Browsers through Collingwood and Wright's The Roman Inscriptions of Britain will be familiar with two short sections at the back labelled Aliena and Falsa, their catalogue entries suffixed with an asterisk to indicate their anomalous nature. The former consist of authentic Roman inscriptions from other provinces found in Britain, and the latter a selection of modern Latin inscriptions, some intended to amuse and others to bemuse. Amongst the latter is RIB I, 2334*, allegedly from an altar, now lost, found at Chichester in the middle of the seventeenth century dedicated by Sallustius Lucullus, a governor of Britain in Domitian's reign, who is independently attested only by Suetonius. At the time of the supposed discovery of the inscription, and even when RIB I was first published in 1965, there was no obvious context for a provincial governor at what was to become one of the lesser civitas capitals of Britain, but the discovery and excavation of a potential governor's palace at Fishbourne in the 1960s changed that.Footnote ²

When first considering the inscription about twenty years ago with a view to proposing an alternative occupier for Fishbourne Palace, I had a vague thought of publishing a note on it but reluctantly concluded that the inscription was most likely a hoax. As it was a long time since anyone had believed otherwise, reiterating this in print hardly seemed newsworthy, and the manuscript lay in a desk drawer for many years. All changed with Miles Russell's Roman Sussex (Reference Russell2006), for a great deal of his speculation about the early history of the province rests on his belief that this inscription is valid and commemorates a provincial governor descended through Amminus from the native British king Cunobelinus.

The inscription was supposedly found during or shortly after 1658 and its text communicated to Samuel Woodford by a certain Henry Babington. Woodford collected inscriptions and the Bodleian Library at Oxford has his manuscript collection of 103 such inscriptions.Footnote ³ All the entries in his manuscript can be independently verified apart from four, of which the Chichester inscription is one. Doubts about the Chichester inscription do not stem from anything intrinsically improbable in the text but from the fact that Henry Babington's other contribution to the corpus, RIB I, 2335*, the text of another altar of unstated provenance, seems to be an undoubted fake. The deities on it are all listed in Camden's Reference Camden and Holland1637 edition of Britannia apart from the first (Tarami), which occurs with slightly different spelling on an altar found in Chester in 1653 (see discussion below). The dedicator, C Verius Fortis, is supposedly tribune of the First Cohort Aelia Brittonum, which is otherwise unrecorded.Footnote ⁴ Of Babington himself, if he existed, nothing seems to be known. His surname recalls the Babington plot against Queen Elizabeth I, but that family was from Derbyshire.

RIB I, 2334* reads as follows:

The first line is a standard abbreviation. Line 2 is more interesting, for the formula pro salute domus divinae occurs in another Chichester inscription, the so-called Cogidubnus stone.Footnote ⁵ Pro salute followed by the name of a specific emperor is attested elsewhere in Britain,Footnote ⁶ but is more commonly followed by a vaguer phrase such as domini nostri, Augustorum nostrorum. Line 3 should in theory have been erased since following his assassination in a.d. 96 the Senate had issued a damnatio memoriae requiring Domitian's name to be deleted from public works. The dedicator, Sallustius Lucullus (line 5), is attested independently as a governor of Britain only in Suetonius,Footnote ⁷ where we learn he was put to death by the emperor for allowing a new type of lance to be named ‘the Lucullan’ after him. His praenomen Caius rests on this inscription alone. Birley gives his dates as a.d. 85/96, and RIB I, c. a.d. 93.Footnote ⁸ The Chichester inscription, if genuine, is the earliest example of a dedication by a governor of Britain and one of only seven examples of acts by provincial governors.Footnote ⁹ Lines 6 and 7 with the formula Leg Aug Pr Pr Prov Britanniae are not paralleled exactly in Roman Britain, though two early third-century inscriptions, RIB I, 1205 (Whitley Castle) and 1706 (Vindolanda), may provide examples of a governor of Britain specifying his province (which was obvious) within Britain, but both are very much restored. Provinciae is usually abbreviated as Provinc,Footnote ¹⁰ but was perhaps further shortened to Prov here because of lack of space. It is sometimes abbreviated to Pr,Footnote ¹¹ but a line reading Pr Pr Pr Britanniae would have been challenging.

