² During excavation directed jointly by Hedley Swain and Harvey Sheldon for the Museum of London, Birkbeck College (University of London) and the Channel 4 ‘Time Team’. For a stamped tile also found, see below. No. 39. Fragments of three other inscriptions were found there in 1902: see RIB 37 (sandstone), 38 and 39 (both marble).

³ Line 1 may well be a dedication [NV]MIN[I AVG] or [NV]MIN[IBVS AVGG], as Guy de la Bédoyère has suggested to us, the imperial divinity/ies being associated with the deity to whom the putative temple was dedicated. (Note that the first surviving line of RIB 38, another thin marble slab from the site, reads NV[…].) Lines 2 and 3 most likely contain a nomen and cognomen in the nominative case, for example Rupilius, Popilius, Turpilius (etc.) and Fuscus, Priscus (etc.), presumably the name of the dedicator. In line 3 there is a serif before the letter C whose form would be consistent with the top terminal of the letter S.

⁴ Identified by Alan Holiday, Head of Geology at Weymouth College, as a shelly Upper Jurassic limestone, probably Portland Limestone rather than Purbeck, and perhaps from the Lulworth / Worbarrow area just south of the find-spot.

⁵ By Mark Vine of Weymouth, in whose possession it remains. He consulted G uy de la Bédoyère, at whose suggestion he sent photographs and full details to RSOT.

⁶ JRS 49 ( 1959), 136, No. 6, Discipulinae Augusti.

⁷ A dedication to Jupiter headed IO M is just possible, but most unlikely. The slab would have been part of a mausoleum, not a free-standing monument in itself. It is only the third Roman tombstone from Dorset (the others being RIB 188, Dorchester, and Britannia 12 (1981), 369, No. 4, Tarrant Hinton). This is surprising, in view of the abundant supplies of excellent freestone, but the assurance of the drawing, and the competence of the cutting, both attest an experienced hand. The rarity of inscriptions from Dorset may thus be an accident of survival rather than alack of ‘epigraphic habit’.

⁸ The width cannot be calculated with certainty, but if D M was centred between two upper hederae linked by a single loop, and not by a meandering line, the surviving width would be about two-fifths of the original. Line 2 evidently ended with a cognomen in -undus. Secundus is the most common (much more common than Iucundus), but Verecundus is also common and, as a ‘Roman’ name containing a popular Celtic name-element, is a more attractive possibility in the Roman ised civil zone of a Celtic-speaking province; it would also fit better into a panel 0.50 m wide. Although it is just possible that the deceased was described as duoviralis (an ex-magistrate, presumably of Dorchester), it is probable that line 3 ended with another cognomen. The only possibility seems to be Liberalis, assuming that E was written informally as II. (This would be very unusual in a monumental text, but the lettering, though quite elegant, is brush-drawn and somewhat informal.) [et Libe]ralis would fit the same line-length as [Verec]undus. This restoration, which is conjectural, would require that two persons, both male, were being commemorated; perhaps they were the small sons of a villa-owner.

⁹ During excavation by the Channel 4 ‘Time Team’ directed by Neil Holbrook, Cotswold Archaeological Trust, who made the stone available to RSOT. For the site see Britannia 30 (1999), 365–6.

¹⁰ There is no trace of any letter to the right of L. Unless the letters were irregularly spaced, therefore, it looks as if FIL is complete. As an abbreviation of films (‘son’) or filia (‘daughter’), in the nominative or an oblique case, it is very common. Below it are two lines intersecting at right-angles, apparently not a letter.

If the stone is part of a larger inscription, it was probably a tombstone or a milestone. Epitaphs occasionally end with a formula incorporating FIL (e.g. RIB 396), and there is a group of milestones (e.g. RIB 2267) erected to Constantine as Caesar (A.D. 306–7) and son of Constantius, … divi Constantipii Augfusti) filio. (The find-spot is c. 3.5 km from the Fosse Way.) But this stone does not look like a fragment. The placing of FIL within equal margins, above a scratched panel(?) and below a carefully tooled top edge, suggests that the inscription is complete. Milestones have been found at villas (e.g. RIB 2220), but it was hardly worth breaking one up and carting the pieces some kilometres to Turkdean, when building stone was available on the spot. In any case, the said group of Constantinian milestone inscriptions all terminate with FILIO (unabbreviated). The Turkdean stone looks more like a grave-marker now broken at the bottom, which belonged perhaps to a funerary group re-used as building material. Conjecturally it was subordinate to a larger stone (also inscribed) in a burial plot, and commemorated a ‘son’ or ‘daughter’. Since a child was not named until it was eight or nine days old, this pair of stones might have commemorated a mother who died in child-birth and her still-born child.

