242 THE RETINA, BY MAX SCHULTZE. 



of granules, which are connected with the external granulated 

 layer by very short fibres. The rod granules and cone granules, 

 and the fibres belonging to them, are here very similar, and 

 break up at the outer granulated layer in the same manner 

 into fine fibres. In Fishes, where the difference in the thick- 

 ness of the more elongated rod and cone fibres is again very 

 apparent, the conical enlargement of the rod fibres is also very 

 similar to that of the cone fibres. In short, everything favours 

 the view that no other essential difference exists between the 

 rod and cone fibres than in their thickness, and that the rod 

 fibre probably always breaks up into a number of fibrils at the 

 external granulated layer, and thus also, like the cone fibre, is 

 in reality a fasciculus of fibrils. In Mammals and in Man the 

 rod fibres are very delicate, but are always several times thicker 

 than the finest fibrils of the optic nerve. This is especially 

 true of the external peripheric part which is turned towards 

 the limitans externa, and which constantly and considerably 

 exceeds the thickness of the internal central portion that extends 

 from the external granule to the external granulated layer. 



Every rod and cone fibre is connected with one of the so- 

 called external granules ; that is to say, each of these fibres 

 has at some point in its course an enlargement in which a 

 nucleus is imbedded, and this is named a rod or cone granule 

 (see fig. 347). If the fibres in question are nerve fibres, the 

 granules must be comparable to small bipolar nerve or ganglion 

 cells. The quantity of the cell substance, however, is very 

 small, yet still somewhat greater in the constantly larger 

 granules of the cones than in those of the rods. The nucleus 

 fills the interior of the granule almost completely, is of hyaline 

 appearance, and contains a bright nucleolus, which is larger in 

 a cone than in a rod granule. Except at the yellow spot, the 

 rod nuclei are much more numerous than those of the cones ; 

 they are superimposed in several tiers upon one another, and 

 are so closely arranged as to be in contact. The cone granules 

 lie just beneath the membrana limitans externa, except where, 

 as at the macula lutea, the cones are so closely placed that the 

 granules belonging to them are necessarily in several layers.* 



* Exceptionally in the more peripheric parts of the retina the interval 



