384 
DR. T. R. MERTON AND MR, S. BARRATT 
on the total amount of Hydrogen present and the pressure of Helium in the discharge 
tube. 
We have extended these results, and with a vacuum tube containing Helium at a 
pressure of 56 mm. and a little Hydrogen, in addition to the phenomena described 
above, it has been observed that on putting in the condenser there is a bright instan¬ 
taneous flash of the Hydrogen lines throughout the capillary before they appear at the 
ends of the capillary only.* In addition to the Hydrogen, the proportion of which 
could be controlled by means of a palladium regulator, the tube showed traces of Mercury, 
Sulphur, Oxygen, the Angstrom Carbon bands and a few other lines due to impurities 
which have not been completely identified. When the uncondensed discharge was first 
passed through the tube the Mercury lines were scarcely visible, but they gradually 
developed, though still very faint and somewhat stronger in the centre of the capillary 
than at the ends. On putting in the condenser the Mercury lines gradually became 
brighter, but appeared only in the centre of the capillary. On cutting out the condenser 
they appeared at once with great brilliance in the centre of the capillary, gradually 
spreading out towards the ends and at the same time becoming fainter. The Mercury 
lines behaved in exactly the opposite way to the Hydrogen lines, and it looked as if the 
effect of the condensed discharge was to collect all the Mercury in the tube to the centre 
of the capillary. The lines due to Sulphur, Oxygen, &c., behaved in the same manner 
as the Mercury lines. In Plate 3 ( a ) shows the appearance of the capillary, as 
photographed in the red and yellow regions of the spectrum while the tube was excited 
by the condensed discharge ; (b) shows the appearance immediately after the condenser 
was cut out, this photograph being obtained by repeatedly putting the condenser in 
and out, and only exposing the plate immediately after the condenser had been cut out, 
and in (c) the lines are seen uniformly distributed throughout the capillary when the 
tube was excited by the uncondensed discharge. In ( d ), ( e ) and (/) respectively the 
same phenomena are shown in a more refrangible region, in which the behaviour of the 
Hydrogen line H/3 and the green Mercury line can be seen. (The Mercury line was too 
weak for reproduction in (d) and (/).) 
The same phenomena can be observed at lower pressures of Helium in the discharge 
tube, but the condition of uniform intensity in the capillary after the condenser is cut 
out is very much less rapidly attained at high pressures. 
The possibility of the removal of Hydrogen by absorption by the glass walls of the 
capillary during the passage of the condensed discharge has been considered ; but it is 
believed that this explanation cannot be upheld, for in this case either it should be 
possible to reach a steady state in which the phenomena are no longer observed when 
the tube has been run for some time, or else the whole of the Hydrogen in the tube 
should rapidly disappear ; but there is no evidence of an approach to a steady state, or 
* It has also been observed that when the quantity of Hydrogen in the discharge tube is sufficiently 
great to show the secondary lines, the latter also appear only at the ends of the capillary when the condenser 
is cut out. 
