¹ With the exceptions noted below, the system of citation employed in this paper conforms, for classical sources, with that of the Oxford Classical Dictionary, 2nd ed. (Oxford, 1970), pp. ix–xxiiGoogle Scholar, and for periodicals with the index of the relevant volume of L'année philologique: Abbott-Johnson, = Abbott, F. F. and Johnson, A. C., Municipal Administration in the Roman Empire (Princeton, 1926)Google Scholar; Comp. Stud. Soc. & History = Comparative Studies in Society and History; Econ. & Pol. Weekly = Economic and Political Weekly; Econ. Hist. Rev. = Economic History Review; ILAlg. = Gsell, S. (ed.), Inscriptions latines de l'Algérie (Paris, 1922–1976)Google Scholar; J. Econ. Hist. = Journal of Economic History; J. Peasant Stud. = Journal of Peasant Studies; Muson. (ed. Hense) = Musonius, Rufus, Reliquiae, ed. Hense, O. (Leipzig, 1905)Google Scholar; PFlor. = Vitelli, G. (ed.), Papiri Fiorentini (Milan, 1906–1915)Google Scholar; Smallwood = Smallwood, E. M., Documents Illustrating the Principates of Gaius Claudius and Nero (Cambridge, 1967)Google Scholar.

The following conversion factors are also utilized:

In the hope that this paper may prove useful to readers unfamiliar with the classical languages, translations have been provided by the author for those passages cited in extenso.

² cf. Columella, , Rust. 2. 9. 17Google Scholar: ‘Panic and millet should also be reckoned among the grain crops, although I have formerly classified them among the legumes, for in many locales the peasants subsist on foods made from them’; Pliny, , HN 18. 101Google Scholar: ‘the Gallic provinces, and especially Aquitania, also make use of panic, as do the districts of Italy along the Po, although they add to it beans without water… The Pontic peoples prefer panic to any other food’.

³ Food riots are recorded in 75 b.c. (Sall. H. fr. 2. 45 M), 57 (Cic. Dom. 6, 10–14; cf. Att. 4. 1. 6; QFr. 2. 3. 2; Har. Resp. 31; Red. Sen. 34; Plut. Pomp. 49. 5) and 40 (App. BCiv. 5. 67–8; Dio Cass. 48. 31), further shortages in 66 (Cic. Leg. Man. 44), 41 (App. BCiv. 5. 22, 25; Dio Cass. 48. 7. 4), 39 (BCiv. 5. 71), 38 (BCiv. 5. 92) and 36 (BCiv. 5. 99).

⁴ In 22 b.c. (Dio Cass. 54. 1. 2–3), a.d. 6 (Dio Cass. 55. 26. 1–3, 27. 1–3), 32 (Tac. Ann. 6. 13), 51 (Ann. 12. 43. 1; Suet. Claud. 18. 2) and 189 (Dio Cass. 73. 13; Hdn. 1. 12. 3–13. 6). Additional shortages are registered under Augustus in a.d. 5 (Dio Cass. 55. 22. 3) and 7 (Dio Cass. 55. 31. 4), under Tiberius in 19 (Ann. 2. 87), Claudius in 42 (Dio Cass. 60. 11. 1), and under Nero (Suet. Ner. 45), Otho (Tac. Hist. 1. 86), Antoninus (S. H. A. M. Ant. 8. 11, 9. 1) and Marcus Aurelius (Marc. 8. 4).

⁵ On which, see esp. Yeo, C. A., ‘Land and Sea Transportation in Imperial Italy’, TAPhA 77 (1946), 221–44Google Scholar; and Burford, A., ‘Heavy Transport in Classical Antiquity’, Econ. Hist. Rev. sec. ser. 13 (1960), 1–18Google Scholar. That land transport charges, while high, were not excessive for a society which relied upon oxcarts and predominantly unpaved roads, has been usefully pointed out by Clark, C. and Haswell, M., The Economics of Subsistence Agriculture, 3rd ed. (London, 1967), pp. 184–9Google Scholar. With the price of frumentum fixed at 100 denarii/castrensis modius, the freight charge of 20 denarii/1200 Roman lb./Roman mile decreed by Diocletian's Edict (17.3) may be expressed as a rate of 4.59 kg/ton/km (since they do not take into account that the Roman mile is only 94.7% of the English, the figure of 4.4 kg/ton/km given by Clark and Haswell must be adjusted upward). The median rate for the 34 examples of cart and wagon transport which they list is 3.4 kg/ton/km.

