Selected chapters in classical

minimal surface theory

by
RABAH SOUAM
The purpose of these notes is to give a short introduction to the theory of minimal surfaces
in R
3
through some selected topics. We present the Weierstrass representation, the solution
to Plateaus problem and the Barbosa-do Carmo stability theorem.
1 Introduction
The theory of minimal surfaces in R
3
is a rich and old subject which has experienced a great
development during the last 30 years and is still an active research eld.
The study of minimal surfaces in R
3
started with Lagrange in 1762 who considered the
problem of determining a graph with least area over a given domain in R
2
and taking pre-
scribed values on the boundary. He derived the partial differential equation to be satised by
the solution. It was Meusnier in 1776 who recognized the geometric meaning of that equa-
tion: the mean curvature vanishes, H 0. Hence the term minimal used for surfaces with
vanishing mean curvature. During the nineteenth century, many authors, among which Bon-
net, Enneper, Riemann and Weierstrass contributed to the study of minimal surfaces. The
strong link with complex analysis was in particular established then through the so-called
Weierstrass representation. The Belgian physicist J. Plateau (1801-1881) observed that min-
imal surfaces can be realized as soap lms. The problem of detemining a minimal disk
bounded by a given Jordan curve was then called Plateaus problem. It was only in 1930
that the problem was solved independently by J. Douglas and T. Rado. During the last 30
years, the theory of minimal surfaces has focused on global properties of complete minimal
surfaces, especially embedded ones.
We present in these notes some classical results in the theory. This is not a survey paper.
We have tried to make this text as elementary and as self-contained as possible and to give in
particular complete proofs of the selected results. This clearly limited the possibilities. We
hope these notes will give some motivations for beginners to go further in the study of the
69
70 G EOM ETRIES ET DYNAMIQUES
rich subject of minimal surfaces in R
3
...and in other ambient spaces. Introductory books to
the subject are [12], [1] and [9]. For more recent results in the theory one may consult the
survey papers [8], [10], [13], [14].
We have organized this text as follows. In Section 2 after recalling some premilinary
facts about the geometry of surfaces, we derive the Weierstrass representation in its local and
global versions and present some classical examples of minimal surfaces. We also derive
the rst variation formula for the area functional. Section 3 is devoted to the solution of
Plateaus problem. We have followed closely the presentation in the book by B. Lawson,[9].
In Section 4, we discuss the stability problem for minimal surfaces. A minimal surface is
stable if it minimizes the area up to the second order for boundary preserving variations. We
give a simplied proof of the Barbosa-do Carmo stability criterion, following an observation
of D. Fischer-Colbrie and R. Schoen, [6].
Acknowledgments: I am grateful to Francisco Martin for allowing me to use some of his images
and to the referee for valuable observations.
2 The Weirestrass representation and the rst variation
formula
We rst reviewquickly some basic denitions about the geometry of immersed surfaces in the
Euclidean space R
3
and x some notations. We denote by , ) the Euclidean inner product.
Let S be a surface and X : S R
3
an immersion, i.e a smooth map such that dX
p
has
rank 2 at each point p S. The rst fundamental form I of X is the Riemannian metric on S
dened as follows:
I
p
(w
1
, w
2
) = dX
p
(w
1
), dX
p
(w
2
)), for p S, w
1
, w
2
T
p
S.
When S is orientable, let N denote a unit normal eld to S, that is to say N : S S
2
R
3
is smooth and N(p) is orthogonal to dX
p
(T
p
S) for each p S. Here and in the sequel S
2
denotes the unit sphere centered at the origin in R
3
. The second fundamental form, denoted
II, is the eld of bilinear symmetric forms dened as follows:
II
p
(w
1
, w
2
) = dN
p
(w
1
), dX
p
(w
2
)), for p S, w
1
, w
2
T
p
S.
The shape operator B of the immersion X (oriented by the choice of the unit normal eld
N) is the eld of symmetric endomorphisms dened by:
II
p
(w
1
, w
2
) = I
p
(Bw
1
, w
2
), for p S, w
1
, w
2
T
p
S.
The mean curvature H of the immersion X and its Gaussian curvature K are dened as
follows:
H =
1
2
trace(B), K = det B.
The principal directions are the eigenvectors of B and the principal curvatures are its eigen-
values.
R. SOUAM 71
Denition 2.1 An immersed orientable surface in R
3
is called minimal if its mean curvature
is identically zero, i.e H 0
Note that, in particular, the Gaussian curvature of a minimal surface is nonpositive, K
0.
2. 1 The minimal surface equation
If the surface is expressed locally as a graph (x, y, h(x, y)), (x, y) of some function
h : R, then minimality means the function h satises in the minimal surface equation:
(1 +h
y
2
)h
xx
2h
x
h
y
h
xy
+ (1 +h
x
2
)h
yy
= 0.
As the reader can check, this may be rewritten as follows:
div
_
h
_
1 +[h[
2
_
= 0.
This is a quasilinear elliptic partial differential equation. A natural problem is to solve the
Dirichlet problem for the minimal surface equation, that is to say solve the equation over
some domain with a given boundary data. This is a well studied problem. The problem has
a solution for any boundary data if and only if the domain is convex. A reference for these
questions is the textbook [7].
2. 2 The Gauss map of a minimal surface
Denition 2.2 A differentiable map : S

S between two Riemannian surfaces is called
almost conformal if there exists a differentiable function on S such that for all p S and
all v
1
, v
2
T
p
S we have
d
p
(v
1
), d
p
(v
2
))
(p)
=
2
(p)v
1
, v
2
)
p
.
If the function is nowhere zero then is said to be conformal.
The geometric meaning of the above denition is that the (unoriented) angles are pre-
served by conformal maps. In fact, let : I S and : I S be two curves in S which
intersect at, say, t = 0. Their angle 0 at t = 0 is given by
cos =

)
[

[[

[
.
A conformal map : S

S maps these curves into curves : I

S, : I

S,
which intersect for t = 0, making an angle 0

satisfying:
cos

=
d(

), d(

))
[d(

)[[d(

)[
=

2

2
[

[[

[
= cos ,
as we claimed. Conversely, a diffeomorphism preserving the angles is conformal. A holo-
morphic map with nowhere zero derivative dened on an open subset of C is conformal.
Conversely, an orientation preserving conformal map between open subsets of C is holomor-
phic. To check these properties the reader is invited to solve the following exercise:
72 G EOM ETRIES ET DYNAMIQUES
Exercise. Let L : V W be a linear map between two Euclidean spaces V, W of the same
dimension. The following properties are then equivalent:
L preserves orthogonality.
L preserves the angles.
There exists a constant > 0 such that for each w
1
, w
2
V, L(w
1
), L(w
2
))
W
=

