Admirable alike for its manifold apt.i.tudes and surprising vitality, the language of Aeschylus, Thucydides, and the other great Athenians outlived all the vicissitudes of the Byzantine empire, during which it was for a time banished from Southern Greece, and even still survives, although in a somewhat degraded form, in the Romaic or Neo-h.e.l.lenic tongue of modern h.e.l.las. Romaic, a name which recalls a time when the Byzantines were known as "Romans" throughout the East, differs far less from the cla.s.sical standard than do any of the Romance tongues from Latin. Since the restoration of Greek independence great efforts have been made to revive the old language in all its purity, and some modern writers now compose in a style differing little from that of the cla.s.sic period.

Yet the h.e.l.lenic race itself has almost perished on the mainland. Traces of the old Greek type have been detected by Lenormant and others, especially amongst the women of Patras and Missolonghi. But within living memory Attica was still an Albanian land, and Fallmerayer has conclusively shown that the Peloponnesus and adjacent districts had become thoroughly Slavonised during the 6th and 7th centuries[1293].

"For many centuries," writes the careful Roesler, "the Greek peninsula served as a colonial domain for the Slavs, receiving the overflow of their population from the Sarmatian lowlands[1294]." Their presence is betrayed in numerous geographical terms, such as _Varsova_ in Arcadia, _Glogova_, _Tsilikhova_, etc. Nevertheless, since the revival of the h.e.l.lenic sentiment there has been a steady flow of Greek immigration from the Archipelago and Anatolia; and the Albanian, Slav, Italian, Turkish, Rumanian, and Norman elements have in modern Greece already become almost completely h.e.l.lenised, at least in speech. Of the old dialects Doric alone appears to have survived in the Tsaconic of the Laconian hills. The Greek language has, however, disappeared from Southern Italy, Sicily, Syria, and the greater part of Egypt and Asia Minor, where it was long dominant.

To understand the appearance of SLAVS in the Peloponnesus we must go back to the Eurasian steppe, the probable cradle of these mult.i.tudinous populations. Here they have often been confused with the ancient Sarmatae, who already before the dawn of history were in possession of the South Russian plains between the Scythians towards the east and the proto-Germanic tribes before their migration to the Baltic lands. But even at that time, before the close of the Neolithic Age, there must have been interminglings, if not with the western Teutons, almost certainly with the eastern Scythians, which helps to explain the generally vague character of the references made by cla.s.sical writers both to the Sarmatians and the Scythians, who sometimes seem to be indistinguishable from savage Mongol hordes, and at others are represented as semi-cultured peoples, such as the Aryans of the Bronze period might have been round about the district of Olbia and the other early Miletian settlements on the northern sh.o.r.es of the Euxine.

Owing to these early crossings Andre Lefevre goes so far as to say that "there is no Slav race[1295]," but only nations of divers more or less pure types, more or less crossed, speaking dialects of the same language, who later received the name of Slavs, borne by a prehistoric tribe of _Sarmatians_, and meaning "renowned," "ill.u.s.trious[1296]." Both their language and mythologies, continues Lefevre, point to the vast region near Irania as the primeval home of the Slav, as of the Celtic and Germanic populations. The Sauromatae or Sarmatae of Herodotus[1297], who had given their name to the ma.s.s of Slav or Slavonised peoples, still dwelt north of the Caucasus and south of the _Budini_ between the Caspian, the Don and Sea of Azov; "after crossing the Tanais (Don) we are no longer in Scythia; we begin to enter the lands of the Sauromatae, who, starting from the angle of the Palus Moeotis (Sea of Azov), occupy a s.p.a.ce of 15 days' march, where are neither trees, fruit-trees, nor savages. Above the tract fallen to them the Budini occupy another district, which is overgrown with all kinds of trees[1298]." Then Herodotus seems to identify these Sarmatians with the Scythians, whence all the subsequent doubts and confusion. Both spoke the same language, of which seven distinct dialects are mentioned, yet a number of personal names preserved by the Greeks have a certain Iranic look, so that these Scythian tongues seem to have been really Aryan, forming a transition between the Asiatic and the European branches of the family.

