ON ARGON AND ITS COMPANIONS. 
57 
This observation introduced a new feature into the work ; for, although no great 
difficulty had been experienced in the separation of the light gas from argon, it 
appeared likely that the separation of two gases so nearly akin as helium and neon 
would be considerably more troublesome. And this proved to be the case. 
The next stage in the fractionation consisted in the separation of fractions 13 
and 14 each into three parts by mixing them with oxygen, and distilling the 
mixture. The oxygen was removed from each fraction separately ; and after 
siiectroscopic examination, the fractions IG, 17, and 19, whicli contained the greatest 
(juantity of helium, were mixed, as were also fractions 18 and 20, which consisted of 
nearly pure neon. The density of the latter, 23, was 8'3. The fractions 8 to 12, 
with 15 and 21, were also submitted to a further fractionation yielding about 
20 cub. centlms. of neon of density 10‘6, containing only a little argon. 
Up to this point, the gas had been divided into five portions 
i. (22), consisting of neon with much helium ; 
il. (23) ,, ,, trace of helium ; (D 
iii. (24) ,, ,, trace of argon; (D : 
iv. (25) ,, ,, much argon; 
V. (26) and (27) residues. 
Following this scheme, fraction 22 was divided into two })arts, and the fraction 
containing least helium (29) was mixed with fractions 23 and 24. This gas, mixed 
with oxygen, was licpiefied and divided into four parts, the lightest containing 
helium, and the heaviest, argon. The two middle fractions, 31 and 32, were mixed, 
and after addition of excess of oxygen, were again fractionated. The lightest 
fraction, amounting to about 10 cub. centims., had a density of 8‘3, and still contained 
helium ; fraction 35, amounting to 5 cub. centims., had the density 9’8 ; and a tliird 
fraction not numbered, for it consisted only of a single small bubble of gas, showed 
distinctly tlie spectrum of argon. We regarded the sample of density 9'8 as the 
purest; l)ut there is no doubt that it still contained Ijoth helium and argon. 
It appeared, therefore, an almost impossible task to separate these gases by frac¬ 
tionation. If it had been possible to licpiefy the mixture without addition of 
oxygen, separation might have been tlius accomplished; but fractionation under 
these circumstances can only be only an incomplete operation ; and we therefore 
resolved to attempt another plan. But for this, a larger quantity of neon was 
required, and a method of separation from the oxygen and nitrogen of the air, which 
we shall now describe, was devised. 
In descrilang the compressor (fig. 1), it was mentioned tliat the air drawn in had 
passed through a low-pressure purifier, filled with trays charged with slaked lime, 
in order to remove carbon dioxide. As a rule, the air which had escaped liquefac¬ 
tion was allowed to blow off through the nozzle Q and mix with the atmosphere. 
But by ca\isiiig it to blow through an india-rubber tube R, the furtlier end of which 
VOL. CXCVII. —A. 
=.• 8-3) 
= 10-G) 
I 
