Animals and plants were free to migrate from Central and Southern Asia to Greece by land for untold ages. The vast accumulation of mammalian bones which have been discovered at Pikermi, and so ably described by Gaudry, are probably to a large extent the remains of Asiatic immigrants to Europe. Many of these resemble forms still living in South Africa, which implies that a highway existed also at that time between Asia and Africa. Among these is a giraffe and antelopes closely allied to African species, and other most interesting mammals.

In still earlier European deposits—the Miocene—we find the ancestors of modern Elephants, which are probably of Asiatic origin. The remains of several kinds of monkeys occur, whose nearest relations are now confined to Southern Asia. Altogether the fauna bears a strong Asiatic facies. Many of our European terrestrial invertebrates probably arrived about this time from Asia. The struggle for existence being keener and the facility for migration much greater in the higher vertebrates, they—or at any rate the mammalian faunas—were subjected to more rapid changes than the invertebrates. I have repeatedly expressed my belief that a great number of our familiar insects and mollusca inhabited Europe long before our present mammals came into existence.[2]

Let us now follow one of the miocene Oriental migrants starting from Central Asia on its way to Europe. Very soon after leaving its home, it must have encountered a sea which extended at that time from the Eastern Mediterranean to the borders of Afghanistan. In following a westward course, the emigrant was compelled to keep along the northern shore of it. We do not know the state of the physical geography of the region between the Black Sea and the Tianshan Mountains, but it seems certain that a considerable extent of dry land enabled a wanderer from Central or Southern Asia to reach the Balkan peninsula by skirting the northern shore of that large miocene sea. No miocene deposits occur north of Teheran or of the Upper Euphrates, nor are they known from the islands of the Ægean Sea or the lands surrounding it. From the Balkan peninsula it was possible for our migrant to reach the European Alps, which were then slowly rising as a peninsula out of the western portion of the great miocene sea. What are now the Alps was then hilly ground, which was being raised from the bottom of the sea. It was no doubt connected with the Balkan peninsula, so that an intercourse of species could take place between this newly-formed peninsula and Central Asia. I say peninsula, because the miocene sea almost completely surrounded it. From the Western Mediterranean a wide gulf extended up the Rhone valley into that of the Rhine as far north as Maintz. Then skirting along the northern outliers of the Tyrol, the gulf can be followed as far east as Transylvania. It is quite probable that it extended much farther east still, but there is as yet no geological evidence forthcoming. At any rate, our Asiatic migrant turning northward from the Balkan peninsula found its farther progress barred once more by an arm of the same sea which in its earlier peregrinations had stopped it from going south (cf. Suess, i., p. 406).

In later miocene times the sea does not seem to have surrounded the Alps to the same extent as it did before, but it certainly extended from the Eastern Alps to the shores of the Sea of Asov, so that the direct northward passage was still more or less barred to the Oriental immigrants. At the same time Alpine species were now able to emigrate to the North European provinces. During the last stages of this epoch, the same sea increased its area very considerably in an eastward direction. One continuous expanse of water now stretched from the Alps as far as the Sea of Aral in Central Asia, perhaps even farther.

