290 More information than in the scanty accounts of Victor and Eutropius is supplied as to this war by the inscribed stones, C. I. L. viii. 2615, 8836, 9045, 9047. According to these the Quinquegentiani may be followed out from Gallienus to Diocletian. The beginning is made by the Baquates who, designated as Transtagnenses, must have dwelt beyond the Shott. Four “kings” combine for an expedition. The most dreaded opponent is Faraxen with his gentiles Fraxinenses. Towns like Mileu in Numidia not far from Cirta and Auzia in the Caesariensis are attacked, and the citizens must in good part defend themselves against the enemy. After the end of the war Maximian constructs great magazines in Thubusuctu not far from Saldae. These fragmentary accounts give in some measure an insight into the relations of the time.

291 Apart from the coins this is proved also by the inscriptions. According to the comparison, for which I am indebted to Herr Euting, the great mass of the old Punic inscriptions, that is, those written probably before the destruction of Carthage, falls to Carthage itself (about 2500), the rest to Hadrumetum (9), Thugga (the famous Phoenico-Berber one), Cirta (5), Iol-Caesarea (1). The new Punic occur most numerously in and around Carthage (30), and generally they are found not unfrequently in the proconsular province, also in Great Leptis (5) and on the islands of Girba (1) and Cossura (1); in Numidia, in and near Calama (23), and in Cirta (15); in Mauretania hitherto only in Portus Magnus (2).

292 The coining in Africa ceases in the main after Tiberius, and thereafter, since African inscriptions from the first century after Christ are before us only in very small numbers, for a considerable period documents fail us. The coins of Babba in the Tingitana, going from Claudius down to Galba, have exclusively Latin legends; but the town was a colony. The Latin-Punic inscriptions of Great Leptis, C. I. L. viii. 7, and of Naraggara, C. I. L. viii. 4636, may doubtless belong to the time after Tiberius, but as bilingual tell rather for the view that, when they were set up, the Phoenician language was already degraded.

293 From the expression in the epitome of Victor, that the emperor Severus was Latinis litteris sufficienter instructus, Graecis sermonibus eruditus, Punica eloquentia promptior, quippe genitus apud Leptim, we may not infer a Punic course of rhetoric in the Tripolis of that time; the late and inferior author has possibly given a scholastic version of the well-known notice.

294 On the statement of the younger Arnobius, writing about 460 (ad Psalm. 104, p. 481 Migne: Cham vero secundus filius Noe a Rhinocoruris usque Gadira habens linguas sermone Punico a parte Garamantum, Latino a parte boreae, barbarico a parte meridiani, Aethiopum et Aegyptiorum ac barbaris interioribus vario sermone numero viginti duabus linguis in patriis trecentis nonaginta et quattuor), no reliance is to be placed, still less upon the nonsense of Procopius, de bello Vand. ii. 10, as to the Phoenician inscription and language in Tigisis. Authorities of this sort were hardly able to distinguish Berber and Punic.

295 In a single place on the Little Syrtis the Phoenician may still have been spoken in the eleventh century (Movers, Phön. ii. 2, 478).

296 More clearly than by the Latin inscriptions found in Africa, which begin too late to illustrate the state of things before the second century A.D., this is shown by the four contracts of patronatus from the time of Tiberius, quoted in next note, concluded by two small places of the proconsular province Apisa maius and Siagu, and two others nowhere else mentioned, probably adjacent, Themetra and Thimiligi; according to which the statement of Strabo (xvii. 3, 15, p. 833) that at the beginning of the last war the Carthaginian territory numbered 300 towns, appears not at all incredible. In each of those four smaller places there were sufetes; even where the old and new Punic inscriptions name magistrates, there are regularly two sufetes. That these are comparatively frequent in the proconsular province, and elsewhere can only be pointed out in Calama, serves to show how much more strongly the Phoenician urban organisation was developed in the former.

297 The contracts of patronatus from the time of Caesar (C. I. L. viii. 10525), of Augustus (ib. 68 comp. 69), and Tiberius (C. I. L. v. 4919–4922), concluded by the senatus populusque of African communities (civitates) of peregrine rights with Romans of rank, appear to have been entered into quite after the Roman fashion by the common council, which represents and binds the community.

298 On the coin undoubtedly struck under Caesar (Müller Num. de l’Afr. ii. 149) with Kar(thago) Veneris and Aristo Mutumbal Ricoce suf(etes), the first two names are probably to be taken together as a Graeco-Phoenician double name, such as elsewhere is not rare (comp. C. I. L. v. 4922: agente Celere Imilchonis Gulalsae filio sufete). Since on the one hand sufetes cannot be assigned to a Roman colony, and on the other hand the conducting of such a colony to Carthage itself is well attested, Caesar himself must either have subsequently changed the form of founding the city, or the founding of the colony must have been carried into effect by the triumvirate as a posthumous ordinance of the dictator (as is hinted by Appian, Pun. 136). We may compare the fact that Curubis stands in the earlier time of Caesar under sufetes (C. I. L. viii. 10525), in the year 709 U.C. as a Caesarian colony under duoviri (ib. 977); yet the case is different, since this town did not, like Carthage, owe its existence to Caesar.

