¹ Hanses, M., ‘The pun and the moon in the sky: Aratus’ ΛΕΠΤΗ acrostic’, CQ 64 (2014), 609–14CrossRefGoogle Scholar. In order to keep notes to a minimum, I refer the reader to Hanses's second footnote, where one will find an up-to-date and very full bibliography of the growing scholarship on this and other ancient acrostics.

² Danielewicz, J., ‘One sign after another: the fifth ΛΕΠΤΗ in Aratus’ Phaen. 783–4?’ CQ 65 (2015), 387–90CrossRefGoogle Scholar.

³ The text is that of D. Kidd (ed.), Aratus Phaenomena, Edited with Introduction, Translation and Commentary (Cambridge, 1997). All translations from Greek are my own.

⁴ This is the order in which they were found by modern investigators. Cf. Danielewicz (n. 2): ‘For centuries, the students of the Phaenomena were aware only of what was visible while reading horizontally, that is, of the two instances of ΛΕΠΤΗ inserted, respectively, in lines 783 and 784.’

⁵ Even if a reader suspects its presence earlier than this, by the appearance of λεπτή in 784 (my 2), only the letters ΛΕ- of the vertical acrostic have been seen, making it extremely unlikely that most readers would have spotted it. The exception, I suppose, would be a reader who habitually scanned vertically down the first letters of lines looking for acrostics ahead of reading the lines themselves first. There may have been such readers, but I doubt there were many.

⁶ Pace Feeney, D. and Nelis, D., ‘Two Virgilian acrostics: certissima signa? ’, CQ 55 (2005), 644–6CrossRefGoogle Scholar, who interpret line 778 as signifying ‘look first at the edges’, that is, at the gamma-acrostic; similarly, Hanses (n. 1), 612. If this view is adopted, I suppose that a first-time reader instantly aware of this double meaning of ἑκάτερθε may find the gamma-acrostic before seeing the second λεπτή that lies in plain, but this scenario seems unlikely to me. Despite my disagreement on this point, Feeney and Nelis's wider discussion of the relationship between Aratus and Virgil seems impeccable to me and particularly valuable with its emphasis on the poets' expectation that their audience would pay attention to number and order in response to textual clues regarding acrostics. For another interpretation of ἑκάτερθε, see Danielewicz, J., ‘Vergil's certissima signa reinterpreted: the Arataean lepte-acrostic in Georgics 1’, Eos 100 (2013), 287–95Google Scholar, where it is taken as a hint to look for an acrostic-telestich combination in this passage.

⁷ That is perhaps not quite the way to put it since it points to a complementary way of reading the clue. Since 783, the first line of our passage, contains the ‘third day’ (περὶ τρίτον ἦμαρ) and we are told that the fourth comes from the third, we can surmise that the fourth λεπτή will start somewhere in that line.

⁸ CQ's anonymous referee remarks: ‘I did also wonder, without coming to a conclusion, if the “bow” could be a figure for something in the sky itself, e.g. the Centaur's bow.’ I mention this because the image of a bow had also suggested itself to me amidst my pondering. The coincidence of our impressions may lead someone else to a solution.

⁹ For the importance of numbers in the context of Aratus' acrostic, cf. Hanses (n. 1), 612: ‘More specifically, the moon is worth a third and a fourth look (αἱ μὲν τρίτῃ, αἱ δὲ τετάρτῃ, 781). This “numeral” element is probably not insignificant, considering how frequently references to “threes” and “fours” recur in the immediate proximity (781, 783, 786, 788, 792, 796, 806, 809, 810, 812).’

¹⁰ Volk, K., ‘Letters in the sky: reading the signs in Aratus' Phaenomena ’, AJPh 133 (2012), 209–40Google Scholar. If I am correct in discerning an actual gamma and delta (see just below) in the acrostic passage here, there is no need to modify her assertion (235) that ‘Aratus does not use the concept of heavenly writing openly, let alone draw an explicit parallel between his own letters and those of the sky’, because the letter forms here are not constellations as such, but it may be worth considering a connection between the possible delta I discuss below and the single instance of a letter-shaped constellation in the rest of the poem, the Δελτωτόν in line 235 (discussed by Volk [this note], 232–5).

¹¹ I discuss κεραία in the sense of a stroke of a letter form below.

¹² There really is no easy way for Aratus to have delineated the bottom stroke of a Δ, and it may be enough merely that the etas of the two instances of λεπτή occur on the same line. Jerzy Danielewicz helpfully suggested to me (per litt.) the argument above by observing that ἄμβλυνται, positioned as it is between the etas, may ‘excuse’ the incomplete letter.

¹³ In actuality, because the lunar month is a little more than 29.5 days long, ancient Greek months alternated between 29 and 30 days. But the round number is conventionally used as the length of a generic month for convenience. Thus Aristophanes' Lysistratus suffers cold and hunger ‘more than 30 days each month’ (Ach. 858–9, πλεῖν ἢ τριάκονθ’ ἡμέρας τοῦ μηνὸς ἑκάστου), and Lucian writes of his parasite (De parasito 15) that he ‘spends the 30 days of the month as holidays’ (ὁ δὲ παράσιτος τοῦ μηνὸς τὰς τριάκονθ’ ἡμέρας ἱερὰς ἄγει) rather than ‘the 29 or 30 days of the month, depending upon precisely which month’.

¹⁴ In the same work (fr. 52 Hultsch) compare ταῖς τοῦ Κ δύο κεραίαις, referring to the two diagonal strokes of the letter kappa (and, in the same passage, the reference to τὴν κάτω κεραίαν of the same letter).