B.A.

PART–III MATHEMATICS : PAPER I

(SEMESTER - V) ABSTRACT ALGEBRA
Department of Distance Education

Lesson No. :

SECTION : A
Punjabi University, Patiala

1.1 : GROUPS – I

1.2 : GROUPS – II
(All Copyrights are Reserved)

1.3 : GROUPS – III

1.4 : GROUPS – IV
B.A. PART–III MATHEMATICS : PAPER I
(SEMESTER - V) ABSTRACT ALGEBRA

LESSON NO. 1.1 AUTHOR : DR. CHANCHAL

GROUPS – I
Objectives
I. Introduction to Groups
II. Examples
III. Elementary Properties of a Group
IV. Semi-Groups and Monoids
V. Problems
VI. Self Check Exercise

I. Introduction to Groups :
A non empty set G, together with a binary composition * is said to form a group, if it
satisfies the following postulates
(i) Associativity: a * (b * c) = (a * b) * c, for all a, b, c  G
(ii) Existence of Identity:  an element e G, such that
a * e = e * a = a for all a G, where e is called identity element of G.
(iii) Existence of Inverse: For every a  G,  a'  G (depending upon a)
such that
a * a' = a' * a = e. Here, a' is called inverse of a
Remarks : (i) Since * is a binary composition on G, therefore for all a, b G, a * b
is a unique member of G. This property is called closure property.
(ii) If, in addition to the above postulates, G also satisfies the commutative
law i.e. a * b = b * a for all a, b G,
then G is called an abelian group or a commutative group.
(iii) We will use the symbol 'e' for identity of a group and 'a–1' for the inverse
of an element 'a' of the group. Further, instead of denoting the
composition as 'a * b' we will simply write 'ab' (which does not mean
multiplication of 'a' and 'b'.)
(iv) If the set G is finite (i.e., has finite number of elements) it is called a
finite group otherwise it is called an infinite group.
(v) Order of a group : It is defined as the number of elements in the finite
group. It is denoted by o(G) or |G|.
1
B.A. Part–III 2 Mathematics : Paper I

II. Examples
Example 1 : The set Z of integers forms an abelian group w.r.t the usual addition
of integers.
It is easy to verify the postualtes in the definition of a group as sum of two integers
is a unique integer (thus closure holds). Associativity of addition is known to us. 0
(zero) will be identity and negatives will be the respective inverse elements.
Commutativity again being obvious.
Example 2 : Similarly, sets Q of rationals, R of real numbers would also form
abelian groups w.r.t. addition.
Example 3 : Set of integers, w.r.t. usual multiplication does not form a group,
although closure, associativity, identity conditions hold.
Note : 2 has no inverse w.r.t. multiplication as there does not exist any integer a
s.t., 2. a = a.2 = 1.
Example 4 : The set G of all +ve irrational numbers together with 1 under
m ul t i p li ca t i on d oe s n o t f o rm a gr ou p a s c l osu re doe s n ot h ol d, b eca u se

3 . 3  3  G, however other conditions in the definition of a group are satisfied

here.
Example 5 : Let G be the set {1, –1}. Then it forms an abelian group under
multiplication. It is again easy to check the properties.
I would be identity and each element is its own inverse.
Example 6 : Set of all 2 × 2 matrices over integers under matrix addition would
be another example of an abelian group.
Example 7 : Set of all non zero complex numbers forms a group under
multiplication defined by
(a + ib) (c + id) = (ac – bd) + i (ad + bc)
Here, 1 = 1 + i.0 will be identity,

a b
and i will be inverse of a + ib.
2 2
a b a  b2
2

Note : a + ib is non-zero means that both a & b are not zero. Thus a2 + b2 0.
Example 8 : The set G of all nth roots of unity, where n is a fixed positive integer
forms an abelian group under usual multiplication of complex numbers.
We know that complex number z is an nth root of unity if zn = 1 and there exist
exactly n distinct roots of unity.
In fact the roots are given by e2ir/n
where r = 1, 2, ....., n and eix = cos x + i sin x.
B.A. Part–III 3 Mathematics : Paper I

If a, b G be any two members, then an = 1, bn = 1 thus (ab) n = an bn = 1.

 ab is an nth root of unity
 ab  G  closure holds.
Associativity of multiplication is true in complex numbers.
Again, since 1.a = a . 1 = a, 1 will be identity.

n
1 1 1
Also for any a  G, will be its inverse as    n  1 .
a  
a a

So, inverse of e2ir/n is e2i(n-r)/n and identity is e2i0/n = 1

Commutativity is obvious. Therefore, G is an abelian group.
In particular if n = 3, then G = {1, , 2}
In order to prove it an abelian group, we use the concept of composition table. We
form the composition table using 3 = 1 as given below :

* 1  2
1 1  
  2 1
2 2 1 

(i) Closure Property : Since all the elements in composition table are elements
of G, so G is closed under multiplication.
(ii) Associalivity : Since the elements of G are complex numbers and
multiplication of complex numbers is associative, so multiplication is associative
in G.
(iii) Existence of identity : Since 2nd row is same as the first row, 1 is left
identity, Also 2nd column is same as the first column, 1 is the right identity. So 1
is the identity of G.
(iv) Existence of inverse : Here each row (column) of the composition table
contains identity element 1 once and only once. So the element left to 1 is the left
inverse of the element above 1. Similarly the element above 1 is the right inverse of
element left to 1.
Thus we see that
1.1 = 1 = 1.1 so 1 –1 = 1.
Also  = 1 = 2., so –1 = 2 and (2)–1 = .
2

(v) Abelian : Since the entries in the composition table are symmetrical about
the principal diagonal.
Hence G is an abelian group under multiplication.
B.A. Part–III 4 Mathematics : Paper I

Example 9 : (i) Let G = {± 1, ± i, ± j, ± k}. Define product on G by usual multiplication

together with

i2  j2  k 2  1, ij   ji  k

jk  kj  i

ki  ik  j
then G forms a group. G is not abelian as ij ji.
This is called the Quaternion Group.
(ii) If set G consists of the eight matrices

1 0   1 0   i 0   i 0   0 1  0 1
0 1  ,  0 1 , 0 i  ,  0 i  ,  1 0  , 1 0  ,
           

0 i   0  i 
 i 0  ,  i 0  , where i  1
   

then G forms a non abelian group under matrix multiplication. (Compare with part
(i)).
Example 10 : Let G = {(a, b) | a, b rationals, a 0}. Define * on G by
(a, b) * (c, d) = (ac, ad + b)
Closure follows as a, c 0 ac 0
[(a, b) * (c, d)] * (e, f) = (ac, ad + b) * (e, f)
= (ace, acf + ad + b)
(a, b) * [(c, d) * (e, f)] = (a, b) * (ce, cf + d)
= (ace, acf + ad + b)
proves associativity.
(1, 0) will be identity and (1/a, –b/a) will be inverse of any element (a, b).
G is not abelian as
(1, 2) * (3, 4) = (3, 4 + 2) = (3, 6)
(3, 4) * (1, 2) = (3, 6 + 4) = (3, 10).
III. Elementary Properties of a Group
Lemma : In a group G,
(1) Identity element is unique.
(2) Inverse of each a G is unique.
(3) (a–1)–1 = a, for all a G, where a–1 stands for inverse of a.
(4) (ab)–1 = b–1 a–1 for all a, b G (Reversal Law)
(5) ab = ac b = c
B.A. Part–III 5 Mathematics : Paper I

ba = ca b = c for all a, b, c G

(called the cancellation laws).
Proof : (1) Suppose e and e' are two elements of G which act as identity.
Then, since e G and e' is identity,
e'e = ee' = e .... (1)
and as e' G and e is identity
e'e = ee' = e' ..... (2)
(1) and (2) e = e'
which proves the uniqueness of identity in a group.
(2) Let a G be any element and let a' and a" be two inverse elements of a, then
aa' = a'a = e
aa" = a"a = e
Now a' = a'e = a'(aa") = (a'a) a" = ea" = a".
Hence, inverse of an element is unique. (By associativity)
(3) Since a–1 is inverse of a
aa–1 = a–1a = e
which also implies a is inverse of a–1
Thus (a–1)–1 = a.
(4) We have to prove that inverse of b–1 a–1 for which we show
(ab) (b–1 a–1) = (b–1 a–1) (ab) = e.
Now (ab) (b–1 a–1) = [(ab) b–1] a–1
= [(a(bb–1)] a–1 (By associativity)
= (ae) a–1 = aa–1 = e
Similarly (b–1 a–1) (ab) = e
and thus the result follows.
(5) Let ab = ac, then
b = eb = (a–1 a) b
= a–1 (ab) = a–1 (ac)
= (a–1 a) c = ec = c
Thus ab = ac b = c
which is called the left cancellation law.
Similarly, one can prove the right cancellation law.
Example 11 : Let X = {1, 2, 3} and let S3 = A(X) be the group of all permutations on
X. Consider f, g, h A(X), defined by
f (1) = 2, f(2) = 3, f(3) = 1
g(1) = 2, g(2) = 1, g(3) = 3
h(1) = 3, h(2) = 1, h(3) = 2
Then, it is easy then to verify that fog = goh
B.A. Part–III 6 Mathematics : Paper I

But f h.
(b) If we consider the group in example 10, we find
(1, 2) * (3, 4) = (3, 6) = (3, 0) * (1, 2)
But (3, 4) (3, 0)
Hence we notice, cross cancellations may not hold in a group.
Theorem 1 : For elements a, b in a group G the equations ax = b and ya = b have
unique solutions for x and y in G.
Proof : Now ax = b
 a–1 (ax) = a–1b
 ex = a–1 b
or x = a–1 b
which is the required solution of the equation ax = b.
Suppose x = x1 and x = x2 are two solutions of this equation, then
ax1 = b and ax2 = b
 ax1 = ax2
 x1 = x2 by left cancellation
Showing that the solution is unique.
Similarly y = ba–1 will be unique solution of the equation ya = b.
Theorem 2 : A non empty set G together with a binary compositioon '.' is a group
if and only if
(1) a(bc) = (ab)c for all a, b, c G
(2) For any a, b G, the equations ax = b and ya = b have solutions in G.
Proof : If G is a group, then (1) and (2) follow by definition and previous theorem.
Conversely, let (1) and (2) hold. To show G is a group, we need to prove existence of
identity and inverse (for each element).
Let a G be any element.
By (2), the equation ax = a
ya = a
have solutions in G.
Let x = e and y = f be the solutions.
Thus  e, f G, s.t., ae = a
fa = a
Let now b G be any element then again by (2)  some x, y in G s.t.,
ax = b
ya = b.
Now ax = b f. (a. x) = f . b
(f . a) . x = f . b
a . x = f . b
B.A. Part–III 7 Mathematics : Paper I

b = f . b
Again y.a = b (y . a) . e = b. e
y . (a . e) = b . e
y . a = be
b = be
thus we have b = fb ... (i)
b = be ... (ii)
for any b G
Putting b = e in (i) and b = f in (ii) we get
e = fe
f = fe
 e = f.
Hence ae = a = fa = ea
i.e., e G, s.t., ae = ea = a
 e is identity.
Again, for any a G, and (the identity) e G, the equations ax = e and ya = e have
solutions.
Let the solutions be x = a1, and y = a2
then aa1 = e, a2a = e
Now a1 = ea1 = (a2a) a1 = a2 (aa1) = a2e = a2.
Hence aa1 = e = a1a for any a G
i.e., for any a G, some a1 G satisfying the above relations a has an inverse.
Thus each element has inverse and, by definition, G forms a group.
IV. Semi-Groups and Monoids
Definition : A non empty set G together with a binary composition '.' is called a
semi-group if
a . (b . c) = (a . b) . c for all a, b, c G
Obviously, every group is a semi-group. But the converse is not true follows by
considering the set N of natural numbers under addition.
Remark : Cancellation law may not hold in a semi-group.
Consider M, the set of all 2 × 2 matrices over integers under matrix multiplication,
which forms a semi-group.

