ON ARGON AND ITS COJ^IPANIONS. 
5!) 
covered the cause of our failure to obtain gas rich in neon. Moi'eover, lit^uid nitrogen 
is very apt to become superheated, and to boil from the surface.* 
The gas contained in the gasholder was freed from oxygen, of which there was 
very little present, by burning with hydrogen, as described later; and from nitrogen, 
of which it chiefly consists, by a red-hot mixture of lime and magnesium. From 
about 120 litres of the light air, about 800 cub. centims. of argon comparatively 
rich in neon was obtained ; it was fractionated, in a manner similar to that described 
on j)- 55, mixing it with nitrogen instead of oxygen; for it was supposed that neon 
might have a somewhat greater solubility in the lower-boiling liquid. No improvement 
was noticed; hence the further fractional solution was carried out with oxygen as a 
solvent, the temperature being lowered to —205°, by causing the oxygen to boil 
under reduced pressure (see p. 52). The gas obtained by the last process, together 
with the lighter fractions from previous operations (34 and 35), were condensed with 
200 cub. centims. of oxygen, and the liquid was evaporated in six fractions (36 to 
41); of these, the last three (39, 40, and 41) were discarded, as consisting mainly 
of argon ; the gas was examined spectrosco})ically after removal of the oxygen with 
phosphorus. The first three portions were divided into five (42 to 46), and the last 
fraction was again discarded after examination. In each case, the last fraction 
consisted of the main bulk of the oxygen. The operation was repeated with the 
first three fractions (42, 43, and 44), giving four portions (47 to 50); tlie last 
fraction was mixed with No. 45. These fractions contained argon, and it was lioped 
that they would contain but little helium. 
The operations described are shown diagrammatically on the table wliicfi follows:— 
rSGi 
r42' 
-53 
Light fractions 
37 
- "1 
43 
r 54 
from air, mixed 
38- 
44 
55 
with 34, 35, and 
36. 
39' 
45- 
-50 - 
56 
1 
to 
Argon; 
-46 discarded. 
-41 
discarded. 
It was supposed that Nos. 53, 54, and 55 conttimed helium and no argon, and 
that 56 contained no helium, but a fair amount of argon. The first three of these 
were again fractionated, and a small quantity of argon was found in tlie oxygen 
remaining after the chief part of the gas had evaporated; this operation was 
repeated, and a trace of argon was again discarded, thus :— 
* We owe our thanks to Messrs. Gray Heuner and Liobury for kindly superintending this 
operation. 
I 2 
