BRITISH WARBLERS
in securing the nest to the supports. At the same time this method in the case of the Marsh Warbler is subject to a variation which is referred to by Mr. Warde Fowler as basket-handles, the nest having thereby the appearance of being slung from the supports, since the dead grass which actually forms the sling is slightly above the top of the nest; but it is by no means always adopted. The nest is frequently placed in a fork from which rise four or five perpendicular stems, and although it would clearly be secure without any actual attachment to the stems, yet the bird takes infinite pains to weave the grass round the supports after the manner adopted by the Reed Warbler. Now some of these special features in the construction of the nest are of considerable importance when we bear in mind the nature of the environment and the peculiarly close relationship of the bird to the Reed Warbler. In my description of the latter species I gave my reasons for believing that the depth of the nest was a factor of no small importance for the welfare of the species, since thereby the young—not the eggs as is sometimes suggested—were prevented from being precipitated into the water. I further stated my ground for the belief that we need have no hesitation in attributing the development of this peculiarity to natural selection, nor in regarding the regularity in this type of structure as the outcome of the same principle. As to whether the depth is to ensure the safety of the young or of the eggs opinions may differ, but no one can well dispute the fact that one or the other must be the cause of the development, and for our present purpose that is sufficient. What then does an examination of the nests of the Marsh Warbler reveal? It reveals all the peculiarities of those of the Reed Warbler—complete adaptations to a special environment—but with this difference that the constancy of type has been replaced by variation, the adaptive characteristics of the structure being in a condition of instability. But when we here speak of the adaptive characteristics, we manifestly make use of a wrong term, since they are not adapted to the
46
