Table of contents

A description is given of edge and screw dislocations in a close-packed cubic lattice, which leads up to an account of the theory of Frank and Read of the origin of slip bands. Consideration is then given to a crystalline grain containing Frank-Read sources; it is shown that one has to expect cross-slip and the formation of deformation bands. It is shown moreover that the formation of vacancies by moving dislocations plays an essential role in these processes. Some speculations on the stability of deformation bands are given. It is suggested that the movement of vacancies (self-diffusion) plays an essential part in polygonization, recovery and steady-state creep, enabling dislocations in deformation bands to move out of their slip planes and so relieve stress. It is probable moreover that in these bands the stresses are several hundred times greater than the applied stress, and that this enables diffusion to occur at temperatures at which it would not be possible otherwise. Finally an account is given of low temperature creep of the type which does not involve recovery.

In this paper, some of the factors which influence the appearance of Lüders bands in mild steel are studied. It is shown that the Lüders band is adequately described by a uniform shear front, spreading over the specimen. In coarse-grained specimens, experiments indicate that this front becomes diffuse; diffuse bands are also present in strain-aged material, but here the diffuse fronts become sharper as the ageing becomes progressively longer.

An attempt is made here to explain the observed phenomena in the yielding and ageing of mild steel, described in two previous papers, in the general terms of a grain-boundary theory. On this hypothesis, a satisfactory explanation of the variation of the lower yield point with grain size may be developed. It is shown that strain-ageing must involve two processes: a healing of the grain-boundary films, coupled with a hardening in the grains themselves. A discussion of the possible nature of the grain-boundary film is also undertaken.

An account is given of investigations on the temperature dependence of the contribution from reversible processes to the magnetization of annealed and strained nickel wires This involves the measurement of reversible susceptibility κ_r over a range of field, and of temperature (-50° C. to +200° C.), and a rapid automatic method of recording (κ_r, H) curves has been developed. The effect of temperature on κ_r is discussed in relation to Becker's strain theory and reasons for the discrepancies are suggested. It is shown that the contribution of reversible processes to the total change in magnetization, integrated over the hysteresis cycle, increases with temperature and strain, from 15% for an annealed specimen at -50° C. to 85% for a strained specimen (17.1 kg. mm^-2) at 150° C. The effect is not uniform over the whole cycle and in the region of the coercive field a considerable part of the change in magnetization must be attributed to irreversible processes.

Determinations of the surface tension of phenyl ether have been made, using Jaeger's method, for the liquid in both the ordinary and the supercooled state.

No abnormal change with temperature, of the type already found for viscosity and dielectric constant, has been detected for surface tension as the liquid enters the supercooled region.

Rau has claimed that, by repeated freezing and thawing, the temperature to which a drop of water supercools before freezing can be lowered and, with special care, the freezing temperature can be reduced to -72° C. This is claimed to be the temperature of spontaneous nucleation of ice. At that temperature the ice forms in rectangular crystals which also melt at -72° C. This note describes a careful attempt to repeat Rau's work. Like Cwilong, we have been unable to do so except by deliberate contamination, and it is presumed that Rau's apparatus was faulty and his drops were contaminated by his cooling fluid. Information is given concerning the reproducibility of the freezing temperature of individual drops when repeatedly frozen and thawed. Rau's rectangular crystals have been observed in strong alcohol solutions but no information concerning their nature has been obtained

When clean, saturated air is subjected to a sufficiently large adiabatic expansion, small water droplets arise spontaneously from the vapour. An equation is derived for the growth rate of a droplet in an atmosphere of given supersaturation and temperature. This is used in conjunction with the Becker-Döring theory of nucleation to calculate the supersaturation and the concentration of droplets at successive time intervals during a cloud-chamber expansion, allowance being made for warming of the air by liberation of the latent heat of condensation. For large expansion ratios, the supersaturation attains a maximum and the droplet concentration a sensibly constant value before the end of the expansion. The computed value of this maximum droplet concentration is in acceptable agreement with the recent observations of Frey. The experimental results appear incompatible with Tolman's recent theory concerning the variation of surface tension with droplet radius. The predicted rates of droplet growth agree well with recent measurements made by Hazen.

