EFFECT OF PRESSURE UPON ARC SPECTRA. 
131 
width chosen gave a definition sufficiently good for the finely reversed portions of the 
lines b 3, 64, cO, &c., Plate 4, to remain quite clear. The width was found to be 
(P022 millim., a larger value than that required for perfect definition, viz., 0'007 millim. 
W. E. Wilson* and Wilson and Fitzgerald! have observed the changes in the 
temperature and brightness of the carbon arc in different gases under pressures 
varying from 1 to 20 atmospheres. No very concordant results were obtained, as 
the absorption of the light by the vapour in the long observing tube was considerable 
and fluctuating, but the evidence tended to show that with the carbon arc there is 
no increase in the intensity of the light emitted from the crater when the pressure 
of the surrounding gases, whether oxygen, nitrogen, or hydrogen, is increased. 
During the course of the present experiments notes were made of the apparent 
intensity of the light at each pressure, and, according to visual observations, the 
brilliance of the image thrown upon the jaws of the slit increased gradually with the 
pressure until at 100 atmospheres it became painful to the eyes to observe it, the 
brightness apparently increasing more rapidly after 50 atmospheres than between 
1 and 50 atmospheres. 
It is to be remarked that with this increase of brightness there is little concomitant 
decrease in The necessary time of exposure. As the pressure was increased the 
spectral lines became very much wider and more diffuse, and in some instances were 
spread out over 15 Angstrom units; consequently, though there is an increase in the 
intrinsic brightness of the arc, the fact that the area over which the energy of 
vibration is spread is increased in a greater proportion, results in the intensity of the 
radiant energy received upon the photographic plate being less than that received 
when the total amount of light is less and the energy concentrated into narrow lines. 
The city mains at 100 volts were used as the source of supply, a reduction being- 
effected by a resistance frame. The current in the arc varied from 12 to 20 amperes ; 
this could not be measured accurately on account of the intermittent nature of the 
arc, nor could a higher voltage than 50 be used across the terminals, because at the 
moment of striking it a short circuit through the molten metal was produced, and the 
wiring of the building was not capable of taking more than 30 amperes. 
An interesting phenomenon, to which reference will be made later (p. 153), was 
observed in connection with the coating of the window with a black deposit derived 
from the arc ; this was very troublesome between 1 and 20 atmospheres, and it 
necessitated the removal and cleaning of the window after each exposure, but between 
20 and 30 atmospheres this ceased, and from 30 to 100 atmospheres the window 
remained perfectly clear. It may be that the particles shot off from the arc cannot, at 
high pressures, penetrate the dense envelope of air, or it may be that the hot vapours 
extend right from the arc to the window and consume the metallic particles. The 
former is the more acceptable hypothesis, since the latter condition can scarcely hold 
* Wilson, ‘Astrophysical Journal,’ II., 213, 1895; ‘Roy. Soc. Proc.,’ 58, 174, 1895. 
t Wilson and Fitzgerald, ‘Astrophysical Journal,’ V., 101, 1897 ; ‘Roy. Soc. Proc.,’ 60, 377, 1896. 
