¹ This translation has been made, in the main, from the Latin text which is fuller than the Greek at several points. However, in l. 7, where the reading is doubtful, I have translated the Greek.

² ILS 5925, fully published at CIL VI 31544 with two further specimens of the text.

³ See Dessau, ad loc.; Wiseman, T. P., New Men in the Roman Senate (1971), 220, 262Google Scholar; Degrassi, A., I Fasti Consolari (1951), 8–9Google Scholar.

⁴ The Roman Revolution (1939), 361, n. 3: ‘that blatant prodigy of nomenclature’. For Sotidii at Canusium see CIL IX 349 and 397; Libuscidii, ibid. 338, 348, 387, 6186. P. M. Fraser and G. E. Bean attempted to restore the names Sextus Sotidius Libuscidius on an inscription of the Rhodian Peraea, and suggested that the senator from Canusium had held a provincial command in Asia or had settled privately there (The Rhodian Peraea (1954), 3–4. no. 3). Given that one or two other Libuscidii are attested on inscriptions of Rhodian origin (cited ad loc.) and a branch of the gens was clearly established locally, this was an implausible suggestion at the time. It is now put firmly out of court by the new text which shows that the man in question was called Libuscidianus, not Libuscidius, which must be the name in the Rhodian inscription.

⁵ This argument is based on a perusal of the indices of ILS, IGR, SIG ³, and OGIS. Supported also by Bureth, P., Les titulatures impériales dans les papyrus, les ostraca et les inscriptions d'Egypte (1966)Google Scholar.

⁶ Strabo XII. 5. 1, 567; 6. 5, 569.

⁷ See Levick, B., Roman Colonies in Southern Asia Minor (1967), 32, n. 3Google Scholar; 164; Ptolemy v. 5.4 indicates that the dividing line lay north of Conana and Seleuceia Sidera, although Apollonia lay in Galatia. Epigraphic evidence confirms this (CIL III 6885; Ramsay, W. M., JRS VI (1916), 132)Google Scholar.

⁸ The best text in Krencker, M. and Schede, M., Der Tempel in Ankara (1936), 52 fGoogle Scholar. Less good in Bosch, E., Quellen zur Geschichte der Stadt Ankara im Altertum (1967), 35 f.Google Scholar, no. 51.

⁹ M. Grant, Num. Chron. 6th ser. x (1951), 43 ff., no. 1.

¹⁰ Similar conclusions are reached by R. K. Sherk in a revised version of his fasti of Galatian legates (see n. 12) to be published in Aufstieg und Niedergang der röm. Welt. I am grateful to Professor Sherk for showing me a typescript of this in advance of publication.

¹¹ See, primarily, Syme, R., ‘Galatia and Pamphylia under Augustus’, Klio XXVII (1934), 122–48Google Scholar. I have argued against Syme for a permanent garrison, in a paper to be published in CQ 1976.

¹² Sherk, R. K., The Legates of Galatia from Augustus to Diocletian (1951), 26–31Google Scholar.

¹³ See Fergus Millar, , ‘The Emperor, the Senate and the Provinces’, JRS LVI (1966), 156–66Google Scholar, esp. 166.

¹⁴ D.3, 17–18.

¹⁵ D.8, 8–10.

¹⁶ D.20, 45–6. The Latin equivalent, licentia, is also found in a passage of Valerius Maximus describing an episode in the triumviral period where a man put on the insignia of a praetor, illegally requisitioned waggons and a ship, and claimed hospitality to which he was not entitled (VII. 3. 9).

¹⁷ D.I.

¹⁸ D.2, 6–9; D.3, 21–5.

¹⁹ D.8.

²⁰ D.7. Note the clauses in Trajan's reply (Ep. 78) particularly aimed at punishing soldiers: ‘si qui autem se contra disciplinam meam gesserint, statim coerceantur; aut, si plus admiserint quam ut in re praesenti satis puniantur, si milites erunt, legatis eorum quod deprehenderis notum facies, aut, si in urbem versus venturi erunt, mihi scribes.’ He does not give any procedure for civilian offenders, although the possibility is envisaged by the clause ‘si milites erunt’.

²¹ D. 10, 8–12.

²² D. 13, 7–10.

²³ For the term see L. Robert, OMS 1, 364–72.

²⁴ See Hirschfeld, O., ‘Die Sicherheitspolizei im röm. Kaiserreich’, Kleine Schriften (1913), 576–612, esp. 596 fGoogle Scholar.; Robert, L., Etudes Anatoliennes (1938), 285Google Scholar.

²⁵ References to these at D.16a, 7, 11; 17, 5; 18, 10, 11, 21; 19, 14, 16; 20, 25, 35, 45; D.21 has much in common with these texts, but does not mention the source of the trouble.

²⁶ D.15, 33 f.

²⁷ D.16, 18–21.

