118 Proceedings of the Royal Society of Edinburgh. [Sess. 
giving saturation currents, or currents which are practically independent 
of the potential difference betweeen the electrodes. (A series of experi- 
ments, hitherto unpublished, made by one of us, has shown that even in 
the highest attainable vacua an increase in the potential difference between 
parallel plane electrodes will produce a slight increase in current under all 
conditions, although this increase is very small in high vacua, and could not 
possibly affect the results of any experiments described below.) 
Series of observations were also taken in gases at atmospheric pressure, 
and at a pressure of about 50 millimetres of mercury, the latter pressure 
being chosen for the reason that saturation (or the nearest approach to it 
possible using the photo-electric effect as the ionising agent) could be easily 
obtained with the voltage at our disposal. These series, although, as ex- 
plained above, scarcely showing the effect of temperature on the actual 
photo-electric discharge, gave some interesting results, and incidentally led 
us to discover the important effect of absorbed gases on the discharge. 
The experiments were made with temperatures ranging from 5° C. to 
500° C. It was found impracticable, with the form of apparatus used, to 
employ higher temperatures than 500°. When the platinum was made 
hotter than this, the sealing-wax joints closing the experimental vessel 
soon became overheated and gave way. The pressure, too, could not be 
maintained as low as 3 Jq millimetre (at which pressure the experiments in 
vacuo were carried out) if higher temperatures were employed, whilst at 
about 500° C. the platinum commenced to give off negative corpuscles due 
to heat alone. This effect increases rapidly with further increase in 
temperature, and soon completely masks the photo-electric discharge. 
Source of Ultra-violet Light. 
The source of ultra-violet light used throughout the series of experiments 
was a discharge between iron terminals in an atmosphere of pure hydrogen, 
as described by one of us in a previous article (Yarley, Phil. Trans, of the 
R.S. London , p. 439, 1903). The spark apparatus is shown diagrammatically 
in fig. 1, in which T T are the iron wire terminals between which the 
sparks pass, the length of the gap being about 6 millimetres. The wires 
T T were enclosed in glass tubes G G, enclosed in turn in wider glass tubes, 
and finally passing through brass side tubes S S, attached to the main 
brass vessel B. At the front of this vessel a quartz window, Q, was 
sealing- waxed on (all joints are made with sealing wax), while two narrow 
tubes, A and E, served to pass a continuous stream of pure hydrogen, 
prepared electrolytically and dried by bubbling very slowly through con- 
