140 'journal of Cojjiparative Neurology and Psychology. 



degenerate nuclei or none at all, and whieb are surrounded bj 

 proliferating connective tissue cells of tbe capsule. 



It is quite remarkable, considering bow large a number of cells 

 disappear from tbe ganglion (52 per cent according to counts on 

 tbe second cervical ganglion of tbe wbite rat), tbat tbe strictly de- 

 generative cbanges are not more obvious in sections of tbe reacting 

 ganglia. One reason wby my preparations bave not given altogetber 

 satisfactory data on tbis point is tbat most of tbem were not counter- 

 stained. Wbile tbis gives tbe best material for study of cbromatolysis, 

 it does not bring out tbe protoplasmic cbanges wbicb indicate tbe 

 disintegration of tbe cell. In tbe preparations of ganglia removed 

 seven and eigbt days after tbe operation, wbicb were counterstained, 

 a small percentage of tbe cells can be seen undergoing deiinite 

 disintegrative cbanges. Two well-marked types of disintegration 

 can be recognized. By far tbe most frequent and im]iortant is tbat 

 in wbicb tbe cell becomes penetrated by proliferating fibroblasts 

 from tbe capsule wbicb form nests of vesicular nuclei suggestive 

 of tbe intracapsular proliferation in rabies. Tbe early stages in 

 sucb a process are sbown in Fig. 5. Tbe small dark capsular nuclei 

 bave given place to large vesicular ones. Tbese are seen grouped 

 about tbe cell. Five bave penetrated into tbe cell. Tbere is no trace 

 of tbe ISTissl-bodies, tbe nucleus presents an indefinite crenated border 

 and contains tbe outlines of a barely recognizable nucleolus. A 

 furtber increase in tbe number of tbe fibroblasts gives rise to tbe 

 formation of nests of sucb cells in wbicb one can just recognize tbe 

 outlines of tbe original spinal ganglion cell. Sucb nests are relatively 

 frequent after tbe seventb day. By tbe transformation of tbese into 

 adult connective tissue witb tbe consequent sbrinkage, tbe spaces 

 formerly occupied by tbe ganglion cells are obliterated so tbat tbe 

 ganglion cells tbat survive are brougbt to lie almost as close togetber 

 •IS in tbe normal ganglion. Tbe amount of intervening connective 

 tissue is surprisingly little after two montbs. 



Tbe otber form of degeneration, vacuolation, is more rare but 

 seems to play a certain part in tbe cell destruction. It may go on 

 to sucb an extent tbat tbe cell is converted into a single large vacuole, 

 surrounded by a tbin ring of protoplasm (see Fig. 6). It is true 



