Sack (3) and Sneddon (4) have considered the problem of a penny-shaped crack in an infinite medium under tension normal to the face of the crack. The same problem has been considered more recently by Green (1) by a different method. In this note we consider the case of the same shaped crack with the medium subjected to a uniform shearing stress parallel to the plane of the crack. It is found that the solution is analogous to that of the corresponding two-dimensional problem considered by Starr (5).

1. In his classical investigation of the motion of a sphere moving (without spin) in a fluid at rest at infinity, Stokes showed (using his method of ‘stream functions') that the sphere experiences a thrust − 6πμau, where a is the radius of the sphere, and μ, u have their usual meanings.

If the average spacing of the resonance levels is very small as compared with the range of energy in which the spacing or width of the levels changes appreciably on the average, one can speak of a statistical distribution of the level spacings and widths. The question then comes up naturally, whether the ‘distribution law’ for width and spacing is different for protons, neutrons, etc., i.e. whether it depends on the long-range interaction of the scattered particle. It is pointed out that while the average width must depend on the long-range interaction (on account of the penetration factor) the distribution of the widths, if these are measured in terms of their average, can be expected to be independent therefrom. In the case of the level spacings, not only the distribution about the average, but the average itself is also independent of the long-range interaction. It is pointed out, incidentally, that the logarithmic derivative of the wave function at the surface of the nucleus will have, on the whole, a positive value between resonances because of the greater effect of the high-energy resonances than that of the bound states.

1. Boltzmann's differentio-integral equation for the molecular velocity distribution function in a perfect gas forms the natural starting-point for a mathematical treatment of the kinetic theory of gases. The classical results of Maxwell and Boltzmann in this theory are well known. They include the proof that, for simple gases, i.e. those in which the molecules have only the three translational degrees of freedom, the only stationary and spatially homogeneous solution is the one which corresponds to the Maxwellian distribution.

The triple velocity correlation, in turbulence produced by inserting a square-mesh grid near the beginning of the working section of a wind tunnel, has been measured for mesh Reynolds numbers of RM = 5300, 21,200 and 42,400 (RM = UM/ν, where U is the mean wind speed in the working section of the tunnel and M is the centre to centre spacing of the rods making up the grid; ν is the kinematic viscosity of air). At the lowest Reynolds number the correlation has been measured at distances downstream of the grid varying from 20 to 120M. This range covers practically all of the initial period of the decay of turbulence, where the turbulent intensity varies as t−1.

Three-dimensional stress distributions in hexagonal aeolotropic materials have recently been considered by Elliott(1, 2), who obtained a general solution of the elastic equations of equilibrium in terms of two ‘harmonic’ functions, or, in the case of axially symmetric stress distributions, in terms of a single stress function. These stress functions are analogous to the stress functions employed to define stress systems in isotropic materials, and in the present note further problems in hexagonal aeolotropic media are solved, the method in each case being similar to that used for the corresponding problem in isotropic materials. Because of this similarity detailed explanations are unnecessary and only the essential steps in the working are given below.

Experimentally it is found for a ferromagnetic substance that the spontaneous magnetization decreases as the temperature increases. At a certain temperature called the Curie point, the magnetization disappears (or substantially disappears) and remains zero for higher temperatures. Associated with these magnetic properties is an anomaly in the specific heat. This quantity is greater than that which would be calculated theoretically for the material were it non-magnetic and becomes large when the temperature approaches the Curie point from below. Just above the Curie point there is a sharp but continuous decrease in the specific heat and the system is said to show critical behaviour. We shall examine the Ising model of the ferromagnet with a simple cubic lattice structure, determine the specific heat anomaly and the corresponding energy excess due to magnetism, and also see how the critical behaviour of the model compares with that actually observed.

In the theory of beta-decay, as is well known, when the electron is described by relativistic wave-functions, e.g. in a Coulomb field, a difficulty arises when these become infinite at the centre of force, r = 0. The wave-functions are then evaluated at the nuclear radius, r = R, and in many formulae there appears (1) a rather sensitive dependence on R. For the capture of an orbital electron by a nucleus the results of calculation depend strongly on the behaviour of the wave-function of the bound electron near the nucleus. The nature of the potential energy near the nucleus will thus affect such quantities as the ratio of K-capture to positron emission. Wannier (2) has discussed a similar problem for the Schrödinger equation, but the singular behaviour of the solutions of the Dirac equation for the ground state makes it desirable to have the solution for the relativistic case. In the following, relativistic wave-functions are derived and used for a potential energy which is Coulombian outside the nucleus and constant inside. Similar problems have been considered by others, Breit et al. (3), Racah (4), in application to the calculation of hyperfine structure, and also by Broch (5).

