506 Mr Horton, The spectrum of the discharge 
metres I repeated both the spiral tube and the vacuum tube 
experiments and again found that the new orange and red lines 
disappeared when the mercury vapour was withdrawn, and that 
they were present in the spectrum of the stuff condensed while the 
discharge was passing in the lime-cathode tube. With helium as 
the residual gas in the apparatus all five lines were seen in the 
spectrum of the condensed substances, whereas with the other 
gases only the orange lines were obtained. 
In some experiments made in helium gas at a low pressure, 
using a platinum anode in the discharge bulb, it was noticeable 
that the mercury spectrum, although bright to begin with, gradually 
faded away as the tube was worked, the orange and red lines 
disappearing at the same time. Since mercury would not be 
absorbed by the platinum anode, it seems possible that it was 
being used up during the passage of the discharge more quickly 
than the vapour could diffuse over from the mercury pump. 
Professor Thomson in a lecture at the Royal Institution on 
January 19, 1906, put forward the view that the luminosity in 
a lime-cathode tube is due to an increase of the internal energy of 
the atoms of the gas in consequence of their bombardment by the 
corpuscles shot out by the glowing lime. When this internal 
energy gets up to a certain critical value the equilibrium of the 
atom becomes unstable and an explosion occurs, resulting in an 
expulsion of corpuscles and such a shaking up of those remaining 
in the atom that these vibrate so vigorously that the energy 
radiated is sufficient to produce luminosity. If this view is correct 
it might mean that some of the mercury atoms would be entirely 
destroyed or converted into new atoms by the removal of a large 
number of corpuscles. This would seem to account for the dis- 
appearance of the mercury from the tube when the discharge is 
passed. On the other hand such an explanation of the luminosity 
might be taken to mean that in addition to the usual mercury 
lines new ones might be seen in the spectrum caused by special 
forms of vibration into which the corpuscles remaining in the atom 
are thrown. It seems probable that this is what occurs, for the 
new lines were never seen apart from the mercury spectrum. 
I attempted to get them by condensing in a vacuum tube with 
platinum electrodes. With such a tube the mercury spectrum was 
always very faint, when visible at all, and no sign of the red or 
orange lines could be seen. 
It seems therefore that these lines are probably due to mercury 
and the fact that they are not given by such observers as Kayser 
and Runge or Eder and Valenta must be due to their absence in 
the ordinary arc or spark spectrum or to their being extremely 
faint. In a recent paper Stark (Ann. der Physik, Vol. XXI. 1905, 
p. 490) gives two lines of small intensity in the vacuum tube 
