¹ Xen., Hell. 6. 1. 11.

² See Section III below for discussion and references.

³ Polyb. 4. 38. 5: . Rostovtzeff, M.The Social and Economic History of the Hellenistic World (1941), 1462, n. 20Google Scholar cites in support of this passage an inscription, which records the importation of grain to Istros, published in Lambrino, S., ‘Fouilles d'Histria’, Dacia 3–4 (1927–1932), 400 ffGoogle Scholar.

⁴ Matley, J. M., ‘The Geographical Basis of Greece’, in Grothusen, K.-D., ed., Griechenland. Sudosteuropea Handbuch, Band III (1980), 208Google Scholar; Admiralty Handbook, Greece (1945), 83.

⁵ Philippson, A., Die Griechischen Landschaften. Teil I. Der Nordosten der Griechischen Halbinsel. Thessalien und die Spercheios-Senke (1950), 127–49Google Scholar.

⁶ A very good discussion of the topography and agroclimate of Thessaly can be found in Halstead, P., ‘From Determinism to Uncertainty: Social Storage and the Rise of the Minoan Palaces’, in Sheridan, A. and Bailey, G., eds., Economic Archaeology. Towards an Integration of Ecological and Social Approaches (1981), 199Google Scholar.

⁷ See Table 1 below.

⁸ de Vooys, A. C., ‘Western Thessaly in Transition’, Tijdschrift van het Koninklijk Nederlandsch Aardrijkskundig Genootschap 75 (1959), 31, 35Google Scholar.

⁹ Levitt, J., Responses of Plants to Environmental Stress. Vol. 2. Water, Radiation, Salt and Other Stresses (1980), 214–18Google Scholar.

¹⁰ de Vooys (see n. 8), 32, 34; Sivignon, M., ‘The Demographic and Economic Evolution of Thessaly (1881–1940)’, in Carter, F. W., ed., An Historical Geography of the Balkans (1977), 388, fig. 7Google Scholar.

¹¹ (1974).

¹² de Vooys (see n. 8), 32: the number of days with frost fluctuates between 11 and 62. We have adopted the following stance in regard to the problem of climatic change and the use of modern data. Because the analysis is focused on the spatial and temporal pattern of variability and the relationship between it and crop yield fluctuations, long series of modern climatic data, incorporating numerous small secular shifts, can be used to calculate the parameters of climate and climatic variability, which may act as guide to past conditions. For a longer discussion of climate, variability and climatic change, see Gallant, T. W., An Examination of Two Island Polities in Antiquity: the Lefkas-Pronnoi Survey (Ph.D. Dissertation, 1982), 2–7Google Scholar.

¹³ Sivignon (see n. 10), fig. 6. Evapotranspiration is the amount of moisture lost from the ground surface through evaporation and from vegetation through transpiration. Potential Evapotranspiration represents the water loss from an extensive, closed, homogeneous cover of vegetation that never suffers from a lack of water. It is a measure of the amount of moisture which could be lost. Actual Evapotranspiration is the amount which is lost. Once Potential Evapotranspiration exceeds precipitation, unless there are other sources of moisture, soil moisture deficits begin. For a general discussion of these terms, see Mather, J. M., Climatology: Fundamentals and Applications (1974), 58Google Scholar.

¹⁴ All the data are gathered from the Στατιστική Ἐπετηρίς τῆς Ἑλλάδος. The higher the figure, the stronger the correlation. A figure of 1 indicates a perfect correlation.

¹⁵ Arnon, I., Crop Production in Dry Regions. Vol. 2. Systematic Treatment of the Principal Crops (1972), 235–7Google Scholar; Theophrastus, de Hist. Plant., 8. 1.4.

¹⁶ Arnon, 237–9.

¹⁷ Arnon, 239–40.

¹⁸ Τσοποτοῦ (see n. 11) 165; Lawless, R. I., ‘The Economy and Landscape of Thessaly during Ottoman Rule’, in Carter, F. W., ed., An Historical Geography of the Balkans (1977), 514Google Scholar.

¹⁹ IG IX, 2, 1104 (Magnesia); AE 1910, col. 345, no. 3 (Larisa); Helly, B., Gonnoi 2 (1973), 41Google Scholar.

²⁰ SEG IX, 2 + = Tod, GHI 11, 196.

²¹ IG IX, 2, 1029, 1093.

²² Xen., Hell. 5. 4. 56.

²³ Livy 42. 64 ff.; cf. App., Mac. fr. 18. 3. Other refs. from the first century B.C. in Westlake, H. D., Thessaly in the Fourth Century B.C. (1935), 6 n. 1Google Scholar.

²⁴ M. Segrè, ‘Grano di Tessaglio a Coo’, Riv. Fil. 12 (1934), 169 ff.; Sherwin-White, S., Ancient Cos (1978), 110 nCrossRefGoogle Scholar.

²⁵ Philostr., vit. soph. 526–7. Cf. Ephippus, fr. 1 (Kock) for an earlier period. Xen., Hell. 6. 1. 11 contains a comparison of Thessaly and Attika.

