1905.] NATURAL SCIENCES OF PHILADELPHIA. 171 



call for some fm-ther explanation. Next to the straight or slightly 

 bent dumbbell-shaped bivalent chromosomes the most frequent form 

 is that of an irregular V, such as those lettered K in Plate IX, figs. 

 32 and 33; in each of these figm'cs one univalent half of each such 

 chromosome is shown black and the other white; p of fig. 32 is such 

 a chromosome seen at right angles to the other views. K of Plate X, 

 fig. 34 shows the separation of the components of such a chromosome. 

 The division of one of the forms of y of Plate IX, fig. 32 is shown bj^ y 

 in Plate X, fig. 34. The division of the dumbbell-shaped chromosomes 

 is clear from the figures. Whereas ring-shaped chromosomes are fre- 

 quent in the preceding late prophases, they are only very exception- 

 ally found in the equatorial plate, so that probably by the pull of 

 the mantle fibres upon them these rings change into the form of the 

 chromosomes lettered K. 



In the anaphase of this reduction division as homologous univalent 

 chromosomes move apart from each other, each opens up in the form 

 of a V (Plate X, figs. 35-37). This opening is the reappearance of the 

 longitudinal split, since it is a cleft along the long axis of each univalent 

 chromosome. In no way can it be considered a transverse split, a space 

 between two whole univalent chromosomes. This split is widest and 

 appears first at the end of the chromosome turned toward the equa- 

 torial plane, and rarely extends quite through the opposite end. Verti- 

 cal (fig. 38) and obliquely lateral (fig. 37) views of a daughter plate of 

 chromosomes, i.e., of the chromosome plate of a second spermatocyte, 

 show without exception ten elements, the same as the number in the 

 first spermatocytes; accordingly all the chromosomes divide in the 

 reduction mitosis. But each of the ten elements of the second sper- 

 matocyte is univalent instead of bivalent, and its cleft or constriction 

 marks the longitudinal split. Witnout any indication of a rest stage 

 the centrosomes of each second spermatocyte wander apart from each 

 other, and each through an angle of 90°, so that the axis of the second 

 maturation spindle comes to lie at right angles to that of the first (fig. 

 39). In the equator of this spindle each of the ten chromosomes be- 

 comes so placed that the fine of its longitudinal split coincides with the 

 equatorial plane. In the ensuing anaphase occurs, then, an equatorial 

 division, separation from each other of longitudinal halves of univalent 

 chromosomes. All ten chromosomes divide, and a pole view of one of 

 the resulting daughter cells (spermatids) shows also ten chromosomes 

 (fig. 40), exactly half the number found in the spermatogom'mn; no 

 exceptions were observed to this numerical relation. 



Exactly how the bivalent heterochromosome comports itself in the 



