242 When the Romans ask from the famous rhetor Proaeresios (end of the third and beginning of the fourth century) one of his disciples for a professorial chair, he sends to them Eusebius from Alexandria; “as respects rhetoric,” it is said of the latter (Eunapius, Proaer. p. 92 Boiss.), “it is enough to say that he was an Egyptian; for this people, no doubt, pursues versemaking passionately, but earnest oratory (ὁ σπουδαῖος Ἕρμης) is not at home among them.” The remarkable resumption of Greek poetry in Egypt, to which, e.g. the epic of Nonnus belongs, lies beyond the bounds of our narrative.

243 A “Homeric poet” ἐκ Μουσείου is ready to sing the praise of Memnon in four Homeric verses, without adding a word of his own (C. I. Gr. 4748). Hadrian makes an Alexandrian poet a member in reward for a loyal epigram (Athenaeus, xv. p. 677 e). Examples of rhetors from Hadrian’s time may be seen in Philostratus, Vit. Soph. i. 22, 3, c. 25, 3. A φιλόσοφος ἀπὸ Μουσείου in Halicarnassus (Bull. de corr. Hell. iv. 405). At a later period, when the circus was everything, we find a noted pugilist figuring (so to say) as an honorary member of the philosophical class (inscription from Rome, C. I. Gr. 5914: νεωκόρος τοῦ μεγά[λου Σαράπιδ]ος καὶ τῶν ἐν τῷ Μουσείῳ [σειτου]μένων ἀτελῶν φιλοσόφων; comp. ib. 4724, and Firmicus Maternus, de errore prof. rel. 13, 3). Οἱ ἐν Ἐφέσῳ ἀπὸ τοῦ Μουσείου ἰατροί (Wood, Ephesus, inscriptions from tombs, n. 7), a society of Ephesian physicians, have relation doubtless to the Museum at Alexandria, but were hardly members of it; they were rather trained in it.

244 Ὁ ἄνω θρόνος in Philostratus, Vit. Soph. ii. 10, 5.

245 Examples are Chaeremon, the teacher of Nero, previously installed in Alexandria (Suidas, Διονύσιος Ἀλεξανδρεύς; comp. Zeller, Hermes, xi. 430, and above, p. 259); Dionysius, son of Glaucus, at first in Alexandria, successor of Chaeremon, then from Nero down to Trajan librarian in Rome and imperial cabinet secretary (Suidas, l.c.); L. Julius Vestinus under Hadrian, who, even after the presidency of the Museum, filled the same positions as Dionysius in Rome (p. 248 note), known also as a philological author.

246 The eunuch of Candace, who reads in Isaiah (Acts of the Apostles, viii. 27) is well known; and a Candace reigned also in Nero’s time (Plinius, H. N. vi. 29, 182).

247 That the imperial frontier reached to Hiera Sycaminos, is evident for the second century from Ptolemaeus, v. 5, 74, for the time of Diocletian from the Itineraries, which carry the imperial roads thus far. In the Notitia dignitatum, a century later, the posts again do not reach beyond Syene, Philae, Elephantine. In the tract from Philae to Hiera Sycaminos, the Dodecaschoinos of Herodotus (ii. 29) temple-tribute appears to have been raised already in early times for the Isis of Philae always common to the Egyptians and Aethiopians; but Greek inscriptions from the Lagid period have not been found here, whereas numerous dated ones occur from the Roman period, the oldest from the time of Augustus (Pselchis, 2 A.D.; C. I. Gr. n. 5086), and of Tiberius (ib. 26 A.D., n. 5104, 33 A.D., n. 5101), the most recent from that of Philippus (Kardassi, 248 A.D., n. 5010). These do not prove absolutely that the place where the inscription was found belonged to the empire; but that of a land-measuring soldier of the year 33 (n. 5101), and that of a praesidium of the year 84 (Talmis, n. 5042 f.), as well as numerous others certainly presuppose it. Beyond the frontier indicated no similar stone has ever been found; for the remarkable inscription of the regina (C. I. L. iii. 83), found at Messaurât, to the south of Shendy (16° 25′ lat., 5 leagues to the south of the ruins of Naga), the most southern of all known Latin inscriptions, now in the Berlin Museum, has been set up, not by a Roman subject, but presumably by an envoy of an African queen, who was returning from Rome, and who spoke Latin perhaps only in order to show that he had been in Rome.

