has yet been suggested, and it is doubtful if any name could be found sufficiently comprehensive to include all the races, long-headed and short-headed, fair and dark, tall and short, that we are at present content to group under this non-committal heading. Undoubtedly the term "Caucasic" cannot be defended on ethnical grounds. "Nowhere else in the world probably is so heterogeneous a lot of people, languages and religions gathered together in one place as along the chain of the Caucasus mountains[1000]." But we are no more called upon to believe that the "Caucasic" peoples originated in the Caucasus, than that the Semites are all descendants of Shem or Hamites of Ham. "Caucasic" has one claim that can never be disputed, that of priority, and it would be well if innovators in these matters were to take to heart the sober language of Ehrenreich, who reminds us that the accepted names are, what they ought to be, "purely conventional," and "historically justified,"

and "should be held as valid until something better can be found to take their place[1001]." It was considerations such as these, weighing so strongly in favour of current usage, that induced me _stare per vias antiquas_ in the _Ethnology_, and consequently also in the present work.

Hence, here as there, the Caucasic Division retains its t.i.tle, together with those of its main subdivisions--Hamitic, Semitic, Keltic, Slavic, h.e.l.lenic, Teutonic, Iranic, Galchic and so on.

The chief exception is "Aryan," a linguistic expression forced by the philologists into the domain of Ethnology, where it has no place or meaning. There was of course a time when a community, or group of communities, existed probably in the steppe region between the Carpathians and the Hindu-Kush[1002], by whom the Aryan mother-tongue was evolved, and who still for a time presented a certain uniformity in their physical characters, were, in fact, of Aryan speech and type. But while their Aryan speech persists in endlessly modified forms, they have themselves long disappeared as a distinct race, merged in the countless other races on whom they, perhaps as conquerors, imposed their Aryan language. Hence we can and must speak of Aryan tongues, and of an Aryan linguistic family, which continues to flourish and spread over the globe. But of an Aryan race there can be no further question since the absorption of the original stock in a hundred other races in remote prehistoric times. Where comprehensive references have to be made, I therefore subst.i.tute for Aryans and Aryan race the expression peoples of Aryan speech, at least wherever the unqualified term Aryan might lead to misunderstandings.

This way of looking at the question, which has now become more th.o.r.n.y than ever, has the signal advantage of being indifferent to any preconceived theories regarding the physical characters of that long vanished proto-Aryan race. How great this advantage is may be judged from the mere statement that, while German anthropologists are still almost to a man loyal to the traditional view that the first Aryans were best represented by the tall, long-headed, tawny-haired, blue-eyed Teutonic barbarians of Tacitus--who, Virchow tells us, have completely disappeared from sight in the present population--the Italian school, or at least its chief exponent, Sergi, was equally convinced that the picture was a myth, that such Aryans never existed, that "the true primitive Aryans were not long, but round-headed, not fair but dark, not tall but short, and are in fact to-day best represented by the round-headed Kelts, Slavs, and South Germans[1003]."

The fact is that the Aryan prototype has vanished as completely as has the Aryan mother-tongue, and can be conjecturally restored only by processes a.n.a.logous to those by which Schleicher and other philologists have endeavoured with dubious success to restore the organic Aryan speech as const.i.tuted before the dispersion.

But here arises the more important question, by what right are so many and such diverse peoples grouped together and ticketed "Caucasians"? Are they to be really taken as objectively one, or are they merely artificial groupings, arbitrarily arranged abstractions? Certainly this Caucasic division consists apparently of the most heterogeneous elements, more so than perhaps any other. Hence it seems to require a strong mental effort to sweep into a single category, however elastic, so many different peoples--Europeans, North Africans, West Asiatics, Iranians and others all the way to the Indo-Gangetic plains and uplands, whose complexion presents every shade of colour, except yellow, from white to the deepest brown or even black.

But they are grouped together in a single division, because of certain common characteristics, and because, as pointed out by Ehrenreich, who himself emphasises these objections, their substantial uniformity speaks to the eye that sees below the surface. At the first glance, except perhaps in a few extreme cases for which it would be futile to create independent categories, we recognise a common racial stamp in the facial expression, the structure of the hair, partly also the bodily proportions, in all of which points they agree more with each other than with the other main divisions. Even in the case of certain black or very dark races, such as the Beja, Somali, and a few other Eastern Hamites, we are reminded instinctively more of Europeans or Berbers than of negroes, thanks to their more regular features and brighter expression.

"Those who will accept nothing unless it can be measured, weighed, and numbered, may think perhaps that according to modern notions this appeal to the outward expression is unscientific. Nevertheless n.o.body can deny the evidence of the obvious physical differences between Caucasians, African Negroes, Mongols, Australians and so on. After all, physical anthropology itself dates only from the moment when we became conscious of these differences, even before we were able to give them exact expression by measurements. It was precisely the general picture that spoke powerfully and directly to the eye[1004]." The argument need not here be pursued farther, as it will receive abundant ill.u.s.tration in the details to follow.

Since the discovery of the New and the Austral Worlds, the Caucasic division as represented by the chief European nations has received an enormous expansion. Here of course it is necessary to distinguish between political and ethnical conquests, as, for instance, those of India, held by military tenure, and of Australia by actual settlement.

