Hostname: page-component-77c89778f8-vpsfw Total loading time: 0 Render date: 2024-07-17T07:26:32.714Z Has data issue: false hasContentIssue false
  • English
  • Français

The Rental Market in Early Imperial Rome

Published online by Cambridge University Press:  24 September 2012

Bruce Woodward Frier
Bruce Woodward Frier
Affiliation:
The University of Michigan
Get access Rights & Permissions [Opens in a new window]

Extract

In Rome of the early empire, most of the residential population lived in rented apartments (cenacula). Only the privileged few could afford single-family dwellings; the character of this minority is vividly illustrated in Juvenal's famous line on Nero's persecutions: ‘rarus venit in cenacula miles’ (10. 18). Almost all of the non-privileged many, if they could afford accommodation, were obliged to dwell in buildings that they did not own, in exchange for rent that our sources agree was exorbitantly high.

However, not everyone in the tenant class was on equal footing, despite recent suggestions to the contrary. The range of accommodation regularly available to the urban masses was, as will be shown below, quite varied, though this range was often effectively limited by the wealth and other social characteristics of the prospective tenant. The rental market of Rome can be reconstructed from three kinds of sources: literary references to rental; the types of rental situations described in legal texts; and the archaeological remains of apartment houses, particularly those in Ostia. These sources converge to suggest a model of the urban rental market that is rather more complex than the one which has appeared in recent scholarship. It must be stressed, however, that the new model suggested below is still just a model, and that variations from it must have been numerous. For instance, no ancient authority conclusively demonstrates that a lease like the typical modern American lease (an apartment taken on a year term for exclusive occupancy of the tenant and his dependants, with monthly payment in advance and many services supplied by the landlord) was legally impossible in antiquity, or was not in fact developed; but the Roman leases actually described in the extant sources are markedly different, and the ruins of Ostian apartment houses seem to accord well with the leases to which these sources refer.

Type
Research Article
Information
The Journal of Roman Studies , Volume 67 , November 1977 , pp. 27 - 37
Copyright
Copyright © Bruce Woodward Frier 1977. Exclusive Licence to Publish: The Society for the Promotion of Roman Studies

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 The model of the rental market proposed in this article is intended to replace the ‘accommodation pyramid’ described by Newbold, R. F., Latomus 33 (1974), 863–5Google Scholar; based on Homo, L., Rome Impériale et l'Urbanisme (1951), 565, 592–3Google Scholar; see also Pöhlmann, R., Die Übervölkerung der Antiken Grossstädte (1884), 107–9Google Scholar. (Newbold's model, which supposes for Rome an efficient disposition of housing space with no de facto segregation of rich and poor, is shaky on legal points, esp. 863–4.) I have frequently cited Enid Gauldie's masterly Cruel Habitations: A History of Working-Class Housing 1780–1918 (1974), on Britain.

2 On the remains, I have mainly used Packer, J. E., The Insulae of Imperial Ostia, MAAR 31 (1971)Google Scholar; on which see however Riemann, H., Gnomon 47 (1975), 186–201Google Scholar; Meiggs, R., Roman Ostia2 (1973), 585–6, 597–8Google Scholar. On similarities with Rome, see Packer, 74–9 (Rome's insulae were for the most part ‘far less methodically planned than any yet found in Ostia’, 77); but see idem, Boll. Com. Rom. 81 (1968/9), 127–48, on the Casa di Via Giulio Romano in Rome. See also McKay, A. G., Houses, Villas and Palaces in the Roman World (1975), 80–99Google Scholar, who adds little in this area.

3 On the modern lease, see ‘Javins v. First National Realty Corp.’, in Federal Reporter 428 (1970), 1071–83Google Scholar, esp. 1074–5 (U.S. Court of Appeals, D.C. Circuit; Wright, J.). In big-city slums, however, this formal contract becomes month-to-month, and in fact tends to be executed informally in a fashion not unlike the contract described in the second section of this paper: Sternlieb, G., The Tenement Landlord (1966), 88Google Scholar (for Newark, N.J.).

