ite eae “Weale He aa EES 7 ie 4 Mr AL AAAS OSS ye Riki SESS RES ANN q . See ARERR Ree Teele: ~ SOSA NERS SNS es: byt SRSA dete. a ai EERE RUE ee ges ae APSA, . AS if ee Pee SERS HE a SSR alien ane SESE E RRS 7 g a Z y , i, 7 “ oy } i ts vo é a ys =~ SUNT ames Nene : RAR RE, z Sik SSS ie ; SENSES as, : » ASO tN e PAE | SSS ; i f SAAS p bles SHE BS ipa Fee) SS LAVA ae Pou: SEARS Ee i SAS NNER ‘ iB Hpi ne Ser SSSR oy inte SSSR Ga SSS SRA x A?) PNAS OORE at ORE OU ‘ Mae ue s SSR euales SSeS NE SEE : elias aN AS sae! “ eae Dae ia NARA SSSSERR year Aisin sf SARS RES Sn tn re ran wy SAE ; es ; XSAN RES : anes Mi RRR Ate: aah SANS SNES 1s ; . NS ios es poe : 8 aes ie tee y Z Be Fert en & 3 Leap ee TOME ; be ie son gree t By ip Meese? Z . “ ; tax i SSS os f Sl ig t Son : ( j a apapereret: add ae eine ite PB iices elaine ae aga Cornell Mniversity Library THE GIFT OF ae Pyeical (Gas A: 4.084 2s/io[9s Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://Awww.archive.org/details/cu31924031221850 A TEXT-BOOK OF ZOOGEOGRAPHY. ZDondon: C. J. CLAY anp SONS, CAMBRIDGE UNIVERSITY PRESS WAREHOUSE, AVE MARIA LANE. AND H. K. LEWIS, 136, GOWER STREET, W.C. a Waa ci ¥ Glasgow: 263, ARGYLE STREET. Deipsig: F. A. BROCKHAUS, Pew Work: MACMILLAN AND CO. ‘suotboy peotdeaboobooz, furmoys deyy oBpriqmeg Uos[lM FT Cambridge Patural Setence Manuals BIOLoGIcaL SERIES. GENERAL Eprror :—Artuur E. Saipiey, M.A. FELLOW AND TUTOR OF: CHRIST'S COLLEGE, CAMBRIDGE. A TEXT-BOOK OF —ZOOGEOGRAPHY BY FRANK E. BEDDARD, M.A. (Oxon.), F.BS. PROSECTOR OF THE ZOOLOGICAL SOCIETY OF LONDON, AND LECTURER ON BIOLOGY AT GUY’S HOSPITAL. CAMBRIDGE: AT THE UNIVERSITY PRESS. 1895 3 [All Rights reserved.} Cambridge: PRINTED BY J. & C. F. CLAY, AT THE UNIVERSITY PRESS. PREFACE. i ie this volume I have attempted to give the principal facts and conclusions of Zoogeography, without an undue profusion of detail. A few examples only have been selected to illustrate the principles. As far as possible I have endeavoured to use instances that have not been made use of by Mr Wallace in his two works upon this subject. I have taken Kerguelen Island and Fernando Noronha as examples of oceanic islands; the facts of distribution and the inferences to be drawn are illustrated largely by the earthworms; and in other places I have availed myself of some of the more recent sources of information. Naturally a large portion of this book, as must be the case with any book upon Geographical Distribution, is based upon the two indispensable treatises of Mr Wallace. I have also found great assistance in Prof. Heilprin’s work upon Distribution in the Inter- national Scientific Series, and in an excellent manual vi PREFACE, upon the same subject by M. Trouessart; I have checked and added to the facts that I had accumulated by a careful comparison of them with the article Geographical Dis- tribution in Prof. Newton’s Dictionary of Birds. Numerous special memoirs have also been laid under contribution, a reference to which will be found in the proper place. FRANK E. BEDDARD. Lonpon. é November, 1894. CONTENTS. CHAPTER I. THE GENERAL FACTS OF THE DISTRIBUTION OF ANIMALS. Introductory, p. 1. Locality and Station, 5. Cosmopolitan groups, 8. Restricted groups, 10. The meaning of a restricted distribution, 12. Discontinuous distribution, 15. Separate areas of the species of a genus, 18. Distribution of Rhea, 19. Distribution of Ibexes, 20. Distribution of the Cassowaries, 23. Classification and Distribution, 25. Distribution of the Gallinaceous birds, 26. Distribution of the Edentata, 31. Distribution of the Cuckoos, 35. Distribution of Chelonia, 38. Distribution of Lizards, 40. Distribu- tion of Crocodiles, 43. Distribution of Snakes, 46. Distribution of Batrachia, 47. Distribution of Scorpions, 49. Distribution of Land Planarians, 53. Distribution of Earthworms, 57. CHAPTER I. ZOOLOGICAL GEOGRAPHY, Mr Sclater’s regions, 72. Mr Huxley's regions, 75. Other sug- gested regions, 76. Mr Sclater’s regions the most convenient, 78. The six Zoological regions of Mr Sclater, 88. The Palzxarctic region, 88. The Nearctic region, 93. The Ethiopian region, 98. The Oriental region, 102. The Neotropical region, 107. The Australian region, 118. Some graphic methods of presenting the facts of Distribution, 118. CHAPTER III. THE CAUSES WHICH INFLUENCE THE DISTRIBUTION OF ANIMALS. Distribution not dependent upon temperature, 124, Distribu- tion of Crustacean Arcturus as illustrative of connection between range and temperature, 125. The country inhabited by an animal Vill CONTENTS, is not necessarily the only one in which it can flourish, 128. Similarities in the faunas of distant countries, 129. Problems of Distribution and Evolution, 131. Means of Dispersal of Animals, 134. The influence of geological terrain upon faunas, 136. Dispersal of Oligocheta, 138. Dispersal of Mollusca, 140. Dispersal of Amphibia, 145. Dispersal of Reptiles, 147. Evidence of capacity for Migration on the part of a given animal, 150. Influence of human interference upon Migration, 151. The existing Distribution of land and sea considered in relation to Zoological Geography, 155. Evidence in favour of Permanenee of Oceans, 156._ Evidence against the view that existing Oceans have not largely changed their areas, 159. Evidence in favour of a formerly more extensive Antarctic Continent, 164. ‘“ Lemuria,” 176. CHAPTER IV. ‘THE FAUNA OF ISLANDS. The Fauna of the British Isles, 183. The Fauna of Madagascar, 187. The Fauna of Fernando Noronha, 190. The Fauna of Kerguelen, 193. The Fauna of the Galapagos, 195. The Fauna of New Zealand, 199. Fauna of the Sandwich Islands, 203. General observations upon the Fauna of Islands, 204. Continental Islands, 205. Oceanic Islands, 207. Anomalous islands, 210. Some pecu- liarities of island animals, 212. CHAPTER V. SOME THEORETICAL CONSIDERATIONS. The bearing of the facts of distribution upon the places of origin of different groups, 220. The place of origin of the Marsupials, 222. Theory of the Polar origin of Life, 227. Map shewing Zoogeographical Regions 3 , Frontispiece ay distribution of Edentata . 3 . to facep. 32 5 95 ay Earthworms . ; » 9» 65 Struthious Birds . ,, ,, 168 bie ” ” Lemurs . . si an. BT ” ” ” CHAPTER I. THE GENERAL FACTS OF THE DISTRIBUTION OF ANIMALS. Introductory. THE entire world both land and sea supports every- where animal life. The extreme cold of the arctic regions is not too intense to permit of the existence of at least a few forms of life, while the warmer regions of the globe have everywhere an abundant fauna, which increases towards the tropics; it is even probable that the icebound antarctic continent, if it could be explored, would be found to possess some inhabitants. . The most elementary knowledge of Zoology is sufficient to enable us to see that, while terrestrial life is as abundant as marine, the kinds of animals inhabiting these diverse situations are different. We can distinguish in fact between purely terrestrial and purely aquatic animals. The latter group can be again divided into two great classes, those which inhabit the sea and those which inhabit the fresh waters of the land. There is however no absolute break between these various groups. The B, Z. 1 2 TERRESTRIAL AND AQUATIC ANIMALS. [cH. I fresh waters themselves gradually pass into salt at the mouths of rivers; while a marsh, particularly