Differential Geometry Note11

AN INTRODUCTION
TO
DIFFERENTIAL GEOMETRY
Philippe G. Ciarlet
City University of Hong Kong
Lecture Notes Series
Contents
Preface
iii
1 Three-dimensional differential geometry

1.1 Curvilinear coordinates . . . . . . . . . . . . . . . . . . . . . . .
1.2 Metric tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Volumes, areas, and lengths in curvilinear coordinates . . . . . .
1.4 Covariant derivatives of a vector field and Christoffel symbols . .
1.5 Necessary conditions satisfied by the metric tensor; the Riemann
curvature tensor . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.6 Existence of an immersion defined on an open set in R3 with a
prescribed metric tensor . . . . . . . . . . . . . . . . . . . . . . .
1.7 Uniqueness up to isometries of immersions with the same metric
tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.8 Continuity of an immersion as a function of its metric tensor . .
5
5
6
9
11
16
18
29
36
2 Differential geometry of surfaces

53
2.1 Curvilinear coordinates on a surface . . . . . . . . . . . . . . . . 53
2.2 First fundamental form . . . . . . . . . . . . . . . . . . . . . . . 57
2.3 Areas and lengths on a surface . . . . . . . . . . . . . . . . . . . 59
2.4 Second fundamental form; curvature on a surface . . . . . . . . . 60
2.5 Principal curvatures; Gaussian curvature . . . . . . . . . . . . . . 66
2.6 Covariant derivatives of a vector field and Christoffel symbols on
a surface; the Gau and Weingarten formulas . . . . . . . . . . . 71
2.7 Necessary conditions satisfied by the first and second fundamental forms: the Gau and Codazzi-Mainardi equations; Gau
theorema egregium . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.8 Existence of a surface with prescribed first and second fundamental forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.9 Uniqueness up to isometries of surfaces with the same fundamental forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
2.10 Continuity of a surface as a function of its fundamental forms . . 92
References
101
PREFACE
The notes presented here are based on lectures delivered over the years by
the author at the Universite Pierre et Marie Curie, Paris, at the University of
Stuttgart, and at City University of Hong Kong. Their aim is to give a thorough
introduction to the basic theorems of Differential Geometry.
In the first chapter, we review the basic notions arising when a threedimensional open set is equipped with curvilinear coordinates, such as the metric
tensor, Christoffel symbols, and covariant derivatives. We then prove that the
vanishing of the Riemann curvature tensor is sufficient for the existence of isometric immersions from a simply-connected open subset of Rn equipped with
a Riemannian metric into a Euclidean space of the same dimension. We also
prove the corresponding uniqueness theorem, also called rigidity theorem.
In the second chapter, we study basic notions about surfaces, such as their
two fundamental forms, the Gaussian curvature, Christoffel symbols, and covariant derivatives. We then prove the fundamental theorem of surface theory,
which asserts that the Gau and Codazzi-Mainardi equations constitute sufficient conditions for two matrix fields defined in a simply-connected open subset
of R3 to be the two fundamental forms of a surface in a three-dimensional Euclidean space. We also prove the corresponding rigidity theorem.
In addition to such classical theorems, we also include in both chapters
very recent results, which have not yet appeared in book form, such as the
continuity of a surface as a function of its fundamental forms.
The treatment is essentially self-contained and proofs are complete. The
prerequisites essentially consist in a working knowledge of basic notions of analysis and functional analysis, such as differential calculus, integration theory
and Sobolev spaces, and some familiarity with ordinary and partial differential
equations.
These notes use some excerpts from Chapters 1 and 2 of my book Mathematical Elasticity, Volume III: Theory of Shells, published in 2000 by NorthHolland, Amsterdam; in this respect, I am indebted to Arjen Sevenster for his
kind permission to reproduce these excerpts. Otherwise, the major part of these
notes was written during the fall of 2004 at City University of Hong Kong; this
part of the work was substantially supported by a grant from the Research
Grants Council of Hong Kong Special Administrative Region, China [Project
No. 9040869, CityU 100803].
Hong Kong, January 2005
iii
Chapter 1
THREE-DIMENSIONAL DIFFERENTIAL
GEOMETRY
1.1
CURVILINEAR COORDINATES
To begin with, we list some notations and conventions that will be consistently
used throughout.
All spaces, matrices, etc., considered here are real.
Latin indices and exponents vary in the set {1, 2, 3}, except when they are
used for indexing sequences, and the summation convention with respect to
repeated indices or exponents is systematically used in conjunction with this
rule. For instance, the relation
g i (x) = gij (x)g j (x)
means that
g i (x) =
3
X
gij (x)g j (x) for i = 1, 2, 3.
j=1
Kroneckers symbols are designated by ij , ij , or ij according to the context.

Let E3 denote a three-dimensional Euclidean space, let a b and a b denote
the
of a, b E3 , and let |a| =
Euclidean inner product and exterior product
3
a a denote the Euclidean norm of a E . The space E3 is endowed with
an orthonormal basis consisting of three vectors b
ei = b
ei . Let x
bi denote the
Cartesian coordinates of a point x
b E3 and let bi := /b
xi .
In addition, let there be given a three-dimensional vector space in which
three vectors ei = ei form a basis. This space will be identified with R3 . Let xi
denote the coordinates of a point x R3 and let i := /xi , ij := 2 /xi xj ,
and ijk := 3 /xi xj xk .
b of E3 and assume that there exist an
Let there be given an open subset
3
b
open subset of R and an injective mapping : E3 such that () = .
b can be unambiguously written as
Then each point x
b
x
b = (x), x ,
5
[Ch. 1
Three-dimensional differential geometry
and the three coordinates xi of x are called the curvilinear coordinates of x

b
(Figure 1.1-1). Naturally, there are infinitely many ways of defining curvilinear
b depending on how the open set and the
coordinates in a given open set ,
mapping are chosen!
g3 (x)
g2 (x)
x3
x
x
e
e2
x2
g1 (x)
E3
e1
x1
e2
e1
b E3 . The three
Figure 1.1-1: Curvilinear coordinates and covariant bases in an open set
b If the three
coordinates x1 , x2 , x3 of x are the curvilinear coordinates of x
b = (x) .
vectors g i (x) = i (x) are linearly independent, they form the covariant basis at x
b = (x)
and they are tangent to the coordinate lines passing through x
b.
Examples of curvilinear coordinates include the well-known cylindrical and

spherical coordinates (Figure 1.1-2).
In a different, but equally important, approach, an open subset of R3
together with a mapping : E3 are instead a priori given.
b := () is open by the inIf C 0 (; E3 ) and is injective, the set
variance of domain theorem (for a proof, see, e.g., Nirenberg [1974, Corollary 2,
b are
p. 17] or Zeidler [1986, Section 16.4]), and curvilinear coordinates inside
unambiguously defined in this case.
If C 1 (; E3 ) and the three vectors i (x) are linearly independent at all
b is again open (for a proof, see, e.g., Schwartz [1992] or Zeidler
x , the set
[1986, Section 16.4]), but curvilinear coordinates may be defined only locally in
this case: Given x , all that can be asserted (by the local inversion theorem)
is the existence of an open neighborhood V of x in such that the restriction
of to V is a C 1 -diffeomorphism, hence an injection, of V onto (V ).
1.2
METRIC TENSOR
Let be an open subset of R3 and let

= i b
ei : E 3
Sect. 1.2]
Metric tensor
E3
E3
Figure 1.1-2: Two familiar examples of curvilinear coordinates. Let the mapping be
defined by
: (, , z) ( cos , sin , z) E3 .
Then (, , z) are the cylindrical coordinates of x
b = (, , z). Note that ( + 2k, , z) or
( + + 2k, , z), k Z, are also cylindrical coordinates of the same point x
b and that is
not defined if x
b is the origin of E3 .
Let the mapping be defined by
: (, , r) (r cos cos , r cos sin , r sin ) E3 .
Then (, , r) are the spherical coordinates of x
b = (, , r). Note that ( + 2k, + 2`, r)
or ( + 2k, + + 2`, r) are also spherical coordinates of the same point x
b and that
and are not defined if x
b is the origin of E3 .
be a mapping that is differentiable at a point x . If x is such that (x+x)

, then
(x + x) = (x) + (x)x + o(x),
where the 3 3 matrix (x) is defined by
1 1 2 1
(x) := 1 2 2 2
1 3 2 3
3 1
3 2 (x).
3 3
Let the three vectors g i (x) R3 be defined by
i 1
g i (x) := i (x) = i 2 (x),
i 3
i.e., g i (x) is the i-th column vector of the matrix (x) and let x = xi ei .
Then the expansion of about x may be also written as
(x + x) = (x) + xi g i (x) + o(x).
If in particular x is of the form x = tei , where t R and ei is one of
the basis vectors in R3 , this relation reduces to
(x + tei ) = (x) + tg i (x) + o(t).
[Ch. 1
A mapping : E3 is an immersion at x if it is differentiable

at x and the matrix (x) is invertible or, equivalently, if the three vectors
g i (x) = i (x) are linearly independent.
Assume from now on in this section that the mapping is an immersion
at x. Then the three vectors g i (x) constitute the covariant basis at the point
x
b = (x).
In this case, the last relation thus shows that each vector g i (x) is tangent
to the i-th coordinate line passing through x
b = (x), defined as the image
b
by of the points of that lie on the line parallel to ei passing through x
(there exist t0 and t1 with t0 < 0 < t1 such that the i-th coordinate line is
given by t ]t0 , t1 [ f i (t) := (x + tei ) in a neighborhood of x
b; hence
f 0i (0) = i (x) = g i (x)); see Figures 1.1-1 and 1.1-2.
Returning to a general increment x = xi ei , we also infer from the expansion of about x that (recall that we use the summation convention):

|(x + x) (x)|2 = xT (x)T (x)x + o |x|2

= xi g i (x) g j (x)xj + o |x|2 .
In other words, the principal part with respect to x of the length between
the points (x + x) and (x) is {xi g i (x) g j (x)xj }1/2 . This observation
suggests to define a matrix (gij (x)) of order three, by letting
gij (x) := g i (x) g j (x) = ((x)T (x))ij .
The elements gij (x) of this symmetric matrix are called the covariant components of the metric tensor at x
b = (x).
Note that the matrix (x) is invertible and that the matrix (gij (x)) is
positive definite, since the vectors g i (x) are assumed to be linearly independent.
The three vectors g i (x) being linearly independent, the nine relations
g i (x) g j (x) = ji
unambiguously define three linearly independent vectors g i (x). To see this, let
a priori g i (x) = X ik (x)g k (x) in the relations g i (x) g j (x) = ji . This gives
X ik (x)gkj (x) = ji ; consequently, X ik (x) = g ik (x), where
(g ij (x)) := (gij (x))1 .
Hence g i (x) = g ik (x)g k (x). These relations in turn imply that

g i (x) g j (x) = g ik (x)g k (x) g j` (x)g ` (x)
= g ik (x)g j` (x)gk` (x) = g ik (x)kj = g ij (x),
and thus the vectors g i (x) are linearly independent since the matrix (g ij (x)) is
positive definite. We would likewise establish that g i (x) = gij (x)g j (x).
The three vectors g i (x) form the contravariant basis at the point x
b = (x)
and the elements g ij (x) of the symmetric positive definite matrix (g ij (x)) are
the contravariant components of the metric tensor at x
b = (x).
Sect. 1.3]
Volumes, areas, and lengths in curvilinear coordinates
To conclude this section, we record for convenience the fundamental relations

that exist between the vectors of the covariant and contravariant bases and the
covariant and contravariant components of the metric tensor:
gij (x) = g i (x) g j (x)
g i (x) = gij (x)g j (x)
1.3
and g ij (x) = g i (x) g j (x),

and g i (x) = g ij (x)g j (x).
VOLUMES, AREAS, AND LENGTHS IN CURVILINEAR COORDINATES
We now review fundamental formulas showing how volume, area, and length
b = () can be expressed either in
elements at a point x
b = (x) in the set
terms of the matrix (x) or in terms of the matrix (gij (x)) or of its inverse
matrix (g ij (x)).
These formulas thus highlight the crucial r
ole played by the matrix (gij (x))
for computing metric notions at the point x
b = (x). Indeed, the metric
tensor well deserves its name!
A domain in Rn is a bounded, open, and connected subset D of R3 with
a Lipschitz-continuous boundary, the set D being locally on one side of its
boundary. All relevant details needed here about domains are found in Necas
[1967] or Adams [1975].
Given a domain D R3 with boundary , we let dx denote the volume
element in D, d denote the area element along , and n = ni b
ei denote the
unit (|n| = 1) outer normal vector along (d is well defined and n is defined
d-almost everywhere since is assumed to be Lipschitz-continuous).
Note also that the assumptions made on the mapping in the next theorem
b ,
b
guarantee that, if D is a domain in R3 such that D , then {D}
b of D
b and D of D are related by
{(D)} = (D), and the boundaries D
b = (D) (see, e.g., Ciarlet [1988, Theorem 1.2-8 and Example 1.7]).
D
If A is a square matrix, Cof A denotes the cofactor matrix of A. Thus
Cof A = (det A)AT if A is invertible.
A mapping : E3 is an immersion if it is an immersion at each
x , i.e., if is differentiable in and the three vectors g i (x) = i (x) are
linearly independent at each x .
Theorem 1.3-1. Let be an open subset of R3 , let : E3 be an injective
b = ().
and smooth enough immersion, and let
b is given in terms of the volume
(a) The volume element db
x at x
b = (x)
element dx at x by
p
db
x = | det (x)| dx = g(x) dx, where g(x) := det(gij (x)).
b x) at
(b) Let D be a domain in R3 such that D . The area element d(b
b
x
b = (x) D is given in terms of the area element d(x) at x D by
q
p
b x) = | Cof (x)n(x)| d(x) = g(x) ni (x)g ij (x)nj (x) d(x),
d(b
10
[Ch. 1
where n(x) := ni (x)ei denotes the unit outer normal vector at x D.

b x) at x
b is given by
(c) The length element d`(b
b = (x)

b x) = xT (x)T (x)x 1/2 = xi gij (x)xj 1/2 ,
d`(b
where x = xi ei .
Proof. The relation db

x = | det (x)| dx between the volume elements
is well known. The second relation in (a) follows from the relation g(x) =
| det (x)|2 , which itself follows from the relation (gij (x)) = (x)T (x).
b x)
Indications about the proof of the relation between the area elements d(b
and d(x) given in (b) are found in Ciarlet [1988, Theorem 1.7-1] (in this formula, n(x) = ni (x)ei is identified with the column vector in R3 with ni (x) as
its components). Using the relations Cof (AT ) = (Cof A)T and Cof(AB) =
(Cof A)(Cof B), we next have:

| Cof (x)n(x)|2 = n(x)T Cof (x)T (x) n(x)
= g(x)ni (x)g ij (x)nj (x).
b x) is
Either expression of the length element given in (c) recalls that d`(b
by definition the principal part with respect to x = xi ei of the length
|(x + x) (x)|, whose expression precisely led to the introduction of the
matrix (gij (x)) in Section 1.2.

The relations found in Theorem 1.3-1 are used in particular for computing
b by means of integrals inside , i.e., in terms
volumes, areas, and lengths inside
b (Figure 1.3-1):
of the curvilinear coordinates used in the open set
3
b
b
Let D be a domain in R such that D , let D := (D), and let fb L1 (D)
be given. Then
Z
Z
p
b
f (b
x) db
x=
(fb )(x) g(x) dx.
b
D
b is given by
In particular, the volume of D
Z
Z p
b
g(x) dx.
vol D :=
db
x=
b
D
b := ()
Next, let := D, let be a d-measurable subset of , let
1
b and let b
b be given. Then
D,
h L ()
Z
Z
q
p
b
b x) = (b
h(b
x) d(b
h )(x) g(x) ni (x)g ij (x)nj (x) d(x).
b
b is given by
In particular, the area of
Z
Z p
q
b :=
b x) =
area
d(b
g(x) ni (x)g ij (x)nj (x) d(x).
b
Sect. 1.4]
11
Covariant derivatives and Christoffel symbols

n(x)
A
x)
d(
d(x)
V
dx
x+x
x
dl(
x)
d
x
(x+x)
R3
(x) = x
E3
f
t
Figure 1.3-1: Volume, area, and length elements in curvilinear coordinates. The elements
b x) at x
b x), and d`(b
b are expressed in terms of dx, d(x), and x at x by
db
x, d(b
b = (x)
means of the covariant and contravariant components of the metric tensor; cf. Theorem 1.3-1.
Given a domain D such that D and a d-measurable subset of D, the corresponding
b = (D) ,
b the area of
b = () D,
b
relations are used for computing the volume of D
b = (C) ,
b where C = f (I) and I is a compact interval of R.
and the length of a curve C
Finally, consider a curve C = f (I) in , where I is a compact interval of R

and f = f i ei : I is a smooth enough injective mapping. Then the length
b := (C)
b is given by
of the curve C
s
Z
Z

d
df i df j
b := ( f )(t) dt =
(t)
(t) dt.
gij (f (t))
length C
dt
dt
I
I dt
This relation shows in particular that the lengths of curves inside the set
() are precisely those induced by the Euclidean metric of the space E 3 .
1.4
COVARIANT DERIVATIVES OF A VECTOR FIELD

AND CHRISTOFFEL SYMBOLS
b of E3 by means of its
Suppose that a vector field is defined in an open subset
b
Cartesian components vbi : R, i.e., this field is defined by its values vbi (b
x)b
ei
b where the vectors b
at each x
b ,
ei constitute the orthonormal basis of E3 ; see
Figure 1.4-1.
b is equipped with curvilinear coordinates
Suppose now that the open set
3
from an open subset of R , by means of an injective mapping : E3
b
satisfying () = .
12
[Ch. 1
v3 (
x)
vi (
x) ei
x3
v2 (
x)
x
v1 (
x)
e3
E3
e2
x2
e1
x1
b the vector
Figure 1.4-1: A vector field in Cartesian coordinates. At each point x
b ,
vbi (b
x)b
ei is defined by its Cartesian components vbi (b
x) over an orthonormal basis of E3 formed
by three vectors b
ei .
How to define appropriate components of the same vector field, but this time
in terms of these curvilinear coordinates? It turns out that the proper way to
do so consists in defining three functions vi : R by requiring that (Figure
1.4-2)
vi (x)g i (x) := vbi (b
x)b
ei for all x
b = (x), x ,
where the three vectors g i (x) form the contravariant basis at x

b = (x) (Section
1.2). Using the relations g i (x) g j (x) = ji and b
ei b
ej = ji , we immediately find
how the old and new components are related, viz.,
vj (x) = vi (x)g i (x) g j (x) = vbi (b
x)b
ei g j (x),
vbi (b
x) = vbj (b
x)b
ej b
ei = vj (x)g j (x) b
ei .
Sect. 1.4]
13
g3 (x)
ui (x)g i (x)
g2 (x)
u3 (x)
u2(x)
x3
x
u1(x)
x
e3
e
e
e3
E3
x2
g1 (x)
x1
Figure 1.4-2: A vector field in curvilinear coordinates. Let there be given a vector field
b by its Cartesian components vbi (b
in Cartesian coordinates defined at each x
b
x) over the
vectors b
ei (Figure 1.4-1). In curvilinear coordinates, the same vector field is defined at each
x by its covariant components vi (x) over the contravariant basis vectors g i (x) in such a
way that vi (x)g i (x) = vbi (b
x)ei , x
b = (x).
The three components vi (x) are called the covariant components of the
vector vi (x)g i (x) at x
b, and the three functions vi : R defined in this fashion
are called the covariant components of the vector field vi g i : E3 .
Suppose next that we wish to compute a partial derivative bj vbi (b
x) at a point
b
x
b = (x) in terms of the partial derivatives ` vk (x) and of the values vq (x)
(which are also expected to appear by virtue of the chain rule). Such a task is
required for example if we wish to write a system of partial differential equations
whose unknown is a vector field (such as the equations of nonlinear or linearized
elasticity) in terms of ad hoc curvilinear coordinates.
As we now show, carrying out such a transformation naturally leads to a
fundamental notion, that of covariant derivatives of a vector field.
Theorem 1.4-1. Let be an open subset of R3 and let : E3 be an
b := (). Given a
immersion that is also a C 2 -diffeomorphism of onto
i b
3
b
b
vector field vbi e : R in Cartesian coordinates with components vbi C 1 (),
let vi g i : R3 be the same field in curvilinear coordinates, i.e., that defined
by
vbi (b
x)b
ei = vi (x)g i (x) for all x
b = (x), x .
Then vi C 1 () and for all x ,
where
and

bj vbi (b
x) = vkk` [g k ]i [g ` ]j (x), x
b = (x),
vikj := j vi pij vp and pij := g p i g j ,

[g i (x)]k := g i (x) b
ek
denotes the i-th component of g i (x) over the basis {b

e1 , b
e2 , b
e3 }.
14
[Ch. 1
Proof. The following convention holds throughout this proof: The simultaneous appearance of x
b and x in an equality means that they are related by
x
b = (x) and that the equality in question holds for all x .
(i) Another expression of [g i (x)]k := g i (x) b
ek .
b x) =
b i (b
b :
b E3 denotes the
Let (x) = k (x)b
ek and (b
x)ei , where
3
b
inverse mapping of : E . Since ((x))
= x for all x , the chain
rule shows that the matrices (x) := (j k (x)) (the row index is k) and
b (b
b x) := (bk
b i (b
x)) (the row index is i) satisfy

b (b
b x)(x) = I,
or equivalently,
j 1 (x)
b i (b
b i (b
b i (b
b i (b
bk
x)j k (x) = b1
x ) 2
x ) 3
x) j 2 (x) = ji .
j 3 (x)

The components of the above column vector being precisely those of the
vector g j (x), the components of the above row vector must be those of the
vector g i (x) since g i (x) is uniquely defined for each exponent i by the three
relations g i (x) g j (x) = ji , j = 1, 2, 3. Hence the k-th component of g i (x) over
b
the basis {b
e1 , b
e2 , b
e3 } can be also expressed in terms of the inverse mapping ,
as:
b i (b
[g i (x)]k = bk
x).
(ii) The functions q`k := g q ` g k C 0 ().
We next compute the derivatives ` g q (x) (the fields g q = g qr g r are of class

C 1 on since is assumed to be of class C 2 ). These derivatives will be needed
in (iii) for expressing the derivatives bj u
bi (b
x) as functions of x (recall that u
bi (b
x) =
uk (x)[g k (x)]i ). Recalling that the vectors g k (x) form a basis, we may write a
priori
` g q (x) = q`k (x)g k (x),
thereby unambiguously defining functions q`k : R. To find their expresb we observe that
sions in terms of the mappings and ,
q`k (x) = q`m (x)km = q`m (x)g m (x) g k (x) = ` g q (x) g k (x).
b q (b
Hence, noting that ` (g q (x) g k (x)) = 0 and [g q (x)]p = bp
x), we obtain
b q (b
q`k (x) = g q (x) ` g k (x) = bp
x)`k p (x) = qk` (x).
b C 1 (;
b R3 ) by assumption, the last relations
Since C 2 (; E3 ) and
q
0
show that `k C ().
Sect. 1.4]
15
(iii) The partial derivatives bi vbi (b

x) of the Cartesian components of the vector
i
1
3
b
b by
field vbi b
e C (; R ) are given at each x
b = (x)
bj vbi (b
x) = vkk` (x)[g k (x)]i [g ` (x)]j ,
where
vkk` (x) := ` vk (x) q`k (x)vq (x),
and [g k (x)]i and q`k (x) are defined as in (i) and (ii).
We compute the partial derivatives bj vbi (b

x) as functions of x by means of the
relation vbi (b
x) = vk (x)[g k (x)]i . To this end, we first note that a differentiable
function w : R satisfies
b x)) = ` w(x)bj
b ` (b
bj w((b
x) = ` w(x)[g ` (x)]j ,
by the chain rule and by (i). In particular then,
b x))[g k (x)]i + vq (x)bj [g q ((b

b x))]i
bj vbi (b
x) = bj vk ((b

= ` vk (x)[g ` (x)]j [g k (x)]i + vq (x) ` [g q (x)]i [g ` (x)]j
= (` vk (x) q`k (x)vq (x)) [g k (x)]i [g ` (x)]j ,
since ` g q (x) = q`k (x)g k (x) by (ii).

The functions
vikj = j vi pij vp
defined in Theorem 1.4-1 are called the first-order covariant derivatives of

the vector field vi g i : R3 .
The functions
pij = g p i g j
are called the Christoffel symbols of the first kind.

