"In the first moon of 1281 some gold, silver, and bank-notes were handed to Puh-lo for the relief of the poor. In the second moon of 1282, just before the assassination of Achmed, the words 'Puh-lo the Minister' (ch'eng-siang) are used in connection with a case of fraud. In the seventh moon of 1282 (after the fall of Achmed) the 'Mongol man Puh-lo' was placed in charge of some gold-washings in certain towers of the then Hu Peh (now in Hu Nan). In the ninth moon of the same year a commission was sent to take official possession of all the gold-yielding places in Yuen Nan, and Puh-lo was appointed darugachi (= governor) of the mines. In this case it is not explicitly stated (though it would appear most likely) that the two gold superintendents were the same man; if they were, then neither could have been Marco, who certainly was no 'Mongol man.' Otherwise there would be a great temptation to identify this event with the mission to 'una citta, detta Carazan' of the Ramusio Text.

"There is, however, one man who may possibly be Marco, and that is the Poh-lo who was probably with Kublai at Chagan Nor when the news of Achmed's murder by Wang Chu arrived there in the third moon of 1282. The Emperor at once left for Shang-tu (i.e. K'ai-p'ing Fu, north of Dolonor), and 'ordered the shu-mih fu-shi Poh-lo [with two other statesmen] to proceed with all speed to Ta-tu (i.e. to Cambalu). On receiving Poh-lo's report, the Emperor became convinced of the deceptions practised upon him by Achmed, and said: "It was a good thing that Wang Chu did kill him."' In 1284 Achmed's successor is stated (chap, 209, p. 9-1/2) to have recommended Poh-lo, amongst others, for minor Treasury posts. The same man (chap. 209, p. 12-1/2) subsequently got Poh-lo appointed to a salt superintendency in the provinces; and as Yang-chou is the centre of the salt trade, it is just possible that Marco's 'governorship' of that place may resolve itself into this.

"There are many other Puh-lo and Poh-lo mentioned, both before Marco's arrival in, and subsequently to Marco's departure in 1292 from, China. In several cases (as, for instance, in that of P. Timur) both forms occur in different chapters for the same man; and a certain Tartar called 'Puh-lan Hi' is also called 'Puh-lo Hi.' One of Genghis Khan's younger brothers was called Puh-lo Kadei. There was, moreover, a Cathayan named Puh-lo, and a Naiman Prince Poh-lo. Whether 'Puh-lo the Premier' or 'one of the Ministers,' mentioned in 1282, is the same person as 'Poh-lo the ts'an cheng,' or 'Prime Minister's assistant' of 1284, I cannot say. Perhaps, when the whole Yuean Shi has been thoroughly searched throughout in all its editions, we may obtain more certain information. Meanwhile, one thing is plain: Pauthier is wrong, Yule is wrong in that particular connection; and M. Cordier gives us no positive view of his own. The other possibilities are given above, but I scarcely regard any of them as probabilities. On p. 99 of his Introduction, Colonel Yule manifestly identifies the Poh-lo of 1282 with Marco; but the identity of his title with that of Puh-lo in 1277 suggests that the two men are one, in which case neither can be Marco Polo. On p. 422 of Vol. I. Yule repeats this identification in his notes. I may mention that much of the information given in the present article was published in Vol. XXIV. of the China Review two or three years ago. I notice that M. Cordier quotes that volume in connection with other matters, but this particular point does not appear to have caught his eye.

"As matters now stand, there is a fairly strong presumption that Marco Polo is once named in the Annals; but there is no irrefragable evidence; and in any case it is only this once, and not as Pauthier has it."

Cf. also note by Prof. E.H. Parker, China Review, XXV. pp. 193-4, and, according to Prof. Pelliot (Bul. Ecole franc. Ext. Orient, July-Sept., 1904, p. 769), the biography of Han Lin-eul in the Ming shi, k. 122, p. 3.

Prof. Pelliot writes to me: "Il faut renoncer une bonne fois a retrouver Marco Polo dans le Po-lo mele a l'affaire d'Ahmed. Grace aux titulations successives, nous pouvons reconstituer la carriere administrative de ce Po-lo, au moins depuis 1271, c'est-a-dire depuis une date anterieure a l'arrivee de Marco Polo a la cour mongole. D'autre part, Rashid-ud-Din mentionne le role joue dans l'affaire d'Ahmed par le Pulad-aqa, c'est-a-dire Pulad Chinsang, son informateur dans les choses mongoles, mais la forme mongole de ce nom de Pulad est Bolod, en transcription chinoise Po-lo. J'ai signale (T'oung Pao, 1914, p. 640) que des textes chinois mentionnent effectivement que Po-lo (Bolod), envoye en mission aupres d'Arghun en 1285, resta ensuite en Perse. C'est donc en definitive le Pulad (= Bolod) de Rashid-ud-Din qui serait le Po-lo qu'a la suite de Pauthier on a trop longtemps identifie a Marco Polo."

Introduction, p. 23.

"The Yuean Shi contains curious confirmation of the facts which led up to Marco Polo's conducting a wife to Arghun of Persia, who lost his spouse in 1286. In the eleventh moon of that year (say January, 1287) the following laconic announcement appears: 'T'a-ch'a-r Hu-nan ordered to go on a mission to A-r-hun.' It is possible that Tachar and Hunan may be two individuals, and, though they probably started overland, it is probable that they were in some way connected with Polo's first and unsuccessful attempt to take the girl to Persia." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 136.)

Introduction, p. 76 n.

With regard to the statue of the Pseudo-Marco Polo of Canton, Dr. B. Laufer, of Chicago, sends me the following valuable note:—

THE ALLEGED MARCO POLO LO-HAN OF CANTON.

The temple Hua lin se (in Cantonese Fa lum se, i.e. Temple of the Flowery Grove) is situated in the western suburbs of the city of Canton. Its principal attraction is the vast hall, the Lo-han t'ang, in which are arranged in numerous avenues some five hundred richly gilded images, about three feet in height, representing the 500 Lo-han (Arhat). The workmanship displayed in the manufacture of these figures, made of fine clay thickly covered with burnished gilding, is said to be most artistic, and the variety of types is especially noticeable. In this group we meet a statue credited with a European influence. Two opinions are current regarding this statue: one refers to it as representing the image of a Portuguese sailor, the other sees in it a portrait of Marco Polo.

The former view is expressed, as far as I see, for the first time, by MAYERS and DENNYS (The Treaty Ports of China and Japan, London and Hong Kong, 1867, p. 162). "One effigy," these authors remark, "whose features are strongly European in type, will be pointed out as the image of a Portuguese seaman who was wrecked, centuries ago, on the coast, and whose virtues during a long residence gained him canonization after death. This is probably a pure myth, growing from an accidental resemblance of the features." This interpretation of a homage rendered to a Portuguese is repeated by C.A. MONTALTO DE JESUS, Historic Macao (Hong Kong, 1902, p. 28). A still more positive judgment on this matter is passed by MADROLLE (Chine du Sud et de l'Est, Paris, 1904, p. 17). "The attitudes of the Venerable Ones," he says, "are remarkable for their life-like expression, or sometimes, singularly grotesque. One of these personalities placed on the right side of a great altar wears the costume of the 16th century, and we might be inclined to regard it as a Chinese representation of Marco Polo. It is probable, however, that the artist, who had to execute the statue of a Hindu, that is, of a man of the West, adopted as the model of his costume that of the Portuguese who visited Canton since the commencement of the 16th century." It seems to be rather doubtful whether the 500 Lo-han of Canton are really traceable to that time. There is hardly any huge clay statue in China a hundred or two hundred years old, and all the older ones are in a state of decay, owing to the brittleness of the material and the carelessness of the monks. Besides, as stated by Mayers and Dennys (l.c., p. 163), the Lo-han Hall of Canton, with its glittering contents, is a purely modern structure, having been added to the Fa-lum Temple in 1846, by means of a subscription mainly supported by the Hong Merchants. Although this statue is not old, yet it may have been made after an ancient model. Archdeacon Gray, in his remarkable and interesting book, Walks in the City of Canton (Hong Kong, 1875, p. 207), justly criticized the Marco Polo theory, and simultaneously gave a correct identification of the Lo-han in question. His statement is as follows: "Of the idols of the five hundred disciples of Buddha, which, in this hall, are contained, there is one, which, in dress and configuration of countenance, is said to resemble a foreigner. With regard to this image, one writer, if we mistake not, has stated that it is a statue of the celebrated traveller Marco Polo, who, in the thirteenth century, visited, and, for some time, resided in the flowery land of China. This statement, on the part of the writer to whom we refer, is altogether untenable. Moreover, it is an error so glaring as to cast, in the estimation of all careful readers of his work, no ordinary degree of discredit upon many of his most positive assertions. The person, whose idol is so rashly described as being that of Marco Polo, was named Shien-Tchu. He was a native of one of the northern provinces of India, and, for his zeal as an apostle in the service of Buddha, was highly renowned."

Everard Cotes closes the final chapter of his book, The Arising East (New York, 1907), as follows: "In the heart of Canton, within easy reach of mob violence at any time, may be seen to-day the life-size statue of an elderly European, in gilt clothes and black hat, which the Chinese have cared for and preserved from generation to generation because the original, Marco Polo, was a friend to their race. The thirteenth-century European had no monopoly of ability to make himself loved and reverenced. A position similar to that which he won as an individual is open to-day to the Anglo-Saxon as a race. But the Mongolian was not afraid of Marco Polo, and he is afraid of us. It can be attained, therefore, only by fair dealing and sympathy, supported by an overwhelming preponderance of fighting strength."

[Dr. Laufer reproduces here the note in Marco Polo, I., p. 76. I may remark that I never said nor believed that the statue was Polo's. The mosaic at Genoa is a fancy portrait.]

The question may be raised, however, Are there any traces of foreign influence displayed in this statue? The only way of solving this problem seemed to me the following: First to determine the number and the name of the alleged Marco Polo Lo-han at Canton, and then by means of this number to trace him in the series of pictures of the traditional 500 Lo-han (the so-called Lo han t'u).

