No. 3.] SEGREGATION OF THE SEX-CELLS. 489 



earliest condition, even if we assume that two or more have 

 been lodged in the gill region, and two in the anterior part of 

 the body. The majority of the cells in this larva are confined 

 to a region only 0.20 mm. long; and if we consider the doubtful 

 cells in sections 26 and 29, the total length over which these 

 cells are distributed is about 0.50 mm. from the anus forward. 

 The sex-cells in this stage measure 9-13 fi. Balfour's admirable 

 account of these "primitive ova" (Elasmobranch Fishes, pp. 

 130-136) might almost be used bodily to describe the same 

 structures in Micrometrns and Abeona, — 2.5 mm. long. He ob- 

 served that the younger ones contain many yolk spherules, and 

 suggests that the cells themselves may have migrated to their 

 position from a peripheral portion of the blastoderm, since "they 

 are the only mesoblast cells filled at this period with yolk spher- 

 ules." He was at a loss as to how they arose, and thought he 

 could detect cells intermediate in size between them and the 

 neighboring cells. As has been seen, the yolk particles simply 

 remain unchanged from the original condition when the sex- 

 cells are segregated. 



In connection with the origin of the primitive cells, an egg 

 containing about 1700 cells, in which no sign of gastrulation 

 (i.e., separation of blastoderm into two layers) can yet be no- 

 ticed, deserves attention. The blastoderm covers the entire 

 yolk, exclusive of the entodermic mass of protoplasm. Many of 

 the features of this egg are abnormal, but these very characters 

 may be instructive. There are in this egg six very large cells 

 in the blastoderm. The relation to the surrounding cells and 

 their relative size may be gathered from Fig. 11, s.c.f A nu- 

 cleus is not present in these cells, but the highly granular pro- 

 toplasm is radially arranged. The granules are more dense near 

 the centre of the cell, so that the peripheral part is less deeply 

 stained with fuchsine, and, in fact, the whole mass of the cell 

 appears less compact than that of neighboring cells, and gives 

 one the impression that these cells are disintegrating. The 

 radial arrangement of protoplasmic granules, would, however, 

 point to cell division. A spindle and dividing nucleus cannot 

 be seen. 



Three of these cells lie in close proximity. Two others are 

 placed near each other, and the sixth lies in their neighborhood. 



The other-features are two multinuclear cells lying in the 



