Hostname: page-component-586b7cd67f-gb8f7 Total loading time: 0 Render date: 2024-11-23T22:28:59.322Z Has data issue: false hasContentIssue false
  • English
  • Français

The Internal Organization of the Arretine Terra Sigillata Industry: Problems of Evidence and Interpretation*

Published online by Cambridge University Press:  14 March 2012

Gunnar Fülle
Gunnar Fülle
Affiliation:
Brasenose College, Oxford
Get access Rights & Permissions [Opens in a new window]

Extract

If the field of ancient economy is a battlefield, arguments based on pottery research certainly belong with the best of the weapons. Among the various kinds of pottery serving ancient historians as sources, red-gloss pottery (terra sigillata) manufactured in several parts of the Roman Empire plays an outstanding role. This special kind of pottery bears inscriptions in the form of stamps referring to persons involved in its production. In combination with the archaeological contexts of stamp finds, such as excavated sites of production, transportation, storage, and consumption, these inscriptions enable us to gain an insight into the structures of production and distribution. An additional reason why Roman red-gloss pottery is of very great interest to students of the ancient economy is that it was mass-produced, and exported to all parts of the Empire. Results of research in this field are, therefore, frequently used as weighty arguments in the discussion of the character of the Roman economy as a whole.

Type
Articles
Information
The Journal of Roman Studies , Volume 87 , November 1997 , pp. 111 - 155
Copyright
Copyright © Gunnar Fülle 1997. 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 cf. the bibliographies by Comfort, H., ‘terra sigillata’, RE Supp. 7 (1940), 1295–1352Google Scholar and Pucci, G., ‘Terra Sigillata Italica’, in Enciclopedia dell'Arte Antica Classica e Orientate, Atlante delle Forme Ceramiche II (1985), 400–4Google Scholar.

2 e.g. Gummerus, H., ‘Industrie und Handel (Beiden Romern)’, RE 9 (1916), 1439–1535Google Scholar, at 1491.

3 cf. Finley, M. I, The Bücher-Meyer-Controversy (1979).Google Scholar

4 Rostovtzeff, M., Social and Economic History of the Roman Empire 2 (1956), 35–6Google Scholar.

5 ibid., 3.

6 ibid.

7 ibid., 36.

8 ibid.

9 ibid., 69–70.

10 Frank, T., An Economic History of Rome to the End of the Republic (1920), and Gummerus, op. cit. (n. 2), 1439–1535Google Scholar; cf. Rostovtzeff, M., Social and Economic History of the Roman Empire (1926), 490 n. 8 and 498 n. 33Google Scholar. Despite the title of Frank's book, it covers some industries of the time of the early Principate. In the German edition of his work, Rostovtzeff referred to the successor of Frank's book, An Economic History of Rome, which is in this part identical with its predecessor; cf. Rostovtzeff, M., Sozial- und Wirtschaftsgeschichte des römischen Kaiserreiches (1931), 241 n. 13 and 251 n. 33Google Scholar, and Frank, T., An Economic History of Rome 2 (1927), 219–22.Google Scholar

11 Gummerus, op. cit. (n. 2), 1487–8.

12 Frank, op. cit. (n. 10, 1920), 167–8.

13 ibid., 165–6.

14 Comfort, op. cit. (n. 1), 1296. In the same decade, Westermann, W. L., ‘Industrial slavery in Roman Italy’, JEH 2 (1942), 149–63CrossRefGoogle Scholar, esp. 158, also took his information from Gummerus' article (op. cit. (n. 2)).

15 Kiechle, F., Sklavenarbeit und technischer Fortschritt im römischen Reich, Forschungen zur Antiken Sklaverei 3 (1969), 702Google Scholar, quoting v Klaveren, J., ‘Die Manufakturen des Ancien Régime’, VSWG 51 (1964), 145–91Google Scholar, esp. 145–6.

16 Pucci, G., ‘La produzione della ceramica aretine. Note sull'industria nella prima etá imperiali’, DArch 7 (1973), 255–93Google Scholar; Peacock, D. P. S., Pottery in the Roman World: An Ethnoarchaeological Approach (1982), 121–2.Google Scholar

17 ibid.

18 Finley, M. I., The Ancient Economy 2 (1984), 26–7Google Scholar.

19 ibid., 137.

20 Carandini, A., ‘Sviluppo e crisi delle manifatture rurali e urbane’, in Società romana e produzione schiavistica II. Merci, mercati e scambi nel Mediterraneo (1981), 249–60Google Scholar, esp. 256–7; cf. idem, Schiavi in Italia, Gli strumenti pensanti dei Romani fra tarda Repubblica e medio Impero, Studi NIS Archeologia 8 (1988), 333–4.

21 cf. e.g. De Martino, F., Wirtschaftsgeschichte des alten Rom 2 (1991), 339Google Scholar; Greene, K., The Archaeology of the Roman Economy (1986), 160Google Scholar; Kloft, H., Die Wirtschaft der griechisch-römischen Welt (1992), 172–3Google Scholar; Aubert, J.-J., Business Managers in Ancient Rome. A Social and Economic Study of Institores 200 B.C.–A.D. 250, Columbia studies in the classical tradition 21 (1994), 296Google Scholar.

22 For details and literature on branch workshops and related problems cf. below Section IX.

23 cf. e.g. Aubert, op. cit. (n. 21), 217; Carandini, op. cit. (n. 20, 1981), 249–60, esp. 256–7 and Carandini, op. cit. (n. 20, 1988), 333–4.

24 cf. Peacock, op. cit. (n. 16), 9.

25 e.g. Gummerus, op. cit. (n. 2), 1491.

26 Peacock, op. cit. (n. 16), 9–10, 43–6.

27 The following considerations take necessary data from Prachner, G., Die Sklaven und Freigelassenen im arretinischen Sigillatagewerbe. Epigraphische, nomen klatorische sowie sozial- und wirtschaftsgeschichtliche Untersuchungen der arretinischen Firmen- und Töpferstempel, Forschungen zur antiken Sklaverei 12 (1980)Google Scholar, hereafter referred to as SFAS, who analysed stamps of twenty-nine selected Arretine firms and produced detailed tables containing potters' stamps and information about the amount and sorts of vessels made by each of them, as well as the places where the remains were found. Including new finds, his data is more comprehensive than that of the Corpus Vasorum Arretinorum, compiled by A. Oxé, edited by H. Comfort (Antiquitas 3 (1968)); hereafter referred to as CVArr. He was also able to include the as yet unpublished substantial finds made in Neuβ, later published in Ettlinger, E., Novaesium IX. Die italische Sigillata von Novaesium, Limesforschungen 21 (1983)Google Scholar, in Haltern, later published in Schnurbein, S. v, Die unverzierte Terra Sigillata aus Haltern, Bodenaltertümer Westfalens 19 (1982)Google Scholar, and in Dangstetten, meanwhile partly published in G. Fingerlin, Dangstetten I. Katalog der Funde (Fundstellen 1–603), Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg 22 (1986). See also SFAS, vii (Vorwort) and 3 with n. 14.

