390 = Prof. Skinner on the Relation of Electrode Fall 
circuit. To confine the discharge solely to the face of the 
electrode, a second circular glass plate properly cut was pressed 
closely on top of the first, covering the tinfoil strips and 
fitting the rim of the electrodes. Leakage from the brush B 
to the gas was avoided by reducing the space between the 
upper face of the electrode-plate and the roof of the chamber 
to about 4 millim. thickness. The discharge-tube was 3 ems. 
in diameter and about 20 cms. long. A fine aluminium 
wire W, sheathed up to about 6 millims. of the end with fine 
- glass tubing (outer diameter 4} millim.), served at position 1 
for obtaining the fall of potential at the anode. If turned 
into position 2, 24 centims. from the electrode, the same wire 
served for obtaining the cathode-fall, and, if desirable, the 
fall in the gas over the intervening space. All permanent 
joints of this apparatus which could not be fused were 
cemented first with water-glass (sodium silicate), then, in 
addition, they were covered on the outside with de Khotinsky’s 
laboratory-cement. The base-plate A, which had to be 
removed to allow for repolishing the electrodes, was sealed 
by simply applying the latter cement on the outside of the 
juncture, care being taken that it did not extend to the 
inner part. This tube proved to be in all cases absolutely 
air-tight. 
The platinum, gold, copper, lead, and aluminium electrodes 
were made from sheets as obtained commercially; silver and 
nickel from plates used in electroplating ; bismuth, antimony, 
tin, cadmium, zinc, and magnesium were cast. The steel 
electrodes were made of good tool-steel not hardened. All 
electrodes were given for each observation, when not other- 
wise designated, a fresh mirror-polish, using new felt as 
a polishing-cloth and mew cotton-cloth for cleaning. As 
polishing material rouge was used for platinum, silver, and 
gold ; but for the rest of the metals infusorial earth mixed 
with tallow proved more satisfactory. After polishing, 
the metals were very carefully cleaned by rubbing lightly 
with a dry cloth. They were then mounted quickly in the 
tube, which was immediately evacuated and flushed several 
times, and finally left standing in open connexion with a 
drying-chamber ina few millimetres of the gas. When the gas 
was dry the metals usually preserved their surface untarnished 
for several days—even in oxygen, lead excepted. At least 
two specimens of each metal were tested, but in no case was a 
notable difference in the different samples observed. 
The gases were carefully prepared and stored in large flasks. 
which had been previously cleaned, then dried and freed from 
occluded gas and moisture by standing highly exhausted 
