THE MUTATED GENE 67 



1. The time at which an event leading to degeneration is still 

 possible is limited, as we saw, to a definite stage in wing differ- 

 entiation. This time would then constitute a limiting value. 

 The action of the gene in question may require so much time to 

 reach the necessary concentration for action, or, in the reciprocal 

 conception, the curve of production of the growth substance may 

 drop below the threshold at so late a time in development, that 

 in either case the event approaches the limiting time. The normal 

 variation in the time of developmental processes may then cause 

 a larger or smaller part of the curve of variation of developmental 

 time to cover the time at which the gene-controlled process 

 takes place. The result is that all individuals belonging to the 

 part of the curve beyond the limiting value for possible changes 

 remain normal. 



2. The same result would be obtained if the time of differentia- 

 tion remained constant and the moment at which the curve drops 

 were variable, or if the threshold concentration for the stuff in 

 question were variable or another member of the system or some 

 or all would vary. 



In general terms, then, the phenomenon would result from the 

 general fact of fluctuation of the particular processes of develop- 

 ment in conjunction with the existence of thresholds for the 

 integrating processes. 



An actual example of this type in which some of the elements 

 are known is the following: In Lymantria dispar, male intersexes 

 of different grades may be produced by combining in appropriate 

 crosses sex genes of typical but different valencies. If, for 

 example, a constant female determiner of high valency, a so-called 

 strong F (which has turned out to be a cytoplasmic quality), is 

 combined with one constant male determiner of low valency or a 

 weak M (which is inherited within the X-chromosome), the second 

 M in the male formula (2X = <? ) may be taken from races of 

 different valencies. In the balanced system of sex determiners, 

 which might be written F /MM, which is the general formula for 

 males, the valencies have been combined as F str /M w M vaT , M var 

 meaning that this M alone will be varied in different combinations 

 built up by crossing. The respective valencies of M in different 

 races are known from other experiments. Introducing then 

 M-genes of the highest valency (= strong) through intermediate 

 stages to the lowest valency (= weak) into individuals of this 



