No. 5.] 



DESMOGNATHUS FUSCA. 



205 



ceeds according to the heterotypic plan, as both Flemming and 

 Meves have already described in Salamandra. The twelve 

 segments lose their position relative to the centrosphere, split 

 longitudinally and incompletely (or, according to Meves, com- 

 pletely but with a subsequent fusion of their ends ; I have not 

 been able to determine this satisfac- 

 torily), and by a shortening and thick- 

 ening become converted into rings, 

 loops, or by twisting into 8's, as has 

 been so often described and figured 

 since the appearance of Flemming's 

 paper in 1887. The split chromatin 

 segments, in the form of rings or 



'=' ^ Fig. 2. 



loops, generally more or less irregular, 



take up a peripheral position in the nucleus under the nuclear 

 membrane. The centrosomes diverge within the centrosphere, 

 the spindle is formed between them, in which the chromatin 

 rings take up their position, forming the figure so characteristic 

 of heterotypic mitosis. In the anaphase, as the daughter- 

 chromosomes are passing to the poles, a second precocious, 

 longitudinal splitting takes place, as recognized by Flemming. 

 As the chromosomes approach the poles, however, they become 

 so closely massed (possibly fused I) that it has been impossible 

 to trace the continued existence of this splitting. In the telo- 

 phase they become again separated from each other, still retain- 

 ing their arrangement in relation to the pole (centrosome) (Fig. 

 2), the apices of the F's all turned in the same direction. 



The nucleus of the spermatocyte of the second order does 



not go into a true resting stage, but the 



chromosomes remain distinct and easily 



distinguishable. Each loop is furthermore 



longitudinally split, so that there are twenty- 



^"^ 3- four (presumably), which, however, are not 



entirely separate and independent, but are joined together in 



twos at the apices of the F's (Fig. 3). 



I believe these pairs are undoubtedly the incompletely split 

 chromosomes of the anaphase in the previous division, although 

 their close massing in the late anaphase has rendered it so far 



