Table of contents

The structure of turbulence in the 80-100 km region above the earth has been investigated by the simultaneous measurement of wind velocities at two points along a meteor trail. The fall off in correlation between the wind velocities at the two points, as their separation is increased, has been used to obtain a measure of the eddy size. The turbulence is found to be distinctly anisotropic, the vertical scale of the large eddies being 7 km whilst their horizontal extent is of the order of 150 km. The size of the smallest eddies is found to be approximately 50 m. Mean wind shears of 10 m sec^-1 km^-1 are found with a maximum of 140 m sec^-1 km^-1. No significant difference between day and night time turbulence is found.

The numbers of ice crystals produced by the explosive expansion of a small volume of moist air have been determined for various values of the terminal temperature and supersaturation. In clean air crystals, formed by spontaneous condensation followed by homogeneous crystallization, appeared only when the supersaturation exceeded 400% and the terminal temperature fell below -40°C. Thereafter, the crystal concentration continuously increased up to a maximum of 4×10⁶/cm³ and not to about 10¹¹/cm³ as reported by Schaefer. The appearance of crystals at temperatures above -40°c in room air may be attributed to the presence of ice nuclei. No evidence was obtained for a sudden increase in the crystal production or a change in their mode of formation at about -65°c as suggested by expansion chamber experiments of other workers.

An investigation has been made of the expansion coefficient of two-dimensional lattices - ionic lattices and the close packed hexagonal lattice with central forces. A survey has been made of the Grüneisen constants γ_ν of a representative set of vibrations in reciprocal space. From these a set of functions are constructed from which the variation of the Grüneisen constant γ can be calculated, and which also permit one to see the reasons for the variation. It is shown that the ionic lattices are characterized by a large spread in γ_ν values, this being in the main due to the lower of the two frequency branches. In the hexagonal lattice the distribution of γ_ν values is, in comparison, very limited and the Grüneisen assumption of a single constant can be justified in this case.

The spectral selective photoelectric effect in alkali metals is considered as a local field phenomenon taking place in a colloidal surface structure. Treating the colloidal particles as spheroids containing a mixture of bound ion core and free valence electrons, an expression is derived for the wavelength for peak photoemission which is compared with the experiments on hydride cells and on thin films.

Measurements of pre-breakdown currents in uniform field conditions in dry air in the E/p range of 25 to 45V cm^-1 (mm Hg)^-1 over the pressure range 60 - 700 mm Hg indicate the presence of appreciable attachment. Assuming dissociative attachment as the mechanism of negative ion formation, it is shown that the growth of current at constant E/p can be represented by the Townsend equation corrected for the removal of electrons from the swarm by attachment. From the semi-logarithmic plots of current against electrode separation, employing this equation the Townsend ionization coefficients α and γ and an attachment coefficient η are evaluated by curve fitting techniques. From the values of attachment coefficient, assuming a Maxwellian distribution, attachment cross sections for various mean electron energies have been evaluated and agree well with earlier measurements. Further, static breakdown potentials have been measured up to pd similar 1200 mm Hg cm. Employing the values of α, γ and η, and the breakdown criterion as derived from the modified Townsend equation, static breakdown potentials have been derived and agree well with measured values, thus confirming the validity of the Townsend build up mechanism.

Measurements have been made with a floating and a biased probe of the low-frequency noise which occurs in hot-cathode discharges through mercury vapour at low pressures. As previously reported by Cobine and Gallagher, and Martin and Woods, the spectrum is continuous but there are no superimposed peaks of oscillation. The spectra agree with that of the tube noise in their general shape. It has been shown that the primary electrons from the cathode are not essential for the noise to be recorded from the probe.

In recent papers static impact tests carried out on different diamond surfaces have been described. Here the quantitative results from these experiments are collected together and are then compared with theoretically derived values for the critical average stress for cracking to be initiated on the different diamond surfaces.

Both theoretical and experimental results show that the octahedral face is less resistant to cracking than the cubic face; but theoretical values obtained are of the order of 100 times greater than the corresponding experimental values.

This, together with other considerations, is thought to indicate experimentally some sort of flaw distribution in diamond.

The thermal changes accompanying the magnetization of specimens of iron and 0.5% silicon-iron and of a single crystal of 2.7% silicon-iron were measured by the direct method in fields between ± 1000 oersteds. Agreement with the Stoner and Rhodes theory was found for the range 1000 to 20 oersteds. The coefficient a of their notation was measured for each specimen; for the single crystal specimen the coefficient b was found from a comparison of the theoretical and the experimental magnetization curves taken at two different temperatures. An estimate made of the effect of including the anisotropy coefficient K₂ in theoretical calculations showed it to have little effect in the case of iron at room temperature.

An alternative, indirect, method of making magnetothermal measurements was devised for use with the single crystal specimen. In low fields measurements by the direct and indirect methods do not agree and it is not possible to give a reliable account of magnetization processes in these regions.

