VincentHunink.nl

Home > ONDERZOEK > VERTALINGEN | EDITIES | PUBL.LIJST | PROJECTEN ||| BRONNEN | INDEX

---

 

commentary on 

LUCAN BELLUM CIVILE III, 3,509-762

(The naval battle near Marseille)

---

published in: Vincent Hunink, Lucan's Bellum Civile Book III, a commentary; (Gieben) Amsterdam 1992

---

509-537 On both sides preparations are made for the naval battle.

(1) Summary:

The Caesareans prepare for battle at sea, using unfinished vessels. The Roman fleet is positioned near the Stoechades islands. On the Massilian side, young and old are deployed on all available ships, even those already scrapped. On a quiet, sunny morning both fleets start approaching each other. Among the various ships the Romans use, the huge ship of their general Brutus is the most conspicuous.

(2) Structure:

In this short scene, the account of the naval battle (509-762) is introdu­ced after having been connected to the foregoing scenes on the siege. In its first part (509-20), both sides equip and crew their ships. The second part (521-37) shows how the fleets come closer, and provides some additio­nal details on the impressive Roman fleet.

(3) Historical material:

From Caes. Civ.1,56-8; 2,4-7; and D.C. 41,21; 41,25 we get the clear impres­sion that the Massilians took the initiative in both naval conflicts, had more ships at their disposal, and were more experienced in nautical matters. As on many other occasions, Lucan seems to have adapted the facts for his tale: he plays down all Massilian resources and qualities except their courage and determination, and suggests that the Caesareans not only started the fight, but also possessed more, stronger and larger vessels; cf. also on 510.

(4) Literary material:

Though the scene seems to be rather technical, several elements reflect epic traditions. The sunrise is a common element to all epic poetry (cf. on 521). Elements like the description of the `battlefield' and the ordering of the troops belong to the sphere of battle scenes, whereas Lucan's interest in ships and nautical affairs gives the scene a very special colour.

Scenes on military preparations at sea are rather exceptional in BC: cf. 4,417-47 (preparation of rafts by Caesareans near Salonae); further 169-297[1]. For military preparations in general, cf. the scenes adduced above on 375-398.

509         spes - abit: `the defeated lost their hope of success on land', a concise but clear expression, in which the poet gives a deliberate distor­tion of the facts, as in the preceding lines; cf. on 498. First, as I have noted before, the siege did not end with the beginning of the naval battle, but was continued during and after the two historical naval battles; cf. e.g. on 509-762 (3). Second, the initiative for the naval battle was probably not just a Roman initiative, especially considering Massilia's famous nautical expertise, for which cf. RE XIV,2,2146-9. Finally, although the Caesareans had faced temporary setbacks, they could hardly be called victi[2] by now. Throughout the Massilia blocks in book 3, Lucan sides with the Massilians, sacri­ficing truth to effect whenever necessary.

profundo...maris: not to be taken together, as BOURGERY a.l. says and others seem to suggest. At sea (profundo) the Caesareans try their fortunam maris. EHLERS is the only one who clearly renders the nuance (`...vor der Küste ihr Glück im Seekrieg zu versuchen'). For fortunam maris OUDENDORP compares Justin. 2,11,19. Profundum and mare equally stand side by side in 651-2, as HOUS­MAN notes.

510         maris: the reading of Z, generally preferred by modern editors (except GRIFFA) to the ablative mari found in the other MSS.

non...: the Caesareans' ships have not been properly finished off and lack the usual ornaments. From Caesar's own account we gain the impression that his ships were made in great haste or improvised (Caes. Civ.1,36,4-5; 1,58,3; 2,5,1). However, Lucan gives no such explanation, but merely conveys the image of rough, massive struc­tures which seem typical of a barbaric tribe rather than of civilized Romans; see also below on 512 and 513. For the negation antithesis non... sed... cf. on 49.

robore picto: some translators think that these words refer to the tutela. This is not impossible, since both the ship and the tutela could be made of wood. But painting in bright colours was used especially for decorative purposes on parts of a ship, as the bow; cf. CASSON 1971,211-2. Moreover, this tutela is said to be fulgens, a word not suggestive of paint (see below on tutela). Finally, it seems more natural to connect robore picto with ornatas...cari­nas than with fulgens (which would leave ornatas standing isolated). Considering these arguments, the words are more likely to refer to the ships.

511              tutela: a statue of the ship's guardian divinity, an essential feature of every Roman ship, placed on the stern; cf. 558; further e.g. Ov. Tr.1,10,1; Petr. 105,4; 108,13; Stat. Theb.10,186; Sil. 14,410. It could be gilded, as in Verg. A.10,171 aurato fulgebat Apolline puppis or made of ivory, as in Sen. Ep.76,13; cf. CASSON 1971,347wn13. This may explain Lucan's epithet fulgens. The ending of the line is imitated in V.Fl. 1,301 fulgens tutela carina.

512         rudis - conseritur: trees in their natural state are fastened together into something more primitive than the raft built in 4,417-26. By their form­lessness and lack of decoration, these constructions recall the barbaric statues in the sacred grove of the Gauls (412-7). The periphrasis qualis procumbit - arbor seems based on Caesar's note that his ships were made from unseasoned wood (ex umida materia Caes. Civ.1,58,3), but may also contain a distant echo of the motif of felling trees, for which see on 440.

513              stabilis - bellis: `a stable site for fighting at sea', a typical Lucanean paradox. By using the raft-like constructions, the Caesareans are in fact continuing the land war at sea, rather than starting a genuine naval battle. In a later stage this stability will prove to be disadvantageous for the Romans; see on 553. For the phrase cf. 6,60 area belli[3]; Ov. Fast.5,707 apta area pugnae; for the dativus finalis dependant on a substantive OBER­MEIER 1886,43-4[4]; further LHS II,99.

514         et iam - classis: a Roman fleet comes down the Rhône. From Caes. Civ.1,36,4-5 we know that Caesar had ordered the construction of 12 warships at Arles. They were completed in no more than 30 days. According to RAMBAUD 1976,857 the fleet had been built between April 29th and May 30th of the year 49 B.C., and arrived near the Stoechades (see below) on June 2nd or 3rd.

turrigeram: `bearing a tower', said of a ship, as in 4,226; Sil. 14,500; cf. also Verg. A.8,693. Towers on ships are attested to already in the Helle­nis­tic period; cf. Plb. 16,3,12. They were used on low lying warships to provide height for missile laun­ching machines. Often they were painted in distinc­tive colours, so as to facili­tate strategic operations; cf. further CASSON 1971,122; VIERECK 1975,26; REDDÉ 1986,95-8. In the present text, the mention of the tower reinforces the link established in 513 between the naval battle and the preceding siege; cf. also Plin. Nat.32,1; for towers in the siege see on 398.

Bruti: D. Iunius Brutus Albinus, Caesar's general who was left in command of the fleet; cf. RE Sup.V,369,30ff; and see also on 761. In book 3, he is the only general who is mentioned by name. He does not occur in Roman poetry apart from BC.

comitata: to be taken with the subject classis, which is postponed until the end of the following line, perhaps to convey the image of the slowness of the vessels.

515         in fluctus: that is, into the sea; cf. OLD s.v. fluctus 1b. The words should not be connected with Rhodani. Rhodani cum gurgite points to the movement downstream on the Rhône; cf. on 514.

