Modern Algebra I: Homework 13

Alejandro Buendia
April 26, 2017

1. (a) Show that a non-abelian simple group cannot be solvable.

Proof. A simple group has two normal subgroups: the trivial group and the entire
group itself. If we have a simple group G, we can create only the subnormal series
G = G0 ⊃ G1 = {1}. But we have that the factor group G0 /G1 ∼ = G is not
abelian by assumption. G is not solvable. 
(b) We proved in class that A5 is simple. Use this and other results (from Gallagher
§12) to show that groups S5 , S6 are not solvable.
Proof. From (a) we know that A5 is not solvable because it is non-abelian and
simple. THM E (Gallagher §12) derives that any subgroup H of a solvable group
G is solvable. Assume toward contradiction that S5 and S6 are solvable. Then
subgroup A5 ⊂ S5 , S6 must be solvable. But since A5 is not solvable, we know S5
and S6 cannot be solvable. 

2. Show that the direct product G × H of two solvable groups is solvable. (Hint: How
can you construct the required chain of subgroups of G × H from those for G and H?)
Proof. Let G = G0 ⊃ G1 ⊃ ... ⊃ Gn = {1} be the subnormal series for solvable group
G, and let H = H0 ⊃ H1 ⊃ ... ⊃ Hm = {1} be the subnormal series for solvable group
H. We can write the subnormal series G × H ⊃ G × H1 ⊃ ... ⊃ G × Hm−1 ⊃ G × {1} ⊃
G1 × {1} ⊃ ... ⊃ Gn−1 × {1} ⊃ {1} × {1}.
We must verify that Ki / Ki−1 for each group in the series. We can see this since if
M / N , then L × M is a normal subgroup of L × N for any groups L, M, N . This
follows trivially from the definition of the direct product. The factor groups Ki−1 /Ki
are isomorphic to factor groups of the subnormal series of G or H. For sub-series
G×H ⊃ G×H1 ⊃ ... ⊃ G×Hm−1 ⊃ G×{1}, each factor group is isomorphic to a factor
group of H, while for sub-series G × {1} ⊃ G1 × {1} ⊃ ... ⊃ Gn−1 × {1} ⊃ {1} × {1},
each factor group is isomorphic to a factor group of G. Hence all factor groups must
be abelian by assumption, and G × H is solvable. 

3. Prove that there are no simple groups of order:

(a) 70 = 2 × 5 × 7
Proof. Consider G with |G| = 2 × 5 × 7. From the third Sylow theorem, we have
S5 (G) ≡ 1 (mod 5) and S5 (G) | 14. This implies that S5 (G) = 1. It follows that

1
 the Sylow 5-subgroup is normal (Theorem 15.10, Judson), and G is not simple.

(b) 64 = 26
Proof. From Gallagher §18, a group G of order n = pa with p prime and a ≥ 2
has a nontrivial center. Hence Z(G) = G or Z(G) = H ⊂ G with H 6= {1}. Since
the center is a normal subgroup, the second case implies that G contains a proper
normal subgroup. In the first case, we know that G is abelian since Z(G) = G
and thus must have every subgroup be normal. Since every group of nonprime
order has a proper subgroup of prime order generated by an element of prime
order, then G must have a proper normal subgroup of prime order. In either case,
G is not simple. 
(c) 100 = 22 × 52
Proof. We have G of the form pa m with p prime and p > m (p = 5, m = 4).
From Exercise 1 of Gallagher §18 (equivalently, Theorem 15.10 of Judson), a
Sylow p-subgroup of G is a proper normal subgroup since there is no divisor
d > p of m with d ≡ 1 (mod p). This also follows from an application of the third
Sylow theorem, since we can see that Sp (G) ≡ 1 (mod p) and Sp (G) | m implies
Sp (G) = 1, from which it follows that the Sylow p-subgroup is normal. Hence G
is not simple. 
(d) 65 = 5 × 13
Proof. We have G of the form pa m with p prime and p > m (p = 13, m = 5).
From Exercise 1 of Gallagher §18 (equivalently, Theorem 15.10 of Judson), a
Sylow p-subgroup of G is a proper normal subgroup since there is no divisor
d > p of m with d ≡ 1 (mod p). This also follows from an application of the third
Sylow theorem, since we can see that Sp (G) ≡ 1 (mod p) and Sp (G) | m implies
Sp (G) = 1, from which it follows that the Sylow p-subgroup is normal. Hence G
is not simple. 
(e) 96 = 25 × 3
Proof. For group G, each subgroup H ⊂ G with index m contains a normal
subgroup K of G with index in G dividing m!. We can see this since we have a
homomorphism φ : G → SG/H (with SG/H having order m!) because G acts on
G/H by left translation. From the first homomorphism theorem, G/K ⊂ SG/H ,
so |G/K| | m!. A corollary of this is that if G has a subgroup H of index m with
m! < |G|, then G has a proper normal subgroup.
Consider G with |G| = 25 × 3. In the corollary, let H be the Sylow 2-subgroup
with index 3. Since 3! is less than the order of G, G has a proper normal subgroup
by the corollary and is thus not simple. 
(f) 80 = 24 × 5
Proof. Consider G with |G| = 24 × 5. From the third Sylow theorem, S5 (G) ≡ 1
(mod 5) and S5 (G) | 16. Then S5 (G) = 1 or S5 (G) = 16. In the first case, the
Sylow 5-subgroup is normal. In the second case, there are 4 × 16 = 64 elements
of order 5. But a Sylow 2-subgroup has order 16 since 16 + 64 = 80. In this case

2
 there can be only one Sylow 2-subgroup, which is then normal. In either case, G
cannot be simple. 

