148 Messrs. Holborn and Austin on Cathode Disintegration 
more than 1 per cent. for hours at a time by regulating the 
electric motor, which was driven by the current from a 
storage-battery. The cathodes were connected to earth 
through a very sensitive milliamperemeter. 
In the course of the investigation, a second spherical tube B, 
12 cm. in diameter, was connected to the first. A cathode 
and its attendant exploring wire were placed approximately 
in the centre of the tube. As this tube was also frequently 
used for regulating the gas-pressure, it was separated from 
the rest of the apparatus by a cock H,. 
2. Observations in Tube A. 
Four cathodes were situated in tube A, and the discharge 
was sent through one cf them at a time for from 15 to 
30 minutes. Before using, the cathodes were left over-night 
in vacuo. The weighings before and after the passage of the 
current gave the loss of weight with a certainty of 0-01 mg. 
In the tables the losses in weight are calculated for a uniform 
time of 30 minutes on the assumption that the loss is pro- 
portional to the time of passage of the current. 
The cathode-fall was varied from its normal value (about 
350 V.) where the negative discharge begins to cover the 
whole cathode, to about 2500 V. Experiments in which 
the potential was higher were inclined to give contradictory 
results, probably on account of irregularities in the discharge. 
These irregularities appeared to begin as soon as the negative 
column began to contract so that it ceased to cover the whole 
of the surface of the cathode. If the potential was forced 
still higher, so that green phosphorescence appeared on the 
glass, the loss of weight of the cathode decreased in a marked 
degree, whereas before it had steadily increased with in- 
creasing potential-difference. The metal deposits on the walls 
of the tubes were not removed between the series of obser- 
vations. The mica disks on the backs of the cathodes, on which 
a deposit was also formed, were removed after each experiment. 
Table I. contains the observations on different metals. 
Under the cathode-fall V are given the means of the readings 
taken every five minutes during the course of the discharge. 
The loss of weight y (for a time of discharge of 30 minutes 
and a current-strength of 0°6x10-? amp.) are given in 
mg. x 10-*, the gas-pressure p in mm. Hg. 
For all the metals, with the exception of gold, the loss of 
weight y from a certain point on increases directly as the 
cathode-fall V (fig. 2, p. 150). All these curves prolonged 
backwards cut the axis of abscissas at the point V=495. 
It is to be observed that the disintegration does not disappear 
at this point, but appears to decrease asymptotically as V 
approaches the normal value. 
