THE CIRCULATION OF CEREBRO-SPINAL SUBSTANCE. 567 



nerves. But we have still to study an important question: i.e., 

 the identity of the intermediary fibrils of neuroglia important 

 in the sense that it has a certain bearing upon the concordance 

 between the older views of Gerlach and the modern observa- 

 tions and conclusions of Golgi. Indeed, if the entire cerebro- 

 spinal axis is made up (as far as true nervous elements go) of 

 these medullated glia-covered nerve-fibers and dendrites, we 

 may well conclude with Gerlach that nerve-cells are united by 

 an intricate net-work of extremely delicate nerve-fibrils. If, 

 on the other hand, the cell-body, its dendrites, and its axon 

 are alone medullated and glia- or neurilemma- covered, the 

 connection with the vascular system being established by non- 

 medullated fibrils, we are in accord with Golgi, who denies the 

 existence of any connection through nervous structures between 

 neurons. 



That Golgi's view prevails is suggested provided our own 

 view that fibrils are plasma-channels is accepted by the fol- 

 lowing lines by Professor Foster: "The larger part of the gray 

 matter consists, besides a neuroglia supporting the nervous 

 elements, of nerve-filaments running in various directions and 

 forming, not a plexus properly so called, but an interlacement 

 of extreme complexity." If the italicized words "nerve-fila- 

 ments" are converted, in accordance with our view, into neu- 

 ro glia- fibrils, the rest of the quotation will lead us to the solu- 

 tion of the question: "These filaments are, on the one hand, 

 the fine medullated fibers spoken of above as being recognized 

 with difficulty, and, on the other hand, non-medullated filaments 

 ranging from fairly wide and conspicuous naked axis-cylinders 

 down to fibrils of extreme tenuity, the latter arising apparently 

 either from the division of axis-cylinders and nerve-fibers pass- 

 ing into or out of the gray matter or from the continued 

 branching of the nerve-cells." 



The solution, it seems to us, lies in the fact that non- 

 medullated fibrils exist at all, and that these range from fairly 

 wide axis-cylinders down to fibrils of extreme tenuity, some of 

 which, at least, appear to originate from dendrites. Indeed, 

 this indicates that these non-medullated fibrils (of neuroglia, 

 as stated by Berkley) represent the continuation of the main, 

 or apical, dendrite (or dendrites, for there are often more than 



