INTRODUCTION TO

DIFFERENTIABLE
MANIFOLDS

spring 2009

Course by Prof. dr. R.C.A.M. van der Vorst

Solution manual Dr. G.J. Ridderbos
1
2

Introduction to differentiable manifolds

Lecture notes version 2.1, February 16, 2009
This is a self contained set of lecture notes. The notes were written by Rob van der
Vorst. The solution manual is written by Guit-Jan Ridderbos. We follow the book
‘Introduction to Smooth Manifolds’ by John M. Lee as a reference text.
This document was produced in LATEX and the pdf-file of these notes is available
on the following website
www.few.vu.nl/˜vdvorst
They are meant to be freely available for non-commercial use, in the sense that
“free software” is free. More precisely:

This work is licensed under the Creative Commons

Attribution-NonCommercial-ShareAlike License.
To view a copy of this license, visit
http://creativecommons.org/licenses/by-nc-sa/2.0/
or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California
94305, USA.

C ONTENTS

I. Manifolds 5
1. Topological manifolds 5
2. Differentiable manifolds and differentiable structures 11
3. Immersions, submersions and embeddings 19

II. Tangent and cotangent spaces 31

4. Tangent spaces 31
5. Cotangent spaces 37
6. Vector bundles 39
6.1. The tangent bundle and vector fields 43
6.2. The cotangent bundle and differential 1-forms 45

III. Tensors and differential forms 48

7. Tensors and tensor products 48
8. Symmetric and alternating tensors 53
9. Tensor bundles and tensor fields 59
10. Differential forms 63
11. Orientations 67

IV. Integration on manifolds 73

 3

12. Integrating m-forms on Rm 73

13. Partitions of unity 74
14. Integration on of m-forms on m-dimensional manifolds. 79
15. The exterior derivative 85
16. Stokes’ Theorem 90

V. De Rham cohomology 95
17. Definition of De Rham cohomology 95
18. Homotopy invariance of cohomology 96

VI. Exercises 100

Manifolds
Topological Manifolds 100
Differentiable manifolds 101
Immersion, submersion and embeddings 102
Tangent and cotangent spaces
Tangent spaces 104
Cotangent spaces 105
Vector bundles 106
Tensors
Tensors and tensor products 107
Symmetric and alternating tensors 107
Tensor bundles and tensor fields 108
Differential forms 109
Orientations 109
Integration on manifolds
Integrating m-forms on Rm 110
Partitions of unity 110
Integration of m-forms on m-dimensional manifolds 110
The exterior derivative 111
Stokes’ Theorem 112
Extra’s 113
De Rham cohomology
Definition of De Rham cohomology 115
4

VII. Solutions 116

 5

I. Manifolds

1. Topological manifolds
Basically an m-dimensional (topological) manifold is a topological space M
which is locally homeomorphic to Rm . A more precise definition is:

Definition 1.1. 1 A topological space M is called an m-dimensional (topological)

manifold, if the following conditions hold:
(i) M is a Hausdorff space,
(ii) for any p ∈ M there exists a neighborhood2 U of p which is homeomorphic
to an open subset V ⊂ Rm , and
(iii) M has a countable basis of open sets.

Axiom (ii) is equivalent to saying that p ∈ M has a open neighborhood U 3 p

homeomorphic to the open disc Dm in Rm . We say M is locally homeomorphic to
Rm . Axiom (iii) says that M can be covered by countably many of such neighbor-
hoods.

M
p U

Rm

F IGURE 1. Coordinate charts (U, ϕ).

1See Lee, pg. 3.

2Usually an open neighborhood U of a point p ∈ M is an open set containing p. A neighborhood

of p is a set containing an open neighborhood containing p. Here we will always assume that a
neighborhood is an open set.
6

p M

Ui
Uj

ϕi ϕj

Vi ϕij = ϕj ◦ ϕ−1
x
i
y Vj

Rm Rm

F IGURE 2. The transition maps ϕi j .

Recall some notions from topology: A topological space M is called Hausdorff

if for any pair p, q ∈ M, there exist (open) neighborhoods U 3 p, and U 0 3 q such
that U ∩U 0 = ∅. For a topological space M with topology τ, a collection β ⊂ τ is
a basis if and only if each U ∈ τ can be written as union of sets in β. A basis is
called a countable basis if β is a countable set.
Figure 1 displays coordinate charts (U, ϕ), where U are coordinate neighbor-
hoods, or charts, and ϕ are (coordinate) homeomorphisms. Transitions between
different choices of coordinates are called transitions maps ϕi j = ϕ j ◦ ϕ−1
i , which
are again homeomorphisms by definition. We usually write x = ϕ(p), ϕ : U →
V ⊂ Rn , as coordinates for U, see Figure 2, and p = ϕ−1 (x), ϕ−1 : V → U ⊂ M, as
a parametrization of U. A collection A = {(ϕi ,Ui )}i∈I of coordinate charts with
M = ∪iUi , is called an atlas for M.
The following Theorem gives a number of useful characteristics of topological
manifolds.
Theorem 1.4. 3 A manifold is locally connected, locally compact, and the union of
countably many compact subsets. Moreover, a manifold is normal and metrizable.
J 1.5 Example. M = Rm ; the axioms (i) and (iii) are obviously satisfied. As for (ii)
we take U = Rm , and ϕ the identity map. I
J 1.6 Example. M = S1 = {p = (p1 , p2 ) ∈ R2 | p21 + p22 = 1}; as before (i) and (iii)
are satisfied. As for (ii) we can choose many different atlases.
3Lee, 1.6, 1.8, and Boothby.
 7

(a): Consider the sets

U1 = {p ∈ S1 | p2 > 0}, U2 = {p ∈ S1 | p2 < 0},
U3 = {p ∈ S1 | p1 < 0}, U4 = {p ∈ S1 | p1 > 0}.
The associated coordinate maps are ϕ1 (p) = p1 , ϕ2 (p) = p1 , ϕ3 (p) = p2 , and
ϕ4 (p) = p2 . For instance
 p 
ϕ−1
1 (x) = x, 1 − x 2 ,

and the domain is V1 = (−1, 1). It is clear that ϕi and ϕ−1i are continuous, and
therefore the maps ϕi are homeomorphisms. With these choices we have found an
atlas for S1 consisting of four charts.
(b): (Stereographic projection) Consider the two charts
U1 = S1 \{(0, 1)}, and U2 = S1 \{(0, −1)}.
The coordinate mappings are given by
2p1 2p1
ϕ1 (p) = , and ϕ2 (p) = ,
1 − p2 1 + p2
which are continuous maps from Ui to R. For example
 4x x2 − 4 
ϕ−1
1 (x) = , ,
x2 + 4 x2 + 4
which is continuous from R to U1 .

p2

Np

(p′1 , p′2 )
p1
(p1 , p2 )

Sp
x x′

F IGURE 3. The stereographic projection describing ϕ1 .

(c): (Polar coordinates) Consider the two charts U1 = S1 \{(1, 0)}, and U2 =
S1 \{(−1, 0)}. The homeomorphism are ϕ1 (p) = θ ∈ (0, 2π) (polar angle counter
clockwise rotation), and ϕ2 (p) = θ ∈ (−π, π) (complex argument). For example
ϕ−1
1 (θ) = (cos(θ), sin(θ)). I
8

J 1.8 Example. M = Sn = {p = (p1 , · · · pn+1 ) ∈ Rn+1 | |p|2 = 1}. Obvious exten-

sion of the choices for S1 . I

J 1.9 Example. (see Lee) Let U ⊂ Rn be an open set and g : U → Rm a continuous

function. Define

M = Γ(g) = {(x, y) ∈ Rn × Rm : x ∈ U, y = g(x)},

endowed with the subspace topology.4 This topological space is an n-dimensional

manifold. Indeed, define π : Rn × Rm → Rn as π(x, y) = x, and ϕ = π|Γ(g) , which
is continuous onto U. The inverse ϕ−1 (x) = (x, g(x)) is also continuous. Therefore
{(Γ(g), ϕ)} is an appropriate atlas. The manifold Γ(g) is homeomorphic to U. I

J 1.10 Example. M = PRn , the real projective spaces. Consider the following
equivalence relation on points on Rn+1 \{0}: For any two x, y ∈ Rn+1 \{0} define

x ∼ y if there exists a λ 6= 0, such that x = λy.

Define PRn = {[x] : x ∈ Rn+1 \{0}} as the set of equivalence classes. One can
think of PRn as the set of lines through the origin in Rn+1 . Consider the natural
map π : Rn+1 \{0} → PRn , via π(x) = [x]. A set U ⊂ PRn is open if π−1 (U) is
open in Rn+1 \{0}. This makes PRn a compact Hausdorff space. In order to verify
that we are dealing with an n-dimensional manifold we need to describe an atlas
for PRn . For i = 1, · · · n + 1, define Vi ⊂ Rn+1 \{0} as the set of points x for which
xi 6= 0, and define Ui = π(Vi ). Furthermore, for any [x] ∈ Ui define
x xi−1 xi+1 xn+1 
1
ϕi ([x]) = ,··· , , ,··· , ,
xi xi xi xi
which is continuous, and its continuous inverse is given by

ϕ−1
 
i (z1 , · · · , zn ) = (z1 , · · · , zi−1 , 1, zi , · · · , zn ) .

These charts Ui cover PRn . In dimension n = 1 we have that PR ∼

= S1 , and in the
dimension n = 2 we obtain an immersed surface PR2 as shown in Figure 4. I

The examples 1.6 and 1.8 above are special in the sense that they are subsets of
some Rm , which, as topological spaces, are given a manifold structure.
Define
Hm = {(x1 , · · · , xm ) | xm ≥ 0},
as the standard Euclidean half-space.

4For a definition see the end of this section.

 9

F IGURE 4. Identification of the curves indicated above yields an

immersion of PR2 into R3 .

Definition 1.12. A topological space M is called an m-dimensional (topological)

manifold with boundary ∂M ⊂ M, if the following conditions hold:
(i) M is a Hausdorff space,
(ii) for any point p ∈ M there exists a neighborhood U of p, which is homeo-
morphic to an open subset V ⊂ Hm , and
(iii) M has a countable basis of open sets.

Axiom (ii) can be rephrased as follows, any point p ∈ M is contained in a neigh-

borhood U, which either homeomorphic to Dm , or to Dm ∩ Hm . The set M is locally
homeomorphic to Rm , or Hm . The boundary ∂M ⊂ M is a subset of M which con-
sists of points p for which any neighborhood cannot be homeomorphic to an open
subset of int (Hm ). In other words point p ∈ ∂M are points that lie in the inverse
image of V ∩ ∂Hm for some chart (U, ϕ). Points p ∈ ∂M are mapped to points on
∂Hm and ∂M is an (m − 1)-dimensional topological manifold.
J 1.14 Example. Consider the bounded cone

M = C = {p = (p1 , p2 , p3 ) ∈ R3 | p21 + p22 = p23 , 0 ≤ p3 ≤ 1},

with boundary ∂C = {p ∈ C | p3 = 1}. We can describe the cone via an atlas

consisting of three charts;

U1 = {p ∈ C | p3 < 1},

with x = (x1 , x2 ) = ϕ1 (p) = (p1 , p2 + 1), and

 q 
ϕ−1
1 (x) = x ,
1 2x − 1, x1
2 + (x − 1)2 ,
2
10

F IGURE 5. Coordinate maps for boundary points.

The other charts are given by

1
U2 = {p ∈ C | < p3 ≤ 1, (p1 , p2 ) 6= (0, p3 )},
2
1
U3 = {p ∈ C | < p3 ≤ 1, (p1 , p2 ) 6= (0, −p3 )}.
2
For instance ϕ2 can be constructed as follows. Define
p p   2q 
1 2 1
q = ψ(p) = , , p3 , σ(q) = , 1 − q3 ,
p3 p3 1 − q2
and x = ϕ2 (p) = (σ ◦ ψ)(p), ϕ2 (U2 ) = R × [0, 12 ) ⊂ H2 . The map ϕ3 is defined
similarly. The boundary is given by ∂C = ϕ−1 −1
2 (R×{0})∪ϕ3 (R×{0}), see Figure
6. I

J 1.16 Example. The open cone M = C = {p = (p1 , p2 , p3 ) | p21 + p22 = p23 , 0 ≤

p3 < 1}, can be described by one coordinate chart, see U1 above, and is therefore a
manifold without boundary (non-compact), and C is homeomorphic to D2 , or R2 ,
see definition below. I

So far we have seen manifolds and manifolds with boundary. A manifold can
be either compact or non-compact, which we refer to as closed, or open manifolds
respectively. Manifolds with boundary are also either compact, or non-compact. In
both cases the boundary can be compact. Open subsets of a topological manifold
are called open submanifolds, and can be given a manifold structure again.
Let N and M be manifolds, and let f : N → M be a continuous mapping. A
mapping f is called a homeomorphism between N and M if f is continuous and
has a continuous inverse f −1 : M → N. In this case the manifolds N and M are said
 11

F IGURE 6. Coordinate maps for the cone C.

to be homeomorphic. Using charts (U, ϕ), and (V, ψ) for N and M respectively,
we can give a coordinate expression for f , i.e. f˜ = ψ ◦ f ◦ ϕ−1 .
Recall the subspace topology. Let X be a topological space and let S ⊂ X be
any subset, then the subspace, or relative topology on S (induced by the topology
on X) is defined as follows. A subset U ⊂ S is open if there exists an open set
V ⊂ X such that U = V ∩ S. In this case S is called a (topological) subspace of X.
A (topological) embedding is a continuous injective mapping f : X → Y , which is
a homeomorphism onto its image f (X) ⊂ Y with respect to the subspace topology.
Let f : N → M be an embedding, then its image f (N) ⊂ M is called a submanifold
of M. Notice that an open submanifold is the special case when f = i : U ,→ M is
an inclusion mapping.

2. Differentiable manifolds and differentiable structures

A topological manifold M for which the transition maps ϕi j = ϕ j ◦ ϕ−1 i for all
pairs ϕi , ϕ j in the atlas are diffeomorphisms is called a differentiable, or smooth
manifold. The transition maps are mappings between open subsets of Rm . Dif-
feomorphisms between open subsets of Rm are C∞ -maps, whose inverses are also
C∞ -maps. For two charts Ui and U j the transitions maps are mappings:

ϕi j = ϕ j ◦ ϕ−1
i : ϕi (Ui ∩U j ) → ϕ j (Ui ∩U j ),

and as such are homeomorphisms between these open subsets of Rm .

12

F IGURE 7. Differentiability of ϕ00 ◦ ϕ−1 is achieved via the map-

pings ϕ00 ◦ (ϕ̃0 )−1 , and ϕ̃0 ◦ ϕ−1 , which are diffeomorphisms since
A ∼ A0 , and A0 ∼ A00 by assumption. This establishes the equiva-
lence A ∼ A00 .

Definition 2.1. A C∞ -atlas is a set of charts A = {(U, ϕi )}i∈I such that

(i) M = ∪i∈I Ui ,
(ii) the transition maps ϕi j are diffeomorphisms between ϕi (Ui ∩ U j ) and
ϕ j (Ui ∩U j ), for all i 6= j (see Figure 2 ).

The charts in a C∞ -atlas are said to be C∞ -compatible. Two C∞ -atlases A and A0

are equivalent if A ∪ A0 is again a C∞ -atlas, which defines a equivalence relation on
C∞ -atlases. An equivalence class of this equivalence relation is called a differen-
tiable structure D on M. The collection of all atlases associated with D, denoted
AD , is called the maximal atlas for the differentiable structure. Figure 7 shows
why compatibility of atlases defines an equivalence relation.

Definition 2.3. Let M be a topological manifold, and let D be a differentiable

structure on M with maximal atlas AD . Then the pair (M, AD ) is called a (C∞ -
)differentiable manifold.
 13

Basically, a manifold M with a C∞ -atlas, defines a differentiable structure on

M. The notion of a differentiable manifold is intuitively a difficult concept. In
dimensions 1 through 3 all topological manifolds allow a differentiable structure
(only one up to diffeomorphisms). Dimension 4 is the first occurrence of manifolds
without a differentiable structure. Also in higher dimensions uniqueness of differ-
entiable structures is no longer the case as the famous example by Milnor shows;
S7 has 28 different (non-diffeomorphic) differentiable structures. The first example
of a manifold that allows many non-diffeomorphic differentiable structures occurs
in dimension 4; exotic R4 ’s. One can also consider Cr -differentiable structures and
manifolds. Smoothness will be used here for the C∞ -case.
Theorem 2.4. 5 Let M be a topological manifold with a C∞ -atlas A. Then there
exists a unique differentiable structure D containing A, i.e. A ⊂ AD .
Proof: Let A be the collection of charts that are C∞ -compatible with A. By the
same reasoning as in Figure 7 we prove that all charts in A are C∞ -compatible with
eachother proving that A is a smooth altas. Now any chart that is C∞ -compatible
with any chart in A is C∞ -compatible with any chart in A and is thus in A and A is
a maximal atlas. Clearly any other maximal altas is contained in A and therefore
A = AD .

J 2.5 Remark. In our definition of of n-dimensional differentiable manifold we

use the local model over standard Rn , i.e. on the level of coordinates we express
differentiability with respect to the standard differentiable structure on Rn , or dif-
feomorphic to the standard differentiable structure on Rn . I

J 2.6 Example. The cone M = C ⊂ R3 as in the previous section is a differentiable

manifold whose differentiable structure is generated by the one-chart atlas (C, ϕ)
as described in the previous section. As we will see later on the cone is not a
smoothly embedded submanifold. I

J 2.7 Example. M = S1 (or more generally M = Sn ). As in Section 1 consider

S1 with two charts via stereographic projection. We have the overlap U1 ∩ U2 =
S1 \{(0, ±1)}, and the transition map ϕ12 = ϕ2 ◦ ϕ−1 1 given by y = ϕ12 (x) = x ,
4

x 6= 0, ±∞, and y 6= 0, ±∞. Clearly, ϕ12 and its inverse are differentiable functions
from ϕ1 (U1 ∩U2 ) = R\{0} to ϕ2 (U1 ∩U2 ) = R\{0}. I

J 2.9 Example. The real projective spaces PRn , see exercises Chapter VI. I

J 2.10 Example. The generalizations of projective spaces PRn , the so-called (k, n)-
Grassmannians Gk Rn are examples of smooth manifolds. I

5Lee, Lemma 1.10.

14

y = ϕ2 (p) Np

S1
ϕ12 p

x = ϕ1 (p) Sp

F IGURE 8. The transition maps for the stereographic projection.

Theorem 2.11. 6 Given a set M, a collection {Uα }α∈A of subsets, and injective
mappings ϕα : Uα → Rm , such that the following conditions are satisfied:
(i) ϕα (Uα ) ⊂ Rm is open for all α;
(ii) ϕα (Uα ∩Uβ ) and ϕβ (Uα ∩Uβ ) are open in Rm for any pair α, β ∈ A;
(iii) for Uα ∩Uβ 6= ∅, the mappings ϕα ◦ ϕ−1
β : ϕβ (Uα ∩Uβ ) → ϕα (Uα ∩Uβ ) are
diffeomorphisms for any pair α, β ∈ A;
(iv) countably many sets Uα cover M.
(v) for any pair p 6= q ∈ M, either p, q ∈ Uα , or there are disjoint sets Uα ,Uβ
such that p ∈ Uα and q ∈ Uβ .
Then M has a unique differentiable manifold structure and (Uα , ϕα ) are smooth
charts.
Proof: Let us give a sketch of the proof. Let sets ϕ−1 n
α (V ), V ⊂ R open, form a
−1
basis for a topology on M. Indeed, if p ∈ ϕ−1
α (V ) ∩ ϕβ (W ) then the latter is again
of the same form by (ii) and (iii). Combining this with (i) and (iv) we establish
a topological manifold over Rm . Finally from (iii) we derive that {(Uα , ϕα )} is a
smooth atlas.

Let N and M be smooth manifolds (dimensions n and m respectively). Let f :

N → M be a mapping from N to M.

Definition 2.12. A mapping f : N → M is said to be C∞ , or smooth if for every

p ∈ N there exist charts (U, ϕ) of p and (V, ψ) of f (p), with f (U) ⊂ V , such that
f˜ = ψ ◦ f ◦ ϕ−1 : ϕ(U) → ψ(V ) is a C∞ -mapping (from Rn to Rm ).

The above definition also holds true for mappings defined on open subsets of
N, i.e. let W ⊂ N is an open subset, and f : W ⊂ N → M, then smoothness on W

6See Lee, Lemma 1.23.

 15

F IGURE 9. Coordinate representation for f , with f (U) ⊂ V .

is defined as above by restricting to pionts p ∈ W . With this definition coordinate

maps ϕ : U → Rm are smooth maps.
The definition of smooth mappings allows one to introduce the notion of dif-
ferentiable homeomorphism, of diffeomorphism between manifolds. Let N and M
be smooth manifolds. A C∞ -mapping f : N → M, is called a diffeomorphism if
it is a homeomorphism and also f −1 is a smooth mapping, in which case N and
M are said to be diffeomorphic. The associated differentiable structures are also
called diffeomorphic. Diffeomorphic manifolds define an equivalence relation. In
the definition of diffeomorphism is suffices require that f is a differentiable bijec-
tive mapping with smooth inverse (see Theorem 2.15). A mapping f : N → M is
called a local diffeomorphism if for every p ∈ N there exists a neighborhood U,
with f (U) open in M, such that f : U → f (U) is a diffeomorphism. A mapping
f : N → M is a diffeomorphism if and only if it is a bijective local diffeomorphism.
J 2.14 Example. Consider N = R with atlas (R, ϕ(p) = p), and M = R with atlas
(R, ψ(q) = q3 ). Clearly these define different differentiable structures. Between
N and M we consider the mapping f (p) = p1/3 , which is a homeomorphism be-
tween N and M. The claim is that f is also a diffeomorphism. Take U = V = R,
then ψ ◦ f ◦ ϕ−1 (p) = (p1/3 )3 = p is the identity and thus C∞ on R, and the same
for ϕ ◦ f −1 ◦ ψ−1 (q) = (q3 )1/3 = q. The associated differentiable structures are
diffeomorphic. In fact the above described differentiable structures correspond to
3 f 3 (p)
defining the differential quotient via limh→0 f (p+h)−
h . I
16

Theorem 2.15. 7 Let N, M be smooth manifolds with atlases AN and AM respec-

tively. For continuous maps the following properties hold.

(i) Given smooth maps fα : Uα → M, for all Uα ∈ AN , with fα |Uα ∩Uβ =

fβ |Uα ∩Uβ for all α, β. Then there exists a unique smooth map f : N → M
such that f |Uα = fα .
(ii) A smooth map f : N → M between smooth manifolds is continuous.
(iii) Let f : N → M be continuous, and suppose that the maps f˜αβ = ψβ ◦ f ◦ϕ−1 α ,
for charts (Uα , ϕα ) ∈ AN , and (Vβ , ψβ ) ∈ AM , are smooth on their domains
for all α, β. Then f is smooth.

Proof: Define f by f |Uα = fα , which is well-defined by the overlap conditioins.

Given p ∈ M, there exists a chart Uα 3 p and f˜ = f˜α which is smooth by definition,
and thus f is smooth.
For any p ∈ U (chart) and choose f (p) ∈ V (chart), then f˜ is a smooth map and
f |U = ψ−1 ◦ f˜ ◦ ϕ : U → V is continuous. Continuity holds for each neighborhood
of p ∈ M.
Let p ∈ Uα and f (p) ∈ Vβ and set U = f −1 (Vβ ) ∩ Uα ⊂ Uα which is open by
continuity of f . Now f˜ = f˜αβ on the charts (U, ϕα |U ) and (Vβ , ψβ ) which proves
the differentiability of f .

M
ϕ Rm
p U
q

V
ϕ̃ = ϕ ◦ ψ −1

x ϕ̃−1 = ψ ◦ ϕ−1
Rm

F IGURE 10. Coordinate diffeomorphisms and their local repre-

sentation as smooth transition maps.

7See Lee Lemmas 2.1, 2.2 and 2.3.

 17

With this theorem at hand we can verify differentiability locally. In order to

show that f is a diffeomorphism we need that f is a homeomorphism, or a bijection
that satisfies the above local smoothness conditions, as well as for the inverse.
J 2.17 Example. Let N = M = PR1 and points in PR1 are equivalence classes
[x] = [(x1 , x2 )]. Define the mapping f : PR1 → PR1 as [(x1 , x2 )] 7→ [(x12 , x22 )]. Con-
sider the charts U1 = {x1 6= 0}, and U2 = {x2 6= 0}, with for example ϕ1 (x) =
tan−1 (x2 /x1 ), then ϕ−1 (θ1 ) = [cos(θ1 ), sin(θ1 )] , with θ1 ∈ V1 = (−π/2, π/2). In

local coordinates on V1 we have

fe(θ1 ) = tan−1 sin2 (θ1 )/ cos2 (θ1 ) ,

and a similar expression for V2 . Clearly, f is continuous and the local expressions
also prove that f is a differentiable mapping using Theorem 2.15. I

The coordinate maps ϕ in a chart (U, ϕ) are diffeomorphisms. Indeed, ϕ : U ⊂

M → Rn , then if (V, ψ) is any other chart with U ∩V 6= ∅, then by the definition the
transition maps ψ ◦ ϕ−1 and ϕ ◦ ψ−1 are smooth. For ϕ we then have ϕ̃ = ϕ ◦ ψ−1
and ϕ̃−1 = ψ ◦ ϕ−1 , which proves that ϕ is a diffeomorphism, see Figure 10. This
justifies the terminology smooth charts.
J 2.18 Remark. For arbitrary subsets U ⊂ Rm and V ⊂ Rn a map f : U ⊂ Rm
→ V ⊂ Rn is said to be smooth at p ∈ U, if there exists an open set U † ⊂ Rm
containing p, and a smooth map f † : U † → Rm , such that f and f † coincide on
U ∩ U † . The latter is called an extension of f . A mapping f : U → V between
arbitrary subsets U ⊂ Rm and V ⊂ Rn is called a diffeomorphism if f maps U
homeomorphically onto V and both f and f −1 are is smooth (as just described
above). In this case U and V are said to be diffeomorphic. I

J 2.19 Example. An important example of a class of differentiable manifolds are

appropriately chosen subsets of Euclidean space that can be given a smooth mani-
fold structure. Let M ⊂ R` be a subset such that every p ∈ M has a neighborhood
U 3 p in M (open in the subspace topology, see Section 3) which is diffeomorphic
to an open subset V ⊂ Rm (or, equivalently an open disc Dm ⊂ Rm ). In this case the
set M is a smooth m-dimensional manifold. Its topology is the subsapce topology
and the smooth structure is inherited from the standard smooth structure on R` ,
which can be described as follows. By definition a coordinate map ϕ is a diffeo-
morphisms which means that ϕ : U → V = ϕ(U) is a smooth mapping (smoothness
via a smooth map ϕ† ), and for which also ϕ−1 is a smooth map. This then directly
implies that the transition maps ψ ◦ ϕ−1 and ϕ ◦ ψ−1 are smooth mappings, see
Figure 11. In some books this construction is used as the definition of a smooth
manifold. In Section ?? we will refer to the above class of differentiable manifolds
18

Rℓ M Rℓ p M
U
p U
W

ϕ ψ
ϕ
V V ψ ◦ ϕ−1 V′
x y
x

Rm
Rm ϕ ◦ ψ −1 Rm

F IGURE 11. The transitions maps ψ ◦ ϕ−1 and ϕ ◦ ψ−1 smooth

mappings establishing the smooth structure on M.

as manifolds in Euclidean space, which are smoothly embedded submanifolds in

Euclidean space. I

J 2.21 Example. Let us consider the cone M = C described in Example 1 (see also
Example 2 in Section 1). We already established that C is manifold homeomrphic
to R2 , and moreover C is a differentiable manifold, whose smooth structure is
defined via a one-chart atlas. However, C is not a smooth manifold with respect to
the induced smooth structure as subset of R3 . Indeed, following the definition in the
above remark, we have U = C, and coordinate homeomorphism ϕ(p) = (p1 , p2 ) =
x. By the definition of smoothq maps it easily follows that ϕ is smooth. The inverse
−1


is given by ϕ (x) = x1 , x2 , x12 + x22 , which is clearly not differentiable at the
cone-top (0, 0, 0). The cone C is not a smoothly embedded submainfold of R3
(topological embedding). I

Let U,V and W be open subsets of Rn , Rk and Rm respectively, and let f : U → V

and g : V → W be smooth maps with y = f (x), and z = g(y). Then the Jacobians
are
   
∂ f1 ∂ f1 ∂g1
∂x1 · · · ∂xn ∂y1 · · · ∂g
∂yk
1

 . .
. . ...  ,
  . .. .. 
J f |x =  . Jg|y =  .
 .   . . . ,
∂ fk ∂ fk ∂gm ∂gm
∂x1 · · · ∂xn ∂y1 · · · ∂yk

and J(g ◦ f )|x = Jg|y= f (x) · J f |x (chain-rule). The commutative diagram for the
maps f , g and g ◦ f yields a commutative diagram for the Jacobians:
 19

V Rk
 A





 A
f
A g J f |x


A Jg|y

A A
A
A




AA
U A

AU
U - W
g◦ f R n - Rm
J(g◦ f )|x

For diffeomorphisms between open sets in Rn we have a number of important

properties. Let U ⊂ Rn and V ⊂ Rm be open sets, and let f : U → V is a diffeomor-
phism, then n = m, and J f |x is invertible for any x ∈ U. Using the above described
commutative diagrams we have that f −1 ◦ f yields that J( f −1 )|y= f (x) · J f |x is the
identity on Rn , and f ◦ f −1 yields that J f |x · J( f −1 )|y= f (x) is the identity on Rm .
Thus J fx has an inverse and consequently n = m. Conversely, if f : U → Rn ,
U ⊂ Rn , open, then we have the Inverse Function Theorem;
Theorem 2.22. If J f |x : Rn → Rn is invertible, then f is a diffeomorphism between
sufficiently small neighborhoods U 0 and f (U 0 ) of x and y respectively.

3. Immersions, submersions and embeddings

Let N and M be smooth manifolds of dimensions n and m respectively, and let
f : N → M be a smooth mapping. In local coordinates f˜ = ψ ◦ f ◦ ϕ−1 : ϕ(U) →
ψ(V ), with respects to charts (U, ϕ) and (V, ψ). The rank of f at p ∈ N is defined
as the rank of f˜ at ϕ(p), i.e. rk( f )| p = rk(J f˜)|ϕ(p) , where J f˜|ϕ(p) is the Jacobian
of f at p:
 ˜
∂ f˜1

∂ f1
· · ·
 ∂x. 1 . ∂xn
.. 
J f˜|x=ϕ(p) =  
.. .. . 
˜
∂ fm ∂ f˜m
∂x1 · · · ∂xn

This definition is independent of the chosen charts, see Figure 12. Via the commu-
tative diagram in Figure 12 we see that f˜ = (ψ0 ◦ ψ−1 ) ◦ f˜ ◦ (ϕ0 ◦ ϕ−1 )−1 , and by the
chain rule J f˜|x0 = J(ψ0 ◦ ψ−1 )|y · J f˜|x · J(ϕ0 ◦ ϕ−1 )−1 |x0 . Since ψ0 ◦ ψ−1 and ϕ0 ◦ ϕ−1
are diffeomorphisms it easily follows that rk(J f˜)|x = rk(J f˜)|x0 , which shows that
our notion of rank is well-defined. If a map has constant rank for all p ∈ N we sim-
ply write rk( f ). These are called constant rank mappings. Let us now consider
the various types of constant rank mappings between manifolds.
20

F IGURE 12. Representations of f via different coordinate charts.

Definition 3.2. A mapping f : N → M is called an immersion if rk( f ) = n, and a

submersion if rk( f ) = m. An immersion that is injective,8 or 1-1, and is a home-
omorphism onto (surjective mapping9) its image f (N) ⊂ M, with respect to the
subspace topology, is called a smooth embedding.

A smooth embedding is an (injective) immersion that is a topological embed-

ding.
Theorem 3.3. 10 Let f : N → M be smooth with constant rank rk( f ) = k. Then
for each p ∈ N, and f (p) ∈ M, there exist coordinates (U, ϕ) for p and (V, ψ) for
f (p), with f (U) ⊂ V , such that
(ψ ◦ f ◦ ϕ−1 )(x1 , · · · xk , xk+1 , · · · xn ) = (x1 , · · · xk , 0, · · · , 0).
J 3.4 Example. Let N = (− π2 , 3π 2
2 ), and M = R , and the mapping f is given by
f (t) = (sin(2t), cos(t)). In Figure 13 we displayed the image of f . The Jacobian is
given by !
2 cos(2t)
J f |t =
− sin(t)
Clearly, rk(J f |t ) = 1 for all t ∈ N, and f is an injective immersion. Since N is
an open manifold and f (N) ⊂ M is a compact set with respect to the subspace
8A mapping f : N → M is injective if for all p, p0 ∈ N for which f (p) = f (p0 ) it holds that p = p0 .
9
A mapping f : N → M is surjective if f (N) = M, or equivalently for every q ∈ M there exists a
p ∈ N such that q = f (p).
10See Lee, Theorem 7.8 and 7.13.
 21

topology, it follows f is not a homeomorphism onto f (N), and therefore is not an

embedding. I

F IGURE 13. Injective parametrization of the figure eight.

