jones roman millitary nw spain

8/6/2019 Jones Roman Millitary Nw Spain

http://slidepdf.com/reader/full/jones-roman-millitary-nw-spain 1/26

The Roman Military Occupation of North-West Spain

Author(s): R. F. J. JonesSource: The Journal of Roman Studies, Vol. 66 (1976), pp. 45-66Published by: Society for the Promotion of Roman StudiesStable URL: http://www.jstor.org/stable/299779

Accessed: 04/12/2009 18:26

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=sprs.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Society for the Promotion of Roman Studies is collaborating with JSTOR to digitize, preserve and extend

access to The Journal of Roman Studies.

http://www.jstor.org

http://www.jstor.org/stable/299779?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=sprs

http://www.jstor.org/action/showPublisher?publisherCode=sprs

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/stable/299779?origin=JSTOR-pdf



THE ROMAN MILITARY OCCUPATION OF NORTH-WEST SPAIN*

By R. F. J. JONES

(Plates IV-VI)

I. INTRODUCTION

The north-west cornerof Spain was long neglectedby Roman archaeologists,who havetended to concentrateon the more spectacularremains to be found in the south and east.However,recently more attentionhas been directed here by workersof severalnationalities,who have now produceda quite extensive literatureon the gold mines,' as well as on wideraspects, chiefly in connection with the activities of the legion VII Gemina.2 Yet there hasbeen little attempt in all this to examinewhy a substantialmilitaryforce was maintained nthe region for so long. This paperaims to review thatproblem o aboutthe end of the secondcentury A.D. The evidence available is almost entirely epigraphic, chiefly consisting ofepitaphs and religious dedications. Building inscriptions are scarce. For convenience allthe epigraphicmaterial romthe north-west of Spain that is relevant o the dispositionof thearmyis collectedin the appendix,and in the main text referencewill be made to the numbers

given there. In additiona few historicalpassages areof importance,but the archaeologicalsite evidence is very slight. The nature of the evidenceis such that most attentionmust bedevoted to the units attested in the regionand their deployment, with little to be said abouttheir actual bases. Previous work on the subject has been dominatedby the late AntonioGarclay Bellido in several masterlypapers.3 However it has tendedto concentratemoreonthe historyof the units themselvesthan on questionsof topographyand the reasonsbehindtheir presence.

* In an early form this paper was submitted as athesis at Manchester University. Further field workin I972 was made possible by a grant from theConsejo Superior de Investigaciones Cientificas ofthe Spanish Government. It is a pleasure to record

my thanks to those who helped with discussion andadvice, in particular D. G. Bird, Professor G. D. B.Jones and Dr. B. Dobson. Professors J. J. Wilkesand F. G. B. Millar and Professor Dr. Gaza Alfbldykindly read a later version of the manuscript andmade many helpful suggestions. All responsibility forthe opinions expressed remains, of course, my own.Unfortunately the recent studies by P. Le Roux, inMelanges de la Casa de Veldzquez (x97z), and by J. M.Roldan Herva's, Hispania y el ej rcito romano.Contribuciona la historia social de la Espaiia Antigua(SalaInanca, I974), were not available at the time ofwriting and therefore could not be referred to.

All epigraphic references are to CIL unless other-wise stated. The following abbeviations are used:Alfoldy I965. G. Alfoldy, Madrider Mitteilungen 6

(I965), I05.Blazquez I962. J. M. Blazquez, 'Estado de laromanizacion de Hispania bajo C6sar y Augusto,'Emerita 30 (-I962), 71.

CMLeo'n. M. G6mez-Moreno, Catdlogo Monu-mental de Espafta: Provincia de Leon ( 925).

CMZamora. M. G6mez-Moreno, Catdlogo Monu-mental de Espanta: Provincia de Zamora (1927).

Cohen I2. H. Cohen, Medailles Imperiales2(1930).Diego Santos. F. Diego Santos, Epigrafia Romana de

Asturias (1959).Fita i904. F. Fita, Boletin de la Real Academia de la

Historia 46 (1904), 8o.Garcia y Bellido I956. A. Garciay Bellido, 'Explora-

ciones arqueologicas en la provincia de San-tander,' Archivo Espantol de Arqueologia 29

(1956), 174.

Garcia y Bellido 1959. Idem, ' Cohors I Gallicaequitata civium Romanorum,' Conimbriga I

(1959), 29.Garcia y Bellido I960. Idem, 'L. Terentius,

figlinarius de la Legio IV Macedonica,' Hom-mages ti Lon Herrfnann (Brussels, I960), 374.

