² RSOT is responsible for this year's ‘monumental’ items, most of which derive from a search of the literature while preparing RIB III (see below. Addendum (a)).

³% With the next two items during excavation by the Kent Archaeological Rescue Unit directed by B.J. Philp. Their present location is presumably the Roman Painted House, Dover. The altar is published by Frank Jenkins in Philp, B.J., The Excavation of the Roman Forts of the Classis Britannica at Dover, 1970-1977 (1981), 146–7, No. 65, with 53 and 144, fig. 30.Google Scholar

⁴ Jenkins in Philp, op. cit. (in note 3), 146. But to judge by the drawing, this reading is far from certain; nor is it even certain, in view of its informality and unusual position, that this faint incision is a literate inscription.

⁵ Philp, op. cit. (in note 3), 173, No. 266, with 172 fig. 45. ‘Marginal traces of two other letters’ were noted, but they are not visible in the published drawing.

⁶ Philp, op. cit. (in note 3), 173, No. 267, with 172 fig. 45.

⁷ During excavation by the Bath Archaeological Trust directed by Prof. Barry Cunliffe, and published as B. Cunliffe and P. Davenport, The Temple of Sulis Minerva at Bath, I (1) The Site (1985), 132, No. Mi3, pl. LXV. (The transmitted scale of 1:10 is evidently an error for 1:5.) The fragment is now in the Roman Baths Museum, Bath, ace. no. 1983.1.231, but has not been located.

⁸ Like the next item during excavations noted by Crummy. They are published as Nina Crummy, Colchester Archaeological Report 2: The Roman Small Finds from Excavations in Colchester 1971-9 (Colchester Archaeological Trust, 1983), 145, Nos 4285 and 4286 respectively, with 146 Fig. 179.

⁹ During excavations directed by R.W. Bartlett, and published by him as Excavations at the Prehistoric and Roman Temple Site at Harlow, Essex 1985-89, forthcoming. The fragments are fully described by I.K. Jones (ibid.), and are now in Harlow Museum (HMB 12240-42), from where Mr Bartlett sent a photograph.

¹⁰ During excavation by the Great Chesterford Archaeological Group directed by Collins, A.E. (Britannia 10 (1979), 309–11).Google Scholar See Miller, T.E., ‘The Romano-British temple precinct at Great Chesterford, Essex’, Proc. Camb. Antiquarian Society 84 (1995), 15–57Google Scholar, which notes (a) at 49 with 47, fig. 22, no. 204. Both letters are now in Saffron Walden Museum, Essex, inv. no. 1989.2/386 and 1986.20/28. Jean Bagnall Smith sent information and a photograph.

¹¹ During excavation directed by Charles Green for the Ministry of Works. See Gurney, D., ‘Leylands Farm, Hockwold cum Wilton: excavations by Charles Green, 1957’, in Gurney, D. (ed.), Settlement, Religion and Industry on the Roman Fen-Edge, Norfolk (East Anglian Archaeology 31, Norfolk Archaeological Unit, Norfolk Museums Service, 1986), 49–92Google Scholar, especially 69, fig. 43, and 70, No. 54. It is now in the Castle Museum, Norwich. Gurney also notes (ibid., 70, No. 55) a fragment of curved sheet metal, ‘possibly part of an O’. The latter is not illustrated, and at the moment is not available.

¹² By metal detector. (Information from Jean Bagnall Smith.) It is now in the Ashmolean Museum, Oxford (inv. no. 1998.40), which sent a photograph.

¹³ By a field-walker, who gave it to the Ashmolean Museum, Oxford (ace. no. 1963.1653). See Oxoniensia 28 (1963). 89.Google Scholar Information from Jean Bagnall Smith, who sent a photograph.

¹⁴ It was apparently not transferred from the temporary site museum to Newport Museum in 1916-17. See Boon, G.C., Monmouthshire Antiquary 3 (1970-1978), 209 with 210, pl. 1.1.Google Scholar

¹⁵ Only a photograph survives, published by Boon (see previous note). It is not clear whether either margin is original. In line 2 Boon reads [.]ANA, but his AN looks more like a damaged N.