During pliocene times especially, the northern parts of the Balkan peninsula were occupied by a series of freshwater lakes, while Greece was joined to Southern Italy, Sicily, and Tunis. Central and Northern Italy were represented by a long narrow peninsula connected in the north with the Alps. Corsica and Sardinia were joined to Sicily, and the Straits of Gibraltar did not exist. When I first published my views regarding these geographical conditions of the Mediterranean area, Professor Depéret was good enough to send me his criticisms from a purely geological standpoint. He is of opinion that though Sicily and Sardinia might at this time have still been connected with Tunis, the Straits of Messina must already have been formed—in other words, Southern Italy and Sicily could no longer have been connected with one another. This opinion is based upon the fact that in the upper strata of the enormously thick Sicilian pliocene deposits are found a number of arctic or subarctic species of mollusca which are entirely foreign to the Mediterranean fauna. It is generally supposed that these reached the Mediterranean area by the newly opened Straits of Gibraltar in later pliocene times, and that the lower Sicilian deposits must therefore have been laid down earlier. So far the deductions are perfectly correct, if we assume the northern mollusca to have arrived in the Atlantic at the time stated. However, they must have reached the Atlantic much later—not till pleistocene times—if we adopt the above-stated suggestions as to the age of the Forest-Bed (cf. p. 125). Moreover, the great similarity between the faunas of Southern Spain and North-western Africa indicate that the formation of the Straits of Gibraltar is of very recent date. The northern mollusca, of course, could not have reached Sicily till later. To suppose that the Sicilian deposits have been uplifted 7000 feet since then is no doubt contrary to all our geological teaching, but we must remember that this is altogether an exceptional case. The area in question has probably ever since been in the immediate neighbourhood of an active volcano, and the rate of the uplift has therefore been immeasurably greater than at other localities with which this one might be compared. The disconnection between Tunis, Sicily, and Southern Italy was evidently produced by a subsidence of the tract of land uniting these countries. If we suppose that this happened in early pliocene times, we have either to take for granted that the terrestrial fauna and flora of these countries are of miocene origin, or that they were joined again during the Pleistocene Epoch. The range of a very large number of animals and plants is such as can only be explained by assuming that Tunis, Sicily, Sardinia, Corsica, and Southern Italy were connected with one another. Of such extensive land-connections subsequent to the arrival of the northern marine mollusca we possess, however, no geological evidence whatsoever; and it is extremely improbable that the land-areas which had sunk were once more raised before again subsiding. The many animals whose presence in the Mediterranean Region bears witness to these ancient land-connections could not have arrived there in miocene times—in fact, they could hardly have lived there before the end of the Pliocene Epoch. On the other hand, it seems difficult to believe, once the Straits of Gibraltar were open and the waters of the Atlantic able to enter the Mediterranean, that the sunken parts between Sicily, Italy, and Tunis could have been raised without affecting the entire area of that sea. Nor is it likely that the junction between these countries could have then been brought about by a general lowering of the Mediterranean waters. As it may be asked what evidences we possess at all for the supposition of such land-connections as I have indicated, also that Southern Italy and Greece were connected, a few of the more salient instances of distribution bearing on this problem may be of interest.

I have already referred to the occurrence of the remains of a small race of Red Deer in the caves of Malta, similar to those still living in North-west Africa, Corsica, and Sardinia. The Black-mouthed Weasel (Mustela boccamela) inhabits Persia, Asia Minor, Greece, South Italy, Sicily, and Sardinia, while Mustela africana is found in Malta and Algiers. The European Porcupine inhabits Asia Minor, the island of Rhodos, Greece, Southern Italy, Sicily, North Africa, and Spain. Then we have the Wild Sheep of Asia Minor, Cyprus, Sardinia, and Corsica, all of which are closely allied. The small shrew-like Crocidura etrusca occurs in South France, Italy, Sicily, and North Africa. Many other mammalia might be quoted, but these are sufficient for our purpose.

There are a good many reptiles and amphibians with a similar distribution. The European Chamæleon (Chamæleon vulgaris) has been found in South Spain, North Africa, and Sicily. The Snake Periops hippocrepis is confined to Spain, Sardinia, and Greece. The worm-like Lizard Blanus cinereus inhabits some of the Greek islands, North Africa, and Spain. Another Lizard belonging to the Scincidæ has also been found in some of the Greek islands, Sicily, Sardinia, Southern Spain, and the Canary Islands. Discoglossus pictus—a toad—occurs in Spain, North-west Africa, Malta, Sicily, Sardinia, and Corsica. A variety of the Tree Frog (Hyla arborea Savignyi) is found in Europe only in Corsica, Sardinia, and the Greek Archipelago.

Eight species of Reptiles and Amphibia—some of which I have just referred to—are enumerated by Dr. Forsyth Major as occurring eastward and westward of the Italian peninsula (and almost all also in North Africa) without being known on the mainland of Italy. And in order to show that Sardinia and Corsica are more closely related to North Africa than to Italy, he indicates the general range of the Reptiles and Amphibians found in these islands. Of the twenty-one species, only twelve inhabit Italy, but at least sixteen North Africa and seventeen Spain. Indeed, he shows that Corsica, Sardinia, Sicily, and North-west Africa form a zoogeographical province, from which Italy, with the exception of a few localities on its west coast, is excluded. It is a remarkable fact that there are a few localities on the west coast of Italy which in their fauna and flora exhibit closer relationship with Corsica and Sardinia than with the mainland. Thus Dr. Major pointed out that the Catena Mettalifera, the Monte Argentario, and Monte Circeo all belong to what we may call the former Tyrrhenian continent. They are to be regarded as its eastern limits, which remained standing, while the central portion—now occupied by the Tyrrhenian Sea—subsided, and is at present covered by deep sea. Subsequently these remnants of the old continent became joined with the newly-formed Italian peninsula, but the plants and animals belonging to the older flora and fauna were mostly destroyed by newer and more vigorous immigrants. A few of the more hardy ones survived, and are a standing testimony of the geographical revolutions of that part of Southern Europe.

