What Sylow guarantees
Lagrange's theorem says the order of a subgroup divides |G|, but it does not promise that a subgroup of a given size exists. The Sylow theorems supply the converse for prime powers. Write |G| = pᵏ·m with p prime and gcd(p, m) = 1. A Sylow p-subgroup is a subgroup of the maximal possible p-power order pᵏ. The three theorems are:
- Existence: Sylow p-subgroups exist. So G has a subgroup of order pᵏ for the full prime power dividing |G| — a genuine p-group sitting maximally inside G.
- Conjugacy: all Sylow p-subgroups are conjugate to one another, and every p-subgroup of G is contained in some Sylow p-subgroup. In particular a Sylow p-subgroup is normal iff it is the only one.
- Counting: the number n_p of Sylow p-subgroups satisfies n_p ≡ 1 (mod p) and n_p divides m. These two congruence-and-divisibility constraints are astonishingly restrictive.
The not-simple playbook
A simple group has no normal subgroups except 1 and itself; these are the atoms of finite group theory. The Sylow counting constraints are the standard weapon for proving a group of a given order is not simple — by forcing some n_p = 1, which makes that Sylow subgroup the unique one, hence normal. The routine:
- Factor |G| = p₁^{a₁} ··· p_r^{a_r}. For each prime, list the candidate values of n_p allowed by n_p ≡ 1 (mod p) and n_p | m.
- If some prime forces n_p = 1, you are done: that Sylow subgroup is normal, so G is not simple.
- Otherwise, count elements. Distinct Sylow p-subgroups of prime order p intersect trivially, so n_p of them contribute n_p·(p−1) elements of order p. Summing these counts over several primes often exceeds |G| unless some n_p collapses to 1 — a contradiction that proves not-simple.
- When counting is not enough, use the coset action: if n_p = t > 1, then G acts on the t cosets/conjugates, giving G → S_t; if |G| does not divide t!, the kernel is a nontrivial proper normal subgroup.
Claim: no group of order 30 is simple.
|G| = 30 = 2 * 3 * 5.
Sylow 5: n_5 ≡ 1 (mod 5) and n_5 | 6 => n_5 ∈ {1, 6}.
Sylow 3: n_3 ≡ 1 (mod 3) and n_3 | 10 => n_3 ∈ {1, 10}.
Suppose G is simple, so n_5 ≠ 1 and n_3 ≠ 1. Then n_5 = 6, n_3 = 10.
Count elements of order 5: six Sylow 5-subgroups, each of order 5,
pairwise intersecting in {e}. => 6 * (5 - 1) = 24 elements of order 5.
Count elements of order 3: ten Sylow 3-subgroups, order 3, trivial pairwise
intersection. => 10 * (3 - 1) = 20 elements of order 3.
But 24 + 20 = 44 > 30 = |G|. Contradiction.
Therefore n_5 = 1 or n_3 = 1: a Sylow subgroup is normal, and G is NOT simple.
(In fact one shows G has a normal subgroup of order 15, and G ≅ a
semidirect product built on Z/15Z.)From Sylow to construction
Sylow does more than forbid — combined with the semidirect product, it lets you build and classify. Once you know one Sylow subgroup N is normal and another H meets it trivially with NH = G, then G ≅ N ⋊ H, and the isomorphism type is pinned down by the action of H on N (a homomorphism H → Aut(N)). For order 30 this strategy delivers exactly four groups: Z/30Z and three non-abelian ones (involving D_5, the dihedral piece, and so on). The orbit-stabilizer engine, the class equation, the isomorphism theorems, and Sylow together turn “classify all groups of order n” from a vague wish into a finite bookkeeping problem.