I. Entoprocta. II. Ectoprocta. {_a._ Chilostomata. 1. GYMNOLÆMATA {_b._ Ctenostomata. {_c._ Cyclostomata. 2. PHYLACTOLÆMATA. 3. PODOSTOMATA (_Rhabdopleura_). ENTOPROCTA. The development of the larvæ of Pedicellina is known from the researches of Hatschek (No. 299) far more completely than that of Loxosoma, though it does not apparently differ from it except in certain details. In both the known Entoproctous genera the segmentation is regular or nearly so, though Hatschek believes that he has detected in Pedicellina a slight difference between the two first segmentation spheres, and regards them as constituting the animal and vegetative poles of the embryo. The segmentation in Pedicellina, to which genus alone the remainder of the description applies, results in the formation of a single-layered blastosphere, with a small segmentation cavity, in which the animal and vegetative poles can readily be distinguished owing to the smaller size of the cells at the animal pole. The hypoblast cells and the vegetative[119] pole become invaginated in the normal manner (fig. 128 A), the blastopore becomes narrowed to a slit with an anteroposterior direction, _i.e._ parallel to the line connecting the mouth and anus in the adult. At the hinder extremity of the blastopore there are present two conspicuously large cells (fig. 128 B, _me_), one on each side of the middle line. These cells give rise to the mesoblast. On the completion of the invagination the mesoblasts become covered by the epiblast (fig. 128 C, _me_). The blastopore then closes, but in the position it occupied the epiblast becomes thickened to form the rudiment of the vestibule, which at this stage constitutes a disc marked off by a shallow groove from the remainder of the body. [119] The succeeding statements about the gastrula are derived from Hatschek. According to Salensky a segmentation cavity is not present, and the hypoblast would seem to be formed by delamination or epibole. Barrois finds a gastrula in both Loxosoma and Pedicellina, but gives no details. Uljanin finds a segmentation cavity in Pedicellina, and Schmidt would appear to have observed a gastrula stage in Loxosoma. None of the accounts we have can be compared in fulness of detail to that of Hatschek. [FIG. 128. THREE STAGES IN THE DEVELOPMENT OF PEDICELLINA ECHINATA. (After Hatschek.) _s.c._ segmentation cavity; _a.e._ archenteron; _ep._ epiblast; _me._ mesoblast; _hy._ hypoblast. A is the commencing gastrula stage from the side in optical section. B is a slightly later stage from above in optical section. It shews the two primitive mesoblast cells. C is a later stage after the closure of the blastopore, viewed from the side in optical section.]