Chapter XX: Part I: The Primary Varieties of the Human Species (17)

The word _po_=_on_, and _Labe_=_Elbe_, so that the Polabic Slavonians means the Slavonians on the Elbe. The importance of this section arises from the fact that at the time of Charlemagne they were, with the exception of the tract occupied by the Saxons of Holstein, and the north-west part of Hanover, not only the occupants of Mecklenburg, and the parts _east_ of that river, but of Lauenburg, Luneburg, Altmark, and a vast section of Germany to the _west_ of it.

To suppose that the Slavonic frontier was not equally extended westwards, in the eighth, seventh, sixth, fifth, fourth, third, second, or first centuries, is, in the first instance, to admit the accuracy of an author like Tacitus.

On the other hand, however, it involves the assumption of so vast an amount of migration, displacement, and other unlikely ethnological processes, that a writer who weighs conflicting probabilities is led to the conclusion that a great historian is more likely to be wrong in the ethnology of countries like Prussia and Poland--countries which could be known to Tacitus only as the interior of Africa can be known to Mr. Hallam or Macaulay--than that, between A.D. 100, and 900, a whole Gothic population, extending from the Niemen to the Elbe, should have been replaced by a Slavonic one, without leaving a single trace of its existence in any intermediate locality; the same encroaching Slavonians, when we first find them mentioned by cotemporary historians, being themselves in a state of displacement by the same previously-displaced Germans.

This, however, is but a very general and superficial view of the difficulties that attend the belief that the Oder and Vistula were originally German. Nevertheless, it is all that room can be found for here.

As to the tribes themselves the chief were--

_The Wagrians._--Occupants of the country between the Trave and the upper portion of the southern branch of the Eyder.

_The Polabi._--Conterminal with the Wagrians and the Saxons of Sturmar, from whom they were separated by the river Bille.

_The Obodriti._--This is a generic rather than a specific term. It means, however, the tribes between the Trave and the Warnow; chiefly along the coast. Zeuss makes Schwerin their most inland locality.

_Varnahi._--This is the form which the name takes in Adam of Bremen. It is also that of the Varni, Varini and Veruni of the classical writers; as well as the Werini of the Introduction to the _Leges Anglorum et Werinorum, hoc est Thuringorum_.

_Linones._--Luneburg. Language spoken during the last century. Known through a pater-noster. Slavonic, modified by German.

Such are the chief western Slavonians of the time of Charlemagne. If they were not also the western Slavonians of the first and second centuries, they must have emigrated between the two periods;[190] "must have done so, not in parts but for the whole frontier; must have, for the first and last time, displaced a population which has ever been the conqueror rather than the conquered; must have displaced it during one of the strongest periods of its history; must have displaced it everywhere, and wholly; and (what is stranger still) that not permanently--since, from the time in question, these same Germans, who, between A.D. 200 and A.D. 800, always retreated before the Slavonians, have from A.D. 800 to A.D. 1800, always reversed the process, and encroached upon their former dispossessors."

3.

MEDITERRANEAN INDO-GERMANS.

_Physical conformation._--Dolikhokephalic, high facial angle;
hair, eyes, and complexion, dark; frame, more slender than
bulky.

When we consider that the aborigines of Spain were Iberic, that they probably extended as far as the Rhone, and that the ancient Ligurians of the Gulf of Genoa are not absolutely known in respect to their ethnological relations, the apparent impropriety of restricting the term _Mediterranean_ to the classical nations of Greece and Italy becomes diminished; to which it may be added that the undoubted civilizational influence of the land-and-water conditions of these two peninsulas requires some term to suggest it. The term, nevertheless, is open to amendment.

So much of what belongs to Greece and Italy is historical, that the brevity of the preceding and following notices may be excused.

MEDITERRANEAN FAMILIES AND NATIONS.

_Localities._--Greece and Italy.

_Area._--Discontinuous.

_Divisions._--1. The Hellenic branch. 2. The Italian (Ausonian)
branch.

_Historical Influence._--Preeminently moral. Material as well.

The discontinuity of the Greek and Italian areas is a difficulty which requires more investigation than it has met with, and is a purely ethnological question.

So is the archæological part of both the Greek and Roman ethnology, _i.e._ the relations of the Hellenes and Latins to the early inhabitants of their respective peninsulas.