Garcia y Bellido i96i. Idem, 'El exercitus hispani-cus desde Augusto a Vespasiano,' ArchivoEspaiiol de Arqueologia 34 (I6i), 114.

Garcia y Bellido 1966. Idem, 'Nuevos documentosmilitares de la Hispania romana,' Archivo

Espaifol deArqueologia 39 (I966), 24.Garcia y Bellido I970. Idem, ' Estudios sobre la legioVII Gemina y su campamento en Le6n,' LegioVII Gemina, 303.

Garcia y Bellido 1971. Idem, 'Parerga de arqueo-logia y epigrafia hispanorromanas (4),' ArchivoEspaiiol de Arqueologz'a 4 (1971), 15 1.

HAE. Hispania Antiqua Epigraphica I-i6 (195o-65).

IRG. InscripcionesRomanas de Galicia I-4 (1949-68).Legio VII Gemina. Legio VII Gemina. (CAtedra de

San Isidoro, Instituto Leon6s de EstudiosRomano-Visig6ticos. Le6n, I970).

de Palol I 969. P. de Palol, ' Una nueva inscripcion deClunia,' Archivo Espafiol de Arquteologfa 42

(I 969), I I8.Syme I970. R. Syme, ' The conquest of north-

west Spain,' Legio VII Gemina, 79.1 D. G. Bird, 'The Roman gold mines of north-

west Spain,' Bonner j7ahrbficher 72 (I97z), 36; P. R.Lewis and G. D. B. Jones, 'Ronman gold-mining innorth-west Spain,' JRS 6o (1970), I69; R. F. J.Jones and D. G. Bird, 'Roman gold-mining innorth-west Spain, II: workings on the Rio Duerna,'YRS 62 (1972), 59; F. de Almeida, ' Minas deOuro na " Gallaecia " portuguesa,' Legio VII Gemina,287; C. Domergue, ' Introduction a 1'6tude desmines d'or du nord-ouest de la p6ninsule iberiquedans l'antiquit6,' Legio VII Gemina, 253; idem,'A propos de Pline, Naturalis Historia, 33, 70-8, etpour illustrer sa description des mines d'or romainesd'Espagne,' Archivo Espaniol de Arqueologia 45-7

(1972-4), 499.2 Legio VII Gemina,passim. See also G. Fabrd, ' Le

tissu urbain dans le nord-ouest de la p6ninsuleiberique,' Latomnus29 (1970), 314; H. Galsterer,Untersuchungen zum r6mischen Stddtewesen af deriberischenHalbinisel(Madrider Forschungen 8, 1971).

3 Especially Garcia y Bellido I96I. Other papersare cited below.



46 R. F. J. JONES

There canbe no doubt now that the wars of conquestunderAugustusweremuch moreserious and extensive than the two campaignswith which he was personally associated, inz6 and 25 B.C.4 In spite of what Augustan propagandawould have had us believe, there isevidence for some unspecifiedcampaigningas early as 39 B.C.,5 although there is nothingthat need definitely have been in the north-west until 29, when M. StatiliusTaurus was in

action against he Cantabri,Vaccaeiand Astures.6 The next decade seems to have seen moreor less continualfighting, including the interventionof Augustus. The recordsuggests thatwhen he took personalcommand n 26 the Romanforces sufferedgravelyat the hands of theresisting tribes and their guerrillatactics, and we may suspect that it was convenient thatillness kept Augustus in Tarraco in 25, leaving the campaigns to his generals, chieflyAntistius.7 After this Augustus proclaimed Spain thoroughly conquered, declared thatpeace reigned throughout the Romanworld, and thereforeclosed the Temple of Janus atRome.8 Unaware of this, the Cantabriand Astures continued to resist their conquerors,with revoltsspecificallyrecorded n 24, 22 and I9, when Agrippawas sent to try to settle theproblem finally.9 He seems to have overcome not only the native tribes, but also someelementsin his own army which were restless, presumably rom having fought for so long.10Only one more disturbance s recordedthereafter, n i6 B.C."

