The tree-creeper, when it flits from tree to tree, generally does so in a downward direction. If trees were of a uniform height, and if the bird usually ascended to the top, or nearly to the top, of each one in succession, one could see the rationale, or even the necessity, of this practice, for the tree-creeper does not—at least not usually—descend the trunk. But in a wood, the top of one tree may not represent half the height of another, and, moreover, a tree will often be abandoned by the bird when it has reached only a moderate height, or is still quite near to the ground; and it is not so easy to see how, under these circumstances, the above-mentioned habit should have arisen. But, now, if the forerunners of the tree-creeper had been birds accustomed to hop about on the ground, and to peer and pry amongst the projecting roots of trees, and if they had, from these, gradually ascended the trunk, getting back to them at first quite soon, but making longer and longer and more and more accustomed excursions, then we can understand how this habit might have become—as one may say—rooted, so as to continue after there was no longer any particular advantage in it. Now, however, it is beginning to weaken. I have on several occasions—which I duly noted down at the time—seen a tree-creeper fly from one tree to another, upon which it clung, in an upward direction. I have little doubt that what is now still a habit will come to be a preference merely, and that, in time, even this will cease to be discernible, and the bird be guided simply by circumstances.
It is said that the tree-creeper never descends the tree it is on, and, also, that it generally proceeds in a spiral direction, by which, I suppose, is meant that
