Herrick, Morphology of Brain of Bony Fishes. 29 



latter. It passes in some cases upon the cephalic aspect of the 

 tuber forming two dorso-lateral cell clusters. In Lucioperca 

 these cells are enormously developed. The function and nature 

 of these highly specialized cells is unknown ; they remind one 

 of the cells described by Fritsch in the dorsal part of the medulla 

 of Lophius, which are often penetrated by blood vessels and are 

 the sources of large fibres passing out with the fifth nerve. How 

 such a perforation could take place is illustrated by Fig. 5, Plate 

 VII. These cells also resemble those of the mesencephalic root 

 of the fifth and the cells of origin of Mauthner's fibres. 



In transverse sections of the eel the large cells of the 

 praeopticus nidulus which lie between the optic tracts give off 

 large, deeply stained, unmeduUated fibres which pursue a lateral 

 course with those of the ependyma toward the lateral walls, 

 i. e. optic tract region. 



The ependyma fibres apparently arch cephalad about the 

 optic tracts and find their ectal terminus on the caudo-lateral 

 aspect of the hemispheres, a course which is due to the fact that 

 the constricting effect of the optic chiasm has separated and dis- 

 torted the relations between the ectal and ental walls of the 

 region in question. 



2. Tuber cinerum a?id tract to ventral commissure. A 

 horizontal section of the infundibulum at the position in- 

 dicated in Fig. 2, Plate IV, shows that the cephalo-ventral wall 

 of the infundibulum consists simply of a thin layer of epithelium 

 upon which the fibres of the hypophysis lie on their way to their 

 definitive locus on the lateral aspects. On the caudal aspect lies 

 the transection of the caudal stalk of the hypophysis from the 

 saccate portion. The caudo-lateral aspects are occupied with 

 the recessi laterales infundibuli, of which more anon. The re- 

 gions adjoining the ventricles are well filled with cells and sup- 

 plied with capillaries. The cells are of the small fusiform sort 

 and send their processes cephalad for the most part. They are 

 especially numerous about the recessus infundibuli. 



In this region also it is possible to observe the origin of 

 clusters of nerve fibres from the cells mentioned and to follow 

 the tracts to their union with the ventral commissure. These 

 disperse bundles collect cephalo-laterad and form a strong bun- 

 dle (tractus tuberi ad com. ventralis.) These fibres do not decussate 



