Darwinism (Wallace)/Chapter XII

by Alfred Russel Wallace

 

CHAPTER XII

 

THE GEOGRAPHICAL DISTRIBUTION OF ORGANISMS

The facts to be explained—The conditions which have determined distribution—The permanence of oceans—Oceanic and continental areas—Madagascar and New Zealand—The teachings of the thousand-fathom line—The distribution of marsupials—The distribution of tapirs—Powers of dispersal as illustrated by insular organisms—Birds and insects at sea—Insects at great altitudes—The dispersal of plants—Dispersal of seeds by the wind—Mineral matter carried by the wind—Objections to the theory of wind-dispersal answered—Explanation of north temperate plants in the southern hemisphere—No proof of glaciation in the tropics—Lower temperature not needed to explain the facts—Concluding remarks.

The theory which we may now take as established—that all the existing forms of life have been derived from other forms by a natural process of descent with modification, and that this same process has been in action during past geological time—should enable us to give a rational account not only of the peculiarities of form and structure presented by animals and plants, but also of their grouping together in certain areas, and their general distribution over the earth's surface.

In the absence of any exact knowledge of the facts of distribution, a student of the theory of evolution might naturally anticipate that all groups of allied organisms would be found in the same region, and that, as he travelled farther and farther from any given centre, the forms of life would differ more and more from those which prevailed at the starting-point, till, in the remotest regions to which he could penetrate, he would find an entirely new assemblage of animals and plants, altogether unlike those with which he was familiar. He would also anticipate that diversities of climate would always be associated with a corresponding diversity in the forms of life.

Now these anticipations are to a considerable extent justified. Remoteness on the earth's surface is usually an indication of diversity in the fauna and flora, while strongly contrasted climates are always accompanied by a considerable contrast in the forms of life. But this correspondence is by no means exact or proportionate, and the converse propositions are often quite untrue. Countries which are near to each other often differ radically in their animal and vegetable productions; while similarity of climate, together with moderate geographical proximity, are often accompanied by marked diversities in the prevailing forms of life. Again, while many groups of animals—genera, families, and sometimes even orders—are confined to limited regions, most of the families, many genera, and even some species are found in every part of the earth. An enumeration of a few of these anomalies will better illustrate the nature of the problem we have to solve.

As examples of extreme diversity, notwithstanding geographical proximity, we may adduce Madagascar and Africa, whose animal and vegetable productions are far less alike than are those of Great Britain and Japan at the remotest extremities of the great northern continent; while an equal, or perhaps even a still greater, diversity exists between Australia and New Zealand. On the other hand, Northern Africa and South Europe, though separated by the Mediterranean Sea, have faunas and floras which do not differ from each other more than do the various countries of Europe. As a proof that similarity of climate and general adaptability have had but a small part in determining the forms of life in each country, we have the fact of the enormous increase of rabbits and pigs in Australia and New Zealand, of horses and cattle in South America, and of the common sparrow in North America, though in none of these cases are the animals natives of the countries in which they thrive so well. And lastly, in illustration of the fact that allied forms are not always found in adjacent regions, we have the tapirs, which are found only on opposite sides of the globe, in tropical America and the Malayan Islands; the camels of the Asiatic deserts, whose nearest allies are the llamas and alpacas of the Andes; and the marsupials, only found in Australia and on the opposite side of the globe, in America. Yet, again, although mammalia may be said to be universally distributed over the globe, being found abundantly on all the continents and on a great many of the larger islands, yet they are entirely wanting in New Zealand, and in a considerable number of other islands which are, nevertheless, perfectly able to support them when introduced.

Now most of these difficulties can be solved by means of well-known geographical and geological facts. When the productions of remote countries resemble each other, there is almost always continuity of land with similarity of climate between them. When adjacent countries differ greatly in their productions, we find them separated by a sea or strait whose great depth is an indication of its antiquity or permanence. When a group of animals inhabits two countries or regions separated by wide oceans, it is found that in past geological times the same group was much more widely distributed, and may have reached the countries it inhabits from an intermediate region in which it is now extinct. We know, also, that countries now united by land were divided by arms of the sea at a not very remote epoch; while there is good reason to believe that others now entirely isolated by a broad expanse of sea were formerly united and formed a single land area. There is also another important factor to be taken account of in considering how animals and plants have acquired their present peculiarities of distribution,—changes of climate. We know that quite recently a glacial epoch extended over much of what are now the temperate regions of the northern hemisphere, and that consequently the organisms which inhabit those parts must be, comparatively speaking, recent immigrants from more southern lands. But it is a yet more important fact that, down to middle Tertiary times at all events, an equable temperate climate, with a luxuriant vegetation, extended to far within the arctic circle, over what are now barren wastes, covered for ten months of the year with snow and ice. The arctic zone has, therefore, been in past times capable of supporting almost all the forms of life of our temperate regions; and we must take account of this condition of things whenever we have to speculate on the possible migrations of organisms between the old and new continents.

The Conditions which have determined Distribution.

When we endeavour to explain in detail the facts of the existing distribution of organic beings, we are confronted by several preliminary questions, upon the solution of which will depend our treatment of the phenomena presented to us. Upon the theory of descent which we have adopted, all the different species of a genus, as well as all the genera which compose a family or higher group, have descended from some common ancestor, and must therefore, at some remote epoch, have occupied the same area, from which their descendants have spread to the regions they now inhabit. In the numerous cases in which the same group now occupies countries separated by oceans or seas, by lofty mountain-chains, by wide deserts, or by inhospitable climates, we have to consider how the migration which must certainly have taken place has been effected. It is possible that during some portion of the time which has elapsed since the origin of the group the interposing barriers have not been in existence; or, on the other hand, the particular organisms we are dealing with may have the power of overpassing the barriers, and thus reaching their present remote dwelling-places. As this is really the fundamental question of distribution on which the solution of all its more difficult problems depends, we have to inquire, in the first place, what is the nature of, and what are the limits to, the changes of the earth's surface, especially during the Tertiary and latter part of the Secondary periods, as it was during those periods that most of the existing types of the higher animals and plants came into existence; and, in the next place, what are the extreme limits of the powers of dispersal possessed by the chief groups of animals and plants. We will first consider the question of barriers, more especially those formed by seas and oceans.

The Permanence of Oceans.

It was formerly a very general belief, even amongst geologists, that the great features of the earth's surface, no less than the smaller ones, were subject to continual mutations, and that during the course of known geological time the continents and great oceans had again and again changed places with each other. Sir Charles Lyell, in the last edition of his Principles of Geology (1872), said: "Continents, therefore, although permanent for whole geological epochs, shift their positions entirely in the course of ages;" and this may be said to have been the orthodox opinion down to the very recent period when, by means of deep-sea soundings, the nature of the ocean bottom was made known. The first person to throw doubt on this view appears to have been the veteran American geologist, Professor Dana. In 1849, in the Report of Wilke's Exploring Expedition, he adduced the argument against a former continent in the Pacific during the Tertiary period, from the absence of all native quadrupeds. In 1856, in articles in the American Journal, he discussed the development of the American continent, and argued for its general permanence; and in his Manual of Geology in 1863 and later editions, the same views were more fully enforced and were latterly applied to all continents. Darwin, in his Journal of Researches, published in 1845, called attention to the fact that all the small islands far from land in the Pacific, Indian, and Atlantic Oceans are either of coralline or volcanic formation. He excepted, however, the Seychelles and St. Paul's rocks; but the former have since been shown to be no exception, as they consist entirely of coral rock; and although Darwin himself spent a few hours on St. Paul's rocks on his outward voyage in the Beagle, and believed he had found some portions of them to be of a "cherty," and others of a "felspathic" nature, this also has been shown to be erroneous, and the careful examination of the rocks by the Abbé Renard clearly proves them to be wholly of volcanic origin.^[1] We have, therefore, at the present time, absolutely no exception whatever to the remarkable fact that all the oceanic islands of the globe are either of volcanic or coral formation; and there is, further, good reason to believe that those of the latter class in every case rest upon a volcanic foundation.