516              Stoechados arva: a periphrasis for the Stoechades, a group of islands commonly identified as the `Iles d'Hyères', east of Toulon. However, the use of the name has been extended to all islands dependant on Massilia; cf. e.g. CLEBERT 1970,117; DIRKZWA­GER 1975,76-80; RIVET 1988,223-4. Lucan's unusual singular Stoechas might point to one specific island[5], but it seems more likely that Stoechados arva is a rather vague designation of place, especially since the poet has fused the two battles fought at different locations, and uses the poetical phrase arva tenere, for which cf. Verg. A.2,209; 6,477; 6,744; 10,741; see further the use of tenere above in 182. Used slightly paradoxically in connec­tion with ships, arva tenere reflects Lucan's special interest in the border between land and sea; cf. on 60.

nec non et: the combination also occurs in 7,56 and 10,486. For nec non see further Schmidt on 10,133; LHS II,778-9.

Graia iuventus: for the ending cf. 355.

517         suum fatis - robur: Lucan presents the Massilians' preparations as a desperate attempt to resist brutal force with courage and determination. In fact, Massilia with its long seafaring tradition was better equipped for a naval war than Caesar, who had to improvise some ships almost on the spot. For Massilia's resources in this sphere, cf. Caes. Civ.1,56; 2,4. Cf. also on 509.

518              grandaevos - ephebis: old men and youths belong to the groups traditionally not taking part in combat, but significantly they are armed here. This effectively turns Lucan's battle into a life-and-death struggle[6], of which the outcome is already certain at the start.

Grandaevos is a solemn, poetical word, also used in 7,371; further e.g. Verg. G.4,392; Aen.1,121. With senes it is used in Sen. Ag.378; Oed.838; Sil. 4,29; Tac. Hist.3,33. On the other hand, ephebus, a Graecism for puer or iuvenis, occurs only rarely in poetry after the early Republican period. Before Lucan cf. only Catul. 63,63; Hor. Ep.2,1,171; Ov. Ars 1,147; Sen. Her.F.853. In BC it is also used in 6,562.

519              stabat: here in the nautical sense of `lying at anchor', as in 8,592; cf. SAINT-DENIS 1935b,104-5; OLD s.v. 8.

520              emeritas...alnos: a second element emphasizing the life-and-death struggle of the Massilians: they patch up scrapped ships taken from their dockyards. This is confirmed by Caes. Civ.2,4,1, who additionally mentions the use of fishing boats. Emeritus is often used of things, cf. e.g. Ov. Fast.1,665; Plin. Nat.17,206; Mart. 10,85,5 (of a ship). With navalibus we must think de, e or a, as in Verg. A.4,593. For navalia cf. on 182; for alnus on 441.

521         ut matutinos...: the second part of the scene 509-37 opens with a poetical descrip­tion of sunrise, a traditional element in epic poetry. This case (like e.g. 7,1-6) clearly shows how its function has shifted from merely denoting time to creating atmo­sphere and tension in preparation for events to come. Cf. already Verg. A.3,521; 588-9; 4,6-7; 4,584-5 a.o.; cf. REEKER 1971,77n177; for sunrise in com­bination with winds subsiding see Verg. A.7,25-8. Lucan may also have been influenced by A. Pers.386-98. His picture is powerful: though the rays of the sun are said to be refrac­ted on the surface of the water (cf. KÖNIG 1957,182-3, comparing the light effect in Verg. A.8,22-5), there is no mention of colour[7]; and though the sea and the air are calm, there is no peace and harmony. Nature is full of grim, silent tension, a calm before the storm perhaps comparable to 5,424-55. On sunrise and sundown in epic see BARDON 1946; further on 40 pronus in undas; in addition see HÜBNER 1976.

spargens: as KÖNIG 1957,183wn1 notices, the word is often used with lumine and an object in the accusative in Vergil's sunrises ; e.g. A.4,584; 9,459; 12,113. Here, more dynamically, the light itself has become the object.

aequora: more than `sea' as DUFF renders it. Considering the following descrip­tion the word has retained its original sense of `level surface' here: the sea is pictured as calm and flat.

522              nubibus aether: for the ending see Lucr. 6,268; Ov. Met.13,582.

523         posito Borea: `as the north wind had dropped'. This medium sense of ponere also occurs in Verg. A.7,27; 10,103; Ov. Ep.7,49. For Boreas see on 69.

pacem - Austris: for the ending cf. 68 cessantibus Austris; for Auster see on 1.

524              servatum bello iacuit: `lay calm and ready for warfare' (HASKINS). The tension is elevated by the ominous calm at sea. The motif will be used again at great length in 5,424-55, where Caesar's ships can no longer sail because of a complete calm.

For servare in the present sense cf. also LUNDQVIST 1907,161; OLD s.v. 8. Iacere is used of water as in 5,434 and 443; (cf. 4,119 and 311); further e.g. Ov. Met.11,747; Man. 1,249; Sen. Ag.449; Tro.199; Nat. 2,6,4; V.Fl. 3,732; Juv. 12,62. The MSS PUV read iacuit bello instead of bello iacuit, which is the commonly accepted word order.

omni: to be taken with statione. Most transla­tors neglect the word, except EHLERS and CANALI.

525         suam: ZMUC read sua, probably an error due to attraction by statio­ne.

paribusque lacertis: on both sides the rowing starts. Though warships were equipped with sails, these were not used during the fighting; cf. CASSON 1971,235‑6. For lacerti referring to rowing cf. Verg. A.5,141 adductis spumant freta versa lacertis; Sil. 14,358. Par can be used to describe opponents in a battle or contest; cf. OLD s.v.11‑12.

526         Graio remige: for the synecdoche in remige cf. 530; 673; 754; 9,149; further Ver­g. A. 4,588; 5,116; Ov. Met.6,445, a.o.. For Graius indicating the Massili­ans, see on 302.

527              tollitur: in the context of rowing, as in Verg. A.10,295; V.Fl. 1,340.

impulsae: in a nautical context, the direct object of impellere can be a ship (e.g. Caes. Civ.3,40,2; Verg. A.5,120; Sil. 1,568); or the sea itself (e.g. Verg. G.1,254; Ov. Met.3,657) or even remos (cf. Verg. A.4,594); see SAINT-DENIS 1935b,70. For the trem­bling of the ship the scholiasts already compare Verg. A.5,198; cf. further V.Fl. 2,77-8.

tonsis: tonsa is an archaic word used in epic as a synonym for `oar', cf. 539; 5,448; further e.g. Enn. Ann.218 Sk (with Skutsch's note); Lucr. 2,554; Verg. A.7,28; 10,299; V.Fl. 1,369; Sil. 6,363. In ZM we find a variant reading contis already mentioned in a scholium in a 12th.ce­ntury MS; cf. CAVAJONI 1979,197.

528              crebraque - puppes: if we translate `and frequent strokes tugged at the tall ships', the phrase adds nothing new to 526-7, `tall' functioning merely as a traditional epithet for ships. Probably the upward movement of the ship[8] due to the impulse of the oars is meant here; cf. Sil. 14,379 with Spaltenstein's note. In that case sublimes would be proleptical: `so that they rose up'; as in EHLERS' `und die Rümpfe bäumten sich, wenn Schlag um Schlag sie vorwärts riss'[9].

529              cornua...: 529-34 briefly picture the general arrangement of the Roman fleet: the larger warships, like triremes and quadriremes, form the wings of the fleet (529-32), which is crescent-shaped (533). In the centre, smaller vessels are positioned (533-4).

