STUDIES ON MENDELIAN FACTORS IN AQUILEGIA VULGARIS 185 
together with the red one might exercise such an influence. The 
difference between the red and dark-blue types in table 8 is 1,7 cm. 
+ 0,41 and D:e(D) —=0,1. In table 9 they have about the same 
average. When examining the height of the light-blue and com- 
paring them with the dark-blue the differences are seen to be about 
the same as between the white and the red, as seen in table 10. Table 
8 gives a difference of 95 cm. + 1,922 with D:&(D) =4,9. Between 
the dark-blue and the white it is 7,0 cm. — 2,13 with D : e(D) = 3,22. 
Thus, the light-blue factor does not exercise any influence on the 
stem-height, even when present together with the red factor. 
Consequently the variation of the stem-height is mainly due to 
the presence or absence of the red factor R. The number of indivi- 
duals in the D3-lines is rather small but as the variation complies with 
the statistical demands, and all the cases examined point in the same 
direction the interpretation given here will no doubt be correct. 
The flower colour factors have also a very striking influence on 
the character of the seed-coat. The dark-blue and the red flowering 
types have a glabrous and shiny seed-coat; the light-blue and white 
ones a dull and dead-black one. According to a microscopical: exa- 
mination made by Mr. P. Orssox the shiny appearance is due to a 
fine curling of the seed-coat; in the dull and dead-black seeds the 
seed-coat was found to be roughly granulated. Another difference is 
the colour of the endosperm of the seed, which is much darker in the 
red and dark-blue plants than in the light-blue and white. 
The differences in seed-coat and endosperm also depend on the 
strong pleiotropical influence of the red factor R, whereas the light- 
blue factor B has no visible effect at all on these characters. 
Tıne Tammes’ (1913) investigations on flax have shown that the 
flower colour sometimes varies in correlation to the size of the seed 
and, therefore, I thought it well to find out whether or not a corre- 
lation was to be found in Aquilegia. Tammers has measured both the 
length and the breadth of the seeds, and this is a very good standard 
of its size, of course. The seeds of Aquilegia, however, are rather 
small and rather troublesome to handie and therefore I considered 
the average wheight of the seeds to be a sufficient good standard of 
the size of the seed. In order to determine the mean of the seed- 
weight I weighed at least 300 seeds of each plant. The result was 
then recounted in the wheight of 1000 grains. 
When examining the variation of the size of the seeds of the 
different types of flower colour I made use of the same method as 
