^{page 53 note 1} Tac. H. i, 8, 76; ii, 58, 59, 65; iii, 65.

^{page 53 note 2} Tac. H. ii, 67, 86, 97; iii, 44.

^{page 54 note 1} Tac. H. ii, 97; iv, 39 (where ‘discessu’ of the MS. is the right reading). Cf. Mommsen, , Hermes iv, p. 319Google Scholar.

^{page 54 note 2} Tac. H. iv, 53; xiv, 3608 = Dessau, 986; P.I.R. P. 363.

^{page 54 note 3} Dessau, 8970. Cf. P.I.R. V, 574; Waddington's latest date for his consulship, 71, is to be accepted, rather than Stobbe and Henzen's (E.E. i, p. 188, n. 2)—68; cf. Cagnat, , L'Armée rom. au Siège de Jerusalem (Paris 1891), p. 19Google Scholar. As a legionary command usually followed the praetorship immediately, it is not likely that Trajan was procos. Baeticae before the Jewish war.

^{page 54 note 4} P.I.R. C 184. 2477 is his earliest record in Spain, of the first half of 79; 4803, 4838, 4847, 4854, 6224, etc. are dated from the first half of 80.

^{page 55 note 1} P.I.R. L 54, P 373. Münzer (Bonn. Jahrb. 104, pp. 103–111) thinks that Largius succeeded Silvanus in 73 as leg. Aug. pro pr. because of his wide-spread activities (Plin. 19, 35; 31, 24). Against this cf. Kornemann in P.W. v, 719 ff; Klio iii, P. 323. In any case a praetorian could hardly interrupt a series of consulars. See p. 87, n. 2.

^{page 55 note 2} 2477 already cited.

^{page 55 note 3} P.I.R. P 501, A 934, argues that he was cos. in 72, not 78, in which view Borghesi and Mommsen (x, p. 840) led the way. Lusitania would thus have a consular governor; but even if this were possible (Mommsen, , Droit, iii, p. 287Google Scholar, n. 1) it would be very unlikely; no parallel is quoted, and Otho had recently governed the province as a quaestorian only. I prefer Hübner's view (5264).

^{page 55 note 4} 189, a fragment from Olisipo, names ‘…. tius Quadratus leg. Aug. pro pr.’ This is more probably Veratius, frater Arvalis 77-91, than his contemporary Antius Iulius Quadratus whose career was entirely eastern. Veratius' consulship is not recorded. No other than these two is known to whom the fragment may be referred.

^{page 55 note 5} vi, 1359. P.I.R. B 8. He was ‘adlectus inter praetorios’ for his services, probably in 73–74.

^{page 55 note 6} Suet. Galba, 10; Tac. H. i, 6. 16; ii, 58, 59, 67; iv, 68; v, 16, 19. Ritterling, de leg. X Gem. p. 36.

^{page 56 note 1} Dio. 55, 24; Suet. l.c.; Tac. H. i, 6; ii, 67, 86; iii, 6, 7, 19–25; iv, 39. cf. Boissevain, de Re Militari provv. Hisp. (1879), p. 31.

^{page 56 note 2} Jos. B.J. vii, 1, 5, 3; Ritterling in Rhein. Mus. 59, pp. 55–62.

^{page 56 note 3} 2477; vi 3538 =Dessau, 2729, names a ‘tribunus militum leg. VII Geminae Felicis in Germania’ who is now proved by a new diploma (A.E. 1906, 99) to have next commanded the Ala Moesica in Germ. Inferior under Rutilius Gallicus in the war of 77–78. Hence Domaszewski (Die Alterthümer unserer heidnischen Vorzeit, 1905, p. 184) supposes the legion to have been present in that war; but Ritterling (Korrespondenzblatt d. Westdeutschen Zeitschrift, 1906, 20–28) argues rather for the earlier operations in Germ. Superior; and his view is strongly confirmed by the later discovery of inscribed tiles of the legion near Mainz (Römgerm. Korrespondenzbl. iv, 1911, p. 37)Google Scholar. The only question now is, whether the legion did not take part in both wars. The evidence that the bestowal of the title ‘felix’ was a special distinction for the later war seems hardly sufficient to justify Domaszewski's inference for the legion (Ph. 56, 1907, p. 167, n. 32), though the inscription from Borbetomagus quoted by Ritterling (xiii, 6212), which reads ‘Leg. VII G F’ and is probably of Vespasian's reign, does not by itself prove that the title was given in Upper Germany; successive commands need not be in the same province or army. On the other hand, we have evidence from the new inscription of Baalbek (A.E. 1903, 368) that the reinforcements for the second war were provided by ‘vexillationes’ from eight other legions, which affords a presumption against the employment of the Leg. VII whether wholly or in part. Also the greater distance and the short interval point to the same conclusion. The whole legion was again in Upper Germany in 88.

^{page 56 note 4} 2425, 2583, 2706, 4157, 5265, 5681, besides 2477, are Flavian records of the legion in Spain, Cf. Ritterling, in Rh. Mus. 59 (1904), p. 55Google Scholar.

^{page 56 note 5} 5674 of 97–8. Boissevain, op. c. p. 47. Hübner, p. lxxxix. Mommsen in Hermes, vii, p. 309, places the ‘canabae’ at Asturica (Astorga) 40 m. from Leon, on account of 2637, which, however, was found half-way between the two places. He pleads also that Asturica had no municipality (a characteristic of legionary head-quarters at least in the early empire); but this, as we shall see, is a mistaken view. Its ‘res publica’ is actually recorded 2636 (second century).

^{page 57 note 1} I agree with Trommsdorf as against Schilling that the legions I Germ, and XV Pr. were both disbanded by Vespasian. For the patriae of Leg. VII cf. indices to ii, also v, 920, 926; viii, 3226, 3245, 3268; Hübner, p. lxxxix; of I Germ., Siebourg in Bonn. Jahrb. 107, pp. 177 ff.; of XV Pr., C.I. Rhen. 479–80. Grotefend in Pauly makes the legion I Germ.; Boissevain, p. 44, proposes the I Macriana Liberatrix. But this must have been disposed of long before, and there is no trace of Africans in Spain. For the title ‘gemina’ cf. Caes. B.C. iii, 4.

^{page 57 note 2} 2552, as emended in A.E. 1910, 3; cf. ib. 5. Cagnat in D.S. s.v. Legio, gives Jan. 11, 68, as the birthday. This is certainly wrong as far as the month is concerned.

^{page 57 note 3} Suet. Galba, 23.

^{page 57 note 4} So Ritterling in Röm.-germ. Korrespondenzbl. iv, 1911, p. 41, n. 8: but the tiles reproduced, p. 38 (four in number, from Rheinzabern near Mainz) read only ‘Leg. VII G.’

^{page 58 note 1} Plin. Pan. 14; Tac. H. iv, 68, v, 19; Dessau, 2648 (res prospere gestae contra Astures) A.D. 66. Mommsen (Provs. i, p. 65, n. I), places two legions, the second prob. the I Ad., in Spain 71–88; but in S. Berl. 1903, p. 821, he places the I Ad. in Germany, adding that possibly it was then in Spain. Jünemann (de leg. I Ad. pp. 36 ff.) and Kornemann (P.W. v, 719 ff.) follow Mommsen's earlier view; Gsell, Domitien, p. 179, places the legion in Spain only from 83 to 88. Ritterling, Domaszewski and Dessau place the VII Gem. alone in Spain: cf. especially Ritterling in Röm.-germ. Korr.-bl. iv, p. 40, where eight newly-discovered tiles of the I Ad. from Rheinzabern are figured. Filow (Die Legionen der Prov. Moesia, p. 17, n. 1, and p. 40, n. 2) revives the two-legion theory, because otherwise Aelianus' appointment would have been retrograde after the command of four legions in Moesia.

^{page 58 note 2} Suetonius, a professed historian, says (Galba 10) that Galba levied ‘legiones’ in Spain; but nobody insists upon the letter of that statement.

^{page 58 note 3} Cf. n. 4, p. 56.