The magnetic susceptibility of three compositions of the ditelluride phase of the Fe-Te system have been measured between 100° and 1200°k. This temperature range can conveniently be divided into three regions: above 700°k the susceptibility variations of all the three compositions can be interpreted in terms of the phase diagram, while below 300°k the variations follow the Curie-Weiss law. Between these temperatures the susceptibilities deviate from the Curie-Weiss values and this has been explained in terms of an increasing concentration of defects. A discussion of the bonding in the three compositions has been given by considering the estimated values of the Curie constants.

The magnetic susceptibility of stoichiometric iron ditelluride has been measured between 15° and 100°K. The material appears to be antiferromagnetic with a Néel temperature at about 85°K. The curve is anomalous in that the susceptibility at 15°K is greater than that at the Néel temperature. This behaviour can be explained by a simple model of an antiferromagnetic containing some non-magnetic ions, provided that the concentration of the latter changes as a result of the ordering process.

The susceptibility of samples of FeTe_1.95 and FeTe_2.10 is also given.

Photoconductive and photoelectromagnetic effects were studied on p-type specimens of InSb with impurity concentrations varying from 10¹⁵ to 2 x 10¹⁷ cm^-3. Results were consistent with a simple theory, which assumes that the carrier relaxation time is independent of energy. Values for carrier lifetime were deduced. The fall in lifetime with increasing impurity concentration is explained as the result of two recombination processes, radiative recombination and recombination via traps.

Spectroscopic observations have been made photographically, over the wavelength range from 3800-8000 Å, upon the luminosity produced by 1/2 and 1 MeV proton beams in N₂, over the pressure range 63 μ Hg to atmospheric, and in O₂ at atmospheric pressure. The spectra excited in N₂ consisted entirely of bands of the N₂ First Positive (B³Π-A³Σ) and Second Positive (C³Π-B³Π) systems, and of the N₂⁺ First Negative (B²Σ-X²Σ) system. No atomic lines of HI, NI, or NII were found. In contrast, the spectra excited in O₂ consisted solely of lines of OI and OII. No bands of O₂ or O₂⁺ were found. The probable excitation mechanisms of these spectral features are discussed.

The angular correlation between the γ-rays produced in Panofsky reactions has been measured. The results exclude one of the two possible ways of introducing the π₀⁰ proposed by Baldin and Kabir. On the conventional assumption that there is only one π⁰, it is found that m_π^- - m_π⁰ = (8.90 ±0.14) m_e; the lifetime of the π⁰ must be greater than 3 × 10^-22 sec.

Broadening of the spark lines from a condensed discharge in neon and argon has been compared with their Stark effects and good correlation obtained, indicating that Stark effect is important as a cause of line broadening for other elements besides hydrogen and helium.

The Hulthén and Kohn variational methods are applied to a further consideration of the effect of polarization in the scattering of slow positrons by hydrogen atoms. A trial function is used which allows for virtual positronium formation. Although the results indicate a greater effect of polarization than those of previous calculations using a mathematically simpler but physically less appropriate trial function the effect is much less than seems to be required from the experimental results on the scattering of slow positrons by rare gas atoms.

If the light scattering by transparent spheres is calculated according to geometrical optics a large part of the scattered flux falls within a cone of half-angle |m - 1| radians around the forward direction, where m is the refractive index of the sphere relative to the surrounding medium. For many purposes it is this light which is of principal interest.

If I(θ) is the intensity scattered at angle theta and ΔE(θ) the flux scattered within this angle, the light externally reflected (I₁) and that transmitted without internal reflection (I₂) normally predominate in the forward-scattered light. For 0.75 ⩽ m ⩽ 2 graphical methods of obtaining I₁ and I₂, and ΔE₁ and ΔE₂, are given: this is made possible by plotting I₁, I₂, and ΔE₁, ΔE₂, against functions of single variables which themselves depend on m and θ. For light internally reflected once (I₃) or twice (I₄) analytical expressions are not given, but the values over which they range are indicated. Light internally reflected more than twice is shown to be negligible.

The accuracy obtainable is within about 1% in each case, and the angular range over which approximation is possible increases as |m - 1| increases as long as m < 1.65: if m > 1.65 the range decreases as m increases. For more accurate computation exact expressions for I₁ and I₂, in terms of m and θ only, are given. Unless m >> 1, I₂(θ) >> I(θ) - I₂(θ) for small values of θ, and consequently if accurate values of I₁ and I₂ are combined with approximations to I₃ and I₄ the overall accuracy is fairly high.

To simplify the computation, from the approximations given, of I and ΔE respectively two tables show the procedures suggested in each case.

An attempt has been made to detect effects due to non-zero pi-meson spin. The distribution of photons from the decay of neutral pi-mesons was measured and from the results it is concluded either that the spin is zero or that the pi-mesons observed were not appreciably polarized.

Corrections to 1958 Proc. Phys. Soc.72 901