X-Rai/s from Boron and Carbon. 159 



connect ionizing potentials 13*5 volts and 25*4 volts with the 

 Ka absorption wave-lengths, considering that in the case of 

 these gases the first line of the series under consideration 

 can be called out at the radiating potential. This is, o£ 

 course, the fundamental difference between the mode of 

 excitation of emission-lines in X-ray spectra and the mode 

 of excitation of the lines of the principal and associated 

 combination series in the ordinary spectra.) 



The breaks at 34*5 volts and 24*5 volts for carbon and boron 

 have been associated with the " L "-radiation from these 

 elements. Fortunately the recent work of Millikan* enables 

 a check to be made in the case of carbon. Millikan found 

 that the ultra-violet spectrum of carbon vapour stops suddenly 

 at X 360*5, this being the shortest line of a small group of 

 lines separated from, and further in the ultra-violet than, 

 any other group of lines. The lines in this group are 

 X 386'4, X 384*4, X 372*1, and X 360*5, and may be regarded, 

 according to Millikan, as the La 3 , Lcq 2 , L/3, and Ly lines 

 in X-ray terminology. It is very significant that our 

 experimental value for the potential corresponding to the 

 L critical absorption wave-length is 34*5 volts, giving a 

 value of X358, which is just shorter than the wave-length 

 (X 360*5) of the Ly emission-line. Now, although in the 

 •ordinary X-ray region there are three critical absorption 

 wave-lengths in the L-series, the La! is the most marked 

 and has a wave-length slightly shorter than that of the 

 Ly emission-line. Thus our experimental result, combined 

 with Millikan's measurements, fits in well w r ith the general 

 relation. Unfortunately no information as to the extreme 

 ultra-violet spectrum of boron is available to compare with 

 our value, 24*5 volts (X505). If this result be correct, 

 boron should show no emission-lines whatever between about 

 X505 (the L critical absorption wave-length) and X83*7 

 {the K-series). Millikan considered it possible that in solid 

 carbon the generation of the L-ring rays might altogether 

 be prevented, as the four electrons which form the second 

 (L) ring in carbon " may form a portion of the space lattice 

 structure." The result of this investigation, however, shows 

 that even in solid carbon the L-rays can be excited. As the 

 L shell of electrons in carbon is the outermost shell, it seems 

 quite likely that its state would be affected to some extent 

 bv the presence or absence of neighbouring atoms ; and we 

 might reasonably expect a difference, perhaps very small, 

 between the type and wave-length of the L emission-lines 

 from gaseous carbon, as in Millikan's experiments, and the 

 * Millikan, Astrophysical Journal, Hi. p. 47 (1920). 