^{page 58 note 4} Hübner, p. lxxxix, rightly says that the I Ad. never returned to Spain after 70, but wrongly assigns to its short sojourn in 69–70 five inscriptions which are all ‘cursus honorum’ of later date and prove nothing. The legion finally left Germany (after several temporary absences) in 107. Cf. Jünemann, op. cit. pp. 18, 27, 110; Ritterling, , Westdeutsche Zeitschrift, xii, 1893, pp. 105Google Scholar ff.; Domaszewski, , Rb. Mus. 46, p. 602Google Scholar, n. 3; Dessau 2277; and n. I supra.

^{page 58 note 5} A.E. 1903, 368. Mommsen, who first commented on this inscription, S. Berl. 1903, 817 ff. referred these vexillationes to a contemporary war in Mauretania; but Domaszewski, in the light of later discoveries, is clearly right in assigning them to Rutilius Gallicus in Germ. Inf. 77–8 (Ph. 66, 1907, pp. 164 ff.). Cf. Ritterling, J. O.A.I. 1904,. Beiblatt, 23 ff. It is to be noted, however, that a little doubt is caused by the reading of the inscription ‘leg(ionum) VIIII’ though the list immediately following specifies only eight by name, Mommsen and Domaszewski agree (against Ritterling) in supposing that VIIII is the stone-cutter's error for VIII (Ph. 1. c. p. 166, n. 26).

^{page 59 note 1} Suet. Galba 10; Dessau 2693, ‘praef. cohortium civium Romanor. quatuor in Hispan.’—P.I.R. C 1082; N. Dig. Occ. p. 119 Boeckhing.

^{page 59 note 2} I have constantly referred to Cichorius' articles P.W. s.vv. Ala and Cohors; also to the appendices in Mr. Cheesman's ‘Auxilia.’ See p. 82 infra.

^{page 59 note 3} E.E. ix, 277 (near Aquae Flaviae), A.E. 1910, 2 and 5 = 2554.

^{page 59 note 4} E.E. viii, 128, p. 408 (Zoelae) with n. Cichorius in P.W. iv, 290.

^{page 59 note 5} A.E. 1910, 3 = 2552; 4 = 2553; Cichorius in P.W. iv, 268. I take it to be the C. Celtibera at Brigantia (Callaecia) of N. Dig. Occ. xlii, 30. The name ‘Baetica’ is a misreading in the Corpus.

^{page 59 note 6} A.E. 1910, 1,2,4 (= 2553), 6 ( = 2556), N. Dig. Occ. xlii, 32.

^{page 59 note 7} E.E. viii, 131, p. 408, with n. Cichorius P.W. iv, 290, distinguishes three Cohortes Iv Gallorum, which Hübner confuses.

^{page 59 note 8} Cichorius, P.W. iv, 311. N. Dig. Occ. xlii, 29.

^{page 59 note 9} Of Mr. Cheesman's units I have excluded Ala I Lemavorum as insufficiently attested; C III Celtiberorum, which seems to be based only on a misreading of A.E. 1910, 4 ( = 2553); and C II Gallica for which see infra. The Ala Sabiniana and C IV Gallorum are outside his time-limit.

^{page 59 note 10} p. 540, and inscriptions 4142 ff.

^{page 60 note 1} xi, 623, ‘praef. orae marit. Hispan. Citerioris bello Actiensi.’

^{page 60 note 2} Digest 47, 9, 7, quoted by Mommsen on x, 6785—a text of Hadrian's reign—indicates wreckers. Plin. Ep. x, 21 and 86, describes the ‘praef. orae Ponticae’ as ‘integer, probus, industrius,’ which suggests responsibility for revenue.

^{page 60 note 3} 4138 = Dessau 2715, Jos. B. J. iii, 486. The identification rests mainly upon the coincidence of name, but the mention of the C IV Thracum (Syriaca) is some confirmation. A good soldier was wanted for the post. 4212 is another instance of a Praef. Coh. IV Thracum from Tarraco perhaps before 80. The cohort had probably left Germany for the East by 67. Cf. xi, 5744 for praef. orae in Mauretania.

^{page 60 note 4} 4225-6 = Dessau 2714, 2714A. The second inscription seems to restrict the sphere of the prefect to Lacetania, N. of Tarraco, which is hardly credible. P.I.R. L 170. This is a good illustration of the rise of a new nobility.

^{page 60 note 5} 4217. Cf. also 4224, 4239, 4264, 4266 (= Dessau 2716–7). The first two of these at least are probably of Flavian date also.

^{page 60 note 6} The title ‘Nova’ was exchanged for ‘Prima’ when C. II was levied.

^{page 60 note 7} E.E. viii, 92, p. 388, is perhaps the oldest record of the legion at Italica (end of first century): there are many later.

^{page 60 note 8} 2224, ‘trib. militum Coh. Maritimae’ at Corduba; cf. Cichorius, P.W. iv, 314. Plin. Ep. iii, 9, 18, refers to a ‘trib. cohortis sub Classico’ before 96; as the unusual tribunus shows, the same cohort is meant.

^{page 61 note 1} 5439, c. 103, for the militia at Urso.

^{page 61 note 2} Assigned to Tarraconensis by Mr. Cheesman (who does not seem to recognise a Lusitanian garrison, at least in the second century). The only record is N. Dig. Occ. xlii, 28; but this indicates a long sojourn, and the C. may have been always in the peninsula, but first in Lusitania.

^{page 61 note 3} 1127 names a [mil. cohor]tis III Gallorum at Italica in the first century; and perhaps 403 (Viseu in Lusitania) which mentions chor. III …. all … belongs to it. But it may also refer to C III Callaecorum, which is not otherwise recorded, but may perhaps be identified with C III Lucensium, which we have already assigned to Tarraconensis; for as C V Callaecorum Lucensium shows, all this series may have had alternative or double titles. E.E. ix, 46, p. 27, names a prefect of some Coh. III who was patronus of Emerita.

^{page 61 note 4} Cichorius, P.W. iv, 312, takes this view; E.E. viii, p. 360, gives an inscription probably of second century, and 5238 confirms. If CIII Lusitanorum lay originally in its native province, it had certainly gone to Germany before 70.

^{page 61 note 5} I take the C I Flavia Ulpia Hisp. mil. C. R. eq. to be the same as C I Fl. Hisp. mil. eq.: it had passed to Dacia and won a new honour. So Cichorius, who also believes in yet another CIFl. Hisp. eq. p. f. (quingenaria) but this is improbable though ‘p. f.’ is not accounted for. Troops from military districts were called by their own tribe-names. For the other Spanish units see infra, p. 82, n. 1.

^{page 62 note 1} Victor, Caesares ix, 8 = Epit. ix. 9; Jos. C. Apionem ii, 4, 40.

^{page 62 note 2} Plin. iii, 30. Mommsen, Stadtrechte der Salpensa und Malaca, p. 400, n. 22.

^{page 62 note 3} De maiore et minore Latio (Berlin 1860), p. 15Google Scholar.

^{page 62 note 4} Zumpt, Stud. Rom. p. 313. Nipp. Opusc. p. 433. Sillig reads ‘iactatus’; later edd. mostly ‘iactatum.“ Harduin conjectures ‘iactatae.’

^{page 62 note 5} So Hirschfeld (Gesch. d. latinischen Rechtes, in Kleine Schriften, p. 303, n. 4): I am indebted to him for the preceding reff. also, and for several that follow.