4 R. Meiggs, op. cit. (n. 2), 247; the ‘bedrooms’ here and elsewhere tend to average c. 20 m2, and thus are quite large.

5 It is Packer Type II C and D: op. cit. (n. 2), 8–10, 18–19; according to p. 91, above thirty-five such ground-floor plans are known. (A remarkable new variant: Veloccia Rinaldi, M. L., Rend. Pont. Acc. Arch. 43 (1970), 165–85Google Scholar.) While all these apartments have the same general floor plan, many (like the Casa a Giardino III, ix, 13–20) are two-storied and were doubtless much more desirable; in general, we also do not know whether ground-floor plans continued in upper floors, as Packer (p. 70) assumes. On Type II C at Rome, see P. Ziçans, in Opusc. Arch. 2 (1941), 191. To this tenant-group should be added the lessees of houses, of large apartments within houses, and of luxury apartments generally. Using information provided by Packer at 69–71 and 88–92 (but after making certain obligatory corrections), I would estimate that these forms of accommodation housed c. 17–1800 persons, both slave and free; of this group, about 40 per cent lived in houses, 20 per cent in luxury apartments, and 40 per cent in Type II C and D cenacula. Ostia's total population within the walls should perhaps be put at 20–35,000.

6 The legal sources are mostly contemporary with Ostian evidence (about A.D. 100 to 225). On the orientation of juristic writings toward the city of Rome, see above all Schulz, F., Principles of Roman Law (1936), 33–4Google Scholar.

7 J. E. Packer, op. cit (n. 2), 69–70; but see R. Meiggs, op. cit (n. 2), 597–8.

8 The main texts concerning such entrepreneurs are Dig. 13. 7. 11. 5; 19. 1. 53 pr.; 19. 2. 7–8, 30 pr., 58 pr., 60 pr. This practice protected the owner against the risk of incomplete occupancy (on which see Suet., Tib. 35. 2), insulated him from the day-today business of handling tenants, and limited the owner's losses in the event that the insula became uninhabitable; see esp. M. Kaser, ZRG 74 (1957), 157–69. Middlemen may have been common; did Crassus act personally as landlord in the huge section of Rome that he owned (Plut., , Crassus 2Google Scholar)? On Cicero's property, see n. 49; in a forthcoming article in the Classical Journal, I will argue that his urban tenements must have been managed in this way. Dig. 19. 2. 7 suggests speculation among entrepreneurs. But observe the presence of slave insularii in the great Julio-Claudian families (CIL VI. 3973–4, 4347, 4446, 6215, 6217, 6296–9, 7291, 7407; cf. VI. 9292, 9479–83, 33863; Juv. 3. 195), and in the familia Caesaris (CIL VI. 8856, XIV. 2769; cf. VI. 8855; Suet., Claud. 38. 2); cf. Boulvert, G., Esclaves et Affranchis Impériaux (1970), 37, 139Google Scholar. As institores, the insularii usually bound the building's owner: Dig. 14. 3. 5. 1. The exactor ad insulas (CIL VI. 9383) collected rents, cf. Dig. 13. 7. 11. 5.

9 For this meaning, see G. Calza, MAAL 23 (1915), 591–5; Calza used almost no secondary sources in this famous study, and so missed the crucial discussion of E. Cuq, Dict. Ant. s.v. ‘locatio conductio (rei)’, 1287 (with no archaeological reference). The text discussed below was therefore ignored until G. Hermansen rediscovered it, Phoenix 24 (1970), 342 f.; but he did not realize its full purport.

10 Ulpian, Dig. 9. 3. 5. 1, quoted below. See in general Amirante, L., in Studi B. Biondi I (1965), 455–65Google Scholar. On the right to sublease, see A. Pernice, ZRG 19 (1898), 94–5; Mayer-Maly, T., Locatio Conductio (1956), 27, 30Google Scholar. The commonness of the practice cannot be established.

11 H. Rowell, CPh 52 (1957), 217–21. The upperclass connotation of cenaculum is implied in Pompeian advertisements for lease: CIL IV. 138, 1136.

12 On the action, see W. Wolodkiewicz, RISG3 12 (1968), 371–9; and the bibliography in Kaser, M., Das römische Privatrecht2 II (1975), 428Google Scholar, nn. 24, 27. On the text of 9. 3. 5. 2, Beseler, G., Studi P. Bonfante II (1930), 71Google Scholar, doubted ‘quod sine captione actoris fiat’, perhaps rightly, although it is good as law: de Visscher, F., La Régime Romaine de Noxalité (1947), 531–2Google Scholar.

