66 Comparison of Characters 



postulate some rearrangement of loci, for with the above orientation 

 of the maps glazed lies near the upper end in one case and the lower 

 end in the other. 



In both of these comparisons the character "short" has been 

 omitted because it is so unreliable for comparative purposes (see 

 above, under willistoni). It is similar in the two species, however, 

 and the sequence of loci is the same when the maps are oriented 

 with respect to yellow, vermilion, and singed. But the short locus 

 is much farther from the others than it is in virilis. Stubby in obscura, 

 with a locus close to that of short, has also been omitted, although it 

 might be compared with forked in virilis. The difficulty here is 

 that stubby does not have forked bristles, and consequently may not 

 be in the same category as the "forked" and "singed" characters. 

 (See special discussion of these above.) 



Comparison with Sex-Linked Characters in Drosophila funebris. 



Comparison with the remaining species, D. funebris, involves only 

 the character forked. As pointed out by Sturtevant (1921a, p. 63), 

 and as noted above, forked in funebris bears a close resemblance to 

 forked in virilis. It might also be compared with singed, however, 

 and since there are no other apparent parallels with which to con- 

 struct a series, both possibilities must be considered. 



COMPARISON OF AUTOSOMAL CHARACTERS. 



As might be expected from the small number involved, the auto- 

 somal characters of D. virilis include too few "parallels," as yet, 

 to permit of a detailed comparison of the linkage groups with those 

 in the other species. Hence the following considerations are given 

 mainly for the sake of completeness. 



The resemblance between confluent in virilis and melanogaster 

 has been pointed out in previous papers (see under description of 

 confluent). The probability of homology here is reduced somewhat 

 by the fact that dominant characters of this type appear to be fairly 

 common. Triangle in virilis, for instance, is similar to confluent, 

 although less extreme and not lethal when homozygous. Extra also 

 resembles confluent in some cases. The veins affected by confluent 

 seem to be particularly susceptible to modification in this direction. 



In D. obscura a dominant, autosomal confluent is known (unpub- 

 lished data of D. E. Lancefield), but this is much more extreme than 

 those in virilis and melanogaster and involves other veins. It is 

 improbable that this is homologous to either of the others. 



Concave in virilis bears a close resemblance to crumpled in mela- 

 nogaster, and gives perhaps the most convincing evidence of homology. 

 Both characters are autosomal recessives; both affect the scutellar 

 bristles and the hairs on the arista in the same fashion; and both 