248 The tropaea Niliaca, sub quibus Aethiops et Indus intremuit, in an oration probably held in the year 296 (Paneg. v. 5), apply to such a rencontre, not to the Egyptian insurrection; and the oration of the year 289 speaks of attacks of the Blemyes (Paneg. iii. 17).—Procopius, Bell. Pers. i. 19, reports the cession of the “Twelve-mile-territory” to the Nubians. It is mentioned as standing under the dominion, not of the Nubians, but of the Blemyes by Olympiodorus, fr. 37, Müll. and the inscription of Silko, C. I. Gr. 5072. The fragment recently brought to light of a Greek heroic poem as to the victory of a late Roman emperor over the Blemyes is referred by Bücheler (Rhein. Mus. xxxix. 279 f.) to that of Marcianus, in the year 451 (comp. Priscus, fr. 27).

249 Juvenal (xi. 124) mentions the elephant’s teeth, quos mittit porta Syenes.

250 According to the mode in which Ptolemy (iv. 5, 14, 15) treats of this coast, it seems, just like the “Twelve-mile-land,” to have lain outside of the division into nomes.

251 Our best information as to the kingdom of Axomis is obtained from a stone erected by one of its kings, beyond doubt in the better period of the empire, at Adulis (C. I. Gr. 5127 b), a sort of writing commemorative of the deeds of this apparent empire-founder in the style of that of Darius at Persepolis, or that of Augustus at Ancyra, and fixed on the king’s throne, before which down to the sixth century criminals were executed. The skilful disquisition of Dillmann (Abh. der Berliner Akademie, 1877, p. 195 f.), explains as much of it as is explicable. From the Roman standpoint it is to be noted that the king does not name the Romans, but clearly has in view their imperial frontiers when he subdues the Tangaites μέχρι τῶν τῆς Αἰγύπτου ὁρίων, and constructs a road ἀπὸ τῶν τῆς ἐμῆς βασιλείας τόπων μέχρι Αἰγύπτου, and further, names as the northern limit of his Arabian expedition Leuce Come, the last Roman station on the Arabian west coast. Hence it follows further, that this inscription is more recent than the Periplus of the Red Sea written under Vespasian; for according to this (c. 5) the king of Axomis rules ἀπὸ τῶν Μοσχοφάγων μέχρι τῆς ἄλλης Βαρβαρίας, and this is to be understood exclusively, since he names in c. 2 the τύραννοι of the Moscophages, and likewise remarks in c. 14, that beyond the Straits of Bab el Mandeb there is no “king,” but only “tyrants.” Thus at that time the Axomitic kingdom did not reach to the Roman frontier, but only to somewhere about Ptolemais “of the chase,” just as in the other direction not to Cape Guardafui, but only as far as the Straits of Bab el Mandeb. Nor does the Periplus speak of possessions of the king of Axomis on the Arabian coast, although he on several occasions mentions the dynasts there.

252 The name of the Aethiopians was associated in the better period with the country on the Upper Nile, especially with the kingdoms of Meroe and Nabata (p. 275), and so with the region which we now call Nubia. In later antiquity, for example by Procopius, the designation is referred to the state of Axomis, and hence in more recent times is frequently employed for Abyssinia.

253 Hence the legend that the Axomites were Syrians settled by Alexander in Africa, and still spoke Syrian (Philostorgius, Hist. Eccl. iii. 6).

254 This is the praefectus praesidiorum et montis Beronices (C. I. L. ix. 3083), praefectus montis Berenicidis (Orelli, 3881), praefectus Bernicidis (C. I. L. x. 1129), an officer of equestrian rank, analogous to those adduced above (p. 249), as stationed in Alexandria.

255 The letter, which the emperor Constantius in the year 356 directs to Aeizanas, the king of the Axomites at that time, is that of one ruler to another on an equal footing; he requests his friendly and neighbourly assistance against the spread of the Athanasian heresy, and for the deposition and delivering up of an Axomitic clergyman suspected of it. The fellowship of culture comes here into the more definite prominence, as the Christian invokes against the Christian the arm of the heathen.

