^8 Journal of Comparative Neurology. 



can be followed into the infundibulum and constitutes the so- 

 called fornix tract described below. 



Fornix tract. Although warned of its apparent rashness, I 

 must again call attention to the fact that a definite bundle on 

 either side of the median line can be traced from just in front of 

 the anterior commissure to the dorsal part of the tuber, where it 

 dips ventrad into the immediate vicinity of the gray matter 

 forming the mammillaria of this paper. The tract can be seen 

 m almost its entire extent in longitudinal perpendicular sections. 

 It seems hard to avoid the identification with the fornix col- 

 umns. 

 III. — The Prcethalamus and Tuber cine reum. 



I. The niduli of the praethalamus have been previously 

 mentioned and seem to belong to a single system. 



The nididus pneopticus lies in the pra^thalamus cephalad of 

 the optic chiasm and is a part of Meynert's ganglion opticum 

 basale. It is composed of a mixture of cells of various sizes and 

 forms. In longitudinal sections of the drum the dorsal part of 

 this nidulus, where it adjoins the peduncles, contains very large 

 deeply staining multipolar cells giving off large fibres. There 

 are more numerous pale cells ventrad and adjoining the ventricle, 

 while a few scattered cells of less than half the size, about as large 

 as those of the fore part of the cerebrum, accompany them. 

 Numerous small " Deiters' nuclei" also cluster in this region. 

 The whole region is filled with blood vessels. In horizontal sec- 

 tions the principal cell cluster is seen to adjoin the ventricle 

 and is crossed dorsally by the fornix tract to the tuber. In the 

 ganoids the same nidulus occurs, but the cells are smaller and 

 less distinct. In Lucioperca the nidulus lies in the narrow 

 isthmus between the strong optic tracts, and the two por- 

 tions are very obvious. The dorsal part, consisting of 

 large multipolar cells, reaches nearly to the habeniilae. The 

 two portions are well shown on Plate VIII, Fig. 4, a and c, and 

 cells from the dorsal portion are figured in Figs. 5 and 5a of the 

 same plate. (See also Fig. i, Plate V; Plate VII, Fig. 7a. 

 Also vol. I, Fig. 5, Plate XX.) 



The nidulus postoptic us is a clustre of large multipolar cells 

 very like those of the dorsal part of the praeopticus and prob- 

 ably morphologically equivalent to a separated portion of the 