The syntax of 529-32 is extremely difficult: we should construct as follows: cornua Romanae classis triremes et... cinxere, with cornua as object of cinxere, and multiplices rates as apposition to the threefold subject (triremes; quasque...commovet; et plures... pinus). Cornua cingere seems a bold expression meaning `form the wings, thereby encircling the centre'; cf. HOUSMAN a.l.; METGER 1957,17wn1-2, comparing Caes. Civ.1,83,2 and Liv. 23,29,3 for this use of cingere. The complex syntax and word order seem deliberate here: they add to the impressive image of the Caesa­rean fleet resulting from the various warships mentioned. Significantly, we hear nothing now about the arrangement of the Massilian fleet: the Caesareans are clearly to be seen as the aggressors.

triremes: the triremis (Gr. trièrès) was one of the dominant types of warships throug­hout antiquity. Surprisingly, no good ancient picture of the vessel is extant. Though it is not certain to what the element of `three' in the name of the ship corresponds, it is now generally accepted that the trireme had three superimposed banks of oarsmen; cf. CASSON 1971,77-80; VIERECK 1975,39-41; WARRY 1980,94; REDDÉ 1986,110-2; for a modern reconstruction of the ship see MOR­RISON/COATES 1986 and WELSH 1988. A more complex problem will be discussed below on 530.

530              quasque - commovet: `and those driven by ranks of rowers rising four times in layers above each other' (HASKINS). Lucan's words seem to allude to four layers of rowers, but this causes a serious problem. With quadrire­mes and other, larger polyremes there is a similar difficulty to that of the triremes: the rowing arrange­ment on these vessels is not known for certain. However, most scholars agree that for technical reasons no Greek or Roman ship could have had more than three superimposed layers of oarsmen. We may reasonably assume that a combination of superimposed banks and several oarsmen to an oar was used. Thus a quadrireme might have three levels of oars (two handled by one oarsman, one by two), two (each oar handled by two men) or even one (handled by four men); cf. CASSON 1971,97-116; VIERECK 1975,46 and 56; WARRY 1980, 94; 98-9; ROUGÉ 1981,93-5; further OLD s.v. quadriremis `...probably with four rowers to every "room"'[10]. Similar explanations may be given for larger ships, like the quiqueremis or the hexeris.

Con­fron­ted with this theory, Lucan might be said to be simply `wrong' here. This is actually stated in RE Suppl.V,922,49ff. However, this has provoked reactions in defence of Lucan. It has been suggested by REDDÉ 1980 (repeated in REDDÉ 1986,55-9) that the present text is a most important piece of evidence against the commonly accepted theory. In his view Lucan is talking of ships with four and more superimposed layers of oars. This same idea is implicit in most translations of the passage, which seem to have overlooked the problem here. But REDDÉ has fallen into a trap, questioning archeolo­gical theories by means of BC, which is a rhetorical poem rather than a handbook of maritime technology.

Lucan's text must be explained in a different way. In lines 530-2 and 535-7 Lucan gives rather vague and poetical periphra­ses for various polyremes. Significant­ly, those larger than the triremes are not even mentioned by name. On the whole, Lucan's descriptions may be reconciled with the commonly accepted theory on these ships; see notes below.

But if we suppose that he actually describes ships with more than three levels of oars, he may also be referring to types of ships of which we know nothing at all. It seems possible that he was influenced by naumachiae here (cf. on 509-762 (4)). Perhaps extraordi­nary vessels with more than three layers of oarsmen were con­structed for these shows.

However, the main aim of the poet is not to give an accurate account of maritime realia, but to enhance the pathos in his text. As with so many other details concerning geography, history, astronomy and other areas, he may have delibe­rately adapted his material to suit his rhetorical purpose. We probably should not ask whether his words are `wrong' or `right', but how they function within their context.

Thus, quater - ordo may be taken as a reference to four oarsmen (cf. remige), who do not necessarily sit in a strictly vertical line and may be handling one large oar (as even REDDÉ 1980,1035; and 1986,56-7 notes) or two or three oars, or even four oars arranged in two adjacent pairs in one `room'. But other explanations of Lucan's words are equally possible. In particular, the poet seems to have adapted his description both to suggest the enormous dimensions of Caesar's ships, and to surpass Verg. A.5,120 terno consurgunt ordine remi who refers to just three layers of oars[11].

531         plures - pinus: `(ships) that plunge more oars into the deep'. This place is con­sidered as the only instance of pinus in the metony­mical sense `oar'; cf. HOUS­MAN a.l.; OLD s.v.2c. We meet the same problem as in 530. If the poet refers to `more than four oars'[12], he may well mean regular polyremes with units of e.g. two and three oars (5), or twice three oars (6), twice two and once three oars (7), etcetera. But he may also be deliberately exaggerating the image of the Caesarean fleet; cf. above on 530. For the ending cf. Stat. Silv.2,6,28; Claud. 1,246.

532              multiplices: either `various' or `numerous' or `complex'. All three senses suit the context here, but the first seems best. Multiplices­...rates is an apposition to the threefold subject of the period, as mentioned above on 529 cornua.... For cingere cf. ibid..

robur: the word is sometimes used for the strongest part of an army or task-force; cf. 7,545; OLD s.v.8.

533              lunata classe recedunt: the ships are lined up in the shape of a half moon, with the heavy ships at the wings, and the lighter ones in second line in the centre. The naval formation is well known; cf. Prop. 4,6,25; Sil. 14,369-70; and especially Veget.4,45[13]. For Lucan's general interest in the moon, see also on 42.

534              ordine - Liburnae: contrary to the other types of warships the liburnian is not of Greek origin. It was a fast galley developed by the Liburnian pirates living on the Dalmatian coast (they are mentioned themselves in 4,530; cf. 8,38). The vessel was adopted by the Romans[14] and widely used in their fleets during the principate; cf. CASSON 1971,141-2; ROUGÉ 1981,124-5; REDDÉ 1986,104-10.

It is generally agreed upon that the liburnian was constructed with two layers of oars, each oar being handled by one oarsman. Lucan gives a poetical periphrasis of this, personalising the liburnians as `content to have risen with two layers'. He may be referring to the oarsmen rather than the oars, as possibly in 530, although both are certainly `double' in the case of the liburnian[15].

535              celsior at cunctis...: the picture of the Roman warships culminates in the huge vessel of its commander Brutus. The ship has already been mentioned as turrigera in 514; in 558-66 it will start the actual fighting.

praetoria puppis: `an admiral's flagship', according to OLD s.v. praetorius 1c, comparing Liv. 26,39,18; Fron. Str.1,1,2; Flor. Epit.1,23,7. Whereas Lucan gives several details about the ship, Caesar merely says that Brutus' ship could be easily identified by its insigne; Caes. Civ.2,6,4. Lucan's description has influenced Sil. 14,384-91.

536              verberibus senis: `with sixfold strokes'. Again, the problem of ancient polyremes emerges. As in the case of the quadrireme, Lucan does not mention a name for the ship, but he is probably referring to the hexeris[16], for which cf. Liv. 29,9,8; 37,23,5; V.Max. 1,8. ext.11. Analogous to what has been said about the quadrireme (above, 530), this text cannot be taken as proof for the theory that this ship had six superimpo­sed banks (REDDÉ 1980,1035 and REDDÉ 1986,57). Lucan's text mentions six strokes in the water, that is, six oars, but this may refer to units of six oars arranged in two adjacent groups of three[17].

But the descrip­tion of Brutus' huge ship seems to have been the result mainly of the poet's imagina­tion. It clearly functions as a climax in the pathos of the section 529-537 (see on 529), and may be far removed from maritime reality. For the phrase cf. Sil. 14,487-8 senis... pu­lsibus.

agitur: for agere used to describe propelling ships by means of oars, see Verg. A.5,116; Ov. Ep.12,7. It may also be used to refer to sailing, as in Lucr. 4,390; Sil. 7,242.

profundo: the passive invehi can govern a dative, as in Liv. 23,47,8; Tac. Ger.40,2. The active form as used here is uncommon; cf. OLD s.v. inveho 4.

