The Geographical Distribution of Animals/Chapter 17

In the preceding part of our work, we have discussed the geographical distribution of animals from the point of view of the geographer; taking the different regions of the earth in succession, and giving as full an account as our space would permit of their chief forms of animal life. Now, we proceed from the standpoint of the systematic zoologist; taking in succession each of the families with which we deal, and giving an account of the distribution, both of the entire family and, as far as practicable, of each of the genera of which it is composed. As in the former part, our mode of treatment led us to speculate on the past changes of the earth's surface; so here we shall endeavour to elucidate the past migrations of animals, and thus, to some extent, account for their actual distribution.

The tabular headings, showing the range of the family in each region, will enable the reader to determine at a glance the general distribution of the group, as soon as he has familiarised himself, by a study of our general and regional maps, with the limits of the regions and sub-regions, and the figures (1 to 4) by which the latter are indicated. Much pains have been taken, to give the number of the known genera and species in each family, correctly; but these numbers must, in most cases, only be looked upon as approximations; because, owing to constant accessions of fresh material on the one hand, and the discovery that many supposed species are only varieties, on the other, such statistics are in a continual state of fluctuation. In the number of genera there is the greatest uncertainty; as will be seen by the two sets of numbers sometimes given, which denote the genera according to different modern authorities.

There is also a considerable difference in the dependence to be placed on the details given in the different classes of animals. In Mammalia and Birds some degree of accuracy has, it is hoped, been attained; the classification of these groups being much advanced, and the materials for their study ample. In Reptiles this is not the case, as there is no recently published work dealing with the whole subject, or with either of the larger orders. An immense number of new species and new genera of snakes and lizards, have been described in the last twenty years; and Dr. Günther—our greatest authority on reptiles in this country—has kindly assisted me in incorporating such of these as are most trustworthy, in a general system; but until entire Orders have been described or catalogued on a uniform plan, nothing more than a general approximation to the truth can be arrived at. Still, so many of the groups are well defined, and have a clearly limited distribution, that some interesting and valuable comparisons may be made.

For Fishes, the valuable "Catalogue" of Dr. Günther was available, and it has rarely been attempted to go beyond it. A large number of new species have since been described, in all parts of the world; but it is impossible to say how many of these are really new, or what genera they actually belong to. The part devoted to this Class is, therefore, practically a summary of Dr. Günther's Catalogue; and it is believed that the discoveries since made will not materially invalidate the conclusions to be drawn from such a large number of species, which have been critically examined and classified on a uniform system by one of our most able naturalists. When a supplement to this catalogue is issued, it will be easier to make the necessary alterations in distribution, than if a mass of untrustworthy materials had been mixed up with it.

For Insects, excellent materials are furnished, in the Catalogue of Mr. Kirby for Butterflies and in that of Drs. Gemminger and Harold for Coleoptera. I have also made use of some recently published memoirs on the Insects of Japan and St. Helena, and a few other recent works; and have, I believe, elaborated a more extensive series of facts to illustrate the distribution of insects, than has been made use of by any previous writer. Several discussions on the bearing of the facts of insect distribution, will also be found under the several Regions, in the preceding part of this work.

Terrestrial Mollusca form a group, as to the treatment of which I have most misgivings; owing to my almost entire ignorance of Malacology, and the great changes recently made in the classification of shells. There is also much uncertainty as to genera and sub-genera, which is very puzzling to one who merely wishes to get at general results. Finding it impossible to incorporate the new matter with the old, or to harmonise the different classifications of modern conchologists, I thought it better to confine myself to the standard works of Martens and Pfeiffer, with such additions of new species as I could make without fear of going far wrong. In some cases I have made use of recent monographs—especially on the shells of Europe, North America, the West Indian Islands, and the Sandwich Islands; and have, I venture to hope, not fallen into much error in the general conclusions at which I have arrived.

The Simiidæ, or Anthropoid Apes, comprehend those forms of the monkey-tribe which, in general organization, approach nearest to man. They inhabit the tropics of the Old World, and are most abundant near the equator; but they are limited to certain districts, being quite unknown in eastern and southern Africa, and the whole peninsula of Hindostan.

The genus Troglodytes (or Mimetes, as it is sometimes named) comprehends the chimpanzee and gorilla. It is confined to the West African sub-region, being found on the coast about 12° North and South of the equator, from the Gambia to Benguela, and as far inland as the great equatorial forests extend. There are perhaps other species of chimpanzee; since Livingstone met with what he supposed to be a new species in the forest region west of Lake Tanganyika, while Dr. Schweinfurth found one in the country beyond the western watershed of the Nile. The gorilla is confined within narrower limits on and near the equator.

We have to pass over more than 70° of longitude before we again meet with Anthropoid Apes, in the northern part of Sumatra—where a specimen of the orang-utan (Simia satyrus) now in the Calcutta Museum, was obtained by Dr. Abel, and described by him in the Asiatic Researches, vol. xv.—and in Borneo, from which latter island almost all the specimens in European museums have been derived. There are supposed to be two species of Simia in Borneo, a larger and a smaller; but their distinctness is not admitted by all naturalists. Both appear to be confined to the swampy forests near the north, west, and south coasts.

The Gibbons, or long-armed apes, forming the genus Hylobates, (7 species) are found in all the large islands of the Indo-Malayan sub-region, except the Philippines; and also in Sylhet and Assam south of the Brahmaputra river, eastward to Cambodja and South China to the west of Canton, and in the island of Hainan.

The Siamang (Siamanga syndactyla) presents some anatomical peculiarities, and has the second and third toes united to the last joint, but in general form and structure it does not differ from Hylobates. It is the largest of the long-armed apes, and inhabits Sumatra and the Malay peninsula.

The Semnopithecidæ, are long-tailed monkeys without cheek-pouches, and with rather rounded faces, the muzzle not being prominent. They have nearly the same distribution as the last family, but are more widely dispersed in both Africa and Asia, one species just entering the Palæarctic region.

The Eastern genus Presbytes or Semnopithecus (29 species), is spread over almost the whole of the Oriental region wherever the forests are extensive. They extend along the Himalayas to beyond Simla, where a species has been observed at an altitude of 11,000 feet, playing among fir-trees laden with snow wreaths. On the west side of India they are not found to the north of 14° N. latitude. On the east they extend into Arakan, and to Borneo and Java, but not apparently into Siam or Cambodja. Along the eastern extension of the Himalayas they again occur in East Thibet; a remarkable species with a large upturned nose (S. roxellana) having been discovered by Père David at Moupin (about Lat. 32° N.) in the highest forests, where the winters are severe and last for several months, and where the vegetation, and the other forms of animal life, are wholly those of the Palæarctic region. It is very curious that this species should somewhat resemble the young state of the proboscis monkey (S. nasalis), which inhabits one of the most uniform, damp, and hot climates on the globe—the river-swamps of Borneo.

Colobus, the African genus (11 species), is very closely allied to the preceding, differing chiefly in the thumb being absent or rudimentary. They are confined to the tropical regions—Abyssinia on the east, and from the Gambia to Angola and the island of Fernando Po, on the west.

This family comprehends all the monkeys with cheek pouches, and the baboons. Some of these have very long tails, some none; some are dog-faced, others tolerably round-faced; but there are so many transitions from one to the other, and such a general agreement in structure, that they are now considered to form a very natural family. Their range is more extensive than any other family of Quadrumana, since they not only occur in every part of the Ethiopian and Oriental regions, but enter the Palæarctic region in the east and west, and the Australian region as far as the islands of Timor and Batchian. The African genera are Myiopithecus, Cercopithecus, Cercocebus, Theropithecus, and Cynocephalus; the Oriental genera, Macacus, and Cynopithecus.

Myiopithecus (1 species), consisting of the talapoin monkey of West Africa, differs from the other African monkeys in the structure of the last molar tooth; in the large ears, short face, and wide internasal septum; in this respect, as well as in its grace and gentleness, resembling some of the American monkeys.

Cercopithecus (24 species), contains all the more graceful and prettily coloured monkeys of tropical Africa, and comprises the guenons, the white-nosed, and the green monkeys. They range from the Gambia to the Congo, and from Abyssinia to the Zambesi.

Cercocebus (5 species), the mangabeys, of West Africa, are very closely allied to the eastern genus Macacus.

Theropithecus (2 species), including the gelada of Abyssinia and an allied species, resemble in form the baboons, but have the nostrils placed as in the last genus.

Cynocephalus (10 species), the baboons, are found in all parts of Africa. They consist of animals which vary much in appearance, but which agree in having an elongated dog-like muzzle with terminal nostrils, and being of terrestrial habits. Some of the baboons are of very large size, the mandrill (C. maimon) being only inferior to the orang and gorilla.

Macacus (25 species), is the commonest form of eastern monkey, and is found in every part of the Oriental region, as well as in North Africa, Gibraltar, Thibet, North China, and Japan; and one of the commonest species, M. cynomolgus, has extended its range from Java eastward to the extremity of Timor. The tail varies greatly in length, and in the Gibraltar monkey (M. innus) is quite absent. A remarkable species clothed with very thick fur, has lately been discovered in the snowy mountains of eastern Thibet.

Cynopithecus (? 2 sp.).—This genus consists of a black baboon-like Ape, inhabiting Celebes, Batchian, and the Philippine Islands; but perhaps introduced by man into the latter islands and into Batchian. It is doubtful if there is more than one species. The tail of this animal is a fleshy tubercle, the nostrils as in Macacus, but the muzzle is very prominent; and the development of the maxillary bones into strong lateral ridges corresponds to the structure of the most typical baboons. This species extends further east than any other quadrumanous animal.

The Cebidæ, which comprehend all the larger American Monkeys, differ from those of the Old World by having an additional molar tooth in each jaw, and a broad nasal septum; while they have neither cheek-pouches nor ischial callosities, and the thumb is never completely opposable. Some have prehensile tails, especially adapting them for an arboreal life. They are divided into four sub-families,—Cebinæ, Mycetinæ, Pitheciinæ, and Nyctipithecinæ. The Cebidæ are strictly confined to the forest regions of tropical America, from the southern part of Mexico to about the parallel of 30° South Latitude. The distribution of the genera is as follows:—

Sub-family, Cebinæ.—Cebus (18 sp.), is the largest genus of American monkeys, and ranges from Costa Rica to Paraguay. They are commonly called sapajous. Lagothrix (5 sp.), the woolly monkeys, are rather larger and less active than the preceding; they are confined to the forests of the Upper Amazon Valley, and along the slopes of the Andes to Venezuela and Bolivia. Ateles (14 sp.), the spider monkeys, have very long limbs and tail. They range over the whole area of the family, and occur on the west side of the Equatorial Andes and on the Pacific coast of Guatemala. Eriodes (3 sp.), are somewhat intermediate between the last two genera, and are confined to the eastern parts of Brazil south of the equator. The three last mentioned genera have very powerful prehensile tails, the end being bare beneath; whereas the species of Cebus have the tail completely covered with hair, although prehensile, and therefore not so perfect a grasping organ.

Sub-family, Mycetinæ, consists of but a single genus, Mycetes (10 sp.), the howling monkeys, characterized by having a hollow bony vessel in the throat formed by an enlargement of the hyoid bone, which enables them to produce a wonderful howling noise. They are large, heavy animals, with a powerful and perfect prehensile tail. They range from East Guatemala to Paraguay. (Plate XIV., vol. ii., p. 24.)

Sub-family, Pitheciinæ, the sakis, have a non-prehensile bushy tail. Pithecia (7 sp.), has the tail of moderate length; while Brachiurus (5 sp.) has it very short. Both appear to be restricted to the great equatorial forests of South America.

Sub-family, Nyctipithecinæ, are small and elegant monkeys, with long, hairy, non-prehensile tails. Nyctipithecus (5 sp.), the night-monkeys or douroucoulis, have large eyes, nocturnal habits, and are somewhat lemurine in their appearance. They range from Nicaragua to the Amazon and eastern Peru. Saimiris or Chrysothrix (3 sp.), the squirrel-monkeys, are beautiful and active little creatures, found in most of the tropical forests from Costa Rica to Brazil and Bolivia. Callithrix (11 sp.), are somewhat intermediate between the last two genera, and are found all over South America from Panama to the southern limits of the great forests.

The Hapalidæ, or marmosets, are very small monkeys, which differ from the true Cebidæ in the absence of one premolar tooth, while they possess the additional molar tooth; so that while they have the same number of teeth (thirty-two) as the Old World monkeys, they differ from them even more than do the Cebidæ. The thumb is not at all opposable, and all the fingers are armed with sharp claws. The hallux, or thumb-like great toe, is very small; the tail is long and not prehensile. The two genera Hapale (9 sp.), and Midas (24 sp.), are of doubtful value, though some naturalists have still further sub-divided them. They are confined to the tropical forests of South America, and are most abundant in the districts near the equator.

The Lemuridæ, comprehending all the animals usually termed Lemurs and many of their allies, are divided by Professor Mivart—who has carefully studied the group—into four sub-families and eleven genera, as follows:—

Sub-family Indrisinæ, consisting of the genus Indris (5 sp.), is confined to Madagascar.

Sub-family Lemurinæ, contains five genera, viz.:—Lemur, (15 sp.); Hapalemur (2 sp.); Microcebus (4 sp.); Chirogaleus (5 sp.); and Lepilemur (2 sp.);—all confined to Madagascar.

Sub-family Nycticebinæ, contains four genera, viz.:—Nycticebus (3 sp.)—small, short-tailed, nocturnal animals, called slow-lemurs,—range from East Bengal to South China, and to Borneo and Java; Loris (1 sp.)—a very small, tail-less, nocturnal lemur, which inhabits Madras, Malabar, and Ceylon; Perodicticus (1 sp.)—the potto—a small lemur with almost rudimentary forefinger, found at Sierra Leone (Plate V., vol. i., p. 264); Arctocebus (1 sp.)—the angwantibo,—another extraordinary form in which the forefinger is quite absent and the first toe armed with a long claw,—inhabits Old Calabar.

Sub-family Galaginæ, contains only the genus Galago (14 sp.), which is confined to the African continent, ranging from Senegal and Fernando Po to Zanzibar and Natal.

The curious Tarsius spectrum, which constitutes this family, inhabits Sumatra, Banca, and Borneo, and is also found in some parts of Celebes, which would bring it into the Australian region; but this island is altogether so anomalous that we can only consider its productions to have somewhat more affinity with the Australian than the Oriental region, but hardly to belong to either. The Tarsier is a small, long-tailed, nocturnal animal, of curious structure and appearance; and it forms the only link of connection with the next family, which it resembles in the extraordinary development of the toes, one of which is much larger and more slender than the rest. (Plate VIII., vol. i. p. 337.)

