494 ORREN LLOYD-JONES AND F. A. HAYS 
TABLE 8 
Relation between size of litter and number of previous services 
| 
NUMBER OF SERVICES 
SIZE OF LITTERS 
Ist 5th 10th 15th 20th Total 
1 1 1 2 1 5 
2 1 1 2 4 8 
3 2, 1 4 7 
4 2 2 
5 3 3 2 1 1 10 
6 7 6 3 16 
Uf 6 2 4 1 2 15 
8 5 3 4 2 1 15 
9 1 4 2 7 
10 3 1 2, 1 | a 
11 1 1 1 3 
12 1 1 
26 20 20 13 17 96 
Correlation coefficient = — 0.2207 + 0.06526. 
Table 8 shows the complete distribution of the litters on the 
basis of size arranged as a correlation table. The coefficient of 
correlation calculated from this table is — 0.2297 — 0.0653, a 
figure which, though over three times its probable error, is yet 
of rather questionable significance. An inspection of the table, 
however, shows up a marked preponderance of small litters in 
the 20th service. If-it is true that the critical number of serv- 
ices, so far as decreasing the number in the litter is concerned, 
is between the 15th and the 20th service, then of course the above 
method of determining the correlation between frequency of 
copulation and number in litter is not valid, for in this case the 
20th, as one group, should be compared with all others as another 
group, and not be considered as one of five coordinate groups. 
In table 9 the first four columns of table 8 are combined and the 
percentage of the total litters in each group which are of a given 
size calculated. 
The preponderance of small litters in the 20th-service group is 
thus revealed. If we condense the data still further into a four- 
fold classification, we find that in the 1 to 15 group 25.32 per cent 
