190 
PHYSICS: TOLMAN AND STEWART 
and should expect the periphery of a rotating disk to become negatively 
charged owing to the action of centrifugal force on the electrons in the 
disk. Such effects would presumably be very small, however, owing 
to the exceedingly small mass probably associated with the electron. 
In the case of electrolytes, where the mass associated with the carriers 
of electricity is much larger than in metals, such effects have already 
been detected by Colley^ and Des Coudres^ using respectively the 
accelerational and the centrifugal method, and have been subjected to a 
more elaborate quantitative investigation by Tolman^ and by Tolman 
and Osgerby.^ 
In the case of metallic conductors. Maxwell,^ himself, was the first, 
not only to discuss the nature of the phenomena which would arise if 
mass of the ordinary kind should be associated with electricity in metals, 
but also to try experiments to detect the possible effects. He had, 
however, no means of predicting the presumable size of such effects if 
they did exist, since this was before the development of the electron 
theory, and he merely states the negative results of his experiments 
without any information as to the dimensions or efficiency of his appa- 
ratus. Lodge^ also reports a negative result for such experiments. 
The first attempts to detect such effects in metals of which we have any 
quantitative information were made by Nichols^ in 1906. He employed 
the centrifugal method, using a rotating aluminum disk and making a 
rubbing contact at the periphery and center with wires, which led to the 
electrical measuring apparatus. Such rubbing contacts, in particular 
the one at the rapidly moving periphery, necessarily introduce larger and 
variable electromotive forces; nevertheless from a series of experiments 
Nichols was able to conclude that the mass of the carrier in metals is 
less than that of the hydrogen atom. 
Since in the case of metals the centrifugal method of attack almost neces- 
sarily involves some disastrous form of rubbing contact, we have been 
at work during the last four years developing the accelerational method 
of attack. In 1913^ we were able to report that the effect in metals, 
if any, was so small that the mass of the carrier in metals was less than 
one two-hundredth part of that of the hydrogen atom. With the help 
of a much more sensitive galvanometer and eliminating one by one a 
number of accidental effects which appear when greater sensitiveness is 
reached, we have now apparently obtained a real effect due to the mass 
of the carrier in metals. 
The apparatus consisted essentially of a coil of insulated copper wire, 
wound on the periphery of a wheel which could be rotated at a speed of 
about 5000 r.p.m. and brought suddenly to rest. The two ends of the 
