SARGENT: THE OPTIC REFLEX APPARATUS OF VERTEBRATES. 217 



missure. The rest of the trunks continue cephalad, and, curving ventrad, 

 enter the ventricle on either side of the ' Schaltsttick,' which is wide 

 and thin. These, joining with other fine trunks from further forward, 

 unite with those from the ventral fissure of the torus to form Reissner's 

 fibre, whose diameter is about 3 micra. Through the posterior portion 

 of the central canal Reissner's fibre gives oft' fine fibril branches. These 

 run for the most part toward the ventral side of the canal, and have a 

 course slightly oblique and caudad (Fig. 66, a). They enter the cord, 

 passing between the large ependymal cells, and are there lost to view, 

 the presence of numerous neuroglia fibres making it impossible to follow 

 them. 



LoPHiiDAE (221). Both young larvae and adult brains of Lophius 

 piscatorius have been studied. In the 8 mm. larvae, one week after 

 hatching, a delicate filament passing from the region of the posterior 

 commissure through the ventricles was identified as Reissner's fibre. 

 In the adult the spinal cord is short and the fibre tapers rapidly in its 

 course through it. In the fourth ventricle the diameter of the fibre 

 is 7 micra; in the central canal 5 cm., and farther, caudad the diameter 

 is 3 micra. 



D. Critical Discussion. 



In the extensive literature of the teleost brain the mesencephalon has 

 not been neglected. Its fibre-tracts and the structure of the tectum 

 opticum have been minutely described by many investigators. How- 

 ever, the morphology of the anterior portion of the roof, including the 

 torus longitudinalis, its finer anatomy, and its relations to surrounding 

 structures, has received very inadequate treatment. 



The torus longitudinalis was first described, in the herring, by Cams 

 ('14), who believed it to be the homologue of the fornix of higher verte- 

 brates, and so named it. Subsequently Gottsche ('35) agreed regarding 

 this homology. Stieda ('68, p. 24) first applied the term torus longi- 

 tudinalis. Baudelot ('70, p. 90) used the terras ' lame commissurale ' 

 and 'eminences commissurales' in referring to it. Fritsch ('78), who 

 believed the optic lobes to represent the cerebral hemispheres, considered 

 the torus longitudinalis to be homologous with the fornix and corpus 

 callosum, and refeired to it as the fornix. 



Sanders ('78, p. 754) , in giving the first description, in Mugil, of the 

 finer structure of the torus, says, "The cells (Figure 19) which constitute 

 the fornix are mostly of a spherical form, consisting almost entirel}'^ of 

 nuclei with only a narrow rim of protoplasm around them. . . . Occasion- 

 ally larger cells occur, which present a triangular shape from the greater 



