1. Dilke, O.A.W., Greek and Roman Maps (London 1998), 62Google Scholar.

2. Dilke (n.1 above), 65.

3. Goodyear, F.R.D., ‘Technical Writing’, in CHCL II (Cambridge 1982), 667Google Scholar, lambasts Mela as derivative, inaccurate and the author of ‘no systematic and professional treatise’.

4. ‘La Chorographie et le déclin de la géographie sous l’Empire’: Silberman, Alain (ed.), Pomponius Mela: Chorographie (Paris 1988), xxvGoogle Scholar.

5. Silberman (n.4 above), xxv-xxvi. Similarly, Rawson, Elizabeth, Intellectual Life in the Late Roman Republic (Baltimore 1985), 252Google Scholar, claims that, after Augustus, ‘descriptive geographers tended to be mere compilers.’

6. E.g. Romm, James, The Edges of the Earth in Ancient Thought (Princeton 1992), 150f.Google Scholar, credits Mela with preferring a ‘bizarre…explanation’ for the Nile’s origin, while ‘the sober-minded Ptolemy…came closest to the truth.’ Long before, Wissowa, Georg, ‘Die Abfassungzeit der Chorographia des Pomponius Mela’, Hermes 51 (1916), 89Google Scholar, had criticised Mela as fundamentally superficial and lacking in the scientific depth of Greek geographers, producing instead an 'inhaltlich recht oberflachliche geographische Handbuch'. Although I consider Strabo to be an Augustan and a Roman writer in terms of cultural outlook, he does also locate himself within the large-scale theoretical debates on the extent and scale of the cosmos (particularly in Books 1 and 2). For a consideration of Strabo’s status and identity see Clarke, Katherine, ‘In Search of the Author of Strabo’s Geography’, JRS 84 (1997), 92–110Google Scholar.

7. This is the interpretation given by Servius ad Aen. 6.848.

8. Caes. BG 1.1, 2.4, 4.1–4, 4.10, 5.12–14, 6.11–28; Sail. Jug. 17.6–19.8, 78.1–79.9, 80.7–8. Similarly Tac. Ag. 10–17.

9. Goodyear (n.3 above), 668.

10. Stahl, William H., Roman Science: Origins, Development and Influence to the Later Middle Ages (Madison 1962), 69Google Scholar.

11. Strabo 1.1.16: (‘the greater part of geography is directed towards political use’). See also 1.1.1; 1.1.14; 1.1.18.

12. Beagon, Mary, Roman Nature: The Thought of Pliny the Elder (Oxford 1992), 7f.Google Scholar, comments that numerous technical handbooks were being written in the early Empire by equestrians, and extrapolates that they were largely intended for an equestrian audience. Pliny’s biography and place in public life (largely from the younger Pliny’s evidence) and his vast, comprehensive knowledge and interests are more well-known. Mela exists in something of a void—but Pliny’s frequent use of his text, and reference to him as a geographical authority (HN 1.3, 1.4, 1.5, 1.8, 1.12, 1.13, 1.21) gives the impression that Mela was part of an equestrian ‘in-group’. Certainly, Romans from Spain were achieving political importance at Rome in this era, as the two Senecas demonstrate (for Spaniards in Rome, see Griffin, Miriam, Seneca, A Philosopher in Politics [Oxford 1976], 249–55Google Scholar).

13. Stahl (n.10 above), 69. Also Rawson (n.5 above), 266: ‘The weaknesses of ancient geography and ethnography are clear….The desire for paradoxa bedeviled serious enquiry.’

14. Silberman (n.4 above), xi; discussed by Dilke, O.A.W., CR 40 (1990), 286Google Scholar.

15. Strabo 3.2.1 (begins with Baetica), 17.3.1–23 (ends with Libya); Pliny HN 3.3 origo ab occasu solis et Gaditano freto (‘the starting point is in the west, at Straits of Gibraltar’); 3.6 prima Hispania (‘first Spain’).

16. 2.5.26: a lengthy panegyric of Europe as the world’s nurturer and leader. Also, Pliny HN 3.5: primum ergo de Europa altrice uictoris omnium gentium populi longeque terrarum pulcherrima… (‘So, to begin from Europe, the nurse of the race which has conquered all peoples, and by far the most beautful of lands…’).

