Archimedes is by common consent one of the greatest of all mathematicians, and his already secure reputation was further enhanced in 1907 when J.L. Heiberg, professor of Greek in the University of Copenhagen and an acknowledged expert on Greek mathematics, published a long article in a German periodical giving an account of a remarkable discovery made in Istanbul.1 A German colleague, H. Sch�ne, had drawn his attention to an entry in the printed catalogue of a manuscript collection in the Ottoman capital belonging to the Metochion of the Holy Sepulchre, the daughter-house of a famous monastery in Jerusalem.2 The author of the catalogue had noted that the lower script of the palimpsest MS 355 included a mathematical text, and he had the good sense to print a few lines of it that he was able to read. This tiny specimen was recognised by Heiberg as coming from Archimedes. Being engaged at the time in revising his edition of Archimedes' works3 he saw at once the potential importance of the palimpsest. After the failure of attempts to have it sent to Copenhagen on loan, a practice not uncommon in the nineteenth century, he went to Istanbul himself. His efforts were rewarded by a spectacular discovery; apart from portions of four known works by Archimedes the palimpsest contained substantial passages of the lost treatise Method and of the essay On floating bodies, the latter previously known only from a medieval version in Latin, plus a fragment of a minor work called Stomachion, another part of which is preserved in Arabic. Of these three items Method was the most striking, since it showed the author taking a vital step towards the invention of calculus.
The importance of the palimpsest can hardly be exaggerated; among those which have a lower script in Greek it is by far the most significant to have come to light. Though Heiberg was able to study it again in 1908, further examination was a desideratum, particularly when the ultra-violet lamp became available as a means of reading palimpsests, removing the temptation to use the drastic and indeed often disastrous technique, all too often employed in the nineteenth century, of painting the leaves with chemicals. But at some unknown date, probably after the First World War, the manuscript left Istanbul in circumstances which remain completely mysterious and eventually turned up in the possession of a French family.
The possibility that a more accurate and complete transcription of the Archimedes text might be feasible became a virtual certainty in 1971 when, acting on a suggestion of my friend Professor G.J. Toomer of Brown University, I visited the Cambridge University Library and inspected a single palimpsest leaf with the shelf-mark MS 1879.23.4 This was stated to contain an unidentified mathematical text. Transcription of almost all the text proved easy enough. The occurrence of a rare technical term proved that the text was Archimedes, and examination of the script showed at once that this was a leaf from the Istanbul palimpsest.
1 Hermes 42 (1907) pp.235-303.
2 A. Papadopoulos-Kerameus, Hierosolymitike Bibliotheke volume IV (St.Petersburg 1899) pp.329-331.
3 Archimedis Opera omnia cum commentariis Eutocii (Leipzig 1880-81); the revised edition appeared in 1910-15.
4 The leaf was acquired from the executors of the eminent German biblical scholar C. Tischendorf, who had seen the manuscript when visiting the library in the 1840s; see his Reise in den Orient (Leipzig 1846), translated into English by W.E. Shuckard (London 1847), where it is mentioned on p.274; see also P.E. Easterling, Scriptorium 16 (1962) pp.302,307. It is hard to avoid the conclusion that Tischendorf stole it.