1 Computation
1 2: compute everything
3 4 5: express as disjoint cycles and lcm them.
2 Computation’
compute
3 Sign
Change of variable (let \(y_i = h(x_i)\)).
Substitution (sub \(x_i = f(i)\)).
4 Odd even
compute
Because \(\operatorname{sgn}\) is a function
because \((12)(12) = \operatorname{id}\)
because \(\left\lvert A_n \right\rvert = \left\lvert C \right\rvert\) and by part 2 each must be half as big as \(S_n\).
5 Only 2 cosets implies normal subgroup
Let any \(n\in N\), then for any \(g \in G\), we have two case,
- \(gn \in N = eN\),
- \(gn \in C\) where \(C\) is the “other” coset.
We consider the term \(gng^{-1}\), in the first case, \[\begin{align*} ng^{-1} &= ( \underbrace{gn^{-1}}_{\text{in }N} )^{-1} \\ gng^{-1} &= g n' \text{ for some }n' \in N \\ &\in N \end{align*}\] and in the second case, we first claim that \(C = gN = g^{-1} N\), for suppose not, then \(g^{-1}N = N\) which entails that \(eN = g(g^{-1}N) = C\). \[\begin{align*} ng^{-1} &= ( \underbrace{gn^{-1}}_{\text{in }C} )^{-1} \\ &= g^{-1} n' \text{ for some }n' \in N \\ gng^{-1} &= n' \in N \end{align*}\] \(N\) is normal in both cases, as the previous question demonstrated that the only cosets of \(A_n\) are itself and \(C\), we have \(A_n\) is a normal subgroup.
6 Conjugates.
Refl, Sym, Trans.
Suppose \(g' \sim g\), then unravel definition of \(C_G(g)\) to show that \(g' \in C_G(g)\).
Straight from part i
7
By normality, the condition \(C_G(g) \subseteq N\) is satisfied.
8 Cycles
…