saegent: the optic reflex apparatus of vertebrates. 173 



ventricle. The two neurites arise at the level of the dorsal surface of 

 the cell from its opposite ends (Plate 3, Fig. 21, c, Plate 2, Fig. 9). 



The tractus tecto-cerebellaris (Fig. 9, trt. td. chl), formed from the 

 posteriorly directed neurites of the ' Dachkern ' cells, consists of several 

 small fascicles of large fibres, and runs caudad through the tectum on 

 either side of the median plane and some distance from it. At the 

 posterior end of the tectum these lateral tracts (Fig. II, trt. fd. cbl.) 

 converge toward the median plane and, passing under the commissure 

 (coins. IV) formed by tlie decussation of the roots of the trochlear 

 nerve, pass upward into the basal fibre-layer of the cerebellum. The 

 fibres of the tractus tecto-cerebellaris, for the most part, decussate 

 (Fig. 11, dec) before entering the cerebellum, but some of them ap- 

 parently pass directly into the cerebellum without decussation. A few 

 cells of the ' Dachkern ' may always be found in Mustelus, ventral and 

 posterior to the trochlear commissure, apparently crowded backward so 

 as to lie almost within the base of the cerebellum (Figs. 9, 11, a). 



The axons of the ' Dachkern ' cells which run cephalad arise either 

 from the anterior end of the cell, or by the division of a single dorsal pro- 

 cess (Fig. 9). They are finer than the cerebellar neurites and less easily 

 followed. Passing dorsad and laterad from the cells, they take a course 

 cephalad, forming the tractus tecto-fibrae Reissneris (td. td.fl>r. Reis.) 

 on either side of the median plane, immediately below the dorsal de- 

 cussation of the tectum (Fig. 9, dec. d.) In the anterior portion of the 

 tectum these tracts converge toward the median plane and pass under 

 the posterior commissure toward the ependymal groove (Figs. 9, 13). 



Among the finer fibres of these tracts may be seen coarser ones 

 (Fig. 13), apparently formed by the fusion of several or many axons. 

 Numerous ependymal fibres here make it difficult to follow the in- 

 dividual nerve-fibres of this tract. Near the upper limit of the epend}''- 

 mal groove the fibres are lost, but apparently they pass between the 

 ependymal cells into the groove. In the process of section-cutting the 

 branches of Reissner's fibre are usually broken from their connections 

 with the brain tissue. However, in one series I have found Reissner's 

 fibre in place within the groove (Fig. 10). The constituent divisions 

 of Reissner's fibre enter the groove through the thickened ependyma; 

 and where the larger divisions enter, the ependyma is indented, forming 

 a conical pocket (Fig. 10, a, a). A direct connection between the fibre- 

 tracts described and Reissner's fibre has not been observed in this 

 species. However, we may conclude that the axons of the tractus 

 tecto-fibrae Reissneris, consolidated into a few main trunks, enter the 



