J. Dewar — Separation of Gases from Air. 293 



gas in D was swept into the sparking tube, which was then 

 sealed off at Gr. 



This tube gave the hydrogen lines C and F, the neon yellow, 

 and some of the orange lines, along with the helium yellow 

 and green quite distinct. With the residuary gas extracted 

 from one liter of air I could see all the helium lines. On the 

 positive pole the neon yellow and the green of helium were 

 alone marked, while the negative pole gave both the neon and 

 helium yellow lines along with the helium green and the F of 

 hydrogen on the continuous spectrum. From this it would 

 appear that the spectroscopic test for helium is as delicate as 

 that for neon, and that 1/50,000 can be recognized. From 3 

 liters of air discharge tubes were obtained, giving the neon 

 and helium spectra associated with a brilliant ruddy glow 

 discharge. 



As 40-50 grains of charcoal can absorb at the temperature 

 of liquid air from 5-6 liters, it is easy to accumulate rapidly 

 the uncondensed gases in considerable quantities for spectro- 

 scopic examination. For this purpose I found it convenient to 

 use two charcoal condensers in circuit as represented in fig. 2. 

 After the charcoal in the first one marked E was saturated, 

 the stop-cock K was closed, while I and J were opened for a 

 short time so as to allow the less condensable gas in E to be 

 sucked into the second vessel of the same type D along with 

 some portion of air. The charcoal condenser E was then taken 

 out of the liquid air, and rapidly heated to 15° C. in order to 

 expel the occluded air. It was thus in a condition to repeat 

 the absorption. In this way 50 liters of air can be treated in 

 a short time. Sparking tubes filled from the accumulated gases 

 in D were very brilliant, showing the complete spectrum of 

 the volatile constituents of air. It is hardly necessary to 

 remark that after the little charcoal receptacle connected to 

 each of the sparking tubes has been cooled and thus all traces 

 of air absorbed, it can be sealed off, leaving the spectroscopic 

 tubes intact. The complete spectroscopic study of the products 

 must be left for further examination with Professor Liveing. 



The method I have described will be equally applicable to 

 the treatment of the gaseous products from minerals containing 

 helium, hydrogen, etc., and also to the radium products of a 

 similar kind. It seems even probable that the separation of 

 the less volatile constituents in air may be improved by a 

 slight modification in the mode of working. The behavior 

 of the gases from the Bath Springs has been examined. When 

 the gas containing l/1000th part of helium in what may be 

 regarded as pure nitrogen is subjected to charcoal absorption 

 exactly in the same way as the air was treated, no high vacuum 



Am. Jour. Sci. — Fourth Series, Vol. XVIII, No. 106. — October, 1904. 

 20 



