Homepage › Solution manuals › David S. Dummit › Abstract Algebra › Problem 5.5.7 (Groups of order 56)

Problem 5.5.7 (Groups of order 56)

This exercise describes thirteen isomorphism types of groups of order 56 . (It is not difficult to show that every group of order 56 is isomorphic to one of these.)

(a)
Prove that there are three abelian groups of order 56 .
(b)
Prove that every group of order 56 has either a normal Sylow 2 -subgroup or a normal Sylow 7 -subgroup.
(c)
Construct the following non-abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup S is as specified:
  • one group when S ≃ Z 2 × Z 2 × Z 2
  • two nonisomorphic groups when S ≃ Z 4 × Z 2
  • one group when S ≃ Z 8
  • two nonisomorphic groups when S ≃ Q 8
  • three nonisomorphic groups when S ≃ D 8

[For a particular S , two groups are not isomorphic if the kernels of the maps from S into Aut ( Z 7 ) are not isomorphic.]

(d)
Let G be a group of order 56 with a nonnormal Sylow 7 -subgroup. Prove that if S is the Sylow 2 -subgroup of G then S ≃ Z 2 × Z 2 × Z 2 . [Let an element of order 7 act by conjugation on the seven nonidentity elements of S and deduce that they all have the same order.]

Answers

Proof. Let G be a group of order 56 = 2 3 ⋅ 7 .

(a)
If G is abelian, by Theorem 5 of Chapter 5, there are three isomorphism types of abelian groups of order 2 3 ⋅ 7 : G ≃ Z 2 3 × Z 7 or G ≃ Z 2 2 × Z 2 × Z 7 or or G ≃ Z 2 × Z 2 × Z 2 × Z 7

(b)
By Sylow’s Theorem, if n p denotes the number of Sylow p -subgroups of G , then n 2 ∣ 7 , n 2 ≡ 1 ( 𝑚𝑜𝑑 2 ) , n 7 ∣ 8 , n 7 ≡ 1 ( 𝑚𝑜𝑑 7 ) ,

so

n 2 ∈ { 1 , 7 } , n 7 ∈ { 1 , 8 } .

Assume n 7 = 8 . Since the 7 -Sylow subgroups have trivial intersection, these subgroups contain one element of order 1 and 8 ⋅ 6 = 48 element of order 7 . It remains only 7 elements, so there is at most one Sylow 2 -subgroup of order 8 = 1 + 7 , so n 2 = 1 . Therefore

n 2 = 1 or n 7 = 1 .

Every group of order 56 has either a normal Sylow 2 -subgroup or a normal Sylow 7 -subgroup.

(c)
Suppose here that the non-abelian group G has a normal Sylow 7 -subgroup H ( n 7 = 1 ), and let S be a Sylow 2 -subgroup of G . The order of an element x ∈ H ∩ S divides 7 and 8 , thus x = 1 , so H ∩ S = { 1 } . Since H ⊴ G , 𝐻𝑆 is a subgroup of G , and | 𝐻𝑆 | = | H | | S | ∕ | H ∩ S | = 56 , thus G = 𝐻𝑆 . Since G = 𝐻𝑆 , H ⊴ G , H ∩ S = { 1 } ,

Theorem 12 shows that

G ≃ H ⋊ φ S ,

where the homomorphism φ : S → Aut ( H ) is associated to the action of S on H by conjugation. There are 5 types of groups of order 8 , so we distinguish five cases.

  • If S ≃ Z 2 × Z 2 × Z 2 , then (in additive notations)

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) .

    where ψ : ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ → Aut ( ℤ ∕ 7 ℤ ) is a homomorphism.

    Since Aut ( ℤ ∕ 7 ℤ ) ≃ ( ℤ ∕ 7 ℤ ) × ≃ Z 6 , there is only one element of order 2 in Aut ( ℤ ∕ 7 ℤ ) , given by

    neg { ℤ ∕ 7 ℤ → ℤ ∕ 7 ℤ x ↦ − x .

    The homomorphism ψ maps the elements ( 1 , 0 , 0 ) , ( 0 , 1 , 0 ) , ( 0 , 0 , 1 ) of order 2 on an automorphism of order 1 or 2 , so ψ ( 1 , 0 , 0 ) ∈ { id , neg } , and so on: there exist η 1 , η 2 , η 3 ∈ ℤ ∕ 2 ℤ such that

    ψ ( 1 , 0 , 0 ) = neg η 1 , ψ ( 0 , 1 , 0 ) = neg η 2 , ψ ( 0 , 0 , 1 ) = neg η 3 .

    Then ψ ( a , b , c ) = ψ ( 1 , 0 , 0 ) a ψ ( 0 , 1 , 0 ) b ψ ( 0 , 0 , 1 ) c , where a , b , c ∈ ℤ ∕ 2 ℤ , thus ψ ( a , b , c ) = neg η 1 a + η 2 b + η 3 c , and so

    ψ ( a , b , c ) ( x ) = ( − 1 ) η 1 a + η 2 b + η 3 c x ( x ∈ ℤ ∕ 7 ℤ ) . (1)

    Conversely, if ψ is defined by (1), then for all x ∈ ℤ ∕ 7 ℤ ,

    [ ψ ( a , b , c ) ∘ ψ ( a ′ , b , ′ , c ′ ) ] ( x ) = ( − 1 ) η 1 a + η 2 b + η 3 c ( − 1 ) η 1 a ′ + η 2 b ′ + η 3 c ′ x = ( − 1 ) η 1 ( a + a ′ ) + η 2 ( b + b ′ ) + η 3 ( c + c ′ ) x ,

    so ψ ( a , b , c ) ∘ ψ ( a ′ , b , ′ , c ′ ) = ψ ( a + a ′ , b + b ′ n , c + c ′ ) = ψ ( ( a , b , c ) + ( a ′ , b ′ , c ′ ) ) and ψ is an homomorphism. This shows that there are 8 homomorphisms defined by (1).

    If ( η 1 , η 2 , η 3 ) = ( 0 , 0 , 0 ) then ψ is the trivial homomorphism, and G ≃ H × S is abelian. There are 7 non trivial homomorphisms, for ( η 1 , η 2 , η 3 ) ≠ ( 0 , 0 , 0 ) .

