196 Davis. 



As in the crosses previously considered the reciprocal hybrids of 

 iiennis and grandiflora in certain characters strongly resemble one or 

 the other of their parents and show corresponding points of contrast 

 when compared with one another. They resemble the pollen parent and 

 are consequently patroclinous (1) in the morphology of the rosettes and 

 in the relative promptness with which the rosettes sent up their central 

 shoots, (2) in the height and branching habit of the mature plants, (3) in 

 the form of the leaves, (4) in the morphology of the inflorescence, (5) in 

 the position of the stigma relative to the anthers, (6) in the time of 

 flowering, (7) in the length of the capsules. A matroclinous tendency 

 was exhibited in the form of the sepal tips, and in the greater amount 

 of anthoeyan pigmentation on the stem and sepals of the hybrids from 

 seed develo])ed on grandiflora. 



Here, as in the other crosses, I have failed to find certain e^ddence 

 that any character of the pollen parent appears in the hybrids in ab- 

 solutely pure form. Apparently there can always be discovered some 

 evidence of the influence of both parents in the modified expression of 

 the characters examined even where patroclinous or matroclinous ten- 

 dencies are most strongly shown. 



The behavior of anthoc.van in the hybrids of biennis and grandifiora 

 was not so sharply defined as in the crosses between biennis and fran- 

 ciscana. The coloration was expressed in varying degrees and was 

 sometimes absent in structures where its presence was to be expected. 

 Thus some rosettes of both reciprocals were almost clear green although 

 those of the grandifiora parent are usually heavily blotched with red. 

 This is difficult to explain on the Mendelian theory of dominance except 

 on the supposition that this race of grandifiora is heterozygous for 

 anthoeyan coloration and forms some gametes carrying few or none of 

 the factors responsible for this character. It is true that grandifiora 

 exhibits much variation in the amount of anthoeyan pigmentation, but 

 the race employed in this study has always produced plants that are 

 well colored. 



The hybrids of iiicnnis and grandifiora arc vigorous and produce 

 an abundant yield of seed over a long flow'ering season. The parents, 

 however, do not lend themselves so readily to the comparison of con- 

 trasting characters and for this purpose the cross is not so favorable 

 as that between biennis and franeiscana. 



