possible, milit(es) l[eg(ionis) or

Transcription

1 R. W. DAVIES, BA, PhD A LOST INSCRIPTION FROM AUCHENDAVY Auchendavy is justifiably famous for the four altars set up to a variety of deities by M. Cocceius Firmus, a centurion of legio II Augusta and former member of the Imperial Mounted Guard (Equites Singulares Augusti)^. It has also produced other one inscriptions; was seen in 1826 by Rev John Skinner, but was unfortunately soon lost. It is described by R. G. Collingwood and R. P. Wright fragment as a of a tombstone and they transcribe it<2): NO... MART. M(arcus) Al[..] P7cft>[r] mil(es) l[eg(ionis) II] Aug(ustae) TALAEEX... VLERNIS From the layout and shape of the stone it seems most unlikely that it was a tombstone. Moreover, the text seems to be enclosed by a circular feature; this looks very much like the victory wreath that on appears various military dedications, including several sites on the Antomne Wall*3). The four letters MART in line 2, regarded by Collingwood and Wright as corrupt, are surely the name of the god Mars in an oblique case, presumably the dative; MAL must be left on one side for the moment. In line 3 Victo could be either Victoria or Victor. It is known from the Roman military religious calendar (Feriale Duranum) that the troops officially worshipped many deities, including Mars and Victory, and this is confirmed by inscriptions*4); thus Victoria Victrix appears on one altar dedicated by Cocceius Firmus, Mars and Victoria on another*5'. Cohors I Baetasiorum was awarded en bloc Roman citizenship for its distinguished conduct (oh virtutem et jidem) in the campaigns during the reoccupation of Scotland*6*. Consequently, the unit set up three altars instead of the customary one on 3 January each year, to Mars Militaris, to Victoria Augusta, and to Iuppiter Optimus Maximus(7). The mark on the Auchendavy inscription after Victo is presumably the remains of the letter 'R' rather than a leafstop. The following letters are expanded by Collingwood and Wright mil(es) l[eg(ionis) as II, though there seems to be sufficient recorded by Skinner to read le[g(ionis). However, as this is probably not a tombstone, it must remain very possible that the expansion should be in the plural as mil(ites) le[g(ionis), or, if ligatured as is quite possible, milit(es) l[eg(ionis) or milite(s) l\eg(ionis). The first three letters of the next line are clearly Aug(ustae). The rest of this line is regarded by Collingwood and Wright as corrupt and they print TALAEEX. This contains, however, the Latin word alae, which seems clearly from the drawing to have a space on either side of it because it was a separate word. The 'T' will be a misread ligatured et with the 'E' reversed, connecting the legio II Augusta and the auxiliary ala. The ex[ at the end of the line will have contained information on the identity of the cavalry regiment. The diploma of 122 lists no less than thirteen alae, but none of them has a name or title beginning ex on this or on any other RIB ILS 4831; E. Birley, Roman Britain and the Roman Army (1953), , hereinafter cited as Birley (1953). Possibly the fragmentary altar to Silvanus (RIB 2178), found with them and the remains of a stone bust and iron hammers, belongs to Firmus; cf. Sir G. Macdonald, The Roman Wall in Scotland (2nd ed., 1934), 285-9, , Pl. LIV. <2> RIB 2179 CIL VII, 1116; Macdonald, op. cit., 437-8, no. 58, fig. 53. W E.g. RIB 2163 (Croy Hill), 2208 (Old Kilpatrick), 2209 (Auchendavy?). Cf. also RIB 1093, PI. XIV and Note the victory wreath (presented by Victoria?) on the Hutcheson Hill slab; Britannia I (1970), , Pis. XVIII and XXXVIII L'Annee Epigraphiquc (1971), 225. <4> R. O. Fink, Roman Military Records on Papyrus (1971), no. 117; Iuppiter Victor, Mars Victor, and Victoria were all worshipped on 3 January, Mars Victor on 1 March also. RIB 1138, , <5> RIB 2176, <«> RIB 2170 (Bar Hill) from the HQ; cf from the aedes there. The title civium Romanorum is not attested on any diploma, including that of 135 (CIL XVI, 82). O RIB 830, 837, 838, 842, 843; R. W. Davies, "Cohors I Hispanorum and the Garrisons of Maryport', in TCWAS, NS 77 (1977), forthcoming. 103

