10 Journal of Comparative Neurology. 



responding layer of the higher vertebrates from stained prep- 

 arations only. For such a homologization the proof was yet to 

 be brought that this layer also contained similar elements and 

 that these elements exhibited the same behavior as in the higher 

 vertebrates. 



I have now succeeded in my preparations in furnishing this 

 proof, viz. : that there are both ' ' small granule cells ' ' {granule 

 cells sensu strictiori) and ^' large granule cells" {so-called Golgi 

 cells of the cerebelluni) in the granular layer of the Selachii. 



Regarding the small granule cells (Plate II, figs. 9, lO and 

 1 1) the cell body and protoplasmic processes behave through- 

 out like those of the higher vertebrates. The relatively small, 

 mostly round or polygonal body sends in all directions a limited 

 number (usually 3 or 4) of delicate protoplasmic processes of 

 which the majority terminate in the way familiar in the other 

 vertebrates, by means of a claw or brush-like structure. 



The extremely thin neuraxon arises in the majority of 

 cases from a protoplasmic process and then winds zig-zag be- 

 tween the other granules, usually proceeding directly to the 

 molecular layer. When it reaches the molecular layer, the axis- 

 cylinder divides in a T i^^ tJi^ manner typical for all vertebrates. 

 The two branches proceed in opposite directions parallel to the 

 surface of the cerebellum and the long axis of its folds and finally 

 (very probably a considerable distance from the point of bifur- 

 cation) break up into their terminal arborizations (Plate II, fig. 

 12). I have unfortunately only once succeeded (notwithstand- 

 ing a most careful examination of my sections), in following one 

 individual axis-cyhnder of a granule cell without interruption from 

 its origin to its point of division in the molecular layer. But I 

 have repeatedly come upon the pieces of these T-shaped bifur- 

 cations, as shown in figure 12 (Plate II). There can scarcely 

 be any doubt as to the connection of these fragments with the 

 axis-cylinders of the granule cells and I have no hesitation in 

 declaring, in spite of any inadequacy of observation, that the 

 behavior of the axis-cylinder of the small gfanule cells in the cere- 

 bellum of the Selachii is identical tJirougliout with that in the other 

 vertebrates. I might mention here a certain peculiarity found 



