The Woodland Reindeer persisted in continental Europe until comparatively recent times, and it has since made its way into Scandinavia across Northern Russia, and probably mingled with the older stock of the Barren-ground form. In the same way, it may have come about that in the English pleistocene deposits the remains of the two races occur.

In a recent contribution to our knowledge of the deer tribe (_c_, p.

88), Mr. Lydekker suggests that the former division of the Reindeer races into the two forms of Woodland and Barren-ground Caribou, no longer holds good. He now recognises no less than six races, as follows:--

1. Rangifer tarandus typicus.

2. " " spitzbergensis.

3. " " caribou.

4. " " terrae-novae.

5. " " grnlandicus.

6. " " arcticus.

I hardly think these can be considered of equal value; indeed, though there may be differences between _R. grnlandicus_, _typicus_, _arcticus_, and _spitzbergensis_, the antlers exhibit a certain much closer relationship among one another than to _R. terrae-novae_ and _caribou_. But the whole subject is by no means as well known as could be wished, and a very careful comparative study of recent and fossil remains of the Reindeer from various parts of the Old and New Worlds is much needed to put our views on a firmer basis.

The presence of the Arctic Hare in Ireland and the absence of the common European Hare (_Lepus europaeus_) can be explained in a somewhat similar manner. The Arctic Hare is the older of the two species--corresponding with the Barren-ground Reindeer--and the European Hare the newer one, a.s.sociating, like the Woodland Reindeer, in its westward migration with Siberian animals, though probably of Oriental origin.

Let us once more refer back again to the map on page 137 indicating the geographical distribution of the Arctic Hare. Its discontinuous range and its isolated position in the Alps, Pyrenees, and the j.a.panese mountains, all tend to show that it is an ancient species. Moreover, its presence in Ireland in the plain as well as in the mountains, clearly points to the fact that, in the British Islands at any rate, the Arctic Hare was the first comer, and that subsequently the European Hare invaded these countries. It probably found Ireland then no longer accessible, having since become separated from England. Again and again do we find the statement repeated, that the presence of the Arctic Hare in Europe is a clear proof of the former prevalence in our continent of an Arctic climate. But if so, why should this Hare at present live and thrive in Ireland, which has a particularly mild climate in winter, and be absent from so many continental stations where the temperature more resembles that of its native home? If we suppose that the European Hare migrated to Europe from the east, after the Arctic Hare had become established in Western Europe, and drove the latter into the mountains or northward whenever the two came into contact, we should have, it seems to me, a better explanation of the range presented by the two species. I was formerly of opinion that the European Hare had come with the Siberian animals from Siberia, but it appears to me more likely now, that it reached our continent with the Oriental migrants, and only then joined the Siberians in Eastern Europe.

The evidence in favour of a former land-connection between Scandinavia and Greenland, rests on many other facts besides those already brought forward. That some form of land-connection formerly existed between Europe and Greenland is now indeed almost universally accepted. That it was situated more to the south between Scotland and Greenland is a supposition which has been actively supported by many leading authorities, but it seems to me that if such a land-bridge existed, it must have been in very early Tertiary times, whilst the northern one, such as I have indicated, may have originated later and persisted until a recent geological date.

The distribution of few groups of animals is now better known than that of the larger b.u.t.terflies and moths (_Macro-lepidoptera_); even those of Siberia have been fairly well investigated. The interesting facts obtainable from their distribution are therefore of special value. No less than 243 species of _Lepidoptera_ are mentioned by Moschler as being common to North America and Europe. It is extremely probable that a fair number of these have either migrated direct from America to Europe or _vice versa_, though many may be of Asiatic origin, and have wandered east and west from their original home. The following twelve species are mentioned by Petersen (p. 38) as occurring in Arctic Europe and also in Arctic North America, but not in Asia:--_Colias nastes_, _Colias hecla_, _Syrichthus centaureae_, _Pachn.o.bia carnea_, _Plusia parilis_, _Anarta Richardsoni_, _Anarta Schonherri_, _Anarta lapponica_, _Anarta Zetterstedti_, _Cidaria frigidaria_, _Cidaria polata_, _Eupithecia hyperboreata_; and these, as he remarks, point to the possibility of a former direct land-connection between Europe and North America.

