characteristic of the whole group, often with metallic reflections; while soft greens, and sometimes metallic greens, occur in the forest regions of tropical Africa and Asia, but rarely anything approaching to crests or other developments of plumage.

But as soon as we reach the Moluccas and New Guinea we find a new type of coloration appearing in both groups. Among the lories we find vivid red and crimson, sometimes with a remnant of green on the wings and tail, but often covering the whole plumage, varied with bands or patches of equally vivid blue or yellow, while the red sometimes deepens into a blackish-purple. Among the cockatoos we have pure whites and deep black, with highly developed crests, often of great beauty, so that in these two families we seem to depart altogether from the usual parrot type of coloration.

Still more remarkably is this the case with the pigeons. In the extensive genus of small fruit-pigeons (Ptilonopus) the usual ground colour is a clear soft green, variegated by blue, purple, or yellow breasts, and crowns of equally brilliant colours. Besides these, we have larger fruit-pigeons almost wholly cream white, while the very large ground pigeons of New Guinea possess flat vertical crests, which are unique in this order of birds. The wonderfully brilliant golden green Nicobar pigeon is probably a native of the AustroMalayan islands, and may have been carried westward by Malay traders, and have become naturalized on a few small islands.

These peculiarities of distribution and coloration in two such very diverse groups of birds interested me greatly, and I endeavoured to explain them in accordance with the laws of natural selection. In the paper on Pigeons (published in The Ibis of October, 1865) I suggest that the excessive development of both these groups in the Moluccas and the Papuan islands has been due primarily to the total absence of arboreal, carnivorous, carnivorous, or egg-destroying mammals, especially of the whole monkey tribe, which in all other tropical forest regions are exceedingly abundant, and are

very destructive to eggs and young birds. I also point out that there are here comparatively few other groups of fruiteating birds like the extensive families of chatterers, tanagers, and toucans of America, or the barbets, bulbuls, finches, starlings, and many other groups of India and Africa, while in all those countries monkeys, squirrels, and other arboreal mammals consume enormous quantities of fruits. It is clear, therefore, that in the Australian region, especially in the forest-clad portions of it, both parrots and pigeons have fewer enemies and fewer competitors for food than in other tropical regions, the result being that they have had freer scope for development in various directions leading to the production of forms and styles of colouring unknown elsewhere. It is also very suggestive that the only other country in which black pigeons and black parrots are found is Madagascar, an island where also there are neither monkeys nor squirrels, and where arboreal carnivora or fruiteating birds are very scarce. The satisfactory solution of these curious facts of distribution gave me very great pleasure, and I am not aware that the conclusions I arrived at have been seriously objected to.

Before I had written these two papers I had begun the study of my collection of butterflies, and in March, 1864, I read before the Linnean Society a rather elaborate paper on "The Malayan Papilionidæ, as illustrating the Theory of Natural Selection." This was published in the Society's Transactions, vol. xxv., and was illustrated by fine coloured plates drawn by Professor Westwood. I reprinted the introductory portion of this paper in the first edition of my "Contributions to the Theory of Natural Selection" in 1870, but in later editions it was omitted, as being rather too technical for general readers, and not easily followed without the coloured plates. I will therefore give a short outline of its purport here.

I may state for the information of non-entomological readers that the Papilionidæ form one of the most extensive families of butterflies, and from their large size, elegant forms,,

and splendid colours were considered by all the older writers to be the princes of the whole lepidopterous order. They are usually known by the English term "Swallowtailed butterflies," because the only British species, as well as a great many of the tropical forms, have the hind wings tailed. They are pretty uniformly distributed over all the warmer regions, but are especially abundant in the tropical forests, of which they form one of the greatest ornaments. In coloration they are wonderfully varied. The ground colour is very frequently black, on which appear bands, spots, or large patches of brilliant colours-pale or golden yellow, rich crimsons or gorgeous metallic blues and greens, which colours sometimes spread over nearly the whole wing surface. Some are thickly speckled with golden green dots and adorned with large patches of intense metallic green or azure blue, others are simply black and white in a great variety of patterns many very striking and beautiful, while others again have crimson or golden patches, which when viewed at certain angles change to quite different opalescent hues, unsurpassed by the rarest gems.