Politically the whole world has become Caucasic with the exception of half-a-dozen states such as China, Turkey, j.a.pan, Siam, Marocco, still enjoying a real or fict.i.tious autonomy. But, from the ethnical standpoint, those regions in which the Caucasic peoples can establish themselves and perpetuate their race as colonists are alone to be regarded as fresh accessions to the original and later (historical) Caucasic domains. Such fresh accessions are however of vast extent, including the greater part of Siberia and adjoining regions, where Slav branches of the Aryan-speaking peoples are now founding permanent new homes; the whole of Australia, Tasmania, and New Zealand, which have become the inheritance of the Caucasic inhabitants of the British Isles; large tracts in South Africa, already occupied by settlers chiefly from Holland and Great Britain; lastly the New World, where most of the northern continent is settled by full-blood Europeans, mainly British, French and German, while in the rest (Central and South America) the Caucasic immigrants (chiefly from the Iberian peninsula) have formed new ethnical groups by fusion with the aborigines. These new accessions, all acquired within the last 400 years, may be roughly estimated at about 28 million square miles, which with some 12 millions held throughout the historic period (Africa north of Sudan, most of Europe, South-West and parts of Central and South Asia, Indonesia) gives an extent of 40 million square miles to the present Caucasic domain, either actually occupied or in process of settlement. As the whole of the dry land scarcely exceeds 52 millions, this leaves not more than about 12 millions for the now reduced domains of all the other divisions, and even of this a great part (_e.g._ Tibetan table-land, Gobi, tundras, Greenland) is barely or not at all inhabitable. This, it may be incidentally remarked, is perhaps the best reply to those who have in late years given expression to gloomy forebodings regarding the ultimate fate of the Caucasic races. The "yellow scare" may be dismissed with the reflection that the Caucasian populations, who have inherited or acquired nearly four-fifths of the earth's surface besides the absolute dominion of the high seas, is not destined to be submerged by any conceivable combination of all the other elements, still less by the Mongol alone[1005].

Where have we to seek the primeval home of this most vigorous and dominant branch of the human family? Since no direct evidence can be cited, the answer necessarily takes the form of a hypothesis, and must rely mainly on the indirect evidence supplied by our vague knowledge of geographical conditions in pleistocene times, on past and present zoological distributions, with here and there, the a.s.sistance of a hint gleaned from archaeological discoveries. We may deal first with the arguments brought forward in favour of Africa north of Sudan. Here were found in quaternary times all the physical elements which zoologists demand for great specialisations--ample s.p.a.ce, a favourable climate and abundance of food, besides continuous land connection at two or three points across the Mediterranean, by which the pliocene and early pleistocene faunas moved freely between the two continents.

Many of the speculations on the subject failed to convince, largely because the writers took, so to say, the ground from under their own feet, by submerging most of the land under a vast "Quaternary Sahara Sea," which had no existence, and which, moreover, reduced the whole of North Africa to a Mauretanian island, a mere "appendix of Europe," as it is in one place expressly called. Then this inconvenient inland basin was got rid of, not by an outflow--being on the same level as the Atlantic, of which it was, in fact figured as an inlet--but by "evaporation," which process is however somehow confined to this inlet, and does not affect either the Mediterranean or the Atlantic itself. Nor is it explained how the oceanic waters were prevented from rushing in according "as the Sahara sea evaporated to become a desert." The attempt to evolve a "Eurafrican race" in such an impossible area necessarily broke down, other endless perplexities being involved in the initial geological misconception.

Not only was the Sahara dry land in pleistocene times, but it stood then at a considerably higher alt.i.tude than at present, although its mean elevation is still estimated by Chavanne at 1500 feet above sea-level.

"Quaternary deposits cover wide areas, and were at one time supposed to be of marine origin. It was even held that the great sand dunes must have been formed under the sea; but at this date it is scarcely necessary to discuss such a view. The advocates of a Quaternary Sahara Sea argued chiefly from the discovery of marine sh.e.l.ls at several points in the middle of the Sahara. But Tournouer has shown that to call in the aid of a great ocean in order to explain the presence of one or two sh.e.l.ls is a needless expenditure of energy[1006]."

At an alt.i.tude of probably over 2000 feet the Sahara must have enjoyed an almost ideal climate during late pliocene and pleistocene times, when Europe was exposed to more than one glacial invasion, and to a large extent covered at long intervals by a succession of solid ice-caps. We now know that these stony and sandy wastes were traversed in all directions by great rivers, such as the Ma.s.sarawa trending south to the Niger, or the Igharghar[1007] flowing north to the Mediterranean, and that these now dry beds may still be traced for hundreds of miles by chains of pools or lakelets, by long eroded valleys and by other indications of the action of running waters.

Nor could there be any lack of vegetable or animal life in a favoured region, which was thus abundantly supplied with natural irrigation arteries, while the tropical heats were tempered by great elevation and at times by the refreshing breezes from sub-arctic Europe.