Mr. Petersen believes that the chief immigration into the Arctic area of Europe is post-glacial and took place from Siberia, since the majority of the species are still to be found in that country at the present day (p. 57). He also draws particular attention to a fact,--which I shall discuss more fully in the next chapter,--namely, that the most characteristically Arctic forms of Northern Europe, which also partly occur in the Alps, are entirely absent from the Caucasus.

Adopting the glacial views of some of our leading geologists, Petersen comes to the logical conclusion that Central Europe could not have possessed any b.u.t.terflies during the height of the Glacial period, but since all evidences seem to point to the chief migration from Siberia having taken place after the Glacial period, he concludes that they must have survived the severe cold of that time in Central Asia. He leaves us, however, to imagine under what possible geographical conditions the climate in Europe could be too severe for a lepidopterous fauna, while at the same time Central Asia could maintain an abundant one.

In a suggestive note on the origin of European and North American Ants, Professor Emery states (p. 399) that a great number of North American ants are specifically identical with European ones; whilst Dr. Hamilton tells us (p. 89), as an instance, that specimens of the beetle _Loricera crulescens_ from Lake Superior and from Scotland do not seem to vary to the extent of a hair on the antennae. He enumerates 487 species of _Coleoptera_ as being common to North America, Northern Asia, and Europe, many of which no doubt have migrated by the Americo-European land-connection.

Arctic Scandinavia or Lapland, according to Sir Joseph Hooker, contains three-fourths of the entire number of species of plants known from the whole circ.u.mpolar area. His view, that the Greenland flora is almost exclusively Lapponian,--having only an extremely slight admixture of American or Asiatic types,--again points to a former more intimate connection between North America and Arctic Europe, and indeed he remarks (p. 252), "It is inconceivable to me that so many Scandinavian plants should, under existing conditions of sea, land, and temperature, have not only found their way to Greenland by migration across the Atlantic, but should have stopped short on its western coast and not crossed to America."

Hooker's view, that the Scandinavian flora is of great antiquity, that, at the advent of the Glacial period, it was everywhere driven southwards, and that during the succeeding warm epoch the surviving species returned north, has been adopted by the great majority of naturalists.

The natural corollary of this theory is that there must have been, between the beginning of the Glacial period and the present time, either two independent land-connections between the Polar Regions and Northern Europe at different epochs to enable animals and plants to travel southwards and once more to regain their former northern home, or, that during the whole of the Glacial period the Polar Regions were uninterruptedly connected with Northern Europe, until the fauna and flora had once more reached their northern goal, after the Polar lands had been desolated by the supposed rigours of that period.

In following the history of the Arctic migration to Europe, it is of great importance to determine the nature and the time of duration of these land-connections. The Greenland flora is a very instructive one in helping us to understand many of the problems connected with the origin of the European plants and animals. To judge from the remarks of Professor James Geikie and Mr. Clement Reid, no flowering plants could have existed in the British Islands during the height of the Glacial period, and one would suppose that the cold in Greenland at that time must have been far more intense than in England. If no flowering plants could exist in the latter country, then very surely none could in Greenland, where the climate was of necessity by far more rigorous. It will be a surprise, therefore, to those who are acquainted only with Professor Geikie's views of the nature of the Glacial period, that two of the most eminent Swedish botanists, who have made a special study of the flora of Greenland, have come to the conclusion that a survival of flowering plants has taken place in Greenland itself from pre-glacial times. According to Professor Nathorst (p. 200), only a few plants could have survived the Glacial period in Greenland. The species now peculiar to that country may perhaps, he thinks, be the remnants of those which existed in pre-glacial times. Mr. Warming, on the other hand, is of opinion that the main ma.s.s of Greenland's present flora survived the Glacial period there (p. 403), and that the remainder was carried from Europe and North America by occasional means of distribution of the nature indicated by Darwin.

Very similar views on the origin of the present Polar flora are expressed by Colonel Feilden, who says, "To my mind it seems indisputable that several plants now confined to the Polar area must have originated there and have outlived the period of greatest ice-development in that region" (_b_, p. 50). No land-connection at all need be supposed to have existed in recent geological times, that is to say, during the Glacial period or after, if Mr. Warming's and Colonel Feilden's views be adopted. A pre-glacial connection would be sufficient to explain the general features of distribution. An admission is thus obtained from these two independent authorities that the climate during the Glacial period must have been vastly less severe in the Polar Regions than is generally conceded. I am of opinion that not only the whole of the present flora, but also the fauna of Greenland survived the Glacial period in that country.

