416 
Cooperatwe Investifiations on Plants 
for plant characters, and group without apparent hesitation individuals into one or 
other categor}'. 
(e) The minimum limit to parental correlation has the average value "33. In 
the best class of characters, however, it lies between '4 and "48, and we believe if 
the disturbing factors could be fully allowed for it would be found to lie between 
•45 and "50, its value as determined for other types of life. 
(/) The Diinimum limit to fraternal correlation has the average value '24. 
This allows for no selection of maternal plants. The real value is undoubtedly 
higher, and is quite consistent with half-sibling resemblance as found for horses 
and cattle, which it closely approaches in the case of colour, where the influence of 
selection and of variation in the plant appears to be least. 
(g) Although very considerable differences exist in the values found for the 
correlations at different centres and for different characters, we do not think these 
differences are sufficient to justify any assertion that the intensity of heredity 
varies with character or with environment. They point rather to the great 
difficulty of dealing with plant life. What is urgently needed is a biometric farm 
where secular experiments could be conducted under nearly uniform conditions ; 
there a continuous recnrd of all flowers of each plant could be made by one and 
the same observer using a standard qualitative scale of groups. 
APPENDIX. CORRELATION TABLES* 
I. RESEMBLANCE OF DAUGHTER AND MOTHER PLANTS. 
DAUGHTER PLANT GROWN FROM SEED TAKEN FROM 
APICAL CAPSULE OF MOTHER PLANT. 
Ph g 
P o 
'I 
A. Farkstone and Crewe. Apical Flowers. 
Parkstone Mother Plant. Number of Stigmata. 
10 
11 
12 
13 
15 
16 
17 
18 
19 
Totals 
9 
10 
11 
12 
IS 
U 
15 
16 
17 
18 
19 
1 
1 
4 
6 
4 
'I 
1 
4 
1 
1 
4 
6 
13 
16 
15 
6 
2 
1 
6 
5 
20 
23 
18 
13 
1 
1 
1 
3 
12 
26 
29 
22 
15 
10 
2 
4 
7 
13 
14 
13 
11 
6 
1 
1 
2 
2 
2 
3 
2 
1 
3 
3 
2 
3 
2 
1 
2 
22 
37 
78 
98 
78 
52 
28 
5 
2 
2 
2 
Totals 
2 
34 
62 
88 
122 
69 
13 
11 
0 
404 
* There is not absolute uniformity in the classification categories of all these tables. Thus one 
observer may have used "wrinkled" entirely and never the extreme "frilled," while a second may 
have placed the few " veined whites " in a category denoting white with very little piuk. 