The following result summarizes properties of covariant derivatives and Christoffel symbols that are constantly used.
Theorem 1.4-2. Let the assumptions on the mapping : E3 be as in
Theorem 1.4-1, and let there be given a vector field vi g i : R3 with covariant
components vi C 1 ().
(a) The first-order covariant derivatives vikj C 0 () of the vector field
i
vi g : R3 , which are defined by
vikj := j vi pij vp , where pij := g p i g j ,
can be also defined by the relations

j (vi g i ) = vikj g i vikj = j (vk g k ) g i .
(b) The Christoffel symbols pij := g p i g j = pji C 0 () satisfy the relations

i g p = pij g j and j g q = ijq g i .
16
[Ch. 1
Proof. It remains to verify that the covariant derivatives vikj , defined in

Theorem 1.4-1 by
vikj = j vi pij vp ,
may be equivalently defined by the relations
j (vi g i ) = vikj g i .
These relations unambiguously define the functions vikj = {j (vk g k )} g i since
the vectors g i are linearly independent at all points of by assumption. To
this end, we simply note that, by definition, the Christoffel symbols satisfy
i g p = pij g j (cf. part (ii) of the proof of Theorem 1.4-1); hence
j (vi g i ) = (j vi )g i + vi j g i = (j vi )g i vi ijk g k = vikj g i .
To establish the other relations j g q = ijq g i , we note that
0 = j (g p g q ) = pji g i g q + g p j g q = pqj + g p j g q .
Hence
j g q = (j g q g p )g p = pqj g p .
Remark. The Christoffel symbols pij can be also defined solely in terms of
the components of the metric tensor; see the proof of Theorem 1.5-1.

If the affine space E3 is identified with R3 and = id , the relation
j (vi g i )(x) = (vikj g i )(x) (Theorem 1.4-2 (a)), reduces to bj (b
vi (b
x)b
ei ) = (bj vbi (b
x))b
ei .
In this sense, a covariant derivative of the first order constitutes a generalization
of a partial derivative of the first order in Cartesian coordinates.
1.5
NECESSARY CONDITIONS SATISFIED BY THE

METRIC TENSOR; THE RIEMANN CURVATURE
TENSOR
It is remarkable that the components gij : R of the metric tensor of an

open set () E3 (Section 1.2), defined by a smooth enough immersion
: E3 , cannot be arbitrary functions.
As shown in the next theorem, they must satisfy relations that take the
form:
j ikq k ijq + pij kqp pik jqp = 0 in ,
where the functions ijq and pij have simple expressions in terms of the functions gij and of some of their partial derivatives (as shown in the next proof,
it so happens that the functions pij as defined in Theorem 1.5-1 coincide with
the Christoffel symbols introduced in the previous section; this explains why
Sect. 1.5]
Necessary conditions satisfied by the metric tensor
17
they are denoted by the same symbol). Note that, according to the rule governing Latin indices and exponents, these relations are meant to hold for all
i, j, k, q {1, 2, 3}.
Theorem 1.5-1. Let be an open set in R3 , let C 3 (; E3 ) be an immersion, and let
gij := i j
denote the covariant components of the metric tensor of the set (). Let the
functions ijq C 1 () and pij C 1 () be defined by
1
(j giq + i gjq q gij ),
2
pij := g pq ijq where (g pq ) := (gij )1 .
ijq :=
Then, necessarily,
j ikq k ijq + pij kqp pik jqp = 0 in .
Proof. Let g i = i . It is then immediately verified that the functions ijq
are also given by
ijq = i g j g q .
For each x , let the three vectors g j (x) be defined by the relations g j (x)
g i (x) = jj . Since we also have g j = g ij g i , the last relations imply that pij =
i g j g p . Therefore,
i g j = pij g p
since i g j = (i g j g p )g p . Differentiating the same relations yields
k ijq = ik g j g q + i g j k g q ,
so that the above relations together give
i g j k g q = pij g p k g q = pij kqp .
Consequently,
ik g j g q = k ijq pij kqp .
Since ik g j = ij g k , we also have

ik g j g q = j ikq pik jqp ,
and thus the required necessary conditions immediately follow.
Remark. The vectors g i and g j introduced above form the covariant and
contravariant bases and the functions g ij are the contravariant components of
the metric tensor (Section 1.2).
18
[Ch. 1
As shown in the above proof, the necessary conditions Rqijk = 0 thus simply constitute a re-writing of the relations ik g j = ki g j in the form of the
equivalent relations ik g j g q = ki g j g q .
The functions
ijq =
and
1
(j giq + i gjq q gij ) = i g j g q = jiq
2
pij = g pq ijq = i g j g p = pji
are the Christoffel symbols of the first, and second, kinds. We saw in
Section 1.4 that the same Christoffel symbols pij also naturally appear in a
different context (that of covariant differentiation).
Finally, the functions
Rqijk := j ikq k ijq + pij kqp pik jqp
are the covariant components of the Riemann curvature tensor of the
set (). The relations Rqijk = 0 found in Theorem 1.4-1 thus express that
the Riemann curvature tensor of the set () (equipped with the metric tensor
with covariant components gij ) vanishes.
1.6
EXISTENCE OF AN IMMERSION DEFINED ON

AN OPEN SET IN R3 WITH A PRESCRIBED METRIC TENSOR
Let M3 , S3 , and S3> denote the sets of all square matrices of order three, of
all symmetric matrices of order three, and of all symmetric positive definite
matrices of order three.
As in Section 1.2, the matrix representing the Frechet derivative at x of
a differentiable mapping = (` ) : E3 is denoted
(x) := (j ` (x)) M3 ,
where ` is the row index and j the column index (equivalently, (x) is the
matrix of order three whose j-th column vector is j ).
So far, we have considered that we are given an open set R3 and a
smooth enough immersion : E3 , thus allowing us to define a matrix field
C = (gij ) = T : S3> ,
where gij : R are the covariant components of the metric tensor of the
open set () E3 .
We now turn to the reciprocal questions:
Given an open subset of R3 and a smooth enough matrix field C = (gij ) :
S3> , when is C the metric tensor field of an open set () E3 ? Equivalently, when does there exist an immersion : E3 such that
C = T in ,
Sect. 1.6] Existence of an immersion with a prescribed metric tensor
19
or equivalently, such that

gij = i j in ?
If such an immersion exists, to what extent is it unique?
The answers to these questions turn out to be remarkably simple: If is
simply-connected, the necessary conditions
j ikq k ijq + pij kqp pik jqp = 0 in
found in Theorem 1.4-1 are also sufficient for the existence of such an immersion. If is connected, this immersion is unique up to isometries in E3 .
Whether the immersion found in this fashion is globally injective is a different
issue, which accordingly should be resolved by different means.
This result comprises two essentially distinct parts, a global existence result
(Theorem 1.6-1) and a uniqueness result (Theorem 1.4-1). Note that these two
results are established under different assumptions on the set and on the
smoothness of the field (gij ).
In order to put these results in a wider perspective, let us make a brief
incursion into Riemannian Geometry. For detailed treatments, see classic texts
such as Choquet-Bruhat, de Witt-Morette & Dillard-Bleick [1977], Marsden &
Hughes [1983], or Gallot, Hulin & Lafontaine [2004].
Considered as a three-dimensional manifold, an open set R3 equipped
with an immersion : E3 becomes an example of a Riemannian manifold
(; (gij )), i.e., a manifold, the set , equipped with a Riemannian metric, the
symmetric positive-definite matrix field (gij ) : S3> defined in this case by
gij := i j in . More generally, a Riemannian metric on a manifold
is a twice covariant, symmetric, positive-definite tensor field acting on vectors
in the tangent spaces to the manifold (these tangent spaces coincide with R3 in
the present instance).
This particular Riemannian manifold (; (gij )) possesses the remarkable
property that its metric is the same as that of the surrounding space E3 . More
specifically, (; (gij )) is isometrically immersed in the Euclidean space E3 ,
in the sense that there exists an immersion : E3 that satisfies the relations gij = i j . Equivalently, the length of any curve in the Riemannian
manifold (; (gij )) is the same as the length of its image by in the Euclidean
space E3 (see Theorem 1.3-1).
The first question above can thus be rephrased as follows: Given an open
subset of R3 and a positive-definite matrix field (gij ) : S3> , when is
the Riemannian manifold (; (gij )) flat, in the sense that it can be locally
isometrically immersed in a Euclidean space of the same dimension (three)?
The answer to this question can then be rephrased as follows (compare with
the statement of Theorem 1.6-1 below): Let be a simply-connected open subset
of R3 . Then a Riemannian manifold (; (gij )) with a Riemannian metric (gij )
of class C 2 in is flat if and only if its Riemannian curvature tensor vanishes
in . Recast as such, this result becomes a special case of the fundamental
theorem on flat Riemannian manifolds, which holds for a general finitedimensional Riemannian manifold.
20
[Ch. 1
The answer to the second question, viz., the issue of uniqueness, can be
rephrased as follows (compare with the statement of Theorem 1.7-1 in the next
section): Let be a connected open subset of R3 . Then the isometric immersions
of a flat Riemannian manifold (; (gij )) into a Euclidean space E3 are unique
up to isometries of E3 . Recast as such, this result likewise becomes a special
case of the so-called rigidity theorem; cf. Section 1.7.
Recast as such, these two theorems together constitute a special case (that
where the dimensions of the manifold and of the Euclidean space are both equal
to three) of the fundamental theorem of Riemannian Geometry. This
theorem addresses the same existence and uniqueness questions in the more
general setting where is replaced by a p-dimensional manifold and E3 is replaced by a (p + q)-dimensional Euclidean space (the fundamental theorem of
surface theory, established in Sections 2.8 and 2.9, constitutes another important special case). When the p-dimensional manifold is an open subset of Rp ,
an outline of a self-contained proof is given in Szopos [2005].
Another fascinating question (which will not be addressed here) is the following: Given again an open subset of R3 equipped with a symmetric, positivedefinite matrix field (gij ) : S3 , assume this time that the Riemannian
manifold (; (gij )) is no longer flat, i.e., its Riemannian curvature tensor no
longer vanishes in . Can such a Riemannian manifold still be isometrically
immersed, but this time in a higher-dimensional Euclidean space? Equivalently,
do there exist a Euclidean space Ed with d > 3 and an immersion : Ed
such that gij = i j in ?
The answer is yes, according to the following beautiful Nash theorem, so
named after Nash [1954]: Any p-dimensional Riemannian manifold equipped
with a continuous metric can be isometrically immersed in a Euclidean space
of dimension 2p with an immersion of class C 1 ; it can also be isometrically
immersed in a Euclidean space of dimension (2p + 1) with a globally injective
immersion of class C 1 .
Let us now humbly return to the question of existence raised at the beginning
of this section, i.e., when the manifold is an open set in R3 .
Theorem 1.6-1. Let be a connected and simply-connected open set in R 3
and let C = (gij ) C 2 (; S3> ) be a matrix field that satisfies
Rqijk := j ikq k ijq + pij kqp pik jqp = 0 in ,
where
1
2
pij := g pq ijq with (g pq ) := (gij )1 .
ijq :=
Then there exists an immersion C 3 (; E3 ) such that

C = T in .
21
Proof. The proof relies on a simple, yet crucial, observation. When a smooth
enough immersion = (` ) : E3 is a priori given (as it was so far), its
components ` satisfy the relations ij ` = pij p ` , which are nothing but
another way of writing the relations i g j = pij g p (see the proof of Theorem
1.5-1). This observation thus suggests to begin by solving (see part (ii)) the
system of partial differential equations
i F`j = pij F`p in ,
whose solutions F`j : R then constitute natural candidates for the partial
derivatives j ` of the unknown immersion = (` ) : E3 (see part (iii)).
To begin with, we establish in (i) relations that will in turn allow us to
re-write the sufficient conditions
j ikq k ijq + pij kqp pik jqp = 0 in
in a slightly different form, more appropriate for the existence result of part (ii).
Note that the positive definiteness of the symmetric matrices (gij ) is not needed
for this purpose.
(i) Let be an open subset of R3 and let there be given a field (gij )
C (; S3 ) of symmetric invertible matrices. The functions ijq , pij , and g pq
being defined by
2
ijq :=
1
2
pij := g pq ijq ,
(g pq ) := (gij )1 ,
define the functions

Rqijk := j ikq k ijq + pij kqp pik jqp ,
p
Rijk
:= j pik k pij + ìk pj` ìj pk` .
Then
p
p
Rijk
= g pq Rqijk and Rqijk = gpq Rijk
.
Using the relations

jq` + `jq = j gq` and ikq = gq` ìk ,
which themselves follow from the definitions of the functions ijq and pij , and
noting that
(g pq j gq` + gq` j g pq ) = j (g pq gq` ) = 0,
we obtain
g pq (j ikq rik jqr ) = j pik ikq j g pq ìk g pq (j gq` `jq )
= j pik + ìk pj` ìk (g pq j gq` + gq` j g pq )
= j pik + ìk pj` .
22

Likewise,
[Ch. 1
g pq (k ijq rij kqr ) = k pij ìj pk` ,
p
and thus the relations Rijk
= g pq Rqijk are established. The relations Rqijk =
p
gpq Rijk are clearly equivalent to these ones.
We next establish the existence of solutions to the system
i F`j = pij F`p in .

(ii) Let be a connected and simply-connected open subset of R3 and let
there be given functions pij = pji C 1 () satisfying the relations
j pik k pij + ìk pj` ìj pk` = 0 in ,
which are equivalent to the relations
by part (i).
0
Let a point x0 and a matrix (F`j
) M3 be given. Then there exists one,
and only one, field (F`j ) C 2 (; M3 ) that satisfies
i F`j (x) = pij (x)F`p (x), x ,
0
F`j (x0 ) = F`j
.
Let x1 be an arbitrary point in the set , distinct from x0 . Since is

connected, there exists a path = ( i ) C 1 ([0, 1]; R3 ) joining x0 to x1 in ;
this means that
(0) = x0 , (1) = x1 , and (t) for all 0 t 1.
Assume that a matrix field (F`j ) C 1 (; M3 ) satisfies i F`j (x) = pij (x)F`p (x),
x . Then, for each integer ` {1, 2, 3}, the three functions j C 1 ([0, 1])
defined by (for simplicity, the dependence on ` is dropped)
j (t) := F`j ((t)), 0 t 1,
satisfy the following Cauchy problem for a linear system of three ordinary differential equations with respect to three unknowns:
dj
d i
(t) = pij ((t))
(t)p (t), 0 t 1,
dt
dt
j (0) = j0 ,
where the initial values j0 are given by
0
j0 := F`j
.
23
Note in passing that the three Cauchy problems obtained by letting ` = 1, 2,

or 3 only differ by their initial values j0 .
It is well known that a Cauchy problem of the form (with self-explanatory
notations)
d
(t) = A(t)(t), 0 t 1,
dt
(0) = 0 ,
has one and only one solution C 1 ([0, 1]; R3 ) if A C 0 ([0, 1]; M3 ) (see, e.g.,
Schwartz [1992, Theorem 4.3.1, p. 388]). Hence each one of the three Cauchy
problems has one and only one solution.
Incidentally, this result already shows that, if it exists, the unknown field
(F`j ) is unique.
In order that the three values j (1) found by solving the above Cauchy
problem for a given integer ` {1, 2, 3} be acceptable candidates for the three
unknown values F`j (x1 ), they must be of course independent of the path chosen
for joining x0 to x1 .
So, let 0 C 1 ([0, 1]; R3 ) and 1 C 1 ([0, 1]; R3 ) be two paths joining x0
to x1 in . The open set being simply-connected, there exists a homotopy
G = (Gi ) : [0, 1] [0, 1] R3 joining 0 to 1 in , i.e., such that
G(, 0) = 0 , G(, 1) = 1 , G(t, ) for all 0 t 1, 0 1,
G(0, ) = x0 and G(1, ) = x1 for all 0 1,
and smooth enough in the sense that
G C 1 ([0, 1] [0, 1]; R3 ) and
G
G
=
C 0 ([0, 1] [0, 1]; R3 ).
t
t
Let (, ) = (j (, )) C 1 ([0, 1]; R3 ) denote for each 0 1 the solution

of the Cauchy problem corresponding to the path G(, ) joining x0 to x1 . We
thus have
j
Gi
(t, ) = pij (G(t, ))
(t, )p (t, ) for all 0 t 1, 0 1,
t
t
j (0, ) = j0 for all 0 1.
Our objective is to show that
j
(1, ) = 0 for all 0 1,
as this relation will imply that j (1, 0) = j (1, 1), as desired. For this purpose,
a direct differentiation shows that, for all 0 t 1, 0 1,
j
Gk Gi
Gi
Gi
= {qij pkq + k pij }p
+ pij p
+ q qij
,
t
t
t
t
24
[Ch. 1
where
j
Gk
pkj p
,
on the one hand (in the relations above and below, qij , k pij , etc., stand for
qij (G(, )), k pij (G(, )), etc.).
On the other hand, a direct differentiation of the equation defining the functions j shows that, for all 0 t 1, 0 1,
j :=
q o Gk
j j n p Gi
Gi
=
+ i kj
p + qkj
+ pij p
.
t
t
t
t
j
Gi
= pij
p , so that we also have
t
t
j j
Gi Gk
Gi
=
+ {i pkj + qkj piq }p
+ pij p
.
t
t
t
t
j
j
Hence, subtracting the above relations and noting that
=
t
t
Gi
Gi
and
=
by assumption, we infer that
t
t
But
j
Gk Gi
Gi
+ {i pkj k pij + qkj piq qij pkq }p
qij
q = 0.
t
t
t
The assumed symmetries pij = pji combined with the assumed relations
j pik k pij + ìk pj` ìj pk` = 0 in show that
i pkj k pij + qkj piq qij pkq = 0,
on the one hand. On the other hand,
j (0, ) =
Gk
j
(0, ) pkj (G(0, ))p (0, )
(0, ) = 0,
since j0 (0, ) = j0 and G(0, ) = x0 for all 0 1. Therefore, for any

fixed value of the parameter [0, 1], each function j (, ) satisfies a Cauchy
problem for an ordinary differential equation, viz.,
dj
Gi
(t, ) = qij (G(t, ))
(t, )q (t, ), 0 t 1,
dt
t
j (0, ) = 0.
But the solution of such a Cauchy problem is unique; hence j (t, ) = 0 for
all 0 t 1. In particular then,
j
Gk
(1, ) pkj (G(1, ))p (1, )
(1, )
= 0 for all 0 1,
j (1, ) =
and thus
25
j
(1, ) = 0 for all 0 1, since G(1, ) = x1 for all 0 1.
For each integer `, we may thus unambiguously define a vector field (F`j ) :
R3 by letting
F`j (x1 ) := j (1) for any x1 ,
where C 1 ([0, 1]; R3 ) is any path joining x0 to x1 in and the vector field
(j ) C 1 ([0, 1]) is the solution to the Cauchy problem
d i
dj
(t) = pij ((t))
(t)p (t), 0 t 1,
dt
dt
j (0) = j0 ,
corresponding to such a path.

To establish that such a vector field is indeed the `-th row-vector field of the
unknown matrix field we are seeking, we need to show that (F`j )3j=1 C 1 (; R3 )
and that this field does satisfy the partial differential equations i F`j = pij F`p
in corresponding to the fixed integer ` used in the above Cauchy problem.
Let x be an arbitrary point in and let the integer i {1, 2, 3} be fixed in
what follows. Then there exist x1 , a path C 1 ([0, 1]; R3 ) joining x0 to
x1 , ]0, 1[, and an open interval I [0, 1] containing such that
(t) = x + (t )ei for t I,
where ei is the i-th basis vector in R3 . Since each function j is continuously
dj
d i
differentiable in [0, 1] and satisfies
(t) = pij ((t))
(t)p (t) for all 0 t
dt
dt
1, we have
dj
( ) + o(t )
dt
p
= j ( ) + (t )ij (( ))p ( ) + o(t )
j (t) = j ( ) + (t )
for all t I. Equivalently,
F`j (x + (t )ei ) = F`j (x) + (t )pij (x)F`p (x) + o(t x).

This relation shows that each function F`j possesses partial derivatives in
the set , given at each x by
i F`p (x) = pij (x)F`p (x).
Consequently, the matrix field (F`j ) is of class C 1 in (its partial derivatives are
continuous in ) and it satisfies the partial differential equations i F`j = pij F`p
in , as desired. Differentiating these equations shows that the matrix field
(F`j ) is in fact of class C 2 in .
In order to conclude the proof of the theorem, it remains to adequately
0
choose the initial values F`j
at x0 in step (ii).
26
[Ch. 1
(iii) Let be a connected and simply-connected open subset of R3 and let

(gij ) C 2 (; S3> ) be a matrix field satisfying
the functions ijq , pij , and g pq being defined by
ijq :=
1
2
pij := g pq ijq ,
(g pq ) := (gij )1 .
0
Given an arbitrary point x0 , let (F`j
) S3> denote the square root of
0
the matrix (gij
) := (gij (x0 )) S3> .
Let (F`j ) C 2 (; M3 ) denote the solution to the corresponding system
i F`j (x) = pij (x)F`p (x), x ,

0
F`j (x0 ) = F`j
,
which exists and is unique by parts (i) and (ii). Then there exists an immersion
= (` ) C 3 (; E3 ) such that
j ` = F`j and gij = i j in .
To begin with, we show that the three vector fields defined by
g j := (F`j )3`=1 C 2 (; R3 )
satisfy
g i g j = gij in .
To this end, we note that, by construction, these fields satisfy
i g j = pij g p in ,
g j (x0 ) = g 0j ,
0
where g 0j is the j-th column vector of the matrix (F`j
) S3> . Hence the matrix
field (g i g j ) C 2 (; M3 ) satisfies
m
k (g i g j ) = m
kj (g m g i ) + ki (g m g j ) in ,
0
(g i g j )(x0 ) = gij
.
The definitions of the functions ijq and pij imply that

k gij = ikj + jki and ijq = gpq pij .
Hence the matrix field (gij ) C 2 (; S3> ) satisfies
m
k gij = m
kj gmi + ki gmj in ,
0
gij (x0 ) = gij
.
27
Viewed as a system of partial differential equations, together with initial

values at x0 , with respect to the matrix field (gij ) : M3 , the above system
can have at most one solution in the space C 2 (; M3 ).
To see this, let x1 be distinct from x0 and let C 1 ([0, 1]; R3 ) be any
path joining x0 to x1 in , as in part (ii). Then the nine functions gij ((t)),
0 t 1, satisfy a Cauchy problem for a linear system of nine ordinary
differential equations and this system has at most one solution.
An inspection of the two above systems therefore shows that their solutions
are identical, i.e., that g i g j = gij .
It remains to show that there exists an immersion C 3 (; E3 ) such that
i = g i in ,
where g i := (F`j )3`=1 .
Since the functions pij satisfy pij = pji , any solution (F`j ) C 2 (; M3 ) of
the system
i F`j (x) = pij (x)F`p (x), x ,
0
F`j (x0 ) = F`j
satisfies
i F`j = j Fì in .
The open set being simply-connected, Poincares theorem (for a proof, see,
e.g., Schwartz [1992, Vol. 2, Theorem 59 and Corollary 1, p. 234235]) shows
that, for each integer `, there exists a function ` C 3 () such that
i ` = Fì in ,
or, equivalently, such that the mapping := (` ) C 3 (; E3 ) satisfies
i = g i in .
That is an immersion follows from the assumed invertibility of the matrices
(gij ). The proof is thus complete.