J 3.6 Example. Let N = S1 be defined via the atlas A = {(Ui , ϕi )}, with ϕ−1 1 (t) =
((cos(t), sin(t)), and ϕ−1
2 (t) = ((sin(t), cos(t)), and t ∈ (− π 3π
,
2 2 ). Furthermore, let
2
M = R , and the mapping f : N → M is given in local coordinates; in U1 as in
Example 1. Restricted to S1 ⊂ R2 the map f can also be described by
f (x, y) = (2xy, x).
This then yields for U2 that f˜(t) = (sin(2t), sin(t)). As before rk( f ) = 1, which
shows that f is an immersion of S1 . However, this immersion is not injective at
the origin in R2 , indicating the subtle differences between these two examples, see
Figures 13 and 14. I

F IGURE 14. Non-injective immersion of the circle.

J 3.8 Example. Let N = R, M = R2 , and f : N → M defined by f (t) = (t 2 ,t 3 ). We

can see in Figure 15 that the origin is a special point. Indeed, rk(J f )|t = 1 for all
t 6= 0, and rk(J f )|t = 0 for t = 0, and therefore f is not an immersion. I
J 3.10 Example. Consider M = PRn . We established PRn as smooth manifolds.
For n = 1 we can construct an embedding f : PR1 → R2 as depicted in Figure 16.
For n = 2 we find an immersion f : PR2 → R3 as depicted in Figure 17. I
22

R R2

F IGURE 15. The map f fails to be an immersion at the origin.

F IGURE 16. PR is diffeomorphic to S1 .

F IGURE 17. Identifications for PR2 giving an immersed non-

orientable surface in R3 .

J 3.13 Example. Let N = R2 and M = R, then the projection mapping f (x, y) = x

is a submersion. Indeed, J f |(x,y) = (1 0), and rk(J f |(x,y) ) = 1. I
 23

J 3.14 Example. Let N = M = R2 and consider the mapping f (x, y) = (x2 , y)t . The
Jacobian is !
2x 0
J f |(x,y) = ,
0 1
and rk(J f )|(x,y) = 2 for x 6= 0, and rk(J f )|(x,y) = 1 for x = 0. See Figure 18. I

F IGURE 18. The projection (submersion), and the folding of the

plane (not a submersion).

J 3.16 Example. We alter Example 2 slightly, i.e. N = S1 ⊂ R2 , and M = R2 , and

again use the atlas A. We now consider a different map f , which, for instance on
U1 , is given by f˜(t) = (2 cos(t), sin(t)) (or globally by f (x, y) = (2x, y)). It is clear
that f is an injective immersion, and since S1 is compact it follows from Lemma
3.18 below that S1 is homeomorphic to it image f (S1 ), which show that f is a
smooth embedding (see also Appendix in Lee). I

J 3.17 Example. Let N = R, M = R2 , and consider the mapping f (t) =

(2 cos(t), sin(t)). As in the previous example f is an immersion, not injective how-
ever. Also f (R) = f (S1 ) in the previous example. The manifold N is the universal
covering of S1 and the immersion f descends to a smooth embedding of S1 . I
11
Lemma 3.18. Let f : N → M be an injective immersion. If
(i) N is compact, or if
(ii) f is a proper map,12
then f is a smooth embedding.
Proof: For (i) we argue as follows. Since N is compact any closed subset X ⊂ N
is compact, and therefore f (X) ⊂ M is compact and thus closed; f is a closed

11Lee, Prop. 7.4, and pg. 47.

12A(ny) map f : N → M is called proper if for any compact K ⊂ M, it holds that also f −1 (K) ⊂ N

is compact. In particular, when N is compact, continuous maps are proper.

24

mapping.13 We are done if we show that f is a topological embedding. By the

assumption of injectivity, f : N → f (N) is a bijection. Let X ⊂ N be closed, then
( f −1 )−1 (X) = f (X) ⊂ f (N) is closed with respect to the subspace topology, and
thus f −1 : f (N) → N is continuous, which proves that f is a homeomorphism onto
its image and thus a topological embedding.
A straightforward limiting argument shows that proper continuous mappings
bewteen manifolds are closed mappings (see Exercises).

Let us start with defining the notion of embedded submanifolds.

Definition 3.19. A subset N ⊂ M is called a smooth embedded n-dimensional

submanifold in M if for every p ∈ N, there exists a chart (U, ϕ) for M, with p ∈ U,
such that

ϕ(U ∩ N) = ϕ(U) ∩ Rn × {0} = {x ∈ ϕ(U) : xn+1 = · · · = xm = 0}.

The co-dimension of N is defined as codim N = dim M − dim N.

The set W = U ∩ N in N is called a n-dimensional slice, or n-slice of U, see

Figure 19, and (U, ϕ) a slice chart. The associated coordinates x = (x1 , · · · , xn ) are
called slice coordinates. Embedded submanifolds can be characterized in terms of
embeddings.

F IGURE 19. Take a k-slice W . On the left the image ϕ(W ) corre-
sponds to the Eucliden subspace Rk ⊂ Rm .

Theorem 3.21. 14 Let N ⊂ M be a smooth embedded n-submanifold. Endowed

with the subspace topology, N is a n-dimensional manifold with a unique (induced)
smooth structure such that the inclusion map i : N ,→ M is an embedding (smooth).

13 A mapping f : N → M is called closed if f (X) is closed in M for any closed set X ⊂ N.

Similarly, f is called open if f (X) is open in M for every open set X ⊂ N.

14See Lee, Thm’s 8.2
 25

To get an idea let us show that N is a topological manifold. Axioms (i) and
(iii) are of course satisfied. Consider the projection π : Rm → Rn , and an inclusion
j : Rn → Rm defined by
π(x1 , · · · , xn , xn+1 , · · · , xm ) = (x1 , · · · , xn ),
j(x1 , · · · , xn ) = (x1 , · · · , xn , 0, · · · , 0).
Now set Z = (π ◦ ϕ)(W ) ⊂ Rn , and ϕ̄ = (π ◦ ϕ)|W , then ϕ̄−1 = (ϕ−1 ◦ j)|Z , and
ϕ̄ : W → Z is a homeomorphism. Therefore, pairs (W, ϕ̄) are charts for N, which
form an atlas for N. The inclusion i : N ,→ M is a topological embedding.
Given slice charts (U, ϕ) and (U 0 , ϕ0 ) and associated charts (W, ϕ̄) and (W 0 , ϕ̄0 )
for N. For the transitions maps it holds that ϕ̄ ◦ ϕ̄−1 = π ◦ ϕ0 ◦ ϕ−1 ◦ j, which are
diffeomorphisms, which defines a smooth atlas for N. The inclusion i : N ,→ M can
be expressed in local coordinates;
ĩ = ϕ ◦ i ◦ ϕ̄−1 , (x1 , · · · , xn ) 7→ (x1 , · · · , xn , 0, · · · , 0),
which is an injective immersion. Since i is also a topological embedding it is thus
a smooth embedding. It remains to prove that the smooth structure is unique, see
Lee Theorem 8.2.
15
Theorem 3.22. The image of an embedding is a smooth embedded submanifold.
Proof: By assumption rk( f ) = n and thus for any p ∈ N it follows from Theorem
3.3 that
fe(x1 , · · · , xn ) = (x1 , · · · , xn , 0, · · · , 0),
for appropriate coordinates (U, ϕ) for p and (V, ψ) for f (p), with f (U) ⊂ V . Con-
sequently, f (U) is a slice in V , since ψ( f (U) satisfies Definition 3.19. By assump-
tion f (U) is open in f (N) and thus f (U) = A ∩ f (N) for some open set A ⊂ M. By
replacing V by V 0 = A ∩V and restricting ψ to V 0 , (V 0 , ψ) is a slice chart with slice
V 0 ∩ f (N) = V 0 ∩ f (U).
Summarizing we conclude that embedded submanifolds are the images of
smooth embeddings.
A famous result by Whitney says that considering embeddings into Rm is not
not really a restriction for defining smooth manifolds.
Theorem 3.23. 16 Any smooth n-dimensional manifold M can be (smoothly) em-
bedded into R2n+1 .
A subset N ⊂ M is called an immersed submanifold if N is a smooth n-
dimensional manifold, and the mapping i : N ,→ M is a (smooth) immersion. This
means that we can endow N with an appropriate manifold topology and smooth
15See Lee, Thm’s 8.3
16See Lee, Ch. 10.
26

structure, such that the natural inclusion of N into M is an immersion. If f : N → M

is an injective immersion we can endow f (N) with a topology and unique smooth
structure; a set U ⊂ f (N) is open if and only if f −1 (U) ⊂ N is open, and the
(smooth) coordinate maps are taken to be ϕ ◦ f −1 , where ϕ’s are coordinate maps
for N. This way f : N → f (N) is a diffeomorphism, and i : f (N) ,→ M an injective
immersion via the composition f (N) → N → M. This proves:
17
Theorem 3.24. Immersed submanifolds are exactly the images of injective im-
mersions.
We should point out that embedded submanifolds are examples of immersed
submanifolds, but not the other way around. For any immersion f : N → M, the
image f (N) is called an immersed manifold in M.
In this setting Whitney established some improvements of Theorem 3.23.
Namely, for dimension n > 0, any smooth n-dimensional manifold can be em-
bedded into R2n (e.g. the embedding of curves in R2 ). Also, for n > 1 any smooth
n-dimensional manifold can be immersed into R2n−1 (e.g. the Klein bottle). In
this course we will often think of smooth manifolds as embedded submanifolds of
Rm . An important tool thereby is the general version of Inverse Function Theorem,

F IGURE 20. An embedding of R [left], and an immersion of S2

[right] called the Klein bottle.

which can easily be derived from the ‘Euclidean’ version 2.22.

Theorem 3.26. 18 Let N, M be smooth manifolds, and f : N → M is a smooth
mapping. If, at some point p ∈ N, it holds that J f˜|ϕ(p)=x is an invertible matrix,
then there exist sufficiently small neighborhoods U0 3 p, and V0 ∈ f (p) such that
f |U0 : U0 → V0 is a diffeomorphism.
As a direct consequence of this result we have that if f : N → M, with dim N =
dim M, is an immersion, or submersion, then f is a local diffeomorphism. If f is a
bijection, then f is a (global) diffeomorphism.
17See Lee, Theorem 8.16.
18See Lee, Thm’s 7.6 and 7.10.
 27

Theorem 3.27. Let f : N → M be a constant rank mapping with rk( f ) = k. Then

for each q ∈ f (N), the level set S = f −1 (q) is an embedded submanifold in N with
co-dimension equal to k.
Proof: Clearly, by continuity S is closed in N. By Theorem 3.3 there are coor-
dinates (U, ϕ) of p ∈ S and (V, ψ) of f (p) = q, such that

fe(x1 , · · · , xk , xk+1 , · · · , xn ) = (x1 , · · · , xk , 0, · · · , 0) = ψ(q) = 0.

The points in S ∩U are characterized by ϕ(S ∩U) = {x | (x1 , · · · , xk ) = 0}. There-

fore, S ∩U is a (n − k)-slice, and thus S is a smooth submanifold in M of codimen-
sion k.

In particular, when f : N → M is a submersion, then for each q ∈ f (N), the

level set S = f −1 (q) is an embedded submanifold of co-dimension codim S = m =
dim M. In the case of maximal rank this statement can be restricted to just one
level. A point p ∈ N is called a regular point if rk( f )| p = m = dim M, otherwise
a point is called a critical point. A level q ∈ f (N) is called a regular level if all
points p ∈ f −1 (q) are regular points, otherwise a level is called a critical level.
Theorem 3.28. Let f : N → M be a smooth map. If q ∈ f (N) is a regular value,
then S = f −1 (q) is an embedded submanifold of co-dimension equal to dim M.
Proof: Let us illustrate the last result for the important case N = Rn and M = Rm .
For any p ∈ S = f −1 (q) the Jacobian J f | p is surjective by assumption. Denote the
kernel of J f | p by ker J f | p ⊂ Rn , which has dimension n − m. Define

g : N = Rn → Rn−m × Rm ∼
= Rn ,

by g(ξ) = (Lξ, f (ξ) − q)t , where L : N = Rn → Rn−m is any linear map which
is invertible on the subspace ker J f | p ⊂ Rn . Clearly, Jg| p = L ⊕ J f | p , which, by
construction, is an invertible (linear) map on Rn . Applying Theorem 2.22 (Inverse
Function Theorem) to g we conclude that a sufficiently small neighborhood of U
of p maps diffeomorphically onto a neighborhood V of (L(p), 0). Since g is a
diffeomorphism it holds that g−1 maps Rn−m × {0} ∩V onto f −1 (q) ∩U (the 0 ∈

Rm corresponds to q). This exactly says that every point p ∈ S allows an (n − m)-
slice and is therefore an (n − m)-dimensional submanifold in N = Rn (codim S =
m).

J 3.30 Example. Let us start with an explicit illustration of the above proof. Let
N = R2 , M = R, and f (p1 , p2 ) = p21 + p22 . Consider the regular value q = 2, then
√
J f | p = (2p1 2p2 ), and f −1 (2) = {p : p21 + p22 = 2}, the circle with radius 2.
We have ker J f | p = span{(p1 , −p2 )t }, and is always isomorphic to R. For example
28

F IGURE 21. The map g yields 1-slices for the set S.

fix the point (1, 1) ∈ S, then ker J f | p = span{(1, −1)t } and define
! ! !
L(ξ) ξ1 − ξ2 1 −1
g(ξ) = = , Jg| p =
f (ξ) − 2 ξ21 + ξ22 − 2 2 2
where the linear map is L = (1 − 1). The map g is a local diffeomorphism and
on S ∩U this map is given by
q !
ξ1 − 2 − ξ21
q
g(ξ1 , 2 − ξ21 ) = ,
0

with ξ1 ∈ (1 − ε, 1 + ε). The first component has derivative 1 + √ ξ1 2 , and there-

2−ξ1


fore S ∩U is mapped onto a set of the form R × {0} ∩V . This procedure can be
carried out for any point p ∈ S, see Figure 21. I
J 3.31 Example. Let N = R2 \{(0, 0)}×R = R3 \{(0, 0, λ)}, M = (−1, ∞)×(0, ∞),
and
! !
x2 + y2 − 1 2x 2y 0
f (x, y, z) = , with J f |(x,y,z) =
1 0 0 0
We see immediately that rk( f ) = 1 on N. This map is not a submersion, but is of
constant rank, and therefore for any value q ∈ f (N) ⊂ M it holds that S = f −1 (q)
is a embedded submanifold, see Figure 22. I
J 3.33 Example. Let N, M as before, but now take
! !
x 2 + y2 − 1 2x 2y 0
f (x, y, z) = , with J f |(x,y,z) =
z 0 0 1

Now rk( f ) = 2, and f is a submersion. For every q ∈ M, the set S = f −1 (q) is a

embedded submanifold, see Figure 23. I
J 3.35 Example. Let N = R2 , M = R, and f (x, y) = 14 (x2 − 1)2 + 12 y2 . The Jacobian
is J f |(x,y) = (x(x2 −1) y). This is not a constant rank, nor a submersion. However,
 29

F IGURE 22. An embedding of a cylinder via a constant rank map-

ping.

F IGURE 23. An embedding circle of an submersion.

any q > 0 is a regular value since then the rank is equal to 1. Figure 24 shows the
level set f −1 (0) (not an embedded manifold), and f −1 (1) (an embedded circle) I

F IGURE 24. A regular and critical level set.

Theorem 3.37. 19 Let S ⊂ M be a subset of a smooth m-dimensional manifold M.

Then S is a k-dimensional smooth embedded submanifold if and if for every p ∈ S
19See Lee, Prop.8.12.
30

there exists a neighborhood U 3 p such that U ∩ S is the level set of a submersion

f : U → Rm−k .
Proof: Assume that S ⊂ M is a smooth embedded manifold. Then for each
p ∈ S there exists a chart (U, ϕ), p ∈ U, such that ϕ(S ∩ U) = {x : xk+1 = · · · =
xm = 0}. Clearly, S ∩U is the sublevel set of f : U → Rm−k at 0, given by fe(x) =
(xk+1 , · · · , xm ), which is a submersion.
Conversely, if S ∩ U = f −1 (0), for some submersion f : U → Rm−k , then by
Theorem 3.28, S ∩U is an embedded submanifold of U. This clearly shows that S
is an embedded submanifold in M.
 31

II. Tangent and cotangent

spaces

4. Tangent spaces
For a(n) (embedded) manifold M ⊂ R` the tangent space Tp M at a point p ∈
M can be pictured as a hyperplane tangent to M. In Figure 25 we consider the
parametrizations x+tei in Rm . These parametrizations yield curves γi (t) = ϕ−1 (x+
tei ) on M whose velocity vectors are given by

d −1
γ0 (0) = ϕ (x + tei ) = Jϕ−1 |x (ei ).
dt t=0

The vectors p + γ0 (0) are tangent to M at p and span an m-dimensional affine linear
subspace M p of R` . Since the vectors Jϕ−1 |x (ei ) span Tp M the affine subspace is
given by

M p := p + Tp M ⊂ R` ,

which is tangent to M at p.

F IGURE 25. Velocity vectors of curves of M span the ‘tangent

space’.
32

The considerations are rather intuitive in the sense that we consider only em-
bedded manifolds, and so the tangent spaces are tangent m-dimensional affine sub-
spaces of R` . One can also define the notion of tangent space for abstract smooth
manifolds. There are many ways to do this. Let us describe one possible way (see
e.g. Lee, or Abraham, Marsden and Ratiu) which is based on the above considera-
tions.
Let a < 0 < b and consider a smooth mapping γ : I = (a, b) ⊂ R → M, such that
γ(0) = p. This mapping is called a (smooth) curve on M, and is parametrized by
t ∈ I. If the mapping γ (between the manifolds N = I, and M) is an immersion,
then γ is called an immersed curve. For such curves the ‘velocity vector’ Jγ̃|t =
(ϕ ◦ γ)0 (t) in Rm is nowhere zero. We build the concept of tangent spaces in order
to define the notion velocity vector to a curve γ.
Let (U, ϕ) be a chart at p. Then, two curves γ and γ̃ are equivalent, γ̃ ∼ γ, if

γ̃(0) = γ(0) = p, and (ϕ ◦ γ̃)0 (0) = (ϕ ◦ γ)0 (0).

The equivalence class of a curve γ through p ∈ M is denoted by [γ].

Definition 4.2. 20 At a p ∈ M define the tangent space Tp M as the space of all

equivalence classes [γ] of curves γ through p. A tangent vector Xp , as the equiva-
lence class of curves, is given by

Xp := [γ] = γ̃ : γ̃(0) = γ(0) = p, (ϕ ◦ γ̃)0 (0) = (ϕ ◦ γ)0 (0) ,



which is an element of Tp M.

M
p
γ

N 0

ϕ
ϕ◦γ

x (ϕ ◦ γ)′ (0)
Rm

F IGURE 26. Immersed curves and velocity vectors in Rm .

20 Lee, Ch. 3.
 33

The above definition does not depend on the choice of charts at p ∈ M. Let
(U 0 , ϕ0 ) be another chart at p ∈ M. Then, using that (ϕ ◦ γ̃)0 (0) = (ϕ ◦ γ)0 (0), for
(ϕ0 ◦ γ)0 (0) we have
h i0
(ϕ0 ◦ γ)0 (0) = (ϕ0 ◦ ϕ−1 ) ◦ (ϕ ◦ γ) (0)
= J(ϕ0 ◦ ϕ−1 ) x (ϕ ◦ γ)0 (0)
= J(ϕ0 ◦ ϕ−1 ) x (ϕ ◦ γ̃)0 (0)
h i0
= (ϕ0 ◦ ϕ−1 ) ◦ (ϕ ◦ γ̃) (0) = (ϕ0 ◦ γ̃)0 (0),
which proves that the equivalence relation does not depend on the particular choice
of charts at p ∈ M.
One can prove that Tp M ∼ = Rm . Indeed, Tp M can be given a linear structure as
follows; given two equivalence classes [γ1 ] and [γ2 ], then
[γ1 ] + [γ2 ] := γ : (ϕ ◦ γ)0 (0) = (ϕ ◦ γ1 )0 (0) + (ϕ ◦ γ2 )0 (0) ,


λ[γ1 ] := γ : (ϕ ◦ γ)0 (0) = λ(ϕ ◦ γ1 )0 (0) .



The above argument shows that these operation are well-defined, i.e. indepen-
dent of the chosen chart at p ∈ M, and the operations yield non-empty equivalence
classes. The mapping
τϕ : Tp M → Rm , τϕ [γ] = (ϕ ◦ γ)0 (0),

is a linear isomorphism and τϕ0 = J(ϕ0 ◦ ϕ−1 )|x ◦ τϕ . Indeed, by considering curves
γi (x) = ϕ−1 (x + tei ), i = 1, ..., m, it follows that [γi ] 6= [γ j ], i 6= j, since
(ϕ ◦ γi )0 (0) = ei 6= e j = (ϕ ◦ γ j )0 (0).
This proves the surjectivity of τϕ . As for injectivity one argues as follows. Suppose,
(ϕ ◦ γ)0 (0) = (ϕ ◦ eγ)0 (0), then by definition [γ] = [eγ], proving injectivity.
Given a smooth mapping f : N → M we can define how tangent vectors in Tp N
are mapped to tangent vectors in Tq M, with q = f (p). Choose charts (U, ϕ) for
p ∈ N, and (V, ψ) for q ∈ M. We define the tangent map or pushforward of f as
follows, see Figure 27. For a given tangent vector Xp = [γ] ∈ Tp N,
d f p = f∗ : Tp N → Tq M, f∗ ([γ]) = [ f ◦ γ].
The following commutative diagram shows that f∗ is a linear map and its definition
does not depend on the charts chosen at p ∈ N, or q ∈ M.
f∗
Tp N −−−−→ Tq M
 
τϕ 
y
τψ
y
J(ψ◦ f ◦ϕ−1 )|x
Rn −−−−−−−→ Rm
34

R
h◦f h

Tp N Xp f Tq M
p f∗ Xp q

N M

ϕ
ψ
f˜

x y
n
R Rm
F IGURE 27. Tangent vectors in Xp ∈ Tp N yield tangent vectors
f∗ Xp ∈ Tq M under the pushforward of f .

Indeed, a velocity vector (ϕ ◦ γ)0 (0) is mapped to (ψ ◦ f (γ))0 (0), and

(ψ ◦ f (γ))0 (0) = (ψ ◦ f ◦ ϕ−1 ◦ ϕ ◦ γ)0 (0) = J f˜ x ·(ϕ ◦ γ)0 (0).

Clearly, this mapping is linear and independent of the charts chosen.

If we apply the definition of pushforward to the coordinate mapping ϕ : N → Rn ,
then τϕ can be identified with ϕ∗ , and J(ψ ◦ f ◦ ϕ−1 )|x with (ψ ◦ f ◦ ϕ−1 )∗ . Indeed,
τϕ ([γ]) = (ϕ ◦ γ)0 (0) and ϕ∗ ([γ]) = [ϕ ◦ γ], and in Rn the equivalence class can be
labeled by (ϕ ◦ γ)0 (0). The labeling map is given as follows

τid ([ϕ ◦ γ]) = (ϕ ◦ γ)0 (0),

and is an isomorphism, and satisfies the relations

τid ◦ ϕ∗ = τϕ , ϕ∗ = τ−1
id ◦ τϕ .

From now one we identify Tx Rn with Rn by identifying ϕ∗ and τϕ . This justifies

the notation
ϕ∗ ([γ]) = [ϕ ◦ γ] := (ϕ ◦ γ)0 (0).

Properties of the pushforward can be summarized as follows:

 35

21
Lemma 4.5. Let f : N → M, and g : M → P be smooth mappings, and let p ∈ M,
then
(i) f∗ : Tp N → T f (p) M, and g∗ : T f (p) M → T(g◦ f )(p) P are linear maps (homo-
morphisms),
(ii) (g ◦ f )∗ = g∗ · f∗ : Tp N → T(g◦ f )(p) P,
(iii) (id)∗ = id : Tp N → Tp N,
(iv) if f is a diffeomorphism, then the pushforward f∗ is a isomorphism from
Tp N to T f (p) M.
Proof: We have that f∗ ([γ]) = [ f ◦ γ], and g∗ ([ f ◦ γ]) = [g ◦ f ◦ γ], which defines
the mapping (g ◦ f )∗ ([γ]) = [g ◦ f ◦ γ]. Now
[g ◦ f ◦ γ] = [g ◦ ( f ◦ γ)] = g∗ ([ f ◦ γ]) = g∗ ( f∗ ([γ])),
which shows that (g ◦ f )∗ = g∗ · f∗ .
A parametrization ϕ−1 : Rm → M coming from a chart (U, ϕ) is a local dif-
feomorphism, and can be used to find a canonical basis for Tp M. Choosing local
coordinates x = (x1 , · · · , xn ) = ϕ(p), and the standard basis vectors ei for Rm , we
define
∂
:= ϕ−1
∗ (ei ).
∂xi p
By definition ∂x∂ i p ∈ Tp M, and since the vectors ei form a basis for Rm , the vectors
∂
∂xi p form a basis for Tp M. An arbitrary tangent vector Xp ∈ Tp M can now be
written with respect to the basis { ∂x∂ i p }:
∂
Xp = ϕ−1
∗ (Xi ei ) = Xi .
∂xi p
where the notation Xi ∂x∂ i p = ∑i Xi ∂x∂ i p denotes the Einstein summation convention,
and (Xi ) is a vector in Rm !
We now define the directional derivative of a smooth function h : M → R in the
direction of Xp ∈ Tp M by
Xp h := h∗ Xp = [h ◦ γ].
In fact we have that Xp h = (h ◦ γ)0 (0), with Xp = [γ]. For the basis vectors of Tp M
this yields the following. Let γi (t) = ϕ−1 (x + tei ), and Xp = ∂
∂xi p , then

∂
0 ∂h̃
(2) Xp h = h = (h ◦ ϕ−1 ) ◦ (ϕ ◦ γi ) (0) = ,
∂xi p ∂xi
in local coordinates, which explains the notation for tangent vectors. In particular,
∂h̃
for general tangent vectors Xp , Xp h = Xi ∂x i
.
21Lee, Lemma 3.5.
36

Let go back to the curve γ : N = (a, b) → M and express velocity vectors. Con-
sider the chart (N, id) for N, with coordinates t, then
d
:= id−1
∗ (1).
dt t=0

We have the following commuting diagrams:

γ γ∗
N −−−−→ M Tt N −−−−→ Tp M
   
 ϕ 
id∗ y
ϕ
idy y y ∗
γ̃=ϕ◦γ γ̃∗
R −−−−→ Rm R −−−−→ Rm
We now define
d 
γ0 (0) = γ∗ = ϕ−1 0


∗ (ϕ ◦ γ) (0) ∈ Tp M,
dt t=0

by using the second commuting diagram.

If we take a closer look at Figure ?? we can see that using different charts at
p ∈ M, or equivalently, considering a change of coordinates leads to the following
relation. For the charts (U, ϕ) and (U 0 , ϕ0 ) we have local coordinates x = ϕ(p)
and x0 = ϕ0 (p), and p ∈ U ∩ U 0 . This yields two different basis for Tp M, namely
∂ ∂
∂xi p , and ∂xi0 p . Consider the identity mapping f = id : M → M, and the push-
forward yields the identity on Tp M. If we use the different choices of coordinates
as described above we obtain
 ∂  ∂h̃ ∂x0j ∂h̃
id∗ h = (ϕ0 ◦ ϕ−1 )∗ 0 = .
∂xi p ∂x j ∂xi ∂x0j

In terms of the different basis for Tp M this gives

∂ ∂x0j ∂
= .
∂xi p ∂xi ∂x0j p

Let us prove this formula by looking a slightly more general situation. Let N, M
be smooth manifolds and f : N → M a smooth mapping. Let q = f (p) and (V, ψ)
is a chart for M containing q. The vectors ∂y∂ j q form a basis for Tq M. If we write
Xp = Xi ∂x∂ i p , then for the basis vectors ∂
∂xi p we have
 ∂  ∂ ∂ g
∂
h ◦ f ◦ ϕ−1


f∗ h = (h ◦ f ) = h◦ f =
∂xi p ∂xi p ∂xi ∂xi
∂ ∂
h̃ ◦ ψ ◦ f ◦ ϕ−1 = h̃ ◦ f˜



=
∂xi ∂xi
∂h̃ ∂ f˜j  ∂ f˜j ∂ 
= = h
∂y j ∂xi ∂xi ∂y j q
 37

which implies that Tp N is mapped to Tq M under the map f∗ . In general a tangent

vector Xi ∂x∂ i p is pushed forward to a tangent vector
" #
∂ ∂ f˜j ∂
(3) Yj = Xi ∈ Tq M,
∂y j q ∂xi ∂y j q

expressed in local coordinates. By taking N = M and f = id we obtain the above

change of variables formula.

5. Cotangent spaces
In linear algebra it is often useful to study the space of linear functions on a
given vector space V . This space is denoted by V ∗ and called the dual vector space
to V — again a linear vector space. So the elements of V ∗ are linear functions
θ : V → R. As opposed to vectors v ∈ V , the elements, or vectors in V ∗ are called
covectors.
Lemma 5.1. Let V be a n-dimensional vector space with basis {v1 , · · · , vn }, then
the there exist covectors {θ1 , · · · , θn } such that

θi · v j := θi (v j ) = δij , Kronecker delta,

and the covectors {θ1 , · · · , θn } form a basis for V ∗ .

This procedure can also be applied to the tangent spaces Tp M described in the
previous chapter.

Definition 5.2. Let M be a smooth m-dimensional manifold, and let Tp M be the

tangent space at some p ∈ M. The the cotangent space Tp∗ M is defined as the dual
vector space of Tp M, i.e.
Tp∗ M := (Tp M)∗ .

By definition the cotangent space Tp∗ M is also m-dimensional and it has a canon-
ical basis as described in Lemma 5.1. As before have the canonical basis vectors
∂ ∗ i
∂xi | p for Tp M, the associated basis vectors for Tp M are denoted dx | p . Let us now
∗
describe this dual basis for Tp M and explain the notation.
The covectors dxi | p are called differentials and we show now that these are
indeed related dh p . Let h : M → R be a smooth function, then h∗ : Tp M → R
and h∗ ∈ Tp∗ M. Since the differentials dxi | p form a basis of Tp∗ M we have that
h∗ = λi dxi | p , and therefore

∂ ∂ j ∂h̃
h∗ = λ j dx j | p · = λ j δi = λi = ,
∂xi p ∂xi p ∂xi
38

and thus
∂h̃ i
(4) dh p = h∗ = dx p .
∂xi
Choose h such that h∗ satisfies the identity in Lemma 5.1, i.e. let h̃ = xi ( h = xi ◦ ϕ
= hϕ, ei i). These linear functions h of course span Tp∗ M, and
h∗ = (xi ◦ ϕ)∗ = d(xi ◦ ϕ) p = dxi | p .
Cotangent vectors are of the form
θ p = θi dxi p .
The pairing between a tangent vector Xp and a cotangent vector θ p is expressed
component wise as follows:
θ p · Xp = θi X j δij = θi Xi .
In the case of tangent spaces a mapping f : N → M pushes forward to a linear
f∗ ; Tp N → Tq M for each p ∈ N. For cotangent spaces one expects a similar con-
struction. Let q = f (p), then for a given cotangent vector θq ∈ Tq∗ M define
( f ∗ θq ) · Xp = θq · ( f∗ Xp ) ∈ Tp∗ N,
for any tangent vector Xp ∈ Tp N. The homomorphism f ∗ : Tq∗ M → Tp∗ N, defined by
θq 7→ f ∗ θq is called the pullback of f at p. It is a straightforward consequence from
linear algebra that f ∗ defined above is indeed the dual homomorphism of f∗ , also
called the adjoint, or transpose (see Lee, Ch. 6, for more details, and compare the
the definition of the transpose of a matrix). If we expand the definition of pullback
in local coordinates, using (3), we obtain
  ∂  ∂ 
f ∗ dy j q · Xi = dy j q · f∗ Xi
∂xi p ∂xi p
" #
j ∂ f˜j ∂ ∂ f˜j
= dy q Xi = Xi
∂xi ∂y j q ∂xi
Using this relation we obtain that

∗ j j
 ∂ ∂ f˜j ∂ f˜j i ∂
f σ dy q · Xi = σj Xi = σ j dx p Xi ,
∂xi p ∂xi ∂xi ∂xi p
which produces the local formula
" # " #
∂ ˜j
f ∂ ˜j
f
(5) f ∗ σ j dy j q = σ j dxi p = σ j ◦ f˜ dxi p .
∂xi ∂xi
x
Lemma 5.3. Let f : N → M, and g : M → P be smooth mappings, and let p ∈ M,
then
 39

(i) f ∗ : T f∗(p) M → Tp∗ N, and g∗ : T(g◦ ∗ ∗

f )(p) P → T f (p) M are linear maps (homo-
morphisms),
(ii) (g ◦ f )∗ = f ∗ · g∗ : T(g◦
∗ ∗
f )(p) P → Tp N,
∗ ∗
(iii) (id) = id : Tp N → Tp N, ∗

(iv) if f is a diffeomorphism, then the pullback f ∗ is a isomorphism from T f∗(p) M

to Tp∗ N.
Proof: By definition of the pullbacks of f and g we have

f ∗ θq · Xp = θq · f∗ (Xp ), g∗ ωg(q)Yq = ωg(q) · g∗ (Yq ).

For the composition g◦ f it holds that (g◦ f )∗ ω(g◦ f )(p) ·Xp = ω(g◦ f )(p) ·(g◦ f )∗ (Xp ).
Using Lemma 4.5 we obtain

(g ◦ f )∗ ω(g◦ f )(p) · Xp = ω(g◦ f )(p) · g∗ f∗ (Xp )

= g∗ ω(g◦ f )(p) · f∗ (Xp ) = f ∗ g∗ ω(g◦ f )(p) · Xp ,

which proves that (g ◦ f )∗ = f ∗ · g∗ .

