of the Fishery Board for Scotland. 



283 



(PI. XIII., fig. 8, cb.), and the cells are arranged as a central patch sur- 

 rounded by peripheral cells slightly larger in size. 



The finer structure of the tectum lobi optici has been distinguished by 

 Stieda* into the following: — (1) a small border of "Grundsubstanz " 

 encloses (2) a layer of nerve fibres — the outer longitudinal layer of fibres ; 

 this again is succeeded on the inside by (3) a broad layer of granular 

 " Grundsubstanz " with spindle-shaped cells running through it and 

 giving it a transversely striated appearance. Proceeding inwards is (4) 

 the inner layer of longitudinal fibres, then (5) a layer of transverse fibres, 

 (6) a layer of small round cells, (7) a layer of net-like " Grundsubstanz," 

 and in the inside of this (8) cylindrical epithelium. Whiie this may be 

 the structure of the tectum of the adult, I have failed to distinguish 

 such an elaborate ground plan in larval or post-larval sand-eels. 



The Hind Brain. — The third primary division of the brain is early 

 marked off from the mid-brain, but at its posterior end it gradually merges 

 into the spinal chord. The deep cleft, which runs transversely between 

 the mid and hind-brain (Plate XI., fig. 2) has been already noticed, but 

 laterally and ventrally, there is no sharp line of demarcation from the 

 lateral and basal portion of the mid brain in the early larva (Plate XI., fig. 

 1, h.b.). The cells of the hind-brain are more closely arranged than those 

 of the mid brain, and this serves to distinguish at the sides the region of 

 each. It is a homogeneous mass of cells, which is only interrupted by 

 the fourth ventricle. Its anterior and dorsal portion— the cerebellum — 

 attains only a minimum development, till post larval stages are reached. 

 In fact, it is a very small structure, and only in the latest post larval 

 stage does its significance become manifest. The posterior portion of the 

 hind brain — the medulla oblongata — is ventral, and it is connected to the 

 other parts of the floor of the brain, though what Stieda * calls the pars 

 commissuralis with the pars peduncularis (taf. II., fig. 32, a.i.k.). 



The posterior portion, or medulla oblongata of a ten days old larva is 

 shown in Plate XL, figs. 7, 11, m.o. It forms the basal surface of the 

 fourth ventricle (v. 4), which is covered only by the thin tectum ventriculi 

 quarti (t.v. 4). The changes that take place in the third primary division 

 of the brain, might be summed up by saying, that in the roof, while the 

 anterior portion becomes greatly, but slowly thickened, the posterior 

 dorsal part remains thin, and separates the fourth ventricle from the 

 tissues above the hind brain, and that the floor is greatly developed. 



The cerebellum in larvae of ten days (Plate XL, figs. 8, 11, cb.), is very 

 small, and only represents the upturned posterior border of the second 

 transverse cleft, of which the fornix forms the anterior. It is at this stage 

 a cellular thin layer of tissue, only two or three cells in depth. In early 

 post-larval forms, the cerebellum is still a thin plate under the posterior 

 border of the tectum lobi optici, and it only arches over the most 

 anterior portion of the fourth ventricle. In specimens 5 mm. in length 

 (Plate XII. figs. 5-9), the cerebellum (cb.), lies close under the fornix, and 

 even its posterior border is all but covered by the posterior tips of the 

 optic lobes. When the sand-eel has attained a size of 12mm., it projects 

 beyond the optic lobes (Plate XIII., fig. 9, cb.), and it has increased in 

 size, especially in thickness. At its anterior end, where it joined to the 

 fornix, it is attached laterally to the base of the brain by tissue, which is 

 partly vesicular and partly fibrous. In Plate XIII., fig. 7, which is a 

 transverse section, through the united fornix and cerebellum, it overlies 

 the small dagger-shaped posterior portion of the aqueduct of sylvius (aq.) 

 and the cerebellar portion of the united parts is indicated on its upper 

 side by a line of cells, which makes its apex look like a mound. Its 

 breadth, when completely separated from the posterior border of the 



* Stieda. Ut supra. 



