of the Yang-tse-kiang; also North Africa and Arabia, to about | Prof. A. Newton) proposed to nuite the Nearctic with the the line of the tropic of Cancer. 2. Ethiopian, including Africa south of the tropic of Cancer, as well as the southern part of Arabia, with Madagascar and the adjacent islands.. 3. Oriental, or Indo-Malay, comprising India and Ceylon, the Indo-Chinese countries and southern China, and the Malay Archipelago as far as the Philippines, Borneo and Java. 4. Australian, composed of the remainder of the Malay Archipelago, Australia, New Zealand and all the tropical islands of the Pacific, as far east as the Marquesas and the Low Archipelago. 5. Neotropical, which comprises South America and the adjacent islands, the West Indies or Antilles, and Central America and Mexico. 6. Nearctic, consisting of temperate and arctic North America, with Greenland. "These six regions," remarks Dr Wallace, "although all of primary importance from their extent, and well marked by their total assemblage of animal forms, vary greatly in their zoological richness, their degree of isolation and their relationship to each other. The Australian region is the most peculiar and the most isolated, but it is comparatively small and poor in the higher animals. The Neotropical region comes next in peculiarity and isolation, but it is extensive and excessively rich in all forms of life. The Ethiopian and Oriental regions are also very rich, but they have much in common. The Palaearctic and Nearctic regions being wholly temperate are less rich, and they too have many resemblances to each other; but while the Nearctic region has many groups in common with the Neotropical, the Palaearctic is closely connected with the Oriental and Ethiopian regions." In Dr Sclater's original scheme the first four of the above regions were bracketed together under the designation of Palaeogaea, and the fifth and sixth, or those belonging to the New World, as Neogaca. T. H. Huxley, in a paper on the distribution of game-birds, published in the Proceedings of the Zoological Society of London for 1868, instead of dividing the world into an eastern and a western division, adopted a northern and a southern division, calling the former Arctogaca, and the latter (which included Australasia and the Neotropical region of Messrs Sclater and Wallace, but not the Ethiopian region) Notogaea. In 1874 Dr Sclater, taking mammals as well as birds into consideration, adopted Huxley's Arctogaea as the major northern division to include the Nearctic, Palaearctic, Oriental and Ethiopian regions; and instead of Huxley's Notogaca recognized three primary divisions, namely, Dendrogaea for the Neotropical region, Antarctogaea for the Australian region (in a somewhat restricted sense), and Ornithogaea for New Zealand and Polynesia. The tendency of these amendments on the original scheme of a simple division into six regions was to recognize three primary divisions of higher rank than such "regions." This view was adopted in 1890 by Dr W. T. Blanford, who proposed to designate these three major divisions of the earth's land surface respectively the Australian, the South American and the Arctogaean regions. A weak point in this scheme is that since the term 66 region " is likewise applied to the subdivisions of Arctogaea, there is a danger of confusion between the primary and secondary divisions. An amendment proposed anonymously in 1893 was to substitute the names Notogaea, Neogaea and Arctogaca for the three primary divisions of Dr Blanford. Yet another emendation, suggested by R. Lydekker and subsequently adopted by Prof. H. F. Osborn, was to designate these three primary divisions as realms," and to reserve the name " region" for their subdivisions. 3 Emendations on the original scheme also included modifications in the limits of the regions themselves. In 1878, for instance, Dr A. Heilprin (in accordance with a suggestion of 6 Palaearctic region under the name of Holarctic; separating at the same time from the former a "transitional " Sonoran, and from the latter a similar Mediterranean, or Tyrrhenian, region, while he also recognized a distinct Polynesian region, distinguished in the main by negative characters... The Sonoran region was subsequently adopted by Dr C. H. Merriam in 1892, and later on by Dr Blanford in the address already cited, the title being, however, changed to Medio-Columbian. A most important proposal was also embodied in Dr Blanford's scheme, namely, the separation from the Ethiopian region of Madagascar and the Comoro islands to form a separate Malagasy region. Another modification of the original scheme was to transfer the island of Celebes, together with Lombok, Flores and Timor, from the Australian to the Oriental region, or to regard them as representing a transitional region between the two. The effect of this change was practically to abolish "Wallace's line" (the deep channel between the islands of Bali and Lombok and thence northward through the Macassar Strait), the deepest channel being really situated to the eastward of Timor. The later evolution of the scheme, as presented by Dr Max Weber, may be tabularized, with some slight alteration, as follows, the realms" being printed in capitals, the regions and sub-regions in ordinary type, and the transitional regions in italics: In the accompanying map the Sonoran and Mediterranean transitional regions are represented as equivalent in value to the main regions, and the Austro-Malayan transitional region is not indicated. The recognition of a Polynesian and still more of a Hawaiian region, is provisional. The most distinct of the three primary realms is undoubtedly Notogaca, the Australian section of which is the sole habitat of egg-laying mammals (Monotremata) and of a variety of marsupials, inclusive of the whole of the great Notogaea. diprotodonts, with the exception of the few (cuscuses) found in the Austro-Malayan transitional region. Apart from monotremes and marsupials, the only indigenous mammals found in Notogaea are rodents and bats, with perhaps a pig in New Guinea; although it is most probable that the latter is introduced, as is almost certainly the dingo, or native dog, in Australia. The rodents are all referable to the family Muridae, and are mostly of peculiar types, lurus, Notomys, &c.); they are, however, in many instances more such as the golden water-rat (Hydromys) and the jerboa-rats (Conior less nearly related to species found in Celebes, the mountains of the Philippines