¹ Volk, K., Manilius and his Intellectual Background (Oxford, 2009), 29–57CrossRefGoogle Scholar provides a detailed discussion of Manilius’ description of the celestial sphere.

² The text of Manilius is quoted from Goold, G.P., M. Manilii Astronomica (Stuttgart and Leipzig, 1998 ² [1985])Google Scholar. I have repunctuated line 463; my reasons for doing so will be explained shortly.

³ After being a regular fixture in the early printed editions, distinguere, a humanist conjecture, gave way to the results of recensio and was replaced, in editions of the nineteenth and twentieth centuries, by disiungere, the reading of the primary manuscripts; disiungere was printed by Jacob, F., M. Manili Astronomicon Libri Quinque (Berlin, 1846)Google Scholar; Bechert, M., Marci Manili Qui Fertur Astronomicon Libri Quinque, in J.P. Postgate, Corpus Poetarum Latinorum, fasc. 3 (London, 1900)Google Scholar; Housman, A.E., M. Manilii Astronomicon Liber I (Cambridge, 1937 ²)Google Scholar and M. Manilii Astronomica (Cambridge, 1932); Breiter, T., M. Manilius: Astronomica (Leipzig, 1908)Google Scholar; van Wageningen, J., M. Manilii Astronomica (Leipzig, 1915)Google Scholar; and Feraboli, S., Flores, E. and Scarcia, R., Manilio: Il poema degli astri (Astronomica) (Milan, 1996), vol. 1Google Scholar. Housman printed disiungere in both of his editions; in his commentary, however, he expressed the reservation that distinguere seems to be said more appropriately of shapes (formas), i.e. on the implied assumption that disiungere is more appropriate to the stars in general, whereas distinguere is the precise term for marking off with stars the distinct patterns recognized as constellations. As the mot juste for denoting the action of marking off the constellation figures with stars, distinguere has been convincingly defended by Goold, G.P., ‘Adversaria Maniliana’, Phoenix 13 (1959), 93–112CrossRefGoogle Scholar, at 108–9, and Caldini, R. Montanari, ‘Le constellazioni in Manilio, ovvero l'imperfezione perfetta’, in D. Liuzzi (ed.), Manilio fra poesia e scienza (Galatina, 1993), 55–78Google Scholar, at 66–7 = A&R 38 (1993), 18–41, at 29–31, which is an important contribution to the understanding of 1.456–82; she compares Cic. Arat. fr. 33.161 et [natura] uario pinxit distinguens lumine formas (J. Soubiran, Cicéron: Aratea, fragments poétiques [Paris, 1993²]). D. Liuzzi prints distinguere in M. Manilio: Astronomica, Libro I (Galatina, 1995²).

⁴ G.P. Goold's indispensable translation, Manilius: Astronomica (Cambridge, MA and London, 1977), was the starting point for my translation. There are two significant differences between the two. First, in 457 Goold translates signa as ‘stars’; this must be a slip for ‘constellations’. Manilius begins his catalogue of constellations by announcing that he will tell of the signa (1.255–6) everywhere in the heavens; he concludes the catalogue by repeating the keyword signa (1.457). The second difference, at lines 463–4, will be discussed above. The rendering of ultima summis | creduntur (466–7) as ‘the out-facing surface [of the figure] is imagined from the in-facing surface’ is intended to represent the two perspectives for viewing the constellations, terrestrial and cosmic, i.e. from outside the celestial sphere looking down on the heavens. In the cosmic perspective, as the representations of the constellations on the Farnese Globe and the Mainz globe show, the figures can be viewed from the backside (ultima) rather than the frontside (summis); for illustrations see Künzl, E., ‘Ein römischer Himmelsglobus der mittleren Kaiserzeit: Studien zur römischen Astralikonographie’, JRGZ 47 (2000), 495–594Google Scholar: on the Mainz globe, Gemini and Orion (plate 36.1), Serpent-Holder (36.2) and Aquarius (36.4); on the Farnese Atlas, Perseus (43.1) and Serpent-Holder (45.1). Dekker, E., Illustrating the Phaenomena: Celestial Cartography in Antiquity and the Middle Ages (Oxford, 2013)Google Scholar, provides detailed discussions of the astronomical theory and data on which the Farnese globe (84–102, 111–15) and the Mainz globe (69–80, 106–111) were constructed. If, as G. Thiele argued in Antike Himmelsbilder (Berlin, 1898), 45–7, Manilius used a celestial globe, then the poet had good reason to mention both views of the figures; his use of a globe is regarded as probable by Dekker, 77 and 97. This explanation of ultima summis | creduntur makes unnecessary conjectures intended to improve the sense. Ellis, R., Noctes Manilianae (Oxford, 1891), 10Google Scholar, proposed infima for ultima, apparently meaning ‘the lower part is inferred from the top part’. This recommendation would not work very well with a bi-form creature like Capricorn, or with Taurus, which has a top but no bottom; and, in any case, the linea defines the shape of the whole figure, top and bottom. The same proposal was made independently by Campbell, A.Y., ‘Manilius I.466–8 and 515–17’, CQ 7 (1957), 186–7CrossRefGoogle Scholar, at 186. Garrod, H.W., ‘Two editions of Manilius. (With some notes on books I and II)’, CQ 2 (1908), 123–131CrossRefGoogle Scholar, at 130, conjectured intima, meaning the interior of the figure, but it is difficult to see how this differs from media in relation to extremis, i.e. what lies in between the outline of the figure. See also Montanari Caldini (n. 3), 64 n. 37 = A&R 27 n. 37.

