123 70 9MB
English Pages 368 [367] Year 2016
Annals of Mathematics Studies Number 103
SEMINAR ON MINIMAL SUBMANIFOLDS
ED ITED B Y
ENRICO BOMBIERI
PRINCETON UNIVERSITY PRESS
PRINCETON, NEW JERSEY 1983
Copyright © 1983 by Princeton University Press A L L RIGHTS RESERVED
The Annals o f Mathematics Studies are edited by William Browder, Robert P. Langlands, John Milnor, and Elias M. Stein Corresponding editors: Phillip A. Griffiths, Stefan Hildebrandt, and Louis Nirenberg
ISBN 0-691-08324-X (cloth) ISBN 0-691-08319-3 (paper)
Printed in the United States o f America by Princeton University Press, 41 William Street Princeton, New Jersey
☆
The appearance o f the code at the bottom o f the first page o f an article in this collective work indicates the copyright owner’ s con sent that copies o f the article may be made for personal or internal use o f specific clients. This consent is given on the condition, however, that the copier pay the stated per-copy fee through the Copyright Clearance Center, 21 Congress Street, Salem, Massa chusetts 01970, for copying beyond that permitted by Sections 107 and 108 o f the United States Copyright Law. This consent does not extend to other kinds o f copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. 0-691-08324-X/83 $.50 + .05 (cloth) 0-691-08319-3/83 $.50 + .05 (paperback) Clothbound editions o f Princeton University Press books are printed on acid-free paper, and binding materials are chosen for strength and durability. Paperbacks, while satisfactory for personal collec tions, are not usually suitable for library rebinding.
Library o f Congress Cataloging in Publication data will be found on the last printed page o f this book
CONTENTS
IN T R O D U C T IO N
v ii
S U R V E Y L E C T U R E S O N M IN IM A L S U B M A N IF O L D S L . Simon
3
O N T H E E X I S T E N C E O F S H O R T C L O S E D G E O D E S IC S A N D T H E IR S T A B I L I T Y P R O P E R T I E S W. B allm an , G. T h o rb e rg s s o n , and W. Z i l le r
53
E X I S T E N C E O F P E R IO D IC M O T IO N S O F C O N S E R V A T I V E SYSTEMS H. G lu ck and W. Z i l le r
65
A R E H A R M O N I C A L L Y IM M E R S E D S U R F A C E S A T A L L L I K E M IN I M A L L Y IM M E R S E D S U R F A C E S ? T . K lotz Milnor
99
E S T IM A T E S F O R S T A B L E M IN IM A L S U R F A C E S IN T H R E E D IM E N S IO N A L M A N IF O L D S R . Schoen
111
R E G U L A R I T Y O F S IM P L Y C O N N E C T E D S U R F A C E S W ITH Q U A S IC O N F O R M A L G A U SS M A P R. Schoen and L . Simon
127
C L O S E D M IN IM A L S U R F A C E S IN H Y P E R B O L I C 3 -M A N IF O L D S K. U h le n b e c k
147
M IN IM A L S P H E R E S A N D O T H E R C O N F O R M A L V A R I A T I O N A L PROBLEM S K. U h le n b e c k
169
M IN IM A L H Y P E R S U R F A C E S O F S P H E R E S W ITH C O N S T A N T SCALAR CURVATURE C . -K . P e n g and C . -L . T ern g
177
R E G U L A R M IN IM A L H Y P E R S U R F A C E S E X IS T O N M A N IF O L D S IN D IM E N S IO N S U P T O SIX J. P itts
199
A F F I N E M IN IM A L S U R F A C E S C . -L . T e rn g
207
T H E M IN IM A L V A R I E T I E S A S S O C IA T E D T O A C L O S E D F O R M F . R e e s e H arvey and H. B la in e L a w s o n , Jr.
217
v
vi
CONTENTS
N E C E S S A R Y C O N D IT IO N S F O R S U B M A N IF O L D S A N D C U R R E N T S W IT H P R E S C R IB E D M E A N C U R V A T U R E V E C T O R R. G u lliv e r
225
A P P R O X IM A T IO N OF R E C T IF IA B L E C U R R E N T S B Y L I P S C H I T Z Q V A L U E D F U N C T IO N S F . J. A lm gren, Jr.