²⁸ Ulpian, Dig. XIX. 2. 15. 2; cf. 13.7; and Paulus, Dig. 1. 18. 6. 3–6. Most of the evidence for soldiers oppressing civilian communities to obtain transport, billets, supplies or simply cash is well discussed by MacMullen, R., Soldier and Civilian in the Later Roman Empire (1964), ch. IVGoogle Scholar.

²⁹ Tac., Ann. XV. 20.1; Plut., Mor. 815a; Ael. Arist. XXVI, 65K; Oliver, J. H., The Ruling Power (1953). 953–8Google Scholar.

³⁰ See n. 20.

³¹ See n. 21.

³² D. 16, 20; cf. A. Schulten, Röm. Mitt. XIII (1898), 245.

³³ D.15, 33–44.

³⁴ D.20, 36–8.

³⁵ JEA 1935, 232, ll. 80 f. This law may only have applied to Egypt.

³⁶ D.4: ‘Cum et colonias et municipia non solum Italiae verum etiam provinciarum et civitatium cuiusque provinciae lebare oneribus vehiculorum praebendorum saepe temptavissem’. The expression ‘cuiusque provinciae’ should indicate that each of Claudius' attempts to enforce his legislation was concerned with all the provinces. Presumably on these grounds, it is regarded as an edict of universal application by W. Williams, ZPE XVII (1975), 44 f. However, on the face of it the particular case in question is simply the latest of Claudius' repeated efforts, concerned, on this occasion, with Tegea.

³⁷ A general clause could have laid down that a city was required to provide transport within its own territory, but the details would have to be worked out in individual cases.

³⁸ For the Lex Iulia see below p. 127 f.

³⁹ Ep. x. 22. 1; 30. 1; 56.3; 96.7; no. 1; 111. 1. The scope of the mandata is discussed by Sherwin-White, A. N., The Letters of Pliny (1966), 547 ffGoogle Scholar. and by Vidman, L., Klio 1959, 217–25Google Scholar and Etude sur la correspondance de Pline le jeune avec Trajan (1960), 45–51Google Scholar.

⁴⁰ C. B. Welles, op. cit., no. 30, 11 (from Ptolemy IV to a governor); no. 33, 20 (also Ptolemy IV); no. 58, 6 (from the high priest at Pessinus to his envoy Menodorus); and p. 336; Sherk, op. cit. no. 58, 78 (from the city of Rhosus to its ambassadors). Cf. Livy XXXVIII. 8. 1, and see IGLS v no. 1998 n. 1.

⁴¹ Sherk, op. cit. no. 11, 6; no. 15, 18, 36, 52; no. 18, 61.

⁴² Dio LIII. 15. 4. Millar, F., JRS LVI (1966), 157–8Google Scholar gives examples sent to imperial legates, procurators in the public provinces and the prefect of Egypt in the first century A.D. An inscription from Cos provides an example of ἐντολαί sent to the proconsul of Asia under Claudius (La Parola del Passato 1975, 102–4).

⁴³ SIG ³ no. 799, 11 f. may provide an example of instructions given by the emperor to Antonia Tryphaena concerning the regulation of the succession in the Thracian dynasty. However, even if we assume stone-cutting errors the text is extraordinarily difficult to understand (see Wilhelm, A., Anatolian Studies pres. to Sir W. M. Ramsay (1923), 427–31Google Scholar).

⁴⁴ I am indebted to Dr. Graham Burton for discussing the question with me, in advance of a study of the subject to be published in ZPE 1976.

⁴⁵ D.6.

⁴⁶ For other Greek inscriptions found at or near Burdur see Ramsay, W. M., The Cities and Bishoprics of Phrygia I (1895), 337 f. nos. 171–9Google Scholar. Burdur appears to have been known as Praetoria in late antiquity, a form preserved by the modern name. See Honigmann, E., Byzantion XIV (1939), 654–5Google Scholar. This suggestion was not noticed by G. E. Bean in his discussion of the problem at AS 1959, 78, but seems more probable than any of the alternatives canvassed there.

⁴⁷ D.3, 8–14 and restoration in 65; D. 10, 34–7; D.12. For the reason, cf. CTh. VIII. 5. 6: ‘iussione nostra cunctis provincialibus intimata’.

⁴⁸ D.6.

⁴⁹ See B. Levick, op. cit. (n. 7), ch. iv–v.

⁵⁰ As I have argued in a paper for the 10th International Congress of Classical Archaeology, Ankara and Izmir 1973.

⁵¹ See Th. Mommsen, Eph. Ep. VII (1892), 436 f., esp. 442; Hatzfeld, J., Les trafiquants italiens dans l'Orient hellénique (1919), 122Google Scholar; MAMA VI, nos. 180, 183; Strabo XII. 8. 15, 577.

⁵² IGR III, no. 325; SEG 11 no. 744.

