SARGENT: THE OPTIC REFLEX APPARATUS OF VERTP^BRATES. 179 



neris of either side, which, entering the ventricle through the charac- 

 teristic ependyma at the anterior end of the optic lobes, fuses with 

 other fascicles of fibres from the habenulae to form the fibre of Reissner. 

 The exact method by which the fascicles enter the ventricle and unite 

 to form Reissner's fibre ditfei's in difi'erent species. The fibre attains its 

 greatest observed diameter (25 micra) in some of the larger selachians. 



Through the posterior portion of its course the fibre gives off numer- 

 ous branches, which enter the walls of the cord. The posterior canal- 

 cells are early developed, and send their axons cephalad to join Reissner's 

 fibre. Each remains, by means of several dendrites, in connection with 

 walls of the cord from which it is derived. 



///. Ganoids. 

 A. Observations. 



" The Ganoids are the most generalized of the true fishes, those 

 nearest the stock from which the Teleosts on the one hand, and the 

 Dipnoi and Batrachia on the other hand, have sprung " (Jordan and 

 Evermann, '96-00, p. 98). Therefore this group off'ers a most interest- 

 ing and fertile field of investigation to the student of comparative 

 neurology. The brain is of a simple generalized type, perhaps nearer 

 to the direct line of brain development than that of any other group. 

 I have been favored in my studies in this group by a full series of larval 

 Amia and adult brains of several other species. The only mention of 

 this apparatus in the literature of the group is the brief description of 

 its cells in Acipenser contained in the papers of Goronowitsch and 

 of Johnston, which will be referred to later, and a brief reference to 

 the description of Goronowitsch in a paper by C. L. Herrick. 



1. Amia calva. a. Development. It was in the study of the devel- 

 opment of Reissner's fibre in Amia that the cellular connections of the 

 fibre were first made out, and the elaborate apparatus of which it forms 

 a part discovered. For this work Amia is the most favorable material 

 I have yet encountered, as the cells are large and the development of 

 the apparatus takes place almost entirely after hatching. Moreover, 

 the detailed studies of Reighard (:03) on the development of functional 

 activities in larval Amia make it possible to correlate the development 

 of structure and function. The development of the optic reflex ap- 

 paratus has been completely followed in a series of larvae of Amia from 

 the time of hatching to thii'ty days after. Some fifteen stages, cut in 

 transverse, sagittal, and frontal sections, covering this period have been 