and Bornco, and apparently represent an ancient fauna. The mammalian fauna of Notogaea is practically limited to the Australian region, its indigenous representatives in New Zealand being only a couple of bats. The monotremes are in all probability the survivors of a group which was widely spread in Jurassic times; while marsupials, as represented by the American opossums (Didelphyidae), had a very wide range even as late as the Oligocene division of the Tertiary period. The diprotodont marregion, or this region conjointly with the Austro-Malayan transisupials may not improbably have originated within the Australian tional region. Notogaca is likewise the home of a number of peculiar types of birds, some of which range, however, into the Austro-Malayan area, that is to say, Celebes and Ceram. In the Australian region the "The Geographical Distribution of Life in North America with special reference to the Mammalia," Proc. Biol. Soc., Washington, vol. vii. pp. 1-64 (1892). 8 See W. L. Sclater, "The Geography of Mammals," part v., Geographical Journal, 1896; M. Weber, "On the Origin of the Fauna of Celebes," Ann. Mag. Nat. Hist., ser. 7, vol. iii. pp. 121-136 (1899), and Der Indo-australische Archipel und die Geschichte seiner Tierwelt (Jena, 1902); Lydekker, Celebes: a Problem in Distribution," Knowledge, vol. xxi. pp. 175-177 (1898); see also Deer of All Lands, p. 168 (1898). Die Säugetiere Jena, 1904), p. 308. peculiar avian families include the birds-of-paradise (Paradiseidae), | Australian marsupials and monotremes reached their present the honeysuckers (Meliphagidae), and the lyre-birds (Menuridae) habitat by means of a land-connexion in the south subsequent to among the perching group, the cockatoos (Cacatuidae) and lories the insulation of New Zealand. This, of course, implies the exist (Loriidae) among the parrots, the mound-builders, or brush-turkeys ence of an extinct southern marsupial fauna of which we have no (Megapodiidae) among the game-birds, and the cassowaries and knowledge except in the case of the Epanorthidae of Patagonia emeus (Casuariidae and Dromaeidae) in the ostrich group. The That Australia formed part of a great equatorial land-belt conpeculiarity of the region is also marked by the absence of certain necting the southern continents in Jurassic times appears to be widely spread family groups, such as the barbets (Megalaemidae), demonstrated by the evidence of the "Gondwana flora." The the otherwise cosmopolitan woodpeckers (Picidae), the trogons question is whether such a connexion-either by way of Antarctica (Trogonidae), and the pheasant and partridge tribe (Phasianidae). or not-persisted in the case of Neogaea long enough to admit of The reptiles, owing probably to their earlier radiation, are much the ancestors of the modern fauna (supposing it all to have come less peculiar, such widely spread types as the monitors (Varanidae) by a southern route) having effected an entrance. The existence and skinks, (Scincidae) being abundant, as are also crocodiles (Croco- of such a land-bridge was suggested by Sir Joseph Hooker in 18471 dilidae). The tortoises belong, however, exclusively to the side- and the idea of a late connexion between Neogaca and Not-gaca necked group (Pleurodira), now restricted to the southern hemi- has been adopted by L. Rütimeyer (1867), Captain F. W. Hutton sphere; among these the most noteworthy being the giant horned (1873), Prof. H. O. Forbes (1893), Mr C. Hedley (1895). Dr H. von tortoise (Miolania) from the Pleistocene of Queensland, which Ihering (1891 and 1900), Prof. H. F. Osborn, who takes an interbelongs to a genus elsewhere known only from the South American mediate view of the extent of the part played by Antarctica (1900), Tertiary. The Australian lung-fish (Ceratodus, or Neoceratodus) is and by Dr A. F. Ortmann (1902). On the other hand, Dr T. Gill the sole survivor of a widely spread Triassic and Jurassic type. (1875) believed in the existence of an "Eogaca" connecting the The salmon tribe (Salmonidae), however, is notable for its absence, three great continents exclusive of Antarctica; and in 1884 Capt. although one peculiar form occurs in New Zealand; and the Hutton, abandoning his former view, suggested the connexion of Cyprinidae, or carps, are wanting throughout the realm, this Australia and South America by means of a mid-Pacific continent. absence extending to Celebes, although in Borneo the group is A summary of these views, with references, is given by Dr Ortmann abundantly represented. in vol. xxxv. pp. 139-142 of the American Naturalist (1901). So far as mammals are concerned, the evidence in favour of a comparatively late land-connexion is weakened by the recent view that certain supposed Patagonian_Tertiary marsupials, such as Prothylacinus, are really creodont Carnivora. On the other hard (putting aside these carnivores), Mr W. J. Sinclair is of optica that the living South American marsupial Caenclestes and its extinct relatives are annectant forms between diprotodonts and poly protodonts, and not far removed from the ancestral stock which gave rise to the Australian phalangers. The occurrence in the Terary of Patagonia of primitive opossums, which cannot be regarded as ancestral to the modern South American forms, is also an important determination., From this, coupled with the testimony afforded by the invertebrate faunas, he considers himself justified in stating that "considerable evidence is now available to show that a landconnexion between Patagonia and the Australian region existed not later than the close of the Cretaceous or the beginning of the Tertiary, and it is possible that at this time the intercharge of marsupials between the two continents was effected. Whether the marsupials originated in South America and migrated thence to Australia, or the reverse, cannot at present be determined." The above-mentioned tortoises of the genus Miolania also appear to afford strong evidence of the persistence of the Jurassic connexion between Notogaea and Neogaea to a comparatively late epoch. New Zealand, here provisionally included in a separate Polynesian region, is characterized by the absence of all indigenous mammals except two bats, each representing a peculiar genus. Among birds, the Neogacic family Meliphagidae includes several peculiar genera, as does also the widely spread starling group (Sturnidae); while the parrots of the genera Stringops and Nestor are likewise peculiar. Still more noteworthy is the abundance