58 McCretianp —Lonization in Atmospheric Air. 
The experiments described in this paper have to do with this slight ionization 
existing in atmospheric air in its normal condition. In addition to papers by 
Wilson, and by Elster and Geitel, the subject has been dealt with by Prof. 
Rutherford* of Montreal, Prof. McLennan and Mr. Burtont of Toronto, and 
by the Hon. R. J. Strutt.¢ 
2.—Deseription of Apparatus. 
A square piece of thick wood ABC (fig. 1) was taken and covered over both 
faces with tinfoil connected to earth. On this tinfoil coating rest blocks of 
parafin £G, supporting a plate LZ, 25°5 cms. by 25°5 ems., made of perforated 
zine. A second similar plate J/is supported by the metal stem H from a clamp 
(not shown in figure), resting on paraffin blocks on the tinfoil coating of ABC. 
Pieces of paraffin Dare inserted in the wooden hase, care being taken that they 
are surrounded by tinfoil in contact with the earthed coating of the base; through 
these blocks wires pass, making connexion with the plates Z and M. 
One plate J is charged, say positively, Z being to earth, and then Z is discon- 
nected from earth, and slowly rises in potential, due to the passage of a current 
through the gas from MW to Z. It is obvious from the arrangement that this rise 
in potential of Z must be due to a current through the gas; a leakage by the 
supports /’@ could not produce the effect, as these supports rest on a conductor 
kept at zero potential. 
A leakage over the supports #G would diminish the rate of increase of 
potential of Z, but could not possibly produce the rise. We are safe, therefore, 
from the error of ascribing to ionization of the air effects due to faulty insulation ; 
the only error that can occur is that of getting too small a value of the ionization. 
This is guarded against by testing the insulation, and in fact it is quite easy 
to get the insulation perfect, as it is only subjected to quite a small electric 
field. 
The rate at which the plate Z gets charged is measured by a quadrant 
electrometer #. One pair of quadrants is permanently connected to earth through 
the mercury cup ¢ ina paraffin block; the other pair of quadrants is permanently 
connected to the plate Z, and can be put to earth, or insulated, by putting in or 
taking out the metal piece d connecting the mercury cups ) and ¢. This connecting 
piece d can be operated by a string from a distance from which the scale is observed 
by a telescope to avoid induction effects produced by the motion of the observer. 
The paraffin block a, and all insulating supports used to carry the wire joining 7 
to LZ, are set on sheets of tinfoil connected to earth, to avoid all possibility of the 
* Phil. Mag., Dec., 1902. + Jbid., June, 1908. t Lbid., June, 1903. 
