LEWIS. — POLYCH^TE ANNELIDS. 237 



Eisig, who believes that such continuity exists, although he did not 

 actually demonstrate it, grants that the structures were originally nervous 

 in nature, although in his opinion they have since become degenerate. 



The ground upon which their nervous nature is denied by von Len- 

 hossek and Retzius — viz. their failure to react to the Golgi method — 

 seems very unimportant, especially when compared with the satisfactory 

 work and positive results obtained by Cerfontaine with methylen blue. 

 The capricious nature of the Golgi method is well known. It is granted 

 by all workers who have employed this method that there are many 

 nervous structures in any preparation which do not react to the silver 

 process. Indeed, one of the things which makes a Golgi preparation 

 especially valuable is the fact that only a small proportion of the nervous 

 elements are affected by this treatment. Considering the exceedingly 

 small number of Leydig's fibres in any preparation, and the limited 

 number of workers who have employed the Golgi method upon them, it 

 is not strange that no reaction has yet been obtained. 



The evidence, however, is not yet sufficient to allow one to draw a safe 

 conclusion as to the particular office of Leydig's fibres in the nervous 

 economy of the worm. 



g. Bearing of the Condition of Leydig's Fibres in Annelids upon 



the Neuron Theory. 



The conditions of Leydig's fibres in Clymene producta and Axiothea 

 torquata, as I have described them, and especially their relations to the 

 great ganglion cells, have an important bearing on the recent neuron 

 theory. We have here dozens of nerve cells sending their processes into 

 the same nerve fibre. Immediately upon entering this fibre, these pro- 

 cesses lose their respective sheaths, and to all appearances fuse with the 

 contents of the fibre, thus losing their individuality. These facts are not 

 easily reconciled with a theory which maintains that the relation of the 

 processes of one nerve cell to those of another nerve cell is that of 

 contiguity alone. 



The neuron theory is now several years old, and it seems a little 

 strange that wheu the generalization was extended to include the nervous 

 system of invertebrates, no account was taken of the difficulty presented by 

 Leydig's fibres, although the relation of such fibres to ganglionic cells had 

 long been described. It was Waldeyer ('91) who first emphasized the 

 principle of the nerve unit ; and introduced the term " Neuron." He says 

 (p. 52 of Separate) : "Das Nervensystem besteht aus zahlreichen interein- 

 ander anatomisch wie genetisch nicht zusammenhaingenden Nervenein- 