So is the analysis of their present representatives, _e. g._ the question as to the amount of Slavonic, Italian, or Albanian blood in the modern Greek, or the determination of the Keltic, Roman, and Gothic elements amongst the French.

Of the sub-divisions of the--

ITALIAN BRANCH

the following classification is, perhaps, the most convenient; to which the previous arrangement of the ethnological elements into _a_, the Original; _b_, the Roman; and _c_, the Superadded, gives precision.

1. _Italians._--_Original Elements_--_a_, Samnite, Etruscan, Keltic(?), Ligurian, &c.; _b_, Roman of Rome; _c_, German.

2. _Hesperians. (Spanish and Portuguese)._--_a_, Iberian, Celtic(?); _b_, Roman of the time of the second Punic war; _c_, Gothic, Arabian.

3. _French._--_a_, Celtic for the North, Iberian for the South; _b_, Roman, chiefly from the time of Cæsar; _c_, German.

4. _Swiss of Graubündten._--_a_, Undetermined; _b_, Roman of an uncertain, though probably late, period; _c_, German.

5. _Wallachians._--_a_, Undetermined; probably Slavonic; _b_, Roman of the time of Trajan; _c_, Turk (Hun, Comanian, and Bulgarian), Slavonic, German, Ottoman, Turk.

II.

IRANIAN INDO-GERMANS.

The whole of this class is hypothetical. Such as it is, however, it comprises the populations of Kurdistan, Persia, Beloochistan, Affghanistan, and Kafferistan.

In order to understand the complications which leave so large a section of the human species in an unsatisfactory ethnological position, a notice of the Sanskrit language, and of the _history of opinion_ concerning it, is necessary.

The language called Sanskrit has a grammar of the same copiousness and complexity as the Greek, and a vocabulary which places it in the Indo-European class of tongues.

It is the language of the religious and literary writings of the _Brahminical_ Hindus; the Ramayana and Mahabharata (epic poems) being referred by Sanskrit scholars to the second century B.C.

A more archaic form of it is the language of the Vedas, referred by some Sanskrit scholars to 1400 B.C.

A form said to approach the archaic character of the Veda. Sanskrit is the language of the arrow-headed inscriptions--so far as they are Persian; the date of these being the reign of Darius.

A form (the Pali) _less_ archaic than the Sanskrit of the Mahabharata has been found upon inscriptions of the æra of the Seleucidæ in Babylon, and as such in records older than that of the Non-Vedaic Sanskrit literature.

The same Pali is the language of the _Buddhist_ religion and literature in India, in Ceylon, in the Trans-Gangetic Peninsula, in Tibet, and in the Sub-Himalayan range.

The Zend, a form closely allied to the proper Sanskrit is the language of the oldest Parsi religious books, the Zendavesta.

Lastly--The inscriptions upon the Indo-Bactrian coins of the successors of Alexander are either Sanskrit or nearly Sanskrit.

It is convenient in speaking of these several forms of speech _as a class_, to designate them by the term Iranian.

It is convenient, also, to indicate the extent to which the approach made by the Persepolitan of a period so late as the reign of Darius, to the Vedaic dialect, _said_ to be about one thousand years older, subtracts from the value of a common argument in favour of the antiquity of the Vedas, _viz._ the extent to which the language is more archaic than the Sanskrit of the Epics.

It is well too, to indicate as a further disturbance to the current opinions, the bearing of the _Pali_ character of the inscriptions; whereby the _oldest_ records are embodied in the _newest_ form of language.

All these, however, are subordinate questions; the main point being the enumeration of the Iranian Indo-Germans.

The Iranian Indo-Germans are those nations and tribes, whatever they may be, who are descended from the speakers of the Iranian languages--be they Sanskrit Proper, the Sanskrit of the Vedas, Pali, Zend, or Persepolitan; languages, which, it must be observed, are, in the present state of our inquiry, _dead_ languages.

What, then, are these tribes and nations? The answer to this gives us the Iranian Indo-Germans.

When the Sanskrit literature of India first commanded attention, the answer to this question was--_all_ the nations of Hindostan.

The first researches (those of Ellis and others) upon the languages of southern India showed that the Tamul tongues, at least, were not in this category.

Further researches (those of Dr. Stevenson and others) gave reasons for making the Mahratta language Tamul rather than Iranian--not that the _vocabulary_ was not Sanskritic, but that the _grammar_ was such as could never have been evolved out of the grammar of that tongue.

