J. A. Harris 
309 
collections. But when the constants are calculated for the repeated organs it 
seems logical to calculate the probable errors on the basis of the actual number 
of such organs, and this was done for the constants calculated for the characters 
of the individual placentae. 
III. Problems Considered. 
1. Type and Variability of the Characters Investifjated in the 
Two Collections. 
The fundamental physical constants are given in Ta.bles 1 and 2. It has 
seemed best to table each collection separately and to put them side by side 
for comparison in a third table. The distributions upon which these constants 
are based will be found as the totals of the correlation tables. The differences of 
the constants to be compared are given with their probable errors in Table 3. 
TABLE 1. 
Constants for 190G Collection of Sanguinaria. 
Average 
Standard 
Coefficient 
Character 
and 
Deviation and 
of 
Probable Error 
Probable Error 
Variation 
Length of Stalk in 5 cm. Units * ... 
17-090+ 058 
2-736+ -041 
16-01 
Length of Fruit in 3 mm. Unit.sf ... 
39-703+ -134 
6-307 ± -095 
15-88 
Ovule.s on First Placenta ... 
14-980 + -097 
4-569+ -069 
30-50 
Ovules on Second Placenta 
14-924 ±-099 
4-658 + -070 
31-21 
Ovules on First and on Second Placenta ... 
14-952 + -070 
4-614 + -049 
30-86 
Total Ovules per Fruit 
29-904 + -193 
9-054 + -136 
30-28 
Seeds on First Placenta 
11 -584 + -083 
3-872 + -058 
33-42 
Seeds on Second Placenta ... 
11-471 ±-085 
3-977 + -060 
34-67 
Seeds on First and on Second Placenta ... 
11-527 + -059 
3-925 + -042 
34-05 
Total Seeds per Fruit 
23-055 + -161 
7-531 + -114 
32-66 
Aborted Ovules on First Placenta... 
3-396 + 060 
2-827 + -043 
83-25 
Aborted Ovules on Second Placenta 
3-453 ±-060 
2-802 + -042 
81-14 
Aborted Ovules on First and on Second Placenta 
3-424+ -042 
2-81 2 + -030 
82-12 
Total Aborted Ovules per Fruit ... 
6-849+ -110 
5-162 + -078 
75-36 
* Constants tabled in units of 1 cm. 
t Constants tabled in units of 1 mm. 
Another constant which is of much service in studies of fertility is not given 
in either of these tables. This is the ratio of the total seeds developing to the 
total ovules formed. 
Several years ago Pearsonj defined fecundity as the ratio of the number of 
ofifspring actually produced to those which might have come into existence 
under the circumstances. His definition was formulated for use in a study of 
I Phil. Trans. Roy. Soc. Lund. A, Vol, cxcii. p. 261, 1899. 
Biometrika vii 
40 
