action on the mercury or the glass, would have made its way out in the same manner. There is every reason for believing that a small quantity of grease round the stoppers would have made them perfectly tight.
Some very singular appearances have been observed by Mr. Gordon, of the Portable-Gas Works, which have led him to believe that chemical changes are occasioned by the sudden expansion of oil-gas, which do not happen when the expansion is gradual; a striking result of the change being the separation of carbon from the gas. The effect referred to is exhibited when oil-gas, compressed into vessels by a power equal to that of thirty atmospheres, is suddenly allowed to escape through a small aperture into the air. It was first observed accidentally, in consequence of the derangement of the valve of a large apparatus, into which the gas had been compressed to twenty-seven atmospheres. The gas escaped with immense velocity, and when an examination took place of what had happened, it was found that all the metallic part of the valve upon which the gas had rushed was covered with a black, moist carbonaceous substance, and the contiguous brick wall with dry, black carbon, the moisture in this case having been absorbed by the brick. Since that time, Mr. Gordon has repeatedly shown the effect, by allowing the gas to rush out with very great violence from a portable lamp against a piece of white paper, which becomes immediately covered with black carbonaceous deposit.
The general conclusion is, that as the gas thus rapidly expands, a partial decomposition takes place and carbon is separated. If this explanation should ultimately prove by further experiments to be true, it will be highly important, as affording an instance of the exertion of mechanical and chemical powers in those circumstances where they most closely verge upon each other. At present, we have but little knowledge of such phenomena, though the announcement in France of the production of several new compound bodies, possessed of peculiar
- ↑ Quarterly Journal of Science, xxiii. 204.
