The Geographical Distribution of Animals/Chapter 13



The Australian is the great insular region of the earth. As a whole it is one of the best marked, and has even been considered to be equal in zoological value to all the rest of the globe; but its separate portions are very heterogeneous, and their limits sometimes ill-defined. Its central and most important masses consist of Australia and New Guinea, in which the main features of the region are fully developed. To the north-west it extends to Celebes, in which a large proportion of the Australian characters have disappeared, while Oriental types are mingled with them to such an extent that it is rather difficult to determine where to locate it. To the south-east it includes New Zealand, which is in some respects so peculiar, that it has even been proposed to constitute it a distinct region. On the east it embraces the whole of Oceania to the Marquesas and Sandwich Islands, whose very scanty and often peculiar fauna, must be affiliated to the general Australian type.

Australia is the largest tract of land in the region, being several times more extensive than all the other islands combined, and it is here that the greatest variety of peculiar types have been developed. This island-continent, being situated in the track of the southern desert zone, and having no central mountains to condense the vapours from the surrounding ocean, has a large portion of its interior so parched up and barren as to be almost destitute of animal life. The most extensive tract of fertile and well-watered country is on the east and south-east, where a fine range of mountains reaches, in the Colony of Victoria, the limits of perpetual snow. The west coast also possesses mountains of moderate height, but the climate is very dry and hot. The northern portion is entirely tropical, yet it nowhere presents the luxuriance of vegetation characteristic of the great island of New Guinea immediately to the north of it. Taken as a whole, Australia is characterized by an arid climate and a deficiency of water; conditions which have probably long prevailed, and under which its very peculiar fauna and flora have been developed. This fact will account for some of the marked differences between it and the adjacent sub-regions of New Guinea and the Moluccas, where the climate is moist, and the vegetation luxuriant; and these divergent features must never be lost sight of, in comparing the different portions of the Australian region. In Tasmania alone, which is however, essentially a detached portion of Australia, a more uniform and moister climate prevails; but it is too small a tract of land, and has been too recently severed from its parent mass to have developed a special fauna.

The Austro-Malay sub-region (of which New Guinea is the central and typical mass) is strikingly contrasted with Australia, being subjected to purely equatorial conditions,—a high, but uniform temperature, excessive moisture, and a luxuriant forest vegetation, exactly similar in general features to that which clothes the Indo-Malay Islands, and the other portions of the great equatorial forest zone. Such a climate and vegetation, being the necessary result of its geographical position, must have existed from remote geological epochs with but little change, and must therefore have profoundly affected all the forms of life which have been developed under their influence. Around New Guinea as a centre are grouped a number of important islands, more or less closely agreeing with it in physical features, climate, vegetation, and forms of life. In most immediate connection we place the Aru Islands, Mysol and Waigiou, with Jobie and the other Islands in Geelvinck Bay, all of which are connected with it by shallow seas; they possess one of its most characteristic groups, the Birds of Paradise, and have no doubt only recently (in a geological sense) been separated from it. In the next rank come the large islands of the Moluccas on the west, and the range terminating in the Solomon Islands on the east, both of which groups possess a clearly Papuan fauna, although deficient in many of the most remarkable Papuan types.

All these islands agree closely with New Guinea itself in being very mountainous, and covered with a luxuriant forest vegetation; but to the south-west we find a set of islands extending from Timor to Lombock, which agree more nearly with Australia, both in climate and vegetation; being arid and abounding in eucalypti, acacias, and thickets of thorny shrubs. These, like the Moluccas, are surrounded by deep sea, and it is doubtful whether they have either of them been actually connected with New Guinea or Australia in recent geological times; but the general features of their zoology oblige us to unite all these islands with New Guinea as forming the Austro-Malay sub-division of the Australian region. Still further west however, we have the large island of Celebes, whose position is very difficult to determine. It is mountainous, but has also extensive plains and low lands. Its climate is somewhat arid in the south, where the woods are often scattered and thorny, while in the north it is moister, and the forests are luxuriant. It is surrounded by deep seas, but also by coralline and volcanic islets, indicating former elevations and subsidences. Its fauna presents the most puzzling relations, showing affinities to Java, to the Philippines, to the Moluccas, to New Guinea, to continental India, and even to Africa; so that it is almost impossible to decide whether to place it in the Oriental or the Australian region. On the whole the preponderance of its relations appears to be with the latter, though it is undoubtedly very anomalous, and may, with almost as much propriety, be classed with the former. This will be better understood when we come to discuss its zoological peculiarities.

The next sub-region consists of the extensive series of islands scattered over the Pacific, the principal groups being the Sandwich Islands, the Marquesas and Society Islands, the Navigators', Friendly, and Fiji Islands. New Caledonia and the New Hebrides have rather an uncertain position, and it is difficult to decide whether to class them with the Austro-Malay Islands, the Pacific Islands, or Australia. The islands of the west Pacific, north of the equator, also probably come into this region, although the Ladrone Islands may belong to the Philippines; but as the fauna of all these small islets is very scanty, and very little known, they are not at present of much importance.

There remains the islands of New Zealand, with the surrounding small islands, as far as the Auckland, Chatham, and Norfolk Islands. These are situated in the south temperate forest-zone. They are mountainous, and have a moist, equable, and temperate climate. They are true oceanic islands, and the total absence of mammalia intimates that they have not been connected with Australia or any other continent in recent geological times. The general character of their zoology, no less than their botany, affiliates them however, to Australia as portions of the same zoological region.

General Zoological Characteristics of the Australian Region.—For the purpose of giving an idea of the very peculiar and striking features which characterise the Australian region, it will be as well at first to confine ourselves to the great central land masses of Australia and New Guinea, where those features are manifested in their greatest force and purity, leaving the various peculiarities and anomalies of the outlying islands to be dealt with subsequently.

Mammalia.—The Australian region is broadly distinguished from all the rest of the globe by the entire absence of all the orders of non-aquatic mammalia that abound in the Old World, except two—the winged bats (Chiroptera), and the equally cosmopolite rodents (Rodentia). Of these latter however, only one family is represented—the Muridæ—(comprising the rats and mice), and the Australian representatives of these are all of small or moderate size—a suggestive fact in appreciating the true character of the Australian fauna. In place of the Quadrumana, Carnivora, and Ungulates, which abound in endless variety in all the other regions under equally favourable conditions, Australia possesses two new orders (or perhaps sub-classes)—Marsupialia and Monotremata, found nowhere else on the globe except a single family of the former in America. The Marsupials are wonderfully developed in Australia, where they exist in the most diversified forms, adapted to different modes of life. Some are carnivorous, some herbivorous; some arboreal, others terrestrial. There are insect-eaters, root-gnawers, fruit-eaters, honey-eaters, leaf or grass-feeders. Some resemble wolves, others marmots, weasels, squirrels, flying squirrels, dormice or jerboas. They are classed in six distinct families, comprising about thirty genera, and subserve most of the purposes in the economy of nature, fulfilled in other parts of the world by very different groups; yet they all possess common peculiarities of structure and habits which show that they are members of one stock, and have no real affinity with the Old-World forms which they often outwardly resemble.

The other order, Monotremata, is only represented by two rare and very remarkable forms, Ornithorhynchus and Echidna, probably the descendants of some of those earlier developments of mammalian life which in every other part of the globe have long been extinct.

The bats of Australia all belong to Old-World genera and possess no features of special interest, a result of the wandering habits of these aerial mammals. The Rodents are more interesting. They are all more or less modified forms of mice or rats. Some belong to the widely distributed genus Mus, others to four allied genera, which may be all modifications of some common Old-World form. They spread all over Australia, and allied species occur in Celebes and the Papuan Islands, so that although not yet known from the Moluccas, there can be little doubt that some of them exist there.

Birds.—The typical Australian region, as above defined, is almost as well characterized by its birds, as by its mammalia; but in this case the deficiencies are less conspicuous, while the peculiar and characteristic families are numerous and important. The most marked deficiency as regards wide-spread families, is the total absence of Fringillidæ (true finches), Picidæ (woodpeckers), Vulturidæ (vultures), and Phasianidæ (pheasants). and among prevalent Oriental groups, Pycnonotidæ (bulbuls), Phyllornithidæ (green bulbuls), and Megalæmidæ (barbets) are families whose absence is significant. Nine families are peculiar to the region, or only just pass its limits in the case of single species. These are Paradiseidæ (paradise-birds), Meliphagidæ (honey-suckers), Menuridæ (lyre-birds), Atrichidæ (scrub-birds), Cacatuidæ (cockatoos), Platycercidæ (broad-tailed and grass-paroquets), Trichoglossidæ (brush-tongued paroquets), Megapodiidæ (mound-makers), and Casuariidæ (cassowaries). There are also eight very characteristic families, of which four,—Pachycephalidæ (thick-headed shrikes), Campephagidæ (caterpillar shrikes), Dicæidæ (flower-peckers), and Artamidæ (swallow-shrikes)—are feebly represented elsewhere, while the other four—Ploceidæ (weaver-finches), Alcedinidæ (kingfishers), Podargidæ (frog-mouths), and Columbidæ (pigeons)—although widely distributed, are here unusually abundant and varied, and (except in the case of the Ploceidæ) better represented in the Australian than in any other region. Of all these the Meliphagidæ (honeysuckers) are the most peculiarly and characteristically Australian. This family abounds in genera and species; it extends into every part of the region from Celebes and Lombock on the west, to the Sandwich Islands, Marquesas, and New Zealand on the east, while not a single species overpasses its limits, with the exception of one (Ptilotis limbata) which abounds in all the islands of the Timorese group, and has crossed the narrow strait from Lombock to Baly; but this can hardly be considered to impugn the otherwise striking fact of wide diffusion combined with strict limitation, which characterizes it. This family is the more important, because, like the Trichoglossidæ or brush-tongued paroquets, it seems to have been developed in co-ordination with that wealth of nectariferous flowering shrubs and trees which is one of the marked features of Australian vegetation. It probably originated in the extensive land-area of Australia itself, and thence spread into all the tributary islands, where it has become variously modified, yet always in such close adaptation to the other great features of the Australian fauna, that it seems unable to maintain itself when subject to the competition of the more varied forms of life in the Oriental region; to which, possessing great powers of flight, some species must occasionally have emigrated. Its presence or absence serves therefore to define and limit the Australian region with a precision hardly to be equalled in the case of any other region or any other family of birds.

The Trichoglossidæ, as already intimated, are another of these peculiarly organized Australian families,—parrots with an extensile brush-tipped tongue, adapted to extract the nectar and pollen from flowers. These are also rigidly confined to this region, but they do not range so completely over the whole of it, being absent from New Zealand (where however they are represented by a closely allied form Nestor), and from the Sandwich Islands. The Paradiseidæ (birds of paradise and allies) are another remarkable family, confined to the Papuan group of Islands, and the tropical parts of Australia. The Megapodiidæ (or mound-builders) are another most remarkable and anomalous group of birds, no doubt specially adapted to Australian conditions of existence. Their peculiarity consists in their laying enormous eggs (at considerable intervals of time) and burying them either in the loose hot sand of the beach above high-water mark, or in enormous mounds of leaves, sticks, earth, and refuse of all kinds, gathered together by the birds, whose feet and claws are enlarged and strengthened for the work. The warmth of this slightly fermenting mass hatches the eggs; when the young birds work their way out, and thenceforth take care of themselves, as they are able to run quickly, and even to fly short distances, as soon as they are hatched. This may perhaps be an adaptation to the peculiar condition of so large a portion of Australia, in respect to prolonged droughts and scanty water-supply, entailing a periodical scarcity of all kinds of food. In such a country the confinement of the parents to one spot during the long period of incubation would often lead to starvation, and the consequent death of the offspring. But the same birds with free power to roam about, might readily maintain themselves. This peculiar constitution and habit, which enabled the Megapodii to maintain an existence under the unfavourable conditions of their original habitat gives them a great advantage in the luxuriant islands of the Moluccas, to which they have spread. There they abound to a remarkable extent, and their eggs furnish a luxurious repast to the natives. They have also reached many of the smallest islets, and have spread beyond the limits of the region to the Philippines, and North-Western Borneo, as well as to the remote Nicobar Islands.

The Platycercidæ, or broad-tailed paroquets, are another wide-spread Australian group, of weak structure but gorgeously coloured, ranging from the Moluccas to New Zealand and the Society Islands, and very characteristic of the region, to which they are strictly confined. The Cockatoos have not quite so wide a range, being confined to the Austro-Malayan and Australian sub-regions, while one species extends into the Philippine Islands. The other two peculiar families are more restricted in their range, and will be noticed under the sub-regions to which they respectively belong.

Of the characteristic families, the Pachycephalidæ, or thick-headed shrikes, are especially Australian, ranging over all the region, except New Zealand; while only a single species has spread into the Oriental, and one of doubtful affinity to the Ethiopian region. The Artamidæ, or swallow-shrikes, are also almost wholly confined to the region, one species only extending to India. They range to the Fiji Islands on the east, but only to Tasmania on the south. These two families must be considered as really peculiar to Australia. The Podargidæ, or frog-mouths—large, thick-billed goat-suckers—are strange birds very characteristic of the Australian region, although they have representatives in the Oriental and Neotropical regions. Campephagidæ (caterpillar-shrikes) also abound, but they are fairly represented both in India and Africa. The Ploceidæ, or weaver-birds, are the finches of Australia, and present a variety of interesting and beautiful forms.

We now come to the kingfishers, a cosmopolitan family of birds, yet so largely developed in the Australian region as to deserve special notice. Two-thirds of all the genera are found here, and no less than 10 out of the 19 genera in the family are peculiar to the Australian region. Another of the universally distributed families which have their metropolis here, is that of the Columbidæ or pigeons. Three-fourths of the genera have representatives in the Australian region, while two-fifths of the whole are confined to it; and it possesses as many species of pigeons as any other two regions combined. It also possesses the most remarkable forms, as exemplified in the great crowned pigeons (Goura) and the hook-billed Didunculus, while the green fruit-pigeons (Ptilopus) are sometimes adorned with colours vying with those of the gayest parrots or chatterers. This enormous development of a family of birds so defenceless as the pigeons, whose rude nests expose their eggs and helpless young to continual danger, may perhaps be correlated, as I have suggested elsewhere (Ibis, 1865, p. 366), with the entire absence of monkeys, cats, lemurs, weasels, civets and other arboreal mammals, which prey on eggs and young birds. The very prevalent green colour of the upper part of their plumage, may be due to the need of concealment from their only enemies,—birds of prey; and this is rendered more probable by the fact that it is among the pigeons of the small islands of the Pacific (where hawks and their allies are exceedingly scarce) that we alone meet with species whose entire plumage is a rich and conspicuous yellow. Where the need of concealment is least, the brilliancy of colour has attained its maximum. We may therefore look upon the genus Ptilopus, with its fifty species whose typical coloration is green, with patches of bright blue, red, or yellow on the head and breast, as a special development suited to the tropical portion of the Australian region, to which it is almost wholly confined.

It will be seen from the sketch just given, that the ornithological features of the Australian region are almost as remarkable as those presented by its Mammalian fauna; and from the fuller development attained by the aërial class of birds, much more varied and interesting. None of the other regions of the earth can offer us so many families with special points of interest in structure, or habits, or general relations. The paradise-birds, the honeysuckers, the brush-tongued paroquets, the mound-builders, and the cassowaries—all strictly peculiar to the region—with such remarkable developments as we have indicated in the kingfishers and pigeons, place the Australian region in the first rank for the variety, singularity, and interest of its birds, and only second to South America as regards numbers and beauty.

Reptiles.—In Reptiles the peculiarity of the main Australian region is less marked, although the fauna is sufficiently distinct. There is no family of snakes confined to the region, but many peculiar genera of the families Pythonidæ and Elapidæ. About two-thirds of the Australian snakes belong to the latter family, and are poisonous; so that although the Crotalidæ and Viperidæ are absent, there are perhaps a larger proportion of poisonous to harmless snakes than in any other part of the world. According to Mr. Gerard Krefft the proportion varies considerably in the different colonies. In Victoria, New South Wales, and Queensland the proportion is about two to one; in West Australia three to one; and in South Australia six to one. In Tasmania there are only 3 species and all are poisonous. The number of species, as in other parts of the world, seems to increase with temperature. The 3 in Tasmania have increased to 12 in Victoria, 15 in South Australia and the same in West Australia; 31 in New South Wales, and 42 in sub-tropical Queensland.

The lizards of Australia have lately been catalogued by Dr. Günther in the concluding part of the "Voyage of the Erebus and Terror," issued in 1875. They belong to 8 families, 3 of which are peculiar; 57 genera of which 36 are peculiar; and about 140 species, all but 2 or 3 of which are peculiar. The scinks and geckoes form the great bulk of the Australian lizards, with a few Agamidæ, Gymnophthalmidæ, and Varanidæ. The three peculiar families are the Pygopodidæ, Aprasiidæ and Lialidæ; comprising only 4 genera and 7 species. The above all belong to Australia proper. Those of the other sub-regions are few in number and will be noticed under their respective localities. They will perhaps bring up the number of genera to 70. West and South Australia seem to offer much peculiarity in their lizards; these districts possessing 12 peculiar genera, while a much smaller number are confined to the East and South-East, or to the North.

