the Turks who, along with the empire, assumed the title of the Seleucidæ. Seljuk the Turk, eo nomine, has no historical existence. If so, the ignorance of the Sultan Mahmud is better than the learning of the historians. Alp Arslan is the nephew of Togrul, and Malek Shah the son of Alp Arslan.

The empire of Malek Shah includes (1) Georgia, Armenia, and Asia Minor on the west; (2) Kashgar on the

east.

A.D. 1092.

It breaks up, however, after his death, into the minor kingdoms of Persia, Kerman, Syria, and Roum.

(1.) Persia.-The little that is known concerning the kingdom of Persia suggests the idea that it was broken up by fresh invasions from Turkestan. At any rate its Sultan Sangiar is defeated and taken prisoner by the Uz.

A.D. 1151.

(2.) Kerman.-The history of Kerman is darker than that of Persia. Not so that of

(3.) Syria, which is, to a great extent, the history of the first and second crusades. It is against not merely the Mahometan but the Turk Mahometan, that the Christian puts on his armour.

In Syria,

A.D. 1127-1174.

The Atabegs (Turks with a Turkish title) make themselves masters of Aleppo, Damascus, Edessa, and Mosul, encroaching on the Kurd frontier, as preliminaries to their conquest of Egypt. Egypt, however, they soon abandoned to the great Kurd, Sultan Saladin; only, however, as rulers. There are still Turks in the valley of the Nile, and of these come

A.D. 1250.

The Baherite Mameluks.-Baherite as opposed to Borgite; the Borgite Mameluks being of Circassian origin. When Selim I. invaded Egypt,

the "military force of the Mamelukes consisted of three classes of warriors; all cavalry superbly mounted and armed, but differing materially in rank. First, there were the Mameluks themselves-properly so called-all of whom were of pure Circassian blood, and who had all been originally slaves. The second corps was called the Djebbans, and was formed principally of slaves brought from Abyssinia. The third, and lowest in rank, was called the Korsans, and was an assemblage of mercenaries of all nations. There were twenty-four beys or heads of the Mameluks, and they elected from among themselves a Sultan, who was called also Emirol Kebir, or Chief of Princes. He reigned over Egypt and Syria, and was also recognised as supreme sovereign over that part of Arabia in which the holy cities of Mecca and Medina are situated."*

(4.) Roum, Anatolia, or Asia Minor.—The Seljukians of Roum are defeated by the Mongols. A fragment, however, of their power is preserved in the little kingdom of Konieh (Iconium); connected with which is the history of

A.D. 1304.

The Chorasmian Turks.-We find formidable bodies of Turks, called Chorasmians or Karismians, in Persia, in Syria, and in Asia Minor, between the death of Malek Shah and the death of Kazan Khan, the last descendant of Tshingiz who governed Persia. They must have come from western rather than eastern Turkestan, from Khiva rather than Kokan or Kashgar. I believe that they were Uz. It was they who dethroned Sanjiar. It was they who, under Mohammed, were attacked and defeated by the Mongols under Tshingiz-khan. It was they who, under Gelaleddin, attempted the recovery of their lost empire after the decline of the power of the Mongols, and who, after the * Creasy's Ottoman Turks, vol. 1.

A.D. 1218-1284.

death of their chief, broke up and separated; some for Syria, some for Palestine, some for the service of the Sultan of Iconium. Amongst these last was Soliman Shah, the father of Ortogrul, the father of Othman, the founder of the dynasty of

The Ottoman Turks, or the Turks of the present Turkish empire-occupants of Rumelia, masters of Bulgaria, Bosnia, Macedonia, Anatolia, Syria, suzerains to Moldavia, Wallachia, Ægypt, Tripoli, Tunis.

The Tokhar.-This is the name of one of the tribes of the north-east. They are expected by the Romans as allies against some of their enemies on the west. A district of Persia took from them the name Tokharis

tan.