537         et summis - remis: the translation `reaches for the water far below with its topmost tier' (DUFF) seems rather awkard: surely all oars are reaching for the water. Only if we put all stress on longe, this interpretation makes sense, the oars of the topmost layers being necessarily longer than the others; thus e.g. HOUSMAN; EHLERS; LUCK; WID­DOWS. In any case, the topmost layer is not the sixth but probably the third; cf. discus­sion above on 530. Less likely, though not impossible is `...with the end of its blades' (cf. BOURGERY; CANALI).

Perhaps we may simply render: `...with its very high oars', with summis compa­ring not the various layers of oars to each other, but the ship and its features as a whole to the surrounding ships (cf. 535 celsi­or - cunctis)[18].

aequora remis: for the ending cf. Verg. A.3,668; Ov. Ep.2,87; 3,65; Sil. 14,538. The whole line is quoted by Serv. on Verg. A.10,207 (reading ferit for petit).  

538-82 The fleets draw up close to each other. After some initial manoeu­vres, the first close combats take place, resulting in many casual­ties.

(1) Summary:

The fleets are moved within each other's range. Some initial naval manoeu­vres take place, but these do not have any effect. The first close combats do not occur until they are provoked by a move of Brutus' flagship. The fighting rapidly becomes intense and causes many casualties.

(2) Structure:

Now that preparations have been made, this second scene of 509-762 descri­bes the initial phases of the battle. It may be divided in four sections: First, (i) the fleets approach each other (538-42); then, (ii) three naval manoeuvres are carried out, ramming (542-5); shooting missiles (545-6); and encircling (547-57), but these have no effect. (iii) After Brutus takes measures to start the fighting (558-6) finally, (iv) close combats and killing begin (567-82).

(3) Historical material:

Lucan's version is pro-Massilian as usual, making the Caesareans responsa­ble for the encircling and the start of the fighting (see on 547 and 558). With the exception of a few other items, such as the different nature of the ships of Massilians and Caesareans (see on 553; 556), his version seems to have little relation with historical sources on the civil war; in general, see on 509-762 (3).

(4) Literary material:

Though the beginning of the scene suggests that the battle will start in traditional epic fashion (see on 538), it is actually postponed until the fourth part (567-82), where attention is immediately focused on unusual and unexpected forms of death.

The tension which has been built up in the prece­ding section (509-37) is therefore not immediately resolved. The rather abstract naval manoeuvres in (2) and the concrete action in (3) function as retarding elements which create the impressi­on of disquiet and continuing tension.

As a whole the section belongs to the scenes of collective fighting. It is inter­rup­ted only by the short intervention of Brutus, which thus stands out even more clearly. Somewhat surprisin­gly, the poet pays relati­vely little attention to the difference between the contesting parties. In the section on naval operations, he does not specify them in any way before 553. The ships damaged due to Brutus' action are likely to be Massili­an, but again it is not made clear to which side the victims falling in 567-82 belong. Attention seems to be shifted from the contrast of the parties (so dominant in 298-508) to the highly pathetic and surprising `special effects', the various ways in which marines get killed; see further on 509-762 (4).

As often, some technical terms are employed (cf. notes on e.g. 545; 547; 548), but taken as a whole the descriptions are either abstract or provide poetical detail (as in 549-52). The scene is clearly not intended to give a faithful or technical account of the events, but to maintain the atmosphere of tension in the scene and prepare for the even more pathetic individual combats to come after 583.

The scene contains many elements founded in or alluding to epic conven­ti­ons; cf. various notes below, e.g. on 538; 540; 542; 543, a.o.. Especial­ly in 567-82 they appear much more frequently than usual; for a complex example see e.g. on 572. In addition, Aeschylus' Persae seems to have been a model in some places; cf. 540; 542; 547; 566). For the traditi­on of the First Punic War as reflected in Plb. 1, see on 509-762 (4).

The section on naval manoeuvres closely matches the initial phases of the land battle described in 455-92 (see on 455-508 (2)). Here too, there is a succession of three operations. They are equally described in such a fragmen­tary and abstract manner that we can hardly gain a good impressi­on of what actually happens. Finally, these naval operations are no more effective than those performed during the siege, and lead up to a climax which does bring about some success (cf. 497-508 and 558-66).

The fierce fighting scene 567-82 is the first of a long series of scenes full of death and destruction; cf. further on 509-762 (2).

538         ut tantum...: both fleets come into close proximity of each other. Lu­can adopts the epic topos of the beginning of battle in an unusual maritime context. In this he may have been particular­ly inspired by Verg. A.5,137-50, descri­bing, perhaps significantly, not the start of a real fight, but of a boat race. Although the present scene appears to develop in a traditi­onal, epic manner, the fighting itself is postponed until much later; cf. on 538-82 (4); further e.g. on 540.

utraque - tonsis: `(so much...) as either fleet could cover by throwing out its oars just once'. A creative variation for the conventi­onal idea of parties being `within range of a missile', as in Verg. A.11,608-9; Ov. Met.4,709-10 (cf. also Bömer's note a.l.); V.Fl. 8,303 (at sea); Stat. Theb.5,361-2; 6,354; Sil. 4,101-2[19]. For excutere HASKINS and OLD s.v. 7 compare Ov. Met.5,596 excussaque bracchia iacto (of a swimmer), but the verb is also often used for throwing missiles, which seems relevant here. For tonsa see on 527.

539              posset: the reading of MYGPEJ preferred by all modern editors to possit of ZABRQUVW; cf. GOTOFF 1971,50.

540              innumerae - tubae: normally, the shouts and battle cries come only after the trumpet has sounded; cf. e.g. 7,475-8; Verg. A.5,139-41; 9,503-4. Here, they do not only precede the official signal, as in A. Pers.386-95, but even drown it. It cannot be heard any more than the sound of the oars. Therefore it does not break the tension as it would normally do. In a different context, the trumpet is drowned out by human voices in 6,165-6, where Scaeva's words rouse a greater fury than a trumpet would do. For the combinati­on innumerae voces cf. Stat. Theb.7,111; 10,147; Silv.1,6,81; Claud. 15,485. For the ending aethera voces cf. Ver­g. A.8,70; V.Fl. 2,241; Rut.Nam. 1,203.

542              tubae: the most popular type of horn. It was made of bronze, and had a straight tube, contrary to the lituus, bucina and cornu. It was already used in 2,690 in a maritime context, as in Sil. 14,373.

tum...: lines 542-57 describe three naval operations carried out by the ships. Here the crews start to row arduously. It is not immedia­tely clear why they are doing so, but considering 544-5 it must be the first move in the manoeuvre of ramming. Significantly, no distinction is made between Massilians and Romans; cf. on 538-82. On naval manoeuvres in BC, cf. SAINT-DENIS 1935,425-35; for ramming cf. A. Pers.408-11 and CASSON 1968,100-1.

caerula verrunt: `sweep the blue waters of the sea' a conventional phrase; cf. Catul. 64,7; Verg. A.3,208; 4,583 (cf. also Verg. A.8,671-4). The underlying model is Enn. Ann.377-8 Sk verrunt extemplo placidum mare: marmore flavo / caeruleum spumat sale conferta rate pulsum. As PATERNI 1987,106-7 has noticed, the tradi­tional contrast of white foam and blue water has been avoided by Lucan, who does not favour bright colours (cf. on 98; further e.g. 238; 503). The colour blue occurs only here and in 2,220 caeruleum aequor, and seems reduced to a mere formulaic element. For caerula used to describe the sea cf. also TLL III,107,27ff.

