THE INTERNATIONAL SCIENTIFIC SERIES VOLUME LYII THE INTERNATIONAL SCIENTIFIC SERIES. EACH BOOK COMPLETE IN ONE VOLUME, 12MO, AND BOUND IN CLOTH. 1. FOKMS OF WATER : A Familiar Exposition of the Origin and Phe- nomena of Glaciers. By J. TYNDALL, LL. D., F. K. S. With 25. Illustrations. $1.50. 2. PHYSICS AND POLITICS ; Or, Thoughts on the Application of the Principles of "Natural Selection" and "Inheritance" to Political Society. By WALTER BAGEHOT. $1.50. 3. FOODS. By EDWARD SMITH, M. D., LL. B., F. R. S. With numerous Illustrations. $1.75. 4. MIND AND BODY: The Theories of their Eelation. By ALEX- ANDER BAIN, LL. D. With 4 Illustrations. $1.50. 5. THE STUDY OF SOCIOLOGY. By HERBERT SPENCER. $1.50. 6. THE NEW CHEMISTRY. By Professor J. P. COOKJS, of Harvard University. With 31 Illustrations. $2.00. 7. ON THE CONSERVATION OF ENERGY. By BALFOUR STEWART, M. A., LL. D., F. R. S. With 14 Illustrations. $1.50. 8. ANIMAL LOCOMOTION; or, Walking, Swimming, and Flying. By J. B. PETTIGREW, M. D., F. R. S., etc. With 130 Illustrations. $1.?5. 9. RESPONSIBILITY IN MENTAL DISEASE. By HENRY MAUDS- LEY, M. D. $1.50. 10. THE SCIENCE OF LAW. By Professor SHELDON AMOS. $1.75. 11. ANIMAL MECHANISM: A Treatise on Terrestrial and Aerial Loco- motion. By Professor E. J. MAREY. With 117 Illustrations. $1.75. 12. THE HISTORY OF THE CONFLICT BETWEEN RELIGION AND SCIENCE. By J. W. DRAPER, M. D., LL. D. $1.75. 13. THE DOCTRINE OF DESCENT AND DARWINISM. By Pro- fessor OSCAR SCHMIDT (Strasburg University). With 26 Illustra- tions. $1.50. 14. THE CHEMICAL EFFECTS OF LIGHT AND PHOTOGRAPHY. By Dr. HERMANN VOGEL (Polytechnic Academy of Berlin). Trans- lation thoroughly revised. With 100 Illustrations. $2.00. New York: D. APPLETOX & CO., 1, 3, & 5 Bond Street. 2 The International Scientific Series. — (Continued.) 15. FUNGI : Their Nature, Influences, Uses, etc. By M. C. COOKE, M. A., LL. D. Edited by the Eev. M. J. Berkeley, M. A., F. L. S. With 109 Illustrations. $1.50. 16. THE LIFE AND GKOWTH OF LANGUAGE. By Professor WILLIAM DWIGHT WHITNEY, of Yale College. $1.50. 17. MONEY AND THE MECHANISM OF EXCHANGE. By W. STANLEY JEVONS, M. A., F. B. S. $1.75. 18. THE N A TUBE OF LIGHT, with a General Account of Physical Optics. By Dr. EUGENE LOMMEL. With 188 Illustrations and a Table of Spectra in Chronio- lithography. $2.00. 19. ANIMAL PARASITES AND MESSMATES. By Monsieur VAN BENEUEN. With 83 Illustrations. $1.50. 20. FEE MENTATION. By Professor SCKUTZENBERGER. With 28 Illus- trations. $1.50. 21. THE FIVE SENSES OF MAN. By Professor BERNSTEIN. With 91 Illustrations. $1.75. 22. THE THEOEY OF SOUND IN ITS EELATION TO MUSIC. By Professor PIETEO BLASERNA. With numerous Illustrations. $1.50. 23. STUDIES IN SPECTEUM ANALYSIS. By J. NORMAN LOCKYER, F. E. S. With 6 Photographic Illustrations of Spectra, and numer- ous Engravings on Wood. $2.50. 24. A HISTOEY OF THE GEOWTH OF THE STEAM-ENGINE. By Professor E. H. THURSTON. With 163 Illustrations. $2.50. 25. EDUCATION AS A SCIENCE. By ALEXANDER BAIN, LL.D. $1.75. 26. STUDENTS' TEXT-BOOK OF COLOE; Or, Modern Chromatics. With Applications to Art and Industry. By Professor OGDEN N. EOOD, Columbia College. New edition. With 130 Illustrations. $2.00. 27. THE HUMAN SPECIES. By Professor A. DE QUATREFAGES, Mem- bre de 1'Institut. $2.00. 28. THE CEAYFISH : An Introduction to the Study of Zoology. By T. H. HUXLEY, F. E. S. With 82 Illustrations. $1.75. 29. THE ATOMIC THEOEY. By Professor A. WURTZ. Translated by E. Cleminshaw, F. C. S. $1.50. New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. The International Scientific Series.— (Continued.) 3 30. ANIMAL LIFE AS AFFECTED BY THE NATURAL CONDI- TIONS OF EXISTENCE. By KAKL SEMPER. With 2 Maps and 101} Woodcuts. $2.00. 31. SIGHT : An Exposition of the Principles of Monocular and Binocular Vision. By JOSEPH LE CONTE, LL. D. With 132 Illustrations. $1.50. 32. GENERAL PHYSIOLOGY OF MUSCLES AND NERVES. By Professor J. ROSENTHAL. With 75 Illustrations. $1.50. 33. ILLUSIONS : A Psychological Study. By JAMES SULLY. $1.50. 34. THE SUN. By C. A. YOUNG, Professor of Astronomy in the College of New Jersey. With numerous Illustrations. $2.00. 35. VOLCANOES : What they Are and what they Teach. By JOHN W. JUDD, F. R. S., Professor of Geology in the Royal School of Mines. With 96 Illustrations. $2.00. 36. SUICIDE : An Essay in Comparative Moral Statistics. By HENRY MORSELLI, M. D., Professor of Psychological Medicine, Royal Uni- versity, Turin. $1.75. 57. THE FORMATION OF VEGETABLE MOULD, THROUGH THE ACTION OF WORMS. With Observations on their Habits. By CHARLES DARWIN, LL. D., F. R. S. With Illustrations. $1.50. 38. THE CONCEPTS AND THEORIES OF MODERN PHYSICS. By J. B. STALLO. $1.75. 39. THE BRAIN AND ITS FUNCTIONS. By J. LUYS. $1.50. 40. MYTH AND SCIENCE. By TITO VIGNOLI. $1.50. 41. DISEASES OF MEMORY: An Essay in the Positive Psychology. By TH. RIBOT, author of " Heredity." $1.50. 42. ANTS, BEES, AND WASPS. A Record of Observations of the Habits of the Social Hyinenoptera. By Sir JOHN LUBBOCK, Bart., F.R.S., D. C.L., LL.D., etc. $2.00. 43. SCIENCE OF POLITICS. By SHELDON AMOS. $1.75. 44. ANIMAL INTELLIGENCE. By GEORGE J. ROMANES. $1.75. 45. MAN BEFORE METALS. By N. JOLY, Correspondent of the Insti- tute. With 148 Illustrations. $1.75. New York : D. APPLETON & CO., 1, 3, & 5 Bond Street. 4 The International Scientific Series. — (Continued.) 46. THE ORGANS OF SPEECH AND THEIR APPLICATION IN THE FORMATION OF ARTICULATE SOUNDS. By G. H. VON MEYER, Professor in Ordinary of Anatomy at the University of Zurich. With 47 Woodcuts. $1.75. 47. FALLACIES : A View of Logic from the Practical Side. By ALFRED SIDGWICK, B. A., Oxon. $1.75. 43. ORIGIN OF CULTIVATED PLANTS. By ALPIIONSE DE CANDGLLE. $2.00. 49. JELLY-FISH, STAR-FISH, AND SEA-URCHINS. Being a Re- search on Primitive Nervous Systems. By GEORGE J. ROMANES. $1.75. 50. THE COMMON SENSE OF THE EXACT SCIENCES. By the late WILLIAM KINGDON CLIFFORD. $1.50. 51. PHYSICAL EXPRESSION : Its Modes and Principles. By FRANCIS WARNER, M. D., Assistant Physician, and Lecturer on Botany to the London Hospital, etc. With 51 Illustrations. $1.75. 52. ANTHROPOID APES. By ROBERT HARTMANN, Professor in the University of Berlin. With 63 Illustrations. $1.75. 53. THE MAMMALIA IN THEIR RELATION TO PRIMEVAL TIMES. By OSCAR SCHMIDT. $1.50. 54. COMPARATIVE LITERATURE. By HUTCHESON MACAULAY Pos- NETT, M. A., LL. D., F. L. S., Barrister-at-Law ; Professor of Clas- sics and English Literature, University College, Auckland, New Zealand ; author of " The Historical Method," etc. $1.75. 55. EARTHQUAKES AND OTHER EARTH MOVEMENTS. By JOHN MILNE, Professor of Mining and Geology in the Imperial Col- lege of Engineering, Tokio, Japan. With 38 Figures. $1.75. 56. MICROBES, FERMENTS, AND MOULDS. By E. L. TROUESSART. With 107 Illustrations. $1.50. New York: D. APPLETON & CO., 1, 3, & 5 Bond Street. V- THE INTERNATIONAL SCIENTIFIC SERIES THE GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION OF ANIMALS BY ANGELO HEILPRIN PROFESSOR OF INVERTEBRATE PALEONTOLOGY AT, AND CURATOR-IN-CHARGE OF. THE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA ; PROFESSOR OF GEOLOGY AT THE WAGNER FREE INSTITUTE OF SCIENCE, PHILADELPHIA; MEMBER OF THE AMERICAN PHILOSOPHICAL SOCIETY, &C. NEW YORK D. APPLETON AND COMPANY 1887 COPYRIGHT, 1886, BY D. APPLETON AND COMPANY. All rights reserved. TO PROFESSOR JOSEPH LEIDY, M. D., LL. D., PRESIDENT OF TIIE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA, AC., WHOSE PROFOUND RESEARCHES HAVE SO LARGELY TENDED TO DEVELOP THE SCIENCE OF BIOLOGY, THIS VOLUME IS RESPECTFULLY DEDICATED. A. H. PEEFACE. IN the preparation of the following pages the author has had two objects in view : that of presenting to his readers such of the more significant facts connected with the past and present distribution of animal life as might lead to a proper conception of the relations of existing faunas ; and, secondly, that of furnishing to the student a work of general refer- ence, wherein the more salient features of the geography and geology of animal forms could be sought after and readily found. The need of such a work has been frequently felt and expressed. As far as he is aware, no work of that kind has as yet appeared, and therefore, to a certain extent, this publication stands alone in the field it is intended to cover. Necessarily, much that it embraces can be found elsewhere, and treated even at considerably greater length ; but the mat- ter is not contained under a single cover, and where a special subject is expounded in extenso the treatment is usually too exhaustive to permit of immediate use by the general reader. This applies particularly to zoogeography. With reference to geological distribution there is little connectedly written — indeed, beyond what is found in text-books largely devoted to cognate subjects, practically nothing. Moreover, what little of connected literature on the subject we do possess is almost Vlll PKEFACE. entirely out of date, and in no way represents the present status of the science. The subject of geographical and geological distribution is so vast that no full treatment of it could be expected in the limited number of pages set apart for it in the present work. The author has, therefore, been obliged to omit, or at least largely ignore, the consideration of some of the less impor- tant animal groups, and, while recognising the deficiencies resulting from such omission, trusts that it will not detract much from the general usefulness of the publication. The plan of treatment followed in the early part of the book (geographical distribution) is largely that so admirably un- folded by Mr. Wallace, to whom, for the constant use of his works, the author is under great obligations. He also wishes to express his special indebtedness to the pioneer workers in this field, Schmarda and Murray, whose writings have laid the foundation of much of our existing knowledge in the premises. No special mention need be made of the numer- ous other authors who have contributed more or less exten- sively to the subject under consideration, and whose works have aided in the preparation of the present volume ; to those, collectively, the author likewise desires to acknowledge his indebtedness. A few words need be said in relation to the zoogeographical regions that are recognised in this work, which differ essen- tially from those generally adopted by naturalists. The rea- sons for uniting the "Nearctic" and " Palaearctic " regions of zoogeographers into a single realm, designated, in accord- ance with a suggestion by Professor Alfred Newton, of Cam- bridge, the " Holarctic," are fully set forth in my paper " On the Value of the Nearctic as one of the Primary Zoological Regions," published in the " Proceedings of the Academy of Natural Sciences of Philadelphia," for December, 1882. Ob- PREFACE. IX jections by Mr. Wallace and Professor Gill to the views there formulated appear in " Nature " of March 22, and June 7, 1883, respectively, and my rejoinders to the criticisms of these gentlemen in " Nature " of April 26, and the " Proceed- ings " of the Philadelphia Academy for November, 1883. To these papers I must refer the reader for a purely technical statement of the case. The classification of the " Transition" tracts is largely that which has been proposed by Forsyth Major, in " Kosmos" for 1884. ACADEMY OF NATURAL SCIENCES, PHILADELPHIA, October, 1886. CONTENTS. PART I. G EO GRAPHIC A L DISTRIB UTION. I. PAGE General principles of zoogeography. — Faunal variation. — Faunas of isola- tion.— Eolations of past and present faunas. — Origination of faunas . 1 II. Areas of specific distribution. — Generic distribution. — Distribution of fam- ilies and orders. — Conterminous and discontinuous areas of distribu- tion 17 III. Conditions affecting distribution. — Climate. — Food-supply. — Earners to migration. — Migrations of mammals and birds. — Dispersal of amphib- ians and reptiles. — Dispersal of insects and mollusks . . . .35 IV. Zoological regions. — Holarctic realm. — Neotropical. — Ethiopian.— Orien- tal.— Australian. — Polynesian. — Tyrrhenian, Sonoran, and Austro- Malaysian transition regions 55 V. Distribution of marine life. — Nature of the deep-sea fauna. — Oceanic pelagic fauna. — Littoral fauna. — Pelagic faunas of lakes. — Deep-lake faunas , . 109 Xll CONTENTS. PART II. G EOL 0 GICAL DISTRIB UTION. PAGE The succession of life. — Faunas of the different geological periods . . 133 II. Appearance and disappearance of species. — Reappearance. — Extinction. — Persistence of type-structures. — Variation. — Geological breaks. — Geographical distribution. — Climatic zones. — Synchronism of geo- logical formations 181 PART III. GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. I. The present and past distribution of individual animal groups. — Forami- nifera.— Corals. — Brachiopoda. — Mollusca generally.— Crustacea,— Insecta, Arachmda, and Myriapoda 234 II. Distribution of the Vettebrata.— Fishes.— Amphibians.— Reptiles.— Birds. -Mammals . 287 PAET I. GEOGRAPHICAL DISTRIBUTION. I. General principles of zoogeography. — Faunas of isolation. — Relations of past and present faunas. EVERYWHERE upon the surface of the earth we meet with mani- festations of animal life. The desert wastes, no less than the trop- ical jungles, the bleak ice-fields of the frozen north, and the most elevated mountain-summits—all have their faunas. The abyss of the sea, no less than its surface, contributes its quota to the animal world ; and in the atmosphere all around us, from the lowest stra- tum not unlikely to the highest, the germs of the organic universe lie everywhere scattered about. In what precise form or guise this life first manifested itself, or how inert matter became endowed with that potentiality which we recognise in vital energy, it seems hopeless to attempt to determine. True science takes cognisance of both fact and theory, but illusory speculation, whose ground- work is a simple outgrowth of the imagination, must find a rest- ing-place without its domain. No one who has paid the smallest amount of attention to the facts of nature as they present themselves can have failed to notice certain peculiarities in the way of the distribution of life, which do not always admit of an immediate or of a satisfactory inter- pretation. Why, for example, one piece of country should differ so essentially in its faunal aspects from another whose physical characteristics are practically identical with its own ; why the sec- ond should differ from a third, and this, again, from a fourth — may not appear comprehensible. Nor any the more comprehensible 2 GEOGRAPHICAL DISTRIBUTION. may appear the circumstance that, in most cases, island faunas are so eminently marked out from those of continental areas. Another peculiarity in faunal distribution is presented in the fact that, while certain animal assemblages enjoy an almost limit- less or universal extension, others, again, without apparent reason, are circumscribed within limits of the opposite extreme. The trav- eller to the most distant shores not infrequently recognises objects that are familiar to him as those of his native home, although pos- sibly, in the interval of his journey, he has completely lost sight of their existence, so different might have been the creatures that successively met his gaze. " When an Englishman travels by the nearest sea-route from Great Britain to Northern Japan, he passes by countries very unlike his own, both in aspect and natural pro- ductions. The sunny isles of the Mediterranean, the sands and date-palms of Egypt, the arid rocks of Aden, the cocoa-groves of Ceylon, the tiger-haunted jungles of Malacca and Singapore, the fertile plains and volcanic peaks of Luzon, the forest-clad mountains of Formosa, and the bare hills of China, pass successively in review; till after a circuitous voyage of thirteen thousand miles he finds himself at Hakodadi, in Japan. He is now separated from his starting-point by the whole width of Europe and Northern Asia, by an almost endless succession of plains and mountains, arid deserts or icy plateaux, yet when he visits the interior of the coun- try he sees so many familiar natural objects that he can hardly help fancying he is close to his home. He finds the woods and fields tenanted by tits, hedge-sparrows, wrens, wagtails, larks, red- breasts, thrushes, buntings, and house-sparrows, some absolutely identical with our own feathered friends, others so closely resem- bling them that it requires a practised ornithologist to tell the differ- ence. If he is fond of insects he notices many butterflies and a host of beetles which, though on close examination they are found to be distinct from ours, are yet of the same general aspect, and seem just what might be expected in any part of Europe. There are also, of course, many birds and insects which are quite new and peculiar, but these are by no means so numerous or conspicuous as to remove the general impression of a wonderful resemblance between the productions of such remote islands as Britain and Yesso."* * Wallace, " Island Life," p. 3. FAUNAE VARIATION. 3 On the other hand, a journey of only very moderate duration will frequently disclose the greatest diversity existing between con- tiguous faunas. The traveller who starts east from the African coast, and who has familiarised himself with the strange produc- tions of the African continent, its elephants, giraffes, rhinoceroses, hippopotami, lions, and antelopes, finds none of these in the island of Madagascar ; the true monkeys have also disappeared, and in their place he meets with forms of half -monkeys (lemurs), a group of animals with which he will have already become acquainted before leaving the mainland. Strange creatures, wholly unlike anything previously known to him, now arrest his attention, and he finds himself in the midst of. what might be termed a peculiar fauna. Likewise, if he leave the shores of Central America or Florida for the Great Antilles, the same marked isolation of the new fauna manifests itself. The larger forms of quadrupeds, such as the jaguar, couguar, tapir, and peccary, are wholly wanting, and even among the smaller and more numerously represented mammalian types many of the more prominent forms will be sought for in vain. On the other hand, he will make the acquaintance of entirely new groups of animals, some of which, like the Centetida?, have their nearest foreign representatives in regions removed by nearly one-half the circumference of the globe. And this diversity in the faunal type is found to permeate to a greater or less extent all the individual groups, birds, reptiles, &c., of the animal king- dom. It might be rashly supposed that the distance separating the regions under comparison would sufficiently account for the pecu- liarities of their respective faunas, or the disparities separating them; but distance alone, without a special relation binding to- gether the principals between which it is supposed to act, can effect nothing. We have, indeed, seen upon what a vast extent of territory the British faunal facies is stamped, and were any further proof needed of the inefficacy of distance, pure and simple, as a prime factor in geographical distribution, we have but to transport ourselves to the Malay Archipelago, and observe how wonderfully diverse are the respective faunas on either side of the very narrow (but deep) channel separating the islands of Bali and Lombok from each other. Mysterious as these various phenomena of distribution may ap- 4: GEOGRAPHICAL DISTRIBUTION. pear, they yet have all their logical explanation. A quarter of a century ago, when the doctrine of independent creation still held sway over the minds of most 'naturalists, and when the organic universe was reflected in the eye of the investigator as an incon- gruous agglomeration of disjointed parts, there was, indeed, no necessity for specially accounting for the facts, since they were conceived to be such by reason of a previous ordination. Now, however, when the full value of the evolutionary process is recog- nised, and animate nature has come to be looked upon as a con- crete whole, bearing special relations to its numberless parts, each individual fact seeks its own explanation, which explanation must of necessity stand in direct harmony with some previously observed fact. When, therefore, wre seek to unravel the tangle of zooge- ography, and to harmonise its apparent incongruities, we must at the outset admit that distribution, such as it is, is the outcome of definite interacting laws — laws which stand in relation to each other as absolutely as they do in any other field of action — and not a hap-hazard disposition, as some would lead us to suppose, setting all enquiry at defiance. The naturalist who in the Western Hemisphere journeys south- ward from the ice-covered fields of British America fails to notice any very sudden or marked alternation in the character of the faunas that successively meet his view. New features are being constantly added, and old ones eliminated, but the interchange is effected so gradually that it becomes difficult to determine the limitations that properly define one fauna from another. The fur- bearing animals of the far north send their representatives into regions which border the habitats of the more exclusively tropical species, or are succeeded by forms which differ but little from them. The skunk, many of whose associates are animals of a distinctively Arctic character, finds its way into Mexico, and the ermine, which penetrates to the farthest northern point reached by mammals generally, still lingers on in some of the Southern United States. The Arctic fox is succeeded by the equally abundant types of the grey and the red fox ; and similarly, the polar bear is followed on the one side of the continent by the grizzly, and on the other by the black bear. Having descended into the middle temperate regions, the traveller still finds about him mostly the forms with which he has already become acquainted. But many of the more FAUNAL VAKIATION". 5 familiar types have either wholly disappeared, or are fast disap- pearing. Such may be the musk-sheep, moose, stag, and reindeer, which will have left as their successors the bisons and the various species of smaller deer which range throughout the remainder of the continent. The grey wolf of the northern forests breaks up into a number of varietal forms more or less distinct from the typical one, and is earned by the coyote into the heart of Mexico. Farther to the south the traveller observes entirely new features gradually appearing. In Arkansas he possibly meets with the pec- cary, the first indigenous member of the pig family with which he will have become acquainted; in Texas, with the armadillo, the first of that group of animals, the Edentata, which, in the past and present history of the South American continent, constitutes such an important element in its fauna ; and, in the States adjoining the Mexican Republic, with an abundant representation of the iguanid lizards, which, by their numbers, so eminently typify the follow- ing region of the tropics. There are as yet neither monkeys, ta- pirs, nor guinea-pigs, but the first appear in Southern Mexico, the second in Central America, and the last in Venezuela or Guiana. The traveller is now in the region of the Equator, and surrounded by an association of animal forms most of which were unknown to him when he entered upon his journey, and which in many respects depart so widely from those with which he was familiar at his start- ing-point as to constitute a distinct fauna. There is no longer either wolf, fox, or catamount, beaver or musk-rat, and of the spe- cifically importan c group of the hares or rabbits but a single species remains. The solitary species of bear is so different from its*north- ern cousins as to be regarded by some naturalists as the type of a distinct genus. The contrast between the successive faunal changes observed on the north and south journey and the faunal identity which so aston- ishes the traveller whose journey is directed eastward from Eng- land to Japan is very great. And yet if the traveller from Britain, instead of proceeding due eastward, were to shape his course a few degrees to the south, much the same kinds of changes as he noticed on his American trip would again present themselves. Along the shores of the Mediterranean he would no longer, or only at rare intervals, meet with his associates of the Arctic north ; on the southern slopes of the Caucasus the tiger, and in Arabia the 6 GEOGRAPHICAL DISTRIBUTION. camel, gazelle, and ostrich, would present to him certain features of a fauna which was in the main unknown to him ; in India the ele- phant, lion, and rhinoceros, and other curious denizens of the jun- gle, the python and crocodile, and the numerous birds of resplendent plumage, would probably crowd from his memory the forms of the creatures ordinarily most familiar to him, and lead a passage to the ultimate goal of his journey, Australia, where he would meet with the most singular and most distinctive fauna on the surface of the earth. Much nearer to his northern home — on opposite sides of the Mediterranean — and with much less travelling, the naturalist will discern scarcely less well-marked faunal differences or peculiarities. To account for the anomalies which the facts of distribution present is the still unsolved problem that is put before the zoogeographer. Granting, with the doctrine of evolution, that all the complex assemblages of existing animal forms are modified derivatives from previously existing forms, and that these are ultimately to be traced back to some common ancestor, it must of necessity follow that any given fauna will depend for the degree of its peculiarity, whether great or small, upon the amount of modification, relative to any other fauna, which it will have undergone. And this modification can be effected in two ways : by inherent modification of the indi- vidual types composing the fauna, and by intermixture with, or immigration from, contiguous or neighbouring faunas. In both cases, manifestly, isolation or its opposite, union of habitation, will constitute the governing factor in determining the amount of varia- tion. A region that is broadly separated from all others will, natu- rally, tend to develop a fauna distinct from any other, since the progressive modifications in its constituent faunal elements must ul- timately lead to divergence ; and the greater the period of isolation the greater, of necessity, will be the amount of this divergence, or the more pronounced the faunal individualisation. Hence it is that in the greater number of the more distantly removed island groups, or in those which are separated by more or less impassable barriers from the nearest land-mass, we meet with such highly specialised faunas. The Galapagos Islands, for example, as will be more fully illustrated farther on, have a fauna very distinct from that of any part of South America, although removed from it by a distance of less than seven hundred miles. The birds are quite distinct, and so are ISLAND FAUNAS. 7 the reptiles, insects, and land mollusks. The island of St. Helena, in the South Atlantic, and the Sandwich Islands, in the North Pacific, present us with similar instances of faunal specialisation, and to a less extent, also, the group of the Azores. In the case of these last, which lie in the course of the storm-winds, a considerable intermixture has been effected with the faunas of Western Africa and Europe, for we find that by far the greater number of the resi- dent land-birds are inhabitants of those two continents as well. The fact that there are so very few peculiar forms is proof either of a recent separation of the islands from the mainland — not sufficient time having been allowed for the development of new species — or of a recent or repeated peopling with old forms from the continents. Even irrespective of considerations connected with the physical geography or geology of the region, it would naturally be inferred, from the prevalence of in-blowing storm-winds, and the known fact that certain birds are transported hither, that the second supposi- tion is the correct one; and that this is the true explanation is proved by evidence of a very positive character furnished by some of the other groups of animals. Thus, the land-Mollusca, which in their distribution are not so readily affected by aerial currents, are eminently distinguished from those of either Europe or Africa, or of any other continental land-mass, proving in their case a long- protracted period of isolation. Further, there is not a single species of fresh- water mollusk known in the entire group ! The Bermuda Islands, which are about equally distant from the mainland, occupy a nearly analogous position with respect of their fauna ; that is, par- tial interchanges have been effected with the fauna of the American continent. In all these cases, necessarily, the amount of faunal specialisation will be the index of the period of isolation. Where faunal immi- gration from a foreign region takes place it not only checks the development of a newly-forming fauna, by infusing into it an ele- ment that does not properly belong there, but also prevents in a measure that variation among individuals which might otherwise obtain. The case of the bobolink of the Galapagos Islands is a well-known example of this kind. It alone, of about thirty species of land-birds inhabiting those islands, is considered to be indis- putably identical with any form occurring on the mainland ; hence it is concluded that this is about the only species of South American 8 GEOGRAPHICAL DISTRIBUTION. bird that ever visits the islands, for, if the case were otherwise, it would be incredible that no more common forms should have been detected there. But the fact that the bobolink has remained abso- lutely identical with the common form of South America, whence, doubtless, most of the species of Galapagos birds have been derived, while all the other birds of the island group have undergone more or less modification since the islands were first tenanted, proves that variation in its case has been prevented by the perpetuation of nor- mal characters through interbreeding with the continental migrants. In other words, the breed has been kept true. Were the migrations of the visitors checked or interrupted, there can be little question that the island breed of bobolinks would undergo the same kind of modification which distinguishes the other birds, and which has developed in them new specific or varietal types. In the conti- nent of Australia, again, we meet with the most remarkable exam- ple of a highly specialised fauna being developed as the result of long-continued isolation. Of all the varied mammalian forms which elsewhere crowd the surface of the earth we have here but the merest trace, for, with the exception of the rodents and bats, none of the ordinary orders — Carnivora, Ungulata, Insectivora, &c. — are represented.* And even of the rodents there is but a single family, that of the mice (Muridae). On the other hand, the implacental mammals — kangaroos, wombats, duck-bill — whose only non-Aus- tralian representatives are the American family of opossums (Didel- phidae), acquire here a wonderful development, and exhibit a diver- sity of type-structure not met with in any other order of mammals. Now, the animals of this class, or such as might be considered most nearly allied to the marsupials, are the first of the Mammalia to ap- pear in geological time, and they alone have thus far been detected in any of the deposits (Triassic, Jurassic) of the middle geological period, or Mesozoic era. They constitute the most primitively or- ganised members of their class, and probably stand not far removed from what may ultimately be proved to be the bottom of the mam- malian series. In order to explain the anomalies of the Australian mammalian fauna we must have recourse to the hypothesis of isolation, for in * The Australian wild-dog, or dingo, may prove to be indigenous, in which case it would represent the Carnivora. AUSTRALIAN FAUNA. 9 no other way could we satisfactorily account for the remarkable development of the marsupial types, and the almost total absence of the commoner forms that are elsewhere so abundant. The oceanic barriers have evidently prevented that diffusion of species which would otherwise have sufficed to render the Australian fauna cosmopolitan in character. That this isolation, further, of the con- tinent has been of very great duration is proved by the long period of time, dating from the Cretaceous epoch, during which the most diverse forms of mammalians have existed, and the high specialisation that its own fauna has acquired. It may appear not a little surprising, in view of what has preceded, that two groups of animals, so widely removed from the rest of the Australian mam- malian fauna as are the mice and bats, should yet constitute a part of this fauna. In the case of the bats it is not difficult to ac- count for their occurrence in the region in question, since their powers of flight have enabled them to overcome such obstacles as to other animals might have proved true barriers to migration. The mice, on the other hand, whose disposition to gnaw into, and conceal themselves among, timber of all kinds, is well known, may have found their way hither from the Asiatic continent or its ad- joining islands through the intermedium of floating masses of vege- tation. Much more inexplicable is the occurrence of the single non- Australian family of marsupials, the opossums, on the American continent, which is removed by a continuous water-way of several thousands of miles, when not a single member of the entire sub-class of implacental mammals is found on any other part of the earth's surface outside the Australian region. The hypothesis that land connection by way of the Antarctic region at one time existed be- tween Australia and South America, and, possibly, also Africa, may or may not be true, but the evidence that has thus far been adduced tending to show that by such connection a transference of one section of the Marsupialia has been effected from one con- tinent to the other is certainly very slim. Yet it is by no means impossible that such may have been the case. The Edentata — armadillos, ant-eaters, pangolins — whose home is preeminently the two great continents of the Southern Hemisphere, and which barely trespass north of the Tropic of Cancer, and the struthious birds, like the rhea, ostrich, and cassowary, offer equal perplexities in the way of an explanation of their anomalous distribution with the 2 10 GEOGEAPHICAL DISTRIBUTION". marsupials, and they have likewise been considered to afford proof of a land connection such as has been indicated. A serious diffi- culty, however, that lies in the way of this explanation is the important fact that none of the characteristic African or South American mammals are found in Australia, for it might justly be contended that if a migration or transferrence was effected in one direction, it could have been effected in the opposite direction as well. But that such reciprocal distribution did not obtain is very nearly certain. It may, indeed, be assumed that at the time of a possible Australian migration the extremities of the southern con- tinents were not yet inhabited ; but this is very unlikely. Or, it may be further assumed, with Rtitimeyer, that the animals under consideration had a polar origin, and that they were distributed northward along continental lines that possibly now lie buried beneath the sea ; but positive evidence in this direction is still wholly wanting. An element in the problem which very materi- ally narrows the issue is the circumstance that marsupial remains have been found in the temperate regions of the Northern Hemi- sphere, and in both Europe and North America in deposits as an- cient as the Triassic period. In this upper tract, therefore, we find a possible and more probable clue towards the explanation of the existing distribution of the animals in question ; and if it be objected that some such living forms ought still to be found in the connecting region, the fact, nevertheless, remains that they did there once exist, but have since become largely extinct. It will be evident that the key to the solution of the more marked peculiarities of modern distribution must be sought in the records of the past, for in the comparison between existing and preexisting faunas alone can we expect to determine the condi- tions upon which present faunas were established, and to ascertain the dates of their respective appearances or antiquity. In most regions of the earth's surface a most intimate relationship links together the existing fauna and the fauna of the geological period or periods immediately preceding. The Pliocene and Post-Pliocene marine shell-fish faunas of the Western United States are practically identical with the equivalent fauna of the (modern) adjoining seas; the Post-Pliocene mammals of Britain are such as still .roam about the land, although they include numerous forms which no longer exist there; in India a large proportion of the mammalian types TERTIARY FAUNAS. 11 that inhabit the region are already represented in deposits of the early Pliocene period ; and in Australia the abundant remains of Marsupialia amply testify to the identity of character which unites the faunas of the past and present periods. A certain amount of antiquity is thus established for the several regional faunas. The farther back in time we proceed, however, the less pronounced ap- pear the common characteristics of past and present periods ; and, finally, they disappear almost altogether. Thus, the Eocene shell- fish fauna of the Atlantic coast of the United States and of France and Great Britain is very unlike that of the seas adjoining those regions at the present day, although, in a measure, it finds its ana- logue in the corresponding fauna of the eastern tropical seas. The Miocene mammals of the American continent are almost wholly unlike those which now inhabit the region, and what little simi- larity still remains completely vanishes with the animals of the more ancient Eocene period. And the same holds good with the European Tertiary fauna. Yet there are a number of existing types which in their own region can be traced through a series of progressive modifications to ancestral forms more or less unlike them, which belong to a comparatively remote geological epoch. The horse of the Old World, for example, has been traced through a number of intermediate forms to the Old Tertiary Palaeotherium, one of the most abundantly represented mammalian genera of the de- posits of Western Europe. The deer of the same region finds early ancestors in the horned and hornless species which occur fossil in the Miocene deposits of France and Germany ; and not unlikely the wolf and fox see their progenitors among the early members of the canine race, whose remains have been traced to the Oligocene, and not impossibly also to the Eocene period. In so far as these ani- mals are concerned, therefore, we have direct evidence of a fauna of considerable antiquity developing in place. In other cases, how- ever, evidence of a very opposite character is often presented ; that is to say, faunas, or their components, are very frequently shown to be in a given region of only brief duration. Thus, although bears are very plentiful at the present time in the North American continent, they are not known to have existed there before the last geological period, the Post-Pliocene. And the same is true of the members of the ox-family (Bovidae) — most of which are, indeed, not represented at all as fossils — of which North America possesses 12 GEOGKAPHICAL DISTRIBUTION. five, in the main, widely-distributed species : two antelopes, two sheep (including the musk-ox), and the bison. The question as to how these animals obtained a foothold in the region which they now inhabit, whether they originated there as derivatives from previously- existing forms, or were introduced as migrants from some land-mass lying without their domain, can only be deter- mined by a reference to the still earlier fauna of not only this, but of other regions as well. In the case of the bears, for example, no immediate ancestors of the tribe have thus far been discovered in the Western Hemisphere antedating the Post-Pliocene epoch ; on the other hand, in the Eastern Hemisphere — Europe — the remains of such animals, and of the true bears themselves, are abundant in deposits of the earlier Pliocene age. Hence, the assumption appears almost unavoidable that the North American fauna received its ursine contingent from the Old World. The same may or may not be also true of the American Bovidae ; but the determination of this question is made difficult, or impossible, through the fact that at least two of the genera — Ovibos and Bison — occur fossil in the Post-Pliocene deposits, and there only, of both the Old and the New World, and consequently appear in the two hemispheres as being of approximately equivalent age. Yet the fact that neither goats, sheep, oxen, nor antelopes have thus far been discovered fossil on the North American continent, while their remains are suf- ficiently abundant in the deposits of Eurasia (Europe- Asia) of Post- Pliocene or even much older age, would seem to indicate that the true home of the Bovidae is the Old World, whence, by gradually spreading, and through the facilities afforded them in the way of a northern land connection, they eventually came 'to occupy a con- siderable portion of the New World as well. The giant sloth-like forms, such as the Megalonyx, Megatherium, and Mylodon, which in North America are associated with the remains of animals of indisputably Post-Pliocene age, occur in South America in an older formation, the Pliocene, and thus seemingly represent an invasion of the north from the latter continent. This conclusion appears further borne out by the circumstance that the Southern Hemi- sphere is the home of the animals of this class, and that, with scarcely a single exception (Moropus, ? Morotherium) no edentate form has thus far been discovered in any North American deposit antedating the period which represents the development of the FAUNAL MIGRATION. 13 South American forms. Similarly, the extinct proboscideans, mam- moth and mastodon, are of later date in America than in Eurasia, and are in all probability to be traced back to the latter region for the place of their birth. The countries of the Old World present to us perhaps no less direct evidence as to the origination of, or the lines of migration taken by, specific groups of organisms. The European mammalian fauna is at the present time not very unlike in its general features that of North America, but in the geological period immediately preceding the present one it numbered a host of forms wholly dif- fering from anything known to have existed in the corresponding period of American history, and, indeed, quite different from any- thing now inhabiting Europe. Such, for example, were the mam- moth, African elephant, hippopotamus, African lion, leopard, the spotted and the striped hyena, several species of rhinoceros, &c , forms the greater number of which are at the present day associated with the region lying south of the Mediterranean. The question that here presents itself is one, perhaps, that cannot be fully an- swered, but yet one whose partial solution is made very nearly certain. Did this fauna become suddenly exterminated, through some agency or other, in the region inhabited by it, or did it migrate elsewhere? There can be but little doubt that both conditions took place. The mammoth and the several species of (fossil) rhinoceros are now all extinct, and there is every reason to believe that their tribes per- ished gradually, without their having accomplished much migration immediately preceding final extermination. The case is, however, different with the other forms, for the fact of their inhabiting the African continent leads one to suspect that they may have found their way thither by way of some land connection no longer remain- ing. That such a connection uniting the two continents did exist within a comparatively recent geological period, permitting of an interchange of the respective faunas, is certain, as is proved by the numerous ties which bind together the faunas of the opposite shores of the Mediterranean. The Barbary ape of the Rock of Gibraltar inhabits Morocco, while the ichneumon of Spain, the porcupine of Italy, and the fallow-deer of the south of Europe generally, are all forms inhabiting the north of Africa as well. These animals evi- dently crossed over the intervening sea by some route or other, and, as has already been stated, in comparatively recent times, otherwise 14 GEOGKAPHICAL DISTRIBUTION. while the type-forms represented on the opposing shores might have been alike, the species would have almost undoubtedly differed. Equally positive proof in this direction is furnished by the similari- ties presented in the reptile and amphibian faunas. The shallow- ness of the channel separating Spain from Morocco renders it prac- tically certain that one such connecting land-mass occupied the position of the present Straits of Gibraltar. On the other hand, the finding of remains of several species of elephant in Sicily and Malta is almost proof positive of a second connection having been formed between Italy and Tunis. An elevation of the present bed of the sea a few hundred fathoms would bring about this result. The Mediterranean would then consist of two land-locked basins. But, doubtless, many of the other islands besides Sicily and Malta were united with the mainland, for otherwise it would be impossible to explain the distribution of several modern animals, the moufflon, for example, which is found in Sardinia, Corsica, Crete, and the mountains of Greece. Granting this connection between Africa and Europe, it appears more than likely that the principal disturbing element which reacted upon the Post -Pliocene European fauna, the great northern ice-sheet and the accompanying cold of the glacial period, rather than caus- ing the complete or sudden extermination of the receding fauna, compelled it to migrate over into regions of a more congenial cli- mate. That such was the fate of many of the forms there can be no reasonable doubt. The Africa'n continent thus became stocked with its existing fauna largely from the more temperate northern regions. But there is every reason to believe that .these same south- ward retreating forms were in great part primarily introduced into Europe from Africa, and over the same routes by which the later southerly migration was effected. Concerning the origin of the African fauna itself we possess little precise information. The paleontology and geology of the region are so imperfectly known that we possess as yet no basis for satisfactory deductions. The absence of sufficient data naturally renders uncertain all speculation relating to the late European fauna as well. It may be considered highly probable, however, that many of its characteristic elements have been derived from the region about India, where a considerable antiquity, extending back to the Miocene or early Pliocene period, is proved for at least a number of the more prominent types. Seve- ORIGINATION OF FAUNAS. 15 ral of the antelopes have related, and apparently ancestral, forms in the Miocene deposits of Greece (Pikerini), which also contain a form not very far removed from the giraffe (Helladotherium), and a species of true giraffe itself (Camelopardalis Attica), so that possibly a contingent of the African fauna may have been derived from this region. Whether the southern or Ethiopian portion of the continent was at one time since the introduction of the placental Mammalia completely severed from the northern part or not there are as yet no means for determining. That Madagascar at one time formed part of the continent is indisputably proved by the character of its fauna ; but that its subsequent isolation is of very ancient date is conclusively shown by the complete absence of all the more distinctive Ethiopian placental mammals. The few examples that have been cited in illustration of the appearance and disappearance of faunas are sufficient to show the character of the investigation that is open to the zoogeographer. While from the data that we now possess much can be done towards shaping our suppositions, it must be confessed that our knowledge is still much too limited to permit of very satisfactory conclusions being drawn therefrom. The principal danger that besets any in- vestigation in the direction here outlined arises from the very natural assumption that the greater antiquity in any one region over another of a given type of animal indicates its prior appearance there, and migration thence to one or more secondary regions. This assumption might be well founded if we were only half con- versant with the past paleontological histories of the regions under consideration ; but where at best our knowledge is still very imper- fect, as it is in the case of Africa, Asia, and South America, it would be, to say the least, highly injudicious. For what evidence have we that animal types not yet found, or dating back only to a com- paratively recent period, might not some day be turned up in abun- dance, and in deposits of such age as to completely overthrow any deductions that may have been based upon their supposed non- occurrence ? A single illustration of this kind will suffice. Pale- ontologists are in the habit of considering the camels a New World family, which by migration finally occupied the region which it now inhabits. This conclusion is based upon the circumstance that numerous cameloid forms (Pliauchenia, Procamelus, Protolabis, Poebrotherium) carry this line of animals back in the North Ameri- 16 GEOGRAPHICAL DISTRIBUTION. can continent to the early Miocene period, whereas such types are almost wholly wranting in the range of equivalent deposits of the Old World. Yet, if this is the true history of the family, it is certainly a surprising fact that the true camel itself (Camelus), which is entirely unknown on the American continent, should al- ready be found fossil in the Miocene (or older Pliocene) deposits of India. Nor is it at all unlikely that ancestral forms leading up to this type may yet be found in deposits of still older age hereafter to be discovered. II. Areas of specific distribution. — Generic distribution. — Distribution of families and orders. — Conterminous and discontinuous areas of distribution. IT is a fact of general observation that a given species of animal is so restricted in its range as to entitle the geographical area princi- pally occupied by it to be considered as its home. This home may be limited in its extent to a very narrowly circumscribed area, possibly not embracing more than a few square miles, or even less, or it may spread out to dimensions coextensive (or nearly so) with the conti- nental boundaries ; or, finally, it may comprise considerable portions of two or more continental areas combined. As examples of animals having a very restricted geographical distribution may be cited the Pyrenean water-mole (Myogale Pyrenaica), a small insectivore found only in a very few localities of the northern valleys of the Pyrenees, and a species of buschbok (antelope, Cephalophus Natalensis), whose habitat is the region about Port Natal, South Africa. Arc- tomys caudata, one of the Asiatic marmots, is confined to the ele- vated valley of Gombur, in India, and to heights exceeding 12,000 feet. Of birds, whose powers for self-distribution are much more fully developed than among mammals, we have equally pointed examples of localisation. The brown-and- white cactus-wren (Cam- pylorhynchus albibrunneus) is confined exclusively to the Isthmus of Panama, where its range is also somewhat limited ; the Bornean yellow-bulbul (Otocampsa mentis) has only been met with on the peak of Kina-Balu, in Borneo ; and the red bird-of-paradise (Para- disea rubra) only within the narrow limits of the island of Waigiou, lying to the northwest of New Guinea. The most remarkable in- stances of localisation are probably afforded by the humming-birds, several species of which would seem to be restricted respectively to the volcanic peaks of Chimborazo and Pichincha, in the equatorial 18 GEOGRAPHICAL DISTRIBUTION. Andes, and to the extinct crater of Chiriqui, in the province of Panama, Colombia. The Loddigesia mirabilis, one of the most beautiful of the TrochilidaB, has been observed thus far only at Chachapoyas, in the Peruvian Andes, and even there it occurs so rarely as to have been obtained but once during the period of forty years following its first discovery.1 Too much stress should not, however, be laid upon what would appear to be the absolute localisation of a species, since such sup- posed localisation is frequently only the expression of our defective knowledge in the premises. In the case of the famous South American oil-bird, or guacharo (Steatornis Caripensis), for example, which was for a long time considered to inhabit solely a cave near Caripe, in the province of Cumana, Venezuela, more recent research has revealed a comparatively broad area of distribution, which embraces Sarayacu and Caxamarca in Peru, Antioquia in Colombia, and the Island of Trinidad. The garden-mouse (Mus hortulanus), which for some twenty years was known only from the botanic gar- dens of Odessa, Russia, has been found in abundance in Kaschau and several other towns of Northern Hungary.2 So, likewise, in the case of the anthropoid apes of the genus Troglodytes, which were formerly supposed to be restricted to the western regions of the African continent, but which the more recent explorations of Schweinfurth, Yon Heuglin, and others have shown to inhabit East Central Africa as well. Of species having a very broad distribution — excluding such as have been transplanted through the agency of man — may be cited the African elephant, whose domain extends over the greater part of the African continent south of the Sahara Desert ; the tiger, whose habitat embraces the entire east and west extent of Asia, from the Caucasus to the Island of Saghalien ; and the ermine, which is found throughout the greater portion of the temperate and boreal regions of the Northern Hemisphere. The leopard ranges over entire Africa and throughout most of Southern Asia, having, with perhaps the exception of the common European wolf, whose identity with the various forms of American wolves is conceded by many natural- ists, and some of the smaller carnivores, the most wide-spread dis- tribution of any mammalian species. There is but little question as to the identity of the North American and European species of brown-bear, Arctic fox, glutton, ermine, weasel, elk, reindeer, and DISTRIBUTION OF SPECIES. 19 beaver,8 all of which have, consequently, a very extended range. The American panther or couguar (Felis concolor) inhabits the territory included between Canada and Patagonia, an extent cover- ing upwards of one hundred degrees of latitude, which probably represents the greatest north and south range of any mammal. As might naturally have been expected from the greater facili- ties for dispersion, we find many more marked instances of broad specific distribution among birds than among mammals. Indeed, when we consider with what apparent facility certain birds accom- modate themselves to the varying conditions of atmospheric pres- sure and climatic changes, and the readiness with v/hich they trav- erse broad expanses of the oceanic waters — e. g., the North Atlantic between Ireland and Labrador — it might at first sight appear as though there ought to be, at least in many cases, no absolute limit to their distribution ; yet, from our present knowledge, it may safely be affirmed that there exist but very few species of birds which are in any way cosmopolitan. The fish-hawk (Pandion haliaetus), with probably the most extensive range of any known bird, inhabits the greater portion of all the continents, with the possible exception of Australia, where its place appears to be sup- plied by a closely-allied (and by many ornithologists considered identical) species, the P. leucocephalus. Scarcely, if at all, less extensive is the range of the common peregrine falcon (Falco com- munis or peregrinus) and the barn-owl (Strix flammea), the former of which is distributed, according to Professor Newton, from " Port Kennedy, the most northern part of the American continent, to Tasmania, and from the shores of the Sea of Okhotsk to Mendoza, in the Argentine territory," and the latter, according to Sharpe, over the entire world, with the exception of New Zealand, and many island groups of Oceania, Malaysia, &c. The common American raven (Corvus corax) has, likewise, a very broad distribution, its range extending from Mexico into the far north, over the whole of Europe and Northern and Central Asia, as far east as the Island of Saghalien. The fishes present scarcely less well-marked examples of broad distribution ; but in such aquatic forms the physical conditions of the medium which they inhabit offer far less obstacles to a very general diffusion than are to be encountered in the case of terrestrial animals. The same holds true with other aquatic animals capable 20 GEOGRAPHICAL DISTRIBUTION. of self-locomotion, and, indeed, in the case of those pelagic forms whose dispersion or * ' migration " is less a matter of volition than the result of an interaction of extraneous physical causes there would seem to be no barriers set to a practically universal distribu- tion. But here, too, Nature has set a limit to the possibilities of migration, and, therefore, even among those lower forms which might be considered best adapted for withstanding the varying physical vicissitudes of their surroundings we meet with but very few species whose distribution might be said to be in any way cosmopolitan. The free-swimming pteropods, or winged -Mol- lusca, and medusoids, although exhibiting individual examples of very broad distribution, are still more or less restricted specifi- cally to well-defined oceanic areas, whose boundaries may in a measure be dependent upon the prevalent surrounding water- currents. Shells of the Spirula Peronii, a member of the two-gilled order of cephalopods, are met with almost all over the oceanic bor- ders, as well in the temperate as in the tropical zones, but, owing to the extreme rarity of the animal itself, which has been observed, perhaps, but a half-dozen times, it is impossible to say what the exact, or even approximate, range of the species is, and, conse- quently, of how much of the area of the distribution of the shell it partakes. The common form of argonaut (Argonauta argo) is found in the tropical parts of the Atlantic, Pacific, and Indian oceans, and in the Mediterranean Sea, and it has been met with as far north in the Atlantic as the New Jersey coast, and as far south as the Cape of Good Hope. The animal might, therefore, be said to be almost cosmopolitan. It may be laid down as a fundamental law in geographical dis- tribution that the areas inhabited by a given species are continuous with each other; in other words, we do not find, except at rare intervals, and under peculiar circumstances, the same species of animal inhabiting distantly-separated localities, in the interval be- tween which no individual of the species is to be met with. Thus, in the entire range of the leopard there occurs no district of any significance where the animal may not be confidently looked for, and which would negatively tend to render its distribution discon- tinuous. And the same may be said of the hundred or more degrees of latitude prowled over by the couguar, an animal whose home is at one place the lowland forests, at another the elevated AEEAS OF HABITATION. 21 mountain plateaus, and at a third the grassy savannas and rolling plains. Naturally, in the case of such animals as are dependent for their existence upon certain physical peculiarities of their en- vironment, or upon particular conditions of food and climate, we shall meet with local areas scattered through the region of dis- tribution of a given species where no individuals of that species are to be met with, an apparent discontinuity being thus pre- sented. For instance, such denizens of the forest as the South American monkeys and the sloths will but very exceptionally be found anywhere else than in their forest homes, and, therefore, the partial destruction of this forest, or its invasion by a grassy savan- na, will tend to render the "home " of those animals discontinuous. Possibilities of such, or a similar, discontinuity may likewise arise in the case of the animals of the plains, marshes, and deserts, since the physical aspects of the earth's surface are constantly subjected to vicissitudes of greater or less magnitude, and, as a matter of fact, we find numerous instances where, in an extensive range, particular animals are restricted in their habitats to certain favoured spots or localities. But in all or most of such instances a former, and comparatively recent, continuity of area, or possibility of mi- gration from one locality to another, can be proved. The chamois, whose range embraces the entire east and west extent of Southern Europe, is found almost exclusively on the higher mountain sum- mits— the Pyrenees, Alps, Carpathians, Caucasus, and the moun- tains of Greece — and would appear, therefore, to occupy several widely-removed habitats. But there can be no reasonable doubt that the peculiar distribution of this animal is the outcome of migration from a central home. The hippopotamus is found in the Nile, Niger, Senegal, and most of the larger rivers of South Africa, between which stretch vast areas where no individuals of the animal have ever been found — regions untenantable by reason of their aridity; but here, as in the case of the chamois, there can be no doubt that a migration or diffusion did take place at a time when the physical aspects of the country were favourable for such a dispersion, and were, consequently, different from what they are at present. One of the most remarkable instances of areal dis- continuity among mammals is that exhibited by the variable hare, whose home, in the Old World, is Eurasia north of the fifty-fifth parallel of latitude. The animal reappears, after skipping the low- 22 GEOGRAPHICAL DISTRIBUTION". lands of Central Europe, in the Pyrenees, Alps, and the Bavarian Highlands, and again in the Caucasus, the last region isolated by fully one thousand miles of non-inhabited country. Equally strik- ing examples were supposed to be aiforded by the fresh-water seals of Lake Baikal and the brackish-water species of the Caspian, which were considered to be identical with the northern Phoca foetida and P. vitulina respectively, but more careful study has shown this identification to be erroneous.4 The critical studies made by Mr. Seebohm of the Central and East Asiatic faunas have disclosed a number of extraordinary instances of discontinuous habitation among birds. One of these is exemplified in the case of a South European variety of the common marsh-tit (Parus palustris), which reappears in an undistinguishable guise in China, although in an intervening tract of some four thousand miles (east of Asia Minor) the variety is entirely wanting, being replaced by one or more closely related forms. Ceryle guttata, a spotted king-fisher, appears to be confined to Japan and the Himalaya Mountains, being com- pletely wanting in China ; and the same is true of a species of crested eagle (Spizaetus orientalis), with the exception that its range embraces the Island of Formosa. Similarly, we have two species of birds, the rufous-breasted fly-catcher (Siphia superciliaris), and the Darjeeling wood-pigeon (Palumbus pulchricollis), which are absolutely confined to the Himalayas and the Island of Formosa. But while individual cases of species inhabiting discontinuous areas do present themselves, they are of comparatively rare occur- rence, and the general law of regional continuity may be recognised. In a region occupied by a given species of animal there is usually an area which is par excellence more thickly inhabited than any other, and which may, consequently, be termed the "metropolis" of that species. From this metropolis there is in most cases a radial distribution of the individuals of the species, with a thinning out towards the periphery. Distinct species of the same genus rarely have coincident geographical distributions ; in other words, they rarely occupy precisely the same areas, but more generally these areas, if at all continuous, overlap each other to a greater or less extent. This fact is beautifully exemplified in the case of the American hares, which are represented by some eleven species, and about as many well-marked varieties. Commencing at the far north, we have the polar or variable hare (Lepus variabilis or L. OVERLAPPING AREAS. 23 timidus, var. Arcticus), whose range extends from the Arctic coast southward to Newfoundland, and in the interior to Fort Churchill, on Hudson's Bay. Along its southern confines it meets and slightly overlaps the boundaries of the northern varying hare (L. Ameri- canus), which, in its several geographical varieties, is distributed from the Barren Grounds in the north southward to a zone which corresponds generally with the isotherm of 52° F. On the Atlantic coast region, the southern limit of this species appears to be Con- necticut ; along the line of the Appalachian highlands, Virginia (or possibly North Carolina) ; and in the Rocky Mountain region, New Mexico. Lepus Americanus is found throughout the northern parts of nearly all the northern tier of States interposed between the Missouri and the Atlantic coast, and over the greater portion of this vast area of distribution, which is continued westward to the Pacific, it forms the sole representative of the family. In the south its habitat overlaps the range of the wood-hare (L. sylvati- cus), which, in its several varietal forms, is distributed along the Atlantic coast from Southern New England to Yucatan. "West- ward, the range of this species extends quite, or very nearly, to the Pacific, keeping, however, to a course south of the isotherm of 45° F. The prairie-hare (L. campestris) is found in the interior region, principally between the isotherms of 56° and 3G°, its range being consequently overlapped on the north by that of Lepus Americanus, and on the south by L. sylvaticus. In the South- eastern United States there are two distinct [species, L. palustris and L. aquaticus, which are almost exclusively confined to the marshy lowlands, and whose habitats, extending to Yucatan on the south, are partially comprised in those of the wood-hare and jackass-hare (L. callotis), the last a western species, whose range descends into the arid interior of the Republic of Mexico. Finally, we have a solitary species of South American hare (L. Brasiliensis), whose reputed range embraces a considerable portion of the con- tinent from Patagonia to Panama, continuing thence into Central America. 6 It frequently happens that the boundaries of a given species are sharply defined against those of another, stopping just where the others begin, and where, consequently, no overlapping takes place. Such cases of specific limitation occur where natural obstacles to a free migration are suddenly encountered, as where mountain or 24 GEOGRAPHICAL DISTRIBUTION. water barriers project themselves into a given region. Thus, it will not rarely be found that a genus of animals is represented by one or several species on one side of a long mountain-slope, and by entirely distinct species on the other. And, similarly, distinct species of a genus may be encountered on opposite sides of a river-bed, although instances of such a nature among the higher animals are probably not .of very frequent occurrence. Mr. Wallace cites the case of cer- tain species of Saki monkey (Pithecia), found on either side of the Amazon River, whose range either southward or northward appears to be limited by that stream. The same naturalist instances among birds species of jacamar (Galbula) and trumpeter (Psophia) which exhibit a similar limitation, particularly the latter, where five dis- tinct species are relegated to as many distinct, but contiguous, geo- graphical areas, separated from each other by the Amazon and some of its tributaries (Negro, Madeira, Tocantins). Of about twelve species of armadillo (separated by some naturalists into several dis- tinct genera), most of which are inhabitants of Brazil, it would seem that not a single species is common to Brazil and the Argen- tine Republic, or the Argentine Republic and Paraguay, the Parana River, with its tributaries, evidently forming an insurmountable barrier to the passage of this animal. The Uruguay River appears in the same way to limit the eastward progression of the viscacha (Lagostomys trichodactylus), an animal allied to the chinchilla, although, as has been pointed out by Mr. Darwin, the trans- Uru- guayan plains are fully as well adapted to the animal as those of its native home. Just as the boundaries of land-animals are in many instances defined by the dominant river-courses, so, in a like manner, but in a much more marked degree, the domains of fresh-water forms are frequently circumscribed by the land surfaces bordering the waters inhabited by them. This fact is beautifully exemplified in the geo- graphical distribution of two American families of fluviatile mol- lusks, the Strepomatida3, or American melanians, and the Unionida3, the fresh-water mussels, where the species of several genera, at least in the Southern United States, are restricted in their habitats to certain individual streams, to the exclusion of all others. In- deed, it would appear that even in such aquatic forms a large river may constitute an almost insuperable barrier to migration, as is shown in the case of the Strepomatidae by the Mississippi (south of DISTRIBUTION OF JAYS. 25 the line of the Ohio River), which but very few members of the family have been able to surmount. According to Tryon, only one species of the family, Goniobasis sordida, is positively known to be common to the region on both sides of that great stream.0 Probably no group of animals, as Mr. Wallace well observes, illustrates in a more striking manner the extreme features of specific distribution than the true jays, birds of the genus Garrulus. About fourteen species are recognised by ornithologists, whose combined domain embraces the entire east and west extent of the continent of Eurasia, from the Bay of Biscay to the Sea of Okhotsk, and also in- cludes the continental British Isles on the west, and the Japanese group on the east. Most of these species occupy independent areas of their own, or areas which but barely overlap on their contiguous borders. Thus, the common jay (Garrulus glandarius) inhabits the greater portion of the semi-continent of Europe, ranging from the Barbary States in Africa northward to about the sixty-fourth paral- lel of latitude (in Scandinavia and Russia), and east to the Ural Mountains. Along its southern border it meets the Algerian jay (G. cervicalis), a distinctly-marked species, and one having but a very limited range. On the southeast, again, its confines meet those of the black-headed jay (G. Krynicki), which occupies a somewhat cir- cular district extending some distance on all sides of the Black Sea. Contiguous with this last is the region inhabited by the Syrian jay (G. atricapillus), a species very closely allied to the preceding, whose domain extends through Syria, Palestine, and Southern Persia. North of this we have the limited area occupied by the Persian jay (G. hyrcanus), which has thus far been found only on the Elbruz Mountains. In an almost direct line east of this region, but separated from it by a considerable area where no jays are to be met with, we pass consecutively over the haunts of the black- throated jay (G. lanceolatus), from the Northwestern Himalayas, the Himalayan jay (G. bispecularis), from the Himalaya Mountains to the eastward of Cashmere, the Chinese jay (G. Sinensis), from South and Central China (and, occasionally, Japan), and the Formosan jay (G. Taivanus). The home of the Burmese jay (G. leucotis) adjoins that of the Himalayan jay on the southeast. North of the belt occupied by the species of southern jay we have a vast region — the desert area of Central Asia, with Thibet, Turkestan, Mon- golia, and Gobi— throughout the greater part of which no jays 26 GEOGRAPHICAL DISTRIBUTION. have as yet been discovered. Bounding this area on the north, and extending from beyond the Ural Mountains (Kazan) to the northern island of the Japanese group, there exists an almost continuous and comparatively broad belt which is tenanted throughout its entire extent, except where it overlaps the habitat of the common Euro- pean G. glandarius, by a solitary species, known as Brandt's jay (G. Brandti). Finally, in the southern island of Japan there are found two species, G. Japonicus and G. Lidthi, the former of which, sin- gularly enough, is the species which is most nearly allied to the common European jay, although separated by the greatest distance from it.7 Generic Distribution. — The laws governing specific distribu- tion are in considerable measure likewise applicable to the dis- tribution of genera. Thus, we have genera that are restricted to very limited areas, and, as a necessary consequence resulting from specific distribution, those whose areas are coextensive with con- tinental boundaries, or embrace portions of two or more continents ; and, again, we have genera of a given family which occupy con- tiguous, overlapping, or discontinuous provinces. The localisation of a genus to an exceptionally narrowly circumscribed area, such as we have seen in the case of the species of humming-birds of the volcanic peaks of South America, can almost necessarily ob- tain only there where the number of species belonging to the genus is also exceptionally limited, or, more nearly, when the genus is coextensive with a single species. Potamogale, which comprises the single species P. velox, a singular otter-like insecti- vore of the west coast of Africa, appears to ,be confined to the region included between Angola and the Gaboon ; Choeropsis, with the single species C. Liberiensis, an animal closely allied to the true hippopotamus, inhabits, as far as is yet known, only the wilds of Liberia ; and, likewise, the singular carnivore constituting the genus Ailurus (A. fulgens) has been met with only in the Southeastern Himalayas. Instances of restriction are much more nu- merously presented in the case of insular than of continental faunas, whether the examples be taken from the class of birds or mam- mals. Genera of very broad, or almost world-wide distribution, are of frequent occurrence, both among the lower and higher animals. Among the latter, in the class of birds, we have numerous examples GEKEKIC DISTRIBUTION". 27 among the swimmers, waders, and birds of prey, whose range covers the greater extent of the primary divisions of the earth's surface, and which may, consequently, be said to have a cosmo- politan distribution. Generic groups with a nearly world-wide distribution among the Mammalia are of much rarer, although of not exactly infrequent, occurrence, and if the Australian dingo, a species of wild dog, be not considered indigenous to the country which it inhabits, there would appear to be, if we except the bats, not a single altogether cosmopolitan genus among that class of ani- mals. Leaving out of consideration the continent of Australia, whose mammalian fauna is deficient in nearly all the orders of the class, we have a considerable number of genera whose range com- prises the greater portion of the habitable globe.* Thus, the mem- bers of the genus Felis (cats) are spread throughout the entire expanse of the continents of both the Eastern and the Western Hemisphere, through regions the extremes of whose temperature may be measured by probably no less than 225 degrees of the Fahrenheit scale. The genus Canis (dogs) has an almost equally broad distribution ; and the same range is exemplified in the case of the weasel genus (Mustela). Ursus, the bear, is met with throughout the greater part of the Northern Hemisphere, and in the continent of South America the genus has one or more rep- resentatives whose habitat is situated considerably to the south of the Equator.f The genus Cervus (deer), in its broader sense, has representatives in both North and South America, Europe, and Asia, with a very limited number of species (fallow-deer, stag) in Africa north of the Sahara. Discontinuous generic areas, like specific areas, are of com- paratively rare occurrence. Among the most remarkable instances of such discontinuity we have that exhibited in the case of the * The only placental animals indigenous to the Australian continent, if we exclude the rather doubtful dingo, which is by most naturalists considered to have been introduced by man, are the Cheiroptera (bats) and Eodentia, the latter represented by the family of mice (Muridse). The implacental mam- mals— kangaroos, wombats, phalangers — have, on the other hand, an extraor- dinary development. t The solitary species of bear inhabiting the continent of Africa appears to be confined to the Atlas Mountains ; it constitutes the genus Helarctos of some authors (H. Crowtheri). 28 GEOGRAPHICAL DISTRIBUTION. genus Myogale, the water-mole, already referred to, which em- braces two species, one of which, M. Pyrenaica, is an inhabitant of the northern valleys of the Pyrenean chain of mountains, and the other, M. Muscovita, the plains of Southeastern Russia skirting the Don and Volga rivers. The pikas (Lagomys), small rodent animals having a rather near relationship with the hares, which are exten- sively distributed along the upper mountain heights from the Ural to Cashmere and the eastern extremity of Siberia, have a single outlier in the Rocky Mountains of North America. The members of the genus Capra — the goats and ibexes — occupy disjointed patches of territory in Europe, Asia, and Africa, mainly confined to the elevated mountain regions, such as the Pyrenees, the Sierras of Spain, the Alps, Caucasus, Himalayas, &c., the intervals between which are deficient in the wild or indigenous representatives of the genus. A similar discontinuity is exhibited in the case of the snow-partridges of the genus Tetraogalius, a bird likewise partial to the elevated mountain-slopes. Numerous other instances of birds occupying discontinuous areas may be cited, and they appear particularly noticeable among families of a more or less tropical habit. Such, for example, are the ja$anas (Parra), which inhabit the tropical regions of both the Old and the New "World, the simi- larly distributed flamingoes (Pho3nicopterus), the wood-ibises of the genus Tantalus, the gerontics (Geronticus), and the marabou storks (Mycteria). Among perching birds a most remarkable in- stance of generic discontinuity has been cited by Wallace in the case of the blue magpies (Cyanopica), which comprise two species, one of which, C. Cookei, inhabits the Spanish Peninsula, and the other, C. cyanus, Eastern Siberia, Japan, and North China, the habitats of the two being removed from each other by an interval of fully 5,000 miles. Still more marked is the case of the bluebirds constituting the genus Sialia, all of whose members, with one exception, inhabit temperate and tropical America ; a solitary form, Sialia (Grandala) coalicolor, singularly enough, crops up again among the Himalaya Mountains, and eastward throughout the mountainous region sepa- rating China from Thibet.8 The most remarkable instance of a mammalian genus occupying two widely - removed areas is fur- nished by Tapirus, the tapir, several species of which are natives of the South American continent, and one, very distinct from the others, of Malacca and Borneo, the group of animals, therefore, DISTRIBUTION OF FAMILIES. 29 appearing at localities separated from each other by nearly half of the earth's circumference. Distribution of Families.— The restriction of families to cer- tain local areas is of comparatively rare occurrence, an almost neces- sary consequence of the number of species and genera of which they are in most cases composed. Among mammals the Cheiromydae, with one genus and one species, the aye-aye (Cheiromys Madagas- cariensis), are confined exclusively to the Island of Madagascar ; the Protelidae, likewise consisting of but a single genus and species, the aard-wolf (Proteles Lalandii), an animal in several respects inter- mediate between the cats and dogs, and considered by some as representing a greatly modified form of hyena, are confined to the extra-tropical regions of South Africa. Occupying pretty nearly the same region, and confined to it, are the Chrysochloridae, or golden-moles, with a single genus and about five species. The Ailuridae, a group of animals having their nearest allies in the coatis and bears, and consisting of one or two species, appear to be restricted to the forest region of Eastern Thibet and the Eastern Himalaya. Among the class Aves we have likewise families that are restricted both as to the number of species comprised by them and the region which they inhabit. The Paictidae, a group of birds, considered by some ornithologists to have affinities with the Ameri- can ant-thrushes (Formicariidfce), and by others with the Old-World pittas, consist of a single genus and two or more species, both of which are confined to the Island of Madagascar. Here, also, ex- clusively belong the Leptosomidae, birds allied to the cuckoos and rollers. The Apterygidae, with one genus (Apteryx) and four species, are strictly confined to the two larger islands of New Zealand ; the Drepanidae, with some four or five genera and ten species, are re- stricted to the Sandwich Island group ; and, finally, the paradise- birds, excluding the bower-birds, which are classed together with them in one family by some ornithologists, with about eighteen genera and thirty species,8* are almost entirely confined to New Guinea and the surrounding islands, only four representatives of the group finding their way into the neighbouring continent of Australia. Mr. Wallace has emphasised the very remarkable case of localisation presented among reptiles by the Uropeltidae, or rough -bellied, bur- rowing snakes, all of whose members appear to be strictly con- fined to Ceylon and the adjacent parts of the Peninsula of India.9 30 GEOGRAPHICAL DISTRIBUTION. Families with restricted ranges, like genera and species, are of infrequent occurrence, broad distribution being with them the rule rather than the exception. Nevertheless, owing to the peculiarity isolated position of the Australian fauna, there are among the land Mammalia only two families which can lay claim to being strictly cosmopolitan. These are the mice (Muridae) among rodents, and the Vespertilionidae among bats, the former being universally dis- tributed throughout the globe, if we except some of the island groups of Australasia. The Vespertilionidse have representatives almost everywhere, being apparently limited, as stated by Wal- lace, only by the necessities of procuring insect food. Among birds, examples of practically cosmopolitan families are presented by the thrushes, warblers, crows, swallows, king-fishers, goatsuckers, and pigeons. The hawks, owls, ducks, and gulls are cosmopolitan par excellence, being found in almost every habitable locality throughout the globe, whether on the mainland or on the most distantly removed oceanic islands. The extensive family Fringil- lidse (finches, buntings), as now generally constituted by ornitholo- gists, with upwards of seventy genera and five hundred species, appears to have no representative in Australia, all the finch-like birds of that continent belonging to the family of the weavers (Ploceidae). As with genera and species, so likewise in the case of families, we have numerous instances of groups occupying discontinuous areas. In the class of Mammalia, for example, the swine (Suidse), which are so extensively distributed throughout both the tropical and temperate regions of the Old World, have no representatives in the New World north of about the thirty-fourth parallel of latitude — the Ked River, in Arkansas — although they have two species (of peccary) in the region south of that line. The Orycteropodidse have a solitary representative in the Cape District, the aard-vark, or Cape ant-eater (Orycteropus Capensis), and another In the interior of Northeast Africa and in Senegal, the form occurring in the latter region being possibly a third species.10 The tapirs, constituting the family Tapiridse, have, as already stated, their representatives on opposite sides of the globe, one species inhabiting the Malay Peninsula and some of the adjacent islands, and the four or five others the tropical forests of Central and South America. The chevrotains, or deer-like animals of the family Tragulidae, abound DIVIDED FAMILIES. 