GERM CELLS IN PEDICELLINA AMERICANA 43 



At the close of the growth period, the most striking elements 

 in the oocyte and spermatocyte nuclei of Pediccllina are the 

 several ring- and their equivalent double rod-figures. The former 

 of these recall the heterotypic mitoses observed in the Amphibia, 

 the Platodes and the flowering plants. In the first of these 

 groups, first Flemming, 'Zj, and later Hermann, '91, Meves, 

 '96, and others, interpreted the rings as having arisen, in the early 

 spermatocytes, through the opening up of the space between the 

 sister chromosomes which had separated in the early longitudi- 

 nal splitting of the spireme. On this interpretation, the space 

 within the ring is thus between like halves and from the relation 

 of the ring to the spindle, the subsequent division is necessarily 

 equational. According to Montgomery, on the other hand, 

 who has lately investigated this subject, the above authors, as 

 well as most others who have worked on Amphibian sperma- 

 togenesis, have overlooked some very essential stages in the 

 origin of these rings. Thus he maintains that what these authors 

 have taken for sister chromosomes in the early spermatocytes, 

 are really not such but different univalent chromosomes which 

 have, as in Peripatiis, united end to end in the previous synapsis, 

 and that the true longitudinal split is of different origin. As has 

 been already observed, this author finds no continuous spireme 

 and consequently no segmentation of the latter into the reduced 

 number of bivalent longitudinally split threads as Flemming and 

 the rest have done. According to him, the young spermatocytes 

 show a reduced number of V-shaped chromosomes. These pres- 

 ent the same characteristic polarity, the connecting band of Hnin 

 at the apices and the wide extension of the angle described for 

 Peripatus. The true longitudinal split occurs very early along 

 each of the arms of the V or the U, soon, however, disappears, 

 only to reappear much later in the dyads of the anaphase of the 

 first maturation mitoses. The free ends of the longitudinally 

 split arms of the V's now unite to form the rings of the meta- 

 phase. The first division would, on this hypothesis, obviously 

 be a reducing one. 



Whatever be the true interpretation of these rings in the 

 Amphibia, they are in the maturation divisions of Pediccllina of 



