Fig. 5.—Map of Europe, with arrows indicating approximately the course taken by the different streams of migration towards the British Islands.

The course of events in the origin of the British fauna might have been therefore somewhat as follows:—In early Tertiary times, when the climate all over Western Europe was moist and semi-tropical, a migration proceeded northward from the south-western corner of Europe. This was strengthened by Oriental migrants which had moved westward along the Mediterranean basin (Fig. 5, No. 1). Owing to geographical changes supervening, the Alpine fauna (No. 2) was then enabled to colonise the British Islands, and subsequently another migration had begun to come in from the south-east (No. 3). The climate had meanwhile gradually become more temperate and drier. About the same time, or even earlier, an Arctic migration commenced to pass southward (No. 4), and finally the Siberian animals (No. 5) poured into our continent. The arrows in the map indicate the directions followed by the different migrants as they travelled to the British Islands. The arrows are not meant to represent the whole nor the full extent of the migrations from any particular centre, but only in so far as they affect our islands. Moreover, it would be impossible to indicate on one map the geographical conditions which obtained during the several migrations. It must be remembered that during the time which elapsed while they passed into the British Islands, these were joined in the north to Scandinavia and in the south to Belgium and France. The various phases of geographical evolution of Europe will be studied in the subsequent chapters, and maps will then be given to show as far as possible in a general way the leading characteristics of these great changes.

I have now given some reasons for the belief that several different migrations of animals entered the British Islands in later Tertiary times. I have also shown why some of them must be looked upon as being older than others, and in so far we have come to a decision as to their relative ages. It still remains for us, however, to examine how their geological ages can be approximately determined. We require for this purpose palæontological aid.

In the fifth chapter will be found the history of the Siberian migration. And since we possess most valuable records of it in the numerous fossil remains discovered in Central and Western Europe, we are able to trace their progress from the east to the west in a very complete and satisfactory manner. In England their first appearance dates from the Forest-Bed, for here we find remains of the Glutton (Gulo luscus), Musk-Ox (Ovibos moschatus), and others (see p. 204). It seems reasonable to suppose, therefore, that the first entry of these Siberian mammals into Europe took place at or just before the Forest-Bed period. But Professor Nehring tells us in his remarkable work on the Tundra and Steppes (p. 222), that in Germany the remains of the same mammals occur in deposits which are certainly more recent than the lower continental boulder clay; and he is inclined to the belief that they migrated into Europe during the inter-glacial phase which is supposed to have separated the earlier from the later stage of the Glacial period. It is evident that in this case the inter-glacial period in Germany would have corresponded to, and be contemporaneous with, our Forest-Bed period. The deposits immediately preceding the Forest-Bed would also be contemporaneous with the lower continental boulder clay. Although this may seem rather a startling statement to make, from the evidence which will be brought forward in the fourth and fifth chapters I am inclined to the belief that such is probably the case.