1 0  0 0  0 0 
If we take A , B  0 2  , C   3 0 
0 0     
B.A. Part–III 8 Mathematics : Paper I

0 0 
then clearly AB  AC   
0 0 

But B C.
Set of natural numbers under addition is an example of a semi-group in which
cancellation laws hold.
Theorem 3 : A finite semi-group in which cancellation laws hold is a group.
Proof : Let G = {a1, a2, ..., an} be a finite semi-group in which cancellation laws hold.
Let a G be any element, then by closure property
aa1, aa2, ..., aan
are all in G.
Suppose any two of these elements are equal
say, aai = aaj for some i j
then ai = aj by cancellation
But ai aj as i j
Hence no two of aa1, aa2, ..., aan can be equal.
These being n in number, will be distinct members of G (Note of (G) = n).
Thus if b G be any element then
b = aai for some i
i.e., for a, b G the equation ax = b has a solution (x = ai) in G.
Similarly, the equation ya = b will have a solution in G.
G being a semi-group, associativity holds in G.
Hence G is a group (by theorem 2).
Theorem 4 : A finite semi-group is a group if and only if it satisfies cancellation
laws.
Proof : Follows by previous theorem.
Definition (Monoid) : A non empty set G together with a binary composition '.'
is said to form a monoid if
(i) a(bc) = (ab)c a, b, c G
(ii) an element e G s. t., ae = ea = a a G
e is then called identity of G. It is easy to see that e is unique.
So all groups are monoids and all monoids are semi-groups.
Notation : Let G be a group with binary composition '.'. If a G be any element
then by closure property a . a G. Similarly (a . a) . a G and so on.
It would be very convenient to denote a . a by a2 and a . (a . a) or (a . a) . a by a3 and
so on. Again a–1.a–1 would be denoted by a –2. And since a.a–1 = e, it would not be
wrong to denote e = a0. It is now a simple matter to understand that
am . an = am+n, (am)n = amn
B.A. Part–III 9 Mathematics : Paper I

where m, n are integers.

In case the binary composition of the group is denoted by +, we will talk of sums
and multiples in place of products and powers. Thus here 2a = a + a, and na = a + a
+ ... + a (n times), if n is +ve integer. In case n is –ve integer then n = –m, where m is
+ve and we define na = –ma = (–a) + (–a) + ... + (–a) m times.
V. Problems
Problem 1 : If G is a finite group of order n then show that for any a G, some
positive integer r, 1 r n, s.t., ar = e.
Solution : Since o(g) = n, G has n elements.
Let a G be any elements. By closure property a2, a3, ....all belong to G.
Consider e, a, a2, .....an
These are n + 1 elements (all in G). But G contains only n elements.
at least two of these elemnts are equal. If any of a, a2, .....,an eqyaks e, out result
is proved. If not then ai = aj for some i, i, 1  i, j  n. Without any loss of generality,
we can take i >j
then ai = aj
ai. a-j = aj. a-j
ai-j =e where 1 i – j  n.
Problem 2 : Show that a finite semi-group in which cross cancellation holds is an
abelian group.
Solution : Let G be the given finite semi-group. Let a, b  G be any elements,
Since G is a semi-group, by associativity
a(ba) = (ab)a
By cross cancellation then ba = ab G is abelian.
Since G is abelian, cross cancellation laws become the cancellation laws. Hence G
is a finite semi-group in which cancellation laws hold.
thus G is a group.
Problem 3 : If G is a group in which (ab)i = aibi forthree consecutive integers i and
any a, b and G, then show htat G is abelian.
Solution : Let n, n + 1, n + 2 be three consecutive integers for which the given
condition holds. Then for any a, b G,
(ab)n = anbn ..(1)
(ab)n + 1 = an+1 bn+1 ...(2)
(ab)n + 2 = an+2 bn+2 ...(3)
Now (ab)n + 2 = an+2 bn+2
 (ab)(ab)n +1 = an+2 bn+2
 (ab)(an+1 bn+1) = an+2 bn+2
B.A. Part–III 10 Mathematics : Paper I

ban+1 = an+1 b (using cancellation) ...(4)

Similarly (ab)n+1 = an+1 bn +1
gives (ab)(ab)n + an+1 bn+1
i.e., (ab)(anbn) = an+1 bn+1
 ban = anb
ban+1 = anba
an+1 b = anba (4)
ab = ba.
Hence G is abelian.
Problem 4 : Let G be a semi-group, Suppose e  G, s.t., ae = a for all a G and
for each a G, a' G, s.t., aa' = e. Show that G is a group.
Solution : We first show that G satisfies the right cancellation law.
Let ac = bc
As given  c' G, s.t., cc' = e
(ac)c' = (bc)c'
a(cc') = b(cc')
ae = be a = b.
We now show that e is left identity
Consider, (ea)a' = e(aa') = e, e = e
Also aa' = e
e is also left identify of G.
Again (a'a)a' = a'(aa') = a'e = a'
and ea' = a'
(a'a)a' = ea'
a'a = e by right cancellation law
a' is also left inverse of a
so, G is a group.
Problem 5 : If in a semi-group S. x2y = y = yx2 x, y. Then show that S is abelian.

Solution : x2y = y x2y2 = y2

yx2 = y x, y S
xy2 = x x, y S
x2y2 = x2
So x2 = y2 x, y S
Now x y = y y y = y y3 = y
2 2
y S
Also yx2y = y2 (i)
Now xy2 = x xy2x = x2 (ii)
By (i) and (ii), xy 2x = yx2y
B.A. Part–III 11 Mathematics : Paper I

Since y = y3 y S, we get

xy = (xy)3 = xy xy xy
= xy xy x3y = x(yx)2x(xy)
= (yx)x2(yx) (xy)
= yx3 yx2y = yxy x2y
= (yx)xy 2x
= y(y2x) (as y = yx2)
= y 3x
= yx (as y 3 = y)
Thus xy = yx x, y S
Hence S is abelian.
Problem 6 : Show that if G is a group then a G is an idempotent if and onlyif a =
e, the identity of G.
Solution : Given G is a group.
Let a G is an idempotent element
 a2 = a
 aa = ae
 a = e. [Using left cancellation law]
Conversely let a = e, the identity of G
 aa = ae
= ee = e = a [  a = e]
 a2 = a
 a is an idempotent element.
VI. Self Check Exercise
1. Check whether the following systems form a group or not
(a) G = set of rational numbers under composition * defined by

ab
a*b , a, b G
2

(b) S e t o f a l l 2 × 2 m a t r i ce s o v er i n t e g er s u n de r m a t r i x
multiplication.

 cos  sin  
(c) Set of all matrices of the form   ,   R, under matrix
  sin  cos  

multiplication.
(d) Q = set of all rational numbers under * where a * b = a + b – ab.
(e) G = {(a,b) | a, b Z} under * defined by
(a, b) * (c, d) = (ac + bd, ad +bc).
B.A. Part–III 12 Mathematics : Paper I

2. Let G be the set {± e, ± a, ± b, ± c} where

1 0  1 0  0 1 0 1 
e  , a  0 1 , b   1 0  , c  1 0  .
 0 1       

Show that G forms a group under matrix multiplication.

3. Show that a group G is abelian iff (ab)2 = a2b2.
4. Prove that a group in which every element is its own inverse is abelian.
5. Show that a monoid is a group if and only if cancellation laws hold in
it.
6. Show that a finite semi-group G with identity is a group iff G contains
only one idempotent.
B.A. PART–III MATHEMATICS : PAPER I
(SEMESTER - V) ABSTRACT ALGEBRA

LESSON NO. 1.2 AUTHOR : DR. CHANCHAL

GROUPS – II
Objectives
I. Subgroups
II. Properties of Subgroups
III. Centre of a Group
IV. Cosets
V. Self Check Exercise

I. Subgroups
We have seen that R, the set of real numbers, forms a group under addition, and Z,
the set of integers, also forms a group under addition. Also Z is a subset of R. Now,
we define a subgroup as :-
Definition : A non empty subset H of a group G is said to be a subgroup of G, if H
forms a group under the binary composition of G.
Obviously, if H is a subgroup of G and K is a subgroup of H, then K is subgroup of
G. If G is a group with identity element e then the subsets {e} and G are the trivial
subgroups of G. All other subgroups will be called non-trivial (or proper subgroups).
Thus it is easy to see that the even integers form a subgroup of (z, +), which is :
subgroup of (Q, +) which is a subgroup of (R, +).
Again the subset {1, –1) will be a subgroup of G = {1, –1, i, –i) under multiplication.
Notice that Z 5 = {0, 1, 2, 3, 4} mod 5 is not a subgroup of Z under addition as
addition modulo 5 is not the composition of Z. Similarly, Z5 is not subgroup of Z6
etc.
We sometimes use the notation H  G to signify that H is a subgroup of G and H < G
to mean that H is a proper subgroup of G.
II. Properties of Subgroups
I. The identity element of a subgroup is same as the identity elements of
the group.
Proof. Let H be a subgroup of a group G.
Let e and e' be the identity elements of G and H respectively
Let a H be any element
13
B.A. Part–III 14 Mathematics : Paper I

a e' = a [  e' is the identity of H]

Also  a  H and H  G a G
 a e= a [  e is the identity of G]
 we have a e = a e'
 e = e' [by left cancellation law]
Hence the identity of a group and that of a subgroup is the same.
II. The inverse of any element of a subgroup is the same as the inverse of
the element regarded as the element of the group.
Proof. Let e be the identity element of G and H.
Let a H be any element.
Since H  G a G.
Let b be the inverse of a in H and c be the inverse of a in G.
b a = e and c a = e
ba = c a
b = c [by right cancellation law]
Hence the inverse of any element of a subgroup is same as the inverse of the same
element regarded as an element of the group.
III. The order of any element in a subgroup is the same as the order the
element regarded as the element of the group.
Proof. Let e be the identity element of G and H.
Let a H such that o(a) = n
an = e and am e for every m < n.
Also a H a G and so an = e G o(a) = n in G.
Hence order of any element in a subgroup is same as the order of element regarded
as the element of the group.
IV. Subgroup of an abelian group is abelian.
Proof. Let H be a subgroup of an abelian group G
H G.
Let a, b H be any two elements
a, b G a b = b a [  G is abelian]
a, b H we have a b = b a
Hence H is an abelien subgroup of G.
The converse of above result is false
i.e., A subgroup may be abelian even if G is not abelian.
For example : (i) The sets {1, –1} and {1, –1, i, – i} are abelian subgroups of the non-
abelian group of Quaternions Q8 under multiplication.
Theroem 1 : A non empty subset H of a group G is a subgroup of G iff
(i) a, b H ab H
(ii) a H a–1  H.
B.A. Part–III 15 Mathematics : Paper I

