] 84 MR. E. C. C. BALY ON THE SPECTRA OF NEON, KRYPTON AND XENON. 
showed an extremely satisfactory agreement with Runge and Paschen’s or Kayser’s 
determinations. 
In apology for the length of time taken oyer the work I must urge certain 
unavoidable delays, of which the chief ones were due to the long exposures necessary 
to obtain the weaker lines and the short lives of the vacuum tubes employed; as wiU 
Ije explained below, these tubes will not withstand the long continued action of 
the electric discharge and therefore they require to be frequently refilled. One of 
the greatest delays arose from the fact that at one time the whole supply of xenon 
was used up and it became necessaiy to prepare a further quantity to finish this 
work. I take this opportunity of thanking Sir William Ramsay for his great 
kindness in undertakine: this for me. 
The Rowland grating has a focal length of 10 feet and is ruled with 14,438 lines to 
tlie inch ; all the measureinents were made in the second and third orders, with the 
exception of certain lines in the red region which were measured in the first order. 
The spectra were photographed upon Lumiere’s plates, the extra rapid and the 
A and B isochromatic series being used for the blue, green and red regions resjiectively. 
ihese plates possess a great advantage over other makes which were tried in that 
they give extremely well defined and fine grained images. 
In every case the gases were illuminated by the passage of the electric discharge 
through them when under reduced pressure, and several forms of vacuum tubes were 
employed, in all of which a capillary portion was viewed “ end on ” through a quartz 
window, which was cemented on with sealing-wax or Chatterton’s compound, 
neither of wdiich give off any vapour wdien cold. The most satisfactory results Avere 
obtained by sealing the electrodes into side tubes on account of the peculiar action of 
these monatomic gases Avhen subjected, under reduced pressure, to an electric dis¬ 
charge between electrodes. The latter become very hot and, unless special care has 
been taken in their manufacture, they rapidly disintegrate ; it is therefore preferable 
to have them hanging down in a vertical jjosition to guard against any possibility of 
their melting and falling against the glass Avails of the vacuum tube. This heating 
of the electrodes is very noticeable under ordinary circumstances AA'hen such metals as 
platinum are used, but in the case of the iicaa^ gases the effect is much more pro¬ 
nounced, for eAmn stout aluminium Avire is readily melted by a moderately strong 
discharge. In making the electrodes it has lieen found necessary to use aluminium 
Avire of at least No. 12 B.W.G., and to carefully guard against there being any soda 
glass sealed to the platinum. As usually made, an electrode possesses a sheath of 
ordinary glass in order to protect the junction betAAmen the platinum and the 
aluminium, this sheath being melted on to the platinum close to the aluminium. 
This, hoAAmAmr, invariably breaks AAdien used for any of the iieAA’ gases. The best AA'ay 
to make an electrode is as folloAA^s : some Amry stout aluminium AA'ire is taken and 
a small hole drilled in one end into whicli the platinum AAure is fixed. A capillary 
tube is made of some blue enamel glass, haAung a bore just sufiiciently large to admit 
