308 PROF. W. A. BONE AND OTHERS ON 



would probably be well worth while to undertake a further extended study of the 

 matter in larger explosion vessels than we have employed, although it would be both 

 a costly and a laborious enterprise. 



Whereas, in the case of hydrogen, the influence of successive increments in x, 

 the volume ratio of the other combustible constituent to methane in the mixture 

 exploded, upon the actual oxygen distribution is proportional to a; 2 , in the case of 

 carbon monoxide, it is more nearly proportional to x. This points to some fundamental 

 difference between the modes of combustion of the two gases in flames ; thus whilst 

 the evidence is strongly in favour of the supposition that hydrogen is burnt directly 

 to steam as the result of trimolecular collisions 2H 2 + 2 = 2H 2 O, the results with 

 mixtures CH 4 +O a + CO seem to require some different supposition, such, for instance, 

 as an intermediary action of steam. 



PART V. EXPERIMENTS WITH MIXTURES OF ETHYLENE, HYDROGEN AND 

 OXYGEN OF THE TYPE C 2 H 4 +O 2 +xH 2 . 



(With Messrs. H. H. GRAY and J. B. DAWSON.) 



It has long been known that when ethylene is exploded with its own volume ol 

 oxygen there is a doubling of the volume in the cooled products, with the formation 

 of principally carbon monoxide and hydrogen, but without any separation of carbon or 

 condensation of steam, substantially in accordance with the empirical equation 



It has also been shown that the addition to such a mixture of hydrogen 

 sufficient to bring its composition up to C 2 H 4 + 2 + H 2 , does not cause any separation 

 of carbon on explosion, although there results a slight condensation of steam on 

 cooling.* 



That the above facts do not really imply a preferential combustion of carbon (as 

 some have supposed) was proved by a study of the behaviour of a mixture 3C 2 H 4 + 2O 2 , 

 which on explosion gives rise to large quantities of both carbon and steam, 

 together with methane, acetylene, hydrogen and oxides of carbon. Indeed, the 

 facts harmonise very well with the hydroxylation theory, which would require the 



H . C . OH 



intermediate formation of monohydroxyethylene , a substance which on 



H.C.H 



thermal decomposition would yield H 2 and j CH residues, the latter subsequently 

 either (i.) undergoing hydrogenation to CH 4 in an atmosphere sufficiently rich in 



* BONE and DRUGMAN, ' Trans. Chem. Soc.' 1906, vol. 89, pp. 669 to 671. 



