² During excavations by the Museum of London's Department of Urban Archaeology, directed by Peter Rowsome. Information on this and the following item was provided by Victoria Cassely. For the site see Britannia xxii (1991), 269–70.Google Scholar

³ The letters MAX suggest part of an imperial title, while the letters ]NIA[ could be part of a cognomen ending -nianus, for which there are many possibilities. This last seems a more likely restoration than a nomen ending in -nia, since in a public building an imperial dedication(?) would more probably be made by a male than a female; note, however, RIB 644, the dedication in a bath house at York by the wife of the legionary legate.

⁴ The second stroke only of the V survives and is somewhat more upright than one might expect, but it is hard to see of what other letter it could be part.

⁵ During excavation by the Museum of London's Department of Greater London Archaeology directed by Bruno Barber, who supplied details and made the object available for inspection. For the site see Britannia xxii (1991), 273.Google Scholar

⁶ At the right-hand edge of the stone is a partially preserved groove which, if it is part of a moulded border (rather than part of another letter), would indicate that the fragment comes from the right-hand bottom corner of an inscribed plaque. If this is correct, then the last line may have given the formula, [h(eres) f(aciendum)] c(uravit), preceded by the heir's name in the nominative in the line above.

⁷ During excavation by the Museum of London's Department of Greater London Archaeology, directed by Kieron Heard. Information from Lyn Blackmore of the DGLA, who supplied a photograph by John Earp.

⁸ The use of lead, the retrograde text, and the care taken to give the exact name, all suggest that this is a ‘curse tablet’.

⁹ During excavation directed by Dr Ann Ellison for the Committee of Rescue Archaeology in Avon, Gloucestershire and Somerset. For other tablets from the site see Britannia xix (1988), 485, No. 2Google Scholar; xx (1989), 327–30, Nos 2 and 3; xxii (1991), 307–8, Corrigendum (a). The present item is briefly noted in RSOT's interim report (see Britannia xx (1989), 327, note 4), in the final report, still forthcoming, where it is No. 72, but the text is published by him here for the first time. It has been communicated to seminars in Oxford and Durham, whose members are thanked again for their helpful comments.Google Scholar

¹⁰ It is the best preserved tablet from the site, with a firm, smooth, brown patination but very little corrosion. The hole at the end of line 4 is probably a casting flaw (a ‘cold shut’). On the reverse are strange marks, short diagonal lines in relief, as if the tablet had been cast on an inscribed surface, but they are not mirror-image letters. After being inscribed, the tablet was folded six times.

¹¹ Commentary

(Tab. Sulis is R.S.O. Tomlin, Tabellae Sulis: Roman Inscribed Tablets of Tin and Lead from the Sacred Spring at Bath (1988), reprinted from B.W. Cunliffe (ed.), The Temple of Sulis Minerva at Bath, II: the Finds from the Sacred Spring (1988). Reference is by page (p.) or item.)

2. conqueror numini tuo: for these formulas see Tab. Sulis, pp. 65–6, and compare genium numinis tui and a maiestate tua below.

2–3. rotas duas et vaccas quattuor: it would be interesting to know whether these went together, a two-wheeled cart hauled by four cows, or even a wheeled plough and its team. Rotas to be worth mentioning may have been spoked wheels stored separately, or perhaps are to be understood as the part for the whole. Cows were used to plough light soils (Varro, R.R. 1.20.4; Columella C1 .22.1), and could have worked as draught animals; the Romans had a wheeled plough, not that it has yet been found in Britain (see Manning in JRS liv (1964), 65).Google Scholar

4. resculas: diminutive of res and very rare. According to Lewis and Short, who are summarising Forcellini, it occurs in this sense only in Plautus and Apuleius (once in each), so it may belong to ‘Vulgar’ rather than Classical Latin.

4–5. de hospitiolo meo: another colloquial term; see Britannia xxii (1991), 309, note 82.

6. rogaverim: ‘potential’ (perfect) subjunctive expressing a possibility, used here for politeness towards the deity; a sophisticated alternative to the usual present indicative rogo (see Tab. Sulis, p. 67).

