Bryn Mawr Classical Review

The first volume of the Roman Inscriptions of Britain, subtitled "Inscriptions on Stone", appeared in 1965.¹ So monumental was the task of processing the ever-increasing corpus of inscriptions that, during its compilation, a cut-off date of 1954 was imposed. The volume finally comprised 2,400 items, each one illustrated, transcribed and translated (in marked contrast to other contemporary epigraphic corpora). As for the hundred or so items that had accrued in the decade prior to publication, "the new material", wrote R.P. Wright at the time, "is reserved for inclusion in some supplement". That supplement is the volume under review.²

As second-named author, R.P. Wright provides a link with the original Roman Inscriptions of Britain project, which he joined in 1938. The tradition had already been established, in 1921, of publishing new epigraphic discoveries annually in the Journal of Roman Studies. Wright continued the work begun there by R.G. Collingwood, and in 1970, he transferred the annual report to the new journal Britannia, now with the assistance of Mark Hassall and (from 1974) Roger Tomlin. (Wright stepped down in 1976, and died in 1992.)

Thus, all 550 items in RIB III have seen previous publication in JRS or Britannia, except for RIB 3030 (a single-letter fragment from Reculver), 3109 (a tombstone fragment from Caerleon), 3178 (a two-letter fragment from Lincoln), 3370 (an incised cross from Vindolanda), 3550 (a non-Roman fragment), and 3019 (a tiny fragment previously believed to belong to 3016).

A set of concordance tables assists in matching the new RIB numbers to each item's primary publication in JRS or Britannia. However, cross-references to L'Année épigraphique for the 200-odd items noted there would have been useful, too. Many scholars have been conscientiously utilising the internationally recognised AE numbers, in preference to the more cumbersome (and accident-prone) JRS or Britannia references. As an example, prior to the publication of this volume, scholars wishing to identify the Westerwood altar had to choose between "JRS 54 (1964), 178, no. 7", or the more concise "AE 1964, 175". Thankfully, they can now use RIB 3504; but those wishing to backtrack from AE 1964, 175 will not have an easy job.

Having said that, RIB III is a model of its type. First of all, Tomlin deserves gratitude for his exemplary front matter, which prospective editors of catalogues, corpora and companions would do well to emulate. Readers are left in no doubt concerning the various conventions used in the volume, and the abbreviations have been expanded usefully (rather than minimally, as is all too often the case).

Each item has been photographed and drawn (wherever possible), and the text has been transcribed, not only literally, letter by letter, but also as a reconstructed text, with abbreviations expanded and lacunae restored, whenever it is safe to do so. Most appealing of all, the convention of translating each item, introduced in volume I in order to "make the work more useful to the general reader", has been retained.³

As in RIB I, each item is accompanied by a short discussion, sometimes explaining the content of the inscription. RIB 3016 (eight fragments from Southwark preserving a list of Roman names) is a good example, where Tomlin's 3-page discussion places the text into the context of known lists of legionaries. Other items receive more cursory treatment. The non-specialist reader consulting RIB 3507 (Hutcheson Hill) may be confused to learn that this distance slab from the turf wall of Antoninus Pius "was originally mounted in masonry". Of course, the presence of cramp-holes simply implies that the slab must originally have been held in a masonry frame, itself presumably set into the turf wall.⁴

Also, the basis for Tomlin's expansions and restorations could, on occasion, have been made plainer. For example, on RIB 3445 (tombstone from Birdoswald), the surviving m[. . .]p.f has been expanded to read m[il(es) leg(ionis) VI Vic(tricis)] p(iae) f(idelis), without explanation. Tomlin's certainty presumably rests on the fact that no other military unit in Britain carried those epithets.

In the past, it has been claimed that Britain's epigraphic record is not of the highest standard, based on the observation that items whose relevance extends beyond Britain are scarce. Admittedly, a large proportion of the inscriptions in RIB III (perhaps as much as 50%) consists of fragments (like RIB 3445, mentioned above), unintelligible letter sequences (like RIB 3033), or single signs (like RIB 3304, a building stone from Hadrian's Wall marked only with an X).

However there are highlights, too. The important tombstone of a legionary veteran from Alchester (RIB 3121) is one of these.⁵ Here, however, Tomlin accepts the expansion of the second-last line as HE(res) C(uravit) without commenting on the peculiar letter spacing. For, as the version recorded as AE 2005, 898 implies, the equidistant H E C are clearly a botched attempt at H(eres) F(aciendum) C(uravit).

