Table of contents

A variational principle devised for the Milne problem without capture has been extended to include capture. The theory is at present restricted to isotropic scattering. The flux in the energy-dependent problem is expressed in terms of the corresponding one-velocity solution having the same asymptotic spatial dependence and the one-velocity half-space Green function for an isotropic plane source. Then a variational principle for a quantity related to the extrapolation length may be written down and the emergent angular distribution determined by iterating the trial function. Results are presented for some simple scattering models.

Photoconductivity has been observed in indium antimonide at 10·6 μm wavelength using a high-power Q-switched laser as a source, and is ascribed to a two-photon absorption process. Variation in photoconductivity over the temperature range 30-300°K has been studied and comparison with the response at 5·3 μm indicates that the laser radiation substantially reduces the carrier lifetime in indium anti-monide.

High-yield photoemission, already reported for GaAs-Cs, is also observed for GaAs-Cs-O. Spectral distributions of yields from both GaAs and tungsten with surface coatings of Cs-O indicate that the work function of such coatings is at least as low as that of GaAs-Cs. Photoemissive yields from GaAs-Cs-O have approximated to the highest observed from GaAs-Cs and have exhibited greater stability. It is concluded that electrons, photoexcited in the GaAs, can diffuse through the coating of Cs-O without a significant loss in number.

Using the phase transformation at 68°C found in materials based on vanadium dioxide, devices exhibiting an S-type current-voltage characteristic have been produced. The characteristic has an unstable differential negative-resistance region which leads to fast switching (20 ns). Devices have been made by the fusion of mixtures of vanadium pentoxide with other oxides on to thin wires in a reducing atmosphere, by depositing vanadium dioxide particles between conductors using a dilute varnish suspension, and by reactive sputtering of vanadium in low oxygen pressures to give thin films.

Theoretical expectations based on a thermal explanation are in good agreement with the observed static and dynamic characteristics. A total cycle time of 700 ns has been obtained, most of which is derived from the turn-off time. On/off resistance ratios over 1000 have been observed.

Using ultra-high vacuum techniques, the sticking probability for chemisorption of hydrogen on a polycrystalline rhenium wire has been measured as a function of surface coverage in the temperature range 296-460°K. At 296°K the initial sticking probability was 0·24, with saturation at about 6×10¹⁴ atoms/cm², these values decreasing with increasing temperature. Slow desorption experiments revealed two binding states. The high-temperature state was found to obey second-order desorption kinetics with a heat of desorption of 30·5±1 kcal mole⁻¹ and a second-order rate constant of (7·5±3)×10⁻⁴ cm² per atom per second.

The suitability of molten salts for thermoelectric generation depends on their thermoelectric potential and the ratio of their electrical to thermal conductivities. The theories and experimental data on these properties have been examined and the thermoelectric potentials of cuprous bromide and cuprous iodide measured as 0·46 and 0·53 mv degk⁻¹. The figure of merit has been estimated for a few of the more promising salts. Molten-salt thermocells have certain advantages over semiconductor thermoelements as there are no contact junction, doping or thermal expansion problems, and they are inexpensive. But the mass transfer between the electrodes forms a major development problem.

Townsend primary and secondary ionization coefficients have been experimentally determined in pure argon and in argon mixed with up to 3% of ethane gas for a total pressure in the range 0·5-700 torr, corresponding to 3less, similarE/p0less, similar2000 v cm⁻¹ torr⁻¹ at 0°C. Ionization enhancement takes place and it is shown analytically that the major contribution to this is due to the Penning process. At an optimum concentration of about 0·66% of ethane in argon the primary coefficient is increased to forty times its value for pure argon. The presence of common impurities increases the coefficient by up to five times, whilst a similar concentration of electron-attaching gases lowers the coefficient by a factor of 10. Because of the quenching nature of ethane it is shown possible to reduce the secondary ionization coefficient of argon by as much as 10⁷ times; with the addition of less than 0·03% of ethane it is possible to increase the secondary coefficient at low E/p0 by a factor of 10 above that for pure argon. At the optimum concentration the sparking voltage is a factor of 2·2 below that for pure argon, whilst the presence of 0·000 03% of ethane still lowers the sparking voltage of pure argon, by a factor of 1·3. All the sparking voltages lie well below that for pure argon.

Recently published experimental results of Ellington and Ralph, showing the effect of the diluent gas on the electrical properties of alkali-metal-seeded rare-gas plasmas, are discussed in terms of an extension of the two-temperature conductivity theory first proposed by Kerrebrock. The theory explains the increase in the plasma electric field required to maintain a given current that is produced when helium replaces argon as diluent, with potassium as seed, over the seed pressure range 0.1-40 torr (the factor of increase falls from about 13 to about 2 between these values). The gas pressure is atmospheric. Neon is shown to occupy an intermediate position between helium and argon.