That the Mediterranean area has undergone such profound geographical changes as I have endeavoured to indicate is no new theory. Many zoologists who have investigated the fauna of that region, and have attempted to explain the faunistic relations, had to acknowledge that the migrations must have taken place under geographical conditions entirely different from those obtaining at present. Rütimeyer long ago remarked that it seemed to him much more probable that Morocco, Algeria, and Tunis were peopled by way of Gibraltar, and perhaps also by Sicily and Malta from Europe, than Southern Europe from Africa. After careful conchological researches in the Western Mediterranean region, Dr. Kobelt came to the conclusion that formerly Southern Spain and Morocco must have been united by a broad land-connection. Sicily and Algeria do not apparently show any very intimate relationship conchologically, but farther west—in the mountains of Tetuan—Dr. Kobelt discovered a colony of Sicilian forms.[3]

"The close relationship," remarks Dr. Major (a, p. 106), "shown in the fauna of Corsica and Sardinia to Africa, permits the supposition that the connection with these islands had persisted to a much more recent date than that with Europe."

Many other authors have pointed out the close similarity existing between the faunas of Southern Europe and North Africa. We need only refer to the writings of Professor Suess, Milne-Edwards, and Boyd Dawkins. Mr. Blanchard went even so far as to say, "a comparer les plantes et les animaux de la Sicile et de la Tunésie, on se croirait sur le même terrain" (p. 1047).

No less than 113 species of phanerogamic plants are enumerated by Professor Engler (p. 53) as occurring in the Mediterranean coast region east and west of Italy without being found in that peninsula, or at least only in the extreme south of it. But he tells us that these species represent only a portion of such plants, which are extremely numerous.

In taking a general survey of these plants, Professor Engler is of opinion that their range implies that a large number of the Mediterranean species have migrated along a line which can be drawn between North Africa, Sicily, Greece, Crete, and Asia Minor, and that from this line the distribution started northward again.

Many of these plants then, and also some of the animals I have referred to, formed part of the older stream of migration which entered Europe from Asia Minor (vide Fig. 5, p. 117). There were only two courses open to them as they arrived on our continent during earlier Tertiary times. They could either go straight west towards Greece, or in a more northward direction to the newly-formed Alps. As the latter were raised, some of the immigrants were modified so as to adapt themselves to the new surroundings. Others became extinct; but a great many have persisted in the Alps to the present day and exhibit discontinuous distribution, having meanwhile disappeared in the intermediate tract between the latter and their original home in Asia. The lowlands of Eastern and Central Europe were either occupied by the sea or by large freshwater lakes, so as effectually to prevent a direct migration northward.

When the newer migrants arrived from Asia not only had the Alps risen to a lofty mountain chain acting as an effectual barrier, but Southern Italy and Greece had become disconnected. Some time after, Sicily and Southern Italy also became separated. Meanwhile the stream of migrants which consisted less and less of typically southern forms, emigrants from Central Asia and even Southern Siberia, mingled with the southern forms on their way to Europe, and these now poured across the newly opened plain of Central and Northern Europe. But it was not until some time after this that the Mediterranean Sea broke across the Ægean region, and that the Northern Sea retired from the plains of Eastern Russia to admit the typical Siberian fauna and flora into our continent (vide pp. 189-241).

I cannot close this chapter without referring to the active distributional centre—or I might say, centre of origin—of species situated in South-eastern Europe. No group of animals is more instructive in elucidating the paths of migration from this centre than the terrestrial mollusca. Wherever the original home of the genus Clausilia may have been in early Tertiary times, it is certain that the most active centre of origin is now, and has been for a considerable time past, in South-eastern Europe. One of the earliest migrants from that modern centre of this interesting genus is Clausilia bidentata, which is the only species found in Southern Spain, and one of the two met with in Ireland, and which has been observed in high altitudes in the Alps and in Scandinavia. As we go eastward from Western Europe the number of species of Clausilia, as we have seen, increases until we reach a maximum in the Balkan peninsula and the region of the Caucasus. Limax, Agriolimax, and Amalia, three genera of slugs, likewise appear to have originated in the same region and spread over Europe from there. Some species like Limax maximus and L. marginatus are very ancient, and probably commenced their wanderings in early Tertiary times. In this manner many animals of European origin have joined the Oriental migrants in their westward and also in their later northward travels. In a similar way species of plants and animals of Alpine origin might have joined these migrants in their northward course, and it is only when we come to carefully analyse the constituent parts of all these members which have come to us in England from the south, that we realise the complexity of their origin. Finally, even the Siberian migrants mingled with the later Oriental ones, and in some cases the decision as to whether a certain species belongs to the former or to the latter migration becomes a matter of great difficulty.

