² During excavations conducted for the Department of Urban Archaeology, Guildhall Museum, by Mr. Charles Hill. For the site see above, p. 347.

³ For other examples of imperial freedmen in Britain see RIB 179 and 643. In l.1 several other expansions e.g. DI]M, compare D(eo) In(yicto) M(ithrae) RIB 1545, or M D]M, M(atri) D(eum) M(agnae) are possible, but on the whole less likely. However, preceding the M, is a diagonal chisel cut, which, while too shallow to be part of the second stroke of the letter A, could be the tail of a centrally placed leaf stop, which would imply that there were only two letters in the first surviving line. In that case [DEO] | [I(NVICTO)] M(ITHRAE) or [MATRI]|[D(EVM>] M(AGNAE), would both be possible. For the spelling of conlabsum in 1.3 with nl compare RIB 979 and 1092, and with b, ibid. 979.

⁴ The letters of the inscription preserve traces of red colouring produced by mixing chalk with crushed tile. The nomen Martiannius is not previously attested, but Martianius with a single N is possible as a formation from the known cognomen Martianus, in which case the second N has been added in error by the stonecutter. On the formation of nomina from cognomina see E. Birley, Roman Britain and the Roman Army (1953), 176. For the spelling of collabsum in 1.9 see above note 3. In 1.7 there may have been an extra letter in the middle of the first word. In 11. 7 and 8 some such phrase as c[um xys]ṭịs would be possible.

⁵ See Petersen, H., ‘Senatorial and Equestrian governors in the third century’, JRS xlv (1955) 47–55. A further difficulty is that equestrian governors will, in normal circumstances, be accorded the title V(ir) P(erfectissimus), not V(ir) E(gregius), although a procurator, acting as the temporary substitute for a senatorial legatus pro praetore would be so described. See Petersen, op. cit. p. 51, and n. 60, Clementius Silvinus, vir egregius agens vices praesidis in Pannonia Inferior. We would like to thank a large number of scholars for discussing the difficulties of this text with us, particularly Prof. E. Birley, Dr. W. Eck, and Dr. J. C. Mann.Google Scholar

⁶ Information on this and the following two items from A. G. Down, who directed excavations for the Department of the Environment and Chichester Excavation Committee. For the site see Britannia vi (1975), 282, and for non-lapidary epigraphic items found elsewhere in Chichester see below Nos. 61–2.Google Scholar

⁷ The size of the letter and the fact that the back of the slab is rough make it unlikely, though not impossible, that this comes from the same inscription as the following item, for the letters could have been of varying heights, and the inscription could have been cut on different slabs.

⁸ In this case AV]G. is an obvious, but not necessary, expansion, since, as noted above, the inscription could have continued on an adjacent slab.

⁹ Reference to Evetts, L. C., ‘The Lettering of Romano-British Inscribed Stones’ in Arch. Ael., Fourth Series, xxvi (1948), 153–71, shows that what remains of the serif is not compatible with either B or D.Google Scholar

¹⁰ The slab, which lacks a moulding, may have been only one of several which contained a lengthy inscription. However, the vacant space between the letters and the edge of the stone shows that l.2 certainly, and l.1 probably, begin new words even if they do not begin new lines. It is, of course, possible that the inscription either lacked a moulded border altogether or that the border was cut from separate pieces of stone, and that this fragment does come from the beginning left-hand edge of an inscription.

¹¹ In l.1 enough survives of the right-hand side of the bottom of the upright to show that the letter was not an E. It is possible that it was an R but other consonants are presumably not in question. In l.2 just enough survives to show that the second letter, presumably a vowel, was an E rather than an 1. The quality of the work suggests an imperial dedication in which case FI[LIO would be possible in the first line.

