The Fitting class of finite soluble π-groups, where π is an arbitrary set of primes, has the property that each complement of an -avoided, complemented chief factor of any finite soluble group G contains an -injector of G. In other words, each -avoided, complemented chief factor of G is -complemented in the sense of Hartley (see [2]).

Magill, Jr. and Yamamuro have been responsible in recent years for a number of papers showing that the property that every automorphism is inner is held by many semigroups of functions and relations on topological spaces. Following [9], we say a semigroup has the Magill property if every automorphism is inner. we say a semigroup has the Magill property if every automorphism is inner. That the semigroup of Fréchet differentiable selfmaps, D of a finite dimensional Banach space E, had the Magill property was shown in [10], while a lengthy result in [6] extended this to the semigroup of k times Fréchet differentiable selfmaps, Dk, of a Fréchet Montel space (FM-space). In the latter paper it was noted that with a little additional effort the semigroup Ck, of k times continuously Fréchet differentiable selfmaps of FM-space, could be shown to possess the Magill property. It is the purpose of this paper to present a simpler proof of this result in the case where the underlying space is finite dimensional.

The purpose of this paper is to give a more precise form of Theorem 1 of [2], which gives a structure theorem for subgroups of HNN groups; we prove the following.

Let H be a subgroup of the HNN group <A, xi;xiU-ixi-1 = Ui>. Then H is an HNN group whose base is a tree product of groups H ∪ wAw-1 where w runs over a set of double coset representatives of (H,A); the amalgamated and associated subgroups are all of the form H ∊ vUiv-l for some v. We can be more precise about which subgroups occur and about the tree product. We will also obtain stronger forms of other results in [1] and [2].

The main result of this paper is the theorem in the title. Only special cases of it seem to be known so far. As an application, we obtain a result on the unique extension of Galois connexions. As a matter of fact, it is only by the use of Galois connexions that we obtain the main result, in its present generality.

This paper is an outgrowth of my old battle with the open sentence problem for the theory of finite groups. The unsolvability of the word problem for groups (cf. [1] and [4]) entails the undecidability of the open sentence problem for the elementary theory of groups and thus strengthens the original undecidability result for this theory (cf. [7]). The fact that the elementary theory of finite groups is also undecidable (cf. [2] and [6]) therefore justifies my interest in the open sentence problem for that theory. This paper contains a construction of groups that might lead to a negative solution.

Throughout this paper we will assume that all groups are contained in some fixed but arbitrary universe. Thus the class of all groups becomes a set. If is a class of groups then we assume that 1 ∈ and if H ≅ G ∈ then H ∈ . German capitals will be used to denote classes of groups.

1. In their fundamental paper of 1949, Higman, Neumann and Neumann proved for the first time that a countable group can always be embedded in some 2-generator group: [1], Theorem IV. Two kinds of improvement of this result have recently appeared. In [4], Theorem 2, Dark has shown that the embedding can always be made subnormally. On the other hand, in [2], Theorem 2.1, Levin has shown that the two generators can be given preassigned orders m > 1 and n > 2; and in [3], Miller and Schupp prove that the 2-generator group can also be made to satisfy several additional requirements, such as being complete and Hopfian.

For sums of independent and identically distributed random variables xn, the HartmanWintner law of the iterated logarithm is equivalent to xn ∈ L2. We show that this is also true when the xn, form a stationary, ergodic martingale difference sequence. This is accomplished by extending a theorem of Volker Strassen to the present context.

This paper presents a basis for the laws which hold in each of the finite simple groups, PSL(2, 2n), n> ≥ 2, thus partially solving a problem raised by Cossey, Macdonald and Street [3]. They considered the more general problem of finding bases for the laws which hold in PSL(2, pn), and succeded in finding a number of general laws, and in completing bases for pn ≤ 11. The solution of the general problem appears to be very difficult.

A weak but still unsolved version of the first half of Problem 7 in Hanna Neumann's book [5] asks whether the product of a nontrivial join-irreducible variety and an arbitrary variety is necessarily also join-irreducible. One of the results of this paper is a positive answer provided either is abelian or the infiniterank free groups of have no nontrivial abelian verbal subgroups. For a general discussion, see [4]; all unexplained notation is as in [5].