S4 



DRS. J. A. MARKER AND P. CHAPPl'ls ON A 



a small water manometer placed in the cork of the apparatus, and must he trans- 

 formed to mercury pressure and added to the reduced barometric height. From this 

 .total pressure the temperature is deduced by means of the tables published by 

 M. BROCH for the temperatures of ebullition of pure water,* part of which is re-printed 

 in the Appendix to this paper, Table III. 



If the small variations of initial pressure during the course of the experiments be 

 taken into account, we obtain, on applying to the observations the formulae indicated 

 above, the following values for the coefficent of expansion of nitrogen under constant 

 volume : 



The general mean of these four groups of determinations has been employed for 

 the calculation of the temperatures in the series of comparisons made about this time, 

 excepting the series with the thermometer K.9, for which the mean of the last group 

 of observations 0'00367 227 was adopted. 



* ' Trav. et. Me'm. du Bureau International,' vol. 1 A, p. 46. 