28 Manacorda, D., ‘Appunti sulla bollatura in etá romana’, in Harris, W. V. (ed.), The Inscribed Economy. Production and Distribution in the Roman Empire in the Light of Instrumentum Domesticum, JRA suppl. ser. 6 (1993), 37–54Google Scholar, esp. 37.

29 Siebert, G., ‘Signatures d'artistes, d'artisans et de fabricants dans l'antiquite classique’, Ktema 3 (1978), 111–31Google Scholar; Manacorda, op. cit. (n. 28), 37; Aubert, op. cit. (n. 21), 201–318.

30 cf. for transport and storage vessels in general Manacorda, op. cit. (n. 28), 37–8. On details esp. about amphorae recently B. Liou and A. Tchernia, ‘L'interprétation des inscriptions sur les amphores Dressel 20’, in Epigrafia della produzione e della distribuzione. Actes de la VIIe Rencontre franco-italienne sur l'épigraphie du monde romain (Rome, 5–6 juin 1992), Collection de l'École française de Rome 193 (1994), 133–56; for dolia cf. also Aubert, op. cit. (n. 21), 246–56, 265–7, 269–74; D. Manacorda and C. Panella, ‘Anfore’, in Harris, op. cit. (n. 28), 55–64; on wooden barrels now Baratta, G., ‘Bolli su botti’, in Epigrafia della produzione e della distribuzione. Actes de la VIIe Rencontre franco-italienne sur l'épigraphie du monde romain (Rome, 5–6 juin 1992), Collection de l'École française de Rome 193 (1994), 555–65.Google Scholar

31 Dig. 18.6.1.2. Manacorda, op. cit. (n. 28), 38.

32 Domergue, C., ‘Production et commerce des métaux dans le monde romain: l'exemple des métaux hispaniques d'après l'épigraphie des lingots’, in Epigrafia della produzione e della distribuzione. Actes de la VIIe Rencontre franco-italienne sur l'épigraphie du monde romain (Rome, 5–6 juin 1992), Collection de l'École française de Rome 193 (1994), 61–91Google Scholar, esp. 62–4, 71–3; Manacorda, op. cit. (n. 28), 38; Colls, D. et al. , ‘Les lingots de plomb de l'épave romaine Cabrera 5 (Ile de Cabrera, Baléares)’, Archaeonautica 5 (1986), 31–80CrossRefGoogle Scholar, esp. 69–70.

33 Manacorda, op. cit. (n. 28), 38; Steinby, M., ‘I senatori e l'industria laterizia urbana’, Tituli 4 (1982), 227–37Google Scholar, esp. 232–3; eadem, ‘L'industria laterizia di Roma nel tardo-impero’, in Giardina, A. (ed.), Societá e impero tardo antico II (1986), 99–164Google Scholar, esp. 100, 106–7, 149–50; eadem, ‘L'organizzazione produttiva del laterizi: un modello interpretativo per l'instrumentum in genere?’, in Harris, op. cit. (n. 28), 139–43.

34 Aubert, op. cit. (n. 21), 303–18; Harris, W. V., ‘Roma terracotta lamps: the organization of an industry’, JRS 70 (1980), 126–45Google Scholar.

35 Zabehlicky-Scheffenecker, S., ‘TK — Zur kommerziellen Verbindung des Magdalensberges mit Aquileia’, in Lebendige Altertumswissenschaft (Festschrift H. Vetters) (1985), 252–4Google Scholar, describes the ligated abbrevation ‘TK’ incised on terra sigillata vessels from Arezzo and the plain of the river Po, on an amphora-fragment, and on a stone-weight, occasionally also in connection with numerals. She suggests that these marks were applied by an Aquileia-based wholesaler, possibly belonging to the prominent gens Kania from Aquileia.

36 cf. Manacorda, op. cit. (n. 28), 38, 44–5.

37 Manacorda, D., ‘Le anfore dell'Italia repubblicana: aspetti economici e sociali’, in Amphores romaines et histoire économique. Dix ans de recherche, Collection de l'École française de Rome 114 (1989), 443–67Google Scholar.

38 Manacorda, op. cit. (n. 28), 39.

39 ibid., 41–4.

40 Steinby, op. cit. (n. 33, 1993), 139–43, stresses that the interpretation of brick stamps as abbreviated locatio conductio contracts excludes their use as part of a guarantee system. She supports her view by pointing out that bricks of very good and of very bad quality were stamped equally. In addition, the user would have had difficulties in deciphering badly applied stamps, abbreviated names, and stamps showing symbols. Thus Steinby suggests that the stamps were used in the context of production and distribution only (ibid., 141).

41 Suggested by Rodríguez-Almeida, E., ‘Graffiti e produzione anforaria della Betica’, in Harris, op. cit. (n. 28), 95–105, esp. 99.Google Scholar

42 Manacorda, op. cit. (n. 28), 43; Aubert, op. cit. (n. 21), 264, 275, 300, and passim. Against the application of this assumption to terra sigillata Pucci, G., ‘I bolli sulla terra sigillata: fra epigraphia e storia economica’, in Harris, op. cit. (n. 28), 73–80, esp. 74.Google Scholar

43 This problem occurs especially when products with identical stamps come from different places, as is the case in the trade of terracotta lamps. It is not possible to decide, if e.g. a provincial producer was a representative of the Italian main manufacturer whose name occurs on provincial products, or simply made unauthorized imitations. Cf. on this problem Aubert, op. cit. (n. 21), 317. Cf. also Marino, F., ‘Appunti sulla falsificazione del marchio nel diritto romano’, ZRG 105 (1988), 771–5Google Scholar.

44 On the use of amphorae Aubert, op. cit. (n. 21), 259–62.

45 Pucci, op. cit. (n. 42), 75. SCOTTIVS FECIT ARETINVM, RVTENVS FEC(it) ARETINVM (C. Bémont, A. Vernhet and F. Beck, La Graufesenque, village de potiers gallo-romains, Catalogo della mostra (1987), 24; Vernhet, A., ‘Centre du production de Millau, Atelier de la Graufesenque’, in Bémont, C. and Jacob, J. P. (eds), La terre sigillée gallo-romaine. Lieux de production de Haut-Empire: implantation, produits, relations, Documents de archéologie française 6 (1986), 100Google Scholar, respectively); ARRETINVM or even ARRET(inum) VERV(m) (CVArr 132); on the origin J.-P. Morel, ‘Artisanat et colonisation dans l'ltalie romaine aux IV et III siècles av. J.C,’ DialArch 1988, 49–63.

46 This is suggested by stamps which have no sense and show a kind of imitation of alphabetical characters (Pucci, op. cit. (n. 42), 75).

47 cf. Pucci, op. cit. (n. 42), 73–9.

48 Pedroni, L., ‘La scomparsa dei bolli sulla ceramica a vernice nera’, Samnium 61 (1988), 1–17Google Scholar.

49 Pucci, op. cit. (n. 42), 73–4. According to another suggestion by Pedroni (op. cit. (n. 48)) the plebiscite of 218 B.C. had the opposite effect, namely the disappearance of stamps that were applied on blackglazed ware from Campania in the third century B.C. The manufacturers are suspected of having dropped the use of stamps in order to conceal certain commercial activities that had been forbidden by the law. There is, however, no evidence for this assumption, which could in any case not explain why stamps were used in this period (Pucci).