Having once arrived at a determination of the exact geological period during which the Siberian mammals invaded our continent, and having also previously determined the relative ages of the various other migrations, we have advanced another step in the direction we are aiming at. Let us suppose that the Siberian migration actually reached the British Islands during the Forest-Bed period. Since the Siberian migration is the most recent of those which entered the British Islands, the others must have commenced their march before the Forest-Bed period. Now it was Professor Boyd Dawkins who first indicated to us, as I have remarked before, the method of research to be adopted in an attempt to determine the geological age of the different migrations in so far as they affected the British Islands. I may be excused, therefore, for again quoting the following important passage in one of his works. "The absence," he says (b, p. xxix), "of the beaver and the dormouse from Ireland must be due to the existence of some barrier to their westward migration from the adjacent mainland, and the fact that the Alpine hare is indigenous, while the common hare is absent, implies that, so far as relates to the former animal, the barrier did not exist." The Beaver, Dormouse, and Common Hare are either Siberians or later migrants from elsewhere, and there can be no doubt that at the Forest-Bed period Ireland was already, or was just being, separated from England. All the southern species, that is to say all the Lusitanian, Alpine, and Oriental forms occurring in Ireland, must therefore be older than that period. I have advocated similar views in a former essay on this subject. Mr. Carpenter recently advanced some interesting and valuable criticisms on these views, which we may examine a little more closely (p. 385). "While, then," he remarks, "I find myself in almost complete agreement with Dr. Scharff with regard to the older sections of our fauna, I think that those widespread species which survived the Glacial period must have been confined to the more southern parts of our area, and have only subsequently spread northwards and westwards to Scotland and Ireland." He suggests, in fact, that the widespread British species belong to a younger or newer section of our fauna than the local ones. In many cases this may be quite true, but we possess also a large number of common and widely-spread forms which bear the impress of antiquity upon them. We have the most positive proof of the antiquity of the very common small circular Snail (Helix rotundata), since it was found in miocene freshwater deposits near Bordeaux. Many other examples might be mentioned to show that, though discontinuous range is generally a proof of antiquity, continuous range is not always a sign of the opposite. Some species, in fact, appear to be short-lived and disinclined to spread, whilst others multiply rapidly even under a change of temperature and climate, and are to be found almost everywhere. But even if we supposed, with Mr. Carpenter, that these widely-ranging species must have been confined during the Glacial period to the more southern parts of England, the idea that they afterwards made their way northwards along the eastern shore of the Irish Sea and then passed into Ireland, does not appeal to me. Southern England was occupied at that very same time by an assemblage of Siberian mammals. Mr. Carpenter thinks these might have been kept out of Ireland by an arm of the sea until the land-connection with North-western England had broken down. But if an arm of the sea could keep out the Siberian mammals it would also keep out the widely-spread British species of the general fauna. On the other hand, I quite admit that my view of the survival in Ireland of the pre-glacial fauna is somewhat difficult to accept, considering that we have such undoubted evidence of a very extensive submergence. The case of Isle of Man, quoted by Mr. Carpenter, can be met, I think, by the supposition that it was connected with Cumberland until quite recently, and quite independently of any connection between England and Ireland; that the Isle of Man, in fact, was always a cape or peninsula of the mainland, and only recently became separated by local subsidences or by the action of the sea.

Part of the history of the British fauna will be referred to again in the next chapter, which deals with the Arctic migration. We need not therefore dwell any longer on this subject here. There is one matter, however, which is of importance in connection with the geographical conditions of the British Islands at the time when the greater portion of our fauna arrived from abroad.

On page 60 will be found a map indicating the physical geography of that part of the ancient continent on which what are now the British Islands were situated. Only one large river has been marked on that map, namely, that flowing out of a lake which occupied part of the Irish Sea. Another probably discharged its waters into the Atlantic midway between France and England, whilst the Thames may have been a tributary of the Rhine, as it emptied itself into the sea near our south-east coast. I have shown in a previous essay that the former presence of a freshwater lake between England and Ireland is indicated by the distribution of the Charrs and also by the various species of British Coregonus. There are three British species of Coregonus, viz., C. clupeoides, C. vandesius, and C. pollan. These are confined to the lakes of North Wales, North-western England, South-western Scotland, and Ireland. All but the latter communicate at present directly with the Irish Sea. The lakes of the latter country, however, must have done so at a time when the west of Ireland stood at a higher level than it does now. The ancestors of the three Coregonus species, and also those of the Charrs, then lived in the large freshwater lake indicated on the map (p. 60), and when the sea gradually crept up the river valley and finally converted the lake into a gulf, the freshwater fish took refuge in the rivers which supplied it with water.

Now as for the continuous sea-shore between the coast of Brittany and the south-west of Ireland, zoological distribution again aids us in proving that such must have actually existed at no very distant geological date. Most of our common shore forms of life migrate along the coast exactly as land animals do—step by step. Their eggs are carefully attached to fixed objects, so as not to be carried away by the waves, whilst the young often remain and grow old in some particular little pool, rarely venturing farther than a few yards from the spot where they first saw the light of day. A number of such shore forms are found on the west coast of France, the same species recurring again on the south-west coasts of England and Ireland, thus clearly indicating a former continuity of coast-line between these points, now separated by deep sea. A very familiar example to British zoologists is the purple rock-boring Sea-urchin (Strongylocentrotus lividus), but there are a great many others, such as the semi-marine Beetles Octhebius Lejolisii and Æpophilus Bonnairei, the Crustaceans Achæus Cranchii, Inachus leptochirus, Gonoplax angulata, Thia assidua, Callianassa subterranea, the Fishes Blennius galerita and Lepadogaster Decandollii, and the Molluscs Otina otis, Donax politus, and Amphidesma castaneum.

Before concluding this chapter, a few words as to my views on the conditions prevailing during the Glacial period will not be out of place. They do not differ very much from those held formerly by most geologists; and even at present there are, as I have mentioned before, a few upholders of those older views.

