The few insect groups peculiar to these islands will be noted when we deal with the entomology of Madagascar.

Extinct fauna of the Mascarene Islands and Madagascar.-Before quitting the vertebrate groups, we must notice the remarkable birds which have become extinct in these islands little more than a century ago. The most celebrated is the dodo of the Mauritius (Didus ineptus), but an allied genus, Pezophaps, inhabited Rodriguez, and of both of these almost perfect skeletons have been recovered. Other species probably existed in Bourbon. Remains of two genera of flightless rails have also been found, Aphanapteryx and Erythromachus; and even a heron (Ardea megacephala) which was short-winged and seldom flew; while in Madagascar there lived a gigantic Struthious bird, the Epyornis. Some further details as to these extinct forms will be found under the respective families, Dididæ, Rallidæ, and Æpyornithidæ, in the fourth part of this work; and their bearing on the past history of the region will be adverted to in the latter part of this chapter. Dr. Günther has recently distinguished five species of fossil tortoises from Mauritius and Rodriguez, all of them quite different from the living species of Aldabra.

Insects. The butterflies of Madagascar are not so remarkable as some other orders of insects. There seems to be only one peculiar genus, Heteropsis (Satyrida). The other genera are African, Leptoneura being confined to Madagascar and South Africa. There are some fine Papilios of uncommon forms. The most interesting lepidopterous insect, however, is the fine diurnal moth (Urania), as all the other species of the genus inhabit tropical America and the West Indian Islands.

The Coleoptera have been better collected, and exhibit some very remarkable affinities. There is but one peculiar genus of Cicindelidæ, Pogonostoma, which is allied to the South American genus, Ctenostoma. Another genus, Peridexia, is common to Madagascar and South America. None of the important African genera are represented, except Eurymorpha; while Meglaomma is common to Madagascar and the Oriental region.

In the Carabidae we have somewhat similar phenomena on a

wider scale. Such large and important African genera as Polyhirma and Anthia, are absent; but there are four genera in common with South Africa, and two with West Africa; while three others are as much Oriental as African. One genus, Distrigus, is wholly Oriental; and another, Homalosoma, Australian. Colpodes, well developed in Bourbon and Mauritius, is Oriental and South American. Of the peculiar genera, Sphærostylis has South American affinities; Microchila, Oriental; the others being related to widely distributed genera.

The Lucanidæ are few in number, and all have African affinities. Madagascar is very rich in Cetoniidæ, and possesses 20 peculiar genera. Bothrorhina, and three other genera belonging to the Ichnostoma group, have wholly African relations. Doryscelis and Chromoptila are no less clearly allied to Oriental genera. A series of eight peculiar genera belong to the Schizorhinidæ, a family the bulk of which are Australian, while there are only a few African forms. The remaining genera appear to have African affinities, but few of the peculiarly African genera are represented. Glyciphana is characteristic of the Oriental region.

The Buprestidæ of Madagascar consist mainly of one large and peculiar genus, Polybothris, allied to the almost cosmopolite Psiloptera. Most of the other genera are both Ethiopian and Oriental; but Polycesta is mainly South American, and the remarkable and isolated genus Sponsor is confined to the Mauritius with a species in Celebes and New Guinea.

The Longicorns are numerous and interesting, there being no less than 24 peculiar genera. Two of the genera of Prionidæ are very isolated, while a third, Closterus, belongs to a group which is Malayan and American.

Of the Cerambycidæ, Philematium ranges to Africa and the West Indies; Leptocera is only found eastward in Ceylon and the New Hebrides; while Euporus is African. Of the peculiar genera, 2 are of African type; 3 belong to the Leptura group, which are mostly Palearctic and Oriental, with a few in South Africa; while Philocalocera is allied to a South American genus.

Among the Lamiidæ there are several wide-ranging and 7

African genera; but Coptops is Oriental, and the Oriental Praonetha occurs in the Comoro Islands. Among the peculiar genera several have African affinities, but Tropidema belongs to a group which is Oriental and Australian; Oopsis is found also in the Pacific Islands; Mythergates, Sulemus, and Coedomaa, are allied to Malayan and American genera.

