STRUCTURE OF THE RETINA. 611 



retina, a similar granular appearance, with a few scattered nuclei in its sub- 

 stance, and in which, in the human retina and in that of most mammals, 

 it is difficult to make out any definite structure. In the horse's retina, 

 however, as Eivolta and others have shown, it is possible, in properly 

 prepared and isolated portions of the layer, to ascertain the existence 

 within it of flattened, irregular or stellate, finely branched, delicately 

 granular cells to which the nuclei in question belong (fig. 421). The 

 cell-offsets, which are excessively fine, form by repeated branching and 

 union a close network or feltwork throughout the layer, the substance 

 of Avhich is in this way formed by them. The nuclei of the cells are 

 clear and distinct, each with a comparatively large nucleolus ; the fine 

 fibres of the network are stated by Golgi and Manfredi to exliibit vari- 

 cosities like nerve-fibrils, but according to Schwalbe the varicose fibrils 

 in all probability do not actually form a part of the network but are 

 intercalated m its meshes. It is, therefore, uncertain whether these 

 cells are to be regarded as of a nervous nature themselves or merely as 

 supporting structures for the true nervous elements. There is little 

 doubt that a similar structure exists in mammalia generally. 



6. Outer nuclear layer.— This (figs.419,422, 6) resembles very closely 

 at first sight the inner nuclear layer, appearing, like that, to consist of 

 several strata of clear, oval or elliptical, nuclear corpuscles (outer gran- 

 ules), from the ends of which delicate fibres are prolonged. They differ, 

 however, essentially from the mner granules, and may be readily dis- 

 tinguished from them. These outer granules are of two kinds, which 

 present well-marked differences and are known respectively as the rod- 

 granules and cone-granules, according as they are connected with the 

 rods or with the cones of the next retinal layer. Those which are con- 

 nected with the rods are, in most parts of the retina, by far the more 

 numerous, and form the mam thickness of the outer nuclear layer. 

 They may be regarded as enlargements or swellings m the course of 

 delicate fibres (rod-fibres), which extend from the inner ends of the rods, 

 at the membrana liraitans externa through the thickness of this layer 

 to the outer molecular layer. The enlargements, of which there is but 

 one to a fibre, situate at any part of its course, are each occupied by 

 an elliptical nucleus which is devoid of a nucleolus, but which, in the- 

 fresh condition, exhibits a remarkable cross-striped appearance (Heule),, 

 the strongly refracting substance which mainly composes it being inter- 

 rupted by bands or disks of a clearer less refracting material, usually 

 two in number, one on each side of the middle line (fig. 422), but occa- 

 sionally single and median (see the left-hand one in fig. 422). The rod- 

 fibres are of excessive fineness, and exhibit frequently minute varicosities 

 m their course : each is directly continuous at the outer end with one 

 of the rods, but at the inner end appears usually to terminate in a 

 somewhat larger varicosity, from the margin of which exquisitely fine 

 fibrils may, in some animals, be traced extending into the substance of 

 the outer molecular layer. 



Those outer granules which are connected with the cones are, 

 in most parts of the retina, much fewer in number than the rod- 

 granules, from which they are distinguished by their shape, which is 

 somewhat pyriform, by the absence of transverse striation, and by their 

 position— for they occupy the part of the outer nuclear layer nearest 

 the membrana limitans externa, and the larger end of each is thus in 

 close proximity to the base of the corresponding cone, with which it is 



