60 



The author next proceeds to consider the influence of rarefaction, 

 produced by heat, upon combustion and explosion. A volume of air 

 at 212 is expanded to 2'25 volumes. At a dull red heat its proba- 

 ble temperature then is 1032, provided it expand equably for equal 

 increments of heat. 



M. Grotthus has concluded that expansion by heat destroys the 

 explosive powers of gases, but Sir H. Davy found that two parts of 

 oxygen and one of hydrogen expanded to 2'5, its original bulk, de- 

 tonated at a red heat, and in another experiment, even at a lower 

 temperature ; whence it appears, that detonating gaseous mixtures 

 have their inflammability rather increased than diminished, by ex- 

 pansion by heat. In prosecuting these inquiries, the author dis- 

 covered that a mixture of oxygen and hydrogen produced water at 

 a temperature below visible redness, and without explosion or even 

 any luminous appearance ; and at a temperature a little above the 

 boiling point of quicksilver, charcoal converts oxygen into carbonic 

 acid without any of the ordinary phenomena of combustion. 



The third section relates to the effect of the mixture of different 

 gases upon explosion and combustion. When 1 part of a mixture of 

 oxygen and hydrogen, in the proportions that form water, is mixed 

 with 8 parts of pure hydrogen, the electric spark does not inflame 

 the mixture ; and its combustion is similarly prevented by 9 parts of 

 oxygen, 1 1 of nitrous oxide, 1 of carburetted hydrogen, 2 of sulphu- 

 retted hydrogen, 0'5 of olefiant gas, 2 of muriatic acid gas, and five 

 sixths of silicated fluoric acid gas. It therefore appears that other 

 causes, besides density and capacity for heat, interfered in these phe- 

 nomena ; for nitrous oxide, which is one third denser than oxygen, 

 and which has a greater capacity for heat, has lower powers of pre- 

 venting explosion ; and hydrogen, though fifteen times lighter than 

 oxygen, has a higher power of preventing explosion ; and olefiant 

 gas, in this respect, precedes the others in an infinitely higher ratio 

 than could have been expected either from its density or capacity. 



The author concludes this paper with some general observations, 

 and practical inferences founded upon the previous detail of facts. 

 Flame may be regarded as gaseous matter, of a temperature above 

 that which is capable of giving to solids a white heat ; for heated air, 

 though not luminous, will communicate that high temperature to 

 solid bodies. When we attempt to pass flame through fine wire 

 gauze, the metal so far cools the gaseous matter that it is no longer 

 luminous. The power of metallic and other tissues to prevent the 

 combustion of explosive gaseous mixtures, will depend upon the 

 heat required for their combustion, as compared with that acquired 

 by the tissue ; and the flame of those bodies which are most readily 

 inflammable, and of those which produce most heat in combustion, 

 will pass through a wire gauze capable of intercepting those flames 

 that produce little heat ; so that the flames of different substances will 

 pass through wire gauze at different temperatures. For instance, a 

 tissue that has 100 apertures in the square inch will intercept the 

 flame of alcohol, but not that of hydrogen ; and a tissue which would 