In his Origin of Species, Darwin further showed that no true oceanic island had any native mammals or batrachia when first discovered, this fact constituting the test of the class to which an island belongs; whence he argued that none of them had ever been connected with continents, but all had originated in mid-ocean. These considerations alone render it almost certain that the areas now occupied by the great oceans have never, during known geological time, been occupied by continents, since it is in the highest degree improbable that every fragment of those continents should have completely disappeared, and have been replaced by volcanic islands rising out of profound oceanic abysses; but recent research into the depth of the oceans and the nature of the deposits now forming on their floors, adds greatly to the evidence in this direction, and renders it almost a certainty that they represent very ancient if not primaeval features of the earth's surface. A very brief outline of the nature of this evidence will be now given.

The researches of the Challenger expedition into the nature of the sea-bottom show, that the whole of the land debris brought down by rivers to the ocean (with the exception of pumice and other floating matter), is deposited comparatively near to the shores, and that the fineness of the material is an indication of the distance to which it has been carried. Everything in the nature of gravel and sand is laid down within a very few miles of land, only the finer muddy sediments being carried out for 20 or 50 miles, and the very finest of all, under the most favourable conditions, rarely extending beyond 150, or at the utmost, 300 miles from land into the deep ocean.^[2] Beyond these distances, and covering the entire ocean floor, are various oozes formed wholly from the debris of marine organisms; while intermingled with these are found various volcanic products which have been either carried through the air or floated on the surface, and a small but perfectly recognisable quantity of meteoric matter. Ice-borne rocks are also found abundantly scattered over the ocean bottom within a definite distance of the arctic and antarctic circles, clearly marking out the limit of floating icebergs in recent geological times.

Now the whole series of marine stratified rocks, from the earliest Palaeozoic to the most recent Tertiary beds, consist of materials closely corresponding to the land debris now being deposited within a narrow belt round the shores of all continents; while no rocks have been found which can be identified with the various oozes now forming in the deep abysses of the ocean. It follows, therefore, that all the geological formations have been formed in comparatively shallow water, and always adjacent to the continental land of the period. The great thickness of some of the formations is no indication of a deep sea, but only of slow subsidence during the time that the deposition was in progress. This view is now adopted by many of the most experienced geologists, especially by Dr. Archibald Geikie, Director of the Geological Survey of Great Britain, who, in his lecture on "Geographical Evolution," says: "From all this evidence we may legitimately conclude that the present land of the globe, though consisting in great measure of marine formations, has never lain under the deep sea; but that its site must always have been near land. Even its thick marine limestones are the deposits of comparatively shallow water."^[3]

But besides these geological and physical considerations, there is a mechanical difficulty in the way of repeated change of position of oceans and continents which has not yet received the attention it deserves. According to the recent careful estimate by Mr. John Murray, the land area of the globe is to the water area as ·28 to ·72. The mean height of the land above sea-level is 2250 feet, while the mean depth of the ocean is 14,640 feet. Hence the bulk of dry land is 23,450,000 cubic miles, and that of the waters of the ocean 323,800,000 cubic miles; and it follows that if the whole of the solid matter of the earth's surface were reduced to one level, it would be everywhere covered by an ocean about two miles deep. The accompanying diagram will serve to render these figures more intelligible. The length of the sections of land and ocean are in the proportion of their respective areas, while the mean height of the land and the mean depth of the ocean are exhibited on a greatly increased vertical scale. If we considered the continents and their adjacent oceans separately they would differ a little, but not very materially, from this diagram; in some cases the proportion of land to ocean would be a little greater, in others a little less.

FIG. 32.

Now, if we try to imagine a process of elevation and depression by which the sea and land shall completely change places, we shall be met by insuperable difficulties. We must, in the first place, assume a general equality between elevation and subsidence during any given period, because if the elevation over any extensive continental area were not balanced by some subsidence of approximately equal amount, an unsupported hollow would be left under the earth's crust. Let us now suppose a continental area to sink, and an adjacent oceanic area to rise, it will be seen that the greater part of the land will disappear long before the new land has approached the surface of the ocean. This difficulty will not be removed by supposing a portion of a continent to subside, and the immediately adjacent portion of the ocean on the other side of the continent to rise, because in almost every case we find that within a comparatively short distance from the shores of all existing continents, the ocean floor sinks rapidly to a depth of from 2000 to 3000 fathoms, and maintains a similar depth, generally speaking, over a large portion of the oceanic areas. In order, therefore, that any area of continental extent be upraised from the great oceans, there must be a subsidence of a land area five or six times as great, unless it can be shown that an extensive elevation of the ocean floor up to and far above the surface could occur without an equivalent depression elsewhere. The fact that the waters of the ocean are sufficient to cover the whole globe to a depth of two miles, is alone sufficient to indicate that the great ocean basins are permanent features of the earth's surface, since any process of alternation of these with the land areas would have been almost certain to result again and again in the total disappearance of large portions, if not of all, of the dry land of the globe. But the continuity of terrestrial life since the Devonian and Carboniferous periods, and the existence of very similar forms in the corresponding deposits of every continent—as well as the occurrence of sedimentary rocks, indicating the proximity of land at the time of their deposit, over a large portion of the surface of all the continents, and in every geological period—assure us that no such disappearance has ever occurred.

Oceanic and Continental Areas.

When we speak of the permanence of oceanic and continental areas as one of the established facts of modern research, we do not mean that existing continents and oceans have always maintained the exact areas and outlines that they now present, but merely, that while all of them have been undergoing changes in outline and extent from age to age, they have yet maintained substantially the same positions, and have never actually changed places with each other. There are, moreover, certain physical and biological facts which enable us to mark out these areas with some confidence.

We have seen that there are a large number of islands which may be classed as oceanic, because they have never formed parts of continents, but have originated in mid-ocean, and have derived their forms of life by migration across the sea. Their peculiarities are seen to be very marked in comparison with those islands which there is good reason to believe are really fragments of more extensive land areas, and are hence termed "continental." These continental islands consist in every case of a variety of stratified rocks of various ages, thus corresponding closely with the usual structure of continents; although many of the islands are small like Jersey or the Shetland Islands, or far from continental land like the Falkland Islands or New Zealand. They all contain indigenous mammalia or batrachia, and generally a much greater variety of birds, reptiles, insects, and plants, than do the oceanic islands. From these various characteristics we conclude that they have all once formed parts of continents, or at all events of much larger land areas, and have become isolated, either by subsidence of the intervening land or by the effects of long-continued marine denudation.

Now, if we trace the thousand-fathom line around all our existing continents we find that, with only two exceptions, every island which can be classed as "continental" falls within this line, while all that lie beyond it have the undoubted characteristics of "oceanic" islands. We, therefore, conclude that the thousand-fathom line marks out, approximately, the "continental area,"—that is, the limits within which continental development and change throughout known geological time have gone on. There may, of course, have been some extensions of land beyond this limit, while some areas within it may always have been ocean; but so far as we have any direct evidence, this line may be taken to mark out, approximately, the most probable boundary between the "continental area," which has always consisted of land and shallow sea in varying proportions, and the great oceanic basins, within the limits of which volcanic activity has been building up numerous islands, but whose profound depths have apparently undergone little change.

Madagascar and New Zealand.

The two exceptions just referred to are Madagascar and New Zealand, and all the evidence goes to show that in these cases the land connection with the nearest continental area was very remote in time. The extraordinary isolation of the productions of Madagascar—almost all the most characteristic forms of mammalia, birds, and reptiles of Africa being absent from it—renders it certain that it must have been separated from that continent very early in the Tertiary, if not as far back as the latter part of the Secondary period; and this extreme antiquity is indicated by a depth of considerably more than a thousand fathoms in the Mozambique Channel, though this deep portion is less than a hundred miles wide between the Comoro Islands and the mainland.^[4] Madagascar is the only island on the globe with a fairly rich mammalian fauna which is separated from a continent by a depth greater than a thousand fathoms; and no other island presents so many peculiarities in these animals, or has preserved so many lowly organised and archaic forms. The exceptional character of its productions agrees exactly with its exceptional isolation by means of a very deep arm of the sea.