The probable explanation is that the Scythians[1299] were a horde which came down from Upper Asia, conquered an Iranian-speaking people, and in time adopted the speech of its subjects. E. H. Minns[1300] suggests that the settled Scythians represent the remains of the Iranian population, and the nomads the conquering peoples. These were displaced later by the Sarmatians, and Scythia becomes merely a geographical term. Skulls dug up in Scythic graves throw no light on racial affinities, some being long, and some short, but in customs there is a close a.n.a.logy with the Mongols, though, as Minns points out, "the natural conditions of steppe-ranging dictated the greater part of them."

Both Slav and Germanic tribes had probably in remote times penetrated up the Danube and the Volga, while some of the former under the name of _Wends_ (Venedi[1301]), appear to have reached the Carpathians and the Baltic sh.o.r.es down the Vistula. The movement was continued far into medieval times, when great overlappings took place, and when numerous Slav tribes, some still known as Wends, others as _Sorbs_, _Croats_, or _Chekhs_, ranged over Central Europe to Pomerania and beyond the Upper Elbe to Suabia. Most of these have long been Teutonised, but a few of the _Polabs_[1302] survive as Wends in Prussian and Saxon Lausatz, while the Chekhs and _Slovaks_ still hold their ground in Bohemia and Moravia, as the _Poles_ do in Posen and the Vistula valley, and the _Rusniaks_ or _Ruthenes_ with the closely allied "Little Russians," in the Carpathians, Galicia, and Ukrania.

It was from the Carpathian[1303] lands that came those _Yugo-Slavs_ ("Southern Slavs") who, under the collective name of Sorbs (Serbs, Servians), moved southwards beyond the Danube, and overran a great part of the Balkan peninsula and nearly the whole of Greece in the 6th and 7th centuries. They were the Khorvats[1304] or Khrobats[1304] from the upland valleys of the Oder and Vistula, whom, after his Persian wars, Heraclius invited to settle in the wasted provinces south of the Danube, hoping, as Nadir Shah did later with the Kurds in Khorasan, to make them a northern bulwark of the empire against the incursions of the Avars and other Mongolo-Turki hordes. Thus was formed the first permanent settlement of the Yugo-Slavs in Croatia, Istria, Dalmatia, Bosnia, and the Nerenta valley in 680, under the five brothers Klukas, Lobol, Kosentses, Mukl, and Khrobat, with their sisters Tuga and Buga.

These were followed by the kindred Srp (Sorb) tribes from the Elbe, who left their homes in Misnia and Lusatia, and received as their patrimony the whole region between Macedonia and Epirus, Dardania, Upper Moesia, the Dacia of Aurelian, and Illyria, _i.e._ Bosnia and Servia. The lower Danube was at the same time occupied by the _Severenses_, "Seven Nations," also Slavs, who reached to the foot of the Hemus beyond the present Varna. Nothing could stem this great Slav inundation, which soon overflowed into Macedonia (Rumelia), Thessaly, and Peloponnesus, so that for a time nearly the whole of the Balkan lands, from the Danube to the Mediterranean, became a Slav domain--parts of Illyria and Epirus (Albania) with the Greek districts about Constantinople alone excepted.

h.e.l.las, as above seen, has recovered itself, and the _Albanians_[1305], direct descendants of the ancient Illyrians, still hold their ground and keep alive the last echoes of the old Illyrian language, which was almost certainly a proto-Aryan form of speech probably intermediate, as above-mentioned, between the Italic and h.e.l.lenic branches. They even retain the old tribal system, so that there are not only two main sections, the northern _Ghegs_ and the southern _Toshks_, but each section is divided into a number of minor groups[1306], such as the Malliesors (Klementi, Pulati, Hoti, etc.) and Mirdites (Dibri, Fandi, Matia, etc.) in the north, and the Toxides (whence Toshk) and the Yapides (Lapides) in the south. The southerners are mainly Orthodox Greeks, and in other respects half-h.e.l.lenised Epirotes, the northerners partly Moslem and partly Roman Catholics of the Latin rite. From this section came chiefly those Albanians who, after the death (1467) of their valiant champion, George Castriota (_Scanderbeg_, "Alexander the Great"), fled from Turkish oppression and formed numerous settlements, especially in Calabria and Sicily, and still retain their national traditions.