The sea, I think, must have gradually crept across England from the east during, or shortly after, the Forest-Bed period, so as to separate the south from the north, whilst Ireland and Scotland were then still connected with one another. At a later stage, the sea also partially invaded Ireland, and this condition is very roughly represented on the accompanying map. Mr. Kendall kindly drew my attention to the fact that several notable areas on which shelly drift has been observed are here placed upon the land; but it must be remembered that one stage only can be shown on the map, and that the sea covered more ground a little later. Many of the smaller islands in the glacial sea, too, are not shown. The map, in fact, is merely meant to give a general idea of the manner in which the great northern sea moved westward and slowly covered a large portion of the British Islands. These peculiar geographical conditions explain, I think, better than anything, the absence from parts of the Midlands and the north of England of such a number of terrestrial invertebrates which are otherwise widely distributed over the British Islands. In spite of the fact that a large portion of the British Islands became submerged, we possessed at that time an extensive area which has since been claimed by the sea, so that there was ample room for the present fauna to survive the Glacial period. The climate during this period was probably much the same as it is at present, though moister, with cooler summers and milder winters.

Fig. 6.—Map of the British Islands, showing approximately in what manner the sea may have invaded the country from the east during, or shortly after, the Forest-Bed period. The darkly shaded parts indicate the areas covered by water, and the lightly shaded and white portions what was land at that time.

It may be asked what proof we have of such an extensive submergence of England and Ireland. My own views are principally based on the general distribution of the fauna in the British Islands, and the belief that nothing but a mild climate during the Glacial period could have brought it about. On purely geological grounds, however, some geologists, notably Mr. Mellard Reade, have come to a similar conclusion. "The whole of Lancashire and Cheshire," he remarks (a, p. 542), "from sea-level up to about 400 feet, and in places 600 feet, is covered by a continuous mantle of boulder-clay and sands." "These clays, as a rule, contain distributed through them, in a greater or less degree, fragments of shells and some perfect ones. I myself have recorded forty-four species." Again he continues (pp. 545 and 546): "A large part of Ayrshire is covered with similar shelly boulder-clays from sea-level to 1061 feet at Dippal. These Ayrshire high-level shells have, in the majority of cases, been taken, not from sand and gravel beds, but from boulder-clay, and in that respect they are most important and unique. In Moel Tryfan the shells are found in sands and gravels at 982 feet; on the range of hills from Miaera to Llangollen from 1000-1200 feet; also in sands and gravels at Gloppa, near Oswestry, at 1100-1200 feet; and near Macclesfield at a level of about 1200 feet. In Ireland marine shells can be traced almost from sea-level to a height of over 1000 feet."

"Again," continues the same author, "if we look broadly at the distribution of these shelly deposits, we find that they occur all round our maritime coasts in Lancashire, Cheshire, and Wales, in Cumberland and Westmoreland, Wigtonshire and Ayrshire, and along the eastern coast of Ireland. The same is to be said of the eastern coasts of England and Scotland."

That a very considerable change of sea-level has taken place in some parts of the British Islands would appear to a zoologist the most logical conclusion after an examination of these "high-level shelly sands and gravels," but the shells contained in them are now generally supposed to have been carried there frozen in the sole of a glacier or pushed up in front of it. The older view, however, which agrees so much better with the facts of distribution, fortunately has not disappeared among geologists. "When we call up," says Mr. Mellard Reade (b, p. 435), "before our mental vision the simple and well-known facts of nature which suffice to explain the marine drifts on the theory of submergence, it seems unnecessary to resort to the ingenious and artificial system of physics elaborated to explain the phenomena of land-ice."

"When we have more knowledge of the glaciers of the Arctic Regions, and facts, in place of ingenious suppositions, to base our reasoning upon, we may possibly have to revise all our glacial conceptions. In the meantime, the submergence theory of the origin of high-level shelly gravels and sands seems to me by far the simpler of the two theories, and the most consistent with the facts and phenomena which the labours of a succession of enthusiastic geologists have made us acquainted with."

Among those geologists, and they form the majority, who hold that Ireland was covered by land-ice, there is a great diversity of opinion as to its extent. Messrs. Close, Kinahan, J. Geikie, and others believe that the ice covered practically everything, whilst others who claim to have examined the ground with equal care, such as Professor Carvill Lewis, were led to believe that the south of Ireland, with the exception of a few local glaciers, was free from ice. The glacial phenomena of the country can therefore be interpreted in different ways, even by those who are convinced that they are due to land-ice and not to icebergs or mud-glaciers.

