90 



T. H. MORGAN 



In the above Fo results the proportions of the males give a 

 measure of the amount of crossing-over between the lethal and 

 the white loci. The non-cross-overs number 404 + 278 or 682. 

 The data from the males of tables 3 and 4 bear on this same point. 

 The combined data give 2 -|- 4 -F 1 or 7 for the number of cross- 

 overs in comparison with 370 -\- 589 + 682 or 1641 for the num- 

 ber of non-cross-overs with -respect to the interval L-W. This 

 gives 0.4 as the percentage of crossing-over. 



The data (table 7) of the cross in which miniature is involved 

 give as the percentage of apparent crossing-over between lethal 

 and miniatm'e 41 which is somewhat higher than the value 

 derived from our other data for this case. 



Similarly the white-eosin daughters (cf . table 6) were mated to 

 their white-eyed, long-winged brothers and gave: 



L93A. 

 L93D. 



195 



195 



94 



98 



66 



53 



The white-eosin daughters that gave the above results should 

 rarely contain lethal (because of linkage) and the result 1 :1 con- 

 firms the expectation. Two classes of males, viz., white minia- 

 ture and eosin long are smaller than the other male classes because 

 they are due to crossing-over between the white-eosin and minia- 

 ture loci. 



The most convincing evidence bearing upon the order of these 

 factors in the linear series is that derived from a consideration 



