50 CHAELES W. METZ 



TYPE II 



Chromosome groups found in Drosophila repleta, Species A and 



Species B 



The type of chromosome group exhibited by these three species 

 differs very decidedly, at first sight, from that just studied. It 

 may be best understood by examining diagram 2 (a and b) and 

 figures 8 to 13. The figures in this case are all but one (fig. 9) 

 taken from fixed and sectioned material, and therefore show more 

 nearly the proper arrangement of the chromosomes. Examin- 

 ing the individual chromosomes, as was done in Type I, two 

 kinds found in that type are seen to be present here also. These 

 are the small ??? -chromosomes and the sex-chromosomes (in this 

 case the longest pair), as indicated in diagram 2. In addition to 

 these two kinds there remain four pairs of short, rod-like auto- 

 somes in place of the two pairs of long, dumb-bell-shaped large 

 autosomes of Type I. These constitute a fourth kind of chro- 

 mosomes and may be characterized as follows : 



4. The small autosomes. Short, rod-like autosomes without 

 any median constriction; almost invariably straight in meta- 

 phase, and lying in a radial position, i.e., pointing from the center 

 toward the margin of the plate (figs. 8, 10, 12); attached to the 

 spindle fibers at one end, thus going to the poles as straight rods 

 (s, diagram 2 a and 2 b). 



The principal interest of Type II as compared with Type I 

 centers around these eight small autosomes, which, it appears, 

 replace the four large autosomes of Type I. This latter fact at 

 once suggests the idea that each of the large autosomes has parted 

 in the middle, at the point of constriction, and has thus given rise 

 to two small ones. In this way the four large dumb-bell-shaped 

 chromosomes could have been readily transformed into the eight 

 small ones. The evidence of size, form, and behavior of the two 

 kinds supports this view in a rather surprising manner. Diagram 

 2 (a and b) are designed to show this relationship schematically. 

 They are purposely made very schematic, but the essential fea- 

 tures are entirely substantiated by actual division figures. The 