^{page 62 note 6} Municipalities of the Rom. Emp. p. 241. But Dr. Reid now accepts iactatum, proposing a new interpretation. He writes to me: ‘This happens to be the only ref. out of a number in Pliny to civil wars, in which the context does not indicate the particular war meant. I think now that iactatum has the sense of “debated,” “talked about,” as opposed to the actual carrying out by Vespasian. The civil war can hardly be any but that between Caesar and Pompey. Caesar showed his readiness to confer Latinitas on cities in Gallia Cisalpina and Sicily. It is not recorded that he thought of granding it to Spain, but it is quite possible. His interest in Spanish municipal affairs is shown by Urso, Gades, and other places. It must be admitted that procellis rei p. as a mere indication of time is awkward after iactatum in the assumed sense. It has crossed my mind of late to suppose that procellis rei p. (which is an odd phrase for pliny to use in the context) may be a gloss due to a misunderstanding of iactatum. Some time-word like olim may have been displaced.’ But, apart from these textual difficulties, it is hard to believe that Caesar could have thought of a general grant, in view of the strong and recurrent hostility of Spain, and the backwardness of most of the Hither and much of the Farther Province. If a partial grant to the more Romanised South is assumed to have been in question, Pliny's universae may be taken as an implied contrast. But evidence for that is lacking; and the exceeding bitterness of Caesar's speech at Hispalis (Bell. Hisp. 42) seems almost to exclude the possibility, whether before or after Munda. Dessau, 103, shows that Baetica was reckoned as pacata (and separated from Lusitania) not much earlier than 2 B.C.

^{page 63 note 1} Suet. Galba 10, 12; Plut. Galba 4.

^{page 63 note 2} Cic. ad Att. xiv, 12; Tac. Ann. xi. 23 ff.; Seneca, Apocol. 3.

^{page 63 note 3} Tac. H. iii, 55 ‘Latium externis dilargiri.’ Ib. i, 78, for Otho's ‘nova iura’ promised to Cappadocia and Africa; in the latter case probably as Hirschfeld thinks a promise of Latium.

^{page 63 note 4} Augustus promoted many Alpini (Plin. iii, 135) and as we shall see Hispani. Tiberius Claudius and Nero (Tac. Ann. xiii, 32) followed his example to some extent, but Vespasian first returned to the policy of Caesar.

^{page 64 note 1} Vell. Paterc. ii, 90. Contrast Varro, R.R. i, 162. Cf. Bouchier, Spain under the Roman Empire, pp. 25–6.

^{page 64 note 2} 1963–4 = Dessau 6088–9. Dr. Hardy in Three Spanish Charters gives a convenient commentary which has been of much use to me. I have not seen the article on Ius Latii recently published in P.W.

^{page 64 note 3} E.E. ix, 261 = c. 67 of Lex Malacitana; cf. Dessau ad loc. and Mommsen ib. p. 10. Dessau suggests that the Salpensan law was never published, since the name of Domitian is not erased as in the Malacan document.

^{page 64 note 4} Mommsen, , Droit vi, 2Google Scholar, p. 24; Willems, Droit Publ. p. 143.

^{page 65 note 1} Gaius i, 95–6. See p. 93 and n. 7.

^{page 65 note 2} So Mommsen, , Droit, vi, 2, p. 263Google Scholar, n. 3.

^{page 65 note 3} 1610 at Igabrum, A.D. 75.

^{page 65 note 4} E.g. 1631. Two inscriptions, 1945, 2096, name the duovirate as the avenue—1610 probably the aedilitas also. There are many citizens with no offices recorded; many too where a complete ‘cursus honorum’ is indicated, though often the duovirate only is mentioned.

^{page 65 note 5} Lex Mal. 54 fixes the age of 25 as the limit for quaestors, aediles, and duoviri alike; so Lex Salp. 25. Pius prescribed ‘ut gradatim honores deferantur.’

^{page 65 note 6} Lex Sal. 21 : the passage is mutilated, but the sense is clear. For the interpretation, cf. Hardy, op. cit. pp. 66, 84. He cites 1286 as an instance of a youthful civis at Salpensa who died at eighteen, and must have gained the franchise through his-father's magistracy.

^{page 66 note 1} E.g. even at Malaca itself, and at Munigua.

^{page 66 note 2} The contemporary Latin charter of Icosium in Mauretania (cf. p. 77, n. 3) bestowed the title of ‘colonia’, viii, 20853, Itin. Ant. p. 15.

^{page 66 note 1} Gellius, xvi, 13; Mommsen, Droit, vi, 2, p. 444, n. 3. Toutain, s.v. Municipium in D.S., rejects Mommsen's view as taking no account of Hadrian. Cicero, pro Balbo, 21, refers to the same right of Latin towns in earlier days: ‘in-numerabiles aliae leases de civili iure sunt latae: quas Latini voluerunt, asciverunt’; cf. Mommsen, op. cit. vi, 1, p. 324.

^{page 67 note 1} Toutain l. c. seems to apply Hadrian's words to all municipia, but he does not discuss the points of doubt. He is wrong in supporting his view from the actual contents of the charters.

^{page 67 note 2} For these points, which are not touched upon in the charters, cf. D.S. s.v. magistrate municipals (Lécrivain), p. 1544a (census 1546b (expenses); also s.v. senatus municipales (Lécrivain), p, 1201, and s.v. honoraria summa (Cagnat); cf. 5439, 70–71, for other expenses. 5232 is a record of a decurion's honorarium in Spain, A.D. 167; there is no other direct evidence on these points for the peninsula, and since the records are so rare and uncertain as compared with Africa it is possible to suppose that the system of ‘honoraria’ was much less developed in Spain.

^{page 67 note 3} Lex Mal. 51, 66, with Dr. Hardy's notes On the general questions, cf. Mommsen, Droit, vi, I, p. 399–401; vi, 2, P. 472–3.

^{page 67 note 4} Lex Sal. 24; Lex Mal. 60, 63, 69, with Dr. Hardy's notes.

^{page 68 note 1} 2021, 2025. Cf. Toutain in D.S. s,v. municipium, p. 2033, who discusses the title with ref. only to three African municipia, Thugga (viii, 1484), Thubursicum Bure (ib. 1427, 1439), and Aulodas (ib. 14355), See also Mommsen, Droit, vi, 2, p. 441, n. 1, and p. 461, nn. 1, 2. Detlefsen, however (Ph. 30, p. 271), thinks the title merely a survival without real force; just as Aventicum Helvetiorum retained its title (‘foederatum’ even when it was probably obsolete.

^{page 68 note 2} Plin. Ep. iii, 4Google Scholar, 9; vii, 33.

^{page 68 note 3} Plin. iii, 6–17 Baetica; 18–30, 76–9, iv, 110–2 Hisp. Citer.; iv, 113–8 Lusitania. For a detailed examination of these passages cf. Detlefsen in Ph. 30, 265–310; 32, 600–68; 36, 111–128.

^{page 68 note 4} Curiga, Bocchorum, and Flaviobriga: see infra.

^{page 68 note 5} Kubitschek, de Romanarum Tribuum originc, etc. p. 156; Mommsen, E.E. iii, pp. 230 ff. While the comitia still retained some power the emperors, did not swamp their own tribes by new members.

^{page 69 note 1} 159, at Ammaia in Lusitania. There were four towns in Mauretania Caesariensis enrolled by Claudius in Quirina; yet Volubilis in Tingitana owes its tribe Claudia to him, and Tingitana was closely connected with Baetica.

^{page 69 note 2} iii, 6419; cf. Kubitschek, pp. 118 ff. 199.

^{page 69 note 3} p. 875; Itin. Ant. 407, 3; Ptol. ii, 6.

^{page 69 note 4} Carmo, Epora, Iulipa, Illiturgicola, [Burguillos]. Detlefsen shows that all or nearly all of the 46 towns are mentioned by Pliny, which makes the distinction easy.

^{page 69 note 5} Strabo, iii, 2, 15.

^{page 70 note 1} In questions of tribe I have consulted Kubitschek, op. cit. and especially his Imp. Rom. tributim discriptum, without, however, accepting all his conclusions. For town identifications I have consulted Müller on Ptolemy, vol. i, as well as Hübner's notes in the C.I.L. which is fundamental. More exact reff. will generally be unnecessary.

^{page 70 note 2} 1065, etc. and p. 138; E.E. ix, p. 74.

^{page 70 note 3} pp. 140, 837; E.E. viii, p. 354 ff. no. 91.

^{page 70 note 4} 1049–50, dedications to Divo Vespasiano Censori and Divo Tito Censori by ‘Municipium Muniguense.’

^{page 70 note 5} 1192 Hübner.

