S.P. Novikov - MATH 740: Spring 2014 S. Novikov - o Ce 3112 (MATH) E-Mails: Novikov@ipst - Umd.edu
S.P. Novikov - MATH 740: Spring 2014 S. Novikov - o Ce 3112 (MATH) E-Mails: Novikov@ipst - Umd.edu
Contents
Lecture 1. Introduction: Textbooks and General Remarks.
Local coordinates: What is Cartesian system of coordinates.
Examples. 5
Homework 1. 21
1
Homework 2. 34
Homework 3. 49
Homework 4. 63
Homework 5. 75
2
Homeworks 2, 3, 4. Solutions. 76
Homework 2. Solutions. . . . . . . . . . . . . . . . . . . . . . 76
Homework 3. Solutions. . . . . . . . . . . . . . . . . . . . . . 78
Homework 4. Solutions. . . . . . . . . . . . . . . . . . . . . . 80
Homework 5. Solutions. 84
Homework 6. 94
Homework 7. 106
Homework 8. 117
3
Homework 6 and Homework 7. Solutions. 118
Homework 6. Solutions. . . . . . . . . . . . . . . . . . . . . . 118
Homework 7. Solutions. . . . . . . . . . . . . . . . . . . . . . 119
Homework 9. 131
4
Lecture 1. Introduction: Textbooks and Gen-
eral Remarks. Local coordinates: What is
Cartesian system of coordinates. Examples.
Textbooks:
Manfredo P. do Carmo. Riemannian Geometry.
Victor Guillemin, Alan Pollack. Differential Topology.
Additional Literature:
J. Milnor. Morse Theory.
S.P. Novikov, I.A. Taimanov. Modern Geometric Structures And Fields.
B.A. Dubrovin, A.T. Fomenko, S.P. Novikov. Modern Geometry - Methods
and Applications: Parts I, II.
f : X →R
5
b) Collection of functions f1 , . . . , fn
fj : X → R
f⃗(x) = f⃗(y) → x = y
Examples.
1) X = R2 , f1 = x, f√ 2 = y.
2) X = R , f1 = ρ = x2 + y 2 , f2 = φ.
2
ρ
a) ρ is not a coordinate in R2 ϕ
but ρ is a coordinate in R2 \0 = X ′
b) φ is not a coordinate in R2 \0 because φ is not a function, it is
multivalued.
“Cartesian Coordinates” in Rn
Rn ↔ (x1 , . . . , xn )
U → Rn
( )
x → x1 (x), . . . , xn (x) = ⃗x(x)
xj : U → R are continuous functions (one-valued!)
⃗x(x) = ⃗x(y) ↔ x = y
6
Lecture 2. Manifolds and Atlases.
Manifolds: = Hausdorff (or metric) spaces such that they are “locally
euclidean”: for every x ∈ M there exists an open set x ∈ U with homeo-
morphism
φU : U → Rn = (x1 , . . . , xn ) (so U ⊂ Rn )
⃗x(x) = ⃗x(y) ↔ x = y
xi (y), i = 1, . . . , n, y k (x), k = 1, . . . , n
∂xi ∂y k
= δji
∂y k ∂xj
Definition: Oriented Atlas is such that all det |∂xi /∂y k | > 0
Oriented manifold: = there exists an oriented Atlas.
Examples:
7
x
n = 2:
S2 is oriented manifold.
RP2 is NOT.
(x, −x) is one point in RP2 .
−x S2
Definitions.
1) C ∞ - function in C ∞ - manifold M with given Atlas of Charts:
M → R is C ∞ in every Chart.
2) C ∞ - map
F
M − → N
(Uα ) (Vβ )
is C ∞ in every Chart. In other words, for every pair Uα , Vβ the corresponding
functions yβk (x1α , . . . , xnα ) are C ∞ in F −1 (Vβ ) ∩ Uα .
Rank of the map F : M → N at the point x ∈ M :
( )
∂yβk Uα −
F
→ Vβ
rkx F = rk where
∂xiα (x) (y)
x
Statement. Rank of the map at the point x does not depend on the choice
of Atlas.
Proof. Let Uα , Vβ be Euclidean domains giving the charts of the manifolds
M and N and the functions ⃗y (⃗x) be represented by the map
F
Uα −→ Vβ
(x) (y)
We have
∂ ỹ i ∂ ỹ i ∂y k ∂xs
= ,
∂ x̃j ∂y k ∂xs ∂ x̃j
where rk |∂ ỹ i /∂y k | ̸= 0, rk |∂xs /∂ x̃j | ̸= 0.
8
Conclusion
∂ ỹ i ∂y i
rk = rk
∂ x̃j ∂xj
Statement is proved.
F
Special Case: M −
→ R (function).
rkx f = 1 − regular point (∇f |x ̸= 0)
rkx f = 0 − critical point (∇f |x = 0)
Example: f = x +y : (x, y) ̸= (0, 0) - regular point, (0, 0) - critical point.
2 2
M = ∪α Uα = ∪α Vα
where Uα ⊂ Vα and there are common coordinates ⃗xα for every Uα , Vα such
that the corresponding domains Uα , Vα in the Euclidean space are defined
by the relations:
V
∑ ∑
Uα : (xjα )2 < 1 , Vα : (xjα )2 < 2 . U
j j
9
φ ≡ 0 for r ≤ 0, 1 ϕ(r)
φ ≡ 1 for r ≤ 1,
φ ≡ 0 for r ≥ 2,
φ′ < 0 for 1 < r < 2.
0 1 r 2
Consider C ∞ - functions on the manifold M with GDA:
∑
x̃jα = φ(rα ) xjα , rα2 = (xjα )2
j
M n = U1 ∪ . . . ∪ UQ = V1 ∪ . . . ∪ VQ
Consider the coordinates ⃗xα (rα < 1 in Uα and rα < 2 in Vα ) and the
corresponding functions φ(rα ) and x̃iα on M n .
Then the map F : M n → RN , N = Q(n + 1):
[ ]
F (x) = x̃11 , . . . , x̃n1 , φ(r1 ), . . . , x̃1Q , . . . , x̃nQ , φ(rQ )
is a C ∞ imbedding (nondegenerate).
Terminology:
Imbedding: = rkx F = n at every point x and F (x) = F (y) → x = y.
Immersion: = rkx F = n = dim M at every point x ∈ M .
rkxF=0
M 1 R2
imbedding immersion
Proof of the Theorem.
I. Let x ∈ Uα , then rk F = n because
[ ] ∂ x̃iα
x̃iα = xiα for x ∈ Uα , F = . . . , x̃1α , . . . , x̃nα , . . . , and rk = n
∂xjα
(unit matrix in Uα ).
10
II. Suppose x ̸= y. Can we still have F (x) = F (y) ?
a) Let x, y ∈ Uα then F (x) ̸= F (y) by the same reason as in I.
b) Let x ∈ Uα , y ∈ Vα (rα > 1). Then φ(rα )|x = 1, φ(rα )|y < 1, so
F (x) ̸= F (y), [ ]
F = . . . , x̃1α , . . . , x̃nα , φ(rα ), . . .
c) Let x ∈ Uα , and y is outside Vα . Then φ(rα )|x = 1, φ(rα )|y = 0, so
F (x) ̸= F (y).
Theorem is proved.
We can easily see the that all the functions ψα (x) represent C ∞ - functions
on M with the following properties
∑
ψα (x) ≥ 0 , ψα (x) ≡ 1
α
Example of Application.
What is an Integral on a Manifold?
∫
I = f (x) dn σ
M
11
We can see now that I is represented as a sum of ordinary integrals
over local domains where we can also put in local coordinates: dn σ =
gα (⃗xα ) dx1α . . . dxnα .
Tangent vector on a C ∞ - manifold M = ∪α (Uα , ⃗xα ):
a) Vector τ is attached to a point x ∈ M .
b) Vector τ is characterized by “components” τ
(τ 1 , . . . , τ n ) in the given system of local coordinates. (x1 ,...,x n)
x
c) Basic vectors ei in the coordinate system ⃗x are identified with the
operators ∂/∂xi :
∂
ei ↔ , τ = τ i ei
∂xi
d) The vector τ is identified with the differential operator τ i ∂/∂xi :
∂
τ ↔ τi ,
∂xi
so τ acts on the functions f (x) at the point x by the formula:
∂f
τ (f ) = τ i
∂xi x
12
By definition, we can write
i
∂f ′j ∂x ∂f
τi = τ ,
∂xi ∂y j ∂xi
so we come to the conclusion
∂xi
τ i = τ ′j
∂y j
(summation over j is assumed).
In the same way, for the inverse transformation ⃗y = ⃗y (⃗x) we can write
∂y j
τ ′j = τ i
∂xi
where
∂xi ∂y j
= δki
∂y j ∂xk
Every C ∞ - map is given locally by its linear part and the smaller terms
∑
n
i i 1 n
y = F (x , . . . , x ) = Const + Aij xj + O(||x||2 )
j=1
(near ⃗x = 0).
Inversion of Map (C ∞ ).
Let us have a map
F : Rn → Rn , x0 → y0
(x) (y)
G : Rn → Rn , y0 → x0 ,
(y) (x)
13
defined in the set U , such that for x ∈ V , y ∈ U we have the relations
G (F (x)) ≡ x , F (G(y)) ≡ y
Implicit Functions.
Let us have a coordinate system (y 1 , . . . , y n+k ) in Rn+k and a system of
k equations z 1 = 0, . . . , z k = 0, z i = z i (y 1 , . . . , y n+k ) with the Condition:
( i)
∂z
rky0 = k
∂y j
(maximal rank).
Statement.
Let us assume that under the above conditions we have the relation
( i)
∂z
det T ≡ det ̸= 0 , j = n + 1, . . . , n + k ,
∂y j y0
Jacobi Matrix
R2 z=c
∂z i ∂z i ∂z i ∂z i y1
z=0
, . . . , , , . . . ,
∂y 1 ∂y n ∂y n+1 ∂y n+k
| {z } n=k=1
nonzero determinant is here y2
y0
Then:
1) There exists an open set U ∋ y0 ∈ Rn+k near the point y0 , where
the values (y 1 , . . . , y n , z 1 , . . . , z k ) represent a coordinate system;
2) The change
( 1 ) F ( 1 )
y , . . . , y n+k −→ y , . . . , yn, z1, . . . , zk
is C ∞ and nondegenerate.
14
Proof.
Easy to see that the Jacobian Matrix of the transformation F can be
written in the form:
1 0 ... 0 ∂z 1 /∂y 1 ... ∂z k /∂y 1
0 1 ... 0 ∂z 1 /∂y 2 ... ∂z k /∂y 2
. .
. . ... .. .. ... ..
. . . . .
n
J = 0 0 ... 1 ∂z 1 /∂y n . . . ∂z /∂y
k
0 0 ... 0 ∂z 1 /∂y n+1 . . . ∂z k /∂y n+1
. . .. .. ..
.. .. ..
. . .
..
. .
1 n+k
0 0 . . . 0 ∂z /∂y ... ∂z k /∂y n+k
We immediately get then det Jy0 = det T ̸= 0. So, we get our statement
from the Inversion Theorem.
Statement is proved.
Under the above conditions, the “Implicit
y z=c Function Theorem” states that on the
0 z=0 submanifold, given by the relations z 1 = 0,
z=c’
. . . , z k = 0, the values (y n+1 , . . . , y n+k ) can
be locally expressed as explicit functions of
the coordinates (y 1 , . . . , y n ):
y n+1 = φ1 (y 1 , . . . , y n ) , . . . , y n+k = φk (y 1 , . . . , y n )
y n+1 = φ1 (y 1 , . . . , y n , z 1 , . . . , z k ) , . . . , y n+k = φk (y 1 , . . . , y n , z 1 , . . . , z k )
15
Every imbedding of manifold
M n → M n+k
(x) (y)
∂y i locally can be given by k nondegenerate
rk ∂xq x0
=n equations z 1 = 0, . . . , z k = 0.
Let ( )
∂y i
det ̸= 0 , i = 1, . . . , n
∂xq x0
According to the Inversion Theorem, the transformation
( 1 ) ( )
x , . . . , xn , y n+1 , . . . , y n+k → y 1 , . . . , y n , y n+1 , . . . , y n+k
We have also ( )
∂z p
det ̸= 0 ,
∂y n+q
so the values (x1 , . . . , xn , z 1 , . . . , z k ) give also a local coordinate system in
Rn+k .
Corollary is proved.
16
Lecture 4. Manifolds and Submanifolds: Im-
plicit functions and Inversion. Vectors and
Covectors.
According to the previous lecture, we can formulate here the following
Statement.
Let us have an imbedding
z1 = 0 , . . . , zk = 0
Conclusion.
Every C ∞ - manifold can be given by a set of
local equations in RN : M n → RN (proved
for compact M n ).
Remark. Not all manifolds can be given by global nondegenerate set of
equations in RN
φ1 (y 1 , . . . , y N ) = 0 , . . . , φN −n (y 1 , . . . , y N ) = 0
(Proof later).
RP2 can not be given by nondegenerate global set of equations.
“Tangent Manifold” T ∗ (M n ).
Let us denote again by τ a tangent vector (τ 1 , . . . , τ n ) in M n attached
to a point x ∈ M n .
‘Tangent Manifold” T ∗ (M n ):
( )
M n = ∪α Uα , x1α , . . . , xnα (Atlas, C ∞ )
17
( )
T ∗ (M n ) = ∪α Uα × Rn , x1α , . . . , xnα , τ 1 , . . . , τ n
Change of coordinates:
Let us put in the intersection of Charts Uα and Uβ : x = xα , y = xβ .
We have then
xi = xi (y) = xi (y 1 , . . . , y n ) (in Uα ∩ Uβ )
The components of the same vector in the coordinates (x) and (y):
(x, τ ) ↔ (y, τ ′ )
(x, η) ↔ (y, η ′ )
∂xi ∂y j
ηj′ = ηi , ηi = ηj′
∂y j ∂xi
( )
T∗ (M n ) = ∪α Uα × Rn , x1α , . . . , xnα , η1 , . . . , ηn
Scalar product (invariant):
⟨τ , η⟩ = τ i ηi
∂
Vector field : fi (Vector fields)
∂xi
Covector field : gi dxi (Differential 1 − forms)
18
Vector Fields = Dynamical Systems (Autonomous)
Integration of 1-forms along the path: let
ω = gi (x) dxi
and the path is given by a piecewise
x(t)
smooth one-parametric curve
( ) B
⃗x(t) = x1 (t), . . . , xn (t) A
Definition.
∫ ∫ B ∫ t1 ( )
i dxi
ω = gi (x) dx = gi (x(t)) dt
x(t) A t0 dt
Properties.
Integral does not depend on the choice of coordinates in M n and “time”
t.
t’
Inner Product of tangent vectors
gij (x) τ i τ̃ j = ⟨τ , τ̃ ⟩
∂xi ∂xj ′
gij (x(y)) = gks (y)
∂y k ∂y s
Let Tensor field gij (x) be symmetric:
19
Types of inner products (“Types of Geometry”):
gij dxi dxj > 0 - Riemannian Metric.
Special Types (p, q) of Pseudoriemannian metric:
p = 0 : - Riemannian
p = 1 : - Lorentzian
p = q : - Ultrahyperbolic
Theorem.
In every C ∞ - manifold there exists a C ∞ Riemannian Metric.
