150 University of California Puhlications in Botany [Vol. 5 



that the particular type which we have considered to be the 

 small-flowered X. acuminata variety (III), was of recent hybrid 

 origin. As has been seen, nothing in the appearance in F^ of 

 any of the crosses which contain this small-flowered variety as 

 a parent suggests that we are dealing with a heterozygote in 

 the case of variety III. Certainly in the result of the crosses 

 where the parents with large and intermediate flowers are con- 

 cerned, there can be practically no doubt as to the homozygous 

 condition of the parental material; and the appearance of the 

 hybrid between these varieties was practically the same as the 

 appearance of the hybrids which resulted from crosses in which 

 the small-flowered parent was concerned. The number of plants 

 brought to maturity in both the experiments reported upon in 

 this combined paper is small but compares favorably with the 

 numbers raised in experiments to the results of which a strict 

 Mendelian interpretation has been applied (Price and Drinkard, 

 1908). It is to be noted, however, that the flowers upon only two 

 generations of the parent plants have been measured and that 

 thus we may not have fulfilled the criterion set by Johannsen 

 when, speaking of "the hybridisation of types that are quan- 

 titatively characterized," he justly states that they "must be 

 pure lines, the constancy (or, if it may be, the mutability, segre- 

 gative capacity, and so on) of which has been previously studied 

 in a sufficient number of generations" (Johannsen, 1906, p. 

 105). To sum up the above discussion, it may, in general, be 

 said that the occurrence of flowers on any one hybrid plant in F^, 

 the corolla diameters of which show the unaffected "dominant" 

 influence, the unaffected "recessive" influence, and all diameters 

 intermediate between the corolla diameters of the parents, makes 

 it difficult to find an explanation for such a type of inheritance 

 according to any of the most widely accepted hypotheses. The 

 discussion of "intermediates" (see Macfarlane, 1895) as given 

 by Bateson (1909, pp. 235-241) and the four important classes 

 within which intermediate types fall, does not seem to offer the 

 necessary explanation. The "presence" of both the unimpaired 

 D and It influences in F^ makes it rather difficult to apply the 

 " presence-and-absence hypotheses" (Shull, 1909) to a discussion 

 of the experimental results herein reported. Again, none of 