In their original homes, located by some between the Bug and the Dnieper, the Slavs have not only recovered from the fierce Mongolo-Turki and Finn tornadoes, by which the eastern steppes were repeatedly swept for over 1500 years after the building of the Great Wall, but have in recent historic times displayed a prodigious power of expansion second only to that of the British peoples. The _Russians_ (Great, Little, and White Russians), whose political empire now stretches continuously from the Baltic to the Pacific, have already absorbed nearly all the Mongol elements in East Europe, have founded compact settlements in Caucasia and West Siberia, and have thrown off numerous pioneer groups of colonists along all the highways of trade and migration, and down the great fluvial arteries between the Ob and the Amur estuary. They number collectively over 100 millions, with a domain of some nine million square miles. The majority belong to Deniker's Eastern race[1307] (a variety of the Alpine type), being blond, sub-brachycephalic and short, 1.64 m. (5 ft. 4-1/2 ins.). The Little Russians in the South on the Black Mould belt are more brachycephalic and have darker colouring and taller stature. The White Russians in the West between Poland and Lithuania are the fairest of all.

We need not be detained by the controversy carried on between Sergi and Zaborowski regarding a prehistoric spread of the Mediterranean race to Russia[1308]. The skulls from several of the old Kurgans, identified by Sergi with his Mediterranean type, have not been sufficiently determined as to date or cultural periods to decide the question, while their dolicho shape is common both to the Mediterraneans and to the proto-Aryans of the North European type[1309]. To this stock the proto-Slavs are affiliated by Zaborowski and many others[1310], although the present Slavs are all distinctly round-headed. Ripley asks, almost in despair, what is to be done with the present Slav element, and decides to apply "the term _h.o.m.o Alpinus_ to this broad-headed group wherever it occurs, whether on mountains or plains, in the west or in the east[1311]."

We are beset by the same difficulties as we pa.s.s with the _Ossets_ of the Caucasus into the Iranian and Indian domains of the proto-Aryan peoples. These Ossets, who are the only aborigines of Aryan speech in Caucasia, are by Zaborowski[1312] identified with the Alans, who are already mentioned in the 1st century A.D. and were Scythians of Iranian speech, blonds, mixed with Medes, and perhaps descendants of the Ma.s.sagetae. We know from history that the Goths and Alans became closely united, and it may be from the Goths that the Osset descendants of the Alans (some still call themselves Alans) learned to brew beer.

Elsewhere[1313] Zaborowski represents the Ossets as of European origin, till lately for the most part blonds, though now showing many Scythian traits. But they are not physically Iranians "despite the Iranian and Asiatic origin of their language," as shown by Max Kowalewsky[1314]. On the whole, therefore, the Ossets may be taken as originally blond Europeans, closely blended with Scythians, and later with the other modern Caucasus peoples, who are mostly brown brachys. But Ernest Chantre[1315] allies these groups to their brown and brachy Tatar neighbours, and denies that the Ossets are the last remnants of Germanic immigrants into Caucasia.

We have therefore in the Caucasus a very curious and puzzling phenomenon--several somewhat distinct groups of aborigines, mainly of de Lapouge's Alpine type, but all except the Ossets speaking an amazing number of non-Aryan stock languages. Philologists have been for some time hard at work in this linguistic wilderness, the "Mountain of Languages" of the early Arabo-Persian writers, without greatly reducing the number of independent groups, while many idioms traceable to a single stem still differ so profoundly from each other that they are practically so many stocks. Of the really distinct families the more important are:--the _Kartweli_ of the southern slopes, comprising the historical Georgian, cultivated since the 5th century, the Mingrelian, Imeritian, Laz of Lazistan, and many others; the _Cherkess_ (Circa.s.sian), the _Abkhasian_ and _Kabard_ of the Western and Central Caucasus; the _Chechenz_ and _Lesghian_, the _Andi_, the _Ude_, the _Kubachi_ and _Duodez_ of Daghestan, _i.e._ the Eastern Caucasus. Where did this babel of tongues come from? We know that 2500 years ago the relations were much the same as at present, because the Greeks speak of scores of languages current in the port of Dioscurias in their time. If therefore the aborigines are the "sweepings of the plains," they must have been swept up long before the historic period. Did they bring their different languages with them, or were these specialised in their new upland homes? The consideration that an open environment makes for uniformity, secluded upland valleys for diversity, seems greatly to favour the latter a.s.sumption, which is further strengthened by the now established fact that, although there are few traces of the Palaeolithic epoch, the Caucasus was somewhat thickly inhabited in the New Stone Age.

