J. A. Harris 63 
TABLE XIII. Correlation for Bracts and Fertility. 
Correlation for 
Correlation for 
Series ot Material 
Bracts and 
Bracts arid 
Number per Locule 
Number per Fruit 
Missouri Botanical Garden, 1905 
v./ v uica ■ ■ ■ ... ••• ... 
- •0753 + -021 2 
- -1057 + - 021 1 
Seeds 
■0287 ± -0213 
- -0039 ± -0213 
Missouri Botanical Garden, 1907 
Ovules 
•0039+ -0116 
■0048+ -0116 
Seeds 
•0603 ± -0115 
•0700 ± -0115 
Missouri Botanical Garden, 1908 
Ovules 
- -0830+ -0138 
- -1053+ -0137 
Seeds 
- -0638 ±-0138 
- -0075 ± -0139 
Cold Spring Harbor, 1908, A 
Ovules 
•0252 + -0171 
- -0336+ -0171 
Seeds 
■0032 ± -0171 
•0041 ±-0171 
Cold Spring Harbor, 1908, B 
Ovules ... * 
■0488 + -0283 
■0683 + -0282 
Seeds 
- -0110 ± -0284 
- -0077 ± -0284 
The constants in this Table show considerable irregularity ; perhaps some 
differ from 0 by an amount greater than can be legitimately attributed to the 
probable errors of random sampling. But from the table as a whole, with most of 
its values insignificant in comparison with their probable errors and both negative 
and positive in sign, only one conclusion is possible — that there is no correlation 
between the number of bracts and fertility. 
Hibiscus manihot, L. 
This was one of the species represented in considerable numbers in my cultures 
at the Station for Experimental Evolution in the summer of 1908. The data are 
given for the relationships for bracts and number of ovules per locule in Tables XIV 
and XV. These give : 
M b = 5173 + -011, <r h = -802 + "008, V h = 15*503, 
M 0 = 17-419 + -014, (t 0 = 2-312 + -010, V 0 = 13.272, 
M s = 12 029 ± 027, a s = 4-276 ± "019, V s = 35-549, 
r b0 = - -1067 + -0138, rjE r = 6-73, 
r bs = - -0733 ± -0139, r/E r = 5 31. 
The statistical significance of the constants cannot close our eyes to the fact 
that they are negative in sign and of a very low numerical order. 
