Karl Pearson and David Heron 
243 
including cases in which one brother was an infant. The second diagram (p. 240) 
worked out by Mr Yule's pseudo-ranks indicates also the heredity within the 
brown group when blues are excluded, but it obviously distorts the whole system 
and gives the spurious appearance of far less correlation. 
Another manner of illustrating the idle character of Mr Yule's assertion that 
the true parental correlation is about 4 may be obtained from a comparison 
of the following tables : 
A 
1+2 
3 
4 
5 + 6 
7 + 8 
Totals 
1+2 
162 
87 
48 
16 
22 
335 
3 
102 
78 
52 
19 
33 
284 
4 
41 
37 
28 
11 
20 
137 
5 + 6 
j 27 
28 
22 
9 
19 
105 
7 + 8 
26 
34 
30 
14 
35 
139 
Totals 
358 
264 
180 
69 
129 
1000 
B 
1+2 
s 
4 
5 + 6 
7 + 8 
Totals 
1+2 
194 
70 
41 
9 
21 
335 
83 
124 
41 
13 
23 
284 
4 
25 
34 
55 
11 
12 
137 
5 + 6 
27 
12 
19 
24 
23 
105 
7+8 
29 
24 
24 
12 
50 
139 
Totals 
358 
264 
180 
69 
129 
1000 
C 
1+2 
3 
4 
5 + 6 
7 + 8 
Totals 
1+2 
193 
84 
38 
10 
10 
335 
3 
100 
85 
54 
19 
26 
284 
4 
34 
39 
31 
12 
21 
137 
5 + 6 
18 
28 
25 
11 
23 
105 
7 + 8 
13 
28 
32 
17 
49 
139 
Totals 
358 
264 
180 
69 
129 
1000 
Now the changes that will convert A into B are given by the scheme : 
B-A 
1+2 
O 
4 
5 + 6 
7 + 8 
1+2 
O 
4 
5 + 6 
7 + 8 
+32 
-19 
-16 
0 
+ 3 
-17 
+ 46 
- 3 
-16 
-10 
- 7 
-11 
+ 27 
- 3 
- 6 
- 7 
- 6 
0 
+ 15 
- 2 
- 1 
-10 
- 8 
+ 4 
+ 15 
That is to say to obtain B from A we must take between | and £ of the 
material of A and transfer it to the diagonal cells from those cells away from the 
diagonal, the only failure of this rule is the + 3 in the cell 1 + 2:7+8. There 