¹⁶ In excavation on the site of a bungalow extension, and deposited in the Roman Legionary Museum, Caerleon, accession no. 94.47H. Julie Reynolds sent a rubbing and photograph.

¹⁷ By Prof. Philip Rahtz, who provided information and a photograph.

¹⁸ There is a natural fissure to the right of the inscription, the edge of which has eroded, so that 1-2 letters may been lost after CRVTC, but on the whole this seems unlikely. The inscription is probably Roman, but doubt remains. Its find-spot adjoins a Roman road (Margary 72b), and its letters look Roman, but not necessarily so. They are much more deeply-cut and long-weathered than those of some names which have been cut further along the rock-face, which are comparatively modern. VS is appropriate as a Roman name-ending, but CRVTC is difficult; perhaps CRIITICVS was intended (i.e. Creticus, RIB II.7, 2501.148 and Britannia 31 (2000), 444, No. 48)Google Scholar, II being cut as V, and I omitted by mistake. Otherwise compare Crotus (RIB 620; 1525 and 1532) and Chrauttius (Tab. Vindol. II 310.1, with note ad loc).

¹⁹ With the next item during excavation by Carlisle Archaeology Ltd for Carlisle City Council's Gateway City Millennium Project (see p. 337 above). Philip Cracknell sent drawings, photographs and full details.

²⁰ Dedications to Aesculapius, when provenanced, seem to be associated with the praetorium or principia: compare RIB 445, 461, JRS 59 (1969), 235Google Scholar, No. 3 (all Chester), and perhaps RIB 1052 (South Shields).

²¹ On the south wall of the principia in a corresponding position, on the second stone from the corner of the building, in the fourth course, there is a phallus carved in relief.

²² The inscription is complete, with no trace of V or VIC for the legion's cognomen, let alone P F for its titles p(ia)flidelis), whether on this stone or the adjacent stone. This is unusual, but for other instances see RIB 1061, 2162, and probably 1038.

There may be a medial point between LEG and VI, but this triangular depression, now partially filled with mortar like the letters LE, is more likely to be the result of dressing the stone before it was inscribed.

²³ During excavations for Tyne and Wear Museums and Earthwatch directed by Paul Bidwell, Nick Hodgson and Grahame Stobbs; for an interim report, see Hodgson, N. in The Arbeia Journal 6–7 (1997-1998), 25–36.Google Scholar The stone (IM97) is now in Arbeia Roman Fort and Museum, from where Ray McBride sent a photograph and drawing.

²⁴ Barrack block II in the eastern quadrant, Period 6B; see The Arbeia Journal 6–7 (1997-1998), 33, fig. 6.Google Scholar

²⁵ The cognomen is common, but has only once been found in Britain (RIB 1039).

²⁶ According to Brian Carter, antiquities dealer of Stroud, by whom it was sold to the Ashmolean Museum, Oxford. The plaque is well-patinated but looks ‘brassy’, like the orichalcum (zinc-copper alloy) of Roman sestertii. Michael Vickers made it available.

²⁷ In 2, L was repeated in error for I. In 3 there is a single redundant punch-mark between MAR and VS, but this does not warrant reading MARCVS. The cognomen Marus is rare, and Kajanto (Cognomina, 176) follows Schulze in seeing it as derived from the Oscan praenomen Marus (cf. CIL x 6555 (Velitrae) = ILS 3697 = Devijver PME s.v. O 15, a tribune of Leg. XIIII Gemina Victrix (sic) perhaps while it was still in Britain). The distinctive cognomen apart, the dedicator's rank makes an Italian origin likely, although it should be noted that -marus is a very common Celtic name-element and that Marus is occasionally found as a cognomen in CIL xiii.

