observed on the Glaciers of Chamouny. 255 



heat in consequence of a hot body being applied to the 

 surface of the liquid, that event is no decisive proof that 

 the heat acquired by the thermometer is communicated 

 by the fluid, from above, downwards, from molecule to 

 molecule, de proche en proche ; so far from being so, it is 

 not even a proof that it is from the fluid that the ther- 

 mometer receives the heat which it acquires ; for it is 

 possible, for aught we know to the contrary, that it may 

 be occasioned by the radiation of the hot body placed at 

 the surface of the fluid. 



In the experiments of which I have given an account 

 in my Essay on the Propagation of Heat in Fluids, 

 great masses, many pounds in weight, of boiling-hot 

 water, were made to repose for a long time (three hours) 

 on a cake of ice, without melting but a very small por- 

 tion of it ; and on repeating the experiment with an 

 equal quantity of very cold water (namely, at the tem- 

 perature of 41 Fahrenheit), nearly twice as much 

 ice was melted in the same time. In these experiments 

 the causes of uncertainty above mentioned did not exist, 

 and the results of them were certainly most striking. 



The conclusions which naturally flow from those 

 results have always appeared to me to be so perfectly 

 evident and indisputable as to stand in no need either 

 of elucidation or of further proof. 



If water be a conductor of heat, how did it happen 

 that the heat in the boiling water did not, in three hours, 

 find its way downwards to the cake of ice on which it 

 reposed, and from which it was separated only by a 

 stratum of cold water half an inch in thickness ? 



I wish that gentlemen who refuse their assent to the 

 opinions I have advanced respecting the causes of this 

 curious phenomenon would give a better explanation 



