440 
DR. W. *AIANSERC4H YAULEY ON THE PHOTO-ELECTRIC 
potential again causes a rapid increase in the current. The experiments were 
conducted solely in air and the results obtained are purely qualitative. 
Owing to the complicated nature of the connection between the photo-electric 
current and the potential dift’erence between the electrodes, it is not possible to 
compare the currents for different pressures or gases under any sunple definite 
conditions, as would have been the case if a saturation current had existed for 
ionisation produced ijy the photo-electric effect as it does for that produced by 
llontgen rays. Under the circumstances the only satisfactory method of comparing, 
for example, the currents in a gas at two different pressures appeared to be to draw 
the complete curve connecting tlie current and the potential difference for each 
pressure, keeping the intensity of the illumination constant throughout the whole 
series of olrservatlons at both pressures. 
Source of Ultra-Violet Light. 
The greatest difficidty encountered during the course of tlie research was that of 
obtaining a source of ultra-violet light which could be relied upon to remain 
approximately constant, at least, in intensity throughout the vdiole course of long 
ranges of observations, often extending over many hours. An arc light or the spark 
between zinc terminals in the secondary of an induction coil worked from accumulators, 
with some form of interrupter in the primary circuit, which are the sources of ultra¬ 
violet light ordinarily used, proved utterly unsuitable, as did also arrangements 
suggested by Keeusler (■Verb, der Phys. Gesel. Berlin,’ Jahrg. 17, p. 898) and 
WuLF (‘ Ann. der Phys.,’ vol. 9, p. 948, 1902). 
The source ultimately adopted and used in the majority of the expeiiments 
descri])ed below was an arc l)etween iron wire terminals. The wires were tapped 
into the ends of long brass screws working in corresponding threads in brass pieces 
attached to the sides of a wooden frame, this arrangement rendering it easy to adjust 
the sparking distance without displacing its position. 
The terminals of the arc were connected to the secondarv terminals of an induction 
4 / 
coil, in parallel with which were also three large Leyden jars. An alternating current 
of aboiit 4 amperes passed through the primary of the coil when a key in the primary 
circuit was closed. 
In order to obtain an arc wliich will work constantlv for a sufficientlv lono’ time it 
«/ « O 
is necessarv to use iron wires of such a diameter, and reo-ulate the current throuo’h 
the primary, so that although the terminals become white hot, yet the heat developed 
is just not sufficient to cause particles of tlie iron to be melted oifi Under favourable 
conditions, when the ends of the terminals had once become properly rounded, the 
intensity of the arc would remain quite constant, without any need for adjusting the 
spark length, when run for as many as fifty or a hundred 10-second periods with 
intervals of a few minutes between each. 
An arc lietween aluminium terminals was also tried, but proved unsatisfactory 
