SECTIONAL TRANSACTIONS.— A. 315 



useful in determining the abundance of the rarer elements in the earth's crust and in 

 meteorites. This method combined with chemical concentration was used in the case 

 where the proportion of the element is less than one in ten thousand. 



Department of Cosmical Physics. 



Dr. H. Knox-Shaw. — Hornsby's Meridian Observations at the Radcliffe 

 Observatory, Oxford. 



The reduction of these observations, which has been proceeding during the last 

 four years with the collaboration of Dr. J. Jackson, is nearing completion. Attention 

 has been concentrated on the two periods 1774—1784 and 1790-1798. The observations 

 in Right Ascension of the Sun and ninety-one standard stars have been reduced ; 

 those of the Moon and Planets have been done for the first period, but with few 

 exceptions they await the computation of tabular places for comparison. The work 

 on the Declinations is much less far advanced, but it is hoped to be able to give definitive 

 results for one period. 



The R.A. observations of the sun indicate progressive changes amounting to 3" 

 in the sun's longitude as compared with Newcomb's tables, but until this can be 

 checked by the observations in declination it must be received with caution. Those 

 of the stars confirm Boss' system of proper motions rather than those of Eichelberger 

 as regards variation both with right ascension and with declination. 



Prof. W. DE Sitter. — Satellites of Jupiter. 



Dr. W. T. LuYTEN.^TAe New Grootfontein Meteorite. 



Friday, August 2. 



Prof. A. Fowler, F.E.S. — The Spectra of Carbon, Nitrogen, Oxygen and 

 Silicon at successive stages of lonisation. 



Lines of these elements appear prominently in the spectra of the hotter stars, and 

 their correct assignment to atoms in various stages of ionisation is of importance in 

 theoretical discussions of stellar atmospheres and in the interpretation of the spectra 

 of nebulae. The paper gives an account of some of the experimental methods by which 

 the lines emitted as an atom loses successive electrons can be identified to some 

 extent, and of the way in which analysis of the spectral structures completes the 

 identifications. In accordance with Bohr's theory, the Rydberg constant, designated 

 by R for neutral atoms, becomes successively 4R, 9R, 16R... when one, two, three... 

 electrons respectively have been removed by the exciting agency. As the series 

 constant increases, the principal lines occur with shorter wave-lengths, and their in- 

 vestigation requires the use of the vacuum spectrograph. In the frequent absence of 

 sufiicient data for the calculation of the series constant, the regular and irregular 

 doublet laws, as shown by MiUikan and Bowen, provide a valuable aid in the classifica- 

 tion of the lines. The final results for applications to stars are expressed in the form 

 of ionisation potentials corresponding with the various spectra, or as excitation poten- 

 tials for individual lines. 



Mr. E. H. Fowler, F.R.S. — The Application of Prof. A. Fowlers Experi- 

 mental Data to Astrophysical Problems. 



The various spectra of Si, C, N, and O are specially important in the study and classi- 

 fication of the B and type stars, which are the hottest types known to us. The 

 maxima of the fines of these spectra which are fixed by analysis of the stellar spectra 

 enable the temperatures of the corresponding stellar atmospheres to be roughly 

 deduced. This work is described, and also more recent work by Prof. MUne which 

 has greatly improved the theoretical basis. A temperature scale of B and type 

 stars is given which is based on the latest theory. The interesting problem of the origin 

 of the so-called nebuhum lines is also discussed, as they appear to be Unes of O and N 

 not emitted under ordinary conditions. 



