244 ©N THE TRANSMISSION OF HEAT. 



mometV?-Not heatecI P articIes mi g h t accumulate, and perhaps at length 

 unlcfs the veflel come in contact with the bulb of the thermometer, raife its 

 ^ n th"s a c S af c W it iCh tejT, P erature > and thus give the appearance of the fluid pof- 

 docsnot. fefling a conducting power, though it had none. I was foon 



fatisfied, however, that this could not be the cafe. Suppofing 

 the caloric to be ever fo flowly abftracled from the hot parti- 

 cles, ftill the itratum of fluid direclly heated, can never ex- 

 tend beneath the folid which is the fource of the caloric. It is 

 from it that the current of heated particles afcend, and they de- 

 scend by the fides of the veflel, where their temperature may 

 be reduced partially or intirely. But this defcending current 

 cannot proceed lower than the point where the afcending one 

 commences, and of courfe no part of the fluid which has thus 

 been direcUy heated can reach the bulb of the thermometer. 

 In glafs veffels this is not the cafe, becaufe in thefe caloric is 

 given out by the fides of the veflel, even beneath the heated 

 folid, which is fufpended within it. But in an ice veflel there 

 is no communication of this kind, as it cannot condud caloric 

 at any temperature above 32 9 . 

 But if tfce ther- The experiment then which has been fiated is fubjecl to 

 the fluid conl* ^ one of thefe fources of fallacy. If the thermometer rife, 

 du&s ; and if there can be no doubt that the fluid conduces caloric. If it do 

 probably anon- not ri ^ e > '* ma y ^e conc l u ded, if not with equal certainty, at 

 conduftor. leaft with the greater! probability, that it has no conducting 



power. 

 A veflel was A quantity of water was frozen in a tin mould, fo as to form, 



formed of ice. ? ho]{ow cylinder A> pj xm Fig> 2 the diamet er of the ca- 

 vity of which was three inches, and the depth 3£ inches, the 

 thicknefs of the ice forming its fides and bottom, being 1 -§ 

 inch. A thermometer B was introduced into it horizontally 

 at the depth of one inch, the bulb being in the axis of the cy- 

 linder, and was frozen in, fo that the veflel was capable of 

 containing any fluid.. It is evident that by pouring a fluid into 

 this ice cylinder, fo as to cover the bulb of the thermometer to 

 a greater or lefs height, and by fufpending in it a heated folid, 

 any propagation of caloric through that part of the fluid beneath 

 the folid, will be afcertained by the rife of the thermometer. 

 Water could not For making this experiment with a view to determine the 

 be ufed in thefe con d U cti n or power of fluids, water is altogether unfit, from the 



experiments, ," ' . ,' r i- • n " 1 r rv 



becaufe it con- Angular property it has ot expanding mitead ot contracting in 

 traces by heat every reduction of temperature from 40° of Fahrenheit to 32. 



Suppofe 