of the ostrich group, represented by the living kiwis (Apteryx), and the moas (Dinornithidae) which have been exterminated within comparatively recent times. Reptiles are scarce, but among them the tuatera lizard (Sphenodon) is especially noteworthy on account of being the sole survivor of an ordinal group (Rhynchocephalia) widely spread during Triassic and Jurassic times. Of the Hawaiian area (whether or no rightly regarded as a distinct region), it must suffice to state that it is the sole habitat of the gorgeously coloured birds known as mamos, or sickle-bills (Drepanididae). With regard to the origin of the modern fauna of Notogaca, and more especially the Australian region, as here restricted, we enter extremely debatable ground. Dr Wallace, who refused to admit the existence of any great inter-continental connexions in the past, was of opinion that Australia received the ancestors of its marsupials and monotremes from Asia by way of the AustroMalayan area (as it certainly has its rodents) "far back in the Secondary period." This view has been endorsed by the present writer who suggested the early Eocene as the most probable date of immigration; and it has also received the assent of Dr Max Weber, who is of opinion that in pre-Tertiary-very likely Creta-comparatively modern group, which did not attain any important ceous-times Australia was united by land with Asia. A_EuroAsiatic fauna inhabited this land, from which during the Eocene a southern portion was cut off by partial submergence, this southern portion being the modern Australia and New Guinea, the home of monotremes, marsupials and ancient forms of other groups, such as cassowaries and birds-of-paradise, while widely distributed specialized types are wanting. Northwards extended a coral-sea, in the islands of which dwelt primitive rodents, insectivores and other ancient groups, with perhaps cuscuses. During the Miocene, great changes of level took place in the archipelago, which attained its present form in the Pleistocene. Celebes was insulated early, Java later. Intermittent land-connexions took place, which allowed of periodical immigrations of Asiatic forms from one side and of Australian types from the other. The question is left undecided whether the cuscuses of the Austro-Malayan islands are remnants of the primitive Euro-Asiatic fauna or later immigrants from Australia. The suggestion is also made that the Australian and Philippine rodents are survivors of the original pre-Tertiary fauna, although it is admitted that the specialization of Hydromys is against this. The author fails to see any evidence in favour of a former connexion of Australasia with either South America or a former large antarctic continent (Antarctica). While admitting that this may be the true explanation, Mr B. A. Bensley considers it possible that opossums (Didelphyidae), which he regards as the ancestral stock of the marsupials, may have effected an entrance into Neogaca by way of Antarctica. In either event, he would place the date of entry as post-Eocene; but against this view is the occurrence of remains of a diprotodont marsupial (Wynyardia) in Tasmanian strata believed to be of Eocene age. Prof. Baldwin Spencer is also of opinion that the Lydekker, Geographical Distribution of Mammals (1896). • Report of Horn Expedition to Central Australia, pp. 187 and 188 (1996). Again, Prof. W. B. Benham, from the evidence of earthworms, is strongly disposed to believe in a late connexion between the areas in question. From their invariable association with angio spermous plants, this author is of opinion that earthworms are a development before the Cretaceous. The ancestral type would appear to have been more cr less nearly related to the existing Notiodrilus, of which the headquarters, if not the birthplace, as the "Melanesian plateau." New Zealand and the neighbouring islands, which possess the most ancient worm-fauna, were separated at an early date from this plateau. From this area the primitive worms travelled in one direction into the Austro-Malayan countries, while in another, by way of Antarctica, they reached South America and Africa. With this brief summary of the chief views, this part of the subject must be dismissed without the writer being com mitted to any definite conclusion. Next to Notogaca the most distinct faunistic continental area, so far at any rate as its present and later Tertiary mammals are concerned, is Neogaea, containing, as we have seen, only Neessa. the Neotropical region. It is remarkable as being, with the exception of Notogaea, the only land-area which contains at the present day more than one living genus of marsupials, and also a large middle Tertiary marsupial fauna. The living marsupials include a large number of true opossums, constituting the far y Didelphyidae and Caenolestes the surviving representative of the Epanorthidae of the Patagonian Tertiaries. The opossums are represented by the genera Chironectes and Didelphys; the latter divisible into a number of sub-genera of which the typical griep alone ranged into North America. Whether the modern opossums belong to the endemic Neogaeic fauna, or whether they are be immigrants from the north (where they were represented in the Oligocene of both hemispheres), is a question in regard to which a definite answer can scarcely at present be given. It appears however, that Microbiotherium and certain allied forms from the middle Tertiary of Patagonia are endemic representatives of the Didelphyidae which did not give rise to the modern types. Epanorthidae, in the opinion of Prof. Max Weber, indicate a subordinal group by themselves; and if this be correct their evidence Proc. Amer. Phil. Soc., xlix. 73 (1905). The in favour of a land-connexion between Neogaea and Notogaca cannot have the weight attributed to it by Mr W. J. Sinclair. The typical Edentata (sloths, anteaters and armadillos) are at the present day practically confined to Neogaca where they have existed from the date of the Santa Cruz beds of Patagonia (which are probably of Miocene age). A few armadillos, however, have penetrated into Texas; and in the Pleistocene epoch several representatives of the extinct ground-sloths (Megatheriidae) and a glyptodon, or giant armadillo, also ranged into North America. The group is, however, essentially Neogaeic. Among the monkeys the Cebidae, or American monkeys, and their relatives the Hapalidae, or marmosets, are likewise peculiar to Neogaca, where they date from the Santa Cruz