ON GASEOUS COMBUSTION. 491 



The first experimental method consisted in maintaining mixtures of 

 each hydrocarbon with varying proportions of oxygen, sealed up in 

 borosilicate glass bulbs of about 60 c.c. capacity, at known constant 

 temperatures between 250° and 350°, for definite periods of time. Subse- 

 quently an apparatus was devised in which large volumes of the reacting 

 mixtures could be circulated at a uniform speed in a closed system 

 comprising (1) a surface of porous porcelain maintained at a constant 

 temperature in a combustion furnace, (2) suitable cooling and washing 

 arrangements for the removal of condensable or soluble intermediate 

 products, and (3) a mercurial manometer for recording pressures. 



By means of these two experimental methods it was proved, as 

 regards slow combustion — (1) that a hydrocarbon is ultimately burnt 

 to a mixture of steam and oxides of carbon without any separation of 

 carbon or liberation of hydrogen at any stage of the process; (2) that 

 the oxidation is marked by a very large intermediate formation of 

 aldehydic products; (3) that the fastest rates of oxidation are (in the 

 cases of the four hydrocarbons examined) always obtained with a ratio 

 of hydrocarbon to oxygen between 2 : 1 and 1 : 1, an excess of oxygen 

 above the equimolecular ratio always having a marked retarding 

 influence; and (4) that a large proportion of carbon dioxide is often 

 found in the products under conditions which preclude all possibility of 

 its formation either by the direct oxidation of the monoxide or by the 

 interaction of the monoxide with steam. 



Finally, the balance of evidence was so overwhelmingly in favour 

 of the supposition that combustion had proceeded by successive stages 

 of ' hydroxylation ' that the following schemes were put forward for 

 the typical hydrocarbons ethane, ethylene, and acetylene: — 



Formic Carbonic 

 Acid Acid 



CH a .CH 3 ->CH a .CH 2 OR-»CH 3 .CH(OH) 2 



Ethane Ethyl Alcohol „ ' , 



CH 3 .CHO + H 2 

 Acetaldehyde 



->CO + H 2 + H.CHO.->H.COOH->C< )(OH) 2 



Formal- , ' . ,— — ' s 



dchydc C0 + H 2 O C0 2 +H 2 



H 2 C:CH 2 ->H 2 C:CH(OH)-»HO.HC:CH.OH ,-* H.COOH -» CO(OH) 2 



Ethylene Vinyl Alcohol ,- l -— ' — — •. ,— — ' . 



2H.CHO 1 CO+H 2 C0 2 +H 2 



Formaldehyde 



HC : CH-»HO.C j CH-*HO.C : C.OH ,-* H.COOH -* CO(OH) 2 



Acetylene ,- ' -, -— ' , . ■ , 



CO + H.CHO 1 CO + H 2 C0 2 +H,0 



In other words, the attack of the oxygen upon the hydrocarbon may be 

 supposed to involve a series of successive ' hydroxy lations, ' the hydroxy- 

 lated molecules either breaking down or undergoing further oxidation, 

 according to their relative stabilities and affinities for oxygen at the 

 particular temperature, substantially as represented by the above 

 schemes. 



It should be mentioned, however, that whilst in many of the ' circu- 

 lation ' experiments a very large proportion of the intermediate alde- 

 hydic products were successfully removed from the sphere of action 

 before undergoing further oxidation — e.g., as much as 92 per cent, of 

 the formaldehyde indicated at the third stage of the ethane scheme and 



