¹ Mauny, R., ‘Tableau géographique de l'ouest africain au moyen âge’, Mém. I.F.A.N., no. 61 (Dakar, 1961), 283.Google Scholar

² The last reference to Libyan war-chariots is of 307 B.C. (Diodorus Siculus, XX. 64, 3); first to Libyan cavalry is of z6i B.C. (Polybius, 1. 19, 2–4).Google Scholar

³ Mauny, R., ‘Une route préhistorique à travers le Sahara occidental’, Bull. I.F.A.N. (1947), 341–57.Google Scholar

⁴ Lhote, H., ‘Le cheval et le chameau dans les gravures et les peintures rupestres du Sahara’, Bull. I.F.A.N. (1953), 1138–228,Google Scholar and ‘The Search for the Tassili Frescoes’, trans. Broderick, A. H. (London, 1959), 124–8.Google Scholar

⁵ Garamantes: Herodotus, IV. 183, 4; Pharusii: Strabo, XVII. 3, 7.Google Scholar

⁶ Especially his belief that the line of oases runs from Thebes. There was a route to Siwa from Thebes (via Kharga: cf. Herod. iii. 26), but it would take about twenty, not ten, days, and it would not follow the ‘ridge’: these data are applicable only to a route from the Nile Delta. (Pliny, NH, V. 50, gives twelve days for the journey from Memphis to Siwa.)Google Scholar

⁷ See, for the earlier part of Herodotus's route, Rhys Carpenter, ‘A trans-Saharan Caravan-route in Herodotus’, Am. J. Archaeol. (1956), 231–42.Google Scholar

⁸ For reference see Rawlinson, C., The History of Herodotus, II (London, 1875), 162, n. 9.Google Scholar

⁹ Ghnema, Djebel ben: Lhote, H. (1953), loc. cit. Latouma: P. Huard, ‘Gravures rupestres des confins nigéro-tchadiens’, Bull. I.F.A.N. (1953), 1573.Google Scholar

¹⁰ Hanno, Periplus, 7.Google Scholar

¹¹ Strabo, XVII. I, 2, etc.Google Scholar

¹² Though Pliny evidently thought he was: cf. V. 45, echoing Herodotus.Google Scholar

¹³ The text of Herod. IV. 184, I actually has ‘Aτλavtεs, but this is impossible, as it duplicates the next people. The reading ‘Arápavtεs is quoted by Eustathius (ad. Dionys. 66) as from Rhianos; Nicolas Damascenus (F 140) has the foim ‘Aφápavtεs.Google Scholar

¹⁴ Carpenter, loc. cit.Google Scholar

¹⁵ There is, in fact, an Adrar massif in Tassili.Google Scholar

¹⁶ πρoς έπέρoυς τóπoυς (Prom. Vinct. 348–50).Google Scholar

¹⁷ Herodotus's statement that the natives called Atlas κíoνa πov ovρaνov (IV. 184, 3) is clearly an echo of Aeschylus (loc. cit.), who speaks of Atlas bearing on his shoulders κíoν ovpaνov πε κaì χωaνóς.Google Scholar

¹⁸ V. 44: ‘quidam solitudinibus interposuerunt Atlantes’.Google Scholar

¹⁹ As assumed by How, W. W. and Wells, J., A Commentary on Herodotus, I (Oxford, 1912), 363.Google Scholar

²⁰ Salt-houses: Mauny (1961), 467; colours: e.g. Leo Africanus gives ‘candidi, rubei, cinericii colons’ (Rawlinson, II, 164, n. 7).Google Scholar

²¹ Mauny (1961), 332, who suggests, with reservations, Erebeb, north-west of Adrar des Iforas.Google Scholar

²² Herod. IV. 172, I, 182.Google Scholar

²³ Mauny (1961), 205–6.Google Scholar

²⁴ Cf. n. 7.Google Scholar

²⁵ Herod. IV. 43, 5.Google Scholar

²⁶ Mauny (1961), 453–4.Google Scholar

²⁷ His source for the journey of the Nasamones was some Cyrenaians, who had heard it from the king of the Ammonioi (II. 32, I).Google Scholar

²⁸ Notably religious: Amen was worshipped at Siwa (Herod. 11. 42) and Audjila (Procopius, Aed. VI. 2, 16–17);Google Scholar Bes and Osiris are apparently depicted in rock engravings in the Fezzan (McBurney, C. B. M., The Stone Age of Northern Africa (Penguin, 1960), 266);Google Scholar bird-headed goddesses in Egyptian style appear in the rock paintings of Tassili (Lhote, The Search for the Tassili Frescoes, p. 202).Google Scholar

²⁹ The Lotophagoi are immediately east of the Machlyes (IV. 178), who are on the shores of ‘Lake Tritonis’ (Gulf of Gabes): the Island of Djerba, opposite Gigthis, was identified with the Island of the Lotophagoi by Eratosthenes (Pliny, V. 41).Google Scholar

³⁰ I do not understand why Gsell, S., Histoire ancienne de l'Afrique du nord, IV (Paris, 1920), 138, assumes that the route in question is that via Bu Njem and Sokna.Google Scholar

