^{page 127 note 1} J.R.S. xi, p. 261.

^{page 128 note 1} De mort. pers. c. 7.

^{page 128 note 2} Gesch. des Untergangs der ant. Welt i, 8; cp. 412.

^{page 128 note 3} In his Opusculi ecclesiastici (Trent), 1742; reprinted in his Opere, vol. xi (Venice), 1790.

^{page 128 note 4} ‘Verzeichniss der römischen Provinzen, aufgesetzt um 297,’ in Abb. d. k. Akad. d. Wiss. zu Berlin, 1862, pp. 489Google Scholarsqq. Reprinted in his Gesammelte Schriften, vol. v.

^{page 128 note 5} In Geograph. Minores, 1878.

^{page 128 note 6} In his ed. of Notitia Dignitatum (1876), pp 247, sqq.

^{page 128 note 7} Particularly Mr.Kuhn, , in N. Jahrb. f. Philologie und Pädagogik, Bd. 115, 1877. See also the article Diöcesis in P.W.Google Scholar

^{page 130 note 1} Mr. Camille Jullian said that the list could only have been drawn up between A.D. 292 and 297. See his learned article in Revue historique xix, p. 331 (1882)Google Scholar.

^{page 130 note 2} Eutropius, 9, 23: Ea lamen occasione ordinauit provide multa et disposuit quae ad nostram aetatem manent. Cp. Seeck, Gesch. des Untergangs der antiken Welt, i, 421.

^{page 130 note 3} Pap. Oxyrh. vi, 896; Archiv. f. Pap.-F. iii, 340.

^{page 130 note 4} See M. Gelzer, Studien zur byz. Verw. Aegyptens, p. 4.

^{page 130 note 5} See Pap. Oxyrh. i, 87, Cod. Theod. 12, 1, 34 and other texts cited by Gelzer, ib.

^{page 131 note 1} Op. cit.

^{page 131 note 2} Die römische Provinz-Liste von 297. Teil i. Duisburg, 1889.

^{page 131 note 3} Ep. 478 (Wolf) = 563 (Förster). For the dates of the letters see Seeck, , Die Briefe des Libanius zeitlich geordnet (1906). Cp. p. 111Google Scholar.

^{page 131 note 4} Ep. 324 W. = 321 F.

^{page 131 note 5} Ep. 318 W. = 315 F.

^{page 132 note 1} Epp. 337, 338 W. 334, 335 F. Eupaterius was the first praeses of Pal. ii.

^{page 132 note 2} Turner, Eccl. Occ. Mon. iuris ant., i, 1, p. 46.

^{page 132 note 3} This agrees with Ammian's account (xiv, 8) of the Orient Diocese, so far as it goes, but he does not mention the second Palestine.

^{page 132 note 4} Op. cit. p. 361. He refers to Vopiscus, V. Probi. 16, 4–17, 1.

^{page 133 note 1} This division cannot have been earlier than A.D. 293, the year in which the Caesars were created, as is shown by the inscription of Aurclius Priscianus v.p. praeses provinciae Ponti, C.I.L. iii, 307.

^{page 133 note 2} Turner, Eccles. Occ. Mon. iuris ant. i, 1. p. 6.sq. It is also mentioned by Himerius, Or. i, § 14 (ὀ Διὸς καλούμϵνος Πόντος).

^{page 133 note 3} C.I.L. iii, 14184³⁴.

^{page 133 note 4} C.I.L. iii, 14184³⁷.

^{page 133 note 5} Maurice, , Numismatique Constantinienne, vol. i, p. cxxx, cxlviii.Google Scholar

^{page 133 note 6} See Basil's letter of that year to Eusebius, bishop of Samosata, Epp. 138, 8 (ed. Benedict. 1839. vol. iii, p. 332).

^{page 133 note 7} Cp. Vopiscus, V. Probi, 16, 3.

