The eastern and the western islands—with which we are now chiefly concerned—would then differ considerably in their vegetation and animal life. The western and more ancient land already possessed, in its main features, the peculiar Australian flora, and also the ancestral forms of its strange marsupial fauna, both of which it had probably received at some earlier epoch by a temporary union with the Asiatic continent over what is now the Java sea. Eastern Australia, on the other hand, possessed only the rudiments of its existing mixed flora, derived from three distinct sources. Some important fragments of the typical Australian vegetation had reached it across the marine strait, and had spread widely owing to the soil, climate and general conditions being exactly suited to it: from the north and north-east a tropical vegetation of Polynesian type had occupied suitable areas in the north; while the extension southward of the Tasmanian peninsula, accompanied, probably, as now, with lofty mountains, favoured the immigration of south-temperate forms from whatever Antarctic lands or islands then existed. This supposition is strikingly in harmony with what is known of the ancient flora of this portion of Australia. In deposits supposed to be of Eocene age in New South Wales and Victoria fossil plants have been found showing a very different vegetation from that now existing. Along with a few Australian types—such as Pittosporum, Knightia, and Eucalyptus, there occur birches, alders, oaks, and beeches; while in Tasmania in freshwater limestone, apparently of Miocene age, are found willows, alders, birches, oaks, and beeches,^[184] all except the latter genus (Fagus) now quite extinct in Australia.^[185] These temperate forms probably indicate a more oceanic climate, cooler and moister than at present. The union with Western Australia and the establishment of an arid interior by modifying the climate may have led to the extinction of many of these forms and their replacement by special Australian types more suited to the new conditions.

At this time the marsupial fauna had not yet reached this eastern land, which was, however, occupied in the north by some ancestral struthious birds, which had entered it by way of New Guinea through some very ancient continental extension, and of which the emu, the cassowaries, the extinct Dromornis of Queensland, and the moas and kiwis of New Zealand, are the modified descendants.

The Origin of the Australian Element in the New Zealand Flora.—We have now brought down the history of Australia, as deduced from its geological structure and the main features of its existing and Tertiary flora, to the period when New Zealand was first brought into close connection with it, by means of a great north-western extension of that country, which, as already explained in our last chapter, is so clearly indicated by the form of the sea bottom (See Map, p. 471). The condition of New Zealand previous to this event is very obscure. That it had long existed as a more or less extensive land is indicated by its ancient sedimentary rocks; while the very small areas occupied by Jurassic and Cretaceous deposits, imply that much of the present land was then also above the sea-level. The country had probably at that time a scanty vegetation of mixed Antarctic and Polynesian origin; but now, for the first time, it would be open to the free immigration of such Australian types as were suitable to its climate, and which had already reached the tropical and sub-tropical portions of the Eastern Australian island. It is here that we obtain the clue to those strange anomalies and contradictions presented by the New Zealand flora in its relation to Australia, which have been so clearly set forth by Sir Joseph Hooker, and which have so puzzled botanists to account for. But these apparent anomalies cease to present any difficulty when we see that the Australian plants in New Zealand were acquired, not directly, but, as it were, at second hand, by union with an island which itself had as yet only received a portion of its existing flora. And then, further difficulties were placed in the way of New Zealand receiving such an adequate representation of that portion of the flora which had reached East Australia as its climate and position entitled it to, by the fact of the union being, not with the temperate, but with the tropical and sub-tropical portions of that island, so that only those groups could be acquired which were less exclusively temperate, and had already established themselves in the warmer portion of their new home.^[186]

It is therefore no matter of surprise, but exactly what we should expect, that the great mass of pre-eminently temperate Australian genera should be absent from New Zealand, including the whole of such important families as, Dilleniaceæ, Tremandreæ, Buettneriacæ, Polygaleæ, Casuarineæ and Hæmodoraceæ; while others, such as Rutaceæ, Stackhousieæ, Rhamneæ, Myrtaceæ, Proteaceæ, and Santalaceæ, are represented by only a few species. Thus, too, we can explain the absence of all the peculiar Australian Leguminosæ; for these were still mainly confined to the great western island, along with the peculiar Acacias and Eucalypti, which at a later period spread over the whole continent. It is equally accordant with the view we are maintaining, that among the groups which Sir Joseph Hooker enumerates as "keeping up the features of extra tropical Australia in its tropical quarter," several should have reached New Zealand, such as Drosera, some Pittosporeæ and Myoporineæ, with a few Proteaceæ, Loganiaceæ, and Restiaceæ; for most of these are not only found in tropical Australia, but also in the Malayan and Pacific islands.

Tropical Character of the New Zealand Flora Explained.—In this origin of the New Zealand fauna by a north-western route from North-eastern Australia, we find also an explanation of the remarkable number of tropical groups of plants found there: for though, as Sir Joseph Hooker has shown, a moist and uniform climate favours the extension of tropical forms in the temperate zone, yet some means must be afforded them for reaching a temperate island. On carefully going through the Handbook, and comparing its indications with those of Bentham's Flora Australiensis, I find that there are in New Zealand thirty-eight thoroughly tropical genera, thirty-three of which are found in Australia—mostly in the tropical portion of it, though a few are temperate, and these may have reached it through New Zealand^[187]. To these we must add thirty-two more genera, which, though chiefly developed in temperate Australia, extend into the tropical or sub-tropical portions of it, and may well have reached New Zealand by the same route.

On the other hand we find but few New Zealand genera certainly derived from Australia which are especially temperate, and it may be as well to give a list of such as do occur with a few remarks. They are sixteen in number, as follows:—

1. Pennantia (1 sp.). This genus has a species in Norfolk Island, indicating perhaps its former extension to the north-west.

2. Pomaderris (3 sp.). One species inhabits Victoria and New Zealand, indicating recent trans-oceanic migration.

3. Quintinia (2 sp.). This genus has winged seeds facilitating migration.

4. Olearia (20 sp.). Seeds with pappus.

5. Craspedia (2 sp.). Seeds with pappus. Alpine; identical with Australian species, and therefore of comparatively recent introduction.

6. Celmisia (25 sp.). Seeds with pappus. Only three Australian species, two of which are identical with New Zealand forms, probably therefore derived from New Zealand.

7. Ozothamnus (5 sp.). Seeds with pappus.

8. Epacris (4 sp.). Minute seeds. Some species are sub-tropical, and they are all found in the northern (warmer) island of New Zealand.

