at the Pvoyal Institvtion, 1 900-1 907, 



;63 



to the stri^ stage it was necessary either to use a larger amount of 

 charcoal or to reduce the pressure of the gas to less than an atmo^- 

 sphere ; but on lowering the temperature of the air bulb to -210 , 

 by boihng off the air, the exhaustion in the tube just reached the 

 beginning of phosphorescence round the cathodes. Helium was 

 absorbed to but a very shght extent ; neon more readily. 



In tubes charged with air, it is easy to observe the gradual 

 disappearance of the spectra of oxygen, nitrogen, etc., in the order 

 of the volatility as the exhaustion proceeds ; the F line of hydrogen 

 and the yellow' line of neon are always noticeable, so that the test 

 for the latter gas is a very delicate one, as the amount of neon in the 

 air cannot well exceed yoio o- 



To obtain a satisfactory spectrum of helium, it is necessary to 

 enrich the air in the sparking tube. Fig. 6 shows the apparatus 

 used for the purpose. In an experiment in which 200 c.c. of air 

 were supplied to the tube D containing 15 grammes of charcoal 

 cooled by liquid air, on passing the residue on to the sparking tube 

 and examining it spectroscopically, the lines seen were the C and F 

 lines of hydrogen, the yellow and some of the orange lines of neon 

 and also *^the yellow and green lines of helium. On using the 

 residuary gas from a litre oi air, all the helium lines were seen, as 

 well as the yellow neon and the F line of hydrogen— from which it may 

 be inferred that sttoVw P^rt of helium by volume may be detected ._ 



As 40 to 50 grammes of charcoal at the temperature of liquid air 

 can absorb from 5 to G litres of air, it is easy to accumulate the more 

 volatile gases for spectroscopic examination by using the two 

 condensers E and D shown in Fig. 6. As soon as the charcoal in E 

 is nearly saturated and the less condensable gas has been transferred 

 to D by closing the cock K and opening J and I, the condenser E 

 is removed and rapidly raised to the air temperature, so as to expel 

 the condensed gas ; it is then replaced in the circuit and used as 

 before. 50 litres of air can be treated in this manner within a short 

 time : sparking tubes charged with the residuary gas show brilliant 

 spectra of all the more volatile constituents of air. 



A variety of interesting demonstrations of a similar kind illus- 

 trating the differential condensation of gases by charcoal have been 

 given" by Sir James Dewar, such as the separation of the gaseous 

 products'^ from minerals and radio-active bodies, the gases dissolved in 

 rain, well and river waters, together with samples taken from the 

 ocean. Speaking generally, the lower the boiling-point, that is to say 

 the less condensable it is, the less a gas is absorbed. 



Krypton and xenon are readily separated from air either by 

 passing a current of air (purified by cooling it with liquid air) over 

 charcoal cooled to - 183" or by covering a few hundred grammes of 

 charcoal with old liquid air and allowing this to evaporate in a 

 silvered vacuum vessel, then allowing a further quantity of the 

 retained air to escape from the charcoal at the temperature of solid 

 carbon dioxide, finally extracting the residual gas from the charcoal 



