quicksand hills, forests, and deserts." The southern limit is not given.

Here we are told, very plainly, that the northern boundary of Cashgar, is a range of mountains, that extends on one side to SнâSH; that is TASHKUND, on the river Sihon (or Jaxartes): and that these mountains branch out from another ridge, that forms the western boundary of the same kingdom. Of course, we are left to understand, that the northern frontier of it, is on a parallel, or nearly so, with the territory of Shâsh: and this is well known to lie on the north of Cogend; and to have its capital, Tashkund, situated in latitude 42°, according to the tables of Abulfeda, Nasereddin, and Ulug Beig. The long ridge of mountains, on the west, is of course, that which extends from the N of Cashmere, to a point beyond the heads of the Sihon; separating, in that part, the countries of Cashgar, and Turkestan. In Sherefeddin's Timur, this ridge is named KARANGOUTAC; and is reckoned inaccessible, in a military point of view (Book V. chap. iv). We must once more call the reader's attention to the IMAUS, which separated the two SCYTHIAS; which this ridge of KARANGOUTAC represents. The two countries, or rather tracts, separated by it, and which answer to the southern parts of the two Scythias, are generally named the two BUCHARIAS, by Europeans. Of these, the western tract, or that of Samarcand, is named GREAT BUCHARIA; and the eastern, LITTLE BUCHARIA: and this includes, amongst other divisions, that of Cashgar and Koten. This tract is also the original country of the Moguls; or MOGULISTAN: and hence this term is applied by Shahnawaz, to the mountains that form its northern boundary.*

No doubt ought to remain, concerning the proximity of Cashgar to Great Bucharia; when we have proved, from the writings

* I apprehend that the term Bucharia, or Bokhara, is derived in the first instance, from the city of Bokhara, near the Jihon (Oxus), which was the emporium of the commerce, carried on by Europeans, in that quarter: and that it was afterwards extended to the adjoining country, beyond it. There are many examples of this kind: in particular, see page xxi of the Introduction, and the note.

of Abul Fazil, and Abdul Humeed, that the south-west extreme of Cashgar, joins to the N E quarter of Cabul: or admitting that part to be no more than a nominal dependency of Cashgar, the real territory itself is not likely to be very remote.

It seems to me that the capital of Cashgar, as well as the mountains on the north of it, which, in D'Anville's map of Asia, extend towards Acsou and Terfan; ought to be removed several degrees to the north-west, and to the neighbourhood of Al Shâsh: these being the mountains of Mogulistan, described by Shahnawaz. This will make a prodigious change in this part of our maps of Asia. Far more consonant to the ideas of Shahnawaz, is this part of Strahlenberg's map of the Rufsian empire; which includes also the western Tartary, and the courses of the rivers Jihon and Sihon (Oxus and Jaxartes). His mount Musart, which passes along the north of Cashgar, and Mogulistan in general, at the height of 43 and 44° of latitude, answers to the Mogulistan mountains of Shahnawaz; and does really join to Shâsh, on the west: having also the long ridge branching out, and forming the western boundary of Cashgar.

Strahlenberg's map is certainly a composition of great merit, for the time in which it appeared (1730); and proves that he had taken a great deal of pains to collect materials, for the tract lying between the Russian borders, and those of India and Persia. Through the want of observations of longitude, his distances are often very faulty: but I am of opinion that his ideas were too much slighted, by some geographers, who came after him; and who have given the preference to matter of much less value, than that which he has exhibited.

Considering this part of Strahlenberg's map, in a very general way, the following are the observations that I have made on it. 1st. Petersburgh, the Caspian sea, and Samarcand, stand nearly in their proper positions, in respect of each other.

2d. Cabul, Cashmere, and Cashgar, although nearly right, in

respect of each other; are from 4 to 5 degrees too far east, in respect of the Caspian sea, and Samarcand.

3d. The head of the Irtish river is by several degrees too near to Cashgar; even as the latter stands in the map.

Hence it follows, that the space between the Irtish and Cashgar is, out of all proportion, contracted; whilst Great Bucharia is too much extended. Few parts of the continent of Asia, appear to have their geography so imperfectly described, as that between the Russian frontier, and Bucharia. I think, too, that our maps are in a great error, with respect to the positions of the countries lying between Bucharia and China: all of which, in my idea, have been made to recede too much from Bucharia, towards China. It is to the Russians that we are to look up, for better information.

