﻿310 Mr. Baly and Dr. Ramsay on the Relations of Pressure, 



values of pv read on the gauges were constant, at least within 

 the limits of experimental error. 



We give a type of the series of readings we obtained : — 



p- 





millim. 



pv. 



24-5 



1-00 



29-5 



1-00 



37-3 



1-01 



49-3 



1-01 



75-2 



1-02 



100-3 



101 



152-3 



1-00 



Many series of values of ]w were obtained, and they prove 

 that hydrogen expands quite normally between the limits 650 

 and 2*5 millim. of mercury pressure. We found it impossible 

 with any accuracy to make readings in the gauges with a 

 pressure smaller than 2*5 millim. These results also prove 

 that hydrogen does not condense on a glass surface between 

 these limits of pressure. By measuring the rate of exhaustion 

 of an apparatus of known volume by a reservoir pump whose 

 volume is also known, we found that hydrogen suffers no 

 condensation on glass down to as low a pressure as '000076 

 millim. 



Let the volume of the apparatus be called V and the volume 

 of the pump V, then the pressure after a stroke of the pump 



V 



will be v , v' ti mes the pressure which existed before the 



stroke. This value may also be found experimentally by 



dividing the pressure after a stroke of the pump by the pre- 

 viously existing pressure. We have called this ratio the 

 Factor of Exhaustion. It is evident that the constancy of 

 this factor of exhaustion dej)ends not on the expansion of the 

 gas but on the possibility of condensed gas coming off the 

 walls of the apparatus, in which case its value would increase. 

 We made many series of measurements to obtain the factors 

 of exhaustion, and found that they were quite constant and 

 equal to theory down to the extreme limit of exhaustion. 



We give here the results tabulated. 



In the column headed ratio of gauge readings are given 

 the ratios of the readings of the two gauges obtained at each 

 measurement. As these, of course, ought to be constant, 

 they may be looked upon as a measure of the experimental 

 inaccuracies of the work, and may be used to correct 

 the deviations observable in the values of the factors of ex- 



V 



haustion. Tn the last column are given the values of-^ — ^ 



