248 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



provided with special means of attachment. These radiculae are seldom 

 traceable into the capsule, because either they are cut off, as often hap- 

 pens in sections, or, as in methylen-blue preparations, they are not stained 

 well at some distance from the cell. I have been able to follow them 

 best in the accessory retina. A methylen-blue preparation of an acces- 

 sory retina contained the cell shown in Figure 36 (Plate 3). This cell 

 gave rise to five radiculae and a neurite. The distal end of the cell 

 points away from the observer, so that the figure shows an oblique, basal 

 view of the cell. One radicula — the one at the left — rises toward the 

 observer as it approaches the ventral wall of the capsule ; the others, at 

 different levels, are directed toward the front face of another part of the 

 capsule. Close to the cell the radiculae stain dark blue ; farther out 

 they grow faint and indistinct and finally branch at points which are 

 slightly enlarged. Examining in a tangential section the middle layer 

 of the capsule next to the accessory retina, I found many very faintly 

 blue branches which are larger than similar branches from the chief 

 retina, for they arise from larger cells. It was in the study of these tan- 

 gential sections that I found the small radicula of the cell shown in 

 Figure 28 (PJate 3). Rising toward the eye, it had escaped notice as I 

 focused sharply on the cell. I have tried to show the diminution, of 

 color toward the end of the radicula farthest from the cell by a decrease 

 in the gray tint. Several other figures (Plate 2, Figs. 14, 17 ; Plate 3, 

 Figs. 24, 29, 34) show similar processes. It is possible that none of 

 these processes are neurites, but from the manner of orientation we may 

 reasonably infer that those which point toward the optic nerve are nerv- 

 ous in function. In Figure 34 the cell is viewed from its distal end, the 

 nucleus being seen endwise. 



The neurite (Fig. 28, n't.) arises from the cell body as a process which 

 at first glance resembles the radiculae, being quite as thick as these. 

 To distinguish this part of the neurite from the more attenuated, prox- 

 imal part, I shall use the term " neurite-process." At some distance 

 from the cell the diameter of the neurites is much reduced (Figs. 28, 

 36), so that even with a magnification of eight hundred diameters they 

 appear as fine lines (Figs. 22, 35). 



The rod, as I have said, represents the distal part of the sensory cell. 

 Together the rods constitute the internal or central zone of the retina. 

 They are cylindrical or slightly tapering, and rounded at the free end. 

 Each rod is composed of two parts ; an axial part, or core, and a periph- 

 eral part, or mantle. In my methylen-blue preparations the mantle 

 is always separated from the core by a considerable space, so that the 



