Manchester Memoirs, Vol. xlviii. {igo^,) No. ^. 3 



In my own experiments I used the method of 

 weighing in air. It was in fact useless to hope for any 

 rehab le result by weighing in water, for the object was 

 cooled by direct immersion in liquid air and necessarily 

 carried a certain quantity of liquid air into the calorimeter, 

 where a larger or smaller quantity was vaporised within 

 the ice-coating. 



My method of procedure was briefly as follows. 



A piece of metal, generally a cylinder with rounded 

 ends, was immersed in liquid air for a time sufficiently 

 long for the establishment of equilibrium of temperature. 

 It was withdrawn by means of a silk thread passing 

 through a hole in a small projection at one end of the 

 axis of the cylinder, and was carried to a calorimeter 

 containing ice-cold water, the transference lasting five 

 seconds or less, and the greater part of the adhering 

 liquid air being shaken off during the transference. It was 

 allowed to remain suspended in the calorimeter for a 

 sufficient time, then, with its ice-coating, was removed to a 

 vessel in which the surface of the ice was dried by contact 

 with ice-cold filter paper, and was finally placed in another 

 vessel and weighed. 



The Method of Cooling the Object. 



The piece of metal was cooled by direct immersion in 

 the liquid air. This procedure has the great advantage 

 that we can be practically certain that the metal, if it has 

 remained in the bath of liquid air for a sufficient time, is, 

 at the moment of extraction, at the same temperature as 

 the liquid air. It has the disadvantage that liquid air 

 adheres to the surface of the metal and is carried over 

 into the calorimeter. As will appear more fully in the 

 sequel, the liquid air carried over, however small in 



