[Illustration: FIG. 28.--_Hydra viridis_, partly in section. M, mouth; O, ovary, or bud containing female reproductive cells; T, testis, or bud containing male reproductive cells. In addition to these buds containing germinal elements alone, there is another which illustrates the process of "gemmation"--i. e. the direct out-growth of a fully formed offspring.] This process clearly indicates very high specialization on the part of germ-cells. For we see by it that although these cells when young resemble all other cells in being capable of self-multiplication by binary division (thus reproducing cells exactly like themselves), when older they lose this power; but, at the same time, they acquire an entirely new and very remarkable power of giving rise to a vast succession of many different kinds of cells, all of which are mutually correlated as to their several functions, so as to constitute a hierarchy of cells--or, to speak literally, a multicellular _co-organization_. Here it is that we touch the really important distinction between the Protozoa and the Metazoa; for although I have said that some of the higher Protozoa foreshadow this state of matters in forming cell-colonies, it must now be noted that the cells composing such colonies are all of the same kind; and, therefore, that the principle of producing different kinds of cells which, by mutual co-adaptation of functions, shall be capable of constructing a multicellular Metazoön,--this great principle of _co-organization_ is but dimly nascent in the cell-colonies of Protozoa. And its marvellous development in the Metazoa appears ultimately to depend upon the highly specialized character of germ-cells. Even in cases where multicellular organisms are capable of reproducing their kind without the need of any preceding process of fertilization (parthenogenesis), and even in the still more numerous cases where complete organisms are budded forth from any part of their parent organism (gemmation, Fig. 28), there is now very good reason to conclude that these powers of a-sexual reproduction on the part of multicellular organisms are all ultimately due to the specialized character of their germ-cells. For in all these cases the tissues of the parent, from which the budding takes place, were ultimately derived from germ-cells--no matter how many generations of budded organisms may have intervened. And that propagation by budding, &c, in multicellular organisms is thus ultimately due to their propagation by sexual methods, seems to be further shown by certain facts which will have to be discussed at some length in my next volume. Here, therefore, I will mention only one of them--and this because it furnishes what appears to be another important distinction between the Protozoa and the Metazoa.