To begin what we propose to do we believe it is necessary to repeat, as a thing that has to be borne in mind, that when we had contracted the original nebula from 6,000,000,000 miles in diameter to 58,000,000 miles, its density was only equal to a barometric pressure of _one-ninth_ of an inch of mercury, and its mean temperature had been increased only _one degree_, that is, from -274° to -273°; and we can add that, although we had given the original nebula ten times that diameter, the result both in density and temperature would have been the same when it was condensed to 58,000,000 miles in diameter. Then, again, we believe it necessary to repeat that by contracting the nebula from 58,000,000 to 9,000,000 miles in diameter its mean density was raised from 1/274th to full atmospheric pressure, and its mean temperature from -273° to zero of the ordinary centigrade scale, i.e. to the temperature of freezing water. These two results strike us, at first sight, as somewhat remarkable, seeing that what looks like almost unlimited condensation to 58,000,000 miles diameter produced only one degree of temperature, while the comparatively insignificant condensation of from 58,000,000 to 9,000,000 miles in diameter produced 273° of heat, _in the way we are accustomed to measure heat_. Following up these two facts gives rise to ideas that have been borne in upon us ever since we stumbled upon them when making the analysis of the nebular hypothesis. One of these notions was that, were it practicable, the most effectual mode of liquefying gases would be by putting any one of them into a sealed vessel, and confining it in another vessel in which a vacuum of 1/274th part of an atmosphere could be produced; no difficult matter as far as the vacuum is concerned, for a good exhausting air-pump would be all that is required. But the practicability? The vessel in which the vacuum is produced would have to be protected so that no extraneous heat could be conveyed or conducted into it in any way whatever. How this could be, or is, done without cutting off every possibility of manipulating the enclosed vessel, we do not see; but it seems evident that some method is available because something presenting the same difficulties has been actually done, as everybody knows. The only degree of vacuum of any use in the exterior vessel would be about one-ninth of an inch of mercury, because that would as we have just said, furnish a temperature of -273°. There would be no necessity for applying pressure to the gas experimented upon. In fact pressure would be an obstacle to the experiment, according to the theory of the air thermometer; and could only be of use by furnishing a larger quantity of liquid to be handled and examined.