J. Dewar — Separation of Gases from Air. 291 



enables the more volatile gases to be trapped and separated. 

 In this way some l/35,000th of the volume of the air liquefied 

 is collected as a gaseous mixture, having the composition 38 

 per cent of nitrogen, 4 per cent of hydrogen, and 58 per cent 

 of mixed helium and neon. After sparking to remove the 

 nitrogen and hydrogen a gaseous mixture of helium and neon 

 containing a little argon was obtained. This mixture had the 

 composition of 16 per cent helium and 84 per cent neon. In 

 both methods of treatment it will be noted the liquefaction of 

 the air was the essential preliminary operation, to be supple- 

 mented in the one case by the use of liquid hydrogen, in the 

 other by sparking to remove the nitrogen. The paper already 

 communicated to the Royal Society,* entitled " The Absorption 

 and Thermal Evolution of Gases Occluded in Charcoal at Low 

 Temperatures," in which the greatly increased power of occlu- 

 sion possessed by charcoal at low temperatures is proved, sug- 

 gested an inquiry into the limits of gaseous pressure reached 

 by such means of condensation. 



With this object a narrow tube CE, tig. 1, was sealed to an 

 ordinary spectroscopic sparking tube AB, and at the end E an 

 enlarged space was blown out capable of holding a few grams 

 of cocoanut charcoal. After the charcoal had been freed from 

 gases by heating and exhaustion and the poles cleared by 

 sparking during this operation, pure and dry gases like oxygen, 

 nitrogen, air, carbonic oxide, hydrogen, neon and helium could 

 be admitted at different pressures and the tube with its charcoal 

 chamber attached sealed off. 



On placing the charcoal end of the apparatus in liquid air, 

 the gas in each case was rapidly absorbed and the vacuum 

 produced reached the phosphorescent stage in all cases with 

 the exception of hydrogen, neon, and helium. A small 

 Crookes' radiometer, full of air at atmospheric pressure, with 

 charcoal tube attached, became quite active to heat radiation 

 when the charcoal was cooled for half a minute in liquid air. 

 To test the amount of exhaustion reached by the use of a given 

 weight of cocoanut charcoal, I sealed on a tube containing 30 

 grams to a large electric discharge tube of 1300 c.c. capacity 

 filled with air at atmospheric pressure. On cooling the char- 

 coal receptacle in liquid air the pressure diminished to 50 mm of 

 mercury. Repeating the same experiment but starting with 

 the tube initially at half an atmosphere, the exhaustion reached 

 was now beyond the striae stage. A further experiment start- 

 ing with one-fourth of an atmosphere gave a vacuum through 

 which no discharge passed. 



Finally, the 30 grams of charcoal were replaced by only 1 

 gram and the initial pressure was reduced to 3 mm of mercury. 

 * This Journal, p. 295 following. 



