It has been contended by some that the medium which conveys the impression of light through transparent, bodies, is necessarily more dense within the body than without; but according to this theory the converse is true. A ray of light is a mechanical impulse, propagated through an elastic medium, and, like a wave in water, tends to the side of least resistance. Within a refracting body the ether is rarefied, not only by the proximity of the atoms of the body (or its density), but also by the motions of those atoms; so that if two _simple_ gases of different specific gravity be made equal in density by compression, their refraction will be approximately as their specific heats. In the case of solids and liquids, or even compound gases, there is a continual absorption of motion to produce the cohesion of composition and aggregation. And the specific heats of compound gases will be found greater than those of simple gases, in proportion to the loss of volume by combination, _ceteris paribus_. If impenetrability be a law of matter, the more a portion of atomic matter is condensed, the less ether will be found in the same space. The same is also true when the natural density or specific gravity of a gas is greater than that of another. And the lighter the gas, the more will this circumstance vitiate the experiments to determine its specific heat. There is, therefore, this great source of fallacy in such experiments, viz.: that the ether permeates all fluids and solids, and that _its specific heat probably far exceeds that of all other matter_. This is a fundamental position of the theory, in support of which we will introduce a fact announced by M. V. Regnault, which was published in the Comptes Rendus of the French Academy for April, 1853. He says: "In the course of my researches I have encountered, indeed, at every step, anomalies which appeared to me inexplicable, in accordance with the theories formally recognized. For the sake of illustration I will quote one instance: 1st, a mass of gas, under a pressure of ten atmospheres, is contained in a space which is suddenly doubled; the pressure falls to five atmospheres. 2d. Two reservoirs of equal capacity are placed in a calorimeter; the one is filled with a gas, under a pressure of ten atmospheres; the second is perfectly empty. In these two experiments, the initial and final conditions of the gas are the same; but this identity of condition is accompanied by calorific results which are very different; for while in the former experiment there is a reduction of temperature, in the second the calorimeter does not indicate the slightest alteration of temperature." This experiment tends to confirm the theory. In the first experiment, the sudden doubling of the space causes the ether also to expand, inasmuch as the sides of the vessel prevent the instantaneous passage of the external ether. In the second, both vessels are full, one of ether, and the other of air mixed with ether; so that there is no