²⁶ Dr Gallis will publish an epigraphical commentary in the Acts of the Conference. With that in mind we have said nothing of an epigraphical nature about the inscription and have printed the text of Gallis without alteration. Our translation, however, reflects a preference for ἐν[δυν]άτωι in 1. 9 (which we owe to Professor Roesch and the Lyon Colloque) and Ẻως for ἐκ in 1. 22.

²⁷ RE s.v. ‘Caecilius’ no. 94.

²⁸ Astin, A. E., The Lex Annalis before Sulla (1958)Google Scholar.

²⁹ RE s.v. ‘Caecilius’ no. 81.

³⁰ Livy 44. 45. 3; 45. 1.

³¹ RE s.v. ‘Caecilius ‘no. 72.

³² In addition to the references that follow see Livy 10. 11. 9; cf. 10. 13; 23. 41. 7; 38. 35. 5; Cic., de off. 2. 17. 58.

³³ Briscoe, J., A Commentary on Livy Books XXXI–XXXIII (1973), ad 31. 4. 6Google Scholar; Rickman, G., The Corn Supply of Ancient Rome (1980), 150Google Scholar.

³⁴ Livy 36. 4. 5–9.

³⁵ Rickman (see n. 33), 36 without the benefit of this inscription thought that aediles were not involved in the procurement of grain. But cf. Flaminius' activity in 196: Livy 33. 42. 8.

³⁶ Brunt, P. A., Italian Manpower (1971), 427–8Google Scholar.

³⁷ App., Lib. 75.

³⁸ Polyb. 36. 2. 1–3, with F. W. Walbank, Commentary, ad loc.; App., Lib. 68 ff. On the chronology, Astin, A. E., Scipio Aemilianus (1967), 49–51, 270–2Google Scholar; Harris, W. V., War and Imperialism (1979), 235 ffGoogle Scholar.

³⁹ Polyb. 6. 38. 3, with R. P. Duncan-Jones, ‘The Choinix, the Artaba and the Modius’, ZPE 21 (1976), 46–7, n. 16.

⁴⁰ Hultsch, F., Metrologicorum Scriptorum Reliquiae (1864–1866) 1, 206, 320Google Scholar; Griechische und Römische Metrologie (1882), 542–4. The small size of the kophinos confirmed by IG VII, 2712. 65: a decree of Akraiphia in honour of Epaminondas, benefactor in the mid-first century A.D. (cf. IG VII, 2711). Among other benefactions (ll. 63–6), at a festival he gave each person present one kophinos of grain (sitos) and I hemina of wine. Attic heminae are equivalent to Roman heminae; Cato gave 1 to 3 heminae per day to his slaves: presumably therefore 1 kophinos of grain was meant only to cover the days of the festival. (If 2 · 5 modii last a man 1 month, then 1 kophinos lasts him only c. 13 days.)

⁴¹ On the Thessalian calendar, see RE s.v. ‘Thessalien’; Helly, , Gonnoi 1 (1973), 137–8Google Scholar; Samuel, A. E., Greek and Roman Chronology (1970), 83Google Scholar. For the safe sailing period, see Casson, L., Ships and Seamanship in the Ancient World (1971), 270–2Google Scholar.

⁴² Brunt (see n. 36), 384.

⁴³ Foxhall, L. and Forbes, H. A., ‘Σιτομετρία: The Role of Grain as a Staple Food in Classical Antiquity’, Chiron 12 (1982), 41–90Google Scholar, especially 51–65. 2 · 5 modii per month would have provided around 72 per cent of the daily caloric requirement of their hypothetical household (see 49 n. 26). Note their conjecture that cereals ‘normally’ provided 70–75 per cent of human energy needs in antiquity (75).

⁴⁴ Livy 36. 4. 5–6. Textual corruption does not permit us to restore with confidence the amount of wheat offered by Carthage: cf. Briscoe, J., A Commentary on Livy Books XXXIV–XXXVII (1981), ad 36. 4. 5 (p. 225)Google Scholar.

⁴⁵ Sivignon (see n. 10), 388–404; Lawless (see n. 18), 515–17; de Vooys (see n. 8), 33–4.

⁴⁶ SIG ³ 543 = Austin, M. M., The Hellenistic World from Alexander the Great to the Roman Conquest (1981), no. 60Google Scholar.

⁴⁷ Habicht, C., ‘Eine hellenistische Urkunde aus Larisa’, in Milojcic, V. and Theocharis, D., eds., Demetrias 1 (1976), 157–74Google Scholar; F. Salviat et C. Vatin, ‘Le Cadastre de Larissa’, BCH 98 (1974), 247–62. Habicht's objections to associating these fragments with the earlier decrees seem to be based primarily on their date–c. 15–20 years later. But it is frequently the case that problems occur following land distribution precisely at this point of intergenerational transmission of property; in addition to the case of nineteenth-twentieth-century Thessaly, see as examples: Kiray, M. and Hinderink, J., ‘Interdependencies between Agroeconomic Development and Social Change: A Comparative Study Conducted in the Cukurova Region of Southern Turkey’, Journal of Development Studies 4 (1968), 497–528Google Scholar; White, C., Patrons and Partisans: a Study of Politics in Two South Italian Towns (1980), 13 ffGoogle Scholar. From the evidence of the best preserved fragment (containing a list comprising a name, patronymic, and a record of land), the mean holding was 6 · 2 ha (70 · 5 plethra), the mode (i.e. the most frequently occurring number) was 4·38 (50 plethra), and the median (i.e. the mid-point between the highest entry, 21·9 ha (250 plethra), and the smallest, 0·87 ha (1 plethra) was 10·5 ha (120 plethra).

