him from the guard, and hastens to the palace, where, obtaining audience of the King, the secret of Bakhtyār's birth is discovered; and the King resigns the throne in favour of his son, and causes the Ten envious Viziers to be put to death.

Such is the frame within which nine different stories are inserted; and although it was doubtless imitated from, it has but a faint likeness to, that of the Book of Sindibad. The work which appears most closely to resemble the Romance of Prince Bakhtyār, in the frame, is a collection of Tales in the Tamul language, entitled, Alakeswara Kathá, in which four ministers of the King of Alakapur are falsely accused of violating the King's private apartments, and vindicate their innocence, and disarm the King's wrath, by relating a number of stories.*

According to M. Deslongchamps, in his learned and elaborate Essai sur les Fables Indiennes, there exist in Oriental languages three versions of the Bakhtyar Nama-Persian, Arabic, and Turki (¿.e., Eastern Turkish-Uygur). Of the Persian version it is said there are numerous manuscripts in the great libraries of England and France; and besides the printed text appended to Sir William Ouseley's English translation, published in 1800, a lithographed text was issued, at Paris, in 1839, probably from a manuscript

* Wilson's Descriptive Catalogue of the Mackenzie MSS. vol. i, p. 220,

in the Royal Library. The Arabian version, under the title of "The History of the Ten Viziers," forms part of the text of the Thousand and One Nights, in 12 volumes, of which Dr Maximilian Habicht edited vols. 1 to 8, published at uncertain intervals, at Breslau, from 1825 to 1838 inclusive, when the work was stopped by Habicht's death. In 1842-3 Professor H. L. Fleischer issued the remaining vols., 9 to 12. The same year when Habicht began the publication of his Arabian text he issued a complete German translation, also at Breslau, in 15 small square volumes, under the title of Tausend und Eine Nacht: Arabische Erzählungen. Zum erstenmal aus einer Tunesischen Handschrift, ergänzt und vollständig übersetzt, von Max. Habicht, F. H. Von der Hagen, und Karl Schall.* But both the number and the order of the tales of our romance are quite different in the translation and the text: the sixth volume of the latter, which contains the romance, was not published till 1834, or nine years after the first issue of the translation; and it would seem that Habicht, in editing his Tunisian manuscript, compared it with other texts, and made very considerable changes. The romance is found in a dislocated form in a work, published at Paris in 1788, entitled, Nouveaux Contes Arabice, ou Supplement aux Mille et Une Nuits, &c., par M.

*The Thousand and One Nights: Arabian Tales. For the first time completely and fully translated from a Tunisian Manuscript, &c.

l'Abbè * * * In this book (which is of little or no value) the several tales are not placed within the frame, or leading story, which, however, appears in connection with one of them. It is also included in the French Continuation of the Thousand and One Nights, translated by Dom Chavis and edited by M. Cazotte, "but singularly disfigured," says Deslongchamps, "like the other Oriental Tales published by Cazotte;" in Caussin de Perceval's excellent edition of the Nights, published, at Paris, in 1806, vol. viii, and in Gauttier's edition, vol. vi. The learned Swede Gustav Knös published, at Goetingen, in 1806, a dissertation on the Romance of Prince Bakhtyār, and the year following the Arabic text, with a Latin translation, under the title of Historia Decem Vizirorum et filii Regis Azad-bacht. He also issued a translation in the Swedish language, at Upsal, in two parts, the second of which appeared in 1814. Of the Turki version M. Amédée Jaubert has furnished, in the Journal Asiatique, Mars 1827, t. x, an interesting account, together with a translation of one of the stories,† from the unique manuscript preserved in the Bodleian Library at Ox

* In 1792 an English translation of this work was published at Edinburgh, in 4 vols., under the title: Arabian Tales. Translated from the original Arabic into French; and from the French into English, by Robert Heron.

† An English rendering of the Turkī version of the story translated into French by M. Jaubert will be found at the end of Notes on Chapter VI, pp. 189-194.

b*

ford, which he describes as very beautifully written, the titles of the several tales and the names of the principal characters being in red ink. Unfortunately the manuscript is imperfect; at present it comprises 294 folia. M. Jaubert remarks that this Turki version is characterised by "great sobriety of ornament and extreme simplicity of style, and the evident intention. on the part of the translator to suppress all that may not have appeared to him sufficiently probable, and all that might justly be taxed with exaggeration."

There is another Oriental rendering, of which M. Deslongchamps was ignorant, in the language of the Malays, with whom the romance is said to be a great favourite, indeed they have at least two very different versions of its frame, if not of the subordinate stories. In Newbold's work on Malacca,* vol. ii, an outline is given of the leading story, or frame, of one Malay version, which exactly corresponds with that of the Persian original, excepting that for Āzād-bakht we find Zād-bokhtin, and that the minister's daughter, who is carried to the city by the King and in our version is nameless, is called Mahrwat. I am indebted to the courtesy of the learned Dr R. Rost, Librarian to the India Office, for the following particulars regarding two other Malay versions, from Van den Berg's account of Malay, Arabic, Javanese and other MSS., published at

*Political and Statistical Account of the British Settlements in the Straits of Malacca. By T. J. Newbold. 2 vols. London, 1839.

Batavia, 1877. One of these (p. 21, No. 132) is entitled "The History of Ghulam, son of Zād-bokhtān, King of Adān, in Persia," and the frame agrees with that of our version, as already sketched in the present section, excepting that the robber-chief who had brought up Ghulam (our Bakhtyar),* "learning that he had become a person of consequence," says Van den Berg, came to his residence to visit him, but finding him imprisoned, he was much concerned, and asked the King's pardon on his behalf, telling him at the same time how he had formerly found Ghulam in the jungle; from which the King knew that Ghulam was his son," and so on. The other version (p. 32, No. 179), though similar in title to the Persian original, History of Prince Bakhtyār," differs very considerably in the frame, which is thus analysed by Van den Berg: "This Prince, when his father was put to flight by a younger brother, who wished to dethrone him, was born in a jungle and abandoned by his parents. A merchant, Idris (Enoch), took charge of him and

66

* Mr J. W. Redhouse has kindly furnished me, as follows, with the various meanings attached to the word Ghulām; which in the Malay romance seems to be employed as a proper name: "Gulām (not Ghulām), an Arabic word, signifies 'a boy,' 'a lad.' The Persians have made it, in their language, signify 'a slave,' and thence 'a life-guardsman,' and 'a king's messenger;' whence any post-messenger who travels on horse-back' -or by rail, now, in some places: all these really mean 'a lad.' The Turks use the word in the first and last senses-'a lad,' 'a Persian post-courier.