The Aye-aye, (Chiromys), the sole representative of this family, is confined to the island of Madagascar. It was for a long time very imperfectly known, and was supposed to belong to the Rodentia; but it has now been ascertained to be an exceedingly specialized form of the Lemuroid type, and must be considered to be one of the most extraordinary of the mammalia now inhabiting the globe. (Plate VI., vol. i., p. 278.)

Not much progress has yet been made in tracing back the various forms of Apes and Monkeys to their earliest appearance on the globe; but there have been some interesting recent discoveries, which lead us to hope that the field is not yet exhausted. The following is a summary of what is known as to the early forms of each family:—

Simiidæ.—Two or three species of this family have been found in the Upper Miocene deposits of France and Switzerland. Pliopithecus, of which a species has been found at each locality, was allied to the gibbons (Hylobates), and perhaps to Semnopithecus. A more remarkable form, named Dryopithecus, as large as a man, and having peculiarities of structure which are thought by Gervais and Lartet to indicate a nearer approach to the human form than any existing Ape, has been found in strata of the same age in France.

Semnopithecidæ.—Species of Semnopithecus have been found in the Upper Miocene of Greece, and others in the Siwalik Hills of N. W. India, also of Upper Miocene age. An allied form also occurs in the Miocene of Wurtemburg. Mesopithecus from Greece is somewhat intermediate between Semnopithecus and Macacus.

Remains supposed to be of Semnopithecus, have also occurred in the Pliocene of Montpellier.

Cynopithecidæ.—Macacus has occurred in Pliocene deposits at Grays, Essex; and also in the South of France along with Cercopithecus.

Cebidæ.—In the caves of Brazil remains of the genera Cebus, Mycetes, Callithrix, and Hapale, have been found; as well as an extinct form of larger size—Protopithecus.

Lemuroidea.—A true lemur has recently been discovered in the Eocene of France; and it is supposed to be most nearly allied to the peculiar West African genera, Perodicticus and Arctocebus.

Cænopithecus, from the Swiss Jura, is supposed to have affinities both for the Lemuridæ and the American Cebidæ.

In the lower Eocene of North America remains have been discovered, which are believed to belong to this sub-order: but they form two distinct families,—Lemuravidæ and Limnotheridæ. Other remains from the Miocene are believed to be intermediate between these and the Cebidæ,—a most interesting and suggestive affinity, if well founded. For the genera of these American Lemuroidea, see vol. i., p. 133.

The most striking fact presented by this order, from our present point of view, is the strict limitation of well-marked families to definite areas. The Cebidæ and Hapalidæ would alone serve to mark out tropical America as the nucleus of one of the great zoological divisions of the earth. In the Eastern Hemisphere, the corresponding fact is the entire absence of the order from the Australian region, with the exception of one or two outlying forms, which have evidently transgressed the normal limits of their group. The separation of the Ethiopian and Oriental regions is, in this order, mainly indicated by the distribution of the genera, no one of which is common to the two regions. The two highest families, the Simiidæ and the Semnopithecidæ, are pretty equally distributed about two equatorial foci, one situated in West Africa, the other in the Malay archipelago,—in Borneo or the Peninsula of Malacca;—while the third family, Cynopithecidæ, ranges over the whole of both regions, and somewhat overpasses their limits. The Lemuroid group, on the other hand, offers us one of the most singular phenomena in geographical distribution. It consists of three families, the species of which are grouped into six sub-families and 13 genera. One of these families and two of the sub-families, comprising 7 genera, and no less than 30 out of the total of 50 species, are confined to the one island of Madagascar. Of the remainder, 3 genera, comprising 15 species, are spread over tropical Africa; while three other genera with 5 species, inhabit certain restricted portions of India and the Malay islands. These curious facts point unmistakably to the former existence of a large tract of land in what is now the Indian Ocean, connecting Madagascar on the one hand with Ceylon, and with the Malay countries on the other. About this same time (but perhaps not contemporaneously) Madagascar must have been connected with some portion of Southern Africa, and the whole of the country would possess no other Primates but Lemuroidea. After the Madagascar territory (very much larger than the existing island) had been separated, a connection appears to have been long maintained (probably by a northerly route) between the more equatorial portions of Asia and Africa; till those higher forms had become developed, which were afterwards differentiated into Simia, Presbytes, and Cynopithecus, on the one hand, and into Troglodytes, Colobus, and Cynocephalus, on the other. In accordance with the principle of competition so well expounded by Mr. Darwin, we can understand how, in the vast Asiatic and African area north of the Equator, with a great variety of physical conditions and the influence of a host of competing forms of life, higher types were developed than in the less extensive and long-isolated countries south of the Equator. In Madagascar, where these less complex conditions prevailed in a considerable land-area, the lowly organized Lemuroids have diverged into many specialized forms of their own peculiar type; while on the continents they have, to a great extent, become exterminated, or have maintained their existence in a few cases, in islands or in mountain ranges. In Africa the nocturnal and arboreal Galagos are adapted to a special mode of life, in which they probably have few competitors.

How and when the ancestors of the Cebidæ and Hapalidæ entered the South American continent, it is less easy to conceive. The only rays of light we yet have on the subject are, the supposed affinities of the fossil Cænopithecus of the Swiss, and the Lemuravidæ of the North American Eocene, with both Cebidæ and Lemuroids, and the fact that in Miocene or Eocene times a mild climate prevailed up to the Arctic circle. The discovery of an undoubted Lemuroid in the Eocene of Europe, indicates that the great Northern Continent was probably the birthplace of this low type of mammal, and the source whence Africa and Southern Asia were peopled with them, as it was, at a later period, with the higher forms of monkeys and apes.

The Pteropidæ, or fruit-eating Bats, sometimes called flying-foxes, are pretty evenly distributed over the tropical regions of the Old World and Australia. They range over all Africa and the whole of the Oriental Region, and northward, to Amoy in China and to the South of Japan. They are also found in the more fertile parts of Australia and Tasmania, and in the Pacific Islands as far east as the Marianne and Samoa Islands; but not in the Sandwich Islands or New Zealand.

The genera of bats are exceedingly numerous, but they are in a very unsettled state, and the synonymy is exceedingly confused. The details of their distribution cannot therefore be usefully entered into here. The Pteropidæ differ so much from all other bats, that they are considered to form a distinct suborder of Chiroptera, and by some naturalists even a distinct order of Mammalia.

No fossil Pteropidæ have been discovered.

The Phyllostomidæ, or simple leaf-nosed Bats, are confined to the Neotropical region, from Mexico and the Antilles to the southern limits of the forest region east of the Andes, and to about lat. 33° S. in Chili. None are found in the Nearctic region, with the exception of one species in California (Macrotus Californicus), closely allied to Mexican and West Indian forms. The celebrated blood-sucking vampyre bats of South America belong to this group. Two genera, Desmodus and Diphylla, form Dr. Peters' family Desmodidæ. Mr. Dobson, in his recently published arrangement, divides the family into five groups:—Mormopes, Vampyri, Glossophagæ, Stenodermata, and Desmodontes.

Numerous remains of extinct species of this family have been found in the bone-caves of Brazil.

The Rhinolophidæ, or Horse-shoe Bats (so-called from a curiously-shaped membranous appendance to the nose), range over all the Ethiopian and Oriental regions, the southern part of the Palæarctic region, Australia and Tasmania. They are most abundant and varied in the Oriental region, where twelve genera are found; while only five inhabit the Australian and Ethiopian regions respectively. Europe has only one genus and four species, mostly found in the southern parts, and none going further north than the latitude of England, where two species occur. Two others are found in Japan, at the opposite extremity of the Palæarctic region.

The genera Nycteris and Megaderma, which range over the Ethiopian and Oriental regions to the Moluccas, are considered by Dr. Peters to form a distinct family, Megadermidæ; and Mr. Dobson in his recent arrangement (published after our first volume was printed) adopts the same family under the name of Nycteridæ. The curious Indian genus Rhinopoma, which, following Dr. J. E. Gray, we have classed in this family, is considered by Mr. Dobson to belong to the Noctilionidæ.

Fossil Rhinolophidæ.—Remains of a species of Rhinolophus still living in England, have been found in Kent's Cavern, near Torquay.

The small bats constituting the family Vespertilionidæ, have no nose-membrane, but an internal earlet or tragus, and often very large ears. They range over almost the whole globe, being apparently only limited by the necessity of procuring insect food. In America they are found as far north as Hudson's Bay and the Columbia river; and in Europe they approach, if they do not pass the Arctic circle. Such remote islands as the Azores, Bermudas, Fiji Islands, Sandwich Islands, and New Zealand, all possess species of this group of bats, some of which probably inhabit every island in warm or temperate parts of the globe.

The genus Taphozous, which, in our Tables of Distribution in vol. i. we have included in this family, is placed by Mr. Dobson in his family Emballonuridæ, which is equivalent to our next family, Noctilionidæ.

Fossil Vespertilionidæ.—Several living European bats of this family—Scotophilus murinus, Plecotus auritus, Vespertilio noctula, and V. pipestrellus—have been found fossil in bone-caves in various parts of Europe.

Extinct species of Vespertilio have occurred in the Lower Miocene at Mayence, in the Upper Miocene of the South of France, and in the Upper Eocene of the Paris basin.

The Noctilionidæ, or short-headed Bats, are found in every region, but are very unequally distributed. Their head-quarters is the Neotropical region, where most of the genera occur, and where they range from Mexico to Buenos Ayres and Chili, while in North America there is only one species in California. They are unknown in Australia; but one species occurs in New Zealand, and another in Norfolk Island. Several species of Dysopes (or Molossus) inhabit the Oriental region; one or two species being widely distributed over the continent, while two others inhabit the Indo-Malayan Islands. A species of this same genus occurs in South Africa, and another in Madagascar and in the Island of Bourbon; while one inhabits Southern Europe and North Africa, and another is found at Amoy in China. It will be seen therefore, that these are really South American bats, which have a few allies widely scattered over the various regions of the globe. Their affinities are, according to Mr. Tomes, with the Phyllostomidæ, a purely South American family. The species which forms the connecting link is the Mystacina tuberculata, a New Zealand bat, which may, with almost equal propriety be placed in either family, and which affords an interesting illustration of the many points of resemblance between the Australian and Neotropical regions.

Dr. Peters has separated this family into three,—Mormopidæ, which is wholly Neotropical, and is especially abundant in the West Indian Islands; Molossidæ, chiefly consisting of the genus Molossus; and Noctilionidæ, comprising the remainder of the family, and wholly Neotropical. Mr. Dobson, however, classes the Mormopes with the Phyllostomidæ, and reduces the Molossi to the rank of a sub-family. In our first volume we have classed Rhinopoma with the Rhinolophidæ, and Taphozous with the Vespertilionidæ; but according to Mr. Dobson both these genera belong to the present family.

Although the bats, from their great powers of flight, are not amenable to the limitations which determine the distribution of other terrestrial mammals, yet certain great facts of distribution come out in a very striking manner. The speciality of the Neotropical region is well shown, not only by its exclusive possession of one large family (Phyllostomidæ), but almost equally so by the total absence of two others (Pteropidæ and Rhinolophidæ). The Nearctic region is also unusually well marked, by the total absence of a family (Rhinolophidæ) which is tolerably well represented in the Palæarctic. The Pteropidæ well characterize the tropical regions of the Old World and Australia; while the Vespertilionidæ are more characteristic of the Palæarctic and Nearctic regions, which together possess about 60 species of this family.

The bats are a very difficult study, and it is quite uncertain how many distinct species are really known. Schinz, in his Synopsis Mammalium (1844) describes 330, while the list given by Mr. Andrew Murray in his Geographical Distribution of Mammalia (1866), contains 400 species. A small number of new species have been since described, but others have been sunk as synonyms, so that we can perhaps hardly obtain a nearer approximation to the truth than the last number. In Europe there are 35 species, and only 17 in North America.

Fossil Chiroptera.—The fossil remains of bats that have yet been discovered, being chiefly allied to forms still existing in the same countries, throw no light on the origin or affinities of this remarkable and isolated order of Mammalia; but as species very similar to those now living were in existence so far back as Miocene or even Eocene times, we may be sure the group is one of immense antiquity, and that there has been ample time for the amount of variation and extinction required to bring about the limitation of types, and the peculiarities of distribution we now find to exist.

The singular and isolated genus Galeopithecus, or flying lemur, has been usually placed among the Lemuroidea, but it is now considered to come best at the head of the Insectivora. Its food however, seems to be purely vegetable, and the very small, blind, and naked young, closely attached to the wrinkled skin of the mother's breast, perhaps indicates some affinity with the Marsupials. This animal seems, in fact, to be a lateral offshoot of some low form, which has survived during the process of development of the Insectivora, the Lemuroidea, and the Marsupials, from an ancestral type. Only two species are known, one found in Malacca, Sumatra, and Borneo, but not in Java; the other in the Philippine islands (Plate VIII. vol. i. p. 337).

The Macroscelides, or elephant shrews, are extraordinary little animals, with trunk-like snout and kangaroo-like hind-legs. They are almost confined to South Africa, whence they extend up the east coast as far as the Zambezi and Mozambique. A single outlying species of Macroscelides inhabits Barbary and Algeria; while the two genera Petrodromus, and Rhynchocyon, each represented by a single species, have only been found at Mozambique.

The Tupaiidæ are squirrel-like shrews, having bushy tails, and often climbing up trees, but also feeding on the ground and among low bushes. The typical Tupaia (7 species), are called ground squirrels by the Malays. They are most abundant in the Malay islands and Indo-Chinese countries, but one species is found in the Khasia Mountains, and one in the Eastern Ghauts near Madras. The small shorter-tailed Hylomys (2 species) is found from Tenasserim to Java and Borneo; while the elegant little Ptilocerus (1 species) with its long pencilled tail, is confined to Borneo; (Plate VIII. vol. i. p. 337). The family is therefore especially Malayan, with outlying species in northern and continental India.

Extinct Species.—Oxygomphus, found in the Tertiary deposits of Germany, is believed to belong to this family; as is Omomys, from the Pliocene of the United States.

The Hedgehogs, comprised in the genus Erinaceus (14 species), are widely distributed over the Palæarctic, and a part of the Oriental regions; but they only occur in the Ethiopian region in South Africa and in the Deserts of the north, which more properly belong to the Palæarctic region. They are absent from the Malayan, and also from the Indo-Chinese sub-regions; except that they extend from the north of China to Amoy and Formosa and into the temperate highlands of the Western Himalayas. The curious Gymnura (1 species) is found in Borneo, Sumatra, and the Malay peninsula.

Extinct Species.—The common hedgehog has been found fossil in several Post-tertiary deposits, while extinct species occur in the lower Miocene of Auvergne and in some other parts of Europe. Many of these remains are classed in different genera from the living species;—(Amphechinus, Tetracus, Galerix.)