⁶ A circumstance most succinctly expressed by Gregory of Nazianzus: ‘there was a famine, the severest in fact in human recollection. The community languished, but help was not forthcoming in any degree nor was there any cure for the calamity. Coastal cities easily endure such shortages, since they of course dispose of what they have in abundance and in turn receive what they lack by sea. But those of us who dwell far from the sea derive no advantage from those things in which we abound nor can we obtain what we lack, as we can neither export what we have nor import what we need’ (Or. 43. 34). In brief but perceptive remarks, however, Kohns, H. P., Versorgungskrisen und Hungerrevolten im spätantiken Rom (Bonn, 1961), pp. 10–11Google Scholar; Jones, A. H. M., The Later Roman Empire 284–602, i (Oxford, 1964), pp. 445–6, ii, pp. 841–5Google Scholar; Brunt, P. A.. Italian Manpower 225 B.C.-A.D. 14 (Oxford, 1971), pp. 703–6Google Scholar; and Pékary, T., Die Wirtschaft der griechisch-römischen Antike (Wiesbaden, 1976), pp. 69–70, 91–2Google Scholar, have all correctly stressed the inadequacy of maritime as well as land transport. The periodic shortages in imperial Rome, and the difficulties which beset such coastal cities as Ariminum (CIL 11. 377) and Aspendus (Philostr. VA 1. 15), demonstrate alike that access to the sea did not guarantee immediate relief from famine.

⁷ cf. Strab. 5. 1. 12; Columella, Rust. 2. 9. 14, 19, 2. 10. 1, 22; Pliny, HN 18. 127, 141.

⁸ Rostovtzeff, M., The Social and Economic History of the Roman Empire, 2nd ed. rev. P. M. Fraser (Oxford, 1957), p. 598 n. 9Google Scholar; MacMullen, R., Enemies of the Roman Order: Treason, Unrest and Alienation in the Empire (Cambridge, Mass., 1967), pp. 249–54Google Scholar.

⁹ Egypt: Joseph. Ap. 2. 63; Tac. Ann. 2. 59; Suet. Tib. 52 (a.d. 19); Johnson, A. C., in Frank, T. (ed.), An Economic Survey of Ancient Rome, ii (Baltimore, 1936)Google Scholar, 18 (a.d. 45–7); P. Flor. 61 (between a.d. 48 and 88); Pliny, Pan. 31 (c. a.d. 100). North Africa: CIL 8. 1648, 9250, 15497, 25703–4, 26121; ILAlg. 1. 2145.

¹⁰ Since the agricultural routine in Egypt was without parallel in the rest of the Empire, the data from that province will be disregarded here.

¹¹ Scramuzza, V. M., in Frank, T. (ed.), Econ. Survey, iii (Baltimore, 1937), p. 260 n. 3Google Scholar, has argued from the high bids of Apronius (216, 000 modii) and Minucius (246,000) for the tithe of Leontini that the yield must actually have approached sixteenfold. Cicero is adamant, however, that Apronius would have suffered a sizeable loss on the contract if his extortionate activities had not enjoyed the support of Verres (2. 3. 110–17).

¹² Rust. 1. 44. 2: ‘they say that at Sybaris in Italy the yield is regularly even one hundredfold, and likewise at Gadara in Syria and Byzacium in Africa’.

