McCieLtanp—Jonization in Atmospheric Air. 59 
potential of Z changing from zero, except by a current through the surrounding 
gas. For measuring the leakage through air in this way a very sensitive 
electrometer is required; but I have had to be content so far with an instru- 
ment giving a deflection of 60 mm. scale divisions for 1 volt difference of potential 
between the quadrants. To get deflections of sufficient magnitude it has been 
necessary to lengthen the time of each observation—in general the deflection 
produced in four minutes has been observed. When care is taken to have the 
& 
Storage Cells Larth 
Ne, Il, 
electrometer free from vibration, and properly shielded from electric disturbances, 
it is quite easy to get the zero sufficiently steady to admit of taking readings 
extending over four minutes or more. Still it involves more time, and, in addition, 
there are many obvious experiments I have been unable to make. 
A glass case PQ, lined inside with tinfoil, can be placed over the plates Z and 
M, resting in a groove in the base ABC, filled with mercury or glycerine making 
the whole practically air-tight. 
M2 