9. Archeria (2 sp.). Minute seeds. A species common to E. Australia and New Zealand.

10. Logania (3 sp.). Small seeds. Alpine plants.

11. Hedycarya (1 sp.).

12. Chiloglottis (1 sp.). Minute seeds. In Auckland Islands; alpine in Australia.

13. Prasophyllum (1 sp.). Minute seeds. Identical with Australian species, indicating recent transmission.

14. Orthoceras (1 sp.). Minute seeds. Identical with an Australian species.

15. Alepyrum (1 sp.). Alpine, moss-like. An Antarctic type.

16. Dichelachne (3 sp.). Identical with Australian species. An awned grass.

We thus see that there are special features in most of these plants that would facilitate transmission across the sea between temperate Australia and New Zealand, or to both from some Antarctic island; and the fact that in several of them the species are absolutely identical shows that such transmission has occurred in geologically recent times.

Species Common to New Zealand and Australia Mostly Temperate Forms.—Let us now take the species which are common to New Zealand and Australia, but found nowhere else, and which must therefore have passed from one country to the other at a more recent period than the mass of genera with which we have hitherto been dealing. These are ninety-six in number, and they present a striking contrast to the similarly restricted genera in being wholly temperate in character, the entire list presenting only a single species which is confined to sub-tropical East Australia—a grass (Apera arundinacea) only found in a few localities on the New Zealand coast.

Now it is clear that the larger portion, if not the whole, of these plants must have reached New Zealand from Australia (or in a few cases Australia from New Zealand), by transmission across the sea, because we know there has been no actual land connection during the Tertiary period, as proved by the absence of all the Australian mammalia, and almost all the most characteristic Australian birds, insects, and plants. The form of the sea-bed shows that the distance could not have been less than 600 miles, even during the greatest extension of Southern New Zealand and Tasmania; and we have no reason to suppose it to have been less, because in other cases an equally abundant flora of identical species has reached islands at a still greater distance—notably in the case of the Azores and Bermuda. The character of the plants is also just what we should expect: for about two-thirds of them belong to genera of world-wide range in the temperate zones, such as Ranunculus, Drosera, Epilobium, Gnaphalium, Senecio, Convolvulus, Atriplex, Luzula, and many sedges and grasses, whose exceptionally wide distribution shows that they possess exceptional powers of dispersal and vigour of constitution, enabling them not only to reach distant countries, but also to establish themselves there. Another set of plants belong to especially Antarctic or south temperate groups, such as Colobanthus, Acæna, Gaultheria, Pernettya, and Muhlenbeckia, and these may in some cases have reached both Australia and New Zealand from some now submerged Antarctic island. Again, about one-fourth of the whole are alpine plants, and these possess two advantages as colonisers. Their lofty stations place them in the best position to have their seeds carried away by winds; and they would in this case reach a country which, having derived the earlier portion of its flora from the side of the tropics, would be likely to have its higher mountains and favourable alpine stations to a great extent unoccupied, or occupied by plants unable to compete with specially adapted alpine groups.

Fully one-third of the exclusively Australo-New Zealand species belong to the two great orders of the sedges and the grasses; and there can be no doubt that these have great facilities for dispersion in a variety of ways. Their seeds, often enveloped in chaffy glumes, would be carried long distances by storms of wind, and even if finally dropped into the sea would have so much less distance to reach the land by means of surface currents; and Mr. Darwin's experiments show that even cultivated oats germinated after 100 days' immersion in sea-water. Others have hispid awns by which they would become attached to the feathers of birds, and there is no doubt this is an effective mode of dispersal. But a still more important point is, probably, that these plants are generally, if not always, wind-fertilised, and are thus independent of any peculiar insects, which might be wanting in the new country.

Why Easily-Dispersed Plants have often Restricted Ranges.—This last consideration throws light on a very curious point, which has been noted as a difficulty by Sir Joseph Hooker, that plants which have most clear and decided powers of dispersal by wind or other means, have not generally the widest specific range; and he instances the small number of Compositæ common to New Zealand and Australia. But in all these cases it will, I think, be found that although the species have not a wide range the genera often have. In New Zealand, for instance, the Compositæ are very abundant, there being no less than 167 species, almost all belonging to Australian genera, yet only about one-sixteenth of the whole are identical in the two countries. The explanation of this is not difficult. Owing to their great powers of dispersal, the Australian Compositæ reached New Zealand at a very remote epoch, and such as were adapted to the climate and the means of fertilisation established themselves; but being highly organised plants with great flexibility of organisation, they soon became modified in accordance with the new conditions, producing many special forms in different localities; and these, spreading widely, soon took possession of all suitable stations. Henceforth immigrants from Australia had to compete with these indigenous and well-established plants, and only in a few cases were able to obtain a footing; whence it arises that we have many Australian types, but few Australian species, in New Zealand, and both phenomena are directly traceable to the combination of great powers of dispersal with a high degree of adaptability. Exactly the same thing occurs with the still more highly specialised Orchideæ. These are not proportionally so numerous in New Zealand (thirty-eight species), and this is no doubt due to the fact that so many of them require insect-fertilisation often by a particular family or genus (whereas almost any insect will fertilise Compositæ), and insects of all orders are remarkably scarce in New Zealand.^[188] This would at once prevent the establishment of many of the orchids which may have reached the islands, while those which did find suitable fertilisers and other favourable conditions would soon become modified into new species. It is thus quite intelligible why only three species of orchids are identical in Australia and New Zealand, although their minute and abundant seeds must be dispersed by the wind almost as readily as the spores of ferns.

Another specialised group—the Scrophularineæ—abounds in New Zealand, where there are sixty-two species; but though almost all the genera are Australian only three species are so. Here, too, the seeds are usually very small, and the powers of dispersal great, as shown by several European genera—Veronica, Euphrasia, and Limosella, being found in the southern hemisphere.

Looking at the whole series of these Australo-New Zealand plants, we find the most highly specialised groups—Compositæ, Scrophularineæ, Orchideæ—with a small proportion of identical species (one-thirteenth to one twentieth), the less highly specialised—Ranunculaceæ, Onagrariæ and Ericeæ—with a higher proportion (one-ninth to one-sixth), and the least specialised—Junceæ, Cyperaceæ and Gramineæ—with the high proportion in each case of one-fourth. These nine are the most important New Zealand orders which contain species common to that country and Australia and confined to them; and the marked correspondence they show between high specialisation and want of specific identity, while the generic identity is in all cases approximately equal, points to the conclusion that the means of diffusion are, in almost all plants ample, when long periods of time are concerned, and that diversities in this respect are not so important in determining the peculiar character of a derived flora, as adaptability to varied conditions, great powers of multiplication, and inherent vigour of constitution. This point will have to be more fully discussed in treating of the origin of the Antarctic and north temperate members of the New Zealand flora.

