^{page 233 note 2} See especially Leo, 's articles, ‘Vergil und die Ciris’, Hermes xxxvii (1902), 14–55Google Scholar (Ausg. kl. Schrift, ii. 29–70Google Scholar), and ‘Nochmals die Ciris und Vergil’, Hermes xlii (1907), 35–77Google Scholar (Ausg. kl. Schrift, ii. 71–112Google Scholar); also Munari, F., ‘Studi sulla Ciris’, Atti Real. Acad. d'Italia, sci. mor. e stor., ser. vii. 4 (1944), 273–314Google Scholar

^{page 233 note 3} See most conveniently F. Munari, loc. cit. 315 ff.

^{page 233 note 4} Hardie's text in the O.C.T. Vitae Vergilianae Antiquae, sect. 17–19.

^{page 233 note 5} The minor poems are not transmitted in any of the great ancient codices of Virgil, but the Catalogue of the Murbach Monastery (for which see below) and the Graz fragment (possibly copied direct from an ancient codex: see Clausen, W., CPh lix [1964], 90 ff.Google Scholar and further bibliography there) suggest that there was a collected edition of the minora in late Antiquity. Such an edition may lie behind the Donatan list, but I suggest below that it is unlikely that Donatus himself at least had much first-hand knowledge of the poems.

^{page 233 note 6} Anyone, it seems to me, who claims authenticity for any part of the Appendix except for one or two poems of the Catalepton, is treading very dubious ground. The works of scholars like Rand, Frank, and Rostagni in this respect are pure eccentricities. Any lingering doubts over the Culex should be dispelled by Fraenkel, 's article on it (JRS xlii [1952] 1–9Google Scholar

^{page 234 note 1} On the question of the relationship of our ‘vita Donati’ to the actual work of Donatus, as well as to the original Suetonian vita, the most convenient summary is in Büchner, K.'s P. Vergilius Maro (Stuttgart, 1961), 2–4Google Scholar (═ RE viiA, 1, pp. 1022–4). Coincidence of vocabulary and method of expression etc. tie up our Donatan life closely with Suetonius, via Donatus. Not unexpectedly, however, there is not absolute agreement about the details of the affinity, and of course the question of the list of minora concerns us acutely. As well as Büchner, cf. e.g. Nettleship, H., Ancient Lives of Vergil (Oxford, 1879)Google Scholar, G. Körtge, In Suetonii de viris illustribus libros inquisitionum capita tria (Diss. phil. Halenses xiv, 1898–1901, pars 3, 189–284), Leo, F., Die griechische –römische Biographie (Leipzig, 1901), 11–13Google Scholar, Naumann, H., ‘Suetons Vergilvita’, RhM lxxxvii (1938), 334–76.Google Scholar

^{page 234 note 2} I am assuming for the moment that the list of minor works in our version of the vita Donati represents the list as it was in the original Donatan Life. For a plausible stemma vitarum see p. xxvii of Hardie's edition (based on H. R. Upson).

^{page 234 note 3} Unless the phrase septem sive octo libros hos is indicative of doubt over Aetna.

^{page 234 note 4} Texts of both these Lives are in Hardie's O.C.T. I cannot help wondering, in view of the conclusions I come to with regard to Suetonius and the list of minora, whether this might not be an indication that these particular Lives had access to a purer, pre-Donatan tradition.

^{page 234 note 5} The Epigrammata remain rather mysterious, nothing surviving under that name. But quite possibly the title originally designated part or all of a collection containing what we now know as the Catalepton and Priapea; cf. Büchner, op. cit. 44 ═ RE viiA, 1, p. 1064. In our MSS, both these are in fact transmitted under the heading Catalepton, but the Priapean poems do closely cohere as a distinct group (cf. Büchner, op. cit. 48 ═ RE viiA, 1, 1068). It is to be noted that Quint. (Inst. 8. 3. 29) implicitly, and Marius Victorinus (Gramm. Lat. Keil, 6. 137. 24) explicitly, refer to a poem of our Catalepton collection as an epigramma.