31 in India and some of the islands of the Malay Archipelago, where they constitute the genus Tragulus ; a solitary representative of the same family, but belonging to a distinct genus (Hyomoschus), is a native of West Africa. The anthropoid apes (Simiidse) are repre- sented in Western (and probably also East Equatorial) Africa by one or more species of gorilla and chimpanzee (Troglodytes), which are almost exclusively confined to the forest region. The form most nearly allied to these man-like apes, and belonging to the same family, is the orang (Simia satyrus), which, as an inhabitant of the islands of Sumatra and Borneo, is encountered after an interval of not less than seventy degrees of longitude. Inhabiting the same region, but with a northward extension to China, and westward to Assam (south of the Brahmaputra River), we find the members of the genus Hylobates, the gibbons. Probably the most striking example of a divided family is furnished by the Camelidse, which in the Old World are represented by the genus Camelus, with two species — the dromedary and the Bactrian camel — whose habitat extends from the Sahara through the desert regions of Western and Central Asia to Lake Baikal ; and in the New World by the genus Auchenia (the llama, alpaca, vicufia, and guanaco), with about four species, all of them restricted to the mountainous and desert regions of Western and Southern South America. We have here, therefore, a family which is not only divided by a vast ocean and the greater mass of two continents, but the members of which, in one hemi- sphere, inhabit the region north of the Equator, and, in the other, the region south of it. Instances of divided families among birds occur as in mammals, although probably to a less marked extent, owing naturally to their increased facilities for dispersion ; such division obtains more espe- cially among the so-called " tropicopolitan " forms, or those whose homes are properly the region of the tropics, or that immediately adjoining it. The flamingoes (Phcenicopteridae), consisting of a solitary genus and about eight species, are about equally distrib- uted as to the number of species throughout the warmer regions of America, Africa, and Asia, some of the forms extending their range to a considerable distance within the bounds of the Tem- perate Zone, as in Southern Europe and South America. The trogons (Trogonidae), comprising many of the most beautifully- arrayed of birds, and with upwards of forty species, are more 32 GEOGRAPHICAL DISTRIBUTION. strictly confined to the tropical regions of the earth's surface, but few forms being found beyond the limits of that zone. They are fairly abundant in the forest region of South America, ranging from Paraguay to Mexico, and less so in South and Southeast Asia, and some of the islands of the Malay Archipelago. In Af- rica the family is represented by but two species. The Psitta- cidae among parrots furnish us with another good example of a divided family, whose members are to be found only in the two great southern continents, Africa and South America, and in some of the adjacent islands. Still more remarkable is the case of the ostriches, of which there are two species (of the genus Struthio) pertaining to the desert regions of Africa and Western Asia (Arabia and Syria), and likewise two (of the genus Rhea, sometimes placed in a distinct family) belonging to temperate South America, whose range extends from Patagonia to the confines of Brazil. Among reptiles similar instances are presented by the tropi- copolitan groups. Thus, we have the Crocodilidae inhabiting the tropical waters of both the Eastern and Western Hemispheres. The Pythonidse, or giant constricting serpents — boas, anacondas, pythons — are, with the exception of the Californian genus Charina, some- times referred to this family, distinctively tropical, but they have representatives in the South American continent, in Africa, Asia, Australia, and in several of the continental and oceanic islands. The family of iguanas (Iguanidae), comprising upwards of fifty gen- era and some three hundred species, is almost distinctively Ameri- can, being distributed from about the fiftieth parallel of south latitude, in Patagonia, to the Canadian boundary-line on the north. No member of the family is known from eith'er of the continents of Eurasia or Africa, yet the family crops up again in a solitary genus — Brachylophus — in the Feejee Islands, and two (doubtfully placed) genera have also been described from Madagascar and Aus- tralia. Distribution of Orders. — The principal features of geographical distribution exhibited by species, genera, and families repeat them- selves in a measure in the case of the higher groups of the animal kingdom known as orders. Very narrowly circumscribed areas of habitation, at least among the orders of higher animals, do not exist ; broad distribution is the rule. Among mammals the most marked instances of semi-localisation, if so it may be termed, are DISTRIBUTION OF OR DEES. 33 furnished by the Monotremata, comprising the two families of duck-bills and echidnas, both restricted to Australia and the Isl- and of Tasmania, and the Hyracoidea, an order consisting of two genera, Hyrax (the coney) and Dendrohyrax, and about a dozen species, all of which are restricted to the continent of Africa and the immediately adjoining parts of Asia (Syria). The only orders of terrestrial mammals which can lay claim to being cosmopolitan are the Cheiroptera (bats) and Rodentia, none of the other orders, except the Marsupialia — unless the dingo, as a member of the Car- nivora, be considered indigenous to the continent it now inhabits — having any representatives in Australia. Among birds we have no instance of an order being restricted to the limits of a single con- tinent. Among reptiles the Crocodilia are almost entirely confined to the tropical and sub-tropical regions, and occur in both the East- ern and Western Hemispheres. The Ophidia (serpents) have what might be called a world-wide extension, although no member of the order has been met with farther to the north than the Arctic Circle. The order Anura (frogs and toads) among amphibians is very nearly cosmopolitan ; the Urodela, on the other hand, com- prising the tailed forms, such as the newts, salamanders, &c., are almost strictly confined to the Northern Hemisphere, a few only of its representatives passing through Central America as far south as Colombia. The entire class of the Amphibia (as indigenous forms) is absent from the vast majority of oceanic islands — New Zealand, New Caledonia, and the Andaman Islands, and possibly the Solo- mon and Seychelles groups, almost alone, according to Darwin, presenting exceptional instances. The marsupials afford a remarkable example of a comparatively large order of animals occupying widely separated and discon- tinuous areas. With the exception of the opossums, of which there are two genera and about twenty species, confined to the two continents of America, and more particularly to the tropi- cal regions of these continents, all the members of this peculiar and lowly-organized order of animals are strictly limited in their range to the Australian continent and its dependent islands, and some of the islands of the Malay Archipelago. In all the broad intervening region — Europe, Africa, and Asia — no representative of the order is to be met with. The Edentata — ant-eaters, armadillos, &c. — are largely confined to the tropical and sub-tropical regions 34 GEOGRAPHICAL DISTRIBUTION. of South America, Asia, and Africa, and are almost completely absent from the vast northern tracts which spread out towards the polar confines, and tend to bring together the terrestrial areas of the Old and the New World. The order is entirely wanting in Europe, and nearly so in North America, the genus Tatusia, an armadillo, alone penetrating within the boundaries of the last into the State of Texas. In Asia no member of the order is found to the north of the Himalaya Mountains. A remarkable example of discon- tinuous habitation among birds is furnished by the Struthiones, or ostrich-like birds, whose members are distributed throughout con- siderable reaches of tropical and sub-tropical South America, Africa, Asia, and Australia, some of the Australian islands, and New Zea- land, and are entirely wanting in Europe and North America. It is a singular circumstance in connection with the distribution of the birds of this very limited order that two genera, so closely allied as are Rhea and Struthio, should occupy areas so distantly removed from each other as Africa and South America. The Psittaci, or parrots, inhabitants of both the New and the Old World, may likewise be considered as being preeminently tropical and sub-tropical, for although a few examples are found whose range in the Southern Hemisphere ascends to the fifty-fourth degree, yet the true home of the order is located in the zone embraced be- tween the thirty-fifth parallels north and south of the Equator. Being absent from Europe and the greater portion of the continent of North America, the distribution of the order is necessarily dis- continuous. III. Conditions affecting distribution. — Climate. — Food-supply. — Barriers to mi- gration.— Migrations of mammals and birds. — Dispersal of insects and mollusks. Of the Conditions which affect or limit Distribution among Animals. — Climate. — It is a common belief that the principal fac- tor limiting or regulating the distribution of animals is constituted by climate ; in other words, certain groups of animals are associated with certain grades or conditions of climate, beyond the reach of whose interacting influence they could no longer maintain an exist- ence. Thus, among quadrupeds, the elephant, camel, and tiger are popularly associated with the hottest climates of the earth's surface ; the reindeer and moose with climates of equal, but opposite, sever- ity. And, similarly, among birds, the ostriches and hummers are considered to be particularly indicative of hot or tropical climates, and the auks, guillemots, puffins, and penguins, as products of the cold northern or southern climes. That climate does regulate dis- tribution, or impose a bar upon the migration of certain forms of life, there can be no manner of doubt ; but that it does not exercise the paramount influence that is generally attributed to it there can likewise be no question. Taking, for example, some of the in- stances that have just been mentioned as indicating the supposed association between animal distribution and given conditions of climate, we find that the tiger, while its home, par excellence, may be considered to be the hot districts of India and the Indian Archi- pelago, is in no way restricted in its range to those regions, or to regions having at all a similar climate. Thus, the animal is found in the elevated regions of the Caucasus and the Altai chain, and in the Himalaya range its footprints are not infrequently found im- pressed in the fields of snow. It is a permanent inhabitant of the 36 GEOGRAPHICAL DISTRIBUTION. cold plains of Manchuria and the Amoor region, as well as of the plains lying north of the Hindu-Kush, in Bokhara, prowling about even in winter along the icy margins of the Aral Sea. As a matter of fact, the range of the tiger extends to about the fifty-third paral- lel of north latitude — or what corresponds to the position of Lake Winnipeg, in British America — in the neighbourhood of Irkutsk and Lake Baikal. Nor can this northern range be taken to represent the range of simply stray individuals, since in the region of South- east Siberia traversed by Radde that traveller affirms that tigers were uncommonly abundant.11 Although at the present time the lion is confined almost exclu- sively to the tropical and sub-tropical regions of Africa and Asia, there can be but little doubt, as appears from the writings of Herod- otus and Aristotle, that as late as the beginning of the historic period that animal still inhabited in Europe a region lying as far north as about the fortieth parallel of latitude — or what corre- sponds in position to the State of Pennsylvania — namely, the region of Thessaly, in Greece. And even at the present day the Tunisian lion is occasionally found in the neighbourhood of the thirty-seventh parallel of north latitude, and until recently the Cape lion was abundant in or about the district of the Cape, extending to the thirty-fifth parallel of south latitude. Although the climate of these latitudes in Africa is of an unusually mild character, yet there are sudden changes of temperature, as between day and night, which may be likened to the changes in the temperature between the summer and winter climates of more temperate regions. We are informed by travellers that in the Kalahari Desert and other dry open districts of South Africa the nights are frequently unpleasant- ly cool, or even cold, the free and rapid radiation of heat from the soil not rarely being accompanied by a freezing of the surface. The formation of ice in the Desert of Sahara is, likewise, not exactly of exceptional occurrence, but in that region of the African continent, except on its immediate borders, lions are only rarely met with. That a restriction to warm climates is likewise not the case with the elephant is almost conclusively proved by the readiness with which, in the Roman period, these animals were made to pass the barrier offered by the lofty Alpine chain. Still more indisputable evidence on this point is, however, afforded by the habits of the Indian elephant, which appears to be equally at home among the CLIMATIC INFLUENCES. 37 cool mountain heights as amidst the hot and jungly lowlands. In Ceylon, according to Sir Emerson Tennent, "the mountain-tops, and not the sultry valleys, are his favorite resort. In Oovah, where the elevated plains are often crisp with the morning frost, and on Pedro-Tella-Galla, at the height of upwards of eight thousand feet, they are found in herds, whilst the hunter may search for them without success in the jungles of the low country. No altitude, in fact, seems too lofty or too chill for the elephant, provided it affords the luxury of water in abundance; and, contrary to the general opinion that the elephant delights in sunshine, he seems at all times impatient of its glare, and spends the day in the thickest depths of the forest, devoting the night to excursions, and to the luxury of the bath, in which he also indulges occasionally by day. "w Mr. Johnston, during his recent explorations of the Kilimanjaro region, encountered elephants, together with buffaloes, and one or more spe- cies of antelope (kudu), at an elevation of thirteen thousand feet.12' The camel is an animal popularly associated with the burning desert regions of Africa and Asia, yet the two-humped or Bactrian species is found throughout the greater portion of Mongolia and Chinese Tartary, in the mountain region as well as in the lowlands, lying between the fortieth and fiftieth parallels of latitude, and it extends its range even considerably beyond the fiftieth parallel into Siberia, as along the borders of Lake Baikal, where it appears to pass the winter season without discomfort. It is a fact worthy of note that the only other existing representatives of the camel family — the llama and llama-like animals of the New World— are strictly adapted to a rigourous winter climate, as is shown by their partiality to the highly-elevated tracts of the South American Andes. The same adaptability to different extremes of climate likewise presents itself in the case of many of the so called Arctic animals. The reindeer, while it habitually prefers for its home a region that en- joys a more or less rigourous climate, and where the soil is for the greater part of the year covered with snow, does not appear to be impatient of the summer heat of comparatively low latitudes, as is proved by the circumstance that in the various zoological gardens of Central Europe it not only develops in good condition, but also breeds freely. Indeed, its restriction to the high northern latitudes appears to be in no way dependent on considerations connected with either cold or snow, but merely upon the presence there in the 38 GEOGKAPHICAL DISTRIBUTION. greatest abundance of its particular food, the reindeer-moss and various lichens, without which it seems incapable of nourishing. There can be little doubt that were individuals of the reindeer transplanted to an elevated mountain region, such as the European Alps, for example, where their own proper nourishment would be again met with, they would thrive very nearly, if not fully, as well as in their true homes north of the fifty-fifth or sixtieth parallel of latitude. Indeed, even in their northern haunts the animals, at least as is shown by the American species or variety, would seem to be impatient of too great a cold, since in the winter they seek the inner recesses of the forests for protection. Turning now to the class of birds, we find that similar illustra- tions of climatic adaptation present themselves. Thus, the usually considered "tropical " or "equatorial " humming-birds are in reality not such at all. While it is true that by far the greater number of species belonging to this family are found within the region em- braced within the tropics, yet the range of the family extends all the way from Cape Horn (Eustephanus galeritus) to Sitka (Selas- phorus rufus), or over a territory covered by no loss than one hun- dred and fifteen degrees of latitude. And even among the strictly tropical forms many of them extend their range to the limits of perpetual snow, some remaining in the cold region permanently. The Oreotrochilus Chimborazo and O. Pichincha have their abode in the equatorial peaks indicated by their respective specific names at an elevation of no less than sixteen thousand feet — or higher than the summit of the Mont Blanc — in a world of almost perpetual snow, hail, and sleet.13 In fact, the elevated Andean slopes are much more thickly visited by humming-birds than the deep low- lands, no matter how luxuriantly these last may be clothed with vegetation. The ostriches constitute another group of animals whose habitat is popularly associated with the burning deserts of the Torrid zone. While it is unquestionable that these birds do delight in just such districts, it may yet be doubted whether the matter of climate has very much to do with the selection of a region, since ostriches are, or have been until recently, equally abundant in all parts of the African continent, in the high table-lands as well as in the low- lands, from Algeria to the Cape, and from the east to the west coast, where the suitable desert conditions present themselves, and CLIMATIC INFLUENCES. 39 where, consequently, as has already been stated, the differences between the temperature of night and day are excessively marked. In the desert region of Western Asia — Persia and the Valley of the Euphrates — the bird ranges or ranged as far north as about the thirty-fifth parallel of latitude, and, indeed, it is not exactly im- probable, as has been maintained by Vambery,14 that even at the present day it exists in limited numbers along the shores of the Sea of Aral, in about the forty-fifth parallel, or what would cor- respond to the position of the southern portion of the State of Maine. In the case of this family — Struthionidae — we also notice the singular fact, analogous to that which has been observed in relation to the distribution of the Camelidae, that the only repre- sentatives of the group other than Struthio (the ostrich proper), constituting the American genus Rhea, are birds belonging almost strictly to the temperate regions, their range extending from Pata- gonia to the southern confines of Brazil. The parrots (Psittaci) may be considered to be preeminently tropical birds, the vast ma- jority of the species being included in a zone bounded by the thirtieth parallel on each side of the Equator, but yet it may be doubted whether this limitation does not depend more upon the nature of the food-supply than upon the character of the climate. In South America a species of Conurus extends its range as far as the Strait of Magellan, and in the Macquarie Islands, in the South Pacific, representatives of the family are met with as high as the fifty-fourth parallel of latitude, corresponding to a position removed by only six degrees from the southern extremity of Greenland. Wallace probably justly refers to the "almost universal distribu- tion of parrots wherever the climate is sufficiently mild or uniform to furnish them with a perennial supply of food. " 15 But while in numerous, and perhaps the majority of, instances the limitation of animal groups to certain geographical regions is de- pendent more upon the physical character of the immediate environ- ment and the nature of the food-supply than upon particular con- ditions of climate, yet it cannot be denied that in very many cases climate appears to exercise a paramount influence upon distribution. This influence is frequently considered to be nowhere more forcibly illustrated than in the migration of birds, both as regards the northern species and those inhabiting the southern climes. That the climatic explanation of the phenomenon of bird migration is a 40 GEOGRAPHICAL DISTRIBUTION. fallacy most ornithologists are now agreed. It is a well-ascer- tained fact that the vast majority of birds are migrants to a greater or less degree, and that non-migration with this class of animals is much more of an exception than the rule. Yet, by reason of their peculiar covering, birds generally, as compared with other vertebrates, are but slightly affected by extremes of either heat or cold, and indeed, as far as we are capable of judging, by most climatic influences, provided only that their food-supply is not affected thereby. The condor in its aerial flight within a few minutes of time accommodates itself to the most varying climatic conditions, the change from the freezing cold of the mountain heights to the scorching heat of the tropical lowland plains seem- ingly having no effect upon the vigour of the bird. There can be but little doubt, as has been insisted upon by Professor Newton, that a deficiency in the food-supply — the necessity for searching for new food — is the most obvious cause or impulse promoting bird migration. Migrations of a somewhat similar character, in- disputably governed, at least in part, by considerations connected with the food-supply, but also in greater part by conditions of climate, manifest themselves among several other classes of ani- mals. Thus, in India, the monkeys habitually ascend the Himalaya Mountains in summer to elevations of ten or twelve thousand feet, and again descend in winter. Semnopithecus schistaceus has been observed at a height of eleven thousand feet, leaping in fir-trees laden with snow wreaths ! Wolves in severely cold weather descend from the mountain-slopes to the lowlands, and bears not infrequently migrate in great numbers to escape the rigours of an extreme winter. The migratory instincts of the northern hares and squirrels, and more particularly of the Norway rat and lemming, which in severe winters move in amazing numbers in direct lines over lake, river, and mountain, overcoming all obstacles that might be placed in their path, are well known. The Kamtchatka rats, under the pressure of numbers, are stated by Pennant to travel westward for a distance of eight hundred miles or more. Similar instances of the force of migration are presented by the hoofed animals. The vast herds of moving buffalo were until recently familiar sights to the traveller on the American plains ; in South Africa countless numbers of antelope, impelled by the necessities of food-supply, pour down upon the more favoured districts lying without the ANIMAL MIGRATIONS. 41 region of parched soils; and similar excursions, although in this case governed by reversed thermometric conditions, are practised by the onager or wild ass of Tartary. Even the reindeer is to an extent a migrant, since in both Russia and Chinese Tartary it de- scends far southward in advance of a rigourous winter, and, indeed, frequently reaches a lower latitude than any part of England, al- though in Scandinavia the animal is rarely seen south of the sixty- fifth parallel. It is not alone among the higher animals that the migratory in- stinct is developed. Turtles, during the ovipositing season, move in considerable numbers from one part of the sea to another, and they are stated to find their way annually to the Island of Ascen- sion, which is distant upwards of eight hundred miles from the nearest continental land-mass.16 Fishes migrate in immense num- bers, but the periodical shifting of the abodes of these animals is directly connected with the processes of reproduction. Certain fishes, as the salmon, shad, and smelt, ascend the waters of fresh- water streams for the purpose of depositing their eggs; others, again, as the herring and mackerel, frequent in immense shoals, during the breeding season, the neighbourhood of the coast-line. The young eel follows the line of the river-courses in myriads, ascending all the tributary streams, and frequently overcoming apparently impassable water - falls by squirming over the moss- covered ledges on either side. Among insects, the devastating migrations of the locust are proverbial, and similar illustrations of the wandering instinct could be cited from other members of the same class of animals. A remarkable example of migration has re- cently been observed in the case of a species of grapsoid crab (Se- sarma ?) off Cape San Antonio, the western extremity of the Island of Cuba. Barriers to Migration, and Facilities for Dispersion.— It has already been remarked that the interposition of extensive and elevated mountain-chains and of large bodies of water, and also sudden changes in the physical character of a country, are insur- mountable obstacles in the way of the migration or dispersion of certain classes of animals. The most serious of these obstacles, as affecting the dispersion of the Mammalia, is of course that of large bodies of water. We are well aware that the most experienced swimmer among this class of animals can accomplish by the nata- 42 GEOGRAPHICAL DISTRIBUTION. torial process but an insignificant journey, and, therefore, it would necessitate the interposition of but a very moderate expanse of water to effectually bar its progress in any given direction. Several members of the cat family are expert swimmers, the jaguar being known to cross the broadest of the South American rivers, the La Plata, as observed by Lieutenant Page. The tiger and elephant are both good swimmers. Deer are likewise prone to take to water, but it may be questioned whether animals of this kind would be apt to trust themselves beyond the sight of land. The domestic pig, even at a very young age, has been known to swim five or six miles, and it is not exactly impossible that the wild-hog, in cases of absolute necessity, might successfully attempt a passage of three or four times this distance. Probably the most remarkable exhibition of the natatorial powers of a land animal is that shown in the case of a polar bear, which was observed by Captain Parry vigourously paddling away in Barrow's Strait at a nearest distance of twenty miles from the shore, with no ice in sight on which it could have secured needed repose. It may safely be conceded, from our present knowl- edge on the subject, that while many of the land Mammalia can effect with safety, and even readiness, such water passages as are most generally to be met with on continental areas, none, probably, would be prompted to undertake a journey across an arm of the sea whose width measured fifty or more miles, or even one much ex- ceeding half that extent.* To these difficulties or impossibilities in the way of dispersion must be attributed the circumstance that the vast number of oceanic islands are deficient, except where man has effected an introduction, in representatives of this particular class of animals. The fact that certain allied, 'or even identical, forms of mammals are found in regions widely removed from each other, and which at the present time are separated by impassable bodies of water of greater or less extent, is practically conclusive evi- * In the case of the polar bear above cited, the absence from view of any ice need not necessarily, or even probably, indicate that there was no ice pres- ent nearer to the swimming subject than the ice of the land- border. From the mast of a vessel, elevated one hundred and fifty feet above the surface of the water, an icebenr rising to the same height could not, owing to the curvature of the earth, be distinguished at a greater distance than thirty-four miles ; flat masses of pack-ice, rising but a few feet above the Avater, at only about half that distance. DISPERSAL OF MAMMALS. 43 dence that in the former periods of the earth's history the surface of the globe must have undergone such vicissitudes as to have at vari- ous times disturbed the general relations existing between land and water. In other words, much of the surface that at one time was occupied by water must have been replaced by land, and, per contra, what was at one time land must at another have been water. And evidences of such variations in terrestrial equilibrium are abundantly afforded by geological landmarks. Had the greater portion of the surface of the globe at one time since the introduction of the Mam- malia consisted principally of dry land, or had there been since that period a general alternation in the relative positions of the land and water areas, the geographical distribution of the Mammalia would have been very different from what we actually find it to be. Hence, it must be assumed that a land and water alternation, such as could have brought about the present result, must have taken place in certain parts of the earth's surface only, and without affect- ing others. There would seem to be very strong grounds for con- cluding that the most recent connection uniting the principal land- areas of the globe was formed in the Northern Hemisphere, as a belt closing off the Arctic Sea (if it then existed) from the Pacific and Atlantic oceans. The only class of terrestrial mammals to which a broad arm of water offers no impediment in the way of migration or disper- sion is that of the bats ; and, singularly enough, just in the case of these animals, as has already been remarked, are we furnished with an example of universal distribution, there being but very few of the habitable oceanic islands which are not tenanted by one or more representatives of the order. But even among the habitually ter- restrial Mammalia there are certain exceptional methods by which dispersion to very considerable distances from the mainland can be effected. In the northern regions the frozen sea constitutes a con- necting bridge between distantly-removed land-masses which is constantly taken advantage of by various forms of Arctic animals. By the breaking up and drifting away of fragments of the northern ice-masses animals that might be temporarily wandering over them could readily be transported to very considerable distances from their true homes ; and, indeed, it is through such means that polar bears are periodically stranded upon the coast of Iceland. In one year alone twelve of such wandering animals made their appearance 44 GEOGRAPHICAL DISTRIBUTION. upon the island.17 The reindeer is stated to cross the Behring Straits by way of the Aleutian Islands and the frozen sea, and in a somewhat similar manner the musk-ox finds its way to Melville Island ; it is, however, singular that the last named, despite its long ice-journeys, never manages to reach either the continent of Asia or Greenland. In regions like the tropics, which support a luxuriant vegetable growth, and which are subject to periodical fluminal overflows, and, consequently, to the uprooting or outwashing action of the inundating waters, it not infrequently happens that islands or ' ' rafts " of considerable magnitude, consisting mainly of inter- laced or matted vegetation — tree-trunks held together by various creepers and climbers, and containing a sufficient quantity of vege- table mould and soil bound together in the roots — are floated down stream into the open sea, where they are at once placed at the mercy of the prevailing oceanic and atmospheric currents. These rafts have been frequently noticed at the mouths of some of the larger streams, as the Mississippi, Amazon, and Ganges, and, in the case of the last named, at a distance of a hundred miles from its mouth. Floating masses of wood, with upright trees growing over them, were mistaken by Admiral Smyth in the Philippine seas for true islands, until their motion made their real nature apparent. Such floating masses not rarely harbour various forms of animal life in their midst, and among these the Mammalia with arboreal hab- its are not inadequately represented. The South American trav- ellers Spix and Martius assert that on different occasions they ob- served monkeys, tiger-cats, squirrels, crocodiles, and a variety of birds, carried down stream (the Amazon) in this manner, and simi- lar observations have been made by other travellers in the case of the Rio Parana. It is asserted that no less than four pumas were landed in one night from such rafts in the town of Montevideo.18 Some of the animals thus conveyed may travel unconcernedly, and without any special disadvantage arising from a change of abode ; others, as the larger quadrupeds, will have been caught up and transported through accident. To what distance such a floating raft with its living cargo may ultimately be carried in safety, and without detriment to its inhabitants, over the oceanic surface there are as yet no data for determining. But there would appear to be no reason for assuming that they could not be transported to a distance of several hundreds of miles, seeing that the upright vege- DISPERSAL OF REPTILES. 45 tation found on many of them would serve with powerful effect in the face of a wind. And while the majority of the animal inhab- itants might be exterminated before the end of the voyage the safe arrival on an island or distant shore of a very limited number of individuals, embracing both males and females, would serve in a short period, under favourable conditions, to stock the new land with the species. That an absolute limit is set, however, to migra- tion as effected in this manner is proved conclusively by the utter absence in most of the oceanic islands of indigenous mammals, ex- cepting bats. The same obstacle that is interposed by the ocean to the disper- sion of the Mammalia presents itself in the case of the vast majority of other terrestrial animals in which the power of flight is not at all, or at best but feebly, developed. Thus, the serpents, although many of them are fairly good swimmers, are, if we except the marine forms, as incapable of passing oceanic barriers as are the quadrupeds, and their transportation from continental areas to regions far remote can only be effected by such or similar accidental means as that just described. As might have been expected, therefore, they are absent from nearly all oceanic islands. The Amphibia (frogs and toads) are no more fortunate in passing broad arms of the sea than are the serpents, despite the circumstance that in their young or larval condition they are strictly aquatic in their habits. Salt water proves fatal both to them and their eggs. Since moisture is a necessary condition for the early existence of this class of animals, it is evident that an extensive desert region will be an effectual barrier to their distribution — in fact, about as much so as an ocean. Lizards, in their adult condition, are as incapable of traversing an oceanic region as are the snakes and amphibians ; but it would ap- pear that in some special way — whether as effected by the oceanic currents themselves or through the agency of birds — their eggs may be transported to very considerable distances out to sea, since this order of animals is sufficiently represented in remote islands where neither snakes nor amphibians have as yet been encountered. That the ocean offers no insuperable obstacle to the broad disper- sion of a very large body of birds is known from almost daily observation. Birds are known to pass several hundreds of miles on the wing without halting, and, indeed, it is not exactly im- possible, or even improbable, that such unassisted flight may ex- 46 GEOGRAPHICAL DISTRIBUTION. tend over one or more thousands of miles. The flights of the wild-goose and the swallow have been estimated to be performed at the almost incredible velocity of from sixty to ninety miles per hour, and the flights of many of the smaller birds at not very much less. A sustained flight of ten or more hours in duration, especially when assisted by a favourable wind, involving an amount of muscular exertion probably within easy command of many birds, would carry them over an enormous stretch of territory, during a period of time which, by its brevity, would render the question of food-supply comparatively unimportant. Land-birds have been en- countered in the North Atlantic at almost all points of the oceanic expanse ; but to what extent these stragglers have received assist- ance in their flight, by taking temporary shelter on board the nu- merous vessels plying between Europe and America, can hardly be determined. There is no question as to such assistance in numerous instances, but whether it is afforded in all or most cases is a matter of pure conjecture. By whatever means or methods the oceanic travel of birds may be effected, it is a matter placed beyond all ques- tion that numerous American birds make their appearance at inter- vals along the European coast. Upwards of sixty species of such foreigners, embracing examples from nearly all the orders of birds, have at different times been noted on the eastern coast of the At- lantic, principally in the British Isles and the Island of Heligoland.19 Singularly enough, no distinctively European birds make their ap- pearance on the American coast, except a few whose journey over is made by way of Greenland and Iceland.* Despite the long- sustained flight of which birds are capable, it may be considered exceedingly doubtful whether many or any of them undertake these protracted journeys as a matter of their own pure choice or volition. It seems hardly possible that an animal would subject itself to such an amount of exertion and privation as would appear to be involved in journeys of this length, when n'o material ad- vantage could in the end be derived therefrom. It therefore ap- pears more than probable, as has been urged by Baird, Wallace, * No account is here taken of the purely pelagic forms, which are found on the opposite border* of the oceanic expanse, and which find suitable rest- ing-places on the surface of the waters. The greenshanks (Totanus glottis) has been obtained once hi Florida, and apparently nowhere else in the United States. MIGRATION OF BIRDS. 47 and Newton, that the oceanic wandering of land-birds must be attributed in most, or nearly all, cases to accidental circumstances — namely, storms, or the prevalence of certain winds — which may have wafted the birds beyond their control off to sea. Winds from the west, as has been shown by Professor Baird, are preva- lent between latitudes 32° and 58° N., and, hence, would be liable to catch such birds as may be passing southward during their au- tumnal migration, especially there where their flight would be at some distance off from the shore, or across broad arms or in- lets of the sea. The dispersal would naturally be facilitated by the interaction of a heavy storm, and it is a most noteworthy confirmatory fact that the appearance of American birds on the European coast is either presaged or accompanied by heavy westerly winds blowing in that quarter. North of the fifty-eighth parallel of latitude the polar winds trend westward, and with them we have the accompanying transferrence of European birds, by way of Iceland and Greenland, to the American continent. That storms or heavy winds do influence the flight of birds in the manner here described, is indisputably proved by the facts that present them- selves in connection with the occurrence of marine birds over con- tinental areas at some distance from the shore-line. The stormy petrel, during and after the prevalence of a northeast storm, has been seen in considerable numbers in the Eastern United States beyond the Alleghany Mountains; the Thalassidroma Leachii has been abundantly killed at or about the city of Washington ; and Professor Baird instances the case of a Pomarine jager (Cataractes Pomarinus), which was killed on the Susquehanna, at Harrisburg, in 1842.20 The golden plovers, in their southerly flight, start di- rect from Nova Scotia or Newfoundland for the West Indies, whence they continue their journey along the South American coast to Patagonia. In this journey but comparatively few in- dividuals touch or rest along the Atlantic States, yet it is known that during heavy northeastern winds, in the month of August, great numbers of the birds may be confidently expected along the New England coast. And it not infrequently happens that un- der similar conditions immense numbers of these and allied birds are driven to very considerable distances in the interior of the continents. In a like manner, during the prevalence of heavy storms, European birds are cast upon the Azores, situated about 4:8 GEOGRAPHICAL DISTRIBUTION. one thousand miles from the nearest continental coast. Among these are the kestrel, hoopoe, oriole, and snow-bunting, and not improbably also swallows, larks, and grebes.* If, then, birds may be drifted by accidental storms to a distance of one thou- sand miles in a direction contrary to that of the prevalent winds, it may be asked, Why may they not be thus drifted, at least after their first landing-place, another one thousand or two thou- sand miles further ? In other words, if European birds are carried to the Azores, why are they not at intervals also transported from there to the American coast? This question can, with our present knowledge, not yet be answered. Three or four species of European birds have been noticed in the Bermuda Islands — the wheat-ear (Saxi- cola cenanthe), the sky-lark (Alauda arvensis), the snipe (Gallinago media), and the land-rail (Crex pratensis) ; but three of these are also found in Greenland or on the North American mainland, while the fourth, the sky-lark, appears to have been brought over in, or to have escaped from, a ship.t In an ocean studded with islands, * Most of the resident land-birds of the Azores are identical with forms found in Europe and North Africa, and it, therefore, becomes impossible to ascertain how many of the individuals actually peopling the i>lands may not have been recently transported from the mainland. It is only under excep- tional circumstances -» barring the case of recognised stragglers — that such "wanderers can be determined. t The total number of European birds known to have found their way across the Atlantic to the American shores (including Greenland) is, accord- ing to Freke (''Zoologist," 1881), thirty-seven, of which Greenland counts about thirty, and the Eastern United States only twelve. This determination naturally excludes all birds that have been artificially introduced. 01' the twelve species occurring in the Eastern United States, s*ix are swimmers and five waders, and only one (and that somewhat doubtful, Buteo vulgar!*, re- ported to have been obtained in Michigan, in October, 1873) is a true land- bird. The wheat-car, referred to as occurring in the Bermudas, is considered a member of the North American fauna. The number of species of American birds crossing the Atlantic in the contrary direction is, according to the same authority (u Proc. Royal Dublin Soc.," 18S1), sixty-nine, of which twenty- two are swimmers, sixteen waders, and no less than thirty-one land-birds. The last include, among other forms, representatives of the genera Turdus (four species i, Galeoscoptes, Regulus, Dendro3ca, Ilirundo, Loxia, Zonotrichia, Ceryle, Coccyzus, Picus, and several species of birds of prey. The bald- headed eagle has been recorded from Sweden. It is significant that, of the forty-seven species of waders and land-birds, only two are known from Ice- land (Falco candicans and Numenius Hudsonicus) and none from the Faroe BIRDS OF THE GALAPAGOS. 49 which afford numerous resting-places, it would not seem difficult to account for the occurrence of land-birds at the remotest dis- tances from the mainland, even without having recourse to the accessory transporting agency of prevalent winds and storms. But even with this favourable condition added, it would appear that most land birds are not disposed to undertake of their own free will extended oceanic journeys, as is proved by the avi-fauna of many of the oceanic islands. Thus, while, as we have already seen, nearly all the representatives of the bird-fauna of the Azores, situated more than one thousand miles from the mainland, are identical with forms inhabiting either Europe or Northern Africa, indicating that the islands were peopled in comparatively recent times from those continents, in the Galapagos, situated only six hundred miles off the west coast of the continent of South Amer- ica, we meet with an entirely different state of things as regards the bird-fauna. Of about thirty species of indigenous land-birds, apparently only one, the common rice-bird (Dolichonyx oryzivorus), which ranges from Canada to Paraguay, is absolutely identical with a form found outside the limits of the island group. In addition to this a species of owl (Asio Galapagoensis) is considered by some authors to be but a mere variety of the cosmopolitan Asio brachy- otus, or short-eared owl, which is distributed from China to Ire- land, and from Greenland to Patagonia.51 We have here, there- fore, positive evidence that migrant stragglers from the South American continent are at the best of but very rare occurrence, and, on the other hand, visitors from the islands to the mainland appear to be equally rare. But since, from the resemblance which the fauna as a whole presents to that of the mainland, it is practi- cally proved that the same was at one time derived by migration from the continental areas — the islands being of volcanic origin — it is manifest that this migration must have taken place at a period sufficiently remote to have permitted the differences separating the two faunas to have been brought about. On the other hand, the absolute identity of the rice-bird with the similar form from the continent, proves, as has been pointed out by Wallace, that the island breed has been kept unaltered only through repeated or fre- Isles. The easterly dispersion is attributed to causes identical with those which have been assigned in explanation of the phenomenon by Professor Baird. 50 GEOGKAPHICAL DISTRIBUTION. quent visits from the specific congeners on the mainland. Not only are, with the one or two exceptions above noted, all the Galapagos land-birds specifically distinct from those found any- where else,* but they also belong largely to distinct genera. Of the fourteen genera represented, four are peculiar to the islands. The rarity of continental visitors to the Galapagos, as compared with the Azores, is to be attributed to the circumstance that these islands are situated in a zone characterised by an absence of storm winds. In the island of Juan Fernandez, situated in latitude 34° S., and only four hundred miles from the Chilian coast, there are but five species of land-birds, and of this number two are peculiar. In the Keeling or Cocos Archipelago, situated in the Indian Ocean at about the same distance from the Sumatran coast as are the Galapagos from the coast of South America, there is not a single species of true (indigenous) land-bird, although snipes and rails of the common Malayan species are sufficiently abundant ; and the same is true in the case of the island of St. Helena, situated eleven hundred miles from the nearest point of the continent of Africa, t Of the twenty species of Passeres, or perching-birds, inhabiting the Sandwich Islands — about the most strictly oceanic of any group of oceanic islands so-called, being situated fully two thousand miles from the nearest continental coast-line, and the same distance from the nearest island groups (Marquesas and Aleutian), if we except the small and almost tenantless shell and coral reefs — all the forms are peculiar ; and, furthermore, in all cases but one or two they belong to genera which are likewise confined to the islands. And even of the twenty-four or more species of aquatic and wading birds that have been observed on or about the islands, five — a coot (Fulica alai), a moor-hen (Gallinula Sandvichensis), a rail (Pennula Millei), and two ducks (Anas Wyvilliana and Bernicla Sandvichensis) — are peculiar.23 All in all there are some fifty species of birds known from the island group, of which about one-half are peculiar. It is evident that migrants (true land-birds) from distantly re- moved countries but rarely arrive here. In the case of the Ber- * The Dendroeca aureola, a species of wood-warbler closely allied to the " golden" or summer warbler of the United States (D. sestiva), is only doubt- fully separable from the D. petechia of the Island of Jamaica. t A small wading-hird of the genus vEgialitis (J2. Sanctse Helenas), allied to a species of plover common in South Africa, is found in the island. DISPERSAL OF BIRDS AND INSECTS. 51 muda Islands, which are distant from seven hundred to eight hundred miles from the nearest coast, we meet with a different order of things. The bird-fauna of these islands consists in all of about one hundred and eighty species, including both the land and aquatic forms, of which number, however, about thirty have been noticed only on one occasion. Of the eighty-five species of land- birds less than ten are permanent residents, the rest making their way principally from the North American continent and the West India islands.23 It is a singular circumstance that most of the foreign invaders are strictly migrating birds, whose course of mi- gration lies along the Atlantic coast, and which in their periodical wanderings frequently pass at some considerable distance out to sea. Entering the region of violent winds and hurricanes, they are liable to be snatched from their track, and to be forcibly trans- ported to some remote shore, where, of necessity, they will be compelled to secure for themselves a new home, and where, through frequent visitations of a like character, the original breeds estab- lished will remain pure and unaltered. Such is the condition of the bird-fauna of the Bermudas at the present time. None of the strictly non-migratory birds are represented in those islands. Two or more species of bat, also North American forms, arc, with the exception of rats and mice, the only indigenous mammals. Dispersal Of Insects.— It is a well-known fact that insects have been found in nearly all parts of the world that have thus far been trod by man, from the extreme limits of the Arctic and Antarctic regions to the Equator, and from the level of the sea to — and con- siderably above — the line of perpetual snow. Butterflies were ob- served by the naturalists of the "Alert" and "Discovery" nearly as far north as the eighty-third parallel of latitude; and Hum- boldt met with insects on Chirnborazo, at an elevation of upwards of 18,000 feet. They are found in fresh and salt waters, freely swimming on the surface — and at very considerable distances from the mainland— as well as below it ; in hot springs, where the water has attained to a moderately high temperature, end in subterranean caves. But, while the members of this class of animals, taken col- lectively, appear to be specially adapted to all the various condi- tions of existence that might be imposed upon them by accidental circumstances, the same does not hold for the individual members composing the class. Thus, certain insects are entirely dependent 52 GEOGRAPHICAL DISTRIBUTION. upon some special vegetable product for their existence, whether it be, as it may happen, the leaf, the flower, or the juice of the plant in question. Again, while in some cases the adult insect may be entirely independent of such a circumscribed food-supply, the larva may still be governed in its diet by a particular kind, without which, consequently, the prolonged reproduction of the species would be impossible. Such instances of limitation are exhibited by numerous forms of caterpillars. Hence, it is not difficult to com- prehend why, in regions which are affected by similar conditions of climate, and which collectively show a general correspondence in the character of the vegetation, certain species of insects should be found at one locality and not at another, even where no physical barrier separating the two should be interposed. In fact, the bar- rier interposed by conditions of vegetable growth is fully as effective in restraining a broad specific distribution as are the barriers re- sulting from the physical conditions of the earth's surface, most of which they are able to overcome, either voluntarily or involun- tarily. The mature insect, from its lightness, is frequently carried away in aerial currents from its native or favourite haunts to regions widely remote, in a manner precisely similar to what obtains in the case of birds. Hawk-moths have been caught on board ship at a distance of two hundred and fifty miles from shore, and a large Indian beetle (Chrysochroa ocellata) was captured some years ago, in the Bay of Bengal, at a distance of two hundred and seventy miles from the nearest land. During Captain King's expedition to the Straits of Magellan dragon-flies flew on board his vessel when still fifty miles out at sea (south of the Rio de la( Plata) ; and Admi- ral Smyth reports that, in the Mediterranean, myriads of flies were brought to his ship by a southerly wind from a region fully one hundred miles distant. A beetle is recorded by Darwin as having been caught aboard the " Beagle" when the vessel was upwards of forty miles distant from the nearest shore ; from what actual dis- tance the insect may have come could, necessarily, not be deter- mined. A locust was observed by the same naturalist three hundred and seventy miles from land; and in 1844 swarms of these insects, "several miles in extent, and as thick as the flakes in a heavy snow-storm, visited Madeira. These must have come with perfect safety more than three hundred miles, and, as they continued flying over the island for a long time, they could evidently have travelled DISPERSAL OF MOLLUSKS. 53 to a much greater distance."24 In addition to this means of aerial dispersion, the distribution of insects may be to a great extent ef- fected in the condition of eggs, which retain a considerable amount of vitality, and which are not infrequently laid in decaying timber and in the living tissues of various plants. When, therefore, float- ing rafts or mats are apt to be formed, and to be floated out to sea, it is almost certain that with them will be carried out a host of in- sects— whether in the perfect form, as grubs, or as eggs — of different species, a fair proportion of which will, doubtless, have retained their vitality even after a protracted sea-voyage of several thousand miles. It is in this manner that many or most of the tropical forms which periodically make their appearance on the British coast have been transported thither, the current of the Gulf Stream, which trends in a general northeasterly direction, being instrumental in drifting tropical log- wood to the trans- Atlantic temperate shores. Dispersal of Mollusks.— The world- wide distribution of the fresh-water and terrestrial Mollusca, and the occurrence of identical or very nearly allied generic forms at opposite quarters of the globe, prove conclusively that the animals of this class are favoured with special instrumentalities by which a broad distribution is effected. Land-snails of the genus Helix are found in all the continental areas, from the polar regions to the Equator, and from the limit of perpetual snow on mountain summits to the level of the sea; they are also found in all the oceanic islands, even the most remote, that have thus far been visited. The exact nature of this distribution has not yet been positively determined, and, in fact, there are sev- eral difficulties in the way of accounting for it. It is well known that these animals cannot survive for any length of time the effects of salt water, and this water is almost immediately fatal to the vitality of the eggs. Hence, only under exceptional conditions is it possible to account for a transferrence over a broad expanse of oceanic surface. But it has been ascertained that such forms as are capable of secreting an epiphragm, and therewith closing up the entrance to the shell, are able to resist the injurious effects of salt water for a very considerable period, in some instances as much as two weeks, or more, as has actually been determined experimentally by the immersion of land-shells in the briny medium. In regard to these, therefore, there will be no difficulty in accounting for a broad distribution, since they, and especially the genus Helix of all others, 54 GEOGKAPHICAL DISTRIBUTION. would be liable to be concealed in and transported away by floating timbers. In this manner they could be drifted away for several hun- dreds of miles, and, under exceptionally favourable circumstances, to possibly one or two thousand, the more readily since some of these animals possess an enormous amount of vital tenacity, even under the most adverse conditions of existence. Thus, a Helix from North Africa (H. desertorum), contained in the British Museum collection, and glued on to a tablet, was found by the conservators to be alive after a period of more than four years. A similar in- stance of resuscitation, although after a less protracted period, has been noted in the case of one of the tabulated snails of the Acade- my of Natural Sciences of Philadelphia. Again, it has been con- clusively shown by Darwin and others that the eggs of pond and other fresh- water bivalve-mollusks are occasionally found attached to the feet of wading-birds — ducks, and the like — visiting such waters, and are by them liable to be carried to very considerable distances from their true homes, and thereby to have their range almost inimitably widened. Such a method of transport, although exer- cised to a much more limited extent, has been observed to be effect- ed even by species of water-beetle, whose legs may have become entrapped between the valves of the shell, as well as by newts and other amphibians. The broad distribution of allied or identical generic and specific forms of fluviatile mollusks over the most ex- tended or widely remote geographical areas receives a partial ex- planation in the circumstance that the physical forces operating upon the earth's crust, causing movements in it of a differential character — i. * 128 GEOGRAPHICAL DISTRIBUTION. which are as transparent as the medium in which they live have held their own." The lacustrine pelagic animals perform daily vertical migrations of the same character as has been noted in the case of the oceanic pelagic fauna, descending to the regions of obscurity during the day-time, and ascending by night. The animals appear to shun the light of the sun, and even of the moon, and hence retire to a depth probably not far from the limits of light penetration ; the fauna is, therefore, one of darkness. The greatest depth whence specimens were obtained by Forel in Lake Geneva was about one hundred and fifty metres ; but at this depth only Diaptomus was found. At a depth of fifty metres, in the Lago d'Orta, Pavesi found a very profuse fauna, represented by seven species ; in the Lago d'Iseo, at five, fifteen, and thirty metres, the catch appears to have been exceedingly abundant (" la pesca fu prodigiosamente ab- bondante ") ; but, at one hundred metres, where the temperature of the water was 19° C., as compared with a surface temperature of 23° C., the fauna was decidedly scanty, although five distinct forms were obtained. To what extent the downward extension of the pelagic fauna is governed by conditions of temperature, or in how far this limitation is dependent principally upon the presence or absence of light as a determining factor in the evolution of plant life, still remains to be ascertained. Forel, in 1874, found that paper sensitised with chlo- ride of silver was still acted upon by the diffused light of the Lake of Geneva at a depth of about forty-five metres in summer and one hundred metres in winter, while ordinary shining objects disap- peared from view at a depth of sixteen to seventeen metres. Asper, in August, 1881, obtained positive results through the use of plates sensitised with an emulsion of bromide of silver at a depth some- what exceeding ninety metres in the Lake of Zurich ; and more re- cently (1884-'85) Fol, Sarasin, Pictet, and others, have been able to detect the penetration of light in Lake Geneva to a maximum depth (in winter) of two hundred metres. In summer the penetra- tion is considerably less. Fol and Sarasin 51 have also demonstrated that, in the Mediterranean, the solar rays penetrate to a depth nearly double that to which they were found to descend in the Swiss lakes, or to four hundred metres, and that at a depth of three hundred and eighty metres the intensity of light is as great as in Lake LAKE FAUNAS. 129 Geneva at one hundred and ninety-two metres. At this depth, however, the impression produced upon the sensitised plates was of no greater value than that which would have been produced, under ordinary conditions, on a clear night, without a moon. A remarkable feature of the lacustrine fauna is the very broad distribution of most of the species. Not only is there a general resemblance between the pelagic faunas of all the European lakes that have thus far been examined, from Scandinavia to Italy, and from Italy to Bohemia and the Caucasus, but a strict identity, at least as far as the species of Entomostraca are concerned. The species that occur in the one lake are also the species of the other lakes, although the respective littoral and deep faunas may be largely distinct. Further, it would appear that the same species are constituents of the pelagic faunas of American lakes as well, and not improbably make up the greater part of them. Professor S. I. Smith,62 in his investigations of the fauna of Lake Superior, determined the presence, in the surface waters, of Daphnia galeata and Leptodora hyalina, common forms in the lakes of both Southern and Northern Europe, and of Daphnia pellucida, which was de- scribed by Muller as a pelagic inhabitant of some of the Danish waters. As to the origin of the pelagic fauna little positive is known. That it is not a direct derivative of the different littoral faunas is very nearly certain, for were this the case we should expect to meet with largely differing assemblages of pelagic forms in all lakes where the littoral or deep faunas likewise differ ; but, as has been seen, this is not the case. Yet there can be little or no question that it really represents a modification of some primary shore- fauna, whose members, through force of circumstances, were compelled to adapt themselves to new conditions of existence. The supposed method of its differentiation and further distribution is thus indi- cated by Forel: "I believe we must find the cause of the differ- entiation of the pelagic fauna in the combination of two different phenomena — namely, the daily migrations of the Entomostraca, and the regular local winds of the great lakes. It is well known that on the borders of great masses of water two regular winds prevail, one of which blows at night from the land towards the water, the other by day from the water to the land. The nocturnal animals of the shore-region, which swim at night at the surface, 7 130 GEOGRAPHICAL DISTRIBUTION. are at this time driven towards the middle of the lake by the sur- face- current of the land-winds, sink during the day, being driven away by the light, into the deep water, and thus escape the surface- current of the lake-winds, which would otherwise have carried them again to the shore. Constantly driven farther every night, they remain confined to the pelagic region, as they are not carried back again during the day. Thus a differentiation takes place by natural selection, until at last, after a certain number of genera- tions, there remain only the wonderfully transparent and almost exclusively swimming animals which we know. When this differ- entiation has once taken place, the pelagic species is conveyed [in the condition of resting eggs] by the migratory water-birds from one country to another, and from one lake into another, where it reproduces its kind if the conditions of existence of the medium are favourable. In this way we may find the pelagic Entomostraca in lakes which are too small to possess the alternation of winds, the animals having been differentiated by the action of the winds in other larger lakes." It might, however, be asked with Pavesi, if the general uni- formity of pelagic faunas has been brought about through a method of distribution such as is here indicated, how has it happened that some lakes should be so largely deficient in pelagic forms as com- pared with other, and nearly contiguous, lakes ? The lakes of Northern Italy may be taken in illustration of a condition of this kind. Seeing that identical forms have been scattered to such widely separated quarters of a continent, as Italy, Scandinavia, and the Caucasus, it certainly appears a little surprising that immedi- ately adjoining districts should have been so irregularly stocked with the distributed material. It might, however, be conceived to be a matter of accident, and, indeed, at first sight the condition appears to be more in the nature of a support to the theory stated than as an argument against it. But if accidental conditions of this kind have happened, why has it not also accidentally happened that some of the lakes should have retained a fauna, formed through modification of their own particular littoral or deep fauna, distinct from that of any other lake ? Still, the objection here raised is not an insuperable one, and offers much less difficulty in the way of the partial solution of the problem than does the circumstance of the oc- currence of identical forms in the lakes of Europe and North America. LAKE FAUNAS. 131 Deep Faunas of Lakes. — The most systematic and thorough investigations that have been made into the nature of deep lacustrine faunas are those of Forel upon the fauna of Lake Geneva.63 As the result of the observations of this naturalist it would appear that the abundant fauna of the floor of this lake comprises representa- tives of nearly all the primary divisions of fresh-water — inhabiting Invertebrata, and that even a fair proportion of the secondary groups are also represented, although by a very limited number of species in nearly all cases. Included in the lowest forms are sev- eral amoebae, and Epistylis, Opercularia, and Acineta among infu- sorians. The hydroids are represented by the common brown hydra (Hydra rubra — to one hundred metres), and the rotifers by Flos- cularia. Three orders of worms are indicated — nematoids, cestoids, and turbellarians — and two of annelids proper, the hirudines and chaetopods (Lombriculus, Tubifex, &c.). The turbellarians (Pla- naria, Mesostomum, Dendroccelum) have no less than eleven species, one of which, Vortex Lemani, is found at all depths between fifteen and three hundred metres. A cestoid was dredged from a depth of two hundred and fifty-eight metres. The crustaceans are repre- sented by a limited number of species belonging to the amphipods (Gammarus caecus), isopods (Asellus csecus), cladoceres (Lynceus), ostracods (Cypris, Candona), and copepods (Cyclops, Canthocamp- tus). Other articulates are four or five species, and as many genera, of arachnids (Arctiscus, Hydrachnella, &c.) and the larvae of some tipuliform insects. The limited number of mollusks inhabiting the depth is not a little remarkable ; of the lamellibranchs there is the single genus Pisidium, with about three species, and of the gastero- pods only the genera Limnaea and Yalvata. Although the Unioni- dae (Anodon) are very abundant in the littoral fauna, they are com- pletely absent below. One species of Limnaea (L. abyssicola) was found to be sufficiently abundant at a depth of two hundred and fifty metres, a circumstance to which Forel calls attention as indi- cating the readiness with which an air-breathing mollusk can ac- commodate itself to conditions largely at variance with those which are considered necessary to conform to certain structural peculiari- ties.* * When brought to what might be considered its proper position, the sur- face of the water, the mollusk almost immediately adapted itself to the new conditions of existence, apparently without undergoing any inconvenience. 132 GEOGRAPHICAL DISTRIBUTION. Although the fauna, taken as a -whole, may be said to possess certain special characters, yet, broadly considered, it is only the representative, by slight modification, of the fauna of the littoral zone. It possesses no really well-defined or abnormal features of its own. Most of the forms are of small size, and a number of them, whose surface representatives are active and good swimmers, appear to have taken to sluggish habits ; neither Cyclops nor Lynceus would rise when placed in an aquarium. Blindness is exceptional, and it is a surprising fact that the animals suffering from this defect (Gammarus cascus, Asellus csecus) are comparatively shallow-water forms (thirty metres), whereas those living at the greater depths, down to three hundred fathoms, are well provided with visual organs. The faunas of different zones of depth do not appear to differ sensibly from one another, except in the elimination or excess of a number of species. The greater number of these would seem to be distinct from their analogues of the littoral zone.* Many of the species found in Lake Geneva are identical with forms found in the other Swiss lakes, and in the lakes of Savoy, as identified by Imhof, and there is good reason for supposing that a general analogy, if not absolute identity, unites the different deep lacustrine faunas of the same region. Professor Smith obtained from deep water (exceeding fifteen fathoms) in Lake Superior64 Hydra carnea (from eight to one hundred and forty-eight fathoms), a Pisidium (from four to one hundred and fifty-nine fathoms), several species of worms (Seenuris, Nephelis, Tubifex, &c.), the larva? of various tipulids and ephemerids, and among crustaceans Mysis relicta and Pontoporeia affinis (from shallow water to one hundred and fifty-nine fathoms). The last two, which .were also found by Stimpson in Lake Michigan, are forms belonging to Lake Wetter in Sweden, supposed by Loven to have been derived by modifica- tion from marine species. * So stated by Forel in his report of 1876, although in 1874 he and Flessis appear to have maintained the opposite view. PAET II. GEOLOGICAL DISTRIBUTION'. I. The succession of life. — Faunas of the different geological periods. THERE is no fact more patent in the history of the organic world than that there has been from first to last a progressive evo- lution from lower to higher forms in the chain of beings that suc- cessively peopled the earth's surface. Casting our eye back over the vast series of rock deposits which together constitute the fossili- ferous scale of geologists, from the Cambrian to the Post-Pliocene, and which together have a maximum development of probably not less than two hundred thousand feet (or forty miles), we re- mark along the most ancient horizon the traces of animals which bespeak the organisation of some of the lowest forms of life with which we are at present acquainted; in the middle distance we note the appearance of forms whose organisation marks a decided advance upon that of their predecessors; and, finally, in the fore- ground, we are brought upon the threshold of those highly com- plicated forms which to-day people the surface of the earth. The simpler forms of life came into existence first ; the most complex last. It must not be implied, however, that with the progressive and steady evolution of higher forms there has been an equally progressive destruction or elimination of the forms of lower or- ganisation; both have kept pace with each other, so that, at the present day, although innumerable groups have completely disap- peared, the lowest is found flourishing side by side with the highest. This inter-association of lower and higher forms has manifested it- 134 GEOLOGICAL DISTRIBUTION. self in all the geological formations that are known to us, from the Cambrian period to the present day, and there can be no doubt that, were a fossiliferous formation discovered of older date than the Cambrian, or immediately underlying it, we should find pre- cisely the same juxtaposition, although to a more limited extent, of organisms of higher and lower development. Only then when we could fathom the first-born deposit, or trench upon the period when life first came into existence, would we, in all probability, be circumscribed in our survey to animals exhibiting a nearly uniform low grade of organisation. Such a point has probably not yet been reached, or, if reached, its existence can only be indicated with doubt, since the oldest rock deposits (the Laurentian) into whose composition an organic element unquestionably largely en- tered have lost all or nearly all traces of their primary fossiliferous character. With this wholesale obliteration have, consequently, disappeared the traces of the earliest and most primitive types of life-forms. If Eozoon and Archa3ospherina, from the Laurentian limestones, be considered actually to represent organic forms, as is maintained by many prominent geologists and naturalists, then, indeed, are we presented with a large series of deposits in which apparently all the organic elements belong to one uniformly low type — the type of the Foraminifera — not yet the lowest, but very nearly it. But even granting the animal nature of the two structures here indi- cated, it would yet be very unsafe to affirm that they represent the only forms of life that tenanted the earliest seas; multitudes of other forms may have flourished and perished, and left no traces behind them, or had their traces completely obliterated at some remote subsequent period. We should then be no wiser for their existence. The succeeding Cambrian period ushers in with it such a host of multiform beings— beings of comparatively high organisation — that it becomes almost impossible to conceive that their ancestry should date back only to a period so little removed from the Cambrian as the Laurentian, unless, indeed, the hiatus separating the Laurentian from the Cambrian is very much greater than is indicated by its stratigraphical position. But if the an- cestral forms of the Cambrian stock already existed in the Laurentian seas, what has become of their remains ? Why is it that in these oldest so-called fossiliferous rocks we meet with only Eozoon and CAMBRIAN FAUNA. 135 Archaeospherina ? Surely it could not have been that there was such a disposition of the remains as to leave nothing but these two forms belonging to the lowest type. It appears far more plausible to assume with those who uphold the mineral nature of Eozoon and Archa3ospherina that we have no traces of this ancient pre- Cambrian fauna remaining, and that, consequently, the destruction was complete. How far back beyond the Laurentian the root of the present existing chain of organisms may have extended it is impossible even to conjecture. Cambrian Fauna. — It is certainly a surprising fact, whichever way it be considered, that, with the formation bringing the first unequivocal evidences of organic life, we should meet with that multiplicity and variety which characterise the faunal assemblage of the Cambrian period. Most of the greater divisions of the ani- mal kingdom, possibly not even excepting the vertebrates, were there represented, and most of these already in the lowest or oldest deposit — protozoans, crelenterates, echinoderms, worms, articulates, and mollusks. And more than this, some of these groups were already represented by a full, or nearly full, com- plement of the orders that have been assigned to them by natu- ralists, and which include all the various forms that have thus far been discovered as belonging to the groups. Thus the Cam- brian echinoderms are represented by forms belonging to three out of the six usually recognised orders — the Cystidea, Crinoidea (ocean-lilies), and Asteroidea (star-fishes). The last two have repre- sentatives living at the present day, whereas the former is entirely extinct. We have here, then, the most ancient ocean-lily and star- fish (Palasterina), and it is interesting to note what distinct rela- tions these two forms hold to their modern representatives. While the Crinoidea attained their maximum development in the seas of the Paleozoic period — Silurian, Devonian, and Carboniferous — since which time they have been pretty steadily declining, until at the present moment they are represented by scarcely more than a half- dozen distinct generic types, the Asteroidea have been just as stead- ily increasing, and, indeed, attain their maximum development in the modern seas. It may appear at first sight anomalous how two groups, so widely dissimilar from each other, and having such varying developments, should have appeared simultaneously in the same period of the 136 GEOLOGICAL DISTRIBUTION. earth's geological history, the Cambrian. But it must be borne in mind that in the Cambrian formation we have only what is seem- ingly the oldest fossiliferous formation, and that the ancestral forms of both ocean-lilies and star-fishes lie buried in rock deposits of undeterminably older age. If, on the hypothesis of evolution, we uphold the inter-derivation, or derivation from one another, of these two forms, then it is but fair to assume that the crinoid, which is structurally the lowest, appeared at a period considerably anterior to the star-fish, which must have required for its specialisation a no inconsiderable lapse of time. . And it is a singular fact, and one strikingly confirmatory of this view of relationship, that we have in both these forms certain peculiarities of structure which effect a sort of transition from the one to the other. Thus, in some of the fixed crinoids the plume or tuft separates from the column after a certain period of existence, and then leads an independent exist- ence, to all appearance a stellarid (the Comatula). Conversely, the officers of the late " Travailleur " deep-sea dredging expedition obtained off the coast of Spain, and from depths respectively of nineteen hundred and sixty and twenty-six hundred and fifty metres, two individuals of a new genus of star-fish (since named by Perrier Caulaster peduncularis), which exhibited on the dorsal surface a true peduncle, demonstrated to be absolutely homologous with the stalk of the crinoid. Yet, despite this obvious relation- ship, it is not a little surprising that no pedunculated star-fish has thus far been found fossil, nor any comatulid crinoid to antedate the Jurassic period (Antedon). The Cambrian Mollusca comprise representatives of five of the six classes that now inhabit the seas, namely,, the Brachiopoda, Acephala, Pteropoda, Gasteropoda, and Cephalopoda. Here again, therefore, we have an apparent simultaneous appearance of lower and higher forms ; but, as before, we must look to a much earlier period for the ancestral traces, if any have been preserved, of the first or most primitive type. The genetic relationships of these various molluscan groups cannot, in the present state of the science, be determined with any degree of certainty; but, if a low degree of organisation indicates antiquity, which certainly appears to be the case with many groups of animals, then it may be fairly assumed that the Brachiopoda were the first to appear. It is a surprising fact in the history of these animals, and one which is, perhaps, not CAMBKIAN FAUNA. 137 repeated to the same extent in any other group, that while hosts of genera, and even complete families of this order, which flourished in the seas intermediate in time between the Cambrian period and our own day, should have successively disappeared, a few individual types seem to have survived from first to last, without having un- dergone any essential modification of structure. Thus, the Lingula, or Lingulella, of the Cambrian rocks is but very little, if at all, dif- ferent from the existing Lingula, and it has indeed been considered doubtful by some authors whether even specific characters could be assigned to distinguish some of the earlier from the later forms, separated by an interval of millions of years. The same persistence of type is represented in the genus Discina. Side by side with these lower molluscan types, but appearing at a somewhat later period, the Upper Cambrian, we find, as has al- ready been stated, forms belonging to the highest order, the Cepha- lopoda (Orthoceras, Cyrtoceras), another apparent contradiction to the doctrine of progressive higher development. Considering the group of the cephalopods by itself, however, we observe that its earliest types belonged to the lower of the two divisions into which the cuttle-fishes have been divided — the tetrabranchiate, or four- gilled order — a division to which the somewhat later appearing, and now probably disappearing, Nautilus also belongs. These primitive cephalopods were succeeded in time by other members of the same order— Gyroceras, Nautilus, Goniatites — until the Tri- assic period was reached, when the first dibranchiate form, Belem- nites, appears. From this period down to the close of the Meso- zoic era both the two-gilled and the four-gilled forms occur in such abundance that it would be almost impossible to state to which group belonged the preeminence. But in the meanwhile a general alteration and succession in the representative cephalopod type had been taking place. The early forms already mentioned, Orthoceras, Gyroceras, Cyrtoceras, and their allies, belonging to the family of the Nautilidae, are succeeded in the Triassic period, where their last traces (excepting Nautilus) are to be met with, by the members of the more complicated group of the Ammonitidae, whose earliest precursors (three or more species from the Carboniferous formations of India, and a solitary species from the Carboniferous of Texas) would seem to have been foreshadowed by Goniatites, a type struc- turally intermediate between the Nautilidae and the Ammonitidae. 138 GEOLOGICAL DISTRIBUTION. Similarly, in the case of the Cephalopoda dibranchiata, the Belem- nitidse are succeeded by forms more nearly resembling the cala- maries, or cuttle-fishes (Teuthidse), of to-day, whose remains are found already in the Jurassic deposits (Onychoteuthis, Teuthop- sis, Belemnosepia). With the beginning of the Tertiary period* we note the final disappearance of the varied group of the Ammoni- tidse, and with them the last traces of all but one of the lower or four-gilled order of cephalopods. The single exception is the Nautilus, which, as a persistent type, almost unaltered from the Silurian to the present period, alone survives to contest the seas with the members of the higher or dibranchiate order. The predominant Mollusca of our modern seas, the Gasteropoda and the Acephala, were but feebly represented in the seas of the Cambrian period; but it seems not improbable that some of the earliest forms — e.g., Capulus, Pleurotomaria, among the snails — belonged to types absolutely identical with those living at the present time. It is not until we have completely passed over the Paleozoic era, which, so far as the Mollusca are concerned, may be said to constitute the age of the Brachiopoda, that these two orders of shell-fish attain any special significance. From the beginning of the Mesozoic era onward they steadily crowd the deposits with their remains, until, finally, with the Tertiary formations, and the forma- tions succeeding these, they constitute the most characteristic and most important invertebrate landmarks to the geologist and paleon- tologist. It has been contended, and with apparent force, that the irregu- lar appearance in time of the Mollusca— i. e., the almost simultane- ous introduction of forms belonging to both the lowest and the highest orders, and the final supremacy in the existing seas of the type of the Acephala, a group of mollusks inferior in organisation to the Cephalopoda, the Gasteropoda, and the Pteropoda — is in- compatible with the doctrine of evolution, which^ as argued, re- quires for its confirmation the introduction first of the lower forms, the development from these of the more advanced, and, ultimate- ly, the appearance of those that are most perfect or specialised in structure. It must be recollected, however, that, as far as the al- most simultaneous introduction of lower and higher forms is con- * A species from the Lower Tertiary of California. CAMBRIAN FAUNA. 139 cerned, the obstacle is more apparent than real, for, as has already been insisted upon, it is impossible to determine how far back be- yond the Cambrian, or first unequivocally fossiliferous* formation, life may have already existed, and, consequently, to what very ancient period the ancestry of the molluscan type may extend. As a matter of fact, the most ancient mollusk, the Lingula, or Lingu- lella, is almost the lowest in structure of any with which we are acquainted, and if in the rock deposits we meet with its remains but barely antedating those of the very much more highly organ- ised Orthoceras, we have yet strong grounds for concluding that its first appearance was very much earlier, only that, through the gen- eral obliteration of all remains in the preceding geological period, direct evidence to that effect has been lost. As to the other objec- tion, that the predominant forms persisting at any given epoch should be those whose structure manifests the highest development, it may be remarked that the evolutionary force requires no such result as the outcome of its operative action. It is among such forms as, in their mutual relations to their surroundings, whatever these may be, are best adapted or fitted for combatting the nu- merous elements that constantly interpose themselves in the path of existence, that we must look for examples of greatest persistence and development — for the survivors in the struggle for existence. Hence, while the highest developed forms in any given series of animals will present themselves in or about the period most re- moved from the birth of that series, yet it need not follow that the higher series will ultimately outlive, or even predominate over, the representatives of a lower parallel series of the same class of animals, whose fitness for struggling in the battle for existence is not infrequently vastly superior to that of the higher class. We need not be, therefore, surprised at finding, in a given class of ani- mals, some of the more perfect forms disappearing from the world's horizon before the less perfect, and these last, consequently, the survivors in the general battle for life. But, while no general law can be formulated regarding the disappearance, as conditioned by the degree of perfection, of the various series of a given class of animals, or, regarding their relative development in any one period of the earth's history, the law of appearance or succession already stated — i. e., the introduction of lower forms before those of a higher order — can very generally be maintained. 140 GEOLOGICAL DISTRIBUTION. An objection to the evolutionary doctrine, similar to that which has been drawn from the distribution of the Mollusca, is also fur- nished by the articulated animals, or, more particularly, by the class of the Crustacea. The members of this class boast of a lineage, as far as has yet been determined, very nearly, if not fully, as ancient as that of the Mollusca, one extending back to the earliest Cambrian period. But, while the most ancient mollusks with which we are acquainted belong in great part to orders, families, and even genera, whose representatives still flourish in the existing seas, the most ancient crustaceans, or at least the majority of them, the Trilobita, have long since become totally extinct ; hence the impossibility of determining their true relationships. However uncertain or obscure this relationship may be, whether it is with the Phyllopods, as claimed by some, or, what is much more likely, with the Xiphosura (king-crabs) and arachnids, as argued by others, there can be no doubt, if homologies of structure can be relied upon, that the mem- bers of this group of animals represent a high grade of structural organisation, and especially if the period of their appearance is taken into consideration. But here, just as in the case of the Mollusca, we have the strongest evidence for concluding that their earliest appearance dates far beyond the Cambrian period, as is proved almost conclusively by the simultaneous appearance in the oldest Cambrian strata of some of the simplest and most compli- cated trilobitic forms, Agnostus and Paradoxides, which are at the same time also among the smallest and the largest forms of the entire order. A further evidence of the pre-Cambrian antiquity of this group is furnished by the circumstance of the abundance in which the earliest remains are found, an abundance which, though perhaps not equal to that characteristic of the succeeding Silurian and Devonian trilobitic faunas, is yet sufficient to impress a dis- tinct individuality upon the fauna of the period. While with the Cambrian trilobites we find associated other forms of crusta- cean animals, such as the phyllopods (Hymenocaris) and ostracods (Primitia, Leperditia), the highest members of the class, the Deca- poda (crabs and lobsters), appear not to have been as yet evolved. Indeed, it is not until the entire Silurian period and a considerable portion of the Devonian are passed that we meet with an example of the ten-legged order of crustaceans. Barely had these higher forms asserted themselves on the field of life ere a decline in the CAMBRIAN FAUNA. 