If we suppose that an extensive centre of origin existed in the Polar area, or we may say in Greenland, both animals and plants would have been able to spread from it into Northern Europe and North America by means of the land-connections which are generally supposed to have existed in pliocene times, that is to say, just before the commencement of the Glacial period. There must have been at this time a connection too between Scotland and Scandinavia, which will be dealt with more fully presently. The important point is to consider what light the Greenland flora and fauna will throw upon the problem of the continuity of the aforesaid land-connection during the Glacial period. We have seen that the Barren-ground Reindeer, a typically Polar species, penetrated as far south as the Pyrenees, the Arctic Hare went as far, while a number of other species of Polar animals and also of plants occur in the Alps. Of these it remains to be seen how many have come direct by way of Northern Europe or from the Polar Regions by way of Asia. At any rate, as the origin of the Alpine animals and plants will be discussed in another chapter, there is no need to dwell on this subject at present.

From the nature of the distribution in Ireland of Arctic plants and animals, which occur mostly on the north and west coasts, it would seem that a stream of migration entered from Scotland, and I have no doubt that that same migration came into Scotland directly from Scandinavia by a route over which now roll the waves of the North Sea. There is, moreover, as I already mentioned on p. 94, a very interesting so-called American element in the north-western European flora, that is to say, plants now found in North-west Europe and North America without occurring in Greenland or any of the islands which might have formed the former highway between the Old World and the New. These are probably some of the more ancient Polar plants which have become extinct in the Arctic Regions and survive in isolated patches in favourable localities.

We find seven species of these American plants in Ireland, almost entirely confined to the north and west coasts. These are _Spiranthes Romanzoviana_, _Sisyrinchium anceps_, _Naias flexilis_, _Eriocaulon septangulare_, _Juncus tenuis_, and _Polygonum sagittifolium_. To them must be added another plant recently discovered by the Rev. Mr. Marshall in the south of Ireland, namely _Sisyrinchium californic.u.m_. As I have mentioned in former writings, there are three species of North American freshwater-sponges in Ireland which have not hitherto been discovered elsewhere in Europe or in Asia. These, namely _Ephydatia crateriformis_, _Heteromeyenia Ryderi_, and _Tubella pennsylvanica_, all occur in some of the lakes near the western coast of Ireland.

There are in all groups of animals instances of species which are confined to Europe and North America, while unknown from the Asiatic continent, but none, as far as is known, have such a very discontinuous range as that of the animals and plants just referred to. In some cases the species still occur in Greenland, and in this way make it still clearer that their migration in former times took place from one continent to the other by way of that country. As an interesting instance of such distribution may be mentioned the Common Stickleback (_Gasterosteus aculeatus_), which is found in Greenland, North America, and Europe, but is quite absent from Asia. Then again, the Nine-spined Stickleback (_Gasterosteus pungitius_) is confined to Western Europe and North America, though an allied species, _Gasterosteus sinensis_, lives in China and has probably penetrated there from the New World across the old Behring Straits land-connection.

The Coleoptera _Diachila arctica_, _Elaphrus lapponicus_, and _Blethisa multipunctata_ are good instances of species which have come to us from North America by way of Greenland. I have already referred to the Lepidoptera, but might add that eleven species of _Anarta_ occur in Scandinavia, eight of which reappear again in Labrador, none of them, however, being met with in Siberia. Then again, take the interesting Crustacean _Lepidurus (Apus) glacialis_. It is found in Greenland, Spitsbergen, Lapland, and Norway; and formerly, as we know from fossil evidence, it ranged into Scotland. Another Phyllopod, viz., _Branchinecta paludosa_, inhabits Greenland, Lapland, and Norway. Mr.

Kennard suggests that the freshwater Snail _Planorbis glaber_ might also belong to the same migration. And there are no doubt large numbers of others.

Professor Emery mentions that Northern Europe possesses one peculiar genus of Ant, viz., _Anergates_. This is closely allied to _Epoccus_, another genus confined to North America. It seems probable, therefore, that both of these have sprung from an Arctic genus which sent two branches southward into the two continents without there being any migration through Asia.

