220 
MESSRS. W. DE LA RUE AND H. W. MULLER ON THE 
at each end; near one of the ends is a small tube, 0'75 inch in diameter, 
fastened to the main tube at right angles, and fitted with a glass stopper, 
in which two stout platinum wires, 0'043 inch diameter, are melted ; there is 
soldered, with gold, to the two platinum wires a spiral of palladium made of 
wire 12 inches long and 0'0125 inch diameter, P d, in the diagram. The 
palladium coil was charged to saturation with hydrogen, by immersing it in 
dilute sulphuric acid and making it the negative pole of a bichromate 
battery of six elements ; after it had been washed in distilled water it was 
dried and inserted in the tube. The two stout wires of platinum, to which 
the palladium coil is attached, are connected to a condenser of 10 '9 m.f., 
charged with 3240 cells. One of the wires leads to K' so that no current 
can pass from the condenser until this key is pressed down, when this is 
done the charge passes and by suddenly igniting the wire drives off the 
hydrogen. 
On first making the experiment, through inadvertence, the condenser was 
not connected with the palladium coil, but the current in the tube heated it 
and drove off the hydrogen, producing the effect to be described. Pressure 
l - 003 m.m., 1320 M, 5120 cells, 300,000 ohms resistance, saucer-shaped strata 
as shown in the diagram. As soon as the palladium coil became heated 
it suddenly drove back the strata about one-third the length of the tube 
from the negative, and the current subsequently became agitated. After 
the liberation of the gas the pressure was increased to l - 088 m.m., 1432 M or 
by 112 M, the pressure before and after discharge being as 1 to 1‘08. 
This experiment was repeated with 4800 cells without external resistance, 
Pressure (U9965 m.m., before the discharge of the condenser, 
,, 1 - 0381 ,, after ,, „ „ 
Difference 0‘0416 „ 55 M, 
double strata were produced from the positive to the palladium coil which 
was on the negative side. On liberating hydrogen by the discharge of the 
condenser these were driven back 14 inches towards the positive, and sub- 
sequently only a confused discharge was produced. 
When the terminal near the coil was positive the same phenomena were 
not produced on the discharge of the condenser. 
When, instead of liberating gas from the palladium coil, calibarted charges 
of hydrogen were let in during the passage of the current, the strata were 
narrowed radially for a few seconds, then again occupied the whole diameter 
of the tube, but were confused and agitated. Each fresh puff of gas was 
accompanied by a fresh radial collapse, and, at the same time, by a repulsion 
