26 H. M. BERNAKD. 



invariably has a large basal vacuole, and it is about the same 

 length as section 1 in type Co- The small refractive globule 

 we now find on the level of the second division of Cg. The 

 third section is generally very distinct, being thick and often 

 deeply stained, but with very delicate walls, usually ill-defined, 

 doubtless owing to the action of the reagents. From this 

 point this second type of cone may again be divided into 

 cones C3 and c^. In the former the delicate tip alone is 

 swollen into a vesicle ; in the latter the whole of the remain- 

 ing length swells into a long, terminal, often shrivelled and 

 generally ruptured vesicle (figs. 2, i, h, and 3, c, d, e, h, i). 



Before discussing these varying forms it will be advisable 

 to examine more closely the measurements which lie at the 

 basis of the scale shown in the figure. Measurements of the 

 cone sections reveal a degree of regularity which requires 

 explanation. In speaking of the measurements of rods and 

 cones we must be careful to note that they can only be relative. 

 The rods are not stiff cuticular structures, but long, cylin- 

 drical sacs filled with certain substances, apparently of 

 varying consistency, which allow the rods to be squeezed up 

 or else pulled out or bent about at all angles. 



The complete forms of the cones are best seen when the 

 rods are drawn out and separated by short intervals. The 

 latter are then about 50 ^u in length, measured from the memb. 

 limitans ext.^ Of this length, the inner section or limb of the 

 normal rod measured on an average 10 ju (fig. 2, e and j), 

 with variations ranging between 8/z and 12 ^u. This leaves 

 the outer limb of the rod about 40 ju in length. 



We have, then, from the membrana limitans externa to the 

 pigment layer about 50 fi, and across this the cones stretch. 

 I have shown in the diagram (fig. 4) the results of many 

 measurements. Passing over c^ for the present, we find that 

 section 1 in Co, which is commonly called the inner limb of 



' This is also the length given by Hoffman (Bronn's ' Thierreicli,' vi, part 

 2, p. 283). I have found rods as much as 60 y, long in well-fixed frogs' 

 retinas. We shall see later on some reasons for believing that the size of the 

 rods varies up to some maximum with age. 



