This cell elongates and its nuclei multiply. After assuming a somewhat columnar form, it divides into (1) a superficial investing layer, and (2) an axial portion.
In the female the superficial layer is only developed distinctly in the median part of the column. In the course of the further development the two ends of the column become the blind ends of the ovary, and the axial tissue they contain forms the germinal tissue of nucleated protoplasm. The superficial layer gives rise to the epithelium of the uterus and oviduct. The germinal tissue, which is originally continuous, is interrupted in the middle part (where the superficial layer gives rise to the uterus and oviduct), and is confined to the two blind extremities of the tube.
In the male the superficial layer, which gives rise to the epithelium of the vas deferens, is only formed at the hinder end of the original column. In other respects the development takes place as in the female.
In the Insecta again the evidence, though somewhat conflicting, indicates that the generative ducts arise very much as in Nematodes, from the same primitive mass as the generative organs. In both of these types it would seem probable that the generative organs were primitively placed in the body cavity, and attached to the epidermis, through a pore in which their products passed out; and that, acquiring a tubular form, the peripheral part of the gland gave rise to a duct, the remainder constituting the true generative gland. It is quite possible that the generative ducts of such forms as the Platyelminthes may have had a similar origin to those in Insecta and Nematoda, but from the analogy of the Mollusca there is nearly as much to be said for regarding them as modified excretory organs.
In the Echinodermata nothing is unfortunately known as to the ontogeny of the generative organs and ducts. The structure of these organs in the adult would however seem to indicate that the most primitive type of echinoderm generative organ consists of a blind sack, projecting into the body cavity, and opening by a pore to the exterior. The sack is lined by an epithelium, continuous with the epidermis, the cells of which give rise to the ova or spermatozoa. The duct of these organs is obviously hardly differentiated from the gland; and the whole structure might easily be derived from the type of generative organ characteristic of the Hydromedusæ, where the generative cells are developed from special areas of the ectoderm, and, when ripe, pass directly into the surrounding medium.
If this suggestion is correct we may suppose that the generative ducts of the Echinodermata have a different origin to those of the majority of[276] the remaining triploblastica.
Their ducts have been evolved in forms in which the generative products continued to be liberated directly to the exterior, as in the Hydromedusæ; while those of other types have been evolved in forms in which the generative products were first transported, as in the Actinozoa, into the gastrovascular canals[277].
[268] E. van Beneden (No. [556]) was the first to discover a different origin for the generative products of the two sexes in Hydractinia, and his observations have led to numerous subsequent researches on the subject. For a summary of the observations on the Hydroids vide Weismann (No. [560]).
[269] The Hertwigs (No. [271]) state that in their opinion the generative cells arise from the lining of the body cavity in all the forms whose body cavity is a product of the archenteron. We do not know anything of the embryonic development of the generative organs in the Echinodermata, but the adult position of the generative organs in this group is very unfavourable to the Hertwigs’ view.
[270] O. Hertwig, Die Chætognathen. Jena, 1880.