DECOMPOSITION OF WATER BY HEAT. 333 



longest time without disintegration, the gas given off contained 

 0*15 of oxyhydrogen gas; from the whole result I beljeve 

 there is an action of the water on the silica (probably forming 

 a hydrate decomposable by heat) which is a bar to satisfactory 

 results. With other oxides, at least such as would bear an in- 

 tense heat, the difficulties were still more insuperable. Priestley 

 has shown that water will corrode glass, and if I mistake not, 

 others have shown the same effect produced on silica. 



Although, as applied to the facts detailed, I attached no 

 further meaning to the title of my paper than that which I 

 have above stated, yet in one or two theoretical inferences I 

 have certainly gone farther ; for instance, when I suppose the 

 possibility or probability of mechanical rarefaction producing 

 the same effects as heat, here (although I do not, indeed I 

 cannot conceive the existence of heat without matter) I cer- 

 tainly abstract from the proposition any consideration of solid 

 matter. In order to ascertain how far this view might be 

 founded on truth, I had thought of making a few experiments 

 on the effect of mechanical rarefaction on the tendency of 

 gases to combine, but (in addition to the interference of neces- 

 sary occupations) I find that M. de Grotthus has already 

 experimented on the point. His experiments, as far as they 

 go, corroborate the views I have put forth. 



He finds * that mixed gases, such as chlorine and hydrogen, 

 or oxygen and hydrogen, when rarefied either by slow incre- 

 ments of heat or by the air-pump, do not take fire (' ne s'enflam- 

 ment pas') by the electric spark. From the context he evidently 

 means that the gases will not detonate or unite in volumes, 

 as he states that a partial combination ensues. Grotthus 

 appears to have considered the combination of gases by the 

 electric spark as an effect of sudden compression or molecular 

 approximation, certain particles being brought within the 

 range of their affinities by the sudden dilatation of others. 

 Although he did not pursue the subject far enough to ascer- 

 tain whether a degree of rarefaction could be reached which 

 would be an actual bar to combination, still his experiments 

 strengthen those views which assimilate mechanical and 



* Annaks de Chitnie^ vol. Ixxxii. 