243
S I M P L E C L O S E D G E O D E S IC S O N O V A L O ID S A N D T H E C A L C U L U S O F V A R IA T IO N S M. S. B e rg e r
261
O N T H E G E H R IN G L I N K P R O B L E M E. B om bieri and L . Simon
271
C O N S T R U C T I N G C R Y S T A L L I N E M IN IM A L S U R F A C E S J. T a y lo r
275
R E G U L A R I T Y O F A R E A -M IN IM IZ IN G H Y P E R S U R F A C E S A T B O U N D A R I E S W ITH M U L T I P L I C I T Y B . White
293
N E W M E T H O D S IN T H E S T U D Y O F F R E E B O U N D A R Y PROBLEM S D . K inderlehrer
303
SOM E P R O P E R T I E S O F C A P I L L A R Y F R E E S U R F A C E S R. F in n
323
B E R N S T E I N C O N J E C T U R E IN H Y P E R B O L I C G E O M E T R Y S. P . W ang and S .W . W ei
339
IN T R O D U C T IO N
T h e present volum e c o lle c t s the p apers w hich w ere presented in the academ ic year 1979-1980 at the Institute for A d v a n ced Study, in the areas of c lo s e d g e o d e s ic s and minimal s u r fa c e s , a s part of the a c t iv itie s of a s p e c ia l y ear in d iffe re n tia l geom etry and d iffe re n tia l eq u atio n s. Starting w ith a su rvey lectu re, they have been arranged acc o rd in g to dim ension and approach, from c l a s s i c a l to that o f geom etric m easure theory. We w is h to extend our s in c e re thanks to a l l contributors, p articu larly for their c o lla b o ra tio n in se n d in g their texts and their re v is io n s a s w e ll as for their p atien ce in w a itin g for th ese notes to ap pear.
Our thanks a ls o to
the N a tio n a l S c ien c e Foun datio n for supporting this s p e c ia l year at the Institute for A d v a n c e d Study. E N R IC O B O M B IE R I
Seminar On Minimal Submanifolds
S U R V E Y L E C T U R E S O N M IN IM A L S U B M A N IF O L D S L e o n Sim on*
Our aim here is to g iv e a gen e ra l (but n e c e s s a rily b r i e f ) introduction to the theory of minimal su b m a n ifo ld s, in clu d in g as many exam p les a s p o s s ib le , and in clu d in g some d is c u s s io n of the c l a s s i c a l problem s (B e r n s t e in ’s , P la t e a u ’s ) w hich h ave provided much of the m otivation for the developm ent of the theory. Our firs t ta sk is to d is c u s s a notion of minimal “ v a r ie t y ” in a s u f f i cien tly gen e ra l s e n s e to in clu de the vario u s c l a s s e s of o b je c ts (e .g . a lg e b r a ic v a r ie t ie s in
C n , co n es over smooth su bm an ifo lds of
“ so a p film li k e ” minimal s u r fa c e s in
Sn ,
R 3 , branched minimal im m ersions,
le a s t area in tegral current re p resen ta tiv e s of hom ology c l a s s e s ) w hich a r is e naturally.
T h is w i l l b e done in §2, after som e c l a s s i c a l introductory
d is c u s s io n of firs t and seco n d variatio n in §1.
In §3 w e present som e of
the p rin cip al c l a s s e s of exam p les of minimal v a rie tie s .
§4
in clu des a
d is c u s s io n of som e of the s p e c ia l properties of minimal su bm an ifo lds of Rn .
In §5 w e g iv e a b rie f survey of the known interior regularity theory,
and in
§6
w e d is c u s s the c l a s s i c a l B ern stein and P la t e a u problem s and
the p resent state of kn ow led ge concern in g them.
F in a lly , in §7, w e d is c u s s
som e s e le c te d a p p lic a tio n s of seco n d variatio n of minimal su bm an ifo ld s in geom etry and to p ology.