2 epigraphic record*8'. Consequendy, one must assume either a hitherto unattested ala in Britain or find a suitable context and explanation for ex. As ex follows immediately after alae, it can hardly be the preposition followed by ex a as placename, the identity of the unit must have occurred between ala and the place. The same reasoning rules out a reference to, for example, ex[peditio. Although the existence of a cohors Aelia expedita known*9', presumably is a specialist cohort of very lightly equipped and highly infantry, mobile the alae were already light and mobile forces, and consequently would not need a title such as ala expedita^. In short, there is only one word ex beginning used in military contexts that can possibly fit: exploratores 'scouts'*11'. There is abundant evidence that troops regularly detached from were their auxiliary units in peacetime and in war for scouting duties on and beyond the frontier*12'; at times units of scouts (numeri exploratorum), apparently of irregulars rather than auxiliaries, were organised*13'; they were mainly under the command ofpraepositi, though occasionally praefecth14). of However, there is good epigraphic evidence to show that some of these could be upgraded to the status of ala exploratorum. Inscriptions from Mauretania Caesariensis reveal the presence in the second quarter of the third century of the ala Pomariensium under the command exploratorum of a praefectus^l; the last word is the adjective derived from the Roman name of Tlemsen and shows that the unit was called 'the regiment of scouts based at Pomarium'. Such a placename not infrequently occurs in the title of exploratores^. This is the only ala attested in the entire Roman Empire at any time in the Principate with a title beginning ex *17'. Inscriptions from Germany refer to a numerus exploratorum Divitiensium, that is a unit of scouts based at the key bridgehead fort at Divitia on (Koln-Deutz) the east bank of the Rhine opposite the capital at Cologne*18'. One inscription, however, refers to T. Flavius Salvianus, expraefecto exploratorum Divitiensium militia quarta^19*; that means that the numerus had subsequently been upgraded to the status of an ala and apparently that it was regarded as the equivalent of an ala milliaria, the elite unit of a province*20'. Scouts, particularly mounted scouts, were expected to be mobile; an inscription from Auzia records that this numerus was at one time transferred from its to Mauretania*21'. It thus base in Germany seems inevitable to suppose that the Auchendavy inscription reveals the presence of an ala exploratorum based on the Antonine Wall; how long it stayed there, is not known; it is not impossible that it was subsequently transferred to Mauretania where it reappeared as the ala exploratorum Pomariensium^. This is, then, a dedication made jointly by the milites of a detachment of legio II Augusta and an ala exploratorum. However, as there is considerable uncertainty as to the length of these lines, it must remain a possibility that the title Augusta was not part of the title of the legion but of an auxiliary unit. It is apparently not borne by any cohors attested in <8> CIL XVI, 69. <9' C. Cichorius, 'Cohors', in RE 4 (1901), (10> Expedita is never attested in the title of any ala anywhere; at this date there were no 'heavy' cavalry units in Britain, so there would be no need for an adjective 'light' for contrast. (11> Cf. the indices in Fink, op. cit. and ILS. <lz> M. Speidel, 'The Captor of Dccebalus: a New Inscription from Philippi', in JRS LX (1970), , hereinafter cited as Speidel (1970); L. Fiebiger, 'Exploratores', in RE 6 (1909), ; Fink, op. cit., nos. I, 2, 63. For the usefulness of exploratores cf. Vegetius III, 6. <13> E.g. CIL XIII, 7054 ILS 2632; VIII, 9059 ILS 2628; ILS 9186, 9187; RIB 1243, 1244, 1262; RIB 1270 ILS <14> Fiebiger, op. cit., <") CIL VIII, 9906 ILS 2634; VIII, 9907 ILS 4492; cf. VIII, 9908, 9909; L'Amtee Epitiraphique (1889), 54. <16> E.g. CIL XIII, 7495 \LS 9185; XI, 3104 ILS 276$; XIII, 6592 ILS See also below, notes 18-19, 21, <") Cf. C. Cichorius, 'Ala', in RE 1 (1898), 1258, and ILS index s.v. <18> E.g. CIL XIII, 7054 ILS <18> CIL XIII, 6814 ILS <20> E. Birley, 'Alae and Cohortes Milliariae', Corolla Memoriae Erich Swoboda Dedicata (Koln-Graz, 1966), It should be noted that there was no normal milliary ala in either German province, so this was a post of considerable importance; ibid. 57. <21> CIL VIII, 9059 ILS 2628; cf. VAnnie Epigraphique (1901), 59 bis ILS Cf. ibid. (1899), 5S ILS 9187a, also from Mauretania, for a splorator [sic] Batavorum, though such people came from the Lower Rhineland (cf. ILS 9186 from Roomburg). (22) The Moors caused considerable trouble (and reinforcements had to be sent to deal with them) in 145, 171, and 178; on the first occasion at least troops were sent from Britain, L'Annee Epigraphique (i960), 28; a vexillatio Brittonum Volubili agentium is attested in Mauretania in the later second century, ibid. (1920),

