Herrick, Morphology of Brain of Bony Fishes. 47 



ganglion layer is obvious, but Fig. i, Plate VII, and Figs, i, 2, 

 Plate IX, sufficiently differentiate them. A comparison with 

 Amphibia shows that the same arrangement prevails there also. 

 Thus Plate X, Figs. 9, 12, indicates the complete distinctness 

 of the epithelium from the ganglion layer. Each ganglion cell 

 has its peripheral process, which passes into the ectal layer of 

 the tectum and there divides dichotomously, mingling with the 

 optic fibres to form a fine reticulum. This zone is perforated at 

 intervals by the lemniscus and decussating cerebellar tracts, 

 whose fibres collect ectad. In many cases actual communica- 

 tion between fibre and cell could be seen, as in Figs, i, 3, 

 Plate IX. But there is also a partial commingling of fibres of 

 tlie next layer at some levels. 4, in Fig. i. Plate VII, is the 

 zone of anterior brachial fibres. These enter with the anterior 

 division of the optic tract and pass caudad immediately ectad of 

 the lemniscus fibre zone, gradually disappearing caudad. These 

 two tracts are approximately at right angles. 



The fifth layer (5, Fig. i, Plate VII,) is a dense band filled 

 with a homogeneous gelatinous substance, through which pass 

 the fibres from the apical processes of the ental ganglion layer. 

 It also contains sporadic cells of that layer as well as possibly 

 those forming" the termini of the anterior brachium fibres; 6 and 

 7 cannot be topographically distinguished with precision ; 6 is 

 chiefly composed of transverse (sylvian commissure) fibres, 

 while 7 is the layer of optic fibres passing more or less obliquely 

 caudad. Within these two layers is contained the ectal layer of 

 ganglion cells of large size and with elongate processes. The apex 

 process terminates in a strong fibre which divides dichotomously 

 and gives off its ultimate branches to the reticulum. Careful 

 study with high high powers showed that these fibrils actually 

 unite with or come into contact with the reticulum. The ap- 

 pearance of free termination, seen in Golgi preparations, seems 

 to be unconvincing, especially in view of the imperfect conser- 

 vation implied by that proceedure. The basal processes of 

 these cell turns cephalad and seem to enter the optic tract, as 

 seen in Figs, i, 2, Plate IX, which are derived from very young 

 fry of Trtitta. These cells are still horizontal, but we believe 

 that they gradually assume a radial position as a function of the 

 thickening of the tectum. 