141 supremacy of their predecessors is made manifest. With the mid- dle of the Devonian period the beginning of the trilobitic decay becomes apparent, and, after the close of that period, i. e., in the Carboniferous, less than a half-dozen types remain, and even these are of comparatively rare occurrence. At the close of this last- named period the trilobites disappear totally and forever from the scene.* Broadly looking over the Cambrian fauna, \ve find it to be dis- tinguished by two important features. One of these is the fact that it is entirely destitute of both land and fresh-water forms, or such as are strictly adapted to breathing directly the oxygen of the atmosphere or that of fresh water. All the forms thus far encoun- tered are, as far as we know, of a strictly marine nature. The ab- sence of land animals will scarcely appear surprising in view of the complete, or nearly complete, absence of a land vegetation, and the correlative want of the nourishing material requisite for that charac- ter of organisms. The absence of fresh-water forms is not so readily accounted for, unless it be that there were formed at that time no fluviatile or lacustrine accumulations of sufficient magnitude to have left their traces behind them. It is not impossible, however, that some, or even many, of the recognised marine fossils, or such as have a marine habit, of the Cambrian formation, are in reality estuarine or brackish forms, as it can scarcely be conceived that all the deposits that were formed at the mouths of the ancient rivers should have been so totally destroyed or covered over as to have left absolutely no vestiges of their former existence. Doubtless, some of these have been preserved, along with their contained fos- sils, although the exact nature of such deposits may be disguised from us by reason of our imperfect knowledge concerning the true habits of their representative organisms. Nor would it be abso- lutely safe to affirm that some of these organisms, undistinguishable from what at the present day are indisputably marine types, may not in reality have been of a purely fresh-water habit in those early days. The other distinguishing feature of the Cambrian fauna is the * Shumard has described representatives of the genera Phillipsia and Proetus in deposits of the Sierra Madre, of the Southern United States, claimed to belong to the Permian period ; the determination of age may be considered to be very doubtful, however. 142 GEOLOGICAL DISTRIBUTION, absence of positive indications of the existence of vertebrated ani- mals. The only objects that have thus far been described as per- taining, with any show of probability, to the members of this highest division of the animal kingdom, are the singular bodies known as conodonts, which, in the opinion of their discoverer, Pander, and of some other naturalists, represent the teeth of fishes belonging to the order of the myxinoids (hags and lampreys), with the exception of the lancelet (Amphioxus) the lowest of the entire class of Pisces. The weight of opinion, however, seems to relegate these problem- atical bodies to the Invertebrata, and not improbably, as has been urged for some of these forms, they represent the jaw-teeth of cer- tain annelids. Silurian Fauna.— The fauna of the Silurian period marks a decided advance upon its predecessor. The chain of organisms which, with the exception of the somewhat doubtful conodonts, was hitherto constituted exclusively by the members of the inver- tebrate series — sponges, echinoderms, mollusks, articulates— exhib- its here for the first time indisputable representatives of the more highly organised group of the vertebrates ; but not until the Upper Silurian deposits are reached. We here meet with the remains of two distinct orders of fishes, the sharks or dog-fishes (Elasmo- branchii), as represented by Onchus and Thelodus, and the bucklered Ganoidei — Pteraspis — the former still very abundant in the modern seas; the latter, which include, among other forms, the sturgeon and alligator gar, probably nearly verging on extinction.* In both these orders the osseous framework or skeleton is frequently in a more or less imperfect condition — complete ossification being the exception rather than the rule — and hence, in so far, these primi- tive vertebrates exemplify a low grade of organisation compared with those — like the bony fishes, and most of the animals above them — in which the vertebral column is completely ossified, or reaches its furthest development. Nor are other characters want- ing proving inferiority of organisation. We have here, therefore, another illustration of the very important fact — a fact sustaining the inference of the progressive evolution of higher from lower * The oldest fishes were, until recently, supposed to belong to the British Ludlow beds ; but the discovery, by Professor Claypole, of ichthyic fragments in deposits below the " Water-Lime" of Pennsylvania would seem to remove them still farther back in the geological scale. SILURIAN FAUNA. 143 forms of life — that the representatives of each class of animals were first ushered in in their simplest or most embryonic forms, and that not until these had attained a considerable development was there a noticeable appearance of the more highly constituted forms. It is a significant (even if not a very remarkable) fact that, prior to the first introduction of this lowest class of the Vertebrata, all the larger divisions of the Invertebrata, as now recognised by natu- ralists, had already come into existence. Of these, the diversity of form in the Silurian deposits, no less than the numerical de- velopment, is very great, and equally so in almost all the classes represented. The most marked feature of the Silurian invertebrate fauna, as contrasted with the Cambrian, is furnished by the corals, which, barring a few forms doubtfully belonging to the Cambrian of Sweden, have here their earliest representatives. These primitive types of the Actinozoa, as well as nearly all others of the Paleozoic series of deposits, have generally been recognised by naturalists to constitute two well-defined groups, the Tabulata (Favosites, Haly- sites, Heliolites, Alveolites, &c.) and the Rugosa, or cup-corals (Cy- athophyllum, Streptelasma, Omphyma, Zaphrentis, &c.), in both of which the calyces are divided up into superimposed chambers by transverse plates or tabulae — the former with very rudimentary septa, the latter with the septa well developed, and the outer calicular wall greatly thickened. In the majority of these cup-corals the septa are disposed in multiples of four (Tetracoralla), whereas in nearly all recent Madreporaria this disposition is effected in multi- ples of six (Hexacoralla). Our recently acquired knowledge of the deep-sea fauna, and a more intimate acquaintance with the anatomy of some of the more aberrant species of coral, tend to show that the supposed sharp delimitation of the Paleozoic actinozoan fauna does not in reality exist. The tabulate corals, for example, whose final extinction with the Paleozoic era has generally been insisted upon as one of the most decisive of geological landmarks, would seem to hold a number of forms more or less closely related to types living in the modern seas, which in themselves combine most diverse features in their organisation. The genera Halysites and Syringopora appear to be not distantly removed from the recent organ-pipe (Tubipora); Favosites is placed among the Poritidae; and Heliolites not impossibly represents an ancestral form of the 144 GEOLOGICAL DISTRIBUTION. group to which the modem Heliopora, recently claimed to be an alcyonarian, also belongs.* The rugose-corals, apart from the very limited number of forms which occur fossil in deposits newer than the Paleozoic — ? Holocystis (Cretaceous), Conosonilia (Tertiary) — have apparently two living representative types in Guynia and Hap- lophyllia. On the other hand, the modern star-corals, if we exclude from this group the Favositida3, have but a feeble development in the earlier deposits, although several recent families are represented (Poritida3, Eupsammidse, Astra3idse). Of the genera belonging to this group, Protaraea, Stylarsea, 1'risciturben, and Calostylis date back to the Silurian period. The remarkable development of corals in the Silurian seas makes it not a little difficult to account for the total, or almost total, ab- sence of their remains in deposits of the preceding Cambrian age. It is scarcely credible that the animals of this class should not have already then existed ; but if so, what has become of them ? In ex- planation of this anomaly some geologists have urged that the strata of Cambrian age which contain recognisable fossils are all of a deep-sea origin, and that in the shallow and littoral deposits, where we might be expected to look for the traces of the organisms in question, organic remains have been completely obliterated through rock-metamorphism of one kind or another. The evi- dence supporting this hypothesis is, however, far from satisfac- tory, and the problem must be considered as one still awaiting solution. The graptolites, a group of organisms whose earliest remains are found in the transition rocks which unite the Cambrian and Silurian formations, and whose organisation appears to be most nearly reflected in that of the recent sertularians or sea-firs of the class Hydrozoa, constitute an important element in the Silurian * The reference of Favosites to the Poritidae, it must be confessed, is based upon rather slender evidence, and perhaps scarcely less so the placing of Heliolites among the Alcyonaria. Homes ("Elcmente der Palaeontologie," 1884) justly emphasises the artificiality of a classification in which the number and disposition of the septa and tentacles are made the basis for a division into primary groups, and in which other equally important characters are com- pletely lost sight of. The relationship existing between past and recent forms, taken in conjunction with the general homogeneousness or' character exhibited by the Tabulata, would seem to imply that the classification of the recent Actinozoa requires serious emendation. SILURIAN FAUNA. 145 fauna, becoming practically extinct with the close of that period.* It is a noteworthy circumstance in connection with the history of this family that the more complicated or double-stemmed forms, such as Diplograptus, Didymograptus, Phyllograptus, and Dichograptus, preceded, in the order of appearance, the simple-stemmed forms, like Monograptus and Rastrites, proving, contrary to what might have been naturally supposed, that the latter were not the ancestral types of the family. On any evolutionary hypothesis the simpler forms appear to have been brought about as the result of degeneration. In modern type hydrozoans the Silurian, as all other Paleozoic, de- posits are very deficient, a circumstance, doubtless, due in con- siderable part to the perishable nature of the organisms belonging to this class. The impressions of jelly-fishes have, however, been indicated in both the Cambrian and Silurian rocks of Sweden. Stromatopora, a very broadly distributed genus, whose affinities are now generally conceded to be with the Milleporida, passes into the Devonian formation. Of other Invertebrata, such as the echinoderms, mollusks, and articulates, there is a vast profusion of forms, which, apart from the mere matter of numbers, are in many respects sharply contrasted with their predecessors of the Cambrian period. The brachiopod mollusks, the predominant forms of which, as Spirifer, Atrypa, Athyris, Strophomena, Rhynchonella, and Pentamerus, belong to the group of the Brachiopoda articulata, are structurally consider- ably in advance of the inarticulate genera Lingula, Lingulella, Discina, and Obolus, which make up almost the whole of the corresponding Cambrian fauna; the latter, as far as is known, con- tains but a single precursor of the articulate division, Orthis. The prodigious development of the Silurian Cephalopoda would of itself be sufficient to distinguish the period from the period preceding. While up to the present time only two species of this class, an Orthoceras and a Cyrtoceras, are positively known from Cambrian deposits, no less than eleven hundred species, referable to a very considerable number of genera of Nautilidse — Orthoceras, Cyrto- ceras, Gyroceras, Endoceras, Gomphoceras, Phragmoceras, Lituites, Nautilus, &c. — have been described from the Silurian basin of Bohe- mia alone. The total number of species of this period may be * The somewhat problematical Dictyonema passes into the Devonian: Triplograptus, if a true graptolite, is Devonian. 146 GEOLOGICAL DISTRIBUTION. estimated at between two and three thousand. The genus Gonia- tites, which effects a partial transition between the Nautilidae and the Mesozoic ammonites, appears for the first time in the later Silurian deposits. It is a surprising fact, considering the remarkable development of the Cambrian trilobitic fauna, that not only are none of the earlier species represented in the Silurian deposits, but that by far the greater number of generic types, and more particularly those which by a special individual or specific development are rendered most important, as Paradoxides, Dikelocephalus, Olenus, Sao, and Conocephalus, should be also wanting. Only seven out of some twenty-seven genera of the primordial zone connect the formations of the two periods. The Silurian Trilobita comprise probably in the neighbourhood of fifteen hundred species, referable to some fifty or more genera ; yet of this vast number there are barely a half-dozen epecics which transgress the boundaries of the formation, passing into the Devonian. Among the more abundantly represented gen- era are Phacops, Dalmania, Calymene, Asaphus, Trinucleus, Aci- daspis, and Chcirurus. Viewed irrespective of numerical development, the most signi- ficant feature connected with the Silurian invertebrate fauna is the introduction of the earliest " air-breathers. n Until recently these were supposed to belong to the period following, the Devonian, but the discovery of a true scorpioid (Palaeophoneus) in the Upper Silu- rian deposits of both Sweden and Scotland, and of an apparent orthopteroid (Palseoblattina) in the nearly equivalent deposits of Calvados, France, proves conclusively that a very considerable differentiation among the air-breathing arthropods had already taken place, and points to a period very much more ancient for the first origination of the group.* Devonian Fauna. — The primitive air-breathing arthropod fauna just referred to finds a somewhat larger extension 'in the rocks of Devonian age, where fragments belonging to some five or six spe- cieo cf insect, possibly representing as many genera — Platephemera, Gerephcmera, Lithentonrarn, Homothetis, Xenoneura, Discrytus — have been discovered. These appear to belong to the modern * The scorpion described "by Professor Whitfield (" Science," July 31, 1835) from the Upper Silurian rocks of Ke~w York may, as suggested by Mr. Pohlinan (" Science," September 4, 1885), prove to be a young euryptcroid. DEVONIAN FAUNA. 147 group of the netted -veins (Pseudoneuroptera and Neuroptera), although by some 'authors they, as well as all other Paleozoic insects, are considered to represent a distinct, and now wholly extinct, type of Insecta, the Palaeodictyoptera. In whichever way their relationship be viewed, there can be little doubt that they represent very nearly the lowest structural type of their class. It is a very remarkable fact that the wing venation of these primi- tive insect forms is practically identical with that which charac- terises the modern insects belonging to the same group or order; the vast lapse of ages between the Devonian period and our own day appears to have effected no essential modification of structure in this particular direction. Besides these neuropterous forms, certain wing fragments have been referred to the members of the higher order Orthoptera, to which the modern grasshopper and cockroach belong. But there seems to be considerable doubt as to the claims of the so-called Devonian cockroach,* and it would, perhaps, be as well to consider its position as still a matter of un- certainty. The marked differentiation exhibited by the Devonian insects indicates that they were far more numerous than would appear from the paucity of their remains, and the inference drawn as to their great antiquity has been confirmed by the discovery of the Upper Silurian form already referred to. Coincidently with the appearance of these early inhabitants of the land surface, we remark the first considerable development of a land vegetation, whose earliest traces are to be met with in the Silurian period. With but very few exceptions (certain forms, as Prototaxites, Ormoxylon and Dadoxylon, considered by some authorities to represent true conifers, the first of their kind) all the Devonian plants belong to the lower or non-flowering division, the Crypto- gamia, comprising a multitude of ferns, tree-ferns, giant representa- tives, like Sigillaria and Lepidodendron, of the modern club-mosses (Lycopodiacese), and scarcely less gigantic forms (Calamites, Cal- amodendron, with Annularia, Asterophyllites, Sphenophyllum) be- longing to the group of the horse-tails (Equisetacese). Of Invcrtebrata other than insects the Devonian fauna is very rich in forms; but these show a marked similarity to those of the * Ecfcrrcd by Ilagcn to the Neuroptcra. The same authority considers the age of the rock formation in which the other insect remains Lave been found as more likely Carboniferous than Devonian. 148 GEOLOGICAL DISTRIBUTION. preceding period. Approximately the same types are represented, and while in the case of certain families a diminution in the number and variety of their representatives is noticeable, in others there is a corresponding increase. The corals, echinoderms, cephalopods, and brachiopods have approximately the same value as in the Silu- rian period, and, indeed, several of the specific forms that are to be met with in the one formation are also seen in the other. Among the Devonian brachiopods we have the first appearance of the genus Terebratula, a form which has continued to flourish, although in constantly diminishing numbers, from that period down to the present time. With it are associated a number of other genera — Strophalosia, Productus, Uncites — which likewise appear here for the first time. The gasteropods are all of the holostomatous or round- mouthed type — Pleurotomaria, Murchisonia, Euomphalus, Loxo- nema, Holopea, Platyceras, &c. — while the lamellibranchs, which show a marked increase, both in numbers and variety of form, over their Silurian predecessors, belong, as far as it has been possible to ascertain, exclusively to the Integropalliata, or such as are devoid of a sinual inflection to the pallial line. The families represented are either Heteromyaria (Aviculidae, Mytilidae) OP Dimyaria (Nucu- lidae, Arcadae, Astartidae, Cardiidae), no true Monomyarian being as yet known. We meet here with the first pulmonate, Strophites, a member of the modern family of snails (Helicidae), and likewise with what appears to be the earliest unequivocal fresh-water inver- tebrate, a mussel of the genus Anodonta, or one very closely allied to it. It is here, therefore, that we have the earliest undoubted traces of a fresh- water formation. The trilobites among Crustacea manifest a very rapid decline, and, indeed, in some regions they appear to have completely died out with the close of this period. The giant eurypterids — Euryp- terus, Pterygotus, Slimonia — the most formidable of all known living and extinct crustaceans — which first appeared in the Upper Silurian formation, linger on into the succeeding Carboniferous period, when they forever disappear. The Devonian deposits, as has already been stated, contain the earliest remains of the highest order of crustaceans, the Decapoda, or ten-footers, which comprise the modern lobster and crab ; the form in question (Paleeopalaemon) belongs to the macrurous, or long-tailed division, and is allied to the shrimps. DEVONIAN FAUtfA. 149 The vertebrate life of this period exhibits a remarkable develop- ment as compared with that of the period preceding. But, as in the latter, all the remains belong to the class of fishes, and indeed principally, or one might say almost exclusively, to the same two orders, the elasmobranchs and ganoids. No animal of a grade higher than fishes had as yet appeared upon the scene, or, if possi- bly it had, no traces of it have thus far been discovered to indicate its existence there. To such an extent was the fish-fauna developed that the term u age of fishes " has not inappropriately been applied by paleontologists to designate this epoch of geological time. The preponderating types are the ganoids, which appear not only in forms that may be considered more or less remotely related to the type of the modern sturgeon (Macropetalichthys), or to the fringe- finned Polypteri of Africa (Holoptychius, Glyptolepis, Dipterus, Osteolepis), and the American alligator-gars (Chirolepis), but in such as have no representatives in any of the succeeding formations. These are the so-called "bucklered ganoids," which, in addition to the enamelled plates characteristic of this group of fishes, had the head and the anterior portion of the body encased in bony plates, more or less firmly united to each other, and serving as a protective armour. To this group, among others, belong Pteraspis, Cepha- laspis, Pterichthys, and Coccosteus, forms which had their forerun- ners already in the Upper Silurian deposits. Generally placed among the ganoids, and closely related to Coccosteus, are the giant Dinich- thys and Titanichthys, which appear to have attained a length of from twenty to thirty feet, and whose dental apparatus closely approximates that of the modern Lepidosiren, one of the lung-fishes (Dipnoi), a group of animals which effect a transition between the true fishes and the amphibians. If this relationship with Lepido- siren be absolutely established, as is claimed to be the case by many of the more prominent anatomists, then it is certainly significant that the advent of the Amphibia (the class of animals immediately above the fishes), in the succeeding Carboniferous period, is preceded by just that group which, in accordance with the principles and work- ings of evolution, we should expect to find interposed — the group which, on the one hand, combines some of the characters of the Amphibia, and, on the other, those of the ichthyic fishes. But, whatever the exact relationship of Dinichthys friay be, there can be little question as to its representing a dipnoan type, or at least a 150 GEOLOGICAL DISTRIBUTION. transition form between the true ganoids and the lung-fishes. A similar position is occupied by some of the other crossopterygian fishes of the period, as Dipterus. Carboniferous Fauna The life of the Carboniferous period is marked by two important features: 1. The introduction for the first time of vertebrate animal forms higher in the scale of organisation than the fishes, i. e., the amphibians; and 2. The great develop- ment of strictly air-breathing or terrestrial animals. Of these last we have at least four distinct types indicated — the Gasteropoda, Insecta, Arachnida, and Myriapoda. Of the first, which have a solitary forerunner in the Devonian formation, we are acquainted with a comparatively limited number of forms (Pupa, Anthraco- pupa, Dawsonella, Zonites), all of them more or less closely related to forms still living at the present day. The insects comprise not only members of the low order of netted-veins, which are the only forms known to be represented in the Devonian deposits, but those of the more highly organised Orthoptera, and not improbably also Coleoptera (beetles), although most of the remains referred to the latter order are now positively known not to belong there. The Orthoptera comprise, among other forms, some sixty or more spe- cies of primitive cockroach, the Palaeoblattarise, which may be con- sidered to represent the ancestral type of the modern social pest (Blatta), whose earliest appearance dates from the Triassic period. To the same order belong the giant walking-sticks recently brought to light from the coal-measures of France, the Titanophasma Fayol- Isi, which measure in length (in one specimen) upwards of twelve inches, and are, therefore, by linear measure, very nearly the largest of recent as well as fossil insects. This extraordinary development of a form, which may be taken to represent the extreme term of specialisation in an insect, in a period so early as the Carboniferous, is certainly not a little re- markable, and argues very strongly for the great antiquity beyond its own period of the origin of this class of animals. It is also not a little surprising that no representatives of the family of walking- sticks (Phasmida), other than those found in the Carboniferous deposits of France — Titanophasma and Protophasma — have as yet been found in a fossil condition, except such as may have been preserved in amber. Of the Neuroptera, the Haplophlebium Bar- nesii, from Nova Scotia, attained an expanse of wing of seven CARBONIFEROUS FAUNA. 151 inches, nearly equal to the expanse of the largest of the living dragon-flies, to which it appears to have been related. The Car- boniferous Arachnida comprise representatives of true spiders (Pro- tolycosa, Anthracomartus), scorpions (Eoscorpius, Cyclophthalmus), and pseudo-scorpions (Microlabis). The scorpions appear to have attained a degree of specialisation very little below that of their modern representatives ; but the true arachnids have all, or nearly all, segmented abdomens, and may be considered to mark a transi- tion between the arthrogastric and anarthrogastric forms. The Myriapoda, which have a solitary forerunner in the Devonian rocks of Scotland (Forfarshire), are represented by both the cheilognath- ous and cheilopodous types, although on account of certain structu- ral peculiarities the greater number of these earlier forms (Eupho- beria, Xylobius, Trichiulus) have been constituted into a special order, the Archipolypoda. Of the remaining invertebrate fauna of the Carboniferous period little need be said. The various groups of the tabulate and rugose corals (Lithostrotion, Syringopora, Cyathophyllum, Amplexus. Za- phrentis), the brachiopods, pteropods, lamellibranchs, gasteropods, and cephalopods, among the mollusks, and the crinoids and blas- toids (Actinocrinus, Platycrinus, Cyathocrinus, Dorycrinus, Batto- crinus, Pentremites, Granatocrinus) of the Echinodermata, have, as in the Devonian formation, abundant representatives; but they belong in considerable part to genera which now appear for the first time, or to such as had but a feeble development hereto- fore. The widely distributed group of trilobites, which, as has al- ready been seen, played such an important part in the faunas, of the Cambrian and Silurian periods, has here barely four gen- eric representatives, Phillipsia, Proetus, Griffithides, and Bra- chymetopus, whose species occur in the main part in the deposits situated below the true coal.* With these forms the trilobites disappear forever from the scene. While the deposits of the pre- ceding Silurian and Devonian formations have shown a fair repre- sentation of at least two of the primary groups of the Echinoder- mata, the Asteroidea and Crinoidea, especially of the latter, it is not until the present period that the urchins themselves (Echinoidea) acquire any significance (Archaeocidaris, Palechinus, Melonites); * Professor Claypole has latterly announced the discovery of Daluiania in the " Waverly group" (Lower Carboniferous) of Ohio. 152 GEOLOGICAL DISTRIBUTION. and here, also, for the first time, if we except the Laurentian rocks, with the hypothetical Eozoon, do the Foraminifera appear to enter largely as rock constituents. The genus Fusulina is developed to an extraordinary extent, and its distribution appears to be but little, if at all, less universal than that of the genera Nummulites and Orbitoides of the Eocene period. A solitary forerunner of the Nummulites has been discovered in the Carboniferous rocks of Belgium. In the remarkable development of the elasmobranch (shark) type of fishes, and in the absence of the bucklered ganoids, the Carboniferous ichthyic fauna is sharply defined from that of the Devonian. With the exception of a considerable number of fin- spines or ichthyodorulites, referred to such genera as Ctenacanthus, Gyracanthus, Oracanthus, &c., whose position is still very doubt- ful, and which may in part belong to the order of ganoids, all the remains of the former appear to have been more or less nearly related to the modern Port Jackson sharks. These remains are in the main in the form of teeth — Psammodus, Helodus, Orodus, Chomatodus, Petalodus, Cochliodus — whose (somewhat distant) resemblance to the pavement teeth of the cestracionts has led to their reference to members of that group ; not impossibly, how- ever, they represent a very distinct type.* The ganoids comprise, in addition to polypteroid forms — Coelacanthus, Rhizodus, Megal- ichthys — representatives of the rhomb-plated Lepidosteidei, which include the American alligator-gar. The most widely distributed and most abundantly represented genera are Palseoniscus and Am- blypterus, the former of which is also one of the most abundant fishes of the succeeding Permian period. The only vertebrates other than fishes which appear in the Car- boniferous period, and now appear for the first time, are the Am- phibia, that group of animals whose members stand immediately next above the fishes in the scale of organisation, and wrhose em- bryonic forms are so clearly ichthyic as to have necessitated the union of the two classes into the one comprehensive division of the Ichthyopsida. It is not a little significant that the appearance of * Mr. Garman has recently described a species of shark from the Japanese seas, Chlamydoselacbus anguineus, which appears to be generically most in- timately related to the Carboniferous Didymodus, and which, accordingly, rep- resents about the most ancient type among living vertebrates. CARBONIFEKOUS FAUNA. 153 these animals should have been foreshadowed in the Devonian dip- teroid ganoids, which, leading up to the lung-fishes on one side, and not impossibly directly to the amphibians on the other, effect a transition to the higher class from the side of the fishes. This succession of higher upon lower types is not a matter of accident, but a direct outcome of the inevitable laws of evolution. Through the application of no other law would the numerous accidental or coincidental occurrences of direct succession, which present them- selves throughout the entire geological series, receive an intelligent explanation. All the Carboniferous amphibians belong to the ex- tinct order of the Labyrinthodontia (Stegocephala), salamandroids of both minute and gigantic frame, whose members were distin- guished by the possession of a dermal (cephalo-dorsal and ventral) armour of sculptured plates, and in many cases by a peculiar laby- rinthine infolding of the enamel of the teeth, a structure unknown among modern amphibians, but which is in great part shared by certain members of the ganoid fishes, as the modern alligator-gars (Lepidosteus) and the genus Rhizodus (Carboniferous). Among the genera are Anthracosaurus, Hylerpeton, Dendrerpeton, Batra- chiderpeton, and the caecilian-like Dolichosoma and Ophiderpeton. No other vertebrate higher in the scale of organisation than these labyrinthodonts is as yet apparent, unless, possibly, the very doubt- ful Eosaurus be proved to be a true reptile. The flora of this period partakes essentially of the character of that of the period preceding, the Devonian. We have here the same ancient representatives of the modern club-mosses and horse- tails, the Lepidodendra and their allies,* and the calamites, the ferns — Neuropteris, Pecopteris, Alethopteris, Sphenopteris, Cyclop- tens — giant tree-ferns, and forms that have been referred to the group of the cycads, an order of plants to which the sago-palms belong, and which appear to be not distantly removed from the conifers. No positive indications of the existence of any true flower-bearing herbaceous plants are yet manifest, and with their absence the total absence of flower-frequenting or nectar-sucking * The recent anatomical investigations of Eenault and Saporta have led these authorities to consider Sigillaria, at least in some of its recognised forms, to be much more closely related to the gymnospermous phanerogams than to the club-mosses ; but Professor Williamson has pretty definitely shown that such a relationship does not exist. 8 154 GEOLOGICAL DISTRIBUTION. insects, the Lepidoptera and Hymenoptera, the two most highly organised orders of insects, is noticeable. The only true trees, or such as are made up principally of woody tissue, of the Carboni- ferous deposits belonged to the coniferous series, the order of plants which embraces the modern pine and its allies. These an- cient evergreens were represented by several distinct genera — Dadoxylon, Palseoxylon, Pinites — which, if the fossil fruit asso- ciated with their remains, and known as Lepidostrobus, be justly attributed to them, had their nearest allies among their modern congeners in the berry-bearing yews. No deciduous leaf-bearing trees, such as the oak, beech, or maple, which make up the great mass of our forest growths, can be positively shown to have existed in these early days. Permian Fauna. — In the formations of the period succeeding the Carboniferous, the Permian,- a considerable advance in the structural type is indicated by the animal remains. While the predominant forms of Ufe of the period preceding pass, although in most cases with very diminished numbers, into the present one — in fact, to such an extent as to have induced many geologists to unite the formations of the two periods into a common whole — we meet here with a class of animals whose representatives had not hitherto been detected. These are the true reptiles, most of whose members belonged to the order Theromorpha (Pelycosauria), rep- tilian forms which in several important characters — the structure of the pectoral and pelvic girdles, humerus, and tarsus — show strong affinities to the lower orders of mammals, the Monotremata and Edentata, of which, not impossibly, they may prove to be the early progenitors. A further approximation to majnmalian structure is found in the character of the dentition, which in many forms ex- hibits a distinct differentiation into incisor and canine teeth. The deposits of the Southern and Western United States, especially of the State of Texas, have yielded a wealth of species and genera belonging to this order (Theropleura, Dimetrodon, Diadectes, Em- pedocles, Clepsydrops), representative of several distinct families. The modern type of lizards had their nearest analogues in the monitor-like Proterosaurus (Germany, England), whose dentition, however, was of the crocodilian type (thecodont). These early reptiles, while exhibiting many points of structure indicative of comparatively high specialisation, yet clearly proclaim a primitive PERMIAN FAUNA. 155 type of organisation in the rudimentary or embryonic condition of the vertebral column, which is in most cases only partially ossified. The Amphibia of the Permian period are by most authors placed in the single group of the Labyrinthodontia, although in certain structural departures from the normal type, as in the very rudi- mentary condition of the vertebral column, and in the absence of the peculiar labyrinthine infolding of the enamel of the teeth, some of the forms may have to be separated from this order. Most of the species were provided with a tail of greater or less length, and the general resemblance to living amphibians appears to have been mainly with the salamandoids, although in several points of struc- ture they more closely approximate the tailless frogs and toads. The relationship with the plated ganoids is well pronounced, and not improbably some of these, as dipneusts, or double breathers, may have been their true ancestors (as well as of the lung-fishes proper, Dipnoi). Among the more prominent genera are Branchio- saurus, Melanerpeton, Urocordylus, Archegosaurus, Eryops, Palaeo- siren, and Ophiderpeton, the last two apparently apodal, and re- calling the coecilians in outline. In Eryops megacephalus, the largest of American amphibians, from the Permian of Texas, the skull measures eighteen inches in length and twelve inches in breadth. The fish -fauna of this period partakes essentially of the char- acter of the fauna of the period preceding, from which it has borrowed most of its types. We have here, however, the first unequivocal remains of the genus Ceratodus (Bohemia and Texas), which represents the most ancient generic type of all existing Vertebrata. Regarding the invertebrate fauna of the period, it may be re- marked that a deficiency in the number of forms is noticeable in nearly all the localities where the Permian deposits are developed, a circumstance due to the peculiar physical conditions under which the deposits were formed, and the subsequent alteration, resulting in the obliteration of the contained organic remains, to which, in many places, the rock-masses were subjected. A very large pro- portion of the known fossils are, as has already been intimated, of clearly Carboniferous types, more especially in the case of the Mollusca. The trilobites, so characteristic of the earlier deposits of the Paleozoic era, are wholly wanting, not a single individual, 156 GEOLOGICAL DISTRIBUTION. apparently, of this very numerous order having survived the Car- boniferous period. Here, also, we have the almost final disappear- ance of the two great groups of the rugose and tabulate corals, which by their numbers so eminently characterise the limestone deposits of the Paleozoic series, from the Silurian to its close. The Permian flora is essentially that of the Carboniferous period, and requires no special consideration. Paleozoic Faunas. — Briefly reviewing the more salient features of the Paleozoic faunas, we find, as far as the invertebrate series is concerned, that with few exceptions all of its recognised classes have their representatives, or, at least, there are representatives of nearly all those classes whose members could reasonably be expected to have been preserved in a fossil state. Thus, of the Protozoa we have the Foraminif era and Spongida ; of the Ccelenterata, the Acti- nozoa and Hydrozoa ; of the Echinodermata, the Echinoidea, Aste- roidea, Ophiuroidea, Crinoidea, Cystidea, and Blastoidea ; of the Mollusca (and Molluscoida), the Polyzoa, Brachiopoda, Acephala, Pteropoda, Gasteropoda, and Cephalopoda; and of the Articulata, the Crustacea, Arachnida, Myriapoda, and Insecta. Of the classes here enumeratedthere are wanting in the Cambrian the Actinozoa, * and possibly also the Hydrozoa; the Echinoidea, Blastoidea, and Ophiuroidea, among the echinoderms ; and the Arachnida, Myria- poda, and Insecta, among the articulates. In the Silurian the number of missing classes is reduced by six, since we have here representatives of both corals and hydroids, blastoids and brittle- tars, insects and arachnids ; but the last two are represented almost by single individuals. In the Devonian the number is further re- duced by one, the class of the Myriapoda, likewise (as is also the case with the insects) represented in almost solitary individuals ; only with the Carboniferous period do all the classes acquire for the first time any marked development. We thus cannot fail to remark the progressive evolution of new forms correlatively with the advance of time. Of the vertebrate series the Paleozoic deposits contain the remains of only three of the five recognised classes, the fishes, amphibians, and reptiles, which appear serially in the order of their progressive organisation, the lowest, or fishes, in the Silurian (or, if the conodonts be fishes, in the Cambrian), the amphibians in the Carboniferous, and the highest, or true reptiles, * Some forma have been doubtfully referred to this period in Scandinavia, TRIASSIC FAUNA. 157 in the Permian. These ancient deposits have as yet yielded no traces of either birds or mammals. Triassic Fauna. — In the first of the Mesozoic series of forma- tions, the Triassic, we enter, as it were, upon an entirely new phase of organic development. Many of the more characteristic groups of organisms of the preceding era have now either completely dis- appeared, or only survived in such diminished numbers as to con- stitute but a very insignificant element in the new fauna. Thus, of the class Brachiopoda but comparatively few of the older generic types are represented ; and the same may be said of the other classes of mollusks, and more especially of the Cephalopoda. Of all the various forms of Paleozoic tetrabranchiates there are barely more than a half-dozen surviving types, and of these one, Orthoceras, itself becomes extinct in this period. But, in the place of these ancient types, we have others of the same class which are no less conspicu- ous for their numbers than for the complexity of form which they subsequently attain, and some of which exhibit a marked advance upon their predecessors in the scale of organisation. The ammo- nites, whose advent appears to have been foreshadowed in the goniatite and the Devonian Clymenia, now for the first time acquire any importance, and, indeed, if we except certain forms from the Carboniferous deposits of India and Texas — Arcestes, Xenodiscus, Sageceras, Medlicottia — now for the first time appear altogether. The numerous species which in some districts, more especially in the region of the Alps, crowd the deposits of this age, belong in principal part to the families Arcestidse and Pinacocera- tidae, as representatives of the leiostracous, or smooth-shelled divi- sion, and the Tropitida3, Ceratitidae, and Clydonitidse (with the somewhat aberrant genera Cochloceras, Rhabdoceras, Choristoceras, and Clydonites), of the Trachyostraca, or forms with strongly sculp- tured shells. In these deposits, also, we meet in the Belemnitidaj with the first unequivocal traces of the dibranchiate or two-gilled, cephalopods, which, if we except the Nautilus, alone of this class of mollusks inhabit the seas of the present day.* The rugose and tabulate * The view entertained by several eminent paleontologists, that the am- monites themselves represent dibranchiate forms, requires further support before it can be fully accepted. The evidence at the present time appears to be fully a.s much, if not more, opposed to this notion as it is in favor of it. 158 GEOLOGICAL DISTRIBUTION. corals have been succeeded by the modern type of the star-corals, Zoantharia perforata and aporosa (Montlivaltia, Thecosmilia, Isas- traea, Thamnastrsea, &c.), whose fragmentary remains build up giant reefs (Alps) ; and, similarly, the more distinctive ancient group of the Echinodcrmata, the crinoids, whose most characteristic repre- sentatives at this period are Encrinus and Pentacrinus, find their successors largely in the more modern Echinoidea, or true urchins (Cidaris, Hemicidaris, Hypodiadema). It is, however, in the vertebrate fauna that we find the most prominent feature separating the life of this period from that of any of the periods preceding. Not only do we meet here with the re- mains of fishes, amphibians, and reptiles, but with those of mam- mals, and not improbably also with the impressions or tracks of birds. Granting these last, which are, however, a little uncertain, it may be assumed that all the classes of the animal kingdom, as now recognised by naturalists, had their representatives. The fishes are still principally referable to the predominant type of the periods preceding, the ganoids, which also in a measure retain the embry- onic heterocercal tail, although a tendency towards homocercality is observable in some of the genera, as in Semionotus. The more numerously represented forms — Ischypterus, Catopterus, Semionotus — belong to the group typified in the American gar-fishes, and may be looked upon as the direct descendants of the Carboniferous and Permian Pala?onisci. The lung-fishes find an abundant repre- sentation in the teeth of Ceratodus, which, as has already been seen, dates from the Permian, and possibly from a still older period. This animal furnishes us with one of those rare instances where a genus of living vertebrates has been founded upon the fossil remains. The amphibians of the Triassic period showr but little advance over the type of their predecessors, all the forms still belong- ing to the single order Labyrinthodontia, some of whose members attained to prodigious dimensions (Mastodonsaurus, Labyrintho- don). To this group are probably referable the singular hand- shaped impressions of the animal known as Cheirotherium, or "hand-beast," originally supposed to have been an animal of the frog-type, but now assumed to have been a salamandroid, or animal allied to the newts, and, like them, provided with a tail, although possessing in the structure of the skull certain features belonging TRIASSIC FAUNA. 159 to the tailless amphibians — frogs and toads. The true reptiles ex- hibit a remarkable variety of form, and, as in the succeeding Juras- sic and Cretaceous periods, constitute the most marked feature of the faunal remains that have been left to us. Hence, by some geologists the collective era of these three periods — the Triassic, Jurassic, and Cretaceous — or what is generally known as the Meso- zoic, has been designated the "era" or " age of reptiles." The modern lizards and crocodiles had both their ancient representa- tives, the former as indicated by the genera Telerpeton, Hypero- dapedon, and Rhynchosaurus, and the latter by Stagonolepis, Belo- don, and Parasuchus. But besides these there flourished a multi- tude of reptiles belonging to several distinct and very widely removed orders, which have left, to our knowledge, no traces whatever of their existence in the present seas. Such are the South African Anomodontia, some of whose members, as Oudenodon, were totally destitute of teeth, and had their beaks encased in horn, after the fashion of the modern turtles (of which they may have been in part the progenitors); while others, as Dicynodon, possessed the horny mandibular apparatus of the former, but were provided, in addition, with a pair of huge and powerful teeth in the upper jaw ; * the Theriodontia (as represented by the South African Galesaurus), reptiles whose dentition partook of the char- acter of that of the ordinary Carnivora, and whose earliest types had already appeared in the deposits of the Permian period ; and the Plesiosauria, a group of essentially sea-inhabiting reptiles, which acquired a very considerable development in the Jurassic seas, and whose best known exponent is the Plesiosaurus. In the animals of this order the extremities of the limbs, both anterior and posterior, were encased in integument, and thus converted into nippers, very much like those of the whale, and admirably adapted for propul- sion through the water. The most characteristic genera of this period were Nothosaurus and Simosaurus. In Placodus the dental armature consisted in principal part of flattened plates, resembling the teeth of the pycnodont fishes, with wThich animals these reptiles were first confounded. * Professor Judd has quite recently (" Nature," October 15, 1885), an- nounced the discovery of dicynodont remains in the Elgin Trias of Scotland ; the group of animals had hitherto been known only from Africa, India and the Ural Mountains. 160 GEOLOGICAL DISTRIBUTION. The most remarkable of all the Triassic reptiles are the Dino- sauria, a group of the greatest importance when viewed from a tele- ological standpoint by reason of the many structural characters which separate them from the typical reptiles, and approximate them to birds. These avian characters are indicated principally in the structure of the powerful pelvic girdle and hind limbs, which depart very broadly from the normal type of reptilian structure. Thus, the pubic bones, in many cases, instead of projecting for- wards as in other reptiles, are directed backwards, more nearly parallel with the ischium, both bones therefore taking a position directed towards the posterior portion of the body, a feature char- acteristic of birds. In the hind limb, again, the ornithic char- acters are seen in the great cnemial ridge which is developed on the tibia, the gradual diminution of the fibula towards the distal extremity, the structure of the astragalus, and in the disposition of the digits, three or more, and their accompanying phalanges. The inner and outer digits are shorter than the rest, or quite rudi- mentary, and the third toe, as in birds in general, is the longest. There are good grounds for concluding that the bones of the limbs, and, doubtless, if this was the case, of some of the other portions of the trunk, were permeated with air-passages, as in birds. The structure of the fore limbs is still only imperfectly understood, but there is no doubt that in many cases they were but very feebly developed, being very much shorter than the hind limbs, and that progression was, either habitually, or at least at times, effected by means of the posterior appendages alone. The remains of these earliest dinosaurs are indicated both by the actual parts pertaining to the skeleton (Zanclodon, Thecodontosaurus, Amphi- saurus, Clepsysaurus, Bathygnathus), and by the foot-prints, many of them three-toed, that have been left implanted in the rock- masses. Some of these, which measure fully a foot, or even con- siderably more, in length, were originally supposed to represent the imprints made by the feet of giant birds, a suspicion strengthened by the serial arrangement in twos in which the tracks are disposed ; but now that the structure of the dinosaurs has been more accu- rately determined, and their ornithic characters and mode of pro- gression recognised, there can be little doubt that they represent the imprints of the reptiles belonging to this order. This view is further strengthened by the circumstance that no actual remains of JURASSIC FAUNA. 161 birds have as yet been discovered in deposits of Triassic age.* But it is not impossible, or even improbable, that some of the smaller foot-prints that are scattered about the larger ones, and which in some instances are disposed in a single series one in advance of the other, indicating a method of progression adopted by certain wad- ing birds, may actually be of an ornithic nature. However this may be, it is certainly a significant fact bearing upon the doctrine of evolution, that no unequivocal traces of birds have thus far been discovered in deposits antedating those which contain the remains of reptiles, which in their several characters most approximate the birds, and in reality effect a transition to them. The progressive evolution of advanced or most specialised types is here clearly indicated. The Mammalia, the highest class of vertebrates, appear for the first time in the deposits of this age. They are indicated by the teeth and fragments of jaws pertaining to two or three genera, Dromatherium, Microlestes, and Hypsiprymnopsis, forms, as nearly as can be determined, belonging to the low type of the Marsupialia, and, probably, more or less closely allied to the modern banded ant-eater (Myrmecobius) and kangaroo-rats (Hypsiprymnus) of Aus- tralia. Jurassic Fauna. — The life-history of the Jurassic period, while combining certain prominent features not hitherto recognised, pre- sents to us primarily an expansion of those characters with which we have just become acquainted. The remarkable group of the dinosaurian reptiles, whose development in the Triassic period had but barely passed beyond its own beginnings, acquires here re- newed importance, apart from the mere matter of numbers, from the circumstance of the gigantic and diverse forms which it includes. Four distinct types of this order are recognised, all of which had representatives in the Jurassic period: 1. The Sauropoda, lizard- footed vegetable-eaters, in which the anterior and posterior pairs of limbs were of nearly equal length, and whose progression was effected on all fours. Among the more important genera of this period belonging to the group are Atlantosaurus, Brontosaurus, Morosaurus, and Cetiosaurus, the first, from the deposits of the * Since the above was written announcement has been made of the dis- covery of the skeletal remains of a track-making dinosaur of the Connecticut Valley. Trans., New York Ac. Sciences, Oct. 26, 1885. 162 GEOLOGICAL DISTRIBUTION. Rocky Mountains, measuring from eighty to one hundred feet in length— the largest land animal with which we are acquainted. 2. The Stegosauria (Stegosaurus, Scelidosaurus), armoured vegeta- ble-feeding dinosaurs, some of them of gigantic frame, whose pro- gression, owing to the feeble development of the anterior pair of limbs, appears to have been in great part effected by means of the hinder extremities alone. In Scelidosaurus Harrisoni, from the Lias of Dorsetshire, the hind foot measured three feet and a half in length. 3. The Ornithopoda, bird-footed herbivores, with a very unequal development of the anterior and posterior appendages, the latter closely approximating the structure found in birds. There can be but little question as to the habitually erect posture assumed by such forms as Camptonotus, Laosaurus, and Iguanodon. Iguanodon Mantelli, a Cretaceous species, measured about thirty feet in length from the tip of the nose to the extremity of the tail. No member of this genus is known from the American deposits. 4. The Theropoda, carnivore forms, whose progression was largely erect, and assisted in many cases, probably, by the greatly developed tail acting as a fulcrum, in the manner of that organ among the kangaroos. This type, which is almost alone represented in the Triassic deposits, includes the most formidable members of the order — Megalosaurus, Allosaurus, Dakosaurus, the former appar- ently attaining a length of fifty feet. The genus Compsogna- thus, represented by a single species (C. longipes) from the Upper Oolite, possesses probably the greatest number of avian characters of the entire order, and is considered to stand in the direct line of the descent of birds. Monsters parallel to those of the land- surf ace inhabited the oceanic waters, such as the finned Plesiosaurus and Pliosaurus, and the not distantly removed Ichthyosaurus and its American tooth- less ally, Sauranodon. The geographical distribution of Ichthyo- saurus is a very remarkable one. While apparently the genus is completely wanting in the deposits of the New World, its range in the Eastern Hemisphere embraced very nearly its whole north and south extent, from Spitzbergen (Ichthyosaurus polaris, Triassic) to Australia (I. australis, Cretaceous). Its greatest development ap- pears to have been in the early part of this period (Lias). From the discovery of fragmentary parts of young individuals within the bodies of more fully developed ones, it has been conjectured that JURASSIC FATJtfA. 163 the animal was viviparous, a supposition in a measure strengthened by the ill-adaptation of its structure to breeding on the land-surface. Not impossibly, however, these animals may have been in the habit of devouring their young, or the young of allied species, as it seems many species of snake do at the present day. In the Jurassic rocks we meet with the first traces of that ex- traordinary group of reptiles, the Pterosauria, which, in the pos- session of a tegumentary membrane stretched between the greatly elongated outer digit of the anterior limbs and the bases of the hinder extremities, resembling in many respects the flying-apparatus of bats, were enabled to navigate the air in the manner of birds. To these last, the pterodactyls, as the members of this order are familiarly designated, were closely related in the general conforma- tion of the skull, the pneumaticity of the bones, and the presence of a well -developed keel to the sternum or breast-plate, a character among recent animals found only in birds and bats. But while possessing these and other avian features, the pterodactyls depart in many important particulars from the bird type, and notably in the presence of true teeth implanted in sockets, as in the Croco- dilia, the structure of the manus, the absence of a feathery integu- ment—the animal having been apparently provided with a naked skin — and the possession of a tail composed of distinct vertebrae.* Despite these important differences, however, it may, perhaps, be deemed doubtful whether the animals in question have not as much right to be considered birds as reptiles, the more so as the one great feature separating them from modern birds, the pres- ence of alveolar teeth, has recently been shown to be character- istic of some, if not of most, of the ancient birds. While, there- fore, it may not be possible to decide upon the exact position oc- cupied by these singular organisms, there can be but little doubt that they, or possibly some closely-allied predecessors with which we are not as yet acquainted, represent the primitive stock whence the type of the modern flying or carinate bird has been evolved. The birds would then have a double line of ancestry, the one here indicated, and another, culminating in the struthious or non-carinate * Professor Marsh lias shown that at least in some forms of pterodactyls (Rhamphorhynchus) the extremity of the tail was provided with a tegu- mentary expansion, or vertical rudder, by means of which the animal doubt- less guided its flight. K;J (|] OLOOK 11 Dl 1TBIB1 HOK 11 III I I),' .Illl.' MM Mil I, |<( ll, .. :M1,| IH :l I. -I in |.,. ,iM\ 11, -I ,h.liiill\ IVIUOVOll II.MH I h,- .liu ,i . .1. I ',MII|' ,"-n il 1m . Mi,- mow Import mil '<" IN ir •-.-m-i •> ,.i ri,-i.« inn i mr Pi.-i.-.l i, i \ hi.. Kh nn|>h,M 1)\ in hn :. ni,l 1 >im,'i )>h>»l,'n. \\lu.h .lillri li.Mnrn.h ,,|h,-i in unl\ in i •'. th.' (I, -nHli, !-< , -l\ i.-l u,-.l l,< lli,- ni,>,l,-i n I I'll,- ,i,<,-,. v n h ,MU, • i| i (•)>(,< .|>,Mi.l\ In ., Hi, -i i,. .n, IniO. i,. Hi, |.Miiiin\r :iut|>lu. ,i-l, MI . ,li\ i .i,-n i l\-l, -,..111111 .. M . nun .». ,M th,' ..- in \\ In, h lli, M. UtioU Oi thr », -lilln i. - I, IM V>. :< . .ll .11,1 Ml, .ll,-,l ll.MIl 111,' III ,,» th,- i 'H i i-'.l. « n'' \\ In, 1» :i|>|»,-:ii h.-ir I.M 1 1\,- lii .1 t iinr. , -\iui-r rknMc UiUVrvm -,1 m ih,-u ,li\.-i.,- I.MHM , ,.,.i|., i i ill. mow M,,I.,.U mi nuidorn •.!,,- t h, -]\ IUUO, \t\\\\\\ ' li, !\,ln,l.r I'll,- .,-, ,-nl .-,-nui tho l%inl«.-,K l>r.l: no :ul,lui,Mi (,< th,- I mm •>) (1- ,11 ,h,-,l IM I hi- IMI, I ' -t nn»->|i- i, Min, I ui tlio CiillUMi V \ ,>l t hr S.Mith t,, in, iii\ ...\ ti.,- 1 1 .t i bird kboul ti»«- y \\hirh i . KM.M\ n IM |wo U\i M.I ui,li\ i.lu il. , ,-inlMii, in inn, -i < -il »li -t mr mil. ,n. Mnn\.-»l tn i\ !«,- ,r". u.l.-.l i • ) i\ i' • t«f «-l i • •'• •• 1'hu-s, ut» of ttl'UrHux- ,»t tho -;UnU :»nil Hunk. MO .-I nin- '«l t'ui UMMM m lli M, iivvnth v .M ii \;. i. i \ i N \. n;;» I.,. ..I .Immiulivr (rclli liilpliililr.l in nni-KH t I In I , . m I It. ..(It. i liui.l, MM mini il. \\lii. Ii \vnii <>l iilxinl (In MI/C <>l V'hll'll \\ 1 1 1 1 |M.U rl Illl \\ III" i. :ll l< <|MMIMMIIV« ilu, p. il .nun" \\linlly 111 111 IIIIMI ipul ptil l.i Mi. .ml. i .,1 Mi. M .1 ii pi ill I ... p. HI, I.. .1 HIIMI il < I'.ol Ii Ml. M "I i MM.) I IK In I I.. ..i ...i r p. - hud fix ii M pi • i MI ill i ' • I).. I < i 111 ,1.1. DM \ll.pllll. I. I'l I.. I I). I MUM III. I \IM|, Illll MM, lll'l Mi. Ill i.i IM ri.M'iini, HK| ii, I* iiiii imp.' ni. i. iii ,i MM i IM.MV (III. IV, in.ll'. lypc «»l Mill uplil ill .. III. II . ,l I.. I ,".(, I. iH'MllfllllM, iMllll, I . pi- 'Ml MM I p. ..I I I I. .1 ||. |'|»|\ ,„,-, MM. I p. .Illl I I" ii i" IMM||»!O origin of I!*' MI.M|.,M pl..i« nhil mi uiii, II.M , ill. !'• ii "pi.,i I li.i-l. M • Ml] • il" I I. •'" 1I(" (l ' • OB "'' ' 'I-'1 -I- I MCI' ..I illll. M Ill li.ll'.M, MM. 1 1- 1 1 lip pill I III I y 1 1 ' • : I on li I. I . I.I. Ih HI Mini i \ hll.llrd |»y I In Mi ill III. pi. 'til Inn. l.ill.M', .,i .hi liilinhi, mill iioliililv I'M. I' "i M ii It, hnvr nl 1 1 MI pi . < I I.. li«.\v Mini MM . mo I mi. i< ill MI m, 111 .1 ..I lli< 'I i iiiMMin Illltl .llll'llftNld |i< ii'.ih w< ic M«I IMM 111 .11 Hpi .ii I. MM < ,M M..W rooognivodf Imi HI.MI,. •, i.iii.lihil. ih Inn I I.I«|CIM, A lliilhci in. iiiMi< i I i, iipml l.y UK Mi . I . MM H «!M mil :ipp. HI I'. In •Ulllli H III IMM Mini , IM liil MM . p ii ,1 I'.M In ii pn.p., i .. Ill ' hi.ii ..I MM II.M. IM" l.'ll '.'. Illl III'' 1. 1> <|.. MUM ,1.1 I. ihlM 'III' < .11' III I ll< Ml. MM I »M I ll Illl (I !>', I l» M IM i.l.i.lv l.y MM ' ' pi. .I'.p'.'l I, I IMI. II, I, ..,M< Ii IM.M i,M,p',.l , i ifii-,11. 1. '.M. ii.., ; ,pmi. MM ., 'I . i iill |M.M;-|I !ill nlln M Mil , .il.iiii.l IM! rid I'M, •-. i I,, in;/ Iliil pun Ml'. Hill iiiipni I.UM '• v. In. Ii 'II -.1 IMIMII I.' li 1 1 ' I 166 GEOLOGICAL DISTRIBUTION. theus, Arietites, Harpoceras, ^Egoceras, Stephanoceras, Lytoceras, Phylloceras), and beletfinite ; but representatives of groups that appear to have been closely related to the modern calamary (Belo- teuthis, Belemnosepia, Teuthopsis) are not exactly wanting. The lamcllibranchs and gasteropods comprise a most varied assemblage of forms, many of them but barely distinguishable from individual forms living at the present day, and by their great numerical de- velopment give a generally modern aspect to the fauna. Of the former the modern families Ostreidoa, Limidse, Mytilidae, Astartidse, Lucinidas, and Cardiida3 are remarkable for their profuse develop- ment, and scarcely less so the now nearly extinct Trigoniadaa and Pholadomyidffi ; of the latter, the more important families are still the non-siphonated ones (Pleurotomariida3, Naticidse, TrochidaB, Actseonidae) ; but a no inconsiderable representation of the Siphon- ata (Cerithiidae, Aporrhaidoe, Strombidas, Purpuridae) is also inter- spersed. The earliest fresh- water univalves belong to this period (Paludina, Melania, Neritina, Planorbis). The corals, which are of the type of existing star-corals, may be considered next in importance to the Mollusca, and, indeed, in some instances, as in the Coralline Oolite, they constitute by their own vast numbers the greater portion of the solid rock, not im- probably the vestiges of ancient reefs. Somewhat less important, but yet very abundant in certain localities, are the fragments of the Crinoidea, which are most distinctively represented by the genera Apiocrinus and Extracrinus, the latter having its modern analogue in the Pentacrinus of the Carribean Sea; but this comparatively little specialised group of the Echinodermata has, ever since the close of the Carboniferous period, been on its decline, and has left its place to be filled by the true urchins (Cidaris, Hemicidaris, Holectypus, Echinobrissus, Clypeus, Collyrites) and brittle-stars (Ophioderma, Ophiurella, Ophioglypha), both of which, but more particularly the former, now for the first time acquire any special importance. The star-fishes are represented, among other forms, by the type of the modern Uraster. Of the articulates there is a considerable development of the Crustacea — crabs, lobsters, and their allies — and in the fine-grained rocks the remains of centipedes, spiders, and true insects are not uncommon, the last comprising representatives of all the recognised modern orders. To this period belong the earliest Diptera, Hymenoptera, and Lepidoptera. JUEASSIC FAUN1 A. 167 Turning to the flora, we find it to be sharply defined from that of any of the Paleozoic periods, although in the abundance of ferns, many of them of ancient type, and in the absence of the higher forms of plants, it shows an interesting correspondence. Its most marked feature is furnished by the group of the Cycads, of which there are numerous genera recognised, and the pines, whose nearest allies appear to be the southern araucarias. The earliest undoubted representatives of endogens are found in the deposits of this age, some of them clearly indicating a close relationship with the Aus- tralian screw-pines (Pandanus). No positive traces of exogenous plants other than conifers have as yet been determined, but it is by no means improbable that they already existed. Comparing the fauna and flora of the Jurassic period with the existing fauna and flora of any portion of the earth's surface, we remark a striking similarity to the conditions presented on the Australian continent. Here, at the present day, is the home of the marsupials, of the Port-Jackson shark, which had its Jurassic representatives in genera like Acrodus, Hybodus, and Strophodus, and of Ceratodus among the lung-fishes, a form which, though of more ancient date, also had its habitat in the seas of the Jurassic period. Only along the Australian coast do we meet at the present time with the lamellibranchiate genus Trigonia, one of the most characteristic and abundant of the Jurassic mollusks. As regards the flora, a no less striking correspondence is apparent. On the Australian land-surface flourish a considerable variety of ferns, tree- ferns, and cycadaceous plants ; likewise, the Araucaria type of Conifer ; and here, principally, do we find the singular plants al- ready referred to as screw-pines (Pandani). Australia is, in fact, that portion of the earth's surface which, as far as its faunal and floral characteristics are concerned, has undergone the least modi- fication since the Jurassic period, and, indeed, it may be said that the present fauna and flora of the continent became differentiated during the interval between the Triassic and Jurassic periods, al- though, as has already been seen, some of the distinctive types date from a more ancient epoch. The retention of an ancient type of fauna and flora clearly indicates that the continent had retained its isolated position through a period probably extending as far back as the Mesozoic era ; otherwise, if connection with some other continental land-mass had existed at some subsequent period, it 168 GEOLOGICAL DISTRIBUTION". would be barely possible that an interchange between the faunal and floral characters of that land mass and Australia should not have been effected to a greater extent than is indicated by the isolated position, especially of the fauna, now existing. Cretaceous Fauna. — The Cretaceous period presents in many respects a marked contrast in its faunal characters to the period preceding. While many of the most important or characteristic of the Jurassic invertebrate types still persist, in many cases with undiminished or even increased force, as, for example, the differ- ent classes of the Mollusca, we meet here with a development of other animal groups which in most of the periods preceding were of comparatively insignificant import. Such are the Fora- minifera, which in their various forms (Globigerina, Rotalia, Tex- tularia, Cristellaria) build up by their remains the great mass of the chalk rocks, whose enormous extension is one of the most im- posing monuments presented to the geologist. The sponges (Si- phonia, Jerea, Thecosiphonia, Ventriculites) here likewise find their greatest development, some of the forms having their analogues in the types that still inhabit the oceanic depths ; and the same has been shown to be the case with the Cretaceous urchins (Echinoidea), which are represented in great multitude and variety — Cidaris, Ananchytes, Galerites, Micraster, Discoidea. The corals are in comparison feebly developed, and can by no means claim that im- portance which they obtained in the Jurassic period. The Belem- nitidce (Belemnites, Belemnitella) and Ammonitidee still constitute the most important of the cephalopod types, the latter especially presenting a very considerable number of characteristic forms, the so-called unrolled ammonites — Crioceras (with Ancyloceras and Toxoceras), Hamites (Ptychoceras), Scaphites, -Turrilites, Helico- ceras, and Baculites. The bivalve and univalve faunas, while largely made up of Jurassic types, show a marked advance over the corresponding faunas of the period preceding in the much greater 'development of the siphonate forms. The Sinuata among the former, which, if we except the very abundant family of the Pholadomyidoe, had hitherto but scattered representatives, now acquire considerable importance, especially in the families Venerida?, Tellinidae, Glycimeridae, Ana- tinid^e, Mactridae, and Myidae. Among the non-sinuate forms the members of the oyster family (Ostrea, Exogyra, Gryphcea) and CRETACEOUS FAUNA. 169 the scallops (Pecten), and the genus Inoceramus among the Hetero- myaria, are distinguished by their numbers ; but the most charac- teristic elements of the lamellibranch fauna are furnished by two families of very inequivalve-shelled mollusks, the Chamidae, with the genera Requienia, Monopleura, Caprina, and Caprotina, and the so long misunderstood Rudistae (Sphserulites, Radiolites, and Hippurites), whose forms so eminently characterise the southern belt of European and American Cretaceous deposits, and which ap- pear and disappear with this period. The siphonate univalves have an almost exclusively modern aspect, and comprise among others representatives of the families FusidaB, StrombidaB, Muricidae, Tri- tonidse, Buccinidse, Cancellariidae, Pleurotomidae, Conidae, Olividae, and Cypraeidae. Turning to the vertebrates, we find in the lowest class, Pisces, the introduction for the first time of teleosts, or true bony fishes, that ichthyic group which at the present day surpasses, both in individual members and variety, all the other orders of fishes put together. These earliest teleosts, although not very abundant, comprise a considerable number of modern types (Clupea, Esox, Osmerus, Beryx) ; but it is not till the Tertiary period that they acquire any well - marked development. No amphibian remains have been detected in any Cretaceous deposit. Reptiles, on the other hand, are exceedingly abundant, and comprise most of the types whose existence has been indicated in the Jurassic seas. Thus, of the modern groups, we have turtles, lizards, and croco- diles (of both the amphiccelous — Hyposaurus — and procoelous types — Holops, Gavialis), and, in addition, the first true serpent (Si- moliophis). The extinct orders Ichthyosauria and Plesiosauria are still represented, and in Elasmosaurus, belonging to the lat- ter, we meet with one of the most formidable types of the finned Reptilia. Here, also, are found some of the most gigantic of the Dinosauria — Iguanodon, Megalosaurus, Hadrosaurus, Camarasau- rus — and the remarkable group of the Pythonomorpha, or "sea- serpents " — Mosasaurus, Leiodon, Clidastes — which in several re- spects united the characters of both serpents and lizards. The largest of the pterodactyls, or flying reptiles, having an expanse of wing of from twenty to twenty-five feet, or even more, occur in deposits of this period, and are represented by the normal- toothed types, and by such, as the American Pteranodon, in which 170 GEOLOGICAL DISTRIBUTION. the jaws appear to have been encased in horn, and to have been entirely edentulous. Bird remains are sufficiently abundant in certain localities, many of them belonging to forms seemingly not very far removed from some of our modern groups. But, in addition to these ordinary forms, we have some of the most extraordinary of any that have ever been described, and which, from the presence of true teeth in their jaws, have received the name of Odontornithes (toothed-birds). In the genus Ichthyornis, as exemplified in I. dispar, which was of about the size of a pigeon, in addition to the peculiarity of alve- olar teeth that of biconcave vertebrae is presented, a structure of the vertebral column characteristic of fishes and many of the ex- tinct reptiles, but not known in modern birds. The wings ap- pear to have been well developed, and in this, and all other respects beyond those just mentioned, the animal conformed strictly to the modern type of bird structure. In the still more remarkable Hes- perornis, which in the species H. regalis attained a height of five or six feet, the teeth, instead of being implanted in distinct sockets, were placed in a continuous groove ; the extremity of the upper jaw appears to have been bent down in the form of a beak, and to have been edentulous. The breastplate was entirely destitute of a keel or ridge for the attachment of the powerful muscles required for the motion of the wings, so that the bird was doubtless completely de- nied the power of flight. The presence, in the same geological period and the same geographical area (Kansas), of two birds so closely related to each other in the presence of jaw-teeth, and yet so distantly removed from each other by other peculiarities of structure, argues strongly for the antiquity of this class of animals, and, though the earliest unequivocal traces of 'birds have thus far been met with in the deposits of the Jurassic period, it is more than probable that their first origin is considerably more ancient. No traces of any mammalian have thus far been discovered in any indisputably Cretaceous deposit, a circumstance in great part attributable to the particular conditions under which most of the deposits of this period, as known to us, were laid down, namely, their marine origin. But there can be no doubt that at some future day such remains will be found, and, indeed, if the deposits of the Laramie age be conceded to be absolutely Cretaceous, as is claimed (although on most contradictory evidence) by many geologists, then CRETACEOUS FAUNA. 171 the first of such remains, the Meniscoessus, has quite recently been discovered. As now generally recognised, the Laramie deposits constitute a series intermediate between the Cretaceous and the Tertiary, the faunal characters, as are principally indicated by the abundant remains of dinosaurian reptiles, pertaining to the former, while the plants point directly to the latter. The angiospermous exogens, whose earliest undoubted remains occur in the Upper Cretaceous deposits, here undergo a very considerable develop- ment, and may, indeed, be said to represent the stock whence the floras of the subsequent Tertiary and existing periods have been derived. We find here many of our most common modern types, such as the oak, beech, poplar, tulip-tree, magnolia, alder, and plane. Tertiary Faunas,— With the close of the Cretaceous period and the beginning of the Tertiary, we note the most marked of all the organic changes that characterise the different geological epochs. Whole series of animals, from the lowest almost to their highest di- visions, suddenly become extinct, or so nearly verge on extinction as to constitute but a very insignificant element in the succeeding fauna ; on the other hand, groups of equal or greater importance, and which had hitherto no (or but very scanty) predecessors, just as suddenly make their appearance. It would seem as though a fresh start had been taken in the peopling of the earth's surface, so different in many respects are the faunas of the Cretaceous and Tertiary periods. But this difference, as it now presents itself, must not be taken to indicate that it in fact even existed as such. The gaps that now separate the one fauna from the other were un- doubtedly filled by animal types of intermediate grade, of whose existence we shall only be made cognisant when the hiatus which here breaks into the continuity of the geological system will be more completely filled in. It is illogical, and directly opposed to the workings of evolutionary force, to conceive of a wide-spread group of animals suddenly appearing and springing into prominence ; and no less illogical to conceive of an equally sudden extermina- tion. Hence, where vast differences in the faunas of any two suc- ceeding geological periods present themselves, we have reasonable grounds for concluding that a long lapse of time has intervened between the close of one period and the commencement of the period (as represented) next succeeding — in other words, that there 172 GEOLOGICAL DISTRIBUTION. is here a geological break. Only there where the continuity of the geological system is complete, or where the imperfection of the record is reduced to insignificance, can we hope to meet with an organic chain whose continuity is likewise complete. No such complete record, or anything approaching it, has as yet been dis- covered, nor is it at all likely that one will ever be discovered. But the gaps in the record that occur in one locality or country may be wanting in another, those present here be absent in the third, and so on ; hence, by a series of comparisons made between several localities, we can in a measure realise a comparatively perfect record, or at any rate one in which the breaks have been materially narrowed, and with it also a comparatively perfect organic chain. Except possibly in one or two regions of the earth's surface, New Zealand and California, nothing that may be said absolutely to link together the Cretaceous and Tertiary deposits, at least those of the marine series, has as yet come to light; the faunas are largely distinct, and their distinctness is the index of the inter- val that separates the outgoing of the one and the incoming of the other. The Tertiary fauna presents to us a clearly modern aspect, and one that characterises all the animal groups represented, from the lowest to the highest. And the farther we advance in this period the more modern becomes the general faunal facies, so that in the Pliocene, or uppermost division, not only are the genera largely identical with existing ones, but (if we exclude the vertebrates) also the species, notably among the mollusks. It may be stated in a general way that all the more comprehensive of the animal groups now existing are represented in the Tertiary deposits, and the majority of these date from the Eocene, or earliest division. We have no longer representatives of those wonderful reptilian orders, the Ichthyosauria, Dinosauria, Pythonomorpha, and Pterosauria, which characterised the greater portion of the Mesozoic era, and continued to its termination ; nor do we find any vestiges of 'the scarcely less wonderful birds of the odontornithic group,* or of the type rep- resented by Archaeopteryx. Both reptiles and birds belong to * An exception may, perhaps, be made in favour of the Odontopteryx, de- scribed by Professor Owen, which has the substance of both jaws developed into well-pronounced serrations (or false teeth), an exaggeration of the character exhibited by ducks and geese, to which the bird appears to have been related. TERTIARY MAMMALIA. 173 the type of existing orders, and the same may be said of the fishes, principally teleosts. The change in the character of the inverte- brate fauna is somewhat less marked than in the case of the verte- brates; but yet certain important differences present themselves. Thus, among the acephalous and gasteropod mollusks, by far the greater number, in fact nearly all the types, are referable to exist- ing families, and even in the oldest division, the Eocene, to exist- ing genera, or to such as are very closely allied to them. Such characteristic families as the Hippuritidae and Caprotinidae, among the bivalves, have completely disappeared, and, if we except some half-dozen or more species found in Australian Tertiary deposits, the same may be said of the Trigoniadse, as well as of the Am- monitidse* and Belemnitidse among cephalopods, about the most distinctive of the invertebrate forms of the entire Mesozoic series. Among the Tertiary invertebrates must be noted the extraordinary development of the foraminiferal forms Nummulites and Orbitoides, which, by their prodigious numbers, make up some of the most stupendous deposits known to us. But that feature of the Ter- tiary fauna which above all others arrests attention is constituted by the class Mammalia. The most striking fact that presents itself in connection with the history of these animals is their very sudden introduction, both as to individual numbers and diversity of form, almost with the beginning of the period, a circumstance of no little significance when it is remembered that, in the period preceding, if we except the doubtfully placed Meniscoessus, not even a trace of their existence has been detected, and that all such forms as have been found in the earlier Jurassic and Triassic deposits belong, as far as we are able to determine, to the single order of the Marsupialia. In the earliest division of the Tertiary, the Eocene, on the other hand, we meet with the remains of individuals belonging to at least one- half of all the recognised orders of the present day.f Thus, we have marsupials of the opossum type (Didelphis), insectivores, rodents (as represented by the Sciurida3, or squirrels), cetaceans * A few ammonitic fragments have been found in the Tertiary deposits of the Tejon group of California, and a Tertiary belemnite is claimed for Aus- tralia. t The Ornithodelphia, Edentata, Proboscidea, Hyracoidea, and possibly also the Sirenia and true Garni vora, are still unknown. 174 GEOLOGICAL DISTRIBUTION. (Zeuglodon), ungulates, both odd-toed and even-toed (among the former the tapiroid Lophiodon and Paluaotherium, and other such forms as Eohippus and Hyracotherium, which, through a series of modified but closely-related types in the Miocene and Pliocene peri- ods— Anchitherium, Hipparion — can be traced genetically to the modern horse ; and among the latter the possible ancestors of some of the modern deer, Xiphodon and Anoplotherium, and the suil- line Anthracotherium and Palaeochoerus), bats, even of existing gene- ra (Yespertilio, Vesperugo), lemurs, or lemurif orm insectivores (Ada- pis, Necrolemur), and not impossibly also the true monkeys. But while most of the forms found in these earlier Tertiary deposits are referable to modern orders, there are others which would appear to have no place in the classification laid down for living forms, and which combine, in many respects, the characters of two or more orders. Thus, it has been convenient to designate an order Ambly- poda for a line of animals which, at the one extremity, stand near- est in their relationship to the Proboscidea, or elephants, and at an- other to the odd- toed ungulates. In it are comprised the Uinta- theria, ponderous tusked-animals, rivalling or exceeding in size the modern elephant, and the coryphodons, considerably smaller ani- mals of a generalised type, the probable progenitors of the last. A still earlier type is embodied in the Condylarthra (Phenacodus), from the very base of the Eocene, which represent the most primi- tive type of known ungulate animals, and which not impossibly are derivatives of some preceding hoofed marsupial. Another order, the Tillodontia, has been established for certain animal forms which, in several respects, combine the characters of the insectivores, ro- dents, and edentates; and, again, a fourth order, the Creodonta, for forms that seem to hold a position intermediate between the insectivores and carnivores, and not unlikely represent the ancestral line of the latter. From the researches of paleontologists it would appear that the primitive type of placental mammal is the insectivore, and that from this original type have descended, by gradual modifica- tion, most of the varied forms that now people the surface of the earth, and those wrhose remains lie buried in the deposits of the Tertiary period. At what precise period in the earth's history the Insectivora first appeared it is impossible to say, for, although no remains occur in any deposits antedating the Eocene, there can be TEKTIARY MAMMALIA. 175 little or no doubt, seeing what modifications of insectivore structure are presented in the earliest deposits containing their remains, that they appeared at a very much earlier epoch. The Tillodontia, Creodonta, and Insectivora appear, as it were, simultaneously in the Lower Eocene deposits, and if, therefore, they represent merely modifications of one and the same structure, as is maintained by Professor Cope, who has united the three groups into the one com- prehensive order of the Bunotheria, then they must point to a com- mon progenitor (foreshadowed in the Jurassic insectivorous marsu- pials) removed far beyond the limits of the Tertiary period. Of the three insectivore types here indicated, the true Insectivora, which may be considered as the main or axial stem, have alone survived to the present time. The Tillodontia and Creodonta both became extinct before the middle of the Tertiary period, the latter, however, by gradual modification passing off into the Carnivora, whose earliest undoubted remains are to be found in the deposits of Oligocene, or Miocene age. From the same group of the Insec- tivora, although apparently at a somewhat later date than the Creo- donta and Tillodontia, appear to have been descended the so-called Prosimia, or primitive monkeys, the lemurs, whose earliest remains occur in deposits of both Lower and Upper Eocene age ; and to these last, again, is doubtless to be traced the direct line of ancestry of the various types of true monkeys that at the present day inhabit the earth's surface, and whose unquestionable traces are first met with in deposits of Miocene age. The most important non-insectivore type of Lower Eocene mammalian is the ungulate, whose remains, be- longing to both the odd-toed and even-toed sub-orders, occur in astonishing abundance, and argue very strongly in favor of a very remote ancestory, one that may not impossibly carry us as far back as the middle of the Mesozoic era. The progressive modifications of structure which can be traced through the more generalised of the Eocene mammalian groups results in greater and greater specialisation the further we advance in the course of time, and hence, in the Miocene period we meet with more of distinctly specialised (or isolated) groups than in the period preceding. In addition to the recent orders that have been enumerated as belonging to the Eocene period we have the Eden- tata (represented by such gigantic forms as Macrotherium and Ancylotherium, whose nearest relationship appears to have been 176 GEOLOGICAL DISTRIBUTION. with the aard-vark), the true Carnivora, Sirenia,* Proboscidea, and Quadrumana. Per contra, most of the older forms have now com- pletely disappeared, and, in fact, no mammalian order, with the possible exception of the Creodonta (Hyaenodon), is indicated which has not its living representatives at the present day. Most of the \ CREODONTA 'MJL PLIOCENE 'TILLODONTIA MIOCENE -(0-H-G-O-C-E-N-E) EOCENE BUNOTHERIA CRETACEOUS DIAGRAM ILLUSTRATING RELATIONSHIP OP TERTIARY MAMMALIA. families are such as still exist, and even many of the genera are identical, so that on the whole the mammalian fauna has a decidedly modern aspect. The Miocene Insectivora comprise, among other forms, representatives of the families of hedgehogs, shrew-mice, and moles ; the Rodentia, porcupines, mice, squirrels, rabbits, beav- ers, &c. ; the Cetacea are represented by true whales and dolphins ; the odd-toed Ungulata by the tapir and a number of allied tapir- oids, and the singular giant forms that have been, referred to the not distantly removed family of the Menodontidse (Symborodon, Titanotherium) ; by true rhinoceroses and other forms (Hyracodon, Aceratherium) closely allied to them ; and by Equidsc— Hipparion, Miohippus (Anchitherium) — which differed from the true horses * Eotheriura Egyptiacum, from the Mokattain nummulitic limestone, is re- ferred to this group by Professor Owen. TERTIARY MAMMALIA. 