epoch. The vampire-bats, or Phyllostomatidae, are likewise peculiar to this realm, and are doubtless also endemic. With the exception of a few shrew-mice, which have evidently entered from the north, continental Neogaea is at the present day devoid of Insectivora. It is, however, very noteworthy that one peculiar family (Solenodontidae) of the order, apparently nearly allied to the Malagasy Centetidae (tenrecs), occurs in the West Indies, while the extinct Necrolestes, believed to be near akin to the African golden moles (Chrysochloridae), is found in the Santa Cruz beds. Rodents of more or less peculiar types are highly characteristic of Neogaea and for the most part date from the Santa Cruz epoch. Among these the Caviidae, Chinchillidae and Octodontidae are peculiar to this realm, while the Capromyidae are common to the Ethiopian region of Arctogaea, but are unknown elsewhere. Ungulates are in the main very poorly represented in Neogaea and include only the llama group (guanaco, &c.), tapirs, and certain small or medium-sized deer related to North American types. Palaeontological evidence tells us that these, like certain peculiar genera of horses now extinct (such as Hippidium) and mastodons, were comparatively recent intruders into the realm from the north. On the other hand, Neogaea at the date of the deposition of the Santa Cruz beds was the home of certain endemic groups of ungulates, such as the Toxodontia and Litopterna, some of the representatives of which (Toxodon and Macrauchenia) flourished during the Pleistocene Pampean epoch. | common to Neogaca and Madagascar. The blind burrowingsnakes of the family Glauconiidae occur throughout the warmer parts of the realm, and are also found in Africa and south-western Asia. The caimans or South American alligators (Caiman) are solely Neogaean; the iguanas (Iguanidae) are mainly peculiar to the realm, although a few inhabit North America, and there are two outlying genera in Madagascar and a third in Fiji. The tejus (Tejidae) are wholly Neogacan. The Xantusiidae are exclusively Central American and Antillean; while the Amphisbaenidae are practically restricted to Neogaca and Africa. On the other hand, Lacertidae, Varanidae and Agamidae are absent. Tailed amphibians are unknown south of Central America; but the region is the home of several peculiar types of toads, such as Pipa (Surinam toad) belonging to an otherwise Ethiopian section, and the majority of the family Cystignathidae, as exemplified by the horned toad and the escuerso (Ceratophrys), the remainder of the group being Australian. Freshwater fishes are very abundant in Neogaea, where they are represented by a number of peculiar generic and certain family types; some of the members have developed the remarkable habit of feeding upon the floating fruits abundant in the rivers of the tropical forest-districts. The electric eels (Gymnotidae) are peculiar to the waters of Neogaea, as are certain other groups, such as the armoured catfishes (Loricariidae), while true cat-fishes (Siluridae) are extremely abundant. Perhaps, however, the most remarkable feature of the fish-fauna of Neogaca is its affinity to that of the Ethiopian region. Among the lung-fishes the family Lepidosirenidae is, for example, restricted to the two areas, with one genus in each, as is also the family Characinidae. Much the same may be said of the Cichlidae, which have, however, representatives in the Malagasy and Oriental regions; and the Cyprinodontidae, which are extremely abundant in Neogaea (where certain of their representatives are separated by some naturalists as a distinct family, Poeciliidae) likewise present the same general type of distribution, although their area includes the southern fringe of the Palaearctic sub-region and a considerable portion of the Oriental region. As regards the past history of Neogaea, Professor Carl Eigenmann, writing in the Popular Science Monthly for June 1906, observes that in the earliest Tertiary tropical America consisted of two land-areas, Archiguiana and Archamazonia, separated by the lower valley of the Amazon, which was still submerged. There was a land-mass, Hellenis, between Africa and South America, possibly in contact with Guiana and some point in tropical Africa. This land-mass, which was inhabited, among other things, by fishes belonging to the families Lepidosirenida (lung-fishes), Poeciliidae, Characinidae, Cichlidae and Siluridae (cat-fishes), sank beneath the surface of the ocean, forcing the fauna in two directions, towards Africa and towards South America, exterminating all types not moved to the east or to the west. From these two rudiments have developed the present diverse faunas of Africa and South America, each reinforced by intrusions from the ocean and neighbouring land-areas, and by autochthonous development within its own border.... The connexion between Africa and South America existed before the origin of present genera, and even before the origin of some of the present families and sub-families, some time before the early Tertiary. There has never been any exchange between Africa and South America since that time.' Of the Carnivora, the civet group (Viverridae) is absent, and the representatives of the dog tribe (Canidae), bears (Ursidae), of which there is only a single existing representative, cats (Felidae), and probably raccoons (Procyonidae), must be regarded as intruders from the north, although several genera of the last-named group are peculiar to the area. In the Santa Cruz epoch the place of these modern specialized Carnivora was taken by marsupial-like creodonts, such as Prothylacinus. In birds Neogaea is especially rich and contains more than a score of family groups unknown elsewhere. Several of these, such as the tyrant-birds (Tyrannidae), manakins (Pipridae), chatterers (Cotingidae), ant-thrushes (Formicariidae), the oven-bird group (Dendrocolaptidae), plant-cutters (Phytotomidae), and wren-thrushes (Pteroptychidae), belong to a low and generalized type of the perching, or passerine, group. Among the so-called picarian birds, which are likewise a generalized type, the big-billed toucans (Rhamphasiidae), puff-birds (Bucconidae), jacamars (Galbulidae), motmots (Momotidae), and the vast assemblage of humming-birds (Trochilidae) are in the main peculiar to this realm, although some of the lastnamed family wander to the northward in summer. The condors (Cathartidae), form a highly characteristic Neogaeic family; while the hoatzin (Opisthocomus) represents