28 H. M. BERNARD. 



emerge from the pigment^ are the longer terminal vesicles in 

 cone C4. Their walls are mostly torn and crumpled, but they 

 may nevertheless be at times found complete (figs. 2, i, h, 

 and 3, h, e, h). The part most commonly seen is the promi- 

 nent end of the vesicle, where the stem of the cone suddenly 

 widens out (fig. 3, c, d). The former of these is on a cone of 

 type C2, which shows that variations occur in the order of 

 development.^ 



From these facts we gather that at the distal ends of the 

 cones there is a terminal vesicle, which increases in size up to 

 20^ in length. 



While this transformation, which appears to take place at 

 the distal ends of the cones, differs widely from that seen in 

 the basal ends, yet it is so far in harmony with it that we have, 

 proximally, an extrusion of matter into the cone, and, distally, 

 an apparent accumulation. 



The facts so far adduced indicate that the different types of 

 cones ai'e due to the successive extrusions of matter through 

 the memb. limitans ext., while the existence of nodes or 

 sections approximately of the same length suggests some 

 periodicity in this process. It is true that this periodicity is 

 not easy to work out in detail; the difficulty lies in the much 

 greater quantity of material in the basal than in the distal 

 sections. We may, perhaps, assume that fig. 2, a, or 4, c^, 

 represents the initial extrusion beyond the memb. limitans 

 ext. of a vesicle containing fluid. This is squeezed up between 

 the rods, and runs as a thin thi-ead, to swell out only in the 

 pigment layer. Into this, or added to this, new matter is forced 

 out by successive discharges, the nodes or breaking points 

 between the sections registering in the earlier, thinner stages 

 of the cone the successive discharges. The contents of the 



1 Here it is worth iiotiug that H. Miiller ('Z. w. Z.,' viii, 1857, p. 48), 

 who had already seen some slight prolongation of the tips of the cones in the 

 frog, found inhuman "cones " an empty prolongation reaching to the pigment 

 layer. The equality in length of rods and "cones " in the human retina has 

 recently been definitely affirmed by Borysiekiewitz (' Weitere Untersuchungen, 

 etc.,' Wien, 1894). 