Remarks. (1) The assumptions
j pik k pij + ìk pj` ìj pk` = 0 in ,
made in part (ii) on the functions pij = pji are thus sufficient conditions for
the equations i F`j = pij F`p in to have solutions. Conversely, a simple
computation shows that they are also necessary conditions, simply expressing
that, if these equations have a solution, then necessarily ik F`j = ki F`j in .
It is no surprise that these necessary conditions are of the same nature as
those of Theorem 1.5-1, viz., ik g j = ij g k in .
(2) The assumed positive definiteness of the matrices (gij ) is used only in
part (iii), for defining ad hoc initial vectors g 0i .
28
[Ch. 1
The definitions of the functions pij and ijq imply that the functions
satisfy, for all i, j, k, p,
Rqijk = Rjkqi = Rqikj ,
Rqijk = 0 if j = k or q = i.
These relations in turn imply that the eighty-one sufficient conditions
Rqijk = 0 in for all i, j, k, q {1, 2, 3},
are satisfied if and only if the six relations
R1212 = R1213 = R1223 = R1313 = R1323 = R2323 = 0 in
are satisfied (as is immediately verified, there are other sets of six relations that
will suffice as well, again owing to the relations satisfied by the functions Rqijk
for all i, j, k, q).
To conclude, we briefly review various extensions of the fundamental existence result of Theorem 1.6-1. First, a quick look at its proof reveals that it
holds verbatim in any dimension d 2, i.e., with R3 replaced by Rd and E3 by
a d-dimensional Euclidean space Ed . This extension only demands that Latin
indices and exponents now range in the set {1, 2, . . . , d} and that the sets of matrices M3 , S3 , and S3> be replaced by their d-dimensional counterparts Md , Sd ,
and Sd> .
The regularity assumptions on the components gij of the symmetric positive
definite matrix field C = (gij ) made in Theorem 1.6-1, viz., that gij C 2 (),
can be significantly weakened. More specifically, C. Mardare [2003a] has shown
that the existence theorem still holds if gij C 1 (), with a resulting mapping
in the space C 2 (; Ed ); likewise, S. Mardare [2004] has shown that the existence
2,
theorem still holds if gij Wloc
(), with a resulting mapping in the space
2,
Wloc
(; Ed ). As expected, the sufficient conditions Rqijk = 0 in of Theorem
1.6-1 are then assumed to hold only in the sense of distributions, viz., as
Z
{ikq j + ijq k + pij kqp pik jqp } dx = 0
for all D().

The existence result has also been extended up to the boundary of the set
by Ciarlet & C. Mardare [2004a]. More specifically, assume that the set
satisfies the geodesic property (in effect, a mild smoothness assumption on
the boundary , satisfied in particular if is Lipschitz-continuous) and that
the functions gij and their partial derivatives of order 2 can be extended by
continuity to the closure , the symmetric matrix field extended in this fashion
remaining positive-definite over the set . Then the immersion and its partial
derivatives of order 3 can be also extended by continuity to .
Sect. 1.7]
Uniqueness of immersions with the same metric tensor
29
Ciarlet & C. Mardare [2004a] have also shown that, if in addition the geodesic
distance is equivalent to the Euclidean distance on (a property stronger than
the geodesic property, but again satisfied if the boundary is Lipschitzcontinuous), then a matrix field (gij ) C 2 (; Sn> ) with a Riemann curvature
e Sn ) defined
tensor vanishing in can be extended to a matrix field (e
gij ) C 2 (;
>
e containing and whose Riemann curvature tensor
on a connected open set
e This result relies on the existence of continuous extensions
still vanishes in .
to of the immersion and its partial derivatives of order 3 and on a deep
extension theorem of Whitney [1934].
1.7
UNIQUENESS UP TO ISOMETRIES OF IMMERSIONS WITH THE SAME METRIC TENSOR
In Section 1.6, we have established the existence of an immersion : R3

E3 giving rise to a set () with a prescribed metric tensor, provided the given
metric tensor field satisfies ad hoc sufficient conditions. We now turn to the
question of uniqueness of such immersions.
This uniqueness result is the object of the next theorem, aptly called a
rigidity theorem in view of its geometrical interpretation: It asserts that, if
e C 1 (; E3 ) share the same metric tensor
two immersions C 1 (; E3 ) and
e
field, then the set () is obtained by subjecting the set () either to a
rotation (together represented by an orthogonal matrix Q with det Q = 1), or to
a symmetry with respect to a plane followed by a rotation (together represented
by an orthogonal matrix Q with det Q = 1), then by subjecting the rotated
set to a translation (represented by a vector c).
Such a geometric transformation is called a rigid deformation of the set
(), to remind that it indeed corresponds to the idea of a rigid one in E3 .
It is also an isometry, i.e., a transformation that preserves the distances.
Let O3 denote the set of all orthogonal matrices of order three.
Theorem 1.7-1. Let be a connected open subset of R3 and let C 1 (; E3 )
e C 1 (; E3 ) be two immersions such that their associated metric tensors
and
satisfy
e T
e in .
T =
Then there exist a vector c E3 and an orthogonal matrix Q O3 such
that
e
(x) = c + Q(x)
for all x .
Proof. For convenience, the three-dimensional vector space R3 is identified

throughout this proof with the Euclidean space E3 . In particular then, R3 inherits
the inner product and norm of E3 . The spectral norm of a matrix A M3 is
denoted
|A| := sup{|Ab|; b R3 , |b| = 1}.
30
[Ch. 1
e : E3 = R3 is
To begin with, we consider the special case where
the identity mapping. The issue of uniqueness reduces in this case to finding
C 1 (; E3 ) such that
(x)T (x) = I for all x .
Parts (i) to (iii) are devoted to solving these equations.
(i) We first establish that a mapping C 1 (; E3 ) that satisfies
(x)T (x) = I for all x
is locally an isometry: Given any point x0 , there exists an open neighborhood
V of x0 contained in such that
|(y) (x)| = |y x| for all x, y V.
Let B be an open ball centered at x0 and contained in . Since the set B is
convex, the mean-value theorem (for a proof, see, e.g., Schwartz [1992]) can be
applied. It shows that
|(y) (x)| sup |(z)||y x| for all x, y B.
z]x,y[
Since the spectral norm of an orthogonal matrix is one, we thus have

|(y) (x)| |y x| for all x, y B.
Since the matrix (x0 ) is invertible, the local inversion theorem (for a
proof, see, e.g., Schwartz [1992]) shows that there exist an open neighborhood
V of x0 contained in and an open neighborhood Vb of (x0 ) in E3 such that
the restriction of to V is a C 1 -diffeomorphism from V onto Vb . Besides, there
is no loss of generality in assuming that V is contained in B and that Vb is
convex (to see this, apply the local inversion theorem first to the restriction of
to B, thus producing a first neighborhood V 0 of x0 contained in B, then to
the restriction of the inverse mapping obtained in this fashion to an open ball
V centered at (x0 ) and contained in (V 0 )).
Let 1 : Vb V denote the inverse mapping of : V Vb . The chain
rule applied to the relation 1 ((x)) = x for all x V then shows that
b 1 (b
x) = (x)1 for all x

b = (x), x V.
b 1 (b
The matrix
x) being thus orthogonal for all x
b Vb , the mean-value
theorem applied in the convex set Vb shows that
|1 (b
y ) 1 (b
x)| |b
yx
b| for all x
b, yb Vb ,
or equivalently, that
|y x| |(y) (x)| for all x, y V.
Sect. 1.7]
31
The restriction of the mapping to the open neighborhood V of x0 is thus

an isometry.
(ii) We next establish that, if a mapping C 1 (; E3 ) is locally an isometry, in the sense that, given any x0 , there exists an open neighborhood V
of x0 contained in such that |(y) (x)| = |y x| for all x, y V , then its
derivative is locally constant, in the sense that
(x) = (x0 ) for all x V.
The set V being that found in (i), let the differentiable function F : V V
R be defined for all x = (xp ) V and all y = (yp ) V by
F (x, y) := (` (y) ` (x))(` (y) ` (x)) (y` x` )(y` x` ).
Then F (x, y) = 0 for all x, y V by (i). Hence
Gi (x, y) :=
1 F
`
(x, y) =
(y)(` (y) ` (x)) i` (y` x` ) = 0
2 yi
yi
for all x, y V . For a fixed y V , each function Gi (, y) : V R is differentiable and its derivative vanishes. Consequently,
`
`
Gi
(x, y) =
(y)
(x) + ij = 0 for all x, y V,
xi
yi
xj
or equivalently, in matrix form,
(y)T (x) = I for all x, y V.
Letting y = x0 in this relation shows that
(x) = (x0 ) for all x V.
(iii) By (ii), the mapping : M3 is differentiable and its derivative
vanishes in . Therefore the mapping : E3 is twice differentiable and
its second Frechet derivative vanishes in . The open set being connected,
a classical result from differential calculus (see, e.g., Schwartz [1992, Theorem
3.7.10]) shows that the mapping is affine in , i.e., that there exists a vector
c E3 and a matrix Q M3 such that
(x) = c + Qox for all x .
Since Q = (x0 ) and (x0 )T (x0 ) = I by assumption, the matrix
Q is orthogonal.
(iv) We now consider the general equations gij = geij in , noting that they
also read
T
e
e
(x)T (x) = (x)
(x)
for all x .
32
[Ch. 1
Given any point x0 , let the neighborhoods V of x0 and Vb of (x0 )

and the mapping 1 : Vb V be defined as in part (i) (by assumption, the
mapping is an immersion; hence the matrix (x0 ) is invertible).
Consider the composite mapping
b :=
e 1 : Vb E3 .
b C 1 (Vb ; E3 ) and
Clearly,
b (b
b x) = (x)
e
b 1 (b
x)
1
e
= (x)(x) for all x
b = (x), x V.
Hence the assumed relations
imply that
T
e
e
(x)T (x) = (x)
(x)
for all x
b (b
b x )T
b (b
b x) = I for all x V.
By parts (i) to (iii), there thus exist a vector c R3 and a matrix Q O3

such that
b x) = (x)
e
(b
= c + Q(x) for all x
b = (x), x V,
hence such that
1
e
(x) := (x)(x)
= Q for all x V.
The continuous mapping : V M3 defined in this fashion is thus locally

constant in . As in part (iii), we conclude from the assumed connectedness of
that the mapping is constant in . Thus the proof is complete.

The special case where is the identity mapping of R3 identified with E3 is
the classical Liouville theorem. This theorem thus asserts that if a mapping
C 1 (; E3 ) is such that (x) O3 for all x where is an open
connected subset of R3 , then there exist c E3 and Q O3 such that
(x) = c + Q ox for all x .
e C 1 (; E3 ) and C 1 (; E3 ) are said to be isometriTwo mappings
e = c + Q in
cally equivalent if there exist c E3 and Q O3 such that
e
, i.e., such that = J , where J := c + Q id is thus an isometry. Theorem
e C 1 (; E3 ) and C 1 (; E3 ) share
1.7-1 thus asserts that two mappings
the same metric tensor field over an open connected subset of R3 if and only
if they are isometrically equivalent.
Remarks. (1) In terms of covariant components gij of metric tensors, parts (i)
to (iii) of the above proof provide the solution to the equations gij = ij in ,
Sect. 1.7]
33
e j
e in ,
while part (iv) provides the solution to the equations gij = i
1
3
e
where C (; E ) is a given immersion.
(2) The classical Mazur-Ulam theorem asserts the following: Let be a
connected subset in Rd , and let : Rd be a mapping that satisfies
|(y) (x)| = |y x| for all x, y .
Then there exist a vector c Rd and an orthogonal matrix Q of order d such
that
(x) = c + Qox for all x .
Parts (ii) and (iii) of the above proof thus provide a proof of this theorem
under the additional assumption that the mapping is of class C 1 (the extension
from R3 to Rd is trivial).

While the immersions found in Theorem 1.6-1 are thus only defined up to
isometries in E3 , they become uniquely determined if they are required to satisfy
ad hoc additional conditions, according to the following corollary to Theorems
1.6-1 and 1.7-1.
Theorem 1.7-2. Let the assumptions on the set and on the matrix field C
be as in Theorem 1.6-1, let a point x0 be given, and let F0 M3 be any
matrix that satisfies
FT0 F0 = C(x0 ).
Then there exists one and only one immersion C 3 (; E3 ) that satisfies
(x)T (x) = C(x) for all x ,
(x0 ) = 0 and (x0 ) = F0 .
Proof. Given any immersion C 3 (; E3 ) that satisfies (x)T (x) =
C(x) for all x (such immersions exist by Theorem 1.6-1), let the mapping
: R3 be defined by
(x) := F0 (x0 )1 ((x) (x0 )) for all x .
Then it is immediately verified that this mapping satisfies the announced
properties.
Besides, it is uniquely determined. To see this, let C 3 (; E3 ) and
C 3 (; E3 ) be two immersions that satisfy
(x)T (x) = (x)T (x) for all x .
Hence there exist (by Theorem 1.7-1) c R3 and Q O3 such that (x) =
c + Q(x) for all x , so that (x) = Q(x) for all x . The relation
(x0 ) = (x0 ) then implies that Q = I and the relation (x0 ) = (x0 )
in turn implies that c = 0.
34
[Ch. 1
Remark. One possible choice for the matrix F0 is the square root of the
symmetric positive-definite matrix C(x0 ).

Theorem 1.7-1 constitutes the classical rigidity theorem, in that both ime are assumed to be in the space C 1 (; E3 ). The next theorem
mersions and
is an extension, due to Ciarlet & C. Mardare [2003], that covers the case where
one of the mappings belongs to the Sobolev space H 1 (; E3 ).
The way the result in part (i) of the next proof is derived is due to Friesecke,
James & M
uller [2002]; the result of part (i) itself goes back to Reshetnyak
[1967].
Let O3+ denote the set of all rotations, i.e., of all orthogonal matrices Q O3
with det Q = 1.
Theorem 1.7-3. Let be a connected open subset of R3 , let C 1 (; E3 ) be
a mapping that satisfies
det > 0 in ,
e H 1 (; E3 ) be a mapping that satisfies
and let
e > 0 a.e. in and T =

e T
e a.e. in .
det
Then there exist a vector c E3 and a rotation Q O3+ such that

e
(x)
= c + Q(x) for almost all x .
Proof. The Euclidean space E3 is identified with the space R3 throughout

the proof.
(i) To begin with, we consider the special case where = id . In other
e H1 () that satisfies (x)
e
words, we are given a mapping
O3+ for
almost all x . Hence
T
T
e
e
e
e
Cof (x)
= (det (x))
(x)
= (x)
for almost all x ,
on the one hand. Since, on the other hand,
e = 0 in (D0 (B))3
div Cof
in any open ball B such that B (to see this, combine the density of C 2 (B)
in H 1 (B) with the classical Piola identity in the space C 2 (B); for a proof of this
identity, see, e.g., Ciarlet [1988, Theorem 1.7.1]), we conclude that
e = div Cof
e = 0 in (D0 (B))3 .
e = (
e j ) (C ())3 . For such mappings, the identity
Hence
e j i
e j ) = 2i
e j i (
e j ) + 2ik
e j ik
e j,
(i
Sect. 1.7]
35
e j = 0 and i
e j i
e j = 3 in , shows that ik
ej =
together with the relations
0 in . The assumed connectedness of then implies that there exist a vector
e
c E3 and a matrix Q O3+ (by assumption, (x)
O3+ for almost all
x ) such that
e
(x)
= c + Q ox for almost all x .
(ii) We consider next the general case. Let x0 be given. Since

is an immersion, the local inversion theorem can be applied; there thus exist
b of (x0 ) satisfying U and
bounded open neighborhoods U of x0 and U
b
{U}
(), such that the restriction U of to U can be extended to a
1
b .
C -diffeomorphism from U onto {U}
1
b
Let U : U U denote the inverse mapping of U , which therefore
1
b 1 (b
b (the notation
b indicates
satisfies
for all x
b = (x) U
U x) = (x)
b Define
that differentiation is carried out with respect to the variable x
b U).
the composite mapping
b :=
e 1 : U
b R3 .
e H1 (U ) and 1 can be extended to a C 1 -diffeomorphism from {U}

b
Since
U
1
3
b H (U
b ; R ) and that
onto U , it follows that
1
b (b
b x) = (x)
e
b 1 (b
e
U x) = (x)(x)
b (see, e.g., Adams [1975, Chapter 3]). Hence

for almost all x
b = (x) U
e > 0 a.e. in , and T =
the assumptions det > 0 in , det
T
e
e a.e. in , together imply that
b (b
b x) O3 for almost all x
b . By
bU
+
3
3
(i), there thus exist c E and Q O+ such that
b x) = (x)
e
b,
(b
= c + Q ob
x for almost all x
b = (x) U
or equivalently, such that
1
e
(x) := (x)(x)
= Q for almost all x U.
Since the point x0 is arbitrary, this relation shows that L1loc ().
By a classical result from distribution theory (cf. Schwartz [1966, Section 2.6]),
we conclude from the assumed connectedness of that (x) = Q for almost all
x , and consequently that
e
(x)
= c + Q(x) for almost all x .
36
[Ch. 1
e H 1 (; E3 ) satisfying the assumptions of

Remarks. (1) The existence of
e C 1 (; E3 ).
Theorem 1.7-3 thus implies that H 1 (; E3 ) and
1
3
e C (; E ), the assumptions det > 0 in and det
e > 0 in
(2) If
are no longer necessary; but then it can only be concluded that Q O3 : This
is the classical rigidity theorem (Theorem 1.7-1), of which Liouvilles theorem
is the special case corresponding to = id .
(3) The result established in part (i) of the above proof asserts that, given
a connected open subset of R3 , if a mapping H 1 (; E3 ) is such that
(x) O3+ for almost all x , then there exist c E3 and Q O3+ such
that (x) = c + Q ox for almost all x . This result thus constitutes a
generalization of Liouvilles theorem.
e is assumed to be instead in the space
(4) By contrast, if the mapping
e
H 1 (; E3 ) (as in Theorem 1.7-3), an assumption about the sign of det
becomes necessary. To see this, let for instance be an open ball centered at the
e
e
origin in R3 , let (x) = x, and let (x)
= x if x1 0 and (x)
= (x1 , x2 , x3 )
1
3
3
e
e
if x1 < 0. Then H (; E ) and O a.e. in ; yet there does
e
= Q ox for all x , since
not exist any orthogonal matrix such that (x)
3
e
()
{x R ; x1 0} (this counter-example was kindly communicated to
the author by Sorin Mardare).
(5) Surprisingly, the assumption det > 0 in cannot be replaced by
the weaker assumption det > 0 a.e. in . To see this, let for instance
be an open ball centered at the origin in R3 , let (x) = (x1 x22 , x2 , x3 ) and let
e
e
(x)
= (x) if x2 0 and (x)
= (x1 x22 , x2 , x3 ) if x2 < 0 (this counterexample was kindly communicated to the author by Herve Le Dret).
(6) If a mapping C 1 (; E3 ) satisfies det > 0 in , then is an
immersion. Conversely, if is a connected open set and C 1 (; E3 ) is an
immersion, then either det > 0 in or det < 0 in . The assumption
that det > 0 in made in Theorem 1.7-3 is simply intended to fix ideas (a
similar result clearly holds under the other assumption).
(7) A little further ado shows that the conclusion of Theorem 1.7-3 is still
e H 1 (; E3 ) is replaced by the weaker assumption
e H 1 (; E3 ).
valid if
loc

Like the existence results of Section 1.6, the uniqueness theorems of this
section hold verbatim in any dimension d 2, with R3 replaced by Rd and Ed
by a d-dimensional Euclidean space.
1.8
CONTINUITY OF AN IMMERSION AS A FUNCTION OF ITS METRIC TENSOR
Let be a connected and simply-connected open subset of R3 . Together, Theorems 1.6-1 and 1.7-1 establish the existence of a mapping F that associates
with any matrix field C = (gij ) C 2 (; S3> ) satisfying
Rqijk := j ikq k ijq + pij kqp pik jqp = 0 in ,
Sect. 1.8]
An immersion as a function of its metric tensor
37
where the functions ijq and pij are defined in terms of the functions gij as in
Theorem 1.6-1, a well-defined element F(C) in the quotient set C 3 (; E3 )/R,
e R means that there exist a vector a E3 and a matrix Q O3
where (, )
e
such that (x) = a + Q(x)
for all x .
A natural question thus arises as to whether there exist natural topologies on
the space C 2 (; S3 ) and on the quotient set C 3 (; E3 )/R such that the mapping
F defined in this fashion is continuous.
Equivalently, is an immersion a continuous function of its metric tensor?
The object of this section, which is based on Ciarlet & Laurent [2003], is to
provide an affirmative answer to this question (see Theorem 1.8-5).
Note that such a question is not only clearly relevant to differential geometry per se, but it also naturally arises in nonlinear three-dimensional elasticity,
where a smooth enough immersion : E3 may be thought of as a deformation of the set viewed as a reference configuration of a nonlinearly elastic
body (although such an immersion should then be in addition injective and
orientation-preserving in order to qualify for this definition; for details, see, e.g.,
Ciarlet [1988, Section 1.4] or Antman [1995, Section 12.1]). In this context, the
associated matrix
C(x) = (gij (x)) = (x)T (x),
is called the (right) Cauchy-Green tensor at x and the matrix
(x) = (j i (x)) M3 ,
representing the Frechet derivative of the mapping at x, is called the deformation gradient at x.
The Cauchy-Green tensor field C = T : S3> associated with a
deformation : E3 plays a major role in the theory of nonlinear threedimensional elasticity, since the response function, or the stored energy function,
of a frame-indifferent elastic, or hyperelastic, material necessarily depends on
the deformation gradient through the Cauchy-Green tensor (see, e.g., Ciarlet
[1988, Chapters 3 and 4]. As already suggested by Antman [1976], the CauchyGreen tensor field of the unknown deformed configuration could thus also be
regarded as the primary unknown rather than the deformation itself as is
customary.
To begin with, we list some specific notations that will be used in this section
for addressing the question raised above. Given a matrix A M3 , we let (A)
denote its spectral radius and we let
|A| := sup
bR
b6=0
denote its spectral norm.
|Ab|
= {(AT A)}1/2
|b|
38
[Ch. 1
Let be an open subset of R3 . The notation K b means that K is

a compact subset of . If g C ` (; R), ` 0, and K b , we define the
semi-norms
|g|`,K = sup | g(x)|
xK
||=`
and kgk`,K = sup | g(x)|,

xK
||`
where stands for the standard multi-index notation for partial derivatives.
If C ` (; E3 ) or A C ` (; M3 ), ` 0, and K b , we likewise set
||`,K =
|A|`,K =
sup | (x)|
and
sup | A(x)|
and
xK
||=`
xK
||=`
kk`,K =
kAk`,K =
sup | (x)|,
xK
||`
sup | A(x)|,
xK
||`
where || denotes either the Euclidean vector norm or the matrix spectral norm.
The next sequential continuity results (Theorems 1.8-1, 1.8-2, and 1.8-3)
constitute key steps toward establishing the continuity of the mapping F (see
n
Theorem 1.8-5). Note that the functions Rqijk
occurring in their statements are
n
meant to be constructed from the functions gij
in the same way that the functions Rqijk are constructed from the functions gij . To begin with, we establish
the sequential continuity of the mapping F at C = I.
Theorem 1.8-1. Let be a connected and simply-connected open subset of
n
n
R3 . Let Cn = (gij
) C 2 (; S3> ), n 0, be matrix fields satisfying Rqijk
= 0 in
, n 0, such that
lim kCn Ik2,K = 0 for all K b .
Then there exist mappings n C 3 (; E3 ) satisfying (n )T n = Cn in

, n 0, such that
lim kn idk3,K = 0 for all K b
where id denotes the identity mapping of R3 , identified here with E3 .

Proof. The proof is broken into four parts, numbered (i) to (iv). The first
part is a preliminary result about matrices (for convenience, it is stated here for
matrices of order three, but it holds as well for matrices of arbitrary order).
(i) Let matrices An M3 , n 0, satisfy
lim (An )T An = I.
Then there exist matrices Qn O3 , n 0, that satisfy

lim Qn An = I.
Sect. 1.8]
39
Since the set O3 is compact, there exist matrices Qn O3 , n 0, such that

|Qn An I| = inf 3 |RAn I|.
RO
We assert that the matrices Qn defined in this fashion satisfy limn Qn An =

I. For otherwise, there would exist a subsequence (Qp )p0 of the sequence
(Qn )n0 and > 0 such that
|Qp Ap I| = inf 3 |RAp I| for all p 0.
RO
Since
lim |Ap | = lim
((Ap )T Ap ) =
p
(I) = 1,
the sequence (Ap )p0 is bounded. Therefore there exists a further subsequence
(Aq )q0 that converges to a matrix S, which is orthogonal since
ST S = lim (Aq )T Aq = I.
q
But then
lim ST Aq = ST S = I,
which contradicts inf RO3 |RAq I| for all q 0. This proves (i).
In the remainder of this proof, the matrix fields Cn , n 0, are meant to be
those appearing in the statement of Theorem 1.8-1.
(ii) Let mappings n C 3 (; E3 ), n 0, satisfy (n )T n = Cn in
(such mappings exist by Theorem 1.6-1). Then
lim |n id|`,K = lim |n |`,K = 0 for all K b and for ` = 2, 3.
As usual, given any immersion C 3 (; E3 ), let g i = i , let gij = g i g j ,

and let the vectors g q be defined by the relations g i g q = iq . It is then
immediately verified that
ij = i g j = (i g j g q )g q =
1
(j giq + i gjq q gij )g q .
2
Applying this relation to the mappings n thus gives

ij n =
1
n
n
n
q gij
)(g q )n , n 0,
(j giq
+ i gjq
2
where the vectors (g q )n are defined by means of the relations i n (g q )n = iq .