2
w
1
, w
2
)
V
.
Furthermore in dimension 2, this is equivalent to say that in orthonormal bases, the matrix of
L is of the form:
_
a b
b a
_
or
_
a b
b a
_
.
Proposition 2.3 Let X : S R
3
be an immersed connected orientable surface in the
Euclidean space. Then its Gauss map: N : S S
2
is almost conformal if and only if X is
minimal or X(S) is a subset of a round sphere.
Proof. Let p S and w
1
, w
2
T
p
S. Since dN
p
(w
1
) dX
p
(T
p
S), we have by dention of
the shape operator:
dN(w
1
), dN(w
2
)) = dN(w
1
), dX(Bw
2
)) = dX(Bw
1
), dX(Bw
2
)) = I(Bw
1
, Bw
2
).
So N is almost conformal if and only if there exists a function such that the shape operator
satises:
B
2
=
2
Id.
Now B is a solution of its characteristic polynomial, so:
B
2
2HB +KId = 0.
So almost conformality is equivalent to the condition that B satises:
2HB = (
2
+K)Id (1)
for some function . This clearly shows minimal surfaces (and spheres) have almost confor-
mal Gauss maps. Conversely, suppose the Gauss map is almost conformal and the surface is
not minimal. Consider a connected component V of the open set H ,= 0. Then (1) shows
the immersion is totally umbilic on V and hence X(V ) is a subset of a sphere (cf. [4]). In
particular the mean curvature of X[
V
is a nonzero constant. This shows V is closed and so
V = S.
Note that since a minimal surface has nonpositive Gaussian curvature, its Gauss map is
orientation reversing when the unit sphere is oriented by the exterior normal (cf. [15]).
R. SOUAM 73
2. 3 The Weierstrass representation
Let X : S R
3
be an immersed surface in the Euclidean space. S is thus endowed with
the induced metric ds
2
. Recall the classsical result that around each point of S we can nd
conformal coordinates, that is coordinates (u, v) on which the metric takes the form:
ds
2
=
2
(du
2
+dv
2
),
for some positive function > 0. We have the following characterization of minimal surfaces
given in conformal coordinates (see [15] for the details).
Proposition 2.4 Let X : S R
3
be an immersed surface oriented by a unit normal N and
let (u, v) be local conformal coordinates on S. Then in these coordinates:
X = 2H
2
N.
In particular the surface is minimal if and only if its coordinates functions X
1
, X
2
, X
3
are
harmonic in any conformal coordinates.
Consider the complex valued functions:

k
(z) =
X
k
u
i
X
k
v
, z = u +iv. (2)
We have the following identities:
3

k=1

2
k
(z) = [
X
k
u
[
2
[
X
k
v
[
2
2i
X
k
u
,
X
k
v
), (3)
3

k=1
[
k
(z)[
2
= [
X
k
u
[
2
+[
X
k
v
[
2
. (4)
As a consequence, we have the:
Proposition 2.5 Let X : (u, v) R
3
dene a minimal immersion with (u, v) conformal
coordinates. Then the functions
k
(z) dened by (2) are analytic functions of z and satisfy
the conditions:
3

k=1

2
k
(z) = 0 (5)
3

k=1
[
k
(z)[
2
,= 0. (6)
Conversely, given
1
(z),
2
(z),
3
(z) analytic functions of z satisfying the conditions (5) and
(6) in a simply connected domain D C. Then:
X(z) = 'e
_
z
z0
(
1
(z),
2
(z),
3
(z))dz, (7)
where z
0
D is any xed point, denes a conformal minimal immersion satisfying (2).
74 G EOM ETRIES ET DYNAMIQUES
Proof. If X is minimal then the statements are true as direct consequences of (2), (3) and
(4). Conversely, given
1
,
2
,
3
analytic functions, then X dened by (7) is well dened:
the integral does not depend on the choice of the path joining z
0
to z since D is simply
connected, and the identities (2), (3) and (4) are satised. By (3) and (5) the coordinates
(u, v) are conformal. By (6) and (4) X is an immersion. Furthermore, (2) and analyticity of

1
,
2
,
3
, imply the harmonicity of the coordinate functions X
1
, X
2
, X
3
. Therefore by the
Proposition 2.4, the immersion X is minimal.
The previous local fact admits the following globalization. Given an orientable Rieman-
nian surface, the local existence of conformal coordinates implies the existence of a complex
structure on the surface compatible with the metric. An orientable Riemannian surface has
therefore an (unique) underlying structure of a Riemann surface compatible with its metric.
This is in particular true for an orientable surface minimally immersed in R
3
. If z = u + iv
is a local conformal coordinate, then one considers the holomorphic 1-forms:

k
=
X
k
u
i
X
k
v
dz = 2
X
k
z
dz, k = 1, 2, 3. (8)
It is straightforward to check that
1
,
2
,
3
are globally dened holomorphic 1-forms.
Given a Riemann surface S, a map X : S R
3
is said conformal if around each point
one can nd a local conformal coordinate z = u +iv, such that:
[
X
k
u
[
2
= [
X
k
v
[
2
and
X
k
u
,
X
k
v
) = 0.
It can be checked this is independent on the choice of the local conformal coordinate. We
have then the following:
Proposition 2.6 Let S be a Riemann surface and X : S R
3
a conformal minimal
immersion. The holomorphic 1-forms dened by (8) satisfy the following:
3

k=1

2
k
= 0, (9)
3

i=1
[
k
[
2
,= 0. (10)
Conversely, given a Riemann surface S and 3 holomorphic 1-forms
1
,
2
,
3
satisfying (9)
and (10) and the period condition:
'e
_

(
1
,
2
,
3
) = 0,
for any closed path in S, then the map:
X(p) = 'e
_
p
p0
(
1
,
2
,
3
),
for p S, with p
0
S any xed point denes a conformal minimal immersion.
R. SOUAM 75
Dene:
g =

3

1
i
2
, =
1
i
2
,
g is then a meromorphic function and holomorphic 1-form and the identity (9) becomes:

1
=
1
2
(1 g
2
),

2
=
i
2
(1 +g
2
),

3
= g
and it can be checked that:
N(z) =
_
2'e(g(z))
1 +[g(z)[
2
,
2m(g(z))
1 +[g(z)[
2
,
1 [g(z)[
2
1 +[g(z)[
2
_
.
This means that the meromorphic function g is the sterographic projection, from the point
(0, 0, 1) of the unit sphere, of the Gauss map N of the immersion X. We recover the fact that
the Gauss map of minimal surfaces is almost conformal (Propostion 2.3).
We have:
3

k=1
[
k
[
2
=
1
2
(1 +[g[
2
)
2
[[
2
, (11)
and (10) means the zeroes of coincide with the poles of g, but with twice order. So the
Proposition 2.6 can be reformulated as follows:
Theorem 2.7 (The Weirerstrass representation) Let X : S R
3
be a minimal immersion
of an orientable surface S. Let g = N be the composition of the stereographic projection
from the point (0, 0, 1) of the sphere to the extended complex plane C , with the
Gauss map N of X. Then g is meromorphic and there exists a holomorphic 1-form on S
such that:
X(p) X(p
0
) = 'e
_
p
p0
_
1
2
(1 g
2
),
i
2
(1 +g
2
), g
_
,
for p, p
0
S, the integration being taken on any path from p
0
to p. Moreover the zeroes of
coincide with the poles of g and have twice order.
Conversely, let S be a Riemann surface and g : S C a meromorphic function
and a holomorphic 1-form on S. Assume the zeroes of coincide with the poles of g and
have twice order. Assume also the following period condition is satised:
'e
_