256 Inland lay the primeval Teimâ, the son of Ishmael of Genesis, enumerated by the Assyrian king Tiglath-Pilesar in the eighth century before Christ among his conquests, named by the prophet Jeremiah together with Sidon, around which gather in a remarkable way Assyrian, Egyptian, Arabian relations, the further unfolding of which, after bold travellers have opened up the place, we may await from Oriental research. In Teimâ itself Euting recently found Aramaic inscriptions of the oldest epoch (Nöldeke, Sitzungsberichte der Berliner Akademie, 1884, p. 813 f.) From the not far distant place Medâin-Sâlih (Hijr) proceed certain coins modelled after the Attic, which in part replace the owl of Pallas by that image of a god which the Egyptians designate as Besa the lord of Punt, i.e. of Arabia (Erman, Zeitschrift für Numismatik, ix. 296 f.) We have already mentioned the Nabataean inscriptions just found there (p. 148, note 3). Not far from thence, near ’Ola (el-Ally) inscriptions have been found, which correspond in the writing and in the names of gods and kings to those of the South-Arabian Minaeans, and show that these had a considerable station here, sixty days’ journey from their home, but on the frankincense-route mentioned by Eratosthenes, from Minaea to Aelana; and alongside of these others of a cognate but not identical South-Arabian stock (D. H. Müller in the Berichte der Wiener Akademie of 17th December 1884). The Minaean inscriptions belong beyond doubt to the pre-Roman period. As on the annexation of the Nabataean kingdom by Trajan these districts were abandoned (p. 152), from that time another south-Arabian tribe may have ruled there.

257 The accounts connected with the trade in frankincense in Theophrastus († 287 B.C.; Hist. plant. ix. 4) and more fully in Eratosthenes († 194 B.C.); in Strabo (xvi. 4, 2, p. 768) of the four great tribes of the Minaeans (Mamali Theophr.?) with the capital Carna; the Sabaeans (Saba Theophr.) with the capital Mariaba; the Cattabanes (Kitibaena Theophr.) with the capital Tamna; the Chatramotitae (Hadramyta Theophr.) with the capital Sabata, describe the very circle out of which the Homerite kingdom developed itself, and indicate its beginnings. The much sought for Minaei are now pointed out with certainty in Ma’in in the interior above Marib and Hadramaut, where hundreds of inscriptions have been found, and have yielded already no fewer than twenty-six kings’ names. Mariaba is even now named Marib. The region Chatramotitis or Chatramitis is Hadramaut.

258 The remarkable remains of this structure, executed with the greatest precision and skill, are described by Arnaud (Journal Asiatique, 7 série, tome 3, for the year 1874, p. 3 f. with plans; comp. Ritter, Erdkunde, xii. 861). On the two sides of the embankment, which has now almost wholly disappeared, stand respectively two stone structures built of square blocks, of conical almost cylindrical form, between which a narrow opening is found for the water flowing out of the basin; at least on the one side a canal lined with pebbles leads it to this outlet. It was once closed with planks placed one above another, which could be individually removed, to carry the water away as might be needed. The one of those stone cylinders bears the following inscription (according to the translation, not indeed quite certain in all its details, of D. H. Müller, Wiener Sitzungsberichte, vol. xcvii, 1880, p. 965): “Jata’amar the glorious, son of Samah’alî the sublime, prince of Saba, caused the Balap (mountain) to be pierced (and erected) the sluice-structure named Rahab for easier irrigation.” We have no secure basis for fixing the chronological place of this and numerous other royal names of the Sabaean inscriptions. The Assyrian king Sargon says in the Khorsabad inscription, after he has narrated the vanquishing of the king of Gaza, Hanno, in the year 716 B.C.: “I received the tribute of Pharaoh the king of Egypt, of Shamsiya the queen of Arabia, and of Ithamara the Sabaean; gold, herbs of the eastern land, slaves, horses, and camels” (Müller, l. c. p. 988; Duncker, Gesch. des Alterthums, ii.⁵ p. 327).

259 Sallet in the Berliner Zeitschrift für Numismatik, viii. 243; J. H. Mordtmann in the Wiener Numism. Zeitschrift, xii. 289.

260 Pliny, H. N. xii. 14, 65, reckons the cost of a camel’s load of frankincense by the land-route from the Arabian coast to Gaza at 688 denarii (= £30). “Along the whole tract fodder and water and shelter and various custom-dues have to be paid for; then the priests demand certain shares and the scribes of the kings; moreover the guards and the halberdiers and the body-guards and servants have their exactions; to which our imperial dues fall to be added.” In the case of the water-transport these intervening expenses were not incurred.

261 The chastising of the pirates is reported by Agatharchides in Diodorus, iii. 43, and Strabo, xvi. 4, 18, p. 777. But Ezion-Geber in Palestine, on the Elanitic gulf, ἣ νῦν Βερενίκη καλεῖται (Josephus, Arch. viii. 6, 4), was so called certainly not from an Egyptian princess (Droysen, Hellenismus, iii. 2, 349), but from the Jewess of Titus.