Among the fresh-water turtles of the family Chelydidæ there are three peculiar genera—Chelodina, Chelemys, and Elseya, all from Australia.

Amphibia.—No tailed amphibians are known from the whole region, but no less than eleven of the families of tail-less Batrachians (toads and frogs) are known to inhabit some part or other of it. A peculiar family (Xenorhinidæ), consisting of a single species, is found in New Guinea; the true toads (Bufonidæ) are only represented by a single species of a peculiar genus in Australia, and by a Bufo in Celebes. Nine of the families are represented in Australia itself, and the following genera are peculiar to it:—Pseudophryne (Phryniscidæ), Pachybatrachus, and Chelydobatrachus (Engystomydæ); Helioporus (Alytidæ); Pelodryas and Chirodryas (Pelodryadæ); Notaden (Bufonidæ).

Fresh-water Fish.—There is only one peculiar family of fresh-water fishes in this region—the Gadopsidæ—represented by a single genus and species. The other species of Australia belong to the families Trachinidæ, Atherinidæ, Mugillidæ, Siluridæ, Homalopteræ, Haplochitonidæ, Galaxidæ, Osteoglossidæ, Symbranchidæ, and Sirenoidei; most of the genera being peculiar. The large and widely-distributed families, Cyprinodontidæ and Cyprinidæ, are absent. The most remarkable fish is the recently discovered Ceratodus, allied to the Lepidosiren of Tropical America, and Protopterus of Tropical Africa, the three species constituting the Sub-class Dipnoi, remains of which have been found fossil in the Triassic formation.

Summary of Australian Vertebrata.—In order to complete our general sketch of Australian zoology, and to afford materials for comparison with other regions, we will here summarize the distribution of Vertebrata in the entire Australian region, as given in detail in the tables at the end of this chapter. When an undoubted Oriental family or genus extends to Celebes only we do not count it as belonging to the Australian region, that island being so very anomalous and intermediate in character.

The Australian region, then, possesses examples of 18 families of Mammalia, 8 of which are peculiar; 71 of Birds, 16 being peculiar; 31 of Reptiles, 4 being peculiar; 11 of Amphibia, with 1 peculiar; and 11 of Fresh-water fish, with 1 peculiar. In all, 142 families of Vertebrates, 30 of which are almost or quite confined to it, or between one-fourth and one-fifth of the whole number.

The genera of Mammalia occurring within the limits of this region are 70, of which 45 are almost, or quite, confined to it.

Of Land-Birds there are 296 genera, 196 of which are equally limited. The proportion is in both cases very nearly five-eighths.

This shows a considerable deficiency both in families of Vertebrates and genera of Mammalia, as compared with the Oriental and Ethiopian regions; while in genera of Birds it is a little superior to the latter in total numbers, and considerably so in the proportion of peculiar types.

We may now consider how far the different classes and orders of vertebrates afford indications that during past ages there has been some closer connection between Australia and South America than that which now exists.

Among Mammalia we have the remarkable fact of a group of marsupials inhabiting South America, and extending even into the temperate regions of North America, while they are found in no other part of the globe beyond the limits of the Australian region; and this has often been held to be evidence of a former connection between the two countries. A preliminary objection to this view is, that the opossums seem to be rather a tropical group, only one species reaching as far as 42° south latitude on the west coast of South America; but whatever evidence we have which seems to require a former union of these countries shows that it took place, if at all, towards their cold southern limits, the tropical faunas on the whole showing no similarity. This is not a very strong objection, since climates may have changed in the south to as great an extent as we know they have in the north. Perhaps a more important consideration is, that Didelphys is a family type unknown in Australia; and this implies that the point of common origin is very remote in geological time. But the most conclusive fact is that in the Eocene and Miocene periods this very family, Didelphyidæ, existed in Europe, while it only appeared in America in the Post-pliocene or perhaps the Pliocene period; so that it is really an Old-World group, which, though long since extinct in its birthplace, has survived in America, to which country it is a comparatively recent emigrant. Primeval forms of marsupials we know abounded in Europe during much of the Secondary epoch, and no doubt supplied Australia with the ancestors of the present fauna. It is clear, therefore, that in this case there is not a particle of evidence for any former union between Australia and South America; while it is almost demonstrated that both derived their marsupials from a common source in the northern hemisphere.

Birds offer us more numerous but less clearly defined cases of this kind. Among Passeres, the wonderful lyre bird (Menura) is believed by some ornithologists to be decidedly allied to the South American Pteroptochidæ, while others maintain that it is altogether peculiar, and has no such affinity. The Australian Pachycephalidæ have also been supposed to find their nearest allies in the American Vireonidæ, but this is, perhaps, equally problematical. That the mound-makers (Megapodiidæ) of the Australian region are more nearly allied to the South American curassows (Cracidæ) than to any other family, is perhaps better established; but if proved, it is probably due, as in the case of the marsupials, to the survival of an ancient and once wide-spread type, and thus lends no support to the theory of a land connection between the two regions. A recent author, Professor Garrod, classes Phaps and other Australian genera of pigeons along with Zenaida and allied South American forms; but here again the affinity, if it exists, is so remote that the explanation already given will suffice to account for it. There remain only the penguins of the genus Eudyptes; and these have almost certainly passed from one region to the other, but no actual land connection is required for birds which can cross considerable arms of the sea.

Reptiles again seem to offer no more support to the view than do mammalia or birds. Among snakes there are no families in common that have not a very wide distribution. Among lizards the Gymnophthalmidæ are the only family that favour the notion, since they are found in Australia and South America, but not in the Oriental region. Yet they occur in both the Palæarctic and Ethiopian regions, and their distribution is altogether too erratic to be of any value in a case of this kind; and the same remarks apply to the tortoises of the family Chelydidæ.

The Amphibia, however, furnish us with some more decided facts. We have first the family of tree-frogs, Pelodryadæ, confined to the two regions; Litoria, a genus of the family Hylidæ peculiar to Australia, but with one species in Paraguay; and in the family Discoglossidæ, the Australian genus Chiroleptes has its nearest ally in the Chilian genus Calyptocephalus.

Fresh-water fishes give yet clearer evidence. Three groups are exclusively found in these two regions; Aphritis, a fresh-water genus of Trachinidæ, has one species in Tasmania and two others in Patagonia; the Haplochitonidæ inhabit only Terra del Fuego, the Falkland Islands and South Australia; while the genus Galaxias (forming the family Galaxidæ) is confined to South Temperate America, Australia, and New Zealand. We have also the genus Osteoglossum confined to the tropical rivers of Eastern South America, the Indo-Malay Islands and Australia.

It is important here to notice that the heat-loving Reptilia afford hardly any indications of close affinity between the two regions, while the cold-enduring amphibia and fresh-water fish, offer them in abundance. Taking this fact in connection with the absence of all indications of close affinity among the mammalia and terrestrial birds, the conclusion seems inevitable that there has been no land-connection between the two regions within the period of existing species, genera, or families. Yet some interchange of amphibia and fresh-water fishes, as of plants and insects, has undoubtedly occurred, but this has been effected by other means. If we look at a globe we see at once how this interchange may have taken place. Immediately south of Cape Horn we have the South Shetland Islands and Graham's land, which is not improbably continuous, or nearly so, with South Victoria land immediately to the south of New Zealand. The intervening space is partly occupied by the Auckland, Campbell, and Macquaries' Islands, which, there is reason to believe are the relics of a great southern extension of New Zealand. At all events they form points which would aid the transmission of many organisms; and the farthest of the Macquaries' group, Emerald Island, is only 600 miles from the outlying islets of Victoria land. The ova of fish will survive a considerable time in the air, and the successful transmission of salmon ova to New Zealand packed in ice, shows how far they might travel on icebergs. Now there is evidently some means by which ova or young fishes are carried moderate distances, from the fact that remote alpine lakes and distinct river systems often have the same species. Glaciers and icebergs generally have pools of fresh water on their surfaces; and whatever cause transmits fish to an isolated pond might occasionally stock these pools, and by this means introduce the fishes of one southern island into another. Batrachians, which are equally patient of cold, might be transported by similar means; while, as Mr. Darwin has so well shown, (Origin of Species, 6th Ed. p. 345) there are various known modes by which plants might be transmitted, and we need not therefore be surprised that botanists find a much greater similarity between the production of the several Southern lands and islands, than do zoologists. It is important to notice that, however this intercommunication was effected, it has continued down to the epoch of existing species; for Dr. Günther finds the same species of fresh-water fish (Galaxias attenuatus) inhabiting Tasmania, New Zealand, the Falkland Islands, and Temperate South America; while another species is common to New Zealand and the Auckland Islands. We cannot believe that a land connection has existed between all these remote lands within the period of existence of this one species of fish, not only on account of what we know of the permanence of continents and deep oceans, but because such a connection must have led to much more numerous and important cases of similarity of natural productions than we actually find. And if within the life of species such interchange may have taken place across seas of greater or less extent, still more easy is it to understand, how, within the life of genera and families, a number of such interchanges may have occurred; yet always limited to those groups whose conditions of life render transmission possible. Had an actual land connection existed within the temperate zone, or during a period of warmth in the Antarctic regions, there would have been no such strict limitations to the inter-migration of animals. It may be held to support the view that floating ice has had some share in the transmission of fish and amphibia, when we find that in the case of the narrow tropical sea dividing Borneo from Celebes and the Moluccas, no proportionate amount of transmission has taken place, but numerous species, genera, and whole families, terminate abruptly at what we have other reasons for believing to be the furthest limits of an ancient continent. We can hardly suppose, however, that this mode of transmission would have sufficed for such groups as tree-frogs, which are inhabitants of the more temperate or even warm portions of the two southern lands. Some of these cases may perhaps be explained by the supposition of a considerable extent of land in the South-Temperate and Antarctic regions now submerged, and by a warm or temperate climate analogous to that which prevailed in the Arctic regions during some part of the Miocene epoch; while others may be due to cases of survival in the two areas of once wide-spread groups, a view supported in the case of the Amphibia by the erratic manner in which many of the groups are spread over the globe.

From an examination of the facts presented by the various classes of vertebrates, we are, then, led to the conclusion, that there is no evidence of a former land-connection between the Australian and Neotropical regions; but that the various scattered resemblances in their natural productions that undoubtedly occur, are probably due to three distinct causes.

First, we have the American Didelphyidæ, among Mammals, and the Cracidæ, among birds, allied respectively to the Marsupials and the Megapodiidæ of Australia. This is probably more a coincidence than an affinity, due to the preservation of ancient wide-spread types in two remote areas, each cut off from the great northern continental masses, in which higher forms were evolved leading to the extinction of the lower types. In each of these southern isolated lands the original type would undergo a special development; in the one case suited to an arboreal existence, in the other to a life among arid plains.

The second case is that of the tree-frogs, and the genus Osteoglossum among fishes; and is most likely due to the extension and approximation of the two southern continents, and the existence of some intermediate lands, during a warm period when facilities would be afforded for the transmission of a few organisms by the causes which have led to the exceptional diffusion of fresh-water productions in all parts of the world. As however Osteoglossum occurs also in the Sunda Islands, this may be a case of survival of a once wide-spread group.

The third case is that of the same genera and even species of fish, and perhaps of frogs, in the two countries; which may be due to transmission from island to island by the aid of floating ice, with or without the assistance of more intervening lands than now exist.

Having arrived at these conclusions from a consideration of the vertebrata, we shall be in a position to examine how far the same causes will explain, or agree with, the distribution of the invertebrate groups, or elucidate any special difficulties we may meet with in the relations of the sub-regions.

The insects of the Australian region are as varied, and in some respects as peculiar as its higher forms of life. As we have already indicated in our sketch of the Oriental region, a vast number of forms inhabit the Austro-Malay sub-region which are absent from Australia proper. Such of these as are common to the Malay archipelago as a whole, have been already noted; we shall here confine ourselves more especially to the groups peculiar to the region, which are almost all either Australian or Austro-Malayan, the Pacific Islands and New Zealand being very poor in insect life.

Lepidoptera.—Australia itself is poor in butterflies, except in its northern and more tropical parts, where green Ornithopteræ and several other Malayan forms occur. In South Australia there are less than thirty-five species, whereas in Queensland there are probably over a hundred. The peculiar Australian forms are few. In the family Satyridæ, Xenica and Heteronympha, with Hypocista extending to New Guinea; among the Lycænidæ, Ogyris and Utica are confined to Australia proper, and Hypochrysops to the region; and in Papilionidæ, the remarkable Eurycus is confined to Australia, but is allied to Euryades, a genus found in Temperate South America (La Plata), and to the Parnassius of the North-Temperate zone.

The Austro-Malay sub-region has more peculiar forms. Hamadryas, a genus of Danaidæ, approximates to some South American forms; Hyades and Hyantis are remarkable groups of Morphidæ; Mynes and Prothoë are fine Nymphalidæ, the former extending to Queensland; Dicallaneura, a genus of Erycinidæ, and Elodina, of Pieridæ, are also peculiar forms. The fine Ægeus group of Papilio, and Priamus group of Ornithoptera, also belong exclusively to this region.

Xois is confined to the Fiji Islands, Bletogona to Celebes, and Acrophthalmia to New Zealand, all genera of Satyridæ. Seventeen genera in all are confined to the Australian region.

Among the Sphingina, Pollanisus, a genus of Zygænidæ, is Australian; also four genera of Castniidæ—Synemon, Euschemon, Damias, and Cocytia, the latter being confined to the Papuan islands. The occurrence of this otherwise purely South American family in the Australian region, as well as the affinity of Eurycus and Euryades noticed above, is interesting; but as we have seen that the genera and families of insects are more permanent than those of the higher animals, and as the groups in question are confined to the warmer parts of both countries, they may be best explained as cases of survival of a once wide-spread type, and may probably date back to the period when the ancestors of the Marsupials and Megapodii were cut off from the rest of the world.

Coleoptera.—The same remark applies here as in the Lepidoptera, respecting the affinity of the Austro-Malay fauna to that of Indo-Malay Islands; but Australia proper is much richer in beetles than in butterflies, and exhibits much more speciality. Although the other two parts of the Australian region (Polynesia and New Zealand) are very poor in beetles, it will, nevertheless, on the whole compare favourably with any of the regions except the very richest.

Cicindelidæ are not very abundant. Therates and Tricondyla are the characteristic genera in Austro-Malaya, but are absent from Australia, where we have Tetracha as the most characteristic genus, with one species of Megacephala and two of Distypsidera, a genus which is found also in New Zealand and some of the Pacific Islands. The occurrence of the South American genus, Tetracha, may perhaps be due to a direct transfer by means of intervening lands during the warm southern period; but considering the permanence of coleopterous forms (as shown by the Miocene species belonging almost wholly to existing genera), it seems more probable that it is a case of the survival of a once wide-spread group.

Carabidæ are well represented, there being no less than 94 peculiar genera, of which 19 are confined to New Zealand. The Australian genera of most importance are Carenum (68 species), Promecoderus (27 species), Silphomorpha (32 species), Adelotopus (27 species), Scaraphites (25 species), Notonomus (18 species), Gnathoxys (12 species), Eutoma (9 species), Ænigma (15 species), Lacordairea (8 species), Pamborus (8 species), Catadromus (4 species),—the latter found in Australia and Celebes. Common to Australia and New Zealand are Mecodema (14 species), Homalosoma (32 species), Dicrochile (12 species), and Scopodes (5 species). The larger genera, confined to New Zealand only, are Metaglymma (8 species), and Demetrida (3 species). The curious genus Pseudomorpha (10 species), is divided between California, Brazil, and Australia; and the Australian genera, Adelotopus, Silphomorpha, and Sphallomorpha, form with it a distinct tribe of Coleoptera. These being all confined to the warmer regions, and having so scattered a distribution, are no doubt the relics of a widespread group. The Australian genus, Promecoderus, has, however, closely allied genera (Cascelius and its allies), in Chili and Patagonia; while two small genera confined to the Auckland Islands (Heterodactylus and Pristancyclus) are allied to a group found only in Terra-del-Fuego and the Falkland Islands, (Migadops); and in these cases we may well believe that a direct transmission has taken place by some of the various means already indicated.

In Lucanidæ, Australia is only moderately rich, having 7 peculiar genera. The most important are Ceratognathus and Rhyssonotus, confined to Australia; Lissotes to Australia and New Zealand; Lamprima to Australia and Papua. Mitophyllus and Dendroblax inhabit New Zealand only; while Syndesus is found in Australia, New Caledonia, and tropical South America.

