in Fluids. 305 



melted which resulted from the cooling of the sides of 

 the jar, but also, and more especially, as that difference 

 was directly contrary to the effects produced by the same 

 means in the experiments with hot water. More ice was 

 melted when the outside of the jar was kept ice-cold 

 than when it was surrounded by air at the temperature 

 of6i. 



All these appearances might, I think, be accounted for 

 in a satisfactory manner on the principles we have as- 

 sumed respecting the manner in which Heat is propa- 

 gated in liquids ; but without engaging ourselves at 

 present too far in these abstruse speculations, let us take 

 a retrospective view of all our experiments, and see what 

 general results may with certainty be drawn from them. 



One of the experiments in which the greatest quantity 

 of ice was melted by hot: water is that No. 36, in which 

 3963 grains were melted in three hours, or 180 minutes. 

 If now from } this quantity we deduct that which, according 

 to the results of the two preceding experiments, must 

 have been melted in the first 30 minutes, namely, 690-3- 

 grains, there will remain 3272^ grains for the quantity 

 melted in the last 150 minutes, which gives 654^- grains 

 for the quantity melted in 30 minutes in the ordinary 

 course of the experiment. 



This quantity, 654^- grains, deducted from that which 

 at a mean of two experiments (those No. 34 and No. 

 35) was found to be actually melted in 30 minutes, 

 namely, 690^- grains, leaves 36 grains for the quantity 

 which in these two experiments was melted in conse- 

 quence of the temporary motions into which the hot 

 water was thrown in the operation of pouring it into the 

 jar. The difference between these two quantities ( 36 

 grains) is very inconsiderable, and shews that the means 



VOL. I. 20 