    Consider two such homomorphisms ψ , ψ ′ defined by

    ψ ( a , b , c ) ( x ) = ( − 1 ) η 1 a + η 2 b + η 3 c x (2) ψ ′ ( a , b , c ) ( x ) = ( − 1 ) η 1 ′ a + η 2 ′ b + η 3 ′ c x , (3)

    where ( η 1 , η 2 , η 3 ) ≠ ( 0 , 0 , 0 ) , ( η 1 ′ , η 2 ′ , η 3 ′ ) ≠ ( 0 , 0 , 0 ) .

    We show ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) ≃ ℤ ∕ 7 ℤ ⋊ ψ ′ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) , by proving ψ ′ = ψ ∘ f , where f ∈ Aut ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) ≃ GL 3 ( 𝔽 2 ) (see “Missing Theorem: precomposition by an automorphism” in the Appendix to Exercise 6).

    We can rewrite (2) and (3) in the form

    ψ ( u ) ( x ) = ( − 1 ) η ( u ) x , ψ ′ ( u ) ( x ) = ( − 1 ) η ′ ( u ) x ,

    where u = ( a b c ) ∈ 𝔽 2 3 , and η = 𝔽 2 3 → 𝔽 2 is a linear form of matrix ( η 1 η 2 η 3 ) . Since η ≠ 0 and η ′ ≠ 0 , there is some M ∈ GL 3 ( 𝔽 2 ) such that

    ( η 1 ′ η 2 ′ η 3 ′ ) = ( η 1 η 2 η 3 ) M ,

    so that η ′ = η ∘ f , where f ∈ Aut ( 𝔽 2 3 ) has matrix M . Then ψ ′ ( u ) ( x ) = ( − 1 ) η ′ ( u ) x = ( − 1 ) η ( f ( u ) ) x for all x ∈ ℤ ∕ 7 ℤ , so that ψ ′ = ψ ∘ f .

    This shows ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) ≃ ℤ ∕ 7 ℤ ⋊ ψ ′ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) .

    In conclusion, G ≃ ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) , where ψ is any of the 7 previous non trivial homomorphisms, for instance

    ψ ( a , b , c ) ( x ) = ( − 1 ) a x ( x ∈ ℤ ∕ 7 ℤ ) .

    Then the law on ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ × ℤ ∕ 2 ℤ ) is given by

    ( x , ( a , b , c ) ) ( x ′ , ( a ′ , b ′ , c ′ ) ) = ( x + ( − 1 ) a x ′ , ( a + a ′ , b + b ′ , c + c ′ ) ) ,

    where x , x ′ ∈ ℤ ∕ 7 ℤ , ( a , b , c ) , ( a ′ , b ′ , c ′ ) ∈ ( ℤ ∕ 2 ℤ ) 3 .

    There is only one type of non abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup is isomorphic to Z 2 × Z 2 × Z 2 .

  • If S ≃ Z 4 × Z 2 , then

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) .

    where ψ : ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ → Aut ( ℤ ∕ 7 ℤ ) is a homomorphism. Since ψ ( 1 , 0 ) 4 = 1 , | ψ ( 1 , 0 ) | divides 4 , and so ψ ( 1 , 0 ) ∈ { id , neg } , and similarly since ψ ( 0 , 1 ) 2 = id , ψ ( 0 , 1 ) ∈ { id , neg } :

    ψ ( 1 , 0 ) = neg η 1 , ψ ( 0 , 1 ) = neg η 2 , ( η 1 , η 2 ∈ { 0 , 1 } ) .

    If a ¯ = [ a ] 4 ∈ ℤ ∕ 4 ℤ and b ¯ = [ b ] 2 ∈ ℤ ∕ 2 ℤ , where a , b ∈ ℤ , then

    ψ ( a ¯ , b ¯ ) = ψ ( 1 , 0 ) a ψ ( 0 , 1 ) b = neg η 1 a + η 2 b ,

    so, for all x ∈ ℤ ∕ 7 ℤ ,

    ψ ( a ¯ , b ¯ ) ( x ) = ( − 1 ) η 1 a + η 2 b x . (4)

    As in the previous item, every map ψ : Z 4 × Z 2 → Aut ( ℤ ∕ 7 ℤ ) defined by (4) is a homomorphism.

    Therefore there are exactly 3 non trivial homomorphisms ψ : ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ → Aut ( ℤ ∕ 7 ℤ ) , defined by

    ψ 1 ( a ¯ , b ¯ ) ( x ) = ( − 1 ) a x , ψ 2 ( a ¯ , b ¯ ) ( x ) = ( − 1 ) b x , ψ 3 ( a ¯ , b ¯ ) ( x ) = ( − 1 ) a + b x .

    Consider the automorphism r ∈ Aut ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) (see the appendix 1) characterized by

    r { α ↦ β β ↦ − α

    where α = ( 1 , 0 ) and β = ( 1 , 1 ) are generators of ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ .

    Since ( a ¯ , b ¯ ) = ( a − b ) ( 1 , 0 ) + b ( 1 , 1 ) = ( a − b ) α + 𝑏𝛽 ,

    r ( a ¯ , b ¯ ) = ( a − b ) r ( α ) + b r ( β ) = − ( a − b ) β − 𝑏𝛼 = ( a − b ) ( 1 , 1 ) − b ( 1 , 0 ) = ( − a ¯ + 2 b ¯ , − a ¯ + b ¯ ) = ( a ¯ − 2 b ¯ , a ¯ − b ¯ ) = ( a ¯ + 2 b ¯ , a ¯ + b ¯ ) .

    Then, for all ( a ¯ , b ¯ ) ∈ ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ , and for all x ∈ ℤ ∕ 7 ℤ ,

    ψ 2 ( r ( a ¯ , b ¯ ) ) ( x ) = ψ 2 ( a + 2 b ¯ , a + b ¯ ) ( x ) = ( − 1 ) a + b x = ψ 3 ( a ¯ , b ¯ ) ( x ) .

    This gives

    ψ 2 = ψ 3 ∘ r , where  r ∈ Aut ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) .

    By the missing Theorem “precomposition by an automorphism” in the Appendix to Exercise 6,

    ℤ ∕ 7 ℤ ⋊ ψ 2 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) ≃ ℤ ∕ 7 ℤ ⋊ ψ 3 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) .

