ACQUIREMENT VERSUS SELECTION 455 
perfect adaptations to its strange environment. We are 
warranted in believing that these plants could not live there 
unless they had some such adaptations, but just where ex- 
treme adaptations are most necessary there is least compe- 
tition to account for them. But may not the environment 
operate selectively nevertheless? Perhaps, but even so it is 
not through the struggle of numerous competing individuals: 
any one of them that is favorably situated and is just able to 
stand the worst of the drought has as good a chance to in- 
crease and multiply as the toughest of all. It may also be 
pointed out that the most marked features of desert plants— 
such for example as extensive root system, succulent or 
greatly reduced foliage, nairy or thickened skin for checking 
loss of water, and an armament of thorns for defense against 
animals—are the ones most easily accounted for as originat- 
ing from direct effect of the environment or individual re- 
sponse; while as regards the defensive armament, which is 
often cited as a most perfect adaptation, it must be said, 
that grave doubts are raised as to its being of much use in 
the very cases where it is most forbidding, since few if any 
large browsing animals are found in those localities where 
the most highly developed armaments occur. In such cases 
any selective influence is hard to imagine. 
Naturalists have found many cases in which a selective 
influence, supposing it to exist, would have no opportunity 
to act upon small fluctuating variations in the Darwinian 
way. The climbing habit of clematis will serve to show 
what is meant. The coiling of the leaf-stalks by unequal 
growth stimulated by contact with a support, is an undoubt- 
edly useful power; but this power, it would seem, can be of 
service only after it is developed; a very slight curving of the 
stalk could not anchor a leaf. Therefore we cannot reason- 
ably suppose that the power to coil was developed from slight 
curving rendered more pronounced in successive generations 
by natural selection. Here it is the first step that counts, 
and it is just this step that the theory of natural selection 
is often unable to take. 
Artificial selection, not being restricted in its operation 
to variations of use to the plant, does not offer so close an 
