1

HOMOTOPY
ABSTRACT :- Two mathematical objects are said to be homotopic if one
can be continuously deformed into the other. The concept of homotopy was
first formulated by Poincare in Analysis Situs (1895), set forth the
fundamentals of homology. In this talk I would like to introduce the concept
of homotopy and then try to explore how it is used in solving one of the
major problem of point set topology namely when the two topological
spaces are homeomorphic to each other.

Definition : (Homotopic maps)

Let f and g be continuous functions from a space X into a space

Y , we say that f is homotopic to g , written as f ≃ g , iff there is a
continuous function H : X × I  Y such that

H( x , 0 ) = f(x) and H( x , 1 ) = g(x)

Where I = [0, 1] , is the unit closed subset of R, the set of real numbers.
The map H is called a homotopy between f and g. We also write H: f ≃ g,
when H is a homotopy between f and g .
This can be shown as below: - .

In other words we have :

Example 1 : Let X = Y = Rn and let f(x) = x and g(x) = 0 for x ε Rn .

 2
Let H : R × I  R defined as H( x , t ) = ( 1-t )f(x) .
n n

Then H is a homotopy between f and g .

Note: There can be more than one homotopy between the two functions f
and g . For example H1 : Rn × I  Rn defined by H1( x , t ) = ( 1-t2 )f(x) is
also a homotopy between f and g .

Example 2 : Let f , g : X R2 be any maps . Then f and g are homotopic

to each other and the Homotopy map H : X × I  R2 , between them is
given by H( x , t ) = ( 1-t )f(x) + t g(x)
such that H(x, 0) = f (x) and H(x, 1) = g (x)
Hence f ≃ g.

Definition : (Null-homotopy map )

A continuous map f : XY is said to be null- homotopic if it is
homotopic to some constant map.

Remark : Null- homotopic maps may not be homotopic . In fact, constant

maps need not be homotopic.

For example: Let X be connected and Y be not connected .Let y1 and y2

be points in distinct components of Y and let f(x) = y1 and g(x) = y2 for all
x ε X . Then f and g are not homotopic because X × I is connected while
Y is not connected and the continuous image of a connected space is
connected .
In certain situation, a restricted type of homotopy is considered.
Under this restriction some subset is required to remain fixed throughout the
deformation.
Let us consider the two simple arcs f(x) and g(x) with the same end points x1
and x2 . Suppose f(x) is deformed continuously into g(x) by the process
described above , but with the condition that every member of the family of
intermediate arcs has x1 and x2 as end points . Under this situation we say
that the mapping defining f(x) and g(x) are homotopic relative to the subset
consisting of x1 and x2 .
 3

We give the formal definition.

Definition: ( Relative Homotopy) Let A ⊆ X and f : X  Y , g : X  Y
be continuous mapping . f and g are said to be homotopic relative to A if
there is a homotopy H : X × I  Y between f and g such that H(x,t) is
independent of t for x ε A . In other words H(x,t) = f(x) for all x ε A , t ε I.
In this situation it is clear that f(x) = g(x) for all x ε A. The homotopy H is
called a homotopy relative to A and is written as f ≃ g(rel A).
If f and g are two paths in X, there is a stronger relation between them
than mere homotopy. It is defined as follows:
Definition. ( Path Homotopy ) Two paths f and g, mapping the interval
I = [0, 1] into X, are said to be path homotopic if they have the same initial
point x0 and the same final point x1 , and if there is a continuous map
H : I I X such that
H(s, 0) = f(s) and H(s, 1) = g(s),
H(0, t) = x0 and H(1, t) = x1 , for each s I and each t I.
We called H is a path homotopy between f and g.
If f is a path homotopic to g , we write f ≃p g.

Remark. The first condition says that H is a homotopy between f and g and
the second condition says that for each t, the path f t defined by the equation
 4
f t (s ) = H(s, t) is a path from x0 and x1 . In other words, the first condition
says that H represents a continuous way of deforming the path f to the path
g, and the second condition says that the ends points of the path remain fixed
during the deformation.
Theorem 1. The relation ≃ is an equivalence relation in the set C(X, Y) of
all continuous maps from X to Y.
Proof. If f C(X, Y), then H : f ≃ f, where H is defined by
H(x, t) = f (x) for all x X and t I = [0, 1]
And if f, g C(X, Y) and H1 : f ≃ g, then H2 : g ≃ f,
where H2(x, t) = H1(x, 1 – t) for all x X and t I
Further, if f, g, h C(X, Y) and H1 : f ≃ g and H2 : g ≃ h, then
H3 : f ≃ h, where
 1
H 1 ( x , 2 t ) 0≤t ≤
2
H3(x, t) =  1
H 2 ( x, 2 t − 1) ≤ t ≤1
 2
is continuous on X I, since it is continuous on each of the closed
 1 1 
subsets X 0, 2 
 
