[ROBERTSON] ELECTRODELESS DISCHARGE 153 



developed, with traces of the Balmer Hnes Ha and H^. On one 

 occasion Ha and H^ seemed to be absent, but as a rule, traces of 

 these lines were visible at this whitish stage. As the gap was gradually 

 lengthened, the whitish appearance gave way to a more and more 

 pinkish shade, while, at the same time, both the Balmer lines and the 

 secondary spectrum increased in intensity. With still greater ex- 

 citation, the shade on the inner side of the ring deepened until at a 

 certain stage (gap of the order of 7 mm.) the ring consisted of two 

 portions, the inner a brick red, the outer pink. Photographs, which 

 were taken with a two-prism Ladd spectrograph of small dispersion 

 directed successively at the red and at the pink portions, gave evidence 

 of the following: (1) Most of the secondary lines were relatively 

 weaker in the red portion of the ring. In particular, the group of 

 lines extending from a wave-length close to Ha (X 6563) to about 

 X 5680 were either absent altogether or extremely faint; while, with 

 the exception of a strong line at X 4928, lines between X 5680 and H^ 

 (X 4861) were very faint. (2) The continuous background, which, in 

 all other photographs extended from about X 4861 well into the violet, 

 was absent. (3) On the other hand, a group of some 15 or 16 second- 

 ary lines running from X 4316 (whgre there was a strong head) to 

 X 4136 came out as strongly in the photograph of the red portion as in 

 that of the pink. Unfortunately, although the exposure given the 

 "red" photograph was 23 minutes as compared with 18 minutes for 

 the pink, the intensity of the former was somewhat weaker. But, 

 while these statements are subject to some reservation on this account, 

 the eVidence points to the conclusions given. This will be evident 

 from an inspection of the accompanying plate, where a is the spectrum 

 of the red portion, h of the pink. The spectra are broken vertically 

 because exposures were made with a neutral-tinted wedge before the 

 slit, so placed that a part of the light passed below it, p>art above, the 

 remainder through it. 



The Blue Discharge 



With conditions which gave rise to the red-pink ring it was noted 

 that the whole of the region inside the ring had a distinctly bluish 

 cast — a purple blue, not the blue so easily obtained when mercury 

 vapour is present. The intensity of this luminosity was very feeble 

 and no attempt was made to photograph its spectrum. That this 

 glow is characteristic of hydrogen, however, there seems no doubt in 

 view of some observations recorded by Masson* a few years ago. 



*I. Masson, Nature, p. 503, Jan. 1, 1914. 



