104 Prof. H. E. Armstrong on the Determination of 



Although acetylene is especially prone to undergo change, I 

 think it is incorrect to speak of it as an " easily decomposable " 

 hydrocarbon (p. 84) ; its apparent instability is probably the 

 direct outcome of its " unsaturatedness " — of the readiness 

 with which it consequently enters into combination. 



(25) I certainly believe, with Lossen and Thomsen, that 

 compounds of the olefine and acetylene type are truly un- 

 saturated bodies ; and that the affinities of the carbon atoms 

 cannot be supposed to have satisfied each other in the manner 

 indicated by the formulae usually employed*. But, on the 

 other hand, I hold it to be both possible and probable that in 

 the formation of the olefines, for example, there is a partial 

 neutralization — a partial outgoing of energy — beyond that 

 which occurs in the formation of paraffins. What appears to 

 be direct confirmation of this view is afforded by Thomsen's 

 observations that the heat of combustion of trimethylene 

 exceeds that of its isomer propylene by 6690 units. Tri- 

 methylene is regarded by many asa u closed chain " hydro- 

 carbon, consisting of three CH 2 groups, while propylene is 

 methylethylene, the formulae being 



CH 2 CH.CH, 



H2O — CH 2 OH2 



Trimethylene. Propylene. 



Trimethylene combines much less readily with bromine than 

 does ordinary propylene ; but, in my opinion, there is no 

 chemical evidence to justify the assumption that the former 

 is a " closed-chain " compound. The properties of trimethy- 

 lene may be explained by regarding it as an open- chain 

 hydrocarbon of the formula CH 2 . CH 2 • CH 2 , which exhibits 

 two of the carbon atoms as possessing each a " free affinity;" 

 and the sluggish behaviour with bromine may be attributed 

 to the fact that the "free affinities" are not associated with 

 contiguous carbon atoms. Thomsen's observation, assuming 

 it to be correct, not only affords evidence in support of this 

 conclusion, but may also, I think, be held to prove that in 

 propylene the affinities of the carbon atoms partially satisfy 

 each other beyond the point which would be typified by the 

 formula H 2 C . CH . CH 3 , as the difference of 6690 units 

 between the heats of combustion of the isomers is considerably 

 below the probable value of v v If this view be correct, the 

 value of v 2 is not determined by deducting the difference 

 between the heats of combustion of a paraffin and the corre- 

 sponding olefine (§ 4) from the heat of combustion of gaseous 

 * See also Bruhl, Liebig's Annalen, 1882, ccxi. p. 162. 



