37. then the limit is Dv (v)(w). If v : IRn + IR is the norm. # 0. y(l) y(t. The same result then holds in allY vector space with a Euclidean metric ( . then dL(&(u» /dulu�o paths for length cannot have kinks. a contradiction. and hence an isometlY.lIv ll .lI v ll I = (v. Y)p = (d) Part (c) shows that 1' : Mp + Mf(p ) is a diffeomorphism. (a) For v. show that hO lim II v + l w ll . Alternately.
TIle Poincare upper half plane with the Riemannian metric Jf' is the manif old {(x. (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. 41. = r"2 = rl. 39. Let g : S" + pn be the map g(p) = [p] = {p. TIlese results show that pH provides a model for "elliptical" nonEuclidean geometry. Q» Show that evcry geodesic y : jR + pn is closed (that is. and let L be a generating line. Y2 (a) Compute that : y > O} ! ri. and that every two geodesiCS intersect exactly or once.�.. = . Investigate the geodesics on a cone (the number of geodesics between two points depends on the angle of the cone. y) E jR' ( . y all other ri� = O. r�. p }. Consider a cone C (without the vertex). = . Consider a cylinder Z C jR3 of radius t. Unfolding C . there is a number a such that y(t + a) = y(l) f aU I). 40.L onto jR 2 produces a map f: C .\' 367 38. and some geodesics may come back to their initial point). )) on pn such that g * « . . Find the metric d induced by the Riemannian metric it acquires as a subset of ffi. (a) Show that there i s a unique Riemannian metric « .L + jR2 which is a local � P(� isometl)� but which is usually not oneone.Ri£1nannian 1\1etfie. ) = dx ® dx + dy ® dy .3.y. The sum of the angles in any triangle is > 1f. )) is the usual Riemannian metric on sn (thc one that makes the inclusion of Sn into jRn+' an isometry).
b .. . A. c) and radius R.: J 3' YN. show that all the geodesics in Jf2 are the (suitably pa rameterized) semicircles with center on the xaxis. .. (d) Show that these geodesics have infinite length in either direction. c. Considering (b) Let C be a semicircle in Jf2 with center at it as a curve t 1+ (t.368 Chapter 9 (0. upper half plane is a model for Lobachevskian nonEuclidean geometry. together with the straight lines parallel to the yaxis. then length B C = length B'C' and the angles at B and B' and at C and C ' are equal ("sideangleside"). show that y'(t) 2 y(t) (c) Using Problem 27.. ad . (f) For those who know a little about conformal mapping (compare with Prob lem IV 76). (e) Show that if y is a geodesic and p ¢ y. so that the upper halfplane is complete. and that we can take any tangent vector at one point to any tangent vector at any other point by some J.bc > O are isometries. d E 'llI. These results show that the Poincare = = A ti c : c . y(t)). The sum of the angles in any triangle is < 7[. Consider the upper halfplane as a subset of the complex num bers <C. Show that the maps z J( ) = az + b cz + d a. 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'. then there are infinitely many geodesics through p which do not intersect y.
46. Show that the Riemannian manifold M is geodesically complete. (a) Show that the normal bundles v" V2 of M n C N n +k defined for two diff erent Riemannian metrics are equivalent. show that the normal bundle of M C E is equivalent to �.00) > M with the initial value y(O) = p. )) is complete if and only if (N. )) is complete. If ( . Let is indeed a kplane bundle. Prove that there is a geodesic y : [0. and (M. 47. show that v M and N be geodesically complete Riemannian manifolds. ( . Show that the Rie mannian manifold M x N is also complete.Riemanni(J}l Metric. This problem presupposes knowledge of covering spaces. ) is a Riemannian metric on N. Let M be a manifold with a Riemannian metric ( . 48. (b) If � = ]f : E > M is a smooth bundle. conclude that TE '" ]f*(�) Ell ]f*(TM) . where N is a Coo manifold. show that E2 ". more generally. (a) Given an exact sequence of bundle maps as in Problem 328. then g'( . Ell E3 .. Let p be a point in a complete noncompact Riemannian manifold M. ) is a Riemannian metric on M. 369 42. Then there is a unique Coo structure on M which makes g an immersion.g'( . (b) Using the notion of Whitney sum Ell introduced in Problem 352. Does it necessarily follow that the Riemannian manifold M is complete? 43. over a paracompact space]. b 1 > M with initial value y(a) = p can be extended to all of JR. Let g : M > N be a covering space. (a) If M n C N n +k is a submanifold of N. 45. having the property that y is a minimal geodesic between any two of its points. show that the normal bundle v M x Ell TM '" (TN)IM. Suppose that every geodesic y : [a. . 49. and give N the Riemannian metric described in Problem 26. E. where the bundles are over a smooth manifold M [or. (b) If � = ]f : E > M is a smooth kplane bundle over M". 44. Let M be a Riemannian manifold such that every two points of M can be joined by a unique geodesic of minimal length. ). and choose a fixed point P E M.
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. any nonorientable manifold contains a "fairly small" nonorientable open submanifold.370 Chapter 9 50. (a) Let M be a nonorient able manifold. it is not hard to show that a bundle over S' is trivial if and only if it is orientable). show that the normal bundle v of S' C M is not orientable.) . Using Problem 47. in fact. (Thus. (b) Use Problem 329 to conclude that there is a neighborhood of some S' C M which is not orientable. but the same conclusions will hold if each (TM) IS' is orientable.
consequently the map (x. b) a 1+ 1+ T preceding chapters. cos2Jfx cos2JfY + sin 2JfX sin 2JfY sin 2Jfx cos2JfY . y) 1+ xy is Coo. y) (x. Thus the map (x. y) 1+ xy.Y 1+ xy.cos2JfX sin 2Jfy. many of the results and concepts of the It will also play an important role in later Volumes. It clearly suffices to aSSume instead that the single map (a. with the operation + . y) is Coo. y) 1+ X . because in the study of these problems the groups of autom orph isms of various structures play a central role. y) 1+ xy' is Coo.y) = Coo . ab aI from G x G to G from G to G are continuous. b) 1+ ab' is continuous. where we are concerned 'V'lith geometric problems. and these groups can be studied by the methods now at our disposal. The circle S' is also a Lie group. A topological group is a space G which also has a group structure (the product of a. it even suffices to assume that the map (x.y) sin2Jf(x . 371 . One way to put a group structure on S' is to consider it as the quotient group JR/Z.CHAPTER 1 0 LIE GROUPS his chapter uses. As a matter offact (Problem I). A Lie group is a group G which is also a manifold with a Coo structure such that (x. It clearly suffices to assume that the map (x.' from S' x S' to S' is also = 1+ 1+ cos2Jf(x .I In In In JR x JR > JR > S' = JR/Z. which can be expressed in coordinates as one of the two maps (x. The simplest example of a Lie group is JRn . The functions x 1+ cos 2n: X and x 1+ sin 2n: x are Coo functions on JR/Z. where Z C JR denotes the subgroup of integers. b E G being denoted by ab) such that the maps (a. and at each point at least one of them is a coordinate system. We will mainly be interested in a very special kind of topological group.I are Coo functions. y) 1+ xy X 1+ X . and illuminates.
considered as a 2 subset of IRn Since the function det : IRn' > IR is continuous (it is a polynomial map). Later in the chapter we will havc another way of proving that 0(11) is a Lie group.' which is Coo on GL(II. IR) = det. is norm presen1ing. Multiplication of matrices is Coo. of dimension 11(11 .b') represent the same element of S' x S' if and only if a ' . IR). generally. the set GL(II. cOJ"isting of all A E GL(n. since the entries of A B are polynomials in the entries of A and B. y) 1+ xy . 1R). In particular. This condition is equivalent to the condition that the rOws [and columns] of A are orthonormal. 1R) with A . where A ij i s t h e matrix obtained from A by deleting row i and column j. f S' C 1R 2 can be considered as the group or . 1R)). A t = I. To show that 0(11) is a Lie group we must show that the map (x.b E Z. which is equivalent to the condition that..' )F = det A ij / det A. 1R) is the group of all nonsingular real II x II matrices. (This gives another proof that S' is a Lie group.{OJ) is open. The argument in the previous paragraph shows. where A t is the transpose of A. that if H e G is a subgroup of G and also a submanifold of G.372 C7wjJler 10 If G and H are Lie groups. By Proposition 21 I . Problem 233 presents a proof that 0(11) is a closed submanifold of GL(II.a E Z and b' . with the product Coo structure.e. The general linear group GL(II. a manifold. Many imporlant Lie groups are matrix groups. then G x H. and in particular. and thus inner product presen1ing. The torus may also be described as the quotient group IR x 1R/(Z x Z). and hence can be given the Coo structure which makes it an open submanifold of IRn' . Smoothness of the inverse map follows similarly fi'om Cramer's Rule: IIIIIIIII!! (A. it suffices to show that it is continuous. is easily seen to be a Lie group. is also Coo as a map from 0(11) x 0(11) to 0(11).1)/2. the matrix A represents a linear transformation which is an "isometry'" i. the torus S' x S' is a Lie group. 1R) is a homeomorphism (since O(n) is a submanifold of GL(II. and the direct product group structure. then H is a Lie group. with respect to the usual basis of jRn. One of the most important examples of a Lie group is the orthogonal group 0(11). b) and (a'. the pairs (a.' (1R . but this is true because the inclusion of 0(11) > GL(n .
Similarly. PROOF If a E K. It is know that these are the only spheres which admit a Lie group structure. we have K.Lie Groups 373 of complex numbers of norm I. a subgroup which is an (imbedded) submanif old is always a Lie subgl"Oup. since b 1+ a . respectively. that a subgroup which is an immersed submanifold is always a Lie subgroup.1 K C K. As we have seen. Since e E bK bI . after some work (Problem 18). we mention £(11). then aI K is connected. if L C lR x lR is a subgroup consisting of all (x. PROPOSITION. which proves that K is a subgroup. the component containing the identity I. then K is an open Lie subgroup. This is not accidental. If G is a topological group. is a subgroup. For any b E G. ex) for e irrational. the two components consist of all A E 0(11) with det A = +1 and det A = . 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 } . and also a Lie group for some Coo structure which makes the inclusion map i : H > G an I l lnl l I . K is dosed since components are always closed. •:. Since this is true for all a E K. we have aI K C K.I b is a homeomor phism of K to itself. which we have already seen is a Lie group. The group SO(2) is just S l . For example. As a final example of a Lie group. but we will not need this fact. then K is also open. we have bKb1 C K. so K is normal. The group 0(11) is disconnected. the group of all Euclidean .) 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. Hence K is a Lie subgroup. so K is a submanifold and a subgroup of G. It even turns out. I\1oreover.I . If G is a Lie group. image of L in S l x S l = lR x lR/(Z x Z) is a dense subgroup. S 3 is the Lie group of quaternions of norm I. Since e = aIa E aI K. it follows similarly that bK b I is connected. since G is locally connected. We define a Lie subgroup H of G to be a subset H of G which is a subgroup of G. then the immersion. If G is a Lie group.
Consequently. A little argument shows (Problem 5) that every element of E(I1) can be written uniquely as A . Notice that La and Ra are both diff eomorphisms. respectively. For any Lie group G. the maps La* : Gb Ra * : Gb 7 Gab 7 Gba for all are isomorphisms.l = A Ta_bB. i. which shows that E(I1) is a Lie group. A Ta(BTb). Now E(I1) is not the direct product 0(11) x IRn as a group. Clearly the component of e E E(I1) is the subgroup of all A T with A E SO(I1).1 = A TaTbB. isometries of IR". La : G > G and Ra : G > G. T where A C 0(11). Consequently. A vector field X on G is called left invariant if La.1 (x) = A(A. Vie can give E(I1) the C"" structure which makes it diff eomorphic to 0(11) x IR" . G. since translations and orthogonal transformations do not generally commute.l . COJlsequently. and T is a translation: T(X) = Ta (X ) = X + a .1 TB(a_b). if a E G we define the left and right translations.1 X + a) = x + A(a) = TA(a) (X) . . b It is easy to see that this is t�ue if we merely hav(' for all a E II E G.X = X Recall this means that for all a . with inverses La .e. so A TaA.374 Chapter 10 motions. given Xe E Geo there is a unique left invariant vector field X On G which has the value Xe at e . E G.. In fact.1 = A B. by La(b) = ab Ra(b) = ba .l and Ra.
PROPOSITION. x(e)). It suffices to prove that X is Coo in a neighborhood of e. X" be the left invari ant vector fields v.. This implies that X is I e where Xe = j =l L cj D X a I e which shows that XXi is Coo .I 375 PROOF. Let (x. PROOF. Coo . . c ' . Choose a basis X' e . Let X" . . c ) . " " X"e for Ge.rith these values at e. Since the map (a. . Then for a E V we have XXi (a) = 2. Every left invariant vector field X on a Lie group G is COO. . ' . so we can define an equivalence G x lR" J (t J=I ciX (a) i ) = " (a. . COROLLARY. • : 3. Choose a neighbor hood V of e so that a. •:. ./f. A Lie group G always has a trivial tangent bundle (and is consequently orientable). Then XI .Lie Grau/}. U) be a coordinate system around e. . Cj a(xi 0 La) � ax} 0 La) j= 1 L c j D"+j Ji (x(a). La.X around a (Prob lem 51). b E V implies ab ' E U. 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). Xn are clearly eveI)"Alhere J : TG > linearly independent.Xe(xi) 0 = Xe (xi La) . XXi (a) = Then " Xe (xi = j= 1 = . . . since the diffeomorphism La then takes X to the C"" vector field La. .
X] = o [[X. ] satisfying [X. ] on the direct sum V = V. Henceforth we will use X. are left invariant and [a/ax . IR) at the identity I can be . etc. since the vector fields a/ax ..376 Chapter 10 A left invariant vector field X is just one that is Larelated to itself for all a . . we write or gl (ll. or Since the [ . to denote elements of Ge. Y](e). This notation requires o some conventions f particular groups. If If. YJ '= 0 f all X. Ell V2 (= V. are Lie algebras with bracket operations [ . YJ = [Y. the tangent space of GL(n . Y. . Clearly £(IRn) is abelian. X S') is also abelian. YJ. 1R) f the Lie algebra of GL(Il. Z] + [[Y. X] + [[Z . The Lie algebra £(S') of s' is I dimensional. X]. Consequently. x V as 2 a set) by It is easy to check that this makes V into a Lie algebra. Y(e) = Y. to denote the left invariant vector fields with X(e) = X. 1R) or f the Lie algebra of 0(11).) We will also use the more customary notation 9 (a German Fraktur g) f r £(G). and that £(G x H) is isomorphic to £(G) x £(H) with this bracket operation. Since GL(n. with a bilinear operation [ . and will be denoted by £(G). Consequently. YJ i s left invariant i f X and Y are. the Lie algebra £(S' x . (Sometimes the Lie algebra of G is defined instead to be the set of left invariant vector fields. Xl Consequently. The structure of gf (11. ] i f i = 1 . or The Lie algebra of IRn is isomorphic as a vector space to IRn. a/axj ] = O. YJ = [X . . then or we can define an operation [ . Proposition 63 shows that [ X . it is also skewsymmetric.. 2. we call a Lie algebra abelian or commutative if [X. ] operation is assumed alternating. ] operation. etc. or 0(11) In general. 1R) is more complicated. Z E V. etc. The vector space Ge . We can then define an operation [ . YJ = 0 'jacobi identity" f all X. Y. and X. 1R) is an open submanifold of IRn ' . is called the Lie algebra of G. a Lie algebra is a finite dimensional vector space V. [X. and consequently must be abelian. ] on Ge by [X. Y. Y. Z] . together with this [ .
lith jRn 2 .Lie Groups 377 identified ". So = L Mi/ A ki = L M.l .(Mx kl ) . 0'=1 n with (constant) partial derivatives � (xkl ax" 0 LA) = A ki {0 j =I J # I.j = L Nij . NlI = � (MN . 1R) as f ollows. M/ (xkl ) = M/ (xkl LA) . a Xl} a = L Nkj Mjl = (NM lkl· j=l From this we see that " . For every A E GL(n. ax l / a k.[M.=) n 0'=1 n •. k=i k # i.j I = MA (xkl ) = LA.NM lkl rJ .IA k Thus.. 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. If \ve use the standard cOOldinates xij on jRn. Now the function xkl Mxkl (A) = L Mij :ij a i.B.2 . 0 1 (x kl 0 LA : GL(n . 1R) i s the linear function LA) (B) 0 = xkl (AB) = L A k .I.kl ) matrix. we have i. So if N is another 11 x a __ .. 1R) > GL(n. N/ (Mx. 1R) corresponding to M. Wc compute the function M xk l on GL(n.kl _ Mjl Mx {0 ax ij 11 . 1R).
then k�i differentiating give. IR)/. ] satisfies the Jacobi identi�: Since 0(11) is a submanifold of GL(I1. For each i. then [ . Mi n 0 ) This subspace has dimension 11(11 . which is exactly the dimension of 0(11). . :2 Thus 0(11)1 C jRn can contain only matrices M= ( M which are skewsymmetric. IR) with IRn' . This subspace may be determined as follows.8) > 0(11) is a curve with A(O) = I. if we define [a. if we identify Chapter 10 gl(I1. If we did not know the dimension of 0(11). the bracket operation is just [M.sin ! sin ! cos t J (rotation in the (i. we can define a cun'e A : IR > 0(11) by A (') � ( j ' cos t . 0 M1 2 M1 3 . which shows that L A ik (t)Ajk (t) = �ih A i/ (O) = . IR) we can consider 0 (11) / as a subspace ' of GL(I1. Notice that i n any ring.Aj/(O) . j with i < j.b] = ab .ba .NM. Mi n M1 2 0 0 M1 3 . we could use the following line of reasoning. If A : (8. so 0(111I must consist exactly of skewsymmetric matrices. N] = MN .378 thus. j)plane) j .1 )/2. and thus identify 0(11) with a certain subspace of IRn . and we denote (A(t))ij by Aij (t). .
j). j) and (j. if Y E and are irelated. i. X(a) X = i. it follows that it is Coo as a map into H. that is.a be the subspace of Ga consisting of all for X E I). •: a X(a) e. Now this map is clearly Coo as a map into G. f] X [ i.. Then the set of all A'(O) span the skewsymmetric ma trices. X La o f i 0 La. THEOREM. It follows immediately that Lb permutes the various maximal integral manifolds of /j. takes H to the maximal integral manifold containing Lb. Let H be the maximal integral manifold of /j. = = La. In particular. and are irelated. He is a subspace of � which is closed under the . He. 1 operation. (H) = H. leaves the distribution /j. X) = X(a). This correspondence between Lie subgroups of G and subalgebras of � turns out to work in the other direction also. Y](e)). which means that [X. ([X. (a.ba. Y](e) Thus. Let PROOF. and O(n) must have dimension n(n . Hence O(n)/ must consist exactly of skewsymmetric matrices.1 )/2. The fact that I) is a subalgebra of � implies that /j. let /j. we can identify He with a subspace of Ge. Since i* : He Ge is an isomorphism into. Any X E He can be extended to a left invariant vector field on H and a left invariant vector field on G. Using Theorem 67.1 . To prove that it is a Lie subgroup we just need to show that 1+ is Coo . b) ab.sin t at (i. is an integrable distribution.Lie Graul" 379 with sin t and . Yl [X. [ G be a Lie group. (i. and let i : H > G be the inclusion. I 's on the diagonal except at (i. then [X. This implies that H is a subgroup of G. if E H. For each E H C G. then clearly Lb. = so Lb. The proof of uniqueness is left to the reader. We do not need any new calculations to determine the bracket operation in o(n). In fact. containing If b E G. among themselves. La. invariant. and I) a subalgebra of �. He with this induced . = La : G > G In. He C Ge = � is a subalgebra of �. Then there is a unique connected Lie subgroup H of G whose Lie algebra is I). other words. we have left translations 7 X X a La : H and So > H. For E G. moreover. so Lb . b (b) e.(/j..a) = /j. and O's elsewhere. Consequently. then Lb. 4. i). i) and (j. consider a Lie subgroup H of any Lie group G. 1 operation is juSt I) = £(H).
and % value 4>*eX at e. and every linear transformation is a group homomorphism. For any a E G we clearly have 4> 0 La = L¢(aj 0 4>. Consequently.4>.. It then follows from Theorem 4 that ever .. Usually. X = %(4)(a)).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 example. we will denote 4>. simply by 4>. the map algebra homomorphism.. so if X E Ge. Later ) on we will be able to obtain a "local" version of this result.. X + b4>. Y]. 4>.aLa. Thus X and % are 4>related. We will soon see to what extent the Lie algebra of G determines G.P. 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). we next consider. If 4> : G > H is a Coo homomorphism. then the condition 4>(s + I) = 4>(s) + 4>(1) implies that d4>(1 + s) ds = 0 gives d4>(s) .. that is. . ds which means that 4>(1) = cl for some c (= 4>'(0)). 4>. Yj = [4>. more generally. Having considered subgroups of Lie groups (and subalgebras of their Lie algebras). homomorphisms between Lie groups....X = L¢(aj. suppose that G = H = JR. then . then 4>. : Ge > He .x is the left invariant vector field on H with 4>*a%(a) = 4>. (aX + bY) = a4>. : � > Ij is a Lie 4>. [X. There are an enormous number of homomorphisms 4> : IR > 1R. because IR is a vector space of uncountable dimension over Q. But if 4> is Coo. Y 4>. X. 1R) for some N. Lie algebra is isomorphic to Ihe lie algebra ifsome Lie group... : � > Ij. We continue the study of Lie groups along the same route used in the study of groups.