¹¹ During the building of the clubhouse of Maryport Golf Club, when more of the slab was found but not reported at the time; it was broken up, and only these pieces were preserved by one of the workmen. They are now in the Senhouse Roman Museum, Maryport, where Ian Caruana provided photographs and full details.

¹² The most likely name is Maximus, but there are other possibilities. More than three letters have been lost before IMI, judging by the likely restoration of line 3, so that the cognomen [MAX]IMI would have been preceded by a nomen, perhaps abbreviated like AVR.

¹³ In a trench 0.50 m wide excavated to a depth of 1.0 m, which was subject to a watching-brief by I. Walker, Tyne and Wear Museums, commissioned by J. Charlton on behalf of Northumbrian Water. Dr N. Hodgson provided photographs, a squeeze and full details. The stone is now in the offices of Tyne and Wear Museums at Jesmond Cemetery Gates, New castle upon Tyne, but will go to the Museum of Antiquities, Newcastle upon Tyne.

¹⁴ N is certain; it is not AN ligatured, let alone M. The same centurion is attested further west at Black Carts (Wall Mile 29) by RIB 1515, (centuria) Helleni. The editors understand the name as the rare nomen Hellenius (Schulze, 173), not as a possible variant of Helenus, which was mostly borne by slaves and freedmen.

¹⁵ In excavation by the Vindolanda Trust directed by Robin Birley, who made it available at Vindolanda Museum.

¹⁶ The numeral was incised as an aid to assembling the arch. For other numbered voussoirs from Vindolanda, see RIB 1720 and Britannia 3 (1972), 354, No. 14.

¹⁷ In consolidation by the Vindolanda Trust directed by Robin Birley, who made it available. The fragments are now built into a wall-display in Vindolanda Museum.

¹⁸ During excavation by the Vindolanda Trust, ditrected by Robin Birley, who made it available at Vindolanda Museum.

¹⁹ Compare the Vindolanda stone illustrated in Britannia 24 ( 1993), 315, fig. 3, and a stone from Birdoswald (Britannia 23(1992), 316, fig. 5).

²⁰ By Robin Birley, who made it available at Vindolanda Museum.

²¹ Presumably the end of the dedicator's name, followed by a dedicatory formula like s(acrum) fec(it) or similar.

²² By Robin Birley, who made it available at Vindolanda Museum.

²³ There is a space between X and II, but presumably they belong to a single incomplete numeral. Conjecturally it is [C]XII or even [P C]XII, [p(edes) C]XII, ‘112 ft’; compare RIB 1818 and 1820, two elaborate inscriptions which record the constructions of ‘112 feet of rampart’ at Carvoran.

²⁴ By Robin Birley, who made it available at Vindolanda Museum.

²⁵ While digging a sewer trench. See Taylor, A., ‘A Roman lead coffin with pipeclay figurines from Arrington, Cambridgeshire’, Britannia 24 (1993), 191–225.CrossRefGoogle Scholar The figurine is published by Miranda Green at 194–6, with fig. 3(i). Alison Taylor sent a sketch of the graffito, and provided information about this and the next two items.

²⁶ See Alexander, J.A. and Pullinger, J., Roman Cambridge [Proceedings of the Cambridge Antiquarian Society 87 (1999)]Google Scholar, Appendix 1,89 and pl. XVI, No. 145.

²⁷ This might be aseries of numerals (either HI, VIII, X, or III, VII, IX), perhaps cut at different times and replacing each other. If it were read inverted, it would be XIIILIII. For similar bone roundels see RIB II.3, 2440.

²⁸ See Alexander and Pullinger, Roman Cambridge, op. cit. (note 26), pl. XVI, No. 144.

