1. It is a well known theorem due to Frobenius that the number of solutions of the equation

Xn = 1

in a finite group G, is a multiple of the greatest common divisor (n, g) of n and the order g of G. Frobenius himself proved later that the number of solutions of the equation

Xn = a

where a is a fixed element of G, is a multiple of (n, ga), ga being the order of the centralizer Z(a) of ain G.

Nemilov and Pisarevskii (1957) have reported a 390 keV γ-ray in coincidence with neutrons from a Po-Li source, and interpret this as evidence for a level in B10 at 390 keV. A search for confirmatory evidence of this γ–ray has been unsuccessful, and it is estimated that such n–γ coincidences would have been detected in the experiment here described if they involved 8 per cent or more of the neutrons from the Po-Li source.

1. The properties of the circulant determinant or the circulant matrix are familiar. The circulant matrix C of order 4 x 4, with elements in the complex field, will serve for illustration.

The four matrix coefficients of c0, c1 c2, c3 form a reducible matrix representation of the cyclic group ℐ4, so that C is a group matrix for this. Let ω be a primitive 4th root of 1. Then Ω as below, its columns being normalized latent vectors of C,

is unitary and symmetric, and reduces Cto diagonal form thus,

where the μr, the latent roots of C, are given by

All of the above extends naturally to the n x n case.

This paper is concerned with the theory of the probability distribution of the total number of electrons in the avalanche produced by the release of a single electron in the gas of a proportional counter. The disagreement between existing theory and experimental results is discussed and a new theory is proposed, based on the fact that fluctuations in the number of electrons at a given point in the avalanche are accompanied by fluctuations in the average electron energy. This aspect of the problem is incorporated directly into a simple one-dimensional model of the multiplication process, and the resulting distribution function has a mathematical foim in agreement with that observed experimentally.

The fluctuation in the number of electrons predicted by this theory is not constant, but is determined by a parameter which, for large mean values, is essentially the mean fraction of electrons in the avalanche having energies above the ionization energy of the counter gas. Limits on the variation of this parameter are obtained by calculations of the mean values using a particular two-dimensional model, in which electrons are divided into two classes according as they have energies above, or below, the ionization energy. The experimentally observed fluctuation lies within the predicted range and close to the lower limit; it is concluded that there is little scope for improvement in the resolution to be obtained from the conventional type of proportional counter.

Let R be a ring and let ΓR be the Jacobson-Perlis radical. It is shown that the radical of the ring of row-finite matrices over R is the ring of row-finite matrices over ΓR if and only if ΓR is right-vanishing. This is done by extending the results of an earlier paper. One interesting consequence of the theorem is noted.

In the analysis of mixed boundary value problems in the plane, we encounter dual integral equations of the type

If we make the substitutions cos we obtain a pair of dual integral equations of the Titchmarsh type [1, p. 334] with α = − 1, v = − ½ (in Titchmarsh's notation). This is a particular case which is not covered by Busbridge's general solution [2], so that special methods have to be derived for the solution.

In previous papers [1, 2] the author has considered the stability of a current-vortex sheet in a non-diffusing incompressible fluid, the magnetic field being parallel to the plane of discontinuity. In [1] a criterion was given for a parallel magnetic field to stabilise a vortex sheet, and in [2] the energy balance of this system was considered and it was shown how the magnetic energy level is increased at the expense of mechanical kinetic energy when the system becomes unstable. In this paper the oscillations on a plane interface are considered when the magnetic field is not parallel to the interface.

Let S be a point set of the Euclidean plane, such that

(i) S is bounded,

(ii) the closure of S has unit Lebesgue measure.

Let P be an arbitrary set of n points contained in S, and let l(P) denote the total length of the shortest system of lines connecting the points of P together. Define ln to be the supremum of l(P), taken over all sets P of n points in S. Beardwood, Halton, and Hammersley [1[ proved that there exists an absolute constant α, independent of S, such that

An expression is found for the biharmonic Green's function, G(x, y; x1, y1), for an infinite area in the x, y-plane bounded internally by a single curve; at all points of the boundary G has a zero of the second order.

