temps passez par la veue tornent en obly, et memorie de homme ne puet mye tot retenir ne comprendre.”  From this passage and from the Latin text:  “Incipit itinerarius a terra Angliae ad partes Iherosolimitanas et in ulteriores transmarinas, editus primo in lingua gallicana a milite suo autore anno incarnacionis Domini m. ccc. lv, in civitate Leodiensi, et paulo post in eadem civitate translatus in hanc formam latinam.” (P. 33 of the Relation des Mongols ou Tartars par le frere Jean du Plan de Carpin, Paris, 1838).  D’Avezac long ago was inclined to believe in an unique French version.  The British Museum, English MS. (Cott., Titus.  C. xvi.), on the other hand, has in the Prologue (cf. ed. 1725, p. 6):  “And zee schulle undirstonde, that I have put this Boke out of Latyn into Frensche, and translated it azen out of Frensche into Englyssche, that every Man of my Nacioun may undirstonde it...."[18]

But we shall see that—­without taking into account the important passage in French quoted above, and probably misunderstood by the English translator—­the English version, a sentence of which, not to be found in the Latin manuscripts, has just been given, is certainly posterior to the French text, and therefore that the abstract of Titus C. xvi, has but a slight value.  There can be some doubt only for the French and the Latin texts.

Dr. Carl Schoenborn[19] and Herr Eduard Maetzner,[20] “respectively seem to have been the first to show that the current Latin and English texts cannot possibly have been made by Mandeville himself.  Dr. J. Vogels states the same of unprinted Latin versions which he has discovered in the British Museum, and he has proved it as regards the Italian version."[21]

“In Latin, as Dr. Vogels has shown, there are five independent versions.  Four of them, which apparently originated in England (one manuscript, now at Leyden, being dated in 1390) have no special interest; the fifth, or vulgate Latin text, was no doubt made at Liege, and has an important bearing on the author’s identity.  It is found in twelve manuscripts, all of the 15th century, and is the only Latin version as yet printed."[22]

The universal use of the French language at the time would be an argument in favour of the original text being in this tongue, if corrupt proper names, abbreviations in the Latin text, etc., did not make the fact still more probable.

The story of the English version, as it is told by Messrs. Nicholson and Warner, is highly interesting:  The English version was made from a “mutilated archetype,” in French (Warner, p. x.) of the beginning of the 15th century, and was used for all the known English manuscripts, with the exception of the Cotton and Egerton volumes—­and also for all the printed editions until 1725.  Mr. Nicholson[23] pointed out that it is defective in the passage extending from p. 36, l. 7:  “And there were to ben 5 Soudans,” to p. 62, l. 25: