240 bulletin: museum of comparative zoology. 



To grant that the network which is described as surrounding the 

 sense cells is nervous in character would not necessitate the abandon- 

 ment of the neurone theory, though such a structure would of course 

 be in perfect accord with Apathy's conception. Kolmer admits, 

 however, that his conclusions are based upon rare observations, 

 and he assumes that in the great number of cells which do not show 

 pericellular networks of neurofibrillae their non-appearance is due to 

 imperfect impregnation. He also assumes that his inability to find 

 evidence of such structures in fishes is due to "some specific physico- 

 chemical characteristic of fish protoplasm which renders these verte- 

 brates unfavorable objects for the study of peripheral neurofibrillae." 

 I know of no evidence for the truth of such an assumption. I have 

 already pointed out good reasons why we should expect fishes to 

 furnish the best material in the vertebrate series for the study of this 

 problem. 



Bielschowsky und Briihl, like Kolmer, base their belief in the 

 existence of pericellular networks • upon rare observations, assuming 

 that the more numerous and more simple structures are the result 

 of imperfect impregnation, and that the rarer pericellular networks 

 represent a universal terminal structure. This appears to me to be a 

 dangerous assumption, and I believe that the safer course would be to 

 look with suspicion upon structures of such rare occurrence, regarding 

 them as probably artefacts. It would perhaps appear presumptuous 

 for me to venture a criticism upon the handling of a method which 

 Bielschowsky himself worked out. Nevertheless, it should be recog- 

 nized that, superior as this method is to the older impregnation proc- 

 esses, it is easy, by overdoing the treatment with the silver oxide 

 solution, to cause almost any tissue to take on the color characteristic 

 of nervous matter in good Bielschowsky preparations. I have several 

 times obtained preparations in which the sense hairs, and the sense cells 

 as Avell, were entirely blackened, and have seen preparations from the 

 lateral line region in which muscle fibres were well impregnated, the 

 cross striations being clearly differentiated. I have no doubt that by 

 excessive treatment with silver oxide solution it would be entirely 

 possible to cause non-nervous intercellular material to take on the 

 appearance of nervous substance. What constitutes successful im- 

 pregnation is a matter which it is necessary for each investigator to 

 determine for himself, and it is this last resort to the personal judgment 

 of the investigator which is responsible for many of the differences 

 of opinion in this field. The controversy between the neuronists and 

 the fibrillists has resolved itself largely into the question whether certain 