²⁹ Birmingham University Field Archaeology Unit sent this and the next two items to RSOT. They will appear, with three minor graffiti not included here, in the final report (in preparation) on Godmanchester, London Road and The Parks.

³⁰ There are many possible names, for example Secundus, Severus, Sextus, and the cognate names Seneca I Senecio I Senecianus.

³¹ The first surviving letter might be F with bottom serif, but this is unlikely. Presumably an abbreviated name, the most likely being [S]EV for [S]ev(erus).

³² Bidwell, P., Snape, M., Croom, A., Hardknott Roman Fort, Cumbria (1999), 121Google Scholar, No. 1, with fig. 56.2.

³³ Presumably an abbreviated name. The angular ‘B’ might possibly be S; an abbreviated name Sat(…) would be more likely than Bat(…).

³⁴ During small-scale excavations by the Colchester Archaeological Trust. Information from Stephen Benfield of the Trust, who made the sherd available.

³⁵ This is the last line of an inscription like No. 32 (London) below, recording the origin and quality of the preserved fish-product COD or CORD, its age, the capacity of the amphora, and concluding with the names of the producers. Two such inscriptions of Proculus and Urbicus have already been found at Pompeii, CIL iv 9368 (with facsimile of a similar, if not the same hand) and 9369. See further, n. 56 below. RSOT.

³⁶ During excavations by Mrs Maureen Bennell in advance of the construction of a pipe line from Ryehill Reservoir to Fairfield Hospital. Information from Mr T.S. Martin of the Planning Department of Essex County Council.

³⁷ In excavation by Essex County Council Archaeology Unit directed by M. Atkinson, who provided information and a drawing by Stewart McNeil. 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

³⁸ Dendro dating of other planks has supplied a felling date of A.D. 149–197. The pottery found in association with the well favours the earlier part of this range for the date of construction.

³⁹ During a watching-brief by Emma Harrison of the Cotswold Archaeological Trust. Information from Emma Harrison and Neil Holbrook of the Trust, who also supplied a copy of a drawing by Richard Morton.

⁴⁰ The stamp is RIB 11.5, 2489.4A. Three other examples were found when the building (Insula X, 4) was excavated in 1922: ibid, xvii-xix. See Darvill, T.C., ‘The ARVERI and TPLF stamped Roman ceramic tiles in the Cotswolds and Severn Valley’, Transactions of the Bristol and Gloucestershire Archaeological Society 100 (1982), 47–63.Google Scholar

⁴¹ In excavation for Gloucester City Museum, where it now is. See Darvill, T.C., ‘Excavations on the site of the early Norman Castle at Gloucester, 1983–84’, Medieval Archaeology 32 (1988) 1–49CrossRefGoogle Scholar, at 28–9, fig. 14 (No. 169 of the SmallFinds by P. Isaac, who reads MSO instead of NASO). Also found were 70 fragments of tile stamped RPG (cf. RIB II.5, 2486), 30 of RPG with duoviri (cf. RIB II.5, 2487) including 13 of type 2487.6, ten of RPG with quinquennales (cf. RIB II.5, 2488), and 86 of unclassifiable RPG. Sheppard Frere sent the reference to RSOT.

⁴² For the type of legend see RIB 11.5,2487, but this example is unique. The decurion's name NASO might be complete and nominative, but we have expanded it conventionally as a genitive.

⁴³ Darvill (see note 41 ), No. 295, fig. 14 (which should probably be inverted).

⁴⁴ During excavations for Transco by Hertfordshire Archaeological Trust directed by Claire Halpin, in advance of a new gas pipe line. Nina Crummy informed us of the discovery initially, and provided a draft of her report on the object in the Roman Finds Group News Letter 20 (2000), forthcoming.

⁴⁵ By Ralph Jackson of the Department of Prehistoric and Romano-British Antiquities at the British Museum.

⁴⁶ The letters TAN (or possibly THAN) are ligatured. The stamp was recognized by Michel Feugère. About 20 instruments stamped by Agathangelus are now known: see Crummy, Nina, ‘Agathangelus, bronzeworker’, Rescue News 81 (2000), 6.Google Scholar

⁴⁷ In excavation of a medieval site by the Canterbury Archaeological Trust. Louise Harrison's report of residual Roman brick and tile in the 21st Annual Report (1996–1997) of the Trust, 38–40, was sent by Sheppard Frere to RSOT.