7. tu(u)i: inserted afterwards, perhaps to correct a mistake in a text being copied. The repetition of u may only be a slip, but it could be a hyper-correction of the ‘Vulgar’ tendency to reduce uu to u (see V. Väänänen, Introduction au Latin Vulgaire (1963), 52).

7–8. qui mihi fraudem fecerit: for this formula see Tab. Sulis, pp. 64 and 69.

8–9. non permittas (etc.): see Tab. Sulis, pp. 65–6. ei has been repeated by oversight. The formula nec iacere nec sedere occurs in Tab. Sulis 54; nec bibere nee manducare in Tab. Sulis 41 and in Britannia xv (1984), 339, No 7 (Pagans Hill).Google Scholar

10–11. si baro (etc.): for these formulas see Tab. Sulis, pp. 67–8.

12. nis(s)i: this ‘Vulgar’ gemination is frequent in British ‘curse tablets’ (see Tab. Sulis 32.7, with note).

13–14. meam concordiam habuerit: presumably a formula, but this is its first occurrence; something similar was intended in Britannia xix (1988), No. 2 (Uley), co(n)scientiam perferat.Google Scholar

14. iteratis pr(a)ecibus: this ‘literary’ formula also occurs in Britannia xv (1984), 339, No. 7 (Pagans Hill); see note ibid. The spelling is a hyper-correction of the ‘Vulgar’ confusion of ae and e.Google Scholar

15. petitio mea: the same term may occur as the heading of Tab. Sulis 9. Like commonitorium in another Uley tablet (Britannia x (1979), 343, No. 3), it supports Professor H.S. Versnel's thesis that the authors of most British ‘curse tablets’ thought of them, not as curses, but as ‘juridical prayers’. See Tab. Sulis, pp. 59, 70–1; and Versnel in C.A. Faraone and D. Obbink (eds), Magika Hiera (1991), 60–106.Google Scholar

16–17. The syntax is confused, although the sense is obvious: compare Tab. Sulis 35, rogo [s]anctissimam maiestatem tuam u[t] vindices, with note ibid. of other instances of ‘vindication’ (whether in the sense of recovery of stolen property or of vengeance on the thief), pareat (to ‘obey’) is inappropriate here, and the author must have meant pariat (to ‘accomplish’), the two verbs being confused by an easy ‘Vulgar’ confusion between unstressed i and e. The infinitive construction me vindicatum esse is awkward and unusual; one might have expected mihi vindicationem.

‘ ’ interlineated

〈〉 redundant letter according to the Classical norm

- word divided between lines

The paragraph is original, but the sentence-punctuation is modern and word-separation has been modernised. (The scribe separated words by elongating the final letter or by leaving a space, especially at the ends of lines; he divided words between the lines only in 5 out of 16 lines, and always by syllable. He did not separate conjunctions and prepositions from the following word, nor possessive pronouns from the preceding word.)

The capitals in line 1 are modern

¹² During excavations by the Central Excavation Unit directed by David Neal, who sent details and a drawing. For the excavations of 1990 see Britannia xxii (1991), 252–3.Google Scholar

¹³ This is not part of the name of an emperor: filiations of emperors refer to the cognomina of the father, as divi Vespasiani f for Titus and Domitian, or Aug f if in the lifetime of the father.

¹⁴ During excavations conducted by the Oxford Archaeological Unit. Information from Paul Booth. For the site see above p. 287.

¹⁵ The pit may have been used to dispose of rubble derived from a structure within the walls, from which rubbish was carted and dumped outside the town. It was sealed by a late Roman building.

¹⁶ If the reading in the last surviving line is PP then it would presumably be for p(ater) p(atriae), preceded by the emperor's name in the genitive, as [Vespasia]ni, [Domitia]ni, [Hadria]ni etc, dependent on some such phrase as pro salute. If it is PR, then it is tempting to take it as part of the titles of a governor, [leg. Aug.] pr(o) [pr(aetore)], whose name occurred in the preceding line, also in the genitive and dependent on some such phrase sub cura or iussu. However, in both cases it is hard to explain the M in the first line.

¹⁷ During excavation by Lancaster University Archaeological Unit under the direction of Nick Hair and Kath Buxton. Janet Lambert sent a drawing and full details. The tombstone will be published by Dr David Shotter in the final report on the excavation, forthcoming. For another tombstone from the site see RIB 756.