¹³ Rust. 3. 3. 4: ‘for we can hardly recall a time when, at least in the greater part of Italy, grain crops responded with a fourfold yield’. Pliny dedicated the Naturalis Historia to Titus in a.d. 77 (praef. 3), and Columella composed the third book of his De Re Rustica c. a.d. 60–5; cf. Reitzenstein, R., De Scriptorum Rei Rusticae Libris Deperditis (Diss. Berlin, 1884), p. 31Google Scholar; Becher, W., De L. Iuni Moderati Columellae Vita et Scriptis (Diss. Leipzig, 1897), p. 11Google Scholar; A. Kappelmacher, ‘Iunius’ no. 104, RE 10 (1917), 1059Google Scholar; and Schanz, M., Geschichte der römischen Literatur, 4th ed. rev. C. Hosius, ii (Munich, 1935), pp. 787–8Google Scholar.

¹⁴ White, K. D., ‘Wheat Farming in Roman Times’, Antiquity 37 (1963), 207–12Google Scholar. The Sicilian harvest in 1959 averaged 10.6 quintals/hectare, that of 70 b.c. 11.2q./ha.; the comparable figures for Tuscany are 19.9 q./ha. (1959) and 17–21 (Varro), depending upon the seed ratio. This argument should be taken very seriously. Deterioration of climate and soil may compensate for whatever technical advances have occurred in the interim; see especially Finley, M. I., Ancient Sicily to the Arab Conquest (London, 1968), pp. 4–5Google Scholar. Nor should such progress itself be taken for granted; cf. K. D. White, supra p. 209; and Frayn, J. M., Subsistence Farming in Roman Italy (London, 1979), p. 13Google Scholar.

¹⁵ Brunt, P. A., JRS 62 (1972), 158Google Scholar. Cf. Duncan-Jones, R., The Economy of the Roman Empire (Cambridge, 1974), p. 49 n. 4Google Scholar, who has placed these passages in useful perspective.

¹⁶ Nor, for that matter, to any other particular region. On the continuing importance of grain production in Apulia, for example, see Varro, , Rust. 1. 2. 6Google Scholar; Strab. 6. 3. 9; Columella, , Rust. 3. 8. 4Google Scholar.

¹⁷ Frank, T., Econ. Survey, v (Baltimore, 1940), p. 141Google Scholar; White, K. D., Antiquity 37 (1963), 209Google Scholar; Roman Farming (London, 1970), p. 49Google Scholar. Cf., on intercultivation, Rust. 2. 9. 6, 5. 6. 11, 7. 3, 9. 7, 9. 12–13, 10. 5, 11. 2. 54; Arbor. 162.

¹⁸ Its exact fate is unknown. For the sources on this episode see Broughton, T. R. S., The Magistrates of the Roman Republic, ii (New York, 1952), p. 168Google Scholar. On the contents of the bill cf. Hardy, E. G., Some Problems in Roman History (Oxford, 1924), pp. 68–98Google Scholar; Afzelius, A., ‘Das Ackerverteilungsgesetz des P. Servilius Rullus’, C&M 3 (1940), 214–35Google Scholar; and Gruen, E., The Last Generation of the Roman Republic (Berkeley, 1974), pp. 388–96Google Scholar.

¹⁹ On Caesar's agrarian legislation of 59 b.c., cf. Cary, M., ‘The Land Legislation of Caesar's First Consulship’, JPh 35 (1920), 174–90Google Scholar; Gruen, E., LGRR, pp. 397–401Google Scholar.

²⁰ That 20,000 new settlers could not have been planted in the ager Campanus has, however, been convincingly argued by Levi, M. A., ‘Una pagina di storia agraria romana’, A&R 3 (1922), 239–52Google Scholar; cf. ‘I confini dell'Agro Campano’, AAT 57 (1921–1922), 604–16Google Scholar. In 51 b.c. land was still available there for distribution (Cic. Fam. 8. 10. 4)-a telling point.

²¹ Cicero returns repeatedly to this theme, even when addressing the assembly (cf. Leg. Agr. 80, 81, 83).

²² The first on record is that of Scipio Africanus at Liternum in 184 b.c. (Livy 38. 52. 1). On the Roman villae here in the second century b.c. see D'Arms, J. H., Romans on the Bay of Naples (Cambridge, Mass., 1970), pp. 1–17Google Scholar.

²³ Including Caesar himself. The evidence is conveniently arranged in D'Arms, op. cit. pp. 171–201; cf. Shatzman, I., Senatorial Wealth and Roman Politics (Brussels, 1975), pp. 440–3, 457Google Scholar.