and X  2 , 1 ,
 
the result follows by pasting lemma.
Hence ≃ is an equivalence relation.
Theorem 2. The relation ≃p is an equivalence relation in the set C(I, X)
of all paths in X having the same initial and terminal points.
Proof. In the above theorem replace the cont. maps f, g, h by paths in X.
Theorem 3. If h, h : X Y are homotopic and k, k : Y Z are
homotopic, then k o h and k o h are homotopic.
In other words composites of homotopic maps are homotopic.
Proof. Since h, h : X Y are homotopic from X to Y
 a continuous map H1 : X I Y such that
H1(x, 0) = h(x) , H1(x, 1) = h (x) ,  x X
Also k, k : Y Z are homotopic from Y to Z
 a continuous map H2 : Y I Z such that
H2(y, 0) = k(y) , H2(y, 1) = k (y), y Y
Clearly, k o H1 : X I Z is a continuous map such that
(k o H1) (x, 0) = k(H1(x, 0)) = k(h(x)) = (k o h) (x)
and (k o H1) (x, 1) = k(H1(x, 1)) = k(h (x)) = (k o h ) (x)
k o H1 is the homotopy between k o h and k o h
i.e. k o H1 : k o h ≃ k o h
 5
Now, By transitivity of the homotopy relation, it remain to construct a
homotopy between k o h and k o h .
Define H : X I Z by
H(x, t) = H2(h (x), t) ,  x X, t I.
Then H is a composite of continuous functions and hence continuous
and
H(x, 0) = H2(h (x), 0) = k(h (x)) = (k o h ) (x)
H(x, 1) = H2(h (x), 1) = k (h (x)) = (k o h ) (x)
H is a homotopy between k o h and k o h
i.e. H:koh ≃k oh
Thus k o H1 : k o h ≃ k o h and H : k o h ≃ k o h
By transitivity of the homotopy relation
koh≃k oh
i.e. k o h and k o h are homotopic.
Definition (Homotopy and path homotopy equivalence classes)
The equivalence classes in C(X, Y) under the relation ≃ are called the
homotopy equivalence classes in C(X, Y).
Similarly, the equivalence class in C(I, X) under the relation ≃p are
called the path homotopy equivalence classes.
Note : If f is a path in X, then its path-homotopy equivalence class is
denoted by [ f ]
Thus [ f ] = { g : [0, 1] X such that g ≃p f }

FUNDAMENTAL GROUP
Now we introduce some algebra into this geometric situation. We define a
certain operation on path-homotopy classes as follow:-

Definition : If f is a path in X from x0 to x1 , and g is a path in X from x1

to x2 . we define the product f * g of f and g to be the path h given by

  1
 f (2 s ) for s ∈ 0, 
 2
h( s ) = 
 g (2 s − 1) for s ∈  1 , 1
  2 

The function h is well-defined and continuous, by pasting lemma; it is a

path in X from x0 to x2. Thus h is a path whose first half is the path f and
whose second half is the path g .
 6
The product operation on paths induces a well-defined operation on
path-homotopy classes, defined by the equation
[ f ] * [ g] = [ f * g]
for, let H1 be a path homotopy between f and f and let H2 be a path
homotopy between g and g . Then
  1
H1 (2 s, t ) for s ∈ 0, 
 2
H(s, t) = 
H 2 (2 s − 1, t ) for s ∈  1 , 1
  2 

is the required path homotopy between f * g and f * g as

H1(1, t) = x1 = H2(0, t) for all t and the map H : I I X is continuous
by pasting lemma.
Remark : (i) The product of two path-homotopy classes is not always
defined.
(ii) The product of two path-homotopy classes [ f ] and [ g] is defined
only when f(1) = g(0).
Definition. ( Loop ) Let X be a topological space, x0 a fixed point in X. A
continuous function f : I X will be called a loop based at x0 iff
f (0) = f (1) = x0 . i.e. a path that begin and end at x0 is called a loop at x0 .

The set of all path homotopy classes of loop based at x0 is denoted by

1(X, x0 ) and it form a group (under the operation *) known as Fundamental
group of X relative to the base point x0 .
Note : We denote the relation of homotopy between two loops f and g by
f ≃ x g and the equivalence class containing f will be denoted by [ f ].
0

Theorem 4. 1(X, x0 ) with the operation * , is a group.

