The Eye of Byblis serrata. 51 



Turning now to the parts of the eye lying proximal to the middle layer 

 of cells, we notice that in all these deeper portions, which apparently cor. 

 respond to the rods, cones, and retinulse of Delia Valle, there seem to be 

 no nuclei, except those lying at the proximal or bottom part of the eye, 

 which is clearly the nuclear region of the retina. The omatidia embrace 

 at least the rods, cones, and reti unite. 



Rods. The rods lie immediately beneath and proximal to the middle 

 layer of cells, from which they are separated by a distinct line. The 

 rods, Fig. 4, r, are somewhat more numerous than the cells in the middle 

 layer. They are columnar, about as tall as the longest cells of the middle 

 layer, but some of the marginal ones are shorter. The rods are coarsely 

 granular. In oblique frontal sections through the chief axis of the eye 

 there is an indication that each rod may possibly be made up of two parts. 



Cones. Beneath and proximal to each rod, and in close connection with 

 it, is a crystalline cone, Fig. 4, c, which has a rounded cubical form and 

 is highly refractive. Each cone is homogeneous except for a white space 

 that usually occurs within its body. These spaces often have the appear 

 ance of more or less spheroidal cavities or vacuoles, but such vacuoles 

 generally indicate the plane of separation between the two component 

 parts of the crustacean cone. This apparent resolution of the cone into 

 two parts seems to be indicated in cross-sections by two opposite sharp 

 indentations of the outline. 



Relinulx. Closely adhering to each cone is a bundle of five fusiform 

 elements, Fig. 4, ret. The bundle at a deep level becomes resolved into 

 its separate elements, and at a still deeper level closely packed nuclei of 

 the retinula cells are found, Figs. 4 and 5, nu. These nuclei, which are 

 completely filled with deeply stained granules, are flask-shaped. A cross- 

 section through a fusiform bundle shows five granular retinula cells clus 

 tered about a highly refractive rhabdome composed of four rhabdomeres. 

 At the place where the bundles are resolved a considerable amount of 

 pigment is seen. In across section five /->- thick each retinal cell contains 

 about two grains of pigment. Nerve fibers have been traced from the 

 optic ganglia to the region of the nuclear layer of the retina, but the exact 

 connection with the retinal cells was not clearly seen. 



CONCLUSIONS. 



The eye of ByUis serrata, with its large lens, humor space, and complex 

 omatidia, seems to be a compound eye built on the general plan of a 

 simple ocellus, but also furnished with a space whose function may be 

 like that of the vitreous humor space of the vertebrate eye. The true 

 significance of this peculiar eye awaits the deft touch of the embryologist, 

 who, in taking up this sense organ, will certainly enter a field where much 

 is to be learned concerning the morphology of the arthropod eye. 



