oO 
PROFESSOR W. RAMSAY AND DR. M. W. TRAVERS 
Density. 
f 61 . . . . 7-24 
r57] 
53] 
1 
1 
58 
54 ^ 
1 
1 
59 J 
55 J 
60 
56 . 
62 . . . . 7-52 
63 ... . 8-27 
64 
Argon; discarded. 
.8'57 
The density of 63 was not greatly dilierent from that of 56, which contained 
argon, hence 63 also obviously contains that gas; these two were accordingly mixed, 
and liquefied along with 500 cub. centims. of oxygen ; the mixture was collected in 
two portions, 65 and 66 ; the first fraction, 65, had the density 7‘24 ; the last, 66, 
contained nearly all the argon in these later samples. 
It became a question whether it was possible to se})arate helium from neon by this 
process. It was obviously })ossible to remove practically all the argon ; and this is 
not surprising, since the boiling-points of argon and oxygen are nearly the same ; 
but an experiment to determine the relative solubilities of helium and of the impure 
neon which we possessed showed that there was no very great difference in favour 
of the latter. The solubility in oxygen at —205° of both gases was approximately 
80 volumes of gas in 100 of liquid oxygen under normal ])ressure. The presence 
of argon greatly increases the solubility, and, indeed, these gases appear to be freely 
soluble in licpiid argon. The results obtained by fractional solution in liquid oxygen 
were not sufficiently promising to induce us to continue the process. 
The next attenq)t to free the neon from helium was made by means of diffusion. 
The density of helium being in round numbers 2, and that of neon presumably 10, 
the rate of diffusion of helium should be about twice as great as that of neon. A 
small diffusion ap])aratus similar to the one described in the ‘ Proceedings ’ (vol. 62, 
}). 318, A in figure) was em])l(_)yed, and a preliminary experiment with oxygen and 
hydrogen showed that it was possible to reduce the quantity of hydrogen consider¬ 
ably in one operati(^n. Collecting the diffusate in three fractions t)f relative volumes, 
50, 45, and 16 })er cent, of the original, the last fraction contained only 16'5 per 
cent. <^f hvdrogen ; the percentage of hydrogen in the original mixture had been 54. 
All the available neon, thereTore, which had been freed from argon by fractional 
solution in litpiid oxygen, was united; the density of the mixed samples was 7'3. 
It was divided by diffusion into three equal amounts, 1,2, and 3, and these were further 
fractionated in the same manner as in conductiim' a fractional distillation ; that is, 
fraction 1 was Introduced into the diffusion apparatus, and the first portion which 
passed over was collected separately ; portion 2 was then introduced, and the first 
portion of the diffusate was addt'd tt) 1 ; the main bnlk was collected in 2 ; and 
fraction 3 was added to tlie gas remaining in the diffusion a])paratus, the first portion 
of the diffusate being collected in tube 2, and the remainder in tube 3. This 
