CROSSINGS IN MELANIUM-VIOLETS 



BY KARl. li. K HI STOFFE US ON 



I.L'NI) 



Already many years ago I l)Ocame interested in the great varia- 

 ^/V tion ol our wild Melaniiim violets, viz. Viola tricolor L. and 

 V. arvensis Murr. In 1912 I begun my genetical investigations in 

 these species and collected a number of different types, partly inland- 

 forms from the neighbourhood of Lund and partly coast-forms from 

 the shores of the Sound and the Baltic. The plants were planted in 

 flower-pots. Selfed seeds were taken from each plant and several 

 crosses were also made. Next year the seeds germinated very well, and 

 I obtained about 50 pure lines and the Figeneration of the 

 crossings. This year the F^-generation of one cross was also 

 obtained; in some years it is possible to grow 2 generations. In 

 1914 the seed also germinated very well, and a large number of 

 isolations and several new crosses were made. Unfortunately I got ill 

 in the autumn and therefore only a part of the seeds could be har- 

 vested. From 1915 the troubles begun; only a few of the lines ger- 

 minated this year and therefore my cultures become very small; I 

 made a number of new crosses, however. The same happened the 

 following year. In 1917 I obtained some Fi- and F:..-generations and 

 also some pure lines. On travels the greater part of that summer I 

 could not make any isolations. As most of the plants wintered in 

 good condition they were made in the following summer, and a 

 large number of seeds of the Fi-plants and of the pure lines was 

 harvested. In 1919 not a single seed germinated though some thou- 

 sands were sowed. Next year some selfings and crossings with un- 

 controlled plants from my experimental field were made, but the 

 seed would not germinate. The experiments were then discontinued. 

 The probable cause of the poor germination in the last years is 

 not easily found. Possibly it depends on too warm storing during 

 winter. I made some germination-experiments with seeds from 

 V. arvensis last summer. As far as could be judged from these pre- 

 liminary trials the seed germiiiated best when sowed immediately 

 upon their harvesting. In 1913 and 1914, however, the seeds ger- 



Hereditas IV. 17 



