¹ Crane, N., Two Degrees West. A Walk along England's Meridian (1999), 350Google Scholar.

² Dorey, T. A., ‘“Agricola” and “Germania”’, in Dorey, T. A. (ed.), Tacitus (1969), 4Google Scholar.

³ O'Gorman, E., ‘No place like Rome: identity and difference in the Germania of Tacitus’, Ramus 22 (1993), 135–54,CrossRefGoogle Scholar sees the Germania as an ‘exploration of a country (Germany) in search of the ideological (Roman) self’ (135). Many of her ideas about the Germania are also relevant to discussion of the Agricola.

⁴ For these periplus texts, see Müller, C. (ed.), Geographici Graeci Minores I (1855)Google Scholar.

⁵ For discussion of the text of the Germania, its relation to other ethnographic works, and later treatments of the land and people, see the excellent commentary by J. B. Rives, Tacitus. Germania (1999).

⁶ See Ogilvie, R. M. and Richmond, I. A. (eds), De Vita Agricolae (1967), 165Google Scholar.

⁷ Text, translation, and commentary are provided by Ramin, J., Le Periple d'Hannon. The Periplus of Hanno (1976)Google Scholar.

⁸ For the suggestion that the voyage is an imagined rather than a real one, see Jacob, C., Géographie et ethnographie en Grèce ancienne (1991), 73–84Google Scholar. It is, of course, also possible, even likely, that not only the voyage, but also its epigraphic commemoration was fictional.

⁹ See Homer, Od. 4.563 ff. on Elysium in the Ocean (see Strabo 1.1.5 for the islands of the blessed in the context of Homeric geography); Hesiod, Works and Days 170–7 describes these islands in some detail. Pindar, Ol. 2.68ff. continues the tradition. Of course the delightful nature of these islands is in contrast with the inhospitable places encountered and described by the explorers, but the abnormality of the island landscape and inhabitants remains a constant theme.

¹⁰ For an extensive and detailed treatment of Scylax, see Peretti, A., Il Periplo di Scilace. Studio sul primo portolano del Mediterraneo, Biblioteca di Studi Antichi 23 (ed. Gabba, E.) (1979)Google Scholar.

¹¹ Amiotti, G., ‘Cerne: “ultima terra”’, CISA 13 (1987) 43–9:Google Scholar ‘it represents not a geographical boundary, but a fantastical limit.’

¹² Cordano, F., La Geographia degli antichi (1992), 107Google Scholar: ‘in ancient literature, the name Thule indicates the northern extremity of the inhabited world.’

¹³ For the close relationship between geographical literature and fiction, see J. S. Romm, The Edges of the Earth in Ancient Thought (1992).

¹⁴ The idea of ‘constructed landscapes’ is interestingly developed in relation to the Germania by O'Gorman, op. cit. (n. 3), especially at 136.

¹⁵ See Strabo 1.4.6 for Eratosthenes' belief that the Western Ocean formed not a limit but a route to the world beyond and ultimately to India.

¹⁶ That Roman commanders from Pompey onwards, and partly in emulation of Alexander the Great, made political mileage out of ‘reaching the Ocean’ is apparent. See, for example, Cicero, de Imperio 33; Sallust, Letter of Mithridates 17; Plutarch, Pompey 38.2–3; Plutarch, Caesar 58.6–7. For Alexander, see Pompeius Trogus, Historiae Philippicae 12.7.4.

¹⁷ Braund, D., Ruling Roman Britain: Kings, Queens, Governors and Emperors from Julius Caesar to Agricola (1996), 12Google Scholar, expresses this neatly: ‘Britain lay both in Ocean and beyond Ocean, so that the conquest of Britain was also the conquest of Ocean itself.’

¹⁸ Dorey, op. cit. (n. 2), 17.

¹⁹ A. R. Burn, ‘Tacitus on Britain’, in T. A. Dorey (ed.), Tacitus (1969), 40. See also Burn, A. R., ‘Mare pigrum et grave’, Classical Review 63 (1949), 94,CrossRefGoogle Scholar for this passage as an accurate description of a ‘phenomenon familiar to sailors of small craft in the Pentland Firth’.

²⁰ Ogilvie and Richmond, op. cit. (n. 6), 173.