¹² See Britannia i (1970), 302.Google Scholar

¹³ During excavation in advance of the Catterick by-pass directed for the Ministry of Works by Mr. J. S. Wacher: see JRS 1 (1960), 217–8, with pl. xxv 3 (the stone in position)Google Scholar. This, and not the tombstone from Catterick described in JRS lvii (1967), 204, No. 8Google Scholar, and JRS lviii (1968), pl. XIX 1, is the stone found in the bath-house. With the two hypocaust arches, it was removed to the Bowes Museum, Barnard Castle, where its character was recognized in 1976 by Mr. R. F. J. Jones, who cleaned off adhering mortar and discussed what was revealed.Google Scholar

¹⁴ Perhaps pine-cones. The M is wrongly placed for the usual D(is) M(anibus) formula, and there is no sign of a funerary text; however, the stone in general resembles memorials of a husband and wife depicted side by side. The lines of the shell-moulding confirm there were two heads and shoulders, but it is not possible to say whether they belonged to two portrait busts, two standing figures, or two figures reclining on a couch.

¹⁵ During excavation for the Department of the Environment directed by Mr. P. R. Scott; see p. 313 above. The stone is now in the Bowes Museum, Barnard Castle, where Mr. R. F. J. Jones provided access and photographs, and discussed the reading.

¹⁶ It may be significant that the well-cut D of l.i seems to be the letter next required in l.2; and that l.3 repeats 1.2 with more success. The intrusive M also suggests this is not a continuous inscription, but rather a piece of practice-lettering, done before the slab was squared off (the M has been cut in half in the process), perhaps on the original quarry-face. A possible expansion of EXORD is ex ord(inato), ‘ex-centurion’, since ordinatus in this sense is already attested at Piercebridge (RIB 1022, 1026).

¹⁷ The stones were deposited on loan by the Newburn Council in the Museum of Science and Engineering, Newcastle, until 1971, where they were seen by Dr. D. J. Smith.

¹⁸ As suggested by Mr. (now Professor) E. B. Birley of the numerals cut on similar ashlars excavated by him in 1929 at Chapel House milecastle (MC 9), Arch. Ael. 4th ser. vii (1930), 160. The stones recorded by Horsley, RIB 1370–2, have since been lost and, despite the coincidence of place and date, cannot be identified with those now at Throckley.Google Scholar

¹⁹ Britannia iv (1973), 329, No. 9. It is 65 (not 61) m west of the west wall of Turret 29a.Google Scholar

³⁰ Britannia v (1974), 462, No. 7. It is 75 (not 72) m west of the west wall of Turret 29a.Google Scholar

²¹ All three stones noticed by Professor D. F. S. Scott, and drawn by R.P.W. See RIB 1370–72 for the fine examples at Milecastle 9 (Chapel House) and RIB 1379 for the group seen by Horsley at Walbottle. See also RIB 1383, 1384, 1393, 1411 and item No. 8 (above).

²² Slightly further west in this sector and just east of Turret 29b (Limestone Bank) two comparable inscriptions were found in 1912, RIB 1517 (reading V) and 1518 (reading X in a rectangular frame).

²³ During restoration directed by Mr. R. Dower, who provided photographs, a squeeze, and full details; information was also received from Mr. I. W. Stuart. The stone has been built

²⁴ The centurion is hitherto unrecorded, unless he is to be identified with the ‘ELIVLIANI’ of a building-stone, now lost, from the Wall near Carrawburgh (RIB 1507). The nomen Sollius is well attested in Gallia Narbonensis (CIL xii), a likely origin for a legionary centurion in the reign of Hadrian.

²⁵ During excavation for the Department of the Environment directed by Mr. L. J. F. Keppie, who provided a photograph and full details, and by Dr. J. M. MacKenzie. The stone is now in the Hunterian Museum, Glasgow. For the site see Britannia vi (1975), 20 ff., and p. 304, above.Google Scholar

²⁶ During excavation for the Edinburgh City Museums, in whose possession the stone remains, directed by Mr. N. M. McQ. Holmes. Mr. Holmes provided a drawing, a squeeze, and full details. For the site see Britannia v (1974), 163 ff., and p. 305, above.Google Scholar