Now consider the coordinate mapping ϕ : U ⊂ M → Rm , which is local diffeo-

morphism. Using Lemma 5.3 we then obtain an isomorphism ϕ∗ : Rm → Tp∗ M,
which justifies the notation
dxi | p = ϕ∗ (ei ).

6. Vector bundles
The abstract notion of vector bundle consists of topological spaces E (total
space) and M (the base) and a projection π : E → M (surjective). To be more
precise:

Definition 6.1. A triple (E, M, π) is called a real vector bundle of rank k over M if
(i) for each p ∈ M, the set E p = π−1 (p) is a real k-dimensional linear vector
space, called the fiber over p, such that
(ii) for every p ∈ M there exists a open neighborhood U 3 p, and a homeomor-
phism Φ : π−1 (U) → U × Rk ;
(a) π ◦ Φ−1 (p, ξ) = p, for all ξ ∈ Rk ;

(b) ξ 7→ Φ−1 (p, ξ) is a vector space isomorphism between Rk and E p .

The homeomorphism Φ is called a local trivialization of the bundle.
40

F IGURE 28. Charts in a vector bundle E over M.

If there is no ambiguity about the base space M we often denote a vector bundle
by E for short. Another way to denote a vector bundle that is common in the
literature is π : E → M. It is clear from the above definition that if M is a topological
manifold then so is E. Indeed, via the homeomorphisms Φ it follows that E


is Hausdorff and has a countable basis of open set. Define ϕ e = ϕ × IdRk ◦ Φ,
and ϕ e : π−1 (U) → V × Rk is a homeomorphism. Figure 28 explains the choice of
bundle charts for E related to charts for M.
For two trivializations Φ : π−1 (U) → U ×Rk and Ψ : π−1 (V ) → V ×Rk , we have
that the transition map Ψ ◦ Φ−1 : (U ∩V ) × Rk → (U ∩V ) × Rk has the following
form
Ψ ◦ Φ−1 (p, ξ) = (p, τ(p)ξ)),

where τ : U ∩ V → Gl(k, R) is continuous. It is clear from the assumptions that

Ψ ◦ Φ−1 (p, ξ) = (p, σ(p, ξ)). By assumption we also have that ξ 7→ Φ−1 (p, ξ =
A(p)ξ, and ξ 7→ Ψ−1 (p, ξ) = B(p)ξ. Now,

Ψ ◦ Φ−1 (p, ξ) = Ψ(A(p)ξ) = (p, σ(p, ξ)),

and Ψ−1 (p, σ) = A(p)ξ = B(p)σ. Therefore σ(p, ξ) = B−1 Aξ =: τ(p)ξ. Continuity
is clear from the assumptions in Definition 6.1.
If both E and M are smooth manifolds and π is a smooth projection, such that the
local trivializations can be chosen to be diffeomorphisms, then (E, M, π) is called
a smooth vector bundle. In this case the maps τ are smooth. The following result
allows us to construct smooth vector bundles and is important for the special vector
bundles used in this course.
 41

F IGURE 29. Transition mappings for local trivializations.

Theorem 6.4. 22 Let M be a smooth manifold, and let {E p } p∈M be a family of k-

dimensional real vector spaces parametrized by p ∈ M. Define E = p∈M E p , and
F

π : E → M as the mapping that maps E p to p ∈ M. Assume there exists

(i) an open covering {Uα }α∈A for M;
(ii) for each α ∈ A, a bijection Φα : π−1 (Uα ) → Uα × Rk , such that ξ →7
−1 k
Φα (p, ξ) is a vector space isomorphism between R and E p ;
(iii) for each α, β ∈ I, with Uα ∩ Uβ 6= ∅, smooth mappings ταβ : Uα ∩ Uβ →
Gl(k, R) such that

Φβ ◦ Φ−1
α (p, ξ) = (p, ταβ (p)ξ).

Then E has a unique differentiable (manifold) structure making (E, M, π) a smooth

vector bundle of rank k over M.
Proof: The proof of this theorem goes by verifying the hypotheses in Theorem
2.11. Let us propose charts for E. Let (Vp , ϕ p ), Vp ⊂ Uα , be a smooth chart for
M containing p. As before define ϕ e p : π−1 (Vp ) → Vep × Rk . We show now, using
Theorem 2.11, that (π−1 (Vp ), ϕ
e p ), p ∈ M are smooth charts, which gives a unique
differentiable manifold structure.

22See Lee, Lemma 5.5.

42

J 6.5 Example. Consider the set E defined as


E = (p1 , p2 , ξ1 , ξ2 ) | p1 = cos(θ), p2 = sin(θ), cos(θ/2)ξ1 + sin(θ/2)ξ2 = 0 .

Clearly, E is a smooth vector bundle over S1 of rank 1, embedded in R4 . For

example if U = S1 \{(1, 0)} (θ ∈ (0, 2π)), then Φ(π−1 (U)) = (p1 , p2 , ξ1 ) is a local
trivialization. This bundle is called the Möbius strip.
To show that E is a smooth vector bundle we need to verify the conditions in
Theorem 6.4. Consider a second chart U 0 = S1 \{(−1, 0)} (θ ∈ (−π, π)), and the
trivialization Ψ(π−1 (U 0 )) = (p1 , p2 , ξ2 ). For the Φ we have that
 cos(θ/2) 
Φ−1 (p1 , p2 , ξ1 ) = p1 , p2 , ξ1 , − ξ1 ,
sin(θ/2)
and thus
 cos(θ/2) 
Ψ ◦ Φ−1 (p1 , p2 , ξ1 ) = p1 , p2 , − ξ1 ,
sin(θ/2)
which gives that τ(p) is represented by τ = − cos(θ/2)
sin(θ/2) , for θ ∈ (0, π), which invert-
ible and smooth in p. I

Mappings between smooth vector bundles are called bundle maps, and are de-
fined as follows. Given vector bundles π : E → M and π0 : E 0 → M 0 and smooth
mappings F : E → E 0 and f : M → M 0 , such that the following diagram commutes
F
E −−−−→ E0
 

πy
 0
yπ
f
M −−−−→ M 0
and F|E p : E p → E 0f (p) is linear, then the pair (F, f ) is a called a smooth bundle
map.
A section, or cross section of a bundle E is a continuous mapping σ : M → E,
such that π ◦ σ = IdM . A section is smooth if σ is a smooth mapping. The space of

F IGURE 30. A cross section σ in a bundle E.

 43

smooth section in E is denoted by E(M). The zero section is a mapping σ : M → E

such that σ(p) = 0 ∈ E p for all p ∈ M.
J 6.7 Remark. By identifying M with the trivial bundle E0 = M × {0}, a section is
a bundle map from E0 = M × {0} to E. Indeed, define σ0 : E 0 → E by σ0 (p, 0) =
σ(p), and let f = IdM . Then π ◦ σ0 = π ◦ σ = IdM . I

6.1. The tangent bundle and vector fields

The disjoint union of tangent spaces
G
T M := Tp M
p∈M

is called the tangent bundle of M. We show now that T M is a fact a smooth vector
bundle over M.
Theorem 6.8. The tangent bundle T M is smooth vector bundle over M of rank m,
and as such T M is a smooth 2m-dimensional manifold.
Proof: Let (U, ϕ) be a smooth chart for p ∈ M. Define
 ∂ 

Φ Xi = p, (Xi ) ,
∂xi p

which clearly is a bijective map between π−1 (U) and U × Rm . Moreover, Xp =

(Xi ) 7→ Φ−1 (p, Xp ) = Xi ∂x∂ i is vector space isomorphism. Let (Uα , ϕα ) be a cov-
p
ering of smooth charts of M. Let x = ϕα (p), and x0 = ϕβ (p). From our previous
considerations we easily see that

Φβ ◦ Φ−1
α (p, X p ) = (p, ταβ (p)X p ),

where ταβ : Uα ∩Uβ → Gl(m, R). It remains to show that ταβ is smooth. We have
e−1
eβ ◦ ϕ
ϕ −1 −1
α = (ϕβ × Id) ◦ Φβ ◦ Φα ◦ (ϕα × Id). Using the change of coordinates
formula derived in Section 4 we have that
!
 ∂x0 
j
e−1 0 0


eβ ◦ ϕ
ϕ α x, (Xi ) = x1 (x), · · · , xm (x), Xj ,
∂xi

which proves the smoothness of ταβ . Theorem 6.4 can be applied now showing that
T M is asmooth vector bundle over M. From the charts (π−1 (U), ϕ e) we conclude
that T M is a smooth 2m-dimensional manifold.
44

Definition 6.9. A smooth (tangent) vector field is a smooth mapping

X : M → T M,

with the property that π ◦ X = idM . In other words X is a smooth (cross) section in
the vector bundle T M, see Figure 31. The space of smooth vector fields on M is
denoted by F(M).

Tp M Xp Xp
TM

p M
p
M
X
F IGURE 31. A smooth vector field X on a manifold M [right],
and as a ‘curve’, or section in the vector bundle T M [left].

For a chart (U, ϕ) a vector field X can be expressed as follows

∂
X = Xi ,
∂xi p
where Xi : U → R. Smoothness of vector fields can be described in terms of the
component functions Xi .
Lemma 6.11. A mapping X : M → T M is a smooth vector field at p ∈ U if and
only if the coordinate functions Xi : U → R are smooth.
Proof: In standard coordinates X is given by
e = (x1 , · · · , xm , Xe1 (x), · · · , Xem (x)),
X(x)

which immediately shows that smoothness of X is equivalent to the smoothness of

the coordinate functions Xi .

In Section 4 we introduced the notion of push forward of a mapping f : N → M.

Under submersions and immersion a vector field X : N → T N does not necessarily
push forward to a vector field on M. If f is a diffeomorphism, then f∗ X = Y is a
vector field on M. We remark that the definition of f∗ also allows use to restate
definitions about the rank of a map in terms of the differential, or pushforward f∗
at a point p ∈ N.
 45

6.2. The cotangent bundle and differential 1-forms

The disjoint union
T ∗ M := Tp∗ M
G

p∈M
is called cotangent bundle of M.
Theorem 6.12. 23 The cotangent bundle T ∗ M is a smooth vector bundle over M of
rank m, and as such T ∗ M is a smooth 2m-dimensional manifold.
Proof: The proof is more identical to the proof for T M, and is left to the reader
as an exercise.
The differential dh : M → T ∗ M is an example of a smooth function. The above
consideration give the coordinate wise expression for dh.

Definition 6.13. A smooth covector field is a smooth mapping

θ : M → T ∗ M,
with the property that π ◦ θ = idM . In order words θ is a smooth section in T ∗ M.
The space of smooth covector fields on M is denoted by F∗ (M). Usually the space
F∗ (M) is denoted by Γ1 (M), and covector fields are referred to a (smooth) differ-
ential 1-form on M.

For a chart (U, ϕ) a covector field θ can be expressed as follows

θ = θi dxi p ,

where θi : U → R. Smoothness of a covector fields can be described in terms of

the component functions θi .
Lemma 6.14. A covector field θ is smooth at p ∈ U if and only if the coordinate
functions θi : U → R are smooth.
Proof: See proof of Lemma 6.11.
We saw in the previous section that for arbitrary mappings f : N → M, a vector
field X ∈ F(N) does not necessarily push forward to a vector on M under f∗ . The
reason is that surjectivity and injectivity are both needed to guarantee this, which
requires f to be a diffeomorphism. In the case of covector fields or differential 1-
forms the situation is completely opposite, because the pullback acts in the opposite
direction and is therefore onto the target space and uniquely defined. To be more
precise; given a 1-form θ ∈ Λ1 (M) we define a 1-form f ∗ θ ∈ Λ1 (N) as follows
( f ∗ θ) p = f ∗ θ f (p) .

23See Lee, Proposition 6.5.

46

Theorem 6.15. 24 The above defined pullback f ∗ θ of θ under a smooth mapping

f : N → M is a smooth covector field, or differential 1-form on N.
Proof: Write θ ∈ Λ1 (M) is local coordinates; θ = θi dyi q . By Formula (5) we
then obtain
" # " #
∂ ˜
f i ∂ ˜
f i
f ∗ θi dyi f (p) = θi y dx j p = θi ◦ f˜ dx j p ,
∂x j ∂x j
x

which proves that f ∗θ ∈ Λ1 (N).

Given a mapping g : M → R, the differential, or push forward g∗ = dg defines

g
an element in Tp∗ M. In local coordinates we have dg p = ∂x
∂e
i dxi p , and thus defines
1
a smooth covector field on M; dg ∈ Λ (M). Given a smooth mapping f : N → M,
it follows from (5) that
g i
∂e g e∂ fei j
∂e
f ∗ dg = f ∗ dx f (p)
= ◦f dy p .
∂xi ∂xi ∂x j
By applying the same to the 1-form d(g ◦ f ) we obtain the following identities

(6) f ∗ dg = d(g ◦ f ), f ∗ (gθ) = (g ◦ f ) f ∗ θ.

Using the formulas in (6) we can also obtain (5) in a rather straightforward way.
Let g = ψ j = hψ, e j i = y j , and ω = dg = dy j |q in local coordinates, then
h ∂ f˜j i i
f ∗ (σ j ◦ ψ)ω = σ j ◦ ψ ◦ f f ∗ dg = σ j ◦ ψ ◦ f d(g ◦ f ) = σ j ◦ f˜




dx p ,
∂xi x
where the last step follows from (4).

Definition 6.16. A differential 1-form θ ∈ Λ1 (N) is called an exact 1-form if there

exists a smooth function g : N → R, such that θ = dg.

The notation for tangent vectors was motivated by the fact that functions on a
manifold can be differentiated in tangent directions. The notation for the cotangent
vectors was partly motivated as the ‘reciprocal’ of the partial derivative. The in-
troduction of line integral will give an even better motivation for the notation for
cotangent vectors. Let N = R, and θ a 1-form on N given in local coordinates by
θt = h(t)dt, which can be identified with a function h. The notation makes sense
because θ can be integrated over any interval [a, b] ⊂ R:
Z Z b
θ := h(t)dt.
[a,b] a

24Lee, Proposition 6.13.

 47

Let M = R, and consider a mapping f : M = R → N = R, which satisfies f 0 (t) > 0.

Then t = f (s) is an appropriate change of variables. Let [c, d] = f ([a, b]), then
Z Z d Z b Z
f ∗θ = h( f (s)) f 0 (s)ds = h(t)dt = θ,
[c,d] c a [a,b]
which is the change of variables formula for integrals. We can use this now to
define the line integral over a curve γ on a manifold N.

Definition 6.17. Let γ : [a, b] ⊂ R → N, and let θ be a 1-form on N and γ∗ θ the

pullback of θ, which is a 1-form on R. Denote the image of γ in N also by γ, then
Z Z Z
θ := γ∗ θ = θi (γ(t))γ0i (t)dt, 25
γ [a,b] [a,b]
the expression in local coordinates.

The latter can be seen by combining some of the notion introduced above:
d d
γ∗ θ · = (γ∗ θ)t = θγ(t) · γ∗ = θγ (t) · γ0 (t).
dt dt
Therefore, γ θ = (γ θ)t dt = θγ(t) · γ (t)dt = θ (γ(t))γ0i (t)dt, and
∗ ∗ 0 i
Z Z Z Z
∗ 0
θ= γ θ= θγ(t) · γ (t)dt = θi (γ(t))γ0i (t)dt.
γ [a,b] [a,b] [a,b]

γ0
If is nowhere zero then the map γ : [a, b] → N is either an immersion or em-
bedding. For example in the embedded case this gives an embedded submanifold
γ ⊂ N with boundary ∂γ = {γ(a), γ(b)}. Let θ = dg be an exact 1-form, then
Z
dg = g ∂γ
= g(γ(b)) − g(γ(a)).
γ
Indeed,
Z Z Z Z b
dg = γ∗ dg = d(g ◦ γ) = (g ◦ γ)0 (t)dt = g(γ(b)) − g(γ(a)).
γ [a,b] [a,b] a
This identity is called the Fundamental Theorem for Line Integrals and is a spe-
cial case of the Stokes Theorem (see Section 16).

25The expressions γ (t) ∈ R are the components of γ0 (t) ∈ T M.

i γ(t)
 48

III. Tensors and differential

forms

7. Tensors and tensor products

In the previous chapter we encountered linear functions on vector spaces, linear
functions on tangent spaces to be precise. In this chapter we extend to notion of
linear functions on vector spaces to multilinear functions.

Definition 7.1. Let V1 , · · · ,Vr , and W be real vector spaces. A mapping T : V1 ×

· · · ×Vr → W is called a multilinear mapping if

T (v1 , · · · , λvi + µv0i , · · · , vr ) = λT (v1 , · · · , vi , · · · vr ) + µT (v1 , · · · , v0i , · · · vr ), ∀i,

and for all λ, µ ∈ R i.e. f is linear in each variable vi separately.

Now consider the special case that W = R, then T becomes a multilinear func-
tion, or form, and a generalization of linear functions. If in addition V1 = · · · =
Vr = V , then
T : V × · · · ×V → R,
is a multilinear function on V , and is called a covariant r-tensor on V . The number
of copies r is called the rank of T . The space of covariant r-tensors on V is denoted
by T r (V ), which clearly is a real vector space using the multilinearity property in
Definition 7.1. In particular we have that T 0 (V ) ∼= R, T 1 (V ) = V ∗ , and T 2 (V ) is
the space of bilinear forms on V . If we consider the case V1 = · · · = Vr = V ∗ , then

T : V ∗ × · · · ×V ∗ → R,

is a multilinear function on V ∗ , and is called a contravariant r-tensor on V . The

space of contravariant r-tensors on V is denoted by Tr (V ). Here we have that
T0 (V ) ∼
= R, and T1 (V ) = (V ∗ )∗ ∼
= V.
J 7.2 Example. The cross product on R3 is an example of a multilinear (bilinear)
function mapping not to R to R3 . Let x, y ∈ R3 , then

T (x, y) = x × y ∈ R3 ,

which clearly is a bilinear function on R3 . I

 49

Since multilinear functions on V can be multiplied, i.e. given vector spaces V,W
and tensors T ∈ T r (V ), and S ∈ T s (W ), the multilinear function
R(v1 , · · · vr , w1 , · · · , ws ) = T (v1 , · · · vr )S(w1 , · · · , ws )
is well defined and is a multilnear function on V r × W s . This brings us to the
following definition. Let T ∈ T r (V ), and S ∈ T s (W ), then
T ⊗ S : V r ×W s → R,
is given by
T ⊗ S(v1 , · · · , vr , w1 , · · · ws ) = T (v1 , · · · , vr )S(w1 , · · · ws ).
This product is called the tensor product. By taking V = W , T ⊗ S is a covariant
(r + s)-tensor on V , which is a element of the space T r+s (V ) and ⊗ : T r (V ) ×
T s (V ) → T r+s (V ).
Lemma 7.3. Let T ∈ T r (V ), S, S0 ∈ T s (V ), and R ∈ T t (V ), then
(i) (T ⊗ S) ⊗ R = T ⊗ (S ⊗ R) (associative),
(ii) T ⊗ (S + S0 ) = T ⊗ S + T ⊗ S0 (distributive),
(iii) T ⊗ S 6= S ⊗ T (non-commutative).
The tensor product is also defined for contravariant tensors and mixed tensors.
As a special case of the latter we also have the product between covariant and
contravariant tensors.
J 7.4 Example. The last property can easily be seen by the following example. Let
V = R2 , and T, S ∈ T 1 (R2 ), given by T (v) = v1 + v2 , and S(w) = w1 − w2 , then
T ⊗ S(1, 1, 1, 0) = 2 6= 0 = S ⊗ T (1, 1, 1, 0),
which shows that ⊗ is not commutative in general. I
The following theorem shows that the tensor product can be used to build the
tensor space T r (V ) from elementary building blocks.
Theorem 7.5. 26 Let {v1 , · · · , vn } be a basis for V , and let {θ1 , · · · , θn } be the dual
basis for V ∗ . Then the set
B = θi1 ⊗ · · · ⊗ θir : 1 ≤ i1 , · · · , ir ≤ n ,


is a basis for the nr -dimensional vector space T r (V ).

Proof: Compute
Ti1 ···ir θi1 ⊗ · · · ⊗ θir (v j1 , · · · , v jr ) = Ti1 ···ir θi1 (v j1 ) · · · θir (v jr )
j j
= Ti1 ···ir δ j11 · · · δ jrr = T j1 ··· jr
= T (v j1 , · · · , v jr ),
26See Lee, Prop. 11.2.
50

which shows by using the multilinearity of tensors that T can be expanded in the
basis B as follows;
T = Ti1 ···ir θi1 ⊗ · · · ⊗ θir ,
where Ti1 ···ir = T (vi1 , · · · , vir ), the components of the tensor T . Linear independence
follows form the same calculation.
J 7.6 Example. Consider the the 2-tensors T (x, y) = x1 y1 + x2 y2 , T 0 (x, y) = x1 y2 +
x2 y2 and T 00 = x1 y1 + x2 y2 + x1 y2 on R2 . With respect to the standard bases θ1 (x) =
x1 , θ2 (x) = x1 , and
θ1 ⊗ θ1 (x, y) = x1 y1 , θ1 ⊗ θ2 (x, y) = x1 y2 ,
θ1 ⊗ θ2 (x, y) = x2 y1 , and θ2 ⊗ θ2 (x, y) = x2 y2 .
Using this the components of T are given by T11 = 1, T12 = 0, T21 = 0, and T22 = 1.
Also notice that T 0 = S ⊗ S0 , where S(x) = x1 + x2 , and S0 (y) = y2 . Observe that
not every tensor T ∈ T 2 (R2 ) is of the form T = S ⊗ S0 . For example T 00 6= S ⊗ S0 ,
for any S, S0 ∈ T 1 (R2 ). I
In Lee, Ch. 11, the notion of tensor product between arbitrary vector spaces is
explained. Here we will discuss a simplified version of the abstract theory. Let
V and W be two (finite dimensional) real vector spaces, with bases {v1 , · · · vn }
and {w1 , · · · wm } respectively, and for their dual spaces V ∗ and W ∗ we have the
dual bases {θ1 , · · · , θn } and {σ1 , · · · , σm } respectively. If we use the identification
{V ∗ }∗ ∼
= V , and {W ∗ }∗ ∼ = W we can define V ⊗W as follows:

Definition 7.7. The tensor product of V and W is the real vector space of (finite)
linear combinations
n o h i
V ⊗W := λi j vi ⊗ w j : λi j ∈ R = vi ⊗ w j i, j ,

where vi ⊗ w j (v∗ , w∗ ) := v∗ (vi )w∗ (w j ), using the identification vi (v∗ ) := v∗ (vi ), and
w j (w∗ ) := w∗ (w j ), with (v∗ , w∗ ) ∈ V ∗ ×W ∗ .

To get a feeling of what the tensor product of two vector spaces represents con-
sider the tensor product of the dual spaces V ∗ and W ∗ . We obtain the real vector
space of (finite) linear combinations
n o h i
V ∗ ⊗W ∗ := λi j θi ⊗ σ j : λi j ∈ R = θi ⊗ σ j i, j ,

where θi ⊗ σ j (v, w) = θi (v)σ j (w) for any (v, w) ∈ V ×W . One can show that V ∗ ⊗
W ∗ is isomorphic to space of bilinear maps from V ×W to R. In particular elements
v∗ ⊗ w∗ all lie in V ∗ ⊗ W ∗ , but not all elements in V ∗ ⊗ W ∗ are of this form. The
isomorphism is easily seen as follows. Let v = ξi vi , and w = η j w j , then for a given
bilinear form b it holds that b(v, w) = ξi η j b(vi , w j ). By definition of dual basis
 51

we have that ξi η j = θi (v)σ j (w) = θi ⊗ σ j (v, w), which shows the isomorphism by
setting λi j = b(vi , w j ).
In the case V ∗ ⊗ W the tensors represent linear maps from V to W . Indeed,
from the previous we know that elements in V ∗ ⊗W represent bilinear maps from
V ×W ∗ to R. For an element b ∈ V ∗ ⊗W this means that b(v, ·) : W ∗ → R, and thus
b(v, ·) ∈ (W ∗ )∗ ∼
= W.
J 7.8 Example. Consider vectors a ∈ V and b∗ ∈ W , then a∗ ⊗ (b∗ )∗ can be iden-
tified with a matrix, i.e a∗ ⊗ (b∗ )∗ (v, ·) = a∗ (v)(b∗ )∗ (·) ∼
= a∗ (v)b. For example let
a∗ (v) = a1 v1 + a2 v2 + a3 v3 , and
 
! ! v
1
a1 b1 v1 + a2 b1 v2 + a3 b1 v3 a1 b1 a2 b1 a3 b1 
Av = a∗ (v)b = =  v2  .

a1 b2 v1 + a2 b2 v2 + a3 b2 v3 a1 b2 a2 b2 a3 b2
v3

Symbolically we can write

! !
∗ a1 b1 a2 b1 a3 b1   b1
A = a ⊗b = = a1 a2 a3 ⊗ ,
a1 b2 a2 b2 a3 b2 b2

which shows how a vector and covector can be ‘tensored’ to become a matrix. Note
that it also holds that A = (a · b∗ )∗ = b · a∗ . I

Lemma 7.9. We have that

(i) V ⊗W and W ⊗V are isomorphic;
(ii) (U ⊗V ) ⊗W and U ⊗ (V ⊗W ) are isomorphic.
With the notion of tensor product of vector spaces at hand we now conclude that
the above describe tensor spaces T r (V ) and Vr (V ) are given as follows;
∗
T r (V ) = V · · ⊗V }∗ ,
| ⊗ ·{z | ⊗ ·{z
Tr (V ) = V · · ⊗V} .
r times r times

By considering tensor products of V ’s and V ∗ ’s we obtain the tensor space of

mixed tensors;
∗
Tsr (V ) := V · · ⊗V }∗ ⊗V
| ⊗ ·{z | ⊗ ·{z
· · ⊗V} .
r times s times

Elements in this space are called (r, s)-mixed tensors on V — r copies of V ∗ , and s
copies of V . Of course the tensor product described above is defined in general for
0
tensors T ∈ Tsr (V ), and S ∈ Tsr0 (V ):
0 0
⊗ : Tsr (V ) × Tsr0 (V ) → Ts+s
r+r
0 (V ).
52

The analogue of Theorem 7.5 can also be established for mixed tensors. In the next
sections we will see various special classes of covariant, contravariant and mixed
tensors.
J 7.10 Example. The inner product on a vector space V is an example of a covariant
2-tensor. This is also an example of a symmetric tensor. I

J 7.11 Example. The determinant of n vectors in Rn is an example of covariant n-

tensor on Rn . The determinant is skew-symmetric, and an example of an alternating
tensor. I

If f : V → W is a linear mapping between vector spaces and T is an covariant

tensor on W we can define concept of pullback of T . Let T ∈ T r (W ), then f ∗ T ∈
T r (V ) is defined as follows:

f ∗ T (v1 , · · · vr ) = T ( f (v1 ), · · · , f (vr )),

and f ∗ : T r (W ) → T r (V ) is a linear mapping. Indeed, f ∗ (T + S) = T ◦ f + S ◦ f =

f ∗ T + f ∗ S, and f ∗ λT = λT ◦ f = λ f ∗ T . If we represent f by a matrix A with
respect to bases {vi } and {w j } for V and W respectively, then the matrix for the
linear f ∗ is given by
∗
· · ⊗ A}∗ ,
|A ⊗ ·{z
r times

with respect to the bases {θi1 ⊗· · ·⊗θir } and {σ j1 ⊗· · ·⊗σ jr } for T r (W ) and T r (V )
respectively.
J 7.12 Remark. The direct sum
∞
∗
M
T (V ) = T r (V ),
r=0

consisting of finite sums of covariant tensors is called the covariant tensor algebra
of V with multiplication given by the tensor product ⊗. Similarly, one defines the
contravariant tensor algebra
∞
M
T∗ (V ) = Tr (V ).
r=0

For mixed tensors we have

∞
M
T (V ) = Tsr (V ),
r,s=0

which is called the tensor algebra of mixed tensor of V . Clearly, T ∗ (V ) and T∗ (V )

subalgebras of T (V ). I
 53

8. Symmetric and alternating tensors

There are two special classes of tensors which play an important role in the
analysis of differentiable manifolds. The first class we describe are symmetric
tensors. We restrict here to covariant tensors.

Definition 8.1. A covariant r-tensor T on a vector space V is called symmetric if

T (v1 , · · · , vi , · · · , v j , · · · , vr ) = T (v1 , · · · , v j , · · · , vi , · · · , vr ),

for any pair of indices i ≤ j. The set of symmetric covariant r-tensors on V is

denoted by Σr (V ) ⊂ T r (V ), which is a (vector) subspace of T r (V ).

If a ∈ Sr is a permutation, then define

a
T (v1 , · · · , vr ) = T (va(1) , · · · , va(r) ),

where a({1, · · · , r}) = {a(1), · · · , a(r)}. From this notation we have that for two
permutations a, b ∈ Sr , b (a T ) = ba T . Define
1
Sym T = ∑ a T.
r! a∈Sr

It is straightforward to see that for any tensor T ∈ T r (V ), Sym T is a symmetric.

Moreover, a tensor T is symmetric if and only if Sym T = T . For that reason
Sym T is called the (tensor) symmetrization.
J 8.2 Example. Let T, T 0 ∈ T 2 (R2 ) be defined as follows: T (x, y) = x1 y2 , and
T 0 (x, y) = x1 y1 . Clearly, T is not symmetric and T 0 is. We have that
1 1
Sym T (x, y) = T (x, y) + T (y, x)
2 2
1 1
= x1 y2 + y1 x2 ,
2 2
which clearly is symmetric. If we do the same thing for T 0 we obtain:
1 0 1
Sym T 0 (x, y) =
T (x, y) + T 0 (y, x)
2 2
1 1
= x1 y1 + y1 x1 = T 0 (x, y),
2 2
showing that operation Sym applied to symmetric tensors produces the same ten-
sor again. I

Using symmetrization we can define the symmetric product. Let S ∈ Σr (V ) and

T ∈ Σs (V ) be symmetric tensors, then

S · T = Sym (S ⊗ T ).
54

The symmetric product of symmetric tensors is commutative which follows di-

rectly from the definition:
1
S · T (v1 , · · · , vr+s ) = ∑ S(va(1) , · · · , va(r) )T (va(r+1) , · · · , va(r+s) ).
(r + s)! a∈Sr+s

J 8.3 Example. Consider the 2-tensors S(x) = x1 + x2 , and T (y) = y2 . Now S ⊗

T (x, y) = x1 y2 + x2 y2 , and T ⊗ S(x, y) = x2 y1 + x2 y2 , which clearly gives that S ⊗
T 6= T ⊗ S. Now compute
1 1 1 1
Sym (S ⊗ T )(x, y) = x1 y2 + x2 y2 + y1 x2 + x2 y2
2 2 2 2
1 1
= x1 y2 + x2 y1 + x2 y2 = S · T (x, y).
2 2
Similarly,
1 1 1 1
Sym (T ⊗ S)(x, y) = x2 y1 + x2 y2 + y2 x1 + x2 y2
2 2 2 2
1 1
= x1 y2 + x2 y1 + x2 y2 = T · S(x, y),
2 2
which gives that S · T = T · S. I

Lemma 8.4. Let {v1 , · · · , vn } be a basis for V , and let {θ1 , · · · , θn } be the dual
basis for V ∗ . Then the set

BΣ = θi1 · · · θir : 1 ≤ i1 ≤ · · · ≤ ir ≤ n ,


the (sub)space Σr (V ) of symmetric r-tensors. Moreover, dim Σr (V ) =

is a basis for !
n+r−1
= (n+r−1)!
r!(n−1)! .
r
Proof: Proving that BΣ is a basis follows from Theorem 7.5, see also Lemma
8.10. It remains to establish the dimension of Σr (V ). Note that the elements in the
basis are given by multi-indices (i1 , · · · , ir ), satisfying the property that 1 ≤ i1 ≤
· · · ≤ ir ≤ n. This means choosing r integers satisfying this restriction. To do this
redefine jk := ik +k −1. The integers j range!from 1 through n+r −1. Now choose
n+r−1
r integers out of n + r − 1, i.e. combinations ( j1 , · · · , jr ), which are
r
in one-to-one correspondence with (i1 , · · · , ir ).

Another important class of tensors are alternating tensors and are defined as
follows.
 55

Definition 8.5. A covariant r-tensor T on a vector space V is called alternating if

T (v1 , · · · , vi , · · · , v j , · · · , vr ) = −T (v1 , · · · , v j , · · · , vi , · · · , vr ),
for any pair of indices i ≤ j. The set of alternating covariant r-tensors on V is
denoted by Λr (V ) ⊂ T r (V ), which is a (vector) subspace of T r (V ).