The alleged Marco Polo Lo-han bears the number 100, and his name is Shan-chu tsun-che (tsun-che being a translation of Sanskrit arya, "holy, reverend"). The name Shan-chu evidently represents the rendering of a Sanskrit name, and does not suggest a European name. The illustration here reproduced is Lo-han No. 100 from a series of stone-engravings in the temple T'ien-ning on the West Lake near Hang Chau. It will be noticed that it agrees very well with the statue figured by M. Cordier. In every respect it bears the features of an Indian Lo-han, with one exception, and this is the curious hat. This, in fact, is the only Lo-han among the five hundred that is equipped with a headgear; and the hat, as is well known, is not found in India. This hat must represent a more or less arbitrary addition of the Chinese artist who created the group, and it is this hat which led to the speculations regarding the Portuguese sailor or Marco Polo. Certain it is also that such a type of hat does not occur in China; but it seems idle to speculate as to its origin, as long as we have no positive information on the intentions of the artist. The striped mantle of the Lo-han is by no means singular, for it occurs with seventeen others. The facts simply amount to this, that the figure in question does not represent a Portuguese sailor or Marco Polo or any other European, but solely an Indian Lo-han (Arhat), while the peculiar hat remains to be explained.

Introduction, p. 92.

THIBAUT DE CHEPOY.

Thibaut de Chepoy (Chepoy, canton of Breteuil, Oise), son of the knight Jean de Chepoy, was one of the chief captains of King Philip the Fair. He entered the king's service in 1285 as squire and valet; went subsequently to Robert d'Artois, who placed him in charge of the castle of Saint Omer, and took him, in 1296, to Gascony to fight the English. He was afterwards grand master of the cross-bow men. He then entered the service of Charles de Valois, brother of Philip the Fair, who sent him to Constantinople to support the claims to the throne of his wife, Catherine of Courtenay. Thibaut left Paris on the 9th Sept., 1306, passed through Venice, where he met Marco Polo who gave him a copy of his manuscript. Thibaut died between 22nd May, 1311, and 22nd March, 1312. (See Joseph PETIT, in Le Moyen Age, Paris, 1897, pp. 224-239.)



THE BOOK OF MARCO POLO.



PROLOGUE

II., p. 6.

SARAI.

"Cordier (Yule) identifiziert den von Pegolotti gewaehlten Namen Saeracanco mit dem juengeren Sarai oder Zarew (dem Sarai grande Fra Mauros), was mir vollkommen untunlich erscheint; es waere dann die Route des Reisenden geradezu ein Zickzackweg gewesen, der durch nichts zu rechtfertigen waere." (Dr. Ed. FRIEDMANN, Pegolotti, p. 14.)

Prof. Pelliot writes to me: "Il n'y a aucune possibilite de retrouver dans Saracanco, Sarai + Kunk. Le mot Kunk n'est pas autrement atteste, et la construction mongole ou turque exigerait kunk-sarai."

XIII., pp. 25-26.

SHANG TU.

See also A. POZDNEIEV, Mongoliya i Mongoly, II., pp. 303 seq.

XV., pp. 27, 28-30. Now it came that Marco, the son of Messer Nicolo, sped wondrously in learning the customs of the Tartars, as well as their language, their manner of writing, and their practice of war—in fact he came in a brief space to know several languages, and four sundry written characters.

On the linguistic office called Sse yi kwan, cf. an interesting note by H. MASPERO, p. 8, of Bul. Ecole franc. Ext. Orient, XII., No. 1, 1912.

XV., p. 28 n. Of the Khitan but one inscription was known and no key.

Prof. Pelliot remarks, Bul. Ecole franc. Ext. Orient, IV., July-Sept., 1904: "In fact a Chinese work has preserved but five k'i-tan characters, however with the Chinese translation." He writes to me that we do not know any k'itan inscription, but half a dozen characters reproduced in a work of the second half of the fourteenth century. The Uighur alphabet is of Aramean origin through Sogdian; from this point of view, it is not necessary to call for Estranghelo, nor Nestorian propaganda. On the other hand we have to-day documents in Uighur writing older than the Kudatku Bilik.



BOOK FIRST.

ACCOUNT OF REGIONS VISITED OR HEARD OF ON THE JOURNEY FROM THE LESSER ARMENIA TO THE COURT OF THE GREAT KAAN AT CHANDU.



BOOK I.

VI., p. 63. "There is also on the river, as you go from Baudas to Kisi, a great city called Bastra, surrounded by woods, in which grow the best dates in the world."

"The products of the country are camels, sheep and dates." (At Pi-ssi-lo, Basra. CHAU JU-KWA, p. 137.)

VI., pp. 63, 65. "In Baudas they weave many different kinds of silk stuffs and gold brocades, such as nasich, and nac, and cramoisy, and many other beautiful tissue richly wrought with figures of beasts and birds."

In the French text we have nassit and nac.

"S'il faut en croire M. Defremery, au lieu de nassit, il faut evidemment lire nassij (necidj), ce qui signifie un tissu, en general, et designe particulierement une etoffe de soie de la meme espece que le nekh. Quant aux etoffes sur lesquelles etaient figures des animaux et des oiseaux, le meme orientaliste croit qu'il faut y reconnaitre le thardwehch, sorte d'etoffe de soie qui, comme son nom l'indique, representait des scenes de chasse. On sait que l'usage de ces representations est tres ancien en Orient, comme on le voit dans des passages de Philostrate et de Quinte-Curce rapportes par Mongez." (FRANCISQUE-MICHEL, Recherches sur le Commerce, I., p. 262.)

VI., p. 67.

DEATH OF MOSTAS'IM.

According to Al-Fakhri, translated by E. Amar (Archives marocaines XVI., p. 579), Mostas'im was put to death with his two eldest sons on the 4th of safar, 656 (3rd February, 1258).

XI., p. 75. "The [the men of Tauris] weave many kinds of beautiful and valuable stuffs of silk and gold."

Francisque-Michel (I., p. 316) remarks: "De ce que Marco Polo se borne a nommer Tauris comme la ville de Perse ou il se fabriquait maints draps d'or et de soie, il ne faudrait pas en conclure que cette industrie n'existat pas sur d'autres points du meme royaume. Pour n'en citer qu'un seul, la ville d'Arsacie, ancienne capitale des Parthes, connue aujourd'hui sous le nom de Caswin, possedait vraisemblablement deja cette industrie des beaux draps d'or et de soie qui existait encore au temps de Huet, c'est-a-dire au XVII'e siecle."

XIII., p. 78. "Messer Marco Polo found a village there which goes by the name of CALA ATAPERISTAN, which is as much as to say, 'The Castle of the Fire-worshippers.'"

With regard to Kal'ah-i Atashparastan, Prof. A.V.W. Jackson writes (Persia, 1906, p. 413): "And the name is rightly applied, for the people there do worship fire. In an article entitled The Magi in Marco Polo (Journ. Am. Or. Soc., 26, 79-83) I have given various reasons for identifying the so-called 'Castle of the Fire-Worshippers' with Kashan, which Odoric mentions or a village in its vicinity, the only rival to the claim being the town of Nain, whose Gabar Castle has already been mentioned above."

XIV., p. 78.

PERSIA.

Speaking of Saba and of Cala Ataperistan, Prof. E.H. Parker (Asiatic Quart. Rev., Jan., 1904, p. 134) has the following remarks: "It is not impossible that certain unexplained statements in the Chinese records may shed light upon this obscure subject. In describing the Arab Conquest of Persia, the Old and New T'ang Histories mention the city of Hia-lah as being amongst those captured; another name for it was Sam (according to the Chinese initial and final system of spelling words). A later Chinese poet has left the following curious line on record: 'All the priests venerate Hia-lah.' The allusion is vague and undated, but it is difficult to imagine to what else it can refer. The term seng, or 'bonze,' here translated 'priests,' was frequently applied to Nestorian and Persian priests, as in this case."

XIV., p. 80. "Three Kings."

Regarding the legend of the stone cast into a well, cf. F.W.K. MUELLER, Uigurica, pp. 5-10 (Pelliot).

XVII., p. 90. "There are also plenty of veins of steel and Ondanique."

"The ondanique which Marco Polo mentions in his 42nd chapter is almost certainly the pin t'ieh or 'pin iron' of the Chinese, who frequently mention it as coming from Arabia, Persia, Cophene, Hami, Ouigour-land and other High Asia States." (E.H. PARKER, Journ. North China Br. Roy. Asiatic Soc., XXXVIII., 1907, p. 225.)

XVIII., pp. 97, 100. "The province that we now enter is called REOBARLES.... The beasts also are peculiar.... Then there are sheep here as big as asses; and their tails are so large and fat, that one tail shall weight some 30 lbs. They are fine fat beasts, and afford capital mutton."

Prof. E.H. PARKER writes in the Journ. of the North China Branch of the Royal Asiatic Soc., XXXVII., 1906, p. 196: "Touching the fat-tailed sheep of Persia, the Shan-hai-king says the Yueh-chi or Indo-Scythy had a 'big-tailed sheep' the correct name for which is hien-yang. The Sung History mentions sheep at Hami with tails so heavy that they could not walk. In the year 1010 some were sent as tribute to China by the King of Kuche."

"Among the native products [at Mu lan p'i, Murabit, Southern Coast of Spain] are foreign sheep, which are several feet high and have tails as big as a fan. In the spring-time they slit open their bellies and take out some tens of catties of fat, after which they sew them up again, and the sheep live on; if the fat were not removed, (the animal) would swell up and die." (CHAU JU-KWA, pp. 142-3.)

"The Chinese of the T'ang period had heard also of the trucks put under these sheep's tails. 'The Ta-shi have a foreign breed of sheep (hu-yang) whose tails, covered with fine wool, weigh from ten to twenty catties; the people have to put carts under them to hold them up. Fan-kuo-chi as quoted in Tung-si-yang-k'au." (HIRTH and ROCKHILL, p. 143.)

Leo Africanus, Historie of Africa, III., 945 (Hakluyt Soc. ed.), says he saw in Egypt a ram with a tail weighing eighty pounds!:

OF THE AFRICAN RAMME.

"There is no difference betweene these rammes of Africa and others, saue onely in their tailes, which are of a great thicknes, being by so much the grosser, but how much they are more fatte, so that some of their tailes waigh tenne, and other twentie pounds a peece, and they become fatte of their owne naturall inclination: but in Egypt there are diuers that feede them fatte with bran and barly, vntill their tailes growe so bigge that they cannot remooue themselves from place to place: insomuch that those which take charge of them are faine to binde little carts vnder their tailes, to the end they may haue strength to walke. I my selfe saw at a citie in Egypt called Asiot, and standing vpon Nilus, about an hundred and fiftie miles from Cairo, one of the saide rams tailes that weighed fowerscore pounds, and others affirmed that they had seene one of those tailes of an hundred and fiftie pounds weight. All the fatte therefore of this beast consisteth in his taile; neither is there any of them to be founde but onely in Tunis and in Egypt." (LEO AFRICANUS, edited by Dr. Robert BROWN, III., 1896, Hakluyt Society, p. 945.)

