Natural Classification of Elementary Substances. 469 



4472 of helium become visible. As the heat and pressure 

 increase, the lines of hydrogen and helium widen out, till at 

 a pressure of 1 5 to 26 inches the helium and sodium lines 

 are nearly of equal width and appear as single lines in the 

 spectrum. 



This experiment shows that, within certain limits, the dis- 

 tance between the components of D 3 is not to be taken as 

 the criterion of the identity of chromospheric and terrestrial 

 helium. 



As solar temperatures are much too high for the formation 

 of chemical compounds, the coincidence of the characteristic 

 lines of chromospheric and terrestrial helium clearly estab- 

 lishes the elementary nature of the gas or gases producing 

 these lines. Lockyer has indicated in several papers * that 

 the new gas obtained from uraninite by his distillation method 

 is a mixture, and that some of the spectral lines common to 

 the chromospheric and terrestrial gas may belong to two or 

 more elementary substances. Profs. Runge and Paschen 

 have made further progress in this direction by showing that 

 the gas from cleveite, after being diffused through asbestos, 

 gives two different spectra and is, consequently, a mixture f. 

 One of these gases, from its more rapid diffusion, is con- 

 sidered to be less dense than the other. From a comparison 

 of these spectra, the German physicists further conclude that 

 the denser of the two gases, producing D 3 , is helium proper, 

 but the lighter constituent has not yet received a new name. 



It may be of interest for me to note, here that, when a 

 strong induction-current is transmitted for a few minutes 

 through a newly filled vacuum-tube of reputed helium, the 

 red line 7066 disappears entirely from the spectrum, and all 

 subsequent differences in the intensity of the discharge fail 

 to reproduce it. 



In my classification of elementary substances in multiple 

 proportions of their atomic weights, each series or family is 

 considered to be condensations of the typical elements H, 

 H2, H3, H4, H5, H6, H7J. That some/if not all, of these 

 elements exist in a gaseous condition, and are of small specific 

 gravity, appears to me to be highly probable. M. Langlet, 

 of the University of Upsala, has found the density of the 

 cleveite gas to be 2*02 (H=l)§; while the more recent de- 

 termination of Pamsay has raised the density of reputed 

 helium to 2*13. Should, however, the conclusions of Lockyer 



* Proc. Koy. Soc, April 25, May 9, June 13, 1895. 



t Paper read before the British Association at Ipswich; 'Nature, 

 September 26, 1895. 



+ Memoirs Manchr. Lit. and Phil. Soc. 1878, 1887, 1895 ; Chem. News, 

 vol. xxxviii. pp. 66, 96, 107. 



§ Comptes JRendus, June 4, 1895 ; Chem. Soc. Journal, June 20, 1895. 



