The Complete Works of Archimedes, Vol. 2

bable to me at this moment is that the Laurentianus in question is the product of a fifteenth-century copyist who had as a model a manuscript of the ninth or tenth century and sought to imitate it scrupulously, even in the tracing of the letters. He therefore arrived at the very same opinion to which I also had arrived by another path. Nor did Wilhelm Gardthausen, to whom I had sent the same image, think this opinion should be rejected, although he himself preferred to attribute the codex to the eleventh century based solely on the form of the letters, unless documents could be sought elsewhere to confirm that opinion; which is what I seem to have done here. Most troublesome, as both those most expert palaeographers have warned, is the form of the letter φ, which appears to have been written in a single stroke (see plate, lines 1, 3, 4, 5, 6, etc.), a thing which, excluding the form $φ$ [distinctive horizontal stroke through the central loop], does not occur before the fifteenth century (Gardthausen: Griech. Palaeogr. p. 208). Furthermore, this opinion is favored by the fact that the writing, which at the beginning of the codex is most diligent and neat, becomes more negligent toward the end and presents a less ancient appearance, and the accents, which are commonly omitted, are here and there, and indeed more toward the end, more frequent, so that the scribe's habit seems sometimes to have overcome his intention of religiously expressing the exemplar. Finally, the membranes themselves, both in kind and species, betray a fairly recent time.

Having weighed all these things, it must be concluded that the Florentine codex is not the codex of Valla, but that the latter is the common source of the codices FBC. Now, regarding our codex F itself, one must accept what Angelo Poliziano writes from Venice to Lorenzo de' Medici in the year 1491 (see Fabronius: Life of Lorenzo II p. 285):

In Venice I have found some books of Archimedes and Hero the mathematician, which we lack... and other good things. So much so that Pope John has something to write for a while.

Since Giorgio Valla was teaching in Venice from the year 1486 to 1499 (Neue Jahrb. Suppl. XII p. 377), and the codex of Valla, as we learned above, contained besides Archimedes also fragments of Hero, we can scarcely doubt that Angelo Poliziano [obtained?] the codex of Valla.