LONDON. Royal Society, June 11.—M. L. E. Oliphant: Electron emission from Langmuir probes and from the cathode of the glow discharge through gases. It is found that for potentials above about 600 volts the rate of increase of current to the probe is greater than that predicted by the original theory of Langmuir and Mott-Smith, and this is ascribed to electron emission from the electrode, which increases with the energy of impact of the positive ions. The energy delivered to the probes as heat has been measured by a compensation method which eliminates all corrections. From this energy and the potential of the probe relative to the surrounding space the positive ion current iv can be obtained. It is found that the ratio. of electron to positive ion current is independent of the energy of the positive ions up to that corresponding to a potential of 600 volts, and thereafter increases. The results are then discussed from the point of view of the angle of impact of the positive ions on the electrode surface, and it is pointed out that there must be an emission of electrons produced by agencies other than the impact of positive ions.—H. E. Watson, G. Gundu Rao, and K. L. Ramaswamy: The dielectric coefficients of gases, I. The dielectric coefficients of the five inactive gases and hydrogen have been measured at 25° and at - 190° or - 78° and compared with that of carbon dioxide. None of these gases except argon has an electric moment detectable by the method of measurement employed, 0.05 × l0-18 being an upper limit for krypton and xenon. For argon the figure appears to be 0.03 × 10-18, but this is probably a spuripiusi result. For the remaining gases it not more ttiah 0.015 × 10-18. Further investigations of possible sources of error have been made, and an approximate method of determining condenser distortion with pressure is described.—G. D. Bengough, A. R. Lee, and F. Wormell: The theory of metallic corrosion, IV. The effect on the corrosion of zinc of faster and slower rates of oxygen supply than those used for previous papers has been studied, and some comparisons made with the corrosion of mild steel. Complete curves are given showing the effect of concentration of potassium chloride and of potassium sulphate on the corrosion rates of zinc in tranquil conditions. The influence of depth of immersion between limits of 0.35 mm. and 100 mm. has been ascertained. An explanation has been found for the departure from linear corrosion rates after prolonged immersion in potassium chloride solution. A micrographic study of the form and distribution of corroded areas has shown that very highly purified zinc yielded eharacteristic etch pits, but no preferential crystal boundary attack; less highly purified zinc rarely showed etch pits of definite shape, but marked preferential attack on the crystal boundaries.