50 Pucci, op. cit. (n. 42), 74.

51 ibid.

52 ibid., 74–5.

53 Peacock, op. cit. (n. 16), 125–6; Marichal, R., ‘Noveaux graffites de la Graufesenque’, REA 76 (1974), 84–110Google Scholar and 266–99, esp. 272–3; Strobel, K., ‘Einige Bemerkungen zu den historisch-archäologischen Grundlagen einer Neuformulierung der Sigil latenchronologie für Germanien und Rätien und zu wirtschaftsgeschichtlichen Aspekten der römischen Keramikindustrie’, MBAH 6 (1987), 75–115Google Scholar, esp. 100–11.

54 Pucci, G., ‘A sigillata kiln in Valdichiana (Central Etruria)’, RCRF 27’28 (1990), 15–23Google Scholar; idem, La fornace di Umbricio Cordo (1992).

55 Pucci, op. cit. (n. 42), 75.

56 ibid.

57 For stamps on black-glazed ware and related theories cf. now Pucci, op. cit. (n. 42), 73–4.

58 Gamurrini, G. F., ‘Di una nuova figulina di vasi neri e rossi, scoperta all'Orciolaia presso Arezzo’, NdS 1890, 63–72Google Scholar, esp. 68–70. Cf. also SFAS, 213 with n. 119.

59 e.g. CVArr 2360, 2333, 2346.

60 Peacock, op. cit. (n. 16), 115.

61 cf. on these questions Kenrick, P. M., ‘Potters' stamps’, in Ettlinger, E.et al., Conspectus Formarum Terrae Sigillatae Italico Modo Confectae, Materialien zur Römisch-Germanischen Keramik 10 (1990)Google Scholar, hereafter referred to as Conspectus, 147–8.

62 cf. SFAS, 2–3.

63 Kenrick, op. cit. (n. 61), 147.

64 The opposite direction of writing is rare, but not unknown on stamps from Arezzo, cf. the stamps EROS AVILI and LIVA/EROS (CVArr 241).

65 I have decided to use the term ‘given name’, since alternative ones such as ‘forename’, ‘first name’ or ‘Christian name’ depend for their meaning on modern name systems, whereas the terms praenomen or cognomen refer to the tria nomina system used for free Roman citizens. The neutral ‘slave name’ means in this context the full form including the master's name. By contrast, the term ‘given name’ reflects the fact that a slave was called by a name of his master's choosing. This could be either what he was called before enslavement, or a new name issued at random by his master. The best modern equivalent would probably be the German Rufname, i.e. the name by which one is usually called. In addition, the use of the term cognomen for a slave's given name would lead to the existence of name forms with two cognomina, since some forms of slaves' names contain the master's tria nomina. Cf. the terminological confusion in Aubert, op. cit. (n. 21), 220–2, where ‘slave name’, ‘personal name’, and cognomen are all used when referring to a slave's given name. Such a complication can easily be avoided by the consistent use of the suggested term ‘given name’ for the name component in question.

66 i.e. either nomen gentile alone, cognomen alone, nomen gentile with cognomen, duo nomina, or tria nomina.

67 Oxé, A., ‘Zur älteren Nomenklatur der römischen Sklaven’, RhM 59 (1904), 108–40Google Scholar.

68 The name forms in question are: master's nomen gentile in the genitive + the slave's given name, e.g. Aureli Eros, and master's duo nomina in the genitive + the slave's given name, e.g. L. Aureli Eros. Oxé, op. cit. (n. 67), 135–40, tentatively considered that these forms refer to freedmen stressing there were no indications that the forms represent slaves or refer to two persons. He backed his interpretation by using the fact that some persons, whose names appear on these stamp-types, are also encountered on stamps which positively show that freedmen are meant, e.g. L.TITI/THYRSI and L.TITI L.L/THYRSI. The stamp P.MESEINV(s)/AMPHIO.S proves that the usually abbreviated second name is not always in the genitive, and that therefore an interpretation like (ex figlinis) Cn. Atei, Hilarus (fecit) can be discounted. He did, however, not discuss the meaning of the abbreviation ‘S’ on this stamp, which normally indicates servile status. Recently Aubert, op. cit. (n. 21), 288, renewed the interpretation of these name forms as (ex officina) Cn. Atei, Hilarus (fecit), but without convincing arguments. The point that in the stamp A.TITI/FIGVL/ARRET the last abbreviation should be read as the name Arretius, who would have been an employee of the figulus A. Titius is not conclusive. FIGVL(us) could have been a job title, or the cognomen of A. Titius (cf. also Salomies, O. and Solin, H., Repertorium nominum gentilium et cognominum Latinorum2, Alpha-Omega 80 (1994), 322Google Scholar; Kajanto, I., The Latin Cognomina, CHL 36.2 (1965), 322Google Scholar. On job titles as cognomina cf. ibid., 82–4. See also Section VI.B below). Furthermore, other plausible readings of ARRET are easily possible, e.g. as a reference to the place of origin of the vessel. Aubert's second example, the exceptional stamp L.SEMPR/L.GELLI, has nothing to do with the forms in question and should, therefore, be discussed separately. Anyhow, the occurrence of stamps with additions like ex officina or fecit in brick production, or in the Gaulish terra sigillata trade, cannot be used as a basis for the reading of stamps of producers from Arezzo without having additional clues.

69 Master's tria nomina in the genitive + the slave's given name, e.g. L. Aureli Cottae Eros. Oxé, op. cit. (n. 67), 139–40, mentions stone inscriptions proving that freedmen did indeed use this form, sometimes with additional indication of status. In one of these inscriptions (CIL II 2093) we find L. VALERI LAETI (et) M. VALERI VETVSTI LIBERTVS VERNA (et) M. VALERI VETVSTI PRIMA VERNAE VX(or). Verna and Prima, both typical slave names, also encountered on Arretine stamps (cf. SFAS, 230), are evidently given names of two freedmen, who used their patrons' tria nomina in the genitive, and before their given names, although they were freed. Prachner, SFAS, 168–9, 207–8, 211, explained the very obscure stamp LVMSC/NOSTA (CVArr 2412) as L. Um(brici) Sc(auri) No(thus) Sta(tuliber). However, the solution of STA as statuliber is not at all positive — the abbreviation does not appear in stone inscriptions, but only on Arretine terra sigillata stamps. In three other cases (CVArr 2040, 2042, 2044) the abbreviation STA or ST follows the nomen gentile of the master and could be read as an abbreviation of the master's cognomen (on the large number of cognomina beginning with ‘Sta’ see Salomies and Solin, op. cit. (n. 68), 407 and 504). The objection that this is not very likely, since we do not have an independent stamp with such a master's name on its own, which is the usual case, is based on the premise that the master of a slave working in pottery production must have been also involved in it (cf. SFAS, 151 with n. 1). This was not necessarily so, because slaves could act economically independently from their master, e.g. for third parties for payment to their master, as is shown below in Section VI.A. We do not know, for instance, a single master's stamp of the large group of dependants of Publius (CVArr1414–45).