13 See esp. Cic., de Or. 3. 17; TLL s.v., 1319 ll. 30–44 (esp. the references to Vitruvius). Exedrai are commonly rented in papyrus leases: see n. 17 below (A. Berger, 360, n. 134). See now Settis, S., Aufstieg und Niedergang I. 4 (1973), 666–71 and 672–5Google Scholar.

14 On the word, A. Herdlitczka, RE Suppl. VI 386; G. Hermansen, op. cit. (n. 9), 345–7. Idem, in Polis and Imperium: Studies E. T. Salmon (1974), 167 f., adds little.

15 Modern authors therefore err in referring to the exedrae as the tablinum and triclinium, and to the medianum as the ‘atrium-hall’; e.g. J. E. Packer, op. cit. (n. 2), 8–10. This text supports Calza's view, op. cit. (n. 9), 595, against the relation of the cenaculum-form to the Italic domus; it goes against those seeking a relationship, e.g. most recently Boyle, B. M., Journ. Soc. of Archit. Hist. 31 (1972), 257–8Google Scholar. A good survey is now in Colognesi, L. Capogrossi, La Struttura della Proprietà II (1976), 286–303Google Scholar.

16 Ulpian, Dig. 43. 32. 1. 4 (citing Labeo); see L. Homo, op cit. (n. 1), 595. Such an instalment payment is a pensio, see Wenger, L., Canon in den römischen Rechtsquellen und in den Papyri, SAWW 220. 2 (1942), 35–8Google Scholar. For annual pensiones see esp. Suet., Nero 44. 2; Dig. 36. 2. 12. 5. For multiple-year leases, Dig. 19. 2. 24. 2 (a domus), and 60 pr.; 43. 32. 1. 4; CIL IV. 1136. Payment after use is presumed in texts concerning tenants' justified withholding of rent, see esp. Dig. 19. 2. 27 pr. (deductio ex mercede), on which Watson, A.The Law of Obligations in the Later Roman Republic (1965), 115–16Google Scholar. For payment before use, there is only Ulpian, Dig. 19. 2. 19. 6. Note, however, that the entrepreneurs discussed above (n. 8) paid in advance: Dig. 19. 2. 7, 30 pr; this accords with the nature of their lease.

17 Lists of published leases, with payment terms, in Berger, A., Zeitschr f. Vergl. Rechtswiss. 29 (1913), 327–30 and 377–90Google Scholar (esp. 378–85); supplemented by Mickwitz, G., Geld und Wirtschaft im römischen Reich des vierten Jahrhunderts n. Chr. (1932), 205–6Google Scholar; more recent lists (only) in Johnson, A. C., Roman Egypt to the Reign of Diocletian (1936), 262Google Scholar; Montevecchi, O., Aegyptus 21 (1941), 287–94Google Scholar; J. Modrzejewski, JJP 7/8 (1953/1954), 217, n. 29; Taubenschlag, R., The Law of Greco-Roman Egypt2 (1955), 364, n. 1Google Scholar; Braunert, H., in Festschrift F. Oertel (1964), 36Google Scholar (Braunert's papyri republished as SB VIII 9931–2); Montevecchi, O., La Papirologia (1973), 217–18Google Scholar; BGU XII 2202, p. 124. Most leases are for houses or parts of houses, and envisage payment by years or half-years. Payment by month is uncommon and seems to be purchased at the cost of signing a multiple-year lease, e.g. BGU IV 1116 (13 B.C.); BGU I 253 (244–9 A.D.); C. P. Herm. 119 (A.D. 253–68); cf. Berger, op. cit. 385–6. Rent is normally paid at the end of the payment period, see Berger, 387–8, and note M. Kaser, ZRG 91 (1974), 162–3; where rent is paid in advance, a lease clause may secure against eviction, P. Rein. 43 (A.D. 102); P. Tebt. II 372 (A.D. 141); see Mickwitz, op. cit. 126, 205 n. 1. There are many similarities with the Roman rental market. Even the start of the rental year is similar: it usually begins in high summer (compare Montevecchi, Pap. 218, with the sources in L. Homo, op. cit. (n. 1), 581–2, on 1 July at Rome. The same contractual form was also probably used in Hellenistic Delos; Molinier, S., Les maisons sacrées de Délos (1914), 58–9Google Scholar. Monthly rent is known from Nicarchos, Anth. Pal. II. 251 (Athens?; cf. I. Casaubon (1592) on Theophr., Char. 10. 2); payment by prytanies, Ammon. 414 (probably for state-owned houses; cf. Xen., , Poroi 4. 19Google Scholar).

