had no connection with each other, which had even been substituted for each
other. The result was a fabricated text, full of contradictions naturally.
But since the edition issued by M. Jannet, the well-known publisher of the
Bibliotheque Elzevirienne, who was the first to get rid of this patchwork,
this mosaic, Rabelais' latest text has been given, accompanied by all the
earlier variations, to show the changes he made, as well as his
suppressions and additions. It would also be possible to reverse the
method. It would be interesting to take his first text as the basis,
noting the later modifications. This would be quite as instructive and
really worth doing. Perhaps one might then see more clearly with what care
he made his revisions, after what fashion he corrected, and especially what
were the additions he made.

No more striking instance can be quoted than the admirable chapter about
the shipwreck. It was not always so long as Rabelais made it in the end:
it was much shorter at first. As a rule, when an author recasts some
passage that he wishes to revise, he does so by rewriting the whole, or at
least by interpolating passages at one stroke, so to speak. Nothing of the
kind is seen here. Rabelais suppressed nothing, modified nothing; he did
not change his plan at all. What he did was to make insertions, to slip in
between two clauses a new one. He expressed his meaning in a lengthier
way, and the former clause is found in its integrity along with the
additional one, of which it forms, as it were, the warp. It was by this
method of touching up the smallest details, by making here and there such
little noticeable additions, that he succeeded in heightening the effect
without either change or loss. In the end it looks as if he had altered
nothing, added nothing new, as if it had always been so from the first, and
had never been meddled with.

The comparison is most instructive, showing us to what an extent Rabelais'