. : IR > IR is just multiplication by c. for X E g. By Theorem 4. P ROOF.(X. If H. Notice that the only continuous homomorphism 4> : S' > IR is the 0 map (since {OJ is the only compact subgroup of 1R). 5. THEOREM. The continuous homomorphisms 4> : IR > S' are clearly of the form once again. Vr: G > H with 4>. We can identify £(IR) with 1R . A neighborhood of the identity e E S' can be identified with a neighborhood of 0 E 1R. but H = S' = IR/Z. : IR > IR is multiplication by c. 7 w. Let G and H be Lie groups. giving rise to an identification of £(S') with 1R.Lie Groups 381 rational numbers). we do have a local result.ab E U. and G is connected. Let I (German Fraktur k) be the subset f C g x l) of all (X. there is an open neigh borhood V of (e. If H2 : G x H > H is projection on the second factor. X = <I>(X).b. there is a unique connected Lie subgroup K of G x H whose Lie algebra is f. 4>. I K. 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. and <1> : g > l) a Lie algebra homomorphism. X E g we have 4>. . Clearly the map 4>.e} Ge is an isomorphism. <I>(X)). = <1>. . then 4> = Vr. For X E g we have SO w* : K(e. Now suppose that G = 1R. Consequently. and w = H. : G x H > G is projection on the first factor.e) E K such that w takes V diff eomorphically onto an open neighborhood U of e E G. <I> (X)) = X .. However. Since <I> is a homomorphism. Consequently. 4>. I is a subalgebra of g x l) = £(G x H). then w : K > G is a Coo homomorphism. a Lie algebra ho momorphism g > I) may not come from any Coo homomorphism 4> : G > H. if there are two Coo homomorphisms Vr. we Can define Moreover.
b in a neighborhood of e. since (Problem 4) mH' .X Given 4>. 1jr (a)). •:. COROLLARY.�O . It follows that G is abelian. By Corollary 6. which implies that 1jr (a) = 4>(a) for all a E G . As for the second. PROOF. A connected Lie group G with an abelian Lie algebra is itself abelian . COROLLARY. 7. Thus G' = K. <I> (X)). so ab = ba for a.X•  . 6. If two Lie groups G and H have isomorphic Lie algebras. then w. if X E g. Since 4>*e = <P is an isomorphism. then they are locally isomorphic. •: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. <I> (X)) = <I>(X). = (X. there is a unique d4> dl Coo homomorphism 1 . neighborhood of e generates G . let 4> be the map given by Theo rem 5. 1jr: G > H.382 Chapter 10 TIle first condition on 4> is obvious. 4> is a diffeomorphism in a neighbor hood of e E G . so 4>. and that all connected Lie groups with a given Lie algebra are covered by the same simplyconnected Lie group. 4> : IR > G such that 8. (X. PROOF. Given an isomorphism <1> : A > I). e: G > G x H by e(a) = (a. COROLLARY. For every X E Ge. define the oneone map = ][2. <I> (X)) = X. G is locally isomorphic to IR". x so £(G') = f. (X. The image G' of e is a Lie subgroup of G x H and for X E A we clearly have e. •:.
they can be extended to all of IR using the method of the first proof.Lie Groups F IRST PROOF. which proves uniqueness. We know that integral cun'es of % exist locally. on some neighbor hood (e. 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.I ) and define 4>(1) = { 4>(e/2) · · . < o. . SECON ( D DIRECT) PROOF.( ) du u=o j' = L¢(r»XU) = %(4)(I))U). . 4>(1) 4>(s + I). . By 111eorem 5. III. The same is so 4> is a homomorphism.X.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. Conversely) if 4> : IR > G is an integral cun'e of %./(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.0 dl ' /(4)(1)4>('')) .:. Irl e/2 r �O ddl4> I Isl. 4>(s + 1) = 4> (s)4> (I ) and Clearly <I> is a Lie algebra homomorphism./(4)(1)) " 1. I 1 Thus 4> must be an integral cun'e of X. Uniqueness also follows from Theorem 5.· 4>(e/2) · 4>(r) . If homomorphism. and < < e 4>: IR > G is a Coo d4> ( f) = lim /(4)(1 + ")) . I = O. ls + 1 1 (:!. then 1 1 is an integral curve of % which passes through 4>(s) at time clearly true for 1+ 1+ 4>(s) . e) of 0 E IR there is a map 4> : (e. 6") > G with .
Consequently: IA + BI IA B I :s: :s: I A I + IB I n I A I ' IBI: I . IR) where · now denotes matrix multiplication.+ +. We have already examined the I parameter subgroups vi Iii . where exponentiation of matrices is defined by exp(A) = I+ This follows from the facts in Problem 56. GL(n.nN! (N + K ) ! :s: nk. Ii 2! 3! hence I A lk I A N+K AN (n I A I) N (nI A I )N+K +". Notice that IR .384 Chapter 1 0 A homomorphism 4> : IR > G i s called a Iparameter sub group o f G .. We thus see that there is a unique I parameter subgroup 4> of G with given tangent vertnr d4>/dl(O) E Ge. some of which will be briefly reca pitulated here. 4> (0) = I. with multiplication as the group operation. with 4> (S + I) = 4>(s)4>(I). IR). The solutions of these equations can be written formally in the same way 4> (1 ) = exp(I4> ' (O)).{OJ."+ > 0 as N > 00.' I A lk :s: nk l A l k . . (N + K ) ! :s: y. 4>(1 ). Then all Coo homomorphisms 4> : IR > IR .{OJ..{OJ is just GL( I . More interesting things happen when we take G to be IR . then clearly A A2 A3 + + +. e¢'(O)r. If A = (a'j ) and I A I = max laij I. must satisfy 4>' (1 ) = 4> ' (0)4>(1 ) 4> (0) The solutions of this equation are 4>(1 ) = = I. All Coo homomorphisms 4> : IR > must satisfy the analogous diff erential equation 4>' (1 ) = 4>'(0) .
Given X E �. then Coo homomorphism We clearly have Ge exp 0 1fr. we now define the "exponential map" exp : � > G as follows. 9.. let </> : lR > G be the unique with d</>/dl(O) = X. and convergence is absolute and uniform in any bounded set. exp(11 + 12 ) X = (exp I . then exp(A + B) Hence.I 385 so the series for exp(A) converges (the J) th entry of the partial sums converge). and 0 is a regular point. if A B = BA. then </>' (1) = = (i. if </>(1) is defined by (* *). The map exp : Ge > G is Coo (note that Ge '" lRn has a natural Coo structure). so that exp takes a neighborhood of 0 E Ge diff eomorphically onto a neighborhood of e E G. so </> does satisfy (*) .Lie Grau/}. PROPOSITION. I] exp(I</>'(O)) +. (exp A ) (exp B)..+O lim Ii + h2 </>'(0) 2 h 2! exp(I</>'(O)) = </>'(0)</>(1). h_O lim exp(I</>'(O) + h</>'(O)) .exp(I</>'(O)) h [exp(h</> (O)) 1 = lim h'l>O h</>'(O) = . h... X) (exp 12 X) exp(IX) = (exp IX). If 1fr : G > H is any Coo homomorphism.I . Moreover (see Problem 56). Then exp(X) = </> ( 1 ) . For any Lie group G. = 1fr 0 expo exp G�H l � He l exp .
then also 4>. So exp. then the curve C(I) = IV in Ge has tangent vector v at o. If 4>.a) o f the Coo manifold G e x G a t the point (X. •: 1 1 1 0. Therefore exp is a diffeomorphism in a neighborhood of O. and hence oneone. let 4> : IR > G be a homomorphism with d4> = X. Every oneone Coo homomorphism 4>: G > H is an immersion (so 4>(G) is a Lie subgroup of H). PROOF.o is the identity. I 1=0 = d dt I 1=0 exp(lv) So exp. . We define a vector field Y on Ge x G by Y(X..:. COROLLARY.X) = Vr 0 4>(1) = Vr (exp X) . . contradicting the fact that 4> is oneone. (tX) = 4>(exp(IX)).p( X (p)) = 0 for some nOllzero X E �.a) = 0 Ell X (a). o.X. dl r �O Then Vr 0 4> : IR > H is a homomorphism with d(Vr 0 4» = Vr. If we identify a vector V E (Ge) o with Ge. which we know is exp X = projection on G of a ( l . Then Y has a flow a: IR x (Ge x G) > Ge x G. a) can be identified with Ge Ell Ga. exp(Vr.. The tangent space (Ge x G)(X.386 Chapter 1 0 PROOF. Coo. (X) = But then e = exp 4>. 0 Ell X). Given Vr : G > H. Since it follows that exp is Coo. dl r�O Consequently.o(v) = = d exp(c(I)) dt v. and X E Ge.
. X. The map 1 1+ 4>(exp IXi) is a continuous homomorphism of IR to H. COROLLARY. ( � IJ . Moreover. .. By continuity. Let U be a starshaped open neighborhood of 0 E G. 4> ( expI. For any a E exp( ! U). PROOF Since X/2. PROOF. 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. t U). X) . 4> (m/2n . Y E U it follows that X/2 = Y. (exp In X. Every continuous homomorphism 4> : G > H is Hence Coo. it follows from the above that 4>(e/2) = exp(X /2). Thus. . and thus everywhere. Every continuous homomorphism 4>: JR > G is Coo. Then on V. Yt l · · . In ) = (exp l.. . e) = 4>(e/2. Xtl · · .I . for all s E [ . ) ) = (exp I. Choose a basis X" . Now choose e > 0 so that 4>(1) E exp( ! U) for I I I � e. i f a = b2 for b E exp( ! U).Lie Groups 387 I I . •:. . . So 4> is Coo at e. I] . .) = exp I Y. COROLLARY. exp X /4 E exp( t U ) . •:. This shows that every a E exp( t U) has a unique square root in the set exp( t U).) · · · (exp ln Xn ) is Coo and clearly (*) and (*) shows that 4> 0 Vr is Coo. if a = exp(X/2) for X E U. (exp In Yn ) . then S o a has a square root i n exp( b = exp(Y/2) for Y E U. so there is Yj E He such that 4> (exp IX. so 2 2 exp(X/2) = a = b = [exp(Y/2)] = exp Y. X E exp( ! U). 4> (se) = expsX 1 2. By induction we have 4> (e/2n ) = exp(X/2n ) .. so X/4 = Y/2. then 2 a = exp(X/2) = [exp(X/4)] . )m = [exp(X/2 n )]m = exp(m/2n . Since 2 [4>( e/2)] = 4>(e) = [exp X/2f . on which exp is oneone. Let 4>(e) = exp X. Now the map Vr : JRn > G given by Vr (l" . X" for G. 4> = (4) o Vr ) o r ' .
extent (0 which this equation fails to hold. It is easy to see that (= �l(n . If G is a Lie group and 1 exp(IX) exp(IY) exp lX exp l Y = exp { 1 2 [X. y] + 0(t 3 )}. since exp(M + N) (exp M)(exp M) So if M is skewsymmetric. the x ver (exp M)(exp M)t = I . the equation exp(X + Y) = exp X exp Y holds whenever [X. (2) (I) exp lX exp l Y = exr l (X + Y) + '2 [X. IR ») > GL(n. bounded for small I. PROOF Apply Corollary 1 1 to the continuous isomorphism and its inverse . 2 The properties of the particular exponential map exp : IRn may !lOW be used to show that O(n) is a Lie group. Y] measures.e. then = (exp M) ( exp N) when MN = NM. Just as in GL(n. there is a diffeo morphism between them which is also a group isomorphism. Then I = A . THEOREM. M = _Mt. In general.1)/2 dimensional submanifold of GL(n . Y E Ge. At = (cxp M)(exp N). [X. IR). so e p N = (exp M) . Thus 0(13) will denote differenl functions at different limes. we will denote it by 0(13). In the following Theorem. Y] + 0(1 3 ) \ j X. up to first order. Moreover. then they are isomorphic as Lie groups. we have = I. COROLLARY.1 = exp(M). IR). i. So exp Mt = At = exp(M). y] = 0 (Problem 1 3). C on sely. Since O(n) is a subgroup. that is. 14.. Let At = exp N. then 12 . It follows that a neighborhood of I in O(n) is an n(n . O(n) is itself a submanifold of GL(n . For sufficiently small M and N this implies that N = M. IR).388 Chapter 10 1 3 . any A E O(n) sufficiently close t o I can be written A = exp M for some M. If G and G' are Lie groups which are isomorphic as topo logical groups. IR) exp(Mt) = ( exp M)t . to indicate that a function c : IR > Ge has the property that C(I)/13 i. hencc Mt = M. exp M E O(n). and in its proof. •:. y] + O(l3)} (3) exp lX exp IY exp(IX) = exp{IY + 1 2 [X.
d d (iii) </> (I) = .o 12 _ _ + "2 [Y ( YJ)](exp sX) + 0(1 3 ) . (i) We have Similarly. du u=o d YJ(a) = J(a · exp uY).J(exp sX exp l Y ) = dt d = ( YJ) (exp sX exp l Y ) . for any F. I J(exp sX exp l Y exp uy) Applying (iii) to YJ instead of J gives (iv) </>" (1) = [Y ( YJ)](exp sX exp I Y). 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 ) . Now Taylor's Theorem says that Suppose that J(e) = (v) J(expsX exp l Y ) = J(exp sX) + I ( YJ)(expsX) _ _ o. 2 . let Then l </>(1) = J(exp sX exp l y). exp uX). d d.Lie Graul)j 389 PROOF. </>(1) = </>(0) + </>' (0)1 + </> ) 1 2 + 0(1 3 ). Similarly. Then we have '�. (ii) d XJ(a) = Xa (J) = La. u u=o by (ii). du u=o _ I For fixed s.X(J) = X(J 0 La) = J(a .
= X + Y. + t 2 Z2 )) + 0(1 3 ) = t ( 2. so by (vi) we have (viii) /(exp Z (t)) = J (exp(t Z .390 Chapter 10 Substituting in (v) for F = J. F = YJ. Z{I) = J(A(t)) + Since w e can take the . If J(e) = 0. (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 ).+ Z2 = .+ X Y + 2 · 2 2 which gives thus proving (I). Applying Taylor's formula to t Z .f's t o be coordinate functions.  X + Y = 2. ] exp tX exp t Y for some gives exp Z (I ) Coo function Z with values in Ge. = Z for some Z" Z 2 E Ge. J)(e) + t 2(22 1)(e) t2 _ _ + 2 [Z. comparison o f (vii) and (viii) gives 2. 2. (Zt /)](e) + 0(1 3 ) . . xx YY . Equation (2) follows immediately from (I). In particular.  Z. + t 2 Z2 + 0(1 3 ). then clearly J(A(t) + 0(1 3 )) 3 O{l ). 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 ) .
(I) = largest integer � Then �I < I. ' 3 J(exp lX exp l Y exp(IX)) = J ( exp(lS. k.Lie Groups 391 To prove (3).X. Let X. Let 1) c Ge be the set of all X E Ge such that exp lX E H f all I . ' 1.X.. then we also have (x) Comparing (ix) and (x) gives the desired result.) = (expl.J)(e) = 1 (.(t) ::: ::: I. More precisely..1 E H. 3 + [ . .YX . Ifwe write 0 ' (1 ) . again choose J with J(e) = O.).:. The work involved in proving Theorem 14 is justified by its role in the f ol lowing beautiful theorem. If G is a Lie group and H e G is a closed subset which is also a subgroup (algebraically) .X)J](e) + 1 + . or Proqf We can assume I. . 1 5.51 J)](e) + 0(1 ). We attempt to reconstruct the Lie algebra of H as f ollows. # O . since exp(I. + I S. . there is a Coo structure on H. Then similar calculations give (ix) J(exp lX exp l Y exp(IX)) = I [(X + Y . Then X E lj. 3 [(XX 2  yy + ""2 + 2 + X Y  XX  xx . let k. Suppose expl. > 0 with each I.)]  (e) ' 3 exp lX exp I Y exp(IX) = exp(lS.51 ( . THEOREI'vl. > X and let I.5. then H is a Lie subgroup of G. I Notice that f ormula (2) is a special case of Theorem 516 (compare also with Problems 516 and 518). E Ge with X. or PROOF. with llle relative topologJ') that makes it a Lie subgroup of G. Assertion J . + I S.)) + 0(1 ) 5. . + 0(1 )). E H f all i. For I > 0.X. > 0. J)(e) + 1 ' ( .
