Table of contents

After a review of the numerous experimental observations in this field which need to be explained, a new theory is presented of nucleus formation when a cold-worked metal is annealed. The nuclei are held to be formed in the most distorted parts of the lattice, in the `local curvatures' postulated by Burgers. At these places the lattice locally regains its perfection by the recently discovered process of diffusion of dislocations, or `polygonization', and the perfect nucleus so created is then able to consume the less perfect crystal in its vicinity. Experimental evidence for this picture is presented. Each potential nucleus is held to have a definite incubation period (as distinct from the fixed probability of becoming active which was the basis of previous theories). By means of certain assumptions as to the distribution of curvatures and the dependence of incubation time on curvature, a quantitative theory is developed, which is found to be capable of accounting for all the experimental data on nucleation rates. The paper concludes by comparing critically the new theory with older ones.

Equations of motion, previously derived by Bloch (1946), for the nuclear magnetic moment in a nuclear induction experiment are applied to the steady state reached when the strong magnetic field H_z is swept sinusoidally about its resonance value. Different approximations for the shapes and sizes of nuclear induction signals are derived for various limiting cases, depending (a) whether the passage of H_z is ` adiabatic ' or ` extremely non-adiabatic ' (very weak radio-frequency field) and (b) whether the relaxation times are shorter or longer than the period of the sweep of H_z. It is found that quite large signals may be induced by even a very weak radio-frequency field if the relaxation times are very long.

On the assumption that τ-mesons are scalar particles and have an appreciable interaction with nucleons, the lifetime of a τ-meson at rest decaying into two photons and a pseudoscalar π-meson, is calculated by means of standard perturbation theory. The lifetime obtained is G × 10^-12 sec. where G is the reciprocal of the product of the dimensionless meson-nucleon coupling parameters. This result is roughly what can be expected from observations and, taken together with the results of a previous paper, seems to indicate that the heavy mesons have spin zero.

A new method for measuring the propagation velocity of the discharge along a Geiger non-proportional counter is described. The results obtained with cylinder grade argon, hydrogen and neon are summarized graphically, and a brief discussion of propagation mechanisms is given.

The Thomas-Fermi statistical theory of the free atom is used to calculate the momentum distribution and the shape of the Compton profile for X-ray scattering, the results being expressed in dimensionless form applicable to all atoms. The mean momentum, which varies as Z^2/3, where Z is the atomic number, agrees well with more accurate wave-mechanical calculations. Numerical results are also reported for argon and krypton using the Dirac modification of the original theory, in which exchange effects are taken into account and the atom has a finite radius. Agreement with experiment is not noticeably better with these refinements.

A quantitative study of the domain structure of single crystals of silicon-iron has been made by the powder pattern method, special examination being made of the patterns formed on a strip specimen whose surface was approximately a (100) plane with a [011] direction parallel to the long edge of the strip. For Mode III magnetization, which occurs when a field acts inside the specimen parallel to the [011] direction, the line deposits formed at right angles thereto were on the average spaced in good accord with Néel's theory in the case of wide strips, but in less good accord for narrow strips. Qualitative results of other measurements on strip and disc specimens are reported and interpreted.

An investigation of the infra-red spectrum of magnesium oxide has been carried out in the range up to 25.5 μ, both in reflection and transmission. The eigen-frequency is found to be at 17.5 ± 0.3 μ and four subsidiary maxima have been found in transmission. The results of Barnes, Brattain and Seitz in the range up to 15.6 μ have not been confirmed and no detailed fine structure has been found. The reflection spectrum of magnesium oxide has been fully investigated up to 25.5 μ, and a broad reflection maximum has been found extending from 17 μ to 21 μ, with subsidiary maxima at 14.02 μ and 23.8 μ.

From the data obtained curves have been drawn to show the behaviour of the two optical constants n and k in the range from 7 μ to 25 μ.