The purpose of this paper is to extend to locally convex spaces and to uncountable systems several well-known results concerning infinite series, biorthogonal sequences, and Schauder bases. Section 2 gives three extensions of the theorem of Orlicz (10) and Pettis (11) and some lemmas that will be needed later. The third section introduces the notions of a summability basis and a summability basis of subspaces, and two main theorems are proved, including a simplification of Retherford and McArthur's proof (12) of a theorem of Nikol'skiĭ (9). Section 4 investigates the positive cone of an uncountable biorthogonal system, particularly conditions equivalent to the regularity of this cone.

In a recent paper, E. T. Copson (2) proves the following result:

Theorem C.Letki >0 (i = 1, …, m), k1 +…+ km = 1, and the real sequence (an) satisfy the inequality

If (an) is bounded, then it must be convergent.

Recently Scheiblich (7) and Munn (3), amongst others, have given explicit constructions for FIA, the free inverse semigroup on a non-empty set A. Further, Reilly (5) has investigated the free inverse subsemigroups of FIA. In this note we generalise two of Reilly's lesser results, and also characterise the surjective endomorphisms of FIA. The latter enables us to determine the group of automorphisms of FIA, and to show that if A is finite then FIA is Hopfian (a result proved independently by Munn (3)). Finally, we give an alternative proof of Reilly's main theorem, which uses Munn's theory of birooted trees.

Let G be a finite group generated by n elements and defined by m relations, then G has a presentation, G = {x1, …, xn | R1, …, Rm} = F/R where F is free on generators x1, …, xn and R is the normal closure in F of R1, …, Rm. The deficiency of this presentation is n − m. Since G is finite the deficiency is non-positive and the deficiency of G is the maximal over the deficiencies of all presentations for G.

Let S denote a compact semitopological semigroup (i.e. the multiplication is separately continuous) and P(S) the set of probability measures on S. Then P(S) is a compact semitopological semigroup under convolution and the weak * topology (4). Let Γ be a subsemigroup of P(S) and where supp μ is the support of μ ∈P(S). In the case in which S is commutative it was shown by Glicksberg in (4) that S(Γ) is an algebraic group in S if Γ is an algebraic group. For a general semigroup S, Pym (7) considered Γ = {η}, η being an idempotent, and established that S(Γ) is a topologically simple subsemigroup of S, i.e. every ideal of S(Γ) is dense in S(Γ). In this note we prove that if Γ is a simple subsemigroup of P(S) (a semigroup is simple if it contains no proper ideal) which contains an idempotent then S(Γ) is a topologically simple subsemigroup of S. We also give an example to show that our conclusion (hence also Pym's) is best possible in the sense that S(Γ) is not simple in general

Let V be a complex algebraic variety. Given integers a1, …, am such that

one defines a (a1, …, am)-flag as a nested system

of subspaces of Sn, the n-dimensional complex projective space. The set of all such flags is called an incomplete flag-manifold in Sn, and is denoted by W(al, …, am). Also let E be a complex n-dimensional vector bundle over V. Then we denote by E(a1, …, am−1, n; V) an associated fibre bundle of E with fibre W(a1 − 1, …, am−1 − 1, n − 1). E(a1, …, am−1 − 1, n; V) is called an incomplete flag bundle of E over V (cf. (2), (3)). In Section 10.3 and Section 14.4 of (1), the generalised Todd genus Ty(W(0, n)) and Ty(W(0, 1, …, n)) of the n-dimensional projective space W(0, n) and the flag manifold W(0, 1, …, n) (or F(n+l)) were calculated. Here we compute Ty(W(a1, …, am)) and also Ty(E(a1, …, am−1, n; V)).

The curve in question is the non-singular intersection Γ of the n − 1 quadric primals

where it is presumed that no two of the n + 1 numbers aj are equal. Define

then it will be seen that the osculating prime of Γ at x = ξ is

In this note we present some rather loosely connected results on Banach algebras together with some illustrative examples. We consider various conditions on a Banach algebra which imply that it is finite dimensional. We also consider conditions which imply the existence of non-zero nilpotents, and hence the existence of finite dimensional subalgebras. In the setting of Banach algebras quasinilpotents figure more prominently than nilpotents. We give an example of a non-commutative Banach algebra in which 0 is the only quasinilpotent; this resolves a problem of Hirschfeld and Zelazko (4).