^{page 70 note 6} 1028=5543. Hübner rejects M. F[lavium] as misread for M. I[ulium]; but a mun. Iulium must have been recorded by Pliny, as Detlefsen shows. Cf. however, 5544–9.

^{page 71 note 1} Plin. iii, 12; 954. Kubitschek, Imp. Rom. p. 169.

^{page 71 note 2} 2041, which Hübner takes as proving a mun. Flavium. Clunia Sulpicia is not quite parallel.

^{page 71 note 3} 1949–54–56. Detlefsen, op. cit. p. 272.

^{page 71 note 4} 2096 of A.D. 75. 2098 seems to indicate Galeria, but Hübner finally agrees with Kubitschek in accepting Quirina (pp. 292, 885; Imp. Rom. p. 172).

^{page 71 note 5} p. 215 calls it mun. Flavium, which title is not given in 1610 of A.D. 75.

^{page 71 note 6} 1648–9 (Hübner).

^{page 72 note 1} p. 211 and nos. 1569, 1572; Quirina in 1597 and A.E. 1903, 233. The confusion is due to the name ‘mun. Contributum.’

^{page 72 note 2} p. 275 Hübner. vi, 2629.

^{page 72 note 3} 1459–60, 5048.

^{page 72 note 4} Hübner on pp. 166, 301, nos. 1258, 2150.

^{page 73 note 1} Detlefsen, Ph. 36, pp. 113, 117–8.

^{page 73 note 2} E.E. viii, 301 (pp. 354 ff.) Hübner.

^{page 73 note 3} 880 ‘Cassius Vegetus Celticoflavensis.’

^{page 73 note 4} Ursin, de Lusitania prov. Rom. p. 89, quotes a coin (Mionnet, Suppl. i, no. 10) to prove Balsa municipium before Vespasian; but the coin is doubtful. For Quirina, I follow Kubitschek, Imp. Rom. p. 184, against Hübner, p. 786.

^{page 73 note 5} 395, 5264.

^{page 73 note 6} E.E. viii, 4 (p. 356) mentions a ‘praefectus Caesarum (Aug. and Tib.) bis’ which might indicate governmental activity in this direction here. The inscription was formerly by error referred to Aquae Flaviae (2479 = 5617).

^{page 74 note 1} Detleften, Ph. 32, 603–7; Plin. iii, 18, 26–7–8, iv, 111–2. He names at least six towns in the other three conventus, and 18 in Cluniensis.

^{page 74 note 2} Strabo, iii, 3, 8, οὐ μόνον εἰρηνικούς, ἀλλά καὶ πολιτικοὺς ήδη τινὰς αὐτῶν ἀπεργασάμενος τυγχάνει.

^{page 74 note 2} Avobriga, Boletum, Clunia, Damania, Dianium, Ilugo, Labitolosa, Lucus Augusti, Mentesa Bastitanorum, Pompaelo, Segobriga, Sigarra, Termes, Uxama, with towns on the site of Albarracin (Lobetum?), Mancha Real and Gandia. The list is formed by a comparison of the inscriptions.

^{page 74 note 4} 3361 gives a II vir in Galeria; probably (Kubitschek) he was rather a citizen of the neighbouring Mentesa.

^{page 74 note 5} 3250 is a dedication to Titus the Censor, decreto decurionum, A.D. 76.

^{page 74 note 6} 4494. But 4495 names a II vir in Galeria, which Kubitschek (Imp. Rom. p. 194) assumes to be the official tribe, supposing that the town ‘Latio vetere utebatur’ though it was reconstituted as a mun. Flavium. If so I can find no certain parallel; yet cf. Cisimbrium, p. 71, n. 4.

^{page 75 note 1} A ‘mun. Flav. Rhodinorum’ was inferred by Hübner (p. 988) from a misread African inscription (viii, 1148 = 14279).

^{page 75 note 2} On 3008 Hübner proposed to read ‘Flavitolosani’ but; Labitolosani' is now accepted (5837).

^{page 75 note 3} 4468 (II vir in Galeria) is difficult, but the office may have been held in another town; cf. Hübner, p. 594.

^{page 75 note 4} E.E. viii, p. 354 ff. no. III = 6291; 2637, 5124 and p. 911 Hübner.

^{page 75 note 5} The ‘magister’ of 3033 is not a non-Latin official, but may be compared with the magistri of a vicus contributus at Asturica, 2636.

^{page 75 note 6} 4318, Palma being in Velina. Detlefsen (Ph. 32, 651) identifies Pliny's Cinium, oppidum Latinum (iii, 77) with Guiuntum: thus its tribe might be referred to Caesar, for which Carales in Sardinia would be a parallel. But a Phoenician town would hardly be chartered so early. Palma and Pollentia were Roman.

^{page 76 note 1} 5796 found 3 m. from the site.

^{page 76 note 2} 2628 = 5649, saec. i (Hübner).

^{page 76 note 3} E.E. viii, p. 354 ff. no. 137.

^{page 76 note 4} Detlefsen, Ph. 32, 639 on 2981.

^{page 76 note 5} ‘Foederatum Bocchorum fuit’; it got a new charter like Malaca. It was of some importance under Augustus, but no later inscriptions record it.

^{page 76 note 6} Transferring Flaviobriga from the first to the second category.

^{page 77 note 1} Plin. iii, 18 = vii, 96. In the second passage there is a variant 776. In the first Pliny seems to exclude Baetica from the total, surely by error. Pliny records only 80 towns for Narbonensis, and Ptolemy only 30. For a commentary on Pompey's claim, cf. Strabo, iii, 4, 13, οί ϕάσκοντες πλείους ἢ χιλίας τὰς τῶν Ἰβήρων ὑπάρξαι πόλεις ὲπὶ τοῦὑτο [τὸ καλλωπίζειν] ϕέρεσθαί μοι δοκοῦσιν, τὰς μεγάλας κώμας πόλεις ὀνομάζοντες

^{page 77 note 2} iv, 117, ‘contributa sunt in eam Castra Servilia, Castra Caecilia.’

^{page 77 note 3} iii, 19; v, 20. The identification is Detlefsen's.

^{page 77 note 4} According to Detlefsen's probable explanation of Plin. iv, 118, ‘praeter iam dictos in Baetica cognomines,’ a ref. to iii, 13. The MSS. have ‘in Baeticae cognominibus’ which may have the same sense.

^{page 77 note 5} De longaevis frag. 29, 1 in F.H.G. iii, p. 608–9 (πόλεως Άπειλοκαρίου).

^{page 77 note 6} More might be added from the Itinerary and other sources, but possibly with less certainty: these suffice for our point.

^{page 78 note 1} 2477.

^{page 78 note 2} Guiuntum. Cf. p. 75, n. 6. Yet if the population was only between 30,000 and 40,000, as Diodorus reports probably of earlier times, the communes were not large. The present population is over 200,000.

^{page 78 note 3} vi, 1463.

^{page 78 note 4} Bocchorum is included. Cf. p. 76, n. 5.

^{page 78 note 5} The total for Citerior is diminished by one—p. 69, n. 2 (Segovia).

^{page 79 note 1} 1049–50, 3250; 2096, 1610.

^{page 79 note 2} 2096. Cf. p. 65, with notes. Mommsen, Droit, v, p. 167, n. 3, says: ‘La mention de Titus à côté de Vespasien pour la concession de la łatinité aux villes espagnoles porte à se demander si cette concession ne se serait pas liée aux pouvoirs censoriens Mais la preuve du contraire est qu'elle a eu lieu en I'an 75, et que le lustre fut accomplidès l'an 74.’ This contradicts his earlier and correct view.

^{page 79 note 3} Lex Sal. c 22 ad init. ‘Qui quaeque ex hac Iege exve edicto imp. Caesaris Aug. Vespasiani imp. ve Titi Caesaris Aug., autimp. Caesaris Aug. Domitiani.… civitatem Romanam consecutus consecuta erit ….’ Cf. 23 ad init.

^{page 79 note 4} Two coins of 78, one of Vespasian and one of Titus (Eckhel. vi, pp. 336, 356) have R/ ‘scrofa cum suculis.’ Hirschfeld comparing 2126—a dedication in the early empire of a ‘scrofa cum porcis triginta’—refers this to new grants of Latium. But the town in question possessed full civitas, not Latium merely.