The Tshagatai.-It is the Tshagatai Turks whose history is so pre-eminently mixed up with that of the Mongols, whose history is that of Timur, and of Baber, conqueror of India, and founder of the Great Mogul dynasty. Yet he was a Turk, notwithstanding. So was Timur. The notice, however, of the Turks of India will appear hereafter.

I now return to

The Khazars. It is in the seventh, eighth, ninth, and tenth centuries that they are conspicuous. The Khanat of Astrakhan is Khazar. The Arab name for the Caspian is the Sea of the Khazars. Meanwhile the Khanat of Kazan is Bulgarian, i. e. Turk and Ugrian.

As the Khazars recede from the field of history, The Petshinegs grow prominent. Wallachia and Moldavia are their area; and they are known to the Hungarians as Bisseni and Bessi, the etymon of Bessarabia. Out of the rougher form of the word has grown the name Budziak.

The Uz.-Whether this be a name of a new group, or

VOL. I.

L L

a name of general application, is uncertain. The Arab form is Ghoz. There were, as has been seen, Uz in Karismia. There were Uz in Moldavia.

The Cumanians.-These have been already noticed. They overrun Volhynia, parts of Galicia, and parts of Hungary. Their language was spoken A.D. 1770-so that they are all but an actually-existing tribe.

The Khazars, the Petshinegs, the Cumanians, and the Osmanli are pre-eminently the Turks of Europe; of which the Danubian Principalities, Bulgaria, Rumelia, and parts of Hungary are (in blood at least) the most Turkish portions.

All the tribes of the preceding list agree in being not only Turks, but Turks of the historical period. Sometimes they are Turks eo nomine. Oftener, however, the place in the Turk family is expressly stated by some competent authority. In the instance of the Cumanians we have a specimen of the language. In all cases, however, the evidence is direct, satisfactory, and generally recognized. With

The Huns, and with

The Scythians there are differences of opinion. That both were of Asiatic origin is universally admitted. That, Mongolia, however, rather than Turkistan was their original area has been not only suggested but insisted on; and that by able writers and influential authorities. Nor can it be denied that the accounts of the physical appearance of both the subjects of Attila and the Herodotæan Skoloti suggest Mongolian affinities to those who only know the Turks as they are represented by the Osmanlis of Constantinople. The Osmanlis, however, to say nothing of the effects of their long residence in Europe, are Georgian, Circassian, and much else beside in blood. As for the identification of the

Huns with the ancestors of the present Magyars, it is a careless inference from the word Hungary. The evidence that they were in the same category with the Avars is as conclusive as the evidence that the Avars themselves were Turks.

Satisfied as I am that the Scythians of Herodotus were members of the Turk family, I am not prepared to say that all to whom the name Scythia applied were the same. I think that some particular divisions of the name may have been Mongol. The further we go east the likelier this is to have been the case. The Persian equivalent to the word Scythæ was Sacæ. Hence what was Scythia in Europe became Sacasthan in Asia. Now, Sok is the name by which, at the present moment, a Tibetan designates a Mongolian. An ancient Persian, I think, most probably applied it to both; oftenest to the Turk. Some few of both the Scythæ and Sacæ may have been Ugrian. As a rule, however, the names applied to Turks.

For the purposes, however, of the present inquiry the details of the difference are unimportant. The minute ethnology of the Turks is one thing. The extent to which tribes akin to the Turk, if not actually Turk, whose original occupancies were the distant steppes of Central Asia, extended themselves southwards and westwards, is another.

Now, I wish to show not only that the Turks have been remarkable amongst migrants and conquerors for their conquests and migrations, but that at the beginning of the historical period the general character of their distribution over the face of the earth was much as it was in the fifth, sixth, and seventh centuries. The degree to which India, Persia, and Syria, and even parts of Africa, were just as Turkish in the reign of Cyrus as they were under the Seljukians will be considered when the proper ethnology of those countries comes under notice.