²¹ Further weight is given to this connection by the observation by J. O. Thompson, History of Ancient Geography (1948), 148–9, that there was an extensive tradition in geographical writings of discovering and describing sluggish seas in remote areas. He refers, for example, to Himilco's exploration of the outer parts of Europe, in which were found seas that were difficult to navigate. Certain phrases in Avienus’ account of Himilco's voyage are particularly revealing. See, Ora Maritima 121: aequoris pigri and 128: navigia lente et languida repentia.

²² Even such sensitive treatments as that of Liebeschuetz, W., ‘The theme of liberty in the Agricola of Tacitus’, Classical Quarterly 16 (1966), 135–6,Google Scholar can find little to link the description of Britain with the surrounding narrative.

²³ See Wiedemann, T., ‘Sallust's Jugurtha: concord, discord, and the digressions’, Greece and Rome 40 (1993), 48–57CrossRefGoogle Scholar. K. Wellesley, ‘Tacitus as a military historian’, in T. A. Dorey (ed.), Tacitus (1969), 69, places the geographical ‘digression’ of the Agricola in the category of origines et situs, which Tacitus discusses at Ann. 4.32ff. as being an integral part of the historiography of imperialism. The description of Britain would thus be motivated by the fact that the island was then for the first time fully conquered. This argument is somewhat supported by Tacitus’ announcement that he is moving on to discuss ‘Britanniae situm populosque’. It is disappointing that Ogilvie and Richmond entitle their commentary on chs 10–12 ‘The Ethnography of Britain’. My aim is in part to reinstate the Britanniae situs as being of at least equal importance to the populi.

²⁴ R. Dion, Aspectes politiques de la géographie ancienne (1977), 254.

²⁵ Burn, op. cit. (n. 19), 37, notes the disproportionate geographical interest in the work devoted to islands. ‘The Romans seem to have had an appetite for British island names’, adducing also Pomponius Mela's mention of 30 Orcades, Thule, and 7 Aemodae.

²⁶ For earlier formulations of the idea that the Ocean might encircle and encompass Roman power, note the ambitions of Pompey (Plut., Pomp. 38.2–3) and Caesar (Plut., Caes. 58.6–7). The flexible definition of where precisely the Ocean ran, beyond Britain or before one reached it from Gaul, could, of course, be useful in justifying either Roman expansion in that direction or the exclusion of the island from imperial aspirations. I owe to Chris Pelling the point that elsewhere Plutarch describes Caesar's British expedition as taking him beyond the confines of the Empire, that is beyond the Oceanic barrier and into a world that was scarcely believable. In the context of nebulous island-worlds, the phrase νῆσον ἀπιστουμένην ὑπὸ μεγέθους (Plut., Caes. 23) is particularly striking. Here, there is no suggestion that Caesar's conquest would threaten the insular nature of Britain; rather, its insularity is proof of Caesar's daring. As Pelling observes, Pompey's allusion to Caesar ‘calling the pools of uncertain depth an Ocean’ (‘Oceanumque vocans incerti stagna profundi’, Lucan, Civil War 2.571) might suggest that Caesar himself made much of the Oceanic adventure in his dispatches concerning the British expedition.

²⁷ The mental division of Britain into a ‘Roman’ and a ‘non-Roman’ section is clearly seen in Appian's preface, in which he outlines Roman acquisitions, including ‘the larger and better portion of it [sc. Britain], since they did not want the remainder at all’ (App., Roman History. Pref. 5). For the unprofitability consequent undesirability of Britain as an element in Roman rhetoric concerning the island's elusiveness from conquest, see Strabo 4.5.3. Braund, op. cit. (n. 17), 149, however, interestingly points to the way in which Scotland, under the name of Caledonia, was used by Flavian poets to express the rulers' aspirations to conquer Britain, telling against the idea of a northern segment of the island which could be mentally detached and dismissed.

²⁸ They may also be, according to R. Martin, Tacitus (1989), 43, used as a means of establishing that the Agricola is to be seen as a historical work. Ogilvie, R. M., ‘An interim report on Tacitus' “Agricola”’, ANRW II 33.3, 1720,Google Scholar also sees in the pair of speeches in the Agricola an echo of those attributed to Hannibal and Scipio in Livy 21.40–4. For a Tacitean parallel, see Annals 12.34, where Tacitus alludes to a prebattle speech given by the British chieftain, Caratacus. Many of the themes are very similar to those raised by Calgacus — libertas, virtus, maiores, the family. Just as in the case of Calgacus, Caratacus is defeated in spite of his brave words. His main speech in oratio recta comes after his defeat, when he has been taken to Rome.