The plaque is genuine, but three anomalies should be noted. (1) the dedicator is a legionary tribune, and thus of equestrian rank, but identifies himself only by his cognomen. His cognomen was distinctive, however, and he may have been pressed for space. (2) Legionary tribunes, since they did not as a rule command detachments, tend to make their dedications at the legion's headquarters; but perhaps the Twentieth Legion was on active service at the time. (3) It is very unusual for the legion's cognomina to be omitted after they had been conferred in A.D. 61 or 83 (see Britannia 23 (1992), 154–8)Google Scholar, and it does not look as if they were omitted here because of pressure of space. In view of (2) and (3), therefore, the transmitted provenance is open to doubt. It is conceivable that the plaque was actually found at one of the legion's early bases, for example Colchester, where four copper-alloy ansate plaques have been found (RIB 191, 194. 195 and RIB II, 2432.8), or even at Cologne, where it was based from A.D. 9 to 43.

²⁸ By Robin Birley, who made it available at Vindolanda Museum.

²⁹ There may be a medial point after VI, but too much has been lost to know if there was one after S. The text concluded with the age at death, either [vixit… anno]s (or anni]s) VI or [vixit… die]s (or diebu]s) VI. The latter, an age expressed in years, months and days, seems more likely than an age of ‘6 years’ only, since the ages of young children are quite often recorded to the nearest day.

³⁰ It was brought for identification to the Ashmolean Museum, Oxford, where Michael Vickers made it available. In colour and style it resembles No. 18 (Benwell area), but we do not know if they share the same provenance.

³¹ In line 2, O is half-size, presumably to fit more closely between T and V; also in line 2, the second T has no cross-bar and thus resembles 1, while the third T has a redundant dot either side of its base; also in line 2, 1 has three redundant dots to the right of its base and thus resembles L. (Compare the second L, a mistake for 1, in line 2 of No. 18 (Benwell area).) These would seem to be copying mistakes by a semi-literate craftsman. In line 3, S is of exaggerated height, and the medial point to indicate abbreviation is level with its foot. In line 4 the leaf-stop, although it balances that in line 1, is redundant since it is preceded by a medial point, and perhaps it was a mistake for M, m(erito).

³² For a (silver) ansate panel dedicated to Mars Toutatis, see RIB 219; and for copper-alloy ansate panels to Mars in other aspects, see RIB 187, 191, 305 and (in gold) 1077.

The personal name Vinoma (whether male or female) is unrecorded, and is difficult to understand as Celtic, unless it is a blundered Vindo- name (but see further Rivet and Smith, Place-Names of Roman Britain, s.v. Vinovia). However, it is surely connected with VINOMATHIVS in RIB 1528 (Carrawburgh), which has been read as Vinomathus, the name of the dedicator of an altar to Coventina. This reduces Coventinae to COVEN, and separates VINOMA from TH with a medial point, prompting a suspicion that the stone-cutter misunderstood the text he was copying; perhaps his TH (or TE, to judge by the photograph in CSIR 1.6, pl. 40, No. 146) really belonged to Coventinae, and his VS anticipated VSLM. With the discovery of this ansate panel, it now seems possible that the text of RIB 1528 contains copying-errors, and that the dedicator's name was actually Vinoma. It is even possible that the two dedicators were the same person, but this possibility could be evaluated better if only the panel's exact provenance were known.

³³ During field-walking by W.F. Cormack in the Churchyard Holm field c. 45 m east of the graveyard, and passed by him to the (1991) excavator of Hoddom, Dr Chris Lowe. Mr Cormack sent information and a photograph. Other stone from Birrens was re-used at Hoddom: see Keppie, L., ‘Roman inscriptions and sculpture from Birrens’, TDGNHAS 69 (1994), 35–51Google Scholar; and for this fragment, 49, with fig. 10.

³⁴ Line 1 is probably […V]IR; the possibilities include [Qu]ir(ina tribu), [sev]ir (etc.), and personal names like [Senov]ir. The letter in line 2 is either P or R, and may be followed by a space.

