﻿336 Prof. Trouton and Mr. Searle on Leakage Currents 



graph, the bands appeared much more distinct. The 'aver of 

 vapour in this case was only a millimetre or two in thickness. 

 The bulb was now opened, and the stem draw 11 down to a very 

 fine capillary, with a bore of perhaps 0*1 mm. It was then 

 re-exhausted and sealed. The capillary was placed in the 

 hottest part of the Maine of the blast-lamp and the rema nder 

 of the bulb heated with a small Bunsen flame. The vapour 

 in the capillary immediately became highly luminous and 

 almost white in colour, owing doubtless to the highei tempe- 

 rature possible with this arrangement. 



The spectrum of the capillary light was found to exhibit 

 strong bands, and resembled closely the fluorescence spectrum. 

 It was photographed with an exposure of 8 hours, the blast- 

 lamp being operated with an improvised water-blower. The 

 bands were not resolved into lines, and it is of great import- 

 ance to study it under higher dispersion. It can doubtlos be 

 photographed with the large concave grating, since an ex- 

 posure of several days can be given if neces-ary, ihe appa- 

 ratus requiring no attention. The fluoresc nee will- give 

 more trouble, as a powerful beam of sun- or arc-light is 

 necessary, either of which requires more or less constant 

 attention. 



The temperature-emission spect um is reproduced on PI. IV. 

 between the fluorescence and- absorption spectrum. 



The bands will be seen to be shifted towards the red, 

 with respect to the corresponding bands in the fluo escence 

 spectrum. 



XL. Leakage Currents in the. Moisture condensed on Glass 

 Surfaces. By Prof. Fred. T. Trouton, F.R.S., and 



C. Searle *. 



Tl^HE leakage currents over the surfaces of non-conductors 

 JL exposed to ordinary atmospheric conditions behave often 

 in a most perplexing way. The current is not only very 

 different on different occasions, but in addition may be 

 very different on the same occasion according to the direction 

 it is applied. This latter effect was exhibited by one of us 

 to the Physical Society during last Session, when it v\as 

 shown that the current may be many times greater when 

 sent in one direction over the surface of glass, tlian « hen sent 



* Communicated by the Authors. 



