Table of contents

The spreading of drops of paraffin oil containing stearic acid on alkaline solutions of calcium chloride, during the course of which the acid molecules escape from the oil lens, is described.

A triode with a cylindrical electrode system had its grid maintained at a positive potential, and its plate at a small negative mean potential, with respect to the filament. If an oscillating potential was superposed on the mean plate potential it was found that plate current flowed, and the magnitude of this plate current was determined as a function of the grid potential. There were maxima of plate current for certain values of grid potential, and it was shown that these could be interpreted in terms of a fundamental or harmonic resonance between the period of the applied oscillating potential and the time of transit of the electrons between different pairs of electrodes in the triode.

A discussion of the discrepancies between Wilson's theory of semiconductivity and recent experiments on the subject is given; and on the basis of Wagner and Schottky's theory of Fehlordnung an extension of Wilson's theory, which may remove the discrepancies in some cases, is proposed.

With assemblies of five and four counters used alternately, measurements of cosmic-ray shower-frequencies from various thicknesses of lead were carried out (a) at sea-level, under a thin roof and under a screen of bricks 1 m. thick, and (b) underground at a depth of 30 m. in a clay soil. Comparison of the results obtained with the two assemblies enables conclusions to be drawn regarding the size of the showers. The results at sea-level under a thin roof are in accordance with the predictions of the cascade theory of shower-production: underground the interpretation of the results is less unequivocal, but it is shown that when allowance is made for the spatial association of the rays incident on the counting assembly (due to the proximity of the roof of the tunnel in which the experiments were carried out) there is no reason to suppose that cascade processes of shower-production do not predominate there, as they do at sea-level. The results obtained at sea-level under a screen of bricks corroborate this conclusion.

The modified (0001) vector projection for an insulin crystal is examined in relation to the C₂ structure proposed for the insulin molecule. It is found that this projection confirms the C₂ structure represented by the discron system 3s+3s'-o provided it is adequately loaded at its slits. Bernal's suggestion of an 18-discron set structure for the insulin molecule is shown to be untenable, since it is inconsistent with the z=o vector section.

An electron-diffraction study of the structure of silver films condensed in vacuo on rocksalt cleavage faces at about 200° c., and of the changes produced by heating these films in vacuo, has shown that the reflection and transmission patterns are in general agreement with Menzer's view of the twinned structure of such films. The films examined yielded patterns of a type indicating strong {111} extensions of the reciprocal-lattice intensity regions, but without marked development of [001] extensions. No conclusive evidence has been found of development of octahedral boundary faces of the films in addition to the twinning which occurs on these planes.

The positions of the diffraction spots and lines in patterns from films which had become distorted during the dissolution of the substrate and the mounting of the film showed that parts of the film had become considerably curved about axes parallel to the cube axes and cube-face diagonal directions in the film plane. These types of curvature are in agreement with the form of the substrate surface and the twinned structure of the films. General equations are derived for use in considering the geometrical features of rotation patterns yielded by cubic crystals.

Heating in vacuo at about 500° c. quickly converts the silver films into normal relatively perfect single crystals in the same orientation as that part of the initial film whose lattice was parallel to that of the rocksalt substrate. Such an extensive atomic rearrangement testifies to a high mobility of the silver atoms at temperatures much below the melting point; this has been inferred by previous investigators, but in this case is shown to occur within the lattice and not merely on its surface.

This paper discusses an objection which has been raised to the law of elastic failure under combined stress which was given in a paper read before the Society in 1900 - namely that, instead of the volumetric stress then used, the average volumetric stress should be employed and that only shear stresses should be associated with it. A resolution of a stress system into uniform tension and shearing stress, of a type which is advocated to-day and appears to offer advantages, is first discussed and is shown to be inapplicable as not being a complete representation. It is then pointed out that there are usually an indefinite number of shear-stress systems and only in certain cases are the shears unique. For the precise expression of a law of failure the indefiniteness must be removed, and the simple limitations available for the purpose are discussed and shown to be inapplicable.