177 principally in size and the polydactyl character of the feet. Among the even-toed ungulates we find the hippopotamus, the true swine, deer, giraffe, and musk-deer, and, of the hollow-horned ruminants, the antelopes — the sheep, goats,* and oxen being still absent, al- though some of the antelopine forms would seem to effect a transition between the true antelopes and goats. The Proboscidea comprise, in addition to the true elephant (which, however, as a Miocene ani- mal is known only from the deposits of the Siwalik Hills of India, now frequently referred to the Lower Pliocene, or Mio-Pliocene), the Mastodon, and the aberrant Dinotherium, which was provided in its lower jaw with two prominent recurved tusks. The Carnivora have yielded representatives of the cats— the true Felis and the related sabre-tooths (Machairodus, Dinictis), the most formidable of all known recent and extinct Carnivora — the weasels, civets, hyenas (with the true hyena), and seals. The dogs are represented by the genus Canis itself, and the more primitive Amphicyon, through which a transition is effected to the ursine Hya3narctos, and to the Pliocene true bears. Finally, the Primates have yielded several genera, as Semnopithecus, Pliopithecus, and Dryopithecus, the last referable to the group of the anthropoid or highest apes, and fully equalling in size the human species. Passing on to the Pliocene period, the mammalian fauna makes a still further approximation to that of the present day in the in- troduction of a number of modern types that had not hitherto made their appearance, or only just appeared. Thus, we have here the camel (in India), the ox, true bear (in Europe), and horse, in addi- tion to most of the types that have been enumerated as belonging to the Miocene period, and it may be broadly stated that the ma- jority of the genera of this period are such as still exist, although the species are in most cases distinct. The regions where Tertiary mammals have been studied are principally the United States, Eu- rope, and India, between whose faunas there is a well-marked cor- respondence. While certain of the animal groups referred to are found in the one region and not in the other, and are therefore specially restricted, it may be said that approximately the same groups are represented throughout, although the date of their ap- pearance, or of that of their individual components, may be different for the different countries. Thus, in the Upper Miocene, or older * Capra Perimensis, from the island of Perim, is possibly a Miocene form. 9 178 GEOLOGICAL DISTRIBUTION. Pliocene, deposits of the Siwalik Hills of India we have the true hippopotamus, bison, bear, and elephant, forms which do not make their appearance until a somewhat later date in Western Europe, Pliocene, or Post-Pliocene. Similarly, the genera Cervus, Hystrix, Felis, Hipparion, and Mastodon, which appear in Western Europe in the Miocene period, are still wanting in the American continent, making their first appearance in the Pliocene. And, likewise, the true bears and oxen in Europe antedate the American forms by one period. It may be stated, as a general rule, that where identical genera of living forms occur in the deposits of both the Old and the New World, those of the Old World are the more ancient ; and the same probably holds good, although to a less extent, with families. From these differences in the dates of appearance of certain animal groups, their presence or absence, we are led to discuss the probable origin of our existing faunas, or portions of them. The existence in Western Europe in Miocene, and especially in Pliocene, times of a fauna consisting of forms which still inhabit the region, and of others as are only to be found at the present time on the continents of Africa and Asia, may appear at first sight somewhat singular. But when we reflect that the climate, during the whole or the greater portion of this period, was probably very much more uniform and warmer than it is at the present time, and possibly not very different from what it now is in the region of the Tropics, the apparent singularity in great measure disappears. Some of the tropical forms, as the giraffe and rhinoceros, may have been indigenous to the region, while others, whose development in South-Central Asia appears to have taken place at an earlier period, not improbably represent immigrants from the heart of that con- tinent. It is practically certain, moreover, that direct land commu- nication existed during a considerable portion of this period with the continent of Africa, with which, consequently, there would have been effected a general interchange of forms. Indeed, it is much more singular that Europe no longer retains its more charac- teristic African forms; but it must be recollected that, with the advent of the Glacial period, an era of cold set in, and that with this inclement climate a general retreat southward took place, the more tropically constituted animals passing over into the conti- nent of Africa, or suffering extermination by the cold. The most northerly animals, passing southward, occupied the region now more ORIGIN OF EXISTING FAUNAS. 179 or less vacated by the tropical forms, and hence, in the deposits of the Glacial and Post-Glacial periods, we meet with the remains of the elk, reindeer, hyena, lion, giraffe, elephant, rhinoceros, and hippopotamus mixed together. At about the same epoch it would appear that the present rupture existing between the African and European continents was effected, a separation which precluded the possibility of a return migration from Africa when the more toler- able climate succeeding the close of the Glacial period set in. Hence the survival only of the more temperate forms of the Euro- pean fauna. While Africa, therefore, retains its strictly African mammalian forms, it may be considered questionable whether these are not direct importations, in great part, from the region lying to the north. As far as the North American continent is concerned, it appears not improbable, from what has already been said with refer- ence to the earlier appearance in Europe of equivalent mammalian types, that a not inconsiderable portion of its later (fossil) fauna was derived from the Old World. Thus, it appears likely that the bears, swine, oxen, sheep, antelopes, and elephants originated in the Old World, whence they were transplanted by way of some land connection existing in the north into the New World. The true dogs, on the other hand, seem not unlikely to have been de- veloped first on the American continent ; and it is also not improb- able that the ancestral line of the camel is to be traced to the West- ern Hemisphere. Our paleontological knowledge of the different countries, even of those more thoroughly explored, is, however, far too insignificant to permit of a definite solution to the problem of the origination of the various mammalian groups, and in the case of most continental faunas but little can be said with positiveness concerning their formation. The North American continent lacks in its Tertiary fauna the giraffe, hyena, and hippopotamus; nor do we find any traces of the group of the Old World monkeys, or, in the greater portion of the region, of the Edentata, whose various forms are now so abundantly represented in the South American fauna. With the Post-Pliocene period the correspondence existing be- tween the fossil and recent faunas of the several geographical regions is in most cases further increased, and not only by the introduction of many new modern genera, but by the presence, in considerable number, of identical specific types. The modern fauna may now 180 GEOLOGICAL DISTRIBUTION. be said to be broadly marked out. Thus, in the Australian Pleis- tocene marsupials, Diprotodon, Nototherium, Thylacoleo, and their allies, we have the forerunners of the various marsupial forms that now characterise the continental fauna ; in the giant birds Palapte- ryx, Dinornis, Mionornis, &c., from New Zealand, and ^Epyornis from Madagascar, the forerunners of the wingless apteryx and the struthious birds from the same or neighbouring regions; and in the giant South American edentates, Glyptodon, Megatherium, My- lodon,* and their allies, the representative, if not the ancestral, forms of the existing sloth, armadillo, and ant-eater. It is re- markable that these last forms also occur in the Post-Pliocene de- posits of the United States ; but there can be but little doubt that their presence there was the result of a temporary migration from the south, since their remains are only exceptionally found in any of the preceding Tertiary formations. We have noted in the Eocene period the presence of a certain generalised group of mammals, from which, by gradual modifica- tions of structure, the more specialised groups of subsequent periods have sprung. The demonstration of this successive evolution of forms is not, however, restricted solely to groups of animals, but it can be indicated with no less positiveness in the case of certain in- dividual members of a group. The most notable instance of evolu- tionary modification in a given line is afforded by the horse. Thus, from the modern horse we can trace downward in the geological scale a gradual series of modifications in the structure of the teeth and limbs which, at the further end of the line, chaiacterise an animal so far removed in general structure from the existing form that, were not the intermediate forms known, or were it to be con- sidered by itself, it would be recognised not only as the type of a distinct family, but of a distinct sub-order. From the solidungu- late type represented in the existing form we reach, by sensible gradations, an animal of the polydactyl type, or one having several toes to each foot. A phylogenetic line, but little less complete, can also be traced in connection with the families Camelida3, Tapiridse, and Felidse, and others, doubtless, will be discerned with the fur- ther progress of paleontological investigation. * The deposits containing these remains have been very generally consid- ered to be Post-Pliocene ; Ameghino and Cope, however, probably correctly refer them either in whole, or in part, to the Pliocene. II. Appearance and disappearance of species. — Reappearance. — Extinction. — Per- sistence of type-structures. — Variation.— Geographical distribution. — Cli- matic zones. — Synchronism of geological formations. IT is assumed by all, or nearly all, geologists, that every species of animal, broadly speaking, had a definite belonging in the geo- logical scale; in other words, that its existence was coincident with a certain period in the development of the earth, and with no other. Thus, the well-known and largely-represented brachiopod, Spirifer disjuncta (Verneuilii), whose occurrence has been noted in North America, throughout the greater extent of continental and insular Europe, in Asia Minor, China, and New South Wales, is everywhere restricted to the Devonian formation, and is, therefore, distinctive of that period. Similarly, the no less widely disseminated Pro- ductus semireticulatus, a member of the same group of animals, is restricted to the Carboniferous formation. So limited, indeed, ap- pears to have been, in most instances, the range in time of a given species, that the inspection of a single well-determined form will frequently fix, not only approximately but absolutely, the horizon of the deposits whence it was obtained. Belemnitella mucronata, one of the squids, characterises a definite horizon of the Upper Cretaceous ; and among the Ammonites we have numerous instances of specific restriction to special " zones " of even the minor divi- sions of a formation. Limitation of range appears to pertain more strictly to the members of the higher groups of animals than to the lower, or to such forms whose complexity of organisation might be supposed to interfere with a ready accommodation to changing physical conditions of the surroundings. Not one of about one hundred and twenty-five species of fish described from 182 GEOLOGICAL DISTRIBUTION. the Old Red Sandstone of Great Britain appears to have lived on into the succeeding Carboniferous period, whereas, of the Inverte- brata, comprising the corals, annelids, echinoderms, crustaceans, and the several classes of the Mollusca, the survivors of the Devo- nian period number no less than twelve per cent. An equally striking case is presented by the Mammalia, where, of the very numerous forms that have been referred to the European and North American Tertiaries, not one is positively known to have passed either from the Eocene division to the Miocene or from the Mio- cene to the Pliocene ; and but a mere handful, if that, from the Plio- cene to the period next succeeding, the Quaternary or Post-Plio- cene. And yet we are aware that, in certain Eocene localities, no less than five per cent, of the molluscan fauna has survived into the present epoch, and as much as thirty to forty per cent, from the Miocene ! Turning to the Invertebrata themselves, we find among the different classes no less striking confirmation of the law of the persistence of the less highly organised specific types over those more highly organised. Taking the sub-kingdom Mollusca, for example, we find, from an examination of the carefully prepared tables of Mr. Etheridge B5 on British Paleozoic fossils, that, of the three great classes whose members are not free-swimmers, and who would be consequently most likely to fall under the influence of special physical conditions, the order of persistence is, in most cases, Brachiopoda, Lamellibranchiata, and Gasteropoda — i. the Coralline Zone (from fifteen to forty or fifty fathoms), the zone of the encrusting Algae (nullipores), and of the large car- nivorous Gasteropoda (Buccinum, Fusus, Pleurotoma, Natica, &c.); 4, the Deep-sea Zone, or that of the Brachiopoda and deep-sea corals (from fifty to two hundred and fifty or three hundred fathoms) ; and, 5, the Abyssal Zone, from three hundred fathoms to the oceanic floor, throughout which the shells are generally of small size, trans- lucent (thin), and white or but feebly coloured. The visual organs are here exceptionally devoid of pigment, and blindness has been noted in a few species of normally seeing gasteropods. It might be doubted, however, whether the last two divisions ought not more properly to constitute a single division, considering the large pro- portion of genera and species which pass from the shallower to the deeper parts. Dall, from the data furnished by the dredgings made in the Gulf of Mexico by the steamer " Blake" (1877-78), affirms that fully twenty per cent, of the molluscan species obtained DISTRIBUTION OF MOLLUSCA GENERALLY. 2G3 from that region enjoy a vertical range extending from less than fifty to two hundred and fifty and two thousand fathoms,* which would then give a very high proportion for those connecting the fourth and fifth zones. At depths of from three hundred to one thousand fathoms mol- lusks are still numerically very abundant, although the number of species very rapidly diminishes. From an extreme depth of 2,435 fathoms (14,610 feet) the "Porcupine" obtained but five species, and from 2,900 fathoms (17,400 feet) the "Challenger" dredged only two — Semele profundorum and Callocardia Pacifica. In the greatest depths the Lamellibranchiata (principally represented by the families Arcadae, Nuculida3, and Pectinidas, the genera Pec- chiolia, Nea:ra, &c.) appear to preponderate over the Gasteropoda, whose dominating forms are tectibranchs and the Scaphopoda, and, among the prosobranchs, the genera Fusus and Pleurotoma. \ The question whether the abyssal fauna is of a generally uni- form type, marked by identical or representative species extend- ing from pole to pole, as was first suggested by Lov6n, and subse- quently admitted by Sir Wyville Thomson, still lacks the necessary data required for its solution. The extremely broad or antipodal ' *"Bull. Mus. Comp. Zool.," vi., 1880. In a supplemental note only fifty-one species (out of four hundred and sixty -two) are recorded whose range covers both the littoral and the abyssal zones, thereby reducing the ratio to eleven per cent. t The researches of Edgar Smith and Boog Watson upon the Mollusca obtained by the " Challenger" reveal some very remarkable instances among this group of animals of ready adaptability to the most varying conditions of depth, and of discontinuous habitation. Silenia Sarsii was dredged about 1,100 miles southwest of Australia in water of 1,950 fathoms, and again off the mouth of the Rio de la Plata, in 2,650 fathoms ; Verticordia Deshayesiana, found off Pernambuco in water of 350 fathoms, was also dredged off Cape York in 155 fathoms ; Petricola lapicida, a well-known West Indian form, recurs off the North Australian coast (seven fathoms) ; and Nuculina ovalis, a fossil of the Suffolk Crag, reappears in the waters of the Cape of Good Hope in twenty fathoms. Venus mesodesma, a shore species, descends to 1,000 fathoms, while the range of Lima multicosta extends from two to 1,075 fathoms, and of Area ptcroessa from 390 to 2,050 fathoms (the West Indies and the North Pacific, respectively). The total number of lamellibranch species ob- tained in depths under 100 fathoms was 384 ; in depths between 100 and 500 fathoms, 148 ; between 500 and 1 ,000 fathoms, 24 (ten stations) ; and "between 1,000 and 2,900 fathoms (thirty-three stations), 70. (" Challenger" Reports, " Zoology ,»xiii., 1885.) 264 GEOGKAPHICAL AND GEOLOGICAL DISTRIBUTION. distribution of many of the species has led to the belief that such is the case, but, on the other hand, certain facts that have recently been brought to light seem to point in the contrary direction. Thus, it has been shown by Mr. Dall that of the species composing the abyssal fauna of the Gulf of Mexico only ten per cent, are such as may be termed boreal, a very small proportion to what might have been expected, were it to be assumed that the peopling of the cold bottom wastes was effected by a descent from the polar regions. On the other hand, thirteen per cent, were found to be tropical, and seventy-five per cent, uncharacteristic, forms. It is concluded from these facts, and from the circumstance that "the tropical forms belong to the same groups as those characteristic of the local littoral mollusk fauna," that in all probability "the abyssal regions have local faunae proper to their various portions, and that a universal exclusive abyssal fauna, so far as mollusks are concerned, does not exist." This conclusion, which was concurred in by the late Mr. Gwyn Jeffreys, receives further support, it is claimed, from the distinctness of the "Challenger" Mollusca as compared with those of the "Blake." The hypsometrical distribution of the Mollusca is governed almost exclusively by conditions of climate and food-supply, the influence of the latter being manifest in the intimate relation which binds many of the species to the plants upon which they habitually feed. Thus, in the Higher Kabylia, Aucapitaine has framed three molluscan zones, each corresponding to a particular plant growth : 1. The zone of the ash, olive, and pomegranate (450 to 2,100 feet); 2. That of the oak and pine (2,100 to 3,600 feet); and, 3. That of the cedar and green turf (3,600 to 7,200 feet). The upper limit to which mollusks attain on the continent of Europe (Alps) is about eight thousand feet, somewhat below the line of perpetual snow ; along the region of the equatorial Andes and the Himalayas the line is placed at about twice this height, also approximating the snow-level. Five species of fresh- water shells, of the genera Planorbis, Paludestrina, and Cyclas, were found by Morelet to inhabit Lake Titicaca at an elevation of nearly 13,000 feet, while irom the Himalayas Anadcnus Schlagintweiti has been obtained at a height of 16,500 feet, and LimnaBa Hookeri at 18,000 feet. Of the North American land shells the representa- tives of extreme hypsometric range appear to be Pupa alticola and GEOLOGICAL DISTRIBUTION OF MOLLUSCA. 2G5 Vallonia pulchella.96 The various hypsometric zones that have been established by conchologists differ at almost all parts of the earth's surface, and are of but local import. The shells of the more elevated mountain-summits are many of them, or mostly, of types which are found at lower levels in regions of reduced average annual temperature, following the well-known law of climatic dispersion which we recognise among plants. Vertigo alpestris, an inhabitant of Scandinavia, reappears in the Alps of Switzerland, although completely wanting in the intermediate regions; and, similarly, many Alpine summits hold identical or representative, species which are wanting in the connecting lowlands. Geological Distribution. — The most salient fact that presents itself in connection with the past distribution of the Mollusca is the reversed order to what might have been expected of the suc- cessive development of its primary classes. Thus, almost every- where, the Cephalophora, or head-bearing mollusks, antedate by one full period the Acephala, or headless forms, which indisput- ably represent a lower grade of organism; and among themselves the first to attain a maximum development are the cuttle-fishes (Cephalopoda),- which are structurally the highest. But two species of this group — an Orthoceras and a Cyrtoceras — are positively known from the Cambrian formation, while in 1868 Bigsby enumerated no less than one hundred and thirteen species of Gasteropoda as belong- ing to the same period of time. On the other hand, the total number of gasteropod species credited by the same author to the Silurian deposits is about eight hundred, whereas Barrande has described upwards of eleven hundred species of Nautilidae from the Upper Silurian deposits of Bohemia alone. As has already been intimated, there is a marked deficiency of Cambrian lamellibranchs, and even in the Silurian formation the number of species is com- paratively limited. Bigsby, in 1808, enumerated, all in all, some six hundred and thirty-six species, or but little more than one-half the number of Upper Silurian cephalopods of the Bohemian basin. The complete differentiation which the different classes of the Mollusca had already attained in the Silurian period argues for a great antiquity beyond that period of the members of this group. To what extent the time measured by the Cambrian period, and the interval intervening between it and the Silurian, may have been effective in bringing about the various changes, cannot be at pres- 266 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. ent determined ; it would appear at first sight as though the length of its duration were sufficient, for how otherwise could we in- telligently explain the total or nearly total absence of the members of at least two of the molluscan group from deposits in which the representatives of other groups are sufficiently abundant ? One fact, however, must not be lost sight of in this connection, and that is, that in these earliest deposits the obliteration of organic remains has been most excessive, and that not improbably the absence of the required forms is to be attributed rather to the destruction of parts than to an actual non-existence in the region. For the present it is impossible to affirm whether the Cephalophora came into existence before the Lamellibranchiata or not, but the evidence scarcely appears sufficient for considering the latter as a race derived by degeneration from the former, as has been pre- sumed by Professor Lankester. The sudden decline of the Cephalopoda (Nautilidaa) after the close of the Silurian period is very remarkable, and scarcely less so their rehabilitation under the form of their successors, the Am- monitidaB, in the deposits of the Mesozoic era ; Barrande enumerates but two hundred and forty- two species from the Devonian forma- tion, more than one- half of which belong to the genus Orthoceras, and the remainder principally to the genera Cyrtoceras, Gyroceras, and Gomphoceras. About an equal number are indicated from the deposits of Carboniferous age, where also we find much the same genera represented, although with different specific relations. The genus Nautilus now for the first time acquires any importance, and it and Orthoceras alone of the limited surviving members of the family of the Permian period transgress the boundaries of the Paleozoic era. The latter genus disappears .early in the Trias, while the former steadily increases in number, until in the Creta- ceous deposits it attains its maximum development, with a repre- sentation of some sixty or more species. The displacement of the Nautilidae by the Ammonitida3 is, if nothing more, certainly an interesting circumstance, and leads one to inquire what special advantage the latter may have possessed over the former in the struggle for existence, by means of which they triumphed over their predecessors. For there can be little question that the Ammonitidse, despite certain peculiarities in their structure, which are not as yet comprehensible to us, are the truly GEOLOGICAL DISTRIBUTION OF CEPHALOPODA. 267 modified descendants of the nautiloids, a transition to which ap- pears to have been effected by way of the genus Goniatites and those forms of the Carboniferous period (India and Texas) which, like Arcestes, have the sutural plication intermediate between what is seen in Goniatites (Silurian-Permian) and Ceratites (Triassic). Just where the embranchment from the nautiloid line took place it has been impossible to determine, but it is significant that the most nautiloid form of the Ammonitidse, the Goniatites, appeared after the Nautilidae had attained their maximum development, and some time after the genus Nautilus had itself appeared. Sutner estimates that there are in the neighbourhood of four thousand species belonging to the group of the ammonoids, nearly all of which have the foliated sutures characteristic of the true ammo- nites. With the exception of the genera Goniatites and Clymenia, and the primitive ammonitic forms of the Carboniferous rocks already referred to (Sageceras, Arcestes, Xenodiscus, Medlicottia, Cyclolobus *), and a single form which occurs in the lowest member of the Californian Tertiaries (the Tejon group), all the species are restricted to the Mesozoic deposits, which by their great numerical development they might be said to characterise. Probably no other group of invertebrates exhibits such a remarkable series of devel- opments corresponding with successive periods of time as do the Ammonitidse, and in none do the species appear to be so distinc- tively characteristic of certain horizons (zones of ammonites). The singularly diversified types of the Triassic period, which combine all the various sutural modification seen in the goniatitic stage (Sageceras, Lobites), the ceratitic (Tyrolites, Celtites), and the ammonitic, from the simplest to the most complex (Pinacoceras), are almost wholly wanting in the Lias, where an entirely new series of forms begins (^Egoceras, Harpoceras, Amaltheus). These in turn are succeeded by groups more or less distinctive of the different Jurassic zones (Oppelia, Stephanoceras, Lytoceras, Phylloceras), which in the main die out before the close of the period. Most of the Cretaceous forms belong to genera or sub-genera which have not hitherto been represented, and here for the first time do we find any great development of the remarkable groups of uncoiled ammonites — Scaphites, Hamites, Turrilites, Crioceras, Ancyloceras (also Juras- sic), and Baculites — whose advent seems to be foreshadowed by the * Cyclolobus has the typical ammonitic sutures. 268 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. Triassic genera Choristoceras, Cochloceras, Rhabdoceras, &c., which, in their departure from the normal ammonitic type, resemble some of these forms, but which differ in the simple character of the sutural foliation. Considering the Ammonoidea to be the modified de- scendants of the Nautiloidea, we have presented the somewhat anomalous fact that while at the beginning of their existence in- volution was the order of development, towards the end this develop- ment was marked by a contrary evolution, with an accompanying approximation in outline to that of the primitive type-forms. No facts with which we are at present acquainted permit us to state what was the underlying principle involved in these reversed changes. It can merely be said that, involution having once set in, any broad departure from the type newly attained must almost necessarily have been accompanied by a certain amount of evolu- tion. The persistence of type -structure among the Nautiloidea, and the relation which the geographical distribution of the Ammonoidea bear to supposed climatic zones, have already been discussed in previous sections. Still more involved in doubt than that of the Ammonoidea is the ancestry of the Belemnitidse, which, as the earli- est representatives of the two-gilled order of cuttle-fishes, first appear in the Trias, and practically disappear with the close of the Meso- zoic era, one speciec only, and that somewhat doubtful — Belem- nites senescens, from Australia — being reported to pass beyond the boundaries of the Cretaceous period.* That the group, however, represents the ancestral line whence the recent Sepiophora (Sepia) have been derived there can be but little question, seeing how close is the relationship between the determining parts — internal skeleton — of the fossil and living species. In the Eocene genus Belosepia the phragmocone is of a somewhat transitional character. The modern pen-bearing cuttle-fishes or calamaries (Chondrophora) ap- pear to have their direct ancestors in the various forms of Teuthidae, whose remains, in a more or less perfect state of preservation, occur in the Liassic and Oolitic deposits. Of the two other classes of Cephalophora, the Gasteropoda and Pteropoda, only the former acquire any geological importance. Beginning with a comparatively limited number of forms in the * The reference of this form to the Belemnitidae is considered more than doubtful by Branco ("• Zeitschrift d. deutsch. geol. Gesellschafl," 1885). GEOLOGICAL DISTRIBUTION OP GASTEROPODA. 269 Cambrian period, they steadily increase in number, until at the present day the number of known species is far in excess of that recorded from any other period of geological time. According to the estimates made by Bronn between the years 1862 and 1866, which may possibly still serve as a basis for the computation of successive ratios, although no longer abreast of the times, the numerical distribution for the several geological eras is as follows : Paleozoic. Mesozoic. Cainozoic. Eeceut. Scapliopoda No. of species. 22 737 141 1 No. of species. 48 1,764 3 152 No. of species. 65 4,622 1 185 530 No. of species. 50 7,500 54 825 5,700 Prosobranchiata Ilcteropoda (inclusive of the Bellerophontidae) . . Opisthobranchiata. . Pulmonata * With very few exceptions (Dentalium, Pleurotomaria, Capulus, Natica, Narica, Emarginula) all the Paleozoic genera are extinct, or at least generally considered to be so, and it is still question- able whether they include even a single siphonate form. The refer- ence of Fusus, Pyrula, and a few other members of the Siphonata to this period, probably rests on unsatisfactory determinations, al- though, indeed, no special reasons can be assigned for the non- extension back of such genera. The absence of prominent characters in many of the species renders their determination difficult or im- possible, and it is by no means improbable that a fair proportion of the genera which have received distinct names are in reality identical with modern genera otherwise designated. The generally holostomate character of the Paleozoic Gasteropoda imparts to the fauna a peculiarity which eminently serves to distinguish it from the similar fauna of the later Mesozoic and Tertiary eras, especially the latter, where the Siphonata largely preponderate. The dominant Paleozoic genera are Pleurotomaria, Murchisonia, Euomphalus, and Loxonema, whose special development throughout the greater part * The Pulmonata are now known to be represented by a limited number of species in the Devonian and Carboniferous deposits. The number of Paleozoic species indicated by Bigsby (1863-'78) is about treble the figure given by Bronn. 270 GEOGRAPHICAL A^D GEOLOGICAL DISTRIBUTION. of the Paleozoic series tends to link the different members to- gether. The earliest great differentiation in the ancient gasteropod fauna is seen in the Jurassic deposits, where a host of new forms, especially of the Siphonata (of the families Cerithiidae, Nerinaeidae, Apor- rhaidie, Strombidae, Buccinidse, Purpuridae, Columbellidae, &c.), are for the first time introduced. One or two of these families appear to have had their representatives already in the Trias, but they were there of insignificant import. It is not, however, until the Cretaceous period that many of the more distinctive of the modern families (Cypraeidae, Cassididae, Ficulidae, Tritonidse, Muricidas, Volutidae, Olividae, Cancellaridae, Terebridae, Pleurotomidae, and Conidae) appear, and of these a fair proportion of the genera date back only to the Eocene period. No recent species is recognised as extending back beyond the Tertiary series, and even in the Eocene the proportion of living to extinct forms is very slight, averaging not more than three to five per cent. ; indeed, it is ques- tionable whether any of the early Tertiary species can be identified with recent forms. The same is also possibly true of the Oligocene, but in the Miocene the percentage ranges as high as thirty-five, and in the Pliocene to seventy-five or more. Practically, all the Post-Pliocene forms are still living. It is impossible to arrive at any absolute estimate of the number of species occurring in each formation. Bigsby " enumerates some seven hundred to eight hun- dred species as belonging to each of the Silurian, Devonian, and Carboniferous periods, or very nearly that which is given by Zittel 98 for the Jurassic forms ; the Permian, which is deficient in nearly all forms of life, has but about thirty. The Eocene-Oligocene Paris basin contains, according to Deshayes, upwards of eighteen hundred species, or more than double the number of the entire Eocene shell fauna of the Eastern and Southern United States. The Miocene basin of Vienna holds upwards of four hundred species of Prosobranchiata. There are but very few truly cosmopolitan species of fossil Gasteropoda, although broad distribution was much more marked in the early periods of the earth's history than now. Thus, while from the American Tertiaries only a very insignificant number of forms could be selected which might in any way be correlated with contemporary European species, and but a fairly representative GEOLOGICAL DISTRIBUTION OF LAMELLIBRASTCIIIATA. 271 series from the Cretaceous, no less than two hundred and fifty out of a round six hundred from the Silurian deposits are stated to occur in Northern and Western Europe. Some seventeen or more species out of the fifty-two recorded by Etheridge from the middle and upper Paleozoic divisions of Australia find their analogues in the equivalent deposits of Europe, and not unlikely this identity will be increased on further comparisons being made. The Tertiary species of the two regions, on the other hand, are almost without exception distinct, and of the recent forms it may be doubted whether there is a single species held in common. The geological distribution of the Lamellibranchiata may be considered to run parallel with that of the Gasteropoda, and in a general way to partake of its peculiarities. Most regions are en- tirely deficient in Cambrian forms, and even in the Lower Silurian formations the number of species is rather limited. Barrande enu- merates upwards of eleven hundred species from the Upper Silurian formation of Bohemia alone, nearly double the number (636) that was assigned by Bigs by in 1868 for the Silurian deposits of the world generally, and considerably over that (918) which was claimed by Bronn in 1862 for the entire Paleozoic series. The oldest forms — i. e., Silurian and Devonian — belong almost exclusively to the Heteromyaria (Aviculidae, Mytilidse) and the Dimyaria (Nuculidae, Arcadae, Astartidaa, Cardiidae), although many of the Devonian forms that have been referred to the Heteromyaria may really be- long to the Monomyaria. The Sinupalliata among the Siphonida appear to be completely wanting, as they are likewise from the Carboniferous deposits, but it is by no means unlikely that some of the commoner genera, as Grammysia, Allorisma, Sanguinolites, Edmondia, &c., in which no sinual impression has been detected, are true members of families whose modern representatives are all furnished with retractile siphons. This supposition, which is based upon external resemblances and habits as deduced from the shell, is in full consonance with the theory of evolution, which would lead us to suppose that the direct ancestors of the Sinupalliata were closely resembling forms devoid of a sinual inflection. The refer- ence of the forms above mentioned to the family Pholadomyida3 may, however, still be considered uncertain. The Carboniferous Lamellibranchiata do not differ very broadly from those of the preceding (Devonian) period, except in so far as 272 GEOGEAPHICAL AND GEOLOGICAL DISTRIBUTION. pertains to the first considerable development of the Monomyaria (Pectinidae, Limidae). Here, too, if we except the still problemati- cal Pracostrea, we meet with the earliest indubitable remains of the oyster (Ostrea). As in the case of the Gasteropoda and Brachio- poda, the number of Permian species is very limited. The lamelli- branch fauna of the lower and middle divisions of the Mesozoic series, Trias and Jura, is characterised by a remarkable develop- ment of the families Ostreidae, Pectinidae, and Limidaa among the monomyarians, the Mytilidae and Pinnidae among the heteromyarians, the Arcadae among the Asiphonida, and the Astartidae, Lucinidse, and Cardiidae among the integropalliate Siphonida. The Phola- domyidae are especially abundant in the Jurassic deposits. Of the two most distinctive Cretaceous groups, the ChamidaB and the Rudistae, only the former have their Jurassic representative (Dice- ras). Beyond the great specialisation of those two families, the Cretaceous fauna does not differ essentially from the Jurassic, al- though the number of true sinupalliate forms (Veneridse, Tellinidae, Solenidae, Glycimeridae) is very much greater. Through the different divisions of the Tertiary, beginning with the Eocene, we see the gradual development which by almost im- perceptible stages leads up to the fauna of the present day. The Monomyaria, which in the Mesozoic period constitute nearly thirty per cent, of the entire lamellibranch fauna, enter upon their decline, and are succeeded, as well as the Heteromyaria, by the Dimyaria, of both the sinuate and non-sinuate types. The relation of extinct to recent forms in the different divisions of the Tertiary holds much the same as with the Gasteropoda. CRUSTACEA GENERALLY. Of the recent orders of Crustacea the only ones that acquire any geological significance are the Phyllopoda, Ostracoda, and Decapoda, although representatives of some of the other orders occur sparingly in formations extending as far back as the Devo- nian period (Praearcturus, among the isopods [?]), and possibly even to the Silurian (Necrogammarus, amphipod). It is a rather surprising fact that of the first two orders, if wre except one or two special types — Leperditia, Beyrichia — all the most abundantly represented genera of the earliest periods, as well as of the periods succeeding, are such as still hold considerable prominence at the DISTRIBUTION OF CRUSTACEA. 273 present day — Estheria, Cythere, Bairdia, Cypridina — thus present- ing one of the most remarkable instances of the persistence of type-structure known in the whole range of the animal kingdom. The genus Estheria dates from the Devonian period, and attains its maximum development in the Trias. In its modern distribu- tion it may be said to be almost cosmopolitan, although it would seem to prefer tjie regions of warm climate, and not to penetrate much beyond the fifty-fifth parallel of latitude. The range of some of the species is extraordinary. Estheria Dahalacensis, which oc- curs as far north as Vienna, is found from Sicily to the island of Dhalak, in the Red Sea, or over an area whose extent is meas- ured by about thirty degrees of longitude, and thirty-two degrees of latitude. The range of E. tetracera comprises fully forty de- grees of longitude (Oran — Kharkov), but is more than equalled by the Carboniferous species E. Leidyi, which has been reported from both England and the State of Pennsylvania. It is a singular circumstance that while all the recent species of Estheria are in- habitants of fresh water, or of water which is but barely brackish, the fossil forms are frequently, or generally, found associated with distinctively marine types of organisms, indicating apparently for these species also a marine habit. While this may have been true, the association with fresh-water forms, which also occurs, tends to show that it was only partially the case. Possibly belonging to the order of the Phyllopoda, but by some authors placed among the Malacostraca, are the singular shield - bearing crustaceans of the Silurian period, Ceratiocaris, Dictyocaris, Discinocaris, Peltocaris, &c., whose affinities have been placed with the modern genera Nebalia (marine) and Apus (fresh- water). In Discinocaris the shield in some individuals measures as much as six inches across. Hymenocaris, which is exclusively Cambrian, represents the oldest type of this order of crustaceans. The genus Apus itself, which is almost universally distributed, and has been observed even in Norway at an elevation exceeding three thousand feet, appears as early as the Carboniferous period. Bran- chipus, although devoid of a head-shield, is stated by Woodward to be preserved as a fossil in the Eocene (Oligocene ?) beds of the Isle of Wight." Of the more important genera of recent Ostracoda, Cythere and Bairdia both date from the Silurian period, and Cypridina from 13 274 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. the Devonian; the fresh- water genus Cypris is unknown prior to the Carboniferous. The species of the first two genera, whose dis- tribution is almost universal, are particularly numerous in the Creta- ceous and Tertiary deposits, and present some remarkable instances of reputed longevity. Two recent species of Bairdia, B. sub- deltoides and B. angusta, are claimed by Gerstaecker 10° to be found also in the Carboniferous deposits, but it is by no means improbable that the identification rests on improper determinations. This is rendered the more likely, seeing how very few, compara- tively, are the recent forms that have been identified as occurring in the Tertiary deposits. Of some seventy species or varieties of Entomostraca described by Rupert Jones from the British Tertiary deposits from the Eocene to the Pliocene inclusive, not more than seventeen or eighteen are recognised by him as constituting a part of the recent fauna. A re-examination of the forms may possibly increase this number somewhat, but it is certainly very remarkable that of two hundred and twenty or more species of Ostracoda dredged by the officers of the "Challenger" expedition only three or four had previously been described by paleontologists. While many of the modern species have a broad distribution, the number of forms which are known to be in any way cosmopolitan is ex- ceedingly limited (species of Halocypris and Cythere). On the other hand, there are some very marked instances of antipodal reappear- ance. Several well-known British and northern forms have been identified from Kerguelen Island and other remote regions of the earth's surface. Although apparently penetrating to the profound- est depths of the sea, the number of both species and individuals rapidly diminishes beyond a comparatively shallow superficial zone. Only fifty-two species were obtained by the /'Challenger" from a depth exceeding five hundred fathoms, and but nineteen from below fifteen hundred. The much greater diversity of the shore fauna as compared with that of the open sea is shown by the fact that among the "Challenger" Ostracoda only twenty-eight genera were represented, whereas on the British coast alone there are at least thirty-one. Among the oldest known representatives of the order are the Primitia prima and Leperditia Cambrensis, from the St. David's (Lower Cambrian) rocks of Wales. The remains of decapod crustaceans in the Paleozoic rocks are exceedingly scanty, as indeed they are also in the earlier part of DISTRIBUTION OF CRUSTACEA. 275 the Mesozoic formations. In North America they have been traced back to the Devonian period (Palseopalremon ; — Macrura), but in the European deposits they are not known before the Carbonifer- ous age (Brachypyge ; — Brachyura ?). Whether the astaciform crus- tacean found in the mountain limestone of Ireland, and described as Astacus Philippi, was of a fresh-water habit, may still be considered as more than doubtful. The same doubt extends to the various As- tacomorpha of the Middle Mesozoic period — Eryma, Pseudastacus — and not until we reach the chalk of Westphalia do we meet with any undoubted remains of the genus Astacus itself. But even here the deposit in which the remains are imbedded is of a marine facies, and seems to argue that up to this time the crayfishes were by nature inhabitants of salt water. Too much weight cannot, how- ever, be attached to the negative evidence afforded by the absence of known fresh-water forms of this group, inasmuch as the chances for their preservation in deposits of this kind among the older rocks is very slim. At the same time, it is not a little singular that the extensive deposits of this nature of Wealden age should be en- tirely barren of their remains. In the Tertiary deposits indispu- tably fresh-water forms are met with, and it is not unlikely that the full differentiation of these types from those of a marine habit was effected somewhere about the close of the Mesozoic era. The lobster or homarine type has probably its oldest representative in the genus Hoploparia, which occurs in the Cretaceous and older Tertiary deposits. Of the modern genera of prawns, PenaBus ap- pears to extend back to the Lias, and Palajmon to the Tertiary period. The most important order of Crustacea, considered from the geological standpoint, is that of the Trilobita, which, apart from the simple fact of numerical development, acquires special signifi- cance from the circumstance of its representing the earliest animal group which attained to any prominence in geological history, al- though in point of actual appearance it would seem to have been preceded by the Brachiopoda and Annelida. Thus, from the basal portion of the St. David's beds of South Wales, which represent the oldest or very nearly the oldest of the fossiliferous rocks that have been thus far discovered, no trilobites are known, and it is not until a full thousand feet in the same series of deposits is passed that their remains are first met with. The number of 276 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. species of trilobites found in the inferior division of the Lower Cambrian of Britain (the Longmynd and Harlech groups) is about ten, representing six genera — Agnostus, Conocoryphe, Paradox- ides, Microdiscus, Palreopyge, Plutonia — and of the entire Cam- brian series one hundred, distributed as follows : Upper Tremadoc 12 (including 2 from the Upper Lingula). Lower Tremadoc 19. Upper Lingula 30 ( " 5 " Lower Lingula). Lower Lingula 22 ( " 14 " Menevian). Menevian 32 ( " 4 " Longmynd). Longmynd and Harlech 10. From the Primordial zone of Bohemia (Cambrian) Barrande recognised in 1871 only twenty-seven species, or but little more than one-tenth the number (252) which he claimed for the Cam- brian deposits of the world generally. This paucity, as compared with the richness of the British fauna, is the more surprising when we consider how largely in excess of the insular forms are the Silu- rian species from the same region. The Bohemian basin contained at the period stated about three hundred and twenty species, whereas the total number of species recognised at about the same time from the entire British Paleozoic series of deposits — i. e., from the Cambrian to Carboniferous inclusive — only slightly exceeded two hundred and twenty. Of the seventeen hundred species, representing seventy-five genera, tabulated by Barrande for the world at large we find (using his data) two hundred and fifty-two relegated to the Cam- brian formations, eight hundred and sixty-six to the Lower Si- lurian, four hundred and eighty-two to the Upper Silurian, one hundred and five to the Devonian, fifteen to the Carboniferous, and one to the Permian.* While these figures would indicate a pronounced culmination of the group in the Lower Silurian period, * The number of species has been very materially increased since the publication of Barrande's paper, but his estimates will still serve as a proper basis for the computation of ratios. The American Carboniferous species seem to fall not far short of a dozen by themselves. Considerable uncer- tainty still attaches to the stratigraphy of the forms that have been referred to the Permian of Germany and the United States (Guadalupe Mountains, in Texas and New Mexico), and it is generally assumed that there are no Per- mian species at all. DISTRIBUTION OF TRILOBITES. 277 it is a significant fact that in Bohemia, which stands next to Scan- dinavia in respect of number of species, the numerical ratio of Lower Silurian to Upper Silurian forms is as one hundred and eighteen to two hundred and five, reversing, apparently, the order of development. The very limited number of generic forms that pass from one major formation to another is remarkable. Barrande enumerates but seven of the twenty-seven Cambrian genera which pass over into the Silurian, and twelve of the fifty -five Silurian genera which reappear in the Devonian. The Carboniferous genera are but three or four in number (Phillipsia, Griffithides, Brachymetopus, Proe- tus).* Of the fifty -five Silurian genera, with three exceptions, all the forms are already represented in the lower division. The number of genera that extend through two or more formations is reduced to two or three (Phillipsia, Proetus). In their geographi- cal relations it may be said that broad distribution is the rule rather than the exception. Thus, of the forty-two genera of the Bohemian basin thirty are held in common with Sweden, and twenty- four with England. More than one-half (seventeen out of thirty) of the North American genera are also trans- Atlantic forms, and the greater number of these are widely distributed over the Euro- pean continent. It has generally been considered that the most widely distributed genera are those which also have the greatest vertical range; but the exceptions to this supposed rule are so numerous — Paradoxides, Agnostus, Trinucleus, Asaphus — that it may be doubted whether any value is to be attached to it. Nor can it be maintained that in all cases the genera having the longest range are those which have the greatest number of specific repre- sentatives, although this is more often the case than otherwise. Probably the greatest number of species represented in any one genus is exemplified in the case of the genus Dalmanites (Lower Silurian-Upper Devonian), of which there were up to 1871 some one hundred and twenty-nine known. The nearest approach to this (some one hundred and fifteen or more) is seen in the genus Asaphus, which is restricted to the Lower Silurian formation. The number of species that transgress the boundaries of any major formation is exceedingly limited. Thus, it is very doubtful * Professor Claypole has more recently described a species of Dalmanites (Dalmunia) from the Waverly Group (Sub-Carboniferous) of the United States. 278 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. whether there is even a single species from the Cambrian which also forms part of the Silurian fauna. There are apparently but three forms — Calymene Blumenbachii, Dalmanites caudatus, Sphac- rexochus minis— which in England connect the Lower and Upper Silurian faunas, and an equally small number which unite the Silurian with the Devonian. In Bohemia the number of Lower and Upper Silurian connecting- forms is somewhat larger; but even here the proportion to the entire fauna — nine out of three hundred and twenty-three — is very small. The limitation in most cases is even more pronounced, fixing the species to definite horizons of the broader formations. With respect to horizontal specific distribution, instances of broad dispersion are not exactly uncommon. A fair proportion of the Bohemian species, for example, are spread throughout Sweden, Italy, Russia, and England, and a number have also been indicated as occurring in the equivalent deposits of the North American con- tinent. Calymene Blumenbachii is the commonest form occurring on both sides of the Atlantic, in both regions being alike a Silurian and Devonian species. Coincidently with the decline of the Trilobita we note the ap- pearance of crustacean forms (Eurypterida) which hold a some- what intermediate position between these and the modern king-crab (Limulus), whose remains are first met with in rocks of Jurassic age. Of this singular order, which comprises the giants of the class, some half-dozen or more genera are recognised, whose com- bined range includes the Upper Silurian and Carboniferous deposits. In the Ludlow (U. Silurian) rocks of Britain there are no less than thirty-two species of this order, representing six genera— Eurypter- us, Pterygotus, Himantopterus, Slimonia, Stylonurus, and Hemi- aspis, the last a transitional type connecting the group with the Carboniferous limuloid forms constituting the family Bellinuridse (Bellinurus, Prestwichia, Euproops). The most ponderous indi- viduals of the order (Pterygotus Anglicus, Slimonia Scoticus), which are at the same time the largest of all known Crustacea, recent or fossil, measuring from five to six feet in length, do not appear until the Devonian period, or not until the group had attained to considerable development. The simultaneous appearance among the trilobites of the largest and smallest forms — Paradoxides, Agnostus — would seem to point to a contrary order of develop- DISTKIBUTIO^ OF INSECTS. 279 ment, but there can be no question that the first appearance of these animals long anteceded the Cambrian period. The total number of eurypteroid forms occurring in the American deposits is twenty-three (representing the genera Eurypterus, Dolichopterus, Pterygotus, and Stylonurus), six of which (Eurypteri) are Car- boniferous, two Devonian, and the remainder Upper Silurian.101 The oldest known limuloid form is the Neolimulus falcatus, from the Upper Silurian rocks of Lanarkshire, whose early appearance would seem to indicate that while the family to which it belongs (Bellinurida3) may stand in its relations intermediately between the eurypterids and the king crabs, its actual origination may be traceable, at least in part, to direct modification from the trilo- bitic type. INSECTS, The number of recognised species of insects is generally con- ceded to be upwards of 100,000, and by some authors is placed as high as 150,000, but it is very questionable whether these represent more than one-tenth of the number actually inhabiting the earth's surface. Probably not less than one-half of the indicated forms belong to the order Coleoptera, or beetles, which is by far the most numerously represented of all the orders. The Lepidoptera, or but- terflies, have thus far yielded some 15,000 species — or about one- thirteenth of the total number (200,000) estimated by Speyer for the world at large — and an equal number may, perhaps, with a cer- tain amount of accuracy, be credited to the Hymenoptera (bees, wasps, and ants), the Hemiptera (bugs), and Diptera (flies). The Orthoptera, or straight-winged insects, which include the locusts, grasshoppers, &c., are considerably less numerous, while the spe- cies of netted-veined forms (Neuroptera) probably do not much ex- ceed 2,000, or perhaps do not even reach this figure. Our knowledge of the general insect fauna of the globe is still too limited to allow of any satisfactory conclusion being drawn as to the geographical distribution of the class as a whole ; enough is known, however, to permit it being stated that practically every portion of the earth's surface harbours a more or less extensive insect fauna, so that the distribution of this class of animals may be said to be universal. While most numerously developed in the warmer or tropical areas, insects are by no means rare in the region of high latitudes, and, indeed, in some of the most north- 280 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. erly points reached by man they appear to be still remarkably abundant. The officers of the British North Pole Expedition, under command of Sir George Nares, brought home a surpris- ingly rich fauna from the region (Grinnell Land) lying between the seventy-eighth and eighty- third parallels of latitude, com- prising no less than forty-five species of true insects and sixteen arachnids, the former distributed as follows : Hymenoptera, five species (two humble-bees) ; Coleoptera, one ; Lepidoptera, thirteen ; Diptera, fifteen; Hemiptera, one; Mallophaga, seven; and Collem- bola, three.102 Among the Lepidoptera are a number of forms belonging to genera common in the temperate zones, such as Co- lias, Argynnis, Lycsena, &c., which appear the more remarkable, seeing that the species of this order are more limited in Green- land (with an insect fauna numbering eighty species), and that no forms are met with either in Iceland or Spitzbergen, although up- wards of three hundred species of insects are represented in the former. On the other hand, M. Bonpland observed butterflies on the slopes of Chimborazo at an elevation of 16,626 feet, or but 1,600 feet below the highest level (18,225 feet) reached by insects (Diptera), as observed by Humboldt, on the same mountain. Insect life flourishes also to a certain extent in waters of high temperature (hot springs) ; and even on the free surface of the ocean, most distantly removed from land, the officers of the * ' Challenger " circumnavigating expedition everywhere obtained one or more spe- cies of Halobates, a member of the Hemiptera, which is stated to live entirely at sea, and to carry its eggs about with it attached to its body. The distribution of insects is determined largely by climatic and general physiographical conditions, and also to, a great extent by the nature of the food-supply, many forms, as has already been seen, being dependent for their sustenance (whether in the larval or mature state) upon the development of a particular vegetable product. Indeed, localisation or restriction appears to be more frequently brought about as the result of changes in the character of the vegetation than as a condition depending upon the interpo- sition of physical barriers which, through their powers of flight, the animals of this class are in great measure able to overcome. The effect of climate is, however, well marked, and is seen in the general restriction of numerous forms to particular climatic zones. DISTRIBUTION OP INSECTS. 281 An extended meridional extension is more frequently to be observed among the tropicopolitan forms than among those which more properly belong to the temperate regions ; and this is especially the case among African insects, where frequently the same spe- cies is found to inhabit both the northern and southern parts of the continent. This condition is also to be observed in the case of northern forms, when mountain-chains, trending in a meridional direction, permit of easy access to regions of very varying physio- graphical features, which in their more elevated parts present conditions more nearly uniform with those exhibited elsewhere on the lowlands. Thus, we find in the Chilian fauna numerous forms that more properly belong to the north temperate zone — the pre- ponderating element among the Carabidso, for example, and the genera Lycsena, Colias, and Argynnis among the butterflies. These not improbably found their way southwards in successive migra- tions along the Andean mountain-system, where suitable habita- tions, corresponding in general physiographical features to those of their northern home, could readily be found. A broad horizon- tal or latitudinal distribution, per contra, characterises the insect fauna of the north temperate zone. A large proportion of the European species, for example, are spread over the far interior of the Asiatic continent, and, indeed, many of them reappear in America. Nothing more strikingly illustrates this broad diffusion of species than the case of the Japanese lepidopterous fauna, which, out of some 1,110 species of Macro-Lepidoptera, contains, accord- ing to Pryer,103 not less than 123 species that are common to Great Britain, or about 16 per cent, of the entire British fauna. Some remarkable and not wholly comprehensible anomalies of distribution are exhibited by the faunas of almost every region, which render unusually intricate the general problems of distribu- tion presented by the class as a whole. One of the most interest- ing and instructive of these is the special relationship which unites the New Zealand and Chilian and Patagonian coleopterous faunas, and the distinctness of the fauna of the first-named region from the Australian,104 a condition which, as Professor Hutton has shown, also characterises some of the other animal groups, and which would seem to argue in favour of some former direct land connection (trans-Pacific) between New Zealand and a portion of the South American continent. 282 GEOGRAPHICAL AtfD GEOLOGICAL DISTRIBUTION. The remains of insects in the older Paleozoic rocks are very scanty, and in the main they occur in such an unsatisfactory state of preservation as to have led to the most divergent views respect- ing their true relationship. The most ancient of these forms, and the only one that is thus far known from the Silurian rocks, is the Palaeoblattina Douvillei, recently described by Brongniart from the Middle Silurian sandstones of Calvados, France, and referred by that naturalist to the orthopteroid group of the cockroaches. No other member of the orthopterous division of insects is positively known prior to the Carboniferous period, and it is not unlikely that the fossil in question, which is represented by a single impression of a wing, may on further investigation prove to belong to the group of netted- veins (Neuroptera or Pseudo- Neuroptera), which alone among the different orders has representatives in the Devonian rocks. These last comprise fragments representing some five or six species, belonging to possibly as many genera — Palephemera, Gerephemera, Lithentomum, Homothetus, Xenoneura, and Dis- crytus. The first two are referred by Hagen to the modern type of the Libellulae (Pseudo-Neuroptera), and the remaining three, not counting the very doubtful and fragmentary Discrytus, to the like- wise modern sialine tribe of the Neuroptera. By Mr. Scudder, on the other hand, several of these earlier hexapods are considered to represent synthetic types, and have accordingly been referred to families specially created for their reception — Palephemeridae, Ho- mothetidae, Xenoneuridse ; indeed, the same authority, following in the footsteps of Goldenberg and Brongniart, insists that these forms, as well as all others, with possibly one exception, from the Paleozoic deposits, in a given departure from modern type-struc- tures, and in the possession of combination ordinal characters, con- stitute a special group or order apart by itself, the Palaeodictyoptera, whose diverging specialisation (effected at about the beginning of the Mesozoic era) outlined the various higher groups or orders now recognised by entomologists.106 In conformity with this view the Paleozoic insect was a synthetic hexapod, in which ordinal differ- entiation had not yet asserted itself. The Carboniferous cock- roaches (Palrcoblattarige), which constitute probably one-half of the entire Paleozoic insect fauna, and of which some sixty or more species are known, and the contemporaneous walking - sticks — Protophasma, the giant Titanophasma from Commentry, France, DISTRIBUTION OF INSECTS. 283 &c. — are accordingly not true Orthoptera, but orthopteroid Palaeo- dictyoptera ; the Carboniferous and Devonian netted- veins not true Neuroptera or Pseudo-Neuroptera, but neuropteroid Palseodictyop- tera, and, similarly, the Hemiptera (Eugereon, Fulgorina), hemipte- roid Palseodictyoptera. Apart from the contradictory conclusions which have been reached from the study of these forms by Hagen, Gerstaecker, Eaton, and others, it may be reasonably doubted whether the extreme specialisation seen among the Palseodictyoptera will carry out the inference that the larger groups of Paleozoic times were more closely related to one another "than any one of them is to that modern group to which it is most allied, and of which it was with little doubt the precursor or ancestral type." 108 Surely, it will not be contended that Palephemera and the highly special- ised Titanophasma are more nearly related to each other than they are to the modern families Libellulidae and Phasmidae, not to men- tion the orders to which these belong ; and if this be so, why should they be referred rather to the one loose comprehensive group than to the several groups which they immediately represent ? The remains of beetles (Coleoptera), if we except the very doubtful Troxites (which is considered by some naturalists to represent the fruit of a plant), are unknown in the Paleozoic depos- its ; but it is by no means unlikely that the members of this order had already existed, since borings in wood, very like those made by the coleopteroid larva, have been discovered in various localities.* The Triassic insect fauna, which is represented by some four or five European species, and by about twenty in America (Colorado), is almost exclusively orthopteroid, and exhibits a distinct passage between the ancient and modern types of cockroaches — Pala3oblat- tariae and BlattariaB. The first indisputable beetle (Chrysomelites), followed by several highly differentiated types in the Rhaetic— Hydrophilites, Buprestites, Curculionites — is found in the deposits of this age. No truly metabolous insects, or those undergoing complete metamorphosis, other than Coleoptera, are known before the Lias, where, however, we have several species of Diptera (Chironomidae, Tipulidse), and at least one representative of the Hymenoptera, an ant (Palseomyrmex prodomus), from Schambelen, * Different coleopteroid species have at various times been described from Carboniferous strata, but these are now known to be mainly referable to the Arachnida or spiders. 284 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. Aargau. The same Swiss deposits have yielded upwards of one hundred and ten species of Coleoptera, in addition to various forms of Orthoptera, Neuroptera, and Hemiptera (Cicadina). A true song-cricket (Cicada Murchisoni) has been described by Brodie from the English Lias. The most diverse types of existing Neu- roptera and Orthoptera, which appear to have been fully differ- entiated at this period, acquire a further development in the suc- ceeding Oolites, where, in addition to representatives of all the other orders that had thus far appeared, we meet with the earliest Lepidoptera (Sphingidae). The singularly deficient fauna of the Cretaceous period is suc- ceeded in the Tertiary, more particularly in the Oligocene and Miocene divisions, by a most prolific development of specific forms. The Miocene deposits of Switzerland, and the immediately ad- joining tracts, have yielded to Dr. Heer no less than eight hun- dred and seventy-six species — eight hundred and forty-four of which are recorded from Oeningen (Baden) alone — distributed as follows: Coleoptera, five hundred and forty-three; Orthoptera, twenty ; Neuroptera, twenty-nine ; Hymenoptera, eighty-one ; Lepi- doptera, three; Diptera, sixty-four; and Rhynchota, one hundred and thirty-six. Scarcely less famous as insect localities than Oen- ingen are Aix in France and Radoboj in Croatia ; but apparently far surpassing either of these in respect to both individual and specific development are the tufa-beds of the Florissant region in Colorado, referred to the Oligocene period. The Hymenoptera from these deposits comprise several species of bees (Apidae and Andrenidae), about thirty species of wasp-like forms (Vespidae, Sphegidae, &c.), fifty species of ants (represented by about four thousand specimens, mainly Formicidae), and some eighty species of Jchneumonidae, be- sides numerous other forms. The Diptera individually make up nearly one-third of all the specimens found in the region (Culicidae, Tipulidae, Bibionidae), and of the heteropterous division of the Hemiptera there appear to be no less than one hundred species. The Coleoptera are likewise exceedingly abundant, and comprise among other forms some thirty species each of the families Carabidae, Staphylinidae, and Scarabeidae, and forty of the Elateridae. The total number of species represented in this order is about three hundred, of which about one hundred and twenty belong to the rhynchophorus division. DISTRIBUTION OF ARACHNIDA. 285 Owing to the very limited nature of insect faunas in general, and the circumstance that only a few localities have thus far yielded insect remains in any abundance, it is impossible to draw any positive conclusions respecting either the geographical dis- tribution or the genealogy of the members of this class of animals. It appears practically certain, however, that the metabolous types were descended from the ametabolous, and that the Iarva3 of the ear- lier forms were all, or mostly all, aquatic in habit. The geological and horizontal range of the majority of the species appears to have been very limited, but it must be recollected that in most instances the number of individuals found representing any one given species is altogether too small to permit of any logical inference being drawn from their occurrence. Thus, of the numerous species of Paleozoic cockroaches, by far the greater number are represented by single specimens, or by specimens coming from a single locality. ARACHNIDA AND MYRIAPODA. The earliest arachnoid remains (scorpions) occur in the Upper Silurian deposits of the island of Gothland, Sweden, and Lanark- shire, Scotland, and in the Helderberg rocks of Waterville, New- York, from each of which regions a single specimen has been ob- tained. The Swedish species, Palseophoneus nuncius, with which the Scotch form appears to have been nearly related, ranks among the largest of its class, measuring about three and a half inches in length ; in its general characters it closely approximates the modern scorpions, although, as has been pointed out by Professor Lindstrom, the structure of the large and pointed thoracic limbs more nearly approaches what is seen in the embryonic forms of other Tracheata and in Campodea. The presence of stigmata and the whole organisation of the body clearly demonstrate the ani- mal to have been an air-breather and an inhabitant of dry land. The American form, Proscorpius Osborni, is less clearly recognisable than the European, and some doubt has been thrown upon its arachnoid nature, which is maintained by both Whitfield and Tho- rell. Excepting these two or three earliest precursors of the Scor- pionidse, whose presence would naturally seem to indicate a rich insect fauna for the period, no traces of arachnoid remains are known antedating the Carboniferous deposits, where, however, several well- marked genera, singularly close in their relationship to modern 286 GEOGRAPHICAL AKD GEOLOGICAL DISTRIBUTION". forms, and representing spiders (Protolycosa, Architarbus, Anthra- comartus), scorpions (Cyclopthalmus, Eoscorpius), and pseudo-scor- pions (Microlabis), are met with. In the Mesozoic deposits but few arachnoid remains, in most cases in an imperfect state of preserva- tion, have thus far been discovered, and it is not until about the middle of the Tertiary series, Oligocene, that they acquire by their numbers any importance. The amber deposits of Europe and the Florissant beds of Colorado have yielded the greatest abundance of such remains. It is not a little remarkable with what degree of persistence the fundamental characters of scorpions have been preserved, for it appears, as has been claimed by Mr. Peach, that the Car- boniferous forms, as represented by Eoscorpius, were as highly or- ganised and specialised towards the beginning of this period as their descendants of the present day. The Paleozoic araneids, on the other hand, appear to have been all, or in the main, possessed of distinctly segmented abdomens, thereby forming a transitionary group between the arthrogastric and non-arthrogastric arachnoids. Both the chilognathous and chilopodous types of Myriapoda may be said to be represented in the Carboniferous rocks, although from certain peculiarities of structure possessed by these early forms — the genera Euphoberia, Xylobius, Acantherpestes, Trichiulus, among the former, and Palseocampa among the latter — which are un- known in their modern representatives, special ordinal groups, the Archipolypoda and Protosyngnatha respectively, have been created for them. It may be doubted, however, whether such knowledge as we possess of the animals in question will permit of the reten- tion of these groups ; indeed, it appears by no means certain that all the forms referred here as Myriapoda are actually such at all. The earliest myriapod remains, referred by Peach 107 to the Chilo- gnatha, occur in the Old Hed Sandstone (Devonian) of Forfarshire, Scotland, and not improbably the problematical Gyrichnites of the nearly equivalent deposit of Gaspe are the belongings of these animals. The Mesozoic rocks are singularly deficient in their traces, and may be said to be almost wholly wanting in them ; excepting the somewhat doubtful Geophilus proavus from the Jurassic deposits it appears that no chilopodous form is met with before the Ter- tiary. II. Distribution of the Vertebrata.— Fishes.— Amphibians.— Keptiles.— Birds.— Mammals. FISHES. THE geographical distribution of fishes is at the present day, and probably has been for a considerable number of past geologi- cal periods, world-wide. Although vastly more abundant, if not individually at least specifically, in the regions of elevated tem- perature than in those of the opposite extreme, both as regards the marine and fresh-water forms, they are still far from wanting in waters of icy coldness, whether these be in high latitudes, the oceanic abysses, or elevated mountain lakes or streams. The offi- cers of the British Polar Expedition, under command of Sir George Nares, obtained specimens of the charr (Salmo arcturus and S. Naresii) from beyond the eighty-second parallel of north latitude, the highest point at which fresh-water fishes have been observed, and from a still higher latitude, the eighty-third, some half-dozen species of shore fishes, among them a bull-head and cod (Cottus quadricornis, Gadus Fabricii, Icelus hamatus, Cyclopteris spinosus, Liparis Fabricii, Gymnelis viridis).108 And were it not for the in- superable obstacles that were interposed in the way of fishing, there can be no doubt that many additional forms would have been dis- covered. The number of forms that descend into, or inhabit, the abyssal waters whose temperature is about that of freezing is very considerable; Gunther109 enumerates thirty-nine species whose range extends to, or passes beyond, the fifteen hundred fathom line, thus penetrating deep into the zone of icy coldness. In the European Alpine region fishes (salmonoids) inhabit the lakes or streams situated at about the level of perpetual snow, and there is very little doubt that the same is the case in nearly all regions of 288 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. elevated mountains. The loach has been found in the Himalayas at an altitude of eleven thousand feet, and in the South American Andes, Lake Titicaca, at an altitude of thirteen thousand feet, has yielded several species. Indeed, in the latter region, some of the Alpine cyprinodonts, as Trichomycterus, penetrate to heights of fifteen thousand feet, and upwards. Fresh- Water Fishes.— It will be evident from the present rela- tion existing between land and water — which in its general features unquestionably dates from a very remote geological period — and the resulting barriers opposed to a free migration, that fresh-water fishes will be much more limited in their range than the fishes of an oceanic type, whose distribution is in main part governed by conditions of food-supply and temperature, and to a certain extent by the nature of oceanic currents. This comparative areal restric- tion among fresh-water forms is exemplified not only in the case of species and genera, but also in that of families, none of which, if we except the cat-fishes (Siluridae), can be considered to be in any way cosmopolitan. Manifestly, the oceanic basins must prove to the animals of this class an obstruction much in the manner that it does to the higher animals, the reptiles and mammals, for ex- ample. Yet, certain peculiar occurrences would seem to indicate that the natural barrier thus formed is not in all cases as effectual in preventing the distribution of fishes as it is with the majority of the animals just mentioned. Thus, several identical species, as the salmon (Salmo salar), perch (Perca fluviatilis), burbot (Lota vul- garis), pike (Esox lucius), and a stickleback (Gasterosteus pungitius), inhabit alike the waters of Europe and Eastern North America; the perch of the Ganges and other East Indian rivers (Lates cal- carifer) is found in the waters of Queensland, Australia; and one of the forms of south temperate so-called " trout" (Galaxias atten- uatus) inhabits Tasmania, New Zealand, the Falkland Islands, and the southern extremity of the continent of South America. In addition to such divided species, representing, naturally, divided genera as well, there are also several generic types whose lim- ited number of representatives (while distinct specifically) be- long to opposite quarters of the globe. The genus Umbra, limited to two species, is represented in the Atlantic States of the American Union by the " dog " or "mud fish " (U. limi), and in the Danubian system of waters by the * ' Hundsfisch " (U. Krameri) ; the shovel- DISTRIBUTION OF FRESH-WATER FISHES. 289 nose sturgeons (Scaphirhynchus), and the paddle-fishes (Polyodon- tidae), both of them restricted to some two or three species, are con- fined respectively to the river systems of Central Asia and the Mis- sissippi, and the Mississippi and the Yangtse-Kiang ; the American suckers (Catostomus) have an outlying representative in Siberia; while the East Indian genus Symbranchus, after skipping Africa, reappears with a single species (S. marmoratus) in the waters of South America. In what precise manner these equivalent types have found their way to such widely removed portions of the earth's surface it has been thus far impossible to determine. That some transference was effected by way of northern waters over land-surfaces now no longer existing is very nearly certain ; hence, the occurrence of identical or representative specific forms in the streams of Northern Eurasia and North America is not very surprising. But that a similar transference was effected over the broader or equatorial parts of the oceanic basins, a hypothesis necessitating the assump- tion of the submergence of vast continental land-masses where no traces of their former existence are visible, is more than doubtful. At any rate, it is very unlikely that any such alternation in the relative positions of land and water took place at a time so recent as satisfactorily to explain such anomalies of distribution as are presented by the genera Lates and Symbranchus, already men- tioned, and by the genus Pimelodus (Africa and South America) among the cat-fishes. It would appear at first sight far more natural to assume that these fishes were originally of a marine type, spread over the oceanic expanses, and that at a later period they accommodated themselves to fresh-water conditions, and grad- ually restricted their habitats to regions where we now find them. This is not unlikely, seeing that some of these (Lates, Symbranchus) freely enter brackish water. That such has been the recent origin of many forms of fresh-water fishes is placed beyond question. Several permanent species of the Northern Baltic, where, through an excessive indraught of fresh water from the surrounding streams, the water has lost most of its salinity, are identical with marine types inhabiting the Arctic seas to the north. Yet the accommo- dation to new conditions of existence was effected since the closing off of the Baltic from the northern ocean, or since the Glacial period. The gobies, blennies, and atherines of the northern lakes 290 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. of Italy present us with other instances of the permanent establish- ment of marine types in fresh waters. Indeed, it is scarcely neces- sary to seek for examples of this kind, seeing how very readily the many marine fishes which periodically ascend the inland streams during the spawning season accommodate themselves to the newly imposed conditions. Of so little importance does a change of medium appear to be in many cases that it is frequently very diffi- cult, or impossible, to indicate whether a given group of fishes is more properly of a marine or fresh- water type. The numerous in- stances wrhere certain species of a genus are of one habit, and other species of the same genus of the opposite habit, render the deter- mination of this question still more difficult. An accommodation similar to that which has been noticed in the case of marine fishes also obtains with many of the more strictly fresh-water forms; i. e., they descend without inconvenience into the briny oceanic medium. This we see in the case of the trout, the charr, in several species of Coregonus, and especially among the toothed carps (cyprinodonts) and sticklebacks. How far these may wander out to sea is not exactly known, but there is no special reason for supposing that they might not proceed, at least in some instances, to very considerable distances. A species of the cyprino- dont genus Fundulus (F. nigrofasciatus) was obtained by the offi- cers of the ''Challenger" from the pelagic fauna of the Atlantic, midway between St. Thomas and Teneriffe. It is manifest, there- fore, that an arm of the sea is not an impassable barrier to certain forms of fresh-water fishes ; and not impossibly some brackish-water forms may occasionally find their way completely across the oceanic expanse. The irregular distribution of certain types or species thus receives a partial, or, at any rate, a possible solution. The total number of species of strictly fresh-water fishes recog- nised by Gunthcr is nearly two thousand three hundred, of which four are lung-fishes, thirty-two ganoids, twelve lampreys, and the remainder teleosts or bony-fishes. Of the last nearly one-third are comprised in the family of the carps (Cyprinidae) and somewhat more than one-fourth in that of the cat-fishes (Siluridae). The Characinida3 and Chromides, forms from tropical America and Africa, are represented by somewhat more than two hundred and fifty and one hundred species respectively, the salmonoids by one hundred and thirty-five species, and the toothed carps (cyprino- DISTRIBUTION OF FRESH-WATER FISHES. 291 donts) by one hundred and ten species. As to their geographical distribution, two primary zones might be recognised : the northern, corresponding largely to North America and Temperate Eurasia, characterised by the presence of sturgeons, salmonoids, pikes, and numerous carps, with only a feeble development of the cat-fishes; and the southern or tropical zone, comprising the Indian, Ethio- pian, Neotropical, and Australian regions of zoogeographers, char- acterised by a special development of the cat-fishes. The fol- lowing scheme for the classification of the southern zone has been proposed by Dr. Gunther : CYPBINOID DIVISION. — Characterised by presence of CyprinidaB and Labyrinthici. 1. Indian Region, 625 Species. — Characterised by Ophiocephalida3 and Mastacembelidse. Cobitoids numerous. 2. African Region, 255 Species. — Characterised by lung-fishes (Protopterus) and ganoids (Polypterus, Calamoichthys). Chromoids and characinoids numerous. Cobitoids absent. ACYPRINOID DIVISION. — Characterised by absence of Cyprinidae and Labyrinthici. 1. Tropical American Region, 672 Species. — Characterised by lung-fishes (Lepidosiren). Chromoids and characinoids numerous. Gymnotidse (electric eels). 