Proof: Let H be a subgroup of G then by definition it follows that (i) and (ii) hold.
Conversely, let the given conditions hold in H.
Closure holds in H by (i).
Again a, b, c H a, b, c G a(bc) = (ab)c
Hence associativity holds in H.
Also for any a H, a–1 H and so by (i)
aa–1 H e H
thus H has identity.
Inverse of each element of H is in H by (ii).
Hence H satisfies all conditions in the definition of a group and thus it forms a
group and therefore a subgroup of G.
Theorem 2 : A non void subset H of a group G is a subgroup of G iff a, b H ab–
1
H .
Proof : If H is a subgroup of G then, a, b  H  ab–1  H (follows easily by using
definition).
Conversely, let the given condition hold in H.
That associativity holds in H follows as in previous theorem.
Let a H be any element (H )
then a, a H aa–1 H e H.
So H has identity.
Again, for any a H, as e H
ea–1 H a–1 H
i.e., H has inverse of each element.
Finally, for any a, b H ,
a, b–1 H
a(b–1)–1 H ab H
i.e., H is closed under multiplication.
Hence, H from a group and therefore a subgroup of G.
Remark : If the binary composition of the group is dehated by '+', the above condition
can be stated as : a, b H a – b H. Note that, the identity e is always in H.
Problem 1 : Show that the sets H = {0, 3} and K = {0, 2, 4} are subgroups of the
group G = {0, 1, 2, 3 , 4, 5} under the operation addition modulo 6.
Sol. Clearly, H and K are non-empty subsets of G. The composition table for H
and K are given below:
Composition table for H Composition table for K
+6 0 3 +6 0 2 4
0 0 3 0 0 2 4
3 3 0 2 2 4 0
4 4 0 2
B.A. Part–III 16 Mathematics : Paper I

From the composition table, it is easy to see that H and K form groups and hence are
subgroups of G.
III. Centre of a Group
Theorem 3 : Centre of a group G is a subgroup of G.
Proof : Let Z(G) be the centre of the group G.
Then Z(G) as e Z(G)
Again, x, y Z(G) xg = gx
yg = gy for all g G
g–1 x–1 = x–1 g–1
g–1 y–1 = y –1 g –1 for all g G
Now g(xy –1) = (gx)y–1 = (xg)y –1
= (xy)y –1 (g–1g)
= xg(y –1 g –1)g = xg (g–1 y–1)g
= x(gg –1) y–1 g
= (xy–1)g for all g G
xy–1  Z(G)
Hence Z(G) is a subgroup.
Remark : obviously, G is abelian iff Z(G) = G.
Definition : Let G be a group, a G be any element. The subset N(A) = {x G/xa =
ax} is called normalizer or centralizer of a in G.
Problem 2 : Find centre of S3.
Solution : We have S3 = {I, (1), (13), (23), (123), (132)}
Centre of S3, Z(S3) = {S3| = for all S3}
Since (12)(13) = (132)
(13)(12) = (123)
We find (12), (13) do not commute.
(12) & (13) do not belong to Z(S3)
Again, (23)(132) = (12)
(13) (12) = (123)
(23), (132) do not belong to Z(S 3)
Also, (123)(12) = (13)
(132)(23) = (13)
Shows (123)  Z(S3)
Hence Z(S 3) contains only l.
Problem 3 : Let G be the group of all 2 × 2 non singular matrices over the reals.
Find centre of G.

a b 
Solution : If   be any element of the centre Z(G) of G then it should commute
c d

with all members of G, In particular we should have

B.A. Part–III 17 Mathematics : Paper I

a b  0 1   0 1  a b 
 c d  1 0    1 0   c d 
     

b = c, a = d

a b  0 1 0 1 a b 
Also  c d  1 1  1 1  c d  gives
     

a  b b   a b 
 c  d d   a  c b  d 
   

a + b = a, b = c = 0

a b  a 0 
Hence any member  c d  of Z(G) turns out to be of the type 0 a  .
   

In other words, members of the centre Z(G) are the 2 × 2 scalar matrices of G.
Problem 4 : Show that N(x–1 ax) = x–1 N(a)x for all a, x G.
Solution: Let y N(x–1 ax)
then (x–1 ax)y = y(x–1 ax)
y–1 x–1 axy = x–1 ax
xy–1 x–1 a = axy –1 x–1
 xy–1 x–1 N(a)
xy–1 x–1 = b N(a)
y–1 = x–1 bx
y = x–1 b–1 x, b–1 N(a) as b N (a)
y x–1 N(a) x
 N(x–1 ax) x–1 N(a) x
Let z x–1 N(a)x z = x-1 cx, c N(a)
 z(x–1 ax) = (x–1 cx) (x–1 ax)
= x–1 cax
= x–1 acx as c N (a)
= (x–1 ax)(x–1 cx)
= (x–1 ax)z
z N(x–1 ax)
x–1 N(a)x N(x–1 ax)
x–1 N(a)x = N(x–1 ax) for all a, x G.
It would be an easy exercise to show that intersection of two subgroups will be a
subgroup.
In fact, one can prove that if {Hi| i I} be any set of subgroups of group G then
B.A. Part–III 18 Mathematics : Paper I

 Hi will be a subgroup of G.
iI

Problem 5 : Show that union of two subgroups may not be a subgroup.

Solution : Let H2 = {2n |n Z}
H3 = {3n | n Z}
where (Z, +) is the group of integers. H2 and H3 will be subgroups of Z and
2n – 2m = 2(n – m) H2
Now, H2 H3 is not a subgroup as 2, 3 H2 H3
But 2 – 3 = –1 H2 H3
Can union of two subgroups be a subgroup ? For this, we prove the following
theorems.
Therorem 4 : Union of two subgroups is a subgroup iff one of them is contained in
the other.
Proof : Let H, K be two subgroups of a group G and suppose H K then H K = K
which is a subgroup of G.
Conversely, let H, K be two subgroups of G s.t., H K is also a subgroup of G. We
show one of them must be contained in the other. Suppose it is not true, i.e.,

H  K, K  H

Then x H s.t., x K

x H s.t., y H
Also then x, y H K and since H K is a subgroup, xy H K
xy H or xy K
Ir xy H, then as x H, x–1 (xy) H y H, which is not true.
Again, if xy K, then as y K, (xy)y–1 K x K which is not true.
i.e., either way we land up with a contradiction.

Hence our supposition that H  K and K  H is wrong.

Thus one of the two is contained in the other.

IV. Cosets
Definition : Let H be a subgroup of a group G. For a, b G, we say a is congruent
to be mod H if ab–1 H.
In notational form, we write a b mod H.
It is easy to prove that this relation is an equivalence relation. Corresponding to
this equivalence relation, we get equivalence classes. For any a G, the equivalence
class of 'a' will be given by
cl(a) = {x G| x a mod H}.
Definition (Coset) : Let H be a subgroup of G and let a  G be any element.
Then Ha = {ha | h H} is called a right coset of H in G.
B.A. Part–III 19 Mathematics : Paper I

We show in the following theorem that any right coset of H in G is an equivalence

class.
Theorem 5 : Ha = {x G| x a mod H} = cl(a) for any a G.
Proof : Let x Ha
Then x = ha for some h H
xa–1 h
 xa–1  H
x a mod H
x cl(a)
thus Ha cl(a).
A gain let x cl(a) be any element.
Then x a mod H
xa–1 H
xa–1 = h for some h H
x = ha Ha
thus cl(a) Ha
and hence Ha = cl(a)
Since right cosets are equivalence classes, therefore we are free to use the results
that we know about equivalence classes. We can, therefore state that any two right
cosets are either equal or have no element in common and also that union of all the
right cosets of H is G will be equal to G.
Remark : Note that a coset is not essentialy a subgroup. If G be the Quaternion
group, then H = {1, –1} is a subgroup of G. Take a = i, then Ha = {i, –i} which is not a
subgroup of G (it doesn't contain identity).
Lemma : There is always a 1 – 1 onto mapping between any two night cosets of H in
G.
Proof : Let Ha, Hb be any two right cosets of H in G.
Define a mapping f : Ha  Hb, s.t.,
f (ha) = hb
Then h 1a = h2a h1 = h 2 h1b = h2b
f(h1a) = f(h2a)
i.e., f is well defined.
f(h1a) = f(h2a) h1a = h2a h1 = h2 h1a = h2a
Showing f is 1 – 1.
That f is onto, is easily seen, as for any hb Hb, ha would be its pre image.
The immediate utility of this lemma is seen, if the group G happens to be finite,
because in that case the lemma asserts that any two right cosets of H in G have the
same number of elements. Since H = He is also a right coset of H in G, this lead us
to state that all right cosets of H in G have the same number of elements as are in
B.A. Part–III 20 Mathematics : Paper I

H(G, being, of course, finite). We are now ready to prove the following result.
Theorem 6 (Lagrange's) : If G is a finite group and H is a subgroup of G, then o(H)
divides o(G).
Proof : Let o(G) = n.
Since corresponding to each element in G, we can define a right coset of H in G, the
number of distinct right cosets of H in G is less than or equal to n.
Using the properties of equivalence classes we know
G = Ho1 Ha2 .....Ha1
where t = no. of distinct right cosets of H in G.
o(G) = o(Ha1) + o(Ha2) + .... + o(Ha1)
(reminding ourselves that two right cosets are either equal or have no element in
common

o(G) = o(H) + o(Ha) ..........  o  H  using the above lemma

t times

o(G) = t. o(H)
or that o(H)| o(G)
Converse of Lagrange's theorem does not hold.
Remarks : (i) If G is a group of prime order, it will have only two subgroups G and
{e}.
(ii) A subset H  G with more than half the elements of G cannot be a subgroup of
G.
We have been talking about right cosets of H in G all this time. Are there left cosets
also?
The answer should be an obvious yes. After all we can similarly talk of
aH = {ah| h  H}, for any a  G, whch would be called a left coset. One can by
defining similarly an equivalence relation (a b mod H a–1 b H) prove all similar
results for left cosets.
Theorem 7 : Let H be a subgroup of G then
(i) Ha = H a H; aH = H a H
(ii) Ha = Hb ab–1 H; aH = bH a–1 b H
(iii) Ha (or aH) is a subgroup of G iff a H.
Proof : (i) Let Ha = H
Since e H, ea Ha ea H a  H.
Let a H, we show Ha = H.
Let x  Ha x = ha for some h H
Now h H, a H ha H x H Ha H
Again, let y H, since a H
ya–1 H
ya-1 h for some h H
B.A. Part–III 21 Mathematics : Paper I

y = ha Ha
H Ha
Hence Ha = H.
(ii) Ha = Hb
 (Ha)b–1 = (Hb)b–1
 Hab–1 = He
Hab–1 = H
ab–1  H (using (i))
(iii) If a H then Ha = H which is a subgroup. Conversely, if Ha is a subgroup
of G then e Ha and thus the right cosets Ha and He have one lement e in common
and hence Ha = He = H a H by (i)
Corresponding results for lelt cosets can be proved similarly.
Definition : Let G be a group and H, a subgroup of G. Then index of H in G is the
number of distinct right (left) cosets of H in G. Is is denoted by iG(H) or [G:H].

O(G)
The proof of Lagrange's theorem suggests that if G is a finite group then iG  H   O(H) .