one that is subject to periodical drying up, is to some extent inter- mediate between the land and the water. Corresponding to this absence of hard and fast lines in inorganic nature we find a similar absence of definiteness in the animals which frequent the waters, the dry land, and the transitional areas. The duck tribe are equally at home when swimming in a lake and when flying from one pool to another. There are even ducks which perch on trees. Closely allied Crustacea may be marine, estuarine, fresh water or terrestrial in habit. Certain sharks, nor- mally marine, ascend rivers for some distance, and the Manatee of the west coast of Africa and the east coast of America is quite as much at home when browsing upon the marine algae of the coasts of those continents as when living in their rivers. Though we may broadly separate animals into terrestrial and aquatic, there is no large group of animals which is exclusively terrestrial. Even insects which approach nearest to this condition have plenty of aquatic representatives; there is even a peculiar genus of bugs, Halobates, which inhabits the open sea far from land. On the other hand, there are groups which are purely aquatic, and even some which are only marine, being never found in fresh water; but these are few. Among Verte- brates fishes are the only group which can be said to be almost absolutely aquatic; and here too there are some slight exceptions. The well-known Climbing Perch, Anabas scandens, can with impunity leave the streams which it CH. 1] HABITATS OF ANIMALS. 3 inhabits; the mud fish of Africa and America has lungs as well as gills, and can suffer with equanimity the drying up of the rivers in which it lives. Eels are said to move from one pool to another across the intervening grass. The curious little fish Periophthalmus voluntarily leaves the sea and hops along its margin on the look-out for the Mollusc Onchidiwm upon which it largely feeds. Among Invertebrates there are more purely aquatic groups; and these are exclusively marine. No Ascidian, no starfish, brittle star, sea urchin, or sea lily has as yet been met with except in the sea. Sponges and Celenterates are without exception aquatic and for the most part marine, their delicate organisation not being able to withstand continued dryness. But these are positively the only large groups which are purely aquatic. Many others are prin- cipally aquatic, such as the Flat worms, with terrestrial allies in the land Planarians, the Crustacea, the Annelids and some others. Not only can we draw these broad distinctions between the habitats of different animals, finding one to be terres- trial, another aquatic and a third amphibious; we can also assign to each a definite place upon the land or in the waters. The Indian Ocean is frequented by creatures which are unknown in the colder waters of the North Sea; the Mississippi has alligators which the Thames, for ex- ample, has not. It is a matter of common knowledge that the tiger is restricted to Asia and the puma to South and certain parts of North America; the elephant is unknown beyond the old world, and it has even there a limited area of range. It would be as surprising to meet with an 1—2 4 DISTRIBUTION OF LAND ANIMALS, [CH. I elephant even in the most secluded of Brazilian forests as to meet with a tiger genuinely at home in the neighbour- hood of a peaceful English village. That branch of Biology which is termed Geographical Distribution, or, when applied to either animals or plants only, Zoogeography and Phytogeography, is concerned with facts such as the above; it has also to do with the solution or attempted solution of the various problems to which these facts give rise. The science is not limited to a consideration of the animals which inhabit dry land. But this volume will only deal with those forms, touching incidentally upon some of the fresh water species, whose distribution is apparently governed by the same laws as those which govern the distribution of the purely terres- trial animals, We shall now commence our survey of the chief facts in the distribution of land animals. It has been said that every animal has its place in the world, which may be wider or narrower. The country which is inhabited by a given animal is called its area of distribution, its habitat or locality. The converse follows that every tract of the earth’s surface is inhabited each by its peculiar set of animals. But they are not met with everywhere in that area. Almost every year a new volume is published giving a list of the birds of a particular county or it may be district; anyone who will take the trouble to inspect and compare a series of these volumes dealing with many different counties will be struck by the fact that a given species of bird may be absent from one list or stated to be uncommon, while numerous records of its discovery will be CH. I] LOCALITY AND STATION. 5 found in another. The same thing holds true of other groups; the badger, for example, will be found in one wood but will be absent from another; nevertheless generally speaking this animal may be said to inhabit the greater part of England, not to mention foreign regions in which it is also found. Were we to put together all that has been recorded of the range of this or any other animal and colour a map of England in correspondence with those facts, we should find that a large map would be coloured by a series of closely set but separate patches of colour ; on the other hand a small map would practically have to be coloured all over to indicate the range of the animal. The reason for this is that the badger can only live in certain kinds of country. It is not at home for instance in bare chalky downs or on the tops of high mountains; it prefers woods and the immediate neighbourhood of woods. Locality and Station. We must carefully distinguish between locality and station. While the area inhabited by a species is usually continuous, it by no means always happens that the station consists of one continuous tract. Animals inhabiting forests or moorland or pools are only found where the suitable circumstances occur. But in such instances the habitat of the animal may be wide, only broken up into a series of stations. We should not regard a case of this kind as one of discontinuous distribution. Often however the local phenomena of distribution have not so clear an interpretation. Every entomologist is aware of the often 6 RANGE OF LEPIDOPTERA. (cH. 1 capricious occurrence of butterflies and moths; we do not now refer to marsh-loving or wood-loving species or to any intelligible restriction of this kind. Out of a dozen fields, to the eye equally well suited for the maintenance of a given species of moth, in which for instance the food plant is equally abundant, two or three or it may be only one will be inhabited by the insect in question, which will there perhaps swarm. Some years since a small moth was found at Folkestone only in a particular tract of grass with no obvious advantages—indeed the reverse, as it was scanty and trodden under the feet of passers by—to the exclusion of neighbouring grassy areas; this insect, known as Tapinostola bondii, was abundant in this particular locality but nowhere else in the neighbourhood. In every local faunal list such and such a wood or field is given as the locality for a particular insect, which would be equally at home in other woods and fields, but is not as a matter of fact found in them. To get the particular insect we have perhaps to journey to another county or even to another part of England; possibly a particular wood is the only part of England where it is to be met with and the next locality will be on the Continent. No doubt some of these apparently capricious variations are to be explained by advancing civilisation. Building, draining and the gradual reclamation of the country are fatal to insect life as a rule. But this will not explain every case of capricious restriction to a few separated stations, such as are so commonly met with among the Lepidoptera of this country. CH. I] WIDE RANGE OF CERTAIN FORMS. 