Proof. Take GDA (Good Double Atlas)
M n = U1 ∪ . . . ∪ UQ = V1 ∪ . . . ∪ VQ
(|⃗xα | < 1 in Uα and |⃗xα | < 2 in Vα ) and the function φ(r) introduced in
Lecture 2.
1 ϕ(r)
0 1 r 2
Consider quadratic forms
∑
n
( )2
Gα = dx̃jα , x̃jα = xjα · φ(rα )
j=1
20
Homework 1.
1. Local coordinates: ρ and φ in R2 .
a) Prove that ρ can not be chosen as a Cartesian coordinate in the
whole domain R2 \0 (whole).
b) Prove that φ can be defined as a one-valued function in any domain
not containing closed paths surrounding 0 .
111111111111111111111111111111
000000000000000000000000000000
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
U 000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
0
000000000000000000000000000000
111111111111111111111111111111
U 000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
000000000000000000000000000000
111111111111111111111111111111
r2 = x2 + y 2 + z 2
Let r = 1 .
N
a) In which domains θ is a good Cartesian
coordinate? r=1
b) In which domains (φ, θ) are good local
coordinates?
S
3. Construct C ∞ functions
1 1
ϕ(r) ψ(r)
0
0 1 r 2 −1 1
using {
0 , x ≤ 0
f (x) =
e−1/|x| , x ≥ 0
2
21
Lecture 5. Manifolds and vectors fields. Im-
portant Examples.
Classes of manifolds:
a) All C ∞ - manifolds M n ⊂ RN for given N .
b) Manifolds defined by the Global Nondegenerate systems of Equations
in RN
f1 = 0 , . . . , fN −n = 0
(Important cases: n = 2, N = n + 1 = 3).
Classification of manifolds:
n=1: S1 , R
↑ ↑
compact noncompact
S2 , T2 , . . . , S2g . 2
Sg =
nonorientable:
b
y x
2 Klien
RP2 , K2 (Klein Bottle), other. RP = K 2= Bottle
a a
−x
−y
(−x,x) RP 2 b
Noncompact 2 - manifolds.
Important manifolds:
RPn - real projective spaces
CPn - complex projective spaces
22
QPn - “quaternionic” projective spaces
Groups: GLn (R), GLn (C), SLn (R), SLn (C), On ⊃ SOn , Op,q ⊃
SOp,q (case p = 1 - Lorentz Groups), Un ⊃ SUn , Up,q ⊃ SUp,q
Stiefel Manifolds
Grassmann Manifolds
Isometry groups of Rn
Isometry groups of Rp,q
where
Euclidean Metric: gij = δij
Lorentzian Metric: gij = diag (1, −1, . . . , −1) , i.e.
1 0 ... 0
0 −1 . . . 0
gij = .. .. . . .
. . . ..
0 0 . . . −1
τ = (τ 1 , . . . , τ n ).
Covectors on Manifold M n : η = (η1 , . . . , ηn ).
T ∗ (M n ) - space of all vectors on M n .
T∗ (M n ) - space of all covectors on M n .
Vector field
τ (x) = (τ 1 (x), . . . , τ n (x))
locally
generates dynamical system:
ẋi = τ i (x) , i = 1, . . . , n
Solution = curve ⃗x(t): ẋi = τ i (x)
23
x(t) → y(t) = (t, y02 , . . . , y0n )
solution const
τ ′ (x0 ) ̸= 0 .
Apply Theorem of Cauchy.
t
One-parametric group generated by vector field.
The equation
ẋ = τ (x)
generates a (local) one parametric group of invertible transformations of (lo-
cal domains) in M n :
St : M n → M n x(t)
x(0) = x0 , x(t) = St (x0 )
x0
Corollary
( )
∂η j ∂ i ∂τ
j
∂ i ∂η
j
i ∂τ
j
∂
[τ , η] = τ i
i j
− η = τ − η
∂x ∂x ∂x ∂xj
i ∂x i ∂x i ∂xj
24
Remark. Sometimes we denote operator τ (f ) acting on scalar functions
by ∇τ f because it coincides in this case with covariant derivative of the scalar
field f along vector field τ .
Statement: Let τ (x0 ) ̸= 0 , η(x0 ) ̸= 0 . Then [τ , η] ≡ 0 ⇒ there exists
local coordinate system (y 1 , . . . , y n ) such that
τ = (1, 0, . . . , 0) , η = (0, 1, 0, . . . , 0)
St′′ ◦ St = St ◦ St′′
Examples.
1. τ = const in Rn : group of shifts x → x + t · const .
2. τ i = aij xj (linear)
25
dim GLn (R) = n2
dim GLn (C) = 2n2
dim GLn (Q) = 4n2
Question: prove that GLn (R) has 2 components. GLn (C) is connected.
2) SLn (R) , SLn (C) : det = 1
Equation : det A = 1 in Rn or in Cn
2 2
dA
= B ∈ Lie Algebra
dt t=0
Another form
dA −1 dA
Lie Algebra → A or A−1
dt dt
Groups SOn , On :
d
⟨A(t)ζ , A(t)η⟩ = ⟨ζ , η⟩ , ⟨A(t)ζ , A(t)η⟩ = 0 ,
dt
26
A(t) = A(0) + Ȧ(0) t + O(t2 ) , A(0) = 1
↑
B
27
evident for the diagonalizable matrices B = S −1 ◦ diag (b1 , . . . , bn ) ◦ S
from the series representation of the matrix eB . Since the set of diagonaliz-
able matrices is dense in the matrix space and the functions exp (Tr B) and
det eB are analytic functions of the matrix entries we actually have
det eB = eTr B
for any matrix B . The statement of the Lemma for small enough B follows
then from the properties of the function ex .
Lemma is proved.
We can see then that in the local coordinate system in GLn (R) , given
by the entries of the matrix B near I ∈ GLn (R) , the equation det A = 1
has the form Tr B = 0 .
Conclusion. This equation is linear and NONDEGENERATE.
Group SOn : Equations for SOn in the coordinate system, given by the
entries of B near I ∈ GLn (R) , have the form B t = −B .
Unitary group A Āt = I.
Un ⊂ GLn (C) , we can put again A = eB near I = GLn (C) and
introduce a local coordinate system, given by the entries of B in Cn =
2
Q : {q = a + ib + jc + kd}
28
i2 = j2 = k2 = −1 , ij = −ji = k , jk = −kj = i , ki = −ik = j
q̄ ≡ a − ib − jc − kd
GL2 (Q) = {q : q q̄ = 1} = {a2 + b2 + c2 + d2 = 1} = S3 = SU2
Pauli matrices:
( ) ( ) ( )
0 1 0 i i 0
i σx = , i σy = , i σz =
−1 0 i 0 0 −i
Correspondence to the basic quaternions (i, j, k):
i σx ↔ i , i σy ↔ j , i σz ↔ k
SO4 = S3 × S3 /Z2
Consider the transformations
q → q1 q q 2 , q1 q̄1 = 1 , q2 q̄2 = 1
Two transformations coincide iff : (q1′ , q2′ ) = ±(q1 , q2 ) .
29
Compact groups: On , SOn , Un , SUn , Tn = Rn /Zn , Spn =
GLn (Q) ∩ SO4n , SO2 = U1 = S1 , SO3 = SU2 /Z2 , SU2 = Sp1 = S3 .
Local coordinates (near I ):
1. SOn : B t = −B , dim SOn = n(n − 1)/2
2. Un : B t = −B̄ , dim Un = n2
Noncompact groups: Rn , Iso (Rn ) −→ Rn (shifts), GLn (R) ,
On
GLn (C) , SLn (R) , SLn (C) , Op,q , Up,q , Sympln (?) .
1 ... 0 0 ... 0
.. . . . .. .. . . . ..
. . . .
0 ... 1 0 ... 0
Rp,q : (e1 , . . . , en ) , ⟨ei , ej ⟩ =
0 . . . 0 −1 . . . 0
. . . .. .. . . .
.. . . . . ..
0 . . . 0 0 . . . −1
R2n : (e′1 , e′′1 , . . . , e′n , e′′n ) , ⟨e′i , e′j ⟩ = 0 , ⟨e′′i , e′′j ⟩ = 0 , ⟨e′i , e′′j ⟩ = δij
0 1 0 0 ... 0 0
−1 0 0 0 ... 0 0
0 0 0 1 ... 0 0
0 −1 0 . . . 0 0
gij = 0
.. .. .. .. . . . ..
. . . . . .. .
0 0 0 0 ... 0 1
0 0 0 0 . . . −1 0
Lie Algebras:
30
1. SOn : B t = −B , (real entries) .
2. Un : B = −B̄ ,
t
(complex entries) .
Matrix B is equal to
31
with definition ( )
g ik = (gki )−1
General definition
Let Riemannian (pseudoriemannian) manifold M n be given with Atlas
(α)
of Charts Uα , (x1α , . . . , xnα ) and “metric” gij (x) in the domain Uα . For 2
tangent vectors attached to x we have
⟨τ , η⟩ = gij (x) τ i η j
⟨ξ , κ⟩ = g ij (x) ξi κj
dh ∂h
= η i (x) i = ⟨∇g h , ∇g f ⟩
dt ∂x
32
Proof.
dh ∂h ∂h ∂f
= η i (x) i = g ij (x) i j
= ⟨dh , df ⟩ = ⟨∇g h , ∇g f ⟩
dt ∂x ∂x ∂x
Lemma is proved.
Corollary 1.
df
g ij = − g ji ⇒ ≡ 0
dt
(Symplectic case).
Corollary 2. For Riemannian metric gij η i η j > 0 we have df /dt > 0
along the gradient system
∂f
ẋi = g ij j
∂x
ij −1
(because (g ) = (gij ) ) .
Important example of Symplectic Manifold is T∗ (M n )
dim T∗ (M n ) = 2n , Atlas : {Uα , (x1α , . . . , xnα , pα1 , . . . , pαn )}
Fix α - number of Chart.
∂ ∂
Tangent basis : e′i = , e ′′
i =
∂xi ∂pi
Inner Product is
⟨e′i , e′j ⟩ = 0 , ⟨e′′i , e′′j ⟩ = 0 , ⟨e′i , e′′j ⟩ = − ⟨e′′j , e′i ⟩ = δij
τ = (e′i ) - tangent vector to M n , p = (e′′j ) - tangent covector to M n ,
⟨τ , p⟩ = τ i pi - natural Invariant Inner Product.
Language of 2-forms
∑ ( )
0 I
Ω = dx ∧ dpi ,
i
gij =
−I 0
((x1 , . . . , xn , p1 , . . . , pn ) - local coordinates in T∗ (M n ) ) .
“Gradient systems” in the Symplectic Manifolds are called “Hamiltonian
systems”.
33
Homework 2.
1. How many projection local coordinate systems for Sn ∈ Rn+1 are needed
to cover all sphere? n = 1, 2, 3, . . . .
2. How many local coordinate systems are needed to cover RPn ? Let
n = 1, 2, 3 .
( )
x ∈ RPn = Rn+1 \0 /x ∼ λx , λ ̸= 0
34
1. Riemannian Case: ||η||2g > 0 , ⟨ei , ej ⟩ = δij .
2. Pseudoriemannian Case: type p, q , ⟨ei , ej ⟩ = ±δij .
3. Symplectic Case: M 2k ,
⟨η , ζ⟩ = gij η i ζ̄ j
(Riemannian Case)
√ √
dn σ = (−1)q det gij , dx1 ⊗ · · · ⊗ dxn = ±g dn x
(Pseudoriemannian Case)
Transformation Rule (“Measure”) :
∂xi
x = x(y) , dn x = dx1 . . . dxn = det dy 1 . . . dy n = |J| dn y
∂y j
35
Important Remark. For manifolds given by Oriented Atlas (J > 0)
we can write dn x = J dn y (differential forms !), J = det ||∂xi /∂y j || .
Lemma 1. Riemannian Metric in the manifold M n defines Riemannian
Metric in every submanifold W k ⊂ M n .
Proof.
Let W k ⊂ M n , locally we have for local coordinates y in M n : y i =
y i (x1 , . . . , xk ) , i = 1, . . . , n, where x represent some local coordinates in
W k.
We define “restriction of metric”
∂y s ∂y k
gij′ (x) = gsk (y(x))
∂xi ∂xj
(restriction of inner product on every linear subspace of tangent space). It
remains positive.
Lemma is proved.
Remark. The analogous lemma is wrong for Pseudoriemannian or Symplec-
tic Geometry because after the restriction to linear subspace metric might
become degenerate.
x0
<η,η>=0
η x2
R1,2 , light cone: ⟨η , η⟩ = 0 .
x1
36
x(t)
B ∫ b√
l(γ) = gij (x(t)) ẋi ẋj dt
γ a
A
“Distance” = minγ l(γ)
a b
Statement. Riemannian metric transforms M n into metric space
S2
b
a
a
“Minimal Time Principle”: Light propagates from the point A to
the point B along the path with minimal time among all piecewise
smooth paths joining these 2 points
37
∫ ∫ b√
b
|dl| dl2
Time = =
a c(x(t)) a c2 (x)
y
√
dl = dx2 + dy 2
x
{
c1 , y > 0
Let c(x) =
c2 , y < 0
a
I(x)= c1
b
So we have a ”Fermat Riemannian Metric”
δij 1 1
gijF = , ||η||2F = ||η||2E , ||η||F = ||η||E
c2 (x) c2 c2
(F = “Fermat”, E = Euclid) .
Speed of light in vacuum c = cvac ∼ 3 · 1010 cm/sec , cmedia < cvacuum
.
How to find geodesics?
Euler - Lagrange (XVIII Century)
Consider more general problem. Let L(x, η) (“Lagrangian”) be a
smooth function in T ∗ (M n ) (tangent manifold). Fix points a, b ∈ M n .
Find “extreme curves” for the action
∫ 1
S(γ) = L (x(t), ẋ(t)) dt
0
38
(η = ẋ) on the piecewise smooth paths: x(t) , x(0) = a , x(1) = b .
b
a x(t)
Examples:
Geometry: action and length functionals
1 1
a) L = gij (x) η i η j = ||η||2
2 2
√
b) L′ = ||η|| = gij η i η j − length
Physics: action functional
1
L = ||η||2 − U (x)
2
(gravity, electric fields)
1 e
L = ||η||2 − e U (x) + Ai (x(t)) ẋi
2 c
- electric field Ei = − ∂U/∂x , magnetic field Bij = ∂Ai /∂xj − ∂Aj /∂xi .
i
b ∫ b√
l(γ) = gij (x(t)) ẋi ẋj dt
a γ a
More general:
39
“Lagrangian” L(x, η) : T ∗ (M n ) → R is given.
“Action functional” is given
∫
S{γ} = L (x(t), ẋ(t)) dt
γ
Examples:
1) L = gij (x) ẋi ẋj /2 = ||ẋ||2 /2
√
2) L′ = gij ẋi ẋj = ||ẋ||
3) L′′ = ||ẋ||2 /2 − U (x) (gravity or electricity)
4) L′′′ = ||ẋ||2 /2 + (e/c) Ai (x) ẋi - magnetic field (its vector-potential).
5) “Relativistic Particle” (?)
Geodesics: Either (1) or (2)
P. Fermat: gij (x) = δij /c2 (x)
“Variation” of path γ
γ + ϵ η(x(t)) = γϵ b
a γ
(locally it makes sense)
η = vector field along (γ) (tangent to M n ).
