¹ Jerome, , Chron. ad ann. 41 (Helm p.159).Google Scholar

² Caesar, , Bell. Alex. 42–7;Google Scholar Dio 48.17, 21 ff.;Appian, , B.C. 3.85, 4.36, 53 ffGoogle Scholar; Livy, , Ep. 123.Google ScholarGanter, F. L., ‘Q. Cornuficius’ Philologus 53 (1894), 132Google Scholar; Stemkopf, W., ‘Die Verteilung der römischen Provinzen vor dem Mutinensischen Krieg’, Hermes 47 (1912), 321Google Scholar; Münzer, F., RE iv.1624 (8)Google Scholar; Tyrrell, and Purser, , The Correspondence of Cicero, iv (Dublin, 1918), ciGoogle Scholar; Syme, R., ‘Observations on the Province of Cilicia’, Anatolian Studies … W. H. Buckler (Manchester, 1939), p.299.Google Scholar Also Drumann-Groebe, ii.531Google Scholar; MRR ii.306.Google Scholar

³ Dio 42.51; Cicero, ad f. 12.17 (46 B.C.): ‘Cicero s.d. Cornificio conlegae’. Sumner, G. V., ‘The Lex Annalis under Caesar’, Phoenix 25 (1971), 357CrossRefGoogle Scholar, for his early praetorship.

⁴ Ganter, , op. cit. n.2Google Scholar; Romanelli, P., Storia delle provincie romane dell' Africa (Rome, 1959), p.142Google Scholar, perhaps ‘in base a disposizioni gia prese da Cesare’. Calvisius was in Rome on the Ides according to Nicolaus of Damascus, but Romanelli thinks he may have returned briefly to Africa thereafter, since Cicero, , Phil. 3.26,Google Scholar says that at the end of November ‘modo ex Africa decesserat’.

⁵ Catullus 38; allocutio in the precise sense of ‘consolation’ is common. It is perhaps Simonides' funerary epigrams, rather than his lyric Threnoi, that were to be imitated. The usual subject for a Consolation in antiquity is the death of a third person, so Catullus' fatal illness or Lesbia's defection, which have been suggested as the occasion of the poem, are not very likely.

⁶ But perhaps the ‘deducta vox’ is the poet's own; whether it means ‘subdued’ or ‘thin’ is hard to say. Mr. R. G. G. Coleman tentatively suggests to me a reference to tenuitas of style. Macrobius, , Sat. 6.4.12Google Scholar compares Cornificius' phrase with passages from comedy (Afranius and Pomponius) and there is also a parallel in Lucilius.

⁷ The fragment of the Glaucus (Macrobius, , Sat. 6.5.13Google Scholar), ‘Centauros foedare bimembres’, conceivably suggests an account in indirect speech of the battle of the Centaurs and Lapiths (the Centaurs are likelier to be subjects than objects of this particular verb, but some read sedare). Centaurs do not seem to play a part in any story involving a Glaucus, so we have here a passing reference, a simile, or even an inset into the epyllion. Bimembris as an adjective for Centaurs had a successful career in Latin poetry (like biformis), but it is a fairly obvious translation of the Greek often used of them, so one cannot be certain that Virgil was consciously following Cornificius.

Of ‘ut folia quae frugibus arboreis tegmina gignuntur’, quoted by Servius (Georg. 1.55) I can make little; Morel reorganizes it into hexameter form. Dactylic elements are clearly visible.

⁸ Ovid, , Trist. 2.436.Google Scholar

⁹ Wissowa, G., ‘Cornificius’ (8) RE iv.1624.Google Scholar

¹⁰ Wissowa, , loc. cit., suggests that the Berne scholiast's idea that Cornificius was an obtrectator of Virgil is a stupid deductioi from Thyrsis' hostile words ‘rumpantur utGoogle Scholar ilia Codri’; but the scholiast himself thinks Cornificius is Thyrsis. In Ecl. 5.10–11 ‘iurgia Codri’, like ‘Phyllidis ignes’ and ‘Alconis laudes’, are subjects for song, but the use of names is not always consistent between the different Eclogues and Codrus may not here be thought of as a poet.

¹¹ Though Frank, Tenney, ‘Cornificius as Daphnis?’, CR 34 (1920), 49Google Scholar, supposes that it would be rash to praise him openly after he fell resisting the triumviral cause. But Cornificius cannot be the Daphnis of Eclogue 5: if no Nestor, he was equally not a puer (though African lions and at a pinch Armenian tigers could be seen as relevant to his career) and, as will be seen, he can hardly be the leader of the neoterics after Cinna's death, as Frank holds.

