Chapter VI (1)

The Monosyllabic Area--the Tʻhay--the Môn and Khô--Tables--the Bʻhot--
the Chinese--Burmese--Persia--India--Tamulian family--the Brahúi--
the Dioscurians--the Georgians--Irôn--Mizjeji--Lesgians--Armenians--
Asia Minor--Lycians--Carians--Paropamisans--Conclusion.

Our plan is now to take up the different lines of migration at the points where they were respectively broken off. This was at their different points of contact with Asia. The first line was--

I. _The American._--In affiliating the American with the Asiatic, the ethnologist is in the position of an irrigator, who supplies some wide tract of thirsty land with water derived from a higher level, but kept from the parts below by artificial embankments. These he removes; his process being simple but effectual, and wholly independent of the clever machinery of pumps, water-wheels, and similar branches of hydraulics. The obstacle being taken away, gravitation does the rest.

The over-valuation of the Eskimo peculiarities is the great obstacle in American ethnology. When these are cut down to their due level, the connexion between America and Asia is neither more nor less than one of the clearest we have. It is certainly clearer than the junction of Africa and north-western Asia; not more obscure than that between Oceanica and the Transgangetic Peninsula; and incalculably less mysterious than that which joins Asia to Europe.

Indeed, there is no very great break, either philologically or anatomically, until we reach the confines of China. Here, the physical conformation keeps much the same: the language, however, becomes _monosyllabic_.

Now many able writers lay so much stress upon this monosyllabic character, as to believe that the separation between the tongues so constituted and those wherein we have an increase of syllables with a due amount of inflexion besides, is too broad to be got over. If speech were a mineral, this might, perhaps, be true. But speech _grows_, and if one philological fact be more capable of proof than another, it is that of a monosyllabic and uninflected tongue being a polysyllabic and inflected one in its first stage of development--or rather in its _non_-development.

The Kamskadale, the Koriak, the Aino-Japanese, and the Korean are the Asiatic languages most like those of America. Unhesitatingly as I make this assertion--an assertion for which I have numerous tabulated vocabularies as proof--I am by no means prepared to say that one-tenth part of the necessary work has been done for the parts in question; indeed, it is my impression that it is easier to connect America with the Kurile Isles and Japan, &c., than it is to make Japan and the Kurile Isles, &c., Asiatic. The group which they form belongs to an area where the displacements have been very great. The Kamskadale family is nearly extinct. The Koreans, who probably occupied a great part of Mantshuria, have been encroached on by both the Chinese and the Mantshús. The same has been the case with the Ainos of the lower Amúr. Lastly, the whole of the northern half of China was originally in the occupancy of tribes who were probably intermediate to their Chinese conquerors, the Mantshús and the Koreans.

That the philological affinities necessary for making out the Asiatic origin of the Americans lie anywhere but on the surface of the language, I confess. Of the way whereby they should be looked for, the following is an instance.

The _Yukahiri_ is an Asiatic language of the Kolyma and Indijirka. Compare its numerals with those of the other tribes in the direction of America. They differ. They are not Koriak, not Kamskadale, by no means Eskimo; nor yet Kolúch. Before we find the name of a single Yukahiri unit reappearing in other languages, we must go as far south along the western coast of America as the parts about Vancouver’s Island. There we find the Hailtsa tongue--where _malúk_ = _two_. Now the Yukahiri term for _two_ is not _malúk_. It is a word which I do not remember. Nevertheless, _malúk_ = _two_ does exist in the Yukahiri. The word for _eight_ is _malúk_ × the term for four (2 × 4).

This phænomenon would be repeated in English if our numerals ran thus:--1. _one_; 2. _pair_; 4. _four_; 8. _two-fours_; in which case all arguments based upon the correspondence or non-correspondence of the English numerals with those of Germany and Scandinavia would be as valid as if the word _two_ were the actual name of the second unit. Indeed, in one respect they would be more so. The peculiar way in which the Hailtsa _malúk_ reappears in the Yukahiri is conclusive against the name being _borrowed_. Whether it is _accidental_ is quite another question. This depends upon the extent to which it is a single coincidence, or one out of many. All that is attempted, at present, is to illustrate the extent to which resemblances may be disguised, and the consequent care requisite for detecting them[27].

II. _The connexion between Oceanica and South-eastern Asia._--The physical conformation of the Malays is so truly that of the Indo-Chinese, that no difficulties lie in this department. The philological ones are a shade graver. They involve the doubt already suggested in respect to the relations between a monosyllabic tongue like the Siamese, and a tongue other than monosyllabic like the Malay.

This brings us to the great area of the monosyllabic tongues itself. _Geographically_, it means China, Tibet, the Transgangetic Peninsula, and the Sub-Himalayan parts of northern India, such as Nepal, Sikkim, Assam, the Garo country, and other similar localities.

_Politically_, it means the Chinese, Nepalese, Burmese and Siamese empires, along with several British-Indian and independent tribes.

The chief _religion_ is Buddhism; the physical conformation unequivocally _Mongolian_.

The transition from _mono_-syllabic to _poly_-syllabic has never created much difficulty with myself: nor do I think it will do so with any writer who considers the greater difficulties involved in the denial of it. What these are will become apparent when we look at the map of Asia, and observe the tongues which come in contact with those of the class in question. Then it will become clear that _unless we allow it to form a connecting link, it not only stands alone itself, but isolates other families_. Thus, it is only through the Transgangetic Peninsula that the _Oceanic_ family can be connected with the _Indian_; a connexion which rests on grounds sufficiently good to have induced careful writers[28] to believe the affiliation to be _direct_ and _immediate_. It is only through this same Transgangetic Peninsula _plus_ Tibet and China that the great Siberian families--Turanian and Japanese--can be similarly connected with the Oceanic. Yet such a connexion really exists, though, from its indirect character, it is but partially recognised. Nevertheless, it _is_ recognised (often, perhaps, unconsciously) by every inquirer who hesitates about separating the Malay from the Mongol.

