² During rescue excavations for Suffolk County Council directed by Judith Plouviez with the support of English Heritage. For the site see Britannia xvii (1986), 404. Ms Plouviez provided full information and a drawing (by Rebecca Archer).

³ By Mr C. Williams, who gave it to the Roman Legionary Museum, Caerleon, where David Zienkiewicz made it available to RSOT.

⁴ Only the bottom survives of the first two letters, which are followed by a round hole (used for fixing?), and then apparently uninscribed space. Only the upright with serif survives of the last letter of all, which could thus be P (for p(osuit)) or even D, but the local formula was evidently FC (see RIB 361–3, 367, 372–3, 375). The formula of age at death was followed by the name(s) of the person(s) responsible for the monument, perhaps a woman's name […] cuda and [coniux ei]us (cf. RIB 371).

⁵ During excavation by the Carlisle Archaeological Unit directed by Ian Caruana, who sent a drawing and other details.

⁶ During consolidation after the excavation directed for Cumbria County Council (Britannia xix (1988), 436- 7, with fig. 13) by Tony Wilmott, who made it available. The stone remains in position.

⁷ The name Iulius without cognomen is also found as a quarry-face graffito (RIB 1950) at Coombe Crag, 2.75 km south-west of Birdoswald.

⁸ By Alan Whitworth, in the course of his measured drawing of the Wall; see further Britannia xix (1988), 494, Nos 14–23, and xx (1989), 333, Nos 6–11. Mr Whitworth provided a photograph and full details.

⁹ During excavations directed by Mrs M.U. Jones for the Ministry of Works. See JRS xlix (1959), 119, with fig. 16 (plan of site). Details of this and the next two items from Charmian Woodfield, who is preparing the site for publication. Mrs Woodfield provided copies of a drawing and rubbing (by Mrs M.U. Jones).

¹⁰ There is a hint of an additional letter at the beginningof the surviving text, joined or ligatured to the V, so perhaps […]NCA, […]NVCA, […]VNCA or […]MVCA.

¹¹ See below under Addenda et Corrigenda, items (c)-(e). Mrs Woodfield provided a copy of a photo by Tom Jones.

¹² With the next nine items during excavation by the Carlisle Archaeological Unit directed by Ian Caruana, who provided drawings and other details. Graffiti where less than three letters survive complete, and marks of identification (single letters, crosses, etc.), have been excluded here, but they will all be published in the forthcoming report on the Annetwell Street excavations.

¹³ The personal name Novixius seems to be unattested, but it is probably a variant of Novicius (RIB 200, etc.); cf. also Novisius (Cod. lust, iv 44.6).

¹⁴ This Latin cognomen is frequent in CIL xiii.

¹⁵ Perhaps the abbreviated name of a century. However, it is not clear whether the letters have become torn; an inverted reading cannot be excluded, ALC or even ALA.

¹⁶ Although Annius is a common Latin nomen, it may also have been a Celtic personal name, judging by the occurrence of ANNIOS as a samian potter's signature (Oswald Index, s.v.). It is also quite frequent as a cognomen, especially in Spain.

¹⁷ The initial letters of both lines have long descenders. The second letter is a vowel, but too little survives to decide between II, I, V and O.

¹⁸ Part of the potter's signature. The script as interpreted here resembles that in waxed tablets of the first and early second centuries: V not II is read because of the upward curl with which the first downstroke terminates; and in M the third stroke is omitted or only implied by an upward termination of the second stroke.

¹⁹ This masculine personal name is Celtic since it is found in CIL xiii and CIL v, and must be cognate with S(a)enus (RIB 67, 685, and AE 1956.249, civis Dumnonius). For the developed form Sennianus see below, No. 49 with note. The name-element is often found superficially romanised in Britain and Gaul as Senilis, Seneca, Senecianus, etc.

²⁰ During excavations for the Colchester Archaeological Trust directed by Philip Crummy. Details from Nina Crummy who submitted the sherd for inspection. For the excavations in 1985 see Britannia xvii (1986), 405.

²¹ There is a greater or lesser degree of uncertainty about the first four letters, especially the third. The first letter could just possibly be R, the second might be C while the fourth might be R; the third letter from the end looks almost certainly to be N but could just conceivably be A without the cross bar followed by I or even the letters RI. The last two letters XI might be a numeral or the ending of a verb in first person singular of the perfect tense, as of fingo, lingo, mingo, pingo or tingo. If this is correct, then the first few letters might be either a personal noun acting as its subject or a substantive acting as its object.