Crossing into Irania we are at once confronted with totally different conditions. For the ethnologist this region comprises, besides the tableland between the Tigris and Indus, both slopes of the Hindu-Kush, and the Pamir, with the uplands bounded south and north by the upper courses of the Oxus and the Sir-darya. Overlooking later Mongolo-Turki encroachments, a general survey will, I think, show that from the earliest times the whole of this region has formed part of the Caucasic domain; that the bulk of the indigenous populations must have belonged to the dark, round-headed Alpine type; that these, still found in compact ma.s.ses in many places, were apparently conquered, but certainly Aryanised in speech, in very remote prehistoric times by long-headed blond Aryans of the IRANIC and GALCHIC branches, who arrived in large numbers from the contiguous Eurasian steppe, mingled generally with the brachy aborigines, but also kept aloof in several districts, where they still survive with more or less modified proto-Aryan features. Thus we are at once struck by the remarkable fact that absolute uniformity of speech, always apart from late Mongol intrusions, has prevailed during the historic period throughout Irania, which has been in this respect as completely Aryanised as Europe itself; and further, that all current Aryan tongues, with perhaps one trifling exception[1316], are members either of the Iranic or the Galchic branch of the family. Both Iranic and Galchic are thus rather linguistic than ethnic terms, and so true is this that a philologist always knows what is meant by an Iranic language, while the anthropologist is unable to define or form any clear conception of an Iranian, who may be either of long-headed Nordic or round-headed Alpine type. Here confusion may be avoided by reserving the historic name of PERSIAN[1317] for the former, and comprising all the Alpines under the also time-honoured though less known name of TAJIKS.

Khanikoff has shown that these Tajiks const.i.tute the primitive element in ancient Iran. To the true Persians of the west, as well as to the kindred Afghans in the east, both of dolicho type, the term is rarely applied. But almost everywhere the sedentary and agricultural aborigines are called Tajiks, and are spoken of as _Parsivan_, that is, _Parsizaban_[1318], "of Persian speech," or else _Dihkan_[1318], that is, "Peasants," all being mainly husbandmen "of Persian race and tongue[1319]." They form endless tribal, or at least social, groups, who keep somewhat aloof from their proto-Aryan conquerors, so that, in the east especially, the ethnic fusion is far from complete, the various sections of the community being still rather juxtaposed than fused in a single nationality. When to these primeval differences is added the tribal system still surviving in full vigour amongst the intruding Afghans themselves, we see how impossible it is yet to speak of an Afghan nation, but only of heterogeneous ma.s.ses loosely held together by the paramount tribe--at present the _Durani_ of Kabul.

The Tajiks are first mentioned by Herodotus, whose _Dadikes_[1320] are identified by Hammer and Khanikoff with them[1321]. They are now commonly divided into Lowland, and Highland or Hill Tajiks, of whom the former were always Parsivan, whereas the Hill Tajiks did not originally speak Persian at all, but, as many still do, an independent sister language called Galchic, current in the Pamir, Zerafshan and Sir-darya uplands, and holding a somewhat intermediate position between the Iranic and Indic branches.

This term Galcha, although new to science, has long been applied to the Aryans of the Pamir valleys, being identified with the _Calcienses populi_ of the lay Jesuit Benedict Goez, who crossed the Pamir in 1603, and describes them as "of light hair and beard like the Belgians."

Meyendorff also calls those of Zerafshan "Eastern Persians, Galchi, Galchas." The word has been explained to mean "the hungry raven who has withdrawn to the mountains," probably in reference to those Lowland Tajiks who took refuge in the uplands from the predatory Turki hordes.

But it is no doubt the Persian _galcha_, a peasant or clown, then a vagabond, etc., whence _galchagi_, rudeness.

As shown by J. Biddulph[1322], the tribes of Galchic speech range over both slopes of the Hindu-Kush, comprising the natives of Sarakol, Wakhan, Shignan, Munjan (with the Yidoks of the Upper Lud-kho or Chitral river), Sanglich, and Ishkashim. To these he is inclined to add the Pakhpus and the Shakshus of the Upper Yarkand-darya, as well as those of the Kocha valley, with whom must now be included the Zerafshan Galchas (Maghians, Kshtuts, Falghars, Machas and Fans), but not the Yagn.o.bis.

All these form also one ethnic group of Alpine type, with whom on linguistic grounds Biddulph also includes two other groups, the Khos of Chitral with the Siah Posh of Kafiristan, and the Shins (Dards), Gors, Chilasi and other small tribes of the Upper Indus and side valleys, all these apparently being long-heads of the blond Aryan type. Keeping this distinction in view, Biddulph's valuable treatise on the Hindu-Kush populations may be followed with safety. He traces the Galcha idioms generally to the old Baktrian (East Persia, so-called "Zend Avesta"), the Shin however leaning closely to Sanskrit, while Khowar, the speech of the Chitrali (Khos), is intermediate between Baktrian and Sanskrit.

