A. Barrington and K. Pearson 
U7 
_o 
■© 
O 
o 
O 
TABLE III. 
Distribution to Sires and Dams of % Offspring. 
Coat- Colour of Sire. 
Offspring 
Ked 
Bed, 
Little White 
Red 
and White 
Roan 
White 
Totals 
Red 
R. 
R. L. W.... 
R. W. ... 
Ro. 
W. 
81 
13 
e 
2 
13 
8 
5 
3 
3 
4 
58 
7 
11 
76 
2 

13 
157 
31 
20 
95 
2 
Group Total 
102 
21 
15 
154 
14 
5 
5 
28 
13 
305 
Red, 
Little 
White 
R. 
R. L. W.,.. 
R. W. ... 
Ro. 
W. 
12 
4 
2 
1 
6 
5 
3 

1 

4 
32 
14 
10 
34 
Group Total 
19 
14 
1 
52 
4 
90 
Red 
and 
White 
R. 
R. L. W. ... 
R. W. ... 
Ro. 
W. 
17 
4 
18 
1 
2 
3 
8 
3 
3 
9 
1 
10 
7 
11 
41 
— 
9 
32 
14 
46 
55 
Group Total 
40 
16 
13 
69 
9 
147 
Roan 
R. 
R. L. W. ... 
R. VV. ... 
Rn 
IVO. ... 
w. 
61 
14 
10 
i o 
12 
4 
6 
1 7 
8 
3 
5 
1 / 
41 
13 
22 
1 "iC, 
1 »5U 
31 
7 
4 
122 
34 
43 
35 
Group Total 
163 
39 
33 
243 
11 
489 
White 
R. 
R. L. W.... 
R. W. ... 
Ro. 
W. 
2 
19 
3 
1 
5 
1 
15 
7 
3 
3 
4 
42 
10 
Group Total 
21 
3 
6 
23 
56 
Grand Total 
345 
93 
G8 
541 
40 
1087 
somewhat overweighted groups of red and roan — amounting to nearly i and \ of 
the total offspring — to deal with. These large sub-groups are likely to reduce the 
mean square contingency coefficients, and we have no means of really testing how 
far we have already approached a limiting value of the mean square contingency 
coefficient. 
In dealing with the material by the 4-fold table method, we considered that 
the best division we could make was between beasts with some whole red and 
