STUDIES IN THE RETINA. 321 



hand, and the rods in which we have every reason to believe 

 that the nerves terminate, on the other. 



As mentioned in the introductory remarks, all attempts to 

 bridge ovei" this intervening space by direct continuations of 

 the nerve-fibres have hitherto failed. After their supposed 

 nnion with the cytoplasm of the " ganglionic cells " they 

 were lost, appearing again, as Max Schultze suggested, in 

 the rod- and cone-fibres, which were, he thought, bundles of 

 primitive nerve-fibrils, such as those found composing the 

 axis-cylinders of meduUated nerves. From here he at one 

 time thought they ran outwards down the rods as striae. We 

 shall see lower down how near these suggestions came to the 

 truth. At the time, however, they were not acceptable, 

 because there still remained the gap between the ganglionic 

 cells and these supposed bundles of nerve-fibrillse, the rod- 

 fibres. The discovery of the protomitomic system pervading 

 the whole retina has put the problem on a new footing. The 

 gaps that now remain to be bridged, are those between the 

 protomitomic system and the nerves, on the one hand, and 

 the same protomitomic system and the rods, on the other. 



The method of union between the nuclear or protomitomic 

 system and the nerves was first discovered in the young 

 stickleback, and once seen, was soon found in other retinas 

 as well, and especially clearly in the preparations of the 

 human retina to which I have already referred (Part V, 

 p. 54). In the stickleback, a nerve-strand, was noticed cut 

 ti'ausversely (see fig. 6) showing the familiar characters, a 

 number of darkly stained points which represented the 

 individual primitive nerve-fibrils scattered on the walls of a 

 delicate framework, and easily seen to be filamentous by 

 focussing up and down. The delicate framework of the 

 strand in such sections can only be regarded as supporting 

 tissue conducting and isolating the fibrils.^ A '^ ganglionic" 



' Tills, I found, was the general structure of the nerve-strands of the 

 Araclinida (see "Comparative Mor|)hology of the Galeodida;," 'Trans. Linn. 

 Soc.,' vol. vi [1896J, p. o4'5). 1 have seen the retinal strands described as 

 medullated, but that is not the case in any retina I have yet examined. 

 VOL. 47, PART 3. — NEW SEKIES. X 