New Zealand possesses no known mammals and only a single species of batrachian; but its geological structure is perfectly continental. There is also much evidence that it does possess one mammal, although no specimens have been yet obtained.^[5] Its reptiles and birds are highly peculiar and more numerous than in any truly oceanic island. Now the sea which directly separates New Zealand from Australia is more than 2000 fathoms deep, but in a north-west direction there is an extensive bank under 1000 fathoms, extending to and including Lord Howe's Island, while north of this are other banks of the same depth, approaching towards a submarine extension of Queensland on the one hand, and New Caledonia on the other, and altogether suggestive of a land union with Australia at some very remote period. Now the peculiar relations of the New Zealand fauna and flora with those of Australia and of the tropical Pacific Islands to the northward indicate such a connection, probably during the Cretaceous period; and here, again, we have the exceptional depth of the dividing sea and the form of the ocean bottom according well with the altogether exceptional isolation of New Zealand, an isolation which has been held by some naturalists to be great enough to justify its claim to be one of the primary Zoological Regions.

The Teachings of the Thousand-Fathom Line.

If now we accept the annexed map as showing us approximately how far beyond their present limits our continents may

The thousand fathom line around the continents

have extended during any portion of the Tertiary and Secondary periods, we shall obtain a foundation of inestimable value for our inquiries into those migrations of animals and plants during past ages which have resulted in their present peculiarities of distribution. We see, for instance, that the South American and African continents have always been separated by nearly as wide an ocean as at present, and that whatever similarities there may be in their productions must be due to the similar forms having been derived from a common origin in one of the great northern continents. The radical difference between the higher forms of life of the two continents accords perfectly with their permanent separation. If there had been any direct connection between them during Tertiary times, we should hardly have found the deep-seated differences between the Quadrumana of the two regions—no family even being common to both; nor the peculiar Insectivora of the one continent, and the equally peculiar Edentata of the other. The very numerous families of birds quite peculiar to one or other of these continents, many of which, by their structural isolation and varied development of generic and specific forms, indicate a high antiquity, equally suggest that there has been no near approach to a land connection during the same epoch.

Looking to the two great northern continents, we see indications of a possible connection between them both in the North Atlantic and the North Pacific oceans; and when we remember that from middle Tertiary times backward—so far as we know continuously to the earliest Palaeozoic epoch—a temperate and equable climate, with abundant woody vegetation, prevailed up to and within the arctic circle, we see what facilities may have been afforded for migration from one continent to the other, sometimes between America and Europe, sometimes between America and Asia. Admitting these highly probable connections, no bridging of the Atlantic in more southern latitudes (of which there is not a particle of evidence) will have been necessary to account for all the intermigration that has occurred between the two continents. If, on the other hand, we remember how long must have been the route, and how diverse must always have been the conditions between the more northern and the more southern portions of the American and Euro-Asiatic continents, we shall not be surprised that many widespread forms in either continent have not crossed into the other; and that while the skunks (Mephitis), the pouched rats (Saccomyidae), and the turkeys (Meleagris) are confined to America, the pigs and the hedgehogs, the true flycatchers and the pheasants are found only in the Euro-Asiatic continent. But, just as there have been periods which facilitated intermigration between America and the Old World, there have almost certainly been periods, perhaps of long duration even geologically, when these continents have been separated by seas as wide as, or even wider than, those of the present day; and thus may be explained such curious anomalies as the origination of the camel-tribe in America, and its entrance into Asia in comparatively recent Tertiary times, while the introduction of oxen and bears into America from the Euro-Asiatic continent appears to have been equally recent.^[6]

We shall find on examination that this view of the general permanence of the oceanic and continental areas, with constant minor fluctuations of land and sea over the whole extent of the latter, enables us to understand, and offer a rational explanation of, most of the difficult problems of geographical distribution; and further, that our power of doing this is in direct proportion to our acquaintance with the distribution of fossil forms of life during the Tertiary period. We must, also, take due note of many other facts of almost equal importance for a due appreciation of the problems presented for solution, the most essential being, the various powers of dispersal possessed by the different groups of animals and plants, the geological antiquity of the species and genera, and the width and depth of the seas which separate the countries they, inhabit. A few illustrations will now be given of the way in which these branches of knowledge enable us to deal with the difficulties and anomalies that present themselves.

The Distribution of Marsupials.

This singular and lowly organised type of mammals constitutes almost the sole representative of the class in Australia and New Guinea, while it is entirely unknown in Asia, Africa, or Europe. It reappears in America, where several species of opossums are found; and it was long thought necessary to postulate a direct southern connection of these distant countries, in order to account for this curious fact of distribution. When, however, we look to what is known of the geological history of the marsupials the difficulty vanishes. In the Upper Eocene deposits of Western Europe the remains of several animals closely allied to the American opossums have been found; and as, at this period, a very mild climate prevailed far up into the arctic regions, there is no difficulty in supposing that the ancestors of the group entered America from Europe or Northern Asia during early Tertiary times.

But we must go much further back for the origin of the Australian marsupials. All the chief types of the higher mammalia were in existence in the Eocene, if not in the preceding Cretaceous period, and as we find none of these in Australia, that country must have been finally separated from the Asiatic continent during the Secondary or Mesozoic period. Now during that period, in the Upper and the Lower Oolite and in the still older Trias, the jaw-bones of numerous small mammalia have been found, forming eight distinct genera, which are believed to have been either marsupials or some allied lowly forms. In North America also, in beds of the Jurassic and Triassic formations, the remains of an equally great variety of these small mammalia have been discovered; and from the examination of more than sixty specimens, belonging to at least six distinct genera, Professor Marsh is of opinion that they represent a generalised type, from which the more specialised marsupials and insectivora were developed.

From the fact that very similar mammals occur both in Europe and America at corresponding periods, and in beds which represent a long succession of geological time, and that during the whole of this time no fragments of any higher forms have been discovered, it seems probable that both the northern continents (or the larger portion of their area) were then inhabited by no other mammalia than these, with perhaps other equally low types. It was, probably, not later than the Jurassic age when some of these primitive marsupials were able to enter Australia, where they have since remained almost completely isolated; and, being free from the competition of higher forms, they have developed into the great variety of types we now behold there. These occupy the place, and have to some extent acquired the form and structure of distinct orders of the higher mammals—the rodents, the insectivora, and the carnivora,—while still preserving the essential characteristics and lowly organisation of the marsupials. At a much later period—probably in late Tertiary times—the ancestors of the various species of rats and mice which now abound in Australia, and which, with the aerial bats, constitute its only forms of placental mammals, entered the country from some of the adjacent islands. For this purpose a land connection was not necessary, as these small creatures might easily be conveyed among the branches or in the crevices of trees uprooted by floods and carried down to the sea, and then floated to a shore many miles distant. That no actual land connection with, or very close approximation to, an Asiatic island has occurred in recent times, is sufficiently proved by the fact that no squirrel, pig, civet, or other widespread mammal of the Eastern hemisphere has been able to reach the Australian continent.

The Distribution of Tapirs.

These curious animals form one of the puzzles of geographical distribution, being now confined to two very remote regions of the globe—the Malay Peninsula and adjacent islands of Sumatra and Borneo, inhabited by one species, and tropical America, where there are three or four species, ranging from Brazil to Ecuador and Guatemala. If we considered these living forms only, we should be obliged to speculate on enormous changes of land and sea in order that these tropical animals might have passed from one country to the other. But geological discoveries have rendered all such hypothetical changes unnecessary. During Miocene and Pliocene times tapirs abounded over the whole of Europe and Asia, their remains having been found in the tertiary deposits of France, India, Burmah, and China. In both North and South America fossil remains of tapirs occur only in caves and deposits of Post-Pliocene age, showing that they are comparatively recent immigrants into that continent. They perhaps entered by the route of Kamchatka and Alaska, where the climate, even now so much milder and more equable than on the north-east of America, might have been warm enough in late Pliocene times to have allowed the migration of these animals. In Asia they were driven southwards by the competition of numerous higher and more powerful forms, but have found a last resting-place in the swampy forests of the Malay region.

