Much of Book 10 of the Aeneid is taken up with scenes of battle. After the rampage of the Latin heroes Clausus, Halaesus and Messapus (10.345–61), the focus switches to Pallas on the Trojan side (10.362–8):

This is the text as printed by Mynors, Geymonat and Harrison,Footnote ¹ but line 366 presents a well-known crux. While most admit that quando postponed to an unparalleled sixth place in its clause is problematic, a further objection has been made by J.M. Trappes-Lomax, whose treatment of the passage is one of the most recent.Footnote ² He describes as ‘not exact’ the parallels quoted by Harrison in support of asper aquis, namely Aen. 4.526–7 aspera dumis | rura,Footnote ³ Sall. Cat. 59.2 planities … rupe asperă and Hor. Carm. 1.5.6–7 aspera | nigris aequora uentis. In the first two cases, he says, dumis and rupe ‘refer to items on top of an otherwise flat surface’; in the third case ‘uentis refers to that which has caused an ordinarily flat surface to be itself rough, whereas in our passage the Arcadians’ difficulties are caused by the saxa and arbusta rather than by any irregularity in the ground itself’. Trappes-Lomax is right in saying that Aen. 4.526–7 is no parallel for aspera aquis, since surface debris, not water-logged ground, is the hazard in Book 10; but what of the two other alleged parallels?

The objection of Trappes-Lomax to the relevance of Horace seems perverse.Footnote ⁴ He says that ‘uentis refers to that which has caused an ordinarily flat surface to be itself rough’, but his statement is exactly applicable, mutatis mutandis, to Virgil's phrase as emended by Madvig: aquis refers to the torrens (10.363) which has caused the ordinarily flat ground to be rough with obstacles posed by the rocks and uprooted trees. As for the passage of Sallust, the quotation implies that aspera qualifies planities, but that is not the case, as will become clear from the clause as a whole: uti planities erat inter sinistros montis et ab dextra rupe aspera. If this, the reading of the principal manuscripts, is correct,Footnote ⁵ it is obvious that aspera must be neuter accusative plural and that the passage is more tricky than we were given to believe. What is its meaning? The neuter plural aspera is frequently used of ‘rough/difficult things’ in a wide variety of contexts (for example Sall. Iug. 89.3 maiora et magis aspera adgredi; TLL 2.812.61–813.9), including places (for example Mela 3.40 in asperiora deuenit, Plin. HN 21.53 asperis et siluestribus; TLL 2.808.81–809.10); rupe, which means ‘a steep rocky cliff, crag’ (OLD), will be an ablative of cause (‘places difficult because of a/the cliff’). On the other hand, it is possible that ab dextra, usually understood as a set phrase = ‘on the right’ (as in Sall. Iug. 50.4; OLD s.v. dextera ¹ 3), is to be taken with rupe and that ab is causal (as in Vitr. De arch. 2.5.2 saxa … lenta sunt ab umore ‘pliant because of the moisture’; OLD 15). It is difficult to know which of these alternatives is preferable, but, if the meaning is ‘because there was level ground between the leftward mountains and the hazards from a cliff on the right’, then the passage seems to offer a reasonable defence of Madvig's aquis.

Trappes-Lomax, who finds problems also with line 367, proposes a radical solution: he regards ‘both lines as interpolated by one who wished to explain why the Arcadians were fighting on foot, although Vergil's intended readers surely did not need telling that cavalry cannot operate over saxa rotantia and arbusta diruta’. No doubt Virgil's readers did not need to be told the kind of ground over which cavalry can and cannot operate, but, if Trappes-Lomax's proposed deletion were to be accepted, it would not be clear from Virgil's text that ‘the Arcadians were fighting on foot’. All we are told in lines 364–5 is that the Arcadians were unused to making attacks on foot (insuetos acies inferre pedestris) and were retreating: since terga can refer to a cavalry retreat (Livy 22.47.3 equites terga uertunt), the natural inference is that the Arcadians, realizing that any advance would have to be made on foot, instead had wheeled their horses and were riding away. Thus, when Pallas a few lines later issues the order fidite ne pedibus (10.372), readers would get a surprise: with lines 366–7 deleted, there had been nothing to tell them that the Arcadians had dismissed their horses. This strongly suggests that lines 366–7 are required by the logic of the narrative.

Now, since the passage of Sallust which is used to defend asper aquis is immediately preceded by a sentence in which reference is made to the dismissal of horses (Cat. 59.1 remotis omnium equis), it is worth asking whether it is simply coincidence that the same two phenomena are juxtaposed in the very two lines of Virgil which Trappes-Lomax wished to delete (10.366–7).

Sallust's two sentences are taken from the final chapters of the Bellum Catilinae describing Catiline's last stand. Realizing that he was trapped outside Pistorium and could neither flee nor expect reinforcements, Catiline decided to engage with the consul Antonius in battle (Cat. 57.5). He therefore assembled his men and addressed them in a pre-battle hortatio of some considerable length (58). When he had finished speaking, he led his troops to a level area and had everyone dismiss their horses quo militibus exaequato periculo animus amplior esset (59.1); his men were deployed as dictated by the nature of the battlefield, with mountains on the left and a cliff on the right (59.2–3). Meanwhile, in another section of the field, Antonius was incapacitated by an ailment to his feet and had handed over command to his deputy, M. Petreius, who followed his deployment of the loyalist troops with a speech of his own (59.4–6). Then both sides came to grips in the battle which would seal Catiline's fate (60).