But differences prevail on these details, which will give occupation to philologists for some time to come.

Speaking generally, all the Galchas of the northern slopes (most of Biddulph's first group) are physically connected with all the other Lowland and Hill Tajiks, with whom should also probably be included Elphinstone's[1323] southern Tajiks dwelling south of the Hindu-Kush (Kohistani, Berraki, Purmuli or Fermuli, Sirdehi, Sistani, and others scattered over Afghanistan and northern Baluchistan). Their type is p.r.o.nouncedly Alpine, so much so that they have been spoken of by French anthropologists as "those belated Savoyards of Kohistan[1324]." De Ujfalvy, who has studied them carefully, describes them as tall, brown or bronzed and even white, with ruddy cheeks recalling the Englishman, black or chestnut hair, sometimes red and even light, smooth, wavy or curly, full beard, brown, ruddy or blond (he met two brothers near Penjakend with hair "blanc comme du lin"); brown, blue, or grey eyes, never oblique, long, shapely nose slightly curved, thin, straight lips, oval face, stout, vigorous frame, and round heads with cephalic index as high as 86.50. This description, which is confirmed by Bonvalot and other recent observers, applies to the Darwazi, Wakhi, Badakhshi, and in fact all the groups, so that we have beyond all doubt an eastern extension of the Alpine brachycephalic zone through Armenia and the Bakhtiari uplands to the Central Asiatic highlands, a conclusion confirmed by the explorations of M. A. Stein in Chinese Turkestan and the Pamirs (1900-8)[1325]. Indeed this Asiatic extension of the Alpine type inclines v. Luschan[1326] to regard the European branch as one offshoot, and the high and narrow ("Hitt.i.te") nosed type as another, or rather as a specialised group, of which the Armenians, Persians, Druses, and other sectarian groups of Syria and Asia Minor represent the purest examples. According to his summary of this complicated region "All Western Asia was originally inhabited by a h.o.m.ogeneous melanochroic race, with extreme hypsi-brachycephaly and with a 'Hitt.i.te' nose. About 4000 B.C. began a Semitic invasion from the south-east, probably from Arabia, by people looking like the modern Bedawy. Two thousand years later commenced a second invasion, this time from the north-west, by xanthrochroous and long-headed tribes like the modern Kurds, half savage, and in some way or other, perhaps, connected with the historic Harri, Amorites, Tamehu and Galatians[1327]."

But the eventful drama is not yet closed. Arrested perhaps for a time by the barrier of the Hindu-Kush and Suliman ranges, proto-Aryan conquerors burst at last, probably through the Kabul river gorges, on to the plains of India, and thereby added another world to the Caucasic domain. Here they were brought face to face with new conditions, which gave rise to fresh changes and adaptations resulting in the present ethnical relations in the peninsula. There is good reason to think that in this region the leavening Aryan element never was numerous, while even on their first arrival the Aryan invaders found the land already somewhat thickly peopled by the aborigines[1328].

The marked linguistic and ethnical differences between Eastern and Western Hindustan have given rise to the theory of two separate streams of immigration, perhaps continued over many centuries[1329]. The earlier entered from the north-west, bringing their herds and families with them, whose descendants are the h.o.m.ogeneous and handsome populations of the Punjab and Rajputana. Later swarms entered by way of the difficult pa.s.ses of Gilgit and Chitral, a route which made it impossible for their women to accompany them. "Here they came in contact with the Dravidians; here by the stress of that contact caste was evolved; here the Vedas were composed and the whole fantastic structure of orthodox ritual and usage was built up.... The men of the stronger race took to themselves women of the weaker, and from these unions was evolved the mixed type which we find in Hindustan and Bihar[1330]."

An attempt to a.n.a.lyse the complicated ethnic elements contained in the vast area of India was made by H. H. Risley[1331], who recognised seven types, his cla.s.sification being based on theories of origin.

1. The TURKO-IRANIAN type, including the _Baloch_, _Brahui_, and _Afghans_ of Baluchistan and the North-West Frontier Provinces, all Muhammadans, with broad head, long prominent nose, abundant hair, fair complexion and tall stature.