Summary and Conclusion on the New Zealand Flora.—Confining ourselves strictly to the direct relations between the plants of New Zealand and of Australia, as I have done in the preceding discussion, I think I may claim to have shown that the union between the two countries in the latter part of the Secondary epoch at a time when Eastern Australia was widely separated from Western Australia (as shown by its geological formation and by the contour of the sea-bottom) does sufficiently account for all the main features of the New Zealand flora. It shows why the basis of the flora is fundamentally Australian both as regards orders and genera, for it was due either to a direct land connection or a somewhat close approximation between the two countries. It shows also why the great mass of typical Australian forms are unrepresented, for the Australian flora is typically western and temperate, and New Zealand received its immigrants from the eastern island which had itself received only a fragment of this flora, and from the tropical end of this island, and thus could only receive such forms as were not exclusively temperate in character. It shows, further, why New Zealand contains such a very large proportion of tropical forms, for we see that it derived the main portion of its flora directly from the tropics. Again, this hypothesis shows us why, though the specially Australian genera in New Zealand are largely tropical or sub-tropical, the specially Australian species are wholly temperate or alpine; for these are comparatively recent arrivals, they must have migrated across the sea in the temperate zone, and these temperate and alpine forms are exactly such as would be best able to establish themselves in a country already stocked mainly by tropical forms and their modified descendants. This hypothesis further fulfils the conditions implied in Sir Joseph Hooker's anticipation that—"these great differences (of the floras) will present the least difficulties to whatever theory may explain the whole case,"—for it shows that these differences are directly due to the history and development of the Australian flora itself, while the resemblances depend upon the most certain cause of all such broad resemblances—close proximity or actual land connection.

One objection will undoubtedly be made to the above theory,—that it does not explain why some species of the prominent Australian genera Acacia, Eucalyptus, Melaleuca, Grevillea, &c., have not reached New Zealand in recent times along with the other temperate forms that have established themselves. But it is doubtful whether any detailed explanation of such a negative fact is possible, while general explanations sufficient to cover it are not wanting. Nothing is more certain than that numerous plants never run wild and establish themselves in countries where they nevertheless grow freely if cultivated; and the explanation of this fact given by Mr. Darwin—that they are prevented doing so by the competition of better adapted forms—is held to be sufficient. In this particular case, however, we have some very remarkable evidence of the fact of their non-adaptation. The intercourse between New Zealand and Europe has been the means of introducing a host of common European plants,—more than 150 in number, as enumerated at the end of the second volume of the Handbook; yet, although the intercourse with Australia has probably been greater, only two or three Australian plants have similarly established themselves. More remarkable still, Sir Joseph Hooker states: "I am informed that the late Mr. Bidwell habitually scattered Australian seeds during his extensive travels in New Zealand." We may be pretty sure that seeds of such excessively common and characteristic groups as Acacia and Eucalyptus would be among those so scattered, yet we have no record of any plants of these or other peculiar Australian genera ever having been found wild, still less of their having spread and taken possession of the soil in the way that many European plants have done. We are, then, entitled to conclude that the plants above referred to have not established themselves in New Zealand (although their seeds may have reached it) because they could not successfully compete with the indigenous flora which was already well established and better adapted to the conditions of climate and of the organic environment. This explanation is so perfectly in accordance with a large body of well-known facts, including that which is known to every one—how few of our oldest and hardiest garden plants ever run wild—that the objection above stated will, I feel convinced, have no real weight with any naturalists who have paid attention to this class of questions.

CHAPTER XXIII

ON THE ARCTIC ELEMENT IN SOUTH TEMPERATE FLORAS

European Species and Genera of Plants in the Southern Hemisphere—Aggressive Power of the Scandinavian Flora—Means by which Plants have Migrated from North to South—Newly moved Soil as Affording Temporary Stations to Migrating Plants—Elevation and Depression of the Snow-line as Aiding the Migration of Plants—Changes of Climate Favourable to Migration—The Migration from North to South has been long going on—Geological Changes as Aiding Migration—Proofs of Migration by way of the Andes—Proofs of Migration by way of the Himalayas and Southern Asia—Proofs of Migration by way of the African Highlands—Supposed Connection of South Africa and Australia—The Endemic Genera of Plants in New Zealand—The Absence of Southern Types from the Northern Hemisphere—Concluding Remarks on the New Zealand and South Temperate Floras.

We have now to deal with another portion of the New Zealand flora which presents perhaps equal difficulties—that which appears to have been derived from remote parts of the north and south temperate zones; and this will lead us to inquire into the origin of the northern or Arctic element in all the south temperate floras.

More than one-third of the entire number of New Zealand genera (115) are found also in Europe, and even fifty-eight species are identical in these remote parts of the world. Temperate South America has seventy-four genera in common with New Zealand, and there are even eleven species identical in the two countries, as well as thirty-two which are close allies or representative species. A considerable number of these northern or Antarctic plants and many more which are representative species, are found also in Tasmania and in the mountains of temperate Australia; and Sir Joseph Hooker gives a list of thirty-eight species very characteristic of Europe and Northern Asia, but almost or quite unknown in the warmer regions, which yet reappear in temperate Australia. Other genera seem altogether Antarctic—that is, confined to the extreme southern lands and islands; and these often have representative species in Southern America, Tasmania, and New Zealand, while others occur only in one or two of these areas. Many north temperate genera also occur in the mountains of South Africa. On the other hand, few if any of the peculiar Australian or Antarctic types have spread northwards, except some of the former which have reached the mountains of Borneo, and a few of the latter which spread along the Andes to Mexico.

On these remarkable facts, of which I have given but the barest outline, Sir Joseph Hooker makes the following suggestive observations:—

"When I take a comprehensive view of the vegetation of the Old World, I am struck with the appearance it presents of there being a continuous current of vegetation (if I may so fancifully express myself) from Scandinavia to Tasmania; along, in short, the whole extent of that arc of the terrestrial sphere which presents the greatest continuity of land. In the first place Scandinavian genera, and even species, reappear everywhere from Lapland and Iceland to the tops of the Tasmanian Alps, in rapidly diminishing numbers it is true, but in vigorous development throughout. They abound on the Alps and Pyrenees, pass on to the Caucasus and Himalayas, thence they extend along the Khasia Mountains, and those of the peninsulas of India to those of Ceylon and the Malayan Archipelago (Java and Borneo), and after a hiatus of 30° they appear on the Alps of New South Wales, Victoria, and Tasmania, and beyond these again on those of New Zealand and the Antarctic Islands, many of the species remaining unchanged throughout! It matters not what the vegetation of the bases and flanks of these mountains may be; the northern species may be associated with alpine forms of Germanic, Siberian, Oriental, Chinese, American, Malayan, and finally Australian, and Antarctic types; but whereas these are all, more or less, local assemblages, the Scandinavian asserts his prerogative of ubiquity from Britain to beyond its antipodes."^[189]

It is impossible to place the main facts more forcibly before the reader than in the above striking passage. It shows clearly that this portion of the New Zealand flora is due to wide-spread causes which have acted with even greater effect in other south temperate lands, and that in order to explain its origin we must grapple with the entire problem of the transfer of the north temperate flora to the southern hemisphere. Taking, therefore, the facts as given by Sir Joseph Hooker in the works already referred to, I shall discuss the whole question broadly, and shall endeavour to point out the general laws and subordinate causes that, in my opinion, have been at work in bringing about the anomalous phenomena of distribution he has done so much to make known and to elucidate.

