96 JOIHNAI. OF CoMl'AliA I 1\ K N El" UOl.CXi Y . 



case of the spinal cord it is impossible to assert that such a 

 relation has been demonstrated. For example, we know that 

 certain parts of the cord are the peculiar seat of sensory 

 activity, while others form centers for motor reactions, yet it 

 would be ver}' difficult to point out exact anatomical distinc- 

 tions characteristic of these two areas. The following is a 

 condensed resume of Golgi's results:(') 



1. All ganglion cells of the spinal cord (those of the dor- 

 sal cornu and Clarke's column not excepted) are provided 

 with a special process connecting with a nerve and differing 

 in physico-chemical peculiarities from all other processes. 

 This process alone is a safe criterion for identifying a cell as 

 nervous. Upon this basis alone I identified numerous 

 ganglion cells in the substantia genlatinosa of Rolando. All 

 the cells of the spinal cord are, therefore, from the stand- 

 point of their specific function unipolar, the single process 

 referred to being the nervous process. 



2. The so-called protoplasmic processes of the cells are 

 neither directly nor indirectly the source of nerve fibres, but 

 they are closely associated with connective tissue cells and 

 blood-vessels. They apparently constitute the avenues by 

 which nourishment reaches these cells from blood vessels 

 and connective elements. 



3. A comparison of the cells from various parts of the 

 grey matter of the spinal cord reveals certain differences in 

 the form, size, and ramifications of the protoplasmic process, 

 yet the only important distinctions are such as relate to the 

 nervous processes. 



4. Upon this basis two sorts of cells may be recognized 

 'in the spinal cord, thus: {a) ganglion cells, in each of which 

 the nervous process divides into minute fibrils, so that it 

 loses its individuality in the formation of a diffuse nervous 

 reticulum: (/^) ganglion cells which give rise to an axis 

 cylinder, though there may be small lateral processes. 



GoLt,]. I'cbcrdeii fcineieu U.iu Ues Ruckcnniai ks; Anat. Aiiz., Xo, 15, 1S90. 