262 This (προσοικειοῦσθαι τούτους—τοὺς Ἄραβας—ἢ καταστρέφεσθαι: Strabo, xvi. 4, 22, p. 780; εἰ μὴ ὁ Συλλαῖος αὐτὸν—τὸν Γάλλον—προὐδίδου, κἂν κατεστρέψατο τὴν Εὐδαίμονα πᾶσαν: ib. xvii. 1, 53, p. 819) was the proper aim of the expedition, although also the hope of spoil, just at that time very welcome for the treasury, is expressly mentioned.

263 The account of Strabo (xvi. 4, 22 f., p. 780) as to the Arabian expedition of his “friend” Gallus (φίλος ἡμῖν καὶ ἑταῖρος, Strabo, ii. 5, 12, p. 118), in whose train he travelled in Egypt, is indeed trustworthy and honest, like all his accounts, but evidently accepted from this friend without any criticism. The battle in which 10,000 of the enemy and two Romans fell, and the total number of the fallen in this campaign, which is seven, are self-condemned; but not better is the attempt to devolve the want of success on the Nabataean vizier Syllaeos by means of a “treachery,” such as is familiar with defeated generals. Certainly the latter was so far fitted for a scapegoat, as he some years afterwards was on the instigation of Herod brought to trial before Augustus, condemned and executed (Josephus, Arch. xvi. 10); but although we possess the report of the agent who managed this matter for Herod in Rome, there is not a word to be found in it of this betrayal. That Syllaeos should have had the design of first destroying the Arabians by means of the Romans, and then of destroying the latter themselves, as Strabo “thinks,” is, looking to the position of the client-states of Rome, quite irrational. It might rather be thought that Syllaeos was averse to the expedition, because the commercial traffic through the Nabataean land might be injured by it. But to accuse the Arabian minister of treachery because the Roman transports were not fitted for navigating the Arabian coast, or because the Roman army was compelled to carry water with it on camels, to eat durra and dates instead of bread and flesh, and butter instead of oil; to bring forward the deceitfulness of the guidance as an excuse for the fact that 180 days were employed for the forward march over a distance overtaken on the return march in 60 days; and lastly, to criticise the quite correct remark of Syllaeos that a march by land from Arsinoe to Leuce Come was impracticable, by saying that a caravan route went thence to Petra, only shows what a Roman of rank was able to make a Greek man of letters believe.

264 The sharpest criticism of the campaign is furnished by the detailed account of the Egyptian merchant as to the state of the Arabian coast from Leuce Come (el-Haura to the north of Janbô, the port of Medina) to the Catacecaumene island (Jebel Taik near Lôhaia). "Different peoples inhabit it, who speak languages partly somewhat different, partly wholly so. The inhabitants of the coast live in pens like the ‘fish-eaters’ on the opposite coast" (these pens he describes, c. 2, as isolated and built into the clefts of the rocks), “those of the interior in villages and pastoral companies; they are ill-disposed men speaking two languages, who plunder the seafarers that drift out of their course and drag the shipwrecked into slavery. For that reason they are constantly hunted by the viceroys and chief kings of Arabia; they are called Kanraites (or Kassanites). In general navigation on all this coast is dangerous, the shore is without harbours and inaccessible, with a troublesome surf, rocky and in general very bad. Therefore, when we sail into these waters, we keep to the middle and hasten to get to the Arabian territory at the island Catacecaumene; from thence onward the inhabitants are hospitable, and we meet with numerous flocks of sheep and camels.” The same region between the Roman and the Homeritic frontiers, and the same state of things are in the view of the Axomite king, when he writes: πέραν δὲ τῆς ἐρυθρᾶς θαλάσσης οἰκοῦντας Ἀρραβίτας καὶ Κιναιδοκολπίτας (comp. Ptolemaeus vi. 7, 20), στράτευμα ναυτικὸν καὶ πεζικὸν διαπεμψάμενος καὶ ὑποτάξας αὐτῶν τοὺς βασιλέας, φόρους τῆς γῆς τελεῖν ἐκέλευσα καὶ ὁδεύεσθαι μετ' εἰρήνης καὶ πλέεσθαι, ἀπό τε Λευκῆς κώμης ἔως τῶν Σαβαίων χώρας ἐπολέμησα.

265 These walls, built of rubble, form a circle of a mile in diameter. They are described by Arnaud (l.c., comp. p. 287, note 1).