3 Britain*23' but no less than three British alae carried this title of honour: ala Petriana, ala Vocontiorum, ala Proculciana. The first two of these regiments seem to have had the title Augusta sandwiched between ala and Petriana or Vocontiorurn^; however, the last named was frequently called simply ala Augusta or ala Augusta ob virtutem appellata^. When this unit changed its title is not known for certain, except that it had taken place by the i8o's<26>; it is conceivable that the dedication was set up by milite[s alae] \ Aug(ustae) et alae ex[ploratorum, though the presence of members of legio II Augusta attested by other inscriptions in the vicinity would point to the dedication being more probably set up by a joint garrison of legionaries and cavalrymen'27'. It is not impossible that Pius decorated some of these alae with the title as Augusta, just he conferred the title Aelia on some cohorts and en bloc on cohors I Baetasiorum^. citizenship Unfortunately, the last line of the inscription remains difficult to interpret and no ready explanation springs to mind. It is not clear whether VLERNIS is one or more words, or whether some letters at the beginning have been lost. The recorded letters do not accord with any ofthe customary formulae or abbreviations to be expected at the end of an inscription*29'. Consequently, it seems most probable that these letters gave further information about the troops. An ethnic noun does occur after the title not normally exploratoresm, but the name of the place where they were stationed very often docs. The various examples from the rest of the Empire have already been mentioned*31'; the principle is very well illustrated by the British examples in the early third century. Inscriptions refer to the exploratores Habitancenses based at Habitancum (Risingham)*32' and the exploratores Bremenienscs based at High Rochester (Bremcnium)'33'; the venatores Bannienses are perhaps scouts based at Bewcastlc, best explained as who travelled widely*34'. Netherby, the fourth outpost fort, because it was the control centre its name from for the scouting patrols, changed Brocara to Castra Exploratorum the 'camp of the scouts'*35'. It might be easiest to suppose that VLERNIS is the placename for the base of the exploratores, perhaps in the locative case or in the genitive agreeing with alae, or, more probably, in the plural and abbreviated agreeing with exploratorum. Unfortunately, the names of only half of the forts along the line of the Antonine Wall are recorded in the Ravenna List*36'. If VLERNIS is (part of) the Roman name for Auchendavy, it does not readily suggest an identification with a site recorded in the Ravenna List, unless it is Litana, apparently meaning 'the expanse'*37'. However, the British and other examples show that the dedications could be made by exploratores some distance from their base, and it is not impossible that the on placename the inscription refers to somewhere else, presumably to the north. One may note as possibilities Ugrulcntum and the corrupt (23> However, cf. RIB 649 ILS 3598 and apparatus for a possible cohors I Augusta Bracarum in Britain, though the unit is never attested in diplomas. <24> RIB 957 (Carlisle). CIL XVI, 69; RIB 2121 (Newstead) of the Antonine I phase and therefore not the Auchendavy unit. <25> E.g. RIB 890, 893-7, 905, 907 (Old Carlisle). Cf. RIB 606 with JRS LIX (1969), 235, no. 4 VAnnie Epigraphiquc ( ), 292 (Lancaster). The earliest securely dated reference is RIB 893 of 188. Note that the longer non-prestigious titles Gallorum Proculciana seem to have been normally used in the diplomas as late as 146; CIL XVI, 69, 82, 88, 93. <27> Note from Shirva farm a dedication-slab from a vexillation (not the whole) of legio II Augusta and a tombstone of an other rank (RIB ); these suggest a garrison rather than merely a building party. It may be noted that it was the other legions who were responsible for building the Antonine Wall in this section (RIB 2173, 2184, 2185). <28> It was very probably Pius, not Hadrian, who gave the title Aelia to cohortes I Dacorum, I Hispanorum, I Classica; Davies, supra, note 