chapter from the Arabian Nights; their subsequent adventures: Marco's participation in the great defeat of the Venetians at Curzola; his captivity at Genoa, and dictation of his memoirs to a fellow-prisoner, Rustician of Pisa; and finally, it suggests how Rustician's notes, jotted down in the Lingua 'franca' in which they were probably communicated, were enlarged, and amended, and annotated, either by Marco himself, or possibly by his uncle Maffeo, who had been his companion throughout his travels; and how from these original notes the various texts were formed which are now extant in seventyfive different manuscript copies of a more or less authentic character.

It is clear that Marco Polo, with little or no preliminary education, must still have possessed considerable natural abilities, since on his arrival at the Mongol court he acquired without difficulty the current languages of the country together with four different modes of writing (probably Mongolian, Ouigour, Persian, and Thibetan*), and further ingratiated himself with the Emperor, so as to be employed by him on confidential affairs of state in preference to the officers of his own household; but it is equally clear that he fully shared in the credulity and superstition of the age; and although Colonel Yule does not scruple to avow his entire confi'dence in the man's veracity,' no one can doubt but that Marco was disposed to exaggeration in his phraseology, and indulged in a very high colouring in all his descriptions. He seems, indeed, mainly to have risen into favour with the Emperor from his skill in bringing back sensational reports of the wonders which he saw when employed on deputation in strange countries such reports contrasting agreeably with the dry matter-of-fact relations of the ordinary commissioners; and

This is Colonel Yule's proposed identification of the four written 'characters' which were learnt by Marco; but instead of Thibetan it is likely enough that he learnt the Baspa alphabet, which was established, by orders of Kublai in 1269, as the official Mongolian character, in contradistinction to the old writing which, like the Ouigour and the Manchu, was of Nestorian origin. At first sight it may seem hazardous to include Persian in this series, as it has no etymological or geographical connexion with Mongolian or Chinese, but Colonel Yule has shown good reason for suspecting that Persian must have been the common tongue of foreigners at the court of the Mongols (vol. i. p. cxxxv). In addition to the examples cited by Colonel Yule of such pure Persian names as Pul-i-sangin, Zar-dandán, &c. used by Marco Polo, it may be of interest to remark that in the famous Kitab-el-Fihrist, recently published, we find the Chinese commander-in-chief in the ninth century to have been named Sir-aspah, which is Persian for head of the army.'

we may well understand that it was this proneness to extravagant talk, this habitual indulgence in travellers' tales,' which gave him the nickname of Master Millions' among his countrymen, and which in fact discredited his general authority. The process of dictation, it may also be suggested, is of itself unfavourable to a very rigid accuracy of description. In telling his stories vivâ voce to Rustician, as he paced the floor of his prison cell at Genoa, he may be forgiven if he occasionally warmed up his flagging memory by a few free touches of lively rodomontade.* That he did not designedly invent or falsify is all, we presume, that Colonel Yule contends for; and for this qualified acquittal there is ample authority in the contemporary evidence that when Marco was asked by his friends on his death-bed to correct the book by removing everything that went beyond the facts, he replied, that he had not told one-half of what he had really seen.'

6

Colonel Yule has allowed himself the fullest latitude in his adoption of a text. He calls his text eclectic,' which means. that he has selected from several types the readings and expressions of which he approves, and has omitted those of which he disapproves. The basis of his translation is the same text which was used by Mons. Pauthier, and which is supposed to represent the version made from Rustician's barbarous patois

The following are a few instances of Marco Polo's proneness to exaggeration in reporting what he heard as well as what he saw :— A ruc's feather brought from Madagascar measured, he was told, 90 spans, while the quill part was 2 palms in circumference; and two boars' tusks from the same place weighed more than 14 lbs. a-piece, the boars themselves being as big as buffalos (ii. 347). In Thibet the bamboos were 3 palms in girth and 15 paces in length, and in burning made a report that could be heard 10 miles off (ii. 26). The Thibet mastiffs, again, were as big as donkeys (ii. 32). The serpents (i.e. alligators) of Carajan were 10 paces in length and 10 palms in girth, with eyes bigger than a great loaf of bread, and a mouth large enough to swallow a man whole (ii. 45). The elephants of Birma carried from twelve to sixteen well-armed fighting men (ii. 63); and the oxen of the same province were as tall as elephants (ii. 78). For 'travellers' tales' we may quote the story of the unicorn (or rhinoceros) of Sumatra which licked its victim to death with its prickly tongue (ii. 227); the tailed men of Lambri, on the same island (ii. 243); the dog-headed men of the Andamans (ii. 251); the famous Ceylon ruby, which was a palm in length and as thick as a man's arm (ii. 254); and especially the couvade of the Zar-dandán or 'golden teeth' (ii. 52), which gave rise to the famous lines in Butler's Hudibras: '

. . . Chineses go to bed

And lie in in the irladies' stead.'

into French of the period, during Marco Polo's life, and subject to his own curtailment, correction, and revision; but he has not slavishly followed this version, of which there are exemplars at Paris, at Berne, and at Oxford. He has admitted variant readings of names, and many expressions of special 'interest and character' from Rustician's original notes, published by the Geographical Society of Paris in 1824; and also in some instances he has borrowed from other versions that were made from that text (apparently during Marco Polo's lifetime), first into Italian, and then into Latin--Pipino's Latin text, under date A.D. 1320, being the type of this class of MSS.; and finally, he has introduced between brackets, as indicative of their supplementary character, a very large number of additional paragraphs, some of the highest interest and importance, which bear internal marks of emanating either from Marco Polo or his uncle, but which are only known at present from their being included, without comment or expla nation, in Ramusio's famous posthumous translation in Italian, which was published in A.D. 1559, nearly 240 years after Marco Polo's decease. It is hardly perhaps consistent with the strict canons of criticism thus to blend several texts into one, culling the best passages of each, and correcting false readings or tedious repetitions à discrétion; but the result is certainly to the advantage of the general reader; and if a thorough dependence can be placed on the knowledge and judgment of the editor, there will be also felt an assurance that the eclectic' text presents what the author said, or would. have desired to say. This, at any rate, is what Colonel Yule has aimed at, and we are bound to say that we think on the whole he has been successful.