18 This lien was a normal lease term, and finally was implied by law in the absence of provision to the contrary (the first such implication is from Neratius, Dig. 20. 2. 4 pr.); see now Schuller, W., Labeo 15 (1969), 267 fGoogle Scholar. Martial 12. 32 gives a lurid picture of its operation.

19 On this aspect of the legal texts, see esp. Daube, D., Roman Law (1969), 71 f.Google Scholar; but also Volterra, E., Riv. Ital. per le Sc. Giurid. II (1967), 239–71Google Scholar. A. Pernice, op. cit. (n. 10), 95, erred on his own evidence in linking low social status with apartment rental; see esp. Cic., Cael. 17, with H. Rowell, op. cit. (n. 11), 219–20; T. Mayer-Maly, op. cit. (n. 10), 227–8; most modern legal literature is still confused, including, for example, Marrone, M., La legittimazione passiva alla ‘rei vindicatio’ (1970), 126–30Google Scholar. Of course, tenants not on long-term leases also fell under the rules of locatio conductio; thus Giton refers to his lodgings (see below) as a conductum, Perron. 9. 4; cf. Mart. 9. 75. 1. But the contractual rules had little useful application to such lodgers.

20 See Girri, G., La Taberna (1956), 37–43Google Scholar, who is widely accepted, for example by J. E. Packer, (op. cit. (n. 2), 69. The statistics cited in n. 5 would suggest that c. 91–5 per cent of Ostia's population inhabited shops or small flats, or slept in the streets (cf., for Rome, Amm. Marc. 14. 6. 25).

21 This is suggested by the pattern of Egyptian commercial leases, which also frequently introduce monthly rent payments; see esp. P. Petrie III 73 (third century B.C.), which may resemble a Roman shop lease: it is for a shop in a synoikia (below, n. 31), with monthly rent payment and renewal. A lease for five years of ‘tabernae, pergulae’ etc. is offered in CIL IV. 1136.

22 For a description, J. E. Packer, op. cit. (n. 2), 177–85, cf. 69 (‘an average living space of two rooms apiece’); plans, 106–7. Archaeologists tend to ignore the use of temporary partitions to divide larger ‘rooms’ into small ‘apartments’, both in these great insulae and in humbler structures like III, i, 12–13; but note Vitruv. 2. 8. 17.

23 Casa di Diana: J. E. Packer, op. cit. (n. 2), 127–34; plan, 94 (cubicles on second floor, rooms 6–7). Caseggiato di Temistocle: 193–5; plan, 110 (cubicles, rooms 10–14).

24 R. Meiggs, op. cit. (n. 2), 249. This goes against the (quite impossible) view of Calza, accepted, for example, by A. G. McKay, op. cit. (n. 2), 96–7, whereby the first floor was a ‘piano nobile’; the ruins are irreconcilable with this thesis.

25 The first sentence, at least, is not interpolated. On it, Amirante, L., Labeo 8 (1963), 207Google Scholar, with bibliography. This passage is crucial in establishing the two alternative modes for renting out space in insulae.

26 This is the widest sense of the word, frequent in inscriptions (e.g. CIL VI. 15640; x. 1450, 3750); see Dig. 32. 91. 4; 47. 10. 5. 5; and below.

27 H. Rowell, op. cit. (n. 11), 211–17. The scene is some Campanian city, probably Puteoli; on the influence of Roman housing patterns in Campania, see J. E. Packer, op. cit. (n. 2), 61–3. For the legal concept of lodging, see The American Law of Property I (1952), 192–4Google Scholar; at Common Law, it is distinct from tenancy (contrast n. 19 above).

28 See Blümner, H., Die römischen Privataltertümer (1911), 455 n. 4Google Scholar; H. Rowell, op. cit (n. 11), 226 n. 4; also de Robertis, F. M., Ann. della Fac. di Giurisprud. Bari 12 (1953), 125Google Scholar n. 4, for legal sources.