³¹ ěθνoς μέλa ìσχυρως, Herod. IV. 283, I.Google Scholar

³² At least in Roman times: Tacitus, Ann. IV. 23; Pliny, VIII. 142; Ptolemy, I. 8, 4; Augustine, Lib. Qu. CXV. 23.Google Scholar

³³ Greek pottery of the seventh century is reported from there (Boardman, J., The Greeks overseas (Penguin, 1964), 222).Google Scholar

³⁴ Livy, XXXIV. 65, 3.Google Scholar

³⁵ E.g. Bovill, E. W., The Golden Trade of the Moors (Oxford, 1958), 22,Google Scholar

³⁶ Herod. IV. 189; cf. Polyb. III. 23, 2; Livy, XXIX. 25; XXXIV. 65.Google Scholar

³⁷ His sources for the passage on carbuncles are Callistratos of Rhodes (fl ca. 155 B.C.) and Archelaos (King of Cappadocia 36 B.C. to A.D. 17).Google Scholar

³⁸ Bovill (1958), 30, n. I.Google Scholar

³⁹ Ptolemy's Γίργiρi, the source-mountain of the Kinyps (IV. 6, 3) must be Garian.Google Scholar

⁴⁰ Several Negroid skulls have been found in Punic cemeteries (Gsell, IV, 174), presumably of slaves. Note also the Carthaginian nurse ‘corpore aquilo’ in Plautus, Poenulus, 1112–13.Google Scholar

⁴¹ There is, of course, no reason to suppose that he ‘crossed the desert’ in the sense of crossing into the Sudan.Google Scholar

⁴² Punica, III. 313–14 and passim.Google Scholar

⁴³ E.g. Gsell, IV, 139.Google Scholar

⁴⁴ ‘Cerné, I'île de Herné, et la question des navigations antiques sur la côte ouest - africaine’, Conférence internationale des africanistes de l'Ouest, IVe Conférence (Fernando Po, 1951), II, 73–80.Google Scholar

⁴⁵ Pliny, II. 169.Google Scholar

⁴⁶ Strabo, XVII. 3, 3.Google Scholar

⁴⁷ The monument of Tin Hinan at Abalessa did produce several gold objects (for ref. cf. n. 94).Google Scholar

⁴⁸ Mauny, ‘Monnaies anciennes trouvées en Afrique au sud du limes de l'empire romain’, C.I.A.O. no. IV (1951), vol. II, pp. 51–70.Google Scholar

⁴⁹ The precise location of the Pharusii is uncertain: Strabo (XVII. 3, 3) says they are thirty days' journey from Lixus; Pliny (v. 43) has them separated by ‘deserta’ from the Gaetuli in the north, and by ‘vastae solitudines’ from the Garamantes in the cast.Google Scholar

⁵⁰ Mauny (1961), 300.Google Scholar

⁵¹ Trimingham, J. S., A History of Islam in West Africa (Cambridge, 1962), 48.Google Scholar

⁵² Polybius, XXX. 2.Google Scholar

⁵³ Sallust, lug. 77, 2.Google Scholar

⁵⁴ See especially Desanges, P., ‘Le triomphe de Cornelius Balbus’, Revue africaine (1957), 5–43.Google Scholar

⁵⁵ The nature of Rome's desert wars is well summarized by Orosius's account of the Gaetulian war of A.D. 6 (vi. 21, 18): ‘Gaetulos latius vagantes Cossus…artatis finibus coercuit atque a Romanis limitibus metu compulit’.Google Scholar

⁵⁶ ‘L'expédition de Cornelius Balbus au Sahara en 19 av. J.-C. d'après le texte de Pline’, Rev. Afr. (1954) 45–83: answered by Desanges, bc. cit.Google Scholar

⁵⁷ The variants are ‘Balsa’ and ‘Balla’: as the preceding name is ‘Galsa’, a corruption from ‘Balls’ to ‘Balsa’ is more likely.Google Scholar

⁵⁸ An inscription from Cyrene (S.E.G. IX. 63) celebrates the end of a Maρμaρiκός πόλεпoς in A.D. 2: but this date raises difficulties for the chronology of Quirinius's career, and Syme, R. (The Roman Revolution (Oxford, 1939), 399, n. I) conjectures ca. 15 B.C.Google Scholar

⁵⁹ Florus, II. 31.Google Scholar

⁶⁰ Tac. Ann. III. 74.Google Scholar

⁶¹ Ibid. IV. 23.

⁶² Ibid. IV. 26.

