a pair has been destroyed. The survivor of any pair might have the home attachment and by bringing a fresh mate create an attachment which would be passed on from mate to mate indefinitely. Again it must not be overlooked that certain sites present advantages to particular species which must be evident to all in search of those advantages; it by no means follows that the occupiers of a nesting site are in any way related, except specifically, to those which occupied it in previous years.
The answer to the argument that birds do not suck fresh nesting places and thus extend their distributional area, is evident when we consider those species which, at the present time, are extending their range. Within the last few years, for instance, the turtle dove and tufted duck have begun to nest regularly in many parts of England in which they were entirely unknown twenty or thirty years ago. The Starling has spread and in some parts is spreading still, and many other similar cases might be cited.
In this manner migration, as we. know it today, may have originated, and as Mr P. A. Taverner expressed it, "however instinctive their habit may now be, there must have been a time when migrations were intelligent movements, intended to escape some danger or secure some advantage" (51).