7 The Range of Animals. In the range of species we meet with every condition, from world-wide distribution to the most restricted habitat. It has been said that man is the only animal (with the exception perhaps of his parasites) which is literally found in every habitable part of the earth’s surface; but a few others are almost as widely scattered. Among mammals only certain bats are in this position, for though the common mouse and the rat are found nearly everywhere, it is very possible that man is responsible for this wide dissemination. The Barn Owl (Strix flammea) occurs in most parts both of the new and old worlds. It is true that in different countries it has received different names ; but the opinion of many is that these are at the most local races which are hardly deserving of being separated as species. The Painted Lady (Vanessa cardut) is an example from another class of animals which has an equally wide range. This butterfly extends from Europe to the Islands of the Pacific and to New Zealand, but is not found in the West Indies and certain parts of South America. The common Red river Worm (Tubifex) seems to be universally spread. At any rate examples from so distant a spot as New Zealand do not differ in any appreciable point from those of England. A species be- longing to another family of Oligocheta, Henlea ventri- culosa, occurs in Europe, in the territory of the Khirghese Tartars, and in New Zealand. The now extinct Large Copper Butterfly was formerly found in abundance in the Cambridgeshire fens, but found 8 ANIMALS WITH RESTRICTED RANGE. [cH. I nowhere else in the world. The limitation of particular species of humming-birds to particular peaks of the Andes, and of snails to particular valleys, are other examples. Francolinus Kirki of the island of Zanzibar is almost as striking an instance of the same phenomenon. The Ven- dace (Coregonus vandesius) restricted to Loch Maben in Dumfriesshire is an animal with a still more limited range, and several other examples might be given, particularly from the faunas of islands. The two extremes are natu- rally connected by numerous intermediate stages, where an animal has a more or less limited habitat. Cosmopolitan groups. There are comparatively\ few groups of terrestrial animals which are truly cosmopolitan, As the groups get smaller the number of those that are cosmopolitan decreases. That is to say, there are more cosmopolitan families than genera, and more genera that are cosmo- politan than species. Mr Wallace enumerates in his work upon Geographical . Distribution no less than sixteen families of birds which are really cosmopolites; they are, Corvide (Crows) and Hirundinide (Swallows) among Passerines; Kingfishers among Picarie; among other land birds, Columbide (Doves), Tetraomde (Grouse), Falconide (Hawks), Stri- gide (Owls); among waders Rallide (Rails), Scolopacidee (Snipe), Charadrie (Plovers), Ardeidw (Herons); among aquatic birds Anatide (Ducks and Geese), Larid@ (Gulls), Procellaride (Petrels), Pelecanide (Pelicans) and Podici- CH. I] COSMOPOLITAN GENERA. 9 pedidee (Grebes). Only one family of mammals are really cosmopolitan, namely, the Vespertilionidw. There are several families of butterflies and moths which are in this position and nearly all the families of beetles. The Helicide alone of the terrestrial Mollusca are universally distributed, and among earthworms the two families Lwm- bricide and Cryptodrilide. It is a significant fact that of these families by far the majority are winged creatures, which are naturally less restricted by barriers. The earth- worms form an apparent exception which is not really so; for it is highly probable that the universal range of these families is due to human agency. Again, it is important to notice that of the cosmopolitan birds at least eight are largely aquatic in habit or frequent the margins of streams and lakes. The conditions of life for aquatic birds are more similar in different parts of the globe than those of purely land birds. Moreover all these families are numerous in genera and species. Cosmopolitan genera are fewer in proportion. Among birds we have Hirundo (Swallow), Pandion (Osprey), Striz {Barn Owl), Rallus, Porzana (Rails), Gallinula, Fulica (Coots), Numenius (Stilt Plovers), Limosa (Godwits) and several other Scolopacide, Charadrius (Plovers), Ardea (Heron), Nycticorax (Night-heron), Anas (Duck), Sterco- rarius (Skua), Larus (Gull), Sterna (Tern), Puffinus (Puffin), Procellaria, Fulmarus (Petrels), Pelecanus (Pelican), Pha- lacrocorax (Cormorant), Podicipes (Grebe). Among butter- flies the genera Pyrameis (Painted Lady), Polyommatus (Blues), Pieris (Whites), Papilio (Swallowtail), Pamphila and Hesperia (Skippers) are universally distributed. 10 COSMOPOLITAN SPECIES. (cH. I Macroglossa, Cherocampa (Hawkmoths), Macrosila (Clear wing) among moths are also cosmopolitan. As to species there are comparatively few that have a world-wide range; the most striking example among birds is the Osprey. Gallinula chloropus and Totanus incanus have the same distribution. Among insects it is more difficult to discount the interference of man, which may be responsible for a world-wide distribution. Anosia plexippus is a butterfly which appears to be found almost everywhere. I have dealt on another page’ with those species of earthworms which are cosmopolitan or nearly so and arrived at the conclusion that it is in those cases really a question of transport by man. The same argu- ment seems according to M. Trouessart to apply to the cosmopolitan Gecko, Platydactylus facetanus. Even the Lepidoptera above mentioned are not above suspicion on this score; the pup could so easily be accidentally conveyed to the most distant countries. The common bee-fly (Eristalis tenax), which so closely resembles to the superficial inspection a honey-bee, has reached as far as New Zealand, doubtless by the same agency. Restricted groups. We have besides Wniversally distributed groups some of which are remarkably limited in their range, Naturally this applies most of all to species and least to families, The majority of species have a more or less restricted range, so much so that it would be impossible to give any 1 y, infra. CH. q] GENERA AND SPECIES OF LIMITED RANGE. 