299 For Africa and Numidia Pliny (H. N., v. 4, 29 f.) numbers in all 516 communities, among which are 6 colonies, 15 communities of Roman burgesses, 2 Latin towns (for the oppidum stipendiarium must, according to the position which is given to it, have been also of Italian rights), the rest either Phoenician towns (oppida), among which were 30 free, or else Libyan tribes (non civitates tantum, sed pleraeque etiam nationes iure dici possunt). Whether these figures are to be referred to Vespasian’s time or to an earlier, is not ascertained; in any case they are not free from errors, for, besides the six colonies specially adduced, six are wanting (Assuras, Carpi, Clupea, Curubi, Hippo Diarrhytos, Neapolis), which are referable, partly with certainty partly with probability, to Caesar or Augustus.

300 Pliny, v. 1, 2, says indeed only of Zulil or rather Zili regum dicioni exempta et iura in Baeticam petere iussa, and this might be connected with the transfer of this community to Baetica as Iulia Traducta (Strabo, iii. 1, 8, p. 140). But probably Pliny gives this notice in the case of Zili alone, just because this is the first colony laid out beyond the imperial frontier which he names. The burgess of a Roman colony cannot possibly have had his forum of justice before the king of Mauretania.

301 Frontinus in the well-known passage, p. 53 Lachm., respecting processes between the urban communities and private persons, or, as it may be, the emperor, appears not to presuppose state-districts de iure independent and of a similar nature with urban territories—such as are incompatible with Roman law—but a de facto refractory attitude of the great land-owner towards the community which makes him liable, e.g. for the furnishing of recruits or compulsory services, basing itself on the allegation that the piece of land made liable is not within the bounds of the community requiring the service.

302 The technical designation gens comes into prominence particularly in the fixed title of the praefectus gentis Musulamiorum, etc.; but, as this is the lowest category of the independent commonwealth, the word is usually avoided in dedications (comp. C. I. L. viii. p. 1100) and civitas put instead, a designation, which, like the oppidum of Pliny foreign to the technical language (p. 331, note), includes in it all communities of non-Italian or Greek organisation. The nature of the gens is described by the paraphrase (C. I. L. viii. 68) alternating with civitas Gurzensis (ib. 69): senatus populusque civitatium stipendiariorum pago Gurzenses, that is, the “elders and community of the clans of tributary people in the village of Gurza.”

303 When the designation princeps (C. I. L. viii. p. 1102) is not merely enunciative but an official title, it appears throughout in communities which are neither themselves urban communities nor parts of such, and with special frequency in the case of the gentes. We may compare the “eleven first” (comp. Eph. epigr. v. n. 302, 521, 533) with the seniores to be met with here and there. An evidence in support of both positions is given in the inscription C. I. L. viii. 7041: Florus Labaeonis f. princeps et undecimprimus gentis Saboidum. Recently at Bu Jelîda, a little westward of the great road between Carthage and Theveste, in a valley of the Jebel Rihan, and so in a quite civilised region, there have been found the remains of a Berber village, which calls itself on a monument of the time of Pius (still unprinted) gens Bacchuiana, and is under “eleven elders”; the names of gods (Saturno Achaiaei [?] Aug[usio]), like the names of men (Candidus Braisamonis fil.), are half local, half Latin. In Calama the dating after the two sufetes and the princeps (C. I. L. viii. 5306, comp. 5369) is remarkable; it appears that this probably Libyan community was first under a chief, and then obtained sufetes without the chief being dropped. It may readily be understood that our monuments do not give much information upon the gentes and their organisation; in this field doubtless little was written on stone. Even the Libyan inscriptions belong, at least as regards the majority, to towns in part or wholly inhabited by Berbers; the bilingual inscriptions found at Tenelium (C. I. L. viii. p. 514), in Numidia westward from Bona in the Sheffia plain, the same place that has furnished till now most of the Berber stone inscriptions, show indeed in their Latin part Libyan names, e.g. Chinidial Misicir f. and Naddhsen Cotuzanis f., both from the clan (tribu) of the Misiciri or Misictri; but one of these people, who has served in the Roman army and has acquired the Roman franchise, names himself in the Latin text in civitate sua Tenelio flamen perpetuus, according to which this place seems to have been organised like a town. If, therefore, success should ever attend the attempt to read and decipher the Berber inscriptions with certainty, they would hardly give us sufficient information as to the internal organisation of the Berber tribes.