266 That the Oriental expedition of Gaius had Arabia as its goal, is stated expressly by Pliny (particularly H. N. xii. 14, 55, 56; comp, ii. 67, 168; vi. 27, 141, c. 28, 160; xxxii. 1, 10). That it was to set out from the mouth of the Euphrates, follows from the fact that the expedition to Armenia and the negotiations with the Parthians preceded it. For that reason the Collectanea of Juba as to the impending expedition were based upon the reports of the generals of Alexander as to their exploring of Arabia.

267 Our only information as to this remarkable expedition has been preserved to us by the Egyptian captain, who about the year 75 has described his voyage on the coasts of the Red Sea. He knows (c. 26) the Adane of later writers, the modern Aden, as a village on the coast (κώμη παραθαλάσσιος), which belongs to the realm of Charibael, king of the Homerites, but was earlier a flourishing town, and was so termed (εὐδαίμων δ' ἐπεκλήθη πρότερον οὖσα πόλις) because before the institution of the direct Indo-Egyptian traffic this place served as a mart: νῦν δὲ οὐ πρὸ πολλοῦ τῶν ἡμετέρων χρόνων Καῖσαρ αὐτὴν κατεστρέψατο. The last word can here only mean “destroy,” not, as more frequently, “subdue,” because the conversion of the town into a village is to be accounted for. For Καῖσαρ Schwanbeck (Rhein. Mus. neue Folge, vii. 353) has proposed Χαριβαήλ, C. Müller Ἰλασάρ (on account of Strabo, xvi. 4, 21, p. 782): neither is possible—not the latter, because this Arabian dynast ruled in a far remote district and could not possibly be presumed as well known; not the former, because Charibael was a contemporary of the writer, and there is here reported an incident which occurred before his time. We shall not take offence at the tradition, if we reflect what interest the Romans must have had in setting aside the Arabian mart between India and Egypt, and in bringing about direct intercourse. That the Roman accounts are silent as to this occurrence is in keeping with their habit; the expedition, which beyond doubt was executed by an Egyptian fleet and simply consisted in the destruction of a presumably defenceless place on the coast, would not be from a military point of view of any importance; about great commercial dealings the annalists gave themselves no concern, and generally the incidents in Egypt came still less than those in the other imperial provinces to the knowledge of the senate and therewith of the annalists. The naked designation Καῖσαρ, in which from the nature of the case the ruler then reigning is excluded, is probably to be explained from the circumstance that the reporting captain, while knowing doubtless the fact of the destruction by the Romans, knew not its date or author.—It is possible that to this the notice in Pliny (H. N. ii. 67, 168) is to be referred: maiorem (oceani) partem et orientis victoriae magni Alexandri lustravere usque in Arabicum sinum, in quo res gerente C. Caesare Aug. f. signa navium ex Hispaniensibus naufragiis feruntur agnita. Gaius did not reach Arabia (Plin. H. N. vi. 28, 160); but during the Armenian expedition a Roman squadron may very well have been conducted by one of his sub-commanders to this coast, in order to pave the way for the main expedition. That silence reigns elsewhere respecting it cannot surprise us. The Arabian expedition of Gaius had been so solemnly announced and then abandoned in so wretched a way, that loyal reporters had every reason to obliterate a fact which could not well be mentioned without also reporting the failure of the greater plan.

268 The Egyptian merchant distinguishes the ἔνθεσμος βασιλεύς of the Homerites (c. 23) sharply from the τύραννοι, the tribal chiefs sometimes subordinate to him, sometimes independent (c. 14), and as sharply distinguishes these organised conditions from the lawlessness of the inhabitants of the desert (c. 2). If Strabo and Tacitus had had eyes as open for these things as that practical man had, we should have known somewhat more of antiquity.

269 The war of Macrinus against the Arabes eudaemones (vita, 12) and their envoys sent to Aurelian (vita, 33), who are named along with those of the Axomites, would prove their continued independence at that time, if these statements could be depended on.

270 The king names himself, about the year 356 (p. 284, note 2), in a document (C. I. Gr. 5128) βασιλεὺς Ἀξωμιτῶν καὶ Ὀμηριτῶν καὶ τοῦ Ῥαειδὰν (castle in Sapphar, the capital of the Homerites; Dillmann, Abh. der Berl. Akad. 1878, p. 207) ... καὶ Σαβαειτῶν καὶ τοῦ Σιλεῆ (castle in Mariaba, the capital of the Sabaeans; Dillmann, l.c.). With this agrees the contemporary mission of envoys ad gentem Axumitarum et Homerita[rum] (C. Th. xii. 12, 2). As to the later state of things comp. especially Nonnosus (fr. hist. Gr. iv. p. 179, Müll.) and Procopius, Hist. Pers. i. 20.