11 catalogue of them. A few of the more remarkable ex- amples may however be referred to. The most striking of these are certain fishes which are limited to a single lake; thus the Lough Killin Charr is confined to the lake of that name. A small moth, Orniz devoniella, was once only found in Devonshire. The monkeys of the genus Brachy- urus, comprising three species, are limited each to a moderately small forest tract in South America. The majority of examples of species with a limited range are of course to be found upon oceanic islands. Not only are there species with a very small range in space but also genera. Here again the most numerous examples are to be found on oceanic islands such as the now extinct Starling, Fregilupus, of Reunion. But there are other cases of genera which have facilities for a wider range, but which for one reason or another (to be considered later) have been unable to extend that range. The genus Opisthocomus (containing by the way only a single species, being the type of a distinct family) is limited to a portion of British Guiana; the Gorilla to the forest tract of the west coast of Africa; here again there is but one species. Among families there are also a few with an exceedingly restricted distribution. The Rhinochetide, containing only one genus and species, R. jubatus, the Kagu, is only found in the island of New Caledonia; the Trumpeters or Psophiide are confined to certain districts of the Amazons, and the four or five species of the genus might be quoted as a case of very limited range in species. In Madagascar the family Chiromyide is represented by but one species. Cases of limited range among groups that may be regarded 12 DISTRIBUTION OF MONOTREMATA. (cH. 1 as higher than families are naturally not common. The most striking perhaps is that of the Rhynchocephalia; this group has but one species, the well-known Hatteria, which is only found in a few small islands off the coast of New Zealand. Not quite so striking from some points of view, but perhaps more so in another, in that the group is a higher one, is the case of the Monotremata; the two or three genera are limited to certain parts of the Australian region. The facts just enumerated lead us to one of the axioms of the science of Zoogeography, which was for- mulated by Mr Sclater? in the following words. “Every species occupies a definite area on the world’s surface; | and: in like manner every genus and family, or other higher assemblage of species, occupies a definite area on the earth’s surface ; or more shortly, locality or existence in a certain spot is quite as much an attribute of animals as — structure or the possession of a certain form or shape.” The meaning of a restricted distribution. It does not, however, by any means follow that this area is now as it always has been. To study Zoogeography properly a knowledge of the extinct forms of life is not only desirable but necessary. By the aid of Paleontology various facts, at first sight dark and meaningless, become clear, or at least clearer. We must imagine each species setting out from its centre of origin and gradually ex- 1 «The Geographical Distribution of Mammals,” Manchester Science Lectures, 1875. CH. 1] AREA INHABITED BY CHIMPANZEE, 18 tending itself by actual or passive migration right and left and in every possible direction from this focus. Gradually the form will be modified, or by competition or for some other reason become extinct, leaving perhaps descendants scattered here and there to tell the tale of a formerly wide range. A guess can be made as to the comparative age of a species or a genus by comparing such facts. In all probability these instances of a restricted distri- bution are to be explained in one of two ways; either the form is a new one or it is an ancient one. A new species recently come into existence would naturally, at least on any theory of evolution, have a limited range because it would have come into being at one locality and not have had time to extend its range, supposing an extension to be possible and not barred by impassable barriers. The former alternative applies probably to most of the examples that have been used, particularly perhaps to the peculiar species often found upon oceanic islands, There are however numerous species, as limited in their range, which are in some cases certainly vestiges of races once universally or widely distributed. The Chimpanzee tribe is at present limited to the forest region of central Africa. Its utmost range is nearly across that continent. But the palaontological records of — India contain a description and figures of a portion of a skull evidently belonging to a chimpanzee which at one time existed in India. Probably that indicated the high- water mark of the extension of the chimpanzee, which has since retired to more restricted boundaries. We know from historical records that the lion used to occur in 14 SIGNIFICANCE OF LIMITED AREAS. [cH. I Greece as well as in India and Africa, where it is now alone met with. And at an earlier period still it was found in this country. The hippopotamus, now limited to Africa, was once found in Madagascar, and probably the same species in Europe. In such cases however the species in question are generally also of generic, even of family, rank. The remarkable lizard Hatteria, as already mentioned, is con- fined to one or two islands off New Zealand; this lizard, as the fossil remains of its allies tell us, is the sole survivor of the Rhynchocephalia, a race of Saurians found in the Mesozoic rocks of this country. The Viverrine Carnivore Cryptoprocta, found at present only in Madagascar, is held by some to be the last surviving remnant of the extinct Creodonta; in any case it is generically distinct from any other carnivorous animal. The same arguments may be applied to the Thylacine of Tasmania, to the Aye-Aye (Chiromys) of Madagascar, and to many other animals. An apparently similar series of facts is therefore probably to be explained in quite a different manner, an instance of extremes meeting. The same mode of distribution is indicative either of great antiquity or of extreme modernity. A comparatively restricted range however may be also due to incapacity for migration. The converse is perhaps more obvious. Widely distributed animals are either flying animals independent of barriers which impede the purely terrestrial species, or possess some special facilities for voluntary or involuntary translation from country to country. An entirely arboreal creature cannot pass across CH. I} DISCONTINUOUS DISTRIBUTION. 15 a level tract of country with no trees; nor can an Amphibian whose skin requires to be kept moist cross an arid desert. This consideration leads to an important matter, the capacities for migration possessed by different animals which will be discussed later. Discontinuous distribution. On the hypothesis that each animal has had its centre of dispersal, that it came into existence once and at a definite place, it is clear that originally at least the area inhabited by a given species must have been perfectly continuous. As a matter of fact it is generally the case ; the remarkable thing appears to be not that there are occasionally breaks in the continuity of the area inhabited by a certain species but that it is so difficult to find instances to illustrate the breaks. Mr Wallace explains the rarity of discontinuous distribution among the species of birds by the suggestion that they are possibly “more rapidly influenced by changed conditions, so that when a species is divided the two portions almost always become modified into varieties or distinct species.” It must be borne in mind also