¹² Presumably Valgius picked the (not very obvious) name of an admired poet from Virgil and applied it to a real man (Valerius Cato?). In thinking that the age of Nestor and Demodocus is significant one need not reject the allusions seen in the lines by Wiseman, T. P., Cinna the Poet (Leicester, 1974), p.57, esp.Google Scholar that the doctus poet Demodocus' tale of Ares and Aphrodite in Odyssey 8 is a model epyllion. Sudhaus, S., ‘Die Ciris und das römische Epyllion’, Hermes 42 (1907), 502Google Scholar, thinks the Ciris could be by Valgius' ‘Codrus’.

¹³ Schmidt, E. A., Zur Chronologie der Eklogen Vergils (Heidelberg, 1974)Google Scholar, for disputes on the date: Büchner, R., RE 2 viii.1. 1224–6Google Scholar puts it around 35; as does Schmidt, accepting Bowersock's dating of no. 8 to that year. R. G. G. Coleman argues in his new edition for 39. Valgius, Horace, , Sat. 1.10.82.Google Scholar

¹⁴ The Berne scholiast suggests that he is the inferior poet Amyntas of Ecl. 2.35 and Servius Auctus the Antigenes of 5.89.

¹⁵ Cicero, , ad f. 12.20Google Scholar attests a visit by Cornificius to Campania. Philargyrius, in the fifth century, ad Ecl. 3.105, says that ‘Cornificius’ recorded a detail that he had heard from Virgil himself. This suggests a man still alive after Virgil's death, but the reading is uncertain, and ‘Cornutus’ perhaps a better one.

¹⁶ ILLRP 439. In other inscriptions of the period the only Camerii are a freedman at Caere and a slave at Praeneste (ibid. 830, 106 d). In CIL they cluster in Central Italy (cf. Cameria, Camerinum) and also in Cisalpine Gaul, though not actually Catullus' Verona (CIL xi.6516–19Google Scholar, Sassina; 7570, Tarquinii; 168, Ravenna; CIL v.2325–6Google Scholar Atria; 2409, Ferrara; 2855, Patavium; 3129, Vicetia). Schulze, W., Lateinische Eigennamen (repr. Berlin, 1933)Google Scholar, perhaps wrongly thinks the name Etruscan. See now Wiseman, T. P., ‘Camerius’, BICS 23 (1976), 15Google Scholar, strongly for origin from ‘domi nobiles’ of Venetia, adding an early inscription from Atria published in Epigraphica 18 (1956), 52Google Scholar, and noting that the Camerii of Vicetia and Patavium are of magisterial status and probably fairly early.

Slater, W. J., ‘Pueri, Turba Minuta’, BICS 21 (1974), 133Google Scholar, misguidedly revives Birt's suggestion that Camerius was Lesbia's delictum or child pet (Camerius is impossible as a slave's name). It is surely unlikely that Catullus would write a poem to such, addressing him as amice. The treatment of Camerius in Neudling, C. L., A Prosopograpby to Catullus (Oxford, 1955), p.46Google Scholar, does not deserve the derision o Slater, who has read it carelessly (that of Cornificius is also sound though incomplete Wiseman rightly protests against Slater's fantastic interpretation.

¹⁷ Cicero, , ad Att. 13.28Google Scholar; mulieres suggests that her mother backed her—her father, as we shall see, was perhaps recently dead.

¹⁸ See above, n.l. Wiseman, , op. cit. n.16Google Scholar, suggests that the inscription, which was on a convex plaque, came from a circular tomb, and that Cornificia, named first, died before her brother, But would he not have been described at his death as proconsul rather than praetor? And were his ashes ever brought back to Rome?

¹⁹ Cicero, , ad f. 8.7.2Google Scholar; Sallust, , Cat. 15.2.Google Scholar

²⁰ Sail. Cat. 35.3.