271 Aristides (Or. xlviii. p. 485, Dind.) names Coptos the Indian and Arabian entrepôt. In the romance of Xenophon the Ephesian (iv. 1), the Syrian robbers resort to Coptos, “for there a number of merchants pass through, who are travelling to Aethiopia and India.”

272 Hadrian later constructed “the new Hadrian’s road” which led from his town Antinoopolis near Hermopolis, probably through the desert to Myos Hormos, and from Myos Hormos along the sea to Berenice, and provided it with cisterns, stations (σταθμοί), and forts (inscription in Revue Archéol. N. S. xxi. year 1870, p. 314). However there is no mention of this road subsequently, and it is a question whether it continued to subsist.

273 This is nowhere expressly said, but it is clearly evident from the Periplus of the Egyptian. He speaks at numerous places of the intercourse of the non-Roman Africa with Arabia (c. 7, 8), and conversely of the Arabians with the non-Roman Africa (c. 17, 21, 31; and after him Ptolemaeus, i. 17, 6), and with Persia (c. 27, 33), and India (c. 21, 27, 49); as also of that of the Persians with India (c. 36), as well as of the Indian merchantmen with the non-Roman Africa (c. 14, 31, 32), and with Persia (c. 36) and Arabia (c. 32). But there is not a word indicating that these foreign merchants came to Berenice, Myos Hormos, or Leuce Come; indeed, when he remarks with reference to the most important mart of all this circle of traffic, Muza, that these merchants sail with their own ships to the African coast outside of the Straits of Bab El Mandeb (for that is for him τὸ πέραν), and to India, Egypt cannot possibly be absent by accident.

274 In Bâmanghati (district Singhbhum) westward from Calcutta, a great treasure of gold coins of Roman emperors (Gordian and Constantine are named), is said to have come to light (Beglar, in Cunningham’s Archaeological Survey of India, vol. xiii. p. 72); but such an isolated find does not prove that regular intercourse extended so far. In Further India and China Roman coins have very seldom been found.

275 The designation Afer does not belong to this series. So far as we can follow it back in linguistic usage, it is never given to the Berber in contrast to other African stocks, but to every inhabitant of the Continent lying over against Sicily, and particularly also to the Phoenician; if it has designated a definite people at all, this can only have been that, with which the Romans here first and chiefly came into contact (comp. Suetonius, vita Terent.). Reasons philological and real oppose themselves to our attempt in i. 162i. 154 to trace back the word to the name of the Hebrews; a satisfactory etymology has not yet been found for it.

276 A good observer, Charles Tissot, (Géogr. de la province romaine de l’Afrique, i. p. 403) testifies that upwards of a third of the inhabitants of Morocco have fair or brown hair, and in the colony of the inhabitants of the Riff in Tangier two-thirds. The women made the impression on him of those of Berry and of Auvergne. Sur les hauts sommets de la chaîne atlantique, d’après les renseignements qui m’ont été fournis, la population tout entière serait remarquablement blonde. Elle aurait les yeux bleus, gris ou “verts, comme ceux des chats,” pour reproduire l’expression même dont s’est servi le cheikh qui me renseignait. The same phenomenon meets us in the mountain masses of Grand Kabylia and of the Aures, as well as on the Tunisian island Jerba and the Canary Islands. The Egyptian representations also show to us the Libu not red, like the Egyptians, but white, and with fair or brown hair.

277 Cyprian, Quod idola dii non sint, c. 2: Mauri manifeste reges suos colunt nec ullo velamento hoc nomen obtexunt. Tertullian, Apolog. 24: Mauretaniae (dei sunt) reguli sui. C. I. L. viii., 8834: Iemsali L. Percenius L. f. Stel. Rogatus v. (s. l. a.), found at Thubusuptu in the region of Sitifis, which place may well have belonged to the Numidian kingdom of Hiempsal. Thus the inscription also of Thubursicum (C. I. L. viii. n. 7* comp. Eph. epigr. v. p. 651, n. 1478) must have rather been badly copied than falsified. Still, in the year 70, it was alleged that in Mauretania a pretender to the throne had ascribed to himself the name of Juba (Tacitus, Hist. ii. 58).

