82 On the Inheritance of the Sex-ratio 
the possibly slight tendency of that sibship to femininity or masculinity. The 
exclusion of even one individual can produce very sensible effects in the case 
of the small families which occur with human beings, although it is of less 
significance in the case of horses or many other mammals. It will thus be seen 
that the vertical and horizontal means and standard deviations of our Tables I. 
and II. could not be expected to be in accordance. To test whether this peculiarity 
has any influence on the result, Table III. was formed. This gives the correlation 
between sibships in the filial generation and all parental sibships. It would seem 
from this table that there really exists a marked difference between the distribu- 
tion of the sex-ratio in the two generations, families which tend wholly or largely 
to one sex being much under-represented in the mated population. 
TABLE III. 
General Correlation, Parental and Filial Sihshij)s. Whitney Family. 
Sex-Ratio of Filial Sibships. 
1 
1 
^1 
1 
1 
1 
1 
1 
i 
1 
1 
-1-00 
Totals 
o 
9> 
©i 
eo 
00 
Cj 
■00— -05 
1-5 
0-5 
2 
1 
1 
0-5 
0-5 
7 
■05— -15 
2 
2 
4 
5 
5 
3 
3 
5 
2 
31 
■15— ^25 
1 
13 
13-5 
14-5 
23-5 
19-5 
12 
15 
1 
6-5 
119-5 
•25— -85 
5 
2 
15-5 
18 
33-5 
48-5 
26-5 
24-5 
22-5 
2 
7-5 
205-5 
■35— ■Jt5 
12 
6 
32-5 
33-5 
67 
93 
74 
46 
33 
12 
11 
420 
•]/)— -55 
7 
3 
39-5 
44-5 
54 
87 
78 
49 
47 
6 
12 
427 
•55— •0-5 
11 
4 
39 
51 
69 
82 
73 
50 
56 
4 
17 
456 
•05— -75 
5 
6-5 
18-75 
27-25 
47 
55-5 
43 
40-5 
26-5 
4-5 
11-5 
286 
■75— -55 
]-5 
10-25 
13-75 
23 
26-5 
22 
13-5 
11-5 
0-5 
7-5 
130 
■85— -95 
1 
1 
5 
9 
8 
14 
12 
13-5 
2-5 
3 
2 
71 
■95-1-00 
0-5 
4-5 
7 
8 
6 
6-5 
4-5 
2 
5 
44 
Totals 
43 
2.5 177-5 
219-5 
330 
444 
358 
259 
224 
35 
82 
2197 
■ rH 
o 
_o 
In Table IV. another method of investigating the problem is considered, based 
also on different data. A sort of mid-parent was used. A joint-parental sibship 
was formed by combining mother's and father's sibships together and taking the 
sex-ratio for the total array. The result is precisely the same as in the previous 
cases. 
In Table V. we have a wholly different method of approaching the problem. 
Here the sex-ratio of the produce of a thoroughbred mare — often reaching 14 to 
18 foals — has been determined and correlated with the produce of one of her 
fillies selected at random. In this case the produce is usually due to a very 
considerable number of sires, or forms a half-sibship, some individuals, however, 
being possibly whole siblings. This method enables us to determine whether the 
individual has any tendency to produce one or other sex which is inherited by 
