¹ Cf. Detlefsen, , ‘Die Anordnung der geogr. Bücher des Plinius u. ihre Quellen’ (Quellen u. Forschungen zur alten Geographie xviii, 1909)Google Scholar, summarising his earlier works (same periodical, vol. xiv; Hermes xxi; Comment, phil. in hon. Th. Mommseni, Berlin, 1877, etc.)Google Scholar; Klotz, , in Quellen … alten Geographie xi (1906)Google Scholar; Kornemann, ‘Die Entstehung der Provinz Lusitanien’ (Beiträge zur Alten Geschichte. Festschrift zu Otto Hirschfelds, 1903, 221 ff.); Rabenhorst, , Quellenstudien zur NH des P (Inaug. Diss. Berlin 1905)Google Scholar; Münzer, , Beitrage zur Quellenkritik der NH des P (Berlin, 1897)Google Scholar; Cuntz, , in Jahrbücher für class. Philol., Supplbd. xvii, 1890Google Scholar; De Augusto Plinii Geogr. auctore (Inaug. Diss. Bonn 1888)Google Scholar; Reitzenstein in Hermes xx; Jullian, in Mélanges de l'École de Rome iii (1883)Google Scholar; Lessert, Pallu de in Mém. Soc. Ant. de France lxvii (1908)Google Scholar; Albertini, Les Divisions administratives de l'Espagne Romaine (1923).

Editions : Detlefsen, , NH ii, 242–vi (vol. ix of Quellen… alten Geographie 1904)Google Scholar; Mayhoff (Teubner), 1906.

² Les divisions administratives de l'Espagne Romaine 56–9.

³ Syme, R., AJPhil. lv, 300Google Scholar, suggests A.D. 9, but prudently adds : ‘It may be earlier.’ It must be, since Pliny's formula, which is subsequent to the transference, dated before A.D. 6. I regret to say I have some responsibility for Mr. Syme's suggestion.

⁴ I agree with Mr. C. H. V. Sutherland (The Romans in Spain, 240) that ILS 103 is a monument of this change of frontier.

⁵ It. Ant. 407, iii, gives ‘Belone Claudiä’. There is no other evidence for a colony at any date.

⁶ ‘Vespasian's Reconstruction of Spain,’ in JRS viii, 69, n. 4. Of McElderry's list of five possible cases, Iulipa is not proven (CIL ii, 2362), and Burguillos is quite uncertain because its ancient name is unknown (CIL ii, 5354). To the certain cases of Carmo, Epora, and Illiturgicola, however, Acinipo should be added (CIL ii, p. 701, which outweighs the doubtful reading ‘Quir’ in CIL ii, 1348). Cf. Kubitschek, Imperium Romanum tributim discriptum.

⁷ Op. cit. ch. iv.

⁸ iii, 166–7.

⁹ NH iv, 110.

¹⁰ Notes on the Coinage of Hispania Citerior 75, n. 20.

¹¹ iii, 195.

¹² Martial x, 103, xii, 21, refers to citizens of Bilbilis as municipes, but the term may mean only ‘fellow-townsmen’, and cannot be pressed.

¹³ Plut. Galba 6 : Suet. Galba 9 : Eckhel, , Doctrina num. vet. i, 46Google Scholar; vi, 294 : CIL ii, 2779.

¹⁴ NH iii, 28; cf. McElderry, op. cit. 89.

¹⁵ McElderry, op. cit. 74, n. 3. He gives seventeen towns, of which two are not proven.

¹⁶ See above, p. 2.

¹⁷ CIL ii, 4223, 2415, 2638 : Klio 1 (1901), 51Google Scholar.

¹⁸ iii, 166–7.

¹⁹ Kornemann in P-W, s.v. ‘Colonia’, 519: Sherwin-White, The Roman Citizenship 176–7. Mr. Sherwin-White understands Dio xliii, 39V, correctly, and leans towards the right answer, but since he has not perceived the evidence in Pliny he is forced to leave the question undecided.

²⁰ xliii, 39V.

²¹ See below, p. 10.

²² Op. cit., loc. cit.

²³ NH iii, 14 (cf. iv, 118). Pliny found the cognomina a convenient means of avoiding the repetition of the names, but does not suggest that they were invented merely in order to distinguish the towns from each other.

²⁴ Except Ilipula, which occurs three times in Baetica. But only one of the three has a cognomen.

²⁵ Pliny's spelling. Strabo iii, 140, gives Νάβρισσα, and CIL ii, 1294, ‘Conoba.’ …

²⁶ Codices: Iulia Fidentia; but it must be Ulia (so Detlefsen, Hübner, Mayhoff).

²⁷ Codices: Illiberri Liberini: ‘Florentini’ from CIL ii, 1572, and others.

²⁸ Codices: ‘Ugultuniacum quae et Curiga nunc est.’ Mayhoff, after Müller, emends ‘Ugultuniae cum qua’. This gives the required dative case, and accounts for the feminine relative pronoun. Pliny makes his relative pronouns agree with what he takes to be the gender of the town name: e.g. ‘Segida quae Augurina, Vesci quod Faventia.’ They are no guide to the question whether the town was a municipium or a colonia. But, if the Spaniards hereabouts knew their own language (which is not certain : cf. Strabo iii, 151), the gender of the town-name will not always account for the gender of the colonial title.

²⁹ Codices : Laeparelia : a dittograph from the following ‘Carisa Aurelia’. Mayhoff emends ‘Regia’, but this may not have been the real title.