7. For cohors I Baetasiorum, supra, notes 6-7. <29> E.g. VSLM, D, DD, F, P, REST. RE s.v. VL- is of no help. (30) As the examples in ILS show, it seems to occur only when the exploratores arc moved from one province to another. (31) Fiebiger (note 12), Supra, notes 15-16, 18-19, 21. <32> RIB 1235; for the locative, RIB (33> RIB 1262; RIB 1270 ILS I34) RIB 1905 ILS I owe this suggestion to Dr Brian Dobson; he argues that a placename would not be needed for hunters, that hunters are not attested in the auxilia, and that a unit of venatores is attested in the Notitia as low-gradetroops. A dedication to Silvanus, the god of the wilds, would be as suitable for scouts as for hunters. Notitia Dignitatum, Occ. XL 25 shows that a numerus exploratorum replaced the thirdcentury auxiliary cohort at Bowes; the outpost exploratores were subsequently replaced by areani, cf. J. C. Mann, 'The Northern Frontier after AD 369', GAJ 3 (1974), <35> Rav. 156; Itin. Ant. 467, 1. <36> Rav ; cf. I. A. Richmond and O. G. S. Crawford, 'The British Section of the Ravenna Cosmography', in Arch. XCIII (1949), 1-50, passim. <3') Rav. 198; Richmond and Crawford, op. cit., 38. IO5

4 Veromo, both sites north of the Antonine Wall, recorded in the Ravenna List*38*. One may also note placenames in England which could easily have reappeared in Scotland, such as Vertis, apparently referring to a 'turning' of a stream or Vernilis, 'flowing amid alders'*39'; the upper reaches of the River Kelvin in fact in a noticeable turn. flow close by Auchendavy This inscription seems to present much interesting information. The choice of gods can hardly be haphazard; there is a reference to deities representing war and victory, both suitable at any time for Roman soldiers and especially in the aftermath of the great campaign ofpius in the reconquest of Scotland that led to his sole acceptance of the acclamation of imperator for his entire reign<40>. Victoria appears on several of the distanceslabs from the Antonine Wall and on the one from Duntocher two Victories are flanked by Mars Victor and Virtus Augusta*41*. no could be the dative ending for the names of several deities (e.g. Iuppiter Dolichenus, Silvanus, Neptunus, Gen(i)us ligatured) or part ofjuno's name. MAL does not appear to be the beginning of the name of any deity, except for Malachbelos, a Palmyrene god, and the goddesses Malvisae*42'; the latter appear on a couple of military dedications set up on the Lower Rhine, one of which was dedicated to Diabus Malvisis et Silvano by an ordinarius Brittonum at Cologne*43*. For MAL it is perhaps best to assume that Skinner misread, for example, Minerva or Matres or Salus; perhaps this was the epithet Militaris, as on the dedications by cohors I Baetasiontm. Many of these deities were regularly worshipped by the Equites Singulares Augusti at Rome*44*. It is perhaps possible that MAL represents territorial some or other aspect of Mars, though there are few attested possibilities*45*. One may note that the town of Malva, perhaps Romula, gave its name to an area of the Lower Danube region*46*. This area provided the bulk of the Equites Singulares Augusti and Birley made a very good case for Cocceius Firmus coming from precisely that part of the Empire*47*. One may perhaps note also the altar found at Colchester to Mars Medocius Campesium, which may refer to the Campsie Fells; it was dedicated by a Caledonian*48*. The identity of the deity MAL must, remain unsolved; while a Celtic epithet is not unfortunately, impossible, the tenor of the inscription might suggest that it is a Roman deity that is involved. Clearly at least one deity must be sought to fill the lacuna in the present first line and there seems to be a line missing at the top of the inscription. For what it is worth, one may note that inscriptions set up by the Equites Singulares Augusti at Rome regularly have the sequence Iuppiter Optimus Maximus, Iuno, Minerva, Mars or Salus, Victoria; it is that this not beyond the bounds of possibility dedication was to deities particularly by favoured the Equites Singulares Augusti(i9K However, a dedication to Mars Militaris et Victor in this Scottish would be equally appropriate context*50*. A mixed garrison of legionaries and mounted scouts or of strike cavalry and mounted scouts*51* would be an ideal one for (38) Rav. 208 and 206; Richmond and Crawford, op. cit., 47, 49- (39> Rav. 64 and 8; Richmond