ON THE TRANSMISSION OP HEAT. 243 



perature of 32°. It appears doubtful, therefore, whether any 

 part of the fluid beneath the heated folid, fuppofing it to be 

 capable of conducting caloric, can have its temperature in 

 fuch a fituation raifed ; and hence it appears uncertain, whe* 

 ther from the circumftance of the thermometer not rifmg in 

 fuch an experiment, it could fairly be concluded, that the 

 fluid is a perfect non-conductor of caloric. 



There are alfo reafons, however, which render it probable, If the heat be 

 that any conduaing power in the fluid if it do exift, may be f^th^icfcL 

 difcovered by an experiment of this kind, by the fife of the abforb it, the 

 thermometer. Suppofing the fluid to pofTefs fuch a power, K^f^S^i 

 is fufficiently poffibie that the caloric may be conveyed from ji UJ d be a prop* 

 the heated folid through the jnterpofed fluid, (efpecially if conduCler. 

 the quantity of the fluid be fmall) to the thermometer, more 

 quickly than it can be cooled by the ice. Two facts render 

 probable this fuppofition. 1 ft. Ice is an imperfect conductor, 

 and therefore abforbs caloric flowly from another body. If a 

 piece of ice be thrown into water at 40 or 50°, a conflderable 

 time elapfes before the temperature of the fluid is reduced to 

 32° ; and 2dly. The celerity with which a hot body gives its ' 

 caloric tp a cold one, is, ceteris paribus, proportioned to the 

 difference between their temperatures. If the hot body, for 

 example, have a temperature 100 degrees higher than the 

 cold one, it will communicate caloric to it with much more 

 celerity, than if it were at a temperature only five degrees 

 higher. This caufe operates in the prefent experiment. The 

 difference between the temperature of the heated folid fuf- 

 pended over the thermometer, and that of the fluid interpofed 

 is very conflderable, and it will therefore give caloric to it 

 rapidly. On the other hand, the difference between the tem- 

 perature of the particles of the fluid that are heated, and that 

 of the ice with which they come in contact muft be much lefs, 

 and therefore they will part with their caloric more flowly. 

 From this circumftance, independent of the flownefs with 

 which ice abforbs caloric, the communication of caloric in this 

 experiment mult be more rapid than its abftraction ; and 

 therefore if the fluid do pofTefs a conducting power, the ther- 

 mometer muff, be heated. 



In reflecting on this fubject it feemed even poffibie, that the whether the 

 fluid might be cooled fo much more flowly by the fides of the Saturn of heated 

 vclTel, than it was heated by the folid fufpended in it, that the betowthVhoT 



R 2 heated °°°yj «id th°* 



