§ 101. The constituents of an aggregate cannot be re-arranged, unless they are moveable: manifestly, they must not be so firmly bound together that the incident force fails to alter their positions. No bodies are, indeed, possessed of this absolute rigidity; since an incident force in being propagated through a body, always produces temporary alterations in the relative positions of its units, if not permanent alterations. It is true also, that even permanent re-arrangements of the units may be thus wrought throughout the interiors of comparatively dense masses, without any outward sign: as happens with certain crystals, which, on exposure to sunlight, undergo molecular changes so great as to alter their planes of cleavage. Nevertheless, since total immobility of the parts must totally negative their re-arrangement; and since that comparative immobility which we see in very coherent matter, is a great obstacle to re-arrangement; it is self-evident that Evolution can be exhibited in any considerable degree, only where there is comparative mobility of parts. On the other hand, those definite distributive changes which constitute Evolution, cannot be extensively or variously displayed, where the mobility of the parts is extreme. In liquids, the cohesion of the units is so slight that there is no permanency in their relations of position to each other. Such re-arrangement as any incident force generates, is immediately destroyed again by the momentum of the constituents moved; and so, nothing but that temporary heterogeneity seen in circulating currents, can be produced. The like still more obviously holds of gases. Thus, while the theoretical limits between which Evolution is possible, are absolute immobility of parts and absolute mobility of parts; we may say that practically, Evolution cannot go on to any considerable extent where the mobility is very great or very little. A few examples will facilitate the realization of this truth.