Hostname: page-component-7d684dbfc8-csfzr Total loading time: 0 Render date: 2023-09-25T02:35:56.647Z Has data issue: false Feature Flags: { "corePageComponentGetUserInfoFromSharedSession": true, "coreDisableEcommerce": false, "coreDisableSocialShare": false, "coreDisableEcommerceForArticlePurchase": false, "coreDisableEcommerceForBookPurchase": false, "coreDisableEcommerceForElementPurchase": false, "coreUseNewShare": true, "useRatesEcommerce": true } hasContentIssue false

Census Records of the Later Roman Empire

Published online by Cambridge University Press:  24 September 2012


There survive from a number of places in Western Asia Minor and the islands of the Aegean inscriptions recording census registrations. They are undated, but were probably engraved in the late third or early fourth century A.D., when Diocletian and his colleagues and successors are known to have been active in carrying out censuses to serve as the basis of their new system of taxation. All are fragmentary, but some are of sufficient length to yield results of some statistical value on the distribution of landed property, on the density of the agricultural population, and on the proportion of slave to free labour. In view of the extreme paucity of any statistical data for the ancient world they are worth analysis.

Research Article
Copyright ©A. H. M. Jones 1953. 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.)


1 IG XII, ii, 79. This was presumably a refinement found too complicated and soon abandoned.

2 Riccobono, , Fontes iuris Romani 2 II (1940), pp. 795–6Google Scholar (§ 121).

3 IG XII, iii, 343.


5 The printed text gives ΡΓ; ΘΓ᾽ would mean 9⅓.

6 IG XII, iii, 180, as read by Déléage, La Capitation dans le Bas Empire pp. 190–4.

7 Keil and Premerstein, ‘Dritte Reise in Lydien,’ Denkschr. Ak. Wien 1914, no. 85.

8 ‘Table of Brigetio’ (311), Riccobono, Fontes iuris Romani I2 (1940) 93, ‘ab annonario titulo duo kapita excusent, id est tam suum quam uxoris suae,’ Cod. Theod. VII, xx, 4 (325), ‘duo capita excusaturis, id est suum atque uxoris.’ The first was published at Brigetio in Pannonia (where the inscription was found), the second at Antioch (the title of the recipient Maximus is wrongly given as PU in the Codex: he was probably vicarius Orientis, see Seeck, Regesten, p. 118).

9 This was the rule of the λαογραφία of the principate (Wilcken, Grundzüge, p. 189). That it remained the rule after Diocletian is a fairly certain inference from the fact that the nine persons declared by Aurelius Sacaon in 310 (Wilcken, Chrestomathie 210, cf. also SB 7673) are all males. The corresponding tax in early Arab times, ἀνδρισμόϛ (the name is now proved by P. Ryl. IV, 658, to date back to the fourth century), certainly fell on males only (see Bell in the introduction to P. Lond. IV, 1419).

10 Cod. Theod. XIII, xi, 2.

11 IG XII, iii, 343, 346; Keil and Premerstein, l.c. (see note 7).

12 Dig. L, xv, 3. This citation from Ulpian was presumably preserved by the compilers of the Digest because still valid. The edict of Aurelius Optatus (SB 7622) shows that Diocletian in 297 laid down lower and upper age limits for Egypt, but they are not known, save that in Aurelius Isidore's declaration (SB 7673) a boy of 3 is exempt (ἀτέλης) but in Aurelius Sacaon's (Chr. 1, 210) one of 12 pays tax (ὑποτέλης) and a man of 55 is still liable.

13 Cod. Theod. XI, iii, 2 (327), ‘mancipia adscripta censibus intra provinciae terminos distrahantur, et qui emptione dominium nancti fuerint inspiciendum sibi esse cognoscant.’

14 Cod. Theod. VII, i, 3 (341), ‘quicumque militum ex nostra auctoritate familias suas ad se venire meruerint, non amplius quam coniugia liberos servos etiam de peculio castrensi emptos neque adscriptos censibus ad eosdem excellentia tua dirigi faciat.’

