and accordingly the makers often put on a row of useless buttons or imitation laces, because habit rendered the appearance of them necessary to us. It is universally admitted that the habits of children and of savages give us the best clue to the habits and mode of thought of animals; and every one must have observed how children at first imitate the actions of their elders, without any regard to the use or applicability of the particular acts. So, in savages, many customs peculiar to each tribe are handed down from father to son merely by the force of habit, and are continued long after the purpose which they originally served has ceased to exist. With these and a hundred similar facts everywhere around us, we may fairly impute much of what we cannot understand in the details of Bird-Architecture to an analogous cause. If we do not do so, we must assume, either that birds are guided in every action by pure reason to a far greater extent than men are, or that an infallible instinct leads them to the same result by a different road. The first theory has never, that I am aware of, been maintained by any author, and I have already shown that the second, although constantly assumed, has never been proved, and that a large body of facts is entirely opposed to it. One of my critics has, indeed, maintained that I admit "instinct" under the term "hereditary habit;" but the whole course of my argument shows that I do not do so. Hereditary habit is, indeed, the same as instinct when the term is applied to some simple action dependent upon a

peculiarity of structure which is hereditary; as when the descendants of tumbler pigeons tumble, and the descendants of pouter pigeons pout. In the present case, however, I compare it strictly to the hereditary, or more properly, persistent or imitative, habits of savages, in building their houses as their fathers did. Imitation is a lower faculty than invention. Children and savages imitate before they originate; birds, as well as all other animals, do the same.

The preceding observations are intended to show, that the exact mode of nidification of each species of bird is probably the result of a variety of causes, which have been continually inducing changes in accordance with changed organic or physical conditions. The most important of these causes seem to be, in the first place, the structure of the species, and, in the second, its environment or conditions of existence. Now we know, that every one of the characters or conditions. included under these two heads is variable. We have seen that, on the large scale, the main features of the nest built by each group of birds, bears a relation to the organic structure of that group, and we have, therefore, a right to infer, that as structure varies, the nest will vary also in some particular corresponding to the changes of structure. We have seen also, that birds change the position, the form, and the construction of their nest, whenever the available materials or the available situations, vary naturally or have been altered by man; and we have, therefore, a right to infer that similar changes have taken place,

when, by a natural process, external conditions have become in any way permanently altered. We must remember, however, that all these factors are very stable during many generations, and only change at a rate commensurate with those of the great physical features of the earth as revealed to us by geology; and we may, therefore, infer that the form and construction of nests, which we have shown to be dependent on them, are equally stable. If, therefore, we find less important and more easily modified characters than these, so correlated with peculiarities of nidification as to indicate that one is probably the cause of the other, we shall be justified in concluding that these variable characters are dependent on the mode of nidification, and not that the form of the nest has been determined by these variable characters. Such a correlation I am now about to point out.

Classification of Nests.

For the purpose of this inquiry it is necessary to group nests into two great classes, without any regard to their most obvious differences or resemblances, but solely looking to the fact of whether the contents (eggs, young, or sitting bird) are hidden or exposed to view. In the first class we place all those in which the eggs and young are completely hidden, no matter whether this is effected by an elaborate covered structure, or by depositing the eggs in some hollow tree or burrow underground. In the second, we group all in which the eggs, young, and sitting bird are

exposed to view, no matter whether there is the most beautifully formed nest, or none at all. Kingfishers, which build almost invariably in holes in banks; Woodpeckers and Parrots, which build in hollow trees; the Icterida of America, which all make beautiful covered and suspended nests; and our own Wren, which builds a domed nest, are examples of the former; while our Thrushes, Warblers, and Finches, as well as the Crowshrikes, Chatterers, and Tanagers of the tropics, together with all Raptorial birds and Pigeons, and a vast number of others in every part of the world, all adopt the latter mode of building.

It will be seen that this division of birds according to their nidification, bears little relation to the character of the nest itself. It is a functional not a structural classification. The most rude and the most perfect specimens of bird-architecture are to be found in both sections. It has, however, a certain relation to natural affinities, for large groups of birds, undoubtedly allied, fall into one or the other division exclusively. The species of a genus or of a family are rarely divided between the two primary classes, although they are frequently divided between the two very distinct modes of nidification that exist in the first of them.

All the Scansorial or climbing, and most of the Fissirostral or wide-gaped birds, for example, build concealed nests; and, in the latter group, the two families which build open nests, the Swifts and the Goat-suckers, are undoubtedly very widely separated from the other families with which they are asso

ciated in our classifications.

The Tits vary much in

their mode of nesting, some making open nests concealed in a hole, while others build domed or even pendulous covered nests, but they all come under the same class. Starlings vary in a similar way. The talking Mynahs, like our own starlings, build in holes, the glossy starlings of the East (of the genus Calornis) form a hanging covered nest, while the genus Sturnopastor builds in a hollow tree. One of the most striking cases in which one family of birds is divided between the two classes, is that of the Finches; for while most of the European species build exposed nests, many of the Australian finches make them dome-shaped.

Sexual differences of Colour in Birds.

Turning now from the nests to the creatures who make them, let us consider birds themselves from a somewhat unusual point of view, and form them into separate groups, according as both sexes, or the males only, are adorned with conspicuous colours.

The sexual differences of colour and plumage in birds are very remarkable, and have attracted much attention; and, in the case of polygamous birds, have been well explained by Mr. Darwin's principle of sexual selection. We can, to a great extent, understand how male Pheasants and Grouse have acquired their more brilliant plumage and greater size, by the continual rivalry of the males both in strength and beauty; but this theory does not throw any light on the causes which have made the female Toucan, Bee-eater, Parro