Herrick, Morphology of Brain of Bony Fishes. 41 



talis tract composed of large fibres which, passing caudo-mesad 

 and then ventrad, unites with the tract next to be mentioned 

 until the level of the nidulus ruber is reached, when it passes 

 into and through that nidulus, as already described. The sec- 

 ond tract arises in the mesal denser part of the nidulus corticalis 

 and consists of finer fibres, which can be distinguished from 

 those of the horizontalis tract even when the two are intimately 

 associated. The point of origin is considerably dorsad of the 

 posterior commissure. Describing a gentle arch ventrad, the 

 tract passes again dorsad into the anterior peduncle of the cere- 

 bellum. It is therefore a direct cerebeller thalamus bundle. In 

 mammals a similar bundle arises in the thalamus, comes likewise 

 into close relations with the red nidulus, and makes up the prin- 

 cipal part of the anterior peduncles of the cerebellum, but it is 

 a crossed bundle. The apparent incompatibility is reconciled 

 when we discover that this tract crosses to the other side within 

 the cerebellum of fishes. 



Plate VI, Fig. i, gives the dorsal course of the tract and 

 indicates by dotted lines it complete extent, while in Fig. 2, the 

 cephalic portion is in section. The region ventrad of the nidulus 

 corticalis contains the origin of Meynert's bundle from the 

 caudal margin of the habenje and the origin of the dorsal and 

 cephalic part of the posterior commissure fibres in the baso-lat- 

 eral part of the region of the corticalis. It seems probable that 

 the dorsal commissure is a decussation of the posterior longitu- 

 dinal fascicle, whose fibres are the only ones entering from the 

 caudo-ventral direction. The whole mass has the appearance 

 of a decussation, but at this writing evidence is not completely 

 satisfactory. 



The posterior longiiuditial fasciculus, after pursuing a course 

 parallel to the ventricle, until it reaches the vicinity of the third 

 nidulus, spreads out into a loose mesh-work passing among a 

 series of very large cells, from which the largest sized fibres of 

 the fasciculus spring. The fibres can be traced with ease from 

 the process of the cells into the tract. At this level fibres from 

 the tectum enter, and seem to communicate with the third and 

 fourth niduli, while others, after decussating, continue toward 

 the base and enter the middle peduncle of the cerebellum. 

 Fibres from ventral levels may associate themselves in this bun- 



