﻿Volume, and Temperature of Rarefied Gases. 325 



0*75 millim. The only possible explanation appears to be that 

 the temperatures of the two gauges must have differed to a 

 small extent, and that the smallest difference in temperature 

 produces quite an unusual disturbance in equilibrium. 



Nitrogen was next experimented upon. The gas was pre- 

 pared by absorbing the oxygen from air, first as far as possible 

 by means of pyrogallate of potassium, and then by passing it 

 over red-hot copper-gauze. 



The experiments were carried out in the usual way, and 

 there was no special difficulty connected with the filling of 

 the apparatus with the gas. 



The great difficulty in working with nitrogen was that 

 experienced in obtaining satisfactory and concordant results 

 in measurement of the ratio between the gauge-readings 

 preliminary to the jacketing. This difficulty increased as the 

 vacuum improved, and it was found impossible to obtain 

 results of any value whatever at a less pressure than *6 millim. 

 The readings for the ratio were always taken with the utmost 

 care, but in spite of this ratios sometimes differed amongst 

 themselves as much as 2 per cent., which means a difference 

 of 9 or more per cent, in the values of the coefficient of 

 expansion. 



The numbers obtained were as follows : — 



Pressure. 



T. 



t. 



5*3 g^. 



132-55 



8-8 



* J ' 303-8" 



132-16 



9 



° 801-6* 



132-42 



8-5 



1*1 - 



x x 304* 



131-7 



14 



•« 1 

 ° 331* 



131-6 



12-5 



f355* ) 



b 1 — . i 



<- 343-5 J 



132-1 



11-5 



•6 {t} 



V 323* J 

 \ 



132-1 



13 



•6 {f } 



<- 327* J 



1 



131-6 



13 



(T ] 132-2 13 



Mean of eight .,-r^ at *6 millim. 



842-56 



These last values of the coefficient of expansion are in four 

 sets, each set being the result of an entire fresh filling of the 