The Centetidæ are small animals, many of them having a spiny covering, whence the species of Centetes have been called Madagascar hedgehogs. The genera Centetes (2 species), Hemicentetes (1 species), Ericulus (1 species), Echinops (3 species), and the recently described Oryzorictes (1 species), are all exclusively inhabitants of Madagascar, and are almost or quite tail-less. The remaining genus, Solenodon, is a more slender and active animal, with a long, rat-like tail, shrew-like head, and coarse fur; and the two known species are among the very few indigenous mammals of the West Indian islands, one being found in Cuba (Plate XVII., vol. ii., p. 67), the other in Hayti. Although presenting many points of difference in detail, the essential characters of this curious animal are, according to Professors Peters and Mivart, identical with the rest of the Centetidæ. We have thus a most remarkable and well-established case of discontinuous distribution, two portions of the same family being now separated from each other by an extensive continent, as well as by a deep ocean.

Extinct Species.—Remains found in the Lower Miocene of the South of France are believed to belong to the genus Echinops, or one closely allied to it.

The genus Potamogale was founded on a curious, small, otter-like animal from West Africa, first found by M. Du Chaillu at the Gaboon, and afterwards by the Portuguese at Angola. Its affinities are with several groups of Insectivora, but it is sufficiently peculiar to require the establishment of a distinct family for its reception. (Plate V., vol. i., p. 264.)

The Chrysochloridæ, or golden moles, of the Cape of Good Hope have been separated by Professor Mivart into two genera, Chrysochloris and Chalcochloris. They are remarkable mole-like animals, having beautiful silky fur, with a metallic lustre and changeable golden tints. They are peculiar to the Cape district, but one species extends as far north as the Mozambique territory. Their dentition is altogether peculiar, so as to completely separate them from the true moles.

The Moles comprise many extraordinary forms of small mammalia especially characteristic of the temperate regions of the northern hemisphere, only sending out a few species of Talpa along the Himalayas as far as Assam, and even to Tenasserim, if there is no mistake about this locality; while one species is found in Formosa, the northern part of which is almost as much Palæarctic as Oriental. The genus Talpa (7 species), spreads over the whole Palæarctic region from Great Britain to Japan; Scaptochirus (1 species) is a recent discovery in North China; Condylura (1 species), the star-nosed mole, inhabits Eastern North America from Nova Scotia to Pennsylvania; Scapanus (2 species) ranges across from New York to St. Francisco; Scalops (3 species), the shrew-moles, range from Mexico to the great lakes on the east side of America, but on the west only to the north of Oregon. An allied genus, Myogale (2 species), has a curious discontinuous distribution in Europe, one species being found in South-East Russia, the other in the Pyrenees (Plate II., vol. i., p. 218). Another allied genus, Nectogale (1 species), has recently been described by Professor Milne-Edwards from Thibet. Urotrichus is a shrew-like mole which inhabits Japan, and a second species has been discovered in the mountains of British Columbia; an allied form, Uropsilus, inhabits East Thibet. Anurosorex and Scaptonyx, are new genera from North China.

Extinct Species.—The common mole has been found fossil in bone-caves and diluvial deposits, and several extinct species of mole-like animals occur in the Miocene deposits of the South of France and of Germany. These have been described under the generic names Dimylus, Geotrypus, Hyporissus, Galeospalax; while Palæospalax has been found in the Pliocene forest-beds of Norfolk and Ostend. Species of Myogale also occur from the Miocene downwards.

The Shrews have a wide distribution, being found throughout every region except the Australian and Neotropical; although, as a species is found in Timor and in some of the Moluccas, they just enter this part of the former region, while one found in Guatemala brings them into the latter. A number of species have recently been described from India and the Malay Islands, so that the Oriental region is now the richest in shrews, having 28 species; the Nearctic comes next with 24; while the Ethiopian has 11, and the Palæarctic 10 species. The sub-genera are Crossopus, Amphisorex, Neosorex, Crocidura, Diplomesodon, Pinulia, Pachyura, Blarina, Feroculus, Anausorex.

Extinct Species.—Several species of Sorex have been found fossil in the Miocene of the South of France, as well as the extinct genera Mysarachne and Plesiosorex; and some existing species have occurred in Bone Caves and Diluvial deposits.

The most prominent features in the distribution of the Insectivora are,—their complete absence from South America and Australia; the presence of Solenodon in two of the West Indian islands while the five allied genera are found only in Madagascar; and the absence of hedgehogs from North America. If we consider that there are only 135 known species of the order, 65 of which belong to the one genus Sorex; while the remaining 26 genera contain only 70 species, which have to be classed in 8 distinct families, and present such divergent and highly specialized forms as Galeopithecus, Erinaceus, Solenodon, and Condylura, it becomes evident that we have here the detached fragments of a much more extensive group of animals, now almost extinct. Many of the forms continue to exist only in islands, removed from the severe competition of a varied mammalian population, as in Madagascar and the Antilles; while others appear to have escaped extermination either by their peculiar habits—as the various forms of Moles; by special protection—as in the Hedgehogs; or by a resemblance in form, coloration, and habits to dominant groups in their own district—as the Tupaias of Malay which resemble squirrels, and the Elephant-shrews of Africa which resemble the jerboas. The numerous cases of isolated and discontinuous distribution among the Insectivora, offer no difficulty from this point of view; since they are the necessary results of an extensive and widely-spread group of animals slowly becoming extinct, and continuing to exist only where special conditions have enabled them to maintain themselves in the struggle with more highly organized forms.

The fossil Insectivora do not throw much light on the early history of the order, since even as far back as the Miocene period they consist almost wholly of forms which can be referred to existing families. In North America they go back to the Eocene period, if certain doubtful remains have been rightly placed. The occurrence of fossil Centetidæ in Europe, supports the view we have maintained in preceding chapters, that the existing distribution of this family between Madagascar and the Antilles, proves no direct connection between those islands, but only shows us that the family once had an extensive range.

The Cats are very widely distributed over the earth—with the exception of the Australian region and the island sub-region of Madagascar and the Antilles—universally; ranging from the torrid zone to the Arctic regions and the Straits of Magellan. They are so uniform in their organization that many naturalists group them all under one genus, Felis; but it is now more usual to class at least the lynxes as a separate genus, while the hunting leopard, or cheetah, forms another. Dr. J. E. Gray divides these again, and makes 17 generic groups; but as this subdivision is not generally adopted, and does not bring out any special features of geographical distribution, I shall not further notice it.

The genus Felis (56 species) has the same general range as the whole family, except that it does not go so far north; the Amoor river in Eastern Asia, and 55° N. Lat. in America, marking its limits. Lyncus (10 species) is a more northern group, ranging to the polar regions in Europe and Asia, and to Lat. 66° N. in America, but not going further south than Northern Mexico and the European shores of the Mediterranean, except the caracal, which may be another genus, and which extends to Central India, Persia, North Africa and even the Cape of Good Hope. The lynxes are thus almost wholly peculiar to the Nearctic and Palæarctic regions. Cynælurus (1 species) the hunting leopard, ranges from Southern and Western India through Persia, Syria, Northern and Central Africa, to the Cape of Good Hope.

Extinct Felidæ.—More than twenty extinct species of true Felidæ have been described, ranging in time from the epoch of prehistoric man back to the Miocene or even the Eocene period. They occur in the south of England, in Central and South Europe, in North-West India, in Nebraska in North America, and in the caves of Brazil. Most of them are referred to the genus Felis, and closely resemble the existing lions, tigers, and other large cats. Another group however forms the genus Machairodus, a highly specialized form with serrated teeth. Five species have been described from Europe, Northern India, and both North and South America; and it is remarkable that they exhibit at least as wide a range, both in space and time, as the more numerous species referred to Felis. One of them undoubtedly coexisted with man in England, while another, as well as the allied Dinictis, has been found in the Mauvaises Terres of Nebraska, associated with Anchitherium and other extinct and equally remarkable forms, which are certainly Miocene if not, as some geologists think, belonging to the Eocene period. These facts clearly indicate that we have as yet made little approach to discovering the epoch when Felidæ originated, since the oldest forms yet discovered are typical and highly specialized representatives of a group which is itself the most specialized of the Carnivora. Another genus, Pseudælurus, is common to the Miocene deposits of Europe and North America.

The Cryptoprocta ferox, a small and graceful cat-like animal, peculiar to Madagascar, was formerly classed among the Viverridæ, but is now considered by Professor Flower to constitute a distinct family between the Cats and the Civets.

The Viverridæ comprise a number of small and moderate-sized carnivorous animals, popularly known as civets, genets, and ichneumons, highly characteristic of the Ethiopian and Oriental regions, several of the genera being common to both. A species of Genetta, and one of Herpestes, inhabit South Europe; while Viverra extends to the Moluccas, but is doubtfully indigenous. The extreme geographical limits of the family are marked by Genetta in France and Spain, Viverra in Shanghae and Batchian Island, and Herpestes in Java and the Cape of Good Hope.

The following are the genera with their distribution as given by Dr. J. E. Gray in his latest British Museum Catalogue:

Sub-family .—Viverra (3 species), North and tropical Africa, the whole Oriental region to the Moluccas; Viverricula (1 species) India to Java; Genetta (5 species), South Europe, Palestine, Arabia, and all Africa; Fossa (1 species), Madagascar; Linsang (2 species), Malacca to Java; Poiana (1 species), West Africa; Galidia (3 species), Madagascar; Hemigalea (1 species), Malacca and Borneo; Arctictis (1 species) Nepal to Sumatra and Java; Nandinia (1 species), West Africa; Paradoxurus (9 species), the whole Oriental region; Paguma (3 species), Nepal to China, Sumatra, and Borneo; Arctogale (1 species), Tenasserim to Java.

Sub-family .—Cynogale (1 species), Borneo; Galidictis (2 species), Madagascar; Herpestes (22 species), South Palæarctic, Ethiopian, and Oriental regions; Athylax (3 species), Tropical and South Africa; Galogale (13 species), all Africa, North India, to Cambodja; Galerella (1 species), East Africa; Calictis (1 species), Ceylon (?); Ariella (1 species), South Africa; Ichneumia (4 species), Central, East, and South Africa; Bdeogale (3 species), West and East Africa; Urva (1 species), Himalayas to Aracan; Tæniogale (1 species), Central India; Onychogale (1 species), Ceylon; Helogale (2 species) East and South Africa; Cynictis (3 species), South Africa.

Sub-family .—Rhinogale (1 species), East Africa; Mungos (3 species), all Africa; Crossarchus (1 species), Tropical Africa; Eupleres (1 species), Madagascar; Suricata (1 species), South Africa.

Fossil Viverridæ.—Several species of Viverra and Genetta have been found in the Upper Miocene of France, and many extinct genera have also been discovered. The most remarkable of these was Ictitherium, from the Upper Miocene of Greece, which has also been found in Hungary, Bessarabia, and France. Some of the species were larger than any living forms of Viverridæ, and approached the hyænas. Other extinct genera are Thalassictis and Soricictis from the Upper Miocene, the former as large as a panther; Tylodon, of small size, from the Upper Eocene; and Palæonyctis from the Lower Eocene, also small and showing a very great antiquity for this family, if really belonging to it.

The curious Proteles or Aard-wolf, a highly-modified form of hyæna, approaching the ichneumons, and feeding on white ants and carrion, is peculiar to South Africa.

The Hyænas are characteristically Ethiopian, to which region two of the species are confined. The third, Hyæna striata, ranges over all the open country of India to the foot of the Himalayas, and through Persia, Asia Minor, and North Africa. Its fossil remains have been found in France.

Extinct Species.—The cave hyæna (H. spelæa) occurs abundantly in the caverns of this country and of Central Europe, and is supposed to be most nearly allied to the H. crocuta of South Africa. Another species is found in some parts of France. The earliest known true hyænas occur in the Pliocene formation in France, in the Red Crag (Older Pliocene) of England, and in the Upper Miocene of the Siwalik hills. In the Miocene period in Europe, quite distinct genera are found, such as Hyænictis and Lycæna from the Upper Miocene of Greece; Ictitherium, supposed to be intermediate between Viverridæ and Hyænidæ; and Thalassictis, uniting the weasels and hyænas.

The Canidæ, comprising the animals commonly known as dogs, wolves, and foxes, have an almost universal range over the earth, being only absent from the island sub-regions of Madagascar, the Antilles, Austro-Malaya, New Zealand, and the Pacific Islands. With the exception of two remarkable forms—the hyæna dog (Lycaon picta), and the great-eared fox (Megalotis Lalandei), both from South Africa—all the species are usually placed in the genus Canis, the distribution of which will be the same as that of the family. Dr. J. E. Gray, in his arrangement of the family (Proc. Zool. Soc., 1868), subdivides it into fifteen genera, the names and general distribution of which are as follows:—

Icticyon (1 species), Brazil; Cuon (4 species), Siberia to Java; Lupus (5 species), North America, Europe, India to Ceylon; Dieba (1 species), North and West Africa; Simenia (1 species), Abyssinia; Chrysocyon (2 species), North and South America; Canis (4 species), India, Australia (indigenous?) Lycalopex (2 species), South America; Pseudalopex (5 species), South America and Falkland Islands; Thous (2 species), South America to Chili; Vulpes (17 species), all the great continents, except South America and Australia; Fennecus (4 species), all Africa; Leucocyon (1 species), Arctic regions; Urocyon (2 species), North America; Nyctereutes (1 species), Japan, Amoorland to Canton (Plate III., vol. i. p. 226). These are all sub-genera according to Professor Carus, except Icticyon. The same author makes Lycaon a sub-genus, while Dr. Gray makes it a sub-family!

Extinct Species.—The dog, wolf, and fox, are found fossil in caverns in many parts of Europe, and several extinct species have been found in Tertiary deposits in Europe, North India, and South America. Two species have been found so far back as the Eocene of France, but the fragments discovered are not sufficient to determine the characters with any certainty. In North America, several species of Canis occur in the Pliocene of Nebraska and La Plata. The genus Galecynus, of the Pliocene of Œninghen, and Palæocyon, of the Brazilian caves, are supposed to belong to the Canidæ. Amphicyon abounded in the Miocene period, both in Europe and North America; and some of the species were as large as a tiger. Other extinct genera are, Cynodictis, Cyotherium, and Galethylax, from the Eocene of France; Pseudocyon, Simocyon, and Hemicyon, from the Miocene; but all these show transition characters to Viverridæ or Ursidæ, and do not perhaps belong to the present family.

The Mustelidæ constitute one of those groups which range over the whole of the great continental areas. They may be divided into three sub-families—one, the Mustelinæ, containing the weasels, gluttons, and allied forms; a second, the Lutrinæ, containing the otters; and a third, often considered a distinct family, the Melininæ, containing the badgers, ratels, skunks, and their allies.

