Hondius and Blaeuw, we find the mountains north of Kabul termed Nochdarizari, in which we cannot miss the combination Nigudar-Hazárah, whencesoever it was got. The Hazáras are eminently Mongol in feature to this day, and it is very probable that they or some part of them are the descendants of the Karáunahs or the Nigudaris, or of both, and that the origination of the bands so called, from the scum of the Mongol inundation, is thus in degree confirmed. The Hazáras generally are said to speak an old dialect of Persian. But one tribe in Western Afghanistan retains both the name of Mongols and a language of which six-sevenths (judging from a vocabulary published by Major Leech) appear to be Mongol. Leech says too that the Hazáras generally are termed Moghals by the Ghilzais. It is worthy of notice that Abul Fazl, who also mentions the Nukdaris among the nomad tribes of Kabul, says the Hazáras were the remains of the Chaghataian army which Mangku Kaan sent to the aid of Hulaku under the command of Nigudar Oghlan. (Not. et Ext. XIV. 284; Ilch. I. 284, 309, &c.; Baber, 134, 136, 140; J. As. ser. 4, tom. iv. 98; Ayeen Akbery, II. 192-3.)

So far, excepting as to the doubtful point of the relation between Karáunahs and Nigudaris, and as to the origin of the former, we have a general accordance with Polo's representations. But it is not very easy to identify with certainty the inroad on India to which he alludes, or the person intended by Nogodar, nephew of Chaghatai. It seems as if two persons of that name had each contributed something to Marco's history.

We find in Hammer and D'Ohsson that one of the causes which led to the war between Barka Khan and Hulaku in 1262 (see above, Frologue, chap. ii.) was the violent end that had befallen three princes of the House of Juji, who had accompanied Hulaku to Persia in command of the contingent of that House. When war actually broke out, the contingent made their escape from Persia. One party gained Kipchak by way of Derbend; another, in greater force, led by NIGUDAR and Onguja, escaped to Khorasan, pursued by the troops of Hulaku, and thence eastward, where they seized upon Ghazni and other districts bordering on India.

But again Nigudar Aghul, or Oghlan, son of (the younger) Juji, son of Chaghatai, was the leader of the Chaghataian contingent in Hulaku's expedition, and was still attached to the Mongol-Persian army in 1269, when Borrak Khan, of the House of Chaghatai, was meditating war against his kinsman, Abaka of Persia. Borrak sent to the latter an ambassador who was the bearer of a secret message to Prince Nigudar, begging him not to serve against the head of his own House. Nigudar, upon this, made a pretext of retiring to his own headquarters in Georgia, hoping to reach Borrak's camp by way of Derbend. He was, however, intercepted, and lost many of his people. With 1000 horse he took refuge in Georgia, but was refused an asylum, and was eventually captured by Abaka's commander on that frontier. His officers were executed, his troops dispersed among Abaka's army, and his own life

spared under surveillance. I find no more about him. In 1278 Hammer speaks of him as dead, and of the Nigudarian bands as having been formed out of his troops. But authority is not given.

The second Nigudar is evidently the one to whom Abu'l Fazl alludes. Khanikoff assumes that the Nigudar who went off towards India about 1260 (he puts the date earlier) was Nigudar the grandson of Chaghatai, but he takes no notice of the second story just quoted.

In the former story we have bands under Nigudar going off by Ghazni, and conquering country on the Indian frontier. In the latter we have Nigudar, a descendant of Chaghatai, trying to escape from his camp on the frontier of Great Armenia. Supposing the Persian historians to be correct, it looks as if Marco had rolled two stories into one.

Some other passages may be cited before quitting this part of the subject. A chronicle of Herat, translated by Barbier de Meynard, says, under 1298: "The King Fakhruddin (of Herat) had the imprudence to authorize the Amir Nigudar to establish himself in a quarter of the city, with 300 adventurers from 'Irak. This little troop made frequent raids in Kuhistan, Sijistan, Farrah, &c., spreading terror. Khodabanda, at the request of his brother Ghazan Khan, came from Mazanderan to demand the immediate surrender of these brigands," &c. And in the account of the tremendous foray of the Chaghataian Prince Kotlogh Shah on the east and south of Persia in 1299, we find one of his captains called Nigudar Bahadur. (Gold. Horde, 146, 157, 164; D'Ohsson, IV. 378 segg., 433 seqq., 513 seqq.; Ilch. I. 216, 261, 284; II. 104; J. A. ser. 5, tom. xvii. 455-6, 507; Khan. Notice, 31.)