. is a dilleomorphism in some neighborhood of O.(t)I. Thus exp IX E H. Choose positive I. we can write. . We now claim that I) c Ge is a vector subspace.D. then H n exp(U) = exp(I) n U). ) = [ exp (t. = I) Ell Ij'. Y E I). . > G defined by t/>(X + X') = exp X exp X' Xl . ). Proq{ Choose a basis Then t/> is given by of G. Alternatively. exp(k. Y] E 1). Y] + O( l /n2) l . [Similarly. so X E I).X exp . sX E I) for all s E IR. . . We clearly also have exp lX E H for 1 < 0. . 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. but we will not even use this fact.) k. using (2) of Theorem 1 4 we see that [X. It dearly suffices to show thaI old if U is small enough. = X + Y + Z(t. for fixed I.] Now let U be an open neighborhood of 0 E Ge on which exp is a diffeomor phism.X. > IX. so that G. (t) > I. so that I) is a subalgebra. 12 211 f taking limits as n > 00 gives exp 1 (X + Y) E H. Clearly X E I) implie. Then Assertion J implies that X + Y E 1).Y 11 11 1 I )n = exp {I(X + Y) + [X. If X.X. since H is closed and exp is continuous. . > 0 and let X. X' E I)' X" . ( exp . t (l) E H.k. QE. . with . Xk . Then exp(1) n U) is a submanif of G.(t)I.392 so Now Olapter 1 0 I.X. Xn X E 1). Choose a subspace Il' C Ge complementary to I). . Assertion 2. Xk a basis for I). . . The map t/> : G.
ax 1jr* (�' I ) 0 = X. it follows fi'om Assertion 1 that X' E I). and a E exp(1) n U) .L7= a. Then a = exp X exp X' X E W. Prorif.D. QE. . . so 0 = X'. If the assertion were false. Clearly exp(1) 11' n U) CH n exp(U). •:.' E 1)' with X/ 0 and exp X/ E H. and 4> of Assertion 2 is a diffeomor phism on W x W'. Choose a neighborhood = W x W' of Ge on which exp is a diffeomorphism. 7 Choosing a subsequence jf necessary. X' E W'.X. let a E H n exp( U). an) is a diffeomorphism of Ge onto jRn . there would be X. We can now complete the proof of the theorem.Lie Grau!}. Since exp(I /I1. . . )(11. Choose an inner product on �' and let K C I)' be the compact set of all X' E IJ' with I ::: IX'I ::: 2. a contradiction. . 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. This is clear. we can assume Xi' I/n. since QE. is a diffeomorphism in a neighborhood of 0 E ]R" .X.D. To prove the reverse inclusion. > 0.\ 393 Since the map . 1 it suffices to show thaI 1+ (a) .)0 X' E K. Since a. There is a neighborhood V' of 0 in I)' such that exp X' ¢ H if 0 # X' E V'. . Choose integers JI j 'V'lith Assertion 3.') E H. exp X E H we obtain exp X' E H. .
Clearly. Y) = X(w(l7)) . cvn (e) span G/'. iJ. La* dw = d(La*w) dw. /\ o/k = L Aj o/ J f certain COI/SWI/Is aj.Y(w(X)) .··· <iJ. of Lie groups are better expressed in terms of f orms. " " Xn E Ge.o} I /\ . invariant Iform (J) and left invariant vector fields X and Y we have dw(X. A f orm w is called left invariant if La'w = w f all a E G. Y) = w(e)([X. The f ormula on page 215 implies that for a ieI'. wn(e) is the dual basis to or then any Coo vector field X can be written Xl . (J)n are left invariant I f orms such that cv I (e). . .w([X. X = 'L. The interplay between left im"ariant and right invariant vector fields is the subject of Problem I I. . . . 1 0 U p t o Ilo\V: w e have concentrated on the left invariant vector fields.. then f a E G we have or CO:: . . H ence dw(e)(X. It f oHows that any left invariant f orm is If (J) is left invariant. . 9]) = w([xj ']). fj Xj j=1 Then f or Coo functions /i so 0/ is Coo. . . then every left invariant kform i� i]<. = so dw i. also left invariant. . . . H ere we consider the case of f orms. This means that or web) = La * [w(ab)]. a left invariant kform w is determined by its value wee) E Qk(Ge) ' Hence� i f cv l � . the bracket being the operation in A. If wi (e). but man� propertie. Y]). .394 Olapte. .. . L ai1 .
then A is abelian (converse of Corollary 7). then all left invariant If orms w are also right invariant. then so o/*w is right invariant. PROPOSITION. (I) Clearly so If (J) is left invariant.(X) = X for X E Ge. It follows fi·om (*) that [X. So 1jr"X is the tangent vector at 1 = 0 of 1 1+ (exptX). we have dw is also left and right invariant.I . Since o/*w and (J) are both left invariant. then 1jr * we = (3) (I) A f orm W is left invariant if and only if (4) If G is abelian. (_J)k we. So it is enough to show that 1jr. 1jr* (we) = ( _ J )k We. So dw = 0 for all left invariant Iforms.! 395 16. The converse is similar. so The f orm But So dw = o. (2) It clearly suffices to prove this for k = J. Yj = 0 for all X. Let 1jr : G > G be 1jr(a) = a. (4) If G is abelian. this tangent vector is just x. then 1jr*w is right invariant. Y E A. then dw = o. (3) If W is a left and right invariant kform. Now X is the tangent vector at t = 0 of the curve 1 1+ explX.1 = exp( I X).Lie Group.. P ROOF. . (2) If We E Qk(Gel. If W is left and right invariant.
. .j � . if G is abelian. = dw is left invariant for any left invariant w. " " Xn of � ). X] + 0(1 3)/1 2 . then for have Hence HX. There are constants Ci� n Since . From skewsymmetry of [ . By Theorem 1 4. . wn(e).L Ci� Wi /\ w) i<j n (2) From (*) on page 394 we obtain = o. it foHows that for a basis w 1 . (J)1I in a neighborhood of 0 E jRn such that . dwk in terms of the 0/ /\ wi . . . . Xl . ll=l dwk = . X)] = L Ci�Xk : k=1 dearly we also have . [Xi. . then we can find evelywherc linearly independent I forms cv 1 . Y E Ge we Allernale proqf qf (4). of of invariant Iforms we can express each such that [Xi . It turns out that (2) is exactly what we obtain fi'om the relation d 2 w k = O. . In fact. . � j k=l n The numbers Ci) are called the constants of structure of G (with respect to the basis XI . . we can prove (Problem 30) that if Ci� are constants satisf ying (I) and (2). Y] + 0(1 3)/1 2 = HY. we obtain [X.396 Ozapter /0 X. . First choose XI .L Ci� wi /\ wi i. . Con dition (2) is thus an integrability condition.:. X)] = " Cik Xk . ] and the Jacobi identity we obtain (I) Ci) = C)� L( C! C/'k + cJk cL + Cl'iC/') ) = . Letting 1 > 0. Y] [Y. ' " X E Ge dual to Wi (e).
. Let G be a Lie group with a basis of left invariant I forms w'.a} where {jk iiJk = 0. 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).iJ n . wIn and constants of structure Ci� . = E > IJ . . .. [i} M "2 : > a. . : r M and r G are each diffeomorphisms in some neighborhood of (p. . /\ wj J) L j .. From this latter fact and (a suitable local version of) Theorem 5 we could immediately deduce the following Theorem. .L Ci� [iJi /\ (iJj _ w j )+ (iJ i _ wi ) /\ wj ]. 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. .[a'. for which we supply an independent proof. iiJn are each linearly inde pendent. Let . Now iJ'.i< Ci� ([iJi /\ iJj ] . . . the sets 8 1 . . Let '" : M x G M and M x G G be the projections.wk Choose G and let r be the folium through (p. . . Let Mn be a differentiable manifold and let e . . the existence of such (J/ implies (Problem 29) that we can define a multiplication (a. THEOREM. • . i<j By Proposition 714. so on J(p. . w'. . .Lie Group. a). en and wi. a} . ." Ck. 397 Moreover.iii ) . . M x G is foliated by IIdimensional manifolds whose tangent spaces at each point are annihilated by all iJk . wn are linearly independent everyvvhere. far 1 7 . Hence . which is the set of vectors in (M x G)(p. 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. = > cI(iJk .. _ " > . e n be everywhere linearly independent I forms on M satisf ying k .
398 Let OUipter 10 j: U > M x G be t h e map j( q ) Since (jk .l ) w (Y2(1)) ( [ Ly. and let fi .j* H2* a/ (H.iiJk = 0 On J. (I) _ ' J. We must show that YI W (Y2(1)) Case 1. (I). we have 0 = j*( {i .' ] d y. 0 j)* e k . J(q)) c r. w (y.li .Ii : M > G be two Coo maps such that . Then fi and . _ == == == It is also po"ible to say by how much any two such maps differ: 1 8 .(H2 ° i)* a/ ek J*w k . let G be a Lie group. M C�2) == Y2* W == = == It follows thai U we regard Yl a� b riven. then it becomes an ordinary dificrclltial equation for dY dt = 2 (d.Ii differ by a left translation.Ii * (w) for all lefi im·ariant I forms w. )* W (Yz (l )) ] . there is a (unique) a E G such that j. . and write this equation out in a coordinate system.(I)y. (I)) (�IJ IR and the fwo maps YI ' Y2 : IR > G satisji Y2· For every left invariant Iform w == (d. we have YI (O) == Y2(O) . == == == j. (I)y. (�I ) [Ly.1 ) YI* w (�IJ * dI · Y2 (of the type considered .*(w) La o . THEOREM.(I)y. j* H tek . Let M be a connected manifold. [ (L y. that is. == PEDESTRIAN PROOF.iii ) == (q.:.(I).
In fact. M = IR. : G x G > G be projection on th e ith factor.(G). Now define h : M > G x G bv h(p) = Uj (p). if t. y. Case 2. In other words. )*(w) = yt ( L:w) = y/(w) = y. aJd p ) = bJ ( p) for all p E M .(O).. . Since M is connected. ../i *(w) = y/(w) .Lie Group. 0 c. s o /2(1') = a · . y. (0). a) : a E G}. . " Then t.lr2*(J/ ). Gelleral case. But this solution is clearly y. we have For any p E M there is a Le t y. it follows that heM) is contained in some left coset of t. then Since La 0 y. are arbitrary'..(O) = a · ydO). By Case 2.. ../i (p). (t) = a .b} = i=l n ker(Jr) *a/ . is an integrable distribution on G x G. b E G with . By assumption. Let po E M.:* . (t) for ali I .. but the maps y" y. it follows fi'om Case 1 that La 0 y. y.*(w) = c* J 2(W) = c* . Choose a E G so that (La 0 y.. ELEGANTPROOF. Choose a E G so that y. 399 in the Addendum to Chapter 5). .(G) c G x G is the diagonal subgroup {(a. w" for the left invariant Iforms.(G). . i n particular for t = I . so it has a unique solution with the initial condition y. .. (0) = y. there are a . = f. = y.*(w). let Ll(a. = y. /2 (p)). Case 3. Let 7[. b) E G x G. arc the left cosets of t. If (J) is a left invariant If orm.(O) = y. For (a.. Then Coo curve c: IR > M with dO) = Po and c ( I ) = p... theu the maximal integral manif olds of t. Choose a basis w ' .
the form Ra*w is left invariant. 20.ourse. For each a E G. we can define Coo an an 7 Since L Ja '" a" is llsLlal1y kept fixed in any discussion. If G is a connected Lie group and J: G > G is a map preserving left invariant forms. and if J : G IR is a Coo function v'lith compact support. ( ) = {I} . PROOF. If G is compact and connected and w is a left invariant Jlform� then (J) is also right invariant.400 Ozapter 10 1 9. == where I note that Lb is an orientation presenTing diffeomorphism. So { G c IR is a compact connected subgroup of IR{O}. left invariancr of implies that a" in other words. of c. . so there is a unique real number with J(a) = Ra'w = l(a)w. COROLLARY. then determines an orientation on G. Clearly. These generally turn out to be quite diff erent from the left invariant IZforms (see the example in Problem 25). ''\'e can. so L Jan = L Lb*uan) = LU o Lb )Lb*a" = LU o Lb )aN• g(a) J(ba). then J for a unique a E G. If is a left invarjant nform. But in one case they coincide. Suppose w # O. PROPOSITION. also consider right invariant l1forms. •:. iG /(a) da L /(ba) da. an left invariant nform� arc a constant multiple of any nonzero one. = La While left im'ariant Iforms play a fundamental role in the study of G. The latter notation has advantages in certain cases. For example. th e left invariant nforms are also very important. this is often abbreviated to L J or fG J(a)dCl. = \\'hich proves the formula]. Hence J G () l(ab) J(ba) = J(a) · J(b).
(I) Since the map 1. ) is any Riemannian metric we can choose a biinvariant IIf orm an and define a biinvariant (( . : Ge > Ge is just multiplication by . then (2) The geodesics y with y(O) = e are precisely the I parameter subgroups of G.. PROOF.) (Y(u .e. F fixed I.l is also an isometry.I (see the proof of Proposi tion 16 (2)). PROPOSITION. the maps I 1+ exp(tX) for some X E A. Clearly Ie reverses geodesics through e.) : 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. G. (I) For any a E G. which reverses geodesics through fa (y(t)) = y (I). Let G be a Lie group with a biinvariant m etric. the map le* : Ga 7 Ga. the map fa : G 2 1 . i.Lie Graul). )) on G by We are finally ready to account for some terminology from Chapter 9. But also .. i. 401 Vife can also consider Riemannian metrics on G.)(y (u)) fy(. In fact: if ( .e. so it is an isometry on Ge. if y > G given by 1a ( b) = ab' a is an isometry is a geodesic and y (O) = a. a.I)) = y(t + u) = y(u + 21). Then y is a geodesic and y(O) fy(ljfe( Y (u)) = = Y(I) . So = fy(. Since for any a E G.. In the case of a compact group G there is always a Riemannian metric on G which is both left and right invariant. Since fa = Ra1e Ra . let or y (1 + u ).
' If VI ..402 so Chapter 10 y(t)y(u)y(t) = y(u + 21). y is a I parameter subgroup. (p)(X" . . . . i=l d w = LO/ ' Vi o i=l a simple calculation shows that this definition does not depend on the choice of basis VI . (J)d such that f X" . . '. ..: i=1 we will write simply ror anv Vvalued kform w we define a Vvalued (k + I)form dw by d dw = L do/ . By continuity. Vd is a basis for V. then y(t' + I") = y(l)n'+n" = y(t')y(t "). •:. . and geodesics through e are determined by their tangent vectors at / = o. . . . since there are I parameter subgroups with any tange nt vector at I = 0. . . x Mp > V. . . . .. If V is a ddimensional vector space. Xk )v. . . y(l1I) = y(t)n for any integer n. 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 . . These are the only geodesics. so y is a homomorphism on Q. 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. Vd for V. then there are ordinary kforms cv l . . It follows by induction that If t f = n'! and t ff = n"t for integers n' and n". . Xk E Mp we have or k times d w(p)(X" " " Xk ) = L w.
Now suppose that Xl " ' " Xn E Ge = A is a basis. • . Vd . Using the bilinear map [ . Uc and 403 p: V" . where V and V have . (J)n is a dual basis of lefi invariant I forms. namely the form w defined by w(a)(X(a)) = X E A. The form w defined by (*) can clearly be written Then (I) . • • • . . . Vj ) i=l j=l is a Wvalued (k+I)form. we have. vf . If w is a Vvalued kform c w = L {J/ ' Ui i= 1 ry = I >j . Uc or VJ . .. • • V x V > W is a bilinear map. and that cvI . /\ ryj · p(u. These concepts have a natural place in the study of a Lie group G. a calculation shows that this does not depend on the choice of bases u" • . for any Avalued kform ry alld any Avalued Iform A on G. Although there is no natural way to choose a basis ofleft invariant Iforms on G.Lie Groups Similarly. . Vd. suppose bases UJ . . j=l d and ry is a Vvalued Iform then c d L L w. We will denote this Wvalued (k + 1) form by p ( w /\ ry). . a new Avalued (k + I)form fry /\ A] on G. • . there iI a natural nvalued If orm on G . respectively. ]: A x A > A.
Y E G" then dw(a)(Xa. k=1 n so (2) Comparing (I) and (2). Y]) = [X. V]). .404 On the other hand. h) = w(e)([X. Y] is left invariant [ . then (Problem 20) we can define a �valued function V(f) on G.: Xk . 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. 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. V) = .. Yo) = 0 . ] in G.0 . V) = V(w( V)) .w(a)([X. On the other hand. 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. Chapter 10 [X" Xj] = L C.w([V. Y] = [W(a)(Xa). 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. by definition of w. V = Y for some X. Y]a) = w(e)([X. I f we choose V = X. w(V) is a �valued function on G. For vector fields V and V we can then define dw(U. 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. I dw(U.[W(V).V(w(V)) .W( V)] · I V and V we have } [X.
\ 405 WARNING: In some books the equation which we have just deduced appears as dOJ(U. dx .Lie Gmu/i. It comes about because some books do not use the factor ( k + /)!/ k! I! in the definition of /\ . The appearance of the factor 4 bere has nothing to do with the 4 in the other f orm of the structure equations.HOJ(U). makes their dOJ equal to � of ours for Iforms OJ. Tbis makes their A /\ ry equal to 4 of ours for Iforms A and ry. OJ(V)]. . ) as L dOJ. V) = . /\ dx . Then the definition of deL OJ.
r for A E O(n) and r a translation. then so is every coset g H. . (b) If H is open. y) 1+ (a) Find I. = (x. xy 2. Let I : IRn > IRn be distance presen'ing. � a= 1 (a) Show that this means that (It is actually clear a priori that M defined in this way is left invariant.f(c) Conclude that G is a Lie group. (a) Show that U n +1 is a neighborhood of u n . Show that the tangent bWldle TG ofa Lie group G can always be made into a Lie group. We have computed that f or M E �l(n. (b) Show that I takes planes to planes. Coco Let G be a group which is also a Coo manifold. l . and suppose that (x. E U. . (c) Show that I is a linear transformation.y) (b) Show that (e. (a) Show that I takes straight lines to straight lines. U a neighborhood of e E G.406 Chapter 10 PROBLEMS is I. 6.ak ax k. (a) If H is open. Let G be a topological group and H e G a subgroup.I when I : G x G > G x G is I(x. (Use Problem 3. or LA since LA is linear. with I(O) = o. Let G be a connected topological group. 3. Let G be a topological group.) (c) If G is locally compact and connected. and Let u n denote all products a l . then G is acompact. 4. e) is a regular point of . xy) . 5. f LA.) (b) Find the right invariant vector field with value M at 1 . and H e G a subgroup.L Mxkl . M(A) = A · M E IRn ' = GL(n. 7. and hence an element of O(n). lRlA = . (d) Show that any element of £(n) can be written A . an for a. . IR) we have where M= .1 n M xkl (A) = " MaI A ka . Show that the closure ii of H is also a subgroup. (b) Conclude that Un u n = G. then H is closed.
then ljr is an isomorphism from G to Go. I I.I . If A is a Lie algebra. Define ( VI . (e) Use this to give another proof that A is abelian when G is abelian. If G is a group. we define the opposite group GO to be the same set with the multiplication . Show that Corollary 7 is false if G is not assumed connected. exp(X + Y) = (exp X)(exp Y).) if ljrl = ljr. 8. provided that U is connected. Show that the set of all ? equivalence classes. and where ljr : V > H satisfies ljr(a) · ljr (b)' = 4>(ab . with operation [ .sin a sin a cos a 13. where V e G is an open neighbor hood of c with V · V. ( �) O I ) . In particular. A= ( 0 a a 0 ) ( = cos a . Y] (� �) = B= O. we define the opposite Lie algebra AO to be the same set with the operation [X. ab E U 9. b E V. ljrd ? ( V" ljr. Let X.· to a homomorphism of G into H . Then (A. then 4> can be extended unique!. ] be the operation on Ge obtained by using right invariant vector fields instead oneft invariant ones. ]) is isomorphic to £( GO). In Theorem 5.I C U. .Lie Groups 407 (a) For each C E G. ]. defined by a. (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 = b · a. (b) More generally. b. show that 4> and ljr are equal even if they are defined only on a neighborhood U of e E G. 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. (a) Use Lemma 513 to show that (exp sX)(exp t Y) = (exp t y)(expsX). consider pairs (V. Y]. and hence to �o. (a) GO is a group. and if ljr : G > G is a 1+ a. use Theorem 5 to show that exp is a homomorphism on the subspace of Ge spanned by X and Y. can be made into a covering space of G. (b) Conclude that if G is simplyconnected. and X 1+ X is an isomorphism of A onto AO.' ) for a. 10. for all c E G. (c) £(GO) is isomorphic to [£(G)]O = AO. ljr). on some smaller neighborhood of c. Y]0 = [X. Y E Ge with [X. (b) AO is a Lie algebra. [ . (d) Let [ . 1 2.