The beautiful Cetoniidæ are poorly represented, there being only 3 peculiar genera;—Schizorhina, mainly Australian, but extending to Papua and the Moluccas; Anacamptorhina, confined to New Guinea, and Sternoplus to Celebes. Lomaptera is very characteristic of the Austro-Malay Islands. This almost tropical family shows no approximations between the Australian and Neotropical faunas.

In Buprestidæ, the Australian region is the richest, possessing no less than 47 genera, of which 20 are peculiar to it. Of these, 15 are peculiar to Australia itself, the most important being Stigmodera (212 species), Ethon (13 species), and Nascio (3 species); Cisseis (17 species), and the magnificent Calodema (3 species), are common to Australia and Austro-Malaya; while Sambus (10 species) and Anthaxomorpha (4 species), with some smaller groups, are peculiarly Austro-Malayan. In this family occur several points of contact with the Neotropical region. Stigmodera is said to have a species in Chili, while there are undoubtedly several allied genera in Chili and South Temperate America. The genus Curis has 5 Australian and 3 Chilian species, and Acherusia has 2 species in Brazil, 1 in Australia. These resemblances may probably have arisen from intercommunication during the warm southern period, when floating timber would occasionally transmit a few larvæ of this family from island to island across the antarctic seas. When the cold period returned, they would spread northward, and become more or less modified under the new physical conditions and organic competition, to which they were subjected.

We now come to the very important group of Longicorns, in which the Australian region as a whole, is very rich, possessing 360 genera, of which 263 are peculiar to it. Of these about 50 are confined to the Austro-Malay Islands, 12 to New Zealand, and the remainder to Australia proper with Tasmania. Of the genera confined to, or highly characteristic of Australia, the following are the most important:—Cnemoplites, belonging to the Prionidæ; Phoracantha, to the Cerambycidæ; Zygocera, Hebecerus, Symphyletes, and Rhytidophora, to the Lamiidæ. Confined to the Austro-Malay Islands are Tethionea (Cerambycidæ): Tmesisternus, Arrhenotus, Micracantha, and Sybra (Lamiidæ); but there are also such Malayan genera as Batocera, Gnoma, Praonetha, and Sphenura, which are very abundant in the Austro-Malay sub-region. A species of each of the Australian genera, Zygocera, Syllitus, and Pseudocephalus, is said to occur in Chili, and one of the tropical American genus, Hammatochærus, in tropical Australia; an amount of resemblance which, as in the case of the Buprestidæ, may be imputed to trans-oceanic migration during the Southern warm period. This concludes our illustrations of the distribution of some of the more important groups of Australian insects; and it will be admitted that we have not met with any such an amount of identity with the fauna of Temperate South America, as to require us to modify the conclusions we arrived at from a consideration of the vertebrate groups.

Land-Shells.—The distribution of many of the larger genera of land-shells is very erratic, while others are exceedingly restricted, so that it requires an experienced conchologist to investigate the affinities of the several groups, and thus work out the important facts of distribution. All that can be done here is to note the characteristic and peculiar genera, and any others presenting features of special interest.

In the great family of the snails (Helicidæ), the only genera strictly confined to the region are, Partula, now containing above 100 species, and ranging over the Pacific from the Solomon Isles on the west, to the Sandwich Islands and Tahiti on the east; and Achatinella, now containing nearly 300 species, and wholly confined to the Sandwich Islands. Pfeifferia is confined to the Philippine Islands and Moluccas; Cochlostyla to the Indo-Malay Islands and Australia; Bulimus occurs in most of the insular groups, including New Zealand, but is absent from Australia.

Among the Aciculidæ, the widely-scattered Truncatella is the only genus represented. Among Diplommatinidæ, Diplommatina is the characteristic genus, ranging over the whole region, and found elsewhere as far as India, with one species in Trinidad. The extensive family Cyclostomidæ, is not well represented. Seven genera reach the Austro-Malay Islands, one of which, Registoma, is confined to the Philippines, Moluccas, New Caledonia, and the Marshall Islands. Omphalotropis is the most characteristic genus, ranging over the whole region; Callia is confined to the Philippines, Ceram, and Australia; Realia to New Zealand and the Marquesas. The genus Helicina alone represents the Helicinidæ, and is found in the whole region except New Zealand. The number of species known from Australia is perhaps about 300; while the Polynesian sub-region, according to Mr. Harper Pease, contains over 600; the Austro-Malay Islands will furnish probably 200; and New Zealand about 100; making a total of about 1,200 species for the whole region.

Few of the great zoological regions comprise four divisions so strongly contrasted as these, or which present so many interesting problems. We have first the Austro-Malay Islands, an equatorial forest-region teeming with varied and beautiful forms of life; next we have Australia itself, an island-continent with its satellite Tasmania, both tropical and temperate, but for the most part arid, yet abounding in peculiar forms in all the classes of animals; then come the Polynesian Islands, another luxuriant region of tropical vegetation, yet excessively poor in most of the higher groups of animals as well as in some of the lower; and lastly, we have New Zealand, a pair of temperate forest-clad islands far in the southern ocean, with a very limited yet strange and almost wholly peculiar fauna. We have now to consider the general features and internal relations of the faunas of each of these sub-regions, together with any external relations which have not been discussed while treating the region as a whole.

The central mass on which almost every part of this sub-region is clearly dependent, is the great island of New Guinea, inhabited by the Papuan race of mankind; and this, with the surrounding islands, which are separated from it by shallow seas and possess its most marked zoological features, are termed Papua. A little further away lie the important groups of the Moluccas on one side and the Eastern Papuan Islands on the other, which possess a fauna mainly derivative from New Guinea, yet wanting many of its distinctive types; and, in the case of the Moluccas possessing many groups which are not Australian, but derived from the adjacent Oriental region. To the south of these we have the Timor group, whose fauna is clearly derivative, from Australia, from Java, and from the Moluccas. Lastly comes Celebes, whose fauna is most complex and puzzling, and, so far as we can judge, not fundamentally derivative from any of the surrounding islands.

Papua, or the New Guinea Group.—New Guinea is very deficient in Mammalia as compared with Australia, though this apparent poverty may, in part, depend on our very scanty knowledge. As yet only four of the Australian families of Marsupials are known to inhabit it, with nine genera, several of which are peculiar. It also possesses a peculiar form of wild pig; but as yet no other non-marsupial terrestrial mammal has been discovered, except a rat, described by Dr. Gray as Uromys aruensis, but about the locality of which there seems some doubt. Omitting bats, of which our knowledge is very imperfect, the Papuan Mammals are as follows:—

We have here no sign of any approach to the Mammalian fauna of the Oriental region, for though Sus has appeared, the Muridæ (rats and mice) seem to be wanting.

In Birds the case is very different, since we at once meet with important groups, either wholly, or almost peculiar to the Papuan fauna. According to a careful estimate, embodying the recent discoveries of Meyer and D'Albertis, there are 350 species of Papuan land-birds comprised in 136 genera. About 300 of the species are absolutely peculiar to the district, while 39 of the genera are exclusively Papuan or just extend into the Moluccas, or into North Australia where it closely approaches New Guinea. In analysing the genera we may set aside 31 as having a wide range, and being of no significance in distribution; such are most of the birds of prey, with the genera Hirundo, Caprimulgus, Zosterops; and others widely spread in both the Oriental and Australian regions, as Dicæum, Munia, Eudynamis, &c. Of the remainder, as above stated, about 39 are peculiar to the Papuan fauna, 50 are characteristic Australian genera; 9 are more especially Malayan, and as much Australian as Oriental; while 7 only, appear to be typically Oriental with a discontinuous distribution, none of them occurring in the Moluccas.

This Papuan fauna is so interesting and remarkable, that it seems advisable to give lists of these several classes of generic types.

I. Genera occurring in the Papuan Islands which are characteristic of the Australian region (89). Those marked with an asterisk are exclusively Papuan.

The chief points of interest here are the richness and specialization of the parrots, pigeons, and kingfishers; the wonderful paradise-birds; the honeysuckers; and some remarkable flycatchers. The most prominent deficiencies, as compared with Australia, are in Sylviidæ, Timaliidæ, Ploceidæ, Platycercidæ, and Falconidæ.

II. The genera which are characteristic of the whole Malay Archipelago are the following (10):—

III. The curious set of genera apparently of Indo-Malayan origin, but unknown in the Moluccas, are as follows:—

The above six birds are very important as indicating past changes in the Austro-Malay Islands, and we must say a few words about each. (1) Eupetes is very remarkable, since the New Guinea birds resemble in all important characters that which is confined to Malacca and Sumatra. They are probably the survivors of a once wide-spread Malayan group. (2) Alcippe or Drymocataphus (for in which genus the birds should be placed is doubtful) seems another clear case of a typical Indo-Malayan form occurring in New Guinea and Java, but in no intervening island. (3) Pomatorhinus is a most characteristic Himalayan and Indo-Malayan genus, occurring again in New Guinea and also in Australia, but in no intermediate island. The New Guinea bird seems as nearly related to Oriental as Australian species. (4) Arachnothera is exactly parallel to Alcippe, occurring nowhere east of Borneo except in New Guinea. (5) Prionochilus, a small black bird, sometimes classed as a distinct genus, but evidently allied to the Prionochili of the Indo-Malay Islands. (6) Eulabes, the genus which contains the well known Mynahs of India, extends east of Java as far as Flores, but is not found in Celebes or the Moluccas. The two New Guinea species are sometimes classed in different genera, but they are undoubtedly allied to the Mynahs of India and Malaya.

We find then, that while the ornithology of New Guinea is preeminently Australian in character and possesses many peculiar developments of Australian types, it has also—as might be expected from its geographical position, its climate, and its vegetation—received an infusion of Malayan forms. But while one group of these is spread over the whole Archipelago, and occasionally beyond it, there is another group which presents the unusual and interesting feature of discontinuous distribution, jumping over a thousand miles of island-studded sea from Java and Borneo to New Guinea itself. It is a parallel case to that of Java in the Oriental region, which we have already discussed, but the suggested explanation in that case is more difficult to apply here. The recent soundings by the Challenger show us, that although the several islands of the Moluccas are surrounded by water from 1,200 to 2,800 fathoms deep, yet these seas form inclosed basins with rims not more than from 400 to 900 fathoms deep, suggesting the idea of great lakes or inland seas which have sunk down bodily with the surrounding land, or that enormous local and restricted elevations and subsidences have here occurred. We have also the numerous small islands and coral banks south of Celebes and eastward towards Timor-Laut and the Aru Islands, indicating great subsidence; and it is possible that there was an extension of Papua to the west, approaching sufficiently near to Java to receive occasional straggling birds of Indo-Malay type, altogether independent of the Moluccas to the north.

Bright Colours and Ornamental Plumage of New Guinea Birds.—One of the most striking features of Papuan ornithology is the large proportion which the handsome and bright-coloured birds bear to the more obscure species. That this is really the case has been ascertained by going over my own collections, made at Aru and New Guinea, and comparing them with my collection made at Malacca—a district remarkable for the number of handsome birds it produces. Using, as nearly as possible, the same standard of beauty, about one-third of the Malacca birds may be classed as handsome, while in Papua the proportion comes out exactly one-half. This is due, in part to the great abundance of parrots, cockatoos, and lories, almost all of which are beautiful; and of pigeons, more than half of which are very beautiful; as well as to the numerous kingfishers, most of which are excessively brilliant. Then we have the absence of thrushes, and the very small numbers of the warblers, shrikes, and Timaliidæ, which are dull-coloured groups; and, lastly, the presence of numerous gay pittas, flycatchers, and the unequalled family of paradise-birds. A large number of birds adorned with metallic plumage is also a marked feature of this fauna, more than a dozen genera being so distinguished. Among the remarkable forms are Peltops, a flycatcher, long classed as one of the Indo-Malayan Eurylæmidæ, which it resembles both in bill and coloration; Machærirhynchus, curious little boat-billed flycatchers; and Todopsis, a group of terrestrial flycatchers with the brilliant colours of Pitta or Malurus. The paradise-birds present the most wonderful developments of plumage and the most gorgeous varieties of colour, to be found among passerine birds. The great whiskered-swift, the handsomest bird in the entire family, has its head-quarters here. Among kingfishers the elegant long-tailed Tanysipteræ are preeminent, whether for singularity or beauty. Among parrots, New Guinea possesses the great black cockatoo, one of the largest and most singular birds in the order; Nasiterna, the smallest of known parrots; and Charmosyna, perhaps the most elegant. Lastly, among the pigeons we have the fine crowned-pigeons, the largest and most remarkable group of the order.



Plate X. Illustrating the Ornithology of New Guinea.—The wonderful ornithological fauna we have just sketched, could only be properly represented in a series of elaborate coloured plates. We are obliged here to confine ourselves to representing a few of the more remarkable types of form, as samples of the great number that adorn this teeming bird-land. The large central figure is the fine twelve-wired paradise-bird (Seleucides albus), one of the most beautiful and remarkable of the family. Its general plumage appears, at first sight, to be velvety black; but on closer examination, and by holding the bird in various lights, it is found that every part of it glows with the most exquisite metallic tints—rich bronze, intense violet, and, on the edges of the breast-feathers, brilliant green. An immense tuft of dense plumes of a fine orange-buff colour, springs from each side of the body, and six of these on each side terminate in a black curled rachis or shaft, which form a perfectly unique adornment to this lovely bird. To appreciate this wonderful family (of which no good mounted collection exists) the reader should examine the series of plates in Mr. Elliot's great work on the Paradiseidæ, where every species is figured of the size of life, and with a perfection of colouring that leaves little to be desired.

Below the Seleucides is one of the elegant racquet-tailed king-hunters (Tanysiptera galatea) whose plumage of vivid blue and white, and coral-red bill, combined with the long spatulate tail, renders this bird one of the most attractive of the interesting family of kingfishers. On a high branch is seated the little Papuan parroquet (Charmosyna papuensis), one of the Trichoglossidæ, or brush-tongued parrots,—richly adorned in red and yellow plumage, and with an unusually long and slender tail. On the ground is the well-known crowned pigeon (Goura coronata), a genus which is wholly confined to New Guinea and a few of the adjacent islands. One of the very few Papuan mammals, a tree-kangaroo (Dendrolagus inustus), is seated on a high branch. It is interesting, as an arboreal modification of a family which in Australia is purely terrestrial; and as showing how very little alteration of form or structure is needed to adapt an animal to such a different mode of life.

Reptiles and Amphibia.—Of these classes comparatively little is at present known, but there is evidence that the same intermixture of Oriental and Australian forms that occurs in birds and insects, is also found here. Dr. A. B. Meyer, the translator of this work into German, and well known for his valuable discoveries in New Guinea, has kindly furnished me with a manuscript list of Papuan reptiles, from which most of the information I am able to give is derived.

Of Snakes, 24 genera, are known, belonging to 11 families. Six of the genera are Oriental,—Calamaria, Cerberus, Chrysopelea, Lycodon, Chersydrus, and Ophiophagus. Four are Australian,—Morelia, Liasis, Diemenia, and Acanthophis; while four others are more especially Papuan,—Dibamus (Typhlopidæ), Brachyorros—a sub-genus of the wide-spread Rhabdosoma (Calamariidæ), found also in Timor; Nardoa and Enygrus (Pythonidæ), ranging from the Moluccas to the Fiji Islands. The rest are either common to the Oriental and Australian regions or of wide range.

Of Lizards also, 24 genera are recorded, belonging to 5 families. Three only are peculiarly Oriental,—Eumeces, Tiaris, and Nycteridium; but another, Gonyocephalus, is Malayan, ranging from Java and Borneo to the Pelew Islands. Three are Australian,—Cyclodus, Heteropus, and Gehyra; while six are especially Papuan,—Keneuxia (extending to the Philippines), Elania, Carlia (to North Australia), Lipinia (to the Philippine Islands), and Tribolonotus,—all belonging to the Scincidæ; and Arua belonging to the Agamidæ. We must add Cryptoblepharus, which is confined to the Australian region, except a species in Mauritius. The other genera have a wider distribution.

The preponderant Oriental element in the snakes as compared with the lizards, is suggestive of the dispersal of the former being dependent on floating trees, or even on native canoes, which for an unknown period have traversed these seas, and in which various species of snakes often secrete themselves. This seems the more probable, as snakes are usually more restricted in their range than lizards, and exhibit less numerous examples of widespread genera and species. The other orders of reptiles present no features of interest.

Of Amphibia only 8 genera are known, belonging to 6 families. Rana, Hylarana, and Hyla are wide-spread genera, the former being, however, absent from Australia. Hyperolius, Pelodryas, Litoria, and Asterophrys are Australian; while Platymantis is Polynesian, with a species in the Philippine Islands. Hence it appears that the amphibia, so far as yet known, exhibit no Oriental affinity; and this is a very suggestive fact. We have seen (p. 29) that salt water is almost a complete barrier to the dispersal of these creatures; so that the wholly Australian character of the Papuan batrachia is what we might expect, if, as here advocated, no actual land connection between the Oriental and Australian regions, has probably occurred during the entire Tertiary and Post-tertiary periods.

