246 bulletin: museum of comparative zoology. 



retain a faint brown tiuge, which enables one to distinguish them from 

 the sensory cells, but is not dark enough to obscure the view. Since 

 acid fuchsin has a special affinity for the cell membrane, the outline of 

 the cells stained can be readily distinguished. After such staining it is 

 ■ seen that the pigment cells are grouped around the sensory cells in a 

 fairly definite way. In Figure 5 is shown in cross section a single group 

 of retinal cells. The group contains five prismatic pigment cells, which 

 together surround a smaller polygonal area, the cross section of a sen- 

 sory cell (cl. sns.). In Figure 12 two such groups are shown, and it 

 is to be observed that certain of the pigment cells belong to both groups. 

 There may be as few as four pigment cells in a group (Fig. 11), but I 

 have never seen fewer than four nor moi'e than five which shared in 

 encircling a sensory cell. Henchman, however, says they number from 

 five to seven. The various groups are not separated from each other by 

 narrow spaces such as Hilger ('85, Tafel 17, Fig. 19) has shown, but 

 are joined to each other closely, as indicated in Figure 12; the narrow 

 spaces in Hilger's figure are represented by the cell membranes in mine. 

 The cells of the retina are so arranged that the sensory cells are never in 

 contact with each other. Thus, these isolated groups strongly suggest 

 ommatidia, though they have not that regularity of arrangement which 

 is usually characteristic of the latter, nor is the recipient apparatus here 

 a differentiation of that side of the pigment cells which is turned toward 

 the centre of the group, as is probably the case in the ommatidia of most 

 arthropods. 



Sections of the retina along the chief axis of the eye when depig- 

 mented (Fig. 4, Plate 1, Fig. 14, Plate 2) give most instructive views of 

 the relation of the pigmented (cl. pig.) to the sensory cells (cl. s?is.). 

 Three groups of cells are shown in Figure 14. The pigment cells are 

 recognized by their coarsely granular appearance. 



There are no neurites arising from the sides or the branches of the pig- 

 ment cells, as some investigators have declared, and it will be shown 

 later that the pigment cells cannot be light-recipient elements. Hitherto 

 no one has traced out the connections of these indifferent cells except 

 Simroth ('76), whose observations contained so many errors that this 

 point seems to have been overlooked by his successors. Inasmuch as the 

 sensory cells in some species (e. g. Patella) are pigmented, much con- 

 fusion has characterized the interpretations placed on the retinal cells of 

 molluscs. Starting from different premises the conclusions of authors 

 have pointed now to the pigmented cells and now to the unpigmented 

 ones as sensory. Reasoning from the structure of the unpigmented cells 



