228 Royal Society : — 



tity of iodide of allyl is rapidly changed into a mass of hydriodates 

 of new compounds.' This crystalhne mass evidently contains a con- 

 siderable quantity of the hydriodate of diallylamine, but to succeed 

 in separatins; it from this very complicated mixture would have re- 

 quired a more corisiderable quantity of substance than we had at our 

 disposal. 



Triallylaimiip, Cj^ Hj^ N. — The oxide of tetrallylammonium, to 

 which we shall presently allude, submitted to the action of heat, is 

 decomposed, with liberation of a basic oil. The perfect analogy of 

 the oxide of tetrallylammonium with the corresponding ethyl-com- 

 pound, leaves no doubt respecting tlie nature of this basic substance. 

 It is obviously triallylamine. When saturated with hydrochloric 

 acid and mixed with bichloride of platinum, it deposits a yellow plati- 

 num-salt, the analysis of which leads to the formula 



CfiH^l N, HCl, PtCI,. 

 CeHj 



Oxide ofTetralhjlammonhm, C04 H^o NO, HO. — The chief product 

 of the action of ammonia upon iodide of allyl is a magnificent crystal- 

 line compound, which is deposited from the solution resulting from 

 the reaction. Iodide of allyl is rapidly attacked even by an aqueous 

 solution of ammonia at the common temperature. By a few days' 

 contact, a large quantity of the iodide dissolves, and the solution be- 

 comes a solid mass. If no deposition of crystals take place from 

 the solution, it is only necessary to add a concentrated solution of 

 caustic potassa, which causes the separation of an oily layer that 

 speedily solidifies. The crystals thus formed are the iodide of te- 

 trallylammonium, which, like the corresponding compound of the 

 ethyl-series, is but slightly soluble in solution of potassa. It is ob- 

 tained pure by exposing it to the air until the potassa is converted 

 into carbonate, and then reerystallizing it from absolute alcoliol. 



The iodide, by treatment with oxide of silver, is immediately trans- 

 formed into the oxide. This forms an alkaline solution, which pos- 

 sesses all the properties of the oxide of tetrethylammouium. 



The iodide contains 



C«H, LNI. 



On taking a rctropective glance at the results obtained in this 

 inquiry, it is obvious that propylene, a homologue of olefiant gas, is 

 susceptible of furnishing a mono-acid alcohol, which bears to it the 

 same relation that is observed between methyl-alcohol and marsh- 

 gas, or between benzoic alcohol and toluol, the analogue of marsh- 

 gas among the derivatives of the aromatic acids. The hydrocarbons 

 homologous and analogous to marsh- gas, the homologues of olefiant 

 gas, and indeed probably a very considerable number of other hydro- 

 carbons, may be regarded as starting-points for the productiou f 



