ON ARGON AND ITS COMPANIONS. 
65 
dealing with these gases ; but later on, it })ecame evident that we had in our hands 
gases possessing very different vapour-jiressures at the temperature of licpiid air ; for 
while argon recpiired additional pressure for its liquefaction, the vapour-pressure of 
krypton did not exceed 160 mihims., and that of xenon appeared to he less than 
could Ije measured. 
On account of the ease with which these gases solidified, we were obliged to take 
certain precautions usually unnecessary. If the bulb and stem of the liquefaction 
bulb were immersed in the liquid air l)efore the introduction of the gas, the gas did 
not solidity in the bulb hut in the stem ; and as the liquid air evaporated, it began to 
volatilise without proper fractionation. To overcome this difficulty, the vacuum vessel 
was held in a clamp, so that the surface of the liquid air just touched the bottom 
of the liquefaction bidb, and when the gas v/as admitted, it condensed in the bulb 
and not in the tube; the vacuum vessel was then raised so that the liquid air com¬ 
pletely covered the bulb. In fractionating krypton and xenon, the fractions of gas 
were taken ofi through the pump, as their vapour-pressure was too low to allow of 
them being collected into the gasholder. 
To return to the fractionation of the gases from which the argon was now nearly 
removed, we had at this point five fractions, viz., Nos. 24, 25, 27, 28, and 29. 
Fraction 25 was unfortunately lost through an accident ; but as the remaining 
fractions consisted of fairly pure krypton and xenon, and amounted to nearly 50 cub. 
centims., the loss was not of vital importance. The following table illustrates the 
method employed in the next series of fractionations. 
24 (D = 32-07)- 
c-30- 
-31- 
27-, r-30- 
—(D = 47-55) 
28- 
L31- 
—(D = 36-6) 
—(D = 41-44) 
r-3lJ 
29 (D= 56-04)—! 
L 32 -(L)=61-7) 
Fractions 27 and 28 were mixed together, and each of the three quantities of 
gas was in turn solidified in the fractionation Indi:), and separated into a ligliter and 
a heavier part. These fractions were mixed as is shown in tlie table, so as to olffain 
three samples of gas, two of which appeared from their sjjectra to I)e nearly pure 
krypton; wliile the third, of density 617, consisted chiefly of xenon. The two 
lighter fractions were separately solidified, and se})arated into two portions :— 
33 (one-third). 
(I) = 36'6) 
34 (two-thirds) (D = 40’82). 
35 (two-thirds) (D = 4073). 
(D = 41-44) 
36 (one-third). 
VOL. CXCVII.-A. 
K 