²⁴ Clark, C. and Haswell, M., Econ. Subsistence Agr., pp. 7–8, 53–4Google Scholar. Their data actually imply specific termini of 188 and 232 kg/person/year (the figures which appear in the text have been rounded off). The normal extraction rate for triticum and unmoistened siligo seems to have been very close to ten per cent; cf. Jasny, N., ‘Wheat Prices and Milling Costs in Classical Rome’, Wheat Studies of the Food Research Institute, Stanford University 20 (1944), 154Google Scholar; and Moritz, L. A., Grain Mills and Flour in Classical Antiquity (Oxford, 1958), pp. 184–209Google Scholar.

²⁵ Econ. Subsistence Agr., pp. 54, 70 and 223, where the nutritional utility of other common food substances relative to wheat has been conveniently tabled.

²⁶ Wild and Cultivated Plants: A Note on the Peasant Economy of Ancient Italy', JRS 65 (1975), 32–9Google Scholar = Subsistence Farming in Roman Italy, pp. 57–72.

²⁷ Περ⋯ εὐχυμ⋯ας κα⋯ κακοχυμ⋯ας τροɸ⋯ν, ed. Kühn, K. G., Medicorum Graecorum Opera, vi (Lipsiae, 1823), pp. 748–9Google Scholar.

²⁸ On the questions of authorship and originality cf. Kroll, W., ‘Moretum’, RE 16 (1933), 298–9Google Scholar; and Schanz-Hosius ii⁴, pp. 89–90, and the literature cited therein.

²⁹ Such crop diversity was by no means peculiar to the Roman peasantry. A similar pattern prevailed, for example, at Kerkeosiris in Ptolemaic Egypt. In 119/18 b.c. wheat accounted for 62 per cent of the cleruchic land under cultivation in the village, lentils 18 per cent and beans 2 per cent. In 116/15 wheat was still the most significant crop (54 per cent), while the relative importance of beans (20 per cent) and lentils (8 per cent) had been reversed; see Crawford, D. J., Kerkeosiris: An Egyptian Village in the Ptolemaic Period (Cambridge, 1971), p. 186Google Scholar.

³⁰ 0·67 Attic medimnus = 3·08 modii; for the conversion factor see n. 1 above.

³¹ Rust. 56. In ch. 149, Cato cites the period 1 September-1 March as appropriate for the leasing of winter pasturage. If this is what he broadly meant by the winter season, then his labourers received 24 modii (162 kg) in the winter and 27(182 kg) in the summer, or 344 kilograms in all.

³² Varro, , Rust. 1. 44. 1Google Scholar; Columella, , Rust. 2. 9. 1–2, 5, 11. 2. 75Google Scholar; Pliny, , HN 18. 198–200Google Scholar.

³³ With a sowing ratio of four modii/iugerum, the return would be marginally higher (5. 25–6. 1: 1).

³⁴ cf. Cic. Leg. Agr. 1. 21, 2. 80; Varro, Rust. 1. 2. 6; Strab. 5. 4. 3; Columella, , Rust. 3. 8. 4Google Scholar; Pliny, , HN 18. 86, 111Google Scholar.

³⁵ The point is not new; cf. Frank, T., Econ. Survey, v, p. 141Google Scholar; and Duncan-Jones, R., Econ. Roman Empire, p. 52 n. 1Google Scholar.

³⁶ cf. Cato, , Rust. 35. 2Google Scholar; Varro, , Rust. 1. 9. 6, 44. 2–3Google Scholar; Columella, , Rust. 2. 9. 15, 10. 4, 10. 6, 10. 31 (continuous cropping)Google Scholar; Verg. G. 1. 73–6; Strab. 5. 4. 3; Pliny, , HN 18. 111, 187, 191Google Scholar. Soil so worked was normally termed ager restibilis.

³⁷ The brief comments of White, K. D., Roman Farming, pp. 126, 136–7Google Scholar; and Brunt, P. A., JRS 62 (1972), 157–8Google Scholar, strike to the core of the problem.