Problem 6 : Give an example to show that an infinite group G can have a subgroup
H G with finite index.
Solution : Let < Z, + > be the group of integers under addition.
Let H = {2n | n Z} then H is a subgroup of Z. We show H has only three right cosets
in Z namely H, H + 1 ; H + 2.
If a Z be any element ( 0, 1, 2) then we can write (bi division algorithm).
a = 3n + r, 0 r < 3
which gives
H + a + H + (3n + r) = (H + 3n) + r = H + r
where 0  r < 3
Hence H has only 3 right cosets in Z and thus has index 3.
Notice, H – 1 = (H + 3) – 1 = H + (3 – 1) = H + 2 etc.
Problem 7 : Show that there exists a one-one onto map between the set of all left
cosets of H in G and the set all right cosets of H in G where H is a subgroup of a
group G.
Solution : Let  = set of all left cosets of H in G.
 = set of all right cosets of H in G.
Define a mapping :    , s.t.,
(aH) = Ha–1 a G
is well defined as aH = bH
a–1 b H
B.A. Part–III 22 Mathematics : Paper I

 Ha–1 = Hb-1
(aH) = (bH)
Taking the steps backwards, we find  is 1 – 1. Again, for any Ha   , a–1 H is the
required pre-image under is onto.
If G is finite, then the above result reduces to saying that number of left cosets of H
in G is same as the number of right cosets of H in G.
Problem 8 : Let H be a subgroup of a group G and N(H) = {a G|aHa–1 = H}. Prove
that N(H) is a subgroup of G which contains H.
Solution : N(H)  subset of G as
eHe–1 = H e N(H)
Let now a, b N(H) be any two elements, then
aHa–1 = H
bHb–1 = H
then bHb = H b–1 (bHb–1) b = b–1 Hb
–1

(b–1 b) Hb–1 b= b–1 Hb

H = b–1 Hb
 aHa–1 = a(b–1 Hb)a–1
 aHa–1 = ab–1 Hba–1
H = (ab–1) H(ab–1)–1
ab–1  N(H) i.e., N(H) is a subgroup of G.
Since h H hHh–1 = H (Ha = H a H etc).
We find h N(H) showing that H N(H).
Problem 9 : Suppose that H is a subgroup of a group G such that whenever Ha 
Hb. Prove that gHg–1 H for all g G.
Solution : It is given that if Ha  Hb then aH bH
thus if aH = bH then Ha = Hb
Let now g G, h H be any elements, then
(g–1 h) (g–1)–1 H (Ha = Hb ab–1  H)
By (1) H(g–1 h) Hg–1
(g–1 h)(g–1)–1 H (Ha = Hb ab–1 H)
g–1 hg H for all h H
g–1 He H.
D e f i ni t i o n : L et H a n d K b e t w o su b g r ou p s o f a g r ou p G . W e d e f i n e
HK = {hk: hH, k K}, then HK will be a non empty subset of G.
Theorem 8 : HK is a subgroup of G iff HK = KH.
Proof : Let HK be a subgroup of G. We show HK = KH
Let x  HK be any element
Then x–1 HK (as HK is a subgroup)
B.A. Part–III 23 Mathematics : Paper I

x–1 = hk for some h H, k K

x = (hk)–1 = k–1 h–1 KH
thus HK KH
Again let y KH be any element
then y = kh for some k K, h H
y–1 = h –1 k–1 HK
y HK (as HK is a subgroup)
KH HK
Hence HK = KH.
Let a, b HK be any two elements, we show ab–1 HK
a, b HK a = h1k1 for some h1, h2 H
b = h 2k2 k1, k2 K
1 1
Then ab–1 = (h1k1)(h2k2)–1 = (h1k1) k 2 h2  
 
1 1
= h1 k 2k 2 h2

Now k k  h
1
1 2
1
2  KH  HK

thus k k  h
1
1 2
1
2  hk for some h  H, k K

then ab–1 = h1(hk) = (h1h)k HK

Hence HK is a subgroup.
Remarks : (i) HK = KH does not mean that each element of H commutes with every
element of K. It only means that for each h  H, k  K, hk = k 1 h 1 for some
k1  K and h1 H.
(ii) If G has binary composition +, we define
H + K = {h + k | h H, k K}.
Theorem 9: If H and K are finite subgroups of a group G then

o  H .o  K 
o  HK  
o H  K 

Proof : Let D = H K then D is a subgroup of K and as in the proof of Lagrange's

theorem,  a decomposition of K into disjoint right cosets of D in K and
K = DK1 Dk2 ......Dk1

o K 
and also t  o D
 
B.A. Part–III 24 Mathematics : Paper I

 1 
Again, HK  H   Dk i  and since D H, HD = H
 i 1 

1
Thus HK   Hk i = Hk1 Hk2 ...... Hki
i 1

Now no two of Hk1, Hk2, ........., Hkt can be equal as if Hki = Hkj for some i, j
1 1 1
then k i k j  H  k i k j  H  K  k i k j  D  Dk i  Dk j

which is not true.

Hence o(HK) = o(Hk1) + (Hk2) + ....+ o(Hkt)
= o(H) + o(H)+ .....+ o(H)
= t . o(H)

o  H  .o  K 
= o HK
 
Which proves the result.
V. Self check Exercise
1. Show that intersection of two subgroups of a group G is a subgroup of
G.
2. If H is a subgroup of G, show that
g–1 Hg = {g–1 hg |h H} is a subgroup of G.
3. Let G be the Quaternion group. Find centre of G. Find also the
normalizer of i in G.
4. Show that normalizer of an element a in a group G is a subgroups of
G.
5. Show that H = {0, 2, 4} is a subgroup of Z6 = {0, 1, 2, 3, 4, 5} addition
modulo 6.
6. If H and K are subgroups whose orders are relatively prime then show
that HK = {e}.

7. Show that for a group G, Z(G) = N  a  .

aG

8. Show that the centralizer C(H) of a subgroup H of a group G is a

subgroup of G.
B.A. PART–III MATHEMATICS : PAPER I
(SEMESTER - V) ABSTRACT ALGEBRA

LESSON NO. 1.3 AUTHOR : DR. CHANCHAL

GROUPS – III
Objectives
I. Cyclic Groups
II. No. of Generators of a finite cyclic group.
II.(a) Euler's Theorem
II. (b) Fermat's Theorem
III. Normal Subgroup
IV. Self Check Exercise

I. Cyclic Groups
Firstly , we define the order of an element:-
Definition (Order of an element) : Let G be a group and a G be any element.
We say a is of order (or period) n if n is the least +ve integer s.t., an = e. If binary
composition of G is denoted by +, it is read as na = 0, where 0 is identity of G.
If it is not possible to find such n, we say a has infinite order. Order of a is be
denoted by o(a) or |a|. It is obvious that o(a) = 1 iff a = e.
Cyclic Group : A group G is called a cyclic group if an element a G, such that
every element of G can be expressed as a power of a. In that case a is called
generator of G. We express this fact by writing G = <a> or G = (a).
Thus G is called cyclic if  an element a  G s.t., G = {an |n  Z}. Again, if binary
composition of G is denoted by +, the words 'power of a' would mean multiple of a.
Note : The number of generators may be more than one. Moreover, if a is generator
so is a-1. A simple example of a cyclic group is the group of integers under addition,
1 being its generator.
Again the group G = {1, –1, i, –i} under multiplication is cyclic as we can express its
members as i, i2, i 3, i4. Thus i (or – i) is a generator of this group.
Example 1 : Consider, Z8 = {0, 1, 2, ....7} addition modulo 8.
Then we can check that 1, 3, 5, 7 will be generators of Z8
Here, 31 = 3, 32 = 3 3 = 6, 33 = 3 3 3 = 1
34 = 3 3 3 3 = 4 and so on
i.e., <3> = {3, 6, 1, 4, 7, 2, 5, 0}
25
B.A. Part–III 26 Mathematics : Paper I

or 3 is a generator of Z8 Observe that 1, 7 and 3, 5 are each others inverses.

Theorem 1 : Order of a cyclic group is equal to the order of its generator.
Proof : Let G = <a> i.e., G is a cyclic group generated by a.
Case (i) : o(a) is finite, say n, then n is the least +ve integer s.t., an = e.
Consider the elements a0 = e, a, a2,......an-1
These are all elements of G and are n is number.
Suppose any two of the above elements are equal
say ai = aj with i > j
then ai. a-j = e ai-j = e
But 0 < i – j n – 1 < n, thus a + ve integer i – j, s.t., ai–j = e and i – j < n, which is a
contradiction to the fact that o(a) = n.
Thus no two of the above n elements can be equal, i.e., G contains at least n
elements. We show it does not contain any other element. Let x G be any elemnt.
Since G is cyclic, generated by a, x will be some power of a.
Let x = am
By division algorithm, we can write
m = nq + r where 0 r < n
Now a = anq+r = (an)q. ar - eq, ar = ar
m

x = ar (where 0 r < n

i.e., x is one of a0 = e, a, a2, .....an–1
or G contains precisely n elements
 o(g) = n = o(a)
Case (ii): o(a) is infinite.
In this case no two powers of a can be equal as if an = am (n > m) then an–m = e, i.e., it
is possible to find a +ve integer n – m s.t., an–m = e meaning thereby that a has finite
order.
Hence no two powers of a can be equal. In other words words G would contain
infinite number of elements.
Problem 1 : If a G be of finite order n and also nm = e then show that n/m.
Solution : Let o(a) = n, then by definition n is the least +ve integer s.t.,
an =e.
Suppose am = e for some m
By division algorithm, m = nq + r, where 0  r < n
am = qnq + r
e = anq. ar = (an)q. ar = eq. ar = ar
where 0 r < n
Since n is such least +ve integer, we must have r = 0
i.e., m = nq or that n/m.
B.A. Part–III 27 Mathematics : Paper I

Problem 2 : If G is a finite abelian group then show that o(ab) is a divisor of

l.c.m. of o(a). o(b).
Solution : Let o(a) = n, o(b) = m, c(ab) = k.
Let l = 1 c.m. (m, n)
then m |l, n|l, l = mr 1, l = nr2
l l l
Now (ab) = a b (G is abelian)

 a nr2 bmrl e.e  e

 o(ab)|l
k|l.
Problem 3 : If in group G, a 5 = e, aba –1 = b 2 for a, b   G then show that
o(b) = 31.
Solution : We have b2 = aba–1
 b4 = (aba–1)(aba–1)
= ab(a–1 a)nr b2 abmrl = ab2a–1
= a(aba–1)a–1
b4 = a2ba–2
 b8 = (a2b–2) (a2ba–2) = a2b2a–2
= a2(aba–1)a–2 = a3ba–3
b16 = a4ba–4 (as above)
b32 = a5ba–5 = b as a5 = e
b31 = e 31 is a multiple of o(b)
Since 31 is a prime number, it is the least +ve integer such that b31 = e
o(b) = 31
We are, of course, taking b e.
Problem 4 : Let G be a finite group with more than one element. Show that G has
no element of prime order.
Solution : Let e a G
Consider a, a2,......a1, ........
Since G is finiter, for some integers i and j, ai = ai, i > j.
So, ai–i = e
an = e, n = i – j > 0
Since a e, n > 1