Proof. Now 1(X, x0 ) = Set of path homotopy classes of loop based at x0 .

Let [ f1 ] and [ f 2 ] be two members of 1(X, x0 ) .
Let f1 ≃ x0 g1 and f 2 ≃ x0 g 2  continuous map H1 , H2 : X I X
s.t. H1(s, 0) = f1 ( s ) , H1(s, 1) = g1 ( s )
H1(0, t) = x0 = H1(1, t) ,  s, t I.
 7
and H2(s, 0) = f 2 ( s ) , H2(s, 1) = g 2 ( s )
H2(0, t) = x0 = H2(1, t) ,  s, t I.
  1
H1 (2 s, t ) for s ∈ 0, 
 2
Then H(s, t) = 
H 2 (2 s − 1, t ) for s ∈  1 , 1
 2 

is the path homotopy between f1 * f 2 and g1 * g 2 such that
H(0, t) = H1(0, t) = x0 and H(1, t) = H2(1, t) = x0
i.e. H(0, t) = x0 = H(1, t)
and so f1 * f 2 ≃ x g1 * g 2
0

Thus [ f1 ] * [ f 2 ] = [ f1 * f 2 ]
* is a binary operation on 1(X, x0 )
For the associatively, it is sufficient to show that ( f * g ) * h ≃ x f * ( g * h)
0

for loops f , g and h based at x0 .

A pictorial approach will make the idea behind the necessary homotopy
easy. In terms of its action on I, (f * g) * h is accomplished by completing
 1 1 1
the action of f in the interval 0, 4  ,
 
the action of g in the interval 4 ,
 2 
and
1 
the action of h in the interval  2 , 1 .
 
This is represented on the top line of the
fig. given below.

vv
The bottom line represents f * ( g * h) . A homotopy between f * ( g * h)
and ( f * g ) * h can then be constructed by allowing the action of f, g and h to
be divided at time t as shown.
For identity, let e x denote the constant loop e x 0 (t ) = x0 for all t I. We claim
0

that [ e x ] is an identity of 1(X, x0). For this it is sufficient to show that

0

f * e xo ≃ x0 f and e xo * f ≃ x0 f for all f 1(X, x0)

To exhibit a homotopy for the first, define for each t I

 8
  2x  2−t
 f   0 ≤ x ≤
H(x, t) =   2 − t  2
x 2−t
≤ x ≤1
 0 2
H is continuous on I I, since it is continuous on each of the closed sets
 2 −t  2 −t
( x, t ) : x ≤  and ( x, t ) : x ≥  the result follows by pasting lemma.
 2   2 
Moveover , H(x, 0) = f(x) and H(x, 1) = ( f * e x ) (x) for all x
0
I

The other relation e xo * f ≃ x0 f can be done similarly.

For the inverse : For each loop f and x define f to be the loop
f ( x ) = f (1 − x ) , 0 ≤ x ≤ 1
and let [ f ]=[ f ]
It is now easy to check that this is well defined. To show that [ f ] is an
inverse of [ f ], it is sufficient to show that

f*f x0 e xo fe * f x0 e xo

≃ and ≃
 1− t
 f ( x) 0≤ x≤
2
 1− t
First, let H(x, t) =  f ( x + t ) ≤ x ≤1− t
 2
 x0 1− t ≤ x ≤1

The function H is continuous on each of the three closed sets which
cover the square and thus continuous.
Clearly H(x, 0) = ( f * f ) ( x) and H(x, 1) = e xo (x) for all x I.
The homotopy showing that f * f ≃ x e xo is similarly constructed.
0
 9
Hence 1(X, x0) with the operation * is a group.
Remark.(1) The fundamental group 1(X, x0) is also called the first
homotopy group of X. There are indeed groups n(X, x0) for all n Z+, but
we shall not study this here. They are part of the general subject called
homotopy theory.
(2) From the structure of 1(X, x0) we may obtain local character of the
space at x0 . For example : if 1(X, x0) is the trivial group i.e. the group
whose only element is the identity e x , then every path at x0 is homotopic to
0

e x and intuitively this mean , for example , that there is no “holes” which
0

prevent a path at x0 from shrinking to the point x0 .

ISOMORPHISM OF FUNDAMENTAL GROUPS

Now, an immediate question arises to what extent the fundamental group
depends on the base point. In this direction we show that if x0 , x1 ε X and
there is a path from x0 to x1, then the groups 1(X, x0) ≅ 1(X, x1) .
For this we first define the following:

Definition. Let be a path in X from x0 to x1.