³⁸ cf. Varro, , Rust. 3. 16. 33Google Scholar; Verg. G. 1. 71–2; Columella, , Rust. 2. 9. 4, 10. 7Google Scholar; Pliny, , HN 18. 187Google Scholar.

³⁹ Chayanov, A. V., Peasant Farm Organization, in Thorner, D., Kerblay, B. and Smith, R. E. F. (eds.), The Theory of Peasant Farming (Homewood, Ill., 1966), pp. 64–5, 68Google Scholar.

⁴⁰ cf., recently, McKinnon, J., ‘Chayanov in the Solomons’, Pacific Viewpoint 17 (1976), 49–60Google Scholar; Shrivkumar, S. S., ‘Family Size, Consumption Expenditures, Income and Land Holding in an Agrarian Economy’, Econ. & Pol. Weekly 11 (1976), 1115–24Google Scholar; and Hunt, D., ‘Chayanov's Model of Peasant Household Resource Allocation’, J. Peasant Stud. 6 (1979), 247–85Google Scholar.

⁴¹ cf. Dion. Hal. Ant. Rom. 9. 51. 6.

⁴² See Powell, J. E., The Rendel Harris Papyri (Cambridge, 1936), pp. 2–5Google Scholar; ‘Musonius Rufus: ΕΙ ΠΑΝΤΑ ΤΑ ΓΙΝΟΜΕΝΑ ΤΕΚΝΑ ΘΡΕΠΤΕΟΝ in P. Harr. 1’, APF 12 (1936), 175–8Google Scholar.

⁴³ Div. Inst. 6. 20. 24–5: ‘for parricides bewail their straitened means and assert that they cannot provide for bringing up several children: as if wealth is truly in the power of its possessors, or as if God does not daily make the rich poor and the poor rich’.

⁴⁴ For additional references to abortion, exporsure and infanticide among the poor or as general practice, cf. Philo, , Spec. Leg. 3. 110–19Google Scholar; Virt. 131; Plut. Mor. 497E; Martyr, Justin, Apol. 27, 29Google Scholar; Athenagoras, , Leg. pro Christ. 35. 6Google Scholar; Clem. Al. Παιδαγωγ⋯ς 2. 10. 96. 1, 3. 3. 21. 5; Tert. Apol. 9. 6–8; Ad Nat. 1. 15. 3–4, 16. 10; Min. Fel. Oct. 30. 2, 31. 4; Lactant. Div. Inst. 5. 9. 15; Hieron. Ep. 22. 13. On the sale of ingenui see Brunt, P. A., JRS 48 (1958), 167–8Google Scholar, and to the list of references given there add Cod. lust. 4. 43. 2. The extent to which family limitation methods were practised in Roman society, and their historical significance, have been frequently discussed in recent years; cf. van Geytenbeek, A. C., Musonius Rufus and Greek Diatribe, trans. Hijmans, B. L. Jr (Assen, 1963), pp. 78–89Google Scholar; Hopkins, M. K., ‘Contraception in the Roman Empire’, Comp. Stud. Soc. & History 8 (1965–1966), 124–51Google Scholar; Brunt, P. A., Italian Manpower, pp. 146–54Google Scholar; Salmon, P., Population et dépopulation dans l'empire romain (Brussels, 1974), pp. 61–76Google Scholar.

⁴⁵ cf., inter alia, Pl. Resp. 372c; Leg. 744d–e; Polyb. 36. 17; Muson. fr. 15B (ed. Hense); Tac. Hist. 5. 5.

