320 



INHERITED LINKAGE VARIATIONS 



'Hi 



c 



II r 



chromosome were mated to females of the constitution 



o p r c ' 



the star not-black offspring must have been of the constitution 

 79 -^r^- -. Nine such females were tested by mating to 



W/J Wlr 



b p r s p males and gave the results shown in the first line of table 17. 

 The data in the second row were obtained in the same way, except 

 that no star had been put in the b p r C H r s p chromosome. The third 

 line represents the offspring of a female (culture 340) of the constitu- 



tion 



C 



II r 



- , produced by mating a male 



pr c 



to a female 



IIr 



. The males in 340 were black purple vestigial arc 

 c Sp 



speck; since no purples were produced the female must have received 

 a b C IIr chromosome from her father; and since she gave 45.2 per 

 cent crossing-over between black and speck, instead of the 9.0 char- 

 acteristic of Cn r females, she must have received from her mother 



II I 



TABLE 17.- 



m 



TABLE 18. 



Cm 



Cnr 



1 These cross-overs are very doubtful. None of them were tested ; and there is a small per- 

 centage of error in classifiying star flies. Similar apparent cross-overs were obtained in working 

 with C tl i, but all were shown, when tested to see if star was really present or not, to be 

 wrongly classified. 



Table 18 and the sixth map of figure 1 summarize the data from 

 these three series of experiments. No second-brood data are avail- 

 able; and the star to purple region gives so few cross-overs that coinci- 

 dence can not profitably be studied. It is, however, very remarkable 

 that all three cross-overs between black and purple were also cross- 

 overs between purple and speck. More data is needed before we can 

 be sure this is a significant result, since purple and speck themselves 

 cross over so frequently (47.4 per cent). 