The general range of the Arctic plants and animals gives no reason to suppose that the Greenland fauna and flora of the present day were exterminated by the Glacial period and then reintroduced into that country. Nor have we any evidence that such a fauna and flora migrated across the British Islands northward. The Greenland animals and plants too are altogether much more like the Lapland ones than those of Scotland. It will also become evident to the reader of this work that no very extensive migrations could have taken place during the post-glacial period, and that almost everything points to a survival of both fauna and flora in northern lat.i.tudes throughout the Glacial period.

If we take into consideration the palaeontological evidence of the two races of Reindeer in Europe, one of which came to us from the north, and that the Arctic Hare and one of the races of the Stoat entered our continent from the same direction--when we, moreover, carefully review the numerous other instances quoted of plants and animals which could only have reached us from the north, the irresistible conclusion is forced upon us that a land-connection existed at no very distant period between Northern Europe and the Arctic Regions of North America. This is not a new hypothesis. Many geologists are of opinion that a land-pa.s.sage did exist within comparatively recent times, uniting Europe, Greenland, and North America. But the position of this old land-bridge, as I have mentioned, has been generally placed somewhat farther south than I should feel inclined to put it.

The fact that very extensive glaciers formerly covered the mountains of Scandinavia on the eastern side, whilst they scarcely reached the sea on the west (Feilden, _a_, p. 721), seems to favour the view of a warm current having washed the western sh.o.r.es. As I shall attempt to show later on (p. 179), the Arctic Ocean extended across Northern Russia at that time from the White Sea to the Baltic--that is to say, to the eastern sh.o.r.es of Scandinavia, which country was then joined to the north of Scotland. The predisposing agents to a copious snowfall existed in Scandinavia, viz., an excessive evaporation of the warm Atlantic waters and unusual precipitation in the form of snow owing to the cold given off by the Arctic waters on the east side of the mountains. It is therefore probable that the land-connection which united Europe and North America was farther north than has been supposed.

If we sail straight across from Northern Scandinavia to Greenland, we traverse an exceedingly deep marine basin; but if we examine the sub-marine bank which runs all along the coast of the former country from south to north, we find that it does not end when the extreme north of the land is reached. The bank extends much farther north, and is continued as far as Spitsbergen. As I have said before, the latter, as well as Bear Island, must be looked upon as the remains of a large ma.s.s of sunken land--the ancient Scandinavia stretching far into the Arctic Circle. Professor Nathorst speaks of Spitsbergen as a northern continuation of Europe, not only geographically, but also botanically and geologically. However, this northern land must have stretched even farther--not perhaps farther north, but farther west. Here lay the old land-connection between Scandinavia, Greenland, and North America (Fig.

13). One of the highest authorities on the geographical distribution of plants, Professor Engler, maintains that the arguments in favour of this Arctic connection of America with Europe are more weighty than those for a land-bridge between Greenland, Iceland, the Faroes, and Great Britain.

Moreover, he is of opinion that a certain number of species of plants belonging to the Alpine flora of Arctic Siberia have travelled from Scandinavia _via_ Greenland and North America to Eastern Asia, and not direct from Scandinavia to Siberia (p. 143).

[Ill.u.s.tration: Fig. 13.--Map of Europe, indicating approximately the distribution of land and water during the earlier stages of the Glacial period--shortly after the period represented in Fig. 12, p. 156. The darkly shaded parts indicate the areas covered by water, and the white portions what was land at the time.]

That this ancient Arctic land-connection existed almost throughout the Glacial period appears to me probable. It has often been suggested that such a land-barrier was one of the princ.i.p.al causes of the production of the glacial phenomena in Europe, and as such it must have existed intact certainly during the earlier stages of the Glacial period. The barrier must then have gradually subsided in one or two places; and once a breach was formed, the complete union between the Atlantic and the Arctic Oceans could not have been long delayed.

The terrestrial fauna and flora, as we have seen, lend strong support to the view of the former connection between Scandinavia and Greenland, but many other facts point in the same direction. It was Edward Forbes who first drew attention to the presence of a number of species of littoral molluscs on the coast of Finmark which also occur on the coast of Greenland, and he expressed the firm conviction that they indicated by their existence on both sides of the Atlantic some ancient continuity of the coast-line. He held that the line of migration of these mollusca was probably from west to east, and that it must have taken place during physical conditions entirely different from those prevailing at present.

If Forbes's view is correct, a current must have existed from the north coast of North America along the northern sh.o.r.e of the ancient land which stretched east as far as Europe. We have also some palaeontological evidence bearing on the existence of such a current (p. 173).