70 The term ‘single-name’ means here stamps containing a single component of a complete name form only, e.g. given name on its own, or cognomen alone, or nomen gentile alone, etc.

71 This suggestion was made by Prachner (SFAS, 205), who adds that such unequivocal and datable sets of stamps are not yet available.

72 CVArr 2343.

73 SFAS, 157.

74 CVArr 2360b.

75 CVArr 2344.

76 CVArr 4345.

77 SFAS, 157–8.

78 This interpretation was suggested by M. Steinby (oral communication).

79 e.g. CVArr 2360a and b, 2333, 2346.

80 P.Oxy. L 3595–7. Cf. for L 3595 Cockle, H., ‘Pottery manufacture in Roman Egypt’, JRS 71 (1981), 87–97Google Scholar; for all three papyri Hengstl, J., ‘Einige juristische Bemerkungen zu drei Töpferei-Mieturkunden’, Studi Biscardi IV (1983), 663–73Google Scholar; Strobel, op. cit. (n. 53); H.-J. Drexhage, Preise, Mieten/Pachten, Kosten und Löhne im römischen Ägypten bis zum Regierungsantritt Diokletions. Vorarbeiten zu einer Wirtschaftsgeschichte des römischen Agypten (1991); Aubert, op. cit. (n. 21), 253–6.

81 In P.Oxy. L 3595 a whole pottery was leased for two years. In P.Oxy. L 3596 one quarter of the pottery in question was leased for one year, in P.Oxy. L 3597 one third for the same period. From a later papyrus, P.Lond. II 944 (A.D. 517), we learn that up to one fourteenth of a pottery could be a subject of a contract; P.Cairo Masp. I 67110 (A.D. 565) shows that a lease could be concluded for life.

82 P.Oxy. L 3597. For details see Strobel, op. cit. (n. 53). 94.

83 cf. n. 81.

84 P.Oxy. L 3595.

85 P.Oxy. L 3596–7. See for details Strobel, op. cit. (n. 53), 94.

86 P. Tebt. II 342. Cf. for details Strobel, op. cit. (n. 53), 96; Drexhage, op. cit. (n. 80), 91–2.

87 P. Mert. II 76. See for details Strobel, op. cit. (n. 53), 96.

88 Strobel, op. cit. (n. 53), 96.

89 cf. the comprehensive treatise on the locatio conductio by Kaser, M., Das römische Privatrecht2, HdAW 10,3,3 (1971), I, 562–72Google Scholar.

90 Strobel, op. cit. (n. 53), 95, follows Hengstl, op. cit. (n. 80), 665–8, in considering the contracts in question as a ‘handwerksspezifische Variante der hellenistisch-römischen Werkverträge’. By contrast, Drexhage, op. cit. (n. 80), 98, pleads for contracts of labour, because the subject of the contract is according to him ‘eine festgelegte Arbeitsleistung in einer festumrissenen Zeit’. However, Aubert, op. cit. (n. 21), 254, excludes this option, on the ground that the potter was allowed to work for his own profit.

91 Aubert, op. cit. (n. 21), 254.

92 Kaser, op. cit. (n. 89), 564; Hengstl, op. cit. (n. 80), 668–70.

93 Peacock, op. cit. (n. 16), 9. Peacock, ibid., 11, also suggests a category of ‘dispersed manufactory’, i.e. the production of a certain product not in a single building, but at the workers' home, a production which requires full-time occupation of the artisans and is centrally organized by the proprietor, who provides materials, sets standards, and buys back the finished product. However, as far as antiquity is concerned it would be extremely difficult to decide whether or not a nucleated workshop industry should be considered a dispersed manufactory. First, in most cases it is impossible to gain a sufficiently deep insight into the legal organization of the production, which is essential for a well-founded decision. Second, even when we have this insight, it can still be difficult to make this decision. For instance, in the case of the Oxyrhynchus papyri only the third-century contracts meet the requirements of this definition without a doubt. In the second-century contracts the problem occurs that the lessee could possibly sell the products himself, when he had paid the rent in kind, or even paid it in money respectively. In addition, he had to lease the source of raw material. The assignation of the production to a certain category would, therefore, depend only on the respective type of locatio conductio. The three theoretically separated types any of which in reality can occur separately, combined with each other, or in mixed variants, which makes it very difficult to find a modern equivalent. Archaeologically no difference can be seen. Third, in most cases of large-scale pottery production in antiquity it is difficult to say if the workshop should be considered the home of the artisan, especially when slaves were involved. In addition, the term manufactory is normally related to the definition given above in Section III, especially to a sub-division of labour higher than in a common workshop. Taking these observations into consideration, the concept of ‘dispersed manufactory’ does not really seem to be a useful category for the classification of pottery mass production in antiquity.

94 For a recent critical reassessment of various interpretions of these inscriptions cf. Liou and Tchernia, op. cit. (n. 30), 133–56.

95 Rodríguez-Almeida, op. cit. (n. 41).

96 On the content of the stamps cf. M. H. Callender, Roman Amphorae with Index of Stamps (1965), xxvi–xxvii; Manacorda and Panella, op. cit (n. 30), 55–64; Liou and Tchernia, op. cit. (n. 30); Aubert, op. cit. (n. 21), 246–8. On the meaning of the term figlinae cf. T. Helen, Organization of Roman Brick Production in the First and Second Centuries A.D., Annales Academiae Scientiarum Fennicae, Dissertationes humanarum litterarum 5 (1975), 33–88 (meaning according to Helen: clay district); against this interpretation Steinby, op. cit. (n. 33, 1986), 99–164, esp. 156–7, and eadem, op. cit. (n. 33, 1993), 141 (meaning: brickyard (‘le figlinae non sono semplici cave di argilla…, bensì organizzazioni produttive’). Cf. also Aubert, op. cit. (n. 21), 236–8.

97 Liou and Tchernia, op. cit. (n. 30), passim; Manacorda, op. cit. (n. 37), 150–5; Peacock, D. P. S. and Williams, D. F., Amphorae and the Roman Economy— An Introductory Guide (1986), 9–10Google Scholar; Aubert, op. cit. (n. 21),247–9.

98 On these stamps and their interpretation Liou and Tchernia, op. cit. (n. 30), 147–8, with literature and a critical reassessment of other suggestions.

99 Liou and Tchernia, op. cit. (n. 30), 147.

100 cf. C.I.N. EX OF / L. LIC. MAG (CIL XV 2972, 3471).

101 Liou and Tchernia, op. cit. (n. 30), 145.

102 Rodríguez-Almeida, op. cit. (n. 41), 98.

103 On La Catria and the interpretation of the finds cf. Liou and Tchernia, op. cit. (n. 30), 145–7, referring to Remesal-Rodríguez, J., ‘La economía oleícola bética: nuevas formas de analisis’, Archivo Español de Arqueologia 50–1 (1977–1978), 87–142Google Scholar, the same article in German: ‘Die Ölwirtschaft in der Provinz Baetica: neue Formen der Analyse’, Saalburg-Jahrbuch 38 (1982), 30–71Google Scholar.