^R esearch w as p artially supported by an N .S .F . grant at the Institute for A d v an ced Study, Princeton. © 1983 by Princeton U n iversity P r e s s S e m in a r on M in im a l S u b m a n ifo ld s 0-691-08324-X/83/003-50 $3.00/0 (clo th ) 0-691-08319-3/83/003-50 $3.00/0 (p ap e rback ) F o r copying information, s e e copyright page.
3
4
LEON SIMON
We w ould here like to recommend the survey a rtic le s [ L I ] , [ N l ] , [0 1 ], [B ], w hich cover topics only touched upon (or not mentioned at a l l ) here. F o r other gen eral re ad in g w e stron gly recommend the works [F H 1 ], [G E 1 ], [L 3 ], [L 4 ], [N 3 ], [0 2 ]. In a ll that fo llo w s ,
N
w ill denote an n-dim en sio n al R iem annian
m anifold without boundary U
(n >
2
),
and
k is an integer with
w ill a lw a y s denote an open s u b s e t of
N.
to co n sid er a “ smooth defo rm atio n ” (i.e . a smooth iso to p y ) of holds everything o u tside a com pact s u b s e t of let
(- 1 ,1 ) x N
be a
C2
( - 1 ,1 ) x N ,
^ : N
s u p p o se
N
X
fix ed .
0 o = i N ,
be defin ed by
K CU
N
w hich
T o be s p e c ific , 0 : ( - 1 , 1 ) x N -> N
t(x ) = 0 ( t , x )
is a diffeom orphism of each
s u p p o se there is a com pact
(0 .1 )
U
be equipped with the product metric, let
map, let
< k < n.
1
We a ls o often have o c c a s io n
for
( t , x )
1=1
with
t-
as in ( 1 .4 ).
It go es without s a y in g (and is e a s ily c h e c k e d ) that a ll th ese d efin itio n s are independent of the particu lar ch o ice of
•••,
We now have from (1 .9 ), (1 .1 0 ), and (1 .1 1 ) that
I (1 .1 2 )
= “
d iv MX
X
a=k+l
= - , and hence
J d iv „X
d „ -
f
d l , „ X T d? - I
M
< X , H > d fi .
M
T o go further, w e assum e that
M
is a c tu a lly com pact with smooth boundary
(9M (p o s s ib ly em pty); then the c l a s s ic a l d iv e rg e n c e theorem te lls us (s in c e
2
i.J=l
^
< R (r i,X i )X i , r i > ) d , i .
1=1
In ch eck in g this w e first show that that
= J i
2
a s in (1 .9 ), (1 .1 0 ) a b o v e ).
I (X ) = I ( X ^ )
and then u se the fact
< r - , V v a X X , v a > = - < B ( r - , t -), X > a = k + iJ i '
( va,
B
F o r further d e ta ils concerning this com putation,
SURVEY LECTURES ON MINIMAL SUBMANIFOLDS
s e e for exam ple [S L 1 ]. X
, w here
=
In the cod im ension 1 oriented c a s e , w e can write
is a smooth unit normal and
v
and the e x p re s s io n for
I(X )
I (X ) =
{!V £ | 2 - C
2
k (| B | 2 + ] £
M
w here
V£
k 2
< R (r -,v )v ,r ->
1=1
1
R ic c i curvature of
is a s c a la r function,
< R ( r i , l, V , r i >) } d ^ ,
i= l
d enotes the gradient (tak en in
N o tic e that
£
becom es
c (1 .1 9 )
11
is ju st
M ) of the s c a la r function
Ric(i/, v ) , w here
R ic
£.
is the
1
N .
We should fin a lly mention the m eaning of the terms “ station ary in and “ s ta b le in
U ” in c a s e
this w e can s u p p o se that
M
is immersed rather than em bedded.
(not n e c e s s a rily an im m ersion), and let J 0 / 0 (x ) = ||Aj^(d ^r)|| , where
TXM 0 ->
induced by
if/ .
of co u rse d efin ed by (1 .2 0 )
A( 0 for every su ch variatio n .
Then
K C U Q,
x e MQ ~
is d efin ed on
if/
0
= if/ and su ch that, for some fix ed com pact
U Q if
if/.