What these Facts Prove.

Now these two cases, of the marsupials and the tapirs, are in the highest degree instructive, because they show us that, without any hypothetical bridging of deep oceans, and with only such changes of sea and land as are indicated by the extent of the comparatively shallow seas surrounding and connecting the existing continents, we are able to account for the anomaly of allied forms occurring only in remote and widely separated areas. These examples really constitute crucial tests, because, of all classes of animals, mammalia are least able to surmount physical barriers. They are obviously unable to pass over wide arms of the sea, while the necessity for constant supplies of food and water renders sandy deserts or snow-clad plains equally impassable. Then, again, the peculiar kinds of food on which alone many of them can subsist, and their liability to the attacks of other animals, put a further check upon their migrations. In these respects almost all other organisms have great advantages over mammals. Birds can often fly long distances, and can thus cross arms of the sea, deserts, or mountain ranges; insects not only fly, but are frequently carried great distances by gales of wind, as shown by the numerous cases of their visits to ships hundreds of miles from land. Reptiles, though slow of movement, have advantages in their greater capacity for enduring hunger or thirst, their power of resisting cold or drought in a state of torpidity, and they have also some facilities for migration across the sea by means of their eggs, which may be conveyed in crevices of timber or among masses of floating vegetable matter. And when we come to the vegetable kingdom, the means of transport are at their maximum, numbers of seeds having special adaptations for being carried by mammalia or birds, and for floating in the water, or through the air, while many are so small and so light that there is practically no limit to the distances they may be carried by gales and hurricanes.

We may, therefore, feel quite certain that the means of distribution that have enabled the larger mammalia to reach the most remote regions from a common starting-point, will be at least as efficacious, and usually far more efficacious, with all other land animals and plants; and if in every case the existing distribution of this class can be explained on the theory of oceanic and continental permanence, with the limited changes of sea and land already referred to, no valid objections can be taken against this theory founded on anomalies of distribution in other orders. Yet nothing is more common than for students of this or that group to assort that the theory of oceanic permanence is quite inconsistent with the distribution of its various species and genera. Because a few Indian genera and closely allied species of birds are found in Madagascar, a land termed "Lemuria" has been supposed to have united the two countries during a comparatively recent geological epoch; while the similarity of fossil plants and reptiles, from the Permian and Miocene formations of India and South Africa, has been adduced as further evidence of this connection. But there are also genera of snakes, of insects, and of plants, common to Madagascar and South America only, which have been held to necessitate a direct land connection between these countries. These views evidently refute themselves, because any such land connections must have led to a far greater similarity in the productions of the several countries than actually exists, and would besides render altogether inexplicable the absence of all the chief types of African and Indian mammalia from Madagascar, and its marvellous individuality in every department of the organic world.^[7]

Powers of Dispersal as illustrated by Insular Organisms.

Having arrived at the conclusion that our existing oceans have remained practically unaltered throughout the Tertiary and Secondary periods of geology, and that the distribution of the mammalia is such as might have been brought about by their known powers of dispersal, and by such changes of land and sea as have probably or certainly occurred, we are, of course, restricted to similar causes to explain the much wider and sometimes more eccentric distribution of other classes of animals and of plants. In doing so, we have to rely partly on direct evidence of dispersal, afforded by the land organisms that have been observed far out at sea, or which have taken refuge on ships, as well as by the periodical visitants to remote islands; but very largely on indirect evidence, afforded by the frequent presence of certain groups on remote oceanic islands, which some ancestral forms must, therefore, have reached by transmission across the ocean from distant lands.

Birds.

These vary much in their powers of flight, and their capability of traversing wide seas and oceans. Many swimming and wading birds can continue long on the wing, fly swiftly, and have, besides, the power of resting safely on the surface of the water. These would hardly be limited by any width of ocean, except for the need of food; and many of them, as the gulls, petrels, and divers, find abundance of food on the surface of the sea itself. These groups have a wide distribution across the oceans; while waders—especially plovers, sandpipers, snipes, and herons—are equally cosmopolitan, travelling along the coasts of all the continents, and across the narrow seas which separate them. Many of these birds seem unaffected by climate, and as the organisms on which they feed are equally abundant on arctic, temperate, and tropical shores, there is hardly any limit to the range even of some of the species.

Land-birds are much more restricted in their range, owing to their usually limited powers of flight, their inability to rest on the surface of the sea or to obtain food from it, and their greater specialisation, which renders them less able to maintain themselves in the new countries they may occasionally reach. Many of them are adapted to live only in woods, or in marshes, or in deserts; they need particular kinds of food or a limited range of temperature; and they are adapted to cope only with the special enemies or the particular group of competitors among which they have been developed. Such birds as these may pass again and again to a new country, but are never able to establish themselves in it; and it is this organic barrier, as it is termed, rather than any physical barrier, which, in many cases, determines the presence of a species in one area and its absence from another. We must always remember, therefore, that, although the presence of a species in a remote oceanic island clearly proves that its ancestors must at one time have found their way there, the absence of a species does not prove the contrary, since it also may have reached the island, but have been unable to maintain itself, owing to the inorganic or organic conditions not being suitable to it. This general principle applies to all classes of organisms, and there are many striking illustrations of it. In the Azores there are eighteen species of land-birds which are permanent residents, but there are also several others which reach the islands almost every year after great storms, but have never been able to establish themselves. In Bermuda the facts are still more striking, since there are only ten species of resident birds, while no less than twenty other species of land-birds and more than a hundred species of waders and aquatics are frequent visitors, often in great numbers, but are never able to establish themselves. On the same principle we account for the fact that, of the many continental insects and birds that have been let loose, or have escaped from confinement, in this country, hardly one has been able to maintain itself, and the same phenomenon is still more striking in the case of plants. Of the thousands of hardy plants which grow easily in our gardens, very few have ever run wild, and when the experiment is purposely tried it invariably fails. Thus A. de Candolle informs us that several botanists of Paris, Geneva, and especially of Montpellier, have sown the seeds of many hundreds of species of exotic hardy plants, in what appeared to be the most favourable situations, but that in hardly a single case has any one of them become naturalised.^[8] Still more, then, in plants than in animals the absence of a species does not prove that it has never reached the locality, but merely that it has not been able to maintain itself in competition with the native productions. In other cases, as we have seen, facts of an exactly opposite nature occur. The rat, the pig, and the rabbit, the water-cress, the clover, and many other plants, when introduced into New Zealand, flourish exceedingly, and even exterminate their native competitors; so that in these cases we may feel sure that the species in question did not exist in New Zealand simply because they had been unable to reach that country by their natural means of dispersal. I will now give a few cases, in addition to those recorded in my previous works, of birds and insects which have been observed far from any land.

Birds and Insects at Sea.