If the inscription is a forgery, there is a good possibility that it is modelled on one or more Romano-British inscriptions known before c. 1658 and, for preference, published. For example, the latter part of the Chichester inscription, in abbreviated form LEG AVG PR PR PROV BRITANNIAE, is echoed closely by RIB I, 335 from Caerleon, found in 1602 and published by Camden,Footnote ¹² which reads LEG AVG PR PR PROVINC CILIC and records the appointment of a probable legate of the Second Legion to the governorship of Cilicia. The only peculiarity is the preference of the Chichester inscription for PROV rather than PROVINC. There is no extant British example of an inscription to Domitian as emperor,Footnote ¹³ but the formula could have been borrowed from coins with the case-ending suitably modified.Footnote ¹⁴ The formula PRO SALUTE followed by the name of a specific emperor is, as noted, rare. Several potential models can be disregarded since they were not discovered until after the Chichester inscription's alleged appearance,Footnote ¹⁵ but Caerleon again provides a possible model in RIB I, 326 found in 1603 and also published by Camden.Footnote ¹⁶ Caerleon can also supply an example of the very common abbreviation IOM,Footnote ¹⁷ but not until ‘a few years before 1835’, which rules it out. A hoaxer with the 1607 edition of Camden could, however, have used an altar from Old Carlisle as a basis for the abbreviation,Footnote ¹⁸ or another inscriptionFootnote ¹⁹ from the same site if equipped with Camden's fifth edition.Footnote ²⁰ The concluding line of the inscription, VSLMQ could, apart from the final letter, be ‘borrowed’ from the Chester inscription found in 1653Footnote ²¹ (as could IOM in line 1) or an inscription from Maryport, published by Camden.Footnote ²² Another inscription from Maryport is a further possibility,Footnote ²³ as are other examples from Binchester,Footnote ²⁴ Risingham,Footnote ²⁵ and Castlesteads.Footnote ²⁶ As the final Q is unparalleled in Britain it may be ‘borrowed’ from a continental original,Footnote ²⁷ or be a quirk of the hoaxer.

The case against the authenticity of the Chichester inscription is, therefore, a strong one. Virtually all its elements could have been culled from published or otherwise known examples by a mischief-making antiquary of the period, and there are two possible ‘mistakes’ — the final Q of line 9 and the abbreviation PROV in line 7 — which add to the doubts. In addition, acts recorded by provincial governors are rare, and no other governor is attested by inscription before the second century, other than Agricola in an inscription from St Albans.Footnote ²⁸ The survival of the dedication to Domitian is also suspect, unless it is assumed that the stone had been buried before his death and damnatio, thus escaping erasure.

Can anything be proposed in its defence? First, it does not tell us anything we did not know from Suetonius (i.e. that Lucullus was a governor during Domitian's reign), whereas it might be supposed that one of the purposes of a hoax would be to impart new (and misleading) information. Second, it in a sense anticipates the Cogidubnus stone and Jupiter column base later found in Chichester, both of which name the emperor's family as beneficiariesFootnote ²⁹ and comprise two out of only fifteen inscriptions in southern Britain to do so.Footnote ³⁰ Third, Chichester had gained ‘the epigraphic habit’ at an early date,Footnote ³¹ so its existence is not anomalous. Fourth, the only other Lucullus known from inscriptions in Britain is also from Chichester;Footnote ³² was Lucullus Ammini fil(ius) so named after the governor because the latter had spent some time in the Chichester area?