SUMMARY OF CHAPTER VI.

Like the last chapter, this deals with the Asiatic migrants. But while the former described the history of the northern invasion, those animals which entered Europe from the south-east are here more particularly referred to. They originated in Central, Southern, and Western Asia. It is not easy to discriminate in all cases between this Oriental migration and the Siberian. To a certain extent, even an entry of Northern Asiatic species has taken place by the southern route, and vice versâ. On the other hand, southern species might have come to Europe by the southern route—that is to say, to the south of the Caspian—and also by the northern, which lay to the north of that great inland sea. The Red Deer is a good example. It arrived on our continent by both routes. However, there is a racial difference in the members of the two migrations. The small race now found in Corsica, Sardinia, North-west Africa, and Western Europe, is probably the older of the two, while the larger one—resembling the American Wapiti Deer—arrived very much later from Siberia.

The Mammoth, Wild Boar, Badger, the Dippers and Pheasants, are all Oriental species which have come to us from the south-east; but there are also Reptiles and Amphibians, and a host of Invertebrates. Not all the animals, for instance, which have reached us in England from the south-east are of Asiatic origin. There is an active centre of distribution in South-eastern Europe itself, from which species radiate out in all directions. This fact is well illustrated by the genus Clausilia. Species from this centre, and also from the Alps, joined the Oriental stream in their northward course.

In reviewing a number of instances of Oriental species in Europe, one is struck by the peculiarity of their having apparently followed two distinct routes. All entered from Asia Minor, which is proved to have been connected with Greece until recent geological times. From here some seem to have proceeded straight west, others northward. Further study reveals the fact that the first route was followed by a much older set of migrants at a time when the Mediterranean area was greatly different from what it is at the present day. Greece was then joined to Southern Italy, Sicily, and Tunis. The latter was also connected with Sardinia and Corsica, and the Straits of Gibraltar did not exist. Under such geographical conditions a direct migration on land from Southern Greece to Spain was not only possible, but was actually undertaken by a very large number of Oriental species.

CHAPTER VII.
THE LUSITANIAN FAUNA.

In discussing the component elements of the British fauna and flora in the third chapter, I have already referred to the distinguishing characters of the Lusitanian migrants and to their distribution. I need only repeat, therefore, that these are now principally confined to the south-western portions of the British Islands. The late Edward Forbes was the first to trace the Lusitanian flora to its native home. In his classical memoir on the geological relations of the existing fauna and flora of the British Isles, he laid the foundations of a new method of research. We are as yet only beginning to realise the far-reaching conclusions obtainable by a careful study of the geographical distribution of animals and plants, though the lines of investigation were indicated by him more than fifty years ago. Forbes was of opinion that the Lusitanian element in the British flora was of miocene age, and that it survived the Glacial period on a now sunken land to the south-west of Ireland. Mr. Carpenter and myself agree in so far that we are both inclined to look upon this Lusitanian flora and the accompanying fauna in Ireland as of pre-glacial origin. But I am not quite satisfied that the Lusitanian migration ceased to come north then. It may have received a temporary check; but the presence, for instance, of the Dartford Warbler (Melizophilus undatus) in the south-east of England would seem to indicate that its northward migration took place in very recent times. It is possible also that the very restricted occurrence of the Dartford Warbler may imply that it is gradually withdrawing towards its centre of origin from a former wider range. Such an eventuality, as we have seen, has actually taken place in a great number of instances.

It is not only in the British Islands that we perceive the influence of the Lusitanian element. Scandinavia, Russia—indeed almost every part of Europe—can boast of some migrants which have originated in South-western Europe or on the mysterious lands which lay beyond it. As a rule, however, we notice a marked decrease of Lusitanian species as we travel eastward from Western Europe. Nevertheless, certain forms have travelled far beyond the confines of our continent, and we certainly meet with them in Asia and Northern Africa.

It is remarkable that we are apt to mistake sometimes for Lusitanian migrants species which are of Oriental origin. In a previous paper I classed such animals which had apparently originated in South-western Europe, but had really come from Asia by a circuitous southern route, with the Lusitanians. However, there is really no reason why the two should not be kept apart, provided we can discriminate between the pseudo-Lusitanians and the true ones. I have already indicated in the last chapter how these pseudo-Lusitanian migrants originated.