XVIII., pp. 97, 100 n.

Dr. B. Laufer draws my attention to what is probably the oldest mention of this sheep from Arabia, in Herodotus, Book III., Chap. 113:

"Concerning the spices of Arabia let no more be said. The whole country is scented with them, and exhales an odour marvellously sweet. There are also in Arabia two kinds of sheep worthy of admiration, the like of which is nowhere else to be seen; the one kind has long tails, not less than three cubits in length, which, if they were allowed to trail on the ground, would be bruised and fall into sores. As it is, all the shepherds know enough of carpentering to make little trucks for their sheep's tails. The trucks are placed under the tails, each sheep having one to himself, and the tails are then tied down upon them. The other kind has a broad tail, which is a cubit across sometimes."

Canon G. Rawlinson, in his edition of Herodotus, has the following note on this subject (II., p. 500):—

"Sheep of this character have acquired among our writers the name of Cape Sheep, from the fact that they are the species chiefly affected by our settlers at the Cape of Good Hope. They are common in Africa and throughout the East, being found not only in Arabia, but in Persia, Syria, Affghanistan, Egypt, Barbary, and even Asia Minor. A recent traveller, writing from Smyrna, says: 'The sheep of the country are the Cape sheep, having a kind of apron tail, entirely of rich marrowy fat, extending to the width of their hind quarters, and frequently trailing on the ground; the weight of the tail is often more than six or eight pounds' (FELLOWS'S Asia Minor, p. 10). Leo Africanus, writing in the 15th century, regards the broad tail as the great difference between the sheep of Africa and that of Europe. He declares that one which he had seen in Egypt weighed 80 lbs. He also mentions the use of trucks which is still common in North Africa."

XVIII., p. 98. "Camadi.—Reobarles.—In this plain there are a number of villages and towns which have lofty walls of mud, made as a defence against the banditti, who are very numerous, and are called CARAONAS. This name is given them because they are the sons of Indian mothers by Tartar fathers."

Mirza Haidar writes (Tarikh-i-Rashidi, p. 148): "The learned Mirza Ulugh Beg has written a history which he has called Ulus Arbaa. One of the 'four hordes' is that of the Moghul, who are divided into two branches, the Moghul and the Chaghatai. But these two branches, on account of their mutual enmity, used to call each other by a special name, by way of depreciation. Thus the Chaghatai called the Moghul Jatah, while the Moghul called the Chaghatai Karawanas."

Cf. Ney ELIAS, l.c., pp. 76-77, and App. B, pp. 491-2, containing an inquiry made in Khorasan by Mr. Maula Bakhsh, Attache at the Meshed Consulate General, of the families of Karnas, he has heard or seen; he says: "These people speak Turki now, and are considered part of the Goklan Turkomans. They, however, say they are Chingiz-Khani Moghuls, and are no doubt the descendants of the same Karnas, or Karavanas, who took such a prominent part in the victories in Persia.

"The word Karnas, I was told by a learned Goklan Mullah, means Tirandaz, or Shikari (i.e. Archer or Hunter), and was applied to this tribe of Moghuls on account of their professional skill in shooting, which apparently secured them an important place in the army. In Turki the word Karnas means Shikamparast—literally, 'belly worshippers,' which implies avarice. This term is in use at present, and I was told, by a Kazi of Bujnurd, that it is sometimes used by way of reproach.... The Karnas people in Mana and Gurgan say it is the name of their tribe, and they can give no other explanation."

XVIII., pp. 98, 102, 165. "The King of these scoundrels is called NOGODAR."

Sir Aurel Stein has the following regarding the route taken by this Chief in Serindia, I., pp. 11-12:—

"To revert to an earlier period it is noteworthy that the route in Marco Polo's account, by which the Mongol partisan leader Nigudar, 'with a great body of horsemen, cruel unscrupulous fellows,' made his way from Badakhshan 'through another province called PASHAI-DIR, and then through another called ARIORA-KESHEMUR' to India, must have led down the Bashgol Valley. The name of Pashai clearly refers to the Kafirs among whom this tribal designation exists to this day, while the mention of Dir indicates the direction which this remarkable inroad had taken. That its further progress must have lain through Swat is made probable by the name which, in Marco Polo's account, precedes that of 'Keshemur' or Kashmir; for in the hitherto unexplained Ariora can be recognized, I believe, the present Agror, the name of the well-known hill-tract on the Hazara border which faces Buner from the left bank of the Indus. It is easy to see from any accurate map of these regions, that for a mobile column of horsemen forcing its way from Badakhshan to Kashmir, the route leading through the Bashgol Valley, Dir, Talash, Swat, Buner, Agror, and up the Jhelam Valley, would form at the present day, too, the most direct and practicable line of invasion."

In a paper on Marco Polo's Account of a Mongol inroad into Kashmir (Geog. Jour., August, 1919), Sir Aurel Stein reverts again to the same subject. "These [Mongol] inroads appear to have commenced from about 1260 A.D., and to have continued right through the reign of Ghiasuddin, Sultan of Delhi (1266-1286), whose identity with Marco's Asedin Soldan is certain. It appears very probable that Marco's story of Nogodar, the nephew of Chaghatai, relates to one of the earliest of these incursions which was recent history when the Poli passed through Persia about 1272-73 A.D."

Stein thinks, with Marsden and Yule, that Dilivar (pp. 99, 105) is really a misunderstanding of "Citta di Livar" for Lahawar or Lahore.

Dir has been dealt with by Yule and Pauthier, and we know that it is "the mountain tract at the head of the western branch of the Panjkora River, through which leads the most frequented route from Peshawar and the lower Swat valley to Chitral" (Stein, l.c.). Now with regard to the situation of Pashai (p. 104):

"It is clear that a safe identification of the territory intended cannot be based upon such characteristics of its people as Marco Polo's account here notes obviously from hearsay, but must reckon in the first place with the plainly stated bearing and distance. And Sir Henry Yule's difficulty arose just from the fact that what the information accessible to him seemed to show about the location of the name Pashai could not be satisfactorily reconciled with those plain topographical data. Marco's great commentator, thoroughly familiar as he was with whatever was known in his time about the geography of the western Hindukush and the regions between Oxus and Indus, could not fail to recognize the obvious connection between our Pashai and the tribal name Pashai borne by Muhammanized Kafirs who are repeatedly mentioned in mediaeval and modern accounts of Kabul territory. But all these accounts seemed to place the Pashais in the vicinity of the great Panjshir valley, north-east of Kabul, through which passes one of the best-known routes from the Afghan capital to the Hindukush watershed and thence to the Middle Oxus. Panjshir, like Kabul itself, lies to the south-west of Badakshan, and it is just this discrepancy of bearing together with one in the distance reckoned to Kashmir which caused Sir Henry Yule to give expression to doubts when summing up his views about Nogodar's route."

From Sir George Grierson's Linguistic Survey of India we learn that to the south of the range of the Hindukush "the languages spoken from Kashmir in the east to Kafiristan in the west are neither of Indian nor of Iranian origin, but form a third branch of the Aryan stock of the great Indo-European language family. Among the languages of this branch, now rightly designated as 'Dardic,' the Kafir group holds a very prominent place. In the Kafir group again we find the Pashai language spoken over a very considerable area. The map accompanying Sir George Grierson's monograph on 'The Pisaca Languages of North-Western India' [Asiatic Society Monographs, VIII., 1906], shows Pashai as the language spoken along the right bank of the Kunar river as far as the Asmar tract as well as in the side valleys which from the north descend towards it and the Kabul river further west. This important fact makes it certain that the tribal designation of Pashai, to which this Kafir language owes its name, has to this day an application extending much further east than was indicated by the references which travellers, mediaeval and modern, along the Panjshir route have made to the Pashais and from which alone this ethnic name was previously known."

Stein comes to the conclusion that "the Mongols' route led across the Mandal Pass into the great Kafir valley of Bashgol and thus down to Arnawai on the Kunar. Thence Dir could be gained directly across the Zakhanna Pass, a single day's march. There were alternative routes, too, available to the same destination either by ascending the Kunar to Ashreth and taking the present 'Chitral Road' across the Lowarai, or descending the river to Asmar and crossing the Binshi Pass."

From Dir towards Kashmir for a large body of horsemen "the easiest and in matter of time nearest route must have led them as now down the Panjkora Valley and beyond through the open tracts of Lower Swat and Buner to the Indus about Amb. From there it was easy through the open northern part of the present Hazara District (the ancient Urasa) to gain the valley of the Jhelam River at its sharp bend near Muzzaffarabad."

The name of Agror (the direct phonetic derivative of the Sanskrit Atyugrapura) = Ariora; it is the name of the hill-tract on the Hazara border which faces Buner on the east from across the left bank of the Indus.

XVIII., p. 101.

Line 17, Note 4. Korano of the Indo-Scythic Coins is to be read Kosano. (PELLIOT.)

XVIII., p. 102.

On the Mongols of Afghanistan, see RAMSTEDT, Mogholica, in Journ. de la Soc. Finno-Ougrienne, XXIII., 1905. (PELLIOT.)

XIX., p. 107. "The King is called RUOMEDAN AHOMET."

About 1060, Mohammed I. Dirhem Kub, from Yemen, became master of Hormuz, but his successors remained in the dependency of the sovereigns of Kerman until 1249, when Rokn ed-Din Mahmud III. Kalhaty (1242-1277) became independent. His successors in Polo's time were Seif ed-Din Nusrat (1277-1290), Mas'ud (1290-1293), Beha ed-Din Ayaz Seyfin (1293-1311).

XIX., p. 115.

HORMOS.

The Travels of Pedro Teixeira, a Portuguese traveller, probably of Jewish origin, certainly not a Jesuit, have been published by the Hakluyt Society:

The Travels of Pedro Teixeira; with his "Kings of Harmuz," and extracts from his "King of Persia." Translated and annotated by William F. Sinclair, Bombay Civil Service (Rtd.); With further Notes and an Introduction by Donald Ferguson, London: Printed for the Hakluyt Society, MDCCCCII, 8 vo. pp. cvii-292.