Aggressive Power of the Scandinavian Flora.—The first important fact bearing upon this question is the wonderful aggressive and colonising power of the Scandinavian flora, as shown by the way in which it establishes itself in any temperate country to which it may gain access. About 150 species have thus established themselves in New Zealand, often taking possession of large tracts of country; about the same number are found in Australia, and nearly as many in the Atlantic states of America, where they form the commonest weeds. Whether or not we accept Mr. Darwin's explanation of this power as due to development in the most extensive land area of the globe where competition has been most severe and long-continued, the fact of the existence of this power remains, and we can see how important an agent it must be in the formation of the floras of any lands to which these aggressive plants have been able to gain access.

But not only are these plants pre-eminently capable of holding their own in any temperate country in the world, but they also have exceptional powers of migration and dispersal over seas and oceans. This is especially well shown by the case of the Azores, where no less than 400 out of a total of 478 flowering plants are identical with European species. These islands are more than 800 miles from Europe, and, as we have already seen in Chapter XII., there is no reason for supposing that they have ever been more nearly connected with it than they are now, since an extension of the European coast to the 1,000-fathom line would very little reduce the distance. Now it is a most interesting and suggestive fact that more than half the European genera which occur in the Australian flora occur also in the Azores, and in several cases even the species are identical in both.^[190] The importance of such a case as this cannot be exaggerated, because it affords a demonstration of the power of the very plants in question to pass over wide areas of sea, some no doubt wholly through the air, carried by storms in the same way as the European birds and insects which annually reach the Azores, others by floating on the waters, or by a combination of the two methods; while some may have been carried by aquatic birds, to whose feathers many seeds have the power of attaching themselves, and some even in the stomachs of fruit or seed eating birds. We have in such facts as these a complete disproof of the necessity for those great changes of sea and land which are continually appealed to by those who think land-connection the only efficient means of accounting for the migration of animals or plants; but at the same time we do not neglect to make the fullest use of such moderate changes as all the evidence at our command leads us to believe have actually occurred, and especially of the former existence of intermediate islands, so often indicated by shoals in the midst of the deepest oceans.

Means by which Plants have migrated from North to South.—But if plants can thus pass in considerable numbers and variety over wide seas and oceans, it must be yet more easy for them to traverse continuous areas of land, whereever mountain-chains offer suitable stations at moderate intervals on which they might temporarily establish themselves. The facilities afforded for the transmission of plants by mountains has hardly received sufficient attention. The numerous land-slips, the fresh surfaces of broken rock and precipice, the debris of torrents, and the moraines deposited by glaciers, afford numerous unoccupied stations on which wind-borne seeds have a good chance of germinating. It is a well-known fact that fresh surfaces of soil or rock, such as are presented by railway cuttings and embankments, often produce plants strange to the locality, which survive for a few years, and then disappear as the normal vegetation gains strength and permanence.^[191] But such a surface will, in the meantime, have acted as a fresh centre of dispersal; and thus a plant might pass on step by step, by means of stations temporarily occupied, till it reached a district where, the general conditions being more favourable, it was able to establish itself as a permanent member of the flora. Such, generally speaking, was probably the process by which the Scandinavian flora has made its way to the southern hemisphere; but it could hardly have done so to any important extent without the aid of those powerful causes explained in our eighth chapter—causes which acted as a constantly recurrent motive-power to produce that "continuous current of vegetation" from north to south across the whole width of the tropics referred to by Sir Joseph Hooker. Those causes were, the repeated changes of climate which, during all geological time, appear to have occurred in both hemispheres, culminating at rare intervals in glacial epochs, and which have been shown to depend upon changes of excentricity of the earth's orbit and the occurrence of summer or winter in aphelion, in conjunction with the slower and more irregular changes of geographical conditions; these combined causes acting chiefly through the agency of heat-bearing oceanic currents, and of snow- and ice-collecting highlands. Let us now briefly consider how such changes would act in favouring the dispersal of plants.

Elevation and Depression of the Snow Line as Aiding the Migration of Plants.—We have endeavoured to show (in an earlier portion of this volume) that wherever geographical or physical conditions were such as to produce any considerable amount of perpetual snow, this would be increased whenever a high degree of excentricity concurred with winter in aphelion, and diminished during the opposite phase. On all mountain ranges, therefore, which reached above the snow-line, there would be a periodical increase and decrease of snow, and when there were extensive areas of plateau at about the same level, the lowering of the snow-line might cause such an increased accumulation of snow as to produce great glaciers and ice-fields, such as we have seen occurred in South Africa during the last period of high excentricity. But along with such depression of the line of perpetual snow there would be a corresponding depression of the alpine and sub-alpine zones suitable for the growth of an arctic and temperate vegetation, and, what is perhaps more important, the depression would necessarily produce a great extension of the area of these zones on all high mountains, because as we descend the average slopes become less abrupt,—thus affording a number of new stations suitable for such temperate plants as might first reach them. But just above and below the snow-line is the area of most powerful disintegration and denudation, from the alternate action of frost and sun, of ice and water; and thus the more extended area would be subject to the constant occurrence of land-slips, berg-falls, and floods, with their accompanying accumulations of débris and of alluvial soil, affording innumerable stations in which solitary wind-borne seeds might germinate and temporarily establish themselves.

This lowering and rising of the snow-line each 10,500 years during periods of high excentricity, would occur in the northern and southern hemispheres alternately; and where there were high mountains within the tropics the two would probably overlap each other, so that the northern depression would make itself felt in a slight degree even across the equator some way into the southern hemisphere, and vice versâ; and even if the difference of the height of perpetual snow at the two extremes did not average more than a few hundred feet, this would be amply sufficient to supply the new and unoccupied stations needful to facilitate the migration of plants. It is well known that all great mountain ranges have undergone such fluctuations, as proved by ice-marks below the present level of snow and ice.

But the differences of temperature in the two hemispheres caused by the sun being in perihelion in the winter of the one while it was in aphelion during the same season in the other, would necessarily lead to increased aërial and marine currents, as already explained; and whenever geographical conditions were such as to favour the production of glaciation in any area these effects would become more powerful, and would further aid in the dispersal of the seeds of plants.