⁵³ Levick, op. cit. (n. 7), ch. xi–xii.

⁵⁴ l. 13 = 31–2. I prefer to follow the order of the Greek in discussing this problem here.

⁵⁵ G. E. Bean, AS IX (1959), 91–7. An inscription from the site (p. 93, no. 42) reads ‘C. Iulius C. f. Papiria natus Cormasa missicius lecinis (sic) VII eques momomentum (sic) f(e)cit sibi et Iulio Iucundo liberto suo’ (followed by a shorter Greek version of the same text). Bean comments that he had probably returned to his home town after discharge from service. Since the evidence for the location of Cormasa accords well enough with the site at EǦneṣ, the identification is probable though still not proven. Against this it might be argued that legio VII had been stationed in south Galatia under Augustus (see n. 11), and that C. Iulius C. f., clearly an early recruit, had retired to a veteran settlement away from his home town.

⁵⁶ art. cit. 84–8, no. 30, citing previously discovered copies of the text. Two more copies have also been found at Yariköy (Robert, L., Hellenica XI/XII (1960), 596)Google Scholar. Statue bases of M. Aurelius and L. Verus and of the tetrarchs set up by Sagalassus have been found at Yaziköy (IGR III, no. 332) and Düwer (IGR III, no. 336) respectively.

⁵⁷ Hadriani, located by Bean at Gâvur ören between Cormasa and Sagalassus, which set up statues of L. Verus and Caracalla, had presumably been a village before it was promoted to city status by Hadrian. See Bean, AS 1959, 108 f. with nos. 79 and 80.

⁵⁸ W. M. Ramsay at one time suggested that Darsa, which appears only in Livy's version and is not otherwise attested, may have been a fiction of Livy (The Cities and Bishoprics of Phrygia I (1895), 327 n. 3Google Scholar, cf. AS 1959, 78). Unless further evidence for the place emerges this view is not without its attractions. The problem is discussed by Bean, AS 1959, 116–17.

⁵⁹ Pol. XXI. 36 (ed. Büttner-Wobst); Livy XXXVIII. 15. For Lysinia see Bean, op. cit. 78 f., and for a discussion of Manlius' march substantially in agreement with the one given here, ibid. 113–16.

⁶⁰ Hirschfeld, O., M-Ber. Akad.Berl. 1879, 315–16Google Scholar; RE XI, 1308 s.v. ‘Konane’ (Ruge).

⁶¹ Robert, L., Hellenica XI/XII, 353, n. 4Google Scholar, 596, citing bibliography.

⁶² RE 11A, 1204 no. 6 (Ruge); Robert, , Hellenica x (1955), 243–4Google Scholar. Robert notes that the ruin-field was still known as Selef in 1948.

⁶³ At Findos. See Robert, , Etudes épigraphiques et philologiques (1938), 281, n. 3Google Scholar; CRAI 1948, 402 = OMS III 1455, citing an unpublished inscription of the imperial period showing that it was a city; Hellenica x, 240, n. 3.

⁶⁴ See Levick, op. cit. (n. 7), 46 f., 48 f., 50 f.

⁶⁵ Robert, OMS III 1455; M. H. Ballance, AS IX (1959), 125–9.

⁶⁶ Levick, 53–4. The location of the site at Barla was first demonstrated by L. Robert, cf. E. W. Gray, CR 1974. 271–2.

⁶⁷ See the argument on p. 120 below. The gradual appearance of stations along the main roads equipped with facilities of all kinds, but above all with remounts, was probably a major factor encouraging this trend. An inscription of the time of Nero mentions tabernae et praetoria constructed along the military roads of Thrace (CIL III 6123, discussed by Pflaum, op. cit. (p. 112), ch. ii). By the third century there was an extensive network of mansiones and mutationes covering most of the main roads of the empire (see the Itineraria Antonini and Burdigalense passim, and the commentary on a Bithynian inscription, which gives details of the personnel of such a station, by Robert, , Hellenica x, 46–66Google Scholar). However, the absence of any mention of such establishments in the new inscription and in the other first century documents suggests that the system had not yet been created. See also p. 127.

⁶⁸ Plut., Galba 8.4; Jones, A. H. M., The Greek City (1940), 328, n. 90Google Scholar also cites SHA, Had. 7.5 and D.8, 5–7. For a similar procedure in the late republic see Caesar, BC 1. 30 (with 11. 18 and 111. 42). Compare the responsibility of city magistrates for collecting the tributum, Brunt, P. A., JRS LXV (1975), 138Google Scholar and Jones, A. H. M., The Roman Economy (ed. Brunt, , 1974), 163, n. 71Google Scholar

⁶⁹ D.5.

⁷⁰ Tac. Ann. XIII. 30. 3; Pflaum, H.G., Journal des Savants 1959, 79 fGoogle Scholar.