that birds are a modern group, and the very difficulty of classifying them satisfactorily indicates that there are but few breaks in the series; they are possibly still in a condition of perpetual modification ; they have not so to speak become fixed and crystallised, like some of the older and in a sense more effete groups of animals. However this may be, Mr Wallace quotes from Mr Seebohm a highly remarkable instance of discontinuous 16 RANGE OF GENUS PERIPATUS. {cH. I distribution in-a species of Tit. The Marsh tit, Parus palustris, has a range nearly co-extensive with the Palearctic region; but it is known throughout this immense tract of country in certainly three varieties. One of these is found in Southern Europe, in Italy, Turkey, Greece, and Asia Minor. The same variety does not crop up in the intervening country but again appears in South China. Whether this variety is entitled to specific rank or not, the fact is remarkable and really of equal value. It almost suggests an explanation that has been sometimes advanced to account for discontinuous distribution in the species of Mollusca. No great diffi- culty could be felt in the assumption that the same variety had been twice produced in localities of a somewhat similar climate. Mr Wallace refers to one or two other examples of a somewhat similar nature; the Reed bunting of this country reappears in Japan, being absent from Asia. To take an instance from another class, Dr Scudder records the existence of the butterfly @neis jutta’ in the Rocky Mountains of British Columbia and in Hudson’s Bay, and its absence from the intervening tract. The genus Peripatus? offers an example of a genus with an exceedingly wide and at the same time discon- tinuous distribution. Peripatus is universally regarded as a very archaic form of Arthropod, which has preserved certain characters of the worm-like ancestor from which it is presumed that the Arthropods have been derived. There are, for example, a series of paired excretory organs 1 Butterflies of HE. United States, 2 Quart. Journ, Mier. Sci, vol. 28, 1888. CH. I] RANGE OF PERIPATUS. 17 like those that occur among the segmented worms. On the other hand it has the trachee of the tracheate division of the Arthropods, and at least rudimentary appendages of the Arthropod type. The genus has been recently the subject of a careful monograph, so that we are in possession of the facts of structure and distribution of a good many species. Mr Sedgwick allows eleven species as well founded, and there may be others. The genus occurs in South Africa, South America and the West Indies, in New Zealand and in Australia; one specimen is found in Sumatra; otherwise it is absent from the Oriental region. The species of the genus are mainly distinguished from each other by the position of the generative aperture, by the number and structure of the legs and by colour. Mr Sedgwick divides them into four groups which correspond to their range; the Australasian species, for instance, form one group, the neotropical another, and so forth. These four main groups are largely separable on account of the varying position of the generative pore, which may be between the last or the penultimate pair of limbs or altogether at the end of the body. There are differences too in the structure of the generative organs, and the eggs show characteristic variations; thus, in the Australasian species the ova are large and full of yolk, in the Cape species though the ova are large the yolk is not abundant, and finally in the neotropical species of Peripatus the ova are quite minute, and without food yolk. The single known species from the Oriental region is more imperfectly known than many of the others; but it seems to resemble the neotropical B. Z. 2 18 DISTRIBUTION OF JAYS. [CH. I rather than the Ethiopian species, which is not a little remarkable. It has, as have the neotropical forms, the generative openings between the legs of the penultimate pair. Separate areas of the species of a genus. It sometimes happens that the area of distribution of a genus is perfectly continuous but traversed by large rivers or other checks to distribution. A genus occupying a group of islands, for example, may be said to have an unbroken range so far as is possible; but here under similar circumstances it is frequently the case that the isolation has been accompanied by the breaking up of the genus into a number of species, perhaps corresponding with the subdivisions of the area. The islands of the group which together constitute the Sandwich Islands are often inhabited by particular species belonging to a genus common to the whole archipelago; the huge tortoises of the Galapagos are in the same condition. A large tract of country is often similarly inhabited by a series of species belonging to the same genus; but each of these keeps rigidly to its own particular territory: an inter- mingling is rendered difficult perhaps by the infertility of the species with each other, and partly also by the fact that, the ground being already taken up, there is no room for the inroad of a closely allied form, which has presumably the same or nearly the same mode of life and would therefore seriously compete. The twelve species of Jays belonging to the genus Garrulus range over the greater part of the Palearctic region; but nearly every species CH. I] DISTRIBUTION OF RHEA. 19 has its own particular habitat, and does not interfere with its neighbours. In the map which Mr Wallace gives in illustration of the facts of distribution of this genus it is seen that two species overlap just at the confines of Europe and Asia, while the former of these, the European Garrulus glandarius, is also overlapped by one species in the south-east of Europe and by another in Algeria. It is far more usual for species to occupy in common a given area, than for a division of the territory to have taken place. Nevertheless the example just quoted is by no means unique, even among birds whose powers of flight set ordinary barriers at defiance. But a rigid partition of the area of a genus is more commonly met with among animals which have not these exceptional means of disposal. This will now be illustrated by three examples. Distribution of Rhea. The distribution of the species of Rhea! illustrates the limitation of the species of the same genus each to its own particular tract. The genus itself occupies a con- siderable area of S. America, to which continent it is absolutely confined. The three species of the genus have been lately subjected to a careful comparison by Dr Gadow, who has plainly differentiated the three recognised species, viz. Rh. americana, Rh. macrorhyncha and Rh. darwint. The anatomical characters which distinguish them are not of course very marked, but they are amply 1 See Gadow, ‘‘On the Anatomical Differences in the three species of Rhea,” P.Z.S,. 1885, p. 308. 