378 SPECIAL HISTOLOGY. 



way that the filaments of the " rods" (fig. 303 l , rr) are con- 

 nected with the " granules" of the outer granular layer, and 

 those of the "cones" (fig. 303 \ r) with the "granules" 

 of the inner layer; in fact, each "cone" or "rod" is 

 in connexion with a granule, the latter also perhaps with 

 several. The filaments which run inwards from both kinds of 

 "grannies," which may be pretty readily traced in vertical 

 sections, are continued in a straight line, or slightly curved, 

 through the layer of grey nerve-substance without any con- 

 nexion with its elements, and enter the expansion of the optic 

 nerve, where, as especially in the posterior part of the eye, in 

 which situation the layer of nerve-fibres is thick, it is easy to 

 perceive that they run in the narrow interstices between the 

 nervous bundles, in a fascicular manner towards the membrana 

 limitans. I have taken much pains in the investigation of 

 their relations at the surface of the retina, and have arrived at 

 the following results. If the inner surface of the retina be 

 examined under a strong magnifying power, in its posterior 

 half, where the fibrous bundles of the optic nerve are still very 

 distinct, a peculiar marking will be observed between them, 

 consisting of series of points, of minute stelliform figures, or of 

 little streaks, which often (also in animals, as the Ox for 

 instance) regularly converge towards each other from the 

 bundles of fibres like the rays of a feather. If these structures 

 are traced in vertical sections, it is easily seen that they are 

 nothing but the extremities of the radiating fibres dipping 

 down between the nervous bundles, and presenting a some- 

 what peculiar aspect. For, whilst in the deeper part of the 

 retina they are simple pale fibres, of at most O'OOOS"' in size, 

 they are here so modified that some of them simply expand, 

 and terminate in a triangular pale corpuscle, 0*0015 — 0-003'" 

 in length and breadth (fig. 303 l « /), from the internal angles 

 of which, one or two horizontal fibres are again given off; 

 whilst the others, without expanding, end in a complete bundle 

 of 5 — 9 or more fine fibres (fig. 303 1< n), which also turn to 

 the sides and continue in the plane of the nervous expansion. 

 What further becomes of these latter, innermost processes of the 

 radiating fibre-system, I have not yet been fortunate enough 

 to observe, however zealously I have investigated the matter, 

 and regret that the decision of this very important point in the 