Prominence being thus given to the non-Sanskritic character of the grammar of one Indo-Gangetic language, the undeniable fact of a vast per-centage of the vocables being Sanskrit, fell in value, as a sign of philological relation.

Thence came an application of the criticism which had unfixed the Mahratta language to the other (apparently) more undoubtedly Iranian dialects of Northern India--the Udiya, the Gujerati, the Hindi, and the Bengali.

The present writer believes that it unfixes these also; an opinion to which he has been led quite as much by what has been said by the defenders as by what has been said by the impugners of their Sanskritic origin. It is not likely any better case will be made out for this, than the one contained in a very able Dissertation of Dr. Max Müllers.[191] Yet it is so unsatisfactory, that it almost proves the question the other way.

Now all this goes to show that Iranian Indo-Germans are not to be looked for in India; except, of course, as a foreign element to the originally Tamul population.

Whether they are to be looked for elsewhere, and (if anywhere) in what quarters, follows the notice of the--

PERSIAN STOCK.

_Physical conformation._--Cranium, dolikhokephalic; complexion,
varied, fair with the mountaineer tribes, dark with those of
the sandy deserts of the south; features, sometimes regular
and delicate, sometimes bold and prominent; in the one case
approaching the character of the high-caste Indians, in the
other Semitic or sub-Semitic.

_Area._--Persia, Beloochistan, Affghanistan, Bokhara,
Kafferistan.

_Languages._--Undeniably Sanskrit in respect to a great per
centage of the vocables. _Not_ undeniably Sanskrit in respect
to their grammatical structure.

The last sentence contains the reason for the provisional character of the present classification. The criticism, or rather scepticism, which has been extended by others to the Indo-Gangetic languages of Hindostan, is extended by the present writer to the Persian.

If so--the nation that is at one and the same time Asiatic and Indo-Germanic, remains to be discovered; it being remembered that it is only Indo-Germanic through its relations with the speakers of the Sanskrit.

The divisions (more or less artificial) of the Persian family are--

1. _The Persians of Northern and Western Persia._--Mahometans. Occupants of elevated plateaux, the alluvial banks of great rivers being exceptional.

2. _The Kurds of Kurdistan._--Mountaineers. Mahometans.

3. _The Beluchi of Beloochistan._--Dark-complexioned, occupants of sandy steppes.

4. _The Patans (Affghans)._--Physiognomy frequently Semitic or sub-Semitic.

5. _The Tajiks of Bokhara._--Here the dominant population is that of the Uzbek Tartars.

6. _The Siaposh._--Fair-complexioned; pagan mountaineers, speaking a language with a great per centage of slightly-altered Sanskrit words.

I have no wish to undervalue the import of this last fact--a fact to which great prominence has been given.

Unaccompanied, however, with any proof that the _grammar_ is Sanskritic, it leaves the question but little altered.

Kafferistan the Siaposh locality, is (roughly speaking) the watershed between the rivers Cabúl and Oxus. In these parts we find conterminous with the Siaposh, and doubtless in the same category--

1. _The Lughmani._--Conterminous with the Affghans.

2. _The Dardoh._--Conterminous with the Cashmirians.

3. _The natives of Wokhan._--On the sources of the Oxus, conterminous with the Turks of Pamer.

More desirous of directing attention to the numerous ethnological difficulties which have arisen, and must yet arise from the adoption of the current opinion respecting the relations between the undoubted Indo-Germans of Europe, and the equivocal Indo-Germans of Asia (meaning thereby a native and aboriginal population), I abstain from any positive expression of opinion as to the quarter from which the Sanskrit language originated. That the language which stands in the same relation to it, as the Italian does to the Latin, has yet to be discovered I firmly believe; to which I may add that, except in Asia Minor or Europe, I do not know where to look for it.

* * * * *

In justice to the classification of the so-called Indian Mongolidæ, I must here remark that the position of the Indo-Gangetic portion of it as Tamulian by no means stands or falls with the relation of its languages to the Sanskrit; since, even if an undeniably Sanskrit origin were proved for them, the evidence of physical form would still justify the inquirer in asking whether they might not still be Tamulians whose language had been replaced by an imported one.

* * * * *

The term quasi-Pulinda now finds an explanation. The key to half the complexities of the ethnology of Hindostan lies in the fact of the Brahminical portion of the population being an invading one, whilst the degree to which it altered the physical and moral character of those who were invaded, has a great range of variation, from a general change to an inappreciable modification.