As regards the route taken by Prince Nogodar in his incursion into India, we have no difficulty with BADAKHSHAN. PASHAI-DIR is a copulate name; the former part, as we shall see reason to believe hereafter, representing the country between the Hindu Kush and the Kabul River (see infra, chap. xxx.); the latter (as Pauthier already has pointed out), DIR, the chief town of Panjkora, in the hill country north of Peshawar. In Ariora-Keshemur the first portion only is perplexing. I will mention the most probable of the solutions that have occurred to me, and a second, due to that eminent archæologist, Gen. A. Cunningham. (1) Ariora may be some corrupt or Mongol form of Aryavartta, a sacred name applied to the Holy Lands of Indian Buddhism, of which Kashmir was eminently one to the Northern Buddhists. Oron, in Mongol, is a Region or Realm, and may have taken the place of Vartta, giving Aryoron or Ariora. (2) "Ariora," Gen. Cunningham writes, "I take to be the Harhaura of Sanscrit-i.e. the Western Panjab. Harhaura was the North-Western Division of the Nava-Khanda, or Nine Divisions of Ancient India. It is mentioned between Sindhu-Sauvira in the west (i.e. Sind), and Madra in the north (i.e. the Eastern Panjab, which is still called Madar-Des). The name of Harhaura is, I think, preserved in the Haro River. Now, the Sind-Sagor Doab formed a portion of the kingdom of Kashmir, and the joint names, like those of Sindhu-Sauvira,

describe only one State." The names of the nine divisions in question are given by the celebrated astronomer Varaha Mihira, who lived in the beginning of the 6th century, and are repeated by Al Biruni (see Reinaud, Mem. sur l'Inde, p. 116). The only objection to this happy solution seems to lie in Al Biruni's remark, that the names in question were in general no longer used even in his time (A.D. 1030).

There can be no doubt that Asidin Soldan is, as Khanikoff has said, Ghaiassuddin Balban, Sultan of Dehli from 1266 to 1286, and for years before that a man of great power in India, and especially in the Panjab, of which he had in the reign of Ruknuddin (1236) held independent possession.

Firishta records several inroads of Mongols in the Panjab during the reign of Ghaiassuddin, in withstanding one of which that King's eldest son was slain, and there are constant indications of their presence in Sind till the end of the century. But we find in that historian no hint of the chief circumstances of this part of the story, viz., the conquest of Kashmir and the occupation of Dalivar or Dilivar (G. T.), evidently (whatever its identity) in the plains of India. I do find, however, in the history of Kashmir, as given by Lassen (III. 1138), that, in the end of 1259, Lakshamana Deva, King of Kashmir, was killed in a campaign against the Turushka (Turks or Tartars), and that their leader, who is called Kajjala, got hold of the country and held it till 1287.* It is difficult not to connect this both with Polo's story and with the escapade of Nigudar about 1260, noting also that this occupation of Kashmir extended through the whole reign of Ghaiassuddin.

We seem to have a memory of Polo's story preserved in one of Elliot's extracts from Wassáf, which states that in 708 (A.D. 1308), after a great defeat of a Mongol inroad which had passed the Ganges, Sultan Ala'uddin Khilji ordered a pillar of Mongol heads to be raised before the Badáun gate, “as was done with the Nigudari Moghuls (III. 48).

We still have to account for the occupation and locality of Dalivar; Marsden supposed it to be Lahore; Khanikoff considers it to be Diráwal, the ancient desert capital of the Bhattis, properly (according to Tod) Deoráwal, but by a transposition common in India, as it is in Italy, sometimes called Diláwar, in the modern State of Bháwalpúr. But Gen. Cunningham suggests a more probable locality in DILAWAR on the west bank of the Jelam, close to Dárápúr, and opposite to Mung. These two sites, Diláwar-Dárápúr on the west bank, and Mung on the east, are identified by Gen. Cunningham (I believe justly) with Alexander's Bucephala and Nicaea. The spot, which is just opposite the battle-field of Chiliánwála, was visited (December 15, 1868) at my request, by my friend Col. R. Maclagan, R.E. He writes: "The present village of Diláwar stands a little above the town

* Khajlak is mentioned as a leader of the Mongol raids in India by the poet Amír Khusrú (A.D. 1289; see Elliot, III. 527.)