²¹ Cn. Aufidius Orestes cos. 71 (Cicero, , dom. 35Google Scholar), and perhaps Mucius Orestinus trib. 63 and favourable to Catiline (see Miinzer, F., RE xvi.423Google Scholar). Sumner, G. V., The Orators in Cicero's Brutus (Toronto, 1973), p.46Google Scholar; Badian, E., ‘Notes on Roman Senators of the Republic’, Historia 12 (1963), 132 n.6Google Scholar; Gruen, E., The Last Generation of the Roman Republic (Berkeley, 1974), p.183 n.74, 218 n.36.Google Scholar Aurelia Orestilla is perhaps as Münzer suggests (RE xiii.1394Google Scholar) the ill-famed Aurelia, of ad f. 9.22.4Google Scholar, of 45 B.C. (where Lollia, joined with her, will be Gabinius' wife and Caesar's mistress). Sal lust's ‘nihil umquam bonus …’ perhaps also suggests that her indiscretions were not confined to early youth?

²² Sallust, , Cat. 35.6.Google Scholar

²³ Ad Att. 12.14.2, cf. 17, ‘etsi reus locuples est’. (Flavius is presumably his procurator.) Neudling, , loc. cit.Google Scholar, suggests that the elder Cornificius has just died and that his son is settling his affairs.

²⁴ Plutarch, , Caesar 51Google Scholar, (cf. 43, where this seems to mean Cornificius).

²⁵ Funaioli, H., GRFGoogle Scholar: Cornificius Longus frag. 4. See Bona, F., Contributo alio studio della composizione del ‘de verborunt significatu’ di Verrio Flacco (Milan, 1964), pp.35 ff.Google ScholarWillers, , De Verrio Flacco (1898)Google Scholar, collected all the glosses in Festus that might on grounds of content come from the De etymis deorum.

²⁶ Funaioli gives them to another work than the de etytmis deorum; this might make against the poet's authorship, as he is perhaps unlikely to have written more than one grammatical work. (The false cognomen, Gallus, produced by the late grammarian Cledonius is explicable as a confusion with Cornelius Gallus.)

²⁷ Caesar, , Bell. Alex. 42.2, 43.4Google Scholar; cf. Cicero, , ad f. 12.19.1Google Scholar ‘et industria et prudentia tua’. On the other hand, Sextius was to find him slow in moving, Dio 48.21.

²⁸ Ad f. 12.17.2, 12.18.1. Recent writings on Atticism have neglected Cornificius, though he figures in older discussions. But he hardly deserves to rank among ‘Les Critiques et les correcteurs des oeuvres de Cicéron’ (Wikarjak, J., Eos 59 (1971), 281).Google Scholar

²⁹ Gruen, E., ‘Cicero and Licinius Calvus’, HSCP 71 (1966), 215.Google Scholar

³⁰ Douglas, A. E., Cicero (Oxford, 1968:Google ScholarGreece & Rome; New Surveys of the Classics, no. 2), p.39.

³¹ Cicero, , Orator 81 ff.Google Scholar

³² Ad f. 12.18.2, 12.25.5. I am not suggesting that Caesar was exactly an ‘Atticizer’.

³³ Cf. 12.23.1, ‘neque enim, quae tu propter magnitudinem et animi et ingeni moderate fers, a te ea non ulciscenda sunt, etiam si non sunt dolenda’. Note the rarity of strictly philosophical jokes and references in the letters to Atticus, a merely nominal Epicurean, compared with those to Varro, Cassius, Torquatus, and others.

³⁴ GRF frag. 6.

³⁵ Ibid.

³⁶ Suetonius, , De gramm. 3.4.Google Scholar

³⁷ Ad f. 12.25.5 ‘conscende nobiscum et quidem ad puppim’.

³⁸ Two legates left in Africa, Cicero, , Phil. 3.26Google Scholar; cf. ad f. 12.30.7 ‘de Venuleio, Latino, Horatio’ who claimed the right to lictors-perhaps to be reduced to two by linking the name Latinus with one of the others, MRR ii.355.Google Scholar Cn. Minucius, who also made trouble, ad f. 12.25a.7, is perhaps an official of some sort.

³⁹ Ad f. 12.25.2.

⁴⁰ Ibid. 12.25.1. Prorogation, Ganter, , op. cit., p.143Google Scholar, MRR ii.345.Google Scholar

⁴¹ Solinus 27.11 (cf. Appian, , Pun. 136,Google Scholar confused but tending to show it was actually founded after Caesar's death). For the various controversies about its early development, see Romanelli, P., op. cit. n.140Google Scholar and refs. The refractory notice in Tertullian, , De Pallio 1.2.Google Scholar is best understood as correctly dating Statilius Taurus' activity in its foundation to his visit in 36, later than Lepidus' presence (for which cf. Dio 52.43.1); but Taurus could have been involved, or hoped to be involved, earlier as well. See also Syme, R., The Roman Revolution (Oxford, 1939), p.237.Google Scholar

⁴² Bell. Aft. 97.2.

