RESEARCHES ON THE VOLATILE HYUROi \RBONS. 149 



sity found and that calculated on the formula C 20 H u is not only more than twice as 

 large as the corresponding difference calculated on the formula C ]s II 12 , but that the 

 error is reversed; being with C2oH u a deficiency, while with Oj S H,.. it is an i.rcess. 

 This circumstance has to my mind a good deal of significance, as it goes strongly to 

 show that the lower formula is the true one. For of the many vapor densities of 

 hydrocarbons which I have determined, I have but rarely met with an instance in 

 which the density found was not greater than the theoretical density. And I have 

 usually observed that the excess of the experimental over the theoretical density is 

 larger in proportion as the boiling-point of the body is higher, a fact which needs 

 explanation. Wurtz* observed a similar difference between the determined and 

 calculated vapor densities of bodies of the formulae C n H n and C n H n+2 , which he 

 accounted for on the ground that his preparations contained an admixture of bodies 

 less volatile, the vapors of which would remain in the balloon, and increase the den- 

 sity. But I cannot accept this explanation for the substances here treated of, since 

 they invariably distil without residue within a range of one degree of temperature. 

 I would rather rely upon the supposition that the high temperature employed causes 

 partial decomposition of the substance, which would be the more liable to occur the 

 higher the boiling-point of the body. I do not, however, offer this as an explanation, 

 but merely make the suggestion. 



PART II- 



HYDROC 



AND (TMIXIC 



The oil of cumin employed in this research was furnished by Messrs. Reed and 

 Cutler, wholesale importers of drugs, etc. of Boston. The package bore the label of 

 Eduard Biittner, manufacturer, of Leipzig, and purported to be a genuine preparation, 

 answering in all of its obvious physical properties — odor, color, etc. — the description 

 given of this oil by Gerhardt and Cahoursf in their original memoir on this substance, 

 who, it appears, also employed a commercial preparation. Its behavior in distillation 

 left no doubt of its being a genuine article ; and this was afterwards confirmed by 

 treatment of the euminole with fused potash, for the production of cuminic acid, its 

 comportment with this reagent being in all respects identical with that described by 



Gerhardt and Cahours. Subjected to repeated series of fractionings by my process of 



* Bulletin de la Societe" Chimique de Paris, 1863, 309. 

 t Annates de Chimin pt de Phvsimie, 1841, 3 e S< : rie, J. GO. 



