§ 22.) The description here given by Xenophon of the preserving their wine in plastered cisterns assists in clearing up a question which has created much discussion among travellers, as to the use of the numerous plastered cisterns, which are so frequent in Kurdistan, Armenia, and northern Syria, and which being in the form of a pear, and the mouth often closed by a single great stone, have been looked upon sometimes as sepulchres, and at others as granaries and reservoirs for water; but which were, no doubt, for the store of wine, when that luxury was more abundant in those countries.

There exist to the present day several large Kurd villages, having goodly wood houses, resembling the better class of Swiss cottages, perched upon the hills and cliffs, of this, the most difficult of the passes of the Tigris; and along the valley of which there is not even a horse or wading road, except at the very dryest part of the summer season; at which time I succeeded in passing it, and hence it is that travellers following the course of the river, are accustomed to cross its stream at the ferry (Chelék,) a few miles to the northward, and to follow a road to Jezireh, which is carried inland beyond the left bank.

6

The following day, the Greeks, although continually harassed by the Kurds, who everywhere occupied the eminences, (and the whole of the road is hilly,) arrived at villages, which were situated, above the plain that extends to the river Centrites, and they thus accomplished a distance of from nine to ten miles. This river is described as 200 feet broad, "and the boundary between Armenia and the country of the Karduchians ;" and also "about six or seven stadia (not quite a mile) from the Karduchian

mountains."

The distances as given in the foregoing narrative identify the Centrites with the Buhtán-chai, which I first determined not to be the river of Betlis, but to flow eastward of Se'rt, and to empty itself into the Tigris, opposite to the ancient Armenian mound and village of Til, celebrated in Armenian history as favoured by Tigranes, and the burial-place of several of the Armenian pontiffs. The history of errors entertained previously, with regard to the Buhtán-chai, form a curious chapter in the progress of geography.

The passage of the Centrites was disputed to the Greeks by Armenians, Mygdonians, and Chaldeans, and they were harassed by Kurds in their rear. It is remarkable that there still exist to

The Kurds pronounce Kelék, a raft or ferry, Chelék.

the present day several villages of Chaldeans in this neighbourhood. The river Centrites is not fordable below Janiminiyáh, where it is hemmed in by hills; and this spot coincides also with the description given of the eminence occupied by the enemy on the opposite side. The army having passed the river about noon, marched at once over the plain of Armenia, intermixed with hills of easy ascent, making no less than fifteen miles, when they arrived at a large village, which had a palace in it belonging to the satrap; this, by the distances given, would correspond to the position of Se'rt, which, like Zákhú, preserves to the present day the character of a large village, with a palace in it. From this place they made in two days' march 30 miles, and were then advanced above the head of the Tigris. Had the Greeks marched by the great road from Sert to Betlis, the distances here given would not have carried them as far as to the head-waters of the Tigris, which are at Básh Khán, not far from lake Van; they must therefore have ascended directly towards the great chain of 'Ali Tágh corresponding to the ancient Niphates; by which proceeding, a journey of 30 miles would have carried them beyond the headwaters of the tributaries of the Tigris, and another 45 miles would have taken them to the valley of the Kárá-sú, which has already been recognized by others as corresponding with the Teleboas of our author, and which flows through the plain of Mush the ancient Moxoene-apparently in remote times, as at present, the seat of numerous towns and villages, and having a large population.

From the river Kárá-sú, the Greeks marched up the plain, and in three days made 45 miles, when they came to a palace surrounded by many villages. The direction followed by the Greeks, after reaching the Kárá-sú, is determined by the time which it took them to arrive at the Euphrates, which they are described as passing over not far from its sources. Had the Greeks pursued a northerly course, they would have arrived at the Murád-sú, or eastern Euphrates, in a day's march or less; but at that point it would not have been fordable; and it must have been for the purpose of arriving above its junction with the river of Khanús, called Bin-göl-sú", that they followed a north-easterly direction, up the plain of Mush, and towards the sites of Perák or Lis. north of lake Nazúk, and in which fertile district the palace and village described by Xenophon appear to have existed. The great

7 "River of a thousand lakes," from the mountain of the same name.

cold experienced by the Greeks in the Armenian uplands, and the sufferings entailed thereby, have been the subject of a variety of conjectures. A positive elevation, amounting on the plain of Mush, by my own barometrical observations, to 4200 feet above the sea; at Khanús to 5200 feet, at Erz-Rúm to 5500 feet, and preserving, if not surpassing, the same elevation, in all the intervening country; and in which the immediate results of a lower temperature induced by elevation, are increased by the openness of the country and the long continuity of high and elevated tracts of land, appear quite sufficient to account for this otherwise curious phenomenon.

After an excursion against Teribazus, the Greeks set forward through a deep snow, and made four marches "through a desert" to the Euphrates, which they passed, "the water coming up to the navel." Allowing a distance of about six miles for the first day's march, in which they are described as only encompassing an eminence possessed by Teribazus, and 15 miles for each of the ensuing days, the point where the Greeks would have forded the river would have been at or near the ford of Melásghird, the first which presents itself above the junction of the Bin-göl-sú.