Variation of Action:
∫ b
S {γ + ϵ η} = L (x(t) + ϵ η(t), ẋ + ϵ η̇) dt
a
40
Lemma (Euler - Lagrange).
Curve γ is extremal iff
( )
d ∂L ∂L
= ,
dt ∂ ẋi ∂xi
∂L/∂xi = fi = “force”
ṗi = fi
41
1
L = gij (x) ẋi ẋj ⇒ pi = gij ẋj
2
T ∗ (M n ) → T∗ (M n )
ẋ → p
velocity momentum
vector (covector)
Equation of Geodesics:
( )
∂L 1 ∂gij
ṗk = k
= k
ẋi ẋj
∂x 2 ∂x
√
Another form (for length) L′ = gij ẋi ẋj :
( )
d pk ∂L′ 1 ∂gij i j
√ = k
= √ ẋ ẋ
dt i
gij (x) ẋ ẋj ∂x 2 gij (x) ẋ ẋ ∂xk
i j
′
Conclusion. Let parameter
√ t for the length functional L is “natural”, i.e.
t = l(γ) (length), dt = gij (x(t)) ẋi ẋj . Then we have same equations
√
for both L and L′ because gij (x(t)) ẋi ẋj = 1 .
Corollaries.
1) Geodesics for Rn , gij = δij , are the b
straight lines. a
x0 =ct timelike
lightlike
2) Relativistic Particles: (x0 , x1 , x2 , x3 ),
gij = diag (1, −1, −1, −1) .
Let x0 = ct, spacelike
η 2 < 0 - “spacelike” vector,
x=x1
η 2 = 0 - “lightlike” vector,
η 2 > 0 - “timelike” vector.
42
x(t)
b
Requirement:
For every real material object we have:
⟨ẋ , ẋ⟩ ≥ 0.
If the mass of the object > 0 then we have:
⟨ẋ , ẋ⟩ > 0 .
“Momentum”:
∂L m ẋi
pi = = √ , i = 1, 2, 3
∂ ẋi 1 − w2
Energy :
∂L
E = ẋi − L
∂ ẋi
Examples :
1) Geodesics
1
L = gij (x) ẋi ẋj ⇒ E = L
2
√
L′ = gij (x) ẋi ẋj ⇒ E ≡ 0
43
2) Gravity or electric field
1 1
L′′ = gij (x) ẋi ẋj − U (x) ⇒ E = gij (x) ẋi ẋj + U (x)
2 2
3) Magnetic field
1 e 1
L′′′ = gij (x) ẋi ẋj + Ai (x) ẋi ⇒ E = gij (x) ẋi ẋj
2 c 2
Equations :
1) Geodesics
( )
1 ∂gij
ṗk = ẋi ẋj , pk = gkj ẋj
2 ∂xk
2) Gravity or electric field
( )
1 ∂gij ∂U
ṗk − ẋi j
ẋ = − , pk = gkj ẋj
2 ∂xk ∂xk
3) Magnetic field
( )
1 ∂gij e ∂Ai i e
ṗk − k
ẋi ẋj = ẋ , pk = gkj ẋj + Ak (x)
2 ∂x c ∂xk c
i.e. ( ) ( )
d ( ) 1 ∂gij e ∂Ai ∂Ak
gkj ẋj − i
ẋ ẋ j
= − ẋi
dt 2 ∂xk c ∂x k ∂xi
Magnetic field
∂Ai ∂Ak
Bik = k
−
∂x ∂xi
44
“Lagrangian”: L(x, η) : T ∗ (M n ) → R .
∫
S{γ} = L (x(t), ẋ(t)) dt (Action)
γ
∫ b( )
dS ∂L d ∂L
δ S (γ, η) → (γ + ϵ η) = i
− η i dt
dϵ ϵ=0 a ∂x dt ∂ ẋi
a η(t) b
Euler - Lagrange equation:
( )
d ∂L ∂L
δS = 0 ↔ =
dt ∂ ẋi ∂xi
Examples
1) L = gij (x) ẋi ẋj /2
√
2) L = gij ẋi ẋj
3) Physics
↑ ↑
gravity, magnetic
electric field
field
“Energy”
∂L ∂L
E = ẋi i
− L = vi i − L
∂ ẋ ∂v
45
Energy Conservation Law
Theorem 10.1. For the Euler - Lagrange System we have the following
conservation law:
dE
= 0
dt
(i.e. E is constant along the trajectories of the Euler - Lagrange System).
Proof.
( ) ( )
dE d i ∂L i ∂L i d ∂L ∂L i ∂L i
= ẋ − L = ẍ + ẋ − ẋ − ẍ =
dt dt ∂ ẋi ∂ ẋi dt ∂ ẋi ∂xi ∂ ẋi
[ ( ) ]
d ∂L ∂L
= − ẋi = 0
dt ∂ ẋi ∂xi
Theorem is proved.
L = L (x2 , . . . , xn , v 1 , . . . , v n )
46
Examples.
1) L = gij (x) ẋi ẋj /2
Energy
∂L
E = ẋi − L = L
∂ ẋi
Ė = 0 ⇒ - parameter along geodesics is NATURAL because
1
E = ||ẋ||2g = const
2
along trajectory.
“Angular Momentum”
∂L
pφ = = const − along geodesics
∂ φ̇
Geometrical meaning - later.
Consider now
1 [ 0 2 ] 1
L′ = a1 · (ẋ ) − (ẋ1 )2 − (ẋ2 )2 − (ẋ3 )2 = a1 · ⟨ẋ , ẋ⟩
2 2
47
√ √
L′′ = a2 · (ẋ0 )2 − (ẋ1 )2 − (ẋ2 )2 − (ẋ3 )2 = a2 · ⟨ẋ , ẋ⟩ , t = x0 /c
Relativity
∂L′
Use L′ : E ′ = L′ , pi
(4)
=
∂ ẋi
time τ ∼ length by theorem above so
( 0 )
(4) ∂L′ dx dt dx1 dt dx3 dt
pi = = a1 · ,− , ..., −
∂ ẋi dt dτ dt dτ dt dτ
dt dτ 1 1
= 1/ = √ ∑3 = √
dτ dt
c2 − (v α )2 c2 − v 2
α=1
where ( )
dx1 dx2 dx3
v = , ,
dt dt dt
4-D Formalism
So we have for the 4-momentum
( )
1 dx1 /dt dx2 /dt dx3 /dt
p (4)
= a1 · √ ,− √ ,− √ ,− √
1 − w2 c 1 − w2 c 1 − w2 c 1 − w2
where w = v/c .
Clearly we have to choose: a1 = mc , where m is the mass of a
particle.
3-D Formalism
Use L′′ . We use time t = x0 /c , x0 = ct .
√
L′′ = a2 · (c2 − (ẋ1 )2 − (ẋ2 )2 − (ẋ3 )2
∂L′′ vα a2 vα
p(3)
α = = − a 2 · √ = − √
∂ ẋα c2 − v 2 c 1 − v 2 /c2
v α = dxα /dt , α = 1, . . . , 3 . We choose a2 = − mc .
So we finally have
∂L′′ mc2
E ′′ = ẋα − L′′
= √ , w = v/c
∂ ẋα 1 − w2
48
E ′′ = “physical energy” E . Let p0 = E/c . We have
vi
v = (v 1 , v 2 , v 3 ) , √ = pi
1 − v 2 /c2
111111111111111111111
000000000000000000000
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
We have “Poincare” Model for 3D Hyperbolic Space 000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
(ball) 000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
Homework 3.
1. Prove that group O(1, 1) consists of transformations:
( )
2 2 ch φ sh φ
P , P = 1 , T , T = 1 ,
sh φ ch φ
( )
1 0
U (1, 1) = Z2 × Z2 × R (topologically) , gij =
0 −1
O(2) ∼
= Z2 × S1 (topologically)
49
3. Prove that SL2 (R) ∼
= R2 × S1 (topologically).
4. Calculate Euclidean Riemannian metric of R2 in polar coordinates.
5. Calculate Riemannian Metric of sphere S2 :
S2 : x2 + y 2 + z 2 = 1 , R3 : dl2 = dx2 + dy 2 + dz 2
x0
H2 in R1,2
x1
50
1
L = gij (x) ẋi ẋj , L(x, η) : T ∗ (M n ) → R
2
Euler - Lagrange Equations:
∂L
ṗi = , pi = gij v j
∂v i
∂gij k j 1 ∂gjk k j
gij (x) ẍj + k
ẋ ẋ = ẋ ẋ
∂x 2 ∂xi
( )
1 ∂gjk ∂gij
ẍj gij = i
− k
ẋk ẋj
2 ∂x ∂x
( )
1 ∂gjk ∂gij
p
ẍ = g pi
i
− k
ẋk ẋj
2 ∂x ∂x
Let us introduce the “Christoffel Symbols”
( )
1 pi ∂gij ∂gik ∂gjk
p p
Γkj = Γjk = g + −
2 ∂xk ∂xj ∂xi
We have then
ẍp + Γpkj ẋk ẋj = 0
dz dz̄
2. S2 : dθ2 + sin2 θ dφ2 ⇒
(1 + |z|2 )2
dz dz̄ dz dz̄
3. H2 : dχ2 + sh2 χ dφ2 ⇒ ⇒
(1 − |z| )
2 2 y2
Lemma 1. For the general metric of the form dl2 = dρ2 + Φ2 (ρ) (dφ)2
every line φ = const is geodesics.
Proof. We have
1 ( 2 )
L = ρ̇ + Φ2 (ρ) φ̇2
2
51
Euler - Lagrange equations:
d ( 2 )
ρ̈ = ΦΦ′ (ρ) φ̇2 , Φ (ρ) φ̇ = 0
dt
So we have: φ̇ = 0 always satisfy to the Euler - Lagrange equations.
Lemma is proved.
Remark. Integrate Euler - Lagrange equations now:
∂V (ρ) 1
so : ρ̈ = − a2 , V (ρ) = ,
∂ρ 2Φ2 (ρ)
∂V (ρ) ρ̇2 a2
so : ρ̈ ρ̇ + a2 ρ̇ or + = b = const
∂ρ 2 Φ2 (ρ)
We have then √
ρ̇ = 2(a2 /Φ2 (ρ) − b)
and
∫
dρ
t = √ ⇒ ρ = ρ(t) , φ̇ = a/Φ2 (ρ)
2(a2 /Φ2 (ρ) − b)
52
Isometry
group
Klein Model:
SL 2(R)
y = Im z > 0 , dl2 = dz dz̄/y 2 .
y=0
How to prove that they represent the same metric?
Geodesic Flow = Dynamical system in T ∗ (M n ) ∼
= T∗ (M n ) .
T ∗ (M n ) : ẋp = v p , v̇ p = Γpij xi xj
∂L
T ∗ (M n ) : ṗi = , ẋi = g ik (x) pk , (pi = gij (x) ẋj )
∂xi
Theorem. Let
∂L
E(x, p) = H(x, p) , E = ẋi − L
∂ ẋi
Then we have
∂H ∂H
ṗi = − , ẋi =
∂xi ∂pi
53
2-form
∑
n
Ω = dxi ∧ dpi
i=1
v
H=const
Example. n = 1 , H = p2 /2 + V (x) ,
L = ẋ2 /2 − V (x) , p = v . x
dH
≡ 0 , (energy conservation)
dt
For geodesics:
1 ij 1
H = g (x) pi pj , L = gij (x) ẋi ẋj ,
2 2
(g ij ) = (gij )−1 , g ik gkj = δji , gij = gji .
54
n
τ
Special case n = 2 (oriented plane) :
2. Curvature of Hypersurfaces
M n ⊂ Rn (x0 , x1 , . . . , xn )
z
τ
Let M n is given by the equation
x0 = z = F (x1 , . . . , xn )
0
x1 ... xn
55
Coordinate system is chosen in such a way that z ⊥ M n at the point
P : (x1 , . . . , xn ) = (0, . . . , 0) , so we have near P :
∂z ∂F
= = 0 , j = 1, . . . , n
∂xj P ∂xj (0,...,0)
K<0
convex K>0
56
( ) ( )
= 1 + Fx2 dx2 + 1 + Fy2 dy 2 + 2 Fx Fy dx dy
For the point P ∈ M n where z ⊥ M n we have gij (x) = δij .
covector field:
(ηi ) = (fxi )
Vector field: ( i) d ( i)
ζ −
→ ∂i ζ
Transformation rule x = x(y)
∂xi
(x) ζ i = ζ ′j (y)
∂y j
Consider now
( i
)
i ∂ ′j ∂x ∂ ∂y p ∂
∂k ζ = ζ , where =
∂xk ∂y j ∂xk ∂xk ∂y p
We have
( )
∂ i ∂y p ∂ ′j ∂x
i
∂y p ∂xi ∂ζ ′j ∂y p ∂ 2 xi ′j
ζ = ζ = + ζ
∂xk ∂xk ∂y p ∂y j ∂xk ∂y j ∂y p ∂xk ∂y p ∂y j
57
Conclusion: Second derivative enters in the change of coordinates rule
( ) ( )
∂ i ∂ ′j ∂y p ∂xi ∂ 2 xi ∂y p ′j
ζ = ζ + ζ
∂xk ∂y p ∂xk ∂y j ∂y p ∂y j ∂xk
∂xi ∂xq
h′pq = hij , h → f ∗ (h)
∂y p ∂y q
h′ ← h
∗
f :
2 − form in N m 2 − form in M n
58
“Covariant derivatives”
∇i η j = ∂i η j + Γjik η k
∇i ηj = ∂i ηj − Γkij ηk
where Γijk are “Christoffel Symbols”.
Requirements: ∇i η j and ∇i ηj are tensors.
Lemma. The set of linear equations (above) is solvable for all “initial
data” Ψ(x0 ) = Ψ0 if and only if
∇i ∇j − ∇j ∇i = Rij = 0
for all x ∈ Rn .
Proof. For every solution Ψ(x) we have
∂i Ψ = Ai Ψ (Ai = matrix)
and
∂i ∂j Ψ = ∂i (Aj Ψ) = ∂j (Ai Ψ)
or
(∂i Aj − ∂j Ai ) Ψ + Aj (∂i Ψ) − Ai (∂j Ψ) = 0
or
(∂i Aj − ∂j Ai − Ai Aj + Aj Ai ) Ψ = 0
59
where
(∂i Aj − ∂j Ai − Ai Aj + Aj Ai ) = Rij (matrix)
So we have Rij Ψ = 0 for all x ∈ Rn . Therefore Rij ≡ 0 .
Lemma is proved.
Rij = [∇i , ∇j ]
Examples.
a) L = n , Ψ - vector field, Aijk = − Γijk .
b) L = n , Ψ - covector field, Aijk = Γijk .
c) L = 1 , Ai - scalar values, Ai Aj − Aj Ai = 0 ,
∂i Aj − ∂j Ai = Rij
60
Other groups are also possible and define the “structural group” of the
bundle.
Our case: X is a C ∞ -manifold and aij (x) are C ∞ -functions.
61
for all x ∈ Rn .
