Heredity Studies in the Morning-Glory 17 



Theoretical consideration of data. — Twenty-three ma<3:enta plants were 

 selfed and yielded an aggregate progeny of loS plants, all of which were 

 magentas. Twenty-one others were selfed and did not breed true. These 

 are discussed separately. 



Among the first group was one that threw 5 magentas and i mauve. 

 This mauve plant was not checked, however, and may have been a minus 

 fluctuation of the magenta type. Another plant that had originally been 

 recorded as magenta threw 4 dark blues only. It may, indeed, have been 

 a dark blue itself, and recorded after it had begun to wither. 



The magenta color may be supposed to be due to an additional gene I, 

 an intensifier which acts upon the full mauve and produces a deeper 

 saturation of the same hue. Its genotypic formula, then, would be 

 CCRRXXII, and such plants when selfed could throv/ only mauves — 

 they would breed true, as in the above cases. 



If a magenta were simplex for one or more of its genes, however, it 



would throw, when selfed, magentas, mauves, and pinks in the ratio of 



9:3:4. The behavior of the other magentas that were selfed can be 



explained on this hypothesis. There were 21 such plants and they gave 



a progeny of 93 offspring, of which 60 were magentas, 19 were mauves, 



and 14 were pinks. No other types were thrown. The theoretical and 



the observed frequencies are here given: 



Theoretical Observed 



Magenta 52.3 60 



Mauve 17.4 19 



Pink 23.3 14 



93 o 93 



Type 6 Light blue (Plate I) 



Description of the type. — The type considered as light blue is, as its 

 name implies, a pale tint of blue. It turns to pink on withering. 



This type may be supposed to be due to the presence of a bluing gene, 

 B, in addition to those present in the pink type, but to lack the gene I 

 of the magenta type. The light blue type does not breed true, and all 

 evidence indicates that it is heterozygous in one or more genes and thus 

 comprises different genotypes. It is conceivable that a pure strain might 

 be obtained which would breed true, but as yet no such plant has been 

 found in the writer's cultures. Such a plant would necessarily be of the 

 constitution CCRRBB. 



Theoretical consideration of data. — Examination of all the data for all 

 the light blue plants that were selfed shows that one group, comprising 

 21 plants, threw only dark blues, light blues, tinged whites, and whites. 

 These plants must have been duplex for the gene B, as they threw no 



i3>S 



