78 Journal of Comparative Neurology. 



which impress the author as axis-cylinder processes are printed 

 in red. 



These axis-cylinder processes may be divided into two dis- 

 tinct types. Either they give off isolated lateral twigs merely, 

 without thereby losing their individuality, until finally they be- 

 come axis-cylinders of medullated nerve fibres (motor type); or 

 the axis-cylinder process breaks up again and again into the 

 finest fibrils without passing into the axis-cylinder of a nerve 

 fibre (sensory type). If, however, in the case of a cell of the 

 sensory type the axis-cylinder process does not pass directly into 

 a nerve fibre, then by that very fact it is deprived of the chief 

 characteristic which makes it an axis cylinder process. The re- 

 lation of the axis-cylinder process is in this case exceedingly 

 similar to that of the so-called protoplasmic- processes. 



That the protoplasmic processes (Dendriten, His) break up 

 by repeated branching into very fine twigs, may easily be shown 

 by a variety of methods, especially by impregnation wuh silver 

 or mercury ; yet the views with reference to their significance 

 are exceedingly diverse. For a long time no one had ventured 

 to doubt their nervous significance. And indeed most of the 

 recent investigators also retain this view, (Ramon y Cajal, Anat. 

 Anzieger, 1890; Fritsch, Akad. d. Wiss., Berlin, 1891) ; while 

 Nansen, and especially Golgi and his school (Sala, Zeitschr f. 

 wiss. Zool., 52 Bd.) deny their nervous significance, KoUiker 

 (Zeits. f. wiss. Zool., 51 Bd.) and Waldeyer (D. med. woclien- 

 schr., 1 891) consider that the matter is still undecided. 



The protoplasmic processes pass out directly from the pro 

 toplasm of the cell body in such a way that weighty reasons had 

 to be adduced in order to raise any doubt as to their nervous 

 nature. These reasons were as follows: protoplasmic processes 

 penetrate into regions which are especially poor in nerve fibres ; 

 and again they attach themselves by means of their ultimate 

 ramifications to neuroglia cells and to the walls of blood-vessels, 

 so that their function might be conceived as chiefly nutritive. 

 Attention may be called to the fact, moreover, that even the 

 facts upon which the views of the non-nerv(jus nature of the 

 ])rotoplasmic processes are based cannot by any means be sus- 

 tained by most other investigators. 



A further question is whether the branchlets terminating the 