543              atque: explicative, since the two notions of 543 merely illustrate and explain the foregoing caerula verrunt.

in transtra cadunt: after bending forward to dip the blades[20], the oarsmen fall back on their benches while pulling the oars. Transtra are `rowers' benches', as in 731; 8,671; 10,495; cf. SAINT-DENIS 1935b,117; CASSON 1971,220. The phrase as a whole is suggestive of arduous rowing and should not be taken too literal[21]. In employing it, the poet may be alluding to the much less violent movement of considere transtris as expressed by Ver­g. A.3,289; 4,573 and 5,136.

remis pectora pulsant: the thought is repeated in an even more visual descrip­tion: the men pull the oars so strongly that they beat their chests. Of course, this is poetical exaggeration once more. The periphrasis for rowing is conventional; cf. Enn. Ann.218 Sk poste recumbite vestraque pectora pellite tonsis and 219 Sk ...exim referunt ad pectora tonsas[22]; cf. further Verg. A.5,141; 8,689-90; Ov. Met.11,461-2; Sen. Ag.437-8; V.Fl. 1,369; Stat. Theb.5,375; Sil. 11,489. For pectora pulsant as verse ending cf. 4,182; 7,128; 608; Ov. Met.12,234.

544         rostris - rostra: a poetic manner of describing a collision between ships. For the phrase cf. 1,6-7 infestisque obvia signis / signa....pila minantia pilis which is an echo of En­n. Ann.582 Sk., a fragment preserved by the Comm.Bern. a.l.; further e.g. 4,783; En­n. Ann.584 Sk.; Verg. A.11,615; Ov. Met.14,301.

The rostrum is the beak of a Roman ship, its most powerful weapon, used for ramming; cf. CASSON 1971,85; VIERECK 1975,22-3 (with illustra­tion); REDDÉ 1986,84-90. It was reinforced with iron or bronze. This explains the sound expressed in crepuere, for which see 657; further 9,288 sonus increpat aeris.

545              puppem: this form is found in PV and Prisc. GLK II,330,2f, specifically quoting it as an example of the accusative on -em. Other MSS (as well as Prisc. GLK III,53,19f have the more regular puppim. Cf. HOUSMAN a.l..

in puppem rediere: `backed astern'. After hitting their adversaries, the ships with­draw, thereby completing the manoeuvre of ramming, as in 659. Since the rostra are said to have hit each other, the operations remain without effect here. For the phrase cf. SAINT-DENIS 1935b,95-6 who compares V.Fl. 5,211 in proram rediit (of a turning move­ment). Following the Comm.Bern. a.l. EHLERS interprets in puppem different­ly, rendering `als die Schiffe eine Bootslänge zurück­gin­gen'. But the phrase indicates the direction of the movement rather than the distance. For in with accusative cf. on 112.

tela: after the attempts at ramming, in a second phase, described only briefly, missiles are launched. This is done either by hand, as in 567, or by means of tormenta as in 716.

546         aera - pontum: a poetical description of a mass of arrows. For covering the air by missiles cf. 7,519 ferro subtexitur aether; Verg. A.11,611 caelumque obtexitur umbra (sc. of tela); 12,578 obumbrant aethera telis; Sen. Epigr.52,3 caelum texere sagittae; Sil. 4,550-1 a.o.[23]. In cadentia we see the conventio­nal comparison of weapons to clouds, snow, hail or rain, as in 2,262; 4,776; 6,134; further e.g. Hom. Il.12,156; Enn. Ann.266 Sk.; 391 Sk.; Verg. A.12,283-4; Ov. Met.5,158; see also HÄUS­SLER 1978,165-7 and above on 482. The image of missiles covering the surface of the sea is a point Lucan seems to have added here (cf. further on 566).

vacuumque: the missiles fall into the sea where there are no ships, and therefore have no effect.

547         iam diductis - proris: `now they deploy their wings by spacing out the ships', a descripti­on of preparations for the third naval manoeuvre in the present passage, that of encircling. At first, it is not made clear which party is actually attempting it. It is only from 553-7 that we learn that it is the Caesarean side. By contrast, the Massilians are the ones who try to encircle the Romans according to Caes. Civ.1,58,1. In addition, Caesar mentions diducere navibus in Caes. Civ.2,6,2 as a move made by the Romans, but not as a preliminary step to encircling as it is here. Lucan seems to have combined the ideas and ascribed them to the Caesareans; cf. also METGER 1957,21-2. For the idea of encircling cf. also A­. Pers.417-8 (by the Greeks); Plb. 1,23,8-9 (by the Cathagi­ni­ans).

Both diducere (proras) and extendere cornua are military technical terms whose use has been broadened to include the sea; cf. SAINT-DENIS 1935b,51 and 61; TLL V,I,1016,76ff and V,II,1971,51ff. For extendere cornua in a maritime context cf. already Liv. 36,44,1. Se­veral MSS read deductis (PVC) and rostris (ZMGC), but these variants for diductis and proris are general­ly rejected. Prora is used as pars pro toto for `ship'.

548              diversae - receptae: `and the ships of the enemy were admitted after the fleet had been extended'. Laxata classe is another military expression applied to the sea; cf. SAINT-DENIS 1935b,78; TLL VII,1071,83ff. It further explains 547, while the rest of the line describes the actual encircling.

549              aestus: considering the nature of the image (see below), the sense `rough sea' (cf. OLD s.v.7) seems the best here.

ut quotiens...: lines 549-52 illustrate the turbulence and inner conflict of the water moved by the oars on both sides. It is compared to the turbulence due to winds, causing the waves at the surface to move in the opposite direction to the deeper levels of the water.

Strictly speaking, the comparison is not accurate: the turbulence on the horizontal level of the water caused by the oars is compared to turbulence on the vertical level caused by a conflict of wind and water. Moreover, in the present situation ships are not likely to move in opposite directions at all: they are more likely approaching or pursuing each other[24]. However, we must not take the text too literally: it is a poetic picture of a dynamic, agitated sea; cf. METGER 1957,22-3. The conflicting streams of the water may be seen as representing the conflicting parties; for a similar transposition of civil war to the scale of nature, see on 60 qua mare....

The main element in the comparison belongs to the same sphere as the element to which it is compared; see on 41 tantum.... Comparisons involving the sea are particu­larly frequent in BC; cf. e.g. 1,100-3; 260; 498-503; 2,189-90; 665-8; 672-7; 715-9; 4,134-6 a.o.; for a full list and analysis cf. SAINT-DENIS 1935,421-5. For the conflict of wind and sea see especially 2,454-60 (with Van Campen's notes) and 9,333-4; further e.g. Ov. Met.8,470-2; Sen. Med.940-3; Ag.138-40; 488-9. On the relation of the present text to full scale storm scenes, see MORFORD 1967,52[25].

Zephyris Eurisque: the west and southeast winds. Only shortly before the Boreas and Auster were named (523). All four quarters of the sky have thus been alluded to in the present passage, emphasizing once again the world wide dimen­sions of the present conflict. For Zephyrus cf. 1,407; 2,676 (Eurum Zephyrumque); 4,72; a.o..; for Eurus see on 232.

Eurisque repugnat: the MSS MZY read Eurusque, while GJ have Eurisve. The right reading Eurisque is found in Z²ABR and the other MSS; cf. GOTOFF 1971,117. Serv. on Verg. A.5,2 quoting 549 and half of 550 reads furiisque. Instead of repugnat PV have the incorrect repugnant.

550         huc - mare: this is one of the rare verses with a bucolic dihaere­sis; SHACKLETON BAILEY's index (p.287) mentions as other examples in book 3 only lines 90 (a rather weak case) and 633. Cf. also on 552.

fluctus: the waves on the surface, agitated by the wind, as opposed to the deeper layers of the sea; see on 549.

illo: after huc we might expect illuc, which is actually found in V and in Servius' quotation mentioned on 549. However, all modern editors prefer the reading adopted in all other MSS.

