of Demonstrating the Ionisation of Air. 687 
The above observations show also the very marked effect of 
an electric field. It will be observed that when v/v; only 
slightly exceeds the critical value, a difference of potential 
of 160 volts entirely prevents the formation of drops. When 
the expansion is made larger a few drops are seen ever in 
presence of the field ; the drops were found to increase in 
number as the expansion was increased. That some drops 
should be formed when the expansion considerably exceeds 
the critical value, even in presence of a strong field, is not 
surprising ; for the supersaturation when nuclei are few or 
absent will exceed the critical value for a finite time, the 
longer the greater the expansion. Any ions, set free while 
the supersaturation exceeds the critical value, will come into 
action as condensation nuclei and give rise to drops before 
the electric field has had time to remove them. 
A difference of potential: of 40 volts between the plates 
was found to reduce the number of drops formed on expansion 
to sensibly the same degree as 1000 volts. Even a potential- 
difference of two volts produced a noticeable diminution. 
In the older small-scale experiments, in which the distance 
between the plates was often considerably less than one cm., 
the strength of the field for a given ditference of potential 
would be about five times as great as in the present experi- 
ments, and the maximum distance the ions had to travel only 
one-fifth as great ; the effect of a given difference of potential 
in removing ions would then be much greater. Itis possible 
that accidental differences of potential may in some cases 
have already largely reduced the number of ions present, so 
that the additional reduction following the application of 
much stronger fields was not noticeable. In the small 
apparatus also the number of ions present per c.c. in the 
absence of an electric field would be less on account of the 
much greater rate of loss of ions by diffusion to the walls of 
the vessel. The number of drops being small a larger share 
of water would fall to each, and they would fall too rapidly 
for variations in their number to be readily detected. 
The total number of ions present when a steady state is 
reached, in the absence of an electric field, is such that the 
number of ions removed per second by recombination and by 
diffusions to the sides is equal to the number produced per 
second. On account of the small rate of production of ions, 
diffusion rather than recombination is in these experiments 
the more important factor in limiting the number of ions even 
with the large apparatus ; with the small apparatus the loss 
of ions by recombination is negligible in comparison with 
that due to diffusion to the walls. Let us assume, for example, 