²⁹ For Lucullus, see Plut., Luc. 23; for Pompey, see Cicero, de imperio Gnaei Pompeii 34–5, in which the anti-pirate command is seen as the spring-board for the granting to Pompey of the command over the war against Mithridates through the Lex Manilla. For Augustus, see his claim at Res Gestae 25 (‘mare pacavi a praedonibus’). This may indeed refer to the war against Sextus Pompeius in Sicily in 39–6 B.C., but the point is made more general in Horace, Odes 4.5.19: ‘pacatum volitant per mare navitae’ and in Suetonius, Life of Augustus 98, in which the Alexandrian sailors hail Augustus: ‘per ilium se vivere, per ilium navigare’.

³⁰ See Vitruvius 6.1.10–11; Strabo 17.2.1, but especially also 6.4.1.

³¹ It is worth remembering that Agricola himself is a Gaul, a detail which further confounds any attempt at strict divisions. On the implications of this, see Richmond, I. A., ‘Gnaeus Iulius Agricola’, JRS 34 (1944), 44Google Scholar. Braund, op. cit. (n. 17), 155, comments that Agricola's up-bringing in Gaul keeps him away from the corruption of Rome itself, a point to which we shall return.

³² See Cicero, De Oratore 1.42 (189) for the definition of definitio as ‘a short piece of description of those features which particularly characterize the thing which we wish to define’.

³³ O'Gorman, op. cit. (n. 3), 147–9, discusses this theme in relation to the Germania, raising the question of whether Germany could be seen as equivalent to primitive Rome, or ‘a reenactment of early Roman history’. In the Germania, just as in the Agricola, this question finds no easy solution.

³⁴ See R. Martin, Tacitus (2nd edn, repr. 1994), 39–49. He discusses the work as one in which the figure of Agricola as hero is maintained partly through acknowledgement of the fact that times have changed since the Republic and a new type of heroism is called for. However, I disagree with his view that Calgacus’ portrayal as worthy opponent is designed primarily to increase the stature of Agricola (44). It seems to me rather that Calgacus actually embodies what Agricola himself might have been like if he had not been a Roman general at the time of Domitian, and subject to the concomitant constraints.

³⁵ Again, Caratacus provides an illuminating parallel. His post-defeat speech is given in Rome, not in Britain, and it draws him even closer to ‘Old Rome’ than even Calgacus can come, since Caratacus not only is eloquent, but directly echoes the pre-death speech of Cremutius Cordus, a historian condemned for his libertas under Tiberius. Caratacus’ claim at 12.37 that execution, the final suppression of his libertas, will bring him eternal memorial is strikingly reminiscent of Cordus at Ann. 4.35. This parallel is all the more resonant given the framing of the Agricola with references to loss of senatorial libertas at Rome.

³⁶ The bibliography on Tacitus' views of libertas is vast, and a discussion of the subject lies far outside the scope of this paper. See, for example, Hammond, M., ‘Res olim dissociabiles: principatus ac libertas’, Harvard Studies in Classical Philology 67 (1963), 93–113CrossRefGoogle Scholar. The classic treatment is C. Wirszubski, Libertas as a Political Idea at Rome during the Late Republic and Early Principate (1950).

³⁷ See Martin, op. cit. (n. 34), 48; Ogilvie, R. M., ‘An interim report on Tacitus' “Agricola”’, ANRW II 33–3, 1718Google Scholar.

³⁸ Martin, op. cit. (n. 34), 41; Sage, M. M., ‘The treatment in Tacitus of Roman Republican history and antiquarian matters’, ANRW II 33.5, 3385–419,Google Scholar esp. 3388.

³⁹ Giua, M. A., ‘Paesaggio, natura, ambiente come elementi strutturali nella storiografia di Tacito’, ANRW II 33.4, 2897,Google Scholar argues that the Agricola reveals that ‘la lontananza e l'isolamento rispetto al cuore della civilità siano considerati garanzia di autentica libertà’.

⁴⁰ See Sage, op. cit. (n. 38), 3388–93, for the framing of the Agricola in a context of Republican style virtus. The opening chapters make clear that a significant contrast is to be drawn between the possibilities for outstanding deeds and their recording in the past and in the present.

