330 WILLIAM LAWRENCE TOWER. 



organism. According to the theory of inhibitors, angora forms 

 have long hair because the inhibitor of hair growth fails to act in 

 one case, giving long hair, and acts in another, giving short hair; 

 and poultry have tails because the inhibitor of tails fails to act 

 in most cases, but acts rarely to give tailless birds, that is, the 

 tail in birds is recessive to no tail, etc. This seems an extreme 

 and unnecessary complication of the present hypothesis. 



Unquestionably, something determines dominance in color, in 

 tall and short peas, etc. Probably the most extreme cases of 

 dominance are those which breed true without segregation in Fj 

 and subsequent generations. A behavior of this type is given 

 in Exp. No. H 410, where the F^ hybrid was exactly like the 

 female parent and continued to breed true during the succeeding 

 generations. 



In Exp. No. H 409/411, there again appeared a dominant 

 type, dominant to the complete exclusion of the other parent, 

 which continued to breed true generation after generation; yet 

 from the same parents there arose individuals which gave a dif- 

 ferent behavior, and both behaviors were based as far as any 

 evidence is available, upon one and the same kind of germ cells. 

 These cases represent dominance of the extreme kind, in which 

 there is a total disappearance of the characters of one of the 

 parents in the subsequent progeny. From these, we pass 

 over such cases as those observed by Mendel and many 

 others, to these F^ hybrids in which the individuals are a blend 

 and intermediate between the two parents. Such are given in 

 several of the experiments cited in the body of this paper, and 

 in some there were a series of gradations from the extreme of 

 total dominance to the other extreme of an intermediate or 

 blending condition between the two attributes. 



In these blends, which are typical heterozygotes, the dominant 

 attribute is diminished in manifestation in F^ but appears un- 

 diminished in F2 — but what shall we say of Exp. No. H 701 B, 

 where in F^ there appeared the undecimlineata type, mid type, 

 and signaticollis type? In this series, from the same parents, the 

 presence dominates the absence, blends therewith, is recessive 

 thereto, in F^ and only the mid types are heterozygous — all from 

 one and the same pair of parents in each repetition of the experi- 

 ment. 