²⁷ Found by Derbyshire County Council workmen excavating a drainage-trench beside the Yeaveley-Alkmonton road on land belonging to Mr. F. Potter of Stydd Hall Farm: currently on loan to Derby Museums. Information from Mr. R. G. Hughes and Miss Josephine Dool of the Museums staff who supplied full notes and squeezes. As they point out, the legend is a fuller version of SOCIOR LVT BR EX ARG on pigs found at Tripontium, Warwicks., JRS lvii (1967), 206, No. 21 and pl. xx; Broomfleet, Yorks., ibid, lviii (1968), 210, No. 31 and Belby, Yorks., ibid, xxxi (1941), 146, No. 17, and of the legend soc LVT BRIT EX ARG on pigs from Ellerker and Brough on Humber, Yorks., ibid, xlviii (1958), 152, No. 12, and xxxi (1941), 146, No. 16. First report by Josephine Dool, Bulletin Peak District Mines Historical Soc. 6 No. 2 (1975), III with pl., sent to R.P.W. by Mr. R. W. P. Cockerton, F.S.A.Google Scholar

²⁸ During excavations by Exeter Archaeological Field Unit, directed by Mr. M. Griffiths. Information from Mr. P. T. Bidwell who made this and the following two items available for inspection. For the site see Britannia iv (1973), 313.Google Scholar

²⁹ In (a) only the top of the second stroke of the V survives and a number of other restorations of the letter are possible. If the reading is correct, the name could have been VIIRI or SIIVIIRI, Veri or Severi. In (b) the first numeral is written above the other two and the whole is written below, and on its side, with respect to (a).

³⁰ The bessalis was two-thirds of a foot, or eight Roman inches (0.197 m) square.

³¹ The ‘horizontal’ strokes of the letters F branch diagonally upwards from the middle of the upright. There are other random markings on the brick, including what looks like a v cut with a knife below the letters IIF.

³² For the site see Britannia vi (1975), 276. The initial letter of the second and third lines are very faint.Google Scholar

³³ Information from Mr. P. Crummy, director of excavations for the Colchester Archaeological Unit; he submitted the stamp for inspection, supplied photographs and made several valuable comments. For the site, see above, p. 343. Mr. Crummy points to possible traces of truncated texts, all but totally erased on the short sides, and it is conceivable that the stamp was originally square and inscribed on all four sides, but was sawn in half after damage to one side.

³⁴ Excavations directed by Mr. J. S. Wacher for the Cirencester Excavation Committee. For the site see Antiq. Journ. xlii (1962), 14. Information from Paul Arthur and the pottery assistant, Valery Rigby, who submitted the sherd for inspection.Google Scholar

³⁵ Excavations by Farnham and District Museum Society were supervised by Mr. G. Cole. Information from Mr. M. Millett, who submitted the object for inspection.

³⁸ Full information from Mr. Grahame Soffe who, with Messrs. G. Moore and J. Nicholls, conducted rescue excavations by the South Hampshire Archaeological Rescue Group for the Department of the Environment. Mr. Soffe supplied a photograph and impression of the stamp, which will be deposited with Portsmouth City Museums. For the site, see above p. 366.

³⁷ The stamp is paralleled by examples from a villa 600 m east of Crookhorn (JRS xvi (1927), 232–3, No. 25), wrongly described as Wymering, and from the villa at Langstone, Havant, about three miles east of Crookhorn, Proc. Hants Field Club and Arch. Soc. x (1926/1930), 286.Google Scholar

³⁸ Information from the finder, Mr. M. Corney, who submitted this and the following item for inspection before presenting both to Reading Museum.

³⁹ For a complete example of the stamp, several of which have been found in the south of England, see CIL vii 1241 (Calne, Wilts).Google Scholar

⁴⁰ Excavations for Winchester Excavation Committee, conducted by Martin Biddle, who made this and the following three items available for inspection. For the site see Antiq. Journ. xlviii (1968), 280–84.Google Scholar

⁴¹ Ibid, xlix (1969), 323–26.

⁴² Several names beginning with these letters are given in A. Holder, Altcelt. Sprachsch.

⁴³ Antiq. Journ. lv (1975), 321–6.Google Scholar

⁴⁴ Presumably part of a name such as COMINIVS.

⁴⁵ Antiq. Journ. lv (1975), 295–303.Google Scholar

⁴⁶ Possibly a mistake for AVR(ELII).

