166 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



uucleolus was present, but eveu then the count is probably too large, 

 as some cells have two nucleoli, and in some cases the nucleolus was 

 probably sectioned. 



In form the cells of the roof nidulus vary greatly, the form of the 

 cell-body and the manner in which neurites are given off being deter- 

 mined by the compactness of the tissue in which they lie and the 

 pressure of surrounding cells. Generally pyriform, they may be spher- 

 ical, oval, or spindle-form (Plate 11, Fig. 71). The two largest pro- 

 cesses may come off in T-form from the tapering end of a pyriform cell, 

 or they may come from the opposite ends of an elongated spindle cell 

 (Fig. 18). These are the two extremes, between which there is every 

 gradation. At the anterior end, where the tectum is thickest and where 

 the cells are crowded together from all sides, they take a generally 

 spheroidal, ovoid, or polygonal form, as determined by the degree of 

 pressure from different directions. Midway in the optic lobe, where 

 they are close to the ventricle and consequently relieved from pressure on 

 that side, the pyriform type predominates. If closely crowded, they 

 may have a cuboidal or polygonal form. At the posterior end of the 

 tectum, where the few scattered cells lie amid the strands of the fibre-tract 

 passing to the cerebellum, the cell necessarily takes an elongated spindle 

 form, the degree of elongation being determined by the closeness of the 

 fasciculi of fibres between which they lie. Similar variations in the 

 form of these tectal cells occur in birds, where the cells have a similar 

 distribution. On the other iiand, in Amia, where the tectal reflex cells 

 are concentrated at the anterior end of the tectum, and the conditions 

 of pressure, etc., are more uniform for all the cells of the group, these 

 cells are almost uniformly spherical, tending to the polygonal outline. 

 These variations merely indicate that the external form of the nerve-cell 

 has little significance, except that it is the expression of all the external 

 forces to which the cell is subjected. 



Numerous capillaries in the roof of the optic ventricle, ramifying 

 among these cells (Figs. 15, 18), supply them with an abundance of 

 blood, and bear evidence of their activity. At the anterior end of this 

 series, near the posterior commissure, I have observed a considerable 

 number of these tectal rfeflex cells which are apparently undergoing 

 atrophy and degeneration, showing all the stages in the process that have 

 been observed in the atrophy of the dorsal giant cells of the spinal cord. 

 This degeneration probably affects only a few of the many cells of this 

 group. As yet I am totally ignorant of its meaning. 



Occasionally, especially in the middle cells of the series, I have seen 