A difficulty of far greater magnitude arises from the following considerations:--There are two principles upon which languages may be classified. According to the first, we take two or more languages as we find them, ascertain certain of their characteristics, and then inquire how far these characteristics coincide. Two or more languages, thus taken, may agree in having a large per-centage of grammatical inflexions, in which case they would agree in certain _positive_ characters. On the other hand, two or more languages may agree in the _negative_ fact of having a small and scanty vocabulary, and an inflexional system equally limited.

The complication here suggested lies in a fact of which a little reflection will show the truth, viz. that _negative points of similarity prove nothing in the way of ethnological connexion_; whence, as far as the simplicity of their respective grammars is concerned, the Siamese, Burmese, Chinese and Tibetan may be as little related to each other, or to a common mother-tongue, as the most unlike languages of the whole world of Speech.

Again--it by no means follows that because all the tongues of the family in question are comparatively destitute of inflexion, they are all in the same class. A characteristic of the kind may arise from two reasons; _non_-development, or loss. There is a stage _anterior_ to the evolution of inflexions, when each word has but one form, and when relation is expressed by mere juxtaposition, with or without the superaddition of a change of accent. The tendencies of this stage are to combine words in the way of composition, but not to go further. Every word retains, throughout, its separate substantive character, and has a meaning independent of its juxtaposition with the words with which it combines.

But there is also a stage _subsequent_ to such an evolution, when inflexions have become obliterated and when case-endings, like the _i_ in _patr-i_, are replaced by prepositions (in some cases by postpositions) like the _to_ in _to father_; and when personal endings, like the _o_ in _voc-o_, are replaced by pronouns, like the _I_ in _I call_. Of the _first_ of these stages, the Chinese is the language which affords the most typical specimen that can be found in the present _late_ date of languages--_late_, considering that we are looking for a sample of its earliest forms. Of the _last_ of these stages the English of the year 1851 affords the most typical specimen that can be found in the present _early_ date of language--early, considering that we are looking for a sample of its latest forms.

Hence--

_a._ How far the different monosyllabic tongues are _all_ in the same stage--is one question.

_b._ Whether this stage be the _earlier_ or the _later_ one--is another; and--

_c._ Whether they are connected by _relationship_ as well as in _external form_--is a third.

In answer to this, it is safe to say (a.) that they are _all_ uninflected, because inflexions have yet to be evolved; not because they have been evolved and lost--as is the case with the English, a language which stands at one end of the scale, just as the Chinese does at the other.

(b.) They are, also, all connected by a _bonâ fide_ ethnological relationship; as can be shown by numerous tables; the Chinese and Tibetans being, apparently, the two extremes, in the way of difference.

As for their geographical distribution, it is a blank-and-prize lottery, with large and small areas in juxtaposition and contrast, just as has been the case in America and in Africa; the Sub-Himalayan parts of British India, Sikkim and Nepâl, and the Indo-Burmese frontier (or the country about Assam and Munipúr) being the tracts where the multiplicity of mutually unintelligible tongues within a limited district is greatest.

Again--whenever the latter distribution occurs we have either a mountain-fastness, political independence, or the primitive pagan creed--generally all three.

The population speaking a monosyllabic language which is in the most immediate contact with the continental tribes of the Oceanic stock, is the Southern Siamese. This reaches as far as the northern frontier of Kedah (Quedah), about 8° N. L. Everything north of this is monosyllabic; with the exception of a Malay settlement (probably, though not certainly, of recent origin) on the coast of Kambogia.

Now the great stock to which the Siamese belong is called Tʻhay. Its direction is from north to south, coinciding with the course of the great river Menam; beyond the head-waters of which the Tʻhay tribes reach as far as Assam. Of these northern Tʻhay, the _Khamti_ are the most numerous; and it is important to know that as many as 92 words out of 100 are common to this dialect and to the classical Siamese of Bankok.

Again, the intermediate tribes of the Upper and Middle Menam--the Lau--speak a language as unequivocally Siamese as the Khamti. If so, the Tʻhay tongue, widely extended as it is in the particular direction from north to south, is a tongue falling into but few dialects; the inference from which is, that it has spread within a comparatively recent period. Consequently, it has encroached upon certain other populations and effected certain displacements.

I think that even in the minuter details that now suggest themselves we can see our way; so far, at least, as to determine in which direction the movement took place--whether it were from north to south or from south to north.

Few classes of tongues can be better studied for ethnological purposes than the monosyllabic. A paper of Buchanan’s, and another of Leyden’s, are amongst the most valuable articles of the Asiatic Researches. One of Mr. Brown’s in the Journal of the Asiatic Society of Bengal gives us numerous tabulated vocabularies for the Burmese, Assamese and Indian frontiers. Mr. Hodgson and Dr. Robertson have done still more for the same parts. Lastly, the chief southern dialects, which have been less studied, are tabulated in the second volume of ‘Crawfurd’s Embassy to Siam.’

Upon looking over these, we find specimens of the two tongues which lie east and west of the southern Siamese; the first being the _Khô_ language of Kambogia, and the second the _Môn_ of Pegu. Each of these is spoken over a small area; indeed the Môn, which is, at present, nearly limited to the Delta of the Irawaddi, is fast giving way before the encroaching dialects of the Burmese class, whilst the Khô of Kambogia is similarly limited to the lower part of the Mekhong, and is hemmed in by the Siamese, the Lau, and the Anamitic of Cochin China.

Now, separated as they are, the Môn and Khô are liker to each other than either is to the interjacent Siamese; the inference from this being that at one time they were connected by transitional and intermediate dialects, aboriginal to the lower Menam, but now displaced by the Siamese of Bankok introduced from the parts to the northwards.

If this be the case, the monosyllabic tongue most closely allied to those of the Malayan Peninsula (which are _not_ monosyllabic) is not the present Siamese, but the language which the present Siamese displaced.

How far this view is confirmed by any special affinities between the Malay dialects with the Môn and Khô is more than I can say. The examination, however, should be made.

