132 journal of Comparative Neurology and Psychology. 



ganglion of the white rat. The small cells constitute about two- 

 thirds of all the cells in the ganglion. They correspond to those 

 described by Dogiel, and according to him possess non-meduUated 

 fibers. A full consideration of the small cells will be found in a 

 previously published paper (Ransou '08). 



2. The Large Cells. The large clear cells are characterized by 

 their clear protoplasm. The size, shape, and arrangement of the 

 chromatic granules presents such an infinite variety that any attempt 

 at classification based on these alone would be largely artificial. We 

 know, however, that among the large cells are some — called by Dogiel 

 cells of Type II, or his new types III, IV, VIII and XI, — which 

 do not send any axon into the peripheral nerve; and, since these 

 would not show axonal reaction after division of the nerve, it 

 should be possible to identify them by their normal appearance 

 in a ganglion in which all the other large cells show chroma- 

 tolysis. Suggestive observations have been made in this connec- 

 tion by Cox, and Warrington and Griffith. The two latter in- 

 vestigators found that after section of the nerve at a point just distal 

 to the spinal ganglion all the cells in the ganglion with the exception 

 of the coarsely granular medium sized cells and the very smallest 

 cells in the ganglion showed chromatolysis. jSTo lesion of the per- 

 ipheral nerve would produce any alteration in these coarsely granular 

 cells. In the cat these cells present a rather constant diameter of 

 from 35 to 50 microns. Similar observations were made by Cox upon 

 the rabbit, but the non-reactive cells which he found differed from 

 those of Warrington and Griffith in the concentric arrangement of 

 the elongated tigroid masses and in the excentric position of the 

 nucleus. In short, they were the cells which constitute his Type II. 

 Cox regards his Type II as identical with Dogiel's Type II, since 

 the component cells are few in number and do not give evidence 

 of any injury 'to their axon after section of the peripheral nerve 

 close to the ganglion. W^arrington and Griffith suggest the same 

 possibility for their coarsely gi'anular cells. 



It seems probable to the writer that these two types are really 

 the same and that in some animals the tendency to concentric arrange- 

 ment is more marked than in others. This is borne out bv the fact 



