supply of water, is reason for suspecting that this is the case. And similarly, though in the more numerous instances where loss of water leads to complete arrest of vital activity, we are unable to say that the immediate cause is a stoppage of molecular changes that results from a diminution of molecular mobility; yet it seems not improbable that this is the rationale of death by thirst. Probably few will expect to find this same condition to Evolution, illustrated in aggregates so widely different in kind as societies. Yet even here it may be shown that no considerable degree of Evolution is exhibited, where there is either great mobility of the parts, or great immobility of them. In such tribes as those inhabiting Australia, we see extremely little cohesion among the units: there is neither that partial fixity of relative positions which results from the commencement of agriculture, nor that partial fixity of relative positions implied by the establishment of social grades. And along with this want of cohesion, we find an absence of permanent differentiations. Conversely, in societies of the oriental type, where accumulated traditions, laws, and usages, and long-fixed class-arrangements, exercise great restraining power over individual actions, we find Evolution almost stopped. Through the medium of institutions and opinions, the forces brought to bear on each unit by the rest, are so great as to prevent the units from sensibly yielding to forces tending to re-arrange them. The condition most favourable to increase of social heterogeneity, is a medium coherence among the parts—a moderate facility of change in the relations of citizens, joined with a moderate resistance to such change—a considerable freedom of individual actions, qualified by a considerable restraint over individual actions—a certain attachment to pre-established arrangements, and a certain readiness to be impelled by new influences into new arrangements—a compromise between fixity and unfixity such as that which we, perhaps as much as any nation, exhibit. * * * * *