As we shall learn presently, the early stages of the Glacial period were accompanied by a marine transgression over Northern Russia and Germany--an overflow, as it were, of the waters of the Arctic Ocean covering a great part of Northern Europe, with the exception of Norway.

One continuous ocean ultimately extended from the east coast of England across Holland, Northern Germany, and Russia to the White Sea (Fig. 12, p. 156). The south of England being at that time joined to France, and Scotland to Scandinavia, there was no direct communication between this large North European Sea and the Atlantic. The glaciers which took their origin in the Scandinavian Mountains discharged icebergs into this sea, and many of them no doubt were stranded on the east coast of England.

The boulders of Scandinavian origin which have been discovered in recent geological deposits on that coast have generally been traced to the action of land-ice, but the supposition that they have been carried by icebergs--the older theory--appears to me the more probable one. Such boulders begin to make their first appearance in the Red Crag, a deposit which is now looked upon as belonging to the newer pliocene series. But whether we call it pliocene or pleistocene really matters little. The important fact is, that glacial phenomena, consisting of the appearance of boulders foreign to the country together with an invasion of Arctic sh.e.l.ls, are now ushered in upon a coast which shortly before teemed with the southern life of a Mediterranean character. Among the new arrivals in these English crags there are no less than eighteen species of North American marine mollusca. Since the German Ocean had then no direct communication with the Atlantic, these mollusca could only have come from the White Sea, and Forbes's _Arctic current_ would offer an explanation of the manner in which they were enabled to migrate there from their original home.

It might be urged that we have no grounds for the supposition that the German Ocean was practically a closed basin; and that these American species probably inhabited at that time the whole of the North Atlantic Ocean. But if such had been the case, we ought to have evidence of the occurrence of some of these species in the newer Tertiary deposits along the west coasts of the British Islands. Such beds exist; there is, however, not a trace in any of them of any American mollusca. In examining the marine deposits of St. Erth, on the coast of Cornwall, which are believed to be of about the same age as the newer crags, Messrs. Kendall and Bell were much struck by the absence of the species characteristic of the latter. The St. Erth fauna led them to believe that the Arctic Ocean could not then have opened into the Atlantic, but that a land-communication had existed between Europe and North America, so as to form a barrier of separation between the two oceans. This again perfectly harmonises with the views I have expressed, and supports them.

Let us now look a little more closely at the history and the fauna of the Baltic and the adjoining lakes, in order to gain additional information as to the geographical changes which have had such lasting influence on the peninsula of Scandinavia. The Baltic is a shallow sea covering an area of 184,496 square miles, and its waters are decidedly brackish. The fauna is a poor one, being too salt for the purely freshwater species and not salt enough for the typical marine forms. The absence of some animals which we should expect to find there is one of the remarkable features about the Baltic, but, on the other hand, some species occur which are altogether strangers to the fauna. And these, moreover, are confined to the extreme northern end of the sea. I need only refer to the Arctic Seal (_Phoca annelata_), which is confined to the Gulf of Bothnia, and to the four-horned sting-fish (_Cottus quadricornis_, Fig. 14, p. 178), neither of which occur on the west coast of Scandinavia. But there are others which point in an equally unmistakable manner to the former existence of a marine connection between the Baltic and the southward prolongation of the Arctic Ocean--known as the White Sea. It is generally admitted now that such a union between these two seas, viz., the Baltic and the White Sea, occurred in recent geological times, but opinions differ as to the duration of this connection. I adhere to the view expressed by Murchison and others, that the boulder-clay is a marine deposit. I am also convinced that the Arctic Ocean, as I have already mentioned, transgressed over the lowlands of Northern Russia at about the time when the newer crags were being deposited on the east coast of England; that the same large sea also covered Northern Germany, Denmark, Holland, and the lowlands of Sweden, and laid down the lower continental boulder-clay which is spread over such vast tracts of land in those countries. I shall have occasion to refer to this again more fully in the next chapter; meanwhile, it should be remembered that this stage was followed by a partial retreat of the northern sea, though Scandinavia did not become joined to the Continent. The date of this retreat of the sea, represented in Fig. 13, corresponds probably to what is known as the inter-glacial phase of the Glacial period, and I think it must have been during this time that the Forest-Bed on the coast of Norfolk was laid down.[2]

None of the Siberian mammals apparently entered Scandinavia at the time when they invaded Central Europe and penetrated as far west as England and Western France. Nor did the great Oriental mammals, like the Mammoth and others, reach Scandinavia; and Professor Pohlig argued, on the strength of these facts, that the latter country was either for a very short time only free from ice, or that it had defective land-communication with the Continent during inter-glacial times. This seems to me scarcely to explain the facts of distribution and account satisfactorily for the absentees. Nor does it, of course, harmonise with the views that I have announced above. Professor Engler's remark (p.

