Comparison of Maps. 71 



XI. COMPARISON OF X-CHROMOSOME MAPS. 



Taking into account all, or practically all, of the characters that 

 bear a resemblance to one another, and omitting the remainder, 

 the X-chromosome maps of the six species considered above aj)])oar 

 as shown in figure 16. These maps are all drawn to the same scale, 

 based on percentages of crossing-over. The first three represent 

 X chromosomes of the short, rod-like tji^e; the third and fourth 

 are of the long V-shaped tj^pe; and the last (D. funebris) is rod-like, 

 but uncertain as to length (see p. 7). Three different grades of 

 characters are represented on these maps. Characters that give 

 considerable indication of homology to one or more of the others 

 are represented by words beginning with capitals; those considered 

 as possible but not probable homologues are represented in small 

 type; while the remainder are represented merely by marks to indicate 

 their respective loci. 



The alinement of the maps is based on the apparent degree of 

 resemblance of the characters as discussed above. One of the most 

 striking features noticeable in a comparison of the maps is shown by 

 the relations of j'-ellow. In the three species having the rod-like 

 type of X chromosome, yellow comes at or near the end of the map ; 

 while in the two possessing long V-shaped X chromosomes it comes 

 near the middle. The probability of homology throughout this 

 series is supported by the presence in three species of a very closely 

 linked character scute ( = scutellar) . On the basis of this resem- 

 blance, the maps are placed so that the locus of yellow corresponds 

 throughout, except in the case of the funebris map, where yellow is 

 not represented. The latter has been oriented with respect to notch 

 and forked (see Sturtevant, 1921a). 



If we compare the total lengths of the different maps (omitting 

 funebris on account of the small number of loci), and consider the 

 size relations of the chromosomes involved, we note at once that the 

 maps are not proportional to chromosome length in all cases. If we 

 use the melanogaster map as a standard of comparison, we find that the 

 simulans map agrees fairly closely, but that the mrilis map is nearly 

 50 per cent too long. Likewise, the obscwa map, which should be 

 approximately twice as long, is actually about three times as long. 

 The willistoni map, on the other hand, is much too short. It should 

 be approximately double that of melanogaster, and equal to that of 

 obscura; but instead it is only slightly longer than that of melatiogaster. 



In each of these cases (excluding /wnc6r/s) the number of characters 

 studied is sufficient to make it probable that most of the chromosome 

 is represented. The difference appears to be due, therefore, to dif- 

 ferences in amount of crossing-over in the various s])ecies. This 

 interpretation is supported also by the distribution of the loci on the 

 maps. In willistoni, where crossing-over is presumed to be low, the 