Detailed measurements have been made of the superconducting transition temperatures and critical magnetic fields of the tin isotopes of mass 116, 120 and 124. The transition temperature varies with isotopic mass according to the law Tc ∞ M−n, with n = 0·462 ± 0·014, a result very similar to that already found in mercury. The critical field curves of isotopes 116 and 124 are geometrically similar, in the sense that both many be represented by the same equation, Hc/Ho = f(T/Tc), with the same ratio Ho/Tc for both. It is deduced that the electronic specific heat in normal tin varies only slowly, if at all, with isotopic mass. The variation of Tc and Ho with M is very close to that predicted by Fröhlich and Bardeen.

In this paper, the theoretical double and triple velocity correlation functions, f(r), g(r) and h(r), which correspond to Heisenberg's spectrum of isotropic turbulence, are obtained numerically for two Reynolds numbers. One set of these correlations is for the limiting case of infinite Reynolds number. In addition, a method is developed for deriving the approximate form of the double correlations for any Reynolds number, which is not too small, from the corresponding correlations for infinite Reynolds number. These theoretical correlations are then compared with the results of experiment.

The equation of steady rectilinear motion of a material of variable viscosity is transformed, when the pressure gradient vanishes, into a second-order differential equation which is linear in the dependent variable. A general solution of the linearized equation can be obtained in the case of a wide class of plastic solids and viscous liquids. By making use of the idealized materials which lend themselves to this method of solution, a theoretical comparison can be made of the behaviour of plastic solids of different types under the same conditions. As an illustration, a plastic solid which yields very sharply when the yield point is first exceeded is compared with one which yields only gradually at the same yield value; the flow caused by the longitudinal motion of an (approximately) elliptic cylinder in an infinite mass of material is investigated in each case.

Measurements of the areas of magnetization curves of superconducting colloids and films indicate that the difference between the surface energies of the normal and super-conducting states at the same temperature is much smaller than has been suggested hitherto. An attempt is made to explain the high critical fields of thin films, which led to the introduction of the surface energy, by the hypothesis that a negative interphase surface energy in films favours the persistence of superconducting threads in fields higher than those normally regarded as sufficient to destroy superconductivity. A model is proposed of the interface between normal and superconducting phases which allows the interphase surface energy to be positive in pure macroscopic specimens but negative in alloys and thin films and at lattice defects. This enables a qualitative explanation to be given of some of the details, such as supercooling, of the phase transition in pure metals and alloys.

1. Introduction. In a recent paper Clark (1) has dealt with the problem of the rotating disk, the material of which is such that the waves of dilatation in this particular material travel with the velocity of light. The material of the disk is supposed to be under an isotropic stress p when in a strained state, and the relation between stress p and the dilatation Δ is found to be connected by an expression

where a = density in the unstrained state, and Δ is given by

Ui (i = 1, 2, 3) are the components of the strain.

On the asymptotic probability distribution for certain Markoff processes 46 (1950), 581.

By Walter Ledermann

Mr H. Schneider has drawn my attention to a mistake which occurs on p. 591 of the above paper.

Both the Ising theory of ferromagnetism and the theory of regular solutions are concerned with systems arranged on a lattice and make the assumption that each system interacts only with its nearest neighbours. Mathematically, there is a close parallel between the two problems (see, for instance, Rushbrooke (1)). In the first half of this present paper the partition functions for these two problems are examined in some detail. Power series expansions of the partition function of the Ising model, valid for low and high temperatures, are obtained. The terms obtained in the power series have been analysed and approximate numerical results obtained. It is hoped to publish these in a second paper.

An investigation by G. L. Clark is found to provide exceptionally instructive illustrations of the calculation of the variously defined ‘densities’ and ‘masses’ occurring in general relativity theory and of their physical significance. In particular, the effects of motion and stress upon gravitational mass are elucidated. An explicit example shows how kinetic energy is included in the gravitational mass.

The object of the present paper is to discuss the Fourier expansion of the Riesz potential. For this purpose a new definition of the electromagnetic potentials, depending upon an arbitrary parameter α is given. It is shown that this definition is a generalization of the Wentzel potentials in the α-plane, whereas that given by Fremberg (3) is a generalization of the Maxwell potentials. The analysis is applied to the problem of eliminating, in a straightforward way, the longitudinal part of the potential describing the electromagnetic field. The problem of the quantization of the field, based on its Fourier expansion, will be considered in another paper. The recent work of Tomonaga, Schwinger and Dyson, and the regularization process of Pauli has lifted the theory of quantum electrodynamics to a much higher level of rigour and fruitful applicability. All the same, a further study of Riesz potential seems to us of some interest in this field.