The answer to the modified question turned out to be positive and in fact the following, more
general, theorem holds.
Theorem 1.1. Let π be a set of primes. Let x ∈ G and suppose that Inda,x(x) is a π-number for any a ∈ G.
Then IndG(x) is a π-number.
The proof of the above theorem is very short and elementary. It is interesting to note that if
IndG(x) is a π-number IndH(x) is not necessarily a π-number for each subgroup H. Let M be the
elementary group of order 8 and A the non-abelian group of order 21. Let A act on M in such
a manner that the subgroup of order 7 permutes the involutions in M transitively. Let G be the
extension of M by A. Choose x to be an involution in M. Then IndG(x) = 7 but there is a non-abelian
subgroup, H of order 24 such that IndH(x) = 3.
Another result obtained in the paper is the following theorem.
Theorem 1.2. Suppose that Inda,b,x(x) is a prime-power for any a, b ∈ G. Then IndG(x) is a prime-power.
The proof of Theorem 1.2 is no longer elementary. In particular it uses the well-known result of
Aschbacher and Guralnick [1] that every non-abelian simple group is 2-generated. This depends on
the classification of finite simple groups. Another important tool used in the proof of Theorem 1.2 is
Flavell’s theorem [3] that x ∈ F2(G) if and only if x ∈ F2(a, x) for any a ∈ G.