Rij = [∇i , ∇j ]
Examples.
a) L = n , Ψ - vector field, Aijk = − Γijk .
b) L = n , Ψ - covector field, Aijk = Γijk .
c) L = 1 , Ai - scalar values, Ai Aj − Aj Ai = 0 ,
∂i Aj − ∂j Ai = Rij
∇i = ∂i − Ai , ∇q = ∂q − Aq ,
∂Aq ∂Ai
[∇i , ∇q ] = Riq = − + + Ai Aq − Aq Ai
∂xi ∂xq
(Matrices in RN ) .
1) For tangent (cotangent) case N = n .
2) For scalar case N = 1 , n is any.
3) For euclidean case ⟨ei , ej ⟩ = δij we should have (Ai )kj = − (Ai )jk
(skew symmetry).
Change of Basis:
ei = bij (x) e′j , [e = B e′ ]
or
Ψi ei = Ψi bji e′j , Ψ′j = Ψi bji
⃗ ′ = B(x) Ψ
Ψ ⃗
Lemma 2. Let
∂Ψj
= Ajik Ψk (x)
∂xi
62
Then for the new basis (e′j ) we have
∂Ψ′j
i
= A′jik Ψ′k (x)
∂x
where
∂B −1
A′i = B Ai B −1 + B
∂xi
∂Ψ′j ∂B ∂B −1 ′
= B Ai Ψ + Ψ = B Ai B −1 Ψ′ + B Ψ
∂xi ∂xi ∂xi
Lemma is proved.
Conclusion.
∂B −1
A′i = B Ai B −1 + B
∂xi
Homework 4.
1. Prove that the group O(1, 1) (connected component) can be written in
the form ( ) ( )
√ 1 √ w
ch ψ sh ψ 1−w2 1−w2
= √ w √ 1
sh ψ ch ψ 1−w2 1−w2
63
4. Find cover of sphere S2g with g handles by 2 systems of local coordinates.
2
Sg =
64
Lecture 14. Vector bundles. Connection
and Curvature. Parallel transport.
Vector bundle=“locally trivial” family of vector spaces RN x , x ∈ X with
“local bases” [e1 (x), . . . , eN (x)], x ∈ U ⊂ X, ∀x ∃U ∋ x.
U U
For x ∈ U ∩ V
eUi (x) = aj,U
i
V
(x)eN
j (x).
“Group” aj,U
i
V
(x) ∈ G ⊂ GLN (R)
Differential geometrical connection (on vector bundle).
Locally matrix-valued functions Ai (x) = (Ai )jk (x) are given for all x ∈
U ⊂ X, and operators ∇Ui = ∂i − AUi acting on functions ΨU : U → RN ,
ΨU = (Ψj,U ),
∇Ui ΨU = ∂i ΨU − AUi ΨU .
In intersection x ∈ U ∩ V we have
ΨU = g U V ΨV , G(x) = g U V (x),
such that:
AVi = G−1 (x)AUi G(x) − G−1 ∂i G
(gauge equivalence).
“Curvature”
U
Rij = ∇Ui ∇Uj − ∇Uj ∇Ui (matrix functions).
Theorem (Later):
V
Rij = G−1 Rij
U
G (!)
ξ-tangent vector to X = M n in the point x ∈ M n , x1 , . . . , xn .
( U )
∂Ψ
∇ξ =
U
i
− Ai Ψ ξ i
U
∂x
65
∇Uγ = ∇Uẋ(t) = ẋi ∇Ui ΨU .
xi (t) γ (t)
x0
Definition. “Parallel” vector field along the line γ(t):
∇Uẋ(t) ΨU (x(t)) ≡ 0 for all t.
“Parallel transport” •+ vector ΨU (x0 ) = Ψ0 along the line γ(t), γ(0) =
x0 .
γ Results for γ1 and γ2 may be
different!
x0 γ’
“Variational geodesics.”
ẍj + Γjik ẋi ẋk = 0
we need
1
Γkij = − (Akij + Akji ) − symmetric part
2
for these 2 equations coincide. Calculating variation, we obtain:
[ [ ]] [ ]
1 ∂gij ∂giq 1 lq ∂gij ∂giq ∂gqj
Γij = Symmetrization of g −
l lq
+ = g − q + + .
2 ∂xq ∂xj 2 ∂x ∂xj ∂xi
66
Calculation of Connection and Curvature: (it is linear operation)
∇ξ f (x) = ξ i ∂i f.
such that
so we see that (2) is true for the “products” (ηl ξl ), so it is true for linear
combinations of products and so it is true fro all tensors of the type (tkl ).
Proof of the formula for Γkij – next lecture.
67
Lecture 15. Connections in tangent bundle.
Curvature. Ricci curvature. Einstein equa-
tion. Spaces of constant curvature.
Differential-geometrical connections in tangent bundle T ∗ (M n ), x1 , . . . , xn .
68
Let us solve the system of linear equations for the triple (ijk) using condition
Γkij = Γkji (3 equations for 3 unknown quantities Γkij = Γkji ) or Γij,k = gks Γsij .
Solution:
1
Γij,k = (∂j gik + ∂i gkj − ∂k gij ) .
2
Our equations are:
∂i gkj = Γij,k + Γik,j
u = ∂i gkj , v = ∂k gij , w = ∂j gki ,
u−v+w
a = Γij,k , ⇒ a = ,
2
1
Γij;k = (∂i gkj − ∂k gij + ∂j gki ) .
2
1
Γkij = g ks Γij;s = g ks (∂i gsj + ∂j gis − ∂s gij ) .
2
For the case Γkij = Γkji , gij = gji we obtain the same formula as for geodesics
(for Calculation of Variations)
Curvature.
∂Γj ∂Γi
∇i ∇j − ∇j ∇i = [∂i + Γi , ∂j + Γj ] = − + Γi Γj − Γj Γi .
∂xi ∂xj | {z }
product of matrices
We have finally:
69
∂Γkij
=? Calculation is required.
∂xl
Conclusions from calculations (later):
s
1. Rij;kl = gis Rj;kl = Rkl;ij .
70
Appendix to Lecture 15.
1
Γkij = g ks Γij;s = g ks (gsj,i + ∂j gis,j − gij,s ) .
2
Our notation:
∂f /∂xk ≡ f, k
∂ Γ̂q ∂ Γ̂i
R̂iq = − + Γ̂i Γ̂q − Γ̂q Γ̂i , where Γ̂i = Γkij .
∂xi ∂xq
Our system:
gij (0) = δij , gij,p (0) = 0, g ij (0) = δij = gij (0), g ij ,p (0) = 0
71
• Curvature Tensor at the most symmetric spaces Rn , S n , H n .
Tji11,...,j
,...,ik
l
(x), dim = nk+l , Vector bundle Ψ : M n → Tensors,
(1) (2)
Operations: linear (sum), product of tensors T IJ · T PQ = TJQ
IP
, permutation
of indexes: gij → gji .
...i...
∑ ...i...
“Trace”: T...i... = i T...i... .
Calculations.
72
Rij = ∇i ∇j − ∇j ∇i = ∂i Γj − ∂j Γi + Γi Γj − Γj Γi .
′
Theorem: Rij = ∇′i ∇′j − ∇′j ∇′i = G−1 Rij G.
Proof: We have ∂i G−1 = −G−1 ∂i G G−1 , because ∂i (G−1 G) = 0 = ∂i G−1 G+
G−1 ∂i G = 0. O.K.
( )
Fxx Fxy
K = det = Fxx Fyy − Fxy
2
, x = 0, y = 0.
Fxy Fyy
Calculate Riemannian Curvature: R12;12 – only this component is non-
trivial.
2 2 2
g12;12 = Fxx Fyy + Fxy , g11;22 = 2Fxy , g22;11 = 2Fxy , x, y = 0, 0.
Using formulas for Rij;kl (0, 0) in these coordinates (See Appendix to the
previous Lecture), we obtain the following results:
p
R = 2R12;12 = 2K, Rij = Ri;pj , R = Rpp .
73
Consider a pair of vector fields ξ, η. Define ∇η = η i ∇i .
Curvature:
R̂ηξ = ∇η ∇ξ − ∇ξ ∇η − ∇[η,ξ] ,
[η, ξ] = [η i ∂i , ξ j ∂j ] = (η i ∂ξ j /∂xi − ξ i ∂η j /∂xi )∂j .
We have
R̂η,ξ = R̂ij η i ξ j (Matrix)
for any vector bundle.
Quadratic Form (tangent bundle, symmetric connection)
Metrics: S n , Rn , H n .
Quadratic form Rij;kl is determined by one constant R. All curvatures if
all 2-directions are the same at all points:
• Positive for S n .
• Negative for H n .
• 0 for Rn .
Symmetry Groups are On+1 for S n , Iso(Rn ) and O1,n for H n . Dimension
of these groups is n(n + 1)/1. Any points can be mapped to any point
(homogeneous space). Any unit tangent vector cam be mapped to any unit
tangent vector.
74
Every pair (x, v) can be mapped to every pair (y, w), where x, y are
points, v, w are tangent vectors at the points x, y respectively, |v| = |w| = 1.
Homework 5.
1. Prove that RP2 \point is diffeomorphic to MOBIUS BAND.
b
x0 N
x 1... x n
0
75
Prove that
∑ ∑
n
2 (dxi )2 2
dl = , r = (xi )2
(1 + r2 )2 i=1
5. Prove that straight lines passing through center are geodesics for the
metric of S2 :
dz dz̄
dl2 = 4
(1 + |z|2 )2
Homeworks 2, 3, 4. Solutions.
Homework 2. Solutions.
1. Projection Coordinates for Sn ⊂ Rn+1 . We need 2(n + 1) coordinate
domains
n=1
- local coordinates, Uj = Rn .
3. SO3 = RP3 : axis + angle ≤ π .
76
11111111111111
00000000000000
π
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111 (φ = π) ∼
= (φ = − π) .
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
00000000000000
11111111111111
−π11111111111111
00000000000000
00000000000000
11111111111111
4. SO1,2 : B̃ij = − B̃ji , Bik = g ks B̃si ,
1 0 0
g ks = 0 −1 0
0 0 −1
6 - 7. Quaternions R4 ∋ q
8.
a) GLn (R) has 2 components only.
b) GLn (C) is connected.
(b) Proof.
Set of matrices with distinct eigen-values is dense. Write linear operators
in basis of eigenvectors for GLn (C):
λ1 . . . 0
.. . . .
. . .. , λj ̸= 0 , λj ∈ C\0
0 . . . λn
it is connected space.
(b) is proved.
77
(a) Proof.
Let our field is R . Choose the same type basis consisting of 1-dim blocks
if λj is real or 2-dim blocks (λ, λ̄) for complex λ. We have for the forms of
blocks
λj ∈ R − (λj )
( )( )
cos bj sin bj ±1 0
λj ∈ C − a j
sin bj sin bj 0 1
Deform λj → ±1 , aj → ±1 , bj → 0 .
Now remember that ( )
−1 0
0 −1
is connected with 1
A
ξ−
→ −ξ
ξ ∈ R2
many times we see that there are 2 components ( det > 0 , det < 0) .
(a) is proved.
Homework 3. Solutions.
⟨ζ , ζ⟩ = 1 , ⟨η , η⟩ = − 1 , ⟨ζ , η⟩ = 0 ⇒
78
|a|2 − |b|2 = 1 , |c|2 − |d|2 = − 1 , a c̄ − b d¯ = 0
solution: ( )
a b
, a ā − b b̄ = 1
b̄ ā
SL2 (R) : e1 , e2 → a e 1 + b e2 , c e 1 + d e 2
ad − bc = 1 , a, b, c, d ∈ R
Take basis
e = e1 + i e2 , ē = e1 − i e2
e1 = (e + ē)/2 , e2 = (e − ē)/2i
In the new basis we have
1 [( ) ( ) ]
e → a + d − i (b − c) e + a − d + i (b + c) ē = q e + p ē
2
1 [( ) ( ) ]
ē → a − d − i (b + c) e + a + d + i (b − c) ē = p̄ e + q̄ ē
2
where
1( )
q q̄ − p p̄ = (a+d) + (b−c) − (a−d) − (b+c)
2 2 2 2
= ad − bc = 1
4
4. SL2 (R) ∼
= S1 × R2 (?)
( ) ( )
cos φ sin φ λ µ
A ∈ SL2 (R) ⇒ A = ×
− sin φ cos φ 0 1/λ
where λ ∈ R+ \0 ∼
= R, µ ∈ R.
Remark. Every “semisimple” Lie Group is topologically a product:
G ∼= K × RN , where K is a “compact group” and RN is a “Borel
subgroup” (upper triangle).
4.
dl2 = dθ2 + sin2 θ (dφ)2 , S2 ⊂ R3
For polar coordinates in R2 : x = ρ cos φ , y = ρ sin φ we have
79
5.
H2 : z 2 − x2 − y 2 = 1
z = ch θ , x = sh θ cos φ , y = sh θ sin φ
− dl2 = dθ2 + sh2 θ (dφ)2
Homework 4. Solutions.
1. We have parametrization (physics)
1 w
ch ψ = √ , sh ψ = √
1 − w2 1 − w2
because
(ch ψ)2 − (sh ψ)2 = 1
v/c 1
x1 = √ x′0 + √ x′1
1 − v /c
2 2 1 − v /c
2 2
So we have
1 v/c2
t = √ t′ + √ x′
1 − v /c
2 2 1 − v /c
2 2
v 1
x = √ t′ + √ x′
1 − v /c
2 2 1 − v /c
2 2
t ≃ t′ , x ≃ x′ + v t′
80
b) Iso (R2 )− ∼
= reflection + shift along the reflexion line .
B’
A
A’
B
81
3. Vector Bundles and Differential-Geometrical (Linear) Connection, Cur-
vature, Gauge Transformation for Connection and Curvature.
Parallel Transport.
f (x) : M n → R.
∂xi j
Ω = f (x(y))dx1 ∧ . . . ∧ dxn , dxi = dy , (summation in j)
∂y j
82
where I = (i1 < i2 . . . < ik ), and for x = x(y) we have
(bilinear, associative.)
Examples:
1.
∑ ∑ ∑ ∑
( ai dxi ) ∧ ( bj dxj ) = ai bj dxi ∧ dxj = (ai bj − aj bi )dxi ∧ dxj .
i j ij i<j
2. Change of coordinates:
∂xi j
x = x(y), dxi = dy , (summation in j)
∂y j
3. Corollary:
(a) k ≤ n.
(b) Ωn = f (x)dx1 ∧ . . . ∧ dxn .
f ∗ : ∧∗ (N m ) → ∧∗ (M n ),
83
For scalar functions we have:
ϕ : N → R, f ∗ ϕ(y) = ϕ(x(y)) by definition.
For forms
∂xi
f ∗ dxi =
(x(y))dy j by definition,
∂y j
products are mapped into products.
The proof is finished.
Homework 5. Solutions.
1.
η η
a
RP 2= RP 2\ Pt= = a a
a
−η −η
2.
b
a a
RP 2 ♯ RP 2 =?.
K2 =
K 2 \ MOBEUS ∼
= MOBEUS
Mobeus b
band (dashed)
1 ψ
ϕ S n \N ∼
= 1Rn n
x ,...,x
1 x l
0 x
84
∑
(dxi )2
dl 2
= const · ?
(1 + |x|2 )2
a) n = 1 (line): x = 2 tg ψ
b) n = 2 preservation of angles
5.
dz dz̄
|z| < 1 , dl2 = 4
(1 + |z|2 )2
f∗ ∂xi j
dxi ⇒ dy (sum).