As before we define
1
Alt T = ∑ (−1)a a T,
r! a∈Sr

where (−1)a is +1 for even permutations, and −1 for odd permutations. We say
that Alt T is the alternating projection of a tensor T , and Alt T is of course a
alternating tensor.
J 8.6 Example. Let T, T 0 ∈ T 2 (R2 ) be defined as follows: T (x, y) = x1 y2 , and
T 0 (x, y) = x1 y2 − x2 y1 . Clearly, T is not alternating and T 0 (x, y) = −T 0 (y, x) is
alternating. We have that
1 1
Alt T (x, y) =T (x, y) − T (y, x)
2 2
1 1 1
= x1 y2 − y1 x2 = T 0 (x, y),
2 2 2
which clearly is alternating. If we do the same thing for T 0 we obtain:
1 0 1
Alt T 0 (x, y) = T (x, y) − T 0 (y, x)
2 2
1 1 1 1
= x1 y2 − x2 y1 − y1 x2 + y2 x1 = T 0 (x, y),
2 2 2 2
showing that operation Alt applied to alternating tensors produces the same tensor
again. Notice that T 0 (x, y) = det(x, y). I
This brings us to the fundamental product of alternating tensors called the wedge
product. Let S ∈ Λr (V ) and T ∈ Λs (V ) be symmetric tensors, then
(r + s)!
S∧T =Alt (S ⊗ T ).
r!s!
The wedge product of alternating tensors is anti-commutative which follows di-
rectly from the definition:
1
S ∧ T (v1 , · · · , vr+s ) = ∑ (−1)a S(va(1) , · · · , va(r) )T (va(r+1) , · · · , va(r+s) ).
r!s! a∈Sr+s

In the special case of the wedge of two covectors θ, ω ∈ V ∗ gives

θ ∧ ω = θ ⊗ ω − ω ⊗ θ.
In particular we have that
56

(i) (T ∧ S) ∧ R = T ∧ (S ∧ R);
(ii) (T + T 0 ) ∧ S = T ∧ S + T 0 ∧ S;
(iii) T ∧ S = (−1)rs S ∧ T , for T ∈ Λr (V ) and S ∈ Λs (V );
(iv) T ∧ T = 0.
The latter is a direct consequence of the definition of Alt . In order to prove these
properties we have the following lemma.
Lemma 8.7. Let T ∈ T r (V ) and S ∈ T s (V ), then

Alt (T ⊗ S) = Alt ((Alt T ) ⊗ S) = Alt (T ⊗ Alt S).

Proof: Let G ∼ = Sr be the subgroup of Sr+s consisting of permutations that only

permute the element {1, · · · , r}. For a ∈ G, we have a0 ∈ Sr . Now a (T ⊗ S) =
a0 T ⊗ S, and thus
1
∑ (−1)a a (T ⊗ S) = (Alt T ) ⊗ S.
r! a∈G
For the right cosets {ba : a ∈ G} we have
 
∑ (−1)ba ba (T ⊗ S) = (−1)b b ∑ (−1)a a
(T ⊗ S)
a∈G a∈G
 
b b
= r!(−1) (Alt T ) ⊗ S .
1
Taking the sum over all right cosets with the factor (r+s)! gives
1
Alt (T ⊗ S) =
(r + s)! ∑ ∑ (−1)ba ba (T ⊗ S)
b a∈G
r! b b
 
= (−1) (Alt T ) ⊗ S = Alt ((Alt T ) ⊗ S),
(r + s)! ∑
b

where the latter equality is due to the fact that r! terms are identical under the
definition of Alt ((Alt T ) ⊗ S).

Property (i) can now be proved as follows. Clearly Alt (Alt (T ⊗S)−T ⊗S) = 0,
and thus from Lemma 8.7 we have that

0 = Alt ((Alt (T ⊗ S) − T ⊗ S) ⊗ R) = Alt (Alt (T ⊗ S) ⊗ R) − Alt (T ⊗ S ⊗ R).

By definition
(r + s + t)!
(T ∧ S) ∧ R = Alt ((T ∧ S) ⊗ R)
(r + s)!t!
(r + s + t)!  (r + s)!  
= Alt Alt (T ⊗ S) ⊗ R
(r + s)!t! r!s!
(r + s + t)!
= Alt (T ⊗ S ⊗ R).
r!s!t!
 57

The same formula holds for T ∧ (S ∧ R), which prove associativity. More generally
it holds that for Ti ∈ Λri (V )
(r1 + · · · + rk )!
T1 ∧ · · · ∧ Tk = Alt (T1 ⊗ · · · ⊗ Tk ).
r1 ! · · · rk !
Property (iii) can be seen as follows. Each term in T ∧ S can be found in S ∧ T .
This can be done by linking the permutations a and a0 . To be more precise, how
many permutation of two elements are needed to change
a ↔ (i1 , · · · , ir , jr+1 , · · · , jr+s ) into a0 ↔ ( jr+1 , · · · , jr+s , i1 , · · · , ir ).
This clearly requires rs permutations of two elements, which shows Property (iii).
J 8.8 Example. Consider the 2-tensors S(x) = x1 + x2 , and T (y) = y2 . As before
S ⊗ T (x, y) = x1 y2 + x2 y2 , and T ⊗ S(x, y) = x2 y1 + x2 y2 . Now compute
1 1 1 1
Alt (S ⊗ T )(x, y) = x1 y2 + x2 y2 − y1 x2 − x2 y2
2 2 2 2
1 1  
= x1 y2 − x2 y1 = 2 S ∧ T (x, y) .
2 2
Similarly,
1 1 1 1
Alt (T ⊗ S)(x, y) = x2 y1 + x2 y2 − y2 x1 − x2 y2
2 2 2 2
1 1  
= − x1 y2 + x2 y1 = −2 T ∧ S(x, y) ,
2 2
which gives that S ∧ T = −T ∧ S. Note that if T = e∗1 , i.e. T (x) = x1 , and S = e∗2 ,
i.e. S(x) = x2 , then
T ∧ S(x, y) = x1 y2 − x2 y1 = det(x, y).
I
J 8.9 Remark. Some authors use the more logical definition
¯ = Alt (S ⊗ T ),
S∧T
which is in accordance with the definition of the symmetric product. This definition
is usually called the alt convention for the wedge product, and our definition is
usually referred to as the determinant convention. For computational purposes the
determinant convention is more appropriate. I
If {e∗1 , · · · , e∗n } is the standard dual basis for (Rn )∗ , then for vectors a1 , · · · , an ∈
Rn ,
det(a1 , · · · , an ) = e∗1 ∧ · · · ∧ e∗n (a1 , · · · , an ).
Using the multilinearity the more general statement reads
β1 ∧ · · · ∧ βn (a1 , · · · , an ) = det βi (a j ) ,


(7)
58

where βi are co-vectors.

The alternating tensor det = e∗1 ∧· · ·∧e∗n is called the determinant function on Rn .
If f : V → W is a linear map between vector spaces then the pullback f ∗ T ∈ Λr (V )
of any alternating tensor T ∈ Λr (W ) is given via the relation:
f ∗ T (v1 , · · · , vr ) = T f (v1 ), · · · , f (vr ) , f ∗ : Λr (W ) → Λr (V ).

In particular, f ∗ (T ∧ S) = ( f ∗ T ) ∧ f ∗ (S). As a special case we have that if f : V →

V , linear, and dimV = n, then
(8) f ∗ T = det( f )T,
for any alternating tensor T ∈ Λn (V ). This can be seen as follows. By multilinearity
we verify the above relation for the vectors {ei }. We have that
f ∗ T (e1 , · · · , en ) = T ( f (e1 ), · · · , f (en ))
= T ( f1 , · · · , fn ) = c det( f1 , · · · , fn ) = c det( f ),
where we use the fact that Λn (V ) ∼
= R (see below). On the other hand
det( f )T (e1 , · · · en ) = det( f )c · det(e1 , · · · , en )
= c det( f ),
which proves (8).
Lemma 8.10. Let {θ1 , · · · , θn } be a basis for V ∗ , then the set
BΛ = θi1 ∧ · · · ∧ θir : 1 ≤ i1 < · · · < ir ≤ n ,


n!
is a basis for Λr (V ), and dim Λr (V ) = (n−r)!r! . In particular, dim Λr (V ) = 0 for
r > n.
Proof: From Theorem 7.5 we know that any alternating tensor T ∈ Λr (V ) can
be written as
T = T j1 ··· jr θ j1 ⊗ · · · ⊗ θ jr .
We have that Alt T = T , and so
1
T = T j1 ··· jr Alt (θ j1 ⊗ · · · ⊗ θ jr ) = T j ··· j θ j1 ∧ · · · ∧ θ jr
r! 1 r
In the expansion the terms with jk = j` are zero since θ jk ∧ θ j` = 0. If we order the
indices in increasing order we obtain
1
T = ± Ti1 ···ir θi1 ∧ · · · ∧ θir ,
r!
which show that BΛ spans Λr (V ).
Linear independence can be proved as follows. Let 0 = λi1 ···ir θi1 ∧ · · · ∧ θir , and
thus λi1 ···ir = θi1 ∧ · · · ∧ θir (vi1 , · · · , vir ) = 0, which proves linear independence.
 59
!
n
It is immediately clear that BΛ consists of elements.
r

As we mentioned before the operation Alt is called the alternating projection.

As a matter of fact Sym is also a projection.
Lemma 8.11. Some of the basic properties can be listed as follows;
(i) Sym and Alt are projections on T r (V ), i.e. Sym2 = Sym, and Alt2 = Alt;
(ii) T is symmetric if and only if Sym T = T , and T is alternating if and only
if Alt T = T ;
(iii) Sym(T r (V )) = Σr (V ), and Alt(T r (V )) = Λr (V );
(iv) Sym ◦ Alt = Alt ◦ Sym = 0, i.e. if T ∈ Λr (V ), then Sym T = 0, and if
T ∈ Σr (V ), then Alt T = 0;
(v) let f : V → W , then Sym and Alt commute with f ∗ : T r (W ) → T r (V ), i.e.
Sym ◦ f ∗ = f ∗ ◦ Sym, and Alt ◦ f ∗ = f ∗ ◦ Alt.

9. Tensor bundles and tensor fields

Generalizations of tangent spaces and cotangent spaces are given by the tensor
spaces
T r (Tp M), Ts (Tp M), and Tsr (Tp M),
where T r (Tp M) = Tr (Tp∗ M). As before we can introduce the tensor bundles:
G
T rM = T r (Tp M),
p∈M
G
Ts M = Ts (Tp M),
p∈M
G
Tsr M = Tsr (Tp M),
p∈M

which are called the covariant r-tensor bundle, contravariant s-tensor bundle,
and the mixed (r, s)-tensor bundle on M. As for the tangent and cotangent bundle
the tensor bundles are also smooth manifolds. In particular, T 1 M = T ∗ M, and
T1 M = T M. Recalling the symmetric and alternating tensors as introduced in the
previous section we also define the tensor bundles Σr M and Λr M.
Theorem 9.1. The tensor bundles T r M, Tr M and Tsr M are smooth vector bundles.
Proof: Using Section 6 the theorem is proved by choosing appropriate local
trivializations Φ. For coordinates x = ϕα (p) and x0 = ϕβ (p) we recall that

∂ ∂x0j ∂ ∂x0j i
= , dx0 j | p = dx | p .
∂xi p ∂xi ∂x0j p ∂xi
60

For a covariant tensor T ∈ T r M this implies the following. The components are
defined in local coordinates by
 ∂ ∂ 
T = Ti1 ···ir dxi1 ⊗ · · · ⊗ dxir , Ti1 ···ir = T ,··· , .
∂xi1 ∂xir
The change of coordinates x → x0 then gives
 ∂x0 ∂ ∂x0jr ∂  ∂x0j1 ∂x0jr
j1 0
Ti1 ···ir = T , · · · , = T j ··· j · · · .
∂xi1 ∂x0j1 ∂xir ∂x0jr 1 r
∂xi1 ∂xir
Define
Φ(Ti1 ···ir dxi1 ⊗ · · · ⊗ dxir ) = p, (Ti1 ···ir ) ,

then !
 ∂x 0 ∂x 0 
j j
Φ0 ◦ Φ−1 p, (T j01 ··· jr ) = p, T j01 ··· jr 1 · · · r


.
∂xi1 ∂xir
The products of the partial derivatives are smooth functions. We can now apply
Theorem 6.4 as we did in Theorems 6.8 and 6.12.
On tensor bundles we also have the natural projection
π : Tsr M → M,
defined by π(p, T ) = p. A smooth section in Tsr M is a smooth mapping
σ : M → Tsr M,
such that π ◦ σ = idM . The space of smooth sections in Tsr M is denoted by Fsr (M).
For the co- and contravariant tensors these spaces are denoted by Fr (M) and Fs (M)
respectively. Smooth sections in these tensor bundles are also called smooth tensor
fields. Clearly, vector fields and 1-forms are examples of tensor fields. Sections in
tensor bundles can be expressed in coordinates as follows:

σ dxi1 ⊗ · · · ⊗ dxir , σ ∈ Fr (M),
 i1 ···ir


σ = σ j1 ··· js ∂x∂j ⊗ · · · ⊗ ∂x∂js , σ ∈ Fs (M),
 1
σ j1 ··· js dx ⊗ · · · ⊗ dxir ⊗ ∂ ⊗ · · · ⊗ ∂ , σ ∈ Fr (M).

i1 ···ir i1 ∂x j1 ∂x js s

Tensor fields are often denoted by the component functions. The tensor and tensor
fields in this course are, except for vector fields, all covariant tensors and covariant
tensor fields. Smoothness of covariant tensor fields can be described in terms of
the component functions σi1 ···ir .
27
Lemma 9.2. A covariant tensor field σ is smooth at p ∈ U if and only if
(i) the coordinate functions σi1 ···ir : U → R are smooth, or equivalently if and
only if
27See Lee, Lemma 11.6.
 61

(ii) for smooth vector fields X1 , · · · , Xr defined on any open set U ⊂ M, then the
function σ(X1 , · · · , Xr ) : U → R, given by

σ(X1 , · · · , Xr )(p) = σ p (X1 (p), · · · , Xr (p)),

is smooth.
The same equivalences hold for contravariant and mixed tensor fields.
Proof: Use the identities in the proof of Theorem 9.1 and then the proof goes as
Lemma 6.11.

J 9.3 Example. Let M = R2 and let σ = dx1 ⊗ dx1 + x12 dx2 ⊗ dx2 . If X = ξ1 (x) ∂x∂1 +
ξ2 (x) ∂x∂2 and Y = η1 (x) ∂x∂1 + η2 (x) ∂x∂2 are arbitrary smooth vector fields on Tx R2 ∼
=
2
R , then
σ(X,Y ) = ξ1 (x)η1 (x) + x12 ξ2 (x)η2 (x),

which clearly is a smooth function in x ∈ R2 . I

For covariant tensors we can also define the notion of pullback of a mapping
f between smooth manifolds. Let f : N → M be a smooth mappings, then the
pullback f ∗ : T r (T f (p) M) → T r (Tp N) is defined as

( f ∗ T )(X1 , · · · Xr ) := T ( f∗ X1 , · · · , f∗ Xr ),

where T ∈ T r (T f (p) M), and X1 , · · · , Xr ∈ Tp N. We have the following properties.

Lemma 9.4. 28 Let f : N → M, g : M → P be smooth mappings, and let p ∈ N,
0
S ∈ T r (T f (p) M), and T ∈ T r (T f (p) M), then:
(i) f ∗ : T r (T f (p) M) → T r (Tp N) is linear;
(ii) f ∗ (S ⊗ T ) = f ∗ S ⊗ f ∗ T ;
(iii) (g ◦ f )∗ = f ∗ ◦ g∗ : T r (T(g◦ f )(p) P) → T r (Tp M);
(iv) id∗M S = S;
(v) f ∗ : T r M → T r N is a smooth bundle map.
Proof: Combine the proof of Lemma 5.3 with the identities in the proof of
Theorem 9.1.

J 9.5 Example. Let us continue with the previous example and let N = M = R2 .
Consider the mapping f : N → M defined by

f (x) = (2x1 , x23 − x1 ).

28See Lee, Proposition 11.8.

62

For given tangent vectors X,Y ∈ Tx N, given by X = ξ1 ∂x∂1 + ξ2 ∂x∂2 and Y = η1 ∂x∂1 +
η2 ∂x∂2 , we can compute the pushforward
!
2 0
f∗ = , and
−1 3x22

∂ ∂
f∗ X = 2ξ1 + (−ξ1 + 3x22 ξ2 ) ,
∂y1 ∂y2
∂ ∂
f∗Y = 2η1 + (−η1 + 3x22 η2 ) .
∂y1 ∂y2

Let σ be given by σ = dy1 ⊗ dy1 + y21 dy2 ⊗ dy2 . This then yields

σ( f∗ X, f∗Y ) = 4ξ1 η1 + 4x12 (ξ1 − 3x12 ξ2 )(η1 − 3x12 η2 )

= 4(1 + x12 )ξ1 η1 − 12x12 x22 ξ1 η2
−12x12 x22 ξ2 η1 + 36x12 x24 x12 ξ2 η2 .

We have to point out here that f∗ X and f∗Y are tangent vectors in Tx N and not
necessarily vector fields, although we can use this calculation to compute f ∗ σ,
which clearly is a smooth 2-tensor field on T 2 N. I

J 9.6 Example. A different way of computing f ∗ σ is via a local representation of

σ directly. We have σ = dy1 ⊗ dy1 + y21 dy2 ⊗ dy2 , and

f ∗ σ = d(2x1 ) ⊗ d(2x1 ) + 4x12 d(x23 − x1 ) ⊗ d(x23 − x1 )

= 4dx1 ⊗ dx1 + 4x12 3x22 dx2 − dx1 ⊗ 3x22 dx2 − dx1

= 4(1 + x12 )dx1 ⊗ dx1 − 12x12 x22 dx1 ⊗ dx2

−12x12 x22 dx2 ⊗ dx1 + 36x12 x24 x12 dx2 ⊗ dx2 .

Here we used the fact that computing the differential of a mapping to R produces
the pushforward to a 1-form on N. I

J 9.7 Example. If we perform the previous calculation for an arbitrary 2-tensor

σ = a11 dy1 ⊗ dy1 + a12 dy1 ⊗ dy2 + a21 dy2 ⊗ dy1 + a22 dy2 ⊗ dy2 .

Then,

f ∗ σ = (4a11 − 2a12 − 2a21 + a22 )dx1 ⊗ dx1 + (6a12 − 3a22 )x22 dx1 ⊗ dx2
+(6a21 − 3a22 )x22 dx2 ⊗ dx1 + 9x24 a22 dx2 ⊗ dx2 ,
 63

which produces the following matrix if we identify T 2 (Tx N) and T 2 (Ty M) with R4 :
 
4 −2 −2 1
 0 6x2 0 −3x2 
f∗ =  2 2 
,

 0 0 6x2 −3x22 
2

0 0 0 9x24
which clearly equal to the tensor product of the matrices (J f )∗ , i.e.
! !
2 −1 2 −1
f ∗ = (J f )∗ ⊗ (J f )∗ = ⊗ .
0 3x22 0 3x22

This example show how to interpret f ∗ : T 2 (Ty M) → T 2 (Tx N) as a linear mapping.

I
As for smooth 1-forms this operation extends to smooth covariant tensor fields:
p = f (σ f (p) ), σ ∈ F (N), which in coordinates reads
( f ∗ σ) ∗ r

( f ∗ σ) p (X1 , · · · Xr ) := σ f (p) ( f∗ X1 , · · · , f∗ Xr ),
for tangent vectors X1 , · · · , Xr ∈ Tp N.
Lemma 9.8. 29 Let f : N → M, g : M → P be smooth mappings, and let h ∈ C∞ (M),
σ ∈ Fr (M), and τ ∈ Fr (N), then:
(i) f ∗ : Fr (M) → Fr (N) is linear;
(ii) f ∗ (hσ) = (h ◦ f ) f ∗ σ;
(iii) f ∗ (σ ⊗ τ) = f ∗ σ ⊗ f ∗ τ;
(iv) f ∗ σ is a smooth covariant tensor field;
(v) (g ◦ f )∗ = f ∗ ◦ g∗ ;
(vi) id∗M σ = σ;
Proof: Combine the Lemmas 9.4 and 9.2.

10. Differential forms

A special class of covariant tensor bundles and associated bundle sections are
the so-called alternating tensor bundles. Let Λr (Tp M) ⊂ T r (Tp M) be the space
alternating tensors on Tp M. We know from Section 8 that a basis for Λr (Tp M) is
given by n o
dxi1 ∧ · · · ∧ dxir : 1 ≤ i1 , · · · , ir ≤ m ,
m!
and dim Λr (Tp M) = r!(m−r)! . The associated tensor bundles of atlernating covariant
tensors is denoted by Λr M. Smooth sections in Λr M are called differential r-forms,
and the space of smooth sections is denoted by Γr (M) ⊂ Fr (M). In particular
29See Lee, Proposition 11.9.
64

Γ0 (M) = C∞ (M), and Γ1 (M) = F∗ (M). In terms of components a differential r-

form, or r-form for short, is given by

σi1 ···ir dxi1 ∧ · · · ∧ dxir ,

and the components σi1 ···ir are smooth functions. An r-form σ acts on vector fields
X1 , · · · , Xr as follows:

σ(X1 , · · · , Xr ) = ∑ (−1)a σi ···i dxi (Xa(1) ) · · · dxi (Xa(r) )

1 r 1 r
a∈Sr
i i
= ∑ (−1)a σi ···i Xa(1)
1 r· · · Xa(r)
1 r
.
a∈Sr

J 10.1 Example. Let M = R3 , and σ = dx ∧ dz. Then for vector fields

∂ ∂ ∂
X1 = X11 + X12 + X13 ,
∂x ∂y ∂z
and
∂ ∂ ∂
X2 = X21 + X22 + X23 ,
∂x ∂y ∂z
we have that
σ(X1 , X2 ) = X11 X23 − X13 X21 .
I

An important notion that comes up in studying differential forms is the notion

of contracting an r-form. Given an r-form σ ∈ Γr (M) and a vector field X ∈ F(M),
then
iX σ := σ(X, ·, · · · , ·),
is called the contraction with X, and is a differential (r − 1)-form on M. Another
notation for this is iX σ = Xyσ. Contraction is a linear mapping

iX : Γr (M) → Γr−1 (M).

Contraction is also linear in X, i.e. for vector fields X,Y it holds that

iX+Y σ = iX σ + iY σ, iλX σ = λ · iX σ.

Lemma 10.2. 30
Let σ ∈ Γr (M) and X ∈ F(M) a smooth vector field, then
(i) iX σ ∈ Γr−1 (M) (smooth (r − 1)-form);
(ii) iX ◦ iX = 0;
(iii) iX is an anti-derivation, i.e. for σ ∈ Γr (M) and ω ∈ Γs (M),

iX (σ ∧ ω) = (iX σ) ∧ ω + (−1)r σ ∧ (iX ω).

30See Lee, Lemma 13.11.

 65

A direct consequence of (iii) is that if σ = σ1 ∧ · · · ∧ σr , where σi ∈ Γ1 (M) =

F∗ (M), then
(10) iX σ = (−1)i−1 σi (X)σ1 ∧ · · · σ
bi ∧ · · · ∧ σr ,
where the hat indicates that σi is to be omitted, and we use the summation conven-
tion.
J 10.3 Example. Let σ = x2 dx1 ∧ dx3 be a 2-form on R3 , and X = x12 ∂x∂1 + x3 ∂x∂2 +
(x1 + x2 ) ∂x∂3 a given vector field on R3 . If Y = Y 1 ∂y∂1 +Y 2 ∂y∂2 +Y 3 ∂y∂3 is an arbitrary
vector fields then
(iX σ)(Y ) = σ(X,Y ) = dx1 (X)dx3 (Y ) − dx1 (Y )dx3 (X)
= x12Y 3 − (x1 + x2 )Y 1 ,
which gives that
iX σ = x12 dx3 − (x1 + x2 )dx1 .
I
Since σ is a 2-form the calculation using X,Y is still doable. For higher forms
this becomes to involved. If we use the multilinearity of forms we can give a simple
procedure for computing iX σ using Formula (10).
J 10.4 Example. Let σ = dx1 ∧ dx2 ∧ dx3 be a 3-form on R3 , and X the vector field
as given in the previous example. By linearity
iX σ = iX1 σ + iX2 σ + iX3 σ,
where X1 = x12 ∂x∂1 , X2 = x3 ∂x∂2 , and X3 = (x1 + x2 ) ∂x∂3 . This composition is chosen
so that X is decomposed in vector fields in the basis directions. Now
iX1 σ = dx1 (X1 )dx2 ∧ dx3 = x12 dx2 ∧ dx3 ,
iX2 σ = −dx2 (X2 )dx1 ∧ dx3 = −x3 dx1 ∧ dx3 ,
iX3 σ = dx2 (X3 )dx1 ∧ dx2 = (x1 + x2 )dx1 ∧ dx2 ,
which gives
iX σ = x12 dx2 ∧ dx3 − x3 dx1 ∧ dx3 + (x1 + x2 )dx1 ∧ dx2 ,
a 2-form on R3 . One should now verify that the same answer is obtained by com-
puting σ(X,Y, Z). I
For completeness we recall that for a smooth mapping f : N → M, the pullback
of a r-form σ is given by
( f ∗ σ) p (X1 , · · · , Xr ) = f ∗ σ f (p) (X1 , · · · , Xr ) = σ f (p) ( f∗ X1 , · · · , f∗ Xr ).
We recall that for a mapping h : M → R, then pushforward, or differential of h
dh p = h∗ ∈ Tp∗ M. In coordinates dh p = ∂x
∂h̃
i
dxi | p , and thus the mapping p 7→ dh p
66

is a smooth section in Λ1 (M), and therefore a differential 1-form, with component

∂h̃
σi = ∂x i
(in local coordinates).
If f : N → M is a mapping between m-dimensional manifolds with charts (U, ϕ),
and (V, ψ) respectively, and f (U) ⊂ V . Set x = ϕ(p), and y = ψ(q), then

f ∗ σdy1 ∧ · · · ∧ dym = (σ ◦ f ) det J f˜|x dx1 ∧ · · · ∧ dxm .

(11)

This can be proved as follows. From the definition of the wedge product and
Lemma 9.8 it follows that f ∗ (dy1 ∧ · · · ∧ dym ) = f ∗ dy1 ∧ · · · ∧ f ∗ dym , and f ∗ dy j =
∂ fej
∂xi dxi = dF j , where F = ψ ◦ f and fe = ψ ◦ f ◦ ϕ−1 . Now

f ∗ (dy1 ∧ · · · ∧ dym ) = f ∗ dy1 ∧ · · · ∧ f ∗ dym

= dF 1 ∧ · · · ∧ dF m ,

and furthermore, using (7),

!
 ∂ ∂   ∂   ∂ fe 
1 m i j
dF ∧ · · · ∧ dF ,··· , = det dF = det ,
∂x1 ∂xm ∂x j ∂x j

which proves the above claim.

As a consequence of this a change of coordinates fe = ψ ◦ ϕ−1 yields

dy1 ∧ · · · ∧ dym = det J f˜|x dx1 ∧ · · · ∧ dxm .

(12)

J 10.5 Example. Consider σ = dx ∧ dy on R2 , and mapping f given by x = r cos(θ)

and y = r sin(θ). The map f the identity mapping that maps R2 in Cartesian coor-
dinates to R2 in polar coordinates (consider the chart U = {(r, θ) : r > 0, 0 < θ <
2π}). As before we can compute the pullback of σ to R2 with polar coordinates:

σ = dx ∧ dy = d(r cos(θ)) ∧ d(r sin(θ))

= (cos(θ)dr − r sin(θ)dθ) ∧ (sin(θ)dr + r cos(θ)dθ)
= r cos2 (θ)dr ∧ dθ − r sin2 (θ)dθ ∧ dr
= rdr ∧ dθ.

Of course the same can be obtained using (12). I

J 10.6 Remark. If we define

∞
M
Γ(M) = Γr (M),
r=0

which is an associative, anti-commutative graded algebra, then f ∗ : Γ(N) → Γ(M)

is a algebra homomorphism. I
 67

11. Orientations
In order to explain orientations on manifolds we first start with orientations of
finite-dimensional vector spaces. Let V be a real m-dimensional vector space. Two
ordered basis {v1 , · · · , vm } and {v01 , · · · , v0m } are said to be consistently oriented if
the transition matrix A = (ai j ), defined by the relation

vi = ai j v0j ,

has positive determinant. This notion defines an equivalence relation on ordered

bases, and there are exactly two equivalence classes. An orientation for V is a
choice of an equivalence class of order bases. Given an ordered basis {v1 , · · · , vm }
the orientation is determined by the class O = [v1 , · · · , vm ]. The pair (V, O) is called
an oriented vector space. For a given orientation any ordered basis that has the
same orientation is called positively oriented, and otherwise negatively oriented.
Lemma 11.1. Let 0 6= θ ∈ Λm (V ), then the set of all ordered bases {v1 , · · · , vm }
for which θ(v1 , · · · , vm ) > 0 is an orientation for V .
Proof: Let {v0i } be any ordered basis and vi = ai j v0j . Then θ(v1 , · · · , vm ) =
det(A)θ(v1 , · · · , vm ) > 0, which proves that the set of bases {vi } for which it holds
that θ(v1 , · · · , vm ) > 0, characterizes an orientation for V .

Let us now describe orientations for smooth manifolds M. We will assume that
dim M = m ≥ 1 here. For each point p ∈ M we can choose an orientation O p for
the tangent space Tp M, making (Tp M, O p ) oriented vector spaces. The collection
{O} p∈M} is called a pointwise orientation. Without any relation between these
choices this concept is not very useful. The following definition relates the choices
of orientations of Tp M, which leads to the concept of orientation of a smooth man-
ifold.

Definition 11.2. A smooth m-dimensional manifold M with a pointwise orienta-

tion {O p } p∈M , O p = X1 , · · · , Xm , is oriented if for each point in M there exists a
 

neighborhood U and a diffeomorphism ϕ, mapping from U to an open subset of

Rm , such that
 
(13) ϕ∗ (X1 ), · · · , ϕ∗ (Xm ) = [e1 , · · · , em ], ∀p ∈ U,
p

where [e1 , · · · , em ] is the standard orientation of Rm . In this case the pointwise

orientation {O p } p∈M is said to be consistently oriented. The choice of {O p } p∈M is
called an orientation on M, denoted by O = {O p } p∈M .
68

If a consistent choice of orientations O p does not exist we say that a manifold is

non-orientable.
J 11.3 Remark. If we look at a single chart (U, ϕ) we can choose orientations O p
that are consistently oriented for p ∈ O p . The choices Xi = ϕ−1 ∂
∗ (ei ) = ∂xi | p are
ordered basis for Tp M. By definition ϕ∗ (Xi ) = ei , and thus by this choice we obtain
a consistent orientation for all p ∈ U. This procedure can be repeated for each chart
in an atlas for M. In order to get a globally consistent ordering we need to worry
about the overlaps between charts. I

J 11.4 Example. Let M = S1 be the circle in R2 , i.e. M = {p = (p1 , p2 ) : p21 + p22 =

1}. The circle is an orientable manifold and we can find a orientation as follows.
Consider the stereographic charts U1 = S1 \N p and U2 = S1 \Sp, and the associated
mappings
2p1 2p1
ϕ1 (p) = , ϕ2 (p) =
1 − p2 1 + p2
 4x x2 − 4   4x 4 − x2 
ϕ−1
1 (x) = , , ϕ−1
2 (x) = , .
x2 + 4 x2 + 4 x2 + 4 x2 + 4
Let us start with a choice of orientation and verify its validity. Choose X(p) =
(−p2 , p1 ), then
!
 2 2p1  −p2 2
(ϕ1 )∗ (X) = Jϕ1 | p (X) = 2
= > 0,
1 − p2 (1 − p2 ) p1 1 − p2

on U1 (standard orientation). If we carry out the same calculation for U2 we obtain

−2
(ϕ2 )∗ (X) = Jϕ2 | p (X) = 1+p 2
< 0 on U2 , with corresponds to the opposite orienta-
tion. Instead by choosing ϕ̃2 (p) = ϕ2 (−p1 , p2 ) we obtain
2
(ϕ̃2 )∗ (X) = J ϕ̃2 | p (X) = > 0,
1 + p2
which chows that X(p) is defines an orientation O on S1 . The choice of vectors
X(p) does not have to depend continuously on p in order to satisfy the definition
of orientation, as we will explain now.
∂
For p ∈ U1 choose the canonical vectors X(p) = ∂x | p , i.e.
16−4x2
!
∂ 2 2
= Jϕ−11 | p (1) =
(x +4)
16x .
∂x p (x2 +4)2
!
−p 2
In terms of p this gives X(p) = 12 (1 − p2 ) . By definition (ϕ1 )∗ (X) = 1
p1
for p ∈ U1 . For p = N p we choose X(p) = (−1 0)t , so X(p) is defined for p ∈ S1
 69

and is not a continuous function of p! It remains to verify that (ϕ̃2 )∗ (X) > 0 for
some neighborhood of N p. First, at p = N p,
!
−1 2
J ϕ̃2 | p = = 1 > 0.
0 1 + p2

∂
The choice of X(p) at p = N p comes from the canonical choice ∂x p with respect
to ϕ̃2 . Secondly,
!
1  −2 2p1  −p2 1 − p2
J ϕ̃2 | p (X(p)) = (1 − p2 ) = ,
2 1 + p2 (1 + p2 )2 p1 1 + p2

for all p ∈ U2 \N p. Note that ϕ̃2 ◦ ϕ−1

1 (x) =
−4
x , and
4
J ϕ̃2 | p (X(p)) = J ϕ̃2 | p (Jϕ−1 −1
1 | p (1)) = J(ϕ̃2 ◦ ϕ1 )| p (1) = ,
x2
which shows that consistency at overlapping charts can be achieved if the transition
mappings have positive determinant. Basically, the above formula describes the
change of basis as was carried out in the case of vector spaces. I

What the above example shows us is that if we choose X(p) = ∂x∂ i | p for all
charts, it remains to be verified if we have the proper collection of charts, i.e. does
{Jϕβ ◦Jϕ−1 −1
 
α (ei )} = [{ei }] hold? This is equivalent to having det(Jϕβ ◦Jϕα ) > 0.
These observations lead to the following theorem.
Theorem 11.5. A smooth manifold M is orientable if and only if there exists an
atlas A = {(Uα , ϕα } such that for any p ∈ M

det(Jϕαβ |x ) > 0, ϕαβ = ϕβ ◦ ϕ−1

α ,

for any pair α, β for which x = ϕα (p), and p ∈ Uα ∩Uβ .