If Lucullus was in or around Chichester, perhaps as early as a.d. 85/86, it is possible to speculate on his purpose. The Cogidubnus stone, as has been pointed out more than once, must date either no later than the end of Nero's reign (a.d. 68) or post-date the accession of Titus in a.d. 79 (or conceivably later still) because Vespasian, in the intervening period, could not claim divine ancestry.Footnote ³³ The phrase domus divina used on the inscription could only be explained as a faux pas if the inscription is Vespasianic. There is no reason why Cogidubnus should not have lived into the mid a.d. 80s if he became king as a young man in a.d. 43 or soon after. If, as some have argued,Footnote ³⁴ the original excavatorFootnote ³⁵ dated the supposed palace at Fishbourne too early, ascribing this degree of longevity to Cogidubnus circumvents the problem of association. Had Cogidubnus died in c. a.d. 85, it could explain the presence in Britain at about that time of C. Salvius Liberalis and L. Javolenus Priscus, two noted jurists whose legal services could have been required to effect a rather smoother absorption of the native kingdom than had been achieved on the death of the Icenian Prasutagus. Lucullus, representing the heir, Domitian, may have occupied the palace at Fishbourne, and this may have marked his downfall if the reason put forward for his execution by Suetonius is not the literal truth (or only a small part of it) and really means ‘he usurped the emperor's privileges’.

It may make a good story, but the balance of probability is that RIB I, 2334* is, as many have suspected, a forgery, an antiquarian joke which misfired or perhaps a malicious attempt to deceive and thus, perhaps, discredit a fellow antiquary. This would be a good point to reiterate Russell's view as to why the stone should be considered genuine, but in practice all he has to say is that there is nothing in the wording, construction or phraseology ‘that screams “hoax”’,Footnote ³⁶ a point which, with reservations, we have already conceded. If the inscription is indeed a compilation from a variety of epigraphic sources, there is no reason why its components should not be valid. The real problem with the inscription is one of guilt by association, since Babington's other supposed contribution to Woodford's corpus, RIB I, 2335*, is undoubtedly false. It is unprovenanced and reads as follows:

The first two lines are reminiscent of an altar found in Foregate Street, Chester in 1653Footnote ³⁷ that begins I O M TANARO and is followed by details of the dedicator, the princeps of the Twentieth Legion. Tanarus is a Celtic thunder god sometimes conflated with Jupiter (in modern Irish toirneach means ‘thunder’). The form Tanaris occurs in Lucan's poem Pharsalia (lines 444–6) and variants of the name are found as far apart as Britain and Dalmatia.Footnote ³⁸ Forms resembling that in RIB I, 2335* include Taranucnus, Taranucus and Taranus, but all are continental and Tarami is most probably a mis-hearing of the form found at Chester. All the other deities can be found in the 1637 edition of Camden's Britannia. The Celtic deities Mounos and Mogons are found at Risingham on two separate altars discovered before 1599.Footnote ³⁹ In both cases the deity's name is followed by CAD…, possibly a territorial epithet, perhaps indicating that Mounos and Mogons are the same god. If this is so, his commemoration twice on the same altar is nonsensical. His/their usual distribution in Britain is along Hadrian's Wall and in continental Europe in Germany and eastern Gaul. The name seems to mean ‘The Great One’ (mago-, mogo-Footnote ⁴⁰ ). RIB I, 1226 also gives a would-be hoaxer the abbreviation VS that, whilst not uncommon, is much more usually extended to VSLM or similar. Belatucadrus (this is the commonest spelling of the name) enjoys a similar northern distribution in Britain and in five cases is equated with Mars.Footnote ⁴¹ The very weathered altar to the god from Kirkby Thore,Footnote ⁴² now at Appleby, was discovered in or before 1601 and is a likely source for RIB I, 2335*. The Syrian Goddess is commemorated at Carvoran on an altar discovered in 1599 or earlier.Footnote ⁴³ A second dedication to her was found nearby in 1816,Footnote ⁴⁴ and a third in the vicinity of Catterick shortly before 1823,Footnote ⁴⁵ both of them too late to be models for RIB I, 2335*. Fortuna and the Mother Goddesses are too widespread to require further comment. Cohors I Aelia Britonnum occurs nowhere in Britain other than, allegedly, on RIB I, 2335*, but the formula cohors I followed by the name of a province is common enough. Cohors I Aelia Brittonum, if it existed, would be unlikely to have served in its home province, nor would it feel the need to placate goddesses from distant Syria if the name of the unit gave any clue to its ethnic origin. Another objection is the form ‘Britonnum’ (with double N) which is not attested: properly it should be ‘Brittones’ (sometimes ‘Britones’), but the form ‘Britonnum’ would have been suggested to a forger by ‘Britannia’ and ‘Britanni’. Tribune C Verius Fortis is equally unattested elsewhere. In brief, the whole inscription could have been, and probably was, cobbled together from the first two lines of a then newly-found inscription from Chester and a selection of others from Hadrian's Wall and its vicinity, its purpose being to suggest that a locally-recruited military unit was stationed in its home province.