∂y j
85
Functorial properties: g : M → N , x = x(y).
y x
∑ ∂xi
G∗ Ω = fI (x(y))dxi1 ∧ . . . ∧ dxik , dxi ⇒ (x(y))dy j .
∂y j
Algebra of forms.
∑
n
∗
∧ (M ) = ∧k (M ),
k=0
g (Ω1 ∧ Ω2 ) = g ∗ Ω1 ∧ g ∗ Ω2 .
∗
Integration of forms.
1. ∫
f (x) = ±f (P ), k = 0, ±P = “point with orientation”.
±P
2. ∫
f (x)dx1 ∧ . . . ∧ dxn = ordinary integral
D
where ∑
ϕq (x) ≡ 1, ϕq is a C ∞ function,
q
86
Properties: Integration of forms does not depend on local coordinates both
in M n and Dk (restrict the k-form to the k-dimensional body and integrate).
Differentiation of forms:
d : ∧k (M n ) → ∧k+1 (M n )
1. df is usual differential.
2. d(dxi1 ∧ . . . ∧ dxik ) = 0.
Example:
∑ ∂fi ∑ ( ∂fj ∂fi
)
d(fi dx ) = dfi ∧ dx =
i i
j
dx ∧ dx =
j i
i
− j dxi ∧ dxj .
i,j
∂x i<j
∂x ∂x
Examples of forms:
Momentum
∑ ∂f i (pi ) in Calculus of Variations. Differential of function
df = ∂xi
dx .
Electric field (Ei )-covector field.
2-forms: Magnetic field in R3 . B = Bαβ dxα dxβ .
Electromagnetic field in R4 = (x0 , x1 , x2 , x3 ).
| ∧ .{z
Ω . . ∧ Ω} − volume forms in T∗ (M n ).
n times
87
Kähler Riemannian Metric in Complex Space (Manifolds) with complex
linear coordinates (z 1 , . . . , z n ).
∑
dl2 = gα,β̄ dz α dz̄ β > 0.
α,β
Examples: n=1.
dzdz̄ : R2
dl2 = gαβ dz α dz̄ β = dzdz̄ S2
:
(1 ± |z|2 )2 H2
i dz ∧ dz̄ dx ∧ dy
Ω= = − Area Form.
2 (1 ± |z| )
2 2 (1 ± (x2 + y 2 ))2
2. d ◦ d = 0.
Proof:
a)
( ) ∂f ∂xi j
d g ∗ f (x(y)) = g ∗ df : d(g ∗ f ) = d(f (x(y))) = dy = g ∗ df.
∂xi ∂y j
b)
d(dxi ) = d(0) ∧ dxi = 0,
( i )
∗ dx j ∂ 2 xi
i
d(g dx ) = d dy = dy q ∧ dy j ≡ 0.
dy j ∂y j ∂y q
88
So d commutes with the map g ∗ for k = 1.
For k > 1 our result is obvious by the following reason: d(dxi1 ∧ . . . ∧
dxik ) = 0. Let us prove now that the Statement 1 of the Theorem. Let
Ωk = f · dxi1 ∧ . . . ∧ dxik , Ωl = g · dxj1 ∧ . . . ∧ dxjl .
[ ]
d(Ωk ∧ Ωl ) = d (f · dxi1 ∧ . . . ∧ dxik ) ∧ (g · dxj1 ∧ . . . ∧ dxjl ) =
Faraday Laws:
n = 3 : a) dB = 0, Magnetic field
1 ∂B α
b) dE = c ∂t , E = Eα dx ,
n = 4 : d(Fij dxi ∧ dxj ) = 0. Electromagnetic field
89
Lecture 19. Differential forms. Cohomol-
ogy.
Consider the algebra of differential forms on a manifold N m . Let Ω ∈
∧k (N m ). Then dΩ ∈ ∧k+1 (N m ). Consider a map f : M → N . Let us
check, that f ∗ dΩ = df ∗ Ω.
Proof. Let ϕ : N → R, f ∗ ϕ(y) = ϕ(x(y)).
( )
∗ ∗ ∂ϕ i ∂ϕ ∗ i ∂ϕ ∂xi j
f dϕ(y) = f dx = (x(y))f (dx ) = (x(y)) dy =
∂xi ∂xi ∂xi ∂y j
∂(f ∗ ϕ) j
j
dy . O.K. f ∗ d = df ∗ ,
∂y
for scalars and dxi . But the product ∧ commutes with f ∗ . Every form is a
combination of ϕ(x) · dxi1 ∧ . . . ∧ dxik . So we have
f ∗ d = df ∗ for all k ≥ 0.
Homology (Cohomology).
H k (M n , R) = Ker d/ Im d in ∧k (M n ).
2. Closed 1-form
∑ ∑ ( ∂ψj ∂ψi
)
ω= i
ψi (x)dx , dω = i
− j dxi ∧ dxj .
i<j
∂x ∂x
90
∂ψj ∂ψi
dω = 0 ⇔ i
− j ≡ 0,
∂x ∂x
∫x γ1
ϕ(x) = ω ⇒ dϕ = ω if ϕ is well-defined. x1
x0 γ2
x0
Necessary condition: dω = 0.
τ
Stokes Formula. Consider an oriented n
manifold M n with boundary
∂M n = W n−1 with “induced” orientation.
What is it? M
W
Consider “external” normal vector n to W ⊂ M and tangent frame τ ton
1. ∫ ∫ ∫ ∫
··· Ω= ··· dΩ,
W n−1 Mn
W n−1 Mn
91
(0,1) (1,1)
x2 W = ∂I n (I n =“ball”).
(0,0) x1 (1,0)
a) Case n = 1. We have
∫b
Ψ′ dx = Ψ(b) − Ψ(a), Ω = Ψ(x). b
a
a
xi =1 ∫ ∫ ∫ ∫
··· Ω(i)
n = ··· Ω(i)
n =
∂I n [xi =0,1]
xi =0
[∫ ∫
= (−1)i−1 ci ∧ . . . ∧ dxn −
· · · Ψi (x1 , . . . , xi = 1, . . . , xn )dx1 ∧ . . . ∧ dx
∫ ∫ ]
− ··· i c
Ψi (x , . . . , x = 0, . . . , x )dx ∧ . . . ∧ dx ∧ . . . ∧ dx =
1 n i 1 n
∫ ∫
= (−1) i−1
··· ∂i Ψi (x1 , . . . , xi , . . . , xn )dx1 ∧ . . . ∧ dxi ∧ . . . ∧ dxn =
In
∫ ∫
= ··· dΩ(i)
n .
In
92
Let n = 2. Orientation for n = 2:
d c
a x1 b
Ω = Ψ1 (x1 , x2 )dx1 + Ψ2 (x1 , x2 )dx2 ,
( )
∂Ψ2 1 2 ∂Ψ1 1 2
dΩ = (x , x ) − (x , x ) dx1 ∧ dx2 ,
∂x1 ∂x2
∫ ∫ 1 ∫ 1 ∫ 0 ∫ 0
1 1 2 2 1 1
Ω= Ψ1 (x , 0)dx + Ψ1 (1, x )dx + Ψ1 (x , 1)dx + Ψ1 (0, x2 )dx2
0 0 1 1
∂I 2
∫1
( )
= dx1 Ψ(x1 , 0) − (Ψ(x1 , 1) .
0
We have ∫ ∫ 1 ∫ 1
Ω= Ψ1 (x , 0)dx −
1 1
Ψ1 (x1 , 1)dx1 .
0 0
∂I 2
0
• ϕ : I k → M n – “singular cube”.
0 1
93
0 2
ψ : σ → M – “singular simplex”.
k n
•
σ k =< 0, 1, . . . , k >. 0 1
0 1
C ∞ − singular cubes ⇒R
C ∞ − singular simplices ⇒ R,
Homework 6.
1. Find all geodesics for Sn , Rn , Hn (for Hn in P - model and in K -
model.
2. Find all closed geodesics in torus T2 with euclidean metric.
3. Find all closed geodesics in RP2 (metric of constant positive curvature).
4. Prove that parallel transport along any path preserves inner product
γ(t) : [xi (t)] ,
ξ (t)
⟨η(t), ζ(t)⟩ = const if ∇ẋ η = 0, ∇ẋ ζ = 0
η(t) γ (t)
(∇s gij ≡ 0 ).
94
Lecture 20. Differential forms and tensors.
Categorial properties. Cohomology and Stocks
formula. Homotopy invariance of Cohomol-
ogy.
Category of C ∞ -manifolds and C ∞ -maps. Functors like spaces of tensors
with low indices. Differential k-forms ∧k (M n ): exterior multiplication ∧,
operator d and integration, their properties. Stokes formula fro integration.
Singular cubes and simplices, boundary operator ∂. Cohomology, forms and
singular complexes.
a) H∧k (M n , R) (forms).
a) Ker
| {z d} / Im d
|{z} = H∧k (M n , R).
closed forms exact forms
∗ ∗
| {z∂} /
b) Ker Im
| {z∂} = H k (M n , R) (cubes, simplices).
cocycles coboundaries
c) |Ker
{z ∂} / Im ∂ = Hk (M n , R) (cubes, simplices).
|{z}
cycles boundaries
95
“Boundary operator:” ∂ : k-chains→ k − 1-chains (linear).
∑
n
k
∂(σ , ϕ) = (−1)i (< 0, 1, . . . , bi, . . . , k >, ϕ),
i=0
“Stokes formula”
∫ ∫
dΩ =< dΩ, b >=< Ω, ∂b >= Ω,
b ∂b
H∧k (M n , R) → Hsimplices
k
(M n , R).
In Algebraic topology:
k
Hsimplices (M n , R) → Hom(Hk (M n ), R).
Isomorphism
H∧k (M n , R) ∼
= Hom(Hk (M n ), R)
was claimed by Poincaré in 1895 and proved by De-Rham in 1930s.
Lemma 1. Cohomology H∧k (M n , R) form a ring (operation ∧) such that
a∧b∼
= (Ω ∧ Ω′ + Im d) in H k+l (M n , R).
96
We have
but
We have:
1
Proof.
du = dy a + dt ∧ ȧ − dt ∧ dy b,
97
∫ 1 ∫ 1 ∫ 1
Ddu = ȧdt − dy bdt = a t=1
−a t=0
− dy bdt,
0 0 0
but
∫ 1 ∫ 1
a t=1
=u t=1
, a t=0
=u t=0
, bdt = Du, dy bdt = dDu.
0 0
O.K.
Finally we have for every closed k-form Ω in M :
Ω → F ∗ Ω → DF ∗ Ω in ∧k−1 (N ).
t N I
dΩ = 0, dF ∗ Ω = 0,
0
so we have
X X
dD(F ∗ Ω) + X
Dd(F ∗
XΩ)
∗
X=F Ω t=1
− F ∗Ω t=0
= g ∗ Ω − h∗ Ω.
So g ∗ Ω = h∗ Ω + Im d,
g ∗ ≡ h∗ : H k (M ) → H k (N ).
Theorem is proved.
Corollary. For every contractible manifold M like point, Ball, space Rn and
so on cohomology are the same:
{
∗ R, k = 0,
H∧ (point) =
0, k ̸= 0.
98
Lecture 21. Homotopy invariance of Co-
homology. Examples of differential forms.
Symplectic and Kähler manifolds.
Homotopy invariance of cohomology.
g:N →M
C ∞ − maps:
h:N →M
F :→ N × I → M, F t=1
= g, F t=0
= h.
Lemma.
Ddu + dDu = u t=1
−u t=0
.
Proof.
du = dy a + dt ∧ ȧ − dt ∧ dy b,
1
∫ 1 ∫ 1
t dy Du = dy bdt = (dy b)dt
0 0
0 y
N
99
Ddu = D(d(a + dt ∧ b)) = D(dy a + dt ∧ ȧ − dt ∧ dy b) =
∫ 1
= a t=1 − a t=0 − dy bdt = a t=1 − a t=0 − dDu.
0
O.K.
Proof of the Theorem.
F :→ N × I → M ,
Ω
F ∗ dΩ = df ∗ Ω = 0,
XX
dD(F ∗ Ω) + X
Dd(F ∗
XΩ)
∗
X=F Ω t=1
− F ∗Ω t=0
= h∗ Ω − g ∗ Ω.
g ∗ ≡ h∗ : H k (M ) → H k (N ).
O.K.
Homotopy equivalent manifolds N , M :
∃ N−
→M −
→N
ϕ ψ
such that
ψ · ϕ is homotopic to 1N : N → N
ϕ · ψ is homotopic to 1M : M → M.
ϕ∗ : H ∗ (M ) → H ∗ (N )
ψ ∗ : H ∗ (N ) → H ∗ (M ).
100
1. k = 0, df = 0 ⇒ f = const (in every component).
H 0 (M ) = Rp , where p = number of components.
101
We have:
(dxi ∧ dpi ) ∧ (dxj ∧ dpj ) = (dxj ∧ dpj ) ∧ (dxi ∧ dpi ).
Step 3. Return to the original basis. Both sides are n-forms and
transform in the same way.
So Theorem is proved.
102
Lecture 22. Symplectic Manifolds and Hamil-
tonian Systems. Poisson brackets. Preser-
vation of Symplectic form by Hamiltonian
System.
∂xi ∂ 2 xi
dṕj = dpi j
+ pi j k
dy k
∂y ∂y ∂y
∑ ∑ ∂xi XXX∂X xXj
2 i
dy j ∧ dṕj = dy j ∧ dpi + dpi X
j ∂y k X
dy ∧
XXX =
dy k
j ij
∂y j ∂y X
∑ ∂xi ∑
= j
dy j ∧ dpi = dxi ∧ dpi .
ij
∂y i
O.K.
More generic simplectic form in T∗ N is
∑ e ∑
dxi ∧ dpi + B = ΩB , B = bij (x)dxi ∧ dxj , dB = 0.
c i<j
103
For k = 3 it is “Magnetic Field” (correction of simplectic form).
Hamiltonian system in: T∗ N = M is defined by “Hamiltonian” H(x, p),
∂H
ẋi =
∂pi
∂H
ṗi = −
∂xi
Another form: Take
∑ covector field (dH). Take inner product given by
simplistic form Ω = dx ∧ dpi .
i
( )
0 1
Ωij = .
−1 0
“Poisson Bracket”.
f˙ = {H, f }.
∂L ∂L
H = ẋi i
− L, pi = .
∂ ẋ ∂ ẋi
Euler-Lagrange equations:
∂L
ṗi = ⇔ ẍi + Γijk ẋj ẋk = 0.
∂xi
104
Theorem 1. In every compact closed simplectic manifold M n (n = 2k)
with form Ω (nondegenerate, we have
H 2j (M ) ̸= 0, j = 1, . . . , k.
∫
Proof. We know that Ωk = k! volume, so Ωk ̸= 0. Therefore H 2j ̸= 0,
because Ω2j can not be exact (if Ω2j is exact then Ω2k is exact.)
Theorem 2. For any simplectic manifold M n and Hamiltonian system with
Hamiltonian H : M n → R simplectic form is preserved by mapping St :
M n → M n (time shift by our system).