Let n  p11 p2 2 .....pr r , where pi's are distinct primes

 p11 p22 ......pr r 1 

So,  a   e
 
B.A. Part–III 28 Mathematics : Paper I

 p1 p2 ......prr 1 

  a 1 2   1 or pr

 

 p1p2 ......prr 1 
If O  a 1 2   pr , then the result follows
 

1 2  1
p ......pr r
Let a p1 2  e , then (proceeding as above, we get an element of prime order
as a e.)
Problem 5 : Suppose that G is a finite group with the property that every non
identity element has prime order. If Z(G) is non trivial, prove that every non identity
element of G has the same order.
Solution : Let e a Z(G). Let o(a) = prime p.
Let b G such that o(b) = prime q.
since a Z(G), ab = ba
So, (ab)pq = apq bpq = e
 o(ab) divises pq
o(ab) = 1, p or q
If o(ab) = 1, then a = b–1
 o(a) = o(b–1) o(b)
p = q .
If o(ab) = p, then (ab)p = e
apbp = e
bp = e
q = o(b)|p q = p.
Similarly, if o(ab) = q, then p = q.
Therefore, every non identity element of G has the same order.
Theorem 2 : A subgroup of a cyclic group is cyclic.
Proof : Let G = <a> and let H be a subgroup of G. If H = {e}, there is nothing to
prove.
Let H {e}. Members of H will be powers of a. Let m be the least +ve integer s.t., am
H. We claim H = <am>.
Let x H be any element. Then x = a4 for some k. By division algorithm, k = mq + r
where 0 r < m
r = k – mq
ar = ak. a–mq = x.(am)–q  H
B.A. Part–III 29 Mathematics : Paper I

But m is the least +ve integer s.t., am H, meaning thereby that r= 0
Thus k = mq
or that x = ak = (am)q H is cyclic, gererated by am.
Theorem 3 : A cyclic group is abelian.
Proof : Let G = <a>. If x, y G be any elements then x = an, y = am for some integers
m, n.
Now xy = an, am = an+m = am+n = am . an = y.x
Hence G is abelian.
Theorem 4 : If G is a finite group, then order of any element of G divides order of
G.
Proof : Let a G be any element,
Let H = {an| n an integer}
Then H is a cyclic subgroup of G, generated by a, as
x, y H x = an, y = am
 xy–1 = an. a–m = an–m H
By Lagrange's theorem o(H)|o(G). But o(H) = o(a)
 o(a)|o(G).
Cor.: If G is a finite group then for any a G, then ao(G) = e
Proof : o(a)|o(G) o(G) = o(a)k for some k
Now ao(G) = ao(a)k = (ao(a))k = ek = e
Thus any element of a finite group, has finite order (which is less than or equal to
the order of the group).
Problem 6 : Show that a group of even order has an element of order 2 and that
the number of elements of order 2 is odd.
Solution : Let o(G) = even
Let H = {x G|x2 = e}, K = {x G| x2 e},
Then G = H K
Now, x E x–1 x also is in K.
 number of elements is K is even and thus number of elements in H will also be
even.
Since, e H(a2 = e), some x H, s.t., x e.
Now, x e, x  H o(x) = 2
G has an element of order 2.
Every element of order 2 is in H, and as e H, o(H) = even, the number of elements
of order 2 is odd.
Theorem 5 : Converse of Lagrange's theorem holds in finite cyclic groups.
B.A. Part–III 30 Mathematics : Paper I

Proof : Let G = <a> be a finite cyclic group of order n,

Then o(G) = o(a) = n
Suppose m|n, We show  a subgroup of G having order m.
Since m|n, k s.t., n = mk
Let H be the cyclic group generated by ak
then H is a subgroup of G and o(H) = o(ak)
We show o(ak) = m
Now (ak)m = akm = an = e, as o(a) = n
Suppose now, that (ak)t = e
Then akt = e
o(a)|kt n|kt
km|kt m|t
thus o(ak) = m
which proves the result.
Problem 7 : Let G be a cyclic group of order n and suppose d divides n. Show that
xd = e has exactly d solutions.
Solution : Let G = <a>, then o(g) = o(a) = n. Since d|n, there exists a unique
subgroup H of G with order d. Let H = <b>
Then o(H) = o(b) = d.
H = {b, b2, ...... bd-1, bd = e}
If bi H be any element, then
(bi)d = (bd)i = e
Thus, every element of H is solution of xd = e, which gives d distinct solutions in G.
Let now c G be any solution of xd = e then cd = e
and, therefore, o(c)|d. If o(c) = m, then m|d = o(H) and thus there exists a subgroup
K of H s.t., o(K) = m. Since K is unique of order m, and <c> is also of order m, K = <c>
or that <c> H as K H and so c H and thus any solution of xd = e is in H.
Hence there exist exactly d solutions.
Theorem 6 : A group G of prime order must be cyclic and every element of G
other than identity can be taken as its generator.
Proof : Let o(G) = p, a prime
Take any a G, a e
and let H = {an|n an integer} then H is a cyclic subgroup of G.
 o(H)|o(G) o(H) = 1 or p
But o(H) 1 as a H, a e,
Thus o(H) = p H = G, i.e., G is a cyclic group generated by a. Since a was taken as
any element (other than e), any element of G can act as its generator.
B.A. Part–III 31 Mathematics : Paper I

Cor.: A group of prime order is abelian.

Theorem 7 : A group G of prime order cannot have any non trivial subgroups.
Proof : If H is any subgroup of G then as o(H)|o(G) = p, a prime
we find o(H) = 1 or p
i.e., H = {e} or H = G.
Theorem 8 : A group of finite composite order has at least one non-trivial
subgroup.
Proof : Let o(G) = n = rs where 1 < r, s < n
Since n > 1, e a G. Consider ar.
Cose (i) : ar = e
then o(a) r, let o(a) = k
then 1 < k  r < n (k > 1, as e)
Let H = {a, a2, a3, .....ak = e}
then H is a non empty. finite subset of G and it is closed under multiplication, thus
H is a subgroup of G. Since o(H) = k < n, we have proved the result.
Case (ii) : ar e, then since (ar)s = ars = an = ao(G) = e
o(ar)  s, Let o(ar) = t then 1 < t  s < n.
If we take K = {ar, 2r, ....atr = e} then K is a non empty finite subset of G, closed under
multiplication and is therefore a subgroup of G. Its order being less than n, it is the
required subgroup.
Theorem 9 : An infinite cyclic group has precisely two generators.
Proof : Let G = <a> be an infinite cyclic group.
As metioned earlier, if a is a generator of G then so would be a–1.
Let now b be any generator of G,
then as b  G, a generates G, we get b = an for some integer n
again as a G, be generates G, we get a = bm for some integer m
a = bm = (an)m = anm
anm–1 = e o(a) is finite and  nm – 1
Since of o(G) = o(a) is infinite, the above can hold only it
nm – 1 = 0 nm = 1

1
 m  or n =1 as m, n are integers.
n
i.e., b = a or a–1
In other words, a and a–1 are precisely the generators of G.
B.A. Part–III 32 Mathematics : Paper I

II. No. of Generators of a finite cyclic group.

Euler's function (or Euler's totient function). as :
For this purpose, we firstly define. For any integer n, we define (1) = 1 and for n >
1, (n) to be the number of +ve integers less than n and relatively prime to n. As an
example (6) = 2, (10) = 4 etc.
Note 1, 5 are less than 6 and relatively prime to 6 and 1, 3, 7, 9 (four in number)
are less than 10 and relatively prime to 10 etc. Obviously, (p) = p – 1, if p is a
prime.
Note : (i) If p1, p2, .....pn are distinct prime factors of n (>1), then

 1  1   1 
(n)  n 1   1   .... 1  
 p1   p2   pk 

(ii) If, m n are co-prime then

(mn) = (m) (n), (m, n 1)
We are now ready to prove
Theorem 10 : Number of generators of a finite cyclic group of order n is (n).
Proof : Let G= <a> be a cyclic group of order n
then o(a) o(G) = n
We claim am is generator of G iff (m, n) = 1, i.e., m, n are relatively prime.
[For instance, if n = 8, then (8) = 4 will be number of generators as we will show a,
a3, a5, a7 will generate G and no other element ca generate G. So here m can have
values 1, 3, 5, 7].
Let now am be a generator of G for some m
since a G, a = (am)i for some i
ami-1 = e o(a)|mi – 1
n|mi – 1
mi – 1 = nj for some integer j
i.e., mi – nj = 1
(m, n) =1
Conversely, let (m, n )= 1
Then integers x and y s.t.,
mx + ny = 1
amx + ny = a
amx. any = a
amx (an)y = a
amx = a as o(a) = n
a = (am)x
B.A. Part–III 33 Mathematics : Paper I

Since every element of G is a power of a and a itself is a power of am, we find am

generates G, which proves our result.
Remark : We thus realize that if a is a generator of a finite cyclic group G of order
n, then other generators of G are of the type am where m and n are coprime.
In fact an integer k will be a generator of Zn if and only if k and n are coprime, and
thus generators of Zn would indeed be the elements of Un.
II.(a) Euler's Theorem
Theorem 11 : Let a, n(n  1) be any integers such that g.c.d. (a, n) = 1.
Then a(n) 1 (mod n).
Proof : Let Un = {x|x is an integer, (x, n) = 1, 1 x < n}
Then U n is a group under multiplication modulo n.
By definition of Euler's –function,
o(Un) = (n).
If n = 1, then (n) = (1) = 1 and a(n) = al  1 (mod 1) (as 1 divides a – 1)
Let n>1
Now by Euclid's algorithm
a = nq + r, for some integers q, r where o r < n.
If r = 0. then a = nq n|a  (a, n) = n > 1, a contradiction
 1 r < n
Also (r, n) = d d|r, d| n d|a-nq, d|nq
d|a, d| n
d|(a, n)= 1
d = 1
(r, n) = 1 and 1 r < n
r Un
Also a = nq + r a  r(mod n)
It follows from Lagrange's theorem that,
r  r .....r = identity of Un = 1 [ao(G) = e]
where is composition multiplication modulo n is Un a (n) is order of group Un.
 r (n) – nq1 =1, for some integer q1
r (n)  1 (mod n)
a(n) 1 (mod n)
so a r (mod n) a(n) r (n) (mod n).
II.(b) Fermat's Theorem
Theorem 12 (Fermat's) : For any integer a and prime p,
a P  a  mod p  .
B.A. Part–III 34 Mathematics : Paper I

Proof : If (a, p) = 1, then by Euler's theorem

 p
a  1  mod p 

 a p 1  1  mod p  as   p   p  1

 a p  a  mod p 

If (a, p) = p, then p|a  p|a p

 p|a p  a

 a p  a  mod p  .
(Note (a, p) = 1 or p as 1 and p are only divisors of p).
Problem 8 : Prove that 3, 5, and 7 are the only three consecutive odd integers
that are prime.
Solution : Suppose p and p + 2 are consecutive primes, p > 3. We show that 12
divides their sum.
p > 3 g.c.d (p, 3) = 1
p2  1 (mod 3) by Fermat's theorem
3|p2 – 1
3|(p – 1)(p + 1)
If 3|p – 1, then p – 1 = 3k p = 3k + 1 p + 2 = 3k + 3 = multiple of 3.
But p + 2 is a prime > 3
So, we get a contradiction
Therefore, 3|p + 1 p + 1 = multiple of 3
Since p is odd, p + 1 is also a multiple of 2
So, p + 1 is a multiple of 6.
Therefore, p + (p + 2) = 2p + 2 = 2(p + 1) = multiple of 12.
12|p + (p + 2)
Problem 9 : Let G be a group.

o(a)
Show that o(an) =  o  a  , n  for all a G
where n is an integer and (0(a), n) = g.c.d (o(a), n).
Solution : Let o(a) = m
m n
Let d = (m, n)  , are integers
d d
B.A. Part–III 35 Mathematics : Paper I

 (an)m/d = (am)n/d = en/d = e

Let (an)r = e anr =e
o(a) |nr
m |nr

m n
 r
d d

m m n
 r as  ,   1
d  d d

m
r
d

  
o an 
m

o(a)
d o a  , n
.