Changes of Climate Favourable to Migration.—It is clear then, that during periods when no glacial epochs were produced in the northern hemisphere, and even when a mild climate extended over the whole polar area, alternate changes of climate favouring the dispersal of plants would occur on all high mountains, and with particular force on such as rise above the snow-line. But during that long-continued, though comparatively recent, phase of high excentricity which produced an extensive glaciation in the northern hemisphere and local glaciations in the southern, these risings and lowerings of the snow-line on all mountain ranges would have been at a maximum, and would have been increased by the depression of the ocean which must have arisen from such a vast bulk of water being locked up in land-ice, and which depression would have produced the same effect as a general elevation of all the continents. At this time, too, aërial currents would have attained their maximum of force in both hemispheres; and this would greatly facilitate the dispersal of all wind-borne seeds as well as of those carried in the plumage or in the stomachs of birds, since we have seen, by the cases of the Azores and Bermuda, how vastly the migratory powers of birds are increased by a stormy atmosphere.

Migration from North to South has been long going on.—Now, if each phase of colder and warmer mountain-climate—each alternate depression and elevation of the snow-line, only helped on the migration of a few species some stages of the long route from the north to the south temperate regions, yet, during the long course of the Tertiary period there might well have arisen that representation of the northern flora in the southern hemisphere which is now so conspicuous. For it is very important to remark that it is not the existing flora alone that is represented, such as might have been conveyed during the last glacial epoch only; but we find a whole series of northern types evidently of varying degrees of antiquity, while even some genera characteristic of the southern hemisphere appear to have been originally derived from Europe. Thus Eucalyptus and Metrosideros have been determined by Dr. Ettingshausen from their fruits in the Eocene beds of Sheppey, while Pimelea, Leptomeria and four genera of Proteaceæ have been recognised by Professor Heer in the Miocene of Switzerland; and the former writer has detected fifty-five Australian forms in the Eocene plant beds of Häring (? Belgium).^[192] Then we have such peculiar genera as Pachychladon and Notothlaspi of New Zealand said to have affinities with Arctic plants, while Stilbocarpa—another peculiar New Zealand genus—has its nearest allies in the Himalayan and Chinese Aralias. Following these are a whole host of very distinct species of northern genera which may date back to any part of the Tertiary period, and which occur in every south temperate land. Then we have closely allied representative species of European or Arctic plants; and, lastly, a number of identical species,—and these two classes are probably due entirely to the action of the last great glacial epoch, whose long continuance, and the repeated fluctuations of climate with which it commenced and terminated, rendered it an agent of sufficient power to have brought about this result.

Here, then, we have that constant or constantly recurrent process of dispersal acting throughout long periods with varying power—that "continuous current of vegetation" as it has been termed, which the facts demand; and the extraordinary phenomenon of the species and genera of European and even of Arctic plants being represented abundantly in South America, Australia, and New Zealand, thus adds another to the long series of phenomena which are rendered intelligible by frequent alternations of warmer and colder climates in either hemisphere, culminating, at long intervals and in favourable situations, in actual glacial epochs.

Geological Changes as Aiding Migration.—It will be well also to notice here, that there is another aid to dispersion dependent upon the changes effected by denudation during the long periods included in the duration of the species and genera of plants. A considerable number of the plants of the Miocene period of Europe were so much like existing species that although they have generally received fresh names they may well have been identical; and a large proportion of the vegetation during the whole Tertiary period consisted of genera which are still living.^[193] But from what is now known of the rate of sub-aërial denudation, we are sure, that during each division of this period many mountain chains must have been considerably lowered, while we know that some of the existing ranges have been greatly elevated. Ancient volcanoes, too, have been destroyed by denudation, and new ones have been built up, so that we may be quite sure that ample means for the transmission of temperate plants across the tropics, may have existed in countries where they are now no longer to be found. The great mountain masses of Guiana and Brazil, for example, must have been far more lofty before the sedimentary covering was denuded from their granitic bosses and metamorphic peaks, and may have aided the southern migration of plants before the final elevation of the Andes. And if Africa presents us with an example of a continent of vast antiquity, we may be sure that its great central plateaux once bore far loftier mountain ranges before they were reduced to their present condition by long ages of denudation.

Proofs of Migration by Way of the Andes.—We are now prepared to apply the principles above laid down to the explanation of the character and affinities of the various portions of the north temperate flora in the southern hemisphere, and especially in Australia and New Zealand.

At the present time the only unbroken chain of highlands and mountains connecting the Arctic and north temperate with the Antarctic lands is to be found in the American continent, the only break of importance being the comparatively low Isthmus of Panama, where there is a distance of about 300 miles occupied by rugged forest-clad hills, between the lofty peaks of Veragua and the northern extremity of the Andes of New Grenada. Such distances are, as we have already seen, no barrier to the diffusion of plants; and we should accordingly expect that this great continuous mountain-chain has formed the most effective agent in aiding the southward migration of the Arctic and north temperate vegetation. We do find, in fact, not only that a large number of northern genera and many species are scattered all along this line of route, but that at the end of the long journey, in Southern Chile and Fuegia, they have established themselves in such numbers as to form an important part of the flora of those countries. From the lists given in the works already referred to, it appears that there are between sixty and seventy northern genera in Fuegia and Southern Chile, while about forty of the species are absolutely identical with those of Europe and the Arctic regions. Considering how comparatively little the mountains of South Temperate America are yet known, this is a very remarkable result, and it proves that the transmission of species must have gone on up to comparatively recent times. Yet, as only a few of these species are now found along the line of migration, we see that they only occupied such stations temporarily; and we may connect their disappearance with the passing away of the last glacial period which, by raising the snow-line, reduced the area on which alone they could exist, and exposed them to the competition of indigenous plants from the belt of country immediately below them.

Now, just as these numerous species and genera have undoubtedly passed along the great American range of mountains, although only now found at its two extremes, so others have doubtless passed on further; and have found more suitable stations or less severe competition in the Antarctic continent and islands, in New Zealand, in Tasmania, and even in Australia itself. The route by which they may have reached these countries is easily marked out. Immediately south of Cape Horn, at a distance of only 500 miles, are the South Shetland Islands and Graham's Land, whence the Antarctic continent or a group of large islands probably extends across or around the south polar area to Victoria Land and thence to Adélie Land. The outlying Young Island, 12,000 feet high, is about 750 miles south of the Macquarie Islands, which may be considered a southern outlier of the New Zealand group; and the Macquarie Islands are about the same distance from the 1,000-fathom line at a point marking the probable southern extension of Tasmania. Other islands may have existed at intermediate points; but, even as it is, these distances are not greater than we know are traversed by plants both by flotation and by aërial currents, especially in such a stormy atmosphere as that of the Antarctic regions. Now, we may further assume, that what we know occurred within the Arctic circle also took place in the Antarctic—that is, that there have been alternations of climate during which some portion of what are now ice-clad lands became able to support a considerable amount of vegetation.^[194] During such periods there would be a steady migration of plants from all southern circumpolar countries to people the comparatively unoccupied continent, and the southern extremity of America being considerably the nearest, and also being the best stocked with those northern types which have such great powers of migration and colonisation, such plants would form the bulk of the Antarctic vegetation, and during the continuance of the milder southern climate would occupy the whole area.

