484 'Journal of Comparative Neurology and Psychology. 



1. The intermediate sized cells agree with small cells in two 

 important characters, the cell-body stains deeply with eosin or 

 erythrosin and the cell-outline is irregular. 



2. Since the arrangement of the stainable substance is rather 

 unstable its difference has significance only when other characters 

 also differ from any other given group under consideration. If 

 the recently proposed hypothesis by Scott ('05) that the stain- 

 able substance is identical with the zymogen granules of the pan- 

 creas turns out to be true, then the size and form of the granules 

 as well as their distribution may vary considerably according to 

 the functional condition of the cell. 



Consequently we have in the spinal ganglion two forms of cells; 

 one which stains deeply and the cell-outline of which is irregular. 

 Such cells are usually small in size. The other the cell-body of 

 which stains lightly the cell-outline being regular. Such cells 

 range from small to large in size. It must be remembered how- 

 ever that the entire cell-population when they are grouped accord- 

 ing to their sizes grade from smallest to largest without showing 

 any interruption. This means of course that there is no definite 

 demarcation line to divide the large from the small cells or vice 

 versa. As a matter of fact, the cells which stain lightly and which 

 also exhibit regular outlines are by no means constant in size. 

 This is also true for the group of the cells which stain deeply and 

 which exhibit an irregular outline, although they are more uni- 

 formly small. For this reason, the size of the cell-body is not a 

 proper criterion by which to classify them. The writer has adopted 

 the staining reaction of the cell for classification because, as has 

 been mentioned above, the ganglion cells fall readily into one of 

 the two classes: that is {a) those with a deeply stained cell-body 

 with irregular cell-outline and {h) lightly stained cell-body with 

 regular cell-outline. 



Although we should expect to find intermediate forms, which 

 must always be present, nevertheless grouping by this method is 

 more definite and practical than by the size and is much simpler 

 than by those proposed by other investigators. In addition, these 

 different histological characters are undoubtedly associated with 

 different physiological states (Hatai '01). 



I have chosen from Nissl's nomenclature two terms by which 

 to designate the two groups of the ganglion cells with the idea 

 that they may aid description. 



