J. A. Harris 
311 
In the present paper the study of variation has been a means to an end. The 
standard deviations have been necessary in the calculation of the coefficients of 
correlation, and the consideration of the interrelationships between different parts 
of the organism as indicated by the coefficient of correlation has been the main 
purpose of the paper. 
TABLE 3. 
Cumparison of Constants for 1906 and 1907 Collections of Sanguinaria. 
Character 
Average 
and 
Probable Error 
Standard 
Deviation and 
Probable Error 
Coefficient 
of 
Variation 
Length of Fruit, 1906 
1907 
Difterence 
Ovules per Placenta, 1906 
1907 
Difference 
Total Ovules per Fruit, 1906 
1907 
Difterence 
Seeds per Placenta, 1906 
1907 
Difference 
Total Seeds per Fruit, 1906 
1907 
Difference 
Aborted Ovules per Placenta, 1906 
1907 
Difference 
Total Aborted Ovules per Fruit, 1906 ... 
„ 1907 ... 
Difference 
39-703±-135 
51-865 ±-268 
6- 307 + -095 
7- 948 ±-190 
15-89 ± -23 
15-33 ± -37 
+ 12-162 ±-300 
14-952+ -070 
12-803 ±-096 
+ 1-641 ±-212 
4-613+ -049 
4-038 ± -068 
- -.56 ± -43 
30- 86+ -51 
31- 64+ -83 
-2-149 ±-119 
29-904 + -193 
25-605 ± -265 
- -575 ±-084 
9-055 + -137 
7-850 ±-187 
+ -78 ± -98 
30-28+ -50 
30-66 ± -80 
-4-299 ±-328 
11 -527 + -059 
9-756 ±-095 
-1-205 ±-232 
3-925 + -042 
3-986 ± -067 
+ -38 ± -94 
34-05+ -57 
40-85 ±1-13 
-1-771±-112 
23-055 + -161 
19-513 ±-258 
+ -061 ± -079 
7-531 + -114 
7-660 ±-183 
+ 6-80 ±1-26 
32-66+ -54 
39-26 ±107 
+ 6-60 ±1-20 
82-12 + 2-69 
84-87 ±3-17 
- 3-542 ± -304 
3-424 + -042 
3-046 ± 062 
+ -129±-215 
2-812 + -030 
2-585 ±044 
- -378 ±-075 
6-849 + -110 
6-093 ±-161 
- -227 ±-053 
5-161 + -078 
4-785 ±-114 
+ 2-75±4-15 
75-36 + 2-43 
78-54 ±2-73 
- -756 ±-195 
- -376 ±-141 
+ 3-18 ±3-65 
The constants for the two years considered individually require no separate 
discussion. The means are not of interest except in comparison with the same 
constants for other races of Sanguinaria. The standard deviations and coefficients 
of variation are of about the order generally found for plant characters. 
The comparison table shows that some of the differences are several times 
their probable errors, and hence clearly significant. For length of fruit this is 
quite obviously due to the somewhat greater maturity of the 1907 series. Here 
I do not care to discuss in detail the interpretation to be assigned to these 
40—2 