2. INDO-ARYAN type in the Punjab, Rajputana and Kashmir, with its most conspicuous members the _Rajputs_, _Khatri_ and _Jats_ in all but colour closely resembling the European type and showing little difference between upper and lower social strata. Their characteristics are tall stature, fair complexion, plentiful hair on face, long head, and narrow prominent nose.

3. ARYO-DRAVIDIAN or Hindustani type in the United Provinces, parts of Rajputana, Bihar, and Ceylon, with lower stature, variable complexion, longish head, and a nose index exactly corresponding to social station.

4. SCYTHO-DRAVIDIAN of Western India, including the _Maratha Brahmans_, _Kunbi_, and _Coorgs_, of medium stature, fair complexion, broad head with scanty hair on the face, and a fine nose.

5. DRAVIDIAN, generally regarded as representing the indigenous element.

The characteristics are fairly uniform from Ceylon to the Ganges valley throughout Madras, Hyderabad, the Central Provinces, Central India and Chota Nagpur, and the name is now used to include the ma.s.s of the population unaffected by foreign (Aryan, Scythian, Mongoloid) immigration. The _Nairs_ of Malabar and the _Santal_ of Chota Nagpur are typical representatives. The stature is short, complexion very dark, almost black, hair plentiful with a tendency to curl, head long and nose very broad[1332].

6. MONGOLO-DRAVIDIAN or Bengali type of Bengal and Orissa, showing fusion with Tibeto-Burman elements. The stature is medium, complexion dark, and head conspicuously broad, nose variable.

7. MONGOLOID of the Himalayas, Nepal, a.s.sam, and Burma, represented by the _Kanet_ of Lahoul and Kulu, the _Lepcha_ of Darjiling, the _Limbu_, _Murmi_ and _Gurung_ of Nepal, the _Bodo_ of a.s.sam and the _Burmese_.

The stature is short, the complexion dark with a yellowish tinge, the hair on the face scanty. The head is broad with characteristic flat face and frequently oblique eyes.

This cla.s.sification while more or less generally adopted in outline is not allowed to pa.s.s unchallenged, especially with regard to the theories of origin implied. Concerning the brachycephalic element of Western India Risley's belief that it was the result of so-called "Scythian"

invasions is not supported by sufficient evidence. "The foreign element is certainly Alpine, not Mongolian, and it may be due to a migration of which the history has not been written[1333]." Ramaprasad Chanda[1334]

goes further and traces the broad-headed elements in both "Scytho-Dravidians" (Gujaratis, Marathas and Coorgs) and "Mongolo-Dravidians" (Bengalis and Oriyas) to one common source, "the _h.o.m.o alpinus_ of the Pamirs and Chinese Turkestan," and attempts to reconstruct the history of the migration of the Alpine invaders from Central Asia over Gujarat, Deccan, Bihar and Bengal. His conclusions are supported by the reports of Sir Aurel Stein of the _h.o.m.o Alpinus_ type discovered in the region of Lob Nor, dating from the first centuries A.D. This type "still supplies the prevalent element in the racial const.i.tution of the indigenous population of Chinese Turkestan, and is seen in its purest form in the Iranian-speaking tribes near the Pamirs[1335]."

But any scheme of cla.s.sification must be merely tentative, subject to modification as statistics of the vast area are gradually collected. And W. Crooke[1336], while acknowledging the value of Risley's scheme[1337]

points out the need of caution in accepting measurements of skull and nose forms applied to the mixed races and half-breeds which form the majority of the people. "The race migrations are all prehistoric, and the amalgamation of the races has continued for ages among a people to whom moral restraints are irksome and unfamiliar. The existing castes are quite a modern creation, dating only from the later Buddhist age."

"The present population thus represents the flotsam and jetsam collected from many streams of ethnical movement, and any attempt to sort out the existing races into a set of pigeon-holes, each representing a defined type of race, is, in the present state of our knowledge, impossible[1338]."

In features, says Dalton, the Kols[1339] show "much variety, and I think in a great many families there is a considerable admixture of Aryan blood. Many have high noses and oval faces, and young girls are at times met with who have delicate and regular features, finely-chiselled straight noses, and perfectly formed mouths and chins. The eyes, however, are seldom so large, so bright, and gazelle-like as those of pure Hindu maidens, and I have met strongly marked Mongolian features.