0 <p. (a) Let (x. IR) > IR .(p)) + <p. there is a neighborhood U of e E G such that every element in U has a unique nth foot in U. Y j = O. (The diagonalizable matrices are dense..}. (e) Usc this fact to give a fancy proof that trace MN = trace NM. (c) For G = S' . (Look at trace(M N . etracc M . (p) = X(ry. Show tbat no matter how small U is. show tbat there is no neighborbood U which has this property for all n. 0) of S' (considered as a subsct of IR'). � dry. ( e ) = O. ( Y(o/. Suppose that [X. show that In other words. (e) Conclude that it holds for all M with complex entries. (Since det. = 0/. 1 5.408 Chapter 10 (a) Show that 14. witbout using (c).) (g) Now usc this result and Proposition 9 to give a fancy proof of (c). V) be a coordinate system around e E G with x. 0 0/. show that for the homomorphism det : GL(n. and trace are homomorphisms. (a) Let U be a neighborhood of the identity ( 1 . {ryrl is generated by X + Y. Problem 1 3 implies that exp t (X + Y) = (exp tX)(exp t Y ) i f [X. (a) If M is a diagonal matrix with complex entries. A more general result holds. = <p.{O}. Y] = 0. : AI(n. 1 6. Let X and Y be vector fields on a Coo manif old M with corresponding local I parameter families of local diffeomorphisms {<p. (p)). (b) Show tbat for each n ::: I . and let ry.) (0 Prove the result in part (d) directly. 1 7.NM) = trace[M. N]. show that det exp M = (b) Show that the same equation holds for all diagonalizable M with complex entries. (b) Using Corollary 51 2.{Oll = IR is just M 1+ trace M. IR) > £(IR . it suffices to look at matrices with only one nonzero entry. compare Problem 715. there are elements a E U which have square roots outside U in addition to their square root in U.) (d) Using Proposition 9. {o/s}.. the map det. Let .
(c) Let CI ' ' ' ' . Hint: Join any a E H to e by a Coo curve c . It is even true that H e G is a Lie subgroup if H is path connected (by not necessarily Coo paths).) (b) Let K e G be the connected Lie subgroup of G with Lie algebra �. The argument of part (a) is essentially equivalent to the initial part of the deduction of (I). Hint: This f ollows immediately from one of our propositions. . then H is a Lie subgroup. Osaka Math. . .Lie Groups for 409 Coo functions J. show that K C H.]. For a E G. isomorphic to G L(n . (d) If H e G is a subgroup and an immersed submanifold. Ck (tk ) . The map Ad is called the adjoint representation. The map is denoted by Ad(a). H is a Lie subgroup of G.l (b). and show that the tangent vectors of c lie in the distribution constructed in the proof of Theorem 4. see Yamabe. . Thus. is the set of all nonsingular linear transformations of the vector space A onto itself (thus. t ) = cd t l ) . Thus we have a homomorphism Ad : G > Aut(�). 19. Ck be curves in H with {c. and let H C G be a subgroup of G (algebraically). . . On an arcwise connected subgroup ifa lie group. the automorphism group of A. usually Ad(a)(X) is denoted simply by Ad(a)X. (b) Show thai exp(Ad(a)X) = a ( exp X )a . (a) Show that I) is a subalgebra of Ge• (Use Theorem 14. By considering or k the map J(t l . where AUI (A). .e) > G are differentiable.'(O)) a basis f �. 2 (1 950). (a) Ad(ab) = Ad(a)oAd(b).' . Show that H C K. 1314.) 18. LaRa . Let � c Ge be the set of tangent vectors to all Coo paths lying in H. show that (b) If Ol. Let G be a Lie group. Show that (e. it suffices to know that exptX exp t Y = exp{t(X + y) + O(t)}. fJ : (01 ' fJ)'(O) (c) Also deduce this result from Theorem 1 4 (1). . consider the map b 1+ aba . IR) if A has dimension n).l = = 01 ' (0) + fJ'(O) . (Not even the full strength of (I) is needed. such that every a E H can be joined to e by a Coo path lying in H.
This tangent space is isomorphic to End(A). Show that H is a normal subgroup of G if and only if I) = £ ( H) is an ideal of � = £(G). 2! (g) Let G be a connected Lie group and H e G a Lie subgroup. we let it operate on Y E A by (Compare with the case A = JR". (O)(Y) = dt dI 1=0 c(Y). � (= Ge ) > . where J = "L1=J Jl . AlIt(�) = GL(n. End(�) = n x n matrices. (A proof may also be given using the fact that [X . . ifand only if [X. (e) Since Ad : G > �. (X)(y) = [X. Vd for V. w e have the map Ad*. .at the Jdenuty map 10 of � to Jtself. Show that this definition is indepen dent of the choice of basis VJ . (a) Let J : M > V. . . . Vl for Jl : M > JR. with basis X(J) = L Xp (Jl) i=] . • . Y j Y 1+ [X.) Use (d) to show that Ad.) (d) Show that Ad(exp tX)Y = Y + t [X.. Y E I). Vd.) The map (ad X) ' Ig + ad X + .nt space of Aut(�) . . (f) Conclude that Ad(exp X) = exp(ad X) = c . LxY. 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. JR) show that J . JR).410 Chapter 1 0 Ad(A)M = A Mr (c) For A E GL(n.JR) and M E � 1 (n . Yj E � for all X E �. Vl. where V is a finite dimensional vector space. . For Xp E Mp. . that is. . Y] + 0 (1 ' ) . Yj is denoted by ad X E End(�). (It suffices to show this for M in a neighborhood of 0. �ange.+ . . Yj. V J . then to regard c'(O) as an element of Aut(A). define Xp (J) E V by d 20. .
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
f* : Hk (M) > Hk(N). .I . �ow suppose that \·ve have an exact sequence of complexes fJ Ci o > C > C2 > C3 > O.422 GiW/)le. ) > Hk(C2 ) and fJ: Hk(C2 ) > Hk (C3)? The nicest thing that could h appen would be for the following dia gram to be exact: fJ Hk(C2 ) > Hk (C3) > o.I (y)). This is nol true . Cik (x) + d/I (Cik . then this map is jus. When Clk = Ck(M).J J for if we change x to x + dl k .I (y) for some y E C/ . ) > Ci u v . and Ci : Ck (M) > Ck (N) is J* for J : N > M. Whal does this imply about the maps Ci: Hk(C.I (y)) = Cik (x) + Cik(dl k . C/ = Ck (N). if V and V arc overlapping portions of S 2 fOJ which there is a deformation retraction of V n V into S l . then Cik (x) is chanl!ed to Cik (X + dl k . For example. then we have an exac i sequen lT 0 > Hk(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 ).
. Moreover.1 (y') . . If 0 > CI > C2 > complexes) then there are linear maps '" fJ C. let x' = fJk . · E d/ . > 0 is a short exact sequence of so that the following infinitelv long sequence is exact (everywhere): 0 > Since ". Let x fJ '" /.k+ l . thus dle)') ". we must check that the result does not depend on the choice of x E cl representing the element of Hk (C3). > Hk (Cr) > Hk (C2) > Hk(C3) > Hk+l (Cr ) > . diagram (*) should be kept at hand. . Ho (Cr) > H0 (C2) > H0 (C3) > H I (Cr) > . Then x = tli (x) = O.EtCU/:r. 0. Throughout the proof. this implies that dl k+1 (0) = 0.k .I (x') = dll fJk . THEOREM. there is y E C.k+1 is oneone. 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.I .I (y').I (x') for x' E C.k with fJk(y) = x. Bv exactness of the middle row of (*) . fJ '" 8 . This means that dl (y) = sO we choose d/ . and hence z = o. So dl(y) E ker fJk+1 = i lnage".I Cr') as . In order to prove that ok is welldefined.k+I (z) for some (unique) Z E C k+I . I ci with = PROOF.le obtain 0 E Hk+I(Cl ) if we start with an element of the form di . . something very nice is true: o II o II IR: 2. Theil o = dl (x) = dl fJk (y) = fJk+1 dl (y) .I (y') = fJkdl .)' 423 an exact sequence II 1'\ evertheless.on in the Realm qfAlgebraic but not 70/l0/0t.J". . In this case. So we have to show that V.
". 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. onto circles. . Moreover. . this is left to the reader. Let x E q k satisfy dl k (X) = 0.� corollarv of Lemma I and Theorem 2. The proof that the sequence is exact consists of 6 similar diagram chases. together with the first part of the proof that 0 is welldefined. � As several of the Problems shO\.I . If M = U ]I. the :Maver. and a deformation retraction or V Viet oris sequence i� n 11 onto 2 circles.. the cohomology of nearly everything can be computed by a suitable application of the MayerVietoris sequence. THEOREM (THE I\1AYERVIETORIS SEQ UENCE). and the open sets 1) and V illustrated below. 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/..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 . ''\'(' consider the torus T = S l X S ' . '''Ie will supply the proof that kerO' C image o. . the definition of 8 i mmed i ately shows that the image of this element under 0 is precisely the cla" represented by x . As a simple example. then we have an exact sequence (eventually ending in O's): o� U HO l M ) . Since there is a deformation retraction of V and r H A (M) � HA (U) Ell HA ( V) > HA (U n )1) 2.'h ere V and V are open. Now So fik. HA +1 (M) > ... 3.·. •:.I (y) represents alI element of HkI (C3). All of Theorem 816 can be derived directly from the followin.
By induction on k. Aswme the theorem for k . = . it induces a map V'.H ' (U n VI � H' (T) � o. Similar rea soning shows that this map also has a I dimensional kernel. 4. So the kernel of this map is I dimensional.dim V2 + dim V3 .a( VI ) > V3 . . this map is oneone. PROPOSITION.I dim Vk . 7 Vk l 7 Vk 7 0 0 = dim h . The reasoning used here can fortunately be systematized. . so its kernel is I I dimensional.. which implie> that VI = O . For k I We have the sequenu Exactness means that {OJ C VI is the kernel of the map VI > 0. .dim V3 + . = PROOF. +:+ 0 = dim V2 /a( Vd . and consequencly the map H I (T) > H I (U) Ell H I (V) has a Idimensional image.dim VI + dim V2 .J . Moreovel. So we have an exact sequence of 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)). then 112 7 .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 . If the sequence o 7 VI + a is exact. " + ( _ I ) k .J vector spac("� hence which proves the theorem for k. Since the map V2 > V3 has kernel a(Yd. Thus the image of the map H (U) Ell H I (Y) > H I (U n V) is Idimensional. It follows that dim H I (T) = 2.
. .::.�+ I L. a ne\\' structur� which we will now define. 1 t is a difficult I heorem that every ('OC manifold has a triangulation. . arising from a «triangulation" of M. then for some k the intersection ani n an} is a k ·lacr of both alii and an}." is defined to be a dillcrent.vard 3simplcxc .X i . Now by a triangulation of a compact nmanifold M we mean a [mite coHection 1 allj} of diffeomorphic ima.". then the image of a kface of /j.n = Jx ) E < < jR'"+1 . for a proof see !\1unkres.=l xi = Ji J. n which cover M and which satisf the following condition: y If a n. n allj #. . set. 0 . l>o \ (In Problem 8S .'cUom of � .. ..!!cs of /:).. Geometric Integration 1 . C M is a dif leomorphic image 0(' some /j. m is called a k face of !:J.. If /j. ''''hitney." is defined as the set /j.. m .k of the coordinates xi equal to 0 is homeomorphic to /j. although homeomorphic. .. and is called a kface of /j. 1 . ��_< 0V ��':he .���:oI1S v � A !1\ � � � � � or 2 4 \f mters(.0.ft ' €SO .j The subsel of /j. Elcmc77la�J' Dijjerf1lial Topology. .. J " " . standard tflangulaooll b' In al id \Z!J"lria.I and . we will use the MaverVietoris sequence to relate the dimensions of H k (M) to an entirely difT�rent set of numbers. . /j." obtained by setting n .426 CizajJter J J Rather than compute the cohomologv of other manifolds. The standard nsim pl ex /j. ( 0 0 1 f S' (afiter SI ' �. . k . �.
I i � H n I (M ) � H . so that Vn_ 1 is a neighborhood of the union of all (n . We let ctk be the number of these ksimplexes..Excunioll ill the Realm rljAlgebraic TO/Jolog} 427 Theor Assuming that our manifold M has a triangulation {an. Then M = V U 1111.l .1 where V n Vn _ 1 has the same cohomology as a di�oint union of Gi copies of sn. each ani an nsimplex of the triangulation.> Hk (U) Ell Hk (Vn_ I i II n II o 11 . n II 0 VnJ l > Hk ( U n Vn. one within each nsimplex ani.I i (3 ) H n'(U II nn Vn.I )simplexes of M. Consider first the case where 11 > 2.I (U) Ell H I ( Vn_d � � H� I (U n Vn_d n � Hn(M) � H ( u ) Ell Hn ( Vn_ J l II . any kface of any an. Now let U be the disjoint union of open balls. will be called a ksimplex of the triangulation .} we will call ). and let Vn_ 1 be the complement ofthe set consisting of the centers of these balls. 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 .> Hk (M) .I.
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.. dim Hk ( Vn . 111 the manifold Vn. . + (I) n an . " k =o n n + ( _ I )n.dim H I (M) + dim H'(M) . defined bv n X (M) = dim Ho(M) . THEOREM.I ) " an · L ( .al + a2 .I )k dim H k ( V. The above equations then imply that X ( Vn .] X ( M) = X ( Vnd + ( . joining the balls of the old U.l [dim H .428 Chapter J J Applying Proposition 4 to these pieces yield.I (M) .I ) "a .t l = n. X ( M) = aO . + (I)" dim H (M).2 L ( .I )face._ t l = X(M) ..I we define a new open set U which consists of a disjoint union of sets diff eomorphic to IR". . � comp ( "eWUonents of (11 = 2) .l ( Vn _ l ) = dim H n.'. . = k �O n. one for each (n .l (M) ..i l = dim Hk ( M) O ::: k � n .I For the case 11 = 2 we easily obtain the same result without splitting up the sequence. .I )k dim Hk(M ) or 5.dim H n (M) + an . . For any triangulation of a compact manifold M we have PROOF. dim H n. We now introduce the Euler characteristic X(M) of M.(..dim H (M) + a.
Lxcuniol1 . . . . . thell V . . joining the centers of the balls in the old U. . . COROLLARY (DESCARTES. .I ) Q'n . while in an other cases we have Combining these equations. ' (11 = 3) An argument precisely like that which proves the equation X(M) = X( Vnl l + ( . • . . .2 ) + [(. we have X(M) = X(Vn!) + (I )"O'n = X(Vn . + ( . + ( ..I )"O'n] n = 0'0 . the last of these is a disjoint union of 0'0 .11 the Realm q[Algebraic 7opolog1 429 vVe wil1 let 1111 2 be the complement of arcs� in the new V. the. ('omplf'ment of � I/n. E edges.I )"O'n] = x(Vo) + [(. . : V 6.3. . /<� .EULER).. .I )"O'n also shows thai Similal'I)� we introduce VIl.. . . .' .I)" IO'n _1 + ( . If a convex polyhedron has vertices. .2 IS . ..1 ) 1 0'1 + . Hence X( Vol = 0'0. each of which is smoothly co Iltractil)le to a point.rt. Vo.0'1 + . � .E+F=2 . and F faces.
we can define a new sequence. To prove that iuT +iv' is OntO. @support w LEMMA. . Then w t t = </>uw + ¢vw E C. I is clear h ac is exact. The sequence PROOF. and let {</>u . To prove . 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. suppose . 7. •:.. in fact.. </>v} be a partition of unity for 'he cover { V .he converse. A. ( V ). A.</>vw) E C. On tho' �'"::. support w U . If V c M is open. we will denme this extended form by i u'(w). V}. If C: (M) den otes the vector space of kforms with compact suppOrt on M. we encounter a very different situation. thi' shows that support i" C V n V and support A. each map iu' and i v' is oneone.(V) satisfies iU'(A t ) + iv'(A2) = 0.. if w is a form with compact support C U. . C V n V. (U) EIl C. It is clear that image Uu' Ell iv') C ker ( i u' + iv'). then w can be extended M by letting it be 0 am side V.hat O" . So (A. let (J) b e a kf())'ffi with compact support on M.. C V.430 ChapteT Jl If we turn from Hk to H. This means that AI = i.) iu' Ell iv' is oneone. ) is the i mage of At E C:(V n V ) . Since support At C V and support A." w other hand. (V) EIl C. is clearly 'he image of (</>uw.
I x IR. i. for all sn x IR"'.+I (UnV)' > H. = O. PROOF. which is diffeomorphic to S. The details will be left to the reader. V open in M. then there is a dual . > H. suppose we want to find H. Ifwe write S" = UuV in the usual way.I x lR .. For example. ( M)' > [H:(U)EIlH:(V)]* > H. (U) Ell H. 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. . so Ihat unv is diffeomorphic 1 0 S. Now suppose A E satisfies a*(A) fJ ) = A 0 (fJ 0 a) = A 0 0 = o. 9.* W3* � W]* 0 � WJ* .'" /.. . (V) > H. So a* 0 fJ * = O. U U V fol' U. We just have to show that if the sequence of linear map' )0 )0 W3 is cxaCi at W2.. . We claim that = fJ 0 t Given a w E W3 which is . > IR wilh A = fJ *( X). then so is the sequence of dual maps and space� For any A E W3* we have a* fJ* (A) = a*(A W. 0 a= O.:' I..+I (U n V) > . for IR" {OJ. .. + 8 H.e. THEOREM (MAYERVIETORIS FOR COMPACT SUPPORTS). . COROLLARY.. . > H: (U n V) > H. If the manifold M = U U V for U. starting with S l X IRm. Apply Theorem 2 10 the short exaci sequence of complexes given by the Lemma .on in the Realm qfAlgelnaic 1O/!% g) 431 8. Ihen sn x IR = (U x IR) U ( V x IR). A IR j . we "vi}} merely record one further resu1t� for later use.. and then proceed to yet anO! her application of Theorem 2. •:.EtcU1. Then A WI � w2 L wJ A there is �: W. (M) This sequence is much harder to work with Ihan the M ayerVietoris sequence. . then there is a long exact sequence . (UnV)' > WI a W2 fJ PROOF. V open in M. If M = long exact sequence .
so A(W) . then l[ 0 j is the identity of 1\'. equivalenl if there is f > i." classes into a Vector space f'/ (N). This makes sense. We therefore have the sequencL C:(M . This defines X on fi( W2 ) C W3.k ( N ) . To circumvent this difficulty. for if fi (W' ) = fi(w"). • • with ni Vi = 1\'. vVe now consider a rather different situation. This sequence is 1lot exact at C! (M): the kernel of i" contains all W E e: (M) which are 0 on N. + j" g. Finally. while the image of e contains all W E e: (M) which are 0 in a neighborhood of N. j such that Wj E Ck(Vj). the "germs of kfonns in a neighborhood of M". and define X to be 0 on W.A(W") = AO'(=) = O.w" = 0'(=) for some z. '>\'e now construct l1 sequence of such neighborhoods V = VI :::J V2 :::J V3 :::J . vVc appeal first to a result fJ"Om the Addendum to Chapter 9. and if j : N > V is the inclusion. Let N C M be a compact sub manifold of M. 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. then w . /I' + e C: (M) + i* Ck (N). we define a map of complex".. it is eas\" to define d : g.432 of the form Chapter Jl fi (w').N) where e is "extension". while j 0 lf is smoothly homotopic to the identity of V. Now consider two forms Wi E Ck(Vi ). Now chooSt W e W) with W3 = fi( W2) Ell W. •:. Moreover. 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.k (N) > gk+ I (N). C: (M) in the obvious way: (J) � the equivalence class o r any w i lli . Then M N is also a manifold. we define �(W) = AW). sO that we obtain a complex g.