2—2 20 THE SPECIES OF RHEA. (CH. I sufficient for the purpose. Rh. darwini and Rh. americana agree to differ from Rh. macrorhyncha in having only 15 (instead of 16) cervical vertebra and in having a broad skull. They differ from each other principally in the fact that the metatarsus in front has scutes on distal half only in Rh. darwini; Bh. macrorhyncha agreeing with Rh. americana in having transverse scutes along the whole length of that part of the leg in front. The range of the three is as follows :—Rh. americana extends from Bolivia through Paraguay into Uruguay and southward to the Rio Negro. Its head-quarters seem to be the pampas of Argentina. Rh. darwini is restricted to the eastern half of Patagonia and to south-eastern Argentina. It overlaps the last species about the Rio Negro. Finally Rh. macrorhyncha occurs in the provinces of Pernambuco and Bahia but does not overlap Rh. americana. Dr Gadow intimates that on the whole Rh. darwin is the best marked species. Hence possibly its overlapping is less remarkable, since the greater difference in organisation may go with a preater difference in habits. Distribution of Ibexes. There are altogether eleven species of Wild Goats as allowed by Mr Sclater’, to which may be added a twelfth®. (1) Capra pyrenaica, the Spanish Ibex, is not only Pyrenean, but is found, slightly altered in character, in the mountain ranges of other parts of Spain and Portugal. 1 P.Z.S. 1886, p. 314. 2 There are probably however more names than species. CH. I] RANGE OF WILD GOATS. 21 “It is curious,” remarks Mr Sclater, “that it is more nearly allied to the Caucasian ibex than to the ibex of the Alps.” (2) C. ibex, the ibex or steinbok found in the Alps and Tyrol, but rare and needing artificial preservation. (3) C. egagrus, the true wild goat, is probably the origin of the domestic variety. It is now found only in Crete and some of the smaller Cyclades as regards Europe, but also extends through Asia Minor and Persia to Sind and Baluchistan. (4), (5) C. caucasica and C. pallasit are restricted to the Caucasus, where they do not appear to overlap greatly. (6) C. sinattica. This ibex is found only in the mountain ranges of Upper Egypt, the Sinaitic peninsula and Palestine. (7) C. walie is a distinct though rare and little known species, from the highest mountain ranges of Abyssinia. (8) C. stbirica. It is remarkable that this species should occur in two such distant localities as the Altai mountains and the Himalayas. But it appears that a thoroughly careful comparison of examples from the two localities has not yet been made. This may very possibly reveal differences. (9) ©. falconerr. This ibex is popularly known as the Markhore, it lives in the Pir-panjab and Sulaiman ranges in Cashmere and Afghanistan. (10) C. jemlanica. The “Tahr” occurs along the whole range of the Himalayas. 22 WIDE DISTRIBUTION OF BEARS. [cH. I (11) C. hylocrius. This species of wild goat. is found in the Neilgherries and some other ranges of southern India. (12) C. severtzowt. This is another Caucasian ibex which Dr Menzbier has added to the two already referred to. Here we have an instance of a genus of tolerably wide distribution, but discontinuous. The discontinuity is entirely due to the mode of life of the genus, which frequents high mountains and cannot tolerate the level plain. The mountains are comparable to oceanic islands which can only be reached from time to time and under favourable circumstances. This naturally results in the isolation of those individuals which have migrated from their original home to a neighbouring mountain range ; and as a consequence of this isolation, which precludes admixture with the parent stock, we have the production of new forms, just as in the case of oceanic islands. With this may. be compared the distribution of such a genus a8 the giraffe, which has, or had until very recently, a range of nearly the same extent as measured by miles. But this range is uninterrupted by many tracts of country that are uninhabitable to the animal; hence there are at most two forms of giraffe. Even the true bears (genus Ursus) may be contrasted. Dr Grevé allows nine species and five varieties, hardly more than there are of goats. But the range is enormously larger—nearly the whole of Asia and Europe and nearly the whole of America— and there is nowhere a gap of any kind, CH. 1] SPECIES OF CASSOWARY. 23 Distribution of the Cassowaries. The species of the genus Casuarius present an ex- cellent instance of the specialisation of a genus when its region is broken up by barriers into detached areas. There appear to be altogether eleven or twelve species of cassowary that are well, ascertained; there may even be one or two more; at any rate there are more than twelve names distributed among the cassowaries. I shall not enter into the characters which distinguish the species beyond remarking that they can be readily defined by the shape of the “casque,” by the presence or absence of wattles depending from the patch of naked skin upon the throat and by the particular tints exhibited by the generally brilliant colouration of the latter. The casso- waries are entirely limited in their distribution to the Australian region and do not range over the whole of that region. They are absent from New Zealand and from many other outlying islands. But although the cas- sowaries are of bulky form and like other Struthious birds, quite incapable of flight, they are by no means limited to the continent of Australia itself. The following is a list of the properly defined species and their range: Casuarius australis, Australia. C. picticollis, C. Edwardsi, | C. Westermanni, New Guinea. C. wuappendiculatus, C. Salvadoria, C. galeatus, Ceram. 24 CASSOWARIES OF NEW GUINEA. [CH. I C. Beccarit, Wokan. C. bicarunculatus, Aru. C. Bennetti, New Britain. . aa Tobi C. occypitalis, This case is analogous in many ways to that of the goats already dealt with. Isolation has led to the differentiation of species from a presumably identical stock. Furthermore, where the area is large it has proved capable of sustaining several species, which is not the case with those islands of limited extent, such as Ceram, which harbour cassowaries. With this fact may be compared the presence of only a single species of Ibex in the com- paratively small tract of country occupied by the Pyrenees, and the presence of this species in the more ' extensive Caucasus. The existence of five out of the ten species in New Guinea marks this large island out as the head-quarters of the group from whence they have migrated elsewhere, or perhaps, if the islands are to be regarded as a broken continent, have been isolated. That New Guinea is to be regarded as the original home of the cassowaries is perhaps also shown by the fact that the species now existing there present in themselves most of the important modifications of structure which the genus exhibits. In his most recent revision of the cassowaries Mr Sclater divides those without watiles from those which have these appendages. Both kinds occur in New Guinea. On the other hand this con- 1 The editor of the Ibis (Oct. 1894, p. 560) is inclined to doubt the distinctness of these two species. CH. 1] DISTRIBUTION AND STRUCTURE. 