A final question must be whether RIB I, 2334* is genuine despite its companion being a hoax. Could Babington himself, if he existed, have been ‘fed’ one authentic inscription, and another composed recently in order to deceive? It could be argued that RIB I, 2334* is a more convincing forgery than RIB I, 2335*, although they have two things in common. One might be called ‘patriotism’, as RIB I, 2334* records a governor of Britain and RIB I, 2335* a British auxiliary unit. Overall, references by name to Britain and its inhabitants are fairly uncommon on insular inscriptions,Footnote ⁴⁶ so the notification of two by the same individual is suspicious. Their other common feature is that both have disappeared. The vast majority of the known inscriptions of Roman Britain survive. It is distinctly odd that these two have not only disappeared, but are attested in one source only.

Having accepted RIB I, 2334* as genuine, Russell goes on to reinterpret RIB I, 90, an incomplete altar found in the middle of Chichester in 1823. Now illegible, it was drawn for publication in RIB I in 1954.Footnote ⁴⁷ By that time all but three letters (emboldened below) had been lost and the text was restored from a photograph. It is nevertheless perfectly credible and is usually read as follows:

It may be expanded as Genio Sacrum Lucullus Ammini filius de suo posuit and translated as ‘Sacred to the Genius (of the place) Lucullus son of Amminus set this up from his own resources’. Another possible extension of the first line is Genio Sancto ‘To the holy Genius’, which means much the same. The altar as a whole is typical of Roman care to placate the omnipresent spirits whose names they did not know, and those responsible for the dedication might be inferred to be unfamiliar with the locality and its spirits. Soldiers on hostile frontiers were amongst the main dedicators of such altars.Footnote ⁴⁸

The surviving part of the altar measures 0.53 m by c. 0.96 m, and its original height may be estimated at about 1.15 m. The sides, as far as is known, were plain, but on similar altars the Genius is frequently depicted as a part-draped female with cornucopia and patera. Some care seems to have been taken in laying out the lettering. The first three lines are justified on the right side and to achieve this there is a wider gap than would be expected between GENIO and S. The left side is not apparently justified, although the first and third lines begin at the same point, and it rather looks as if in front of LVCVLLVS on line 2 there may have been a single letter and a gap. DSP in the final line is symmetrical about the centre line of the altar. The style of lettering has been dated to the end of the first century,Footnote ⁴⁹ but no reassessment is now possible. The restriction of the text to the upper half of the altar may mean that it was designed for a location where any lettering on the lower part would have been obscured.

Russell, having accepted RIB I, 2334* as genuine, notes that the only two occurrences of the name Lucullus in Britain are at Chichester.Footnote ⁵⁰ This, he maintains, is an improbable coincidence and both are most likely the same person. The true reading of line 1, he suggests, may be GENIO C S the ‘C’ occupying a spot damaged before the stone was photographed, the correct reading being:

and meaning ‘To the Genius, Caius Sallustius Lucullus, son of Amminus, set this up from his own resources’.