^{page 234 note 6} Text in Diehl, , Die Vitae Vergilianae (Bonn, 1911, kleine Texte 72), 26–37.Google Scholar For this life, see also Büchner, op. cit. 4 ═ RE viiA, 1, p. 1024.

^{page 235 note 1} My intention is to counter the impression that Vollmer, for example, tries to create in ‘Die kleineren Gedichte Vergils’, Sitz. der philos.-philol. u. der hist. Klasse der K.B. Akad. der Wiss. zu München (1907), 335–74Google Scholar, that these poems are well attested as Virgilian in antiquity. I shall only deal with the ones that appear in the Donatan list (for Copa and Moretum, see Vollmer, op. cit. 349 ff.). In fact Munari (loc. cit. 246–72) gives quite a good review of most of the ancient testimonia to them, and makes some important points. Indeed it was only through prejudging the issue of the linguistic parallels between the Ciris and the Silver epicists that he missed the full significance of some of these points, as regards the date of the Ciris.

I shall not here consider in detail the indications concerning acceptance of these poems that can be gleaned from apparent verbal echoes in later poets—which do not in any case build up a very clear picture. e.g. I have noticed several probable echoes of the Ciris in Claudian, and a case can be made (with varying degrees of success) for echoes of other poems in the Appendix by Claudian and Ausonius (cf. Steele, R. B., TAPhA lxi (1930), 203 ff.Google Scholar) and other poets.

^{page 235 note 2} To be more accurate, no collection with the title Epigrammata is mentioned: cf. p. 234 n. 5 above.

^{page 235 note 3} Talking of metres, Diomedes says: Priapeum quo Vergilius in prolusionibus suis usus fuit, tale est …, and he quotes a verse, the metre of which is certainly used in the third poem of our collection of Priapea, but which itself does not appear there.

^{page 235 note 4} See Goodyear, 's ed. maior of the Aetna, p. 56 note 2.Google Scholar The only other sign of recognition of the Aetna that Vollmer can muster is that line 320 is adduced s.v. suffocat by the Exempla Vaticana (Keil).

^{page 235 note 5} Pliny, while defending his own versiculos severos parum (5. 3. 2)—apparently the hendecasyllabas … quibus … iocamur ludimus amamus dolemus querimur, etc. of 4. 14. 2 f.— appeals to the example of (among others) P. Vergilius (5. 3. 6). So apparently he knows of Virgilian occasional verses, probably some of them erotic or risqué, as this seems to be a main concern in Epist. 5. 3. Bearing in mind Quintilian's evidence, we are probably right to imagine he is thinking of some of the Catalepton; this certainly seems more likely than that he could have had in mind the three Priapea, at least exclusively—but it is not impossible that the Priapea were from the start in one collection with some or all of our Catalepton.

Pliny's invocation of Virgil should be contrasted with Ovid's reaction in a similar situation; if Ovid had known any of the erotic or risqué Catalepton as Virgilian, we might expect them to figure in his list of Roman predecessors in the sphere of amorous poetry at Trist. 2. 422 ff. et Romanus habet multa iocosa liber … Their absence (as that of the Ciris) at Trist. 2. 533 ff. is even more noticeable. This ties in with the universal lack of recognition of any of the minora during Virgil's lifetime or immediately after it: cf. especially also Ov. Am. 1. 15. 25–6, Prop. 2. 34. 67 ff., and note of course Virg. Georg. 4. 565–6. (On all this see Munari, loc. cit. 252 ff.)

^{page 236 note 1} facit praeterea versum iambicum trimetrum tam ‘Simonides’ quam ‘negotio’ repetitum ter, haud alias, quam ut aiunt fecisse Vergilium nostrum iambico epigrammate, thalassio thalassio thalassio ( ═ Catal. 12. 9).

^{page 236 note 2} See Boerma, Westendorp's commentary on Catalepton i, p. xliiGoogle Scholar. For further possible indications of knowledge of this collection, see Westendorp Boerma, i, p. xliii.

