Exerc. 8.1.13 in D.J.S. Robinson, A Course in the Theory of Groups, 2d edition, p. 222, is as follows : Let F be a field, let $n$ be a natural number, let $G$ be a subgroup of $GL(n,F)$. Assume that the class number (number of conjugacy classes) of $G$ is finite. Then $G$ is finite.

I think I found a proof, but I wonder if it is the best possible. I use two lemmas :

Lemma 1. Let H be a group with finite class number $c$. Let $N$ a normal subgroup of finite index of $H$. Then the class number of $N$ is finite (and $\leq c \ [H:N]$).

Lemma 2. Let H be a soluble group wih finite class number $c$. Then $H$ is finite (and $\vert H \vert \leq c^{2^{s}-1}$, where $s$ is the derived length of $H$).

Now, I sketch the proof of the satement of the exercise. Without loss of generality, we may assume that $F$ is algebraically closed. The statement is equivalent to the same statement for a subgroup $G$ of $GL(V)$, where $V$ is a $F$-vector space of dimension $n$. We use induction on $n$. The case $n=0$ being trivial, we asume $n > 0$. Assume first that $G$ is irreducible. Since the class number of $G$ is finite, there is only a finite number of possible values for the traces of the elements of $G$. This, with the irreducibility of $G$, implies that $G$ is finite (Robinson, 8.1.9, p. 220). Thus the statement is true if $G$ is irreducible. If not, there is a subspace $U$ of $V$, invariant by $G$, with $0 < U < V$. Let $L_{1}$ denote the subgroup of $G$ formed by the elements of $G$ whose restriction to $U$ is identiy. Let $L_{2}$ denote the subgroup of $G$ formed by the element of $G$ acting on $V/U$ as the identity (with an evident definition). Put $L = L_{1} \cap L_{2}$. As in the proof of Robinson, 8.1.11, (i), p. 221, $L$ is a normal subgroup of $G$ and the induction hypothesis implies that $G/L$ is finite; moreover, $L$ is isomorphic to a subgroup of the group of unitriangular matrices in $GL(n,F)$, thus $L$ is nilpotent. By lemma 1, the class number of $L$ is finite. By lemma 2 (and the fact that $L$ is nilpotent and thus soluble), $L$ is finite, thus $G$ is finite.

I wonder if the lemmas are really needed, because Robinson doesn't speak of them, alhough they don't seem trivial to me.

Do you know a simpler proof ? Thanks in advance.


Your line of reasoning is perfectly right, see also a spelled out proof in the book of T.Y. Lam, A First Course in Non-Commutative Rings, Theorem (9.5).


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.