STUDIES IN THE RETINA. 305 



Eemak, Max Scliultze, and others, tliey have been accepted for 

 the last forty years without question. Tliey are represented 

 as multipolar bodies with large vesicular nuclei, each with one 

 proximal process, called the axis-cylinder process, con- 

 tinuous with one of the fibres of the nerve-layer, and one or 

 more distal processes, which plunge into the inner reticular 

 laj'er, where they ramify and are lost. These processes are 

 all apparently prolongations of the cytoplasm of the cell- 

 body, the conspicuous nucleus, so far as one can see, taking 

 no part in their production. Our knowledge of the finer 

 structure of this cytoplasm, through which we are left to 

 conjecture that the stimulus must pass from the retina into 

 the optic nerve, may be summed up as follows : — It consists of 

 the fibrillar substance of Flemming with the interfibrillar 

 liyaloplasma, further complicated by (1) the presence of the 

 staining clumps ("I^issl's Schollen"), often arranged in 

 irregular concentric rings round the nucleus ; and (2) the 

 occurrence of one or more refractive staining globules, 

 " centrosomes," each in the centre of a clear field. Add to 

 these the fact that the nucleus contains a nucleolus suspended 

 on an intra-nuclear reticulum, and we have all the accepted 

 constituents of the ganglionic cells of the retina. 



It must be admitted that this familiar diagram seemed to 

 meet some of the needs of the problem as to the passage of 

 the nerve-stimuli, for it at least shows the fibres of the optic 

 nerve bent down at right angles into the retina, i.e. towards 

 the layer of rods. The one great difficulty was, however, to 

 find how the nerves bridged over the interval between the 

 processes of these ganglionic cells, assuming at least one of 

 these processes to be nervous, and the rods. This has been, 

 if I may say so, the burning problem of the retina, viz. : 

 which of the two or three distal processes of the ganglionic 

 cells carried the stimulus? Why were they all, sooner or 

 later, lost to view ? Why was not at least one of them trace- 

 able light through to the rod-layer ? No solution of this 

 problem seemed to be forthcoming until the elaboration of 

 the metal impregnation methods, especially in the hands of 



VOL. 47, PART 3. — NEW SERIES. U 