⁴⁸ The impression was damaged when the unfired tile was combed for keying, but the stamp is probably RIB II.5, 2481.82.

⁴⁹ The stamp is probably RIB II.5, 2481.43.

⁵⁰ In excavation by the Museum of London Archaeological Service directed by Dick Bluer and Robin Nielsen. Fiona Seeley of the Museum of London Special Services section provided information on this and the next three items.

⁵¹ A third possibility is that the letters are the end of the name [FEL]IX, or of the slogan [VTERE FEL]IX.

⁵² The end is possibly complete. Conceivably it is part of a date: [a(nle)] d(iem) IHNon[as…], ‘On the third day before the Nones of…’

⁵³ This might suggest a funerary formula dated by (a), but the vessel is unlikely to have been a funerary urn since the find-spot lies within the circuit of the city walls and the cemeteries will all have been extramural. Memor and Sacer may therefore be cognomina, as slogans wishing good fortune to (named) individuals are occasionally found.

⁵⁴ The last two letters are very faint, but this personal name is common.

⁵⁵ During excavations by the Museum of London Archaeological Service directed by Peter Rowsome. It is now in the Museum of London, where Jenny Hall made it available to RSOT.

⁵⁶ The text can be read and understood by comparing it with the painted inscriptions on other amphoras of this type: those from Colchester (No. 19, above), Lincoln (RIB II.6, 2492.39, misunderstood and now lost) and Chester (RIB 11.6, 2492.11) are fragmentary, but there is a good selection in CIL iv 5634–52. The last four lines survive of a five-line text. (1) The letters SVMAVR, sometimes difficult to read exactly, and of uncertain meaning, regularly qualify the best COD or CORD, which has been convincingly explained (see Bernard Liou in Archaeonautica 7 (1978), 68) as a preserved fish-product made from young tunny (cordula). This industry was located on the coast of southern Spain, so as to exploit the seasonal migration of the fish through the Straits of Gibraltar. (2) Its product is variously said to be (matured) ‘two years’ (AHA) or ‘four years’ (AAAA and AIIIIA), and it is not clear which should be understood here. AIAA has actually been written, the last three letters resembling VM in SVMAVR. Perhaps AIIA was intended. (3) The numeral ‘80’ is unexplained. Other examples of LXXX are known (e.g. CIL iv 5637), and also LXXXIV and LXXXV. One might have expected a note of capacity, but two such amphoras found in Colchester held 16.5 litres and 18 litres, which neither by volume (30 and 33 sextarii) nor by weight (55 and 60 librae, assuming a specific gravity of 1.1) suggest the numeral ‘80’. (4) Lastly the producer's name in the genitive case. Gaius Asicius Probus seems to be hitherto unattested, but his distinctive Italian nomen is found outside Italy precisely at Gades (Cadiz): see CIL ii 1772, the tombstone of L. Asicius Suavis. Many of these amphoras come from the Cadiz region.

⁵⁷ Initial D is certain, but other letters are ambiguous and perhaps defective. Presumably a personal name in the genitive case, somewhat resembling DOMITI(?) in CIL iv 5640.

⁵⁸ During rescue excavations by Pre-Construct Archaeology directed by Robin Taylor Wilson. Information on this and the following four items from Malcolm Lyne.

⁵⁹ In excavation by Pre-Construct Archaeology directed by Mark Beasley.

⁶⁰ In excavation directed by Tony Johnson for the London and Middlesex Archaeological Society. Information from Fiona Seeley of the Museum of London Special Services section.

⁶¹ Saevus, although a common Latin adjective (‘fierce’), is not attested as a cognomen. There is one instance of Saevius as a nomen (CIL xiii 11407).

⁶² In excavation directed jointly by Hedley Swain and Harvey Sheldon for the Museum of London, Birkbeck College (University of London) and the Channel 4 ‘Time Team’. For the fragment of a marble inscription also found, see above, No. 1.