^{page 236 note 3} I assume that the text of the Culex that we possess is the one accepted by 1st-century literary figures as a genuine Virgilian poem. This of course is where echoes in other authors become very important: cf. e.g. Spaeth, J. W., TAPhA lxi (1930), 24.Google Scholar Some scholars (see e.g. Plesent, C., Le Culex: Étude sur l'Alexandrinisme latin [Paris, 1910], ch. 1)Google Scholar, puzzled by the existence on the one hand of superficially good external evidence for a Virgilian Culex, but on the other of overwhelming internal evidence against our present poem's authenticity, have sought a compromise solution in the theory that a genuine Virgilian Culex has been lost (an unlikely occurrence, surely), to be replaced by the present poem. They see some support for this idea in the fact that Donatus' summary of the Culex (quoted above) does not exactly fit our present text: a main point is the omission in Donatus of the long Hades digression—a reasonable economy for a summarizer, we might feel. I shall be returning to the summary in the Donatan Life below (it is, I think, almost certainly later than Suetonius), but it does seem to me that it is very probably not based on a recent reading, or perhaps any first-hand knowledge of the poem at all. In this it is interestingly similar to the Servian references to the Aetna and the Ciris (see above).

^{page 236 note 4} … ut praefatione quadam aeiatem et initia sua cum Vergilio comparans ausus sit dicere ‘et quantum mihi restat ad Culicem ?’

^{page 236 note 5} Vollmer's other purported signs of recognition (op. cit. 349) are not cogent.

^{page 236 note 6} Vollmer's plea that the minora did not form part of the school edition just is not sufficient. If these works existed and were thought of as Virgilian juvenilia, they would surely have been a bountiful source of interest and remark in the post-Virgilian period, whether they were read in schools or not.

^{page 237 note 1} We will remember, too, how Suetonius rejected on stylistic grounds forged Horatian elegi and a letter of self-commendation purportedly from the poet to Maecenas: see the vita Horati (in Klingner's Teubner text of Horace, p. 3, lines 20 ff.). If Suetonius was capable of doing this, can we really imagine him accepting the Ciris or the Dirae as Virgilian? Even the Culex, we might think, would give him pause. Donatus' apparent blanket-approval is puzzling enough, but in the light of the rarity of contemporary testimonia and the possible indications of sketchiness in his knowledge of the Culex, I have my doubts whether he actually knew many of his list of minora at first hand. Cf. above, p. 236 n. 3.

^{page 237 note 2} In the case of the Aetna at least, a fairly convincing terminus ante quem of composition is the year A.D. 79. See Goodyear, 's ed. maior, p. 59.Google Scholar

^{page 237 note 3} Loc. cit. 258.

^{page 237 note 4} The Culex, which Suetonius probably did include, offers no distinctive biographical detail; nor do Catal. 2 and 12, the only poems of that collection for which we have specific ancient testimonia. As for Catal. 5 and 8, which have generally been considered genuine Virgilian, I think the association with Siro alluded to therein, and other biographical details which are completely missing from the Donatan Life, indicate one of two things: either the poems are post-Suetonian and thus presumably intentional forgeries (like the more obvious Catal. 14, for which see Westendorp Boerma ad loc), or, if they are pre-Suetonian, even Virgilian, Suetonius did not know them: otherwise the biographical material they contain would feature in the Life that basically emanates from him. I think too that if these poems were known as Virgilian in the first century, the absence of specific citations from them, especially the attractive Catal. 5, becomes very surprising.

Virgil (Georg. 4. 563 f.) himself mentions that he wrote the Georgics at Naples, and Donatus refers to a ‘Campanian retreat’: habuitque domum Romae … quamquam secessu Campaniae Siciliaeque plurimum uteretur (sect. 13); cf. sect. 11 … ut Neapoli Parthenias vulgo appellatus sit. The Servian Life says he studied at Naples (Hardie, line 6). Poems, thus, which are apparently Virgilian and imply a close connection between that poet, in whose earlier works a strong dash of Epicureanism is discernible, and Siro, the well-known contemporary Epicurean philosopher and teacher, who functioned at Naples (see Westendorp Boerma, op. cit. i. 99 f.; Frank, T., Vergil [Blackwell, Oxford, 1922], 48)Google Scholar, are not implausible. And certainly Catal. 5 at least would be the work of a very inventive and imaginative forger, who would be creating a motif apparently without authority. It is not impossible, however, to see Catal. 8 as a forgery based on knowledge of Catal. 5 and other elements of the biographical tradition, and perhaps with echoes of Horace's Sabine farm. Catal. 5 then, at least, appears to be either genuine Virgilian, or at least contemporary with him, but it disappeared, only to re-emerge much later.