551              sulcato - tractus: `as the ships drew various tracks by ploughing the water'. `Ploug­hing the sea' is a conventional poetic metapher; cf. e.g. Verg. A.2,780; 5,142; 158; 10,197; Ov. Met.4,707 (with Bömer's note); Am.2,10,33; Stat. Theb.8,18; Sil. 14,362. But here it is used not merely as a poeticism for sailing: it takes up a more literal sense of agitating the water, the key element in the passage. The result is made visible in duxerunt tractus, a phrase recalling Lucr. 2,207 (of a comet).

552         quod: to be connected with aequor. As MORFORD 1967,52 points out, the verse is built up with a quadruple chiasm, reinfor­ced by allitte­ration of the nouns in the centre. The balanced rhythm of lines 550 and 552 sharply contrasts with the strong inner turbulence they describe.

tulit remis: the meaning of tulit and rettulit is not fully clear, but I take them to refer to a horizontal movement of the water[26]; cf. on 549. For remis HAS­KINS has `with its prow', which is plainly wrong.

haec: the reading of PUV,generally preferred to hoc of ZMG.

553         sed Grais habiles...: it is only in lines 553-7 that a difference is made between the parties (cf. on 547). With their light ships, the Massili­ans are able both to provoke the Caesare­ans, and to escape them. Therefore, it seems likely that the Massilians are the ones who are encircled. Earlier, in the course of the siege, they were at a disadvantage because of the lack of flexibility of their war machines; cf. on 478. Now a similar disadvanta­ge arises for the Romans because of the sluggishness of their ships (see also 513).

Caesar confirms both the manoeuvrability of the Massilian ships and the slowness of those of his own troops, thereby adding further weight to the victory of the latter; Caes­. Civ.1,58,1 and 3. Here, the contrast of the two types of ships merely indicates the ineffectiveness of the encircling manoeuvre. Cf. further on 556.

Grais: a possessive dative, with ellipse of the verb esse.

habiles: on this adjective four infinitives are dependant: lacesse­re; temptare; frangere and cedere. Habilis constructed with infinitive is extremely rare; the only other case is Stat­. Theb.4,225. The constructions with gerundi(v)um or with ad and accusative are more regular.

lacessere: the generally adopted reading of ZM, all other MSS reading capessere. The combination with pugnam is quite normal, cf. Liv. 37,16,9; Verg. A.5,429; Stat. Theb.1,413; cf. also BC 4,720[27].

pinus: `ship'; cf. OLD s.v. 2a.

554         nec longo - cursum: `to break off their course with a short turn'. Nec must be taken with gyro only. For the use of frangere cf. Stat. Theb.10,183; 12,232; further OLD s.v. 6c.

555         tarde: one of the rare places where DUFF departs from the text of HOUSMAN and, for that matter, all editors: he writes tardae following the MSS GU. It would then be constructed with an infinitive, like habiles (which would have three instead of four infinitives dependant upon it). However, this would split up 553-5 into two parts, creating an image far less suggestive of speed and manoeuvrability. The MSS authority is clearly in favour of tarde, which should be taken with nec just like nec longo in the preceding line.

clavo: the helm, a bar used for moving a steering oar (gubernacu­lum); cf. 9,345; Enn. Ann.508 Sk.; Verg. A.5,177; 10,218; a.o.. Ancient rudders were not installed at the stern but both at the port and starboard; cf. in general on rudders SAINT-DENIS 1935b,64-7; CASSON 1971,224-8; REDDÉ 1986,80-4. Flectenti...clavo is only paralleled by Sil. 14,403 following the present text.

556         at Romana - usum: `but the Roman ships were more reliable in providing a stable bottom to the fighters, and a use similar to land'. This has already been alluded to in 513 and will re­turn later in 566. For the idea cf. also 5,708 ut terrestre, coit consertis puppibus agmen.

The slug­gish, heavy ships of the Caesare­ans recall those first used by Romans in the First Punic War (264-241 B.C.) against lighter Cartha­ginian vessels; cf. Plb. 1,20; 22-3; 51; esp.1,23,6-7 paraplèsion gar pezomachias sunebaine ton kin­dunon apoteleis­thai; see also below, and in gene­ral on 509-762 (4).

Romana ratis: according to OPELT 1957,438 this expression reveals patriotic pathos overshadowing Lucan's anti-Caesarean attitude. However, I cannot see patriotism here any more than in 529 Romanae classis. The mere allitterati­on can hardly count as a patriotic signal. The echoes of the First Punic War (see above) rather bring out an ironical contrast between the glori­ous Ro­man achie­vements of the past and Cae­sar's impious acts in the present civil war.

558         tunc in...: after the three successive operati­ons (542-57) have led to nothing, Caesar's gene­ral Brutus initiates close combat by a provocati­ve action: he orders his helmsman to expose the sides of his ship to the enemy's attacks. The closest model for the passage is Caes. Civ.2,6,4-5[28], where two triremes assail Brutus' ship; it rapidly moves away, causing the assailants to crash. Lu­can's version is markedly different: first, Brutus provokes the conflict; second, his ship is actually rammed by some of the enemy's ships; although these are caught, it can hardly have got away un­damaged itself; finally, the incident is used as a starting point for the close combat.

Within the context of the mass scenes of 538-82, the concrete image of Brutus and his ship stands out clearly[29]. It brings the tension and pa­thos to a climax, and resolves it. Characteristical­ly, Lucan does not tell what happened further to Brutus' ship: once the first attacks have occurred, the text develops into a rather abstract mass scene again.

signifera: obviously an epithet of puppe. It has been taken as referring to the image of a ship's tutela (see on 511); cf. HASKINS; OLD s.v. signifer 1 `adorned with images'. But a tutela was present on many ships, whereas Brutus' ship is repeatedly mentioned as unique (514; 535-7). Probably the word refers to military signa, in this case the ensign of the flagship; cf. also SAINT-DENIS 1935b,113.

magistro: poetical for the steersman of a ship, as in e.g. Ver­g. A.1,115; 5,176; 6,353; Ov. Ars 1,6. The normal term was guberna­tor; cf. CAS­SON 1971,310. For the ending puppe magistro 1,501; further Stat. Theb.8,269; Silv.5,3,127; Sil. 1,687; 3,153 a.o.; for steersmen cf. further on 593.

559         ait: a full and neutral introduction of a speech, as in 435. Ait is remarkable here, since it is mostly used in the parenthetical type of intro­duction, as in 38 and 716; cf. SANGMEISTER,65-6. For the short speech star­ting in the middle of 559, see on 38.

pateris...: Brutus' words seem to display a traditional Roman fighting spirit, averse to tricks and strategems. But ironically, the Roman ships took the initiative, at least in the third manoeu­vre (547-57); see also on 560.

acies: TLL V,808,64ff explains acies as classes, comparing other examples of errare referring to ships, Verg. A.5,867; Ov. Fast.2,391 and B.Afr.44,1. However, the abstract meaning `battle' seems more to the point in this context of the beginning of a battle scene and produces a stronger effect; cf. WID­DOWS' free but apt transla­tion `why allow this to become wide ranging, a battle of movement?'.

560              artibus - pelagi: interpreted by most transla­tors as a dative dependant on certare expressing the field of rivalry: `against naval manoeu­vres'. In this case Brutus would be formally disclaiming any respon­sibility for the manoeuvres. But we may also take it with CANALI and WIDDOWS as an ablative expres­sing the field of rivalry: `in naval manoeu­vres'. This gives a subtler, more ironical sense: like a principled Roman, Brutus is condem­ning the manoeu­vres - after they have been tried in vain. For both con­structi­ons cf. OLD s.v. certo 1a-b.