In the discussion of the position of Cashgar, I laid out of the question, entirely, the Chinese and Tartarian geography in Du Halde. At the same time, I would not be understood to impeach the truth of the Chinese geography of Tartary, in general; because there is an error in a particular part of it. It must stand or fall hereafter, according to its own merits. But the great error respecting the latitude of the upper part of the Ganges, in the same geography, ought to make us receive with caution the remaining particulars of it.

Having noticed an error or two of M. D'Anville's, which fell in my way, I feel it a duty even to go out of my way, in order to add my opinion to his, on the subject of Serica; and the limits of the world, as known to the ancients.* I cannot hesitate a moment, after examining the evidence, to determine with him, that the Sera Metropolis of Ptolemy, was situated at the NW extremity of the present empire of China: and very near to the parallel assigned it by that geographer. Of course, the knowledge of Ptolemy ended at this point, eastward. The rivers of Serica, al

* See the Supplement to D'Anville's Antiq. Geog. de l'Inde.

though described to run to the parallel of 55°, and upwards; a parallel far beyond the heads of the Siberian rivers, yet are evidently meant, by the context, for the rivers of Eygur (Yugure), and Tangut. I differ in one particular from M. D'Anville; which is, that the CHARDE, not the ITHAGURI, represent the YUGURES: I think the position clearly points it out. As for the error in Ptolemy's latitude, as well as his longitude, it must be placed to the account of badness of materials. Had Ptolemy lived in the present times, he might have expressed his wonder, that, considering the advantages we pofsefs, OUR maps of this part should be so incorrect; when the tables of Abulfeda, Nasereddin, and Ulug Beig, and the History of Timur, by Sherefeddin, have been so long amongst us, in an European language.*

After so many digrefsions, for which I solicit the reader's indulgence, I shall close the account of the small map, with an observation or two, respecting some geographical misconceptions which I have observed to prevail. The first is, that the modern Bucharia (or Bokhara), is the same with the ancient Bactria. This is so far from being the case, that Bucharia is situated beyond the river anciently called the Oxus, or the modern Jihon: and is the country anciently named Sogdiana; from Sogd, the valley: that is, the beautiful valley, in which Samarcand (anciently Maracanda) is situated. Bactria, or Bactriana, on the contrary, lay on the south of the Oxus; and comprehended the present provinces of Balk and Gaur; and probably part of Korasan. Mawer-ul-nere, is also applied to the country beyond the Jihon; and between the lower parts of the courses of that river, and the Sihon, or ancient

The tables were translated about the middle of the last century, by Dr. John Graves, of Oxford: and the History of Timur, by M. Petis de la Croix (the younger; not the compiler of the Life of Gengiz Cawn); early in the present century. The many references that I have made to the works, thus brought to our knowledge, by the well applied learning of these gentlemen; furnish the best eulogy in my power to bestow, both on the works themselves, and on the translators of them.

It is to be regretted, that Abulfeda's Persian geography, was not one of the subjects made choice of, by the learned gentlemen, who have favoured the world with translations of certain MSS. in the library of the King of France.

Jaxartes: Mawer-ul-nere signifying the country beyond the river; or Transoxiana.

The other misconception respects ancient Parthia. Very inaccurate ideas prevail, concerning the local situation of that country. Those, whose knowledge of it, is collected chiefly from its wars with the Romans, conceive Parthia to be only the country bordering on the Euphrates and Tigris; as the Parthian boundary, on the extension of their empire westward, met that of the Romans. Strabo has either been mistaken in this point, or has not fully expressed himself, where he describes the Parthians who defeated Crafsus, as the descendants of those Carduchians, who gave so much trouble to Xenophon, during the celebrated retreat of the Greeks. It is pro. bable, or at least possible, that the Parthians might have had in their army at that time, some detachments from among those hardy mountaineers; as the Carduchi were then numbered among their subjects: but the bulk of the Parthian army, came from Persia, their proper country. Whoever considers the slight subjection in which the Carduchians were held, even during the vigorous reigns of the first Persian emperors, will not expect that the Parthians had many recruits from that quarter. The history of the Parthian geography is briefly this: Parthia proper, was a small province, very near to the south-east extreme of the Caspian sea; which territory, after the division of Alexander's empire, fell to the share of the Seleucidæ, kings of Syria, and of the east; about 300 years before our æra. About 50 years after, Parthia rebelled; and together with Hyrcania, and other adjoining provinces, became an independent state, under Arsaces. As the empire of the Seleucida grew weaker, the Parthians extended their country westward; and the fine province of Media (now Irak-Ajami) fell to them and within a century after the foundation of their state, it had swallowed up all the countries from the Indus to the Euphrates, Bactria included: and this province had thrown off the yoke of the Seleucidæ, long before Parthia. The Parthian conquests