640] T^4i?7.4r/O.V AXD IIKREDITY IS LUFINUS 443 



blues out of 150 seeds. Taking into account that four of 

 the eight cultures had only one or two plants and so 

 might have produced dark blues also, this 8 : 27 is prob- 

 ably as close an approximation to a 3 : 1 ratio as could 

 be expected. The data now available would appear to 

 justify the conclusion that the light-blue color is due to 

 a single factor difference and that heterozygotes and ho- 

 mozygotes are phenotypically indistinguishable. 



Owing to the fact that a single cross yields only five 

 or six seeds, it has not seemed profitable to attempt hy- 

 bridizing the light blues and striped whities either with 

 one another or with other forms, until a techinque has 

 been perfected that insures a higher percentage of ger- 

 mination. In nanus, which is naturally crossed, certain 

 observations have been made which indicate something 

 of the relations of certain factors to one another. Seeds 

 of a white-flowered plant collected last year from un- 

 protected flowers were grown this year and produced 

 both whites and blues apparently like normal wild ones. 

 This would indicate that the mother plant had been vis- 

 ited by a hoo wliicli liad offoctod ])olIination in part with 

 its own ])()lleii and in })art with that of a neighboring 

 dark blue. If this hv the true exi)lanation this white was 

 a recessive one. It might have been a heterozygous domi- 

 nant white, of course, which would be in agreement with 

 the nature of the whites in apricus, Pedigree V. Seeds 

 of pink nanus collected last year yielded only a half 

 dozen plants, all dark blue. This would point to the con- 

 clusion that pink is also recessive to blue. However, 

 these results are too meager to have more than a sug- 

 gestive value. 



ance. Dark-hlnc npnrHs plants invariably have dark 

 seeds. Imt the particular \y\)o of marking .liffrrs accM.rd- 

 ing to the origin. It is suspect tnl that the two tvpes of 

 seeds found in Pedi-ree V, striiMMl wliites, indicate a 

 genetic dilTereuce, possibly distinguish striped whites 



