158 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



through the brain tissue, perhaps a reminiscence of its past history. 

 The fact that in Amphioxus the giant axis-cylinder in question occupies 

 a position ventral to the neurocoele, while in the Craniota it lies within 

 the central canal, is not the insuperable objection to homologizing the 

 two that might at first be supposed, since in Amphioxus the neurocoele 

 is not so definitely bounded as is the central canal in craniotes. More- 

 over, in that portion of the central nerve cord of Amphioxus which is 

 immediately below the neurocoele, the nerve elements are even more 

 loosely aggregated than elsewhere. It is easy to imagine that the giant 

 axis-cylinder in question, in growing backward, has taken a course 

 through this loosely organized nerve tissue rather than through the 

 lumen of the indefinitely limited neurocoele. 



On the other hand, it may be that the position of the median giant 

 nerve fibre of Amphioxus within the nerve tissue ventral to the neuro- 

 coele presents, not a secondary and degenerate condition, but the primi- 

 tive position of the original elements from wliich Reissner's fibre has 

 become diiferentiated, and that the fibre has only secondarily, in the 

 Craniota, shifted its position to the central canal. This hypothesis gains 

 in importance, if we recognize the generally accepted view of Rohde 

 and others that the giant nerve-fibres of Amphioxus and the giant 

 nerve-fibres of the ventral nerve cord of invertebrates — of chaetopods 

 in particular — are homologous. It is possible, then, that the optic 

 reflex apparatus may be found in yet more primitive form in inverte- 

 brates, — that Reissner's fibre and the cells which give rise to it are 

 represented by elements in the invertebrate nervous system. 



Known facts, both anatomical and physiological, make probable such 

 a relation between the optic reflex apparatus of vertebrates and certain 

 of the giant fibres and cells of chaetopods. 



(1) It has been shown by Spengel ('81), Rohde ('87), Friedlander 

 ('88), Cerfontaine ('92), Vejdovsky ('88-92), Lewis ('96), and Ham- 

 aker ('98 , that the giant nerve-fibres of chaetopods frequently are 

 formed by the union of the neurites of two or more cells, sometimes a 

 large number of cells. 



(2) Hamaker ('98) and others have shown that the giant fibres of 

 chaetopods extend for long distances through the ventral nerve cord, 

 and by their intimate relation with other centrifugal fibres are in 

 connection with the musculature of every segment of the body. 



(.3) In Nereis the giant fibres are at times parallel and in close 

 apposition, according to Hamaker ('98, p. 115), "and nervous relation 

 with other fibres is established directly between the axis-cylinders " 