another. Of the higher forms of perching-birds the quit-quits (Coerebidae), greenlets (Vireonidae), the hang-nests and many other representatives of the Icteridae, and the tanagers (Tanagridae) are exclusively Neogaeic; while crows, starlings, thrushes, warblers and flycatchers are either rare or wanting, although the finches are abundant. Parrots are numerous, and represented by peculiar forms such as the macaws (Ara) and conures or ordinary South American parrots (Conurus). Very characteristic of the realm, and unknown elsewhere are the curassows and guans (Cracidae) among the game-birds, the chajas, or screamers (Palamedeidae), the trumpeters (Psophiidae), sunbitterns (Eurypygidae), and the seriema (Cariamidae). Allied apparently to the last is Phororhachos, a giant extinct bird from the Santa Cruz beds with a skull nearly as large as that of a pony. The tinamous (Tinamidae), possibly an annectant type between game-birds and the ostrich group, and the rheas or American ostriches (Rheidae) are likewise exclusively Neogacic. It may be added that the distribution of all the members of the ostrich group affords a strong argument in favour of a former union of the southern continents, especially as their earliest known representative is African. Among reptiles, the tortoises, with the exception of representa-sloths, ground-sloths, ant-eaters, armadillos, glyptodonts, toxodonts, tives of the terrestrial genus Testudo, all belong to the Pleurodira, and include several peculiar generic types such as Chelys (matamata) and one, Podocnemis, common to Madagascar. The occurrence in the Tertiary of Patagonia of a representative of Miolania, elsewhere known only from the Pleistocene of Queensland, has been already mentioned. A number of snakes of the boa group (Boinae) occur in the realm, to which the genus Eunectes (anacondas) is restricted; but Boa itself, like Podocnemis among the tortoises, is This connexion between Neogaea and Africa was doubtless a continuation of the old Jurassic equatorial land-belt to which allusion has been already made; freshwater fishes being probably a group of earlier radiation than mammals. Perhaps the distri bution of the reptilian genera common to Neogaea and Madagascar may be explained in the same manner, although tortoises apparently identical with Podocnemis occur in the Eocene of Europe (as well as in that of Africa and India), so that this group may have radiated from the north. Whether the evidence of the Cystignathidac among the amphibians and of the extinct Miolania among chelonians is also evidence of the persistence of the Jurassic connexion between Neogaca and Notogaea till a considerably later epoch must, for the present, be left an open question. The distribution of other families of lizards is, however, not in favour of such a connexion, the Lacertidae and Agamidae being confined to the Old World, inclusive of Australia but exclusive of Madagascar, while the cosmopolitan Scincidae, so abundant in Notogaea, are extremely scarce in Neogaea. Reverting to the mammalian fauna, its evidence, combined with that of geology, indicates that during the greater portion of the Tertiary period South America was isolated from North America, and inhabited by its autochthonous fauna of monkeys, marmosets, macrauchenias (together with certain other peculiar ungulates), rodents, marsupials and creodonts, as well as by Phororhachos, rheas, tinamous and probably some of the other groups of birds now peculiar to the area. This state of things continued till the later Miocene or Pliocene epoch, during some portion of which a connexion was established with North America by way of the isthmus of Darien. By means of this new land-bridge a certain proportion of the autochthonous fauna of Neogaea was enabled to effect an entrance into North America, as is exemplified by the occurrence there of ground-sloths and glyptodonts. Simultaneously a large immigration of northern forms took place into Neogaca; these invaders from Arctogaca, including cats and sabre-toothed tigers, bears, fox-like dogs, raccoons, llamas, horses, tapirs, deer, mastodons and perhaps opossums. While representatives of most of these invaders have persisted to the present day, some groups, such as horses and mastodons, have entirely disappeared, as has also a large portion of the autochthonous fauna. Here it may be well to notice that the evidence for the insulation of Neogaea during a large portion of the Tertiary period does not by any means rest only on that supplied by mammals. C. H. Gilbert and E. C. Starks, for instance, in a work on the fishes of the two sides of the isthmus of Darien, wrote as follows:-" The ichthyological evidence is overwhelmingly in favour of the existence of a former open communication between the two oceans, which must have become closed at a period sufficiently remote from the present to have permitted the specific differentiation of a very large majority of the forms involved.... All evidence concurs in fixing the date of that connexion at some time prior to the Pleistocene, probably in the early Miocene." This, it will be observed, agrees almost precisely with the conclusions drawn from the fossil mammalian faunas of North and South America, which indicate that land-communication between those two continents was interrupted during a considerable portion of the Tertiary epoch, and only re-established (or [?] established for the first time) either towards the close of the Miocene or the early part of the Pliocene epoch. The South American mammalian fauna, as we now know it, is, then, a complex, consisting of an original autochthonous element and of a large foreign infusion from the north. As to the origin of the latter, there is no difficulty; but some degree of obscurity still prevails with regard to the source of the autochthonous fauna. According to Prof. Eigenmann's interpretation of the evidence of the fresh-water fishes the early Tertiary Atlantic "Hellenis may have been in contact with Guiana on the one side and tropical Africa on the other. That such a connexion did really exist in Tertiary times is the conclusion reached by Dr C. W. Andrews, as the result of his studies of the Tertiary vertebrate fauna of the Fayum district of Egypt, as expressed in the following passage:— "Speaking generally, it appears that (1) probably in Jurassic times Africa and South America formed a continuous land-mass; (2) in the Cretaceous period the sea encroached southwards over this land, forming what is