    Therefore

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ 1 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ )  or  G ≃ ℤ ∕ 7 ℤ ⋊ ψ 2 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) ,

    where

    ψ 1 ( a ¯ , b ¯ ) ( x ) = ( − 1 ) a x , ψ 2 ( a ¯ , b ¯ ) ( x ) = ( − 1 ) b x ,

    The laws of these two groups are defined by

    ( x , ( a , b ) ) ( x ′ , ( a ′ , b ′ ) ) = ( x + ( − 1 ) a x ′ , ( a + a ′ , b + b ′ ) ) , ( x , ( a , b ) ) ( x ′ , ( a ′ , b ′ ) ) = ( x + ( − 1 ) b x ′ , ( a + a ′ , b + b ′ ) )

    ( x , x ′ ∈ ℤ ∕ 7 ℤ , a , a ′ ∈ ℤ ∕ 4 ℤ , b , b ′ ∈ ℤ ∕ 2 ℤ ) .

    It remains to verify that these two groups are not isomorphic, by computing the kernels of ψ 1 and ψ 2 . If ( a ¯ , b ¯ ) ∈ ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ,

    ψ 1 ( a ¯ , b ¯ ) = id ⟺ ∀ ⁡ x ∈ ℤ ∕ 7 ℤ , ( − 1 ) a x = x ⟺ ( − 1 ) a = 1 ⟺ a ¯ ∈ { 0 , 2 } ⟺ ( a ¯ , b ¯ ) ∈ { ( 0 , 0 ) , ( 0 , 1 ) , ( 2 , 0 ) , ( 2 , 2 ) } .

    So

    ker ⁡ ( ψ 1 ) = { ( 0 , 0 ) , ( 0 , 1 ) , ( 2 , 0 ) , ( 2 , 2 ) } ,

    Since all elements have order 2 ,

    ker ⁡ ( ψ 1 ) ≃ Z 2 × Z 2 .

    Similarly,

    ψ 2 ( a ¯ , b ¯ ) = id ⟺ ∀ ⁡ x ∈ ℤ ∕ 7 ℤ , ( − 1 ) b x = x ⟺ ( − 1 ) b = 1 ⟺ b ¯ = 0 ⟺ ( a ¯ , b ¯ ) ∈ { ( 0 , 0 ) , ( 1 , 0 ) , ( 2 , 0 ) , ( 3 , 0 ) } = ⟨ ( 1 , 0 ) ⟩ .

    Therefore

    ker ⁡ ( ψ 2 ) ≃ ℤ 4 .

    Since ker ⁡ ( ψ 1 ) and ker ⁡ ( ψ 2 ) are not isomorphic, ℤ ∕ 7 ℤ ⋊ ψ 1 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) and ℤ ∕ 7 ℤ ⋊ ψ 2 ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) are not isomorphic by Theorem A given in the Appendix 2.

    There are two types of non abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup is isomorphic to Z 4 × Z 2 .

  • If S ≃ Z 8 , then

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ ℤ ∕ 8 ℤ .

    where ψ : ℤ ∕ 8 ℤ → Aut ( ℤ ∕ 7 ℤ ) ≃ Z 6 is a homomorphism.

    Since 1 has order 8 in ℤ ∕ 8 ℤ , then ψ ( 1 ) 8 = id , thus ψ ( 1 ) 2 = id and ψ ( 1 ) ∈ { id , neg } . Therefore ψ ( 1 ) = neg η for some η ∈ { 0 , 1 } , and for all x ∈ ℤ ∕ 7 ℤ ,

    ψ ( a ¯ ) ( x ) = ( − 1 ) 𝜂𝑎 x , ( a ∈ ℤ , η ∈ { 0 , 1 } ) .

    There is only one non trivial homomorphism ψ 0 , given by

    ψ 0 ( a ¯ ) ( x ) = ( − 1 ) a x .

    Therefore

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ 0 ℤ ∕ 8 ℤ ,

    where ψ 0 is given by

    ψ 0 ( a ¯ ) ( x ) = ( − 1 ) a x .

    The law on ℤ ∕ 7 ℤ ⋊ ψ 0 ℤ ∕ 8 ℤ is defined by

    ( x , y ) ( x ′ , y ′ ) = ( x + ( − 1 ) y x ′ , y + y ′ ) .

    There is only one type of non abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup is isomorphic to Z 8 .

  • If S ≃ Q 8 , then

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ Q 8 .

    where ψ : Q 8 → Aut ( ℤ ∕ 7 ℤ ) ≃ Z 6 is a homomorphism.

    By Exercise 6.3.7,

    Q 8 = ⟨ i , j ∣ i 2 = j 2 , i − 1 𝑗𝑖 = j − 1 ⟩ .

    Since i 4 = j 4 = 1 , then ψ ( i ) 4 = ψ ( j ) 4 = 1 , thus ψ ( i ) ∈ { id , neg } and ψ ( j ) ∈ { id , neg } , so ψ ( i ) = neg η 1 and ψ ( j ) = neg η 2 , for some η 1 , η 2 ∈ { 0 , 1 } .

    Conversely, if η 1 , η 2 ∈ { 0 , 1 } , then I = neg η 1 and J = neg η 2 satisfy I 2 = J 2 ( = id ) and I − 1 𝐽𝐼 = neg − η 1 neg η 2 neg η 1 = neg η 2 = J = J − 1 . Therefore I and J satisfy the relations of the presentation of Q 8 , so there is a homomorphism ψ : Q 8 → Aut ( ℤ ∕ 7 ℤ ) such that ψ ( i ) = I = neg η 1 and ψ ( j ) = J = neg η 2 .

    In conclusion, there are four homomorphisms ψ : Q 8 → Aut ( ℤ ∕ 7 ℤ ) characterized by

    { ψ 0 ( i ) ( x ) = x ψ 0 ( j ) ( x ) = x { ψ 1 ( i ) ( x ) = − x ψ 1 ( j ) ( x ) = x { ψ 2 ( i ) ( x ) = x ψ 2 ( j ) ( x ) = − x { ψ 3 ( i ) ( x ) = − x ψ 3 ( j ) ( x ) = − x

    for all x ∈ ℤ ∕ 7 ℤ .

    Then ψ 0 is the trivial homomorphism, and the corresponding group G 0 is the direct product G 0 ≃ ℤ ∕ 7 ℤ × Q 8 , which is not abelian.

