1. All the early stages of fertilization which previously I was unable to find have been described.
2. The ripe spermatozoon within the collar-cells of Sycon ciliatum has been found.
3. The spermatozoon in all cases examined lies towards the basement side of the cell (figs. 5 and 6, Pl. 19).
4. The enlargement of this flagellate filiform spermatozoon into the oval structure found in the sperm nurse-cell (fig. 7, Pl. 19) has been followed (figs. 3, 4, 6, 8, Pl. 19).
5. The cytoplasmic elements of the sperm nurse-cell have been described (fig. 7, Pl. 19).
6. A later stage in fertilization showing the male pronuclear centrosomes has been-figured (fig. 12, Pl. 20).
Cell Granules of Sycon and Grantia
7. The ‘chromidia’ of Dendy are regarded as mitochondria (fig. 8, Pl. 19; figs. 9, 12, Pl. 20).
8. A remarkable process of fragmentation of choanocyte ‘chromidia’ or mitochondria to form peculiar spheres is described (figs. 21, 22, 24, 26, Pl. 21).
9. The granules of the eggs of Sycon ciliatum have been described (fig. 1, Pl. 19; fig. 18, Pl. 21).
10. Some new facts regarding the Golgi elements have been ascertained (figs. 14, 15, and 16, Pl. 21).
Spermatogenesis of Grantia
11. It is claimed that spermatogenesis in Grantia compressa may take place in two ways. First, from definite pockets of cells, lined by large cells, lying in the matrix, as described by Görich and myself. Such areas to be regarded as testes. Second, from a direct rapid metamorphosis of collar-cells into spermatocytes.
The behaviour of the granules in such cases is figured in figs. 21-6, Pl. 21. In fig. 23, Pl. 21, is a supposed pachytene stage.
12. This metamorphosis involves the whole flagellated cavity, as in fig. 19, Pl. 21, each cell being like the one drawn in fig. 22, Pl. 21. The nucleus of this is like that of spermatocyte, and unlike a collar-cell nucleus. The collar is lost, but the flagellum remains.