¹⁸ Coniuci seems to have been inscribed for coniugi by oversight. The cognomina, though superficially ‘Roman’, probably incorporate Celtic name-elements. Sentica seems to be unattested: ostensibly it derives from the Latin nomen Sentius, but should probably be regarded as another of the Sen- names popular in Britain. Likewise Verulus, though in form a diminutive of Verus, may be compared with the place-name Verulamium; Verus and names in Ver- like Verecundus are popular in Gaul and Britain. Verulus itself is well attested in Gaul, but this is the first instance from Britain; however, compare JRS xlviii (1958), 152, No. 10 (e), Verullus.Google Scholar

¹⁹ By Christine Clayborough of Sheffield University, while preparing a dissertation on the church. This incorporates Roman ashlar, and she notes that whereas Millstone Grit is not available locally, the old course of the Ouse ran within 50 yards of the church and York is only 23 miles upstream. When the stone was photographed (see PL. XIX), fragments of plaster still adhered to some of the letters, but had mostly been removed when RSOT examined the stone. His drawing represents the remains of the squared face and the letters now visible. Miss Clayborough provided photographs and other details.

²⁰ Line 1: the wide gap between the first two letters suggests that the first letter is T (L can probably be eliminated), but M or even I are possible. Line 2: the first letter resembles C in the photograph, but this is an illusion due to the broken edge of the stone. What remains of the letter is straight, not curving, and there is no trace of the lower return of ‘C’. Line 4: the diagonal ligatured to M never reached E, nor was it followed by a vertical stroke (for N), so A ligatured must be read. Line 5: the final letter might be M, but the resulting letter would be wider than the considerably taller M above, so N must be read. The uninscribed space below is greater than that between the lines, so this may be the bottom line of the inscription.

Unfortunately too little survives for useful comment. SSAB is probably […u]s Sab[inus], and it is tempting to see a reference to Octavius Sabinus, governor of Lower Britain in the period 263–8 and possibly attested at York (Britannia xvii (1986), 436, No. 10), but the other letters lend no support. In particular the sequence SPRCON, which contains at least one abbreviation, should be distinctive, but we have not been able to identify it.

²¹ During excavation for the Bowes Museum directed by Dr R.F.J. Jones and Mr S. Clarke: see Britannia xxii (1991), 238. The stone is now in the Bowes Museum, where John Pickin made it available.Google Scholar

²² Lines 1 and 6 are conjectural, but altars dedicated by units and their commanding officer are usually dedicated to Jupiter, and in this formulation. After I(ovi) O(ptimo) M(aximo), the next line regularly begins with the name of the unit. Since [al]a must be restored in line 2, this determines how much text is missing to the left. At the beginning of line 3, therefore, there is only enough space for [V]M, so line 2 must have ended in VET[TON], which determines how much text is missing to the right. There were no medial points: the mark between C and P is only casual damage.

The ala Vettonum is already known to have garrisoned Binchester (RIB 1028, 1035), but in this text its title C(ivium) R(omanorum) was evidently omitted, as in RIB 730. The restoration of [R]ufus as the cognomen of its prefect is certain. His nomen ended in -onius, but the exact number of letters lost cannot be established, not least because we do not know whether his praenomen was specified. Nomina in -onius are fairly common, but Antonius is much the most common, so that Antonius Rufus is likely but not certain. It is the only possibility in Devijver, Prosopographia Militiarum Equestrium, where it may be noted that, although the career of C. Antonius Rufus (A 143) excludes an identification, there seem to have been two procurators of this name in the second century: see Supplementum I, A 143, and especially Fitz, J., ‘Les deux C. Antonius Rufus’, Alba Regia xx (1983), 273–4 = AE 1983, 769.Google Scholar

²³ During excavation for the Tyne and Wear Museums Service directed by Paul Bidwell, who made it available.

²⁴ All four letters are incomplete; the first two are 58 mm high, the other two can be estimated to have been c. 50 mm. Only part of the first stroke of the second M survives, but its forward rake (parallel with that of the first M) makes the reading likely. The combination -MM- might suggest the pro pietate ac devotione communi formula, but this has already been found at South Shields (Britannia xvi (1985), 325, No. 11), and to accommodate ED in the line above entails a very large inscription like RIB 1235, where this combination occurs. Since the fragment is so small, no restoration is possible.