Clearly e is oneone. (M) + i* fJ.k (N)) PROOF. This follows easily from the fact that j* : Hk ( Vi) isomorphism for each Vi . Then there is an exact sequence + o H. > H. + e C. this means that A Wi = 0 for some i . This means that i*e(w) = O. Hk (fJ. (M . compact and n. (M).k (N).N has compact support C M . which we will need later.f�).k (N) > 0 I I . and consequently W = 0 on Vi . •:. The sequence 433 0 > C. THEOREM (THE EXACT SEQ UENCE OF A PAIR). and then use Lemma I I . Hence AIM . Conversely.N) > . THEOREM. 1 3 . then there is an exact sequence N c . (M .N). •:. any element of fJ. .k (N) as 1 ] .k (N) is represented by a form � on some V. 1] be a Coo function which is 1 on v. and A = e(AIM . ! .N) > H.ExcU):rion in the Realm ifAlgebraic Topolog). . PROOF.) of the complex {fJ.N) is exact. 1 0. Apply Theorem 2 to the exact sequence of complexes given by Lemma In the proof of this theorem. But in the next theorem. Since N i.N). Details are left to the reader. •:. suppose A E C/ (M) satisfies i*(A) = O.N. consequently this element is i*(. Let M be a manifoldwithboundary. The cohomology vector spaces are isomorphic to Hk (N) for all k. PROOF. the de Rham cohomology of the manifoldwith boundary V. then W '= 0 in some neighborhood U of N. entered only as an intermediary (and we could have replaced the Vi by their interiors). having support I C + interior Vi . Let I : M > [0. +I (M . Vi = N. (M) > Hk (N) 10. . LEMMA. Then l� E C. with compa�t bound ary aM. it is the object of primary interest . and l� represents the same element of fJ. > Hk (N) is an M is a 1 2 . LEMMA. If W E C. If compact submanifold of M. B y definition o f fJ. there is some i such that Vi C U. Finally. (M .
n+1 . Then Theo rem 13 may be used to compute the cohomology of S" X sm . M is the boundary of each component. PROOF. I\1oreover.'(�"+ I ) � H"(M) � H.M has exactly 2 components. Using Theorem 817.M = dim H" (M) + 1 . For our next application we wiIJ seek bigger game. and ffi. •:. by choosing M to be the closed ball B in �". of aM in M. U R n H.M) � H:+I (�"+I ) � H"+I (M). using tubular neighborhoods V. dim H"(M) + 1 ? 2 . then M is ori entable. with H. M) gives H.I ) .I = a(S" x closed ball in �m). . Just like the proof of Theorem 1 2. we can rederive H. (�") from a knowledge of Hk (S" . ( B) "" H k ( B) = 0 for k '" o. by considering the pair (S" x �m. {p) X �m).M ? 2 . The reader may use Theorem 12 to compute O R O It follows thaI (*) number of components of �"+I .'+I (�"+I . the sequence of the pair (�"+ I . From (*) and (**) we obtain 14.434 C hapter J J PROOF. pact "hypersurface" of �"+I). Let M C �"+ I be a compact ndimensional submanifold of �n+ 1 (a com As a simple application of Theorem 1 3 . THEOREM. But we also know (Problem 825) thaI (**) number of components of�"+ 1 . (S" x �m). If M c �"+I is a compact hypersurface.
The proof in Problem 825 shows that every point of M is arbitrarily close to points in different components of jRn+1 . If the radial function p of a starshaped open set U C jRn is Coo. Consider a bounded open set U C jRn which is starshaped with respect to o. . 1 5. This produces difficulties at 0.Excunillll ill the Realm ifAlgelna. The basic idea of the proof is to take Ix E B to p(X)1 .. then H shows that jRn+1 . so every point of M is in the boundal"\' of each of the h¥O components. then we can prove that U is diff eomorphic to the open ball B of radius 1 in jRn.I > JR . x E U. which we will have to dispose of first. so a modification is necessary.t 7ojJO/ogy 435 Since dim Hn(M) is either 0 or 1. •:. However. then jRn+1 . 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. then U is diffeomorphic to the open baJJ B of radius 1 in jRn. so M is orientable. COROLLARY (GENERALIZED [COOl JORDAN CURVE THEOREM). we conclude that dim Hn (M) = 1. We will call p the radial function of U. 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.M has two components.M has exactly two components. 1 7.. 16. and M is the boundary of each. If p is Coo. COROLLARY. LEMMA.M. If M C jRn+1 is a submanifold homeomorphic to sn. there are a number of technicalities involved.
there is an open ball B with x E B C U. I < e. Since U is open. / : [0. ° /(J) I . LEMMA.436 eha/ller J J PROOF. with a nonzero Jacobian. > U by x E sn. At each point x E sn.I . . it might not even be continuous. This means that for . •:.1 )/(I)]X. so it is Coo. = Clearly h is a oneone map of B onto U. the radial function p of a starshaped open set V C ffi. Choose IX E U with p(x ) PROOF. that p [0. and hence in U.11 is "lower semicontinuous": for every E: > 0 there is a neigh borhood W of x in sn.1' E W the point IJ' is in B.e for all Y E W. At any other point the same conclusion follows from the fact that I r+ 1 + (p(x) . 1] be a Coo function with = Define h : B / ° in a neighborhood of ° /' 2':. In general.J' E W we have p (y) 2':. However� the discontinuities of p can be of a certain form only. •:.I . 1 8 . 1] > We can assume. I t is the identity i n a neighborhood of 0.I such that p()') > p(x) . I Coo. There is clearly a neighborhood W of x with the property that for . I > p(x) . without loss of generality. at 0.e . (he function p need 110t be h(lx) = [I + (p(x) .I . 1L 0 ::: 2':. Let I < 1. 1 on sn.1)/(1) is a Coo function with strictly positive derivative.
Oll . . Since K is closed and p is lower semi K C V = {IX : x E SnI and 0 :s I < p(x)}. . it follows that p(x) < p(x). p. it looks as if U should be diffeomorphic to IR". .I such that Ix may also be used as 1. then Hk (U) "" H k (IR") and H. •:. Then PI+dx). . . The prooffor Hk is clear. For each x E sn. . Proving this turns out to be quite a f eat. (U ) "" IR "" H.1' for all Y E W. let 'PI . Since 1>dx) + . . . . (IRn) for all k. (U) "" H. Let IVx" . + 1>1 (x) = 1. lxl is < p(x). say WX1 " ' " Wx/ for convenience..EXCU1J. By Theorem 817. and we will be content with proving the following. be a partition of unity subordinate to this cover. K C V. (U) = 0 for O :S k < Il. Each Ix" . PROOF. continuous. . If U is an open starshaped set in IRn. LEMMA. and define Any point X E snI is in a ceI"tain sllbcollection of the J..I . . 1>. . . . We also know that H. (lRn). IVx. cover S.1lx. there is a neighborhood Wx of x in sn. Similarly. 437 Even when p is discontinuous. and consequently w = d� where � has compact support contained in V. .I . since U is smoothly contractible to a point. we just have to show that H. We claim that there is a Coo function p : snI > IR such that p < p and This will prove the Lemma. 19. . . for then V is diffeomorphic to IRn . and hence in U. .11 the Realm ifAlgebmic ToJ)olog). Let w be a closed kform with compact support K C U. (x) are o. 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.
Suppose i t i s true for a certain 1'. 3. . According to Problem 932. In general. . and consider a nice cover {V" . . (M) are finite dimensional for all k . n· . 2. and . . Ii follows from Lemma 19 that V has the same Hk and H: as IRn. where V is a disjoint union of balls or around 1 . V. . H' (M) (fJ H' ( V) II 0 � H ' (M n V) � H2(1R2). We obtain H' (1R2) II 0 � . . If p E V. and choose a Riemannian metric for M. . this shows Ihat H' (M) is infinite dimensional (see Problem 7 for more informalion abOUI the cohomology of M). If any V = Vi. If M has finite type. is nonempty. II 0 where M n V has the same H' as a disjoint union of infinitely many copies of S ' .. such a cover will be called nice. /II is not of finite type. 20. a manifold M will be called of finite type if there is a finite cover {V" . . . } as a subset of 1R2. n Vi. . V" V) of M. 3. then expp establishes a diffeomorphism of V with an open starshaped set in Mp . V. . Let M be a compaci manifold. . } be a finite cover by such open sets. clear for PROOF B v induction o n the number o f open sets r in a nicc cover.438 Chapter II We can applv this last Lemma in the following way. . every point has a neighborhood V which is geodesically convex. Then the theorem is true for V = V. then V is clearly geodesically convex. then M = 1R2 .} such that each nonempty intersection has the same Hk and H: as IRn. It is fairlv clear that if we consider /II = { I .:. 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. . Ii i s r = 1. . To prove this rigorously. 2. U · · · U V. . PROPOSITION. . Let {V" . we first use the MayerVielDris sequence f 1R2 = M U V. then Hk (M) and H. On the other hand.
(M) + /J. One of the fundamental theorems of manifold theory states that P is always D an isomorphism. So Hk (M) must be finite dimensional.k (M)* by f3 r+ 1 (M. since a! 1\ � has compact support if � does. .) � = J.(M) represented by any � E C. since this has the nice cover {V n V. Now suppose that Mn is connected and oriented. in order not to plague ourselves with additional technical details. /J. by the same f ormula. As with most big theorems of algebraic topology. There is then a unique element of H. M we can define (see Problem 831) the cup product map Hk (M) x HI (M) "::' Hk+/ (M) r+ by ( [a!].. Now every a E Hk (M) determines an element of the dual space H. > Hk. The proof f H. but we shall restrict the theorem to manif olds of finite type. the main part of the proof is called a Lemma. Now consider the MayerVietoris sequence . [�]) We can also define [a! I\ �].ExcuIJion in the Realm ifAlgebraic 7ojJolog} 439 for U. We are all set up to prove this fact. It is also true for V n V. •:. . (M) and the isomorphism H�'(M) > IR which takes this element to I E 1R. . (M) with It is convenient to also use /J. with orientation IJ. and the theorem itself is a simple corollar\'. . V n V. (M) is similar. to denote both this element of H. and the kernel of a is also finite dimensional. or For any manif old + a Hk (V) (fJ Hk ( V) > ' " . (a v f3). 1R. . . . a We denote this element of that we have a map H. the "Poincare dual" of a. so P D(a) (f3) = v f3 E H.p.l (V n V) Ii H k (M) + The map a maps Hk (M) onto a finite dimensional vector space. .k (M)* by PD(a) . ] .
+I(U) H.440 C1zapter J J H'IW) H'I(V) H'IW V) H'(M) H'(U) H'(V) HkW V) j PD j PDEIlPD 1PD 1PDEIlPD j PD [H. It is not hard to check (Problem 8) that every square in this diagram commutes up to sign� so that by changing �ome of the vertical isomorphism� to their negatives: we obtain a commutative diagram.<U v)· Ell � � Let I = 11 . 1>4 3 = "3 (X) = 0. = = 1>dw) w E VI . except possibly the middle one. 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. and is left original Lemma . This proves the . ffi n � ffi n 1>1 . 1>5 Supposc f some or Then (x) so is an isomorphism. o. Hence "THE FIVE LEMMA". in which the top row is the MayerVietoris sequence. aU vertical maps. for all k on U. and the bottom row is the dual of the Mayer Vietoris sequence for compact supports. PD V PD � � Ell � � n By as�umption. 1>2. x E V3. and U n V.H. V. 1>4.(M)· [H. and that are isomorphisms. LEMMA.+I(U V)+ H. So is oneonc.<VW H.+I(Vn. Vve now forget all about our manifold and use a purely algebraic resull. are isomor phisms. : for some Then which implies that y = "dw)." y E 1>2(Y)V2 f3d:) x = E WI . 1>4 0 = 1>3(X) = 1>3"2(Y) = f321>2 V3. Consider the following commutative diagram of vec tor spaces and linear maps."2(Y) . 1>3 the reader. If M = U U for open sets U and and is also an isomorphism for all k on M. Suppose that the rows are exact.k. •:. Thc proof that is onto is similar. Then is also an isomorphism. n � is an isomorphism 2 1 .(U) H. then V PROOF. 1>3(X) = 0 f331>3 (x) = 0. Consider the following diagram. 1>3 x= "2(Y) = "2 "dw)) = ( (Q o. By exactness at there i. Hence since Thus (Y) · Hence with for some Z Moreover.
•:. We begin by considering a smooth kdimensional orientabie vector bundle S = 1f : E > M over M. . the terms (l)k dim Hk(M) and ( _ I )"k dim Hnk (M) cancel in pairs. 23. THEOREM (THE POINCARE DUALITY THEOREM). Suppose it is true for a certain r. noting that V' is isomorphic to V if V is finite dimensional ./II) and H� (IR2 . since E is locally a product. 441 22. The theorem is true for V. Problem 7 shows that H' (IR2 . then X(M) = O. 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". PROOF Use the Theorem and Proposition 19. then a slight modification of the proof for Lemma 1 9 . . •:. f M and v f S give an or or orientation /J. COROLLARY. . Vr } is a nice cover of M by geodesically convex sets so small that each bundle S I Vi is trivial. U · · · U Vr . . Vr ./II) have different (infinite) dimensions. If M is a connected oriented nmanifold of finite type. This completes the induction step. If { V" . 25.ExeulJion in Ilze Realm ifAlgebraic Topolog). Corollary 23 does not. By the Lemma. •: .k ( M) ' is an isomorphism f all k . then the map PD : Hk (M) > H. In fact. If M is a connected oriented nmanif old of finite type. Ell v for the (n + k )manif old E. V} of M. or PROOF B y induction on the number r o f open sets in a nice cover o f M . 24. . The theorem is clearly true for r = 1 . If M is a compact orientable odddimensional manifold.k (M) have the same dimension. If M is a compact connected orientable nmanifold. . COROLLARY. Let V = V. PROOF In the expression for X(M). and consider a nice cover {V" . and for V n V (as in the proof of Proposition 19). V. it is true for M. Even though the Poincare Duality Theorem holds for manifolds which are not of finite type. Orientations /J. then Hk (M) and Hnk (M) have the same dimension. then H k (M) and H.
On the other hand. (This condition means that f where U is the class of the closed form w. 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. Pick some closed form w E c� (E) representing U.vp) jp'w = l . . Our definition of U states that ( I) Let A C M be an open set which i s diffeomorphic to IRn. This class U is called the Thorn class of $. Notice also that for the maps s = Osection E. Also choose A so that there is an equivalence (Fp.' (Ur)} is a nice cover of finite type. . 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. and $ = 1f : E > M an oriented kplane bundle over M with orientation v.). 1f . 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. p. 26. Let (M. the orientation v for $ determines an orientation vp for Fp. ) � = 1. L 1f'� l\ w = 1. /L) be a compact connected oriented manifold. and let jp : Fp > E be the inclusion map. and let � E Cn(M) be a form representing /L. Let Fp = 1f' (p) be the fibre of $ over any point p E M. Our first goal will be to find a simpler property to characterize U. and hence an element vp E H�(Fp). so that J( M. . . Since jp is proper. there is an element jp' U E H� (Fp).442 Chapter J J shows that {1f ' (U. . so that A i s smoothly contractible to any point p E A . THEOREM.) PROOF.
Under this identifica y tion.e) : q E A . under the identification of 1f . 1] > A x Fp such that H(e. or which we will continue to denote by jp. the support of wl1f1 (A) is contained in {(q. Consequently. We will also use 1f2 : A x Fp > Fp to denote projection on the second factor. By choosing a smaller A if necessary. I) 'f.I (A) with A x Fp. the map jp : Fp > 1f1 (A) corresponds to the map e r+ (p. support w if lie II ? K . A glance a t the definition o f J (page 224) shows that the . For the H constructed in this way it follows that m e. we can assume that there is some K > 0 such that. 1) = (p. 1f2 (e)) = }i> (1f2 (e)). e) f e E Fp. Let II II be a norm on Fp. O) = e me. lleII < K ) .EtC/miDlI in lize Realm ifAl gebraic ToJ)ology 443 This equivalence allows us to identif 1f1 (A) with A x Fp. wejust pull the fibres along the smooth homotopy which makes A contractible to e. support w Using the fact that A is smoothly contractible to p. I) : lIell < K}. the form H ' w o n (A x Fp) x [0. e. it is easy to see that there is a smooth homotopy H : (A x Fp) x [0. 1] has support contained i n {(q.
Now. Thm (2 ) So support A c {(q.io*(H*w) ° = = d(lH*w) + I(dH*w) d(lH*w). Since we have where Jr* /L 1\ A has compact support on A' x Fp by (2) = = ± 0 .  I ) dimen .444 Chapter II lIell < K}. on the one hand we have (Problem 817) (4) On the other hand. (4).*(H*w) . Combining (3). " learly suffices to prove that the integral is 0 over A' x Fp for any closed ball A' C A. Theo form IH*w on A x F has support contained in {(q. To prove this. [ 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). e) p rem 714 shows that (jp Jr2 ) 'W . (5) we See tha.w = i. we claim that the last integral in (3) is O. i. JaA1XF/. e) : lIell < K J .
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.
then it holds for M. . "3 . (It will be necessary to recall how various maps are defined. V u.E. Vo is u nion of shaded VI is union of un shaded C�·. "2. and U2 . . (Mi) and all but finitely many "i = 0 E H.l l . xcwsioll ill lize Realm ifA lgebraic Tapolog). this "direct product" consisting of ali sequences ("1 . (Mi). (M) "" EEli H. (c) Show that if the Poincare duality theorem holds for each Mi. . except for the square Hk. ( Mi).) 9. which is a good exercise. Show that H. ) with "i E Hk (Mi). VI. (a) Let M = MI U M2 U M3 U · · be a disjoint union of oriented IImanif olds. the only slightly difficult maps are the ones involved in the above diagram. this "direct sum" consisting of all sequences (" I . Show that the squares in the diagram in the proof of Lemma 21 commute. "3. (b) Show that Hk (M) "" n i Hk(Mi).I (u n V) H�+I (U n V)' lPD > Hk (M) > H� (M)' lPD which commutes up to the sign ( . (d) The figure below shows a decomposition of a triangulated 2manifold into three open sets Vo. " ) with "i E H. 455 8. is union ofshaded & . . "2 . Use an analogous decomposition in n dimen sions to prove that Poincare duality holds for any triangulated manifold.