See Appendix A. A Short Narrative of the Origin of the Kingdom of Harmusz, and of its Kings, down to its Conquest by the Portuguese; extracted from its History, written by Torunxa, King of the Same, pp. 153-195. App. D. Relation of the Chronicle of the Kings of Ormuz, taken from a Chronicle composed by a King of the same Kingdom, named Pachaturunza, written in Arabic, and summarily translated into the Portuguese language by a friar of the order of Saint Dominick, who founded in the island of Ormuz a house of his order, pp. 256-267.

See Yule, Hobson-Jobson, s.v. Ormus.

Mr. Donald Ferguson, in a note, p. 155, says: "No dates are given in connection with the first eleven rulers of Hormuz; but assuming as correct the date (1278) given for the death of the twelfth, and allowing to each of his predecessors an average reign of thirteen years, the foundation of the kingdom of Hormuz would fall in A.D. 1100. Yule places the founding somewhat earlier; and Valentyn, on what authority I know not, gives A.D. 700 as the date of the founder Muhammad."

XIX., I., p. 116; II., p. 444.

DIET OF THE GULF PEOPLE.

Prof. E.H. Parker says that the T'ang History, in treating of the Arab conquests of Fuh-lin [or Frank] territory, alludes to the "date and dry fish diet of the Gulf people." The exact Chinese words are: "They feed their horses on dried fish, and themselves subsist on the hu-mang, or Persian date, as Bretschneider has explained." (Asiatic Quart. Rev., Jan., 1904, p. 134.)

Bretschneider, in Med. Researches, II., p. 134, n. 873, with regard to the dates writes: "Wan nien tsao, 'ten thousand years' jujubes'; called also Po-sze tao, or 'Persian jujubes.' These names and others were applied since the time of the T'ang dynasty to the dates brought from Persia. The author of the Pen ts'ao kang mu (end of the sixteenth century) states that this fruit is called k'u-lu-ma in Persia. The Persian name of the date is khurma."

Cf. CHAU JU-KWA, p. 210.

XXII., p. 128 n.

TUN-O-KAIN.

Major Sykes had adopted Sir Henry Yule's theory of the route from Kuh-benan to Tun. He has since altered his opinion in the Geographical Journal, October, 1905, p. 465: "I was under the impression that a route ran direct from Kubunan to Tabas, but when visiting this latter town a few months ago I made careful inquiries on the subject, which elicited the fact that this was not the case, and that the route invariably followed by Kubunan-Tabas caravans joined the Kerman-Ravar-Naiband route at Chah-Kuru, 12 miles south of Darbana. It follows this track as far as Naiband, whence the route to Tabas branches off; but the main caravan route runs via Zenagan and Duhuk to Tun. This new information, I would urge, makes it almost certain that Ser Marco travelled to Tun, as Tabas falls to the west of the main route. Another point is that the district of Tabas only grows four months' supplies, and is, in consequence, generally avoided by caravans owing to its dearness.

"In 1893 I travelled from Tun to the south across the Lut as far as Chah Kuru by this very route, and can testify to the general accuracy of Ser Marco's description,[1] although there are now villages at various points on the way. Finally, as our traveller especially mentions Tonocain, or Tun va Kain, one is inclined to accept this as evidence of first-rate importance, especially as it is now corroborated by the information I gained at Tabas. The whole question, once again, furnishes an example of how very difficult it is to make satisfactory inquiries, except on the spot."

It was also the opinion (1882) of Colonel C.E. Stewart, who says: "I was much interested in hearing of Kuh Banan, as it is one of the places mentioned by Marco Polo as on his route. Kuh Banan is described as a group of villages about 26 miles from the town of Rawar, in the Karman district. I cannot help thinking the road travelled by Marco Polo from Karman to Kain is the one by Naiband. Marco Polo speaks of Tun-o-Cain, which, Colonel Yule has pointed out, undoubtedly means Tun and Kain. At present Tun does not belong to the Kain district, but to the Tabbas district, and is always spoken of as Tun-o-Tabbas; and if it belonged, as I believe it formerly did, to the Kain district, it would be spoken of as Tun-o-Kain, exactly as Marco Polo does. Through Naiband is the shortest and best road to either Tun or Kain." (Proc. Royal Geog. Soc., VIII., 1886, p. 144.)

Support to Yule's theory has been brought by Sven Hedin, who devotes a chapter to Marco Polo in his Overland to India, II., 1910, Chap. XL., and discusses our traveller's route between Kuh-benan and Tabbas, pp. 71 seq.:

"As even Sykes, who travelled during several years through Persia in all directions, cannot decide with full certainty whether Marco Polo travelled by the western route through Tebbes or the eastern through Naibend, it is easy to see how difficult it is to choose between the two roads. I cannot cite the reasons Sir Henry Yule brings forward in favour of the western route—it would take us too far. I will, instead, set forth the grounds of my own conviction that Marco Polo used the direct caravan road between Kuh-benan and Tebbes.

"The circumstance that the main road runs through Naibend is no proof, for we find that Marco Polo, not only in Persia but also in Central Asia, exhibited a sovereign contempt for all routes that might be called convenient and secure.

"The distance between Kerman and Kuh-benan in a direct line amounts to 103 miles. Marco Polo travelled over this stretch in seven days, or barely 15 miles a day. From Kuh-benan to Tebbes the distance is 150 miles, or fully 18 miles a day for eight days. From Kuh-benan via Naibend to Tun, the distance is, on the other hand, 205 miles, or more than 25 miles a day. In either case we can perceive from the forced marches that after leaving Kuh-benan he came out into a country where the distances between the wells became much greater.

"If he travelled by the eastern route he must have made much longer day's journeys than on the western. On the eastern route the distances between the wells were greater. Major Sykes has himself travelled this way, and from his detailed description we get the impression that it presented particular difficulties. With a horse it is no great feat to ride 25 miles a day for eight days, but it cannot be done with camels. That I rode 42-1/2 miles a day between Hauz-i-Haji-Ramazan and Sadfe was because of the danger from rain in the Kevir, and to continue such a forced march for more than two days is scarcely conceivable. Undoubtedly Marco Polo used camels on his long journeys in Eastern Persia, and even if he had been able to cover 205 miles in eight days, he would not be obliged to do so, for on the main road through Naibend and Duhuk to Tun there are abundant opportunities of procuring water. Had he travelled through Naibend, he would in any case have had no need to hurry on so fast. He would probably keep to the same pace as on the way from Kerman to Kuh-benan, and this length he accomplished in seven days. Why should he have made the journey from Kuh-benan to Tun, which is exactly double as far, in only eight days instead of fourteen, when there was no necessity? And that he actually travelled between Kuh-benan and Tunocain in eight days is evident, because he mentions this number twice.

"He also says explicitly that during these eight days neither fruits nor trees are to be seen, and that you have to carry both food and water. This description is not true of the Naibend route, for in Naibend there are excellent water, fine dates, and other fruits. Then there is Duhuk, which, according to Sykes, is a very important village with an old fort and about 200 houses. After leaving Duhuk for the south, Sykes says: 'We continued our journey, and were delighted to hear that at the next stage, too, there was a village, proving that this section of the Lut is really quite thickly populated.' [Ten Thousand Miles in Persia, p. 35.] This does not agree at all with Marco Polo's description.

"I therefore consider it more probable that Marco Polo, as Sir Henry Yule supposes, travelled either direct to Tebbes, or perhaps made a trifling detour to the west, through the moderate-sized village Bahabad, for from this village a direct caravan road runs to Tebbes, entirely through desert. Marco Polo would then travel 150 miles in eight days compared with 103 miles in seven days between Kerman and Kuh-benan. He therefore increased his speed by only 4 miles a day, and that is all necessary on the route in question.

"Bahabad lies at a distance of 36 miles from Kubenan—all in a straight line. And not till beyond Bahabad does the real desert begin.

"To show that a caravan road actually connects Tebbes with Bahabad, I have inserted in the first and second columns of the following table the data I obtained in Tebbes and Fahanunch, and in the third the names marked on the 'Map of Persia (in six sheets) compiled in the Simla Drawing Office of the Survey of India, 1897.'

From Tebbes to Bahabad From Fahanunch to Bahabad 1. Kurit . . . . . . . 4 2. Moghu . . . . . . . . 4-1/2 2. Moghu . . . . . . . 9 3. Sefid-ab . . . . . . 6 3. Sefid-ab . . . . . 6 4. Belucha . . . . . . . 5 4. Burch . . . . . . . 5 5. God-i-shah-taghi . . 6 5. God . . . . . . . . 5 6. Rizab . . . . . . . . 5 6. Rizab . . . . . . . 6 7.{Teng-i-Tebbes . . . . 4-1/2 7. Pudenum . . . . . . 8 {Pudenun . . . . . . . 4-1/2 8. Ser-i-julge . . . . 4 8. Kheirabad . . . . . . 4 9. Bahabad . . . . . . 4 9. Bahabad . . . . . . . 4 Farsakh . . . . . 51 Farsakh . . . . . . 43-1/2

Map of Persia. 2. Maga . . . . . . . Salt well. 3. Chashma Sufid . . " " 4.{Khudafrin . . . . Sweet spring. {Pir Moral . . . . Salt well. 5. God Hashtaki . . . " " 6. Rezu . . . . . . . " "

"These details are drawn from different authorities, but are in excellent agreement. That the total distances are different in the first two columns is because Fahanunch lies nearer than Tebbes to Bahabad. Two or three discrepancies in the names are of no importance. Burch denotes a castle or fort; Belucha is evidently Cha-i-beluch or the well of the Baluchi, and it is very probable that a small fort was built some time or other at this well which was visited by raiders from Baluchistan. Ser-i-julge and Kheirabad may be two distinct camping grounds very near each other. The Chasma Sufid or 'white spring' of the English map is evidently the same place as Sefid-ab, or 'white water.' Its God Hashtaki is a corruption of the Persian God-i-shah-taghi, or the 'hollow of the royal saxaul.' Khudafrin, on the other hand, is very apocryphal. It is no doubt Khuda-aferin or 'God be praised!'—an ejaculation very appropriate in the mouth of a man who comes upon a sweet spring in the midst of the desert. If an Englishman travelled this way he might have mistaken this ejaculation for the name of the place. But then 'Unsurveyed' would hardly be placed just in this part of the Bahabad Desert.

