710 Prof. R. W. Wood on Hydrogen Spectra 



process exhibits unique spectra when exhausted and charged 

 with other gases. It' we introduce a little pure nitrogen or 

 air, sufficient to give a dark-space o£ say 2 mm., we obtain 

 a discharge of a very pare golden-ye low colour, which 

 resembles in colour the discharge in pure helium. Spectro- 

 grams show that what is commonly known as the second 

 positive spectrum (violet and ultra-violet bands) has almost 

 completely disappeared, while the first positive spectrum, 

 consisting o£ strong bands in the red. yellow, and green, 

 appears at great intensity. It is this peculiarity that gives 

 the discharge its unique brilliancy and colour ; the red and 

 green are in the proper proportion to produce the sensation 

 of subjective yellow, to which is added the sensation pro- 

 duced by the real yellow. This remarkable yellow discharge 

 usually appears when a small crack develops in one of the 

 bulbs. The colour is absolutely different from the purple- 

 yellow discharge which occurs if a leak develops in an 

 ordinary vacuum tube. The visible spectrum of the tube in 

 the white stage is the same as that of an ordinary hydrogen 

 tube, except for the omission of the Balmer lines. In my 

 earlier communication I drew attention to what appeared to 

 to be a pair of lines in the ultra-violet between H y and H 6 ~, 

 which appeared only when the tube was in the white stage. 

 The only photographs of this region available at the time of 

 writing the p;iper were made with a small quartz spectro- 

 graph. 



Spectrograms made with the grating show that we are in 

 reality dealing with a very complicated group of lines, which 

 at two points are closely bunched together, giving the 

 appearance of lines under low dispersion. 



This group is shown on PI. XXI. fig. 2, the group 

 being marked with a ^ — '. 



Hy and Hj appear at the extreme ends of the spectrum. 

 The two upper spectra (a and b) show the spectrum just 

 before the white stage is reached, with the secondary spec- 

 tra n very strong ; spectrum c the white stage, with Hy and 

 H<* absent, and the new group, which we may call the 

 X group, strongly developed ; spectrum d, taken a few 

 minutes later, *.\ith the X group much weakened and the 

 B diner lines reappearing. Pig. 3 shows the X group (in 

 coincidence with the iron* spectrum) more highly enlarged, 

 the position of Hd being indicated by a black murk. 



I can find no mention of this group of lines in the litera- 

 ture of the spectroscope, and as it is of great importance to 

 ascertain its origin, in connexion with any theory of what 

 goes on in the tube when the white stage arrives, I hope 