278 This is attested for the year 70549 B.C. as regards both by Dio, xli. 42 (comp. Suetonius, Caes. 54). In the year 70747 B.C. Bogud lends assistance to the Caesarian governor of Spain (Bell. Alex. 59, 60), and repels an incursion of the younger Gnaeus Pompeius (Bell. Afric. 23). Bocchus, in combination with P. Sittius, in the African war makes a successful diversion against Juba and conquers even the important Cirta (Bell. Afr. 23; Appian, ii. 96; Dio, xliii. 3). The two obtained in return from Caesar the territory of the prince Massinissa (Appian, iv. 54). In the second Spanish war Bogud appears in the army of Caesar (Dio, xliii. 36, 38); the statement that the son of Bocchus had served in the Pompeian army (Dio, l.c.) must be a confusion, probably with Arabio the son of Massinissa, who certainly went to the sons of Pompeius (Appian, l.c.). After Caesar’s death Arabio possessed himself afresh of his dominion (Appian, l.c.), but after his death in the year 71440 B.C. (Dio, xlviii. 22) the Caesarian arrangement must have again taken effect in its full extent. The bestowal on Bocchus and Sittius is probably to be understood to the effect that, in the western part of the former Numidian kingdom otherwise left to Bocchus, the colony of Cirta to be founded by Sittius was to be regarded as an independent Roman town, like Tingi subsequently in the kingdom of Mauretania.

279 If, according to Dio, xl. 43, Caesar in the year 721 33 B.C. after the death of Bocchus, nominates no successor, but makes Mauretania a province, and then (li. 15) in the year 72430 B.C., on occasion of the end of the queen of Egypt, there is mention of the marriage of her daughter with Juba and his investiture with his father’s kingdom, and, lastly (liii. 26), under the year 72925 B.C. there is reported Juba’s investiture with a portion of Gaetulia instead of his hereditary kingdom, as well as with the kingdoms of Bocchus and Bogud; only the last account confirmed by Strabo, xvii. 3, 7, p. 828, is correct. The first is at least incorrect in its way of apprehending the matter, as Mauretania evidently was not made a province in 721 33 B.C., but only the investiture was held in abeyance for the time being; and the second partly anticipates, since Cleopatra, born before the triumph about 719 (Eph. epigr. i., p. 276), could not possibly be married in 724, and is partly mistaken, because Juba certainly never got back his paternal kingdom as such. If he had been king of Numidia before 729, and if it had been merely the extent of his kingdom that then underwent a change, he would have counted his years from the first installation and not merely from 729.

280 That Balbus carried on this campaign as proconsul of Africa, is shown in particular by the triumphal Fasti; but the consul L. Cornelius of the year 732 must have been another person, since Balbus, according to Velleius ii. 51, obtained that consular governorship, ex privato consularis, i.e. without having filled a curule office. The nomination, therefore, cannot have taken place according to the usual arrangement by lot. To all appearance he fell into disgrace with Augustus for good reasons on account of his Spanish quaestorship (Drumann ii. 609), and was then, after the lapse of more than twenty years, sent, as an extraordinary measure, to Africa, on account of his undoubted aptitude for this specially difficult task.

281 The tribes whom Tacitus names in his account of the war, far from clear, as always, in a geographical point of view, may be in some measure determined; and the position between the Leptitanian and the Cirtensian columns (Ann. iii. 74) points for the middle column to Theveste. The town of Thala (Ann. iii. 20) cannot possibly be sought above Ammaedara, but is probably the Thala of the Jugurthan war in the vicinity of Capsa. The last section of the war has its arena in western Mauretania about Auzia (iv. 25), and accordingly in Thubuscum (iv. 24) there lurks possibly Thubusuptu or Thubusuctu. The river Pagyda (Ann. iii. 20) is quite indefinable.

282 Ptolemaeus, iv. 3, 23, puts the Musulamii southward from the Aures, and it is only in accord therewith that they are called in Tacitus ii. 52, dwellers beside the steppe and neighbours of the Mauri; later they are settled to the north and west of Theveste (C. I. L. viii. 270, 10667). The Nattabutes dwelt according to Ptolemaeus l.c. southward of the Musulamii; subsequently we find them to the south of Calama (C. I. L. viii. 484). In like manner the Chellenses Numidae, between Lares and Althiburus (Eph. epigr. V. n. 639), and the conventus (civium Romanorum et) Numidarum qui Mascululae habitant (ib. n. 597), are probably Berber tribes transplanted from Numidia to the proconsular province.

