2
$\begingroup$

I'm trying to prove the following result. I think I'm not understanding it correctly. Just to clarify, my knowledge extends to Sylow's theorem corollaries but I don't know about solvable groups.

Let $G$ be a group of order $pqr$, $p>q>r$. Prove that $|G| \geq 1 + n_p(p-1) + n_q(q-1) + n_r(r-1)$

Here's what I have thought.

What I understood is that equation tells me to prove that there are more elements of the group $G$ than elements that belong to any Sylow subgroup. Isn't this always valid? Which means that I'm looking at the case when the inequality is a strict one?. This confuses me. Anyways, in that case if I find an element of order, say $pq$, then that element will not be an element of any Sylow subgroup. So the equation will be valid with a strict inequality.

I know that in this group there is a normal group of order $p$ which means that $n_p = 1$. From this I can conclude that there is a normal subgroup of order $pq$. If I knew that $p \neq 1\mod{q}$ , then the subgroup would be a cyclic subgroup which means that I can find an element of order $pq$ and the equation will be valid. Bu there are no restrictions on either $p$ or $q$.

I would appreciate any hints. Sorry if it has been asked before.

Edit: Maybe the question could be rephrased as is there a cyclic subgroup of order bigger than $p$? If that's true then the problem is solved.

$\endgroup$
  • $\begingroup$ You are right, in this case this inequality just says that $G$ has more elements than in Sylow's subgroups, and this is always true. $\endgroup$ – SMM Jul 8 '18 at 15:31
3
$\begingroup$

Since all Sylow subgroups of $G$ are of prime order, the intersection of any two distinct Sylow subgroups is trivial. Hence there are $n_p (p − 1)$ elements of order $p$, $n_q (q − 1)$ elements of order $q$, $n_r(r − 1)$ elements of order $r$, and $1$ element of order $1$ in G. Therefore the inequality follows.

$\endgroup$
  • $\begingroup$ I see. Thanks for the answer! Do you know when the inequality is strict? I have tried some small cases and found that the inequality is strict but couldn't quite prove it in the general case. $\endgroup$ – Leo Lerena Jul 8 '18 at 17:33
  • 2
    $\begingroup$ The inequality is strict, because all such groups have elements of composite order. The proof depends on the non-obvious (and not that easy) theorem that if all Sylow subgroups are cyclic, then $G$ is a semidirect product of two cyclic groups. Obviously one of those factors must be a cyclic subgroup of composite order, so it has elements of composite order. $\endgroup$ – C Monsour Jul 8 '18 at 18:25
  • $\begingroup$ @CMonsour Thanks a lot for the answer. Just I was thinking how to prove that theorem in the special case that the group order is $231=3*7*11$, i.e that it's isomorphic to a semi direct product of $\mathbb{Z}_{33}$ and $\mathbb{Z}_7$. $\endgroup$ – Leo Lerena Jul 8 '18 at 18:46
  • 1
    $\begingroup$ Well, the 7 and 11 Sylows need to be normal, and the cyclic group of order 11 has no automorphisms of order 3, so the only non-abelian possibility is $\Bbb{Z}/(7)\rtimes\Bbb{Z}/(33)\cong\Bbb{Z}/(77)\rtimes\Bbb{Z}/(3)\cong F_{21}\times \Bbb{Z}/(11)$, where $F_{21}$ is the non-abelian group of order 21. $\endgroup$ – C Monsour Jul 8 '18 at 18:54
  • $\begingroup$ @CMonsour Thanks a lot!! You make it look so easy. $\endgroup$ – Leo Lerena Jul 8 '18 at 19:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.