32 ■ Journal of Comparative Neurology. 



which have the aspect of small non-medullated nerve fibres. 

 These fibres occur in the form of a cylinder composed of sepa- 

 rate strands about a cavity. This cavity communicates by means 

 of a narrow horizontal tube in the substance of the cinereum 

 with the ventriculus infundibulus at the point where the latter 

 closely adjoins the anterior part of the hypophysis proper. 

 Tracing the peduncle of the saccate part of the hypophysis dor- 

 sad its fibres are followed to a point where they are associated 

 with similar fibres from the mammillary. Of the course of the 

 fibres, however, nothing sufficiently definite could be discov- 

 ered. They doubtless turn ectad. The ventral course of the 

 peduncular fibres is as in Didelphys. They pass ventrad and 

 diverge laterad, forming two irregular series on either side of a 

 median cleft. Beneath this level the sac becomes a compli- 

 cated maze of convolutions due to the folding of the walls and 

 the filling of the interspaces with a dense reticulum of blood 

 vessels. The numiber of vessels crowded into these narrow 

 lamellae is wonderful. The fibres lie m separate bundles within 

 the convolutions surrounded by blood capillaries. 



The walls of the convolutions are composed of a single 

 layered epithelium with large spherical nuclei. 



The fibres constituting the bundles above mentioned have 

 scarcely the aspect of connective tissue but stain densely and 

 assume a granular appearance. They are devoid of nuclei ex- 

 cept small granules which sometimes appear to enter a sheath 

 about the bundles but this seems to have no general significance. 

 In properly orientated sections we at last succeeded in demon- 

 strating that these fibres spring from the long apical process of 

 the cells of the epithelium, which therefore is exactly like the 

 similar layer of ependyma cells throughout the nervous system. 

 It should be observed that the process is turned from the ven- 

 tricular surface and the fibre passes among the blood vessels 

 which lie morphologically between the ependyma and the ectal 

 surface. The ectal surface corresponding to this epithelium lies 

 somewhere above in the thalamus and the blood vessels lie 

 morphologically between the two. The relation is difficult to 

 explain but if one will imagine a portion of the wall of the 

 thalamus consisting of an epithelium with fibres radiating to the 

 surface, while a number of blood vessels lie between the two, and 