From these well-watered and fertile lands, some of which continued even in Roman times to be the granary of the empire, came that succession of southern animals--hippopotamus, hyaena, rhinoceros, elephant, cave-lion--which made Europe seem like a "zoological appendix of Africa." In a.s.sociation with this fauna may have come man himself, for although North Africa has not yet yielded evidence of a widespread culture comparable to that of the Palaeolithic Age in Europe, yet the negroid characters of the Grimaldi skeletons have been held to prove an early connection between the opposite sh.o.r.es of the Mediterranean. The hypothesis of African origin is supported by archaeological evidence of the presence of early man all over North Africa from the sh.o.r.es of the Mediterranean through Egypt to Somaliland. Thus one of J. de Morgan's momentous conclusions was that the existence of civilised men in Egypt might be reckoned by thousands, and of the aborigines by myriads of years. These aborigines he identified with the men of the Old Stone Age, of whom he believed four stations to have been discovered--Dahshur, Abydos, Tukh, and Thebes[1008].

Of Tunisia a.r.s.ene Dumont declared that "the immense period of time during which man made use of stone implements is nowhere so strikingly shown." Here some of the flints were found in abundance under a thick bed of quaternary limestone deposited by the waters of a stream that has disappeared. Hence "the origin of man in Mauretania must be set back to a remote age which deranges all chronology and confounds the very fables of the mythologies[1009]."

The skeleton found in 1914 by Hans Reck at Oldoway (then German East Africa) was claimed to be of Pleistocene Age, but according to A. Keith "the evidence ... cannot be accepted as having finally proved this degree of antiquity[1010]."

The doctrine of the specialisation of the dolichocephalic European types in Africa, before their migrations northwards, lies at the base of Sergi's views regarding the African origin of those types. Arguing against the Asiatic origin of the Hamites, as held by Prichard, Virchow, Sayce and others, he points out that this race, scarcely if at all represented in Asia, has an immense range in Africa, where its several sub-varieties must have been evolved before their dispersion over a great part of that continent and of Europe. Then, regarding Hamites and Semites as essentially one, he concludes that Africa is the cradle whence this primitive stock "spread northwards to Europe, where it still persists, especially in the Mediterranean and its three princ.i.p.al peninsulas, and eastwards to West Asia[1011]."

The theory of an African cradle for the dolichocephalic Mediterranean type does not lack supporters, but when, relying on the undeniable presence of brachycephals, some writers would derive the Alpine type from the same area, the larger aspect of continental migrations appears to be overlooked (see pp. 451-2 below). To const.i.tute a distinct race, says Zaborowski, a wide geographical area is needed, such as is presented by both sh.o.r.es of the Mediterranean "with the whole of North Africa including the Sahara, which was till lately still thickly peopled[1012]." Then to the question by whom has this North African and Mediterranean region been inhabited since quaternary times, he answers "by the ancestors of our Libyans, Egyptians, Pelasgians, Iberians"; and after rejecting the Asiatic theory, he elsewhere arrives at "the grand generalisation that the whole of North Africa, connected by land with Europe in the Quaternary epoch, formed part of the geographical area of the ancient white race, of which the Egyptians, so far from being the parent stem, would appear to be merely a branch[1013]."

Coming to details, Bertholon[1014], from the human remains found by Carton at Bulla-Regia, determined for Tunisia and surrounding lands two main long-headed types, one like the Neandertal (occurring both in Khumeria, and in the stations abounding in palaeoliths), the other like the later Cro-Magnon dolmen-builders, whom De Quatref.a.ges had already identified with the tall, long-headed, fair, and even blue-eyed Berbers still met in various parts of Mauretania, and formerly represented in the Canary Islands[1015]. Bertholon agrees with Collignon that the Mauretanian megalith-builders are of the same race as those of Europe, and besides the two long-headed races describes (1) a short round-headed type in Gerba Island and East Tunisia[1016] representing the Libyans proper, and (2) a blond type of the Sahel, Khumeria, and other parts, whom he identifies with the Mazices of Herodotus, with the "Afri," whose name has been extended to the whole continent, and the blond Getulians of the Aures Mountains.

It has been objected that, as established by de Lapouge and Ripley, there are three distinct ethnical zones in Europe:--(1) Nordic: the tall, fair, long-headed northern type, commonly identified by the Germans with the race represented by the osseous remains from the "Reihengraber," _i.e._ the "Germanic," which the French call Kymric or Aryan, for which de Lapouge reserves Linne's _h.o.m.o europaeus_, and to which Ripley applies the term "Teutonic," because the whole combination of characters "accords exactly with the descriptions handed down to us by the ancients. Such were the Goths, Ostrogoths, Visigoths, Vandals, Lombards, together with the Danes, Nors.e.m.e.n, Saxons.... History is thus corroborated by natural science." (2) Mediterranean: the southern zone of short, dark, long-heads, _i.e._ the primitive element in Iberia, Italy, South France, Sicily, Corsica, Sardinia, and Greece, called Iberians by the English, and identified by many with the Ligurians, Pelasgians, and allied peoples, grouped together by Ripley as Mediterraneans[1017]. (3) Alpine: the central zone of short, medium-sized round-heads with light or chestnut hair, and gray or hazel eye, de Lapouge's and Ripley's _h.o.m.o alpinus_, the Kelts or Kelto-Slavs of the French, the Ligurians or Arvernians of Beddoe and other English writers. Here belong the tall Armenoids, the Armenians being descendants of the Hitt.i.tes.