104 Rodríguez-Almeida, op. cit. (n. 41), 99.

105 This has already been pointed out by Strobel, op. cit. (n. 53), 97–100.

106 Manacorda, op. cit. (n. 37), 150–5.

107 Tassaux, F., ‘Laecanii. Recherches sur une famille sénatoriale d'Istrie’, MEFRA 94 (1982), 227–69CrossRefGoogle Scholar, esp. 254–7. Aubert, op. cit. (n. 21), 257–8, has obviously misunderstood the table of persons given by Tassaux, for he misinterprets all persons as slaves attached to the workshop near Fasana. Tassaux, however, says that some of them were more probably freedmen, and only fourteen were active at the site near Fasana.

108 Tassaux, op. cit. (n. 107), 255–6.

109 cf. Aubert, op. cit. (n. 21), 257.

110 op. cit. (n. 107), 257.

111 D. Manacorda, ‘Produzione agricola, produzione ceramica e proprietà della terra nella Calabria romana tra Repubblica e Impero’, in Epigrafia della produzione e della distribuzione. Actes de la VII eRencontre franco-italienne sur l'épigraphie du monde romain (Rome, 5 à 6 juin 1992), Collection de l'École française de Rome 193 (1994), 3–59, esp. 4. The identification of Visellius as the owner of the fundus on which the kiln site was located is valueless as long as it rests exclusively on the amphora-stamps. As it stands, Visellius could have been merely the owner or lessee of the kiln site itself, or a contractor. The same applies to attempts to identify Visellius as a member of the urban élite. As long as we do not know his cognomen we cannot be sure that he was more than a freedman, or even a descendant of a freedman of a forefather of the Visellii we know from other sources.

112 Manacorda, op. cit. (n. 111), 5. The connection with Visellius is epigraphically attested for only four persons, cf. D. Manacorda, ‘Le fornaci di Visellio a Brindisi. Primi risultati dello scavo’, Vetera Christi anorum 27 (1990), 375–415, esp. 381 n. 21.

113 Manacorda, op. cit. (n. 111), 5.

114 Manacorda, op. cit. (n. 112, 1990), 384.

115 ibid., 380–5.

116 The firing of a kiln filled with 160 amphora-like oil jars on the isle of Djerba, which were considerably larger than the amphorae produced at Giancola, took about ten to fifteen days (Peacock, op. cit. (n. 16), 42, quoting J. L. Combès and A. Louis, Les potiers de Djerba (1967)).

117 Manacorda, op. cit. (n. 112, 1990), 380.

118 cf. the plans of modern potteries Peacock, op. cit. (n. 16), 30, fig. 11, especially the plan of a pottery in Orei, Euboea, Greece (fig. 11,2), with two wheels, where the arrangement of the firing area is very similar to the installations at Giancola.

119 Peacock, op. cit. (n. 16), 41–3, quoting Combès and Louis, op. cit. (n. 116).

120 120 if the amphorae were fired standing upright; 180 if they were fired in three layers (it is likely that more layers would have led to breakages among the unfired heavy vessels). On the size of the kilns Manacorda, op. cit. (n. 112, 1990), 378 n. 11, for the size of the amphorae cf. ibid., figs 8 and 9.

121 Manacorda, op. cit. (n. 111),8–9.

122 cf. Manacorda, op. cit. (n. 112, 1990), 381 n. 21.

123 Manacorda, op. cit. (n. 111), 7.

124 CIL III 6634, 17.

125 Dressel, H., CIL XV (1891), p. 4Google Scholar, Helen, op. cit. (n. 96); Steinby, op. cit. (n. 33, 1982), 232–3; Steinby, op. cit. (n. 33, 1986), 100, 106–7, 1449–50; Steinby, op. cit. (n. 33, 1993), 139–43; Aubert, op. cit. (n. 21), 222–38.

126 Aubert, op. cit. (n. 21), 227 n. 81.

127 Steinby, op. cit. (n. 33, 1982), 232–3; Steinby, op. cit. (n. 33, 1986), 100, 106–7, 149–50; Steinby, op. cit. (n. 33, 1993), 139–43.

128 Aubert, op. cit. (n. 21), 232–5.

129 Strobel, op. cit. (n. 53), 91–114, esp. 96–7, 100–14.

130 Peacock, op. cit. (n. 16), 122–8, Aubert, op. cit. (n. 21), 208–11.

131 See Appendix, Table 1. Some of the C.ANNI stamps in CVArr 82' are stamps on decorated vessels and moulds, and were therefore probably mere parts of slave names, of which elements were integrated into the decoration on different areas of the vessel.

132 For examples see Fig. 1 and Appendix, Table 1. All stamps are collected under CVArr 83. C.AN(n)I/CHRESI(mus?) (CVArr 831), C.AN(ni)/EROS (CVArr 83r), C.ANNI/FELIX (CVArr 83s), and C.ANNI/OPILLI(o) (CVArr 83z) used the doubtful form ‘master's duo nomina in the genitive + the slave's given name’ (cf. above n. 68), if these stamps are not to be read from the bottom to the top line; this is a fundamental problem, which is discussed in detail above in Section IV.B.

133 CVArr 83b, c, d, e, f, h, k, l, m, r, t, u, w, x, y, aa + bb, hh, ii.

134 CVArr 83h and y. We have to distinguish between the number of stamps and the number of variants. Variants are stamps which bear the same name, but are not identical in form. Since each variant points towards a certain number of vessels stamped with it, the number of different variants is more meaningful for a comparison of the output than the total number of stamps. Unfortunately, we do not know how many vessels could have been stamped with one stamp.

135 Ulpianus, (29 ad ed.), Dig. 15.1.7.4.Google Scholar

136 Florentinus, (11 inst.), Dig. 15.1.39.Google Scholar

137 Ulpianus, (29 ad ed.), Dig. 15.1.7.1Google Scholar. The libera administratio was first mentioned by Proculus (7 epist.), Dig. 46.3.84, and gave the right of free disposal of the peculium to the holder.

138 (28 ad ed.), Dig. 14.3.7.1.

139 (28 ad ed.), Dig. 14.3.3.

140 (28 ad ed.), Dig. 14.3.5.1–15. Cf. for details Aubert, op. cit. (n. 21), 6–9.

141 Strobel, op. cit. (n. 53), 110–11. To reinforce the parallel between amphorae production in Oxyrhynchus and terra sigillata manufacture in La Graufesenque, he argues against Cockle, op. cit. (n. 80), 96, that the large number of amphorae made in the Oxyrhynchite potteries would indicate that not all vessels were used by the owner of the pottery himself, but were also sold to smaller neighbouring estates, who could not afford to run their own potteries (ibid., 93 and 95).