29 Compare E. Gaudie, op. cit. (n. 1), 97–100. Encolpius' room was ‘furnished’. On locks, see also Mart., 7.20. 20–21.

30 The suggestion of G. Bagnani, AJPh 79 (1958), 441–2, that deversitor means ‘bartender’, should be rejected. At 95. 8, insularii means ‘building attendants’ (as always in Latin), not tenants, contra H. Rowell, op. cit. (n. 11), 223–4.

31 At Athens, synoikiai are lodging houses accommodating on short-term leases the lower classes (Is. 6. 19–21) or travellers (Ps.-Xen., Ath. Pol. I. 17; Aeschin. 1. 43); they are often run by entrepreneurial middlemen called naukleroi: Is. 6. 19; Harp., Phot., Hesych., s.v.; Ammon. 330; Poll. I. 75. As to the papyri, the subject has not been treated in secondary literature. These sources are helpful. It is a building with one (P. Petrie I 12. 7, P. Petrie II, p. 23: 238/7 B.C.; PSI x 1159. 19: second century A.D.) or more (BGU VII 1573. 25: 141/2 A.D.) owner, for whom it comes to be named (P. Petrie III 73: third century B.C.; P. Fouad III 59. 2: after 75/6; P. Fayum 37. 3: third century A.D.; SB VIII 9902 col. B II. 3: fourth century A.D.; compare, for Athens, Aeschin. I.125). It is managed by a superintendent (P. Petrie III 73) and has many tenants (P. Fouad III 59. 2; P. Mich. VIII 481. 34: early second century A.D.; P. Fayum 37. 3, a police report on a tenant), renting flats (P. Fayum 31. 13, with a note: c. A.D. 129; P. Berl. Leihgabe 16 col. B II, col. C 10: A.D. 161) or stores (P. Petrie III 73). A tenant pays six months' rent for another person (BGU II 362 col. XIII 5: c. A.D. 215). Synoikiai are often associated with Alexandria (P. Petrie I 12. 7, with note above; BGU IV 115. 16, 19: 13 B.C.; P. Mich. VIII 481. 34). If BGU II 362 col. XIII 5 can be generalized, synoikiai may include apartment houses; there is no exact Greek translation of cenaculum in the sense of apartment. SB x 10233 (fifth century A.D.) may be a rent-receipt register from a lodging house.

32 See Kleberg, T., Hôtels, restaurants et cabarets dans l'Antiquité romaine (1957), 19–25Google Scholar, who cites (129 n. 54) Cassiodorus, ad Psalm. 14. 1: ‘Maiores nostri domus pauperum tabernas appellaverunt …’. Other probable examples of this sense (rather than ‘shop’) are Cic., Att. 14. 9. 1; Ascon. p. 37 C (‘dormientem in taberna’); Tac., Hist. I. 86. 2 (‘in tabernis et cubilibus’: ‘in tenements and hovels’); so too, perhaps, Suet., Nero 37. 1; Juv. 1. 105. Taberna deversoria: Plaut., Men. 436; Truc. 697; Varro, RR 1. 2. 23; Suet., , Nero 27. 3Google Scholar (invariably an inn).

33 See T. Kleberg, op. cit. (n. 32), 11–14, with other examples of ‘inn’. Note especially CIL IV. 807 (Pompeii): ‘hospitium hie locatur/triclinium cum tribus lectis’. Deversorium equals hospitium: Nonius p. 60; but hospitium is often used very generally in Latin, like ‘quarters.’

34 On this passage see H. Rowell, op. cit. (n. 27), 220–1. The meaning of pergulae is uncertain; see G. Calza, op. cit. (n. 8), 586–7. The Insula Feliculae (or Felicles), also mentioned in the Notitia, is one of several named insulae in Rome; most are named for an owner (cf. n. 31), but here the name is probably euphonic (cf. the Insula Eucarpiana, CIL VI. 10250).

35 Caupo provides the crucial link to the urban deversoria of Petronius: see 39. 12, 61. 6, 62. 12, 98. 1. Latin apparently had no other common word for this occupation. For the Edictal provisions, see O. Lenel, Edictum Perpetuum 3 (1927), 131 (receptum); 205 (damnum iniuria datum); 333–4 (furtum).