The question is, Can all these have come from North Africa? We have seen that this region has yielded the remains of one round-headed and two long-headed prehistoric types. Henri Malbot pointed out that, as far back as we can go, we meet the two quite distinct long-headed Berber types, and he holds that this racial duality is proved by the megalithic tombs (dolmens) of Roknia between Jemmapes and Guelma, possibly some 4000 or 5000 years old. The remains here found by L. L. C. Faidherbe belong to two different races, both dolichocephalic, but one tall, with prominent zygomatic arches and very strong nasal spine (it reads almost like the description of a brawny Caledonian), the other short, with well-balanced skull and small nasal spine[1018]. The earliest (Egyptian) records refer to brown and blond populations living in North Africa some 5000 years ago, and it has been claimed that the raw materials, so to say, were here to hand both of the fair northern and dark southern European long-heads.

These different races were represented even amongst the extinct Guanches of the Canary Islands, as shown by a study of the 52 heads procured in 1894 by H. Meyer from caves in the archipelago[1019]. Three distinct types are determined: (1) Guanche, akin to the Cro-Magnon, tall (5 ft. 8 in. to 6 ft. 2 in.), robust, dolicho (78), low, broad face; large eyes, rather short nose; fair, reddish or light chestnut hair; skin and eyes light; ranged throughout the islands, but centred chiefly in Tenerife; (2) "_Semitic_," short (5 ft. 4 or 5 in.), slim, narrow mesocephalic head (81), narrow, long face, black hair, light brown skin, dark eyes; range, Grand Canary, Palma, and Hierro; (3) _Armenoid_, akin to von Luschan's pre-Semitic of Asia Minor; shorter than 1 and 2; very short, broad, and high skull (hyperbrachy, 84); hair, skin and eyes very probably of the West Asiatic brunette type; range, mainly in Gomera, but met everywhere. Many of the skulls had been trepanned, and these are brought into direct a.s.sociation with the full-blood Berbers of the Aures Mts. in Algeria, who still practise trepanning for wounds, headaches, and other reasons. This type is scarcely to be distinguished from Lapouge's short brown _h.o.m.o alpinus_, which dates from the Stone Ages, and is found in densest ma.s.ses in the Central Alpine regions, but the true Armenoids are differentiated by their taller stature[1020].

How numerous were the inhabitants of France at that time may be inferred from the long list of no less than 4000 neolithic stations given for that region by Ph. Salmon. Of the 688 skulls from those stations measured by him, 57.7 per cent. are cla.s.sed as dolicho, 21.2 as brachycephalic, and 21.1 as intermediate. This distinguished palethnologist regards the intermediates as the result of crossings between the two others, and of these he thinks the first arrivals were the round-heads, who ranged over a vast area between Brittany, the Channel, the Pyrenees, and the Mediterranean, 60 per cent. of the graves. .h.i.therto studied containing skulls of this type[1021]. Belgium also, where a mixture of long- and round-heads is found amongst the men of Furfooz, must be included in this neolithic brachy domain, which can be traced as far westward as the British Isles[1022]. Attempts have been made, as indicated above, to derive these brachycephals, as well as the dolichocephals, from North Africa, in accordance with the view that the latter region was the true centre of evolution and of dispersion for all the main branches of the Caucasic family, but this theory has few supporters at the present time. Sergi recognised the Asiatic origin of the neolithic round-heads and regarded them as "peaceful infiltrations[1023]," forerunners of the great invasions of the later Metal Ages. Verneau points out[1024] that when all the neolithic stations in which brachycephalic skulls have been discovered are plotted out on a map of Europe it is easy to recognise a current running almost directly from east to west. Moreover towards the west this current divides, being clearly separated by zones of dolichocephaly.

Evidence of the presence in early times of tall blond peoples in Africa, side by side with a short dark population, and of brachycephals together with dolichocephals, proves that even in the Stone Age ethnic mixtures had already taken place, and racial purity--if indeed it ever existed--must be sought for in still remoter periods.

With Sergi's view which traces the neolithic inhabitants of the northern sh.o.r.es of the Mediterranean (Iberians, Ligurians, Messapians, Siculi and other Itali, Pelasgians), to North Africa, most anthropologists agree[1025]. Also that all or most of these were primarily of a dark (brown), short, dolicho type, which still persists both in South Europe and North Africa, and in fact is the race which Ripley properly calls "Mediterranean," although in the west they almost certainly ranged into Brittany and the British Isles. But there are some who hold that the migration was in the opposite direction, and derive the North African branch from Europe, rather than the European branches from Africa.

"Anthropologists who have specially studied the question of the Berbers or Kabyles have concluded that they are descendants of prehistoric European invaders who occupied the tracts that suited them best[1026]."

In France the neolithic "Mediterranean type" has been regarded as lineally descended from palaeolithic predecessors _in situ_[1027]. Some would even go further still, and claim Europe as the place of origin not only of the Mediterranean but also of the Alpine and Northern branches.

"The so-called three races of Europe are in the main the result of variation from a common European stock, a variation due to isolation and natural selection[1028]."

Without making any claim to finality the following perhaps best represents orthodox opinion at the present time. It may be a.s.sumed that man evolved somewhere in Southern Asia in pliocene times, and that the early groups possessed a tendency to variability which was directed to some extent by geographical conditions and became fixed by isolation.

