the Air^Thermometer . . 339 



, be infinitely remote, and yet the quantity of heat finite. In- 

 deed, this opinion is itself a contradiction in terms, with re- 

 gard to the heat in a thermometer, considered as measuring its 

 own heat. 



Mr Dalton supposed that mercury expands as the squares of 

 the true temperatures, reckoned from its freezing point; and 

 that, relatively to the degrees of this scale, taken in arithmeti- 

 cal progression, the expansions of the gases proceed in geome- 

 trical progression. But the researches of M. Gay Lussac, Dr 

 Ure, and of MM. Dulong and Petit, have proved fatal to this 

 hypothetical law of temperature, and have shewn, that mercury 

 and the gases observe the same law through a great range, only 

 at length diverging very slowly. So that after Mr Dalton''s er- 

 roneous progression in the expansion of air relatively to the mer- 

 curial thermometer is corrected, the relation between his scale 

 and that expansion is curiously metamorphosed. It is thus evi- 

 dent, that his scale bears a totally different relation to the com- 

 mon mercurial thermometer from that of the law I have inves- 

 tigated: the latter only differing about half the quantity that 

 Mr Dal ton's does from the old scale, between the freezing and 

 boiling points of water ; especially since Mr Dalton's scale, as 

 actually constructed, differs less from the old one than his theory 

 strictly allows. The more scientific part of chemistry certainly 

 owes much to Mr Dalton ; but though his law of temperature 

 had happened to be that of nature, it was still to be considered 

 unknown, and entitled to no confidence, so long as nothing sa- 

 tisfactory was advanced in its behalf. 



The specific heat of a given weight of air, is cateris paribus 

 independent of its density or pressure. 



For^ in the former figure, let the temperature of this air cor- 

 respond to the point E, and let EF x 1° and EG x 1° be re- 

 spectively the specific heats under a constant volume and con- 

 stant pressure ; suppose the air now to be condensed till its tem- 

 perature rise to H ; then HI x 1% which is less than EF x 1% • 

 will be its specific heat under a constant volume relatively to 

 the common scale * ; but whilst the temperature sinks to E, 



• It is obvious that the specific heat of air, relatively to the true scale, 

 must be independent of the temperature. 



y2 