36 Interpretation is made difficult by the notorious problems concerning the development of the Edictal liabilities of the caupo; but this text is considered essentially classical even by the very critical Solazzi, S., Scritti di Diritto Romano III (1960), 506–8Google Scholar. See also Sargenti, M., in Studi E. Albertario I (1953), 555–8Google Scholar; F. M. de Robertis, op. cit. (n. 28), 134–43; and now Wolodkiewicz, W., Riv. It. Sc. Giur.3 14 (1970), 210–13Google Scholar, with further literature. The interpolation of the other two texts cited is arguably evident; the above authors discuss the matter at length. Compare Dig. 47. 10. 5. 5.

37 On hotels at Ostia, T. Kleberg, op. cit. (n. 32), 45–8, 53–6; R- Meiggs, op. cit (n. 2), 428–30. The Casa delle Volte Dipinte (III, V, I) may be a hotel: J. E. Packer, op. cit. (n. 2), 170. For Rome, compare Livy 45. 22. 2; Sidon., Ep. 1. 5. 9; et al.

38 Broadly speaking, the distinction would help mark a typical boundary between upper and lower classes: see Sjoberg, G., The Preindustrial City (1960), 108 f.Google Scholar, esp. 123–33. Observe the story of Vitellius who, at a time of personal financial stress, leased his domus and placed his family in a cenaculum meritorium (Suet., Vit. 7. 2). Lodging houses are also characteristic of Victorian city life, see E. Gauldie, op. cit. (n. 1), 152, 241–6; but I argue that in Rome, due to extreme overcrowding produced by inadequate transportation (Sjoberg, 92), they became the dominant form of housing for the lower classes. It may be noted that the tiered Roman market was closely reproduced in Athens and in Egypt (above, nn. 17, 31); in Athens, urban immoveables were sharply divided into oikiai and synoikiai (Ar., Thesm. 272–73; Thuc. 3. 74. 2; Is. 2. 27; Aeschin. I. 105, 124; Harp. s.v. ‘naukleros’). What is apparently novel at Rome is the emergence of a relatively wealthy tenantry for apartments, and this despite the fact that ‘there was probably something dubious or vulgar about renting’; Rawson, E., in Studies in Roman Property (ed. Finley, M. I., 1976), 87Google Scholar; see n. 19 above.

39 Some statistics on luxury rent are collected by Friedländer, L., Darstellungen aus der Sittengeschichte Roms9 II (1922), 331–5Google Scholar. Many senators rented in Rome, usually at higher than HS 6,000 a year: Vell. 2. 10. 1; cf. Cic., Pis. 61; Cael. 17; Dio 46. 31. 3. Rents rose primarily because of land prices at Rome: see R. Duncan-Jones, PBSR 33 (1965), 224–5; compare Isaeus II. 42, who gives a ratio, for two houses, of rent per year to value: 8·6 per cent. The government struggled to deal with resultant market problems: Philipps, E. J., Latomus 32 (1973), 86–95Google Scholar.

40 This is a ‘base rate“ for housing of below-market quality, similar to rates occurring in the nineteenth century: E. Gauldie, op. cit. (n. 1), 159. One should distinguish ‘physiological subsistence level’ from ‘cultural subsistence level’: Heichelheim, F., Wirtschaftliche Schwankungen (1930), 100Google Scholar.

41 Frank, T., An Economic Survey of Ancient Rome I (1933), 385Google Scholar, cf. 189; thus, about 35 per cent of a labourer's wages. On subsistence wages, MacMullen, R., Roman Social Relations (1974), 183 n. 1Google Scholar, with bibliography; the Rosc. Com. figure is a minimum daily wage, see Crawford, M., Roman Republican Coinage II (1974), 622–3Google Scholar. In 1885, half the London working class paid between one-quarter and one-half of their wages in rent: E. Gauldie, op. cit. (n. 1), 164.

42 Gell., NA 15.1.3; Herodian 7. 12. 6; cf. n. 51.

43 Dig. 19. 2. 7–8, 30 pr. (the latter text refers to the subleasing of cenacula); see R. F. Newbold, op. cit. (n. 1), 864–5. Similar practices in the nineteenth century netted 100 per cent profits: E. Gauldie, op. cit. (n. 1), 159.