⁷¹ τῆς ἀνγαρείας ὑμᾶς τὸ λοιπὸν ἀπολύω παρὲξ ὦν διὰ τῆς ὑμετέρας χώρας.

⁷² For the Thasian Peraea see Dunant and Pouilloux, op. cit. (p. III ) I and II, index s.v. The map opposite 1, 8 appears to show the course of the Via Egnatia.

⁷³ Tac. Ann. XII. 62. 2, cf. 63. 3.

⁷⁴ D.7.

⁷⁵ D.9, A. l. 9 f.

⁷⁶ D. 11, 50 f.

⁷⁷ D. 14.

⁷⁸ ll. 4–7.

⁷⁹ ll. 8–13.

⁸⁰ Untersuchungen zu den röm. Reichsstrassen, 148 f. esp. 153; compare the evidence from Egypt cited by Zucker, F., Sitzb. Berl. Akad. 1911, 805–6Google Scholar.

⁸¹ However, the question is somewhat obscured by the fact that both villages appear to have been on an imperial estate and may have lain outside city jurisdiction in any case. See Millar, F., Historia XIII (1964), 186Google Scholar; Brunt, P. A., Latomus XXV (1966), 483 f.Google Scholar; Boulvert, G., Esclaves et affranchis impériaux (1970), 417 fGoogle Scholar.

⁸² Since c. 40 B.C. there had been a succession of major and minor conflicts in Pisidia and Isauria. See Syme, R., Klio 1934, 122 fGoogle Scholar. for campaigns under Augustus, and Anatolian Studies pres. to W. H. Buckler (1938), 299–332Google Scholar for the earlier period. Also Levick op. cit. (n. 7), ch. ii–iii and Appendix 5. Later Q. Veranius, the first governor of Lycia, 43–8, had to deal with rebellious tribesmen in Cilicia Tracheia (Gordon, A. E., Q. Veranius, consul A.D. 49 (1950), 248 f.Google Scholar) and the Cilician Cietae were a constant menace (Tac., Ann. VI. 41; XII. 55). For the late empire see Rougé, J., ‘L'Histoire Auguste et Isaurie dans le IV^e siècle’, REA LXVIII (1966), 282–315CrossRefGoogle Scholar.

⁸³ References to the Via Sebaste are collected by Levick, op. cit. (n. 7), n. 1. For republican roads in Asia see Magie, D., Roman Rule in Asia Minor (1950) 11, 1048–9Google Scholar.

⁸⁴ 11. 6.

⁸⁵ Strabo XVII. 1. 24, 804; XI. 14. 11, 530.

⁸⁶ NH VI. 124; XII. 53.

⁸⁷ Geometrica XXIII. 20, 43. See also Daremberg, Saglio, Dictionnaire des Antiquités, s.v. ‘Schoenus’.

⁸⁸ It may thus be added to the other evidence for the survival of Persian institutions in Graeco-roman Asia Minor. See, most recently, L. Robert, CRAI 1975. 306–30.

⁸⁹ See n. 82.

⁹⁰ The relevant portions of the road were traced on the ground in July 1975. I am very grateful to David French for making this information available to me in advance of detailed publication.

⁹¹ See p. 117 f.

⁹² See p. 120.

⁹³ The Greek text, which is less tightly worded than the Latin here, allows the alternatives of ζευκτά for κάρρα and νωτοφόροι for ἡμίονοι, translating waggons and mules respectively, but the general sense and the Latin version show that these were exact equivalents.

⁹⁴ For the gender see TLL s.v. citing Nonius Marcellus p. 195. 6.

⁹⁵ See TLL s.v. and Vigneron, P., Le cheval dans l'antiquité gréco-romaine I (1968), 151 fGoogle Scholar.

⁹⁶ Note especially Caesar, BG I. 26. 1; IV. 14. 4; VII. 18. 3; VIII. 14. 2.

⁹⁷ XV. 38 f. and XVII. 4 (ed. Lauffer, 1970); see also the new fragment from Aphrodisias in JRS 1973, 102, l. 4, with comment on p. 106. P. Vigneron, op. cit. 1, 152 with II. figs. 89a and b, assumes that it was generally a two-wheeled vehicle, of the type illustrated on Trajan's column. I can find no basis for this assumption.

⁹⁸ CTh. VIII. 5. 47.

⁹⁹ For the reliability of the Theodosian Code as a guide to the capacity of waggons see Sion, J., Annales 1935, 628–9Google Scholar.

