ORIGIN OF THE TRABECULE OF THE OPTIC NERVE. 481 



stalk ; . . . this is a matter of some morphological, and cer- 

 tainly of pathological, importance." 



All my specimens undoubtedly confirm the observations of 

 the authorities quoted by Robinson (7), and his own state- 

 ments, which are based upon observations made on human 

 embryos, cats, ferrets, sheep, rabbits, rats, and mice, viz. that 

 the ingrowing nerve-fibres lie within the membrana limitans 

 externa, throughout the whole of their course in the optic 

 stalk, and that they enter the stalk along the ventral wall; 

 though Froriep (3 a) has lately stated that, in his specimens 

 of rabbit embryos, the earliest bundles of nerve-fibres grow 

 in higher up on each side of the ventral wall, and that the 

 nuclei that lie above the ingrowing nerve-fibres are pushed 

 up towards the lumen of the stalk, whilst those that lie below 

 are pushed still further down, as the number of nerve-fibres 

 increases. 



In tadpoles of 6 mm. in length I have invariably found the 

 earliest bundles of nerve-fibres, as they issue from the optic 

 cup, occupying a central position just within the membrana 

 limitans externa of the ventral wall of the stalk, and^ as they 

 approach the brain, getting more and more towards the pos- 

 terior side of it, though in 8*5 mm. tadpoles they seem first 

 of all to travel a little anteriorly for a very short distance, 

 just after leaving the optic cup. These observations are in 

 agreement with Froriep's (3 a), figs. 237 — 239, taken from 

 tadpoles. 



II. Cellular Segmentation. 



Robinson (7) has referred to the difficulty of obtaining in- 

 dications of definite cell-territories in the early stages of the 

 embryonic optic stalk of the rat. 



In tadpoles of 4'5 mm. in length cell limits are certainly 

 recognisable (fig. 1), but in those of 6 mm. in length they 

 can rarely be distinguished from the pigmented fibrils of 

 the protoplasm that encircles the granules of food-yolk, or 