4. Determine the number of ways to color vertices of a regular pentagon using 4 colors,
up to the symmetries of the pentagon (the symmetry group is D5 ). First derive the
formula when the number of colors is n, and then specialize to n = 4.
Let X be the set of vertices. Then D5 acts on X, and the number of orbits of D5 is
the number of colorings up to symmetries. Let k be the number of colorings up to
symmetries. Using Burnside’s Theorem, we have:
1 X
k= |Xg |
|D5 | g∈D
5

We create a table to count the number of colorings fixed by g, in which we are using
n colors:

g 1 r r2 r3 r4 s sr sr2 sr3 sr4

number of colorings n5 n n n n n3 n3 n3 n3 n3

The identity fixes each vertex, so there are n choices of color for all 5 vertices, giving
n5 possible colorings.
For any rotation, all vertices must have the same color. Hence there are n colorings
corresponding to the n different possible colors. This applies to each element of the
form ri , i ≥ 1.
For any reflection, we must have the two pairs of opposite vertices be the same color
and we allow the fifth vertex along the line of symmetry to be any color. There are
n choices for each pair and the singleton vertex, giving n3 possible colorings. Hence
there are n3 colorings for each element of the form sri , i ≥ 0.
This gives a general formula in terms of the number of colors n:

1 5
k(n) = (n + 4n + 5n3 )
10

For n = 4 colors, we find k(4) = 136 .

5. (a) How many necklaces can you arrange out of ten red and two green beads, up to
dihedral symmetries?
We have the symmetry group D12 of order 24. Let X be the set of vertices
(beads). Then D12 acts on X, and the number of orbits of D12 is the number of
arrangements up to symmetries. Again we use Burnside’s Theorem:
1 X
k= |Xg |
|D12 | g∈D
12

3
  
12
The identity fixes each vertex, so there are = 66 arrangements correspond-
2
ing to selecting 2 vertices out of 12 to be green.
Only the rotation r6 ∈ D12 preserves the coloring of the two green beads if they are
positioned opposite each other. This gives 6 possible arrangements corresponding
to the 6 pairs we can choose from the 12 beads.
Consider the two possible reflections:
i. sri with i odd: In this case, the axis of symmetry lies in between two pairs of
beads. We can choose any of the opposite pairs across the axis of symmetry
to be the two green beads. This gives 6 possible arrangements.
ii. sri with i even: In this case, the axis of symmetry lies along a pair of beads.
We may have the two green beads lie on the axis of symmetry, giving 1
arrangement. We may also have the green beads be any of the 5 opposite
pairs across the axis of symmetry. This gives 6 possible arrangements.
There are 12 reflections in D12 . In total, we have 6×12 = 72 arrangements arising
1
from the reflections. Hence, we have k = (66+6+72) = 6 total arrangements.
24
(b) Same question, but out of six red, three green, and three blue beads.
We proceed as above.   
12 6
The identity fixes each vertex. There are = 18480 ways to first choose
6 3
the 6 red beads and then the 3 green beads from the remaining 6 beads.
Consider the rotations. Only r4 , r8 ∈ D12 give rise to valid arrangements. In this
case, we construct the necklace as follows: GRBRGRBR.... We must have the
rotation be a multiple of 4 so that the triples of green and blue beads remain
the same color. Since we must preserve this structure, there are 12 choices for
which to place the first bead in the ordering, giving 24 arrangements for the two
elements in D12 .
Consider the two possible reflections:
i. sri with i odd: In this case, the axis of symmetry lies in between two pairs
of beads. We cannot find any valid arrangement since if we try to find pairs
across the axis of symmetry to have the same color, at least one pair will hold
two different color beads.
ii. sri with i even: In this case, the axis of symmetry lies along a pair of beads.
We must put one green and one blue bead on the axis of symmetry to give
rise to a valid arrangement. There are two choices for which color to put on
which end of the axis of symmetry. Once this is determined, we must put
the two pairs of green and blue beads among the five pairs of opposite beads.
There are 5×4 ways to do this. This gives rise to 2×5×4 = 40 arrangements.
There are 6 reflections of the form sri with i even, hence there are 40 × 6 = 240
arrangements arising from the reflections. Using Burnside’s Theorem, this gives
1
(18480 + 24 + 240) = 781 total arrangements.
24