In the first group (Mustelinæ) the genera Martes and Putorius (13 species), range over all the Palæarctic region, and a considerable part of the Oriental, extending through India to Ceylon, and to Java and Borneo. Two species of Martes (= Mustela of Baird) occur in the United States. The weasels, forming the genus Mustela (20 species), have a still wider range, extending into tropical Africa and the Cordilleras of Peru, but not going south of the Himalayas in India. The North American species are placed in the genus Putorius by Professor Baird. An allied genus, Gymnopus (4 species), is confined to the third and fourth Oriental sub-regions. Gulo (1 species), the glutton, is an arctic animal keeping to the cold regions of Europe and Asia, and coming as far south as the great lakes in North America. Galictis (2 species), the grisons, are confined to the Neotropical region.

The Otters (Lutrinæ) range over the whole area occupied by the family. They have been subdivided into a number of groups, such as Barangia (1 species), found only in Sumatra; Lontra, containing 3 South American species; Lutra (7 species), ranging over the whole of the Palæarctic and Oriental regions; Nutria (1 species), a sea-otter confined to the west coast of America from California to Chiloe; Lutronectes (1 species), from Japan only; Aonyx (5 species), found in West and South Africa, and the third and fourth Oriental sub-regions. Hydrogale (1 species), confined to South Africa; Latax (2 species), Florida and California to Canada and British Columbia; Pteronura (1 species), Brazil and Surinam; and Enhydris (1 species), the peculiar sea-otter of California, Kamschatka and Japan. The last two are the only groups of otters, besides Lutra, admitted by Professor Carus as genera.

The Badgers and allies (Melininæ) have also a wide range, but with one exception are absent from South America. They comprise the following genera: Arctonyx (1 species), Nepal to Aracan; Meles (4 species), North Europe to Japan, and China as far south as Hongkong (Plate I., vol. i., p. 195); Taxidea (2 species), Central and Western North America to 58° N. Lat.; Mydaus (1 species), mountains of Java and Sumatra; Melivora (3 species), Tropical and South Africa and India to foot of Himalayas; Mephitis (12 species), America from Canada and British Columbia to the Straits of Magellan (Plate XX., vol. ii., p. 136). Ictonyx (2 species), Tropical Africa to the Cape; Helictis (4 species), Nepal to Java, Formosa and Shanghai (Plate VII., vol. i. p. 331).

Fossil Mustelidæ.—Species of otter, weasel, badger, and glutton, occur in European bone caves and other Post-tertiary deposits; and in North America Galictis, now found only in the Neotropical region, and, with Mephitis, occurring in Brazilian caves.

Species of Mustela have been found in the Pliocene of France and of South America; and Lutra in the Pliocene of North America.

In the Miocene deposits of Europe several species of Mustela and Lutra have been found; with the extinct genera Taxodon, Potamotherium, and Palæomephitis; as well as Promephitis in Greece.

In the Upper Miocene of the Siwalik Hills species of Lutra and Mellivora are found, as well as the extinct genera Enhydrion and Ursitaxus.

The family appears to have been unknown in North America during the Miocene period.

The Procyonidæ are a small, but very curious and interesting family of bear-like quadrupeds, ranging from British Columbia and Canada on the north, to Paraguay and the limits of the tropical forests on the south.

The Racoons, forming the genus Procyon, are common all over North America; a well-marked variety or distinct species inhabiting the west coast, and another, most parts of South America. The genus Nasua, or the coatis (5 species?), extends from Mexico and Guatemala to Paraguay. The curious arboreal prehensile-tailed kinkagou (Cercoleptes candivolvus) is also found in Mexico and Guatemala, and in all the great forests of Peru and North Brazil. Bassaris (2 species), a small weasel-like animal with a banded tail, has been usually classed with the Viverridæ or Mustelidæ, but is now found to agree closely in all important points of internal structure with this family. It is found in California, Texas, and the highlands of Mexico, and belongs therefore as much to the Nearctic as to the Neotropical region. A second species has recently been described by Professor Peters from Coban in Guatemala, in which, country it has also been observed by Mr. Salvin.

Fossil Procyonidæ.—A species of Nasua has been found in the bone caves of Brazil, and a Procyon in the Pliocene or Post-pliocene deposits of Illinois and Carolina.

The Panda (Ælurus fulgens), of the forest regions of the Eastern Himalayas and East Thibet, a small cat-like bear, has peculiarities of organization which render it necessary to place it in a family by itself. (Plate VII. vol. i. p. 331). An allied genus, Æluropus, a remarkable animal of larger size and in colour nearly all white, has recently been described by Professor Milne-Edwards, from the mountains of East Thibet; so that the family may be said to inhabit the border lands of the Oriental and Palæarctic regions. These animals have their nearest allies in the coatis and bears.

The Bears have a tolerably wide distribution, although they are entirely absent from the Australian and Ethiopian, and almost so from the Neotropical region, one species only being found in the Andes of Peru and Chili. They comprise the following groups, some of which are doubtfully ranked as genera.

Thalassarctos, the polar bear (1 species) inhabiting the Arctic regions; Ursus, the true bears (12 species), which range over all the Nearctic and Palæarctic regions as far as the Atlas Mountains, the Indo-Chinese sub-region in the mountains, and to Hainan and Formosa; Helarctos, the Malay or sun-bear (1 species) confined to the Indo-Malayan sub-region; Melursus or Prochilus, the honey-bear (1 species), confined to the first and second Oriental sub-regions, over which it ranges from the Ganges to Ceylon; and Tremarctos, the spectacled bear—commonly known as Ursus ornatus—which is isolated in the Andes of Peru and Chili, and forms a distinct group.

Fossil Ursidæ.—Two bears (Ursus spelæus and U. priscus) closely allied to living species, abound in the Post-tertiary deposits of Europe; and others of the same age are found in North America, as well as an extinct genus, Arctodus.

Ursus arvernensis is found in the Pliocene formation of France, and the extinct genus Leptarchus in that of North America.

Several species of Amphicyon, which appears to be an ancestral form of this family, are found in the Miocene deposits of Europe and N. India; while Ursus also occurs in the Siwalik Hills and Nerbudda deposits.

The Otariidæ, or Eared Seals, comprehending the sea-bears and sea-lions, are confined to the temperate and cold shores of the North Pacific, and to similar climates in the Southern Hemisphere, where the larger proportion of the species are found. They are entirely absent from the North Atlantic shores. Mr. J. A. Allen, in his recent discussion of this family (Bull. Harvard Museum) divides them into the following genera:—

Otaria (1 species), Temperate South America, from Chili to La Plata; Callorhinus (1 species), Behring's Straits and Kamschatka; Arctocephalus (3 species), temperate regions of the Southern Hemisphere; Zalophus (2 species), North Pacific, from California to Japan, and the shores of Australia and New Zealand; Eumetopias (1 species), Behring's Straits and California.

Fossil Otariidæ.—Remains supposed to belong to this family have been found in the Miocene of France.

The Morse, or Walrus (Trichechus rosmarus), which alone constitutes this family, is a characteristic animal of the North Polar regions, hardly passing south of the Arctic circle except on the east and west coasts of North America, where it sometimes reaches Lat. 60°. It is most abundant on the shores of Spitzbergen, but is not found on the northern shores of Asia between Long. 80° and 160° E., or on the north shores of America from 100° to 150° west.

Its remains have been found fossil in Europe as far south as France, and in America as far as Virginia; but the small fragments discovered may render the identification uncertain.

The earless or true Seals are pretty equally divided between the Northern and Southern Hemispheres, frequenting almost exclusively the temperate and cold regions, except two species said to occur among the West Indian islands. The genus Phoca and its close allies, as well as Halichœrus and Pelagius, are northern; while Stenorhynchus and Morunga, with their allies, are mostly southern. The genera admitted by Dr. Gray in his catalogue are as follows:—

Callocephalus (3 species), Greenland, North Sea, also the Caspian Sea, and Lakes Aral and Baikal; Pagomys (2 species), North Sea, North Pacific, and Japan; Pagophilus (2 species), North Pacific and North Atlantic; Halicyon (1 species), North West coast of America; Phoca (2 species), North Atlantic and North Pacific, Japan; Halichœrus (1 species), Greenland, North Sea, and Baltic; Pelagius (2 species), Madeira, Mediterranean, Black Sea; Stenorhynchus (1 species), Antarctic Ocean, Falkland Islands, New Zealand; Lobodon (1 species), Antarctic Ocean; Leptonyx (1 species), Antarctic Ocean, South Australia, East Patagonia; Ommatophoca (1 species), Antarctic Ocean; Morunga (2 species), California, Falkland Islands, Temperate regions of Southern Ocean; Cystophora (2 species), North Atlantic, Antilles.

Fossil Seals.—Remains of living species of seals have been found in Post-tertiary deposits in many parts of Europe and in Algeria, as well as in New Zealand. Pristiphoca occitana is a fossil seal from the Pliocene of Montpellier, while a species of Phoca is said to have been found in the Miocene deposits of the United States.

Terrestrial Carnivora.—For the purposes of geographical distribution, the terrestrial and aquatic Carnivora differ too widely to be considered in one view, their areas being limited by barriers of a very different nature. The terrestrial Carnivora form a very extensive and considerably varied group of animals, having, with the doubtful exception of Australia, a world-wide distribution. Yet the range of modification of form is not very great, and the occurrence of three families consisting of but one species each, is an indication of a great amount of recent extinction. One of the most marked features presented by this group is its comparative scarcity in the Neotropical region, only four families being represented there (not counting the Ursidæ, which has only one Andean species), and both genera and species are few in number. Even the Procyonidæ, which are especially South American, have but two genera and six species in that vast area. We might therefore, from these considerations alone, conclude that Carnivora are a development of the northern hemisphere, and have been introduced into the Neotropical region at a comparatively recent epoch. The claim of the Nearctic region to be kept distinct from the Palæarctic (with which some writers have wished to unite it) is well maintained by its possession of at least six species of Mephitis, or skunk, a group having no close allies in any other region,—and the genera Procyon and Bassaris,—for the latter, ranging from the high lands of Guatemala and Mexico to Texas and California, may be considered a Nearctic rather than a Neotropical form. In the other families, the most marked feature is the total absence of Ursidæ from the Ethiopian region. The great mass of the generic forms of Carnivora, however, are found in the Oriental and Ethiopian regions, which possess all the extensive group of Viverridæ (except a few species in the fourth Palæarctic sub-region) and a large number of Felidæ and Mustelidæ.

Aquatic Carnivora.—The aquatic Carnivora present no very marked features of distribution, except their preference for cold and temperate rather than tropical seas. Their nearest approximation to the terrestrial group, is supposed to be that of the Otariidæ to the Ursidæ; but this must be very remote, and the occurrence of both seals and bears in the Miocene period, shows, that until we find some late Secondary or early Tertiary formation rich in Mammalian remains, we are not likely to get at the transition forms indicating the steps by which the aquatic Carnivora were developed. The most interesting special fact of distribution to be noticed, is the occurrence of seals, closely allied to those inhabiting the northern seas, in the Caspian, Lake Aral, and Lake Baikal. In the case of the two first-named localities there is little difficulty, as they are connected with the North Sea by extensive plains of low elevation, so that a depression of less than 500 feet would open a free communication with the ocean. At a comparatively recent epoch, a great gulf of the Arctic ocean must have occupied the valley of the Irtish, and extended to the Caspian Sea; till the elevation of the Kirghiz Steppes cut off the communication with the ocean, leaving an inland sea with its seals. Lake Baikal, however, offers much greater difficulties; since it is not only a fresh-water lake, but is situated in a mountain district nearly 2,000 feet above the sea level, and entirely separated from the plains by several hundred miles of high land. It is true that such an amount of submergence and elevation is known to have occurred in Europe so recently as during the Glacial period; but Lake Baikal is so surrounded by mountains, that it must at that time have been filled with ice, if at anything like its present elevation. Its emergence from the sea must therefore have taken place since the cold epoch, and this would imply that an enormous extent of Northern Asia has been very recently under water.

We are accustomed to look on Seals as animals which exclusively inhabit salt water; but it is probably from other causes than its saltness that they usually keep to the open sea, and there seems no reason why fresh-water should not suit them quite as well, provided they find in it a sufficiency of food, facilities for rearing their young, and freedom from the attacks of enemies. As already remarked in vol. i. p. 218, Mr. Belt's ingenious hypothesis (founded on personal examination of the Siberian Steppes), that during the Glacial period the northern ice-cap dammed up the waters of the northward flowing Asiatic rivers, and thus formed a vast fresh-water lake which might have risen as high as Lake Baikal, seems to offer the best solution of this curious problem of distribution.

Range of Carnivora in Time.—Carnivora have been found in all the Tertiary deposits, and comprise a number of extinct genera and even families. Several genera of Canidæ occur in the Upper Eocene of Europe; but the most remarkable fact is, that even in the Lower Eocene are found two well-marked forms, Palæonyctis, one of the Viverridæ, and Arctocyon, forming a distinct family type of very generalized characters, but unmistakably a carnivore. This last has been found at La Fère, in the north-east of France, in a deposit which, according to M. Gaudry, is the very lowest of the Lower Eocene formation in Europe. Arctocyon is therefore one of the oldest, if not the very oldest, of the higher forms of mammal yet discovered.

—Temperate and Cold Seas of both Northern and Southern Hemispheres.

This family comprises the whalebone or "right" whales, the best known species being the Greenland whale (Balæna mysticetus). Allied species are found in all parts of the southern seas, as far north as the Cape of Good Hope; while some of the northern species are found off the coast of Spain, and even enter the Mediterranean. As most of the species indicated are imperfectly known, and their classification by no means well settled, no useful purpose will be served by enumerating the genera or sub-genera.

—Cold and Temperate Seas of both Hemispheres.

This family comprises the finner whales and rorquals, and are characterised by possessing a dorsal fin and having the baleen or whalebone less developed. They are abundant in all northern seas, less so in the southern hemisphere, but they seem occasionally to enter the tropical seas. The best known genera are Megaptera (7 species); Physalus (4 species); and Balænoptera (2 species); all of which have species in the North Sea.

—All the Tropical Oceans, extending north and south into Temperate waters.

This family, comprising the cachalots or sperm whales, and black-fish, are separated from the true whales by having teeth in the lower jaw and no whalebone. They are pre-eminently a tropical, as distinguished from the two preceding which are arctic and antarctic families. The spermaceti whale (Catodon macrocephalus) abounds in the Pacific Ocean and in the deep Moluccan Sea, and also in the Indian Ocean and the Mozambique Channel. In the Atlantic it is scarce, although it occasionally comes north as far as our shores.

The genera of Catodontidæ as given by Dr. Gray are, Catodon (2 species?), Warm Eastern Oceans; Physeter (1 species), "the black fish," North Sea; Cogia (2 species), South Temperate Oceans; Euphysetes (1 species), Coast of Australia.

—Atlantic, Mediterranean, Indian Ocean, and Southern Ocean.

