The spermatogenesis of Peripatus (Peripatopsis) balfouri. 335 



4. The Second Maturation Division. 



In each spermatocyte of the 2ii(l order, the product of the di- 

 vision of the spermatocyte of the 1st order, there are 14 univalent 

 chromosomes grouped closely together near the centrosome pole 

 (Fig. 199). The two centrosomes of a pair do not begin to wander 

 apart from one another until this late stage in the dyaster, unlike the 

 case in Penfatoma (Montgomery, 1898) where this movement commences 

 as early as the metakinesis. The two centrosomes wander apart in 

 a line perpendicular to the axis of the previous spindle (Figs. 200 - 202) ; 

 during this process I could not determine the presence of a central 

 spindle (w^hich is clearly marked in Fentatoma at this stage), and pole 

 radiations are exceedingly faint. The centrosomes have become very 

 small in comparison with earlier stages , and are scarcely larger 

 than those of the sperraatogonic mitoses. When the centrosomes se- 

 ])arate it can be clearly seen that from each of them there passes 

 one mantle fibre to each one of the chromosomes, so that each centro- 

 some forms the apex of a pyramid of very delicate mantle fibres : this 

 is quite corroborative of my previous result, that in the monaster stage 

 of the foregoing division there are attached to each chromosome two 

 mantle fibres from each pole of the spindle. I have found only a few 

 cases showing this early separation of the centrosomes, so that it 

 l)robably proceeds with relative rapidity. Finally the centrosomes 

 reach points where they occupy the ends of a diameter of the cell 

 l)laced at right angles to the axis of the previous spindle. 



Thereby the univalent chromosomes become gradually moved into 

 the equator of this 2nd maturation spindle in such a way that their 

 constrictions (longitudinal splits) come to lie in the plane of the 

 equator (Figs. 203, 204). These chromosomes are univalent, and no 

 ring-forms are found ; many of them have the form of an elongated 

 dumbbell, others of a flattened dumbbell; their constrictions are the 

 marks of the longitudinal splits, and are usually clearly apparent at 

 this time. 



Now if the original spindle of the 1st maturation division were 

 still present we would find that the axes of the constrictions on these 

 chromosomes would be approximately parallel to the axis of that 

 spindle. Hence though the achromatic spindle of the 2nd maturation 

 mitosis has come to occupy a position at right angles to that of the 

 first, the chromosomes themselves have undergone no such revolution 

 upon their axes, but each univalent chromosome now lies in the same 