As well as being missing from the Donatan Life, Siro is also unmentioned in the vita Servii, the vita Probiana (though cf. lines 10 f. (Hardie) liberali in otio secutus Epicuri sectam), and the excerpta of Jerome. He emerges in the commentaries (Serv. Ecl. 6. 13, Aen. 6. 264, Schol. Veron. Ecl. 6. 10), and in the vita Focae (63) and the vulgate Donatus auctus. Presumably these latter references are based on knowledge of Catal. 5 and/or 8.

^{page 238 note 1} Cf. Munari, loc. cit. 257.

^{page 238 note 2} ‘Passages in the life of Vergil’, TAPhA lvii (1926), III.Google Scholar

^{page 238 note 3} Geer in fact (op. cit. 110–12) thinks that the whole section from deinde Catalepton … to the end of the quotation from the Culex is un-Suetonian in style, and thus that the summary must be an addition. I have shown above that it is possible that Suetonius would have included the Culex as a Virgilian work, so we may give this a moment's more thought.

Stylistic arguments are less valuable in this sort of case than when considering a specific feature like the list of titles: we could easily imagine Donatus altering the phraseology of a summary—say, if he wanted to abbreviate it. But I still think it highly unlikely that Suetonius would have ever had such a thing: for (a) there is nothing comparable elsewhere in Suetonius, (b) the Culex seems to have been sufficiently well known in Suetonius’ time to make a summary otiose anyway. It seems to me to belong much more naturally to a later period when knowledge of all the minora, including the Culex and Catalepton, was very sparse. Thus if information was available for one of these rather mysterious poems, it would be natural to work it in. I do not actually think the summary is based on first-hand knowledge of the poem: cf. above, p. 236 n. 3. It ought to be mentioned that Munari, loc. cit. 257, in the face of the un-Suetonian character of the style, sees at least one un-Donatan feature.

^{page 238 note 4} Cf. Munari's list, loc. cit. 248; these are scarcely significant on their own, but since they are largely different from the ones I discuss, they provide a further useful indication of the extent of the connection.

^{page 239 note 1} From common sense alone it is hard to see how the relationship could be the other way round. The only conceivable reason why the Silver epic poets should be influenced by the Ciris is that they regarded it as a genuine Virgilian work—which from the evidence set out above seems scarcely possible. (Contrast e.g. Steele, R. B. (AJPh li (1930), 173Google Scholar) who argues thus: because Statius imitated the Ciris—premiss unquestioned— he must have considered it a genuine Virgilian work).

^{page 239 note 2} Fraenkel, Cf, JRS xlii (1952), 8.Google Scholar

^{page 239 note 3} André, Cf, Les Termes de couleur dans la langue latine (Paris, 1949), 111 f.Google Scholar

^{page 239 note 4} See Cahen's note on Call. Hymn 4. 302 f. In Latin literature the image is not common: cf. Avienus, Arat. 1628, Octavia 3, and a few more times; but note too the use of crinis (as κόµη) to denote the flaming trail of comets or shooting stars.

^{page 240 note 1} Hypallage involving roseus in contexts of dawn is common: cf. Lucr. 5. 976 dum rosea face sol inferret lumina caelo, Tib. 1. 3. 93 f., Virg. Aen. 6. 535, etc.

^{page 240 note 2} Commentators sometimes quote Catull. 64. 309 at roseae niveo residebant vertice vittae as a possible influence behind the choice of adjective. I doubt there being any direct or conscious connection, but Professor Skutsch does point out to me that since the author of the Ciris was fond of the ‘golden line’ which here could not be attained simply with purpureus (because of the necessity for et), the Catullan ‘golden’ line may have suggested to him a way round the difficulty.