Proof. Locally we choose such coordinates (x1 , . . . , xk , p1 , . . . , pk ) that Ω =
∑
k
dxi ∧ dpi (Darboux).
i=1
Our system is:
∂H ∂H
ṗi = − , ẋi =
∂xi ∂pi
Small time shift is:
St : pi → pi + ṗi t + O(t2 )
xi → xi + ẋi t + O(t2 )
We have
∑ ∑[ ∂H
] [
∂H
]
St∗ (dxi ∧dpi ) = i i
(dx +ẋ t)∧(dpi +ṗi t) = i
dx + td ∧ dpi − td i =
i i
∂pi ∂x
∑ [ ]
∂H ∂H
= dx ∧ dpi + t d
i
∧ dpi − dx ∧ d i =
i
i
∂pi ∂x
∑ [ ∂ 2H
∂ 2
H ∂ 2
H
∂ 2
H
]
j ∧ dpi +
dp dx ∧ dpi − dx∧
j i
dx − dx ∧ i
j i
∂xi ∂xj
= Ω+ t ∂p dpk =
∂p i j ∂p i ∂x j ∂x ∂pk
ij
=Ω
105
Theorem is proved.
Remark 1. “Multivalued Hamiltonian”= closed 1-form dH.
Remark 2. “Poisson structure”: Ω may be degenerate, Ω∗ is well-defined
but degenerate (Ω∗ = Ωij ).
Remark 3. The Poisson Structure is a symplectic inner product of
covectors which defines a Lie Algebra of functions (the Poissoon Bracket):
{f, g} =< df, dg >Ω
{{f, g}, h} + {{g, h}, f } + {{h, f }, g} = 0.
Here we have Ωij = (Ω)−1 if it is nondegenerate. This definition implies
in that case dΩ = 0, Ω = Ωij dxi ∧ dxj .
Remark 4. According to the Darboux Theorem, every nondegenerate
Symplecttic Geometry with dΩ = 0 is always ”flat”, i.e. there exists system of
local coordinates such that Ω is constant (and reduced to standard canonical
form.)
Homework 7.
1. Let action functional is
∫ ∫
S {γ} = i j
gij ẋ ẋ dt + Ai (x) dt
γ γ
107
Theorem. Let group G be compact. There exists Riemannian Metric in
M = G/H s.t. g ∗ gij (x) = gij (x), g ∈ G, g : M → M , g(x)- action of the
group G.
Remark. We assume that there exists a G-invariant volume form Ω in
the group G. We call it dσ(g).
Proof of the Theorem. Take any Riemannian Metric gij (x) in M . Con-
sider family of metrics g ∗ gij (x) = gijh (x) in M , depending on g ∈ G as
parameter. Integrate:
∫
gijg (x)dσ(g) = Gij (x).
G
G G G
108
1. This is an invariant form.
H 0 (S n ) = H n (S n ) = R.
1, Ω, Ω2 , . . . , Ωn .
109
1. Rn , dx1 , . . . , dxn , A ∈ SOn .
∑
A(dxi ) = aij dxj , AAt = 1.
j
2. Cn , dz 1 , . . . , dz n , dz̄ 1 , . . . , dz̄ n , A ∈ Un .
∑ i j
A(dz i ) = aj dz ,
∑ i j
j
A(dz̄ i ) = āj dz
j
For n = 1.
R2 = C1 : dz, dz̄, A = (eiϕ ) = U1 .
dz → eiϕ dz
dz ∧ dz̄ → dz ∧ dz̄.
dz̄ → e−iϕ dz̄
110
1. S n = SOn+1 /SOn .
S 2n−1 = Un /Un−1 .
S 4n−1 = Spn /Spn−1 .
3. Stiefel Manifolds.
Vn,k = SOn /SOn−k (orthogonal k-frames in Rn ).
C
Vn,k = Un /Un−k (complex version).
4. Grassmann Manifolds.
Gn,k = SOn /SOk × SOn−k .
GCn,k = Un /Uk × Un−k
111
2. Operator of “Averaging” maps H k (M ) as a unit homomorphism:
∫
1
ω= h∗ ω dσ(h), h ∈ G,
|G|
G
H ∗ (M ) ∼ ∗
= Hinv (M ) ← calculated with G − invariant forms only.
∧0 = R, ∧1 = R3 , ∧2 = R3 , ∧0 = R,
d : ∧1 → ∧2 is an isomorphism.
For n = 1.
dz → eiϕ dz
⇒ invariant form dz ∧ dz̄.
dz̄ → e−iϕ dz̄
∑
n
k ≥ 1: invariant form are dz i ∧ dz̄ i and all Ωk ,k = 1, . . . , n.
i=1
112
Cohomology rings.
1. Tori T n = S 1 × . . . × S 1 = Rn /Zn
H ∗ (T n ) = ∧(v1 , . . . , vn ), dim Vj = 1.
H 0 (S n ) = H n (S n ) = R.
3. S n1 × . . . × S nk , G = SOn1 +1 × . . . × SOnk +1 .
4. CP n .
H ∗ ( , R) =?
113
Approximation. Every continuous map of compact manifolds f : M n →
x
N m , y(x) can be approximated by homotopic C ∞ -map g : M n → N m :
y
max |f (x), g(x)| < ϵ.
x
V
Lemma. For ϵ > 0 small enough maps f and g are homotopic.
Proof. Take Riemannian Metric gij (y) in N m .
Our homotopy it such that every point moves from f (x) to g(x) along
this short geodesics homogeneously and reaches its end at t = 1. O.K.
Corollary. Homotopy class (C ∞ ) for c∞ maps M → N are the same as C 0
homotopy classes.
Transversality. Let W m−k ⊂ N m be a C ∞ -submanifold. A map f : M n →
N m is called “transversal along W ” iff for every x ∈ f −1 (W ) map of
tangent spaces
(π ◦ df ) : Txn
df
/ Tm π / T m /T m−k ∼
= Rk
f (x) normal projection 6N W
normal plane
has rank k.
Examples.
2. n = k:
114
(a) f 1 (W ) are isolated points x1 , . . . , xl , (df )xk has rank k.
(b) Special case: m = k, dim W = 0 (point).
a) 2
N 1 b) m=k=n
1 W
f(M )
...
x1 x2 xl W
k=1, m=1, n=2 Txi ∼ = R = Rm
= TN ∼ n
115
These equations are NONDEGENERATE because of TRANSVER-
SALITY long W (i.e. at the points f (x) ∩ W ).
So we have nondegenerate submanifold
V n−k ⊂ M
and map
V n−k → W n−k
is isomorphic along normal k-planes by definition of TRANSVERSAL-
ITY.
f
Examples. Function Rn − → R ⊃ W = point. Level f (x) = y0 is transversal
iff df ̸= 0 for all f (y). Topological invariants.
1
1. Let n = 1.
N +−N−= 0 N +−N−= 1
f −1 (point) = N+ ∪ N− ,
3. Intersection index.
f
f(M 1) Mn −
→ N n+p ⊃ W p .
1 2
W N
(M n , f ) ◦ W p ∈ Z
if oriented, otherwise modulo 2.
n=p=1 n+p=2
116
IMPORTANT 1ST CASE: “Empty set transversality”
f : M n → N m ⊃ W m−k , n < k, f 1 (Ω) = ∅ (transversal map).
Homework 8.
∑
1. Prove that the expression i pi dxi (locally in T∗ (N k )) defines correctly
1-form in T∗ (N k ) .
2. Prove that geodesic
∑ flow (Hamiltonian H(x, p) = g ij (x) pi pj / 2 ) pre-
i n
serves the form i pi dx at the level H = const in the manifold T∗ (M ) .
111111111111
000000000000
000000000000
111111111111
000000000000
111111111111 x0
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111 Disk
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
000000000000
111111111111
5. Prove that the flow ẋi = ξ i (x) preserves the volume form
Ω = dx1 ∧ · · · ∧ dxn if
∑n
∂ξ i (x)
= 0 ( ⇔ St∗ Ω = Ω )
i=1
∂xi
117
Homework 6 and Homework 7. Solutions.
Homework 6. Solutions.
1. Geodesics for Sn , Rn , Hn . Metric
m
y y = x + const
n
(0,0) x (1,0)
RP 2
η
ξ
3. Closed geodesic in RP2 (all nongomotopic to zero)
−ξ
−η
118
4. Parallel transport along the path γ(t) , x1 (t), . . . , xn (t) , connection
∇ gij ≡ 0 .
ξ (t)
η(t) γ (t)
d d ( )
⟨η(t) , ζ(t)⟩ = gij η i (t) ζ j (t) =
dt dt
( ) ( ) ( )
= ∇ẋ gij η (t) ζ (t) + gij ∇ẋ η (t) ζ j (t) + gij η i (t) ∇ẋ ζ j (t) = 0
i j i
(Leibnitz Identity).
Homework 7. Solutions.
1. ∫ ∫
S {γ} = γ
gij ẋi ẋj dt + γ
Ai (x) dt
↑ ↑
S0 (γ) S1 (γ)
A = Ai (x) dxi is one-form in M n .
( )
d ∂L1 ∂L1
EL(i) = EL0(i) + i
− = 0
dt ∂ ẋ ∂xi
( )
∂L1 d ∂L1 ∂Ai k ∂L1 ∂Aj j
i
= Ai (x) , i
= k
ẋ , i
= ẋ
∂ ẋ dt ∂ ẋ ∂x ∂x ∂xi
∑ ∂Ai ∑ ∂Aj
EL(i) = EL0(i) + k
ẋk
− i
ẋj =
k
∂x j
∂x
( )
∂Ai ∂Ak
= EL0(i) + − ẋk = EL0(i) + Bik ẋk
∂xk ∂xi
(dA = B).
119
2.
dφ
∫ z }| {
γ
k(s) ds = 2π n ?
γ - closed
φ - rotation of normal (and tangent)
γ (t) vector.
dẋ(s)
k(s) =
ds
s - natural parameter.
Another form:
(check!)
3. Area form
√
R2 : dx ∧ dy = r dr ∧ dφ , r = det gij
( )
1 0 √
dl2 2 2 2
= dr + r dφ , gij = , g = r
0 r2
4 dz ∧ dz̄
S2 : = sin θ dθ ∧ dφ
(1 + |z|2 )2
( )
2 4 dz dz̄ 1 0
dl = = dθ2 + sin2 θ dφ2 , gij =
(1 + |z|2 )2 0 sin2 θ
4 dz ∧ dz̄
H2 : = sh χ dχ ∧ dφ
(1 − |z|2 )2
( )
4 dz dz̄ 1 0
dl 2
= = dχ2 + sh2 χ dφ2 , gij =
(1 − |z| )
2 2 0 sh2 χ
4.
xdy − ydx y
dφ = const · , φ = arctg
x2 + y 2 x
120
y
ϕ
x
M n = ∪α Uα , x1α , . . . , xnα
Conclusion:
d AUReal = AVReal = H
(same for all U α ).
Obviously d H = 0 .
Remark 1.
a) H ∈ H 2 (M n , R) is the first Chern class of a line bundle.
b) It does NOT depend on the choice of connection.
Proof. Let us have ÃU = AU + iΨU , ÃV = AV + iΨV . We have then
in U ∩ V : ΨU = ΨV = Ψ . So we get, that Ψ is a globally defined 1-form
in M n and
H̃ = H + dΨ
where dΨ is an exact form in M n .
121
c) ∫
1
H
2π 2−cycle
is integer ( ∈ Z).
Transversality.
f : M n → N m ⊃ W m−k
(along W )
df π m−k ∼ k
→ Tfm(x) −
Txn − → TNm /TW =R ,
rk(π ◦ df ) = k for all x ∈ f −1 (W ).
Example. n < k ⇒ f −1 (W ) = ∅.
Applications (imbeddings and immersions).
Theorem 1. Every map f : M n → RN can be approximated by NONDE-
GENERATE imbedding if N ≥ 2n + 1 (M n is compact C ∞ manifold).
Proof.
ϕ ψ
Step 1. Consider any C ∞ -map M n − → RN and C ∞ imbedding M n −
→ RQ .
Their product ϕ × ψ : M n → RN +Q = RP gives us C ∞ imbedding Φ :
M ⊂ RN +Q = RP , N + Q > 2n + 1.
n
122
Lemma 1. For all vectors ±l outside of the image R : M n × M n \∆ → S P −1
x−y
R(x, y) = ∈ S P −1 , ±l ̸∈ R(M n × M n \∆).
|x − y|
the projection map πl ◦ Φ is imbedding.
Proof. πl (x) = πl (y) ⇒ ±l ∈ T (M × M ).
Lemma 2.
τ Consider the map T1∗ (M n ) −→
R 1
S P −1 , where
x (x, τ ) ∈ T1∗ (M n ), |τ | = 1, τ is tangent to M n at x.
R1 (x, τ ) = τ .
Let ±l ̸∈ Im R1 . Then the projection πl ◦ Φ : M n → RP −1 is an nonde-
generate immersion.
Proof. No one tangent vector to M n ∈ RP belongs to the kernel of the
projection if l not parallel to τ for all (x, τ ).
Lemma is true. O.K.
Proof of the Theorem 1. Let Φ be any imbedding M n ⊂ RN where
N > 2n + 1. We project it along the direction l, ±l ̸∈ Image(M × M \∆).
Almost all vectors have this property because dim Image(M ×M \∆) = 2n,
less then dim S N −1 ⊂ RN , N − 1 > 2n.
Here Transversality works. We apply transversality to the pair
M × M \∆ → S N −1 ⊃ (point).
123
Example.
n=1
N = 2n + 1. Projection to RN −1 may
•
be not imbedding. N=2
Singularities of projection:
R3
π: R1 R1 (projection from R2 )
R2
knot diagrams
πl
R1
124
N τW
W • τN - oriented frame to N .
f(M)
• τM - oriented frame to M .
τM
• τW - oriented frame to W .
1. Intersection number:
∑
M ◦ W in N = (−1)sign(xj )
xj ∈f −1 (W )
Degree of map.
f : M n → N n ⊃ W = (point), k = n = m.
f -transversal along W :
f −1 (W ) = x1 ∪ . . . ∪ xp ∈ M n ,
( )
sign(xj ) : τN /τM = sign det df xj
,
df xj : Rntangent to M → Rntangent to N.
| {z }
linear map
degW f ∈ Z2 .
125
Theorem 1. Degree of map and Intersection Index are the same for homo-
topic maps.
Proof. Consider homotopy F such that
F : M × I → N ⊃ W, F t=0
= f, F t=1
= g, F ∈ C ∞ ,
1. sign(x1 ) = − sign(x2 ).
x1 x2
x1 x2
2. sign(x1 ) = − sign(x2 ).
3. no point for t = 0, 1.
126
Conclusion. Total sum remains unchanged.
O.K.
Corollary (N = S n ): degree of map does not depend on the point W ⊂ N n .
Proof. Let N = S n . Rotate sphere S n ϕt , such that ϕ0 = 1, ϕ1 (W1 ) = W2 ,
A
γ1
S 1 ⊂ R2 (C ∞ -imbeddings). Two paths γ1 ,
B γ2
γ2 connecting a pair of points A, B.
127
3. F : M n → N n – map of closed oriented manifolds, m = n = k, W -
point.
Theorem.
∫ ∫
∗
f (Ω) = (deg f ) Ω, (Why?)
| {z }
M integer N
and ( )
∫ ∫ ∑
f ∗Ω = Ω · s
(−1) j .
j
f −1 (U ) U
128
We have
∑
deg f = (−1)sj , (deg f calculated at the point w ∈ U ),
j
by definition.