and Crawford, op. at., 49. One should perhaps not forget Verterae (Valteris) meaning 'summit'; Rav. 127, Richmond and Crawford, op. at., 48. <40> Birley (1953), 32. Ibid. 100: 'all the Auchendavy dedications with the exception of that to Victoria Victrix occur regularly on the altars of the equites singulares'; the context of Pius's victory surely explains the exception. Perhaps note RIB 2122, ob prosperos eventus. <"> RIB 2193 and PI. XIX, 2200 and PI. XIX, 2208; the victorious campaigns are also depicted on the Bridgeness (RIB slab 2139 and PI. XVIII). For Victoria (rather than Britannia) on the Hutcheson Hill slab, see note 3. Cupids occur on RIB 2198 and 2206 (cf. also 2163), which could conceivably be misinterpreted Victories; cf. Current Archaeology 15 (July 1969), 108 JRS LIX (1969), PI. XV, 1. <42> RE s.v. Mai cites only Malachbelos (824-8) and Malvisae (927-8). ILS index s.v. Mai cites no other deities. <43> CIL XIII, 8208, probably of third century. <44> Birley (1953), ; M. Speidel, Die Equites Singulares Augusti (Bonn, 1965), 68-78, hereinafter cited as Speidel (1965). <45> RE s.v. Mars cites only five Celtic examples of Mars with an epithet beginning with 'M' (1952); only Marmogius and Mullo, both from the Continent, have any resemblance with the letters Mai ; there is only one Germanic epithet possible, Halamardus (1963). ILS index cites no further examples. <4e> M. Speidel, 'Numerus Syrorum Malvensium. The Transfer of a Dacian Army Unit to Mauretania and its Implications', Dacia, NS 17 (1973), l47> Birley (1953), <48> RIB 191 ILS <4B> Speidel (1965), 69. Perhaps restore: [I(ovi) O(ptimo) M(aximo) Iu]jno(ni) [Minervae] j Mart(i) Sal[uti] / Victor(iae). (so) por Mars Militaris supra, note 7, and Birley (19S3), 97; for Mars Victor, supra, note 4, and Speidel (1965), 69. (") Note that Newstead had a mixed garrison of legionaries and equites alares; RIB and references. 106

5 a site such as Auchendavy at the middle of the Antonine Wall*52'. Birlcy showed overwhelming evidence that Cocceius Firmus had earlier in his career been a member of the Equites Singiilarcs August^53); such a man, when subsequently commissioned as a legionary centurion, would be a most suitable officer to be put in charge of the special force at Auchendavy*54'. A legionary centurion was similarly in command of the two legionary cohorts and the auxiliary cavalry stationed at Newstead, the control centre for the hinterland of the Antonine Wall*55'. The main difference between the northern frontier in Hadrian's time and in the third century that the later was version had an excellent wideranging patrol system of scouts beyond the physical frontier*56'. There are strong hints that this was foreshadowed in the later second century (Hadrian's Wall IB)*57'. The new evidence of this inscription raises the question as to whether the credit for the creation of a patrolling and scouting system beyond the frontier in Britain should not be assigned to the Antonine Wall and that this was one of the refinements and improvements subsequently incorporated into the southern frontier at a later date*58'. (52) jjje tombstones of the Semitic trader Salmanes and the a route centre civilian Verccunda point to Auchendavy being (RIB ). (53> Birley (1953), Cf. Speidel (1965), 47. (54) q JJAnnie Epigraphique (1895), 20, for a legionary centurion praepositus Brittomim et exploratorum. <55> C. Arrius Domitianus, a centurion of legio XX Valeria Victrix, dedicated altars to Diana and Silvanus, both favourites of Danubian Equites Singulares Augusti, and also to I.O.M. (RIB ; Birley (1953), 102); conceivably he, too, had had similar experience earlier in his career. For another centurion, RIB 2120 with 1725, to Apollo, another deity in this category. (56> Supra, notes 32-5; D. J. Breeze and B. Dobson, 'The Development of the Mural Frontier in Britain from Hadrian to Caracalla', PSAS 102 ( ), 118. <57> B. R. Hartley, "The Roman Occupation of Scotland: the Evidence of Samian Ware', Britannia III (1972), 40-1, One may add occupation at Nctherby as late as 177 (RIB 975) and Risingham in IB (RIB 1227). (58> E.g. the addition of the military way, streamlining of milecastle and turret system, abolition of vallum and coastal system; Breeze and Dobson, op. cit., Note that H. SchSnberger, Kastell Kunzing-Quintana (Berlin, 1975), , has recently suggested that exploratores were based in a spare barrack-block at Kiinzing. H 107