Let K denote an arbitrary compact subset of . On the one hand,
n
n
n
q gij
|0,K = 0,
lim |j giq
+ i gjq
40
[Ch. 1
n
n
since limn |gij
|1,K = limn |gij
ij |1,K = 0 by assumption. On the other
hand, the norms |(g q )n |0,K are bounded independently of n 0; to see this,
observe that (g q )n is the q-th column vector of the matrix (n )1 , then that
|(n )1 |0,K = |{((n )T (n )1 )}1/2 |0,K
n 1 1/2
n 1
= |{((gij
) )} |0,K {|(gij
) |0,K }1/2 ,
and, finally, that

n
n 1
lim |(gij
) I|0,K = 0 = lim |(gij
) I|0,K = 0.
Consequently,
lim |n id|2,K = lim |n |2,K = 0 for all K b .
Differentiating the relations i g j g q = 21 (j giq + i gjq q gij ) yields

ijp = ip g j = (ip g j g q )g q

1
(jp giq + ip gjq pq gij ) i g j p g q g q .
=
2
n
n
Observing that limn |gij
|`,K = limn |gij
ij |`,K = 0 for ` = 1, 2 by
q n
assumption and recalling that the norms |(g ) |0,K are bounded independently
of n 0, we likewise conclude that
lim |n id|3,K = lim |n |3,K = 0 for all K b .
e n C 3 (; E3 ) that satisfy (
e n )T
e n = Cn
(iii) There exist mappings
in , n 0, and
n
e id|1,K = 0 for all K b .

lim |
Let n C 3 (; E3 ) be mappings that satisfy ( n )T n = Cn in ,

n 0 (such mappings exist by Theorem 1.6-1), and let x0 denote a point in the
set . Since limn n (x0 )T n (x0 ) = I by assumption, part (i) implies
that there exist orthogonal matrices Qn (x0 ), n 0, such that
lim Qn (x0 ) n (x0 ) = I.
e n C 3 (; E3 ), n 0, defined by
Then the mappings
satisfy
e n (x) := Qn (x0 ) n (x), x ,
e n )T
e n = Cn in ,
(
Sect. 1.8]
41
e C 2 (; M3 ) satisfy
so that their gradients
e n |0,K = lim |
e n |2,K = 0 for all K b ,
lim |i
by part (ii). In addition,
e (x0 ) = lim Qn n (x0 ) = I.

lim
Hence a classical theorem about the differentiability of the limit of a sequence

of mappings that are continuously differentiable on a connected open set and
that take their values in a Banach space (see, e.g., Schwartz [1992, Theorem
e n uniformly converge on every compact
3.5.12]) shows that the mappings
subset of toward a limit R C 1 (; M3 ) that satisfies
n
e (x) = 0 for all x .

i R(x) = lim i
n
This shows that R is a constant mapping since is connected. Consequently,

e n (x0 ) = I. We have therefore
R = I since in particular R(x0 ) = limn
established that
e n id|1,K = lim |
e n I|0,K = 0 for all K b .
lim |
(iv) There exist mappings n C 3 (; E3 ) satisfying (n )T n = Cn

in , n 0, and
lim |n id|`,K = 0 for all K b and for ` = 0, 1.
The mappings
clearly satisfy

n
e {
e n (x0 ) x0 } C 3 (; E3 ), n 0,
n :=
(n )T n = Cn in , n 0,
lim |n id|1,K = lim |n I|0,K = 0 for all K b ,
n (x0 ) = x0 , n 0.
Again applying the theorem about the differentiability of the limit of a sequence of mappings used in part (iii), we conclude from the last two relations
that the mappings n uniformly converge on every compact subset of toward
a limit C 1 (; E3 ) that satisfies
(x) = lim n (x) = I for all x .
n
42
[Ch. 1
This shows that ( id) is a constant mapping since is connected. Consequently, = id since in particular (x0 ) = limn n (x0 ) = x0 . We have
thus established that
lim |n id|0,K = 0 for all K b .
This completes the proof of Theorem 1.8-1.
We next establish the sequential continuity of the mapping F at those matrix

fields C C 2 (; S3> ) that can be written as C = T with an injective
mapping C 3 (; E3 ).
Theorem 1.8-2. Let be a connected and simply-connected open subset of R 3 .

n
Let C = (gij ) C 2 (; S3> ) and Cn = (gij
) C 2 (; S3> ), n 0, be matrix fields
n
satisfying respectively Rqijk = 0 in and Rqijk
= 0 in , n 0, such that
lim kCn Ck2,K = 0 for all K b .
Assume that there exists an injective mapping C 3 (; E3 ) such that

T = C in . Then there exist mappings n C 3 (; E3 ) satisfying
(n )T n = Cn in , n 0, such that
lim kn k3,K = 0 for all K b .
Proof. The assumptions made on the mapping : R3 E3 imply that

b := () E3 is open, connected, and simply-connected, and that
the set
b :
b E3 R3 belongs to the space C 3 (;
b R3 ). Define
the inverse mapping
n
b S3> ), n 0, by letting
the matrix fields (b
gij
) C 2 (;
n
n
b
(b
gij
(b
x)) := (x)T (gij
(x))(x)1 for all x
b = (x) .
b of ,
b let K := (
b K).
b Since limn kg n
Given any compact subset K
ij
gij k2,K = 0 because K is a compact subset of , the definition of the functions
n
b R and the chain rule together imply that
gbij
:
n
lim kb
gij
ij k2,K
b = 0.
b let bi = /b
b n denote the functions conGiven x
b = (b
xi ) ,
xi . Let R
qijk
n
structed from the functions gbij in the same way that the functions Rqijk are
constructed from the functions gij . Since it is easily verified that these funcb n = 0 in ,
b Theorem 1.8-1 applied over the set
b shows that
tions satisfy R
qijk
n
3 b
3
b
there exist mappings C (; E ) satisfying
such that
b n bj
b n = gbn in ,
b n 0,
bi
ij
c b = 0 for all K
b n idk
b b ,
b
lim k
3,K
Sect. 1.8]
43
where c
id denotes the identity mapping of E3 , identified here with R3 . Define
the mappings n C 3 (; S3> ), n 0, by letting
b n (b
b x) .
n (x) =
x) for all x = (b
b := (K). Since limn k

bn
Given any compact subset K of , let K
n
c
idk3,K
b = 0, the definition of the mappings and the chain rule together
imply that
lim kn k3,K = 0,
n
on the one hand. Since, on the other hand, (n )T n = Cn in , the proof

is complete.

We are now in a position to establish the sequential continuity of the mapping
F at any matrix field C C 2 (; S3> ) that can be written as C = T with
C 3 (; E 3 ).

n
Let C = (gij ) C 2 (; S3> ) and Cn = (gij
) C 2 (, S3> ), n 0, be matrix fields
n
respectively satisfying Rqijk = 0 in and Rqijk
= 0 in , n 0, such that
Let C 3 (; E3 ) be any mapping that satisfies T = C in (such

mappings exist by Theorem 1.6-1). Then there exist mappings n C 3 (; E3 )
satisfying (n )T n = Cn in , n 0, such that
Proof. The proof is broken into four parts. In what follows, C and Cn
designate matrix fields possessing the properties listed in the statement of the
theorem.
(i) Let C 3 (; E3 ) be any mapping that satisfies T = C in .
ThenSthere exist a countable number of open balls BSr , r 1, such that
r
= r=1 Br and such that, for each r 1, the set s=1 Bs is connected and
the restriction of to Br is injective.
Given any x , there exists anSopen ball Vx such that the restriction
of to Vx is injective. Since = x Vx can also be written as a countable
union of compact subsets of , there already
exist countably many such open
S
balls, denoted Vr , r 1, such that =
r=1 Vr .
Let r1 := 1, B1 := Vr1 , and r2 := 2. If the set Br1 Vr2 is connected,
let B2 := Vr2 and r3 := 3. Otherwise, there exists a path 1 in joining
the centers of Vr1 and Vr2 since is connected. Then there exists a finite set
I1 = {r1 (1), r1 (2), , r1 (N1 )} of integers, with N1 1 and 2 < r1 (1) < r1 (2) <
< r1 (N1 ), such that
[
1 V r1 V r2
Vr .
rI1
44
[Ch. 1
Furthermore there exists a permutation 1 of {1, 2, . . . , N1 } such that the sets

Sr
SN 1
Vr1 ( s=1 V1 (s) ), 1 r N1 , and Vr1 ( s=1
V1 (s) ) Vr2 are connected.
Let
Br := V1 (r1) , 2 r N1 + 1, BN1 +2 := Vr2 ,

r3 := min i {1 (1), . . . , 1 (N1 )}; i 3 .
S 1 +2
If the set ( N
r=1 Br ) Vr3 is connected, let BN1 +3 := Vr3 . Otherwise, apply
the same argument as above to a path 2 in joining the centers of Vr2 and
Vr3 , and so forth.
The iterative procedure thus produces a countable number of open
Sballs
Br , r 1, that possess the announced properties. In particular, = r=1 Br
since, by construction, the integer ri appearing at the i-th stage satisfies ri i.
en
(ii) By Theorem 1.8-2, there exist mappings n1 C 3 (B1 ; E3 ) and
2
C 3 (B2 ; E3 ), n 0, that satisfy
(n1 )T n1 = Cn in B1
e n )T
e n = Cn in B2
(
2
2
and
and
lim kn1 k3,K = 0 for all K b B1 ,
e k3,K = 0 for all K b B2 ,

lim k
2
and by Theorem 1.7-1, there exist vectors cn E3 and matrices Qn O3 , n 0,

such that
e n (x) = cn + Qn n (x) for all x B1 B2 .
2
1
Then we assert that
lim cn = 0 and lim Qn = I.
Let (Qp )p0 be a subsequence of the sequence (Qn )n0 that converges to
a (necessarily orthogonal) matrix Q and let x1 denote a point in the set B1
e p (x1 ) Qp 1 (x1 ) and limn
e p (x1 ) = limn p (x1 ) =
B2 . Since cp =
2
2
1
p
(x1 ), the subsequence (c )p0 also converges. Let c := limp cp . Thus
e p2 (x)
(x) = lim
p
= lim (cp + Qp p1 (x)) = c + Q(x) for all x B1 B2 ,

p
on the one hand. On the other hand, the differentiability of the mapping
implies that
(x) = (x1 ) + (x1 )(x x1 ) + o(|x x1 |) for all x B1 B2 .
Note that (x1 ) is an invertible matrix, since (x1 )T (x1 ) = (gij (x1 )).
Let b := (x1 ) and A := (x1 ). Together, the last two relations imply
that
b + A(x x1 ) = c + Qb + QA(x x1 ) + o(|x x1 |),
Sect. 1.8]
45
and hence (letting x = x1 shows that b = c + Qb) that

A(x x1 ) = QA(x x1 ) + o(|x x1 |) for all x B1 B2 .
The invertibility of A thus implies that Q = I and therefore that c = bQb = 0.
The uniqueness of these limits shows that the whole sequences (Qn )n0 and
(cn )n0 converge.
(iii) Let the mappings n2 C 3 (B1 B2 ; E3 ), n 0, be defined by
n2 (x) := n1 (x) for all x B1 ,
n
e (x) cn ) for all x B2 .

n2 (x) := (Qn )T (
2
Then
(n2 )T n2 = Cn in B1 B2
(as is clear), and
lim kn2 k3,K = 0 for all K b B1 B2 .
The plane containing the intersection of the boundaries of the open balls B1
and B2 is the common boundary of two closed half-spaces in R3 , H1 containing
the center of B1 , and H2 containing that of B2 (by construction, the set B1 B2
is connected; see part (i)). Any compact subset K of B1 B2 may thus be written
as K = K1 K2 , where K1 := (K H1 ) B1 and K2 := (K H2 ) B2 (that
the open sets found in part (i) may be chosen as balls thus play an essential r
ole
here). Hence
lim kn2 k3,K1 = 0 and lim kn2 k3,K2 = 0,
the second relation following from the definition of the mapping n2 on B2 K2

e n k3,K2 = 0 (part (ii)) and limn Qn = I
and on the relations limn k
2
and limn cn = 0 (part (iii)).
(iv) It remains to iterate the procedure described
in parts (ii) and (iii). For
Sr
some r 2, assume that mappings nr C 3 ( s=1 Bs ; E3 ), n 0, have been
found that satisfy
(nr )T nr = Cn in
r
[
Bs ,
s=1
lim knr k2,K = 0 for all K b
r
[
Bs .
s=1
Since the restriction of to Br+1 is injective (part (i)), Theorem 1.8-2 shows
e nr+1 C 3 (Br+1 ; E3 ), n 0, that satisfy
that there exist mappings
e n )T
e n = Cn in Br+1 ,
(
r+1
r+1
n
e
lim k
r+1 k3,K = 0 for all K b Br+1 ,
46
[Ch. 1
S
and since the set r+1
s=1 Bs is connected (part (i)), Theorem 1.7-1 shows that
there exist vectors cn E3 and matrices Qn O3 , n 0, such that
e n (x) = cn + Qn n (x) for all x
r+1
r
r
[
s=1

Bs Br+1 .
Then an argument similar to that used in part (ii) shows that limn Qn = I
and limn cn = 0, and an argument similar to that used in part (iii) (note
that the ball Br+1 may intersect more
Sr than one of the balls Bs , 1 s r)
shows that the mappings nr+1 C 3 ( s=1 Bs ; E3 ), n 0, defined by
nr+1 (x) := nr (x) for all x
satisfy
r
[
Bs ,
s=1
e n (x) cn ) for all x Br+1 ,

nr+1 (x) := (Qn )T (
r
lim knr+1
n
k3,K = 0 for all K b
r
[
Bs .
s=1
Then the mappings n : E3 , n 0, defined by

n (x) := nr (x) for all x
r
[
s=1
Bs , r 1,
possess all the required properties:SThey are unambiguously defined since for all
r
s > r, ns (x) = nr (x) forSall x s=1 Bs by construction; they are of class C 3
r
n
since the mappings r : s=1 Bs E3 are themselves of class C 3 ; they satisfy
n T
n
n
n
(
r satisfy the same relations
Sr ) = C in since the mappings
n
in s=1 Bs ; and finally, they satisfy limn k
S k3,K = 0 for all K b
since any compact subset of is contained in rs=1 Bs for r large enough. This
completes the proof.

It is easily seen that the assumptions Rqijk = 0 in are in fact superfluous
in Theorem 1.8-3 (as shown in the next proof, these relations are consequences
n
of the assumptions Rqijk
= 0 in , n 0, and limn kCn Ck2,K = 0 for
all K b ). This observation gives rise to the following corollary to Theorem
1.8-3, in the form of another sequential continuity result, of interest by itself. The
novelties are that the assumptions are now made on the immersions n , n 0,
and that this result also provides the existence of a limit immersion .
Theorem 1.8-4. Let be a connected and simply-connected open subset of
R3 . Let there be given immersions n C 3 (; E3 ), n 0, and a matrix field
C C 2 (; S3> ) such that
lim k(n )T n Ck2,K = 0 for all K b .
Sect. 1.8]
47
e n C 3 (; E3 ), n 0, of the form
Then there exist mappings
e n = c n + Q n n , c n E 3 , Qn O 3 ,
e n )T
e n = (n )T n in for all n 0, and there
which thus satisfy (
exists a mapping C 3 (; E3 ) such that
n
e k3,K = 0 for all K b .

T = C in and lim k
n
n
Proof. Let the functions Rqijk
, n 0, and Rqijk be constructed from the
n
components gij and gij of the matrix fields Cn := (n )T n and C in the
n
usual way (see, e.g., Theorem 1.6-1). Then Rqijk
= 0 in for all n 0, since
these relations are simply the necessary conditions of Theorem 1.5-1.
We now show that Rqijk = 0 in . To this end, let K be any compact subset
of . The relations
Cn = C(I + C1 (Cn C)), n 0,

together with the inequalities kABk2,K 4kAk2,K kBk2,K valid for any matrix
fields A, B C 2 (; M3 ), show that there exists n0 = n0 (K) such that the matrix
fields (I + C1 (Cn C))(x) are invertible at all x K for all n n0 . The same
relations also show that there exists a constant M such that k(Cn )1 k2,K M
for all n n0 . Hence the relations
(Cn )1 C1 = C1 (C Cn )(Cn )1 , n n0 ,
together with the assumptions limn kCn Ck2,K = 0, in turn imply that the
components g ij,n , n n0 , and g ij of the matrix fields (Cn )1 and C1 satisfy
lim kg ij,n g ij k2,K = 0.
With self-explanatory notations, it thus follows that

p
lim knijq ijq k1,K = 0 and lim kp,n
ij ij k1,K = 0,
n
limn kRqijk
hence that
Rqijk k0,K = 0. This shows that Rqijk = 0 in K,
hence that Rqijk = 0 in since K is an arbitrary compact subset of .
By the fundamental existence theorem (Theorem 1.6-1), there thus exists a
mapping C 3 (; E3 ) such that T = C in . Theorem 1.8-3 can now
e n C 3 (; E3 ) such that
be applied, showing that there exist mappings
n
e )T
e = Cn in , n 0, and lim k
e k3,K for all K b .
(
n
Finally, the rigidity theorem (Theorem 1.7-1) shows that, for each n 0,
e n = cn + Qn n in because
there exist cn E3 and Qn O3 such that
n
e and n share the same metric tensor field and the set is
the mappings
connected.
48
[Ch. 1
It remains to show how the sequential continuity established in Theorem

1.8-3 implies the continuity of a deformation as a function of its metric tensor
for ad hoc topologies.
Let be an open subset of R3 . For any integers ` 0 and d 1, the space
C ` (; Rd ) becomes a locally convex topological space when its topology is defined
by the family of semi-norms kk`,K , K b , defined earlier. Then a sequence
(n )n0 converges to with respect to this topology if and only if
lim kn k`,K = 0 for all K b .
Furthermore, this topology is metrizable: Let (Ki )i0 be any sequence of subsets
of that satisfy
Ki b and Ki int Ki+1 for all i 0, and =
Ki .
i=0
Then
lim kn k`,K = 0 for all K b lim d` (n , ) = 0,
where
d` (, ) :=
X
1 k k`,Ki
.
i 1 + k k
2
`,Ki
i=0
For details, see, e.g., Yosida [1966, Chapter 1].

Let C3 (; E3 ) := C 3 (; E3 )/R denote the quotient set of C 3 (; E3 ) by the
e R means that there exist a vector
equivalence relation R, where (, )
3
3
e
c E and a matrix Q O such that (x) = c + Q(x)
for all x . Then
3
3
it is easily verified that the set C (; E ) becomes a metric space when it is

equipped with the distance d3 defined by
= inf d3 (, ) = inf d3 (, c + Q),
)
d3 (,
3
cE
QO3
denotes the equivalence class of modulo R.

where
We now show that the announced continuity of an immersion as a function
of its metric tensor is a corollary to Theorem 1.8-1. If d is a metric defined on
a set X, the associated metric space is denoted {X; d}.

Let
C02 (; S3> ) := {(gij ) C 2 (; S3> ); Rqijk = 0 in },
and, given any matrix field C = (gij ) C02 (; S3> ), let F(C) C3 (; E3 ) denote
the equivalence class modulo R of any C 3 (; E3 ) that satisfies T =

C in . Then the mapping
F : {C02 (; S3> ); d2 } {C 3 (; E3 ); d3 }
defined in this fashion is continuous.
Sect. 1.8]
49
Proof. Since {C02 (; S3> ); d2 } and {C3 (; E3 ); d3 } are both metric spaces, it
suffices to show that convergent sequences are mapped through F into convergent sequences.
Let then C C02 (; S3> ) and Cn C02 (; S3> ), n 0, be such that
lim d2 (Cn , C) = 0,
i.e., such that limn kCn Ck2,K = 0 for all K b . Given any
F(C), Theorem 1.8-3 shows that there exist n F(Cn ), n 0, such that
limn kn k3,K = 0 for all K b , i.e., such that
lim d3 (n , ) = 0.
Consequently,
lim d3 (F(Cn ), F(C)) = 0.

As shown by Ciarlet & C. Mardare [2004b], the above continuity result can
be extended up to the boundary of the set , as follows. If is bounded and
has a Lipschitz-continuous boundary, the mapping F of Theorem 1.8-5 can be
extended to a mapping that is locally Lipschitz-continuous with respect to the
topologies of the Banach spaces C 2 (; S3 ) for the continuous extensions of the
symmetric matrix fields C, and C 3 (; E3 ) for the continuous extensions of the
immersions (the existence of such continuous extensions is briefly commented
upon at the end of Section 1.6).
Another extension, motivated by three-dimensional nonlinear elasticity, is
the following: Let be a bounded and connected subset of R3 , and let B
be an elastic body with as its reference configuration. Thanks mostly to
the landmark existence theory of Ball [1977], it is now customary in nonlinear
three-dimensional elasticity to view any mapping H 1 (; E3 ) that is almosteverywhere injective and satisfies det > 0 a.e. in as a possible deformation
of B when B is subjected to ad hoc applied forces and boundary conditions. The
almost-everywhere injectivity of (understood in the sense of Ciarlet & Necas
[1987]) and the restriction on the sign of det mathematically express (in
an arguably weak way) the non-interpenetrability and orientation-preserving
conditions that any physically realistic deformation should satisfy.
As mentioned earlier, the Cauchy-Green tensor field T L1 (; S3 )
associated with a deformation H 1 (; E3 ) pervades the mathematical modeling of three-dimensional nonlinear elasticity. Conceivably, an alternative approach to the existence theory in three-dimensional elasticity could thus regard
the Cauchy-Green tensor as the primary unknown, instead of the deformation
itself as is usually the case.
Clearly, the Cauchy-Green tensors depend continuously on the deformations,
since the Cauchy-Schwarz inequality immediately shows that the mapping
H 1 (; E3 ) T L1 (; S3 )
50
[Ch. 1
is continuous (irrespectively of whether the mappings are almost-everywhere

injective and orientation-preserving).
Then Ciarlet & C. Mardare [2005a] have shown that, under appropriate
smoothness and orientation-preserving assumptions, the converse holds, i.e., the
deformations depend continuously on their Cauchy-Green tensors, the topologies
being those of the same spaces H 1 (; E3 ) and L1 (; S3 ) (by contrast with the
orientation-preserving condition, the issue of non-interpenetrability turns out to
be irrelevant to this issue). In fact, this continuity result holds in an arbitrary
dimension d, at no extra cost in its proof; so it will be stated below in this more
general setting. The notation Ed then denotes a d-dimensional Euclidean space
and Sd denotes the space of all symmetric matrices of order d.
This continuity result is itself a simple consequence of a nonlinear Korn
inequality, which constitutes the main result of ibid.: Let be a bounded
and connected open subset of Rd with a Lipschitz-continuous boundary and let
C 1 (; Ed ) be a mapping satisfying det > 0 in . Then there exists
a constant C() with the following property: For each orientation-preserving
mapping H 1 (; Ed ), there exist a d d rotation R = R(, ) (i.e., an
orthogonal matrix of order d with a determinant equal to one) and a vector
b = b(, ) in Ed such that
1/2
k (b + R)kH 1 (;Ed ) C()kT T kL1 (;Sd ) .

