UNDER THE INFLUENCE OF CHEMICAL ACTION. 
17 
These data show that the experimental conditions under which chlorine has been 
tested cover a wide range as in the case of COCl 2 . Thus the chemical saturation current 
is varied between the limits 500/min. with 0*04mfd., equal to 5-9 X 10~ 10 amperes, 
and about 3 -5 X 10 -13 amperes. The photo-electric saturation current varies between 
about 3 X 10~ 11 amperes and 10~ 12 amperes. The chemical emission increases and 
decreases with time in the different series ; sometimes it is almost constant, at others 
it is changing rapidly. The pressure of gas in the apparatus ranges from 0 -002 mm. 
to 0 TOO mm. The chemical effect is nearly 60 times the photo-electric in one series, 
whereas it is only about one-sixth of the photo-electric in another, and this ratio has 
intermediate values in other cases. In five of the series the gaseous products of the 
reactions occurring were allowed to accumulate, whereas in one of them the pumps 
were run continuously. The rate at which the drops fell was varied between 1 in 10 
seconds and 1 in 110 seconds. 
In spite of this wide variation in the conditions, all the points fall on the smooth 
curve m fig. 7 to the degree of accuracy which it seems reasonable to expect. An excep¬ 
tion is perhaps furnished by the 3 points x of series I. at — 2 -5, — 3 and — 4 volts. 
It is unlikely that chance errors, which would have to be rather large in any event, 
would make these three points, which were determined in succession at the beginning 
of the series, lie so persistently to the right of the graph required by the rest. It seems 
likely that we have here another example of a change in the contact potential taking 
place during the actual course of the experiments, as in the case of series VI. with COCl 2 . 
Some of the points in I. x , II. ©, and VI. ® might seem to suggest a slower voltage 
approach to saturation than the normal, and it was thought that this effect might be 
VOL. CCXXII.-A, 
D 
