384 Danvlnism, Biometrji and Some Recent Biology. I. 
is the question of a.<;e. Were the bigger rats on the average the older, and do the 
older rats have on the average larger litters ? If anything of this kind takes place 
in rats, we should not be comparing the fertility of large and small rats, but the 
fertility of rats in diiferent stages of development, which is a very diverse point. 
I am inclined to think that the real relation between number of young and weight 
of rat, although overlooked by Captain Lloyd, actually has this simple origin. It 
is suggested, but not demonstrated by the relation which is manifest in mice 
between size of litter and number of litter. Thus in 700 litters of mice recorded 
by the late Professor Weldon, I find : 
Number Average size 
First litter 307 5-46 
Second „ 254 5 -.57 
Third „ 139 5-7G 
It will be seen that the size of the litter increased uniformly with the increasing 
age of the mouse, and there is little doubt that the weight of the mouse also 
increased with age. Only a direct experiment in which age and number of litter 
were taken into account could settle the point. In Captain Lloyd's experimental 
weighing, there was no record of these characters. I have small doubt that 
the very sensible correlation between weight and number of young shown by his 
data is only a secondary result of the relations between age and weight, and again 
between age and number of young. It is not evidence of a real correlation between 
fertility and a somatic character, and thus has no bearing on genetic selection, and 
does not oppose the general principle that fertility is not highly correlated with 
inherited characters. 
(9) The reader may ask : Why have I considered in the same paper such 
diverse work as that of Jennings, Hanel and Lloyd ? The reason lies in this : 
We have been told recently in an ex cathedra fashion by a distinguished biologist* 
that : 
" Of the so-called investigations of heredity pursued by extensions of Galton's 
non-analytical method and promoted by Professor Pearson and the English 
Biometrical School it is now scarcely necessary to speak A preliminary 
acquaintance with the natural history of heredity and variation was sufficient to 
throw doubt on the foundations of these elaborate researches. To those who 
hereafter may study this episode in the history of biological science, it will appear 
inexplicable that work so unsound in construction should have been respectfully 
received by the scientific world." 
It seems therefore, however regrettable, needful to re-emphasise the point from 
which this Journal started more than eight years ago. Biology requires the help 
of a more exact logic than it appears to possess, above all of an adequate statistical 
theory to enable it to interpret its observations and test its own theories. I have 
* W. Bateson : McndeVn Priiicijilcs of Ilcri'dity, Cambridge, 1909, pp. 6—7. Luckily the doctrine 
of infallibility is not yet accepted in the world of science. 