Captain D. Fullarton of the ship Timaru recorded in his log the occurrence of a great number of small land-birds about the ship on 15th March 1886, when in Lat. 48° 31' N., Long. 8° 16' W. He says: "A great many small land-birds about us; put about sixty into a coop, evidently tired out." And two days later, 17th March, "Over fifty of the birds cooped on 15th died, though fed. Sparrows, finches, water-wagtails, two small birds, name unknown, one kind like a linnet, and a large bird like a starling. In all there have been on board over seventy birds, besides some that hovered about us for some time and then fell into the sea exhausted." Easterly winds and severe weather were experienced at the time.^[9] The spot where this remarkable flight of birds was met with is about 160 miles due west of Brest, and this is the least distance the birds must have been carried. It is interesting to note that the position of the ship is nearly in the line from the English and French coasts to the Azores, where, after great storms, so many bird stragglers arrive annually. These birds were probably blown out to sea during their spring migration along the south coast of England to Wales and Ireland. During the autumnal migration, however, great flocks of birds—especially starlings, thrushes, and fieldfares—have been observed every year flying out to sea from the west coast of Ireland, almost the whole of which must perish. At the Nash Lighthouse, in the Bristol Channel on the coast of Glamorganshire, an enormous number of small birds were observed on 3d September, including nightjars, buntings, white-throats, willow-wrens, cuckoos, house-sparrows, robins, wheatears, and blackbirds. These had probably crossed from Somersetshire, and had they been caught by a storm the larger portion of them must have been blown out to sea.^[10]

These facts enable us to account sufficiently well for the birds of oceanic islands, the number and variety of which are seen to be proportionate to their facilities for reaching the island and maintaining themselves in it. Thus, though more birds yearly reach Bermuda than the Azores, the number of residents in the latter islands is much larger, due to the greater extent of the islands, their number, and their more varied surface. In the Galapagos the land-birds are still more numerous, due in part to their larger area and greater proximity to the continent, but chiefly to the absence of storms, so that the birds which originally reached the islands have remained long isolated and have developed into many closely allied species adapted to the special conditions. All the species of the Galapagos but one are peculiar to the islands, while the Azores possess only one peculiar species, and Bermuda none—a fact which is clearly due to the continual immigration of fresh individuals keeping up the purity of the breed by intercrossing. In the Sandwich Islands, which are extremely isolated, being more than 2000 miles from any continent or large island, we have a condition of things similar to what prevails in the Galapagos, the land-birds, eighteen in number, being all peculiar, and belonging, except one, to peculiar genera. These birds have probably all descended from three or four original types which reached the islands at some remote period, probably by means of intervening islets that have since disappeared. In St. Helena we have a degree of permanent isolation which has prevented any land-birds from reaching the island; for although its distance from the continent, 1100 miles, is not so great as in the case of the Sandwich Islands, it is situated in an ocean almost entirely destitute of small islands, while its position within the tropics renders it free from violent storms. Neither is there, on the nearest part of the coast of Africa, a perpetual stream of migrating birds like that which supplies the innumerable stragglers which every year reach Bermuda and the Azores.

Insects.

Winged insects have been mainly dispersed in the same way as birds, by their power of flight, aided by violent or long-continued winds. Being so small, and of such low specific gravity, they are occasionally carried to still greater distances; and thus no islands, however remote, are altogether without them. The eggs of insects, being often deposited in borings or in crevices of timber, may have been conveyed long distances by floating trees, as may the larvae of those species which feed on wood. Several cases have been published of insects coming on board ships at great distances from land; and Darwin records having caught a large grasshopper when the ship was 370 miles from the coast of Africa, whence the insect had probably come.

In the Entomologists' Monthly Magazine for June 1885, Mr. MacLachlan has recorded the occurrence of a swarm of moths in the Atlantic ocean, from the log of the ship Pleione. The vessel was homeward bound from New Zealand, and in Lat. 6° 47' N., Long. 32° 50' W., hundreds of moths appeared about the ship, settling in numbers on the spars and rigging. The wind for four days previously had been very light from north, north-west, or north-east, and sometimes calm. The north-east trade wind occasionally extends to the ship's position at that time of year. The captain adds that "frequently, in that part of the ocean, he has had moths and butterflies come on board." The position is 960 miles south-west of the Cape Verde Islands, and about 440 north-east of the South American coast. The specimen preserved is Deiopeia pulchella, a very common species in dry localities in the Eastern tropics, and rarely found in Britain, but, Mr. MacLachlan thinks, not found in South America. They must have come, therefore, from the Cape Verde Islands, or from some parts of the African coast, and must have traversed about a thousand miles of ocean with the assistance, no doubt, of a strong north-east trade wind for a great part of the distance. In the British Museum collection there is a specimen of the same moth caught at sea during the voyage of the Rattlesnake, in Lat. 6° N., Long. 22½° W., being between the former position and Sierra Leone, thus rendering it probable that the moths came from that part of the African coast, in which case the swarm encountered by the Pleione must have travelled more than 1200 miles.

A similar case was recorded by Mr. F.A. Lucas in the American periodical Science of 8th April 1887. He states that in 1870 he met with numerous moths of many species while at sea in the South Atlantic (Lat. 25° S., Long. 24° W.), about 1000 miles from the coast of Brazil. As this position is just beyond the south-east trades, the insects may have been brought from the land by a westerly gale. In the Zoologist (1864, p. 8920) is the record of a small longicorn beetle which flew on board a ship 500 miles off the west coast of Africa. Numerous other cases are recorded of insects at less distances from land, and, taken in connection with those already given, they are sufficient to show that great numbers must be continually carried out to sea, and that occasionally they are able to reach enormous distances. But the reproductive powers of insects are so great that all we require, in order to stock a remote island, is that some few specimens shall reach it even once in a century, or once in a thousand years.

Insects at great Altitudes.

Equally important is the proof we possess that insects are often carried to great altitudes by upward currents of air. Humboldt noticed them up to heights of 15,000 and 18,000 feet in South America, and Mr. Albert Müller has collected many interesting cases of the same character in Europe.^[11] A moth (Plusia gamma) has been found on the summit of Mont Blanc; small hymenoptera and moths have been seen on the Pyrenees at a height of 11,000 feet, while numerous flies and beetles, some of considerable size, have been caught on the glaciers and snow-fields of various parts of the Alps. Upward currents of air, whirlwinds and tornadoes, occur in all parts of the world, and large numbers of insects are thus carried up into the higher regions of the atmosphere, where they are liable to be caught by strong winds, and thus conveyed enormous distances over seas or continents. With such powerful means of dispersal the distribution of insects over the entire globe, and their presence in the most remote oceanic islands, offer no difficulties.

The Dispersal of Plants.

The dispersal of seeds is effected in a greater variety of ways than are available in the case of any animals. Some fruits or seed-vessels, and some seeds, will float for many weeks, and after immersion in salt water for that period the seeds will often germinate. Extreme cases are the double cocoa-nut of the Seychelles, which has been found on the coast of Sumatra, about 3000 miles distant; the fruits of the Sapindus saponaria (soap-berry), which has been brought to Bermuda by the Gulf Stream from the West Indies, and has grown after a journey in the sea of about 1500 miles; and the West Indian bean, Entada scandens, which reached the Azores from the West Indies, a distance of full 3000 miles, and afterwards germinated at Kew. By these means we can account for the similarity in the shore flora of the Malay Archipelago and most of the islands of the Pacific; and from an examination of the fruits and seeds, collected among drift during the voyage of the Challenger, Mr. Hemsley has compiled a list of 121 species which are probably widely dispersed by this means.

A still larger number of species owe their dispersal to birds in several distinct ways. An immense number of fruits in all parts of the world are devoured by birds, and have been attractively coloured (as we have seen), in order to be so devoured, because the seeds pass through the birds' bodies and germinate where they fall. We have seen how frequently birds are forced by gales of wind across a wide expanse of ocean, and thus seeds must be occasionally carried. It is a very suggestive fact, that all the trees and shrubs in the Azores bear berries or small fruits which are eaten by birds; while all those which bear larger fruits, or are eaten chiefly by mammals—such as oaks, beeches, hazels, crabs, etc.—are entirely wanting. Game-birds and waders often have portions of mud attached to their feet, and Mr. Darwin has proved by experiment that such mud frequently contains seeds. One partridge had such a quantity of mud attached to its foot as to contain seeds from which eighty-two plants germinated; this proves that a very small portion of mud may serve to convey seeds, and such an occurrence repeated even at long intervals may greatly aid in stocking remote islands with vegetation. Many seeds also adhere to the feathers of birds, and thus, again, may be conveyed as far as birds are ever carried. Dr. Guppy found a small hard seed in the gizzard of a Cape Petrel, taken about 550 miles east of Tristan da Cunha.

Dispersal of Seeds by the Wind.