"The information I obtained about the road from Tebbes to Bahabad was certainly very scanty, but also of great interest. Immediately beyond Kurit the road crosses a strip of the Kevir, 2 farsakh broad, and containing a river-bed which is said to be filled with water at the end of February. Sefid-ab is situated among hillocks and Burch in an upland district; to the south of it follows Kevir barely a farsakh broad, which may be avoided by a circuitous path. At God-i-shah-taghi, as the name implies, saxaul grows (Haloxylon Ammodendron). The last three halting-places before Bahabad all lie among small hills.

"This desert route runs, then, through comparatively hilly country, crosses two small Kevir depressions, or offshoots of one and the same Kevir, has pasturage at at least one place, and presents no difficulties of any account. The distance in a direct line is 113 miles, corresponding to 51 Persian farsakh—the farsakh in this district being only about 2.2 miles long against 2.9 in the great Kevir. The caravans which go through the Bahabad desert usually make the journey in ten days, one at least of which is a rest day, so that they cover little more than 12 miles a day. If water more or less salt were not to be found at all the eight camping-grounds, the caravans would not be able to make such short marches. It is also quite possible that sweet water is to be found in one place; where saxaul grows driftsand usually occurs, and wells digged in sand are usually sweet.

"During my stay in Tebbes a caravan of about 300 camels, as I have mentioned before, arrived from Sebsevar. They were laden with naft (petroleum), and remained waiting till the first belt of Kevir was dried after the last rain. As soon as this happened the caravan would take the road described above to Bahabad, and thence to Yezd. And this caravan route, Sebsevar, Turshiz, Bajistan, Tun, Tebbes, Bahabad, and Yezd, is considered less risky than the somewhat shorter way through the great Kevir. I myself crossed a part of the Bahabad desert where we did not once follow any of the roads used by caravans, and I found this country by no means one of the worst in Eastern Persia.

"In the above exposition I believe that I have demonstrated that it is extremely probable that Marco Polo travelled, not through Naibend to Tun, but through Bahabad to Tebbes, and thence to Tun and Kain. His own description accords in all respects with the present aspect and peculiarities of the desert route in question. And the time of eight days he assigns to the journey between Kuh-benan and Tonocain renders it also probable that he came to the last-named province at Tebbes, even if he travelled somewhat faster than caravans are wont to do at the present day. It signifies little that he does not mention the name Tebbes; he gives only the name of the province, adding that it contains a great many towns and villages. One of these was Tebbes."

XXII., p. 126.

TUTIA.

"It seems that the word is 'the Arabicized word dudha, being Persian for "smokes."' There can be little doubt that we have direct confirmation of this in the Chinese words t'ou-t'ieh (still, I think, in use) and t'ou-shik, meaning 'tou-iron' and 't'ou-ore.' The character T'ou [Chinese] does not appear in the old dictionaries; its first appearance is in the History of the Toba (Tungusic) Dynasty of North China. This History first mentions the name 'Persia' in A.D. 455 and the existence there of this metal, which, a little later on, is also said to come from a State in the Cashmeer region. K'ang-hi's seventeenth-century dictionary is more explicit: it states that Termed produces this ore, but that 'the true sort comes from Persia, and looks like gold, but on being heated it turns carnation, and not black.' As the Toba Emperors added 1000 new characters to the Chinese stock, we may assume this one to have been invented, for the specific purpose indicated.'" (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, pp. 135-6.) Prof. Parker adds the following note, l.c., p. 149: "Since writing the above, I have come across a passage in the 'History of the Sung Dynasty' (chap. 490, p. 17) stating that an Arab junk-master brought to Canton in A.D. 990, and sent on thence to the Chinese Emperor in Ho Nan, 'one vitreous bottle of tutia.' The two words mean 'metropolis-father,' and are therefore without any signification, except as a foreign word. According to Yule's notes (I., p. 126), tutia, or dudha, in one of its forms was used as an eye-ointment or collyrium."

XXII., pp. 127-139. The Province of Tonocain "contains an immense plain on which is found the ARBRE SOL, which we Christians call the Arbre Sec; and I will tell you what it is like. It is a tall and thick tree, having the bark on one side green and the other white; and it produces a rough husk like that of a chestnut, but without anything in it. The wood is yellow like box, and very strong, and there are no other trees near it nor within a hundred miles of it, except on one side, where you find trees within about ten miles distance."

In a paper published in the _Journal of the R. As. Soc._, Jan., 1909, Gen. Houtum-Schindler comes to the conclusion, p. 157, that Marco Polo's tree is not the "Sun Tree," but the Cypress of Zoroaster; "Marco Polo's _arbre sol_ and _arbre seul_ stand for the Persian _dirakht i sol, i.e. the cypress-tree. If General Houtum Schindler had seen the third edition of the _Book of Ser Marco Polo_, I., p. 113, he would have found that I read his paper of the _J.R.A.S._, of January, 1898."

XXII., p. 132, l. 22. The only current coin is millstones.

Mr. T.B. CLARKE-THORNHILL wrote to me in 1906: "Though I can hardly imagine that there can be any connection between the Caroline Islands and the 'Amiral d'Outre l'Arbre Sec,' still it may interest you to know that the currency of 'millstones' existed up to a short time ago, and may do so still, in the island of Yap, in that group. It consisted of various-sized discs of quartz from about 6 inches to nearly 3 feet in diameter, and from 1/2 an inch to 3 or 4 inches in thickness."

XXV., p. 146.

OLD MAN OF THE MOUNTAIN.

Regarding the reduction of the Ismaelites, the Yuaen Shi tells us that in 1222, on his way back after the taking of Nishapur, Tuli, son of Genghis, plundered the State of Mu-la-i, captured Herat, and joined his father at Talecan. In 1229 the King of Mu-lei presented himself at the Mongol Court.... The following statement is also found in the Mongol Annals: "In the seventh moon [1252] the Emperor ordered K'i-t'ah-t'eh Pu-ha to carry war against the Ma-la-hi.'" (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 136.)

XXVI., p. 149. "On leaving the Castle [of the Old Man], you ride over fine plains and beautiful valleys, and pretty hill-sides producing excellent grass pasture, and abundance of fruits, and all other products.... This kind of country extends for six days' journey, with a goodly number of towns and villages, in which the people are worshippers of Mahommet. Sometimes also you meet with a tract of desert extending for 50 or 60 miles, or somewhat less, and in these deserts you find no water, but have to carry it along with you.... So after travelling for six days as I have told you, you come to a city called Sapurgan...."

Sven Hedin remarks: "From this it is apparent that the six days' journey of fine country were traversed immediately before Marco Polo reached Sapurgan. Sir Henry Yule says in a note: 'Whether the true route be, as I suppose, by Nishapur and Meshed, or, as Khanikoff supposes, by Herat and Badghis, it is strange that no one of those famous cities is mentioned. And we feel constrained to assume that something has been misunderstood in the dictation, or has dropped out of it.' Yule removes the six days of fine country to the district between Sebsevar and Meshed, and considers that for at least the first day's marches beyond Nishapur Marco Polo's description agrees admirably with that given by Fraser and Ferrier.

"I travelled between Sebsevar and Meshed in the autumn of 1890, and I cannot perceive that Marco Polo's description is applicable to the country. He speaks of six days' journey through beautiful valleys and pretty hillsides. To the east of Sebsevar you come out into desert country, which, however passes into fertile country with many villages.[2] Then there comes a boundless dreary steppe to the south. At the village Seng-i-kal-i-deh you enter an undulating country with immense flocks of sheep. 'The first stretch of the road between Shurab and Nishapur led us through perfect desert..; but the landscape soon changed its aspect; the desert passed by degrees into cultivated lands, and we rode past several villages surrounded by fields and gardens.... We here entered the most fertile and densely peopled region in Khorasan, in the midst of which the town of Nishapur is situated.' Of the tract to the east of Nishapur I say: 'Here are found innumerable villages. The plain and slopes are dotted with them. This district is extraordinarily densely inhabited and well cultivated.' But then all this magnificence comes to an end, and of the last day's journey between Kademgah and Meshed I write: 'The country rose and we entered a maze of low intricate hillocks.... The country was exceedingly dreary and bare. Some flocks of sheep were seen, however, but what the fat and sleek sheep lived on was a puzzle to me.... This dismal landscape was more and more enlivened by travellers.... To the east stretched an undulating steppe up to the frontier of Afghanistan.'

"The road between Sebsevar and Meshed is, in short, of such a character that it can hardly fit in with Marco Polo's enthusiastic description of the six days. And as these came just before Sapurgan, one cannot either identify the desert regions named with the deserts about the middle course of the Murgab which extend between Meshed and Shibirkhan. He must have crossed desert first, and it may be identified with the nemek-sar or salt desert east of Tun and Kain. The six days must have been passed in the ranges Paropamisus, Firuz-kuh, and Bend-i-Turkestan. Marco Polo is not usually wont to scare his readers by descriptions of mountainous regions, but at this place he speaks of mountains and valleys and rich pastures. As it was, of course, his intention to travel on into the heart of Asia, to make a detour through Sebsevar was unnecessary and out of his way. If he had travelled to Sebsevar, Nishapur, and Meshed, he would scarcely call the province of Tun-o-Kain the extremity of Persia towards the north, even as the political boundaries were then situated.

"From Balkh his wonderful journey proceeded further eastwards, and therefore we take leave of him. Precisely in Eastern Persia his descriptions are so brief that they leave free room for all kinds of speculations. In the foregoing pages it has been simply my desire to present a few new points of view. The great value of Marco Polo's description of the Persian desert consists in confirming and proving its physical invariableness during more than six hundred years. It had as great a scarcity of oases then as now, and the water in the wells was not less salt than in our own days." (Overland to India, II., pp. 75-77.)

XXVII., p. 152 n.

DOGANA.

"The country of Dogana is quite certain to be the Chinese T'u-ho-lo or Tokhara; for the position suits, and, moreover, nearly all the other places named by Marco Polo along with Dogana occur in Chinese History along with Tokhara many centuries before Polo's arrival. Tokhara being the most important, it is inconceivable that Marco Polo would omit it. Thus, Poh-lo (Balkh), capital of the Eptals; Ta-la-kien (Talecan), mentioned by Hiuan Tsang; Ho-sim or Ho-ts'z-mi (Casem), mentioned in the T'ang History; Shik-nih or Shi-k'i-ni (Syghinan) of the T'ang History; Woh-k'an (Vochan), of the same work; several forms of Bolor, etc. (see also my remarks on the Pamir region in the Contemporary Review for Dec., 1897)." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 142.)

XIX., p. 160.

BADAKHSHAN.