Having crossed the eastern Euphrates at the first feasible point, the Greeks appear to have wished to regain a more direct course to the sea, and they made in three days' march 45 miles, over a plain covered with deep snow, and then another day, also in the snow; when they came to villages, near to which, or about three miles distance, was the palace of the satrap. This corresponds closely with the position of Khanús Kal'ehsi (Jihán Numá, p. 425). This wild castle, built upon more ancient foundations, is situated on a branch of the Bin-göl-sú, in the midst of the Armenian uplands, three miles from 'Arúz, the nearest village, and beyond which, again, is the fertile portion of the district, containing, in the present day, eighteen Armenian villages. This isolation of the castle or palace from the villages is a remarkably distinctive fact. The description given by Xenophon of the Armenian houses, has been recognized in all its simple yet forcible truth, by almost every traveller in these countries.

At these villages they obtained a guide, who after conducting them three days over an uninhabited country, ran away from them; and they made after that seven more marches, before they arrived at the river Phasis, where it is described as being 100 feet in width (iv. 6. § 4). The distance here travelled of 150 miles would carry the Greeks to the Arás, north of Mount Ararat. In giving the name of Phasis, applied by Strabo, Pliny, and many other

authorities, to the Colchian river, now called Rhión, to the river Arás or Araxes, Xenophon appears to have followed a tradition belonging to earlier times than the imagined discovery of the Phison or Phasis, as a tributary to the Euxine; and to have identified the Arás with the Phison of the Scriptures, which sprang from the same locality as the Euphrates, and the Hiddekel or Tigris. Rennell, Delisle, and others, have advocated this identity; and it is remarkable that the upper part of the Arás is still called Pásín-chai. Having crossed the Phasis or Arás, north of Mount Ararat, the Greeks would have before them the redoubtable chain called the Kapán Tágh, the Coraxii of Pliny, and which they reached in two marches making 30 miles; and having carried the pass against a mixed army of natives, designated as xáλußes kaì Táoxoi Kai Paσavoi, they marched down into the plain, where they found villages well stored with all sorts of provisions.

From these villages the Greeks advanced through the country of the Taochians, or that part of Georgia which extends between the Arús and the Kur or Cyrus, and through which they travelled five days, making 90 miles, during which they suffered many privations, for their provisions began to fail them, the Taochians inhabiting mountains and fastnesses; from which custom they probably derived their name-living indeed like the fowls of the air; Taöke in the original language of the Georgians and Tatars of the Kúr, signifying a fowl or bird. At the end of the five marches they arrived at a strong place without houses, but where a great number of men and cattle had assembled, and which, from the description of its inaccessible cliffs, and the distances travelled, would appear to be represented by the modern Russian fort of Tzalka, or Tzarskie as it is called by Strüve in his astronomical positions, and the same as the Tahoskari of D'Anville, and Taochir of Delisle.

How the Greeks after advancing 90 miles into Georgia were led to change their direction, and return the same distance, and also that quantity added to it, which would make the difference between where they entered the country, from the Arás, and issued from it by the 'Arpá-chaï (Apraσos), or in all 150 miles, does not appear, but it was connected with the capture of the fort of the Taochians, towards which they may have only moved in search of the positive necessaries of life, or they may have obtained there information calculated to mislead them with regard to the direct road to the sea. Be this as it may, we have travelled too far with our historian, and found a too constant accuracy in

all his details, to venture to express doubts as to that accuracy now that the amount of journeys and extent of countries traversed almost exceeds credibility, and could indeed scarcely be believed, were it not for this tested accuracy of the historian and the connecting points, which even here, in the most obscure portion of the whole narrative, still serve, like links in a chain, to illustrate these otherwise extraordinary and almost inexplicable movements of the Greeks.

The leading fact to be observed in this portion of the narrative is, that the Greeks must have necessarily crossed the Harpasus ('Arpá-chai,) low down in its course, from the account given of its width; and also that they crossed it from east to west; for crossing it, they came into the country of the Ekvool, whose country, after gaining Gymnias, and after a journey of six days, directly towards the sea, they found separated by a river from that of the Macrones, and hence it was to the westward of the Harpasus.

From the Harpasus the Greeks next travelled 60 miles to villages, which would appear to have been thus situated at the head-waters of the river of Kárs, and from whence to have gained the city of Gymnias they had to effect the passage of the Soghánli Tágh. From these villages they travelled 60 miles in four days' marches, to the city of Tupvías, called Tvpvaoía by Diodorus Siculus, and which appears to correspond to the "citadel of the Romans or Greeks," as its modern name Erz-Rúm signifies, and whose city must have had a name among its previous possessors, anterior to that given to it by the Arabs. But this is one of the few unsatisfactorily determined identifications in this memoir.

Here the Greeks obtained a guide, who led them in five days to the holy mountain eńxns, from whence, to their infinite joy, they first saw the sea. This spot is on the Kóp Tágh, the chain which separates the Kárá-sú from the Tchorúk-sú. Five marches certainly appear considerable to have been required to arrive. at this distance from Erz-Rúm, and the length of the marches is not given, but that of the three subsequent marches, through the country of the Macrones, is given, and amounts to 10 miles per diem. If the Greeks only marched at the same rate from Gymnias to Theches, the five marches would be at once explained, for by the road I travelled, there were 67 miles from Erz-Rúm to Baïbûrt, on the Tchorúk-sú; and that they were on the northern side of the chain, or that which is above the Tchorúk-sú when they first distinguished the sea, is evidenced by the fact that there only remained part of a day's journey to gain the river.