In colour they vary greatly, the copper tints being about the most common [though the Mirzapur Kols are very dark]. Eyes dark brown, hair black, straight or wavy [as all over India]. Both men and women are noticeable for their fine, erect carriage and long, free stride[1340]."

The same variations are found among the Dravidians, where, as should be expected, there are many aberrant groups showing divergences in all directions, as amongst the _Kurumba_ and _Toda_ of the Nilgiris, the former approximating to the Mongol, the latter to the Aryan standard. W.

Sikemeier, who lived amongst them for years, notes that "many of the Kurumbas have decided Mongoloid face and stature, and appear to be the aborigines of that region[1341]." The same correspondent adds that much nonsense has been written about the Todas, who have become the trump card of popular ethnographists. "Being ransacked by European visitors they invent all kinds of traditions, which they found out their questioners liked to get, and for which they were paid." Still the type is remarkable and strikingly European, "well proportioned and stalwart, with straight nose, regular features and perfect teeth," the chief characteristic being the development of the hairy system, less however than amongst the Ainu, whom they so closely resemble[1342]. From the ill.u.s.trations given in Thurston's valuable series one might be tempted to infer that a group of proto-Aryans had reached this extreme limit of their Asiatic domain, and although W. H. R. Rivers has cleared away the mystery and established links between the Todas and tribes of Malabar and Travancore, the problem of their origin is not yet entirely solved[1343].

The Dravidians occupy the greater part of the Deccan, where they are const.i.tuted in a few great nations--Telugus (Telingas), Tamils (numbers of whom have crossed into Ceylon and occupied the northern and central parts of that island, working in the coffee districts), Kanarese, and the Malayalim of the west coast. These with some others were brought at an early date under Aryan (Hindu) influences, but have preserved their highly agglutinating Dravidian speech, which has no known affinities elsewhere, unless perhaps with the language of the Brahuis, who are regarded by many as belated Dravidians left behind in East Baluchistan.

But for this very old, but highly cultivated Dravidian language, which is still spoken by about 54 millions between the Ganges and Ceylon, it would no longer be possible to distinguish these southern Hindus from those of Aryan speech who occupy all the rest of the peninsula together with the southern slopes of the Hindu-Kush and parts of the western Himalayas. Their main divisions are the Kashmiri, many of whom might be called typical Aryans; the Punjabis with several sub-groups, amongst which are the Sikhs, religious sectaries half Moslem half Hindu, also of magnificent physique; the Gujaratis, Mahratis, Hindis, Bengalis, a.s.samis, and Oraons of Orissa, all speaking Neo-Sanskritic idioms, which collectively const.i.tute the Indic branch of the Aryan family. Hindustani or Urdu, a simplified form of Hindi current especially in the Doab, or "Two waters," the region between the Ganges and Jumna above Allahabad, has become a sort of _lingua franca_, the chief medium of intercourse throughout the peninsula, and is understood by certainly over 100 millions, while all the population of Neo-Sanskritic speech numbered in 1898 considerably over 200 millions.

Cla.s.sification derives little help from the consideration of caste, whatever view be taken of the origin of this inst.i.tution. The rather obvious theory that it was introduced by the handful of Aryan conquerors to prevent the submergence of the race in the great ocean of black or dark aborigines, is now rejected by many investigators, who hold that its origin is occupational, a question rather of social or industrial pursuits becoming hereditary in family groups than of race distinctions sanctioned by religion. They point out that the commentator's interpretation of the _Pancha Ksitaya_, "Five Cla.s.ses," as _Brahmans_ (priests), _Kshatriyas_ (fighters), _Vaisya_ (traders), _Sudra_ (peasants and craftsmen of all kinds), and _Nishada_ (savages or outcasts) is recent, and conveys only the current sentiment of the age.

It never had any substantial base, and even in the comparatively late Inst.i.tutes of Manu "the rules of food, connubium and intercourse between the various castes are very different from what we find at present"; also that, far from being eternal and changeless, caste has been subject to endless modifications throughout the whole range of Hindu myth and history. Nor is it an inst.i.tution peculiar to India, while even here the stereotyped four or five divisions neither accord with existing facts, nor correspond to so many distinct ethnical groups.

All this is perfectly true, and it is also true that for generations the recognised castes, say, social pursuits, have been in a state of constant flux, incessantly undergoing processes of segmentation, so that their number is at present past counting. Nevertheless, the system may have been, and probably was, first inspired by racial motives, an instinctive sense of self-preservation, which expressed itself in an informal way by local cla.s.s distinctions which were afterwards sanctioned by religion, but eventually broke down or degenerated into the present relations under the outward pressure of imperious social necessities[1344].

