STRUCTURE OF THE RETINA. C09 



end of the cell, which is frequently imbedded in the granular substance 

 of the succeeding layer, one or more much thicker processes extend out- 

 wards for a variable distance into that stratum, and after branching 

 dichotomously once or twice become lost in its substance. The number 

 of nerve cells and consequently the thickness of the ganglionic layer in 

 the different regions of the retina varies exceedingly. Over the greater- 

 part of the retina they form a single stratum, but in the neighbourhood 

 of the yellow spot they are placed two or three deep. At the spot itself 

 (fig. 426, 2) they are very thickly set (from eight to ten deep) ; the cells 

 are also much smaller here, and are bipolar. Towards the ora serrata, 

 on the other hand, there is but a single stratum, and that frequently 

 incomplete. 



o. Inner molecular layer. — Next in order to the ganglionic layer 

 comes a comparatively thick stratum of a granular-looking substance, 

 which in the perfectly fresh condition presents, under high powers of the 

 microscope, the appearance of a pale, homogeneous matrix with numerous 

 minute clear globules or granules, imbedded in it. The fibres of jMiiller 

 pass through the substance composing this layer without being directly 

 connected with it ; the off"sets of the ganglion-cells can also be traced 

 into it for a greater or less distance ; and, finally, the fine, varicose 

 central processes of the nuclear bodies of the layer next to be 

 described, can be followed a short distance into the layer in question 

 (fig. 422), passing in the direction of the ganglionic and nervous layers. 

 Flattened cells, similar to those noticed in the nerve-fibre and ganglionic 

 layers, are also said to occur in the inner molecular layer, especially on 

 its surfaces (Golgi and Manfredi). 



The exact nature of the substance composing this layer is still a matter of 

 doubt. By Max Schultze it was described as a reticulating mass with numerous 

 fine intercommunicating meshes, supported in which was supposed to be a fine 

 network formed by ramifications of the ganglion-cell processes, which become 

 lost to view in this layer. Schultze further described it as being of the same 

 nature as and intimately connected with the system of Miillerian fibres. Schwalbe, 

 however, has shown that these pass through the inner molecular layer sometimes 

 with a smooth contour and always without giving off lateral offsets to it ; and, 

 moreover, that they are of a different chemical natm-e from the substance of the 

 molecular layer. 



4. Inner nuclear layer. — Tliis is mainly composed of several strata 

 of characteristic transparent nucleus-like bodies, which are frequently 

 known collectively as the " inner granules," but are nevertheless of three 

 or four distinct kinds. Those of one kind, few in number, are connected 

 with the fibres of Miiller as these pass through the layer, and will most 

 conveniently be described with those fibres. Those of the second kind 

 (fig. 422, 4), by far the most numerous, are prolonged at either end into 

 a delicate fibre, and contain each a clear, round or oval nucleus, with 

 a distinct nucleolus. These last (the nucleus and nucleolus) resemble 

 very closely the analogous structures found in ganglion-cells, and indeed 

 the bodies in question are commonly regarded as bipolar nerve-cells, 

 but with the protoplasm of the cell very small in amount, devoid of 

 granules, and principally collected at either end of the cell. 



Of the cell processes or fibres which proceed from these ends, the inner 

 one, or that extending into the inner molecular layer towards the gan- 

 glionic and nerve-fibre layers, is finer than the other, is always unbranched, 

 and commonly exhibits minute varicosities similar to those on the ultimate 



