That compounding and re-compounding through which we pass from the simplest inorganic substances to the most complex organic substances, has several concomitants. Each successive stage of composition presents us with molecules that are severally larger or more integrated, that are severally more heterogeneous, that are severally more unstable, and that are more numerous in their kinds (_First Principles_, § 151). And when we come to the substances of which living bodies are formed, we find ourselves among innumerable divergent groups and sub-groups of compounds, the units of which are large, heterogeneous, and unstable, in high degrees. There is no reason to assume that this process ends with the formation of those complex colloids which constitute organic matter. A more probable assumption is that out of the complex colloidal molecules there are evolved, by a still further integration, molecules which are still more heterogeneous, and of kinds which are still more multitudinous. What must be their properties? Already the colloidal molecules are extremely unstable--capable of being variously modified in their characters by very slight incident forces; and already the complexity of their polarities prevents them from readily falling into such positions of equilibrium as results in crystallization. Now the organic molecules composed of these colloidal molecules, must be similarly characterized in far higher degrees. Far more numerous must be the minute changes that can be wrought in them by minute external forces; far more free must they remain for a long time to obey forces tending to re-distribute them; and far greater must be the number of their kinds. Setting out with these physiological units, the existence of which various organic phenomena compel us to recognize, and the production of which the general law of Evolution thus leads us to anticipate; we get an insight into the phenomena of Genesis, Heredity, and Variation. If each organism is built of certain of these highly-plastic units peculiar to its species--units which slowly work towards an equilibrium of their complex proclivities, in producing an aggregate of the specific structure, and which are at the same time slowly modifiable by the re-actions of this aggregate--we see why the multiplication of organisms proceeds in the several ways, and with the various results, which naturalists have observed. Heredity, as shown not only in the repetition of the specific structure but in the repetition of ancestral deviations from it, becomes a matter of course; and it falls into unison with the fact that, in various inferior organisms, lost parts can be replaced, and that, in still lower organisms, a fragment can develop into a whole.