40 Linkage Group III. 



escoped 0. The 58 wild-type flies here should be in the scaly class 

 but do not show the eye modification. The double-recessive class 

 is small, as usual, on account of low viability. 



Scaly and garnet. — Linkage between scaly and garnet was shown by 

 the usual back-cross test. This gave (M 289, 311) 42 scaly, 49 garnet, 

 3 wild-type (probably genetically scaly), and no scaly garnet. 



Hunch, telescoped, and spread. — The following F2 counts from a 

 mating of hunch telescoped by spread indicate that spread probably 

 belongs in the third linkage group; hunch telescoped 86, spread 110, 

 wild-type 227, hunch 9, telescoped 22, hunch telescoped spread (4?), 

 hunch spread 0, telescoped spread 0. The four doubtful hunch 

 telescoped flies had wings slightly spread, but this condition is occa- 

 sionally found in hunch telescoped flies, and the present cases are 

 probably due to hunch. Tests of spread with members of groups II 

 and IV show no linkage, and the same is probably true in the case 

 of group V, although the data are not conclusive as yet. 



Cross-over Values in Group III. 



Scaly and hunch. — Two back-crosses using heterozygous females 

 gave 35 cross-over flies and 99 non-cross-overs (experiment 57). 

 The wild-type and hunch classes are probably enlarged at the expense 

 of the other two through the occasional failure of scaly to manifest 

 itself, but the error due to this is probably not as great as that due to 

 viability. As they stand, the data indicate approximately 26 per 

 cent crossing-over. 



Scaly and telescoped. — Two back-crosses (experiment 58), using 

 heterozygous females by telescoped males, gave 237 cross-overs to 313 

 non-cross-overs. These counts, like those in the preceding case, 

 are subject to some error due to viability and possibly to errors in 

 classifying scaly flies, but they leave no doubt that the cross-over 

 value is high (43 per cent here). 



Scaly and spread. — Spread has only been tested with scaly and 

 with garnet, and hence its locus is not yet accurately placed. Five 

 back-crosses of females heterozygous for scaly and spread are sum- 

 marized under experiment 56. These give a cross-over value of 

 approximately 21 per cent (counting only scaly flies). This is 

 subject to considerable error, but the value is sm.aller than that 

 given by any of the other three characters when used with scaly and 

 indicates that the locus of spread is probably between scaly and 

 hunch, or else above scaly. 



Scaly, hunch, telescoped. — Under experiment 60 are given the results 

 of 8 back-crosses of heterozygous females with hunch and telescoped 

 in one chromosome and scaly in the other. The cross-over values 

 here are approximately 41 units between scaly and hunch and 19 

 units between scaly and telescoped. These values are only approxi- 



