138 ON FLUIDS. 



a quarter of a minute, much more when to feveral minutes, 

 we may fairly conclude, that there is a difference in conducting 

 power. 



In ah thefe experiments the fides of the apparatus fhould 

 have produced equal increments, had this been the caufe of 

 the rife of the thermometer ; and it is evident that currents 

 downwards could not affect it. That the fides of the veffel 

 could not communicate the temperature to the thermometer, 

 nor even the radiant caloric affect it in the manner obferved, 

 the eleventh experiment (which by the way arole from an error 

 in the mode of conducting the trial with fulphate of foda) fuffi- 

 ciently demonftrates. From an inflection of the table, it will 

 be feen, that the aqueous folutions of different falls differ ma- 

 terially from each other in the celerity with which caloric is pro- 

 pagated through them. 



I attempted to meafure the conducting powers of feveral of 

 the weaker acids, but I was foon convinced that their action on 

 the iron might invalidate the accuracy of the refults. 

 The fluids arc It will be unneceflary to obferve that if we find the thermo- 

 propercon uc- meter re q U j r ; n g different times for its elevation, in fuch cafes, 

 we muft afcribe it to the conducling powers of the medium 

 between it and the heated body. 



If I am not deceived, we may conclude from what I have 

 above adduced, that liquids as well as folios are conductors of 

 caloric ; that the tranfmifiion of it through them follows a par- 

 ticular law depending on the properties of the particular liquid, 

 but which is not in the exact ratio of any of their mechanical 

 properties, though nearer that of their dcnjlty than any other. 

 The Count's Such, Sir, are the principal arguments that feem to militate 



wdUxplained againft Count Rumford's hypothefis, which he has, with that 

 by the flow con- ingenuity which diftinguifhes his refearches, applied to the 

 *f fluids*^" y* fo! ut ' on °f many important phenomena of nature. Thefe, 

 its negation. however, may be equally well explained by fuppofmg liquids 

 V ,-ry flow cur- very bad conductors of caloric ; and, if the currents caufed in 

 " nt *£ !1 " P k* ln liquids ^y changes in temperature, have even a very inferior 

 Cbcicre of velocity to what he fuppofes, we may, I think, account iutri- 

 Chamouni. ciently well for the appearance he obferved on the Glaciere of 

 Chamouni, which he propofes as a left of his opinions, by 

 the decreafe in denfity of water while its temperature de- 

 fcends from 42° to 32°, (a facr which the Count's late expe. 

 rir.ients confirm) without affenting to his opinions with regard 



tQ 



