BLACKMAN : THE SPERMATOGENESIS OF SCOLOPENDKA. 47 



At the time when the quadripartite chromosomes are completely 

 formed, but are yet of a granular consistency, the archoplasm exhibits 

 its first marked activity (Fig. 59). Previous to this, it had become re- 

 arranged in a denser mass upon one side of the nucleus, and the centro- 

 somea had moved through this mass Inward the nucleus. Now, however, 

 these structures undergo more striking changes. The mass of faintly 

 staining archoplasm which has surrounded the centrosomes since the 

 telophase of the spermatogonium loses its staining capacity, and the 

 centrosomes now seem to bei,enclosed in a vacuole of hyaloplasm (Figs. 

 57, 58). The sphere containing the centrosomes moves toward the 

 nucleus, and as it passes through the archoplasm, this substance is re- 

 organized and converted into astral rays. When the centrosomes reach 

 the nucleus, they immediately begin to move apart along its membrane 

 (Fig. 59), and as this separation takes place the astral rays continue to 

 become more marked and the reticular archoplasm gradually disappears. 

 The centrosomes continue to diverge until they have reached points 

 upon the nuclear membrane about 100° to 120° apart, at which time the 

 membrane begins to disintegrate and the spindle begins to be formed. 

 In the type of small spermatocytes I have noticed no case in which the 

 centrosomes have reached the full distance of 180° apart, as is usually 

 true in the majority of germ cells. The reason for this phenomenon I 

 cannot state, but it may have to do with the greater relative volume of 

 the chromatic structures in this type of spermatocyte. I have never 

 observed it in the larger cells. 



Another peculiarity of the behavior of the cell at this time concerns 

 the dissolution of the nuclear membrane. As I announced in my first 

 paper on Scolopendra (Blackmail, :Ol), that part of the membrane over 

 which the centrosomes have not passed in their migration is the first to 

 be dissolved, while the remaining portion persists for some time, as shown 

 by the numerous cells in which it is still to be found (Fig. G2). To me 

 this seems inexplicable, and is exactly the reverse of what I should 

 expect. The centrosomes seem to exert a powerful influence upon the 

 archoplasm with which they are in contact; this seems to cause the archo- 

 plasm to dissolve, and later to reappear in fibrillar form. They appear, 

 however, to have an exactly opposite effect upon the nuclear membrane, 

 as is shown by the fact that the part of the membrane with which 

 they have come into contact persists for a considerable time after the 

 rest of the membrane has disintegrated. This would seem to indicate 

 that the two structures, nuclear membrane and archoplasm, are different 

 in their chemical properties. 