^{page 240 note 3} Clearly lines 496–7 refer to specific features of the pre-metamorphosis Scylla, which then concrescere in unum coepere; thus the inclusion of the present example of frons in the TLL under the heading bestiarum is misleading and erroneous (vi. 1359. 53).

^{page 240 note 4} Professor Skutsch points out to me that the author, in order to emphasize the change that Scylla's face underwent in the metamorphosis, may be meaning to stress the relative broadness of her original form compared with her beaked bird's face. Thus patulae may be intended to contrast with concrescere in unum. I think this helps to explain the choice of phrase, though I still believe one has to be charitable to forgive the bald patulae frontis species. There is at least, I think, enough awkwardness in it to suggest the priority of the Culex.

^{page 241 note 1} For convenience's sake I have chosen the following authors to provide a reasonable survev: Enn., Lucr., Catull., Virsil and Vergiliana, Hor., Corp. Tib., Prop., Ovid, Sen. Trag., Luc.,Sil., Val. Flacc, Stat., Mart., and Juv.

^{page 241 note 2} Cf. Ovid, Met. 4. 267 luridus pallor, of a deathly pallor resulting from love-induced starvation.

^{page 241 note 3} Cf. also Virg. Ecl. 2. 49 f. tum casia atque aliis intexens suavibus herbis / mollia luteola pingit vaccinia caltha, and other comparable uses of pingo.

^{page 241 note 4} The appearance of aureolus in Culex 144, in the light of the likelihood that the Ciris is indebted to this line for viridis pallor, makes Housman's conjecture of aureolam at Cir. 151 even more likely: the author may well have been stimulated to its use hence as well as by the occurrence of the word in Catullus.

^{page 242 note 1} Cf. also Cir. 198 vosque, adeo, humanos mutatae corporis artus and Virg. Ecl. 6. 78 aut ut mutatos Terei narraverit artus. The identical metrical pattern of the conclusions of Culex 128 and Ecl. 6. 78 suggest to me that the author of the Culex had the latter line in mind when he wrote his line 128, whilst the closer similarity of grammatical form points to a direct connection between Culex 128 and Cir. 482.

^{page 242 note 2} We might, however, compare how πєντєτηρίς means both a five-year period and a festival celebrated every fifth year (see LSJ s.v.).

^{page 243 note 1} Steph., however, reads coronarum for coronatorum.

^{page 243 note 2} At Mart. 4. 1. 7 ingenti … lustro is directly designating, I think, the ludi saeculares, and thus the use of lustrum parallels the one we have noticed in Statius; and it should be noticed that neither here, nor at Stat. Silv. 3. 1. 45, is the festival in question even five-yearly.

^{page 243 note 3} See above, p. 241 n. 1.

^{page 243 note 4} But cf. also Stat. Silv. 4. 4. 45; TLL (annus) adduces only Cir. 45 and Ovid, Met. 7. 295. Cf. too anni iuvenales, Ovid, Met. 8. 632, Stat. Theb. 1. 486, etc.

^{page 243 note 5} Teneri anni is quite a common phrase and thus (it might be feit) not particularly desirable.

^{page 244 note 1} Cf., however, Ilias Lat. 882 lyrae graciles extenso pollice chordas / percurrit; and also Seneca Ag. 338 licet et chorda graviore sones and Stat. Silv. 1. 4. 36 nec … sperne coli tenuiore lyra, in both of which the epithet is explained by the context and a hypallage similar to that at Silv. 4. 4. 53 is operating.

^{page 244 note 2} See Sudhaus, S., Hermes xlii (1907), 482.Google Scholar

^{page 244 note 3} This is not absolutely certain. Ovid for example in his Io (Met. 1. 583 ff.) tells the wanderings of the heroine very elliptically and has her goaded by a terrifying apparition, thus line 725 horriferam … Erinyn, 726 stimulos … caecos. But it is Ovid's habit to introduce changes, and he would naturally vary the Calvan account.

^{page 244 note 4} See Sudhaus, loc. cit.

^{page 244 note 5} Cf. also Juv. 4. 123, Nemesian. Cyn. 3.

^{page 244 note 6} And even the apparently original Graecisms of diction in the Ciris could easily be taken from neoteric sources now lost.