142 CIL XI 2.1, p. 1082. Cf. the map shown as Fig. 2. (The map is a redrawn and improved version of the map depicted in CIL XI 2.1, p. 1082. The locations are numbered as in the CIL, the assignment of the workshops to the locations shown on the map is that of Ihm, CIL XI 2.1, p. 1082. I have added the location of the workshop of Cn. Ateius, which makes up the only new findspot. The extension of the ancient town at the end of the first century B.C. is shown according to Maetzke, G., s.v. Arezzo, Enciclopedia dell'Arte Antica Classica e Orientale I (1958), 617–18Google Scholar, fig. 798. For the redrawing the Atlante dei siti archeologici della Toscana, Biblioteca di studi e materiali 1(1992), Tav. 114 and the Guida d'ltalia del Touring Club Italiano 114 (Toscana) (1974), city map of Arezzo p. 837, were used as correctives.)

143 cf. Fig. 2. At eight locations, more than one group was present (nos 1, 2, 4, 6, 7, 9, 11, 12), at six of these three or four groups (nos 1, 4, 6, 7, 9, 11).

144 cf. SFAS, 66 and 198. Prachner emphasizes rightly that it is very difficult to explain why vessels with ‘Firmenstempeln’ are in the majority. However, his cautiously formulated suggestion that potters had to stamp their own name on production above target only, and were allowed to offer this over-production on their own behalf when the production with ‘Firmenstempeln’ was already sold out, is debatable. On the contrary, one can much more easily imagine that slaves’ stamps could have been used to control whether or not a potter had already reached a production target.

145 A.VIBI/FIGVLI (CVArr 2324); SENTI/FIGVL (CVArr 1731);SESTI/FIGVLI/OPT(CVArr 1794); A.TITI/FIGVLI and A.TITI/FIGVL/ARRET(CVArr 2002).

146 SESTI/FIGVL/OPT (CVArr 1794); A.TITI/FIGVL/ARRET (CVArr 2002).

147 A.VIBI/FIGVLI (CVArr 2324); A.TITI/FIGVLI and A.TITI/FIGVL/ARRET (CVArr2002).

148 Aubert, op. cit. (n. 21), 295.

149 See below Section VIII.

150 A look at the stamp group tables in SFAS makes that fact perfectly clear. To give a typical example, in table No. 1 (SFAS, 9–16) 61 dependants of four Arretine stamp groups are represented by 149 different variants of stamps. The average number is 2.44 variants per person. Of 39 of them we have only one or two different variants of stamps. In Arezzo 45 out of 61 names were found, which means that more than one quarter is represented only by stamps from outside the assumed production place. See also the Appendix here.

151 Aubert, op. cit. (n. 21), 295.

152 cf. below Section VIII.

153 Aubert, op. cit. (n. 21), 220. Cf. also his argumentation against Helen (op. cit. (n. 96)) in the question of whether or not officinatores could work as potters themselves (Aubert, op. cit. (n. 21), 224–5).

154 Steinby, op. cit. (n. 33, 1993), 142.

155 On these methods and on problems occurring in their application see below Section IX.

156 Aubert, op. cit. (n. 21), 295.

157 Gummerus, op. cit. (n. 2), 1488, reckons with more than 100 simultaneous potters in P. Cornelius' ‘firm’. De Martino, op. cit. (n. 21), 339, mentions only the total number of potters of the biggest ‘firms’ and the necessity of auxiliary staff. He seems to assume that the number cited as belonging to any single pottery owner represents the number he owned at a given time. Kloft, op. cit. (n. 21), 172–3, claims the production of terra sigillata in Arretium ‘kannte bereits Betriebsgrößen, die sich der neuzeitlichen Manufaktur (60–70 Arbeiter) nähern’.

158 SFAS, 1.

159 See Appendix, Table 1.

160 For the period of production see SFAS, 19.

161 See SFAS, 21 and here Appendix, Table 2. Prachner knew of eight such persons, we can add three new finds (cf. Appendix, Tables 1 and 2).

162 See Appendix, Table 5; for the period of production see SFAS, 146.

163 SFAS, 146, for up-to-date numbers see Appendix, Table 5.

164 Published by U. Pasqui, ‘Nuove scoperte di antiche figuline della fornace di M. Perennio’, NdS 1896, 453–66. Used as an argument by e.g. H. Comfort, ‘Terra sigillata’, in T. Frank (ed.), An Economic Survey of Ancient Rome (1940), V, 188–94, 190; Kiechle, op. cit. (n. 15), 73; De Martino, op. cit. (n. 21), 339; Peacock, op. cit. (n. 16), 121; Aubert, op. cit. (n. 21), 296.

165 Aubert, op. cit. (n. 21), 296.

166 Peacock, op. cit. (n. 16), 54.

167 ibid.

168 ibid., 38–40.

169 ibid., 9.

170 Hoffmann, B. and Juranek, H., ‘Versuche zur Rekonstruktion von Terra sigillata’, Archäologie in Deutschland (1993, no. 1), 32–5, esp. 32.Google Scholar

171 ibid.

172 cf. the sketch and the description by Pasqui, op. cit. (n. 164), 455–6.

173 cf. Winter, A., Die antike Glanztonkeramik. Praktische Versuche, Keramikforschungen 3 (1978), 7–11Google Scholar; Conspectus 34, with more bibliographical references.

174 Callender, op. cit. (n. 96), 41, mentions a find of a large number of amphorae ‘full of very finely washed and levigated clay’ (referring to Dressel, H., ‘Di un grande deposito di anfore rinvenuto nel nuovo quartiere del Castro Pretorio’, BCAR 7 (1879), 143–96Google Scholar, esp. 193).

175 SFAS, 191.

176 SFAS, 245.

177 Maiuri, A., ‘Due singolari dipinti Pompeiani’, MDAI (R) 60/61 (1953/1954), 88–99Google Scholar, esp. 90–1, Taf. 31.2, explicitly excludes the possibility of regarding the picture as depicting an officina vasaria; by contrast, Rieth, A., ‘Zur Frage der Römischen Töpferscheibe’, Fundberichte aus Schwaben, n.s. 17 (1965), 153–5Google Scholar, esp. 155, interprets the picture as a pottery employing four stick-driven potters’ wheels. However, the construction of the more than 40 cm high feet of the depicted round tables looks quite unstable, so that it is doubtful if they could bear a potter's wheel heavy enough to function as a stick-driven fly-wheel.

178 cf. Section V.

179 cf. n. 143.

180 CIL XI 6702; Oxé, A., ‘Die Töpferrechnungen von der Graufesenque’, BJ 130 (1925), 38–99Google Scholar, esp. 51. Some very fragmentary graffiti from Arezzo are preserved. The best example is this vessel stamped by Rufus C. Anni, in which the names Archilaus, Epapra, Onesimus, and Ampio are engraved. With the exception of Ampio all these persons are known as slaves of C. Annius by stamps. Unfortunately, all graffiti from Arezzo are so fragmentary that the kind of co operation remains obscure.

181 V. Funghini, Degli Antichi Vasi Fittili Aretini. Estratto del Catalogo generate dell’ Esposizione di Ceramica ed arti affini, Roma 1889, 4a Edizione con molto aggiunte (1893), 18.

