Ancient and Mediaeval Science 19 



the Latin edition of Tartaglia^ (Venice 1543) — the first printed Archi- 

 medes in any language — ; his translation of both books was printed by 

 Troianus Curtius (Venice 1565) and by Federico Commanding (Bo- 

 logna 1565) . The Greek text of the "floating bodies" was lost until 1906. 

 In that year the Danish philologist, J. L. Heiberg, discovered it in a 

 Constantinople palimpsest below a twelfth to fourteenth century eucho- 

 logionJ The same palimpsest concealed other Archimedian texts, the 

 most precious of all being the Method (IcpoStov), the existence of which 

 was known only through a remark of Suidas (X-2).^ That method is 

 one of the most important books of antiquity. We have it!, but remem- 

 ber that it was preserved only in the most erratic way — as a palimpsest 

 — , that is, it was preserved in spite of its being deliberately cancelled, 

 and that its recovery happened only within our own lifetime, in 1906. 

 An Archimedian monograph on the regular heptagon was preserved in 

 the Arabic translation of Thabit ibn Qurra (IX-2) and this was dis- 

 covered in a Cairo MS. and published in 1926 by Carl Schoy.^ 



In other words, lost treatises of Archimedes were revealed only in 

 1906 and 1926. It is possible that other lost treatises may still be dis- 

 covered, chiefly in the second manner. The Greek palimpsests have 

 been pretty well examined and there is little hope of repeating Heiberg's 

 stroke of genius and luck, but there is much hope on the contrary of find- 

 ing Arabic translations of lost Greek scientific books, because many Ara- 

 bic libraries are still unexplored and many Arabic MSS, undescribed. 

 Some of the classics of Greek science have been revealed in that way, 

 notably books V to VII of Apollonios' Conies and various treatises of 

 Galen.i" 



" The Latin tradition of some other Archimedian treatises was different. Nicho- 

 las V (pope from 1447 to 1455), one of the early patrons of humanism, founder 

 of the Vatican Library, caused an Archimedian MS. to be translated into Latin by 

 one Jacopo da S. Cassiano of Cremona. A copy of that translation was made c. 

 1461 by Regiomontanus, who added marginal glosses derived from Greek MSS. 

 Regiomontanus' copy, preserved in Nuremberg, was the source of the Latin version 

 added to the Greek princeps by Thomas Gechauff (Basel 1544). 



''A palimpsest is a "rewritten" MS., the first writing having been erased to make 

 room for the new one. An euchologion is a book of the Orthodox Church con- 

 taining liturgies, etc. As writing materials (parchment or paper) were expensive 

 and difficult to obtain, monks would rub off texts of no interest to them to replace 

 them by the texts which they needed. We would do the same under similar cir- 

 cumstances. Chemical and optical means make it possible to read the erased text. 



® SumAS remarked that Theodosios of Bithynia (I-l B.C.) wrote a commentary 

 on the Method. Three propositions are quoted from it in the Metrica of Heron 

 OF Alexandria, but the Metrica itself was discovered only in 1896, in a Constanti- 

 nople MS., by R. Schone; it was first published in 1903 by the discoverer's son, 

 Hermann Schone. 



*Carl Schoy: Graeco-Arabische Studien (Isis, 8, 21-40, 1926). 



^° The Arabic translation of books V to VII of the Conies by Thabit ibn Qurra 

 (IX-2) was revised by Abu'l-Fath Mahmud ibn Muhammad al-Isfahani (X-2); 

 it was first published in Latin version by Abraham Ecchellensis and Giacomo 

 Alfonso Borelli (Florence 1661), then again in Edmund Halley's monumental 

 edition of Apollonios (Oxford 1710). The seven books of Galen's anatomy were 