Now--where the invaded have been so little changed as to preserve both their original habits and their original language, they are full or true Pulindas; whilst, where they have lost their language, but retained enough of their habits to show their probable Pulinda relations, they are called _quasi_-Pulindas.

* * * * *

The "original[192] population of the country which now separates the nearest point of the Dioscurian area from the Seriform" must, in its earliest epoch, have been intermediate or transitional between the two stocks. However, long before the dawn of history, this was displaced. By what nations? Most probably, by one of the two following--The Turks, by means of a southern, the Persians by means of a northern extension.

* * * * *

UNPLACED STOCKS.

In the present state of our knowledge it is safest to leave the following stocks unplaced.

1.

ARMENIANS.

That the Armenian language has Indo-Germanic elements is undoubted. Whether, however, they are sufficient to make it Indo-Germanic is questionable.

Sub-Semitic in appearance, and conterminous with the Semitic area, the Armenian has much in common with the tribes with which he is so often and so naturally associated, the Dioscurian Georgians; and it is through the Armenian that the transition from the Mongolidæ to the Atlantidæ is most likely to be recognized.

2.

IBERIANS.

_Native Name._--Euscaldunac.

_Localities._--The provinces of Biscay and Navarre, in Spain;
the department of the Basses Pyrenees in France. Conterminous
with the _French and Spanish_.

Compared with the Spanish and Portuguese of the Peninsula, and (to a certain extent) with the French of France, the Basque language has the same relation as the Welsh has to the English. It is the remains of the ancient language of the whole country.

Considering its mountain locality and its position at the north-western extremity of the country, on the one hand, and the undeniably recent origin of the present Spanish and Portuguese, on the other, this is no more than is expected _à priori_.

Further proof, however, has been supplied by the researches of ethnographical philologists, most especially by those of W. Humboldt. In an elaborate essay, first published in Vater's Appendix to the _Mithridates_, that writer analyzes the names of the ancient Spanish rivers, mountains, and tribes, and shows that, whenever they have a meaning at all, that meaning is to be found in the Basque.

He shows more, _viz._ that not only Spain and Portugal, but that the Aquitanian province of Southern Gaul was Basque as well; in other words, that the present language of Bilbao and Navarre was extended southwards, and that of Les Basses Pyrenees northwards. Thus far the views of Humboldt have been generally received.

The extension of the Basque to Sardinia and Corsica, to Sicily and part of Italy, is more problematic. Nevertheless, it has been suggested; and, _in the way of colonization_, although _not as an aboriginal language_, it is probable.

A geographical extension, however, is not necessary to create an interest in the Basque language. Its antiquity is that of the oldest tongues of Europe. Before Rome, before Greece, before Tyre or Carthage had been attracted by the mineral wealth of the far west, the mother-tongue of the Basque was spoken on the Douro, the Tagus, the Ebro, and the Guadalquiver. Afterwards it was the language of those who defended Numantia and Saguntum; of those who dealt with the Greeks at Emporiæ, and of those who bought and sold with the Phœnicians at Gades and Tartessus. The Lusitani, the Turdetani, the indomitable Cantabri, were Euskaldunac. It is better, however, when speaking of the Basque in its oldest form to call it _Iberic_ or _Iberian_.

That the general ethnological relations of the Basque are undetermined is denoted by the place it takes in the present volume. The principle, however, which is most likely to determine it deserves to be noticed. It arises out of a bold conception of (I believe) Arndt's, adopted in its fullest extent by Rask, and, serving, at the present moment, as one of the best _methods_ which honourably characterize the Scandinavian school of ethnology.

Just as, in geology, the great primary strata underlie the more recent super-imposed formations, so does an older and more primitive population represent the original occupancy of Europe and Asia, previous to the extension of newer, and (so to say) secondary--the Indo-Germans.

And just as, in geology, the secondary and tertiary strata are not so continuous but that the primary formations may, at intervals, show themselves through them, so also do fragments of the primary population still exist--discontinuous, indeed, but still capable of being recognized.

With such a view--the earliest European population was once comparatively homogeneous, from Lapland to Grenada, from Tornea to Gibraltar. But it has been overlaid and displaced; the only remnants extant being the Finns and Lapplanders, protected by their arctic climate, the Basques by their Pyrenæan fastnesses, and, perhaps the nation next in order of notice.

