t^i iLxamlnation of the recai'Vi'd b'Mr'wcs 



\vheii in equilibrium it rufTers no reftraint whatever, for theft 

 every individual fubftance has its ovn proper quantity of th6 

 common (lock. 



I ftiall now examine for a little feveriil fa6i:s whiL'h appeat 

 to pro've, if duly weighed and applied, that heat poirefTes an- 

 other chara6lei'iftic of inatter, I mean gravity. If thi;' 

 fa6l can be eflabliflied, we ihall have another llrong proof of 

 the fuhilailtiality of hea,t; arid tlie fafts I mean to bring to 

 your recollection are, I think, fufficient to eftablifli it as 1 

 truth — a truth which has been forcing itfelf on the riotice of 

 men, fof ilianv centuries, with much Orongef evidence than 

 manv others vvhic'h have received generaladmiffion, though 

 it has not only befen overlooked, but many experiments havd 

 been brought forward to eftablifli the oppofite. 



Heat poured into bodies lefTens their fpfccific gravity ; and 

 yet heat may riot be weiglied, even comparatively ! ! There 

 appears to me to be foniethinsJ; fo extrerfielv repugnant to 

 reafon in this aflertion, that ever fince T have dated to reafon 

 for myfelf I have found myfelf forced to refufe my aflcnt 

 to it. If heat leflens the fpecific gravity of bodies, (and we 

 can evefl determhie, in many cafes, the ratio in which this 

 lakes place,) I think W'e iire then weighing the heat itfelf, 

 or, rather, thediffeffenccof the quantity iri a body at one tem- 

 perature and at another^ 



Tn the Various direct attempts that have becri made tof 

 weigh heatj I fear philofophers have been following a plart 

 juil about as rational as it would be in the hih&bitants of thef 

 ocean to attempt to weigh water by employing a balance 

 hilpcnded in the medium that furrounds them, arid putting 

 hito one (liell a fubftance that to them flrOuM feCm wet, and 

 into the other a fubltance which they might call dry ! 



If we could abftracSl heat entirely from a l^odv, as we can 

 t\r from a glafs balloon, We ihotild be at rio lofs in weighing 

 it : but is there tK) other way c^f weighing air but that one > 

 If T take a bladdct containing afi unknown quantity of air, 

 1 cafily find out how nuich W'cight is required to fink it in 

 <r.'ater : if I afterwards pour into the bladder a known bulk of 

 jiir,- f can come at a knowledge of the weight of the air fo, 

 added by attendiivg tO* the quantity of weight now riecellliry 



4 IQ* 