2. Tropical Pacific Region, 36 Species. — Characterised by lung- fishes (Ceratodus Forsteri and C. miolepis, from the waters of Queensland, Australia). Chromoids and characinoids absent. The eastern and western divisions of the northern zone, with some three hundred and sixty and three hundred and forty species respectively, are very closely related to each other, not only through the preponderance of types that are common to both regions, but in the possession, as has already been seen, of a number of identical species. Two genera of ganoids not found in Eurasia, Lepidosteus and Amia, serve to characterise the American ichthyic fauna, which is further distinguished from its trans- Atlantic correspondent in the special development of the suckers (CatostomidaB) and in the absence of cobitoids and barbels (Barbus). An Antarctic zone, made to include New Zealand, Tasmania (with a portion of Southeast Australia), the Falkland Islands, Tierra del Fuego, Patagonia, and Chili, whose faunas are very intimately related to one another, is recognised by some authors, but the mini- 292 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. ber of species here represented is too limited to permit of much importance being attached to a negative region of this kind. Galaxias attenuatus, a species of southern ' ' trout, " and one of the three species of lamprey are found in New Zealand and the Tas- manian and Fuegian tracts. No fresh-water fishes are thus far known to inhabit any of the islands situated south of the fifty-fifth parallel of south latitude. The paucity of the fish-fauna of the tropical Pacific region is very remarkable, and only receives a partial explanation through the circumstance that the greater part of the tract belonging to it, the Australian, is deficient in water-courses. The island of Celebes, for example, which, as has been well urged by Gunther, would seem to offer favourable conditions for the development of a fresh- water fauna, has thus far offered barely more than a half-dozen species, all of them common Indian forms; nor has New Guinea shown itself to be much more prolific. Long-continued isolation has apparently prevented much of the tract from receiving the necessary supply from the fresh waters of continental areas, al- though the identity of existing forms with such as are found else- where would seem to indicate a recent migration and peopling of the waters. The smaller islands of the Pacific are inhabited prin- cipally by such forms, as eels, atherines, gobies, mullets, which can readily exchange fresh water for salt water, and to which, con- sequently, the oceanic basin constitutes no insuperable barrier to a free migration. A species of Arius, a siluroid, inhabits the Sand- wich Islands. The marine fishes may be conveniently divided into three cate- gories : shore fishes, or such as habitually frequent the coast-lines, and which rarely descend to a greater depth of water than three hundred fathoms; pelagic fishes, which inhabit the waters of the open sea, the majority of them spawning there also ; and deep-sea fishes, or the fishes of the greater oceanic depths, where the in- fluence of light and surface temperature is but little felt. No sharp line of separation between these several classes is permissible, however, since the habits normally belonging to the members of one class are occasionally assumed by members of the other classes as well, as must necessarily follow from the different conditions governing their distribution. DISTRIBUTION OF SHORE FISHES. 293 Shore Fishes. — The fishes of the first category number upwards of three thousand five hundred species. Their range in the north extends to, or beyond, the eighty-third parallel of latitude, but in the Southern Hemisphere no species have been found to pass be- yond the sixtieth parallel, although, doubtless, they exist along some of the more southerly shore-lines. The Arctic fauna, or the fauna occurring north of the sixtieth parallel of latitude, is, as far as we are warranted in believing, a strictly homogeneous one, iden- tical types largely characterising both the Old and the New World divisions. The more extensively represented families are, among the spiny-rayed fishes, the bull-heads (Cottida3 — Cottus, Icelus, Tri- glops), the Agonida3, lump-suckers (Discoboli), and blennies (Blen- niidae— Anarrichas, wolf -fish) ; and among the anacanths the cod- fishes, with the cod (Gadus), hake (Merlucius), arid ling (Molva). Among the physostomous fishes, or those in which the air-bladder is provided with a pneumatic duct, the herring (Clupea) is repre- sented by a limited number of species. The cartilaginous fishes are very scarce ; indeed, thus far only one species, the Greenland shark (LaBmargus), is known to penetrate north of the Arctic cir- cle. The chimera, spiny dog-fish (Acanthias), and ray, are met with along the southern borders of this tract. The Antarctic shore fauna is in many respects closely related to the Arctic, although nearly one-third of the generic types are peculiar. As in the north, the cartilaginous fishes are scarce, being represented by a single species of shark (Acanthias), and one or more species of ray (Raja, Psammobatis). The Scorpa3nida3 and Agonidai among bony-fishes have each one genus, Sebastes and Agonus respectively, which is held in common with the Arctic fauna. The lophobranchs have in addition to the northern pipe- fish (Syngnathus) the remarkable Protocampus, represented by a single species of the Falkland Islands (P. hymenolomus). A most interesting fact connected with the Antarctic fauna is the recur- rence of types belonging to the far north, which are wanting in the intermediate region. This we see in the single species of spiny dog-fish (Acanthias vulgaris), which is a member of the Arctic and north temperate faunas, but is absent from the equa- torial- region. The hakes comprise two species, one of which is restricted to the northern waters and the other to the southern; and a similar separation is found among the species of the Arctic 294 GEOGKAPHICAL AND GEOLOGICAL DISTRIBUTION. and Antarctic genus Lycodes. The gadoids are in both regions accompanied by the parasitic hag (Myxine). The shore fishes of the north temperate sea are largely identi- cal in their general types in both the Atlantic and Pacific basins, as well as on opposite sides of these basins. Thus, of the fishes frequenting the British seas we find, among other genera, the fol- lowing types also represented on the east coast of America : the bass (Labrax), sea-perch (Serranus), porgy (Pagrus), bull-head (Cot- tus), angler (Lophius), wolf -fish (Anarrichas), Zoarces, cod (Gadus), hake (Merlucius), ling (Molva), and rockling (Motella) ; and among physostomes the smelt (Osmerus), herring (Clupea), and conger. The surmullets (Mullus), gurnards (Trigla), John Dorys (Zeus), and some of the breams are wanting, or are but rarely met with.* The cartilaginous fishes show as common to the East and West Atlantic C5 the "hounds" (Mustelus), rays, sting-rays (Trygon), and the elec- tric rays (Torpedo) ; the true dog-fishes (Scyllium) appear to be wanting on the East American coast, and Chimera has thus far been found only in deep water. Most of the genera of British fishes also frequent the Mediter- ranean Sea, whose fauna effects a passage to the fauna of the equatorial zone. The number of peculiar genera is very limited. Among the newly appearing forms are the beryces (Beryx), star- gazers (Uranoscopus), umbrines (Umbrina), barracudas (Sphyrana), horse-mackerels (Caranx), and sea-horses (Hippocampus), members of the (more southerly) American fauna as well (elements borrowed from the fauna of the West Indies). The flat fishes— turbots, plaices, flounders, soles (Rhombus, Pleuronectes, Solea, &c.) — ex- hibit an increased development, while the gadoids rapidly diminish in numbers. A most remarkable correspondence exists between the Mediterranean fishes and those of the Japanese province — i. e. , the coast of Asia between the thirtieth and thirty-seventh parallels of latitude ; indeed, viewed from a generic standpoint, this corre- spondence may be almost said to amount to identity. More than one half of all the generic types represented are fishes of the south of Europe, and in many cases even the species are identical with * The occurrence of Trivia cuculus on the American coast is considered very doubtful by Jordan and Gilbert; on the other hand, Mullus barbatus, supposed to be absent, has during late years been reported from Pengacola, Florida, and Wood's Holl, Massachusetts. (Smith's " Misc. Coll.," 1883.) DISTKIBUTIOtf OF SHORE FISHES. 295 European forms. no Several of the berycoid genera inhabit the Japa- nese and Mediterranean waters exclusively, while others, as the red mullets, John Dorys, and trumpet-fishes (Centriscus), occurring in these two districts and elsewhere, are wanting on both the East and West American coasts.* How the transference of the Mediter- ranean fauna to the East Asiatic coast, or the converse, was effected, whether by means of a comparatively recent open water-way be- tween the two regions, as has been supposed by some, our present knowledge does not permit us to say. The shore fishes of the south temperate zone, which extends northward from the Antarctic faunal belt to about the thirtieth parallel of south latitude, are most intimately related to those of the north temperate, although very distinct from the forms which occupy the intermediate or equatorial zone. Nor is this corre- spondence restricted to generic types alone, since we find a consid- erable number of northern species which, skipping the intermediate tract, reappear here without having undergone even varietal modi- fication. Such are the chimaera (Chimera monstrosa), two species of dog-fish (Acanthias vulgaris and A. Blainvillii), the monk-fish (Rhina squatina), John Dory (Zeus faber), angler (Lophius piscato- rius), bellows-fish (Centriscus scolopax), anchovy (Engraulis en- crasicholus), sprat (Clupea sprattus), and conger (Conger vulgaris). Among the marked differences separating the two regions may be mentioned the substitution of the cottoids by the Nototheniae (fam- ily Trachinidse) and the Discoboli by the Gobiesocidae.f Of the four south temperate provinces, that of the Cape of Good Hope, the South Australian (with New Zealand), the Chilian, and the Patagonian, the Australian is by far the richest, numbering in its fauna not less than one hundred and twenty genera and two hundred species. Two-thirds of all the genera occurring on the coasts of Southeastern Australia and Tasmania are also represented in the shore fauna of New Zealand, which comprises some one hundred species. The New Zealand genera not represented in the Australian coast are about twenty-six in number, with about an equal number of species. The most marked difference between the two nearly contiguous faunas is the absence of gadoids in the Australian element, while the group is represented by not less than * Centriscus scolopax is said to be accidental on the American coast, t The Gobiesocidae have also northern representatives. 296 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. six genera (Gadus, Merlucius, Lotella, Motella, &c.) in the fauna of New Zealand. — The Chilian fauna is generically closely related to the South Australian, and contains but few types that are strictly peculiar to it. One of its more remarkable features is the posses- sion of a species of Polyprion (P. Kneri, from Juan Fernandez), one of the Percidae, the only other known species of the genus (P. cernium) being a member of the European fauna.* The number of fishes thus far obtained from the Patagonian province is very limited, and scarcely permits us to formulate any definite opinion as to the characters of the entire fauna. The equatorial shore fishes far exceed those of the temperate zones both in number and diversity of form, agreeing in this re- spect with the condition presented by equatorial land and fluviatile faunas generally. It is here, too, that we find grotesqueness of outline combined with a most varied and beautiful colouring, an almost infinite arrangement of lively tints, scarlet, yellow, blue, black, &c., imparting to the fishes of this tract an indescribable brilliancy. Throughout the greater part of the equatorial or tropi- cal belt there is manifest a strong faunal identity, rendering the institution of ichthyic provinces practically impossible. Thus, the greater number of the dominant types of the Atlantic Ocean are also represented in the Pacific, and vice versa, and in many instances even the species representing these types are identical. The fishes of the tropical waters of the Indian and Pacific oceans are very intimately related, the number of identical species ranging from the Red Sea far into the Polynesian Archipelago being very great. A limit to the eastward extension of the Indo-Pacific fauna appears to have been set, however, by the cold current sweeping northward along the western coast of South America, aqd as a consequence we find a more or less distinct, or individualised, fauna along the east- ern border of the great ocean. Among the more largely represented generic or family groups of the equatorial zone are the sea-perches (Serranus), snappers (Mesoprion), mullet-kings (Apogon), blow- fishes (Chaetodon, &c.), Scorpaenidae, horse-mackerels (Caranx), coral-fishes (Pomacentridse), Julidinae, flat-fishes (Pleuronectidae), herrings, and Mursenidae, several of which are more or less distinc- tive of the zone. In all of these groups the number of species in * Obtained also in deep water by the United States Fish Commission. (Jordan and Gilbert, " Synopsis Fishes of North America," 18S2.) PELAGIC FISHES. 297 the Indo-Pacific basin is very much greater than in the Atlantic. The extensive development of reef-structures through the former area, presenting unusually favourable conditions for existence, has, doubtless, much to do with this comparative superabundance. Pelagic Fishes. — Our knowledge respecting the fishes which spend a considerable, or the greater, part of their existence on the free surface of the oceans, borne resistlessly in the course of the oceanic current, or inhabiting masses of floating sea-weed, although still very meagre, is sufficient to indicate that the number of such types is very limited. As with the other groups of fishes, they are most numerous in the regions of high temperature, diminishing rapidly as we proceed either north or south from the Equator. Most of the tropical genera are also met with in the temperate zones, and probably the converse is also true, although a number oi exceptions have been indicated. The warm-water fishes become rare beyond the fortieth parallel, and decline very rapidly with the decline of the temperate fauna itself. Almost the only pelagic fish of the Arctic Ocean is the Greenland shark (Laemargus borealis). Among the better known bony-fishes that enter into the com- position of the pelagic fauna are the flying-gurnards (Dactylopterus), mackerels, tunny, bonitos, dolphins (Coryphsena), skip-jacks, pilot- fishes, sword-fishes, frog-fishes, scopelids, skippers, flying-herrings (Exoccetus), sea-horses, porcupine-fishes, and sun-fishes (Orthago- riscus). The cartilaginous fishes are preeminently pelagic in their habits, and contribute largely to this fauna. The sharks are repre- sented by a number of genera (Carcharias, Galeocerdo, Zygsena, Lamna, Notidanus, &c.), and by forms which are not only the largest of their tribe, but approximately the largest of known fishes. The basking-shark (Selache), the largest shark of the North At- lantic, attains a length of more than thirty feet ; Carcharodon Rondeletii, a tropical or sub-tropical species, the most formidable of all sharks, has been known to measure forty feet, and Rhinodon typicus, a species of the Indian Ocean, fifty to sixty feet. Rivals to these monsters of the deep are the sea-devils or eagle-rays (My- liobatidae), many of whose forms (Dicerobatis, Ceratoptera) appear to attain a length of twenty feet or more. The similarity existing between the pelagic faunas of the At- lantic and Indo-Pacific basins is very great, extending not only to genera, but to species. A number of the forms are nocturnal in 14 298 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. their habits, living in a deeper zone during the day, and appearing at the surface only at night. The majority of these are provided with phosphorescent organs, a structure largely prevalent among deep-sea fishes, to which these fishes effect a passage. Deep-Sea Fishes. — Much uncertainty still remains as to the character of the abyssal ichthyic fauna, owing to the difficulty of determining in many or most cases the depths whence specimens caught in the net were obtained. But there can be no doubt that many forms inhabit very nearly the greatest depths that have been reached by the dredge, Gunther, in his "Introduction to the Study o£ Fishes," enumerates upwards of fifty forms which are supposed to have been obtained from depths exceeding 1,000 fathoms; twenty-six from depths exceeding 2,000 fathoms; and nine from depths of 2,500 fathoms or over. Additional forms have since been obtained in the dredgings of the "Talisman" and "Al- batross." Bathyophis ferox and Halosaurus rostratus are reported to have been dredged by the "Challenger" in water of 2,750 fathoms (5,019 metres), and Gonostoma microdon at an extreme limit of 2,900 fathoms. Considerable doubt, however, attaches to the last, since the fish is abundant in water of only moderate depth, and may have been taken down in the descent of the dredge, or captured only in its ascent. A similar doubt attaches to many of the other forms with a reputed very broad bathymetrical range, as it is hardly to be supposed that animals, organised specially to meet the conditions (pressure, &c.) of life in the oceanic abyss, should at the same time be so constituted as to endure with impunity the very different conditions governing life near the surface. The "Talisman" obtained Alepocephalus rostratus at depths stated to range between 868 and 3,650 metres, Scopelus Maderensis between 1,090 and 3,655 metres, and Macrurus affinis between 590 and 2,220 metres ; similarl}', it is claimed that the ' ' Albatross " obtained Cyclothone (Gonostoma) lusca and Scopelus Mulleri in depths vary- ing from 560 to 5,394 metres (2,949 fathoms— latitude 37° 12' north; longitude 69° 39' west), or over a vertical extent of 16,000 feet. Mr. Tarlton H. Bean justly considers the deep catch as doubt- ful. Less doubt attaches to the position of Aleposomus (?) Copei and Mancalias uranoscopus, both of which were obtained in the 5,394 metre haul.111 The deepest recorded find of the "Talis- man " was in 4,255 metres (Bythites crassus)."8 DEEP-SEA FISHES. 299 The Gadidae, Ophidiidas, Macruridaa, and Scopelidos make up a very large proportion of the deep-sea fauna, which has thus far yielded some ninety to one hundred or more genera. The eels (Muraenidse) are largely represented, and have been dredged from depths extending to 2,500 fathoms. Very few cartilaginous fishes were obtained by the ''Challenger," and these (Scyllium, Centro- phorus, Raja) were restricted to water not much exceeding six hun- dred fathoms, although the same group of fishes yielded specimens to the "Talisman" at nearly twice this depth (coast of Portugal). Whether or not different zones of ichthyic life can be recognised in the oceanic abyss may perhaps still be considered an open ques- tion. Dr. Gunther maintains, as the result of his studies, that "as far as the observations go at present, no distinct bathymetrical regions which would be characterised by peculiar forms can be defined," and that, "if the vertical range of deep-sea fishes is actually as it appears from the ' Challenger ' lists, then there is no more distinct vertical than horizontal distribution of deep-sea fishes." As a result of the explorations of the "Talisman," Filhol arrives at the opposite conclusion, and believes that a series of more or less distinct zones can be indicated. Although in its gen- eral features the abyssal fauna ("Bassalian" of Gill) is closely related to that of the superficial tracts, yet a number of distinctive elements, arising in part from certain remarkable abnormalities of structure, or from the presence of types not elsewhere represented, serve in a measure to define it. Professor Gill indicates twenty- eight families of deep-loving fishes.113 Of the two lowest orders of fishes, the Pharyngobranchii, as rep- resented by the Amphioxus or lancelet, and the Marsipobranchii, lampreys and hags, no unequivocal fossil remains have as yet been discovered. Possibly, however, some of the singular tooth-like bodies described as conodonts from the Cambrian and Silurian de- posits, which by many authors are referred to the dental armature of annelids, may belong to fishes more or less nearly related to the modern hag. The Elasmobranchii (sharks, rays) have their oldest representa- tives in the Upper Ludlow horizon of the Upper Silurian formation, being preceded in time by a bucklered ganoid, Pteraspis (Scaphas- pis) Ludensis, from the Lower Ludlow. These earliest remains, belonging to the family of the true sharks (Squalidse), are in the 300 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. form of teeth (Thelodus) and fin-spines or ichthyodorulites (On- chus, Ctenacanthus). The order is but very scantily represented in the Devonian deposits, but in those of the succeeding Carboniferous age (sub Carboniferous limestone) it acquires a profuse develop- ment, the pavement-teeth of forms generally considered to be allied to the modern " Port- Jackson " shark (Cestracion) being particu- larly abundant — Helodus, Orodus, Chomatodus, Petalodus, Cochlio- dus, Psammodus. The relationships of these genera are, however, still very obscure, and not impossibly they constitute a group apart by themselves — Psaminodontes. Less doubtful representatives (Strophodus, Acrodus, Hybodus) of the type of cestraciont fishes appear in the early and middle Mesozoic periods, Triassic and Jurassic, and in the chalk the teeth of Ptychodus are still very abundant. But in the latter period the more modern type of sabre-toothed sharks and dog-fishes, largely represented by existing genera, Notidanus, Scyllium, Lamna, Carcharias, Hemipristis, Gale- ocerdo (Corax, Otodus — extinct), appear to have gained the ascend- ency, which they retained throughout the subsequent Tertiary pe- riods. The most important of the later genera, and especially distinctive of the Miocene deposits, is Carcharodon. The earliest unequivocal traces of rays occur in the deposits of Jurassic age, although not impossibly some of the hypothetically placed forms of the later Paleozoic periods may belong to this group, or, at any rate, effect a union between it and the sharks. Of this nature appear to be the Jurassic Thaumas and Squaloraja, both of them nearly allied to the modern (and Cretaceous) Squatina. The true- and sting-rays are abundantly represented, particularly by frag- ments of their dental armatures, throughout the Tertiary deposits, the modern genera (Trygon, Myliobatis, ^Etobatis, Zygobatis, Raja) predominating. The saw-fish (Pristis) and torpedo-ray date from the Eocene, and not impossibly from the Cretaceous period. Appearing almost simultaneously with the selachians, but at- taining a much earlier considerable development, are the ganoids, a class of fishes which in the earlier geological periods exhibit a remarkable diversity of structure, but at the present day are com- prised within a very limited number of genera, whose members, with the exception of the partially marine sturgeon (Accipenser), are all inhabitants of fresh water. Three or more distinct types, based upon characters drawn from the structure of the dermal PALEOZOIC FISHES. 301 armor, may be conveniently recognised : the osseous plated ganoids, as represented by the sturgeon ; the enamel rhomb-plated ganoids, typified in the American alligator-gars (Lepidosteus) and the Afri- can Polypterus ; and the cycloid scaled ganoids, exemplified in the American Amia, which may in a measure be said to connect these fishes with the herring among the teleosts. None of these types appear to be represented in the Silurian period, unless, in- deed, the group of the bucklered ganoids, to which the ancient Scaphaspis Ludensis, and the greater number of the Devonian forms (Cephalaspis, Pteraspis, Pterichthys, Coccosteus) belong, be considered to be nearly related to the sturgeons. This relation- ship, however, requires further demonstration before it can be accepted as a fact; indeed, it has recently been attempted to show that some of these most ancient ichthj7oid forms — e. g.y Pterichthys — are not fishes at all, but members of what may, perhaps, be con- sidered to be a degenerated group of the lower Vertebrata, the tunicates. The almost total obliteration of the bucklered type of ganoid with the close of the Devonian period is very remarkable, and has not yet received a satisfactory explanation. It has been conjectured by some that these fishes early withdrew to fresh water, and that, in the absence of fresh-water deposits of any magnitude in the period succeeding, they have necessarily left behind but scanty traces of their existence. It must be confessed, however, that this explanation is more in the nature of an assumption than anything else, and has but little positive to bear it out. In how far the Devonian fishes were of a fresh-water habit still remains to be determined, but it seems more than probable that they were largely of this, or at least of a brackish-water, character. The re- mains of the sturgeon are not known prior to the Eocene period, although a direct forerunner (Chondrosteus), uniting this family with the Spatularidaj, occurs as low down as the Lias. The non-bucklered ganoids of the Devonian period (Holop- tychius, Glyptolepis, Dipterus, Osteolepis, Diplopterus) belong principally to the type of the fringe -finned or crossopterygian Polypteri, which effect a passage to the lung-fishes (Dipnoi). This series, which is represented by both scaled and plated forms, is continued into the Carboniferous period (Rhizodus, Dendrodus, Megalichthys, Ccelacanthus) ; but here, as in the succeeding Permian period, they are already largely replaced by the lepidosteoid type 302 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. (Palseoniscus, Amblypterus), which has its forerunner in the Devo- nian Chirolepis. The primitive heterocereal tail which characterises the ganoid fishes of the Paleozoic era, and which is still borne by many of the rhomb-plated fishes of the early Mesozoic period — Palseoniscus, Ischypterus, Catopterus — is converted by gradual modification into the higher homocereal type, which distinguishes the more important genera of Jurassic fishes (Tetragonolepis, Dape- dius, Lepidotus, and the teleostoid Leptolepis). In Semionotus, which ranges from the Trias into the Lias, the tail is of a well- marked transitional character. Very few of the Jurassic Ganoidei survive into the Cretaceous period, which practically marks the final collapse of this important order of animals, henceforward suc- ceeded by the more highly constituted bony-fishes. The most im- portant remaining group is that of the pycnodonts, or bean-toothed ganoids, whose numerous closely related forms have been referred to several distinct genera, which range collectively from the Car- boniferous period (Platysomus) to the Eocene. Among the Jurassic genera are Microdon, Mesodon, Gyrodus, and Pycnodus, the last, a remarkable example of a persistent type, surviving the close of the Cretaceous period into the Tertiary. Of the group of the cycloid scaled ganoids (Ganoidei cycliferi), represented at the present day by Amia, probably the earliest unequivocal remains are those of the genus Amia itself, which appear in the Cretaceous deposits ; by many naturalists, however, several Paleozoic forms, as Holopty- chius, Asterolepis, Bothriolepis, and Ccelacanthus, are referred to this group, and with them also the ccelacanthine Cretaceous genus Macropoma. The lung-fishes (Dipnoi), which at the present day are repre- sented by the three very isolated genera Lepidosiren (Brazil), Protopterus (Africa), and Ceratodus (Australia), have left un- doubted traces of their existence as far back as the Permian period, when the genus Ceratodus itself appears (Bohemia, Texas), presenting us with the most remarkable instance of persistence in the whole range of vertebrate animals. Not unlikely the genus may be found to be of still older date, and to have been nearly contemporaneous with its formidable predecessor, the Dinichthys. Remains of Ceratodus have been found throughout the entire series of Mesozoic deposits, from the Trias to the Cretaceous inclusive. Lepidosiren and Protopterus are not known in a fossil condition. FOSSIL FISHES. 303 The osseous fishes proper appear for the first time in Cretaceous strata, being immediately preceded by the teleostoid group of ga- noid fishes of the family Leptolepidse, which effects a passage to them. Indeed, by many authors the genus Leptolepis and its Jurassic allies (Caturus, Thrissops, &c.) are classed with the for- mer, and placed near the herring, with which they appear to have been most nearly related. Although both the physoclist and phy- sostome types, or those in which the swimming-bladder is closed off from, or remains connected with, the gullet, appear very nearly simultaneously in the same deposits, and consequently by their occurrence give no evidence as to their respective antiquity, there can be no question that the physostome is the more ancient type, the severance of the bladder in the physoclists being the result of the disuse of parts. This is further proved by the existence of a connecting air-bladder among the recent (and, doubtless, also among the ancient) ganoids. The Cretaceous teleosts belong largely to existing types — Clupea (herring), Osmerus (smelt), Esox (pike), Beryx ; but it is not until the Eocene period that we find a repre- sentative modern ichthyic fauna. To this, and the succeeding Tertiary period, most of the more prominent existing types date their first appearance. In addition to a very large representation of both the arthropterous and anarthropterous forms, the Eocene deposits contain remains of the Lophobranchii (Syngnathus, pipe- fish), Plectognathi (Diodon, porcupine-fish; Ostracion, trunk-fish), and Apodes (Anguilla, eel). The study of the distribution of fossil fishes renders evident two important facts : First, that there has been a progressive modi- fication and evolution from less to more highly organised types; and, secondly, that among the almost innumerable forms of com- paratively recent origin, our existing ichthyic fauna still holds the wreck of a past fauna, whose period of decline belongs already to the earlier part of the earth's history. It, moreover, reveals an extraordinary persistence on the part of some of the individual types. The occurrence of the recent genus Ceratodus in deposits as ancient as the Permian is certainly very remarkable, but it does not argue, as some would lead us to believe, that this particular fish has undergone no modification since the period of its first introduction. The fact that its remains have been found fossil in the deposits of Europe and America, whereas at the present time it 304 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. is found exclusively in Australia, would seem to imply a different distribution of land and water masses than now exists. But there appears to be no doubt, from their association with a marine fauna, that the early members of the genus were of an oceanic character, and that the fresh- water habit was obtained as the result of later modifications. The peculiar distribution would then be readily accounted for. It appears more than probable that from the order of fishes represented by this genus have been descended at least some of the earlier amphibians ; if this be true, it would seem that the lungs of the dipnoans had been developed when the animals were still more or less strictly marine. The genetic relationship of the two remaining genera, the South American Lepidosiren and the African Protopterus, to Ceratodus still remains to be deter- mined ; but the development of an additional lung in these forms, coupled with the circumstance of their broad geographical isola- tion, would indicate an ancient differentiation of the two groups of the Dipneumones and Monopneumones. In tracing the phylogeny of the class of fishes as a whole we are presented with certain difficulties which in the present state of our knowledge prove an insuperable obstacle to the solution of the problem. Paleontology thus far offers no positive clue as to what might have been the direct ancestors of the animals in question, and until it does so inferences drawn from purely zoological char- acters will be largely in the nature of pure hypotheses. The earli- est fishes that appear, although obviously of a much less perfect structural type, exhibit very nearly the amount of specialisation seen in the modern forms, and are evidently far removed from the period of their first origination. Whether, therefore, the tunicates, which unquestionably possess many points o£ structural relation- ship with the fishes, are their true ancestors or not, or whether, instead of representing primitive vertebrate types, they are merely the degenerated remains of a more highly constituted ichthyic stock, must still be considered an open question. Nor would it be safe to affirm that the most ancient representative of the class was a form either closely or remotely allied to amphioxus, the lancelet, or that the latter is an ancient type at all. If the views recently set forth by Professor Cope m as to the tunicate affinities of Pterichthys and its allies be proved to be correct, then, indeed, the presumptive evidence would be very great for concluding that what have hither- AMPHIBIA. 305 to been generally considered to constitute an abnormal type of the Mollusca are in reality the true progenitors of the fishes. The facts in the case, however, require further substantiation. The phylogenetic relationship existing between the selachians and ganoids is equally obscure as that which exists between fishes generally and the other classes of animals. Both types are known to us in their oldest forms from very nearly the same horizon, and consequently give no indication as to priority of birth. Much more positive indication in this direction is afforded with respect to the lung-fishes and teleosts, the former of which appear to be clearly related to, and to have been derived from, the dipteroid ganoids (of the type of Dipterus), and the latter to have held a similar relationship to the rhomb-scaled ganoids, possibly of the type of Leptolepis. From the former, apparently, have descended the amphibians, while the latter have continued to develop as the dominant fish-fauna of existing waters. That the fresh-water forms are modified descendants of types originally inhabiting the seas there can be no reasonable doubt. It is impossible to state when the earliest differentiation of marine and fresh-water forms was effected, but there is every reason for supposing that it dates back far into the Paleozoic era, and that some, if not many, of the Devonian fishes were of a strictly fresh- water habit. AMPHIBIA. The most salient facts that present themselves in connection with the geographical distribution of the Amphibia are, first, their almost complete absence from oceanic islands — the Seychelles, New Caledonia, and the Feejee and Solomon Islands forming island groups exceptional to a general rule — and, secondly, the very nearly universal limitation of the tailed forms, sirens, newts, salamanders, &c., to the Northern Hemisphere. The nature of the first condition has already been discussed in treating of the dispersal of animals generally. The total number of known forms are comprised, ac- cording to the latest researches of Boulenger, in somewhat more than one hundred and forty genera and nine hundred species, of which only twenty- seven genera and seventy species belong to the urodelous or tailed division, eleven genera and thirty-one species are coecilians (Apoda), and the remainder, one hundred and five genera and eight hundred species, frogs and toads (Anura). 306 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. The ccecilians are tropical forms belonging to the East Indies, Africa (with the Seychelles — Hypogeophis rostratus), and America. The American species, including all of the genus Ccecilia itself, are about twenty in number, and range from Mexico to Peru and Brazil.* Remarkable instances of divided genera are presented by Dermophis, which possesses five American species and one from West Africa (D. Thomensis), and Urseotyphlus, represented by two species in Malabar and likewise one in West Africa. The urodele amphibians are comprised in four families : The Sirenidae or sirens, with two or three species, inhabiting the South- eastern United States ; the Proteid®, with two genera, Proteus and Menobranchus (or Necturus), the former confined to the subter- ranean waters of Carinthia, Carniola, and Dalmatia, and the latter to the streams of Eastern and Central United States and Canada; the Amphiumidae, with three genera, two of which, Amphiuma and Menopoma, represent North American forms, while the third, Sie- boldia (Cryptobranchus or Megalobatrachus), which is closely re- lated to the menopomas, is confined to Japan and China; and the Salamandridse (newts, salamanders, &c.), comprising upwards of ninety species, very extensively distributed throughout temperate Eurasia and North America, with some fifteen or more species in tropical America (from Mexico southward — Amblystoma, Spelerpes), a limited number in North Africa, and two (Tylotriton) in the Himalayas. The North American forms belong principally to the genera Plethodon, Desmognathus, Diemyctylus, Amblystoma (with Axolotl), and Spelerpes, the first two of which appear to be restricted to the Western Hemisphere. f Spelerpes has one species (S. fuscus) in the south of Europe, and Amblystoma one (A. persimile) in Siam, remarkable instances of separation in genera. The urodele Amphibia of North America, north of the Mexican boundary, num- ber about fifty species. The permanent larval forms of one or more species of Amblystoma (A. tigrinum, A. mavortium), known as * Boulcnger gives tlie range of Cbthonerpeton indistinctum as extending to Buenos Ay res ; but this is considered doubtful by Peters (u Monatsb. Beii. Akad.," 187H, p. 940). t American zoologists recogrise the Plethodontidse (with Spelerpes), Des- mognatMdoe. and Amblystorr.idse as distinct families ; Diemyctylus, repre- senting the Pleuroddidse, is by Boulen^er considered to be synonymous with the Eurasiatic Molge (Triton of Laurenti). AMPHIBIA. 307 axolotls, occur in various parts of Mexico and the Western United States (California, Wyoming). The greater number of the Old World salamandroids belong to the genus Molge (or Triton), whose range extends from Great Britain (M. cristata; M. palmata) to China and Japan (M. pyr- rhogastra; M. Sinensis), and south to Syria and the Mediterranean coast of Africa. The species having the most extended range ap- pear to be M. cristata and M. vulgaris, both of which are distrib- uted throughout the greater part of Europe, and largely also over temperate Asia. The most northerly point reached by any species seems to be about 63° 30' (M. vulgaris or aquatica, in Norway). The Alpine triton (M. alpestris) ascends the Alps, according to Fatio, to an elevation of about 8,000 feet (2,500 metres), while a Mediterranean species (M. montana) inhabits the Lago d'Argento, on Monte Cinto, in Corsica, at an altitude of 6,000 feet. The genus Salamandra has three species, which collectively inhabit the greater part of Central and Southern Europe, the Caucasus, Asia Minor, and Algeria. Salamandra atra, the black or rain salaman- der, inhabits the mountain - regions of Savoy, Switzerland, and Austria between altitudes of 2,500 and 10,000 feet. The anurous, or tailless, amphibians (frogs and toads), which, as has already been seen, comprise not less than eight hundred species, enjoy a much broader distribution than the tailed forms, being absent only from the regions of high northern and south- ern latitudes, and the remote oceanic islands. The genera Rana and Hyla are each represented by a single species in the Solo- mon Islands, and Cornufer (Ranidse) by three species (C. dor- salis, C. Vitianus, C. unilineatus) in the Feejee Islands. The only family that is entitled to be considered in any way cosmopolitan is that of the toads (Bufonidse), which are only absent, apart from local areas and the strictly oceanic islands, from Madagascar, New Guinea, and New Zealand. The genus Bufo, which in itself com- prises nearly eighty out of a total of some ninety species belonging to the family, covers the entire range, with the exception of Aus- tralia, where it is replaced by the genera Pseudophryne, Notaden, and Myiobatrachus. The most broadly distributed species of the genus is the common European toad or paddock (B. vulgaris), whose range comprises practically the whole of Europe, Asia as far east as Japan, and Northern Africa ; in Switzerland it ascends 308 GEOGRAPHICAL A>TD GEOLOGICAL DISTRIBUTION the Alps to a height of nearly 7,000 feet. Bufo calamita and B. viridis are likewise distributed throughout the greater part of Europe, the latter extending its range eastward to Turkestan. — The greater number of the American species occurring north of the Mexican boundary belong to the Sonoran transition-tract, where some six or seven species are met with. The common form of the Eastern and Southern United States is the Carolina toad (Bufo lentiginosus), of which several distinct varieties are recognised. A number of bufonine species are found in the West Indies, and Bufo (Chilophryne) dialophus is said to inhabit the Sandwich Islands. Next to the toads the most broadly distributed family is that of the true frogs (Ranidse), which are most abundantly developed in the Oriental and Ethiopian tracts, but are almost entirely absent from Australia. Of some two hundred species (representing eighteen genera), recognised as belonging to this group, somewhat more than half belong to the genus Rana itself, whose distribution is practi- cally that of the family. The genus is absent from the southern parts of South America — in the whole of which continent there have been determined thus far only three or four species — and from New Zealand, but is represented by a single species (Rana Papua) in North Australia. A solitary species (Rana Krefftii) is also found in the Solomon Islands. The most broadly diffused Old World form is the green or edible frog (Rana esculenta), whose habitat extends from England and Scandinavia to North Africa, and east- ward through Central Asia to China and Japan ; the species is wanting in the island of Sardinia.* Somewhat less broadly dis- tributed through Eurasia is the common frog (R. temporaria), which is the most northerly of known species, ranging in Norway (var. platyrhina) to beyond the seventieth parallel of latitude. In the Alps it still frequents the waters at an elevation of 8,000 feet. The two commonest species of Eurasian frog have their American representatives in the shad- or leopard-frog (R. halecina) and wood- frog (R. sylvatica) — the latter by some authors considered to be identical with R. temporaria — both of which are widely distributed in the United States. The largest American species of the genus, which alone represents the family north of the Mexican frontier, * Schrciber affirms that the species is also wanting in Great Britain ; but the British Museum Is in possession of a specimen from Cambridgeshire (Boulenger, " British Museum Catalogue," 1882). AMPHIBIA. 309 is the common bull-frog (R. Catesbiana). Among the more im- portant remaining genera of Ranidee are Rhacophorus (with Polype- dates, according to Boulenger), whose thirty or more species inhabit Japan, the Philippines, Southeast Asia, India (with Ceylon), and Madagascar; Txalus, with about twenty -five species, restricted to the East Indies ; and Rappia, with a nearly equal number of species, inhabiting tropical Africa. The tree-frogs (Hylidae), with upwards of one hundred and sixty species, find their greatest development in the Neotropical region, which contains somewhat more than one hundred species. The genus Hyla itself is represented by nearly ninety species, or by nearly three-fourths of all the known forms. The species of the North American fauna are comprised in the genera Hyla, Acris, and Chorophilus. Temperate Eurasia has but a solitary representa- tive of the family, the common tree-frog (Hyla arborea), which, in its several varietal forms, is distributed from Great Britain and the Canary Islands to Japan. Hyla Chinensis and H. annectens, the latter from North India, are the only other Asiatic species. The genus Hyla is wanting in the Ethiopian realm, but is represented by several species on the continent of Australia, whose amphibian fauna is made up almost exclusively of the families Cystignathidas (about twenty species), Bufonidae (six species), and Hylidae (eleven species — Hyla and Hylella). Scarcely inferior in point of specific development to the tree- frogs are the Cystignathidee, whose one hundred and fifty or more species are almost entirely restricted to Australia (with Tasmania) and South America, a few species penetrating northward into Mexico and the West Indies, and three or four into the Southern and Western United States (Florida, Texas, California). The family may, there- fore, be said to be distinctive of the Southern Hemisphere. The most abundantly represented of its numerous genera is Hylodes (forty-five species, tropical America), peeping-frogs, many of whose species partake of the habit of the common tree-frogs. Collectively the species are very broadly distributed, and penetrate far beyond the region of elevated temperatures. Hylodes leptopus, about the most southerly of all known species of frog, descends to the Strait of Magellan, while H. Whymperi was obtained by Mr. Whymper on the slopes of Chimborazo at an altitude of 13,200 feet. Paludicola marmorata (Leiuperus viridis), a member of the same family, was 310 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. found by Tschudi in the Peruvian Andes at an elevation of nearly 16,000 feet. Among the more distinctive forms of Cystignathidae are the horned frogs (Ceratophrys), which inhabit tropical America from Guiana to Uruguay. A remarkable instance of a divided family among the Anura is furnished by the Dendrobatidse, which comprise two genera and ten species, one genus, Mantella, being confined to Madagascar and Nossi Be, and the other, Dendrobates, to tropical South America. The intermediate tracts are entirely devoid of representatives of the family. The number of families restricted to a single zoogeographi- cal region is five, of which four characterise the Neotropical realm — Dendrophryniscidse, Amphignathodontidse, Hemiphractidae, and Pipidae — and one, the Dactylethridae, the Ethiopian. The most limited of all the families is the Pipidae, which is restricted to a single species, the Surinam toad (Pipa Americana), an inhabitant of Guiana and Brazil. Of the four primary groups to which the animals of this class are referable, the Stegocephala (corresponding to the Labyrintho- dontia of most authors), Gymnophiona (coecilians), Urodela (sala- manders, tritons), and Anura (frogs and toads), the first acquires special geological importance from the fact that all, or very nearly all, of the older forms are comprised within it. Remains of Urodela are only doubtfully known from the Paleozoic deposits, while the anurous type does not appear before the Tertiary epoch ; no fossil ccecilian has as yet been discovered. The now wholly extinct order Stegocephala, which comprises salamandroid and ophidian forms more or less covered with a pro- tecting armour of bony (ganoid) plates, dates from the Carbon- iferous period (Hylerpeton, Batrachiderpeton, Pelion, Dolichosoma, Ophiderpeton), when, or at a still considerably earlier era, they appear to have become differentiated from the type of lung-fishes (Dipnoi) or of the dipteroid ganoids. A further development of types, with a partial persistence of Carboniferous genera, is mani- fest in the Permian deposits, where, as in the older strata, the forms are principally referable to the division Ganocephala (Archae- gosaurus, Dendrerpeton, Branchiosaurus, Protriton,* Hylonomus, * Protriton Petrolei is by Dcichmuller considered to be identical with Branchiosaurus gracilis. AMPHIBIA. 311 Limnerpeton, Melanerpeton), in which the peculiar labyrinthine in- folding of the teeth, distinctive of the true labyrinthodonts, is largely absent. The apodal and coecilian-like division Aistopoda is represented among other forms by the Carboniferous genera Dolichosoma and Ophiderpeton, and by Palseosiren and the Ameri- can Molgophis. Contemporaneously with these types we have also the true labyrinthodonts, whose earliest member appears to be Baphetes, from the Carboniferous deposits of Pictou, Nova Scotia. The full development of this group does not obtain, however, be- fore the Triassic period, at the close of which the entire order of animals seems to have become extinct in most regions.* Among the more distinctive genera of this period are Labyrinthodon, Mas- todonsaurus, Trematosaurus, and Metopias, to one or several of which probably belong the foot-prints of the fanciful animal desig- nated Cheirotherium. Of the perennibranchiate division of the Urodela, in which ex- ternal gills are retained throughout the entire existence of the animal (Siren, Proteus), we have as yet no positive indications in any of the rock- formations. The Cryptobranchia, which retain a gill-opening after the absorption of the gills — the American Am- phiuma and Menopoma, and the giant salamander of Japan, Cryp- tobranchus Japonicus — seem to have one or more fossil representa- tives in the genus Andrias (Cryptobranchus of some authors), from the Miocene deposits of Oeningen, Germany, to which are referred the remains presumed by Scheuchzer to be those of earliest man (Homo diluvii testis). It is certainly a very remarkable fact in dis- tribution that the only link uniting the so widely separated, but closely related, genera Menopoma and Cryptobranchus, should be this extraordinary form from the middle Tertiary period. Its ex- istence would seem to indicate a former much broader diffusion of this particular group of animals, and a very different distribution of land and water areas than now obtains. — The caducibranchiate urodeles (salamanders, tritons), which in their transformation to lung- * Brachyops, from the Damucla beds of India, and one or two other genera of Inbynnthodonts have been indicated as belonging to the Jurassic period, but it may be questioned whether the age of the deposits in which these remains occur has been as yet satL-factorily determined. Excepting these somewhat doubtfully placed forms, no amphibians are known from Jurassic strata. 812 GEOGEAPHICAL AND GEOLOGICAL DISTRIBUTION. breathers pass one stage beyond the cryptobranchs in the oblitera- tion of the gill-aperture, date from the Eocene period, when forms more or less nearly allied to recent types appear. Both Triton and Salamandra are represented. Remains of tailless amphibians (Anu- ra), more or less nearly allied to modern forms, have been obtained from the Tertiary lignitic and fresh-water strata (Oligocene, Mio- cene) of Western and Central Europe. Palseobatrachus diluvianus, one of the oldest knowTn forms, is from the lignitic strata of Orsberg, near Bonn, Germany. The lacustrine deposits of Oeningen have yielded several extinct genera, among which are Latonia (related to the Brazilian horned toad, Ceratophrys), Palseophrynus (a bufonine type), and Pelophilus, the last not impossibly a true Bombinator. Among recent genera, Rana, Bufo, and Pipa have also Tertiary representatives. The paucity of remains of existing types of amphibians, com- bined with the circumstance of their very late appearance, renders impracticable the determination of the phylogenetic relationships which bind together the various groups. Equally uncertain are the stages which mark the differentiation of the modern fauna from that of the Palaeozoic and the early Mesozoic periods, nor is it likely that any progress towards the solution of this problem will be effected until the void which is caused by the almost total absence of amphibian remains from the deposits of Jurassic and Cretaceous age will have been in great part filled. That the animals in question are derived either in whole or in part from the dipnoan type of fishes there is very little doubt, but the immediate connecting link or links between their ichthyic pro- genitors, whatever these may have been, and the earliest stego- cephalic forms are still wanting. The apparently sudden disap- pearance with the Triassic period of the largely represented order which contained all, or very nearly all, the earlier forms of amphib- ians, without leaving in the modern fauna any positive indications of its former existence, is not a little surprising, but it appears not unlikely that the coecilians, which in many points of structure resemble the ophidian labyrinthodonts, represent at least a part of this ancient stock. Again, by many geologists the crocodiles are' assumed to be the modified descendants of the true labyrintho- donts. TURTLES. 313 KEPTILES. Chelonia. — The total number of known species of chelonians is estimated by Hoffmann (1880) to be somewhat more than two hun- dred and fifty. Of these only five are marine forms, the rest being inhabitants of the land and its fresh waters. The former, com- prised in the genera Dermatochelys (or Sphargis), Chelone, and Thalassochelys, are very broadly distributed throughout the tropi- cal and sub-tropical or temperate waters of both the Old and the New World, most of the species being cosmopolitan, or nearly so. Dermatochelys coriacea, the leathery turtle, is found along the American border from Brazil to South Carolina and Massachusetts, exceptionally on the European coast, and in the Indian and Pacific oceans, from Africa to Chili. The green turtle (Chelone viridis), which is held in such high estimation as an article of food, has an equally extended range, although it is but very rarely found on the European coast (England to the Mediterranean). Of still rarer oc- currence in the European seas is Chelone imbricata, the hawk's bill, which yields the tortoise-shell of commerce, and whose habitat embraces nearly the whole circumference of the globe. The log- gerhead (Thalassochelys corticata) is abundant on both sides of the Atlantic, and in the Mediterranean, and is at rarer intervals also met with in the Indo-Pacific basin. The land and fresh- water chelonians have a very unequal dis- tribution, being most abundant in the region of the tropics, and rapidly diminishing as we pass either north or south into the tem- perate zones. The greatest number of forms belong to tropical and sub-tropical America, and the smallest number to Australia and temperate Eurasia, each of which possesses some fifteen species. The northern limit reached by these animals in the Western Hemi- sphere is about the fiftieth parallel of north latitude (Chelydra serpentina), and not improbably the same parallel marks the cor- responding general limit in the Eastern Hemisphere, although in Europe Cistudo lutaria or Europaea, the most wide-spread species of the continent, is found as far north as the fifty -fourth parallel (Mecklenburg), and possibly still farther. The total number of European species is five, most of which more properly belong to the region about the Mediterranean. No species is known from Great Britain, the Scandinavian Peninsula, Denmark, Holland, or 314 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION". Belgium. Cistudo lutaria and Testudo Graeca, the latter intro- duced, inhabit the waters of Southern France; the first of these also inhabits Switzerland, but it is only doubtfully indigenous to that country. The number of species occurring in North America north of the Mexican boundary is about forty, nearly one-half of which properly belong to the Southern United States. Among the commoner or better known forms are the box-turtles (Cistudo, Cinosternum), wood-turtles (Chelopus), painted-turtles (Chrysemys), marsh -turtles (Malacoclemmys), terrapins (Pseudemys), musk-turtles (Aromochelys), snappers (Chelydra), and soft-shells (Aspidonectes), all of which are very broadly distributed, especially in the Eastern and Southern United States. The species having the most ex- tended range is the common snapper (Chelydra serpentina), which is found from Canada to Ecuador. Two other species, the common wood-turtle (Chelopus insculptus) and the painted-turtle (.Chrysemys picta), range as far north as Canada. Most of the species of Chelonia are restricted to a single faunal region, and where identical species are found in more than one continent, the range of the species on the continent not properly its home is, as a rule, very limited. Two species are known to be common to Europe and (North) Africa — Testudo nemoralis and Cistudo lutaria (C. EuropaBa); one species, Pyxis arachnoides, is common to the continent of Africa (with Madagascar, and some of the neighbouring islands) and India; and likewise one, Manauria fusca, common to the East Indies (Java, &c.) and Australia. Tur- tles are wanting in the true oceanic islands, but they are sufficiently abundant in many of the continental islands, even where these are distant several hundred miles from the nearest mainland. Two of the most ponderous representatives of the order belong to such isl- and groups : the Galapagos turtle (Testudo nigra) and the elephant turtle (T. elephantina), the latter, whose weight is known to reach five hundred pounds, inhabiting the Seychelles and some of minor island groups of the Mozambique Channel. Of the four more generally recognised families of land and fresh- water turtles, the Testudinidae, Emydse, Chelydae, and Triony- chida3, only the first has representatives in all the major divisions of the earth's surface. Australia is lacking in both the Emydas and Trionychidae, the latter being also absent from South America, while the three southern continents are almost the sole possessors LIZARDS. 315 of the Chelydae. Remarkable instances of discontinuous genera are seen in Hydromedusa, one species of which inhabits the Oriental realm and the remainder the continent of South America, and in Podocnemis, whose species are divided between South America and Australia. The earliest chelonian remains occur in deposits of Jurassic age (Switzerland, Germany, France), in which a well-marked differentia- tion of the modern families Emydse Cas seen in the genera Thalas- semys, Eurysternum, Tropidemys, Helemys, the last supposed to have been closely related to the American snapper) and Chelydaa (Plesiochelys, Idiochelys, Craspedochelys) already appears. The number of forms is materially increased in the succeeding Creta- ceous deposits, where, in addition to the representatives of the two families already indicated (e. g., Platemys, Pleurosternum, Adocus, Euclastes, Osteopygis), we have those of the Trionychidas and Cheloniidas (Trionyx, New Jersey ; Chelone, Maestricht chalk, and greensand of New Jersey). Protostega gigas, a marine turtle from the deposits of this age of Kansas, attained a length of upwards of twelve feet. Many of the recent genera, as Testudo, Chelydra, Emys, Cistudo, &c. , appear as fossils in the early or middle Tertiary deposits. The most extraordinary of all extinct forms is the giant land-tortoise of the Siwalik Hills of India, Colossochelys atlas, which measured apparently not less than fifteen to twenty feet in length. Of somewhat less than one-half these dimensions was the Macrochelys mira, from the molasse of Southern Germany (Ober- kirchberg, naar Ulm), whose modern representative is the Missis- sippi snapper (Macrochelys lacertina). Lacertilia. — The number of known species of. lizard is esti- mated by Giinther to be about seventeen hundred, of which by far the largest part is confined to the warmer regions of the earth's surface. But comparatively few forms are found to pass beyond the fortieth parallel of latitude, and at about the sixtieth parallel (north) the order practically disappears. The most northerly spe- cies is Lacerta vivipara, whose range comprises nearly the whole of Europe, and extends northward to the seventieth parallel (in Nor- way) ; it is accompanied as far as Lapland by the no less broadly distributed blind- worm (Anguis fragilis). In the Western Hemi- sphere the northward extension of the order is much more limited than in the Eastern, and it would appear that only one species, a 316 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. Gerrhonotus, passes beyond the fiftieth parallel ; in the Middle United States the northern skink (Eumeces septentrionalis) pene- trates into Minnesota, and along the Atlantic border Eumeces fas- ciatus, a species, singularly enough, also found in Japan, forms part of the Massachusetts fauna. The most southerly range of any species is that of Liolsemus Magellanicus, which reaches the Strait of Magellan. In the whole of Europe north of the forty-fifth parallel of lati- tude, or what might be considered to be Central and Northern Europe, there are scarcely more than a dozen species of lizard, of which nearly one-half belong to the genus Lacerta, or common lizard. Scandinavia, Great Britain, and Denmark have each three (and the same) species: Lacerta vivipara, L. agilis, and Anguis fragilis. An additional species, the wall-lizard (L. muralis), be- longs to Belgium and Holland, and a fifth one, the green lizard (L. viridis), which has also found a congenial home on the island of Guernsey, to Germany. All of these species form part of the southern or Mediterranean fauna, which in Europe comprises some thirty-five or more species, many, or most of them, of a distinctively African type. The affinities with the tropical faunas are seen in the development of the geckotine type (Hemidactylus verruculatus, the common gecko of the houses of Southern Europe ; Gymno- dactylus, Phyllodactylus, Platydactylus) and the agamas (Agama, Stellio— South Russia and the Balkan Peninsula), the Old World representatives of the American iguanas. One species of chamseleon (Chamseleo vulgaris) is found in Andalusia. In temperate North America lizards are even more scarce than in the equivalent region of the Old World. Indeed, in the whole of the continent north of a line that might be considered to unite San Francisco with Galveston in Texas there are probably less than twenty species, of which more than one-half belong to the Old World genus of skinks, Eumeces. A distinctive feature separating the saurian fauna of this tract from the European is the absence of the group to which all the commoner European forms (Lacerta) belong, although the genus Xantusia, from the Pacific coast, is by some authors doubtfully referred to the Lacertida3. On the other hand, a distinct Old World relationship is established in the glass- snake (Ophiosaurus — from Tennessee southward and westward), a near ally of which is the glass-snake (Pseudopus) of Southern Eu- LIZAKDS. 317 rope (Dalmatia, Hungary, Russia) and West-Central Asia.* In the region lying south of the San Francisco-Galveston line, which is largely in the form of parched or desert tracts, the prevalence of a considerable number of tropical or South American types imparts a distinct individuality, or non-North American character, to the fauna, which is best expressed in the family of iguanas (Iguanidae). This group is represented by not less than forty species, the greater number of which belong to the genera Sceloporus and Phrynosoma (u horned-toad "), one species of the latter genus penetrating as far north as Dakota, f Inhabiting the same tracts, but extending its range to Tehuantepec, is the venomous Heloderma. A single spe- cies of amphisbsenian, the "thunder-worm" (Rhineura Floridana), is known from Florida. The so-called chamseleon of the Southern United States is the green goitred lizard Anolis. The more distinctive or most largely represented tropical fami- lies of lizards are the iguanas, agamas, monitors, geckos, amphis- ba3nians, and chamseleons. The first of these is almost exclusively American, and is represented by probably not less than three hun- dred species, of which nearly, or fully, one-third belong to the genus Anolis, whose members especially abound in the West India islands. The genus Iguana is more properly South American, al- though also found in some of the West Indies, and penetrating northward into Mexico. Basiliscus, the basilisk, likewise ranges into Mexico. In addition to the forms that have already been in- dicated as belonging to the United States, may be mentioned Uta, Callisaurus, and Holbrookia, the last of which is sufficiently abun- dant in certain parts of Texas and the transition-region to the northwest. The most remarkable member of the family is the Galapagos leguan (Amblyrhynchus), which is partially marine in its habits. Brachylophus inhabits the Feejee Islands. No iguanian * A variety of this species also occurs in Morocco. M. Boulenger has re- cently attempted to show ("Ann. and Mag. Nat. Hist.," Aug., 1885) that the North and South American lacertiiian faunas are, strictly speaking, one, the Neogean, a conclusion which is not borne out by the facts of distribu- tion. The misconception arises from the incorporation of the tract lying south of the line indicated above with the North American faunal region proper, while in reality it is a transition-tract more nearly Neotropical in character than " Nearctic." t Phrynosoma orbiculare was found by Mr. Geddes on the plateau of Mexico at an altitude of 7,500 feet. 318 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. is found on any of the continental divisions of the Old World, but two genera of terrestrial habits, Hoplurus and Chalarodon, appear in Madagascar. The true Old World representatives of the iguanas are the agamas (Agamidre), which might be said to present a parallel series of forms to the iguanian types of the New AVorld. Their distribu- tion covers the greater part of the continent of Africa, the warmer tracts of Asia, especially the islands of the East Indian Archipelago, and much of Australia. No species has thus far been obtained from New Zealand. A limited number of species is found in Asia north of the Himalayas (Trapelus, Phrynocephalus — Tartary to Afghanistan), and their occurrence in Southern Europe (Agama, Stellio *) has already been noted. Several genera have representa- tives in the Andaman and Feejee Islands groups. Among the more remarkable forms of the family are the flying-lizards (Draco) — in- habitants of the East Indies (except Ceylon) — which are provided with a tegumentary expansion specially adapted for sailing through the air; the frilled lizard of Queensland, Australia (Chlamydosaurus Kingii), which is ornamented with a broad fan-like collar nearly encircling the head and neck ; and the spine-covered Moloch hor- ridus of Southern and Western Australia. The agamas proper range throughout Africa, and eastward to India. The geckos (Geckotidae), which, with the exception of the cosmopolitan skinks, have the broadest distribution of all the lacertilian families, number about two hundred species. They occur in the hotter parts of all the continental regions, and are largely represented even in the more distant oceanic islands — Ma- deira, Ascension, the Seychelles, New Zealand; the Solomon, An- daman, and Sandwich Islands groups, &c. -"-evidently possessing some special means for dispersion which is wanting in other rep- tiles. Several of the more largely represented genera, as Gymno- dactylus, Phyllodactylus, and Hemidactylus, have practically the range of the entire family ; Gonatodes is found in tropical America and East India, but is wanting in Africa. The genus Gecko, as restricted, has about seven species, which are confined to China, Japan, the Papuan Islands, and the islands of the East Indian Archipelago. Most of the geckos are nocturnal in* their habits, * Boulenger (" Catalogue of Lizards," British Museum, 1885) considers Trapelus and Stellio as synonyms of Agama. LIZARDS. 319 and it would appear that the different species intentionally keep apart from each other. Colonel Tytler observes that "although several species of geckos may inhabit the same locality, yet, as a general rule, they keep separate and aloof from each other ; for in- stance, in a house the dark cellars may be the resort of one species, the roof of another, and crevices in the walls may be exclusively occupied by a third species. However, at night they issue forth in quest of insects, and may be found mixed up together in the same spot; but on the slightest disturbance, or when they have done feeding, they return hurriedly to their particular hiding-places."116 Remarkable instances of broad specific range are presented by Hemi- dactylus mabouia, which inhabits Brazil, San Domingo, Eastern Africa, and Madagascar, and Gehyra mutilata, whose range extends from the Mascarene Islands through India, the Malay Peninsula, and New Guinea to Mexico. With the exception of the common chamseleon (Chamaeleo vulgaris), whose range extends from Anda- lusia through North Africa eastward to India and Ceylon, all the species of the family are restricted to the African continent and the neighbouring islands (Madagascar, Bourbon, and Fernando Po, the first with nearly one-half the total number of species). The monitors, or water-lizards (Varanidse), which range over the greater part of Africa, East India, Australia, and the Austro-Malay- an islands, comprise the largest Old World members of the class, some of the species measuring, or -exceeding, six feet in length. The common monitor of the Nile (Monitor Niloticus) is found in the neighbourhood of all the more important streams of tropical Africa. Psammosaurus scincus, a North African species, is strictly terrestrial in its habits. The amphisbaenians, or footless lizards, which by Dr. Gray are elevated to the rank of a distinct order, are principally tropical American forms, although a considerable number of species are known from the African continent, and a few, of the genus Blanus, from the Mediterranean districts of Europe and Asia. In America the species range from the Argentine Republic through the West Indies to Florida (Rhineura [Lepidosternon] Floridana). Chirotes lumbricoides, which is provided with the anterior pair of appen- dages, is a Mexican species. A distinctively American family of lizards is the Teiida3, or teguexins, which may be said to replace the Old World Lacertida3, and whose range extends from Patagonia 320 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. to Montana and Virginia. Upwards of a hundred species have been described. New Zealand (or rather the small islands off the northeast coast) possesses a remarkable lizard in the genus Hatteria or Sphenodon, which in many points of structure departs from the type of true lizards, and approximates it to an ancient lost form from the Trias, the genus Hyperodapedon. The earliest known representative of the Lacertilia is Protero- saurus, from the Permian deposits of Germany and England, which appears to be most nearly related to the monitors, from which, however, it differs in its thecodont dentition.* Hyperodapedon (which, with the contemporaneous Rhynchosaurus, and the recent Hatteria, is by some authors constituted into a distinct order, Rhynchocephala) and the acrodont genus Telerpeton (Elgin lime- stones of Scotland) appear in the Trias, and are succeeded in the deposits of Jurassic age by a number of more or less obscurely defined genera (Geosaurus, Homceosaurus, Acrosaunis, Anguisau- rus), whose relationships with modern forms are in most cases not clearly indicated. Lacertilian remains are not abundant in the Cre- taceous deposits, and such as have been preserved are mainly in a fragmentary condition ; the recent genus Hydrosaurus, one of the monitors, is indicated. In Tertiary strata the remains become nu- merous, and belong in considerable part to modern types. Frag- mentary skeletons from the European Miocene deposits have been referred to Iguana and Lacerta, and, doubtfully, also to Scincus and Anguis. — No true lacertilians are known from American de- posits older than the Eocene. The western lake-basins of this age have yielded numerous remains, which are referable to a number of distinct genera — Glyptosaurus, Iguanavus, Oreosaurus, Tino- saurus, Saniva — and some of which appear to have survived into the Miocene. Among the very limited number of forms of this period may be mentioned Peltosaurus, doubtfully referred to the Gerrhonotidae, and Cremastosaurus, the latter of about the size of the horned-toad. Ophidia. — The distribution of the Ophidia is very similar to that of the Lacertilia, the order being most numerously represented * Professor Seeley believes it probable that Protcrosaurus is a dinosaur. (Phillips, " Manual of Geology," edited by Etheridge and Seeley, 1885.) SERPENTS. 321 in the tropical regions of the earth's surface, and rapidly diminish- ing toward either pole. Excepting, however, members of the family of water-snakes (Hydrophidse), which are especially abun- dant in the Australian and Indian seas— ranging westward to Mada- gascar, and eastward to Panama — the order is only exceptionally represented in the strictly oceanic islands, in this respect differing from the lizards and agreeing with the amphibians. Evidently, the animals of this class, like the Amphibia, possess no facilities for traversing broad arms of the sea. With our deficient knowledge of many of the more favoured regions of the globe it is impossible to arrive at any estimate of the numerical extent of the order, but it may be safely assumed that there are considerably more than one thousand clearly defined species known to naturalists, of which very nearly one-half are found in British and Farther India, and the East Indian Archi- pelago. Mr. Blanford places the number of species from British India and its dependencies alone at two hundred and seventy-four.* In Europe, north (and inclusive) of the Alps, there are some fifteen or more species, of which three, the common viper or adder (Vipera [Pelias] berus), the grass or ringed snake (Tropidonotus natrix), and the Coronella Austriaca (Isevis), penetrate beyond the fifty-fifth parallel of latitude. These are the only species found in Scandi- navia, the British Isles, Denmark, Holland, and Belgium. The most northerly of all serpents is the common viper, whose range embraces the whole of Europe and Northern Asia, and which in Scandinavia extends to the Arctic circle; in the Alps it is occa- sionally met with at an altitude of nine thousand feet. The north- ern limit of the ringed snake appears to be the sixty-fifth parallel. Germany has in all six or seven species,116 the three above men- tioned, and Tropidonotus tessellatus, Elaphis flavescens (TEsculapii), Zamenis viridiflavus (doubtful), and Vipera aspis (the asp), the last very largely distributed throughout the whole of France and Switz- erland, and the commonest of the venomous serpents of Italy. It does not appear to ascend the Alps to elevations much excee*ding * The census of the other Reptilia is as follows : Chelonia fifty-four, Cro- codilia four, Lacertilia one hundred and eighty-two. The Amphibia com- prise about one hundred species, of which one only belongs to the tailed division, and five to the Pseudophidia (Ccecilia). "Journ. Asiatic Soc. Bengal," Dec., 1881. 15 322 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. five thousand feet. The species, like many of the other Mediter- ranean forms, is found also in Algeria, but seems to be absent from Morocco.117 All of the German species occur in Switzerland, which, however, numbers one additional form, Tropidonotus viperrinus. The Mediterranean fauna, or what might be considered to be the fauna of Southern Europe, comprises about thirty species, in which are included probably all the forms that occur elsewhere in Europe ; the Iberian Peninsula numbers twelve species, 118 Italy about fifteen, and Greece fourteen.119 In the western half of this region the similarity existing between the ophidian faunas of the several countries amounts almost to identity, but eastward, from the Balkan Peninsula to the Crimea, a gradual exchange of species is effected, so that in both Turkey and Russia nearly, or fully, one-half of the species (about fifteen in each country) are distinct. Somewhat more than one-half of the Italian and Iberian species are also found in the region south of the Mediterranean — Algeria and Morocco. The North American serpents, or those found north of the Mexican boundary, belong in the main to two families, the colubers (Colubridre) and pit- vipers or rattlesnakes (Crotalidae), the former numbering some one hundred and ten or more species, and the latter about twenty. In addition to these there are a limited num- ber of representatives of three or four other families. Thus, the worm or burrowing snakes (Typhlopidae), whose species are abun- dantly distributed over the tropical regions of both hemispheres, occur sparingly in California and Texas (Stenostoma) ; the Erycida3, a limited family of Old and New World serpents allied to the boas, are represented on the west coast by two species of Charina (Cali- fornia to Puget Sound) ; and the venomous Elapidae, to which very nearly two-thirds of all the Australian snakes', and the deadly cobra (Naja), Bungarus, and Ophiophagus of India belong, are repre- sented by the harlequin-snake (Elaps fulvius) in the Southern United States (east of the Mississippi), and by Elaps euryxanthus in Ari- zona. The greater number of these forms can scarcely be said to constitute a part of the North American ophidian fauna proper, inasmuch as they occur principally in a border tract whose gen- eral faunal relationship is more nearly with the region lying to the south than the north. The North American colubrine snakes are comprised principally in five or six groups or genera : 1. Tropidonotus, water-snakes, whose SERPENTS. 323 range is coextensive with the whole United States, and whose best known exponents are the ribbon-snake (T. [Eutcenia] saurita), water- snake or adder (T. sipedon), and garter (T. [Eutaenia] sirtalis), the first two abundant in the region east of the Mississippi, and the last found almost everywhere from Canada and Nova Scotia to Mexico and Panama. 2. Coluber, whose range is no less extensive than that of the water-snakes, and which embraces among other forms the most broadly distributed black-snake or constrictor (C. [Basca- nium] constrictor) and the coachwhip-snake (C. flagelliformis) of the Southern States. 3. Pityophis, pine-snakes. 4. Elaphis, to which the spotted racer (E. [Scotophis] guttatus), chicken-snake (E. quadrivittatus), and pilot (E. obsoletus) belong, the first two prin- cipally from the Southern States, and the last generally distributed over the Atlantic border, from New England to Alabama. 5. Ophi- bolus, king-snakes, whose species are widely diffused throughout the United States, and whose best known representatives are the southern chain-snake (O. getulus) — with a western variety known as the king-snake (O. Sayi) — the red-snake (O. doliatus), and the very common milk-snake or spotted adder (O. triangulus), whose range extends from the Atlantic border to the Mississippi, and northward to Canada ; and, 6. Diadophis, ring-necked snakes, rang- ing nearly through the entire continent south of the Canadian line. The genus Cyclophis comprises two common species of green- or grass-snake, the summer-snake (C. sestivus) and spring-snake (C. vernalis), both of which have a very extensive distribution. Two species of hog-nose snake (Heterodon) occupy a considerable part of the United States, and are locally known as blowing-vipers or adders. The remaining colubrine forms are embraced in genera largely limited as to the number of species, and which in many cases, as in Contia, Tantilla, Sonora, &c., are confined to the transition-tract which unites with the Neotropical realm. The North American crotaloids are comprised in three or more genera: Crotalus, the rattlesnakes proper, Sistrurus (or Crotalophorus), the prairie or grass rattlesnakes, which are confined principally to the Central and Southern United States, and Ancistrodon, the copperheads and moccasins. Of the last there are three species: A. contortrix, the copperhead, whose habitat is the greater part of the region east of the Mississippi; A. piscivorus, the true or water moccasin, which 324 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. inhabits the waters of the Southern States from South Carolina to Texas; and A. atrofuscus, the highland moccasin, found in the mountain-region south of Virginia, and by many authors considered to be only a variety of the last. The rattlesnakes proper are repre- sented by some ten or more species, most of which are found in the region of the Southwestern United States. Five species (or varieties) are known east of the Mississippi, of which the common or banded rattlesnake (C. horridus), which is still abundantly dis- tributed between Texas and New England, has the most extended range. The diamond - rattlesnake (C. adamanteus) inhabits the Southern States. Although the greater number of species of North American non- venomous Ophidia belong to genera or groups which are also largely developed in, and are equally characteristic of, the Old World, as Tropidonotus (EutaBnia), Coluber (Bascanium), and Elaphis (Scoto- phis), types but barely represented in the Neotropical realm — thus clearly indicating the Old World affinities of the so-called " Nearc- tic " fauna, it appears that all the species are distinct.* This is not very surprising in view of the limited northern range, especially in the Western Hemisphere,! of the majority of the species, which are incapable, and have been incapable for a long period past, of tra- versing the chilled northern tracts by which at one time, doubtless, a union was effected between the two hemispheres. As a result of this isolation new species have been formed. It is more remarkable that the most northern of all ophidian genera, Viperus, the viper, whose appearance on the American continent might have been con- fidently looked for as a result of its extended range, is completely wanting. Other anomalies of distribution are presented by the distinctively American genera Heterodon ami Dromicus, both of which have representatives in the island of Madagascar, and the family of pit- vipers (Crotalidae), which is largely developed in the Oriental realm, but is wanting in Africa. Of the more important families of tropical and sub-tropical * By most American herpetologists Eutamia, Bascanium, and Scotophis are considered to be distinct from the O'd World genera with which they have been united by the greater number of European naturalists. t Several species arc found in British Columbia along the Canadian bound- ary-line, but it is doubtful whether any penetrate much beyond the fiftieth parallel of latitude. SERPENTS. 325 snakes — indeed, of all snakes — the colubers take first rank, num- bering probably fully one-fourth of all known species of Ophidia. They are, strictly speaking, the most cosmopolitan of all the vari- ous groups, and are represented, in addition to genera whose dis- tribution embraces several of the zoogeographical regions, by a number of distinct genera in each of the great zoogeographical regions except Australia, where the family is but feebly developed (Tropidonotus, Coronella). Next in importance, and more strictly tropical, are the venomous colubrine snakes (Elapidre), with probably upwards of one hundred species, about one-half of which are con- fined to Australia and the neighbouring islands. The family, which is almost wholly wanting in the north temperate region — repre- sented by the genus Callophis in Japan and by the harlequin-snakes (Elaps) in the United States — comprises many of the most deadly of the Thanatophidia, as the cobra (Naja tripudians), Bungarus, and Ophiophagus of India and some of the eastern islands. Callo- phis bilineatus appears to be the only poisonous snake of the Philip- pines. The genus Elaps embraces all or most of the American species of the family, including the much -dreaded Brazilian coral- snake (Elaps corallinus).* Partaking very nearly of the distribution of the last family are the burrowing-snakes (Typhlopida?), whose numerous members, belongirg chiefly to the genus Typhlops, are found in nearly all the warmer regions of the earth's surface. One species of the genus, Typhlops lumbricalis, is found in Greece and on some of the Grecian islands. The tree-snakes proper (Dendrophida3) are found in all the tropical regions ; the nocturnal tree-snakes (Dip- sadidse) and the arboreal whip-snakes (Dryiophida?) are also essen- tially tropical, but they are either wholly, or almost wholly, wanting in Australia. The boas or pythons (Boida? ; Pythonidse) are one of the most distinctively tropical families, comprising some fifty or more species. The pythons proper (genus Python) are distributed throughout nearly the whole of the Oriental region — the islands as well as the * Many travellers and naturalists, and notably Maximilian, Prince ofWied, have denied the venomous nature of this animal. The researches of Ihenng, however, conclusively demonstrate this nature in Elaps Marcpravii, and would seem, consequently, to uphold the common notion concerning E. corallinus (" Zoologischer Anzeiger," August, 1881). 326 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. mainland — and over the greater part of the continent of Africa, although by some naturalists the Ethiopian species are placed in a distinct genus, Hortulia. The netted python (P. reticulatus) in- habits nearly all the islands of the Malay Archipelago, besides portions of the mainland (Farther India), where it shares in part the habitat of the common Indian species, P. molurus. Among the African species are the royal python of the western forests (P. regia), Seba's python, or the fetich-snake (P. Sebae), whose distribution is much more general, and the Natal rock-snake (P. [Hortuiia] Natalensis). The Australian Pythonida? are included in the genera Morelia, Aspidiotes, Liasis (islands of the Arafura Sea), and Nardoa, to the first of which belong the diamond-snake (M. spilotes) and the carpet-snake (M. variegata) of the colonists. In the New World the pythons are replaced by the boas and anacon- das, which by many naturalists have been constituted into a distinct family, Boidtc, and whose habitat is principally the warmer parts of the South American continent. Boa constrictor, whose home is more properly the equatorial forest region, is represented by several closely allied forms in Central America and Mexico, as B. isthmica, B. imperator, and B. Mexicana, which are by some authorities con- sidered to be mere varieties of the common southern constrictor, and by others as distinct species. A fourth species, the yellow boa (Chilabothrus inornatus), whose home is the West Indies, is doubtfully said to inhabit Central America and Mexico as well.