¹⁰⁰ This makes it almost impossible to estimate the size of a normal waggon load. The best we can do is to compare the rate of hire for a carrum, ten asses per schoenum, with four asses per schoenum for a mule. This suggests that the carrum carried between 2½ and 3 times as much as a mule. The carrying capacity of a mule is variously estimated. C. A. Yeo, in a study of land transport in Roman Italy, suggests that a ‘pack ass’ (presumably a mule) could carry about 250 English pounds (TAPA LXXVII (1946), 225), while Vigneron suggests an upper limit of 150 kilograms (op. cit. I, 147–9). The carrum might therefore be expected to carry between 625 and 950 pounds. These figures, however, are wretchedly imprecise and, combined with the variable length of the schoenus already discused, preclude any attempt to assess even the approximate cost of the service in comparison with transport prices on the open market (for which see Duncan-Jones, R. P., The Economy of the Roman Empire (1974), 366 f.Google Scholar).

¹⁰¹ D.6, 11, 30–1. The ὑποζύγιον of D. 1, 13 could be a carriage drawn by mules or other draft animals.

¹⁰² D. 14, 4. Pekáry, op. cit. (p. 112), 136, makes the attractive suggestion that these were used to transport marble from the nearby quarries at Docimium.

¹⁰³ D. 16, 21. Cf. P. Oxy. 3109 for oxen requisitioned for military purposes under Valerian and Gallienus. However, these purposes may have been confined to ploughing (ἐπιτηδειοτάτας πρὸς ἄροσιν, l. 27). See also CTh. VIII. 5. 1, 11.

¹⁰⁴ Fougères, G., Journal des Savants 1924, 220–32Google Scholar; Burford, A., ‘Heavy Transport in Classical Antiquity’, Ec. Hist. Rev. 2nd ser. XIII (1960), 1–18Google Scholar; Vigneron, op. cit. (n. 95) 1, 153–5.

¹⁰⁵ In mountainous districts, such as Pisidia, mules might have been more readily available than oxen, while the reverse might be the case in agricultural plain land. It has been suggested that the ox-drawn cart is a feature of steppe or plateau land, not of the mountainous fringes of the Mediterranean. The hypothesis might be worth testing in Anatolia, by comparing the distribution of the ox-drawn solid-wheeled cart with that of lighter waggons.

¹⁰⁶ See, above all, Sion, J., ‘Quelques problèmes de transport dans l'antiquité’, Annales d'Histoire Economique et Sociale 1935, 628–33Google Scholar, criticizing the view of Lefebvre des Noettes, L'attelage, le cheval de selle a travers les âges ² (1931). See also Vigneron, op. cit. (n. 95) 1, 147–9; Braudel, F., The Mediterranean and the Mediterranean World in the Age of Philip II (Eng. trans, of 1966 ed., 1972) I, 282–4Google Scholar.

¹⁰⁷ D. 1, 19–21; D.3, 5–7; D.6, 11.

¹⁰⁸ Lesquier, J., L'armée romaine d'Egypte (1919), 369–74Google Scholar.

¹⁰⁹ See below p. 127 f.

¹¹⁰ See the notes on the text. Perhaps read the passive Χρησθή[σ]ονται which could either have been used in place of the expected middle form, or show that the author had also been bewildered by the construction of his own sentence.

¹¹¹ cf. Davies, R. W., Britannia II (1971), 123Google Scholar.

¹¹² Strabo III. 4. 20, 167: οἱ διανέμοντες τὰ χρήματα τοῖς στρατιώταις εἰς τὴν διοίκησιν τοῦ βίου. The last phrase suggests that goods and services as well as money were involved.

¹¹³ IGR IV no. 914, ll. 11–15, discussed by Magie, D., Studies in Roman Social and Economic History in Honour of A. C. Johnson (1951), 152–4Google Scholar. Rostovtzeff, SEHRE ², 700, n. 21 assumes that Tiberius Nicephorus had been the procurator, which seems more probable than Magie's suggestion that he was a minor treasury official.

¹¹⁴ D.6. The procurator was certainly the procurator of Syria, not a domanial official concerned with imperial estates, see Millar, F., JRS 1963, 199Google Scholar.

¹¹⁵ IGR III no. 738 = TAM 11. 3 no. 908. sect. IVa. See Millar, loc. cit.

¹¹⁶ Welles, C. B., Fink, R. O. and Gilliam, J. F., The Excavations at Dura Europos V. I. The Parchments and Papyri (1959), 222 ff.Google Scholar, no. 60. See M. I. Rostovtzeff, CRAI 1933, 315–23; Pflaum, H.-G., Les Procurateurs Equestres sous le Haut-Empire Romain (1950). 156–7Google Scholar.

¹¹⁷ Modestinus, Dig. XXVII. 1. 6. 8, 14.

¹¹⁸ If the stone cutter had been responsible we would have expected him to have been correct in at least one version.

¹¹⁹ For similar limits to the amount that could be obtained from any one place, see CTh. VIII. 5. 35.

¹²⁰ My orientation here owes much to several suggestions made to me by Professor Brunt.