    Explicitly, ψ 1 , ψ 2 , ψ 3 are given by

    y 1 − 1 i − i j − j k − k
    ψ 1 ( y ) id id neg neg id id neg neg

    x 1 − 1 i − i j − j k − k
    ψ 2 ( x ) id id id id neg neg neg neg

    x 1 − 1 i − i j − j k − k
    ψ 3 ( x ) id id neg neg neg neg id id

    By the presentation of Q 8 , since j and i in place of i and j satisfy the relations of the presentation of Q 8 , there is a unique automorphism f ∈ Aut ( Q 8 ) such that f ( i ) = j and f ( j ) = i . Then

    ( ψ 1 ∘ f ) ( i ) = ψ 1 ( j ) = id = ψ 2 ( i ) , ( ψ 1 ∘ f ) ( j ) = ψ 1 ( i ) = neg = ψ 2 ( j ) .

    Therefore ψ 2 = ψ 1 ∘ f , and so, by the missing Theorem “precomposition by an automorphism” in the Appendix to Exercise 6,

    ℤ ∕ 7 ℤ ⋊ ψ 1 Q 8 ≃ ℤ ∕ 7 ℤ ⋊ ψ 2 Q 8 .

    Similarly, there exists a unique g ∈ Aut ( Q 8 ) such that g ( i ) = k and g ( j ) = i . Then

    ( ψ 1 ∘ g ) ( i ) = ψ 1 ( k ) = neg = ψ 3 ( i ) , ( ψ 1 ∘ g ) ( j ) = ψ 1 ( i ) = neg = ψ 3 ( j ) .

    Therefore ψ 3 = ψ 1 ∘ g , and so

    ℤ ∕ 7 ℤ ⋊ ψ 1 Q 8 ≃ ℤ ∕ 7 ℤ ⋊ ψ 3 Q 8 .

    Moreover, G 0 ≄ G 1 since the kernel of the trivial homomorphism ψ 0 is Q 8 , and ker ⁡ ( ψ 1 ) = { 1 , − 1 , j , − j } ≄ Q 8 .

    The laws are defined by

    G 0 : ( a , x ) ( a ′ , x ′ ) = ( a + a ′ , x x ′ ) , G 1 : ( a , x ) ( a ′ , x ′ ) = ( a + 𝜀 a ′ , x x ′ ) , 𝜀 = − 1  if  x ∈ { ± i , ± k } , 𝜀 = 1  otherwise .

    There are two types of non abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup is isomorphic to Q 8 .

  • If S ≃ D 8 , then

    G ≃ ℤ ∕ 7 ℤ ⋊ ψ D 8 .

    where ψ : D 8 → Aut ( ℤ ∕ 7 ℤ ) ≃ Z 6 is a homomorphism.

    By definition, D 8 = ⟨ r , s ⟩ , where r 4 = s 2 = 1 and 𝑠𝑟 s − 1 = r − 1 , and every element x ∈ D 8 is of the form x = r k s l for some k ∈ { 0 , 1 , 2 , 3 } and some l ∈ { 0 , 1 } .

    Since ψ is a homomorphism, ψ ( r ) 4 = id and ψ ( r ) 2 = id , thus ψ ( r ) ∈ { id , neg } and ψ ( r ) ∈ { id , neg } , so ψ ( r ) = neg η 1 and ψ ( s ) = neg η 2 , for some η 1 , η 2 ∈ { 0 , 1 } . For all x ∈ ℤ ∕ 7 ℤ ,

    ψ ( r ) ( x ) = ( − 1 ) η 1 x , ψ ( s ) ( x ) = ( − 1 ) η 2 x ,

    therefore, for all x ∈ ℤ ∕ 7 ℤ , and for 0 ≤ k < 4 , 0 ≤ l < 2 ,

    ψ ( r k s l ) ( x ) = ( − 1 ) η 1 k + η 2 l x ( η 1 , η 2 ∈ { 0 , 1 } ) . (5)

    Conversely, if ψ is defined by (5) for some η 1 , η 2 ∈ { 0 , 1 } , then ψ : ℤ ∕ 8 ℤ → Aut ( ℤ ∕ 7 ℤ ) is a homomorphism: using ( s l r k ) ( s l ′ r k ′ ) = r k + ( − 1 ) l k ′ s l + l ′ , we obtain for x ∈ ℤ ∕ 7 ℤ ,

    ψ ( r k s l ) ( r k ′ s l ′ ) ( x ) = ψ ( r k + ( − 1 ) l k ′ s l + l ′ ) x = ( − 1 ) η 1 ( k + ( − 1 ) l k ′ ) + η 2 ( l + l ′ ) x [ ψ ( r k s l ) ψ ( r k ′ s l ′ ) ] ( x ) = ( − 1 ) η 1 ( k + k ′ ) + η 2 ( l + l ′ ) x .

    Since ( − 1 ) η 1 ( − 1 ) l k ′ = ( − 1 ) η 1 k ′ (because ( − 1 ) u = ( − 1 ) − u for all u ), we obtain ψ ( r k s l ) ( r k ′ s l ′ ) = ψ ( r k s l ) ψ ( r k ′ s l ′ ) , so ψ is a homomorphism.

    (Alternatively, this is a consequence of the presentation of D 8 .)

    Therefore there are 4 homomorphisms ψ : ℤ ∕ 8 ℤ → Aut ( ℤ ∕ 7 ℤ ) , defined, for x ∈ ℤ ∕ 7 ℤ , by

    ψ 0 ( r k s l ) ( x ) = x , ψ 1 ( r k s l ) ( x ) = ( − 1 ) k x , ψ 2 ( r k s l ) ( x ) = ( − 1 ) l x , ψ 3 ( r k s l ) ( x ) = ( − 1 ) k + l x .

    Since ψ 0 is the trivial homomorphism, the corresponding group G 0 = ℤ ∕ 7 ℤ ⋊ ψ 0 D 8 is the non abelian direct product ℤ ∕ 7 ℤ × D 8 .

