Page:BirdWatchingSelous.djvu/361

Rh part, or intended to be part, of the process of preening themselves. This close pressing seems to be a pleasure in itself, independently of the result of warmth, for sometimes they will come unstuck, as it were, and move a little away from each other along the twig, in order to press and squeeze again. For a little, then, their tails may be separate, but soon they rejoin, and, the heads being now quiet—for they are going to sleep—and tucked closely in amongst the feathers of the breast, their outlines, never very salient, are entirely lost, and the two birds have become one perfectly globular one, without a head and with a long tail. Thus two of these long-tailed tits have returned again to roost in the same place, but the other pair do not come to the bush."

It is interesting to watch sand-martins building their nests, or, rather, excavating the tunnels in which they will afterwards be built. To see one enter one of these whilst it is yet but a few inches long, and then to see the dust powdering out at the aperture, as from the mouth of an ensconced cannon, is pretty. The sand is scratched out backwards with the feet, but the bird also uses its bill as a pickaxe, often making a series of rapid little blows with it, almost like a woodpecker, the wings, which quite cover the body, quivering at the same time. Both sexes work at the hole, and both often fly together to it, one remaining clinging at the edge whilst the other scratches out the sand from inside. I have seen one sitting just in the embrasure, quietly regarding the outer world and, thus, impeding the entrance of his partner, who at last squeezed by him with great difficulty. Sometimes three or four will descend upon