⁶³ The stamp runs across the break between the two fragments, the first part of the stamp, including the vertical stroke of L, being on (a) and the rest of the stamp, including the horizontal stroke of L, being on (b).

⁶⁴ The stamp is from the same die as RIB 11.5, 2485.5, which is illustrated by a composite drawing of two examples, fragmentary but complementary. We follow the editors' reading and expansion here, but autopsy actually suggests PPBRLCvV, the die having suffered damage, a triangular piece of wood between the apex of A and the cross-bar having flaked away. This reading would prompt the expansion Lo(ndini) Au(gustae) (‘… at London Augusta’), another testimony to the award of the title Augusta to London at some date; for the other evidence, see Rivet, A.L.F. and Smith, C., The Place-Names of Roman Britain (1979), 260.Google Scholar But difficulties remain.

⁶⁵ Displayed in Vindolanda Museum, where Robin Birley made it available.

⁶⁶ During excavation for the Vindolanda Trust directed by Robin Birley, who made it available. It is now in Vindolanda Museum.

⁶⁷ By metal-detectorists, but excavated by the Trent and Peak Archaeological Unit, University of Nottingham, directed by Lee Elliott. Darrel Garton, Lee Elliott and Steve Malone provided full details, including the interim report by Elliott, and Malone, in Transactions of the Thoroton Society of Nottinghamshire 103 (1999), 88–89Google Scholar, with pl.

⁶⁸ This is the usual form of construction; see Guy, C.J., ‘Roman circular lead tanks in Britain’, Britannia 12 (1981), 271–6.CrossRefGoogle Scholar

⁶⁹ F in FELIX is like an E, but with an elongated upper cross-bar; it is thus closely similar to that on the lead tank found at East Stoke, c. 5 km to the north-west (RIB II.2, 2416.8 with pl. VI B). This detail, and the general similarity of the two, suggests that they are the products of the same artisan or workshop.

⁷⁰ During excavation by Bath Archaeological Trust whose Finds Officer, John Clarke, made it available to RSOT.

⁷¹ Probably part of a personal name ending in -llus. There are many possibilities, ranging from conventional cognomina like Gallus and Marcellus to diminutives in -illus.

⁷² With the next three items during excavation by Birmingham University Field Archaeology Unit directed by Iain Ferris, who made them available to RSOT.

⁷³ S has been scratched over a thick diagonal stroke, and is followed by a ‘letter’ which can be variously interpreted: it may be part of an earlier graffito (perhaps A) scratched out, or II (i.e. E) scratched out because it was inadvertently written twice. Whatever the explanation, this ‘letter’ should be disregarded, since the succeeding letters UNICI are certain, and Senicus belongs to the large group of Seno- names found in Celtic-speaking provinces. In Britain, it occurs (as Senica) in RIB 374. In this form it is rare; but the developed form Senicianus is common.

⁷⁴ The two graffiti are differently aligned, and (a) is much larger. In (b) the alternative reading IE[…] is excluded by the sequence of letters, and the space available would permit only 1–2 more letters, now lost. For other examples of owners' names abbreviated to TER, see RIB II.7, 2501.537, and RIB II.8, 2503.424. Tertius is the most likely, then Tertulias, but there are other possibilities.

⁷⁵ Victor is much the most likely name in (c), but there are other possibilities. T and (?)TR in (a) and (b) would be the abbreviated names of previous owners.

⁷⁶ The sherds are in private possession, and only a photograph was available. Possibly C was lost by abrasion, but there is no evidence of this on the photograph.

⁷⁷ In excavation by the Littlehampton Natural Science and Archaeology Society, by whom it was given to Littlehampton Museum (accession No. A1281). Information from Dr Ian Friel, Museum and Community Arts Officer, Littlehampton, who sent photographs and a drawing to RSOT.

⁷⁸ The name may be Celtic; Creticus, the victorious cognomen assumed by Q. Caecilius Metellus (cos. 69 B.C.), would be a coincidental Latin homonym. For instances see Holder (s.v.), and especially CIL xiii 3183.8 and 9, Creticus Runatis (filius). In Britain, see RIB II.7, 2501.148 (Caerleon).