Insects.—The general character of the Papuan insects has been sufficiently indicated in our sketch of the Entomology of the region. We will here only add, that the metallic lustre so prevalent among the birds, is also apparent in such insects as Sphingnotus mirabilis, a most brilliant metallic Longicorn; Lomaptera wallacei and Anacamptorhina fulgida, Cetonii of intense lustre; Calodema wallacei among the Buprestidæ; and the elegant blue Eupholi among the weevils. Even among moths we have Cocytia durvillii, remarkable for its brilliant metallic colours.

The Moluccas.—The islands of Gilolo, Bouru, and Ceram, with several smaller islands adjacent, together with Sanguir, and perhaps Tulour or Salibaboo to the north-west, and the islands from Ke to Timor-Laut to the south-east, form the group of the Moluccas or Spice-Islands, remarkable for the luxuriance of their vegetation and the extreme beauty of their birds and insects. Their Mammalia are of Papuan character, with some foreign intermixture. Two genera of the New Guinea marsupials, Belideus and Cuscus, abound; and we have also the wide-spread Sus. But besides these, we find no less than five genera of placental Mammals quite foreign to the Papuan or Australian faunas. These are 1. Cynopithecus nigrescens, found only in the small island of Batchian, and probably introduced from Celebes, where the same ape occurs. 2. Viverra tangalunga, a common Indo-Malayan species of civet, probably introduced. 3. Cervus hippelaphus, var. Moluccensis, a deer abundant in all the islands, very close to a Javan species and almost certainly introduced by man, perhaps very long ago. 4. Babirusa alfurus, the babirusa, found only in the island of Bouru, and perhaps originally introduced from Celebes. 5. Sorex sp., small shrews. With the exception of the last, all these species are animals habitually domesticated and kept in confinement by the Malays; and when we consider that none of the smaller Mammalia of Java and Borneo, numbering at least fifty different species, are found in any of the Moluccas, we can hardly suppose that such large animals as the deer and ape, could have reached them by natural means. There is every reason to believe, therefore, that the indigenous Mammalia of the Moluccas are wholly of Papuan stock, and very limited in number.

The birds are much more varied and interesting. About 200 species of land-birds are now known, belonging to 85 genera. Of the species about 15 are Indo-Malayan, 32 Papuan, and about 140 peculiar. Of the genera only two are peculiar,—Semioptera, a paradise bird, and Lycocorax, a singular form of Corvidæ; but there is also a peculiar rail-like wader, Habroptila. One genus, Basilornis, is found only in Ceram and Celebes; another, Scythrops, is Australian, and perhaps a migrant. About 30 genera are characteristic Papuan types, and 37 others, of more or less wide range, are found in New Guinea and were therefore probably derived thence. There remains a group of birds which are not found in New Guinea, and are either Palæarctic or Oriental. These are 13 in number as follows:—

Of these the Monticola, found only in Gilolo, appears to be a straggler or migrant from the Philippine islands. Acrocephalus, of which four species occur, is a wide-spread group; one of the Moluccan birds is an Australian and another a North-Asian species, which perhaps indicates that there has long been some migration southward from island to island, across the Moluccas. Cisticola is a genus of very wide range, extending to Australia. Hypolais is probably a modified form of a Chinese or Javanese species. Criniger is a pure Indo-Malay form, represented here by three fine species. Butalis is a Chinese species, no doubt straggling southward. Budytes and Corydalla are widespread Oriental and Palæarctic species or slight modifications of them. Hydrornis is a Malayan form of Pittidæ. Batrachostomus is a distinct representative of a purely Indo-Malay genus. Loriculus is Malayan, and especially Philippine, but it reaches as far as Mysol. Treron is here at its eastern limit, and is represented in Bouru and Ceram by one of the most beautiful species. Neopus, a Malayan eagle, is said to occur in the Moluccas. We find then only three characteristic Indo-Malay types in the Moluccas,—Criniger, Batrachostomus, and Treron. All are represented by distinct and well marked species, indicating a somewhat remote period since their ancestors entered the district, but all are birds of considerable powers of flight, so that a very little extension of the islands in a south-westerly direction would afford the means of transmission, but this could not well have been by way of Celebes, because the two former genera are unknown in that island.

It is evident, therefore, that the Moluccas are wholly Papuan in their zoology; yet they are no less clearly derivative, and must have obtained their original immigrants under conditions that rendered a full representation of the fauna impossible. Such remarkable and dominant types as the eleven genera of Paradiseidæ, with Cracticus, Rectes, Todopsis, Machærirhynchus, Gerygone, Dacelo, Podargus, Cyclopsitta, Microglossus, Nasiterna, Chalcopsitta, and Goura,—all characteristic Papuan groups, found in almost all the islands and most of them very abundant, are yet totally absent from the Moluccas. Taking this, in conjunction with the absence of the two genera of Papuan kangaroos and the other smaller groups of marsupials, and we must be convinced that the Moluccas cannot be mere fragments of the old Papuan land, or they would certainly, in some one or other of their large and fertile islands, have preserved a more complete representation of the parent fauna. Most of the Moluccan birds are very distinct from the allied species of New Guinea; and this would imply that the entrance of the original forms took place at a remote period. The two peculiar genera with clearly Papuan affinities, show the same thing. The cassowary, found only in the large island of Ceram and distinct from any Papuan species, would however seem to have required a land connection for its introduction, almost as much as any of the larger mammalia.

Taking all the facts into consideration, I would suggest as the most probable explanation, that if the Moluccas ever formed part of the main Papuan land, they were separated at an early date, and subsequently so greatly submerged as to destroy a large proportion of their fauna. They have since risen, and have probably been larger than at present, and rather more closely approximated to the parent land, whence they received a considerable immigration of such animals as were adapted to cross narrow seas. This gave them several Papuan forms, but still left them without a number of the types more especially confined to the forest depths, or powerful enough to combat the gales which often blow weaker flyers out to sea. Most of the birds whose absence from the Moluccas is so conspicuous belong to one or other of these classes.

Among the most characteristic birds of the Moluccas are the handsome crimson lories of the genera Lorius and Eos. These are found in every island (but not in Celebes or the Timor group); and a fine species of Eos, peculiar to the small islands of Siau and Sanguir, just north of Celebes, obliges us to place these with the Moluccas instead of with the former island, to which they seem most naturally to belong. The crimson parrots of the genus Eclectus are almost equally characteristic of the Moluccas, and add greatly to the brilliancy of the ornithology of these favoured islands.

Reptiles.—The Reptiles, so far as known, appear to agree in their distribution with the other vertebrates. In some small collections from Ceram there were no less than six of the genera peculiar to the Australian region, and which were before only known from Australia itself. These are, of snakes, Liasis and Enygrus, genera of Pythonidæ; with Diemenia and Acanthophis (Elapidæ); of lizards, Cyclodus, a genus of Scincidæ; and of Amphibia, a tree-frog of the genus Pelodryas.

Insects—Peculiarities of the Moluccan Fauna.—In insects the Moluccas are hardly, if at all, inferior to New Guinea itself. The islands abound in grand Papilios of the largest size and extreme beauty; and it is a very remarkable fact, that when the closely-allied species of the Moluccas and New Guinea are compared, the former are almost always the largest. As examples may be mentioned, Ornithoptera priamus and O. helena of the Moluccas, both larger than the varieties (or species) of Papua; Papilio ulysses and deiphobus of Amboyna, usually larger than their allies in New Guinea; Hestia idea, the largest species of the genus; Diadema pandarus and Charaxes euryalus, both larger than any other species of the same genera in the whole archipelago. It is to be noted also, that in the Moluccas, the very largest specimens or races seem always to come from the small island of Amboyna; even those of Ceram, the much larger island to which it is a satellite, being almost always of less dimensions. Among Coleoptera, the Moluccas produce Euchirus longimanus, one of the largest and most remarkable of the Lamellicornes; Sphingnotus dunningi, the largest of the Austro-Malayan Tmesisterninæ; a Sphenura, the largest and handsomest of an extensive genus; an unusually large Schizorhina (Cetoniidæ); and some of the most remarkable and longest-horned Anthotribidæ. Even in birds the same law may be seen at work,—in the Tanysiptera nais of Ceram, which has a larger tail than any other in the genus; in Centropus goliath of Gilolo, being the largest and longest-tailed species; in Hydrornis maximus of Gilolo, the largest and perhaps the most elegantly and conspicuously coloured of all the Pittidæ; in Platycercus amboinensis, being pre-eminent in its ample blue tail; in the two Moluccan lories and Eos rubra, being more conspicuously red than the allied New Guinea species; and in Megapodius wallacei of Bouru, being the only species of the genus conspicuously marked and banded.

All these examples, of larger size, of longer tails or other appendages, and of more conspicuous colouring, are probably indications of a less severe struggle for existence in these islands than in the larger tract of New Guinea, with a more abundant and more varied fauna; and this may apply even to the smaller islands, as compared with the larger in the immediate vicinity. The limited number of forms in the small islands compared with a similar area in the parent land, implies, perhaps, less competition and less danger; and thus allows, where all other conditions are favourable, an unchecked and continuous development in size, form, and colour, until they become positively injurious. This law may not improbably apply to the New Guinea fauna itself, as compared with that of Borneo or any other similar country; and some of its peculiarities (such as its wonderful paradise-birds) may be due to long isolation, and consequent freedom from the influence of any competing forms. The difference between the very sober colours of the Coleoptera, and in a less degree of the birds, of Borneo, as compared with their brilliancy in New Guinea, always struck me most forcibly, and was long without any, even conjectural, explanation. It is not the place here to go further into this most curious and interesting subject. The reader who wishes for additional facts to aid him in forming an opinion, should consult Mr. Darwin's Descent of Man, chapters x. to xv.; and my own Contributions to the Theory of Natural Selection, chapters iii. and iv.

Timor Group.—Mammalia.—In the group of islands between Java and Australia, from Lombok to Timor inclusive, we find a set of mammals similar to those of the Moluccas, but some of them different species. A wide-spread species of Cuscus represents the Papuan element. A Sorex and a peculiar species of wild pig, we may also accept as indigenous. Three others have almost certainly been introduced. These are, (1.) Macacus cynomolgus, the very commonest Malay monkey, which may have crossed the narrow straits from island to island between Java and Timor, though it seems much more probable that it was introduced by Malays, who constantly capture and rear the young of this species. (2.) Cervus timoriensis, a deer, said to be a distinct species, inhabits Timor, but it is probably only a variety of the Cervus hippelaphus of Java. This animal is, however, much more likely to have crossed the sea than the monkey. (3.) Paradoxurus fasciatus, takes the place of Viverra tangalunga in the Moluccas, both common and wide-spread civets which are often kept in confinement by the Malays. The Felis megalotis, long supposed to be a native of Timor, has been ascertained by Mr. Elliot to belong to a different country altogether.

Birds.—The birds are much more interesting, since they are sufficiently numerous to allow us to determine their relations, and trace their origin, with unusual precision. There are 96 genera and 160 species of land-birds known to inhabit this group of islands; and on a careful analysis, they are found to be almost equally related to the Australian and Oriental regions, 30 genera being distinctly traceable to the former, and the same number to the latter. Their connection with the Moluccas is shown by the presence of the genera Mimeta, Geoffroyus, Cacatua, Ptilopus, and Ianthœnas, together with Megapodius and Cerchneis represented by Moluccan species. Turacœna shows a connection with Celebes, and Scops is represented by a Celebesian species. The connection with Australia is shown by the genera Sphæcothera, Gerygone, Myiagra, Pardalotus, Gliciphila, Amadina, and Aprosmictus; while Milvus, Hypotriorchis, Eudynamis, and Eurystomus, are represented by Australian species. Other genera confined to or characteristic of the Australian region, are Rhipidura, Monarcha, Artamus, Campephaga, Pachycephala, Philemon, Ptilotis, and Myzomela.

We now come to the Indo-Malay or Javan element represented by the following genera:

Such genera as Merops and Strix, which are as much Australian as Oriental, are inserted here because they are represented by Javan species. The list is considerably swelled by genera which have reached Lombok across the narrow strait from Baly, but have passed no further. Such are Zoothera, Orthotomus, Pycnonotus, Pericrocotus and Strix. A much larger number (12) stop short at Flores, leaving only 13, indicated in the list by (T) after their names, which reach Timor. It is evident, therefore, that these islands have been stocked from three chief sources,—the Moluccas (with New Guinea and Celebes,) Australia, and Java. The Moluccan forms may well have arrived as stragglers from island to island, aided by whatever facilities have been afforded by lands now submerged. Most of the remainder have been derived either from Australia or from Java; and as their relations to these islands are very interesting, they must be discussed with some detail.

Origin of the Timorese Fauna.—We must first note, that 80 species, or exactly one-half of the land-birds of the islands, are peculiar and mostly very distinct, intimating that the immigration commenced long enough back to allow of much specific modification. There is also one peculiar genus of kingfishers, Caridonax, found only in Lombok and Flores, and more allied to Australian than to Oriental types. The fine white-banded pigeons (s. g. Leucotreron) are also almost peculiar; one other less typical species only being known, a native of N. Celebes. In order to compare the species with regard to their origin, we must first take away those of wide distribution from which no special indications can be obtained. In this case 49 of the land-birds must be deducted, leaving 111 species which afford good materials for comparison. These, when traced to their origin, show that 62 came from some part of the Australian region, 49 from Java or the Oriental region. But if we divide them into two groups, the one containing the species identical with those of the Australian or Oriental regions, the other containing allied or representative species peculiar to the islands, we have the following result:

This table is very important, as indicating that the connection with Australia was probably earlier than that with Java; since the majority of the Australian species have become modified, while the majority of the Oriental species have remained unchanged. This is due, no doubt, in part to the continued immigration of fresh individuals from Java, after that from Australia, the Moluccas and New Guinea had almost wholly ceased. We must also notice the very small proportion of the genera, either of Australia or Java, that have found their way into these islands, many of the largest and most wide-spread groups in both countries being altogether absent. Taking these facts into consideration, it is pretty clear that there has been no close and long-continued approximation of these islands to any part of the Australian region; and it is also probable that they were fairly stocked with such Australian groups as they possess before the immigration from Java commenced, or a larger number of characteristic Oriental forms would have been able to have established themselves.

On looking at our map, we find that a shallow submerged bank extends from Australia to within about twenty miles of the coast of Timor; and this is probably an indication that the two countries were once only so far apart. This would have allowed the purely Australian types to enter, as they are not numerous; there being about 6 Australian species, and 10 or 12 representatives of Australian species, in Timor. All the rest may have been derived from the Moluccas or New Guinea, being mostly wide-spread genera of the Australian region; and the extension of Papua in a south-west direction towards Java (which was suggested as a means of providing New Guinea with peculiar Indo-Malay types not found in any other part of the region) may have probably served to supply Timor and Flores with the mass of their Austro-Malayan genera across a narrow strait or arm of the sea. Lombok, Baly, and Sumbawa were probably not then in existence, or nothing more than small volcanic cones rising out of the sea, thus leaving a distance of 300 miles between Flores and Java. Subsequently they grew into islands, which offered an easy passage for a number of Indo-Malay genera into such scantily stocked territories as Flores and Timor. The north coast of Australia then sank, cutting off the supply from that country; and this left the Timorese group in the position it now occupies.

The reptiles and fishes of this group are too little known to enable us to make any useful comparison.

Insects.—The insects, though not numerous, present many fine species, some quite unlike any others in the Archipelago. Such are—Papilio liris, Pieris læta, Cirrochroa lamarckii and C. leschenaultii among butterflies. The Coleoptera are comparatively little known, but in the insects generally the Indo-Malay element predominates. This may have arisen from the peculiar vegetation and arid climate not being suitable to the Papuan insects. Why Australian forms did not establish themselves we cannot conjecture; but the field appears to have been open to immigrants from Java, the climate and vegetation of which island at its eastern extremity approximates to that of the Timorese group. The insects are, however, so peculiarly modified as to imply a very great antiquity, and this is also indicated by a group of Sylviine birds here classed under Oreicola, but some of which probably form distinct genera. There may, perhaps, have been an earlier and a later approximation to Java, which, with the other changes indicated, would account for most of the facts presented by the fauna of these islands. One deduction is, at all events, clear: the extreme paucity of indigenous mammals along with the absence of so many groups of birds, renders it certain that the Timorese islands did not derive their animal life by means of an actual union with any of the large islands either of the Australian or the Oriental regions.

Celebes Group.—We now come to the Island of Celebes, in many respects the most remarkable and interesting in the whole region, or perhaps on the globe, since no other island seems to present so many curious problems for solution. We shall therefore give a somewhat full account of its peculiar fauna, and endeavour to elucidate some of the causes to which its zoological isolation may be attributed.

