the Constituents of Cleveite Gas. 299 



this we find to be the case in both our systems (see the 

 figure). We can distinguish in each of them two Subordinate 



o © o o o 



© O to <M O O 



io oq r— , — j ■— qo 



l/\ 



I 1 l'l fsl'l Tlalol |'| |'i|s|'| I | aia | I I | alp | I |'|3[o| ) |' | g|s I | | U|o| 1 | '| | 



Hy"d.Toge:a 



J... I I 



4? 



1.1. 







Series diverging from the same situation, of which the 

 brighter is the more closely spaced. Moreover, each system 

 contains a Principal Series whose lines are stronger than those 

 of the Subordinate Series, and which extends to shorter wave- 

 lengths. 



From the similarity of the two systems to the spectra of 

 the alkalies, we do not mean to infer a chemical relationship. 

 But we seem justified in ascribing the two systems, of series 

 to distinct constituents in the gas. Following the astronomers, 

 we should give the name of Helium to that constituent only 

 which emits the yellow line D 3 . 



If these views are correct, we may venture a further con- 

 jecture (based upon the analogy of other spectra) as to the 

 place of the two constituent gases in the series of chemical 

 elements. Since the gas is always present in the sun's 

 chromosphere, it is at all events of low density. This follows 

 also from the determinations of the density of the gas which 

 have been up to the present made, although they have led to 

 discordant results. Cleve has found 2*02, Ramsay 8'89 *, 

 referred to the density of hydrogen as the unit. Ramsay 

 regards the gas as monatomic, on account of the velocity of 

 sound in it : and, assuming this, the atomic weight would be 

 4*04 according to Cleve's determination, and 7'78 according 



Note hy Translator. — More recent determinations by Professor Ramsay, 

 after careful purification of the gas, are given in a paper read before the 

 Chemical Society on the 20th of June, but were not accessible at the 

 time Professors Runge and Paschen wrote their paper. The mean of 

 these determinations gives 2*181 as the density, with a probable error not 

 exceeding 0-05. See ' Nature,' of August 1st, 1895, p. 332. 



X2 