This family consists of the beaked whales, which have no permanent teeth in the upper jaw. The genera, according to Dr. Gray, are, Hyperoodon (2 species) "bottle-nosed whales," North Sea; Lagenocetus (1 species), North Sea; Epiodon (2 species), North and South Atlantic; Petrorhynchus (2 species), Mediterranean Sea and Southern Ocean; Berardius (1 species), New Zealand; Xiphius (1 species) North Atlantic; Dolichodon (1 species), Cape of Good Hope; Neoziphius (1 species) Mediterranean; Dioplodon (1 species), Indian Ocean.

The "Narwhal" (Monodon monoceros) which constitutes this family, is placed by Dr. Gray along with the "white whales," in his family Belugidæ. It inhabits the North Sea.

—All Oceans, Seas, and Great Rivers of the globe.

This family, including the Porpoises, Dolphins, White Whales, &c., may be described as small, fish-shaped whales, having teeth in both jaws. According to Dr. Gray they form seven families and 24 genera; according to Professor Carus, four sub-families and 8 genera, but as these groups appear to be established on quite different principles, and often differ widely from each other, I shall simply enumerate Dr. Gray's genera with their distribution as given in his British Museum Catalogue.

Platanista (2 species), long-snouted porpoises, inhabiting the Ganges and Indus; Inia (1 species), a somewhat similar form, inhabiting the upper waters of the Amazonian rivers: Steno (8 species), Indian Ocean, Cape of Good Hope, and West Pacific; Sotalia (1 species), Guiana; Delphinus (10 species), all the oceans; Clymenia (14 species), all the oceans; Delphinapterus (1 species), South Atlantic; Tursio (7 species), Atlantic and Indian Oceans; Eutropia (2 species), Chili, and Cape of Good Hope; Electra, (8 species), all the oceans; Leucopleurus (1 species), North Sea; Lagenorhynchus (1 species), North Sea; Pseudorca (2 species), North Sea, Tasmania; Orcaella (2 species), Ganges; Acanthodelphis (1 species), Brazil; Phocæna (2 species), North Sea; Neomeris (1 species), India; Grampus (3 species), North Sea, Mediterranean, Cape of Good Hope; Globiocephalus (14 species), all the oceans; Sphærocephalus (1 species), North Atlantic; Orca (9 species), Northern and Southern Oceans; Ophysia (1 species), North Pacific; Beluga (6 species), Arctic Seas, Australia; Pontoporia (1 species), Monte Video.

Remains of Cetacea are tolerably abundant in Tertiary deposits, both in Europe and North America. In the Lower Pliocene of England, France, and Germany, extinct species of five or six living genera of whales and dolphins have been found; and most of these occur also in the Upper Miocene, along with many others, referred to about a dozen extinct genera.

In the Post-pliocene deposits of Vermont and South Carolina, several extinct species have been found belonging to living genera; but in the Miocene deposits of the Eastern United States cetacean remains are much more abundant, more than 30 species of extinct whales and dolphins having been described, most of them belonging to extinct genera.

The Zeuglodontidæ, an extinct family of carnivorous whales, with double-fanged serrated molar teeth, whose affinities are somewhat doubtful, are found in the older Pliocene of Europe, and in the Miocene and Eocene of the Eastern United States. Zeuglodon abounds in the United States, and one species reached a length of seventy feet. A species of this genus is said to have been found in Malta. Squalodon occurs in Europe and North America; and in the latter country four or five other genera have been described, of which one, Saurocetes, has been found also at Buenos Ayres.

The Sea-cows are herbivorous aquatic animals living on the coasts or in the great rivers of several parts of the globe. Manatus (2 species) inhabits both shores of the Atlantic, one species ranging from the Gulf of Mexico to North Brazil, and ascending the Amazon far into the interior of the continent; while the other is found on the west coast of Africa. Halicore (2 species?), the Dugong, is peculiar to the Indian Ocean, extending from Mozambique to the Red Sea, thence to Western India and Ceylon, the Malay Archipelago and the north coast of Australia. Rhytina (1 species), supposed to be now extinct, inhabited recently the North Pacific, between Kamschatka and Behring's Straits.

Fossil Sirenia.—Extinct species of Manatus have been found in the Post-pliocene deposits of Eastern North America from Maryland to Florida; and an extinct genus, Prorastomus, in some Tertiary deposits in the Island of Jamaica.

In Post-pliocene deposits in Siberia, remains of Rhytina have been found; while several species of the extinct genus Halitherium, perhaps intermediate between Manatus and Halicore, have been found in the older Pliocene and Upper Miocene of France and Germany.

The Horses, Asses, and Zebras form a highly specialized group now confined to the Ethiopian and Palæarctic regions, but during the middle and later tertiaries having a very extensive range. The zebras (3 species) inhabit the greater part of the Ethiopian region, while the asses (4 species) are characteristic of the deserts of the Palæarctic region from North Africa and Syria to Western India, Mongolia, and Manchuria. The domestic horse is not known in a wild state, but its remains are found in recent deposits from Britain to the Altai Mountains, so that its disappearance is probably due to human agency.

Extinct Equidæ.—Extinct forms of this family are very numerous. The genus Equus occurs in Post-pliocene and Pliocene deposits in Europe, North America, and South America. In North America the species are most numerous. An allied genus Hipparion, having rudimentary lateral toes, is represented by several species in the Pliocene of North America, while in Europe it occurs both in the Older Pliocene and Upper Miocene. Various other allied forms, in which the lateral toes are more and more developed, and most of which are now classed in a distinct family, Anchitheridæ, range back through the Miocene to the Eocene period. A sufficient account of these has already been given in vol. i. chap. vi. p. 135, to which the reader is referred for the supposed origin and migrations of the horse.

The Tapirs form a small group of animals whose discontinuous distribution plainly indicates their approaching extinction. For a long time only two species were known, the black American, and the white-banded Malay tapir, the former confined to the equatorial forests of South America, the latter to the Malay peninsula, Sumatra, and Borneo (Plate VIII. vol. i. p. 337). Lately however another, or perhaps two distinct species (or according to Dr. J. E. Gray, four!) have been discovered in the Andes of New Granada and Ecuador, at an elevation of from 8,000 to 12,000 feet; while one or perhaps two more, forming the allied genus Elasmognathus, have been found to inhabit Central America from Panama to Guatemala.

Extinct Tapirs.—True tapirs inhabited Western Europe, from the latest Pliocene back to the earliest Miocene times; while they only occur in either North or South America in the Post-pliocene deposits and caves. The singular distribution of the living species is thus explained, since we see that they are an Old World group which only entered the American continent at a comparatively recent epoch. An ancestral form of this group—Lophiodon—is found in Miocene and Eocene deposits of Europe and North America; while a still more ancient form of large size is found in the Lower Eocene of France and England, indicating an immense antiquity for this group of Mammalia. There are many other extinct forms connecting these with the Palæotheridæ, already noticed in chapter vi. (vol. i. pp. 119-125).

Living Rhinoceroses are especially characteristic of Africa, with Northern and Malayan India. Four or perhaps five species, all two-horned, are found in Africa, where they range over the whole country south of the desert to the Cape of Good Hope. In the Oriental region there are also four or five species, which range from the forests at the foot of the Himalayas eastwards through Assam, Chittagong, and Siam, to Sumatra, Borneo and Java. Three of these are one-horned, the others found in Sumatra, and northwards to Pegu and Chittagong, two-horned. The Asiatic differ from the African species in some dental characters, but they are in other respects so much alike that they are not generally considered to form distinct genera. In his latest catalogue however (1873), Dr. Gray has four genera, Rhinoceros (4 species), and Ceratorhinus (2 species), Asiatic; Rhinaster (2 species), and Ceratotherium (2 species), African.

Extinct Rhinocerotidæ.—Numerous species of Rhinoceros ranged over Europe and Asia from the Post-pliocene back to the Upper Miocene period, and in North America during the Pliocene period only. The hornless Acerotherium is Miocene only, in both countries. Other genera are Leptodon from Greece, and Hyracodon from Nebraska, both of Miocene age. More than 20 species of extinct rhinoceroses are known, and one has even been found at an altitude of 16,000 feet in Thibet.

The Hippopotamus inhabits all the great rivers of Africa; a distinct species of a smaller size being found on the west coast, and on some of the rivers flowing into Lake Tchad.

Fossil Hippopotami.—Eight extinct species of Hippopotamus are known from Europe and India, the former Post-pliocene or Pliocene, the latter of Upper Miocene age. They ranged as far north as the Thames valley. An extinct genus from the Siwalik Hills, Merycopotamus, according to Dr. Falconer connects Hippopotamus with Anthracotherium, an extinct form from the Miocene of Europe, allied to the swine.

The Swine may be divided into three well-marked groups, from peculiarities in their dentition. 1. The Dicotylinæ, or peccaries (1 genus, Dicotyles). These offer so many structural differences that they are often classed as a separate family. 2. The true swine (3 genera, Sus, Potamochœrus, and Babirusa); and, 3. The Phacochœrinæ, or wart hogs (1 genus, Phacochœrus). These last are also sometimes made into a separate family, but they are hardly so distinct as the Dicotylinæ.

The Peccaries (2 species), are peculiar to the Neotropical region, extending from Mexico to Paraguay. They also spread northwards into Texas, and as far as the Red River of Arkansas, thus just entering the Nearctic region; but with this exception swine are wholly absent from this region, forming an excellent feature by which to differentiate it from the Palæarctic.

Sus (14 species), ranges over the Palæarctic and Oriental regions and into the first Australian sub-region as far as New Guinea; but it is absent from the Ethiopian region, or barely enters it on the north-east. Potamochœrus (3 species?), is wholly Ethiopian (Plate V. vol. i. p. 278). Babirusa (1 species), is confined to two islands, Celebes and Bouru, in the first Australian sub-region.

Phacochœrus (2 species), ranges over tropical Africa from Abyssinia to Caffraria.

Dr. J. E. Gray divides true swine (Sus) into 7 genera, but it seems far better to keep them as one.

Fossil Suidæ.—These are very numerous. Many extinct species of wild hog (Sus), are found in Europe and North India, ranging back from the Post-pliocene to the Upper Miocene formations. In the Miocene of Europe are numerous extinct genera, Bothriodon, Anthracotherium, Palæochœrus, Hyotherium, and some others; while in the Upper Eocene occur Cebochœrus, Chœropotamus, and Acotherium,—these early forms having more resemblance to the peccaries.

None of these genera are found in America, where we have the living genus Dicotyles in the Post-pliocene and Pliocene deposits, both of North and South America; with a number of extinct genera in the Miocene. The chief of these are, Elotherium, Perchœrus, Leptochœrus, and Nanohyus, all from Dakota, and Thinohyus, from Oregon. One extinct genus, Platygonus, closely allied to Dicotyles, is found in the Post-pliocene of Nebraska, Oregon, and Arkansas. Elotherium is said to be allied to the peccary and hippopotamus. Hyopotamus, from the Miocene of Dakota, is allied to Anthracotherium, and forms with it (according to Dr. Leidy) a distinct family of ancestral swine.

It thus appears, that the swine were almost equally well represented in North America and Europe, during Miocene and Pliocene times, but by entirely distinct forms; and it is a remarkable fact that these hardy omnivorous animals, should, like the horses, have entirely died out in North America, except a few peccaries which have preserved themselves in the sub-tropical parts and in the southern continent, to which they are comparatively recent emigrants. We can hardly have a more convincing proof of the vast physical changes that have occurred in the North American continent during the Pliocene and Post-pliocene epochs, than the complete extinction of these, along with so many other remarkable types of Mammalia.

According to M. Gaudry, the ancestors of all the swine, with the hippopotami and extinct Anthracotherium, Merycopotamus, and many allied forms,—are the Hyracotherium and Pliolophus, both found only in the London clay belonging to the Lower Eocene formation.

The Camels are an exceedingly restricted group, the majority of the species now existing only in a state of domestication. The genus Camelus (2 species), is a highly characteristic desert form of the Palæarctic region, from the Sahara to Mongolia as far as Lake Baikal. Auchenia (4 species), comprehending the Llamas and Alpacas, is equally characteristic of the mountains and deserts of the southern part of South America. Two species entirely domesticated inhabit the Peruvian and Bolivian Andes; and two others are found in a wild state, the vicuna in the Andes of Peru and Chili (Plate XVI. vol. ii. p. 40), and the guanaco over the plains of Patagonia and Tierra del Fuego.

Extinct Camelidæ.—No fossil remains of camels have been found in Europe, but one occurs in the deposits of the Siwalik Hills, usually classed as Upper Miocene, but which some naturalists think are more likely of Older Pliocene age. Merycotherium, teeth of which have been found in the Siberian drift, is supposed to belong to this family.

In North America, where no representative of the family now exists, the camel-tribe were once abundant. In the Post-pliocene deposits of California an Auchenia has been found, and in those of Kansas one of the extinct genus Procamelus. In the Pliocene period, this genus, which was closely allied to the living camels, abounded, six or seven species having been described from Nebraska and Texas, together with an allied form Homocamelus. In the Miocene period different genera appear,—Pœbrotherium, and Protomeryx,—while a Procamelus has been found in deposits of this age in Virginia.

In South America a species of Auchenia has been found in the caves of Brazil, and others in the Pliocene deposits of the pampas, together with two extinct genera, Palæolama and Camelotherium.

We thus find the ancestors of the Camelidæ in a region where they do not now exist, but which is situated so that the now widely separated living forms could easily have been derived from it. This case offers a remarkable example of the light thrown by palæontology on the distribution of living animals; and it is a warning against the too common practice of assuming the direct land connection of remote continents, in order to explain similar instances of discontinuous distribution to that of the present family.

The Tragulidæ are a group of small, hornless, deer-like animals, with tusks in the upper jaw, and having some structural affinities with the camels. The musk-deer was formerly classed in this family, which it resembles externally; but a minute examination of its structure by M. Milne-Edwards, has shown it to be more nearly allied to the true deer. The Chevrotains, or mouse-deer, Tragulus (5 species), range over all India to the foot of the Himalayas and Ceylon, and through Assam, Malacca, and Cambodja, to Sumatra, Borneo, and Java (Plate VIII., vol. i. p. 337). Hyomoschus (1 species), is found in West Africa.

Extinct Tragulidæ.—A species of Hyomoschus is said to have been found in the Miocene of the South of France, as well as three extinct genera, Dremotherium (also found in Greece), with Lophiomeryx from the Upper Miocene, said to be allied to Tragulus; and Amphitragulus from the Lower Miocene, of more remote affinities, and sometimes placed among the Deer. There seems to be no doubt, however, that this family existed in Europe in Miocene times; and thus another case of discontinuous distribution is satisfactorily accounted for.