_
1
2
(1 g
2
),
i
2
(1 +g
2
), g
_
= 0,
for every closed path on S. Then the map: X : S R
3
, dened by:
X(p) = 'e
_
p
p0
_
1
2
(1 g
2
),
i
2
(1 +g
2
), g
_
,
where p
0
is any xed point in S, is a conformal minimal immersion. Moreover the Gauss map
of X is N =
1
g.
76 G EOM ETRIES ET DYNAMIQUES
We can express all the geometric quantities associated to a minimal immersion in terms
of its Weierstrass data (g, ). If z is a local conformal coordinate with = f(z)dz, then, in
particular, the metric and the Gaussian curvature are expressed as follows:
ds
2
=
1
4
[f(z)[
2
(1 +[g(z)[
2
)
2
[dz[
2
, (12)
K(z) =
_
4[g

(z)[
[f(z)[(1 +[g(z)[
2
)
2
_
2
, (13)
(12) follows from (11) and (4). For (13) see [12].
2. 4 Classical examples
The Weierstrass representation is an efcient tool to construct minimal surfaces. From the
geometric point of view one is interested in complete surfaces. This implies some condi-
tions on the behaviour at innity of the Weierstrass data (g, ) in view of equation (12). The
compatibility condition between the zeroes and poles of the Weierstrass data (g, ) is easy to
check. The period condition is usually the most difcult condition to verify in practice. How-
ever on a simply connected Riemann surface it is automatically satised. We present here
some examples of complete minimal surfaces and refer to the literature for more examples
(in particular [1] which contains a detailed description of many examples).
The plane: S = C, g = 0, = dz
Ennepers surface: (gure 1) S = C, g(z) = z, = dz.
Figure 1: Ennepers surface
The catenoid: (gure 2) S = C 0, g(z) = z, =
dz
z
2
.
The helicoid: (gure 3) S = C, g(z) = e
z
, = ie
z
dz.
Costas surface (gure 4) S = (C/Z Z) 0,
1
2
,
i
2
, g(z) =
2

2P(
1
2
)
P

, =
Tdz, where T is the Weierstrass function of the unit square.
R. SOUAM 77
Figure 2: The catenoid
Figure 3: The helicoid
78 G EOM ETRIES ET DYNAMIQUES
Figure 4: Costas surface
2. 5 The rst variation formula
We now see that minimal surfaces are critical points of the area functional thus justifying
their name. Let X : S R
3
be an immersion of an orientable surface in R
3
. Let D S be
a compact domain with smooth boundary in S. By a variation X
t
of X
|D
we mean a smooth
mapping:
F : D (, ) R
3
such that X
t
:= F(t, .) is an immersion for each t and X
0
= X.
The variation vector eld associated to the variation X
t
is the vector eld along the
immersion dened by:
(p) =
X
t
t
[
t=0
(p), p D.
The variation is said normal if is orthogonal to the surface, i.e = N for some function
C

(D). Finally we say that the deformation preserves the boundary if, for all t, X
t
= X
on D. Let (

0
(D) denote the set of functions on D which are smooth up to the boundary
and vanishing on D. The variation vector eld associated to a normal boundary preserving
variation is of the form = N, where (

0
(D).
Conversely, given (

0
(D), consider the variation X
t
= X+N. For t small enough,
X
t
is an immersion by compactness of D and continuity. It is, moreover, clearly normal and
boundary preserving. So at the innitesimal level a normal boundary preserving variation is
the same thing as a function (

0
(D).
Denition 2.8 (Divergence) Let X : S R
3
be an immersion and V : S R
3
a (smooth)
vector eld along X, that is to say:
V (p) dX
p
(T
p
S),
R. SOUAM 79
for all p S. For w T
p
S, denote by dV (w)

the orhtogonal projection of dV (w) on

dX
p
(T
p
S). The divergence of V, denoted div V is the trace of this map:
TS TS
w dV (w)

.
Otherwise said: for p S, let e
1
, e
2
be an orthonormal basis of T
p
S, then
div V (p) = dV (e
1
), dX(e
1
)) +dV (e
2
), dX(e
2
)).
If now V is a vector eld on S, then its divergence is by denition the divergence of dX(V ).
Let denote the area form of the surface X (see [15]). Recall that for w
1
, w
2
TS, we
have
(w
1
, w
2
) = det(dX(w
1
), dX(w
2
), N).
Let V be a vector eld along the immersion X, we dene a 1-form i
V
as follows:
i
V
(w) = det(V, dX(w), N).
Proposition 2.9 d(i
V
) = (div V ) .
Proof. We show the equation is veried at each point of S. Let p S and let (u, v) be local
coordinates in S around p compatible with the orientation. In these coordinates the 1-form
i
V
writes:
i
V
= i
V
(X
u
) du +i
V
(X
v
) dv = det(V, X
u
, N) du + det(V, X
v
, N) dv.
So:
d(i
V
) =
_

u
det(V, X
v
, N)

v
det(V, X
u
, N)
_
du dv.
We have:

u
det(V, X
v
, N) = det(V
u
, X
v
, N) + det(V, X
vu
, N) + det(V, X
v
, N
u
)
= det(V
u
, X
v
, N) + det(V, X
vu
, N),
because N
u
is tangent to the surface. Similarily:

v
det(V, X
u
, N) = det(V
v
, X
u
, N) + det(V, X
uv
, N).
Therefore:
d(i
V
) = det(V
u
, X
v
, N) det(V
v
, X
u
, N) du dv.
We may assume the coordinates (u, v) are such that the basis X
u
, X
v
is orthonormal and
positively oriented at p. This is always possible, take any coordinates and perform a linear
change of coordinates. Then at p we have:
d(i
V
)(p) = V
u
, X
u
)(p) +V
v
, X
v
)(p) det(X
u
(p), X
v
(p), N(p))du dv(p).
So:
d(i
V
)(p) = div V (p) (p).
80 G EOM ETRIES ET DYNAMIQUES
We are now in position to prove the:
Proposition 2.10 (The rst variation formula) Let X
t
be a boundary preserving variation of
an immersion X : D R
3
, where D is a compact orientable surface, oriented by a choice
of a unit normal eld N. Call A(t) the area of X
t
. Then:
dA
dt
(0) = 2
_
D
HdA,
where H is the mean curvature of the immersion and =
Xt
t
[
t=0
, N).
Proof. We denote by N
t
the unit normal eld associated to the immersion X
t
and depending
smoothly on t, with N
0
= N and by dA
t
its area element. We have:
A(t) =
_
D
dA
t
,
so we need to compute the derivative of the area element dA
t
. Let p D and let (u, v) be
local coordinates around p in S compatible with the orientation. Then:
dA
t
= det((X
t
)
u
, (X
t
)
v
, N
t
) dudv.
Set