²² During excavations for Hampshire County Council and the University of Reading directed by Professor M.G. Fulford, who provided details and submitted the sherd for inspection.

²³ The first two letters are not certain, the A, if that is what it is, lacking a cross bar. The character interpreted here as ligatured IO takes the form of a circle with a vertical line through the centre. While this could, formally, represent the dipthong OI, IO is much more likely, especially given the letter L following, the three letters together being part of the diminutive suffix -ioluslalum. Such an ending could be either part of a substantive or a personal name (for the latter see the section in Kajanto, The Latin Cognomina, on diminutive forms, especially pp. 124–5, which deals with the suffix -ulusla, replaced by the form -olus after vowels). However, there is no obvious short personal name or substantive with stem ending in R of which this would be the diminutive.

²⁴ During excavations directed by Mrs Niblett and Mr C.S. Saunders for Verulamium Museum. For the excavations in 1986 see Britannia xviii (1987), 329. Full details from Dr S. Greep, who submitted the sherd for inspection.

²⁵ Though the initial C is largely missing, reference to I. Marriott's unpublished letter group index of personalnames in RIB and CIL xiii shows that its restoration here is certain. Cupitusla is found 118 times in CIL, of which 79 examples come from Celtic areas. For examples of the nomen Cupitius from Britain see Britannia x (1979), 347–8, No. 20 (Binchester), and abbreviated to the first three letters, RIB 344 (Caerleon) and Britannia xii (1981), 369–70, No. 4 (Tarrant Hinton, Dorset).

²⁶ By the farmer, Mr Welberry. Information from Miss R.H. Healey, who submitted the sherd for inspection.

²⁷ This could be part of a personal name either beginning Novi-, or, if the last vertical character was originally followed by a second (i.e. NOVI[I…]), Nove-, as in the relatively common name Novellius.

²⁸ By Mr A.H. Gammon, who made them available to RSOT.

²⁹ p(rae)p(ositus), the acting-commander of a unit, cannot be excluded, but where sealings specify the issuing officer's rank it is always that of centurion (in alae of course decurión). Marcus Ulpius Max(…) was presumably the chief centurion of one of the three British legions in the second or third centuries, but the name is too colourless for closer identification. An Ulpius Maximus is attested probably as centurion of Leg. V Macedonica in 238/44 (AE 1971.386 = Inscriptiile Daciei Romane III.3, 237).

³⁰ The constellation Leo contains fifteen bright stars, but they include a distinct ‘sickle’ formed by Regulus (a Leonis) and five other stars, so the die may have been intended to have astrological significance, the Moon in Leo. Regulus and a crescent moon would suggest an imperial horoscope: Firmicus Maternus, Mathesis vi. 2.2.

³¹ During the observation of contractors' work after the conclusion of excavations conducted by the Museum of London's Department of Urban Archaeology under the direction of Julian Ayre. The vessel is similar in form to that illustrated in R.E.M. Wheeler, London in Roman Times, pl. lv, No. 5, itself similar to Haltern 70, vessels produced in Baetica in the mid first century, a date confirmed on palaeographic grounds in the present example. Information from Beth Richardson. We are grateful to Pedro Paulo Funari and to Paul Sealey for interpreting and discussing this text with us.

³² For oliva(e) alba(e) cf. CIL iv 2610 and 9437 and xv, 2 4802, perhaps an amphora of similar form. Funari observes that Columella (xii. 49) also refers to the practice of using amphorae as containers for green olives in must. For an amphora of similar form found full of olives in the Thames estuary in 1983 see Sealey, P. and Tyers, P. in Antiq. Journ. lxix (1989), 53–72.CrossRefGoogle Scholar The numeral CCL cannot refer to weight in pounds since the total weight of the vessel and its contents will have fallen far short of 250 Roman pounds, 81.85 kg. It is possible that it represents a measure of volume, perhaps cyathi: the amphora from the Thames estuary is estimated to have had a capacity of something over 12 litres. The cyathus was the equivalent of 0.0456 litres and so 250 would be the equivalent of 11.4 litres. For the expansion of the nomen in C.L.A., Funari draws our attention to the painted inscriptions on two Dressel 20s, found with material dating to the first half of the first century in the ditch of the Castra Praetoria, which carry the painted inscription M LOCILLI ALEXANDRI (CIL xv, 2 3660–1), and to the painted inscription from the same provenance, also on a Dressel 20, with the inscription C LVCILI… (CIL xv, 2 3662).

³³ During excavations by the Department of Urban Archaeology of the Museum of London directed by J. Schofield.