⁵ Manilius’ scenario of a superabundance of stars that would generate excessive heat and cause a universal conflagration appears to be without parallel, with the exception of a similar scenario in 5.740–5, where he imagines a conflagration precipitated by the stars of the Milky Way; see W. Hübner, Manilius, Astronomica, Buch V (Berlin and New York, 2010), 2.449 on 5.744. Scaliger, J., In M. Manili Quinque Libros Astronomicon Commentarius et Castigationes (Heidelberg, 1590), 51Google Scholar, cited Aristotle's Meteorologica 1.35 (340a) to illustrate Manilius’ hypothetical conflagration. But there Aristotle, who maintains that the aether and the stars are not composed of fire, is making the argument that if the celestial bodies and the intervals between them were in fact composed of fire, then the other elements would have disappeared. The story of Phaethon's errant ride in the chariot of the sun and the ensuing conflagration illustrates a different cosmic event, the sun's deviating from its proper course (Man. 1.735–49 and 4.834–7; see P. Glauthier, ‘Repurposing the stars: Manilius, Astronomica 1, and the Aratean tradition’, AJPh 138 [2017], 267–303, at 285–90). Montanari Caldini (n. 3), 76 = A&R 39, suggests that the scenario of a universal conflagration caused by a superabundance of stars is original to Manilius. A related notion is present in Ovid's explanation of Jupiter's decision not to destroy humankind with his thunderbolts for fear that they would set the heavens ablaze: sed timuit ne forte sacer tot ab ignibus aether | conciperet flammas longusque ardesceret axis (Met. 1.254–5).

⁶ On the meaning of the terms stella, astrum, sidus and signum in Manilius, Housman (n. 3 [1937²]), on 1.465 offers a concise and reliable formulation: ‘nam apud Manilium stella corpus lucidum significat, signum figuram e pluribus stellis formatam quam hodie appellamus constellationem, astrum et sidus utrumuis.’ For a more detailed discussion see Liuzzi, D., ‘Stella, astrum, signum, sidus negli Astronomica di Manilio’, CCC 7 (1986), 43–51Google Scholar; and for an overview of the use of these terms in Latin prose and poetry, A. Le Bœuffle, Le noms latins d'astres et de constellations (Paris, 1977), 5–40, and on ignis, 41. In her review of Le Bœuffle's monograph, Caldini, R. Montanari, ‘La terminologia latina dei corpi celesti’, A&R 24 (1979), 156–71Google Scholar conducts an independent analysis of the terminology, with brief mention of Manilius, 1.465 at 166. There can be no doubt that ignes in 1.466 means individual stars; cf. 5.733, where ignes refers to the countless stars that fill the heavens.

⁷ Bentley, R., M. Manilii Astronomicon (London, 1739)Google Scholar. Housman (n. 3 [1937²]) regarded Bentley's punctuation and conjecture as an improvement in sense but offered his own conjecture damnis, for flammis, a more plausible palaeographical modification of flammis that achieves the same sense, i.e. nature spared herself from catastrophe; he first proposed damnis in ‘Emendations in the first book of Manilius’, Journal of Philology 26 (1889), 60–3, at 62 = J. Diggle and F.R.D. Goodyear (edd.), The Classical Papers of A.E. Housman (Cambridge, 1972), 2.492–4, at 493. Without flammis, however, the text does not yield a clear explanation of nature's method of populating the firmament with stars.

