In Different Species. 67 



involve an irregular series of wing modifications that ranges from the 

 normal condition through various long, narrow, curved, or shortened 

 forms (see figs. 5 to 7 in plate 3) to a condition in which the wing is 

 very short and broad and is wavy or crumpled. In both cases the 

 wings are frequently asymmetrical and are often held at an angle 

 from the body. Crumpled is in the third-chromosome series in 

 melanogaster at approximately 93 (unpublished data of C. B. Bridges), 

 and concave is in the second-chromosome group of virilis (fig. 7). 

 If we consider the genes homologous, we would compare the virilis 

 second-chromosome group with the third of melanogaster. The 

 only other third-chromosome characters that resemble any in 

 virilis are ascute (at about the middle of the map) and spread (at 

 about 65). The latter may be left out of consideration for the 

 present, because the nature of the character makes it of little value 

 for comparison, (i. e., spread- wing mutants are too numerous). 

 Ascute resembles hunch in virilis, which is in the third-chromosome 

 series. This suggests that chromosome III of melanogaster may 

 represent chromosomes II and III in virilis combined. 



The presence in D. ohscura, however, of two characters like ascute, 

 in different linkage groups, casts considerable doubt on the reliability 

 of ascute for comparative purposes. Likewise, it should be noted 

 that confluent is in the same linkage group as concave in virilis, 

 while the characters resembling them in melanogaster are in different 

 groups (II and III). It seems probable that most of the above 

 cases involve mimicry rather than homology. 



The other cases of resemblance among the autosomal characters 

 are equally doubtful, except in the case of bent and net, which have 

 been discussed fully above (pp. 51-53). Telescoped resembles fur- 

 rowed melanogaster in its effect on the thorax, but the eyes are affected 

 very differently in the two cases, and the bristles are short in furrowed 

 and not in telescoped, so that the resemblance is probably superficial. 

 A closer resemblance is shown by the sex-linked character compressed 

 in D. ohscura. It resembles telescoped in nearly all respects (see 

 D. E. Lancefield, 1922). This resemblance recalls the fact that hunch 

 in virilis is in the same linkage group (III) as telescoped and that in 

 ohscura ascute, which resembles hunch, is sex-linked. Thus the 

 two sex-linked characters in the one case resemble the two third- 

 chromosome characters in the other, suggesting that the V-shaped 

 X chromosome of ohscura represents the rod-like X of virilis, plus 

 chromosome III. This is made improbable, however, by the presence 

 in ohscura of an autosomal character like ascute, and also by the fact 

 that the loci of compressed and ascute are in the part of the map 

 that seems to correspond to the X-chromosome map of virilis. 



Only two other characters need be noted in this connection. The 

 first of these is approximated. It resembles several characters in 