That a vector b and a rotation R should appear in the left-hand side of such
an inequality is of course reminiscent of the classical rigidity theorem (Theorem
e C 1 (; Ed ) and C 1 (; Ed )
1.7-1), which asserts that, if two mappings
e
satisfying det > 0 and det > 0 in an open connected subset of
Rd have the same Cauchy-Green tensor field, then the two mappings are isometrically equivalent, i.e., there exist a vector b in Ed and a d d orthogonal
e
matrix R (a rotation in this case) such that (x)
= b + R(x) for all x .
More generally, we shall say that two orientation-preserving mappings
e H 1 (; Ed ) and H 1 (; Ed ) are isometrically equivalent if there ex
ist a vector b in Ed and a d d orthogonal matrix R (again a rotation in this
case) such that
e
(x)
= b + R(x) for almost all x .
One application of the above key inequality is the following sequential continuity property: Let k H 1 (; Ed ), k 1, and C 1 (; Ed ) be orientationpreserving mappings. Then there exist a constant C() and orientation-prese k H 1 (; Ed ), k 1, that are isometrically equivalent to
erving mappings
k
such that
k
1/2
e kH 1 (;Ed ) C()k(k )T k T k 1
k
.
L (;Sd )
e k ) converges to in H 1 (; Ed ) as k if the
Hence the sequence (
k=1
T
1
d
sequence ((k )T k )
k=1 converges to in L (; S ) as k .
Sect. 1.8]
51
Should the Cauchy-Green strain tensor be viewed as the primary unknown

(as suggested above), such a sequential continuity could thus prove to be useful when considering infimizing sequences of the total energy, in particular for
handling the part of the energy that takes into account the applied forces and
the boundary conditions, which are both naturally expressed in terms of the
deformation itself.
They key inequality is first established in the special case where is the
identity mapping of the set , by making use in particular of a fundamental
geometric rigidity lemma recently proved by Friesecke, James & M
uller [2002].
It is then extended to an arbitrary mapping C 1 (; Rn ) satisfying det >
0 in , thanks in particular to a methodology that bears some similarity with
that used in Theorems 1.8-2 and 1.8-3.
Such results are to be compared with the earlier, pioneering estimates of
John [1961], John [1972] and Kohn [1982], which implied continuity at rigid body
deformations, i.e., at a mapping that is isometrically equivalent to the identity
mapping of . The recent and noteworthy continuity result of Reshetnyak [2003]
for quasi-isometric mappings is in a sense complementary to the above one (it
also deals with Sobolev type norms).
Chapter 2
DIFFERENTIAL GEOMETRY OF SURFACES
2.1
CURVILINEAR COORDINATES ON A SURFACE
In addition to the rules governing Latin indices that we set in Section 1.1, we
henceforth require that Greek indices and exponents vary in the set {1, 2} and
that the summation convention be systematically used in conjunction with these
rules. For instance, the relation
(i ai ) = (| b 3 )a + (3| + b )a3
means that, for = 1, 2,
3
X
i=1
2
2
X

X
i ai =
(| b 3 )a + 3| +
b a3 .
=1
=1
Kroneckers symbols are designated by , , or according to the context.

Let there be given as in Section 1.1 a three-dimensional Euclidean space E 3 ,
equipped with an orthonormal basis consisting of three vectors b
ei = b
ei , and let
a b, |a|, and a b denote the Euclidean inner product, the Euclidean norm,
and the vector product of vectors a, b in the space E3 .
In addition, let there be given a two-dimensional vector space, in which two
vectors e = e form a basis. This space will be identified with R2 . Let y
denote the coordinates of a point y R2 and let := /y and :=
2 /y y .
Finally, let there be given an open subset of R2 and a smooth enough
mapping : E3 (specific smoothness assumptions on will be made later,
according to each context). The set
is called a surface in E3 .
b := ()
53
54
[Ch. 2
Differential geometry of surfaces
If the mapping : E3 is injective, each point yb

b can be unambiguously written as
yb = (y), y ,
and the two coordinates y of y are called the curvilinear coordinates of yb

(Figure 2.1-1). Well-known examples of surfaces and of curvilinear coordinates
and their corresponding coordinate lines (defined in Section 2.2) are given in
Figures 2.1-2 and 2.1-3.
a2 (y)
(y)
a2 (y)
a1 (y)
a1 (y)
=()
E3
y2
R2
y1
Figure 2.1-1: Curvilinear coordinates on a surface and covariant and contravariant bases of
the tangent plane. Let
b = () be a surface in E3 . The two coordinates y1 , y2 of y are
the curvilinear coordinates of yb = (y)
b . If the two vectors a (y) = (y) are linearly
independent, they are tangent to the coordinate lines passing through yb and they form the
covariant basis of the tangent plane to
b at yb = (y). The two vectors a (y) from this tangent
plane defined by a (y) a (y) = form its contravariant basis.
Naturally, once a surface

b is defined as
b = (), there are infinitely many
other ways of defining curvilinear coordinates on
b , depending on how the domain and the mapping are chosen. For instance, a portion
b of a sphere
may be represented by means of Cartesian coordinates, spherical coordinates, or
stereographic coordinates (Figure 2.1-3). Incidentally, this example illustrates
the variety of restrictions that have to be imposed on
b according to which kind
of curvilinear coordinates it is equipped with!
Sect. 2.1]
55
Curvilinear coordinates on a surface
y
x
x
u
Figure 2.1-2: Several systems of curvilinear coordinates on a sphere. Let be a sphere
of radius R. A portion of contained in the northern hemisphere can be represented by
means of Cartesian coordinates, with a mapping of the form:
: (x, y) (x, y, {R2 (x2 + y 2 )}1/2 ) E3 .
A portion of that excludes a neighborhood of both poles and of a meridian (to fix
ideas) can be represented by means of spherical coordinates, with a mapping of the form:
: (, ) (R cos cos , R cos sin , R sin ) E3 .
A portion of that excludes a neighborhood of the North pole can be represented by
means of stereographic coordinates, with a mapping of the form:
2R2 u
2R2 v
u2 + v 2 R 2
: (u, v)
,
,
R
E3 .
u2 + v 2 + R 2 u2 + v 2 + R 2
u2 + v 2 + R 2
The corresponding coordinate lines are represented in each case, with self-explanatory
graphical conventions.
1
56
[Ch. 2
Figure 2.1-3: Two familiar examples of surfaces and curvilinear coordinates. A portion
b
of a circular cylinder of radius R can be represented by a mapping of the form
: (, z) (R cos , R sin , z) E3 .
A portion
b of a torus can be represented by a mapping of the form
: (, ) ((R + r cos ) cos , (R + r cos ) sin , r sin ) E3 ,
with R > r.
The corresponding coordinate lines are represented in each case, with self-explanatory
graphical conventions.
Sect. 2.2]
2.2
First fundamental form
57
FIRST FUNDAMENTAL FORM
Let be an open subset of R2 and let

= i b
ei : R2 () =
b E3
be a mapping that is differentiable at a point y . If y is such that (y +y)

, then
(y + y) = (y) + (y)y + o(y),
where the 3 2 matrix (y) is defined by
1 1 2 1
(y) := 1 2 2 2 (y).
1 3 2 3
Let the two vectors a (y) R3 be defined by
1
a (y) := (y) = 2 (y),
3
i.e., a (y) is the -th column vector of the matrix (y) and let y = y e .
Then the expansion of about y may be also written as
(y + y) = (y) + y a (y) + o(y).
If in particular y is of the form y = te , where t R and e is one of
the basis vectors in R2 , this relation reduces to
(y + te ) = (y) + ta (y) + o(t).
A mapping : E3 is an immersion at y if it is differentiable at
y and the 3 2 matrix (y) is of rank two, or equivalently if the two vectors
a (y) = (y) are linearly independent.
Assume from now on in this section that the mapping is an immersion
at y. In this case, the last relation shows that each vector a (y) is tangent
to the -th coordinate line passing through yb = (y), defined as the image
by of the points of that lie on a line parallel to e passing through y
(there exist t0 and t1 with t0 < 0 < t1 such that the -th coordinate line is
given by t ]t0 , t1 [ f (t) := (y + te ) in a neighborhood of yb; hence
f 0 (0) = (y) = a (y)); see Figures 2.1-1, 2.1-2, and 2.1-3.
The vectors a (y), which thus span the tangent plane to the surface
b at
yb = (y), form the covariant basis of the tangent plane to
b at yb; see
Figure 2.1-1.
Returning to a general increment y = y e , we also infer from the expansion of about y that
|(y + y) (y)|2 = y T (y)T (y)y + o(|y|2 )
= y a (y) a (y)y + o(|y|2 ).
58
[Ch. 2
In other words, the principal part with respect to y of the length between
the points (y + y) and (y) is {y a (y) a (y)y }1/2 . This observation
suggests to define a matrix (a (y)) of order two by letting

a (y) := a (y) a (y) = (y)T (y) .
The elements a (y) of this symmetric matrix are called the covariant
components of the first fundamental form, also called the metric tensor,
of the surface
b at yb = (y).
Note that the matrix (a (y)) is positive definite since the vectors a (y) are
assumed to be linearly independent.
The two vectors a (y) being thus defined, the four relations
a (y) a (y) =
unambiguously define two linearly independent vectors a (y) in the tangent
plane. To see this, let a priori a (y) = Y (y)a (y) in the relations a (y)
a (y) = . This gives Y (y)a (y) = ; hence Y (y) = a (y), where
(a (y)) := (a (y))1 .
Hence a (y) = a (y)a (y). These relations in turn imply that
a (y) a (y) = a (y)a (y)a (y) a (y)
= a (y)a (y)a (y) = a (y) = a (y),
and thus the vectors a (y) are linearly independent since the matrix (a (y))
is positive definite. We would likewise establish that a (y) = a (y)a (y).
The two vectors a (y) form the contravariant basis of the tangent plane
to the surface
b at yb = (y) (Figure 2.1-1) and the elements a (y) of the
symmetric matrix (a (y)) are called the contravariant components of the
first fundamental form, or metric tensor, of the surface
b at yb = (y).
Let us record for convenience the fundamental relations that exist between
the vectors of the covariant and contravariant bases of the tangent plane and
the covariant and contravariant components of the first fundamental tensor:
a (y) = a (y) a (y)
a (y) = a (y)a (y)
and a (y) = a (y) a (y),

and a (y) = a (y)a (y).
A word of caution. The presentation in this section closely follows that

of Section 1.2, the mapping : R2 E3 replacing the mapping :
R3 E3 . There are indeed strong similarities between the two presentations,
such as the way the metric tensor is defined in both cases, but there are also
sharp differences. In particular, the matrix (y) is not a square matrix, while
the matrix (x) is square!
Sect. 2.3]
2.3
59
Areas and lengths on a surface
AREAS AND LENGTHS ON A SURFACE
We now review fundamental formulas expressing area and length elements at

a point yb = (y) of the surface
b = () in terms of the matrix (a (y)); see
Figure 2.3-1.
These formulas highlight in particular the crucial r
ole played by the matrix (a (y)) for computing metric notions at yb = (y). Indeed, the first
fundamental form well deserves metric tensor as its alias!
A mapping : E3 is an immersion if it is an immersion at each
y , i.e., if is differentiable in and the two vectors (y) are linearly
independent at each y .
dS(
y)
(y+y) A
y)
dl(
(y) = y
E3
y y+y
dy
A
R2
C
t
b y ) at
Figure 2.3-1: Area and length elements on a surface. The elements db
a(b
y ) and d`(b
yb = (y)
b are related to dy and y by means of the covariant components of the metric
tensor of the surface
b ; cf. Theorem 2.3-1. The corresponding relations are used for computing
b = (A)
b = (C)
the area of a surface A
b and the length of a curve C
b , where C = f (I)
and I is a compact interval of R.
Theorem 2.3-1. Let be an open subset of R2 , let : E3 be an injective

and smooth enough immersion, and let
b = ().
(a) The area element db
a(b
y ) at yb = (y)
b is given in terms of the area
element dy at y by
db
a(b
y) =
a(y) dy, where a(y) := det(a (y)).
60
[Ch. 2
b y ) at yb = (y)
(b) The length element d`(b
b is given by

b y ) = y a (y)y 1/2 .
d`(b
Proof. The relation (a) between the area elements is well known. It can
b x) and
also be deduced directly from the relation between the area elements d(b
d(x) given in Theorem 1.3-1 (b) by means of an ad hoc three-dimensional
extension of the mapping .
b y ) is by definition
The expression of the length element in (b) recalls that d`(b
the principal part with respect to y = y e of the length |(y + y) (y)|,
whose expression precisely led to the introduction of the matrix (a (y)).

The relations found in Theorem 2.3-1 are used for computing surface integrals and lengths on the surface
b by means of integrals inside , i.e., in terms
of the curvilinear coordinates used for defining the surface
b (see again Figure
2.3-1).
Let A be a domain in R2 such that A (a domain in R2 is a bounded,
open, and connected subset of R2 with a Lipschitz-continuous boundary; cf.
b := (A), and let fb L1 (A)
b be given. Then
Section 1.3), let A
Z
Z
p
fb(b
y ) db
a(b
y ) = (fb )(y) a(y) dy.
b
A
b is given by
In particular, the area of A
Z
Z p
b :=
area A
db
a(b
y) =
a(y) dy.
b
A
Consider next a curve C = f (I) in , where I is a compact interval of R

and f = f e : I is a smooth enough injective mapping. Then the length
b := (C)
of the curve C
b is given by
s
Z
Z

df df
d
b := ( f )(t) dt =
length C
a (f (t))
(t)
(t) dt.
dt
dt
I dt
I
The last relation shows in particular that the lengths of curves inside the
surface () are precisely those induced by the Euclidean metric of the space E3 .
For this reason, the surface () is said to be isometrically imbedded in E3 .
2.4
SECOND FUNDAMENTAL FORM; CURVATURE

ON A SURFACE
While the image () E3 of a three-dimensional open set R3 by a smooth

enough immersion : R3 E3 is well defined by its metric, uniquely
up to isometries in E3 (provided ad hoc compatibility conditions are satisfied by
the covariant components gij : R of its metric tensor ; cf. Theorems 1.6-1
Sect. 2.4]
Second fundamental form; curvature on a surface
61
and 1.7-1), a surface given as the image () E3 of a two-dimensional open

set R2 by a smooth enough immersion : R2 E3 cannot be defined
by its metric alone.
As intuitively suggested by Figure 2.4-1, the missing information is provided
by the curvature of a surface. A natural way to give substance to this otherwise vague notion consists in specifying how the curvature of a curve on a
surface can be computed. As shown in this section, solving this question relies
on the knowledge of the second fundamental form of a surface, which naturally
appears for this purpose through its covariant components (Theorem 2.4-1).
2
Figure 2.4-1: A metric alone does not define a surface in E3 . A flat surface
b0 may be
deformed into a portion
b1 of a cylinder or a portion
b 2 of a cone without altering the length
of any curve drawn on it (cylinders and cones are instances of developable surfaces; cf.
Section 2.5). Yet it should be clear that in general
b 0 and
b1 , or
b0 and
b2 , or
b1 and
b2 ,
are not identical surfaces modulo an isometry of E3 !
Consider as in Section 2.1 a surface

b = () in E3 , where is an open
2
2
3
subset of R and : R E is a smooth enough immersion. For each
y , the vector
a1 (y) a2 (y)
a3 (y) :=
|a1 (y) a2 (y)|
is thus well defined, has Euclidean norm one, and is normal to the surface
b at
the point yb = (y).
62
[Ch. 2
Remark. The denominator in the definition of a3 (y) may be also written as

|a1 (y) a2 (y)| =
where a(y) := det(a (y)).
a(y),
Fix y and consider a plane P normal to

b at yb = (y), i.e., a plane that
b = P
contains the vector a3 (y). The intersection C
b is thus a planar curve
on
b.
b at
As shown in Theorem 2.4-1, it is remarkable that the curvature of C
yb can be computed by means of the covariant components a (y) of the first
fundamental form of the surface
b = () introduced in Section 2.2, together
with the covariant components b (y) of the second fundamental form of
b.
The definition of the curvature of a planar curve is recalled in Figure 2.4-2.
b at yb is 6= 0, it can be written as 1 , and R is
If the algebraic curvature of C
R
b at yb. This means
then called the algebraic radius of curvature of the curve C
b at yb is the point (b
that the center of curvature of the curve C
y + Ra3 (y));
see Figure 2.4-3. While R is not intrinsically defined, as its sign changes in any
system of curvilinear coordinates where the normal vector a3 (y) is replaced by
its opposite, the center of curvature is intrinsically defined.
b at yb is 0, the radius of curvature of the curve C
b at yb
If the curvature of C
is said to be infinite; for this reason, it is customary to still write the curvature
1
as
in this case.
R
1
Note that the real number
is always well defined by the formula given in
R
the next theorem, since the symmetric matrix (a (y)) is positive definite. This
implies in particular that the radius of curvature never vanishes along a curve
on a surface () defined by a mapping satisfying the assumptions of the next
theorem, hence in particular of class C 2 on .
It is intuitively clear that if R = 0, the mapping cannot be too smooth.
Think of a surface made of two portions of planes intersecting along a segment,
which thus constitutes a fold on the surface. Or think of a surface () with
0 and (y1 , y2 ) = |y1 |1+ for some 0 < < 1, so that C 1 (; E3 ) but
/ C 2 (; E3 ): The radius of curvature of a curve corresponding to a constant

y2 vanishes at y1 = 0.
Theorem 2.4-1. Let be an open subset of R2 , let C 2 (; E3 ) be an injective
immersion, and let y be fixed.
Consider a plane P normal to
b = () at the point yb = (y). The inb on
b = (C) of a curve
tersection P
b is a curve C
b , which is the image C
C in the set . Assume that, in a sufficiently small neighborhood of y, the restriction of C to this neighborhood is the image f (I) of an open interval I R,
where f = f e : I R is a smooth enough injective mapping that satisfies
df
(t) e 6= 0, where t I is such that y = f (t) (Figure 2.4-3).
dt
Sect. 2.4]
63
(s)
p(s) + R(s)
(s + s)
(s + s)
(s)
p(s + s)
(s)
p(s)
Figure 2.4-2: Curvature of a planar curve. Let be a smooth enough planar curve,
parametrized by its curvilinear abscissa s. Consider two points p(s) and p(s + s) with
curvilinear abscissae s and s + s and let (s) be the algebraic angle between the two
normals (s) and (s + s) (oriented in the usual way) to at those points. When s 0,
(s)
has a limit, called the curvature of at p(s). If this limit is non-zero, its
the ratio
s
inverse R is called the algebraic radius of curvature of at p(s) (the sign of R depends on
the orientation chosen on ).
The point p(s) + R(s), which is intrinsically defined, is called the center of curvature
of at p(s): It is the center of the osculating circle at p(s), i.e., the limit as s 0 of the
circle tangent to at p(s) that passes through the point p(s + s). The center of curvature is
also the limit as s 0 of the intersection of the normals (s) and (s + s). Consequently,
the centers of curvature of lie on a curve (dashed on the figure), called la developpee in
French, that is tangent to the normals to .
Then the curvature
1
b at yb is given by the ratio
of the planar curve C
R
df df
(t)
(t)
1
dt
dt
=
,
R
df
df
a (f (t))
(t)
(t)
dt
dt
b (f (t))
where a (y) are the covariant components of the first fundamental form of
b
at y (Section 2.1) and
b (y) := a3 (y) a (y) = a3 (y) a (y) = b (y).
64
[Ch. 2
a3(y)
y =(y)
y + Ra3(y)
E3
e2
e1
R2
y = f (t)e
df
(t)e
dt
f = f e
t
Figure 2.4-3: Curvature on a surface.

Let P be a plane containing the vector
a1 (y) a2 (y)
a3 (y) =
, which is normal to the surface
b = (). The algebraic curvature
|a1 (y) a2 (y)|
1
b =P
of the planar curve C
b = (C) at yb = (y) is given by the ratio
R
df df
b (f (t))
(t)
(t)
1
dt
dt
=
,
R
df df
a (f (t))
(t)
(t)
dt
dt
where a (y) and b (y) are the covariant components of the first and second fundadf
mental forms of the surface
b at yb and
(t) are the components of the vector tangent to
dt
1
b
the curve C = f (I) at y = f (t) = f (t)e . If
6= 0, the center of curvature of the curve C
R
at yb is the point (b
y + Ra3 (y)), which is intrinsically defined in the Euclidean space E3 .
1
of
Proof. (i) We first establish a well-known formula giving the curvature
R
a planar curve. Using the notations of Figure 2.4-2, we note that
sin (s) = (s) (s + s) = {(s + s) (s)} (s + s),
so that
1
(s)
sin (s)
d
:= lim
= lim
= (s) (s).
s0 s
s0
R
s
ds
Sect. 2.5]
65
(ii) The curve ( f )(I), which is a priori parametrized by t I, can be

also parametrized by its curvilinear abscissa s in a neighborhood of the point yb.
There thus exist an interval I I and a mapping p : J P , where J R is
an interval, such that
s J.
( f )(t) = p(s) and (a3 f )(t) = (s) for all t I,
By (i), the curvature
1
b is given by
of C
R
1
d
= (s) (s),
R
ds
where
d(a3 f ) dt
df dt
d
(s) =
(t)
= a3 (f (t))
(t) ,
ds
dt
ds
dt
ds
dp
d( f ) dt
(s) =
(s) =
(t)
ds
dt
ds
df dt
df dt
= (f (t))
(t)
= a (f (t))
(t) .
dt
ds
dt
ds
Hence
df df
1
= a3 (f (t)) a (f (t))
(t)
(t)
R
dt
dt
To obtain the announced expression for
dt
ds
2
1
, it suffices to note that
R
a3 (f (t)) a (f (t)) = b (f (t)),

by definition of the functions b and that (Theorem 2.3-1 (b))

1/2 n
df df o1/2
ds = y a (y)y
(t)
(t)
dt.
= a (f (t))
dt
dt
The knowledge of the curvatures of curves contained in planes normal to

b
suffices for computing the curvature of any curve on
b . More specifically, the
e at yb of any smooth enough curve C
e (planar or not) on the
radius of curvature R
cos
1
surface
b is given by
= , where is the angle between the principal
e
R
R
1
e
normal to C at yb and a3 (y) and
is given in Theorem 2.4-1; see, e.g., Stoker
R
[1969, Chapter 4, Section 12].
The elements b (y) of the symmetric matrix (b (y)) defined in Theorem
2.4-1 are called the covariant components of the second fundamental form
of the surface
b = () at yb = (y).
66
2.5
[Ch. 2
PRINCIPAL CURVATURES; GAUSSIAN CURVATURE
The analysis of the previous section suggests that precise information about the
shape of a surface
b = () in a neighborhood of one of its points yb = (y)
can be gathered by letting the plane P turn around the normal vector a3 (y)
and by following in this process the variations of the curvatures at yb of the
corresponding planar curves P
b , as given in Theorem 2.4-1.
As a first step in this direction, we show that these curvatures span a compact
interval of R. In particular then, they stay away from infinity.
Note that this compact interval contains 0 if, and only if, the radius of
curvature of the curve P
b is infinite for at least one such plane P .
Theorem 2.5-1. (a) Let the assumptions and notations be as in Theorem 2.41. For a fixed y , consider the set P of all planes P normal to the surface
b = () at yb = (y). Then the set of curvatures of the associated planar

h 1
1 i
curves P
b , P P, is a compact interval of R, denoted
.
,
R1 (y) R2 (y)
(b) Let the matrix (b (y)), being the row index, be defined by
b (y) := a (y)b (y),
where (a (y)) = (a (y))1 (Section 2.2) and the matrix (b (y)) is defined as
in Theorem 2.4-1. Then
1
1
+
= b11 (y) + b22 (y),
R1 (y) R2 (y)
1
det(b (y))
= b11 (y)b22 (y) b21 (y)b12 (y) =
.
R1 (y)R2 (y)
det(a (y))
1
1
6=
, there is a unique pair of orthogonal planes P1 P
R1 (y)
R2 (y)
and P2 P such that the curvatures of the associated planar curves P1
b and
1
1
P2
b are precisely
and
.
R1 (y)
R2 (y)
(c) If
Proof. (i) Let (P ) denote the intersection of P P with the tangent plane
b ) denote the intersection of P with
T to the surface
b at yb, and let C(P
b . Hence
b ) at yb
(P ) is tangent to C(P
b.
b )
In a sufficiently small neighborhood of yb the restriction of the curve C(P
b
to this neighborhood is given by C(P ) = ( f (P ))(I(P )), where I(P ) R
is an open interval and f (P ) = f (P )e : I(P ) R2 is a smooth enough
df (P )
injective mapping that satisfies
(t)e 6= 0, where t I(P ) is such that
dt
y = f (P )(t). Hence the line (P ) is given by

d( f (P ))
(P ) = yb +
(t); R = {b
y + a (y); R} ,
dt
Sect. 2.5]
Principal curvatures; Gaussian curvature
67
df (P )
(t) and e 6= 0 by assumption.
dt
b ))
Since the line {y+ e ; R} is tangent to the curve C(P ) := 1 (C(P
3
at y (the mapping : R is injective by assumption) for each
such parametrizing function f (P ) : I(P ) R2 and since the vectors a (y)
are linearly independent, there exists a bijection between the set of all lines
(P ) T , P P, and the set of all lines supporting the nonzero tangent
vectors to the curve C(P ).
Hence Theorem 2.4-1 shows that when P varies in P, the curvature of the
b = C(P
b ) at yb takes the same values as does the ratio
corresponding curves C

b (y)
when := ( ) varies in R2 {0}.
a (y)
where :=
(ii) Let the symmetric matrices A and B of order two be defined by

A := (a (y)) and B := (b (y)).
Since A is positive definite, it has a (unique) square root C, i.e., a symmetric
positive definite matrix C such that A = C2 . Hence the ratio
T C1 BC1
b (y)
T B
=
= T
, where = C,
a (y)
T
A
is nothing but the Rayleigh quotient associated with the symmetric matrix
C1 BC1 . When varies in R2 {0}, this Rayleigh quotient thus spans the
compact interval of R whose end-points are the smallest and largest eigenvalue,
1
1
respectively denoted
and
, of the matrix C1 BC1 (for a proof,
R1 (y)
R2 (y)
see, e.g., Ciarlet [1982, Theorem 1.3-1]). This proves (a).
Furthermore, the relation
C(C1 BC1 )C1 = BC2 = BA1
shows that the eigenvalues of the symmetric matrix C1 BC1 coincide with
those of the (in general non-symmetric) matrix BA1 . Note that BA1 =
(b (y)) with b (y) := a (y)b (y), being the row index, since A1 =
(a (y)).
Hence the relations in (b) simply express that the sum and the product of
the eigenvalues of the matrix BA1 are respectively equal to its trace and to its
det(b (y))
determinant, which may be also written as
since BA1 = (b (y)).
det(a (y))
This proves (b).
(iii) Let 1 = (1 ) = C 1 and 2 = (2 ) = C 2 , with 1 = (1 ) and
2 = (2 ), be two orthogonal ( T1 2 = 0) eigenvectors of the symmetric matrix
1
1
and
, respectively.
C1 BC1 , corresponding to the eigenvalues
R1 (y)
R2 (y)
Hence
0 = T1 2 = T1 CT C 2 = T1 A2 = 0,
68
[Ch. 2
since CT = C. By (i), the corresponding lines (P1 ) and (P2 ) of the tangent
plane are parallel to the vectors 1 a (y) and 2 a (y), which are orthogonal
since

1 a (y) 2 a (y) = a (y)1 2 = T1 A 2 .
1
1
6=
, the directions of the vectors 1 and 2 are uniquely
R1 (y)
R2 (y)
determined and the lines (P1 ) and (P2 ) are likewise uniquely determined.
This proves (c).