N =
dNt
dt
[
t=0
. We have:
d
dt
det((X
t
)
u
, (X
t
)
v
, N
t
)[
t=0
= det(
u
, X
v
, N) + det(X
u
,
v
, N) + det(X
u
, X
v
,

N).
It follows from N
t
, N
t
) = 1 that

N is tangent to the surface and so the last term in the
right-hand side of the previous equation vanishes. Write:
= N +V
where V is the tangent part to the immersion of the deformation vector eld . Then:

u
= d(

u
)N +N
u
+V
u
,
v
= d(

v
)N +N
v
+V
v
.
Therefore:
d
dt
det((X
t
)
u
, (X
t
)
v
, N
t
)[
t=0
= det(N
u
, X
v
, N) + det(X
u
, N
v
, N)
+det(V
u
, X
v
, N) + det(X
u
, V
v
, N).
Now as in the previous Propositon we may assume the basis X
u
, X
v
is orthonormal and
positive at the point p and so:
d
dt
det((X
t
)
u
, (X
t
)
v
, N)[
t=0
(p) = (p) N
u
, X
u
)(p) +N
v
, X
v
)(p)
+V
u
, X
u
)(p) +V
v
, X
v
)(p)
= 2(p)tr(dN(p)) + div V (p)
= 2(p)H(p) + div V (p).
R. SOUAM 81
It follows that:
dA
t
dt
(p) = 2(p)H(p) + div V (p) dA(p).
Since this is true for every p D, we conclude that:
dA
dt
(0) =
_
D
2H + div V dA.
Since V = 0 on D, the 1-form i
V
vanishes on D. By Stokes theorem:
_
D
div V dA =
_
D
d(i
V
) =
_
D
i
V
= 0,
which completes the proof.
Corollary 2.11 An immersion X : S R
3
is minimal if and only if for every variation X
t
of X with compact support, the derivative of the area A(t) vanishes at t = 0 :
dA
dt
(0) = 0.
3 The Plateau problem
Let be a Jordan curve in R
n
, i.e a continuous curve which is homeomorphic to the circle
S
1
. The Plateau problem asks to nd the surface of least area spanning . The problem was
solved independently by Douglas and Rado, in 1930, for disk-type surfaces. In order to
formulate the result, we introduce some notations: B = (u, v) R
2
, u
2
+ v
2
1 will
denote the closed unit disk in the Euclidean plane.
A map X : B R
n
is said to be piecewise C
1
if it is continuous and if it is C
1
except
along B and along a nite number of regular C
1
arcs and points in

B . A continuous map
: B is said to be monotone if it is such that if B is traversed once in the positive
direction, then is traversed once also in a given direction, although we allow arcs of B to
map to single points of . Otherwise said, a monotone map is a continuous map of degree 1
such that for each p ,
1
(p) is connected.
Let (() = X : B R
n
, X is piecewise C
1
and X[B is a monotone parameterization
of We dene the area functional A : (() R by the following (maybe improper)
integral:
A(X) =
_
B
[X
u
X
v
[dudv
where [X
u
X
v
[
2
= [X
u
[
2
[X
v
[
2
X
u
, X
v
)
2
. Let
/

= inf
XC()
A(X).
Therefore our problem is to nd a X (() such that A(X) = /

.
Note that we have to restrict ourselves to curves satisfying /

< since there are

examples of Jordan curves bounding no surfaces of nite area. If is rectiable, for instance
C
1
, then it is true that /

< (cf.[12]).
82 G EOM ETRIES ET DYNAMIQUES
The idea is to try to apply the direct method of the calculus of variations, that is take a
sequence X
i
(() such that A(X
i
) /

and try to extract a subsequence wich converges

to an X

(() with A(X

) = /

. However this cannot work this way. The reason is

that the area functional is invariant under a big group of transformations: for any piecewise
C
1
diffeomorphism of the disk B, and any X /

we have A(X ) = A(X). We can

nd a sequence of smooth homeomorphisms
i
of the unit disk B such that
i
(re
i
)
e
i
uniformly on compact subsets of B 0 and
i
(0) 0. To get such an example
just construct such a sequence of diffeomorphisms of the unit interval and then use polar
coordinates. Take for instance the sequence of homeomorphisms of the unit interval:
_

_
l
i
(t) = t if 0 t
1
i+1
,
l
i
(t) = (i
2
i 1)t
i
2
i2
i+1
if
1
i+1
t
1
i
,
t+i2
i1
if
1
i
t 1,
which is piecewise smooth and perturb it a bit to make it smooth. Then assuming X
i
con-
verges uniformly to X

one can check that no subsequence of X

i

i
converges to a con-
tinuous map.
So we need to control the parameterizations of our minimizing sequences. In the one-
dimensional case, i.e curves in a Riemannian manifold, this is done by minimizing the energy
integral. Here minimizing sequences of curves tend to geodesics which not only minimize
the length integral but which are forced to be parameterized by a multiple of arc-length. In
the case of surfaces the corresponding energy is the so-called Dirichlet integral:
D(X) =
_
B
([X
u
[
2
+[X
v
[
2
)dudv.
Note that for any vectors v, w R
n
we have:
[v w[
2
= [v[
2
[w[
2
v, w)
2
[v[
2
[w[
2

1
4
([v[
2
+[w[
2
)
2
,
and equality holds if and only if [v[ = [w[ and v, w) = 0. Therefore, for any X ((),
A(X)
1
2
D(X)
with equality if and only if
[X
u
[ = [X
v
[ and X
u
, X
v
) = 0
almost everywhere in B. Wherever [X
u
[ > 0, such a map is conformal and induces a metric
on B of the form
ds
2
= (du
2
+dv
2
)
where = [X
u
[ = [X
v
[. The parameters (u, v) are thus conformal coordinates for the
surface. A theorem of fundamental importance is the following:
Theorem 3.1 (Conformal representation of disk-type domains) Let X : B R
n
be a
continuous map such that X[

B is an immersion of class C
k
, 1 k (or real analytic).
Then there exists a homeomorphism : B B where [

B is of class C
k
(or real analytic)
such that the parameterized mapping X

= X is conformal.
R. SOUAM 83
Consider a Jordan curve R
n
and dene:
T

= inf
XC()
D(X).
Clearly:
/

1
2
T

.
Now take a sequence X
k
(() such that A(X
k
) /

. The X
k
are not necessarily im-
mersions so we cannot apply directly the Theorem 3.1. The trick is to consider the sequence
of mappings:

X
k
: B R
n+2
, dened by

X
k
(u, v) = (X
k
(u, v),
1
k
u,
1
k
v).
Clearly the

X
k
are C
1
immersions (even embeddings) and so there exist homeomorphisms

k
of B given by the Theorem 3.1 such that the

X
k

k
are conformal. Using the invariance
of area under reparametrizations and some elementary estimates, we get:
T

D(X
k

k
) D(

X
k

k
) = 2A(

X
k

k
) = 2A(

X
k
) 2A(X
k
) +
4
k
and it follows that
/

=
1
2
T

.
Corollary 3.2 Let R
n
be a Jordan curve with /

< . Then for any X ((),

D(X) = T

if and only if A(X) = /

and X is almost conformal.