The Cervidæ, or deer tribe, are an extensive group of animals equally adapted for inhabiting forests or open plains, the Arctic regions or the Tropics. They range in fact over the whole of the great continents of the globe, with the one striking exception of Africa, where they are only found on the shores of the Mediterranean which form part of the Palæarctic region. The following is the distribution of the genera.

Alces (1 species), the elk or moose, ranges all over Northern Europe and Asia, as far south as East Prussia, the Caucasus, and North China; and over Arctic America to Maine on the East, and British Columbia on the west. The American species may however be distinct, although very closely allied to that of Europe. Tarandus (1 species), the reindeer, has a similar range to the last, but keeps farther north in Europe, inhabiting Greenland and Spitzbergen; and in America extends farther south, to New Brunswick and the north shore of Lake Superior. There are several varieties or species of this animal confined to special districts, but they are not yet well determined. Cervus (40 species), the true deer, have been sub-divided into numerous sub-genera characteristic of separate districts. They range over the whole area of the family, except that they do not go beyond 57° N. in America and a little further in Europe and Asia. In South America they extend over Patagonia and even to Tierra del Fuego. They are found in the north of Africa, and over the whole of the Oriental region, and beyond it as far as the Moluccas and Timor, where however they have probably been introduced by man at an early period. Dama (1 species), the fallow deer, is a native of the shores of the Mediterranean, from Spain and Barbary to Syria. Capreolus (2 species), the roe-deer, inhabits all Temperate and South Europe to Syria, with a distinct species in N. China. Cervulus (4 species), the muntjacs, are found in all the forest districts of the Oriental region, from India and Ceylon to China as far north as Ningpo and Formosa, also southward to the Philippines, Borneo, and Java. Moschus (1 species), the musk-deer, inhabits Central Asia from the Amoor and Pekin, to the Himalayas and the Siamese mountains above 8000 ft. elevation. This is usually classed as a distinct family, but M. Milne-Edwards remarks, that it differs in no important points of organisation from the rest of the Cervidæ. Hydropotes (1 species) inhabits China from the Yang-tse Kiang northwards. This new genus has recently been discovered by Mr. Swinhoe, who says its nearest affinities are with Moschus. Other new forms are Lophotragus, and Elaphodus, both inhabiting North China; the former is hornless, the latter has very small horns about an inch long.

Extinct Deer.—Numerous extinct species of the genus Cervus are found fossil in many parts of Europe, and in all formations between the Post-pliocene and the Upper Miocene. The Elk and Reindeer are also found in caves and Post-pliocene deposits, the latter as far south as the South of France. Extinct genera only, occur in the Upper Miocene in various parts of Europe:—Micromeryx, Palæomeryx, and Dicrocerus have been described; with others referred doubtfully to Moschus, and an allied genus Amphimoschus.

In N. America, remains of this family are very scarce, a Cervus allied to the existing wapiti deer, being found in Post-pliocene deposits, and an extinct genus, Leptomeryx, in the Upper Miocene of Dakota and Oregon. Another extinct genus, Merycodus, from the Pliocene of Oregon, is said to be allied to camels and deer.

In South America, several species of Cervus have been found in the Brazilian caves, and in the Pliocene deposits of La Plata.

It thus appears, that there are not yet sufficient materials for determining the origin and migrations of the Cervidæ. There can be little doubt that they are an Old World group, and a comparatively recent development; and that some time during the Miocene period they passed to North America, and subsequently to the Southern continent. They do not however appear to have developed much in North America, owing perhaps to their finding the country already amply stocked with numerous forms of indigenous Ungulates.

The Camelopardalidæ, or giraffes, now consist of but a single species which ranges over all the open country of the Ethiopian region, and is therefore almost absent from West Africa, which is more especially a forest district. During the Middle Tertiary period, however, these animals had a wider range, over Southern Europe and Western India as far as the slopes of the Himalayas.

Extinct Species.—Species of Camelopardalis have been found in Greece, the Siwalik Hills, and Perim Island at the entrance to the Red Sea; and an extinct genus, Helladotherium, more bulky but not so tall as the giraffe, ranged from the south of France to Greece and North-west India.

This large and important family, includes all the animals commonly known as oxen, buffaloes, antelopes, sheep, and goats, which have been classed by many naturalists in at least three, and sometimes four or five, distinct families. Zoologically, they are briefly and accurately defined as, "hollow-horned ruminants;" and, although they present wide differences in external form, they grade so insensibly into each other, that no satisfactory definition of the smaller family groups can be found. As a whole they are almost confined to the great Old World continent, only a few forms extending along the highlands and prairies of the Nearctic region; while one peculiar type is found in Celebes, an island which is almost intermediate between the Oriental and Australian regions. In each of the Old World regions there are found a characteristic set of types. Antelopes prevail in the Ethiopian region; sheep and goats in the Palæarctic; while the oxen are perhaps best developed in the Oriental region.

Sir Victor Brooke, who has paid special attention to this family, divides them into 13 sub-families, and I here adopt the arrangement of the genera and species which he has been so good as to communicate to me in MSS.

Sub-family I. (6 genera, 13 species). This group is one of the best marked in the family. It comprises the Oxen and Buffaloes with their allies, and has a distribution very nearly the same as that of the entire family. The genera are as follows: (1 sp.), now represented by our domestic cattle, the descendants of the Bos primigenius, which ranged over a large part of Central Europe in the time of the Romans. The Chillingham wild cattle are supposed to be the nearest approach to the original species. Bison (2 sp.), one still wild in Poland and the Caucasus; the other in North America, ranging over the prairies west of the Mississippi, and on the eastern slopes of the Rocky Mountains (Plate XIX., vol. ii., p. 129). Bibos (3 sp.), the Indian wild cattle, ranging over a large part of the Oriental region, from Southern India to Assam, Burmah, the Malay Peninsula, Borneo, and Java. Poephagus (1 sp.), the yak, confined to the high plains of Western Thibet. Bubalus (5 sp.), the buffaloes, of which three species are African, ranging over all the continental parts of the Ethiopian region; one Northern and Central Indian; and the domesticated animal in South Europe and North Africa. Anoa (1 sp.), the small wild cow of Celebes, a very peculiar form more nearly allied to the buffaloes than to any other type of oxen.

Sub-family II. : (3 genera, 11 species). The Bovine Antelopes are large and handsome animals, mostly Ethiopian, but extending into the adjacent parts of the Palæarctic and Oriental regions. The genera are: Oreas (2 sp.), elands, inhabiting all Tropical and South Africa. Tragelaphus (8 sp.), including the bosch-bok, kudu, and other large antelopes, ranges over all Tropical and South Africa (Plate IV., vol. i., p. 261). Portax (1 sp.) India, but rare in Madras and north of the Ganges.

Sub-family III. : (2 genera, 5 species). Oryx (4 sp.) is a desert genus, ranging over all the African deserts to South Arabia and Syria; Addax (1 sp.) inhabits North Africa, North Arabia, and Syria.

Sub-family IV. (1 genus, 3 species). The Sable Antelopes, Hippotragus, form an isolated group inhabiting the open country of Tropical Africa and south to the Cape.

Sub-family V. (6 genera, 23 species). This is a group of small or moderate-sized animals, most abundant in the deserts on the borders of the Palæarctic, Oriental, and Ethiopian regions. Gazella (17 sp.) is typically a Palæarctic desert group, ranging over the great desert plateaus of North Africa, from Senegal and Abyssinia to Syria, Persia, Beloochistan, and the plains of India, with one outlying species in South Africa. Procapra (2 sp.), Western Thibet and Mongolia to about 110° east longitude. Antilope (1 sp.) inhabits all the plains of India. Æpyceros (1 sp.) the pallah, inhabits the open country of South and South-east Africa. Saiga (1 sp.) a singular sheep-faced antelope, which inhabits the steppes of Eastern Europe and Western Asia from Poland to the Irtish River, south of 55° north latitude. (Plate II., vol. i., p. 218.) Panthalops (1 sp.) confined to the highlands of Western Thibet and perhaps Turkestan.

Sub-family VI. (1 genus, 1 species), Antilocapra, the prong-horned antelope, inhabit both sides of the Rocky Mountains, extending north to the Saskatchewan and Columbia River, west to the coast range of California, and east to the Missouri. Its remarkable deciduous horns seem to indicate a transition to the Cervidæ. (Plate XIX., vol. ii., p. 129.)

Sub-family VII. (5 genera, 21 species). This group of Antelopes is wholly confined to the continental portion of the Ethiopian region. The genera are: Cervicapra (4 sp.), Africa, south of the equator and Abyssinia; Kobus (6 sp.), grassy plains and marshes of Tropical Africa; Pelea (1 sp.), South Africa; Nanotragus (9 species), Africa, south of the Sahara; Neotragus (1 sp.) Abyssinia and East Africa.

Sub-family VIII. (2 genera, 24 species), Africa and India; Cephalophus (22 sp.), continental Ethiopian region; Tetraceros (2 sp.) hilly part of all India, but rare north of the Ganges.

Sub-family IX. (2 genera, 11 species), large African Antelopes, one species just entering the Palæarctic region. The genera are: Alcephalus (9 sp.) all Africa and north-east to Syria; Catoblepas (2 sp.), gnus, Africa, south of the Equator.

Sub-region X. (1 genus, 2 species) Budorcas inhabits the high Himalayas from Nepal to East Thibet.

Sub-family XI. (1 genus, 2 species) the Chamois, Rupicapra, inhabit the high European Alps from the Pyrenees to the Caucasus. (Plate I., vol. i., p. 195.)

Sub-family XII. (2 genera, 10 species). These goat-like Antelopes inhabit portions of the Palæarctic and Oriental regions, as well as the Rocky Mountains in the Nearctic region. Nemorhedus (9 sp.) ranges from the Eastern Himalayas to N. China and Japan, and south to Formosa, the Malay Peninsula and Sumatra. Aplocerus (1 sp.), the mountain goat of the trappers, inhabits the northern parts of California and the Rocky Mountains.

Sub-family XIII. (2 genera, 23 species). The Goats and Sheep form an extensive series, highly characteristic of the Palæarctic region, but with an outlying species on the Neilgherries in Southern India, and one in the Rocky Mountains and California. The genera are Capra (22 sp.) and Ovibos (1 sp.). The genus Capra consists of several sub-groups which have been named as genera, but it is unnecessary here to do more than divide them into "Goats and Ibexes" on the one hand and "Sheep" on the other—each comprising 11 species. The former range over all the South European Alps from Spain to the Caucasus; to Abyssinia, Persia, and Scinde; over the high Himalayas to E. Thibet and N. China; with an outlying species in the Neilgherries. The latter are only found in the mountains of Corsica, Sardinia, and Crete, in Europe; in Asia Minor, Persia, and in Central and North-Eastern Asia, with one somewhat isolated species in the Atlas mountains; while in America a species is found in the Rocky Mountains and the coast range of California. Ovibos (1 sp.), the musk-sheep, inhabits Arctic America north of lat. 60; but it occurs fossil in Post-glacial gravels on the Yena and Obi in Siberia, in Germany and France along with the Mammoth and with flint implements, and in caves of the Reindeer period; also in the brick earth in the south of England, associated with Rhinoceros megarhinus and Elephas antiquus.

Extinct Bovidæ.—In the caverns and diluviums of Europe, of the Post-Pliocene period, the remains are found of extinct species of Bos, Bison, and Capra; and in the caverns of the south of France Rupicapra, and an antelope near Hippotragus. Bos and Bison also occur in Pliocene deposits. In the Miocene of Europe, the only remains are antelopes closely allied to existing species, and these are especially numerous in Greece, where remains referred to two living and four extinct genera have been discovered. In the Miocene of India numerous extinct species of Bos, and two extinct genera, Hemibos and Amphibos, have been found, one of them at a great elevation in Thibet. Antelopes, allied to living Indian species, are chiefly found in the Nerbudda deposits.

In North America, the only bovine remains are those of a Bison, and a sheep or goat, in the Post-pliocene deposits; and of two species of musk-sheep, sometimes classed in a distinct genus Bootherium, from beds of the same age in Arkansas and Ohio. Casoryx, from the Pliocene of Nebraska, is supposed to be allied to the antelopes and to deer.

In the caves of Brazil remains of two animals said to be antelopes, have been discovered. They are classed by Gervais in the genera Antilope and Leptotherium, but the presence of true antelopes in S. America at this period is so improbable, that there is probably some error of identification.

The extinct family Sivatheridæ, containing the extraordinary and gigantic four-horned Sivatherium and Bramatherium, of the Siwalik deposits, are most nearly allied to the antelopes.

From the preceding facts we may conclude, that the great existing development of the Bovidæ is comparatively recent. The type may have originated early in the Miocene period, the oxen being at first most tropical, while the antelopes inhabited the desert zone a little further north. The sheep and goats seem to be the most recent development of the bovine type, which was probably long confined to the Eastern Hemisphere.

With the exception of the Australian region, from which this order of mammalia is almost entirely wanting, the Ungulata are almost universally distributed over the continental parts of all the other regions. Of the ten families, 7 are Ethiopian, 6 Oriental, 5 Palæarctic, 4 Neotropical, and 3 Nearctic. The Ethiopian region owes its superiority to the exclusive possession of the hippopotamus and giraffe, both of which inhabited the Palæarctic and Oriental regions in Miocene times. The excessive poverty of the Nearctic region in this order is remarkable; the swine being represented only by Dicotyles in its extreme southern portion, while the Bovidæ are restricted to four isolated species. Deer alone are fairly well represented. But, during the Eocene and Miocene periods, North America was wonderfully rich in varied forms of Ungulates, of which there were at least 8 or 9 families; while we have reason to believe that during the same periods the Ethiopian region was excessively poor, and that it probably received the ancestors of all its existing families from Europe or Western Asia in later Miocene or Pliocene times. Many types that once abounded in both Europe and North America are now preserved only in South America and Central or Tropical Asia,—as the tapirs and camels; while others once confined to Europe and Asia have found a refuge in Africa,—as the hippopotamus and giraffe; so that in no other order do we find such striking examples of those radical changes in the distribution of the higher animals which were effected during the latter part of the Tertiary period. The present distribution of this order is, in fact, utterly unintelligible without reference to the numerous extinct forms of existing and allied families; but as this subject has been sufficiently discussed in the Second Part of this work (Chapters VI. and VII.) it is unnecessary to give further details here.

The elephants are now represented by two species, the African, which ranges all over that continent south of the Sahara, and the Indian, which is found over all the wooded parts of the Oriental region, from the slopes of the Himalayas to Ceylon, and eastward, to the frontiers of China and to Sumatra and Borneo. These, however, are but the feeble remnants of a host of gigantic creatures, which roamed over all the great continents except Australia during the Tertiary period, and several of which were contemporary with man.

Extinct Elephants.—At least 14 extinct species of Elephas, and a rather greater number of the allied genus Mastodon (distinguished by their less complex grinding teeth) have now been discovered. Elephants ranged over all the Palæarctic and Nearctic regions in Post-Pliocene times; in Europe and Central India they go back to the Pliocene; and only in India to the Upper Miocene period; the number of species increasing as we go back to the older formations.