In the preceding cases we have been able to obtain direct evidence of transportal; but although we know that many seeds are specially adapted to be dispersed by the wind, we cannot obtain direct proof that they are so carried for hundreds or thousands of miles across the sea, owing to the difficulty of detecting single objects which are so small and inconspicuous. It is probable, however, that the wind as an agent of dispersal is really more effective than any of those we have hitherto considered, because a very large number of plants have seeds which are very small and light, and are often of such a form as to facilitate aerial carriage for enormous distances. It is evident that such seeds are especially liable to be transported by violent winds, because they become ripe in autumn at the time when storms are most prevalent, while they either lie upon the surface of the ground, or are disposed in dry capsules on the plant ready to be blown away. If inorganic particles comparable in weight, size, or form with such seeds are carried for great distances, we may be sure that seeds will also be occasionally carried in the same way. It will, therefore, be necessary to give a few examples of wind-carriage of small objects.

On 27th July 1875 a remarkable shower of small pieces of hay occurred at Monkstown, near Dublin. They appeared floating slowly down from a great height, as if falling from a dark cloud which hung overhead. The pieces picked up were wet, and varied from single blades of grass to tufts weighing one or two ounces. A similar shower occurred a few days earlier in Denbighshire, and was observed to travel in a direction contrary to that of the wind in the lower atmosphere.^[12] There is no evidence of the distance from which the hay was brought, but as it had been carried to a great height, it was in a position to be conveyed to almost any distance by a violent wind, had such occurred at the time.

Mineral Matter carried by the Wind.

The numerous cases of sand and volcanic dust being carried enormous distances through the atmosphere sufficiently prove the importance of wind as a carrier of solid matter, but unfortunately the matter collected has not been hitherto examined with a view to determine the maximum size and weight of the particles. A few facts, however, have been kindly furnished me by Professor Judd, F.R.S. Some dust which fell at Genoa on 15th October 1885, and was believed to have been brought from the African desert, consisted of quartz, hornblende, and other minerals, and contained particles having a diameter of 1/500 inch, each weighing 1/200,000 grain. This dust had probably travelled over 600 miles. In the dust from Krakatoa, which fell at Batavia, about 100 miles distant, during the great eruption, there are many solid particles even larger than those mentioned above. Some of this dust was given me by Professor Judd, and I found in it several ovoid particles of a much larger size, being 1/50 inch long, and 1/70 wide and deep. The dust from the same eruption, which fell on board the ship Arabella, 970 miles from the volcano, also contained solid particles 1/500 inch diameter. Mr. John Murray of the Challenger Expedition writes to me that he finds in the deep sea deposits 500 and even 700 miles west of the coast of Africa, rounded particles of quartz, having a diameter of 1/250 inch, and similar particles are found at equally great distances from the south-west coasts of Australia; and he considers these to be atmospheric dust carried to that distance by the wind. Taking the sp. gr. of quartz at 2.6, these particles would weigh about 1/25,000 grain each. These interesting facts can, however, by no means be taken as indicating the extreme limits of the power of wind in carrying solid particles. During the Krakatoa eruption no gale of special violence occurred, and the region is one of comparative calms. The grains of quartz found by Mr. Murray more nearly indicate the limit, but the very small portions of matter brought up by the dredge, as compared with the enormous areas of sea-bottom, over which the atmospheric dust must have been scattered, render it in the highest degree improbable that the maximum limit either of size of particles, or of distance from land has been reached.

Let us, however, assume that the quartz grains, found by Mr. Murray in the deep-sea ooze 700 miles from land, give us the extreme limit of the power of the atmosphere as a carrier of solid particles, and let us compare with these the weights of some seeds. From a small collection of the seeds of thirty species of herbaceous plants sent me from Kew, those in the above table were selected, and small portions of eight of them carefully weighed in a chemical balance.^[13] By counting these portions I was able to estimate the number of seeds weighing one grain. The three very minute species, whose numbers are marked with an asterisk (*), were estimated by the comparison of their sizes with those of the smaller weighed seeds.

No	Species.	Approximate No. of Seeds in one Grain	Approximate Dimensions.	Remarks.
			in.   in.   in.
1	Draba verna	1,800	1/60 x 1/90 x 1/160	Oval, flat.
2	Hypericum perforatum	520	1/30 x 1/80	Cylindrical.
3	Astilbe rivularis	4,500	1/50 x 1/100	Elongate, flat, tailed, wavy.
4	Saxifraga coriophylla	750	1/40 x 1/75	Surface rough, adhere to the dry capsules.
5	Oenothera rosea	640	1/40 x 1/80	Ovate.
6	Hypericum hirsutum	700	1/30 x 1/100	Cylindrical, rough.
7	Mimulus luteus	2,900	1/60 x 1/100	Oval, minute.
8	Penthorum sedoides	8,000*	1/70 x 1/150	Flattened, very minute.
9	Sagina procumbens	12,000*	1/120	Sub-triangular, flat.
10	Orchis maculata	15,000*	---	Margined, flat, very minute.
11	Gentiana purpurea	35	1/25	Wavy, rough, with this coriaceous margins.
12	Silene alpina	---	1/30	Flat, with fringed margins.
13	Adenophora communis	---	1/20 x 1/40	Very thin, wavy, light.
	Quartz grains	25,000	1/250	Deep sea ... 700 miles.
	Do.	200,000	1/500	Genoa ... 600 miles.

If now we compare the seeds with the quartz grains, we find that several are from twice to three times the weight of the grains found by Mr. Murray, and others five times, eight times, and fifteen times as heavy; but they are proportionately very much larger, and, being usually irregular in shape or compressed, they expose a very much larger surface to the air. The surface is often rough, and several have dilated margins or tailed appendages, increasing friction and rendering the uniform rate of falling through still air immensely less than in the case of the smooth, rounded, solid quartz grains. With these advantages it is a moderate estimate that seeds ten times the weight of the quartz grains could be carried quite as far through the air by a violent gale and under the most favourable conditions. These limits will include five of the seeds here given, as well as hundreds of others which do not exceed them in weight; and to these we may add some larger seeds which have other favourable characteristics, as is the case with numbers 11-13, which, though very much larger than the rest, are so formed as in all probability to be still more easily carried great distances by a gale of wind. It appears, therefore, to be absolutely certain that every autumnal gale capable of conveying solid mineral particles to great distances, must also carry numbers of small seeds at least as far; and if this is so, the wind alone will form one of the most effective agents in the dispersal of plants.

Hitherto this mode of conveyance, as applying to the transmission of seeds for great distances across the ocean, has been rejected by botanists, for two reasons. In the first place, there is said to be no direct evidence of such conveyance; and, secondly, the peculiar plants of remote oceanic islands do not appear to have seeds specially adapted for aerial transmission. I will consider briefly each of these objections.

Objection to the Theory of Wind-Dispersal.

To obtain direct evidence of the transmission of such minute and perishable objects, which do not exist in great quantities, and are probably carried to the greatest distances but rarely and as single specimens, is extremely difficult. A bird or insect can be seen if it comes on board ship, but who would ever detect the seeds of Mimulus or Orchis even if a score of them fell on a ship's deck? Yet if but one such seed per century were carried to an oceanic island, that island might become rapidly overrun by the plant, if the conditions were favourable to its growth and reproduction. It is further objected that search has been made for such seeds, and they have not been found. Professor Kerner of Innsbruck examined the snow on the surface of glaciers, and assiduously collected all the seeds he could find, and these were all of plants which grew in the adjacent mountains or in the same district. In like manner, the plants growing on moraines were found to be those of the adjacent mountains, plateaux, or lowlands. Hence he concluded that the prevalent opinion that seeds may be carried through the air for very great distances "is not supported by fact."^[14] The opinion is certainly not supported by Kerner's facts, but neither is it opposed by them. It is obvious that the seeds that would be carried by the wind to moraines or to the surface of glaciers would be, first and in the greatest abundance, those of the immediately surrounding district; then, very much more rarely, those from more remote mountains; and lastly, in extreme rarity, those from distant countries or altogether distinct mountain ranges. Let us suppose the first to be so abundant that a single seed could be found by industrious search on each square yard of the surface of the glacier; the second so scarce that only one could possibly be found in a hundred yards square; while to find one of the third class it would be necessary exhaustively to examine a square mile of surface. Should we expect that one ever to be found, and should the fact that it could not be found be taken as a proof that it was not there? Besides, a glacier is altogether in a bad position to receive such remote wanderers, since it is generally surrounded by lofty mountains, often range behind range, which would intercept the few air-borne seeds that might have been carried from a distant land. The conditions in an oceanic island, on the other hand, are the most favourable, since the land, especially if high, will intercept objects carried by the wind, and will thus cause more of the solid matter to fall on it than on an equal area of ocean. We know that winds at sea often blow violently for days together, and the rate of motion is indicated by the fact that 72 miles an hour was the average velocity of the wind observed during twelve hours at the Ben Nevis observatory, while the velocity sometimes rises to 120 miles an hour. A twelve hours' gale might, therefore, carry light seeds a thousand miles as easily and certainly as it could carry quartz-grains of much greater specific gravity, rotundity, and smoothness, 500 or even 100 miles; and it is difficult even to imagine a sufficient reason why they should not be so carried—perhaps very rarely and under exceptionally favourable conditions,—but this is all that is required.