44 See E. Gauldie, op. cit. (n. 1), 82–92, with special reference to the problems created by lodgers.

45 The legal texts variously concern themselves with flammability, poor construction, and prostitution.

46 cf. Syme, R., Emperors and Biography (1971), 256Google Scholar; the source is late, the anecdote is apocryphal and does not concern Rome, and Diocletian was not a civilian. Probably bed and board were being paid for; cf. Polyb. 2. 15. 6 (inns in Cisalpine Gaul charge half an as daily for board). For such a billing, see CIL IX. 2689 (Aesernia, an inn). Diocletian's price edict omits to regulate rents.

47 The word meritoria suggests straightforward exchange of money for services: Isid. 10. 182; cf. Livy 45. 22. 2. Such an arrangement could be swiftly ended by unilateral action of either party; see Gallo, F., in Synteleia V. Arangio-Ruiz II (1964), 1201–6Google Scholar. The main advantage of the short-term contract was that it gave deversoria a sponge-like ability to absorb a fluctuating number of lower-class tenants; for seasonal fluctuations in Rome's population, see R. F. Newbold, op. cit. (n. 1), 864–5.

48 Most rents cited by E. Gauldie, op. cit. (n. 1), 157–68, are weekly, but not a few are daily. On payment of Roman workers, see de Robertis, F. M., I rapporti di lavoro (1946), 140–3Google Scholar; also the discussion in Dig. 19. 2. 51. 1.

49 E. Gauldie, op. cit. (n. 1), 165–6; such allowance for arrears tends to raise rent overall. Cicero insisted on punctual payment (Att. 12. 32. 2), but still faced problems in collecting (Att. 15. 17. 1; 20. 4); on his urban holdings at Rome (including several insulae), Shatzman, I., Senatorial Wealth and Roman Politics (1975), 403–4Google Scholar; also Walcot, P., Greece and Rome 22 (1975), 122–8CrossRefGoogle Scholar. Note the proverbially speedy caupo compilatus: Petron. 62, 12; cf. Mart., 7. 92. 5–6.

50 Tenants apparently obtained quarters on the information of crude advertising (e.g. Petron. 38. 10; cf. n. 33) or through oral report (e.g. Petron. 6–8). Dig. 19. 2. 60 pr. mentions the ‘showing’ of apartments to prospective sublessees; cf. Mart. 12. 32. 23–4.

51 See above, nn. 8, 42, 49; on such investment, Cic., 2 Verr. 3. 199; Fin. 2. 83; also Att. I. 14. 7 (an insula owned by Quintus); Cael. 17 (Clodius); Off. 3. 66 (lower nobility); cf. CIL VI. 65, 67 (M. Vettius Bolanus, prob. suff. A. D. 66); Dig. 5. 3. 27. 1; note also the commonness of women as owners (Petron. 95. 3). Regular return is emphasized in Cic., Att. 14.10. 3, 11. 2; 16. 1. 5; cf. Duncan-Jones, R., The Economy of the Roman Empire (1974), 296, n. 3Google Scholar, for an isolated example of alimentary funds invested in urban properties; and note the early imperial extension of superficies to private property, cf. Paladini, A., Nov. Dig. Ital. s.v. ‘superficie’ (1971), 941–4Google Scholar, and Pastori, F., in Studi G. Donatuti II (1973), 871–96Google Scholar. Freedmen are known to have inherited insulae, at least: Petron. 71. 2; Dig. 37. 7. 7; CIL VI. 10248, 29791. A more diversified ownership is implied by Vell. 2. 130. 2; Tac., Ann. 15. 43. 2. We might expect freedmen at the entrepreneurial level (compare, for Athens, Isaeus 6. 19). On urban investment, see now P. Garnsey, in Roman Property, (op. cit. in n. 38), 123–132; add to the sources above Nepos, Att. 14. 3; Mart., 3. 31. 2; 4. 37. 4.

52 See R. F. Newbold, op. cit (n. 1), 866–9. These rents would have been due 1 July (above, n. 17). Suetonius clearly implies that landlords were not to be paid, and notes the fiscus' insistence on fresh coin. For the economic background, see Thornton, M. K., Festschrift J. Vogt (Aufstieg und Niedergang) II. 2 (1975), 160–71Google Scholar, very superficial.

