CHAP, in.] SIGHT. 65 



ganglionic cells with the optic fibres. On theoretical grounds we 

 may conclude that these three layers are functionally continuous, 

 that changes set going in the (inner limbs of the) rods and cones 

 sweep through the inner nuclear layer, and issue along the fibres of 

 the optic nerve as nervous impulses ; but, with at least our present 

 knowledge, we cannot demonstrate a structural continuity between 

 them. Two conspicuous breaks occur at the two molecular layers. 

 We can trace the rod fibres and the cone fibres to the outer 

 molecular layer, and there we lose them. We can trace the opftc" 

 fibres through or apart from the ganglionic cells to the inner 

 molecular layer, and there we lose them too. We can trace 

 the processes of the cells of the inner nuclear layer on the one 

 hand to the inner and on the other to the outer molecular layer, 

 and there these too are lost. These two molecular layers, which 

 repeat in the outlying part of the brain which we call the 

 retina some of the characteristic features of the brain itself, 

 are obviously of no little importance in the development of visual 

 impulses ; but for a proper understanding of their nature we must 

 await the results of further inquiry. At present all we can per- 

 haps say is that each layer consists of a network of fine nervous 

 fibrils imbedded in neuroglia, but that, as in corresponding cerebral 

 structures, we cannot accurately distinguish neuroglial from 

 nervous elements, much less trace out the exact disposition of the 

 latter. In the outer molecular layer among the tangle of fibrils, 

 nervous and neuroglial, we may distinguish small branched cells, 

 lying flatwise in the plane of the layer ; these are probably 

 neuroglial cells whose branched processes become neuroglial 

 fibrils. In the inner molecular layer such cells are absent or 

 at least inconspicuous ; the layer seems to consist on the one 

 hand of nervous fibrils derived from the branching processes of 

 nerve-cells and on the other hand of neuroglial fibrils, all im- 

 bedded in a peculiar ground-substance which stains deeply with 

 osmic acid, and indeed is of a nature in some respects allied to the 

 medulla of a nerve fibre. 



We have reason to think that the molecular changes which 

 light induces in the inner limbs of the rods and cones differ very 

 considerably in character from the molecular changes in the fibres 

 of the optic nerve which constitute a nervous impulse, and that the 

 transformation from the one set of changes to the other is effected 

 through some or other of the retinal structures which we have 

 described. But we cannot attribute definite functions to the 

 several elements ; and here, as in the case of the brain and spinal 

 cord, we may hesitate to assign too much to ' cellular elements. 

 We may, perhaps, in conformity with what we have urged 

 elsewhere ( 579), regard the cells of the ganglionic layer as 

 being largely concerned in nutritive labours, and may even apply 

 the same view to the nuclear layers ; if this be so, no small part 

 of the work of the retina in transforming the first crude effects 