The tall fair blue-eyed dolichocephals (Northern Race) and the short dark dolichocephals (Mediterranean Race) may be regarded as two varieties of a common stock, the former having their area of characterisation in the steppes north of the plateaus of Eur-Asia, and migrating eastwards and westwards as the country dried after the last glacial phase. The southern branch, entering East Africa from Southern Asia, spread all over North Africa; those in the east were the archaic Egyptians; to the west were the Libyans whose descendants are the Berbers; those who crossed the Mediterranean formed the European branches of the Mediterranean race. With regard to the third type, while the central plateaus of Asia were the centre of dispersal for the true Mongols the western plateaus were the area of characterisation of a non-Mongolian brachycephalic race, which includes short and tall varieties. This is the Alpine race, which extends from the Hindu Kush to Brittany, and formerly spread further westwards into the British Isles[1029].

The problem of European origins has often in the past been obscured rather than enlightened by an appeal to linguistics, but linguistic factors cannot altogether be ignored. No doubt the earliest populations of the Mediterranean sh.o.r.es during the Stone Age spoke non-Aryan languages, but it is only here and there that traces--mostly indecipherable--can be discovered. On the African side we have the Berber language still in its full vigour; and apparently little changed for thousands of years. But in Europe the primitive tongues have everywhere been swept away by the Aryan (h.e.l.lenic, Italic, Keltic) except in the region of the Pyrenees. In Italy Etruscan is the only language which can with safety be called non-Aryan[1030], though the place of Ligurian is still under dispute[1031]. Of Pelasgian, nothing survives except the statement of Herodotus, a dangerous guide in this matter, that it was a barbaric tongue like the peoples themselves[1032], but Ridgeway considers it Indo-European[1033]. Further east, in Asia Minor, neither Karian inscriptions and glosses nor occasional Lydian[1034] and Mysian glosses afford any safe basis for establishing relationships[1035]; the fuller evidence of Lycian leaves its position indeterminate[1036] and the Cretan script is still undeciphered[1037].

But in Iberia besides the Iberian inscriptions, which, so far, remain indecipherable[1038], there survives the Basque of the western Pyrenees, which beyond question represents a form of speech which was current in the peninsula in pre-Aryan times, and on the a.s.sumption of a common origin of the populations on both sides of the Strait of Gibraltar might be expected to show traces of kinship with Berber. In a posthumous work on this subject[1039], the eminent philologist G. von der Gabelenz goes much further than mere traces, and claims to establish not only phonetic and verbal resemblances, but structural correspondences, so that his editor Graf von der Schulenberg was satisfied as to the relationship of the two languages[1040]. This conclusion has not, however, met with general acceptance[1041] and the affinities of Basque with Finno-Ugrian cannot be overlooked[1042]. A study of the physical features of the modern Basques adds complexity to the problem. Most observers are agreed that a distinct Basque type exists, and this physical and linguistic singularity has led to various more or less fanciful theories "connecting the Basques with every outlandish language and bankrupt people under the sun[1043]," while G. Herve[1044] would regard them as forming by themselves a separate ethnic group, "a fourth European race."

On the other hand Feist[1045] has grounds for claiming that the Basques are not, in anthropological respects, essentially different from their Spanish or French neighbours (p. 357) and Jullian[1046] denies them more than a superficial unity. These apparently conflicting opinions are reconciled by the conclusions of R. Collignon[1047], himself one of the best authorities on the subject. "The physical traits characteristic of the Basques attach them unquestionably ('indiscutablement') to the great Hamitic branch of the white races, that is to say, to the ancient Egyptians and to the various groups commonly comprised under the collective name of Berbers. Their brachycephaly, slight as it is, cannot outweigh the aggregate of the other characters which they present.... It is therefore in this direction and not amongst Finns or Esthonians that is to be sought the parent stem of this paradoxical race. It is North African or European, a.s.suredly not Asiatic." Collignon's explanation of the Basque type is that it is a sub-species of the Mediterranean stock evolved by long-continued and complete isolation, and in-and-in breeding, primarily engendered by peculiarity of language. The effects of heredity, aided perhaps by artificial selection, have generated local peculiarities and have developed them to an extreme[1048].

"The Iberian question," says Rice Holmes, "is the most complicated and difficult of all the problems of Gallic ethnology[1049]." From the testimony of Greek and Roman authors, he draws the following conclusions. "The name Iberian was probably applied, in the first instance, only to the people who dwelt between the Ebro and the Pyrenees. The Iberians once occupied the seaboard of Gaul between the Rhone and the Pyrenees; but Ligurians encroached upon this part of their territory. They also probably occupied the whole eastern region of the Spanish peninsula. But," he adds, "we must bear in mind that the data are both insufficient and uncertain" (p. 288). Later (p. 301), reviewing the evidence collected by philologists and by craniologists, he continues, "it seems to me probable that the Iberians comprised both people who spoke, or whose ancestors had spoken, Basque, and people who spoke the language or languages[1050] of the 'Iberian' inscriptions; that to observers who had not learned to measure skulls and knew nothing of scientific methods, they appeared to be h.o.m.ogeneous; that the prevailing type was that which is now called Iberian and is seen at its purest in Sardinia, Corsica and Sicily; but that a certain proportion of the whole population may have been characterised by physical features more or less closely resembling those which the modern Basques--French and Spanish--possess in common, and which, as MM. Broca and Collignon tell us, distinguish them from all other European peoples. Finally it seems probable that the true Iberians were the people who spoke the languages of the inscriptions, and that Basque was spoken by a people who occupied Spain and Southern Gaul before the Iberians arrived. But unless and until the key to those appalling inscriptions is found, the problem will never be solved."