III. Normal Subgroup :

Definition : A subgroup H of a group G is called a normal subgroup of G if
Ha = aH for all a G
Clearly G and {e} are normal subgroups of G known as the trivial normal subgroups.
A group G  {e} is called a simple group if the only normal subgroups of G are {e}
and G. Any group of prime order is simple.
It is easy to see that if H is a normal subgroup of G and K is a subgroup of G s.t., H
 K  G then H is normal in K. Again, if G is abelian, all its subgroups will be
normal. We use the notation H  G to convey that H is normal in G.
For Example : H = {1, –1} is a normal subgroup of the Quaternion group G.
Indeed Ha = {a, –a} = aH for any a G.
Theorem 13 : A subgroup H of a group G is normal in G iff g–1 Hg = H for all g G.
Proof : Let H be normal in G
then Hg = gH for all g G
g–1 Hg = g–1 (gH) = (g–1 g) H = H.
Conversely, let g –1Hg = H for all g G
Then g(g–1 Hg) = gH
(gg –1)Hg = gH
Hg = gH.
Hence H is normal.
B.A. Part–III 36 Mathematics : Paper I

Theorem 14 : A subgroup H of a group G is normal in G iff g–1 hg  H for all

h H , g G.
Proof : Let H be normal in G, then
Ha = aH for all a G
Let h H, g G be any elements, then
hg Hg = gh
hg = gh1 for some h1 H
g–1hg = h1 H
which proves the result,
Conversely, let a  G be any element,
Then a–1ha H for all h H
a(a–1 ha) aH for all h H
g–1 hg = h1 H
Ha aH
Taking b = a , we note, as b G
–1

b–1 hb H h H

aha–1 H for all h H
(aha–1)a Ha for all h H
ah Ha for all h H
aH Ha.
Hence Ha = aH, showing H is normal.
Theorem 15 : A subgroup H of a group G is normal subgroup of G iff product of
two right cosets of H in G is again a right coset of H in G.
Proof : Let H be a normal subgroup of G.
Let Ha and Hb be any two right coses of H in G.
then (Ha)(Hb) = H(aH)b
= H(Ha)b
= HHab
= HaB ab G
Conversely, we are given that product of any two right cosets of H in G is again a
right seet.
To shwo H is normal, let g G be any element.
Then Hg and Hg-1 are two right cosets of H in G. Thus HgHg –1 is also a right coset of
H G.
We claim HgHg –1 = He
egeg –1  HgHg–1
 e  HgHg–1
Also e H
B.A. Part–III 37 Mathematics : Paper I

Thus H and HgHg–1 are two right cosets having one element common. Recalling the
properties of equivalence classes, we know that any two right cosets are either
equal or have no element in common. Thus, (as e is common element)
H = HgHg–1
Now hgh1g–1 HgHg–1 for all h, h1 H, g G
hgh1g–1 H for all h, h 1 H, g G
h–1(hgh1g–1)h–1 H
gh1g–1 H for all h1 H, g G
H is normal in G.
Hence the result.
Remark : Let H be a subgroup of a group G. Define
g–1Hg = {g–1hg|h H}
then g–1Hg forms a subgroup of G.
Again, if we define a mapping f : H g–1Hg, by
f(h) = g–1hg
then f will be a 1 – 1 onto mapping.
In case G is finite, this would mean that both H and g–1Hg (for any g G) will have
same number of elements.
Using this result we have thus proved that if H be a subgroup of a finite group G
s.t., there is no other subgroup of G having the same number of elements as H has,
then H is normal in G. After all, H and g –1Hg (for any g G) have same number of
elements which mean that they are equal and H = g–1Hg means H is normal.
Problem 10 : Prove that a non empty subset H of a group G is normal subgroup
of G for all x, y H, g G, (gx)(gy)–1 H.
Solution : Let H be normal subgroup of G.
Let x, y H, g G be any elements,
then (gx)(gy)–1 = (gx)(y–1 g–1) = g(xy –1)g–1 H
as xy–1 H, g G, H is normal in G.
Conversely, we show H isnormal subgroup of G.
Let x, y H be any elements,
then xy–1 = exy –1 e = (ex)(ey)–1 H as e G
i.e., H is a subgroup of G.
Again, let h H, g G be any elements
Then as (gh)(ge)–1 H
we get (gh)(eg–1)  H
ghg–1 H
H is normal.
Problem 11 : Show that the normaliser N(a) of a in a group G may not be a normal
B.A. Part–III 38 Mathematics : Paper I

subgroup of G.
Solution : Let G = S3 and a = (23), then Na(a) = N((23)) = (S3|(23) = (23)} = {I,
(23)}
Since, N(a)(12) = {(2), (132)}
and (12)N(a) = {(12), (123)}
we find N(a)(12) (12)N(a) or that N(a) is not normal.
Problem 12 : If N is a normal subgroup of order 2, of a group G then show that N
Z(G), then centre of G.
Solution : Let N = {a, e}.
Since e Z(G) (centre being a subgroup contains e) all that we want to show is that
a Z(G)
i.e., ag = ga for all g G
or g–1ag = a for all g G
Let g G be any element then as a N and N is normal, g–1 ag N = {a, e}
g–1ag = a or g –1 ag = e
since g ag = e ag = ge ag = eg a = e, which is not true
–1

we get g–1 ag = a a Z(G)

or N Z(G).
Problem 13 : Show that a subgroup of index 2 in a group G is a normal subgroup
of G.
Solution : Let H be a subgroup of G, with index 2 then number of distinct right
(left) cosets of H is G is 2 and also then G is union of these two right (left) cosets.
Let g G be arbitary.
Case (i) : g H, then Hg = gH (=H)
Hence H is normal.
Case (ii) : g H then gH H, Hg H
Thus Hg and H = He are the two distinct right cosets of H in G and
G = Hg H
Similarly, G = gH H
Hg H = gH H
Hg = gH (as Hg H = gH H = )
H is normal in G.
Remarks : Converse is not true. Indeed H = {1, –1} has index 4 in the Quaternion
group and is normal.
Problem 14 : Let H be a subset of a group G. Let N(H) = {x G|Hx = xH} be the
normalizer of H is G.
We have already shown that N(H) is a subgroup of G. Show
B.A. Part–III 39 Mathematics : Paper I

(i) If H is a subgroup of G then N(H) is the largest subgroup of G in which H is normal.

(ii) If H is a subgroup of G then H is normal in G iff N(H) = G.
(iii) Show by an example, the converse of (ii) fails if H is only a subset of G.
(iv) If H is a subgroup of G and K is a subgroup of N(H) then H is normal subgroup
of HK.
Solution : (i) We show H is normal in N(H).
Since Hh = hH for all h H, we find
h N(H) for all h H
Thus H N(H).
Again by definition of N(H), Hx = xH for all x N(H)
 H is normal in N(H)
To show that N(H) is the largest subgroup of G in which H is normal, suppose K is
any subgroup of G such that H is normal in K.
Then k–1 Hk = H for all k K
Hk = kH for all k K
k N(H) for all k K
K N(H).
(ii) Let H be a normal subgroup of G
then N(H) G (by definition)
Let x G be any element,
then xH = Hx as H normal in G.
x N(H) G N(H)
Hence G = N(H)
Conversely, let G = N(H), H is a subgroup of G (given)
Let h H, g G be any elements
Then g N(H) as N(H) = G
gH = Hg
H is normal in G.
(iii) Consider G = <a> = {e, a, a2, a3}
then G being cyclic is abelian group.
Take H ={a}
thenf H is a subset and not a subgroup of G (e H)
Also N(H) = G as G is abelian.
(iv) Let K be a subgroup of N(H)
then k K k N(H)  Hk = kH
i.e., Hk = kH for all k K
HK = KH
HK is subgroup of N(H)
Note, h H Hh = hH (=H)
B.A. Part–III 40 Mathematics : Paper I

H N(H) Also K N(H)

Again H HK N(H)
Hence H is a subgroup of HK
H is normal subgroup of HK
[a HK a N(H) Ha = aH].

o G
Problem 15 : Let H be normal in G such that o(H) and o  H are co-prime. Show

that H is unique subgroup of G of given order.

o G
Solution : Let o(H) = m, o  H   n , Suppose K is a subgroup of G of order m.

m.m
Then o  HK   , where d = o(H K)
d

Since H is normal, HK  G
Thus o(HK)|o(G)

m m
 m. |m.n  |n
d d

m
d |dn  m|dn
d

m|d as (m, n) = 1
But d|m as H K H
Thus d = m and hence
o(H K) = o(H) = o(K)
H = K.
IV. Self Check Exercise
1. Find order of each element in the group G = {1, 1} under multiplication.
2. Show that a finite cyclic group with three or more elements has even number
of generators.
3. If order of a group G is pq, where p, q are primes, then show that every
proper subgroup of G is cyclic.
4. Let G be a finite group whose order is not divisible by 3. Suppose (ab)3 = a3b3
for all a, b G, then show that G is abelian.
B.A. Part–III 41 Mathematics : Paper I

5. Let H be a subgroup of G and let N  x xHx 1 then show that N is a normal

G

subgroup of G.
6. Show that every subgroup of a cyclic group is normal.
7. Show that intersection of two normal subgroups is a normal subgroup.
8. Every subgroup of an obelian group is normal. Prove that converse is not
true. (Consider Quaternion group).
9. Prove that centre of a group is a normal subgroup.
10. Show that C(H) is a normal subgroup of N(H), where H G.
B.A. PART–III MATHEMATICS : PAPER I
(SEMESTER - V) ABSTRACT ALGEBRA

LESSON NO. 1.4 AUTHOR : DR. CHANCHAL

GROUPS – IV
Objectives
I. Quotient Groups
II. Examples of Quotient Groups
III. Homomorphisms – Isomorphisms
IV. Kernel of Homomorphism
IV.(a) First Theorem of Isomorphism
IV.(b) Second Theorem of Isomorphism
IV.(c) Third Theorem of Isomorphism
V. Self Check Exercise

I. Quotient Groups
Let G be a group of N a normal subgroup of G. Let us collect all the right cosets of N

G
in G and form a set to be denoted by or G/N. Since N is normal in G, product of
N
any two right cosets of N will again be a right coset of N in G, i.e., we have a well

G G
defined binary composition (Prove it). We now show that this set forms a group
N N
under this product as its binary composition.

G G
For Na, Nb  , NaNb = Nab 
N N

G
If Na, Nb, Nc  be any members, then
N
Na(NbNc) = Na(Nbc) = Na(bc) = N(ab)c = NabNc = (NaNb) Nc.

G G G
Again Ne  will act as identity of and for any Na  , Na–1 will be the inverse
N N N

G
of Na. Thus forms a group, called the Quotient group or the factor group of G by
N
N.
42
B.A. Part–III 43 Mathematics : Paper I

It is easy to see that if G is abelian then so would be any of its quotient groups as
NaNb = Nab = Nba = NbNa.
Converse of this result may not hold.