J( N
T h e area a s s o c ia t e d with su ch a
M
B y a variation of
T o do
M Q is any com pact k-dim en sio n al Riem annian
m anifold (w ith or without boundary) and let
T h at is ,
U ”
n Xt = i f / f a )| £ dt ^ t v ‘t = 0
T N ) then one can
12
LEON SIMON
d eriv e e x p re s s io n s for firs t and seco n d v ariatio n s (i.e . for d2 — d t2
and
^
^
) w hich are the sam e a s the right s id e s of ( 1 . 5 ), ( 1 .6 ), t= 0
but in w hich the vario u s quan tities must now be appropriately interpreted. (N e a r points hood
W
of
x Q e M Q , w here x Q into
N
J 0 A ) (x o) ^ 0 ,
if/
embeds a sm all n eigh b o r
and one com putes the form ulae for firs t and
seco n d v ariatio n by com puting the t-d e riv a tiv e s of the relevan t J aco b ian a s befo re.
On the other hand, the points w here
contribute nothing to the fo rm u la e.) sio n , it fo llo w s that if and only if
if/
near points w here
of co u rse
Of co u rse from the previou s d is c u s
U Q H dMQ = 0 ,
lo c a lly em beds
J 0 A )(x o) = 0
then
if/
is stationary in
U Q if
M Q a s a ze ro mean curvature subm an ifold
J (if/) ^ 0 .
§2 . k-varifolds, k-currents An exam ination of the d is c u s s io n in §1 w ill show that the form ulae (1 .5 ), (1 .6 ) (and their d e riv a tio n s) remain v alid even in the p resen ce of s erio u s s in g u la ritie s in
M.
T o make this statement p re c is e , w e first
need to introduce some term inology. We henceforth let
M b e a countably k-r e ctifia b le B o r e l s e t in
N .
oo T h at is , C
1
M
is B o re l,
subm an ifolds of
M C U M- , where j= l J N
( Mj
NL are (o p e n ) k-dim en sion al J
not n e c e s s a rily com plete nor p a irw is e
d is jo in t). ^ We a ls o a llo w the introduction of a multiplicity function of s p e c ific a lly , w e let w here
0
p be the m easure on
is a non-n egative lo c a lly
multiplicity function) on rather than ju st on
M,
of co u rse often have
N.
N
d efin ed by
M;
dp = ^ d K ^ ,
K ^-sum m able function (c a lle d the
(We take
6 to be d efined on a ll on
N ,
for re ason s of purely te ch n ical co n ven ien ce; w e
6 = 0 on
N ~
M .)
^ M o d u lo sets of K k -m easure zero, this is equivalent to the requirement that M is contained in a countable union of im ages, under L ip sch itz maps, of compact subsets of R k .
(By [F H 1 , 3 .1 .1 6 ].)
SURVEY LECTURES ON MINIMAL SUBMANIFOLDS
T h e pair
(M, p )
w ill then be c a lle d a k -varifo ld in
N .
13
(T h is co rre
spo n d s e x a c tly to the d efin itio n of k -dim en sio n al re c tifia b le v arifo ld as d efin ed by A lla r d [A W 1 ]; one s e e s this by virtue of [AW 1, 3 .5 (1 ), 2 . 8 (5 )]. / * ^ is c a lle d an integral k -varifo ld if the m ultiplicity function
QA, p)
is
0
in tege r-v a lu ed . T he support, denoted (2 .1 )
spt(M , p ) ,
spt(M , p ) = N -
w here the union is over a l l open spt(M , f ) C M
(c lo s u re of
true.
spt(M , p )
In fac t
d efin ed by
/x L M .
p L M (A ) = p Q A f \ A ) . )
1
su ch that
subm an ifold of
denoted
s in g (M ,/ z ),
(2 .2 )
(/xLM
p (M )
is the (o u te r) m easure on N
is c a lle d the mass of
(M , p) .
is d efin ed to be the s et of
x
in
is a properly em bedded k-dim en sion al
for som e open
W
co ntaining
x.
