CROSSINGS IN MELANIUM-VIOLETS 269 



Fa-generation would have shown the vahdity of this assumption but 

 the seeds did not germinate. 



Rasmuson (1916) accounts of a monohyhrid segregation in Vitis 

 with regard to the albomaculata-characlcr. Cohhkns (1920) found the 

 same case in Ipomca and Tropœolum. Dahlgiu:n (1921) recorded a 

 differing mode of inheritance in liarhdren. When crossed with normal 

 green the /'\-generation becomes green. l\, showed dihybrid segrega- 

 tion in most families, though the number of the white-spotted plants 

 was too high. Two Fo-families, however, showed a monohybrid segre- 

 gation. Dahlgren presumes that this monohybrid segregation would 

 be due to the fertilization by polh'ii grains, which had nuitaled, 

 and which therefore were carrier of only one Mendelian factor for 

 normal green. However, he has not grown the Fa-generation and the 

 15:1 ratio may therefore be uncertain, especially as the plants were 

 standing very close together. It seems very probable that a selection 

 of the albomaculata-forms might have taken place. 



THE SIZE OF THE PETALS. 



My pure lines of wild Melanium violets represented two diffeient 

 types as to the size of the corolla. The lines of V. tricolor and one 

 of the lines of V. arvensis had relatively large petals; in most of the 

 other lines of V. arvensis they were considerably smaller, as is seen 

 in table 9. 



The seasonal modification is very great. In spring and autumn 

 the corolla of the first mentioned type was rather large; in the height 

 of summer it was diminished to about "/s of the size of the spring- 

 flower. The second type, the one most commonly found in V. ar- 

 vensis, became so small-flowering in the height of summer that the 

 petals scarcely covered the stamens and the pistil in the bud. 



The difference in soil does not influence the size of corolla in so 

 high a degree, as Wittrock (1897) states. So far as I could see the 

 size of the corolla was not influenced in a visible degree when a plant 

 was transferred from a sandy locality to fertile manured soil. I 

 made no measurements, however. The number of flowers on tlu^ 

 plant, on the contrary, became much larger. The leaves, stipulae 

 and on the whole the vegetative parts became much stronger deve- 

 loped by such a transferring. 



This behaviour of the petals, viz. the small influence of unequalities. 

 in the soil upon their size, would make the Melanium violets to an exeN 