^{page 245 note 1} Prolegomena in Cirin (Diss. Utrecht, 1882), 48.

^{page 245 note 2} Cf. Virgil's use of substantival abruptum at Aen. 3. 422, 12. 687.

^{page 245 note 3} For nimium … avidum cf. Lucr. 4. 594 humanum genus est avidum nimis auricularum.

^{page 246 note 1} In spite of the gloss CGL iii. 17. 36 κύκνος anser olor’, anser and olor always seem to refer to different species, and to be distinguished—cf. e.g. Virg. Ecl. 9. 36 (videor) argutos inter strepere anser olores; anser in fact means goose, not swan. Professor Goodyear points out to me how similar this reference to the divine swan as a goose is to the kind of comic depreciation of poetic–mythological figures which we find in satire, especially Juvenal: cf. 1. 54 fabrumque volantem (Daedalus), 10. 257 Ithacum … natantem (Odysseus), Pers. Prologus 1 fonte … caballino, etc.

^{page 246 note 2} Heinsius in fact proposed to read Amyclaeae at Cir. 489. But it seems to me that the fact that Leda happened to give birth to her sons at this town is a very good reason for calling them ‘Amyclaean’ (cf. Virg. Georg. 3. 89 Amyclaei.… Pollucis, Stat. Theb. 7. 413) or for calling Amyclae ‘Ledaean’ (cf. Sil. 2. 434), but not really sufficient to warrant Leda herself being called Amyclaean. (At Stat. Silv. 2. 1. III Amyclaea mater is probably meant to mean any Spartan mother—cf. Vollmer ad. loc.—though the use of the word mater does make the Leda-identification possible.) Thus at Cir. 489 the poet did indeed write Amyclaeo, and I believe very probably under the influence of Stat. Silv. 1. 2. 142.

^{page 247 note 1} The meaning of classis here is discussed below.

^{page 247 note 2} See Torr, C., Ancient Ships (Cambridge, 1895), 103.Google Scholar

^{page 247 note 3} Classis in the singular does occasionally seem to be used where we would more naturally expect navis (cf. Hor. Carm. 3. 11. 48), and more frequently classes seems to occur where ‘ships’ makes the best sense: see TLL 3. 1283. 60 ff., though by no means all the examples cited there ought to be. In our present line classis not only seems to be used for naves, but the plural is most conveniently taken as ‘poetic’. It could be, though, that cumba is due to the poet's thinking of a specific, classis a general, occasion

^{page 247 note 4} It appears additionally awkward when we remember that Minos’ fleet was called a classis in the main narrative at 459, and when we reflect that the classis in the simile turns out to be representing Minos’ particular ship!—see below.

^{page 247 note 5} The reference to the actual suspension of Scylla (389) is so elliptical that it is in fact only from the present simile that we infer that the poet is adopting the version of the story in which the girl was hung from the stern, rather than from the prow (for which see Tzetz. ad Lycophr. Alex. 650).

^{page 247 note 6} The use of classis to denote ‘ship’ in 479 (see above) was indeed rather obscure; but we should note how Minos’ own vessel, the item in the main narrative to which it apparently corresponds, is also in the ‘poetic’ plural: see line 389 de navibus altis.

^{page 248 note 1} Cf. the unextended simile at Catull. 25. 12 f.

^{page 248 note 2} To mention one detail, Statius’ use of the small/big contrast (immensae … minor) in the simile plays an integral role in his scheme, whilst the emphatic juxtaposition of parvula and magnas in the Ciris’ simile has little point as regards the main narrative.

^{page 248 note 3} I am a little surprised that the poet abandoned conexa or a close equivalent in favour of the vague sequitur—one would have thought he might have enjoyed drawing the exact parallel between Scylla fastened to the stern of a ship, and the dinghy likewise fastened. But this does of course highlight the grotesqueness of the comparison being made or at least implied, and we must assume that the poet appreciated this, and chose to take refuge in a vaguer substitute.

^{page 249 note 1} firmo with an acc. and abl. in this kind of sense is common.

^{page 249 note 2} Cf. note 4 below.

