and the Black Oxidt of Iron at a white heat. 253 



combustions. The hot wire or the electric spark kindles only the 

 portions of gas immediately adjacent to it, but the combustion of 

 those sets fire to the molecules contiguous to them, and these in their 

 turn to their neighbours, till all are made to burn. Thus the flame 

 travels after the original cause of combustion has ceased to operate 

 directly, and the momentary action of a small spark, or the transient 

 heat of a red-hot capillary wire may suffice to fire an infinitely large 

 mass of hydrogen and oxygen. There is no provision for a similar 

 propagation of decomposition through water or steam when either is 

 made white-hot; the absolute amount accordingly of disunion of the 

 elements of water occasioned is very small. 



If allowance, however, be made for the apparent difference in ex- 

 tent of effect which heat shews in uniting and in disuniting the ele- 

 ments of water, the phenomena otherwise seem referable solely to 

 the intensity of the temperature to which hydrogen and oxygen are 

 exposed. The opposite processes might go on simultaneously, union 

 or disunion being determined simply by the different temperatures to 

 which different portions of the gases were raised. At least it seems 

 not improbable that if a mixture of steam and of hydrogen and oxy- 

 gen were exposed to electric discharge, decomposition of the steam 

 and combination of the hydrogen and oxygen might be effected by 

 the same spark, provided the volume of steam were not large. In 

 the track of the spark, decomposition would occur, so long as a white 

 heat prevailed. When the temperature fell, combination would hap- 

 pen where the spark had passed, if it had not already commenced in 

 the neighbourhood of its direct route. Similar remarks apply, mutatis 

 mutandis, to the action of a hot platinum wire on a mixture of steam 

 with oxygen and hydrogen. 



It may be objected to this view, that Mr Grove decomposes steam 

 in his eudiometer, and obtains a permanent bubble of gas, consisting 

 of hydrogen and oxygen. The bubble however obtained in this way 

 is very small, and could not probably be greatly increased. Mr 

 Grove has not mentioned how large a volume of hydrogen and oxygen 

 he could obtain in the same eudiometer, by alternately boiling the 

 water till the steam produced caused the liquid to fall below the wire, 

 and allowing the steam to condense till the water rose above the metal. 

 But I venture to say that no large volume of permanent gas could 

 be procured by this process if the same eudiometer were employed 

 many times successively. The combining action of the wire might 

 not take effect on the hydrogen and oxygen when their quantity was 

 small, and they were diluted through a large volume of steam, for in 

 virtue of the law of diffusion, the molecules of hydrogen and oxygen 

 would be separated from each other by molecules of water vapour ; 

 but when the latter diminished in bulk, it seems impossible to doubt 

 that kindling of the gases would occur. 



Mr Grove's experiments then do not appear to prove that heat of 