Join these by a ray going to infinity. (E) and U' E H.) 1 I ..X at p is ( _ I )n times the index of X at p. (c) it is a fact that a Coo map f : sn. then there is a . (d) If M is connected and not compact. : TM > M fOl' M of odd dimension. then there is a vector field X on M with only one singu larity.k (E). with nov\. (E') are the Thorn classes. such that all Pi E interior D. (a) Let P I . (Using the notation of Problem 323. providing anotllCl proof that X(M) = 0 in this case. (a) If U E H. and push everything off to infinity. : E > M and r = . show that this is 0 for k odd. p. (Begin with a triangulation to obtain a vector field with a discrete set of zeros. Using this. X(j*(�))·) (e) If M is a connected manifoldwithboundary. we can form U v U E H. Conclude that U represents 0 E Hk (E). then there is a nowhere 0 vector field on M. (a) Let � = . . Let � = . thaI X(�) = 0 when � = . we have f*(x(�)) = 12. show that the index of .456 Chapter JJ 10.I > snI of degree 0 is smoothly ho motopic to a constant map. Q») Since we can also consider U as being in Hk (E). : E > M be an oriented kplane bundle over an oriented manifold M. then j* ( U) = U' . in particular. This p rovides another proof that X(M) = 0 for odd 11. Using anticommutativity of /\.here zero vector field on M. 1 3. (b) . show that there is a subsel D C M difTeomorphic to the closed ball.. If a vector field X has an isolated singulaI"ity at p E Mn . then there is a nowhere 0 vector field on M. It follows. with Thorn class U. prove the Thorn Isomorphism Theorem: The map HI (E) > H:+k (E) given by " r+ " V U is an isomorphism for all I. show that if X(M) = 0. and (j.. E M. 0» If M is compact. ' : E' > M be oriented kplane bundles. over a compact oriented manifold M. so that X(�) = O. f) a bundle map from �' to I: which is an isomorphism on each fibre. Using Problem 826. enclose this ray in a cone. . aM '" 0. . .. Using Poincare duality.f*(X(n) = xW).
This Problem proves de Rham's Theorem. and U n V. while i 0 r is chain homotopic to the identity of Sd M) [basically. . Basic knowledge of singular cohomology is required.) (e) Check that the cup product defined using 1\ corresponds to the cup product defined in singular cohomology. (a) If w is a closed kform on M. and 19 will not be necessary for this.EXCU1JiOl1 in tlze Realm ifA( �ebmic TopologJ' 457 1 4. let Rh(w) E Hom(S'k ( M). 1 8. then it is an isomorphism for U U V. to show that if Rh is an isomorphism for U. using the MayerVietoris sequence for singular cohomology. it follows that Rll is an isomorphism f any triangulated or manifold.) (c) Imitate the proof of Theorem 2 1. For a manifold M. (d) Conclude that Rh is an isomorphism if M is of finite type. (Using the method of Problem 9. (b) Show that Rh is an isomorphism on a smoothly contractible manifold (Lem mas 17. 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). 1R). 1 w. We will denote the group of singular kchains of X by SdX). and let i : S'k(M) > Sd M) be the inclusion. IR) be Rh(w)(c) = Show that Rh i s a chain map from { Ck (M)} t o {Hom(S'k ( M). IR)}. V. This means that we obtain the correct singular cohomology of M if we consider the complex Hom(S'k (M). (Hillt: Stokes' Theorem.) It follows that there is an induced map Rh from the de Rham cohomology of M to the singular cohomology of M. r is approximation by a Coo chain]. we let S'k( M) denote the Coo singular kchains.
.
a Haus dorfflocally compact space is regular.iust like neighborhoods of o.) 459 U n JR is open. this argument does not work f "infinite dimensional" manifolds. THEOREM. (By the way. Although nonHausdorff manifold. We bave just seen that the Hausdorff property is not a "local property". a compact manifold is metrizable. (2) For each x E M there is a neighborhood such that U is homeomorphic to lR:". JR. (c) M is metrizable. but local compactness is. . 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. which are lo or cally like Banach spaces. these need not be regular even if they are Hausdorff) On the other hand. Before exhibiting nonmetrizable manifolds. (I) U of x and an integer n Condition (I) is necessary. we first note .APPENDIX A CHAPTER I Following the suggestions in this chapter. The following properties are equivalent for any manifold M: (a) Each component of M is a compact. (b) Each component of M is second countable (has a countable base for the topology). Ev ery manifold is also clearly locally connected. that almost all "nice': properties of a � anifold are equivalent. then (U n JR) U {OJ is a neighborhood of 0 (in JR). we will now define a manifold to be a topological space M such that (I) M is Hausdorff. there are manifolds which are not normal (Pt·oblem 6). (d) M is paracompact. so every manifold is locally compact. for there is even a I dimensional "manifold" which is not Hausdorff. and thus a manifold itself. we will not consider them. (In particular. with the following topology: A set U is open if and only if (2) If :>: 0 Thus the neighborhoods of * look . I t consists of JR U {*} where * 'f. are important in certain cases. so every component is open. :Moreover. * E U. so every manifold is regular.
If Vo is one ofthe�e. Let M = C. U . ha. Un':!. where each C. for if x is in the closure. It lurWi out thaI there are even l manifolds "dlich are not paracompact. which are briefly explained here. . then x must be in the closure of a finite union of these Vi. Similarly Do U VI U . an open neighborhood VI "·. = Uo U .460 Appendix A FIRST PROOF. paracompact space is a compact. is compact. (Ord1nal numbers will nol be needed for a 2d1111ensional ex ample 10 come later. U V". l. : and so On.. II is easy to show iiom th1s that M is paracompact. . Clearly C. (cJ =} (d) because anv metric space is paracompact (Kelley. U · . U C. ./. because x has a neighborhood \vhich intersects only finitely man}: Thus x is 1n the union. locally compact. U .) oflhe others. c E A (tl"ansitivit\·. The serond proof does not rei v on this difficult theorem. (b) =} (c) follows (rom the Urvsohn metrization theorem. . . U . . U . · . is (a) is Theorem 12.'hos('" union contains all Cj. . (a I =} (b) follows immediateh' fi·om the simple proposition that a acompact locally second countable space is second countable. A connected. pg. U Un .s countable union of C0111pact set�. . U [in . is clearly open. then Vo can intersect onlya finite number lJ1 • . Since the space 1S connected. . This pro\'es the Lemma and the Theorem. we obta1n open sets Vi with Vi compact and Vi C Vi+ l . . 160)..:. \". b . . \l\'1th compact closure_ Continu1ng in th1s way. Gellem/ 7ojJo/agT. and hence =} SECOND PROOF (0) (a) =} (b) =} (c) and (d) =} (a) as before. (a) =} (d). The unioll Va U " . The cOllSl ruct10n of J11Cse eXalnples requires the ordinal numbers. Then VI U C2 has an open nl'i�hborhood U:!.. (d) =} (a) is a consequence of the following. ' . PlOOf.ith compact closure. ' U U". LEMMA. There is a locall\' finite COver of the space by open sets with compact clo sure. it equals thi. 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 . . U V'I I imersects only Un l+1 • . . M. It is also dosed. . or .
. .· <w·2 « ··<w·3 < . and II is easily seen to be an order isomorphism also. .· <w<w+l <w+2< . .firs/ clement. one and onlv One of the following holds: (trichotomy).··<w·2 <w·2+ 1 < . for the prese]]!. < w . that is. O < 1 < 'J < · · . <) 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. = (O. an element b such that b ::5 b' for all b' E B. An orderinf! < On A lS a wellordering jf every nonelnpty subset B C A has a. < w < w + l < w + 2 < · · · < w · ] < w <w+1 <w+2 < . < w . 2)\ 0 < 1 <2<3< . . just a set distinct from those already mentionedl 0 < 1 < 2 < . 'J < . o< 1 < ' " < w 1. 3 . < w < . .Appendi. 0<1 <2< 0 < 1 < 2 < · · · < w < w + l <w+2 < ·· · < w · 2 < w · 2+ 1 < ··· < w O< 1 <2 < ··· < w < . I ) ) CO. .. 1 . . . O < l < '2 < · · · < w 0 < 1 <2 < ..I' 0. .. ) (w + I is. <) where < is an ordering on A . (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. 3 < . b 461 E A.· · < w <w+1 (w is some set . < . <) and (B. 2. > An ordered set is just a pair (A .· < · '] < .J (2) For all a. . .. 1 . Some wellordered sets are illustrated below: in this scheme we do not list any of the < relations which are consequences of the ones ab'eady 1isted. Two ordered sets (A...
the only (wder isomorphism f)'om B to a segmeni of A is 1111 ldentit:: > f(h) = b. <) and (B. ..B. It is easy to show that any tWo such order isomorphisms agrC'(. For example. then one i. consider order isomorphisms from segments of A onto segments of B. If (A. <w and 0 < I < . then B is not order iSOlnol· phl( to A.. then either B = A or else there 1S some a E A such that In fact. on the smaller of their two domains (just cOllsider the smaJJesl element \... It is easy to see that if B is " seglncm of A . But there lS a much more general propOSltlOll which will settle all cases at once: J. of course. And so Oll . PROPOSI TION. If f : B B' C A is all order isomorphism and B' is a segment of A . Notice that each set on OUf list 1S <1 . the "wth H . . <w <w+ I arc nOl order isomorphic because the second has both a last and a next to la�l eien1cnt. the second with the second. If B '" A is a segment of A. TI. which makes the study of wellordered set' simply delightful.c.462 AfJPclldix A Any �ub�et of a wellordered set is. a i s the first element of A . 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. also a wenordered set with tbt same ordering. etc. It 1S not hard to sec that no two sets on our lis1 are order isomorphic. . We match the first element of A with the first of B. while the first does not.rith the "w th". < ) are wellordered sets. etc. B = {a' E A : a' < a J : o < I < .en feb') must be b'. until we run out of one .c?ment of the succeeding ones. Proposition I has a com pall ion. •:. PROOF. order isomorphic to a segment of the other.. . ('ven for the "wth" eienlellt (the first one after the first: second: third..srI. where b' is the second element.. just consider the first element of {b E B : feb) '" b): an outright contradiction appears almost immediately.)! The way we prove this ligorouslv is amazingly simple: If feb) '" b for some b E B.. then for the firs! clement b of B (and hence of A) we clearly mus! haw 2 .. . PROPOSITION.hen' PROOF. In particular. To do Ihls rigorously. In fact..
...+ I the smallest ordinal after " (if " is represented by the wellordered set ( A .. I\otice that if ex is an o)dinal number.Appendix A 463 thev don't). If " is an ordinal numbe. v·:hkh can be f ound in (he Appendix to Kelley. <)". instead of '''lith the wellordered sets themselves.A. we will denote by .'ellordered set with just one more element.gnore elements 2:. represented by a wellordered set (A .A. Consider the least of these a's. . then the wellordered set of all ordinals fJ < " has a particularly simpk representation: it is order isomorphic to the set (A . larger than all member!' of A). <) order isomorphic to (B. ordi nal !lumbers. 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. •:. <). which is c1earlv the largest of all.ularly elegant way to do this. <) when (A . To produce a smallest element of . These older lsomorphism classes arc called ordinal numbers. This fJ is the smallest element of . PROPOSITION. <r! must be replaced by "(A . <) is order isomorphic to a proper segment 01 (B. So all such order isomorphisms can be put together to give another.A of ordinal numbers. Each ordinal number ls a horribly large set. 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 . There is a panic.oll1orphism class. due to von Neu mann .A \'\'C can obviously i. and Propositions I and 2 show that we al most have trichotomy. Notice that some ordinals are not of the form " + I for any ". < ) ! Roughly speaking: An ordinal number is order isomorphic to the set of an ordinals less than 1t. <). while those of the form " + I are called Successor * Only one feature mars the beauty of the ordinal numbers as presented here.'enordered sets. < is a wellordering of the ordinal numbers.rl from each ordrr l:. let ( A . <) = (B. Given a nonemptv set . <) be a well ordered set representing one of its elements ex. To obviate this difficulty V'!e' need only '''lark with order isomorphism classes of . .. Transitivity of < is obvious. If it is defined on all of A we are done. c¥. General Topolog). ''Almost'': because the condition H(A. and dcfine thesr speclfic sets to be thr PROOF. <). <) < (B. They are beautiful:' 3. Suppose wc define a relation < between wellordered sets by stipulating thaI (A . these are called limit ordinals. It determines a segmen t wh1ch represents somt' fJ E .:. then " + I is represented by a \". it would be much !llcer to choose one specific well o)'dered :.
For . any member of SI is countable. If it were countable.464 Appendix A ordinals. 2. would appear (svmbols like w3 + w2 . . 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. We have thus established the existence of an uncountable ordinal. Then SI is uncountable. w . is clearly the first uncountable ordinal. 3 . •:. fO I· 011e would have to go uncountably far.tI. There are indeed an enormous nUlnber of countable v·. . OUl· list of wellordered sets onlv begins to suggest the complexity which well ordered Sets can achieve. i.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. . Our sv example. one can see how the symbol> w3. and henct' is order isomorphic to an initial s<'glllt'nt of ()."ts llWlll loned so fal" are countable. etc. <) is a wellordered set. We will also denote some ordinals by the symbols appearing before: 0. . this ". . . By the remark after Proposition 3. . A leap of faith is required. ( SI . After one comes to and this 1S only the beginnln�� A ll the wellordered Sf.e. (h is hopeless to tl'Y 10 "reach" n by C'o 'ltinuing the llsting of well ordered �ets begun above. see Kelley's Appendix.. il would represent a countable ordinal " E SI. PROPOSITION. 1 . w' . w + 1.wuld mean that n is order i�omorphi(' to the C'oJknion of ordinals < Q . . re presented by n. a countable wellordered set). 4. to a proper segment o f itsdL contradicting* Prop osition J . so it represents an ordinal ex E (htl. and ell counter �rts with an ullcountable nmnber of degree� of complexity. With a little thought.I resohllioll of this paradox.) Although the countable ordinals exhibit ul1countably many degrees of COIll pkxit�: they are each simple in one way: rihe PROOF By Proposition 3. . Let SI bc the collection of aJ! countable ordinals (ordinal. and consequentlv has onlv countably many pre decessors. .
0) and a could bv represented by a subset of � . PROPOSITION. 11S1 \vhich comes after fill ' (. 0)). . and L + = n x [0. So by Proposition 5. and we can easily arrange that the subset representing f3i is a segment of the subset representing f3j for i < j. < a.g. with the order < defined a s follows: (a. thus also by a subset of (00. Then f3. ).0) .0) )' (2. aside f)'om the line and the circ1e� there are 110 other connected Imanifolds. < a cofinal in a .0) . •:. Since a is countable. O)} is the (open) long ray. 1) . This is impossible . Suppose there were one. namely those bel\'I'eCll the points representln. 1) with the order topology (a subbase consists of sets of the form {s : X < xo l and (x : x > .O) (W+I..AJ)pelldix A 465 5. Howevel. . . < f3) < . . < f32 < f33 < . there is a sequence f3 . This can be pictured as follows: (0.{(O. i). ) tw·2. the names "halflong linel ' alld "long line" may also be used. a contradiction. (1. If a E n is a limit ordinal. If a subset of � were order isomorphic to n. 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.Q and a + 1 for all a E n. . . ). To disllnguish L + and L. and hence a smallest.0) (w. It cannot happen that a = f3 + 1. 1 ) .o)  The set n x [0. Consider n x [0. PROOF. COROLLARY. . PROOF. s ) < (f3 . for then f3 would be represented by a subset of �.O) (w+2. The disjoint union of two copies of the closed long ray with their origins identified is the long line L. YJ .\'0 ) ) is called the closed long ray (with "origin" (0. no subset of � is order isomorphic to n . such that every f3 < a satisfies f3 < f3n for some 11 (we sa�' that (f3nl is "cofinal" in a). is represented by a subset of ( 00. If a E n. then a is represented bv some wellordered subset of �. a E n not represented bv some subset of � . The first example of a nonmtU'izable manifold is defined in terms of n. 6. then there is a sequence f3 t < f3. The union of all these sets would then represent ex. The Corollarv to Proposition 5 implies easilv' that the long ray and the long line are I dimensional manifolds. then there would be uncoul1\ablv mallY disjoint intervals in � . I ) if a < f3 or if a = f3 and s < I .