^{page 249 note 3} Cf. also Cic. Fam. 6. i. 3 quantum in cuiusque animo roboris est atque nervorum, Hor. Ars 26 f. (in a context similar to that of the Ciris) sectantem levia nervi / deficiunt animique, etc.

^{page 249 note 4} Cf. Drac. Laud. Dei 2. 739 tenues adiecto robore firmas. For firmo alone in such contexts, cf. Ovid, Pont. 1. 3. 27 bene firmarunt animum praecepta iacentetn.

^{page 249 note 5} ‘Nochmals …’, Hermes xlii (1907), 44 f. ═ Ausg. Kl Schrift, ii. 79–81.

^{page 250 note 1} For Val. Flacc. 2. 57, see Langen, ad loc, and cf. Virg. Georg. 1. 430 f.

^{page 250 note 2} senectae occurs at the end of a line twice in the Ciris—at 287 and 314.

^{page 250 note 3} e.g. Rem. 247.

^{page 250 note 4} If the Ciris is imitating Silius here, we might compare how nullus in ore rubor was perhaps borrowed from a completely different context: see above.

^{page 251 note 1} The poet could have thought some or all of the Catalepton collection genuine: testimonia exist for certain of the poems as early as the first century B.c.: see above, pp. 235 f. There are in fact other parallels between the Ciris and poems of the Catalepton indicative of possible direct connection.

^{page 251 note 2} See above, p. 241 n. 1.

^{page 251 note 3} Lachmann implausibly wished to restore it also to Prop. 2. 1.5.

^{page 251 note 4} Cf. Martial's use of sophos and sophōs, and what Friedländer has to say about it (at 7. 32. 4 and 1. 3. 7).

^{page 252 note 1} See p. 241 n. 1.

^{page 252 note 2} Editors with some plausibility have wanted to emend it to superum sedes.

^{page 252 note 3} Not a very happy use of the phrase; the author of the play, who it is generally agreed was not Seneca, has perhaps lifted it from the Herc. F.

^{page 252 note 4} Pristes, which is obviously correct, was suggested by Barth in 1608 for pestes H (also G) and pisces AR, and is in fact presented, presumably by conjecture, by Corsinianus 43 F. III. 21.

^{page 252 note 5} Suet. Tib. 70.

^{page 252 note 6} His hexameter Orpheus for instance (of which a few lines survive) may well have been an epyllion. It seems to have been confined to one book, and the style of the fragments is very Alexandrian.

^{page 252 note 7} See especially lines 93 ff.

^{page 253 note 1} Cf., e.g., how at 4. 14 he compares himself and his work to Catullus, whilst associating Silius with Virgil. And, interestingly, he inveighs against some of the turgid contemporary epic writing: see Spaeth, J. W., ‘Martial and Vergil’, TAPhA lxi (1930), 21.Google Scholar His views are, however, not easily compartmentalized; there is some interesting information on this in Preston, K., ‘Martial and Former Literary Criticism’, CPh xv (1920), 340–52.Google Scholar

^{page 253 note 2} Cf. Epist. i. 16, 4. 27.

^{page 253 note 3} See Schanz-Hosius iii. 21 ff.

^{page 253 note 4} See Schanz-Hosius iii. 45. A cento in fact is only a linguistic Spielerei on the level of the metrical games of the poetae novelli.

^{page 253 note 5} People have speculated how this might happen: it is not an unparalleled phenomenon for anonymous works to gather around a great name, and we have, too, to reckon with very simple possibilities—e.g. that a scribe might fill up spare Ieaves at the end of a codex with an odd poem or two. The Virgilian lines in a work like the Ciris might suggest the association with the great poet. We might compare the Virgilian corpus to the collection of poems in the Tibullan manuscripts, though there is much more reason for their association in a single edition. One thing is certain: odd attributions in our sources should cause us no surprise. The inclusion of the Eleg. Maec. as a Virgilian work is far more immediately outrageous than that of the Ciris; and we can remember too, e.g., that the Laus Pis. was assigned, in a manuscript now lost, to Virgil.

^{page 253 note 6} According to the list in RE, the Valerii Messallae are fast dying out in the second century A.D.