In North America two or three species of Mastodon are Post-pliocene and Pliocene; and a species is found in the caves of Brazil, and in the Pliocene deposits of the pampas of La Plata, of the Bolivian Andes, and of Honduras and the Bahamas. In Europe the genus is Upper Miocene and Pliocene, but is especially abundant in the former period. In the East, it extends from Perim island to Burmah and over all India, and is mostly Miocene, but with perhaps one species Pliocene in Central India.

An account of the range of such animals as belong to extinct families of Proboscidea, will be found in Chapters VI. and VII.; from which it will be seen that, although the family Elephantidæ undoubtedly originated in the Eastern Hemisphere, it is not improbable that the first traces of the order Proboscidea are to be found in N. America.

The genus Hyrax, which alone constitutes this family, consists of small animals having the appearance of hares or marmots, but which more resemble the genus Rhinoceros in their teeth and skeleton. They range all over the Ethiopian region, except Madagascar; a peculiar species is found in Fernando Po, and they just enter the Palæarctic as far as Syria. They may therefore be considered as an exclusively Ethiopian group. In Dr. Gray's last Catalogue (1873) he divides the genus into three—Hyrax, Euhyrax and Dendrohyrax—the latter consisting of two species confined apparently to West and South Africa.

No extinct forms of this family have yet been discovered; the Hyracotherium of the London clay (Lower Eocene) which was supposed to resemble Hyrax, is now believed to be an ancestral type of the Suidæ or swine.

The Muridæ, comprising the rats and mice with their allies, are almost universally distributed over the globe (even not reckoning the domestic species which have been introduced almost everywhere by man), the exceptions being the three insular groups belonging to the Australian region, from none of which have any species yet been obtained. Before enumerating the genera it will be as well to say a few words on the peculiarities of distribution they present. The true mice, forming the genus Mus, is distributed over the whole of the world except N. and S. America where not a single indigenous species occurs, being replaced by the genus Hesperomys; five other genera, comprehending all the remaining species found in South America are peculiar to the Neotropical region. Three genera are confined to the Palæarctic region, and three others to the Nearctic. No less than twelve genera are exclusively Ethiopian, while only three are exclusively Oriental and three Australian.

Mus (100-120 sp.) the Eastern Hemisphere, but absent from the Pacific and Austro-Malayan Islands, except Celebes and Papua; Lasiomys (1 sp.) Guinea; Acanthomys (5-6 sp.) Africa, India and N. Australia; Cricetomys (1 sp.) Tropical Africa; Saccostomus (2 sp.) Mozambique; Cricetus (9 sp.) Palæarctic region and Egypt; Cricetulus (1 sp., Milne-Edwards, 1870) Pekin; Pseudomys (1 sp.) Australia; Hapalotis (13 sp.) Australia; Phlæomys (1 sp.) Philippines; Platacanthomys (1 sp., Blyth, 1865) Malabar; Dendromys (2 sp.) S. Africa; Nesomys (1 sp. Peters, 1870) Madagascar; Steatomys (2 sp.) N. and S. Africa; Pelomys (1 sp.) Mozambique; Reithrodon (9 sp.) N. America, Lat. 29° to Mexico, and south to Tierra del Fuego; Acodon (1 sp.) Peru; Myxomys (1 sp.) Guatemala; Hesperomys (90 sp.) North and South America; Holochilus (4 sp.) South America; Oxymycterus (4 sp.) Brazil and La Plata; Neotoma (6 sp.) U.S., East coast to California; Sigmodon (2 sp.) Southern United States; Drymomys (1 sp.) Peru; Neotomys (2 sp.) S. America; Otomys (6 sp.) S. and E. Africa; Meriones = Gerbillus (20-30 sp.) Egypt, Central Asia, India, Africa; Rhombomys (6 sp.) S. E. Europe, N. Africa, Central Asia; Malacothrix (2 sp.) South Africa; Mystromys (1 sp.) South Africa; Psammomys (1 sp.) Egypt; Spalacomys (1 sp.) India; Sminthus (1-3 sp.) East Europe, Tartary, Siberia; Hydromys (5 sp.) Australia and Tasmania; Hypogeomys (1 sp., Grandidier, 1870) Madagascar; Brachytarsomys (1 sp., Günther, 1874) Madagascar; Fiber (2 sp.) N. America to Mexico; Arvicola (50 sp.) Europe to Asia Minor, North Asia, Himalayas, Temp. N. America; Cuniculus (1 sp.) N. E. Europe, Siberia, Greenland, Arctic America; Myodes (4 sp.) Europe, Siberia, Arctic America, and Northern United States; Myospalax = Siphneus (2 sp.) Altai Mountains and N. China ; Lophiomys (1 sp.) S. Arabia, and N. E. Africa; Echiothrix (1 sp.) Australia.

Extinct Muridæ.—Species of Mus, Cricetus, Arvicola, and Myodes, occur in the Post-Pliocene deposits of Europe; Arvicola, Meriones, and the extinct genus Cricetodon, with some others, in the Miocene.

In North America, Fiber, Arvicola, and Neotoma, occur in caves; an extinct genus, Eumys, in the Upper Miocene of Dakota, and another, Mysops, in the Eocene of Wyoming.

In South America Mus, or more probably Hesperomys, is abundant in Brazilian caverns, and Oxymycterus in the Pliocene of La Plata; while Arvicola is said to have occurred both in the Pliocene and Eocene deposits of the same country.

The Spalacidæ, or mole-rats, have a straggling distribution over the Old World continents. They are found over nearly the whole of Africa, but only in the South-east of Europe, and West of Temperate Asia, but appearing again in North India, Malacca, and South China. Ellobius (1 sp.), is found in South Russia and South-west Siberia; Spalax (1 sp.), Southern Russia, West Asia, Hungary, Moldavia, and Greece (Plate II., vol. i. p. 218); Rhizomys (6 sp.), Abyssinia, North India, Malacca, South China; Heterocephalus (1 sp.), Abyssinia; Bathyerges (= Orycterus 1 sp.), South Africa; Georychus (6 sp.), South, Central, and East Africa; Heliophobus (1 sp.), Mozambique.

The Jerboas, or jumping mice, are especially characteristic of the regions about the eastern extremity of the Mediterranean, being found in South Russia, the Caspian district, Arabia, Egypt, and Abyssinia; but they also extend over a large part of Africa, and eastward to India; while isolated forms occur in North America, and the Cape of Good Hope. Dipus = Gerbillus (20 sp.), inhabits North and Central Africa, South-East Europe, and across Temperate Asia to North China, also Afghanistan, India, and Ceylon; Pedetes (1 sp.), South Africa to Mozambique and Angola; Jaculus = Meriones (1 sp.), North America, from Nova Scotia and Canada, south to Pennsylvania and west to California and British Columbia (Plate XX., vol. ii. p. 135).

Extinct Dipodidæ.—Dipus occurs fossil in the Miocene of the Alps; and an extinct genus, Issiodromys, said to be allied to Pedetes of the Cape of Good Hope, is from the Pliocene formations of Auvergne in France.

The Dormice (Myoxus), are small rodents found over all the temperate parts of the Palæarctic region, from Britain to Japan; and also over most parts of Africa to the Cape, but wanting in India. Some of the African species have been separated under the name of Graphidurus, while those of Europe and Asia form the sub-genera Glis, Muscardinus, and Eliomys.

Extinct Myoxidæ.—Myoxus ranges from the Post-pliocene of the Maltese caverns to the Miocene of Switzerland and the Upper Eocene of France; and an extinct genus Brachymys is found in the Miocene of Central Europe.

The Saccomyidæ, or pouched rats, are almost wholly confined to our second Nearctic sub-region, comprising the Rocky Mountains and the elevated plains of Central North America. A few species range from this district as far as Hudson's Bay on the north, to South Carolina on the east, and to California on the west, while one genus, doubtfully placed here, goes south as far as Honduras and Trinidad. The group must therefore be considered to be pre-eminently characteristic of the Nearctic region.

The genera are,—Dipodomys (5 sp.), North Mexico, California, the east slope of the Rocky Mountains to the Columbia River, and one species in South Carolina; Perognathus (6 sp.), North Mexico, California, east slope of the Rocky Mountains to British Columbia; Thomomys (2 sp.), Upper Missouri, and Upper Columbia Rivers to Hudson's Bay; Geomys (5 sp.), North Mexico, and east slope of Rocky Mountains to Nebraska (Plate XIX., vol. ii. p. 129); Saccomys (1 sp.), North America, locality unknown; Heteromys (6 sp.), Mexico, Honduras, and Trinidad. Geomys and Thomomys constitute a separate family Geomyidæ, of Professor Carus; but I follow Professor Lilljeborg, who has made a special study of the Order, in keeping them with this family.

In the Post-Pliocene deposits of Illinois and Nebraska, remains of an existing species of Geomys have been found.

The Beavers, forming the genus Castor, consist of two species, the American (Castor canadensis) ranging over the whole of North America from Labrador to North Mexico; while the European (Castor fiber) appears to be confined to the temperate regions of Europe and Asia, from France to the River Amoor, over which extensive region it doubtless roamed in prehistoric times, although now becoming rare in many districts.

Extinct Castoridæ.—Extinct species of Castor range back from the Post-pliocene to the Upper Miocene in Europe, and to the Newer Pliocene in North America. Extinct genera in Europe are, Trogontherium, Post-Pliocene and Pliocene; Chalicomys, Older Pliocene; and Steneofiber, Upper Miocene. In North America Castoroides is Post-Pliocene, and Palæocastor, Upper Miocene. The family thus first appears on the same geological horizon in both Europe and North America.

The Squirrel family, comprehending also the marmots and prairie-dogs, are very widely spread over the earth. They are especially abundant in the Nearctic, Palæarctic, and Oriental regions, and rather less frequent in the Ethiopian and Neotropical, in which last region they do not extend south of Paraguay. They are absent from the West Indian islands, Madagascar, and Australia, only occurring in Celebes which doubtfully belongs to the Australian region. The genera are as follows:—

Sciurus (100-120 sp., including the sub-genera Spermosciurus, Xerus, Macroxus, Rheithrosciurus, and Rhinosciurus), comprises the true squirrels, and occupies the area of the whole family wherever woods and forests occur. The approximate number of species in each region is as follows: Nearctic 18, Palæarctic 6, Ethiopian 18, Oriental 50, Australian (Celebes) 5, Neotropical 30. Sciuropterus (16-19 sp.), comprises the flat-tailed flying squirrels, which range from Lapland and Finland to North China and Japan, and southward through India and Ceylon, to Malacca and Java, with a species in Formosa; while in North America they occur from Labrador to British Columbia, and south to Minnesota and Southern California. Pteromys (12 sp.), comprising the round-tailed flying squirrels, is a more southern form, being confined to the wooded regions of India from the Western Himalayas to Java and Borneo, with species in Formosa and Japan. Tamias (5 sp.), the ground squirrels, are chiefly North American, ranging from Mexico to Puget's Sound on the west coast, and from Virginia to Montreal on the Atlantic coast; while one species is found over all northern Asia. Spermophilus (26 sp.), the pouched marmots, are confined to the Nearctic and Palæarctic regions; in the former extending from the Arctic Ocean to Mexico and the west coast, but not passing east of Lake Michigan and the lower Mississippi; in the latter from Silesia through South Russia to the Amoor and Kamschatka, most abundant in the desert plains of Tartary and Mongolia. Arctomys (8 sp.), the marmots, are found in the northern parts of North America as far down as Virginia and Nebraska to the Rocky Mountains and British Columbia, but not in California; and from the Swiss Alps eastward to Lake Baikal and Kamschatka, and south as far as the Himalayas, above 8,000 feet elevation. Cynomys (2 sp.), the prairie-dogs, inhabit the plains east of the Rocky Mountains from the Upper Missouri to the Red River and Rio Grande (Plate XIX., vol. ii. p. 129). Anomalurus (5 sp.), consists of animals which resemble flying-squirrels, but differ from all other members of the family in some points of internal structure. They form a very aberrant portion of the Sciuridæ, and, according to some naturalists, a distinct family. They inhabit West Africa and the island of Fernando Po.

Extinct Sciuridæ.—These are tolerably abundant. The genus Sciurus appears to be a remarkably ancient form, extinct species being found in the Miocene, and even in the Upper Eocene formations of Europe. Spermophilus goes back to the Upper Miocene; Arctomys to the Newer Pliocene. Extinct genera are, Brachymys, Lithomys and Plesiarctomys, from the European Miocene, the latter said to be intermediate between marmots and squirrels.

In North America, Sciurus, Tamias, and Arctomys occur in the Post-pliocene deposits only. The extinct genera are Ischyromys, from the Upper Miocene of Nebraska; Paramys, allied to the marmots, and Sciuravus, near the squirrels, from the Eocene of Wyoming.

Here we have unmistakable evidence that the true squirrels (Sciurus) are an Old World type, which has only recently entered North America; and this is in accordance with the comparative scarcity of this group in South America, a country so well adapted to them, and their great abundance in the Oriental region, which, with the Palæarctic, was probably the country of their origin and early development. The family, however, has been traced equally far back in Europe and North America, so that we have as yet no means of determining where it originated.

The genus Haploodon or Aplodontia, consists of two curious rat-like animals, inhabiting the west coast of America, from the southern part of British Columbia to the mountains of California. They seem to have affinities both with the beavers and marmots, and Professor Lilljeborg constitutes a separate family to receive them.

The Chinchillidæ, including the chinchillas and viscachas, are confined to the alpine zones of the Andes, from the boundary of Ecuador and Peru to the southern parts of Chili; and over the Pampas, to the Rio Negro on the south, and the River Uruguay on the east. Chinchilla (2 sp.), the true chinchillas, are found in the Andes of Chili and Peru, south of 9° S. lat., and from 8,000 to 12,000 feet elevation (Plate XVI. vol. ii. p. 40); Lagidium (3 sp.), the alpine viscachas, inhabit the loftiest plateaus and mountains from 11,000 to 16,000 feet, and extend furthest north of any of the family; while Lagostomus (1 sp.), the viscacha of the Pampas, has the range above indicated. The family is thus confined within the limits of a single sub-region.

Extinct Chinchillidæ.—Lagostomus has been found fossil in the caves of Brazil, and in the Pliocene deposits of La Plata. The only known extinct forms of this family are Amblyrhiza and Loxomylus, found in cavern-deposits in the island of Anguilla, of Post-Pliocene age. These are very interesting, as showing the greater range of this family so recently; though its absence from North America and Europe indicates that it is a peculiar development of the Neotropical region.

