299 STUDY OF NATURAL HISTORY. 
while we see and admit it in her grander features: 
besides, it is not to be supposed that such forms as 
we have elsewhere cited (150.), are scattered indis- 
criminately in their respective groups, without being 
accompanied by others, equally representing each 
other, and therefore implying, in the strongest 
possible manner, the existence of strict uniformity. 
We may, then, safely conclude, that if the number 
of our genera in a sub-family disagrees with the 
number of divisions in our genera, the fault lies 
with ourselves. We must again retrace our steps, 
perhaps abandon altogether the number first assumed 
as definite, and adopt some other more in unison 
with the facts before us. If, on the contrary, we 
can, in these new and higher groups, demonstrate 
the same prevalence of a determinate number, the 
strength of our theory is doubled. It has been well 
observed *, that, ‘* whatever error we commit in a 
single determination, it is highly improbable we 
should always err in the same way; so that, when we 
come to take an average of a great number of de- 
terminations (unless there be some constant cause 
which gives a bias one way or the other), we cannot 
fail, at length, to attain a very near approximation 
to truth ; and, even allowing a bias, to come much 
nearer to it than can fairly be expected from any 
single observation, liable to be influenced by the 
same bias.’ This useful and valuable property of 
the average of a great many observations —that it 
brings us nearer to the truth, —that is, to the deter- 
mination of a prevalent number,—than any single 
* Hersch. Discourse, p. 215. 
ti 