"The Chinese name for 'Badakhshan' never appears before the Pa-ta-shan of Kublai's time." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 143.)

XXX., pp. 164-166. "You must know that ten days' journey to the south of Badashan there is a province called PASHAI, the people of which have a peculiar language, and are Idolaters, of a brown complexion. They are great adepts in sorceries and the diabolic arts. The men wear earrings and brooches of gold and silver set with stones and pearls. They are a pestilent people and a crafty; and they live upon flesh and rice. Their country is very hot."

Sir A. STEIN writes (Ancient Khotan, I., pp. 14-15 n.): "Sir Henry Yule was undoubtedly right in assuming that Marco Polo had never personally visited these countries and that his account of them, brief as it is, was derived from hearsay information about the tracts which the Mongol partisan leader Nigudar had traversed, about 1260 A.D., on an adventurous incursion from Badakhshan towards Kashmir and the Punjab. In Chapter XVIII., where the Venetian relates that exploit (see Yule, Marco Polo, I., p. 98, with note, p. 104), the name of Pashai is linked with Dir, the territory on the Upper Panjkora river, which an invader, wishing to make his way from Badakhshan into Kashmir by the most direct route, would necessarily have to pass through.

"The name Pashai is still borne to this day by a Muhamadanized tribe closely akin to the Siah-posh, settled in the Panjshir Valley and in the hills on the west and south of Kafiristan. It has been very fully discussed by Sir Henry Yule (Ibid., I., p. 165), who shows ample grounds for the belief that this tribal name must have once been more widely spread over the southern slopes of the Hindu kush as far as they are comprised in the limits of Kafiristan. If the great commentator nevertheless records his inability to account for Marco Polo's application of 'the name Pashai to the country south-east of Badakhshan,' the reason of the difficulty seems to me to lie solely in Sir Henry Yule's assumption that the route heard of by the traveller, led 'by the Dorah or the Nuksan Pass, over the watershed of Hindu kush into Chitral and so to Dir.'

"Though such a route via Chitral would, no doubt, have been available in Marco Polo's time as much as now, there is no indication whatever forcing us to believe that it was the one really meant by his informants. When Nigudar 'with a great body of horsemen, cruel unscrupulous fellows' went off from Badakhshan towards Kashmir, he may very well have made his way over the Hindu kush by the more direct line that passes to Dir through the eastern part of Kafiristan. In fact, the description of the Pashai people and their country, as given by Marco Polo, distinctly points to such a route; for we have in it an unmistakable reflex of characteristic features with which the idolatrous Siah-posh Kafirs have always been credited by their Muhammadan neighbours.

"It is much to be regretted that the Oriental records of the period, as far as they were accessible to Sir Henry Yule, seemed to have retained only faint traces of the Mongol adventurer's remarkable inroad. From the point of view of Indian history it was, no doubt, a mere passing episode. But some details regarding it would possess special interest as illustrating an instance of successful invasion by a route that so far has not received its due share of attention." [See supra, pp. 4, 22-24.]

XXX., p. 164.

"The Chinese Toba Dynasty History mentions, in company with Samarcand, K'a-shi-mih (Cashmeer), and Kapisa, a State called Pan-she, as sending tribute to North China along with the Persian group of States. This name Pan-she [Chinese] does not, to the best of my belief, occur a second time in any Chinese record." (PARKER, Asiatic Quart. Rev., Jan., 1904, p. 135.)

XXX., p. 164. "Now let us proceed and speak of another country which is seven day's journey from this one [Pashai] towards the south-east, and the name of which is KESHIMUR."

This short estimate has perplexed Sir Henry Yule, l.c., p. 166. Sir Aurel Stein remarks in a note, Serindia, I., p. 12: "The route above indicated [Nigudar's route] permits an explanation. Starting from some point like Arnawal on the Kunar River which certainly would be well within 'Pashai,' lightly equipped horsemen could by that route easily reach the border of Agror on the Indus within seven days. Speaking from personal knowledge of almost the whole of the ground I should be prepared to do the ride myself by the following stages: Dir, Warai, Sado, Chakdara, Kin kargalai, Bajkatta, Kai or Darband on the Indus. It must be borne in mind that, as Yule rightly recognized, Marco Polo is merely reproducing information derived from a Mongol source and based on Nigudar's raid; and further that Hazara and the valley of the Jhelam were probably then still dependent on the Kashmir kingdom, as they were certainly in Kalhana's time, only a century earlier. As to the rate at which Mongols were accustomed to travel on 'Dak,' cf. Yule, Marco Polo, I., pp. 434 seq."

XXXII., pp. 170, 171. "The people [of Badashan] are Mahommetans, and valiant in war.... They [the people of Vokhan] are gallant soldiers."

In Afghan Wakhan, Sir Aurel Stein writes:

"On we cantered at the head of quite a respectable cavalcade to where, on the sandy plain opposite to the main hamlet of Sarhad, two companies of foot with a squad of cavalry, close on two hundred men in all, were drawn up as a guard of honour. Hardy and well set up most of them looked, giving the impression of thoroughly serviceable human material, in spite of a manifestly defective drill and the motley appearance of dress and equipment.

"They belonged, so the Colonel explained to me afterwards, to a sort of militia drafted from the local population of the Badakhshan valleys and Wakhan into the regiments permanently echeloned as frontier guards along the Russian border on the Oxus. Apart from the officers, the proportion of true Pathans among them was slight. Yet I could well believe from all I saw and heard, that, properly led and provided for, these sturdy Iranian hillmen might give a good account of themselves. Did not Marco Polo speak of the people of 'Badashan' as 'valiant in war' and of the men of 'Vokhan' as gallant soldiers?" (Ruins of Desert Cathay, I., p. 66.)

XXXII., pp. 170 seq.

In Chap. III., pp. 64-66, of his Serindia, Sir Aurel Stein has the following on Marco Polo's account of Wakhan:—

"After Wu-k'ung's narrative of his journey the Chinese sources of information about the Pamirs and the adjoining regions run dry for nearly a thousand years. But that the routes leading across them from Wakhan retained their importance also in Muhammedan times is attested by the greatest mediaeval travellers, Marco Polo. I have already, in Ancient Khotan [pp. 41 seq.], discussed the portion of his itinerary which deals with the journey across the Pamirs to 'the kingdom of Cascar' or Kashgar, and it only remains here to note briefly what he tells us of the route by which he approached them from Badakhshan: 'In leaving Badashan you ride twelve days between east and north-east, ascending a river that runs through land belonging to a brother of the Prince of Badashan, and containing a good many towns and villages and scattered habitations. The people are Mahommetans, and valiant in war. At the end of those twelve days you come to a province of no great size, extending indeed no more than three days' journey in any direction, and this is called VOKHAN. The people worship Mahommet, and they have a peculiar language. They are gallant soldiers, and they have a chief whom they call NONE, which is as much as to say Count, and they are liegemen to the Prince of Badashan.' [Polo, I., pp. 170-171.]

"Sir Henry Yule was certainly right in assuming that 'the river along which Marco travels from Badakhshan is no doubt the upper stream of the Oxus, locally known as the Panja.... It is true that the river is reached from Badakhshan Proper by ascending another river (the Vardoj) and crossing the 'Pass of Ishkashm, but in the brief style of our narrative we must expect such condensation.' [Polo, I., pp. 172-3.] Marco's great commentator was guided by equally true judgment when he recognized in the indications of this passage the same system of government that prevailed in the Oxus valleys until modern times. Under it the most of the hill tracts dependent from Badakhshan, including Ishkashim and Wakhan, were ruled not direct by the Mir, but by relations of his or hereditary chiefs who held their districts on a feudal tenure. The twelve days' journey which Marco records between Badashan and 'Vokhan' are, I think, easily accounted for if it is assumed that the distance from capital to capital is meant; for twelve marches are still allowed for as the distance from Baharak, the old Badakhshan capital on the Vardoj, to Kila Panja.

"That the latter was in Marco's days, as at present, the chief place of Wakhan is indicated also by his narrative of the next stage of his journey. 'And when you leave this little country, and ride three days north-east, always among mountains, you get to such a height that 'tis said to be the highest place in the world! And when you have got to this height you find [a great lake between two mountains, and out of it] a fine river running through a plain.... The plain is called PAMIER.' The bearing and descriptive details here given point clearly to the plain of the Great Pamir and Victoria Lake, its characteristic feature. About sixty-two miles are reckoned from Langar-kisht, the last village on the northern branch of the Ab-i-Panja and some six miles above Kila Panja, to Mazar-tapa where the plain of the Great Pamir may be said to begin, and this distance agrees remarkably well with the three marches mentioned by Marco.

"His description of Wakhan as 'a province of no great size, extending indeed no more than three days' journey in any direction' suggests that a portion of the valley must then have formed part of the chiefship of Ishkashim or Zebak over which we may suppose 'the brother of the Prince of Badashan' to have ruled. Such fluctuations in the extent of Wakhan territory are remembered also in modern times. Thus Colonel Trotter, who visited Wakhan with a section of the Yarkand Mission in 1874, distinctly notes that 'Wakhan formerly contained three "sads" or hundreds, i.e., districts, containing 100 houses each' (viz. Sad-i-Sar-hadd, Sad Sipang, Sad Khandut). To these Sad Ishtragh, the tract extending from Digargand to Ishkashim, is declared to have been added in recent times, having formerly been an independent principality. It only remains to note that Marco was right, too, in his reference to the peculiar language of Wakhan; for Wakhi—which is spoken not only by the people of Wakhan but also by the numerous Wakhi colonists spread through Mastuj, Hunza Sarikol, and even further east in the mountains—is a separate language belonging to the well-defined group of Galcha tongues which itself forms the chief extant branch of Eastern Iranian."

XXXII., pp. 171 seq., 175, 182.

THE PLATEAU OF PAMIR.

"On leaving Tash-kurghan (July 10, 1900), my steps, like those of Hiuan-tsang, were directed towards Kashgar.... In Chapters V.-VII. of my Personal Narrative I have given a detailed description of this route, which took me past Muztagh-Ata to Lake Little Kara-kul, and then round the foot of the great glacier-crowned range northward into the Gez defile, finally debouching at Tashmalik into the open plain of Kashgari. Though scarcely more difficult than the usual route over the Chichiklik Pass and by Yangi-Hisar, it is certainly longer and leads for a considerably greater distance over ground which is devoid of cultivation or permanent habitations.