Define a map
αˆ : π 1 ( X, x0 ) → π 1 ( X, x1 ) by αˆ ([ f ]) = [α ] * [ f ] * [α ]
The map α̂ , known as ‘‘ -hat’’, is well-defined, for the operation * is
well-defined. If f is a loop based at x0 , then α * ( f * α ) is a loop based at x1 .
Note : The map α̂ depends only on the path-homotopy class of .
Theorem 5. If x0 , x1 ε X and be a path in X from x0 to x1. Then
the map α̂ : 1(X, x0) 1(X, x1) is a group isomorphism.

Proof. Let [ f ] and [ g] be any two elements from 1(X, x0).

To show that α̂ is a group isomorphism, we need to show

(i) αˆ ([ f ] * [ g ]) = αˆ ([ f ]) * αˆ ([ g ]) and
(ii) α̂ is one-one and onto.
Now αˆ ([ f ]) * αˆ ([ g ]) = ([α ] * [ f ] * [α]) * ([α ] * [ g ] * [α])
= [α ] * [ f ] * ([α ] * [α ]) * [ g ] * [α ]
 10
= [α ] * [ f ] * (ex ) * [ g ] * [α ]
0

= [α ] * [ f ] * [ g ] * [α ]
= αˆ ([ f ] * [ g ])
For second, we show that if denotes the path α , which is the reverse
of , then β̂ is an inverse of α̂ .
We compute for each element [h] of (X, x1),
1

βˆ ([h]) = [ β ] * [h] * [ β ] = [α ] * [h] * [α ] , then

αˆ ( βˆ [h]) = αˆ ([α ] * [h] * [α ])
= [α ] * ([α] * [h] * [α ]) * [α]
= ([α ] * [α]) * [h] * ([α ] * [α])
= e x * [ h] * ex
1 1

= [h]
Similarly, we can show that
βˆ (αˆ ([ f ])) = [ f ] , for each [ f ] 1 (X, x0)
⇒ β̂ is the inverse of α̂ and vice a versa and so α̂ is one – one and onto.
Hence α̂ is an isomorphism.
Corl. If X is an arcwise connected space (or path connected space) and
x0 and x1 are two points of X, then 1(X, x0) is isomorphic to 1(X, x1).
Note : (1) From above , we notice that in case of arcwise connected space
(or path connected space) X, the fundamental group of X do-not depends
upon the base point. Thus in this case we can speak of fundamental group
1(X) of X. This will cause us no difficulty here but create problem over
simplification in a deeper study of fundamental group.
(2) The above corollary is not true if the condition of space being path
connected is omitted. Even if X is connected , the corollary need not be true.
Theorem 6. Let be a path in X from x0 to x1, let be a path in X
from x1 to x2 and if = * , then γˆ = βˆ o αˆ i.e. α * β = βˆ o αˆ .
Proof. Since be a path in X from x0 to x1 and is a path in X
from x1 to x2 and = * . Therefore is a path in X from x0 to x2 .
To show that γˆ = βˆ o αˆ .
Let [ f ] 1 (X, x0) be any element.
( βˆ o αˆ ) [ f ] = βˆ (αˆ ([ f ]))
= βˆ ([α ] * [ f ] * [α ])
= [ β ] * [α ] * [ f ] * [α ] * [ β ]
 11
= [ β * α ] * [ f ] * [α * β ]
= [ α * β ] * [ f ] * [α * β ]
= [γ ] * [ f ] * [γ ]
= γˆ ([ f ])
βˆ o αˆ = γˆ.
As we know that the map α̂ depends only on the path-homotopy class of .
Now the question arises “ is there any situation where the α̂ is independent
of . To this direction we have the following theorem:
Theorem 7. If be a path in X from x0 to x1 and if
α̂ : 1(X, x0) 1(X, x1) , be the isomorphism determined by the path .
Then α̂ is independent of if and only if 1(X, x0) is abelian group .
Proof. Firstly , let 1(X, x0) be an abelian group
Let and be any paths from x0 and x1

Since α * β is a loop based at x0.

∴ for each [ f ] ε 1(X, x0) , we have
[ α * β ] *[ f ] = [ f ] *[ α * β ]
⇒ [ α*β * f ]=[f * α*β ]
Pre Multiplying by [α ] and post multiplying by [ β ] , we get
[α ] * ([ α * β * f ]) * [ β ] = [α ] * ([ f * α * β ] )* [ β ]
⇒ [α * α * β * f * β ] = [α * f * α * β * β ]
⇒ [ β * f * β ] = [α * f * α ]
⇒ β̂ ( [ f ] ) = α̂ ( [ f ] ) , ∀ [ f ] ε 1(X, x0)
⇒ β̂ = α̂
Conversely , let α̂ = β̂ . To show that 1(X, x0) is an abelian group
Let [ f ] , [g ] ε 1(X, x0) be any two elements
Since f is a loop based at x0 and β is a path from x0 to x1
 12

∴ f * β is also a path from x0 to x1 .

