The ovum so constituted is developed either (1) from one cell out of an aggregation or layer of cells all of which have the capacity of becoming ova; or (2) from one of a number of cells segmented off from a polynuclear mass of protoplasm, not divided into separate cells. In both cases the cells which have the capacity of becoming ova may be spoken of as germinal cells, and in the case where the ova are ultimately developed from a polynuclear mass of protoplasm the latter structure may be called a germogen. In some cases the whole of the germinal cells eventually become ova, but as a rule only a small proportion of them have this fate, the remainder undergoing various changes to be spoken of in the sequel. Extended investigations have shewn that the distinction between germinal cells which are independent cells from the first, or derived from a germogen in which the nucleated protoplasm is not divided into cells, is an unimportant one; and closely allied forms may differ in this respect. It is moreover probable that a germogen of nucleated protoplasm is less common than is often supposed: it being a matter of great difficulty to determine the structure of the organs usually so described. A germogen is stated to be found in most Platyelminthes, Nematoidea, Discophora, Insecta, and Crustacea. A more important distinction in the origin of the germinal cells is that afforded by their position. In this respect three groups may be distinguished. (1) The germinal cells may form the lining of a sack or tube, having the form of a syncytium or of an epithelium of separate cells (Platyelminthes, Mollusca, Rotifera, Echinodermata, Nematoidea, Arthropoda). (2) Or they may form a specialized part of the epithelium lining the general body cavity (Chætopoda, Gephyrea, Vertebrata). (3) Or they may form a mass placed between the two elsewhere contiguous primitive germinal layers (Coelenterata[12]). [12] In all the Metazoa the generative organs are placed between the primitive germinal layers; and the peculiarity of their position in the Coelenterata depends on the absence of a body cavity and of a distinct mesoblast. Types of transition between the first and second group are not uncommon. Such types, properly belonging to the second group, originate by a special membranous sack continuous with the oviduct being formed round the primitively free patch of germinal cells. Examples of this are afforded by the Discophora, the Teleostei, etc. It is very probable that all the cases which fall under the first heading may have been derived from types which belonged to the second group.