ON GASEOUS COMBUSTION. 4 ( J3 



an earlier stage lliau in slow combustion, but they do not precede the 

 onslaught of the oxygen upon the hydrocarbon, as was formerly sup- 

 posed, but arise in consequence of it. 



In considering explosive combustion, therefore, it is necessary to take 

 into account the possible modes of decomposition of the hydroxy 

 derivatives formed in the first two stages of slow combustion, because 

 these derivatives are so manifestly unstable at the high temperatures of 

 flames that they would at once break down into simpler products. Ethyl 

 alcohol, it is known, decomposes into ethylene and steam; acetaldehyde 

 into methane and carbon monoxide, or into carbon, hydrogen, methane, 

 and carbon monoxide, according to the temperature; formaldehyde is 

 resolved into carbon monoxide and hydrogen without the slightest separa- 

 tion of carbon. Thus: — 



Ethyl Alcohol Acetaldehyde Formaldehyde 



C 2 H 5 .OH CH,.CHO H.CHO 



C. 2 H,+ H,0 J CH 4 +CO ! CO+H. 2 



l C+2Ha+CO' 



Therefore, taking as examples the typical cases of ethane and ethylene, 



the scheme for explosive combustion becomes 



1 2 



H,C.CH, -* H, t C.CH,OH -> H«C.CH(OH). 2 



0,117+^0 HaC.CHO + H.,6 



| CH 4 +CO L 

 l C+Ha+CO' 



1 2 



H 2 C:CH 2 -> H 2 C:GHOH -> HO.HC:CH.OH 



20+H. 2 +H 2 O 2CH 2 



2CO+2H. 2 



with the proviso that, in a sufficient supply of oxygen, the transition 

 from the hydrocarbon to the dihydroxy stage is so rapid that no breaking 

 down of the molecular structure occurs in passing through the inono- 

 hydroxy stage. When, however, the supply of oxygen is reduced below 

 the equimolecular proportion it is evident that the initial monohydroxy 

 product cannot all be further oxidised to the dihydroxy stage ; some of 

 it must therefore decompose, yielding, usually, steam as one product. 

 It will be immediately perceived that the above theory affords a complete 

 explanation of the well-known fact that either ethylene or acetylene 

 on explosion with its own volume of oxygen yields carbon monoxide 

 and hydrogen, without any separation of carbon or steam formation, 

 because in each case the principal intermediate product is formaldehyde, 

 a substance which decomposes straight into carbon monoxide and 

 hydrogen at high temperatures. 



There are, moreover, two groups of facts relating to explosive 

 combustion which, whilst they completely subvert the notion 

 of a preferential combustion of carbon, harmonise very well 

 with the new view of ' hydroxylation.' They are, briefly, as follows: 

 (1) That whereas mixtures of olefines and oxygen corresponding to 



oa 



C ! n H ;ft +-0. 2 behave on explosion like ethylene, inasmuch as they yield 



1910. K K 