⁸ This interpretation of subduxit is consistent with celant in 468.

⁹ Goold (n. 4) translates: ‘Whatever nature has removed from such fires she has subtracted from a burden to which she would have proved unequal.’ The translation makes the quidquid-clause the object of pepercit, renders pepercit as a synonym for subduxit and treats oneri as shared between pepercit and the participle succubitura. The chief difficulty here is that ‘subtracted’ is not an accurate rendering of pepercit; ‘to subtract’ is not the same thing as ‘to be sparing with’. In fact, on this interpretation, the meaning would be more accurately represented by the following, ‘Nature was sparing in whatever it removed from the fires for the burden’, i.e. nature's procedure was to be economical in removing stars and was trying to maintain the burden rather than reduce it, a sense opposite to the one required, namely that she was sparing in the number of stars that she placed in the heavens in order to reduce the burden. Nature's minimalist method of stellar distribution is confirmed by 464–5.

¹⁰ Goold's translation (n. 4). I have modified his translation of 392–3 to reflect more clearly the syntax; Orion is the subject, caput, an accusative of respect with immersus, and uultu a local ablative, though it is usually treated, together with subducto, as a loosely appended ablative of attendant circumstance. The combination of adversative at and the repetition of Orion's name redirects the observer's view from the large and conspicuous torso to the dimmer and less well-defined head and face, both of which are described in terms (immersus, subducto) suggesting the enormous height of the constellation. For a different view of 1.392–3 see Bailey, D.R. Shackleton, ‘The Loeb Manilius’, CPh 74 (1979), 158–69Google Scholar, at 162–3. In [Eratosth.], Cat. 32, cited by Housman (n. 3 [1937²]) on 1.393), the three stars in Orion's head are described as dim (ἀμαυρούς). It is to be noted, however, that in J. Pàmias i Massana and A. Zucker (edd.), Ératosthène de Cyrène: Catasterismes (Paris, 2013), Pàmias i Massana prints L. Robert's conjecture λαμπρούς instead of ἀμαυρούς (Eratosthenis Catasterismorum Reliquiae [Berlin, 1878], 166, in the apparatus), on the evidence of Hyg. De astr. 3.33 and the scholia to German. Arat. (A. Breysig, Germanici Caesaris Aratea Cum Scholiis [Berlin, 1867], 94.1–2, 166.15). The evidence of the Latin Aratus-tradition is not sufficient, in my view, to support the alteration of the Greek text and the astronomical record it represents; see further Kidd, D., Aratus: Phaenomena (Cambridge, 1997)Google Scholar, note on 322–5.

¹¹ Housman, (n. 3 [1937²]), on 1.468.

¹² Garrod, (n. 4), 130.

¹³ Goold (n. 3), 108–9, cited respondent (2.414) in defense of respondent at 1.467. The contexts, however, are very different. In 2.414 Manilius is discussing the conjunctions of the zodiacal signs, where the geometrical correspondence between signs is clear; the poet is not discussing the disposition of individual stars as in 1.467. respondent in 1.467 is genuine; the problem lies elsewhere in the sentence. See also Montanari Caldini (n. 3), 64 n. 37 = A&R 27 n. 37.

¹⁴ D.A. Sutton, ‘“Something about fire” in Manilius’ Astronomica 1.466 f. and 1.515 f.’, Latomus 74 (2015), 689–98, at 696–7, disagrees with the interpretation of ignis as star; he translates ‘Indeed the fires (relating to the flaming constellations in 462–3) correspond to the fiery flames’; and explains, ‘… the line is addressing a fusion of entities or a type of celestial conjugation’.