⁴⁷ For the treasure, now acquired by the British Museum, see above, p. 333; briefly mentioned by Wild, J. P., Durobrivae 3 (1975), 30Google Scholar. We are grateful to Mr. Painter, K. S. for photographs, drawings and report in Rivista di Archeologia Cristiana li (1975), 333–45, ‘A fourth-century Christian silver treasure found at Water Newton, England, in 1975’. Later see his mono-graph to be published by the British Museum.Google Scholar

⁴⁸ For the use of altare in this sense see Painter, op. cit., 344.

⁴⁹ In (a) omega is inverted and interchanged with alpha. In (b) not only the ligatured N in ANCILLA, but the N in CONPLEVIT are written retrograde. For tua ancilla compare E. Diehl, Inscriptiones Latinae Veteres, 3871. The expression ancilla dei is not uncommonly used on Christian epitaphs (see Diehl, index vii). Painter interprets it as a personal name, Anicilla, though the first stroke of the initial surviving letter makes it V.

⁵⁰ Excavations directed by Miss Jean Mellor for Leicester Museums and the Department of the Environment. Information on this and the following five items from Miss Mellor and the small-finds officer, Mr. Patrick Clay, who supplied photographs. Cohors I Aquitanorum is attested at Carrawburgh under Hadrian (RIB 1550) and Brough-on-Noe under Antoninus Pius (RIB 283). For a tile-stamp of the unit from Brancaster see Britannia vi (1975), 288, No. 25.Google Scholar

⁵¹ EE ix 1296d=JRS xxix (1939), 226, No. 5.Google Scholar

⁵² See Richmond, I. A., ‘Roman Leaden Sealings from Brough-under-Stainmore’, Cumb. and Westm. ser. 2, xxxvi (1936), 104–25.Google Scholar

⁵³ For the site see Britannia vi (1975), 246.Google Scholar

⁵⁴ If, as seems likely, the reverse represents the name of a centurion, despite the absence of centurial sign, IV(LIVS) AQ(VILA) may be intended although there are other possible expansions.

⁵⁵ For a recent discussion of this form of Chi-Rho see Greene, Kevin, ‘A Christian monogram from Richborough, Kent’, Britannia v (1974), 393–5CrossRefGoogle Scholar. For the elongation of the tail of the loop compare Wright, R. P., ‘A Roman Christian monogram from York Minster’, Antiq. Journ. lv (1975), 129–30, where, however, the tail of the loop, and not the upright of the Rho, is used as one of the strokes of the Chi.CrossRefGoogle Scholar

⁵⁶ JRS xv (1925) 250, No. 17 b (A. Smith reported it as IIBIIRIISTO). Interpreted as in the text (in 1945 by R.P.W. and) by Mr. A. J. White, of Lincoln Museum. He has found the contemporary report of the find in the Rutland and Stamford Mercury, 22 Jan. 1926. It was one of a sequence of connected tiles supplying water presumably to the bath-house of the villa. Mr. White suggests that this might be ‘a legal formula for the manumission of slaves’. But as the text would be underground and out of sight it seems better to regard it as merely a casual slogan. R.P.W.Google Scholar

⁵⁷ During excavations directed by Brian Hobley for the Guildhall Museum, Department of Urban Archaeology. Information from the finds research officer, Mr. M. Rhodes. For the site see Britannia vi (1975), 265.Google Scholar

⁵⁸ Excavations for the Southwark Archaeological Excavation Committee by Laura Schaaf and Mike Dennis, who provided full details. Thanks are due to Mr. J. Price of the Ancient Monuments Laboratory of the Department of the Environment who supplied photographs and a drawing by Mr. J. C. Thorn of the Laboratory.

⁵⁹ Compare the branded outer leaf of the tablet from London, issued by the imperial procurators of the province of Britain, JRS xxvi (1936), 265, No. 5Google Scholar, but the present example could derive from a large private business concern. If the reading is correct and a name is intended, CN(AIVS) N(I)P(IVS)·P(…) is just conceivable. For the nomen compare the lead pig from Carmel, Flints., stamped c. NIPI ASCANI, JRS xli (1951), 142, No. 8.Google Scholar

⁶⁰ Excavations directed by Mr. A. Rogerson for the Norfolk Archaeological Unit who submitted this and the following item for inspection. For the site see Britannia v (1974), 439.Google Scholar