    Since D 8 ≃ ℤ ∕ 4 ℤ ⋊ ℤ ∕ 2 ℤ , we can write the laws of these tree groups G 1 , G 2 , G 3 in the form

    G 1 : ( x , k , l ) ( x ′ , k ′ , l ′ ) = ( x + ( − 1 ) k x ′ , k + ( − 1 ) l k ′ , l + l ′ ) , G 2 : ( x , k , l ) ( x ′ , k ′ , l ′ ) = ( x + ( − 1 ) l x ′ , k + ( − 1 ) l k ′ , l + l ′ ) , G 3 : ( x , k , l ) ( x ′ , k ′ , l ′ ) = ( x + ( − 1 ) k + l x ′ , k + ( − 1 ) l k ′ , l + l ′ ) ,

    where ( x , k , l ) , ( x ′ , k ′ , l ′ ) ∈ ℤ ∕ 7 ℤ × ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ .

    We compute the kernels of ψ 1 , ψ 2 , ψ 3 :

    r k s l ∈ ker ⁡ ψ 1 ⟺ ∀ ⁡ x ∈ ℤ ∕ 7 ℤ , ( − 1 ) k x = x ⟺ ( − 1 ) k = 1 ⟺ r k s l ∈ { 1 , s , r 2 , r 2 s } .

    So

    ker ⁡ ( ψ 1 ) = { 1 , s , r 2 , r 2 s } ≃ Z 2 × Z 2 .

    Similarly

    r k s l ∈ ker ⁡ ψ 2 ⟺ ∀ ⁡ x ∈ ℤ ∕ 7 ℤ , ( − 1 ) l x = x ⟺ ( − 1 ) l = 1 ⟺ r k s l ∈ { 1 , r , r 2 , r 3 } .

    So

    ker ⁡ ( ψ 2 ) = ⟨ r ⟩ ≃ Z 4 .

    Finally,

    r k s l ∈ ker ⁡ ψ 3 ⟺ ∀ ⁡ x ∈ ℤ ∕ 7 ℤ , ( − 1 ) k + l x = x ⟺ ( − 1 ) k + l = 1 ⟺ r k s l ∈ { 1 , 𝑟𝑠 , r 2 , r 3 s } .

    So

    ker ⁡ ( ψ 3 ) = { 1 , 𝑟𝑠 , r 2 , r 3 s } ≃ Z 2 × Z 2 .

    This shows that ℤ ∕ 7 ℤ ⋊ ψ 1 D 8 is not isomorphic to ℤ ∕ 7 ℤ ⋊ ψ 2 D 8 , but we cannot conclude for ψ 3 . Note that ψ 1 and ψ 3 are characterized by

    ψ 1 ( r ) ( x ) = − x , ψ 1 ( s ) ( x ) = x , ψ 3 ( r ) ( x ) = − x , ψ 3 ( s ) ( x ) = − x ( x ∈ ℤ ∕ 7 ℤ ) .

    Since

    r 4 = ( 𝑟𝑠 ) 2 = 1 , ( 𝑟𝑠 ) r ( 𝑟𝑠 ) − 1 = r ( 𝑠𝑟 s − 1 ) r − 1 = r r − 1 r − 1 = r − 1 ,

    the ordered pair ( r , 𝑟𝑠 ) satisfy the relation of the presentation of D 8 :

    D 8 = ⟨ r , s ∣ r 4 = s 2 , 𝑠𝑟 s − 1 = r − 1 } .

    Therefore there is a unique homomorphism f : D 8 → D 8 such that f ( r ) = r and f ( s ) = 𝑟𝑠 . Since D 8 = ⟨ r , 𝑟𝑠 ⟩ , f : D 8 → D 8 is surjective, where D 8 is finite, so f is an automorphism. Moreover, for all x ∈ ℤ ∕ 7 ℤ ,

    ( ψ 1 ∘ f ) ( r ) ( x ) = ψ 1 ( r ) ( x ) = − x , ( ψ 1 ∘ f ) ( s ) ( x ) = ψ 1 ( 𝑟𝑠 ) ( x ) = ψ 1 ( r ) ( ψ 1 ( s ) ) ( x ) ) = − x .

    Therefore ψ 1 ∘ f = ψ 3 , where f ∈ Aut ( D 8 ) , and so, by the precomposition by an automorphism in the Appendix to Exercise 6,

    ℤ ∕ 7 ℤ ⋊ ψ 1 D 8 ≃ ℤ ∕ 7 ℤ ⋊ ψ 3 D 8 .

    Moreover, since ψ 0 is the trivial homomorphism, with kernel D 8 , G 0 ≃ ℤ ∕ 7 ℤ × D 8 is not isomorphic to G 1 , nor to G 2 .

    There are three types of non abelian groups of order 56 which have a normal Sylow 7 -subgroup and whose Sylow 2 -subgroup is isomorphic to D 8 .

(d)
Let G be a group of order 56 with a nonnormal Sylow 7 -subgroup. By part (b), the Sylow 2 -subgroup S is normal in G ( n 2 = 1 ). Since H ∩ S = { 1 } , we obtain G ≃ S ⋊ ψ H , H ≃ Z 7 ,

where ψ : H → Aut ( S ) is a homomorphism. Since | ker ⁡ ( ψ ) | divides | H | = 7 , | ker ⁡ ( ψ ) | = 1 or | ker ⁡ ( ψ ) | = 7 . If | ker ⁡ ( ψ ) | = 7 , then ψ is the trivial homomorphism, thus G ≃ S × H , and H ⊴ G , which is in contradiction with the hypothesis. Therefore

ker ⁡ ( ψ ) = { 1 } .

Since ψ is injective, H ≃ im ( ψ ) ≤ Aut ( S ) , therefore

7 ∣ | Aut ( S ) | .

The group S of order 8 is isomorphic to Z 2 × Z 2 × Z 2 , Z 4 × Z 2 , Z 8 , D 8 or Q 8 . The following array gives the order of | Aut ( S ) | in each case:

S Z 2 × Z 2 × Z 2 Z 4 × Z 2 Z 8 D 8 Q 8
| Aut ( S ) | 168 = 7 ⋅ 6 ⋅ 4 8 4 8 24

Since 7 divides | Aut ( S ) | , the only possibility is

S ≃ Z 2 × Z 2 × Z 2 .

(e)
By part (d) (in additive notations), G ≃ ( ℤ ∕ 2 ℤ ) 3 ⋊ φ ℤ ∕ 7 ℤ ,

where φ : ℤ ∕ 7 ℤ → Aut ( ( ℤ ∕ 2 ℤ ) 3 ) ≃ GL 3 ( 𝔽 2 ) is a homomorphism.