¹⁵ The noun insigne can also be used of celestial bodies in the sense of ‘recognizable object/form’: Cic. Arat. fr. 2.2, praeclara insignia caeli, here either stars or constellations (Soubiran, n. 3); Nat. D. 1.100 insignia, referring to sun, moon and stars; Lucr. 5.700 radiatum insigne diei, the sun). Cf. Aratus’ use of ἀγάλματα (‘images / figures’, Phaen. 453) for the constellations; Kidd (n. 10) rightly observes that the word emphasizes the recognizability of the constellations. In explaining how the constellations were formed and named, Varro, Ling. 7.73 remarks that in the distant past country peasants identified for the first time certain constellations (signa) in the sky, which were remarkable in appearance (insignia) beyond the rest (arbitror antiquos rusticos primum notasse quaedam in caelo signa, quae praeter alia erant insignia …); for text and commentary see De Melo, W.D.C., Varro: De Lingua Latina (Oxford, 2019), 2Google Scholar vols. Here Varro articulates a basic principle that undergirds attempts at reconstructing the formation and naming of the constellations: the quality of being easily recognizable (insignis) in the sky is constitutive of what is identified as a signum. And, no doubt, Varro was aware of the connection on the linguistic level between a signum and its defining quality as insignis. Verg. Aen. 11.89 has insignibus (‘trappings’) in the same position in the line, post bellator equus positis insignibus Aethon; cf. Ov. Met. 9.776 tuaque haec insignia uidi (prayer to Isis). One may well wonder whether Vergil is slyly giving a veiled Latin gloss, -ignibus, on the horse's Greek name Aethon, ‘fiery/bright one’. If the name Aethon is interpreted as the participle αἴθων, then ignibus αἴθων can be construed to mean ‘burning with fires’. Cf. Verg. Aen. 7.281 semine ab aetherio spirantis naribus ignem, on which see J.J. O'Hara, True names: Vergil and the Alexandrian Tradition of Etymological Wordplay (Ann Arbor, 2017²), 166 and 189; Cic. Arat. fr. 34.88 and De consulatu suo fr. 2.1 (Soubiran, n. 3); and Ov. Met. 1.254, quoted in n. 5.

¹⁶ In addition to 1.466, the line-ending ignibus ignes occurs in Manilius at 4.67 and 4.531 (ignis), parallels which might have induced the alteration of the text from insignibus ignes, and it is found several times in other hexameter poets, Lucr. 6.225 (ignem), Ov. Met. 2.313, 4.509, Fast. 6.439, and Trist. 4.3.65; and Il. Lat. 73. Ovid uses twice the line-ending to describe Jupiter's use of the thunderbolt to strike the chariot of the sun and its driver, Phaethon, when it threatened to burn up the earth (Met. 2.313 and Trist. 4.3.65). The Manilian context of cosmic conflagration may have prompted a misguided attempt to harmonize the text of 1.466 with the action of Ovid's Jupiter who fought fire with fire when the occurrence of a similar conflagration was imminent. This process of harmonization can be seen in Met. 11.523 where ignibus is followed by the variants ignes (Marcianus) and undae (cett.). Although editors generally adopt ignes, W.S. Anderson, P. Ovidii Nasonis Metamorphoses (Leipzig, 1993²) printed undae, a reading forcefully defended by R. Helm in his review of H. Magnus's P. Ovidi Nasonis Metamorphoseon libri XV, in GGA 177 (1915), 505–54, at 542–3. Housman, A.E., ‘Emendations in Ovid's Metamorphoses’, Transactions of the Cambridge Philological Society 3 (1890), 140–53Google Scholar, at 151 = J. Diggle and F.R.D. Goodyear (edd.), The Classical Papers of A.E. Housman (Cambridge, 1972), 1.162–72, at 170 conjectured imbres. See further on Met. 11.523, F. Bömer, P. Ovidius Naso: Metamorphosen. Buch X–XI (Heidelberg, 1980), 5.377.

¹⁷ This type of didactic paronomasia, in which sound-play is employed as a meaningful part of the argument, has well-known precedent in Lucretius, e.g., mater-materies (1.167–71) and ignis-lignum (1.911–12), on which see J.M. Snyder, Puns and Poetry in Lucretius’ De Rerum Natura (Amsterdam, 1980), 39–42 and 90–108; for a survey of paronomasia in Latin poetry, J.J. O'Hara (n. 15), 60–4; on wordplay and word patterning in Manilius, Takeshita, T., ‘Symmetrical wordplay in the first book of Manilius’ Astronomica’, AClass 64 (2021), 317–21Google Scholar, and R.M Colborn, ‘Manilius on the nature of the universe: a study of the natural-philosophical teaching of the Astronomica, with select commentary’ (Diss., Oxford, 2015), 107–28 (I owe the latter reference to the anonymous reader). In his description of the nameless stars, German. Arat. 371–8 employs the same word play, inter signa ignes (377), but to different effect; while the phrase inter signa ignes, taken by itself, may be seen as an instruction to the reader to recognize that the syllable ign- is found in the word signa, in the larger context of the nameless stars the poet makes the point that these particular ignes are not organized as constellations because they lack a form and a name, and are recognized by the stars of a nearby constellation ([ignes] per appositi noscuntur lumina signi, 378); thus these ignes, though themselves sources of light, are paradoxically in need of light (lumina) in order to be identified by the observer.