Abstract
LONDON. Royal Society, March 11.—J. J. Thomson, president, in the chair.—W. G. Duffield, T. H. Burnham, and A. A. Davis: The pressure upon the poles of metallic arcs, including alloys and composite arcs. In a previous communication (Phil.. Trans., A, ccxx., p. 209, 1919) the authors showed that the poles of a carbon arc behaved as though they repelled one another, and methods were described by which the pressure upon each pole could be measured. Reasons were given for attributing this effect to the reaction consequent upon the emission of electrons from the poles under the influence of thermionic or photo-electric action. The present experiments relate to arcs between iron, copper, and silver terminals, the rate of variation of the pressure with current density being measured for the anodes and cathodes. The pressures were greater than in the carbon arc, that within the copper arc being the largest. Assuming that the pressure is due to the projection of electrons, a comparison between the kinetic energy of the electron and that of the metallic atom at the temperature of the poles showed sufficient agreement of suggest that the electrons before projection were in thermal equilibrium with the metaf of the pole. The reactions upon electrodes composed of an alloy of silver and copper were also measured, likewise those within an arc between a silver and a carbon pole. In this case the pressure was determined mainly by the material of the pole under examination. The problem of the mechanism whereby a gas may be heated is briefly discussed. Some account is also given of the variation in the potential difference between the poles when the material of one is altered.—J. H. Vincent: Further experiments on the variation of wave-length of the oscillations generated by an ionic valye due to changes in filament current. Eccles and Vincent have found that in an oscillatory circuit maintained by a thermionic valve with a grid coil coupling, the wavelength has a maximum value for a certain filament current. This effect is studied further in this paper. In order to vary the filament current, rheostats were designed and used in which the change of resistance was unaccompanied by any sensible change in the self-induction of the filament circuit. The methods of measuring the change of wave-length due to the variation of filament current were different from that emploved by Eccles and Vincent, but it was found that the results obtained were independent of the particular method by which the wave-length was studied. It is suggested that changes in several of the variables of a valye-maintained circuit produce effects of the same sien on the wave-length and the amplitude of the oscillations. The wave-length and amplitude decrease with the decrease of the grid voltage or of the plate voltage. They also decrease when the coupling of the grid coil with the main oscillator coil decreases. Increasing the resistance in either the condenser branch or the induction branch of the main oscillating circuit lessens the amplitude and wave-length; while altering the filament curretit in either direction from that giving the maximum wave-length gives also a decreased amplitude.—H. A. Daynes: The theory of the katharometer. A historical introductory note by Dr. G. A. Shakespear gives a description of the katharometer and an account of its development by him for hydrogen purity measurements and similar work in connection with lighter-than-air craft. The paper discusses the conditions which determine the temperature of the hot wire in the katharometer cell, and shows that loss of heat by conduction through the gas is the most important factor, convection and radiation being quite unimportant. Equations are given expressing the experimental law of heat loss in a single katharometer wire, and these are applied to the case of two wires in parallel in the arms of a Wheatstone bridge. These equations are then used to show what are the conditions for greatest sensitiveness and precision in various cases arising in practice.—H. A. Daynes: The process of diffusion through a rubber membrane. The nature of diffusion of gases through rubber membranes is discussed in the light of some recent work. This all points to a simple process, determined by the case of diffusion through the rubber, and by the absorption of the gas by the rubber. This Is introduced mathematically into the problem of diffusion through a membrane. The unsteady state is considered, in which the membrane, after being exposed to air, is suddenly exposed on one side to, say, hydrogen, and the rate of emission of hydrogen from the other side calculated. The passage of gas through the material is treated purely as a diffusion problem, the boundary conditions only being determined by absorption. It is shown that measurements of the permeability of a membrane and of the lag on reaching a steady state are sufficient for the determination of both absorption and diffusion constants. Experiments are described in which these conditions are fulfilled. The measurements of the diffusion are made by means of a katharometer. From these experiments the constants of diffusion and absorption for hydrogen, nitrogen, oxygen, carbon dioxide, nitrous oxide, and ammonia are determined. Temperature coefficients for the constants are gfiven for liydrogen, and the high temperature coefficient of permeability of rubber is shown to be due chiefly to the high temperature coefficient of the diffusion constant. The extraordinarily high permeability of rubber to carbon dioxide, ammonia, etc., is shown to be due entirely to the high absorption. A relation is also suggested between absorption and critical temperature of the gas.
Article PDF
Rights and permissions
About this article
Cite this article
Societies and Academies. Nature 105, 121–123 (1920). https://doi.org/10.1038/105121a0
Issue Date:
DOI: https://doi.org/10.1038/105121a0
This article is cited by
-
Zur Frage der Blutdrucksteigerung
Zeitschrift für Die Gesamte Experimentelle Medizin (1924)