The _southern_ Tʻhay dialects are not only less like the Môn and Khô than is expected from their locality, but the _northern_ ones are less like those of the Indo-Burmese frontier and Assam than the geographical contiguity prepares us to surmise; since the per-centage of words common to the Khamti and the other dialects of Munipur and Assam is only as follows[29].

Siamese. Khamti.
0 1 per cent. with the Aka.
0 1 „ „ Abor.
3 5 „ „ Mishimi.
6 8 „ „ Burmese.
8 8 „ „ Karien.
3 3 „ „ Singpho.
10 10 „ „ Jili.
1 3 „ „ Garo.
3 3 „ „ Munipúri.
1 1 „ „ Songphu.
0 0 „ „ Kapwi.
1 1 „ „ Koreng.
0 0 „ „ Maram.
0 0 „ „ Kamphung.
0 0 „ „ Luhuppa.
0 0 „ „ North Tankhul.
0 0 „ „ Central Tankhul.
0 0 „ „ South Tankhul.
0 0 „ „ Khoibu.
0 0 „ „ Maring.

This shows that their original locality is to be sought in an _eastern_ as well as in a _northern_ direction.

If the Tʻhay dialects are less like the Burmese than most other members of their class, they are more like the Bʻhot of Tibet.

_English_ boat.
Ahom _ru._
Khamti _hu._
Lau _heic._
Siamese _reng._
W. Tibetan[30] _gru._
S. Tibetan[30] _kua._

_English_ bone.
Khamti _nuk._
Lau _duk._
Siamese _ka-duk._
S. Tibetan _ruko._

_English_ crow.
Ahom _ka._
Khamti _ka._
Lau _ka._
Siamese _ka._
W. Tibetan _kha-ta._

_English_ ear.
Khamti (3) _hú._
W. Tibetan _sá._
S. Tibetan _amcho._

_English_ egg.
Ahom _khrai._
Khamti _khai._
Lau _khai._
Siamese _khai._

_English_ father.
Ahom (3) _po._
W. Tibetan _phá._
S. Tibetan _pálá._

_English_ fire.
Ahom (3) _fai._
W. Tibetan _má._
S. Tibetan _mé._

_English_ flower.
Ahom _blok._
Khamti _mok._
Lau _dok._
Siamese _dokmai._
W. Tibetan _me-tog._
S. Tibetan _men-tok._

_English_ foot.
Ahom _tin._
W. Tibetan _{r}kang-pa._
S. Tibetan _kango._

_English_ hair.
Ahom _phrum._
Khamti _phom._
Lau _phom._
Siamese _phom._
W. Tibetan _skra._
---- _spu._
S. Tibetan _ta._
---- _kra._

_English_ head.
Ahom _ru._
Khamti _ho._
Lau _ho._
Siamese _hoa._
W. Tibetan _mgo._
S. Tibetan _go._

_English_ moon.
Siamese _tawan._
W. Tibetan _{z}lava._
S. Tibetan _dawa._

_English_ mother.
Ahom (4) _me._
Tibetan _ama._

_English_ night.
Khamti (3) _khün._
W. Tibetan _m tshan-mo._
S. Tibetan _chen-mo._

_English_ oil.
Ahom _man grá._
Khamti _nam._
---- _man._
Lau (2) _nam._
---- _man._
S. Tibetan _num._

_English_ road.
Ahom (2) _táng._
Siamese _tháng._
W. Tibetan _lami._
S. Tibetan _lani._

_English_ salt.
Ahom _klu._
Khamti _ku._
Lau _keu._
---- _keou._
Siamese _kleua._

_English_ skin.
Ahom _plek._
W. Tibetan _pag-spa._
S. Tibetan _pag-pa._

_English_ tooth.
Ahom _khiu._
Khamti _khiu._
Lau _khiau._
Siamese _khiau._
Tibetan _só._

_English_ tree.
Ahom _tun._
Khamti _tun._
Lau _tón._
Siamese _tón._
W. Tibetan _l. jon-shing._
S. Tibetan _shin dong._

_English_ three.
Ahom (3) _sam._
W. Tibetan _q-sum._
S. Tibetan _sum._

_English_ four.
Ahom (3) _si._
W. Tibetan _bzhi._
S. Tibetan _zhyi._

_English_ five.
Ahom (3) _ha._
W. Tibetan _hna._
S. Tibetan _gna._

_English_ six.
Ahom _ruk._
Siamese (3) _hok._
W. Tibetan _druk._
S. Tibetan _thú._

_English_ nine.
Ahom (3) _kau._
W. Tibetan _d-gu._
S. Tibetan _guh._

_English_ in, on.
Ahom _nu._
Khamti _nau._
Lau _neu._
Tibetan _la, na._

_English_ now.
Ahom _tinai._
Khamti _tsang._
Lau _leng._
W. Tibetan _deng-tse._
S. Tibetan _thanda._

_English_ to-morrow.
Ahom _sang-manai._
Tibetan _sang._

_English_ drink.
Siamese _deum._
W. Tibetan _{p}thung._
S. Tibetan _thung._

_English_ sleep.
Ahom (2) _non._
W. Tibetan _nyan._
S. Tibetan _nyé._

_English_ laugh.
Ahom _khru._
Khamti _khó._
Lau _khóa._
Siamese _hoaro._
W. Tibetan _{b}gad._
S. Tibetan _{f}gá._

[30] S. means the _spoken_, W. the _written_ Tibetan. The collation has been made from a table of Mr. Hodgson’s in the Journal of the Asiatic Society of Bengal. The Ahom is a Tʻhay dialect.

The Bʻhot itself is spoken over a large area with but little variation. We anticipate the inference. It is an intrusive tongue, of comparatively recent diffusion. What has been its direction? From east to west rather than from west to east; at least such is the deduction from its similarity to the Tʻhay, and from the multiplicity of dialects--representatives of a receding population--in the Himalayas of Nepâl and Sikkim. This, however, is a point on which I speak with hesitation.

Dialects of the Bʻhot class are spoken as far westward as the parts about Cashmír and the watershed of the Indus and Oxus. This gives us the greatest extent eastwards of any unequivocally monosyllabic tongue.