³⁴ During excavations by the Upper Nene Archaeological Society directed by Roy Friendship-Taylor, who provided full details and illustrations of this and the next six items. For the site itself 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, published by the Upper Nene Archaeological Society (1989).

³⁵ The first character could be the Roman numeral / with a short superscript bar, or possibly the letter T, although the combination of consonants -tc- in personal names at least hardly exists.

³⁶ The foot of the letter Fis provided with an exaggerated descending serif and so is, overall, 15 mm high.

³⁷ If the man commemorated bore the imperial nomen Claudius, then the praenomen Tiberius was written in full, since the traces of the letter following C are consistent with the letter E but not the letter C.

³⁸ There is some doubt as regards the first surviving letter. If the reading is correct, there is no obvious expansion and it is just possible that the genitive of the common name Respectus was intended.

³⁹ During excavations for the South Staffordshire Archaeological Society and H.B.M.C., directed by Frank and Nancy Ball, who made it available to RSOT.

⁴⁰ The numeral is incised with a thicker point on a slightly different alignment, so is not necessarily part of the same graffito. […] alis is presumably a personal name. Only the bottom half survives of the preceding letters, as it were C intersecting the first of two close-set vertical strokes, which can be restored in many different ways.

⁴¹ During excavations directed by Judith Plouviez for Suffolk County Council with the additional support of English Heritage. For the excavations of 1985 see Britannia xvii (1986), 404, and for a bronze letter from the same site, see above, No. 1. Ms Plouviez sent details of this and the next four items, provided drawings (of the lead sealings by Rebecca Archer and of the graffiti by Donna Sullivan), and submitted the objects for inspection.

⁴² There seems to be no attested nomen beginning Noce-. Schulze includes the nomen Occius, but the initial letter, though somewhat blurred, does not seem to be M, as in M(arci) Occ(ii) / Vari.

⁴³ The top horizontal bar of the final letter on the obverse hardly extends to the left of the upright and so would be more compatible with F, but there is no clear second horizontal stroke below it, so that the letter may after all be T. There are a number of possible expansions. If Ruf- is a nomen then Rufius and Rufinius are possible expansions; if a cognomen (but this is less likely since Agr- is probably an abbreviated cognomen), then Rufus or Rufinus would be likely, though other expansions are possible.

⁴⁴ There are no common nomina beginning Agr-, so this is probably an abbreviated cognomen such as Agricola, Agrippa or Agrippinus, all relatively common.

⁴⁵ During extraction of gravel by mechanical excavator in Shepperton Ranges gravel pit, as already noted in Britannia xix (1988), 478. Details from R. Poulton, Surrey County Council Planning Department (Conservation and Archaeology Section), who made the plate available to RSOT.

⁴⁶ The surface of the plate is worn, and there are other surface scratches (not drawn), so the reading of the dotted letters is not certain. The name is unattested, but Celtic personal names in -genus ('born of) are common. Mallus is well attested, but not Maints. Above the graffito, where the flange curves to join the body of the plate, there may be traces of a second graffito in cursive letters only 2–3 mm high, but they cannot be distinguished from the marks of wear at this exposed angle.

⁴⁷ During excavation for the Ministry of Works directed by Mr (now Professor) J.S. Wacher (see JRS 1 (1960), 217–8). This and the next item were identified by Mr A.D. Hooley of the York Archaeological Trust, who made them available with other details.

⁴⁸ During the excavations directed by Mr (now Professor) J.S. Wacher for the Ministry of Works, in which the previous two items were found, and for the Department of the Environment (see Britannia iv (1973), 279–80). The items were made available by Dr Jerry Evans and Peter Wilson of the Central Excavation Unit.

⁴⁹ They will all be published, with line-drawings, in the forthcoming final report on the excavations, to which reference should also be made for fuller discussion of readings, individual names, and the significance of the collection as a whole. Graffiti already published by RPW in JRS 1 (1960), 240–2, and by RSOT in Britannia x (1979), 355, have been excluded from this selection unless a new reading or interpretation is offered; likewise graffiti which are only marks of identification (a cross, etc.), a single letter, or incomplete and less than three consecutive letters.

⁵⁰ RPW reads APA, but the name is doubtfully attested, and the second ‘A’ is better read as a triangular O.

⁵¹ RPW reads FELIIA?, but ignores the name-ending VS which follows.

⁵² This cognomen is one of several derived from Mars and is already well attested in Gaul and Spain, but apparently not in Britain. In form it is a diminutive of Martius.

⁵³ The cognate Remnius could also be restored, but is much rarer. The nomen Remmius is probably of Etruscan etymology (W. Schulze, Eigennamen, 219).