25. sideration is to be qualified by the fact that no casso- wary with a laterally compressed casque (another character made use of by Mr Sclater) exists in New Guinea. In all the New Guinea species the casque is transversely compressed. Classification and Distribution. The facts of distribution are constantly liable to be misunderstood through ignorance of classification. Not only is a serious error in the actual facts of the distribution of a particular group caused by wrongly assigning to it some individual genus or species, but the significance of the facts is by this largely, sometimes totally, obscured. A knowledge of comparative anatomy is absolutely essential to the student of distribution. It used to be supposed that the central American Carnivore Bassaris was a member of the family Viverride; the genus therefore was believed to be the only Viverrine found in the New World, a singular anomaly in the distribution of the group. But Sir William Flower in his paper upon the skull in the Carnivora showed that this animal is really an ally of the Raccoons, which are purely an American family. Everybody is acquainted with the fact that monkeys are found both in the old world and in the new. But the fact gets a far larger significance when it is realised that the new world monkeys form a group by themselves which differs from that of the old world monkeys in a number of important anatomical characters. The wide distance and the absence of means of transit 26 WIDE RANGE OF ANCIENT FORMS. [cH. I within recent times has brought about the great diver- gence which is now seen between the two sections of the Primates, the new world Platyrrhines and the old world Catarrhines. It has been already pointed out in dealing with the distribution of the archaic Peripatus that the species of different parts of the world form natural assemblages separable from those of other parts of the world by definite anatomical characters. This is the case too with many other genera and families of animals. We invariably find ‘that when a group, which Palzontology— or in the absence of direct evidence from fossils, other considerations derived from anatomy or embryology— proves to be an ancient group, has a tolerably wide and discontinuous distribution, marked differences in structure distinguish its representatives in different parts of the world. This is more marked still in the case of a group which has but limited powers of dispersal. We shall now illustrate the connection between distribution and ana- tomical structure by a few examples, which are of course few among many; in other pages other instances have been or will be treated of and reference may be made to those places for further illustration of the general fact. Distribution of the Gallinaceous birds. The Gallinaceous birds (Alectoromorphe of Huxley) offer an exceedingly instructive example of the connection between anatomical structure and geographical distri- bution. There can be no doubt that this group is a natural one. It is divisible into the following seven cH. 1] THE GALLINACEOUS BIRDS. 27 families!:—Grouse, Turkeys, Guinea-fowls, Pheasants, Megapodes, Curassows and the aberrant Hoatzin (Opis- thocomus). They are thus distributed: 1. Tetraonide (Grouse). Palearctic and Nearctic. Phasianide (Pheasants). Oriental. Numidide (Guinea-fowls). Ethiopian. Meleagridaw (Turkeys). Southern Nearctic. Cracidee (Curassows). Neotropical. Megapodidee (Mound-builders). Australian. 7. Opisthocomide (Hoatzin). Neotropical’. They are not, however, in every case absolutely confined to these regions as defined by Mr Sclater. Thus among the Megapodes one species gets into the Indian region, and the Phasianide stray into the Palearctic. The Tetraonide are really almost cosmopolitan, though mainly massed in the two northern regions of the earth’s surface. The Curassows extend into the southern parts of the Nearctic, occurring as they do in Mexico and in California. Such briefly are the facts of the distribution of this group of birds; it now remains to enquire into the mutual relationships of the several families or subfamilies. Mr Huxley unites the Megapodes with the Cracide into a group Peristeropodes, and separates them from all the rest which constitute his Alectoropodes. The former division has a sternum with less deeply marked notches, the vomer is well-developed, and the hallux is attached to the foot on a level with the other toes. These characters look as if they were more primitive than the deeply O om wo Lo 1 P,Z.S. 1868. 2 These terms are explained later; see Chap. II. 28 STRUCTURE OF GALLINACEOUS BIRDS. [cH. I notched sternum, the more rudimentary vomer, and the abnormal position of the hallux in the Alectoropodes. This is very likely so, but as Fiirbringer’ has pointed out, the structure of the soft parts of the Megapodes are more different from what is found in the Curassows. The latter often have a convoluted windpipe, which does not occur in the Megapodes, but is met with in some Guinea fowls and in the Grouse (Tetrao urogallus). The Megapodes have lost one of the two carotid arteries, and their oil gland has not the tuft of feathers found in other Galli- naceous birds; in fact, as regards internal structure other than that of the skeleton, the Cracide are not so very near to the Megapodes. All these structural features will seem perhaps of small moment to the student of invertebrate anatomy; but it must be remembered that birds form a very circumscribed group; the anatomist is glad of the smallest characters upon which to found differences; and the differences enumerated are not small, considering the characteristics of the order. Apart, how- ever, from these differences it does appear the two families Cracidee and Megapodide are the most primitive Gallin- aceous birds now in existence; not only do the two points referred to above tend to show this, but we might also perhaps urge the “reptile-like habit” which the Megapodes have of laying their eggs in a heap of dead leaves and abandoning them to the kindness of nature— a habit which of course recalls that of nearly all Reptilia. Moreover neither the Curassows nor the Mound-builders show to anything like so great a degree that difference in 1 Untersuchungen zur Morphologie der Vigel. CH. I] FOSSIL REPRESENTATIVES, 29 plumage between the sexes which is often developed to so extraordinary an extent in the other members of the group. In many birds which are presumed to be of an ancient type, for instance the Ostrich tribe, there is the same absence of strongly marked secondary sexual characters in colouration. In this particular enquiry we cannot unfortunately get any assistance whatsoever from Paleontology; the only fossil Megapode recorded in Lydekker’s Catalogue of Fossil Birds is Talegalla lathami, a species now living; there is no information as to extinct Cracide. The Alectoropodes on the other hand, are much more nearly connected among themselves. Fiirbringer, indeed, does not divide them further‘. The Guinea-fowls perhaps are the most distinct group; but the Argus, Pheasant and the Peacock are looked upon by Fiirbringer as somewhat intermediate between them and the more typical Phasianide. About this group of Gallinacee there is some paleontological information ; a few existing species (Lagopus albus, Francolinus pictus) have been described from the Pleistocene of Europe and India respectively; the extinct genus Palwortyz, “ Partridge-like birds,” containing eight species, occurs in the Eocene and Miocene of Europe; three species of an allied genus, Palwoperdix, are also found in the Miocene of Europe; five species of Phasianus have been found in the Pliocene and Miocene of Europe; four species of Gallus occur in the “superficial deposits” of New Zealand, the “Cavern-deposits of the Lahn valley, Germany” and the Pliocene of France; of greater interest is the genus Tao- 1 Nor Gadow with any confidence, see Bronn’s ‘‘ Thierreich.” Aves. 30 CRACIDEZ AND MEGAPODID. {cH. I perdix (one species only), from the Eocene of France, which is said to present affinities with Nuwmida and Meleagris, ie. with African and American forms. The facts which have so far been enumerated enable us to draw some interesting conclusions; the first is indis- putable; each of the great divisions of the globe is tenanted by a special group of Gallinaceous birds, which is with the exception of the nearly cosmopolite Tetraonide, confined to that particular region, There are some reasons for considering that the cosmopolitan Tetraonide are of a less ancient stock than the restricted Cracide and Megapodide. There is a closer structural connection between the Gallinaceous birds of the three great con- tinents of Europe, Africa and Asia, than between any one of them and the Gallinaceous inhabitants of South America or remote Australia. The two latter regions, being truly the ends of the earth, are populated by the two most ancient types of Gallinaceous bird, which how- ever are not very closely allied. In a very tentative way we may point out another possible conclusion. We may presume that the earth was possessed, as regards Gallinaceous birds, by an ancient stock of which the Cracide and Megapodide are the only survivors; later on, from the ancient stock, arose other families which increased and multiplied so much as to drive their forerunners into the more remote corners, where an inroad of the sea preserved them from further competition; as the remnants of the more ancient race came thus to be widely separated and exposed to divergent conditions they would naturally get to be more CH.I] RANGE OF SLOTHS AND ANTEATERS. 31 and more unlike each other; hence the differences between the Cracide and Megapodide. We find in fact the presumably younger race spreading over the whole earth, while the remnants of the older race are limited to the more remote parts. All this fits in well, as will be remarked later, with the Polar theory of the origin of life. Distribution of the Edentata. The Edentata are a group of mammals in which the distribution has a very strong relation to anatomical structure. That there is this intimate connection has been shown by recent anatomical investigation, which is summed up and its purport explained in a paper by Sir William Flower‘, from which the information upon the subject can be most conveniently got. Formerly the members of this group were divided into families not at all consistent with deep lying structure but rather with superficial modification depending upon similar habits and ways of life. Before examining the rational classifi- cation of the group as proposed by Sir William Flower, it will be convenient to briefly pass in review the different genera and families into which the Edentata may be divided. The living members of the group readily separate into five families, which are the following :—Bradypodide, or sloths, containing the two genera Bradypus and Cholepus. Myrmecophagide, or anteaters, with the genera Myr- mecophaga, Tamandua and Cycloturus. Dasypodide, or Armadillos, with the six genera, 1 P.Z.S. 1882, p. 358. 32 OLD WORLD EDENTATES. (cH. I Tatusia, Dasypus, Xenurus, Priodon, Tolypeutes and Chlamydophorus. Manide, or scaly anteaters, with really only one genus Manis, though more have been allowed by some system- atists. Orycteropodide, or Cape anteaters, containing but a single genus Orycteropus. Now it has been customary to associate together the anteaters of both the old and the new world, sepa- rating them on the one hand from the sloths and on the other from Orycteropus. We thus get a group ranging over South America, the greater part of the Oriental region, and a large tract of the Ethiopian region, for Manis is found in both of the Jast named regions. The Manis of the old world has a strong superficial likeness to the anteaters of the new world. The same long tongue and well-developed salivary glands are present in both, while neither of them have any teeth; correlated with this likeness in structure is the fact that both feed upon ants. The Australian anteater, Echidna, was on this account placed by Linneus in the same preat group as that containing the Edentata; it has in the same way a long tongue and well developed salivary glands. There is, however, of course no intimate connection between the animals; we have here merely a case of modification to the same end, the utilising of an abundant ant or termite supply. The Woodpecker and the Chameleon show a remarkably analogous modification of the alimentary organs. This is really the only reason, apart perhaps from a general similarity in form, which has led to the ‘'VIVINSGS JO NOILNSIYLSIC CH. I] STRUCTURE OF EDENTATES. 33 uniting of the Old World with the New World anteaters. But an examination of other structures does not show any likeness between the Manis and the Myrmecophaga, but does reveal an unmistakeable resemblance between the latter and the sloth, to which it is so unlike in external form and in habits. In the American anteater the vertebre to a large extent interlock with each other by an additional series of articular processes, not found in other mammals, excepting only the Dasypodide, and the extinct Megatheriide, and which is to a less extent but still obvious in the sloths. This structural resem- blance found in animals of such diverse habit must have a significance in considering their affinities. The fact that the processes in question are in a rudimentary condition in the sloths is correlated with the fact that in those animals which depend from the branches of trees, and use the muscles of the back but little, the articular processes generally of the spine are poorly developed; the fact is they are of even greater importance as evidence of real blood relationship. As Sir William Flower says, the fact may be almost said to prove “that the sloths are descended from animals in which they existed in a fully developed form.” Neither in Manis nor in Orycteropus’ are there the slightest vestiges of these additional articu- lar processes. It is also pointed out that the shape of the sternum is characteristic of the New World and the Old World Edentates respectively. When we come to other details of structure there is the same alliance to be noted between the various families of Edentata found in America, more particularly is this to be seen in the anatomy of the B. Z. 3 34 EXTINCT EDENTATES. [cH. I reproductive organs. On the other hand, the Oryctero- podide are not so nearly allied to Manis as are the different genera of New World Edentates among them- selves. But though this is the case it does not show any special affinities with the New World forms. It may be temporarily regarded as a distinct family having (at present) problematic relations to the Manide, which it resembles by what are principally negative characters. The placenta is an organ which is of great assistance in bringing out affinities between various groups of mammals; unfortunately it is not known whether the placenta of Orycteropus is deciduate or the reverse; if non-deciduate there is a likeness to the Manis and not to the New World Edentates. At any rate it is clear that Manis, even if it be not a close ally of Orycteropus, is still further removed from the Myrmecophagide, Dasypodide or Bradypodide.