³⁰ Ptolemy (ii, 5, viii) cites Metellinum with a fine confusion of genders as Καικιλία Γεμλλινον ἣ Μετίλλινα, which may preserve the two stages of its history : (a) as a praesidium settled by Metellus during the Sertorian wars, and (b) as a full Augustan colony.

³¹ NH iv, 119. But the whole problem of the twin foundations of Gades cannot be discussed in this paper.

³² It was an important place on the Guadalquivir, and early adopted Roman ways : cf. the interesting Republican or early imperial record of the building of an arch and gates by Urchail Atitta F. Chilasurgun (CIL ii, 1087).

³³ Ὄνοβα Αἰστούρια in Ptolemy ii, 4 v : Ὄνοβα Αἰστούρια in Marcian. Heracl. Periplus maris ext. ii, 9, p. 546 (Müller).

³⁴ CIL ii,, 2105, 2114.

³⁵ Ibid. 2081.

³⁶ Ibid. 2121.

³⁷ Ibid. 2126, 2129–2132.

³⁸ Ibid. 2188, cf. 2186.

³⁹ Ibid. 1536–7.

⁴⁰ Ibid. 5363.

⁴¹ Ibid. 1029.

⁴² Ibid. 972 and Brarnbach 1150, 1151, 1160 (no Ilvir, but four citizens in tribus Galeria).

⁴³ Ibid. 727–9 et al.

⁴⁴ Ibid. 1256.

⁴⁵ McElderry, op. cit. 70–3.

⁴⁶ Cf. above, p. 2, n. 6.

⁴⁷ Detlefsen, , Philologus 30 (1870), 265CrossRefGoogle Scholar, had noted what he called the double Roman and native names of the towns in question, and had conjectured that they implied a municipal status. He did not, however, date the grants or confirm his hypothesis by inscriptions, but only attempted to support it from Pliny's grouping, which is too inconsistent to be convincing. Nor did it occur to him that the ‘Roman names’ were colonial titles. His list differs in some particulars from mine, chiefly owing to differences in the reading of Pliny's text, which he had not edited at the date of the article.

⁴⁸ Bell. Hisp. iii, 3.

⁴⁹ NH iv, 117.

⁵⁰ NH iii, 18–30. Mommsen's emendation (codices ‘Caesarivenales’). (Sutherland, op. cit. 116, asserts that Castulo and Saetabis had Latium under the Republic, apparently because of Pliny's description ‘Latini veteres’, which implies no such thing.) Besides the four towns mentioned, the Segisamaiulienses seem to have a title, but they are not mentioned as Latin—perhaps because Pliny gives no details of the towns of the conventus Cluniensis, where Segisama was. But I doubt a Julian grant, on topographical grounds.

⁵¹ Cf. above p. 1 and Albertini, op. cit. 37.

⁵² Hill, op. cit. and pl. xi.

⁵³ Livy xxxiv, 9 : Pliny, NH iii, 22Google Scholar : Hill, op. cit. 12. Livy describes the post-Republican history of Emporiae in three stages: (1) Caesar settled some coloni Romani there; (2) afterwards the Spaniards were incorporated in a Roman municipium; (3) finally the Greeks joined in this municipium. Now the coins suggest that the Roman municipium began at stage (2) — i.e. the Spaniards were in at its first foundation—because the word municipium occurs in a mixture of the Latin and Iberian alphabets : a mistake which Roman colonists would not make. I think it most likely that Caesar left the colonists simply as an informal conventus civium Romanorum and that Augustus created the municipium of Romans and Spaniards, adding the Greek population later.

⁵⁴ McElderry, op. cit. 74, n. 3, and cf. pp. 3 and 11, nn. 15 and 62.

⁵⁵ Cf. above p. 2, n. 6.

⁵⁶ Kornemann, P-W s.v. ‘Colonia’, 517–18 : Jullian, , Histoire de la Gaule iv, 343 ffGoogle Scholar.

⁵⁷ Incidentally, my findings show that Caesar, like Augustus, used the tribus Galeria for Spanish enrolments.

₅₈ ILS 6092.

⁵⁹ Cf. above, n. 53.

⁶⁰ Including Castulo : cf. above, p. 9.

⁶¹ E.g. Asido, and others perhaps (cf. below, n. 68).

⁶² In Baetica, of 37 Roman and Latin municipia, only 15 are independently confirmed by inscriptions. If this proportion held in Citerior, the–15 Augustan grants known only from inscriptions would represent a total of 37; which would bring Augustus's total in Citerior to 64, and in all Spain to 78. This is mathematics, not history; but it is certain that many Augustan grants remain unknown, whereas Caesar's grants cannot much exceed the number I have given.

⁶³ Caesar, Bell. Civ. i, 61Google Scholar; ii, 18.

⁶⁴ iii, 151. Strabo refers to the Guadalquivir region. Elsewhere, Latin was not necessarily the popular tongue in the Julian and Augustan municipia, but was understood and spoken in them.

⁶⁵ Pliny, NH iii, 14Google Scholar.

⁶⁶ Hill, op. cit. 44.

⁶⁷ ILS 8888. The turma Salluitana consists of men from the Ebro and Segre regions and thereabouts, whatever view is taken of the meaning of Salluitana (cf. Cichorius, Röm. Stud. (1922) 133; and for confirmation of the older view that the epithet is local or tribal, Gomez Moreno, ‘Sobre los Iberos y su lengua,’ in Homenaje a Menéndez Pidal (1925) iii, 486).

⁶⁸ See below Note (1).

⁶⁹ See below Note (2).

⁷⁰ Carthago Nova, Tarraco, Celsa (colonies); Dertosa and ?Emporiae.

⁷¹ See below Note (3).