INTENSITY RELATIONS IN THE SPECTRUM OF HELIUM. 139 



series of Hydrogen is brought about by the admixture of Neon. The importance of 

 the mutual effects of gases on the intensity distribution in their spectra is 

 considerably enhanced by the fact that, in celestial spectra, the radiation from a pure- 

 gas is never in question, and if indeed the spectrum of any single element were mani- 

 fest, its presence would not disprove the presence of other elements which, though not 

 giving rise to perceptible radiations peculiar to themselves, might nevertheless exert 

 an influence on the distribution of intensity in lines due to other elements. 



A third and most significant condition which affects the relative intensities of 

 spectrum lines is the pressure of the gas from which they are produced. In Helium, 

 as is well known, this is peculiarly conspicuous ; the colour of the discharge, for 

 example, being green at low pressures. The existence of this phenomenon has, in 

 fact, been familiar for many years, and indeed was responsible at one time for the 

 erroneous view that Helium was a mixture of two gases. This misconception was 

 only removed by the demonstration that the effect in question was due to variations 

 of the pressure of the gas in the tube, but there has since been no quantitative 

 investigation of the nature of the changes which are known to occur. 



It is thus evident that there are a number of circumstances which modify the 

 distribution of intensity in the spectrum of an element, and that in order to obtain 

 further information it is desirable to investigate the simplest possible cases in which 

 the nature of any changes introduced into the method of excitation of the spectrum 

 can be followed in some detail. Such considerations have been the determining 

 factor in the particular conditions which have been selected for study in the work 

 described in the present communication. 



(II.) The Cathode Glow. 



A. source of light in which we already have some definite information with regard 

 to the electrical conditions is obviously presented by the glow around the cathode of 

 a vacuum tube. The radiations obtained from this source in the case of Helium are 

 of especial interest, for they include at once the " arc " lines, the spark line at " 4686," 

 and also the band spectrum which FOWLER has shown recently* to be of an unusual 

 type, inasmuch as the heads of the bands are not related by the law of Deslandres 

 appropriate for the usual band spectra, but by the Rydberg formula which had been 

 regarded hitherto as applying exclusively to line series. The presence of the " 4686 " 

 line in the same source is also interesting, as the appearance of a characteristic 

 " spark " line in company with a band spectrum is perhaps somewhat surprising, 

 although the existence of such a phenomenon shows clearly that, though the 

 conditions necessary for the production of these radiations may be different, they are 

 evidently at the same time not incompatible. 



* ' Roy. Soc. Proc.,' A, vol. 91, p. 208, 1915. 

 X 2 



