The spermatogenesis of Peripatus (Peripatopsis) balfouri. 317 



of these radiations seemiug to pass gradually over into the cyto- 

 reticulum ; from this stage on, until the time of disappearance of the 

 nuclear membrane (when the pairs of centrosomes have reached op- 

 posite points on the surface of the nucleus), these polar radiations 

 are much more strongly developed than at any later stage (compare 

 the figures on Plates 21 and 22). The pairs of centrosomes seem to 

 wander in a straight line to the surface of the nucleus; and when 

 they have reached the latter they attain their greatest size, which is 

 many times greater than that of the spermatogonic centrosomes, and 

 stain most deeply. When they have reached the nuclear membrane, 

 an axis joining the two centrosomes of a pair is at first usually per- 

 pendicular to the nuclear membrane. 



After the two pairs of centrosomes have reached the nucleus, 

 they migrate along the outer surface of the latter apart from one 

 another, one to the right and the other to the left of a line connecting 

 the central and distal poles of the cell, until finally they stop on the 

 surface of the nucleus at points about half way between its central 

 and distal poles (Figs. 161—165, Plate 21 ; Figs. 166—169, Plate 22). 

 They here mark the poles of the definitive mitotic spindle, soon to 

 form; and thus the spindle of the 1st maturation mitosis comes to 

 be formed in a position perpendicular to the axis of the last sper- 

 matogonic spindle. There can be no doubt about this conclusion, for 

 in the last spermatogonic division the pole of the spindle marks what 

 has been termed the central pole of the cell, the equator of the 

 spindle, where the Zwischenkörperchen plate came to be formed, the 

 distal pole; these two poles of the cell, as we have seen, can be 

 positively recognized up to the present prophases; and in these pro- 

 phases we see one pair of centrosomes wandering to the right, the 

 other to the left, of an axis joining the central and distal poles of 

 the cell. Even at this late stage the distal pole of the cell is clearly 

 marked by its greater amount of cytoplasm and yolk substance, and 

 by the persisting portions of the idiozome cup. 



In none of the stages just described is any trace of a central 

 spindle to be seen, nor do any of the polar radiations seem to connect 

 the two pairs of centrosomes. Very probably, however, there is some 

 organic connection between the two centrosomes of a pair while on 

 their migration, even though it be invisible, for otherwise we could 

 not well explain how the centrosomes preserve their close proximity 

 during their migration. But between the two pairs of centrosomes 

 there is neither any visible organic connection, nor yet would any 