"It is the latter fact which makes me believe that Professor H. Cordier was right in tracing by this very route Marco Polo's itinerary from the Central Pamirs to Kashgar. The Venetian traveller, coming from Wakhan, reached, after three days, a great lake which may be either Lake Victoria or Lake Chakmak, at a 'height that is said to be the highest place in the world.' He then describes faithfully enough the desert plain called 'Pamier,' which he makes extend for the distance of a twelve days' ride, and next tells us: 'Now, if we go on with our journey towards the east-north-east, we travel a good forty days, continually passing over mountains and hills, or through valleys, and crossing many rivers and tracts of wilderness. And in all this way you find neither habitation of man, nor any green thing, but must carry with you whatever you require.'

"This reference to continuous 'tracts of wilderness' shows clearly that, for one reason or another, Marco Polo did not pass through the cultivated valleys of Tash-kurghan or Tagharma, as he would necessarily have done if his route to Kashgar, the region he next describes, had lain over the Chichiklik Pass. We must assume that, after visiting either the Great or Little Pamir, he travelled down the Ak-su river for some distance, and then crossing the watershed eastwards by one of the numerous passes struck the route which leads past Muztagh-Ata and on towards the Gez defile. In the brief supplementary notes contributed to Professor Cordier's critical analysis of this portion of Marco Polo's itinerary, I have pointed out how thoroughly the great Venetian's description of the forty days' journey to the E.N.E. of the Pamir Lake can be appreciated by any one who has passed through the Pamir region and followed the valleys stretching round the Muztagh-Ata range on the west and north (cf. Yule, Marco Polo, II., pp. 593 seq.). After leaving Tash-kurghan and Tagharma there is no local produce to be obtained until the oasis of Tashmalik is reached. In the narrow valley of the Yaman-yar river, forming the Gez defile, there is scarcely any grazing; its appearance down to its opening into the plain is, in fact, far more desolate than that of the elevated Pamir regions.

"In the absence of any data as to the manner and season in which Marco Polo's party travelled, it would serve no useful purpose to hazard explanations as to why he should assign a duration of forty days to a journey which for a properly equipped traveller need not take more than fifteen or sixteen days, even when the summer floods close the passage through the lower Gez defile, and render it necessary to follow the circuitous track over the Tokuk Dawan or 'Nine Passes.' But it is certainly worth mention that Benedict Goez, too, speaks of the desert of 'Pamech' (Pamir) as taking forty days to cross if the snow was extensive, a record already noted by Sir H. Yule (Cathay, II., pp. 563 seq.). It is also instructive to refer once more to the personal experience of the missionary traveller on the alternative route by the Chichiklik Pass. According to the record quoted above, he appears to have spent no less than twenty-eight days in the journeys from the hamlets of 'Sarcil' (Sarikol, i.e. Tash-kurghan) to 'Hiarchan' (Yarkand)—a distance of some 188 miles, now reckoned at ten days' march." (Stein, Ancient Khotan, pp. 40-42.)

XXXII., p. 171. "The Plain is called PAMIER, and you ride across it for twelve days together, finding nothing but a desert without habitations or any green thing, so that travellers are obliged to carry with them whatever they have need of."

At Sarhad, Afghan Wakhan, Stein, Ruins of Desert Cathay, I., p. 69, writes: "There was little about the low grey houses, or rather hovels, of mud and rubble to indicate the importance which from early times must have attached to Sarhad as the highest place of permanent occupation on the direct route leading from the Oxus to the Tarim Basin. Here was the last point where caravans coming from the Bactrian side with the products of the Far West and of India could provision themselves for crossing that high tract of wilderness 'called Pamier' of which old Marco Polo rightly tells us: 'You ride across it ...' And as I looked south towards the snow-covered saddle of the Baroghil, the route I had followed myself, it was equally easy to realize why Kao Hsien-chih's strategy had, after the successful crossing of the Pamirs, made the three columns of his Chinese Army concentrate upon the stronghold of Lien-yuen, opposite the present Sarhad. Here was the base from which Yasin could be invaded and the Tibetans ousted from their hold upon the straight route to the Indus."

XXXII., p. 174.

"The note connecting Hiuan Tsang's Kieh sha with Kashgar is probably based upon an error of the old translators, for the Sita River was in the Pamir region, and K'a sha was one of the names of Kasanna, or Kieh-shwang-na, in the Oxus region." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 143.)

XXXII., I. p. 173; II. p. 593.

PAONANO PAO.

Cf. The Name Kushan, by J.F. Fleet, Jour. Roy. As. Soc., April, 1914, pp. 374-9; The Shaonano Shao Coin Legend; and a Note on the name Kushan by J. Allan, Ibid., pp. 403-411. PAONANO PAO. Von Joh. Kirste. (Wiener Zeit. f. d. Kunde d. Morg., II., 1888, pp. 237-244.)

XXXII., p. 174.

YUE CHI.

"The old statement is repeated that the Yueeh Chi, or Indo-Scyths (i.e. the Eptals), 'are said to have been of Tibetan origin.' A long account of this people was given in the Asiatic Quart. Rev. for July, 1902. It seems much more likely that they were a branch of the Hiung-nu or Turks. Albiruni's 'report' that they were of Tibetan origin is probably founded on the Chinese statement that some of their ways were like Tibetan ways, and that polyandry existed amongst them; also that they fled from the Hiung-nu westwards along the north edge of the Tibetan territory, and some of them took service as Tibetan officials." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 143.)

XXXII., pp. 178-179.

BOLOR.

We read in the Tarikh-i-Rashidi of Mirza Haidar (Notes by Ney Elias; translated by E.D. Ross, 1895), p. 135, that Sultan Said Khan, son of Mansur Khan, sent the writer in the year 934 (1528), "with Rashid Sultan, to Balur, which is a country of infidels [Kafiristan], between Badakhshan and Kashmir, where we conducted successfully a holy war [ghazat], and returned victorious, loaded with booty and covered with glory."

Mirza Haidar gives the following description of Bolor (pp. 384-5): "Balur is an infidel country [Kafiristan], and most of its inhabitants are mountaineers. Not one of them has a religion or a creed. Nor is there anything which they [consider it right to] abstain from or to avoid [as impure]; but they do whatever they list, and follow their desires without check or compunction. Baluristan is bounded on the east by the province of Kashgar and Yarkand; on the north by Badakhshan; on the west by Kabul and Lumghan; and on the south by the dependencies of Kashmir. It is four months' journey in circumference. Its whole extent consists of mountains, valleys, and defiles, insomuch that one might almost say that in the whole of Baluristan, not one farsakh of level ground is to be met with. The population is numerous. No village is at peace with another, but there is constant hostility, and fights are continually occurring among them."

From the note to this passage (p. 385) we note that "for some twenty years ago, Mr. E.B. Shaw found that the Kirghiz of the Pamirs called Chitral by the name of Palor. To all other inhabitants of the surrounding regions, however, the word appears now to be unknown....

"The Balur country would then include Hunza, Nagar, possibly Tash Kurghan, Gilgit, Panyal, Yasin, Chitral, and probably the tract now known as Kafiristan: while, also, some of the small states south of Gilgit, Yasin, etc., may have been regarded as part of Balur....

"The conclusions arrived at [by Sir H. Yule], are very nearly borne out by Mirza Haidar's description. The only differences are (1) that, according to our author, Baltistan cannot have been included in Balur, as he always speaks of that country, later in his work, as a separate province with the name of Balti, and says that it bordered on Balur; and (2) that Balur was confined almost entirely, as far as I am able to judge from his description in this passage and elsewhere, to the southern slopes of the Eastern Hindu Kush, or Indus water-parting range; while Sir H. Yule's map makes it embrace Sarigh-Kul and the greater part of the eastern Pamirs."

XXXIII., p. 182. "The natives [of Cascar] are a wretched, niggardly set of people; they eat and drink in miserable fashion."

The people of Kashgar seem to have enjoyed from early times a reputation for rough manners and deceit (Stein, Ancient Khotan, p. 49 n). Stein, p. 70, recalls Hiuan Tsang's opinion: "The disposition of the men is fierce and impetuous, and they are mostly false and deceitful. They make light of decorum and politeness, and esteem learning but little." Stein adds, p. 70, with regard to Polo's statement: "Without being able to adduce from personal observation evidence as to the relative truth of the latter statement, I believe that the judgements recorded by both those great travellers may be taken as a fair reflex of the opinion in which the 'Kashgarliks' are held to this day by the people of other Turkestan districts, especially by the Khotanese. And in the case of Hiuan Tsang at least, it seems probable from his long stay in, and manifest attachment to, Khotan that this neighbourly criticism might have left an impression upon him."

XXXVI., p. 188.

KHOTAN.

Sir Aurel Stein writes (Ancient Khotan, I., pp. 139-140): "Marco Polo's account of Khotan and the Khotanese forms an apt link between these early Chinese notices and the picture drawn from modern observation. It is brief but accurate in all details. The Venetian found the people 'subject to the Great Kaan' and 'all worshippers of Mahommet.' 'There are numerous towns and villages in the country, but Cotan, the capital, is the most noble of all and gives its name to the kingdom. Everything is to be had there in plenty, including abundance of cotton [with flax, hemp, wheat, wine, and the like]. The people have vineyards and gardens and estates. They live by commerce and manufactures, and are no soldiers.' Nor did the peculiar laxity of morals, which seems always to have distinguished the people of the Khotan region, escape Marco Polo's attention. For of the 'Province of Pein' which, as we shall see, represents the oases of the adjoining modern district of Keriya, he relates the custom that 'if the husband of any woman go away upon a journey and remain away for more than twenty days, as soon as that term is past the woman may marry another man, and the husband also may then marry whom he pleases.'

"No one who has visited Khotan or who is familiar with the modern accounts of the territory, can read the early notices above extracted without being struck at once by the fidelity with which they reflect characteristic features of the people at the present day. Nor is it necessary to emphasize the industrial pre-eminence which Khotan still enjoys in a variety of manufactures through the technical skill and inherited training of the bulk of its population."

Sir Aurel Stein further remarks (Ancient Khotan, I., p. 183): "When Marco Polo visited Khotan on his way to China, between the years 1271 and 1275, the people of the oasis were flourishing, as the Venetian's previously quoted account shows. His description of the territories further east, Pein, Cherchen, and Lop, which he passed through before crossing 'the Great Desert' to Sha-chou, leaves no doubt that the route from Khotan into Kan-su was in his time a regular caravan road. Marco Polo found the people of Khotan 'all worshippers of Mahommet' and the territory subject to the 'Great Kaan', i.e. Kublai, whom by that time almost the whole of the Middle Kingdom acknowledged as emperor. While the neighbouring Yarkand owed allegiance to Kaidu, the ruler of the Chagatai dominion, Khotan had thus once more renewed its old historical connexion with China."