283 In the year 70 the troops of the two Mauretanias amounted together, in addition to militia levied in large numbers, to 5 alae and 19 cohortes (Tacitus, Hist. ii. 58), and so, if we reckon on the average every fourth as a double troop, to about 15,000 men. The regular army of Numidia was weaker rather than stronger.

284 Inscription C. I. L. viii. 8369 of the year 129: Termini positi inter Igilgilitanos, in quorum finibus kastellum Victoriae positum est, et Zimiz(es), ut sciant Zimizes non plus in usum se habere ex auctoritate M. Vetti Latronis pro(curatoris) Aug(usti) qua(m) in circuitu a muro kast(elli) p(edes) D. The Zimises are placed by the Peutingerian map alongside of Igilgili to the westward.

285 If the praefect of a cohort doing garrison duty in Numidia held the command at the same time over six Gaetulian tribes (nationes, C. I. L. v. 5267), men that were natives of Mauretania were employed as irregulars in the neighbouring province. Irregular Mauretanian horsemen frequently occur, especially in the later imperial period. Lusius Quietus under Trajan, a Moor and leader of a Moorish troop (Dio lxviii. 32), no Λίβυς ἐκ τῆς ὑπηκόου Λιβύης, ἀλλ' ἐξ ἀδόξου καὶ ἀπῳκισμένης ἐσχατιᾶς (Themistius, Or. xvi p. 250 Dind.), was without doubt a Gaetulian sheikh, who served with his followers in the Roman army. That his home was formally independent of the empire, is not affirmed in the words of Themistius; the “subject-territory” is that with Roman organisation, the ἐσχατιά its border inhabited by dependent tribes.

286 To the inscriptions, which prove this (C. I. L. viii. p. xviii. 747), falls now to be added the remarkable dedication of the leader of an expeditionary column from the year 174, found in the neighbourhood of Géryville (Eph. epigr. v. n. 1043).

287 The tumultus Gaetulicus (C. I. L. viii. 6958) was rather an insurrection than an invasion.

288 Ptolemy certainly takes as boundary of the province of Caesarea the line above the Shott, and does not reckon Gaetulia as belonging to it; on the other hand he extends that of Tingis as far as the Great Atlas. Pliny v. 4, 30, numbers among the subject peoples of Africa “all Gaetulia as far as the Niger and the Ethiopian frontier,” which points nearly to Timbuctoo. The latter statement will accord with the official conception of the matter.

289 Already in Nero’s time Calpurnius (Egl. iv. 40) terms the shore of Baetica trucibus obnoxia Mauris.—If under Pius the Moors were beaten off and driven back as far as and over the Atlas (vita Pii, 5; Pausanias viii. 43), the sending of troops at that time from Spain to the Tingitana (C. I. L. iii. 5212–5215) makes it probable that this attack of the Moors affected Baetica, and the troops of the Tarraconensis marching against these followed them over the straits. The probably contemporary activity of the Syrian legion at the Aures (p. 320) suggests moreover that this war extended also to Numidia.—The war with the Moors under Marcus (vita Marci, 21, 22; vita Severi, 2), had its scene essentially in Baetica and Lusitania.—A governor of Hither Spain under Severus had to fight with the “rebels” by water and by land (C. I. L. ii. 4114).—Under Alexander (vita, 58) there was fighting in the province of Tingi, but without mention of Spain in the case.—From the time of Aurelian (vita Saturnini, 9) there is mention of Mauro-Spanish conflicts. We cannot exactly determine the time of a sending of troops from Numidia to Spain and against the Mazices (C. I. L. viii. 2786), where presumably not the Mazices of the Caesariensis but those of the Tingitana on the Riff (Ptolem. iv. 1, 10), are meant; perhaps with this is connected the fact that Gaius Vallius Maximianus, as governor of Tingitana, achieved in the province Baetica (according to Hirschfeld, Wiener Stud. vi. 123, under Marcus and Commodus) a victory over the Moors and relieved towns besieged by them (C. I. L. ii. 1120, 2015); these events prove at least that the conflicts with the Moors on the Riff and the associates that flocked to them from the country lying behind did not cease. When the Baquates on the same coast besieged the pretty remote Cartenna (Tenes) in the Caesariensis (C. I. L. viii. 9663), they perhaps came by sea. Where the wars with the Moors under Hadrian (vita, 5, 12) and Commodus (vita, 13) took place is not known.