⁴⁶ A partible system of inheritance, with equal division of the estate among all the sons, seems to have prevailed throughout archaic and classical Greece. The unhappy condition of our sources, however, permits only an occasional glimpse of the resultant tension within the family, while overt displays of the pressure to control the number of their offspring which many heads of household must have experienced are rarer still – as anyone who reads Walcot, P., Greek Peasants, Ancient and Modern (Manchester, 1970), pp. 45–56Google Scholar, or Burford-Cooper, A., ‘The Family Farm in Greece’, CJ 73 (1977/1978), 162–75Google Scholar, will readily appreciate. Roman law, by means of such practices as bonorum possessio contra tabulas testamenti and querella inofficiosi testamenti, also encourged partibility, indeed mandated it in the event of intestacy, which Daube, D., ‘The Preponderance of Intestacy at Rome’, Tulane Law Review 39 (1964–1965), 253–62Google Scholar, has persuasively argued to have been the rule rather than the exception. If strictly applied, as in Columella, , Rust. 4. 3. 6Google Scholar, the impact of such regulations upon the Roman peasantry would have been grievous indeed. Law and custom, however, do not necessarily coincide, and in Dio Chyrs. Or. 7. 10 we may observe one of several possible compromises between the conflicting demands of law, natural sentiment and economic necessity: the co-residence of two families on one plot of land (a joint-family household). If the ancient evidence for this problem is intractable (the Roman is worse than the Greek), that from peasant societies of more recent vintage may be profitably examined to gauge its dimensions. The classical historian might begin with Homans, G., English Villagers of the Thirteenth Century (Cambridge, Mass., 1941)Google Scholar; Habakkuk, H. J., ‘Family Structure and Economic Change in Nineteenth-Century Europe’, J. Econ. Hist. 15 (1955), 1–12Google Scholar; Bourdieu, P., ‘Célibat et condition paysanne’, Etudes rurales 5–6 (1962), 32–135Google Scholar; ‘Les stratéges matrimoniales dans le système de reproduction’, Annales (ESC) 27 (1972), 1105–27Google Scholar; Pingaud, M.-C., ‘Terres et families dans un village du Châtillonnais’, Etudes rurales 42 (1971), 52–104Google Scholar; Berkner, L. K., ‘Inheritance, Land Tenure and Peasant Family Structure: A German Regional Comparison’, in Goody, J., Thirsk, J. and Thompson, E. P. (eds.), Family and Inheritance in Rural Western Europe: 1200–1700 (Cambridge, 1976), p. 71–95Google Scholar; ‘Peasant Household Organization and Demographic Change in Lower Saxony (1689–1766)’, in Lee, R. (ed.), Population Patterns in the Past (New York, 1976), pp. 53–69Google Scholar; and Berkner, L. K. and Mendels, F. F., ‘Inheritance Systems, Family Structure, and Demographic Patterns in Western Europe, 1700–1900’, in Tilly, C. (ed.), Historical Studies of Changing Fertility (Princeton, 1976), pp. 209–23Google Scholar.

⁴⁷ It is unclear whether the initiative was taken by Nerva, (Epit. de Caes. 12. 4)Google Scholar or Trajan (S. H. A. Hadr. 7. 8; Pert. 9. 3).

⁴⁸ Epit. de Caes. 12. 4: ‘he [Nerva] ordered girls and boys born to indigent parents in the municipalities in Italy to be brought up at public expense’.

⁴⁹ If, as Szilágyi, J., ‘Prices and Wages in the Western Provinces of the Roman Empire’, AAntHung. 11 (1963), 336–7Google Scholar; and Duncan-Jones, R., Econ. Rom. Empire, pp. 145–6, 345–7Google Scholar, have recently argued, the market-price of wheat in Italy of the first century a.d. was regularly HS 3–4/modius, then the allowance for puellae was equivalent to 243–324 kg/year (36–48 modii)-which is again in conformity with the modern definition of subsistence. For the mechanics of the alimenta, and particularly for the complex and only partially comprehensible estate valuation and loan allocation procedures, see now Veyne, P., ‘La table des Ligures Baebiani et l'institution alimentaire de Trajan’, MEFR 69 (1957), 81–135Google Scholar; 70 (1958), 177–241; Bourne, F. C., ‘The Roman Alimentary Program and Italian Agriculture’, TAPhA 91 (1960), 47–75Google Scholar; and Duncan-Jones, R., supra pp. 294–315Google Scholar. To the list of 49 programmes compiled by Duncan-Jones (p. 340), add now Trebula Suffenas (AE 1972, 167) and Cherchio (AE 1975, 295).

