418 THE CAUSES OF THE 3 XI 
variety which had arisen in this manner—the 
variation might be perpetuated, as far as we can 
see, indefinitely. 
The next question, and it is an important one 
for us, is this: Is there any limit to the amount 
of variation from the primitive stock which can 
be produced by this process of selective breeding ? 
In considering this question, it will be useful to 
class the characteristics, in respect of which 
organic beings vary, under two heads: we may 
consider structural characteristics, and we may 
consider physiological characteristics. 
In the first place, as regards structural charac- 
teristics, I endeavoured to show you, by the 
skeletons which I had upon the table, and by 
reference to a great many well-ascertained facts, 
that the different breeds of Pigeons, the Carriers, 
Pouters, and Tumblers, might vary in any of their 
internal and important structural characters to a 
very great degree ; not only might there be changes 
in the proportions of the skull, and the characters 
of the feet and beaks, and so on; but that there — 
might be an absolute difference in the number of 
the vertebree of the back, as in the sacral vertebree 
of the Pouter; and so great is the extent of the 
variation in these and similar characters that I 
pointed out to you, by reference to the skeletons 
and the diagrams, that these extreme varieties 
may absolutely differ more from: one another in 
their structural characters than do what naturalists 