Mammalia.—The following is the list of the mammalia of Celebes as far as at present known, though many small species may yet be discovered.

The first—a large black ape—is itself an anomaly, since it is not closely allied to any other form of quadrumana. Its flat projecting muzzle, large superciliary crests and maxillary ridges, with the form and appearance of its teeth, separate it altogether from the genus Macacus, as represented in the Indo-Malay islands, and ally it closely to the baboons of Africa. We have already seen reason to suppose that it has been carried to Batchian, and there is some doubt about the allied species or variety (C. niger) of the Philippines being really indigenous there; in which case this interesting form will remain absolutely confined to Celebes. (2.) The tarsier is a truly Malayan species, but it is said to occur in a small island at the northern extremity of Celebes and on some of the Philippine Islands. It might possibly have been introduced there. (3) and (4)—a civet and a deer—are, almost certainly, as in the Moluccas, introduced species. (5.) Anoa depressicornis. This is one of the peculiar Celebesian types; a small straight-horned wild-bull, anatomically allied to the buffaloes, and somewhat resembling the bovine antelopes of Africa, but having no near allies in the Oriental region. (6.) Sus Celebensis; a peculiar species of wild-pig. (7.) Babirusa alfurus; another remarkable type, having no near allies. It differs in its dentition from the typical Suidæ, and seems to approach the African Phacochœridæ, The manner in which the canines of the upper jaw are reversed, and grow directly upwards in a spiral curve over the eyes, is unique among mammalia. (8.) Five squirrels inhabit Celebes, and all are peculiar species. (9.) These are forest rats of the sub-genus Gymnomys, allied to Australian species. 10. Cuscus. This typical Australian form is represented in Celebes by two peculiar species.

Leaving out the Indo-Malay species, which may probably have been introduced by man, and are at all events comparatively recent immigrants, and the wild pig, a genus which ranges over the whole archipelago and which has therefore little significance, we find two genera which have come from the Australian side,—Cuscus and Mus; and four from the Oriental side,—Cynopithecus, Anoa, Babirusa, and Sciurus. But Sciurus alone corresponds to Cuscus, as a genus still inhabiting the adjacent islands; the other three being not only peculiar to Celebes, but incapable of being affiliated to any specially Oriental group. We seem, then, to have indications of two distinct periods; one very ancient, when the ancestors of the three peculiar genera roamed over some unknown continent of which Celebes formed, perhaps, an outlying portion;—another more recent, when from one side there entered Sciurus, and from the other Cuscus. But we must remember that the Moluccas to the east, possess scarcely any indigenous mammals except Cuscus; whereas Borneo and Java on the west, have nearly 50 distinct genera. It is evident then, that the facilities for immigration must have been much less with the Oriental than with the Australian region, and we may be pretty certain that at this later period there was no land connection with the Indo-Malay islands, or some other animals than squirrels would certainly have entered. Let us now see what light is thrown upon the subject by the birds.

Birds.—The total number of birds known to inhabit Celebes is 205, belonging to about 150 genera. We may leave out of consideration the wading and aquatic birds, most of which are wide-ranging species. There remain 123 genera and 152 species of land-birds, of which 9 genera and 66 species are absolutely confined to the island, while 20 more are found also in the Sula or Sanguir Islands, so that we may take 86 to be the number of peculiar Celebes species. Lord Walden, from whose excellent paper on the birds of Celebes (Trans. Zool. Soc. vol. viii. p. 23) most of these figures are obtained, estimates, that of the species which are not peculiar to Celebes, 55 are of Oriental and 22 of Australian origin, the remainder being common to both regions. This shows a preponderant recent immigration from the West and North, which is not to be wondered at when we look at the long coast line of Java, Borneo, and the Philippine islands, with an abundant and varied bird population, on the one side, and the small scattered islands of the Moluccas, with a comparatively scanty bird-fauna, on the other.

But, adopting the method here usually followed, let us look at the relations of the genera found in Celebes, omitting for the present those which are peculiar to it. I divide these genera into two series:—those which are found in Borneo or Java but not in the Moluccas, and those which inhabit the Moluccas and not Borneo or Java; these being the respective sources from which, primâ facie, the species of these genera must have been derived. Genera which range widely into both these districts are rejected, as teaching us nothing of the origin of the Celebesian fauna. In a few cases, sub-genera which show a decided eastern or western origin, are given.

These tables show a decided preponderance of Oriental over Australian forms. But we must remember that the immediately adjacent lands from whence the supply was derived, is very much richer in the one case than in the other. The 24 genera derived from Borneo and Java are only about one fourth of the characteristic genera of those islands; while the 15 Moluccan and Timorese genera are fully one third of their characteristic types. The proportion derived from the Australian, is greater than that derived from the Oriental side.

We shall exhibit this perhaps more clearly, by giving a list of the important groups of each set of islands which are absent from Celebes.

If we reckon the absent families to be each represented by only two important genera, we shall find the deficiency on the Oriental side much the greatest; yet those on the side of the Moluccas are sufficiently remarkable. The Meliphagidæ are not indeed absolutely wanting, since a Myzomela has now been found in Celebes; but all its larger and more powerful forms which range over almost the entire region, are absent. This may be balanced by the absence of the excessively abundant Timaliidæ of the Indo-Malay islands, which are represented by only a single species; and by the powerful Phasianidæ, represented only by the common Malay jungle fowl, perhaps introduced. The entire absence of Pycnonotidæ is a very anomalous fact, since one of the largest genera, Criniger, is well represented in several islands of the Moluccas, and one has even been found in the Togian islands in the great northern inlet of Celebes; but yet it passes over Celebes itself. Ceyx, a genus of small kingfishers, is a parallel case, since it is found everywhere from India to New Guinea, leaving out only Celebes; but this comes among those curiosities of the Celebesian fauna which we shall notice further on. In the list of genera derived from Borneo or Java, no less than 6 are represented by identical species (indicated by sp. after the name); while in the Moluccan list 5 are thus identical. These must be taken to indicate, either that the genus is a recent introduction, or that stragglers still occasionally enter, crossing the breed, and thus preventing specific modification. In either case they depend on the existing state of things, and throw no light on the different distribution of land and sea which aided or checked migration in former times; and they therefore to some extent diminish the weight of the Indo-Malay affinity, as measured by the relations of the peculiar species of Celebes.

From our examination of the evidence thus far,—that is, taking account firstly, of the species, and, secondly, of the genera, which are common to Celebes and the groups of islands between which it is situated, we must admit that the connexion seems rather with the Oriental than with the Australian region; but when we take into account the proportion of the genera and species present, to those which are absent, and giving some weight to the greater extent of coast line on the Indo-Malay side, we seem justified in stating that the Austro-Malay element is rather the most fully represented. This result applies both to birds and mammals; and it leads us to the belief, that during the epoch of existing species and genera, Celebes has never been united with any extensive tract of land either on the Indo-Malay or Austro-Malay side, but has received immigrants from both during a very long period, the facilities for immigration having been rather the greatest on the Austro-Malay or Australian side. We have now to consider what further light can be thrown on the subject by the consideration of the peculiar genera of Celebes, and of those curiosities or anomalies of distribution to which we have referred.

Nine genera of birds are altogether peculiar to Celebes; three more are found only in one other island, and seem to be typically Celebesian; while one is found in the Sula islands (which belongs to the Celebes group) and probably exists in Celebes also. The following is a list of these 13 genera:

Of the above, Artamides, Monachalcyon, Cittura, and Megacephalon, are modifications of types characteristic of the Australian region. All are peculiar to Celebes except Cittura, found also in the Sanguir islands to the northward, but which seems to belong to the Moluccan group. Streptocitta, Charitornis, and Gazzola, are peculiar types of Corvidæ; the two former allied to the magpies, the latter to the jackdaws. Charitornis is known only from the Sula islands east of Celebes, and is closely related to Streptocitta. There is nothing comparable to these three groups in any of the Malay islands, and they seem to have relations rather with the Corvidæ of the old-world northern continent. Basilornis, Enodes, and Scissirostrum, are remarkable forms of Sturnidæ. Basilornis has a beautiful compressed crest, which in the allied species found in Ceram is elongated behind. Enodes has remarkable red superciliary streaks, but seems allied to Calornis. Scissirostrum seems also allied to Calornis in general structure, but has a very peculiarly formed bill and nostrils. We can hardly say whether these three forms show more affinity to Oriental or to Australian types, but they add to the weight of evidence as to the great antiquity and isolation of the Celebesian fauna. Scissirostrum has been classed with Euryceros, a Madagascar bird, and with Buphaga, an African genus; but the peculiar beak and nostrils approximate more to Cracticus and its allies, of the Australian region, which should probably form a distinct family. Ceycopsis is undoubtedly intermediate between the Malayan Ceyx and the African Ispidina, and is therefore especially interesting. Meropogon is a remarkable form of bee-eater, allied to the Indo-Malayan Nyctiornis. Prioniturus (the raquet-tailed parrots) of which two species inhabit Celebes, and one the Philippines, appears to be allied to the Austro-Malayan Geoffroyus.

We must finally notice a few genera found in Celebes, whose nearest allies are not in the surrounding islands, and which thus afford illustrations of discontinuous distribution. The most remarkable, perhaps, is Coracias, of which a fine species inhabits Celebes; while the genus is quite unknown in the Indo-Malay sub-region, and does not appear again till we reach Burmah and India; and the species has no closer affinity for Indian than for African forms. Myialestes, a small yellow flycatcher, is another example; its nearest ally (M. cinereocapilla) being a common Indian bird, but unknown in the Malay islands. The Celebesian bird described by me as Prionochilus aureolimbatus, is probably a third case of discontinuous distribution, if (as a more careful examination seems to show) it is not a Prionochilus, but congeneric with Pachyglossa, a bird only found in the Himalayas. The fine pigeon, Carpophaga forsteni, belongs to a group found in the Philippines, Australia, and New Zealand; but the Celebes species is very distinct from all the others, and seems, if anything, more allied to that of New Zealand.

The Sula islands (Sula-mangola, Sula-taliabo, and Sula-besi) lie midway between Celebes and the Moluccas, being 80 miles from the nearest part of Celebes, with several intervening islands, and 40 miles from Bouru, all open sea. Their birds show, as might be expected, a blending of the two faunas, but with a decided preponderance of that of Celebes. Out of 43 land birds which have been collected in these islands, we may deduct 6 as of wide range and no significance. Of the 37 remaining, 21 are Celebesian species, and 4 are new species but allied to those of Celebes; while there are 10 Moluccan species and 2 new species allied to those of the Moluccas. It is curious that no less than 3 Moluccan genera, quite unknown in Celebes itself, occur here,—Monarcha, Pachycephala, and Criniger; but all these, as well as several other of the Moluccan birds, are rather weak flyers, and such as are likely to have been carried across by strong winds. Of the genera, 23 are from Celebes, 10 from the Moluccas. These facts show, that the Sula islands form part of the Celebes group, although they have received an infusion of Moluccan forms, which will perhaps in time spread to the main island, and diminish the remarkable individuality that now characterises its fauna.

Insects.—Of the reptiles and fishes of Celebes we have not sufficient information to draw any satisfactory conclusions. I therefore pass to the insects of which something more is known.

The Butterflies of Celebes are not very numerous, less than 200 species in all having been collected; but a very large proportion of them, probably three-fourths of the whole, are peculiar. There is only one peculiar genus, Amechania, allied to Zethera (a group confined to the Philippine Islands), with which it should perhaps be united. Most of the genera are of wide distribution in the archipelago, or are especially Malayan, only two truly Australian genera, Elodina and Acrophthalmia, reaching Celebes. On the other hand, 7 peculiar Oriental genera are found in Celebes, but not further east, viz., Clerome, Adolias, Euripus, Apatura, Limenitis, Iolaus, and Leptocircus. There are also several indications of a direct affinity with the continent rather than with Malaya, as in the cases already enumerated among birds. A fine butterfly, yet unnamed, almost exactly resembles Dichorragia nesimachus, a Himalayan species. Euripus robustus is closely allied to E. halitherses of N. India; there are no less than 5 species of Limenitis, all quite unlike those found in other parts of the archipelago. The butterflies of Celebes are remarkably distinguished from all others in the East, by peculiarities of form, size, and colour, which run through groups of species belonging to different genera. Many Papilionidæ and Pieridæ, and some Nymphalidæ, have the anterior wings elongated, with the apex often acute, and, what is especially remarkable, an abrupt bend or shoulder near the base of the wing. (See Malay Archipelago, 3rd Ed. p. 281, woodcut.) No less than 13 species of Papilio, 10 Pieridæ, and 4 or 5 Nymphalidæ, are thus distinguished from their nearest allies in the surrounding islands or in India. In size again, a large number of Celebesian butterflies stand preeminent over their allies. The fine Papilios—adamantius, blumei, and gigon—are perfect giants by the side of the closely-allied forms of Java; while P. androcles is the largest and longest-tailed, of all the true swallow-tailed group of the Old World. Among Nymphalidæ, the species of Rhinopalpa and Euripus, peculiar to Celebes, are immensely larger than their nearest allies; and several of the Pieridæ are also decidedly larger, though in a less marked degree. In colour, many of the Celebesian butterflies differ from the nearest allied species; so that they acquire a singularity of aspect which marks them off from the rest of the group. The most curious case is that of three butterflies, belonging to three distinct genera (Cethosia myrina, Messaras mæonides, and Atella celebensis) all having a delicate violet or lilac gloss in lines or patches, which is wholly wanting in every allied species of the surrounding islands. These numerous peculiarities of Celebesian butterflies are very extraordinary; and imply isolation from surrounding lands, almost as much as do the strange forms of mammals and birds, which more prominently characterise this interesting island.

Of the Coleoptera we know much less, but a few interesting facts may be noted. There are a number of fine species of Cicindela, some of peculiar forms; and one Odontochila, a South American genus; while Collyris reaches Celebes from the Oriental region. In Carabidæ it has one peculiar genus, Dicraspeda; and a species of the fine Australian genus Catadromus. In Lucanidæ it has the Oriental genus, Odontolabris. In Cetoniidæ it has a peculiar genus, Sternoplus, and several fine Cetoniæ; but the characteristic Malayan genus, Lomaptera, found in every other island of the archipelago from Sumatra to New Guinea, is absent—an analogous fact to the case of Ceyx among birds. In Buprestidæ, the principal Austro-Malay genus, Sambus, is found here; while Sponsor, a genus 8 species of which inhabit Mauritius, has one species here and one in New Guinea. In Longicorns there are four peculiar genera, Comusia, Pytholia, Bityle, and Ombrosaga; but the most important features are the occurrence of the otherwise purely Indo-Malayan genera Agelasta, Nyctimene, and Astathes; and of the purely Austro-Malayan Arrhenotus, Trysimia, Xenolea, Amblymora, Diallus, and Ægocidnus. The remaining genera range over both portions of the archipelago. In the extensive family of Curculionidæ we can only notice the elegant genus, Celebia, allied to Eupholus, which, owing to its abundance and beauty, is a conspicuous feature in the entomology of the island.

Origin of the fauna of Celebes.—We have now to consider, briefly, what past changes of physical geography are indicated by the curious assemblage of facts here adduced. We have evidently, in Celebes, a remnant of an exceedingly ancient land, which has undergone many and varied revolutions; and the stock of ancient forms which it contains must be taken account of, when we speculate on the causes that have so curiously limited more recent immigrations. Going back to the arrival of those genera which are represented in Celebes by peculiar species, and taking first the Austro-Malay genera, we find among them such groups as Zonœnas (s.g.), Phlogœnas, Leucotreron (s.g.), and Turacœna, which are not found in the Moluccas at all; and Myzomela, found in Timor and Banda, but not in Ceram or Bouru, which are nearest to Celebes. This, combined with the curious absence of so many of the commonest Moluccan genera, leads to the conclusion that the Austro-Malay immigration took place by way of Timor and the southern part of New Guinea. It will be remembered, that to account for the Indo-Malayan forms in New Guinea, we suggested an extension of that country in a westerly direction just north of Timor. Now this is exactly what we require, to account for the stocking of Celebes with the Australian forms it possesses. At this time Borneo did not approach so near, and it was at a somewhat later period that the last great Indo-Malay migration set in; but finding the country already fairly stocked, comparatively few groups were able to establish themselves.

Going back a little farther, we come to the entrance of those few birds and insects which belong to India or Indo-China; and this probably occurred at the same time as that continental extension southward, which we found was required to account for a similar phenomenon in Java. Celebes, being more remote, received only a few stragglers. We have now to go much farther back, to the time when the ancestors of the peculiar Celebesian genera entered the country, and here our conjectures must necessarily be less defined.

