¹ Hammond, M., ‘Composition of the senate, A.D 68–235’, JRS 47 (1957), 75Google Scholar;Hopkins, Keith, Death and Renewal (1983), ch. 3Google Scholar.

² On the cost of having children see Pliny, , Ep. 4. 15. 3Google Scholar; Musonius Rufus, fr. 15b (ed. O. Hense).

³ Goody, J. and Tambiah, S. J., Bridewealth and Dowry (1973), 17–47Google Scholar. See also his more general work, Production and Reproduction (1976).

⁴ Schaps, D., Economic Rights of Women in Ancient Greece (1979), ch. 6 and App. IGoogle Scholar.

⁵ Goody, , Production, 6–7Google Scholar. Goody recognizes that the two means of transmission of property to daughters carry with them some noteworthy differences. Gortyn code iv. 31–54.

⁶ For the European evidence I have depended on Cooper, J. P., ‘Patterns of inheritance and settlement by great landowners from the fifteenth to the eighteenth centuries’ in Family and Inheritance: Rural Society in Western Europe, 1200–1800, ed. Goody, J., Thirsk, J., Thompson, E. P. (1976), 192–327 (especially 249, 269, 283, 286, 301)Google Scholar; Forster, R., The Nobility of Toulouse in the Eighteenth Century (1960), ch. 6Google Scholar; Davis, J. C., A Venetian Family and its Fortune 1500–1900 (1975), 106 ff.Google Scholar; Stone, L., The Crisis of the Aristocracy, 1558–1641 (1965), 175Google Scholar; Litchfield, R. B., ‘Demographic characteristics of Florentine patrician families, sixteenth to nineteenth centuries’, Journal of Economic History 29 (1969), 203CrossRefGoogle Scholar.

⁷ I am concerned here exclusively with well-to-do classes of Rome and Italy (from the municipal aristocracies up). To the best of my knowledge no evidence exists concerning dowry exchange among members of the lower orders.

⁸ Watson, Alan, Roman Private Law around 200 B.C. (1971), 17Google Scholar, together with his The Law of Persons in the Later Roman Republic (1967), 29 ff.

⁹ Corbet, P. E.. The Roman Law of Marriage (1930), 90 f.Google Scholar; Watson, , Persons, 19Google Scholar.

¹⁰ Corbett, , Law of Marriage, 152–4Google Scholar.

¹¹ Ibid. 179 ff. For example, a law of Augustus prohibited the husband from alienating Italian praedia in the dowry without the wife's consent.

¹² A title in the Digest (23. 4) is devoted to such pacts.

¹³ Corbett, , Law of Marriage, 182–201Google Scholar. Schulz, F., Classical Roman Law (1951), 126–8Google Scholar, offers a brief summary of the rules. The rules are set out in Ulp. Reg. 6. 3–13 (in FIRA ii. 269 f.).

¹⁴ It is possible to use life tables to arrive at a rough estimate of the chances that a paterfamilias who has provided a dos profecticia will die before his daughter, with the consequence that the dos will remain with the husband on her death (assuming no divorce). The numbers are based on the assumption that the daughter was born when the father was 30 and married 15 years later. At the age of marriage roughly one-third of the brides would already have lost their fathers through death and as a result their dos would necessarily have been adventicia, thus reverting to the husband on their death. Of those who married with their fathers still alive, roughly three in every four would have outlived their fathers. Thus, the reversion of dos profecticia to the paterfamilias was a rarity. The husband was much more likely to restore the dos upon divorce. (For age of marriage see Hopkins, M. K., ‘The age of Roman girls at marriage’, Population Studies [1965], 315 ffGoogle Scholar. I have used the life-table of Frier, Bruce, ‘Roman life expectancy: Ulpian's evidence’, HSCP 86 [1982], 245Google ScholarPubMed, with no pretence of being able to judge its relative merits in comparison with other tables on the grounds that any high mortality table is likely to produce roughly similar results.)

¹⁵ Crook, J., Law and Life of Rome (1967), 105Google Scholar.

¹⁶ In the title on pacta dotalia 23. 4 see 2, 12 pr., 23, 24, 26. pr., 26. 2, 30. For further discussion and references in other titles see Humbert, M., Le remariage à Rome (1972), 284 ffGoogle Scholar.

¹⁷ Humbert, , Remariage, 284 n. 6Google Scholar.

¹⁸ Personal Patronage under the Early Empire (1982), 121. Cicero showed great reluctance about using legal remedies to recover Tullia's dowry from Dolabella (Ad Att. 12. 12, 14. 18–19, 16. 15).