Hence to solve the Plateau problem it is sufcient to nd a map X (() which mini-
mizes the Dirichlet integral. We will use the
Theorem 3.3 (Dirichlets Principle) Let : B R
n
be a continuous map and dene
(() = X : B R
n
: X is piecewise C
1
and X[B =
Assume that the number
T

= inf
XC

D(X)
is nite. Then there exists a unique map X

(() such that D(X

) = T

. The function
X

is harmonic in

B and represents the solution to the boundary value problem X =
0, X[B = .
We can thus consider a minimizing sequence X
k
((), for the Dirichlet functional
satisfying, X
k
is hamonic in

B . Still we cannot extract a converging subsequence, the rea-
son is that the Dirichlet energy is invariant under the action of the group of conformal dif-
feomorphisms of the disk B. Recall that the group G of conformal orientation preserving
diffeormorphisms of the disk:
G = g(z) = e
i0
a +z
1 + az
, a C, [a[ < 1,
0
R (14)
84 G EOM ETRIES ET DYNAMIQUES
is not compact. Indeed take a sequence a
k
C, [a
k
[ < 1, a
k
1, then the sequence in G,
g
k
(z) =
a
k
+z
1+ a
k
z
converges uniformly on compact subsets of B1 to the constant 1. So we
need to normalize the mappings X ((). This is done by imposing a three-point condition:
choose three distinct points p
1
, p
2
, p
3
and three distinct points z
1
, z
2
, z
3
B, and
dene
(

() = X (() : X(z
k
) = p
k
, k = 1, 2, 3.
By the invariance of the Dirichlet integral under G, we have that
infD(X), X (

() = T

,
and thus we may solve the Plateau problem by minimizing in this smaller class. We thus take
a minimizing sequence X
k
(

(), we claim that the boundary values X

k
[B : B R
n
form an equicontinuous family. The key to this crucial result is the following
Lemma 3.4 (The Courant-Lebesgue lemma) Suppose X C
0
(B, R
n
) C
1
(

B, R
n
) and
satises
D(X) M (15)
for some M with 0 M < . For each z
0
B and r > 0 we dene C
r
to be the
intersection of B with the circle of radius r centered at z, and we denote by s the arc length
parameter of C
r
. Then for each 0 < < 1, there exists a number with

such
that
_
C
[X
s
[
2
ds
2M
log(
1

)
. (16)
Proof. For 0 < < 1, consider the integral
I =
_

_
Cr
[X
s
[
2
dsdr D(X) M,
and express I as
I =
_

p(r)
1
r
dr where p(r) = r
_
Cr
[X
s
[
2
ds.
Then by the mean value theorem for the measure d(log r), there exists a number [,

],
such that
I = p()
_

d(log r) = p()
1
2
log(
1

).
Thus, p() 2M/ log(
1

) as claimed.
We can now prove the following:
Proposition 3.5 Let M be a constant > T

. Then the family of functions

T = X[B : X (

() and D(X) M
is equicontinuous on B. Thus, by Ascolis theorem T is compact in the topology of uniform
convergence.
R. SOUAM 85
Proof. For z B and 0 < r < 1, denote by l(C
r
) the length of the curve C
r
dened in the
Lemma 3.4. It follows from the Schwarz inequality that
l(C
r
)
2
= (
_
Cr
[X
s
[ds)
2
2r
_
Cr
[X
s
[
2
ds.
Let 0 < < 1 be a xed number and X T. It follows from the Lemma 3.4 that there exists

such that
l(C
r
)
2

2M
log(
1

)
.
Let now > 0 be a given number. By an easy topological argument one sees that there
exists a number d > 0 such that for all p, q with 0 < [[p q[[ < d, one of the
two components of p, q will have diameter < . (Recall that the diameter of a set
S R
n
is diam(S) = sup[[p q[[, p, q S.) We now choose 0 < < 1 such that
_
4M/ log(
1

) < d and such that the disks of radius

centered at z
1
, z
2
, z
3
are pairwise
disjoint. We may assume without loss of generality that < min
i=j
[[p
i
p
j
[[. Then for
any z B, there exists by the Lemma 3.4 a real

such that l(C

) < d. B
is now divided by C

into two arcs: a small arc A

containing z and its complement A

containing two of the points z

1
, z
2
, z
3
. By the monotonicity of X[B, the images X(A

) and
X(A

) determine a partition of into two arcs and since l(C

) d one of them has diameter

< . However, since X(A

) contains at least two of the points p

1
, p
2
, p
3
, its diameter is > .
Hence, diameter (X(A

)) < . This immediately implies that for z, z

B, [z z

[ < ,
we have [[X(z) X(z

)[[ < and depends only on and the points z

1
, z
2
, z
3
, p
1
, p
2
, p
3
.
This shows the equicontinuity of the family X(A

)
Now by Ascolis theorem each sequence in X(A

) has a convergent subsequence for the

uniform convergence. To conclude note that a limit of monotone maps is also monotone.
Theorem 3.6 (Douglas, Rado) Let be a Jordan curve in R
n
such that /

< . Then
there exists a continuous map X : B R
n
such that
(1) X[B maps B monotonically onto ,
(2) X[

B is harmonic and almost conformal,

(3) D(X) = /

.
Proof. Let X
k
(

() be a minimizing sequence, lim

k
D(X
k
) = T

. By the Theorem
3.3 we can assume the X
k
harmonic in

B . By the Proposition 3.5, we can nd a subsequence,
still denoted X
k
, such that X
k
[
B
converges uniformly to a monotone parametrization :
B . Denote by X

the hamonic extension of to B given by the Dirichlet principle

(Theorem 3.2). Then the maximum principle for harmonic functions ensures that X
k
X

uniformly on B as k . Harnacks principle ensures that X

k
X

as k
uniformly on compact subsets of

B. Let then K be any compact subset of

B. Then:
_
K
[X

[
2
= lim
k
_
K
[X
k
[
2
liminf
k
D(X
k
).
86 G EOM ETRIES ET DYNAMIQUES
Letting K tend to B, we get:
D(X