Such a homomorphism is characterized by the image φ ( 1 ) of 1 ∈ ℤ ∕ 7 ℤ , of order 7 . So φ ( 1 ) 7 corresponds to a matrix A ∈ GL 3 ( 𝔽 2 ) which satisfies A 7 = I . The identity I has order 1 , and there are 48 elements of order 7 in GL 3 ( 𝔽 2 ) (see the sagemath computation below), one of them being

A = ( 0 1 0 0 0 1 1 1 0 ) .

     sage: liste= []
     sage: G = GL(3, GF(2))
     sage: G.order()
     168
     sage: I = G(1);
     sage: for A in G:
     ....:     if A^7 == I:
     ....:         liste.append(A)
     ....:
     sage: liste
     [
     [1 0 0]  [0 1 0]  [0 0 1]  [1 1 1]  [1 1 0]             [1 1 1]
     [0 1 0]  [0 0 1]  [1 0 0]  [0 1 1]  [1 0 1]    (...)    [0 1 1]
     [0 0 1], [1 1 0], [0 1 1], [1 1 0], [0 1 0],            [1 0 1]
     ]
     sage: len(l)
     49

Thus there are exactly 49 homomorphisms φ : ℤ ∕ 7 ℤ → Aut ( ( ℤ ∕ 2 ℤ ) 3 ) ≃ GL 3 ( 𝔽 2 ) .

If φ is the trivial homomorphism corresponding to I , then G ≃ S × H , so H ⊴ G , which is in contradiction with the hypothesis. Each of the 48 matrices of order 2 gives a non trivial homomorphism, and the group G ≃ ( ℤ ∕ 2 ℤ ) 3 ⋊ φ ℤ ∕ 7 ℤ is non abelian.

In particular, the group G A , whose law is given, for ( a b c ) , ( a ′ b ′ c ′ ) ∈ 𝔽 2 3 and x , x ′ ∈ ℤ ∕ 7 Z by

( ( a b c ) , x ) ( ( a ′ b ′ c ′ ) , x ′ ) = ( ( a b c ) + ( 0 1 0 0 0 1 1 1 0 ) x ( a ′ b ′ c ′ ) , x x ′ )

is a group of order 56 such that the Sylow 7 -Sylow subgroup is not normal. Therefore this group is not isomorphic to the previous groups.

If A is replaced by another matrix B ∈ GL 3 ( 𝔽 2 ) of order 7 , we show that G A ≃ G B . The corresponding homomorphisms φ A and φ B are given by

φ A { ℤ ∕ 7 ℤ → Aut ( 𝔽 2 3 ) x ¯ ↦ φ A ( x ¯ ) { 𝔽 2 3 → 𝔽 2 3 ( a b c ) ↦ A x ( a b c ) φ B { ℤ ∕ 7 ℤ → Aut ( 𝔽 2 3 ) x ¯ ↦ φ B ( x ¯ ) { 𝔽 2 3 → 𝔽 2 3 ( a b c ) ↦ B x ( a b c )

(Since A 7 = B 7 = I , the maps φ A ( x ¯ ) and φ A ( x ¯ ) are well defined and are automorphisms of 𝔽 2 3 .)

The 48 elements of order 7 in G are in two conjugacy classes of GL 3 ( 𝔽 2 ) , the classes of

A = ( 0 1 0 0 0 1 1 1 0 ) and A 1 = ( 0 0 1 1 0 0 0 1 1 ) .

     sage: G = GL(3, GF(2))
     sage: A = matrix([[0,1,0], [0,0,1], [1,1,0]])
     sage: A1 = matrix([[0,0,1], [1,0,0], [0,1,1]])
     sage: A = G(A); A1 = G(A1); A, A1
     
     (
     [0 1 0]  [0 0 1]
     [0 0 1]  [1 0 0]
     [1 1 0], [0 1 1]
     )
     sage: A.conjugacy_class().list()
     
     [
     [0 1 0]  [0 1 1]  [0 1 0]  [1 0 1]  [0 1 1]  [1 1 1]  [1 1 1]  [1 1 0]
     [0 0 1]  [0 0 1]  [1 0 1]  [1 1 1]  [1 1 0]  [1 0 1]  [1 0 0]  [1 1 1]
     [1 1 0], [1 0 0], [1 0 0], [1 1 0], [1 1 1], [0 1 1], [1 0 1], [0 1 0],
     
     [0 1 0]  [0 0 1]  [0 0 1]  [1 0 1]  [0 0 1]  [1 1 0]  [1 1 1]  [0 1 0]
     [1 1 1]  [0 1 1]  [1 0 1]  [0 0 1]  [1 0 0]  [0 0 1]  [1 1 0]  [0 1 1]
     [0 1 1], [1 1 1], [0 1 0], [1 1 1], [1 1 0], [1 0 1], [1 0 0], [1 0 1],
     
     [1 1 0]  [0 1 1]  [1 0 1]  [1 1 1]  [0 0 1]  [0 1 1]  [1 1 0]  [1 0 1]
     [0 1 1]  [1 0 0]  [1 1 0]  [0 1 1]  [1 1 0]  [1 1 1]  [1 0 1]  [1 0 0]
     [1 0 0], [0 1 0], [0 1 0], [1 1 0], [0 1 1], [1 0 1], [1 1 1], [0 1 1]
     ]
     sage: A1.conjugacy_class().list()
     
     [
     [0 0 1]  [1 1 1]  [1 0 1]  [0 0 1]  [1 0 1]  [0 1 0]  [1 0 1]  [0 1 1]
     [1 0 0]  [1 1 0]  [1 0 0]  [1 1 0]  [1 1 1]  [1 0 1]  [1 1 0]  [1 0 0]
     [0 1 1], [0 1 1], [0 1 0], [0 1 0], [0 1 1], [0 1 1], [1 1 1], [1 0 1],
     
     [1 0 1]  [1 1 0]  [1 1 0]  [0 1 1]  [1 1 1]  [0 0 1]  [0 1 0]  [0 1 1]
     [0 0 1]  [1 0 1]  [0 0 1]  [1 1 0]  [1 0 0]  [0 1 1]  [0 1 1]  [1 1 1]
     [1 1 0], [0 1 0], [1 0 0], [1 0 0], [1 1 0], [1 1 0], [1 0 0], [0 1 0],
     