As regards the second objection, it has been observed that orchideae, which have often exceedingly small and light seeds, are remarkably absent from oceanic islands. This, however, may be very largely due to their extreme specialisation and dependence on insect agency for their fertilisation; while the fact that they do occur in such very remote islands as the Azores, Tahiti, and the Sandwich Islands, proves that they must have once reached these localities either by the agency of birds or by transmission through the air; and the facts I have given above render the latter mode at least as probable as the former. Sir Joseph Hooker remarks on the composite plant of Kerguelen Island (Cotula plumosa) being found also on Lord Auckland and MacQuarrie Islands, and yet having no pappus, while other species of the genus possess it. This is certainly remarkable, and proves that the plant must have, or once have had, some other means of dispersal across wide oceans.^[15] One of the most widely dispersed species in the whole world (Sonchus oleraceus) possesses pappus, as do four out of five of the species which are common to Europe and New Zealand, all of which have a very wide distribution. The same author remarks on the limited area occupied by most species of Compositae, notwithstanding their facilities for dispersal by means of their feathered seeds; but it has been already shown that limitations of area are almost always due to the competition of allied forms, facilities for dispersal being only one of many factors in determining the wide range of species. It is, however, a specially important factor in the case of the inhabitants of remote oceanic islands, since, whether they are peculiar species or not, they or their remote ancestors must at some time or other have reached their present position by natural means.

I have already shown elsewhere, that the flora of the Azores strikingly supports the view of the species having been introduced by aerial transmission only, that is, by the agency of birds and the wind, because all plants that could not possibly have been carried by these means are absent.^[16] In the same way we may account for the extreme rarity of Leguminosae in all oceanic islands. Mr. Hemsley, in his Report on Insular Floras, says that they "are wanting in a large number of oceanic islands where there is no true littoral flora," as St. Helena, Juan Fernandez, and all the islands of the South Atlantic and South Indian Oceans. Even in the tropical islands, such as Mauritius and Bourbon, there are no endemic species, and very few in the Galapagos and the remoter Pacific Islands. All these facts are quite in accordance with the absence of facilities for transmission through the air, either by birds or the wind, owing to the comparatively large size and weight of the seeds; and an additional proof is thus afforded of the extreme rarity of the successful floating of seeds for great distances across the ocean.^[17]

Explanation of North Temperate Plants in the Southern Hemisphere.

If we now admit that many seeds which are either minute in size, of thin texture or wavy form, or so fringed or margined as to afford a good hold to the air, are capable of being carried for many hundreds of miles by exceptionally violent and long-continued gales of wind, we shall not only be better able to account for the floras of some of the remotest oceanic islands, but shall also find in the fact a sufficient explanation of the wide diffusion of many genera, and even species, of arctic and north temperate plants in the southern hemisphere or on the summits of tropical mountains. Nearly fifty of the flowering plants of Tierra-del-Fuego are found also in North America or Europe, but in no intermediate country; while fifty-eight species are common to New Zealand and Northern Europe; thirty-eight to Australia, Northern Europe, and Asia; and no less than seventy-seven common to New Zealand, Australia, and South America.^[18] On lofty mountains far removed from each other, identical or closely allied plants often occur. Thus the fine Primula imperialis of a single mountain peak in Java has been found (or a closely allied species) in the Himalayas; and many other plants of the high mountains of Java, Ceylon, and North India are either identical or closely allied forms. So, in Africa, some species, found on the summits of the Cameroons and Fernando Po in West Africa, are closely allied to species in the Abyssinian highlands and in Temperate Europe; while other Abyssinian and Cameroons species have recently been found on the mountains of Madagascar. Some peculiar Australian forms have been found represented on the summit of Kini Balu in Borneo. Again, on the summit of the Organ mountains in Brazil there are species allied to those of the Andes, but not found in the intervening lowlands.

No Proof of Recent Lower Temperature in the Tropics.

Now all these facts, and numerous others of like character, were supposed by Mr. Darwin to be due to a lowering of temperature during glacial epochs, which allowed these temperate forms to migrate across the intervening tropical lowlands. But any such change within the epoch of existing species is almost inconceivable. In the first place, it would necessitate the extinction of much of the tropical flora (and with it of the insect life), because without such extinction alpine herbaceous plants could certainly never spread over tropical forest {{hws|low|lowlands} lowlands; and, in the next place, there is not a particle of direct evidence that any such lowering of temperature in inter-tropical lowlands ever took place. The only alleged evidence of the kind is that adduced by the late Professor Agassiz and Mr. Hartt; but I am informed by my friend, Mr. J.C. Branner (now State Geologist of Arkansas, U.S.), who succeeded Mr. Hartt, and spent several years completing the geological survey of Brazil, that the supposed moraines and glaciated granite rocks near Rio Janeiro and elsewhere, as well as the so-called boulder-clay of the same region, are entirely explicable as the results of sub-aerial denudation and weathering, and that there is no proof whatever of glaciation in any part of Brazil.

Lower Temperature not needed to Explain the Facts.

But any such vast physical change as that suggested by Darwin, involving as it does such tremendous issues as regards its effects on the tropical fauna and flora of the whole world, is really quite uncalled for, because the facts to be explained are of the same essential nature as those presented by remote oceanic islands, between which and the nearest continents no temperate land connection is postulated. In proportion to their limited area and extreme isolation, the Azores, St. Helena, the Galapagos, and the Sandwich Islands, each possess a fairly rich—the last a very rich—indigenous flora; and the means which sufficed to stock them with a great variety of plants would probably suffice to transmit others from mountain-top to mountain-top in various parts of the globe. In the case of the Azores, we have large numbers of species identical with those of Europe, and others closely allied, forming an exactly parallel case to the species found on the various mountain summits which have been referred to. The distances from Madagascar to the South African mountains and to Kilimandjaro, and from the latter to Abyssinia, are no greater than from Spain to the Azores, while there are other equatorial mountains forming stepping-stones at about an equal distance to the Cameroons. Between Java and the Himalayas we have the lofty mountains of Sumatra and of North-western Burma, forming steps at about the same distance apart; while between Kini Balu and the Australian Alps we have the unexplored snow mountains of New Guinea, the Bellenden Ker mountains in Queensland, and the New England and Blue Mountains of New South Wales. Between Brazil and Bolivia the distances are no greater; while the unbroken range of mountains from Arctic America to Tierra-del-Fuego offers the greatest facilities for transmission, the partial gap between the lofty peak of Chiriqui and the high Andes of New Grenada being far less than from Spain to the Azores. Thus, whatever means have sufficed for stocking oceanic islands must have been to some extent effective in transmitting northern forms from mountain to mountain, across the equator, to the southern hemisphere; while for this latter form of dispersal there are special facilities, in the abundance of fresh and unoccupied surfaces always occurring in mountain regions, owing to avalanches, torrents, mountain-slides, and rock-falls, thus affording stations on which air-borne seeds may germinate and find a temporary home till driven out by the inroads of the indigenous vegetation. These temporary stations may be at much lower altitudes than the original habitat of the species, if other conditions are favourable. Alpine plants often descend into the valleys on glacial moraines, while some arctic species grow equally well on mountain summits and on the seashore. The distances above referred to between the loftier mountains may thus be greatly reduced by the occurrence of suitable conditions at lower altitudes, and the facilities for transmission by means of aerial currents proportionally increased.^[19]

Facts Explained by the Wind-Carriage of Seeds.