It hadtaken all of thirteen difficultyears to win full controlof the region. It was thus tobe expectedthat subsequentlya powerfulforcewould be necessary o maintain hat control,andindeedthat even such a force could not hopeto be successfulat all times. Yet no historyof persistentuprisings s recorded. Onlytwo more disturbancesareknown,one under Neroand one underCommodus. The latter s merely based upon one sentenceof Herodianwhichlinks Spain with the revolt of Maternus n Gaul in i87.12 The other comes from an inscrip-tion fromAriminum n A.D. 66, giving the careerof M. VettiusValens.13 A few years earlier,probably in the late fifties, he had been decorated with torques, halerae et armillaewhileserving as primuspilus of VI Victrix in Spain, ' ob res prosper(e) gest(as) contra Astures.'The decorationsawardeddo not suggest that the fighting in which Vettius Valens distin-guished himself was of too greatseriousness.'4

Thus there is little actual evidence for a turbulent history in the region, save for thepresence of the army itself. Admittedlythis is an argumente silentio,but the view tends tobe confirmedby other materialwhich suggests an area rapidly adaptingto Roman customs.It was describedin very unflattering erms by Strabo, who thought the inhabitants wildbarbarians, ackingall the refinementsof Mediterranean ulture. Their 'rough and savagemanners were attributednot only to their warlikenatures, but to their isolation too.15 Thelatter complaint at least was soon rectified by the building of a road network which theitinerariesand milestones, as well as the tracing of roads on the ground, suggest was veryintensive.'6 Between the conventus capitals BracaraAugusta and Asturica Augusta no less

4 General discussions of the conquest can be found

in the following works: D. Magie, 'Augustus' warin Spain,' Classical Philology I5 (I920), 323;

R. Syme, 'The Spanish war of Augustus,' Am.3ourn.Phil. 55 (I934), 293; Syme 1970; A. Schul-ten, Los Cantabros y Astures y su Guerra con Roma(1942). For furtherreferences,see Syme 1970.

5 Under Cn. Domitius Calvinus: Dio xlviii, 41,7-42. His triumph in 36 B.C.: Ins. Italiae XIII. I,

p. 569.6 Dio li, 20, 5.7The main sources are Orosius v, 21, 1-1;

Florus ii, 33, 46-60; Dio liii, 25, 5-26, I. For a fullcollection of sources for the Astures, see J. M. RoldainHervas, 'Fuentes antiguas sobre los Astures,'Zephyrus1-2 (1970-I), 171-238.

8 Dio liii, 26, 5; Orosius vi, 21, II.9 24 B.C.: Dio lii, 29, 1-2. 22 B.C.: idem liv, 5,

1-3. I9 B.C.: idem liv, II, 2-6.10 Dio liv, II, 3-4.11Dio liv, 20, 3.12 Herodian i, IO, 2. Cf. E. A. Thompson, ' Revolts

in late Roman Gaul and Spain,' Past and Present 2

(I952), reprinted in M. I. Finley (ed.), Studies in

Ancient Society (1974), 304; G. Alf6ldy, 'Bellum

desertorum,' Bonner Jahrbiicher 171 (1971), 367.13 ILS 2648 (= XI 395).14 I am grateful to Dr. Brian Dobson for this point.15 Strabo iii, 3, 8.16 Antonine Itinerary: 387,4-395,4; 422,2-425,5;

427,4-431,3; 439,5-443,2; 448,2-452,5; 453,4-456,6. Itinerario de Barro: Diego Santos, 246 f.(The authenticity of all the routes in this document,except the second from Lugo to Iria and Docionum,has now been rejected by J. M. Roldan Herva's, ' Lastablas de barro de Astorga, una falsificaci6nmoderna?,' Zephyrus 23-4 (1972-3), 221-32. Mile-stones: II 4773, 4774, 4778, 4803, 6zI5, 6224, 6344;EE VIII 236 (= II 4838); IRG I,2 (= II 6234), II,5; CMLeon, 87 f. Modern treatments: M. EstefaniaAlvarez, ' Vias romanas de Galicia,' Zephyrus iI

(I960), 5, including further references to milestones;J. Rodriguez, ' Las vias militares romanas en la actualprovincia de Le6n,' Legio VII Gemina, 401; E.Loewinsohn, ' Una calzada y dos campamentosromanos del conventus Asturum,' Archivo EspainoldeArqueologia 38 (I965), 26; J. M. Roldan Hervas, Iterab Emerita Asturicam: El Camino del Plata (I97I).



