ON AEGON AND ITS COMPANIONS. 
G3 
gas containing nitrogen slowly over this mixture, heated to dull redness, else the 
temperature may rise so high as to cause the tube to fuse. There is no limit to the 
rapidity of the absorption, which is conditioned merely by the melting-point of the 
glass, for great heat is liberated by the reaction. The nitrogen is completely 
removed by this method, no trace being left l^ehind ; but as it is impossible to exclude 
all traces of carbonate or hydroxide, the gas collected always contains some hydrogen 
and carbon monoxide. These must he subsequently removed by passing tlie gas over 
copper oxide at a red heat, and absorbing in soda-lime the carbon dioxide produced. 
The first preparation of krypton was described In the ‘Proceedings,’ vol. G3, 
p. 405 ; the density of the sample of argon showing the spectrum of krypton 
brilliantly was 22’5. But we were surprised to find that on refractionating, the 
density of the gas, instead of rising, fell. This we found subsequently due to the 
fact that the gas \vas punq^ed out of the fractionating hidh without removing the 
jacket of liquid air; and under these circumstances, the xenon, of which the density 
is 64, remained in the bulb. On one occasion, the jacket was removed ; and the 
small remaining quantity was collected separately ; it turned out to possess a spec¬ 
trum totally dilfering from that of kryptou, and was subsequently named “ xenon.” 
The spectrum of krypton is one which is remarkal)le for its brilliancy, and for the 
fact that a small amount of gas, present along with argon, is easily recognised by the 
spectroscope; and this led us to believe tliat krypton possessed a density not 
differing greatly from that of argon. Ptepeated fractionation, moreover, did not 
appear to produce any marked change in its character or brilliancy. 
It would he tedious to relate in detail all the unsuccessful experiments which were 
made Ijefore the truth was discovered, that krypton is a gas present in very small 
cpiantity, l)ut possessing a density of 40‘8. Suffice it to state that the gas was 
fractionated more than twelve times, and that 24 determinations of density were 
made. We will, therefore, at once })roceed to descril)e tlie method l)y whicli krypton 
and xenon were prepared in a pure state. 
The gas remaining after the evaporation of the air used in various experiments, 
which we estimate as having been not less than 30 litres in all, was purified in the 
manner already described from oxygen and nitrogen. The remaining gas, of which, 
of course, argon constituted the main part, was liquefied and fractionated, so as to 
separate the heavier portion. The process employed in fractionation was as follows ;— 
The heavy gas was liquefied in the apparatus descrilied on page 51, at the temperature 
of liquid air. Wlien it was ricli in argon, some external pressure was necessary for 
licpiefaction ; hut later, the gases licpiefied, and when nearly pure, solidified, even 
under reduced pressure. I'lie lighter portion was allowed to boil liack into the gas¬ 
holder, and the remainder collected in one portion ; this operation was repeated, and 
the remaining gas is alluded to as fraction 1. 
The whole of this fraction, amounting to about 300 cub. centims., was introduced 
into the gasholder of the liquefaction ajiparatus, and was condensed at the 