17. Mela (3.90f.) and Pliny (HN 2.169) cite Hanno’s journey as proof that Africa is circumnavigable, whereas Hanno sailed only part way down the west coast before turning back. Romm (n.6 above), 122, discusses possible reasons for this discrepancy. Pliny (HN 6.201) also finds Africa to be ungraspable: on its west coast omnia circa incerta sunt…nec Mauretania insularum certior fama est (‘all is in doubt…nor is there a more definite story with regard to the islands of Mauretania’); on the east side (6.173) in ora Aethiopiae sinus incognitus (‘there is an unexplored bay on the Ethiopian coastline’). It was not until Nero’s expedition down the Nile that disputes over distances in Ethiopia were resolved (6.183–4)—at the time of the Chorographia’s composition even this remained unexplored by Rome. Mela finds the Nile geographically confusing, as it is difficult to navigate and its name fluctuates: desertis Africae missus, nee statim nauigari facilis nee statim Nilus est (‘flowing from the deserts of Africa, it is not at the beginning easy to navigate, nor is it [called] the Nile at the beginning’); and the river divides into two around Meroe (1.50): et cum diu simplex saeuusque descendit, circa Meroen late patentem insulam…diffunditur, alteraque parte Astabores altera Astape dictus est (‘and when it flows down at last as one violent stream, it divides around Meroe, a widely-spreading island…and one stream is called Astabores, the other is called Astape’).

18. The existence of two distinct groups of Ethiopians had a long tradition, as far back as Homer (Od. 1.22–25). Situating them at opposite ends of Africa was one way of accommodating their divided location (see Silberman [n.4 above], ad loc).

19. Africa: 17.3.1–23 (the closing section of the entire work) of which 17.3.2–22 covers the north coast, 17.3.23 the whole of the interior and the other two coasts (Strabo perceives Africa as a triangle, whereas Mela sees it as rectangular). A great deal of space is devoted to Carthage and its relations with Rome (17.3.14–15), as is consistent with Strabo’s emphasis on mapping Rome’s history on to the land.

20. The contrast between the north, which houses cities and possesses a recorded history, and the remainder of the continent, which is mountainous and desert, and has only the occasional settlement, is established earlier (17.3.9).

21. This involved, for example, the accession of Juba II, a Roman citizen brought up in Rome and imposed as monarch of Mauretania by Augustus in 25 BCE. Juba rebuilt the capital, Iol, as a showpiece Roman city, and renamed it Caesarea. For a discussion of the extent to which north Africa was Romanised, see Raven, Susan, Rome in Africa (London 1993), esp. 106–27 and 152–56Google Scholar; and Broughton, T.R.S., The Romanization of Africa Proconsularis (New York 1968Google Scholar).

22. BG 5.12–14.

23. It is not certain whether Mela refers to the emperor Gaius or Claudius here, which is obviously relevant to fixing the precise date at which Mela was writing; see Parroni, P., Pompeio Mela: De chorographia libri tres (Rome 1984), 16–22Google Scholar, for a discussion of the date of composition. Most scholars prefer 43 CE under Claudius, as the circumstances surrounding Caligula’s invasion are hazy. Brodersen, Kai, Pomponius Mela: Kreuzfahrt durch die alte Welt (Darmstadt 1994), 1Google Scholar, and Romer, F.E., Pomponius Mela’s Description of the World (Ann Arbor 1998), 3CrossRefGoogle Scholar, argue for a pun on Claudius’ name in the description of Britain as tamdiu clausam.

24. Romm (n.6 above), 123.

25. The Irish as cannibals: Diodorus Siculus 5.32.2; Strabo 4.5.4 (he also claims they practise incest); although at 2.1.13 and 2.1.17, Ireland is uninhabitable because of the cold.

26. The form uirtutium is rare, but also found in Mela at 2.45, and Sen. Dial. 3.6.3.

27. For the savagery of the Scythians, see Book 4, passim, particularly 4.59–75. Elsewhere Herodotus tells of related, but specialised, dietary customs, such as the geriatric sacrifice and cannibalism of the nearby Massagetae (1.216), the funerary cannibalism of the Scythian Issedones (4.26) and the Indian Padaei (3.99), a nomadic tribe, who live on raw meat, and cannibalise the sick early on in illness before they waste away and the meat spoils.

28. Strabo 7.3.6: they sacrifice and eat the flesh of strangers; Pliny HN 7.9 esse Scytharum genera, et quidem plura, quae corporibus humanis uescerentur indicauimus (‘we have already pointed out that there are Scythian tribes, indeed many of them, who feed themselves on human bodies’). Elsewhere (4.81), Pliny locates the Scythians in remote and unknown zones: extremi gentium…ignoti prope ceteris mortalibus (‘the most remote of peoples…almost unknown to other humans’).