The singular set,
is d efin ed by
sin g (M , p ) = s p t (M , p )
G iv en
N o tic e that
N ), but eq u a lity may be far from
reg(M ,^z),
spt(M , p ) fl W N
^ ( M f lW ) = 0 .
is ju st the support (in the u su a l s e n s e of m easures
T he regular set, denoted spt(M , p)
is d efin ed by
U W
W su ch that
M taken in
[F H 1 , Ch. 2 ]) of the m easure
C
of the k -v arifo ld (M, p )
a k -v arifo ld
(M, p )
in
N , and a smooth map
other Riem annian m an ifo ld ) such that the image varifold i/a#(M, p )
(2 .3 )
reg(M , p ) .
i/r|spt(M, p )
if/ : N
N
( N
any
is p r o p e r / ^ w e d efin e
to be the k -varifo ld in
N
given by
(z).
zcif/ 1(y)riM F or an alternative description of A lla r d ’s work, s ee [S L 5 ]. ^T hat is,
is compact whenever
K C N
is compact.
if/
if/^p 6 (y) =
is
LEON SIMON
14
N o tic e that if
is a diffeom orphism then
if/
m ultiplicity function
^ #(M, p )
has the
6 = 6°ifz~^ .
We then have (by virtue of the area formula for k -re c tifia b le s e ts — [F H 1 , § 3 .2 ]), that (w ith
U
open such that
l0 t# (M n u ,/ z )|
=
ft(M D U ) < °o )
f
] ( t f x )d p
MflU w h en ever
| TXN
v a lidity
if/ one c h o o s e s to u s e ).
fd K k ,
a diffeom orphism
( 2 -4 ) is independent of
O f co u rse
here
d x ifj denotes
if/ .
U s in g stan dard, m easu re-theoretic fa c ts about d e n s itie s ([F H 1 ,
2 .1 0 .1 9 (4 )] and the fact that
alm ost a l l
f e A , in c a s e
meas lim ---------------- p - ------ = 1 P±° m e a s (B ^ ))
AC
for L e b e s g u e
is L e b e s g u e m easu rab le), together
w ith the d efin itio n of coun tably k -re c tifia b le s et given a b o v e, one e a s ily ch e ck s that such an approxim ate tangent s p a c e e x is t s
p - a .e . in
We thus h av e , by e x ac tly the com putations of §1, that
M.
SURVEY LECTURES ON MINIMAL SUBMANIFOLDS
15
^ l * t# ) d ^ .
to be stationary in
U
( re s p e c t iv e ly
U ) if the integral on the right of ( 2 .5 ) v a n is h e s (re s p e c t iv e l y if
the right s id e of (2 . 5 ) v a n i s h e s and the right of (2.6 ) > 0 ) for a l l smooth vector fie ld s
X
w hich have compact support in
U.
N o t ice a l s o that a g a in the first term on the right of ( 2 .6 ) ca n a l w a y s be d ele ted in c a s e
(M , n )
is stationary in
U.
Fo r later ap p lic a tio n s it is conven ient here to introduce a s lig h t gen eraliz atio n of the notion of “ s ta t io n a r y ’ ’ . has g en e ra liz ed m ean-curvature v ecto r
I
d i v MXd/x = -
M
whenever
X
Evid ently
(M, /i)
M
P
in U
if
H c L^/i, U )
and if
M
is statio nar y in U,
U
U.
if and only if it has z e ro gen eraliz ed
and (by the c l a s s i c a l formula (1.13 )), in
is a compact smooth manifold with
dMHU = 0
fi = K k L M ), the g e n e ra liz e d mean curvature vector of c o in c id e s with the c l a s s i c a l mean curvature vector of on
(M,/x)
< X , H > d ii
is a smooth vect or field with compact support in
mean curvature vector in case
H
N am ely, w e s a y that
(and in c a s e (M ,K k L M )
M a s a submanifold
N . Our main reason for introducing this notion is the fo llo w in g remark.
LEON SIMON
16
REMARK 2.1.