¹²¹ It is preferable to translate the clauses ‘et iis … et iis …’ as ‘both those who … and those who …’ in this way, rather than to treat the ‘et … et’ as Unking three independent categories comprising 1) militantes 2) ‘ii qui diplomum habebunt’ and 3) ‘ii qui … commeabunt’. If this had been the meaning intended we would expect that only the last element would be attached with a -que. This rule is not invariable, and l. 24 of the Latin text provides a good counter-example (‘omnibus in comitatu nostro et militantibus … et princ. opt. libertis et servis et iumentis eorum’); but the Greek version supports the interpretation I have adopted. If three categories were intended the Greek would not run τοῖς στρατευομένοις καὶ … καὶ, but τοῖς τε στρατευομένοις καὶ … καί, which is the construction we find in the Greek version of 1. 24 (ll. 49–51). However, Prof. R. G. M. Nisbet points out that this grammatical argument is unsound, which therefore casts doubt on the interpretation offered.

¹²² Pliny, Ep. X. 45–6, 64, 83, 120–1 with Sherwin-White's notes. See also Pflaum, op. cit. (p. 112), 231 f.

¹²³ In Pisonem 90.

¹²⁴ Other references to diplomas at this date are not very helpful on this point (Cic., ad Att. x. 17. 4; ad Fam. VI. 12. 3). However, Cato in 164 B.C. alludes to the right he had had as a governor to issue evectiones, which seem to be equivalent to diplomas (quoted in Fronto 11. 44 (Haines) = Malcovati, ORF ² 71, fr. 173). Evectio reappears with this meaning in the fourth century A.D. (CTh. VIII. 5. passim).

¹²⁵ See principally Suet., Aug. 50, which, despite Pflaum, op. cit., ch. ii, need not refer exclusively to diplomas for the postal or transport service; cf. Seneca, De Clem. 1. 10. 3; Suet., Gains 38; Nero 12.

¹²⁶ Although the word diploma is not used, it is worth noting that both Germanicus and his friend Baebius could authorize transport requisitions in Egypt in A.D. 19 (D. 1, 13–15).

¹²⁷ Tac., Hist. 11. 65.

¹²⁸ Possibly as much to provide them with a date as to give diem authority, cf. Sherwin-White on Pliny, Ep. X. 45–6. Compare Plutarch, , Otho 3Google Scholar and Tac., Hist. 11. 54.

¹²⁹ Plutarch, , Galba 8Google Scholar.

¹³⁰ D.6, 18–19. Compare CTh. VIII. 5. 5.

¹³¹ D.6, 10–12.

¹³² For the seals on diplomas note especially Plut., , Galba 8. 4Google Scholar, τὰ καλούμενα διπλώματα σεσημασμένα, and Suet., Aug. 50, showing that Augustus used his own seals on all diplomas, epistulae and libelli (cf. Pliny, NH XXXVII. 10 and Dio LI. 3. 4 f.). Pliny, who used his own personal seal to guarantee a package sent to Trajan (Ep. X. 74. 3), was apparently required to counter-stamp the imperial diplomas which he distributed (Ep. X. 45; ‘Vereor enim, ne in alterutram pattern ignorantia lapsus aut inlicita confirmem aut necessaria impediam’). For official seals in general see L. Wenger, RE IIa, 2361–2448, esp. 2440 f. s.v. ‘signum’.

¹³³ Modestinus, Dig. XLVIII. 10. 27. 2. There are also references to unauthorized documents in D.6, 20–1 and in D.4, 11 where the word falsa appears.

¹³⁴ Dio LII. 42. 6–7, cf. Tac., Ann. XII. 23. See Mommsen, Th., Römisches Staatsrecht III, 2, 912–3Google Scholar citing other references.

¹³⁵ Dig. XLVIII. 10. 27. 2; see n. 16 above. Some confirmation for the interpretation of the text proposed here comes from the argument advanced by Pflaum that an imperial diploma not only gave the holder permission to claim transport but also specified the number of animals and waggons to which he was entitled (op. cit. 321–2, citing CIL VIII 1027). This short sepulchral epigram set up by a government messenger (tabellarius) for himself contains the phrase ‘diploma circavi totam regionem pedestrem’, meaning apparently ‘with the help of a diploma I have covered the whole region on foot’. Since the tabellarius evidently used neither carts nor animals, Pflaum argues that the diploma gave details of the hospitium and other basic necessities to which he was entitled, and suggests that such documents would, where appropriate, give similar particulars of the means of transport required. If this is correct, it adds support to the view that the clauses beginning at ‘ita ut’ only qualify the second category of official traveller. These details would be otiose if they were already listed on a valid diploma.

¹³⁶ CTh. VIII. 5. 38.

¹³⁷ For the effect of the high cost of overland transport see Rostovtzeff, SEHRE ², 590–600, 700–1; Jones, A. H. M., The Later Roman Empire (1964), 841 fGoogle Scholar. esp. 844–5; Brunt, P. A., Italian Manpower (1971), 703 fGoogle Scholar. I suspect, however, that the case may be overstated, and that land transport of bulky, lowcost materials was commoner than they suggest.

