632 
IMPROVEMENT OF AGRICULTURAL HORSES. 
v’hich they certainly would not willingly submit to so long 
as any kind of food could be obtained. In this manner, 
without any sudden or violent change, the digestive organs^ 
as well as the external form, might be supposed to experi¬ 
ence remarkable alterations ; the strongest animals would first 
obtain an easy prey, then would come war between the 
strongest survivors for the mastery; the victors would inevit¬ 
ably be parents to the greatest number of progeny as well as 
the strongest in physical constitution ; and thus, by natural 
selection and changed external circumstances, herbivorous or 
herb-eating animals might possibly beconm carnivorous or 
flesh-feeders, in every respect most opposite to their proge¬ 
nitors. 
In this hypothetical example there is nothing new or ori¬ 
ginal, nor does it assume more than may be fairly granted 
without doing violence to the knowledge we already possess 
of the changes that have taken place to a minor extent under 
favorable circumstances. It is not probable that the condi¬ 
tions assumed will ever occur; but the position, however con¬ 
jectural, tends to illustrate what is meant by the influence of 
the conditions of existence and natural selection in the pro¬ 
duction of new varieties. Allowing the influence of natural 
circumstances to be so decided, we are led to inquire into the 
comparative effects of artificial changes, or the voluntary and 
intentional modification of the conditions of existence, when 
natural agencies are insufficient to effect what we require, 
or may even tend to the opposite results. Isolated instances 
are on record showing how horses have been gradually edu¬ 
cated to feed on flesh. Dogs, we know, naturally carnivorous, 
often live mainly on herbaceous food. Stock-breeders are 
well aware that a peculiar excellence or particular defect is 
likely to be perpetuated in the progeny; but all these illus¬ 
trations are barren, because unconnected; they concern the 
life of an individual, and consequently produce little or no 
permanent result. Artificial modifications depend uj)on 
individual judgment or caprice. Conditions intentionally 
changed are usually temporary in duration; but in nature 
the variations affecting the climate, the products, and conse¬ 
quently the animal life of a country, are slowly developed and 
constant, and their influence is beyond all comparison more 
decided and extensive. Notwithstanding that natural 
changes cannot be adequate^ represented by any intentional 
modifications established for the purpose of inducing such 
alterations of animal structures and qualities as we may wish 
to effect, there is much to be said in favour of the svstem of 
artificial selection; not merely the application of a single 