now the South Atlantic. How far this depression had advanced southwards at the end of the Secondary period is not clear, but it appears certain that the final separation of the two continents did not take place till Eocene times, and that there may have been a chain of islands between the northern part of Africa and Brazil which persisted even till the Miocene.' By this route, as was suggested considerably carlier by Prof. W. B. Scott and subsequently by the present writer, Neogaca may have received a considerable portion of its autochthonous mammalfauna. Further reference to this point is made later; but it may be added that the evidence of the land-faunas is supplemented by that of the shallow-water marine faunas on the two sides of the Atlantic, which present a striking similarity. ...we In an address to the British Association at the meeting in 1905 in South Africa Mr G. A. Boulenger expressed himself, however, as by no means satisfied with the evidence of a Tertiary connexion between Africa and South America. "It is undeniable," he observed, that the hypothesis of a South Atlantic land-communication in the Eocene has much in its favour, and when this is really established, all difficulty in explaining the distribution of the Cichlidae will have disappeared. In the meanwhile, must not construct bridges without being sure of our points of attachment." In this connexion it may be mentioned that those who explain the distribution of certain forms of life by the former existence of a land-connexion between the southern continents by way of "Antarctica," have attached some importance to the existence of fishes of the genus Galaxias in the freshwaters of New Zealand, Australia, South America and the Cape. This evidence has been shattered by Mr Boulengor's description (in a memoir of the fishes of the Congo) of a marine representative of the genus in question from the Southern Ocean. For the zoological subregions of Neogaca the reader must refer, as in the case of most of the other regions, to special works on zoological distribution. As Arctogaca includes the whole of the rest of the land-surface of the globe (with the exception of Antarctica) it is almost impossible to give any general diagnosis even of its mammalian Arctogaea. fauna. It may be mentioned, however, that at the present day monotremes are wholly wanting, while marsupials are represented only by one or two species of opossums (Didelphys) in North America and by cuscuses (Phalanger) in the AustroMalayan subregion. The true or typical Edentata are, if we except late wanderers from Neogaca into North America, absent from this realm at the present date and during the Pleistocene; the alleged occurrence of a ground-sloth in the Pleistocene of 1 Mem. Californian Academy, vol. iv. (1904). 2 Catalogue of the Tertiary Vertebrala of the Fayum (London, 1906). Madagascar being probably due to a misinterpretation. On the other hand, this region, and more especially its eastern half, is the great home of the ungulate mammals. Indeed rhinoceroses may be considered absolutely characteristic of Arctogaea, since at one tire or another they have ranged over the whole area, except Madagascar, and are quite unknown elsewhere. The modern land Carnivora are likewise an essentially Arctogaeic group, which only found its way into Neogaca at a comparatively recent epoch; and the realm may be said to have been the birthplace of most of the higher groups of placental mammals. The tortoises of the family Trionychidae form an exclusively Arctogaean group, once ranging all over the realm, although long since extinct in Europe. If Madagascar be excepted, the Ethiopian region (or Ethiopia) is the most distinct of all the regions of Arctogaea. So distinct is it that, on the evidence of the distribution of moths, Ethioplas Dr H. S. Packard' has suggested that it should be sepa- regioa. rated from Arctogaea to form a realm by itself, under the name of Apogaca. The mammalian fauna, even exclusive of the Tertiary one of Egypt, does not, however, countenance such a separation. By Sclater and Wallace, Madagascar was included in the Ethiopian region, but that island was subsequently made a region by itself by Dr Blanford. This separation of Madagascar to form a Malagasy region has met with general acceptance; but in the opinion of Mr R. I: Pocock, who bases his conclusion on the distribution of trapdoor-spiders (which in other respects accords curiously well with that of mammals), it is not justified. The mammalian evidence appears, however, to be overwhelmingly strong in its favour; and it also receives support from reptilian distribution. All are agreed that the Ethiopian region should exclude that part of Africa which lies, roughly speaking, northward of the tropic of Cancer. By Sclater and Wallace the region was taken to include that portion of Arabia lying to the south of the same tropic; but Mr Pocock has pointed out that this separation of Arabia into two portions is not supported by the distribution of scorpions, and he would refer the whole of it to the Mediterranean transitional region. The occurrence of a takr. goat (Hemitragus) in Oman lends some support to this proposal since that genus has no representative in Africa, and occurs else where only in the Himalaya and the mountains of southern India. Other writers have not accepted Mr Pocock's emendation; and the reference of the northern half of Arabia to the Mediterranean and of the southern half to the Ethiopian region is usually followed. The area is admittedly a meeting-ground of at least two fauras. Discoveries in the Fayum district of Egypt have conclusively proved that during the early (Eocene) part of the Tertiary period Ethiopia was a great centre of development, and subsequently of dispersal, instead of having received (as was formerly supposed) the whole of its higher modern mammalian fauna from the north. In this Ethiopian centre were developed the ancestors of the elephants (Proboscidea) and of the hyraxes (Hyracoidea); the latter group being represented by species of much larger size than the existing forms, some of the former of which ranged into southern Europe during the later Tertiary. It was also the home of a peculiar subordinal group of ungulates (Barypoda), typified by Arsinöitherium, and may likewise have been the birthplace of the swine (Suidae) as the earliest known representative of that group (Geniohyus) occurs in the Fayum Eocene. The hippopotamers (Hippopotamidae), which appear to be descended