12 



The values are near together, and when we remember that 

 all the errors of each determination are accumulated in this one 

 final difference, the agreement is surprisingly good. The one 

 bad item in the table is the case of carbon tetrachloride, and this 

 has already been noted as questionable. 



The rule, then, for the thirty-three compounds now under 

 consideration, is simple. To Jind the absolute heat of formation 

 of any compound in this group, multiply the number of atomic 

 unions in its molecule by zj ^00 calories, and the product is the 

 value desired. The factor 27 400, of course, is subject to some 

 adjustment later, as the magnitude of the henotherm becomes 

 better known. For other series of compounds other funda- 

 mental "constants hold, as we see already in the estimates for 

 carbon dioxide and water. This question need not be con- 

 sidered any more fully just now. 



One other application of the law now before us is worth not- 

 ing, although the requisite data relate, not to gases, but to solu- 

 tions. Many organic compounds unite directly with bromine to 

 form addition products, heat being evolved. Each bromine 

 atom taken up means one more atomic linking, and therefore a 

 liberation of 27400 calories. On the other hand, each bromine 

 molecule requires 27 400 calories for its dissociation. When 

 Brj acts upon an olefine, then, 27 400 calories are lost, and twice 

 27 400 are evolved. The gain in heat, therefore, should be 

 27 400 calories. The following data by Louguinine and Kab- 

 lukoff* will serve to illustrate this relation. The reactions 

 were effected upon substances dissolved in carbon tetrachloride. 

 I include with these measurements the evidence in the case of 

 ethylene, as determined by Berthelot," who employed gaseous 

 substances. 



1 Compt. Rend., ii6, 1197, and 124, 1303. 

 *Essai de Mec. Chim., i, 343. 