^{page 79 note 5} 3695, Bocchorum A.D. 6; 5048, Ostippo A.D. 15–20; 1953, Cartima A.D. 53–4; 2633, Zoelae A.D. 27: 172, Aritium Vetus A.D. 37; 1569, Ipsca A.D. 46; 1343, Lacilbula A.D. 5; 5346, Augustobriga Lusitaniae, probably under Augustus —eight instances. Cf. p. 75 n. 5; Mommsen, Provs. i, p. 73, n. 1.

^{page 80 note 1} A II vir at Consabura is enrolled in Hadrian's tribe, Sergia; but since he was a civis under Trajan, that tribe cannot be the official one of the town, which was stipendiaria in Plin. iii, 25, and 4211. 159 records a II vir in Quirina at Ammaia, who was ‘viritim a divo Claudio civitate donatus’: the town was thus a municipium under Vespasian. But 167 records a private citizen in Papiria, Trajan's tribe, which cannot be official. Kubitschek, de orig. trib. Rom. p. 156, refers this man to Emerita, probably rightly, though Hübner. p. 809, objects.

^{page 80 note 2} 3271 ‘….fisci et curatori divi Titi in Bae | tica prae Galleciae pref. fisci | Germaniae Caesarum Imp. tribu | no leg. VIII flamini Augustali | in Baetica primo ….’ The original was broken at both ends; it has disappeared, and its genuineness has been suspected, probably wrongly, but its importance is such that any doubt is unfortunate. It is accepted by Hübner, Kornemann, and others; Hirschfeld accuses it of interpolation or misreading (Verwaltungsb.² p. 6, n. 2) and Dessau omits it from his selection. It has no certain mark of forgery.

^{page 80 note 3} 1041 of Vespasian's reign (Hübner). Mommsen ad loc. thinks the pagi were added to Contributa Iulia. Hübner in P.W. iv, 1836, assigns Quirina to Curiga from 1042 rightly, though Kubitschek is doubtful. Plin. iii, 14 is corrupt: read with C. Müller (on Ptol. ii, 4, 10, p. 124 n.) ‘Contributa Iulia Ugultuniae (cognomen adicitur) cum qua et Curiga nunc est.’ Cod. Leid. has ‘Ugultuniacum quae ….’ This is probably a contemporary note of Pliny's.

^{page 80 note 4} Detlefsen in Ph. 30, p. 300, wrongly refers Ipsca to Contributa Iulia. Cf. Hübner on 1572.

^{page 80 note 5} Plin. iii, 10; 2322; coins prove its former importance.

^{page 80 note 6} 1423. Hübner puts the date one year too late.

^{page 80 note 7} Direct taxes were rare if found at all in municipia; town lands, market tolls, trade licenses and the honoraria of the decurions, magistrates and Augustales were the usual sources of revenue. Cf. Liebenam, Stadteverwaltung, p. 255.

^{page 81 note 1} At Ilipula Minor and at Aeso (1470, 4466) iv viri are recorded who afterwards became II viri: both towns had Flavian charters. But neither inscr. can be certainly dated. It is possible that IV vir may = aedilis, since aediles were once the colleagues of the II viri, the name thus surviving; or that the iv viri = aediles and quaestors.

^{page 81 note 2} 2349, 5510.

^{page 81 note 3} Hübner, pp. 860–1–6. Dessau, ii, p. 515 n. The charter was first issued about 44 B.C.

^{page 81 note 4} 1727)(1313, cf. 1731, 1315)(1314; with Hübner's notes.

^{page 81 note 5} 4616. ‘L. Marcius Gal. Optatus | aedil. Tarracone II vir Ilurone | et II vir quinquennalis primus, | praefectus Asturiae tribun. milit | legionis secundae Augustae, | annor. xxxvi in Phrygia decessit.’ ‘Litterae optimae saeculi secundi,’ Hübner. But Iluro was an old oppidum c. R. an d the claim to have been II vir quinq. primus must refer to a transition from a IV virate to a duoviral system. Hence I would date the inscription from Vespasian's reign or soon after.

^{page 81 note 6} E.E. viii, IO4 (PP. 354 ff.) = 2213 with Hübner's note.

^{page 81 note 7} Tac. H. i, 78; Frontinus, de Controv. Agr. ii, p. 52, Lachmann (a second and third settlement); 490–1, vertarns of Leg. VI; 488–9 E.E. viii, 27 (pp. 354 ff.)—of Flavian date —veterans of Leg. VII. 656 a terminus of Domitian, 70 m. from Emerita: was the terminus of Domitian, 70 m. Emerita: was the territoty continuous? Cf. 5212n.; Dessau, 5972 n.

^{page 82 note 1} Fourteen Spanish alae and 66 cohorts, some of double strength, are recorded in the Roman army lists, nearly all in post-Flavian times.

^{page 82 note 2} Vita Hadriani, xii, 4, ‘omnibus Hispanis Tarraconem in conventum vocatis delectumque ioculanter retractantibus Italicis…’; Vita M. Aurelii, II,7, ‘Hispanis exhaustis Italica adlectione contra [Hadriani?] Traianique praecepta verecunde consuluit’ Cf. Mommsen, Provs. i, 69.

^{page 82 note 3} xi, 6344= Dessau 2693; the date is Hirschfeld's. The occasion of such a command is not recorded, Cf. Domaszewski, Rangordnung, p. 119; p. 59 supra, n. 1.

^{page 82 note 4} Cf. Krascheninnikoff, , ‘Uber die Einführung des provinzialen Kaisercultus im römischen Westen’ in Ph. 53 (1894), pp. 147–189Google Scholar, especially 178–183; Kornemann, s.v. Concilia in P.W. iv, 801–20, and in Klio, i, 95–136, esp. 123–4.

^{page 83 note 1} Tac. Ann. iv, 37; Krasch. 178 ff. Kornemann (Klio, p. 124) thinks it possible to argue from Tac. that the worship began soon after A.D. 25; but he ignores the inscription. In P.W. iv, p. 811, he refers it to Nero. 1663 ’flamen col. immunium provinciae Baeticae’—a worship probably dating from the early empire (Hirschfeld, S. Berl. 1888, p. 840, n.35).

^{page 83 note 2} Kornemann, P.W. iv, 810 ff.; Klio, 117.

^{page 83 note 3} suet. Claudius, 45.

^{page 83 note 4} Krasch. 178 thinks concilia may have existed before cultus; Kornemann, P.W. iv, 804, rightly disagrees. Cf. Schmidt on viii, 12039.

^{page 83 note 5} Plin. Ep. vii, 33, 4; iii, 9, 4 ‘in Classicum tota provincia incubuit.’ Digest, xlvii, 14, 1.

^{page 84 note 1} 5264; 3732 (Valencia) styles Titus during Vespasian's life ‘conservator pacis Aug.’

^{page 84 note 2} 4188–4260; of these 4188 may belong to Nero's reign; 4212, 4217, 4225–6, 4251 are Flavian.

^{page 84 note 3} Plin. Ep. ii, 13, 2; cf. 3866 with Mommsen's note and Plin. Ep. x, 4, 2.

^{page 84 note 4} E.g. 4192, a citizen of Iuliobriga honoured by the province ‘ob causas utilitatesque publicas fideliter et constanter defensas’; 4208, ‘ob legationem censualem gratuitam summo opere gestam’; 4248, ‘ob curam tabulari censualis fideliter administratam.’

^{page 84 note 5} 1570; 5217 with Hübner's n.; 6095, 4212, 3349.

^{page 84 note 6} Cf. Clement of Rome 5 with Lightfoot's n.

^{page 84 note 7} See Bouchier's Spain under the Rom. Emp. c xi, especially p. 174. Tradition makes him land at Gades; if so he may have returned overland through Gaul (II Tim. iv, 10). Tertullian (adv. Judaeos vii) says that Christianity was spread through, all Spain; and Prudentius the Spaniard says there were Spanish martyrs in all the persecutions.