Thus α and f * β are two paths from x0 to x1

α̂ = f * β

∴ α̂ ( [g ] ) = f * β ( [g ] )

( [α * g * α ] ) = [ f * β * g * f * β ]

( [α * g * α ] ) = [ β * f * g * f * β ]
Pre Multiplying by [α ] and post multiplying by [α] , we get
[α ] *( [α * g * α ]) * [α
] = [α ] *( [ β * f * g * f * β ] )* [α ]

[α * α * g * α * α ] = [α * β * ( f * g * f ) * β * α ]
[g ] = [e x * f * g * f * e x 0 ]
0

[g ] = [ f * g * f ]
Pre Multiplying by [ f ] , we get
[ f ]*[ g] = [ f ] * [ f * g * f ]
[ f ]*[ g] = [ f * f * g * f ] = [g * f ] = [g ] *[ f ]
Hence 1(X, x0) be an abelian group

Remark : From the above theorem we conclude that if x0 and x1 be points

of the path connected space X . Then 1(X, x0) be an abelian group iff for
each pair α and β of paths from x0 to x1 , we have α̂ = β̂ .

Definition.(Simply connected space) A space X is said to be simply

connected if it is a path connected space and 1(X, x0) is the trivial (one
element) group for some x0 X, and hence for every x0 X.
The trivial group is expressed by writing 1(X, x0) = 0.
Theorem 8. Any two paths, in a simply connected space X, having the
same initial and final points are path homotopic.
 13
Proof. Let and be two paths from x0 and x1.

Then * β is a loop in X based at x0.

Since X is simply connected, this loop is path homotopic to the constant
loop ex at x0 . i.e. α * β ≃p e x
0 0

⇒ [α * β ] = [ex0 ] and so
[α * β ] * [ β ] = [e x0 ] * [ β ] = [e x0 * β ] = [ β ]

Also [α * β ] * [ β ] = [α * β * β ] = [α * ex0 ] = [α ]
Thus , we have [ ] = [ ] ⇒ α ≃p β .
The two paths, having the same initial and final points in a simply
connected space are path homotopic.

HOMOMORPHISM OF FUNDAMENTAL GROUPS

If X and Y two topological spaces, we now investigate how the fundamental
groups of X and Y are affected by a continuous transformation from X into
Y . In this direction we show that if f : X Y be any continuous , then the
fundamental group in X may not be isomorphic to its transformed form in Y.
However, there is a homomorphism between the groups as the following
theorem show:
Remark. Suppose that h : X Y is a continuous map that carries the point x0
of X to the point y0 of Y. We write this fact by
h : (X, x0) (Y, y0)
If f is a loop in X based at x0, then the composite map h o f : I Y is a
loop based at y0 .
 14
Thus the correspondence f h o f, gives rise to a map carrying
1(X, x0) into 1(Y y0). We define it as follows

