840 THE AMERICAN NATURALIST. (VoL. XXXII. 
above, is shown in Fig. 4. One looks down into the boat- 
shaped retinas of ocelli I and III, the shape of which in cross- 
sections is well shown in Fig. 2. In these ocelli the flattened 
ends of the rods (shown on a larger scale in Fig. 1—'%) are seen _ 
as short, nearly parallel lines, the retinidial fibrils, not repre- 
sented in the figures, running at right angles to them. The 
median band of pigment and the double row of giant rods are 
indicated by heavy lines. The retina of ocellus V is seen end- 
Fic. 3. — Section through ocellus V of Acilius. 
wise, as though retina III were rotated 90° on its long axis 
and then stood on end, bringing the row of giant cells in a 
vertical position; see also Fig. 3. We therefore view the rods 
flatwise and lengthwise. Retinas II and IV are like those of 
eyes I and III, but smaller. They are placed so that the space 
between the two rows of giant cells lies in the plane of the 
paper, consequently we are looking at the giant rods lengthwise 
and edgewise. 
In Lycosa the retinas I, II, and III show a similar orientation 
to the three planes of space. These three retinas are in every 