General Remarks on the Insect-fauna of Madagascar.-Taking the insects as a whole, we find the remarkable result that their affinities are largely Oriental, Australian, and South American: while the African element is represented chiefly by special South African or West African forms, rather than by such as are widely spread over the Ethiopian region.1 In some families as Cetoniidæ and Lamiida-the African element appears to preponderate; in others as Cicindelida-the South American affinity seems strongest; in Carabidæ, perhaps the Oriental; while in Buprestidæ and Cerambycidæ the African and foreign elements seem nearly balanced. We must not impute too much importance to these foreign alliances among insects, because we find examples of them in every country on the globe. The reason they are so much more pronounced in Madagascar may be, that during long periods of time this island has served as a refuge for groups that have been dying out on the great continents; and that, owing to the numerous deficiencies of a somewhat similar kind in the series of vertebrata in Australia and South America, the same groups have often been able to maintain themselves in all these countries as well as in Madagascar. It must be remembered too, that these peculiarities in the Malagasy and Mascarene insect-fauna are but exaggerations of a like phenomenon on the mainland. Africa also has numerous affinities with South America, with the Malay countries, and with Australia; but they do not bear anything like so large a proportion to the whole fauna, and do not, therefore, attract so much attention. The special conditions of existence and the long-continued isolation of Madagascar, will account for much of this difference; and it will evidently not be necessary

1 There are also some special resemblances between the plants of Madagascar and South Africa, according to Dr. Kirk.

to introduce, as some writers are disposed to do, a special land connection or near approach between Madagascar and all these countries, independently of Africa; except perhaps in the case of the Malay Islands, as will be discussed further on.

Land-shells.-Madagascar and the adjacent islands are all rich in land-shells. The genera of Helicidæ are Vitrina, Helix, Achatina, Columna (peculiar to Madagascar and West Africa), Buliminus, Cionella (chiefly Oriental and South American, but not African), Pupa, Streptaxis, and Succinea. Among the Operculata we have Truncatella (widely scattered, but not African); Cyclotus (South American, Oriental, and South African); Cyclophorus (mostly Oriental, with a few South African); Leptopoma (Oriental); Megalomastoma (Malayan and South American); Lithidion (peculiar to Madagascar, Socotra, and South-West Arabia); Otopoma (with the same range, but extending to West India and New Ireland); Cyclostomus (widely spread but not African); and Omphalotropis (wholly Oriental and Australian). We thus find the same general features reproduced in the landshells as in the insects, and the same remarks will to a great extent apply to both. The classification of the former is, however, by no means so satisfactory, and we have no extensive and accurate general catalogues of shells, like those of Lepidoptera and Coleoptera, which have furnished us with such valuable materials for the comparison of the several faunas.

On the probable Past History of the Ethiopian Region.

Perhaps none of the great zoological regions of the earth present us with problems of greater difficulty or higher interest than the Ethiopian. We find in it the evidence of several distinct and successive faunas, now intermingled; and it is very difficult, with our present imperfect knowledge, to form an adequate conception of how and when the several changes occurred. There are, however, a few points which seem sufficiently clear, and these afford us a secure foundation in our endeavour to comprehend the rest.

Let us then consider what are the main facts we have to account for.-1. In Continental Africa, more especially in the south

and west, we find, along with much that is peculiar, a number of genera showing a decided Oriental, and others with an equally strong South American affinity; this latter more particularly showing itself among reptiles and insects. 2. All over Africa, but more especially in the east, we have abundance of large ungulates and felines-antelopes, giraffes, buffaloes, elephants, and rhinoceroses, with lions, leopards, and hyænas, all of types now or recently found in India and Western Asia. 3. But we also have to note the absence of a number of groups which abound in the abovenamed countries, such as deer, bears, moles, and true pigs; while camels and goats-characteristic of the desert regions just to the north of the Ethiopian-are equally wanting. 4. There is a wonderful unity of type and want of speciality in the vast area of our first sub-region extending from Senegal across to the east coast, and southward to the Zambezi; while West Africa and South Africa each abound in peculiar types. 5. We have the extraordinary fauna of Madagascar to account for, with its evident main derivation from Africa, yet wanting all the larger and higher African forms; its resemblances to Malaya and to South America; and its wonderful assemblage of altogether peculiar types.

Here we find a secure starting-point, for we are sure that Madagascar must have been separated from Africa before the assemblage of large animals enumerated above, had entered it. Now, it is a suggestive fact, that all these belong to types which abounded in Europe and India about the Miocene period. It is also known, from the prevalence of Tertiary deposits over the Sahara and much of Arabia, Persia, and Northern India, that during early Tertiary times a continuous sea from the Bay of Bengal to the British Isles completely cut off all land communication between Central and Southern Africa on the one side, and the great continent of the Eastern hemisphere on the other. When Africa was thus isolated, its fauna probably had a character somewhat analogous to that of South America at the same period. Most of the higher types of mammalian life were absent, while lemurs, Edentates, and Insectivora took their place. At this period Madagascar was no doubt united with Africa,