⁵⁰ Thus recently Veyne, P., MEFR 70 (1958), 228Google Scholar; ‘Les “Alimenta” de Trajan’, Les empereurs romains d'Espagne (Paris, 1965), pp. 169–70Google Scholar; Duncan-Jones, R., ‘The Purpose and Organisation of the Alimenta’, PBSR n.s. 19 (1964), 127, 130Google Scholar; Econ. Rom. Empire, p. 295; Garnsey, P., ‘Trajan's Alimenta: Some Problems’, Historia 17 (1968), 381Google Scholar; Brunt, P. A., Italian Manpower, p. 5Google Scholar; Salmon, P., Population et dépopulation, p. 57Google Scholar.

⁵¹ cf. Bourne, F. C., TAPhA 91 (1960), 68–72Google Scholar; Ford, G. B., ‘The Letters of Pliny the Younger as Evidence of Agrarian Conditions in the Principate of Trajan’, Helicon 5 (1965), 386–7Google Scholar.

⁵² cf. Veyne, P., Emp. rom. d'Espagne, p. 171Google Scholar; and Duncan-Jones, R., Econ. Rom. Empire, pp. 306–7Google Scholar.

⁵³ Duncan-Jones, R., Econ. Rom. Empire, p. 301Google Scholar. This may also account for the competition of private (CIL 14. 4450) and public (14. 298, 4664) initiatives at Ostia.

⁵⁴ cf. also the fragmentary inscription CIL 9. 5825 (Auximum).

⁵⁵ cf. Pan. 26. 5–6, 27. 1 and 28. 5.

⁵⁶ A point so far rendered explicit only by Heitland, W. E., Agricola (Cambridge, 1921), p. 273Google Scholar; and Kahrstedt, U., Kulturgeschichte der römischen Kaiserzeit, 2nd ed. rev. (Bern, 1958), pp. 56–7Google Scholar. Among scholars who have examined the alimenta in depth, the accounts of Sirago, V. A., L'Italia agraria sotto Traiano (Louvain, 1958), pp. 276–89Google Scholar; and Hands, A. R., Charities and Social Aid in Greece and Rome (London, 1968), pp. 108–15Google Scholar, offer the most prudent discussions of this question.

⁵⁷ For a sobering introduction to this topic see Brunt, P. A., ‘Charges of Provincial Maladministration under the Early Principate’, Historia 10 (1961), 189–227Google Scholar, who has argued convincingly that the transition from Republic to Principate did little to ameliorate Rome's abusive standards of provincial administration.

⁵⁸ Cic. Phil. 3. 31: ‘he empties storerooms, and slaughters herds of cattle and whatever other beasts he obtains; his soldiers feast, while he himself, imitating his brother, besots himself with wine. The fields are laid waste and villas plundered, mothers of families, virgins and free-born boys are carried off and handed over to the soldiers. These same things, wherever he led his army, were done by Marcus Antonius’; S. H. A., Aurel. 7. 5–6Google Scholar: ‘if you wish to be a tribune – rather, if you wish to stay alive, restrain the hands of your soldiers. No one should seize another's fowl or lay hands upon his sheep. No one should carry off grapes, thresh out corn, extort oil, salt or firewood: each should be content with his own allowance. The soldiers should make their living from the spoils of the enemy, not from the tears of the provincials.’

⁵⁹ The principal military highway from Sirmium to Aquileia; see Mócsy, A., ‘Pannonia’, RE supp. 9 (1962), 661Google Scholar.

⁶⁰ If the chronology of events in a.d. 238 remains a vexed problem, this point at least seems secure; cf. Townsend, P. W., ‘The Chronology of the Year 238 a.d..’, YCIS 1 (1928), 231–8Google Scholar; van Sickle, C. E., ‘Some Further Observations on the Chronology of the Year 238 a.d.’, CPh 24 (1929), 285–9Google Scholar; Vitucci, G., ‘Sulla cronologia degli avvenimenti dell 238 d. C.’, RFIC 32 (1954), 372–82Google Scholar; Carson, R. A. G., Coins of the Roman Empire in the British Museum, vi (London, 1962), 98, 101Google Scholar; and Loriot, X., ‘Les Fasti Ostienses et les Dies Imperii de Gordien III’, Mélanges d'historie offerts à William Seston (Paris, 1974), pp. 297–312Google Scholar; ‘Les premières années de la grande crise du III^e siècle: De l'avènement de Maximin le Thrace (235) à la mort de Gordien III (244)’, in Temporini, H. and Haase, W. (eds.), Aufstieg und Niedergang der römischen Welt, ii, 2 (Berlin, 1975), 720–2Google Scholar.