of Darapur (I mean on higher ground), looking down on Darapur and on the river, and on the cultivated and wooded plain along the river bank. The remains of the Old Dilawar, in the form of quantities of large bricks, cover the low round-backed spurs and knolls of the broken rocky hills around the present village, but principally on the land side. They cover a large area of very irregular character, and may clearly be held to represent a very considerable town. There are no indications of the form of buildings, . . . . but simply large quantities of large bricks, which for a long time have been carried away and used for modern buildings. . . . . After rain coins are found on the surface. . . . . There can be no doubt of a very large extent of ground, of very irregular and uninviting character, having been covered at some time with buildings. The position on the Jelam would answer well for the Dilawar which the Mongol invaders took and held. . . . The strange thing is that the name should not be mentioned (I believe it is not) by any of the well-known Mahomedan historians of India. So much for Dilawar. . . . . The people have no traditions. But there are the remains; and there is the name, borne by the existing village on part of the old site." I had come to the conclusion that this was almost certainly Polo's Dalivar, and had mapped it as such, before I read certain passages in the History of Ziyáuddín Barni, which have been translated by Professor Dowson for the third volume of Elliot's India.' When the comrades of Ghaiassuddin Balban urged him to conquests, the Sultan pointed to the constant danger from the Mongols, saying: "These accursed wretches have heard of the wealth and condition of Hindustan, and have set their hearts upon conquering and plundering it. They have taken and plundered Lahor within my territories, and no year passes that they do not come here and plunder the villages. . . . . They even talk about the conquest and sack of Dehli." And under a later date the historian says: "The Sultan . . . marched to Lahor, and ordered the rebuilding of the fort which the Mughals had destroyed in the reigns of the sons of Shamsuddin. The towns and villages of Lahor which the Mughals had devastated and laid waste he repeopled." Considering these passages, and the fact that Polo had no personal knowledge of Upper India, I now think it probable that Marsden was right, and that Dilivar is really a misunderstanding of "Città di Livar" for Laháwar or Lahore.

The Magical darkness which Marco ascribes to the evil arts of the Caraunas is explained by Khanikoff from the phenomenon of Dry Fog which he has often experienced in Khorasan, combined with the Dust Storm with which we are familiar in Upper India. In Sind these

* Professor Cowell compares the Mongol inroads in the latter part of the 13th and beginning of the 14th century, in their incessant recurrence, to the incursions of the Danes in England. A passage in Wassáf (Elliot, III. 38) shows that the Mongols were, circa 1254-55, already in occupation of Sodra on the Chenab, and districts adjoining.

phenomena often produce a great degree of darkness. During a battle fought between the armies of Sindh and Kachh in 1762, such a fog came on, obscuring the light of day for some six hours, during which the armies were intermixed with one another and fighting desperately. When the darkness dispersed they separated, and the consternation of both parties was so great at the events of the day that both made a precipitate retreat. In 1844 this battle was still spoken of with wonder. (J. Bomb. Br. R.A.S. I. 423.)

Major St. John has given a note on his own experience of these curious Kerman fogs (see Ocean Highways, 1872, p. 286): "Not a breath of air was stirring, and the whole effect was most curious, and utterly unlike any other fog I have seen. No deposit of dust followed, and the feeling of the air was decidedly damp. I unfortunately could not get my hygrometer till the fog had cleared away."

The belief that such opportune phenomena were produced by enchantment was a thoroughly Tartar one. D'Herbelot relates (art. Giogathai) that in an action with a rebel called Mahomed Tarabi, the Mongols were encompassed by a dust-storm which they attributed to enchantment on the part of the enemy, and it so discouraged them that they took to flight.

NOTE 5.-The specification that only seven were saved from Marco's company is peculiar to Pauthier's Text, not appearing in the G. T.

Several names compounded of Salm or Salmi occur on the dry lands on the borders of Kerman. Edrisi, however (I. p. 428), names a place called KANÁT-UL-SHÁM as the first march in going from Jiruft to Walashjird. Walashjird is, I imagine, represented by Galashkird, Major R. Smith's third march from Jiruft (see my map of routes from Kerman to Hormuz); and as such an indication agrees with the view taken below of Polo's route, I am strongly disposed to identify Kanát-ulShám with his castello or walled village of Canosalmi.

The raids of the Mekranis and Biluchis long preceded those of the Karaunas, for they were notable even in the time of Mahmud of Ghazni, and they have continued to our own day to be prosecuted nearly on the same stage and in the same manner. About 1721, 4000 horsemen of this description plundered the town of Bander Abbasi, whilst Capt. Alex. Hamilton was in the port; and Abbott, in 1850, found the dread of Bilúch robbers to extend almost to the gates of Ispahan. A striking account of the Bilúch robbers and their characteristics is given by General Ferrier (see Hamilton, I. 109; J.R.G.S. XXV.; Khanikoff's Memoire; Macd. Kinneir, 196; Caravan Journeys, p. 437 seq.)