132 s. iKENo : 



4. All Cß-plauts are white. There are thi-ee kiuds of whites, which I 

 caU /, II and /// respectively. The toUte-I (ccrrhh) possesses perfectly 

 white floral parts, while the lohite-II {ccRRBli) may produce few magenta 

 stripes or spots ou petals and few magenta filaments. The wMle-III is 

 externally psrfectly similar to the lahite-I, but is characterised by producing 

 in certain crosses a number of the so-called " pseudo-white," in all probabihty 

 according to Meudehau rule ; its genotypic constitution is yet unsettled. 



5. The factors JB and J5 are generally in complete linkage or " coujjHug," 

 and act Hke one single factor, but sometimes it changes into a paHicd, causing 

 the production of a certain number of imexpected red individuals. 



6. The colour of vegetative organs and floral parts are correlated to 

 each other. 



7. A small iiuiiil)er of imexpected individuals of several kinds nuiy 

 often be produced, thus for instance, magenta or red, etc. is found among the 

 ofifepriug of orange or white parent (selfed ! ) I have tried to explain such 

 phenomena chiefly (but not all) by means of " reverse mutations." 



8. Bud-mutations were also observed. 



AU expenses incuiTed in conducting the experiments described in this 

 paper have been deh-ayed in part by grants from the Minish-y of Etluca- 

 tion and from the " Keimeikwai " (Society whose object is to give pecuniaiy 

 assistance to scientific investigators, etc., etc.) Thanks are due to the authori- 

 ties of both. 



In x^ractical work invaluable assistance was given by MM. M. Ando, K. 

 KiMUEA, S. NoHAEA and Y. Tanihaka ; I am much pleased to thank sincerely 

 all these gentlemen for their kind assistance. 



