142 



/•// YSIOLOOICAL GENETICS 



dished the curve. If [nfrabar is actually a dosage difference like 

 Bar, a calculation of X front all these points for the Infrabar 



series must give identical values. This has been done in the 

 second section of the table (fifth and sixth columns), e.g., for 

 B+/fii 3 _ IX = 2.4 A' = 0.6; i.e., Infrabar is a 40 per cent 

 addition to the Wild locals. 



Table 16 



Number 



B + /B< 



B</B' 



B + /B'B' 



B'fB'B' 



B'B'/B'B' 



B/B> 



B+/BB' 



B</BB 



B/B'B* 



B</BB> 



B/BB' 



RB' BB< 



B<B< BB 



BB'BB' 



BB/BB' 



Per cent 

 facet 

 loss 





 54 

 |91 

 >94 

 95 

 97 



59 

 74 

 82 

 95 



91 

 93 

 94 

 95 

 95 

 95 

 96 

 97 

 97 

 97 



73 . 5 

 50.5 

 41.8 

 38.3 



37.8 



37 



27.9 



26.7 



26.7 



24 1 



The values are seen to vary between 40 and 75 per cent addi- 

 tion; B l then would be about l 1 2 Wild or :i i Bar dosage. This 

 ealeulation is made in a rough way without curve fitting, but the 

 result looks positive. 



The next consequence of this analysis would be that the com- 

 bined dosages of B and B' ought also to fit into the series, i.e., give 

 identical values for X. The same calculations have been made 