29. Arens, William, The Man-Eating Myth (Oxford 1979), 69Google Scholar.

30. Arens (n.29 above), passim; see especially 167f., for the provocative claim that anthropology needs cannibalism to survive because it ensures that the field always has ‘bizarre customs’ to interpret. Arens’s view is controversial, and has been criticised by certain archaeologists and physical anthropologists, e.g. Brown, Paula and Tuzin, Donald, The Ethnography of Cannibalism (Washington DC 1983Google Scholar); see also Lindenbaum, Shirley, Kuru Sorcery (Mayfield 1979Google Scholar), concerning cannibalism in Papua New Guinea. Recent work continues the debate, with biological anthropologist Christy Turner becoming a media sensation by claiming that ritual cannibalism was practised by the Anasazi in New Mexico (see Turner, Christy and Turner, Jacqueline, Man Corn: Cannibalism and Violence in the Prehistoric American South-West [Salt Lake City 1999]Google Scholar), while cultural historians Peter Hulme and Gananath Obeyesekere, and Arens himself, continue to question the assumptions and discourses which underlie accusations of cannibalism (see Barker, Francis, Hulme, Peter and Iversen, Margaret, Cannibalism and the Colonial World [Cambridge 1998]Google Scholar). But it should be noted that Arens does not argue that cannibalism has never existed—rather that anthropologists and archaeologists have been overeager to attribute anthropophagy on the basis of scanty evidence and to the exclusion of other interpretations of what evidence does exist.

31. Arens (n.29 above), 64–79, discusses the example of Fray Diego Durán and Fray Bernado de Sahagún, who wrote sixteenth-century historical and ethnographic accounts of the Aztecs, and solidified the idea that they were cannibals; Arens believes that the Aztecs in fact practised human sacrifice, and that this was conflated with cannibalism by the missionaries. See also Marina Warner’s eloquent’ demonstration of how ethnographic and literary representations of cannibals enact stratagems which designate difference as barbarism: Six Myths of Our Time: Little Angels, Little Monsters, Beautiful Beasts and More (London 1994), 65–79Google Scholar.

32. Augustus sent Aelius Gallus (prefect of Egypt) to explore Ethiopia, Arabia Felix and other places in 25 BCE (Strabo 16.4.22–24). For Nero’s expedition see Pliny HN 6.181, 184; Seneca NQ 6.8.3f.; see Romm (n.6 above), 152–56, for discussion of Roman exploration in Egypt, particularly of Lucan, who attributes to Julius Caesar the ambivalent desire to discover the source of the Nile (BC 10.191f.).

33. Egypt (‘has more wonders than all other lands’), and Egyptian customs oppose most others. Herodotus’ account focuses on religious differences, the reversal of gender roles, personal appearance, mourning customs, spatial distinctions, diet, clothing and hygiene. More generally, Egypt has its own brand of climate and geography, which, it is implied, are causes of its otherness. Cf. also 2.64, 84–6.

34. Although the passage continues (nisi…) by explaining all the ways in which the coastal Africans are not like the Romans (language, religion, diet, clothing—the standard ways that Roman ethnographers differentiate races); presumably the difference is relative, as the more extraordinary peoples follow. The coastal cities were often Greek in origin (see Silberman [n.4 above], ad loc.) and therefore more likely to be perceived as familiar to Roman writers.

35. See Romm (n.6 above), 79, on language as ‘a defining human attribute’ for Greco-Roman antiquity; also Dierauer, U., Tier und Mensch im Denken der Antike (Amsterdam 1977), 33fGoogle Scholar.

36. Romer (n.23 above), 48.

37. duBois, Page, Centaurs and Amazons: Women and the Pre-History of the Great Chain of Being (Ann Arbor 1991), 31CrossRefGoogle Scholar.

38. For a reassessment of Lucian, see Bracht Branham, R., Unruly Eloquence (Cambridge MA 1989CrossRefGoogle Scholar).

39. OLD s.v. effigies 3a.

40. The Trogodytes are located near to the satyrs in Pliny’s text, appearing just before the quoted passage (HN 7.23), but without the attribution of unintelligible screeching.

41. Mela concentrates on Indian dietary and mortuary habits, basing much of his description on Herodotus (3.97–104). ‘Indography’ (particularly Greek writing) is discussed by Romm (n.6 above), 94–109, where he also cites Strabo’s rejection of Hellenistic India-myths (2.1.9, 15.1.37) and a consistent later tradition of wonders attributed to India, which persisted until the Renaissance. Africa and India were physically connected by some Greek geographers: Polybius and Ptolemy actually mapped them as one continent, though the existence of a sea between them was well known (see Romm [n.6 above], 82).

42. Strabo and Pliny concentrate on the limited diet of nomadic peoples (Strabo 3.5.11, 7.4.6; Pliny HN 6.160,6.189, 6.190), a practice which goes back to Homer (11. 13.5f.); like Sallust, both clearly view this lifestyle as primitive, as it is either a practice long past (Jug. 18–19, Strabo 2.5.33, Pliny HN 5.22, of the Numidians) or a surprising vestige of an obsolete mode (Strabo 7.3.7, ; of the Scythians, who are also cannibals; Pliny HN 6.143, et ipsi uagi, of the Scenitae of Arabia). Note that for Strabo (7.4.6) the Black Sea Chersonese nomads can also represent a warlike but just people, who counter decadence and luxury, indicating that nomadism can inhabit the ambivalent space which incorporates the primitive and the virtuous.

