490 REPORTS ON THE STATE OF SCIENCE. 



and also when Smithells and Ingle, about the same time, in their 

 researches on the Structure and Chemistry of Flames, discovered 

 hydrogen in the interconal gases of aerated hydrocarbon flames, x the 

 dogma of the preferential combustion of hydrogen became untenable. 

 Attempts were then made to revive an idea originally put forward by 

 Kersten in 1861 — namely, that ' before any part of the hydrogen is 

 burnt all the carbon is burnt to carbonic oxide, and tJiat the excess of 

 oxygen (if any) divides itself between the carbonic oxide and hydrogen. ' 

 H. B. Dixon was himself inclined to this view, and Smithells spoke of 

 the ' preferential ' burning of carbon both in his 1892 paper and in his 

 lecture on ' Flame ' at the Nottingham meeting of the Association in 

 1893, although he has since disclaimed any intention of exalting the 

 idea into a general doctrine. 



The idea of a ' selective ' combustion, whether of carbon 

 or of hydrogen is, however, so repugnant to well-established 

 principles that it could hardly be expected to meet with general 

 acceptance in any final or complete sense, and there were many 

 sceptics as to its validity. Moreover, whilst in the cases of acety- 

 lene and ethylene the assumption of a direct transition from the system 



C„H,» + — O a to nCO +— H., implies a simple transaction from 



the kinetic standpoint, an extension of the idea to the cases of such 

 hydrocarbons as propane or propylene would obviously raise serious 

 difficulties, 



2C 1 H«+30 2 =6CO + 8H 2> 



2C 3 H +3O 2 =6CO + 6H 2 . 



It therefore seemed necessary to consider whether the solution of the 

 problem might not lie in the assumption of an initial association of 

 the hydrocarbon and oxygen forming an unstable ' oxygenated ' or 

 ' hydroxylated ' molecule, as suggested by II. E. Armstrong many 

 years ago, and it was with some such possibility in view that the writer 

 undertook the re-investigation of the subject about ten years ago. 



Previous workers had confined their attention to combustion at 

 such high tempera lu res as prevail in hydrocarbon flames and the ex- 

 plosion wave — conditions highly unfavourable to the detection or isola- 

 tion of unstable intermediate ' oxygenated ' products, if such are really 

 formed. It was therefore decided first of all to make a systematic 

 study of hydrocarbon combustion at temperatures below the ignition- 

 points of the mixtures used, where the rate of oxidation would 

 be much slower and more controllable, and intermediate products 

 more stable. The hydrocarbons selected for investigation were 

 methane, ethane, ethylene, and acetylene, and at the outset of 

 the work it was fortunately discovered that all four gases com- 

 bine with oxygen at temperatures much below those at which either 

 hydrogen or carbon monoxide begin to be oxidised with any appreciable 

 velocity, or at which either carbon reduces steam or the reversible 

 reaction CO+OH 2 = COi+H a could have any influence whatever 

 upon the result. Conditions were thus established which precluded the 

 interference of secondary processes with the main line of change. 

 £ ' Smithells and Ingle, Tram. Chem. Soc, 1892, 61, 209. 



