CROSSINGS IN MELANILM-MOLETS 



25» 



pollination. The small number of seeds in Fi was at least partly 

 due to the weakness ot the Fiplants at the time they were investigated. 



TABLE 3. The nuinbcr of seeds in the purent Unes and in F,. 



Table -t shows how small the number of seeds was in eaeh cap- 

 sule of these sterile» plants. Some of the capsules had no seeds 

 at all and other only very few. A flower of one plant (N:o Vg) had 

 by chance become richly pollinated and therefore developed a normal 

 number of seeds, viz. 69. This capsule was not included in the cal- 

 culation of the mean of this plant. 



An investigation of the fertility of these plants by the way of 

 crossing and germination-experiments with the pollen was also made. 

 When crossed the development of the seeds proved to be a normal. 



In the germination-experiments with pollen an 1 % agar solution 

 to which 25 — 30 % glucose had been added was used. In this excep- 

 tionally high concentration a time of about 24 hours was required to 

 bring about the complete development of the pollen. The pollen of 

 these plants had a germination-power of about 95 % . and this was 

 also the case in all my pure lines of V. trieolor and V. arvensis as 

 well as in Fi and Fo of crosses between these species. 



The eventuality remains, however, that these plants were allo- 

 gamous. I pollinated several flowers with their own pollen, but was 

 forced to break of the investigation before the capsules were quite 

 developed. In plant N:o '^ -, the ovary had begun to swell indicating 

 fertilization. The fact that it has been possible to fertilize all violets 

 of the section Mekmium here investigated with their own pollen also 

 argues against the assumption of allogamy. I have cultivated aliout 

 60 lines belonging to different species, and each and all of them 

 could be self-fertilized. The same was the case with 15 pansy-plants. 

 (The pansies are, as is well known, descendants of hybrids between 

 M ekiniiini-\\o\eis) . 



