The names also do not always agree in both versions. Thus we find «Generyse», «S» 1139, for «Garin», «F» 438; «Mapyn», «S» 2325, for «Maubrun», «F» 3046; «Alagolofur», «S» 2135, for «Agolafre», «F» 4290 or «Golafre», «F» 4267, 4383; «Bryer», «S» 2604, for «Basin», «F» 3313; «Maragounde», «S» 1563, for «Marabunde», «F» 2196; «Boloyne», «S» 3238, for «St. Denis», «F» 6199; «Barokke», «S» 2939, and «Espiard», «S» 2145, are not mentioned at all in the French «Fierabras», nor does« Belmore», «S» 3122, occur in the «Fierabras», either in the corresponding passage, «F» 5867, or elsewhere. On the fact that the names of the twelve peers (see above, p. xxvii) differ in the «Sowdan» from those mentioned in the «Fierabras», too much stress need not, I think, be laid, as it might be explained by the simple inadvertence of the composer. The poet in freely reproducing his source, which he generally followed pretty closely as far as relates the course of events, well remembered the names of the principal French knights; but having forgotten those of less important characters, some of whom do not appear again in the poem, and being obliged to fill up their number of twelve, might have placed any names which he remembered having met with somewhere [p-xxxi] as included in the list of the douzeperes. By an oversight he omitted to mention Richard, whom however we see appear afterwards.[63] Similarly the names of «Laban» and «Ferumbras» for «Balan» and «Fierabras» afford no convincing proof of the impossibility of the French «Fierabras» being the original of the second part of the «Sowdan», as the poet, having found those spellings in the «Destruction», the source of the first portion of his romance, might simply have retained them for the whole poem. But reviewing all the facts of the case, and taking into account those passages which relate incidents omitted in the «Fierabras», and which the author of the «Sowdan» therefore cannot have taken from that poem—and further taking into account the several differences between the two versions, which, it may be admitted, generally speaking, are only slight ones—the French «Fierabras», «i. e.» the version edited by MM. Krœber and Servois, which represents the group «w» (see before, p. xix, footnote), cannot have been the original of the second part of the «Sowdan». Proceeding now to a comparison of the «Sowdan» with the Escorial MS.,[64] we have not found any passage where «S» differing from «F» agrees with «E», as «E» and «F» generally have in those places the same reading. Therefore the Escorial MS. cannot be regarded as the original of the «Sowdan».