The notion of a group G having periodic cohomology after k steps was introduced by Talelli in [10], and is equivalent to having the functors Hm(G, —) and Hm+q(G, —) naturally isomorphic for some q ≥ 1 and all m ≥k + 1. It extends to infinite groups the long-understood phenomenon of cohomological periodicity for finite groups (for which k = 0).

In [6] B. H. Neumann proved the following beautiful result: if a group G is covered by finitely many cosets, say G = xiHi, then we can omit from the union any xiHi, for which |G|Hj| is infinite. In particular, |G:Hj| is finite, for some j ∈ {l,…,n}.

In an unpublished result R. Baer characterized the groups covered by finitely many abelian subgroups, they are exactly the centre-by-finite groups [8]. Coverings by nilpotent subgroups or by Engel subgroups and by normal subgroups have been studied, for example, by R. Baer (see [8]), L. C. Kappe [2,1], M. A. Brodie and R. F. Chamberlain [1], and recently by M. J. Tomkinson [9].

Let R be a commutative Noetherian ring and G a group of elements acting on R as automorphisms. In this note, we are concerned with the structure of the lattice of invariant ideals of R. In particular we shall compute the Krull dimension of this lattice. Our group is an arbitrary group. There are none of the usual assumptions of some sort of algebraic action.

Recently, P. Kaplan and K. S. Williams [10] considered (as an example) the representation of primes by binary quadratic forms of discriminant –768. These forms fall into 4 genera, each consisting of two classes. In particular, they considered the forms

F=3X2+642 and G = 12X2+12XY+19Y2.

It follows from genus theory (as explained in [10]) that every prime p ≡ 19 mod 24 is represented by exactly one of the forms F and G. Based on numerical data, they conjectured that a prime p ≡ 19 mod 24 is represented by

where

Vo = 2, V1 = -4, Vn+2=-4Vn+1 -Vn (n∨0).

We prove that if n>0 is an integer and r>0 is a real number, then

The upper bound is best possible. Inequality (*) is a converse of a result of G. Bennett who proved that Qn(r)>l.