But it is not this grand development of size and colour that constitutes the attraction of these insects to the student of evolution, but the fact that they exhibit, in a remarkable degree, almost every kind of variation, as well as some of the most beautiful examples of polymorphism and of mimicry. Besides these features, the family presents us with examples of differences of size, form, and colour, characteristic of certain localities, which are among the most singular and mysterious phenomena known to naturalists. A short statement of the nature of these phenomena will be useful to show the great interest of the subject.

In all parts of the world there are certain insects which, from a disagreeable smell or taste, are rarely attacked or devoured by enemies. Such groups are said to be "protected," and they almost always have distinctive and conspicuous colours. In the Malay Archipelago there are several groups of butterflies which have this kind of protection; and one group is coloured black, with rich blue glosses and

VOL. I.

2 D

ornamented with white bands or spots. These are excessively abundant, and, having few enemies, they fly slowly. Now there are also several different kinds of papilios, which in colour are so exactly like these, that when on the wing they cannot be distinguished, although they frequent the same places and are often found intermingled. Other protected butterflies are of paler colours with dark stripes, and these are also closely imitated by other papilios. Altogether there are about fifteen species which thus closely resemble protected butterflies externally, although in structure and transformations they have no affinity with them. In some cases both sexes possess this resemblance, or "mimicry," as it is termed, but most frequently it is the female only that is thus modified, especially when she lays her eggs on low-growing plants; while the male, whose flight is stronger and can take care of himself, does not possess it, and is often so different from his mate as to have been considered a distinct species.

This leads us to the phenomenon of dimorphism and polymorphism, in which the females of one species present two or three different forms. Several such cases occur in the Malay Archipelago, in which there are two distinct kinds of females, sometimes even three, to a single male, which differs from either of them. In one case four females are known to one male, though only two of them appear to occur in one locality. These have been almost always described as distinct species, but observation has now proved them to be one, and it has further been noticed that each of the females, which are very unlike the male, resembles more or less closely some "protected " species. It has also been proved by experimental breeding that eggs laid by any one of these females are capable of producing butterflies of all the different forms, which in the few cases recorded are quite distinct from each other, without intermediate gradations.

The local diversities of form are illustrated by outline figures (as regards two species of papilio from Celebes) in my "Malay Archipelago" (p. 216), and similar local peculiarities of colour, both in papilio and other groups, are described in my "Natural Selection and Tropical Nature"

(pp. 384, 385), while extraordinary development of size in Amboyna is referred to at p. 307 of my "Malay Archipelago."

This brief outline of the paper will, perhaps, enable my readers to understand the intense interest I felt in working out all these strange phenomena, and showing how they could almost all be explained by that law of "Natural Selection" which Darwin had discovered many years before, and which I had also been so fortunate as to hit upon.

The only other groups of insects upon which I did any systematic work were the families of Pieridæ among butterflies and Cetoniidæ among beetles. Of the former family, which contains our common whites, our brimstone and orange tip butterflies, I gave a list of all known from the Indian and Australian regions, describing fifty new species, mostly from my own collection. This paper is in the "Transactions of the Entomological Society for 1867," and is illustrated by four coloured plates. The other paper, which is contained in the same volume, is a catalogue of the Cetoniidæ (or Rosechafers, named after our common species) of the Malay Archipelago, in which I described seventy new species, the majority of which were collected by myself, and it is illustrated by four coloured plates, beautifully executed by the late Mr. E. W. Robinson, in which thirty-two of the species are figured. These two papers, filling about 200 pages of the society's "Transactions," occupied me for several months, and if I had not had wider and more varied interests— evolution, distribution, physical geography, anthropology, the glacial period, geological time, sociology, and several others-I might have spent the rest of my life upon similar work, for which my own collection afforded ample materials, and thus settled down into a regular " species-monger." For even in this humble occupation there is a great fascination; constant difficulties are encountered in unravelling the mistakes of previous describers who have had imperfect materials, while the detection of those minute differences, which often serve to distinguish allied species, and the many curious modifications of structure which characterize genera or their subdivisions, become intensely interesting, especially