³⁵ During excavations directed by P. Halkon, M. Millett and J. Taylor for the University of Durham. See Britannia 28 (1997), 417–19Google Scholar, and 29 (1998), 387. Prof. Millett provided information, and Louise Revell made the fragment available.

³⁶ The cognomen is very common. The bath-house is of third/fourth-century date, and the brick is thought to be contemporary with it, but the lettering would suit an earlier date.

³⁷ With the next ten items during excavations by Essex County Council Field Archaeology Unit for English Heritage directed by Mark Atkinson. For the site see Atkinson, M. and Preston, S.J., ‘The Late Iron Age and Roman settlement at Elms Farm, Heybridge, Essex, excavations 1993-5: an interim report’, Britannia 29 (1998), 85–110.CrossRefGoogle Scholar Edward Biddulph and Joyce Compton of the Unit provided information and drawings (by Stewart MacNeil and Ian Bell). Since some of the sherds may be earlier than A.D. 43 (as indicated in the descriptions above), we have waived our usual criterion for inclusion of ‘three clear letters’. The graffiti were all made after firing.

³⁸ The final letter resembles ∑.

³⁹ Perhaps a name incorporating the Celtic element Att-, for which see Alföldy, G., ‘Epigraphisches aus dem Rheinland II’, Epigraphische Studien 4 (1967), 10–16.Google Scholar They often occur in the Cologne-Bonn region.

⁴⁰ Either a letter or a numeral (‘5’ or more); or, if inverted, an open A.

⁴¹ Conceivably for feci, ‘I made (this)’, but the reading is very uncertain.

⁴² […]ARM I I[…] is an alternative reading.

⁴³ The top of the first stroke of the V extends in a flourish around the foot (or neck) of the vessel.

⁴⁴ ‘4 (or more)’; unless inverted, […]AI[…].

⁴⁵ ‘6 (or more)’; unless inverted, IA[…].

⁴⁶ During an evaluation by the Cotswold Archaeological Trust conducted by Tim Harvard. Information from the Director of the Trust, Neil Holbrook, and from Emma Harrison, who was then its Finds and Archives Officer.

⁴⁷ The letters measure c. 52 by 20 mm. For the tile-stamp LHS, see RIB II.5, 2489.21, but there seems to be no direct match.

⁴⁸ During excavations funded by HBG Construction (Midlands) Ltd, before the building of a new magistrates’ court, directed by Darren Vyce of Archaeological Excavations Ltd. The stamp (Silvanus ii of Lezoux, Die 3c) is dated by Brenda Dickinson to c. A.D. 150. Mr Vyce made the sherd available to RSOT.

⁴⁹ With the next two items during field walking in the course of a metal detector survey by Mr Ian Pinder, in whose possession they remain. Information from Christopher Lydamore of Harlow Museum, who made them available.

⁵⁰ […]CII in line 3 may be a numeral, ‘102 (or more)’, and thus a record of the tile-maker's output.

⁵¹ By a metal detectorist, from whom it was bought by Mr Alan Harrison, Winterton, who sent details to Donald Mackreth for his index of Romano-British brooches. Mr Mackreth provided information and a drawing, and discussed the significance of the place-name with us.

⁵² This moulded inscription is identical to that of RIB II.3, 2421.44 (unknown provenance) except for the enamel background, which was red. See below, Corrigendum (e).

