238 MR. I, H. BURKILL ON SOME VARIATIONS IN 
from under trees in Cambridge. Such being the case, it appears 
as if the soil might influence the proportion of stamens to carpels, 
if not also the number of these. 
Correlations.—The foregoing Tables show that there is a pro- 
portion between stamens and carpels, probably constant under 
the same conditions, and at any rate unaffected by the decrease 
in number of parts which occurs as the flowering period passes 
over (cf. Table XIII.). But beyond this the number of stamens 
and carpels possesses a definite relation to the number of 
petals, most decidedly shown throughout the flowers which I 
examined *, The following will serve as instances :— 
TABLE XV. 
Correlation in increase in number of Petals, Stamens, and 
Carpels in Flowers of Ranunculus Ficaria. 
3 sepals, | No. of 
| 
Cambridge No. 1. Ashbourne. Cayton Cliffs, Top, No.1. 
Average | Average No. of Average |Average No. of Average | Average 
flowers. of of of of of of 
stamens. | carpels. flowers. Stamens. carpels. flowers. stamens. | carpels. 
7 petals...| 11 22:54 | 12:91 1 230 | 160 4 370 | 390 
18 | 2355 | 1288 | 53 3045 | 16:77 60 3673 | 3133 
0 | ...... | ...... 19 31:05 | 18-21 19 40°26 | 36:68 
0 
esc lone. 4 24:0 99:0 | 18 49:54 4223 | 
But with regard to the sepals there is no such correlation. In 
the flowers obtained from Beresford Dale there was a considerable 
number with more sepals than three, the petals usually being 
reduced in number corresponding with the increase in number 
ofthe sepals. Table XVI. shows this. 
* H. Müller has given for Myosurus minimus, ‘Nature, xxvi. p. 81, 1882 
(and x. p. 129, 1874), the results of the examination of approximately 400 
flowers. His figures show, but not very distinetly, some correlation between 
the numbers of petals and stamens ; and he remarks that the greater number of 
organs was accompanied by greater size of the flower. 