SUMMARY OF CHAPTER III.

The history of the British fauna is not only of interest to us from a sentimental point of view, it is a convenient starting-point in the study of the larger European problem. The fauna, broadly speaking, is composed of three foreign elements, viz., the northern, eastern, and southern, to which may be added a small endemic one. Examples are given of the more noteworthy forms belonging to each of these. This leads us to the subject of the natural divisions of the British Islands according to their animal inhabitants. Zoologists attempted at first to subdivide these countries, on the lines laid down by botanists, into a large number of provinces. Forbes proposed ten such divisions for mollusca, and subsequently five, which were ultimately reduced by others to two or three.

The opinions of biologists are almost unanimous in attributing the bulk of the British fauna and flora to migrations by land from the Continent, but two other theories, viz., those of Professor Cole and Messrs. Kinahan and Lamplugh, are also referred to. The first believes in a possible migration eastward from Western Europe, and the latter support the view of the former existence of ice-bridges to assist the fauna in their migrations.

An endeavour is next made to determine at what geological periods the various migrations entered the British Islands. There is considerable difference of opinion on this subject. Some believe that the British fauna is altogether post-glacial; a few think that it is partly so and the remainder glacial; others again hold that a portion is pre-glacial and the rest glacial and post-glacial. Those who have studied the subject most closely feel convinced that the south-western or Lusitanian fauna, and also the flora, must have arrived before the Glacial period and survived the latter in these Islands. It seems reasonable to suppose, therefore, that the climate cannot have been very severe during the so-called Ice-Age. This Lusitanian fauna must be looked upon as the oldest portion of the British fauna. The Alpine and Oriental migrations arrived next. After these came the Arctic, and finally the Eastern or Siberian. As the fossil evidence is most complete with regard to the last, we are able to determine with precision not only the direction whence this migration came, but approximately its geological age. It arrived in Germany from the east after the deposition of the lower boulder-clay. Since the boulder-clay is looked upon as a glacial deposit, the Siberian migration reached Central Europe after the first portion of the Glacial period had passed. In England it makes its first appearance in the Forest-Bed, which would therefore correspond to the "Loess" formation of Central Europe. All the other migrations are older than the Siberian. They must therefore have come to Great Britain during the earlier part of the Glacial period or before it.

The chapter concludes with a short statement on the physical geography of the British Islands during the time when these migrations entered them. That there existed a continuous coast-line between France and Ireland is proved by the occurrence of a considerable number of identical shore species, whilst the former existence of a freshwater lake on the site of the present Irish Sea is indicated by the distribution of some freshwater fishes.

CHAPTER IV.
THE ARCTIC FAUNA.

There are six typical Polar Land-mammals, one of which, the Polar Bear, is semi-aquatic. The Reindeer (Rangifer tarandus) occurs upon almost all the polar lands, and it has often been a source of speculation in what manner it has reached such remote islands as Spitsbergen and Novaya Zemlya—the former of the two being so remote from a continent. There is no doubt that Reindeer are great wanderers, owing to the difficulty of finding sufficient food-supply for the large herds in which they are accustomed to travel; and for this reason they can cross, and have been known to cross, distances of from ten to twenty miles on ice. The Behring Straits, when frozen over in winter, is frequently traversed by them. But I quite agree with Dr. Brauer (p. 260) that it is impossible to account for their presence in Spitsbergen by an immigration from either Novaya Zemlya, Greenland, or Scandinavia, under the present geographical conditions. The seas between the former island and the other land-masses referred to are rarely entirely frozen over. Even if this should occur, the distances between Spitsbergen and Greenland, Novaya Zemlya, or Scandinavia are so great, that a migration across ice is quite excluded from the range of possibilities, since Reindeer could not subsist without food during the time it would take to travel from one to the other. The manner in which it did reach Spitsbergen and Greenland will be discussed more fully below, and I will therefore proceed to mention the other Arctic mammals.

One of the most important and most typical species is the Polar Bear (Ursus maritimus), the greater part of whose life is spent on the ice and in the sea. The fact that its favourite nourishment consists of seals proves its excellent and keen faculties of sight and hearing, and its facility in swimming. But it is not a dainty feeder, and lives upon almost all animals which come within its reach; birds, land-mammals, or fish are not despised in times of scarcity. Its fur throughout the year is coloured white, though in old bears it assumes a more yellowish hue.

Fig. 7.—The Musk-Ox (Ovibos moschatus). (From Flower & Lydekker's Mammals, p. 358. London: Adam & Chas. Black.)