48 R. F. J. JONES

structure of the region. Throughout, mountainmasses and river crossings exercisea closecontrol over anyone seekingto transportanything in bulk, whether people or goods. Thiswas discovered n northern Spain by Romangenerals,PeninsularWar generals,and roadorrailway engineers of every period.23 After the wars of conquest the small hill-top site atCoafia, near the north coast of Asturias, continued in occupation and indeed may have

expanded.24Such a patternseems likely at other less remote places,althoughas yet there islittle evidence. It may now be accepted that a recognizable level of Mediterranean-stylecivilization was attainedquite widely by the Flavian period.25

The fundamentalcause of whatever prosperity there was in the north-west was thegold mines. They were the chief economic activityand certainlywhat Asturia and Callaeciabecamefamous for in the Romanworld.26Evidence of mining on a largescale hasnow beenfound throughout he region, though the most importantareasseem to have been the south-easternside of the Montes de Leon and the valleyof the Rio Sil. Full accountsof the minesare now being published.27 It appears that mining activity began very soon after theconquest, especially f Syme's argumentsareaccepted that the figures which Pliny gives forgold production date from the Augustan period.28 Heavy productionmust have continuedat least until the end of the second century, when a mining official, Aurelius Firmus, Aug.

lib. met., is recorded in i9i at Villalis (94).In this area, which appears o have been prosperingand developingeconomically,the

military history falls conveniently into two parts, divided by the civil wars of 69-70. Thefirst is made up of the movementsand gradualreductionof the army of conquest, the secondof the activities of the single legion, VII Gemina, and various auxiliaryunits.

z. THE CONQUEST AND LEGIONARY MOVEMENTS TO 69 29

The army of conquest consisted of seven legions: I Augusta, V Alaudae, IX Hispana,II Augusta, IV Macedonica, X Gemina and VI Victrix. Of these, four seem to have beenwithdrawnrelativelyquickly, without leaving much trace.

I Augusta. A legioI was mentioned on a coin issue from ColoniaIulia Gemella Acci and anunspecified egioAugustawas punished in Spain by Agrippa n I9 B.C., at least by the loss ofits title.30 It seems likely that the two referenceswere made to the same legion, I Augusta,which will have been removed from the peninsulain I9.

V Alaudae. V Alaudae was in Spain with Julius Caesarat the battle of Munda and laterwasrecordedon Augustancoin issues fromEmeritaAugusta.31A veteranof both V AlaudaeandX Gemina s commemoratedat Hispalis in Baetica,32 nd an important nscripttionromItaly gives the career of a centurion, Sabidius, who served in V Alaudae as well as inX Gemina and VI Victrix, all in Spain.33 V Alaudae had been withdrawn to Germany byA.D. I4, but it may well haveleft Spain as early as i8 or I9 B.C.,perhapsat the same time asI Augusta.34

IX Hispana. The only links with Spain known for IX Hispana are its cognomen and thefact that tombstones from other provinces recordsoldiers of the legion who originallycamefrom Spain. Stones fromAquileia suggest that the legion was there early.35It had certainly

23 For Peninsular War campaigns, see C. Hibbert,Corunna (I96I) and M. Glover, Wellington's Penin-sular Victories (I963).

24 A. Garcia y Bellido, 'El castro de Coafia,'Archivo Espaiiol de Arqueologia I5 (1942), 2 I6.

25 See Blaszquez 2962.26 Josephus By ii, i6, 4; Florus ii, 33, 6o;

Diodorus v, 35-8.27 See above, n. i.

28 Pliny gives 20,o00 pounds of gold a year fromAsturia, Callaecia and Lusitania (NH xxxiii, 78).Syme has argued that these figures do not belong toPliny's own procuratorship in Tarraconensis, but tothe Augustan period, when Asturia and Callaeciaformed a part of Lusitania (HSCP 73 (I969), 2I8).

29 References to inscriptions from north-westSpain which give military information are made tothe catalogue in the appendix below. Numbers inbrackets refer to this list.

30 Cohen 12, 152, 632-4; Dio liv, ii, 5.31Bellum Hispaniense, 30, 7; A. Vives, La Moneda

Hispanica (1926), Iv, 63, 23 if.; M. Grant, FromImperium to Auctoritas (1946), 221.

32 II x76.33 IX 4122 (= ILS 2644).34 cf. R. Syme, 'Some notes on the legions under

Augustus,' JRS 23 (X933), I5, with the referencesgiven there.