We know that deg f does not depend on w: it is same for all transversal
points w ∈ N , and their measure is 1.
So we calculated ∫ ∫
∗
f (Ω) = (deg f ) · Ω.
M N
Theorem is proved.
Corollaries.
f : (M, ∂M ) → (N, ∂N ) f
N
M
assuming that f (∂M ) ⊂ ∂N .
M N
In this case deg fM = deg f∂M (Why?) (Homework 10).
f = Pn (x) : R → R , y = a0 xn + . . . + an .
(x) (x)
129
a) {M n , W k } = M n ◦ B k+1 , ∂B k+1 = W k in Rn+k+1 .
b) {M n , W k } = Z n+1 ◦ W k , ∂Z k+ = M n in Rn+k+1 .
c) {M n , W k } = Z n+1 ◦ B k+1 in R+
n+k+2
.
Z B
M
n+k+2
W R+
M W M W
(here n−k=0)
2. Self-intersection (M n = W n ).
M n ◦ M n =? in N 2n .
Perturb M = (M̃ ).
M
M ◦ M = M̃ ◦ M .
M
Corollary. M n ◦ M n = 0 = −M n ◦ M n , n = 2k + 1.
Modulo 2 (nonorientable case).
Example.
RP 2 ξ
−η
M 1 ◦ M 1 = 1.
η
−ξ
130
Fixpoint (Lefschetz) problem:
f : M n → M n (orientable).
Solutions: f (x) = x.
Algebraic number of solutions: Let ∆ ⊂ M ×M , (x, x) ∈ ∆. “Lefschetz
Number” (x, f (x)) ∈ ∆f .
L(f ) = ∆f ◦ ∆.
Homework 9.
1. Prove that for all open 2-manifolds we have H 2 (M 2 , R) = 0 .
2. Prove that degree of map f : S1 → S1 can be calculated by
f (x + 2π) = f (x) + 2πm , m ∈ Z , m = deg f
(2π - circles).
3. Prove that intersection index of 2 closed curves in any domain U ⊂ R2
is equal to 0.
4. Prove that every map M n → Sm \W m−k is homotopic to zero for n <
k−1 .
5. Prove that every map Sn → Sm is homotopic to zero for n < m .
6. Calculate cohomology ring H ∗ (SO4 , R) = ?
7. Calculate H 1 (M 2 , R) = ? for M 2 = R2 \(∗1 ∪ ∗2 ) :
R2
*1 *2
131
For f : M n → N n we have deg f.
They have the following properties:
2. M n ◦ W k = (−1)kn W k ◦ M n .
M = ∂Z, W = ∂B, M ∩ W = ∅, {M n , W k } = M ◦ B = Z ◦ W.
M W M W
Gauss formula:
Let M = S 1 ∈ R3 , W = S 1 ∈ R3 .
M
dγ 1 dγ 2 dγ 3
W det dx 1
dx 2
dx3
γ 1 − x1 γ 2 − x2 γ 3 − x3
∥
γ (t) I I
x(s) 1 (d⃗γ (t) × d⃗x(s), ⃗γ (t) − ⃗x(s))
{M, W } =
M = γ(t), W = x(s), 4π |⃗γ (t) − ⃗x(s)|3
M W
γ(t) ̸= x(s),
ϕ = {γ(t), x(s)},
Is it closed 2-form in R3 × R3 \∆?
ϕ : T 2 → R3 × R3 \∆.
f : M n → M n , f (x) = x (?)
132
Algebraic (Lefschetz) number of fixpoints:
∆ = ∆f ⇒ f (x) = x.
∆ ◦ δf = 1.
Sn D n− g = ψ ◦ f ◦ ϕ ∼ 0 (obvious).
S’
X
St (X) : M n → M n , X − vector field,
St (X) ∼ 1, M n ⊂ T ∗ (M n ) M
133
n
M =(M ,tX)
χ(M n ) = M̃ ◦ M =?
(M n,0)
T∗ (M ) ∼
= T ∗ (M ), we use T∗ (M ).
M̃ ◦ M =?
Critical point: (df ) xk
= 0. Critical point xk is nondegenerate iff the
quadratic form ( )
2 ∂ 2f
df= dxi dxj
∂xj ∂xj xk
is non-degenerate.
“Sing” of xk is equal to the number of negative squares in the form
(d2 f )xk , which is the “Morse index” n(xk ).
134
T*(M) τ’
M
M
M M
xk
τ
135
Lecture 30. Two applications of differential
forms: Degree of Map and Hopf Invariant.
Two applications of differential forms.
f
I) Homotopy theory: Degree of map S n −
→ S n (spheres may be replaced
by any closed orientable manifolds:)
∫
A) Take n-form Ω ∈ ∧n (S n ), Ω = 1.
Sn
Definition. ∫
deg f = f ∗ (Ω)
Sn
Proof.
Let F be a homotopy from f to g: F : S n ×I → S n , (0 ≤ t ≤ 1), F (x, 0) =
f , F (x, 1) = g. Consider F ∗ (Ω)). We have dF ∗ (Ω) = F ∗ (dΩ) = 0.
∫ ∫ ∫
∗ ∗
F (Ω) − F (Ω) = dF ∗ (Ω) = 0.
x x
(S n ,1)
(S n ,0)
(S n ×I)
∗ ∗
g Ω f Ω
∫ ∫
Ω → Ω + dω ⇒ Ω + dω = Ω.
(S n ,t) (S n ,t)
Theorem is proved.
∫
We calculated integral f ∗ Ω geometrically and proved geometrically
( ) S n
∫ ∗ ∫
that f Ω = Ω · deg f .
Sn Sn
∫
B) Consider S 3 → S 2 and fix 2-form Ω in ∧2 (S 2 ) such that Ω = 1.
S2
Let f ∗ Ω = dω, ω ∈ ∧1 (S 3 ).
136
Theorem. H(f ) is homotopy invariant.
∫
Proof. Consider F : S 3 × I → S 2 , Ω = 1. Take F ∗ (Ω) ∈ ∧2 (S 3 × I).
S2
Take
We have
So ∫ ∫ ∫
∗ ∗
ω ∧ F (Ω) − ω ∧ F (Ω) = (F ∗ Ω)2 = 0.
t=1
x t=0
x S 3 ×I
∫ ∫
′′ ∗
ω ∧g Ω ω′ ∧ f ∗Ω
S3 S3
Here
F t=1
= g, F t=0
= f.
z
x0 = z(x1 , . . . , xn ), P = (0, 0, . . . , 0), (dz) P
= 0.
n(P)
137
Definition dσ – volume element at S n .
K ∗ = G∗ dσ.
138
The volume element of sphere at N is:
Map:
(1, −∇z )
⃗u = nQ = √ ,
1 + (∇z )2
1 −∂z/∂xj
w=√ , uj = √ , j = 1, . . . , n.
1 + (∇z )2 1 + (∇z )2
Calculate Jacobian:
( )
∂uj
J0 = det , x = 0, 0, . . . , 0 .
∂xk T
So we have
( )
∗ ∂ 2z
KdσM = G (dσ) = (−1) · det n
dx1 ∧ . . . ∧ dxn .
∂xi ∂xi (0,...,0)
Theorem is proved.
Conclusion: ∫ ∫∫
Kdσ = (deg G) · dσ.
M2 S2
139
How to calculate deg G in R3 (the same consideration can be done for
any compact M n ⊂ Rn+1 ).
M 2 ⊂ R3 , G : M 2 → S 2 , h : M 2 → R.
z
N
Lemma 1. Let S, N ∈ S 2 (the
South and the North poles) be
transversal points for G. Then
∇h = 0 iff Pj ∈ G−1 (N ) or Pj ∈
G−1 (S).
S
G−1 (N )
signSj = (−1)n+Morse Index in
Calculation:
∑
n {
n j n 0, n = 2k + 1,
χ(M ) = (−1) mj (z) = degN G+(−1) degS G =
2 deg G, n = 2k.
j=0
140
D.C. “Differential Structure”, Hausdorff.
Our Course: “Atlas of Charts”, C ∞ manifolds, metric spaces, “Dou-
ble good Atlas”.
Local coordinates - same.
Imbedding and Immersions M n = RN (proof of existence is missing in
D.C.).
Partition of unity (proof is missing in D.C.) - we proved for compact
manifolds.
Tangent vectors, basis of Tτn as ∂i = ∂/∂xi in local coordinates
(x1 , . . . , xn ) .
∑ i
Vector field X = a (x) ∂i (locally).
Our notation X = (ai ) (index is “upper”). Action of vector field on
functions f (x) :
∑ ∂f ∂f
D.C. : X (f ) = ai (x) i
= ai i
∂x ∂x
Commutator
[X , Y ] = X (Y (t)) − Y (X(t))
( i i
)
j ∂b j ∂a
i i i
X = (a ) , Y = (b ) , [X , Y ] = a −b
∂xj ∂xj
f
C ∞ -maps : M n −
→ N m , differential of map, rank :
Implicit function theorem / local inversion theorem of the map (no proof
in D.C. and in our course).
Approximation and Transversality are missing in D.C. or presented in the
highly reduced form - see C.P.
Examples of manifolds - similar.
Orientation.
i
Covectors∑(basis (dxi ) in local coordinates) are missing in D.C. Covector
fields ω = i ui (x) dx (1-forms).
Manifolds T ∗ (M n ) (vectors), Atlas, Charts.
Manifolds T∗ (M n ) (covectors, missing in D.C.)
141
Riemannian and Pseudoriemannian Metric = inner product (symmetric,
nondegenerate) is ⟨X , Y ⟩ , gij = ⟨∂i , ∂j ⟩ .
Symplectic inner products and differential forms are missing in
D.C.
√
Volume element dn σ = det gij dx1 ∧ · · · ∧ dxn (locally).
Induced Metric for submanifolds M n ⊂ RN or M n ⊂ N k (good only
for positive Riemannian Metrics).
Lie Groups, right and left invariant vector fields, right and left invariant
Riemannian Metric on Compact Lie Groups, bi-invariant metric and inner
product of right-invariant vector fields (D.C.).
Riemannian Metric and length of curves. Existence of Riemannian metric
in every manifold (proved by partition of unity in D.C.; another proof -
imbedding in RN ). Volume provided by Riemannian metric.
Existence of biinvariant volume on a compact group G. Existence of
invariant metric in the spaces with action of compact group G (“averaging
procedure” - both in D.C. and our course).
Connections (“affine”) - considered in D.C. for tangent bundles only:
vector field X, X = ai (x) ∂i ,
∑
∇X Y = ai ∇ i Y , Y = bj (x) ∂j
i
( )
∂bj ∂bj
∇i Y = b (∇i ∂j ) +
j
∂j = + bj Γkij ∂j
∂xi ∂xi
for components, ∇i (∂j ) = Γkij ∂k .
Symmetric Connection” Γkij = Γkji {∇i ∂j = ∇j ∂i } .
Curve
c(t) = {xj (t)} , X = (ẋ)
Let V be a vector fields, V = (v j ) , X(t) - tangent to c.
( j )
d dv
(V ) = ∇X V = q k j
+ v ẋ Γqk ∂k
dt dt
142
1. Scalars (type (0, 0)) - f (x) .
∑ i
2. Vectors (type (1, 0)) - (ui ) u ∂i .
∑
3. Covectors (type (0, 1)) - (vi ) vi dxi .
4. Inner products of vectors gij (type (0, 2)) .
5. Inner products of covectors g ij (type (2, 0)) .
6. Operators (aij ) (type (1, 1)) .
General Tensors (type (k, l)) .
Components Tji11...j ...ik
l
(x)
e1 , . . . , en - basis of tangent vectors
e1 , . . . , en - basis of covectors
ei1 ⊗ ei2 ⊗ . . . ⊗ eik ⊗ ej1 ⊗ ej2 ⊗ . . . ⊗ ejl - basis in tensor space
(k, l).
Operations : Linear, Tensor product
(k, l) ⊗ (p, q) ⊂ (k + p, l + q)
′ ′′ ′ ′′
Tj′i′ ⊗ Tj′i′′ = Tji′ ji ′′ (x)
Permutation of indices (lower or upper)
Trace
′ ′′ Trace ′ ′′
∑
n
′ ′′
Tji′ jji i′′ −−−→ Tji′ ji ′′ = Tji′ ii ji ′′
i=1
∂
∇X ⟨V , W ⟩ = ui ⟨V , W ⟩ = ⟨∇X V , W ⟩ + ⟨V , ∇X W ⟩
∂xi
143
where X = (ui ) ,
( )
∇k ⟨V , W ⟩ = ∇k gij v i wj = ⟨∇k V , W ⟩ + ⟨V , ∇k W ⟩
∇k = ∇Xk , Xk = ∂k .
For parallel Transform
d dV dW
⟨V , W ⟩ = ⟨ , W ⟩ + ⟨V , ⟩
dt dt dt
dV dW d⟨V , W ⟩
= 0, = 0 ⇒ = 0
dt dt dt
“symmetric”: ∇k ∂j = ∇j ∂k ↔ ∇X Y − ∇Y X = [X , Y ] .
Homework 10.
1. Prove that for complex polynomial
w = Pn (z) : S2 → S2
we have deg f = n .
2. Calculate degree of rational map
Pn (z)
w = , (irreducible fraction)
Qk (z)
y = a0 xn + a1 xn−1 + . . . + an : R → R , aj ∈ R
144
γ (t) f(x)
f
Ω = x dy ∧ dz − y dx ∧ dz + z dx ∧ dy
Prove that ( )
Ω
d = 0
(x + y + z 2 )3/2
2 2
Prove that
Ω|unit sphere = area element in S2
which is SO3 - invariant.
Homeworks 8, 9. Solutions.
Homework 8. Solutions.
∑
1. Prove that i pi dxi is 1-form in T∗ (N k ) (local coordinates (xi , pj )).
Change of coordinates x = x(y) ,
∂xi
(system x) pi = p′j (system y)
∂y j
So we have
∑ ∑ ∂xi ∑
p′j dy j = pi dy j
= pi dxi
j i,j
∂y j i
∑
Remark. dpi ∧ dxi is a well-defined 2-form also (done before).
145
ij
2. Prove that geodesic
∑ flowi with Hamiltonian H(x, p) = g (x) pi pj / 2
preserves
∑ the form i pi dx [correction due to student: it preserves the
form i pi dxi restricted to the (2n−1) - dimensional submanifold T1 (M n ) ,
|ẋ| = 1 (unit tangent vectors)].
Proof. Calculation shows that for the shift St along the flow we have for
geodesic flow ( )
d ∗ ∑
St pi dxi = dH
dt i t=0
For geodesic flow have
1 ij 1
H(x, p) = g (x) pi pj = gij (x) ẋi ẋj
2 2
so T1 (M n ) is given by the condition H = 1 and dH = 0 on T1 (M n ) .
Calculation: We have
d ∗ d ∗ i
S (pi ) = − Hxi , S (x ) = Hpi
dt t t=0 dt t t=0
d ∗ i
S (dx ) = dHpi = Hpi pj dpj + Hpi xj dxj
dt t t=0
Finally:
( )
d ∗ ∑ [ ]
St pi dxi = − Hxi dxi + pi Hpi pj dpj + Hpi xj dxj =
dt i t=0
( )
∂H
= d pi − H = dL
∂pi
where L is a “Lagrangian”.