     [0 1 0]  [0 0 1]  [1 1 0]  [0 1 1]  [1 1 0]  [1 1 1]  [0 1 0]  [1 1 1]
     [0 0 1]  [1 0 1]  [1 1 1]  [0 0 1]  [0 1 1]  [1 0 1]  [1 1 1]  [0 1 1]
     [1 0 1], [1 1 1], [1 0 1], [1 1 1], [1 1 1], [1 0 0], [1 1 0], [1 0 1]
     ]
                                                                  

                                                                  
     

Note that if Q = ( 0 0 1 0 1 1 1 0 0 ) , then

A 1 3 = ( 1 1 1 0 1 1 1 0 1 ) = ( 0 0 1 0 1 1 1 0 0 ) ( 0 1 0 0 0 1 1 1 0 ) ( 0 0 1 0 1 1 1 0 0 ) − 1 = 𝑄𝐴 Q − 1 ,

so

A 1 3 = 𝑄𝐴 Q − 1 .

Let 𝒞 A and 𝒞 A 1 denote the conjugacy classes of A and A 1 .

  • If B ∈ 𝒞 A , then B = 𝑃𝐴 P − 1 for some P ∈ GL 3 ( 𝔽 2 ) .
  • If B ∈ 𝒞 A 1 , then B = R A 1 R − 1 for some R ∈ GL 3 ( 𝔽 2 ) . Therefore

    B 3 = R A 1 3 R − 1 = R ( 𝑄𝐴 Q − 1 ) R − 1 = ( 𝑅𝑄 ) A ( 𝑅𝑄 ) − 1 = 𝑃𝑄 P − 1 where  P = 𝑅𝑄 .

In both cases there is some integer k ( k = 1 or k = 3 ) and some P ∈ GL 3 ( 𝔽 2 ) such that

𝑃𝐴 P − 1 = B k .

If σ ∈ Aut ( 𝔽 2 3 ) is defined by σ ( X ) = 𝑃𝑋 , then for all X ∈ 𝔽 2 3 (where X ′ = 𝑃𝑋 = σ ( X ) ), and all x ¯ ∈ ℤ ∕ 7 ℤ ,

( φ B k ( x ¯ ) ∘ σ ) ( X ) = φ B k ( x ¯ ) ( X ′ ) = B 𝑘𝑥 X ′ = P A x P − 1 X ′ = P A x X = ( σ ∘ φ A ( x ¯ ) ) ( X ) .

Therefore φ B k ( x ¯ ) ∘ σ = σ ∘ φ A ( x ¯ ) for all x ¯ ∈ K = ℤ ∕ 7 ℤ , thus

φ B k ( K ) = σ φ A ( K ) σ − 1 ,

where K = ℤ ∕ 7 ℤ is cyclic.

By Exercise 6,

( ℤ ∕ 2 ℤ ) 3 ⋊ φ A ℤ ∕ 7 ℤ ≃ ( ℤ ∕ 2 ℤ ) 3 ⋊ φ B ℤ ∕ 7 ℤ .

There is a unique group of order 56 with a nonnormal Sylow 7 -subgroup, up to isomorphism.

We have proved that there are exactly 13 types of groups of order 56 . □

APPENDIX 1. Automorphisms of ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ .

Let S = ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ . Then

S = { ( 0 , 0 ) , ( 0 , 1 ) , ( 1 , 0 ) , ( 1 , 1 ) , ( 2 , 0 ) , ( 2 , 1 ) , ( − 1 , 0 ) , ( − 1 , 1 ) } .

The elements of order 4 in S are

α = ( 1 , 0 ) , − α = ( − 1 , 0 ) , β = ( 1 , 1 ) , − β = ( − 1 , 1 ) .

Note that ⟨ α , β ⟩ contains five distinct elements 0 , α , − α , β , − β , and so, by Lagrange’s Theorem,

S = ⟨ α , β ⟩ .

Let f ∈ Aut ( S ) . Since α , β are distinct non opposite elements of order 4 , it is the same for f ( α ) , f ( β ) . Therefore there are 8 possibilities for f ( α ) , f ( β ) , given by the following array:

f ( α ) α α − α − α β β − β − β f ( β ) β − β β − β α − α α − α (6)

This shows that there are at most 8 automorphisms.

Conversely, if γ , δ are distinct non opposite elements of order 4 , there is a homomorphism f characterized by f ( α ) = γ , f ( β ) = δ . Indeed, if ( a ¯ , b ¯ ) ∈ S , where a , b ∈ ℤ , then ( a ¯ , b ¯ ) = ( a − b ) ( 1 , 0 ) + b ( 1 , 1 ) = ( a − b ) α + 𝑏𝛽 . We define

f ( a ¯ , b ¯ ) = ( a − b ) γ + 𝑏𝛿

( f is well defined, because γ and δ are of order 4 ).

Then, for all ( a ¯ , b ¯ ) , ( c ¯ , d ¯ ) ∈ S ,

f ( a ¯ , b ¯ ) + f ( c ¯ , d ¯ ) = ( ( a − b ) γ + 𝑏𝛿 ) + ( ( c − d ) γ + 𝑑𝛿 ) = ( ( a + c ) − ( b + d ) ) γ + ( b + d ) δ = f ( a + c ¯ , b + d ¯ ) = f ( ( a ¯ , b ¯ ) + ( c ¯ , d ¯ ) ) .

So f is a homomorphism, and f ( α ) = γ , f ( β ) = δ .

In conclusion, there are exactly 8 automorphisms of ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ , characterized by the array (3).

Among these homomorphisms, consider r and s defined by

r { α ↦ β β ↦ − α s { α ↦ α β ↦ − β

Then Aut ( S ) = ⟨ r , s ⟩ , because ⟨ r , s ⟩ contains 5 distinct automorphisms id , r , r 2 , r 3 , s .

Moreover r 4 = s 2 = 1 , and

( 𝑠𝑟 s − 1 ) ( α ) = ( 𝑠𝑟 ) ( α ) = s ( β ) = − β = r − 1 ( α ) ( 𝑠𝑟 s − 1 ) ( β ) = ( 𝑠𝑟 ) ( − β ) = s ( α ) = α = r − 1 ( β ) ,

so

𝑠𝑟 s − 1 = r − 1 .

