The Scholarship of the Palimpsest

As far as Archimedes is concerned the results which may be expected from the project can be summarised as follows. It is obvious that they will not be of equal importance for all the treatises. One does not anticipate substantial improvements to the text of the four that were already well known, having been transmitted by what in effect are two sources. The first of these was a Byzantine MS of probably the ninth century, which was still extant in the Renaissance and belonged to the Venetian humanist Giorgio Valla (d.1501); though lost now, it can be reconstructed from four copies made in the fifteenth and sixteenth centuries. It is referred to as A, and the copies are D (MS Florence, Laur. 28.4), E (MS Venice Marcianus gr. 305), G (MS Paris gr.2360) and H (MS Paris gr.2361). Of these D is especially interesting, because the scribe seems to have tried to imitate the archaic script of his exemplar.¹⁶ The second source is a Latin version by William of Moerbeke, made in 1269 when he was living in Viterbo. His autograph is MS Vaticanus Ottob.lat. 1850. The translation is extremely literal and in most passages permits a precise reconstruction of the Greek text that he had in front of him. In these treatises Heiberg noted¹⁷ that the palimpsest, while exhibiting a number of errors peculiar to itself, nevertheless was of value for establishing the text in many passages.¹⁸ This means that it is worth looking closely at the points in the text where Heiberg declared himself unable to determine what the palimpsest reads. However, any improvements in the text of the four previously known treatises are bound to seem modest in comparison with what can be achieved for the three that made Heiberg's discovery so notable.

First comes On floating bodies, which had been available in the Latin version. Heiberg recovered most of the Greek. It is now possible to make further advances. As a minor example I cite a passage from folio 169 verso.¹⁹ The context deals with an object being immersed in water, so that part of it is moistened. Heiberg's Greek includes the word temnesthai �cut�, which neither the context nor the faintly visible script seem to justify. After lengthy examination of the passage I realised that the correct transcription is brekhesthai �moistened�, and on checking the Latin was delighted to find that it indeed offers humectetur, which corresponds perfectly. But such refinements are not the whole story. Heiberg failed to read some folios because they were already too badly damaged. Significant progress is possible, partly because faint traces that by themselves scarcely permit reliable transcription can be seen as parts of words which can be predicted from the Latin version.

Back

Continue

¹⁶ It is the subject of a paper by J. Irigoin, Bollettino della Badia greca di Grottaferrata 54 (2000) pp.307-317. I do not follow his statement that the scribe of D began his work on the verso of the initial folio, leaving the recto blank because his exemplar was in very poor condition. The script on the recto is very faint, but can clearly be identified as the hand of John Thessalos Scutariota, as was proposed by Charles Graux in a letter to Heiberg, which the latter quotes in Philologus 42 (1884) p.435. I examined the MS in Florence on June 3 2003. I infer that the scribe's patron requested a copy which would resemble the old codex as far as possible, and that this wish was communicated just after the scribe had begun his work, so that only the first page is in his normal hand.

¹⁷ III p.lxxxviii-xc.

¹⁸ At III pp.lv-lvi he cites cases where the version has a superior reading to that of MS A; some of them were confirmed by the palimpsest.

¹⁹ II p.408 line 13.