For each non-empty subset Λ of the complex plane, let (Λ) be the set of all those operators (on a fixed Hilbert space H) whose spectrum is included in Λ. The problem of spectral approximation is to determine how closely each operator on H can be approximated (in the norm) by operators in (Λ). The problem appears to be connected with the stability theory of certain differential equations. (Consider the case in which Λ is the right half plane.) In its general form the problem is extraordinarily difficult. Thus, for instance, even when Λ is the singleton {0}, so that (Λ) is the set of quasinilpotent operators, the determination of the closure of (Λ) has been an open problem for several years (3, Problem 7).

Let A denote a complex Banach algebra with unit, Inv(A) the set of invertible elements of A, Sp(a) and r(a) the spectrum and spectral radius respectively of an element a of A. Let Γ denote the set of elements of A whose spectra contain non-negative real numbers, i.e.

Recently Passman (attributing the origin of the idea to Zalesskii) has defined the following ideal in a ring, (2).

Definition. For a unitary ring R,

N * R = {α ∈ R | αS is nilpotent for all finitely generated subrings S of R}.

For a group algebra KG over a field K of characteristic p ≠ 0, he has proved the radical property:

A basic set (formerly basic sequence) ℬ is a set of pairs (a, b) of positive integers satisfying

(1) if (a, b) ∈ ℬ, then (b, a) ∈ ℬ,

(2) (a, bc) ∈ ℬ if and only if (a, b) ∈ ℬ and (a, c) ∈ ℬ,

(3) (1, k ∈ ℬ, k = 1, 2, ….

Some familiar examples of basic sets are

, where Sk = {(1, k), (k, 1)},

= {(a, b)| a and b are relatively prime positive integers},

ℒ = {(a, b)| a and b are any positive integers}.

This paper concerns criteria for assuring that every solution of a real fourth order nonselfadjoint differential equation

is oscillatory at x = ∞. Our technique is a generalisation of that used by Whyburn (1) for the study of the selfadjoint equation,

combined with the theory of H-oscillation of vector equations as introduced by Domšlak (2) and studied by Noussair and Swanson (3). Whyburn's technique consists of representing (1.2) as a dynamical system of the form

and then studying (1.3) in terms of polar coordinates in the y, z-plane. In Section 2 below we show how to represent (1.1) as a dynamical system of the form

Throughout, A denotes an associative ring with identity and “module” means “left, unitary A-module”. In (3), it is proved that A is semi-simple, Artinian if A is a semi-prime ring such that every left ideal is a left annihilator. A natural question is whether a similar result holds for a (von Neumann) regular ring. The first proposition of this short note is that if A contains no non-zero nilpotent element, then A is regular iff every principal left ideal is the left annihilator of an element of A. It is well-known that a commutative ring is regular iff every simple module is injective (I. Kaplansky, see (2, p. 130)). The second proposition here is a partial generalisation of that result.

The abstract theory of positive compact operators (acting in a partially ordered Banach space) has proved to be particularly useful in the theory of integral equations. In a recent paper (2) it was shown that many of the now classical theorems for positive compact operators can be extended to certain classes of non-compact operators. One result, proved in (2, Theorem 5), was a fixed point theorem for compressive k-set contractions (k<l). The main result of this paper (Theorem 3.3) shows that some of the hypotheses of (2, Theorem 5) are unnecessary. We use techniques based on those used by M. A. Krasnoselskii in the proof of Theorem 4.12 in (4), which is the classical fixed point theorem for compressive compact operators, to obtain a complete generalisation of this classical result to the k-set contractions (k < 1). It should be remarked that J. D. Hamilton has extended the same result to A-proper mappings (3, Theorem 1). However apparently it is not known, even in the case when we are dealing with a Π1-space, whether k-set contractions are A-proper or not.