Beyond the mainland and Ceylon no Caucasic peoples of Aryan speech are known to have ranged in neolithic or prehistoric times. But we have already followed the migrations of a kindred[1345], though mixed race, here called INDONESIANS, into Malaysia, the Philippines, Formosa, and the j.a.panese Archipelago, which they must have occupied in the New Stone Age. Here there occurs a great break, for they are not again met till we reach Micronesia and the still more remote insular groups beyond Melanesia. In Micronesia the relations are extremely confused, because, as it seems, this group had already been occupied by the Papuans from New Guinea before the arrival of the Indonesians, while after their arrival they were followed at intervals by Malays perhaps from the Philippines and Formosa, and still later by j.a.panese, if not also by Chinese from the mainland. Hence the types are here as varied as the colour, which appears, going eastwards, to shade off from the dark brown of the Pelew and Caroline Islanders to the light brown of the Marshall and Gilbert groups, where we already touch upon the skirts of the true Indonesian domain[1346].

A line drawn athwart the Pacific from New Zealand through Fiji to Hawaii will roughly cut off this domain from the rest of the Oceanic world, where all to the west is Melanesian, Papuan or mixed, while all to the right--_Maori_, some of the eastern _Fijians_, _Tongans_, _Samoans_, _Tahitians_, _Marquesans_, _Hawaiians_ and _Easter Islanders_--is grouped under the name POLYNESIAN, a type produced by a mixture of Proto-Malayan and Indonesian. Dolichocephaly and mesaticephaly prevail throughout the region, but there are brachycephalic centres in Tonga, the Marquesas and Hawaiian Islands. The hair is mostly black and straight, but also wavy, though never frizzly or even kinky. The colour also is of a light brown compared to cinnamon or cafe-au-lait, and sometimes approaching an almost white shade, while the tall stature averages 1.72 m. (5 ft. 7-3/4 ins.).

Migrating at an unknown date eastwards from the East Indian archipelago[1347], the first permanent settlements appear to have been formed in Samoa, and more particularly in the island of _Savaii_, originally _Savaiki_, which name under divers forms and still more divers meanings accompanied all their subsequent migrations over the Pacific waters. Thus we have in Tahiti _Havaii_[1348], the "universe,"

and the old capital of Raiatea; in Rarotonga _Avaiki_, "the land under the wind"; in New Zealand _Hawaiki_, "the land whence came the Maori"; in the Marquesas _Havaiki_, "the lower regions of the dead," as in _to fenua Havaiki_, "return to the land of thy forefathers," the words with which the victims in human sacrifices were speeded to the other world; lastly in _Hawaii_, the name of the chief island of the Sandwich group.

The Polynesians are cheerful, dignified, polite, imaginative and intelligent, varying in temperament between the wild and energetic and politically capable Maori to his indolent and politically sterile kinsmen to the north, who have been unnerved by the unvarying uniformity of temperature. Wherever possible, they are agriculturalists, growing yams, sweet potatoes and taro. Coconuts, bread-fruit and bananas form the staple food in many islands. Scantily endowed with fertile soil and edible plants the Polynesians have gained command over the sea which everywhere surrounds them, and have developed into the best seamen among primitive races. Large sailing double canoes were formerly in use, and single canoes with an outrigger are still made. Native costume for men is made of bark cloth, and for women ample petticoats of split and plaited leaves. Ornaments, with the exception of flowers, are sparingly worn. The bow and arrow are unknown, short spears, clubs and slings are used, but no shields. The arts of writing, pottery making, loom-weaving and the use of metals were, with few exceptions, unknown, but mat-making, basketry and the making of _tapa_ were carried to a high pitch, and Polynesian bark-cloth is the finest in the world.

Throughout Polynesia the community is divided into n.o.bles or chiefs, freemen and slaves, which divisions are, by reason of _tabu_, as sharp as those of caste. They fall into those which partic.i.p.ate in the divine, and those which are wholly excluded from it. Women have a high position, and men do their fair share of work. Polygyny is universal, being limited only by the wealth of the husband, or the numerical preponderance of the men. Priests have considerable influence, there are numerous G.o.ds, sometimes worshipped in the outward form of idols, and ancestors are deified.