278 
IOWA ACADEMY OF SCIENCE 
applied in order to bring the plates into contact at one edge, much 
greater than was applied to the speculum metal plates. However, there 
was considerable irregularity in the pressure required to cause contact, 
and it may have been that after heating the speculum metal plates were 
separated only by the thickness of the mica. At any rate one of the 
mica flakes measured two and a half wave lengths, another one two, 
and a third one only one wave length. 
I observed after the second heating that the mica flakes on speculum 
surfaces showed the same deep green color that they exhibited when 
free. This rather indicates close contact between the mica and the spec- 
ulum. Also the fact that long continued heating permitted the electrical 
charges to become closer together, when separated by mica, indicates 
that heating either drives off something that conducts and separates, or 
alters the structure of the surfaces. If the mica flakes were in contact 
with the speculum, then there can be no conducting atmosphere of more 
than two wave lengths of light, for this I regard as about the safe accur- 
acy of my experiments, but if on the other hand there exists a conducting 
atmosphere which is at the same time a mechanically separating medium, 
then it would not be safe to set such a low limit on the thickness of a 
possible electron atmosphere. The question will not, be definitely settled 
until the distances are measured simultaneously by the two methods. 
We propose to do this by using very .thin films of platinum deposited 
on plane glass surfaces. It is hoped thereby to finally decide to a fraction 
of a wave length whether or not an electron atmosphere exists,, and also 
to obtain information on the role played by insulating materials of 
small dimensions. 
Inasmuch, as heating certain insulators such as mica" and quarts re- 
stored their insulating properties, it is certain that the conduction ob- 
tained by "Wood at such long distances was due t,o the presence of the- 
insulating material, and it is presumed that moisture films and con- 
ducting particles were responsible for it. 
In conclusion it may be said that while the electrical method was 
satisfactory as here used, it can be more accurately adapted to the prob- 
lem in hand by more suitably choosing the suspension for the elec- 
trometer needle and by using a larger capacity in parallel with the 
electrometer. 
The State University of Iowa. 