If
We are now in a position to state several fundamental definitions:

The elements b (y) of the (in general non-symmetric) matrix (b (y)) defined
in Theorem 2.5-1 are called the mixed components of the second fundamental form of the surface
b = () at yb = (y).
1
1
The real numbers
and
(one or both possibly equal to 0) found
R1 (y)
R2 (y)
in Theorem 2.5-1 are called the principal curvatures of
b at yb.
1
1
If
=
, the curvatures of the planar curves P
b are the same in
R1 (y)
R2 (y)
1
1
all directions, i.e., for all P P. If
=
= 0, the point yb = (y) is
R1 (y)
R2 (y)
1
1
called a planar point. If
=
6= 0, yb is called an umbilical point.
R1 (y)
R2 (y)
It is remarkable that, if all the points of
b are planar, then
b is a portion
of a plane. Likewise, if all the points of
b are umbilical, then
b is a portion of
a sphere. For proofs, see, e.g., Stoker [1969, p. 87 and p. 99].
Let yb = (y)
b be a point that is neither planar nor umbilical; in other
words, the principal curvatures at yb are not equal. Then the two orthogonal
lines tangent to the planar curves P1
b and P2
b (Theorem 2.5-1 (c)) are
called the principal directions at yb.
A line of curvature is a curve on
b that is tangent to a principal direction
at each one of its points. It can be shown that a point that is neither planar
nor umbilical possesses a neighborhood where two orthogonal families of lines
of curvature can be chosen as coordinate lines. See, e.g., Klingenberg [1973,
Lemma 3.6.6].
1
1
If
6= 0 and
6= 0, the real numbers R1 (y) and R2 (y) are called
R1 (y)
R2 (y)
1
the algebraic principal radii of curvature of
b at yb. If, e.g.,
= 0,
R1 (y)
the corresponding radius of curvature R1 (y) is said to be infinite, according
to the convention made in Section 2.4. While the algebraic principal radii of
curvature may simultaneously change their signs in another system of curvilinear
coordinates, the associated centers of curvature are intrinsically defined.
1
1
1
The numbers
+
and
, which are the principal
R1 (y)
R2 (y)
R1 (y)R2 (y)
invariants of the matrix (b (y)) (Theorem 2.5-1), are respectively called the
mean curvature and the Gaussian, or total, curvature of the surface
b
at yb.
Sect. 2.5]
Principal curvatures; Gaussian curvature
69
Figure 2.5-1: Different kinds of points on a surface. A point is elliptic if the Gaussian
curvature is > 0 or equivalently, if the two principal radii of curvature are of the same sign;
the surface is then locally on one side of its tangent plane. A point is parabolic if exactly one
of the two principal radii of curvature is infinite; the surface is again locally on one side of its
tangent plane. A point is hyperbolic if the Gaussian curvature is < 0 or equivalently, if the
two principal radii of curvature are of different signs; the surface then intersects its tangent
plane along two curves.
70
[Ch. 2
A point on a surface is elliptic, parabolic, or hyperbolic, according as its

Gaussian curvature is > 0, = 0 but it is not a planar point, or < 0; see Figure
2.5-1.
An asymptotic line is a curve on a surface that is everywhere tangent to
a direction along which the radius of curvature is infinite; any point along an
asymptotic line is thus either parabolic or hyperbolic. It can be shown that,
if all the points of a surface are hyperbolic, any point possesses a neighborhood
where two intersecting families of asymptotic lines can be chosen as coordinate
lines. See, e.g., Klingenberg [1973, Lemma 3.6.12].
As intuitively suggested by Figure 2.4-1, a surface in R3 cannot be defined
by its metric alone, i.e., through its first fundamental form alone, since its
curvature must be in addition specified through its second fundamental form.
But quite surprisingly, the Gaussian curvature at a point can also be expressed
solely in terms of the functions a and their derivatives! This is the celebrated
Theorema egregium (astonishing theorem) of Gau [1828]; see Theorem 2.6.2
in the next section.
Another striking result involving the Gaussian curvature is the equally celebrated Gau-Bonnet theorem, so named after Gau [1828] and Bonnet
[1848] (for a modern proof, see, e.g., Klingenberg [1973, Theorem 6.3-5] or
do Carmo [1994, Chapter 6, Theorem 1]): Let S be a smooth enough, closed,
orientable, and compact surface in R3 (a closed surface is one without
boundary, such as a sphere or a torus; orientable surfaces, which exclude for
instance Klein bottles, are defined in, e.g., Klingenberg [1973, Section 5.5]) and
let K : S R denote its Gaussian curvature. Then
Z
K(b
y ) db
a(b
y ) = 2(2 2g(S)),
S
where the genus g(S) is the number of holes of S (for instance, a sphere
has genus zero, while a torus has genus one). The integer (S) defined by
(S) := (2 2g(S)) is the Euler characteristic of
b.
According to the definition of Stoker [1969, Chapter 5, Section 2], a developable surface is one whose Gaussian curvature vanishes everywhere. Developable surfaces are otherwise often defined as ruled surfaces whose Gaussian
curvature vanishes everywhere, as in, e.g., Klingenberg [1973, Section 3.7]). A
portion of a plane provides a first example, the only one of a developable surface
all points of which are planar. Any developable surface all points of which are
parabolic can be likewise fully described: It is either a portion of a cylinder,
or a portion of a cone, or a portion of a surface spanned by the tangents to a
skewed curve. The description of a developable surface comprising both planar
and parabolic points is more subtle (although the above examples are in a sense
the only ones possible, at least locally; see Stoker [1969, Chapter 5, Sections 2
to 6]).
The interest of developable surfaces is that they can be, at least locally,
continuously rolled out, or developed (hence their name), onto a plane,
without changing the metric of the intermediary surfaces in the process.
Sect. 2.6]
2.6
71
COVARIANT DERIVATIVES OF A VECTOR FIELD

AND CHRISTOFFEL SYMBOLS ON A SURFACE;
THE GAUSS AND WEINGARTEN FORMULAS
As in Sections 2.2 and 2.4, consider a surface

b = () in E3 , where :
2
3
R E is a smooth enough injective immersion, and let
a3 (y) = a3 (y) :=
a1 (y) a2 (y)
,
|a1 (y) a2 (y)|
y .
Then the vectors a (y) (which form the covariant basis of the tangent plane to
b at yb = (y); see Figure 2.1-1) together with the vector a3 (y) (which is normal
to
b and has Euclidean norm one) form the covariant basis at yb.
i (y)ai (y)
3 (y)
a3 (y)
a2 (y)
2 (y) a1 (y)
1 (y)
= ( )
E3
y2
R2
y1
Figure 2.6-1: Contravariant bases and vector fields along a surface. At each point yb =
(y)
b = (), the three vectors ai (y), where a (y) form the contravariant basis of the
a1 (y) a2 (y)
, form the contravariant
tangent plane to
b at yb (Figure 2.1-1) and a3 (y) =
|a1 (y) a2 (y)|
basis at yb. An arbitrary vector field defined on
b may then be defined by its covariant
components i : R. This means that i (y)a i (y) is the vector at the point yb.
Let the vectors a (y) of the tangent plane to

b at yb be defined by the relations a (y) a (y) = . Then the vectors a (y) (which form the contravariant
basis of the tangent plane at yb; see again Figure 2.1-1) together with the vector a3 (y) form the contravariant basis at yb; see Figure 2.6-1. Note that the
72
[Ch. 2
vectors of the covariant and contravariant bases at yb satisfy

ai (y) aj (y) = ji .
Suppose that a vector field is defined on the surface

b . One way to define
such a field in terms of the curvilinear coordinates used for defining the surface
b consists in writing it as i ai : R3 , i.e., in specifying its covariant

components i : R over the vectors ai of the contravariant bases. This
means that i (y)ai (y) is the vector at each point yb = (y)
b (Figure 2.6-1).
Our objective in this section is to compute the partial derivatives (i ai )
of such a vector field. These are found in the next theorem, as immediate consequences of two basic formulas, those of Gau and Weingarten. The Christoffel
symbols on a surface and the covariant derivatives on a surface are also naturally introduced in this process.
Theorem 2.6-1. Let be an open subset of R2 and let C 2 (; E3 ) be an
immersion.
(a) The derivatives of the vectors of the covariant and contravariant bases
are given by
a = C
a + b a3 and a = C
a + b a3 ,
3
a3 = a = b a = b a ,
where the covariant and mixed components b and b of the second fundamental
form of
b are defined in Theorems 2.4-1 and 2.5-1 and
C
:= a a .
(b) Let there be given a vector field i ai : R3 with covariant components

i C 1 (). Then i ai C 1 () and the partial derivatives (i ai ) C 0 () are
given by
(i ai ) = ( C
b 3 )a + ( 3 + b )a3
= (| b 3 )a + (3| + b )a3 ,
where
| := C
and 3| := 3 .
Proof. Since any vector c in the tangent plane can be expanded as c =

(c a )a = (c a )a , since a3 is in the tangent plane ( a3 a3 = 21 (a3
a3 ) = 0), and since a3 a = b (Theorem 2.4-1), it follows that
a3 = ( a3 a )a = b a .
This formula, together with the definition of the functions b (Theorem 2.51), implies in turn that
a3 = ( a3 a )a = b (a a )a = b a a = b a .
Sect. 2.6]
73
Any vector c can be expanded as c = (c ai )ai = (c aj )aj . In particular,
a = ( a a )a + ( a a3 )a3 = C
a + b a3 ,
by definition of C
and b . Finally,
a = ( a a )a + ( a a3 )a3 = C
a + b a3 ,
since
a a3 = a a3 = b a a = b .
That i ai C 1 () if i C 1 () is clear since ai C 1 () if C 2 (; E3 ).

The formulas established supra immediately lead to the announced expression
of (i ai ).

The relations (found in Theorem 2.6-1)
a = C
a + b a3 and a = C
a + b a3
and
a3 = a3 = b a = b a ,
respectively constitute the formulas of Gau and Weingarten. The functions
(also found in Theorem 2.6-1)
| = C
and 3| = 3
are the first-order covariant derivatives of the surface vector field i ai :

R3 , and the functions
C
:= a a = a a
are the Christoffel symbols of the first kind.
Remarks. (1) The Christoffel symbols C

can be also defined solely in
terms of the covariant components of the first fundamental form; see the proof
of Theorem 2.7-1
(2) The notation C

is preferred here instead of the customary notation
, so as to avoid confusion with the three-dimensional Christoffel symbols

pij introduced in Section 1.4.
The definition of the covariant derivatives | = of a vector

field defined on a surface () given in Theorem 2.6-1 is highly reminiscent of
the definition of the covariant derivatives vikj = j vi pij vp of a vector field
defined on an open set () given in Section 1.4. However, the former are
more subtle to apprehend than the latter. To see this, recall that the covariant
derivatives vikj = j vi pij vp may be also defined by the relations (Theorem
1.4-2)
vikj g j = j (vi g i ).
74
[Ch. 2
By contrast, even if only tangential vector fields a on the surface ()

are considered (i.e., vector fields i ai : R3 for which 3 = 0), their covariant
derivatives | = satisfy only the relations
| a = P { ( a )} ,
where P denotes the projection operator on the tangent plane in the direction
of the normal vector (i.e., P(ci ai ) := c a ), since
3
( a ) = | a + b
a
for such tangential fields by Theorem 2.6-1. The reason is that a surface has in
3
general a nonzero curvature, manifesting itself here by the extra term b
a .
This term vanishes in if

b is a portion of a plane, since in this case b = b =
0. Note that, again in this case, the formula giving the partial derivatives in
Theorem 2.9-1 (b) reduces to
(i ai ) = (i| )ai .
2.7
NECESSARY CONDITIONS SATISFIED BY THE

FIRST AND SECOND FUNDAMENTAL FORMS:
THE GAUSS AND CODAZZI-MAINARDI EQUATIONS; GAUSS THEOREMA EGREGIUM
It is remarkable that the components a : R and b : R of the first

and second fundamental forms of a surface (), defined by a smooth enough
immersion : E3 , cannot be arbitrary functions.
As shown in the next theorem, they must satisfy relations that take the
form:
C C + C
C C
C = b b b b in ,
b b + C
b C
b = 0 in ,
where the functions C and C

have simple expressions in terms of the
functions a and of some of their partial derivatives (as shown in the next
proof, it so happens that the functions C

as defined in Theorem 2.7-1 coincide
with the Christoffel symbols introduced in the previous section; this explains
why they are denoted by the same symbol).
These relations, which are meant to hold for all , , , {1, 2}, respectively constitute the Gau, and Codazzi-Mainardi, equations.
Theorem 2.7-1. Let be an open subset of R2 , let C 3 (; E3 ) be an immersion, and let

a := and b :=
n o
1
2
|1 2 |
Sect. 2.7] Necessary conditions satisfied by the fundamental forms
75
denote the covariant components of the first and second fundamental forms of
the surface (). Let the functions C C 1 () and C

C 1 () be defined
by
1
( a + a a ),
2
:= a C where (a ) := (a )1 .
C :=
Then, necessarily,
C C + C
C C
C = b b b b in ,
b b + C
b C
b = 0 in .
Proof. Let a = . It is then immediately verified that the functions

C are also given by
C = a a .
a1 a2
and, for each y , let the three vectors aj (y) be defined
|a1 a2 |
by the relations aj (y) ai (y) = ij . Since we also have a = a a and a3 = a3 ,
the last relations imply that C

= a a , hence that
Let a3 =
a = C
a + b a3 ,
since a = ( a a )a + ( a a3 )a3 . Differentiating the same relations

yields
C = a a + a a ,
so that the above relations together give
a a = C
a a + b a3 a = C
C + b b .
Consequently,
a a = C C
C b b .
Since a = a , we also have
a a = C C
C b b .
Hence the Gau equations immediately follow.

Since a3 = ( a3 a )a + ( a3 a3 )a3 and a3 a = b =
a a3 , we have
a3 = b a .
Differentiating the relations b = a a3 , we obtain
b = a a3 + a a3 .
76
[Ch. 2
This relation and the relations a = C

a + b a3 and a3 = b a
together imply that
a a3 = C
b .
Consequently,
a a3 = b + C
b .
Since a = a , we also have
a a3 = b + C
b .
Hence the Codazzi-Mainardi equations immediately follow.
Remark. The vectors a and a introduced above respectively form the

covariant and contravariant bases of the tangent plane to the surface (), the
unit vector a3 = a3 is normal to the surface, and the functions a are the
contravariant components of the first fundamental form (Sections 2.2 and 2.3).

As shown in the above proof, the Gau and Codazzi-Mainardi equations
thus simply constitute a re-writing of the relations a = a in the form
of the equivalent relations a a = a a and a a3 = a a3 .
The functions
C =
1
( a + a a ) = a a = C
2
and
C
= a C = a a = C
are the Christoffel symbols of the first, and second, kind. We recall that
the same Christoffel symbols C

also naturally appeared in a different context
(that of covariant differentiation; cf. Section 2.6).
Finally, the functions
S := C C + C
C C
C
are the covariant components of the Riemann curvature tensor of the

surface ().
Remark. Like the notation C

vs. pij , the notation C is intended to
avoid confusions with the three-dimensional Christoffel symbols ijq introduced in Section 1.4.
Letting = 2, = 1, = 2, = 1 in the Gau equations gives in particular

S1212 = det(b ).
Consequently, the Gaussian curvature at each point (y) of the surface ()
can be written as
S1212 (y)
1
=
, y ,
R1 (y)R2 (y)
det(a (y))
Sect. 2.8] Existence of a surface with prescribed fundamental forms
77
1
det(b (y)
=
(Theorem 2.5-1). By inspection of the function
R1 (y)R2 (y)
det(a (y))
S1212 , we thus reach the astonishing conclusion that, at each point of the surface,
a notion involving the curvature of the surface, viz., the Gaussian curvature,
is entirely determined by the knowledge of the metric of the surface at the
same point, viz., the components of the first fundamental forms and their partial
derivatives of order 2 at the same point! This startling conclusion naturally
deserves a theorem:
since
Theorem 2.7-2. Let be an open subset of R2 , let C 3 (; E3 ) be an immersion, let a = denote the covariant components of the first fundamental form of the surface (), and let the functions C and S1212 be defined
by
1
( a + a a ),
2
1
:= (212 a12 11 a22 22 a11 ) + a (C12 C12 C11 C22 ).
2
C :=
S1212
Then, at each point (y) of the surface (), the principal curvatures
and R21(y) satisfy
1
S1212 (y)
=
, y .
R1 (y)R2 (y)
det(a (y))
1
R1 (y)
Theorem 2.7-2 constitutes the famed Theorema Egregium of Gau [1828],

so named by Gau who had been himself astounded by his discovery.
2.8
EXISTENCE OF A SURFACE WITH PRESCRIBED

FIRST AND SECOND FUNDAMENTAL FORMS
Let M2 , S2 , and S2> denote the sets of all square matrices of order two, of all
symmetric matrices of order two, and of all symmetric, positive definite matrices
of order two.
So far, we have considered that we are given an open set R2 and a
smooth enough immersion : E3 , thus allowing us to define the fields
(a ) : S2> and (b ) : S2 , where a : R and b : R
are the covariant components of the first and second fundamental forms of the
surface () E3 .
Note that the immersion need not be injective in order that these matrix
fields be well defined.
We now turn to the reciprocal questions:
Given an open subset of R2 and two smooth enough matrix fields (a ) :
S2> and (b ) : S2 , when are they the first and second fundamental
forms of a surface () E3 , i.e., when does there exist an immersion :
E3 such that
n o
1
2
a := and b :=
in ?
|1 2 |
78
[Ch. 2
If such an immersion exists, to what extent is it unique?

The answers to these questions turn out to be remarkably simple: If is
simply-connected, the necessary conditions of Theorem 2.7-1, i.e., the Gau
and Codazzi-Mainardi equations, are also sufficient for the existence of such an
immersion. If is connected, this immersion is unique up to isometries in E3 .
Whether an immersion found in this fashion is injective is a different issue,
which accordingly should be resolved by different means.
Following Ciarlet & Larsonneur [2001], we now give a self-contained, complete, and essentially elementary, proof of this well-known result. This proof
amounts to showing that it can be established as a simple corollary to the fundamental theorem of three-dimensional differential geometry (Theorems 1.6-1
and 1.7-1).
This proof has also the merit to shed light on the analogies (which cannot
remain unnoticed!) between the assumptions and conclusions of both existence
results (compare Theorems 1.6-1 and 2.8-1) and both uniqueness results (compare Theorems 1.7-1 and 2.9-1).
A direct proof of the fundamental theorem of surface theory is given in
Klingenberg [1973, Theorem 3.8.8], where the global existence of the mapping
is based on an existence theorem for ordinary differential equations, analogous
to that used in part (ii) of the proof of Theorem 1.6-1. A proof of the local
version of this theorem, which constitutes Bonnets theorem, is found in, e.g.,
do Carmo [1976].
This result is another special case of the fundamental theorem of Riemannian geometry alluded to in Section 1.6. We recall that this theorem asserts that
a simply-connected Riemannian manifold of dimension p can be isometrically
immersed into a Euclidean space of dimension (p + q) if and only if there exist
tensors satisfying together generalized Gau, and Codazzi-Mainardi, equations
and that the corresponding isometric immersions are unique up to isometries in
the Euclidean space. A substantial literature has been devoted to this theorem
and its various proofs, which usually rely on basic notions of Riemannian geometry, such as connections or normal bundles, and on the theory of differential
forms. See in particular the earlier papers of Janet [1926] and Cartan [1927]
and the more recent references of Szczarba [1970], Tenenblat [1971], Jacobowitz
[1982], and Szopos [2005].
Like the fundamental theorem of three-dimensional differential geometry,
this theorem comprises two essentially distinct parts, a global existence result
(Theorem 2.8-1) and a uniqueness result (Theorem 2.9-1), the latter being also
called rigidity theorem. Note that these two results are established under different assumptions on the set and on the smoothness of the fields (a ) and
(b ).
These existence and uniqueness results together constitute the fundamental theorem of surface theory.
Theorem 2.8-1. Let be a connected and simply-connected open subset of R 2
and let (a ) C 2 (; S2> ) and (b ) C 2 (; S2 ) be two matrix fields that satisfy
79
the Gau and Codazzi-Mainardi equations, viz.,
C C + C
C C
C = b b b b in ,
b b + C
b C
b = 0 in ,
where
1
( a + a a ),
2
:= a C where (a ) := (a )1 .
C :=
Then there exists an immersion C 3 (; E3 ) such that

n o
1
2
a = and b =
in .
|1 2 |
Proof. The proof of this theorem as a corollary to Theorem 1.6-1 relies
on the following elementary observation: Given a smooth enough immersion
: E3 and > 0, let the mapping : ], [ E3 be defined by
(y, x3 ) := (y) + x3 a3 (y) for all (y, x3 ) ], [ ,
where a3 :=
1 2
, and let
|1 2 |
gij := i j .
Then an immediate computation shows that

g = a 2x3 b + x23 c and gi3 = i3 in ], [ ,
where a and b are the covariant components of the first and second fundamental forms of the surface () and c := a b b .
Assume that the matrices (gij ) constructed in this fashion are invertible,
hence positive definite, over the set ], [ (they need not be, of course; but
the resulting difficulty is easily circumvented; see parts (i) and (viii) below).
Then the field (gij ) : ], [ S3> becomes a natural candidate for applying
the three-dimensional existence result of Theorem 1.6-1, provided of course
that the three-dimensional sufficient conditions of this theorem, viz.,
can be shown to hold, as consequences of the two-dimensional Gau and
Codazzi-Mainardi equations. That this is indeed the case is the essence of the
present proof (see parts (i) to (vii)).
By Theorem 1.6-1, there then exists an immersion : ], [ E3 that
satisfies gij = i j in ], [. It thus remains to check that := (, 0)
indeed satisfies (see part (ix))
n o
1
2
in .
a = and b =
|1 2 |
80
[Ch. 2
The actual implementation of this program essentially involves elementary,

but sometimes lengthy, computations, which accordingly will be omitted for the
most part; only the main intermediate results will be recorded.
For clarity, the proof is broken into nine parts, numbered (i) to (ix).
(i) Given two matrix fields (a ) C 2 (; S2> ) and (b ) C 2 (; S2 ), let the
matrix field (gij ) C 2 ( R; S3 ) be defined by
g := a 2x3 b + x23 c and gi3 := i3 in R
(the variable y is omitted; x3 designates the variable in R), where
c := b b and b := a b in .
Let 0 be an open subset of R2 such that 0 is a compact subset of . Then
there exists 0 = 0 (0 ) > 0 such that the symmetric matrices (gij ) are positive
definite at all points in 0 , where
0 := 0 ]0 , 0 [ .
Besides, the elements of the inverse matrix (g pq ) are given in 0 by
X
g =
(n + 1)xn3 a (Bn ) and g i3 = i3 ,
n0
where
(B) := b and (Bn ) := b1 bn1 for n 2,
i.e., (Bn ) designates for any n 0 the element at the -th row and -th
column of the matrix Bn . The above series are absolutely convergent in the
space C 2 (0 ).
P
Let a priori g = n0 xn3 h

n where hn are functions of y 0 only, so
that the relations g g = read
h
0 a + x3 (h1 a 2h0 b )
X
+
xn3 (h
n a 2hn1 b + hn2 c ) = .
n2
It is then easily verified that the functions h

n are given by
h
(Bn ) , n 0,
n = (n + 1)a
so that
g =
n0
(n + 1)xn3 a b1 bn1 .
It is clear that such a series is absolutely convergent in the space C 2 ( 0

