ON ARGON AND ITS COMPANIONS. 
(II 
operation was repeated three times; the lightest portion showed the helium spectrum 
witli brilliancy. During the third fractionation, the diffusate was collected in four 
fractions. The density of the heaviest was 8‘90, and of the lightest, G'24. The 
heaviest fraction was again diffused, and the first quarter was rejected ; the density 
of the remainder had risen to 9T8. In each case the gas was sparked with oxygen 
before it was weighed, so as to remove any accidental contamination. It was thus 
evident that 9T8 was a minimum density for neon. 
As it was evident that no complete separation of neon from Iielium could be 
effected by means of any process of diffusion or solution, these attempts were given 
up. The separation was finally accomplished by means of liquid hydrogen, the 
ojieration being conducted in precisely the same manner as the separation of argon 
from krypton and xenon, and the latter two gases from each other by condensing 
them with liquid air. 
The liquid hydrogen was made in an apparatus constructed by one of us 
(M. W. T.) wuth the valuable assistance of the laboratory mechanician, Mr. J. 
Holding. As it is described in detail elsewhere,suffice it to say that hydrogen, 
made from zinc and sulphuric acid, and submitted to no furtlier purification 
than by a rapid passage through a wash-hottle charged witli a solution of potas¬ 
sium permanganate, was compressed by the Whitehead compressor into a set of coils 
cooled by liquid air lioiliug under low" pressure ; the iiydrogen was allow"ed to ex})and 
through a jet to atmospheric pressure, at the end of a regenerator-coil, similar to that 
in the Hampson’s apparatus. The operation presents no special difficulty to those 
used to the Hampson liquefier, and the yield is rajiid and ample. Any impurities can 
be easily separated by filtration. For complete separation of neon and helium, 
100 cub. centims. is sufficient ; for licpiid hydrogen is fairly permanent, and does not 
evaporate much more ([iiickly tlian liquid air. 
On admitting the mixed gases into a bulb cooled by liquid hydrogen, all the gas at 
once entered, even under reduced jiressure, there'being sufficient neon present to 
dissolve and liquefy tlie helium. The 2 )ressure in the mercury gasholder w"as made 
very low", by depressing the inservoir, and the first fraction of gas I'e-entered the 
gasholder. It consisted of helium, mixed w"itli a trace of neon. It w"as set aside as 
not further needed. Tlie residue was removed through the pump, and re-liquefied. 
The stop-cock to the pump being opened, a few bubbles were removed. The second 
fraction, w"hich had a va})our-])ressure of only a few millimetres at the boiling-point 
of liquid hydrogen, w"as removed by pumping ; the residue, amounting to a few 
bubbles only, being kept separate, for it might possibly have contained a little argon, 
the vapour-pressure of winch w"as practically zero at that temperature. It was 
pumped oft* after the removal of tlie liquid hydrogen. It is evident from the very low 
vapour-pressure of neon at the boiling-point of hydrogen that helium must have been 
completely removed by this process. It is true that a little neon was lost by 
* ‘Phil. Mag.,’ lyOl, vol. 1, p. 411. 
