112 University of California Publications in Botany [Vol. 5 



ratios and interpretations which have been advanced, that the re- 

 sults obtained upon tabulating the combinations of characters ex- 

 hibited by the F, generation plants are susceptible of a Mendelian 

 interpretation. As is noted in the following, however, the small 

 number of plants grown may be the cause of the situation. Grant- 

 ing that this is the cause, we may briefly set down the situation 

 we should encounter upon endeavoring to arrange our experi- 

 mental data according to a Mendelian scheme of analysis. The 

 two parental forms differ from one another in respect to at least 

 seven definite pairs of characters. The hybrid in F^, which re- 

 sults from a cross between the two parents, exhibits an approx- 

 imately complete dominance for one parent. There was, how- 

 ever, no difficulty in distinguishing between the heterozygote in- 

 dividuals and the plants of the dominant parent. To observe 

 among the F^ generation plants even an approximation of the 

 expected Mendelian ratios, at least 15,000 plants would not only 

 have to be grown but would need to be critically examined in 

 order that they might be arranged into the more than 2000 

 possible combinations that might appear. Making our calcula- 

 tions along this line we should not be surprised to find that among 

 the 175 plants grown in this experiment not a single true 

 homozygote could be found. It has, I think, elsewhere been 

 noted that the ordinary plant-breeder, without unlimited field 

 space and a number of trained assistants, is forced to group the 

 various appearances of the individuals which constitute his 

 limited F^, F-j, etc., generations under general headings, when 

 he attempts an investigation of the inheritance of a polyhybrid 

 character in which the power of )i is greater than 3 or 4. As has 

 been seen, such a method of grouping has been adopted in this ex- 

 periment (see Baur, 1911, p. 213). Certain plants in the F2 gen- 

 eration resembled, usually rather closely, the general appearance 

 of one parent or the other, while many other plants produced a 

 typical flower of one parent along with a typical leaf of the other 

 parent, or appeared to possess flowers and leaves in which the dis- 

 tinguishing characters of both parents were hopelessly mingled. 

 Thus we have classified the plants in F^ as "resembling" one 

 parent or the other, as "intermediates" or as "blends." We 

 certainly have no basis for either definitely affirming or denying 



