22 THE ROYAL SOCIETY OF CANADA 



was probably hydrogen. Apart, however, from certain considerations 

 of intensity which he has brought forward there seems to be no con- 

 clusive evidence against the probability of these wave-lengths really 

 existing in the radiations capable of being emitted by helium atoms, 

 and in the case of X = 972 A.U. of this wave-length being emitted by 

 hydrogen atoms as well. As the wave-lengths X = 904.7 Â.U. and 

 X = 977.9 A.U. were obtained by us with the vacuum carbon arc and 

 not with the spark discharge in helium and hydrogen it would seem 

 that Millikan's view that these wave-lengths belong to the spectrum 

 of carbon is the correct one. 



Since the wave-lengths at or near X = 1037 Â.U., X = 1177 Â.U., 

 X = 1278 Â.U. and X = 1335 Â.U. were obtained by Millikan with the 

 carbon spark and by us with the spark discharge in helium and with the 

 vacuum carbon arc, but not by us with the spark discharge in hydrogen 

 and since from evidence adduced above there was probably a minute 

 contamination of the helium used by us with carbon in some form, it 

 seems practically certain that the radiation of these wave-lengths had 

 its origin in the atoms of carbon. It may be, however, that the wave- 

 lengths X = 1278 Â.U. and X = 1335 A.U. originated in mercury. This 

 would mean, however, that the helium was contaminated with 

 mercury vapour while the hydrogen was not. The radiations 

 having the wave-lengths X= 1134.7 Â.U., X = 1199.7 Â.U., X = 1494.4 

 Â.U.,X = 1647.2 Â.U.,X = 1649.9Â.U.,X = 1742.6 Â.U.,X= 1744.9 Â.U. 

 X = 1849.3 A.U. and X = 1942.4 A.U. it will be seen all come out with 

 strong intensity on our plates of the spectra of the spark discharge 

 in helium, but were not obtained on our plates of the spectra of 

 hydrogen and carbon. Neither were they observed by Millikan in 

 the spark spectrum of carbon. Moreover, wave-lengths X= 1134.7 

 A.U., X = 1199.7 A.U. and X = 1494.4 A.U. had previously been ob- 

 tained by McLennan^ with strong intensities in the spectrum of the 

 helium arc. 



In seeking for an explanation of the origin of these wave-lengths 

 one is led again to suspect mercury. A Gaede Mercury pump, it 

 will be recalled, formed part of the circuit of the circulatory system, 

 and it is possible that the liquid air-cooled charcoal traps Q and R 

 did not entirely prevent the mercury vapour from reaching the dis- 

 charge tube. A confirmation of this view is found in the fact that 

 wave-lengths are recorded for the spark spectrum of mercury by 

 Handke» at X = 1942 A.U. and by Lyman" at X = 1849.6 A.U., X = 1745.2 



2 McLennan, Proc. Roy. Soc. A, Vol. 98, p. 114, 1920. 



* Handke, Inaugural Dissertation, Berlin, Aug. 1909. 



* Lyman, The Spectroscopy of the Extreme Ultra-violet, p. 118. 