²⁵ By Jim Crow, who pointed out that the fort wall has been rebuilt here using squared blocks larger than the usual ashlar of the fort wall and Hadrian's Wall itself. A reworked stone is therefore quite likely: see further, next note.

²⁶ The diagonal of N is clear, followed by faint but sufficient trace of the second upright. There is no space after the D, and it would be an odd text which ended with -ND. This makes it look as if it has been cut from a longer text; note also that the axis of the letter diverges from that of the stone, though not decisively. Possibilities include a ‘centurial stone’, say of the centurion lulius Candidus attested hereabouts (RIB 1632, 1646), although the letters are rather large for this; or an eccentric building stone like RIB 1712; or a much larger slab, where the sequence -ND- is not distinctive and [Alexa]nd[er Severus] is only a guess.

²⁷ By Alan Whitworth, who sent photographs and full details of this and the next two items. For similar stones, usually thought to mark quarry-batches, see Britannia xxii (1991). 297, No. 8 with note 17.Google Scholar

²⁸ By Alan Whitworth, who sent a photograph and full details.

²⁹ The top of the circie has been removed by subsequent tooling. It is an elaboration of the crosses cut on Wall facing stones: compare Britannia xxi (1990), 366, No. 5. Usually they take the form of X and are thought to mark quarry-batches: compare Nos 13–15 above, and see note 27. A cross within a circle or rectangle sometimes occurs as a graffito on pottery, and is obviously a mark of identification, although it has sometimes been identified as ‘Christian’. That it also occurs on a stone presumably of Hadrianic date rules out this hypothesis.Google Scholar

³⁰ Near Britannia xxii (1991), 297, No. 3, and during the same excavation directed by Tony Wilmott for Cumbria County Council.Google Scholar

³¹ The site of other quarry-inscriptions (RIB 1946–52), which have now also become detached, and must lie among the fallen debris. This new inscription was first identified by David Parsons, University of Leicester, who sent photographs and other details.

³² For the animal compare No. 17 above. Both look like horses, but they have horns. They are not obviously bulls rather than stags, but they do look like the charging ‘bull’ depicted on a tile from Dover inscribed TAVRI (Britannia ix (1978), 478, No. 44 = RIB 2491, forthcoming). The quarry-inscriptions from Coombe Crag do not name the legion concerned, but Legio VI Victrix was active at the nearby Lodge Crag quarry (RIB 1953). The bull may have been its emblem, but the question must remain open: see note to RIB 2411.34.Google Scholar

The surface immediately to the left of SIA is uninscribed, but before that there is a natural ‘chamfer’ in which letters may have been lost. No letters were inscribed to the right of BLEI. After the second S and removed from it, there is a short vertical stroke (drawn) which could be read as I, but is here taken to be casual. The letters are ill-formed, and the reading of those dotted is particularly open to question. There is no obvious sense to be drawn from them, but blundered personal names including a variant of Sums are possible.

³³ With the next item during excavation of the Anglian monastery directed by Dr Christopher Lowe for Historic Scotland. Dr Lowe and Dr David Breeze provided photographs and squeezes. Radiocarbon-dating of the timber superstructure of the building in which the two stones were re-used suggests it was built some time between the late fifth and the early seventh centuries. Other inscribed stones from Birrens were also re-used at Hoddom: see RIB 2097 and 2113.

³⁴ The letters, carefully drawn and rather square, are typical of the inscriptions of Cohors II Tungrorum at Birrens, during the Antonine II period (from c. 158). In line 2, T should be read, not C, since there is no trace of the return; tenet could be read, but […]tente admits of more possibilities. D which follows is unusually narrow, but there is no sign of the loop appropriate to B, P or R it is followed by a serifed upright, E, I or R. In line 3, the loop is too tight for D, so the first letter must be B, P or R.

The Numina Augustorum are often coupled with other gods, e.g. Mercury at Birrens (RIB 2103), but it is not possible to say whether the text began with them. […] tente is the ending of a present participle in the ablative singular, which might also be an adjective or a cognomen (e.g. polens). In default of parallels, no restoration is possible of what was probably a long inscription.