Proof: Obvious from the previous example and Lemma 11.1.

As we have seen for vector spaces m-forms can be used to define an orientation
on a vector space. This concept can also be used for orientations on manifolds.
Theorem 11.6. 31 Let M be a smooth m-dimensional manifold. A nowhere van-
ishing differential m-form θ ∈ Γm (M) determines a unique orientation O on M for
which θ p is positively orientated at each Tp M. Conversely, given an orientation O
on M, then there exists a nowhere vanishing m-form θ ∈ Γm (M) that is positively
oriented at each Tp M.

31See Lee, Prop. 13.4.

70

Proof: Let θ ∈ Γm (M) be a nowhere vanishing m-form, then in local coordinates

(U, ϕ), θ = f dx1 ∧ · · · ∧ dxm . For the canonical bases for Tp M we have
 ∂ ∂ 
θ ,··· , = f 6= 0.
∂x1 ∂xm
Without loss of generality can be assumed to be positive, otherwise change ϕ by
means of x1 → −x1 . Now let { ∂y∂ i } be canonical bases for Tp M with respect to a
chart (V, ψ). As before θ = gdy1 ∧ · · · ∧ dym , g > 0 on V . Assume that U ∩V 6= ∅,
then at the overlap we have
 ∂ ∂     ∂ ∂ 
−1
0<g = θ ,··· , = det J(ϕ ◦ ψ ) θ ,··· ,
∂y1 ∂ym ∂x1 ∂xm
 
= f det J(ϕ ◦ ψ−1 ) ,
 
which shows that det J(ϕ ◦ ψ−1 ) > 0, and thus M is orientable.
For the converse we construct non-vanishing m-forms on each chart (U, ϕ). Via
a partition of unity we can construct a smooth m-form on M, see Lee.
This theorem implies in particular that non-orientable m-dimensional manifolds
do not admit a nowhere vanishing m-form, or volume form.
J 11.7 Example. Consider the Möbius strip M. The Möbius strip is an example
of a non-orientable manifold. Let us parametrize the Möbius strip as an embedded
manifold in R3 :
 
r sin(θ/2) cos(θ) + cos(θ)
g : R × (−1, 1) → R3 , g(θ, r) =  r sin(θ/2) sin(θ) + sin(θ)  ,
 
r cos(θ/2)
where g is a smooth embedding when regarded as a mapping from R/2πZ ×
(−1, 1) → R3 . Let us assume that the Möbius strip M is oriented, then by the
above theorem there exists a nonwhere vanishing 2-form σ which can be given as
follows
σ = a(x, y, z)dy ∧ dz + b(x, y, z)dz ∧ dx + c(x, y, z)dx ∧ dy,
where (x, y, z) = g(θ, r). Since g is a smooth embedding (parametrization) the
pullback form g∗ σ = ρ(θ, r)dθ ∧ dr is a nowhere vanishing 2-form on R/2πZ ×
(−1, 1). In particular, this means that ρ is 2π-periodic in θ. Notice that σ =
iX dx ∧ dy ∧ dz, where
∂ ∂ ∂
X = a(x, y, z) + b(x, y, z) + c(x, y, z) .
∂x ∂y ∂z
∂ ∂ ∂ ∂ ∂ ∂
For vector fields ξ = ξ1 ∂x + ξ2 ∂y + ξ3 ∂z and η = η1 ∂x + η2 ∂y + η3 ∂z we then have
that
σ(ξ, η) = iX dx ∧ dy ∧ dz(ξ, η) = X · (ξ × η),
 71

where ξ × η is the cross product of ξ and η. The condition that θ is nowhere

vanishing can be translated to R2 as follows:

g∗ σ(e1 , e2 ) = σ(g∗ (e1 ), g∗ (e2 )).

Since g∗ (e1 ) × g∗ (e2 )|θ=0 = −g∗ (e1 ) × g∗ (e2 )|θ=2π , the pullback g∗ σ form cannot
be nowhere vanishing on R/2πZ × R, which is a contradiction. This proves that
the Möbius band is a non-orientable manifold. I

Let N, M be oriented manifolds of the same dimension, and let f : N → M be

a smooth mapping. A mapping is called orientation preserving at a point p ∈ N
if f∗ maps positively oriented bases of Tp N to positively oriented bases in T f (p) M.
A mapping is called orientation reversing at p ∈ N if f∗ maps positively oriented
bases of Tp N to negatively oriented bases in T f (p) M.

F IGURE 32. Out(in)ward vectors and the induced orientation to

∂M.

Let us now look at manifolds with boundary; (M, ∂M). For a point p ∈ ∂M we
distinguish three types of tangent vectors:
(i) tangent boundary vectors X ∈ Tp (∂M) ⊂ Tp M, which form an (m − 1)-
dimensional subspace of Tp M;
(ii) outward vectors; let ϕ−1 : W ⊂ Hm → M, then X ∈ Tp M is called an out-
ward vector if ϕ−1
∗ (Y ) = X, for some Y = (y1 , · · · , ym ) with y1 < 0;
(iii) inward vectors; let ϕ−1 : W ⊂ Hm → M, then X ∈ Tp M is called an inward
vector if ϕ−1
∗ (Y ) = X, for some Y = (y1 , · · · , ym ) with y1 > 0.
72

Using this concept we can now introduce the notion of induced orientation on ∂M.
Let p ∈ ∂M and choose a basis {X1 , · · · , Xm } for Tp M such that [X1 , · · · , Xm ] = O p ,
{X2 , · · · , Xm } are tangent boundary vectors, and X1 is an outward vector. In this case
[X2 , · · · , Xm ] = (∂O) p determines an orientation for Tp (∂M), which is consistent,
and therefore ∂O = {(∂O) p } p∈∂M is an orientation on ∂M induced by O. Thus for
an oriented manifold M, with orientation O, ∂M has an orientation ∂O, called the
induced orientation on ∂M.
J 11.8 Example. Any open set M ⊂ Rm (or Hm ) is an orientable manifold. I
J 11.9 Example. Consider a smooth embedded co-dimension 1 manifold
M = {p ∈ Rm+1 : f (p) = 0}, f : Rm+1 → R, rk( f )| p , p ∈ M.
Then M is an orientable manifold. Indeed, M = ∂N, where N = {p ∈ Rm+1 : f (p) >
0}, which is an open set in Rm+1 and thus an oriented manifold. Since M = ∂N the
manifold M inherits an orientation from M and hence it is orientable. I
 73

IV. Integration on manifolds

12. Integrating m-forms on Rm

We start off integration of m-forms by considering m-forms on Rm .

Definition 12.1. A subset D ⊂ Rm is called a domain of integration if

(i) D is bounded, and
(ii) ∂D has m-dimensional Lebesgue measure dµ = dx1 · · · dxm equal to zero.

In particular any finite union or intersection of open or closed rectangles is a

domain of integration. Any bounded32 continuous function f on D is integrable,
i.e.
Z
−∞ < f dx1 · · · dxm < ∞.
D

Since Λm (Rm ) ∼
= R, a smooth m-form on Rm is given by

ω = f (x1 , · · · , xm )dx1 ∧ · · · ∧ dxm ,

where f : Rm → R is a smooth function. For a given (bounded) domain of integra-

tion D we define
Z Z Z
ω := f (x1 , · · · , xm )dx1 · · · dxm = f dµ
D
ZD D

= ωx (e1 , · · · , em )dµ.
D

An m-form ω is compactly supported if supp(ω) = cl{x ∈ Rm : ω(x) 6= 0}

is a compact set. The set of compactly supported m-forms on Rm is denoted by
Γm m m m
c (R ), and is a linear subspace of Γ (R ). Similarly, for any open set U ⊂ R
m

we can define ω ∈ Γm m m m
c (U). Clearly, Γc (U) ⊂ Γc (R ), and can be viewed as a
linear subspace via zero extension to Rm . For any open set U ⊂ Rm there exists a
domain of integration D such that U ⊃ D ⊃ supp(ω) (see Exercises).

32Boundedness is needed because the rectangles are allowed to be open.

 74

Definition 12.2. Let U ⊂ Rm be open and ω ∈ Γm c (U), and let D be a domain of

integration D such that U ⊃ D ⊃ supp(ω). We define the integral
Z Z
ω := ω.
U D

If U ⊂ Hm open, then
Z Z
ω := ω.
U D∩Hm

The next theorem is the first step towards defining integrals on m-dimensional
manifolds M.
Theorem 12.3. Let U,V ⊂ Rm be open sets, f : U → V an orientation preserving
diffeomorphism, and let ω ∈ Γm
c (V ). Then,
Z Z
ω= f ∗ ω.
V U

f ∗ ω.
R R
If f is orientation reversing, then U ω=− V
Proof: Assume that f is an orientation preserving diffeomorphism from U to
V . Let E be a domain a domain of integration for ω, then D = f −1 (E) is a do-
main of integration for f ∗ ω. We now prove the theorem for the domains D and
E. We use coordinates {xi } and {yi } on D and E respectively. We start with
ω = g(y1 , · · · , ym )dy1 ∧ · · · ∧ dym . Using the change of variables formula for inte-
grals and the pullback formula in (11) we obtain
Z Z
ω = g(y)dy1 · · · dym (Definition)
E
ZE
= (g ◦ f )(x) det(J f˜|x )dx1 · · · dxm
D
Z
= (g ◦ f )(x) det(J f˜|x )dx1 ∧ · · · ∧ dxm
D
Z
= f ∗ ω (Definition).
D

One has to introduce a − sign in the orientation reversing case.

13. Partitions of unity

We start with introduce the notion partition of unity for smooth manifolds. We
should point out that this definition can be used for arbitrary topological spaces.
 75

Definition 13.1. Let M be smooth m-manifold with atlas A = {(ϕi ,Ui )}i∈I . A
partition of unity subordinate to A is a collection for smooth functions {λi : M →
R}i∈I satisfying the following properties:
(i) 0 ≤ λi (p) ≤ 1 for all p ∈ M and for all i ∈ I;
(ii) supp(λi ) ⊂ Ui ;
(iii) the set of supports {supp(λi )}i∈I is locally finite;
(iv) ∑i∈I λi (p) = 1 for all p ∈ M.

Condition (iii) says that every p ∈ M has a neighborhood U 3 p such that only
finitely many λi ’s are nonzero in U. As a consequence of this the sum in Condition
(iv) is always finite.
Theorem 13.2. For any smooth m-dimensional manifold M with atlas A there
exists a partition of unity {λi } subordinate to A.
In order to prove this theorem we start with a series of auxiliary results and
notions.
Lemma 13.3. There exists a smooth function h : Rm → R such that 0 ≤ h ≤ 1 on
Rm , and h|B1 (0) ≡ 1 and supp(h) ⊂ B2 (0).
Proof: Define the function f1 : R → R as follows
(
e−1/t , t > 0,
f1 (t) =
0, t ≤ 0.
One can easily prove that f1 is a C∞ -function on R. If we set
f1 (2 − t)
f2 (t) = .
f1 (2 − t) + f1 (t − 1)
This function has the property that f2 (t) ≡ 1 for t ≤ 1, 0 < f2 (t) < 1 for 1 < t <
2, and f2 (t) ≡ 0 for t ≥ 2. Moreover, f2 is smooth. To construct f we simply
write f (x) = f2 (|x|) for x ∈ Rm \{0}. Clearly, f is smooth on Rm \{0}, and since
f |B1 (0) ≡ 1 it follows that f is smooth on Rm .
An atlas A gives an open covering for M. The set Ui in the atlas need not be
compact, nor locally finite. We say that a covering U = {Ui } of M is locally finite
if every point p ∈ M has a neighborhood that intersects only finitely Ui ∈ U. If
there exists another covering V = {V j } such that every V j ∈ V is contained in some
V j ⊂ Ui ∈ U, then V is called a refinement of U. A topological space X for which
each open covering admits a locally finite refinement is called paracompact.
Lemma 13.6. Any topological manifold M allows a countable, locally finite cov-
ering by precompact open sets.
76

F IGURE 33. The functions f1 and f2 .

F IGURE 34. A locally finite covering [left], and a refinement

[right].

Proof: We start with a countable covering of open balls B = {Bi }. We now

construct a covering U that satisfies
(i) all sets Ui ∈ U are precompact open sets in M;
(ii) Ui−1 ⊂ Ui , i > 1;
(iii) Bi ⊂ Ui .
We build the covering U from B using an inductive process. Let U1 = B1 , and
assume U1 , · · · ,Uk have been constructed satisfying (i)-(iii). Then
Uk ⊂ Bi1 ∪ · · · ∪ Bik ,
where Bi1 = B1 . Now set
Uk+1 = Bi1 ∪ · · · ∪ Bik .
Choose ik large enough so that ik ≥ k + 1 and thus Bk+1 ⊂ Uk+1 .
From the covering U we can now construct a locally finite covering V by setting
Vi = Ui \Ui−2 , i > 1.
 77

F IGURE 35. Constructing a nested covering U from a covering

with balls B [left], and a locally finite covering obtained from the
previous covering [right].

Next we seek a special locally finite refinement {W j } = W of V which has spe-

cial properties with respect to coordinate charts of M;
(i) W is a countable and locally finite;
(ii) each W j ⊂ W is in the domain of some smooth coordinate map ϕ j : j → Rm
for M;
(iii) the collection Z = {Z j }, Z j = ϕ−1
j (B1 (0)) covers M.
The covering W is called regular.
Lemma 13.8. For any open covering U for a smooth manifold M there exists a
regular refinement. In particular M is paracompact.
Proof: Let V be a countable, locally finite covering as described in the previous
lemma. Since V is locally finite we can find a neighborhood Wp for each p ∈ M
that intersects only finitely many set V j ∈ V. We want to choose Wp in a smart
way. First we replace Wp by Wp ∩ {V j : p ∈ V j }. Since p ∈ Ui for some i we then
replace Wp by Wp ∩ Vi . Finally, we replace Wp by a small coordinate ball Br (p)
so that Wp is in the domain of some coordinate map ϕ p . This provides coordinate
charts (Wp , ϕ p ). Now define Z p = ϕ−1 p (B1 (0)).
For every k, {Z p : p ∈ Vk } is an open covering of Vk , which has a finite subcov-
ering, say Zk1 , · · · , Zkmk . The sets Zki are sets of the form Z pi for some pi ∈ Vk . The
associated coordinate charts are (Wk1 , ϕ1k ), · · · , (Wkmk , ϕm k i
k ). Each Wk is obtained via
the construction above for pi ∈ Vk . Clearly, {Wki } is a countable open covering of
M that refines U and which, by construction, satisfies (ii) and (iii) in the definition
of regular refinement. It is clear from compactness that {Wki } is locally finite. Let
us relabel the sets {Wki } and denote this covering by W = {Wi }. As a consequence
M is paracompact.
78

F IGURE 36. The different stages of constructing the sets Wp col-

ored with different shades of grey.

Proof of Theorem 13.2: From Lemma 13.8 there exists a regular refinement W
for A. By construction we have Zi = ϕ−1
i (B1 (0)), and we define

Zbi := ϕ−1
i (B2 (0)).

Using Lemma 13.3 we now define the functions µ: M → R;

(
f2 ◦ ϕi on Wi
µi =
0 on M\Vi .
These functions are smooth and supp(µi ) ⊂ Wi . The quotients

bλ = µ j (p) ,
∑i µi (p)
are, due to the local finiteness of Wi and the fact that the denominator is positive
for each p ∈ M, smooth functions on M. Moreover, 0 ≤ bλ j ≤ 1, and ∑ j bλ j ≡ 1.
Since W is a refinement of A we can choose an index k j such that W j ⊂ Uk j .
Therefore we can group together some of the function bλ j :

λi = ∑ bλ j ,
j : k j =i

which give us the desired partition functions with 0 ≤ λi ≤ 1, ∑i λi ≡ 1, and

supp(λi ) ⊂ Ui .
Some interesting byproducts of the above theorem using partitions of unity are.
In all these case M is assumed to be an smooth m-dimensional manifold.
Theorem 13.10. 33 For any close subset A ⊂ M and any open set U ⊃ A, there
exists a smooth function f : M → R such that
33See Lee, Prop. 2.26.
 79

(i) 0 ≤ f ≤ 1 on M;
(ii) f −1 (1) = A;
(iii) supp( f ) ⊂ U.
Such a function f is called a bump function for A supported in U.
By considering functions g = c(1− f ) ≥ 0 we obtain smooth functions for which
g−1 (0)can be an arbitrary closed subset of M.
Theorem 13.11. 34 Let A ⊂ M be a closed subset, and f : A → Rk be a smooth
mapping.35 Then for any open subset U ⊂ M containing A there exists a smooth
mapping f † : M → Rk such that f † |A = f , and supp( f † ) ⊂ U.

14. Integration on of m-forms on m-dimensional manifolds.

In order to introduce the integral of an m-form on M we start with the case of
forms supported in a single chart. In what follows we assume that M is an oriented
manifold with an oriented atlas A, i.e. a consistent choice of smooth charts such
that the transitions mappings are orientation preserving.

F IGURE 37. An m-form supported in a single chart.

Let ω ∈ Γm c (M), with supp(ω) ⊂ U for some chart U in A. Now define the
integral of ω over M as follows:
Z Z
ω := (ϕ−1 )∗ ω,
M ϕ(U)

where the pullback form (ϕ−1 )∗ ω

is compactly supported in V = ϕ(U) (ϕ is a
diffeomorphism). The integral over V of the pullback form (ϕ−1 )∗ ω is defined in
34See Lee, Prop. 2.26.
35 A mapping from a closed subset A ⊂ M is said to be smooth if for every point p ∈ A there exists

an open neighborhood W ⊂ M of p, and a smooth mapping f † : W → Rk such that f † |W ∩A = f .

80
R
Definition 12.2. It remains to show that the integral M ω does not depend on the
particular chart. Consider a different chart U 0 , possibly in a different oriented atlas

F IGURE 38. Different charts U and U 0 containing supp(ω), i.e.

supp(ω) ⊂ U ∩U 0 .

A0 for M (same orientation), then

Z Z Z
(ϕ0−1 )∗ ω = (ϕ0−1 )∗ ω = (ϕ0 ◦ ϕ−1 )∗ (ϕ0−1 )∗ ω
V0 ϕ0 (U∩U 0 ) ϕ(U∩U 0 )
Z Z
−1 ∗ 0 ∗ 0−1 ∗
= (ϕ ) (ϕ ) (ϕ ) ω= (ϕ−1 )∗ ω
ϕ(U∩U 0 ) ϕ(U∩U 0 )
Z
= (ϕ−1 )∗ ω,
V

which show that the definition is independent of the chosen chart. We crucially do
use the fact that A ∪ A0 is an oriented atlas for M. To be more precise the mappings
ϕ0 ◦ ϕ−1 are orientation preserving.
By choosing a partition of unity as described in the previous section we can now
define the integral over M of arbitrary m-forms ω ∈ Γm c (M).

Definition 14.3. Let ω ∈ Γm c (M) and let AI = {(Ui , ϕi }i∈I ⊂ A be a finite subcov-
ering of supp(ω) coming from an oriented atlas A for M. Let {λi } be a partition
of unity subordinate to AI . Then the integral of ω over M is defined as
Z Z
ω := ∑ λi ω,
M i M
R
where the integrals M λi ω are integrals of form that have support in single charts
as defined above.
 81

We need to show now that the integral is indeed well-defined, i.e. the sum is
finite and independent of the atlas and partition of unity. Since supp(ω) is compact
AI exists by default, and thus the sum is finite.
Lemma 14.4. The above definition is independent of the chosen partition of unity
and covering AI .
Proof: Let A0J ⊂ A0 be another finite covering of supp(ω), where A0 is a com-

F IGURE 39. Using a partition of unity we can construct m-forms

which are all supported in one, but possibly different charts Ui .

patible oriented atlas for M, i.e. A ∪ A0 is a oriented atlas. Let {λ0j } be a partition
of unity subordinate to A0J . We have
Z Z   Z

M
λi ω =
M
λ0
∑ j i ∑ λ0j λi ω.
λ ω =
M
j j

By summing over i we obtain ∑i M λi ω = ∑i, j M λ0j λi ω. Each term λ0j λi ω is sup-

R R

ported in some Ui and by previous independent of the coordinate mappings. Sim-

ilarly, if we interchange the i’s and j’s, we obtain that ∑ j M λ0j ω = ∑i, j M λ0j λi ω,
R R

which proves the lemma.

J 14.6 Remark. If M is a compact manifold that for any ω ∈ Γm (M) it holds that
supp(ω) ⊂ M is a compact set, and therefore Γm m
c (M) = Γ (M) in the case of com-
pact manifolds. I

So far we considered integral of m-forms over M. One can of course integrate

n-forms on M over n-dimensional immersed or embedded submanifolds N ⊂ M.
Given ω ∈ Γn (M) for which the restriction to N, ω|N has compact support in N,
then Z Z
ω := ω|N .
N N
82

As a matter of fact we have i : N → M, and ω|N = i∗ ω, so that N ω := N ω|N =

R R
R ∗ R
N i ω. If M is compact with boundary ∂M, then ∂M ω is well-defined for any
(m − 1)-form ω ∈ Γm−1 (M).
Theorem 14.7. Let ω ∈ Γm
c (M), and f : N → M is a diffeomorphism. Then
Z Z
ω= f ∗ ω.
M N
Proof: By the definition of integral above we need to prove the above statement
for the terms Z Z
λi ω = f ∗ λi ω.
M N
Therefore it suffices to prove the theorem for forms ω whose support is in a single
chart U.

F IGURE 40. The pullback of an m-form.

Z Z Z
f ∗ω = (ϕ−1 )∗ f ∗ ω = (ϕ−1 )∗ f ∗ ψ∗ (ψ−1 )∗ ω
U ϕ(U) ϕ(U)
Z Z
= (ψ−1 )∗ ω = ω.
ψ(U 0 ) U0

By applying this to λi ω and summing we obtain the desired result.

For sake of completeness we summarize the most important properties of the
integral.
Theorem 14.9. 36 Let N, M be oriented manifolds (with or without boundary) of
dimension m , and ω, η ∈ Γm
c (M), are m-forms. Then
R R R
(i) M aω + bη =a M ω + b M η;

36See Lee Prop. 14.6.

 83

(ii) if −O is the opposite orientation to O, then

Z Z
ω=− ω;
M,−O M,O
R
(iii) if ω is an orientation form, then M ω > 0;
(iv) if f : N → M is a diffeomorphism, then
Z Z
ω= f ∗ ω.
M N
For practical situations the definition of the integral above is not convenient since
constructing partitions of unity is hard in practice. If the support of a differential
form ω can be parametrized be appropriate parametrizations, then the integral can
be easily computed from this. Let Di ⊂ Rm — finite set of indices i — be compact
domains of integration, and gi : Di → M are smooth mappings satisfying
(i) Ei = gi (Di ), and gi : int (Di ) → int (Ei ) are orientation preserving diffeo-
morphisms;
(ii) Ei ∩ E j intersect only along their boundaries, for all i 6= j.
Theorem 14.10. 37 Let {(gi , Di )} be a finite set of parametrizations as defined
above. Then for any ω ⊂ Γm
c (M) such that supp(ω) ⊂ ∪i Ei it holds that
Z Z
ω=∑ g∗i ω.
M i Di

Proof: As before it suffices the prove the above theorem for a single chart U,

F IGURE 41. Carving up supp(ω) via domains of integration for

parametrizations gi for M.

i.e. supp(ω) ⊂ U. One can choose U to have a boundary ∂U so that ϕ(∂U) has

37See Lee Prop. 14.7.

84

measure zero, and ϕ maps cl(U) to a compact domain of integration K ⊂ Hm . Now

set
Ai = cl(U) ∩ Ei , Bi = g−1
i (Ai ), Ci = ϕi (Ai ).
We have Z Z Z
−1 ∗ ∗ −1 ∗
(ϕ ) ω = (ϕ ◦ gi ) (ϕ ) ω = g∗i ω.
Ci Bi Bi

Since the interiors of the sets Ci (and thus Ai ) are disjoint it holds that
Z Z Z
ω = (ϕ−1 )∗ ω = ∑ (ϕ−1 )∗ ω
M K i Ci
Z Z
= ∑ g∗i ω =∑ g∗i ω,
i Bi i Di

which proves the theorem.

R
J 14.12 Remark. From the previous considerations computing M ω boils down
to computing (ϕ−1 )∗ ω, or g∗ ω for appropriate parametrizations, and summing the
various contrubutions. Recall from Section 12 that in order to integrate one needs
to evaluate (ϕ−1 )∗ ωx (e1 , · · · , em ), which is given by the formula

(ϕ−1 )∗ ω x (e1 , · · · , em ) = ωϕ−1 (x) (ϕ−1 −1

∗ (e1 ), · · · , ϕ∗ (em )).

For a single chart contribution this then yields the formula

Z Z
ω = ωϕ−1 (x) (ϕ−1 −1
∗ (e1 ), · · · , ϕ∗ (em ))dµ
U ϕ(U)
Z  ∂ ∂ 
= ωϕ−1 (x) ,··· , dµ
ϕ(U) ∂x1 ∂x1
or in the case of a parametrization g : D → M:
Z Z
ω= ωg(x) (g∗ (e1 ), · · · , g∗ (em ))dµ.
g(D) D

These expression are useful for computing integrals. I

J 14.13 Example. Consider the 2-sphere parametrized by the mapping g : R2 →

S2 ⊂ R3 given as
   
x sin(ϕ) cos(θ)
g(ϕ, θ) =  y  =  sin(ϕ) sin(θ)  .
   
z cos(ϕ)

This mapping can be viewed as a covering map. From this expression we derive
various charts for S2 . Given the 2-form ω = zdx ∧ dz let us compute the pullback
 85

form g∗ ω on R2 . We have that

∂ ∂ ∂
g∗ (e1 ) = cos(ϕ) cos(θ) + cos(ϕ) sin(θ) − sin(ϕ) ,
∂x ∂y ∂z
∂ ∂
g∗ (e2 ) = − sin(ϕ) sin(θ) + sin(ϕ) cos(θ) ,
∂x ∂y
and therefore

g∗ ω(e1 , e2 ) = ωg(x) (g∗ (e1 ), g∗ (e2 )) = − cos(ϕ) sin2 (ϕ) sin(θ),

and thus
g∗ ω = − cos(ϕ) sin2 (ϕ) sin(θ)dϕ ∧ dθ.
The latter gives that S2 ω = − 02π 0π cos(ϕ) sin2 (ϕ) sin(θ)dϕdθ = 0, which shows
R R R

that ω is not a volume form on S2 .

If we perform the same calculations for ω = xdy ∧ dz + ydz ∧ dx + zdx ∧ dy, then
Z Z 2π Z π
ω= sin(ϕ)dϕdθ = 4π.
S2 0 0

The pullback form g∗ ω = sin(ϕ)dϕ ∧ dθ, which shows that ω is a volume form on
S2 . I

15. The exterior derivative

The exterior derivative d of a differential form is an operation that maps an
k-form to a (k + 1)-form. Write a k-form on Rm in the following notation

ω = ωi1 ···ik dxi1 ∧ · · · ∧ dxik = ωI dxI , I = (i1 · · · ik ),

then we define

(14) dω = dωI ∧ dxI .

Written out in all its differentials this reads

  ∂ω
I
dω = d ωi1 ···ik dxi1 ∧ · · · ∧ dxik = dxi ∧ dxi1 ∧ · · · ∧ dxik .
∂xi
Of course the Einstein summation convention is again applied here. This is the
definition that holds for all practical purposes, but it is a local definition. We need
to prove that d extends to differential forms on manifolds.
J 15.1 Example. Consider ω0 = f (x, y), a 0-form on R2 , then
∂f ∂f
dω0 = dx + dy.
∂x ∂y
86

For a 1-form ω1 = f1 (x, y)dx + f2 (x, y)dy we obtain

∂ f1 ∂ f1 ∂ f2 ∂ f2
dω1 = dx ∧ dx + dy ∧ dx + dx ∧ dy + dy ∧ dy
∂x ∂y ∂x ∂y
∂ f1 ∂ f2 ∂ f
2 ∂ f1 
= dy ∧ dx + dx ∧ dy = − dx ∧ dy.
∂y ∂x ∂x ∂y
Finally, for a 2-form ω2 = g(x, y)dx ∧ dy the d-operation gives
∂g ∂g
dω2 = dx ∧ dx ∧ dy + dy ∧ dx ∧ dy = 0.
∂x ∂y
The latter shows that d applied to a top-form always gives 0. I
J 15.2 Example. In the previous example dω0 is a 1-form, and dω1 is a 2-form.
Let us now apply the d operation to these forms:
 ∂2 f ∂2 f 
d(dω0 ) = − dx ∧ dy = 0,
∂x∂y ∂y∂x
and
d(dω1 ) = 0,
since d acting on a 2-form always gives 0. These calculations seem to suggest that
in general d ◦ d = 0. I
As our examples indicate d 2 ω = 0. One can also have forms ω for which dω = 0,
but ω 6= dσ. We say that ω is a closed form, and when ω = dσ, then ω is called
an exact form. Clearly, closed forms form a possibly larger class than exact forms.
In the next chapter on De Rham cohomology we are going to come back to this
phenomenon in detail. On Rm one will not be able to find closed forms that are
not exact, i.e. on Rm all closed forms are exact. However, if we consider different
manifolds we can have examples where this is not the case.
J 15.3 Example. Let M = R2 \{(0, 0)}, and consider the 1-form
−y x
ω= dx + dy,
x 2 + y2 x 2 + y2
which clearly is a smooth 1-form on M. Notice that ω does not extend to a 1-form
on R2 ! It holds that dω = 0, and thus ω is a closed form on M. Let γ : [0, 2π) → M,
t 7→ (cos(t), sin(t)) be an embedding of S1 into M, then
Z Z 2π Z 2π
∗
ω= γ ω= dt = 2π.
γ 0 0

Assume that ω is an exact 1-form on M, then ω = d f , for some smooth function

f : M → R. In Section 5 we have showed that
Z Z Z 2π
ω= df = γ∗ d f = ( f ◦ γ)0 (t)dt = f (1, 0) − f (1, 0) = 0,
γ γ 0
 87
R R
which contradicts the fact that γω = γdf = 2π. This proves that ω is not exact.
I

For the exterior derivative in general we have the following theorem.

Theorem 15.4. 38 Let M be a smooth m-dimensional manifold. Then for all k ≥ 0
there exist unique linear operations d : Γk (M) → Γk+1 (M) such that;
(i) for any 0-form ω = f , f ∈ C∞ (M) it holds that

dω(X) = d f (X) = X f , X ∈ F(M);

(ii) if ω ∈ Γk (M) and η ∈ Γ` (M), then

d(ω ∧ η) = dω ∧ η + (−1)k ω ∧ dη;

(iii) d ◦ d = 0;
(iv) if ω ∈ Γm (M), then dω = 0.
This operation d is called the exterior derivative on differential forms, and is a
unique anti-derivation (of degree 1) on Γ(M) with d 2 = 0.
Proof: Let us start by showing the existence of d on a chart U ⊂ M. We have
local coordinates x = ϕ(p), p ∈ U, and we define

dU : Γk (U) → Γk+1 (U),

via (14). Let us write d instead of dU for notational convenience. We have that
d( f dxI ) = d f ∧ dxI . Due to the linearity of d it holds that

d( f dxI ∧ gdxJ ) = d( f gdxI ∧ dxJ ) = d( f g)dxK

= gd f dxK + f dgdxK
= (d f ∧ dxI ) ∧ (gdxJ ) + f dg ∧ dxI ∧ dxJ
= (d f ∧ dxI ) ∧ (gdxJ ) + (−1)k f dxI ∧ dg ∧ dxJ
= d( f dxI ) ∧ (gdxJ ) + (−1)k ( f dxI ) ∧ d(gdxJ ),

which proves (ii). As for (iii) we argue as follows. In the case of a 0-form we have
that
∂f ∂2 f  j
d(d f ) = d dxi = dx ∧ dxi
∂xi ∂x j ∂xi
 ∂2 f ∂2 f  i
= ∑ − dx ∧ dx j = 0.
i< j ∂x i ∂x j ∂x j ∂xi

38See Lee Thm. 12.14

88

Given a k-form ω = ωI dxI , then, using (ii), we have

 
d(dω) = d dωI ∧ dxI

= d(dωI ) ∧ dxI + (−1)k+1 dωI ∧ d(dxI ) = 0,

since ωI is a 0-form, and d(dxI ) = (−1) j dxi1 ∧ · · · ∧ d(dxi j ) ∧ · · · ∧ dxik = 0. The

latter follows from d(dxi j ) = 0. The latter also implies (iv), finishing the existence
proof in the one chart case. The operation d = dU is well-defined, satisfying (i)-
(iv), for any chart (U, ϕ).
The operation dU is unique, for if there exists yet another exterior derivative deU ,
which satisfies (i)-(iv), then for ω = ωI dxI ,
e I ∧ dxI + ωI d(dx
e = dω
dω e I ),

where we used (ii). From (ii) it also follows that d(dx e I ) = (−1) j dxi1 ∧· · ·∧ d(dx
e i j )∧
· · · ∧ dxik = 0. By (i) d(ϕ(p)i j ) = dxi j , and thus by (iv) d(dx
e i j ) = de◦ d(ϕ(p)
e i j ) = 0,
which proves the latter. From (i) it also follows that dωI = dωI , and therefore
e

e I ∧ dxI + ωI d(dx
e = dω
dω e I ) = dωI ∧ dxI = dω,

which proves the uniqueness of dU .