Another large mammal, perhaps less well known, is the Musk-Ox (Ovibos moschatus, Fig. 7), which resembles in size the smaller varieties of Oxen, but in structure and habits is closely allied to the Sheep. As is implied by the specific name, it exhales a musky odour; this does not, however, appear to be due to the secretion of a special gland, as is the case in other animals with a similar smell. The skin is covered with long brown thickly-matted hair, interspersed with white. It is confined to the most northerly parts of North America and the American Arctic islands, and to North Greenland. Though not now living in the Old World, it seems formerly to have been abundant in Siberia, and, as we shall learn later on, it was one of the species which took part in the great Siberian invasion of Europe. Its remains have been found not only in Germany and France, but also in the south of England.

The Polar Fox (Canis lagopus) occurs throughout the Polar Regions, and on islands where even the Reindeer and the Musk-Ox are unknown. Beyond the Polar Circle, its range extends into Northern Asia, to the extreme north of North America, and the mountains of Scandinavia. Like its congeners, it had in pleistocene times a more southerly extension, and fossil remains have been met with in various parts of continental Europe and in England.

The Stoat (Mustela erminea), which is known and much valued in commerce under the name of Ermine, was formerly believed to occur only in Arctic America and the northern parts of the Old World, but in more recent years it has been discovered in a number of the northern islands, such as Saghalien, in the islands of the Behring Straits, the Aleutian islands, and also in Greenland and Spitsbergen. In Europe, it is found as far south as the Arctic Hare, or perhaps even farther, and it flourishes in the Alps up to a height of 9000 feet. It offers a parallel to the Arctic Hare in the fact that in some countries, such as Ireland, it only rarely turns white in winter. The Irish form of the Stoat differs so much from the English, that Messrs. Thomas and Barrett-Hamilton are of opinion that it is specifically distinct, as I mentioned in speaking of the divisions of the British fauna (p. 90).

The Arctic Hare (Lepus variabilis) is almost the only one of the typical Arctic mammals which still inhabits the British Islands, and for that reason it is to most of us more familiar than any of the preceding species. Hares have been described from Greenland by the name of Lepus glacialis, from the European Alps as Lepus alpinus, and under other names from Arctic North America; but though slight differences in the fur and even in the skull can be pointed out, there is no doubt that all these are only varieties or races of what, in the British Islands, is known as the Irish or the Scotch Mountain Hare, Lepus variabilis. In the Arctic Regions this Hare remains white throughout the year, but in Scandinavia and some other parts its fur becomes brown in the summer, and in Ireland it frequently remains entirely brown during the whole year, and never, or only in very rare cases, becomes entirely white in winter. Besides Scandinavia, Scotland, and Ireland, it is found in Northern Russia, and also in the Pyrenees, the Alps, and the Caucasus. In Asia it occurs not only on the mainland of Siberia, but it has been obtained on the Akita Mountains in Japan and on the Mioko San Mountain, and also on the island of Saghalien. It had in former times a more extensive range, and its remains have been discovered in England and in a number of places on the continent of Europe. The peculiarity of its range, which will be explained more fully directly, lies in the fact of the occurrence of isolated colonies in the mountains of Europe, in Ireland and Scotland, and in the mountains of Japan (Fig. 8). From a distributional point of view, it is one of the most interesting species of mammals, and its history throws a flood of light on the geographical changes which have occurred in former times.

Fig. 8.—Map of the northern hemisphere, to show the geographical distribution of the Arctic Hare (Lepus variabilis) indicated in black.

One more species must be mentioned, and that is the Banded Lemming (Cuniculus torquatus), which occurs chiefly in Arctic America, Northern Siberia, and Greenland. Though frequently mistaken for the Scandinavian Lemming, there is a striking difference in the character of the teeth, which has induced zoologists to put them into distinct genera. The Arctic Lemming, moreover, is distinguished from the Scandinavian by the absence of external ears, the densely furred feet, and by the great length of the two middle claws in the fore-feet. There are two species of the true Lemming, namely, the one just referred to, Myodus lemmus, and Myodus obensis. These may be looked upon as more or less Arctic species, since they occur within the Polar Circle, but they are not so exclusively confined to that region as the Banded Lemming (Cuniculus torquatus). The remains of both Cuniculus torquatus and of Myodus lemmus have been found in British pleistocene deposits.

