1878.] in spectrum analysis. 209 



an induction coil without a Leyden jar were passed through them, 

 gave strong hydrogen lines when a large jar was interposed. A 

 bulb tube with magnesium wires filled with hydrogen at low 

 pressure gave in one half scarcely any spectrum but the F line of 

 hydrogen, while the other half gave the spectrum of acetylene. 

 They generally found hydrogen lines, and flashes of sodium (no 

 doubt from the glass) in tubes very much exhausted ; and they 

 conclude that impurities enter such tubes from sources hitherto 

 unsuspected. Tubes filled with oxygen obtained from silver iodate 

 have been found to give the spectrum of iodine, pointing to the 

 conclusion that chemical reactions occur at very low pressures 

 which are not produced under other circumstances. Generally the 

 authors conclude that the spectrum, of a gas in a rarified state 

 affords the most delicate test of its purity, and that it is to the 

 chemical problem of obtaining pure gases that attention needs to 

 be specially directed. 



November 18, 1878. 

 Peofessor G. D. Liveing, President, in the Chair. 



The following communication was made to the Society : — 



Dr Arthur Schuster, Ph. D., F.KA.S. Some results of the 

 last two total solar eclipses. (Siam, 1875, and Colorado, 1878.) 



Every scientific investigation passes through a preliminary state 

 in which a general survey of the facts is taken and by means of 

 which the most hopeful line for future inquiry is determined. The 

 important investigation on eclipses, which can only be carried on 

 at intervals of several years, for a few minutes at a time, may be 

 said to have just passed through that preliminary stage. The 

 present is therefore a fitting time for a general survey of what has 

 been done, and a discussion of what remains to be done. If, as 

 the title of this paper indicates, I shall refer chiefly to the two 

 last eclipses, it is because on them I can speak from, my own 

 experience, and because it is chiefly during the late eclipse that 

 certain changes were proved to go on in the sun's surroundings, 

 which unmistakably point to the line of inquiry which in future 

 will have to be adopted. 



1. Spectroscopic ohservations. 



The first observations on the spectrum of the corona were made 

 during the eclipse of 1868 ; but the results were not correctly 

 interpreted until after the eclipse of 1869, during which the 