15 At Chios (Déléage, o.c, pp. 182–6) and Lesbos (IG XII, ii, 76d, 78c).

16 At Tralles (BCH 1880, 336–8).

17 At Thera (IG XII, iii, 343).

18 Cod. Just. XI, xlviii, 7 (371), ‘quemadmodum originarios absque terra ita rusticos censitosque servos vendi omnifariam non licebit.’

19 Cod. Theod. XI, i, 14 (371), ‘penes quos fundorum dominia sunt, pro his colonis originalibus quos in locis isdem censos esse constabit vel per se vel per actores proprios recepta compulsionis sollicitudine implenda munia functionis agnoscant. sane quibus terrarum erit quantulacumque possessio, qui in suis conscribti locis proprio nomine libris censualibus detinentur, ab huius praecepti communione discernimus. eos enim convenit propriae commissos mediocritati annonarias functiones sub solito exactore cognoscere.’

20 Cod. Theod. XI, xxiv, 6 (415).

21 At Chios (Déléage, o.c. pp. 182–6) some farms have παροίκων κεφαλαί only (coloni, clearly including their animals, since they must have had some), others have παροίκων κεφαλαί and δούλων κεφαλαί and ζᾡων κεφαλαί. At Tralles (BCH 1880, PP' 336–8) the figures of κεφαλαί attached to each farm clearly include animals, since they contain small fractions. Some owners record above their farms δούλων καὶ ζᾡων κεφαλαί and some ζᾡων κεφαλαί as well. The former are animals in charge of the slaves, the latter presumably animals supplied for the use of the coloni. In this list some owners have a fractional figure of ζᾡων κεφαλαί immediately after their names. These are presumably animals kept at the owner's residence in town. The other animals are specified as being on this or that farm.

22 At Thera (IG XII, iii, 343, 346). In Lesbos (IG XII, ii, 76) few farms have animals. Those registered without comment presumably belonged to the landowner. Other entries of the form βοῦς δ᾿ Ἐλπιδηφόρος· ἵππον α’ Κυζίκιος καὶ Ἐλπιδηφόρος· presumably mean that Elpidephorus declared 4 oxen and he and Cyzicius 1 horse. Elpidephorus is known as tenant of another farm (IG XII, ii, 79) which had no pasture. He presumably kept his beasts on a farm of which he was neither owner nor tenant.

23 IG XII, iii, 343.

24 IG XII, iii, 345. The words at the end of line 1 should be read Ἀττάλου λαμπ(ροτάτου). Excluding the heading 16 lines survive, and some lines seem to contain two farms (ll. 9, 10, 12, and probably 2 and 3). I have omitted lines 2–3 and 14 from my calculations as their figures are incomplete.

25 IG XII, iii, 346.

26 IG XII, ii, 76.

27 IG XII, iii, 180, as read by Déléage, o.c, pp. 190–4.

28 BCH 1880, 336–8.

29 Kern, Inschriften von Magnesia am Maeander, no. 122. In the calculations which follow I have followed the editor's interpretation of the figures except as specified in note 48 below (p. 54).

30 On block b (probably As) there are on the right the initial Bs of a lost column of farms.

31 Valerianus son of Romus (a I, e 17–8), Variana (a 10, b 15), Patroina (b 4, e 14, g 3), Paulus philosebastos (a 7, f 7–8, g 4), Pisticus (b 12, 18, d 16–7), Priscillianus v.c. (a 12–3), Severianus the tribune (b 1–2, 9, d 5, e 7), Tychicus son of Eugnomonius (e 4–5, f 6), Tyrannus (e 1 and 3, 12), Tyrannus the Asiarch (e 10–11), Eutychis of Ephesus (d 12, e 13), Philip of Tralles (b 7, g 1).

32 Trallians, b 7, 14, 16, c 3, d 3; Ephesians, d 4, 8, 12, e 13, 15, f 9, h 2; Colophonian, h 5.

33 Capitolinus (d 6), Eutychus (d 15), Priscillianus (a 12–3), Aristocleia (b 3), Hermonactiane (d 9).

34 c 2; of his name only -nus survives and he may be identical with Capitolinus (d 6).

35 Maximianus (a 9), Metrodorus (a 5).

36 Tyrannus (e 10–1).

37 Heracleides (g 2), Mandrogenes (a 9), Paulus (a 7, f 7–8, g 4), Phanius (f 4), Pollio (d 1), Tychicus (a 8). The council of Magnesia is styled ἡ φιλοσέβαστος βουλή in nos. 179 and 193. Unless Philosebastos means member of the council, the absence of decurions would be strange.

