SAKGENT : THE OPTIC REFLEX APPARATUS OF VERTEBRATES. 185 



Follownng is a tabulated record of the number and size of the optic 

 reflex cells of larval Amia, at different stages : 



(2) Posterior Canal-Cells. By the third day of larval development 

 in Amia, Eeissuer's fibre has grown backward through the fourth ven- 

 tricle and the whole length of the central canal. In earlier stages care- 

 ful search fails to show any trace of the fibre, except at the extreme 

 posterior end of the canal. Here I early saw a delicate fibre, but could 

 not trace it forward to the middle and anterior portion of the canal. 

 Tliis was the cause of much perplexity until I discovered at the extreme 

 end of the canal the cells from which it arose. 



At this stage the posterior end of the cord (Plate 4, Fig. 31) is 

 slightly enlarged at its tip, and within this enlargement the canal 

 dilates to form the ventriculus terminalis. There is no exterior open- 

 ing in this ventricle, such as the sinus which has been described in 

 Petromyzon by Studnicka ('95), nor is there evidence of a persistent 

 remnant of the neureuteric canal, such as I have figured in Squalus 

 (Plate 3, Fig. 22). 



At the time of hatching one may find in the extreme end of the 

 canalis centralis a number of small cells, 3 to 4 micra in diameter, lying 

 in the lumen of the canal and in the ventriculus terminalis (Plate 4, 

 Fig. 27). The walls of the cord surrounding this part of the lumen are 

 made up of a single layer of loosely aggregated cells, some of them in 