⁴³ Curubis, , ILLRP 580.Google ScholarRomanelli, , op. cit., p.140Google Scholar, lists probable and less probable Caesarian foundations.

⁴⁴ Ad f. 12.26, 27, 29; cf. 30.1.

⁴⁵ Pbil. 13.30.

⁴⁶ Bell. Civ. 2.28.

⁴⁷ Editors of the Philippics identify him and Gsell, S., Histoire ancienne de l'Afrique du nord viii (Paris, 1928), 187 n.lGoogle Scholar has him, but the passage is made to refer to Attius Varus and an earlier date by Klebs, , RE ii.2257Google Scholar and MRR, ii.311Google Scholar (quite impossible in the context of Antony's letter).

⁴⁸ Veil. Pat. 2.71.2. Syme, R., CP 50 (1955), 127Google Scholar, cf. MRR Suppl., p.52.Google ScholarGruen, , The Last Generation, p. 194Google Scholar, still thinks the quaestor of 49 can be traced no further The L. Varus holding Rhodes for Cassius (Appian, , BC 4.74Google Scholar) must be another man, though pace Syme the praenomen is found among the patrician Quinctilii. Alfenus Varus was probably praetor c.43, but his legal and poetical interests perhaps do not suggest a man twice captured in war; his political outlook is obscure (Syme, , RR, p.235Google Scholar). Theoretically possible as the object of Antony's complaint is also the Varus proscribed in 43, Appian, B.C. 4.28Google Scholar, described as an ex-consul, which must either be an error for ex-praetor or (so Syme, , ‘Missing Persons’, Historia 5 (1956), 208Google Scholar) refer to Caesar's grant of consular insignia to some praetorii.

⁴⁹ Appian, , B.C. 3.85Google Scholar; but ad f. 10.24.4 and 8 shows the legions must have been summoned to Italy before Antony's junction with Lepidus: possibly at the time of Cornificius' prorogation in March? Whether Cornificius already had legionary troops is unclear.

⁵⁰ Appian, , B.C. 5.26Google Scholar, represents Sextius as even after the defeat of Cornificius dependent on veterans (from Caesar's colonies?), natives, and mercenaries serving local kings. For his vicissitudes then, Dio 48.21.2, Appian, , B.C. 5.26Google Scholar; an accensus of his, a freedman of the Caelii (known to have interests in Africa) ILS 1.1945.

⁵¹ Crawford, M. H., Roman Republican Coinage, i (Cambridge, 1974), no. 509, 1–5Google Scholar; illegally he says, ibid. ii.604; but the Senate had authorized him to requisition, ad f. 12.28.2—by the March s.c. rather than the normal one for all governors? Cicero urged him alternatively to raise money by a loan (ibid. 28.3) but this may have been hard to do after the fines levied by Caesar on pro-Pompeian conventus, especially Utica. Difficulty in getting hold of bullion, or his death, may explain the tiny issue-and conceivably the plated denarii that have been found, RRC i.565Google Scholar: Crawford is prepared to concede that in such circumstances these might be official, rather than private forgeries, p.560.

⁵² Accepted recently by Syme, R., Historia 4 (1955), 61Google Scholar, Taylor, L. R., Vofiwg Districts of the Roman Republic (Rome, 1960), p.208Google Scholar, Wiseman, T. P., New Men in the Roman Senate (Oxford, 1971), p.227Google Scholar, Gruen, E., op. cit., p.137.Google ScholarDrumann-Groebe, briefly denied it (ii.530).Google Scholar

⁵³ Festus 56L: ‘Corniscarum divarum locus (lucus?) erat trans Tiberim cornicibus dicatus quod <in> Iunonis tutela esse putabantur’ (from the De etymis deorum?). Wissowa, G., Religion und Kultus der Romer (Munich, 1912), p.189 n.lGoogle Scholar; CIL i ².2.975Google Scholar, with Kajanto, I., ‘Contributions to Latin Morphology’, Arctos 5 (1967), 67Google Scholar—really only one ‘diva Cornisca’, and no relation to crows.

