ORGANS OF THE BODY. 591 



roof of the fourth ventricle. Here we encounter ganglion cells of consider- 

 able size, varying from 0*045 to 0*067 in diameter, and of brownish tint, 

 disseminated through the white substance (substanfia ferruglnea superior 

 of Kuelliker). 



The nucleus dentatus of the cerebellum is of great interest on account 

 of its relationship to the specific nucleus of the same name in the olivary 

 bodies. In its plicated capsule of grey matter numerous ganglionic cor- 

 puscles of medium size (0*018-0*036 mm.) are to be met with disposed 

 in three layers, an external and an internal of fusiform cells, and middle 

 of multipolar elements. The bodies of these also are coloured, as a rule, 

 with pigment. Between them a maze of fibres is to be seen. 



Our knowledge of the course of the fibres in the cerebellum is at pre- 

 sent exceedingly imperfect. Nervous bundles of the crura cerebelli ad 

 med. oblong, are said to advance as far as the crumpled plate of grey 

 matter of the corpus dentatum, and to terminate in its ganglion cells pro- 

 visionally. Then, again, efferent fibrous bands are stated to make their 

 way out of the organ, at that point where the corrugated capsule of grey 

 substance is imperfect, passing into the crura cerebelli ad, corpora guadri- 

 gemina (Rutkowsky). The fasciculi of the- last-named crura, however, 

 have also been described as radiating from the corpus dentatum towards 

 the surface of the cerebellum; they are said to .connect this cortical layer 

 with the so-called corona radiata fibres (Stabkranzfaserung) of the hemi- 

 spheres of the cerebrum ( 299). But the bundles of fibres of the crura 

 cerebelli ad medullam ollongatarri are also described as entering the roof 

 nucleus of Stilling, and spreading from thence to the cortex. Some 

 speak also of a system of arching fibres here (similar to those we shall 

 have to consider presently in the hemispheres of the cerebrum), serving to 

 connect adjacent convolutions of the cortex with one another. 



In short, the whole matter is still very obscure. 



The cortical layer of the cerebellum has, to be sure, been recently made 

 the subject of very earnest study, and, to some extent, the older and newer 

 investigations have thrown a light on the subject. 



And first, as regards its coarser structure, it presents two layers, an 

 internal of a rusty brown colour, and an external grey. The first of these 

 is not so deep as the latter. 



It was at one time supposed (Gerlach, Hess, Rutkowsky) that the nerves 

 of the white substance undergo repeated division, radiating at the same 

 time in a brush-like manner, and forming eventually, with further sub- 

 division, a retiform plexus of fibres of only 0*0023 mm. in thickness, in 

 whose course the numerous nuclear structures of the rust -brown layer are 

 interpolated (Gerlach). This view, however, has not since been confirmed. 



The rmt-brown stratum, which has a thickness of from 1 to 0*5 mm. 

 (least at the bottom of the sulci), is by no means sharply defined against 

 the white matter beneath it. In it we find densely aggregated those 

 nuclear structures already mentioned, which also occur in the white 

 stratum (the "granules" of Gerlach). These are of rounded form and 

 diameter of about 0*0067 mm. on an average ; each presents also one or 

 two nucleoli (fig. 561, below). Whether we have before us cells or nuclei 

 here is a point very difficult of decision. One fact, however, in regard to 

 them can hardly escape even the superficial observer, namely, that they 

 possess considerable resemblance to certain elements of the retina, i.e., those 

 of the granular layers of the latter. 



Many of these elements present very fine filiform processes, of which 