The Octodontidæ include a number of curious and obscure rat-like animals, mostly confined to the mountains and open plains of South America, but having a few stragglers in other parts of the world, as will be seen by our notes on the genera. The most remarkable point in their distribution is, that two genera are peculiar to the West Indian islands, while no species of the family inhabits the northern half of South America. The distribution of the genera is as follows:—Habrocomus (2 sp.), Chili; Capromys (3 sp.), two of which inhabit Cuba, the third Jamaica (Plate XVII. vol. ii. p. 67); Plagiodontia (1 sp.), only known from Hayti; Spalacopus, including Schizodon (2 sp.), Chili, and east side of Southern Andes; Octodon (3 sp.), Peru, Bolivia, and Chili; Ctenomys (6 sp.), the tuco-tuco of the Pampas, the Campos of Brazil to Bolivia and Tierra del Fuego; Ctenodactylus (1 sp.), Tripoli, North Africa; Pectinator (1 sp.), East Africa, Abyssinia, 4,000 to 5,000 feet.

Capromys and Plagiodontia, the two West Indian genera, were classed among the Echimyidæ by Mr. Waterhouse, but Professor Lilljeborg removes them to this family.

Extinct Octodontidæ.—Species of Ctenomys have been found in the Pliocene of La Plata, and an extinct genus Megamys, said to be allied to Capromys, in the Eocene of the same country. In Europe, Palæomys and Archæomys from the lower Miocene of Germany and France, are also said to be allied to Capromys.

The Echimyidæ, or spiny rats, are a family, chiefly South American, of which the Coypu, a large beaver-like water-rat from Peru and Chili is the best known. Two of the genera are found in South Africa, but all the rest inhabit the continent of South America, East of the Andes, none being yet known north of Panama. The genera are as follows:—Dactylomys (2 sp.), Guiana and Brazil; Cercomys (1 sp.), Central Brazil; Lasiuromys (1 sp.), San Paulo, Brazil; Petromys (1 sp.), South Africa; Myopotamus (1 sp.), the coypu, on the East side of the Andes from Peru to 42° S. lat., on the West side from 33° to 48° S. lat.; Carterodon (1 sp.), Minaes Geraes, Brazil; Aulacodes (1. sp.), West and South Africa; Mesomys (1 sp.), Borba on the Amazon; Echimys (11 sp.), from Guiana and the Ecuadorian Andes to Paraguay; Loncheres (10 sp.), New Granada to Brazil.

Fossil and Extinct Echimyidæ.—The genus Carterodon was established on bones found in the Brazilian caves, and it was several years afterwards that specimens were obtained showing the animal to be a living species. Extinct species of Myopotamus and Loncheres have also been found in these caves, with the extinct genera Lonchophorus and Phyllomys.

No remains of this family have been discovered in North America; but in the Miocene and Upper Eocene deposits of France there are many species of an extinct genus Theridomys, which is said to be allied to this group or to the next (Cercolabidæ). Aulacodon, from the Upper Miocene of Germany, is allied to the West African Aulacodes; and some other remains from the lower Miocene of Auvergne, are supposed to belong to Echimys.

The Cercolabidæ, or arboreal porcupines, are a group of rodents entirely confined to America, where they range from the northern limit of trees on the Mackenzie River, to the southern limit of forests in Paraguay. There is however an intervening district, the Southern United States, from which they are absent. Erethizon (3 sp.), the Canadian porcupine, is found throughout Canada and as far south as Northern Pennsylvania, and west to the Mississippi (Plate XX., vol. ii. p. 135); an allied species inhabiting the west coast from California to Alaska, and inland to the head of the Missouri River; while a third is found in the north-western part of South America; Cercolabes (12 sp.), ranges from Mexico and Guatemala to Paraguay, on the eastern side of the Andes; Chætomys (1 sp.), North Brazil.

Extinct Cercolabidæ.—A large species of Cercolabes has been found in the Brazilian caves, but none have been discovered in North America or Europe. We may conclude therefore that this is probably a South American type, which has thence spread into North America at a comparatively recent epoch. The peculiar distribution of Cercolabes may be explained by supposing it to have migrated northwards along the west coast by means of the wooded slopes of the Rocky Mountains. It could then only reach the Eastern States by way of the forest region of the great lakes, and then move southward. This it may be now doing, but it has not yet reached the Southern States of Eastern North America.

The true Porcupines have a very compact and well-marked distribution, over the whole of the Oriental and Ethiopian regions (except Madagascar), and the second Palæarctic sub-region. There is some confusion as to their sub-division into genera, but the following are those most usually admitted:—Hystrix (5 sp.), South Europe to the Cape of Good Hope, all India, Ceylon, and South China; Atherura (5 sp.), "brush-tailed porcupines," inhabit West Africa, India, to Siam, Sumatra, and Borneo; Acanthion (2 sp.), Nepal and Malacca, to Sumatra, Borneo, and Java.

Extinct Hystricidæ.—Several extinct species of Hystrix have been found in the Pliocene and Miocene deposits of Europe, and one in the Pliocene of Nebraska in North America.

The Cavies and Agoutis were placed in distinct families by Mr. Waterhouse, in which he is followed by Professor Carus, but they have been united by Professor Lilljeborg, and without pretending to decide which classification is the more correct I follow the latter, because there is a striking external resemblance between the two groups, and they have an identical distribution in the Neotropical region, and with one exception are all found east of the Andes. Dasyprocta (9 sp.), the agouti, ranges from Mexico to Paraguay, one species inhabiting the small West Indian islands of St. Vincent, Lucia, and Grenada; Cælogenys (2 sp.), the paca, is found from Guatemala to Paraguay, and a second species (somewhat doubtful) in Eastern Peru; Hydrochœrus (1 sp.), the capybara inhabits the banks of rivers from Guayana to La Plata; Cavia (9 sp.), the guinea-pigs, Brazil to the Straits of Magellan, and one species west of the Andes at Yça Peru; Kerodon (6 sp.), Brazil and Peru to Magellan; Dolichotis (1 sp,), the Patagonian cavy, from Mendoza to 48° 30′ south latitude, on sterile plains.

Extinct Caviidæ.—Hydrochœrus, Cælogenys, Dasyprocta, and Kerodon, have occurred abundantly in the caves of Brazil, and the last-named genus in the Pliocene of La Plata. Hydrochœrus has been found in the Post-Pliocene deposits of South Carolina. Cavia and Dasyprocta are said to have been found in the Miocene of Switzerland and France. No well-marked extinct genera of this family have been recorded.

If the determination of the above-mentioned fossil species of Cavia and Dasyprocta are correct, it would show that this now exclusively South American family is really derived from Europe, where it has long been extinct.

The Lagomyidæ, or pikas, are small alpine and desert animals which range from the south of the Ural Mountains to Cashmere and the Himalayas, at heights of 11,000 to 14,000 feet, and northward to the Polar regions and the north-eastern extremity of Siberia. They just enter the eastern extremity of Europe as far as the Volga, but with this exception, seem strictly limited to the third Palæarctic sub-region. In America they are confined to the Rocky Mountains from about 42° to 60° north latitude.

Extinct Lagomyidæ.—Extinct species of Lagomys have occurred in the southern parts of Europe, from the Post-Pliocene to the Miocene formations. Titanomys, an extinct genus, is found in the Miocene of France and Germany.

The Hares and Rabbits are especially characteristic of the Nearctic and Palæarctic, but are also thinly scattered over the Ethiopian and Oriental regions. In the Neotropical region they are very scarce, only one species being found in South America, in the mountains of Brazil and various parts of the Andes, while one or two of the North American species extend into Mexico and Guatemala. In the Nearctic region, they are most abundant in the central and western parts of the continent, and they extend to the Arctic Ocean and to Greenland. They are found in every part of the Palæarctic region, from Ireland to Japan; three species range over all India to Ceylon, and others occur in Hainan, Formosa, South China, and the mountains of Pegu; the Ethiopian region has only four or five species, mostly in the southern extremity and along the East coast. An Indian species is now wild in some parts of Java, but it has probably been introduced.

Extinct Leporidæ.—Species of Lepus occur in the Post-Pliocene and Newer Pliocene of France; but only in the Post-Pliocene of North America, and the caves of Brazil.

With the exception of the Australian region and Madagascar, where Muridæ alone have been found, this order is one of the most universally and evenly distributed over the entire globe. Of the sixteen families which compose it, the Palæarctic region has 10; the Ethiopian, Nearctic, and Neotropical, each 9; and the Oriental only 5. These figures are very curious and suggestive. We know that the rodentia are exceedingly ancient, since some of the living genera date back to the Eocene period; and some ancestral types might thus have reached the remote South American and South African lands at the time of one of their earliest unions with the northern continents. In both these countries the rodents diverged into many special forms, and being small animals easily able to conceal themselves, have largely survived the introduction of higher Mammalia. In the Palæarctic and Nearctic regions, their small size and faculty of hibernation may have enabled them to maintain themselves during those great physical changes which resulted in the extermination or banishment of so many of the larger and more highly organised Mammalia, to which, in these regions, they now bear a somewhat inordinate proportion. The reasons why they are now less numerous and varied in the Oriental region, may be of two kinds. The comparatively small area of that region and its uniformity of climate, would naturally lead to less development of such a group as this, than in the vastly more extensive and varied and almost equally luxuriant Palæarctic region of Eocene and Miocene times; while on the other hand the greater number of the smaller Carnivora in the tropics during the Pliocene and Post-Pliocene epochs, would be a constant check upon the increase of these defenceless animals, and no doubt exterminate a number of them.

The Rodents thus offer a striking contrast to the Ungulates; and these two great orders afford an admirable illustration of the different way in which physical and organic changes may affect large and small herbivorous Mammalia; often leading to the extinction of the former, while favouring the comparative development of the latter.

The Sloths are a remarkable group of arboreal mammals, strictly confined to the great forests of the Neotropical region, from Guatemala to Brazil and Eastern Bolivia. None are found west of the Andes, nor do they appear to extend into Paraguay, or beyond the Tropic of Capricorn on the east coast. The genera as defined by Dr. Gray in 1871 are:—Cholœpus (2 sp.), "Sloths with two toes on fore limbs, sexes alike," Costa Rica to Brazil; Bradypus (2 sp.), "Sloths with three toes on fore limbs, sexes alike," Central Brazil, Amazon to Rio de Janeiro; Arctopithecus (8 sp.), "Sloths with three toes on fore limbs, males with a coloured patch on the back," Costa Rica to Brazil and Eastern Bolivia (Plate XIV., vol ii. p. 24).

Extinct Bradypodidæ.—In the caves of Brazil are found three extinct genera of Sloths—Cælodon, Sphenodon, and Ochotherium. More distantly allied, and probably forming distinct families, are Scelidotherium and Megatherium, from the caves of Brazil and the Pliocene deposits of La Plata and Patagonia.

The Manididæ, or scaly ant-eaters, are the only Edentate Mammalia found out of America, They are spread over the Ethiopian and Oriental regions; in the former from Sennaar to West Africa and the Cape; in the latter from the Himalayas to Ceylon, and Eastward to Borneo and Java, as well as to South China, as far as Amoy, Hainan, and Formosa. They have been sub-divided, according to differences in the scaly covering, into five groups, Manis, Phatagin, Smutsia, Pholidotus and Pangolin, the three former being confined to Africa, the last common to Africa and the East, while Pholidotus seems confined to Java. It is doubtful if these divisions are more than sub-genera, and as such they are treated here.

No extinct species referable to this family are yet known.

The Dasypodidæ, or armadillos, are a highly characteristic Neotropical family, ranging from the northern extremity of the region in south Texas, to 50° south latitude on the plains of Patagonia. The distribution of the genera is as follows:—Tatusia (5 sp.), has the range of the whole family from the lower Rio Grande of Texas to Patagonia; Prionodontes (1 sp.), the giant armadillo, Surinam to Paraguay; Dasypus (4 sp.), Brazil to Bolivia, Chili, and La Plata; Xenurus (3 sp.), Guiana to Paraguay; Tolypeutes (2 sp.), the three-banded armadillos, Bolivia and La Plata; Chlamydophorus (2 sp.), near Mendoza in La Plata, and Santa Cruz de la Sierra in Bolivia.

Extinct Armadillos.—Many species of Dasypus and Xenurus have been found in the caves of Brazil, together with many extinct genera—Hoplophorus, Euryodon, Heterodon, Pachytherium, and Chlamydotherium, the latter as large as a rhinoceros. Eutatus, allied to Tolypeutes, is from the Pliocene deposits of La Plata.

The Aard-vark, or Cape ant-eater (Orycteropus capensis) is a curious form of Edentate animal, with the general form of an ant-eater, but with the bristly skin and long obtuse snout of a pig. A second species inhabits the interior of North-East Africa and Senegal, that of the latter country perhaps forming a third species (Plate IV. vol. i. p. 261).

Extinct Orycteropodidæ.—The genus Macrotherium, remains of which occur in the Miocene deposits of France, Germany, and Greece, is allied to this group, though perhaps forming a separate family. The same may be said of the Ancylotherium, a huge animal found only in the Miocene deposits of Greece.

The true ant-eaters are strictly confined to the wooded portions of the Neotropical region, ranging from Honduras to Paraguay on the East side of the Andes. The three genera now generally admitted are: Myrmecophaga (1 sp.), the great ant-eater, Northern Brazil to Paraguay; Tamandua (2 sp.), 4-toed ant-eaters, Guatemala, Ecuador to Paraguay (Plate XIV. vol. ii. p. 24); Cyclothurus (2 sp.), 2-toed ant-eaters, Honduras and Costa Rica to Brazil.

Extinct Ant-eaters.—The only extinct form of this family seems to be the Glossotherium, found in the caves of Brazil, and the Tertiary deposits of Uruguay. It is said to be allied to Myrmecophaga and Manis.

These singular animals are almost confined to South America, where they constitute an important part of the fauna. In Africa, two family types are scantily represented, and one of these extends over all the Oriental region. In Pliocene and Post-Pliocene times the Edentata were wonderfully developed in South America, many of them being huge animals, rivalling in bulk, the rhinoceros and hippopotamus. As none of these forms resemble those of Africa, while the only European fossil Edentata are of African type, it seems probable that South Africa, like South America, was a centre of development for this group of mammalia; and it is in the highest degree probable that, should extensive fluviatile deposits of Pliocene or Miocene age be discovered in the former country, an extinct fauna, not less strange and grotesque than that of South America, will be brought to light. From the fact that so few remains of this order occur in Europe, and those of one family type, and in Miocene deposits only, it seems a fair conclusion, that this represents an incursion of an ancient Ethiopian form into Europe analogous to that which invaded North America from the south during the Post-Pliocene epoch. The extension of the Manididæ, or scaly ant-eaters, over tropical Asia may have occurred at the same, or a somewhat later epoch.

For a summary of the Numerous Edentata of North and South America which belong to extinct families, see vol. i. p. 147.