The theory developed is used to give an estimate, based on experiment, of the effect on K-He, K-Ne and K-Ar plasmas of varying the diluent pressure from 0.01 to 20 atmospheres. At low diluent pressures the electric field values required to produce a given current in the three mixtures converge and level out, while at high pressures they diverge and increase rapidly with rising pressure. At high pressures helium requires field values about 20 times as high as for argon, and neon about 4 times as high. It is concluded that argon and neon seeded with potassium (or caesium) appear to be suitable for use as the working fluid in a magnetohydrodynamic generator operating at pressures up to 10 or 20 atmospheres, but that helium appears to be unsuitable as a diluent at pressures above 2 or 3 atmospheres.

In this paper the authors have investigated analytically the use of the photoelectric effect as a means of enhancing the electrical conductivity of dust-laden gases and have shown that it is not a very efficient process for the range of temperatures (> 900°K) useful for magnetohydrodynamic generation of power. Some numerical results, for the special case of argon laden with barium oxide dust, have also been presented in the discussion.

Measurements on the scattering of visible light in the angular range θ=8-145° using monodisperse spherical particles of polystyrene and polyvinyltoluene in aqueous suspension are reported for x values of 1, 5, 10, 15 and 20, where x is the ratio of the sphere perimeter to the wavelength of the incident radiation in the surrounding medium. There is, in general, good agreement between the shape of the experimental curves and the curves obtained from theory, but there are apparently significant differences in the absolute values, although these differences are in accord with the results of other workers. The dependence of angular position of a scatter peak upon x is investigated and used to determine particle size.

Operation of an electron microscope as a high-resolution microdiffraction camera allows investigation of the fine structure of the zero-order beam produced on transmission through a small polyhedral crystallite. The advantages of this type of observation for the direct determination of the Fourier coefficients of lattice potential are discussed.

The fraction of magnetic thin-film flux which links a rectangular coupling loop is calculated for rectangular, square and circular film configurations. The calculation is based on a magnetic line charge model for the film external flux distribution. The results are presented in a graphical dimensionless form.

Various types of fracture are discussed on the basis of energy considerations. A model of a crack with associated plastic zones extending from the tip of the crack and localized in a narrow layer is developed. According to this model the work lost in the plastic deformation process, due to crack extension, replaces the work done against cohesive forces; this appears to be the most important factor determining fracture. This plastic work is not a material constant, as is surface tension in the Griffith theory, since it depends on the crack length as well as on the properties of the material and on the way in which the external load is applied (three different schemes of loading are considered).

After the consideration of the potential energy is introduced, some interesting features of fracture, such as deceleration of crack spread, fracture arrest and `stickslip' motion of a crack in a non-homogeneous medium, are discussed.

The observed output z(x) of a physical apparatus is assumed to be a weighted integral of the input f(x), the weighting function g(x) of the apparatus being known. This paper describes a convenient numerical method of determining f(x), based on expanding z(x) and g(x) in Fourier series, having regard to the noisy nature of actual observations.

Experiments on the prompt neutron-induced cavitation effect in tetrachloroethylene are described. Cavitation bubbles were detected by the fast acoustic signals they emit, probably during the collapse phase. Using a pulsed fast-neutron source phase locked to the acoustic field, and a time analyser, information was obtained on the history of bubbles which were nucleated at a known phase of the sound pressure field. Some other properties of the neutron-induced cavitation effect were also investigated using both the pulsed source and a Po/Be source.

Experimental data have been collected from individual cavitation bubbles in degassed tetrachloroethylene in an attempt to elucidate the bubble motion. These results are compared with those computed from a modified Noltingk-Neppiras equation.

As a result of collisions between excited Hg atoms in the ³P₁ and ³P₀ states positive molecular ions Hg₂⁺ and electrons are produced. From the observed saturation currents the associative ionization cross section is found to be about 5×10⁻¹⁴ cm², which compares favourably with theoretical estimates. Using this result the cross section for the reverse process, dissociative recombination into excited atoms, has been determined thermodynamically. The corresponding recombination coefficient agrees with that found in earlier experiments in a decaying Hg plasma.

The sound field in a rectangular enclosure was investigated at frequencies above and below the lowest natural modes with varying wall absorption. The results obtained illustrate the substantial non-uniformity existing at frequencies below the lowest natural mode, and the increase in uniformity produced by walls with low absorption.