The Chinese seem to have effected displacements as remarkable for both breadth and length as the Tʻhay were for length. We get at their original locality by the exhaustive process. On the northern and western frontier they keep encroaching at the present moment--at the expense of the Mantshús and Mongolians. For the provinces of Chansi, Pe-tche-li, Chantung, Honan, &c., indeed, for four-fifths of the whole empire, the uniformity of speech indicates a recent diffusion. In Setshuen and Yunnan the type changes probably from that of the true Chinese to the Tibetan, Tʻhay and Burmese. In Tonkin and Cochin the language is like but different--like enough to be the only monosyllabic language which is placed by any one in the same section with the Chinese, but different enough to make this position of it a matter of doubt with many. Putting all this together, the south and south-eastern provinces of China appear to be the oldest portions of the present area.

In fixing upon these as the parent provinces, the evidence of ethnology on the one side, and that of the mass of tradition and inference which passes under the honourable title of Chinese history on the other, disagree. This latter is as follows:--

At some period anterior to 550 =B.C.=, the first monarch with whom the improvement of China began, and whose name was Yao, ruled over a small portion of the present empire, viz. its _north-west_ district; and the first nations that he fought against were the Yen and Tsi, in Pe-tche-li and Shantong respectively.

Later still, Honan was conquered.

=B.C.= 550. All to the south of the Ta-keang was barbarous; and the title of King of Chinese was only _Vang_ or _prince_, not _Hoang-te_ or _Emperor_.

At this time Confucius lived. Amongst other things he wrote the _Tschan-tsen_, or Annals of his own time.

=B.C.= 213. Shi-hoang-ti, the first Emperor of all China, built the great wall, colonized Japan, conquered the parts about Nankin, and _purposely destroyed all the previously existing documents upon which he could lay hand_.

=B.C.= 94. Sse-mats-sian lived. What Shi-hoang-ti missed in the way of records, Sse-mats-sian preserved, and, as such, passes for the Herodotus China.

A destruction of the earlier records, with a subsequent reconstruction of the history which they are supposed to have embodied, is always suspicious; and when once the principle of reconstruction is admitted, no value can be attached to the intrinsic probability of a narration. It may be probable. It may be true. It cannot, however, be historical unless supported by historical testimony; since, if true, it is a guess; and if probable, a specimen of the tact of the inventor. At best, it can but be a _tradition_ or an _inference_, the basis of which may be a certain amount of fact--little or great according to the temperament of the investigator.

Now, in the previous notice of the history of Chinese civilization, we have placed its claims to a high antiquity under as favourable a point of view as is allowable. They bear the appearance of truth--so much so, that if we had reason to believe that there were any means of recording them at so early an epoch as 600 years =B.C.=, and of preserving them to so late a one as the year ’51, scepticism would be impertinent. But this is not the case. An historical fact must be taken upon evidence, not upon probabilities; and to argue the antiquity of a civilization like the Chinese from the antiquity of its history, and afterwards to claim an historical value for remote traditions on the strength of an early civilization, is to argue in a circle.

Without saying that _all_ argument upon the antiquity of the Chinese Empire is of this sort, it may fairly be said that _much_ of it has been so--so much as to make Confucius as mythological a character as Minos, and to bring the earliest reasonable records to an epoch subsequent to the introduction of Buddhism from India. Even this antiquity is only probable.

A square block of land between the Ganges and Upper Irawaddi is occupied by one dominant, and upwards of thirty subordinate sections of one and the same population--the _Burmese_. Some of these are mountaineers, and have retreated before the Indians from the south and west--encroachers upon the originally Burmese countries of Assam, Chittagong and Sylhet. Others are themselves intruders, or (what is much the same) consolidators of conquered countries. Such are the Avans of the Burmese Empire, properly so called, who seem to have followed the course of the Irawaddi, displacing not only small tribes akin to themselves, but the Môn of Pegu, as well. Lastly, the Kariens emulate the Tʻhay in the length of their area and in its north-and-south direction, being found in the southern part of the Tenasserim Provinces (in 11° N. L.) and on the very borders of China (in 23° N. L.).

No great family has its distribution so closely coincident with a water-system as the one in question. The plateau of Mongolia and the Himalayas are its boundaries. It occupies the whole[31] of all the rivers which rise within these limits, and fall into either the Bay of Bengal or the Chinese Sea; whereas (with the exception of the Himalayan portions of the Indus and the Ganges) it occupies none of the others. The lines of migration with the Indo-Chinese populations have generally followed the water-courses of the Indo-Chinese rivers; and civilization has chiefly flourished along their valleys. Yet, as these lead to an ocean interrupted by no fresh continent, the effect of their direction has been to isolate the nations who possess them. I imagine that this has much more to do with peculiarities of the Chinese civilization than aught else. Had the Hoang-ho fallen into a sea like the Mediterranean, the Celestial Empire would, probably, have given and taken in the way of social and political influence, have acted on the manners of the world at large, and have itself been reacted on. Differences should only be attributed to so indefinite and so impalpable a force as _race_ when all other things are equal.

Upon the principle of taking the questions in the order of complexity, so as to dispose of the simplest first, I pass over, for the present, the connexion between Africa and South-Western Asia, and take the easier of the two European ones.

_The Turanians._--The line which, beginning at Lapland, and, after exhibiting the great Turanian affiliations, ends at the wall of China, comprising the Ugrians, Samoeids[32], Yeniseians[32], Yukahiri[32], Turks, Mongols, and Tungusians[33], is connected with the area of the monosyllabic languages in different degrees of clearness according to the criterion employed. The physical conformation is nearly identical. The languages differ--the Turanian, like the Oceanic and the American, being inflected and polysyllabic[34]. With this difference, the complexities of the affiliation begin and end. Their amount has been already suggested.