On the Australian side we have to account for Megacephalon, and the other genera of purely Papuan type. It may perhaps be sufficient to say, that we do not yet know that these genera, or some very close allies, do not still exist in New Guinea; in which case they may well have entered at the same time with the species, already referred to. If, on the other hand, they are really as isolated as they appear to be, they represent an earlier communication, either by an approximation of the two islands over the space now occupied by the Moluccas; or, what is perhaps more probable, through a former extension of the Moluccas, which have since undergone so much subsidence, as to lead to the extinction of a large proportion of their ancient fauna. The wide-spread volcanic action, and especially the prevalence of raised coral-reefs in almost all the islands, render this last supposition very probable.

On the Oriental side the difficulty is greater; for here we find, what seem to be clear indications of a connection with Africa, as well as with Continental Asia, at some immensely remote epoch. Cynopithecus, Babirusa, and Anoa; Ceycopsis, Streptocitta, and Gazzola (s. g.), and perhaps Scissirostrum, may be well explained as descendants of ancestral types in their respective groups, which also gave rise to the special forms of Africa on the one hand, and of Asia on the other. For this immigration we must suppose, that at a period before the formation of the present Indo-Malay Islands, a great tract of land extended in a north-westerly direction, till it met the old Asiatic continent. This may have been before the Himalayas had risen to any great height, and when a large part of what are now the cold plateaus of Central Asia may have teemed with life, some forms of which are preserved in Africa, some in Malaya, and a few in Celebes. Here may have lived the common ancestor of Sus, Babirusa, and Phacochœrus; as well as of Cynopithecus, Cynocephalus, and Macacus; of Anoa and Bubalus; of Scissirostrum and Euryceros; of Ceyx, Ceycopsis, and Ispidina. Such an origin accounts, too, for the presence of the North-Indian forms in Celebes; and it offers less difficulties than a direct connection with continental Africa, which once appeared to be the only solution of the problem. If this south-eastward extension of Asia occurred at the same time as the north-eastward extension of South Africa and Madagascar, the two early continents may have approached each other sufficiently to have allowed of some interchange of forms: Tarsius may be the descendant of some Lemurine animal that then entered the Malayan area, while the progenitors of Cryptoprocta may then have passed from Asia to Madagascar.

It is true that we here reach the extremest limits of speculation; but when we have before us such singular phenomena as are presented by the fauna of the island of Celebes, we can hardly help endeavouring to picture to our imaginations by what past changes of land and sea (in themselves not improbable) the actual condition of things may have been brought about.

A general sketch of Australian zoology having been given in the earlier part of this chapter, it will not be necessary to occupy much time on this sub-region, which is as remarkably homogeneous as the one we have just left is heterogeneous. Although much of the northern part of Australia is within the tropics, while Victoria and Tasmania are situated from 36° to 43° south latitude, there is no striking change in the character of the fauna throughout the continent; a number of important genera extending over the whole country, and giving a very uniform character to its zoology. The eastern parts, including the colonies of New South Wales and Queensland, are undoubtedly the richest, several peculiar types being found only here. The southern portion is somewhat poorer, and has very few peculiar forms; and Tasmania being isolated is poorer still, yet its zoology has much resemblance to that of Victoria, from which country it has evidently not been very long separated. The north, as far as yet known, is characterised by hardly any peculiar forms, but by the occurrence of a number of Papuan types, which have evidently been derived from New Guinea.



Mammalia.—The Australian sub-region contains about 160 species of Mammalia, of which 3 are Monotremata, 102 Marsupials, 23 Chiroptera, 1 Carnivora (the native dog, probably not indigenous), and 31 Muridæ. The north is characterised by a species of the Austro-Malayan genus Cuscus. Phascolarctos (the koala, or native bear) is found only in the eastern districts; Phascolomys (the wombat) in the south-east and Tasmania; Petaurista (a peculiar form of flying opossum) in the east. Thylacinus (the zebra-wolf), and Sarcophilus (the "native devil"), two carnivorous marsupials, are confined to Tasmania. West Australia, the most isolated and peculiar region botanically, alone possesses the curious little honey-eating Tarsipes, and the Peragalea, or native rabbit. The remarkable Myrmecobius, a small ant-eating marsupial, is found in the west and south; and Onychogalea, a genus of kangaroos, in West and Central Australia. All the other genera have a wider distribution, as will be seen by a reference to the list at the end of this chapter.

''Plate XI. A Scene in Tasmania, with Characteristic Mammalia.''—As some of the most remarkable Mammalia of the Australian region are now found only in Tasmania, we have chosen this island for the scene of our first illustration of the fauna of the Australian sub-region. The pair of large striped animals are zebra-wolves (Thylacinus cynocephalus), the largest and most destructive of the carnivorous marsupials. These creatures used to be tolerably plentiful in Tasmania, where they are alone found. They are also called "native tigers," or "native hyænas;" and being destructive to sheep, they have been destroyed by the farmers and will doubtless soon be exterminated. In the foreground on the left is a bandicoot (Perameles gunnii). These are delicate little animals allied to the kangaroos; and they are found in all parts of Australia, and Tasmania, to which latter country this species is confined. On the right is the wombat (Phascolomys wombat), a root-eating marsupial, with large incisor teeth like those of our rodents. They inhabit south-east Australia and Tasmania. In the foreground is the porcupine ant-eater (Echidna setosa), belonging to a distinct order of mammalia, Monotremata, of which the only other member is the duck-billed Ornithorhynchus. These animals are, however, more nearly allied to the marsupials, than to the insectivora or edentata of the rest of the world, which in some respects they resemble. An allied species (Echidna hystrix) inhabits south-east Australia.

Birds.—Australia (with Tasmania) possesses about 630 species of birds, of which 485 are land-birds. Not more than about one-twentieth of these are found elsewhere, so that it has a larger proportion of endemic species than any other sub-region on the globe. These birds are divided among the several orders as follows:

The Psittaci, we see, are very richly represented, while the Picariæ are comparatively few; and the Columbæ are scarce as compared with their abundance in the Austro-Malay sub-region.

Birds seem to be very evenly distributed over all Australia; comparatively few genera of importance being locally restricted. In the eastern districts alone, we find Origma, and Orthonyx (Sylviidæ); Sericulus and Ptilorhynchus (Paradiseidæ); Leucosarcia (Columbidæ); and Talegalla (Megapodiidæ). Nectarinia, Pitta, Ptilorhis, Chlamydodera, and Sphecotheres, range from the north down the east coasts. Nanodes (Psittacidæ), and Lipoa (Megapodiidæ), are southern forms, the first extending to Tasmania; which island appears to possess no peculiar genus of birds except Eudyptes, one of the penguins. West Australia has no wholly peculiar genus except Geopsittacus, a curious form of ground parroquet; the singular Atrichia, first found here, having been discovered in the east. In North Australia, Emblema (Ploceidæ) is the only peculiar Australian genus, but several Austro-Malayan and Papuan genera enter,—as, Syma and Tanysiptera (Alcedinidæ); Machærirhynchus (Muscicapidæ); Calornis (Sturnidæ); Manucodia, Ptilorhis, and Ælurœdus (Paradiseidæ); Megapodius; and Casuarius. The presence of a species of bustard (Eupodotis) in Australia, is very curious, its nearest allies being in the plains of India and Africa. Among waders the genus Tribonyx, a thick-legged bird somewhat resembling the Notornis of New Zealand, though not closely allied to it, is the most remarkable. The district where the typical Australian forms most abound is undoubtedly the eastern side of the island. The north and south are both somewhat poorer, the west much poorer, although it possesses a few very peculiar forms, especially among Mammalia. Tasmania is the poorest of all, a considerable number of genera being here wanting; but, except the two peculiar carnivorous marsupials, it possesses nothing to mark it off zoologically from the adjacent parts of the main land. It is probable that its insular climate, more moist and less variable than that of Australia, may not be suitable to some of the absent forms; while others may require more space and more varied conditions, than are offered by a comparatively small island.

The remaining classes of animals have been already discussed in our sketch of the region as a whole (p. 396).

''Plate XII. Illustrating the Fauna of Australia.''—In this plate we take New South Wales as our locality, and represent chiefly, the more remarkable Australian types of birds. The most conspicuous figure is the wonderful lyre-bird (Menura superba), the elegant plumage of whose tail is altogether unique in the whole class of birds. The unadorned bird is the female. In the centre is the emu (Dromæus novæ-hollandiæ), the representative in Australia, of the ostrich in Africa and America, but belonging to a different family, the Casuariidæ. To the right are a pair of crested pigeons (Ocyphaps lophotes), one of the many singular forms of the pigeon family to which the Australian region gives birth. In every other part of the globe pigeons are smooth-headed birds, but here they have developed three distinct forms of crest, as seen in this bird, the crowned pigeon figured in Plate X., and the double-crested pigeon (Lopholaimus antarcticus). The large bird on the tree is one of the Australian frog-mouthed goat-suckers (Podargus strigoides), which are called in the colony "More-pork," from their peculiar cry. They do not capture their prey on the wing like true goat-suckers, but hunt about the branches of trees at dusk, for large insects, and also for unfledged birds. A large kangaroo (Macropus giganteus) is seen in the distance; and passing through the air, a flying opossum (Petaurus sciureus), a beautiful modification of a marsupial, so as to resemble in form and habits the flying squirrels of the northern hemisphere.

Although the area of this sub-region is so vast, and the number of islands it contains almost innumerable, there is a considerable amount of uniformity in its forms of animal life. From the Ladrone islands on the west, to the Marquesas on the east, a distance of more than 5,000 miles, the same characteristic genera of birds prevail; and this is the only class of animals on which we can depend, mammalia being quite absent, and reptiles very scarce. The Sandwich Islands, however, form an exception to this uniformity; and, as far as we yet know, they are so peculiar that they ought, perhaps, to form a separate sub-region. They are, however, geographically a part of Polynesia; and a more careful investigation of their natural history may show more points of agreement with the other islands. It is therefore a matter of convenience, at present, to keep them in the Polynesian sub-region, which may be divided into Polynesia proper and the Sandwich Islands.



Polynesia proper consists of a number of groups of islands of some importance, and a host of smaller intermediate islets. For the purpose of zoological comparison, we may class them in four main divisions. 1. The Ladrone and Caroline Islands; 2. New Caledonia and the New Hebrides; 3. The Fiji, Tonga, and Samoa Islands; 4. The Society, and Marquesas Islands. The typical Polynesian fauna is most developed in the third division; and it will be well to describe this first, and then show how the other islands diverge from it, and approximate other sub-regions.

Fiji, Tonga, and Samoa Islands.—The land-birds inhabiting these islands belong to 41 genera, of which 17 are characteristic of the Australian region, and 9 more peculiarly Polynesian. The characteristic Australian genera are the following: Petroica (Sylviidæ); Lalage (Campephagidæ); Monarcha, Myiagra, Rhipidura (Muscicapidæ); Pachycephala (Pachycephalidæ); Rectes (Laniidæ); Myzomela, Ptilotis, Anthochæra (Meliphagidæ); Amadina, Erythrura, (Ploceidæ); Artamus (Artamidæ); Lorius (Trichoglossidæ); Ptilopus, Phlogœnas (Columbidæ); Megapodius (Megapodiidæ).

The peculiar Polynesian genera are:—Tatare, Lamprolia (Sylviidæ); Aplonis, Sturnoides (Sturnidæ); Todiramphus (Alcedinidæ); Pyrrhulopsis, Cyanoramphus, (Platycercidæ); Coriphilus (Trichoglossidæ); Didunculus (Didunculidæ).

The wide-spread genera are Turdus, Zosterops, Hirundo, Halcyon, Collocalia, Eudynamis, Cuculus, Ianthœnas, Carpophaga, Turtur, Haliæetus, Astur, Circus, Strix, Asio. The aquatic birds are fifteen in number, all wide-spread species except one—a form of moor-hen (Gallinulidæ), which has been constituted a new genus Pareudiastes.

Society, and Marquesas Islands.—Here, the number of genera of land-birds has considerably diminished, amounting only to 16 in all. The characteristic Australian genera are 5;—Monarcha, Anthochæra, Trichoglossus, Ptilopus, and Phlogœnas. The Polynesian genera are 4;—Tatare, Todiramphus, Cyanoramphus, Coriphilus, and one recently described genus, Serresius, an extraordinary form of large fruit pigeon, here classed under Carpophaga. These remote groups have thus all the character of Oceanic islands, even as regards the rest of Polynesia, since they possess hardly anything, but what they might have received by immigration over a wide extent of ocean.

Ladrone, and Caroline Islands.—These extensive groups of small islands are very imperfectly known, yet a considerable number of birds have been obtained. They possess two peculiar Polynesian genera, Tatare and Sturnoides; one peculiar sub-genus, Psammathia (here included under Acrocephalus); and ten of the typical Australian genera found in Polynesia,—Lalage, Monarcha, Myiagra, Rhipidura, Myzomela, Erythrura, Artamus, Phlogœnas, Ptilopus, and Megapodius, as well as the Papuan genus Rectes, and the Malayan Calornis;—so that they can be certainly placed in the sub-region. Genera which do not occur in the other Polynesian islands are, Acrocephalus, (s.g. Psammathia) originally derived perhaps from the Philippines; and Caprimulgus, a peculiar species, allied to one from Japan.

New Caledonia, and the New Hebrides.—Although these islands seem best placed with Polynesia, yet they form a transition to Australia proper, and to the Papuan group. They possess 30 genera of land-birds, 18 of which are typical of the Australian region; but while 13 are also Polynesian, there are 5 which do not pass further east. These are Acanthiza, Eopsaltria, Gliciphila, Philemon, and Ianthœnas. The peculiar Polynesian genus, Aplonis, of which three species inhabit New Caledonia, link it to the other portions of the sub-region. The following are the genera at present known from New Caledonia:—Turdus, Acanthiza, Campephaga, Lalage, Myiagra, Rhipidura, Pachycephala, Eopsaltria, Corvus, Physocorax (s.g. of Corvus, allied to the jackdaws), Gliciphila, Anthochæra, Philemon, Zosterops, Erythrura, Aplonis, Artamus, Cuculus, Halcyon, Collocalia, Cyanoramphus, Trichoglossus, Ptilopus, Carpophaga, Macropygia, Ianthœnas, Chalcophaps, Haliastur, Accipiter. The curious Rhinochetus jubatus, forming the type of a distinct family of birds (Rhinochetidæ), allied to the herons, is only known from New Caledonia.

It thus appears, that not more than about 50 genera and 150 species of land-birds, are known from the vast number of islands that are scattered over the Central Pacific, and it is not probable that the number will be very largely increased. Some of the species, as the Eudynamis taitensis and Tatare longirostris, range over 40° of longitude, from the Fiji Islands to the Marquesas. In other genera, as Cyanoramphus and Ptilopus, each important island or group of islands, has its peculiar species. The connection of all these islands with each other, on the one hand, and their close relation to the Australian region, on the other, are equally apparent; but we have no sufficient materials for speculating with any success, on the long series of changes that have brought about their existing condition, as regards their peculiar forms of animal life.

Sandwich Islands.—This somewhat extensive group of large islands, is only known to contain 11 genera and 18 species of indigenous land-birds; and even of this small number, two birds of prey are wide ranging species, which may well have reached the islands during their present isolated condition. These latter are, Strix delicatula, an owl spread over Australia and the Pacific; and Asio accipitrinus, a species which has reached the Galapagos from S. America, and thence perhaps the Sandwich Islands. Of the remaining 8 genera, one is a crow (Corvus hawaiensis), and another a fishing eagle (Pandion solitarius), of peculiar species; leaving 7 genera, which are all (according to Mr. Sclater) peculiar. First we have Chasiempis, a genus of Muscicapidæ, containing two species (which may however belong to distinct genera); and as the entire family is unknown on the American continent these birds must almost certainly be allied to some of the numerous Muscicapine forms of the Australian region. Next we have the purely Australian family Meliphagidæ, represented by two genera,—Moho, an isolated form, and Chætoptila, a genus established by Mr. Sclater for a bird before classed in Entomyza, an Australian group. The four remaining genera are believed by Mr. Sclater to belong to one group, the Drepanididæ, altogether confined to the Sandwich Islands. Two of them, Drepanis and Hemignathus, with three species each, are undoubtedly allied; the other two, Loxops and Psittirostra, have usually been classed as finches. The former seem to approach the Dicæidæ; and all resemble this group in their coloration,

The aquatic birds and waders all belong to wide-spread genera, and only one or two are peculiar species.

The Sandwich Islands thus possess a larger proportion of peculiar genera and species of land-birds than any other group of islands, and they are even more strikingly characterised by what seems to be a peculiar family. The only other class of terrestrial animals at all adequately represented on these islands, are the land shells; and here too we find a peculiar family, sub-family, or genus (Achatinella or Achatinellidæ) consisting of a number of genera, or sub-genera,—according to the divergent views of modern conchologists,—and nearly 300 species. The Rev. J. T. Gulick, who has made a special study of these shells on the spot, considers that there are 10 genera, some of which are confined to single islands. The species are so restricted that their average range is not more than five or six square miles, while some are confined to a tract of only two square miles in extent, and very few range over an entire island. Some species are confined to the mountain ridges, others to the valleys; and each ridge or valley possesses its peculiar species. Considerably more than half the species occur in the island of Oahu, where there is a good deal of forest. Very few shells belonging to other groups occur, and they are all small and obscure; the Achatinellæ almost monopolising the entire archipelago.