from the Tertio Anthracotheriidae, may likewise be of Ethiopian origin, and the same may turn out to be the case with the giraffe group (Giraf¿€ although definite evidence with regard to the latter point is wanting. The occurrence of an ostrich-like flightless bird in the Fayua Eocene the oldest known representative of that group is su tive that the Ratitae, originated in Ethiopia, which would accord well with their distribution both in the present and the past. A giant land-tortoise (Testudo) is likewise known from the Favim beds, and as it is allied to the species recently or still inhabiting Madagascar and the Mascarene islands, there is a strong probability that Ethiopian Africa was likewise the centre of development and dispersal of that group. Turning to its existing mammalian fauna, Ethiopia possesses a number of peculiar family or generic groups, and is also nearly equally well characterized by the absence of others. As remarked by Wallace, one of its characteristics is the great number of species of large size. Among the Primates, it is the home of the typol group of the Negroid branch of the human species, whose northern limits coincide approximately with the boundary of the rege a itself, being replaced in northern Africa by races of the Caucasian stock. Gorillas and chimpanzees (Anthropopithecus) are pecular to the region, as are also baboons (Papio and Therapithecus), á southern Arabia be included. Monkeys abound, and although in most cases nearly allied to those of the Oriental region, are generically Science, ser. 2, vol. xix. p. 221 (1904). Dr Packard greens Notogaea and Neogaea in a single realm under the name Artantogaca. Some other writers, such as Dr H. Gadow, take Notogara to include all the three southern continents, and employ the terma Arctogaca for the rest of the world. Proc. Zool. Soc., London, 1903, pp. 340-368. As regards its past history, Ethiopian Africa was in connexion with India during the Triassic and Jurassic periods, the two areas collectively forming "Gondwanaland," which doubtless constituted a portion of the equatorial land-belt referred to as existing during the epochs in question. Gondwanaland was the home of a large section of the anomodont reptiles from which mammals have sprung; and it is quite probable that the evolution of the latte: group took place within the present area. Between the Trias and the Eocene little or nothing is known of the vertebrate palaeontology of Ethiopia; and in Egypt there is also a long gap between the lower Miocene and certain Pliocene beds in the Wadi Natrun. The Tertiary deposits of southern Europe and northern India indicate, however, that Ethiopian Africa was in free communication with these countries during the upper Miocene and Pliocene epochs. There occur, for instance, either in south-eastern Asia or southern Europe, or both, during the latter period numerous genera of antelopes now restricted to Ethiopia, as well as giraffes, okapi-like ruminants (Palacotragus), elephants and rhinoceroses of an African type, probably zebras, hippopotamuses, baboons, chimpanzees and ostriches. Owing to imperfect knowledge of Pliocene Africa, it is impossible to say whether these types were first developed in Ethiopia or to the north-east, and consequently whether or not Professor Huxley was right in his theory that the modern higher mammalian fauna of Ethiopia came from the north. It has, however, been suggested that while the Bovidae are an autochthonous Ethiopian group, the Cervidae originated in either the Holarctic or the Oriental region; a theory which if confirmed will materially aid in explaining the absence of the latter group from Ethiopia. It is supported to some extent by the fact that we are acquainted with primitive ancestral deer in the European Tertiary, while the ancestors of the Bovidae are at present unknown. Whatever be the truth on this.point, it is manifest that whether the middle Tertiary Bovidae migrated from Ethiopia to Asia or in the opposite direction, there must have been some cause which barred the entrance by the same route into the latter area of all members of the deer-tribe (as well as bears). It should be added that although the ancestral Proboscidea were Ethiopian, the passage from the mastodons into the true elephants appears to have taken place in Asia; a circumstance which would imply the Asiatic origin of the African elephant. distinct. The Prosimiae, or lemuroids, include the galagos (Galago) | present article, it may be mentioned that Ethiopia is remarkable and pottos (Perodictycus), of which the latter are akin to the Oriental for the total absence of fresh-water cray-fishes. lorises, while the former are quite distinct from the Malagasy lemurs. Among the Carnivora, the aard-wolf (Proteles), the hunting-dog (Lycaon) and the long-eared fox (Olocyon) are peculiar generic types, as are several forms of mungooses (Herpestinae); while the spotted hyaena forms a subgenus by itself. The bear family (Ursidae), on the other hand, is totally absent. In the great ungulate order the African elephant is widely sundered from its Asiatic cousin, as are the two species of rhinoceros from their representatives in the Oriental region; indeed each group is subgenerically distinct. The hyraxes, forming the suborder Hyracoidea, are, with the exception of a single outlying Syrian species, confined to Ethiopia. Zebras and true wild asses are likewise peculiar to the region. More remarkable is the extraordinary number of peculiar genera of antelopes, a few of which range, however, into North Africa, Syria and Arabia; the African buffaloes are markedly different from those of Asia; and sheep and goats are absent from the region, with the exception of intruding into it to some extent in the mountains of the Sudan and Abyssinia. The giraffe-family (Giraffidae), as represented by giraffes (Giraffa) and the okapi (Ocapia), is absolutely confined to this region, from which the deer-tribe (Cervidae) is completely absent. Chevrotains, or mouse-deer, are represented by the peculiar genus Dorcatherium (or Hyomoschus); in the pigs the wart-hogs (Phacochoerus), foresthogs (Hylochoerus), and the bush-pigs (subgenus Potamochoerus), with the exception of one Malagasy species, are now unknown elsewhere, as are also hippopotamuses. Rodents include a number of peculiar types, among which may be noticed the scaly-tailed squirrels (Anomaluridae), the jumping-hares (Pedetes), the strandmoles (Bathyergidae), the crested-rats (Lophiomys), and the