) liminf
k
D(X
k
) = T

.
The map X is called a classical solution to the Plateau problem for .
Remark 3.1. The solution to Plateaus problem obtained this way is a branched minimal
surface, i.e. a non constant continuous map X : B R
n
which is almost conformal and
harmonic inside B. The Proposition 2.6, which easily extends to minimal surfaces in R
n
,
holds for such a map exactly as for immersed minimal surfaces except that the condition (10)
is lacking. The points where the map is not an immersion are called its branch points. They
are the zeroes of the holomorphic functions
1
, . . . ,
n
and so are isolated. R. Osserman
proved that, for n = 3, the solution is free of branch points inside B (the proof was incom-
plete and was completed by R. Gulliver). This is not the case for n 4, there are examples
where the solution has branch points. Another issue is the regularity at the boundary. Many
authors contributed to settle the problem. For more information one may see [12] or [9] and
the references therein.
It remains to check that the classical solution to Plateaus problem is a homeomorphism
on B. To see this we will need the following fact which is of independent interest:
Proposition 3.7 (The reection principle) Let X : (u, v) R
2
, u
2
+ v
2
< , v >
0 R
n
be a branched minimal surface. Suppose there exists a line L in R
n
such that
X(u, v) L when v 0. Then X can be extended as a branched minimal surface dened
in the whole disk (u, v) R
2
, u
2
+v
2
< by reection across L.
Proof. We may assume, after a rotation in space, that L is the x
n
axis. By the reection
principle for harmonic functions, the functions X
1
, X
2
extend to harmonic functions on the
whole disk by setting X
k
(u, 0) = 0, X
k
(u, v) = X
k
(u, v); k = 1, . . . , n 1. Thus the
functions

k
=
X
k
u
i
X
k
v
, k = 1, . . . , n 1
are holomorphic on the whole disk and imaginary on the real axis. The equation:

2
n
=
2
1
. . .
2
n1
,
shows
n
extends continuously to the real axis and takes nonnegative real values there.
Therefore
n
extends continuously to the real axis and has real values there. By the re-
ection principle for holomorphic functions, we can extend
n
to the whole disk by setting

n
( z) =

n
(z). Integrating X
n
(z) = 'e(
_
z
0

n
(z)dz), we see that X
n
extends to the whole
disk by the relation:
X
n
(u, v) = X
n
(u, v).
We can now prove the following:
R. SOUAM 87
Proposition 3.8 For each solution X : B R
n
to the Plateau problem for , the map
X[B : B is a homeomorphism.
Proof. Since X[
B
is monotone, the only way it fails to be a homoeomorphism is
that it sends an arc of B to a single point. Suppose this is the case. After a preliminary
conformal map of B onto the upper half-plane we can apply the reection principle (Proposi-
tion 3.7) and extend the surface to a branched surface beyond a segment of the real axis. The
branched surface thus obtained would send the whole segment to a point. This contradicts
the isolatedness of branch points of branched minimal surfaces (cf. Remark 3.1).
The solution to Plateaus problem and the reection principle give a way to construct
complete surfaces. One may start with a polygonal curve and solve the Plateau problem and
then use the reection principle to reect the solution across the edges on the boundary. One
can then do again reections across the new boundary edges and so on. Of course the surface
may be singular at the vertices of the initial boundary (and their images after reections). If
one chooses carefully the polygonal curve one can obtain a complete minimal surface which
is immersed everywhere. An example obtained starting from an appropiate quadrilateral in
space is a famous surface discovered by Schwarz (gure 5). It is triply periodic (i.e invariant
under three independent translations) and embedded.
The solution to Plateaus problem gives a surface which minimizes area among disk-
type surfaces. This does not mean the solution obtained has the smallest area among all the
possible surfaces bounded by the curve. The proof of the existence of a surface minimizing
the area among all possible surfaces without restriction on their topological type requires
the techniques of geomeric measure theory. This is a difcult but very powerful theory for
establishing existence and regularity of volume minimizing submanifolds and applies to more
general manifolds. A readable introduction to this subject is [11].
Figure 5: The Schwarz surface
88 G EOM ETRIES ET DYNAMIQUES
4 The stability of minimal surfaces
4. 1 The second variation formula
We are now interested in the second derivative of the area functional for minimal surfaces in
order to study those which are minimizing up to the second order. It is not true in general that
a minimal surface minimizes the area with respect to its boundary even among neighbooring
surfaces. We will see this happens for instance for some catenoid slices as a consequence
of the discussion to follow. However it can be proved that each point on a minimal surface
admits a neighborhood which is minimizing with respect to its boundary.
We will now derive the second variation formula. We rst need to recall some denitions.
Denition 4.1 (The gradient and the Laplacian) Let (S, , )) be a Riemannian surface and
f : S R a C
1
function on S. The gradient of f, denoted f, is the vector eld on S such
that:
f(p), w)
p
= df
p
(w),
for every p S and w T
p
S.
Let f C
2
(S). The Laplacian of f, denoted f, is the function dened by:
f = div(f).
With this denition, the Laplacian on the plane for the Euclidean metric ds
2
0
= du
2
+dv
2
is
0
=

2
u
2
+

2
v
2
. If the metric on a surface S is given in conformal coordinates by:
ds
2
=
2
(du
2
+dv
2
),
then the associated Laplacian is related to the Euclidean Laplacian as follows:
=
1

2

0
. (17)
We can now state the second variation formula. To simplify the presentation, we restrict
ourselves to normal variations, however the same formula holds in the general case.
Proposition 4.2 (The second variation formula) Let X
t
be a normal boundary preserving
variation of a minimal immersion X : D R
3
, where D is a compact orientable surface,
oriented by a choice of a unit normal eld N. Call the deformation vector eld of X
t
and
set = , N). Then the second derivative of the area is:
d
2
A
dt
2
(0) =
_
D
2K
2
dA
=
_
D
[[
2
+ 2K
2
dA.
Proof. The second equality follows from Stokes theorem and the esaliy checked identity:
div() = [[
2
+.
R. SOUAM 89
Denote by N
t
the unit normal eld to the immersion X
t
depending smoothly on t, with
N
0
= N and let H
t
and dA
t
be respectively the mean curvature and the area element of X
t
.
Finally put:
t
=
Xt
t
[
t
.
By the rst variation formula (Proposition 2.10), for each t:
dA
dt
(t) = 2
_
D

t
, N
t
)H
t
dA
t
.
Taking into account that the immersion X is minimal, i.e H
0
= 0, we get:
d
2
A
dt
2
(0) = 2
_
D

dH
t
dt
[
t=0
dA.
It remains to compute
dHt
dt
[
t=0
. We rst compute the derivative

N =
dNt
dt
[
t=0
. Let p S and
(u, v) local coordinates around p such that the basis X
u
, X
v
is orthonormal at p. For each
t we have:
N
t
, (X
t
)
u
) = N
t
, (X
t
)
v
) = 0,
so taking the derivatives:

