tn the Discharge through Gases at Low Pressures. 151 
were taken is indicated by the numbers in the first column in 
the two tables. A difference in gas-pressure for the same 
-cathode-fall could not be discovered in the two cases. 
In the two groups of metals, the loss of weight y of the 
cathode, due to the passage of a current of 0°6x 10-° amp. 
for half an hour is represented very closely by the equations 
y=0-00163 A/n(V—495), or 
y=0°00186 A/n(V—495) me. x 10-2. 
In a voltameter during the same time, the same current 
would deposit 0°0112 A/n mg. according to the Faraday law, 
where A/n denotes the chemical equivalent. This quantity 
is disintegrated in the discharge at low pressures (tube A) at 
a cathode-fall of 1180 V. or 1090 V. respectively. A second 
effect must be assumed in addition to the Faraday law to 
explain the change of disintegration with the change of 
potential. 
A similar curve has been calculated for gold as a univalent 
metal. The calculated values are all larger than the observed, 
and the difference increases with increasing cathode-fall. 
Several easily oxidized metals of low melting-point were 
also observed. At low potentials, zinc yvave fairly good 
results, on the assumption that it is a di-valent metal. We 
do not lay great stress on this, however, as the cathode was 
strongly oxidized as soon as the cathode-fall was increased 
above 1600 volts. After this experiment, the loss of weight 
was found to be abnormally large, and the cathode and walls 
of the tube were entirely covered with white oxide. The 
surface also showed a crystalline structure which denoted a 
high temperature. Cadmium, lead, and tin were also tried, 
but as they are still more easily oxidized the results are not 
taken into consideration, as we desired to investigate the 
disintegration under conditions as far as possible free from 
sublimation and ordinary oxidation. Many of the metals given 
in the tables were coloured on account of slight oxidation, 
especially after the discharge at low potentials. Even 
platinum showed the Priestley rings at the point opposite the 
exploring wire. Sometimes however, this point remained 
unchanged, and the rim of the cathode showed the coloured 
rings. 
We must also mention the experiments with aluminium, 
iron, and steel, all of which show very little disintegration. 
In the case of steel we examined both a hard and a soft 
specimen, to see whether the hardness of the material was a 
cause of difference. The loss of weight in both cases was so 
small, however, that the difference fell within the limits of 
