THE EVOLUTION OF INSTINCT 135 
and in the case of many instincts it is not difficult to picture 
to one’s self how natural selection may have brought them to 
their present state of perfection. To find, as we can in many 
cases, instincts in various grades of development in allied 
species of animals, while it shows the general course of their 
evolution, tells us little of their method of evolution. That 
instincts aré variable, that they are often imperfect, that their 
general course of development has been along adaptive lines, 
and that, as Darwin emphasized, they are always primarily 
of value to the species possessing them and only incidentally 
of service to others, are facts indicative of the potency 
of natural selection in the evolution of instinctive be- 
havior. 
Evidence for the potency of selection is furnished by the 
study of the striking modifications of instinct which have 
taken place in animals under domestication. While only 
rarely has the attempt been made to modify instincts along 
particular lines by selection, yet they have doubtless been 
modified as the incidental result of selection on other lines. 
A sort of unconscious selection has probably played a part 
in modifying especially the emotional characteristics of 
animals. Ugly and vicious dispositions in dogs, for instance, 
would tend to be eliminated, and the qualities of affection, 
fidelity and other traits which commend the animals to the 
good graces of their keepers would be fostered. The 
useful instincts of the pointer and the setter, however they 
may have made their beginning, have certainly been devel- 
oped to a considerable degree by continued selection. The 
curious instincts of tumbling and pouting which, as Whitman 
has shown, have their basis in traits of behavior found in 
pigeons in general, have been developed by the efforts of 
fanciers to an almost monstrous degree. The Indian sub- 
breed of tumblers which has been bred for at least two-hundred 