38 Seen. 31.

39 See n. 37.

40 See n. 28.

41 See n. 31.

42 Seen. 31.

43 IG XII, ii, 76a, 79.

44 IG XII, iii, 343.

45 IG XII, ii, 79. Elpidephorus' animals appear on no. 76e.

46 IG XII, iii, 180. There is a mysterious fourth column of figures on the right which Déléage (o.c, pp. 190–4) interprets as capita animalium. I cannot explain them, but I cannot accept Déléage's theory both for the reason given in the text and because the items are clearly labelled iuga or iuga vel capita (ΖΚ or ΖΥΓΚ).

47 BCH 1880, 336–8. In Tatian's estate I read the figure of capita for Ἀγρὸς Ἄραρα (line 24) at not (ΓΛ´Ρ´ for Γ´Λ´Ρ´) as there cannot have been less than ½ a caput of coloni. In the text as printed the fact that the stone is broken away on the right towards the bottom is not obvious, but it should be noted that in lines 35 and 41 the symbol қ is followed by no figure and that some fractions must be missing in Latron's figures of iugatio to make up the total.

48 These are a 1, a 12, c 1, c 4, c 6, d 1–2 (where by the position of the iuga figures there were probably capita figures following as in d 12), d 5 and 16 (when the jumble of confused symbols probably conceals capita figures), e 3, e 18, f 9 (where again there is a jumble of figures as in d 5 and 16), g 1, h 1. In b 18 I read as iuga 4½ ¼ capita 11⅓ (not 91⅓, which would be absurd); that is, I assume that the capita figure was ΙΑΓ and the loop a fault in the stone. The following have no figures at all: b 1–6, d 17–8, e 1, 2, h 2. I read b 10, ΖḰ as iuga (the editor wrongly writes 20) and a 5, ΚϛΧι´ as capita (not the editor wrongly renders it).

49 g 4, e 11, b 12. Other farms with an excessive capitation are a 5, 6, 9, b 7, e 4.

50 C 2, e 14, c 5.

51 d 9, f 8, f 7, e 8, h 4, e 13, e 17.

52 Keil and Premerstein, o.c., no. 87,

Κινάμουρα σὺν τοῖς ἐνκεκτημ᾿ · ζυ…

Διδείφυτα σὺν τοῖς ἐνκεκτημ᾿…

53 o.c., no. 86.

54 o.c., no. 85.

55 Lactantius, de Mort. Persec. 23, ‘hominum capita notabantur, in civitatibus urbanae et rusticae plebes adunatae, fora omnia gregibus familiarum referta, unusquisque cum liberis cum servis aderant’ (speaking of Galerius): Cod. Theod. XIII, x, 2 (313). ‘plebs urbana, sicut in Orientalibus quoque provinciis observatur, minime in censibus pro capitatione sua conveniatur, sed iuxta iussionem nostram immunis habeatur, sicuti etiam sub domino et parente nostro Diocletiano Seniore Augusto eadem plebs immunis fuerat’ (Licinius to the governor of Lycia-Pamphylia after the fall of Maximin).

56 IG XII, iii, 343, 346.

57 Livy, xxxv, 40 (10 at Vibo in 192), XXXIX, 44 (6 at Potentia and Pisaurum in 184), XXXIX, 55 (5 at Mutina, 8 at Parma, 10 at Saturnia in 183), XL, 29 (5 at Graviscae in 181), XLII, 4 (10 to citizens, 3 to allies in Cisalpine Gaul in 173).

58 Livy, XXXVII, 57 (50 at Bononia in 189), XL, 34 (50 at Aquileia in 180), XLI, 13 (51½ at Luna in 180). The settlers here may well have been intended to be farmers employing (slave) labour. Thi s is suggested by the fact that at Bononia equites got larger plots (70) and at Aquileia centurions and equites got 100 and 140. At Vibo also equites received a double allowance (20).

59 Suet. Julius, 20, 3Google Scholar; Cicero, Ep. ad. Att. II, xvi, 1Google Scholar. These allotments were on the ager Campanus and Stellas, famous for their fertility.