Remarks on the probable past history of the Sandwich Islands.—The existence of these peculiar groups of birds and land-shells in so remote a group of volcanic islands, clearly indicates that they are but the relics of a more extensive land; and the reefs and islets that stretch for more than 1,000 miles in a west-north-west direction, may be the remains of a country once sufficiently extensive to develope these and many other, now extinct, forms of life.

Some light may perhaps be thrown on the past history of the Sandwich Islands, by the peculiar plants which are found on their mountains. The peak of Teneriffe produces no Alpine plants of European type, and this has been considered to prove that it has been always isolated; whereas the occurrence of North Temperate forms on the mountains of Java, accords with other evidence of this island having once formed part of the Asiatic continent. Now on the higher summits of the Sandwich Islands, nearly 30 genera of Arctic and North Temperate flowering plants have been found. Many of these occur also in the South Temperate zone, in Australia or New Zealand; but there are others which seem plainly to point to a former connection with some North Temperate land, probably California, as a number of islets are scattered in the ocean between the two countries. The most interesting genera are the following:—Silene, which is wholly North Temperate, except that it occurs in S. Africa; Vicia, also North Temperate, and in South Temperate America; Fragaria, with a similar distribution; Aster, widely spread in America, otherwise North Temperate only; Vaccinium, wholly confined to the northern hemisphere, in cold and temperate climates. None of these are found in Australia or New Zealand; and their presence in the Sandwich Islands seems clearly to indicate a former approximation to North Temperate America, although the absence of any American forms of vertebrata renders it certain that no actual land connection ever took place.

Recent soundings have shown, that the Sandwich Islands rise from a sea which is 3,000 fathoms or 18,000 feet deep; while there is a depth of at least 2,000 fathoms all across to California on one side, and to Japan on the other. Between the Fiji Islands, New Caledonia, the Solomon Islands, and Australia, the depth is about 1,300 fathoms, and between Sydney and New Zealand 2,600 fathoms; showing, in every case, a general accordance between the depth of sea and the approximation of the several faunas. In a few more years, when it is to be hoped we shall know the contour of the sea-bottom better than that of the continents, we shall be able to arrive at more definite and trustworthy conclusions as to the probable changes of land and sea by which the phenomena of animal distribution in the Pacific have been brought about.

Reptiles of the Polynesian Sub-region.—The researches of Mr. Darwin on Coral Islands, proved, that large areas in the Pacific Ocean have been recently subsiding; but the peculiar forms of life which they present, no less clearly indicate the former existence of some extensive lands. The total absence of Mammalia, however, shows either that these lands never formed part of the Australian or Papuan continents, or if they did, that they have been since subjected to such an amount of subsidence as to exterminate most of their higher terrestrial forms of life. It is a remarkable circumstance, that although Mammalia (except bats) are wanting, there are a considerable number of reptiles ranging over the whole sub-region. Lizards are the most numerous, five families and fourteen genera being represented, as follows:—

The first five are wide-spread genera, represented mostly by peculiar species; but sometimes the species themselves have a wide range, as in the case of Ablepharus pœcilopleurus, which (according to Dr. Günther) is found in Timor, Australia, New Caledonia, Savage Island (one of the Samoa group), and the Sandwich Islands! Gehyra and Heteronota are Australian genera; while Lophura has reached the Pelew Islands from the Moluccas. The remainder (printed in italics), are peculiar genera; Brachylophus being especially interesting as an example of an otherwise peculiar American family, occurring so far across the Pacific.

Snakes are much less abundant, only four genera being represented, one of them marine. They are, Anoplodipsas, a peculiar genus of Amblycephalidæ from New Caledonia; Enygrus, a genus of Pythonidæ from the Fiji Islands; Ogmodon, a peculiar genus of Elapidæ, also from the Fiji Islands, but ranging to Papua and the Moluccas; and Platurus, a wide-spread genus of sea-snakes (Hydrophidæ). In the more remote Sandwich and Society Islands there appear to be no snakes. This accords with our conclusion that lizards have some special means of dispersal over the ocean which detracts from their value as indicating zoo-geographical affinities; which is further proved by the marvellous range of a single species (referred to above) from Australia to the Sandwich Islands.

A species of Hyla is said to inhabit the New Hebrides, and several species of Platymantis (tree-frogs) are found in the Fiji Islands; but otherwise the Amphibians appear to be unrepresented in the sub-region, though they will most likely be found in so large an island as New Caledonia.

From the foregoing sketch, it appears, that although the reptiles present some special features, they agree on the whole with the birds, in showing, that the islands of Polynesia all belong to the Australian region, and that in the Fiji Islands is to be found the fullest development of their peculiar fauna.

The islands of New Zealand are more completely oceanic than any other extensive tract of land, being about 1,200 miles from Australia and nearly the same distance from New Caledonia and the Friendly Isles. There are, however, several islets scattered around, whose productions show that they belong to the same sub-region;—the principal being, Norfolk Island, Lord Howe's Island, and the Kermadec Isles, on the north; Chatham Island on the east; the Auckland and Macquarie Isles on the south;—and if these were once joined to New Zealand, there would have been formed an island-continent not much inferior in extent to Australia itself.

New Zealand is wholly situated in the warmer portion of the Temperate zone, and enjoys an exceptionally mild and equable climate. It has abundant moisture, and thus comes within the limits of the South-Temperate forest zone; and this leads to its productions often resembling those of the tropical, but moist and wooded, islands of the Pacific, rather than those of the temperate, but arid and scantily wooded plains of Australia. The two islands of New Zealand are about the same extent (approximately) as the British Isles, but the difference in the general features of their natural history is very great. There are, in the former, no mammalia, less than half as many birds, very few reptiles and fresh-water fishes, and an excessive and most unintelligible poverty of insects; yet, considering the situation of the islands and their evidently long-continued isolation, the wonder rather is that their fauna is so varied and interesting as it is found to be. Our knowledge of this fauna, though no doubt far from complete, is sufficiently ample; and it will be well to give a pretty full account of it, in order to see what conclusions may be drawn as to its origin.

Mammalia.—The only mammals positively known as indigenous to New Zealand are two bats, both peculiar to it,—Scotophilus tuberculatus and Mystacina tuberculata. The former is allied to Australian forms; the latter is more interesting, as being a peculiar genus of the family Noctilionidæ, which does not exist in Australia; and in having decided resemblances to the Phyllostomidæ of South America, so that it may almost be considered to be a connecting link between the two families. A forest rat is said to have once abounded on the islands, and to have been used for food by the natives; but there is much doubt as to what it really was, and whether it was not an introduced species. The seals are wide-spread antarctic forms which have no geographical significance.

Birds.—About 145 species of birds are natives of New Zealand, of which 88 are waders or aquatics, leaving 57 land-birds belonging to 34 genera. Of this latter number, 16, or nearly half, are peculiar; and there are also 5 peculiar genera of waders and aquatic birds, making 21 in all. Of the remaining genera of land-birds, four are cosmopolite or of very wide range, while the remainder are characteristic of the Australian region. The following is a list of the Australian genera found in New Zealand: Sphenæacus, Gerygone, Orthonyx (Sylviidæ); Graucalus (Campephagidæ); Rhipidura (Muscicapidæ); Anthochæra (Meliphagidæ); Zosterops (Dicæidæ); Cyanoramphus (Platycercidæ); Carpophaga (Columbidæ); Hieracidea (Falconidæ); Tribonyx (Rallidæ). Besides these there are several genera of wide range, as follows:—Anthus (Motacillidæ); Hirundo (Hirundinidæ); Chrysococcyx, Eudynamis (Cuculidæ); Halcyon (Alcedinidæ); Coturnix (Tetraonidæ); Circus (Falconidæ); Athene (Strigidæ).

Most of the above genera are represented by peculiar New Zealand species, but in several cases the species are identical with those of Australia, as in the following: Anthochæra carunculata, Zosterops lateralis, Hirundo nigricans, and Chrysococcyx lucidus; also one—Eudynamis taitensis—which is Polynesian.

We now come to the genera peculiar to New Zealand, which are of especial interest:

We have thus a wonderful amount of speciality; yet the affinities of the fauna, whenever they can be traced, are with Australia or Polynesia. Nine genera of New Zealand birds are characteristically Australian, and the eight genera of wide range are Australian also. Of the peculiar genera, 7 or 8 are undoubtedly allied to Australian groups. There are also four Australian and one Polynesian species. Even the peculiar family, Nestoridæ is allied to the Australian Trichoglossidæ. We have therefore every gradation of similarity to the Australian fauna, from identical species, through identical genera, and allied genera, to distinct but allied families; clearly indicating very long continued yet rare immigrations from Australia or Polynesia; immigrations which are continued down to our day. For resident ornithologists believe, that the Zosterops lateralis has found its way to New Zealand within the last few years, and that the two cuckoos now migrate annually, the one from Australia, the other from some part of Polynesia, distances of more than 1,000 miles! These facts seem, however, to have been accepted on insufficient evidence and to be in themselves extremely improbable. It is observed that the cuckoos appear annually in certain districts and again disappear; but their course does not seem to have been traced, still less have they ever been actually seen arriving or departing across the ocean. In a country which has still such wide tracts of unsettled land, it is very possible that the birds in question may only move from one part of the islands to another.

We will here notice the smaller islands belonging to the sub-region, as it is chiefly their birds that possess any interest.

Norfolk Island.—The land-birds recorded from this island amount to 15 species, of which 8 are Australian, viz.: Climacteris scandens, Symmorphus leucopygius, Zosterops tenuirostris and Z. albogularis, Halcyon sanctus, Platycercus pennanti, Carpophaga spadicea, Phapspicata and P. chalcoptera. Of the peculiar species three belong to Australian genera; Petroica, Gerygone, and Rhipidura; one to a cosmopolitan genus, Turdus. So far the affinity seems to be all Australian, and there remain only three birds which ally this island to New Zealand,—Nestor productus, Cyanoramphus rayneri, and Notornis alba. The former inhabited the small Phillip Island (close to Norfolk Island) but is now extinct. Being a typical New Zealand genus, quite incapable of flying across the sea, its presence necessitates some former connexion between the two islands, and it is therefore perhaps of more weight than all the Australian genera and species, which are birds capable of long flights. The Cyanoramphus is allied to a New Zealand broad-tailed parroquet. The Notornis alba is extinct, but two specimens exist in museums, and it is even a stronger case than the Nestor, as showing a former approximation or union of this island with New Zealand. A beautiful figure of this bird is given in the Ibis for 1873.

Lord Howe's Island.—This small island, situated half-way between Australia and Norfolk Island, is interesting, as containing a peculiar species of the New Zealand genus Ocydromus, or wood-hen (O. sylvestris). There is also a peculiar thrush, Turdus vinitinctus. Its other birds are wholly of Australian types, and most of them probably Australian species. The following have been observed, and no doubt constitute nearly its whole indigenous bird fauna. Acanthiza sp., Rhipidura sp., Pachycephala gutturalis, Zosterops strenuus and Z. tephropleurus, Strepera sp., Halcyon sp., and Chalcophaga chrysochlora. The two species of Zosterops are peculiar. The Ocydromus is important enough to ally this island to New Zealand rather than to Australia; and if the white bird seen there is, as supposed, the Notornis alba which is extinct in Norfolk Island, the connection will be rendered still more clear.

Chatham Islands.—These small islands, 450 miles east of New Zealand, possess about 40 species of birds, of which 13 are land-birds. All but one belong to New Zealand genera, and all but five are New Zealand species. The following are the genera of the land-birds: Sphenæacus, Gerygone, Myiomoira, Rhipidura, Zosterops, Anthus, Prosthemadera, Anthornis, Chrysococcyx, Cyanoramphus, Carpophaga, Circus. The peculiar species are Anthornis melanocephala, Myiomoira diffenbachi and M. traversi, Rhipidura flabellifera, and a peculiar rail incapable of flight, named by Captain Hutton Cabalus modestus. It is stated that the Zosterops differs from that of New Zealand, and is also a migrant; and it is therefore believed to come every year from Australia, passing over New Zealand, a distance of nearly 1,700 miles! Further investigation will perhaps discover some other explanation of the facts. It is also stated, that the pigeon and one of the small birds (? Gerygone or Zosterops) have arrived at the islands within the last eight years. The natives further declare, that both the Stringops and Apteryx once inhabited the islands, but were exterminated about the year 1835.

The Auckland Islands.—These are situated nearly 300 miles south of New Zealand, and possess six land-birds, of which three are peculiar,—Anthus aucklandicus, Cyanoramphus aucklandicus, and C. malherbii, the others being New Zealand species of Myiomoira, Prosthemadera, and Anthornis. It is remarkable that two peculiar parrots of the same genus should inhabit these small islands; but such localities seem favourable to the Platycercidæ, for another peculiar species is found in the remote Macquarie Islands, more than 400 miles farther south. A peculiar species and genus of ducks, Nesonetta aucklandica, is also found here, and as far as yet known, nowhere else. A species of the northern genus Mergus is also found on these islands, and has been recently obtained by Baron von Hügel.



''Plate XIII. Illustrating the peculiar Ornithology of New Zealand.''—Our artist has here depicted a group of the most remarkable and characteristic of the New Zealand birds. In the middle foreground is the Owl-parrot or Kakapoe (Stringops habroptilus), a nocturnal burrowing parrot, that feeds on fern-shoots, roots, berries, and occasionally lizards; that climbs but does not fly; and that has an owl-like mottled plumage and facial disc. The wings however are not rudimentary, but fully developed; and it seems to be only the muscles that have become useless for want of exercise. This would imply, that these birds have not long been inhabitants of New Zealand only, but were developed in other countries (perhaps Australia) where their wings were of use to them.

Beyond the Kakapoe are a pair of the large rails, Notornis mantelli; heavy birds with short wings quite useless for flight, and with massive feet and bill of a red colour. On the right is a pair of Kiwis (Apteryx australis), one of the queerest and most unbird-like of living birds. It has very small and rudimentary wings, entirely concealed by the hair-like plumage, and no tail. It is nocturnal, feeding chiefly on worms, which it extracts from soft earth by means of its long bill. The genus Apteryx forms a distinct family of birds, of which four species are now known, besides some which are extinct. They are allied to the Cassowary and to the gigantic extinct Dinornis. On the wing are a pair of Crook-billed Plovers (Anarhynchus frontalis), remarkable for being the only birds known which have the bill bent sideways. This was at first thought to be a malformation; but it is now proved to be a constant character of the species, as it exists even in the young chicks; yet the purpose served by such an anomalous structure is not yet discovered. No country on the globe can offer such an extraordinary set of birds as are here depicted.

Reptiles.—These consist almost wholly of lizards, there being no land-snakes and only one frog. Twelve species of lizards are known, belonging to three genera, one of which is peculiar, as are all the species. Hinulia, with two species, and Mocoa, with four species (one of which extends to the Chatham Islands), belong to the Scincidæ; both are very wide-spread genera and occur in Australia. The peculiar genus Naultinus, with six species, belongs to the Geckotidæ, a family spread over the whole world.

The most extraordinary and interesting reptile of New Zealand is, however, the Hatteria punctata, a lizard-like animal living in holes, and found in small islands on the north-east coast, and more rarely on the main land. It is somewhat intermediate in structure between lizards and crocodiles, and also has bird-like characters in the form of its ribs. It constitutes, not only a distinct family, Rhynchocephalidæ, but a separate order of reptiles,. It is quite isolated from all other members of the class; and is probably a slightly modified representative of an ancient and generalised form, which has been superseded in larger areas by the more specialized lizards and saurians.

The only representatives of the Ophidia are two sea-snakes of Australian and Polynesian species, and of no geographical interest.

Amphibia.—The solitary frog indigenous to New Zealand, belongs to a peculiar genus, Liopelma, and to the family Bomburatoridæ, otherwise confined to Europe and temperate South America.

Fresh-water Fishes.—There are, according to Captain Hutton, 15 species of fresh-water fish in New Zealand, belonging to 7 genera; six species, and one genus (Retropinna), being peculiar. Retropinna richardsoni belongs to the Salmonidæ, and is the only example of that family occurring in the Southern hemisphere, where it is confined to New Zealand and the Chatham Islands. The wide distribution of Galaxias attenuatus—from the Chatham Islands to South America—has already been noticed; while another species, G. fasciatus, is found in the Chatham and Auckland Isles as well as New Zealand. A second genus peculiar to New Zealand, Neochanna, allied to Galaxias, has recently been described. Prototroctes oxyrhynchus is allied to an Australian species, but belongs to a family (Haplochitonidæ) which is otherwise South American. An eel, Anguilla latirostris, is found in Europe, China, and the West Indies, as well as in New Zealand! while the genus Agonostoma ranges to Australia, Celebes, Mauritius, and Central America.