This suggestion is, the author believes, quite improbable for a number of reasons. To begin with, the altar is a modest example of its class. Although quite tall, it is hardly the sort of thing a first-century governor would put his name to, or indeed compound the insult by claiming to have paid for it from his own deep purse unless he intended it as some kind of snub to the deity, and even governors needed friends in the supernatural world. Normal practice would be for the deity to have his own line on the inscription, and the ‘S’ is almost certainly for sancto, sacrum or similar. The dedicator's name might begin on line 2 and occupy as many lines as necessary to identify him. There is no obvious pressure of space to justify condensing the writing as the whole of the text is confined to the upper half of the altar. The tria nomina Footnote ⁵¹ are not usually abbreviated in the way that Russell suggests. Caius Sallustius Lucullus might well occur on inscriptions as C Sallustius Lucullus or even C Sallustius L, but C S Lucullus is most improbable. Another problem is that the Lucullus on this altar, if provincial governor, would have said so, but he instead identifies himself as the son of someone with a Celtic name. In Roman conventions of naming, a Roman citizen who was the son of a peregrine would surely have used the sequence ‘C. Sallustius Ammini f. Lucullus’. Russell should have explained, or at least hypothesised, how the son of a client king gained Roman citizenship as ‘Sallustius’; by all analogies, he would have borne an imperial gentilicium, presumably ‘Tib. Claudius’.

Amminus is not found on any other inscribed stone in Britain. The name recalls, and may be identical with, the name spelled Adminius by Suetonius. This native prince, having been banished by his father the British king Cunobelinus, fled to the Continent and surrendered to Caligula with a few followers, allowing the emperor to claim that the whole island had yielded to him.Footnote ⁵² He may also have been one of the ‘certain deserters’ whose return the Britons demanded from Claudius.Footnote ⁵³ His coin distribution puts Adminius in east Kent, where there are coins inscribed Amminus, A and Am, and he seems to have ruled as a sub-king in the 30s before his surrender to Caligula at Mainz.Footnote ⁵⁴ There are two identical silver minims from excavations in north-west Chichester marked ‘A’ within a star and with a bird holding a bunch of grapes on the reverse,Footnote ⁵⁵ one from Hayling IslandFootnote ⁵⁶ and a fourth found in 1938 from between Harting Down and Beacon Hill,Footnote ⁵⁷ now lost, which, it has been suggested, are stylistically close to the Kentish issues and may reflect Amminus in exile from Kent before his eventual escape overseas.Footnote ⁵⁸ Yet these hardly make a cluster, for Hayling Island is about 15 km from Chichester and South Harting a similar distance from both, and we have no certain evidence of a focus at Chichester until some time after a.d. 43 when the ‘new town’ Noviomagus was founded.

Russell's contention, in brief, is that RIB I, 90 commemorates King Amminus and his son Lucullus, both of whom had fled to the court of Caligula in a.d. 40, the latter (a thoroughly Romanised grandson of Cunobelinus) returning as governor of the province under Domitian. Henig and NashFootnote ⁵⁹ originally proposed the idea that Lucullus was the son of an Iron Age prince, but his equation with Domitian's governor is Russell's own contribution. My own view is that a monument erected by two such distinguished individuals would have been more worthy of their status and, most importantly, have trumpeted the high office of the donors. Even if they were ‘between jobs’ (e.g. an ex-kinglet and his yet-to-be-appointed son), the addition of VIR CLARISSIMVS or similar would have served to distinguish one or other from the normal run of men. In other words, the ‘inscription’ that puts Sallustius Lucullus at Chichester is a hoax, and the genuine RIB I, 90 commemorates two relatively humble men with distinguished namesakes.