Until recently no Lemming remains had been found to the south of France, but Mr. Barrett-Hamilton announced to us a short time since that Dr. Gadow had discovered some skeletons with their skins still preserved in a cave in Northern Portugal. These were found to belong to the Scandinavian Lemming (M. lemmus), and the author incidentally expressed the opinion that there was some possibility of this species still inhabiting the mountains of Spain.

The Lemming multiplies with great rapidity under favourable conditions. In speaking of his experiences in Siberia Dr. Brehm says (p. 79): "All the young of the first litter of the various Lemming females thrive, and six weeks later at the most these also multiply. Meanwhile the parents have brought forth a second and a third litter, and these in their turn bring forth young. Within three months the heights and low grounds of the tundra teem with lemmings, just as our fields do with mice under similar circumstances. Whichever way we turn we see the busy little creatures, dozens at a single glance, thousands in the course of an hour. But the countless and still increasing numbers prove their own destruction. Soon the lean tundra ceases to afford employment enough for their greedy teeth. Famine threatens, perhaps actually sets in. The anxious animals crowd together and begin their march, hundreds join with hundreds, thousands with other thousands, the troops become swarms, the swarms armies. They travel in a definite direction, at first following old tracks, but soon striking out new ones; in unending files—defying all computation—they hasten onwards; over the cliffs they plunge into the water. Thousands fall victims to want and hunger; the army behind streams on over their corpses; hundreds of thousands are drowned in the water or are shattered at the foot of the cliffs; the remainder speed on; other hundreds and thousands fall victims to the voracity of Arctic and red foxes, wolves and gluttons, rough-legged buzzards and ravens, owls and skuas which have followed them; the survivors pay no heed. Where these go, how they end, none can say; but certain it is, that the tundra behind them is as if dead, that a number of years pass ere the few who have remained behind and have managed to survive slowly multiply and visibly re-people their native fields." This eloquent passage reminds us of the manner in which migrations of all kinds of animals have taken place in former times, and are still taking place. It is principally want of food which compels them to search for new homes.

On page 91 I have referred to some birds which have come to us from the north. One of these, the Snow Bunting (Plectrophenax nivalis), is a typically Arctic species. In summer it is widely distributed, and is found in Spitsbergen, Novaya Zemlya, Siberia, and the Arctic Regions generally. In winter it migrates down into North America, into Japan, Northern China, Turkestan, Southern Russia, and occasionally even across Europe into North Africa. Very characteristic Arctic birds are the Eider Ducks belonging to the genus Somateria. Three species have visited the British Islands. The common Eider Duck (S. mollissima), which is of such high commercial value, is abundant in Norway and northward, throughout the Polar Regions. The appearance of the King Eider (S. spectabilis) on our coasts is an extremely rare occurrence, and even in Norway it is only known as a visitor, but on Novaya Zemlya and along the Arctic shores of Siberia, in Greenland and Arctic North America, it is known to breed. The third species, Steller's Eider (S. Stelleri), seems to be still rarer, and only in the Aleutian islands and in the north of Alaska can it be said to be at all abundant. It is probable that the famous Great Auk (Alca impennis, Fig. 9) also was a typical Arctic species. Its range extended to both sides of the Atlantic. In Newfoundland and on the coast of Iceland it is known to have been met with in considerable numbers within historic times; and no doubt, like all Arctic species, it extended farther southwards at a more remote period.

Fig. 9.—The Great Auk (Alca impennis).

The members of the genus Lagopus, including the various species of Grouse, are likewise of northern origin. The British Red Grouse (L. scoticus), which may be looked upon as a form of the Scandinavian Willow Grouse (L. albus) (compare p. 91), constitutes in some respects a curious case of parallelism with the Arctic Hare, since the latter, in its more southern station, generally retains the summer fur throughout the year. The allied Ptarmigan (L. mutus) inhabits Scandinavia, the Ural Mountains, and some of the Asiatic mountain ranges. It is also found in the European Alps and in the Pyrenees. The North European range of the Ptarmigan suggests that we are dealing with an ancient species which came south from the Arctic Regions at about the same time as the Arctic Hare; but it is more probable, as I have shown in a subsequent chapter (p. 334), that this species has entered Europe more recently with the Siberian migrants from Central Asia, where indeed the genus had its original home. The Black Cock (Tetrao tetrix) and the Capercaillie (Tetrao urogallus) have also come to us from the east, and have even penetrated into Ireland. They are therefore some of the few instances of members of the Siberian invasion having become temporarily established there.