^{page 84 note 8} The first record of the tradition is in S. Aldhelm. of Malmesbury about 700 A.D. Cf. Cathlic Encyclopaedia s.v. ‘Spain.’ It played a great part in the medieval history of Leon: and has caused the invention of inscriptions: 505* of inscriptiones falsae (Iria Flavia) reads: ‘Neroni et Caes. Aug. pont. max. ob provinciam latronibus et is qui novam generis humani superstitionem inculcarunt purgatam’ = 231* (Clunia).

^{page 85 note 1} Cf. Müller on Ptol. ad loc.; Kiepert in Formae Orbis Antiqui shows the alternative boundaries at this point; Spruner-Sieglin follows Pliny.

^{page 85 note 2} Plin. iii, 19, assigns ‘Vettones’ to Citerior; iv, 112, seems to confirm this ‘disterminatis ab Asturia Vettonibus, a Lusitania Callaecis’ (of the R. Douro); iv, 116, places ‘circa Tagum Vettones’ in Lusitania. Cf. Ptol. ii, 5, 7.

^{page 85 note 3} Cf. Kornemann, ‘die Entstehung der Prov. Lusitanien’ in Festschrift Hirschfeld, pp. 221 ff.

^{page 85 note 4} Strabo, iii, 4, 20; cf. ib. 1, 6; 3, 2 and 3; 4 12.

^{page 85 note 5} Front. de Agror. Qual. p. 4, 4, Lachm.; 485 (Hadrian's reign), 1178 (A.D.150), 484(third cent.).

^{page 85 note 6} If the Ocelenses of Plin. iv, 118 = Ptolemy's “οκελον = Ocelodurum (Zamora on the Douro) this view of a northern and southern district is untenable; but Ptolemy's “οκελον was 42 m. south of Salmantica, and therefore in the southern district. The name was a common one, and the epithetdurum or Duri may have been used for clearness’ sake; cf. Ptol. ii, 6, 22, for another “οκελον. Detlefsen first raised this question of boundary (Ph. 32, p. 666) but thinks the complete transfer of the Vettones was after A.D. 100, because of the title of the ‘Ala Hispanorum Vettonum C. R.’ recorded in A.D. 103. But ‘Hispanus’ could be used of all Spain: e.g. vii, 52; in any case the name might be an historical relic. The map-makers ignore the difficulty, assigning all Vettonia to Lusitania, and differing only about ths province of Ocelum Duri.

^{page 85 note 7} Caes. B.C. i, 38, 1, ‘Vettonum ager Lusitaniaque’ under Varro.

^{page 86 note 1} Verwaltungsb.² p. 377 and n. 3.

^{page 86 note 2} iii, 1919; cf. ib. p. 1030.

^{page 86 note 3} v, 534; ii, 1855 = Dessau 1379-80. ‘Provincia’ does not prove a completely independent government; cf. the contemporary case of the Hellespont. Hirschfeld seems to bracket these two procc. with Arruntius Maximus (l.c. ‘anscheinend unter Vespasian und wohl mindestens bis auf Traian’) and is followed by Mispoulet (Rev. de Philologie 34, pp. 313 ff.) who thinks that the proc. was presently replaced by the legatus iuridicus (p. 315). He argues also that a financial proc. Asturiae could not exist alongside of the proc. Hisp. citerioris, who controlled the whole province.

^{page 86 note 4} 4204 (Aquae Flaviae), 4215 (Limici), 4247 (Abobriga), 4248 (Bergidum Flavium), 4257 (Bracara), 6094 (Brigaecium)—all records of flamens at Tarraco.

^{page 86 note 5} 3271, 4616, quoted supra, p. 80, n. 2, and p. 81, n. 5.

^{page 87 note 1} Strabo, iii, 4, 20, εἰσί δὲ καὶ έπίτροποι τοῦ Καίσαρος (in Citerior) ίππκοὶ ἄνδρες, οἰ διανέμοντες τὰ χρήματυ τοῖς ατρατιώταις εἰς τὴν διοίκησιν τοὗ βίον. For this purpose one would sitll be needed for the north-west. 2661.

^{page 87 note 2} xiv, 3995; viii, 2747. Domaszewski, (Rhein. Mus. xlv, 1890, p. 10Google Scholar) assigns the institution to Domitian, supposing withdrawal of one leg. legionis in 88; cf. Kornemann, P.W. s.v. ‘Dioecesis,’ v, 719 ff. Largius (p. 55, n. 1) may have been the first iuridicus. Cf. contemporary praetorii in Galatia-Cappadocia.

^{page 87 note 3} Cf. pp. 73–4, 78. Detlefsen, Ph. 32, 606–8.

^{page 88 note 1} Mommsen, Provs. i, p. 73—he agrees that the communes were non-urban; cf. Plin. iii, 138 for Cisalpine Gaul.

^{page 88 note 2} At an earlier stage of the conquest the Roman policy was different, many towns being broken up into villages. Strabo, iii, 3, 5 … κώμας ποιήσαντες τὰς πόλεις αὐτῶν τὰς πλείστας ἐνίας δε καὶ συνοικίζοντες βέλτιον. Cf. also Appian, Iberica, 44, οί δὲ ἒλεγον άπηγορεῦσθαι Κελτίβηρσιν ὑπὸ Τράκχον μὴ κτίζειν πόλεις

^{page 88 note 3} viii, 3226, 3268 prove Galeria for Lucus (Kubitschek, Hübner). Abobrica was ‘oppidum insigne’ according to one reading of Pliny; in Galeria 4247, E.E. ix, 390.

^{page 89 note 1} Kornemann, Klio, i, p. 117, 119–21, gives all the inscriptions. Cf. p. 82.

^{page 89 note 2} E.E. viii, 131 (pp. 354 ff.) ‘litt. saeculi circ. I exeuntis,’ Hübner ad. loc.; Cichorius in P.W. s.v. ‘Cohors.’

^{page 89 note 3} de Longaevis, F.H:G. iii, p. 608 ff.

^{page 89 note 4} Sillig's text gives a total of 575,000, but the proportion to the civitates of the three conventus favours the variant.

^{page 90 note 1} 4799 ff., 4814 with Hübner's notes. Itin. Ant. p. 428.

^{page 90 note 2} Hübner on 2477. A Roman bridge still stands there, but as 2478 records that Aquae Flaviae ‘pontem lapideum de suo f. c.’ in A.D. 104, perhaps some other work may be referred to.

^{page 90 note 3} Hübner on 4697. Strabo iii, 4, 9. Itin. Ant. p. 403. The inscription quoted by Mr. Bouchier, op. cit. p. 134, recording the completion by Domitian of a road ordered by Vespasian but neglected ‘nequitia publicanorum’ cannot be accepted as genuine. Cf. p. 44*, no. 446*, also 444*, 445*, 446a* (inscriptions falsae).

^{page 90 note 4} Dessau 233, 8902, referring to the same person, who offers vows for the safety of Nero and Poppaea.

^{page 90 note 5} Plin. iii, 18, 25, cf. 76, where the distances given refer to the journeys of litigants to Carthago Nova; ib. viii, 218—the islanders petitioned Augustus for military aid against a plague of rabbits. There is no other record of a garrison.

^{page 91 note 1} Tac. Ann. xiii, 43Google Scholar; Suet. Galba 10. One earlier exile is recorded, in Tiberius' reign: cf. Tac. Ann. iv, 43Google Scholar, with Furneaux' n.

^{page 91 note 2} Tac. H. i, 77, 90Google Scholar; ii, 92, for the action of Galba, Otho, and Vitellius. Titus banished informers ‘in asperrimas insularum’ (Suet. 8) but hardly hither; Domitian, too, found room elsewhere.

^{page 91 note 3} Pliny's ‘foederatum’ as applied to Ebusus may, as Detlefsen thinks, refer rather to a union with Colubraria separated by a narrow strait.