⁵⁴ Festus 462L (as argues, Bona, op. cit., p.39Google Scholar, from the De etymis deorum): ‘sispitem Iunonem, quam vulgo Sospitem appellant, antiqui usurpabant, cum ea vox ex Graeco videatur sumpta, quod est ’ Possible real origin of the name, Dumézil, G.‘Iuno S.M.R.’, Eranos 52 (1954), 105.Google Scholar

⁵⁵ Fears, J. Rufus, ‘The Coinage of Q. Cornificius and Augural Symbolism on late Republican Denarii’, Historia 24 (1975) 592Google Scholar. Carthaginian Juno's ‘currus et arma’, see my ‘Scipio, Furius, Laelius and the Ancestral Religion’, JRS 63 (1973), 161Google Scholar. For the iconography of Juno Sospita Hafner, G., ‘Der Kultbildkopf einer Göttin im Vatikan’. JDAI 81 (1966), 186.Google Scholar There was a cult of Juno Sospita in Rome itself, Livy 32.30.10, 34.53.4, Ovid, , Fasti 2.55.Google Scholar For Juno/Tanit possibly on the obverse of Cornificius' coins see below.

⁵⁶ RRC no. 489, 1–4. For crows and augury, Cicero, De n.d. 3.14, Isidore, , Etym. 12.7.44Google Scholar, Festus 214L. Antony shows lituus and jug as well, however.

⁵⁷ RRC no. 379, 1–2, must refer to Lanuvium as it shows the snake connected with the cult there, but it is not proved that L. Procilius came thence; so nos. 480 2a, 23, and 28, with snake and girl (cf. Propertius 4.7.3 ff.), but their author M. Mettius is of unknown origin (Wiseman, , op. cit., no. 253Google Scholar, on weakish grounds suggests Gallic). But see RRC, nos. 384 and 472, two moneyers called L. Papius showing the goddess' head, with Asconius 53C for a Papius from Lanuvium; no. 316, L. Thorius Balbus, cf. Cicero, De fin. 2.63Google Scholar (Lanuvinus) and also Grant, M., From Imperium to Auctoritas (Cambridge, 1946), p.384Google Scholar, on Thorius Flaccus.

⁵⁸ The actor came from Lanuvium, Cicero, , De div. 1.79Google Scholar; two republican Roscii in its tribe, Maecia, ILLRP 1262; Fabatus, L. Roscii, RRC, no. 412Google Scholar, shows Juno Sospit with snake. Syme, R., ‘Missing Senators’, Historia 4 (1955), 61.Google Scholar

⁵⁹ Ad f. 12.25.3. There appears to be no direct evidence from Rhegium: a man who held a magistracy there came from the Pomptina, but seems to originate from Grumentum, which certainly had this tribe (CIL x.228Google Scholar).

⁶⁰ Fears, , op. cit. in n.55Google Scholar, drawing on Alföldi, A., ‘The Main Aspects of Political Propaganda on the Coinage of the Roman Republic’, Essays in Roman Coinage Presented to Harold Mattingly (Oxford, 1956), p.84Google Scholar: ‘the iconographical prototype of Cornificius’ coin is a royal investiture, the first example for this act in the whole Roman source material as far as I can see’. Contra, showing the lack of real evidence for the lituus meaning anything but possession of the augurate, Badian, E., ‘sulla's Augurate’, Arethusa 1 (1968), 41 n.2.Google Scholar However, coins seem to stress the augurate more than the pontificate, and it may possibly be relevant that, as Cicero, says in De div. 2.76–7Google Scholar, promagistrates as opposed to magistrates could not take the auspices. Some, e.g. Weinstock, S., RE xvii.2.1727Google Scholar s.v. obnuntiatio, think that augurs always had the right to auguria impetrativa as well as oblativa—though see Cicero, , Phil. 2.81.Google Scholar

⁶¹ Dio 48.21. Why this stress on cattle? Caesar's legions had a bull on their standards, Ritterling RE xii.1549 ff.Google Scholar, Weinstock, S., Divus Julius (Oxford, 1971), p.119.Google Scholar Or, if Sextius was attracting natives rather than veterans, it seems (from the coins of Numidia and Mauretania, Mazard, J., Corpus Nummorum Munidiae Mauretan-iaeque (Paris, 1955), nos. 224–6, 277–9, 392–3, 503Google Scholar) that Egyptian veneration for cows and bulls may have been widespread in Africa, N. (perhaps a native Mauretanian cult, Mazard nos. 107–12Google Scholar; cf. Müller, L., Numismatique ancienne de l'Afrique du nord (Copenhagen, 1860–1874), nos. 356–8Google Scholar, Libyan). Bull on African coins of Octavian, Grant, , op. cit., p.60.Google Scholar

