M. E. Goldstein on Spectra of Gases. 343 



If to a constant quantity of hydrogen air was admitted, so 

 that the partial pressure remained constant, then its lines grew 

 broader. 



Experiments having an exterior similarity to those just de- 

 scribed were made by Stearn and Lee, and were described in the 

 Philosophical Magazine for 1873 (vol. xl. p. 406). But while 

 the leading thought of the authors, which they believe finally 

 verified by their experiments, is that the cause of the enlarge- 

 ment of lines depends only on the tension of electricity which 

 passes at a given moment — therefore only on the absolute amount 

 of the total resistance in the closing-arc, — my experiments re- 

 sulted from the supposition that the above idea was wrong; 

 and this I believe they prove. With an equal total of resistances 

 the jar gives much stronger enlargements when air, as carrying 

 the greatest resistance, is mixed with the hydrogen, than when 

 the spark of the tube passes through pure hydrogen and the 

 greatest resistance in the shape of air is placed outside. In this 

 way I have obtained with a single jar still considerable dimness 

 of the lines with partial pressures in the hydrogen of less than 

 -j-^-g- millim. Pure hydrogen of such minimal density offers 

 insuperable resistance to the currents of induction-apparatus 

 hitherto constructed; in the above experiments, therefore, the 

 nitrogen, which gives a bright spectrum, must not be regarded 

 as a supplementary arc-closer, but as the sole conductor, through 

 which the hydrogen present is only heated. 



I conclude from the experiments that any required state of 

 the spectrum can be obtained at any density, however small, if 

 only the gas be exposed to a sufficiently high temperature. 



Wiillner has described in detail the constitution of a series of 

 gas-spectra. On account of the theoretical importance which 

 the discoverer attaches to a number of his results, I may be per- 

 mitted to enumerate here a few differing data, first of all for 

 hydrogen. According to Wiillner, the spectrum of the enlarged 

 lines, and the continuous spectrum resulting from this by raising 

 the temperature, stretches from Ha to Hy, and ends suddenly 

 at the latter line, which is somewhat enlarged. A further exten- 

 sion of the spectrum with further increase of the density of the 

 gas Wiillner even thinks improbable. I have seen very distinctly 

 H8, which Wiillner only mentions in the historical introduction. 

 This line, situated beyond Hy, is the first one which enlarges at 

 increasing density or temperature. Between Ha and HS and 

 beyond the latter line a perfectly distinct and quite continuous 

 spectrum appears when the enlargements are not yet at all 

 strong, perhaps corresponding at its most refrangible end to the 

 enlargement-limit of H S, although theoretically there is no suf- 

 ficient reason for this supposition. In any case the spectrum 