¹⁹ Humbert, , Remariage, 284 ffGoogle Scholar.

²⁰ Buckland, W., Text-book of Roman Law, 325 fGoogle Scholar.

²¹ P. 199

²² While stressing onera matrimonii Corbett, (Law of Marriage, 178)Google Scholar recognizes that ‘there were cases of legal separation of dos from the onera’. Schulz, , Classical Roman Law, 124 fGoogle Scholar. and Kaser, M., Das römische Privatrecht² i (1971), 332 fGoogle Scholar. stress onera matrimonii.

²³ Remariage, 275. The State's interest in remarriage certainly explains the restrictions on dotal pacts to the detriment of the wife.

²⁴ Wolff, H. J., ‘Zur Stellung der Frau im klassischen römischen Dotalrecht’, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte 53, 2 (1933), 297–371Google Scholar; Koschaker, Paul, ‘Unterhalt der Ehefrau und Früchte der Dos’, Studi Bonfante iv. 3–27Google Scholar.

²⁵ This is the conclusion of Koschaker in the article cited in the previous note. Wolff gives close attention to the problem of interpolation in his previously cited article, particularly in n. 1, p. 369.

²⁶ Fundi and praedia are frequently mentioned in the Digest titles concerning dowry, e.g. from title 23. 3: 6. 1, 10. 1, 32, 47, 50pr., 52.

²⁷ Though this letter does not include the word dos, from the context it seems to me all but certain that Pliny's contribution was intended for the dowry. (Duncan-Jones, R., The Economy of the Roman Empire [1974], 28, includes this among contributions to dowries.)Google Scholar The context and language (confero) are parallel to Pliny, Ep. 2. 4, which explicitly involved dowry. Digest 23. 3. lOpr. makes it clear that the wife's clothing, mentioned by Pliny, might be included in a dos aestimata (though it was to the husband's disadvantage to have it valued in this way). A Latin dotal agreement from Egypt of c. A.D. 100 has been preserved on papyrus and includes clothing and jewelry as a part of the dos aestimata, but unfortunately this is not very sound evidence for the custom in Rome and Italy (FIRA iii, no. 17).

²⁸ Ep. 2. 4, 6. 32.

²⁹ Apol. 71, 77, 92; 91 for the description of the dos as modica.

³⁰ Apol. 92.

³¹ Ann. 2. 86. It has been suggested to me that Tiberius simply contributed to the dowry, but the Latin says that Tiberius consoled the loser with a dos, not a contribution to one. Furneaux implies in his comment on the passage an understanding similar to mine. This passage and others cited below have been used as evidence that 1,000,OOOHS was a customary size for dowries: see e.g. Balsdon, J. P. V. D., Roman Women, 187 and Mayor's commentary on Juvenal 10. 335Google Scholar.

³² Cons, ad Helv. 12. 6. Seneca makes it clear in this passage that he is not referring to the modestly rich but to the very rich.

³³ Epig. 11. 23; the same figure appears in 2. 65 and 12. 75.

³⁴ Sat. 6. 136; 10. 335; 2. 117.

³⁵ Duncan-Jones, , Economy, 32Google Scholar.

³⁶ Schaps, , Economic Rights of Women, 78Google Scholar, makes this point for Athenian dowries on the basis of better evidence than is available for Rome.

³⁷ Apol. 91–2.

³⁸ Apol. 91 (it should be noted that there is no evidence that young, beautiful brides from wealthy families went without dowries); see also Seneca, , Ben. 4. 22Google Scholar. 4 andMartial, , Epig. 7. 10Google Scholar.

³⁹ It seems probable to me that daughters usually received a large fraction (a quarter or more) of their father's estates, since (1) aristocratic families of this period were typically quite small, (2) only a third of the children born outlived their fathers, and (3) there was a strong feeling that daughters deserved an equal or at least substantial share of the inheritance. The assumption of small families is based on the opposition to Augustan legislation penalizing those with fewer than three children and literary evidence such as that cited in n. 2. For (2) see Goody, , Production and Reproduction, 133 fGoogle Scholar. J. Crook has argued for (3) in a paper entitled ‘Women in Roman succession’ (as yet unpublished). Though a daughter's rights were not quite on a par with her brother's, she had a claim to an equal share according to the civil law rules on intestacy; further, the praetor in the late Republic recognized her claim in granting bonorum possessio contra tabulas even if she had been emancipated; finally, unless disinherited by her father (though not necessarily by name as her brother had to be) she could undertake a querela inofficiosi testamenti in order to receive her intestate share of the estate (see Pliny, , Ep. 6. 33)Google Scholar. This last development probably occurred in the early Principate and shows a continuing feeling that daughters deserved a substantial share of the estate. (See Schulz, , Classical Roman Law. 270–9Google Scholar.) I am grateful to Professor Crook for permission to read and to use his paper before publication.Pomeroy, S. B., ‘The relationship of the married woman to her blood relatives in Rome’, Anc. Soc. 7 (1976), 223 f.Google Scholar, misstates the rule regarding succession of filiae, neglecting the distinction between sui heredes and agnati (Gaius, , Inst. 3. 1–16)Google Scholar.