⁶² Jenkins, G. K. and Lewis, R. B., Carthaginian Gold and Electrum Coins (London, 1963), p.11Google Scholar, regard as possible ‘a secondary assimilation to Demeter in a type primarily intended to convey the chief Carthaginian deity Tanit’. In both Cornificius' coins and those of Carthage the goddess looks left and wears a pendant ear-ring, while the wreath is rather similar in treatment.

⁶³ RRC, nos. 321,351, 378:82, 97 (9, 16, 23), 99 (2a), 414,427, 467, 494 (44a). Cf. Müller, , op. cit. ii, no. 375Google Scholar, from Hippo Diarrhytus.

⁶⁴ Alföldi, A., ‘Commandants de la flotte romaine à Cyréne’, Mélanges Carcopino (Paris, 1966), p.25Google Scholar, followed by Nicolet, C., L'Ordre équestre à l'époque républicaine ii (Paris, 1974), no. 106.Google Scholar Alfondi's idea that they were struck in the same mint as those of Q. Oppius and C. Clovius is rejected by Crawford, , RRC i.94.Google Scholar The find-spots of Cornificius' very rare coins do not, so far as I can discover them, throw any light on where they were struck.

⁶⁵ Robinson, E. S. G., Catalogue of the Greek Coins of Cyrenaica (British Museum, 1927), ccxxxiiiGoogle Scholar. The heads on these two issues also look r. not 1., unlike those of Cornificius.

⁶⁶ Mazard, , op. cit., nos. 90, 92Google Scholar (cf. 123 Octavian, and 335–6, Juba II) and p.51. Müller, L., op. cit. i.103aGoogle Scholar, attributes a poor copy of Cyrenaic types to Libyan natives of the Syrtes area. Lucan's oracle of Ammon (9.511 ff.) appears to be on Cato's line of march from Berenice to Leptis, and this need not be wild poetic licence: the geographers refer to a grove or springs of Ammon somewhere along this coast, perhaps in the area ruled by Juba. For this and other evidence for the worship of Ammon here and further west see Gsell, , op. cit. iv.286 n.3, 287 n.l, vi.143–4.Google Scholar He was possibly identified with the Punic Baal Hammon. The famous shrine at the Siwah Oasis was of course far away to the southeast, near the Egyptian border. A few non-African Greek cities occasionally show Zeus Ammon on their coins.

⁶⁷ Mazard, , op. cit., no. 89, cf. 94–8, 103, 118Google Scholar. Roman coin of 43 B.C., perhaps connected with the arrival of Sextius' legions in Italy, RRC, no. 491. Earlier examples, nos. 402, 461.

⁶⁸ Grant, , op. cit., pp.45 ff.Google Scholar; and for a brief triumviral colony founded by one Cosconius, P., p.260Google Scholar. Reynolds, J., ‘Four Inscriptions from Roman Cyrene’, JRS 49 (1959), 97Google Scholar, records a suggestion by R. Syme that Augustan Sestii in Cyrene reflect the presence of L. Sestius P.f., quaestor to Brutus in 43—a very long shot.

⁶⁹ Romanelli, , op. cit., p.118 n.2Google Scholar (and La Cirenaica romana (Rome, 1941), p.60Google Scholar) accepts, as from Livy, Lucan's figure of two months from Berenice to Leptis; Gsell viii.3 prefers Strabo 17.3.20, round the Syrtis from Berenice in thirty days. Plutarch, , Cato 56.4Google Scholar, speaks of six days of march through the desert. Lucan 9.302 ff. exaggerates in regarding the sea voyage as virtually impossible, but Cato, having apparently lost some ships near Berenice, clearly preferred land. His only precedent seems to be Ophelias' march to join Agathocles, on which he lost many men; Kebric, R. B., ‘Lucan's Snake Episode’, Latomus 35 (1976), 380Google Scholar, thinks Lucan closely based on Diodorus' account of this march (20.42, perhaps from Duris?), including his chronology. Cf. Aumont, J., ‘Caton en Libye’, REA 70 (1968), 318.Google ScholarGoodchild, R. G., (Libyan Studies (London, 1976), p.145)Google Scholar notes that Cyrenaica and Tripolitania were never united except for a brief period in the sixth century A.D. In 1865 George Dennis complained that the camel-mail from Benghazi to Tripoli took three weeks and storms along the coast cut communication by sea for long periods (Rhodes, D. E., Dennis of Etruria (London, 1973), p.87)Google Scholar. Cornificius had no camels, and inferior ships, if any.