Definition. Let h : (X, x0) (Y, y0) be a continuous map that carries the
point x0 of X to the point y0 of Y.
Define h* : 1(X, x0) 1(Y, y0) by
h* ([ f ]) = [h o f ]
The map h* is called the homomorphism induced h, relative to the base
point x0.
Remark :The map h* is well-defined, for if H: f ≃p f′ be the paths
homotopy between f and f′ , where f and f′ are paths at x0 . Let g and g′
be paths at y0 defined by g = h o f and g′ = h o f .
The we show that h o H : g ≃p g′ .
As both h : X Y and H : I×I X are continuous and so is the
composite map h o H : I×I Y is also continuous .
Moreover , h o H( x, 0) = h(H( x, 0)) = h( f (x)) = h o f (x) = g(x) 0.
and h o H( x, 1) = h(H( x, 1)) = h( f′ (x)) = h o f ′(x) = g′ (x)
h o H( 0, t) = h(H( 0, t)) = h( x0)) = y0
h o H( 1, t) = h(H( 1, t)) = h( x0)) = y0
Thus , we have h o H( x, 0) = g(x) , h o H( x, 1) = g′(x) and
h o H( 0, t) = h o H( 1, t) = y0 .
and so h o H : g ≃p g′ .
Also h* is a homomorphism , for let [ f ] , [ g ] ε 1(X, x0) . Then
h* ([ f ] * [ g ]) = [h o ( f * g )] =[ (h o f) * (h o g)]
= [h o f ] *[ h o g]
= h* ([ f ]) * h* ([ g ])
Note : The homomorphism h* depends not only on the map h : X Y but
also on the choice of base point x0 (once x0 is chosen, y0 is determined by
h). To avoid notational difficulty we use the notation.
(hx0 )* : π 1 ( X, x0 ) → π 1 (Y, y0 )
Theorem 9. If h : (X, x0) (Y, y0) and k : (Y, y0) (Z, z0) are
continuous, then (k o h)* = k* o h* .
If i : (X, x0) (X, x0) is the identity map, then i* is the identity
homomorphism.
Proof. Since h : (X, x0) (Y, y0) and k : (Y, y0) (Z, z0) are continuous
∴ k o h : (X, x0) (Z, z0) is also continuous . Then by definition
 15
(k o h)* ([ f ]) = [(k o h) o f ]
Also, (k* o h* ) ([ f ]) = k* (h* ([ f ]) = k* ([h o f ])
= [k o (h o f )]
= [(k o h) o f ]
Thus (k o h)* ([ f ]) = (k* o h* ) ([ f ]) ,  [ f ] 1(X, x0)
(k o h)* = k* o h*
Moreover, i* ([ f ]) = [i o f ] = [ f ]
i* ([ f ] * [ g ]) = i* ([ f * g ]) = [ f * g ] = [ f ] * [ g ]
= i* ([ f ]) * i* ([ g ])
Hence i* is the identity homomorphism.
Corl. If h : (X, x0) (Y, y0) is a homeomorphism of X with Y, then
h* : 1(X, x0) 1(Y, y0) is an isomorphism.

Proof. Let i, j be the identity map of (X, x0) and (Y, y0) respectively. Then
i* , j* be their respective identity homomorphism of the groups 1(X, x0) and
1(Y, y0). Let k : (Y, y0) (X, x0) be the inverse of h : (X, x0) (Y, y0).
Then, we have
k* o h* = ( k o h)* = i* and
h* o k* = ( h o k )* = j*
This imply, that k* is the inverse of h* .
h* is one-one and onto.
For homomorphism, let [ f ], [ g] 1(X, x0) be two elements

Then h* ([ f ] * [ g ]) = h* ([ f * g ])
= [h o (f * g)]
= [(h o f) * (h o g)]
= [h o f] * [h o g]
= h* ([ f ]) * h* ([ g ])
Hence h* is an isomorphism.
Corl. If i : (X, x0) (Y, y0) be the identity map and
j : (Y, y0) (X, x0) be the identity map, then
(i o j )* = i* o j* , is the identity on 1(X, x0)
and ( j o i )* = j* o i* , is the identity on 1(Y, y0)
Definition. (Homotopically equivalent spaces )
 16
Two topological spaces X and Y are homotopically
equivalent or of same homotopy type iff there exist continuous mapping h
: X Y and k : Y X such that h o k ≃ iY and k o h ≃ iX

We observe that two homeomorphic spaces are of the same homotopy type
For, suppose X and Y are homeomorphic and let h :X Y be a
-1
homeomorphism . Then the inverse map h : Y  X is continuous and
satisfies the condition that h-1 o h ≃ i X and h o h-1 ≃ iY . This mean h is a
homotopy equivalence, i.e. X and Y are of the same homotopy type .
The converse is not true.
For example:-Consider the unit disk Ď2 (open or closed) and a point x0 εĎ2.
Let i : P = { x0 } Ď2 be the inclusion map and e x : Ď2  P be a constant
0

map . Then clearly e x o i = IP . On the other hand , the map H : Ď2× I  Ď2

0

defined by H (x , t) = (1-t) x + t x0 , is a homotopy from I Ď2 to i o e x .

0

Thus, Ď2 is of the same homotopy type as a point P and these are clearly not
2
homeomorphic . More-over 1(Ď , x0) = 1(P, x0) =0.

Theorem 10. If (X, x0) and (Y, y0) are homotopically equivalent, then
1(X, x0) and 1(Y, y0) are isomorphic.