182 Gamurrini, G. F., ‘Nuove scoperte di antichità’, NdS 1887, 438–9.Google Scholar

183 This term means the stamping tool used to press the inverted motives or patterns into the mould.

184 Stenico, A., ‘Sulla produzione di vasi con rilievi di C. Cispius’, Athenaeum n. s. 33 (1955), 173–216Google Scholar, at 205.

185 cf. SFAS, 113. Prachner shows this by way of an example of the jointly owned slaves of Rasinius and Memmius.

186 CIL XI 2.1, p. 1082. Cf. Fig. 2 and n. 143.

187 SFAS, 191.

188 CIL XI 2.1, p. 1082.

189 cf. Ihm, M., ‘Die arretinischen Töpfereien’, BJ 102 (1898), 106–26Google Scholar. Unfortunately neither the Italian excavators, nor the Corpus Vasorum Arretinorum by Oxé and Comfort (CVArr), nor Stenico, A., Revisione critica delle pubblicazioni sulla ceramica arretina (1960)Google Scholar, give any usable information on the state of the sherds bearing stamps. Therefore, it is virtually impossible to say if they were in a good state, or production waste.

190 See Appendix, Table 3. Cf. SFAS, Tab. 6; for the period of production SFAS, 67–8. We can add four new finds to Prachner's list of sixty-five names, namely FELIX/P.CORNELI, OLVM(pus)/P.COR(neli), P.CORN(eli)/PILADI, and P.CORNE/PRIMIG(geni?); all in rectangular frames (information by courtesy of P. M. Kenrick).

191 cf. Fig. 2, locations 11 and 12.

192 CIL XI 6700. 204, 209, 220, 244.

193 See Appendix, Table 3. Cf. CIL XI 6700. 205–59.

194 See Appendix, Table 3. Cf. the table CIL XI 6700. 204; CIL XI 6700. 205, 208a, 209a + c, 213a–c, 215, 216a–c, 220a–d, 221a + b, 227, 234, 235a, 237, 238, 242, 244b–d, 245b, 247a–c, 249a + b, 56, 59b (these data are collected in the table CIL XI 6700. 204).

195 See Appendix, Table 3. Cf. the table CIL XI 6700. 204; CIL XI 6700. 207a + b, 210a + b, 212, 214, 217, 218a + b, 219, 222, 223a + b, 226a, 229, 230, 232, 233, 239, 248a+ b, 252–4. The conclusion already drawn there by Ihm was: ‘Itaque dubium non est, quin his fere locis, prope Cincelli et pontem vicinum a Buriano, officinae P. Corneli fuerint’ (my emphasis).

196 See Appendix, Table 3. Cf. the table CIL XI 6700. 204; CIL XI 6700. 228, 231, 240, 243.

197 Even if this were the case, it would not be against the existence of different sites, for P. Cornelius could have moved some of his dependants from site to site.

198 SFAS, 66. The number had to be changed due to new finds. The up-to-date state is represented in the Appendix, Table 3.

199 cf. Appendix, Table 3.

200 CIL XI 6700. 204i–z, aa–dd.

201 Three finds are reported (CIL XI 6700. 204f–h), but the reading of 204f as a stamp of P. Cornelius is in my opinion very questionable. Cf. also the report of G. F Gamurrini, ‘Nuove frammenti di vasi aretini scoperti nel sito di un’ antica fabrica presso ponte a Buriano’, NdS 1893, 138–42, esp. 41.

202 CIL XI 6700. 204a–c. Cf. Fig. 2, location 5.

203 CIL XI 6700. 259a, b.

204 SFAS, 64.

205 See Appendix, Table 4. For the period of production cf. SFAS, 112.

206 cf. Fig. 2, location 4.

207 CIL XI 2.1, p. 1082.

208 CIL XI 6700. 520–2 Cf. Fig. 2.

209 See Appendix, Table 4. Cf. CIL XI 6700. 521, 523–5, 527, 534, 537, 540, 541, 543, 547–9.

210 The discovery of fragments of forms with the master's name (CIL XI 6700. 520b; cf. also A. Stenico, La ceramica arretina I, Rasinius, Collana di testi e documenti per lo studio dell'antichità 4 (1960), nos 1, 25, 101, 114, 126, 151, 188, 213) does not count, since such fragments of moulds or moulded ware normally show only a part of a full name form of a dependant (cf. Section IV.B and the commentary on such finds in CVArr 1486).

211 CIL XI 6700. 520i–m. Cf. Fig. 2 for the location.

212 cf. CIL XI 6700. 520.

213 Lasfargues, A. and J. and Vertet, H., ‘Les estampilles sur sigillée lisse de l'atelier augustéen de la Muette à Lyon’, Figlina 1 (1976), 39–87Google Scholar.

214 cf. Fig. 2, location 1.

215 CIL XI 6700. 31–61. Cf. Fig. 2, locations 11 and 5.

216 Ihm, op. cit. (n. 189), 118. One site is noted on the map as the principal place, but with a query. Apart from sherd finds in this area, mentioned on the map as situated inter viam Guido Monaco et viam Tolleta (cf. Fig. 2, location 3), stamped sherds were unearthed near a place called Fonte Pozzolo, a long way from the principal provenance, with only three of them being master's stamps (CIL XI 6700. 696a1, f + h. Cf. Fig. 2, location 6). Moreover, some sherds were also found near S. Maria in Gradi and ad fluvium Castro, with only one of them a master's stamp (CIL XI 6700. 696a2. Cf. Fig. 2, location 5 (S. Maria in Gradi))

217 Gummerus, op. cit. (n. 2), 1485; Rostovtzeff, op. cit. (n. 4), 36; Oxé, A., Arretinische Reliefgefäße vom Rhein (1933), 8Google Scholar; Kiechle, op. cit. (n. 15), 70; Peacock, op. cit. (n. 16), 122; SFAS, 194; Aubert, op. cit. (n. 21), 296.

218 SFAS,192; Peacock, op. cit. (n. 16), 122; Aubert, op. cit. (n. 21), 296.

219 Kiechle, op. cit. (n. 15), 70; While Oxé, op. cit. (n. 217), 8, assumed that the manufacturing of moulded terra sigillata was done by at least three hands, namely of the poinçon maker, of the mould producer, and of the moulded-vessel potter, Prachner, SFAS, 194 n. 30, tries to show that e.g. the vessels' bases could have been added by a fourth hand.

220 e.g. Aubert, op. cit. (n. 21), 204: ‘It would have been economically senseless to waste the talent and expertise of a specialised mould designer by employing him in other unrelated activities’; cf. V. Klaveren, op. cit. (n. 15), 145.

221 Winter, op. cit. (n. 173), 7–11.

222 SFAS, 221.

223 SFAS, 223.

224 83k, 83r, 155, 157–60, 164, 166/167, 168, 170, 172, 176/177, 180/181, 500, 532, 1280, 1498, 1521, 1534, 1988, 2061, 2082/2084, 2086 plus one new find cf. SFAS, 29, Tab. 2; (all data according to the tables in SFAS).

225 cf. Section VII.C.

