SECOND SERIES ON LATENT HEAT. 83 



from the slowness with which he worked his pump — from 

 ten to five strokes a minute — it may be inferred that he was 

 alive to it, and conceived that he had eliminated it. 



This he seems very nearly, but not quite, to have done, 

 for the first series of experiments gave him a mechanical 

 force of 82 3 lb. falling one foot to raise lib. of water one degree, 

 and a repetition with slower pumping gave 795. Both these 

 quantities are above the true value 772, as he finally deter- 

 mined it, and the difference may very well have been due 

 to a small excess of temperature in the pump. 



The second series of these experiments were made to 

 determine the change in temperature due to change in the 

 latent heat of the air with change of volume. He does not 

 state his object, which, however, is clear from the experi- 

 ments. Two equal receivers are connected by a pipe with 

 a cock in it, initially closed. Into one vessel he compresses 

 air, and the other he exhausts, then immerses the whole in 

 water until they are at an equal temperature with the 

 water, which temperature he carefully measures ; then he 

 opens the intervening cock and allows the air to flow 

 from the full vessel into the empty, until they are at equal 

 pressure. Now, from his first series of experiments, it would 

 appear that the air, as it is compressed into the empty vessel, 

 must rise in temperature, which, if there is no compensating 

 action by the expansion of air in the full vessel, will raise 

 the temperature of the water, but since the work done by 

 the air in the full vessel is all expended on the air in the 

 empty vessel, it would follow that an amount of heat equi- 

 valent to this work must be abstracted from the full vessel, 

 and so that there should be an exact compensation, unless, 

 indeed, there were some change in the latent heat. By this 

 means he has eliminated every possible error in testing the 



