260 



KARL B. FvRISTOFFERSON 



TABLE 4. Till' number of seeds in the partially sterile plants. 



The probable cause of this singular sterility may be the direction 

 of the inlet to the stigmatic chamber; it is antrorse and turned 

 upwards instead of downwards, and a little backwards. 



The difference in the number of seeds in these »sterile» plants 

 and the normal ones is important: the latter yield more than twenty 

 times as many seeds as the former. The difference as to the power of 

 reproduction may be still greater, however. Capsules with a very small 

 number of seeds did not open at all. They remained, as a rule, un- 

 opened on the plant until it withered and the seeds become spoiled. This 

 singular sterility is evidently a selective factor of great importance as 

 these forms probably very soon break down in struggle for life. 



THE COLOUR OF THE PETALS. 



Three types of flower colour in V. tricolor have been in culture. 

 In line 10 — belonging to V. tricolor geniiina versicolor — the corolla 

 was violet all through when fully developed. The just opened flower 

 had only the two upper petals violet; the three lower were light 

 yellow. The flowers were much less coloured in the middle of the 

 summer than in spring or autumn. Most flowers had at that time 

 about the same colour as the flowers just opening in spring. Some- 

 times, however, the lower petals get a violet margin. 



Line 41 belongs to V. tricolor coniophila. Its corolla is of a 

 pure blue colour. Line 40 belongs to V. tricolor ammotropha and has 

 the corolla red (red violet) coloured. The modification in these types 

 is about the same as in line 10. 



All the lines of V. aroensis had the corolla yellow coloured. In 

 spring or autumn it is rather highly coloured; in the height of summer 

 it is light yellow or almost white. 



