Page:BirdWatchingSelous.djvu/281

Rh tree-fluttering. Now, as it appears to me, though it might be easier for a bird to creep up a tree by going round it, it could more easily flutter up it perpendicularly, in the way I have described, and, if so, we can understand a bird that is only in process of becoming a tree-creeper, commencing, as it were, at the most advanced end. For it would first have fluttered up perpendicularly, then have both crept and fluttered so, and finally, when it could creep without fluttering, it would do so at first on the old fluttering lines. Then it might begin to adopt the spiral method, but as the effort required became less and less, and structural modification—as seen, for example, in the shape and stiffness of the tail-feathers of the tree-creeper—came to its assistance, this would cease to be a help, and become a habit merely, and when once a habit has lost its rationale, it is in the way of being broken, even in good society. Thus the perpendicular ascents of the tree-creeper may be the final stage in a long process, and the return in ease to what was before done in toil.

The tree-creeper is assisted in its climbing by the stiff, pointed feathers of the tail, which act as a prop, and also by its small size, which may possibly have been partly gained by natural selection. The great green woodpecker is possessed of the first of these advantages, but not of the second, and it is, I believe, the case that he much more adheres to the spiral mode of ascent than does the tree-creeper, who, as it seems to me, has almost discarded it. It would be interesting, therefore, to observe if the smaller spotted