Other highlights include the well-known Verulamium inscription (AE 1957, 169), discovered during construction work in 1955 and now re-christened as RIB 3123. The Hutcheson Hill distance slab (AE 1971, 225) becomes RIB 3507, and the altars from Westerwood (AE 1964, 175) and Old Kilpatrick (AE 1971, 226) become RIB 3504 and 3509. Curiously, Tomlin discusses the latter in terms of "Antonine I" and "Antonine II", terms which have largely been discredited in favour of a single Antonine occupation.⁶ (If he has something else in mind, he has not explained it.)

RIB 3364 offers a revised version of a tombstone from Vindolanda, of which only the left-hand side survives.⁷ Tomlin takes a minimalist approach, assuming that only half of the text is missing, whereas the reconstruction originally suggested by A.R. Birley assumes about two-thirds. And, in fact, Birley's extrapolation of the deceased as a centur[io leg(ionis) . . . praepositus coh(ortis)] is, perhaps, more attractive, since Tomlin's version (centur[io cohortis]) implies that a fairly elaborate monument was erected by an auxiliary centurion, rather than Birley's legionary. And one may be entitled to expect a mention of the theatre of war in connection with a man who was in bell[o . . . inter]fectus, which again requires a wider inscription.

Other items that caught my interest include:

RIB 3027, the dedication of a shrine in the headquarters at Reculver. The reading su[b A(ulo) Triar]io Rufino, enshrined in the literature as JRS 51, 1961, 191, no. 1 (whence AE 1962, 258), is now officially rejected, with strong support for su[b Q(uinto) A]r[ad]io Rufino and a consequent redating.⁸

RIB 3056, a welcome elucidation of the curious altar to Mercury discovered during excavations in Uley in 1978. The garbled dedication, reported as "]OVERNIVSEYXOSIFMVSLMT" in Britannia 12 (1981), 370 no. 5 (omitted from AE), can now be seen, from the excellent interpretative drawing here, to read simply [L]overnius pos(u)it v(otum) s(olvit) l(ibens) [m(erito)].

RIB 3072, an improved reading of the poignant tombstone of a slave boy, aged only 14, from Gloucester. It now seems that Martialis' master was named Gaius Cilonius, rather than the "--]lonius", originally given in Britannia 36 (2005), 474 no. 2 (whence AE 2005, 896, not noted by Tomlin, as it no doubt appeared after RIB III's cut-off).

RIB 3185, the almost complete tombstone of an auxiliary cavalryman, with splendidly detailed sculpture, from Lancaster, published as Britannia 37 (2006), 468, no. 3 (whence AE 2006, 750, not noted by Tomlin).

RIB 3299, a fragment of a dedicatory slab from Chesters, published as Britannia 36 (2005), 480, no. 8 (whence AE 2005, 923, not noted by Tomlin). The apparent reference to sy]mmachar[ii at Chesters fort on Hadrian's Wall will no doubt generate further speculation, and the consular date (AD 286) is noteworthy as the latest of its type in Britain.

And, most amusing of all, RIB 3358, a quarry-face graffito of a phallus, with the message H P III, inventively expanded by Tomlin to read h(abet) p(edes) III, "it is three feet long".

1.   R.G. Collingwood and R.P. Wright, The Roman Inscriptions of Britain, Volume I, Inscriptions on Stone (Oxford, 1965). Individual items are conventionally identified by the abbreviation RIB followed by the item number.
2.   In the meantime, Wright's promise to issue a Volume II containing the instrumentum domesticum ("personal belongings and the like") was fulfilled in 1990-1995, when RIB 2401-2505 appeared in eight fascicles. In Volume III, Tomlin has made the shrewd decision to begin numbering at 3001, thus assisting future researchers in locating the correct volume, when faced with an RIB number.
3.   By and large, the translations are accurate, as expected. But I noted RIB 3499 (lower half of an altar, Inveresk), where ex nuntio dic(atam) ar(am) pos(uit) is better translated as "set up the altar, which is dedicated as a result of a (divine) message", rather than the rather clumsy "(he) dedicated (this) altar as a result of a Message (and) erected it", offered by Tomlin.
4.   As explained (e.g.) by L.J.F. Keppie, Roman Inscribed and Sculptured Stones in the Hunterian Museum, University of Glasgow (London, 1998), 53.
5.   Fully published by Eberhard Sauer in Britannia 36 (2005), 101-133.
6.   N. Hodgson, "Were there two Antonine occupations of Scotland?", Britannia 26 (1995), 29-49.
7.   A.R. Birley, "A new tombstone from Vindolanda", Britannia 29 (1998), 299-306, whence AE 1998, 835.
8.   As already suggested by Richard Harper, Anatolian Studies 14 (1964), 166, which Tomlin notes in his exhaustive bibliography.