Proof. Since (X, x0) and (Y, y0) are homotopically equivalent
 mapping h : (X, x0) (Y, y0) and k : (Y, y0) (X, x0) such that
h o k ≃ iY and k o h ≃ i X
h o k is homotopic to the identity on Y and k o h is homotopic to the
identity on X.
Now h* : 1(X, x0) 1(Y, y0) and k* : 1(Y, y0) 1(X, x0) are
homomorphism Also by above corl., we have
h* o k* = ( h o k )* is the identity on 1(X, x0)
and k * o h* = ( k o h) * is the identity on 1(Y, y0)
Thus h* and k * are isomorphism and hence proved the theorem.
Remark : From the above example and theorem we see that if the
fundamental groups of two spaces are isomorphic , we can not say that the
two spaces are homeomophic to each other , however when the two spaces
are homeomorphic to each other then their fundamental groups are
isomorphic . This gives us a very important tool to say that two topological
spaces are not homeomorphic to each other.
 17
Theorem 11. Let X and Y be two spaces with base x0 ε X and y0 ε Y,
respectively . Then
π1( X × Y , ( x0 , y0 ) ) ≅ π1( X, x0 ) × π1( Y, y0 )
i.e. the fundamental group of the product space is isomorphic to the product
of the fundamental groups of factors.
Proof . Suppose α : I  X × Y is a loop based at ( x0 , y0 ) and let
p1 : X × Y  X , p2 : X × Y  Y are the projection mapping . Then
p1 , p2 induces group homomorphism
p1* : π1( X × Y , ( x0 , y0 ) ) π1( X, x0 ) and

p 2 * : π1( X × Y , ( x0 , y0 ) ) π1( Y, y0 ) given by

p1* ( [α ] ) = [p1  α ] and p 2 * ( [α ] ) = [p 2  α ]

∴ by the definition of the direct product of groups , we get a group

homomorphism

ρ : π1( X × Y , ( x0 , y0 ) ) ≅ π1( X, x0 ) × π1( Y, y0 )

defined by ρ( [α ] ) = ( [p1  α ] , [p 2  α ] ).
Now it is sufficient to show that ρ is a bijection .
For this , first we note that given a pair of continuous maps f : I  X ,
g: I  Y , there is , by definition of product topology , a continuous map
( f , g ) : I  X × Y defined by ( f , g )(t) = ( f (t) , g(t) ) , ∀ t ε I
conversely , any continuous map h : I  X × Y defines a pair of continuous
maps p1  α : I  X and p 2  α : I  Y . Moreover there is one-to –one
correspondence between the set of all continuous maps h : I  X × Y and
the set of pairs ( p1  α , p 2  α ) of continuous maps .
Now , let ( [α ] , [ β ] ) ε π1( X, x0 ) × π1( Y, y0 ) . Then the pair (α , β) of
loops based at x0 and y0 respectively corresponds to the loops
(α , β) : t (α(t) , β(t)) in X × Y based at (x0 , y0 ) and defines a map

η : π1( X, x0 ) × π1( Y, y0 )  π1( X × Y , ( x0 , y0 ) ) given by

η ( [α ] , [ β ] ) = ( [α , β] ) .
To see that η is well defined , note that if H1: α1 ≃ α2 ; H2: β1 ≃ β2
 18
are homotopies , then the map H : I × I  X × Y defined by
H( s , t ) = ( H1(s,t) , H2(s,t) )
is a homotopy from (α1 , β1) and (α2 , β2) relative to the set { 0 , 1} .
Clearly , the map ρ and η are inverse of each other and so ρ is an
isomorphism . This complete the proof of the theorem .
Note : Let { Xα : α ε I } be a family of spaces and for each α ε I , xα ε Xα
be a base point . Then the above theorem can be generalized as
π1( ∏ Xα , xα) ≅ ∏ π1( Xα, xα )

FUNDAMENTAL GROUPS OF SOME SPACES

In this section, we prove a few results which would be of help in computing

the fundamental groups of some familiar subspaces of the Euclidean space
ℝn , etc. Most importantly we will show that the fundamental group of the
unit circle S1 , π1(S1 ) is isomorphic to the additive group ℤ of integers and
that of a torus T ( a surface ), π1( T ) is isomorphic to ℤ × ℤ .
Example 1 . Let Ďn ( n ≥ 1) be the n-dimensional unit disk in ℝn .
Since Ďn is path connected , we can compute its fundamental group at any
point . For simplicity , let us choose origin as its base point . Then the
homotopy H: Ďn × I  Ďn defined by
H(x,t) = (1-t)x
n
Shrinks every point of Ď to the origin without moving the origin at all.
Therefore, each loop in Ďn based at origin is path homotopic to the constant
loop. Whence π1( Ďn ) =0.
More generally , if S is any convex subset of ℝn and s0 ε S is a fixed point
, then the homotopy H( x,t) = (1-t)x + t s0
Shrinks each point of S to the point s0 without moving the point s0 . Therefore
, each loop in S based at s0 is null homotopic , and π1(S ) =0.