When the cold returned and the land again became ice-clad, these plants would be crowded towards the outer margins of the Antarctic land and its islands, and some of them would find their way across the sea to such countries as offered on their mountain summits suitable cool stations; and as this process of alternately receiving plants from Chile and Fuegia and transmitting them in all directions from the central Antarctic land may have been repeated several times during the Tertiary period, we have no difficulty in understanding the general community between the European and Antarctic plants found in all south temperate lands. Kerguelen's Land and The Crozets are within about the same distance from the Antarctic continent as New Zealand and Tasmania, and we need not therefore be surprised at finding in each of these islands some Fuegian species which have not reached the others. Of course, there will remain difficulties of detail, as there always must remain, so long as our knowledge of the past changes of the earth's surface and the history of the particular plants concerned is so imperfect. Sir Joseph Hooker notes, for example, the curious fact that several Compositæ common to three such remote localities as the Auckland Islands, Fuegia, and Kerguelen's Land, have no pappus or seed-down, while such as have pappus are in no case common even to two of these islands. Without knowing the exact history and distribution of the genera to which these plants belong it would be useless to offer any conjecture, except that they are ancient forms which may have survived great geographical changes, or may have some peculiar and exceptional means of dispersion.

Proofs of Migration by way of the Himalayas and Southern Asia.—But although we may thus explain the presence of a considerable portion of the European element in the floras of New Zealand and Australia, we cannot account for the whole of it by this means, because Australia itself contains a host of European and Asiatic genera of which we find no trace in New Zealand or South America, or any other Antarctic land. We find, in fact, in Australia two distinct sets of European plants. First we have a number of species identical with those of Northern Europe or Asia (of the most characteristic of which—thirty-eight in number—Sir Joseph Hooker gives a list); and in the second place a series of European genera usually of a somewhat more southern character, mostly represented by very distinct species, and all absent from New Zealand; such as Clematis, Papaver, Cleome, Polygala, Lavatera, Ajuga, &c. Now of the first set—the North European species—about three-fourths occur in some parts of America, and about half in South Temperate America or New Zealand; whence we may conclude that most of these, as well as some others, have reached Australia by the route already indicated. The second set of Australo-European genera, however, and many others characteristic of the South European or the Himalayan flora, have probably reached Australia by way of the mountains of Southern Asia, Borneo, the Moluccas, and New Guinea, at a somewhat remote period when loftier ranges and some intermediate peaks may have existed, sufficient to carry on the migration by the aid of the alternate climatal changes which are known to have occurred. The long belt of Secondary and Palæozoic formations in East Australia from Tasmania to Cape York continued by the lofty ranges of New Guinea, indicates the route of this immigration, and sufficiently explains how it is that these northern types are almost wholly confined to this part of the Australian continent. Some of the earlier immigrants of this class no doubt passed over to New Zealand and now form a portion of the peculiar genera confined to these two countries; but most of them are of later date, and have thus remained in Australia only.

Proofs of Migration by way of the African Highlands.—It is owing to this twofold current of vegetation flowing into Australia by widely different routes that we have in this distant land a better representation of the European flora, both as regards species and genera, than in any other part of the southern hemisphere; and, so far as I can judge of the facts, there is no general phenomenon—that is, nothing in the distribution of genera and other groups of plants as opposed to cases of individual species—that is not fairly accounted for by such an origin. It further receives support from the case of South Africa, which also contains a large and important representation of the northern flora. But here we see no indications (or very slight ones) of that southern influx which has given Australia such a community of vegetation with the Antarctic lands. There are no less than sixty genera of strictly north temperate plants in South Africa, none of which occur in Australia; while very few of the species, so characteristic of Australia, New Zealand, and Fuegia, are found there. It is clear, therefore, that South Africa has received its European plants by the direct route through the Abyssinian highlands and the lofty equatorial mountains, and mostly at a distant period when the conditions for migration were somewhat more favourable than they are now. The much greater directness of the route from Northern Europe to South Africa than to Australia; and the existence even now of lofty mountains and extensive highlands for a large portion of the distance, will explain (what Sir Joseph Hooker notes as "a very curious fact") why South Africa has more very northern European genera than Australia, while Australia has more identical species and a better representation on the whole of the European flora—this being clearly due to the large influx of species it has received from the Antarctic Islands, in addition to those which have entered it by way of Asia. The greater distance of South Africa even now from any of these islands, and the much deeper sea to the south of the African continent, than in the case of Tasmania and New Zealand, indicating a smaller recent extension southward, is all quite in harmony with the facts of distribution of the northern flora above referred to.

Supposed Connection of South Africa and Australia.—There remains, however, the small amount of direct affinity between the vegetation of South Africa and that of Australia, New Zealand, and Temperate South America, consisting in all of fifteen genera, five of which are confined to Australia and South Africa, while several natural orders are better represented in these two countries than in any other part of the world. This resemblance has been supposed to imply some former land-connection of all the great southern lands, but it appears to me that any such supposition is wholly unnecessary. The differences between the faunas and floras of these countries are too great and too radical to render it possible that any such connection should have existed except at a very remote period. But if we have to go back so far for an explanation, a much simpler one presents itself, and one more in accordance with what we have learnt of the general permanence of deep oceans and the great changes that have taken place in the distribution of all forms of life. Just as we explain the presence of marsupials in Australia and America and of Centetidæ in Madagascar and the Antilles, by the preservation in these localities of remnants of once wide-spread types, so we should prefer to consider the few genera common to Australia and South Africa as remnants of an ancient vegetation, once spread over the northern hemisphere, driven southward by the pressure of more specialised types, and now finding a refuge in these two widely separated southern lands. It is suggestive of such an explanation that these genera are either of very ancient groups—as Conifers and Cycads—or plants of low organisation as the Restiaceæ—or of world-wide distribution, as Melanthaceæ.

