In the Sphingidea the same general features recur in a less marked degree, the Neotropical being the richest region; but here we have one family (Castniidæ) which appears to be confined to the two southern regions,—the Australian and Neotropical.
The distribution of the genera affords us some facts of special interest, which must be briefly noticed. There are several genera typically characteristic of the North Temperate regions which have a few species widely scattered on mountains, or in the temperate parts of the Southern Hemisphere. Chili possesses representatives of four of these genera—Argynnis, Lycæna, Colias, and Deilephila; and this has been thought by some naturalists to be of such importance as to outweigh the purely Neotropical character of a large portion of the Chilian fauna, and to render it advisable to join it on, as an outlying portion of a great North Temperate zoological region. But when we remember that Argynnis occurs also in Java, and Lycæna in New Zealand, while Colias ranges to Southern Africa, Malabar, and the Sandwich Islands, we can hardly admit the argument to be a sound one. For a fuller discussion of this question see Vol. II., pp. [43]-[47]. The remarkable fact of the existence of the otherwise purely Neotropical genus, Urania, in Madagascar is even more striking, supported as it is by the Antillean, Solenedon, belonging to a family of Mammalia otherwise confined to Madagascar, and by one or two Coleopterous genera, to be noticed farther on as common to the two countries. Our view as to the true explanation of this and analogous phenomena will be found at Vol. I., p. 284.
The division of the Castniidæ (a family almost confined to the Tropics), between the Neotropical and Australian regions, is also a very curious and important phenomenon, because it seems to point to a more remote connection between the two countries than that indicated by the resemblance between the productions of South Temperate America with those of Australia and New Zealand; but we have already shown that the facts may be explained in another way. (See Vol. I., pp. 398 and 404).
The division of the Malay Archipelago between the Oriental and Australian regions is clearly marked in the Lepidoptera, and it is very curious that it should be so, for in this, if in any group of animals, we should expect an almost complete fusion to have been effected. Lepidoptera fly readily across wide tracts of sea, and there is absolutely no climatal difference to interfere with their free migration from island to island. Yet we find no less than 10 genera abundant in the Indo-Malayan sub-region which never cross the narrow seas to the east of them; 6 others which only pass to Celebes; and 2 more which have extended from Java along the closely connected line of islands eastwards to Timor. On the other side, we find 5 strictly Austro-Malayan genera, and 2 others which have a single representative in Java. The following is a list of these genera:—
Indo-Malayan Genera:—Amathusia, Thaumantis, Tanæcia, Eurytela, Ilerda, Zemeros, Taxila, Aphneus, Prioneris, Dercas, Clerome, Adolias, Apatura, Limenitis, Iolaus, Leptocircus, (the last six reach Celebes); Discophora, Thestias; (the last two reach Timor.)
Austro-Malayan Genera:—Hamadryas, Hypocista, Mynes, Dicallaneura, Elodina, Hyades, Prothoë (the last two reach Java).
The most characteristic groups, which range over the whole Archipelago and give it a homogeneous character, are the various genera of Danaidæ, the genus Elymnias, and Amblypodia with a few other Lycænidæ. These are all abundant and conspicuous groups, but they are nevertheless exceptions to the general rule of limitation to one or other of the regions. The cause of this phenomenon is probably to be found in the limitation of the larvæ of many Lepidoptera to definite species, genera, and families of plants; and we shall perhaps find, when the subject is carefully investigated, that the groups which range over the whole Archipelago feed on genera of plants which have an equally wide range, while those which are limited to one region or the other, have food-plants belonging to genera which are similarly limited. It is known that the vegetation of the two regions differs largely in a botanical sense, although its general aspect is almost identical; and this may be the reason why the proportion of wide-ranging genera is greater among such insects as feed upon dead wood, than among those which derive their support from the juices of the living foliage. This subject will be again discussed under the various families of Coleoptera, and it will be well to bear in mind the striking facts of generic limitation which have been here brought forward.
Fossil Butterflies, apparently of existing genera, occur in the Miocene and Eocene formations, and an extinct form in the Lower Oolite; but these cannot be held to give any adequate idea of the antiquity of so highly specialised a group, which, in all probability, dates back to Palæozoic times, since one of the Bombycidæ,—a group almost as highly-organised—has been discovered in the coal formation of Belgium. (See Vol. I. p. 168.)
Order—COLEOPTERA.
Geodephaga, or Carnivorous Ground Beetles.