MR. E. C. C. BALY ON THE SPECTRA OE NEON, KRYPTON AND XENON. 185 
the platinum wire ; this capillary tube is then joined to a piece of ordinary glass 
tubing which is able to slip easily over the aluminium wire. In this way a covering 
is made into which the electrode is placed and then the blue enamel capillary is 
melted on to the platinum wire, but this must not be done within half-an-inch of the 
platinum aluminium junction. The glass sheathing over the^ aluminium is cut to 
a convenient length and the electrode is finished off and sealed into the vacuum tube 
in the usual way. 
The extraordinary heating of the electrodes forms a source of annoyance in woiking 
with these gases on account of the great cjuantity of hydrogen evolved from them. 
It is well known that in the process of filling a vacuum tube with any of the ordinary 
gases a considerable quantity of hydrogen is evolved from the electrodes, which veiy 
often masks the spectrum of the gas to be experimented with. This hydrogen can 
readily enough be removed by continued exhaustion while the electric dischaige is 
passing, and, if necessary, by washing the tube out with a small quantity of the gas in 
question. When this has been efficiently carried out entire freedom from contamination 
by hydrogen is secured under ordinary circumstances. If now into a vacuum tube, 
which has been carefully treated in this way, a small amount of one of the new 
monatomic gases be Introduced, a furtlier great quantity of hydrogen will be given up 
by the electrodes, and it is absolutely necessary that this be removed, as otherwise 
the spectrum of the new gas will be entirely masked by the hydrogen spectrum- 
This hydrogen can only be removed by alternately exhausting and admitting small 
quantities of one of the monatomic gases, this being done until the spectrum lines of 
hydrogen begin to weaken; three or four repetitions of this with argon are generally 
sufficient, provided that the electrodes are not very .^tout. The first tune a vacuum 
tube is filled, naturally the greatest trouble in removing the liydrogen is to be met 
with, although similar precautions must be taken whenever a tube is refilled. 
A curious effect is to be noticed in the splashing or volatilization of the aluminium 
electrodes with these gases. This is common enough with platinum and similai 
electrodes and has been termed liy Sir William Crookes electrical evaporation; as 
far as I am aware this has not been previously noticed with aluminium electrodes. 
It is this phenomenon which shortens the lives of the vacuum tubes containing these 
gases, both on account of the disintegration of the electrodes themselves and also on 
account of the slow absorption of the gas by the aluminium mirror deposited upon 
the walls of the tube immediately surrounding the electrode. This absorption of the 
gas when the tube is in continual use necessitates frequent refilling, especially 
because these gases must be illuminated under very reduced pressures, as will be 
presently explained. 
As regards the spectra of the gases, they all consist of bright well-defined lines 
similar to those of argon and helium. The most striking is the spectrum of neon, 
which consists almost entirely of very bright lines in the orange and red regions, the 
colour of the electric discharge through the gas is a magnificent orange. When a 
2 B 
VOL. coil.—A. 