60 Cato, de Agr. II.

61 Columella 11, 12, 7. This is not extravagant seeing that he estimates that each iugerum requires 4 man-days to plough (3 times), 1 man-day to harrow, 3 to hoe, 1 to weed, 1½ to harvest (11, 13, 1) as well as 1 man-day to cut the straw after harvest (XI, 2, 54). This work alone would occupy his 8 men 300 days in the year, without allowing for sickness or bad weather, or carting, repairs, and other odd jobs.

62 Déléage, o.c, 182–6.

63 IG XII, iii, 343.

64 BCH 1880, 336–8.

65 IG XII, ii, 79.

66 IG XII, ii, 76, 78.

67 In what follows I have ignored fractions smaller than a half. I cannot vouch for the absolute accuracy of my arithmetic; for, apart from my incapacity for adding long columns of figures, I find Greek numerical notation troublesome, and missing or mutilated figures add to the confusion. In dealing with them I have exercised my discretion, eliminating those where the element of doubt is large, but including those where the missing figure is relatively unimportant (e.g. I have reckoned in ρπ[ ] as 180 +, but ignored [ ]β).

68 This appears from P. Princeton 134 and P. Strassb. 45, discussed at the end of this paper.

69 LI. 143–4, 170–1. 186–9, 210–3, 235–6, 241–253, 304–8, 314–7, 424–6, 450–3. 454–7, 461–2, 536–8, 547–9, 583–8, 589–591, 596–9, 716–8.

70 L. 88, ἐπ᾿ ὀνόματος Διογένους Πάριδος under Ἀμμωνίων Ἀντωνίνου; l. 251, καὶ ὄνομ᾿ Σιλβανοῦ Ἑρμαπόλλωνος under Ἡρακλέων Ὑπερεχίου; l. 280, καὶ ὄνομ᾿ Ὡρίωνος under Ἱεροκλῆς Ἑλλαδίου. Silvanus son of Hermapollon occurs elsewhere out of place, over his brother ( ?) Pamunis, son of Hermapollon (ll. 416–8). Another name which appears out of place is Ἑρμαπόλλων Μαικηνᾶ, who follows κλ(ηρονόμοι) Μαικηνᾶ Φιβίωνος. Σέρηνο[ς υ]ἱός follows his father in l. 782.

71 Ll. 154, 375, 463, 533, 547, 581, 611, 636, 639–640 (brothers), 755, 774.

72 Ll. 747–752 (Antinoite); 127, 137–8, 474–5 (Hermopolite).

73 See l. 72.

74 Ll. III, 286, 561, 660; there is also one case (526) where an owner who holds land in several pagi has public land only in one pagus.

75 See A. H. M. Jones, Cities of the Eastern Roman Provinces 345, 482, n. 64. SB 8942 has since confirmed my conjecture that Cusae was already a city.

76 Chr. 1, 29.

77 Ll. 521, 747–751. 566, 592.

78 e.g. ll. 143–4, 186–9, 210–3, 304–8, 314–7, 450-3, 461–2, 583–8, 596–9.

79 Ll. 186–9.

80 Ll. 77–8, 128, 159, 183, 214, 247, 267, 325, 342–3, 364, 380, 395, 438, 460.

81 L. 299 (κλ᾿ Ἀμμωνίου Υπερεχίου).

82 Ll. 64 (Ἀκύλας Ὀλυμπιοδώρου), 123 (Γεννάδιος Διοκλέους) 129 (Διοσκουρίδης Αἰλιανοῦ), 241 (Ἡρακλέων Ὑπερεχίου), 274 (Ἱεροκλῆς Ἑλλαδίου), 408 (Πινουτίων Ὀλυμπιοδώρου).

83 Ll. 515, 612, 625, 697, 707, 714. There is also a wealthy primipilaris (over 206 arurae) at Hermopolis l. 60). Primipilaris at this date, of course, means former princeps of the provincial officium.

84 Ll. 680, 791.

85 Ll. 509, 546, 550, 604; in the Hermopolite list (ll. 160, 380) there are another beneficiarius and an actuarius, holding 65½ and 60 arurae.

86 L. 589.