⁶¹ For the name compare M. Ulpius Novantico from Leicester, CIL xvi, supp. 160, but Celtic names ending in -co and -ico are common, see these suffixes in de Schaetzen, Index des terminaisons… sur terra sigillata, Brussels (1956), though none is given there ending in -TICO.Google Scholar

⁶² For the Nene Valley Research Committee Mr. A. Challands published it in Durobrivae 3 (1975), 21, fig. 9. Mr. M. Todd reports that it does not match the L(eg.) v stamps found in Lincoln. Mr. Challands considers that it is probably a civilian product. R.P.W.Google Scholar

⁶³ During rescue excavation in Buddle Street directed for the Department of the Environment by Mr. C. M. Daniels, who made this and the following eight objects available, and identified them. They will go to the North Tyneside Museum at Wallsend. For the site, see pp. 306 ff.

⁶⁴ Difficult to read because unevenly impressed on a surface since corroded. The first stroke of the M is abnormally long and curved. The die cannot be paralleled, but in form resembles the centurial marks impressed on the reverse of some cohortal sealings from Brough-under-Stainmore (Cumb. Westm. n.s. xxxvi (1936), 117 and 119 (Nos. 17–21)).Google Scholar

⁶⁵ During rescue excavation in Buddle Street, directed for the Department of the Environment by Mr. C. M. Daniels, who made them available. They will go to the North Tyneside Museum at Wallsend.

⁶⁹ Those found at Burnswark had certainly been fired from the surviving ballista-platforms (Davies, R. W., Historia xxi (1972), 101), and resemble missiles from Masada (Y. Yadin, Masada (1966), 163). But those found in large numbers in forts on the central sector of the Antonine Wall (G. Macdonald, The Roman Wall in Scotland ²(1934), 257, 270, 275, 287; and S. N. Miller, The Roman Fort at Balmuildy (1922), 98) do not occur in so definite a context.Google Scholar

⁶⁷ Parallels are hard to find, the closest being an inscribed stone ball from Corbridge (Arch. Ael. 3rd ser. vi (1910), 269), now lost. Mr. L. J. F. Keppie reports that there are no such marks on the c. 120 stone balls in the Hunterian Museum, mostly from Bar Hill, whose diameters range from 30 to 120 mm. The ‘numerals’ on the Wallsend stones may have recorded the work of different men: the balls are not finished uniformly, but there is similarity of finish between the two stones in each pair with the same mark.Google Scholar

⁶⁸ JRS lvii (1967), 208, No. 30aGoogle Scholar; JRS lviii (1968), 211, No. 45c. The stamp is too damaged for certainty.Google Scholar

⁶⁹ Arch. Ael. 4th ser. iv. (1927), 183, and pl. xxxvm, 1B.Google Scholar

⁷⁰ For a similar statement of the obvious, cf. mortario scratched before firing on the side of a samian mortarium from Corbridge (Arch. Ael. 3rd ser. ix (1913), 270, No. 8). No other nomen ends in- urelius (unless an unattested variant of Turellius/Turrelius), but Aurelius is comparatively rare until the accession of Marcus Aurelius (A.D. 161), and its occurrence on a Hadrianic sherd in a Hadrianic context is noteworthy. In so far as the sherd can be reconstructed, Aurelius seems to have been used without either praenomen or cognomen, which is certainly most unusual after A.D. 161.Google Scholar

⁷¹ For wine and honey (separately) in the Roman military diet, see R. W. Davies, Britannia ii (1971), 131.Google Scholar

⁷² Sent for study by Professor J. J. Wilkes. For a personal name IVLIVS[… at the military tilery at Holt see EE vii 1133 (from Chester), and for a complete example Grimes, Holt, 141, fig. 59, 16.

⁷³ During excavation for the Department of the Environment and the University of New-castle, directed by Mr. C. M. Daniels, who made the fragment available, and by Mr. J. P. Gillam. It will go to the Housesteads Museum. For the site, see p. 309.