⁴⁸ Sivignon (see n. 10), 379, 382.

⁴⁹ Parenthetically, a comparison of the amount of wheat sent only to Larisa, Atrax and Meliboia in the 320s with the amount Thessaly sent to Rome in 151–150 (53,100 compared to 32,250 quintals) provides a clear indication of the possible magnitude of a drought-induced deficit in Thessaly.

⁵⁰ Stählin, F., Die hellenische Thessalien (1924)Google Scholar; Stillwell, R., MacDonald, W. L. and McAllister, M. H., eds., The Princeton Encyclopedia of Classical Sites (1976)Google Scholar; Simpson, R. Hope and Dickinson, O. T. P. K., A Gazetteer of Aegean Civilization in the Bronze Age. Vol. 1. The Mainland and the Islands (1979), 272–98Google Scholar.

⁵¹ Livy 32. 37. 2.

⁵² Livy 31. 19. 4. Cf. Briscoe (see n. 33), ad 31. 19. 4 (p. 108): ‘Massinissa himself pays for the transport, presumably the ships are his own. Sage mistranslates curavit as “supervised”’. Also Livy 43. 6. 11–14.

⁵³ For example, at around the same time as the present episode, there is evidence that the Thessalian polis of Krannon had to resort to public subscription to pay off its debts; Moretti, L., Iscrizioni Storiche Ellenistiche 11 (1976), 99Google Scholar; Austin (see n. 46), no. 103.

⁵⁴ Personal communication from Professor Franke, referring to his article in Schweiz. Münzblätter 35 (1959), 61 ff.Google Scholar, at 67. We are extremely grateful for his expert advice.

⁵⁵ For wheat prices in the Aegean area in the second century B.C. we follow the table of prices given by Heichelheim, ‘Sitos’, RE suppl. VI (1935), 819–92. We assume that the medimnoi were all on the Attic-Sicilian standard, holding c. 40 kg of wheat, that the drachmai were all on the Attic-Alexandrian standard, containing 4·37 g silver, and that the Megalopolitan stater was on the Korinthian standard, equivalent to an Attic didrachm.

The average of all the prices given is 76·48 g of silver per quintal of wheat; the average discounting the four high prices from Delos is 48 · 07 g of silver. The fine, equivalent to 220 g of silver per quintal, was probably over nine times the subsidized selling price at Rome.

This conclusion is based on the prices given in Livy in the texts (a)–(e) above; converted into the weight of silver which was needed to purchase one quintal (1 modius holding 6·67 litres, 1 denarius = 10 asses, 1 denarius = c. 4·0 g) of wheat, they come to 23·99 g (203 and 201 B.C.), 11·99 g (200 and 196 B.C.), and less than 11 · 99 g (202 B.C.). We have no other evidence for the price of wheat at Rome until we come to the frumentationes of G. Gracchus in 123 or 122. At these wheat was sold at 6–3 asses per modius. Around 141 B.C. the as had been re-tariffed at 16 to the denarius (Crawford, M. H., Roman Republican Coinage 11 611–14Google Scholar), and the denarius now had a projected weight of 3 · 85 g (based on Crawford, 11 594, cf. 592). Converted for comparison with the earlier prices, these figures indicate a cost of 22–85 of silver for 1 quintal of wheat. The ‘famine ‘price at Rome of 25 asses ( = 2 · 5 denarii) per modius in 211–210 (Polyb. 9. 11A), at a time when the denarius had a target weight of 4·5 g (Crawford, 11 595), produces a cost of 168·67 of silver for a quintal of wheat.

⁵⁶ cf. App., Lib. 75. For a cargo of 360,000 modii, 36 ships of 10,000 modii, or 65 tonnes burden, would be needed, or 8 ships of 50,000 modii or 325 tonnes burden (or half the number of ships for half the quantity of grain). The Romans commonly built ships in the range of 200–300 tonnes in the following century; see Pomey, P. and Tchernia, A., ‘Le tonnage maximum des navires de commerce romains’, Archaeonautica 2 (1978), 233–51CrossRefGoogle Scholar.

⁵⁷ Livy 36. 4. 5 ff. The Sicilians in text (d) may be assumed to have expected some payment from Flaminius. Another episode, from 169 B.C., if not strictly a gift, may still be a closer parallel for the inscription: see Livy 44. 16. 2. Note the delayed fixing of the price at Rome.

⁵⁸ Pliny, Pan. 31; cf. Rickman (see n. 33), 115.