82 
PEOFESSOR W. EAMSAY AND DE. M. W. TEAYEES 
the other after fractionating it from xenon (II.). The ratio between the refractivities 
of (I.) and (II.) were I'OOS to 1 ; the mean of all observations is 1’450. 
Spectra of the Inactive Gases. 
It is not intended to give a complete account of the spectra of the inactive gases; 
exceedingly accurate measurements have been made by Mr. E. C. C. Baly, with a 
IiOWLAXd’s grating, which will shortly he published. But to the general reader, it 
is of interest to gain some idea of the visual spectra, and these have been reproduced 
in Plate 3 from drawings made under the supervision of Professor Erdmaxx, from 
tubes furnished to him by us. 
The spectra of neon, krypton, and xenon cannot be detected in that of crude argon, 
even by photographic methods. 
The change of spectrum produced in argon by the introduction of a jar and sj)ark- 
gap is well known ; but less attention has been paid to the fact that helium, too, 
alters its spectrum, though much less markedly. The effect of the jar and sjDark-gap 
with kiwpton is almost to obliterate the spectrum ; the brilliant 3 "ellow and green 
lilies disappear. The spectrum of neon is hard!}" changed; but that of xenon is 
enormously intensified; indeed the plate shows the jar and spark-gap spectra of 
xenon. With the ordinary discharge, it is far less luminous, and the blue lines 
become only faintly visible; the lilue and green are much intensified Iw the inter¬ 
position. 
The colour of these tubes is very striking ; the yellow of helium and the red- 
purple of argon are now well known ; neon has a brilliant flame-colour, as intense as 
the yellow of helium ; kiypton has a pale violet tint, and is not .striking ; and xenon 
is .sky-blue, and witli the jar acquires a greener tint. 
The question of the identit}" of the .s])ecti'um of kiypton and that of the aurora 
borealis can be settled after the publication of Mr. Baly’s work. 
Conclusions, and Resume of Results. 
The densities of the inactive gases are respectively :— 
Heliiim. Yeon. Argon. -Krypton. Xenon. 
1-98 9-9G 19-96 40-78 64-0 
There are two lines of argument which show that these numbers, multiplied by 2, 
give approximately the atomic weights of the elements. 
(1.) The ratio between their .specific heats at con.stant pressure and at constant 
volume is 1-66. It is true that no measurements have been made \\-ith pure neon. 