Before giving the defining d for ω ∈ Γk (M) we should point out that dU trivially
satisfies (i)-(iii) of Theorem 15.5 (use (14). Since we have a unique operation dU
for every chart U, we define for p ∈ U

(dω) p = (dU ω|U ) p ,

as the exterior derivative for forms on M. If there is a second chart U 0 3 p, then by

uniqueness it holds that on

dU (ω|U∩U 0 ) = dU∩U 0 (ω|U∩U 0 ) = dU 0 (ω|U∩U 0 ),

which shows that the above definition is independent of the chosen chart.
The last step is to show that d is also uniquely defined on Γk (M). Let p ∈ U, a
coordinate chart, and consider the restriction

ω|U = ωI dxI ,

where ωI ∈ C∞ (U). Let W ⊂ U be an open set containing p, with the additional

property that cl(W ) ⊂ U is compact. By Theorem 13.10 we can find a bump-
function g ∈ C∞ (M) such that g|W = 1, and supp(g) ⊂ U. Now define
e = gωI d(gxi1 ) ∧ · · · ∧ d(gxik ).
ω

Using (i) we have that d(gxi )|W = dxi , and therefore ω̃|W = ω|W . Set η = ω
e − ω,
then η|W = 0. Let p ∈ W and h ∈ C (M) satisfying h(p) = 1, and supp(h) ⊂ W .
∞
 89

Thus, hω ≡ 0 on M, and

0 = d(hω) = dh ∧ ω + hdω.

This implies that (dω) p = −(d f ∧ ω) p = 0, which proves that (d ω e )|W = (dω)|W .
If we use (ii)-(iii) then


e = d gωI d(gxi1 ) ∧ · · · ∧ d(gxik )
dω


= d(gωI ) ∧ d(gxi1 ) ∧ · · · ∧ d(gxik ) + gωI d d(gxi1 ) ∧ · · · ∧ d(gxik )
= d(gωI ) ∧ d(gxi1 ) ∧ · · · ∧ d(gxik ).

It now follows that since g|W = 1, and since (d ω

e )|W = (dω)|W , that
∂ωI i
(dω)|W = dx ∧ dxI ,
∂xi
which is exactly (14). We have only used properties (i)-(iii) the derive this ex-
pression, and since p is arbitrary it follows that d : Γk (M) → Γk+1 (M) is uniquely
defined.

The exterior derivative has other important properties with respect to restrictions
and pullbacks that we now list here.
Theorem 15.5. Let M be a smooth m-dimensional manifold, and let ω ∈ Γk (M),
k ≥ 0. Then
(i) in each chart (U, ϕ) for M, dω in local coordinates is given by (14);
(ii) if ω = ω0 on some open set U ⊂ M, then also dω = dω0 on U;
(iii) if U ⊂ M is open, then d(ω|U ) = (dω)|U ;
(iv) if f : N → M is a smooth mapping, then

f ∗ (dω) = d( f ∗ ω),

i.e. f ∗ : Γk (M) → Γk (N), and d commute as operations.

Proof: Let us restrict ourselves here to proof of (iv). It suffices to prove (iv) in a
chart U, and ω = ωI dxI . Let us first compute f ∗ (dω):

f ∗ (dω) = f ∗ dωI ∧ dxi1 ∧ · · · ∧ dxik

= d(ωI ◦ f ) ∧ d(ϕ ◦ f )i1 ∧ · · · ∧ d(ϕ ◦ f )ik .

Similarly,

d( f ∗ ω) = d (ωI ◦ f ) ∧ d(ϕ ◦ f )i1 ∧ · · · ∧ d(ϕ ◦ f )ik

= d(ωI ◦ f ) ∧ d(ϕ ◦ f )i1 ∧ · · · ∧ d(ϕ ◦ f )ik ,

which proves the theorem.

 90

16. Stokes’ Theorem

The last section of this chapter deals with a generalization of the Fundamental
Theorem of Integration: Stokes’ Theorem. The Stokes’ Theorem allows us to
R
compute M dω in terms of a boundary integral for ω. In a way the (m − 1)-form
act as the ‘primitive’, ‘anti-derivative’ of the dω. Therefore if we are interested in
R
M σ using Stokes’ Theorem we need to first ‘integrate’ σ, i.e. write σ = dω. This
is not always possible as we saw in the previous section.
Stokes’ Theorem can be now be phrased as follows.

Theorem 16.1. Let M be a smooth m-dimensional manifold with or without bound-

ary ∂M, and ω ∈ Γm−1
c (M). Then,
Z Z
(15) dω = j∗ ω,
M ∂M

where j : ∂M ,→ M is the natural embedding of the boundary ∂M into M.

In order to prove this theorem we start with the following lemma.

Lemma 16.2. Let ω ∈ Γm−1

c (Hm ) be given by

ω = ωi dx1 ∧ · · · ∧ dx
ci ∧ · · · ∧ dxm .

Then,

(i) HM dω = 0, if supp(ω) ⊂ int (Hm );

R

(ii) Hm dω = ∂Hm j∗ ω, if supp(ω) ∩ ∂Hm 6= ∅,

R R

where j : ∂Hm ,→ Hm is the canonical inclusion.

Proof: We may assume without loss of generality that supp(ω) ⊂ [0, R] × · · · ×

[0, R] = IRm . Now,

Z Z
∂ωi
dω = (−1)i+1 dx1 · · · dxm
Hm IRm ∂xi
Z  
i+1
= (−1) m−1
ωi |x i =R − ω i |x i =0 dx1 · · · dx
ci · · · dxm .
IR

If supp(ω) ⊂ int IRm , then ωi |xi =R − ωi |xi =0 = 0 for all i, and

Z
dω = 0.
Hm
 91

If supp(ω)∩∂Hm 6= ∅, then ωi |xi =R −ωi |xi =0 = 0 for all i ≤ m−1, and ωm |xm =R = 0.
Therefore,
Z Z
∂ωi
dω = (−1)i+1 dx1 · · · dxm
Hm IRm ∂xi
Z  
i+1
= (−1) m−1
ωi |xi =R − ω i |x i =0 dx1 · · · dx
ci · · · dxm
I
ZR  
= (−1)m+1 m−1
−ωi |xm =0 dx1 · · · dxm−1
IR
Z  
= (−1)m m−1
ωi xm =0 dx1 · · · dxm−1 .
|
IR

The mapping j : ∂Hm → Hm , which, in coordinates, is given by j˜ = j ◦ϕ : (x1 , · · · , xm−1 ) 7→

(x1 , · · · , xm−1 , 0), where ϕ(x1 , · · · , xm−1 , 0) = (x1 , · · · , xm−1 ). The latter mapping is
orientation preserving if m is even and orientation reversing if m is odd. Under the
mapping j we have that

j∗ (ei ) = ei , i = 1, · · · , m − 1,

where the ei ’s on the left hand side are the unit vectors in Rm−1 , and the bold face
ek are the unit vectors in Rm . We have that the induced orientation for ∂Hm is
obtained by the rotation e1 → −em , em → e1 , and therefore

∂O = [e2 , · · · , em−1 , e1 ].

Under ϕ this corresponds with the orientation [e2 , · · · , em−1 , e1 ] of Rm−1 , which is
indeed the standard orientation for m even, and the opposite orientation for m odd.
The pullback form on ∂Hm , using the induced orientation on ∂Hm , is given by

( j∗ ω)(x1 ,··· ,xm−1 ) (e2 , · · · , em−1 , e1 )

= ω(x1 ,··· ,xm−1 ,0) j∗ (e2 ), · · · , j∗ (em−2 ), j∗ (e1 )
= (−1)m ωm (x1 , · · · , xm−1 , 0),

where we used the fact that

dx1 ∧ · · · ∧ dxm−1 j∗ (e2 ), · · · , j∗ (em−1 ), j∗ (e1 ))
= dx1 ∧ · · · ∧ dxm−1 (e2 , · · · , em−1 , e1 )
= (−1)m dx2 ∧ · · · ∧ dxm−1 ∧ dx1 (e2 , · · · , em−1 , e1 ) = (−1)m .

Combining this with the integral over Hm we finally obtain

Z Z
dω = j∗ ω,
Hm ∂Hm

which completes the proof.

92

Proof of Theorem 16.1: Let us start with the case that supp(ω) ⊂ U, where
(U, ϕ) is an oriented chart for M. Then by the definition of the integral we have
that
Z Z Z Z  
−1 ∗ −1 ∗
dω = (ϕ ) (dω) = (ϕ ) (dω) = d (ϕ−1 )∗ ω ,
M ϕ(U) Hm Hm

where the latter equality follows from Theorem 15.5. It follows from Lemma 16.2
that if supp(ω) ⊂ int (Hm ), then the latter integral is zero and thus M ω. = 0. Also,
R

using Lemma 16.2, it follows that if supp(ω) ∩ ∂Hm 6= ∅, then

Z Z   Z
dω = d (ϕ−1 )∗ ω = j0∗ (ϕ−1 )∗ ω
M Hm ∂Hm
Z Z
= (ϕ−1 ◦ j0 )∗ ω = j∗ ω,
∂Hm ∂M
where j0 : ∂Hm → Hm is the canonical inclusion as used Lemma 16.2, and j =
ϕ−1 ◦ j0 : ∂M → M is the inclusion of the boundary of M into M.
Now consider the general case. As before we choose a finite covering AI ⊂ A
of supp(ω) and an associated partition of unity {λi } subordinate to AI . Consider
the forms λi ω, and using the first part we obtain
Z Z Z
∗ ∗
j ω = ∑ j λi ω = ∑ d(λi ω)
∂M i ∂M i M
Z  
= ∑ dλi ∧ ω + λi dω
i M
Z   Z  
= d ∑ λi ∧ω+ ∑ λi dω
M i M i
Z
= dω,
M
which proves the theorem.
J 16.3 Remark. If M is a closed (compact, no boundary), oriented manifold mani-
fold, then by Stokes’ Theorem for any ω ∈ Γm−1 (M),
Z
dω = 0,
M
R
since ∂M = ∅. As a consequence volume form σ cannot be exact. Indeed, M σ > 0,
R R
and thus if σ were exact, then σ = dω, which implies that M σ = M dω = 0, a
contradiction. I
Let us start with the obvious examples of Stokes’ Theorem in R, R2 , and R3 .
J 16.4 Example. Let M = [a, b] ⊂ R, and consider the 0-form ω = f , then, since
M is compact, ω is compactly supported 0-form. We have that dω = f 0 (x)dx, and
Z b Z
f 0 (x)dx = f = f (b) − f (a),
a {a,b}
 93

which is the fundamental theorem of integration. Since on M any 1-form is exact

the theorem holds for 1-forms. I
J 16.5 Example. Let M = γ ⊂ R2 be a curve parametrized given by γ : [a, b] → R2 .
Consider a 0-form ω = f , then
Z Z
fx (x, y)dx + fy (x, y)dy = f = f (γ(b)) − f (γ(a)).
γ {γ(a),γ(b)}

Now let M = Ω ⊂ R2 , a closed subset of R2 with a smooth boundary ∂Ω, and

consider a 1-form ω = f (x, y)dx + g(x, y)dy. Then, dω = gx − fy )dx ∧ dy, and
Stokes’ Theorem yields
∂g ∂ f 
Z  Z
− dxdy = f dx + gdy,
Ω ∂x ∂y ∂Ω

also known as Green’s Theorem in R2 . I

J 16.6 Example. In the case of a curve M = γ ⊂ R3 we obtain the line-integral for
0-forms as before:
Z
fx (x, y, z)dx + fy (x, y, z)dy + fz (x, y, z) = f (γ(b)) − f (γ(a)).
γ

If we write the vector field

∂ ∂ ∂
F=f +g +h ,
∂x ∂y ∂z
then
∂f ∂ ∂f ∂ ∂f ∂
grad f = ∇ f = + + ,
∂x ∂x ∂y ∂y ∂z ∂z
and the above expression can be rewritten as
Z
∇ f · ds = f |∂γ .
γ

Next, let M = S ⊂ R3 be an embedded or immersed hypersurface, and let ω =

f dx + gdy + hdz be a 1-form. Then,
 ∂h ∂g   ∂ f ∂h   ∂g ∂ f 
dω = − dy ∧ dz + − dz ∧ dx + − dx ∧ dy.
∂y ∂z ∂z ∂x ∂x ∂y
Write the vector fields

 ∂h ∂g  ∂  ∂ f ∂h  ∂  ∂g ∂ f  ∂
curl F = ∇ × F = − + − + − .
∂y ∂z ∂x ∂z ∂x ∂y ∂x ∂y ∂z
Furthermore set    
dx dydz
ds =  dy  , dS =  dzdx  ,
   
dz dxdy
94

then from Stokes’ Theorem we can write the following surface and line integrals:
Z Z
∇ × F · dS = F · ds,
S ∂S
which is usually referred to as the classical Stokes’ Theorem in R3 . The version in
Theorem 16.1 is the general Stokes’ Theorem. Finally let M = Ω a closed subset
of R3 with a smooth boundary ∂Ω, and consider a a 2-form ω = f dy ∧ dz + gdz ∧
dx + hdx ∧ dy. Then
 ∂ f ∂g ∂h 
dω = + + dx ∧ dy ∧ dz.
∂x ∂y ∂z
Write
∂ f ∂g ∂h
div F = ∇ · F = + + , dV = dxdydz,
∂x ∂y ∂z
then from Stokes’ Theorem we obtain
Z Z
∇ · FdV = F · dS,
Ω ∂Ω
which is referred to as the Gauss Divergence Theorem. I
 95

V. De Rham cohomology

17. Definition of De Rham cohomology

In the previous chapters we introduced and integrated m-forms over manifolds
M. We recall that k-form ω ∈ Γk (M) is closed if dω = 0, and a k-form ω ∈ Γk (M)
is exact if there exists a (k − 1)-form σ ∈ Γk−1 (M) such that ω = dσ. Since d 2 = 0,
exact forms are closed. We define
Z k (M) = ω ∈ Γk (M) : dω = 0 = Ker(d),


Bk (M) = ω ∈ Γk (M) : ∃ σ ∈ Γk−1 (M) 3 ω = dσ = Im(d),



and in particular
Bk (M) ⊂ Z k (M).
The sets Z k and Bk are real vector spaces, with Bk a vector subspace of Z k . This
leads to the following definition.

Definition 17.1. Let M be a smooth m-dimensional manifold then the de Rham

cohomology groups are defined as
k
(16) HdR (M) := Z k (M)/Bk (M), k = 0, · · · m,
where B0 (M) := 0.

It is immediate from this definition that Z 0 (M) are smooth functions on M that
are constant on each connected component of M. Therefore, when M is con-
0 (M) ∼ R. Since Γk (M) = {0}, for k > m = dim M, we have
nected, then HdR =
k
that HdR (M) = 0 for all k > m. For k < 0, we set HdR k (M) = 0.

J 17.2 Remark. The de Rham groups defined above are in fact real vector spaces,
and thus groups under addition in particular. The reason we refer to de Rham
cohomology groups instead of de Rham vector spaces is because (co)homology
theories produce abelian groups. I
An equivalence class [ω] ∈ HdRk (M) is called a cohomology class, and two form

ω, ω0 ∈ Z k (M) are cohomologous if [ω] = [ω0 ]. This means in particular that ω and
ω0 differ by an exact form, i.e.
ω0 = ω + dσ.
Now let us consider a smooth mapping f : N → M, then we have that the pull-
back f ∗ acts as follows: f ∗ : Γk (M) → Γk (N). From Theorem 15.5 it follows that
 96

d ◦ f ∗ = f ∗ ◦ d and therefore f ∗ descends to homomorphism in cohomology. This

can be seen as follows:

d f ∗ ω = f ∗ dω = 0, and f ∗ dσ = d( f ∗ σ),

and therefore the closed forms Z k (M) get mapped to Z k (N), and the exact form
Bk (M) get mapped to Bk (N). Now define

f ∗ [ω] = [ f ∗ ω],

which is well-defined by

f ∗ ω0 = f ∗ ω + f ∗ dσ = f ∗ ω + d( f ∗ σ)

which proves that [ f ∗ ω0 ] = [ f ∗ ω], whenever [ω0 ] = [ω]. Summarizing, f ∗ maps

cohomology classes in HdR k (M) to classes in H k (N):
dR

f ∗ : HdR
k k
(M) → HdR (N),

Theorem 17.3. Let f : N → M, and g : M → K, then

g∗ ◦ f ∗ = ( f ◦ g)∗ : HdR
k k
(K) → HdR (N),

Moreover, id ∗ is the identity map on cohomology.

Proof: Since g∗ ◦ f ∗ = ( f ◦ g)∗ the proof follows immediately.

As a direct consequence of this theorem we obtain the invariance of de Rham

cohomology under diffeomorphisms.
k (M) ∼ H k (N).
Theorem 17.4. If f : N → M is a diffeomorphism, then HdR = dR
Proof: We have that id = f ◦ f −1 = f −1 ◦ f , and by the previous theorem

id∗ = f ∗ ◦ ( f −1 )∗ = ( f −1 )∗ ◦ f ∗ ,

and thus f ∗ is an isomorphism.

18. Homotopy invariance of cohomology

We will prove now that the de Rham cohomology of a smooth manifold M is
even invariant under homeomorphisms. As a matter of fact we prove that the de
Rham cohomology is invariant under homotopies of manifolds.
 97

Definition 18.1. Two smooth mappings f , g : N → M are said to be homotopic if

there exists a continuous map H : N × [0, 1] → M such that
H(p, 0) = f (p)
H(p, 1) = g(p),
for all p ∈ N. Such a mapping is called a homotopy from/between f to/and g. If
in addition H is smooth then f and g are said to be smoothly homotopic, and H is
called a smooth homotopy.

Using the notion of smooth homotopies we will prove the following crucial
property of cohomology:
Theorem 18.2. Let f , g : N → M be two smoothly homotopic maps. Then for k ≥ 0
it holds for f ∗ , g∗ : HdR
k (M) → H k (N), that
dR

f ∗ = g∗ .
J 18.3 Remark. It can be proved in fact that the above results holds for two homo-
topic (smooth) maps f and g. This is achieved by constructing a smooth homotopy
from a homotopy between maps. I
Proof of Theorem 18.2: A map h : Γk (M) → Γk−1 (N) is called a homotopy map
between f ∗ and g∗ if
(17) dh(ω) + h(dω) = g∗ ω − f ∗ ω, ω ∈ Γk (M).
Now consider the embedding it : N → N × I, and the trivial homotopy between
i0 and i1 (just the identity map). Let ω ∈ Γk (N × I), and define the mapping
Z 1
h(ω) = i ∂ ωdt,
0 ∂t

which is a map from Γk (N × I) → Γk−1 (N). Choose coordinates so that either

ω = ωI (x,t)dxI , or ω = ωI 0 (x,t)dt ∧ d x I 0 .
In the first case we have that i ∂ ω = 0 and therefore dh(ω) = 0. On the other hand
∂t
 ∂ω ∂ωI i 
I
h(dω) = h dt ∧ dxI + dx ∧ dxI
∂t ∂xi
Z 1
∂ωI  I
= dt dx
0 ∂t
ωI (x, 1) − ωI (x, 0) dxI = i∗1 ω − i∗0 ω,


=
which prove (17) for i∗0 and i∗1 , i.e.
dh(ω) + h(dω) = i∗1 ω − i∗0 ω.
98

In the second case we have

 ∂ω 0

I
h(dω) = h dxi ∧ dt ∧ dxI
∂xi
Z 1
∂ωI 0

= i ∂ dxi ∧ dt ∧ dxI dt
0 ∂dxi ∂xi
Z 1 ∂ω  0
I
= − dt dxi ∧ dxI .
0 ∂x i
On the other hand
Z 1  0
dh(ω) = d ωI 0 (x,t)dt dxI
0
∂  1
Z  0
= ωI 0 (x,t)dt dxi ∧ dxI
∂xi 0
Z 1 ∂ω  0
I
= dt dxi ∧ dxI
0 ∂xi
= −h(dω).
This gives the relation that
dh(ω) + h(dω) = 0,
and since i∗1 ω = i∗0 ω = 0 in this case, this then completes the argument in both
cases, and h as defined above is a homotopy map between i∗0 and i∗1 .
By assumption we have a smooth homotopy H : N × [0, 1] → M bewteen f and
e = h ◦ H ∗ . Using the
g, with f = H ◦ i0 , and g = H ◦ i1 . Consider the composition h
relations above we obtain
h(dω)
e + d h(ω)
e = h(H ∗ dω) + dh(H ∗ ω)
= h(d(H ∗ ω)) + dh(H ∗ ω)
= i∗1 H ∗ ω − i∗0 H ∗ ω
= (H ◦ i1 )∗ ω − (H ◦ i0 )∗ ω
= g∗ ω − f ∗ ω.
If we assume that ω is closed then
g∗ ω − f ∗ ω = dh(H ∗ ω),
and thus
0 = [dh(H ∗ ω)] = [g∗ ω − f ∗ ω] = g∗ [ω] − f ∗ [ω],
which proves the theorem.
J 18.4 Remark. Using the same ideas as for the Whitney embedding theorem one
can prove, using approximation by smooth maps, that Theorem 18.2 holds true for
continuous homotopies between smooth maps. I
 99

Definition 18.5. Two manifolds N and M are said to be homotopy equivalent, if

there exist smooth maps f : N → M, and g : M → N such that
g◦ f ∼
= idN , f ◦g ∼
= idM (homotopic maps).
We write N ∼ M., The maps f and g are homotopy equivalences are each other
homotopy inverses . If the homotopies involved are smooth we say that N and M
smoothly homotopy equivalent.

J 18.6 Example. Let N = S1 , the standard circle in R2 , and M = R2 \{(0, 0)}. We

have that N ∼ M by considering the maps
f = i : S1 ,→ R2 \{(0, 0)}, g = id/| · |.
Clearly, (g ◦ f )(p) = p, and ( f ◦ g)(p) = p/|p|, and the latter is homotopic to the
identity via H(p,t) = t p + (1 − t)p/|p|. I

Theorem 18.7. Let N and M be smoothly homotopically equivalent manifold, N ∼

M, then
k
HdR (N) ∼ k
= HdR (M),
and the homotopy equivalences f g between N and M, and M and N respectively
are isomorphisms.
As before this theorem remains valid for continuous homotopy equivalences of
manifolds.
 100

VI. Exercises

A number of the exercises given here are taken from the Lecture notes by J. Bochnak.

Manifolds

Topological Manifolds

+ 1 Given the function g : R → R2 , g(t) = (cos(t), sin(t)). Show that f (R) is a mani-
fold.

+ 2 Given the set T2 = {p = (p1 , p2 , p3 ) | 16(p21 + p22 ) = (p21 + p22 + p23 + 3)2 } ⊂ R3 ,
called the 2-torus.
(i ) Consider the product manifold S1 × S1 = {q = (q1 , q2 , q3 , q4 ) | q21 + q22 = 1, q23 +
q24 = 1}, and the mapping f : S1 × S1 → T2 , given by


f (q) = q1 (2 + q3 ), q2 (2 + q3 ), q4 .
2
Show that f is onto and f −1 (p) = pr1 , pr2 , r − 2, p3 , where r = |p| 4+3 .

(ii ) Show that f is a homeomorphism between S1 × S1 and T2 .

+ 3 (i ) Find an atlas for T2 using the mapping f in 2.

(ii ) Give a parametrization for T2 .

4 Show that
A4,4 = {(p1 , p2 ) ∈ R2 | p41 + p42 = 1},
is a manifold and A4,4 ∼
= S1 (homeomorphic).

+ 5 (i ) Show that an open subset U ⊂ M of a manifold M is again a manifold.

(ii ) Let N and M be manifolds. Show that their cartesian production N × M is also
a manifold.
 101

+ 6 Show that
(i ) {A ∈ M2,2 (R) | det(A) = 1} is a manifold.
(ii ) Gl(n, R) = {A ∈ Mn,n (R) | det(A) 6= 0} is a manifold.
(iii ) Determine the dimensions of the manifolds in (a) and (b).

7 Construct a simple counterexample of a set that is not a manifold.

8 Show that in Definition 1.1 an open set U 0 ⊂ Rn can be replaced by an open disc
D n ⊂ Rn .

9 Show that PRn is a Hausdorff space and is compact.

10 Define the Grassmann manifold Gk Rn as the set of all k-dimensional linear sub-
spaces in Rn . Show that Gk Rn is a manifold.

11 Consider X to be the parallel lines R × {0 ∪ R × {1. Define the equivalence relation

(x, 0) ∼ (x, 1) for all x 6= 0. Show that M = X/ ∼ is a topological space that satisfies
(ii) and (iii) of Definition 1.1.

+ 12 Let M be an uncountable union of copies of R. Show that M is a topological space

that satisfies (i) and (ii) of Definition 1.1.

13 Let M = {(x, y) ∈ R2 : xy = 0}. Show that M is a topological space that satisfies

(i) and (iii) of Definition 1.1.

Differentiable manifolds

+ 14 Show that cartesian products of differentiable manifolds are again differentaible

manifolds.

+ 15 Show that PRn is a smooth manifold.

16 Prove that PR is diffeomorphic to the standard unit circle in R2 by constructing a

diffeomorphism.

+ 17 Which of the atlases for S1 given in Example 2, Sect. 1, are compatible. If not
compatible, are they diffeomorphic?

+ 18 Show that the standard torus T2 ⊂ R3 is diffeomorphic to S1 × S1 , where S1 ⊂ R2

is the standard circle in R2 .

19 Prove that the taking the union of C∞ -atlases defines an equivalence relation (Hint:
use Figure 7 in Section 2).
 102

+ 20 Show that diffeomorphic defines an equivalence relation.

+ 21 Prove Theorem 2.15.

Immersion, submersion and embeddings

+ 22 Show that the torus the map f defined in Exer. 2 of 1.1 yields a smooth embedding
of the torus T2 in R3 .

23 Let k, m, n be positive integers. Show that the set

Ak,m,n := {(x, y, z) ∈ R3 : xk + ym + zn = 1},
is a smooth embedded submanifold in R3 , and is diffeomorphic to S2 when these
numbers are even.

24 Prove Lemma 3.18.

+ 25 Let f : Rn+1 \{0} → Rk+1 \{0} be a smooth mapping such that f (λx) = λd f (x),
d ∈ N, for all λ ∈ R\{0}. This is called a homogeneous mapping of degree d.
Show that fe : PRn → PRk , defined by fe([x]) = [ f (x)], is a smooth mapping.

+ 26 Show that the open ball Bn ⊂ Rn is diffeomorphic to Rn .

+ 27 (Lee) Consider the mapping f (x, y, s) = (x2 + y, x2 + y2 + s2 + y) from R3 to R2 .

Show that q = (0, 1) is a regular value, and f −1 (q) is diffeomorphic to S1 (stan-
dard).

28 Prove Theorem 3.24 (Hint: prove the steps indicated above Theorem 3.24).

29 Prove Theorem 3.26.

30 Let N, M be two smooth manifolds of the same dimension, and f : N → M is a

smooth mapping. Show (using the Inverse Function Theorem) that if f is a bijec-
tion then it is a diffeomorphism.

31 Prove Theorem 3.28.

+ 32 Show that the map f : Sn → PRn defined by f (x1 , · · · , xn+1 ) = [(x1 , · · · , xn+1 )] is
smooth and everywhere of rank n (see Boothby).

+ 33 (Lee) Let a ∈ R, and Ma = {(x, y) ∈ R2 : y2 = x(x − 1)(x − a)}.

(i ) For which values of a is Ma an embedded submanifold of R2 ?
(ii ) For which values of a can Ma be given a topology and smooth structure so that
Ma is an immersed submanifold?
 103

+ 34 Consider the map f : R2 → R given by f (x1 , x2 ) = x13 + x1 x2 + x23 . Which level sets
of f are smooth embedded submanifolds of R2 .

35 Let M be a smooth m-dimensional manifold with boundary ∂M (see Lee, pg. 25).
Show that ∂M is an (m − 1)-dimensional manifold (no boundary), and that there
exists a unique differentiable structure on ∂M such that i : ∂M ,→ M is a (smooth)
embedding.

+ 36 Let f : N → M be a smooth mapping. Let S = f −1 (q) ⊂ N is a smooth embedded

submanifold of codimension m. Is q necessarily a regular value of f ?

+ 37 Use Theorem 3.27 to show that the 2-torus T2 is a smooth embedded manifold in
R4 .

38 Prove Theorem 3.37.

39 Show that an m-dimensional linear subspace in R` is an m-dimensional manifold.

40 Prove that the annulus A = {(p1 , p2 ) ∈ R2 : 1 ≤ p21 + p22 ≤ 4} is a smooth manifold

two dimensional manifold in R2 with boundary.
 104

Tangent and cotangent spaces

Tangent spaces

+ 41 Let U ⊂ Rk , y ∈ Rm , and f : U → Rm a smooth function. By Theorem 3.28 M =

f −1 (y) = {p ∈ Rk : f (p) = y} is a smooth embedded manifold in Rk if rk(J f )| p =
m for all p ∈ M. Show that
Tp M ∼
= {Xp ∈ Rk : Xp · ∂ pi f = 0, i = 1, · · · , m} = ker J f | p ,
i.e. the tangent space is the kernel of the Jacobian. Using this identification from
now on, prove that
M p = p + Tp M ⊂ Rk
is tangent to M at p.

+ 42 Given the set M = {(p1 , p2 , p3 ) ∈ R3 : p31 + p32 + p33 − 3p1 p2 p3 = 1}.

(i ) Show that M is smooth embedded manifold in R3 of dimension 2.
(ii ) Compute Tp M, at p = (0, 0, 1).

+ 43 Given the set M = {(p1 , p2 , p3 ) ∈ R3 : p21 − p22 + p1 p3 − 2p2 p3 = 0, 2p1 − p2 +

p3 = 3}.
(i ) Show that M is smooth embedded manifold in R3 of dimension 1.
(ii ) Compute Tp M, at p = (1, −1, 0).

44 Prove Lemma 4.5.

+ 45 Express the following planar vector fields in polar coordinates:

∂ ∂
(i ) X = x ∂x + y ∂y ;
∂ ∂
(ii ) X = −y ∂x + x ∂y ;

(iii ) X = (x2 + y2 ) ∂x
∂
.

46 Find a vector field on S2 that vanishes at exactly one point.

47 Let S2 ⊂ R2 be the standard unit sphere and f : S2 → S2 a smooth map defined as

a θ-degree rotation around the polar axis. Compute f∗ : Tp S2 → T f (p) S2 and give a
matrix representation of f∗ in terms of a canonical basis.
 105

Cotangent spaces

+ 48 (Lee) Let f : R3 → R be given by f (p1 , p2 , p3 ) = |p|2 , and g : R2 → R3 by

 2x
1 2x2 |x|2 − 1 
g(x1 , x2 ) = , , .
|x|2 + 1 |x|2 + 1 |x|2 + 1
Compute both g∗ d f , and d( f ◦ g) and compare the two answers.

+ 49 (Lee) Compute d f in coordinates, and determine the set points where d f vanishes.
p1
(i ) M = {(p1 , p2 ) ∈ R2 : p2 > 0}, and f (p) = |p|2
— standard coordinates in R2 .

(ii ) As the previous, but in polar coordinates.

(iii ) M = S2 = {p ∈ R3 : |p| = 1}, and f (p) = p3 — stereographic coordinates.

(iv) M = Rn , and f (p) = |p|2 — standard coordinates in Rn .

+ 50 Express in polar coordinates

(i ) θ = x2 dx + (x + y)dy;

(ii ) θ = xdx + ydy + zdz.

51 Prove Lemma 5.3.

+ 52 (Lee) Let f : N → M (smooth), ω ∈ Λ1 (N), and γ : [a, b] → N is a smooth curve.

Show that Z Z
∗
f ω= ω.
γ f ◦γ

+ 53 Given the following 1-forms on R3 :

4z 2y 2x
α=− dx + dy + dz,
(x2 + 1)2 y2 + 1 x2 + 1
4xz 2y 2
ω=− dx + 2 dy + 2 dz.
(x2 + 1)2 y +1 x +1

R R
(i ) Let γ(t) = (t,t,t), t ∈ [0, 1], and compute γα and γ ω.

(ii ) Let γ be a piecewise smooth curve going from (0, 0, 0) to (1, 0, 0) to (1, 1, 0) to
(1, 1, 1), and compute the above integrals.

(iii ) Which of the 1-forms α and ω is exact.

(iv) Compare the answers.

 106

Vector bundles

+ 54 Let M ⊂ R` be a manifold in R` . Show that T M as defined in Section 6 is a smooth

embedded submanifold of R2` .

55 Let S1 ⊂ R2 be the standard unit circle. Show that T S1 is diffeomorphic to S1 × R.

56 Let M ⊂ R` be a manifold in R` . Show that T ∗ M as defined in Section 6 is a smooth

embedded submanifold of R2` .

57 Show that the open Möbius band is a smooth rank-1 vector bundle over S1 .
 107

Tensors

Tensors and tensor products

+ 58 Describe the standard inner product on Rn as a covariant 2-tensor.

+ 59 Construct the determinant on R2 and R3 as covariant tensors.

 
1 !
2
60 Given the vectors a =  3 , and b = . Compute a∗ ⊗ b and b∗ ⊗ a.
 
4
0
+ 61 Given finite dimensional vector spaces V and W , prove that V ⊗W and W ⊗V are
isomorphic.

62 Given finite dimensional vector spaces U, V and W , prove that (U ⊗ V ) ⊗ W and

U ⊗ (V ⊗W ) are isomorphic.

+ 63 Show that V ⊗ R ' V ' R ⊗V .

Symmetric and alternating tensors

+ 64 Show that for T ∈ T s (V ) the tensor Sym T is symmetric.

65 Prove that a tensor T ∈ T s (V ) is symmetric if and only if T = Sym T .

66 Show that the algebra

∞
Σ∗ (V ) =
M
Σk (V ),
k=0
is a commutative algebra.

