632 William Patten 



The outer parts of the retinulae seem to terminate with the knob- 

 like swellings containing- the nuclei; but this Ì8, in reality, not so, for 

 they are continued outwards as extremely delicate membranes, similar 

 to those already described for the retinulae of Arca. At the base of the 

 style, and especially at that point where it becomes continuous with 

 the pedicel, the membrane is so closely applied to the former that it is 

 with difficulty observed (figs. 83 and 84) : but farther toward the ex- 

 terior, the united terminal membranes of each group of retinulae form 

 a delicate sh ea th loosely surrounding the style, from which it is often 

 separated by a clear and narrow area. As the styles expand iuto the 

 closely packed, octagonal calyces, it is no longer possible to distinguish 

 the she ath, between the adjacent walls of the former, which are so 

 closely placed that their walls bave apparently fused to form one line. 

 But the smaller faces of the calyces enclose quadrilateral areas, left 

 for the passage of other elements , and in each of them a delicate 

 membranous tube may he distinguished (figs. 79 — 81 rt.s.). 



From theoretical considerations, it is probable that contiuuations 

 of the membrane exist between the walls of the calyces, but are in- 

 visible for reasons already suggested. It is also probable that the sac- 

 like membrane, in the open Spaces between the calyces, is formed by 

 the fusionof parts of four adjacent sheaths, producing the 

 closed tubes seen in cross section (figs. 79—81), and which would, ac- 

 cordingly, consist of parts of the terminal membranes of seven retinu- 

 lae, belonging to four dififerent ommatidia. This fact will become more 

 intelligible when we consider the outer ends of the membranes and their 

 arrangement around the ommatidia. 



Just beneath the inner edge of the pigmented band (fig. 69 pg.c), 

 the Square spaces bave attained their greatest size, and it is bere that 

 the circular membrane is most clearl}^ seen (fig. 79 7-(.s.) ; beyond this 

 point, towards the exterior, the spaces become completely filled with 

 the pigmented cells , and it is then a matter of the greatest difficulty 

 to distinguish the membranous sheath in cross section. In the following 

 sections, through that part of the ommatidium between the pigmented 

 band and the corneal facets, the sheath may once more be distinguished, 

 in the same form as before. Drawings of the sections bave not been 

 given, since the relation of the parts is sufficiently well represented in 

 those macerated specimens, in which the outer ends of the retinophorae 

 and various portions of the surrounding pigment cells, as well as the 

 sheath of the calyx, remain attached to the inner surface of the cornea 

 (figs. 76 and 77). In such preparations, ali combinations of the above 