³⁵ Half the foliate pelta survives, its midpoint coinciding with the top of the surviving letters; half the text, vertically speaking, must therefore have been lost. Sense demands that f(ecerunt) was the only word in line 5; and since F was presumably centred, its position makes it possible to calculate the original width of the panel and thus the approximate number of missing letters. This calculation is confirmed by the certain restoration of LEG in lines 3 and 4.

³⁶ There is a suprascript bar (for a numeral) over XXII, which should be restored over VIII in the line above. Since both numerals occur in exactly the same place in the line, LEG only should be restored in lines 3 and 4; there was no connecting et in line 4. (For this formulation compare RIB 2146.) The restoration of lines 1 and 2 is more difficult, since the stone is a fairly brief building-record, and comparable inscriptions (e.g. RIB 2119, 2171, 2180, 2216) limit themselves to the name of the unit responsible. VEXILLATIONES is not long enough to fill lines 1 and 2; perhaps the structure was specified, or more likely, the emperor. If the latter, even in (very) abbreviated form his name would have compelled the abbreviation of vexillationes to VEX. The emperor was probably Antoninus Pius: see below.

This is the first evidence from Britain of vexillations of the two legions of Upper Germany, VIII Augusta and XXII Primigenia, being brigaded together, but they are already attested separately in RIB 2426.1 (probably Hadrianic, compare ILS 2726) and RIB 2216 (undated) respectively. Milliary vexillations of both legions and of Legio VII Gemina were brigaded together in Britain under Hadrian: see ILS 2726. For the dating of this expeditio Britannica, mentioned also in ILS 2735, see Jarrett, in Britannia vii (1976), 146–9, who convincingly reasserts the traditional date of 122. The present inscription, however, does not belong to this occasion, since stylistically it is Antonine: there is no exact parallel from Britain, but the motif of pelta and rosettes is frequent on the ‘distance slabs’ of the Antonine Wall. The contemporary (Antonine I) rebuilding of the fort at Birrens is characterised by ‘beautifully finished masonry’, and the fort's turf rampart on its stone base closely resembles the Antonine Wall. This has all been regarded as the work of legionaries, unlike the ‘second-rate building methods’ of the Antonine II rebuilding: see A.S. Robertson, Birrens (Blatobulgium) (1975), 280 and 284. These legionaries can now be identified as belonging to the two legions of Upper Germany, which evidently sent vexillations to reinforce the British legions engaged on the Antonine Wall itself.Google Scholar

The governor responsible for the Wall, Lollius Urbicus, like other second-century governors charged with major campaigns in north Britain, came straight from Lower Germany, and it is no surprise to find him being reinforced from the Rhine legions. Platorius Nepos had brought the whole of Legio VI Victrix with him from Lower Germany in 122 (ILS 1094 and 1100; compare RIB 1319, 1320, 1427), as well as the vexillations from Upper Germany already mentioned. Iulius Verus in c. 155 also received legionary reinforcements from both German armies (RIB 1322, its exact interpretation a problem: see Wilkes, in ZPE lix (1985), 291—5, and Speidel in Britannia xviii (1987), 233–7). Calpurnius Agricola, who was sent ‘against the Britons’ in c. 161, came from Upper Germany, and seems to have brought troops with him (RIB 589 with Speidel, loc. cit.). RIB 2216 could belong to any of these occasions.Google Scholar

³⁷ During excavation directed for Cumbria County Council by Tony Wilmott, who made them available. Together with two fragmentary graffiti on samian of only two letters, they will be published with line-drawings in the final report.

³⁸ The imperial gentilicium lulius, although commonly borne by soldiers, sometimes occurs on its own without a cognomen: e.g. Britannia xxi (1990), 365, No. 4 with note 7, also from Birdoswald.Google Scholar

³⁹ The same text occurs on a samian vessel at Rocester: Britannia Xxii (1991), 306, No. 45.Google Scholar

⁴⁰ The incised marks between V and I seem to be casual. The reading otherwise is certain, and the graffito should not be seen as a blundered Aurelius. Aurelius is often abbreviated to AVREL, and Victor and Vitalis are two of the most common cognomina in Roman Britain. There are four instances of Aurelius Victor in the RIB Index, at three of the Wall forts and at Risingham.