The Endemic Genera of Plants in New Zealand.—Returning now to the New Zealand flora, with which we are more especially concerned, there only remains to be considered the peculiar or endemic genera which characterise it. These are thirty-two in number, and are mostly very isolated. A few have affinities with Arctic groups, others with Himalayan, or Australian genera; several are tropical forms, but the majority appear to be altogether peculiar types of world-wide groups—as Leguminosæ, Saxifrageæ, Compositæ, Orchideæ, &c. We must evidently trace back these peculiar forms to the earliest immigrants, either from the north or from the south; and the great antiquity we are obliged to give to New Zealand—an antiquity supported by every feature in its fauna and flora, no less than by its geological structure, and its extinct forms of life^[195]—affords ample time for the changes in the general distribution of plants, and for those due to isolation and modification under the influence of changed conditions, which are manifested by the extreme peculiarity of many of these interesting endemic forms.

The Absence of Southern Types from the Northern Hemisphere.—We have now only to notice the singular want of reciprocity in the migrations of northern and southern types of vegetation. In return for the vast number of European plants which have reached Australia, not one single Australian plant has entered any part of the north temperate zone, and the same may be said of the typical southern vegetation in general, whether developed in the Antarctic lands, New Zealand, South America, or South Africa. The furthest northern outliers of the southern flora are a few genera of Antarctic type on the Bornean Alps; the genus Acæna which has a species in California; two representatives of the Australian flora—Casuarina and Stylidium, in the peninsula of India; while China and the Philippines have two strictly Australian genera of Orchideæ—Microtis and Thelymitra, as well as a Restiaceous genus. Several distinct causes appear to have combined to produce this curious inability of the southern flora to make its way into the northern hemisphere. The primary cause is, no doubt, the totally different distribution of land in the two hemispheres, so that in the south there is the minimum of land in the colder parts of the temperate zone and in the north the maximum. This is well shown by the fact that on the parallel of Lat. 50° N. we pass over 240° of land or shallow sea, while on the same parallel of south latitude we have only 4°, where we cross the southern part of Patagonia. Again the three most important south temperate land-areas—South Temperate America, South Africa, and Australia—are widely separated from each other, and have in all probability always been so; whereas the whole of the north temperate lands are practically continuous. It follows that, instead of the enormous northern area, in which highly organised and dominant groups of plants have been developed gifted with great colonising and aggressive powers, we have in the south three comparatively small and detached areas, in which rich floras have been developed with special adaptations to soil, climate, and organic environment, but comparatively impotent and inferior beyond their own domain.

Another circumstance which makes the contest between the northern and southern forms still more unequal, is the much greater hardiness of the former, from having been developed in a colder region, and one where alpine and arctic conditions extensively prevail; whereas the southern floras have been mainly developed in mild regions to which they have been altogether confined. While the northern plants have been driven north or south by each succeeding change of climate, the southern species have undergone comparatively slight changes of this nature, owing to the areas they occupy being unconnected with the ice-bearing Antarctic continent. It follows, that whereas the northern plants find in all these southern lands a milder and more equable climate than that to which they have been accustomed, and are thus often able to grow and flourish even more vigorously than in their native land, the southern plants would find in almost every part of Europe, North America or Northern Asia, a more severe and less equable climate, with winters that usually prove fatal to them even under cultivation. These causes, taken separately, are very powerful, but when combined they must, I think, be held to be amply sufficient to explain why examples of the typical southern vegetation are almost unknown in the north temperate zone, while a very few of them have extended so far as the northern tropic.^[196]

Concluding Remarks on the Last Two Chapters.—Our inquiry into the external relations and probable origin of the fauna and flora of New Zealand, has thus led us on to a general theory as to the cause of the peculiar biological relations between the northern and the southern hemispheres; and no better or more typical example could be found of the wide range and great interest of the study of the geographical distribution of animals and plants.

The solution which has here been given of one of the most difficult of this class of problems, has been rendered possible solely by the knowledge very recently obtained of the form of the sea-bottom in the southern ocean, and of the geological structure of the great Australian continent. Without this knowledge we should have nothing but a series of guesses or probabilities on which to found our hypothetical explanation, which we have now been able to build up on a solid foundation of fact. The complete separation of East from West Australia during a portion of the Cretaceous and Tertiary periods, could never have been guessed till it was established by the laborious explorations of the Australian geologists; while the hypothesis of a comparatively shallow sea, uniting New Zealand by a long route with tropical Australia, while a profoundly deep ocean always separated it from temperate Australia, would have been rejected as too improbable a supposition for the foundation of even the most enticing theory. Yet it is mainly by means of these two facts, that we are enabled to give an adequate explanation of the strange anomalies in the flora of Australia and its relation to that of New Zealand.

In the more general explanation of the relations of the various northern and southern floras, I have shown what an important aid to any such explanation is the theory of repeated changes of climate, not necessarily of great amount, given in Chapters VIII. and IX.; while the whole discussion justifies the importance attached to the theory of the general permanence of continents and oceans, as demonstrated in Chapter VI., since any rational explanation based upon facts (as opposed to mere unsupported conjecture) must take such general permanence as a starting-point. The whole inquiry into the phenomena presented by islands, which forms the main subject of the present volume has, I think, shown that this theory does afford a firm foundation for the discussion of questions of distribution and dispersal; and that by its aid, combined with a clear perception of the wonderful powers of dispersion and modification in the organic world when long periods are considered, the most difficult problems connected with this subject cease to be insoluble.

CHAPTER XXIV

SUMMARY AND CONCLUSION

The Present Volume is the Development and Application of a Theory—Statement of the Biological and Physical Causes of Dispersal—Investigation of the Facts of Dispersal—of the Means of Dispersal—of Geographical Changes Affecting Dispersal—of Climatal Changes Affecting Dispersal—The Glacial Epoch and its Causes—Alleged Ancient Glacial Epochs—Warm Polar Climates and their Causes—Conclusions as to Geological Climates—How far Different from those of Mr. Croll—Supposed Limitations of Geological Time—Time Amply Sufficient both for Geological and Biological Development—Insular Faunas and Floras—The North Atlantic Islands—The Galapagos—St. Helena and the Sandwich Islands—Great Britain as a Recent Continental Island—Borneo and Java—Japan and Formosa—Madagascar as an Ancient Continental Island—Celebes and New Zealand as Anomalous Islands—The Flora of New Zealand and its Origin—The European Element in the South Temperate Floras—Concluding Remarks.

