uges the Avignon anti-popes were recognized by some
persons to the very last, the latest date at which these passages could
have been written is the year 1417. It is not easy to understand how it
was possible for Caxton to leave uncorrected these references to a state
of things which he must have known had long ceased to exist. The only
explanation of the fact seems to be that, as has been suggested above,
he sent his many years old MS. to the press without going over it again.
It may be remarked that one of the Avignon passages does not occur in
the text as printed by Michelant. As it would be absurd to suppose that
it was introduced by Caxton himself, the inference is clear that his
copy of the original work was fuller than that contained in the Paris
MS. Probably Caxton may have added a few lines here and there--the
mention of certain English towns and fairs on pp. 18-19, and that of
English bishoprics on p. 23, for instance, were most likely inserted by
him. But by far the greatest portion of the matter which is peculiar to
Caxton's form of the dialogues may be confidently ascribed to his
original, on account of the frequent occurrence of passages in which,
while the French is quite correct, the English translation shows
imperfect understanding of the sense.

One of the most remarkable differences between Caxton's form of the
dialogues and that which is preserved in the Paris MS. consists in the
transposition of several of the sections in that portion of the work to
which the title 'Le Livre des Mestiers' is most properly applicable (pp.
24-44 of Caxton's edition). In both versions the sections in this
portion are arranged in the alphabetical order of the Christian names of
the persons referred to; but the names connected with particular
employments are not always the same in the two versions. Thus in
Michelant the bowyer is called Filbert, in Caxton he is Guillebert; in
Michelant the carpenter is Henri, in Caxton Lambert; in Michelant the
tiler is Martin, in Caxton Lamfroy; and so on. The resulting
transpositions render it somewhat difficult at first sight to perceive
the substantial identity of the matter in the two books. If an editor
wished to print Caxton's text and that of the Paris MS. in parallel
columns, he would need to have recourse to the ingenious device adopted
by Professor Skeat in the Clarendon Press edition of the three
recensions of _Piers Plowman_; that is to say, all the sections in which
the nam