32 H. W. REllNARD. 



qnently much thinner than that of the fully formed rod ; 

 several measurements gave from 4 to 5 /i for the outer limbs 

 of Schwalbe's rods, and 6 to 7 ^ for those of normal rods. 

 On the other hand, they may be found of the same thickness. 

 I note in some microphotographs which I have had made 

 of the rods and cones of the Frog, in one case a Schwalbe's 

 rod is about two thirds as thick, while in another there occurs 

 one nearly if not quite as thick as the fully developed rods 

 near it. 



We have, then, in the diagram a series of structures con- 

 nected together by a scale of measurements. What evidence 

 is there that this series is really developmental ? We have 

 already found strong reasons for believing that the three-cone 

 stages are mere phases in the development of new rods ; the 

 retina, therefore, should show us stages between the last cone 

 with its long terminal vesicle and the fully formed rod. The 

 only variations on the normal rod which we can find are 

 those with the long inner limbs (Schwalbe's rods). There 

 can, I think, be no reasonable doubt that these are the inter- 

 mediate stages we require. The outer limb of the (pre- 

 sumably) youngest rod stage is of the same average length 

 as the terminal vesicle of the oldest cone stage, viz. 20 ^, and 

 further, it has apparently but continued the line of develop- 

 ment by becoming filled up and distended. On examining 

 the change of shape which has taken place, the most striking 

 change between this rod {i\) and cone (c^) is that the matter 

 which was formei'ly in the basal sections of the cone has 

 moved outward to form a swelling in contact with, but in 

 some way separated from, the terminal vesicle or outer limb, 

 although it may be important to note that the separation 

 between inner and outer limb does not seem to be in this 

 stage nearly so sharp as it is in the later normal rod. The 

 next step (iv) is a further filling up of the outer limb and its 

 extension through the length of 10 /z; at the same time the 

 inner limb, as it were, is drawn back and shortened to 20 ^u. 

 The last step is the extension of the outer limb to its full 

 length, about 40 ju, and the corresponding shortening of the 