If we compare Sallust's narrative, which extends to three pages of the Oxford Classical Text, with Virgil's account of both Latin and Trojan activity (10.354–79), the results look like this:

certare in the impersonal passive and concurrunt are commonplaces of battle narrative, and the verb tendere can be expected to be used in any context where mental or physical endeavour is described. Likewise, references to spes and patria are conventional in pre-battle speeches.Footnote ⁶ Plunging into the enemy ranks is a generic feature of heroes (one thinks of the practice of deuotio),Footnote ⁷ and accendere is the mot juste for raising morale.Footnote ⁸ Taken individually, these similarities between the two passages seem insignificant; but the clustering within single episodes is suggestive: one could argue, for example, that in line 379 Virgil condensed two Sallustian references into one, substituted densos for confertissumos and prorumpit for incurrit, and subtly changed medios to the less obvious medius.

Rather different is the coincidence of At ex altera parte in Sallust and At parte ex alia in Virgil. Although ex altera parte is itself a very common expression in various forms, when preceded by at it is found only in the Bellum Alexandrinum (40.2 at … ex altera parte) and Livy (3.11.1, 10.29.3 at ex parte altera), all three examples being earlier than Virgil.Footnote ⁹ Likewise, parte ex alia in various forms is common, but preceded by at is found before Virgil only in Catullus (64.251 at parte ex alia) and Cicero (Arat. 367 at parte ex alia): although Virgil's debt to Poem 64 of Catullus is well known, the coincidence of the same phrase in Cicero's Aratea strongly suggests that all three poets were adopting a phrase from Ennius’ Annales which is no longer extant.Footnote ¹⁰ If that is so, has Sallust adopted the same phrase but changed alia to altera to suit his own context? We should remember that certare is an Ennian verb (six times in the extant Annales) and that the first appearance of concurrunt is also in Ennius (Ann. 144 Sk.). Livy has numerous references to making one's way with the sword (for example 4.28.5 ferro uia facienda est),Footnote ¹¹ but he also has ui uiam faciunt (4.38.4), which may be compared with Virgil's fit uia ui (Aen. 2.494). Although Horsfall on this last passage says that use of Ennius is ‘far from proved’,Footnote ¹² the frequency of the motif strongly suggests a common source: if so, it could be argued that this source, presumably Ennius, also lies behind Sallust's ferro iter aperiundum est.

The expression fortia facta, which occurs three times elsewhere in Sallust's works and once elsewhere in the Aeneid (1.641), seems a strong candidate for an Ennian reminiscence;Footnote ¹³ but we should note that, when Sallust uses the expression in our passage, he couples it with persons (59.6 ipsos factaque eorum fortia), as does Virgil (10.369 uos et fortia facta), and that Virgil's words uos et … facta are an exact reproduction of Sallust's uos … et … facta (58.18). These are just the kinds of detailed interconnection that one would expect when one author has another in mind. We have already seen that, although both Sallust (59.1) and Virgil (10.366–7) refer to the dismissal of horses, the reason in Sallust is the raising of morale but in Virgil it is necessitated by the state of the ground: this too is the kind of adjustment that an author makes when he is alluding to another. Catiline begins his hortatio with the conventional statement that words cannot induce courage (58.1 uerba uirtutem non addere),Footnote ¹⁴ but Virgil introduces Pallas’ hortatio with a statement of the opposite (10.368 dictis uirtutem accendit), a classic case of what has been called oppositio in imitando.

The polyptoton pede pes in Virgil's line 361 originates in Latin with Ennius’ Annales, where it is coupled with another (Ann. 584 Sk. premitur pede pes atque armis arma teruntur): this makes it very probable that Virgil's double polyptoton in the same line (uiro uir) is also Ennian, although in extant verse the only exact parallel for this particular doubling is Furius Bibaculus (fr. 10C/75H pressatur pede pes, mucro mucrone, uiro uir).Footnote ¹⁵ Such features support the notion that Ennius lies behind Virgil's battle description, whose numerous similarities to the narrative in Sallust could thus be explained by the reliance of both authors upon a common source; yet, as we have seen, a case can certainly be made that Sallust himself is ‘a possible model’ for Virgil here.Footnote ¹⁶

Sallust tells us that the sedition of Catiline and his followers was felt in Cisalpine Gaul (Cat. 42.1), and Eden comments that this was ‘perhaps the first political event to impress itself on the boyhood of Virgil (about seven at the time)’.Footnote ¹⁷ The Eclogues were written under the patronage of Asinius Pollio,Footnote ¹⁸ who shared with Sallust both a literary consultant and a Thucydidean style,Footnote ¹⁹ and in the Aeneid Virgil depicted Catiline, pendentem scopulo, and the younger Cato, dantem iura, alongside each other on Aeneas’ shield (8.668–70).Footnote ²⁰ Several recent scholars have suggested various ways in which Virgil may have been indebted to Sallust,Footnote ²¹ and it would not be at all surprising if, as he wrote about the conflict between Etruscans and Latins in Book 10, the events of a more recent civil war came into his mind.