The Euskaldune is only one of the isolated languages of Europe. There is another--the Albanian.

The notion that the Albanian is a mere mixture of Greek and Turkish, has long been superseded by the conviction that, although mixed, it is essentially a separate substantive language. The doctrine, also, that it is of recent introduction into Europe has been similarly abandoned. There is every reason for believing that, as Thunmann suggested, it was, at dawn of history, spoken in the countries where it is spoken at the present moment.

If so, it is easily identified with either the ancient Illyrian, or the ancient Epirote; and, as it is by no means certain that these two languages were essentially different, it is possible that the Albanian may represent both. Hence, it would certainly be spoken by a portion of the soldiers of Pyrrhus, and, most probably, by the whole army of Teuta and Gentius. At present, however, it is enough to insist upon its independent character as a separate substantive language.

ALBANIANS.

_Native Name._--Skipetar=Mountaineer.

_Turkish._--Arnaout.

_Locality._--The ancient Illyria and Epirus. Albanian settlers
in Greece, Turkey, and Calabria.

_Conterminous_ with the Greek, Turk, Slavonic, and Italian
languages; and containing numerous words borrowed from each of
them.

_Religion._--Imperfect Christianity and Mahometanism.

_Social Constitution._--Division and sub-division into tribes
and families.

EXTINCT STOCKS.

Is there reason to believe that any definite stock, or division of our species has become either _wholly_ extinct, or so incorporated as to be virtually beyond the recognition and analysis of the investigator? With the vast majority of the _so-called extinct_ stocks this is not the case--_e. g._ it is not the case with the old Gauls of Gallia; who, though no longer extant, have extant congeners--the Welsh and Gaels.

To an extinction of this kind amongst the better-known historical nations of Europe and Asia--for in America such extinction, or the tendency towards it, is the normal condition of the majority of the aboriginal populations--the nearest approach is to be found in the history of--

1.

THE PELASGI.

_Æra._--In the time of Herodotus, known only in two--

_Localities._--Chreston and Plakiæ.

_Area._--As then known, discontinuous.

_Language._--Unintelligible to an Hellenic Greek.

I follow Mr. Grote, in his masterly separation of the wheat of contemporary evidence from the chaff of tradition in respect to the Pelasgi; but do not follow him in the inference from the dissimilarity between their language and that of the Hellenes. The two sections might still be as closely allied as the Greek and Roman. On the other hand, the difference might be as great as that of the Hebrew and English.

The point of most importance is the nature of their two unconnected points of occupancy at the time of Herodotus.

1. If these represented _parts of the original area_, the intermediate portions whereof had been overlaid by a permanent invasion, the evidence would be in favour of the Pelasgi having been in the same category with the Thracians; and, as such, _perhaps_ Slavonic.

2. On the other hand--if they represented two separate _colonizations_ such a distribution would indicate an origin in _a._ Asia Minor; _b._ the Ægean Islands; or _c._ Continental Greece.

A sanguine scholar may, perhaps, hope that an investigation of the present dialects of the two Herodotean localities may reward the minute analyst with some Pelasgic glosses.--_Optandum magis quam sperandum._

2.

THE ETRUSCANS.

_Æra of their maximum development._--The earlier centuries of
the Roman Republic. Veii taken 360 A.U.C.

_Historical Influences._--Upon early Rome.

_Social Development._--Agricultural, architectural, religious,
commercial, artistic. Partially self-developed. Probably,
chiefly of Greek origin.

_Alphabet._--Derived from the Greek.

_Language._--Extant, only in hitherto untranslated (or
imperfectly translated) fragments. Considered, by Lipsius, as
Indo-Germanic.

The reason in favour of the descent of the Etruscans from the Rhætian Alps has not been put, even by Niebuhr, so strongly as it might have been.

What we find in Livy is something more than an _opinion_ to that effect. It is an express statement that the Rhætian and Etrurian languages were alike.

If so, we have a discontinuous area; an area which--considering that the Cisalpine Kelts were preeminently the tribes of an encroaching frontier--was, most likely, originally continuous.

I believe, then, that the Etrurians represented the _maximum_ civilization, and the Rhætian mountaineers the _maximum_ rudeness of one and the same stock--a stock originally indigenous to Northern Italy, but subsequently broken-up by Keltic and other permanent invasions. Such, at least, is the ethnological view of the question--based upon the general phænomena of ethnological distribution.