If
N
is (without lo s s of generality, by [ N J ] ) as sum ed to
be is omet rically embedded
in RP ,
consid ered as a k-varifold
in N ,
k-varifold in
RP ,
at each point
x
(M, fi)
of
has
U C RP
is open,
is stationary in
and
U fl N ,
if
then,
gen eraliz ed meancurvature vector
(M, f i ) , as a
H which
M fl U s a t i s f i e s |H| < k|BN | ,
(2.7 )
where
|BN | denotes the length of the second fundamental form of
a subm anifold of
X
on N
vecto r fie ld on
N
(as
RP ).
T o s e e th is, w e note that fie ld
if
f
M
d iv MX dfj. = 0 for any smooth vecto r
with com pact support in
U H N . Thus if
RP with com pact support in
(2 .8 )
I
U,
X
is a smooth
then
d i v u ( X - X 1 ) d fi = 0 ,
M w here
X^
.
P
X
=
denotes the component of
2 i s a , where a -n + l
normal s e t of vecto rs in
RP
X
normal to
N . Then (s in c e
w e have
a -
Thus, lo c a lly ,
i/P is a sm oothly varyin g ortho-
normal to
i= l
N.
n+1
k
P
i=l
a=n+l k
-< x . 2 V i ^ i ^ 1=1
= 0 ),
SURVEY LECTURES ON MINIMAL SUBMANIFOLDS
w here
BN
is the seco n d fundam ental form on
N
17
a s a subm an ifold of
RP.
T h u s (b y (2 .8 ))
J d i v MX
= - J d fi ,
k and hence
B lsJ(V -,r -)
2
i= l QA,fi)
1
is the ge n e ra liz e d mean curvature vector of
1
(a s a k -v arifo ld in
RP ) on
U.
(2 .7 ) evid en tly fo llo w s .
We w ill a ls o have o c c a s io n s u b seq u e n tly to co n sid er k-currents. k-current in
N
is sim ply an oriented integral k -varifo ld in
an orientation a
•••
£. at
a
H ere, by an orientation ^-alm o st a ll points
x (M
£ w e mean that
is
f(x ) =
(w here
orthonormal b a s is for the approxim ate tangent s p a c e £
That is , a
(M, (j l ) , with in tege r-va lu ed m ultiplicity function, together with
k-v arifo ld
± ^
N.
A
is any TXM ), and a ls o that
^ -m easu rab le w hen co n sid ered lo c a lly (v ia coordinate charts of
a s a m apping from
W fl M
( W
a coordinate neighborhood of
N )
N ) into
A k (R n) . The notion of k-current c o in c id e s exa c tly with the notion of k -d im en sio n al lo c a lly re c tifia b le current in the s e n s e of F e d e re r and F le m in g [ F F ] , and F e d e re r [ F H l ] .
T h is becom es cle a re r (it is form ally
proved from [ F H l , 4 .1 .2 8 ]) when w e note that, a s s o c ia t e d w ith each k-current
(
M
smooth k-forms
i
n
N,
w e have a lin ear fu n ction al
co with com pact support in
(2 .9 )
T(fi>) =
I
N
T
d efin ed on the
by
< (* ,£ > d m
M
here
( x ) = < t u ( x ) , f ( x ) >
denotes the d u al p airin g betw een
k-co vectors and k-vectors on
TXN .
thus id en tify in g the k-current
(M,/z, £ )
tion al on k-form s.
We w ill often w rite
T = ( M , / * ,£ ),
w ith the a s s o c ia t e d linear fu n c
B y an a lo gy w ith S to k e s ’ Theorem , w e can d efin e
LEON SIMON
18
another fu n ctio n al (c a lle d the boundary of
(2.10)
T ) by
(< 3 T )(w ) = T ( d o )
for a l l smooth (k -l)-fo r m s
co w ith compact support in
N.
N o tic e that
(9 T , s o d efin ed , may or may not represent a (k -l)-c u r r e n t on
N.
In fact
there is a very u s e fu l theorem due to F e d e re r and F lem in g ( [ F H l , 4 .2 .1 6 ]) w hich a s s e rt s that if is fin ite and if
T = (M, p, f )
is a k-current and if the mass of
dT
spt