Complete Extant Works of Hero of Alexandria, Vol. 2

XXI

INTRODUCTION.

errors, to the extent that he relegated good readings of the manuscript to the footnotes. For example, p. 23 of the separate edition, note 4, p. 66 notes 1 and 2, p. 93 note 2, p. 107 note 1, and others. In general, his critical work is directed less at the technical content, which as an engineer might have been closer to him, than at the linguistic side of the text, although he was not always entirely successful in the punctuation, for example. That does not mean much, however, because the French editor is clearly more proficient in the modern language than in the older one. One could excuse the fact that he did not change anything in his text for technical reasons, in consideration of the first edition being based on a single manuscript. Unfortunately, however, his translation also offers no indication in some difficult questions, for example—to pick just one—the manufacture of the screw nut (III, 21), that he had become clear about the technical aspects in this paragraph. Yet these small criticisms should not diminish de Vaux's merit in having been the first to turn to this difficult subject and to have brought to light and made accessible the writing long lost in the original.

Following this, I would like to mention a case from the first edition that shows how one can be tempted by harmless reading errors to overlook the obvious correct reading and to cause misunderstandings. In Mech. I, 24, a name appears which is written Bōsidōmōs in the Leiden manuscript (only the ō are written fully). Although the first editor correctly resolved this as Posidonius in the Arabic text, since it is a common scribal error in Arabic manuscripts to contract n + j into m, just as conversely m is often expanded into n (b etc.) + j, he nevertheless made Praxidamas out of it in the translation, which Clermont-Ganneau only later conjectured back into the correct Posidonius. Through the three remaining