The present volume has gone over a very wide field both of facts and theories, and it will be well to recall these to the reader's attention and point out their connection with each other, in a concluding chapter. I hope to be able to show that, although at first sight somewhat fragmentary and disconnected, this work is really the development of a clear and definite theory, and its application to the solution of a number of biological problems. That theory is, briefly, that the distribution of the various species and groups of living things over the earth's surface, and their aggregation in definite assemblages in certain areas, is the direct result and outcome of a complex set of causes, which may be grouped as "biological" and "physical." The biological causes are mainly of two kinds—firstly, the constant tendency of all organisms to increase in numbers and to occupy a wider area, and their various powers of dispersion and migration through which, when unchecked, they are enabled to spread widely over the globe; and, secondly, those laws of evolution and extinction which determine the manner in which groups of organisms arise and grow, reach their maximum, and then dwindle away, often breaking up into separate portions which long survive in very remote regions. The physical causes are also mainly of two kinds. We have, first, the geographical changes which at one time isolate a whole fauna and flora, at another time lead to their dispersal and intermixture with adjacent faunas and floras—and it was here important to ascertain and define the exact nature and extent of these changes, and to determine the question of the general stability or instability of continents and oceans; in the second place, it was necessary to determine the exact nature, extent and frequency of the changes of climate which have occurred in various parts of the earth,—because such changes are among the most powerful agents in causing the dispersal and extinction of plants and animals. Hence the importance attached to the question of geological climates and their causes, which have been here investigated at some length with the aid of the most recent researches of geologists, physicists, and explorers. These various inquiries led on to an investigation of the mode of formation of stratified deposits, with a view to fix within some limits their probable age; and also to an estimate of the probable rate of development of the organic world; and both these processes are shown to involve, so far as we can judge, periods of time less vast than have generally been thought necessary.

The numerous facts and theories established in the First Part of the work are then applied to explain the phenomena presented by the floras and faunas of the chief islands of the globe, which are classified, in accordance with their physical origin, in three groups or classes, each of which are shown to exhibit certain well-marked biological features.

Having thus shown that the work is a connected whole, founded on the principle of tracing out the more recondite causes of the distribution of organisms, we will briefly indicate the scope and object of the several chapters, by means of which this general conception has been carried out.

Beginning with simple and familiar facts relating to British and European quadrupeds and birds, I have defined and shown the exact character of "areas of distribution," as applied to species, genera, and families, and have illustrated the subject by maps showing the peculiarities of distribution of some well-known groups of birds. Taking then our British mammals and land-birds, I follow them over the whole area they inhabit, and thus obtain a foundation for the establishment of "zoological regions," and a clear insight into their character as distinct from the usual geographical divisions of the globe.

The facts thus far established are then shown to be necessary results of the "law of evolution." The nature and amount of "variation" is exhibited by a number of curious examples; the origin, growth, and decay of species and genera are traced, and all the interesting phenomena of isolated groups and discontinuous generic and specific areas are shown to follow as logical consequences.

The next subject investigated is the means by which the various groups of animals are enabled to overcome the natural barriers which often seem to limit them to very restricted areas, how far those barriers are themselves liable to be altered or abolished, and what is the exact nature and amount of the changes of sea and land which our earth has undergone in past times. This latter part of the inquiry is shown to be the most important as it is the most fundamental; and as it is still a subject of controversy, and many erroneous views prevail in regard to it, it is discussed at some length. Several distinct classes of evidence are adduced to prove that the grand features of our globe—the position of the great oceans and the chief land-areas—have remained, on the whole, unchanged throughout geological time. Our continents are shown to be built up mainly of "shore-deposits"; and even the chalk, which is so often said to be the exact equivalent of the "globigerina ooze" now forming in mid-Atlantic, is shown to be a comparatively shallow-water deposit formed in inland seas, or in the immediate vicinity of land. The general stability of continents has, however, been accompanied by constant changes of form, and insular conditions have prevailed over every part in succession; and the effect of such changes on the distribution of organisms is pointed out.

We then approach the consideration of another set of changes—those of climate—which have probably been agents of the first importance in modifying the specific forms as well as the distribution of animals. Here again we find ourselves in the midst of fierce controversies. The occurrence of a recent glacial epoch of great severity in the northern hemisphere is now universally admitted, but the causes which brought it on are matter of dispute. But unless we can arrive at these causes, as well as at those which produced the equally well demonstrated mild climate in the Arctic regions, we shall be quite unable to determine the nature and amount of the changes of climate which have occurred throughout past ages, and shall thus be left without a most important clue to the explanation of many of the anomalies in the distribution of animals and plants.

I have therefore devoted three chapters to a full investigation of this question. I have first given such a sketch of the most salient facts as to render the phenomena of the glacial epoch clear and intelligible. I then review the various suggested explanations, and, taking up the two which alone seem tenable, I endeavour to determine the true principles of each. While adopting generally Mr. Croll's views as to the causes of the "glacial epoch," I have introduced certain limitations and modifications. I have pointed out, I believe, more clearly than has hitherto been done, the very different effects on climate of water in the liquid and in the solid state; and I have shown, by a variety of evidence, that without high land there can be no permanent snow and ice. From these facts and principles the very important conclusion is reached, that the alternate phases of precession—causing the winter of each hemisphere to be in aphelion and perihelion each 10,500 years—would produce a complete change of climate only where a country was partially snow-clad; while, whenever a large area became almost wholly buried in snow and ice—as was certainly the case with Northern Europe and America during the glacial epoch—then the glacial conditions would be continued and perhaps even intensified when the sun approached nearest to the earth in winter, instead of there being at that time, as Mr. Croll maintains, an almost perpetual spring. This important result is supported by reference to the existing differences between the climates of the northern and southern hemispheres, and by what is known to have occurred during the last glacial epoch; and it is shown to be in complete harmony with the geological evidence as to interglacial mild periods.

Discussing next the evidence for glacial epochs in earlier times, it is shown that Mr. Croll's views are opposed by a vast body of facts, and that the geological evidence leads irresistibly to the conclusion that during a large portion of the Secondary and Tertiary periods, uninterrupted warm climates prevailed in the north temperate zone, and so far ameliorated the climate of the Arctic regions as to admit of the growth of a luxuriant vegetation in the highest latitudes yet explored. The geographical condition of the northern hemisphere at these periods is then investigated, and it is shown to have been probably such as to admit the warm tropical waters freely to penetrate the land, and to reach the Arctic seas by several channels; and, adopting Mr. Croll's calculations as to the enormous quantity of heat that would thus be conveyed northwards, it is maintained that the mild Arctic climates are amply accounted for. With such favourable geographical conditions, it is shown, that changes of excentricity and of the phases of precession would have no other effect than to cause greater differences of temperature between summer and winter; but, wherever there was a considerable extent of very lofty mountains the snow-line would be lowered, and the snow-collecting area being thus largely increased a considerable amount of local glaciation might result. Thus may be explained the presence of enormous ice-borne rocks in Eocene and Miocene times in Central Europe, while at the very same period all the surrounding country enjoyed a tropical or sub-tropical climate.