Reptiles and amphibia are altogether unknown in the Polar Regions, but a large number of fish, chiefly marine, have taken their origin there. The Salmon family are of Arctic origin, as also are the Sticklebacks and the Perches, many of the Cod family, the Herrings, and several of the Flat fish.

It would lead me too far to refer to the invertebrate fauna of the Polar Regions, but a few remarks on the Arctic plants may not be out of place.

The principal Arctic genera are Salix, Ranunculus, Draba, Pedicularis, Potentilla, Saxifraga, Carex, Juncus, Luzula, Eriophorum, and others.

Among the most characteristic Arctic plants may be mentioned Dryas octopetala, to which I have already referred as occurring in the west of Ireland; Saxifraga oppositifolia, another British species, occurs in the higher mountains of Scotland, Ireland, and Wales; Braya alpina, Papaver nudicaule, Lychnis apetala, Diapensia lapponica, and Lobelia Dortmanna, which is found in the lakes of Scotland and Ireland. The dwarf birch (Betula nana) also, which still occurs in Scotland and the North of England, and which had formerly a wider range in the British Islands, should be included among these; but there are other plants probably of Arctic origin, though not now occurring in the Arctic Regions, and to these may be classed the so-called American species of plants which are found on the northern and western coasts of Ireland, in the Hebrides, in Scotland, and in North America. These are no doubt the relics of an Arctic flora which flourished in high latitudes in past times when the climate there was more temperate. A list of these species will be found on page 166.

As none of them occur in Siberia, they must either have found their way to North America and to Europe from the Arctic Regions, or have travelled from North America across the latter to Europe. In any case a former land-connection between the two continents must have existed. This becomes the more evident when we examine the remarkable results obtained by the late Professor Heer, who first described the Tertiary plant-beds in North Greenland. No less than 282 species of plants have been described by this eminent botanist from these deposits. A large number of the plants found were trees belonging to the genus Sequoia, Thujopsis, and Salisburia, besides beeches, oaks, planes, poplars, limes, and magnolias. That they grew on the spot is proved by the fruits, which have been obtained from these beds in various stages of growth.

From a similar deposit in Spitsbergen a large number of fossil plants have also been brought to light, many of which are identical with those found in Greenland; and some of the Greenland forms (such as Taxodium distichum and Sequoia Langsdorfii) have been found too in Alaska, showing that there was probably a continuity of land between Spitsbergen and North America by way of Greenland. Two species of Sequoias, namely, S. sempervirens and S. gigantea, the well-known Californian giant trees, are very closely allied to the Greenland forms discovered by Professor Heer.

Heer assigned the Arctic plant-bearing beds to the Miocene epoch, but doubts have been recently thrown upon this opinion by Mr. Starkie Gardner, who brought forward arguments in support of his theory of their being of the Eocene age. Professor Heer, however, was able to meet these criticisms, and he is ably supported in his views by Professor Engler and other eminent continental botanists.

It is evident that under the present conditions of temperature none of those plants could have flourished in Greenland. The climate must have been much milder than it is at present. Professor Heer estimated from the general aspect of the fossil flora that the mean annual temperature of North Greenland was at least nine degrees centigrade, and that the mean winter temperature was not below zero.

It will hardly be necessary for me to review here the various theories which have been advanced by geologists and botanists to account for this remarkably high temperature in such northern latitudes. Any one who has read the writings of the late Dr. Croll cannot help being struck by the facts he adduces to show the importance of ocean currents in relation to the distribution of heat over the globe, and it seems to me that the view which attributes the mild climate prevailing in former times in Greenland to warm ocean currents reaching the Polar Circle is the one least open to serious objections. If we suppose that the North Atlantic Ocean was bridged by a land-connection between Scandinavia and Greenland by way of Spitsbergen, and between Greenland and North America, the Polar Ocean would be practically a closed sea. If, then, a wide passage existed somewhere about Behring Straits to allow a warm current to enter and circulate within the Arctic Seas, we should have the southern shores of Greenland washed by the warm Atlantic current and the northern shores by a warm Pacific current, which combination would undoubtedly produce the effect of raising the temperature throughout the Polar Regions very considerably; and especially would that be the case with regard to Greenland and the neighbouring islands.

It might be urged that the constant darkness during winter must have had an injurious action upon the flora, but it is found that in countries such as Northern Russia, where southern plants are housed during winter in greenhouses, the light being almost entirely excluded by a covering of straw, no serious damage is done thereby to the plants.