131), that Scandinavia probably projected above the glacial sea as an island, is more in accordance with these views, though the term island is scarcely applicable to that country, since it was always, as I said, indirectly joined to the Continent (_vide_ Fig. 13, p. 170). The fauna of Scandinavia, both fossil and recent, points to a direct isolation of that country from the continent of Europe during a considerable period.

Another proof that Northern Russia and the lowlands of Sweden were covered by the sea comes to us from a study of the fauna of the relict lakes--the "Reliktenseen" of Leuckart. This name was first applied by Leuckart to lakes containing marine organisms, which are supposed to have been flooded by, or to have been in close communication with the sea at some former period, like the lakes Ladoga and Onega in Russia.

His views have been worked out subsequently in greater detail by Loven and O. Peschel, who gave them their strong adherence. Many leading zoologists, such as Professor Sars and others, have since adopted them, and though discredited by Professor Credner, the theory still offers the best explanation for the origin of marine animals in freshwater lakes.

Professor Credner's contention, that marine mollusca are always absent from these relict lakes, seems at first sight a stumbling-block to the theory. But the explanation is really simple enough. It is to Dr. Sollas that we owe a very ingenious explanation of the origin of freshwater faunas. He showed that all freshwater organisms in their early stages of development are provided either with some process enabling them to attach themselves to a foreign object, or that they pa.s.s this period within the body of the parent. This is a provision of nature to prevent freshwater organisms from being floated out to sea, where they would perish, until they reach maturity and can cope with floods and currents.

Had Professor Credner been aware of Dr. Sollas's views, no doubt he would have modified his criticisms, for, as most marine mollusca have free-swimming larvae, they would have little chance of becoming permanent residents of lakes. During their larval stage, marine molluscs are quite a prey to the currents of the sea. They have practically no swimming organs, and only move by lashing to and fro the tender cilia with which they are provided.

[Ill.u.s.tration: Fig. 14--The Four-horned Sting-fish (_Cottus quadricornis_), reduced from Professor Smitt's figure in the _Fishes of Scandinavia_.]

This disposes, therefore, of Professor Credner's main criticisms. As for the fauna of the relict lakes, we are now only concerned with those of Northern Russia, Finland, and Sweden. In the lakes Wetter and Wener in the latter country occurs the four-horned sting-fish (_Cottus quadricornis_, Fig. 14), which, as we have learned, also inhabits the northern part of the Baltic, and, as was suggested, migrated there at a time when the latter was connected with the White Sea. The princ.i.p.al food of this little fish consists in a marine Crustacean called _Idotea entomon_, an animal allied to our common woodlouse. This is a typical marine species, but it occurs also in the relict lakes of the countries mentioned above, as well as in the Baltic and the Caspian. Perhaps the best known form with a similar range is the Schizopod crustacean _Mysis relicta_[3] (Fig. 15), which is clearly a descendant of the Arctic marine _Mysis oculata_, of which it was formerly considered a mere variety. The two Amphipods _Gammaracanthus relictus_ and _Pontoporeia affinis_ and the Copepod _Limnocala.n.u.s macrurus_, are three additional well-known Arctic crustaceans whose range differs but little from those above-mentioned.[4]

[Ill.u.s.tration: Fig. 15--Mysis relicta, a small shrimp-like Crustacean, after Sars (enlarged).]