A great part of Northern Europe, Independent Tartary, Siberia, Mongolia, Tibet, China, and the Transgangetic Peninsula, has now been disposed of. Nevertheless, India, Persia, Asia Minor, and Caucasus remain; in size inconsiderable, in difficulty great--greatly difficult because the points of contact between Europe and Asia, and Africa and Asia, fall within this area; greatly difficult because the displacements have been enormous; greatly difficult because, besides displacement, there has been intermixture as well. Lest any one undervalue the displacement, let him look at Asia Minor, which is now Turk, which has been Roman, Persian and Greek, and which has no single unequivocal remnant of its original population throughout its whole length and breadth. Yet, great as this is, it is no more than what we expect _à priori_. What families are and have been more encroaching than the populations hereabouts--Turks from the north, Arabs from the south, and Persians from the east? The oldest empires of the world lie here--and old empires imply early consolidation; early consolidation, premature displacement. Then come the phænomena of intermixture. In India there is a literary language of considerable age, and full of inflexions. Of these inflexions not one in ten can be traced in any modern tongue throughout the whole of Asia. Yet they are rife and common in many European ones. Again, the _words_ of this same language, _minus_ its inflexions, are rife and common in the very tongues where the inflexions are wanting; in some cases amounting to nine-tenths of the language. What is the inference from this? Not a very clear one at any rate.

Africa has but one point of contact with Asia, _i.e._ Arabia. It is safe to say this, because, whether we carry the migration over the Isthmus of Suez or the Straits of Babel-Mandeb, the results are similar. The Asiatic stock, in either case, is the same--Semitic. But Europe, in addition to its other mysteries, has two; perhaps three. One of these is simple enough--that of the Lap line and the Turanian stock. But the others are not so. It is easy to make the Ugrians Asiatic; but by no means easy to connect the other Europeans with the Ugrians. The Sarmatians, nearest in geography, have never been very successfully affiliated with them. Indeed, so unwilling have writers been to admit this relationship, that the Finnic hypothesis, with all its boldness, has appeared the better alternative. Yet the Finnic hypothesis is but a guess. Even if it be not so, it only embraces the Basks and Albanians; so that the so-called Indo-Europeans still stand over.

For reasons like these, the parts forthcoming will be treated with far greater detail than those which have preceded; with nothing like the detail of _minute_ ethnology, but still slowly and carefully.

All that thus stands over for investigation is separated from the area already disposed of by that line of mountains which is traced from the Garo Hills in the north-east of Bengal to the mouth of the Kuban in the Black Sea. First come the Eastern Himalayas, which, roughly speaking, may be said to divide the Indian kingdoms and dependencies from the Chinese Empire. They do not do so exactly, but they do so closely enough for the present purpose.

They may also be said, in the same way, to divide the nations of the Hindu from those of more typically Mongolian conformation.

They may also be said, in the same way, to divide the Indian tongues from the monosyllabic.

On the _north_ side of this range, languages undoubtedly, monosyllabic are spoken as far westwards as Little Tibet. On the _south_ there are Hindu characteristics both numerous and undoubted as far in the same direction as _Cashmír_.

Then comes a change. To the north and west of Cashmír is a _Kohistan_, or _mountain-country_, which will soon require being described in detail. The line, however, which we are at present engaged upon is that of the northern boundary of the Valley of the Kabúl River, the mountains between Cabul and Herat, and the continuation of the same ridge from Herat to the south-eastern corner of the Caspian. _North_ of this we have--roughly speaking--the Uzbek and Turcoman Turks; south of it, the Afghans and Persians Proper. Bokhara, however, is Persian, and the _Kohistan_ in question is _not_ Turk--whatever else it may be.

To proceed--this line runs nearly parallel to the southern shore of the Caspian. Of the provinces to the north of it, Asterabad is partly Turk and partly Persian; Mazenderan and Ghilan, Persian. From Ghilan northwards and westwards, the valleys of the Cyrus and Araxes form the chief exception--but, saving these, all is mountain and mountaineership. Indeed, it is Ararat and Armenia which lie on our left, and the vast and vague Caucasus which rears itself in front.

The simplest ethnology of the parts between this range, the Semitic area, and the sea, is that of the Persian province of Khorasan. With Persia we are so much in the habit of connecting ideas of Eastern pomp and luxury, that we are scarcely able to give it its true geographical conditions of general sterility. Yet it is really a desert with oases--a desert with oases for the far greater part of its area. And of all its provinces few are more truly so than Khorasan. Here we have a great elevated central table-land; pre-eminently destitute of rivers; and with but few towns. Of these Yezd is the chief in interest: the head-quarters of remains of the old fire-worship: Yezd the city of the Parsees, more numerous there than in all the others in Persia besides. Perhaps, too, it is the ethnological centre of the Persian stock; since in a westerly direction they extend to Kurdistan, and in a north-eastern one as far as Badukshan and Durwaz on the source of the Oxus.

The northern frontier is Turcoman, where the pastoral robbers of the parts between Bokhara and the Caspian encroach, and have encroached.

As far south as Shurukhs they are to be found; and east of Shurukhs they are succeeded by the Hazarehs--probably _wholly_, certainly _partially_, of Mongolian blood.

Abbasabad on the north-west is a Georgian colony. On the line between Meshed and Herat are several Kurd colonies. In Seistan we have Persians; but further south there are Biluch and Brahúi. Due east the Afghans come in.

Kerman is also Persian; and that to a greater degree than Khorasan. Fars is the same; yet west of Fars the population changes, and Arabian elements occur. They increase in Khuzistan; and in Irak Arabi we, at one and the same time, reach the rich alluvia of the Tigris and Euphrates and a doubtful frontier. Whether this was originally Arab or Persian is a matter of doubt.

From Irak we must subtract Laristan, and the Baktyari Mountains, as well as the whole north-western half. Hamadan is the ancient Ecbatana; the ancient Ecbatana was Median--but that the Medes and Persians were as closely allied in blood as we suppose them to have been in their unalterable laws, is by no means a safe assumption. The existence of a _third_ language in the arrow-headed inscriptions yet awaits a satisfactory explanation.

On the other hand, Mazenderan is wholly Persian; and so is Ghilan Proper. The Talish, however, to the north of that province, are, possibly, of another stock. Asterabad, as stated above, is a frontier province.

