506 University of California Publications in Zoology [Vol. 13 



many families the arrangement with expectation on sucli an hypothesis 

 is close, but it is not clear what significance is to be attached to the 

 wide departures from these ratios, such as 5 :1, 7 :1 and 25 :1. In his 

 work on Oenothera rubricalyx Gates (1915) obtained similar aberrant 

 ratios, particularly 5 :1. 



The ratios 15 to 1 and 63 to 1 are shown most frequently. This 

 might, from the Mendelian standpoint, indicate that we have two 

 or three equivalent factors for the normal condition vs. extra bristles. 

 The triplicate condition might be thought of as resulting from similar 

 changes in a particular locus of three chromosomes in Drosophila, or 

 by matings of individuals in which different chromosomes had under- 

 gone this change. Since the inheritance is apparently not sex-linked 

 in any case, we cannot assume that the sex chromosomes carry this 

 factor. Similar results have been obtained by Nilsson-Ehle (1909) 

 in w^heat, by Shull (1914) in Capsella, and by Gates (1915) in 

 Oenothera. 



The fact that in the Fg generation (table IX) the ratios are 63 to 1 

 and 15 to 1, representing three and two equivalent factors, seems 

 weighty evidence in favor of the factorial theory. After the first 

 change from the normal, one might, perhaps, expect further changes 

 or successive mutations in the same direction as a result of crossing. 



In MacDowell's (1915) extensive experiments he has postulated 

 an inhibiting unit-factor which prevents the development of extra 

 bristles, and accessory factors which, in the absence of the main re- 

 stricting factor, produce flies with reduced numbers of extra bristles. 

 These accessory factors are therefore different from the equivalent 

 factors postulated in this paper, the former being used to explain 

 the occurrence of flies with more than one or two extra bristles, while 

 the latter is used to explain the ratios in which flies having a single 

 extra bristle occur. In my cultures the extra flies, almost without 

 exception, had only a single extra bristle. The difference in Mac- 

 Dowell's strain, some of which had as many as eight extra bristles, 

 will be referred to again, but it may be pointed out here that this high- 

 grade strain which was obtained at Woods Hole, IMassachusetts. dif- 

 fered from all other known strains in having a number of extra 

 bristles. 



The normal condition of four bristles in most individuals is the 

 dominant condition, yet extra flies appear in all my cultures. These 

 must be judged either as new mutations, or as a result of imperfect 

 dominance or as the expression of a heterozygous condition. Obviously, 



