454 EDWIN CARLETON MacDOWELL 



dominance steps in and determines the proportion of extras 

 that appears, then steps out when these extras, no longer reces- 

 sives, are bred and give from 71.5 per cent to 95 per cent nor- 

 mals. However much the dominance may alternate in the 

 parent flies, it is with peculiar consistency that the normals 

 appear predominatingly among the progeny. 



The case has obviously not been analyzed; one may suppose 

 that the persistence of the extra flies in the race indicates some 

 relationship between the extra bristle and the germ plasm of the 

 race. Beyond this the data indicate that the race in question 

 would offer excellent material for clearing up just such types of 

 characters whose inheritance seems to be irregular and weak. 

 The statement is made that the results agree with those of 

 Gates ('15) and Shull ('14). Although Gates did find irregular 

 ratios in his later rubricalyx crosses, there is no fundamental 

 likeness between the inheritance of this extra bristle and that of 

 Oenothera or of Bursa: in both these cases the dominance in the 

 first generation was complete and the recessive form did not give 

 dominants. 



The introduction of linkage relations as a method of analyzing 

 complicated genetic situations has occasioned a great advance in 

 the solution of selection problems. However, it is only in such 

 extensively investigated material as Drosophila that this method 

 can be used. Since the first cases of the employment of this 

 method (Dexter, '14, and Altenburg and MuUer, see Morgan, 

 Sturtevant Muller Bridges, '15, p. 191, where the case of truncate 

 wings is presented) several difficult cases have been cleared up 

 by this method. Of these, the most complicated and far reach- 

 ing in its bearing, is the case of beaded wings; first investigated 

 by Morgan, then taken up by Dexter, and finally fully cleared 

 up by Muller ('18). Muller points out very clearly (p. 426) that 

 the action of this character in inheritance could have been used 

 to give strong support to the theory of factorial inconstancy, yet 

 the linkage relations support his hypothesis of balanced lethals 

 so strongly that the conception of factorial inconstancy is con- 

 clusively eliminated. A clear summary of this work is given by 

 Morgan ('18, pp. 386-390). 



