The spermatogenesis of Peripatus (Peripatopsis) baltouri. 331 



cytoplasmic substance, such as Auerbach (1874) and Conklin (1899) 

 have described. 



3. Metakiiicsis and Dyaster. 



In the equatorial plate each of the 14 bivalent chromosomes is 

 so arranged, that for each of them the central linin fibre (that which 

 joins the central ends of the two univalent chromosomes) lies, as far 

 as is compatible with the form of chromosome, in the plane of the 

 equator of the spindle (Figs. 179 — 182, 184). In the metakinesis each 

 bivalent chromosome becomes so divided that one whole univalent 

 chromosome is separated from the other, whereby the central linin 

 fibre becomes stretched out to form a connective fibre (Figs. 185, 180, 

 189 — 193). This is a true transverse or reduction division, exactly 

 as in the 1st maturation division of the spermatogenesis in Pentntoma 

 (Montgomery, 1898). Any-one who has taken the pains to read care- 

 fully all tlie chromosomal changes from the time of the last spermato- 

 gonic mitosis, and who has compared the figures, will, I think, 

 grant the correctness of this conclusion, and will not be inclined to 

 consider it as an equation division of a heterotypic character. It is 

 further interesting to note, that the two univalent chromosomes which 

 compose a bivalent one are not always of the same volume; so that 

 in the metakinesis chromosomes of different volumes are separated 

 from one another; this is readily explainable when it is recalled that 

 a bivalent chromosome represents two chromosomes of a spermatogonic 

 mitosis, and that the latter vary in size. 



In this metakinesis the distal ends of the separating univalent 

 chromosomes point to the poles of the spindle, due to the fact that 

 the contracting mantle fibres are attached to these distal ends. Thus 

 the univalent chromosome has reversed the position taken in the last 

 spermatogonic metakinesis: in the latter the central ends of the 

 chromosomes come to lie nearest the spindle poles. Chromosomes 

 which have the more fiequent bent-dumbbell shape have their univalent 

 portions separated more or less in the direction of a straight line. 

 Chromosomes of an annular or flattened shape become usually so 

 halved, that first the distal ends of the univalent chromosomes become 

 separated, then the central ends. 



When the daughter (univalent) chromosomes are thus sej)arating, 

 the longitudinal split becomes again apparent in them. At first this 

 split appears usually as a constriction of the surface of the chromo- 

 some, in a line more or less parallel to the axis of the spindle 