^{page 91 note 4} Mommsen on x, 6785, is not clear; he groups him with the praefecti orae, who were not independent, and with the other island governors, of whom at least some were; viewing him apparently as a permanent officer. Domaszewski (Rangordnung, p. 137) says ‘der praef. insularum Balearum [as well as the other praefecti orae] wird die Strand-polizei ausgeübt haben.’ Cf. p. 60 supra, with notes; But Hirschfeld, Verwaltungsb.² p. 390 and n. 2, refers him to the exiles, ‘in the first century of the empire.’ He views him merely as a ‘Stellvertreter’ of the legatus and not as independent. But there is no parallel for such a use of ‘pro legato.’

^{page 91 note 5} 4197, 4205 Palma, 4218 Guiuntum.

^{page 91 note 6} Suet. Vesp. 16. Plin. Pan. 37 applies the word to the ‘vicesima,’ which was strictly a vectigal.

^{page 92 note 1} Plin. Pan. 37 ff.; cf. 39, ‘Laeti ergo adite honores, capessite civitatem.’

^{page 92 note 2} 964, 1949, 2214, 4184, etc.: Hirschfeld, Verwaltungsb.² p. 102.

^{page 92 note 3} 5064 with Mommsen's note. Hirschfeld op. cit. p. 79.

^{page 92 note 4} v, 7209 viii, 12656: cf. Rostowzew, Staatspacht, p. 400, and Hirschfeld op. cit. p. 83 and notes.

^{page 92 note 5} Suet. Galba 12; Tac. H. i, 51, 53, 65Google Scholar.

^{page 92 note 6} Ulpian. de censibus Dig. L. 15, 4 classifies land thus for census purposes: arvum, vinea, olivae, pratum, pascua, silvae; cf. Hyginus (p. 205 in Römische Feldmesser) ‘certa enim pretia agris constituta sunt, ut in Pannonia arvi primi, arvi secundi, prati, silvae glandifcrae, silvae vulgaris, pascuae.’ The Flavian policy conerning Italian subseciva is in point.

^{page 93 note 1} Phlegon of Tralles op. cit. refers to centenarians in Lusitania.

^{page 93 note 2} Sir W. Ramsay, Bearing of Recent Discovery on the Trustworthiness of the N. T. p. 259 f. and p. 273 f.

^{page 93 note 3} vi, 1463: Unger, de Censibus, p. 28.

^{page 93 note 4} Cf. p. 84, n. 4 supra. 4208 at least must refer to the imperial census; and even if 4248 refers primarily to a municipal census (which I do not think likely) the office must have been co-ordinated with the imperial service.

^{page 93 note 5} Apuleius, Apol. 101 proves this for a quaestor of Oea in Africa. Cf. Mommsen, Droit, vi, 1, p. 390, n. 3; Hirschfeld, Verwaltungsb.² p. 74 and n. 6. Mommsen thinks that the tax was directly levied in the imperial provinces, but he cites no evidence for this distinction, which seems improbable.

^{page 93 note 6} Papinian, Dig. L. 1, 17, 7 (about A.D. 200) ‘exigendi tributi inter sordida munera non habetur et ideo decurionibus quoque mandatur.’ Cic. in Verrem, iii 84, 88, 99, indicates perhaps the first beginnings of the system. Decuriones had to possess a certain valuation: e.g. HS 100,000 at Comum.

^{page 93 note 7} Yet A.E. 1902, 164 (Gigthis) seems to show that ‘maius Latium’ was not readily granted, two successive embassies being required to secure it. Hirschfeld wrote before the discovery of this inscr

^{page 93 note 8} Only two instances are traceable in the rest of the empire—Stobi and Coela (Mommsen, Droit, vi, 2, p. 458, n. 1).

^{page 94 note 1} Plin. xv, 1, 17, 18, xvii, 31, for oil; xiv, 71, Silius iii, 369, Martial xiii, 118, for wine; Plin. xviii, 95, xvii, 94, for wheat: xix, 10, Silius iii, 373, for flax, and linen from Tarraco and Saetabis; Plin. xix, 30, for esparto grass, with which the campus Spartarius, near New Carthage, supplied the whole world.

^{page 94 note 2} ‘Theldones’ and the smaller ‘asturcones’ Plin. vii, 166; Silius, iii, 335; Martial, xiv, 199. Cf. the tale of the Lusitanian breed whose sire was the west wind.

^{page 94 note 3} xvi, 32, coccum, or cusculium, ‘pensionem alteram tributi pauperibus Hispaniae donat’; cf. Strabo iii, 2, 6, for the great export trade in this from Baetica in his day.

^{page 94 note 4} xv, p. 562. Amphorae recovered date from A.D. 140 to 255. Cf. Hirschfeld op. cit. p. 141; 1180 (Hispalis) ‘adiutor praef. annon(ae) ad oleum Afrum et Hispanum recensendum…. item vecturas naviculariis exsolvendas’ (after A.D. 160); also vi, 29809.

^{page 94 note 5} Plin. XV, 91. ‘Flaccus Pompeius eques Romanus qui cum eo [Vitellio] militabat’ must have been a proc. Cf. 2029, ‘proc. ad Falernas vegetandas’ (saec. 2).

^{page 94 note 6} So it is said nowadays that the oil of Persia may ultimately repay our national debt, and that the exploitation of the Ural and Altai mountains would similarly clear Russia.

^{page 94 note 7} Strabo, iii, 2, 9, from Poseidonius, ‘ καθόλου δʹ ἂν εἷπε, ίδών τις τοὑς τόπους, θησαυροὑς εἷναι ϕύσεως άενάους ἢ ταμιεῖον ἡγεμονίας άνέκλειπτον ’ Cf. Ezekiel, xxvii, 12, for the Tyrian trade with Tarshish. Plin. iii, 30, iv, 112.

^{page 94 note 8} Plin. xxxiii, 118, ‘celeberrimo Sisaponensi regione in Baetica miniario metallo vectigalibus populi Romani.’ Cf. Hirschfeld, op. cit. p. 149.

^{page 95 note 1} Dio, 52, 28, 4: Maecenas puts first among the resources of the exchequer ὅσα ἔκ τε μεταλλείας καὶ εἰ δή ποθεν ἄλλοθεν βεβαίως δύναται προσιέναι. Strabo, iii, 2, 10; Suet. Tib. 49; Hirschfeld, op. cit. pp. 146–150 with notes. 5132 (Lusitania), ‘Ti. Claudius Rufus Iovi O.M. ob reperta auri p. cxxii, v.l.s.’ seems to be an example of private enterprise in the latter half of the first century.

^{page 95 note 2} For the first fragment, cf. pp. 793–801, with Hübner's and Mommsen's commentary; Hirschfeld, op. cit. p. 159 with notes; Ardaillon in D.S. s.v. ‘Metalla.’ For the second, Cagnat in Journ. des Savants, 1906, p. 442; A.E. 1907, 151; Cuq in Mélanges Gérardin, 1907, pp. 86–133; Mispoulet in Nouv. Revue Historique de Droit Fr. et Etr. 1907, 345 ff., 491 ff.; Rostowzew, Gesch. des röm. Kolonates, pp. 353–362.

^{page 95 note 3} Cuq retains the earlier dating of the first fragment; Rostowzew seems to incline the same way, but leaves the p int uncertain.

^{page 95 note 4} Rostowzew, op. cit 320–5, and especially 355.

^{page 96 note 1} Of the word ‘vicus’ in the law only the first letter remains.

^{page 96 note 2} On the same principle whereby half of treasure trove was claimed by the government.

^{page 96 note 3} Diod. v, 35–38, esp. 36, 3–4 and 38, 1; he speaks of silver mines chiefly. Mommsen, R.H. iii, p. 308, ed. 1894.

^{page 96 note 4} Tabella x in iii, p. 948; cf. nos. ix and xi, from Vicus Pirustarum (Verespatak) in Dacia, A.D. 164.

^{page 97 note 1} Rostowzew, op. cit. p. 360–1. Cuq, p. 126 ff. The system obviously would not suit all mines.

^{page 97 note 2} Cf. several instances to be mentioned infra. The silver-lead mines at Coto Fortuna in Murcia (Mons Argentarius) still show a Roman gallery over three miles long. See infra, p. 100, n. 6.