For
1 ij 1
H(x, p) = g (x) pi pj = gij (x) ẋi ẋj
2 2
we have H = L and also dH = dL .
146
2 is proved.
3. Locally every closed form is exact: ω = dν in the domain U ⊂ M n .
Let U is the ball |x|2 < ϵ . Multiply ν by function φ (C ∞ , φ ≡ 1 in U
and φ ≡ 0 outside of the ball |x|2 > 2ϵ ).
Our form is ω − d(φν) .
4. Let ∫ (∑ )
S {γ} = pi (t) ẋ (t) dt − H(p, x) dt
i
,
γ i
∑
γ = {x(t), p(t)} , L = pi ẋi − H(x, p) . We have from the Euler
Lagrange equations
( )
d ∂L
∂L ∂H
ṗi = = − i =
dt ∂ ẋi
∂x i ∂x
( )
∂H ∂L d ∂L
ẋi − = = = 0
∂pi ∂pi dt ∂ ṗi
Remark. We can write
∫
( )
S {γ} = d−1 (Ω) − H dt
γ
147
Homework 9. Solutions.
1. Let M 2 = R2 \(point) . It is homotopy equivalent to S1 , so H 2 (M 2 ) =
H 2 (S1 ) = 0 .
Every open 2-manifold is homotopy equivalent to “graph” like
111111111
000000000
000000000
111111111
000000000
111111111
000000000
111111111
000000000
111111111
000000000
111111111
00000
11111
000000000
111111111
00000000
11111111
00000
11111
000000000
111111111
00000000
11111111
00000
11111
000000000
111111111
00000000
11111111
00000
11111
B2 D2
00000000
11111111
00000000
11111111
00000000
11111111
00000000
11111111
00000000
11111111 D2
00000000
11111111
00000000
11111111
1111111111
0000000000
0000000000
1111111111 strip
0000000000
1111111111
0000000000
1111111111
0000000000
1111111111
0000000000
1111111111
0000000000
1111111111
0000000000
1111111111
0000000000
1111111111
0 1111111111
0000000000
Proof.
4π 4π
Inverse
a image
deg=2
2π 2π of point (a)
a appears twice
0 2π 0 2π
2 is proved.
3. Deform one circle far: S1 (t) , 0 ≤ t ≤ 1 , S2 (0) = S2 . We have
S1 ◦ S2 (0) = S1 ◦ S2 (1) = Ø
148
S1(0) S1(1)
S2(1)
S2(0)
F (M n × I) ∩ W n−k = Ø
H0 = R , H3 = R ⊕ R , H6 = R .
7. We have
1111111111
0000000000 000000000
111111111
000000000
111111111
111111111111
000000000000 0000000000
1111111111
00000
11111
000000000
111111111
000000000000
111111111111 0000000000
1111111111
00000
11111
0000000000
1111111111 000000000
111111111
000000000000
111111111111
000000000000
111111111111 00000
11111
0000000000
1111111111 000000000
111111111
R2 \(∗1 ∪ ∗2 ) ∼ 000000000000
111111111111 ∼ 00000
11111
=1111111111
0000000000 000000000
111111111
00000
11111
= 000000000000
111111111111 0000000000
1111111111 000000000
111111111
000000000000
111111111111
000000000000
111111111111 0000000000 111111111
1111111111 000000000
2
D D 2 B2
Use arguments same as in Problem 3 above for calculation of H 2
( )
H 1 R2 \(∗1 ∪ ∗2 ) = R ⊕ R
149
Lecture 32. Geodesics. Gauss Lemma.
1. Geodesics: smooth curves γ(t) = {xi (t)} such that
D d⃗x
≡ 0 ( or ∇ẋ ẋ = 0 )
dt dt
Equation
ẍi + Γikj ẋk ẋj = 0
Calculus of variation:
a) Length
∫ b ∫ b
l(γ) = |ẋ(t)| dt = L0 (x, ẋ) dt
a a
where √
L0 (x, ẋ) = gij (x) ẋi ẋj
is a “Lagrangian”.
b) “Action”
∫ b ∫ b
1
S(γ) = gij (x) ẋi ẋj dt = L1 (x, ẋ) dt
a 2 a
Lemma 1.
D ∂s D ∂s
=
∂v ∂u ∂u ∂v
where
∂s ∂xi ∂s ∂xj
= ∂i = X , = ∂j = Y
∂u ∂u ∂v ∂v
or ∇X Y = ∇Y X at s.
Proof. Vector fields X, Y commute at the surface s, so [X, Y ] = 0 at s.
But we have
∇X Y − ∇Y X − [X , Y ] = 0
at s.
Lemma is proved.
151
Gauss Lemma. Let expP (v) be defined for v ∈ TR = Rn , and w ∈ TP .
Then we have
w
dv
w = , v(0) = v, |v(s)| = const
ds s=0
v
v(s)
For small ϵ > 0 all u = t v(s) , |s| < ϵ , have well-defined expP (u) ,
0 ≤ t ≤ 1.
Consider surface f : A → M n , parametrized by (t, s) :
∂f ∂f
⟨ , ⟩ = ⟨(d expP )v (w) , (d expP )v (v)⟩
∂s ∂t t=1,s=0
We have also
∂ ∂f ∂f D ∂f ∂f ∂f D ∂f D ∂f ∂f
⟨ , ⟩ = ⟨ , ⟩ + ⟨ , ⟩ = ⟨ , ⟩ =
∂t ∂s ∂t dt ∂s ∂t ∂s dt ∂t dt ∂s ∂t
D ∂f ∂f 1 ∂ ∂f ∂f
= ⟨ , ⟩ = ⟨ , ⟩ = 0 ,
ds ∂t ∂t 2 ∂s ∂t ∂t
∂f ∂f
so ⟨ , ⟩ is t − independent.
∂s ∂t
Calculate it for t → 0 :
We have
∂f
(t, 0) = (d expP )tv (tw)
∂t
152
and
∂f
(t, 0) → 0 , t → 0 .
∂s
∂f ∂f ∂f ∂f
So ⟨ , ⟩ ≡ 0 for all t , and ⟨ , ⟩ = 0
∂s ∂t ∂s ∂t t=1,s=0
Lemma is proved.
Theorem. Let ϵ > 0 is small enough and expP (Bϵ (0)) ⊂ M n is a geodesic
ball. Let c(t) be a smooth curve joining γ(0) and γ(1) where γ is a
geodesics started in P .
γ (t)
P
c(t)
Then length of c(t) is: l (c) ≥ l (γ) , and l (c) = l (γ) implies c = γ .
Proof. Let c ⊂ Bϵ (round geodesic ball). Curve c(t) can be written as
where v(t) is a curve in tangent space TP , and r(t) is such that 0 < r ≤ 1 .
Let c(t1 ) ̸= P for all t1 ∈ (0, 1] and r(t) > 0 for all t1 ∈ (0, 1] .
We have for
dc ∂f ′ ∂f
= r (t) +
dt ∂r ∂t
By Gauss Lemma (above) we have
∂f ∂f
⟨ , ⟩ = 0
∂r ∂t
We have also |∂f /∂r| = 1 , so
2 2
dc ′ 2 ∂f
= |r (t)| +
dt ∂t
153
and ∫ 1 ∫ 1
dc
dt ≥ |r′ (t)| dt ≥ r(1) − r(0)
ϵ dt ϵ
For ϵ → 0 we have r(0) = 0 . So we get
l (c) ≥ r(1) = l (γ)
At the same time the condition ∂f /∂t ≡ 0 means v(t) = v(0) = const ,
i.e. c = γ (geometrically for r′ (t) ≥ 0 , otherwise: l (c) > l (γ) ) .
Theorem is proved.
γ
Proof. Move f (x) to g(x) monotonically along g(x)
unique minimal geodesics joining these points
f(x)
Corollary is proved.
Corollary 2. Let M n be compact (or metrically complete) Riemannian
manifold. For every 2 points P, Q with distance ρ (P, Q) = d there exists
geodesics γ , l(γ) = d , joining P and Q .
Proof. Let us remind that d(P, Q) = minγ̃ l (γ̃) , γ̃ joins P and Q and
γ̃ is smooth or piecewise smooth curve:
γ
Q
P
Let us remind you the “Arcellát Principle” claiming that the set of piece-
wise smooth curves of length ≤ Γ is “precompact”. It means in particular
that every “Cauchy Sequence” of curves for which length is well-defined.
154
Consider sequence of piecewise smooth curves γn such that l (γn ) → d ,
d = ρ (P, Q) . Limit curve γ∞ is such that l (γ∞ ) = d , but it can be not
smooth (even not piecewise smooth).
Take 2 points P1 , P2 at γ∞ such that ρ (P1 , P2 ) < ϵ (small enough).
Q
P P1 P2 γ (t)
X = (ui ) , Y = (v j ) , Z = (z k ) , W = (wl ) .
( )
R (X, Y ) = ∇Y ∇X − ∇X ∇Y − ∇[X,Y ]
- curvature operator.
Index notations
( )
Rkl ji
= (∇j ∇i − ∇i ∇j ) = R̂ji = − R̂ij = (Rji )jk
(Here [∂i , ∂j ] = 0 , X = ∂i , Y = ∂j , Z = ∂k , W = ∂l ).
R (X, Y ) − linear 0-order operator on tangent space (matrix).
155
operator
k
R (X, Y ) = Rl,ij ui v j = R̂X,Y
vector
k
R (X, Y ) Z = Rl,ij ui v j z l
number
⟨R (X, Y ) Z , W ⟩ = Rl,ij
k
ui v j z l wp gkp
l
Formulas for symmetric connection were given, and for Rk,ij through Γkij .
(Especially for coordinate system such that gij (P ) = δij , ∂k Γij |P = 0).
p
Homework 11.
1. Compact Lie groups, bi-invariant Riemannian metric g → h1 g h−1
2 .
At ∼ 3 ∼
Prove that e are geodesics (for SO3 = RP and SU2 = S ). A is
3
156
Lecture 34. Riemannian Curvature. Exam-
ples.
Consider M n , ψ ∈ Rm
x .
Curvature tensor.
Connection: ∇X , X = ∂i , ∇X ψ = ∂i ψ + Γ̂i ψ
Curvature:
∇X ∇Y − ∇Y ∇X − ∇[X,Y ] = R̂
X = ∂i , Y = ∂j , R̂ij = ∇i ∇j − ∇j ∇i : Rm
x → Rx
m
A = Γ̂i dxi
(Matrix Multiplication).
1
R̂ = dA + [A ∧ A]
2
- same; written through the commutator - Lie Algebra multiplica-
tion.
Gauge Transformation: ψ = G(x) φ(x)
157
Symmetric connection compatible with Riemannian metric (tangent bun-
dle)
∇i ⟨X , Y ⟩ = ⟨∇i X , Y ⟩ + ⟨X , ∇i Y ⟩
Corollary: ( )
1 ks ∂gij ∂gis ∂gsj
Γkij = g − s + j
+
2 ∂x ∂x ∂xi
Geodesic Equation
2) ∇ẋ ẋ = 0 .
Calculation of Curvature: Choose coordinate system such that
in such point
∂ s ∂ s
s
Rijk = Γik − Γ
∂xj ∂xi jk
s
Symmetries of Curvature: Let Rij,kl = gis Rj,kl then
Bianchi:
Rij,ks + Rjk,is + Rki,js = 0
R (X, Y, X, Y )
K (X, Y ) =
Area (X, Y )
158
X, Y - vectors,
√
Area (X, Y ) = |X ∧ Y | = |X|2 |Y |2 − ⟨X, Y ⟩2 =
Y
√ ( ) 111111111111111111111
000000000000000000000
000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
⟨X, X⟩ ⟨X, Y ⟩ 000000000000000000000
111111111111111111111
= det 000000000000000000000
111111111111111111111
⟨X, Y ⟩ ⟨Y, Y ⟩ 000000000000000000000
111111111111111111111
000000000000000000000
111111111111111111111
X
X Y
X ∧ Y ∈ Λ2 TP = Λ2 Rn
R (X ∧ Y , X ∧ Y ) - quadratic form on Λ2 Rn .
K > 0 - “Positive Curvature” (like Sn )
K < 0 - “Negative Curvature” (like Hn )
Ricci Curvature: (see D.C.)
s
Rji = Rij = Ri,sj
1
Ric = Rij
n−1
Notation (D.C.)
1 ∑
RicP (X) = ⟨R (X, Zi ) X , Zi ⟩
n−1 i
2 - dimensional Manifolds
159
R = R12,12 - the only nonzero component.
K<0
K>0
R = Gaussian Curvature M 2 ⊂ R3 .
3 - dimensional Manifolds
Rij determines all Curvature Tensor.
4 - dimensional Manifolds
Einstein Equation:
1
Rij − R gij = λ Tij + µ gij
2
where the term λ Tij represents “Matter” and µ gij is the “cosmological
term” (or “dark energy”).
Lie Groups and “symmetric” spaces (“constant curvature tensor”)
( i )
∇s Rj,kl ≡ 0
zn = 1
160
(Fraction is irreducible).
Take point w0 such that roots are distinct. We have deg f = max (deg P, deg Q).
3. Real polynomial f = a0 xn + . . . + an , a0 ̸= 0 . Let n = 2k + 1 .
Then:
a0 >0 a0 <0
a0 > 0 : deg f = 1 ,
a0 − 0 : deg f = −1 .
n
dτ = k(s) ds = dφ - obvious.
τ
5. 2-form
Ω = x dy ∧ dz − y dx ∧ dz + z dx ∧ dy
1. Rotate in (y, z) - plane:
x → x , dy ∧ dz → dy ∧ dz
2. Rotate in (x, y) - plane:
z → z , dx ∧ dy → dx ∧ dy
3. Rotate in (x, z) - plane:
y → y , dx ∧ dz → dx ∧ dz
161
Similar arguments we have for other relations.
So our form is invariant under all 3 rotations. Therefore it is invariant
under the whole SO3 .
Right shift maps right invariant metric into itself, and left shift maps
right invariant metric into another right invariant metric g̃ij .
For h2 = h−11 metric g̃ij can be obtained from gij at the Lie Alge-
bra by conjugation. So g̃ij = gij for inner product ⟨A, A⟩ invariant under
conjugation. For SOn it is ⟨A, A⟩ = − Tr A2 , At = − A .
162
3. Geodesics in SUn are eAt (if it starts at I for t = 0). So we need:
eAT = −I for some T . We have the following basis for su2 :
( ) ( ) ( )
i 0 0 −1 0 i
, ,
0 −i 1 0 i 0
Let us put ( )
ia 0
A = , a ∈ R
0 −ia
Fix time moment T such that eiaT = − 1 . We have then eAT = − I .
All other solutions can be obtained by the change of basis in C2 (same T ) :
( ) ( )
i 0 i 0
→ g g −1 ,
0 −i 0 −i
0 1 ... 0 0
1 0 ... 0 0
.. .. . . ....
M 2 = S2g : . . . . . over Z2
0 0 ... 0 1
0 0 ... 1 0
1 0 ... 0
0 1 ... 0
M 2 = RP2 # . . . #RP2 : .. .. . . .. k times
. . . .
0 0 ... 1
( )
2 2 0 1
M = K :
1 1
163
K 2 # RP2 = T2 # RP2 = RP2 # RP2 # RP2 (?)
164