N, X
u
) = N,
v
),

N, X
v
) = N,
u
).
Now,
u
= (N)
u
= d(

u
)N +N
u
and
v
= (N)
v
= d(

v
) +N
v
. So:

N, X
u
) = d(

u
),

N, X
v
) = d(

v
).
Moreover since N
t
, N
t
) = 1 we have:

N, N) = 0. It follows that at the point p :

N(p) = (p).
Since this is true for every p S, we conclude that:

N = (18)
Denote by B
t
the shape operator of the immersion X
t
, i.e the eld of endomorphisms of
TS dened by:
I
t
(B
t
w
1
, w
2
) = II
t
(w
1
, w
2
),
which means:
dX
t
(B
t
w
1
), dX
t
(w
2
)) = dN
t
(w
1
), dX
t
(w
2
)).
Put

B =
dBt
dt
[
t=0
. Taking the derivatives at t = 0 :
d(Bw
1
) +dX(

Bw
1
), dX(w
2
)) +dX(Bw
1
), d(w
2
)) = d(

N)(w
1
), dX(w
2
))
dN(w
1
), d(w
2
)).
90 G EOM ETRIES ET DYNAMIQUES
Writing = N and using (18), we obtain after simple calculations:
dN(Bw
1
), dX(w
2
)) +dX(

Bw
1
), dX(w
2
)) +dX(Bw
1
), dN(w
2
)) =
= d()(w
1
), dX(w
2
)) dN(w
1
), dN(w
2
)).
Finally:
dX(

Bw
1
), dX(w
2
)) = d()(w
1
), dX(w
2
)) +dX(Bw
1
), dX(Bw
2
)).
Take any point p S and a basis e
1
, e
2
of T
p
S formed of principal eigenvectors with
associated principal curvatures k and k (remember the surface is minimal). It follows that:
2
dH
t
dt
[
t=0
(p) = tr(

B)(p)
= dX(

Be
1
), dX(e
1
)) +dX(

Be
2
), dX(e
2
))
= div()(p) + 2k
2
(p).
This is true for any p S, so nally:
2
dH
t
dt
[
t=0
= 2K.
A deformation X
t
of an immersion is said trivial if X
t
= X, for every t.
Denition 4.3 A compact domain D S on a minimal immersion X : S R
3
is said
stable if for every nontrivial normal boundary preserving variation X
t
on D :
d
2
A
dt
2
(0) > 0.
Equivalently:
_
D
[[
2
+ 2K
2
dA > 0,
for all C

0
(D), not identically 0.
This denition means that the domain is minimizing up to the second order for boundary
preserving variations. For instance, a domain on a plane is clearly stable since the curvature
vanishes identically.
4. 2 The Barbosa-do Carmo stability criterion
The Barbosa-do Carmo theorem ([2]) gives a nice and easy to verify criterion for the stability
of minimal surfaces.
Theorem 4.4 (The Barbosa-do Carmo stability theorem) Let D be a regular compact do-
main on a minimal surface. Assume the area of the Gaussian image N(D) on the sphere is
less than 2. Then D is stable.
R. SOUAM 91
To prove the theorem we need to recall some basic facts about the fundamental Dirichlet-
eigenvalue of Schrodinger operators on surfaces. These notions make sense on Riemannian
manifolds of any dimension but we will restrict ourselves to the two-dimensional case. More
details can be found in [3] and [5].
Proposition 4.5 Let S be a Riemannian surface, q C

(S) a smooth function on S and

S an open regular domain with compact closure in S. Then the operator + q has a
discrete spectrum for the Dirichlet eigenvalue problem on . The eigenfunctions are smooth
up to the boundary of . The rst eigenvalue
1
( +q, ) is simple and is characterized by
the property:

1
( +q, ) = inf
fC
2
0
(), f=0
_

[f[
2
qf
2
dA
_

f
2
dA
. (19)
Moreover the associated nontrivial eigenfunctions realize the inmum in (19) and are
characterized, among other eigenfunctions, by the fact that they do not vanish inside D.
Furthermore, the rst eigenvalue enjoys a monotonicity property: if

S are two
open regular domains with compact closure and ,=

, then:

1
( +q, ) >
1
( +q,

).
Let us nd for instance the fundamental eigenvalue of the Laplacian on a hemisphere
on S
2
. We denote by
1
the Laplacian on S
2
. It sufces to consider the hemisphere S
2
+
=
(x
1
, x
2
, x
3
) S
2
, x
3
0. Then since S
2
has mean curvature 1 with respect to the out-
ward unit normal,the restriction h of the coordinate function x
3
to S
2
satises (cf. Proposition
2.4 and (17)):
_

1
h + 2h = 0 in S
2
+
,
h = 0 on S
2
+
.
The Proposition 4.5 then shows:

1
(
1
, S
2
+
) = 2. (20)
We will need the following result:
Theorem 4.6 (Faber-Krahn) Let be a regular domain on the unit sphere S
2
and let

be
a round disk having the same area as . Then:

1
(
1
, )
1
(
1
,

)
with equality if and only if is a round disk.
This theorem was proved by Faber and Krahn in the Euclidean case but the proof extends
to the spherical case (and to the hyperbolic one too). For the proof see [3] or [5].
The Theorem 4.6, the monotonicity property (cf. Proposition 4.5) and (20) then give the:
Corollary 4.7 Let be a domain in S
2
. Assume the area of is less than 2. Then:

1
(
1
, ) > 2.
92 G EOM ETRIES ET DYNAMIQUES
We now use the following result due to D. Fischer-Colbrie and R. Schoen [6] (which
holds in any dimension). We give an independent proof.
Lemma 4.8 Consider a Riemannian surface S and denote by the associated Laplacian.
Let q be a smooth function on S and D S be a compact domain in S with smooth boundary.
Assume there exists a function u C
2
(D) C
0
(

D) satisfying:
u +qu 0 in D and u > 0 on

D.
Then
1
( +q, D) > 0.
Proof. Set w = u + qu. Let h C

0
(D), since u > 0 we can write: h = u with
C
2
0
(D). Using Stokes formula, we can write:
_
D
[h[
2
dA =
_
D
hhdA =
_
D
u(u)dA
=
_
D
u(u +u + 2u, ))dA
=
_
D
u
2
+
2
u(w qu) + 2uu, )dA
So:
_
D
[h[
2
qh
2
dA =
_
D
u
2
+ 2uu, )dA
_
D

