OF GASES AT HIGH EXHAUSTION'S. 
393 
the pump. A tube packed with gold-leaf was used in the early experiments, but as it 
did not entirely stop the passage of mercury vapour, it was discarded in favour of 
sulphur, which, whilst not quite perfect, answers fairly well. As it was thought possible 
that sulphur vapour might be given off, the copper tube was interposed : at first a 
tube of oxide of lead was used for this purpose, but copper seems to answer better. At 
to a similar system of six tubes is interposed in the path of the entering gas. These 
will be further alluded to when the experiments with gases other than air are 
described. The tube o is the phosphoric anhydride reservoir belonging to the pump ; 
x is the McLeod measuring apparatus; y is a small radiometer, having a standard 
candle at a fixed distance from it; h is a spectrum tube with capillary bore furnished 
with aluminium poles ; l is a resistance tube having aluminium terminals with ends 
4 millims. apart (663, 664, 665). At z a series of spectrum tubes is attached, so that 
in course of experiment one tube can be sealed off at any particular pressure for 
further reference. 
642. The pump employed has already been described (355). The measuring appa¬ 
ratus is similar to that described by Professor McLeod'" before the Physical Society, 
June 13th, 1874. As it contains several improvements shown by experience to be 
necessary when working at very high vacua, a detailed description may be useful. 
The instrument shown in fig. 4 consists essentially of a globe, a, a volume tube, b, 
* Philosophical Magazine, vol. xlviii., p. 110, August, 1874. 