¹³⁸ See TLL s.v. ‘hospitium’, ‘mansio’; Oxford Latin Dictionary s.v. ‘mansio’, 1; Pflaum, op. cit. (p. 112), 339 f.

¹³⁹ Terence, , Phormio 1012Google Scholar; Cic., ad Fam. IV. 4. 5; ad Att. VIII. 15. 2, IX. 5. 1.

¹⁴⁰ NH XII. 65; Suet., , Titus 10. 1Google Scholar is ambiguous.

¹⁴¹ See TLL and W. Kubitschek RE XIV, 1233–52 s.v. ‘mansio’.

¹⁴² See n. 67. Μονή appears in this technical sense on the Phrygian angareia inscription, D. 14, 3. For mansio see also MAMA VIII no. 305, 23 (Orcistus, reign of Constantine).

¹⁴³ ad Att. v. 10. 2 (on the journey to his province); 16. 3; 21. 5.

¹⁴⁴ The locus classicus for abuses of the hospitality owed to a governor and his staff is Cic., II Verr. I. 63–70 (Verres at Lampsacus). See also Cic., De Imp. Cn. Pompei 13; ad Qu.fr. 1. 1. 9; and Livy XLII. I. 7–12 (173 B.C.). Much of the evidence for abuse, and exploitation of the inhabitants of Asia Minor under the republic is collected by Broughton, T. R. S., Roman Asia Minor in An Economic Survey of Ancient Rome vol. IV, ed. Frank, T. (1938), 562–78, esp. 574 fGoogle Scholar.

¹⁴⁵ Plutarch, , Sulla 25Google Scholar; cf. Broughton, loc. cit. and MacMullen, op. cit. (n. 28) ch. IV for evidence from the empire. For immunities see ILS 38 col. II. 10–17 (Lex Antonia de Termessis); M. Segré, Riv. fil. XVI (1938), 253 f. (on the letter from Sulla granting privileges to the Dionysiac artists of Cos, citing much of the other evidence on the subject); EJ no. 302 (privileges given to veterans by Octavian); Sherk, op. cit. no. 58, 34 f. (grants of immunity to Seleucus of Rhosus, including apparent references to the Lex lulia and the Lex Atilia de repetundis); OGIS no. 262 = IGLS VII no. 4028, iii (immunity for the temple at Baetocaece).

¹⁴⁶ D. 1, 15 f.; D.3, 20; D.6, 11–12 (reading ξε[νιῶν]ὀχλὴσεσιν with Mihailov, IGBulg. IV p. 226); D. 10; D. 15. 40 f., 47 f.

¹⁴⁷ Ulpian, Dig. L. 1. 16. 4; 6. 3; Modestinus, Dig. L. 1. 18. 1 f.

¹⁴⁸ Columella could recommend that a villa should not be built too close to a military road lest it be blighted by having to provide too much hospitality for the soldiers passing by (1. 5. 6–7). There may also be echoes of the terms of the Lex lulia in Hor., Sat. 1.5. 45–6 and even in the edict which Aurelian is alleged to have made concerning military abuses (SHA, Aurelian 7). Also compare the regulations on the supplies which should be provided to παραφύλαες; at Hierapolis (OGIS no. 527).

¹⁴⁹ The reference in the edict of Vergilius Capita to τῶν ὑπὸ Μαξίμου σταθέντων (D.3, 27) should indicate refinements introduced during the reign of Augustus. M. Magius Maximus had been prefect of Egypt between A.D. II and 14 (Stein, A., Die Präfekten von Ägypten (1950), 22–3Google Scholar; O. W. Reinmuth, BASP IV (1967), 77–8; Brunt, P. A., JRS 1975, 143Google Scholar; against, unconvincingly, Fraser, P. M., Ptolemaic Alexandria 11 (1972), 1109Google Scholar). It would appear from this reference, the anecdote cited at n. 172 and the new document, that the last years of Augustus and the first years of Tiberius were particularly notable for attempts to suppress corruption in the administration.

¹⁵⁰ See nn. 16, 123–4.

¹⁵¹ Suet., Aug. 49. 5–50. Discussed in scrupulous, perhaps over-scrupulous, detail by Pflaum, op. cit. (p. 112) ch. ii.

¹⁵² Pflaum, ch. i. For the Ptolemaic post see Preisigke, F., ‘Die Ägyptische Schnellpost’, Klio VII (1907), 241–77Google Scholar. For requisitioning in the Hellenistic world both for the postal system and for transport see Rostovtzeff, M. I., ‘Angariae’, Klio VI (1906), 249–58Google Scholar; O. Seeck, RE 1, 2184–5 s.v. ‘angaria’; Rostovtzeff, SEHHW, 315, 1391, n. 115 and the index. For republican courier systems see Caesar, BC III. 101. 3, cf. Suet., Caesar 57.