⁶¹ Extensive damage also ensued during the course of the siege itself (Hdn. 8. 4. 8).

⁶² Hdn. 8. 6. 5–8; S. H. A., Max. 24. 4–8Google Scholar.

⁶³ In a.d. 69 the eminent Othonian officer Suetonius Paulinus advised that emperor to decline battle with the Rhine legions which had invaded Italy in support of L. Vitellius' bid for the throne. His reasoning, in this context, is especially revealing: ‘Italy north of the Po, shut in by the Alps and without any possibility of relief from the sea, has been exhausted by the very passage of their army; nowhere is there any corn for their forces, and an army cannot be maintained without supplies’ (Tac. Hist. 2. 32).

⁶⁴ OGI 665 = IG Rom. 1. 1262 = White, H. G. Evelyn and Oliver, J. H., The Temple of Hibis in El Khargeh Oasis, Part II: Greek Inscriptions (New York, 1938), pp. 1–19Google Scholar = SEG 20. 694 = Smallwood 382.

⁶⁵ cf., inter alia, P London 1171 = Abbot-Johnson 162 = Smallwood 381 (a.d. 42); and PSI 446 = Hunt, A. S. and Edgar, C. C., Select Papyri, ii (Cambridge, Mass., 1934), no. 221 (a.d. 133–7)Google Scholar.

⁶⁶ See Robert, L., ‘Sur un papyrus de Bruxelles’, RPh 17 (1943), 111–19Google Scholar.

⁶⁷ cf. OGI 609 = IG Rom. 3. 1119 = Abbott-Johnson 113 (Phaena, Syria).

⁶⁸ CIL 3. 12336 = IG Rom. 1. 674 = Abbott-Johnson 139 (a.d. 238).

⁶⁹ Abbott-Johnson 142.

⁷⁰ cf. CIL 3. 14191 = OGI 519 = IG Rom. 4. 598 = Abbott-Johnson 141 (a.d. 244/7); Abbott-Johnson 143, 144.

⁷¹ SEHRE ², p. 495, but the theme is recurrent; cf. pp. xiii, 459, 463–6, 495–501. For the evidence so far reviewed, cf. the masterly notes to these chapters, and MacMullen, R., Soldier and Civilian in the Later Roman Empire (Cambridge, Mass., 1963), pp. 85–9Google Scholar. The problem of civilian attitudes to the Roman army has recently been resumed by Alföldy, G., ‘Soziale Konflikte im römischen Kaiserreich’, Heidelberger Jahrbücher 20 (1976), 111–25Google Scholar; and Kolb, F., ‘Der Aufstand der Provinz Africa Proconsularis im Jahr 238 n. Chr.’, Historia 26 (1976), 440–78Google Scholar; ‘Wirtschaftliche und soziale Konflikte im römischen Reich des 3 Jahrhunderts n. Chr.’, Bonner Festgabe Johannes Straub (Bonn, 1977), pp. 277–95Google Scholar.

⁷² Περ⋯ εὐχυμ⋯ας κα⋯ κακοχυμ⋯ας τροɸ⋯ν (Kühn, K. G., Medic. Graec. Opera, vi, pp. 748–9Google Scholar). For the date, see Ilberg, J., ‘Ueber die Schriftstellerei des Klaudios Galenos’, RhM 51 (1896), 187Google Scholar.

⁷³ I wish to express my appreciation to the University of Minnesota for the award of a leave grant which greatly facilitated the appearance of this paper, and to the faculty of the Seminar für Alte Geschichte der Universität Heidelberg, particularly Professor G. Alföldy, for their patience and advice.