Jan., 1908.] Variation in Syndesmon and Hepatica. 
435 
4. The number of sepals and carpals is greater on the ter¬ 
minal flowers than on the lateral ones. 
5. The number of carpals and sepals is greater according as 
the number of flowers on the plant is greater. 
TABLE VII. CORRELATION OF POSITION AND NUMBER OF FLOWERS WITH 
NUMBER OF SEPALS AND CARPELS IN THE LOT FROM A. MADISON. 
MAY 2, 1902. 
No. of Flowers 
on Plant 
No. of Sepals 
No. of Carpels 
n 
terminal 
n 
lateral 
n 
terminal 
n 
lateral 
2 . 
4 
6.0 
4 
5.5 
2 
9.5 
2 
10.0 
3 . 
31 
6.3 
64 
5.8 
19 
13.0 
38 
10.2 
4 . 
11 
7.1 
33 
6.2 
10 
16.2 
30 
12.6 
5 . 
3 
7.7 
12 
6.25 
2 
16.5 
8 
15.2 
All plants. 
49 
6.5 
111 
6.0 
33 
14.0 
78 
11.7 
6. There are probably changes in the number of the several 
parts through the season, but the data is too conflicting to permit 
generalizations as to the direction of these changes in each case. 
7. These results agree with those of Shull on Aster that place 
modes of floral parts cannot be used for geographical comparison 
without the most careful discrimination as to the other causes 
which affect the number of parts. 
HEPATICA. 
A study of Table VIII leads to the following conclusions: 
1. Hepatica acuta (Pursh) Britton has a larger and more 
variable number of sepals than Hepatica Hepatica (L.) Karst. 
2. The proportion of the individuals having flowers of the 
several colors differs greatly from one general locality to another 
and even from grove to grove in one locality. 
3. There is no consistent correlation of appreciable magni¬ 
tude between the number of flowers or number of sepals and 
the color of sepals. 
4. The number of sepals per flower is regularly greater where 
the number of flowers per plant is greater. 
5. The number of sepals has a skew variation towards 
increased number of sepals in conformity with most floral 
variation. 