The Ligurian question is still more complex than the Iberian. For while no facts can be brought forward in direct contradiction of the a.s.sumption that the Iberians were a short dark dolichocephalic population occupying the Iberian peninsula in the Stone Age, and speaking a non-Indo-European language, no such generalisations with regard to race, physical type, culture, geographical distribution or language are accepted for the Ligurians. Some, with Sergi[1051], consider the Ligurians merely as another branch of the Mediterranean race. Others, with Zaborowski[1052], tracing their presence among the modern inhabitants of Liguria, regard them as representing the small, dark, brachycephalic race at its purest. While many who recognise the Ligurians as belonging to the Mediterranean physical type deny their affinity with the Iberians. Meyer[1053] considers such a relationship "not improbable," but Dechelette[1054] shows that it is absolutely untenable on archaeological grounds. The geographical range is equally uncertain. C. Jullian[1055] distributes Ligurians not only over the whole of Gaul, but also throughout Western Europe, and attributes to them all the glories of neolithic civilisation; A. Bertrand[1056] thinks that they played even in Gaul merely a secondary role; Dechelette[1057], on archaeological evidence, proves that the Ligurian period was _par excellence_ the Age of Bronze, and Ridgeway[1058] identifies it with the Terramare civilisation. Finally, if we follow Sergi, the Ligurians must have spoken a non-Indo-European language; but the most eminent authorities are in the main agreed that such traces of Ligurian as remain show affinities with Indo-European[1059]. With regard to their physical type Sergi puts forward the view that the true Ligurians were like the Iberians, a section of the long-headed Mediterranean (Afro-European) stock. From prehistoric stations in the valley of the Po he collected 59 skulls, all of this type, and all Ligurian; history and tradition being of accord that before the arrival of the Kelts this region belonged to the Ligurian domain. "If it be true that prehistoric Italy was occupied by the Mediterranean race and by two branches--Ligurian and Pelasgian--of that race, the ancient inhabitants of the Po valley, now exhumed in those 59 skulls, were Ligurian[1060]."

These Ligurians have been traced from their homes on the Mediterranean into Central Europe. From a study of the neolithic finds made in Germany, in the district between Neustadt and Worms, C. Mehlis[1061]

infers that here the first settlers were Ligurians, who had penetrated up the Rhone and Saone into Rhineland. In the Kircherian Museum in Rome he was surprised to find a marked a.n.a.logy between objects from the Riviera and from the Rhine; skulls (both dolicho), vases, stone implements, mill-stones, etc., all alike. Such Ligurian objects, found everywhere in North Italy, occur in the Rhine lands chiefly along the left bank of the main stream between Basel and Mainz, and farther north in the Rheingau at Wiesbaden, and in the Lahn valley.

The Ligurians may of course have reached the Riviera round the coast from Illiberis and Iberia; but the same race is found as the aboriginal element also at the "heel of the boot," and in fact throughout the whole of Italy and all the adjacent islands. This point is now firmly established, and not only Sergi, but several other leading Italian authorities hold that the early inhabitants of the peninsula and islands were Ligurians and Pelasgians, whom they look upon as of the same stock, all of whom came from North Africa, and that, despite subsequent invasions and crossings, this Mediterranean stock still persists, especially in the southern provinces and in the islands--Sicily, Sardinia, and Corsica. Hence it seems more reasonable to bring this aboriginal element straight from Africa by the stepping stones of Pantellaria, Malta, and Gozzo (formerly more extensive than at present, and still strewn with megalithic remains comparable to those of both continents), than by the roundabout route of Iberia and Southern Gaul[1062]. This is a simple solution of the problem, but it is a question if it is justifiable to extend the name Ligurian to all that branch of the Mediterranean race which undoubtedly forms the substratum of population in Italy and parts of Gaul, ignoring the presence or absence of "Ligurian" culture or traces of Ligurian language.

Dechelette[1063], relying chiefly upon archaeological and cultural evidence, sums up as follows: we must consider the Ligurians as Indo-European tribes, whose area of domination had its centre, during the Bronze Age, in North Italy, and the left bank of the Rhone. They were enterprising and energetic in agriculture and in commerce. Together with neighbouring peoples of Illyrian stock they engaged in an indirect but nevertheless regular trade with the northern regions where amber was collected. Among the Ligurians, as among the Illyrians and Hyperboreans, a form of heliolatry was prevalent, popularising the old solar myths in which the swan appears to have played an important role. Rice Holmes[1064] defines more closely their geographical range. "Ligurians undoubtedly lived in South-eastern Gaul, where they were found at least as far north as Bellegarde in the department of the Ain; and, mingled more or less with Iberians, in the departments of the Gard, Herault, Aude and Pyrenees-Orientales. Most probably they had once occupied the whole eastern region as far north as the Marne, but had been submerged by Celts: and perhaps they had also pushed westward as far as Aquitania." He continues, "Were it possible to regard the theory of MM.

d'Arbois de Jubainville and Jullian as more than an interesting hypothesis, we should have to conclude that the Ligurians were simply the long-headed and short-headed peoples who, reinforced perhaps from time to time by hordes of immigrants, had inhabited the whole of Gaul since the Neolithic Age, and of whom the former, or many of them, were descended from palaeolithic hunters; in other words that they were the same people who, after they had been conquered by, or had coalesced with, the Celtic invaders, called themselves _Celtae_: but to say which of them were first known as Ligurians or introduced the Ligurian language would be utterly hopeless. Finally the little evidence we possess tends to show that the people called Ligurians, when they became known to the Greek writers who described them, were a medley of different races."