A method is developed for finding the distribution of velocity, density, pressure and magnetic field behind an expanding strong cylindrical shock wave in an infinitely conducting fluid in the presence of a magnetic field.

In the flow of such a fluid there are two distinct methods for producing a strong shock:

(i) by imposing the usual density ratio across the shock, as in the non-magnetic cnse, and

(ii) by imposing a large magnetic field, such that the Alfvén velocity is very much larger than the speed of sound. The distribution of the various physical quantities and the velocity of propagation of the shock are discussed for both cases, and numerical results given.

Let p be a prime and let F be a polynomial in one variable with coefficients in GF(p), the field of p elements. Let d be the degree of F, and let r+1 denote the number of distinct values F(µ) as µ. ranges over GF(p). A generalization of the Waring problem modulo p leads to the question the determination of a lower bound for r.

In an earlier paper (cf. [1]) I had given a generalization of the concept of an absolute discriminant to arbitrary finite number fields K as base fields. In a second paper (cf. [2] 2. 3, see also [3] 1. 3) it was shown that the discriminant δ(Λ/K) of a finite extension Λ of K determines the structure of the ring ς of algebraic integers in Λ qua module over the ring ο of algebraic integers in K. The purpose of the present note is to establish a corresponding result for an arbitrary Dedekind domain ο, and finite separable extensions Λ of its quotient field K. The general theory of discriminants and module invariants developed in [1] and [2] for algebraic integers applies in principle to arbitrary Dedekind domains, as already pointed out in the earlier papers. It is usually evident what further hypotheses—if any—have to be imposed to ensure the validity of any particular theorem. For the quoted result of [2] this is, however, not at all clear. In fact the proof involves the proposition:

I. If an element of K is a square everywhere locally then it is a square in K.

In a recent paper [1], we have given some account of theories of equivalence and intersection on a singular algebraic surface and have shown that such theories share many of the simple properties enjoyed by corresponding theories on non-singular surfaces. Another paper [2], now in preparation, will extend this work to singular varieties of arbitrary dimension. In the meantime, Zobel [3] has drawn attention to some suspect arguments of Samuel [4] concerning the specialization of intersections on a singular variety.

In ring theory there is the following theorem (cf. [1], p. 39):

If R is a ring satisfying the descending chain condition for left ideals, then the following three conditions are equivalent: (i) R is primitive, (ii) R is simple, and (iii) R is isomorphic to the complete ring of linear transformations in a finite dimensional vector space over a division ring.

The object of this note is to extend Dini's theorem about (monotonic) sequences of continuous functions on a compact topological space to the case where the underlying domain is an abstract group which is free from topological restrictions. Continuous functions are replaced by almost periodic real-valued functions and the main result may be stated as follows: If a monotonically increasing sequence (fn) of almost periodic real-valued functions on a group G converges pointwise to an almost periodic function f on G, then the sequence converges to f uniformly. The basic idea of the present (elementary) proof is due to v. Kampen [2] and A. Weil [4], i.e., every almost periodic function on a group induces a kind of compact topology in it, relative to which the function is continuous. We modify this idea with the aid of the mean-value of an almost periodic function and obtain a pseudometric topology. This topology facilitates convergence proofs greatly. Moreover, it turns out to be equivalent with the previous one (Lemma 1). No use will be made of the theory of bounded matrix representations. This is significant as any use of the ‘Approximation Theorem’ [3, p. 66, see also p. 226] would violate the claim of an elementary proof.

Let f(x, y, z) be an indefinite ternary quadratic form of signature (2, 1) and determinant d ≠ 0. Davenport [3] has shown that there exist integral x, y, z with, the equality sign being necessary if and only if f is a positive multiple of f1(x, y, z) = x2 + yz.