purposely omits a portion of the narrative which comes between these, as pointed out by the footnote on p. 5. This is, however, not quite all. The scribe was determined not to lose the curious account of the trees which grew every day while daylight lasted, but disappeared every night; and, thinking that this short account would seem out of place if merely added at the end of the Letters, boldly inserted it in the middle; at ll. 111–136. If this be not quite the right history of the matter, it is perhaps as nearly so as we can guess, and is quite sufficient for the purpose of understanding the present state of the text.
§ 6. I have said that the French romance follows, in the main, one form of the story, and the English romances another. The French romance is all printed, as explained above, and may now be dismissed, as we have nothing more to do with it. The three English fragments are all connected, and are founded mainly on the same Latin version. Repeating from p. xxxvii of my Introduction to William of Palerne and Alisaunder, I may remind the reader that the principal basis of these fragments is the Greek text known as the Pseudo-Callisthenes, whence three principal Latin versions are derived. These are (1) that by Julius Valerius; (2) the Itinerarium Alexandri (relating to Alexander's wars); and (3) that by the Archpresbyter Leo, which is also known as the "Historia de preliis." It is with the third of these that the English fragments have most to do. This version begins with the words—"Sapientissimi egiptii scientes mensuram terre;" and an edition of it was printed in 1490, which has been my guide throughout, and from which I have given numerous citations. It is from this edition that the Latin text is quoted which appears at the foot of pages 1–42.
§ 7. All three English fragments are founded mainly on this Latin version, but the manner of translation is not the same in all. Fragment C may be taken first, as it is much the easiest to understand. This is a close translation of the Latin, with a brief original prologue of 22 lines only. It is of great length, extending to 5680[1] and is only slightly imperfect at the end.[2] As a result, it contains