⁵³ This is first regio in Roman Britain to be named explicitly, but three others can be deduced from RIB 152 (Bath, (centurio) reg(ionarius)), RIB 587 (Ribchester, praep(ositus) n(umeri) et regi[onis]), and Tab. Vindol. II 250.9-10 (Carlisle, (centurioni) regionario Luguvalio) = Tab. Vindol. I 22.9-10 (with commentary). Lagitium, the place-name implied here, can reasonably be identified as Castleford, where in the Roman period there was ‘an extensive manufactory including enamel on copper alloy’: see Bayley and Budd in Cool, H.E.M. and Philo, C., Roman Castleford, Excavations 1974-1985, I (1988), 195–224Google Scholar (a reference we owe to Donald Mackreth), Hitherto the transmitted place-name has been either Lagetium (from the Antonine Itinerary's LAGECIO) or Lagentium (Ravenna Cosmography); see Rivet, and Smith, , The Place-Names of Roman Britain, 383Google Scholar , s.v. Lagentium. It now seems that Lagitium was the original form, the stressed short i in the second syllable becoming e by the third century; examples of this process are collected by Smith, in Temporini, H. and Haase, W. (eds), Aufstieg und Niedergang der romischen Welt 11.29, 902.Google Scholar

⁵⁴ By metal-detector, and noted with a photograph in Treasure Hunting, March 2001, 45-6 with fig. 7. Martin Henig informed RSOT.

⁵⁵ For other examples of TOT rings, see RIB 11.3, 2422.36-40; Henig, M. and Ogden, J., Antiq. J. 67 (1987), 366–7.Google Scholar The ring somewhat resembles RIB 11.3, 2422.36 (Lincoln).

⁵⁶ During excavation in advance of the new Wolfson Wing of Guy's Hospital directed by Christopher Pickard of Pre-Construct Archaeology Ltd under contract to Watkins Gray International. Information from Mr Pickard and Jenny Hall, Roman Curator at the Museum of London, who together published a preliminary note in the news sheet of the Museum of London, Archaeology Matters No. 12 (December 2000). Jenny Hall made the die available to RSOT.

⁵⁷ The total number of letters on each pair of opposing faces is seven, like the numerical ‘pips’ on ancient and modern dice: thus P is opposite ITALIA, VA opposite VRBIS, and EST opposite ORTI. An almost identical die was found at Autun in 1836 (CIL xiii 10035.24): I |VA|EST|ORTI|CAIVS|VOLO TE. It has been published with two more dice of identical type, also from Autun, by Alain Rebourg in Autun Augustodunum, Capitate des Eduens: ouvrage réalisé à partir de l'exposition qui s'est tenue à l'Hotel De Ville d'Autun du 16 mars au 27 octobre 1985 (Autun 1987), 100-101, No. 207 (a reference we owe to Hilary Cool). The others read: I|VA|EST|ORTI|VRBIS|ITALIA, and + |VA|EST|ORTI|VRBIS|IITALIA. Rebourg mentions a fourth example found at Lyon, but gives no reference or transcription. He does not cite a fifth example without provenance, now in the Hungarian National Museum, Budapest, which was published by Tóth, E. in Folia Archaeologia 32 (1981), 154, No. 45, with 152, fig. 6. This reads: + |VA|EST|ORTI|VALLA|VIANNA. These dice all share the sequence VA|EST|ORTI, but the nature of the game has not been elucidated.Google Scholar

⁵⁸ During excavations for Bloor Homes (Newbury) Ltd by Wessex Archaeology directed by Rachael Seager Smith who, with Andrew Fitzpatrick the project manager, provided details and a drawing by S.E. James.

⁵⁹ By Richard Grasby. in the course of a detailed re-examination of the tomb and its inscription which he is preparing for publication.

⁶⁰ The only instance seems to be from Gloucester, Britannia 29 (1998), 434Google Scholar , No. 2 with pl. XXXIII (above the first V in LVS[I]VS), but others may have gone unrecorded. For its use in the Vindolanda Tablets, see Adams, J.N. in JRS 85 (1995), 97–8.Google Scholar

⁶¹ Acc. no. 1963.1652. Information from Jean Bagnall Smith.

⁶² By Prof. Roger Wilson, who provided information directly and via Lindsay Allason-Jones and David Sherlock. David Sherlock also sent photographs by Peter Hill.

⁶³ The surface is pitted, but none of the marks is a medial point. The letters are in any case too close together, and S.E.V is almost impossible as the initial letters of tria nomina.

