of lime, and the uniform tenacity of solutions of soap in water, as shown in experiments with soap bubbles, both illustrate how equably substances held in solution are diffused. The evaporation of fluids into air, like the dissolution of solids in water, is in its results, at any rate, a case of assimilation to the character of the surrounding or adjacent medium; so that, the more we heat a bar of iron, the more progress does it make toward that vaporous condition in which it can easily be diffused through its environment. The mixing of metals also illustrates diffusion; for, whether it results in amalgams or alloys, whether the mixture be a merely mechanical association of the parts brought together or a chemical combination of those parts, the fact remains that for given areas, which may be large or small, the average degree of diffusion is the same. This is shown with especial clearness in those compounds of carbon and manganese with iron needed for a variety of industrial pur]30ses, since such compounds would have none of their present commercial value were it not for the uniform diffusion through the iron of the substance employed to modify it.
Here, then, our treatment of the subject must draw to a close. While necessarily brief, it has been complete enough to reveal a process far reaching in its scope and of cosmical significance. We have seen how like units everywhere tend to be associated and unlikes dissociated; how unlikes, held in forcible association, tend to be more or less profoundly assimilated to one another; and how disturbances of prevailing uniformity tend to be equably distributed through the several media in which they occur. But we have also noted that the power impelling to these multifarious acts of assimilation, to these movements of association and dissociation, is not the power of the units themselves, but the power of the system to which they immediately belong; and we are thus warned of the important bearing which our law has upon two problems of the utmost generality in physics—namely, the problems of chemical affinity and gravitation. It is true that we have as yet no formula for explaining these manifestations of power by a single principle; that we do not yet know the real structure of ether; and that there is still needed a definite account of gravitation as an intelligible mechanical process. Nevertheless, the causal connection of both gravitative and chemical actions with the ether system is already obvious to physicists. That the power which accomplishes these actions does not reside in matter alone, but resides also in the ether system—is, in fact, a function of that system regarded as including both ether and matter—seems to be increasingly pointed to by the trend of recent physical research. Basing our final conjecture, therefore, on generalization from a wide array of facts, it may not be premature to