G
Remarks : (i) In , as N is normal, it does not matter whether we use the word right
N
cosets or left cosets, as Na = aN for all a.
Theorem 1 : If G is a finite group and N is a normal subgroup of G then

 G  o G
o  
 N  o N

Proof : Since G is finite, using Lagrange's theorem

o G
o  N  = number of distinct right cosets of N in G

G
 o 
N

Theorem 2 : Every quotient group of a cyclic group is cyclic.

Proof : Let G = <a> be a cyclic group.
Then G is abelian, so every subgroup of G is normal. Let H be any subgroup of G.

G G
We show is cyclic. In fact we claim is generated by Ha.
N N

G
Let Hx  be any element.
N
Then x G = <a>, i.e., x will be some power of a
Let x = am
Then Hx = Ham = Ha a ......a (m times)
= Ha Ha ...... Ha (m times)
= (Ha)n

G G G
i.e., any element Hx of is a power of Ha Ha generates or that is cyclic.
N N N

Remarks : (i) The above result is proved for m > 0. In case m 0, the proof follows
similarly. Notice am = a–n = (a–1)n where n > 0 and remember that Ha–1 = (Ha)–1 and so
(Ha–1)n = (Ha)–n = (Ha)m,
B.A. Part–III 44 Mathematics : Paper I

(ii) If G = <a> is cyclic and H  G, then o(G/H) is the least +ve integer m, s.t.,
amH.
Also, G/H = <Ha>. So o(G/H) = o(Ha) = m
(iii) The converse of above result is not true.
II. Examples of Quotient Groups

a b 
Example 1 : Let G be the set of 2 × 2 matrices over reals of the type 0 d  where
 

ad 0. Then it is easy to see that G will form a group under matrix multiplication

1 b 
a 
1 0  ad  a b 
 
0 1  will be identity,  1  will be inverse of any element 0 d  Also G is not
  
0  
d 

abelian.

1 b 
Let N be the subset containing members of the type 0 1  . Then N is a subgroup
 

of G (Prove). Also it is normal as the product of the type

1 b 
  b
 a b  1 k   a ad  1 akd  bd  
0 d  0 1     d N
   0 1   
  0 1 
d 

G G
So, we get the quotient group . We show is abelian.
N N

G
Let Nx, Ny  be any elements, then x, y G.
N

a b  c e
Let x    ,y   
0 d  0 f 

G
will be abelian iff NxNy = NyNx
N
Nxy = Nyx
xy (yx)–1 N
 xyx–1 y–1  N
B.A. Part–III 45 Mathematics : Paper I

We can easily check that the product

1 b  1 e
 
a b   c e   a ad   c cf  1 t 
 0 d   0 1   
   0 1   1  is a matrix of the type 0 1 .

0  
d   f 

Thus we can have an abelian quotient group.

Example 2 : Let < Z, + > be the group of integers and let N = {3n| n Z}, then N is
a normal subgroup of Z.

Z
will consist of members of the type N + a, a Z
N

Z
We show contains only three elements. Let a  Z be any element, where
N
a 0, 1, 2, then we can write, by division algorithm,
a = 3q + r where 0 r  2
N + a = N + (3q + r) = (N + 3q) + r = N + r as 3q N
but r can take values 0, 1, 2
Hence N + a will be one of N, N + 1, N + 2

Z
or that contains only these three members.
N

Remarks : (i) This example shows that in case of cosets, Ha = Hb may not
necessarily mean a = b. For instance, N + 4 = N + 1, but 4 1 in above example.
[N + 4 = (N + 3) + 1 = N + 1]
(ii) It also serves as an example of an infinite group which has a subgroup N having
finite index in G.

G
Problem 1 : If G is group such that Z  G  is cyclic, where Z(G) is centre of G then

show that G is abelian.

G
Solution : Let us write Z(G) = N. Then is cyclic. Suppose it is generated by Ng.
N
Let a, b G be any two elements,

G
then Na,Nb 
N
B.A. Part–III 46 Mathematics : Paper I

Na = (Ng)n, Nb = (Ng)m for some n, m

Na = Ng. Ng ......Ng = Ngn
Nb = Ngm
ag –n N, bg–m N
ag –n = x, bg–m = y for some x, y N
a = xg n, b = yg m
ab = (xgn)(ygm) = x(gny) gm
= x(ygn)gm as y N = Zn (G)
= xygn g m
= xyg n+m
Similarly, ba = (yg m)(xgn) = y(g mx) gn = y(xgm) gn
= (yx)gm+n
ab = ba as xy = yx as x, y Z(G)
Showing that G is abelian.

G
Remarks : (i) We are talking about Z  G  assuming, therefore, that Z(G) is a normal

subgroup of G, a result which can be proved easily.

(ii) One can, moving on same lines as in the above solution prove that if G/H is
cyclic, where H is a subgroup of Z(g), then G is abelian.
(iii) If G is a non abelian group then G/Z(G) is not cyclic.

G
(iv) If is cyclic for some normal subgroup H of G then G may not be abelian. Take
H

8
G = Quaternion group and H = {1, i}, then o(G/H) =  2 , a prime. So G/H is
4
cyclic, but G is not abelian.
Problem 2 : Let G be a non-abelian group of order pq where p, q are primes then
o(Z(G)) = 1.

G
Solution : Since G is non-abelian, by Problem 1, is not cyclic.
Z G 

Now, o(Z(G))|o(G) = pq
o(Z(G)) = 1, p, q or pq
o(Z(G)) = pq Z(G) = G
G is abelian which is not so.
B.A. Part–III 47 Mathematics : Paper I

pq
o(Z(G)) = p o(G/Z(G)) = p  q , a prime, meaning G/Z(G)) is cyclic which is also not

true.
Similarly, o(Z(G)) = q cannot hold and we are left with the only possiblility that
o(Z(G)) = 1.
III. Homomorphisms – Isomorphisms
Definition : Let < G, * > and < G', o > be two groups.
A mapping f : G G' is called a homomorphism if
f (a * b) = f(a) o f(b) a, b G
We shall use the same symbol (.)for both binary composition.
With that as notation, we find a map f : G  G' is a homomrphism if
f(ab) = f(a) f(b)
If, in addition, f happens to be one -one, onto, we say f is an isomorphism and in
that case, we write G G'.
An onto homomorphism is called epimorphism.
A one-one homomorphism is called monomorphism.
A homomorphism from a group G to itself is called an endomorphism of G.
An isomorphism from a group G to itself is called automorphism of G.
If f : G  G' is onto homomorphism, then G' is called homomorphic image of G.
Example 1 : Let <Z, + > and < E, + > be the groups of integers and even integers.
Define a map f : Z  E, s.t.,
f(x) = 2x for all x Z
then f is well defined as x = y 2x = 2y  f(x) = f(y)
Also f is 1 – 1 (can be proved by taking the steps backwards.)
Now, f is a homomorphism as
f(x + y) = 2(x + y) = 2x + 2y = f(x) + f(y)
Also f is onto as any even integer 2x would have x as its pre-image.
Hence f is an isomorphism.
In fact this example shows that a subset can be isomorphic to its superset.
Example 2 : Let f be a mapping from <Z, + > the group of integers to the group G
= {1, –1} under multiplication defined as
f : Z  G, s.t.,
f(x) = 1 if x is even
= –1 if x is odd
then f is clearly well defined. We check, if it is a homomorphism.
Let x, y  Z be any elements.
B.A. Part–III 48 Mathematics : Paper I

Case (i) : x, y are both even, then x + y is even and as

f(x + y) = 1, f(x) = 1, f(y) = 1
we notice f(x + y) 1 = 1.1 = f(x). f(y)
Case (ii) : x, y are both odd, then x + y is even and
f(x + y) + 1 = (–1)(–1) = f(x) f(y)
Case (iii) : x is odd, y is even, then x + y is odd and
f(x + y) = –1 = (–1)(1) = f(x) f(y)
thus in all case f(x + y) = f(x) f(y)
Showing thereby that f is a homomorphism. Is it an isomorphism?
Obviously, f is onto. But f is not 1 – 1 as f(x) = f(y) does not necessarily mean x = y.
Indeed f(2) = f(4) but 2 4.
Example 3 : Let R+ be the group of positive real numbers under multiplication
and R be the group of all real numbers under addition. Then the map
: R+  R s.t., (x) = log x
is an isomorphism,
 is clearly well defined.
Now, (x) = (y)
log x = log y
 elog x = elog y
x = y
  is, one-one.
Since (xy) = log xy = log x + log y = (x) + (y)
 is homomorphism.
Finally, if y R be any member, then
Since ey R+ and (ey) = y, therefore  is onto and hence on isomorphism (The map f
: R  R + , s.t., f(a) = ea can also be considered).
Example 4 : Let G be a group and N, a normal subgroup of G. Define a map
G
f :G  s.t.,
N
f(x) = Nx, x G
then f is clearly well defined. Again
f(xy) =Nxy = NxNy = f(x) f(y)
shows f is a homomorphism.
It is sometimes called the natural (or canonical) homomorphism. Also, f is onto.
Remark : The relation of isomorphism in group is an equivalence relation. Thus
whenever a group G is isomorphic to another group G', G' will be isomorphic to G.
So we can say that G and G' are isomorphic and denote it by G G'.
B.A. Part–III 49 Mathematics : Paper I

Theorem 1 : If f : G G' is a homomorphism then

(i) f(e) = e'
(ii) f(x–1) = (f(x))–1
(iii) f(xn) = [f(x)]n, n an integer.
where e, e' are identity elements of G and G' respectively.
Proof : (i) We have
e. e = e
f(e. e) = f(e)
f(e). f(e) = f(e)
f(e). f(e) = f(e). e'
f(e) = e' (Left cancellation law)
(ii) Again xx–1 = e = x–1x
f(xx–1) = f(e) = f(x–1x)
f(x) f(x–1) = e' = f(x–1) f(x)
(f(x))–1 = f(x–1)
(iii) Let n be a +ve integer.

f(x.x.........x)
f(xn) = (n times)

= f(x). f(x) ......f(x) (n times)

= (f(x))n
If n = 0, we have the result by (i). Incase n is –ve integer, result follows by using (ii).
Problem 1 : Sh ow th at <Q, +> cannot be isomo rph ic to < Q*, > , w here
Q* = Q – {0} and Q = rationals.
Solution : Suppose f is an isomophism from Q to Q*. Then as 2 Q*. Then as 2 
Q*, f is onto, < Q, + >, s.t., f() = 2

 
f  2
2 2

 
or  f   f    2
2 2


x2 = 2 where x  f    Q *
2

But that is a contradiction as there is no rational no. x.s.t., x2 = 2. Hence the result
follows.
B.A. Part–III 50 Mathematics : Paper I

Z Z
Problem 2 : Find all the homomorphisms from to .
4Z 6Z

Z Z
Solution : Let f :  be a homomorphism.
4Z 6Z

Then f(4Z + n) = n f(4Z + 1)

So, f is completely known if f(4Z + 1) is known.
Now order of (4Z + 1) is 4 and so o(f(4Z + 1)) divides 4
Also o(f(4Z + 1)) divides 6 and thus o(f(4Z + 1)) = 1 or 2

Z
If o(f(4Z + 1)) = 1, then f(4Z + 1) = 6Z = zero of
6Z

Hence f(4Z + n) = zero

If o(f(4Z + 1)) = 2, then f(4Z + 1) = 6Z + 3
f(4Z + n) = 6Z + 3n
Also f(4Z + n + 4Z + m) = f(4Z + n + m)
= 6Z + 3(n + m)
= (6Z 3n) + (2Z + 3m)
= f(4Z + n) + f(4Z + m)
Thus there are two choices for f and it can be defined as

Z Z
f:  s.t.,
4Z 6Z

f(4Z + n) = 6Z + 3n
Notice 4Z + n = 4Z + m
n – m 4Z
3(n – m) 12Z 6Z
3(n – m) 6Z
6Z + 3n 6Z + 3m
i.e., f is well defined.