Supposing an Oriental species had left Asia for Europe in miocene times, it would on its arrival in Greece have had to decide between two courses. It could either advance into the newly-formed Alpine peninsula and there remain, or at once push on westward into Southern Italy, Sicily, and Tunis, by means of the old land-connections, and thence into Southern Spain. The Atlantic communicated at that time with the Mediterranean across the valley of the Guadalquivir; but that connection ceased to exist towards the end of the Miocene Epoch, when the Oriental migrants were free to ramble through Spain and the whole of the North European plain. I have indicated on a previous occasion (a, p. 484) that the earliest members of the Red Deer migration, which have left their traces in the caves of Malta, and whose descendants still live in Corsica, Sardinia, and North Africa, may have found their way to Northern Europe in this manner. Many other Asiatic mammals probably reached the British Islands in a similar way.

I cannot call to mind any large species of mammal which we might reasonably suppose to have originated in South-western Europe. Even among the smaller ones, few give us any definite clue in this respect. For instance, the present range of the genus Myogale—a small Insectivore belonging to the Mole family (Talpidæ)—teaches us nothing. The two living species show discontinuous distribution, and are almost confined to Europe. Myogale occurs fossil in French miocene deposits, but is unknown beyond the confines of our continent. It is therefore probably of West European origin. The gap between the South Russian M. moschata and the Spanish M. pyrenaica is bridged over in so far as we know from fossil evidence that the former had a much wider range in pleistocene times, being then found in England, Belgium, and Germany. Talpa, too,—to which genus our common Mole belongs,—seems to be a West European genus, since it occurs in French miocene deposits. However, it would be difficult to name many more recent genera which could be included in the area which I propose to investigate in this chapter. The genus Lepus is probably not of Lusitanian origin, but the sub-genus Oryctolagus—to which our common Rabbit belongs—has no doubt had its original home in that region. Only two species of Lepus (Oryctolagus) are known, one of which—Lepus lacostei—has been met with in French pliocene deposits. The other is the Rabbit (L. cuniculus). Though generally considered to have been introduced into the British Islands, no reason can be brought forward in favour of such a supposition, especially as it is known to have spread into Germany in pleistocene times from South-western Europe. It occurs in France, the Spanish peninsula, North-western Africa, and on some of the Mediterranean islands. Its nearest living relatives, as we should almost expect, are found in South America.

Of the Lusitanian Birds I have already mentioned the so-called Dartford Warbler (Melizophilus undatus), which ranges from the south of England to the extreme south-west of Europe. A second species occurs on the Balearic Islands and on Corsica, Sardinia, and Sicily. The Andalusian Bush-quail (Turnix sylvatica) is probably of North African origin, and has subsequently spread into Southern Spain and Portugal, and eastward as far as Sicily. It is an instance of a migrant utilising the old Mediterranean land-connections in the opposite direction from that described in the last chapter.

Two of our British Wagtails are very closely related, so much so that it requires a very critical eye to distinguish them even at close range. They also frequently interbreed. In their distribution, however, there is a considerable difference between the White Wagtail (Motacilla alba) and the Pied Wagtail (M. lugubris). While the former ranges almost all over Europe and Asia, the latter is a local form resident in the British Islands, Southern Scandinavia, and France, and a winter visitor to Spain and North-west Africa. The genus Motacilla is probably Oriental in its origin, but it seems as if the Pied Wagtail was a Lusitanian species which had gradually spread northward, only to return to South-western Europe in severe weather for shelter.

The Bearded Titmouse (Panurus biarmicus)—the only representative of the family Panuridæ—may possibly be a Lusitanian bird. The fact of its being absent from Scandinavia and Northern Russia is suggestive of a southern origin. It is doubtful whether the bird occurs on the south side of the Mediterranean, but it is common in the south of France and Spain, and has also been observed in Sicily, Greece, and Asia Minor. In Central Europe it is found sparingly, and eastward its range extends as far as Turkestan.

The genus Fringilla, which belongs to the great family of the Finches, appears to be not only of European origin, but, if the range of the species counts for anything, I should feel inclined to locate their home in the south-west. Altogether, five species are known. One of them, viz., Fringilla teydea, is confined to the Island of Teneriffe; another, F. madeirensis, is found in Madeira, the Canaries, and the Azores; a third, F. spodiogenys, inhabits North-west Africa. The two remaining species have a much wider range. F. cœlebs—the common Chaffinch—occurs in Europe, while its range extends eastward to Western Siberia, Persia, and Turkestan. The other—F. montifringilla, known as the Brambling—is more common in Northern Europe, and generally frequents the more northern latitudes of Asia as far as Japan.