For Sicily, with which may practically be included the south of Italy, we have the conclusions of G. Patroni based on years of intelligent and patient labours[1065]. To Africa this archaeologist traces the palaeolithic men of the west coast of Sicily and of the caves near Syracuse explored by Von Adrian[1066]. "We are forced to conclude that man arrived in Sicily from Africa at a time when the isthmus connecting the island with that Continent still stood above sea-level. He made his appearance about the same time as the elephant, whose remains are a.s.sociated with human bones especially in the west. He followed the sea coasts, the sh.e.l.ls of which offered him sufficient food[1067]." He was followed by the neolithic man, whose presence has been revealed by the researches of Paolo Orsi at the station of Stentinello on the coast north of Syracuse.

To Orsi is also due the discovery of what he calls the "Aeneolithic Epoch[1068]," represented by the bronzes of the Girgenti district. Orsi a.s.signs this culture to the _Siculi_, and divides it into three periods, while regarding the neolithic men of Stentinello as _pre-Siculi_. But Patroni holds that the aeneolithic peoples have a right to the historic name of _Sicani_, and that the true Siculi were those that arrived from Italy in Orsi's second period. It seems no longer possible to determine the true relations of these two peoples, who stand out as distinct throughout early historic times. They are by many[1069] regarded as of one race, although both ([Greek: Sikanos, Sikelos]) are already mentioned in the Odyssey. But the evidence tends to show that the Sicani represent the oldest element which came direct from Africa in the Stone Age, while the Siculi were a branch of the Ligurians driven in the Metal Age from Italy to the island, which was already occupied by the Sicani, as related by Dionysius Halicarna.s.sus[1070]. In fact this migration of the Siculi may be regarded as almost an historical event, which according to Thucydides took place "about 300 years before the h.e.l.lenes came to Sicily[1071]." The Siculi bore this national name on the mainland, so that the modern expression "Kingdom of the Two Sicilies"

(the late Kingdom of Naples) has its justification in the earliest traditions of the people. Later, both races were merged in one, and the present Sicilian nation was gradually const.i.tuted by further accessions of Phoenician (Carthaginian), Greek, Roman, Vandal, Arab, Norman, French and Spanish elements.

Very remarkable is the contrast presented by the conditions prevailing in this ethnical microcosm and those of Sardinia, inhabited since the Stone Ages by one of the most h.o.m.ogeneous groups in the world. From the statistics embodied in R. Livi's _Antropologia Militare_[1072] the Sards would almost seem to be cast all in one mould, the great bulk of the natives having the shortest stature, the brownest eyes and hair, the longest heads, the swarthiest complexion of all the Italian populations.

"They consequently form quite a distinct variety amongst the Italian races, which is natural enough when we remember the seclusion in which this island has remained for so many ages[1073]." They seem to have been preserved as if in some natural museum to show us what the Ligurian branch of the Mediterranean stock may have been in neolithic times. Yet they were probably preceded by the microcephalous dwarfish race described by Sergi as one of the early Mediterranean stocks. Their presence in Sardinia has now been determined by A. Niceforo and E. A.

Onnis, who find that of about 130 skulls from old graves thirty have a capacity of only 1150 c.c. or under, while several living persons range in height from 4 ft. 2 in. to 4 ft. 11 in. Niceforo agrees with Sergi in bringing this dwarfish race also from North Africa[1074].

With remarkable cranial uniformity, similar phenomena are presented by the Corsicans who show "the same exaggerated length of face and narrowness of the forehead. The cephalic index drops from 87 and above in the Alps to about 75 all along the line. Coincidently the colour of hair and eyes becomes very dark, almost black. The figure is less amply proportioned, the people become light and rather agile. It is certain that the stature at the same time falls to an exceedingly low level: fully 9 inches below the average for Teutonic Europe," although "the people of Northern Africa, pure Mediterranean Europeans, are of medium size[1075]."

In the Italian peninsula Sergi holds not only that the aborigines were exclusively of Ligurian, _i.e._ Mediterranean stock, but that this stock still persists in the whole of the region south of the Tiber, although here and there mixed with "Aryan" elements. North of that river these elements increase gradually up to the Italian Alps, and at present are dominant in the valley of the Po[1076]. In this way he would explain the rising percentage of round-heads in that direction, the Ligurians being for him, as stated, long-headed, the "Aryans" round-headed.

Similarly Beddoe, commenting on Livi's statistics, showing predominance of tall stature, round heads, and fair complexion in North Italy, infers "that a type, the one we usually call the Mediterranean, does really predominate in the south, and exists in a state of comparative purity in Sardinia and Calabria; while in the north the broad-headed Alpine type is powerful, but is almost everywhere more or less modified by, or interspersed with other types--Germanic, Slavic, or of doubtful origin--to which the variations of stature and complexion may probably be, at least in part, attributed[1077]."

