A fuller study by Vogels (_Handschriftliche Untersuchungen über die Englische Version Mandeville's_, Crefeld 1891) brings to light a new fact: the two Bodleian MSS., E Museo 116 and Rawlinson D 99, contain an English translation (say L) made from a Latin text of the _Travels_. Vogels also shows that E is based on D, because the characteristic lacuna of D is filled in E by a passage which is borrowed from L and is not homogeneous with the rest of E. So far there is no conflict with the view of Nicholson and Warner. But, after adducing evidence in favour of the contention that C, D, and E are at base one translation, Vogels concludes that D derives from C, arguing thus: There is good evidence that C is a direct translation from the French, and if it be granted that a single translation from the French is the base of C and D, it follows that D derives from C. In short, the one party maintains that C is an expansion of D, the other that D is an abridgement of C; and this flat opposition results from the acceptance of common ground: that C and D represent in the main one translation and not two translations. To return to our interpolation: (1) Vogels's first piece of evidence that C, D, and E are at base one translation is the appearance in all of this interpolation, which is absent from the MSS. in French. But a passage so remarkable might spread from one to the other of two independent English texts; or if the interpolation originated in England in a MS. of the French text since lost, it might be twice translated. (2) Vogels assumes that the interpolation first appeared in type C. But C is the form in which it would be least likely to originate, because here the contradiction of statement is sharpest owing to the rendering at ll. 313-14: _and now I am comen hom_, which is peculiar to C (see the French). (3) If, in order to eliminate individual peculiarities, we take two MSS. of the D type—say Harley 2386 and Royal 17 C. XXXVIII—we find that their text of the interpolation is identical with that of E. This is consistent with Vogels's finding that the body of E derives from D; and it confirms the evidence of all the defective MSS. that the interpolation in this particular form was an integral part of the D type. (4) But between the text of the interpolation in D and that in C there are differences in matter, in sentence order, and in phrasing, which, while they do not exclude the possibility of interdependence, do not suggest such a relation. In D the passage is a naked attempt at authentication; in C it is more artfully though more shamelessly introduced by the touch of piety conventional in epilogues. And as the signs of a French original that appear in C are absent from D, it is unlikely that the text of the interpolation in C derives from D.