I think that there is good reason for believing Ajerbijan to have been, originally, other than Persian.

In Balkh and Bokhara, the older--but not necessarily the oldest--population appears to be Persian under recently immigrant Uzbek masters. Beyond these countries, the Persians reappear as the chief population, _i.e._ in Badukshan and Durwaz.

Here the proper Persian population ends--but not either wholly or abruptly.

Three modifications of it occur--

1. In Biluchistan to the south-east.
2. In Kurdistan to the west.
3. In Afghanistan to the east.

Besides which, there are Persians encroaching upon the Armenian and Caucasian area in Shirvan, Erivan, and Karabagh--in all of which countries, as well as in Ajerbijan, I believe it to have been intrusive.

_The Biluch._--East and south-east of the proper Persians of Kerman come the Biluch, of Biluchistan. There is certainly a change of type here. Physically, the country is much like the table-land of Kerman. India, however, is approached; so that the Biluch are frontier tribes. To a certain extent they are encroachers. We find them in Sind, in Múltan, and in the parts between the Indus and the Sulimani Mountains, and in the middle part of the Sulimani Mountains themselves. They style themselves _Usul_ or _The Pure_, a term which implies either displacement or intermixture in the parts around. Their language is a modified (many call it a _bad_) Persian. Philologically, however, it may be the older and more instructive dialect--though I have no particular reasons for thinking it so. Hindu features of physiognomy now appear. So do Semitic elements of polity and social constitution. We have tribes, clans, and families; with divisions and sub-divisions. We have a criminal law which puts us in mind of the Levites. We have classes which scorn to intermarry; and this suggests the idea of _caste_. Then we have pastoral habits as in Mongolia. The religion, however, is Mahometan, so that if any remains of the primitive Paganism, available for the purposes of ethnological classification, still exist, they lie too far below the surface to have been observed.

Captain Postans distinguishes the Biluch from the Mekrani of Mekran; but of this latter people I know no good description. They are, probably, Kerman Persians. The hill-range between Jhalawan and Sind is occupied by a family which has commanded but little notice; yet is it one of the most important in the world, the Brahúi.

_The Kurds._--A line drawn obliquely across Persia from Biluchistan towards the north-west brings us to another frontier population; a population conterminous with the Semitic Arabs of Mesopotamia, and the unplaced Armenians. These are mountaineers--the Kurds of Kurdistan. Name for name, they are the _Carduchi_ of the Anabasis. Name for name, they are the _Gordyæi_. Name for name, they are, probably, the _Chaldæi_ and _Khasd-im_--a fact which engenders a difficult complication, since the Chaldæi in the eyes of nine writers out of ten--though not in those of so good an authority as Gesenius--are Semitic. The Kurd area is pre-eminently irregular in outline. It is equally remarkable for its physical conditions. It is a range of mountains--just the place wherein we expect to find old and aboriginal populations rather than new and intrusive ones. On the other hand, however, the Kurd form of the Persian tongue is not remarkable for the multiplicity and difference of its dialects--a fact which suggests the opposite inference. Kurds extend as far south as the northern frontier of Fars, as far north as Armenia, and as far west as the head-waters of the Halys. Have they encroached? This is a difficult question. The Armenians are a people who have generally given way before intruders; but the Arabs are rather intruders than the contrary. The Kurd direction is vertical, _i.e._ narrow rather than broad, and from north to south (or _vice versâ_) rather than from east to west (or _vice versâ_), a direction common enough where it coincides with the valley of a river, but rare along a mountain-chain. Nevertheless it reappears in South America, where the Peruvian area coincides with that of the Andes.

_The Afghans._--The Afghan area is very nearly the water-system of the river Helmund. The direction in which it has become extended is east and north-east; in the former it has encroached upon Hindostan, in the latter upon the southern members of a class that may conveniently be called the Paropamisan. In this way (I think) the Valley of the Cabul River has become Afghan. Its relations to the Hazareh country are undetermined. Most of the Hazarehs are Mongolian in physiognomy. Some of them are Mongolian in both physiognomy and language. This indicates intrusion and intermixture--intrusion and intermixture which history tells us are subsequent to the time of Tamerlane. Phænomena suggestive of intrusion and intermixture are rife and common throughout Afghanistan. In some cases--as in that of Hazarehs--it is recent, or subsequent to the Afghan occupation; in others, it is ancient and prior to it.

_Bokhara._--I have not placed the division containing the Tajiks of Balkh, Kúnduz, Durwaz, Badukshan, and Bokhara, on a level with that containing the Afghans, Kurds and Biluch, because I am not sure of its value. Probably, however, it is in reality as much a separate substantive class as any of the preceding. Here the intrusion has been so great, the political relations have been so separate, and the intermixed population is so heterogeneous as for it to have been, for a long time, doubtful whether the people of Bokhara were Persian or Turk. Klaproth, however, has shown that they belong to the former division, though subject to the Uzbek Turks. If so, the present Tajiks represent the ancient Bactrians and Sogdians--the Persians of the valley and water system of the Oxus. But what if these were intruders? I have little doubt about the word _Oxus_ (_Ok-sus_) representing the same root as the _Yak_ in _Yaxsartes_ (_Yak-sartes_), and the _Yaik_, the name of the river flowing into the northern part of the Caspian. Now this is the _Turanian_ name for _river_, a name found equally in the Turk, Uguari, and Hyperborean languages. At any rate, Bokhara is on an ethnological frontier.

But Bactria and Sogdiana were Persian at the time of Alexander’s successors; they were Persian at the very beginning of the historical period. Be it so. The historical period is but a short one, and there is no reason why a population should not encroach at one time and be itself encroached upon at another.

All the parts enumerated, and all the divisions, are so undoubtedly Persian, that few competent authorities deny the fact. The most that has ever been done is to separate the Afghans. Sir W. Jones did this. He laid great stress upon certain Jewish characteristics, had his head full of the Ten Tribes, and was deceived in a vocabulary of their languages. Mr. Norris also is inclined to separate them, but on different grounds. He can neither consider the Afghan language to be Indo-European, nor the Persian to be otherwise. His inference is true, if his facts are. But what if the Persian be other than Indo-European? In that case they are both free to fall into the same category.