It might be urged that the peculiar little blue Magpie of Spain—Cyanopolius Cooki—should find a place among the Lusitanian species, since there is no bird like it anywhere else in Europe. But in Eastern Siberia there lives a bird so closely allied as to be barely distinguishable from it. Nevertheless, since there are some distinguishing characters, it has received a distinct name—C. cyanus. This is a most interesting and remarkable case of discontinuous distribution, which may perhaps be explained by the supposition that the genus is of Oriental origin, and has died out at its former headquarters in Southern Asia and all along the line of migration, except at the extreme limits of the range in both directions—east and west.

As we go down in the scale of life—among the lower vertebrates and invertebrates—we meet with a greater number of prominent members of the Lusitanian migration. The Bullfinch, Dipper, and Chough, which might be thought to be of Lusitanian origin, are, as I have shown in the last chapter, Asiatic.

The European snakes seem to be all of eastern origin, unless Tropidonotus viperinus might be claimed as a Lusitanian form. Of very great interest from a zoogeographical point of view is our only European member of the South American and African family Amphisbænidæ. This species—Blanus cinereus—is of the size and shape of an ordinary earth-worm, from which, however, it may be distinguished by its snake-like wriggling motions. It lives under stones in Spain and Portugal, North-west Africa, and Greece. It has, therefore, a somewhat similar distribution to that of many of the animals and plants referred to in the last chapter. But here we have an animal which has evidently utilised the old Mediterranean route described on p. 271, from west to east. Two other species of Blanus inhabit Asia Minor and Syria, but most of its nearest relations either live in South America or tropical Africa. In migrating to North and West Africa, its ancestors probably made use of the land-bridge which spanned the Atlantic in early Tertiary times. Another Lusitanian Lizard—belonging not to an aberrant group, but to the typical Lacertidæ—is Psammodromus hispanicus. It is rather variable in colour—generally of a brown or green—and grows to a length at about four or five inches. It occurs throughout the Spanish peninsula and also in Southern France. One of the handsomest European Lizards, which reaches almost a foot in length,—of an olive colour with greenish or mother-of-pearl reflection, and with two yellow stripes along each side of the body,—is an allied species (P. algirus). From the Spanish peninsula it passes into Southern France and North Africa. Two other species of the genus are confined to North-west Africa.

It is quite possible that the genus Pelobates is of south-western origin. Of the two known species of this genus of Toads, one is found in the Central European plain and the other on the Spanish peninsula and in France. The closely allied Pelodytes punctatus, too, is confined to this south-western district, and their nearest relations are found in Mexico. Similarly, the genus to which the Midwife Toad (Alytes obstetricans) belongs may have its original home in that part of Europe. Of the two species, one is confined to France, Switzerland, Belgium, and Western Germany, and the other, viz., Alytes cisternasii, to Spain. Discoglossus pictus—a well-known and conspicuous Toad in Southern Europe—inhabits Spain, Algiers, and Tunis, the islands of Malta, Sicily, Sardinia, and Corsica. From the general range of the family Discoglossidæ, as given in Mr. Boulenger's excellent catalogue, it appears that nowhere in the vast space between China and New Zealand has any member of the family been discovered. The peculiar genus of Salamander—Chioglossa—is quite confined to the Spanish peninsula.

The Butterflies Nemeobius lucina and Charaxes jasius may also have had their home in that south-western district. To this migration also seems to belong the genus Gonepteryx, which has so peculiar a range in the British Islands. The only British species, known as the Brimstone Butterfly (Gonepteryx rhamni), occurs in the south of England and in the south and west of Ireland. It is met with over the greater part of Europe, and its range extends into Asia Minor and Northern India, and then it reappears again in distinct varieties in Japan and the Amur district. Three other species of Gonepteryx are known from Tibet and India, and one (G. cleopatra) from Southern Europe and Northern Africa. All the remaining species inhabit the west, viz., Brazil, Mexico, and Venezuela. That the genus has migrated from America eastward to Europe appears to be more probable than a migration in the opposite direction. At any rate, that an exchange of species between the south-western portion of the Holarctic Region and the Neotropical area took place is indicated by the fact, not only that a variety of G. cleopatra has been found in Madeira, but also that the Canary Islands possess a distinct form of Gonepteryx, viz., G. cleobule.

Dr. Kobelt has given us such an exhaustive memoir on the characteristic Mollusca of the different zoogeographical provinces of Europe, that we are particularly well informed as regards that group of Invertebrates. He tells us that the group Torquilla of the genus Pupa—which is a small chrysalis-like snail—is especially characteristic of the Pyrenees, Spain, and Portugal. In a certain measure they replace there the Clausiliæ which, as we have seen in the last chapter, have come from the east and are almost entirely absent in the south-west of Europe. Of about seventy species of Torquilla, the larger number are confined to this district, and some, which like Pupa (Torquilla) granum, range eastward, have travelled along the old Mediterranean highway, viâ Algiers, Sicily and Greece, to Asia Minor. They are still found along the whole of this route.