* The anaconda (Eunectes murinus) is found in the tropical waters. The most remarkable instance of a localised family of any ex- tent is presented by the earth-snakes, or rough-tailed burrowing- snakes, as they are sometimes called, the Uropeltidee, whose thirty- five or more species are confined almost entirely to Ceylon and the southern part of the Indian Peninsula, or to the tract constituting the Cingalese sub-region of the Oriental realm. Their headquarters on the peninsula are the western mountain-ranges between Canara and Cape Comorin, only one species, according to Beddome,120 be- ing found in the mountains of the east coast, and but three on the west, whose range extends northward beyond Kudra Mukh in South Canara. Several species of Silybura ascend the Neilgherries to an * The naturalists of the United States Fish Commission steamer " Alba- tross" found a species of boaeform serpent on the island of New Providence, Bahamas ("Science," June, 1886). CROCODILES. 327 elevation of seven thousand feet, and Plectrurus Perrotetii is found between five thousand and eight thousand feet. Fossil remains of serpents are not numerous, and only one spe- cies, the Simoliophis Rochebruni, from the Upper Cretaceous de- posits of the Charente, France, is known to antedate the Tertiary period. Several species of Palaeophis, considered by some authors to have been closely related to the boas, which they rivalled in size, and by others to constitute the type of a distinct family, have been found in the Lower Eocene deposits (Londonian) of England, France, and Italy ; two or three species have been likewise described from the nearly equivalent deposits of the State of New Jersey.* Bose- form serpents appear to be indicated by the Python Euboeicus, from Kumi, in the island of Euboea, and by the remains from the Eocene fresh-water deposits of the Western United States which have been referred to the genera Boavus, Lithophis, and Limno- phis. The genus Coluber is represented by several species from the Miocene fresh- water deposits of the continent of Europe (Oen- ingen, &c.). Fossil Toxicophidia, or venomous serpents, appear to be still less abundantly represented than the non-venomous types. A form supposed to be related to the rattlesnakes has been described from Salonica as Laophis crotaloides, and one, related to the cobra, from Steinheim, as Naja Suevica. The most ancient remains of Ophidia in the New World appear to be those of Hela- gras prisciformis, from the Puerco Eocene, which was of about the size of the black constrictor (Coluber constrictor). The paucity of ophidian remains leaves very uncertain any specu- lations as to the origin or evolution of this order of animals. Whether or not they are in part the modified descendants of the lacertilian pythonomorphs, which they seem to approximate in cer- tain points of structure, still remains to be determined. Grocodilia. — Of the four orders of existing reptiles the Croco- dilia are numerically the least important, and at the same time the most restricted in their distribution. Some twenty-five more or less well-defined species, inhabiting the tropical and sub-tropical regions of the earth's surface, are known to naturalists, by whom three distinct groups or families are recognised : the gavials, croco- diles proper, and alligators. The gavials are exclusively Old World forms, and the alligators forms belonging to the New World. The * Palaeophis littoralis, P. Halidanus, P. (Dinophis) grandis. 328 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. former, as understood by most systematists, are comprised in two genera, Gavialis (with a single species, G. Gangeticus), restricted to the waters of the Indian Peninsula, and Tomistoma, a Bornean form, whose range probably extends to North Australia. The true crocodiles, of which some authors recognise two gen- era, Crocodilus and Mecistops,* inhabit nearly all the larger streams (and many of the lakes) of Africa, India, and the north coast of Australia. Although for a long time supposed to be entirely want- ing in the New World, they are now known to inhabit the waters of tropical America on both sides of the Andes (Ecuador, Colombia, the Orinoco, &c.), extending their range to Mexico and the West India Islands (Cuba, San Domingo, Jamaica). Crocodilus Ameri- canus enters some of the streams of Florida. The species having the broadest distribution appear to be Crocodilus porosus, whose range embraces the area included between the North Australian coast, the Indian Peninsula, and China, and C. vulgaris, the com- mon African form, which is found throughout the greater part of the continent, and which has been reported, although doubtfully, also from Palestine. Two species of crocodile, C. robustus and C. Madagascariensis, the one related to the common Indian form and the other to the African, are found on the island of Madagascar. The alligators (Alligator), also known as caymans and jacares, and comprising, according to some authors, not less than ten dis- tinct species, are confined to the waters of tropical and sub- tropical America, ranging from the Argentine Republic to Tennessee. The single species of the United States is the Alligator Mississippiensis. It is not a little surprising, seeing the presence there of crocodiles, that alligators should be almost wholly absent from the West In- dies; one species (A. latirostris) is said to inhabit the island of Guadeloupe. Geologically the crocodiles represent an ancient group, dating their first appearance, as far as is yet known, from the Triassic pe- riod. Three genera of this age are recognised : S'tagonolepis, from the Elgin sandstones of Scotland, Belodon, from Wurtemberg, the * Dr. Gray, in his " Catalogue of the Shield Reptiles of the British Mu- seum" (1872), makes seven crenera, of which Oopholis is Asiatic and Austra- lian, Bombifrons Asiatic, Palinia and Molinia American, and the remainder, Crocodilus, Halcrosia, and Mecistops, African. It is questionable whether any of these forms is entitled to generic distinction. CROCODILES. 329 Eastern United States, and India, and Parasuchus, from India. In the deposits of the succeeding Jurassic age the number of distinct types and species is very largely increased. No less than forty species, belonging in the main to the genera Mystriosaurus, Teleo- saurus, Steneosaurus, Metriorhynchus, and Dakosaurus, are known from British strata alone. m Many of these are also found in the deposits of the continent of Europe, which comprise a considerable number of additional types. The amphiccelous, or biconcave, type of vertebra, distinctive of the Triassic and Jurassic crocodilians, is retained in a measure by the Cretaceous forms, as in Goniopholis and the American Hyposaurus,* but we now also meet, and for the first time, with the type of the modern procoelian crocodile. Gavialis and Crocodilus, abundantly developed as Tertiary forms, both oc- cur in the Upper Cretaceous beds of Europe, and are represented in the nearly equivalent American deposits by the gavialine genera Holops and Thoracosaurus. Tomistoma is found in the Miocene of Malta and Lower Austria. Alligator does not appear before the Tertiary (Eocene) period (Europe and America). A gavialine form from the Siwalik deposits of India, Rhamphosuchus crassidens, is supposed to have attained a length of from fifty to sixty feet. The origin of the crocodilian line is involved in much obscurity. Whether or not the animals of this group stand in direct genetic relation with some of the earlier labyrinthodonts, as is maintained by some paleontologists, our present knowledge does not permit us to determine. Among themselves, however, the different croco- dilian types exhibit a remarkable gradational series of structural peculiarities, which connect the most ancient and the modern forms, and place them in an almost unbroken sequence. Professor Huxley has indicated the line of succession as passing from the Parasuchia — the Triassic forms, in which neither the palatine nor pterygoid bones enter into the formation of secondary posterior nares — through the Jurassic and Cretaceous Mesosuchia, in which the palatines alone are produced to form the;e nares, to the modern and Upper Cretaceous (procoelian) Eusuchia, in which both bones are similarly produced. M. Dollo recognises in Bernissartia, a recently discov- * Hyposaurus Ro^ersi, from the " grcensands " of the Eastern United States, was until recently the only known species of the genus; a second species, II. Dcrbianus, has been described by Professor Cope from the Province of Pernambuco, Brazil ("Trans. Am. Phil. Soc.," Jan., 1886). 330 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION". ered form from the Cretaceous deposits of Belgium, the ancestral type of the short nosed modern crocodilians — i. e., the crocodile and alligator. BIRDS. The principal features connected with the geographical distribu- tion of birds having been discussed in the early part of this work, only the geological distribution of the class will be considered here. The earliest known birds are the Archaeopteryx, whose remains have thus far been found only in the Solenhofen limestone (Up- per Oolite) of Bavaria, and the Laopteryx priscus, from a nearly equivalent horizon of the Western United States (Wyoming Terri- tory). The latter, which was of about the size of the great blue heron (Ardea Herodias), is apparently a member of the hetero- geneous group designated by Marsh the Odontornithes, or toothed- birds, to which the more remarkable of the American (Middle) Cretaceous birds, Ichthyornis, Hesperornis, and Apatornis, belong. Ornithic remains, with somewhat doubtful relationships, and re- ferred to the genera Graculavus, Laornis, Palasotringa, and Tel- matornis, have also been obtained from a somewhat higher horizon (Upper Cretaceous) in the Eastern United States (New Jersey). Almost the only clearly determined bird-remains of this period occurring in Europe are those of Enaliornis (Pelagornis ; Upper Greensand of Cambridge), which appears to have had some resem- blance to a penguin. In the Tertiary deposits remains of this class are very much more numerous, and there is a close approximation to modern type- structures. Thus, in the Eocene deposits of the Paris Basin and elsewhere in France (Auvergne, Provence, Langucdoc) we find the remains of the true quail (Coturnix), grouse (Tetrao), cormorant, godwit, rail, sandpiper, nuthatch, and falcon, associated with which are a number of forms whose relationships have not in all cases as yet been absolutely determined. The most remarkable of these is probably Gastornis Parisiensis, a bird of about 'the stature of the African ostrich, but possessing so many well-marked anatine char- acters as to have induced some naturalists to class it with the ducks and geese.* Agnopterus and Elornis appear to have repre- * Gastornis Klaasscni, a bird apparently exceeding the ostrich in size, has re:ently been described by Mr. Newton from the Lower Eocene strata of Croydon, England ("Proc. Geol. Assoc.," Feb., 1886). BIKDS. 331 sented the flamingoes, and Palaeocircus and Palseortyx, as is indi- cated in their names, the raptorial and gallinaceous birds respec- tively. It is not a little remarkable that Leptosomus, the type of a small family now absolutely restricted to the island of Madagascar, should constitute a part of this fauna. The deposits of the Swabian Alps have yielded a limited number of bird-remains (harrier, cor- morant), and so have those of Glarus, Switzerland, whence was obtained the nearly perfect skeleton of the passerine form known as Protornis or Osteornis. The equivalent, or nearly equivalent, deposits of the London Basin, the island of Sheppey, and of Hempstead, in the Isle of Wight, have also yielded a number of avian forms, some of which appear to have been most intimately related to types now living, such as the herons, gulls, and kingfishers (Halcyon or Halcyornis). But here, as in the Paris Basin, there occur several distinct types whose position among living forms it is very difficult or impossible to establish. Such are the Megalornis, a bird somewhat smaller in size than the emu ; Dasornis, which apparently combines true struthious characters with those of the recently exterminated moas of New Zealand; Macrornis, also with struthious characters; and the very singular anatine Odontopteryx toliapicus, recalling in its dental armature the Cretaceous toothed-birds of America. All the older Tertiary bird-remains that have thus far been described from the American continent are from the Western United States, and belong in principal part to the gruiform genus Aletornis (Wy- oming), some of whose species appear to have attained to nearly the stature of the sand-hill crane. A true owl (Bubo leptosteus), about two-thirds as large as the great horned-owl (B. Virginianus), represents the birds of prey, and the passerine Palseospiza bella the songsters; the former is from Wyoming (Eocene), and the latter from the insect-bearing shales of Florissant, Colorado (Oligocene?). A giant struthious bird, combining some of the characters of the extinct moas, has been described by Professor Cope from the Eocene deposits of New Mexico, as Diatryma gigantea, a form not un- likely generically identical with the European Gastornis. Ornithic remains are much more abundant in the Miocene de- posits than in the Eocene, and there is a corresponding further approximation to modern type structures. From the lacustrine deposits of Central and Southern France, whence the greatest 332 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. number of distinct types has been obtained, upwards of fifty spe- cies have been described, the greater number of which are referable to the modern genera Aquila (eagle), Haliaetus (fishing eagle), Milvus (kite), Bubo (owl), Picus (woodpecker), Corvus (crow), Motacilla (wagtail), Passer (sparrow), Columba (pigeon), Rallus (rail), Phoenicopterus (flamingo), Grus (crane), Ardea (heron), Ibis, Totanus (tattler), Numenius (curlew), Tringa (sandpiper), Lams (gull), Phalacrocorax (cormorant), Sula (ganuet), Pelecanus, and Anas (duck). The occurrence of a parrot (Psittacus) and of sev- eral species of pheasant (Phasianus; also in Greece) is rather re- markable, since the former is no longer an inhabitant of the European continent, or of any adjoining tract, and the latter is generally conceived to have been a modern introduction from Asia. Several of the generic types found in France have also been recognised in the deposits of South Germany (Steinheim, &c.). The Siwalik Hills formation of India has yielded the remains of two species of pelican, a cormorant, stork (Leptoptilus), merganser, ostrich (Struthio Asiaticus) and emu (Dromseus? Sivalensis). With reference to the occurrence in India of the last named bird, whose relationship with its living Australian congener, Dromaeus Novae- Hollandise, is very intimate, Mr. Lydekker says : ' ' The former occurrence in India of a large struthioid closely allied to the emu is one more instance of the originally wide distribution of the struthioid birds ; and it not improbably indicates that the home of the group of which the cassowaries, emus, and moas are diverging branches, was originally somewhere in the neighbourhood of the Indian region, whence a migration took place during some part of the Tertiary period towards the southeast, where the group, in regions more or less completely free from the larger mammals, sub- sequently attained its greatest development."123 The American Miocene birds are limited to some four or five species, a turkey (Meleagris antiquus ; Colorado), nearly as large as the common wild species (M. gallopavo), gannet, shearwater, and guillemot. The Pliocene and Post-Pliocene birds of the continent of Europe are much less numerous than the Miocene, and in the greater num- ber of cases do not admit of absolute determination. Several species of waders, swimmers, and gallinaceous birds (Gallus, Scolo- pax, Anas, Anser), more or less intimately related to existing forms, have been described from England, France, and Germany. The MONOTREMES. 333 mallard (Anas boschas) and grey lag-goose (Anser cinereus) both appear to be represented in the later deposits. Among the cave deposits of France have been discovered the remains of the snowy- owl (Nyctea Scandiaca) and willow-grouse (Lagopus albus), north- ern forms which appear to have followed the southward migration of the reindeer, and of a large extinct species of crane (Grus primi- genia). Two or more species of swan have been found in the os- siferous cavern of Zebbug, in the island of Malta, one of which, Cygnus Falconeri, an extinct form, exceeded by about one-third the dimensions of the common C. olor. Of the group of sub-fossil birds, or those whose remains belong to a comparatively very recent period, are the giant struthious birds of New Zealand, known as "moas" (Dinornis and Mionornis, with some seven or more species), and the palapteryxes (Palapteryx and Euryapteryx) ; the ^Epyornis maximus of Madagascar; and the Australian Dromaeornis australis, the precursor of the modern emu. A giant goose (Cnemiornis), associated with which are the remains of several remarkable ralline forms (Aptornis and Notornis — the latter surviving up to our own period), and a number of other birds, also occur in the newer deposits of New Zealand. In this connection may be mentioned, although not strictly falling under the category of fossils or sub-fossils, the recently exterminated didine birds of the Mascarene Islands — the dodo (Didus ineptus) of Mauritius, and the solitaire (Pezophaps solitarius) of Rodriguez ; the crested parrot of Mauritius (Lophopsittacus Mauritianus), and the Aphanapteryx, an abnormal ralline species, from the same isl- ands. MAMMALIA. Mo.'iotremata. — This, the most limited, order of terrestrial Mammalia, forming the sub-class Ornithodelphia of most natural- ists, comprises two families, the Ornithorhynchidae, or duck-bills, and Echidnidre, or Australian hedgehogs, the former of which is restricted to the continent of Australia and Tasmania, and the lat- ter to the same region with the addition of New Guinea. The duck-bills are represented by a single species, the platypus or water- mole of the colonists (Ornithorhynchus paradoxus or anatinus). No fossil remains referable to this genus have as yet been dis- covered. The Echidnidae comprise two recent genera : Echidna, with 334 GEOGRAPHICAL AKD GEOLOGICAL DISTRIBUTION". three or four more or less clearly defined species (E. hystrix or aculeata, E. cetosa, E. acanthion*) inhabiting Australia and Tas- mania, and a single one (E. Lawesii) New Guinea; and Acantho- glossus, represented by A. Bruijnii, from Northern New Guinea. Fossil remains of this family are not numerous, and belong exclu- sively to the Post-Pliocene deposits of the Australian continent. Echidna Oweni is founded upon a portion of a humerus from Darling Downs, and indicates an animal considerably larger than the com- mon recent species. E. Ramsayi, from a breccia cave in Wellington Valley, is likewise founded upon a humerus. Marsupialia, — All the existing members of this order, if we except the single family of American opossums (Didelphidse), are restricted to the Old World, and are in the main confined to the Australian continent and New Guinea, a limited number of forms finding a habitat in the debatable tract between the Australian and Oriental realms. The general features of their distribution are discussed in the chapter treating of the Australian realm. The order is not represented in either of the continents of Eurasia or Africa. The opossums comprise a considerable number of species, the majority of which are confined to South and Central America; two species, Didelphys Virginiana and D. Calif ornica, are found in the United States, the former, the common American species, rang- ing from the Gulf border to the State of New York. An aberrant web-footed form, the yapock (Chironectes), inhabits South and Central America. Marsupial remains in deposits older than the Tertiary are not abundant, and are in the main comprised in a number of genera whose exact relationships have not as yet been absolutely deter- mined. Indeed, it is not a little doubtful whether the earliest forms usually referred to this order — those from the Trias — actually belong here, or represent an even more primitive type of mammal. To this category of uncertain forms may be referred the Microlestcs antiquus, from the Keuper of Germany, M. Moorei and Hypsiprym- * Described by Collet* from North Queensland (" Forli. Selsk. Christiania," 1884). Lutken indicates the possible existence of a fourth Australian species (" Proc. Zool. Soc.," London, 1884, p. 150), while Dubois defines a supposed new species, named Proechidna villosissima, from New Guinea ("Bull. Mus. Belg.," in., p. 109). MARSUPIALS. 335 nopsis Rhaeticus, from the Rhsetic deposits of Somersetshire, Eng- land, and Dromatherium sylvestre, from the Chatham coal-fields of North Carolina. Of less doubtful affinity are Tritylodon, from the Triassic deposits of South Africa, and the numerous forms whose fragments have been obtained from the British Oolites and the island of Purbeck — Amphitherium, Phascolotherium, Stereogna- thus, Spalacotherium, Amblotherium, Stylodon, Triconodon, Tria- canthodon, Plagiaulax — and from the nearly equivalent deposits of the Western United States (Diplocynodon, Stylacodon, Tinodon, Triconodon, Dryolestes, Ctenacodon). Many, or most, of these forms appear to depart to a certain extent from the normal type of marsupial structure — approximating the Insectivora — hence, by some naturalists, as Professor Marsh, they are relegated to distinct groups— Pantotheria and Allotheria — supposed to have no living representatives.* The Marsupialia are not represented in the Creta- ceous deposits ; Meniscoessus, a form whose closest relationship appears to be with the Jurassic Stereognathus, occurs (in associa- tion with dinosaurian remains) in the Laramie formation of the Western United States, the position of which in the geological scale, as has already been intimated, is more properly with the Cainozoic than with the Mesozoic series. The Tertiary marsupial remains of the Northern Hemisphere belong principally to the earlier periods, beginning with the oldest Eocene ; in Europe they have not been recognised higher than the middle Miocene, and in North America, if we except the pygmy opossum (Didelphys pygmsea) frbm the Miocene of Chalk Bluffs, Colorado, no representative is known from deposits newer than the Oligocene (White River beds). Barring the opossums,! whose earliest remains have been found in the Eocene deposits of both * According to Professor Seeley, Ilvpsiprymnopsis, which appears to be most intimately related to the modern kangaroo-rat (Ilypsiprymnus >, is founded on the premolar toetli of Microlestes. The same authority recognises in Am- pliithcriurn and Phascolotherium a strong combination of marsupial and in- scctivore characters, and the indications of a "generalised insectivorous type, modified from a monotreme stock in the direction of the marsupial plan " (Phillips's "Manual of Geology," i., p. 520, 1*85). t Separated by some authors from the genus Didelphys as Peratherium and Amphiperatherium ; Peratlicrium, whose range extends into the Miocene, is represented by five or more species in the White River deposits of Colorado, the largest of which about equals in size the mole (Scalops aquaticus). 336 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION". France and England, none of the recent families are indicated. The herbivorous type seems to be entirely wanting in the European deposits, but in America, the forms that have been described from the Puerco formation of New Mexico as Polymastodon are re- ferred to this type by Professor Cope. Neoplagiaulax, from the basal Eocene beds of France and New Mexico, and its near ally Ptilodus, represent the Jurassic Plagiaulacidse, and appear to effect a partial transition from these to the Pliocene or Post-Pliocene Thylacoleo of Australia. The remains of true kangaroos (Macropodidre), some of them, as Palorchestes, considerably exceeding in size the largest of the modern representatives of the family, occur in the newer Pliocene or Post-Pliocene deposits of the Australian continent. Associated with these are a number of remarkable forms whose precise affini- ties still remain to be determined, although in the general character of their dentition they approximate the kangaroos and phalangers. Diprotodon australis, with less disproportionate limbs than in the kangaroos, appears to have exceeded the rhinoceros in size. Of somewhat smaller dimensions are the species of Nototherium. Thylacoleo carnifex, described as "one of the fellest and most destructive of predatory beasts," is held by many naturalists to have been an herbivore. Edentata, — The animals of this order are at the present day confined almost wholly to the southern continents — indeed, it might be said principally to the continent of South America (with Central America), which possesses more than three-fourths of all the known species. Of the five recognised families, the sloths (Bradypodidse), ant-eaters (Myrmecophagidae), and armadillos (DasypodidaB), are exclusively American ; the aard-varks (with two or three species — Orycteropus Capensis, the Cape ant-eater, O. ^Ethiopicus, from Northeast Africa, and a possible third species from Senegal) are African; and the scaly ant-eaters or pangolins (Manidida3), both African and Asiatic. The species of the last, some eight or more, are properly referable to a single genus, Manis, although several sections, by some authors considered to be of generic value, have been constituted to receive certain well-marked, but unimportant, peculiarities of structure. The common pangolin (Manis penta- dactyla) inhabits the Indian peninsula and the island of Ceylon, sharing in part the distributional area of the Chinese species (M. ANT-EATERS, ARMADILLOS. 337 aurita), whose range extends from North India to the island of Formosa. The Javan pangolin is a native of Burmah, the Malay Peninsula, and the larger islands — Java, Borneo, Sumatra — of tlie Eastern Archipelago. A limited number of species are known from Western Africa, one of which, M. (Pholidotus) gigantea, measures about five feet in length to the tip of the tail. The most aberrant form of the family is M. (Smutsia) Temminckii, from the southern and eastern portions of the African continent. Of the American groups the most restricted in point of numbers are the ant-eaters, whose range, collectively, embraces the greater portion of the Neotropical realm included between Mexico and Paraguay, east of the Cordilleras. The better known species — in- deed, the only species admitted by most authors — are the great ant- eater (Myrmecophaga jubata), the tamandua (Tamandua tetradac- tyla), and the little or two-toed ant-eater (Cycloturus didactylus), whose individual ranges coincide largely with the range of the en- tire family. The sloths, of which some authors recognise not less than a dozen fairly well-marked species or varieties, occupy much the same area as the ant-eaters, although they do not appear to enter Paraguay. They are inhabitants of the forest region, which limits their distribution. Two genera, founded upon the number of toes on the fore -feet, are generally admitted : Bradypus, the three-toed sloths, and Cholcepus, two-toed sloths, both of which are very extensively distributed.* The armadillos (Dasypodidas), which comprise nearly twenty clearly defined species, are the most broadly distributed of the American edentates, their range extending from the most northern limits of the Neotropical realm to the fiftieth parallel in Patagonia. A single species, the peba or seven-banded armadillo (Tatusia septemcincta), which is found in South America as far south as Paraguay, enters the United States in Texas. Among the better known species are the six-banded armadillo or encoubert (Dasypus sexcinctus), an inhabitant of Brazil and Paraguay; the tatouay or cabassou (Xenurus unicinctus), with much the same range as the last, but extending into Guiana ; the three-banded armadillo or * Arctopithocus appears to have no distinctive generic characters. A specimen of Bradypus tridactylus in the museum of the Royal College of Surgeons, of London, corresponds, according to Professor Flower, with Gray's Arctopithecus gularis. 338 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. apar (Tolypeutes tricinctus), which, with the remaining members of the generic group to which it belongs, has the power of rolling itself into a complete ball ; and the great armadillo (Priodon gigas), an inhabitant of the forests of Brazil and Guiana, the largest living representative of the family, measuring upwards of three feet from the tip of the nose to the root of the tail. Fossil remains of edentate animals are not numerous, and are in the main confined to the Pliocene and Post-Pliocene deposits of the New World, especially South America (Pampean formation of the Argentine Republic ; bone-caves of Brazil). The oldest known form is Ancylotherium priscum, from the phosphorites of Quercy, Fiance (Oligocene), a generalised type of animal, considered by some au- thors to stand intermediate between the Edentata and Ungulata; the same genus (A. Pentelici) occurs in the Miocene deposits of Pikermi, Greece. An apparently allied form, Macrotherium, whose remains indicate a possible climber of gigantic proportions, with compara- tively feebly developed hinder extremities, is represented by several species in the Miocene deposits of both France and Germany. No New World forms are known to antedate the middle Tertiary period. Moropus, from the Miocene and Pliocene deposits of the Western United States, comprises animals ranging in size between the tapir and rhinoceros, but with uncertain affinities; equally uncertain is the position to be assigned to the Pliocene Morotherium, whose re- mains have been found at various localities in Idaho and California. The South American edentate fauna (Pliocene and Post-Pliocene) comprises, according to Gervais and Ameghino, some eighty or more species, the greater number of which belong to genera now no longer living. The better known forms are referable to the families Megatheriida? and Glyptodontidae ©r Hoplophoridse, the former of which appear to hold an intermediate position between the modern sloths and ant-eaters— combining the head and denti- tion of the one with the trunk and appendages of the other — while the latter, in the presence of a carapace, approach the armadillos. Included in the family Megatheriidae, besides other forms, are the genera Megatherium, writh animals of the size of the rhinoceros,* * Megatherium Amerieanum, from the Argentine Republic and Paraguay, was only inferior in size to the elephant, far surpassing all other land animals. A mounted skeleton measures eighteen feet in length from the fore part of the head to the tip of the tail. SEA-COWS. 339 Ccelodon, Mylodon, Lestodon, Scelidotherium, Plationyx, and Me- galonyx, the majority of which embrace species of very robust dimensions. The species are mainly found in the bone-caves of Brazil, and in the Pampean and diluvial deposits of the Argentine Republic and Patagonia. Megalochnus rodens, a diminutive spe- cies, is from the island of Cuba. Nothropus priscus, from the Argentine Republic, appears to have possessed arboreal habits, in this respect agreeing more closely with the modern sloths than any of the other forms. Of the North American members of this family the best known species are Megatherium mirabile, a some- what smaller form than the M. Americanum, from the superficial deposits of the Southern United States ; Megalonyx Jeffersoni, originally described from a cave in Virginia ; and Mylodon Harlani, from the Western and Southern United States. A peculiar genus from the deposits of Natchez, on the Mississippi, has been described as Ereptodon. The glyptodons embrace a considerable number of Pampean species, which by Burmeister and other authors are referred to sev- eral distinct genera — Hoplophorus, Panochthus, Doedicurus, Eury- urus, Glyptodon, and Schistopleurum. The best known species are Glyptodon typus and clavipes. Sirenia (Sea-cows). — This order is at the present day limited to some half-dozen species, referable to two genera : Manatus, the ma- natees, and Halicore, the dugongs, the former of which is common to both the Eastern and Western Hemispheres, while the latter is strictly confined to the Old World. Of the two American species of Manatus the West Indian sea-cow (Manatus latirostris) inhabits the creeks, lagoons, and estuaries of the north of South America, the West Indies, and Florida, and the Brazilian sea-cow (M. Americanus or inunguis) the South American coast-line to about the twentieth parallel of south latitude, and the more important Brazilian rivers, very nearly to their sources.* The only Old World form (M. Senegalensis) inhabits the West African coast for about ten degrees on either side of the Equator, and the interior as far as, or farther than, Lake Tchad. Of the three species of Halicore, one (Halicore tabernaculi) is * The identity of the coast species and that of the Upper Amazon and Ori- noco Rivers has not yet been absolutely established, but is considered highly probable by Hartlaub (Spcngel's "Zool. Jahrb.," 1886). 340 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. restricted to the East African coast and the Red Sea, another (Hali- core dugong) inhabits the Indian and Pacific oceans, eastward from the home of the last to the Philippines, and the third (Halicore australis) the waters of Eastern and Northern Australia. Fossil remains of sirenians, although not very numerous, occur throughout all the Tertiary formations, from the Eocene to the Pliocene, inclusive. The earliest forms are the Eotherium ^Egyptia- cum and Manatus Coulombi, from the Mokattam limestone of Egypt, Hemicaulodon effodiens, from the basal Tertiary beds of Shark River, New Jersey, and the somewhat doubtful Halitherium du- bium, from the deposits of the Gironde, France. In the last-named genus are included a considerable number of species from the Mio- cene deposits of Germany, France, Belgium, and Italy, and a single undetermined (?) form from the Isthmus of Suez. Felsinotherium and Chirotherium are Pliocene forms from Central and Northern Italy, and Rhytiodus Capgrandi a species from the nearly equivalent deposits of the Garonne, France. The American fossil sirenians comprise, in addition to the Eocene Hemicaulodon, two or more species from the Miocene deposits of South Carolina, Manatus an- tiquus, M. inornatus, and Dioplotherium Manigaulti ; the first also occurs in New Jersey and Virginia. Of the Post -Pliocene forms the best known is the Rhytina gigas or Stelleri, " Steller's sea- cow," an animal which appears to have been fairly abundant about Behring and Copper Islands as late as the second half of the last century, but which is now apparently entirely extinct. The im- bedded remains occur principally in the raised beaches and peat- mosses of Behring Island. Woodward calls attention to the significant fact that, if we " take the belt of the tropics, that is, 23£° N. and 23 J° S. of the Equator (or, better still, say 30° N. and S. of the Equator), we shall cover the geographical distribution of all the living sirenians. If we take another belt of 30° north beyond the Tropic of Cancer, we shall embrace the whole geographical area in which fossil remains of sirenians have been met with. Assuming, as I think we may, that the Sirenia at the present day belong exclusively to the tropi- cal regions of the earth, and that Rhytina, in its boreal home, was simply a surviving relic from the past (a sort of geological ' out- lier,' as of a stratum elsewhere entirely denuded away), we must conclude that the presence of about twelve genera and twenty- WHALES. 341 seven species of fossil Sirenia, as widely distributed then as the recent forms are at the present day, but with a range from the Tropic of Cancer up to CO0 of north latitude, affords a most valua- ble piece of evidence (if such were needed) attesting the former northern extension of subtropical conditions of climate which must have prevailed over Europe, Asia, and North America, in Eocene and Miocene times, and in the older Pliocene also."123 Cetacea (Whales, &c.).— The animals of this order are distrib- uted throughout almost the entire oceanic expanse, and a limited number of forms, the members of the family Platanistidae, and some delphinoids, are also found in fresh or estuarine waters. Platanista Gangetica, an inhabitant of the waters of Northern India — Ganges, Brahmaputra, and Indus, and their tributaries— is en- tirely fluviatile, never being known to pass out to sea The only absolutely fluviatile form occurring in America is Inia Geoffrensis, from the Upper Amazonian water system ; Pontoporia Blainvillii inhabits the estuary of the Rio de la Plata, but is not positively known to ascend that stream into fresh water. Of the marine cetaceans two distinct types are usually recog- nised by naturalists : the whalebone or toothless whales (Mystaco- ceti), as represented by the right-whales (Balacna), rorquals, or fin- whales (Balaenoptera), and the humpbacks (Megaptera). and the toothed-whales (Odontoceti), which comprise the sperm-whales, dolphin, porpoise, grampus, &c. The right-whales, which are confined principally to the northern and southern seas, have been divided into some half-dozen species, which, however, so closely resemble one another that not improbably they represent only vari- etal forms of one and the same species. The best known is the Greenland right-whale (Balsena mysticetus) of the Arctic seas ; other .northern forms are B. Biscayensis and B. Japonica. The southern species or varieties are B. australis, of the South Atlantic, and B. antipodarum and B. Novse-Zelandiaa, of the South Pacific. A nearly equal uncertainty attaches to the different varieties or species of rorquals and humpbacks, especially the latter, some forms of which are found in almost every sea. Four or more ap- parently distinct types of rorqual inhabit the northern seas, but whether these are absolutely separable from their antipodal con- geners still remains to be determined. Much confusion exists as to the synonymy of the species ; hence, the great difficulty of their iden- 342 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. tification. The members of this genus are probably the largest of all living animals, some of the forms, as Balamoptera Sibbaldii and B. sulfurea, attaining a length of eighty or perhaps even a hun- dred feet. The smallest of the known species of whalebone whales is the rare Neobalsena marginata, from the Australian and New Zealand seas, which attains a greatest length of about twenty feet. Of the toothed-whales, other than the members of the family Delphinidfe (dolphins, porpoises, &c.), the best known and prob- ably most widely distributed species is the cachalot or sperm-whale (Physeter macrocephalus), a giant form measuring upwards of sixty feet in length, whose habitat is more properly the tropical and sub- tropical seas, the animal but rarely appearing in the polar waters. More or less closely related forms of the same family (Physeterida?) are Kogia, Ziphius, and Mesoplodon, the species in each group of which have either individually or collectively a very broad exten- sion.* Two species of bottle-nose whale (Hyperoodon rostratus and H. latifrons) inhabit the North Atlantic. Of the delphinoid type of cetaceans the most numerously repre- sented genus is Delphinus, the dolphin, or, as it is frequently mis- called, porpoise, the numerous species of which are distributed throughout most seas, a limited number even habitually ascending some of the larger streams, as the Amazon. The type-form of the genus is the common or Mediterranean dolphin, the hieros ichthys or sacred fish of the ancients (D. delphis), which is also abundant in the Atlantic, and of which closely allied, if not identical, forms are found in the Australian seas (D. Forsteri) and in the North Pacific (D. Bairdii). ™* One of the most northerly species of dolphin is the tursio, or nesarnak of the Greenlanders (D. tursio), which inhabits the Atlantic between Greenland and' the European shores. Modifications of the ordinary delphinoid type are seen in the long- beaked forms of the group Steno, and in a South Sea species, Leu- corhamphus (Delphinapterus) Peronii, in which there is no dorsal fin. The bottle-heads (Globicephalus) are inhabitants of nearly all * Much diversity of opinion exists as to the number of species belonging to the different genera. Thus, while Gray recognised 'not less than six species of Ko^'ia, founded upon about as many individual specimens, only one (Kogia breviceps, found in the South and North Pacific oceans; is admitted by Flower (" Encycl. Brit.," 9th ed., xv., p. 396). The same authority likewise considers the species of the other genera as being in great part founded upon insufficient characters. DOLPHINS, PORPOISES. 343 seas, especially the north and south temperate, and exhibit a marked specific identity between the most widely removed forms (Australia and North Atlantic). The type of the genus is the pilot-whale (G. melas ; Delphinus globiceps), the grindhval of the Faroe-Islanders, whose distribution is practically coextensive with the northern seas. Related to the preceding are the so-called grampuses of the genus Grampus, of which only one species (G. griseus), remarkable for the variability of its colouring, has been thus far clearly determined ; it inhabits the northern ocean, more rarely descending into the Medi- terranean. A second form, from the Cape of Good Hope, has been described as G. Richardsoni. The true grampuses, also known as " killers," from their preda- cious habits, constitute the genus Orca, and are more nearly related to the true porpoises than to the dolphins proper. They are dis- tributed over the greater portion of the oceanic expanse, from Green- land to Tasmania ; but neither the relationships of the different so- called species, nor the limitations of their respective habitats, have as yet been determined. Orca gladiator, the common grampus, is more properly a northern form, and is fairly abundant in the polar seas. Pseudorca crassidens, a much rarer form of grampus, found on the Danish coast, appears to be identical with a species from the Australian waters. The genus Orcella is represented by two species, one of which (O. brevirostris) inhabits the Bay of Bengal, and the other (O. fluminalis), a fluviatile form, the Irrawaddy River, at a distance of several hundreds of miles from its mouth. Of the true porpoises of the genus Phocsena, whose limited species are confined principally to the waters of the Northern Hemisphere,* the best known and probably most widely distributed form is P. communis, the common porpoise, which inhabits in shoals or schools the North Atlantic, from Britain to Greenland and the American coast, frequently ascending the outflowing streams for a considerable distance above their mouths. They have been observed on the Thames, at London, and appear to have occasionally penetrated up the Seine as far as Paris. The animal does not seem to enter the Mediterranean. By some natu- ralists the common porpoise of the Atlantic coast of the United States is considered to be a distinct species, to which the name P. * Phocaena spinipennis has been described from the mouth of the Rio de la Plata. 344 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. Americana has been applied. A peculiar form of porpoise, from the Indian Ocean (?) and the Japanese coast, destitute of the dorsal fin, has been described as Neomeris phocaeniformis. Of the remaining delphinoids the narwhal, or sea-unicorn (Mo- nodon monoceros), inhabits the Arctic Ocean, rarely passing south of the sixty-fifth parallel ; the beluga, or white whale (Delphinap- terus leucas), closely related to the last, is also an inhabitant of the Arctic Ocean, descending on the American coast to the St. Lawrence River, and more rarely, in European waters, to the shores of Scot- land. Excepting the Palaeocetus Sedgwicki, from the boulder clay of Roswell Pit, Ely, England, whose remains were encased in a matrix supposed to be Upper Jurassic (Kimmeridgian), the earliest ceta- ceans of whose organisation we know anything are the zeuglodons, which apparently represent a type intermediate between the toothed and toothless forms of the present day. They occur in the Upper Eocene deposits (Jacksonian) of the Southern United States; one species (Zeuglodon Wanklyni) has also been discovered in the equivalent Barton sands of England, and two, corresponding to the American forms — Z. macrospondylus and Z. brachyspondylus — in the deposits (Eocene or Oligocene) of Birket-el-Keroun, Egypt.* Closely related in dental characters to the zeuglodons, but differing in well-marked cranial features, are the squalodons, whose remains are abundantly scattered throughout the Miocene and Pliocene de- posits of many parts of continental and insular Europe (Vienna Basin, France, Antwerp and Suffolk Crags, &c.). They have also been noted from nearly contemporaneous strata in North America (Squalodon Atlanticus, from Shiloh, New Jersey t) and Australia. The oldest known form of modern-type cetacean is Balaenoptera Juddi, from the Oligocene (Headon) beds of the Hampshire basin, England ; no other representative of an existing genus has thus far been found to antedate the Miocene period. Bo.th the toothed and * Dames suggests that the two forms of Miiller may only represent the male and female of a single species, which would then be the Zeuglodon cetoides of Owen (" Sitzungsb. Berl. Ak.," 1883). t The forms described from the Eocene deposits of the Ashley River, South Carolina, are considered by Van Beneden an.d Gervais to be only doubtfully referable to this genus. INSECTIVORA. 345 toothless whales are represented in the deposits of this age (Mio- cene), the latter, however, apparently only by the rorquals or their immediate allies (Balsenopterce) ; Cetotherium seems to have occu- pied a position intermediate between the Mystacoceti and Odonto- ceti. Of the latter the earliest representatives appear to have been the genera Ziphius and Mesoplodon, although not improbably the true dolphin was an immediate contemporary. The baleen whales proper are not known before the Pliocene, when, in addition to Balrana, and possibly Neobal&na, we meet with a number of extinct types more or less closely related to these — Bala3nula, Idiocetus, Plesiocetus. Balrenotus, from the Antwerp Crag, is a connecting form between the baleen whales and the rorquals, and Bartinop- sis between the rorquals and humpbacks. Cetacean remains are abundant in the Post-Pliocene deposits, and comprise a variety of recent types; the narwhal has been indicated from the deposits of England and Siberia. The Miocene deposits of the Eastern United States have yielded a number of delphinoid remains to which the generic names Priscodelphinus, Tretosphys, Zarachis, Lophocetus, Rhabdosteus, and Ixacanthus have been applied. Insect ivora. — The animals of this order, which comprises barely more than one hundred and thirty to one hundred and forty living species, of which about one-half are true shrews (Soricidae), are distributed over the greater part of the earth's surface, but are absent from both South America and Australia. The greater num- ber of the nine generally recognised families are limited to com- paratively narrowly circumscribed distributional areas, which in some cases are only co-extensive with the sub-regions or provinces of the main zoogeographical divisions. The GaleopithecidaB, or flying-lemurs, which were formerly referred to the true lemurs, and by some naturalists to the bats, are the most aberrant forms of the order, and constitute the type of a distinct sub-order, Insecti- vora dermoptera. But two species are known, Galeopithecus volans and G. Philippinensis — the former an inhabitant of the forests of the Malay Peninsula, Sumatra, and Borneo, and the latter of the similar districts of the Philippine Islands. The squirrel- or tree- shrews (Tupaiida3), small arboreal insectivores resembling squirrels in outline and habits, are restricted to the Oriental region, where they range from the Khasia Hills, in India, to Java and Borneo. Of the two genera, Tupaia and Ptiloccrcus, the latter contains but 346 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. a single species, the very interesting Bornean pentail (P. Lowii), remarkable for its long quill-shaped tail. — The elephant-shrews (MacroscelidaB), small leaping animals furnished with a trunk-like snout, are confined principally to South Africa; a single species of Macroscelides, to which genus the greater number of species belong, is found north of the Atlas Mountains. Several fossil forms, appar- ently referable to genera belonging to this family (Oxygomphus, Parasorex, Echinogale), have been described from the Miocene de- posits of France and Germany. Likewise restricted to the African continent are the potamogales and golden-moles (Chrysochloridffi) — the former, represented by a single species (Potamogale velox), inhabiting the territory about the Gaboon River, and the latter the region south of the Equator, but more particularly the Cape of Good Hope districts. Neither family is known by fossil repre- sentatives. The Centetidae occupy two widely separated regions of the earth's surface, namely, Madagascar, with the adjoining isl- ands (possibly introduced into Mauritius and Reunion) and the two larger islands of the Antilles, Cuba and Hayti. Most of the forms, with the genus Centetes itself, the Madagascar hedge-hog, occupy the former region ; the American species (two) belong to the genus Solenodon, which in certain anatomical points differs so essentially from its nearest allies as to constitute in the opinions of some sys- tematists 126 the type of a distinct family, Solenodontidae. Of the three remaining families of insectivores, the shrews (Soricidas), moles (Talpidae), and hedge-hogs (Erinaceidae), the first, which, as has already been stated, comprise about one-half of all the species of the order, have the broadest distribution, embracing, in fact, the entire tract covered by the Insectivora gen- erally. By some naturalists but a single genus (Sorex), apart from the very remarkable web-footed Nectogale from Thibet, is recog- nised, which is differentiated into a number of more or less well- marked sub-genera, founded upon the number and colour of the teeth, and the bristles on the tail — Crocidura (Old World), Sorex (the entire range), Blarina (Canada to Costa Rica), Neosorex, and Crossopus, the last two amphibious forms of the New and the Old World respectively. Several species referable to the genus Sorex (and Crocidura) have been described from the Miocene formations of France, and similar remains occur in the Quaternary cavern de- posits and breccias of both Europe and Asia. One or two extinct HEDGE-HOGS, MOLES. 347 genera or sub-genera, Mysarachne and Plesiosorex,* have also been indicated (Miocene). The hedge-hogs comprise two genera, Gymnura and Erinaceus (the hedge-hog proper) — the former of which inhabits the Malay Peninsula and some of the islands of the Indian Archipelago, and the latter, with about nineteen species, the greater part of Europe, Asia, and Africa, although wanting in Madagascar, Ceylon, Bur- mah, Siam, the Malay Peninsula, and the Archipelago. The range of the common European species, Erinaceus Europseus, extends from Ireland and the Shetland Isles (possibly introduced) to East- ern China, and from the sixty-third parallel of latitude in the Scandinavian Peninsula to Southern Italy, Asia Minor, and Syria, ascending the Alps and Caucasus to elevations of 6,000 and 8,000 feet respectively. In view of this very remarkable range, its absence from the New World is not a little surprising. Re- mains of several species of hedge-hog have been found in the Miocene deposits of France and Germany, but none of the recent species, except E. Europaeus, which forms part of the Quaternary cave fauna, are known as fossils. The genus Neurogymnurus, which is probably closely related to Gymnura, is represented by a single species (G. Cayluxi) in the French Eocene. The moles are generically the most numerous of the Insectivora, although the number of species is comparatively limited. The greater number of the ten or twelve recognised genera are repre- sented by but one or two species. The family belongs almost ex- clusively to the Holarctic region, through which it is very generally distributed, only a very limited number of species passing beyond the confines of that region into the Oriental tract. The moles proper (Talpa), with about four species, are found throughout nearly the whole of the Eurasiatic tract, the range of the common species, Talpa Europaea, extending from Britain to Japan, and from Scandinavia and Siberia to Italy and the southern slopes of the Himalayas. The water-moles (Myogale) are comprised in two spe- cies, one of which, M. Pyrenaica, inhabits the northern valleys of the Pyrenees, and the other, M. Muscovitica, the region of the Don and Volga rivers. The American moles belong in principal part to three genera: Condylura (the star-nosed mole, which in- * Eeferred by Trouessart to the Talpidae (" Catalogue Mamra. Viv. et Foss.," 1881). 348 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. habits the Atlantic slope from Nova Scotia to South Carolina, and westward to Oregon), Scapanus, and Scalops, the last (Scalops aquaticus) the common American form, whose range covers the greater part of the North American continent east of the Rocky Mountains. A somewhat aberrant form, Neurotrichus Gibbsii, found in the Western United States (Cascade Mountains to Texas), has its analogue in the Urotrichus of Japan, to which genus it has generally been referred, and from which it differs mainly in the dental formula. — Fossil remains of Talpidse date back to the Eo- cene period (Prototalpa, Quercy, France ; Talpavus, Wyoming), but the genus Talpa itself is not known prior to the Miocene ; the common European species is found in the Quaternary deposits. Myogale (with Palseospalax and Galeospalax) occurs in the Miocene and Pliocene deposits of France and England respectively. None of the existing American genera have as yet been found in a fossil state. Other insectivorous forms, however, known principally in a fragmentary condition, and not impossibly referable, at least in part, to the type of insectivorous Marsupialia, have been described from the Eocene of Wyoming and New Mexico (Passalacodon, Centetodon, Entomodon, Entomacodon, Triacodon, Esthonyx), and, together with a number of Miocene forms from Dakota and Colo- rado (Leptictis, Isacis, Ictops), constitute a distinct family, Lep- tictidae. Insectivorous Forms of Doubtful Position.— Numerous insec- tivorous animals, known largely by portions of their dental armature alone, are found in the older (principally Eocene) Tertiary deposits of France and the western territory of the United States (Wyoming, New Mexico). They are not improbably, as Professor Cope sug- gests, referable in part to the lemurs, although from their imperfect state of preservation it is in many or most cases impossible to de- termine their true relationship, whether with the class of animals just mentioned or with the true insectivores. By Trouessart they are all ranged with the Insectivora as the group of the protolemurs. Among the better known of these forms are Microsyops, Pataco- don, Hyopsodus, Sarcolemur, Tomitherium, Notharctus, Necro- lemur (supposed by Filhol to have its nearest ally in the galago of Africa), Adapis (Pala3olemur), and Protoadapis, the last three from the Eocene of France, and representing distinctively lemuroid types. Galerix is from the Miocene deposits of the same country. In the BATS. 349 North American lower Eocene genus Anaptomorphus (A. semulus and A. homunculus), which comprises animals of about the size of the ground-squirrel (Tamias), and whose dentition approximates that of the higher apes and man, Professor Cope recognises the most simian type of lemur yet discovered, and believes that it "represents the family from which the anthropoid monkeys and men were derived. Its discovery is an important addition to our knowledge of the phylogeny of man." 12e Cheiroptera (Bats). — Bats are practically of world- wide dis- tribution, being found almost everywhere over the continental tracts where there is a sufficient supply of insect food. The number of species is very much greater in the region of the tropics than in the temperate zone — probably three times as great— the specific and individual diminution corresponding to a marked elevation of lati- tude being very rapid. Vesperugo noctivagans alone among the American bats appears to reach the fifty-fifth parallel, but in Eu- rasia one or more forms (Vesperugo borealis) penetrate to the Arctic circle. Most of the species inhabiting the region of ele- vated mountain-chains do not seem to ascend to any very great altitude, preferring to remain in the basal zone of from 4,000 to 6,000 feet; a few instances are noted of habitation at nearly twice this height. Vesperugo montanus and Molossus rufus have been observed on the Peruvian Andes (Huasampilla) at an elevation of 9,000 feet; Vespertilio muricola on the Himalayas at 8,000 feet; and Vespertilio oxyotus on the slopes of Chimborazo at nearly 10,000 feet. Vesperugo maurus is in Europe found chiefly in the region of the higher Alps. Bats, differing from all other mammals, are found in most of the oceanic islands, but none have so far been observed in either Iceland, St. Helena, the Galapagos, Kerguelen Land, or the islands of the Low Archipelago.127 Three species inhabit the Bermudas — Vesperugo noctivagans, Atalapha cinerea, and Trachyops cirrhosus (the last a vampyre *) — but only a single one the Azores — Vesperugo Leisleri. It would thus appear that the members of this order of animals were specially endowed with the power of crossing broad arms of the sea, standing, in this respect, next to the birds. It is more than probable, however, that, in the case of the species in- * An individual of Molossus rufus, var. obscurus, is catalogued by Dob- son as coming from Bermuda. 350 GEOGRAPHICAL A^D GEOLOGICAL DISTRIBUTION. habiting the last two named island groups, their appearance there is an accidental circumstance, depending upon the prevalence of certain storm-winds by means of which an unlooked-for transport has been effected ; and even in the case of the species found in the Pacific islands, it is not exactly unlikely that the broad distribu- tion has been brought about in a manner not necessarily indicating sustained flight, or at a period when the physical configuration of that portion of the earth's surface — the relation of land to water — was different from what it is at the present day. For, as Mr. Dobson suggestively points out, if the ' l Cheiroptera possess great powers of dispersal, it is certain that quite nine-tenths of the spe- cies avail themselves of them in a very limited degree indeed, and it is significant that the distribution of the species is limited by barriers similar to those which govern it in the case of other spe- cies of mammals."* Thus, it is shown that out of a total of some- what more than four hundred recognised species only about twenty- five pass beyond the confines of the regions to which they properly belong — i. e., about ninety -five per cent, are characteristic regional forms ; and of this number more than two-thirds belong to the pre- eminently wandering family Vespertilionidse, which has by far the broadest geographical distribution of any family of the order. If, however, it be urged that this restriction is not so much due to any inability on the part of the animals to migrate, but to considerations connected with altered conditions of food and climate — the influence of which must be very marked — we have the still more salient fact presented that of the numerous flying-foxes (Pteropus) which in- habit Madagascar and the Comoro Islands, not a single species is found on the east coast of Africa, the narrow channel of one hun- dred and eighty to two hundred miles which intervenes between the continent and Great Comoro Island seemingly being sufficient to form an effectual barrier to a westward migration. Still, in this special instance we are, perhaps, not presented with a just criterion of the actual powers of dispersion possessed by this class of ani- mals, since the slow and laboured flight of the large flying-foxes can scarcely be compared with that of the smaller insectivorous species ; indeed, there is a striking similarity between the insectivo- rous bat fauna of Africa and Madagascar. Instances of very broad specific distribution among the Cheirop- * "Kept. Brit. Assoc.," 1878. BATS. 351 tera are numerous, and perhaps most notably so in the case of the genus Vesperugo, of the family Vespertilionidae. Yesperugo noc- tula, the noctule, is distributed throughout the greater part of the Old World, from England to Japan, and from the Scandinavian peninsula to Southern Africa ; it extends through India to Ceylon and the islands of the Malay Archipelago. Vesperugo abramus, whose home is primarily the Oriental region, extending from Japan to Northern Australia, is found during the summer months through- out Middle Europe, and even as far north as Sweden ; the species furnishes us with a remarkable example of a true migrant. The range of V. maurus extends from the Canary Islands (Palma, Tene- riife) through Central Europe (Switzerland, the Tyrol) to China and Java, and that of Miniopterus Schreibersii from Southern Europe (Spain, Italy) to Japan and the Philippine Islands, and throughout the whole of Africa (with Madagascar) eastward to Australia. The last is probably the most widely distributed of all known species of bats, with the exception of the little serotine (Vesperugo sero- tinus), whose distributional area covers nearly the whole of Eurasia, Northern and Central Africa, and, in the New World, the American continent from Lake Winnipeg to Central America, and the West Indian islands. This is the only species that has been thus far positively identified as being common to both the Eastern and Western Hemispheres.* Of the strictly American species Atalapha Noveboracensis, in its several varietal forms, appears to be the most widely distributed, ranging from the Aleutian Islands to Chili. Of the six families into which the Cheiroptera have been divided only two, the Vespertilionida3 and the Emballonuridae, are common to both the Eastern and Western Hemispheres; the former com- prise some one hundred and sixty or more species, fully three- quarters of which are confined to the Old World, over which they are very extensively distributed. This is the most broadly dis- tributed of all the families, and is that which has the most north- erly range. Of its sixteen or more genera, at least five of which — Antrozous, Nycticejus, Atalapha, Natalus, and Thryoptera — are peculiar to America, only two, Vesperugo and Vespertilio, the * Vesperugo abramus is thought by Dobson to be possibly identical with a species (Scotopliilus hesperus of Allen) from Vancouver's Island ("Cat. Cheir. Brit. Mus.," p. 229). 352 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. former with about fifty species, and the latter with about forty, approach cosmopolitanism. Of the Emballonuriciae but a single genus, Nyctinomus, is common to both hemispheres, and its early differentiation is shown in the fact that, while all the American forms are closely related to one another, they depart widely from their European representatives. The Phyllostomida3, or simple leaf-nosed bats, which comprise the vampyres, number upwards of fifty species, all of them very closely related, and, with one exception — Trachyops cirrhosus, which has been noted from the Bermudas (and also doubtfully recorded from South Carolina) — confined to the Neotropical realm, over the forest-covered tracts of which they range from Mexico to about the thirtieth parallel of south latitude.* Vampyrus spectrum, the best known species of vampyre — whose habits appear to be mainly frugivorous — and the largest of all the American bats, is distributed over the greater portion of the tract covered by the entire family. Of the strictly Old World families of bats the Pteropodidse, fruit-eating bats or flying-foxes, are specifically the most numerous, comprising about seventy species, distributed between the Aus- tralian, Oriental, and Ethiopian realms, and some of the intervening tracts, with a preponderance of species in the first-named region. They are restricted almost wholly to the region of the tropics, where a continuous supply of tree-fruits might be obtained; no species has thus far been noted from either New Zealand or Tas- mania. Cynonycteris, alone of the genera, has the distribution of the entire family. The most largely represented genus is Pteropus, which includes more than one-half of all the recognised species be- longing to the family; its range extends from the Comoro Islands on the west to the Navigators' Islands, in the Pacific, on the east, and through much the greater portion of tne Oriental and Aus- tralian regions ; but few of the island groups of the Pacific — Sandwich Islands, Low Archipelago, Gilbert's and Ellice's groups — are deficient in the members of this genus, to which the largest known forms of bats belong. Pteropus edulis, which inhabits the islands of the Malay Archipelago, measures five feet in expanse of wing. Only one species, Pteropus medius, the common flying-fox, * Macrotus Californicus or "Waterhousii just enters the United States (Fort Yuma, California), but at a point which more properly belongs to the Neo- tropical than to the Ilolarctic tract. BATS. 353 has thus far been obtained from the Indian peninsula, and this, sin- gularly enough, is more nearly related to the Madagascan P. Ed- wardsii than to any of the more eastern species, although separated from its habitat by a continuous water-way of upwards of one thousand miles. The rarity of species in the Indian peninsula is not readily accounted for, seeing how numerous are the individuals belonging to the single species of the genus.* Of the two remaining families of bats, the Nycterida3 and Rhino- lophidae, the former, comprising some twelve species, are almost wholly restricted to the Oriental and Ethiopian realms, while the latter, numbering about fifty species, are spread over the greater portion of the Old World, from Ireland eastward to Japan and New Ireland, and southward to the Cape of Good Hope. No spe- cies appears to have been thus far positively identified from any of the Polynesian islands. Nearly all the species are included in the genera Rhinolophus and Phyllorhina, the former of which has prac- tically the range of the whole family ; Rhinolophus ferrum-equi- num, the common horseshoe bat, is distributed over almost the entire tract included between the south of England, Japan, and the Cape of Good Hope. The species of Phyllorhina are confined principally to the tropical and sub-tropical parts of the Old World ; Phyllorhina armigera, the most northerly species, has been found at Amoy, China, and at Mussoree, on the Himalayas, at an eleva- tion of five thousand feet. * Dr. J. Anderson thus describes the appearance of these animals (" Cat. Mamm. Ind. Mus.," 1881, p. 101, Part I): "This species has been flying for the last few days from the north to the south of the city [Calcutta] in im- mense numbers, immediately after sunset. The sky, from east to west, has been covered with them as far as the eye could reach, and all were flying with an evident purpose, and making for some common feeding-ground. Over a transverse area of two hundred and fifty yards as many as seventy bats passed overhead in one minute ; and as they were spread over an area of great breadth and could be detected in tte sky on both sides as far as could be seen, their numbers were very great, but yet they continued to pass over- head for about half an hour. This is not the first time I have observed this habit in this species ; indeed, it was more markedly seen in August, 1864, while I was residing in the Botanical Gardens, Calcutta. The sky, imme- diately after sunset, was covered with this bat, travelling in a steady manner from west to east, and spread over a great expanse, all evidently making for one goal, and travelling, as it were, like birds of passage, with a steady 354 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. Fossil remains of Cheiroptera, although of rare occurrence, are found as far back as the upper Eocene, from the deposits of which period a limited number of forms, representing in the main modern genera, have been described. The oldest known forms are Ves- perugo Parisieusis, much resembling the broadly distributed sero- tine, from the gypsum of Montmartre, Vespertilio Morloti from the nearly equivalent deposits of Switzerland, and Vespcrugo velox, V. priscus, and Nyctilcstes serotinus from the United States. In this family of placental mammals alone do representatives of exist- ing genera extend to such an ancient epoch as the Eocene. A species of Rhinolophus, R. antiquus, has been described from the phosphorites of Quercy, France — usually referred to the Oligocene period — while a generically related form, Palceonycteris, occurs in the Miocene. The Miocene deposits of both France and Germany also contain several species of the genus Vespertilio. Post -Pliocene or late Pliocene cave-remains closely approximate the forms now living in the equivalent or adjoining region. The early specialisa- tion of this class of animals, concerning whose differentiation prac- tically nothing is known, indicates a most ancient line of ancestry, which must be traced far into the Mesozoic era. Rodentia, — With the exception of the bats the rodents are the only order of terrestrial mammals which can be said to have a nearly universal distribution, being found in all the primary zoo- geographical regions but the Polynesian. On the continent of Australia, however, only a single family, that of the mice (Muridse), is represented, and that by a comparatively limited number of spe- cies ; the squirrels have a few representatives in the Austro-Malay- sian transition- tract. Of the four great divisions into which th,e rodents are divided, the myomorphs, or mouse-forms, are numerically much the most important, the family of mice alone comprising considerably more than one-third of the total number of species — some eight hundred or more— belonging to the order. The geographical distribution of this family is practically coextensive with that of the order. The mice proper (Mus), of which there are upwards of one hundred species known, are restricted exclusively to the Old "World, over which they are very extensively distributed ; they are almost wholly wanting in the Pacific and the greater number of the Austro-Malay- sian islands, but are found sparingly in both Australia and New ..RATS, MICE. 355 Guinea. Tasmania has several species, and one or more forms (M. Novse-Zelandiae, M. Maorium) appear also to be indigenous to New Zealand ; M. Salamonis inhabits the island of Ugi, in the Solomon group. The better known members of the group, the Norway or brown rat (Mus decumanus), and black rat (Mus rattus), whose origi- nal home seems to have been Southern or Central Asia, and the com- mon mouse (Mus musculus), probably a native of India, have been spread through man's intervention over the greater portion of the inhabited globe, rapidly displacing in many quarters the indigenous races of similar or allied forms that originally occupied the con- quered territory. The black rat, which appears to have been the earliest intruder, is now largely supplanted by the brown species; in England there would appear to be but a single colony left.* The wood-mouse (Mus sylvaticus) and harvest-mouse (M. minutus), the latter the smallest of the European species of mice, are distributed over the greater portion of Europe and Northern Asia. The largest member of the rat tribe is the great bandicoot or pig-rat of the Indian peninsula (Nesokia bandicota), which frequently exceeds one foot in length. Other distinctive rat-like forms of the Old World are the spiny-mice (Acanthomys), which are confined prin- cipally to Syria (Palestine), and the east coast of Africa ; the jump- ing-mice (Meriones or Gerbillus) of the continent of Africa, the warmer tracts of Southern and Southwestern Asia, and the steppe region about the Caspian Sea ; and the jumping-rats of Australia (Hapalotis), which recall in general appearance the jerboas. The Old World forms of the genus Mus are represented in the New World by the vesper-mice (Hesperomys), which very closely resemble them in general character, but differ in certain peculiari- ties of dental structure, which likewise serve to distinguish most of the American rats; some seventy or more species and varieties have been described, ranging collectively from the Arctic regions to the Strait of Magellan. Hesperomys leucopus, the white-footed or deer mouse, inhabits the greater portion of the North American * While in most regions where the black and the brown rat have been in- troduced the latter has been rapidly driving out or exterminating the for- mer, it would appear that in some parts of Central Germany the reverse phe- nomenon is presented — that is to say, the black rat is regaining its ascendancy over the brown species (Magnus, " Sitzungsber. d. (-resell, naturf. Freunde," 1883, p. 47). 356 GEOGRAPHICAL AKD GEOLOGICAL DISTRIBUTION. continent. The wood-rats, constituting the genus Neotoma, are the largest of the American murine forms, and inhabit the greater portion of the region included between Guatemala and Canada. The cotton-rat (Sigmodon hispidus) is confined principally to the Southern United States, Mexico, and Central America, occasionally, it appears, wandering into the northern portions of South America. In the genus Reithrodon are included a limited number of remarka- ble leporine forms, which, though differing very essentially in gen- eral appearance, do not seem to be distantly removed from the North American harvest-mice of the genus Ochetodon; they are confined principally to the extremity of the South American con- tinent and to Tierra del Fuego. — Interesting modifications of struct- ure or habit are seen in some of the murine forms, as in the par- tially web-footed water-rats of the Australian region (Hydromys), and in their Brazilian analogues of the genus Holocheilus ; again, in the arboreal dormouse-like forms that have been referred to the genera Dendromys (Ethiopian) and Rhipidomys (American). Of the less murine or rat-like forms of the mouse family the voles or meadow-mice (Arvicolae), which in a measure replace the true mice in the far north, and on elevated mountain-summits, and whose distribution embraces nearly the whole of temperate and Arctic Eurasia and North America, are probably the most nu- merous specifically, while in point of individual numbers they far exceed any other mammal, with the possible exception of the closely related lemming. Many of the species enjoy a very broad distribution, but none are known to be common to both the Eastern and Western Hemispheres. Arvicola arvalis or agrestis, the com- mon meadow-mouse, which ascends the Alps to a height of 7,000 feet, is distributed over nearly the whole of Europe (including Italy) and Siberia, its range corresponding approximately with that of A. amphibia (water-vole) ; A. alpina or nivalis inhabits the re- gion of the higher Alps, between 5,000 and 12,000 feet elevation; on the Finster-Aarhorn it lias been observed at an altitude of 4,000 metres. The most broadly distributed of the American species is the common meadow -mouse (A. riparia), whose range extends from the Atlantic to the Pacific, and from the Carolinas to the Hudson Bay territory. A single species, A. quasiater, is known from Mex- ico. Evotomys rutilus, a form very closely related to the arvicoles, inhabits the circumpolar regions of both hemispheres. HAMSTERS, LEMMINGS, DOKMICE. 35? The hamsters (Cricetus) inhabit the greater portion of Europe and Central and Northern Asia, the range of the common species (C. frumeiitarius or vulgaris) extending from the Rhine to the Obi, and from the Obi and Irtish southward to Persia and the Caucasus ; other species inhabit the elevated steppes of Mongolia. Pouched rats allied to the hamsters (Cricetomys, Saccostomus) are also found in various parts of Africa. The lemmings, which are readily dis- tinguished from the field-mice by the hairy covering on the soles of the feet, and their sickle-shaped claws, are the most strictly northern forms of all Rodentia. The better known species are the Scandinavian or Norwegian lemming (Myodes lemmus), the Si- berian lemming (M. Obensis), which inhabits the boreal regions of both hemispheres, and the Hudson Bay lemming (Cuniculus tor- quatus or Hudsonius), an inhabitant of Arctic America, Greenland, and corresponding latitudes in the Eastern Hemisphere ; it is also found in Nova Zembla. A strictly American genus of MuridaB is Fiber, of which the only recognised species is the musk-rat (F. zibethicus), whose range embraces practically the whole of North America. A closely re- lated, but considerably smaller, form is the recently described Neo- fiber Alleni, from Brevard County, Florida. Of the non-murine families of myomorphs the dormice (Myoxida3) and mole-rats (Spalacidae) belong to the Old "World exclusively, the pouched rats (Saccomyida?) are American, and the true jumping-mice or jerboas (Dipodidse) both Old and New World forms. The dor- mice are scattered over the greater portion of temperate Eurasia, from Britain to Japan, and southward over almost the whole of Africa; they appear to be wanting in the warmer parts of Asia (India). The common northern species, Muscardinus avellanarius, is more generally replaced in the south by Glis vulgaris, the ' ' seven- sleeper" of the Germans, whose range extends eastward to the Volga River and Georgia. Of the mole-rats, which are confined almost wholly to the African continent and the tracts comprised in the Oriental region, only a very limited number of species (Spalax) pass within the European boundaries, and these are restricted largely to the southeastern districts (Southeast Russia, Greece, Hungary). Bathyergus maritimus, the " great rodent-mole," in- habits the sand dunes of the Cape coast of Africa. The distribu- tion of the distinctively North American family of pouched rats 358 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION". has already been adverted to in the general treatment of the North American fauna. The jerboas or true jumping-mice are distributed from the eastern confines of the Mediterranean to India, and south- ward over nearly the whole of Africa. Alactaga jaculus extends its range westward from the Altai Mountains to the Danube, and Dipus sagitta from the far east of Mongolia to the Volga. The largest representative of the family is the Cape jumping-hare (Pede- tes Caffer), whose range extends from Mozambique and Angola to the southern extremity of the African continent. Only one trans- Atlantic species is known, the American jumping-mouse (Zapus Hudsonius), which is distributed over almost the whole of the North American continent between Labrador and Mexico. Of the squirrel forms, or sciuromorphs, the family of squirrels (Sciurida3), which comprises, in addition to the ordinary and flying- squirrels, also the marmots and prairie-dogs, embraces very nearly all the species belonging to the group. The true squirrels (Sciurus), of which there are probably not less than one hundred species, are extensively distributed over all of the continental divisions of the globe with the exception of the Australian, being limited only or primarily by a deficiency in the forest growth. The headquarters of the genus might be said to be the Oriental region, which holds nearly one-half of all the recognised species ; no species are known from either Madagascar or the West Indian islands, although sev- eral forms inhabit the larger islands of the Malay Archipelago — Java, Borneo, Sumatra, and Celebes. In the whole of Europe, excluding the Caucasus, there is but a single species of squirrel, Sciurus vulgaris, whose range extends from the extreme north to the Mediterranean, and eastward through- out Siberia. In the Engadine it ascends the Alps to an elevation of 7, 000 feet. North America north of the Mexican boundary possesses six species (and about an equal number of well-marked varieties de- pending upon size and colouration), of which the most familiar form is the common chickaree (Sciurus Hudsonius), in its several varieties — eastern chickaree, western chickaree, Fremont's chickaree, and Richardson's chickaree — whose range extends from the northern limit of forest vegetation to the highlands of Georgia and Alabama ; on the Atlantic border its southern limit appears to be Delaware Bay. This is the only species of squirrel found north of the Canadian boundary. The flying-squirrels are usually separated into two dis- SQUIRRELS, MARMOTS. 359 tinct groups — the flat-tailed forms (Sciuropterus), which are abun- dantly distributed over the northern parts of the North American continent, and whose range in Eurasia extends from Lapland to China and Java; and the round-tailed forms (Pteromys), constitut- ing a more southerly group, whose home is the wooded districts of tropical Southeast Asia, Japan, and some of the Malaysian islands. By some authors the separation into two generic groups is not recognised. The spermophiles or pouched marmots (Spermophilus) and ground-squirrels (Tamias) are spread over the greater portion of the temperate and boreal regions of the Northern Hemisphere, but find their greatest numerical development in the New World. The former, which in a measure connect the true marmots with the squirrels, although sufficiently abundant in the far north, of Siberia and on the most elevated slopes of the Caucasus (S. musi- cus), appear to be wanting in the Alps. The American species occupy the western portion of the continent, ranging from the Arctic seas to the plains of Mexico ; none are found east ot the central plains or prairies. Of the ground-squirrels there are some four or five recognised species, all of which are represented in North America; the northern ground-squirrel or chipmunk (Tamias Asi- aticus) is common to both the Eastern and Western Hemispheres, ranging in America from Lake Superior and Arizona to the Barren Lands, and in Eurasia from Saghalien and Japan through Siberia to the Dwina River. In the eastern portions of the American con- tinent this species is replaced by the common chipmunk or striped squirrel (Tamias striatus). The marmots (Arctomys) are restricted to the middle and north- ern portions of the Northern Hemisphere, and comprise some ten or more species, three of which are American. Of the last the best- known form is the woodchuck (A. monax), whose habitat ex- tends from the Carolinas to the sixty-second parallel of latitude, and from the Atlantic border to Minnesota. Of the two European species the bobac (A. bobac) is more properly an Asiatic form, ranging from Kamtchatka to the German frontier. The true mar- mot, the Murmelthier of the Germans (A. marmota or Alpina), inhabits the mountain-tracts of Southern Europe — Pyrenees, Alps, Carpathians — between elevations of 5,000 and 10.000 feet. The American u prairie" or u barking dogs," more properly marmots (Cynomys), of which there are two species known, appear to be 360 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. restricted to the parks and plains of the Rocky Mountain platsau region. The remaining sciuromorphs comprise the sewellels (Haplodon), leporine rodents, somewhat of the habit of the musk-rat, inhabit- ing the Northwestern United States ; the singular anomalures from Western Africa, which, in the possession of a lateral cutaneous expansion adapted to aerial sailing, recall the fly ing- squirrels; and the beavers (Castoridae), of which the only species (Castor fiber or Canadensis) inhabits the northern parts of both the East- ern and Western Hemispheres. In its American home the beaver is still met with in tolerable abundance west of the Mississippi in the entire tract included between Alaska and Mexico ; to what extent its range extends into the last-named country has not yet been ascertained. East of the Mississippi it is now but sparingly found south of the Great Lakes, a limited number being still harboured in the Maine and Adirondack wildernesses, and a still smaller number probably finding their way along the thinly settled districts southward to Alabama and Mississippi. In some portions of Virginia and Pennsylvania they appear to be still fairly numerous. The present range of the beaver in Europe is even more restricted than in America, the animal being almost wholly confined to Russia (and Poland), specially the streams of the Ural Mountains and those emptying into the Caspian Sea, and, in iso- lated colonies, to the Rhone, Weser, Elbe, and Danube rivers. The animal is now extinct in Great Britain, and appears, also, to have completely disappeared from Scandinavia. The hystricomorphs embrace a number of families whose repre- sentatives depart widely from one another in many essential char- acters ; their greatest development is in the Neotropical realm, which alone possesses the chinchillas, the agoutis, and the cavies, besides the greater number of the partially Ethiopian family of spiny -rats (Echiomyidee). The best-known representative of the last is the coypu (Myopotamus coypu), a large beaver-like animal found only in Chili, measuring nearly two feet in length. Of scarcely smaller dimensions is the arboreal Capromys pilorides, indigenous to Cuba, where it constitutes the largest native mammalian. Pla- gicdontia radium, a member of the same family, also found in San Domingo, appears to be the only indigenous mammal of the island of Jamaica, except the bats and mice (the latter probably introduced). CHINCHILLAS, AGOUTIS, PORCUPINES. 361 The chinchillas comprise a limited number of species which are restricted to the Alpine zones of the Peruvian and Chilian Andes (the true chinchillas, C. lanigera, 8,000 to 12,000 feet; Alpine vis- cachas, Lagidium Peruanum, 10,000 to 16,000 feet), and the pampas between the Rio Negro and the Uruguay (true viscachas, Lagosto- mus trichodactylus). All the agoutis and cavies (or Guinea-pigs) are, as has already been stated, restricted to the Neotropical realm, over which (principally east of the Andes) they are very extensively distributed. The range of the agoutis proper (Dasyprocta) extends from Mexico to Paraguay, one species (D. cristata) finding its way into some of the smaller West Indian islands — St. Vincent, Santa Lucia, Grenada; the paca (Caelogenys paca), the largest member of the family, inhabits the river-bottom forests over almost the entire tract covered by the remainder of the species. The cavies proper (Cavia) are spread throughout nearly the whole of the South American continent, from Guiana to the Strait of Magellan, and from the lowlands to the plateau region of perpetual snow ; one or more doubtful species are said to occur west of the Andes. Brazil is most favoured as to number of species, from one of which (Cavia aperea) appears to have descended the domestic Guinea- pig. The Patagonian cavy (Dolichotis Patagonica), an animal measuring nearly three feet in length, inhabits the plains between Mendoza and the forty-ninth parallel. The capybara (Hydrochrerus capybara), the largest of all living rodents, inhabits the whole of South America east of the Andes and north of the Rio de la Plata, wherever water is found ; its range at one time appears to have ex- tended as far south as the Salado, or even farther. Of the remaining hystricomorphs the porcupines, of which some authors recognise two distinct families, the true porcupines or por- cupines of the Old "World (Hystricidae) and the tree-porcupines, or the species of the New World (Cercolabidse), comprise a consider- able number of forms, which though closely related to one another in point of anatomical structure, affect most diverse conditions of habit. The American species range from the northern limits of trees to Paraguay, but the South American forms are generically distinct from those inhabiting North America (except Mexico). Two well-marked varieties of a single species, the Canada porcu- pine (Erethizon dorsatus), inhabit the forest region of North America. The eastern porcupine, or Canada porcupine proper. 