⁴⁰ Remariage, 76–112.

⁴¹ Cooper (above, n. 2), 301.

⁴² Polyb. 31. 27; cf. Polyb. 18. 35 for the dowry received by Aemilius Paulus.

⁴³ See n. 8.

⁴⁴ Production and Reproduction, 61.

⁴⁵ Polyb. 31. 27; Cicero, Ad Att. 11. 2. 2; 11. 23. 3. Cicero had some difficulty in making the dowry payments for Tullia and mentions the possibility of borrowing from Atticus, but he clearly attributes the difficulty to mismanagement of his property. Ad Att. 11. 2. 2 indicates that Cicero expected to be able to pay the second instalment, at least in part, out of current income from his estates.

⁴⁶ Cooper (above, n. 2), 301.

⁴⁷ Remariage, 99 f.

⁴⁸ Apol. 91.

⁴⁹ Though Humbert refers in the text (Remariage, 99 f.) to enrichment through acquisition of a large dowry, in fact none of his references concerns dos. In Ad Att. 13. 28. 4 Cicero does not say that Nicias Talna was trying to marry Cornificia for her money; even if that was the case, there is no hint that the enrichment would have been in the form of a dos, as opposed to inheritance on her death. In any case Talna suffered what must often have been the fate of would-be fortune hunters: his advances were discouraged because he had an estate worth a mere 800,OOOHS. Quintilian, Inst. 6. 3. 73 again says nothing of financial motives for marriage and not a word about dowry. The motives of the quaestor who divorced his wife after being assigned his province (Suet. Tib. 35. 2) are not stated, but the reference to the assignment of the province implies that the quaestor kept his wife just long enough to take advantage of Augustus' marriage laws favouring candidates with wives and children (nothing to do with the wife's money or dowry). Finally, Humbert refers to the important financial questions raised by Cicero's divorce and remarriage. The restoration of Terentia's dowry caused Cicero some discomfort in a time of civil war and disarray in his own estate management. Whether these financial problems prompted the marriage to Publilia is an impossible question to answer (Bailey, D. R. Shackleton, Cicero [1971], 202 f)Google Scholar. If it was a ‘mariage d'argent’, it is unclear whether Cicero was interested in her dowry or her personal fortune. (Publilia's father had died, so she would have owned property in her own right.) The last point to be made about this case is that even if Cicero had married Publilia to solve financial difficulties, the dowry was small enough to be repaid without great discomfort (one instalment being repaid before the due date according to Ad Att. 16. 2. 1). Better evidence for Humbert's argument, though fictional, are Martial's Epig. 2. 65 and 10. 15, which refer to the husband acquiring a large dowry, but it is important to note that the husband is said to enjoy the windfall only in the wife's death rather than from the beginning of the marriage. Carcopino, J., DailyLife in Ancient Rome, transl. Lorimer, E. O. (Penguin, ed., 1956), 103Google Scholar, portrays Roman husbands as moving from wife to wife in search of ever larger dowries – an exaggerated view of ‘mariages d'argent’ for which he produces no adequate evidence.

⁵⁰ Horace, , Carm. 3. 24Google Scholar; Seneca, , Matrim., fr. 87Google Scholar; Martial, , Epig. 8. 12Google Scholar; Juvenal, , Sat. 6. 136Google Scholar.

⁵¹ Duncan-Jones, , Economy, 25 ffGoogle Scholar.

⁵² See above n. 2.

⁵³ Epig. 8. 12.

⁵⁴ A preliminary version of this paper was read at a seminar of the London Institute of Classical Studies. I am grateful to the participants for their suggestions and to Professor J. Crook, Sir Moses Finley, Dr P. Garnsey, Mr G. Herman, Dr B. Shaw, Dr A. Wallace-Hadrill and Mr C. R. Whittaker for reading and commenting on the paper.