80 . Mr Zeleny, On the Conditions | 
to 0:0543 cm., when the oscillations in question and the currents) 
from the points were observed to stop. Since the current from a 
point also starts at this stage it was then thought that the values 
-of f obtained were those requisite for starting a discharge, but 
it is known from the considerations given in this paper that 
the oscillations indicate rather the beginning of instability at the | 
surface, although with water the two actions start at potentials 
which are not greatly different. 
The results, in the table referred to, show that f Va=a constant, | 
and putting f 0 ue , we get V?=a x a constant, which is the relation | 
in expression (10). ) 
Taking the average value of the constant C from Table L i 
expression (10) becomes V?= 140a7' and for the sake of comparison ! 
with the expressions obtained in § 13 and 14, this may be put mi 
1l67aT 
pe 
the form V2 = 0:36 
It was found experimentally (loc. cit., p. 74) that the sunfale ! 
density on the end of a hemispherical drop attached to the end of | 
a tube, with a plate at a distance of 1:5 cms., is only about six-tenths 
of that on an isolated sphere of the same diameter and having the | 
same potential. The square of this number, or 0°36, must be intro- | 
duced into the denominators of the theoretical expressions in § 13! 
and 14 to permit of their comparison with experimental results on) 
drops at the ends of tubes. The experimental coefficient 16 is seen 
to be between the two coefficients 24 and 12 obtained in §§ 13 and 
14 respectively, and this may be taken to indicate that the experi- 
mental drop changes form in a way intermediate between that of | 
keeping a constant volume and that of keeping a constant cross- 
section. 
) 
Discussion. 
18. The relation of the potential at which the surface of a | 
charged drop becomes unstable, to that at which an electrical dis- 
charge starts, will now be considered. A study of a table giving 
surface tensions indicates that in air at atmospheric pressure the | 
limit for instability is reached on the surface of drops of all sub- 
stances which are liquid at room temperature, with the exception | 
of mercury, before a potential is reached at which an electrical” 
discharge through the air begins. But in hydrogen for example 
the potential for an electric discharge from a water surface is” 
smaller than the potential fer the beginning of instability. 
The size of the tube on which the drop is formed is of no great | 
consequence for the relationship in question, since it has been shown™ 
* J. Zeleny, Physical Review, Vol. 25, p. 317, 1907. 
