440 
MR. J. T. BOTTOMLEY ON THERMAL 
it which maintained it at a good red heat; and each morning and evening the traces of 
gas that had come off were removed by working the pump. In spite of this, however, 
the supply was by no means exhausted, although an improvement had certainly been 
made, and in the course of ten minutes’ heating of the wire the vacuum would become 
sensibly deteriorated. I was at first inclined to attribute these effects to some 
extremely minute leakage, such as one may get from electric perforation of a glass 
tube or from a nearly perfect sealing of a platinum wire or of a joint. The pump, 
however, and all the connections were absolute]y faultless so long as the wire was 
maintained at lower temperatures, say below dull-red heat; but, on being raised perhaps 
to 700° C. or so, I might make a determination of the radiation, and, on making a second 
determination one or two minutes later, might find that a distinct cooling influence 
had set in. The rapidity with which the change supervenes proves at any rate that 
the minute quantity of gas is generated in the immediate proximity of the wire, and it 
is not till a considerable time later that the McLeod gauge begins to be affected. 
On the other hand, with regard to pressures higher than about 50 mm., I was 
unable to obtain satisfactory series of results on account of the incessant disturbance 
of the temperature of the wire at high temperatures by the air-currents which its own 
temperature produces. Special experiments were, however, made at not very high 
temperatures (see p. 445) in order to bridge over to some extent the interval between 
normal air-pressures and the highest pressure to which these curves refer. 
As the results of the experiments just described are shown on the curves, the actual 
points experimentally found being marked, it would serve no useful purpose for me to 
quote the long Tables of numbers from which the curves are plotted. It may, however, 
be of interest that I should put down a single specimen of a series observed and 
calculated. I therefore insert here an extract from my laboratory note-book of 
June 22, 1886, and show side by side with it the results obtained by calculation 
from the observations. 