These facts all go to prove that the sea formerly covered the lowlands of Sweden, Finland, and Northern Russia. The fauna of Scandinavia, as we have seen, indicates that during the greater part of the Glacial period the country was not directly connected with continental Europe as it is now. It seems that the barrier of separation probably consisted of a broad expanse of ocean on which floated numerous icebergs, which originated from the Scandinavian glaciers as they reached the sea. This was a cold sea, whilst Western Scandinavia was washed by the Gulf Stream (_vide_ Fig. 12, p. 156). We might look upon the boulder-clay which covers such vast tracts of country in Northern Germany, Russia, and Holland as deposits formed by this sea rather than the ground-moraine of a huge Scandinavian glacier. I shall refer to this subject again in the next chapter; meanwhile it may be remembered that the boulder-clay of Northern Europe exactly resembles in all important particulars the similar acc.u.mulations met with in the British Islands. They resemble one another also in the occasional occurrence of sea-sh.e.l.ls, the frequent appearance of bedded deposits, and the often inexplicable course taken by boulders from their source of origin. There occurs often a singular mixture and an apparent crossing of the paths of boulders in the boulder-clay. Professor Bonney remarks (p. 280) that these are less difficult to explain on the hypothesis of distribution by floating ice than on that of transport by land-ice, because, in the former case, though the drift of winds and currents would be generally in one direction, both might be varied at particular seasons. So far as concerns the distribution and thickness of the glacial deposits, he says there is not much to choose between either hypothesis; but on that of land-ice it is extremely difficult to explain the intercalation of perfectly stratified sands and gravels and of boulder-clay, as well as the not infrequent signs of bedding in the latter. Two divisions are generally recognisable in the continental boulder-clay--a lower and an upper. An inter-glacial phase characterised by a less severe climate is a.s.sumed to have intervened between the deposition of the two. In Russia no such division can as a rule be made out, and sea-sh.e.l.ls are either entirely absent or extremely scarce. It has been pointed out by Professor J. Geikie that the erratics--a name applied to boulders in boulder-clay--in the upper division have travelled in a different direction from those contained in the lower. Taking for granted that the boulder-clay is a marine deposit, this phenomenon seems to indicate that the current which prevailed during the early part of the Glacial period in this North European ocean was different from the prevailing current during the latter part. I have attempted to explain this circ.u.mstance by the supposition that during the early part of the Glacial period the Northern Sea had a connection with the Ponto-Caspian Sea--a sea formed by the junction of the Black Sea and the Caspian (Fig. 12, p. 156).

There is geological evidence, as will be explained in the following chapter, that the area of these two seas was considerably larger in glacial times than it is now, and that they were joined across the valley of the Manytch. After the inter-glacial phase of the Glacial period, the North European Ocean became connected with the Atlantic Ocean across the north of England (Fig. 6, p. 126), the junction between the former and the Ponto-Caspian having meanwhile become dry land (Fig.

13, p. 170). A fresh current, now flowing westward, was set up in the North European Ocean, which accounts for the fact just cited that the erratics in the upper continental boulder-clay have travelled in a different direction from those in the lower. The boulder-clay laid down by the sea on the midland and northern counties of England, just as was the case with the similar deposit on the Continent, is generally accredited to the action of land-ice. It is by most geologists looked upon as the ground-moraine, partly of the huge Scandinavian glacier which is supposed to have impinged upon the English coast, partly of local British glaciers.

But renewed geological investigations on this point throw doubts upon these theories. Thus Mr. Harmer remarks in a recent contribution to glacial literature (p. 775), that "it is difficult to see how the Baltic glacier could have reached East Anglia, though ice-floes with Scandinavian boulders might easily have done so, while had the Norwegian ice filled the North Sea and overflowed the county of Norfolk, some evidence of its presence ought to be found in the glacial beds of Holland."

All the phenomena of distribution of the British fauna and flora are, as we have seen, much more easily explained by the supposition of a damp, temperate climate, such as might have been produced by the proximity of a cold sea on one side and of a warm one at the other, than by invoking an arctic climate with enormous glaciers. Most of the living animals and plants would have been exterminated under the latter conditions.

Palaeontological evidence in Great Britain clearly indicates that southern species migrated first to these islands, that Arctic species were then driven south from their native lands,--probably owing insufficient food-supply and climatic changes in the north,--that finally eastern species invaded the country--all this without the annual temperature of Europe being apparently much affected. For we find in the British pleistocene deposits--and Mr. Lydekker draws particular attention to this remarkable fact--a curious intermingling of southern and northern mammals, which undoubtedly lived side by side. Everybody knows that northern and Arctic species can live perfectly well in a temperate climate, but that it is almost impossible to acclimatise southern animals in an Arctic or even temperate one. We have in this circ.u.mstance almost a proof, therefore, that the climate cannot have been very cold. Though a cold sea bathed the sh.o.r.es of Eastern England, and even eventually invaded a portion of Northern England, the warm ocean on the west must have effectually prevented any great lowering of temperature.