ON AEGON AND ITS COMPANIONS. 
55 
After weighing the globe, it was again attached to the apparatus, and when the 
tube C had been exhausted, the stop-cock e was opened, and the gas was removed 
from the globe through the pump. The globe was then weighed empty, and from 
the difference of the two weighings the weight of the gas was determined. With a 
globe of the small size of those employed it was unnecessary to apply any correction 
for shrinkage under atmospheric pressure. 
The Preparation of Neon. 
The jDrocess which led to the discovery of neon was the fractionation of a large 
quantity of licpiid argon. A partial description of this has already been given in the 
‘ Proceedings,’ vol. G4, p. 188, where we have described the preparation and some of the 
properties of pure argon; and also in a note “ On the Companions of Argon,” vol. G3, 
p. 437. But for the sake of completeness, we shall briefly recapitulate the main points. 
The whole of the argon which we had made, probably about 15 litres, was liquefied 
in a bnlb immersed in liquid air Ijoiling under reduced pressure. The bull) was then 
placed in communication with a series of mercury reservoirs, and successive fractions 
of gas were collected ; the remaining liquid was then allowed to evaporate, and the 
gas was returned to the gasometer in which it had originally been stored. The first 
fraction of gas turned out to be rich in neon, and was afterwards found to contain 
helium. After rejjeated fractionation its density was reduced to 14. 
The main quantity of the argon was submitted to a second fractionation, and a 
further quantity of neon was obtained. A third liquefaction of the argon yielded a 
first distillate, of which the density was only slightly lower than that of argon, viz., 
19'05. After one-sixth of the argon had evaporated, the remaining fractions 
manifested no further change in density; and it was therefore assumed that the 
middle fractions consisted of pure argon. 
The gas of density 14 was refractionated by means of the apparatus shown in fig. 2. 
The gas was introduced from the gasholder a into the bulb b, which was immersed in 
liquid air, boiling under reduced pressure. The apparatus communicated with a 
Topler pump through the stop-cock; before admitting the gas, all air had been 
removed. On raising the reservoir attached to «, and applying an additional 
atmosphere’s pressure, the mixture of neon and argon condensed in the bulb h. The 
stop-cock on the gasholder a was then closed, and the liquid was allowed to stand for 
a few minutes, so that it might become homogeneous. 
On lowering the mercury reservoir, and opening the stop-cock, a portion of the gas, 
about one-fifth of the wliole, was allowed to boil oft' into the gasholder, the 
temperature of the bulb l)eing kept as low as possible. This fraction of the gas, 
which contained the greater quantity of the lower-l)oiling constituents, was trans¬ 
ferred to a tube, and a second, third, and fourth fraction were taken off* in the same 
manner. The last fraction was removed through tlie pump, and collected as usual. 
These five fractions, consisting of approximately equal quantities of gas, formed 