3.

POPULATIONS OF ASIA MINOR.

How numerous these may once have been is difficult to determine. Thus much, however, may safely be assumed;--

1. That the languages represented by the western dialects of the Georgian had _some_ extension beyond their present frontier--possibly as far as Bithynia.

2. That the languages represented by the Lycian of the Lycian inscriptions had _some_ extension beyond Lycia--possibly (though there are several difficulties to be reconciled) as far as the Hellespont.

3. That on some portion of the coast, a language intelligible to some portion of the Thracians on the one hand, and the Armenians on the other, was spoken.

Such are a few of the details of an important section of our subject.--They are given, however, more for illustrating the nature of the difficult question of _Descent_ than for exhausting the subject.

The same applies to the complex subject of--

HYBRIDISM (EXTREME INTERMIXTURE).

Of this just enough will be said to illustrate the form which the present classification of the primary divisions of mankind renders necessary.

I.

IAPETIDÆ AND MONGOLIDÆ.

A. _Kelts with Mongolidæ._--The infusion of Keltic blood takes place when the Welsh, Irish, or Scotch of England, like the--

B. _Goths with Mongolidæ_, come as 1. _English or Americans_, in contact with--_a_, Chinese; _b_, Malays; _c_, Polynesians; _d_, Australians; _e_, Eskimo; _f_, American Indians; _g_, East Indians.

2. _High-Germans with_--_a_, American Indians; _b_, Finns.

3. _Dutch with_--_a_, Chinese; _b_, Malays; _c_, East Indians; _d_, South Americans (Guiana).

4. _Scandinavians with_--_a_, Eskimo; _b_, Ugrians.

C. _Slavonians with Mongolidæ_--chiefly Russians with--_a_, Siberians; _b_, Eskimo; _c_, North-east Americans; _d_, Turanians; _e_, Dioscurians.

D. _Mediterranean Indo-Germans with Mongolidæ_--chiefly with--

1. _French with_--_a_, North Americans; _b_, South Americans (Guiana).

2. _Spaniards with_--_a_, Malays (the Philippines); _b_, North Americans (Mexico, &c.); _c_, South Americans (Peru, Buenos Ayres, Guiana, Venezuela, &c.)

3. _Portuguese with_--_a_, Chinese; _b_, East Indians; _c_, Brazilian Americans.

II.

IAPETIDÆ AND ATLANTIDÆ.

A. _Kelts with Atlantidæ._--Under the same conditions as English Goths.

B. _Goths with Atlantidæ._

1. English and Americans with Africans.

2. Dutch with Hottentots--_Griquas._

C. _Mediterranean Indo-Germans with Atlantidæ._ Spanish and Portuguese with Africans.

III.

ATLANTIDÆ AND MONGOLIDÆ.

1. North American Negroes with Native Indians--_Zambos_.

2. South American Negroes with Native Indians--_Mamelucos_.

It is only when two extreme sections of two of the primary divisions meet that there is true Hybridism. With intermediate and transitional forms, such as the Arab and Indian, and others, there is merely--

SIMPLE INTERMIXTURE.

This is a point of minute ethnology. To take a few of the European populations as instances, it attempts to determine the amount of foreign elements in--

1. _The English._--These being Keltic, Roman, Danish, Anglo-Norman, &c., anterior to, or engrafted on, a Saxon foundation.

2. _The French._--Foundation, Roman; other elements, Keltic, German, &c.

3. _The Spanish._--Foundation, Roman; other elements, Iberic, Goth, Arab.

4. _The Germans._--Foundation, Gothic; other elements, Slavonic, Keltic.

5. _The Slavonians._--Non-Slavonic elements, Ugrian, Turk, Mongol, Dioscurian, &c.

6. _The Hungarians._--Non-Majiar elements; Roman, Turk, Mongol, Slavonic, German.

And so on throughout most countries of the world.

Intermediate between simple and extreme intermixture (or Hybridism), but at points where it is difficult to draw a line of demarcation, are such half-breeds as those of the Turk and Mongol, Turk and Persian, Turk and Georgian, Persian and Georgian, &c.--the difference between the parent stocks lying within a small compass.

FOOTNOTES:

[188] See p. 537.

[189] Mithridates. Vol. iii. p. 639.

[190] Philological Transactions, No. 93.

[191] Transactions of British Association for the Advancement of Science, 1847.

[192] See page 123.