144 'Journal of Comparative Neurology and Psychology. 



Table II shows that the decrease in the number of cells is due 

 chiefly to the loss of cells under 20 microns in diameter. There is 

 also no doubt a slight loss of cells of the larger size, but this is 

 compensated by the fact that the remaining large cells lie closer 

 together, so that a single section shows about the same number as 

 there are in a similar section of a normal ganglion. But the point 

 to be emphasized is that the cell destruction affects chiefly the 

 smaller cells. That this fact escaped the attention of all the previous 

 investigators is probably due to the slow repair of the large cells 

 in their specimens, which permitted a considerable amount of atrophy 

 to occur before the cells again became of normal appearance. 



TABLE II. 

 Showing the proportion of large and small cells in the normal and "operated" 

 second cervical ganglion of the white rat 20 days after section of the ramus pos- 

 terior of the corresponding nerve. 



It is necessary in making such a differential count to make it 

 at just the proper period after the operation. The careful measure- 

 ments of Lugaro and Fleming have shown that there is during the 

 phase of active chromatolysis a considerable swelling of the cells, 

 and that repair is accompanied by marked atrophy which continues 

 to progress for some time and leaves the cells very nmch reduced 

 in size. It is clear that cither swelling or atrophy would tend to 

 obscure the relations brought out in Table II. In making the 

 counts it was necessary to select a period when the repair of the 

 cells was just complete and tlieir a?dematous condition had subsided, 

 but before they had begun to show marked atrophy. For this purpose 

 the preparations from the rat killed 20 days after the operation 