XXVI., p. 190.

"A note of Yule's on p. 190 of Vol. I. describes Johnson's report on the people of Khoten (1865) as having 'a slightly Tartar cast of countenance.' The Toba History makes the same remark 1300 years earlier: 'From Kao-ch'ang (Turfan) westwards the people of the various countries have deep eyes and high noses; the features in only this one country (Khoten) are not very Hu (Persian, etc.), but rather like Chinese.' I published a tolerably complete digest of Lob Nor and Khoten early history from Chinese sources, in the Anglo-Russian Society's Journal for Jan. and April, 1903. It appears to me that the ancient capital Yotkhan, discovered thirty-five years ago, and visited in 1891 by MM. de Rhins and Grenard, probably furnishes a clue to the ancient Chinese name of Yu-t'ien." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 143.)

XXXVII., p. 190 n.

Stein has devoted a whole chapter of his Sand-buried Ruins of Khotan, Chap. XVI., pp. 256 seq. to Yotkan, the Site of the Ancient Capital.

XXXVII., p. 191, n. 1.

PEIN.

"It is a mistake to suppose that the earlier pilgrim Fa-hien (A.D. 400) followed the 'directer route' from China; he was obliged to go to Kao ch'ang, and then turn sharp south to Khoten." (E.H. PARKER, Asiatic Quart. Rev., Jan., 1904, p. 143.)

XXXVII., p. 192.

I have embodied, in Vol. II., p. 595, of Marco Polo, some of the remarks of Sir Aurel Stein regarding Pein and Uzun Tati. In Ancient Khotan, I., pp. 462-3, he has given further evidence of the identity of Uzun Tati and P'i mo, and he has discussed the position of Ulug-Ziarat, probably the Han mo of Sung Yun.

XXXVII., p. 191; II., p. 595.

"Keriya, the Pein of Marco Polo and Pimo of Hwen Tsiang, writes Huntington, is a pleasant district, with a population of about fifteen thousand souls." Huntington discusses (p. 387) the theory of Stein:

"Stein identifies Pimo or Pein, with ancient Kenan, the site ... now known as Uzun Tetti or Ulugh Mazar, north of Chira. This identification is doubtful, as appears from the following table of distances given by Hwen Tsiang, which is as accurate as could be expected from a casual traveller. I have reckoned the 'li,' the Chinese unit of distance, as equivalent to 0.26 of a mile.

Distance according to Names of Places. True Distance. Hwen Tsiang. Khotan (Yutien) to Keriya (Pimo) 97 miles. 330 li 86 miles. Keriya (Pimo) to Niya (Niyang) 64 " 200 " 52 " Niya (Niyang) to Endereh (Tuholo) 94 " 400 " 104 " Endereh (Tuholo) to Kotak Sheri? (Chemotona) 138? " 600 " 156 " Kotak Sheri (Chemotona) to Lulan (Nafopo) 264? " 1000 " 260 "

"If we use the value of the 'li' 0.274 of a mile given by Hedin, the distances from Khotan to Keriya and from Keriya to Niya, according to Hwen Tsiang, become 91 and 55 miles instead of 86 and 52 as given in the table, which is not far from the true distances, 97 and 64.

"If, however, Pimo is identical with Kenan, as Stein thinks, the distances which Hwen Tsiang gives as 86 and 52 miles become respectively 60 and 89, which is evidently quite wrong.

"Strong confirmation of the identification of Keriya with Pimo is found in a comparison of extracts from Marco Polo's and Hwen Tsiang's accounts of that city with passages from my note-book, written long before I had read the comments of the ancient travellers. Marco Polo says that the people of Pein, or Pima, as he also calls it, have the peculiar custom 'that if a married man goes to a distance from home to be about twenty days, his wife has a right, if she is so inclined, to take another husband; and the men, on the same principle, marry wherever they happen to reside.' The quotation from my notes runs as follows: 'The women of the place are noted for their attractiveness and loose character. It is said that many men coming to Keriya for a short time become enamoured of the women here, and remain permanently, taking new wives and abandoning their former wives and families.'

"Hwen Tsiang observed that thirty 'li,' seven or eight miles, west of Pimo, there is 'a great desert marsh, upwards of several acres in extent, without any verdure whatever. The surface is reddish black.' The natives explained to the pilgrim that it was the blood-stained site of a great battle fought many years before. Eighteen miles north-west of Keriya bazaar, or ten miles from the most westerly village of the oasis, I observed that 'some areas which are flooded part of the year are of a deep rich red colour, due to a small plant two or three inches high.' I saw such vegetation nowhere else and apparently it was an equally unusual sight to Hwen Tsiang.

"In addition to these somewhat conclusive observations, Marco Polo says that jade is found in the river of Pimo, which is true of the Keriya, but not of the Chira, or the other rivers near Kenan." (Ellsworth HUNTINGTON, The Pulse of Asia, pp. 387-8.)

XXVIII., p. 194. "The whole of the Province [of Charchan] is sandy, and so is the road all the way from Pein, and much of the water that you find is bitter and bad. However, at some places you do find fresh and sweet water."

Sir Aurel Stein remarks (Ancient Khotan, I., p. 436): "Marco Polo's description, too, 'of the Province of Charchan' would agree with the assumption that the route west of Charchan was not altogether devoid of settlements even as late as the thirteenth century.... [His] account of the route agrees accurately with the conditions now met with between Niya and Charchan. Yet in the passage immediately following, the Venetian tells us how 'when an army passes through the land, the people escape with their wives, children, and cattle a distance of two or three days' journey into the sandy waste; and, knowing the spots where water is to be had, they are able to live there, and to keep their cattle alive, while it is impossible to discover them.' It seems to me clear that Marco Polo alludes here to the several river courses which, after flowing north of the Niya-Charchan route, lose themselves in the desert. The jungle belt of their terminal areas, no doubt, offered then, as it would offer now, safe places of refuge to any small settlements established along the route southwards."

XXXIX., P. 197.

OF THE CITY OF LOP.

Stein remarks, Ruins of Desert Cathay, I., p. 343: "Broad geographical facts left no doubt for any one acquainted with local conditions that Marco Polo's Lop, 'a large town at the edge of the Desert' where 'travellers repose before entering on the Desert' en route for Sha chou and China proper, must have occupied the position of the present Charklik. Nor could I see any reason for placing elsewhere the capital of that 'ancient kingdom of Na-fo-po, the same as the territory of Lou-lan,' which Hiuan Tsang reached after ten marches to the north-east of Chue-mo or Charchan, and which was the pilgrim's last stage before his return to Chinese soil."

In his third journey (1913-1916), Stein left Charchan on New Year's Eve, 1914, and arrived at Charkhlik on January 8, saying: "It was from this modest little oasis, the only settlement of any importance in the Lop region, representing Marco Polo's 'City of Lop,' that I had to raise the whole of the supplies, labour, and extra camels needed by the several parties for the explorations I had carefully planned during the next three months in the desert between Lop-nor and Tunhuang."

"The name of LOB appears under the form Lo pou in the Yuan-shi, s.a. 1282 and 1286. In 1286, it is mentioned as a postal station near those of K'ie-t'ai, Che-ch'an and Wo-tuan. Wo-tuan is Khotan. Che-ch'an, the name of which reappears in other paragraphs, is Charchan. As to K'ie-t'ai, a postal station between those of Lob and Charchan, it seems probable that it is the Kaetaek of the Tarikh-i-Rashidi." (PELLIOT.)

See in the Journ. Asiatique, Jan.-Feb., 1916, pp. 117-119, Pelliot's remarks on Lob, Navapa, etc.

XXXIX., pp. 196-7.

THE GREAT DESERT.

After reproducing the description of the Great Desert in Sir Henry Yule's version, Stein adds, Ruins of Desert Cathay, I., p. 518:

"It did not need my journey to convince me that what Marco here tells us about the risks of the desert was but a faithful reflex of old folklore beliefs he must have heard on the spot. Sir Henry Yule has shown long ago that the dread of being led astray by evil spirits haunted the imagination of all early travellers who crossed the desert wastes between China and the oases westwards. Fa-hsien's above-quoted passage clearly alludes to this belief, and so does Hiuan Tsang, as we have seen, where he points in graphic words the impressions left by his journey through the sandy desert between Niya and Charchan.

"Thus, too, the description we receive through the Chinese historiographer, Ma Tuan-lin, of the shortest route from China towards Kara-shahr, undoubtedly corresponding to the present track to Lop-nor, reads almost like a version from Marco's book, though its compiler, a contemporary of the Venetian traveller, must have extracted it from some earlier source. 'You see nothing in any direction but the sky and the sands, without the slightest trace of a road; and travellers find nothing to guide them but the bones of men and beasts and the droppings of camels. During the passage of this wilderness you hear sounds, sometimes of singing, sometimes of wailing; and it has often happened that travellers going aside to see what these sounds might be have strayed from their course and been entirely lost; for they were voices of spirits and goblins.'...

"As Yule rightly observes, 'these Goblins are not peculiar to the Gobi.' Yet I felt more than ever assured that Marco's stories about them were of genuine local growth, when I had travelled over the whole route and seen how closely its topographical features agree with the matter-of-fact details which the first part of his chapter records. Anticipating my subsequent observations, I may state here at once that Marco's estimate of the distance and the number of marches on this desert crossing proved perfectly correct. For the route from Charklik, his 'town of Lop,' to the 'City of Sachiu,' i.e. Sha-chou or Tun-huang, our plane-table survey, checked by cyclometer readings, showed an aggregate marching distance of close on 380 miles."

XXXIX., p. 196.

OF THE CITY OF LOP AND THE GREAT DESERT.

"In the hope of contributing something toward the solution of these questions [contradictory statements of Prjevalsky, Richthofen, and Sven Hedin]," writes Huntington, "I planned to travel completely around the unexplored part of the ancient lake, crossing the Lop desert in its widest part. As a result of the journey, I became convinced that two thousand years ago the lake was of great size, covering both the ancient and the modern locations; then it contracted, and occupied only the site shown on the Chinese maps; again, in the Middle Ages, it expanded; and at present it has contracted and occupies the modern site.