GASEOUS COMBUSTION AT HIGH PRESSUEES. 295 



stage, and that, therefore, the subequent thermal decompositions are restricted to 

 those of such hydroxylated molecules as CH 3 .OH and CH,(OH) 2 which do not involve 

 any separation of carbon. 



The above view of the matter is also consistent with another notable. fact, namely, 

 that in the case of mixtures intermediate between 2CH 4 + 2 , and CH 4 + 2 (vide 

 Table III.) the minimum proportion of the oxygen appearing as steam in the products 

 is found, not with 2CH 4 + O 2 , as might at first appear likely, but with 3CH 4 + 20 2 . 



During the relatively long cooling period, which follows the attainment of the 

 maximum explosion temperature, the following secondary interactions may come into 

 play, namely: (a) the reversible change CO + OH 2 ^Z^lC0 2 + H 2 ; and, in the case of 

 mixtures containing less oxygen than 3CH 4 + 20 2 , (fr) the interaction of steam and 

 carbon C + OH 2 = CO + H 2 ; or, possibly (c) some slight interaction between methane 

 and steam. 



In this connection the recently published work of G. W. ANDREW on the " Water 

 Gas Equilibrium in Hydrocarbon Flames,"* may be cited as proving that in a system 

 containing only CO 2 , CO, H 2 , and H 2 0, rapidly cooling down from the high 



temperatures prevailing in hydrocarbon flames, the equilibrium ratio -= ^ ~ adjusts 



itself automatically with the temperature until a point between 1500 C. and 

 1600 C. on the cooling curve is reached (corresponding to a value K = 4'0, approxi- 

 mately), after which no further re-adjustment occurs. He also found that the 

 adjustment of the equilibrium is not greatly influenced even when relatively large 

 quantities of methane and carbon are found in the final products. 



The results of our experiments, 1 to 3 inclusive, with the mixture 2CH 4 + O 2 , where 

 the initial pressures, 8'5 to 16'5 atmospheres were not sufficient for detonation, and in 



which the ratios j- 2 in the final products varied between 4'45 and 376 only, 



_tI 2 



confirm ANDREW'S conclusions. But in experiments 4 and 5, where with initial 

 pressures 2T69 and 31'25, respectively, the mixtures detonated, the much lower ratio 



CO x OH 



= 2'56 and 2'24, respectively, indicated some intervention of the separated 



x - 



carbon in the chemical interaction during the cooling period ; there is nothing, 

 however, in the results suggestive of any appreciable intervention of methane. 



This conclusion is also borne out by the following figures relating to the mixture 

 2CH 4 +0 2 , which show the ratio "B," which the sum of the oxygen as H 2 O, 

 plus half the oxygen as C0 2 in the products (assuming the C0 2 to have arisen wholly 

 from the interaction CO + OH 2 ~^I C0 2 + H 2 ) bears to the total oxygen originally 

 present in the mixtures fired in our experiments. The deviation of this ratio from 

 0'5 may be regarded as a measure of the participation of separated carbon in the 

 secondary interactions during the respective cooling periods. 



* 'Trans. Chem. Soc.,' 1914, TO!. 105, pp; 444-456. 



