and rapidity of the stream are most fit. There is a tree upon the bank,
and food and materials for his work in the vicinity. If a man should
attempt to build a beaver's dam, he would abstractly consider all these
elements of fitness. The outward manifestations of the quality of
abstraction are equally observable in either case. But we must not
hastily conclude, because the beaver in one instance acts in a manner
apparently reasonable, that he has any reason of his own; for, when we
come to study the habits of this animal, we find that he displays all
the characteristics of the instinctive principle. If animals are
endowed with instincts which apparently act so much like reason in the
ordinary course of their operations, we should not at once conclude
that there is any need of endowing them with a modicum of reason to
account for their deviations from this course, which do not outwardly
resemble the acts of reason any more strongly. And besides, it is said,
that, if we refer the variations to an intelligent principle, we must
refer the ordinary conduct to the same principle. To use an old
illustration,--if a bird is reasonable and intelligent, when, on
perceiving the swollen waters of the stream approach her half-finished
nest, she builds higher up the bank, she was intelligent while making
her first nest, and was always intelligent; for how otherwise, it is
asked, could she know when to lay down instinct and take up reason?

Instinct aims at certain definite ends; but these ends cannot always be
reached by the same means, especially when places and circumstances are
not the same. Accommodation is necessary, or it could not always
produce the effects for which it is intended. Would the instinct of the
spider be complete, if, after it has guided her to spin a web so neat
and trim and regular, it did not also lead her to repair her broken
snare, when the cords have been sundered by the struggles of some
powerful captive? But this pliancy of the spider's instinct is no more