Similarly, we are told by the same author, that Gonostoma—a group of the large genus Helix—has a number of species in the same south-western district, while only one, viz., Helix obvoluta, occurs in England and Germany, and two in the Alps. Southward we again find many representatives crossing over to North Africa, among which Helix lenticula has a similar range to Pupa granum, which I have just referred to. The Alpine sub-genus Campylaea is quite absent in the Lusitanian district.

Among our own British testaceous Land Mollusca, several Helices, viz., Helix pisana, ericetorum, virgata, acuta, fusca, rotundata, aculeata, and probably many others, have come to us from the south-west. The species of Hyalinia are undoubtedly of very remote origin, and it would be futile at the present state of our knowledge to speculate as to their home. Some of our species may possibly be of British origin. Balea perversa is probably a south-western species, and certainly Pupa anglica, which is quite confined to Western Europe.

Fig. 18.—The Spotted Slug (Geomalacus maculosus).

Much more characteristic of South-western Europe, however, than these land-shells are some of the slugs. The peculiar genus Geomalacus is almost entirely confined to Portugal. One species, which I have had several occasions to refer to in illustration of the term "discontinuous distribution," ranges far beyond the confines of that country. This is Geomalacus maculosus (Fig. 18), first discovered in the south-west of Ireland, and more recently also in Portugal. Although careful search has been made for it in other parts of the British Islands, this slug has only been found in the portion of Ireland just indicated. Within the last few years I have taken it, up to a height of over a thousand feet, on the promontory north of the Kenmare River, also from sea-level up to a considerable height near Glengariff, and more recently Messrs. Praeger and Welch discovered it in abundance near the town of Kenmare. But beyond this rather circumscribed area in the counties of Cork and Kerry it does not occur (vide Fig. 19). Several Portuguese species of this interesting genus have since been added to science by Dr. Simroth and others. Dr. Simroth, too, has promulgated the view that the genus Arion—to which our common brown garden slug belongs—is of Lusitanian origin. Indeed, the number of species of Arion diminishes as we leave that province, though one extends beyond the borders of Europe into Siberia. The same number of species, viz. five, occur in Germany and in England. Testacella—a slug-like mollusc—which lives underground on earthworms, and of which genus three species, viz. T. maugei, T. haliotidea, T. scutulum, are known to inhabit the British Islands, is another Lusitanian animal. All the species are confined to Western Europe and North Africa; they do not even reach Germany or Switzerland.

Fig. 19.—Map of the British Islands on which the geographical distribution of Geomalacus maculosus is indicated in black.

I have had occasion to mention once before an extremely interesting genus of blind Woodlouse, viz., Platyarthrus. Like Testacella, it lives underground, and also resembles it in its general range. Its distribution is therefore of particular interest. It is difficult to conceive that Platyarthrus, from its peculiar mode of life could have crossed any formidable barrier, such as even a narrow straits of sea. Its occurrence in Spain and North Africa indicates, therefore, that the Straits of Gibraltar did not exist at the time when it undertook the migration southward, just as the English Channel and the Irish Sea could not have been there when it wandered to England and Ireland. The species which occurs in the south of England has a wide range in Ireland, and reaches in Scotland its most northern European limit of distribution. Platyarthrus is only one of the Lusitanian genera of woodlice. In Ireland—chiefly on the west coast—we also find a brilliantly coloured Woodlouse, which is absent from Great Britain, viz. Metoponorthus cingendus. It reappears again on the Continent in the south of France. Its range is therefore suggestive of a Lusitanian origin; and indeed, when we examine the general distribution of the genus Metoponorthus, we find that out of the forty-four known species, fully one-half are confined to Western Europe and North Africa.

My friend and colleague, Mr. Carpenter, informs me that among the Irish Spiders he is acquainted with, the following are to be looked upon as Lusitanian species:—

• Dysdera crocota.
• Oonops pulcher.
• Tegenaria hibernica.
• Theridion aulicum.
• Lasæola inornata.
• Agroeca celans.
• do.gracilipes.
• Teutana grossa.
• Cnephalocotes curtus.
• Porrhomma myops.

Of the Coleoptera, the genera Trichis, Glycia, and Singilis, all belonging to the Running Beetles (Carabidæ), are almost confined to the Spanish peninsula.