Insects.—The great poverty of this class is well shown by the fact, that only eleven species of butterflies are known to inhabit New Zealand. Of these, six are peculiar, and one, Argyrophenga (Satyridæ), is a peculiar genus allied to the Northern genus Erebia. The rest are either of wide range, as Pyrameis cardui and Diadema bolina; or Australian, as Hamadryas zoilus; while one, Danais erippus, is American, but has also occurred in Australia, and is no doubt a recent introduction into both countries. Only one Sphinx is recorded, and no other species of the Sphingina except the British currant-moth, Ægeria tipuliformis, doubtless imported. Coleoptera are better represented, nearly 300 species having been described, all or nearly all being peculiar. These belong to about 150 genera, of which more than 50 are peculiar. No less than 14 peculiar genera belong to the Carabidæ, mostly consisting of one or two species, but Demetrida has 3, and Metaglymma 8 species. Other important genera are Dicrochile, Homalosoma, Mecodema, and Scopodes, all in common with Australia. Mecodema and Metaglymma are the largest genera. Even the Auckland Islands have two small genera of Carabidæ found nowhere else.

Cicindelidæ are represented in New Zealand by 6 species of Cicindela, and 1 of Distypsidera, a genus peculiar to the Australian region.

The Lucanidæ are represented by two peculiar genera, Dendroblax and Oxyomus; two Australian genera, Lissotes and Ceratognathus; and by the almost cosmopolite Dorcus.

The Scarabeidæ consist of ten species only, belonging to four genera, two of which are peculiar (Odontria and Stethaspis); and two Australian (Pericoptus and Calonota). There are no Cetoniidæ.

There is only one Buprestid, belonging to the Australian genus Cisseis. The Elateridæ, (about a dozen species,) belong mostly to Australian genera, but two, Metablax and Ochosternus, are peculiar.

There are 30 species of Curculionidæ, belonging to 22 genera. Of the genera, 12 are peculiar; 1 is common to New Zealand and New Caledonia; 5 belong to the Australian region, and the rest are widely distributed.

Longicorns are, next to Carabidæ, the most numerous family, there being, according to Mr. Bates (Ann. Nat. Hist., 1874), about 35 genera, of which 26 are peculiar or highly characteristic, and 7 of the others Australian. The largest and most characteristic genera are Æmona and Xyloteles, both being peculiar to New Zealand; few of the remainder having more than one or two species. Demonax extends to the Moluccas and S. E. Asia. A dozen of the genera have no near relations with those of any other country.

Phytophaga are remarkably scarce, only two species of Colaspis being recorded; and there is only a single species of Coccinella.

The other orders of Insects appear to be equally deficient. Hymenoptera are very poorly represented, only a score of species being yet known; but two of the genera are peculiar, as are all the species. The Neuroptera and Heteroptera are also very scarce, and several of the species are wide-spread forms of the Australian region. The few species of Homoptera are all peculiar. The Myriapoda afford some interesting facts. There are nine or ten species, all peculiar. One genus, Lithobius, ranges over the northern hemisphere as far south as Singapore, and probably through the Malay Archipelago, but is not found in Australia. Henicops occurs elsewhere only in Tasmania and Chili. Cryptops, only in the north temperate zone; while two others, Cermatia and Cormocephalus, both occur in Australia.

Land-Shells.—Of these, 114 species are known, 97 being peculiar. Three species of Helix are also found in Australia, and five more in various tropical islands of the Pacific. Nanina, Lymnæa, and Assiminea, are found in Polynesia or Malaya, but not in Australia. Amphibola is an Australian genus, as is Janella. Testacella and Limax belong to the Palæarctic region.

From the Chatham Islands, 82 species of shells are known, all being New Zealand species, except nine, which are peculiar.

The Ancient Fauna of New Zealand.—One of the most remarkable features of the New Zealand fauna, is the existence, till quite recent times, of an extensive group of wingless birds,—called Moas by the natives—many of them of gigantic size, and which evidently occupied the place which, in other countries, is filled by the mammalia. The most recent account of these singular remains, is that by Dr. Haast, who, from a study of the extensive series of specimens in the Canterbury museum, believes, that they belong to two families, distinguished by important differences of structure, and constitute four genera,—Dinornis and Meiornis, forming the family Dinornithidæ; Palapteryx and Euryapteryx, forming the family Palapterygidæ. These were mostly larger birds than the living Apteryx, and some of them much larger even than the African ostrich, and were more allied to the Casuariidæ and Struthionidæ than to the Apterygidæ. No less than eleven species of these birds have been discovered; all are of recent geological date, and there are indications that some of them may have been in existence less than a century ago, and were really exterminated by man. Remains have been found (of apparently the same recent date) of species of Apteryx, Stringops, Ocydromus, and many other living forms, as well as of Harpagornis, a large bird of prey, and Cnemiornis, a gigantic goose. Bodies of the Hatteria punctata have also been found along with those of the Moa, showing that this remarkable reptile was once more abundant on the main islands than it is now.

The Origin of the New Zealand Fauna.—Having now given an outline sketch of the main features of the New Zealand fauna and of its relations with other regions, we may consider what conclusions are fairly deducible from the facts. As the outlying Norfolk, Chatham, and Lord Howe's Islands, are all inhabited (or have recently been so) by birds of New Zealand type or even identical species, almost incapable of flight, we may infer that these islands show us the former minimum extent of the land-area in which the peculiar forms which characterise the sub-region were developed. If we include the Auckland and Macquarie Islands to the south, we shall have a territory of not much less extent than Australia, and separated from it by perhaps several hundred miles of ocean. Some such ancient land must have existed to allow of the development and specialization of so many peculiar forms of birds, and it probably remained with but slight modifications for a considerable geological period. During all this time it would interchange many of its forms of life with Australia, and there would arise that amount of identity of genera between the two countries which we find to exist. Its extension southwards, perhaps considerably beyond the Macquaries, would bring it within the range of floating ice during colder epochs, and within easy reach of the antarctic continent during the warm periods; and thus would arise that interchange of genera and species with South America, which forms one of the characteristic features of the natural history of New Zealand.

Captain F. W. Hutton (to whose interesting paper on the Geographical relations of the New Zealand Fauna we are indebted for some of our facts) insists upon the necessity of former land-connections in various directions, and especially of an early southern continental period, when New Zealand, Australia, Southern Africa, and South America, were united. Thus he would account for the existence of Struthious birds in all these countries, and for the various other groups of birds, reptiles, fishes, or insects which have no obvious means of traversing the ocean,—and this union must have occurred before mammalia existed in any of these countries. But such a supposition is quite unnecessary, if we consider that all wingless land-birds and some water-birds (as the Gare-fowl and Steamer Duck) are probably cases of abortion of useless organs, and that the common ancestors of the various forms of Struthiones may have been capable of a moderate degree of flight; or they may have originated in the northern hemisphere, as already explained in Chap. XI. p. 287. The existence of two, if not three, distinct families of these birds in New Zealand, proves that the original type was here isolated at a very early date, and being wholly free from the competition of mammalia, became more differentiated than elsewhere. The Hatteria is probably coeval with these early forms, and is the only relic of a whole order of reptiles, which once perhaps ranged far over the globe.

Still less does any other form of animal inhabiting New Zealand, require a land connection with distant countries to account for its presence. With the example before us of the Bermudas and Azores, to which a great variety of birds fly annually over vast distances, and even of the recent arrival of new birds in New Zealand and Chatham Island, we may be sure that the ancestors of every New Zealand bird could easily have reached its shores during the countless ages which elapsed while the Dinornis and Apteryx were developing. The wonderful range of some of the existing species of lizards and fresh-water fish, as already given, proves that they too possess means of dispersal which have sufficed to spread them, within a comparatively recent period, over countries separated by thousands of miles of ocean; and the fact that a group like the snakes, so widely distributed and for which the climate of New Zealand is so well adapted, does not exist there, is an additional proof that land connection had nothing to do with the introduction of the existing fauna. We have already (p. 398), discussed in some detail the various modes in which the dispersal of animals in the southern hemisphere has been effected; and in accordance with the principles there established, we conclude, that the New Zealand fauna, living and extinct, demonstrates the existence of an extensive tract of land in the vicinity of Australia, Polynesia, and the Antarctic continent, without having been once actually connected with either of these countries, since the period when mammalia had peopled all the great continents. That event certainly dates back to Secondary, if not to Palæozoic, times, because so dominant a group must soon have spread over the whole continuous land-area of the globe. We have no reason for believing that birds were an earlier development; and certainly cannot, with any probability, place the origin of the Struthiones before that of Mammals.

Causes of the Poverty of Insect-life in New Zealand: its Influence on the Character of the Flora.—The extreme paucity of insects in New Zealand, to which we have already alluded, seems to call for some attempt at explanation. No other country in the world, in which the conditions are equally favourable for insect-life, and which has either been connected with, or is in proximity to, any of the large masses of land, presents a similar phenomenon. The only approach to it is in the Galapagos, and in some of the islands of the Pacific; and in each of these cases the absence of mammals leads us to infer, that no connection with a continent has ever taken place. Yet the fauna of New Zealand evidently dates back to a remote geological epoch, and it seems strange that an abundance of indigenous insects have not been developed, especially when we consider the vast antiquity that most of the orders and families, and many of the genera, of insects possess (see p. 166), and that they must always have reached the country in greater numbers and variety than any of the higher animals. The undoubted fact that such an indigenous insect-fauna has not arisen, would therefore lead us to conclude, that insects find the conditions requisite for their development only in the great continental masses of land, in strict adaptation to, and dependance on, a varied fauna and flora of ever-increasing richness and complexity. A small number of widely-separated forms, introduced into a country where the fauna and flora are alike scanty and unrelated to them, seem to have little tendency to vary and branch out into that vast network of insect-life which enriches all the great continents and their once connected islands.

It is a striking confirmation on a large scale, of Mr. Darwin's beautiful theory—that the gay colours of flowers have mostly, or perhaps, wholly been produced, in order to attract insects which aid in their fertilization—that in New Zealand, where insects are so strikingly deficient in variety, the flora should be almost as strikingly deficient in gaily-coloured blossoms. Of course there are some exceptions, but as a whole, green, inconspicuous, and imperfect flowers prevail, to an extent not to be equalled in any other part of the globe; and affording a marvellous contrast to the general brilliancy of Australian flowers, combined with the abundance and variety of its insect-life. We must remember, too, that the few gay or conspicuous flowering-plants possessed by New Zealand, are almost all of Australian, South American, or European genera; the peculiar New Zealand or Antarctic genera being almost wholly without conspicuous flowers. In the tropical Galapagos the same thing occurs. Mr. Darwin notices the wretched weedy appearance of the vegetation; and states that it was some time before he discovered that most of the plants were in flower at the time of his visit! And the insect-life was correspondingly deficient, consisting mainly of a few terrestrial beetles.

The poverty of insect-life in New Zealand must, therefore, be a very ancient feature of the country; and it furnishes an additional argument against the theory of land-connection with, or even any near approach to, either Australia, South Africa, or South America. For in that case numbers of winged insects would certainly have entered, and the flowers would then, as in every other part of the world, have been rendered attractive to them by the development of coloured petals; and this character once acquired would long maintain itself, even if the insects had, from some unknown cause, subsequently disappeared.

After the preceding paragraphs were written, it occurred to me, that if this reasoning were correct, New Zealand plants ought to be also deficient in scented flowers; because it is a part of the same theory, that the odours of flowers have, like their colours, been developed to attract the insects required to aid in their fertilization. I therefore at once applied to my friend Dr. Hooker, as the highest authority on New Zealand botany; simply asking whether there was any such observed deficiency. His reply was:—"New Zealand plants are remarkably scentless, both in regard to the rarity of scented flowers, of leaves with immersed glands containing essential oils, and of glandular hairs." There are a few exceptional cases, but these seem even more rare than might be expected, so that the confirmation of theory is very complete. The circumstance that aromatic leaves are also very scarce, suggests the idea that these, too, serve as an attraction to insects. Aromatic plants abound most in arid countries, and on Alpine heights; both localities where winged insects are comparatively scarce, and where it may be necessary to attract them in every possible way. Dr. Hooker also informs, me that since his Introduction to the New Zealand Flora was written, many plants with handsome flowers have been discovered, especially among the Ranunculi, shrubby Veronicas, and herbaceous Compositæ. The two former, however, are genera of wide range, which may have originated in New Zealand by the introduction of plants with handsome flowers, which the few indigenous insects would be attracted by, and thus prevent the loss of their gay corollas; so that these discoveries will not much affect the general character of the flora, and its very curious bearing on the past history of the islands through the relations of flowers and insects.

In judging of the relation here supposed to exist, it must be remembered, that if the New Zealand insects have been introduced from the surrounding countries by chance immigrations at distant intervals, then, as we go back into the past the insect fauna will become poorer and poorer, and still more inadequate than at present to lead to the development of attractive flowers and odours. This quite harmonizes with the fact of the ancient indigenous flora being so remarkably scentless and inconspicuous, while a few of the more recently introduced genera of plants have retained their floral attractions.

We have already discussed in some detail, the various relations of the Australian sub-regions to the surrounding Regions, and the geographical changes that appear to have taken place. A very few observations will therefore suffice, on the supposed early history of the Australian region as a whole.

It was probably far back in the Secondary period, that some portion of the Australian region was in actual connection with the northern continent, and became stocked with ancestral forms of Marsupials; but from that time till now there seems to have been no further land connection, and the Australian lands have thenceforward gone on developing the Marsupial and Monotremate types, into the various living and extinct races we now find there. During some portion of the Tertiary epoch Australia probably comprised much of its existing area, together with Papua and the Solomon Islands, and perhaps extended as far east as the Fiji Islands; while it might also have had a considerable extension to the south and west. Some light has recently been thrown on this subject by Professor McCoy's researches on the Palæontology of Victoria. He finds abundant marine fossils of Eocene and Miocene age, many of which are strikingly similar to those of Europe at the same period. Among these are Cetaceans of the genus Squalodon; European species of Plagiostomous fishes; mollusca and corals closely resembling those of Europe and North America of the same age,—such as numerous Volutes closely allied to those of the Eocene beds of the Isle of Wight, and the genus Dentalium in great abundance, almost or quite identical with European tertiary species. Along with these, are found some living species, but always such as now live farther north in tropical seas. The Cretaceous and Mesozoic marine fossils are equally close to those of Europe.

The whole of these remains demonstrate that, as in the northern so in the southern hemisphere, a much warmer climate prevailed in the Eocene and Miocene periods than at the present time. This is a most important result, and one which strongly supports Mr. Belt's view, before referred to, that the warmer climates in past geological epochs, and especially that of the Miocene as compared with our own, was caused by a diminution of the obliquity of the ecliptic, leading to a much greater uniformity of the seasons for a considerable distance from the equator, and greatly reducing the polar area within which the sun would ever disappear during an entire rotation of the earth. During such a period, tropical forms of marine animals would have been able to spread north and south, into what are now cool latitudes; and identical genera, and even species, might then have ranged along the southern shores of the old Palæarctic continent, from Britain to the Bay of Bengal, and southward along the Malayan coasts to Australia,

Numerous Miocene plant-beds have also been found in Victoria, containing abundance of Dicotyledonous leaves, which are said generally to resemble those of the Asiatic flora, and of the Miocene plant-beds of the Rhine. It is to be hoped these beds will be more closely examined for remains of insects, land-shells, and vertebrates, and that the plants will be carefully preserved and critically studied; for here probably lies hidden the key, that will solve much of the mystery that attaches to the past history of the Australian fauna.

In drawing up these tables, showing the distribution of the various classes of animals in the Australian region, the following sources of information have been relied on, in addition to the general treatises, monographs, and catalogues used in compiling the 4th Part of this work.

Mammalia.—Gould, Mammals of Australia; Waterhouse on Marsupials; Dr. J. E. Gray's List of Mammalia of New Guinea; Müller, Temminck and Schlegel on Mammals of the Moluccas; papers by Dr. Gray; and personal observations by the Author.

Birds.—Gould's Birds of Australia; Buller's Birds of New Zealand; G. R. Gray's Lists of Birds of Moluccas, &c.; Hartlaub and Finsch on Birds of Pacific Islands; Sclater on Birds of Sandwich Islands; papers by Haast, Hutton, Meyer, Salvin, Schlegel, Sclater, Travers, Lord Walden and the Author.

Reptiles.—Krefft, Catalogue of Snakes; Gunther, List of Lizards in Voyage of Erebus and Terror (1875); and numerous papers.

Names in italics show families which are peculiar to the region.

Names inclosed thus (...) show families which only just enter the region, and are not considered properly to belong to it.

Numbers correspond to the series of numbers to the families in Part IV.