Reviewing Steele'sJ. M. contribution to Ptolemy in perspective, Mercier writes that the paper is “spoiled by Steele's acceptance of Theon's ‘observation’ in Alexandria of the solar eclipse of 364 June 16, but here he simply follows a habit that goes back to Fotheringham”, and he goes on to state categorically but mistakenly that Theon “never reported his observations as such”, and that this is “quite clear to any one who troubles to read the whole Greek text of Theon's commentary on Book 6 of the Almagest”. Fotheringham certainly did read it, since he provides a page reference to the Basel edition (p. 322) and loosely translates the crucial passage, which is close to the beginning of Theon's discussion of the eclipse; see FotheringhamJ. K., “A solution of ancient eclipses of the Sun”, Monthly notices of the Royal Astronomical Society, lxxxi (1920), 104–26, p. 114. Theon's observation report is omitted in another version of the calculation of the eclipse, existing both as an independent text and as an insertion in Theon's Little commentary on the Handy Tables, edited in TihonA., “Le calcul de l'éclipse du Soleil du 16 juin 364 p.C. et le ‘Petit Commentaire’ de Théon”, Bulletin de l'Institut Historique Belge de Rome, xlvi–xlvii (1976–77), 35–79.
RomeA., “The calculation of an eclipse of the Sun according to Theon of Alexandria”, in Proceedings of the International Congress of Mathematicians, Cambridge, Mass., 1950 (Providence, 1952), i, 209–19. Of course Fotheringham could not have seen the text in Laur. plut. 28.18. He perhaps ought to have been troubled by the conflicting phrase about “seasonal and apparent time”, but he chooses to translate this inaccurately as “civil and apparent time”, thus concealing the clue.
7.
It is not clear to me whether Theon intends to indicate that the time of the beginning was observed more precisely than the time of the end, as Rome, “Calculation” (ref. 6) asserts; the adverb (“approximately”, or literally “most nearly”) does not signify roughness but usually that a stated number is as close to the accurate value as the chosen precision allows.