It seems probable that a similar gradual refrigeration of climate in northern latitudes has taken place after Miocene times as has been proved to have occurred in Europe.

Some years ago Dr. Haacke propounded the hypothesis that the centre of creation of all the larger groups of animals was situated in the region of the North Pole, and that the newly originated groups must always push the older ones farther and farther south into the most remote corners of the earth. As instances of the correctness of his view he quotes the fact that the more ancient mammals, such as Monotremes, Marsupials, Lemurs, Edentates, and Insectivores, all inhabit the more southerly parts of the world. The Apteryx, Moa, Rhea, and the Ostrich, as well as Æpyornis, which is only recently extinct, are found in the same regions. But we have no palæontological evidence in favour of these extravagant views. Fossil Edentates and Marsupials are almost entirely confined to the Southern Hemisphere, and the supposition that because these primitive mammals inhabit the extreme south of our great continental land-masses, they therefore came from the north, cannot be said to be an argument. Nevertheless, I am quite with Dr. Haacke in considering that the North Pole, or, we might say, the lands within the Arctic Circle, have been the place of origin of some of our European mammals, and there can be no doubt that certain species in other groups, among invertebrates and also plants, have originated in the Polar Regions. The facts of geographical distribution teach us that in these regions there has been a centre of origin within comparatively recent geological times. I have on a previous occasion drawn attention to the range of the Reindeer: that it lives almost throughout the Polar lands, and that it spreads into North America, Northern Europe, and Northern Asia. We have, again, fossil proof that its range extended down to the Pyrenees in Europe in pleistocene times. But there is not a scrap of evidence that it ever during any time occurred farther south, either in Europe, Asia, or North America. Its original home must therefore have been in the Polar Regions, for if it had originated either in Central Europe, Asia, or America, there is no reason why it should not, in the natural course of events, have extended its range to the south as well as to the north.

The Arctic Hare presents us with a very similar case of distribution. Like the Reindeer, it inhabits, as we have learned, the Polar Regions and the northerly parts of the Old World and the New; but while we have only fossil evidence of the former, more southerly, extension of the range of the Reindeer, the Arctic Hare furnishes us with a still stronger proof of its past southward range in the survival of small isolated colonies in some of the southern mountain ranges of Europe and Asia. It is generally believed that the occurrence of the Arctic Hare in these southern mountains is a standing testimony to the severity of the climate at the time when it commenced its southerly increase of range, but I have already shown that the climate of Europe at that time was not necessarily colder than it is at present, but that it may have been somewhat milder (p. 80). I think that a vast increase of ice in the Polar Regions has taken place only at a comparatively recent date, and that both the Reindeer and the Arctic Hare originated there during a much more temperate climate than obtains at present. A great sensation was produced among European zoologists and anthropologists when the discovery was first announced that the remains of the Reindeer had been found in the Pyrenees, and it naturally gave rise to many speculations as to the nature of the climate at the time when its range extended so far south.[1] The greater number of our best authorities are still of opinion that the existence of the Reindeer in Southern Europe points to the prevalence of an arctic climate in that region. It is generally overlooked, however, that the Reindeer-remains occur in company with many typically southern animals, which, if they had been found alone, would have been held to be a certain indication of a warm climate. The French geologist Professor Lartet, indeed, was of opinion that the temperature during the time when the Reindeer lived in the Pyrenees must have been rather milder than it is at present (compare pp. 71-75). Similarly, Mr. Harlé argues, that the extremely cold climate probably did not extend to South-western France, since that area only received occasional visits from some of the representatives of the Arctic fauna.

Long ago North American zoologists recognised the existence in their country of two well-marked races of the Reindeer (Caribou)—a smaller one with rounded antlers (Fig. 10), and a larger one in which the antlers are more or less flattened out (Fig. 11). Two somewhat similar races can also be traced in the fossil remains of the Reindeer in Europe. It was, I think, Gervais who first pointed out that the Reindeer remains from the north of France differed from those found in the south; and Lartet referred to the fact that the southern remains were more like what, in America, is called the Barren-ground Caribou, while those from Central European deposits all belonged to the Siberian variety, which is more like the Woodland Caribou of North America. In Ireland, Professor Leith Adams also drew attention to the curious fact that all the Irish Reindeer remains resemble the Norwegian variety rather than the Siberian; and Mr. Murray was so much struck by the close resemblance between the Spitsbergen and Greenland forms with the Barren-ground Caribou, that he based some speculations on a former land-connection between these countries on this circumstance.