⁵⁴ The graffito as it stands is a woman's name, Sennia, but is not necessarily complete; and all the other Catterick graffiti, where this can be ascertained, are of men's names. For Sennianus see JRS xxx (1940), 190, No. 30. The name is developed from Senna (see above, No. 17 with note) and S(a)enus (ibid.), cf. ILS 2572, D. Senius Vitális, civis Brit.

⁵⁵ Interpunctuation by leaf stop is uncommon in graffiti.

⁵⁶ Graffiti incised before firing on amphoras are usually names and dates, evidently the potter's ‘signature’. This graffito contains two cognomina, Fructuosus, well attested in Spain and Africa, and Phileros, assimilated here to the Latin masculine termination in -us, which is a personal name of Greek etymology well attested in Italy and the western provinces. Fruttuose, however, is in the vocative case, and this ‘signature’ is apparently cast in the form of a greeting (possibly [bene valea] or [felix viva]s) addressed to Fructuosus by Philerus.

⁵⁷ The personal name Borus and its two cognates are attested in CIL xiii, and must be of Celtic etymology. Two instances in particular suggest that the writer may have been an auxiliary solider, perhaps from Raetia: CIL xiii 8320 (Cologne), Borissus, father of a soldier of Coh. I Vindelicorum; EE iv, p. 502 (Regensburg, diploma dated A.D. 153), Borus, father-in-law of a veteran of Ala II Flavia milliaria p.f.

⁵⁸ The space after T and the unlikely sequence TVEL justify the expansion t(urma). Velox is a common Latin cognomen, but this seems to be the first instance in Britain. The letter after CA is probably M, N or P. This graffito is the most direct evidence there is of the Roman garrison of Catterick, an ala or cohors equitata. The siting of Catterick on the lower Swale, in country suitable for cavalry just south of a major road-junction, makes it an obvious candidate for an ala; however, if Wacher is right in deducing that the early forts were only 4.12 acres in area, the garrison must have been a cohors equitata. Some support is given by the mention of a century on a mortarium (see below, No. 64), but there is no knowing whether this was auxiliary, not legionary, nor indeed whether it is contemporary.

⁵⁹ Graffiti deeply incised on the rim of an amphora after firing are usually notes of capacity (m(odii) and numeral), but this seems to be part of a personal name.

⁶⁰ RPW reads […]IONIS, but the end of T can be seen in the broken edge of the sherd. Several personal names terminate in -tio, the likeliest being the Latin cognomina which form a ‘numerical’ sequence, Quartio, Quintio, etc.

⁶¹ Numerals are quite often found incised on the neck or handle of amphoras, presumably a note of weight or capacity. ‘IX’ occurs on the handles of amphoras (unpublished) from Richborough and Corbridge; ‘VIIII’ on the handle of an amphora from Ambleside (Cumb. Westm. n.s. xv (1915), 57, with fig. 27).

⁶² Perhaps evidence of the garrison of Catterick: see above, n. 58.

⁶³ Gerontius is well attested in CIL xiii, and was current in Britain at the end of the Roman period, which it survived as Geraint: see Birley, A., The People of Roman Britain (1979), 211 n.Google Scholar This is one of the earliest Greek inscriptions from northern Britain, but cf. RIB 662–3.

⁶⁴ By Mr J.G. Hunt, who made it available through Mr A.D. Hooley of the York Archaeological Trust.

⁶⁵ There is no exact parallel, but cf. RIB II 2411.24.

⁶⁶ By Richard Grasby, before cutting a replica in the manner of his Roman predecessor. A full account will be published by him and David Zienkiewicz.

⁶⁷ Bell drew them, together with RIB 1455 and 1464 (both found at Chesters before 1840), without note of date and provenance but in the same style and on identical paper as stones from Risingham (glossed ‘Found at Risingham’) published by him in AA¹ iii (1844). See Bodleian MS Top. Northumberland C. I, ff. 315–8 and 311–4 respectively.

⁶⁸ Full details from Charmian Woodfield who examined the material in the course of preparing the report on the excavations directed by Mrs M.U. Jones for the Ministry of Works. For a plan of the site showing the location of the four principal buildings, see JRS xlix (1959), 119, fig. 6. For examples of three other tile stamps from the same site, see above, Nos 6–8.

⁶⁹ As suggested by Dr J.R. Rea. There is probable trace of the crossbar of the first Tin the grain of the tablet, and Domitius is a common nomen. It is particularly frequent in Spain and Narbonensis (Syme, R., Tacitus (1958), 783–4).Google Scholar