^{page 134 note 1} De C. Caelii Saturnini titulo, in Nuove memorie dell' Istituto di Corrispondenza archeologica, 298 sqq. (Leipzig, 1865)Google Scholar. In the two texts cited by Mommsen from Anon. Vales. Pars 1, viz. 5 § 18 and § 21, it is not at all clear that Moesia means the diocese; in the first the natural meaning is Moesia inferior.

^{page 134 note 2} Mommsen, op. cit. p. 306. For comites = vicarii, cp. Cod. Theod. 11, 34, 1 (A.D. 331), etc.

^{page 134 note 3} Brev. c. 8.

^{page 134 note 4} Cp. Vopiscus, Vit. Aurel. 39; Eutropius, ix, 15.

^{page 134 note 5} ‘Die Teilung des aurelianischen Dakiens’ in Klio (1912), pp. 12, 234Google Scholarsqq.. The statement of Festus Brev. c. 8) is per Aurelianum duae Daciae factae sunt. In the list of Illyric provinces which follows, Förster has rightly inserted <Dardaniam> after Moesiam. This is guaranteed by the text of Jordanes, Rom. 218, who copied from Festus.

^{page 135 note 1} This (not 346) is the date of the Council of Sardica(see Schwartz, E., ‘Zur Gesch. des Athanasios’ in Gött. Nachrichten, 1904, p. 341Google Scholar). Dacia ripensis appears in the subscriptions to that Council (Mansi, Concil. iii, p. 42).

^{page 136 note 1} Hist. ecc. ii, 8.1. p. 101 (ed. Parmentier). In the Libellus Synodicus (Mansi, iii, 73) it is Ἠπϵίρου.

^{page 136 note 2} As Mommsen pointed out, Pannonia is used to designate this Diocese in Anon. Val. Pars i, 4 § 9; huic Seuero Pannoniae et Italae urbes et Africae contigerunt.

^{page 136 note 3} And so Mommsen took it, not observing that the order might also be taken as geographical.

^{page 136 note 4} It may be noted that Dalmatia was already a praesidial province in A.D. 277, C.I.L. iii, Supp. 8707, Aurelius Marcianus v. p. praeses prov. Dalmatia (so A.D. 280, ib. 1805); cp. Vopiscus, V. Carini 16, 6, praesidiatus Dalmatiae.

^{page 137 note 1} Cambridge Historical Journal, vol. i (1923) p. 4.Google Scholar

^{page 138 note 1} It seems probable that Lugd. iii was also formed Maximus. These two provinces are not mentioned by Ammian. The motive for the alteration of provincial divisions was in many, perhaps in most, cases some disturbance due to rebellion or invasion, and we may conjecture that Maximus was prompted to a reorganization by troubles in Armorica. We do not know the date of the inscription of Epetum (a few miles from Salonae), which records that L. Artotius [Ju ?]stus, prefect of leg. vi victrix (York), was dux of an expedition consisting of Britannic troops against the Armoricans (C.I.L. iii, 1919). It seems to me more likely to belong to the third century than on the fourth. But I may take the opportunity of observing that the expedition probably involved operations by sea, off the north Gallic coast, and the reason for appointing the prefect of the York legion to the command was that he had had naval experience, having been once praepositus classis Misenatium.

^{page 138 note 2} Cp. also Mansi, Conc. iii, 491, Synodic. to the bishops per Gallias et quinque prouincias (A.D. 374). We have, however, septem prouinciae in Cod. Theod. 15, 1, 15, A.D. 400.

^{page 138 note 3} Cp. the fifth century inscription on a statue to Acilius Glabrio Sibidius who had been uicarius per Gallias Septem prouinciarum, C.I.L. vi. 1678.

^{page 139 note 1} Mansi, Conc. ii, 476.

^{page 139 note 2} Aquitanien in der Römerzeit (S.B. of Berlin Academy, 1896, i, 429 sqq.), p. 437.

^{page 139 note 3} Cp. Homo, Aurélien, p. 167.

^{page 140 note 1} E.g. the temporary union of the two Mauretanian provinces (C.I.L. viii, 8475); that of Europa and Thracia (C.I.L. vi, 1690) under one governor.

