The Spermatogenesis in Pentatoma up to the Formation of the Spermatid. 57 



(by retraction of its substance into the centrosomes) and its place is 

 taken by a delicate central spindle, which is at first of an ovoid 

 shape {c.sp.3, Figs. 193—195—202); at the earliest stage (Figs. 193 

 195, 198, 199) only the boundaries of this spindle are to be seen, 

 and these probably represent a central spindle sheath (cf. Lauter- 

 born, '97), but later the spindle becomes elongated and then is seen 

 to be composed of a number of delicate fibres, which pass continuously 

 from centrosome to centrosome. This, which seems to have no genetic 

 connection with the secondary central spindle of the monaster stage, 

 may be termed the tertiary central spindle of the reduction divisions, 

 since it is the third one produced. It is seen in these stages with 

 great distinctness in all cells. The proof that it has no genetic con- 

 nection with the secondary central spindle is brought by the fact, that 

 the latter still persists in the dyaster stage connecting centrosomes of 

 opposite pairs (c.sp.2, Figs. 201, 202); but the possibility remains, 

 that the tertiary central spindle might be formed from the pro- 

 ximal ends of the fibres of the secondary one. Perhaps the 

 secondary central spindle and the connecting fibres become changed 

 into pole fibres in the 2nd spermatocytes. 



From the time of disappearance of the nuclear membrane up to 

 the metakinesis a pair of mantle fibres is attached to each end of each 

 chromosome and of the chromatin nucleolus, that is, each pair of 

 centrosomes forms the point of attachment for 8 pairs of mantle fibres 

 (Figs. 160 — 169, 172 — 175). In these stages the centrosomes of each 

 pair are so closely apposed, that it is difficult to determine by ob- 

 servation alone whether each centrosome receives only one of the two 

 mantle fibres of each pair. But the study of the subsequent meta- 

 kinesis shows that this must be the case, since when the centrosomes 

 of a pair separate it may be clearly seen that a single mantle fibre 

 passes from each centrosome to each chromosome (Figs. 192 — 195, 197 

 — 201, 203, 204). Before the centrosomes have widely separated, the 

 mantle fibres of the two intersect one another (Figs. 192— 195, 197 — 200). 

 But when the centrosomes have reached opposite points of the cell, i. e. 

 in the monaster stage of the 2Qd spermatocytic division, they no longer 

 intersect one another, but a single mantle fibre is attached to each 

 end of each chromosome (Figs. 201 — 205). Since two mantle fibres, 

 however, were originally attached to one end of a chromosome, 

 one of them must have retained its original position, and the other 

 have become moved to the opposite end of the chromosome. What 

 is fully determined by the observation of these interesting phenomena, 