561              Phocaicis: referring to Phocaia in Asia Minor, the mother city of Massilia; hence poeti­cal for `Massilian', as in 583 and 728; see on 172.

rostris: the steersman is ordered to expose the sides of the ship to the enemy[30], so as to provoke attempts at ramming. This move is in­tended to take away the enemy's advantage of mobility. The double hyperbaton in 561 may suggest the confrontation itself.

562              obliquas - alnos: an explanation of medias - carinas. For the phrase cf. Liv. 37,30,6 obliquas se ipsae ad ictus praebebant; 28,30,10; Sen. Phaed.1073 ne det obliquum latus.

563              tum...: the passage on Brutus' ship is con­cluded by a pathetic climax: it is rammed by Massilian vessels, but these cannot withdraw after hitting it and get caught, either directly, or through grappling-irons and chains, or because their oars get entangled. For the model and the function of the passage see on 558 tunc in...; for its pathos see NOWAK 1955,85-6.

robora Bruti: an allusion both to the mate­rial from which the Roman ships were made (510-3) and to the strength they represent (see on 529).

564              percussae capta cohaesit: `got stuck in the (ship) it had hit, and was caught'. An explanati­on of the striking paradox ictu victa suo. OUDEN­DORP compares Liv. 37,30,10 and Sil. 14,381-4 for the idea. The MSS provide several variants for percussae (found in MZQY): percussa et (Z²ABR­UVEWJ), percussaque (G) and percussa est (P); cf. GOTOFF 1971,117. But cohaereo is regularly constructed with the dative (cf. Ov. Met.4,553; 5,125; further OLD s.v.1), and the commonly accepted text makes excellent sense.

565         ast: the archaic form of at. In poetry it remained in use as a metrically convenient vari­ant; cf. Austin on Verg. A.2,467; TLL II,942,83ff. In BC 3 it is also used in 754.

manicae: normally the word denotes hand­cuffs, but here it is used in an excep­tional way as a syno­nym of harpago or manus ferrea (cf. Caes. Civ.1,58,4 and 635): a `grappling-iron'; TLL VIII,301,83f gives no paral­lel for this meaning.

teretes: `tightly twisted', `firmly woven', a rare meaning, given only by Lewis/Short s.v. I. There seems to be no Latin parallel for its use with catenae, although one may compare exam­ples like Hor. Carm.1,1,28 (plagae); Ov. Fast.2,320 (zonam) and Sen. Phaed.45-6 (laquei)[31]. Possibly Lucan is imita­ting the Homeric eustrephès, used to describe ropes and cables in Od.10,147 and 14,346[32].

566         seque - remis: the ships' oars get entangled, probably with the oars of Caesarean ships. For remis the MSS PC have remi, a less likely, though not impossible reading.

tecto...aequore: another paradox: the sea is covered with ships, as it was with missiles in 546 and will be with blood in 572-3 and with corpses in 575 (cf. also on 652 clauso...mari). Only shortly before, in 546, the sea was entirely free of ships. This is the sort of extreme contrast Lucan likes very much.

The idea of the sea being covered by ships is a topos dating back to A. Pers.419-20 and Hdt. 7,45[33]. In Latin literature cf. e.g. Verg. A.4,582 latet sub classibus aequor; Man. Poet.3,20; Eleg.Maec.45; Sen. Suas.2,3,8; (Sen.) Oct.42; Sen. Ag.41; 434; further Man. Poet.1,776; Juv. 10,175-6. The combinati­on of the two para­doxes in 566 may have been inspired by Liv. 26,39,12-3.

stetit...bellum: `the fight became stationary' (HASKINS). Now the paradox already alluded to in 513 (stabilis - bellis) and 556-7 (stabi­lem - usum) has come to a climax: finally, the Romans have achieved their aim in making the fight like a battle on land. For the use of stare cf. Sil. 14,519 steteruntque rates ad proelia nexae. In addition, it may recall the sense `to last, to continue' of Livian phrases like stetit... pugna (e.g. Liv. 7,7,7; 27,2,6; 29,2,15).

567         iam non...: after Brutus has provoked the first clashes, close combat starts taking place. The mass scene 567-82, formed by seven successive paradoxes, brings out the violence and novelty of the fighting. For the priamel in 567-70 see on 101; for catalogues of horrors see on 342; similar cases involving horrors of war are e.g. 6,169-79; 7,619-30; 764-76; 789-94. Cruel and violent deaths are not merely part of the war, as in Homer or Vergil, but seem to have become the main theme; see METGER 1957,27-9; see also on 509-762 (4).

excussis...lacertis: `as the arms were shot out'. Excutere is used for the movement of the arms in throwing missiles, cf. OLD s.v. 7 and above on 538. For the phrase cf. 1,424 excusso...lacerto; 4,386; Ov. Ep.4,43; Pont.2,9,57; Sen. Ben.2,6,1; Stat. Theb.10,745.

tela lacertis: for the ending cf. 2,502 (torserunt tela lacerti); further e.g. Ov. Met.12,79; Stat. Theb.5,378; 9,904.

568         nec - ferro: not merely a paraphrase of 567: the notions of distance (longinqua) and of hitting (cadunt...vulnera) are added.

cadunt...vulnera: considering the verb, vulnera must be rendered as `wounding weapons' here, as in 314. Cf. also 7,514 cadunt mortes.

vulnera ferro: for the ending cf. 9,678; Ov. Hal.61 and further SCHUMANN 1983, s.v..

569              miscenturque manus: `but they fought hand to hand'. -Que is adversative: after the notion of fighting at a distance has been repeated once again in 568 (cf. 538-9; 545-6), here it is turned into its opposite. For the pathos cf. NOWAK 1955,21-2.

570         ensis: the sword plays a major role in the naval battle. This paradox illustrating the foregoing miscenturque manus brings the recurrent idea of a land battle at sea (cf. on 566) to a new climax. It is closely imitated in Sil. 14,521 et gladio terrestria proelia miscent. For ensis and gladius see on 323; for the sword as a special weapon in epic associated with force and furor, see METGER 1957,33-6.

stat - puppis: `each man leaned forward from the bulwark of his own ship' (DUFF). The Latin is concise, stat indicating a firm but not fully upright position.

571              pronus - ictus: `exposing himself to hostile blows' (cf. HASKINS, LUCK and CANALI). An active translation, `eager to deal blows to the enemy', is possible as well, and many scholars (starting with the Comm.Bern.) have chosen that approach. But in view of the constant stress in book 3 on victims rather than those who kill, on opposition and defiance of weapons rather than the use of them, the first seems more appropriate. For the idea cf. e.g. 707-8; 4,480 admoto occurrere fato and HÜBNER 1975,206n45; in general RUTZ 1960.

nulli: the generally accepted reading of ZM, for which the other MSS read multi.

perempti: not surprisingly, words and expressions for `killing' are frequent in any battle scene. Lucan's vocabulary in this sphere shows subtle changes in relation to Vergil's. For instance, Vergil's favourite words sternere and caedere (29 and 27 times) are used less frequently (7 and 16 times) in favour of synonyms like perimere (8 in Vergil, 14 in Lucan); cf. WEBER 1969,45-52.

572              cecidere: in a regular naval battle, men struck by missiles tend to fall on their own ships or in the water. Here, paradoxically, they fall onto the enemy's ships. The poet focuses on the unusual character of this falling by using a negation (nulli - in suis) and adding no other detail. It seems relevant to note that cadere is a common word in epic poetry for `to die'; cf. WEBER 1969,52. Lucan may well be playing with this sense here.

cruor - spumat: `blood foamed deep upon the sea'. The basic epic topos is that of water colouring red due to blood[34]. For blood see on 124; for its flowing during the battle, cf. on 589 and 639; for the imagery of blood and water see below on 577.