226 Hoffmann, B., Die Rolle handwerklicher Verfahren bei der Formgebung reliefverzierter Terra Sigillata, D.Phil, thesis, Munich (1983), 6Google Scholar.

227 Hoffmann, op. cit. (n. 226), 28.

228 Hoffmann, op. cit. (n. 226), 66–7, 92 and passim.

229 Klumbach, H., ‘Materialien zu P. Cornelius’, Jb. d. RGZM 22 (1975) (= Festschrift Hundt II), 47–61, esp. 48Google Scholar; Stenico, op. cit. (n. 184), 173–216, esp. 205; Stenico, op. cit. (n. 210), 22; Dragendorff, H. and Watzinger, C., Arretinische Reliefkeramik. Mit Beschreibungder Sammlung in Tübingen (1948), 162–3.Google Scholar

230 SFAS, 239.

231 This applies to Attius (CVArr 205, 212), Rasinius (CVArr 1485, 1543), C. Sentius (CVArr 1729, 1730, 1732), Thyrsus (CVArr 2062, 2068), and Cerdo C.Anni (CVArr 83), who are represented at the site in question by masters' stamps and'or dependants' stamps, cf. Lasfargues and Vertet, op. cit. (n. 213), 65–9.

232 Pucci, G., ‘Terra Sigillata Italica’ in Enciclopedia dell'Arte Antica Classica e Orientale, Atlante delle Forme Ceramiche II (1985), 365–406Google Scholar, esp. 368.

233 On scientific provenancing in general G. Schneider and B. Hoffmann, ‘Chemische Zusammensetzung italischer SigiUata’, Conspectus, 27–37.

234 On XRF and NAA respectively cf. G. Schneider, ‘X-ray fluorescence analysis and the production and distribution of terra sigillata and Firmalampen’, and J. T. Peña, ‘Two studies of the provenience of Roman pottery through neutron activation analysis’, both in Harris, op. cit. (n. 28), 129–37, 107–20 respectively.

235 Cn. Ateius is in some regards exceptional. His group includes more than twenty dependants (SFAS, 26–9), of whom only a handful is represented in Arezzo (two without doubt, two questionable, cf. SFAS, 30). In addition, the master's stamps prevail not only in Arezzo, but also in production elsewhere. Many of his dependants were clearly freedmen, who had their own slaves (SFAS, 32). The most up-to-date discussions of the Ateius-problem are found in SFAS, 30–6 and Aubert, op. cit. (n., 21), 280–7.

236 Marchini, P. Taponeco, ‘La fabbrica pisana di Ateio’, Antichità Pisane I (1974), 3–9Google Scholar. The workshop was located outside the actual town, cf. Atlante dei siti archeologici della Toscana, Biblioteca di studi e materiali I (1992), Tav. 104.

237 Picon, M.et al., ‘Recherches sur les céramique d'Ateius trouvées en Gaul’, RCRF 14/15 (1972/1973), 128–35Google Scholar;, Wiedemann, F.et al., ‘A Lyons branch of the pottery-making firm of Ateius of Arezzo’, Archaeometry 17 (1975), 45–59CrossRefGoogle Scholar.

238 Marichal, R., ‘Nouvelles fouilles et nouveaux graffites de la Graufesenque’, CRAI (1981), 244–72Google Scholar, esp. 251. This assignation is questionable because so far it is based on the analysis of a single sherd. Schneider and Hoffmann, op. cit. (n. 233), 32, mention five fragments of local products with dubious Ateius-stamps.

239 Maetzke, G., ‘Notizie sulla esplorazione dello scarico della fornace di CN. ATEIVS in Arezzo’ RCRF 2 (1959), 25–7Google Scholar.

240 Aubert, op. cit. (n. 21), 282.

241 Picon, op. cit. (n. 237).

242 Ettlinger, E., ‘Vorbemerkungen zu einer Diskussion des Ateius-Problems’, RCRF 4 (1962), 27–44Google Scholar. SFAS, 32–5.

243 Peña, op. cit (n. 234), 114–15 with n. 27; Pucci, op. cit. (n. 54, 1990), 23 with n. 13; Pucci, op. cit. (n. 54, 1992), 148–54.

244 CVArr468.

245 Pucci, op. cit. (n. 54, 1990); Pucci, op. cit. (n. 54, 1992).

246 Schneider, op. cit. (n. 234), 132; Pucci, op. cit. (n. 54, 1992), 149; A. Manneius had four or five dependants, cf. CVArr 946–53, as for Camurius it is debatable whether his name was Camurius or C. Amurius, cf. CVArr 397.

247 Schneider, op. cit. (n. 234), 130–2.

248 S. Sertorius Ocella (CVArr 1775–85), A. Titius (CVArr 1998–2003), P. Attius (CVArr 209), A. Sestius (CVArr 1792–1819), C. Sentius (CVArr 1792–1819). Zabehlicky-Scheffenegger, S., ‘Frühe padanische Filialen einiger arretinischer Töpfereien’, RCRF 29/30 (1991), 95–104Google Scholar.

249 Pucci, op. cit. (n. 54, 1990), 19.

250 On the production of clay artefacts in the context of the villa economy in general cf. Peacock, op. cit. (n. 16), 129–35.

251 Pucci, op. cit. (n. 54, 1990), 19.

252 Di Caprio, N. Cuomo, ‘Proposta di classificazione delle fornaci per ceramica e laterizi nell'area italiana dall preistoria a tutta l'epoca romana’, Sibrium 11 (1972), 371–464Google Scholar, esp. 397. As known from finds in Gaul, terra sigillata was fired in kilns in which the hot gases were conducted through clay tubes (tubuli) so that the vessels could not get into contact with exhaust fumes and fire. Cuomo Di Caprio suggests that comparable results could be reached by plugging the central holes in the floor, and placing the terra sigillata piles in the shelter of other vessels so that they were not touched by flames.

253 cf. Section V.

254 cf. on this method Peacock, op. cit. (n. 16), 65, referring to evidence from Holt reported by Grimes, W. F., ‘The works depot of the XXth Legion at Holt’, Y Cymmrodor 41 (1930), 182Google Scholar, and from Tarsus by H. Goldman, Excavations at Gözlü Kule, Tarsus (1950), respectively.

255 Pucci, op. cit. (n. 54, 1990), 23.

256 A summary of this is given by Peacock, op. cit. (n. 16), 118–19.

257 P. M. Kenrick, ‘Cn. Ateius — the inside story’, a paper presented at the international congress of the Rei Cretariae Romanae Fautores at York in September 1996; it will be published in the proceedings of the congress (RCRF).

258 cf. e.g. recently Aubert, op. cit. (n. 21), 277–84.

259 For the concept of ‘domestic’ and ‘traditional’ use of slave labour as opposed to its qualitatively innovative use cf. Rathbone, D. W., ‘The slave mode of production in Italy’ (review article of Giardina, A. and Schiavone, A. (eds), Società Romana e Produzione Schiavistica (1981)), JRS 73 (1983), 160–8, at 167, and Carandini, op. cit. (n. 20), 250–3Google Scholar.

260 SFAS, 182, referring to Gummerus, op. cit. (n. 2), 1497.