Breeding experiments which show that hybridisation and mutation etc. 273 



1. The results of the crosses 0. graiidi flora (^) :: nibricalyx and 

 its reciprocal rubricalyx :■. grandiflora(^) in F^ and Fo are 

 described, as well as various back-crosses, double-reciprocal 

 crosses, and collateral experiments. Since grandiflora and 

 ruhricalyx differ throughout in a large number of characters, 

 and do not exactly agree in any single point, these crosses may 

 be considered interspecific. 



2. The Fl hybrids were in both reciprocal crosses uniform and 

 intermediate between their parents, but in certain features they 

 resembled more tlie male parent and were therefore patroclinous. 



3. The inheritance of the red pigmentation character, R, which 

 originated by a mutation, and distinguishes ruhricalyx from 

 rubrincrvis, is considered at length. In the F^ this character 

 is more or less completely dominant in both crosses, and in 

 F2 there is segregation into R {nibricalyx pigmentation) and r 

 {rubrincrvis pigmentation). The most sharp and characteristic 

 •distinction between R and r is the presence in the former of 

 red and in the latter of green hypanthia (flower stalks), though 

 the differences in pigmentation extend to every part of the 

 plant. 



4. An important feature of the results is that, although the nibri- 

 calyx parent in both crosses was heterozygous for R, splitting 

 in a Mendelian monohybrid ratio of 3:1, and giving approxi- 

 mately the anticipated i : i ratio in Fi crosses; yet in different 

 F2 families different ratios of R : ;' were obtained. 



5. Thus in certain families the ratio was almost exactly 3 : i, 

 in certain others approximately 5:1, and in one case nearly 

 10 : I. This ratio was moreover found to be constant for the 

 different families of offspring of a given Fj individual in the 

 cases tested. 



6. The conclusion is therefore reached, that different Fi indi- 

 viduals possess different degrees of prepotency as regards the 

 percentage of R plants which will appear in their offspring. 

 The various ratios obtained cannot be explained on a Mendelian 

 basis by assuming the presence of more than one factor for R, 

 but it is also recognized that the term prepotency in itself 

 does not furnish an explanation. 



(1) A race o£ grandiflora derived from its native locality in Alabama, U. S. A. 



(■-) A slightly different race of gnnidiflora from a naturalized locality near 

 Birkenhead, England. 



Induktive Abstaiiimungs- mid Vererljuiigslelire. XI. IS 