Consider the disjoint union of �� and ��. L x S ' (long cylinder). 8 ) in the product of the closed long ray and S' produces another 2manifold. for each a E �. (big plancj. Y ) o ) = o« (xy. y)o. + to JR2 . + But consider now. the following diagram shows two open sets homeomorphic to �2. homeomorphic �F='. We begin with the open upper halfplane �� = {(x. y).\". y). which might be called the "big disc". we will denote this set by �6. Define a map /0 : (��) > �� b. 0) by (x. Deline j� : (�.) > �� b�' + j� «X. y) E �2 : r > OJ and another copy �2 x {OJ of the plane." � /_. in fact. Identifying all points « 0.' f . with P E (��) and f (p) E JR � o + identified. y.:.\' . we could have thro\vn away �� to begin with slnce it is ldcntified by a homeomorphism with (��) . say �2 x {a}. There is another way of producing a nonmcuizable 2manifold which doe� not use n at all. and denote the point (x. (halflong strip..)')a) = (a + y. (big quadrant). L x IR (long strip)." '" + q�. The manifold itself is. 0)..:. which we will denote by �. . another copy of �2. L x L+ (big halfplane). This is a Hausdorff manifold.466 Appendix A Quite a few new 2manifolds can now be constructed: L+ X S ' L + x IR LxL L+ x L+ (halflong cvlinder).
developed in the problems. for it has an uncountable discrete subset.Y)a and (X'. (a) L + and L are not metrizable. and related man]folds� have some very strange properties.e. Consequently. This space is still a Hausdorff manifold. show that one of them is equivalent to (can be put in oneone correspondence with) a subset of the other.y')b for which y = y' > 0 and . (b) ]( we denote (a + 1 ) + 1 by a + 2.\. a E IR we wish to identify each p E (��) + with f a(P) E ��.. are a space homeomorphic to ��� so we will consider �� a subset of the resulting space. but it cannot be second countable. a wellordering < on a subset Y of X will be called "distinguished" if for all Y E Y. any sequence has a convergent subsequence (but L + i� not compact!). The equivalence c1asses� of course.\3 :s n polnt.y+a = x'.ApjJendix A 467 In the disjoint union of �� and all ��.. one is an extension of the oth. then {x } converges to some (b) If Xl � X2 :s: .l'+b. y = e(Y . and in the disjoint union of all �� identify each (x. (Zorn's Lemma may be deduced from this fact fairly easily. then any a equals f3 + II for a unique limit ordinal f3 and integer II � O. (a + 2) + 1 by a + 3. (d) Show that on any infinite set th"'e is a wellordering which represents a limit ordinal. 3. namely the set { (O. is a sequence in L +. (Thus one can define even and odd ordinals. PROBLEMS (a) A wenordered set cannot contain a decreasing infinite sequence X l > X3 > . c(A ) is defined for each set A '" 0. Given a set X. and Problem I. We mav dispense with �� completely. i. O)a ] . Let c be a "choice function". show that if X and Y are disjoint equivalent infinite sets.{y' E Y : / < y}). This manifold. (e) From (d).: (b) Show that there is a wellordering on X. then X U Y 1S equivalent to 1'. etc. .) L X2 > 2. the Priifer manifold.) (c) Given two sets. and c(A ) E A for all A.. (a) Show that of any two distingulshed wellorderings.
cations of L x L. and r > s. + (and also L) are normal. then one of the sets f' « oo. n . any order topology is normal (completely different proof!. X Sl ) = z.s]) and . (a) Show that the Proler manifold P is Hausdorff (b) P does not have a countable dense subset. L + x L +. b) x (0. Show that U include. then f is eventually constant. 4. 6. L+ x S l and L x S' are nOl homeomorphic. L + x L" . " paracompact ends'·.. Sillee IR: = Q U Un A . I] L + with H (x. defin. (Use (c)). a whole rectangle of the form J (a. (b) L + x III and L x � are not homeomorphic. (e) Of the 2manifolds constructed from L + or L with HI = 0 and one para compact end.' ([r. show that for every I we have ( H (x. then n n both sequences converge to the same point. (d) The StoneCech compactif.O) for every irrational a. 0) x for all x. (a) L + is not contractible. 00)) is countablt. (d) L (e) More generally. (c) Let U be an open set in �� \vhich 1S the union of "wedges" centered a t (a. Similarly for L + x �. only L + x � has the homotopy type of L +. &). L x �.468 Appendix A (c) If {X } and {Y } are sequences in L + with Xn S Yn :5 Xn +! for all 11. (Using Problem 3(g). L x L.I)} L +. L x L +. (b) HI(L +) Ht {L) O . L + x L. (a) L + and L are not homeomorphic. (f) If f : L + > � is continuous.t ech compactificatlon�. some A n 1S not nowhere denst . Hint: Given H : L + x [0. Hint: Let A n = {a : the wedge centered at a has width � I /n } . one can explicitly construct these Slone. (gJ If f : L + > � is continuous.r. Hint: Imitating Problem 119. and the big disc arc all distinct. (c) HI(L + x S') = = Ht {L = = > = 5.
is a manifold which is not metrizable. ( g) The dis joint union o f two copies o f P' . . 1) E �i U {(O. Its fundamental group is uncountable. Let M be a connected Hausdorff manifold which is not a point. Alternatively. (a) If A C M has cardinality c (the cardinality of �). Y)a. 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 ) . O)a a irrational) : : joint Show that C. � )a J'\ I 1 + sy l . = 469 C2 = { (0. and C2 are closed. with con·esponding points o n tht boundary identified.Appendix A (d) Let C" C2 C P be C. O)a a rational ) .{ (x . (e) Define H : P x [0. (b) If C e M is closed and has cardinality c. 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 . Conclude that P is contractible. then C has an open neighborhood with cardinality (. and such thai f lO) = { p ) (Consider functions defined on inidal segments of n vvith these same propertle�! and apply Zorn's Lemma. cardinality r. then the closure A ha. y)a y < 0) is a manifoldwithboundary P'. (c) Let p E M. but which has a countable dense subset. whose boundary is a disjoint union of uncountably many copies of JR . (f) P . There is a function f: n > (set of subsets of M ) such that f(rx) has cardinality c for all rx E n. { (O. 1( x 1 s + sy �' y  (x �. Describe a homeomorphism of this manifold onto �2 8.s + sy a ) if y if y > 0 : � o. 7.O)aJ for all p E P. s ) = Show that H is welldefined and that H(p. but that thev are not contained in dis open sets. 1 ] > P by H « x .
(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. 1 ] ) with the properties of tbe func tion in part (c) is eventually constant. (Given p' E M.470 Appendix A order.) (d) A function f : n > (set of subsets of [0.) 9. is homeomorphic to either the real line. Then there is a unique f with the required properties. (e) M has cardinality r. then M is homeomorphic to S' . (a) A connected I manifold whose topology is the order topology for some . of UP<a f(f3) with cardinality C. consider an arc from p to p'. the long line. or the halflong line. This is an example of defining a function by "transfinite induction" .
.L+p is clearly homeomor phic to L +. How ever. the method used fOI" obtaining a CW structUl'e on L + will not yield a complex ana!wic Jlruc/ure on L + x L +. the problem is that a complex analytic structure on �2 mav bc conformally equivalent to the disc. and then it is easy to see P I that we must also have f(/(q)) > f(q). but this implies that f(x) = x for all x. this follows immediately from the fact that the maps f used for identifying points in various (��) + with a. we must have f(q) > q.L +p. we can assume f(q J ) > q l . Now f cannot be the identity on all points > Xo (for then it would be the identity ever)'\'Vhere. but it . 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. It is knOvvll that there are uncountably many inequivalent CW structures on L + . This sequence has a limit in L + . f(/(/(q ))) > f(/(q)). a contradiction. and hence for L +. and consequently that the structure on M can be extended if M is a proper subset of L +. in other words. So for some ql > Xo we have f(ql ) '" ql. points in ��. . Using the CW structure on L + .Appendix A 471 CHAPTER 2 The long ray L + can be given a Coo structure.L +p. 1 ). then it yields a CW structure for L + . has a limit point Xo E L + . there is no CW map q > p with a CW inverse. f(/(q)). These are all distinct. since we can consider fI in the cOlltrary case. . In fact. and even a CW structure. I do not know whether every 2manifold has a Coo struCtUlT. are all CW. etc.L +p. . and hence extendable. 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. with f(xn } = Xn . f(q). A CW stmcture exists on the Prufer manifold.e can get a CW structure on L + x L +. since it is CW). \". Continuing i n this wa)� we obtain Xo < Xl < X2 < . Reasoning as before. . and I(xo) = "0 . If (!) is a CW structure for L +. The increasing sequence q. This implies that it is also diffeomorphic to (0. then L + . If p E L +. and L+p denotes all points S p. we obtain Xl > Xo with f(x} ) = X l .
L+q . Arner.iihlbarkeil. a mod ification of the Priifer manifold yields a nonmetrizable manifold of complex dimension 2. pp. Sic. E X are contained In a compact connected C C X. I 25115 (1958). 4 (1953). Of the Yalious manifolds mentioned in the previous section. Complex AnaD'lic Manifolds wilhoul Counlable Base. Kneser. try to deter mine v·. H. (b) Let d(x. Soc.472 Appendix A may also be equivalent to the complex plane. Fennicae Se ries A. ). In fact. Ann. On the other hand. and not extendable. (a) Every twO points X. Proc. Acad. where X is H ausdorfl connected. 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 + . p) be a metric space and let f : X > Y be a continuous locall" oneone map.hich can be immersed in which. Prove that lor I I . pp. it is a classical theorem of Rado that every Riemannian surface (2manifold with a complex analytic structure) is second countable. PROBLEMS L+ . Let (Y. 335340. 10. and locall" compact. References to these matters are to be found 111 Calabi and Rosenlicht. . 18.L +p > 1 2. locally connected. y) be the greatest lower bound of the diameters of f(C) (in the pmetric) lor all compact connected C containing x and y. Math. AnaiJ'lische Slrucklur und A b<.
2a) E A with y < O.z) E �3 : y '" OJ. y. y .ala E �� with Y > O. For. y . Z )a (x . z) a for y > O for y < 0 with with (a + yx.ala) contains the points (x. y. This example is due to M. y. II (1960).al a with y < O. and these are identified with the set of points (x. y. Now the folium containing { (x.z". O) : y < O} C A . y. The equivalence classes form a 3dimensional H ausdorff manifold M. manifold there is an obvious function ". c . pp. c . Archiv. 280281. The remarkable fact about this 2dimensional manifold N is that it is connected. Beispiel eiller dimensio'nserhohenden anab'li. Kneser. We will now describe a 3dimensional version with a twist. o) a . the set of points (x. z + a ) (a + yx. points in �� being denoted by (x. and the sets z = constant form a foliation of M by a 2dimensional manifold N. In the di�oint union of A and all ��. Math. y. a E IR we identify (X.a).Appendix A 473 CHAPTER 6 Problem A6(g) describes a nonparacompact 2manifold in which two open halfplanes are a dense set.. and for each a E � let �� be a copy of �J. y . y. c . c) E A with y > 0 is identified with the set of points (x. . On thi. Let A = {(x. c . y. Since we can choose a = c/2. z . we see that all leaves of the foliation are the same as the ieaf containing { (x. y.· chell A bbildung "wischen iiberiib"ahlbaren Mannigfaltigkeilen.
2. they cannot be extended to non· paracompact manifolds. a locally compact connected topological group is acompacl (Problem 104.474 CH APTERS 7. The converSe also holds: slnce a Riemannlan metric determines an ordinary metric. 6. since an equivalence would determine a Riemannian metric. Although the results in the Addendum to Chapter 9 can be extended to closed. Y. 10 I. 5. nor is the tangent bundle of the Pnjfer manifold.. a Riemannlan metric on M. even though the Prufer manifold is contractible. Theorem 79 fails for L. Compare pg. but a much easier proof is now available: Let ( . if f: N M lS an immerslon. Notice also that if M is nonparacompact. as can be seen by considering the Odimensional sub· manifold {(O. We have seen that any paracompact Coo manifold has a Riemannian metric. 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. . Thus the tangent bundle of the long line is not trivial. ) b. (On the other hand. a basic result about bundles says that a bundle over a paracompact contractible space is trivial. More generall\'. Thus. also metrizable. Problem AI I implies that a manifold N immersed in a paracompact mani fold M is paracompact. for the folium is a submanifold. Appendix A 9. then N has the Riemannian metric f* ( . then TM lS definitely not equivalent to T* M.) 4. 3. So there are at least two inequivalent nontrivial bundles over M. contradicting the fact that the tangem bundle is not trivial. Since there is no Riemannian metric on a nonparacompact manifold M. and hence paracompact. 2 7 2. 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. the tangent bundle TM cannot be trivial. not necessarily compact. since its tangent bundle is trivial. submanifolds. ). O)a) of the Prufer manifold. The tangent bundle of the long line L is clearly orientable. A Lie group is automatically paracompact.
PROBLEM 13. 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).Ajijlclldix A 475 7. Addendum 1). Is there a nowhere zerO 2form on the various nonparacompact 2mani folds which have been described? . TIlis will b.
.
�.NOTATION I NDEX CHAPTER ] CHAPTER 2 M p2n pn sn aM JRn 8(X) lHln 23 19 4 S' 11 19 1 6 19 .i) Till" T' Xp J . 61 34 34 28 34 35 61 6] v" v { f) Iv) . . � Ell 1. }.· 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. p I'm F 65 72 83 65. W ) Tk < v ) Tk(i' ) T V) d T. VnJ IA � Ix. ax ..'(V) Tt' (� ) 7ik ( V ) I' J p' End(V) df d. 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. 1 1 9 1 13 131 109 1 16 1 16 1 20 121 121 1 22 I ii . v p a . . U ) aJ ar I ax. (p). . �) SO(n ) (x. 'U) SL(n. �) O (n) R(n) (�n . 75 65 101 T'M T®S 109 1 10 1 21 107.i)p np TM T(M. x �2 � Mp (M.· ' CHAPTER 3 E"(X) J. i. 1 1 6.
. . II P II Lx w S.) / ')" r" . 234 23. 10. . 8fi.h.. 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 5.I/.il.478 N olation Inde> T (I.. 234 202 Ali curl X All I dx ' /\ .n OIx(l) = 01 (1 . w(X) Iv t' �j l . 1'·* V' 1 23 10..�:��/ . Vk ) (v" .«) ('R. 235 2 10.. .i l . signp J w(p) Zk ( M ) . (. " . Vk ) IA I (/(1 2 ) exp A Lx A Lx f Lx ) " Lx w [X. 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. /\ dxn l de den d8ta . X I.c) f' ' . 2 1 3. 1 29 1 34 1 34 1 08 1 09 1 0.J w <PI a· /\ a· CHAPTER ("" Q(M) Q k (V) QO(l ') (v. x) (01. g ) mj '[La) wa d f l(LI) i1·w Iw dw M. . 1 0.wl (V) n.f Hk (M ) H� ( M ) I' deg f div A" de t(f.wI 7i� ( V ) T°(V) 231 231 231 231 20 1 227 206 202 227 215 20 1 201 203 v. . II. CHAPTER q" 7 202 205 238 238 2 19. 224 215 22.
J ) exp d. ).!. ).' // / J dx + g dy /w Lw # (' (q) u [O.. 257. I Jo CHAPTER A ( ( ( ( ( ( ( « I II II rti . 248.(M) 6n a' 0' [wJ al' di ( 268 285 29 1 264.1 IAI A i) Ad( ) ad X Aul( g) a Euc(V) Euc(�) E(y ) J' ( . �) . 356 356 314 314 31 I 309 309 324 334 302 306 32G 312 344 sinb tanh x'( v) xi r . )' . . 246. 259.N olalion il1dc. } .. I II II. 291 264 264 263 248. 285 285 260 246 239 243.. 288 294 246 299 299 266 9 322. 245. )r . HI ii(u) oj 0. }/ ./J L� Mp" exp(A) GL(II.) . 404 373 410 385 385 372 407 407 376 CHAPTER d(p.\ 479 Z. 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. ) (g. q) ds d l' ("osh ("osh . �) g[(". )' .u E(II) Elld(nl exp O C O B C' 'J [ij.
APPENDIX A k=d L W' U! to Q Old <> + 1 464 463 464 464 461 46 1 463 .] [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 " .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. u .480 i.1 R" SO(n) X . 'II p(W A ry) [ .: <jJ. £(G) o(n) P Ck (M) 9!(N) PD RI. 410 403 376 403 400 400 400 403 CHAPTER I I L. L. 0(13 ) .
1 9 long ray. 391 submanifold. 20 subgroup of a Lie group. 348. 466 plane. ] ] I tangerlt. 252 geodesic. 248. 23(l CauchyRiemann equations. generalized. 218. R Antipodal map. 3 9 . 283 Choice function. I CJ Boy's Surface. 378 in o(n. 19.k(p). 19J Besicovitch. 7 1 normal. 35. 293 AJgebraic inequaJities. 4!=l . 252 Bounded manifold. L. 108 fibre.. 45 I sphere bundle. 379 Bundle cotangent. 7 J BuraliForti Paradox. 1 07 for Mp'. J. 248. I I Arclength. 344 of contra\'ariant tensors. 1 9 subgroup of a Lie group. 7 ) Basi� dual. 20i multilinear function. 31 2 function. 360 Cartan's Lemma.INDEX Abelian Lie algebra. 36 i halfspace. J O� dual. 20J Alternation. 246 Associated disc bundle. 7 2 . 466 Biinvariant metric. 1 1 1 Chart. 309 induced. A S" 1 7 9 Big disc. 46() Calabi. 7 3 nplane. 395 Adams. Fundamentru Theorem of. 200 Cayley numbers) 100 Chain. 38 Change) infinitely small. �). 1 54 in gl(n. 28 Choice. l Oll Adjoint T* of a linear transf ormation T. ] 20 of covariant tensors. 465 manifold.. principle of irrelevance of. 2 1 0 vector. 233 Alternatinc ("ovaria� t tensor field. 7 i trivial. L n. 285 Chain Rule. 46� Circle. 27�: point. 60 Bracket. 28 maximal. 380 Alexander's Horned Sphere. 285. 3 3 2 Arcwise connected. 20S f g. 34 Annihilator. 3 1 3. 6 Closed form. 46b half plane. J 45 Base space. 55 AJgebra. 1 03 Ado. 3 1 6 Cartan. 2� Auslander. F. 1 0 1 map. 409 Area. 202 Analytic manifold. 208 or Belongs to a distribution. �). 376. Elie. 228 Annulus. 106 quadrant. £" 472 Calculus of variations. 40J Boundary. 464 Banach space. 382. 451 Atlas.
83 Derived set. 130 tensor field. 133. Uy. 38i Contractible. 32 at a point. 279 Degenerate. 286 Degree. 465 Cohomology. 28 Cotangent bundle. 32 . 439 Curl. 1 20 vector field. 283 integral. 60 refinement of. 65. 3i3 Conjugate. J O �l Covariant functor. 209 odd scalar. 30 oneparameter group of. 39. 228 Definition. 30 Dilleomorphism. 299 Cube. 39b Continuous homomorphism. 429 Determinant. I I . geodesic'. 1 1 3 COWl locally finite. 284 Debauch of indices. 28 Darbom: integrable. 27. 236 Contraction. 82 Coorelated. 275 mod 2. 34 CO manifold. 238 Cvlinder. ] 2] . 4 19 d e Rham. 123 Decomposable. 263 group of M with real coefficient:. 5(1 Cramer's Rule. ] 30 tensor field.. 3ib Complete. ]7 9 form. 429 Coordinate lines. vector field. 8 C. singular. 39.10 pointfinite. 214 Deformation retraction. 28. 34 C� distribution. 22i Lemma. 372 Critical poim. 139. 32 manifold. . 30 on the long line. 358 Constants of structure. 220. 3 1 manifold. 421 analytic structure. 133 Derh<ation. 3 1. 259 relative scalar. 232 DiAeomorphk. 32 Riemannian metric. 47 1 on pn. 295 Densin· eve. 40 i ll the calculus of \·ariations. 308 structure on TM. 1 39 Contravariant functor. 225. 15� Coordinate system. 341 Complex. 42 [ Commutative diagram. 78 of a ring.482 Closed (contiuued. 320 Critical value. up to first order. 283 Darboux's Theorem. 207 function. 1 33. 14 Crossproduct. 15H Coordinates. 299. manifold. 24b index Cup product. 25 DescartesEuler Theorem. 29 structure. 420 Commutative Lie algebra. 263 of a complex. 40 Cross section. invariant. scalar. J 6{1 Cofinal. 28. 29 manifo]dwithboundary. 227 Crosscap. 23 J scalar. 4 i ) numhers of norm 1 . I 1 3 Convex geodesicaU)" 363 polyhedron. 148 DiAerentiable.