This shows

Aut ( ℤ ∕ 4 ℤ × ℤ ∕ 2 ℤ ) = ⟨ r , s ⟩ ≃ D 8 ,

where

r { α ↦ β β ↦ − α s { α ↦ α β ↦ − β ( α = ( 1 , 0 ) , β = ( 1 , 1 ) ) .

APPENDIX 2: We prove the following result, used in part (c):

Theorem A. Let H and K be finite groups. Assume that H is abelian and that g . c . d ( | H | , | K | ) = 1 . Let φ 1 , φ 2 : K → Aut ( H ) be two group isomorphisms.

If ker ⁡ ( φ 1 ) ≄ ker ⁡ ( φ 2 ) then H ⋊ φ 1 K ≄ H ⋊ φ 2 K .

Proof. We put G i = H ⋊ φ i K , i = 1 , 2 . We consider the two subgroups of G i given by

H ~ i = H × { 1 } ≃ H , K ~ i = { 1 } × K ≃ K .

( H 1 ~ and H 2 ~ are the same set, but subgroups of distinct groups.)

Assume that H ⋊ φ 1 K ≃ H ⋊ φ 2 K , so that there is an isomorphism f : G 1 → G 2 .

  • The isomorphism f sends H ~ 1 on H ~ 2 .

    Since g . c . d ( | H | , | K | ) = 1 , then g . c . d ( | H ~ i | , | G i : H ~ i | = 1 | , so H i ~ is a normal Hall subgroup of G i (see Exercise 3.3.10).

    A normal Hall subgroup of order d is unique: Suppose that N and M are two normal subgroups of same order d , and indices n ∕ d relatively prime to d . The prime decompositions of d , n ∕ d and n are of the form

    d = p 1 α 1 p 2 α 1 ⋯ p k α k , n d = q 1 β 1 q 2 β 1 ⋯ q l β l ( { p 1 , p 2 , … , p k } ∩ { q 1 , q 2 , … , q l } = ∅ ) ,

    and

    n = p 1 α 1 p 2 α 1 ⋯ p k α k q 1 β 1 q 2 β 1 ⋯ q l β l .

    Then 𝑁𝑀 is a subgroup of G , of order | 𝑁𝑀 | = | N | | M | | N ∩ M | , thus | 𝑁𝑀 | is a divisor of d 2 , so | 𝑁𝑀 | = p 1 γ 1 p 2 γ 2 … p k γ k ( γ i ≥ 0 ) . Since 𝑁𝑀 ≤ G , | 𝑁𝑀 | divides n = | G | , thus 0 ≤ γ i ≤ α i .

    Moreover N ≤ 𝑁𝑀 , and so d ∣ | 𝑁𝑀 | . Therefore α i ≤ γ i . This shows that α i = γ i , thus | 𝑁𝑀 | = | N | , where N ⊆ 𝑁𝑀 , so 𝑁𝑀 = N and similarly 𝑁𝑀 = M . This shows N = M .

    Since f ( H 1 ~ ) and H 2 ~ are normal Hall subgroups of G 2 of same order, we obtain

    f ( H ~ 1 ) = H ~ 2 . (7)
  • Centralizer of H ~ i .

    Let G = H ⋊ φ K , where φ : K → Aut ( H ) is a homomorphism. Here H ~ = H × { 1 } .

    Since H is abelian, for all ( h , k ) ∈ G and all x ∈ H ,

    ( h , k ) ( x , 1 ) ( h , k ) − 1 = ( h φ k ( x ) h − 1 , 1 ) = ( φ k ( x ) , 1 ) .

    Hence, for all ( h , k ) ∈ G ,

    ( h , k ) ∈ C G ( H ~ ) ⟺ ∀ ⁡ x ∈ H , ( φ k ( x ) , 1 ) = ( x , 1 ) ⟺ ∀ ⁡ x ∈ H , φ k ( x ) = x ⟺ φ k = id H ⟺ k ∈ ker ⁡ φ ⟺ ( h , k ) ∈ H × ker ⁡ ( φ ) ,

    ( H × ker ⁡ ( φ ) is here the set { ( h , k ) ∣ h ∈ H , k ∈ ker ⁡ ( φ ) } )

    so that

    C G ( H ~ ) = H × ker ⁡ ( φ ) ,

    where the law on the subgroup H × ker ⁡ ( φ ) is the restriction of the law of G , so that

    C G ( H ~ ) = H ⋊ φ ′ ker ⁡ ( φ ) ,

    where φ ′ = φ ∣ ker ⁡ ( φ ) is the restriction of φ to ker ⁡ ( φ ) , so φ ′ is the trivial homomorphism, and the law on H × ker ⁡ ( φ ) is the direct product.

    C G ( H ~ ) = H ⋊ φ ′ ker ⁡ ( φ ) ≃ H × ker ⁡ ( φ ) ≃ H ~ × ker ⁡ ( φ ) .

    Hence

    C G ( H ~ ) ∕ H ~ ≃ ker ⁡ ( φ ) .

    If we apply this result to G 1 and G 2 , we obtain

    C G ( H ~ i ) ∕ H ~ i ≃ ker ⁡ ( φ i ) . (8)
  • The kernels are isomorphic.

    Since f ( H ~ 1 ) = H ~ 2 by (7), the isomorphism sends centralizers to centralizers:

    f ( C G 1 ( H ~ 1 ) ) = C G 2 ( H ~ 2 ) .

    Then f induces an isomorphism

    C G 1 ( H ~ 1 ) ∕ H ~ 1 ≃ C G 2 ( H ~ 2 ) ∕ H ~ 2 ,

    the well defined isomorphism being

    f ¯ { C G 1 ( H ~ 1 ) ∕ H ~ 1 → C G 2 ( H ~ 2 ) ∕ H ~ 2 h H ~ 1 ↦ f ( h ) H ~ 2 .

    Then (8) shows that

    ker ⁡ ( φ 1 ) ≃ C G 1 ( H ~ 1 ) ∕ H ~ 1 ≃ C G 2 ( H ~ 2 ) ∕ H ~ 2 ≃ ker ⁡ ( φ 2 ) .

    In conclusion,

    ker ⁡ ( φ 1 ) ≄ ker ⁡ ( φ 2 ) ⇒ H ⋊ φ 1 K ≄ H ⋊ φ 2 K .

□
User profile picture
2026-09-14 12:45
Comments