⁶⁴ Donald Mackreth tells us that he drew it from a photograph without access to the original, and that the photograph must have been tacitly enlarged for the sake of clarity. The two brooches are almost certainly the same size. For expansion and elucidation of the inscription, see above, No. 39 (Alford area).

⁶⁵ It was sold by Garry Edwards, antiquities dealer of Newark, to an Australian collector, Rob Kirk, who showed it to RSOT. The combination of the formula D(is) M(anibus) S(acrum) with the formula h(ic) s(ita) est, and the formula p(ia)/p(ius) v(ixit), are both typically African. See (passim) Lassère, J.-M., ‘Recherches sur la chronologie des épitaphes païennes de l'Africa’, Antiquités Africaines 7 (1973), 7–151.CrossRefGoogle Scholar The name Donatus/a, although common elsewhere, is typically African: Kajanto (Cognomina, 298) notes 368 instances out of 832.

⁶⁶ The deceased were evidently mother and son, in view of their relative ages, the age-rounding of the woman's age compared with the precision of the boy's age, the derivation of his cognomen from her nomen, and their association both on the same stone and apparently by the same hand.

⁶⁷ With the next item in unknown circumstances; to judge by the weathering and discoloration, they were in a garden. The marble looks like Carrara Marble; this, and the texts of both, suggest that they are modern imports from Italy (‘Grand Tour souvenirs’), most likely from Rome or its vicinity. They are now stored by the City of Edinburgh Council Archaeology Service, where John Lawson made them available.

⁶⁸ The dedicator's name is not very distinctive, but he may be the same as the dedicator of CIL vi. 13890 (Rome): Dis Manibus | C. Caelius Epaphroditus | sibi et | Calpurniae | Prepusae | uxori suae.

⁶⁹ DEMETRVS is a variant of Demetrius, if not an error. SEPETESVCV in Thrace seems to be otherwise unattested, but -sucu is a Thracian place-name element (see Detschew, D.. Die thrakischen Sprachreste (1957), 469Google Scholar , s.v.), and for the first element compare Sipte (ibid., 448, citing Pausanias 5, 27, 12). THRAAEC is quite badly damaged, but the reading is acceptable; the double A is not M, since both A’ s have cross-bars and do not resemble M in 2. Thra<a>ec(ius), usually Thraex, is a variant of Thracius commonly applied to the ‘Thracian’ gladiator (so called from his equipment), but here its significance is only ethnic. The unnamed PROCVRATOR may have been the agent of the deceased in his lifetime, the usual meaning of the term, but here it is likely that he was appointed under his will; for an explicit example see Arangio-Ruiz, V., Fontes luris Romani Anteiustiniani, III (1969)Google Scholar, No. 47, where the testator, a soldier, appoints a comrade his procurator to collect his ‘property at the fort’ and to give it to his widow in trust for his son and heir. Compare also ILS 2354, her(es) ips(ius) sive proc(urator).

⁷⁰ Information from Garry Edwards (see above, note 65), who sold it to Rob Kirk. Mr Kirk showed it to RSOT.

⁷¹ The altar resembles other crude altars from Roman Britain, and its style, letter-forms and surface patination are not obviously modern. It is possibly genuine, therefore, but it may be impugned for the following reasons: (1) the somewhat vague provenance; (2) the ease with which such an altar could have been faked (and other forgeries have come from the Bath area); (3) the anomalous dedication to genius as if the name of a god, without specifying what it was the ‘genius’ of (compare RIB 119 and 646, the only dedications from Britain deo genio); (4) the anomalous name of the dedicator, apparently a praenomen and a (very rare) nomen, without any cognomen; (5) the confusion between the termination of the nomen and the VSLM formula. This last might be a genuine mistake, and it might also be argued that VIRIDI is genitive; but there is no space between it and VS, whereas Q is separated by a space from VIRIDIVS, and a dedication to the ‘genius’ of a named individual would itself be anomalous