The beetles Rhopalomesites Tardyi, Eurynebria complanata, and Otiorrhynchus auropunctatus also belong to this fauna, as also the Earthworms Allolobophora veneta and A. Georgii, and the Millipede Polydesmus gallicus.

It will be evident to every one from these few instances of Lusitanian species, that somewhere in South-western Europe and North-western Africa, and also, perhaps, in a larger now submerged western land-area, there existed an active centre of development, from which animals spread in all directions.

If the presence of Platyarthrus in North-west Africa proves that the Straits of Gibraltar had come into existence after its southward migration, it also suggests that the ancestral home of this woodlouse was in the Spanish peninsula. Whether this supposition is correct or not, does not affect the Straits of Gibraltar problem, for in a migration northward into Spain from Morocco a land-connection would be equally necessary. Almost every group of vertebrates and invertebrates furnishes instances of species which must have crossed the Straits on dry land. Many naturalists have come to this conclusion, and have clearly expressed their views on the subject. At the commencement of the present period, says Mr. Bourguignat (p. 354), the north of Africa was a peninsula of Spain, the Straits of Gibraltar did not exist, and the Mediterranean communicated by the Sahara with the Atlantic.

The faunas of North-west Africa and the south-western portion of our continent are so closely related, that an uninterrupted intercourse by land must have existed for a very long period. The Mediterranean, however, throughout the Tertiary period—at any rate since miocene times—must have had almost constant communication with the Atlantic. According to Professor Suess, this was the case. The Atlantic was joined with the Mediterranean across the valley of the Guadalquivir during the Miocene Epoch, so that Andalusia must have belonged to North Africa in those days. The Straits of Gibraltar are supposed to have been formed in the next epoch. I have already expressed my disagreement with that theory from a zoogeographical point of view. The old Guadalquivir connection probably persisted much longer,—though interrupted by temporary periods of a partial retreat—so as to uncover sufficient land to allow of an interchange during miocene as well as pliocene times between the European and North African faunas. It is in this way, perhaps, that some of the members of the Alpine fauna have reached Spain by way of Corsica, Sardinia, and North-western Africa, and vice versâ. The Balearic Islands were then connected with Spain; and we find there many curious survivals which have long ago become extinct on the mainland.

That the Straits of Gibraltar are only of recent formation has been suggested on zoogeographical evidence by Bourguignat, Simroth, Kobelt, and many others. Dr. Kobelt believes that the former land-connection between the south of Spain and Morocco was much wider than is generally assumed, and that the coast-line stretched from Oran in Algeria straight across to Cartagena in Spain (b, ii., p. 228).

My allusions to the lands lying beyond the Lusitanian province, refer chiefly to the Canary Islands and Madeira. Whatever doubts Dr. Wallace had on the subject of their former connection with Morocco, it cannot be denied that they used to be of much larger extent, especially in miocene and pliocene times. It seems extremely probable that these islands formed part of the mainland of North Africa until comparatively recently, and that they are the last traces of a sunken continent which united Africa and South America. A discussion of this problem, however, must be deferred, as it is a complicated one, and one which would lead me altogether outside the scope of this little volume. I hope I shall have an opportunity to publish my views on this subject before long, meanwhile the reader must content himself with this mere statement.

During the greater portion of the Miocene, and I think for part of the Pliocene Epoch too, the advance of the Lusitanian species eastward was barred on the continent of Europe by an arm of the sea which stretched northward along the Rhone valley from the Mediterranean. The Lusitanian forms which originated in Southern Spain were able to travel east during these times by way of North-west Africa, Sicily, Southern Italy, and Greece; it is possible that some may have reached the Alps in this manner, and Eastern Europe generally. That the Lusitanian centre was never a very active one compared with, for instance, the Oriental is indicated by many distributional facts. It is difficult to understand, however, why the Oriental species, on the whole, have migrated so far west, while few Lusitanians have gone very far east. This seems to have been noted particularly in the case of the flora. Mr. Bonnet drew attention to the fact that in Tunis there are none of the absolutely characteristic plants of Morocco and Spain, while the Oriental flora is represented by a good many species. Lusitanian species have spread chiefly southward into North Africa, and northward into France, the British Islands, and even Scandinavia. As I have mentioned in the third chapter, there are a good many species of Lusitanian origin in the British Islands. However, we have only a mere remnant of what we ought to have, had the climate been less trying. It is probable, too, that the submergence destroyed a good many plants and the insects dependent on them. That the Lusitanian fauna is very ancient in the British Islands is proved by the fact of the discontinuous distribution of so many species. A greater number survived in Ireland than in England.