But Lucan goes further, using several other motifs: the water does not simply turn red, but is actually hidden from view by thick layers of blood, a variation of the motif of the covered sea (see on 566), possibly inspired by Achilles' words in Acc. Trag.313 W Scamandriam undam salso sanctam obtexi sanguine. Furthermore, the normal foam of the waves is replaced by foam of blood, recalling Enn. Trag.118 Joc. maria salsa spumant sanguine; Verg. A.6,87; 9,456 (spumanti sanguine rivos) and 9,700-1; cf. also BC 4,758; 7,699. Finally, this motif is given a new twist: the foam starts to clot, which recalls Verg. A.2,277; Ov. Met.12,270; 13,492; 14,201[35].

altus: with blood, as in 1,329 and 2,214. It is not impossible to interpret cruor altus with METGER 1957,48n1 as `Herzensblut', coming from the chests defiantly exposed to the enemy in 571. Still, it seems prefera­ble to take altus as a predicative adjective indica­ting the quantity of blood (cf. also Verg. A.11,633). This is a better explanation for both the foam (spumat) and the clotted blood (concreto).

573         et obducti - fluctus: this image explains the foregoing difficult state­ment, and adds a final point; see on 572. Lines 572-3 have been closely imitated in the medieval Vita Willibror­di by Thiofrid, 4,200-1, cf. ROSSBERG 1883,152.

obducti concreto: with obducti we must think sunt. In some MSS (PV) this was apparently misunderstood and instead of concreto they wrote concres­cunt.

574         quas: sc. puppes (575).

immissi - ferri: `the chains of the iron launched upon them', a curious­ly exact periphra­sis for the instruments described in 565. For the ending vincula ferri cf. 2,72; 4,466.

575              conferta: `packed closely together'. This reading, found in ZGVPE is generally preferred to conserta of MZ²ABRQYUWJ; cf. GOTOFF 1971,50. Cf. 4,490 confer­ta... corpora. HOUSMAN a.l. compares Lucr.6,1263; V.Fl. 3,274; Juv. 10,186. The word may also be taken as grimly ironical, since it is often used to describe the close order of living troops (e.g. Verg. A.2,347); for this view see LYNE 1989,117. The idea of water having become impenetrable through piled up corpses is an epic topos, cf. 2,209-20; Hom. Il.21,218-20; Catul. 64,359-60; Verg. A,5,806-8; Stat. Theb.9,436-7; Sil. 1,45-9; Claud. 21,186-7.

cadavera: the word is avoided in classical poetry, as it had a very negative ring. In epic it was usually replaced by corpus; cf. AXELSON 1945,49-50; Norden on Verg. A.6,149[36]. Lucan has a particular fancy for this crude word: no other Latin author shows a more frequent use of it (36 cases); Seneca comes closest (with 18 cases). HÜBNER 1976b,303 has observed that the sense `fallen' is more accurate here than `corpse'[37]. The ambiguity inherent to the first is similar to that of cadere (above, 572).

576              semianimes: a word indicating the vague condition between life and death. Cf. e.g. Enn. Ann. 484 Sk.; Verg. A.4,686; 10,396; 404; Ov. Met.5,105 7,577; see ESPOSITO 1987,40wn4. Lucan has 5 other cases of it.

subiere: in a nautical context, subire is normally used in the technical sense of `making one's access into a place', such as a harbour or a coast; cf. OLD s.v. 10; SAINT-DENIS 1935b,107. But Lucan has returned to the literal sense of a vertical movement into the water; for which cf. Ov. Met.15,358. Here subiere will become even more ominous through periere of 579 which closely echoes it in metrical position, etymology and sound.

vastum...profundum: the heavy u-sounds as well as the hyperbaton seem to suggest the deepness and vastness of the sea. Cf. also 651.

577         suo - pontum: the imagery of water defiled by blood belongs to the epic tradition; cf. above on 572-3; it will return in 638-41 and 661[38]. The horrific image here echoes Man. Poet.5,666 inficiturque suo permixtus sanguine pontus; and Ov. Met.12,326 where a man's blood pours into his drinking cup (the former has been noted by SCHWEMMLER 1916,10; the latter by METGER 1957,47); cf. also Ov. Met.4,728-9 mixtos cum sanguine fluctus / ore vomit. Lucan's phrase has been imitated without much fantasy by Sil. 4,593.

578              luctantem - trahentes: `drawing their last breath which struggled with slow death'. For luctantem animam (cum morte) cf. Verg. A.4,695; Sen. Phoen.142-3; further e.g. Sen. Oed.344; Apoc.3,1; Stat. Silv.1,4,107 pugnantem animam. Lucan has effectively combined this with the notion of breathing one's last; cf. OLD s.v. anima 1b.

579              periere: cf. on 576 subiere. The men who were already struggling with death (after having been injured or threatened with drowning) now finally get killed in another form of death. Similar duplications occur later in the battle; e.g. 687-90; 748-51; see also on 587.

periere ruina: for the ending cf. 5,637; 9,969; for ruina see on 290.

580         irrita - caedes: `weapons which missed their target did some killing of their own in the water', a paradox illustrated and clarified in 581-2. Lucan gives a new twist to the motif of demoniac weapons behaving as living beings, a form of animism common in epic. Cf. e.g. Hom. Il.4,126; 521; 11,574; 21,70; Verg. A.11,804; see FUHRMANN 1968,38. For missiles striking at random cf. e.g. 6,78-9; 7,485-8; 514-7; further e.g. Sen. Phaed.816-9; for irrita tela cf. 722; Verg. A.2,459; 11,735; Prop. 4,9,40 a.o..

581              frustrato: almost a synonym of irrita in 580; cf. also TLL VI,1,1440,36ff with further parallels.

pondere ferrum: for the ending cf. 725; 4,776; further e.g. Ov. Am.3,8,37; Sil. 6,355.

582              exceptum: sc. by a victim; cf. 601.

invenit vulnus: this particular idea of a weapon `finding wounds' (explaining suas - caedes) forms the climax of the section 567-82, esp. lines 575-6. It may also have been inspired by texts like Liv. 26,39,13 ita in arto stipatae erant naves, ut vix ullum telum in mari vanum interci­deret (quoted by METGER 1957,70n1). After Lucan, it returns in Stat­. Theb.8,526; 11,512; Sil. 4,140-1.

583-646 In a series of single combats, several warriors meet their death.

(1) Summary:

While displaying courage and vigour, the Roman soldier Catus perishes in an uncom­mon way. So do the Massilians Telo and Gyareus, as well as one of a pair of twins. The ship his body falls on is sunk. Lycidas meets an equally violent end.

(2) Structure:

In this third part of 509-762 the battle is no longer described in scenes of mass fighting, as in the previous part, but in a sequence of individual fights. It may be divided according to its protagonists who are all clearly distinguished from each other: (i) Catus (583-91); (ii) Telo (592-9); (iii) Gyareus (600-2); (iv) a pair of twins (603-34); (v) Lycidas (635-46). The central part is clearly (iv): it is the longest and most elaborate scene.

(3) Historical material:

Considering the nature of the scenes, we should not expect them to have any historical basis. Still, the actions of one of the twin brothers in 603-34 echo among other things the heroic behaviour of the Caesarean soldier Acilius during the naval battles near Massilia; for Lucan's adaptations see on 609.

(4) Literary material:

The section consists of a number of relatively small, isolated scenes. In each of them, an individual warrior, identified by his skills or actions and often by a name, is represented in his final hour. All of them behave heroically, literally fighting to the death if they are given the opportunity. Individual combats are part of the epic tradition; so are extensive and detailed descriptions of wounds and violent fighting, and `frozen movements'; see on 509-762 (4).