114 bulletin: museum of comparative zoology. 



to the axial filament, as he has described for Lithobius. In some stages 

 appearances which might lead to such an opinion are occasionally met 

 with, but in the great majority of cells there is at no time a condition 

 which would warrant this conclusion. In early stages each of the fibres 

 ends free in the cytoplasm, but later they become united to form a mem- 

 brane, which likewise ends free in the cytoplasm without having any 

 connection with the axial filament. 



The similarity of this differentiation of the archoplasm to the "Faser- 

 korb " of Meves cannot be doubted. The structure has nearly the same 

 history in Scolopendra as in the guinea pig. In both it first appears in 

 the form of fibres. Later these fuse to constitute a membrane extending 

 backward from the nucleus. For a time this membrane grows in length, 

 and then it gradually disappears. I think it is possible that this struc- 

 ture may aid in the formation of the tail membranes, as su<zaested bv 

 von Lenhossek ('98) and Niesing (:00), but I have not been able to find 

 any trace of it in the mature spermatozoon. 



C. Axial Filament. 



The origin of the axial filament in the spermatids of myriapods has 

 been explained in one way by Tonniges (:02), in another by Meves (:02 b ). 

 According to the account of Tonniges the filament in Lithobius arises in 

 very much the same manner as here described for Scolopendra. It 

 appears during the course of the migration of the centrosomes from the 

 periphery of the cell toward the nucleus. Although it always arises in 

 close relation to the distal centrosome, Tonniges contends (by implication 

 at least) that it is formed, not from centrosome substance by an elon- 

 gation of one of these bodies, but rather by a metamorphosis of the 

 cytoplasm. 



Meves (:02 b , p. 501), on the contrary, from a study of the same animal 

 concludes, " dass der intracellular Faden (Achsenfaden, Tonniges) von 

 Centralkorpersubstanz gebildet wird." In the early stages two centro- 

 somes (Boveri's centrioles) lie with their axis perpendicular to the cell 

 membrane. " Beide Centralkorper riicken dann auf den Kern zu. 

 Dabei bleibt der distale durch einen Faden, der aus Centralkorper- 

 substanz ausgesponnen wird, mit der Zellperipherie in Verbindung." 



From preparations of Lithobius which I possess, I can confirm the 

 observations of Tonniges. For in Lithobius mordax, as well as in Scolo- 

 pendra, the axial filament arises during the migration of the centmsomes 

 away from the cell wall but is not connected with the cell membrane. 