43. Russo, Mary, The Female Grotesque: Risk, Excess and Modernity (New York 1994), 3Google Scholar (my emphasis).

44. Precise dates for wall-painting styles are subject to debate and difficult to standardise, but Third Style is usually said to run approximately from 15 BCE to 40 CE or later; see Ling, Roger, Roman Painting (Cambridge 1991), 52ffGoogle Scholar.

45. The word ‘grotesque’ is derived from ‘grotto-esque’, as used in sixteenth-century writings on the architectural and painted designs discovered in Nero’s Domus Aurea, which did indeed include fantastical images combining plant, animal and human forms. See Russo (n.43 above), 3, and Barton, Carlin, The Sorrows of the Ancient Romans (Princeton 1993), 87Google Scholar.

46. Barton (n.45 above), 86.

47. Book 7 of Pliny’s Historia Naturalis begins with a section on humankind (principium iure tribuetur homini, 7.1), but almost immediately falls to describing abnormalities (cannibals, 7.9f., 7.12; deformed peoples, 7.1 If.; androgynous peoples, 7.16; dwarves and giants, 7.74–76). Many of these monstra are found in Africa (7.14, 7.16, 7.31).

48. Russo (n.43 above), 5

49. uastus is again used of Africa; compare the uastitas of the continent (3.89), discussed above (p.58).

50. Perhaps it is also implied that there is a military threat from these hybrid beings: they inhabit the campi—which implies ‘field of battle’.

51. Bakhtin, Mikhail, Rabelais and his World, tr. Iswolsky, Helene (Bloomington 1984), 27Google Scholar.

52. Silberman (n.4 above), 317, considers that the variations are due to the use of different sources, claiming that Mela uses Nepos, whereas Pliny uses an unknown source.

53. E.g. Parroni (n.23 above), ad loc.

54. Ethnographically the power to effect injury through one’s own body is related to gender and to dietary perversions (also sites of disquiet): Pliny reports Cicero’s claim that feminas quidem omnes ubique uisu nocere quae duplices pupillas habeant (‘all women who have double pupils do harm by their glance’, HN 7.18); and he connects cannibals’ bodies to those which are shot through with poison: adeo naturae, cum ferarum morem uescendi humanis uisceribus in homine genuisset, gignere etiam in toto corpore et in quorundam oculis quoque uenena placuit (‘when nature implanted in humans the wild beasts’ habit of devouring human flesh, she also thought fit to implant poisons in the whole of the body, and with some poisons in the eyes as well’, ibid.). Pliny’s examples of the evil eye and related sorceries (HN 7.16–18) are concentrated in Africa and the far north-east (Scythia, on which see below); these forces are brought together by the fact that that the all-female ‘Gorillas’ live on the island once inhabited by the terrifying Gorgons. See Barton (n.45 above), 93–95, for a discussion of the aggressive evil eye in antiquity.

55. Freud, Sigmund, ‘Medusa’s Head’, in Collected Papers, vol. 5, tr. Strachey, James (London 1950), 106Google Scholar.

56. Theweleit, Klaus, Male Fantasies, Volume 1 (Minneapolis 1987), 201Google Scholar.

57. In the eastern part of the continent, the Nile and the land on its banks also form a space which represents mysterious and excessive femininity, with its supernaturally fertile banks, whose mud spontaneously produces life: terra expers imbrium, mire tamen fertilis et hominum aliorumgue animalium perfecunda generatrix. Nilus efficit, amnium in Nostrum mare permeantium maximus…glaebis etiam infundat animas, ex ipsaque humo uitalia effingat (‘the land lacks rainfall, however it is miraculously fertile and an extremely productive generator of humans and other living creatures. The Nile brings it about, the greatest of the rivers which pass into Our Sea…it even pours breath into clods of earth, and creates living things out of the very soil’, 1.49–52).

58. Richlin, Amy, ‘Pliny’s Brassiere’, in Hallett, Judith P. and Skinner, Marilyn B. (eds.), Roman Sexualities (Princeton 1997), 202–05Google Scholar.

59. Other strange medical phenomena are attached to Africa by Pliny: e.g. the Psylli tribe, whose bodies contain a substance which is poisonous to snakes (HN 7.14).

60. Barton (n.45 above), 101.

61. The best evidence we have for this map is the elder Pliny’s account (HN 3.16f., 6.139). See Nicolet, Claude, Space, Geography and Politics in the Early Roman Empire (Ann Arbor 1991), 95–122CrossRefGoogle Scholar.