PRECONTACT CONDUCTION CURRENTS 
233 
There are various interpretations of the results. The most 
obvious is that of metallic conduction, based on the presence of 
Ohm ’s Law. The changing resistance could hardly be attributed 
to the variable temperature, as this would call for a negative 
temperature coefficient of large value. Or the resistance varia- 
tion might be explained on the ground of variable contact area. 
But computation of the small areas needed to account for the 
observed currents shows that they are exceedingly minute. For 
a single approach of the plates an extremely small contact area 
might be obtained, but to be able to get so nearly the same re- 
sults three times in succession, with the relatively crude means 
of adjustment, so as to give contact areas varying from 6.02x 
10" 12 cms. 2 to 3.62x10 * 12 cms. 2 , is difficult of belief. Expressed 
in terms of the diameter of a circular contact area the range 
would be from 1/20 to 1/25 wave length, sodium light. The dis- 
tance was constant to within 0.25 wave length, while the diam- 
eter of the contact area, supposed circular, varies within 1/100 
wave length. 
The presence of Ohm’s Law, while being a necessary condi- 
tion of metallic conduction, is not a sufficient condition. It 
might be expected to hold if the conductors were bridges of for- 
eign matter of high resistance, such as dust particles between 
the plates, or if the conduction was due to ionisation currents 
of values considerably below saturation, or again if the conduc- 
tion was electrolytic in nature, as suggested by Prof. G. W. Stew- 
art, without polarization. It did not appear difficult to clear 
the film surfaces of dust particles of any appreciable size. Also 
there was no known constant source of ionisation as is the case 
with the ordinary ionisation current curve. The question arises 
whether there could be ionisation by collision at atmospheric 
pressure and ordinary room temperature under relatively low 
potential gradients when the thickness of a given volume of 
gas is very small compared with its other dimensions. In 
the present case, if the depth of the volume between the films is 
represented by 1 mm., the length and breadth of the same volume 
would be represented by about 1 % m. The currents may conceiv- 
ably be due to a coalition of surface films of moisture, gases, etc. 
There is also a possible application of the theory of electron at- 
mospheres as advanced by Prof. R. W. Wood 2 , but according to 
Philosophical Magazine, p. 316, Aug., 1912. 
