THE PULSATION THEORY OF CEPHEID VARIABILITY.—The Monthly Notices for November last contain a paper on this subject by Prof. Eddington, who selects fourteen Cepheids, the light-curves of which are well known, to test the theory. The absolute magnitudes are deduced from the periods, using a diagram given by Mr. Shapley, and the effective temperatures and densities from a former paper of his own. Prof. Eddington finds that all the stars are in a gaseous state throughout their volume except the two of lowest absolute magnitude; he connects this with the fact that Mr. Shapley's diagram shows a linear relation between period and magnitude for the brighter stars, but a curve for the fainter ones. The radius of Y Ophiuchi, the brightest star on the list (abs. mag. -4), is given as 42,000,000 km., the mass being thirteen times the sun's; on the average, the semi-amplitude of the pulsation is 1/13 of the radius. Assuming an effective temperature proportional to the fourth root of the luminosity, the semi-amplitude of the temperature fluctuation is 1/12 of the whole. Prof. Eddington also deduces that with period 4.5 days should correspond spectral type F8½, and with period 30.8 days type G3½. These deductions are in fair accordance with Mr. Shapley's latest observational results.