⁷⁰ Romanelli, , op. cit., p.130;Google Scholar cf. Gsell viii.163.

⁷¹ Dio 48.17.6. Grant, , op. cit., p.234Google Scholar, argues that coins of Malta evoking those of the Liberators suggest that for a time it was controlled by their admiral L. Staius Murcus-surprising if Cornificius had a fleet, though strictly Malta was attached to Sicily. Hadas, M., Sextus Pompey (New York, 1930), p.74,Google Scholar supposes Cornificius himself had beaten off from Lilybaeum the supporters of Antony (or mere pirates?) of ad f. 12.28.1. More likely he only arrested them on their return to or arrival in Africa. Some think, by contrast, that Sextus visited Cornificius: the confused chronology of Tertullian, , De Pallio 1.2,Google Scholar has suggested that the Pompeius who set up ‘trinas aras’ at or near Carthage is Sextus, not Magnus (so Dessau, H., Klio 8 (1908), 460CrossRefGoogle Scholar).

⁷³ Appian, , BC 4.55Google Scholar, Cicero, ,Phil. 3.10.26.Google ScholarSydenham, , Roman Republican Coinage (London, 1952), p.212,Google Scholar suggested Carthage. Crawford, , op. cit.,Google Scholar does not commit himself, but on p.738 suggests that Metellus Scipio had showed the city goddess of Utica on his coins.

⁷³ Appian, , B.C. 4.54Google Scholar; Dio 48.22.4.

⁷⁴ Ad f. 12.30.4. his need of money ‘ad rem militarem’.

⁷⁵ CIL viii. 1441Google Scholar; Thibursicum Bure perhaps a pagus of Carthage, Pflaum, H. -G.‘La Romanisation de I'ancien territoire de la Carthage '’, Antiquités africaines 4 (1970), 75.CrossRefGoogle ScholarMünzer, RE iv.1627.Google Scholar

⁷⁶ CIL x.8314–15Google Scholar, Münzer, , RE iv.1627.Google Scholar Other Comificii in Africa may derive from L. Cornificius, relationship unknown, cos. for Octavian in 3 5 and subsequently governor of the province: RE iv.1623Google Scholar (s.v. Cornificius 5); possibly the man who wished to prosecute Milo for ambitus, Asconius, In Mil. 38–9 C, and did so de vi, 54 C. Perhaps not the P. Cornificius who attacked Milo in the Senate in 52 (36 C)? Syme, R., Historia 4 (1955), 60–1.Google Scholar

Who is the Cornificius of ad f. 12.25.1? He cannot be a son of Quintus, at least by OrestUla's daughter, so early, and ad Att. 12.14.2 and ad f. 8.7.2 rule out a brother.

The name is rare, though there was a Cornificius who was scriba to Verres in Rome and Sicily, crooked according to Cicero, and granted the gold ring by his master (In Verr. 2.1.150, 157; 2.3.181–7). How the name got into the family of the Emperor M. Aurelius, who had a sister and a daughter called Cornificia, we do not know.

⁷⁷ Appian, , B.C. 4.56Google Scholar (none of the proscribed is identifiable).

⁷⁸ Dessau, , op. cit. in n.71.Google Scholar He notes the hot springs at Carp is.

⁷⁹ Münzer, , RR xii.411 (Laelius6).Google Scholar

⁸⁰ Appian, , B.C. 4.53.Google Scholar

⁸¹ Ad f. 12.22a.4, 23.1, 25a.2, 30.2 and 5; ad Q.f. 1.1.14.

⁸² Ad f. 12.17, 27.1, 30.7; cf. 25.7, 22.4.

⁸³ I am grateful to Mr. R. G. G. Coleman, Miss J. M. Reynolds, and Professoi T. P. Wiseman for reading and criticizing this article, and to Mr. T. Volk for helping me to obtain the numismatic literature.