Z Z
So there are two homomorphisms from  . In fact, in general, there are
4Z 6Z

Z Z
homomorphisms from  where d = g.c.d (m, n)
mZ nZ

IV. Kernel of Homomorphism

Definition : Let f : G G' be a homomorphism. The Kernel of f, (denoted by Ker f)
B.A. Part–III 51 Mathematics : Paper I

is defined by
Ker f = {x G|f(x) = e'}
where e' is identity of G'.
Theorem 2 : If f : G  G' be a homomorphism, then Ker f is a normal subgroup of
G.
Proof : Since f(e) = e', e Ker f, thus Ker f .
Again, x, y  Ker f f(x) = e', f(y) = e1
Now f(xy–1) = f(x) f(y–1) = f(x) (f(y))–1 = e'. e'–1 = e'. e' = e'
xy–1  Ker f
Hence it is a subgroup of G.
Again, for any g G, x Ker f
f(g–1xg) = f(g–1)f(x) f(g)
= (f(g)) –1 f(x) f(g) = (f(g))–1 e' f(g)
= (f(g)) –1 f(g) = e'
g–1 xg Ker f
Hence, it is a normal subgroup of G.
Theorem 3 : A homomorphism f : G  G' is one-one iff Ker f = {e}.
Proof : Let f : G  G' be one-one.
Let x Ker f be any element
then f(x) = e' and as f(e) = e'
f(x) = f(e) x = e as f is 1 – 1
Hence Ker f = {e}.
Conversely, let Ker f contain only the identity element.
Let f(x) = f(y)
Then f(x) (f(y))–1 = e'
f(xy–1) = e'
xy–1 Ker f = {e}
xy–1 = e
x = y or that f is one-one.
Problem 3 : Let f : G  G' be a homomorphism. Let a G be such that o(a) = n and
o(f(a)) = m. Show that o(f(a))|o(a) and f is 1 – 1 iff m = n.
Solution : Since o(a) = n
we find an = e f(an) = f(e)
f(a. a........a) = f(e)
(f(a))n = e'
o(f(a))|n = o(a)
Again, let f be 1 –1
B.A. Part–III 52 Mathematics : Paper I

since o(f(a)) = m
we find (f(a))m = e'
f(a). f(a) .....f(a) = e'
f (a. a.....a) = e'
f(am) = e' = f(e)
am = e (f is 1 – 1)
i.e., o(a)|m or n|m, but already m|n
Hence m = n.
Conversely, let o(a) = o(f(a)).
Then f(x) = f(y)
f(x) (f(y))–1 = e'
f(xy–1) = e'
o(f(x–1)) = 1
o(xy –1) = 1 xy–1 = e x = y
f is 1–1.
Remark: Under an isomorphism, order of any element is preserved.
Problem 4 : Show that the group <R, +> of real numbers cannot be isomorphic to
the group R* of non zero real numbers under multiplicaiton
Solution : –1 R* and order of –1 is 2 as (–1)2 = 1. But R has no element of order
2. As if x R is of order 2 then 2x = x + x = 0. But this does not hold in <R, +> for any
x except x = 0.
By above remark, under an isomorphism order of an element is preserved. Thus
there cannot be any isomorphism between R and R*.
Problem 5 : Let G be a group and f : G  G s.t. f(x) = x–1 be a homomorphism.
Show that G is abelian.
Solution : Let x, y G be any elements.
 xy = (y–1 x–1)-1 = f(y–1x–1)
= f(y–1) f(x–1)
= yx, hence G is abelian.
IV.(a) First Theorem of Isomorphisms
Theorem 4 (Fundamental theorem of group homomorphism) : If f : G 
G
G' be an onto homomorphism with K = Ker f, then  G' .
K
In other words, every homomorphic image of a group G is isomorphic to a quotient
group of G.
B.A. Part–III 53 Mathematics : Paper I

G
Proof : Define a map  :  G ' s.t.,
K

(Ka)  f(a), a  G
We show is an isomorphism.
The  is well defined follows by
Ka = Kb
ab–1 K = Ker f
f(ab–1) = e'
f(a)(f(b))–1 = e'
(Ka) = (Kb)
By retracing the steps backwards, we can prove that is 1 –1.
Again as (KaKb) = (Kab) = f(ab) = f(a) f(b)
= (Ka) (Kb)
Therefore,  is a homomorphism.
To check that is onto, let g' G' be any element. Since f : G G' is onto, g G,
s.t.,
f(g) = g'
Now (Kg) = f(g) = g'
Which shows that Kg is the required pre-image of g' under .
Hence  is an isomorphism.
Remark : The above theorem is also called first theorem of isomorphism. It can
also be stated as:

G
If f : G G' is a homomorphism with K = Ker f, then Ker f  f(G) .

IV.(b) Second Theorem of Isomorphism

Theorem 5 : Let H and K be two subgroup of a group G. where H is normal in G
then

HK K

H HK

Proo f : It is easy to see th at H  K will be a normal subgroup of K as

H  HK G, H will be normal in HK.

HK
Define f :K  s.t., f(k) = Hk
H
As k1 = k2 Hk1 = Hk2 f(k1) = f(k2)
B.A. Part–III 54 Mathematics : Paper I

So, f is well defined.

Again f(k1k2) = Hk1k2 = Hk1 Hk2 = f(k1)f(k2)
Which shows f is a homomorphism.
Now obviously f is onto and thus using Fundamental theorem, we have

HK K

H Ker f

Since k Ker f f(k) = H

Hk = H
K H
k H K (k K as Ker f K)
We find Ker f = H K
Hence Proved.
IV.(c) Third Theorem of Isomorphism
Lemma : If H and K are two normal subgroups of a group G such that H K, then
K G
is a normal subgroup of , and conversely.
H H

K G
Proof : is non empty subset of (by definition)
H H

K
For any Hk1, Hk2 
H

 
(Hk1)(Hk2)–1 =  Hk1  Hk 2  Hk1k 2 
1 1 K
H

K
i.e., is a subgroup.
H

K G
Again, for any Hk  and Hg  , we have
H H
(Hg)–1 (Hk)(Hg) = Hg–1 HkHg

1 K
= Hg kg 
H
as g G, k K, K is normal in G gives g –1 kg K.
The converse part is left as an exercise for the reader.
Theorem 6 : (Third theorem of isomorphism) : If H and K are two normal
B.A. Part–III 55 Mathematics : Paper I

subgroups of G such that H K then

G G /H

K K /H

K G
Proof : The above lemma ensures that is a normal subgroup of and, therefore,
H H

G /H
we can talk of K /H .

G G
Define a map f:  s.t.,
H K
f(Ha) = Ka, a G
f is well defind as
Ha = Hb
ab–1 H K
Ka = Kb
f(Ha) - f(Hb)
f is a homomorphism as
f(HaHb) = f(Hab) = Kab = KaKb = f(Ha) f(Hb).
Obviously, f is onto.
Using Fundamental theorem of group homomorphism, we can say

G G /H

K Ker f

K
We claim Ker f 
H

G
A member of Ker f will be some member of .
H
Now Ha Ker f f(Ha) = K (identity of G/K)
Ka = K
a K

K
 Ha 
H

G G /H
Hence we find 
K K /H
B.A. Part–III 56 Mathematics : Paper I

Which proves our result, it is also called Freshman's theorem.

K K G
Remark : Since = Ker f, we notice that is a normal subgroup of .
H H K

Problem 6 : Let G be the group of all non zero complex numbers under

 a b
multiplication and let G' be the group of all real 2 × 2 matrices of the form  b a  ,
 

where not both a and b are zero, under matrix multiplication, show that G G'.
Solution : Define a map
: G G', s.t.,

 a b
(a + ib) =  b a 
 

 is clearly well-defined,
Also

 ac  bd ad  bc 
[(a + ib) (c + id)] = [(ac – bd) + i(ad + bc)] =  ad  bc ac  bd 
 

and

 a b   c d   ac  bd ad  bc 
(a + ib) (c + id)  b a   d c   ad  bc ac  bd 
   

shows that  is a homomorphism.

 a b
Also for   , the required pre-image is (a + ib).
 b a 

Thus  is onto
Also, (a + ib) = (c + id)

 a b  c d 
  
 b a   d c 

a = c, b = d a + ib = c + id
Hence  is an isomorphism.
Problem 7 : Suppose G is a group of order p 2 , where p is a prime. Let
: G H be an onto homomorphism, where H is a group. Then show that either is
an isomorphism or maps each element x of G onto the identity e of H and H = {e}
or else, for each y H, exactly p elements x of g such that (x) = y.
B.A. Part–III 57 Mathematics : Paper I

Solution : : G H is an onto homomorphism

o(G) = p2.
Since Ker is a subgroup of G, by Lagrange's theorem o(Ker )|o(G) = p2
o(Ker ) = 1, p or p2
Case (i) : o(Ker ) = 1 Ker = {e}
is 1 – 1.
Hence  is an isomorphism.
Case (ii) : o(Ker ) = p2
Ker = G
for all x G, x Ker (x) = e for all x G
Since is onto, each element of H has pre-image, but all members of G are mapped
to e.
 H = {e}
Case (iii) : o(Ker ) = p
Let y H be any element then as is onto, x G, s.t., (x) = y
Let Ker  = {a1 = e, a2, a3, ......ap}
We claim xa1, xa2, ......., xap are distinct.
Suppose xai = xaj. then ai = aj
Which is not true.
Thus xa1, xa2, ............, xap are distinct members of G.
Now (xai) (x)(ai) i = 1, 2, ......., p,
= ye = y for all i (ai Ker )
thus y has p pre-images x = xa1, ax2, .............., xap
To show that y does not have more than p pre-images, let x' be any other pre-image
of y under 
Then (x') = y = (x)
((x))–1 (x') = e
(x–1 x') = e
 x–1x' Ker 
x–1 x' = ai for some i
x' = xai for some i
i.e., it is one of the p pre-images, we have considered.
Hnece y has exactly p pre-images.
V. Self Check Exercise
1. For a fixed element a in a group G, define
f a : G G, s.t., fa(x) = a–1 xa, x G
Show that f a is an isomorphism.
B.A. Part–III 58 Mathematics : Paper I

2. Let f, g be homomorphisms from G  G'. Show that

H = {x G|f(x) = g(x)} is a subgroup of G.
3. Show that the relation of isomorphism in groups is an equivalence
relation.
4. Show that 2Z  3Z by considering the mapping 2x 3x. Generalise.
5. Show that homomorphi image of
(a) an abelian group is abelian
(b) a cyclic group is cyclic.
(c) a finite group is finite.

G
6. Let N be a normal subgroup of G then show that is abelian iff
N
xyx–1 y–1 N, for all x, y G.

G
7. Show that a subgroup H of a group G is normal in G iff the set of all
H
its right cosets is closed under multiplication.
8. If H and K are two normal subgroups of G such that (G/H) and (G/K)

G
are abelian then show that is abelian.
HK

Q
9. Show that is an infinite group and is not cyclic.
Z
10. Let N be a normal subgroup of a group G. Show that o(Na)|o(a) for any
a G.