THE GENETICS OF HABROBRACON JUGLANDIS ASHMEAD 



This second method, the method of reverse 

 crosses, does not compensate for linked lethals , 

 but comparison may be made of the six ratios 

 obtainable (two by first method in case reverse 

 crosses are made and four by second) and in 

 case of differing results presence of linked 

 lethals determined. 



Muller's method does not compensate for so- 

 matic overlaps or for difficulty in determina- 

 tion of types, in which case phenotypic ratio 

 involves not only deficiency of certain types 

 below gametic ratio but also excess of other 

 types. If viability difference be absent be- 

 tween A and a and if overlapping be from a to A 

 in the same proportion whether B or b be pres- 

 ent, then the phenotypic distribution AB, aB, 

 Ab, ab, may be corrected as follows: Decrease 

 AB by subtracting ( (AB + Ab - aB - ab)/2(AB + 

 Ab + aB + ab)) X aB + ab and increase aB by the 

 same amount. Also decrease Ab by subtracting 

 ((AB + Ab - aB - ab)/2(AB + Ab + aB + ab) ) 

 X ab + aB and increase ab by the same amount. In 

 other words each of the two numerically defi- 

 cient groups should be increased at the expense 

 of its corresponding excess group by a propor- 

 tion of itself equal to one-half the difference 

 between the sum of the deficient groups and the 

 sum of the excess groups to the total. 



The gametic ratio calculated after this cor- 

 rection is made will in case of linkage always 

 give fewer recombination types than the origi- 

 nal data. If overlapping of one difference is 

 always in the same direction and unaffected by 

 the other difference or correlated differences, 

 the gametic ratio obtained should be correct. 

 Even if overlapping is in both directions (from 



118 