+ 67 Prove that for any T ∈ T s (V ) the tensor Alt T is alternating.

68 Show that a tensor T ∈ T s (V ) is alternating if and only if T = Alt T .

     
1 0 1
69 Given the vectors a =  2 , b =  1 , and c =  1 . Compute a ∧ b ∧ c,
     
4 3 2
and compare this with det(a, b, c).

70 Given vectors a1 , · · · , an show that a1 ∧ · · · ∧ an = det(a1 , · · · , an ).

 108

71 Prove Lemma 8.11.

Tensor bundles and tensor fields

72 Let M ⊂ R` be an embedded m-dimensional manifold. Show that T r M is a smooth

r
manifold in R`+` .

s
73 Similarly, show that Ts M is a smooth manifold in R`+` , and Tsr M is a smooth
r s
manifold in R`+` +` .

74 Prove that the tensor bundles introduced in Section 7 are smooth manifolds.

+ 75 One can consider symmetric tensors in Tp M as defined in Section 8. Define and

describe Σr (Tp M) ⊂ T r (Tp M) and Σr M ⊂ T r M.

+ 76 Describe a smooth covariant 2-tensor field in Σ2 M ⊂ T 2 M. How does this relate to

an (indefinite) inner product on Tp M?

77 Prove Lemma 9.2.

+ 78 Given the manifolds N = M = R2 , and the smooth mapping

q = f (p) = (p21 − p2 , p1 + 2p32 ),

acting from N to M. Consider the tensor spaces T 2 (Tp N) ∼

= T 2 (Tq M) ∼
= T 2 (R2 )
and compute the matrix for the pullback f ∗ .

+ 79 In the above problem consider the 2-tensor field σ on M, given by

σ = dy1 ⊗ dy2 + q1 dy2 ⊗ dy2 .

(i ) Show that σ ∈ F2 (M).

(ii ) Compute the pulback f ∗ σ and show that f ∗ σ ∈ F2 (N).

(iii ) Compute f ∗ σ(X,Y ), where X,Y ∈ F(N).

 109

Differential forms

+ 80 Given the differential form σ = dx1 ∧ d x2 − dx2 ∧ dx3 on R3 . For each of the fol-
lowing vector fields X compute iX σ:
∂
(i ) X = ∂x1 + ∂x∂2 − ∂x∂3 ;
(ii ) X = x1 ∂x∂1 − x22 ∂x∂3 ;
(iii ) X = x1 x2 ∂x∂1 − sin(x3 ) ∂x∂2 ;

+ 81 Given the mapping

f (p) = (sin(p1 + p2 ), p21 − p2 ),
acting from R2 to R2 and the 2-form σ = p21 dx1 ∧ dx2 . Compute the pullback form
f ∗ σ.

+ 82 Prove Lemma 10.2.

+ 83 Derive Formula (10).

Orientations

+ 84 Show that Sn is orientable and give an orientation.

85 Show that the standard n-torus Tn is orientable and find an orientation.

+ 86 Prove that the Klein bottle and the projective space PR2 are non-orientable.

+ 87 Give an orientation for the projective space PR3 .

88 Prove that the projective spaces PRn are orientable for n = 2k + 1, k ≥ 0.

89 Show that the projective spaces PRn are non-orientable for n = 2k, k ≥ 1.
 110

Integration on manifolds

Integrating m-forms on Rm

+ 90 Let U ⊂ Rm be open and let K ⊂ U be compact. Show that there exists a domain
of integration D such that K ⊂ D ⊂ Rm .

91 Show that Definition 12.2 is well-posed.

Partitions of unity

92 Show that the function f1 defined in Lemma 13.3 is smooth.

+ 93 If U is open covering of M for which each set Ui intersects only finitely many other
sets in U, prove that U is locally finite.

94 Give an example of uncountable open covering U of the interval [0, 1] ⊂ R, and a

countable refinement V of U.

Integration of m-forms on m-dimensional manifolds

+ 95 Let S2 = ∂B3 ⊂ R3 oriented via the standard orientation of R3 , and consider the
2-form
ω = xdy ∧ dz + ydz ∧ dx + zdx ∧ dy.
Given the parametrization
   
x sin(ϕ) cos(θ)
F(ϕ, θ) =  y  =  sin(ϕ) sin(θ)  ,
   
z cos(ϕ)

for S2 , compute
R
S2 ω.

+ 96 Given the 2-form ω = xdy ∧ dz + zdy ∧ dx, show that

Z
ω = 0.
S2
 111

+ 97 Consider the circle S1 ⊂ R2 parametrized by

! !
x r cos(θ)
F(θ) = = .
y r sin(θ)

Compute the integral over S1 of the 1-form ω = xdy − ydx.

98 If in the previous problem we consider the 1-form

x y
ω= p dy − p dx.
x 2 + y2 x2 + y2
represents the induced Euclidian length of S1 .
R
Show that S1 ω

+ 99 Consider the embedded torus

T2 = {(x1 , x2 , x3 , x4 ) ∈ R4 : x12 + x22 = 1, x32 + x42 = 1}.

Compute the integral over T2 of the 2-form

ω = x12 dx1 ∧ dx4 + x2 dx3 ∧ dx1 .

+ 100 Consider the following 3-manifold M parametrized by g : [0, 1]3 → R4 ,

 
  r
r  s 
 s  7→  .
   
 t 
t 2
(2r − t)
Compute Z
x2 dx2 ∧ dx3 ∧ dx4 + 2x1 x3 dx1 ∧ dx2 ∧ dx3 .
M

The exterior derivative

+ 101 Let (x, y, z) ∈ R3 . Compute the exterior derivative dω, where ω is given as:

(i ) ω = exyz ;

(ii ) ω = x2 + z sin(y);

(iii ) ω = xdx + ydy;

(iv) ω = dx + xdy + (z2 − x)dz;

(v) ω = xydx ∧ dz + zdx ∧ dy;

(vi ) ω = dx ∧ dz.
 112

+ 102 Which of the following forms on R3 are closed?

(i ) ω = xdx ∧ dy ∧ dz;
(ii ) ω = zdy ∧ dx + xdy ∧ dz;
(iii ) ω = xdx + ydy;
(iv) ω = zdx ∧ dz.

+ 103 Verify which of the following forms ω on R2 are exact, and if so write ω = dσ:
(i ) ω = xdy − ydx;
(ii ) ω = xdy + ydx;
(iii ) ω = dx ∧ dy;
(iv) ω = (x2 + y3 )dx ∧ dy;

+ 104 Verify which of the following forms ω on R3 are exact, and if so write ω = dσ:
(i ) ω = xdx + ydy + zdz;
(ii ) ω = x2 dx ∧ dy + z3 dx ∧ dz;
(iii ) ω = x2 ydx ∧ dy ∧ dz.

105 Verify that on R2 and R3 all closed k-forms, k ≥ 1, are exact.

106 Find a 2-form on R3 \{(0, 0, 0)} which is closed but not exact.

Stokes’ Theorem

107 Let Ω ⊂ R3 be a parametrized 3-manifold,i.e. a solid, or 3-dimensional domain.

Show that the standard volume of Ω is given by
1
Z
Vol(M) = xdy ∧ dz − ydx ∧ dz + zdx ∧ dy.
3 ∂Ω

108 Let Ω ⊂ Rn be an n-dimensional domain. Prove the analogue of the previous prob-
lem.

+ 109 Prove the ‘integration by parts’ formula

Z Z Z
f dω = fω− d f ∧ ω,
M ∂M M
where f is a smooth function and ω a k-form.

110 Compute the integral

Z
x2 ydx ∧ dz + x3 dy ∧ dz + (z − 2x2 )dx ∧ dy,
S2
 113

where S2 ⊂ R3 is the standard 2-sphere.

111 Use the standard polar coordinates for the S2 ⊂ R3 with radius r to compute
xdy ∧ dz − ydx ∧ dz + zdx ∧ dy
Z
,
S2 (x2 + y2 + z2 )3/2
and use the answer to compute the volume of the r-ball Br .

Extra’s

112 Use the examples in Section 16 to show that

curl grad f = 0, and div curl F = 0,

where f : R3 → R, and F : R3 → R3 .

+ 113 Given the mapping f : R2 → R3 , ω = dx ∧ dz, and

 
cos(s) sin(t)
 √2 2 
f (s,t) =  s +t .
st
Compute f ∗ ω.

+ 114 Given the mapping f : R3 → R3 , ω = xydx ∧ dy ∧ dz, and

 
s cos(t)
f (s,t, u) =  s sin(t)  .
 
u
Compute f ∗ ω.

+ 115 Let M = R3 \{(0, 0, 0)}, and define the following 2-form on M:

1

ω= xdy ∧ dz + ydz ∧ dx + zdx ∧ dy .
(x2 + y2 + z2 )3/2
R2
(i ) Show that ω is closed (dω = 0) (Hint: compute S ω).
(ii ) Prove that ω is not exact on M!
On N = R3 \{x = y = 0} ⊃ M consider the 1-form:
−z xdy − ydx
η= .
(x2 + y2 + z2 )1/2 x2 + y2

(iii ) Show that ω is exact as 2-form on N, and verify that dη = ω.

 114

+ 116 Let M ⊂ R4 be given by the parametrization

 
s
 t +u 
(s,t, u) 7→  , (s,t, u) ∈ [0, 1]3 .
 
 t 
s−u
R
(i ) Compute M dx1 ∧ dx2 ∧ dx4 .
2
R
(ii ) Compute M x1 x3 dx1 ∧ dx2 ∧ dx3 + x3 dx2 ∧ dx3 ∧ dx4 .

+ 117 Let C = ∂∆ be the boundary of the triangle OAB in R2 , where O = (0, 0), A =
(π/2, 0), and B = (π/2, 1). Orient the traingle by traversing the boundary counter
clock wise. Given the integral
Z
(y − sin(x))dx + cos(y)dy.
C
(i ) Compute the integral directly.
(ii ) Compute the integral via Green’s Theorem.

· dS, where Σ = ∂[0, 1]3 , and

R
118 Compute the integral ΣF
 
4xz
F(x, y, z) =  −y2 
 
yz
(Hint: use the Gauss Divergence Theorem).
 115

De Rham cohomology

Definition of De Rham cohomology

119 Prove Theorem 17.3.

S
120 Let M = j M j be a (countable) disjoint union of smooth manifolds. Prove the
isomorphism
k
HdR (M) ∼ k
= ∏ HdR (M j ).
j

k (M)×H ` (M) → H k+` (M), called the cup-product,

Show that the mapping ∪ : HdR
121 dR dR
and defined by
[ω] ∪ [η] := [ω ∧ η],

is well-defined.

122 Let M = S1 , show that

0
HdR (S1 ) ∼
= R, 1
HdR (S1 ) ∼
= R, k
HdR (S1 ) = 0,

for k ≥ 2.

123 Show that

0
HdR (R2 \{(0, 0)}) ∼
= R, 1
HdR (R2 \{(0, 0)}) ∼
= R, k
HdR (R2 \{(0, 0)}) = 0,

for k ≥ 2.

1 (R2 \{(0, 0)}).

Find a generator for HdR
124

125 Compute the de Rham cohomology of the n-torus M = Tn .

126 Compute the de Rham cohomology of M = S2 .

 116

VII. Solutions

1 Note that (x, y) ∈ g(R) if and only if x2 + y2 = 1. So g(R) is just the circle S1 in
the plane. The fact that this is a manifold is in the lecture notes.

2 We will show that f has an inverse by showing that the funcrion suggested in the
exercise is indeed the inverse of f . It follows that f is a bijection and therefore it
is onto. By continuity of f and its inverse, it follows that it is a homeomorphism.
We first show that f maps S1 × S1 into T2 : We use the parametrisation of S1 × S1
given by
(s,t) 7→ (cos s, sin s, cost, sint),

where s,t ∈ [0, 2π). Suppose that q = (cos s, sin s, cost, sint) ∈ S1 × S1 . Then

f (q) = (cos s(2 + cost), sin s(2 + cost), sint),

so f (q) ∈ T2 if and only if LHS = RHS where

(LHS) = 16(cos2 s(2 + cost)2 + sin2 s(2 + cost)2 ),

(RHS) = (cos2 s(2 + cost)2 + sin2 s(2 + cost)2 + sin2 t + 3)2 .

We have

LHS = 16(2 + cost)2

RHS = ((2 + cost)2 + sin2 t + 3)2
= (4 + 4 cost + cos2 t + sin2 t + 3)2
= (4(2 + cost)2 = 16(2 + cost).

This shows that f maps S1 × S1 into T2 . Now we let g : T2 → S1 × S1 be the

suggested inverse; i.e.
p p 
1 2
g(p) = , , r − 2, p3 ,
r r
2
where r = |p| 4+3 . We will first show that g maps T2 into S1 × S1 . So we let p ∈ T2
and we will show that g(p) ∈ S1 × S1 . But this is the case if and only if p21 + p22 = r2
and (r − 2)2 + p23 = 1. The first equality follows from the definition of r and the
 117

fact that p ∈ T2 . For the second equality we note that

(r − 2)2 + p23 = r2 − 4r + 4 + p23
= p21 + p22 − 4r + 4 + p23
= (p21 + p22 + p23 + 3) − (|p|2 + 3) + 1
= 1.
This shows that g maps T2 into S1 × S1 . We will now show that g is the inverse of
f ; let p ∈ T2 and q ∈ S1 × S1 . Then
p p 
1 2
f ◦ g(p) = f , , r − 2, p3
 p r rp 
1 2
= · r, · r, p3 = p
r r
g ◦ f (q) = g(q1 (2 + q3 ), q2 (2 + q3 ), q4 )
 q (2 + q ) q (2 + q ) 
1 3 2 3 0
= , , r − 2, q 4
r0 r0
q1 (2 + q3 ) + q2 (2 + q3 )2 + q24 + 3
2 2 2
where r0 =
4
= /4 · (2 + q3 ) + q24 + 3
2

1

= 1/4 · 4 + 4q3 + q23 + q24



= 1/4 · 4(2 + q3 ) = 2 + q3
and therefore g ◦ f (q) = (q1 , q2 , q3 , q4 ) = q.
This show that f is a bijection, so in particular it is onto. Since f and g are both
continuos, it follows that f is a homeomorphism.

3 (i ) Given co-ordinate charts (U, ϕ) and (V, ψ) for S1 , we define a new co-ordinate
chart for T2 as follows; W = f (U) × f (V ), ξ : W → ϕ(U) × ψ(V ) is defined by
ξ(p) = (ϕ(x), ψ(y)) where (x, y) = f −1 (p) and p ∈ W . This shows how to make
an atlas for T2 using an atlas for S1 .
(ii ) A parametrisation for S1 × S1 was already given in the solution of 2. Using the
map f , we obtain the following parametrisation for T2 ;


(s,t) 7→ cos s · (2 + cost), sin s · (2 + cost), sint ,
where s,t ∈ [0, 2π).

5 (i ) If (V, ψ) is a co-ordinate chart for M, then let W = U ∩V . Let ξ : W → ψ(W )

be defined by ξ(x) = ψ(x), i.e. ξ is just the restriction of ψ to W . Then clearly ξ
is a homeomorphism from W to an open subset of ψ(V ). In particular, if ψ(V ) is
m-dimensional, then so is ξ(W ). This shows how to make an atlas for U, given an
atlas for M.
 118

(ii ) Let (U, ϕ) and (V, ψ) be charts for N and M respectively. The map ξ : U ×V →
ϕ(U) × ψ(V ) is defined by ξ(x, y) = (ϕ(x), ψ(y)). This is clearly a homeomor-
phism. This shows how to make an atlas for N × M, given atlases for N and M.
Note that if N is n-dimensional and M is m-dimensional, then N × M is n + m-
dimensional.

6 (i ) We always let A be a matrix of the form

!
a b
c d

We let M = {A ∈ M2,2 (R)| det A = 1} and U = {A ∈ M : a 6= 0}. Note that U is an

open subset of M. We now define ϕ : U → (R \ {0}) × R × R by

ϕ(A) = (a, b, c).

The inverse of ϕ is given by

!
x y
ϕ−1 (x, y, z) = 1+yz .
z x

Cleary, ϕ and its inverse are continuous, and therefore ϕ is a homeomorphism. So

(ϕ,U) is a co-ordinate chart for M. In a similar way we can define charts on sets
in M where b 6= 0, c 6= 0 and d 6= 0. These four sets cover all of M and so we
obtain an atlas for M which shows that M is a manifold. Since the range of ϕ is
3-dimensional, it follows that M has dimension 3.

(ii ) Note that det : Mn,n (R) → R is a continuous function and that Gl(n, R) =
det−1 (R \ {0}). It follows that Gl(n, R) is an open subset of the space Mn,n (R).
2
But Mn,n (R) is homeomorphic to Rn , which is a manifold, and therefore Gl(n, R)
is an open subset of a manifold. It follows from 5 that this space is a manifold. The
dimension of Gl(n, R) is n2 .

12 The space M is not second countable; observe that every second countable space
is separable. To see that M is not separable, note that it contains an uncountable
collection of non-empty pairwise disjoint open subsets.

14 Let M and N be differentiable manifolds with maximal atlases A = {(Ui , ϕi )}i∈I

and B = {(Vk , ψk )}k∈K respectively. A typical co-ordinate chart for M × N is given
by ϕi × ψk : Ui ×Vk → Rm+n .

Now, for i, j ∈ I and k, l ∈ K, consider the transition map (ϕi × ψk ) ◦ (ϕ j × ψl )−1 .

This os just the product of the maps ϕi ◦ ϕ−1 −1
j and ψ j ◦ ψk . Since these maps are
diffeomorphisms, their product is also a diffeomorphisms.
 119

This shows that the atlas for M × N given by {(Ui × Vk , ϕi × ψk )i∈I,k∈K is a C∞ -

atlas. Taking a maximal extension of this atlas endows M × N with a differentiable
structure.

15 As usual, π : Rn+1 \ {0} → PRn is the quotient mapping, given by π(x) = [x]. For
i = 1, . . . , n + 1, we have Vi = {x ∈ Rn+1 \ {0} : xi 6= 0} and Ui = π(Vi ). For [x] ∈ Ui
we define x
1 xi−1 xi+1 xn+1 
ϕi ([x]) = ,..., , ,..., ,
xi xi xi xi
and its inverse is given by
ϕ−1
i (z1 , . . . , zn ) = [(z1 , . . . , zi−1 , 1, zi , . . . , zn )].

We will show that for i < j, the transition map ϕi ◦ ϕ−1

j : ϕ j (Ui ∩U j ) → ϕi (Ui ∩Ui )
−1
is a diffeomorphism. We compute ϕi ◦ ϕ j ;

ϕi ◦ ϕ−1
j (z1 , . . . , zn ) = ϕi ([z1 , . . . , z j−1 , 1, z j , . . . , zn ])
z
1 zi−1 zi+1 z j−1 1 z j zn 
= ,..., , ,..., , , ,... .
zi zi zi zi zi zi zi
This function is clearly a diffeomorphisms. Note that
ϕi (Ui ∩U j ) = {z ∈ Rn : z j−1 6= 0},
ϕ j (Ui ∩U j ) = {z ∈ Rn : zi 6= 0}.

17 All the atlases are compatible. We will compute some of the transition maps. We
list three co-ordinate charts, one from each atlas;

(1) ϕ : U → U 0 , where U = {p ∈ S1 : p2 > 0} and U 0 = (−1, 1).

p
ϕ(p) = p1 and ϕ−1 (x) = (x, 1 − x2 ).
(2) ψ : V → V 0 , where V 0 = {p ∈ S1 : p 6= (0, 1)} and V 0 = R.
2p1  4x x2 − 4 
ψ(p) = and ψ−1 (x) = 2 , .
1 − p2 x + 4 x2 + 4
(3) ξ : W → W 0 , where W = {p ∈ S1 : p 6= (1, 0)} and W 0 = (0, 2π).
ξ(p) = arccos p1 and ξ−1 (θ) = (cos θ, sin θ).

Now we compute some of the transition maps;

(1) The map ψ ◦ ϕ−1 : (−1, 1) \ {0} → (−∞, −2) ∪ (2, ∞) is given by
2x
x 7→ √ .
1 − 1 − x2
 120

(2) The map ϕ ◦ ψ−1 : (−∞, −2) ∪ (2, ∞) → (−1, 1) \ {0} is given by
4x
x 7→ 2 .
x +4
(3) The map ϕ ◦ ξ−1 : (0, π) → (−1, 1) is given by
θ 7→ cos θ.

All the functions mentioned above are C∞ -functions. If we check this for all tran-
sition maps, then we have shown that the atlases are compatible.

18 We have proved in 15 that the product of differentiable manifolds is again a dif-

ferentiable manifolds. The C∞ -atlas constructed there gives a C∞ -atlas for S1 × S1 .
We have seen in 2 that S1 × S1 is homeomorphic to the torus T2 and the map
f : S1 × S1 → T2 is a homeomorphism. In this exercise we are asked to show that f
is even a diffeomorphism. The differentiable structure of T2 is the structure which
it inherits from R3 .
Note that using the map f we can endow T2 with a differentiable structure inherited
from S1 ×S1 . So we have two differentiable structures on T2 ; the structure inherited
from R3 and the structure inherited from S1 × S1 . This exercise amounts to saying
that T2 endowed with the structure inherited from S1 × S1 is a smooth embedded
submanifold of R3 .
Fixing appropriate charts for S1 × S1 and T2 one can show that f is a diffeomor-
phism. For example, consider U ⊂ S1 × S1 given by
{q ∈ S1 × S1 | q1 > 0 & q3 > 0 }
with ϕ : U → (−1, 1) × (−1, 1) given by ϕ(q) = (q2 , q4 ). For T2 we fix a co-
ordinate chart (V, ψ) such that V ⊂ f (U) and ψ is a projection onto the 2nd and 3rd
co-ordinate. Computing ψ ◦ f ◦ ϕ−1 gives;
p
ψ ◦ f ◦ ϕ−1 (x, y) = (x(2 + 1 − y2 ), y).
This map is a diffeomorphism. For other co-ordinate charts this is similar.

20 Let M, N and O be differentiable manifolds. To show that ‘diffeomorphic’ is an

equivalence relation, we prove reflexivity, symmetry and transitivity.
For reflexivity: M is diffeomorphic to M since the transition maps are diffeomor-
phism. So the identity is a diffeomorphism from M onto M.
For symmetry: The inverse of a diffeomorphism is a diffeomorphism.
For transitivity: Suppose f : M → N and g : N → O are diffeomorphisms and let
p ∈ M. Fix charts (U, ϕ), (V, ψ), (V 0 , ψ0 ) and (W, ξ) such that p ∈ U, f (p) ∈
V ∩ V 0 and g ◦ f (p) ∈ W . Since f and g are diffeomorphisms, we may asssume
 121

that ψ ◦ f ◦ ϕ−1 and ξ ◦ g ◦ (ψ0 )−1 are diffeomorphims. Also note that since N is a
differentiable manifold, the map ψ0 ◦ ψ−1 is a diffeomorphism. Now note that
ξ ◦ g ◦ f ◦ ϕ−1 = (ξ ◦ g ◦ (ψ0 )−1 ) ◦ (ψ0 ◦ ψ−1 ) ◦ (ψ ◦ f ◦ ϕ−1 ).
The map on the right hand side is a composition of diffeomorphisms, so it is again a
diffeomorphism. It follows that the map on the left hand side is a diffeomorphism.
Since p ∈ M was arbitrary, this show that the composition of diffeomorphisms
between manifolds is again a diffeomorphism; i.e. M and O are diffeomorphic. In
this proof we have implicitly used part (iii) of Theorem 2.14.

21 See Lee, pg 33&34.

22 Recall that f : S1 × S1 → R3 . We fix a co-ordinate chart (U, ϕ) on S1 × S1 where

U = {q = (q1 , q2 , q3 , q4 )|q2 > 0 & q4 > 0},
and ϕ : U → V is defined by ϕ(q1 , q2 , q3 , q4 ) = (q1 , q3 ) where V = (−1, 1)×(−1, 1).
The chart (ψ, R3 ) is just the identity. We express f in local co-ordinates, i.e.
f˜ = ψ ◦ f ◦ ϕ−1 where
p p
f˜(x, y) = x(1 + y), 1 − x2 (2 + y), 1 − y2 .

We have that
 
1+y x
J f˜x=ϕ(p) =  −x(2 + y)(1 − x2 )1/2 (1 − x2 )1/2 
 
0 −y(1 − y2 )1/2

It is not hard to verify that the matrix J f˜x=ϕ(p) has rank 2 for all p ∈ U. For example,
if p = ϕ−1 (0, 0), then
 
1 0
J f˜(0,0) =  0 1 
 
0 0
This shows that f is of rank 2 at all p ∈ U. Of course, we can also prove this for
other similar charts. So it follows that rk( f ) = 2, so it is an immersion. Since f is
a homeomorphism onto its image, it follows that f is a smooth embedding.

25 We compute g in local co-ordinates. So fix the usual charts (Ui , ϕ) and (U j , ψ j )

for PRn and PRk respectively. For z ∈ Rn , we let zi be the point in Rn+1 given by
(z1 , . . . , zi−1 , 1, zi , . . . zn ). If f j (zi ) 6= 0, then
g̃(z) = ψ j ◦ g ◦ ϕ−1 i
i (z) = ψ([ f (z )])
 f (zi )
1 f j−1 (zi ) f j+1 (zi ) fn (zi ) 
= , . . . , , , . . . , .
f j (zi ) f j (zi ) f j (zi ) f j (zi )
 122

By assumption, f (zi ) 6= 0, so there is some j such that f j (zi ) 6= 0. The fact that g̃ is
smooth follows from the smoothness of all co-ordinate functions fk of f .

26 The mapping f : Bn → Sn defined by

1
f (x) = · x,
1 − |x|
is a diffeomorphism. Note that one can chose charts for Bn and Rn such that the co-
ordinate maps are just the identity. So the expression f˜ of f in local co-ordinates
is equal to f .

27 We first note that !

2x 1 0
J f(x,y,s) =
2x 2y + 1 2s
Now consider a point p = (x, y, s) ∈ f −1 (q). Then x2 + y = 0 and y2 + s2 = 1. If
s 6= 0, then the last two columns of J f p are independent. If s = 0, then y ∈ {−1, 1}
and since x2 + y = 0, we have in fact that y = −1 and x2 = 1. In this case the first
two colums on J f p are independent. Sow e have shown that if p ∈ f −1 (q), then
rk(J f p ) = 2 and it follows that q is a regular value.

To show that f −1 (q) is diffeomorphic to S1 , consider the set

A = {(x, y, s) ∈ R3 | x4 + s2 = 1 & y = 0}.

Then A is diffeomorphic to S1 , see 23. Now consider the map g : A → R3 defined

by
f (x, y, s) = (x, −x2 , s).
Then g is a constant rank map with rk(g) = 1 and g is injective. So g is a smooth
embedding of A (and thus of S1 ) into R3 . It follows that g[A] = f −1 (q) is diffeo-
morphic to S1 .

32 Let Ui = {x ∈ Sn : xi > 0} and ϕ be the projection on all co-ordinates but the ith .
We have that q
ϕ−1 (z) = (z1 , . . . , zi−1 , 1 − |z|2 , zi , . . . , zn ).
where |z| < 1. Next we let Vi = π[Ui ] and
x xi−1 xi+1 xn+1 
1
ψi ([x]) = ,..., , ,..., .
xi xi xi xi
We calculate f in local co-ordinates, so f˜ = ψ ◦ f ◦ ϕ−1 ,
1
f˜(z) = p · (z1 , . . . , zi−1 , zi+1 , . . . , zn ).
1 − |z|2
 123

We note that
1+2zi

 (1−|z|2 )3/2
i= j
∂ zj 
p =
∂zi 1 − |z|2  2zi

(1−|z|2 )3/2
i 6= j
so  
1 + 2z1 2z1 ··· 2z1
2z2 1 + 2z2 · · · 2z2
 

3/2 
J f˜z = 1 − |z|2

 .. .. .. .. 

 . . . .


2zn ··· 2zn 1 + 2zn
For all z with |z| < 1, this matrix has rank n.

33 We let fa : R2 → R be given by fa (x, y) = y2 − x(x − 1)(x − a) and note that:

fa (x, y) = y2 − x3 + (a + 1)x2 − ax.

Of course, Ma = f −1 (0). We also have that:

J f˜a |(x,y) = (−3x2 + 2(a + 1)x − a, 2y).

We compute the rank of this matrix for various values of a and (x, y) ∈ Ma ;

If x 6∈ {0, 1, a} : then y 6= 0 → the rank is 1

x = 0 : −3x2 + 2(a + 1)x − a = −a → the rank is 1 iff a 6= 0

2
x = 1 : −3x + 2(a + 1)x − a = a − 1 → the rank is 1 iff a 6= 1
2
x = a : −3x + 2(a + 1)x − a = a(a − 1) → the rank is 1 iff a(a − 1) = 0

So we conclude that if a 6∈ {0, 1}, then Ma is an embedded submanifold of R2 .

For (ii ) : this is possible for all a ∈ R.

34 So the question really is; which values q ∈ R are regular values of f ? For this, we
compute the Jacobian;

J f˜|(x,y) = (3x2 + y, 3y2 + x).

Let q ∈ R and (x, y) ∈ f −1 (q). The rank of J f |(x,y) is 0 if and only if:

3x2 + y = 3y2 + x = 0.

In this case we have

y = −3x2
y2 = −x/3
 124

It follows from these equations that y3 = x3 and therefore x = y. Furthermore:

0 = 3x2 + y = 3(−3y2 )2 + y =
= 27y4 + y = y(27y3 + 1).

We conclude that (x, y) is either (0, 0) or (−1/3, −1/3). So q ∈ {0, 1/27}.

We conclude that if q 6∈ {0, 1/27}, then q is a regular value of f and therefore

f −1 (q) is a smooth embedded submanifold of R2 . The level sets corresponding to
q = 0 and q = 1/27 are not embedded submanifolds.

36 NO; consider the mapping f : R3 → R2 given by f (x, y, z) = (x2 + y2 , z2 ). We let

q = (1, 0) and S = f −1 (q). Note that

S = {(x, y, z) ∈ R3 |x2 + y2 = 1 & z = 0}.

So S is just a circle embedded in R3 . The co-dimension of S is 2. The Jacobian of

f in p = (x, y, z) is given by
!
2x 2y 0
J f |p =
0 0 2z

So for example, at the point p = (1, 0, 0) ∈ S, this matrix is of rank 1 and not of
rank 2. So q is not a regular value of f .

37 Let N = R4 \ {p|p21 + p22 = p23 + p24 = 0} and consider the mapping f : N → R2

given by f (p) = (p21 + p22 , p23 + p24 ). The Jacobian of f at the point p is given by
!
2p1 2p2 0 0
J f |p =
0 0 2p3 2p4

On N, this matrix is always of rank 2, so f is a constant rank mapping with rk( f ) =

2. The 2-torus in R4 is the level set f −1 (1, 1), so it follows from Theorem 3.27 that
this space is a smooth embedded submanifold of N. Since N is an open subset of
R4 , it follows that N, and hence the 2-torus, is a smooth embedded submanifold of
R4 .

41 Fix p ∈ f −1 (y) and as in the proof of Theorem 3.28, construct a map g : Rn →

Rn−m × Rm where
g(ξ) = (Lξ, f (ξ) − y),
where L : Rn → Rn−m is invertible on ker J f | p ⊂ Rn . We have that

Jg| p = L ⊕ J f | p ,
 125

and this map is onto. As in the proof of Theorem 3.28, g is invertible on a neigh-
bourhood V of (L(p), 0) such that
g−1 : Rn−m × {0} ∩V → f −1 (y).

This gives a local co-ordinate map ϕ with

ϕ−1 : Rn−m ∩V |n−m → f −1 (y) ∩U,
where ϕ(ξ) = Lξ for all ξ ∈ f −1 (y) ∩U. It follows that ϕ−1 = L−1 and
Jϕ−1 | p = JL−1 | p = L−1 .
So we have that JL−1 | p (Rn−m ) = ker J f | p by construction, so it follows that
Tp M = ker J f | p ,
and this is what we wanted to show.

42 Let f : R3 → R be defined by f (p1 , p2 , p3 ) = p31 + p32 + p33 − 3p1 p2 p3 , so that

M = f −1 (1). The Jacobian of f at the point p is given by
 
J f | p = 3p21 − 3p2 p3 , 3p22 − 3p1 p3 , 3p23 − 3p1 p2

So we have that rk(J f )| p = 0 if and only if:

p21 = p2 p3 & p22 = p1 p3 & p23 = p1 p2 .
So it follows that if rk(J f )| p = 0, then f (p) = 0. We conclude that q = 1 is a regular
value of f and therefore M is 2-dimensional smooth embedded submanifold of R3 .
We now compute Tp M at the point p = (0, 0, 1). We use 41, so Tp M = ker J f | p . We
have:  
J f |p = 0 0 3 ,
so it follows that
Tp M = {(x, y, z) ∈ R3 : z = 0}.
So this is just the (x, y)-plane.

43 Let f : R3 → R2 be defined by f (p1 , p2 , p3 ) = (p21 − p22 + p1 p3 − 2p2 p3 , 2p1 − p2 +

p3 ) so that M = f −1 (0, 3). The Jacobian of f in p is given by:
!
2p1 + p3 −2p2 − 2p3 p1 − 2p2
J f |p =
2 −1 1

Note that if p ∈ f −1 (0, 3), then p21 + p1 p3 = p22 + 2p2 p3 and therefore
p1 (p1 + p3 ) = p2 (p2 + 2p3 )
2p1 + p3 = p2 + 3.
 126

It is left to the reader to verify that in this case, the rank of the Jacobian of f at p is
2.