[0 , 0 ]) if 0 > 0 is small enough.
81
(ii) The functions C

being defined by
C
:= a C ,
where
(a ) := (a )1 and C :=
1
( a + a a ),
2
define the functions

b | := b + C
b C
b ,
b| := b C
b C
b = b| .
Then
b | = a b| and b| = a b | .
Furthermore, the assumed Codazzi-Mainardi equations imply that
b | = b | and b| = b| .
The above relations follow from straightforward computations based on the
definitions of the functions b | and b| . They are recorded here because they
play a pervading r
ole in the subsequent computations.
(iii) The functions gij C 2 (0 ) and g ij C 2 (0 ) being defined as in part (i),
define the functions ijq C 1 (0 ) and pij C 1 (0 ) by
ijq :=
1
(j giq + i gjq q gij ) and pij := g pq ijq .
2
Then the functions ijq = jiq and pij = pji have the following expressions:
= C x3 (b | a + 2C
b ) + x23 (b | b + C
c ),
3 = 3 = b x3 c ,
33 = 33 = 33q = 0,
X
= C
xn+1
b | (Bn ) ,
3
n0
3 = b x3 c ,
X
3 =
xn3 (Bn+1 ) ,
n0
33
p33
= 0,
where the functions c , (Bn ) , and b | are defined as in parts (i) and (ii).
All computations are straightforward. We simply point out that the assumed
Codazzi-Mainardi equations are needed to conclude that the factor of x3 in the
function is indeed that announced above. We also note that the computation of the factor of x23 in relies in particular on the easily established
relations
c + C
c .
c = b | b + b | b + C
82
[Ch. 2
(iv) The functions ijq C 1 (0 ) and pij C 1 (0 ) being defined as in

part (iii), define the functions Rqijk C 0 (0 ) by
Rqijk := j ikq k ijq + pij kqp pik jqp .
Then, in order that the relations
Rqijk = 0 in 0
hold, it is sufficient that
R1212 = 0,
R23 = 0,
R33 = 0 in 0 .
The above definition of the functions Rqijk and that of the functions ijq
and pij (part (iii)) together imply that, for all i, j, k, q,
Rqijk = Rjkqi = Rqikj ,
Rqijk = 0 if j = k or q = i.
Consequently, the relation R1212 = 0 implies that R = 0, the relations
R23 = 0 imply that Rqijk = 0 if exactly one index is equal to 3, and finally,
the relations R33 = 0 imply that Rqijk = 0 if exactly two indices are equal
to 3.
(v) The functions
R33 := 33 3 3 + p3 3p p33 p
satisfy
R33 = 0 in 0 .
These relations immediately follow from the expressions found in part (iii)
for the functions ijq and pij . Note that neither the Gau equations nor the
Codazzi-Mainardi equations are needed here.
(vi) The functions
R23 := 23 3 2 + p2 3p p23 p
satisfy
R23 = 0 in 0 .
The definitions of the functions g (part (i)) and ijq (part (iii)) show that
23 3 2 = (2 b b2 ) + x3 ( c2 2 c ).
Then the expressions found in part (iii) show that
p2 3p p23 p = 3 2 23
b
= C
b2 C2
+ x3 (b2 | b b | b2 + C2
c C
c2 ),
83
and the relations R33 = 0 follow by making use of the relations
c = b | b + b | b + C
c + C
c
together with the relations
2 b b2 + C
b2 C2
b = 0,
which are special cases of the assumed Codazzi-Mainardi equations.

(vii) The function
R1212 := 1 221 2 211 + p21 21p p22 11p
satisfies
R1212 = 0 in 0 .
The computations leading to this relation are fairly lengthy and they require
some care. We simply record the main intermediary steps, which consist in
evaluating separately the various terms occurring in the function R1212 rewritten
as
R1212 = (1 221 2 211 ) + (12 12 11 22 ) + (123 123 113 223 ).
First, the expressions found in part (iii) for the functions 3 easily yield
123 123 113 223 = (b212 b11 b22 )
+ x3 (b11 c22 2b12 c12 + b22 c11 ) + x23 (c212 c11 c22 ).
Second, the expressions found in part (iii) for the functions and
yield, after some manipulations:
12 12 11 22 = (C12
C12
C11
C22
)a

+ x3 {(C11 b2 |2 2C12
b1 |2 + C22
b1 |1 )a
+ 2(C11
C22
C12
C12
)b }
+ x23 {b1 |1 b2 |2 b1 |2 b1 |2 )a

+ (C11
b2 |2 2C12
b1 |2 + C22
b1 |1 )b
+ (C11
C22
C12
C12
)c }.
Third, after somewhat delicate computations, which in particular make use

of the relations established in part (ii) about the functions b | and b| , it is
found that
1 221 2 211 = 1 C221 2 C211

x3 {S1212 b
+ (C11 b2 |2 2C12 b1 |2 + C22 b1 |1 )a
+ 2(C11
C22
C12
C12
)b }
+ x23 {S 12 b1 b2 + (b1 |1 b2 |2 b1 |2 b1 |2 )a

b1 |1 )b
+ (C11
b2 |2 2C12
b1 |2 + C22
+ (C11
C22
C12
C12
)c },
84
[Ch. 2
where the functions
S := C C + C
C C
C
are precisely those appearing in the left-hand sides of the Gau equations.
It is then easily seen that the above equations together yield
R1212 = {S1212 (b11 b22 b12 b12 )}
x3 {S1212 (b11 b22 b12 b12 )b
}
+ x23 {S 12 b1 b2 + (c12 c12 c11 c22 )}.
Since
S 12 b1 b2 = S1212 (b11 b22 b21 b12 ),
c12 c12 c11 c22 = (b11 b12 b11 b22 )(b11 b22 b21 b12 ),
it is finally found that the function R1212 has the following remarkable expression:
R1212 = {S1212 (b11 b22 b12 b12 )}{1 x3 (b11 + b22 ) + x23 (b11 b22 b21 b12 )}.
By the assumed Gau equations,
S1212 = b11 b22 b12 b12 .
Hence R1212 = 0 as announced.
(viii) Let be a connected and simply-connected open subset of R2 . Then
there exist open subsets ` , ` 0, of R2 such that ` is a compact subset of
for each ` 0 and
[
=
` .
`0
Furthermore, for each ` 0, there exists ` = ` (` ) > 0 such that the symmetric matrices (gij ) are positive definite at all points in ` , where
` := ` ]` , ` [ .
Finally, the open set
:=
`0
is connected and simply-connected.

Let
S ` , ` 0, be open subsets of with compact closures ` such that
= `0 ` . For each `, a set ` := ` ]` , ` [ can then be constructed in
the same way that the set 0 was
S constructed in part (i).
It is clear that the set := `0 ` is connected. To show that is simplyconnected, let be a loop in , i.e., a mapping C 0 ([0, 1]; R3 ) that satisfies
(0) = (1) and (t) for all 0 t 1.
85
Let the projection operator : be defined by (y, x3 ) = y for all

(y, x3 ) , and let the mapping 0 : [0, 1] [0, 1] R3 be defined by
0 (t, ) := (1 )(t) + ((t)) for all 0 t 1, 0 1.
Then 0 is a continuous mapping such that 0 ([0, 1][0, 1]) , by definition of
the set . Furthermore, 0 (t, 0) = (t) and 0 (t, 1) = ((t)) for all t [0, 1].
The mapping
e := C 0 ([0, 1]; R2 )
e (0) = ((0)) = ((1)) =

e (1) and
e (t) for
is a loop in since
all 0 t 1. Since is simply connected, there exist a mapping 1
C 0 ([0, 1] [0, 1]; R2 ) and a point y 0 such that
and
e and 1 (t, 2) = y 0 for all 0 t 1

1 (t, 1) =
1 (t, ) for all 0 t 1, 1 2.
Then the mapping C 0 ([0, 1] [0, 2]; R3 ) defined by

(t, ) = 0 (t, )
(t, ) = 1 (t, )
for all 0 t 1, 0 1,
for all 0 t 1, 1 2,
is a homotopy in that reduces the loop to the point (y 0 , 0) . Hence the

set is simply-connected.
(ix) By parts (iv) to (viii), the functions ijq C 1 () and pij C 1 ()
constructed as in part (iii) satisfy
j ikq k ijq + pij kqp pik jqp = 0
in the connected and simply-connected open set . By Theorem 1.6-1, there thus
exists an immersion C 3 (; E3 ) such that
gij = i j in ,
where the matrix field (gij ) C 2 (; S3> ) is defined by
g = a 2x3 b + x23 c and gi3 = i3 in .
Then the mapping C 3 (; E3 ) defined by
(y) = (y, 0) for all y ,
satisfies
a = and b =
n o
1
2
in .
|1 2 |
86
[Ch. 2
Let g i := i . Then 33 = 3 g 3 = p33 g p = 0; cf. part (iii). Hence there

exists a mapping 1 C 3 (; E3 ) such that
(y, x3 ) = (y) + x3 1 (y) for all (y, x3 ) ,
and consequently, g = +x3 1 and g 3 = 1 . The relations gi3 = g i g 3 =
i3 (cf. part (i)) then show that
( + x3 1 ) 1 = 0 and 1 1 = 1.
These relations imply that 1 = 0. Hence either 1 = a3 or 1 = a3
in , where
1 2
.
a3 :=
|1 2 |
But 1 = a3 is ruled out since
{1 2 } 1 = det(gij )|x3 =0 > 0.

Noting that
a3 = 0 implies a3 = a3 ,
we obtain, on the one hand,
g = ( + x3 a3 ) ( + x3 a3 )
= 2x3 a3 + x23 a3 a3 in .
Since, on the other hand,

g = a 2x3 b + x23 c in
by part (i), we conclude that
a = and b = a3 in ,
as desired. This completes the proof.
Remarks. (1) The functions c = b b = a3 a3 introduced in part (i)

are the covariant components of the P
third fundamental form of the surface ().
(2) The series expansion g = n0 (n + 1)xn3 a (Bn ) found in part (i)
is known; cf., e.g., Naghdi [1972].
(3) The Gau equations are used only once in the above proof, for showing
that R1212 = 0 in part (vii).
The definitions of the functions C

and C imply that the sixteen Gau
equations are satisfied if and only if they are satisfied for = 1, = 2,
= 1, = 2 and that the Codazzi-Mainardi equations are satisfied if and only
Sect. 2.9] Uniqueness of surfaces with the same fundamental forms
87
if they are satisfied for = 1, = 2, = 1 and = 1, = 2, = 2 (other

choices of indices with the same properties are clearly possible).
In other words, the Gau equations and the Codazzi-Mainardi equations in
fact reduce to one and two equations, respectively.
The regularity assumptions made in Theorem 2.8-1 on the matrix fields (a )
and (b ) can be significantly relaxed in several ways. First, C. Mardare [2003b]
has shown by means of an ad hoc, but not trivial, modification of the proof
given here, that the existence of an immersion C 3 (; E3 ) still holds under
the weaker (but certainly more natural, in view of the regularity of the resulting immersion ) assumption that (b ) C 1 (; S2 ), all other assumptions of
Theorem 2.8-1 holding verbatim.
In fact, Hartman & Wintner [1950] have shown the stronger result that the
existence theorem still holds if (a ) C 1 (; S2> ) and (b ) C 0 (; S2 ), with a
resulting mapping in the space C 2 (; E3 ). Their result has been itself super1,
seded by that of S. Mardare [2004], which asserts that if (a ) Wloc
(; S2> )
2
and (b ) Lloc (; S ) are two matrix fields that satisfy the Gau and CodazziMainardi equations in the sense of distributions, then there exists a mapping
2,
Wloc
() such that (a ) and (b ) are the fundamental forms of the surface ().
2.9
UNIQUENESS UP TO ISOMETRIES OF SURFACES

WITH THE SAME FUNDAMENTAL FORMS
In Section 2.8, we have established the existence of an immersion : R2

E3 giving rise to a surface () with prescribed first and second fundamental
forms, provided these forms satisfy ad hoc sufficient conditions. We now turn
to the question of uniqueness of such immersions.
This is the object of the next theorem, which constitutes another rigidity
theorem, called the rigidity theorem for surfaces. Like its three-dimensional
counterpart (Theorem 1.7-1), it asserts that, if two immersions C 2 (; E3 )
e C 2 (; E3 ) share the same fundamental forms, then the surface ()
e
and
is obtained by subjecting the surface () to a rotation (represented by an
orthogonal matrix Q with det Q = 1), then by subjecting the rotated surface
to a translation (represented by a vector c). Such a rigid transformation is
thus an isometry in E3 .
As shown by Ciarlet & Larsonneur [2001] (whose proof is adapted here),
the issue of uniqueness can be resolved as a corollary to its three-dimensional
counterpart, like the issue of existence. We recall that O3 denotes the set of all
orthogonal matrices of order three and that O3+ = {Q O3 ; det Q = 1} denotes
the set of all 3 3 rotations.
e C 2 (; E3 ) be two immersions such that their associated first and second
and
fundamental forms satisfy (with self-explanatory notations)
a = e
a and b = eb in .
88
[Ch. 2
Then there exist a vector c E3 and a rotation Q O3+ such that

e
(y) = c + Q(y)
for all y .
Proof. Arguments similar to those used in parts (i) and (viii) of the proof
of Theorem 2.8-1 show that there exist open subsets ` of and real numbers
` > 0, ` 0, such that the symmetric matrices (gij ) defined by
g := a 2x3 b + x23 c and gi3 = i3 ,
where c := a b b , are positive definite in the set
[
:=
` ]` , ` [ .
`0
e C 1 (; E3 ) defined by (with
The two immersions C 1 (; E3 ) and
self-explanatory notations)
e
e
e 3 (y)
(y, x3 ) := (y) + x3 a3 (y) and (y,
x3 ) := (y)
+ x3 a
for all (y, x3 ) therefore satisfy
gij = geij in .
By Theorem 1.7-1, there exist a vector c E3 and an orthogonal matrix

Q O3 such that
e
(y, x3 ) = c + Q(y,
x3 ) for all (y, x3 ) .
Hence, on the one hand,
e
det (y, x3 ) = det Q det (y,
x3 ) for all (y, x3 ) .
On the other hand, a simple computation shows that

q
det (y, x3 ) = det(a (y){1 x3 (b11 + b22 )(y) + x23 (b11 b22 b21 b12 )(y)}
for all (y, x3 ) , where
b (y) := a (y)b (y), y ,

so that
e
det (y, x3 ) = det (y,
x3 ) for all (y, x3 ) .
Therefore det Q = 1, which shows that the matrix Q O3 is in fact a

rotation. The conclusion then follows by letting x3 = 0 in the relation
e
(y, x3 ) = c + Q(y,
x3 ) for all (y, x3 ) .
89
As a preparation to our next result, we note that the second fundamental

form of the surface () can still be defined under the weaker assumptions that
a1 a2
C 1 (; E3 ) and a3 =
C 1 (; E3 ), by means of the definition
|a1 a2 |
b := a a3 ,
which evidently coincides with the usual one when C 2 (; E3 ).
Theorem 2.9-1 constitutes the classical rigidity theorem for surfaces, in the
e are assumed to be in the space C 2 (; E3 ).
sense that both immersions and
Following Ciarlet & C. Mardare [2004a], we now show that a similar ree1 a
e2
a
e H 1 (; E3 ) and a
e 3 :=
sult holds under the assumptions that
e2 |
|e
a1 a
H 1 (; E3 ) (with self-explanatory notations). Naturally, our first task will be to
e 3 , which is not necessarily well defined a.e. in for
verify that the vector field a
e H 1 (; E3 ), is nevertheless well defined a.e. in for
an arbitrary mapping
e that satisfy the assumptions of the next theorem. This fact
those mappings
e 3 are likewise well defined
will in turn imply that the functions eb := e
a a
a.e. in .

be an immersion that satisfies a3 C 1 (; E3 ). Assume that there exists a vector
e H 1 (; E3 ) that satisfies
field
e
a = a a.e. in ,
e 3 H 1 (; E3 ),
a
and
eb = b a.e. in .
Then there exist a vector c E3 and a matrix Q O3+ such that

e
(y)
= c + Q(y) for almost all y .
Proof. The proof essentially relies on the extension to a Sobolev space setting
of the three-dimensional rigidity theorem established in Theorem 1.7-3.
(i) To begin with, we record several technical preliminaries.
First, we observe that the relations e
a = a a.e. in and the assumption
that C 1 (; E3 ) is an immersion together imply that
e2| =
|e
a1 a
q
q
det(e
a ) = det(a ) > 0 a.e. in .
e 3 , and thus the functions eb , are well defined

Consequently, the vector field a
a.e. in .
Second, we establish that
b = b in and eb = eb a.e. in ,
90
[Ch. 2
e a
e3 = a
e a
e 3 a.e. in . To
i.e., that a a3 = a a3 in and a
this end, we note that either the assumptions C 1 (; E3 ) and a3 C 1 (; E3 )
together, or the assumptions H 1 (; E3 ) and a3 H 1 (; E3 ) together, imply
that a a3 = a3 L1loc (), hence that a3 D0 ().
Given any D(), let U denote an open subset of R2 such that supp U
and U is a compact subset of . Denoting by X 0 h, iX the duality pairing
between a topological vector space X and its dual X 0 , we have
Z
a3 dy
D 0 () h a3 , iD() =
Z
Z
=
(a3 ) dy ( ) a3 dy.
Observing that a3 = 0 a.e. in and that

Z
Z
(a3 ) dy =
(a3 ) dy
H 1 (U ;E3 )
h ( ), a3 iH01 (U ;E3 ) ,
we reach the conclusion that the expression D0 () h a3 , iD() is symmetric

with respect to and since = in D0 (U ). Hence a3 =
a3 in L1loc (), and the announced symmetries are established.
Third, let
e 3 a
e 3 and c := a3 a3 .
e
c := a
Then we claim that ce = c a.e. in . To see this, we note that the matrix
fields (e
a ) := (e
a )1 and (a ) := (a )1 are well defined and equal a.e. in
since is an immersion and e
a = a a.e. in . The formula of Weingarten
(Section 2.6) can thus be applied a.e. in , showing that ce = e
a ebeb a.e.
in .
The assertion then follows from the assumptions eb = b a.e. in .
(ii) Starting from the set and the mapping (as given in the statement
of Theorem 2.9-2), we next construct a set and a mapping that satisfy the
assumptions of Theorem 1.7-2. More precisely, let
(y, x3 ) := (y) + x3 a3 (y) for all (y, x3 ) R.
Then the mapping := R E3 defined in this fashion is clearly continuously differentiable on R and
q
det (y, x3 ) = det(a (y)){1 x3 (b11 + b22 )(y) + x23 (b11 b22 b21 b12 )(y)}
for all (y, x3 ) R, where
b (y) := a (y)b (y), y .

Let nS
, n 0, be open subsets of R2 such that n is a compact subset of
and = n0 n . Then the continuity of the functions a , a , b and the
91
assumption that is an immersion together imply that, for each n 0, there

exists n > 0 such that
det (y, x3 ) > 0 for all (y, x3 ) n [n , n ].
Besides, there is no loss of generality in assuming that n 1 (this property
will be used in part (iii)).
Let then
[
:=
(n ]n , n [).
n0
Then it is clear that is a connected open subset of R3 and that the mapping
C 1 (; E3 ) satisfies det > 0 in .
Finally, note that the covariant components gij C 0 () of the metric tensor
field associated with the mapping are given by (the symmetries b = b
established in (i) are used here)
g = a 2x3 b + x23 c ,
g3 = 0,
g33 = 1.
e (as given in the statement of Theorem

(iii) Starting with the mapping
e that satisfies the assumptions of Theorem
2.9-2), we construct a mapping
e : E3 by letting
1.7-2. To this end, we define a mapping
e
e
e 3 (y) for all (y, x3 ) ,
(y,
x3 ) := (y)
+ x3 a
e H 1 (; E3 ), since ]1, 1[.

where the set is defined as in (ii). Hence
e = det a.e. in since the functions eb := e
Besides, det
aeb , which are
well defined a.e. in , are equal, again a.e. in , to the functions b . Likewise, the
components geij L1 () of the metric tensor field associated with the mapping
e satisfy geij = gij a.e. in since e
a = a and eb = b a.e. in by
assumption and ce = c a.e. in by part (i).
(iv) By Theorem 1.7-2, there exist a vector c E3 and a matrix Q O3+
such that
e
e 3 (y) = c + Q((y) + x3 a3 (y)) for almost all (y, x3 ) .
(y)
+ x3 a
Differentiating with respect to x3 in this equality between functions in H 1 (; E3 )

e
e 3 (y) = Qa3 (y) for almost all y . Hence (y)
shows that a
= c + Q(y) for
almost all y as announced.

e H 1 (; E3 ) satisfying the assumptions
Remarks. (1) The existence of
e C 1 (; E3 ) and a
e 3 C 1 (; E3 ), and that
of Theorem 2.9-2 implies that
1
3
1
3
H (; E ) and a3 H (; E ).
(2) It is easily seen that the conclusion of Theorem 2.9-2 is still valid if the
e H 1 (; E3 ) and a
e 3 H 1 (; E3 ) are replaced by the weaker
assumptions
1
3
1
e H (; E ) and a
e 3 Hloc
assumptions
(; E3 ).

loc
92
2.10
[Ch. 2
CONTINUITY OF A SURFACE AS A FUNCTION

OF ITS FUNDAMENTAL FORMS
Let be a connected and simply-connected open subset of R2 . Together, Theorems 2.8-1 and 2.9-1 establish the existence of a mapping F that associates
to any pair of matrix fields (a ) C 2 (; S2> ) and (b ) C 2 (; S2 ) satisfying the Gau and Codazzi-Mainardi equations in a well-defined element
e R means that
F ((a ), (b )) in the quotient set C 3 (; E3 )/R, where (, )
3
3
e
there exists a vector c E and a rotation Q O+ such that (y) = c + Q(y)
for all y .
A natural question thus arises as to whether there exist ad hoc topologies on
the space C 2 (; S2 ) C 2 (; S2 ) and on the quotient set C 3 (; E3 )/R such that
the mapping F defined in this fashion is continuous.
Equivalently, is a surface a continuous function of its fundamental forms?
The purpose of this section, which is based on Ciarlet [2003], is to provide
an affirmative answer to the above question, through a proof that relies in an
essential way on the solution to the analogous problem in dimension three given
in Section 1.8.
Such a question is not only relevant to surface theory, but it also finds
its source in two-dimensional nonlinear shell theories, where the stored energy
functions are often functions of the first and second fundamental forms of the
unknown deformed middle surface (for an overview of nonlinear shell theories,
see, e.g., Ciarlet [2000]). For instance, the well-known stored energy function
wK proposed by Koiter [1966, Equations (4.2), (8.1), and (8.3)] for modeling
nonlinearly elastic shells made with a homogeneous and isotropic elastic material
takes the form:
wK =

3
a
(e
a a )(e
a a ) + a (eb b )(eb b ),
2
6
where 2 is the thickness of the shell,

a :=
4
a a + 2(a a +a a ),
+ 2
> 0 and > 0 are the two Lame constants of the constituting material, a
and b are the covariant components of the first and second fundamental forms
of the given undeformed middle surface, (a ) = (a )1 , and finally e
a and
eb are the covariant components of the first and second fundamental forms of
the unknown deformed middle surface.
An inspection of the above stored energy functions thus suggests a tempting
approach to shell theory, where the functions e
a and eb would be regarded
as the primary unknowns in lieu of the customary (Cartesian or curvilinear)
components of the displacement. In such an approach, the unknown components
e
a and eb must naturally satisfy the classical Gau and Codazzi-Mainardi
equations in order that they actually define a surface.
To begin with, we introduce the following two-dimensional analogs to the
notations used in Section 1.8. Let be an open subset of R3 . The notation b
Sect. 2.10]
A surface as a function of its fundamental forms
93
means that is a compact subset of . If f C ` (; R) or C ` (; E3 ), ` 0,

and b , we let
kf k`, := sup | f (y)| ,
y
||`
kk`, := sup | (y)|,

y
||`
where stands for the standard multi-index notation for partial derivatives and
|| denotes the Euclidean norm in the latter definition. If A C ` (; M3 ), ` 0,
and b , we likewise let
kAk`, = nsup | A(y)|,
y
||`
where || denotes the matrix spectral norm.

