NEW LAW IN THERMOCHEMISTRY 21 



coefficients of c,, the latter, like other similar coefficients, hav- 

 ing been determined by trial, empirically. Now, in the general 

 formula, — c^ means one henotherm, — ^c^ four henotherms, 

 and — 5*:, five henotherms, to be subtracted from the divisor 

 for each pair of oxygen atoms contained in the substance 

 burned. These quantities correspond to the varying thermal 

 values of the internal oxygen, and they find their precedent in 

 the two values which nitrogen has given us. They represent 

 additions to the normal heats of formation of the various com- 

 pounds, as determined by the rules already laid down, when all 

 atomic unions are considered as being thermally equal ; but 

 their application in detail involves uncertainties which need 

 fuller discussion than I am now prepared to give them. As 

 an example I may point out one difficulty, as follows : 



The absolute heat of formation of water, H — O — H, as 

 shown in the formulae already developed, is six henotherms. 

 But the neutralization reaction, H — OH, as [interpreted by the 

 theory of solutions, gives only one henotherm. Hence it would 

 seem as if the formation of hydroxyl, supposing that it could be 

 isolated, would develop five henotherms. Now consider methyl 

 alcohol, CH3— OH. The three hydrogen unions with carbon 

 are doubtless normal, and equal to two henotherms each ; but 

 what values have the other bonds in the compound? If normal, 

 they also should liberate two henotherms apiece, but if they 

 follow the analogies suggested by water, they ought to give one 

 plus five, or six henotherms, making, with the methyl group, 

 twelve in all. The normal value for methyl alcohol is ten ; the 

 actual value is certainly higher, but the amount and distribution 

 of the excess are in doubt. If, from the coefficient of c^ we 

 infer that the increase is one, then the true heat of formation of 

 methyl alcohol should be eleven henotherms. On this suppo- 

 sition we should have four unions giving the usual two heno- 

 therms each, and one, the hydroxyl union, giving three. The 

 latter quantity is exactly half the absolute heat of formation of 

 water, and its assumption satisfies most of the conditions of the 

 problem. At all events, whenever hydroxyl is present in an 

 organic compound, the heat of combustion is modified by a 

 quantity which is represented by c^ in the general formula for 