362 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. whose range formerly extended southward to Virginia (and pos- sibly also to Kentucky), is now largely restricted to its northern habitat, although it is still found in certain portions of the State of Pennsylvania. According to Dr. Allen it is but rarely met with in New England south of Central Maine and Northern New Hamp- shire. The western variety (E. dorsatus, var. epixanthus) occupies the western half of the continent, ranging between Alaska and the Mexican frontier. The greater number of the South American species are included in the genus Cercolabes, tree-porcupine, whose combined range extends from Paraguay to Mexico ; Chsetomys sub- spinosus inhabits the warmer parts of Brazil. Of the Old World forms the best known is the crested or common porcupine (Hystrix cristata), whose home is the Mediterranean districts of both Europe and Africa, with some considerable reaches of territory also in the western part of the latter continent. In South Africa this species is replaced by Hystrix Africre-australis, and in India by the hairy- nosed porcupine (H. leucura), which is found from the Himalayas to the extreme south of the peninsula. Other species of the genus Hystrix and of the brush-tailed porcupines (Atherura) are distrib- uted over the Oriental realm from Nepaul to Borneo ; a species of atherure is also found in Western Africa. The last division of the rodents, the rabbit-forms or lagomorphs, embraces the rabbits or hares, and the pikas (Lagomys), small Guinea-pig-like animals which are restricted almost wholly to the elevated mountain districts (11,000 to 15,000 feet) of Northern and Central Asia, with a single species found in Southeastern Europe, and another, Lagomys princeps, in the Rocky Mountain region of the Western United States and Canada. The rabbits and hares (Lepus) include some twenty or more species, which are almost entirely confined to the Northern Hemisphere, where they occupy very nearly the whole of North America and Eurasia, and also Northern Africa. A single species (Lepus Brasiliensis) is found in South America, while several are known from South Africa, al- though in the vast interior of the last-named continent the genus does not appear to be represented. The distribution of the prin- cipal American species has already been discussed in the general consideration of the North American fauna. Only one of these, the polar or Arctic hare (Lepus glacialis or timidus), whose range extends over Greenland and the Barren Grounds to the Arctic HARES. 363 coast, is common to the Old World, where its range extends over the greater portion of Europe, from Scotland to the Ural Mountains and the Caucasus. Singularly enough, the species ap- pears to be wanting in Scandinavia. The variable hare (L. vari- abilis), which is by many authors considered to be identical with the last, or at best only a varietal race, inhabits Eurasia (from Ire- land to Japan) north of the fifty-fifth parallel of latitude, but reap- pears on the more elevated mountain regions of the south where the climatic conditions are approximately those of the northern lowlands. Thus, we find the animal in the Swiss, Bavarian, and Austrian Alps, in the Pyrenees, and in the Caucasus, although in much or most of the intervening lowland it is completely wanting. Unlike the last species, which in the Alpine region occupies prin- cipally the basal tracts, rarely ascending above 5,500 feet,* the variable hare more properly frequents the elevated summits, up to 10,000 feet or more, and only occasionally descends below the level of 4, 000 feet. The species is wanting in the Jura Mountains. Much uncertainty attaches to the true home of the semi-domesticated rab- bit (Lepus cuniculus), which is at the present time so extensively distributed throughout Europe, and the contiguous parts of Asia and Africa. Until recently supposed to have been introduced from Spain, the discovery of its remains in the Quaternary depos- its north of the Alps would seem to throw considerable doubt upon the accuracy of this hypothesis. The most complete analysis of the extinct rodent fauna of the Northern Hemisphere is furnished by Schlosser ("Palseontograph- ica," 1884), who recognises about seventy well-characterised species from the Tertiary deposits of Europe alone. Of these the most ancient, or those of the Eocene and Oligocene periods, belong in principal part to types that are either entirely extinct or have their nearest analogues among forms living at the present day in tropical America. Such are the genera Nesokerodon, Theridomys, and Protechimys, from the French phosphorites, the first of which ap- pears to be closely related to the South American cavies, and the last two to the spiny-rats, and more distantly to the chinchillas. The squirrels and dormice are represented by the modern gen- * Theobald affirms the presence of the common hare at an altitude of 7,000 feet in the Orisons (Fatio, "Faune des Vertebres do la Suissc," p. 250). 364 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. era Sciurus and Myoxus respectively, while Plesiarctomys presents us with an ancestral form of the true marmots. Among the extinct genera we find some remarkable suggestions of marsupial structure ; thus, Pseudosciurus and Sciurodon, in the character of their den- tition, approach the Australian koala (Phascolarctos cinereus), and Sciuroides recalls the phalangers and the kangaroo-rats (Hypsi- prymnus). The Miocene rodents, principally represented in France and Germany, although still retaining a number of the older types, as Theridomys, show a much closer approximation to the modern fauna. Among living genera we find (in the newer deposits) repre- sentatives of the true squirrels, dormice, porcupines, and pikas or tailless hares (Lagomys). The rabbits or hares (Lepus) are en- tirely wanting, as, indeed, they are from the whole of the Euro- pean Tertiary series, and likewise the true mice (Mus), if we ex- cept the single species, Mus (Acomys) Gaudryi, from Pikermi, Greece. The genus Cricetodon, whose earliest appearance is in the Oligocene deposits of France, was without doubt the near ally of the hamsters (Cricetus). One or more species of marmot (Arctomys) have been indicated as belonging to the Miocene de- posits of both France and Germany, but it is a little doubtful whether the horizons whence the remains were obtained have been correctly identified. Arctomys primigenia, from Eppelshcim, is not improbably a comparatively recent species, and not impos- sibly identical with either A. marmotta or A. bobac. The most abundant forms of this period are Myolagus Meyeri and Steneofiber Jageri, both from the Upper Miocene of Germany and France, the last replacing the more modern true beavers (Castor). Of the older Miocene genera ArchaBomys stands intermediate between the still earlier Protechimys and the chinchillas, and Issiodoromys be- tween Nesokerodon and the cavies. The Pliocene rodent fauna does not differ essentially from the Upper Miocene, of which it may be considered to be a mere amplification. The recent genera occurring here are already in the main represented in the period preceding, although a limited number of new types, such as the beaver (Castor Issiodorensis, possibly identical with the recent Castor fiber; Puy-de-D6me, France) and the Arvicolida3, are for the first time introduced. The forms related to the South Ameri- can fauna have, on the other hand, completely disappeared from ELEPHANTS. 365 the continent. The remains of the existing species of porcupine and beaver (cave of Gailenreuth) are found in the Quaternary deposits. The rodent fauna of the American (western) Tertiaries is very closely related to the European, a large number of identical, or analogous, genera being represented. This is especially the case with the forms belonging to the Miocene period, where, in addi- tion to a considerable number of extinct types, we find such forms as Steneofiber, true beavers (Castor — several species), squirrels (Sci- urus), vesper-mice (Hesperomys), and not impossibly also the true porcupine (Hystrix). Eumys does not appear to differ essentially from Cricetodon, while Ischyromys represents Sciuromys. A dis- tinctive feature separating the American from the European fauna is the introduction of the hares, which are not only represented by forms now no longer living (Palaeolagus), but also by the modern genus Lepus. In the Pliocene fauna there is a still further approxi- mation to the fauna of the present day in the appearance of an additional number of recent genera — Erethizon (Canada porcupine), Geomys (gopher). The last genus is also found in the Quaternary deposits, as also other members of the same family (Saccomys), and the vole, musk-rat, wood-chuck, ground-squirrel, wood-hare (Lepus sylvaticus), beaver, and a form of capybara (HydrochaBrus ^Esopi). Castoroides Ohioensis, the largest of all known rodents, recent or fossil, appears to have been of the dimensions of the black bear. Proboscidea (Elephants).— At the present day there are but two living species of this order known — the one being the Asiatic elephant, Elephas Indicus, which inhabits the forest lands of In- dia and Southeast Asia generally, with the islands of Ceylon, Suma- tra, and (?) Borneo, and the other the African elephant, E. (Loxodon) Africanus, a native of the greater part of the continent of Africa south of the Sahara. The insular Asiatic form is by some authors considered to represent a distinct species, to which the name E. Sumatranus has been applied. Although now restricted in a gen- eral way to the warmer parts of the earth's surface, there can be no doubt that the range of the species was very much greater at an earlier period of the earth's history than it is at present, seeing how very broad was the distribution of the genus. The remains of elephants undistinguishable from the African form have been 366 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. discovered in the Post-Pliocene deposits of Algeria, Sicily, and Spain. The extinct species of elephants are numerous, and their re- mains are largely distributed over the continents of North America (Alaska to Mexico) and Eurasia. None date back in time beyond the Pliocene period, if we except the forms from the Siwalik de- posits of the peninsula of India and the island of Perim, whose horizon is somewhat doubtfully placed by geologists as Mio-Plio- cenc. The best-known species is the mammoth (Elephas primi- genius), which was very closely related to the Indian elephant, and whose range covered the greater part of Northern Eurasia (extend- ing as far south as Santander in Spain and Home in Italy) and Northwest America. Its remains are found most abundantly along almost the entire Arctic shore of Siberia. The species belongs ex- clusively to the Post-Pliocene period, and was doubtless contempo- raneous with man in many of the regions inhabited by it. Other well-known and somewhat earlier species are the European E. antiquus and E. meridionalis, and the American E. Americanus. Elephas Melitensis, from the island of Malta, is the smallest known species, barely exceeding three feet in height when adult ; by Pohlig it is considered to represent only a diminutive variety of E. antiquus. The closely related genus Mastodon antedates the true elephants by one period, appearing in Europe in the Middle Miocene (Plio- cene in America).* Its extinction in the Old World appears to have been effected in the Pliocene period, although in America several species, and more particularly the commonest form, M. Ohioticus or M. giganteus, survived into the late Post-Pliocene. Remains of true elephants have been found in China and Japan,f and it appears not unlikely, from a fragment' of tusk recently de- scribed by Professor Owen as Notelephas, that a proboscidean, * The Loup Fork beds, in which several of the American species occur, are "by Professor Cope considered to be of Upper Miocene age. This author- ity recognises nine species of North American mastodon', to which a tenth one, M. Floridanus, has recently been added by Dr. Leidy. t At least two of the Indian forms — E. Cliftii and E. insignis — referred to the group Stcgodon, which in dental characters stands intermediate between the mastodons and true elephants, have been identified by Koken as occurring in the Pliocene deposits of China, and by Naumann in the nearly equivalent series of Japan. MASTODONS. 367 whether elephant or mastodon, also existed in Australia. The mastodon is also known from South America (M. Andium, M. Hum- boldtii). Of the recognised true proboscideans the genus Dinotherium may be considered to represent the earliest type, inasmuch as its remains are thus far known only from the Miocene and Mio-Plio- cene deposits (Europe and Asia). Extinct Animals related to the Proboscidea. — Numerous extinct animal forms, of both small and gigantic dimensions, ex- hibiting more or less intimate relationship with the Proboscidea, have been described from the Eocene deposits of both Europe and America. Among the best known of these is the genus Coryphodon (order Amblypoda of Cope), with about fourteen species, vegetable feeders, ranging in size from the dimensions of a tapir to that of an ox, and, judging from the skeleton, most nearly resembling among living animals the bear in outward appearance. The structure of the foot was largely that of the elephant. All the species are Lower Eocene (England, France, America). Belonging to the same order are the American Dinocerata, animals equalling or surpassing in size the modern elephants, to which they bore many points of structural resemblance. The upper jaw was provided with a pair of vertically descending canine tusks. The type genus of this group is Uintatherium, and seemingly the other forms which have been described under Loxolophodon, Eobasileus, Dinoceras, and Tinoceras also belong here. Some twenty-nine species are known, all of them of the Middle Eocene period. — A form uniting the coryphodons with the Dinocerata has recently been discovered by Professor Scott in the Bridger (Middle Eocene) beds of Wyoming, and named Elachoceras parvum.128 The members of the amblypod order of Mammalia, as well as the more recent Proboscidea and Hyracoidea (conies), are traced back by Professor Cope to a type of ungulate animals which largely preceded these in the order of their development, and which are by that naturalist considered to represent the primitive hoofed forms whence the modern even- and odd-toed ungulates, the Artio- dactyla (deer, ox, camel, &c.) and Perissodactyla (horse, rhinoceros, tapir), ultimately descended. The teeth were tuberculated (of the bunodont or hog type), and the feet largely plantigrade, provided with five toes, both front and rear, constituting the generalised 368 GEOGKAPHICAL AND GEOLOGICAL DISTRIBUTION. mammalian foot. The best-known genus of the order (Condy- larthra), which comprised animals intermediate in size between the opossum and tapir, is Phenacodus, from the Puerco and Wasatch Eocene. TJngulata Perissodactyla (Odd-toed hoofed-animals).— The only modern representatives of the odd-toed ungulates are the rhinoceros, horse (including the zebra and ass), and tapir. Of the first some five or six species are known, which are generally re- ferred to a single genus Rhinoceros, although by some authors several distinct genera are recognised. The African species (R. bicornis, R. keitloa(?), and R. simus, the so-called white rhinoceros), all two-horned, occupy the greater part of the continent south of the desert, while in Asia the species, both single- and double- horned, range from the forest-covered foot-hills of the Himalayas through Farther India and the Malay Peninsula to Borneo, Java, and Sumatra. The common Indian species, R. unicornis or In- dicus, is now restricted in its range almost wholly to the terai re- gion of Nepaul and Bhotan, and to the upper valley of the Brah- maputra. Many species of rhinoceros, in part referable to the genus or genera which contain the modern forms, are found fossil in Europe and India in deposits dating from the Upper Miocene ; Rhinoceros (Ceratorhinus) Schleiermacheri is Middle Miocene. The hornless genus Aceratherium, which, on the American continent, is preceded in the Upper Eocene by the somewhat rhinocerotic Amyn- odon, and which may be considered as the first true rhinoceros, ap- pears in the Lower Miocene deposits of both the Old and the New World, and is looked upon by many as the ancestral type whence, through migration and subsequent development, the existing and Post-Pliocene (R. tichorhinus, &c.) species have been derived. The American genus Aphelops, likewise hornless, belongs to a series of deposits (Loup Fork) which by some authors are referred to the Upper Miocene, and by others to the Lower Pliocene. Hyracodon, a genus in many respects allied to the rhinocerps, but possessing only three toes to the foot, is Lower Miocene or Oligocene (North American). Singularly enough, no rhinocerotic form is found in the New World above the Pliocene. — None of the existing species of rhinoceros antedate the Post-Pliocene period, although the African bicorn type is actually represented in the earlier deposits of Greece (R. pachygnathus) and the Siwalik Hills of India, which TAPIRS. 369 last also contain the remains of several forms closely allied to the existing Indian species (R. Sivalensis, R. palaeindicus). Of the tapirs (Tapiridae) there are at present five or six recog- nised species, one of which, the Malay tapir (Tapirus Malayanus), inhabits the Malay Peninsula and the islands of Borneo and Suma- tra, and the remainder the forest regions of South and Central America, one or more of the species ascending the Andean slopes to heights of from 10,000 to 12,000 feet. The Central American forms (T. Bairdi, T. Dowi) have been referred by Gill to a distinct genus, Elasmognathus. The genus Tapirus itself, which is not known in North America previous to the Post-Pliocene period, extends back in Europe to the Upper or Middle Miocene, and has continued with but slight modification of form from that time up to the present day. Its precursor appears to haver been the Listriodon (Middle Miocene), which united it with the somewhat tapiroid group of the lophiodons (Lophiodontidae— Eocene), the earliest group of known perissodactyls, comprising animals ranging in size from the rabbit to the ox. It is difficult, or impossible, to determine just whether the tapirs constitute a primarily Old World or New World group of animals, for, despite the intimate relationship which is established between them and the European Lophiodon through Listriodon, an equally close connection unites them on the western side of the Atlantic with genera — Helaletes (Tapirulus?), Desmato- therium, and Hyrachyus — which appear to have been contempora- neous with Lophiodon, and, indeed, may have preceded it. Nor is it exactly impossible, as Professor Vogt has suggested, that a parallel development on opposite sides of the Atlantic may have evolved similar forms from slightly different ancestors. The scanty remains of tapirs in the American Miocene formation are referred by Professor Marsh to the genus Tapiravus ; between these and those of Post - Pliocene age there intervenes a complete hi- atus. Of the horses (Equidae) there are usually recognised three groups : the horses proper, the asses, and the zebras. By most zoologists these are all placed in the one genus Equus, the characters defining Asinus (the asses) not being considered to be of generic value. Until Poliakof quite recently made known the existence of a new species of horse (E. Przevalskii) from the desert wilds of Cen- tral Asia it was generally supposed that the domestic animal (E. 17 370 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. caballus) was the only living representative of the caballine section of the family, and that no truly wild stock any longer existed on the surface of the earth. Whether Przevalski's horse proves to be a good species or not, there can be little or no question as to the normally wild state in which it occurs. The domestic animal has been spread through the agency of man over the greater part of the globe, where in nearly all localities it has nourished to a remarkable degree. That America was wholly, or in great part, deficient in horses at the time of the Spanish conquest, is proved beyond doubt, but at the same time it is equally proved, from the number of fossil remains that have been found between Escholz Bay in the north and Patagonia in the south, that the animal not only inhabited, but abounded on, the continent during a period com- paratively recent preceding. There is, further, very little question as to the contemporaneity of the horse and man on the American continent, and, indeed, it would appear not exactly improbable, from certain references contained in old narratives, that at least in South America the animal still lingered on even after the advent of the Europeans. What led to its general extermination, when under apparently similar physical conditions the introduced animal has been able to thrive to such a wonderful degree, is a problem which still awaits solution. The species of ass appear to be more numerous than those of the horse, although not unlikely one or more of the forms usually considered distinct will have to be classed as mere varietal types. Zoologists are practically agreed that the domestic animal (E. asi- nus) is either identical with, or only a feebly modified derivative from, the wild ass of Abyssinia (E. tseniopus), the only African species, which it very closely resembles. Three generally recog- nised species of ass roam over the wilds of West-Central Asia, the Syrian ass (E. hemippus), the onager (E. onager), from Persia and Northwest India, and the kiang or dziggetai (E. hemionus), the most horse-like in appearance, which inhabits the high table-lands of Thibet, at elevations of 15,000 feet and upwards. Two species of zebra— the quagga (E. quagga) and dauw or Burchell's zebra (E. Burchellii) — inhabit the plains of South Africa, while a third species, the mountain zebra (E. zebra), frequents the mountainous districts of the same region. A fourth form (E. Grevyi) has re- cently been described by Milne-Edwards from the land of Shoa, HOESES. 371 and is probably identical with the form seen by Speke and Grant during their journey to the lake regions. Of the fossil forms of Equus there have been thus far described some twenty or more species, which date back in both hemi- spheres to the Pliocene period. Only two of the recent species are positively known in the fossil state — the E. asinus, or ass, which is Post-Pliocene, and the horse (E. caballus), which appears to be first known from the Upper Pliocene ; it is not improbable, how- ever, that some of the Post Pliocene equine remains of Central Europe belong to the dziggetai. The Pliocene species seemingly most nearly related to the modern horse are the E. major or E. Americanus, from the deposits of North America, and the E. Ste- nonis, from the Val d'Arno, Italy. The only other genus of Equidse besides Equus is Hippidium (Pliohippus ?), which occurs fossil in the Pliocene deposits of both North and South America. Nehring, from a careful study of the numerous fossil remains of horses found in Germany and elsewhere, arrives at the conclusion that the present European animal — at least, as representing some of the races — instead of being, as is commonly supposed, a recent introduction from Asia, is in reality indigenous to the region which it now inhabits in a domesticated state, and that it has been a con- tinuous inhabitant of Central Europe, then largely in the form of a steppe country, supporting a steppe fauna, ever since the early part of the Quaternary epoch. Extinct Animals related to the Horse and Tapir.— In no group of mammals, probably, is the difficulty of drawing family boundaries greater than among the perissodactyl ungulates, a cir- cumstance due chiefly to the perfection with which many of the lines of descent have been traced out, and to the intimate rela- tionship which the animals of one line bear to the animals of one or more other lines. Thus the primitive ancestors of the horse — the four-toed and three-toed (fore and aft) Eohippus and Orohippus (Hyracotherium) from the Eocene, the three-toed Mesohippus and Miohippus (Anchitherium) from the Miocene, and the Pliocene Hipparion (also Miocene) and Protohippus — which form an almost continuous chain connecting Eohippus and the Equidae, belong to two or three families, Lophiodontida3, Palaeotheridae,* Anchithe- * To the type genus of this family, the Eocene Pala?otherium, many paleon- tologists have traced the ancestral line of the European horse — Equus Stenonis, 372 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. ridse, one of which, at least, lies as much in the direct line of descent of the tapir as it does of the horse. It will thus be seen that from the evolutionary standpoint, or from the point which views rela- tionship by descent as of equal importance with that which unites forms solely through a community of general characters, family lines could be traced among these earlier ungulates at points other than where they have actually been drawn, and with fully as much reason. For it can scarcely be gainsaid that the direct ancestors of the horse, for example, would form as natural a group among themselves as they now form three or four groups in the way they have been scattered about. Other tapiroid forms, more or less closely related to the lophio- dons, are Limnohyus and Palseosyops, both from the Eocene ; Chalicotherium, whose remains have been found in Oregon, and in Eurasia from France to China, appears in the Oligocene and Mio- cene. Approximately contemporaneous with the last, and embrac- ing animals of the dimensions of the elephant, or even larger, with certain resemblances to the rhinoceros, were the Menodon- tidoe, whose remains have been found in both hemispheres. Sev- eral genera of this family — Menodus, Titanotherium, Symborodon, Brontotherium — have been described, but it would appear that not all of these are entitled to generic recognition. A group of highly specialised and abnormal forms of Perisso- dactyla, concerning which there has been much diversity of opinion expressed, and whose position in the zoological scale has not yet been definitely established, is that of the Macrauchenidae, with the Hipparion, Anchitherium. Palaeotherium. — Dr. Max Schlosser, in an elaborate review of the phylogenetic relationships of the Ungulata (" Morphologisches Jahrbuch," 1886), considers the equine line of descept to pass through Phena- codus (as earliest form), Hyracotherium, Anchitherium, Merychippus, Hippa- rion, and Plioluppus. Hyracotherium and Hipparion, and possibly also An- chitherium, are assumed to be by origination American forms, which subse- quently wandered over to Europe, but the relative appearance of these forms on the two continents scarcely warrants this supposition. The same author- ity identifies the American Mesohippus and Miohippus with Anchitherium, and Protohippus with Hipparion ; Eohippus is considered to be probably identical with Hyracotherium. By Mr. Wortman, on the other hand, who first recognised in Phenacodus the earliest ancestor of the horse, and who unequivocally identifies Orohippus with Hyracotherium, both Mesohippus and Protohippus are considered to represent types of distinct generic value (" Eevue Scientifique," June, 1883). HIPPOPOTAMI. 373 two species Macrauchenia Patachonica and M. Boliviensis, remnants of the remarkable South American Pliocene fauna. With certain characters approximating it to the camel and horse, it is claimed by Burmeister that the animal was provided with a proboscidiform trunk. Ungulata Artiodactyla (Even-toed hoofed-animals). — The members of this sub-order, both recent and fossil, are conveniently divided into two groups — those which, like the hog, have the grinding surfaces of the molar teeth tuberculated (Bunodonta), and those, in which these surfaces are crescentically ridged, as in the sheep, ox, deer (Selenodonta). The first of these groups, the Bunodonta, comprises but two families, the swine (Suidas) and the hippopotami (HippopotamidaB). Of the hippopotami there are, as generally recognised, only two species, the common form (H. amphibius), which until re- cently inhabited most of the larger streams of the continent of Africa, from the Congo, Senegal, and Zambesi to the Nile, but whose domain has of late been rapidly narrowing (completely ex- cluded from the Egyptian Kile), and the West African Choeropsis Liberiensis, a comparatively small animal, differing primarily from the first in the possession of only a single pair of incisors in the lower jaw instead of two pairs. A third species, whose remains have been found sub-fossil in the swamp deposits of the island of Mada- gascar (in association with the giant JEpyornis), and which may consequently be classed with the recent period, has been described by Goldberg (1883) under the name of H. Madagascariensis. The common species of hippopotamus represents one of the most an- cient of the mammalian types entering into the formation of the modern fauna, it being one of the very few forms which survived the Pliocene period up to the present day. Its range was formerly very much greater than it now is, and even as late as the Post-Plio- cene it appears to have inhabited Europe as far north as Northern Wales, where its remains have been found associated with human implements. Seven other species of the genus have been de- scribed from the Pliocene and Post-Pliocene deposits of Europe and India, but none have so far been recorded from America. — The pigmy hippopotamus of Malta (H. minutus) appears to have been closely allied to, if not identical with, the living Liberian spe- cies. 374 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. The recent hogs are commonly divided into three more or less distinct groups : the peccaries (genus Dicotyles), whose range comprises the region included between Arkansas and Paraguay ; the wart-hogs (Phacochcerus) of East and South-Central Africa; and the true swine, under which are ranged the hog proper (Sus), the babyrousa (Babirusa), and river-hog (Potamochoerus). The last is exclusively West African, while the babyrousa is confined to Celebes, and some of the smaller islands of the Eastern Archipel- ago (Bouro, &c.). Much diversity of opinion still exists as to the number of species that are comprised in the genus Sus. Some fifteen or more have at various times been recognised by zoolo- gists, but by Forsyth Major,1'*9 who has probably enjoyed better opportunities for making a critical study of the group than any of his predecessors, the number is restricted to four, with a probable fifth one, concerning which we know little, but which appears to occupy a considerable part of the Ethiopian region : Sus vittatus, whose distributional area extends from Sardinia to New Guinea, and from Japan to Damara-Land (Southwest Africa) ; S. verrucosus, from Java and Celebes; S. barbatus, from Borneo; and S. scrofa, the boar, or common hog, whose domain extends, or did extend before man had greatly narrowed its limits, over the greater part of temperate Europe and Asia. This species, which was an early inhabitant of Britain, as is indicated by the remains found in the forest-bed (Post-Pliocene) of Norfolk, was completely exter- minated in that region a number of centuries ago. Several species of the genus have been found fossil in the Miocene and Pliocene deposits of France, Italy, Germany, and Greece, and five are described from the Siwalik Hills of India, one of which, S. Titan, is considered by Lydekker to have attained in extreme specimens a height to the shoulder of forty-nine inches or more. None of the genera of recent bunodont Artiodactyla are found fossil in America with the exception of Dicotyles (peccary), which, in association with a nearly related genus, Platygonus (also Plio- cene), occurs in the Post -Pliocene deposits. Closely related to the last are the Miocene Thinohyus, Choenohyus, and Hyotherium (Palaeochrerus ?), the last a genus also abundantly represented in Europe in deposits of both Upper Eocene and Miocene age. The oldest representatives of the suilline tribe appear to be the forms that have been described as Eohyus and Achanodon (Parahyus ?), CAMELS. 375 of Lower and Middle Eocene age, the former of which is stated by Marsh to have had at least four functional toes, while the latter united with its predominant suilline characters certain peculiar car- nivore modifications of the skull. Some very remarkable buno- donts of a hog-like character, found in the phosphorite deposits of Quercy, France, and named by Filhol the Pachysimia, are consid- ered by that author to possess some striking structural features allying them with the Primates, and rendering it not exactly im- probable that the last may find their earliest ancestors in these ancient types. Of somewhat doubtful position, but with distinctively suilline affinities, are Chceropotamus (Eocene) and Anthracotherium and Hyopotamus (Eocene — Miocene). Artiodactyla Selenodonta. — Among recent forms this section com- prises the camels (Tylopoda), chevrotains (Tragulina), and true ruminants (Ruminantia), with such well-known forms as the ox, goat, deer, and antelope. The camels, constituting the family Camelida3, comprise, as gen- erally recognised, two genera— the Old World Camelus, the camel proper, and the New World Auchenia, the llamas (guanaco, vicufia, alpaca). Of the former there are two species, the dromedary, or one-humped animal (C. dromedarius), a native of the deserts of Arabia, whence it has spread eastward to India, and the Bactrian, or two-humped camel (C. Bactrianus), which occupies the region of Central Asia from the Black Sea to China, and from the Himalayas to beyond the Siberian boundaries. Although the camel is gener- ally considered to be a belonging of the hotter regions of the earth's surface, it is well known that the Bactrian species thrives admira- bly in the northern districts of Mongolia, and that even as far north as the southern extremity of Lake Baikal, on the fifty-third parallel of latitude, it passes the rigours of a Siberian winter with- out apparent discomfort. That the dromedary, which is now one of the distinctive animals of North Africa, was unknown to the ancient Egyptians is proved by the absence of representations of it from all monumental inscriptions. The American representatives of the Camelida3 are the llama, alpaca, guanaco, and vicuna, con- stituting the genus Auchenia, the first two of which, inhabitants of the Peruvian and Bolivian Andes, exist now only in a state of domestication, while the vicufia inhabits the Andean slopes of Peru 376 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. and Chili, and the guanaco the plains of Patagonia and Tierra del Fuego. The Tertiary deposits of Europe have thus far yielded no traces of the Camelida3, and if we except the Camelus Sivalensis and C. antiquus, from the Siwalik Hills of India, and C. Thomasi, from the Quaternary deposits of Algeria, the same may be said of the Eastern Hemisphere generally. On the other hand, animals refer- able in part to this family, and, again, others closely related to these, are abundant in America, where they form a connecting series or chain almost as complete as that which has been established for the horse. The cameline line of descent has been traced by Professor Cope from the Oligocene or Miocene Poebrotherium, in which, as well as in the succeeding genus, the metapodial bones were dis- tinct, and the mouth was furnished with a complete series of in- cisor teeth, through Protolabis, Procamelus, and Pliauchenia (Plio- cene), the last standing in the relation of its dentition intermedi- ately between Procamelus and the camel. Auchenia, the llama, which may be considered to terminate the series, is late Pliocene and Post-Pliocene ; with it occur associated several related forms, as Protauchenia, Palseolama, &c. Professor Marsh indicates the Eocene Parameryx as the probable most ancient ancestor of the camels, whereas by Scott, Osborn, and Speir this place is given to a contemporaneous genus Ithygrammodon. It would appear, therefore, that the camels are a New World family, but this is by no means proved to be the case ; the absence of the true camel in America and its occurrence in India in deposits as ancient as the older Pliocene, render it very probable that an- cestral cameline forms will be found in the Old World as well as in the New. The chevrotains, or mouse-deer (Tragulida3), which comprise some of the smallest of known ungulates, and which in structure stand in a measure intermediate between the deer and hog, are ranged under two genera, Tragulus and Hyaemoschus, the former restricted to Southern and Southeast Asia, and the larger islands of the Eastern Archipelago, and the latter to West Africa. The family dates from the Miocene period, of which the genus Hyasmoschus is a belonging. Of the true ruminants, the CamelopardalidsB, or giraffes, con- stitute perhaps the most peculiar group. Only one species, Caraelo- ANTELOPES. 377 pardalis giraffa, is known, which ranges over the greater part of the grass-covered plains of East -Central and South Africa. An extinct species of the genus, C. Attica, which occurs in the Mio- cene deposits of Greece, appears to have rivalled or fully equalled the modern form in size. Other allied species have been described from the Siwalik Hills of India. The Helladotherium, an animal of less elevated proportions than the giraffe, but closely related to it, roamed during the Miocene, or early Pliocene, epoch over the south of Europe, from France to Greece, and across to India. With the same family are possibly to be placed also the Siwalik genera Brahmatherium, Vislmutherium, and Sivatherium, the last a huge antelopine form, referred by most zoologists to the true antelopes. The antelopes, whose special distribution has been considered in connection with the several zoogeographical regions of the earth's surface, constitute by far the most extensive group of the Ungulata, there being probably not less than one hundred distinct species. The greater number of these belong to the continent of Africa, where they inhabit as well the desert tracts as the open plains and forests, from the Sahara to the Cape, and from Senegal to the Nile. Among the better known of these forms are the springbok, blesbok, bontebok, hartebeest, buschbok, waterbok, koodoo, oryx, gemsbok, klipspringer, gnu, and eland, the last a bubaline form equalling in size a large ox. The opposite extreme in the series is presented by the western guevi, which barely exceeds the dimensions of a rabbit. The fifteen (?) or more Asiatic species, whose combined range com- prises very nearly the entire extent of the continent, with the islands of Japan, Formosa, and Sumatra, are nearly all distinct from the African, and even the generic types that are held in common are limited almost exclusively to such forms, as Oryx, Addax, and Gazella, whose domain embraces the almost contiguous desert tracts of Northeast Africa and Arabia. Europe has but two antelopine species, the Alpine chamois and the saiga, or steppe antelope, the latter of which may perhaps with more propriety be considered an Asiatic species, whose range ex- tends over Russia to the confines of Poland. North America is equally deficient with Europe, holding likewise but two species — the prong-horn (Antilocapra) and Rocky Mountain goat (Aplocerus) — while in South America the family or group is entirely wanting. The remains of antelopes, if we except certain doubtful forms 378 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. from the cave deposits of Brazil, which have been referred to Anti- lope and Leptotherium, are completely absent from America. In Southern and Western Europe they date from the Miocene period, continuing through the Pliocene and Post-Pliocene. The richest antelopine deposit is that of Pikermi, Greece, whence Professor Gaudry has described the genera Gazella, Pala3otragus, Tragoceros, Palaaoryx, and Palseoreas, the last two most intimately related to the recent Oryx (gemsbok) and Oreas (koodoo) respectively. Several, or most, of these types have also been recognised in the more or less equivalent deposits of France, Spain, Italy, and the Vienna Basin ; Antilope cristata of the Middle Miocene of Switzer- land appears to have had its nearest ally among recent forms in the chamois. African-type antelopes are still met with in the late Pliocene, or Post-Pliocene, volcanic deposits of the Auvergne (France). The only unequivocal antelopine remains from Britain are those of Gazella Anglica, recently described by Mr. Newton from the newer Pliocene beds of Norwich, England ; not impossi- bly, however, the saiga is also represented in the English fauna. Some eight or more species of the family, principally referable to modern genera — Oreas, Palseoryx, Portax (nylghau), Gazella, An- tilope, and Alcelaphus — have been described from the early Plio- cene of the Siwalik Hills. The bubaline or bovine ruminants proper (Bovina) comprise about thirteen recent species, which arc distributed over the greater part of Eurasia, Africa, and North America. The buffaloes are represented by four species, two African, Bubalus Caffer (Cape buffalo) and B. brachyceros, the former of which roams over the greater part of Southern and Central Africa, and two Asiatic, the Buffelus Sondaicus and B. Indicus, from the last of which has descended the domesticated variety which hak been so extensive- ly acclimatised in North Africa, Italy, Greece, and Hungary. A form related to the buffaloes, but differing in certain important essentials, is the dwarf wild-cow of the island of Celebes (Anoa or Probubalus Celebensis), whose early representatives occur fossil in the Pliocene deposits of the Siwalik Kills of India. Two species of buffalo, referable to the genus Buffelus, likewise occur fossil in the Indian deposits (Pliocene and Post-Pliocene), and one species, B. Pallasii, in the Quaternary of Danzig, Germany. Associated with the former are the bubaline forms Amphibos (Hemibos) and Lepto- CATTLE, SHEEP, GOATS. 379 bos. The African buffalo is thus far known only from the Qua- ternary of that continent (Bubalus antiquus, from Setif, Algeria). The wild cattle (ghaurs) of India, including, according to some authorities, the Thibetan yak * (Poephaga grunniens), and constitut- ing the genus Bibos, range from Southern India through the Malay Peninsula to Java and Borneo. The earliest representative of this group appears to be the Etruscan bull (Bos or Bibos Etruscus), from the Pliocene deposits of France and Italy. The bisons (genus Bison) are at present comprised in two spe- cies, the American (B. Americanus), which until recently inhabited the greater part of the continent of North America, but which is now restricted to a few hundred individuals, and the European (B. Europseus), also known as the aurochs, which up to the pe- riod of the Roman Empire appears to have been sufficiently abun- dant in South-Central Europe, but which is now limited to the imperial preserves of Lithuania and the wilds of the Ural and Cau- casus. Its immediate precursor was the Bison priscus, whose remains are distributed throughout the Quaternary deposits of al- most every country in Europe and of Siberia ; they have also been found at Escholtz Bay, Alaska. The Pliocene B. Sivalensis would seem to be a closely allied form, and, according to Rutimeyer, nearer to it than to the living American species, or to its Post- Pliocene predecessors, the B. latifrons and B. antiquus. Of the taurine genus Bos, which comprises the domestic cattle, several well-marked varieties are recognised, all of which are by most authorities referred back in their descent to the urus (Bos primigenius), which was abundant in Central Europe in the time of the early Roman emperors, but is now wholly extinct. By Wilckens, on the other hand, it is claimed that at least some of the breeds of cattle are the descendants of the European bison. The sheep and goats constitute an almost exclusively Old World group of hollow-horned ruminants, of which there are but two in- digenous representatives in the Western Hemisphere (North Amer- ica). One of these is the Rocky Mountain sheep (Ovis montana), which is very closely related to the argali of East-Central Asia, and the musk-ox (Ovibos moschatus), which inhabits the region of Arctic America north of the sixtieth parallel of latitude, or there- abouts, but whose fossil remains are met with as well in the * Frzcvalski describes a second species of yak as P. mutus. 380 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. Quaternary deposits of Eurasia — England, France, Germany, Si- beria, &c. — as in America. The range of this species, or of forms closely allied to it (O. bombifrons, O. cavifrons), on the American continent at one time extended to the confines of Arkansas. The Old World Caprina, whose special distribution has already been discussed in the zoogeographical part of this work, comprises some twenty or more species, which, with two exceptions — a Neil- gherry goat and an Abyssinian ibex — are confined to the Holarctic region, or to this and the Mediterranean transition tract. It is a somewhat surprising circumstance, in view of the broad distribu- tion of the goats, that the sheep, which have obtained such a firm foothold in the mountainous and intermountainous regions of Asia, should be almost entirely wanting from the continent of Europe — indeed, from the entire Eurafrican region, if we except the islands of Corsica, Sardinia, and Crete, the Balkan Peninsula, and some isolated spots on the Atlas Mountains. No unequivocal remains of the goat or sheep, except Ovibos, have thus far been discovered in any American formation, and in Europe such remains are confined almost exclusively to the Quaternary cave and brec- cia deposits (France, Italy). In India, however, they have been traced back to an older period ; Capra Sivalensis, a form closely re- lated to the recent Iharal of the Neilgherries, has been described from the Pliocene of the Siwalik Hills, and Capra Perimensis, whose remains would seem to have been associated with those of Dinotherium and Aceratherium, animals indicative of the Miocene period, from the island of Perim. A remarkable hornless form, to which Riltimeyer has given the name of Bucapra Daviesii, also belongs to the Siwalik fauna, The most important group of ungulates after the antelopes is constituted by the deer (Cervida3), which comprises some sixty or more species — excluding the Central Asiatic musk (Moschus) and the giraffe which are referred here by some authors (Riiti- meyer) — distributed over the greater portion of both the Old and the New World. Australia, as in nearly all other mammalian groups, is entirely deficient, and Africa counts but two species, the fallow-deer and a stag, which inhabit the Mediterranean re- gion. The South American forms, whose domain extends com- pletely across the continent to Tierra del Fuego, have been placed in the genera (or sub-genera) Pudu, Coassus, Furcifer, Blastocerus, DEER. 381 . and Cariacus, the last of which comprises all the North American deer north of the Mexican boundary, with the exception of the Canada stag or wapiti (Cervus Canadensis), the moose (Alces mach- lis) and reindeer (Rangifer tarandus), the last two of which are cir- cumpolar, and inhabit the whole northern portion of the Eurasiatic continent, from Norway to China. The European forms are com- prised under the three groups Cervus, which includes the stag or red-deer (C. elaphus), whose range embraces, or until recently em- braced, the whole of Europe and a large part of Northern Asia ; Dama, the fallow-deer, a native of the Mediterranean districts of Europe, Asia, and Africa; and Capreolus, the roe, which was at ona time extensively distributed over nearly the whole of Europe, with the exception of the greater part of Russia. Among the bet- ter-known Asiatic forms are Axis (peninsula of India, Ceylon, China) and Rusa, the latter containing some of the largest of the cervine tribe, several species of which inhabit the hotter regions of Hither and Farther India, and the islands of the Eastern Archipelago. The muntjacs (Cervulus), which seem to connect the true deer with the musks, inhabit the forest tracts of the Oriental region, from India to China, and from Formosa to the Philippines, Java, and Bor- neo. The earliest cervine animals, in the strict sense of the term, with which we are acquainted, are met with in the Middle Miocene de- posits of France and Germany, where forms showing evident rela- tionship with the muntjacs, and possessing the simplest kind of horn structure — a simple bifurcated stem — have been variously described as Procervulus, Prox, Dicrocerus, Palaeomeryx, and Micromeryx. These are all united by Rutimeyer into the one genus Palseomeryx, to which is also added the supposed differing Dremotherium from the same, and a possibly lower (Lower Miocene), horizon. Of equiv- alent age is the hornless Amphitragulus. The progressive devel- opment from the simple- formed antler to the more complex has been traced through numerous forms of Cervus from the Upper Miocene to the Post-Pliocene, the most complex structure known being that exhibited by C. dicranios, from the Pliocene of the Val d'Arno, Tuscany. Professor Boyd Dawkins thus sums up his ob- servations on this point : "We may gather from the study of the fossil Cervida3 the important fact that in the Middle Miocene age the cervine antler consisted of a simple forked crown only. In 382 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. the Upper Miocene it becomes more complex, but is still small and erect, like that of the roe. In the Pliocene it becomes larger and longer, and altogether more complex and differentiated, some forms, such as the Cervus dicranios of Nesti, being the most complicated antlers known either in the living or fossil state. These succes- sive changes are analogous to those which are to be observed in the development of the antlers in the living deer, which begin with a simple point and increase their number of tines until their limit is reached. It is obvious, from the progressive diminution in size and complexity of the antlers in tracing them back from the Pliocenes into the Upper and Middle Miocenes of Europe, that in the latter period we are approaching the zero of antler devel- opment. In the Lowrer Miocenes I have failed to meet with evi- dence that the deer possessed any antlers." 13° The roe, stag, elk, and reindeer occur fossil in the Quaternary deposits, together with a giant form, the Irish stag (Cervus me- gaceros), whose extinction appears to have been effected long after the region inhabited by it was also inhabited by man. The stag (wapiti), elk, and reindeer also occur fossil in the American Post- Pliocene deposits, the last, as in Europe, in latitudes very much lower than it now occupies. The genus Cervus dates from the Pliocene. A Quaternary form intermediate in many points of structure between the true deer and the elk, and originally re- ferred to the latter, has recently been re-described by Scott as Cer- valces (C. Americanus). Its remains have thus far been met with only in New Jersey and Kentucky. The hornless Miocene genus Leptomeryx, which by Leidy and Marsh is placed near the Cervidae, is by Kutimeyer considered to more nearly approach the camels. Of undeterminable position among the Ruminantia, but more or less closely related to each other, and showing certain analogies of structure with the Tragulina, are the Old World Lower and Middle Tertiary genera, Anoplotherium, Xiphodon, Dichobune, and Caino- therium. An equally aberrant type of American ruminants, the " ruminating hogs " (Oreodontidae) of Dr. Leidy, whose remains are exceedingly abundant in the Western Territories, appears to have been nearly related to the Anoplotheridae. Thirty-five species, with three exceptions (Merychyus — Pliocene), all belonging to the Miocene period, are referred to this family by Professor Cope.131 CATS. 383 One of the genera of the family, Agriochoerus, seems to be also represented in the deposits of the Siwalik Hills. Carnivora. — The members of this order may be conveniently classed under four primary groups, defined in their broadest sense as the cats (>^Eluroidea), dogs (Cynoidea), bears (Arctoidea), and seals (Pinnipedia). The first of these embrace the cats proper (Fe- lidaB), civets (Viverridae), the South African aard-wolf (Proteles Lalandii), and the hyenas (Hyaenida)). The true cats, which by many authorities are considered to be comprised within the single genus Felis, have an almost world-wide distribution, but are most abundantly developed in regions of elevated temperature. No species occurs in either the Australian region or Madagascar. The better-known American forms are the jaguar (F. onca), whose range comprises the entire region included between Patagonia and Texas; the cougar or puma (F. concolor), with probably the most extended north and south range of any mammalian species — Patagonia to the sixtieth parallel of north latitude ; the ocelot (F. pardalis), which, in one or other of its several varieties, ranges from Arkansas through Texas and Mexico to Patagonia ; the nearly equally distributed margay (F. tigrina — Mexico to Paraguay), and several allied species of small intertropical "tiger-cats;" the jaguarundi (F. yaguarundi) and eyra (F. Eyra), unspotted cats ranging from Paraguay to the northern boundary of Mexico, the Chilian colocollo (F. colocollo), the pampas-cat (F. pajeros), and the lynx. The last, of which several species or varie- ties have been described, in whole or in part identical with the common European species (F. lyncus or mfa), inhabits the greater part of the American continent north of Mexico. Of the Old World cats, besides the lion and tiger, whose range has been specially considered in the zoogeographical portion of this work, the better-known forms are the Felis pardus, leopard or pan- ther, which may represent several distinct species, inhabiting the greater part of Africa and the warmer regions of Asia, from Pales- tine to Japan; the ounce or irbis (F. uncia), of about the size of, and somewhat resembling, the leopard, a native of the elevated mountain-tracts of Central Asia (Thibet — Siberia), where it ascends to heights of from 15,000 to 18,000 feet; the spotted or clouded tiger (F. macroscelis), an arboreal species, indigenous to the forest regions of Southeast Asia and the adjoining islands of Formosa, 384 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. Sumatra, Java, and Borneo ; the serval (F. serval), from the greater part of the African continent ; and the cheetah or hunting leopard (F. or Cynaelurus jubatus), whose domain covers nearly the whole of the African continent and a very considerable part of South- ern and Western Asia. The lynx or lynxes range from the polar regions to the Mediterranean, whence they are continued by an allied form, the caracal, over a large part of both Asia and Africa. Felis catus, the wild- cat proper, which is met with in both insular and continental Europe, is not, as is frequently supposed, the an- cestor of the domestic animal ; this place is now generally conceded to the Egyptian and West Asiatic F. maniculata. Numerous species of the family, referable in considerable part to the genus Felis of most authors, and differing but little from forms still living, .are found fossil in the Post-Pliocene, Pliocene, and Miocene deposits of Europe, Asia, and America. Among the best known of these is the cave-lion (F. spelaea), a species but barely if at all distinguishable from the F. leo, whose remains are abun- dantly met with in the Post-Pliocene cave deposits of continental Europe and England. During the same period the existing lion appears to have hunted its prey as far north as Yorkshire and on the frontiers of Poland, and the leopard or panther among the Mendip Hills. Felidse, allied to the panther and lynx, have been discovered in the Pliocene strata of the Siwalik Hills of India. Felis angustus, from the North American Pliocene deposits, was in- termediate in size between the jaguar and tiger, while the later F. atrox, which may be considered to represent in the New World the European cave-lion, appears to have surpassed in this respect both the lion and the tiger. In association with the modern type-forms of Felidee there occur others which depart very widely from these, the most remarkable of which, as representing the most highly special- ised forms of the family, and as strictly the most carnassial of all known Carnivora, were the so-called sabre-tooths. The animals of this group are characterised by a prodigious development of the , upper canines, which in some instances appear to have measured as much as seven or nine inches in length. The best-known genus is Machairodus (Drepanodon), whose remains have been found in both the Miocene and Pliocene deposits of Europe (Pliocene of India), and whose immediate American representative appears in Smilodon (Pliocene or Post-Pliocene of Buenos Ayres and Texas), a contem- CIVET-CATS. 385 porary of the giant sloths and glyptodons. A supposed species of the last (S. gracilis) has also been described from a cave deposit in the State of Pennsylvania. Other allied forms are the American genera Dinictis, Nimravus, Pogonodon, and Hoplophoneus, from the Lower Miocene deposits of the Western United States, whose members were intermediate in size between the lynx and tiger. Eusmilus bidentatus, from the phosphorites of France (Upper Eocene or Oligocene), although the oldest-known form, is, singu- larly enough, in many respects also the most specialised. Contem- poraries with it were Pseudselurus and ^Elurogale, the former the representative of the group denominated by Professor Cope the " primitive " non -specialised cats. The Viverridae, or civet-cats, comprise some one hundred or more species of moderate -sized Garni vora, which are in the main restricted to the Ethiopian and Oriental regions. The better-known types are the true civets (Viverra), from North and Tropical Africa, India, China, and the Malay Peninsula ; the genets (Genetta), from Africa (the entire continent), Southwest Asia, and Southern Europe (France, Spain); and the ichneumons or mongooses (Herpestes), which are widely distributed over the continent of Africa and Indo- Malaysia. One species of the last is also found in Spain. Among other genera of the family are Viverricula, the rasses, and Para- doxurus, palm-civets (both from Indo-Malaysia), Cynogale, the otter-civets (Borneo), and Cryptoprocta, the last sometimes con- sidered the type of a distinct family, and the largest of the Mada- gascan carnivores. Numerous genera, more or less closely allied to recent forms, carry this family back to the early Tertiary period, where (in the phosphorites of Quercy, France) we find two or more species repre- senting the genus Viverra itself. Others of the same genus (or possibly Genetta) are found in the French Miocene, and in the Pliocene of the Siwalik Hills. A remarkable viverrine form, show- ing intermediate relationships between the civets and the hyenas, has been described by Professor Gaudry, from the Middle Tertia- ries of Greece and elsewhere, as Ictitherium. Three well-differentiated species of hyena are recognised by zoologists — the striped hyena (H. striata), from Africa generally and Southern Asia; the spotted hyena (H. crocuta) — Africa, south of the desert, sometimes placed in a distinct genus, Crocuta ; and 386 GEOGEAPHICAL AND GEOLOGICAL DISTRIBUTION. the brown hyena (H. brunnea), from South Africa. Although now restricted to the continents of Asia and Africa, the numerous re- mains found in the European Post-Pliocene deposits indicate that this animal, as well as the lion and other semi-tropical species, was an abundant form in the north temperate regions at a comparatively recent period, and that from those parts the Ethiopian realm has drawn much of its existing distinctive fauna. The widely distrib- uted cave-hyena (H. spela^a), whose range embraced a part of the British Isles, was most nearly related to, if not identical with, the H. crocuta, and was without doubt its direct ancestor. The striped hyena may be traced back to the older (Pliocene) H. Arvernensis of Central France, and the brown form not improbably to the Miocene (or Pliocene) H. eximia of Pikermi, Greece. The aberrant form Hysenictis, described by Gaudry from Pikermi, and showing cer- tain viverrine relationships, is considered by Lydekker to repre- sent at most only a sub-genus of hyena. No representatives of this family, either recent or extinct, have thus far been discovered in America, unless, indeed, the Miocene ^Elurodon prove to be a distant relative. The Cynoidea, or canine division of the Carnivora, comprises but a single family, the dogs (Canidae), whose numerous representa- tives enjoy a nearly world- wide distribution. Apart from the hunt- ing or hyena dog of South Africa (Lycaon picta*), the long-eared fox (Otocyon megalotis), from the same region, and the bush dog (Icticyon venaticus), from Brazil, all the species— some fifty or more — may be conveniently grouped in the single genus Canis, whose range would then be coextensive with that of the family. If the dingo, or wild-dog of Australia, be proved to be indigenous to that continent, then the genus will be the most nearly cosmopolitan of any of the terrestrial Mammalia. Two clearly defined sections of the genus may be recognised, the lupine and the vulpine, to the former of which belong the wolves, jackals, dogs proper, and a number of not readily classifiable forms which have a general canine aspect ; and to the latter the foxes and fennecs. The origin of the various breeds or races of the domestic dog is involved in much uncertainty, and whether their progenitors are to be sought in a * A fragment of a jaw from the Post- Pliocene deposits of Glamorganshire, Wales, has been referred to the genus Lycaon by Lydekker (L. Anglicus). DOGS, FOXES. 387 single one of the feral forms now living, as the wolf or jackal, or in several such forms as are denominated wild-dogs, or in the union of both, still remains to be determined. The researches of Nehring seem to indicate that a race of wild-dogs, akin to the existing do- mestic one, inhabited a considerable part of Central Europe during prehistoric times. Various forms of wild-dog, as the dhole and buansuah (sub-genus Cuon), range over the greater part of Asia, from Siberia to Java and Sumatra, where they in great measure replace the wolf of the more strictly northern regions. The last (Canis lupus) is found throughout the whole of Europe and North- ern Asia, from the Atlantic to the Pacific, and also in Nova Zembla and Japan. There can be little question as to the identity with this form of the corresponding American species (Canis occiden- talis), which, in its numerous varieties, covers the entire North American continent, from Mexico to the Arctic Ocean. The coy- ote, or American prairie-wolf, is by some authors considered to be intermediate between the wolf and fox. South America is wholly deficient in wolves, as in foxes, their place being taken by the fox- like forms which have been referred to the groups Lycalopex, Pseudalopex, and Thous. The most broadly distributed of these is Azara's dog (C. Azarse), which ranges over the greater part of Brazil, and southward to Patagonia. The most southerly species of the family is the Antarctic dog (C. Magellanicus), which inhab- its Chili, Patagonia, and Tierra del Fucgo. Of the remaining lupine forms the most widely distributed are the jackals, of which several species are recognised, whose combined ranges embrace the whole of Africa and much of Southern and Western Asia. The vulpine section of the Canida3 includes the fennecs and foxes, the former all African, the latter — with ten to fifteen species — spread over the whole of North America and Europe, and largely also over Asia and Africa. There can be no question (as in the case of the wolves) that several of the forms that have generally been recognised as distinct species are in reality only varietal types, whose inter-relationship is made manifest when full geographical suites are made use of for comparison. The identity between the common European fox (C. vulpes) and the American red-fox (C. fulvus) may be considered as established. Recognised as one species, the habitat of the common fox of the Northern Hemisphere may be said to embrace the whole of Europe, North Africa, North 388 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. and Central Asia, and practically the whole of the North American continent, although it appears to be absent from the immediate Pacific coast. A well-marked variety of this form, by many natu- ralists considered to be a distinct species, is the broadly distributed silver-fox (C. Virginianus), which alone of the different varieties is represented in Central America. The burrowing-fox (C. velox) is an inhabitant of the interior region included between the (low- er) Missouri and Saskatchewan rivers and the Cascade Mountains. Over a considerable part of North-Central Asia — Tartary, Mon- golia, Siberia — the common fox is replaced by the corsac or steppe- fox (C. corsac), a closely related species. The most northerly species of fox is the Arctic fox (C. lagopus), a circumpolar mainland form, occurring also in Iceland and Spitzbergen. The Canidse appear to date from the (Upper) Eocene period, and not impossibly the genus Canis is itself represented in the form that has been described by Cuvier as Canis Parisiensis. Less doubt attaches to the C. Filholi, from the phosphorites of Central France, whose position, however, still remains somewhat uncertain. Barring these two forms, the oldest representatives of the family are seen in the genera Galecynus and Amphicyon, which appear in Europe in the Upper Eocene and Lower Miocene respectively. In America Galecynus is unknown prior to the Oligocene Cor Lower Miocene — White River formation), to which period must be referred the most ancient undoubted remains of the family in the New World. To this genus, whose representatives appear to have been very abundant during the Miocene epoch, the existing species of dog are referred in their line of descent by Filhol and Cope. Canis, which in the Miocene is associated with a number of allied generic forms — Tem- nocyon, Oligobunis, ^Elurodon (with feline andg hysenoid relation- ships), in addition to those above named — becomes the dominating, if not the only, type in the Pliocene, where also we meet with the earliest existing species, the wolf and coyote (Western Territories of the United States). The Post-Pliocene deposits, contain the re- mains of the wolf, fox, and dog.* * Mr. J. A. Allen has recently described a species of extinct dog (Pachy- cyon robustus) from Ely Cave, Lee County, Virginia, which in many respects departs widely from the type of any of the ordinary wild or domesticated races. In its general proportions, the shortness of the legs, &c., it mere nearly approaches the badgers. Its geological horizon has not been absolutely determined. BEARS. 389 The Arctoidea, or ursine division of the Carnivora, includes the bears (Ursidae), weasels (Mustelidae), raccoons (Procyonidae), and the singular panda (Ailurus fulgens), from the Himalaya Mountains, whose connection with the true bears is established by the Thibetan Ailuropus. The bears, whose range embraces practically the whole of the continents of North America and Eurasia, comprise some ten or more species, most of which fall under the division Ursus proper. Among the better known of these are the American grizzly (Ursus horribilis), from the Western United States and British Columbia; the circumpolar white or polar bear (U. or Thalassarctos mariti- mus) ; the brown bear (U. arctos), the common species of Europe and Asia, which also appears to be identical with the common American or black bear (U. Americanus), and to which the (Hima- layan) Isabelline and Syrian bears (U. Isabellinus and U. Syriacus) are nearly related; the Japanese bear (U. Japonicus); the Indo- Malaysian sun-bear (U. or Helarctos Malayanus); and the sloth- bear (Melursus labiatus), from India and Ceylon. A solitary species (U. Crowtheri) is found on the African continent (Atlas Mountains), and likewise but a single one in South America, the spectacled bear (Tremarctos ornatus), from the Peruvian and Chilian Andes. The last is, according to Giinther,132 undistinguishable in its dental characters from a species inhabiting the island of Formosa. Two species of bear, the Ursus Arvernensis and U. Etruscus, are known from the Pliocene deposits of Europe, which, with the forms described from the Siwalik Hills, constitute the oldest mem- bers of the genus with which we are acquainted. The complete absence of ursine remains from the American Tertiary deposits would seem to indicate that the existing New World representa- tives of the family were a recent introduction, a supposition strengthened by the discovery of the remains of the grizzly, as a contemporary of the great cave-bear (U. spelaaus), in the European Post-Pliocene deposits. Of forms most nearly related to the true bears are Arctotherium (Pliocene or Post-Pliocene of Buenos Ayres) and Hyaenarctos (Upper Miocene and Pliocene of Europe and India), which last, by way of Dinocyon and Amphicyon, would seem to effect a direct transition to the dogs. It thus appears that Amphi- cyon lies at the converging point of both family lines. The Procyonidae are all members of the New World fauna, and 390 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. comprise the raccoons (Procyon), from most parts of North and South America; the coatis (Nasua), Mexico to Paraguay; kinkajou (Cercoleptes), Mexico to Peru and Brazil; and the bassarids (Bas- saris), from the warmer regions of the United States and Mexico. The family, as represented in the genus Procyon, dates from the Pliocene period. The Mustelida3, whose numerous representatives are spread over the greater part of all the continental areas with the exception of the Australian, comprise among better-known forms: Lutra, the otters, whose range embraces nearly the whole of the continents of Eurasia and North America, with parts of South America and Africa; Enhydris, the North Pacific sea- otter (California — Japan); Nutria, the South American west coast sea-otter (California to Chiloe) ; Meles, the true badger ("North Europe to Japan and China) ; Taxidea, American badger ; Mellivora, the African and Indian ratels ; Mephitis, the skunk, whose range comprises the entire tract included between Canada and the Strait of Magellan ; Ictonyx, the African zorilla; Mustela, the martens, boreal forms of both the Eastern and Western Hemispheres ; * Putorius, the weasels, which are distributed over the greater part of the Northern Hemisphere, and enter into tropical Africa and South America; and Gulo, the wolverine or glutton, whose habitat in America extends from about the fortieth to the seventy-fifth parallel (Melville Island), and in Eurasia from Lithuania to Kamtchatka and the Arctic tundras. In the martens are included the Asiatic sable (Mustela zibellina) and the American sable (M. Americana), the range of the latter ex- tending over the greater part of the American continent north of the fortieth or forty-fifth parallel of latitude. Other well-known forms are the Eurasiatic pine-marten (M. martes), and the pekan or fisher (M. Pennanti), which is still extensively distributed over the American continent north of the fortieth parallel of latitude. Of the weasels proper (Putorius) the true or common weasel (P. vulgaris) and ermine (P. erminea) are held in common by the northern regions of Europe, Asia, and America, the true ferret or polecat (P. fcetid'is) is Eurasiatic, and the mink, comprising the two species, P. lutreola and P. vison, both Eurasiatic and American. The total number of fossil forms referable to the Mustelidse is * Martes flavigula, the Indian marten, is distributed from the southern slopes of the Himalayas to Ceylon and Java. SEALS, SEA-LION'S. 391 very limited. The glutton, badger, otter, marten, acd ermine oc- cur in the Post-Pliocene deposit of Europe, while in the equivalent American series we have species of Galictis and Mephitis. Taxidea, Lutra, and Mustela are Pliocene in North America, and Mellivora in India (Siwalik Hills). The fourth group of the Carnivora, the Pinnipedia or seals, comprise three very distinct families: 1. Otaridae, the eared or fur- seals, sea-lions, with about nine species, whose habitat is the tem- perate and cold waters of the southern oceans (as far north as the Galapagos Islands) and the North Pacific (south to California). No species is known from the North Atlantic. To this group belongs the highly -prized northern fur-seal (Callorhinus ursinus), which was at one time abundant along the greater part of the American coast between Alaska and Lower California, but is now rapidly disap- pearing, although still very abundant among the Prybilov or Fur- Seal Islands ; the species appears to be found also along the coasts of Kamtchatka and the island of Saghalien. Of the sea-lions, commonly so-called, the best-known species are the northern sea- lion (Eumetopias Stelleri), whose range extends from Behring Strait to California and Japan, and the Californian or black sea- lion (Zalophus Calif ornianus), the familiar animal of the harbour of San Francisco. The common species of the west coast of South America is the southern sea-lion (Otaria jubata). Other species of the family, popularly known as sea-bears, whose domain covers much of the southern seas, from South America to Africa and New Zealand, are referred to the genus Arctocephalus. 2. The Tri- chechidaB, walruses, containing a single species, the walrus or morse (Trichecus rosmarus), whose habitat is the icy waters of the Arctic regions of North America (from Labrador), Europe, and Asia. The animal appears not to exist on the American coast between the ninety-seventh and one hundred and fifty-eighth meridians of longi- tude, nor on the Eurasiatic coast between the one hundred and thirtieth and one hundred and sixtieth meridians. East of the Yenisei it is very rare. The North Pacific form is by some authors considered to be a distinct species (T. obesus). 3. Phocidce, the earless or true seals, which are almost universally distributed over the temperate and colder portions of the globe. One species, the monk-seal (Monaclms albi venter) inhabits the Mediterranean and the 392 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. waters of the Canary Archipelago, and an allied form (M. tropicalis) the shores of the West India islands and Florida. The species of the Caspian Sea (Phoca Caspica) is by many authors identified with the common harbour seal of the North Atlantic and Pacific oceans (P. vitulina), and the seal of Lake Baikal (P. Sibirica) with the northern ringed-seal, P. fostida. The greater number of the south- ern forms are generically distinct from the northern, of which last about one-half have a circumpolar distribution. It has been thus far impossible to determine the exact range of the different species of seal, but it appears that, of the northern forms, the harbour-seal is the most widely distributed. On the American coast it has been noted as far south as New Jersey and the Santa Barbara Islands, California, and is reported to have been also observed near Beaufort, North Carolina. Along the European coast it is not rare off the coasts of Spain and France, and is even said to occasionally enter the Mediterranean. The most northerly species appears to be the ringed-seal, which has been met with considerably beyond the eighty-second parallel of latitude. The Greenland or harp-seal (P. Groenlandica) appears to be a permanent inhabitant of the St. Lawrence River. The more aberrant forms of the family are the hooded-seal (Cystophora cristata), from the colder parts of the North Atlantic,* and the sea-elephants (Macrorhinus), of which there are two generally recognised species, one of which (M. an- gustirostris) appears to be confined to the coasts of California and Western Mexico, and the other, the southern sea-elephant (M. elephantinus or leoninus), to the waters of the southern oceans (Patagonia, Juan Fernandez, Kerguelen Land, Macquarie Island). The former species is now almost completely exterminated. With the exception of some doubtful fragments described from European museums, the only unequivocal remains of Otaridae have thus far been described from the Pliocene deposits of Victoria, Australia, and the Post-Pliocene of New Zealand. The remains of the walrus have been found in the Post-Pliocene deposits of various parts of North America — south to New Jersey and South Carolina — while in Europe its representatives appear to be traced back to the Pliocene, or even late Miocene, period. Most of the so-called trichecoid remains, however, have been shown by Van Beneden to * An individual of this species was obtained in November, 1883, at Spring Lake, New Jersey (Brown, "Am. Naturalist," xvii., p. 1191). MONKEYS. 393 belong to animals having no close relationship with the walruses. The only undoubted phocine fragment found in any American formation of older date than the Quaternary belongs to a form described from the Miocene of Virginia as Phoca Wyrnani. In Europe, especially in the Antwerp Basin, remains of the family are abundant, and in living and extinct genera (Phoca, Palaeo- phoca, Callophoca, Gryphoca, Monatherium, Prophoca) extend back through the Pliocene to the later Miocene period. Leith Adams has described a species of Phoca from the Miocene calcareous strata of Gozo, near Malta. Extinct Carnivora of Uncertain Position. — Under the order Creodonta Professor Cope has united a number of generalised Euro- pean and American carnivore types, which differ in many essentials — greatly reduced cerebral hemispheres, absence of scapho-lunar bone, ungrooved astragalus — from the true Carnivora, and whose direct affinities are at once with these last, the marsupials and insectivores. They are regarded as the primitive carnivores, inasmuch as from these two at least of the great modern groups — the ^Eluroidea (cats) and Cynoidea (dogs) — are claimed to be directly derived. Of the four extinct families that are referred here, the Miacidse (Miacis, Didymictis), Oxyeenida? (Oxysena, Palaeonyctis), Mesony- chida3 (Mesonyx), and Hyrenodontidse (Hya3nodon), the first three are restricted to the Eocene period (beginning with the Lower Eocene), while Hysenodon is both Upper Eocene and Lower Mio- cene. The Miacidae and Oxysenida3 are considered to be the ances- tral forerunners of the dogs and cats respectively, a union between the latter and the civets being seemingly effected by the genus Proviverra. Here, perhaps, may also be referred the European Arc- tocyon, one of the oldest forms of Tertiary mammals known. Primates (Monkeys and Man *).— Naturalists usually recog- nise three distinct groups of the quadrumanous section of the Primates : the monkeys or apes of the New World (Platyrhina), the apes of the Old World (Catarhina), and the lemurs or half- monkeys (Lemuroidea), inhabitants of both continental and insular Asia and Africa. In the broader aspects of their distribution the members of this order may be said to be restricted to a zone included between the thirtieth parallels of north and south lati- tude, although a limited number of forms pass slightly beyond * The consideration of man is not entered into in this work. 18 394 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. these boundaries ; they are, therefore, essentially tropical in habit. To what extent, however, climate alone is efficient in determin- ing this distribution still remains to be ascertained, as it is well known that certain forms, most intimately related to species in- habiting the torrid lowlands, appear to habituate themselves to regions of opposite climatic conditions, or where a fairly rigourous winter prevails. Semnopithecus schistaceus has been observed in the Himalayas at an elevation of 11,000 feet, sporting among the garlands of a winter's snow, while a second species of the same ge- nus, 8. Roxellanoe, and a macaque (Macacus Thibetanus), steadily inhabit the snow-clad mountains of Moupin, Thibet, at a nearly equal altitude. The most northerly apes known are the two spe- cies last mentioned, a species from Japan (Macacus speciosus), and the Barbary ape of the Hock of Gibraltar (Macacus inuus), but it is a little doubtful whether the last is truly indigenous to the region which it now inhabits. The southern limit in the Old World is the region about the Cape of Good Hope, the home of the chac- mas. In the New World no form is positively known to pass north of the twentieth parallel of latitude in Southern Mexico (Ateles vel- lerosus, a species of spider-monkey), but it is by no means improb- able that a more northerly extension may be reached by some spe- cies; * the American forms being exclusively arboreal in habit, their southern extension will necessarily be determined by the limit of forest growth, which, excepting along the Andean slopes, is in about the thirtieth parallel of latitude, beyond which line no monkeys are known. It is a circumstance of some little importance, as bearing upon geographical distribution in general, that certain continental isl- ands, as the West Indies, apparently so well adapted in their natural physical conditions to the development of the members of this group of animals, should be entirely deficient in them ; the same holds true with New Guinea, and, indeed, with the entire continent of Australia. Madagascar, while largely supplied with the lemurs, or half-mon- keys, is wholly wanting in the apes proper. Long-continued isola- tion of the tracts under consideration is, doubtless, the primary, if * According to a statement of M. Salle", made some twenty -five years ago, monkeys (probably a species of Atelcs) were found as far north as the upper Tampico, or up to about latitude 23° (Sclater, " Nat. Hist. Keview," 1861, p. 509). AMERICAN MONKEYS. 395 not the only, cause of this deficiency. At the same time, it is not quite as easy to account for the present northern limitation. In Mexico, for example, as far as we are able to judge of the general character of the environment, or of the physical conditions govern- ing it, a much more northerly extension might have been assumed than is actually found ; but, possibly, the matter of a particular form of food-supply may have something to do with restriction in this quarter. The semi-continent of Europe, again, whose only simian inhabitant is the Barbary ape already mentioned, in view of the fact that at one time it was the home of various forms of ape, offers another instance of a region apparently suited to the wants of these animals, yet practically entirely deficient in them ; but here, doubtless, the extermination was a part of the general extermina- tion which removed so many of the more distinctive types of Mio- cene and Pliocene mammals into other regions, whatever the exact cause may have been. The New World monkeys are generally all included in two fam- ilies : the Cebidse, with several sub-families, monkeys with thirty- six teeth, and the Hapalida3, or marmosets, monkeys with thirty- two teeth. The former comprise the sapajous (Cebus), which may be taken as the representative genus of American monkeys, the woolly monkeys (Lagothrix), spider-monkeys (Ateles, and the re- lated Eriodes), howlers (Mycetes), sakis (Pithecia and Brachyurus), night-monkeys or douroucoulis (Nyctipithecus), squirrel-monkeys or saimiris (Chrysothrix), and the related Callithrix. The total number of species known is between seventy and eighty, of which about twenty are sapajous, fifteen spider-monkeys, ten howlers, and about an equal number members of the genus Callithrix. The extensive equatorial forests of the Amazon and Orinoco, and their tributaries, constitute par excellence the home of the American monkeys, but the majority of the genera have a very ex- tended range, appearing in one or more species throughout the greater portion of the tract covered bypthe entire family. This is more particularly the case with the sapajous, spider-monkeys, howlers, and callithrixes. The range of the species, on the other hand, is not infrequently very sharply defined, as, for example, when a natural barrier, offering insurmountable obstacles to fur- ther migration, suddenly interposes itself. Examples of such limitation, as brought about by the dominant water-courses of the 396 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. equatorial forests, have already been noticed in treating of specific distribution in general (pp. 23, 24). The number of species found in, and north of, the Isthmus of Panama is ten, of which only one, the spider-monkey already referred to, extends into Mexico ; My- cetes villosus, the Guatemalan howler, or mono, has thus far been found only in Guatemala and Honduras. It is a little surprising that the range of only two of the species — the black-faced spider- monkey (Ateles ater) and one of the night-apes (Nyctipithecus vo- ciferans) extends beyond Colombia in South America. None of the South American monkeys appear to pass west of the Andean chain of mountains south of Ecuador, and even north of the Peruvian boundary the number of such transgressional forms is very limited. Indeed, even among the wooded slopes; a habita- tion along the basal line of the mountain axis seems to be much preferred. The greatest altitude at which monkeys were observed by Tschudi in Peru was 3, 000 feet (Lagothrix Humboldtii) ; Ateles ater and Cebus robustus were found at 2,500 feet. On the other hand, Godman and Salvin state that in the district of Vera Paz, in Guatemala, the mono or howler is most abundant at an elevation of 6,000 feet ; and on the volcano of Atitlan, in the same country, Mr. Salvin found troops of the Mexican spider-monkey (Ateles vellero- sus) in the forest region of 7,000 feet elevation. The range of the marmosets and oustitis (Hapalida?) is nearly coextensive with that of the monkeys proper, but no form is thus far known to pass beyond the Isthmus of Panama ; * Midas Geof- froyi alone inhabits the Isthmus. The species, of which there are some thirty or more referable to two genera (or sub-genera), Midas and Hapale, are most numerous in the equatorial forests Of the Old World Quadrumana, the anthropoid apes (Simiims), which include the gorilla, chimpanzee, gibbon, and orang, acquire special importance by reason of their high structural organisation. In the sum of all their characters, the gorilla probably stands the highest, although by many naturalists this place is conceded to the chimpanzee. Only one species of gorilla (Troglodytes gorilla) has thus far been positively determined, but not impossibly other forms may inhabit the interior of the African continent. The rec- ognised habitat of the species is the west coast of Africa a few de- * Midas rufiventer, erroneously described as coming from Mexico, is a Bra- zilian species (M. labiatus). GOKILLA, CHIMPANZEE, BABOONS. 