The next sequential continuity result constitutes the key step towards establishing the continuity of a surface as a function of its two fundamental forms in
ad hoc metric spaces (see Theorem 2.10-2).
Let (a ) C 2 (; S2> ) and (b ) C 2 (; S2 ) be matrix fields satisfying the Gau
and Codazzi-Mainardi equations in and let (an ) C 2 (; S2> ) and (bn )
C 2 (; S2 ) be matrix fields satisfying for each n 0 the Gau and CodazziMainardi equations in . Assume that these matrix fields satisfy
lim kan a k2, = 0 and lim kbn b k2, = 0 for all b .
Let C 3 (; E3 ) be any mapping that satisfies

a = and b =
n o
1
2
in
|1 2 |
(such mappings exist by Theorem 2.8-1). Then there exist mappings n

C 3 (; E3 ) satisfying
an = n n and bn = n
n n n o
1
2
in , n 0,
|1 n 2 n |
such that
lim kn k3, = 0 for all b .
Proof. For clarity, the proof is broken into five parts.

n
(i) Let the matrix fields (gij ) C 2 ( R; S3 ) and (gij
) C 2 ( R; S3 ), n 0,
be defined by
g := a 2x3 b + x23 c
n
g
:= an 2x3 bn + x23 cn
and gi3 := i3 ,
n
and gi3
:= i3 , n 0
94
[Ch. 2
(the variable y is omitted, x3 designates the variable in R), where

cn
c := b b , b := a b , (a ) := (a )1 ,
n
,n n
:= b,n
b,n
b , (a,n ) := (an )1 , n 0.
b ,
:= a
Let 0 be an open subset of R2 such that 0 b . Then there exists 0 =

0 (0 ) > 0 such that the symmetric matrices
n
C(y, x3 ) := (gij (y, x3 )) and Cn (y, x3 ) := (gij
(y, x3 )), n 0,
are positive definite at all points (y, x3 ) 0 , where

0 := 0 ]0 , 0 [ .
The matrices C(y, x3 ) S3 and Cn (y, x3 ) S3 are of the form (the notations
are self-explanatory):
C(y, x3 ) = C0 (y) + x3 C1 (y) + x23 C2 (y),
Cn (y, x3 ) = Cn0 (y) + x3 Cn1 (y) + x23 Cn2 (y), n 0.
First, it is easily deduced from the matrix identity B = A(I + A1 (BA))
and the assumptions limn kan a k0,0 = 0 and limn kbn b k0,0 =
0 that there exists a constant M such that
k(Cn0 )1 k0,0 + kCn1 k0,0 + kCn2 k0,0 M for all n 0.
This uniform bound and the relations
C(y, x3 ) = C0 (y){I + (C0 (y))1 (2x3 C1 (y) + x23 C2 (y))},
Cn (y, x3 ) = Cn0 (y){I + (Cn0 (y))1 (2x3 Cn1 (y) + x23 Cn2 (y))}, n 0,
together imply that there exists 0 = 0 (0 ) > 0 such that the matrices C(y, x3 )
and Cn (y, x3 ), n 0, are invertible for all (y, x3 ) 0 [0 , 0 ].
These matrices are positive definite for x3 = 0 by assumption. Hence they
remain so for all x3 [0 , 0 ] since they are invertible.
2
(ii)
S Let ` , ` 0, be open subsets of R such that ` b for each ` and
= `0 ` . By (i), there exist numbers ` = ` (` ) > 0, ` 0, such that the
n
symmetric matrices C(x) = (gij (x)) and Cn (x) = (gij
(x)), n 0, defined for all
x = (y, x3 ) R as in (i), are positive definite at all points x = (y, x3 ) ` ,
where ` := ` ]` , ` [, hence at all points x = (y, x3 ) of the open set
[
:=
` ,
`0
which is connected and simply connected. Let the functions Rqijk C 0 () be

defined from the matrix fields (gij ) C 2 (; S3> ) by
Sect. 2.10]
95
where
ijq :=
1
(j giq +i gjq q gij ) and pij := g pq ijq , with (g pq ) := (gij )1 ,
2
n
and let the functions Rqijk
C 0 (), n 0 be similarly defined from the matrix
n
2
3
fields (gij ) C (; S> ), n 0. Then
n
Rqijk = 0 in and Rqijk
= 0 in for all n 0.
That is connected and simply-connected is established in part (viii) of the

n
proof of Theorem 2.8-1. That Rqijk = 0 in , and similarly that Rqijk
= 0 in
for all n 0, is established as in parts (iv) to (viii) of the same proof.
n
(iii) The matrix fields C = (gij ) C 2 (; S3> ) and Cn = (gij
) C 2 (; S3> )
defined in (ii) satisfy (the notations used here are those of Section 1.8)
Given any compact

subset K of , there exists a finite set K of integers
S
such that K `K ` . Since by assumption,
lim kan a k2,` = 0 and lim kbn b k2,` = 0, ` K ,
n
it follows that
lim kCnp Cp k2,` = 0, ` k , p = 0, 1, 2,
where the matrices Cp and Cnp , n 0, p = 0, 1, 2, are those defined in the proof
of part (i). The definition of the norm kk2,` then implies that
lim kCn Ck2,` = 0, ` K .
The conclusion then follows from the finiteness of the set K .

(iv) Conclusion.
Given any mapping C 3 (; E3 ) that satisfies
a = and b =
n o
1
2
in ,
|1 2 |
let the mapping : E3 be defined by

(y, x3 ) := (y) + x3 a3 (y) for all (y, x3 ) ,
where a3 :=
1 2
, and let
|1 2 |
gij := i j .
96
[Ch. 2
Then an immediate computation shows that

g = a 2x3 b + x23 c and gi3 = i3 in ,
where a and b are the covariant components of the first and second fundamental forms of the surface () and c = a b b .
In other words, the matrices (gij ) constructed in this fashion coincide over
the set with those defined in part (i). Since parts (ii) and (iii) of the above
proof together show that all the assumptions of Theorem 1.8-3 are satisfied
n
by the fields C = (gij ) C 2 (; S3> ) and Cn = (gij
) C 2 (; S3> ), there exist
n
n T
n
3
3
mappings C (; E ) satisfying ( ) = Cn in , n 0, such
that
n
We now show that the mappings

n () := n (, 0) C 3 (; E3 )
indeed satisfy
an = n n and bn = n
n n n o
1
2
in .
n
|1 2 n |
Dropping the exponent n for notational convenience in this part of the proof,
let g i := i . Then 33 = 3 g 3 = p33 g p = 0, since it is easily verified that
the functions p33 , constructed from the functions gij as indicated in part (ii),
vanish in . Hence there exists a mapping 1 C 3 (; E3 ) such that
(y, x3 ) = (y) + x3 1 (y) for all (y, x3 ) .
Consequently, g = + x3 1 and g 3 = 1 . The relations gi3 = g i g 3 = i3
then show that
( + x3 1 ) 1 = 0 and 1 1 = 1.
These relations imply that 1 = 0. Hence either 1 = a3 or 1 = a3
in . But 1 = a3 is ruled out since we must have
{1 2 } 1 = det(gij )|x3 =0 > 0.
Noting that
a3 = 0 implies a3 = a3 ,
we obtain, on the one hand,
g = ( + x3 a3 ) ( + x3 a3 )
= 2x3 a3 + x23 a3 a3 in .
Sect. 2.10]
97
Since, on the other hand,

g = a 2x3 b + x23 c in ,
we conclude that
a = and b = a3 in ,
as desired.
It remains to verify that
lim kn k3, = 0 for all b .
But these relations immediately follow from the relations

lim kn k3,K = 0 for all K b ,
combined with the observations that a compact subset of is also one of ,

that (, 0) = and n (, 0) = n , and finally, that
kn k3, kn k3, .

Remark. At first glance, it seems that Theorem 2.10-1 could be established
by a proof similar to that of its three-dimensional counterpart, viz., Theorem
1.8-3. A quick inspection reveals, however, that the proof of Theorem 1.8-2 does
not carry over to the present situation.

In fact, it is not necessary to assume in Theorem 2.10-1 that the limit matrix fields (a ) and (b ) satisfy the Gau and Codazzi-Mainardi equations (see
the proof of the next theorem). More specifically, another sequential continuity
result can be derived from Theorem 2.10-1. Its interest is that the assumptions
are now made on the immersions n that define the surfaces n () for all n 0;
besides the existence of a limit surface () is also established.
For each n 0, let there be given immersions n C 3 (; E3 ), let an and bn
denote the covariant components of the first and second fundamental forms of
the surface n (), and assume that bn C 2 (). Let there be also given matrix
fields (a ) C 2 (; S2> ) and (b ) C 2 (; S2 ) with the property that
en C 3 (; E3 ) of the form
Then there exist immersions
en = cn + Qn n , cn E3 , Qn O3
98
[Ch. 2
en () and n ()
(hence the first and second fundamental forms of the surfaces
are the same for all n 0) and an immersion C 3 (, E3 ) such that a and
b are the covariant components of the first and second fundamental forms of
the surface (). Besides,
n
e k3, = 0 for all b .

lim k
Proof. An argument similar to that used in the proof of Theorem 1.8-4 shows
that passing to the limit as n is allowed in the Gau and Codazzi-Mainardi
equations, which are satisfied in the spaces C 0 () and C 1 () respectively by the
functions an and bn for each n 0 (as necessary conditions; cf. Theorem
2.7-1). Hence the limit functions a and b also satisfy the Gau and CodazziMainardi equations.
By the fundamental existence theorem (Theorem 2.8-1), there thus exists an
immersion C 3 (; E3 ) such that

1 2
a = and b =
.
|1 2 |
Theorem 2.10-1 can now be applied, showing that there exist mappings (now
en C 3 (; E3 ) such that
denoted)
an
and
e n
en and bn =
en
=
e n 2
en
1
in , n 0,
e n 2
en |
|1
e k3, = 0 for all b .

lim k
Finally, the rigidity theorem for surfaces (Theorem 2.9-1) shows that, for
each n 0, there exist cn E3 and Qn O3+ such that
en = cn + Qn n in ,
en () and n () share the same fundamental forms and the

since the surfaces
set is connected.

It remains to show how the sequential continuity established in Theorem
2.10-1 implies the continuity of a surface as a function of its fundamental forms
for ad hoc topologies.
Let be an open subset of R2 . We recall (see Section 1.8) that, for any
integers ` 0 and d 1, the space C ` (; Rd ) becomes a locally convex topological
space when its topology is defined by the family of semi-norms kk`, , b ,
and a sequence ( n )n0 converges to with respect to this topology if and only
if
lim kn k`, = 0 for all b .
n
Sect. 2.10]
99
Furthermore, this topology is metrizable: Let (i )i0 be any sequence of

subsets of that satisfy
i b and i int i+1 for all i 0, and =
i .
i=0
Then
lim kn k`, = 0 for all b lim d` ( n , ) = 0,
where
d` (, ) :=
X
1 k k`,i
.
i 1 + k k
2
`,i
i=0
Let C3 (; E3 ) := C 3 (; E3 )/R denote the quotient set of C 3 (; E3 ) by the

e R means that there exist a vector c E3
equivalence relation R, where (, )
3
e
and a matrix Q O such that (y) = c + Q(y)
for all y . Then the
3
3
set C (; E ) becomes a metric space when it is equipped with the distance d3

defined by
)
:= inf d3 (, ) = inf d3 (, c+Q),
d3 (,
3
cE
QO3
where denotes the equivalence class of modulo R.

The announced continuity of a surface as a function of its fundamental forms
is then a corollary to Theorem 2.10-1. If d is a metric defined on a set X, the
associated metric space is denoted {X; d}.
Theorem 2.10-3. Let be connected and simply connected open subset of R 2 .
Let
C02 (; S2> S2 ) := {((a ), (b )) C 2 (; S2> ) C 2 (; S2 );
C C + C
C C
C = b b b b in ,
b b + C
b C
b = 0 in }.
Given any element ((a ), (b )) C02 (; S2> S2 ), let F (((a ), (b )))

C3 (; E3 ) denote the equivalence class modulo R of any C 3 (; E3 ) that
satisfies
a = and b =
n o
1
2
in .
|1 2 |
Then the mapping

F : {C02 (; S2> S2 ); d2 } {C3 (; E3 ); d3 }
defined in this fashion is continuous.
100
[Ch. 2
Proof. Since {C02 (; S2> S); d2 } and {C 3 (; E3 ); d3 } are both metric spaces,
it suffices to show that convergent sequences are mapped through F into convergent sequences.
Let then ((a ), (b )) C02 (; S2> S2 ) and ((an ), (bn )) C02 (; S2>
2
S ), n 0, be such that
lim d2 (((an ), (bn )), ((a ), (b ))) = 0,
i.e., such that

Let there be given any F (((a ), (b ))). Then Theorem 2.10-1 shows
that there exist n F (((an ), (bn ))), n 0, such that
lim kn k3, = 0 for all b ,
i.e., such that
lim d3 (n , ) = 0.
Consequently,
lim d3 (F (((an ), (bn ))), F (((a ), (b )))) = 0,
and the proof is complete.
The above continuity results have been extended up to the boundary of the
set by Ciarlet & C. Mardare [2005b].
References
Adams, R.A. [1975]: Sobolev Spaces, Academic Press, New York.

Antman, S.S. [1976]: Ordinary differential equations of non-linear elasticity I: Foundations of the theories of non-linearly elastic rods and shells, Arch. Rational Mech.
Anal. 61, 307351.
Antman, S.S. [1995]: Nonlinear Problems of Elasticity, Springer-Verlag, Berlin
(Second Edition: 2004).
Ball, J.M. [1977]: Convexity conditions and existence theorems in nonlinear elasticity, Arch. Rational Mech. Anal. 63, 337403.
Bonnet, O. [1848]: Memoire sur la theorie generale des surfaces, Journal de lEcole
Polytechnique 19, 1146.
Cartan, E. [1927]: Sur la possibilite de plonger un espace riemannien donne dans un
espace euclidien, Annales de la Societe Polonaise de Mathematiques 6, 17.
do Carmo, M.P. [1976]: Differential Geometry of Curves and Surfaces, Prentice-Hall,
Englewood Cliffs.
do Carmo, M.P. [1994]: Differential Forms and Applications, Universitext, Springer-Verlag, Berlin (English translation of: Formas Diferenciais e Aplico
es, Instituto
da Matematica, Pura e Aplicada, Rio de Janeiro, 1971).
Choquet-Bruhat, de Witt-Morette & Dillard-Bleick [1977]: Analysis, Manifolds and Physics, North-Holland, Amsterdam (Revised Edition: 1982).
Ciarlet, P.G. [1982]:
Introduction a
` lAnalyse Numerique Matricielle et a
`
lOptimisation, Masson, Paris (English translation: Introduction to Numerical
Linear Algebra and Optimisation, Cambridge University Press, Cambridge, 1989).
Ciarlet, P.G. [1988]: Mathematical Elasticity, Volume I: Three-Dimensional Elasticity, North-Holland, Amsterdam.
Ciarlet, P.G. [2000]: Mathematical Elasticity, Volume III: Theory of Shells, NorthHolland, Amsterdam.
Ciarlet, P.G. [2003]: The continuity of a surface as a function of its two fundamental
forms, J. Math. Pures Appl. 82, 253274.
Ciarlet, P.G.; Larsonneur, F. [2001]: On the recovery of a surface with prescribed
first and second fundamental forms, J. Math. Pures Appl. 81, 167185.
Ciarlet, P.G.; Laurent, F. [2003]: Continuity of a deformation as a function of its
Cauchy-Green tensor, Arch. Rational Mech. Anal. 167, 255269.
Ciarlet, P.G.; Mardare, C. [2003]: On rigid and infinitesimal rigid displacements
in three-dimensional elasticity, Math. Models Methods Appl. Sci. 13, 15891598.
Ciarlet, P.G.; Mardare, C. [2004a]: On rigid and infinitesimal rigid displacements
in shell theory, J. Math. Pures Appl. 83, 115.
101
102
References
Ciarlet, P.G.; Mardare, C. [2004b]: Recovery of a manifold with boundary and

its continuity as a function of its metric tensor, J. Math. Pures Appl. 83, 811843.
Ciarlet, P.G.; Mardare, C. [2005a]: Continuity of a deformation in H 1 as a
function of its Cauchy-Green tensor in L1 , J. Nonlinear Sci. to appear.
Ciarlet, P.G.; Mardare, C. [2005b]: Recovery of a surface with boundary and its
continuity as a function of its two fundamental forms, Analysis and Applications,
to appear.
as, J. [1987]: Injectivity and self-contact in nonlinear elasticity,
Ciarlet, P.G.; Nec
Arch. Rational Mech. Anal. 19, 171188.
ller, S. [2002]: A theorem on geometric rigidFriesecke, G.; James, R.D.; Mu
ity and the derivation of nonlinear plate theory from three dimensional elasticity,
Comm. Pure Appl. Math. 55, 14611506.
Gallot, S.; Hulin, D.; Lafontaine, J. [2004]: Riemannian Geometry, Third Edition, Springer, Berlin.
Gauss, C.F. [1828]: Disquisitiones generales circas superficies curvas, Commentationes societatis regiae scientiarum Gottingensis recentiores 6, G
ottingen.
Hartman, P.; Wintner, A. [1950]: On the embedding problem in differential geometry, Amer. J. Math. 72, 553564.
Jacobowitz, H. [1982]: The Gauss-Codazzi equations, Tensor, N.S. 39, 1522.
Janet, M. [1926]: Sur la possibilite de plonger un espace riemannien donne dans un
espace euclidien, Annales de la Societe Polonaise de Mathematiques 5, 3843.
John, F. [1961]: Rotation and strain, Comm. Pure Appl Math. 14, 391413.
John, F. [1972]: Bounds for deformations in terms of average strains, in Inequalities,
III (O. Shisha, Editor), pp. 129144, Academic Press, New York.
Klingenberg, W. [1973]: Eine Vorlesung u
ber Differentialgeometrie, SpringerVerlag, Berlin (English translation: A Course in Differential Geometry, Springer-Verlag, Berlin, 1978).
Kohn, R.V. [1982]: New integral estimates for deformations in terms of their nonlinear
strains, Arch. Rational Mech. Anal. 78, 131172.
Koiter, W.T. [1966]: On the nonlinear theory of thin elastic shells, Proc. Kon. Ned.
Akad. Wetensch. B69, 154.
Mardare, C. [2003a]: On the recovery of a manifold with prescribed metric tensor,
Analysis and Applications 1, 433453.
Mardare, C. [2003b]: Private communication.
Mardare, S. [2004]: On isometric immersions of a Riemannian space with little
regularity, Analysis and Applications 2, 193226.
Marsden, J.E.; Hughes, T.J.R. [1983]: Mathematical Foundations of Elasticity,
Prentice-Hall, Englewood Cliffs (Second Edition: 1999).
Naghdi, P.M. [1972]: The theory of shells and plates, in Handbuch der Physik,
gge & C. Truesdell, Editors), pp. 425640, Springer-Verlag,
Vol. VIa/2 (S. Flu
Berlin.
Nash, J. [1954]: C 1 isometric imbeddings, Annals of Mathematics, 60, 383396.
as, J. [1967]: Les Methodes Directes en Theorie des Equations Elliptiques, MasNec
son, Paris.
References
103
Nirenberg, L. [1974]: Topics in Nonlinear Functional Analysis, Lecture Notes,

Courant Institute, New York University (Second Edition: American Mathematical
Society, Providence, 1994).
Reshetnyak, Y.G. [1967]: Liouvilles theory on conformal mappings under minimal
regularity assumptions, Sibirskii Math. J. 8, 6985.
Reshetnyak, Y.G. [2003]: Mappings of domains in Rn and their metric tensors,
Siberian Math. J. 44, 332345.
Schwartz, L. [1966]: Theorie des Distributions, Hermann, Paris.
Schwartz, L. [1992]: Analyse II: Calcul Differentiel et Equations Differentielles,
Hermann, Paris.
Stoker, J.J. [1969]: Differential Geometry, John Wiley, New York.
Szczarba, R.H. [1970]: On isometric immersions of Riemannian manifolds in Euclidean space, Boletim da Sociedade Brasileira de Matem
atica 1, 3145.
Szopos, M. [2005]: On the recovery and continuity of a submanifold with boundary,
Analysis and Applications, to appear.
Tenenblat, K. [1971]: On isometric immersions of Riemannian manifolds, Boletim
da Sociedade Brasileira de Matem
atica 2, 2336.
Whitney, H. [1934]: Analytic extensions of differentiable functions defined in closed
sets, Trans. Amer. Math. Soc. 36, 6389.
Yosida, K. [1966]: Functional Analysis, Springer-Verlag, Berlin.
Zeidler, E. [1986]: Nonlinear Functional Analysis and its Applications, Vol. I: FixedPoint Theorems, Springer-Verlag, Berlin.

Differential Geometry Note11

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Differential Geometry Note11

Uploaded by

Copyright:

Available Formats

AN INTRODUCTION

Lecture Notes Series

1 Three-dimensional differential geometry

2 Differential geometry of surfaces

gij (x)g j (x) for i = 1, 2, 3.

Kroneckers symbols are designated by ij , ij , or ij according to the context.

Three-dimensional differential geometry

and the three coordinates xi of x are called the curvilinear coordinates of x

Examples of curvilinear coordinates include the well-known cylindrical and

Let be an open subset of R3 and let

be a mapping that is differentiable at a point x . If x is such that (x+x)

Let the three vectors g i (x) R3 be defined by

Three-dimensional differential geometry

A mapping : E3 is an immersion at x if it is differentiable

= g ik (x)g j` (x)gk` (x) = g ik (x)kj = g ij (x),

Volumes, areas, and lengths in curvilinear coordinates

To conclude this section, we record for convenience the fundamental relations

and g ij (x) = g i (x) g j (x),

VOLUMES, AREAS, AND LENGTHS IN CURVILINEAR COORDINATES

Three-dimensional differential geometry

where n(x) := ni (x)ei denotes the unit outer normal vector at x D.

Proof. The relation db

Covariant derivatives and Christoffel symbols

Finally, consider a curve C = f (I) in , where I is a compact interval of R

COVARIANT DERIVATIVES OF A VECTOR FIELD

Three-dimensional differential geometry

where the three vectors g i (x) form the contravariant basis at x

Covariant derivatives and Christoffel symbols

vikj := j vi pij vp and pij := g p i g j ,

denotes the i-th component of g i (x) over the basis {b

Three-dimensional differential geometry

x)) (the row index is i) satisfy

We next compute the derivatives ` g q (x) (the fields g q = g qr g r are of class

Covariant derivatives and Christoffel symbols

(iii) The partial derivatives bi vbi (b

vkk` (x) := ` vk (x) q`k (x)vq (x),

We compute the partial derivatives bj vbi (b

by the chain rule and by (i). In particular then,

b x))[g k (x)]i + vq (x)bj [g q ((b

= (` vk (x) q`k (x)vq (x)) [g k (x)]i [g ` (x)]j ,

since ` g q (x) = q`k (x)g k (x) by (ii).

defined in Theorem 1.4-1 are called the first-order covariant derivatives of

are called the Christoffel symbols of the first kind.

can be also defined by the relations

(b) The Christoffel symbols pij := g p i g j = pji C 0 () satisfy the relations

Three-dimensional differential geometry

Proof. It remains to verify that the covariant derivatives vikj , defined in

NECESSARY CONDITIONS SATISFIED BY THE

It is remarkable that the components gij : R of the metric tensor of an

Necessary conditions satisfied by the metric tensor

ik g j g q = k ijq pij kqp .

Since ik g j = ij g k , we also have

Three-dimensional differential geometry

EXISTENCE OF AN IMMERSION DEFINED ON

Sect. 1.6] Existence of an immersion with a prescribed metric tensor

or equivalently, such that

Three-dimensional differential geometry

Then there exists an immersion C 3 (; E3 ) such that

Sect. 1.6] Existence of an immersion with a prescribed metric tensor

define the functions

Using the relations

= j pik + `ik pj` `ik (g pq j gq` + gq` j g pq )

= j pik + `ik pj` .

Three-dimensional differential geometry