¹ In many English translations the verb is translated as ‘call back’ or ‘recall’, but this is too weak for the context. Thomann's, translation in the Der Bibliothek der alten Welt series (Zurich, 1961)Google Scholar is even weaker: ‘Den vermesseren Brüdern entgleite das Königtum und falle ihnen als Verbannten wieder zu’.

² On line 33 Tarrant says ‘the language fits both Atreus and Thyestes, each of whom returned to power from exile’. The language fits even better if profugi is read.

³ For another example of ‘regnum repetere’ in Seneca cf. Oed. 794 ‘repetam paterna regna’.

⁴ This is the interpretation of many editors and translators (e.g., Weise, Francken, Ehlers, and Luck). Fantham refers to Bourgery for this interpretation, but in the Budé Bourgery takes sollicitant in the sense ‘wins over’ or ‘appeals to’ and the object of the verb to be ‘us’ (‘nous sollicitent’) in a general sense (see his note, ad loc). In her recent translation Braund renders sollicitant as ‘are incitements’ (with no object expressed).

⁵ The waters are muddied by the scholiastic tradition: ‘SOLLICITANT te scilicet, ut hoc sequaris’ (see Endt, Adnotationes); ‘SOLLICITANT commovent scilicet te ad bellum’ (Cavajoni, G. A., ‘Scholia inediti a Lucano del codex bernensis litt. 45 saec. x’, Acme 28 [1975], p. 95Google Scholar); ‘sollicitant, inquid, nos senatus consul et proceres alii, quod quasi omnes sub Pompeio esse videantur’ (Comm. Bern., ad 2.277).

⁶ Micyllus gives it as the second of two ways of construing sollicitant: ‘De Catone intelligendum videtur, non de Bruto, ut sis sensus: senatus et consules sollicitant et instigant te, (scilicet Catonem), ut una capessas bellum et arma, quibus si accesseris, iam nemo restabit liber praeter Caesarem, etc. Aut certe, ut in genere de omnibus accipiatur quasi dicat senatus et coss. communiter sollicitant et impellunt omnes ad arma secum capessenda: quibus si tu quoque accesseris, etc.’ Oudendorp is more forthright: ‘sed recte Micyllus monuit in genere accipiendum esse, sollicitant omnes, ut solet passim fieri quando casus verbi omittitur’.

⁷ The organisation of phrases with crescendo effect pervades Lucan's poem. Even when he departs from it, the final element often has some point. At 2.592–3 (‘Cappadoces mea signa timent et dedita sacris | incerti Iudaea dei mollisque Sophene’) the last subject is geographically more remote than the other two, being to the east of the Euphrates, in territory that the Parthians once thought should be theirs (cf. Plut. Pomp. 33).

⁸ It is an antithesis that appears elsewhere in the speech, for at 246–7 Brutus contrasts himself with alii at the same time as he contrasts Cato with Pompey and Caesar (‘namque alii Magnum vel Caesaris arma sequantur, | dux Bruto Cato solus erit’).

⁹ Much of what has just been said, however, rests on taking the antecedent of quibus to be the unexpressed object of sollicitant. This seems to have been the view of Housman (he says in his note that the object of sollicitant that is to be understood is ‘ceteros homines’, i.e. everyone other than Cato) and, with less certainty, of Micyllus (see above, note 6). The alternative, and at first sight the more obvious, antecedent is the multiple subject of sollicitant (‘pars magna senatus…consul… proceresque alii’)—add Cato to these (whom Brutus sees as already Pompey's slaves) as a further inducement and all men will be enslaved. I cannot see any objection to this interpretation, but, if the paradosis is retained, I prefer the other: the antithesis between Cato and Caesar is more pointed if Cato is linked, not with the leaders of the Pompeian side, but with those who are currently being won over by them.

¹⁰ For the insertion of -que to remove metrical problems as a result of corruption or wilful change see, for example, Morgan, J. D., CQ 36 (1986), pp. 197f.Google Scholar, on Propertius 4.6.74.

¹¹ In this case, the omission of -que in M and Z would reflect an intermediate stage between what I suggest was the original reading and what is now the accepted text: ‘proceres alios’ > ‘proceres alii’ > ‘proceresque alii’. Cf. Håkanson, L., PCPhS 25 (1979), p. 37Google Scholar (re 2.554): ‘This is, by the way, an instance where the quality of ZM is conspicuous; they retain the slight corruption quod and the correct hosti, whereas the other MSS have an interpolation’.

¹² For this part of the text Z and M are derived from a common source; see Gotoff, H., The Text of Lucan in the Ninth Century (Cambridge, MA, 1971), pp. 52–5.Google Scholar