304 That the Gaetulian purple is to be referred to Juba is stated by Pliny, H. N. vi. 31, 201: paucas (Mauretaniae insulas) constat esse ex adverso Autololum a Iuba repertas, in quibus Gaetulicam purpuram tinguere instituerat; by these insulae purpurariae (ib. 203) can only be meant Madeira. In fact the oldest mention of this purple is that in Horace, Ep. ii. 2, 181. Proofs are wanting as to the later duration of this manufacture, and, as the Roman rule did not extend to these islands, it is not probable, although from the sagum purpurium of the tariff of Zarai (C. I. L. viii. 4508) we may infer Mauretanian manufactures of purple.

305 The tariff of Zarai set up at the Numidian customs-frontier towards Mauretania (C. I. L. viii. 4508) from the year 202 gives a clear picture of the Mauretanian exports. Wine, figs, dates, sponges, are not wanting; but slaves, cattle of all sorts, woollen stuffs (vestis Afra), and leather wares play the chief part. The Description of the earth also from the time of Constantius says, c. 60, that Mauretania vestem et mancipia negotiatur.

306 According to an epitaph found in Mactaris in the Byzacene (Eph. epigr. v. n. 279), a man of free birth there, after having been actively engaged in bringing in the harvests far around in Africa, first throughout twelve years as an ordinary reaper and then for another eleven as a foreman, purchased for himself with the savings of his pay a town and a country house, and became in his turn a member of council and burgomaster. His poetical epitaph shows, if not culture, at least pretensions to it. A development of life of this sort was in the Roman imperial period doubtless not so rare as it at first may seem, but probably occurred in Africa more frequently than elsewhere.

307 How far our Latin texts of the Bible are to be referred to several translations originally different, or whether, as Lachmann assumed, the different recensions have proceeded from one and the same translation as a basis by means of manifold revision with the aid of the originals, are questions which can scarcely be definitely decided—for the present at least—in favour of either one or the other view. But that both Italians and Africans took part in this work—whether of translation or of correction—is proved by the famous words of Augustine, de doctr. Christ. ii. 15, 22, in ipsis autem interpretationibus Itala ceteris praeferatur, nam est verborum tenacior cum perspicuitate sententiae, over which great authorities have been perplexed, but certainly without reason. Bentley’s proposal, approved afresh of late (by Corssen, Jahrb. für protestant. Theol. vii. p. 507 f.), to change Itala into illa and nam into quae, is inadmissible alike philologically and in substance. For the twofold change is destitute of all external probability, and besides nam is protected by the copyist Isidorus, Etym. vi. 4, 2. The further objection that linguistic usage would require Italica, is not borne out (e.g. Sidonius and Iordanes as well as the inscriptions of later times, C. I. L. x. p. 1146, write Italus by turns with Italicus), and the designation of a single translation as the most trustworthy on the whole is quite consistent with the advice to consult as many as possible; whereas by the change proposed an intelligent remark is converted into a meaningless commonplace. It is true that the Christian Church in Rome in the first three centuries made use throughout of the Greek language, and that we may not seek there for the Itali who took part in the Latin Bible. But that in Italy outside of Rome, especially in Upper Italy, the knowledge of Greek was not much more diffused than in Africa, is most clearly shown by the names of freedmen; and it is just to the non-Roman Italy that the designation used by Augustine points; while we may perhaps also call to mind the fact that Augustine was gained for Christianity by Ambrosius in Milan. The attempt to identify the traces of the recension called by Augustine Itala in such remains as have survived of Bible translations before Jerome’s, will at all events hardly ever be successful; but still less will it admit of being proved that Africans only worked at the pre-Hieronymian Latin Bible texts. That they originated largely, perhaps for the most part, in Africa has certainly great probability. The contrast to the one Itala can only in reason have been several Afrae; and the vulgar Latin, in which these texts are all of them written, is in full agreement with the vulgar Latin, as it was demonstrably spoken in Africa. At the same time we must doubtless not overlook the fact that we know the vulgar Latin in general principally from African sources, and that the proof of the restriction of any individual linguistic phenomenon to Africa is as necessary as it is for the most part unadduced. There existed side by side as well vulgarisms in general use as African provincialisms (comp. Eph. epigr. iv. p. 520, as to the cognomina in -osus); but that forms like glorificare, nudificare, justificare, belong to the second category, is by no means proved from the fact that we first meet with them in Africa, since analogous documents to those which we possess, e.g. for Carthage in the case of Tertullian, are wanting to us as regards Capua and Milan.

308 The arguments of Mr. B. W. Henderson (English Hist. Review, 1903, 1–23) for a different advance seem to me to be based on a misconception of some of the evidence. Thus, there is no tile of Leg. ix. at Leicester, nor any trace yet noted of Leg. ii. Aug. at Cirencester or Gloucester.

INDEX