OF ARTS AND SCIENCES. 07 



merely as a mixture of butyl hydride and amyl hydride, C 5 H 12 , 

 which latter boils at about 30°. 



It is needless to remark that to call this a mixture, in the broad 

 sense that Ronalds has done, does not accord with my experience. 

 The boiling-point, 8°-9°, that I found for butyl iso-hydride was 

 the result of an actual determination in the usual manner. The 

 material employed for this purpose had been repeatedly and care- 

 fully distilled through my regulated condenser; but it had finally 

 become so reduced in quantity as to necessitate suspension of the 

 process of fractionation before the separation was so nearly com- 

 plete as is claimed for the bodies of higher boiling-point. Never- 

 theless, it is believed to have been sufficiently pure to justify the 

 conclusion — supported also by analogy — that there is a constit- 

 uent of the petroleum boiling at about 8° to 9°, as I had stated. 

 I would not, however, be understood to maintain that this body, 

 as obtained, contained no admixture of butyl hydride, or even of 

 amyl hydride, for doubtless it contained both of them to some ex- 

 tent. I only contend that probably neither was present in suffi- 

 cient quantity to materially, if indeed perceptibly, influence the 

 boiling-point, and consequently that the elevation of the boiling 

 temperature above 0°, or that of butyl hydride, was not due to the 

 presence of amyl hydride, as we should infer from Ronald's state- 

 ment, but rather to the fact that it was mainly composed of butyl 

 iso-hydride, which he overlooked. 



I may remark in this connection that a similar view is held in 

 regard to the purity of the other bodies that I have separated from 

 the petroleum; although the purification of most or many of these 

 was carried to the extreme practicable limit, they doubtless still 

 remained mixtures, but probably in such relative proportions as to 

 admit of correct, or at least very nearly correct, determinations 

 of their boiling-points. That such was the probable condition of 

 these bodies is not merely a theoretical opinion based on my own 

 observations. The experiments of Berthelot * show that mixtures 

 of neutral liquids whose boiling-points differ by 20°-30°, being 

 mixed in such proportion that the least volatile amounts to 8 or 

 10 per cent, do not admit of separation by distillation under ordi- 

 nary pressure, and will frequently, if not always, be found to com- 

 port themselves like homogeneous substances. 



Berthelot's results have since been confirmed by Alluard, f who 



* Comptes Rendus, LVII. 430. t Comptes Rendus, LVIII. 84. 