397 grees on either side of the Equator, or the forest region drained by the Gaboon, Muni, Fernand-Vaz, and Ogowai rivers. Much uncer- tainty still remains as to the number of species of chimpanzee, but most naturalists seem inclined to unite all the variously designated forms, either actually found living or reported to be such, into a single species, Troglodytes niger, whose habitat extends from the west coast (Gambia — Benguela) through the heart of the continent to the central lake region. The Asiatic anthropoid apes are the gibbons (Hylobates) and orang (Simia satyrus), the latter restricted to the forests of the islands of Borneo and Sumatra. The gibbons, or long-armed apes, which probably comprise a dozen or more species, are confined to South-Eastern Asia, and some of the larger islands of the Eastern Archipelago. On the continent they range from the Brahmaputra River, in Assam, to the Malay Peninsula, and eastward to the region about Canton, China (Hylobates pileatus); the Chinese species is also found in the island of Hainan. The better-known forms are the siamang (H. siamanga), the largest member of the genus, from Su- matra, hoolock (H. hoolock), the most northern form (Assam, Ben- gal), and lar (H. lar), from Siam, the Malay Peninsula, and Su- matra. Of the non-anthropoid Quadrumana of the Old World the most numerous in point of species are the dog-apes (Cynopithecinae) — green-monkeys, macaques, drills, baboons, &c. The long-tailed forms of the genus Cercopithecus are exclusively African, and com- prise all the more graceful monkeys of the continent that have been variously designated guenons, green-monkeys, and white-nosed monkeys. Collectively, they range over the greater part of the tracts included between the Gambia and Congo Rivers on the west and Abyssinia and the Zambezi on the east. The mangabeys, some- times separated as a distinct genus (Cercocebus), are West African, as is also the talapoin (Miopithecus talapoin). Almost equally dis- tinctive of the African region are the dog-faced baboons of the genus Cynocephalus, which have a very general distribution throughout the continent, extending also into the adjoining tracts of Asia (Arabia). Among the better-known members of this group are the mandrill and drill (C. maimon or mormon, and C. leuco- phseus), both from the west coast (Guinea) ; the baboons proper (C. babuin), whose habitat extends from Abyssinia and Kordofan 398 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. into the wilds of the interior ; the nearly related and rock-inhabit- ing chacma (C. porcarius) and sphinx (C. Sphinx), from the south and west of the continent respectively ; and the hamadryas (C. hamadryas), whose home is constituted principally by the coast mountains of Abyssinia and Southern Nubia, and the littoral of Western Arabia. A somewhat aberrant form, the gelada (C. gela- da), differing from the other baboons in the non-terminal position of the nostrils, and hence sometimes constituted into a distinct ge- nus (Theropithecus), inhabits the highlands of Abyssinia at an elevation, according to Schimper, of from 10,000 to 13,000 feet. The macaques, if we exclude the Barbary ape or magot, whose habitat in the north of Africa and on the Rock of Gibraltar has al- ready been noted, are exclusively Asiatic, ranging on the continent from the Himalayas to Japan, and southward to the extremity of the Malay Peninsula. One or more species of the group are found on nearly all the more important islands of the Malay Archipelago, from Sumatra to Timor. The best-known form is the common macaque (Macacus cyuomolgus), wrhose habitat comprises nearly the whole of Southeast Asia, and the islands of Sumatra, Banca, Java, Bor- neo, Celebes, Bali, Lombok, Flores, Sumbawa, and Timor. In Java, where it ascends to a height of 5,000 feet, it is one of the commonest of animals, and has been brought into a general condi- tion of domestication. Other well-known forms of the group are the Rhesus-monkey (M. Rhesus), whose home is British India, especially the wooded tracts of the lower Himalayas, and the wanderoo (M. Silenus), from the forest region of Malabar. The former has been observed to ascend the Himalayas to an eleva- tion of upwards of 10,000 feet, and even during the winter it is said to dwell habitually in the snow-clad forests about Simla. A remarkable and somewhat aberrant form of this group, the black macaque (M. niger), whose relationship with the African baboons is more intimate than that of any of the other species, inhabits Celebes (and Batchian ?) ; it is frequently recognised as the type of a distinct genus, Cynopithecus. The remaining types of Old World monkeys are usually included in the genera Semnopithecus and Colobus, constituting the sub- family SemnopithecinaB, the former of which, with probably not less than twenty-five species, are. exclusively Asiatic, and the latter, considerably less numerous, African. Of the genus Semnopithecus, OLD WORLD MONKEYS. 399 whose distributional area extends from Ceylon and the snow-bound heights of Thibet (S. Roxellanse) to the islands of the Malay Archi- pelago, which properly constitute its headquarters, the better- known species are the sacred entellus, or hoonuman (S. entellus), from the Gangetic provinces, and the proboscis monkey (S. nasi- cus) of Borneo. A form related to the last, characterised by an excessively up-turned nose, is the proboscis monkey of Thibet above mentioned (S. Roxellanse). The African Colobi are slender, long-tailed monkeys like the Semnopitheci, from which they are barely separable, but differ in the complete absence of the thumb. Of probably not more than a dozen species, whose combined habitat embraces the greater part of the African continent, from the west coast to Abyssinia and Zanzibar, the best known, and, at the same time, probably the most graceful and beautiful of all monkeys, is the guereza (C. gue- reza), whose home appears to be the highlands of Abyssinia, at elevations of from 7,000 to 10,000 feet. A closely related form is Colobus Angolensis. The total number of apes inhabiting the islands of the Malay Archipelago is, according to Rosenberg,133 twenty-five, distributed among the different islands as follows : Sumatra, twelve ; Banca, four; Borneo, eleven ; Java, five; Celebes, two; and Bali, Lombok, Flores, Sumbawa, and Timor, each one. The rapid diminution in the direction of the Australian continent, which is entirely deficient in the animals of this class, is very marked. Only one form, the common macaque, is common to all the islands; Sumatra holds only one species in common with Java, whereas, surprisingly enough, four of its species are represented in Borneo. The greater number of the species are restricted to individual islands. No unequivocal remains of true monkeys are known to antedate the Miocene period, and in America they do not appear before the late Pliocene or Fost-Pliocene. Several forms, referred to the South American genera Cebus, Callithrix, and Hapale, and one representing an extinct type, Protopithecus, probably allied to the howlers, have been described by Lund from the cavern deposits of Brazil. Protopithecus Bonseriensis is founded upon a number of incisor teeth obtained by Ameghino in the neighbourhood of the city of Buenos Ayres. No quadrumanous remains other than those 400 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. referable to the lemurs, or to a type, Laopithecus (Miocene), stand- ing intermediate between these and the Cebidse, have as yet been discovered on the North American continent. To the extent of our present knowledge, therefore, the type of the Old World mon- keys appears to have had no representatives in the Western Hemi- sphere. Numerous remains of Quadrumana are found in the Tertiary de- posits of France, Germany, Switzerland, Italy, and Greece, indicat- ing for these animals a much broader distribution in past periods of time than they now enjoy. Several of these forms are referred to existing genera, such as the Miocene Colobus grandsevus, from Steinheim, Wtirtemberg, and the Pliocene Macacus priscus and Semnopithecus Monspessulanus, from Montpellier, France. A macaque (Macacus Pliocasnus) has also been cited from Essex, Eng- land, and several forms have been described from the Val d'Arno, Italy. But the greater number of the more ancient species still remain with undetermined relationships. One of the most remark- able of these is the Dryopithecus Fontani, from the Middle Miocene deposits of St. Gaudens, France, and the Swabian Alps, which in stature appears to have rivalled the largest of the existing anthro- poid apes, although probably more nearly related to the gibbon than to any other living member of this group. Two other appar- ently anthropoid forms of somewhat smaller dimensions have been described from the nearly equivalent deposits of Sansan, France, and Elgg, in the Canton of Zurich, Switzerland, as Pliopithecus an- tiquus and P. platyodon respectively ; these are by some authors considered to be more nearly related to the group of the Semno- pithecina3, or even to the macaques, while by others, as Rutimeyer and Lydekker, they are referred to the modern genus Hylobates. Of still more doubtful relationship are the singular Mesopithecus Pentelici, from the Mio-Pliocene of Pikermi, Greece, which in its cranial and dental features most nearly approaches Semnopithecus, while in the structure of the limbs it approximates the macaques, and the probably still less simian Oreopithecus Bambolii, from Monte Bamboli, Tuscany. Several species of fossil ape have been described from the Siwa- lik Hills (Pliocene), and are by Lydekker referred to the genera Palseopithecus, Semnopithecus, Macacus, and Cynocephalus. If the determination in the case of the last-named genus be cor- LEMURS. 401 rectly made, it is interesting as proving the former much further extension of the baboons than we find at the present time. Cyno- cephalus Atlanticus occurs in the late Pliocene or Post-Pliocene de- posits of Algeria. The lemurs, or half-monkeys, constitute a well-differentiated group of the Primates, differing, indeed, in so many essential points of structure from the type of this order as to have induced many naturalists to elevate them to an order apart by themselves, the Lemuroidea or Prosimise. Their non-simian affinities are at once with the Insectivora and Ungulata, to which they appear to -be united by many connecting ties, both recent and fossil. Upwards of fifty more or less clearly defined (recent) species, representing a dozen or more genera, have been referred to this group, more than one-half of which, embracing all the typical lemurs of the genera Lemur (some fifteen species), Hapaleinur, and Lepilemur, are abso- lutely confined to the island of Madagascar, where they inhabit the forest region. The indrises (Indris, Propithecus), which are like- wise confined to the Madagascan region, comprise some of the largest of the lemurs, Indris brevicaudatus measuring upwards of two feet in length. The sub-family of the galagos numbers probably not less than twenty species, distributed under the two genera Chirogaleus and Galago, the former of which is restricted to Madagascar, while the latter inhabits the scattered wooded tracts of the interior of the African continent, from Senegambia to Abyssinia, and southward to Natal ; no species is known from Madagascar. All the Asiatic lemurs, if we except the very remarkable tarsier (Tarsius spectrum), the type of a distinct family, which inhabits some of the larger isl- ands of the Malay Archipelago (Sumatra, Borneo, Celebes) and the Philippines, and which differs from the lemurs proper, apart from other general characters, in the large eyes and unusual elongation of the tarsal elements of the foot, belong to the sub-family of the lories. They constitute a limited group of small nocturnal animals, desti- tute of a tail, and distinguished, as far as their habits are concerned, by their exceedingly slow movements. Hence they are frequently termed the ''slow lemurs." Of the two Oriental genera, Nyctice- bus, the typical slow-lemur, is distributed over Cochin China, Siam, the Malay Peninsula, and the larger islands of the adjoining 402 GEOGRAPHICAL AXD GEOLOGICAL DISTRIBUTION. archipelagos — Sumatra, Java, Borneo. Loris, with a single spe- cies, the graceful loris (L. gracilis), is a native of Ceylon. Two other species of the same sub-family, resembling the last in their habits, but provided with a rudimentary tail, and with a greatly reduced index-finger, inhabit the west coast of Africa : the potto (Perodicticus potto) is a native of the Gaboon region and the ter- ritory of Sierra Leone, and the awantibo (P. [Arctocebus] Calaba- rcnsis) of Old Calabar. The most aberrant form of lemur is the Madagascan aye-aye (Chiromys Madagascariensis), an animal of about the size of a cat, with a rodent-like dentition, and singularly elongated fingers fur- nished with pointed claws. For a long time the position of this remarkable animal was misunderstood, it having been placed alter- nately with the lemurs, insectivores, and rodents. It constitutes the type of a distinct family, ChiromyidaB. The somewhat anomalous distribution of this group of animals, taken as a whole — their headquarters in Madagascar, with a thin- ning out towards the west on the African continent, and their re- appearance in Ceylon and the mainland of Asia — has suggested to some naturalists the notion that at a former, and fairly ancient, pe- riod of the earth's history direct land connection existed between these various points, bridging over the chasms that now separate them in the way of water, and permitting of ready migration from one region to another. For this hypothetically assumed, now sunken, continent, Mr. Sclater has proposed the name "Lemuria." In how far such a connecting land-mass may have existed in fact, or in how far, if it actually existed, it was directly concerned with the present distribution of the lemurs, still remains to be deter- mined. The earliest lemuroid remains of the Old World are probably those of Caenopithecus lernuroides, described by Rutimeyer from the Eocene deposits of Egerkingen, Jura Mountains, and supposed by their discoverer to represent an animal of intermediate relation- ships between the true lemurs and the American monkeys. This form is not unlikely identical with a species described from the gypsum deposits of the Paris Basin and the phosphorites of Quercy (Oligocene) as Adapis Parisiensis — for a long time supposed to represent an ungulate — with which, also, the PalaBolemur Betillci, LEMURS. 403 from Beduer, France, has been identified. The animals here re- ferred to appear to have their nearest analogues among the African lories or galagos. From certain peculiarities in the structure of the cranium, which are supposed to represent similar structures seen in the Ungulata, Filhol recognises in these animals an extinct zoological type, designated the Pachylemur, which stands inter- mediate between the true lemurs and the pachyderms. Necrolemur antiquus and N. Edwardsi, on the other hand, from the phospho- rites of Quercy, are considered to be true lemurs. Lemuroid forms do not appear to be represented in any of the Tertiary formations newer than the Lower Miocene or Oligocene. Numerous forms, referable to the same group of animals, have been described from the Lower Tertiaries of the Western United States (Lemuravus, Limnotherium, Microsyops, Hyopsodus, Mixo- dectes, Anaptomorphus, &c.). These in part indicate a transition to the hoofed animals, while others, again, are so closely linked with the Insectivora that they are barely, if at all, separable from them. A reference to some of these forms will be found in the section following the discussion of the Insectivora. As in Europe, no lemuroid forms are known from either the American Miocene or Fliocene formations. EEFEEENCE NOTES. 1. On the authority of Wallace. It would appear, however, from the observations of Taczanowski (" Ornithologie du Perou," vol. i, p. 321, 1884), that the bird is less rare in the region indicated than has been generally supposed. 2. Jeittcles, " Verhandl. d. zool. bot. Gesell. Wien," 1862, p. 262. 3. To this list might also be added the chipmunk, Arctic hare, lynx, wolf, walrus, several seals, &c. 4. Allen, " North American Pinnipeds," pp. 609 et seq. 6. Allen, " North American Rodentia," " U. S. Geol. Survey," vol. xi. 6. " Monograph of the Strepomatidae," p. xli, Smithson. Misc. Pub., 253. 7. " Island Life," pp. 20-22. 8. Seebohm, " Catalogue of Birds," British Museum, v, p. 328. 8a. Since the preparation of the text a large number of additional species, and several genera, of paradise-birds have been described from New Guinea by Finsch, Meyer, Forbes, and others. 9. The family comprises some thirty-Sve or more species (Beddomc, "Ann. Mag. Nat. History," January, 1886). 10. Gray, "Catalogue of Edentate Mammalia," British Museum, 1869, p. 389. 11. Brehm, " Thierleben," i, p. 391. 12. " Ceylon," ii, p. 287. 12a. "Proc. Zool. Soc.," London, pp. 221, 222. 13. Newton, "Encycl. Brit.," article "Humming-Bird," ninth ed., xii, p. 359. 14. Mosenthal and Hartinjr, " Ostriches and Ostrich Farming," p. 28. 15. " Geograph. Distrib. of Animals," ii, p. 330. 16. Lyell, " Principles of Geology," eleventh ed., ii, p. 309. 17. " Encycl. Brit,," ninth ed., iii, p. 461. 18. Lyell, " Principles of Geology," eleventh ed., ii, p. 366. REFERENCE NOTES. 405 19. Baird, "Am. Journ. Science," 1866 (xli), p. 340. 20. Baird, "Am. Journ. Science," 1866 (xli), p. 346. 21. Sharpe, "Catalogue of Birds," British Museum, ii, p. 238. 22. Wallace, " Island Life," p. 296. 23. Wallace, " Island Life," p. 253. 24. Wallace, " Geographical Distribution of Animals," i, p. 32. The occurrence was noted by Mr. Lowe, who communicated the facts to Sir Charles Lyell. 25. Wallace, "Geograph. Distrib.," i, p. 180. 26. Wallace, " Geograph. Distrib.," ii, p. 13. 27. Coues, " Key to North American Birds," p. 317. 28. Gould, " Birds of Australia," ii, p. 1. 29. " Ibis," 1884, p. 471. Two hundred sheep are said to hare been killed in a single night by a flock of these birds, at a station on Wanaka Lake. 30. " Proc. Zool. Soc.," London, 1864, p. 456. 31. Elwes, "Proc. Zool. Soc.," 1873, p. 648. 32. Forsyth Major, " Zoogeographische Ubergangsregionen," " Kosmos," 18~84, p. 106. 33. Forsyth Major, " Kosmos," 1884, p. 109. Forsyth Major is in error in quoting from Bottgcr that twenty-seven out of the forty reptilian and am- phibian species inhabiting Morocco are also found in Spain ; the number stated is twenty-two (" Abhandl. d. Senckenb. Naturf.-Gesellsch.," xiii, 1863, p. 146). 34. Cope, " Bull. U. S. National Museum," 1875 ; Heilprin, " Proc. Acad. Nat. Sciences," Philadelphia, 1882. 35. Moseley, " * Challenger » Reports," "Zoology," ii, pp. 188, 189. 36. Wyville Thomson, " Depths of the Sea," p. 454. 37. Lyman, " ' Challenger ' Reports," " Zoology," v, p. 327. 38. A. Agassiz, " ' Challenger ' Reports," " Zoology," iii, p. 30. 39. A. Agassiz, op. oil. 40. Milne-Edwards, "Contes Rendus," December 17, 1883. 41. Wyville Thomson, " Voyage of the ' Challenger,' » ii, p. 350. 42. " Nature," March 20, 1884. 43. Gunther, " Study of Fishes," p. 305. 44. Wyville Thomson, " Voyage of the ' Challenger,' " ii, pp. 352, 353. 45. "Annals and Magazine of Natural History," January, 1883; "Ver- handl. d. k. k. geol. Reichsanstalt," 1882, No. 4. 46. " Nature," xxvi, p. 560. 47. " Bull. Museum Comp. Zoology," Cambridge, vi, p. 153. 48. " Nature," September 3, 1885. 406 REFERENCE NOTES. 49. "Bull. Soc. Vaud.," xiv, p. 211, 1876. 50. " Kendic. R. Istit. Lomb.," ser. 2, xii, p. 694. 50a. Asper and Ileuscher have quite recently shown, through the use of a " pelagic " net, that the pelagic faunas of lakes are far more prolific in mi- croscopic animal forms than has hitherto been supposed. A drop of water from Lake Zurich was estimated to contain ten individuals of Anurasa foli- acea, eight of Anuraea longispina, sixty of Ceratium hirundinella, and mill- ions of Einobryon forms and Asterionellse, besides various heliozoans, roti- fers, and crustaceans. Identical results were obtained under all the most varied conditions of light (darkness) and water, in the open lake and along the shallower shore-line (" Zoologischer Anzciger," June 19, 1886). 61. "Nature," June 11, 1885. 62. "Am. Journ. Science," 1871, p. 161. 53. " Bull. Soc. Vaud.," xiii, xiv, 1874, 1876. Dr. Henri Blanc enumer- ates the following twelve species of Rhizopoda as entering into the compo- sition of the deep-water fauna of Lake Geneva (seventy to one hundred and twenty metres) : Amoeba proteus, A. verrucosa, A. radiosa, Difflugia pyri- formis, D. urceolata, D. globulosa, Ilyalosphenia cuneata, Arcella vulgaris, Centropyxis aculeata, Pamphagus hyalinus, Actinophrys sol, and an unde- termined large Difflugia. All or most of the above forms have been ob- served by Leidy in the surface-waters of the United States, and it is re- marked that the species indicated to be rare by Leidy are also rare in the deep waters of the lake (" Bull. Soc. Vaud.," scr. 2, xx, p. 287). 54. "Am. Journ. Science," 1871. 55. "Anniversary Address, London Geol. Soc.," 1881. 55a. Probably the most striking and convincing evidence indicating con- vergent modification is presented by the Australian fauna, where, among the numerous implacental forms, we have such remarkable reproductions of the distinctive types seen among the Placentalia, aithcugh based upon an entirely different type of structure, and arising independently of the other. t 56. " Paleontographical Soc. Reports," 1884. 57. "Ann. Mag. Nat. Hist.," 1874, xiii, p. 222. 58. Heilprin, "Proc. Acad. Nat. Sci.," Philadelphia, March 4, 1884. 59. "Anniversary Address, London Geol. Soc.," 1881. 60. Medlicott and Blanford, " Geology of India," part i, p. 282. 61. Seelcy, " Q. Journ. Geol. Soc.," London, 1883. 62. Dawson, " Am. Journ. Science," third ser., xx, pp. 403 et scq. 63. Anodonta Jukesii, from the Old Red Sandstone of Ireland. Pro- fessor Hall recognises in the Cypricardites Catskillensis of Vanuxem, from the Oneonta Sandstones of the State of New York (Middle Devonian), a REFERENCE NOTES. 407 fresh-water mussel having the general characters of Anodonta (" Science," New York, December 11, 1880). 64. " Review of the Non-Marine Fossil Mollusca of North America," " U. S. Gcol. Surv.," third annual report, 1881-'82. 65. Smith, "Proc. Zool. Soc.," London, 1881 ; Crosse, "Journal de Con- chyliologie," 1881 ; Bourguignat, " Mollusques Terr, et Flur. du Lac Tan- ganyika," August, 1885. 66. Marshall, "Ann. Mag. Nat. History," December, 1883. 67. Richthofcn, " China," iv. 68. " Q. Journ. Geol. Soc.," London, 1878, pp. 568 et seq. 69. " Manuel de Conchyliologie," p. 144 (1881). 70. Tryon, " Structural and Systematic Conchology," i, p. 159. 71. Carpenter, on the authority of Fischer, op. cit., p. 168. 72. " Palaeontologia Indica," ser. ix, 1875. 73. " Palaeontologia Indica," 1865. 74. " Kreidebildungen von Texas," 1852. 75. "Zeitschr. d. dcutsch. geol. Ges.," 1870, pp. 191-251. 76. " Jahrb. d. k. k. geol. Reichsanstalt," 1871, p. 524. 77. " Palaeontologia Indica," " Cephalopoda of Kutch," ser. ix, p. 237. 78. " Kreideb. v. Texas," pp. 22-25. 79. " Smithsonian Miscell. Pub.," vii. 80. " Jahrb. d. k. k. geol. Reichsanstalt," 1878, p. 44. 81. "British Assoc. Reports," 1884, p. 555. 82. "Anniversary Address, Gcol. Soc.," London, 1862, p. xiv. 83. Bronn's " Klassen und Ordnungen d. Thier-Reichs, Protozoa." 84. From the data furnished by Brady, " ' Challenger ' Reports," " Zool- ogy," ix. 85. From the data furnished by Brady, " ' Challenger ' Reports," " Zool- ogy," ix. 86. " < Challenger ' Reports," "Zoology," vi, p. 132. 87. " ' Challenger ' Reports," " Zoology," ii, " Corals," pp. 132, 133. 88. Duncan, " Q. Journ. Geol. Soc.," London, xxvii, p. 437. 89. Duncan, " Q. Journ. Geol. Soc.," London, xxix, p. 561. 90. Duncan, " Q, Journ. Geol. Soc.," London, xxvi, p. 313. 91. " Denkschr. d. k. k. Akad.," Vienna, 1872, p. 199. 92. "Q. Journ. Geol. Soc,," London, xxvi, p. 311. 93. " Q. Journ. Geol. Soc.," London, xxxii, p. 343. 94. Zittel, " Handbuch der Palaeontologie," i, pp. 717-720. 95. " Manuel de Conchyliologie," pp. 173 et seq. 96. Ellsworth Call, " Bull. U. S. Geol. Survey," xi (1884), p. 43. 97. Thesaurus Siluricus ; Thesaurus Devonico-Carboniferus. 408 REFERENCE HOTES. 98. " Handbuch der Palaeontologie," ii, p. 320. 99. " Encycl. Brit.," ninth ed., vi, p. 663 (" Crustacea"). 100. Bronn's " Klassen und Ordnungen des Thier-Reichs," v, p. 1073. 101. Hall, " Pennsylvania Second Geol. Survey," 1884, PPP, p. 29. 102. McLachlan, in Nares's "Voyage to the Polar Sea," ii, p. 234. 103. " Trans. Asiatic Soc. of Japan," November, 1885. 104. " Nature," December 24, 1885. 105. " Mem. Boston Soc. Nat. History," April, 1885. 106. Scuddcr, " Mem. Boston Soc. Nat. History," April, 1885, p. 355. 107. " Nature," January 29, 1885, p. 297. 108. " Voyage to the Polar Sea," ii, p. 1220. 109. "Study of Fishes," pp. 307-311. 110. Giinther, " Study of Fishes," pp. 270, 271. 111. "Science," May 23, 1884. 112. Filhol, in "Science," May 23, 1884. 113. "Science," May 23, 18S4. 114. "American Naturalist," March, 1885, p. 289. 115. " Journ. Asiatic Soc. Bengal," 1864, p. 544. 116. Leydig, "Abhandl. d. Senckenb. Naturf.-Gesell," xiii, 1883. 117. Bottger, "Abhandl. d. Senckenb. Naturf.-Gesell.," xiii, 1883. 118. Bosca, " Bull. Soc. Zool. de France," 1880. 119. Bedriaga, "Bull. Soc. Natural.," Moscow, 56, 1881. 120. "Ann. Mag. Nat. Hist.," January, 1886. 121. Woodward, " Gcol. Magazine," November, 1885. 122. " Palaeontologia Indica," ser. x, iii, p. 146. 123. " Q. Journ. Geol. Soc.," London, August, 1885. 124. "Encycl. Brit.," ninth ed., xv, p. 399. 125. Dobson, " Monograph of the Insectivora," 1882. 126. " American Naturalist," May, 1885, p. 467. 127. Dobson, " Catalogue of the Cheiroptera, British Museum," p. xxx. 128. "Am. Journ. Science," April, 1886. 129. " Zoologischer Anzciger," June, 1883. 130. " Q. Journ. Geol. Soc.," London, 1878, p. 419. 131. "Trans. Am. Philos. Society," January 18, 1884. 132. "Proc. Zool. Soc.," London, p. 443. 133. "Zoologischer Garten," xxiii (1882), pp. 111-115. INDEX. Aard-wolf, 383. Abyssal zone, 262. Acanthomys, 355. Accipenser, 68, 69. Aceratheiium, 368. Acervularia, 249. Achoenodon, 374. Achntinella, 261. Acidaspis, 146. Acomys, 364. Acris, 309. Acrobata, 99. Acrodus, 300. Acrosaurus, 320. Acrotreta, 255. Actaxmidae, 166. Actinia, 240, 242. Actinocrinus, 151. Adapis, 348, 402. Addax, 377. Adder, 321. Adocus, 315. ^Egoceras, 166, 267. ^Elurodon, 386, 388. J21 .rogale, 385. ^Epyornis, 333. JEtobatis, 300. Agama, 316, 318. A^aricia, 248. Agelaius, 66. Agnostus, 140, 276-278. Agouti, 361. Agriochoerus, 383. Ailuropus, 389. Ailurus, 389. Alactaga, 358. Alaudidae, 66. Alca, 69. Alcedinidse, 66. Alces, 381. Alcyonaria, 251. Alepocephalus, 298. Alctornis, 331. Alk, 69. Alligators, distribution of, 328, 329. Alligator-gar (Lepidosteus), 301. Allorisma, 271. Allosaurus, 162. Alpaca, 375. Alveolites, 143. Amadine, 88. Araaltheus, 267. Amblypoda, 367. Amblypterus, 302. Amblyrhynchus, 317. Amblystoma, 306. Amia, 68, 301. Amiurus, 68. Ammodiscus, 240. Ammonites, 137, 166, 192, 200, 222- 224, 266, 267. Amphibia, 45, 305. Amphibos, 378. Amphicyon, 388, 389. 410 IlSTDEX. Amphioxus, 299. Amphipoda, 272. AmphLsaurus, 160. Amphisbaenia, 319. Amphitherium, 335. Amphitragulus, 381. Amphiuraa, 300, 311. Amplexus, 151. Amynodon, 368. Ananchytes, 168. Anaptomorphus, 349, 403. Anatinidae, 168. Anchitheriura, 371, 372. Anchovy (Engraulis), 295. Ancistrodon, 323. Ancyloceras, 168, 267. Ancylotherium, 333. Ancylus, 261. Andrcnidse, 234. Anemone, 242. Angler (Lophius), 294, 295. Annelida (deep-sea), 109. Anoa, 378. Anodonta, 148, 209. Anolis, 317. Anomia, 211. Anomodontia, 159. Anoplotheriurn, 382. Ant, 283. Ant-eater, 337. Antechinus, 99. Antedon, 136. Antelopes, 84, 377, 3*8. Anthracomartus, 151, 286. Anthracopupa, 150, 208. Anthracotherium, 375. Anthropoid apes, 396. Antilocapra, 377. Antipathidae, 242. Antrozous, 351. Anura, distribution of, 307. Apatornis, 330. Apes, 396. Aphanapteryx, 333. Aphelops, 368. Apidae, 284. Apiocrinidae, 111. Apiocrinus, 166. Aploccrus, 377. Aporrhaidoe, 166. Apteryx, 29, 101. Aptornis, 333. Apus, 207, 273. Aquila, 79. Aragarie, 78. Arachnactis, 242. Arachnecthra, 102. Arachnids (of lakes'), 131 ; geol. dis- tribution of, 285. Arcadae, 148, 263, 271, 272. Arcestes, 156, 267. Archaegosaurus, 310. Archaeocidaris, 151. Archaeomys, 364. Archseopteryx, 330. Archaeospherina, 134. Archipolypoda, 286. Arctoccbus, 402. Aretocephalus, 391. Arctogale, 92. Arctomys, 359, 364. Arctonyx, 97. Arctotherium, 389. Arechnothera, 94, 102. Argiope, 255. Argonauta, 20, 122, 226. Argus, 95, 97. Argynnis, 280, 281. Arietites, 166.. Arius, 104. Armadillo, 5, 24, 337- Aromochelys, 314. Artamidae, 104. Artemia, 212. Arvicola, 356, 365. Asaphus, 146, 277. Aspidonectes, 314. Ass, 370. Assiminea, 261. Astacus, 207, 275. INDEX. 411 Astartidse, 148, 166, 271, 272. Asteroidea (deep-sea), 111. Asterolepis, 302. Astrsea, 247-249. Astweidse, 144. Astrangia, 248. Atalapha, 349, 351. Atax, 126. Ateles, 394-396. Atherines, 104. Atherura, 362. Athyria, 145. Atlanta, 120, 122. Atlantosaurus, 161. Atrypa, 145, 214. Auchenia, 375, 376. Aurelia, 122. Aurochs, 379. Australian fauna, anomalies of, Australian realm, 97. Aviculidae, 148, 271. Axis, 381. Axolotl, 306. Aye-aye, 402. Azores, birds of the, 7, 48. Babbling-thrush, 72, 94, 107. Baboon, 397. Babyrousa, 374. Baculites, 168, 267. Badger, 390. Baikal, Lake, fauna of, 212. Bairdia, 207, 273, 274. Balana, 341. Balffinoptera, 341, 344. Batenotus, 345. Baltimore bird, 66, 79. Bandicoot, 99. Barbary ape, 394, 395. Barbel, 68, 291. Barbus, 68. Barracuda (Sphyrsena), 294. Barramunda, 103. Earners affecting migration, 41. Bascanium, 323, 324. Basilisk, 317. Bass (Labrax), 294. Bassalian province, 299. Bassaris, 390. Bathmoceras, 192. Bathyactis, 110, 243. Bathycrinus, 111. Bathyergus, 357. Bathygnathus, 160. Bathyophis, 298. Bats, distribution of, 349. Battocrinus, 151. Bear, 4, 27, 389. Beaver, 360, 364, 365. Bee-eater, 87, 94, 106. Bees, 279, 284. Beetles, 150, 279, 280, 283, 284. Belernnitella, 168. Belemnites, 137, 156, 168, 268. Belemuosepia, 138, 166. Belideus, 99, 100. Bell-bird, 78. Belliuurus, 278, 279. Bellows-fish (Centriscus), 295. Beloteuthis, 166. Beluga, 344. Bermudas, birds of the, 48, 51. Bernissartia, 329. Beroe, 240, Beyrichia, 272. Beryx, 294, 303. Bibos, 379. Big-horn, 379. Birds, American, common to Europe, 46 ; European, common to Amer- ica, 48 ; dispersal of, 45 ; migra- tions of, 47 ; oceanic journeys of, 47 ; Arctic, 70 ; geological dis- tribution of, 330 ; of the Holarctic realm : of the Eurasiatic division, 65; of the North American, 66; of the Neotropical realm, 77-79 ; of the Ethiopian realm, 87, 88 ; of the Oriental realm, 94, 95 ; of the Australian realm, 100-102; 412 IXDEX. of the Polynesian realm, 104 ; of the Mediterranean region, 105, 100. Bison, 379. Blarina, 346. Blastocerus, 380. Bluttarioe, 283. Blenny (Blennius), 293. Blcssbok, 84. Blind-worm (Anguis), 315, 320. Blow-fish (Chsetodon), 296. Blue-bird, 28, 66. Boa, 325-327. Boar, 374. Boat-bill, 81. Boavus, 327. Bobolink, 66. Bombinator, 312. Bonasa, 66. Bonito, 297. Bos, 379. Bothriolepis, 302. Bower-bird, 101. Brachiopoda, age of, 138 ; deep-sea, 113 ; distribution of, 214, 252. Brachymetopus, 151, 277. Brachyops, 311. Brachypyge, 275. Brachyura, 112, 275. Bradypus, 337. Brahmathcrium, 377. Branchiosaurus, 310. Branchipus, 212. Breaks, geological, 192. Brisinga, 111. Brittle-stars, 111,166. Broad-bill, 95. Brontosaurus, 161. Brontotherium, 372. Brush-turkey, 95, 101. Buansuah, 387. Bubalus, 378. Bucapra, 380. Buccinidse, 169. Buccinum, 262. Buffalo, 378. Buffelus, 378. Bufo, 307, 312. Bugs, 279, 280, 284. Bulbul, 94, 107. Buiimus, 261. Bull-finch, 65, 70. Bull-head (Cottus), 68, 287, 293, 294. Bungarus, 322, 325. Bunodonta, 373. Buuting, 66, 70. Buprestites, 283. Burbot (Lota), 288. Bushbok, 84, 89. Bustard, 66, 72. Butterflies, 279, 280, 284 ; Arctic, 51, 70 ; on Chimborazo, 51. Buzzard, 66. Cacatua, 1C2. Cacatuidse, 88, 102. Cachalot, 342. Cselogenys, 361. Caenopithecus, 402. Cainotherium, 382. Calamary, 268. Calamoichthys, 89, 291. Calliste, 78. Callithrix, 395, 399. Callophoca, 393. Callorhinus, 391. Calostylis, 144. Calymene, 146, 278. Calyptorhynchus, 102. Camarasaurus, 169. Cambrian fauna, 135. Camel, 15, 16, 31. 37, 375, 376. Camelopardalis, 376, 377. Campephagidse, .104. Camptonotus, 1G2. Canace, 66. Cancellaridae, 169, 270. Cancroma, 81. Canis, 386-388. Capercaillie, 66. INDEX. 413 Capito, 81. Capra, 380. Capreolus, 381. Caprina, 169. Capromys, 360. Caprotina, 169. Capulus, 138, 207, 269. Capybara, 361, 365. Carabidae, 69, 281, 284. Caracal, 384. Carboniferous fauna, 150. Carcharias, 297, 300. Carcharodon, 297, 300. Cardiid®, 148, 166, 271, 272. Cariacus, 381. Cariama, 88. Carinaria, 122. Carnivora, distribution of, 383. Carp, 68, 89, 290. Caryophyllia, 243, 244. Cassicus, 79. Cassididse, 270. Cassidulina, 235. Cassowary, 101, 332. Castor, 360, 364, 365. Castoroides, 365. Cat, 27, 383. Catarhina, 393. Caterpillar-eater, 72, 94, 104. Cat-fishes (Siluridae), 68, 80, 89, 288- 290. ' Cathartinse, 66. Catopterus, 302. Catostomus, 68. Cattle, 379. Caulaster, 136. Cave-bear, 389. Cave-lion, 384. Cavy (Cavia), 361. Cebus, 395, 399. Ccntetes, 346. Centetodon, 348. Centipedes, 285, 286. Centrarcbldse, 68. Centrocercus, 66. Cephalaspis, 149, 301. Cephalopoda, deep-sea, 113 ; geologi- cal distribution, 265. (See MOL- LUSCA.) Cephalophus, 89. Cephalopterus, 78. Ceratiocaris, 273. Ceratites, 267. Ceratodus, 103, 291, 302, 304. Ceratophrys, 310, 312. Ceratoptera, 297. Ceratorhinus, 368. Cercocebus, 397. Cercolabes, 362. Cercoleptes, 390. Cercopithecus, 397. Cerianthus, 242. Ceriornis, 66. Cerithiidae, 166. Certhiadse, 107. Certhiola, 78. Cervalces, 382. Cervulus, 381. Cervus, 380, 381. Ceryle, 106. Cetacea, distribution of, 341. Cetiosaurus, 161. Cetotherium, 345. Chacma, 398. Chaffinch, 70. Chalicotherium, 372. Chalk, nature of, 237. Chamsea, 73. Chameleon, 316, 318. Chamidse, 272. Chamois, 21, 377. Characinidfe, 80, 290. Charina, 322. Chasmorhynchus, 78. Chatterer, 78. Cheetah, 384. Cheiroptera, distribution of, 349. Cheirurus, 146. Chelone, 313, 315. Chelonia, distribution of, 313. 414 INDEX. Chclopus, 314. Chelydse, 314. Clielydra, 313-315. Cbevrotain, 376. Chilabothrus, 326. Ciiimoera, 293-295. Chimpanzee, 397. Chinchilla, 361. Chipmunk, 359. Chirogaleus, 401. Chirolepis, 149. Chiromys, 402. Chironectes, 334. Chironomidse, 283. Chiton, 207, 260, Chlamydosaurus, 318. Choenohyus, 374. Chceropotamus, 375. Choeropsis, 26. Chceropus, 99. Choloepus, 337. Chomatodus, 300. Choristoceras, 156, 268. Chromides, 290. Chrysaetos, 79. Chrysemys, 314. Chrysochloridae, 346. Chrysothrix, 395. Cicada, 284. Cidaridse (deep-sea), 111. Cidaris, 157, 166, 168. Cinosternum, 314. Cistudo, 313-315. Civet-cats, 385. Cladocera (of lakes), 126. Cladocora, 248. Cladophyllia, 247. Cleodora, 122. Clepsydrops, 154. Clepsysaurus, 160. Clidastes, 169. Climate, effect of, upon distribution, 35-40. Climatic zones (geological), 222. Cnemiornis, 333. Clotho, 88. Clydonites, 156. Clymenia, 156, 267. Clypeus, 166. Coassus, 380. Coati, 390. Cobitoids, 291. Cobitus, 68. Cobra (Naja), 322, 325. Coccosteus, 149, 301. Cochliodus, 300. Cochloceras, 156, 268. Cockatoo, 88, 95, 101. Cock-of-the-rock, 78. Cockroach, 147, 282, 283. Cod (Gadus), 287, 293, 294, 296, 2 Coecilia, distribution of, 306, 310. Coelacanthus, 301, 302. Coelodon, 339. Ccerebidse, 78. Coleoptera, 96, 279, 280, 283, 234. Colias, 280, 281. Collyrites, 166. Colobus, 399, 400. Colocollo, 383. Colonies, geological, 231. Colossochelys, 315. Colubers, 67, 322-324, 327. Columba, 102. Columbidae, 66. Colymbus, 69. Compsognathus, 162. Condor, 79, 80. Condor, Caliibrninn, 73 Condylarthra, 368. Condylura, 347. Coney, 33, 367. Conger, 294, 295. Conidse, 169, 270. Conocephalus, 146. Conocoryphe, 276. Conocyathus, 246. Conodonts, 299. Contia, 323. Conurus, 78, 88. 415 Copepoda (of lakes), 126. Copperhead, 323. Coral-fishes (Pomaccntridoe), 296. Coralline zone, 202. Coralliurn, 252. Coral-reefs, ancient, 249. Corals, deep-sea, 110, 242 ; distribu- tion of, 240. Coral- snake, 325. Corax, 300. Corbicula, 211. Corbula, 211. Cormorant (fossil), 330-332. Cornufer, 307. Coronella, 321, 325. Corsac, 388. Corvidae, 65. Corvus, 79. Coryphodon, 367. Corythaix, 87. Cotingidse, 78. Cottidae, 68. Cotton-rat, 356. Couguar, 19, 383. Cow-bird, 66. Coyote, 388. Coypn, 360. Crabs (deep-sea), 112. Cracidae, 78. Crane (fossil), 332, 333. Crania, 207, 255, 257. Craspedocephalus, 79. Creeper, 78, 107. Cremastosaurus, 320. Creodonta, 393. Cretaceous fauna, 168. Cricetodon, 365. Cricetornys, 357. Cricetus, 357, 364. Cricket, 284. Crinoidea, 135; deep-sea, 111. Crioceras, 168, 2C7. Cristellaria, 188. Crocidura, 346. Crocodiles, distribution of, 327, 328. Crossbill, 65. Crotalus, 322, 323. Crow, 65, 79, 104 ; fossil, 332. Crustacea, deep-sea, 112 ; pelagic, 120,122; of lakes, 126,131; dis- tribution of, 272. Cryptobranchus, 306, 311. Cryptoprocta, 385. Ctenacanthus, 152, 300. Cuckoo, 104. Cuckoo, ground, 73. Cuon, 387. Cupido, 66. Curassow. 78. Curculionites, 283. Curlew (fossil), 332. Cuscus, 100. Cyanea, 122. Cyathocrinus, 151. Cyathophylluin, 143, 151, 249. Cyclas, 264. Cyclolobus, 267. Cycloclypeus, 234. Cyclophis, 323. Cyclophthaluius, 151. Cycloturus, 337. Cynocephalus, 397, 400, 401. Cynogale, 92, 385. Cynomys, 359. Cynopithecus, 398. Cypraeidse, 169, 270. Cypridina, 207, 273. Cyprinidce, 68, 89. Cyprinodon, 80. Cypris, 274. Cypselidse, 66. Cyrtoceras, 139, 145, 190, 192, 265, 266. Cystignathidae, 309. Cystophora, 392. Cythere, 207, 273, 274. Dactylethridae, 310. Dakosaurus, 162, 329. Dalmania, 146, 199, 277, 278. 416 I^DEX. Dama, 3S1. Dupedius, 302. Dasornis, 331. Dasyprocta, 361. Dasypus, 337. Dasyuridae, 99. Dawsonella, 150, 208. Decapoda, 274. Deep-sea fauna, 109. Deep-sea zone, 262. Deer, 27, 380. Delphinapterus, 342. Delphinus, 342. Deltocyathus, 243-246. Dendrerpeton, 310. Den dro bates, 310. Dcndrocolaptidae, 78. Dendrodus, 301. Dendrohyrax, 33, 87. Dendrolagus, 99. Dendromys, 356. Dendrophidae, 325. Dcndrophyllia, 248. Dentaliuni, 207, 269. Dermatochelys, 313. Desmatotherium, 369. Dcsmognathus, 303. Devonian fauna, 146. Dhole, 387. Diaclectes, 154. Diadophis, 323. Diatryma, 331. DicsHdse, 94, 104. Diceras, 272. Dichobune, 382. Dichograptus, 145. Dicotyles, 374. Dicroccrus, 381. Dictyocaris, 273. Dicynodon, 159. Didelphys, 334, 335. Didunculidae, 104. Didymictis, 393. Didymograptus, 145. Diemyctilus, 306. Dikelocephalus, 146. Dimetrodon, 154. Dimorphodon, 164. Dinichthys, 149, 302. Dinictis, 385. Dinoceras, 367. Dinocyon, 389. Dinornis, 333. Dinosauria, 160, 161, 169, 172, 203. Dioplotherium, 340. Diplograptus, 145. Diploria, 247-249. Dipnoi. 302. Diprotodon, 336. Dipsadidse, 325. Diptera, 279, 280, 283, 284. Dipterus, 149, 301. Dipus, 358. Discina, 145, 207, 255, 257. Dischmca, 253. Discinocaris, 273. Discoidea, 168. Discrytus, 146, 282. Diver, 69. Dodo (Didus), 333. Dodo-pigeon, 104. Dog, 27, 386, 387. Dog-fish (Acanthias), 293-295, 300. Dolichonyx, 66. Dolichosoma, 310, 311. Dolichotis, 361. Dolium, 260. Dolphin, 342, 343. Dolphin (Cpryphaena), 297. Dormice, 357. Dorycrinus, 151. Douroucouli, 395. Draco, 318. Drepanidae, 104. Drepanodon, 384. Dromaeus, 101. Dromasornis, 333. Dromatherium, 335. Dromicia, 99. Dryiophidae, 325. INDEX. 417 Dryolestes, 335. Dryopithecus, 400. Duck (fossil), 332, 333. Duck-bill, 333. Dugong, 339. Dziggetai, 370. Eagle, 66, 79, 88 ; fossil, 332. Echidna, 333, 334. Echinidse (deep-sea), 111. Echinobrissus, 166. Echinoderms, 111, 166. Echinogale, 346. Echiomyidae, 360. Edentata, distribution of, 336. Edmondia, 271. Edwardsia, 242. Eels, 104, 292, 296, 303. Eider-duck, 69. Elachoceras, 367. Eland, 84. Elaphis, 321, 323, 324. Elaphodus, 72. Elapidse, 322, 325. Elasmobranchii, 299. Elasmognathus, 369. Elateridse, 284. Electric eel (Torpedo), 80, 294. Elephant. 18, 36, 37, 206; distribu- tion of, 365, 366. Elephant-shrews, 346. Elk, 381, 382. Emarginula, 269. Emballonura, 351. Emberiza, 66, 70. Empedocles, 154. Emu, 101 ; fossil, 332. Emydse (Emys), 314, 315. Enaliornis, 330. Encrinus, 157. Endoceras, 145, 190. Enhydris, 390. Entellus-monkey, 399. Entomodon, 348. Eobasileus, 367. 19 Eohippus, 371, 372. Eohyus, 374. Eoscorpius, 151, 286. Eotherium, 340. Eozoon, 134, 196. Equus, 369-371. Erethizon, 362. Erinaceus, 347. Eriodes, 395. Ermine, 4, 390. Erycidse, 322. Eryops, 155. Esocidse, 68. Esox, 69. Estheria, 207, 273. Esthonyx, 348. Estrild, 88. Ethiopian realm, 82. Euclastes, 315. Eugereon, 283. Eulabes, 94. Euinetopias, 391. Euomphalus, 148, 269. Euphoberia, 151, 286. Euphonia, 78. Euplectella, 110. Euplocamus, 97. Euproops, 278. Eupsammidse, 144. Eurasiatic region, fauna of, 60. Euryapteryx, 333. Euryla?midae, 95. Eurypterus, 148, 278, 279. Eurysternum, 315. Eusmilus, 385. Eutsmia, 323, 324. Evotomys, 656. Exogyra, 168. Extracrinus, 166. Falcon, 19, 66, 69, 104 ; fossil, 330. Fallow-deer, 380, 381. Families, distribution of, 29. Faunas, variation in, 3-6 ; migrations of, 13; origination of, 15, 178; 418 IKDEX. relation of past to present, 11, 12. Favia, 247. Favosites, 143, 215, 249. Fells, 383, 384. Felsinotherium, 340. Ferret, 390. Fiber, 357. Finch, 65, 88, 100. Fin-whale, 341. Fisher, 390. Fishes, distribution of, 287 ; shore, 293 ; pelagic, 297 ; deep-sea, 298 ; phylogeny of, 305 ; earliest, 142 ; of the Holarctic realm, 68 ; of the Neotropical realm, 80 ; of the Ethiopian realm, 88, 89 ; of the Australian realm, 103 ; of the Polynesian realm, 104. Fish-hawk, 19. Flabellum, 243-246. Flamingo, 31. 106 ; fossil, 331, 332. Flat-fishes, 294. Flies, 279, 280. Florissant, insects of, 284. Flounder, 294. Flower-pecker, 94, 104. Fly-catcher, 65, 79, 88, 100. Flying-foxes, 350, 352. Flying-garnard (Dactylopterus), 297. Flying- herring (Exocoetus), 297. Flying-lemurs, 345. Flying-lizard, 318. Flying-opossum, 99. Flying-squirrels, 358, 359. Foraminifera (deep-sea), 109 ; dis- tribution of, 234. Formicidse, 284. Fowl (Gallus, fossil), 332. Fox, 4, 387, 388. Francolin, 88, 106. Fringillidae, 66, 88, 100. Frog-fish, 297. Frogs and toads, dispersal of, 45 ; distribution of, 307, 312. Fulgorina, 283. Fundulus, 290. Fungia, 247-249. Fungiacyathus, 242. Furcifer, 380. Fusidae, 169. Fusulina, 236. Fusus, 262, 263, 269. Galago, 401. Galapagos, birds of, 6, 49. Galaxias, 288, 292. Galbulidae, 78. Galecynus, 388. Galeocerdo, 297, 300. Galeopithecus, 345. Galeospalax, 248. Galerites, 168. Galerix, 348. Galesaurus, 159. Galictis, 90. Galidia, 90. Gallus, 95. Gannet (fossil), 332. Ganocephala, 310. Ganoids, 291, 299, 300. Garden-mouse, 18. Gasteropoda, deep-sea, 113 ; pelagic, 120 ; geological distribution of, 268. (See MOLLUSCA.) Gasterosteidae, 68. Gastornis, 330, 331. Gavial, 328, 329. Gazelle, 84, 377, 378. Gecko, 316, 318. Gclacla, 398. Gemsbok, 84, 377. Genera, distribution of, 26. Genet, 385. Gcococcyx, 73. Geomys, 365. Geophaps, 102. Geophilus, 286. Geosaurus, 320. Gerbillus, 355. ItfDEX. 419 Gerephemera, 146, 282. Ghaur, 379. Gibbon, 3G7. Giraffe, 376, 377. Girvanella, 240. Glandaria, 188. Glass-snake, 316. Glaucus, 120. Glis, 357. Globicephalus, 342. Globigerina, 109, 188, 235, 236. Globigerina ooze, 236, 237. Glutton, 390. Glycimeridse, 168, 272. Glyptodon, 338, 339. < Glyptolepis, 140, 301.' Glyptosaurus, 320. Gnu, 84, 377. Goats. 28, 379, 380. Goatsucker, 104. Gobiesocidse, 295. Goby, 104, 292. God\vit (fossil). 330. Golden-niole, 346. Golden-pheasant, 66. Goldfinch, 65, 70. Gomphoceras, 145, 266. Goniastraea, 247. Goniatites, 137, 145, 267. Goniobasis, 69. Goose (fossil), 332, 333. Gopher, 365. Gorgonella, 251. Gorgonia, 251. Gorilla, 396. Goura, 102. Graculavus, 330. Grammysia, 271. Grampus, 343. Granatocrinus, 151. Graphularia, 251. Graptolites, 144, 208, 215. Greenland, insects of, 70. Griffithidcs, 151, 277. Grizzly-bear, 389. Grosbeak, 65. Ground-pigeon, 102. Ground-squirrel, 359, 365. Grouse, 66, 72 ; fossil, 330, 333. Gryphsea, 168. Gryphoca, 393. Guacharo, 18. Guan, 78. Guanaco, 375. Guenon, 397. Guereza, 399. Guevi, 84. Guinea-fowl, 88. Guinea-pig, 361. Gull (fossil), 331, 332. Gulo, 390. Gurnard (Trigla), 294. Guynia, 144, 240. Gymnodactylus, 316, 318. Gymnotus, 80. Gymnura, 347. Gypattus, 71. Gyps, 106. Gyrae-antlms, 152. Gyrichnites, 286. Gyroceras, 137, 266. Gyrodus, 302. Gyroporella, 237. Hadrosaurus, 169. Hag (Myxine), 294, 299. Hake (Merlucius), 294, 296. Halcampa, 242. Halcyon, 106. Haley orn is, 331. Haliaetus, 79. Halicore, 340. Halitherium, 340. Halobates, 120. Halysites, 143, 215, 249. Hamadryas, 398. Hamites, 168, 267. Hamster, 357. Hang-nest, 66, 79. Hapale, 396, 399. 420 I]STDEX. Hapalemur, 401. Haplodon, 360. Haplophlebium, 150. Haplophorus, 339. Haplophyllia, 144, 240. Hares, 21-23, 70, 362, 364, 365. Harlequin snake, 322. Harpoceras, 166, 267. Harpy, 79. Harrier (fossil), 331. Ilartebeest, 84. Hatteria, 320. Haw-finch, 65. Hawk, 66. Hedgehogs, 347. Hclagras, 327. Helaletes, 369. Helarctos, 369. Heliastraa, 247. Ilelicidoe, 69. Hclicoceras, 168. Heliolites, 143, 249. Helix, 261, 262. Helladotherium, 377. Heloderma, 317. Helodus, 300. Hcmiaspis, 278. Hemicaulodon, 340. Hemicidaris, 158, 166. Hcmidactylus, 316, 318. Hemiptera, 279, 280, 283, 284. Hermit-crabs (deep-sea), 112. Heron, 104; fossil, 331, 332. Herpestes, 385. Herring (Clupea). 293, 294, 296, 303. Hesperomys, 355, 365. Hesperornis, 330. Hcterodon, 90, 323, 324. Hetcrodryas, 90. Hill-tit, 94, 107. Hipparion, 371, 372. Hippidiutn, 371. Hippopotamus, 26, 373. Hippuntes, 169, 208. Hirundinidse, 65. Hoazin, 81. Hog, 374. Holarctic realm, 57, 73. Holbrookia, 317. Holectypus, 166. Holocystis, 144. Holopea, 148. Holops, 329. Holoptychius, 149, 301. Holothuroidea (deep-sea), 109, 112. Horuseosaurus, 320. llomotaxis, 227. Homothetus, 146, 282. Honey-guide, 87. Honey-sucker, 72, 87, 102, 104, 106, 107. Hoolock, 397. Hoonuman, 399. Hoopoe, 106. Hoplophoneus, 385. Horned-toad, 317. Horse, 369 ; extinction of, in Ameri- ca, 204, 206. Horse-mackerel (Caranx), 294, 296. Horseshoe bat, 353. Hortulia, 326. Hound (Mustelus), 294. Howler, 395, 396. Humming-birds, 17, 38, 73, 77. Humpback, 341. Hytemoschus, 376. Hya?narctos, 389. Hyaenodon, 393. Hyalonema; 110. Hybodus, 300. Hydrobia, 209. Hydrochoerus, 361, 365. Hydromys, 356. Ilydrophidffi, 321. Hydrophilites, 283. Hydropotes, 72. Hydrosaurus, 320. Hyena, 385, 386. Hyla, 307. Hylerpeton, 810. INDEX. Hylobates, 397, 400. Hylonomus, 310. Ilymcnocaris, 140, 273. Hymenoptera, 279, 280, 283, 284. Hyopotamus, 375. Hyopsodus, 348, 403. Hyotherium, 374. Hyperamina, 240. Hyperodapedon, 320. Hypodiadema, 158. Hyposaurus, 329. Hypsiprymnopsis, 334. Hypsiprymnus, 99. Hyrachyus, 369. Hyracoidea, 87, 367. llyracodon, 368. Hyracotherium, 371, 372. Hyrax, 33, 87. Hystrix, 362, 365. lantliina, 120, 122. Ibex, 28, 380. Ibis (fossil), 332. Iceland, insects of, 70. Icelus, 293. Ichneumon, 385. Ichthyornis, 330. Ichthyosauria, 162, 172. Icteridse, 66, 79. Icticyon, 386. Ictitherium, 385. Ictonyx, 390. Idiocetus, 345. Iguana, 317, 320. Iguanavus, 320. Iguanodon, 162, 169. Impeyan, 66, 95. Indicatoridse, 87. Indris, 401. Infusoria (pelagic), 124; of lakes, 127, 131. Inia, 341. Insectivora, distribution of, 345. Insects, dispersal of, 51 ; distribution of, 279 ; earliest, 146. lo, 69. Irrisoridse, 87. Isacis, 348. Isastrsea, 157. 247. Ischadites, 236. Ischypterus, 302. Ischyromys, 365. Island faunas, 6, 7. Isolation, effects of, on faunas. 6. Isophyllia, 249. Isopoda, 272. Issiodoromys, 364. Ithygrammodon, 376. Jacamar, 78. Jackal, 387. Jaguar, 383. Jaguar undi, 383. Jararaca, 79. Jays, 25, 26. Jerboa, 358. John Dory (Zeus), 294, 295. Juan Fernandez, birds of, 50. Jungle-fowl, 95. Jurassic fauna, 161. Kangaroo, 99, 336. Kangaroo-rat, 99. Keeling Islands, birds of the, 50. Kiang, 370. King- fisher, 66, 104, 106 ; fossil, 331. Kinglet, 66. Kinkajou, 390. Kite, 66 ; fossil, 332. Klipspringer, 84. Koala, 99. Kogia, 342. Koninckia, 257. Koodoo, 84, 377. Kutorgina, 255. Labyrinthodontia, 310. Lacerta, 315, 320. Lacertilia, distribution of, 315. Lachesis, 79. 422 INDEX. Lagena, 236, 240. Lagidium, 361. Lagomys, 362, 364. Lagopus, 66, 69, 70. Lagostomus, 361. Lagothrix, 39,", 396. Lake thuna, 126. Lamollibranchiata (deep-sea), 113 ; geological distribution of, 271. (See MOLLUSCA.) Laminarian zone, 262. Lammergeier, 71. Lamna, 297, 300. Lamprey, 68, 299. Lanitda?, 65. Laophis, 327. Laopithecus, 400. Laopteryx, 330. Laornis, 330. Laosaurus, 162. Lar, 397. Laramie formation, 203, 210. Lark, 66. Lark-bunting, 65. Latcs, 288, 289. Latimoeandra, 247. Laurentian (2) fauna, 134. Leaf-nosed bats, 352. Leda, 207. Leiodon, 169. Leipoa, 101. Leiotricludaj, 94, 107. Lemming (Myodes), 357. Lemuravus, 403, Lemuria, 87, 402. Lemuroidca, 393. Lemurs, distribution of, 401. Leopard, 18, 383, 38L Leperditia, 140, 272. Lepidoptera, 96, 279, 280, 284. Lepidosiren, 80, 302, 304. Lepidosternon, 319. Lepidosteus, 68. Lepidotus, 302. Lepilemur, 401. Leptictis, 348. ^ Leptobos, 378. Lcptolepis, 302., 303. Leptosomus (fossil), 331. Lepus, 362, 364, 365. Lestodon, 339. Libellulidse, 283. Light, penetration of, into water, 118, 128. Lima, 207. Liinacina, 260. Limidag, 166, 272. Limn«3a, 69, 261, 264. Limnohyus, 372. Liinnophis, 327. Limnotlierium, 403. Limulus, 207, 278. Ling (Molva), 69, 294. Lingula, 137, 139, 145, 253, 255, 257. Lingulella, 137, 139, 145. Lingulcpis, 255. Linnet, 65, 70. Linota, 70. Lion, 36, 384. Listriodon, 369. Lithentomum, 146, 282. Lithophis, 327. Lithostrotion, 151, 215, 249. Littoral fauna, 124. Littoral zone, 262. Lituites, 145. Lizards, dispersal of, 45 ; distribu- tion of, 315. Llama, 375. Loach, 288. Lobites, 207. Lophiodon, 369. Lophohelia, 243, 244. Lophophorus, 66'. Lophopsittacus, 333. Lophotragus, 72. Loris, 402. Lota, 69. Loxodon, 365. Loxolophodon, 367. INDEX. 423 Loxonema, 148, 269. Lucinidse, 166, 272. Lump-sucker (Discobolus), 293. Lung-fish, 80, 89, 103, 291, 302. Luscinia, 65, 70. Lutra, 390. Lyoena, 280, 281. Lycalopex, 387. Lycaon, 386. Lynx, 383, 384. Lyre-bird, 101. Lytoceras, 166, 207. Macaque (Macacus), 394, 398, 400, Macaw, 78, 88. Machairodus, 384. Mackerel, 297. Macrauchenia, 373. Macropetalichthys, 149. Macropodidse, 98, 336. Macropoma, 302. Macropus, 99. Macrorhinus, 392. Macrornis, 331. Macroscelides, 346. Macrotherium, 338. Macrura, 275. Mactridae, 108. Madrepora, 242, 247-249. Ma?andrina, 247, 248. Magpie, 28. Malacoclcmrnys, 314. Mammalia, dispersal of, 43; distri- bution of, 333 ; Arctic, 70 ; of the Holarctic realm : of the Eurasiatic division, 59-61, 69-72 ; of the American division, 62-65, 72, 73 ; of the Neotropical realm, 74-77, 80, 81 ; of the Ethiopian realm, 83-87, 89, 90; of the Oriental realm, 91-94, 96, 97; of the Aus- tralian realm, 98-100 ; of the Polynesian realm, 103, 104 ; of the Mediterranean region, 105. Mammalian descent, line of, 176. Mammoth, 205, 366. Manakin, 78. Manatee, 339, 340. Mandrill, 397. Mangabey, 397. Manicinn, 248. Manis, 336. Manucodia, 101. Marmoset, 396. Marmot, 359, 364. Marsipobranchii, 299. Marsupial ia, distribution of, 334. Marten, 390. Mastodon, 366. M'doqua, 84. Meadow-mice, 356. MedKcottia, 156, 267. Medusae (deep-sea), 110 ; pelagic, 120. Megalichthys, 301. Megalonyx, 339. Megalornis, 331. Megalosaurus, 162, 169. Mcgapodius, 95, 101, 104. Mcgaptera, 341. Megatherium, 338. Megerlia, 255. Melanerpeton, 311. Melania, 166, 209. Meleagris, 66. Meles, 390. Meliphagidse, 102, 104. Mellivora, 390. Melonites, 151. Meniscoessus, 335. Menobranchus, 306. Menodus, 372. Menoporaa, 306, 311 Menuridse, 101. Mephitis, 390. Merganser (fossil), 332. Meriones, 355. Merops, 87, 94, 106. Merychyus, 382. Mesohippus, 371, 372. Mesonyx, 393. 424 INDEX. Mesopithecus, 400. Mcsoplodon, 342, 345. Metridium, 242. Metriorhynchus, 329. Miacis, 393. Mice, 354, 355, 364. Micraster, 168. Microlabis, 151. Microlestes, 334. Micromeryx, 381. Microsyops, 348, 403. Midas, 396. Migration of birds, 39, 40 ; of fishes, 41 ; of reptiles, 41 ; of quadru- peds, 40, 41. Minivet, 94. Mink, 390. Miohippus, 371, 372. Mionornis, 333. Miopithecus, 397. Mixodectes, 403. Mniotiltidee, 66. Mole (Talpidae), 347. Mole-rats, 357. Molgophis, 311. Mollusca, dispersal of, 53 ; longevity of, 54 ; Arctic, 70 ; of lakes, 131 ; fresh -water, earliest, 209 ; of American coast, 216, 217; of British coast, 216 ; of French coast, 216 ; of Iberian coast, 216 ; of Norwegian coast, 216 ; Japan- ese, 216, 217 ; Mediterranean, 216, 218 ; Indo-Pacific, 217 ; Ant- arctic, 219; distribution of, 258; indicating climate, 224. Molluscan provinces, 258. Moloch, 318. Molossus, 349. Molothrus, 66. Monachus, 391. Mongoose, 385. Monitor, 319. Monkeys, 75 ; distribution of, 393. Monk-fish (Kbina), 295. Monoceros, 344. Monograptus, 145. Monopleura, 169. Monotreniata, distribution of, 333. Montlivaltia, 158. Moose, 381. Morelia, 326. Mormyridae, 89. Moropus, 338. Morosaurus, 161. Mosasaurus, 169. Moschus, 380. Motacillidse, 66. Motmot, 78. Mound-builder, 95, 101, 104. Mouse-deer, 376. Mullet, 104, 292. Mullet-king (Apogon), 296. Muntjac, 381. Murchisonia, 148, 269. Muricidae, 169, 270. Mus, 354, 355, 364. Muscardinus, 357. Muscicapidse, 65, 79. Musk-deer, 380. Musk-ox, 206, 379. Musk-rat, 357, 365. Musophaga, 87. Mustela, 390. Myacites, 209, 395. Myidse, 168. Myliobatis, 297, 300. Mylodon, 339. Mynah, 94. . Myodcs, 357. Myogalc, 28, 347, 348. Myolagus, 364. Myopotamus, 360. Myoxidae, 357, 364. Myriapoda, geological distribution of, 285. Myiinecobiidae, 99. Myrmecobius, 99. Myrmecophaga, 337. Mysarachne, 347. INDEX. 425 Mystriosaurus, 329. Mytilidae, 148, 166, 207, 271. Naiadites, SOD. Narica, 269. Narwhal, 344. Nasua, 390. Natica, 166, 207, 262, 269. Nautilus, 137, 145, 192, 207, 213, 265- 267. Nesera, 263. Necvolemur, 348, 403. Nectarinea, 87, 94, 102, 106. Nectarophila, 94. Necturus, 306. Nematocarcinus, 112. Neobalaena, 345. Neolimulus, 279. Neophron, 106. Neoplagiaulax, 336. Neotropical realm, 73. Neritina, 166, 209, 211. Nesokerodon, 363. Ncsokia, 355, Nestor, 102. Neurogymnurus, 347. Neuroptera, 279, 282, 283. Neurotrichus, 348. Newt, 306. Night-ape, 395, 396. Nightingale, 65,70. Nimravus, 385. Noctiluca, 122. Notelephas, 366. Nothosaurus, 159. Notidanus, 297, 300. Notornis, 333. Nototherium, 336. Nuculidae, 148, 207, 263, 271. Numidinoe, 88. Nummulites, 152, 173, 189. Nuthatch (fossil), 65, 330, Nutria, 390. Nycticebus, 401. Nycticejus, 351. Nyctiornis, 94. Nyctipithecus, 395, 396. Nylghau, 378. Obolella, 255. Obolus, 145, 257. Oceanic basins, permanency of, 220. Ocelot, 383. Oculina, 248, 249. Odontopteryx, 331. Odontornithes, 330. Oeningen, insects of, 284. Olenus, 146. Oligobunis, 388. Olividae, 169, 270. Omphyma, 143. Onager, 370. Onchus, 142, 300. Onychogale, 96. Onychoteuthis, 138. Ophibolus, 323. Ophiderpeton, 310, 311. Ophidia, distribution of, 320. Ophidiidae, 299. Ophioderma, 166. Ophioglypha, 166. Ophiophagus, 322, 325. Ophiosaurus, 316. Ophiura, 111. Ophiurclla, 166. Ophiuroidca (deep-sea). 111. Opisthocomus, 81. Opo.ssum, 334. Oppelia, 267. Oracanthus, 152. Orbicella, 248. Orbiculina, 234. Orbitoides, 173. Orbitolites, 234. Orbulina, 109, 235-237. Orca, 343. Orcella, 343. Orders, distribution of, 32. Oreodon, 382. Oreopithecus, 400. 426 INDEX. Oreortyx, 66. Orcosaurus, 323. Organist, 78. Oriental realm, 00. Oriole, 106. Ornithopoda, 162. Ornithorhynchus, 333. Orodus, 300. Orohippus, 371, 372. Orthis, 145, 255. Orthisina, 255. Orthoccras, 137, 139, 145, 156, 190, 192, 265, 266. Orthoptcra, 279, 283, 284. Orthotomus, 94. Orycteropus, 336. Oryx, 84, 377. O-scillutoriiB, 120. Osteolepis, 149, 301. Osteopygis, 315. Osteornis, 331. Ostracoda, 126, 274. Ostrea, 166, 168, 211, 272. Ostrich, 32, 38, 88; American, 79; fossil, 332. Otaria, 391. Otis, 66. Otocyon, 386 Otodus, 300. Otter, 390. Oudenoclon, 159. Ounce, 383. Oustiti, 396. Ovibos, 379, 380. Owl, 19, 66, 69, 104 ; fossil, 331-333. Oxyaena, 393. Oyster, earliest, 272. Paca. 3d. Pachycyon, 388. Pachy lemur, 403. Pachysimia, 375. Pacific, temperature of, 116. Palsencodon, 348. Palaearctic region, 56. Palaemon, 275. Palseoblattina, 146, 282, 283. Palaeocampa, 286. Palaeocetus, 344. Paloeochcerus, 374. Palseocircus, 331. Palseodictyoptera, 282. Palseolagu;, 365. Paloeolama, 376. Palaeolemur, 348, 402. Pala3omc^x, 381. Palseoniscus, 302. Palceonyctis, 393. Palasopalsemon, 148, 275. Palaeophis, 327. Palseophoca, 393. Palseophoncus, H6, 285. Palaeopithecus, 400. Palseorbis, 208. Palseoreas, 378. Palaeornis, 88. Pala?ornithidae, 95. Palseortyx, 331. ' Palasoryx, 378. Palaeosaurus, 155. Palaeosiren, 311. Palaeospalax, 348. Pala?ospiza, 331. Palseosyops, 372. Palseotherium, 371. Palseotragus, 378. Palseotringa, 330. Palaeozoic faunas, 156. Palamedea, 8J. Palapteryx, 333. Palasterina, 135. Palechinus, 151. Palephemera, 282, 283. Palestine, birds of, common to Eu- rope, 106. Palinuru*, 122. Paladestrina, 264. Paludina, 69, 166. Pangolin, 336, 337. Panochtlius, 339. INDEX. 427 Panther, 383, 384. Paractis, 241. Paracyathus, 245. Paradise-bird, 100, 107. . Paradiseinjfi, 100, 140, 146, 276-278. Paradoxurus, 385. Parahyus, 374. Parakeet, 78, 95, 101. Parameles, 99. Parameryx, 376. Parasmilia, 244. Parasorex, 346. Paridse, 65, 107. Parrots, 39, 88, 95, 101 ; fossil, 332. Partridge, 66, 69, 72. Pastor, 94, 106. Pauxi, 81. Pavo, 97. Pavonaria, 251. Pavonia, 247, 248. Peachia, 242. Peacock, 95, 97. Peccary, 5, 374. Pecchiolia, 263. Pecten (Pectinidae), 169, 207, 263, 272. Pedetes, 358. Pelagic fauna (oceanic), 119 ; of lakes, 127-130. Pelagomis, 330. Pelecan (fossil), 332. Pclias, 71. Pelion, 310. Pcltocans, 273. Pcnseus, 275. Pennatulids, 251. Pentacrinus, 157, 1C6. Pentamerus, 145. Pentremites, 151. Perch (Perca), 68, 69, 288. Perdix, 66. Pericrocotus, 94. Permian fauna, 154. Perodicticus, 402. Persia, birds of, common to Europe, 106. Petalodus, 300. Petaurista, 99. Petrogale, 99. Petromyzon, 68. Phacochoerus, 374. Phacops, 146. Phalanger, 99. Phalangistidae, 99. Pharyngobranchii, 299. Phascolarctos, 99. Phascolomyidse, 100. Phascolotherium, 335. Phasianidse, 95, 100. Phasianus, 66. PhasmidjE, 283. Pheasant, 66, 95, 100 ; fossil, 332. Pheriacodus, 368, 372. Phillipsia, 141, 151, 277. Philodryas, 90. Phoca, 392, 393. Pholadomyid«, 166, 168, 272. Phragmoceras, 145. Phrynosoma, 317. Phylloceras, 166, 267. Phyllodactylas, 316, 318. Phyllograptus, 145. Phyllopoda, 274. Phyllornithidse, 94. Phyllostomidae, 352. Physa, 69, 261. Physalia, 120. Physeter, 342. Picidse, 66, 100. Pie, 65. Pigeon, 66, 102, 104 ; fossil, 332. Pig-rat, 355. Pika, 28, 362. Pike (Esox), 68, 69, 288, 303. Pilot-fish, 297. Pilot-whale, 343. Pimelodus, 89, 289. Pinacoceras, 267. Pinna, 207. Pipa, 310, 312. Pipe-fish (Syngnathus), 293, 303. 428 INDEX. Pipridse, 78. Pithecia, 395. Pityophis, 323. Placodus, 159. Plagiaulax, 335. Plagiodontia, 360, Plaice, 294. Planorbis, 69, 166, 209, 261, 264. Plantain-eater, 87. Platanista, 341. Platephemera, 146". Platidactylus, 316. Platyceras, 148. Platycercus, 101. Platycrinus, 151. Platydia, 253, 255. Platygonus, 374. Platyrhina, 393. Plectrophanes, 70. Plesiarctomys, 364. Plesiocetusr 345. Plesiochelys, 164, 315-. Plesiosaurus, 159, 162. Plesiosorex, 347. Plethodon, 306. Pleurosternum, 31 5c Pleurotoma, 262, 263.. Pleurotomaria, 138,, 148, 166y 207, 269. Pleurotomidaev 169, 270. Pliauchenia, 376. Pliohippus, 371, 372. Pliopithecus, 400. Pliosaurus, 162. Ploceus, 83, 100, 104. Plutonia, 276. Pocillopora, 247-249. Poebrottierium, 376. Poephaga, 379. Pogonodon, 385. Polar-bear, 389. Polecat, 390. Poly mastodon, 336. Polynesian realm, 103. Polyplectron, 97. Polypterus, 89, 291, 301. Polysiphonia, 241. Polystomella, 235. Pompadour, 78. Pontoporia, 341. Porcupine, 361, 362, 365. Porcupine-fish, 297, 303. Porgy (Pagrus), 294. Porites, 144, 247-249. Porpitn, 120. Porpoise, 343. Portax, 378. Portuguese man-of-war, 120. Petamochcerus, 374. Potamogale, 26, 346. Potoroo, 99. Potto, 402. Pouched-rat, 357. Pourtalesia, 112. Pourtalcs Plateau, 115. Praeostrea, 272. Prairie-dog, 359. Prestwichia, 278. Primates, distribution of, 393. Primitia, 140. Priodon, 338. Prionitidse, 78. Prisciturben, 144. Priscodelphinus, 345. Proboscidea, distribution of, 365. Proboscis-monkey, 399. Procamelus, 376. Procervolus, 381. Procyon, 390. * Productus, 148, 214. Proetus, 141, 151, 277. Prong-horn, 377. Prophoca, 393. Propithecus, 401. Proscorpius, 285. Protareea, 144. Protechimys, 363. Protelcs, 383. Proterosaurus, 320. Proteus, 306, 311. INDEX. 429 Protolabis, 376. Protolycosa, 151, 286. Protophasma, 150, 282. Protopithecus, 399. Protopterus, 89, 291, 302, 304. Prototalpa, 348. Protornis, 331. Protriton, 310. Provivera, 393. Prox, 381. Psammodus, 300. Psammophis, 71. Pseudselurus, 3-85. Pseudalopex, 387. Pseudastacus, 275. Pseudemys, 314. Pseudo-Neuroptera, 282, Pseudopus, 316. Psophia, 81. Psittacus, 88. Ptarmigan, 66, 70. Ptcranodon, 169. Pteraspis, 142, 149, 299, 301. Pterichthys, 149, 301. Pterocles, 71. Pterodactyl, 163, 164, 169. Pteromys, 359. Pteropoda (pelagic), 120. Pteropus, 352, 353. Pterosauria, 163, 172. Pterygotus, 148, 278, 279. Ptiloccrcus, 345. Ptilopus, 102. Ptiloris, 101. Ptilotis, 102. Ptychoccras, 168. Ptychodus, 300. Puff-adder, 88. Pullenia, 235, 236. Pultnonata, 269. Pulvinulina, 235. Puma, 383. Pupa, 150, 208, 212, 264. Purpuridae, 166. Putorius, 390. Pycnodontidae, 94. Pycnodus, 302. Pycnonotidae, 107. Pyranga, 78. Pyrocystis, 122. Pyrosoma, 120. Pyrrhula, 70. Pyrula, 269. Python, 325, 326. Pythonomorpha, 169, 172. Quagga, 370. Quail, Californian, 66. Quail (fossil), 330. Quinqueloculina, 237. Babbit, 363, 364. Kaccoon, 390. Radiolaria, 109. Kadiolites, 169. Rail (fossil), 330, 332. Rana, 307, 308, 312. Rangifer, 381, 382. Rasse, 385. Rastrites, 145. Rat, 198, 355, 364. Ratel, 390. Rattlesnakes, 67, 322, 323. Raven, 19, 79. Ray (Raja). 293, 294, 299, 300. Receptaculites, 236. Red Sea, color of, 120. Redstart, 65. Red-wing, 66. Regulus, 66. Reindeer, 37, 70, 381, 382. Reptiles, dispersal of, 45 ; age of, 159 ; distribution of (see CHELONIA, CROCODILIA, LACEBTILTA, OPHI- DIA) ; of the Holarctic realm, 66, 67, 70, 71; of the Neotropical realm, 79, 80 ; of the Ethiopian realm, 88, 90; of the Oriental realm, 95, 96 ; of the Australian realm, 102; of the Polynesian 430 realm, 104 ; of the Mediterranean region, 106. Kequienia, 169. Ehabdoceras, 156, 268. Khabdophyllia, 247. Ehamphastidse, 78. Ehamphorhynchus, 163, 164. Ehamphosuchus, 329. Ehea, 80. Bhesus-monkey, 398. Ehinoceros, 368. Bhinochetidse, 104. Elrinodon, 297. Ehinolophus, 352, 354. Ehipidomys, 356. Elrizocrinus, 111. Ehizodus, 301. Bhynchocephala, 320. Bhynchonella, 145, 165, 188, 207, 252, 255. Ehynchosaurus, 159. Ehytina, 340. Ehytiodus, 340. Eifle-bird, 101. Eight-whale (Balsena), 341. Bivcr-hog, 374. Eobin, 65. Eockling (Motella), 294, 296. Eocky Mountain goat, 377. Eodentia, distribution of, 354. Eoe, 381, 382. Borqunl, 341, 342. Eotalia, 237. Eotifera (of lakes), 127, 131. Budistse, 169, 208, 272. Bugose corals, 143. Eupicola, 78. Eusa, 381. Sable, 390. Saccamina, 236. Saccomys, 365. Saccostomus, 357. Sagcceras, 156, 267. Saiga, 377. Saki, 395. Salamander, 306, 312. Salenidse (deep-sea), 111. Salmon (Salmo), 68, 69, 288. Salpa, 120. Sandpiper (fossil), 330, 3S2. Sandwicb Islands, birds of the, 50. Sanguinolites, 271. Saniva, 320. Sao, 146. Sapajou, 395. Sarcolemur, 348. Sarcorhamphus, 73, 79. Sauranodon, 162. Sauropoda, 161. Saw-bill, 78. Saw-fish (Pristis), 300. Scalops, 348. Scapanus, 348. Scaphaspis, 299, 301. Scaphites, 168, 267. Scaphopoda, 2G3. Scarabeidse, 284. Sceleporus, 317. Scelidosaurus, 162. Scelidotherium, 339. Sclmopoda (deep-sea), 112. Scincus, 320. Sciurodon, 364. Sciuromys, 365. Sciuropterus, 359. Sciurus, 358, 364, 365. Scopelidae, 297-299. Scorpsenidse, 296. Scorpions, 285. Scotophis, 323, 324. Screamer, 81. Sculpin, 68. Scyllium, 294, 299, 300. Sea-anemones, distribution of, 240. Sea-cows, 339. Sea-devils, 297. Sea-elephant, 392. Sea-fans, 251. Sea-horse (Hippocampus), 294, 297. INDEX. 431 Seal, 22, 391. Sca-lioa, 391. Sea-otter, 390. Sea-perch (Serranus), 294, 296. Sea-urchins, 111, 168. Secretary-bird, 88. Selache, 297. Sclasphorus, 78. Salenodonta, 373, 375. Scmele, 263. Scmioptera, 107. Semnopithecus, 394, 398, 399. Serpcntarius, 88. Serpents, dispersal of, 45 ; distribu- tion of, 320. Serval, 384. Sewellel, 360. Shark, 293, 297, 299, 300. Sheep, 379, 380. Shrew, 345, 346. Shrike, 65, 100. Sialia, 66. Siamang, 397. Sidcrastrsea, 249. Siderina, 248. SieboJdia, 306, 311. Sigmodon, 356. Silurian fauna, 142. Siluridae, 68, 80, 89. Simosaurus, 159. Siphonia, 168. Siren, 306, 311. Sirenia, distribution of, 339. Siskin, 70. Sistrurus, 323. Sittidae, 65, 107. Sivatherium, 377. Skink (Eumcces), 316. Skip-jack, 297. Skipper, 297. Skunk, 4, 390. Slimonia, 148, 278. Sloths, 337. Smelt (Osmerus), 294, 303. Snapper (Mesoprion), 296. Snipe (fossil), 332. Snow-partridge, 66. Solenidae, 272. Solenodon, 346. Soles (Rhombus, Pleuronectes, Solea), 294, 296. Solitaire (Pezophaps), 333. Sornateria, 69. Sonoran transition region, 106. Sorex, 346. Spalacidse. 357. Sparrow, 65, 197 ; fossil, 332. Species, appearance of, 190 ; extinc- tion of, 197 ; distribution of, 17- 26 ; ancient migrations of, 220, 229. Spelerpes, 306. Spermophile, 359. Sperm-whale, 341, 342. Sphserexochus, 278. Sphserulites, 169. Sphargis, 313. Sphenodon, 320. Sphenotrochus, 243. Spheroidina, 109, 235, 236. Spice-bird, 78. Spider-monkey, 394, 395. Spiders, 285. Spiny-mice, 355. Spiny-rats, 360. Spirifer, 145. Spirit'erina, 165. Spirula, 20. Sponges (deep-sea), 110 ; Cretaceous, 168. Sprat (Clupea), 295. Springbok, 84, 377. Squalodon, 344. Squaloraja, 300. Squatina, 300. Squirrel, 358, 365. Squirrel-monkey, 395. Stag, 381, 382. Stagonolepis, 328. Standard-wing, 107. 432 INDEX. Staphylinidse, 69, 284. Star-fishes, 111, 135, 166. Star-gazer (Uranoscopus), 294. Starling, 66, 79, 94. Steatornis, 18. Stegocephala, 310. Stegodon, 366. Stegosauras, 162. Stellarida, 135. Stcllio, 71, 316. Steneofiber, 364, 365. Steneosaurus, 329. Stephanoceras, 166, 267. Stcphanophyllia, 243, 245. Stereognathua, 335. St. Helena, birds of, 50. Stickleback ( Gasterosteus), 68, 288. Stone-hatch, 107. Stork (fossil), 332. Strepomatidae, 24. Streptclasma, 143. Strigopidae, 102. Stromatopora, 145. Strombidae, 166, 169. Strophalosia, 143. Strophites, 148, 208. Strophodus, 300. Strophomena, 145. Struthio, 79, 88. Sturgeon (Accipenser), 66, 69, 289, 300. Sturnidse, 66, 79. Stylacodon, 335. Stylaraea, 144. Stylina, 247. Stvlodon, 335. Stylonurus, 278, 279. Sucker, 68. Sucker (Catostomus), 289, 291. Sugar-bird, 78. Summer-redbird, 78. Sun -bird, 94. Sun-fish (Orthagoriscus), 68, 297. Surmullets (Mullus), 294. Surnia, 69. Sus, 374. Swallow, 65, 104. Swallow-shrike, 104. Swan (fossil), 333. Swift, 66, 101. Sword-fish, 297. Sylviadse, 65, 88* 94. Symborodon, 372. Symplectes, 88. Synchronism, geological, 227. Syringopora, 143, 151, 249. Syrrhaptes, 71. Tabulate corals, 143, 251. Tailor-bird, 88, 94. Talapoin, 397. Talpa, 347, 348. Talpavus, 348. Tamandua, 337. Tamias, 359, 365. Tanagcr, 78. Tanganyika, Lake, fauna of, 212. Tantilla, 323. Tapir (Tapirus), 28, 369. Tapiravus, 309. Tapirulus, 309. Tarsier (Tarsius), 401. Tarsipes, 100. Tatouay, 337. Tattler (fossil), 332. Tatusia, 337. Taxidea, 390. Tectonarchinse, 101. Tciidaa, 319. , Tdegallus, 101. Telerpeton, 320. Teleosaurus, 329. Tellinidse, 168, 272. Tcmnocyon, 388. ' Terebratella, 165, 252, 255. Terebratula, 148, 165, 188, 207, 252- 255, 260. Terebratulina, 188, 252, 253, 255, 2CO. Tcrebridae, 270. Terrapin, 314. INDEX. 433 Tertiary fauna, 11, 171. Testudo, 314, 315. Tetragonolepis, 302. Tetrao, 66. Tetraogallus, 66. Teuthidse, 2C8. Teuthopsis, 138, 166. Textor, 88. Textularia, 237. Thalassarctos, 389. Thalassemys, 315. Thalassochelys, 313. Thamnastnea, 157, 247. Thaumalia, 66. Thecidium, 253, 255, 257. Thecocyathus, 244. Theeodontosaurus, 160. Thecosiphonia, 168. Thecosmilia, 157, 247. Thelodus, 142, 300. Theridomys, 363, 364. Theriodontia, 159. Theropithecus, 398. Tberoplcura, 154. Theropoda, 162. Thinohyus, 374. Thoracosanrus, 329. Thous, 387. Thrasaetos, 79. Thrush, 65, 100, 104. Thunder-worm, 317. Thylacinus, 99. Thylacoleo, 336. Tiger, 5, 18, 35, 36. Timalidse, 94, 107. Tinamida?, 78. Tinamou, 78. Tinoceras, 367. Tinoporus, 234. Tipulidse, 283, 284. Tit, 22, 65, 107. Titanichthys, 149. Titanophasma, 150, 282. Titanctherium, 372. Toad, 307, 312 Tolypeutes, 338. Tomistoma, 328. Tomitherium, 348. Toothed-carps (Cyprinodonts), 290. Toucan, 78. Toxoceras, 168. Tragoceros, 378. Tragopan, 66, 95. Tragulus, 376. Tree-frog, 307, 309. Tremarctos, 389. Triacodon, 348. Triassic fauna, 156. Trichecus, 391. Trichiulus, 151, 286. Trichodesmium, 120. Trichoglossus, 101. Triconodon, 335. Trigonia, 199. Trigoniadse, 166. Trigonocephalus, 71. Trilobites, appearance and disappear- ance of, 200; distribution of, 275. Trinucleus, 146, 277. Trionychidse, 314. Trionyx, 315. Triton (Amphibia), 306, 307, 312. Triton (Mollusca), 260. Tritonidse (Mollusca), 169, 270. Tritylodon, 335. Trochidae, 166. Trochilus, 78. Trochoceras, 192. Trochocyathus, 243, 245. Trochosmilia, 245. Troglodytes, 18, 396, 397. Trogon, 31, 78, 95. Tropidonotus, 67, 321, 322, 324, 325. Trout, 290. Troxites, 283. Trumpeter, 81. Trumpet-fish (Centriscus), 295. Trunk-fish (Ostracion), 303. 434 INDEX. Trygon, 294, 300. Tubipora, 143, 247. Tunicata (pelagic), 120. Tunny, 297. ] Tupaia, 345. J Turaco, 87. 1 Turbinolidas, 242. Turbo, 207. Turbot, 294. Turdidse, 65. Turkey, 66 ; fossil, 332. Turkey-buzzard, 66. Turnix, 106. Turrilites, 168, 267. Turtle-dove, 102. Turtles, distribution of, 313. Type-structure, persistence of, 207. Typhlopidje, 322, 325. Tyrannus (Tyrannidse), 6G, 79. Tyrant-shrike, 66, 79. Tyrolites, 267. Tyrrhenian transition region, 105. Tze-tze, ravages or, 206. Uintatherium, 367. Umbra, 288. Umbrella-bird, 78. Umbrinc, 294. Uncites, 148. Ungulata Artiodactyla, 373. Ungulata Pcrissodactyla, 368. Unio (Unionidas), 69, 209, 211, 261. Upupa, 106. Uraster, 166. Urocordylus, 155. Urodela, distribution of, 306. Uromys, 100. Uropeltidse, 326. Urotrichus, 348. Ursus, 389. Urus, 379. Uta, 317. Vallonia, 265. Vulvata, 69, 209. Vampyres, 349, 352. VaranidiK, 319. Velella, 120. Veneridse, 168, 272. Vcntriculites, 168. Vertigo, 265. Vesper-mice, 355. Vespertilio, 349, 351, 354. Vesper ugo, 349, 351, 354. Vespidae, 284. Vicuna, 375. Vidua, 88. Viper (Viperus), 67, 71, 321, 324. Viscacha, 24, 361. Viverra, 385. Viverricula, 385. Vivipara, 209, 211. Vole, 356, 365. Volutidae, 270. Vulture, 66, 88, 100, 106. Wagtail, 66 ; fossil, 332. Waldheimia, 255, 257. Walrus, 391. Wanderoo, 398. Wapiti, 381, 382. Warbler, 65, 88, 94, 100, 104. Wart-hog, 374. Wasps, 279, 284. Water- buck, 84. Water-mole, 347. Weasel, 390. Weaver-bird, 72, 88, 100, 104. Whales, distribution of, 341. White-whale, 344. Whydah, 88. Wild -cat, 384. Wolf, 18, 387. Wolf-fish (Anarrichas), 293, 294. Wolverine, 390. Wombat, 100. Woodchuck, 359, 365. Woodpecker, 66, 100 ; fossil, 332. Wood- warbler, 66. Worms (of lakes), 131. INDEX. 435 Xenodiscus, 156, 267. Zapus, 358. Xenoneura, 146, 282. Zarachis, 345. Xiphodon, 382. Zebra, 370. Xylobius, 151, 286. Zebu, 92. v Zeuglodon, 344. * „' , Ziphius, 342, 345. \ ellow-liammer, 70. Zoarccs, 294. Zalophus, 391. Zonites, 150, 208. Zamenis, 321 . Zoological regions, definition of, 56. Zanclodon, 160. Zorilla, 390. Zaphrentis, 143, 151. Zygobatis, 300. THE END. ANIMAL INTELLIGENCE By GEOBGE J. BOMANES, F.R.S., Zoological Secretary of the Linnsean Society, etc. 12MO. CLOTH, $1.75. "My object in the work as a whole is twofold: First, t have thought it de- airable that there should be Hometbing resembling a text-book of the facts of Comparative Psychology, to which men of science, and also metaphysicians, may turn whenever they have occasion 10 acquaint themselves with the particular Jevel ot intelligence to which this or that species of animal attains. My second and much more important object is thar, of considering the facts of animal intel- ligence in their relation to the theory of descent.'1 — from the Preface. " Unless \ve are greatly mistaken, Mr. Romanes's work will take its place as one of the most attractive volumes of the INTERNATIONAL, SCIENTIFIC SEEIES. Some persons may. indeed, be disposed to say that it is too attractive, that it feeds tne popular taste for the curious and marvelous without supplying any commensurate discipline in exact scientific reflection ; but the author has, we think, fully justified himself in his modest preface. The result is the a- pearance of a collection cf facts which will be a real boon to the student of Comparative Psychology, for this is the first attempt to present systematically well-assured observations on the mental ILe of animals. " — Saturday Review. "The author believes himself, not without ample cau*e, to have completely bridged the suppled gap between instinct and reason by the authentic proofs here mar haled of remarkable intelligence in some of the higher animals. It is the seemingly com-.Iusive evidence of reasoning powers furnished by the adapta- tion of means to ends in cases which can not be explained on the theory of inner- itei aptitude or habit." — New York Sun. "The high standing of the author as an original investigator is a sufficient giiarantee that his task has been conscientiously carried out. His subject is one of absorbing interest. He has collected and classified an enormous amount of information concerning the mental attributes of the animal world. The result is astonishing. We find marvelous intelligence exhibited not only by animals \yhich are known to be clever, but by others seemingly without a glimmer of light, like the snail, fur instance. Some animals display imagination, others affection, and so on. The psychological portion of the discussion is deeply in- teres?t;ng.v— JVew; York Herald. " The chapter on monkeys closes this excellent work, and perhaps the most instructive portion of it is that devoted to the life-history of a monkey."— Ntw York Time.f. " Mr. Romanes brings to hi* work a wide information and the best of scientific methods. He has carefully culled and selected an immense mass of data, choos- ing with admirable skill those, facts which are really significant, and rejecting those which lacked sustaining evidence or relevancy. The contents of the volume arw arranged with reference to the principles which they seem to him to estab- lish. The volume is rich and suggestive, and a model in its way.11— Boston ( ourier. " It presents the facts of animal intelligence in relation to the theory of de- scent, supplementing Darwin and Spencer in tracing the principles which are coiicerned in the genesis of mind.11— Boston ( ommonwealtii. " One of the most interesting volumes of the series.'1— New York Christian at Work. " Few subjects have a greater fascination for the general reader than that vrith which this book is occupied/1 — Good Literature, New York. For sale by all booksellers ; or sent by mail, post-paid, on receipt of price. New York: D. APPLSTON & CO., 1, 3, an3 5 Bond Street. ANTHROPOID APES By ROBEBT HARTMANN, Professor in the University of Berlin. With. 63 Illustrations - 12mo, cloth, $1.75. "The anthropoid, or manlike or tailless, apes include tho gorilla and chimpnnzee of tropical Africa, the oran-r of Borneo :ind Sumatra, and the gibbons ol the East Indies. India, and some other pirts of Asia. ') be author of the present work h:i3 given much attention to the group. Like most living zoologists he is an evolutionist, and holds thut man can not have descended from any of the fossil species which have hitherto co.ne under our notice, nor yet from any of the species now extant: it is more prob-ible that both types have been produced from a common ground-form which has become extinct/'— T,ie Nation. " This Berlin professor is constrained, after a full presentation of the opinions and arguments of scientists and philosophars, and a careful collection :md analysis of recent facts and observation, to declare : " A great chasm between man and anthro- poids is constitu ed, as I believe, by t e fact that the human race is capable of educa- tion, and is able to acquire the highest mental culture, whiie the most intelligent anthropoid can only receive a certain mechanical training." — J\ew }ork Observer. "It will be found, by those who follow the author's exegesis with the heed and candor it deserves, that t ie simian ance-try of man does not :is yet rest upon such solid and perfected proofs as to warrant the assumption of absolute certainty in which materialists indulge.1' — New York Sun. "'The International Scientific Series' has now reached its fifty-recond volume. Started as a venture, the res ilt of which was very f'oubtful. the series ha* made its own way into the colleges, academies, ana public and private libraries of the country. Its secure position is dua to th.3 uniform excellence or the works which bear its name, and to the faith, energy, and capital of 1). Appletpn & «'o. This ho-ise knows by long experience that it pays to publish first- cl.iss scientific works. If the tone of such books see:ns at first to ba too high for the pubiic taste, then it only remains to educate the people up to them. This has been successfully done in the case ot 'The Inter- national Scientific Series.' One of its murked characteristics is the fullness of treat- ment accorded to every subject in every volume. Thus in the fifty-second issue re- lating to ' Anthropoid Apes,' the author, Professor Hnrtmann. of the University of Berlin, tells everything tlr.it one could possibly care to know at out the ap?s whose physical structure most nearly resembles that of man. It contains all that's in the libraries, plus a mass of the author's original observations. The goi ilia, chimpanzee, oranir-outang, and gibbon, undergo a minute and profound examination — in wi.d life, in caiitivity. and in the dissecting-room. There are more than 3 0 pages of this novel and interesting matter, accompanied by sixty-three illustrations. When the attentive r.-ader has finished the bonk he possesses all that science has yet discovered about the nature aud habits of anthropoid apes."— New York Journal of Commerce. " The most able and satisfactory summary of our knowledge upon this important brauch of science which has yet appeared." — Boston Courier. "The work is necessarily less complete than Huxley's monograph on 'The Craw- fish.' or Mivart's on 'The Cat.' but it is a worthy companion of those brilliant works; aad in saying this we bestow praise equally hi^h and deserved."— Boston Gazette. '•The arrangement of the work is most satisfactory. Th.3 volume is one of tha moat entertaining of the series."— Hartford Evening Pod. New York: D. APPLETON & CO., 1, 3, & 5 Bond Street. THE MAMMALIA IN THEIR RELATION TO PRIMEVAL TIMES. Ey Professor OSCAR SCHMIDT, Author of "The Doctrine of Descent and Darwinism." With 51 Woodcuts - - - - 12mo. Cloth, $1.50. l; Professor Schmidt was one of the best authorities on the subject which he hap here treated with the knowledge derived from the studies of a liletime. We use the past tense in speaking of him. because, since this book was prii.ted, its accomplished author has died in the fullness of his powers. Although he pre- pared it nominally for the use of advanced students, there are few if any pages in his book which can not be readily understood by the ordinary reader. As the title implies. Professor Schmidt has traced the Jinks of connection between existing mammalia and those types of which are known to us only through the dtsclosi.res of ecology. Pigs, camels, deer, horses, y that his line of speculation takes him through mai;y misty labyrinths of thought. While, however, the wriier d^als much with intricate hypotheses, he devotes much space to the anatomical structure and other physical peculiarities of the mammalia. This phase of bis woik gives it value apart from his theory." — New York Herald. "The work is an excellent and discriminating treatise upon one of the most important branches of what is by far the most ptiipenopns scientific problems of the day. It is marked l>y ripe t-cholarshin. keen intuition into the value and relations of tacts, and by that c'earness whkh c«n only result Irom a perfect maetery of the subject on the part of the author.'" — Boston Courier. "The author presents this ns furrisHng 'proofs of the ne< es^ty, the truth, and the value of D irwinism as the foundation for the theory of descent,' within the limited field des< nbed by the title. The work is supplemental to the au- thor's tn-atisc on the 'Doctrine of Descent and Darwinism,' published in the same series, but it is complete in itself, and includes the results of the latest BcieLtitic research in this field."— Boston Journal. " Professor Schmidt offers this work as ' a pugaestive i introduction to that portion of the animal kingdom which stai d« closest to anthropology.' He com- pares in detail living mammalia with their paleontoloerical ancestors, paying particular attention to the Pti nature of the teeth. Professor Schmidt asserts .that man's teeth have decreased in number during Hs development, and are likelv to deer ase in future. He believes that man will retain his present com- plement of fingers and toes, but that the race will eventually become bald. The work indicates minute-ness of research, and the subject-matter is ably pre- sented."— Albany Evening Journal. "As presenting the results of a considerab'e amount of original work the volume should meet with a wide welcome at the hands of students of natural history as a science of development. It doos not deal with (what human vanity has chosen to calli the highest type? of the animal kingdom merely as subjects for description, or even lor comparison with other forms, but considers them in their relations to surrounding facts. This inv Ives a study of the changes in organism due to the alteration of thoc-e conditions through the lapse of geologic time." — Chicago Tribune. u The history of the development of animals and tV>e history of the earth and geosrranhy are made to confirm one another. The book is Illustrated with wood- cut-, which will prove both interesting and instructive. It tells of living mam- malia, pi-js. hippopotami, camels, deer, antelopes, oxen, rhinoceroses, horses, elephants, sea-cows, whales, do^s, seals, insect-eaters, rodents, bats, semi apes, apes and their ancestors, and the man of the future."— Syracuse (N. Y.) Herald. New York : D. AITLETON & CO., 1, 3, & 5 Bond Street. EARTHQUAKES AND OTHER EARTH MOVEMENTS. By JOHN MILNE, Professor of Mining and Geology in the Imperial College of Engineering, Tokio, Japar. With 33 Illustrations. ... l£mo, cloth, $1.75. An attempt is made in this volume to give a systematic acconrt of various Earth Movements. These comprise Earthquakes, or the sudden violent move- ments of the ground ; Earth Tremors, or minute movements which escape our attention by the smallness of their amplitude ; Earth Pulsations, or movements which are overlooked on account of the length of their period; and Earth Oscilla- tions, or movements of long period and large amplitude. '• Having chosen Japan as the center of active seismic energies, Mr. "Milne has had the fullest opportunity of studying earthquakes, and this volume gives a sys- tematic account of various earth movements. Disturbances at sea, magnitude of waves, velocity of propagation, records of tides gauges, all find their place in this volume. The m my questions of a cosmical character are all ably treated by Professor Milne. One would have thought that from expeiience the Japanese would have built earthquake-proof houses, but Professor Milne says they Lave not."— New York Times, "In this little book Professor Milne has endeavored to bring together a1! that is known concerning; the nature and causes of earthquake movements. His task was one of much difficulty. Professor Milne's excellent work in the science of seismology has been done in Japan, in a region of incessant shocks of sufficient enemy to make observation possible, yet, with rare exceptions, of no disastrous effects. He has had the good fortune to be aided by Mr. Thomas Gray, a gentle- man of gre it constructive skill, as well as by Professors J. A. Ewing, W. S. Chap- lin, and his other colleagues in the scientific colony which has gathered about the Imperial University of Jap-.ni. To these gentlemen we owe the best of our sci- ence of seismology, for before their achievements we had nothing of value con- cerning the physical conditions of earthquakes except the great works of Kobert Mallet; and Mallet, with all his genius and devotion to the subject, had but lew chances to observe the actual shocks, and &o failed to understand many of their important features.1' — The Nation. " This volume contains a treat deal in the way of result? of recent observation that has never before been given to the reading public. A large part of fie ma- terial used watt obtained from experiment and criminal investigation durintr sm eight years' residence in Japan, where the author had an opportunity of observ- ing an earthquake on an average of once a week."— Aew York Christian Vrdon. " The author considers the primary causes of earthquakes to be telluric heats, solar heat, and variations in gravitating influences. Among the secondary causes are expansions and contractions of the earth's crusf, variations in temperature, barometrical pressure, rain, wind, etc. Some are due to explosions of steam beneath the crust of the earth, others to chemical action forming caverns in the earth which give way, and still others to volcanic evieceralion. The subject in all its bearings is exhaustively treated in the light of the latest researches, ar.d affords a very interesting study of a class of natural phenomena which have al- ways been involved in more or less obscurity." — Chicago Eiening Journal. " Although it. is addressed to a special class of readers, it has an interest which may be said to be universal. It will surprise readers to be told that nearly two thousand works have been published on the particular subject of earthquakes. In China a commission was appointed more than 1700 years ago to implicate the causes of these phenomena, and sixty-five works exist in the Japanese language devoted to their scientific consideration. The first part of this work deals With the various movements, oscillations, and tremors of the eaith, with their effects ; the later chapters being devoted to the theories of various writers on the phe- nomena. The volume is well illustrated."— fioston Evening Transcript, New York: D. APPLETOX & CO., 1, 3, & 5 Bond Street. O INI •H; +> I o •H University of Toronto Library 3 I 1 a !« & DO NOT REMOVE THE CARD FROM THIS POCKET Acme Library Card Pocket LOWE-MARTIN CO. LIMITED