⁴¹ See Pelling, C. B. R., ‘Tacitus and Germanicus’, in Woodman, A. J. and Luce, T. J. (eds), Tacitus and the Tacitean Tradition (1993), 59–85Google Scholar.

⁴² On the landscape of Germanicus' campaigns, see Giua, op. cit. (n. 39), 2887–90.

⁴³ However, it is worth noting the shift in the image of Ocean as no longer the proof of a great campaigning exploit (as for Caesar), but now a muted barrier, or not one at all. So, the traditional historiographical topos of the commander surpassing a mighty boundary may be toned-down in accordance with the times, and the new geographical image of the encompassed Britain may be appropriate for the new muted heroism. I owe this point to Chris Pelling.

⁴⁴ For imperial invidia against Agricola, see Ag. 39; for that against Germanicus, Ann. 2.22 (metu invidiae); 2.26. I. Shatzman, ‘Tacitean rumours’, Latomus 33 (1974), 574, suggests a further parallel in that rumours surrounding both Agricola and Germanicus, unusually for Tacitus, almost always cast them in a more favourable light than reality suggests.

⁴⁵ Braund, op. cit. (n. 17), 63.

⁴⁶ For Tacitus' belief in the opportunity for virtue and heroism even in a wicked age, see Percival, J., ‘Tacitus and the Principate’, G&R 27 (1980), 127–9Google Scholar. Another key example is Lepidus in Annals 4.20.

⁴⁷ The importance of Germany for our understanding of the Agricola is a recurrent theme. Not only does it provide the subject for Tacitus' other work in monograph form, but it acts as a foil in all kinds of ways. Germanicus' exploits are notoriously unfulfilled (Ann. 2.26 for the recall of Germanicus on the cusp of victory); Domitian's own expedition was incomplete (Ag. 39).

⁴⁸ We should perhaps in any case recall Calgacus' designation of the Romans as raptores orbis (30.4).

⁴⁹ However, the fact that there could be a fairly thin line to be drawn between great deeds and transgressive ones was neatly encapsulated in the tales surrounding Alexander the Great, whose conquests and ambitions could be seen as both mighty and hybristic, although this still does not match the transgressions of the Usipi. However, the point that greatness could be dangerous is very relevant to the question posed by the Agricola of how to be a good man under a bad emperor.

⁵⁰ Note the contrast drawn earlier between the Germans, for whom the only defence from Rome was a river, and the Britons, who could hide behind an Oceanic barrier. Here, the Usipi benefit from the geographical advantages afforded by Britain, and denied to their own people.

⁵¹ A similar disclaimer relating to discussion of the Ocean is given by Strabo 2.3.3, referring almost without question to Posidonius. If, as Ogilvie and Richmond suggest, Tacitus knew Posidonius' works, then it is quite possible that he was entering here into a well-worn topos of avoiding overlap with Posidonius' monumental work On Ocean.

⁵² It is striking, given other similarities between Agricola and Germanicus, that, when Germanicus is recalled by Tiberius, the reason given (namely to give Drusus a share in the glory) is considered by Germanicus fictitious (fingi: Ann. 2.26).

⁵³ Ogilvie, R. M., ‘An interim report on Tacitus' “Agricola”’, ANRW II 33.3, 1117–18,Google Scholar argues that Tacitus deliberately belittles the achievements of Agricola's predecessors in Britain so as to increase Agricola's stature.

⁵⁴ See Classen, C. J., ‘Tacitus — historian between Republic and Principate’, Mnemosyne 41 (1988), 93–116,CrossRefGoogle Scholar for the realism of Tacitus' vision; also for the argument that even the definition of what constitutes moderatio has changed since the time of Cicero.

⁵⁵ On the new political virtue of quietism and moderation, see Liebeschuetz, W., ‘The theme of liberty in the Agricola of Tacitus’, Classical Quarterly 16 (1966), 126–39CrossRefGoogle Scholar. It has, however, been argued by Luce, T. J., ‘Tacitus on “History's highest function”: praecipuum munus annalium (Ann. 3.65)’, ANRW II 33.4, 2904–27,Google Scholar that Tacitus is not, by praising characters in his works, necessarily setting them up as models for emulation in the manner of Livy. Tacitus is far more interested in commemoration than in paradigms.