We now compute Tp M at the point p = (1, −1, 0) ∈ M. In this case we have

!
2 2 3
J f |p =
2 −1 1

So, using the fact that Tp M = ker J f | p , we have

Tp M = {(x, y, z) ∈ R3 : 2x + 2y + 3z = 0 & 2x − y + z = 0}

This is the line spanned by the vector (5, 4, −6).

45 We recall that
∂ ∂
∂x = (1, 0) ∂r = (cos θ, sin θ)

∂ ∂
∂y = (0, 1) ∂θ = (−r sin θ, r cos θ).

So if r 6= 0, then
∂ 1
∂x = cos θ · ∂r∂ − r
∂
sin θ · ∂θ

∂ 1
∂y = sin θ · ∂r∂ + r
∂
cos θ · ∂θ .
Also recall that x = r cos θ and y = r sin θ. we obtain the following: (i )
∂ ∂ ∂ 1 ∂
x +y = r cos θ(cos θ · − sin θ · )
∂x ∂y ∂r r ∂θ
∂ 1 ∂
+r sin θ(sin θ · + cos θ · )
∂θ r ∂θ
∂ ∂
= r cos2 θ · + r sin2 θ ·
∂r ∂r
∂
= r
∂r
(ii )
∂ ∂ ∂ 1 ∂
−y +x = −r sin θ(cos θ · − sin θ · )
∂x ∂y ∂r r ∂θ
∂ 1 ∂
+r cos θ(sin θ · + cos θ · )
∂r r ∂θ
∂ ∂
= sin2 θ · + cos2 θ
∂θ ∂θ
∂
=
∂θ
 127

(iii )
∂ ∂ 1 ∂
(x2 + y2 ) = (r2 cos2 θ + r2 sin2 θ)(cos θ · − sin θ · )
∂x ∂r r ∂θ
∂ ∂
= r2 cos θ · − r sin θ ·
∂r ∂θ

48 We let α = u2 + v2 and β = (α + 1)2 . We first compute d( f ◦ g). Note that f ◦ g :

R2 → R is given by
4u2 + 4v2 + (u2 + v2 − 1)2
f ◦ g(u, v) =
β
4α + α2 − 2α + 1
=
β
α2 + 2α + 1 (α + 1)2
= = = 1.
β β
So we have that d( f ◦ g) = 0. Next we compute g∗ d f . Note that d f = 2xdx +
2ydy + 2zdz. We have the following;
2u  2u  2v  2v  α−1 α−1
g∗ d f = 2 d +2 d +2 d
α+1 α+1 α+1 α+1 α+1 α+1
We first make the following computations:
 2u  2(α + 1) − 4u2 −4uv
d = du + dv
α+1 β β
2(v2 − u2 + 1) −4uv
= du + dv
β β
 2v  −4uv 2(α2 + 1) − 4v
d = du + dv
α+1 β β
−4uv 2(u2 − v2 + 1)
= du + dv
β β
 α2 − 1  2u(α2 + 1) − 2u(α2 − 1) 2v(α2 + 1) − 2v(α2 − 1)
d = du + dv
α+1 β β
4u 4v
= du + dv
β β
Combining these results with the above expression for g∗ d f , one can verify that
g∗ d f = 0; i.e. we have that d( f ◦ g) = g∗ d f .

49 (i )
p1 x
f (p) = f˜(x, y) = .
p1 + p22
2 x 2 + y2
 128

∂f ∂f
df = dx + dy
∂x ∂y
2
y −x 2 −2xy
= 2 2 2
dx + 2 dy
(x + y ) (x + y2 )2
(ii )
p1 r cos θ cosθ
f (p) = f˜(r, θ) = =
p1 + p22
2 r2 r

∂ f˜ ∂ f˜
df = dr + dθ
∂r ∂θ
− cos θ − sin θ
= 2
dr + dθ.
r r

(iv)
f (p) = |p|2 f˜(x1 , . . . , xn ) = x12 + . . . + xn2

df = 2xi dxi

50 Recall that x = r cos θ and y = r sin θ and note that

dx = cos θdr − r sin θdθ

dy = sin θdr + r cos θdθ

(i ) We obtain:

ω = r2 cos2 θ(cos θdr − r sin θdθ) + (r cos θ + r sin θ)(sin θdr + r cos θdθ)
= (r2 cos3 θ + r cos θ sin θ + r sin2 θ)dr + (r2 cos2 θ + r2 sin θ cos θ − r3 cos2 θ sin θ)dθ

(ii ) We obtain:

ω = r cos θ(cos θdr − r sin θdθ) + r sin θ(sin θdr + r cos θdθ)
= rdr.

52 We use the fact that ( f ◦ γ)∗ = γ∗ f ∗ , see Lemma 6.3. We have the following se-
quence of equalities:
Z Z Z Z
f ∗ω = γ∗ f ∗ ω = ( f ◦ γ)∗ ω = ω.
γ [a,b] [a,b] f ◦γ
 129

53 (i )
−4t 2t 2t
Z Z Z
∗
α= γ α = dt + dt + dt
γ [0,1] [0,1] (t 2 + 1)2 t2 + 1 t2 + 1
−4t 4t
Z Z
= 2 2
dt + 2
dt
[0,1] (t + 1) [0,1] t + 1
2 1 1
= 2 + 2 ln(t 2 + 1) = 2 ln 2 − 1.
t +1 0 0
and:
−4t 2 2t 2
Z Z Z
ω= γ∗ ω = 2 2
dt + 2 dt + 2 dt
γ [0,1] [0,1] (t + 1) t +1 t +1
2(t 2 + 1) − 4t 2 2t
Z Z
= 2 2
dt + 2
dt
[0,1] (t + 1) [0,1] t + 1
2t 1 1
= 2 + ln(t 2 + 1) = 1 + ln 2.
t +1 0 0

(ii ) We can split γ into three pieces; γ1 , γ2 and γ3 all ranging from [0, 1] to R3 where
γ1 (t) = (t, 0, 0), γ2 (t) = (1,t, 0) and γ3 (t) = (1, 1,t). We now compute the integrals
as follows:
Z Z Z Z
α = α+ α+ α
γ γ1 γ2 γ3
4·0 2t 2
Z Z Z
= 2 2
dt + 2
dt + dt
[0,1] (t + 1) [0,1] t + 1 [0,1] 1 + 1
2t 1 1
Z Z
2
= 2
+ 1dt = ln(t + 1) + t = ln 2 + 1.
[0,1] t + 1 [0,1] 0 0

and:
2t
Z Z Z Z Z Z
ω = ω+ ω+ ω= 2
dt + 1dt = ln 2 + 1.
γ γ1 γ2 γ3 [0,1] t + 1 [0,1]

(iii ) and (iv) : Recall that a 1-form is exact if it is of the form dg for some smooth
function g. The Fundamental Theorem for Line Integrals states that the integral of
an exact 1-form over a path γ only depends on the end-points of γ. So if either α
or ω is exact, then the answers in (i) and (ii) do not differ. So we can conclude
immediately that α is not exact. So what about ω? Well, we guess that it is exact,
but we have to find a function g : R3 → R such that ω = dg. This function g is
given by
2z
g(x, y, z) = 2 + ln(y2 + 1).
x +1

54 Fix p ∈ M and a homeomorphism ϕ : U → V where U,V ⊂ Rl and

ϕ(U ∩ M) = V ∩ (Rm × {0}).
 130

Suppose (x1 , . . . , xl ) is a co-ordinate representation for ϕ. Note that Jϕ−1 | p (Rl ) is

just Rl . Now define a map ϕ̃ : U × Rl → V × Rl as follows;
 ∂ 
ϕ̃ q, vi i = x1 (q), . . . , xl (q), v1 , . . . , vl ,


∂x q
where ∂
|
∂xi q
= Jϕ−1 | p (ei ). Now note that
 ∂ 
ϕ̃−1 (x, v1 , . . . , vl ) = ϕ−1 (x), vi i .
∂x ϕ−1 (x)

So ϕ̃ is a homeomorphism from U × Rl onto V × Rl . Since ϕ is a slice chart for M,

we also have that
h  i h i
ϕ̃(π−1 [U]) = V ∩ Rm × {0} × Rm × {0} ⊂ Rl × Rl
where π : T M → M is the natural projection map. So ϕ̃ maps a neighbourhood of
(p, v) ∈ T M onto a 2m-slice in R2l . This shows how to make co-ordinate charts
for T M. It remains to verify that if ψ̃ is another such chart, then the transition map
ψ̃ ◦ ϕ̃−1 is smooth. For this, see Lee pg81.

58 Let {θ1 , . . . , θn } be the standard basis for (Rn )∗ . So we have that for x ∈ Rn , θi (x) =
xi . Then the inner product on Rn is given by
θ1 ⊗ θ1 + . . . + θn ⊗ θn .

59 We show how to do the determinant on R2 . Again, {θ1 , θ2 } is the standard basis

for (R2 )∗ . The determinant is a covariant 2-tensor which is given by
θ1 ⊗ θ2 − θ2 ⊗ θ1 .

61 It is explained and proved in the lecture notes that V ∗ ⊗ W ∗ is isomorphic to the

space of all bilinear maps from V × W to R. Using the fact that (V ∗ )∗ ∼ = V , it
follows that V ⊗W is isomorphic to the space of all bilinear maps from V ×W ∗ to
∗

R and similarly for W ⊗V . Now it is cleat that V ∗ ×W ∗ ∼= W ∗ ×V ∗ and this gives

way of finding an isomorphism from V ⊗W onto W ⊗V .

63 In light of 61, it suffices to show that V ⊗ R ∼= V . Let {v1 , . . . , vn } be a basis for

V , and e1 = 1 be the standard basis for R. The space V ⊗ R consists of all linear
combinations
V ⊗ R = λi vi ⊗ e1 : λi ∈ R .


This gives a way of defining a map from V ⊗ R to V as follows; f : V ⊗ R → V is

defined by
f (λi vi ⊗ e1 ) = λi vi .
 131

One can verify easily that this defines an isomorphism.

64 We recall the definition of Sym T ,

1
Sym T = ∑ T σ,
r! σ∈Sr

where T σ (v1 , . . . , vn ) = T (vσ(1) , . . . , vσ(r) ). Note that if λ, µ ∈ R and S, T ∈ T r (V ),

then

(λS + µT )σ = λSσ + µT σ ,

and that if σ, π ∈ Sr , then (T σ )π = T σπ . From these obervations it follows that

(Sym T )σ = Sym T σ .

Next, if π ∈ Sr , note that {σ : σ ∈ Sr } = {πσ : σ ∈ Sr } and this gives the following:

1 1
Sym T π = ∑ (T π )σ = r! ∑ T σπ
r! σ∈Sr σ∈Sr
1 1
= ∑ T σπ = r! ∑ T σ = Sym T.
r! σπ∈Sr σ∈Sr

symmetry of Sym T : Using the equalities we have proved, we get that

(Sym T )π = Sym T π = Sym T.

Using π = (i, j), it follows that Sym T is symmetric.

T is symmetric if and only if Sym T = T : If Sym T = T , then T is symmetric

because Sym T is. On the other hand, if T is symmetric, then T π = T for all
π ∈ Sr , so that
1 1
Sym T = ∑ Tσ = ∑ T = T.
r! σ∈Sr r! σ∈Sr

This completes the exercise.

67 We recall the definition of Alt T ,

1
Alt T = ∑ (−1)σ T σ ,
r! σ∈Sr
 132

−1
where (−1)σ is 1 if σ is even and −1 if σ is odd. Recall that (−1)σ = (−1)σ and
that (−1)σπ = (−1)σ (−1)π . As in the symmetric case, note that,
1
(Alt T )π = ∑ (−1)σ T π σ
r! σ∈Sr

1
= ∑ (−1)π (−1)π (−1)σ T π σ
r! σ∈Sr

= (−1)π Alt T.
1
Alt T π = ∑ (−1)σ T σ π = . . . = (−1)π Alt T.
r! σ∈Sr

So we have that (Alt T )π = Alt T π .

Alt T is alternating: Use π = (i, j) and (Alt T )π = (−1)π Alt T to show that Alt T
is alternating, since (i, j) is an odd permutation.
T = Alt T if and only if T is alternating: If Alt T = T , then T is alternating
because Alt T is. Conversely, if T is alternating, show by induction that T π =
(−1)π T and thus (−1)π T π = T . It follows that
1 1
Alt T = ∑ (−1)π T π = r! ∑ T = T.
r! π∈Sr π∈Sr

This completes the exercise.

75 The set Σr (Tp M) consists of all symmetric r-linear forms on Tp M and Σr M is the
union of these forms over all p ∈ M, so
Σr M = {(p, σ) : p ∈ M and σ ∈ Σr (Tp M)}.

76 Let σ ∈ Σ2 M. Then for every p ∈ M, σ p is a symmetric bilinear form on the vector

space Tp M. In other words, σ p defines an (indefinite) inner product on Tp M. So we
may view σ as a smooth choice of inner products on the vector spaces Tp M for all
p ∈ M.

78 Recall that the matrix for the pull-back is just the transpose of the Jacobian matrix
for f in p, so we have
!∗ !
∗ ∗ 2p1 −1 2p1 1
f = (J f | p ) = =
1 6p22 −1 6p22

79 We compute the pullback f ∗ σ. For vectors ξ, ϕ ∈ Tq (M), we have

σq (ξ, ϕ) = ξ1 ϕ2 + q1 ξ2 ϕ2 .
 133

So for q = f (p), and ξ, ϕ ∈ Tp (N), we have

f ∗ σ p (ξ, ϕ) = σq ( f∗ ξ, f∗ ϕ)
! !!
2p1 ξ1 − ξ2 2p1 ϕ1 − ϕ2
= σq ,
ξ1 + 6p22 ξ2 ϕ1 + 6p22 ϕ2
= (2p1 ξ1 − ξ2 )(ϕ1 + 6p22 ϕ2 ) + (p21 − p2 )(ξ1 + 6p22 ξ2 )(ϕ1 + 6p22 ϕ2 )
= (p21 + 2p1 − p2 )ξ1 ϕ1 + (6p22 p21 − 6p32 + 12p1 p22 )ξ1 ϕ2 +
(6p22 p21 − 6p32 − 1)ξ2 ϕ1 + (36p42 p21 − 36p52 − 6p22 )ξ2 ϕ2

so we have the following:

f ∗ σ = (p21 + 2p1 − p2 )dx1 ⊗ dx1 + (6p22 p21 − 6p32 + 12p1 p22 )dx1 ⊗ dx2 +
(6p22 p21 − 6p32 − 1)dx2 ⊗ dx1 + (36p42 p21 − 36p52 − 6p22 )dx2 ⊗ dx2

∂ ∂ ∂ ∂
For vector fields X = X1 ∂x + X2 ∂y and Y = Y1 ∂x +Y2 ∂y , we get that

f ∗ σ(X,Y ) = (p21 + 2p1 − p2 )X1Y1 + (6p22 p21 − 6p32 + 12p1 p22 )X1Y2 + etc.

We can also compute f ∗ alternatively as follows;

f ∗ σ = d(p21 − p2 ) ⊗ d(p1 + 2p32 ) + (p21 − p2 )d(p1 + 2p32 ) ⊗ d(p1 + 2p32 )

= (2p1 dx1 − dx2 ) ⊗ (dx1 + 6p22 dx2 ) + (p21 − p2 )(dx1 + 6p22 dx2 ) ⊗ (dx1 + 6p22 dx2 )
= (p21 + 2p1 − p2 )dx1 ⊗ dx1 + (6p22 p21 − 6p32 + 12p1 p22 )dx1 ⊗ dx2 +
(6p22 p21 − 6p32 − 1)dx2 ⊗ dx1 + (36p42 p21 − 36p52 − 6p22 )dx2 ⊗ dx2

The fact that both σ and f ∗ σ are smooth follows from the fact that the component
functions of these 2-tensor fields are smooth.

80 We first recall that

σ = dx1 ∧ dx2 − dx2 ∧ dx3

= dx1 ⊗ dx2 − dx2 ⊗ dx1 − dx2 ⊗ dx3 + dx3 ⊗ dx2 .

We now compute iX σ for the given vector fields:

(i ) iX σ = σ(X, ·) = dx2 − dx1 − dx3 + dx2 = −(dx1 + dx3 ), (ii ) iX σ = σ(X, ·) =

x1 dx2 −x22 dx2 = (x1 −x22 )dx2 , (iii ) iX σ = σ(X, ·) = x1 x2 dx2 +sin(x3 )dx1 +sin(x3 )dx3 .
 134

81

f ∗ σ = sin2 (p1 + p2 ) d(sin(p1 + p2 )) ⊗ d(p21 − p2 ) − d(p21 − p2 ) ⊗ d(sin(p1 + p2 ))

h
= sin2 (p1 + p2 ) cos(p1 + p2 )dx1 + cos(p1 + p2 )dx2 ⊗ 2p1 dx1 − dx2

i
− 2p1 dx1 − dx2 ⊗ cos(p1 + p2 )dx1 + cos(p1 + p2 )dx2


h
= sin2 (p1 + p2 ) 2p1 cos(p1 + p2 )dx1 ∧ dx1 + 2p1 cos(p1 + p2 )dx2 ∧ dx1
i
− cos(p1 + p2 )dx1 ∧ dx2 − cos(p1 + p2 )dx2 ∧ dx2
h i
= sin2 (p1 + p2 ) cos(p1 + p2 ) 2p1 dx2 ∧ dx1 − dx1 ∧ dx2
= − sin2 (p1 + p2 ) cos(p1 + p2 )(1 − 2p1 )dx1 ∧ dx2 .

82 (i ) Consider the component functions of iX σ and note that these are all smooth.

(ii ) let X1 , . . . , Xr−2 ∈ F(M), then

(iX ◦ iX σ)(X1 , . . . , Xr−2 ) = σ(X, X, X1 , . . . , Xr−2 )

= −σ(X, X, X1 , . . . , Xr−2 ) = −(iX ◦ iX σ)(X1 , . . . , Xr−2 ).

The second equality follows from the fact that σ ∈ Γr (M) and hence it is alternat-
ing. We have just proved that for all X1 , . . . , Xr−2 ∈ F(M), we have

(iX ◦ iX σ)(X1 , . . . , Xr−2 ) = 0.

But this means that all component functions of iX ◦ iX σ are identical 0, and hence
iX ◦ iX σ = 0. Since σ was arbitrary, it follows that iX ◦ iX = 0.

(iii ) We first derive an intermediate result, where X = X1 :

1
∑ (−1)a σ(Xa(1) , . . . , Xa(r) )ω(Xa(r+1) , . . . , Xa(r+s) )
r!s!
a ∈ Sr+s
a−1 (1) ∈ {1, . . . , r}
1
= ∑ (−1)a(1,a(1)) σ(1,a(1)) (Xa(1) , . . . , Xa(r) )ω(Xa(r+1) , . . . , Xa(r+s) )
r!s!
a ∈ Sr+s
a−1 (1) ∈ {1, . . . , r}
r
= ∑ (−1)a σ(X1 , Xa(2) . . . , Xa(r) )ω(Xa(r+1) , . . . , Xa(r+s) )
r!s! a∈S{2,...,r+s}
= (iX σ) ∧ ω(X2 , . . . , Xr+2 ).
 135

Similarly we have that

(iX ω) ∧ σ(Xr+1 , . . . , Xr+s , X2 , . . . , Xr )

1
= ∑ (−1)a σ(Xa(1) , . . . , Xa(r) )ω(Xa(r+1) , . . . , Xa(r+s) )
r!s!
a ∈ Sr+s
a−1 (1) ∈ {r + 1, . . . , r + s}

Combining these results we get that:

iX (σ ∧ ω)(X2 , . . . , Xr+s ) = (σ ∧ ω)(X1 , . . . , Xr+s )

1
= ∑ (−1)a σ(Xa(1) , . . . , Xa(r) )ω(Xa(r+1) , . . . , Xa(r+s) )
r!s! a∈Sr+s

= (iX σ) ∧ ω(X2 , . . . , Xr+2 ) + (iX ω) ∧ σ(Xr+1 , . . . , Xr+s , X2 , . . . , Xr )

 
= (iX σ) ∧ ω + (−1)(r−1)s (−1)(r+s−1)(r−1) σ ∧ (iX ω) (X2 , . . . , Xr+s )
 
= (iX σ) ∧ ω + (−1)(r−1) σ ∧ (iX ω) (X2 , . . . , Xr+s )

83 First recall that

bi ∧ . . . σr = (−1)−1 σ1 ∧ . . . ∧ σi (X) ∧ . . . σr .
σi (X) ∧ σ1 ∧ . . . ∧ σ

Now the proof is by induction on r using (iii) of Lemma 10.2. If r = 2, then this
is just (iii) of Lemma 10.2. Now assume that it is true for r and consider the case
r + 1. Using (iii) of Lemma 10.2 we have

iX σ = iX (σ1 ∧ . . . ∧ σr ∧ σr+1 )
= iX (σ1 ∧ . . . ∧ σr ) ∧ σr+1 + (−1)r σ1 ∧ . . . ∧ σr ∧ (iX σr+1 )
= (−1)i−1 σi (X) ∧ σ1 ∧ . . . ∧ σ
bi ∧ . . . σr+1 .

84 We assume that n ≥ 2, the case for n = 1 is done in the lecture notes. Using
stereographic projections, it is possible to find two charts (U, ϕ) and (V, ψ) for Sn
which together form an atlas for Sn and such that U ∩ V is connected. Since the
determinant is a continuous function, it follows that the function f : U ∩ V → R
defined by
f (p) = det J(ψ ◦ ϕ−1 ) p .
 136

is also continuous. Note that f (p) 6= 0 for all p and since U ∩ V is connected it
follows that either f (p) > 0 or f (p) < 0 for all p ∈ U ∩V . In the first case we are
done. If f (p) < 0 for all p ∈ U ∩V , then let ψ0 be defined by

ψ0 (x1 , x2 , . . . , xn ) = ψ(−x1 , x2 , . . . , xn ).

It is now easily verified that the Jacobian of the transition map ψ0 ◦ ϕ−1 has positive
determinant everywhere, hence Sn is orientable.

86 We only do the case for the Klein bottle. Note that if a manifold M is an open
subset of an orientable manifold N, then M is orientable. So non-orientability of
the Klein bottle follows from the fact that the Möbius strip is a non-orientable open
subspace of the Klein bottle.

87 We shall give an atlas for which the Jacobian of the transition maps has positive
determinant everywhere. This gives an induced orientation. So recall that PR3 is
a quotient space of R4 \ {0} with projection map π. We let Ui = {x ∈ R4 : xi 6= 0}
and Vi = π(Ui ). The homeomorphisms ϕi : Vi → R3 are given by (THESE ARE
NOT CORRECT: FIX THIS)
 
x2 −x3 −x4
ϕ−1
 
ϕ1 ([x]) = ,
x1 x1 , x1 1 (z) = (−1, −z1 , z2 , z3 )

 
x1 −x3 −x4
ϕ−1
 
ϕ2 ([x]) = x2 , x2 , x2 2 (z) = (−z1 , −1, z2 , z3 )

 
x1 −x2 −x4
ϕ−1
 
ϕ3 ([x]) = x3 , x3 , x3 3 (z) = (−z1 , z2 , −1, z3 )

 
x1 −x2 −x3
ϕ−1
 
ϕ4 ([x]) = x4 , x4 , x4 4 (z) = (−z1 , z2 , z3 , −1)

We compute the Jocabian of the transition map ϕ1 ◦ ϕ−1

2 , first of all note that
1 z z 
2 3
ϕ1 ◦ ϕ−1
2 = , , .
z1 z1 z1
So that
 1 
z21
0 0
 
 
 z2 1
Jϕ1 ◦ ϕ−1

2 z  − z21
= z1 0 

 
 
− zz32 0 1
z1
1

This matrix has positive determinant. The computations for the other transition
maps are naturally left to the reader.
 137

90 Since the open rectangles form a basis for the topology on Rm , we may fix an open
covering U of K such that U ⊂ U and every member of U is a rectangle in Rm .
S

Since K is compact, we may pick a finite collection V ⊂ U such that V also covers
K. Now let D = V. Then D is the finite union of rectangles and therefore its
S

boundary has measure 0.

93 Let U be such a covering. For every x ∈ M, we need to find an open neighbourhood

Ux of x such that the set
{U ∈ U : U ∩Ux 6= 0},
/

is finite. But since U is a covering of M, for an arbitrary x ∈ M, we may pick an

element V ∈ U with x ∈ V . Then V intersects only finitely many members of U,
hence V witnesses the fact that U is locally finite (in x).

95 For completeness, note that F : D → S2 where D = [0, π]×[0, 2π]. We first compute
F ∗ ω. Note that

F ∗ dx = cos ϕ cos θdϕ − sin ϕ sin θdθ,

F ∗ dy = cos ϕ sin θdϕ + sin ϕ cos θdθ,
F ∗ dz = − sin ϕdϕ.

We now compute F ∗ ω as follows:

F ∗ ω = sin ϕ cos θ(cos ϕ sin θdϕ + sin ϕ cos θdθ) ∧ (− sin ϕdϕ)
+ sin ϕ sin θ(− sin ϕdϕ) ∧ (cos ϕ cos θdϕ − sin ϕ sin θdθ)
+ cos ϕ(cos ϕ cos θdϕ − sin ϕ sin θdθ) ∧ (cos ϕ sin θdϕ + sin ϕ cos θdθ)
= − sin3 ϕ cos2 θdθ ∧ dϕ + sin3 ϕ sin2 θdϕ ∧ dθ
− cos2 ϕ sin ϕ sin2 θdθ ∧ dϕ + cos2 ϕ sin ϕ cos2 θdϕ ∧ dθ
= sin ϕdϕ ∧ dθ.

So we now have the following:

Z Z Z
∗
ω = F ω= sin ϕdϕ ∧ dθ
S2 D D
Z 2π Z π Z 2π π
= sin ϕdϕdθ = − cos ϕ dθ
0 0 0 0
Z 2π
= 2dθ = 4π.
0
 138

96 We first compute F ∗ ω.
F ∗ ω = sin ϕ cos θ(cos ϕ sin θdϕ + sin ϕ cos θdθ) ∧ (− sin ϕdϕ)
cos ϕ(cos ϕ sin θdϕ + sin ϕ cos θdθ) ∧ (cos ϕ cos θdϕ − sin ϕ sin θdθ)
= − sin3 ϕ cos3 dθ ∧ dϕ − cos2 ϕ sin2 θ sin ϕdϕ ∧ dθ
cos2 ϕ sin ϕ cos2 θdθ ∧ dϕ
= (sin3 ϕ cos2 θ − sin ϕ cos2 ϕ)dϕ ∧ dθ.

and thus:
Z Z Z 2π Z π
ω = F ∗ω = sin3 ϕ cos2 θ − sin ϕ cos2 ϕdϕdθ
S2 D 0 0
Z 2π  π π
2 1 3 1
= − cos θ cos ϕ − cos ϕ + cos3 ϕ dθ
0 3 0 3 0
Z 2π
4 2
= cos2 θ − dθ
0 3 3
θ 2ϕ 4
 
4 1
= sin 2θ + −
3 4 2 0 3π
= 0.
We can avoid this calculation if we use Stokes’ Theorem: first note that
dω = dx ∧ dy ∧ dz + dz ∧ dy ∧ dx = 0,
so we obtain: Z Z Z
ω= dω = 0 = 0.
S2 B3 B3

97 First note that

F ∗ dx = − sin θdθ,
F ∗ dy = cos θdθ.
So that:
F ∗ ω = cos2 θdθ + sin2 θdθ = dθ.
Computing the integral is easy:
Z Z Z
ω= F ∗ω = dθ = 2π.
S1 [0,2π] [0,2π]

99 The torus is parametrized by the mapping F : D → R4 given by

F(θ, ϕ) = (cos θ, sin θ, cos ϕ, sin ϕ)T ,
 139

where D = [0, 2π] × [0, 2π]. We have the following:

F ∗ dx1 = − sin θdθ F ∗ dx3 = − sin ϕdϕ
F ∗ dx2 = cos θdθ F ∗ dx4 = cos ϕdϕ.
So we have:
F ∗ ω = cos2 θ(− sin θdθ ∧ cos ϕdϕ) + sin θ(− sin ϕdϕ ∧ − sin θdθ)
= − sin θ cos2 θ cos ϕdθ ∧ dϕ + sin ϕ sin2 θdϕ ∧ dθ
= (sin θ cos ϕ cos2 θ + sin ϕ sin2 θ)dϕ ∧ dθ.
Z Z
ω = F ∗ω
T2 D
Z 2π Z 2π
= (sin θ cos ϕ cos2 θ + sin ϕ sin2 θ)dϕdθ
0 0
Z 2π 2π 2π
= sin θ cos2 θ sin ϕ − cos ϕ sin2 θ dθ
0 ϕ=0 ϕ=0
Z 2π
= 0 = 0.
0

100 We have
g∗ dx1 = dr g∗ dx3 = dt
g∗ dx2 = ds g∗ dx4 = 2(2r − t)(2dr − dt).
Let ω be the form to be integrated. We have
g∗ ω = sds ∧ dt ∧ (2(2r − t)(2dr − dt)) + 2rtdr ∧ ds ∧ dt
= 4s(2r − t)ds ∧ dt ∧ dr + 2rtdr ∧ ds ∧ dt
= (8rs − 4st + 2rt)dr ∧ ds ∧ dt.
We now compute the integral:
Z Z
ω = g∗ ω
M D
Z 1Z 1Z 1
= (8rs − 4st + 2rt) drdsdr
0 0 0
Z 1Z 1
= (4s − 4st + t) dsdt
0 0
Z 1
1
= (2 − t) dt = 1 .
0 2

101 (i ) dω = yzexyz dx + xzexyz dy + xyexyz dx, (ii ) dω = 2xdx + z cos ydy + sin ydx,
(iii ) dω = 0 and ω = dσ, where σ = x2 +y2 , (iv) dω = dx ∧dy −dx ∧dz, (v) dω =
 140

xdy ∧ dx ∧ dz + dz ∧ dx ∧ dy = (1 − x)dx ∧ dy ∧ dz, (vi ) dω = 0 and ω = dσ where

σ = dx ∧ dz.

102 All of these forms are closed.

103 (i ) dω = dx ∧ dy − dy ∧ dx = 2dx ∧ dy, so ω is not even closed. (ii ) ω = dσ where

σ = xy. (iii ) ω = dσ where σ = xdy. (iv) ω = dσ where σ = (x3 /3 + y3 x)dy.

104 All of these forms are exact: (i ) σ = 12 (x2 + y2 + z2 ), (ii ) σ = 13 x3 dy + xz3 dz,
(iii ) σ = 31 x3 ydy ∧ dz.

109 Note that d( f ω) = d f ∧ ω + f dω, so by Stokes’ Theorem we have:

Z Z Z Z
fω = d( f ω) = d f ∧ω+ f dω.
∂M M M M

Rearranging these equalities gives the desired formula.

113 First note that

f ∗ dx = cos s costdt − sin s sintds,

f ∗ dz = tds + sdt.

So we obtain:

f ∗ ω = (cos s costdt − sin s sintds) ∧ (tds + sdt)

= −(s sin s sint + t cos s cost)ds ∧ dt.

114
f ∗ ω = s2 cost sint [(costds − s sintdt) ∧ (sintds + s costdt) ∧ du]
= s3 cost sint(sin2 t + cos2 t)ds ∧ dt ∧ du
= s3 cost sintds ∧ dt ∧ du.

115 First note that

(x2 + y2 + z2 )3/2 − 3x2 (x2 + y2 + z2 )1/2
 
∂ x
=
∂x (x2 + y2 + z2 )3/2 (x2 + y2 + z2 )3
y2 + z2 − 2x2
= .
(x2 + y2 + z2 )5/2
 141

So now we compute dω as follows:

∂ ∂ ∂
dω = . . . dx ∧ dy ∧ dz + . . . dy ∧ dz ∧ dx + . . . dz ∧ dx ∧ dy
∂x ∂y ∂z
1 h
2 2 2


= y + z − 2x dx ∧ dy ∧ dz+
(x2 + y2 + z2 )5/2
i
+ x2 + z2 − 2y2 dx ∧ dy ∧ dz + y2 + x2 − 2z2 dx ∧ dy ∧ dz

= 0.

This shows that ω is closed. Now, to see that ω is not exact, note that by 107 we
have Z
ω = 3 · Vol(B3 ) > 0.
S2
Now it follows from Stokes’ Theorem that if ω is exact, this integral is zero, so it
follows that ω is not exact.

116 Denote the parametrization mapping with f . The pullback of the first form is given
by

f ∗ ω = ds ∧ (dt + du) ∧ (ds − du)

= −ds ∧ dt ∧ du.

So we obtain
Z Z
dx1 ∧ dx2 ∧ dx4 = − ds ∧ dt ∧ du = −1.
M [0,1]3

117 For the direct computation, let f : [0, π/2] → R2 be given by f (t) = (t, 2/πt), i.e.,
f is a (reversed) parametrization of the line BO. Let

ω = (y − sin x)dx + cos ydy,

and note that 2  2 2 

f ∗ω = t − sint dt + cos t dt
π π π
So we have:
Z Z Z Z
ω = − sin xdx + cos ydy − f ∗ω
C OA AB OB
Z 1
2 2 2 
Z π/2 Z π/2
= −sinxdx + cos ydy − t − sint + cos t dt
0 0 0 π π π
 2 
π/2 1 t π/2 π/2 2 π/2
= cos x + sin x − + cost + sin t
0 0 π 0 0 π 0
π π
= −1 + + 1 = .
4 4
142

Using Green’s Theorem, note that dω = dy ∧ dx = −dx ∧ dy, so we obtain:

Z Z Z
π
ω= dω = − dxdy = − .
C OAB OAB 4