canerats (Thryonomys, or Aulacodus); the last being nearly allied to South American forms. In the Insectivora, moles (Talpidae) are absent, the jumping-shrews (Macroscelididae) are solely African, although ranging north of the Sahara, while the golden moles (Chrysochloridae) and the Potamogalidae are exclusively Ethiopian. Lastly, the ant-bears, or aard-varks (Orycteropodidae), represent a suborder of the Edentata unknown elsewhere; while the African pangolins (Manidae) differ markedly from their Oriental The Ethiopian birds are less peculiar. The ostrich (Struthio) ranges, in suitable localities, all over the region, thus entering the Mediterranean transition-region in the north. The guineafowls (Numidinae) form a subfamily confined to Ethiopia and Madagascar, where true pheasants are unknown. Other peculiar types are plantain-eaters (Musophagidae), colics (Coliidae), woodhoopoes (Irrisoridae), barbets (Megalaemidae), ground-hornbills (Bucorvus), secretary-birds (Serpentariidae), glossy starlings (Lamprotornis), ox-peckers (Buphaga), the genera Laniarius and Telephorus, as well as a number of others, all of which are unknown in Madagascar. In addition to true pheasants, wrens (Troglodytidae) and water-ousels (Cinclidae) are unknown in the Ethiopian region. kindred. The evidence in favour of the continuation of the Mesozoic landbridge between Ethiopia and Neogaea has been discussed under the heading of the latter area. If the arguments in favour of such a connexion are valid, it is to the old mammal fauna of Ethiopia that we must probably look for the progenitors of the Santa Cruz fauna of Patagonia. Very noteworthy is the alleged occurrence of remains of primitive armadillos in the Oligocene beds of southern Europe in association with those of pangolins and aard-varks; since, if these fossils be rightly determined, there at once arises the probability of Africa having been the original home of the entire Edentate order. In the case of an island lying so close to the African continent as does Madagascar the natural expectation would be that its fauna should be intimately related to that of the former. Malagasy As a matter of fact-in the case of mammals and birds, region. at any rate-it is much more distinct from the Ethiopian fauna than is the latter from the fauna of either the Oriental or the Holarctic region. The evidence-from the above-mentioned fact positively overwhelming, while it is also supported in some degree by the distribution of groups other than those named. In place of the Ethiopian assemblage of apes, monkeys, baboons, galagos and pottos, Madagascar (together with the Comoro islands) possesses an absolutely unique fauna of lemurs, constituting the family Lemuridae, which, as now understood, is confined to this island, where it is represented by the three subfamily-groups of sifakas (Indrisinae), true lemurs (Lemurinae), and aye-ayes (Chiromyinae). All these animals agree with one another in the characters of the tympanic region of the skull; thereby differing from the African and Oriental Prosimiae, but agreeing with the European Oligocene Adapis, which must apparently be regarded as the ances tral form. This is a striking confirmation of the theory advanced many years ago by Huxley that Madagascar received its lemuroid fauna from Europe at a very early date, since which time, at any rate, it has been isolated from Africa. Some of the Pleistocene Malagasy lemurs were much larger than any of the living forms, rivalling in this respect a chimpanzee. The Carnivora are represented only by a small number of species, mostly referable to peculiar genera, of Viverridae, among which the fossa (Cryptoprocta) is the largest. In the ungulates there are only two extinct species of hippopotamus and a living bush-pig, the ancestors of all three of which probably crossed the Mozambique channel by swimming; and Edentata are equally conspicuous by their absence. Insectivora, on the other hand, are represented by the tenrecs (Centetidae), with numerous generic types, whose nearest relatives Apart from the widespread Trionychoidea (of which there are two genera peculiar to the region), the Ethiopian fresh-water tortoises belong to the section Pleurodira; the two genera Pelomedusa and Sternothaerus being commcn to Africa and Madagascar, and unknown elsewhere. The Amphisbaenidae are common to Neogaea and Ethiopia, to the exclusion of Madagascar; but the Gerrhosauridae and Zonuridae, on the other hand, are restricted to the present region and Madagascar, which also form the head-groups-in favour of recognizing a distinct Malagasy region is in quarters of chameleons. In contrast to the latter community is the absence in Madagascar of Agamidae and Varanidae, which are common in Ethiopia. The absence of slow-worms and their kindred (Anguidae) is a marked negative feature of the present region. As regards batrachians, the region has no salamanders or other tailed forms, but, in common with India, possesses caecilians (Apoda); while it shares the group of tongueless toads (Aglossa) with Neogaea, its peculiar family being the Xenopodidae, in contradistinction to the South American Pipidae. The Pelobatidae are absent, and true toads are few, but frogs are abundant. Among fishes, Africa south of the Sahara possesses a number of peculiar types. With Neogaea it shares the possession of the typical lung-fishes (Lepidosirenidae), while it is the habitat of the species of bichir (Polypterus) and Calamoichthys, the sole survivors of the ancient group of fringe-finned ganoids (Crossopterygii). The other families peculiar to Ethiopia are the Mormyridae (proboscis-fishes), Pantodontidae, and Phractolaemidae; the two latter being represented only by a single species each. The Notopteridae, Ophiocephalidae, Anabantidae, Osphromenidae and Mastacembelidae are common to Ethiopia and the Oriental region. In addition to the Lepidosirenidae, the Characinidae are peculiar to this region and Neogaca. The Cichlidae occur in Madagascar, Ethiopia, the Oriental region and Neogaca; and the Osteoglossidae are common to the last three of these regions, as well as Australia, while the Nandidae are Ethiopian, Oriental and Neotropical. On the whole, the affinities of the fish-fauna of Ethiopia are nearest to that of the Oriental region, and, secondly, to that of South America. Although invertebrates do not come within the scope of the 1 The fossils of the Uitenhage beds, now generally classed as Jurassic, consist chiefly of invertebrates and plants. |