2
uwdA

_
D
u
2
+ 2uu, )dA.
Using again Stokes formula:
_
D
u
2
dA =
_
D
div(u
2
) (u
2
), ) dA
=
_
D
(u
2
), ) dA
=
_
D
2uu, ) +u
2
[[
2
dA.
Therefore:
_
D
[h[
2
qh
2
dA
_
D
u
2
[[
2
dA 0.
Moreover if the leftmost-hand side vanishes then is constant and so = 0 (since vanishes
on D) and thus also h = 0. We conclude then by the characterization (19).
Proof of Barbosa-do Carmos theorem. As we observed above, a domain on a plane is stable,
so we will assume our domain does not lie on a plane. We will prove the existence of function
u C
2
(D) C
0
(

D), which is positive on

D and satisfying: u 2Ku 0. The Lemma
R. SOUAM 93
4.8 will then imply that
1
(2K, D) > 0, where is the Laplacian for the metric ds
2
on
the surface. Taking into account the Proposition 4.8, this then means D is stable.
Set = N(D) S
2
. Since N is conformal and non constant, is an open set with
piecewise smooth boundary. By hypothesis area() < 2. We may consider an open domain

with smooth boundary such that

and area(

) < 2. By the Corollary 4.7 we have:

1
(
1
,

) > 2.
Let v a nontrivial eigenfunction associated to
1
(
1
,

). We may assume v > 0 inside

(cf.
Proposition 4.5). Then for the spherical Laplacian
1
, we have:

1
v + 2v <
1
v +
1
(
1
,

)v = 0 in and v > 0 on

.
On the set D
0
= D K = 0, the Gauss map is a local diffeomorphism, so we may
consider the pull-back metric
ds
2
2
= N

ds
2
1
.
For the Laplacian
2
associated to ds
2
2
, there holds the relation:

2
(v N) =
1
v.
It follows that the function u := v N :

D R satises on D :

2
u + 2u < 0.
Furthermore, since N is conformal, we have:
ds
2
2
= N

ds
2
1
= Kds
2
.
It is straightforward to check the relation between the Laplacians:

2
=

(K)
.
It follows (since K < 0) that u satises:
u 2Ku < 0 in D
0
.
It follows from (13) that on a nonplanar minimal surface the zeroes of the Gaussian curvature
are isolated. Therefore we have by continuity:
u 2Ku 0 in D.
Moreover, by construction, the function u is positive on

D.
Remark 4.1. As it should be clear, the proof shows more generally that if the rst eigenvalue
of the Gaussian image N(D) for the Laplacian with Dirichlet condition is bigger then 2 then
the domain D is stable. This is the case for instance if area(N(D)) = 2 and N(D) is not a
hemisphere.
94 G EOM ETRIES ET DYNAMIQUES
Corollary 4.9 Let D be a regular domain on a minimal surface in R
3
. Assume:

_
D
K dA

< 2.
Then D is stable.
Proof. The Gauss map of a nonplanar minimal surface is an orientation reversing local diffeo-
morphismexcept at isolated points which are the zeroes of the Gaussian curvature. Therefore:
area(N(D)) =
_
N(D)
dA
1

_
D
N

(dA
1
),
dA
1
being the area element on S
2
. The area element dA of the surface is related to dA
1
as
follows (cf. [15]): N

dA
1
= KdA. Therefore:
area(N(D)) =
_
N(D)
dA
1

_
D
N

(dA
1
) =
_
D
(K)dA =

_
D
K dA

and the result follows from Barbosa-do Carmos theorem.

It should be clear from the previous considerations that in case the Gauss map is injective,
which happens for instance on the catenoid and Ennepers surface, then stable domains are
precisely those having absolute total curvature less than 2.
A question which naturally comes to mind is to determine the complete minimal surfaces
(without boundary) which are globally stable. A complete minimal surface is said globally
stable if all its compact subdomains are stable. For instance a plane is globally stable. A
theorem proved independently by do Carmo-Peng, Pogorelov and by Fischer-Colbrie and
Schoen [6] states that the planes are the only globally stable complete (orientable) minimal
surfaces without boundary in R
3
.
References
[1] J. Barbosa and G. Colares.: Minimal surfaces in R
3
. Translated from the Portuguese.
Lecture Notes in Mathematics, 1195. Springer-Verlag, Berlin, (1986).
[2] J.L. Barbosa and M.P. do Carmo.:On the size of a stable minimal surface in R
3
. Amer.
J. Math. 98 (1976), no. 2, 515528.
[3] M. Berger; P. Gauduchon and E. Mazet.: Le spectre dune vari et e riemannienne. Lec-
ture Notes in Mathematics, Vol. 194 Springer-Verlag, Berlin-New York, (1971).
[4] M. P. do Carmo.: Differential Geometry of Curves and Surfaces. Prentice-Hall Inc.,
Englewood Cliffs,N.J., (1976).
[5] I. Chavel.: Eigenvalues in Riemannian geometry. Including a chapter by Burton Ran-
dol. With an appendix by Jozef Dodziuk. Pure and Applied Mathematics, 115. Aca-
demic Press, Inc., Orlando, FL, (1984).
R. SOUAM 95
[6] D. Fischer-Colbrie and R. Schoen. The structure of complete stable minimal surfaces
in 3-manifolds of nonnegative scalar curvature. Comm. Pure Appl. Math. 33 (1980),
no. 2, 199211.
[7] D. Gilbarg and [Link].: Elliptic partial differential equations of second order.
Grundlehren der Mathematischen Wissenschaften, Vol. 224. Springer-Verlag, Berlin-
New York, (1977).
[8] D. Hoffman and H. Karcher,.:Complete embedded minimal surfaces of nite total cur-
vature. Geometry, V, 593, 267272, Encyclopaedia Math. Sci., 90, Springer, Berlin,
(1997).
[9] H. B. Lawson.:Lectures on minimal submanifolds. Vol. I. Second edition. Mathematics
Lecture Series, 9. Publish or Perish, Inc., Wilmington, Del., (1980).
[10] F. Lopez and F. Martin.:Complete minimal surfaces in R
3
. Publ. Mat. 43 (1999), no.
2, 341449.
[11] F. Morgan.:Geometric measure theory. A beginners guide. Academic Press, Inc.,
Boston, MA, (1988).
[12] R. Osserman.: A survey of minimal surfaces. Second edition Dover Publications, Inc.,
New York, (1986).
[13] J. Perez and A. Ros.:Properly embedded minimal surfaces with nite total curvature.
The global theory of minimal surfaces in at spaces (Martina Franca, 1999), 1566,
Lecture Notes in Math., 1775, Springer, Berlin, (2002).
[14] H. Rosenberg.:Some recent developments in the theory of properly embedded minimal
surfaces in R
3
. Seminaire Bourbaki, Vol. 1991/92. Asterisque No. 206, (1992), Exp.
No. 759, 5, 463535.
[15] R. Souam.: Basic geometry of curves and surfaces. These proceedings.
Rabah Souam
Institut de Math ematiques de Jussieu - CNRS UMR 7586
Universit e Paris Diderot - Paris 7, G eom etrie et Dynamique
Case 7012 75205 - Paris Cedex 13
e-mail: souam@[Link]