EXPEEIMENTAL KNOWLEDGE OF THE PBOPERTIES OF MATTER. 493 



nearly identical heat-capacities (of 22-32 litres) whether at constant 

 pi'essure or at constant volume ; and these are the gases for which 

 Regnault found that the specific heats did not vary much with variation 

 of pressure or of temperature ; they also obey the laws of Boyle and Gay- 

 Lussac very well. Of the remaining four, chlorine deviates more than 

 the others from agreement with the laws mentioned ; while ammonia 

 behaves in accordance with these laws, and methane does not deviate 

 much ; but the heat-capacity of a given volume of ammonia is very 

 different from that of an equal volume of either of the first six gases in 

 the table, as also are those of the gases methane, ethylene, chlorine, and 

 others ; while the heat-capacities of equal volumes of vapours of a large 

 number of liquids examined were still farther removed from those of the 

 same volume of the six gases mentioned in the table ; of which the heat- 

 capacities of equal volumes are equal to that of air within 1 percent., with 

 the exceptions of oxygen and nitric oxide, in which the difference is just 

 above 1 per cent, and of hydric chloride in which the difference is a little 

 less than 2 per cent. Now the aberrations of chlorine and ammonia from 

 the rule of constant heat-capacities of equal volumes of bodies in the 

 gaseous state are, as seen by comparison with the six gases, glaringly 

 wide ; the rule does not hold even approximately for chlorine and 

 ammonia. The deviations of these two gases from conformity with 

 Boyle's and Gay-Lussac's laws seem an inadequate explanation ; in the 

 case of ethylene, again, the rule of heat-capacities entirely fails. This 

 rule is therefore not supported by these wide investigations of Regnault, 



E. Wiedemann on Specific Heats of Gases. 



E. Wiedemann ^ repeats a number of Regnault's experiments on these 

 specific heats, with improved apparatus and methods, and with a modifi- 

 cation which in the case of vapours avoids condensation in the calorimeter, 

 60 that the specific heat of the vapour is determined directly in one experi- 

 ment and not by the difference of two experiments, as in Regnault's method; 

 the results differ slightly from Regnault's and are probably more accurate, 

 but go no nea,rer to finding any general law. The chief addition made by 

 Wiedemann is the determination of the variation of specific heat with 

 temperature in the case of a considerable number of gases and vapours. 

 With reference to the hint thrown out by Regnault, as to the possibility 

 of an ideal law applicable to ideal gases, and corresponding to Dulong and 

 Petit's law for solids, Wiedemann finds the true specific heats of various 

 gases, including nitrous oxide, carbon dioxide, and ammonia, and finds 

 that the specific heat of ammonia varies less with the temperature than 

 either of the other two, while it approaches more nearly to a perfect gas 

 in its elasticity or in its conformity with Boyle's law. Thus Regnault 

 gives -2 iorpv Ip'v', wherep' is about=2j7, the values 1-00651 and 1-01881 

 for nitrous oxide and ammonia respectively, showing nitrous oxide as 

 nearer to the condition of a perfect gas than ammonia. Thus of the two 

 gases nitrous oxide and ammonia the more nearly perfect gas, nitrous 

 oxide, has a more variable specific heat than the less perfect gas ammonia. 

 But these two gases, as also carbonic acid gas, are too far from perfect 

 gases to justify any inference on that score alone as to specific heats ; it 

 is still possible that, if all three gases could have their specific heats 



' Pogtf. An7i. 157, pp. 1-42; and Mled, Ann. 2, 1877, p. 195. 

 ^ Mem. de VAcad. t. xxvi. pp. 251, 252. 



