MATURATION MITOSIS IN AMPHIBIA. 263 



It is then apparent that the split along each arm of a bivalent 

 chromosome, is a longitudinal split of each univalent chromosome, 

 and is a preparation for the second maturation mitosis. It is 

 equally evident that the space marked ,r in the figures does not 

 represent any longitudinal split, but a space between two uni- 

 valent chromosomes. Hence in the typical chromosomes of the 

 ring form, as shown in Fig. 8, the space enclosed by the chro- 

 mosome is the space between two univalent chromosomes, and 

 has nothing to do with the longitudinal split. The thickenings 

 so frequently found upon the rings, as in Figs. 7 and 8, are the 

 points of union of the ends of two univalent chromosomes, as 

 becomes clear from the intermediate stages shown in Figs. 6 and 

 7, and are not, as generally interpreted, portions of the chromo- 

 some where the longitudinal split is least evident. The reader 

 may follow the spaces marked .r in the chromosomes from the 

 stage of Fig. 8 back through successive stages to Fig. 2, to be 

 certain of this fact. And it is decisive that this space becomes 

 wider the earlier the prophase, as a comparison of Figs. 8 and 2 

 shows, and does not lead back to a simple longitudinal splitting. 



So the form of the bivalent chromosomes is explained, and the 

 fact established that the heterotypic mitosis, the first maturation 

 mitosis, is not an equation division but separates entire univalent 

 chromosomes, while the second maturation mitosis is equational. 



These views are in disagreement with the conclusions of most 

 other workers on amphibian spermatogenesis, because the writers 

 have mostly assumed that the space marked x in my figures, the 

 space enclosed by the ring, is a longitudinal split. Flemming 

 first described the heterotypic mitosis and named it ; he overlooked 

 the true longitudinal split in each arm of a bivalent chromosome, 

 and concludes the space enclosed by the definitive chromosome 

 to be the longitudinal split. Vom Rath described in Sala- 

 niandra a reduction division on the basis of tetrad formation with 

 spherical chromosomes, but later observers have demonstrated 

 that he studied abnormal mitoses. Meves essentially corroborates 

 Flemming's interpretations, but he did not note the polarity of 

 the spermatocytes as evidenced in the arrangement of the chro- 

 mosomes, though it is shown in his Fig. 47, PI. IV., and hence 

 leaves untouched the question of the origin of the bivalent 