Incidentally at the outset of Marco Polo's narrative, a geographical question arises which well deserves a little careful consideration, since it involves the existence, or non-existence, at that period of history of one of the great inland seas of Central Asia. The elder Poli, in their first journey to the East, in A.D. 1260, are said to have passed directly from the Volga. to Bokhárá by a route which, according to the present physical. configuration of the country, must have led them along the northern, or the southern, border of the Sea of Aral; yet neither in Marco's brief notice of this journey, nor in any other part of his work, is there the slightest allusion to the sea.

Colonel Yule mentions as a literary curiosity of some interest an Irish version which was made with an astounding freedom' from this Latin text, and which is included in the famous Book of Lismore,, written about A.D. 1450.

in question; and a doubt therefore naturally arises in the reader's mind as to whether the Aral could have been in existence in the thirteenth century. Colonel Yule does not enter on the discussion of this curious question in either of his great works, Marco Polo' or 'Cathay;' but in another place he has casually considered it, and the result of his investigation is that he supports the opinion of his distinguished relation, Sir Roderick Murchison, to the effect that-notwithstanding certain admitted temporary deviations of the Oxus, and notwithstanding much ambiguity both of nomenclature and description, which is due, they think, to the carelessness or ignorance of the early geographers--the relative condition of the Caspian and Aral has in reality never materially varied during the historic period. A strong array of authorities, including the honoured names of Saint Martin, Malte Brun, Hugh Murray, Baillie Fraser, and Burnes, are even more positive in their opinions, maintaining that any such variation has been simply impossible, since the Oxus and Jaxartes have never changed their course, but from time immemorial have disembogued into the Aral Sea, precisely as is the case at the present day.

On the other hand, it has been roundly asserted by geographers of almost equal weight, that the Aral has fluctuated at different periods of history between the condition of a great inland sea and that of a reedy marsh, according to the varying course of its two feeders, the Oxus and Jaxartes; and has sometimes even, when the supply of water from those feeders has been entirely cut off for a lengthened period, disappeared altogether from the map of Asia. We have reason to believe that this curious question of physical hydrography, which has been already partially ventilated before the Geographical Society of London,* will, in the course of their ensuing session, be subjected to further rigid inquiry, and will receive probably a definite solution; but, in the meantime, a brief recapitulation of the changes which the Aral is said to have undergone, and of the evidence on which those asserted changes depend, may not perhaps be out of place as an introduction to Marco Polo's own view of the geography of Central Asia.

Among the ancients, then, Herodotus and Strabo are the only authors who can be supposed from their writings to have had any cognisance of the existence of the Aral; and their description applies, not to a large independent sea, but rather

* See 'Proceedings of the Royal Geographical Society,' vol. xi. no. iii. p. 114; and Journal of the Royal Geographical Society,' vol. xxxvii. President's Address, p. 140.,

to a series of reedy swamps, fed by the overflow of the Jaxartes, the main arm of which river, however, found its way to the Caspian. All other writers simply bring the Oxus and Jaxartes into the Caspian without any allusion to the deflexion or bifurcation of either stream, estimating the distance between the mouths of the two rivers at about eighty parasangs; and when we consider the extent of information at the disposal of the Greek and Roman geographers, when we remember that Greek princes ruled for some centuries in the countries between Persia and the Indian Caucasus, that Greek admirals navigated the Caspian, and Greek commanders penetrated beyond the Jaxartes, while the merchants who followed the caravan routes from India to the Mediterranean brought their journals and road-books to Rome, it seems impossible to doubt but that we have in such standard works as those of Strabo and Pliny and Ptolemy, a representation of the true hydrography of Eastern Persia for about 500 years before and after the Christian era. As late, indeed, as A.D. 570, when Zemarchus returned from his mission to the Turkish Khagan, then encamped in the Ak-Tagh, north of Samarcand, and crossed the Oech (or Vakh,' probably the right arm of the Oxus), nearthe city of Urganj, he found the Aral, not yet developed to the condition of an inland sea, but still bearing the character of a large reedy morass;* and it was not probably till thirty or forty years later, during the reign of Khusrú Parvíz, that the great change took place which cut off the water of the Oxus entirely from the Caspian, and turned the full stream into the Aral-the sea of Kardar, which was the south-western portion of the present Abugír Lake, and which had been probably fed, up to that date, by the Urganj branch of the river, being at the same time desiccated, and a treasure-city (the modern

6

* The geography of the expedition of Zemarchus has, we think, been quite misunderstood by Colonel Yule (Cathay, vol. i. p. clxiii). The camp of Dizabulus was in the country of the Sogdians, and certainly, therefore, not beyond the Jaxartes. The name of Talas, indeed, which has led Colonel Yule astray, did not apply in the seventh century to the town and river which bore that title in later times. The Choliatæ (or Khalaját?) probably dwelt on the left bank of the old bed of the Jaxartes. At any rate the Roman party must have struck the Aral marshes at their south-eastern corner, and thus skirted round their southern edge, the short desert-track which was followed by George and his party on their return to Byzantium being the direct line from Urganj to Asterabad, and so on by the north of Persia to Asia Minor. This short cut, indeed, is quite inexplicable if we suppose Zemarchus to have passed to the north of the Aral marshes.