^{page 140 note 2} Compare prouincia Aquitanica = both Aquitanicae, in Ammian, 15, 11, 13–15

^{page 140 note 3} Mansi, iii, p. 615, a letter addressed to the bishops prouinciae Viennensium et Narbonensium primae et secundae.

^{page 140 note 4} The question of Narbonensis has been much debated. See Kuhn, , Neue Jahrh. f. Philol u. Pädag. Band115, 1877Google Scholar; Czwalina, , Ueber das Verzeichniss der röm. Provinzen vom J. 297 (Wesel, 1881)Google Scholar; Ohnesorge, op. cit. Mommsen accepts the two Narbonenses for Diocletian's reign (Ges. Schr. v. 583).

^{page 141 note 1} For the general study of the Italian regions and provinces Mommsen's Essay Die libri coloniarum (Ges. Schriften, v, 146, sqq.) is important. Cp. also Die italischen Regionen (ib. 269, sqq.)

^{page 141 note 2} ‘Odeinat et Vaballat, rois de Palmyre et leur titre romain de corrector,’ in Revue biblique 1920, pp. 382–419Google Scholar.

^{page 142 note 1} Hist. Aug. Vit. xxx tyr, 14. In A.D. 283 or 284 Caeonius Rufius Volusianus was corrector Italiae per annos octo, C.I.L. x, 1655 and vi, 1707. For the correctores of Italy, cp. Mommsen, Staatsrecht ii, 1086, and Eph. epigr. i, 138; L. Homo, Aurélien, pp. 144-5.

^{page 142 note 2} e.g. under Diocletian, T. Flavius Postumius Titianus was corrector Italiae regionis transpadanae, C.I.L. v, 1418.

^{page 142 note 3} Cp. Mommsen, chronica Minora, i, p. 532Google Scholar.

^{page 142 note 4} Cp. Vopiscus, V. Probi 16, 1Google Scholar, Retias sic pacatas reliquit, etc.

^{page 142 note 5} Gesch. d. Untergangs d. ant. Welt, ii, p. 498.

^{page 143 note 1} Mommsen's latest view, however, was that Zeugitana should be deleted (C.I.L. viii, praef. p. xvii, note 5). But I do not think that there is any necessity for considering Zeugitana an interpolation. It may have been added in the margin by the original author of the List, who intended it to follow and qualify proconsularis, and then been inserted in the text by the first copyist of the List, after bizacina.

^{page 144 note 1} C.I.L. viii, 7002, 4578, 2529, 2530 and 2643.

^{page 144 note 2} C.I.L. vi, 1690. The inscription is subsequent to A.D. 340, the year of the consulship of Aradius. He was Prefect of Rome in 337, and, before that, proconsul of Africa discharging the duties of vicar of the Diocese, viz. per prouincias proconsularem et Numidiam Byzacium ac Tripolim itemque Mauretaniam Sitifensem et Caesariensem.

^{page 144 note 3} C.I.L. viii, Praef. p. xvii–xviii.

^{page 144 note 4} Probably Fl. Victor Calpurnius v.e. praeses prov. Tripol. (viii, 22672), to whom Leptis Magna decreed a statue, functioned in Diocletian's reign or soon after.

^{page 144 note 5} For the part of Numidia (including Hippo Regius, Calama, etc.) which was under the proconsul and may be distinguished as Num. proconsularis, see Mommsen, C.I.L. viii, praef. p. xvi.

^{page 145 note 1} Atlas Antiquus, Lief. V. Nr. 27. Not a good map.

^{page 145 note 2} Mileuitana would be an easy emendation, but it would involve a very improbable division of provinces.

^{page 145 note 3} Praeses prou. Maur. Cae. et Sitifensis. The dedication is to Fl. Julius Constantius Caesar, who was created Caesar on Nov. 8, A.D. 323. Flavius Augustianus, p. Maur. Sitif., may have come between Flavianus and Terentianus. His stone (C.I.L. viii, 8475) is dedicated Flavio Claudia Constantio notilissimo Caes., where Constantio must be an error for Constantino (viz. Constantine II).