¹⁵³ See Sherwin-White, op. cit. (n. 39), 546 f.

¹⁵⁴ ibid. 628.

¹⁵⁵ J. Lesquier, op. cit. (n. 108), 369 f., cf. P. Oxy. 3109.

¹⁵⁶ For Ptolemaic requisitioning procedures see Préaux, C., L'Economie royale des Lagides (1939), 387 ffGoogle Scholar. esp. 393, n. 2, and for the Hellenistic world in general, Rostovtzeff, , Klio 1906, 249 ffGoogle Scholar. and SEHHW index s.v. ἀγγαρεῖαι, εἰσφορά, ἐπίταγμα, ἀνεπισταθμεία, ἐπισταθμεία, σταθμός, παρουσία. For transport requisitioned by the Seleucids see Josephus, AJ XIII. 52; I Macc. 10. 33. Evidence from Judaea at the time of the New Testament, Matth. 5. 41; 27. 32; Mark 15.21 (Luke 23. 26). In the late republic P. Ventidius and other equestrians provided mules and carriages to provincial governors and military commanders on a commercial basis, perhaps indicating that there was then no regular universal requisitioning procedure (Cicero, ad fam. X. 18. 3; Pliny, NH VII. 135; Aul. Gell. XV. 4. 3; Syme, R., ‘Sabinus the Muleteer’, Latomus XVII (1958), 73–80)Google Scholar.

¹⁵⁷ See van Berchem, D., L'Annone militaire dan l'empire romain au III^e siecle (1937)Google Scholar.

¹⁵⁸ Lesquler, op. cit. (n. 108) 352 f.; Wilcken, U.Grundzüge und Chrestomathie der Papyrusurkunde 1.1 (1912), 359 f.Google Scholar; MacMullen, op. cit. (n. 28), 85 f. Compare also the evidence of ‘Hunt's Pridianum (re-published by Fink, R. O., JRS XLVIII (1958) 102–16Google Scholar) which shows that soldiers were sent off to collect supplies for their unit both inside and outsid the province.

¹⁵⁹ n. 73.

¹⁶⁰ See nn. 101–3.

¹⁶¹ See nn. 18–19, 23–8.

¹⁶² Following the calculations of Davies, R. W., Britannia 11 (1971), 123Google Scholar.

¹⁶³ Provisioning and transport for large forces was known as παραπομπή or prosecutio, for which see Rostovtzeff, SEHRE², 723; Jones, A. H. M., The Greek City, 328, n. 89Google Scholar; Millar, F., A Study of Cassiu Dio (1964), 20–1Google Scholar. Careful plans were made for such movements, as they were for the passage of the emperor and his court. See the detailed account, valid at least for the late empire, in SHA, Sev. Alex. 47, comparing Ambrose, ad Ps. 118. 1, which may be approximately contemporary with the writing of the Historia Augusta. However, the practice certainly went back to the early empire (Suet., Tib. 38; Statius, , Silvae IV. 9. 17–19Google Scholar; Pliny, Pan. 20; Siculus Flaccus, de cond. agr. p. 169), and even to the republican period (Livy XLII. 1. 7–12). Cf. U. Wilcken, op. cit. (n. 158), 358, and Lesquier, op. cit. (n 108), 350 f. D. van Berchem has plausibly argued that the Antonine Itinerary was based on an edict of Caracalla designed to secure provisioning in this way for his journey to the east, op. cit. (n. 156), 164–87, reaffirmed recently in Actes du IX^e Congrès international d'études sur les frontières romaines (1972, publ. 1974), 301–7.

¹⁶⁴ See n. 158. Even when the orders to make requisitions came from above, soldiers would be responsible for executing them.

¹⁶⁵ cf. Marshall, A. J., Phoenix XX (1966), 231–46CrossRefGoogle Scholar; Burton, G. P., JRS LXV (1975), 92–106Google Scholar.

¹⁶⁶ See, e.g., Oliver, J. H., AJPhil LXXXVII (1966), 75–80Google Scholar, commenting on an inscription from Samothrace which lists the proconsul of Macedonia and his accompanying staff visiting the sanctuary of the Cabeiroi.

¹⁶⁷ D. 1.

¹⁶⁸ D. 13, 7–8; D. 15, 34; D. 16, 17–20; cf. Robert, OMS 1, 351.

¹⁶⁹ See n. 146.

¹⁷⁰ Eck, W., ‘Die Laufbahn eines Ritters aus Apri in Thrakien’, Chiron V (1975), 365–92Google Scholar.

¹⁷¹ See A. Stein, op. cit. (n. 149), 23–4. By coincidence he was clearly related to the author of D.2.

¹⁷² Dio LVII. 10. 5; Suet., Tib. 32. 2. Compare Tacitus' general observations at Ann. IV. 6.