LONDON. Royal Society, March 17.—Prof. C. S. Sherrington, president, in the chair.-Lord Rayleigh: The colour of the light from the night sky. Photographic exposures were made under coloured media selected for isolating various parts of the spectrum. Comparison with direct sunlight or moonlight showed that the night sky was of the same quality as these. Visual comparisons through coloured films showed that a blue film, which was equally bright with a yellow one against the night sky, was brighter against the twilight sky. These comparisons were not embarrassed by colour differences, because the light was so faint as to give purely monochromatic.vision. The requirements as regards colour and polarisation of the light would be satisfied if we regarded it as coming from an unresolved background of stars. They would equally be satisfied if we regarded it as due to sunlight scattered by meteoric matter.—R. O. Street: The dissipation of energy in permanent ocean currents, with some relations between salinities, temperatures, and currents. On the assumption of slow, non-turbulent motion a formula for the mean rate of energy dissipation in permanent ocean currents is obtained which, when integrated over the whole of the oceans, gives a dissipation at the mean rate of approximately 3 ×1018 iols ergs per second. Simple relations between the strength of the current, the salinity, and the temperature of the water are also found; satisfactory estimates of the currents in mid-ocean can thus be made.—S. Datta: The vacuum arc spectra of sodium and potassium. Definite improvements in the measures for the spectra of sodium and potassium have been obtained by the use of sodium and potassium vapour lamps as sources. With potassium an interesting combination pair indicating satellites to the diffuse series has been observed. The presence of potassium in the sun has been established, and some additional sodium lines have been identified with solar lines.— W. E. Garner and C. L. Abernethy: Heats of combustion and formation of nitro-compounds. Part i.: Benzene, toluene, phenol, and methvlaniline series. In this paper the heats of combustion of all the isomerides of the mono-, di-, and tri-nitro-toluenes and -benzenes, together with a number of nitro-derivatives of phenol and methvlaniline, have been determined, and the heats of formation and nitration calculated. The heats of formation and nitration of the isomerides of the di- and tri-nitro-toluenes and -benzenes show considerable variation, the values tending to a minimum when the nitro-groups are adjacent to one another or to a methyl group. The heats of formation in anv series increase to a maximum value with the introduction of the nitro-groups, which is reached in the toluene, phenol, and methvlaniline series (when svmmetrical substitution takes place) at the dinitro-derivative. The introduction of the methyl group into benzene modifies only slightly the shape of the curves showing the heats of formation of the derivatives, but the hvdroxyl or methvlaniline group has a much greater effect.—E. K. Rideal: The catalytic dehydrogenation of alcohols. Application of the approximation formula of the Nernst heat theorem to the equilibria: