The spermatogeuesis of Peripatus (Peripatopsis) balfuuri. 337 



the mother cell ; aud since the latter were fouud to be morphologically 

 univalent (with relation to the chromosomes of the spermatogonic 

 mitoses) the chromosomes of the spermatids are therefore semivalent. 

 This valence in the different generations of the spermatogenesis is then 

 as follows: 



spermatogonia 28 univalent chromosomes 



1st spermatocytes 14 bivalent chromosomes 

 2nd spermatocytes 14 univalent chromosomes 

 spermatids 14 semivalent chromosomes. 



Following the metakinesis of the 2nd maturation division the 

 daughter chromosomes become pulled apart to opposite poles of the 

 now elongated spindle (Figs. 208 — 210). The connective fibres which 

 become stretched out between them may be regarded as derivatives 

 of the linin (axial threads) within the chromosomes. It is difficult to 

 determine whether the centrosomes are now double or single at each 

 pole, owing to their extreme minuteness ; in some cases there appeared 

 to be two at each pole of the spindle, in some cases only one. Neither 

 in these stages nor in any of the preceding ones do the centrosomes 

 stand in connection with extracellular flagella, though I have searched 

 carefully for such structures in the free-lying cells of the seminal 

 vesicle. 



Exactly as in the 1st maturation division, the cell body divides 

 by an annular equatorial constriction, without any formation of a cell 

 plate (Figs. 209, 210). A spindle plate (Zwischenkörperchen-plate) is 

 however formed, apparently by equatorial thickenings of the connective 

 fibres as in the spermatogonic mitoses ; these thickenings become 

 gradually pushed together by the equatorial constriction of the cell 

 body (Fig. 210). The ends of the connective fibres which are attached 

 to the chromosomes are the first to disappear. 



5. Structure of the Spermatozoon. 



So far only have I followed the spermatogenesis. The chief in- 

 terest in the metamorphosis of the spermatid into the spermatozoon 

 is found in the fate of the centrosomes; and since in the spermatid 

 the centrosomes are exceedingly minute, Peripatus seemed an un- 

 favorable object for this study. 



But a final word as to the structure of the mature spermatozoon, 

 in so far as it could be determined by sections of fixed and stained 

 nuiterial without disassociation preparations (Figs. 211, Plate 22). The 

 head (H) is long and slender, tapering very gradually to a point; it 



ZooL Jahrb. XIV. Abth. f. Morph. 22 