⁷⁴ Carlisle: CIL vii 1248 (Hodgson, J., Hist, of Northumberland, Pt. 11, vol. iii (1840), 221)Google Scholar; J. C. Bruce, Lapidarium Septentrionale (1875), No. 501, illustrated pp. 18 and 254. Bainbridge: JRS xlviii (1958), 154, No. 29, where R.P.W. now considers the difference in letter-height is due only to shrinkage. The die consists of an M C. 65 mm high, with boldly serified verticals, the loop of p ligatured to the right-hand vertical, and the central v set unusually low and interrupted at the foot.Google Scholar

⁷⁵ For the three types see Wright, , Britannia vii (1976), 232.Google Scholar

⁷⁶ Sent by Dr. A. W. J. Houghton for the Shropshire Roman Research Group. Other finds included unguentaria and a fragment of a bronze circular mirror. For the work on the road see Britannia vi (1975), 248.Google Scholar

⁷⁷ Excavations by the Chichester Excavation Committee for the Department of the Environment were directed by A. G. Down. For the site see Britannia ii (1971), 284–5. For monumental inscriptions from Chichester, see above Nos. 3–5.Google Scholar

⁷⁸ For Roman weights see F. Hultsch, Griechische und römische Metrologie, ed. 2 (Berlin 1882), especially Table xiii.

⁷⁹ I. Kajanto, The Latin Cognomina, 240, points out that, of the nine examples of FLAOCINUS/A in CIL, six come from Spain. At the end of l.1, two vertical strokes have been traced between the C and the edge of the weight. They do not appear to be the symbol for a quadrans which, as three unciae, was =—.

⁸⁰ For the excavations of 1973 see Britannia v (1974), 457.Google Scholar

⁸¹ During excavation for the Department of the Environment directed by Dr. D. J. Breeze, who sent the sherd. For the site see Britannia vi (1975), 228, and p. 302 above.Google Scholar

⁸² During excavation for the Department of the Environment directed by Mr. L. J. F. Keppie, who provided a rubbing, and by Dr. J. M. MacKenzie. The sherd is in the Hunterian Museum. For the site see Britannia v (1974), 163 ff., and p. 304 above.Google Scholar

⁸³ During excavation for the Department of the Environment directed by Mr. P. J. Casey, who made the fragment available, and by Mr. J. L. Davies. The final letter is either o or c. There may be a space between AB and so which would make ab a preposition; otherwise, this may be a record of work done (absolvi, etc.), but there are other possibilities.

⁸⁴ Information from Dr. W. H. Manning who directed excavations for the Department of the Environment and University College, Cardiff. For the site see Britannia v (1974), 401.Google Scholar

⁸⁵ Pelveis: In early Latin the dipthong ei may, strictly speaking, only be written for long i; here it takes the place of a short i probably as a hypercorrective archaism. This graffito indicates that W. Hilgers was wrong in concluding (Lateinische Gefässnamen, BJ Beiheft xxxi (1969), 73), that the word pelvis was not used for mixing-bowls, which were always called mortaria (cf. EE ix, p. 674 No. 1354 b, a Drag. 45 from Corbridge with a graffito MORTARIO, an example not cited by him).Google Scholar

⁸⁶ Annual Report of the Council of the Society of Antiquaries of Newcastle upon Tyne for 1975, 1976, p. 14.

⁸⁷ Wright, and Jackson, , Antiq. Journ. xlviii (1968), 296.CrossRefGoogle Scholar

⁸⁸ Dr. A. W. J. Houghton sent this correction to R.P.W. (9 October 1973) and published it in WMANS No. 16 (1973), 17; see Britannia v (1974), 429Google Scholar. G. Webster, The Cornovii (1975), 114, adopted it. The site lies at grid ref. SJ 569091, east of the counterscarp of the town wall about 40 m south-east (not west) of Horseshoes Lane or Watling Street in Field 394 (on 25-in. O.S. map) where other tombstones have been found. For plans see Kenyon, Kathleen M., Archaeologia lxxxviii (1940), pl. LXVIII, and Atkinson, Wroxeter, 329 fig. 50, pl. 72. The distance to north-east of the north-east angle of the Forum should be corrected from 575 to about 600 yards (549 m).Google Scholar

⁸⁹ Notebook labelled on the spine ‘Roman and Saxon 36’, compiled by A. W. Franks, first Keeper of the Department of British and Medieval Antiquities.