Similar relations prevail in the Balkan peninsula, where the Mediterranean stock is represented by the "Pelasgic[1078]" substratum.

Invented, as has been said, for the purpose of confounding future ethnologists, these Pelasgians certainly present an extremely difficult racial problem, the solution of which has. .h.i.therto resisted the combined attacks of ancient and modern students. When Dionysius tells us bluntly that they were Greeks[1079], we fancy the question is settled off-hand, until we find Herodotus describing them a few hundred years earlier as aliens, rude in speech and usages, distinctly not Greeks, and in his time here and there (Thrace, h.e.l.lespont) still speaking apparently non-h.e.l.lenic dialects[1080]. Then Homer several centuries still earlier, with his epithet of [Greek: dioi], occurring both in the _Iliad_ and the _Odyssey_[1081], exalts them almost above the level of the Greeks themselves. It would seem, therefore, almost impossible to discover a key to the puzzle, one which will also fit in both with Sergi's Mediterranean theory, and with the results of recent archaeological researches in the Aegean lands. The following hypothesis is supported by a certain amount of evidence. If the pre-Mykenaean culture revealed by Schliemann and others in the Troad, Mykenae, Argos, Tiryns, by Evans and others in Crete, by Cesnola in Cyprus, be ascribed to a pre-h.e.l.lenic rather than to a proto-h.e.l.lenic people, then the cla.s.sical references will explain themselves, while this pre-h.e.l.lenic race will be readily identified with the Pelasgians, as this term is understood by Sergi.

It is, I suppose, universally allowed that Greece really was peopled before the arrival of the h.e.l.lenes, which term is here to be taken as comprising all the invading tribes from the north, of which the Achaeans were perhaps the earliest. On their arrival the h.e.l.lenes therefore found the land not only inhabited, but inhabited by a cultured people more civilised than themselves, who could thus be identified with Sergi's Pelasgian branch of the Mediterranean or Afro-European stock, whom the proto-h.e.l.lenes naturally regarded as their superiors, and whom their first singers also naturally called [Greek: dioi Pelasgoi][1082]. But in the course of a few centuries[1083] these Pelasgians became h.e.l.lenised, all but a few scattered groups, which lagging behind in the general social progress are now also looked upon as barbarians, speaking barbaric tongues, and are so described by contemporary historians. Then these few remnants of a glorious but forgotten past are also merged in the h.e.l.lenic stream, and can no longer be distinguished from other Greeks by contemporary writers. Hence for Dionysius the Pelasgians are simply Greeks, which in a sense may be true enough. All the heterogeneous elements have been fused in a single h.e.l.lenic nationality, built upon a rough Pelasgic substratum, and adorned with all the graces of h.e.l.lenic culture.

Now to make good this hypothesis, it is necessary to show, first, that the Pelasgians were not an obscure tribe, a small people confined to some remote corner of h.e.l.las, but a widespread nation diffused over all the land; secondly, that this nation, as far as can now be determined, presented mental and other characters answering to those of Sergi's Mediterraneans, and also such as might be looked for in a race capable of developing the splendid Aegean culture of pre-h.e.l.lenic times.

On the first point it has been claimed that the Pelasgians were so widely distributed[1084] that the difficulty rather is to discover a district where their presence was unknown. They fill the background of h.e.l.lenic origins, and even spread beyond the h.e.l.lenic horizon, to such an extent that there seems little room for any other people between the Adriatic and the h.e.l.lespont. Thus Ridgeway[1085] has brought together a good many pa.s.sages which clearly establish their universal range, as well as their occupation especially of those places where have been found objects of Mykenaean and pre-Mykenaean culture, such as engraved gems, pottery, implements, buildings, inscriptions in pictographic and syllabic scripts. In Crete they had the "great city of Knossos" in Homer's time[1086]; not only was Mykenae theirs, but the whole of Peloponnesus took the name of Pelasgia; the kings of Tiryns were Pelasgians, and Aeschylus calls Argos a Pelasgian city; an old wall at Athens was attributed to them, and the people of Attica had from all time been Pelasgians[1087]. Orchomenus in Boeotia was founded by a colony from Pelasgiotis in Thessaly; Lesbos also was called Pelasgia, and Homer knew of Pelasgians in the Troad. Their settlements are further traced to Egypt, to Rhodes, Cyprus, Epirus--where Dodona was their ancient shrine--and lastly to various parts of Italy.

Moreover, the Pelasgians were traditionally the civilising element, who taught people to make bread, to yoke the ox to the plough, and to measure land. It would appear from these and other allusions that there were memories of still earlier aborigines, amongst whom the Pelasgians appear as a cultured people, introducing perhaps the arts and industries of the pre-Mykenaean Age. But the a.s.sumption, based on no known data, is unnecessary, and it seems more reasonable to look on this culture as locally developed, to some extent under eastern (Egyptian, Babylonian, Hitt.i.te?) influences[1088]. Here it is important to note that the Pelasgians were credited with a knowledge of letters[1089], and all this has been advanced as sufficient confirmation of our second postulate.