The geometrical-optical treatment of Nielsen, which referred to a spatially extended source of refractive index unity, is extended to sources of any refractive index. Provided that the direction of emergence of the light from the source is nearly normal, it is shown that a change of refractive index from 1 to n modifies the light flux by the factor 1/n². This result is independent of whether or not a condensing lens is used between source and spectrograph. Its relevance in connection with observed intensities in Raman spectra is discussed.

Measurements have been made of the refractive index of amorphous selenium in the red and near infra-red spectral region. The absorption coefficient of amorphous selenium has been measured in the ultra-violet, visible and near infra-red spectral regions. The transmission coefficients of several single crystals of selenium have been measured and an estimate made of the absorption coefficient in the wavelength range 0.68 to 2.0 μ. The results show that the absorption edge of the crystalline material occurs at about the same wavelength as that for the amorphous form. A discrepancy between the value of the near infra-red refractive index and the dielectric constant of amorphous selenium indicates a further absorption band in the infra-red.

The focus of rays from a lens with spherical aberration appears as a spot surrounded by a discrete halo ring if a diaphragm with fine circular aperture is placed across the beam in front of the focus. The diameter of the halo allows an estimate of the magnitude of the aberration involved. The geometry of rays forming the halo is explained here by schematic drawings. The practical application of the halo test is illustrated by some examples. In particular, the negative spherical aberration produced by an electronic space charge in a saddle field lens is demonstrated.

The Langmuir probe technique is not suitable for measuring ion densities above 1 mm Hg pressure because the probe dimensions approach those of the ionic and electronic mean free paths. In addition to the invalidity of the Langmuir theory for this case, such a probe also causes a great disturbance of the discharge.

In this paper a detailed examination is made of the possibility of using a probe collecting positive ions as a means of finding ion densities

It has been found possible to calculate the potential distribution, outside a space-charge sheath around a spherical probe, if the radius of the sheath is known From this the current of positive ions to the probe may be found Curves are given to facilitate this calculation.

It is found that, depending on the ion concentration and pressure, there are two rather different sets of conditions around the probe. If the ion density is high (greater than about 10⁸/mean free path (in cm)) a thin sheath will form on the probe and a complete solution of the problem is possible Under such conditions the energies of the diffusing ions in the neighbourhood of the probe greatly exceed the thermal energies of the gas particles At the lower ion densities normally encountered, a thick `diffusion' sheath occurs and the problem can now only be solved if the sheath thickness is known The ion energies may or may not greatly exceed the thermal energies, depending on conditions.

These results contradict one of the basic assumptions made by Davydov and Zmanovskaja in their approach to the same problem They assume a thin sheath with the ions in thermal equilibrium with the surrounding gas As a result their conclusion is in conflict with the conclusion of this paper, that in most circumstances it is not possible to use a negative probe to measure ion densities unless an additional means of determining the sheath thickness is available.

The properties of a number of simple models of the upper atmosphere are tabulated. A study is made of the thermal equilibrium in the region of the F layers. The rate at which energy is gained from ionizing photons is estimated. Various loss processes are examined. It is found that the most important is conduction, and the next most important is probably emission by the magnetic dipole connecting the two low levels of the ground term of atomic oxygen. These are so effective indeed, that the energy source first considered appears inadequate to maintain the high temperature the upper atmosphere is generally supposed to possess. Alternatives are investigated. Sufficient energy might conceivably be supplied either by band absorption by nitric oxide, or by incoming interstellar matter: but neither source is attractive. Attention is drawn to the possibility that the original estimate of the contribution from ionizing photons may be much too small. This estimate is based on radio measurements It is shown that these only give a lower limit to the electron production rate: for certain ions may be removed so rapidly by recombination that they would escape detection even if their formation rate were greater than that of the dominant ions in the layers. The suggestion is tentatively made that heat is supplied to the upper atmosphere mainly by non-observed ionization. It is pointed out that owing to the possible existence of such ionization, the measurements of the radio scientists cannot be assumed to give the intensity of the solar emission beyond the Lyman limit. The temporal variation of the temperature is briefly discussed.