But it may be urged that the change of density, in these experiments, has not been carried far enough to justify the enunciation of a law of molecular physics. The condensation into less than one-third of the space does not, it may be said, quite represent the close file of men across Pall Mall. Let us therefore push matters to extremes, and continue the condensation till the vapor has been squeezed into a liquid. To the pure change of density we shall then have added the change in the state of aggregation. The experiments here are more easily described than executed; nevertheless, by sufficient training, scrupulous accuracy, and minute attention to details, success may be insured. Knowing the respective specific gravities, it is easy, by calculation, to determine the condensation requisite to reduce a column of vapor of definite density and length to a layer of liquid of definite thickness. Let the vapor, for example, be that of sulphuric ether, and let it be introduced into our 38 inch tube till a pressure of 7.2 inches of mercury is obtained. Or let it be hydride of amyl, of the same length, and at a pressure of 6.6 inches. Supposing the column to shorten, the vapor would become proportionally denser, and would, in each case, end in the production of a layer of liquid exactly one millimeter in thickness.[1] Conversely, a layer of liquid ether or of hydride of amyl, of this thickness, were its molecules freed from the thrall of cohesion, would form a column of vapor 38 inches long, at a pressure of 7.2 inches in the one case, and of 6.6 inches in the other. In passing through the liquid layer, a beam of heat encounters the same number of molecules as in passing through the vapor layer: and our problem is to decide, by experiment, whether, in both cases, the molecule is not the dominant factor, or whether its power is augmented, diminished, or otherwise overridden by the state of aggregation. [Footnote 1: The millimeter is 1-25th of an inch.]