312 Breeding Experiments with Foxgloves 
The general results may be summarized in the form of a table : 
Relationship) 
No. of 
mid- 
parents 
No. of 
off- 
spring 
No. of 
W .1 O 
tandard 
deviation 
of mid- 
parent 
Standard 
deviation 
of 
offspring 
Coefficient 
of 
correlation 
Mid-parent and offspring (including 
15 
176 
1027 
10-509 
8-500 
-563 
self-fertilized families) 
Mid-parent and offspring (excluding 
11 
127 
773 
6-610 
7-790 
-545 
seli-fertilized families) 
Male parent and offspring (including 
15 
176 
1027 
16-733 
8-595 
-431 
self-fertilized families) 
Male parent and offspring (excluding 
11 
127 
773 
15-249 
7-790 
•320 
self-fertilized families) 
With the small number of parents available the constants are bound to be 
erratic, and the results can only be regarded as giving a rough indication of the 
strength of inheritance. The general tendency is for a low parental correlation: 
the mean of the two paternal correlations is -38. The average of some 52 correla- 
tion coefficients of parental heredity obtained from a variety of different animal 
species with various characters has been given by Prof. Karl Pearson as -45, while 
the mean of the parental coefficients for a series of characters in the Shirley Poppy 
was -33. 
4. General Colouration of Corolla. 
We have seen that the general colouration of the corolla is due to coloured 
sap in the epidermal cells of both the inside and outside of the corolla. The 
intensity of this colouration was estimated, as already described, by a colour-scale 
obtained by viewing by transmitted light varying depths of a standard tinted 
solution. There is discontinuity in the colour of the flowers of the plants. The 
flowers can be divided into (1) purple and (2) not purple (white). For this sub- 
division a purple flower must be defined as one in which the general surface of the 
corolla is coloured. Tlie colour may be quite pale or very dark. A white flower 
is one in which the general surface is white, but there may be a few extremely 
faint blushes of purple on the corolla, especially on the inner surface. 
Such definitions may seem empirical and unsatisfactory, but nevertheless 
it was found in practice to work quite easily owing to the fact that exceedingly 
faint general colouration of the whole corolla did not occur. There was a clear 
gap in the series between faintly "purple flowers" and the "white flowers." In 
the whole series the palest purple flower was represented by a depth of 15 mm. of 
coloured fluid and there was only one such plant. 
In the accompanying table the families are arranged in the order of the mid- 
parental colour — the mean of the colours of the male and female parents. 
