The Complete Works of Archimedes, Vol. 2

PROLEGOMENA.

Concerning the Archimedean codices.

When, two years ago, I was inquiring into the correct assessment of the Archimedean codices—the greater and more important part of which discussion is [found in] Quaestiones Archimedeae (Copenhagen 1879, chapter VI)—it could not escape me that the Florentine codex is the most excellent of all, a fact which the reviewer of the Torelli edition (Jenaer Literaturzeitung 1795, p. 610 ff.) had already understood, and [that it is] closely linked by the tightest bond to the very ancient codex of Giorgio Valla. Therefore, since I had found out nothing concerning the age of this codex, except that Bandini had assigned it to the XIII century, I was necessarily led to conclude that the Florentine codex was that very ancient codex of Valla from which the Paris codices B and C were copied. But difficulties remained, both in the readings of individual passages and, primarily, in explaining how that codex had finally arrived at the Laurentian library; also, that letter prefixed to the first book of On the Sphere and Cylinder seemed to be better preserved in codex B (Quaest. Arch. p. 130). These difficulties I then attempted to explain as best I could (Quaest. Arch. p. 132 ff.). But even then I had begun to doubt the judgment of Bandini. And after I myself examined the Florentine codex and collated it diligently, I persuaded myself that this codex could by no means be the same as Valla's codex, but rather a copy of it, transcribed with the greatest diligence, such that the scribe also imitated the form of the letters with generally laborious care. Therefore, I have decided to retract this entire passage.