Abstract
The following paper forms part of an investigation of the relation between the variations of animals and the conditions under which they live. It appears to me necessary that any investigation of this problem should be begun by the examination of cases in which difference in environment is known to exist, and that variations should then be sought for among the forms of life subjected to these conditions. If by this examination any variations can be shown to occur regularly with the change of conditions, or in any way in proportion to their intensity, it is so far evidence that there is a relation of cause and effect between them. By thus first approaching the question from the point of view of the conditions, many difficulties are obviated which occur in any attempt which begins by ascertaining the variations in the animal, in the hope of afterwards finding an environmental change to which they may be traced. Such attempts to trace back variations to some environmental cause have often been made, and have, in general, been unsuccessful. In the case of species which have varied in isolated situations not apparently differing from each other, the failure to find points of environmental difference has been held to be evidence that the variations in question did not arise from such causes at all. This appears likely, and is probably true of the variations in question; but it must be borne in mind that the fact that no palpable difference can be found between the conditions in the several localities is no proof that they do not exist. While these differences in condition are usually evasive and hard to detect, it is best to begin to investigate their relation to variations in animals by selecting cases in which the change in conditions is unequivocal, and proceed from this starting point to seek for correlated variation in the forms of life subjected to them.

This publication has 0 references indexed in Scilit: