BLACKMAN: SPERMATOGENESIS OF THE MYRIAPODS. 67 



The peculiar thing is that the mantle fibers no longer converge to- 

 ward the centrosomes, but to the points which the centrosoraes occu- 

 pied before their last migration. From this center of convergence, 

 which I will call the apical point, parallel strands of linin extend to 

 the centrosorae. These linin bands are, I believe, but the continua- 

 tion of the mantle fibers. Radiating fibers are now seen extending 

 out into the cytoplasm, from the apical point as well as from the cen- 

 trosome proper, although those from the latter are much more distinct. 

 The astral rays connecting the centrosome and the cell membrane are 

 still more pronounced at this stage. The centrosome is still conical 

 in form. 



The centrosomes again move apart and take up their final position 

 upon the cell membrane, where they are flattened out into rather large 

 hemispherical bodies. Well-marked astral radiations still extend out 

 into the cytoplasm. By these last movements of the centrosomes the 

 chromosomes have been drawn apart, and, by the synchronous con- 

 traction of the mantle fibers, have been drawn toward the poles, and 

 come to rest at the point at which the mantle fibers converge, where 

 they are arranged in a densely packed mass. (Fig. 17.) 



When the chromosomes have taken up their final positions the 

 cell- walls begin to constrict. As this proceeds the fibers connecting 

 the two daughter masses of chromatin are crowded together into a 

 bundle, which finally presents the appearance represented in fig. 20. 

 A midbody, or zwischenkorper, is formed at the point where the con- 

 stricting wall approaches this bundle of persisting fibers. This is 

 composed of a number of small, darkly staining bodies about equal in 

 number to the chromosomes. These bodies are arranged close together 

 upon the periphery of the bundle in such a manner as to form a ring. 

 ( Fig. 20. ) A similar appearance has been described by McGregor'^ 

 in Ainphiuma. In Scolopendra, however, this ring does not i^ersist, 

 as is the case in Ainphiuma. 



The division figures of the second spermatocytes are very similar 

 to those of the first. Indeed, the two generations can be distinguished 

 only by the character of the chromosomes. A short spindle is formed 

 ( fig. 21 ), and the cell undergoes changes similar to those in the first 

 spermatocytes. In the early telophase the cell presents the appear- 

 ance represented in fig. 22. The daughter masses of chromatin are 

 smaller than in the corresponding stage of the first division. The 

 chromosomes are so closely massed together that the outlines of the 

 individual elements cannot be distinguished, although, of course, they 

 do not lose their individuality. Connecting fibers extending from 

 one daughter plate to another are jjlainly visible, and astral radiations 



5. McGregor, J. H. : "The Spermatogenesis of Anipliiuma," Jour. Morpb., vol. XV. Sup- 

 plement. 