But the complexities of the Persian population are not complete. There is the division between the _Tajiks_ and the _Iliyats_; the former being the settled occupants of towns and villages speaking Persian, the others pastoral or wandering tribes speaking the Arab, Kurd, and Turk languages. That _Tajik_ is the same word as the root _Taoc_, in _Taoc-ene_, a part of the ancient country of Persis (now _Fars_), and, consequently, in a pre-eminent Persian locality, is a safe conjecture. The inference, however, that such was the original locality of the Persian family is traversed by numerous--but by no means insuperable--difficulties. In respect to their chronological relations, the general statement may be made, that wherever we have Tajiks and Iliyats together, the former are the older, the latter the newer population. Hence it is not in any Iliyat tribe that we are to look for any nearer approach to the aborigines than what we find in the normal population. They are the analogues of the Jews and gipsies of Great Britain rather than of the Welsh--recent grafts rather than parts of the old stock. In Afghanistan this was not so clearly the case. Indeed, the inference was the other way.

The antiquities and history of Persia are too well-known to need more than a passing allusion. The creed was that of Zoroaster; still existent, in a modified (perhaps a corrupted, perhaps an improved) form, in the religion of the modern Parsis. The language of the Zoroastrian Scriptures was called Zend. Now the Zend is Indo-European--Indo-European and highly inflected. The _inflexions_, however, in the modern Persian are next to none; and of those few it is by no means certain that they are Zend in origin. Nevertheless, the great majority of modern Persian words _are_ Zend. What does this mean? It means that the philologist is in a difficulty; that the grammatical structure points one way and the vocabulary another. This difficulty will meet us again.

_India._--In the time of Herodotus, and even earlier, India was part of the Persian empire. Yet India was not Persia. It was no more Persia in the days of Darius than it is English now. The original Indian stock was and is peculiar--peculiar in its essential fundamentals, but not pure and unmodified. The vast extent to which this modification implies encroachment and intermixture is the great key to nine-tenths of the complexities of the difficult ethnology of Hindostan. Whether we look to the juxtaposition of the different forms of Indian speech, the multiform degrees of fusion between them, the sections and sub-sections of their creeds--legion by name,--the fragments of ancient paganism, the differences of skin and feature, or the institution of caste, intrusion followed by intermixture, and intermixture in every degree and under every mode of manifestation, is the suggestion.

And now we have our duality--viz. the primitive element and the foreign one--the stock and the graft. Nothing is more certain than that the graft came from the north-west. Does this necessarily mean from Persia? Such is the current opinion; or, if not from Persia, from some of those portions of India itself nearest the Persian frontier. There are reasons, however, for refining on this view. Certain influences foreign to India may have come _through_ Persia, without being Persian. The proof that a particular characteristic was introduced into India _viâ_ Persia is one thing: the proof that it originated in Persia is another. They have often, however, been confounded.

In the south of India the foreign element is manifested less than in the north; so that it is the south of India which exhibits the original stock in its fullest form. Its chief characteristics are referable to three heads, physical form, creed, and language. In respect to the first, the southern Indian is darker than the northern--_cæteris paribus_, _i. e._ under similar external conditions; but not to the extent that a mountaineer of the Dekhan is blacker than a Bengali from the delta of the Ganges. Descent, too, or caste influences colour, and the purer the blood the lighter the skin. Then the lips are thicker, the nose less frequently aquiline, the cheek-bones more prominent, and the eyebrows less regular in the southrons. The most perfect form of the Indian face gives us regular and delicate features, arched eyebrows, an aquiline nose, an oval contour, and a clear brunette complexion. All this is Persian.

Depart from it and comparisons suggest themselves. If the lips thicken and the skin blackens, we think of the Negro; if the cheek-bones stand out and if the eye--as it sometimes does--become oblique, the Mongol comes into our thoughts.

The original Indian creeds are best characterized by negatives. They are neither Brahminic nor Buddhist.

The language, for the present, is best brought under the same description. No man living considers it to be _Indo-European_.

In proportion as any particular Indian population is characterized by these three marks, its origin, purity, and indigenous nature become clearer--and _vice versâ_. Hence, they may be taken in the order of their outward and visible signs of aboriginality.

First come--as already stated--the Southrons of the Continent[35]; and first amongst these the mountaineers. In the Eastern Ghauts we have the Chenchwars, between the Kistna and the Pennar; in the Western the Cohatars, Tudas, Curumbars, Erulars, and numerous other hill-tribes; all agreeing in being either imperfect Brahminists or Pagans, and in speaking and languages akin to the Tamul of the coast of Coromandel; a language which gives its name to the class, and introduces the important philological term _Tamulian_. The physical appearance of these is by no means so characteristic as their speech and creed. The mountain _habitats_ favour a lightness of complexion. On the other, it favours the Mongol prominence of the cheek-bones. Many, however, of the Tudas have all the regularity of the Persian countenance--yet they are the pure amongst the pure of the native Tamulian Indians.

In the _plains_ the language is Tamulian, but the creed Brahminic; a state of evidence which reaches as far north as the parts about Chicacole east, and Goa west.

In the _South_, then, are the chief samples of the true Tamulian aborigines of Indian; the characteristics of whom have been preserved by the simple effect of distance from the point of disturbance. Distance, however, alone has been but a weak preservative. The combination of a mountain-stronghold has added to its efficiency.

In _Central_ India one of these safeguards is impaired. We are nearer to Persia; and it is only in the mountains that the foreign elements are sufficiently inconsiderable to make the Tamulian character of the population undoubted and undeniable. In the Mahratta country and in Gondwana, the Ghonds, in Orissa the Kols, Khonds, and Súrs, and in Bengal the Rajmahali mountaineers are Tamulian in tongue and Pagan in creed--or, if not Pagan, but imperfectly Brahminic. But, then, they are all mountaineers. In the more level country around them the language is Mahratta, Udiya, or Bengali.

