This formation of a heterogeneous aggregate with its variously adapted parts, which while influenced by the whole are mainly self-formed, goes on among units of essentially the same natures, inherited from units who belonged to similar societies. And now, carrying this conception with us, we may dimly perceive how, in a developing embryo, there may take place the formation, first of the great divisions--the primary layers--then of the outlines of systems, then of component organs, and so on continually with the minor structures contained in major structures; and how each of these progressively smaller divisions develops its own organization, irrespective of the changes going on throughout the rest of the embryo. So that though all parts are composed of physiological units of the same nature, yet everywhere, in virtue of local conditions and the influence of its neighbours, each unit joins in forming the particular structure appropriate to the place. Thus conceiving the matter, we may in a vague way understand the strange facts of autogenous development disclosed by the above named experiments. § 97f. "But how immeasurably complex must be the physiological units which can behave thus!" will be remarked by the reader. "To be able to play all parts, alike as members of the whole and as members of this or that organ, they must have an unimaginable variety of potentialities in their natures. Each must, indeed, be almost a microcosm within a microcosm." Doubtless this is true. Still we have a _consensus_ of proofs that the component units of organisms have constitutions of extremely involved kinds. Contemplate the facts and their implications. (1) Here is some large division of the animal kingdom--say the _Vertebrata_. The component units of all its members have certain fundamental traits in common: all of them have proclivities towards formation of a vertebral column. Leaving behind the great assemblage of Fishes with its multitudinous types, each having special units of composition, we pass to the _Amphibia_, in the units of which there exist certain traits superposed upon the traits they have in common with those of Fishes. Through unknown links we ascend to incipient Mammalian types and then to developed Mammalian types, the units of which must have further superposed traits. Additional traits distinguish the units of each Mammalian order; and, again, those of every genus included in it; while others severally characterize the units of each species. Similarly with the varieties in each species, and the stirps in each variety. Now the ability of any component unit to carry within itself the traits of the sub-kingdom, class, order, genus, species, variety, and at the same time to bear the traits of immediate ancestors, can exist only in a something having multitudinous proximate elements arranged in innumerable ways. (2) Again, these units must be at once in some respects fixed and in other respects plastic.