⁷⁹ With the next item during excavation by Falkirk Local History Society directed by Geoff Bailey, who made it available. They will go to Falkirk Museums. Only the tip survives of the cross-bar of T, but it does not seem to be casual damage. In view of the previous item, MAT[…] is likely, although the hand is different; the name may have been written in full, or abbreviated to MAT.

⁸¹ Perhaps in 1959, when Discovery and Excavation in Scotland 1959, p. 40, reports the discovery of ‘Agricolan’ samian, but there were other finds of Roman pottery at the Gothic Iron Foundry, Glasgow Road, which gave this item to Falkirk Museums in 1986. Information from Geoff Bailey, who made it available.

⁸² The use of QV for CV is quite common (for examples see ILS III, p. 831, especially qum < cum), the confusion being due to the tendency to reduce [kw] to [k] in Vulgar Latin. The Celtic name Cunavo seems to be unattested, but the name-elements Cuno- and -avos are frequent, and the name Avo occurs in RIB 931.

⁸³ During excavation ahead of a sewer trench by Frances Murray for Falkirk Museums. Geoff Bailey sent a drawing.

⁸⁴ Colyer, C., Gilmour, B.J.J. and Jones, M.J., edited by Jones, M.J., The Defences of the Lower City: Excavations at The Park and West Parade 1970–2, and a Discussion of Other Sites Excavated up to 1994 (The Archaeology of Lincoln Vol. VII–2, CBA Research Report 114, 1999), 47Google Scholar with pl. 28, and 50, n. 1. RSOT.

⁸⁵ Adams, J.N., ‘Two notes on RIB’, ZPE 123 (1998), 235–6.Google Scholar For the formula see Tab. Sulis, pp. 65–6. PETMITTAS for permittas is a copying-mistake due to confusing a cursive R for T; compare BRVCERI for Bruceti in Tab. Sulis 10,2. RSOT

⁸⁶ Adams, op. cit. (in previous note).

⁸⁷ The cognomen PRIMI is preceded by an abbreviated nomen now VL, an error for IVL, VLP or VAL. AU are possible with the accidental omission of one letter, but the nomina Julius and Valerius are much more common than Ulpius among centurions, especially in the reign of Hadrian. There is no trace of ligatured VA, but between (I)VL and V(A)L there is little to choose; however, the century of Iulius Primus is already recorded just east of Birdoswald URS 48 (1958), 152, No. 10(f), and probably also by RIB 1369 (west of Benwell).

⁸⁸ By R.P. Wright for future publication in RIB III as ‘monumental’, which is why his drawing was overlooked when the item was included in RIB II as ‘instrumentum domesticum’.

⁸⁹ It belongs to the collection of antiquities from Waddon Hill formed by James Rails of Bridport before 1895. Like the previous item it was classed as ‘monumental’ by R.P. Wright, and his draft corrigendum was overlooked when RIB II was compiled. We reproduce the drawing he made for RIB III, showing the embossed dots above and below AL[…]. He noted that there was no trace of a third letter.

⁹⁰ In I ine 4, NI was inscribed as N, perhaps by oversight. In lines 4 and 5, there is no trace of a medial point after C, but the surface is rather worn. Line 6 is badly worn, and the full reading is restored from CIL vi 30907 (see below, note 95).

⁹¹ Mr Clayton made it available to RSOT. For the published details see the catalogue of Bonhams antiquities sale, 7 December 1993, Lot 16, and Fox & Co (Yeovil), antiquities catalogue No. 5, p. 6. Other information from David Brown, then of the Ashmolean Museum, Oxford, and from the two previous owners.

⁹² RIB 762, cf. RIB 658 (the temple at York). This objection is not decisive, since the cult is quite well attested in Britain; see Henig, M., Religion in Roman Britain (1984), 114–15.Google Scholar

⁹³ CIL vi 30907 = ILS 3433, first published in Notizie degli Scavi 1888, p. 571, and Bulletino della commissione archeologico comunale di Roma 1888, p. 405.

⁹⁴ Information from Robin Birley, Director of the Vindolanda Trust.

⁹⁵ Information from Siobhan Ratchford, Dumfries Museum.

⁹⁶ Information from W.B. Griffiths, Arbeia Roman Fort and Museum, South Shields.