Example 2 The fundamental group π 1(S1 ) of the unit circle is

isomorphic the additive group ℤ of integers.
Recall that the unit circle S1= { (x , y) ε ℝ2  x2 + y2 = 1 } is a subspace of
ℝ2 .It can also be described in terms of complex numbers, viz.,
S1= { z ε ℂ : |z| = 1 }. Clearly S1 is a topological group under the
multiplication of complex numbers. We choose (1 ,0) ε S1 as the base point
and denote this point by po . Now we will show that π1(S1, po ) ≅ ℤ .
 19
Let p : ℝ S be the exponential map defined by p(t) = e , t ε ℝ ,
1 2π it

clearly , this is a continuous map which simply wraps the real line ℝ onto
the circle S1 infinite number of times

More over it has a nice property that each path α : I S1 with initial point
α(0) = po ε S1 can be lifted to a unique path α′ : I ℝ s.t. α′ (0) = 0 ε ℝ.

The path α′ need not be a path at 0, that is α′(1) need not be 0. But because
poα′= α , α′(1) must belong to the kernel of p , i.e. α′(1) ε ℤ . The integer
α′(1) is the number of complete revolution of S1(counterclockwise if α′(1)
is positive and clockwise if α′(1) is negative)made by any element of [α]
this integer is called the degree of the loop α .so deg([α]) = deg( α) .
This defines a map f : π1(S1, po )  ℤ defined by f ([α] ) = deg( α)
Now we shall show that f is an isomorphism.
For Homomorphism : Let [α] , [β] be any two elements of π1(S1, po ) .
Let α′ , β ′ : I  ℝ be the unique paths starting from origin of ℝ which are
lifts of α and β, respectively . Define a new path ω : I  ℝ by
 '  1
α (2t ) for t ∈ 0, 
 2
ω (t) = 
α ' (1) + β ' (2 t − 1) for t ∈  1 , 1
  2 

Then obviously ω also begin at origin and

 20
  1
 poα ( 2t ) for t ∈ 0, 
'

 2
p o ω (t) = 
( poα ' )(1) + .( poβ ' )(2 t − 1) for t ∈  1 , 1
  2 

  1
α (2 t ) for t ∈ 0, 
 2
=
 β ( 2 t − 1) for t ∈  1 , 1
  2 

= (α* β)( t ) , ∀ t ε I.
Thus , ω is a lift of α* β starting from the origin .
∴ deg([ α]*[β ] ) = deg ([ α* β ] ) = ω(1) = α′(1) + β′(1)
= deg([α] ) + deg( [β] ) .
and thus , we have
f ( [ α]*[β ] ) = f ( [ α* β ] ) = deg(α* β) = deg(α) + deg(β)
= f ([α]) + f ([β])
f is one-one map:Let [α] , [β] ε π1(S1, po ) such that deg([α] ) = deg( [β] )
This mean that if α′ : I  ℝ , β ′ : I  ℝ are lift of α , β
respectively starting from origin , then α′(1) = β′(1).
Define a homotopy H : I × I ℝ by
H(s ,t ) = (1-t) α′(s) + t β′(s) , (s , t) ε I × I
Then H is a homotopy between α and β′ relative to {0,1}.
′

This imply that p o H : I × I  S1 is a homotopy between α and β relative

to { 0,1} , and so [ α] = [β ].
f is onto : Let n ε ℤ . Define a loop γ : I  S1 by γ (t) = e2π ni t .
The path γ ′ : I  ℝ defined by γ ′(t) = nt , , starting at the origin of ℝ and
lifts the path γ .
∴ deg( [γ ] ) = γ ′(1) = n .
Thus f is an isomorphism and so π1(S1, po ) ≅ ℤ .
Remark : We have already proved that the fundamental group of two
homeomorphic spaces are isomorphic . Consequently, any space which is
homeomorphic to the unit circle S1 will have its fundamental group as the
additive group ℤ of integers.
For example all closed curve which have the subspace topology induced
from ℝ2 and , as shown below:
 21

have ℤ as their fundamental group . Infact , take a pencil , put it at a point

p in the plane ℝ2 , move your hand tracing some curve not crossing the
curve already traced anywhere , finally come back to the point p . Then the
curve so described is homeomorphic to the circle S1 and will have its
fundamental group isomorphic to the additive group ℤ of integers.
Example 3 : Fundamental group of Torus
We know that a torus T ( a surface) is homeomorphic to the product S1×S1
of two circles S1 , whereas the n- dimensional torus T n ≅ S1×S1… S1
(n-copies) . Each is a path connected spaces . Therefore , we have
π1(T) ≅ ℤ×ℤ and π1(T n ) ≅ ℤ×ℤ×……× ℤ ( n- copies).