Finally it is shown that the adoption of the suggested scheme could only be applicable to such materials as have a failure limit in simple compression less than three times that in simple tension. Thus it is incompatible with a law of failure which is to be of general application.

The rapid summation of Fourier series is becoming a more and more urgent problem in X-ray crystallography. A machine is proposed in which the fundamental sine curve of any amplitude may be represented by a number of sets of impulses. A switch can arrange these sets of impulses to correspond to a wave of any wave-number, and the sets of impulses then go into counters. Successive operations of the generator and arranging-switch result in the addition of Fourier terms, the terms being added for all values of the variable at the same time.

The reflection of the radiation emitted from an x-ray tube by a crystal in an x-ray goniometer gives origin to a pattern of lines standing on a general background. This background is constituted by white radiation emitted by the tube and regularly reflected from the crystal, and by radiation of different origins which has not been subjected to spectroscopic separation. The paper describes a method by which these two constituents of the background can be assessed separately, so that an accurate measurement of the intensity-distribution of the reflected white radiation becomes possible. This method is applied to examine the discontinuities of the background which one of us in collaboration with A. Baxter has indicated as a source of error in the evaluation of the intensities of reflection lines. Microphotometer records illustrating the conditions discussed in a previous paper are reproduced. The method is further discussed in its application to the determination of the dispersive change of the atomic scattering factor in the vicinity of an absorption edge; by its use f values for wave-lengths very close to the edge can be obtained. The results of experiments on the long-wave side of the zinc K absorption edge are compared with the wave-mechanical dispersion curve.

Experiments have been carried out to determine the sensitivity of photographic films to x radiation when exposures are made with the films maintained at temperatures obtainable with liquid nitrogen, hydrogen, and helium. The sensitivity at any temperature T is defined as the ratio of the exposure times required to produce images of equal density (0.25) by exposures made at room temperature and at the temperature T, the intensity of the radiation being the same for each exposure.

It is found that the sensitivity appears to decrease uniformly down to about 100° K., but below that temperature the decrease becomes less rapid, until, at temperatures below 20° K., the sensitivity remains practically constant at a value considerably higher than that found for visible light by Berg and Mendelssohn. The results obtained for X rays are compared with those found by Berg and Mendelssohn for visible light, and it is suggested that the relatively high value found for the limiting sensitivity in the case of X rays indicates that a considerable number of free electrons must be formed by the absorption of the X radiation while the film is at low temperature.

A new method of determining the electrical properties of dielectrics at centimetre wave-lengths is described. Although this is a free-wave method it requires only relatively small amounts of the substance and is particularly applicable to measurements on solid materials. The method has been employed at a wave-length of 12.0 cm. to measure the refractive indices and the absorption coefficients of some common laboratory substances. The results obtained are given and briefly discussed.

The investigation deals with the reproducibility of the standard platinum couple at the three points, viz. the freezing point of gold, silver and antimony, at which it is calibrated for the realization of the international temperature scale from 660° to 1063° C. The factors limiting reproducibility are considered under three heads, namely, electrical measurement, temperature conditions in the furnaces containing the ingots, and homogeneity of the thermocouple wires. The electrical measuring unit is found to be adequate for an accuracy of the order of ±0.01° C., whilst in connexion with the other two factors improvements have been made which render such an increased accuracy just attainable as an upper limit. In conclusion it is considered that the quality of the thermocouple wires, particularly in relation to their state of strain, constitutes the main obstacle to an increase in reproducibility beyond the normal limit of ±0.1° C.

A systematic study of the problem of the simultaneous excitation of two band systems in the same source has been made by measuring the gross intensity-distribution in the AlO blue-green and CN violet systems in the carbon arc. The results have been applied to the interpretation of the relation between the vibration temperature and rotation temperature in this particular source, in the light of the results obtained by Ornstein and Brinkman. The ratios of the excitation temperatures of the CN system to those of the AlO system measured simultaneously in the same source have been examined and compared with corresponding ratios for C₂ to CH derived by Lochte-Holtgreven under similar conditions. It has been concluded that in the case of simultaneous excitation, so long as the molecules remain the same, the ratio of their temperatures remains practically constant, irrespective of the conditions.