But if we altogether reject aerial transmission of seeds for great distances, except by the agency of birds, it will be difficult, if not impossible, to account for the presence of so many identical species of plants on remote mountain summits, or for that "continuous current of vegetation" described by Sir Joseph Hooker as having apparently long existed from the northern to the southern hemisphere. It may be admitted that we can, possibly, account for the greater portion of the floras of remote oceanic islands by the agency of birds alone; because, when blown out to sea land-birds must reach some island or perish, and all which come within sight of an island will struggle to reach it as their only refuge. But, with mountain summits the case is altogether different, because, being surrounded by land instead of by sea, no bird would need to fly, or to be carried by the wind, for several hundred miles at a stretch to another mountain summit, but would find a refuge in the surrounding uplands, ridges, valleys, or plains. As a rule the birds that frequent lofty mountain tops are peculiar species, allied to those of the surrounding district; and there is no indication whatever of the passage of birds from one remote mountain to another in any way comparable with the flights of birds which are known to reach the Azores annually, or even with the few regular migrants from Australia to New Zealand. It is almost impossible to conceive that the seeds of the Himalayan primula should have been thus carried to Java; but, by means of gales of wind, and intermediate stations from fifty to a few hundred miles apart, where the seeds might vegetate for a year or two and produce fresh seed to be again carried on in the same manner, the transmission might, after many failures, be at last effected.

A very important consideration is the vastly larger scale on which wind-carriage of seeds must act, as compared with bird-carriage. It can only be a few birds which carry seeds attached to their feathers or feet. A very small proportion of these would carry the seeds of Alpine plants; while an almost infinitesimal fraction of these latter would convey the few seeds attached to them safely to an oceanic island or remote mountain. But winds, in the form of whirlwinds or tornadoes, gales or hurricanes, are perpetually at work over large areas of land and sea. Insects and light particles of matter are often carried up to the tops of high mountains; and, from the very nature and origin of winds, they usually consist of ascending or descending currents, the former capable of suspending such small and light objects as are many seeds long enough for them to be carried enormous distances. For each single seed carried away by external attachment to the feet or feathers of a bird, countless millions are probably carried away by violent winds; and the chance of conveyance to a great distance and in a definite direction must be many times greater by the latter mode than by the former.^[20] We have seen that inorganic particles of much greater specific gravity than seeds, and nearly as heavy as the smallest kinds, are carried to great distances through the air, and we can therefore hardly doubt that some seeds are carried as far. The direct agency of the wind, as a supplement to bird-transport, will help to explain the presence in oceanic islands of plants growing in dry or rocky places whose small seeds are not likely to become attached to birds; while it seems to be the only effective agency possible in the dispersal of those species of alpine or sub-alpine plants found on the summits of distant mountains, or still more widely separated in the temperate zones of the northern and southern hemispheres.

Concluding Remarks.

On the general principles that have been now laid down, it will be found that all the chief facts of the geographical distribution of animals and plants can be sufficiently understood. There will, of course, be many cases of difficulty and some seeming anomalies, but these can usually be seen to depend on our ignorance of some of the essential factors of the problem. Either we do not know the distribution of the group in recent geological times, or we are still ignorant of the special methods by which the organisms are able to cross the sea. The latter difficulty applies especially to the lizard tribe, which are found in almost all the tropical oceanic islands; but the particular mode in which they are able to traverse a wide expanse of ocean, which is a perfect barrier to batrachia and almost so to snakes, has not yet been discovered. Lizards are found in all the larger Pacific Islands as far as Tahiti, while snakes do not extend beyond the Fiji Islands; and the latter are also absent from Mauritius and Bourbon, where lizards of seven or eight species abound. Naturalists resident in the Pacific Islands would make a valuable contribution to our science by studying the life-history of the native lizards, and endeavouring to ascertain the special facilities they possess for crossing over wide spaces of ocean.

1. See A. Agassiz, Three Cruises of the Blake (Cambridge, Mass., 1888), vol. 1. p. 127, footnote.
2. Even the extremely fine Mississippi mud is nowhere found beyond a hundred miles from the mouths of the river in the Gulf of Mexico (A. Agassiz, Three Cruises of the Blake, vol. i. p. 128).
3. I have given a full summary of the evidence for the permanence of oceanic and continental areas in my Island Life, chap. vi.
4. For a full account of the peculiarities of the Madagascar fauna, see my Island Life, chap. xix.
5. See Island Life, p. 446, and the whole of chaps. xxi. xxii. More recent soundings have shown that the Map at p. 443, as well as that of the Madagascar group at p. 387, are erroneous, the ocean around Norfolk Island and in the Straits of Mozambique being more than 1000 fathoms deep. The general argument is, however, unaffected.
6. For some details of these migrations, see the author's Geographical Distribution of Animals, vol. i. p. 140; also Heilprin's Geographical and Geological Distribution of Animals.
7. For a full discussion of this question, see Island Life, pp. 390-420.
8. Géographie Botanique, p. 798.
9. Nature, 1st April 1886.
10. Report of the Brit. Assoc. Committee on Migration of Birds during 1886.
11. Trans. Ent. Soc., 1871, p. 184.
12. Nature (1875), vol. xii. pp. 279, 298.
13. I am indebted to Professor R. Meldola of the Finsbury Technical Institute, and Rev. T.D. Titmas of Charterhouse for furnishing me with the weights required.
14. See Nature, vol. vi. p. 164, for a summary of Kerner's paper.
15. It seems quite possible that the absence of pappus in this case is a recent adaptation, and that it has been brought about by causes similar to those which have reduced or aborted the wings of insects in oceanic islands. For when a plant has once reached one of the storm-swept islands of the southern ocean, the pappus will be injurious for the same reason that the wings of insects are injurious, since it will lead to the seeds being blown out to sea and destroyed. The seeds which are heaviest and have least pappus will have the best chance of falling on the ground and remaining there to germinate, and this process of selection might rapidly lead to the entire disappearance of the pappus.
16. See Island Life, p. 251.
17. Mr. Hemsley suggests that it is not so much the difficulty of transmission by floating, as the bad conditions the seeds are usually exposed to when they reach land. Many, even if they germinate, are destroyed by the waves, as Burchell noticed at St. Helena; while even a flat and sheltered shore would be an unsuitable position for many inland plants. Air-borne seeds, on the other hand, may be carried far inland, and so scattered that some of them are likely to reach suitable stations.
18. For fuller particulars, see Sir J. Hooker's Introduction to Floras of New Zealand and Australia, and a summary in my Island Life, chaps. xxii. xxiii.
19. For a fuller discussion of this subject, see my Island Life, chap. xxiii.
20. A very remarkable case of wind conveyance of seeds on a large scale is described in a letter from Mr. Thomas Hanbury to his brother, the late Daniel Hanbury, which has been kindly communicated to me by Mr. Hemsley of Kew. The letter is dated "Shanghai, 1st May 1856," and the passage referred to is as follows:—
    "For the past three days we have had very warm weather for this time of year, in fact almost as warm as the middle of summer. Last evening the wind suddenly changed round to the north and blew all night with considerable violence, making a great change in the atmosphere.
    "This morning, myriads of small white particles are floating about in the air; there is not a single cloud and no mist, yet the sun is quite obscured by this substance, and it looks like a white fog in England. I enclose thee a sample, thinking it may interest. It is evidently a vegetable production; I think, apparently, some kind of seed."
    Mr. Hemsley adds, that this substance proves to be the plumose seeds of a poplar or willow. In order to produce the effects described—quite obscuring the sun like a white fog,—the seeds must have filled the air to a very great height; and they must have been brought from some district where there were extensive tracts covered with the tree which produced them.