^{page 147 note 1} We learn this from Procopius, De Aed. iii, 1, who has confounded Theodosius I and Theodosius II, so that Mommsen was misled into dating the province of Armenia Maior to A.D. 441. Cp. Chapot, , La frontière de l'Euphrate (1907), p. 169Google Scholar.

^{page 147 note 2} Mr. Adonts, in his useful work Armenia v epokhu Iustiniana (Petersburg, 1908)Google Scholar, touches on this question (pp. 116–7), and, assuming that the date of the Notitia was A.D. 410–3 and that the comes Armeniae was a later institution, he finds no difficulty.

^{page 147 note 3} Ohnesorge, op. cit. p. 14.

^{page 148 note 1} Mr. Seeck has conjectured Gaul, on insufficient grounds.

^{page 148 note 2} Op. cit.

^{page 149 note 1} The authoritative text will be found in vol. i of Mommsen's Chronica Minora. It is reprinted in Seeck's Notitia Dignitatum.

^{page 149 note 2} Cp. C.I.L. x, 1125.

^{page 149 note 3} To what I have said in support of this date in J.R.S. x, 153, I may add here that the appeal of the Britons to Aëtius in A.D. 446, recorded by Gildas, is more natural and intelligible if the evacuation occurred a few years before than if it had occurred thirty-five years before. I may add that Mr. Collingwood's article in J.R.S. xii, 74, sqq. has not shaken my belief that this date is approximately right. His contention that a date later than 410 is impossible rests mainly on coin evidence. It was known to most people that no Roman coins or hardly any (the exceptions are negligible) of later date than issues of Arcadius and early issues of Honorius have been found in Britain, and Mr. Collingwood has performed the service of collecting and arranging this evidence as a whole. This minor premiss of his argument is unimpeachable; but the major premiss required for his conclusion implies assumptions which seem to me altogether erroneous. The historical value of the coin evidence is not that it throws any light on the evacuation—it is equally compatible with either date—but that it illustrates the fact that during the reign of Honorius the Channel became so unsafe, through the operations of the Saxon pirates, that the trade of Britain with the continent declined and presently ceased altogether. I must thank Mr. Collingwood for making it clear to me that I was quite mistaken in what I suggested as to the signal-stations on the Yorkshire coast.

^{page 150 note 1} My reason for this conjecture is that are no observations on the southern provinces of Gaul, like those which occur on the northern, e.g. Belgica prima in qua est Treveris, Belgica secunda de qua transitur ad Brittanniam, etc. But of course these notes (and those on the other Dioceses) may have been in the List which Polemius copied and not be due to him.

^{page 150 note 2} We do not know how far provincial arrangements in Spain may have been altered as a result of the barbarian invasions and wars from A.D. 409 to 429. In the latter year the peninsula was freed from most of these intruders; only the Suevians were left in Callaecia. There is little doubt that their presence there had been regularised by the Imperial government, and it may be presumed that the contract allotted to them a portion of the land (as the cases of the Visigoths and Burgundians in Gaul), If so, their settlements would not have involved the extinction of the Roman province of Callaecia.

^{page 150 note 3} J.R.S. x, p. 135. The Silvian List places Galatia under Asia instead of Pontus.

^{page 150 note 4} It is interesting to find Haemimontus designated by another name, Thracia secunda, in the place where it appears among the Thracian provinces.

^{page 151 note 1} Tillemont, H.E. v. 699; Mommsen, Chron. Min. i, 533.

^{page 151 note 2} It may be noted that in another work (Brev. Temp. in Chron. Min. i, p. 547), Polemius seems to have had a different list of Dioceses before him, for he enumerates the parts of the Empire exactly as he does in his Laterc. Provinc., except that he omits Egypt. This points to a source prior to A.D. 380.