﻿Volume, and Temperature of Rarefied Gases. 313 



the pressure-tube (B). Then follow the volumes represented 

 by the reading-points v' } and lastly the product of pressure 

 and volume. In addition to this, if the original pressure in 

 the apparatus before reading the gauge is over *1 millim. or 



thereabouts, this must be taken into account as the pressures 

 found will be too small by this amount. The original pressure 

 must therefore be found from the expression 



Hi 



li 



as before shown and added to the values given in column 

 headed p above. 



When the calibration of the gauges was finished they were 

 ready for use. In fitting up the apparatus it was absolutely 

 necessary to seal all joints with a hand blowpipe so as to 

 prevent any possibility of leakage. The tap K for the 

 admission of the gas to be experimented on was an exception- 

 ally good one of oblique bore. It was lubricated with gummy 

 phosphoric acid, and though in constant use for over a year 

 never once showed the faintest trace of any leak whatsoever. 



During the experiments the mercury of the gas volume- 

 metre Q was kept at a known point #, so that the volume of 

 the apparatus to be exhausted was from the point x to the 

 T-piece c of the mercury-pump, the tubes ee, together w T ith 

 the two gauges and the vacuum-tube J. The gauges when 

 read had their volume-tubes jacketed with cold water ; and 

 fitted on gauge E on the diagram is shown the jacket-tube 

 used in the experiments at high temperatures. 



Whilst working with these gauges we noticed the following 

 points amongst others. First, when the gauge is not in 

 use, the mercury should be kept as close as practicable to the 

 trap, for if it be kept some way below, the gas between the 

 mercury and the trap will, on raising the mercury, inevitably 

 be driven into the gauge and considerably disturb the accu- 

 racy of reading, even as much as seven per cent, error having 

 occurred from this cause. 



Secondly, the ' sticktion ' of the mercury in the capillary 

 tubes of the gauges is very apt to introduce large errors into 

 the results, especially at high vacua, when the pressures and 

 volumes to be read are very small. To overcome this difficulty 

 the tubes were continually tapped before reading, and many 

 readings were taken both after ascent and descent of the 

 mercury. That this was necessary may be shown by the fact 

 that the before-mentioned capillary corrections when measured 



Phil. Mag. 8. 5. Vol. 38. No. 232. Sept. 1894. Y 