Now the Mahratta, Udiya[36] and Bengali are _not_ unequivocally and undeniably Tamulian. They are so far from it, that they explain what was meant by the negative statement as to the Tamulian tongues not being considered _Indo-European_. This is just what the tongues in question _have_ been considered. Whether rightly or wrongly is not very important at present. If rightly, we have a difference of language as _primâ facie_--but not as _conclusive_--evidence of a difference of stock. If wrongly, we have, in the very existence of an opinion which common courtesy should induce us to consider reasonable, a practical exponent of some considerable difference of some sort or other--of a change from the proper Tamulian characteristics to something else so great in its _degree_ as to look like a difference in _kind_. With the Bengali--and to a certain extent with the other two populations--the foreign element approaches its _maximum_, or (changing the expression) the evidence of Tamulianism is at its _minimum_. Yet it is not annihilated. The physical appearance of the Mahratta, at least, is that of the true South Indian. Even if the language be other than Tamulian, the Hindús of northern India may still be of the same stock with those of Mysore and Malabar, in the same way that a Cornishman is a Welshman--_i. e._ a Briton who has changed his mother-tongue for the English.

Intermediate to the Khonds and the Bengali, in respect to the evidence of their Tamulian affinities, are the mountaineers of north-western India. Here, the preservative effects of distance are next to nothing. Those, however, of the mountain-fastnesses supply the following populations--Berdars, Ramusi, Wurali, Paurias, Kulis, Bhils, Mewars, Moghis, Minas, &c. &c., speaking languages of the same class with the Mahratta, Udiya, and Bengali, but all imperfectly Brahminic in creed.

The other important languages of India in the same class with those last-mentioned, are the Guzerathi of Guzerat, the Hindú of Oude, the Punjabi of the Punjab, and several others not enumerated--partly because it is not quite certain how we are to place them[37], partly because they may be sub-dialects rather than separate substantive forms of speech. They take us up to the Afghan, Biluch, and Tibetan frontier.

These have been dealt with. But there is one population, belonging to these selfsame areas, with which we have further dealings, Bilúchistan has been described; but not in detail. The Bilúch that give their name to the country have been noticed as Persian. But the Bilúch are as little the only and exclusive inhabitants of it, as the English are of Great Britain. We have our Welsh, and the Bilúch have their Brahúi.

Again--the range of mountains that forms the western watershed of the Indus is not wholly Afghan. It is Bilúch as well. But it is not wholly Bilúch. The Bilúch reach to only a certain point southwards. The range between the promontory of Cape Montze and the upper boundary of Kutch Gundava is _Brahúi_. There is no such word as _Brahúistan_; but it would be well if there were.

_Now the language of the Brahúi belongs to the Tamulian family._ The affinity by no means lies on the surface--nor is it likely that it should. The nearest unequivocally Tamulian dialect on the same side of India is as far south as Goa--such as exist further to the north being either central or eastern. Supposing, then, the original continuity, how great must have been the displacement; and if the displacement have been great, how easily may the transitional forms have disappeared, or, rather, how truly must they once have been met with!

However, the Brahúi affinities by no means lie on the surface. The language is known from one of the many valuable vocabularies of Leach. Upon this, no less a scholar than Lassen commented. Without fixing it, he remarked that the numerals were like those of Southern India. They are so, indeed; and so is a great deal more; indeed the collation of the whole of the Brahúi vocabularies with the Tamul and Khond tongues _en masse_ makes the Brahúi Tamulian.

Is it original or intrusive? All opinion--_valeat quantum_--goes against it being the former. The mountain-fastness in which it occurs goes the other way.

* * * * *

Our sequence is logical rather than geographical, _i. e._ it takes localities and languages in the order in which they are subservient to ethnological argument rather than according to their contiguity. This justifies us in making a bold stride, in passing over all Persia, and in taking next in order--Caucasus, with all its conventional reminiscences and suggestions.

The languages of Caucasus fall into a group, which, for reasons already given, would be inconveniently called _Caucasian_, but which may conveniently be termed _Dioscurian_[38]. This falls into the following five divisions:--1. The Georgians; 2. the Irôn; 3. the Mizjeji; 4. the Lesgians; and 5. the Circassians.

1. _The Georgians._--It is the opinion of Rosen that the central province of Kartulinia, of which Tiflis is the capital, is the original seat of the Georgian family; the chief reasons lying in the fact of that part of the area being the most important. Thus, the language is called _Kartulinian_; whilst the provinces round about Kartulinia are considered as additions or accessions to the Georgian domain, rather than as integral and original portions of it--a fact which makes the province in question a sort of _nucleus_. Lastly, the Persian and Russian names, _Gurg-istan_ and _Gr-usia_, by which the country is most widely known, point to the valley of the Kur.

To all this I demur. The utmost that is proved thereby is the greater political prominence of the occupants of the more favoured parts of the country; as the middle course of the Kur really is.

Of the two sides of the watershed that separates the rivers of the Black Sea[39] from those of the Caspian[40], it is the _western_ which has the best claim to be considered the original _habitat_ of the Georgians. Here it is that the country is most mountainous, and the mountains most abrupt. Hence it is, too, that a population would have both the wish and power to migrate towards the plains rather than _vice versâ_.

More weighty still is the evidence derived from the dialects. The Kartulinian is spoken over more than half the whole of Georgia: whereas, for the parts _not_ Kartulinian, we hear of the following dialects:--

1. The _Suanic_, on the head-waters of the small rivers between Mingrelia, and the southern parts of the Circassian area--the Ingur, the Okoumiskqual, &c. This is the most northern section of the Georgian family.

2, 3. The _Mingrelian_ and the _Imiritian_.

4, 5. The _Guriel_ and _Akalzike_ in Turkish Georgia.

6. The _Lazic_.--This is the tongue of the most western dialects. The hills which form the northern boundary of the valley of the Tsorokh are the Lazic locality; and here the diversity has attained its _maximum_. Small as is the Lazic population, every valley has its separate variety of speech.