53 See R. F. Newbold, op. cit. (n. 1), 861–3; exceptional: Tac., Ann. 15. 43. 2. A Neronian measure encouraging the building of private homes (Gaius 1. 33) hardly represents a responsible approach to the problem of insufficient housing (pace Newbold); on competition in house-building among the wealthy, see esp. Strabo 5. 3. 7. The building codes looked mainly to outward appearance and public security. Vogt, J., Das Erbbaurecht (1950), 5–19Google Scholar, though exaggerated, is still interesting on the management of public urban properties.

54 Remissions occur in 48 (CIL XIV. 4531, from Ostia; Suet., Caes. 38. 2; Dio 42. 41. 1) and in 41 (Dio 48. 9. 5); see Yavete, Z., Latomus 17 (1958), 515–17Google Scholar; Frederiksen, M., JRS 56 (1966), 133–5Google Scholar, establishing the date of Caesar's remission; Royer, J. P., Rev. hist. de droit 45 (1967), 191–240 and 441–50Google Scholar. On the revolutionary nature of these remissions, Caes., BC 3. 21. 1; Dio 42. 22. 3–4; 32. 2; on consequent anger among landlords, Cic., Off. 2. 83, with Frederiksen, 138–9. I have no access to Yavetz, Z., The Plebs Urbana and the Abolition of Debts (1958), 149 f.Google Scholar, in Hebrew; but cf. idem, Plebs and Princeps (1969), 45. Contractual remission, which seems to be associated with unexpected economic hardship (see most recently Thomas, A., in Studi G. Donatuti III (1973), 1271–7Google Scholar), occurred also in the case of urban leases (Dig. 19. 2. 5), a fact not often observed, but cf. Benöhr, J., Das sogennante Synallagma (1965), 105–6Google Scholar.

55 As is obvious from my remarks above, the juristic sources faithfully reflect rental institutions. Modern scholarship on Roman jurisprudence has inclined to the view that it was reflective of and sympathetic to its society and social structure, but internally oriented in its argumentation and (unlike modern law) not even in part aiming for the control and alteration of society. For bibliography, see Kaser, M., Das römische Privatrecht2 (1971), 10–11, 183, 210–14Google Scholar; esp. Lombardi, L., Saggio sul Diritto Giurisprudenziale (1967), 36–54Google Scholar; Vonglis, B., La Lettre et l'Esprit de la Loi (1968), 200–1Google Scholar; further bibliography in Kaser, M., Privatrecht2 II (1975), 569, 576–9Google Scholar; and note the comments of Wieacker, F., in Festschrift M. Kaser (1976), 3–27, esp. 5–6Google Scholar. What is needed is a view of Roman jurisprudence as an instrument of social control, in a framework which goes beyond the trivializing ‘ideologies’ currently fashionable in Italy, but which is not too all-embracing, either.

56 cf. T. Mayer-Maly, op. cit. (n. 10), 153–6. This warranty, which goes beyond the normal duty to furnish the premises, is naturally limited by the state of Roman technology; however, it included, beyond the continuing basic inhabitability of the dwelling (19. 2. 27 pr., 28 pr.- 1; cf. 39.2.13.6, 43.1), its physical security and the non-obstruction of its lights—19. 2. 25. 2, on which Rodger, A., Owners and Neighbours (1972), 87–9Google Scholar. Gross violation justified the tenant in abandoning the premises; smaller violations perhaps allowed a deduction from the rent (arg. ex Dig. 19. 2. 27 pr., 28 pr.-1), but see Nicosia, G., Riv. It. Sc. Giur. 9/10 (1957/1958), 424–6Google Scholar. Note also Dig. 43. 10. 1. 3: ‘repair and deduct’ authority for tenant's performance of a duty publicly imposed on the owner.

57 In the case cited above, n. 3; see also ‘Marini v. Ireland’, in New Jersey Reports 56 (1970), 130–47Google Scholar (Supreme Court of New Jersey; Haneman, J.); however, both these cases admit narrower interpretations than their general language suggests. In some jurisdictions (for example, in Great Britain) legislation replaces Common Law in this area. See generally Donahue, C., Modern Law Review 37 (1974), 242–63CrossRefGoogle Scholar.

58 I would like to thank Profs. P. J. Sijpesteijn (Amsterdam) and H. C. Youtie (Michigan) for help with papyri; and Prof. Charles Donahue (Michigan Law School) for advice on Common Law. Prof. Max Kaser (Salzburg) kindly commented on my typescript. This article was completed in the Institut für Römisches Recht, Salzburg.