THE EYE OF PERIPATUS 167 
rather large, the latter are arranged at different levels in the 
cells. It is the nuclei of the rod-cells which collectively have 
been mistaken for an optic ganglion. 
Proximally the rod-cells are continued as nerve fibres, 
which form the very short optic nerve. The distal portions 
of the rod-cells are hexagonal in section, so that all fit together 
closely to form a mosaic (fig. 4). 
The rods are projections from the rod-cells, but the main 
part, the axis, of the rod is composed of a rather non-staining 
material. Thus in longitudinal sections the axes of the rods 
lie between the stained column-like bodies, whilst in trans- 
verse sections the rods would be the meshes of the grating 
(see fig. 8). The next question is, naturally, what is the 
‘orating’ itself, the part so easily mistaken for the rods in 
longitudinal section. It would appear as if this staiming 
substance was simply the peripheral portions of the rods. 
Each rod can be seen in maceration preparations to bear 
peripheral ‘ Stiftchen ’"—short processes very characteristic of 
invertebrate visual cells. These * Stiftchen’ clothe each rod 
completely, and it is the ‘Stiftchen’, or the ‘Stiftchen ’- 
borders, of the rods which stain up so readily and actually 
appear to be the rods in longitudinal sections. This explains 
why they show up as a kind of grating when cut transversely, 
for the ‘ Stiftchen ’-borders of adjacent rods touch each other 
(see figs. 1 and 8). 
Underlying the layer of rod-cells is a collecting region of 
nerve fibres—the prolongations of the sensory cells. These 
collect to form a short optic nerve (fig. 1, Op. N.) which enters 
the brain. The optic tract is traceable for some distance within 
the ‘ Punktsubstanz’. A delicate layer of connective tissue 
forms a capsule bounding both retina and optic nerve. 
COMPARISON OF THE EYE oF PERIPATUS WITH THAT OF OTHER 
ARTHROPODS, AND WITH THE PoLycHAETE [HYE. 
The Eye of Peripatus is in reality a very simple struc- 
ture compared with some insect ocelli. It is developed, 
as was discovered by Sedgwick (4), as a simple vesicular 
