300 Profs. C. Runge and F. Paschen on 



to Ramsay's. We regard Cleve's determination as the more 

 trustworthy, since his Geissler tnbes showed no argon lines, 

 and since Ramsay's material, as he states, was 1 not wholly 

 free from nitrogen ; and we shall assume that the atomic 

 weights of both the constituents lie at all events between 

 those of hydrogen and lithium. 



So far as it has been possible hitherto to analyse the spectra 

 of elements into series, the series, in the spectra of those 

 chemical elements which follow one another according to 

 atomic weights in a row of the Mendeleeff table, seem on the 

 whole to proceed with increasing atomic weights to shorter 

 wave-lengths. The reverse is the behaviour within a group 

 of related elements as, for example, Li, !Na, K, Rb, Cs. We 

 shall in the diagram place the spectra of our two systems 

 between the spectra of hydrogen and lithium in such an order 

 that in going from hydrogen to lithium the series shall proceed 

 to shorter wave-lengths. We should then say that the system 

 next to the series of hydrogen corresponds with the smaller 

 atomic weight rather than the system next lithium. 



Helium would, on this view, be the heavier of the two con- 

 stituents. This is in accordance with an observation we 

 made when filling the Geissler tube. The connecting tube 

 contained a plug of asbestos to keep from the exhausted tube 

 the heavier impurities, which diffused more slowly through 

 the plug. It now appeared, on first opening the cock, that a 

 greenish light was radiated, and in a small direct-vision 

 spectroscope the line 5016 was seen equal in brightness to 

 the yellow line. But, as more gas streamed in, the tube 

 became yellow, and the line 5016 was outshone by 5876. The 

 lighter constituent, whose brightest line in the visible part of 

 the spectrum is 5016, had diffused quicker through the plug. 

 Moreover, Deslandres has already expressed the opinion that 

 the two lines 5016 and 5876 belong to different elements, on 

 account of the varying ratio of their intensities in different 

 parts of his Geissler tubes. 



We see a further confirmation of the conclusion that the 

 two systems correspond to different elements in the circum- 

 stance that those lines of our spectrum which appear with the 

 most persistence in the chromosphere of the sun. all belong to 

 one system, viz. to the helium system ; whereas those lines of 

 the other system which up to the present have likewise been 

 observed in the chromosphere, occur much less frequently 

 according to Young. 



From the series, as presented in the diagram, a conjecture 

 may be formed as to the magnitudes of the atomic weights. 