^{page 97 note 3} Mispoulet, op. cit. p. 491 ff.; Rostowzew, op. cit. p. 360–1.

^{page 97 note 4} Cuq, op. cit. pp. 97, 111. Cf. Floras, ii, 25, of the Dalmatians, ‘Augustus perdomandos Vibio mandat, qui efferum genus terras fodere coegit aurumque venis repurgare; quod alioquin gens omnium cupidissima eo studio, ea diligentia anquirit, ut illud in usus suos eruere videatur.’ See xiii, 1550, for silver mines in Aquitania worked by imperial slaves under Tiberius. When this paper was read before the Classical Society of Trin. Coll. Dublin, Prof. Alison Phillips expressed a doubt as to whether the existence of excellent Roman laws securing good conditions for the workers in mines could be taken as proof that these conditions were in fact good. He pointed out that in some of the modern Latin-American states excellent laws for safeguarding the rights and liberties of the peons are on the statute books, but that these do not prevent the virtual enslavement almost everywhere of the workers under the ‘peonage system,’ some of the worst abuses of which were exposed during the enquiry into the Putumayo atrocities. This caveat should be recorded, though the extant evidence for Spain under the early empire points rather in the opposite direction, and it may not be quite safe to argue from Latin-America to ancient Rome. Ammianus 31, 6, 6, records that ‘sequendarum auri venarum periti non pauci’ welcomed a barbarian inroad, ‘vectigalium perferre posse non sufficientes sarcinas graves,’ but Thrace in the fourth century hardly justifies an inference for Spain in the first.

^{page 97 note 5} Apud Strab. iii, 2, 10.

^{page 97 note 6} Plin. xxxiii, 78; Silius, 1, 231–3; Statius, Silv. iv, 7, 13, imitates the conceit.

^{page 98 note 1} Silvae, iii, 3, 89.

^{page 98 note 2} Plin. xxxiii, 80; 2598 ‘proc. (libertus) metall. Alboc.’ Albocela in Ptol. ii, 6, 49, is different, being a municipium.

^{page 98 note 3} Plin. xxxiv, 156. Haverfield in Mélanges Boissier, p. 249 ff.

^{page 98 note 4} Plin. xxxiv, 163, ‘India nec aes neque plumbum habet gemmisque ac margaritis suis haec permutat’: Confirmed by the Periplus Maris Erythraei, 7, 28, 49, 56.

^{page 98 note 5} Besnier, s.v. ‘Stannum’ in D.S. p. 1459B, with reff. Bérard, les Phéniciens et l'Odyssée i, p. 444–5.

^{page 98 note 6} Poseidonius apud Strab. iii, 2, 9, ὲξαννθεῖν ϕησιν τὴν γῆν άργυρίῳ, καττιτέρῳ, χρυσίῳ λενκῷ, άργυυρομιγὲς γάρ έστι

^{page 98 note 7} 2423.

^{page 98 note 8} Dio. 37, 53; cf. Plut. Caes. 12.

^{page 99 note 1} Itin. Ant. p. 423–5.

^{page 99 note 2} Orosius 2, 3. 2559 with Hübner's n. Strabo, iii, 1, 9. Mela's ‘turris Augusti titulo memorabilis’ (iii, 11) cannot be identified with the turris Herculis, being apparently on a different site.

^{page 99 note 3} E.E. viii, 307–8, cf. 309; N. Dig.

^{page 99 note 4} xi, 395 = Dessau 2648, ‘donis donato ob resprosper(e) gest(as) contra Astures ….’ ‘Milites are expressly made free of the baths by the law of Vipasca. Cf. Hirschfeld, op. cit. pp. 160–1 and especially 172–3 with notes.

^{page 99 note 5} 5752, among the northern Astures, ‘domo Flaviobrigensis’ but the reading is uncertain.

^{page 100 note 1} Plin. xxxiv, 149, 144; E.E. vii, 1212; Martial, i, 49, iv, 55. Cf. Plutarch, de Garrulitate, 17, for a contemporary ref. to Celtiberian iron-work. Besides Flaviobriga, Pliny mentions on the north coast Portus Victoriae Iuliobrigensium (Santona?) Portus Blendium (Santander?) and Portus Vereasueca as already existing.

^{page 100 note 2} Plin. xxxiii, 96–7 Diodorus and Strabo (Poseidonius) speak chiefly of the older silver mines; Plin. xxxiv, 158, ‘Non fit in Gallaecia nigrum (plumbum) cum vicina Cantabria nigro tantum abundet.’

^{page 100 note 3} Plin. xxxiv, 164. These are the usual identifications, but they are uncertain. The name Ovetum may well represent the modern Oviedo, and it is not surprising that a mining ‘vicus’ should be unrecorded by the geographers: but the reading is only conjectural, Pliny's MSS. having Iovetanum or the like: also the neighbourhood of Oviedo was rather a tin district. And there does not seem to be any other evidence of lead from Capraria or Oleastrum.

^{page 100 note 4} Plin. xxxiv, 165. The numerals are confused in the MSS.; another reading makes the Santarensian rise from 45,000 to 200,000 denarii (Besnier, s.v. Plumbum, in D.S.) Polybius apud Strab. iii, 2, 10, says that the silver mines of New Carthage yielded a daily revenue (net?) of 25,000 drachmae to the Roman people.

^{page 100 note 5} Plin. xxxiv, 164, ‘Nigro plumbo … laboriosius in Hispania eruto … sed in Britannia summo terrae corio adeo large ut lex ultro dicatur ne plus certo modo fiat.’

^{page 100 note 6} The ‘silver mountain’ mentioned by Strabo, iii, 2, 11, yielded much lead : one pig found at Rome and five at Coto Fortuna in Murcia are inscribed ‘Societ. Mont. Argent. Ilucro.’ xv, 7916 (prob. saec. 1), A.E. 1907, 135. [This does not imply private ownership.] Yet in some Spanish mines the silver did not repay the cost of refining from the lead, Strabo, iii, 2, 10.

^{page 101 note 1} Plin. xxxiv, 4, ‘Summa gloriae nunc in Marianum conversa, quod et Cordubense dicitur.’ The context implies a copious output. ‘Aurichalci bonitatem imitatur in sestertiis dupondiariisque, Cyprio suo assibus contentis.’

^{page 101 note 2} Diod. v, 36, 2; Strabo, iii, 2, 9 (after Poseidonius).

^{page 101 note 3} 1179: Hübner strangely says ‘litteris optimis aevi Antoniniani.’ xiv, 52 (Ostia) ‘proc. massae Marianae.’ Hirschfeld, op. cit. p. 159 and n. 1.

^{page 101 note 4} 956; Hirschfeld l. c.

^{page 101 note 5} This may be inferred from the ‘et’ in l. 1 of 3271, quoted supra, p. 80, n. 2.

^{page 101 note 6} D.S. s.v. ‘Argentum Oscense’; Head, Hist. Numorum ² p. 3–5.

^{page 101 note 7} 4206; Hirschfeld, op. cit. p. 182 and n. 2. For the, preceding statements I am endebted to an article by Mr. Mattingly in Num. Chron. 1914, p. 110 ff.; which he has very kindly supplemented by letter. Cf. also his article, ‘Mints of the Early Empire,’ J.R.S. vii, 59, ff.; on page 68 he seems to consider that a branch of the Roman mint remained at Lugdunum.

^{page 102 note 1} Plin. xxxiii, 145; Plut. Galba 4, τᾧν ἀλιτηρίων ἐπιτρὀπων; cf. Suet. Galba 9 and 12.

^{page 102 note 2} Suet. Vesp. 16.

^{page 102 note 3} Suet. l.c.; about £400,000,000, ‘ut res publica stare posset.’ Cf. Tac. H. iv, 47Google Scholar, ‘actum in senatu ut sescentiens sestertium a privatis mutuum acciperetur’.

^{page 102 note 4} Plin. E. ii, 13, 4.

^{page 102 note 5} So Haverfield, s.v. Spain in Encyc. Brit. ¹¹

^{page 102 note 6} i, 63–6.