In Different Species. 65 



melanogaster does not agree in position (with respect to yellow, etc.) 

 with the vermihon in virilis (see fig. 16). In the former it is between 

 singed and forked, while in the latter it is between crossveinless and 

 singed, at a point very close to crossveinless. The order in willistoni 

 agrees with that in virilis, suggesting that in these the vermilions 

 may be homologous. If crossveinless in willistoni should prove 

 to be "located" between yellow^ and vermilion, the parallehsm would 

 be complete.^ 



The only other case of close resemblance between sex-linked 

 characters in these two species is that of short. Short virilis is 

 similar to, but less extreme than, the least extreme of the short 

 allelomorphs in willistoni (cf. Lancefield and Metz, 1922, fig. 13). 

 Its locus with respect to the other characters considered above, 

 however, is very different (fig. 16). This indicates that the two are 

 merely mimics or else that a rearrangement of genes is probably 

 involved. The former interpretation seems more probable, for 

 mutants with short veins appear frequently — indeed a second 

 "short" is now known in willistoni which is due to a mutation in 

 another locus in the X chromosome (unpublished data). 



Comparison with Sex-Linked Characters in Drosophila obscura. 



(Figure 16.) 



When the sex-linked characters of D. ohscura are considered in 

 relation to those of virilis, two different comparisons may be made, 

 according to the manner of orientation of the map. These have 

 been discussed by D, E. Lancefield (1922, p. 377), and need only a 

 brief review here. Yellow, vermilion, and singed are fairly similar in 

 the two species and their loci are in the same sequence. It is also 

 the same sequence as that of yellow, vermilion, and forked (probably 

 singed) in willistoni. The map-distances in the three species differ 

 considerably, but this appears to be correlated with general differ- 

 ences in crossing-over in the three. The parallelism of these series 

 is suggestive of homology throughout, especially since an additional 

 character, scute, is present and coincides in its locus in ohscura and 

 willistoni. The case for willistoni and obscura is further strengthened 

 by the fact that both possess V-shaped X chromosomes and that 

 the series begins (with scute and yellow) at approximately the 

 middle of the map. 



The alternative comparison of the virilis and obscura sex-linked 

 groups is based on the resemblance between the two "glazed" 

 characters. Glazed in obscura agrees with glazed in virilis both in 

 appearance and in the sterility of the females. If these characters 

 are considered as parallels, it is necessary to reorient the maps or to 



* Subsequent data indicate that this is the case. 



