Karl Pearson 
383 
average rat. If w stands for weight, / for number of young, then we have the 
following results : 
1003 Poona Eats 
Mean Weight w =113-92 
Mean Number of Young /= 5 '49 
(T„= 24-86 
crf= 1-68 
n = 
•1670 
•1733 
lOOG Belgaum Eats 
128-84 
5-38 
25-35 
1-29 
-1498 
■1634 
As the above results indicate, the regression curves of fertility on weight are 
almost linear, and the average number of offspring increase at an almost uniform 
rate from the dwarf to the giant rats. This is well illustrated on the accompany- 
ing diagram. It will be seen at once that Captain Lloyd's view is not supported 
Weight in grs. of Mother. 
Regression Curves of Number of Young on Mother's Weight in Eats. 
AA Poona Eats. BB Belgaum Eats. 
by his own data. There is a very sensible relationship between the weight of the 
rat and the number of its offspring. But it is not a relation in which the type rat 
is the most fertile. In other words the case seems at first sight precisely one in 
which the effect of reproductive selection would cause, without regard to natural 
selection, a progressive increase in the size of rats. But is this inference really 
justified ? It seems so very important to determine whether any such cases of 
genetic selection actually occur that we may be pardoned for hoping that further 
information may be forthcoming. It is conceivable that in certain stages of 
pregnancy even the size of the litter might have some effect on the weight, but 
this is hardly likely to produce all the difference noted. Another important point 
49—2 
