On several occasions Arabic words are given, but are not always recognizable, owing perhaps to the carelessness of copyists in such matters. Thus, we find (p. 50) the names (not satisfactorily identified) of the wood, fruit and sap of the balsam plant; (p. 99) of bitumen, "alkatran" (_al-Katran_); (p. 168) of the three different kinds of pepper (long pepper, black pepper and white pepper) as _sorbotin_, _fulful_ and _bano_ or _bauo_ (_fulful_ is the common Arabic word for pepper; the others have not been satisfactorily explained). But these, and the particulars of his narrative for which no literary sources have yet been found, are too few to constitute a proof of personal experience.
Mandeville, again, in some passages shows a correct idea of the form of the earth, and of position in latitude ascertained by observation of the pole star; he knows that there are antipodes, and that if ships were sent on voyages of discovery they might sail round the world. And he tells a curious story, which he had heard in his youth, how a worthy man did travel ever eastward until he came to his own country again (p. 183). But he repeatedly asserts the old belief that Jerusalem was in the centre of the world (79, 183), and maintains in proof of this that at the equinox a spear planted erect in Jerusalem casts no shadow at noon, which, if true, would equally consist with the sphericity of the earth, provided that the city were on the equator.
The sources of the book, which include various authors besides those whom we have specified, have been laboriously investigated by Dr Albert Bovenschen[8] and Dr G. F. Warner,[9] and to them the reader must be referred for more detailed information on the subject.
The oldest known MS. of the original--once Barrois's, afterwards the earl of Ashburnham's, now Nouv. Acq. Franç. 4515 in the Bibliothèque Nationale, Paris--is dated 1371, but is nevertheless very inaccurate in proper names. An early printed Latin translation made from the French has been already quoted, but four others, unprinted, have been discovered by Dr J. Vogels.[10] They exist in eight MSS., of which seven are in Great Britain, while the eighth was copied by a monk of Abingdon; probably, therefore, all these unprinted translations were executed in this country. From one of them, according to Dr Vogels,[11] an English version was made which has never been printed and is now extant only in free abbreviations, contained in two 15th century MSS. in the Bodleian Library, Oxford--MS. e Museo 116, and MS. Rawlinson D. 99: the former, which is the better, is in Midland dialect, and may possibly have belonged to the Augustinian priory of St Osyth in Essex, while the latter is in Southern dialect.