⁴¹ This is a common cognomen, already attested at Birdoswald (RIB 1907).

⁴² The first letter could be N, but looks more like V. This is presumably a cognomen in -io, e.g. Calvio or Silvio.

⁴³ During excavation by Carlisle Archaeological Unit directed by Ian Caruana, who made it available. The object is probably not a ‘curse tablet’ (compare No. 4 above). The name is not central or written at the top of the ‘page’, and is more likely to be the owner's name inscribed on what was once a much larger sheet.

⁴⁴ During excavations by Birmingham University for the then Ministry of Public Buildings and Works directed by Professor R.A. Tomlinson and Dr L.H. Barfield, who sent a photograph and details of this and the following item. For the excavations in 1971, see Britannia iii (1972), 317.Google Scholar

⁴⁵ In the first line, the second letter could also be an N, or M, but the combination of consonants then resulting is not very likely.

⁴⁶ During excavations directed by Dr Barfield, for which see Britannia iv (1973), 287–8.Google Scholar

⁴⁷ By the Lower Medway Archaeological Group under the supervision of Mike Freeman, who sent details and a rubbing. For the site, see Kent Archaeological Review 101 (Autumn 1990), 1—2, with a drawing of the sherd (ibid. tig. 1).

⁴⁸ The second letter, which consists of a short horizontal stroke branching off from near the top of a vertical upright, could also be a P or an R.

⁴⁹ During excavation by the Museum of London's Department of Greater London Archaeology, for which see Britannia xviii (1987), 337. Lyn Blackmore provided information and a photograph (by A. Fulgoni).Google Scholar

⁵⁰ During excavation by the Museum of London's Department of Urban Archaeology, directed by Peter Rowsome. Information on this and the next item was provided by Victoria Cassely of the Department. For the site see Britannia xxii (1991), 269—70; and for two fragments of monumental inscriptions from the same site, see above, Nos 1 and 2.Google Scholar

⁵¹ By the Museum of London's Department of Urban Archaeology during trial trenching under the direction of Michael Shea, preliminary to the full scale excavations in 1989.

⁵² Given the uncertainty of the reading, any attempt at interpretation, such as to connect it with the place name Ocric(u)lum, 44 miles up the Tiber from Rome, would be highly speculative.

⁵³ During excavation carried out by the Museum of London's Department of Urban Archaeology directed by D.M.B. Lees and A. Woodger. Information from Dr Roberta Tomber, who tells us that the sherds are from an amphora of Dressel Form 2—4 of probably Italian but not Campanian fabric.

⁵⁴ Dr Walter Cockle, who has interpreted the dipinto for us, points out that the eta is intermediate in form between the capital (H) as found in texts of the Greco-Roman period and the cursive form (η) found first in the Byzantine period.

⁵⁵ By Mr S. Dunthorne using a metal detector, in whose possession it remains. Details, photographs and drawing (by K.J. Penn) were supplied by David Gurney of the Norfolk Archaeological Unit, who submitted the object for inspection. We are most grateful to Dr M.M. Roxan for discussing the diploma with us and in particular for drawing our attention to the indications of date.

⁵⁶ At the beginning of surviving line 2 is a short oblique stroke which could be the tail of a Q and at the end of the line is what is probably the top of a letter I, so that it would be possible to read […]QVI QVI[…]

⁵⁷ The diploma was almost certainly issued before 24 September 105 since on that date the relative clause qui quina et vicena plurave stipendia meruerunt was replaced by the ablative absolute quinis et vicenis pluribusve stipendiis emeritis. Compare RMD 9.

⁵⁸ Up to A.D. 91, Maias is often written in full (RMD 4). From some time Between this date and A.D. 100. it is normally abbreviated, as here, to MAI (CIL XVI. 46, dated to the latter year, gives the month in abbreviated form). This means that the diploma should be dated to later than AD 91.

⁵⁹ See preceding two notes. In fact, as Dr Roxan shows in RMD III forthcoming, there are slight grounds for arguing that this diploma belongs to the same issue as the Sydenham and Middlewich diplomas (CIL XVI. 51 = RIB 2401.2 and RMD 8 = RIB 2401.3), and that all three should date either to 4 or 5, or 12 or 13 May 105.

⁶⁰ During excavation by Upper Nene Archaeological Society directed by Roy Friendship-Taylor, who supplied details and drawings of this and the following item, and made them available. For the site see R.M. and D.E. Friendship-Taylor, Iron Age and Roman Piddington: an Interim Report on the Excavation of a Late Iron Age Settlement and Romano-British Villa in Northamptonshire (The Upper Nene Archaeological Society, 1989).

⁶¹ For the excavations of 1989 see Britannia xxi (1990), 323–33.Google Scholar

⁶² During excavations for the Department of the Environment directed by P.A. Barker and Dr K.B. Pretty who provided details of this and the next item and submitted the sherds for inspection.

⁶³ In the first line there is no bar to the A, which is joined at the bottom to the last stroke of the M. The T could be read as a C. but there is no attested name Marcurus. In the second line the F has the form of a vertical followed by a C.

The graffito could be understood as the signature of two potters. Maturus and Fuscus, but there is no connecting et, and one potter would have been enough. Maturius Fuscus is an acceptable name: see Britannia xix (1988), 490, note 11.Google Scholar

⁶⁴ For the excavations of 1966–74, see Britannia vi (1975). 106–17.Google Scholar

⁶⁵ For the restoration suggested, compare MIXST which occurs as part of a inscribed text on an amphora found in Pompeii (CIL IV. 6914), and contrast the motto that occurs on some Rhenish beakers (e.g. JRS li (1961), 197, No. 40 (a)): da merum, ‘serve unmixed (wine)’.Google Scholar

⁶⁶ During excavation for the Tyne and Wear Museums Service directed by Paul Bidwell. who provided photographs. He notes that it is the first sealing of the Fifth Cohort of Gauls to be found at South Shields since the nineteenth century: for the others see RIB 2411.100–105.

⁶⁷ During excavation for Cadw and the Manpower Services Commission by the Glamorgan Gwent Archaeological Trust under the direction of Dr E.M. Evans. Details on this and the following three items were provided by Steve Sell and D.R. Evans, who submitted the objects for inspection. For the excavations of 1985–6 see Britannia xvii (1986), 366–9 and xviii (1987), 307.Google Scholar

⁶⁸ Compare the fourth-century cut and engraved bowl with a Bacchic scene from Colliton Park, Dorchester: J.M.C. Toynbee, Art in Roman Britain (1962), 185, No. 141 and pl. 159.

⁶⁹ Ian Marriott's unpublished Index of letter groups in personal names gives only Cintusmus as a possible expansion of the first line, but this name, though found in the Celtic north-western provinces, is not so far attested in Spain, and there is no certainty anyway that the letters are part of a personal name.

⁷⁰ During excavations for by the Trent and Peak Archaeological Trust directed by Christopher Drage. Details and a drawing were supplied by Maurice Brassington. For the site see Britannia xx (1989), 283–6. Present location of object unknown.Google Scholar

⁷¹ The finder was the late Mr George Greggains, and the fabric red sandstone. RPW took it to be part of a tombstone, and read […]M-IIX ∣ […]TER, a reference to the age of the deceased and to his pater or frater. Autopsy (and indeed JRS lvii (1967), pl. xviii, 3) gives no support to M and T, both dotted by RPW in any case, and we consider the format and the surviving text to be more appropriate to a building stone, possibly one of Conors III Nerviorum.Google Scholar

⁷² The reverse is an incomplete merchant's mark, consisting of the typical ‘four’ symbol on a staff with initials either side; it might have been used to identify a bale of cloth. See F.A. Girling, English Merchants’ Marks (1964), especially 11 ff. and no. We are grateful to Ian Caruana for pointing this out.

⁷³ Palaeographically this reading is easier than Rigiano: the second and third strokes must originally have converged as V, and the fourth stroke (vertical rather than curved) is better seen as L than G. Above all, Iulianus is a common cognomen. We are grateful to Dr J.R. Rea for pointing this out.

⁷⁴ Information from the Assistant Curator, Richard Halliwell.