⁶³ Tac. Hist. IV. 50.Google Scholar

⁶⁴ Pliny, V. 38.Google Scholar

⁶⁵ Tac. Hist. IV. 50.Google Scholar

⁶⁶ Pliny's main source was evidently the campaign of Balbus (cf. v. 36).Google Scholar

⁶⁷ For example, Bovill (1958), 34.Google Scholar

⁶⁸ Ptolemy also has a Φaμavίa somewhere in Africa (IV, 7, 10), but is completely vague about its location.Google Scholar

⁶⁹ Pliny, V. 38.Google Scholar

⁷⁰ Zonaras, XI. 19.Google Scholar

⁷¹ Victor, Caes. 20, 79; S.H.A. Severus, 18, 3.Google Scholar

⁷² Goodchild, R. G. and Perkins, J. B. Ward, ‘The “Limes Tripolitanus” in the light of recent discoveries’, Journal of Roman Studies (1949), 81–95.Google Scholar

⁷³ Goodchild, ‘Oasis forts of Legio III Augusta on the routes to the Fezzan’, Papers of the British School at Rome 1954, 56–68.Google Scholar

⁷⁴ Mauny, ‘Monnaies anciennes…’, loc. cit. There are also civilian gravestones at Ghadames, with inscriptions in both Latin and Libyan (J. M. Reynolds ‘Three inscriptions from Ghadames in Libya’, P.B.S.R. (1958), 135–6).Google Scholar

⁷⁵ Procop. Aed. VI. 3, II.Google Scholar

⁷⁶ Goodchild, , Roman Roads and Milestones in Tripolitania (Tripoli, 1948), 54–20, 29.Google Scholar

⁷⁷ A find of ancient coins is reported at Waddan in the Oasis of Djofra (Mauny, ‘Monnaies anciennes…’, loc. cit.).Google Scholar

⁷⁸ Pace, Sergi and Caputo, ‘Scavi Sahariani’, Monumenti Antichi 1951, coll. 151–552;Google ScholarSir Wheeler, M., Rome Beyond the Imperial Frontiers (London,1954), 97–107.Google Scholar

⁷⁹ Loc. cit. 105–6.Google Scholar

⁸⁰ Ibid. 178.

⁸¹ From information given by Miss O. Brogan.Google Scholar

⁸² Gsell, I (1913), p. 60, n. 8, suggests that the Seven also introduced the camel for desert travel in Tripolitania: camels were certainly known earlier, but they did not become numerous till the third century (for a different interpretation of the evidence, see O. Brogan, ‘The camel in Roman Tripolitania’, P.B.S.R. (1954) 126–31, and Bovill (1958), 39–43).Google Scholar

⁸³ Monumenti Antichi 1951, coll. 313–14 (and figs. 106–7).Google Scholar

⁸⁴ For which cf. Strabo, XVII. 3, 18 (Zouchis), Pliny IX. 127 (Djerba Is.).Google Scholar

⁸⁵ Aurigemma, S., ‘L'elefante di Leptis Magna e il commercio dell'avorio e delle ‘Libycae ferae’ negli Emporia Tripolitani’, Africa Italiana (1940), 67–86.Google Scholar

⁸⁶ An elephant was brought across the Sahara to Marrakech at the end of the sixteenth century (Bovill (1958), 184).Google Scholar

⁸⁷ For references see Gsell, I, 302, n. 2.Google Scholar

⁸⁸ Anthologia Latina (ed. Bücheler and Riese), no. 183.Google Scholar

⁸⁹ The excrement of the Garamantes has come into our world, and the black slave rejoices in his pitch-coloured body. If the voice emitted from his lips did not prove him human, people would be frightened at the sight of this terrible ghost. May the Underworld, O Hadrumetum, carry off your monster for itself! Pluto's house should have him as its guardian.Google Scholar

⁹⁰ For example, Bovill (1958), 40, identifies Ptolemy's Báρδητoν and Mέοχη with Bardai and Miski in Tibesti.Google Scholar

⁹¹ Mauny (1961), 265.Google Scholar

⁹² Ibid. 435.

⁹³ Mauny, ‘Monnaies anciennes…’ loc. cit.Google Scholar

⁹⁴ Reygasse, M., Monuments funéraires préislamiues de l'Afrique du nord (Paris, 1950), 88–117;Google ScholarWheeler (1954), 107–11.Google Scholar

⁹⁵ Mauny, ‘Monnaies anciennes…’, loc. cit.Google Scholar

⁹⁶ In the forest, a coin of Constantine has been found at Buea in Cameroun (Bovill (1958), 41, n.).Google Scholar

⁹⁷ Unless they are credited with the introduction of iron-working into West Africa: cf. Mauny, ‘Essai sur l'histoire des métaux en Afrique occidentale’, B.I.F.A.N. (1952), 545–95, and against this Huard, ‘Contribution à l'étude du cheval, du fer, et du chameau au Sahara occidental. I. Le fer’, Bull. I.F.A.N. (1960), 134–78.Google Scholar

⁹⁸ Procopius, Bell. Vand. I. 8, cf. II. 21.Google Scholar

⁹⁹ Aed. VI. 2, 18–20.Google Scholar

¹⁰⁰ Ibid. VI. 3, 9–10.

¹⁰¹ Trimingham (1962), 15, n. I.Google Scholar

¹⁰² Trimingham, ibid.; Bovill (1958), 51–2.Google Scholar

¹⁰³ Especially in the cemetery of Tag'lit (Monumenti Antichi 1951, coll. 381–4).Google Scholar

¹⁰⁴ Ibid.coll. 416–19.

¹⁰⁵ Cf. n. 91.Google Scholar