194 Force field. Dual basis. 205 type. 445 Euler's Equation. 32. 89 Fundamental Theorem of Algebra. 429 characteristic. 32(1 Even ordinal. 2 1 0 equation. 374 nspace. 433 of vector bundles. 432 Extremal. 1 92. 384 Extension. 3 1 S motion. I Faith. 201 of a function. 3 19. 4. 207 differential. 4 Direct sum. 96 Euclidean metric. 72 weak. 233 Disjoint union. 358 Equations depending on parameter�. leap of.Inde. Five Lemma. 254 . 2 113 sequence. 2 1 5 o n torus.ntion. 1 69 Equations of structure. 461 First variation. 254 Functor.. 400 Irelated. 308 Elliptical nonEuclidean geometry. 422 of a pair. 440 Fixed point. 428 class. 10. 133. 209 Exact form. 23 paracompact. 68. 64. 7 1 Finitt> characteristic. 421 Disc bundJe. 134.. 1 36. 20 Distribution. 374 right invariant. 464 Fibre.' Differentiable (cOTlliTlurd (structure con1iTllled.. 29 on s n . 385 . 293 Fundamental Theorem of Calculus. 23 1 relative tensor. 324 Envelope. 3 Du Bois Revmond's Lemma. 1 8 1 ideal of. 39 Elements of norm 1 . l OB Euler. ponential map. J 2 1 Energy. 1 0 7 vector bundle. 285. 334. 320 483 Einstein summation conv(. 1 8ll Divergence. 23 1 scalar density. 4� End. 190 Frobenius Integrability Theorem. 164 depending on parameters. 1 03 E. 194 Folium. 468 Endomorphism. 305. 45 1 Discriminant. 419. associated. 109 Dimension. 467 relative scalar. 438 First element. 139 Foliation. 215 Fubini's theorem. 30 Differential. Exponential of matrices. 240 Form. 355 . on �n . space. 169 linear. 238 Theorem. 36i Embedding. 16S forms. 179. 201 left invariant. 404 Equivalence (or vector bundles). 1 30 Functorites. 352 Domain.
H. 84 Group Lie. 40i orthogonaJ. 3 1 4 Infinitesimal generator. 1 9 Handle. 372 usual. 23:. reversing map. principle of irrelevance of algebraic. H Hardy. 3 7 1 matrix. 337 General linear group. 387 of Lie algebras. 283 Riemann. 283 . 3 1 4 Initial conditions. 104. 61. 401 Geodesically complete. 459 Homogeneous. 462 Inner product. 356 sine. 194 Gradient. 1 0 Imbedding. 3 1 2 Infinitely small cbange. 304 Gmk. 60 Indefinite metric. 14 Immersed submanif old. 30l preserving. 3 33 Germs of kforms. 14. 466 Halfspace. 46 Implicit function theorem. 44i on �n . 46 lOpologi �al. 148 Infinitesimals. 192 Integrable functioll Darboux.. 380 Homotopic. 101 orientation. 23i GramSclllnidt orthonormalization process. 49 topological. 3 i 2 opposite. 356 Ideal of a Lie algebra. 37� Generalized area. 189 Integrable distribution. independent. 2 1 Integrability conditions. 432 Global theory of integral manifold�. 104. 466 line. 36. 277 Hopf. 106 HahnBanach theorem. 410 Identification. 342. 46S strip. 136 of integral ClJrve. 342 Inde> Hyperboli( cosine. 7 Homomorphism continuous.. 145 Hair. 350 Independent infiniteBimals. 47 Immersion. 227. 44 6 Indite:debauch of. 84 Hausdorff. G. H. 341 Geodesicallv convex. 277 Homotopy. 3 7 J Guillemin. 333 closed. 356 tangent. 372 topological. Sylvester's Law of. I nertia. 69 HalfIonp cylinder. 314 Index of inner product.484 Gauss's Lemma. 363 ' Geodesy. 304. 2 3 Heiukin. 450 Hopf Ril1owde Rham Theorem. 349 Infinite volume. 260 Inequalities. l 36 Initial segment. 246 Geodesic. 351 Induced bundle. 123 raising and lowering. I I I Infinitely small displacements. 301 Inside.. 1 7 9 Ha s one end. 349 Index of vector field on a manifold. 39. V W. Robert A.
136. 179.. 245 Integration. 226. 3 76 abelian. 472 Kneser.. locally: 382 lsomorphism. 233 Isometry. 4 1 5 oneparameter group of local diffeo morphism. 409 closed subgroup of. 415 third. 165 svstems of. 194 of a differential equation. 374 Left translation. l O P Isotopic.. 103 contraction of. 243 Linear differential equations. 1 1 9 Lobachevskian nonEuclidean geome try.. S. 7 1 Local theory o f integral manifoIds� 190 Kelley. 1 8 1 maximal. ]. 18. 460. 43S 485 vector field. 150 Lie group. 243. 410 opposite. 58 Lav�' of ] nertia.ket in any ring. :M. 128. 21. 239 In\'ariance of Domain. 394 n form. Invariant. 179 triviality. 60 Line integral. 34 9 Leaf. 463 ' set. 382. . natural. 366 Klein bottle. 373 Lie's fundamental theorem first. 464 Kink. 283 surface. 296 Lipschitz condition. 380 ideal of.. 243 manif old. 410 topologically isomorphic. principle of.. 1 7 J Lin� ar transformation ac\joint of. 374 Length. Integral curve.inde. J44 Lie group. 148 spanned locally. 400 . 395 commutative. 232 definition. 435 Kneser. 136 Darboux. 121 positive definite. 239. 138 Littlewood. 376 for the blac. 407 Lie derivative. 283 line. 194 Leap of f aith.. 376 homomorphism of.]. 340 Isomorphic Lie groups. 305. 4 1 6 Limit ordinal. E. 414 second. 368 Local flow. 104 positive semidefinite. H. 378 Jacobian matrix. Sylvester's. 391 local. 40 J ordan Curve Theorem. 376. 463. 145 Laplace's expansion. 3 7 1 arcwise connected subgroup of. 239. 179 Lives at points. 415 normal su bgloup of. 2 J 4 Irrelevance of algebraic inegualitie�. 104 Linking number. 294 Jacobi identity. 464 Left iJwariant form. 155. 3 1 2 o f a curve) 59 Lie algebra. 230 Laplacian. 473 Lang. 1 36 Riemann. 388 Lie $ubgroup.
1 3 path wise connected. 46] open. E. 345 Newman.1 Mobius strip. 19 Nondegenerate. 1�1 C' 34 Co: 34 Coo. 283 index MayerVietoris Sequence. 29. 4 5 9 analvtic. 344 . I . V. J 9 closed. 382 Lipschitz. tubular. 2�1 Manifold. 46" ray. i Natural gvalued Iform. 2�1 dimension of. 20 connected. 3 I\'latrix grou ps. 466 line. 4 ( 1 Massey.<. i ntorus. 403 Natural isomorphi. 20 Milnor. H . 34. 208 Multilinear function.. 367 Lobachevskian.486 Locall) compact. Ht) smooth. W. 401 Euclidean. 73 rank of. 2 8 boundary o C 1 �l ' bounded. 4 J Mesh. 304. 350 Riemannian. 20 finite cover. 40. 4 6G orielltation of. J 15 Munkres.. 30 I NonEuclidean geometr>" elliptical. 46:} Lower sum. J06 MacKenzie. 4 nplane bundle. 86 manifold.]. :)7�' Maximal intcgral manifold. 4 . 36H Nonmctrizable manifold. 194 ndimensional. 2 1 4 Manifold.'. ](l generalized. 1 39 oneone. 3 4 ' atlas for. M .vitllbnundarv. 37:2 Normal bundle. 3 1 2 spaces. 424 for compact supports. 179. 20 Long cylinder. 42 Mod 2 degree. 465. 108 Neighborhood. 400 nholed torus. S.. 50 isomorphic Lie groups. 3� Map betweell complexes. left invariant. 1 9 4 nOllmctrizable. 8(1 Norm. I�l . 4 imbedding in �N. R. 3 1 1 usual.m. 52 integral. 3 Nice cover. 4 nforms. 438 Nonbounded. 303 preserving. 466 N onorientablc bundle. A . disjoint union of. 239 Metric biinvariant.]. Ion Multiindex. l OCi Magic. 308. 2�1 diflerentiable. R. 305. 465. 460 closed. 7 1 nsphere. 43 J :Measure zero. 9 nmanifold. Coo.. 1 8 1 maximal. 42 1 bundle. 3 1 5 indefinite.
yellow. 98 outward. 304. 384 Opm long ray.. 148 of local diff eomorphisms. 248 reversing. 468 Parameter curves. 225 Paracompact. 104. 133. ] 34. 60 submanifold. 35 Partition. S. 465 map. 450 Poincare dual. 86 manif old. GramSchmidt. 209 Outside. 450 Poincare Lemma. 86 Oriel1latim. 248 Orthogonal group. 367 Point inward. 259 Onedimensional distribution. 17!=l Onedimensional sphere: 6 Oneparameter group of diff eomorphisms. 407 Lie algebra. 3 1 2 Pig. 36 integration in. 461 topology. local. 84. 372 Orthonormal. 30. isomorphic. 461 Ordering. 351 Odd ordinal.index 487 Normal (cantinued) space. 102 tensor.. 85 of a manif old. 225 Poincare upper half plane. 86 of a vector space. 434 Poincare. 304 Pollack. 60 Polar coordinates. 88. A. 463 Orientablc:bundle. 61. 104 Product of vector bundles. 284 Palais. 88 Proper map. 441 PoincareHopf Theorem. 459 subgroup of a Lie group. 52 Path wise connected. 460 Ordinal numbers. 301 Positive element of norm I . 308 Positive semidefinite. 304 Osgood's Theorem. 266 Polarization. 465 Ordered set. 98. of a bundle. special. 100. 19. 3 1 3 Partial derivatives.. 95 . 88. 32 Projective plane. H. 348 Orthonormalization process. 275 Prufer manif old� 467 Pseudometric. R. 148 Oneparameter subgroup. 7. I I. 260 Pointderivation. 84 preserving. 407 Orde. Opposite group. 245 of unity. 260 Outward unit normal. 21 Outward pointing. 459 end. 461 isomorphism. 351 Nowhere zero section. l l6 Projection. 85. 167 Parameterized by arclength . J06 Positive definite. 84. 105. 39 Pointfinite cover. 60. 20 Piecewise smooth. 239. 288 scalar density. 467 relative tensor. 435 space. 410 outward unit. 210. 439 Poincare Duality Theorem.
relative. 45!=l Section of a vector bundle. 194 Slice maps. 62 . 382 Singular cube. 283 Riemannian metric. 40.. T. 7 Sphele bundle. 422 of vector bundles. 40 space. 103 MayerVietoris. 4 7'2 Rotation group.. 285 Skewsymmetric. initial. 303. 362 Second countable. 451 RadiaJ f unction. 6(J Shuffle permutation. 277 manif old. 294 Solid angle. 472 Rank of a f orm. 342 Roman surf ac{'. 433 Shrinking Lemma. 134. sum. 59 Rdlnement of a cover. 427 singular. 248 Retraction.. 17'1 Special linear group. 321 de Rham. 22� of a map. 22i Simplex of a triangulation.. 1 90 Relative scalar. 246 simplex. 457 RicmaJ1Jl integrable. 28:. 50 Regular point. 40 Related vector fields. 29 piecewise. 98 Rectifiable.. 8. 54 Smooth. 42. 419. 2 3 1 tensor. 288 Reparameterizatiol1. 9 6 Segment. 263. 28 homotopy. 435 Rado. 231 Scalar density. 104 Semidefinite. 3 J J usual. 342 de Rham cohomology v{'ctor spaces. 290 Space filling curve.'. 43J or of a pair. 400 Right translatioll.connected. 100 of norm 1. 220 homotopic. 58 Spanned locally. M. 20J. 268 de Rham's Theorem. \fl.488 Quaternions. 462 Selfadjoint linear transf ormation. 134. 263 with compact supports. 3i 3 Sard's Theorem. 95 Sequence exact. positive. 424 f compact supports . 7 3 zero. surface of. 62 Sphere. 6 J Special orthogonal group . 3 1 2 Right invariant nform. 231. 133 Schwarz. 308. 374 Rinow. associated. 459 value. 26 Roscnlichl. 285 Simply.. 3 1 2 Smoothly contractible. 134. 244. H. 5 I Shrinking Lemma. 3iB Slice. 264 def ormation. J7. 354 Schwarz inequality. 104 Separate points and dosed sets. 294 Scalar. 279 Revolution. 277 isotopic. 283 integral. 287 Lie group. G.
147 Surface. 246 Starshaped. 221 Steiner's surf ace. 352 Stonee ech compactification. 134. 8 nholed.. 101 Support. 42. 1 1 3 even relative. 42 7 Trichotomy. 171 acompact. 285. 374 right. 3 7 1 imbedding. 396 Subalgebra of a Lie algebra. 50 Subgroup Lie. 106 Stokes' Theorem. 120 covariant. Standard nsimplex. 374 Triangle inequ ality. 122 Tensor product.Inrie. 4 . 33. 4 7 open. 231 odd relative. 3 Wedge product. 63. 7. 461 Trivial vector bundle. 17. 14. 49 Coo.\'hitney sum. 301 System of linear diff erential equations. H. 8. 354 integral. 442 Thorn lsomorphism Theorem. 459 489 Tensor contravariant. 64 outward pointing. 426 singular cube. :2 Su ccessor ordinal. 4�) immersed.. 46 Topologically isomorphic Lie groups. 321 Sylvester's Law of Inertia. 72 Tubular neighborhood. 388 Torus. 373 oneparameter. 456 Topological group. 253. 345 Twoholed torus. 3 7� Subbundle. 66 Vick. 123 contravariant. 349 Symmetric bilinear f orm. 239 of revolution. 1 2 1 . 468 Structure constants. 8 Tangent bundle. 77 Tangent space of �Ii. 116 Thorn class.. 98. 303 Triangulation. 7 1 Totally disconnected. 1 1 3 mixed. 101 . 261. 9 Total space. Whitney. 1 4 immersion. 76 of �". 120 covariant. 49 closed. 7 area. 260 to a curve. 426 simpJex of. 26 Sternberg. 106 '. 384 Submanif old. J W. 464 Sum of vector bundles. 198 Subcover. 460 Translation left. 134. 64 Tangent vector inward pointing. 203 Whitney. 25 Transitivity. S. 98 of a manif old. 7. 288 Tensor field classical definition of.
.
The elegant swashes of the italic ). The bold face supplied by Monotype� even the "semibold": is obtru sively extended. and special parentheses and other punctuation symbols were also required.. results from the linear scaling. The text font is I I point Monotype Baskervillethough the emdash has been modified together with its italic. The figures were produced with Adobe Illustrator. as well as the fact that the upper case Baskerville letters are of somewhat heavier weight than the lower case. Knuth's 'lEX typesetting system. A thicker set ofnumerals was constructed for use with the upper case lettering in chapter headings and statements of theorems. which was created in three essential sizes: 9. including bold symbols. as well as for the extension font and its bold version. On the other hand. so a special gy ligature was added. The somewhat bold appearance of chapter headings.hese books were typeset using Donald E. 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.. The italic fonts of these three families were used as the basis for creating the three separate " math italic" fontsfor use at ordinary size. while blackboard bold letters were made by hollowing out these bold letters.in. and additional special symbols. DVIPSONE PostScript driver. bold type was useful in special circumstancesmainly for indicating defined terms. The Adobe Mathematical Pi 2 font was used for the ordinary sized German Fraktur letters. Numerous other modifications of this sort.e Rime qfthe Ancient Ma. 7 and 5� point. 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. . The proportions and weights for these were then used for the three sizes of the symbol font and the other mathematics fonts. and J cause problems in words like topologl' and apology. were made for various pu rposes. including additional kerns and alterations of set widths. with su itably modified versions used for su perscripts. The bold letters for mathematics come directly from the bold fonts of the Times families. painted by the author in his spare moments. special gg and gf ligatures were also required. . Although a Baskerville bold face is unhistorical. so a nonextended version was created. are loosely based on Samuel Tavlor Coleridge's poem TI. in superscripts. in 16 point type. The covers. script letters. and in second order superscripts. since simple scaling would have made them u npleasantly heavy. and new or modified fonts were created using FontographeI. the tall initial letters beginning each chapter were designed specially.
. Problem 30: Change part (f) and add part (g): pg. M. Problem 29(d). 78: the third display should read: 0 = e(O) pg. JR2 . . Xk . 22: Replace the top left figure with IHI" such that M and is denoted by 4>(x) = aM. fibre by fibre. Xk) is a Coo f unction for all COO vector fields XI . v(p )) for some v(p) E JR . '� . and aM itself is always a manifold JR" (but only one homeomorphic V pg. Xdp)). . then is Coo. 14: relabel the lower left part o f the central figure as di(x. the lower rectangle in the matrix ( a(Ui ) o f) ax} = pg. homeomorphically onto 2 T(M. 103. 3. Equivalently. ( p). �3 . If there were a way to map + N for a disconnected manifold N. . 43: Replace the last line and displayed equation with the f ollowing: Since rank f =k in a neighborhood of p.CORRECTIONS FOR VOLUME I pg. A (p) (X. . 1 1 8: Add the following to the statement of the theorem : If . corresponding to a M x JR2 . . . . the function J can be made Coo if (g) The same is true if M has at most countably many components. . + M is a Coo manif old. 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).y) < 1 to di (x. pg. JR is replaced with J : M. liP) is called the boundary of 4> : V + pg. 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. restate part (c) as f ollows: (c) This is f alse if J : (f) If M is a connected manifold. A is also. 60. . Add the hypothesis that = e(fh) = J(p)e(h) + h(p)e(f) = 0 + e(f). in the sense that A(X) .� �\ B: :B A. .i). = pg. there is a proper map J : M + JR. . Xk)(p) If A pg. Problem 32: For clarity. theorem of topology). A(X" .A is Coo.y) ::: 1. 1 1 7: After the next to last display. by using the up would pg. . x E aM if and only if there is a neighborhood of x and O. 6 1 . and we could continuously pick w (p) E criterion that w(p) should make a positive angle with v(p). . add: M is orientable. then each dashed vector. then A is Coo. . then aM = 0. line 3: change pg.. If M is actually a manif old. . . A.
226.1) x (n . A follows from the fact that the function Ail . In Problem + . pg. . pg. . Aut should be replaced with Aut . must also assume that each /'. we 17. 408410. For X Y E Mp and . 1. End with End . part (c) should read: . what are here called "odd scalar densities" should probably simply be called uscalar densities". . when the transformation rule involves (det (7) A)W. 133.] is integrable. Problem 9: Let F be a covariant functor from V. h component of VI x ··· X VnI in terms of the (n .. V. 133. 134. In Problem 20. 237. 177. O. 375 of V olume V pg. Extending the changed terminology f rom pg. 0<2 be two maps on some open interval 1 such that 0<1(1). what are called "even scalar densities" might best be called "signed scalar densities". . scalar densities In part (c) of Problem 10. Problem (d) Let f: M pg. In Problem 26. we should be considering the h of part (a).. In the comutative diagram. omitting the modifier «signed" when it involves I det A l w. In part (g) of H is a connected Lie subgroup. . 292. . 1 1 9 : Add the f ollowing at the end of the proof: Smoothness of pg. . replace parts (b) and (c) with: I (b) Determine the . For consistency with standard usage.. . 408. add the hypothesis that pg. . the lower right entry should be "Iforms on pg. The hypothesis o f Theorem Let pg. 4 1 1 .p N. Ell /'.i. Though there is considerable variation in terminology. is . 2 C V. 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. show that n Vj # 0 should be CJ n V. . 3 should b e changed s o that i t reads: x and E V and let 0<1 . p N". pg. pg. part (d) should begin: = for some to E I. 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. . The reference "pg. and suppose that f 5.pg. restricting densities to those of weight 1). ajaXI. 233. 143. " . and then replace Problem 19. .1) submatrices of the matrix In particular. the condition V. # 0. p. 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.A(ajaxl" . The display in Problem 21 . 198. pg.. ). Problem 16 (b) should read: "For any Lie group G. for 1R3 . 1 3 1 . V375" refers to pg.+I pp. show that .