3 V 906, 91 I, 947.



THE ROMAN MILITARY OCCUPATION OF NORTH-WEST SPAIN 49

reachedPannoniaby A.D. I4.36 The actual date of its departure rom Spainis unknown, butthe lack of evidence for its presence suggests that it was very early.37

II Augusta. II Augusta appearedearly on the coinage of Acci, alongside what was probablyI Augusta.38 Its presence in the peninsula s attested by severalinscriptions,39 hough only

one links it with the north-west. A dedication found at Aquae Flaviae was set up byL. Aelius Flaccus, a signifer 5). It must have left Spain quite earlyfor Germany,perhaps nA.D. io in the aftermathof the Varian disaster.40

None of these four legionshas left any significantguide to where theirbasesmight havebeen, if indeedthey everestablishedanything recognizableas permanentfortresses. Thereis a lack of topographicalprecision in all discussions of the campaignsof conquest, arisingfromthe conflictingand confusingaccounts eft us by Orosius, Florus and Dio.41 Little newcan be added to Syme's most recent treatment,where the variousmodern nterpretations reconsidered.42Although we may suspect flaws even in Syme's theories on the basis of thepresent evidence, any further argumentwould still be conducted in the murky confusionofthe historical accounts until new light can be brought by reliable archaeological videnceabout the actual ocationsof campaignbases.43 Still it is clear that within at most thirty-five

years, and probablymuch sooner, the conquest army had been reducedfrom seven to threelegions. This force was to remain stable for another thirty years. These legions also leftratherfuller information.

IV Macedonica. IV Macedonica is perhaps the best documented of all the legions apartfromVII Gemina,as the position of its base is at least approximately ixed. There is generalevidencefor its presencein Spainin the Tiberian coinage of Caesaraugusta nd in inscrip-tions from other provinces.44The most useful material, however, comes from a series ofstones defining the boundariesbetween its prata legionisand the lands of the neighbouringcivitates, Segisama and Juliobriga(95-Io8).45 Moreover at Herreradel Pisuerga severalsherds of pottery have been found bearing the stamps of L. Terentius and legio IVMacedonica,presumably he products of a legionary workshop (35). The actual site of the

fortress has not been precisely identified, although it has been suggested that it was atAguilarde Campoo, north of Herrera n the valley of the Pisuerga. Yet recent excavationsat Aguilar have only produced potteryin native styles, whilst at Herreramuch ArretineandJulio-Claudianmaterialhas been found, in addition to the stampedsherds alreadynoted.46

36 Tacitus, Ann. i, 23; i, 30.37 Syme suggested I3 B.C. (_JRS 23 (I933), 23);

H. M. D. Parker, The Roman Legions 2 (I958), 268preferred A.D. 6. Garcia y Bellido proposed that itreceived its cognomen immediately after the Can-tabrian wars and was then moved to Pannonia (I96I,

125).38 Cohen 12, 152, 632-4.39 At Olisipo, near Lisbon (II 266), and at

Burguillos, Baetica (II 985). A veteran recorded atStrasbourg came originally from Norba (modernCaiceres) (XIII 5975).

40 Ann. i, 37.41 See works cited above, n. 4 and n. 7.42 Syme 1970.43 Syme concludes that the ' tria agmina ' in which

Augustus' armies set out from Segisama in 26 B.C.,

according to both Orosius (vi, 21, 3) and Florus (ii,33, 48), could not have been intended to move againstall the north-west, including Asturia and Callaecia,but must have been confined to Cantabria. This restspartly on the assumption that it was far too rash astrategy for Augustus, and partly on the identificationof ' Bergida ' in Florus, where a battle took place (ii,33, 49), not as Bergidum Flavium in the Sil valley,but as the 'Vellica' in Cantabria noted by Ptolemy(ii, 6, 50). This place is given as 'Attica' in Orosius(Vi, 2I, 5) and as ' Belgica' in anothermanuscriptofFlorus. Although by Syme's own admission (1970,

go) the Codex Bambergensis, which gives 'Bergida',is generally the most reliable manuscript, it is thereading ' Belgica ' that he amends to read ' Vellica ',

which he fixes in Cantabria. Schulten, op. cit. (n. 4),followed by A. Brancati, Augusto e la Guerra diSpagna (I963) and by W. Schmitthenner, Historia i i

(I962), 54-70, accepted 'Bergida' as BergidumFlavium and went on to posit a massive triple advancefrom Bracara, Asturica and Segisama. With hind-sight, such a plan certainly seems too ambitious towork, as Syme rightly points out, but it is clear thatAugustus' campaign of 26 was not at all a success

(Dio liii, 25, 6-7).Incidentally, the strategic importance of the siteof Bergidum at the crossing of the Rio Cua wasrecognized in Moore's retreat to Corufia in i8o8,when one of the few stands took place at Cacabelos,immediately below the site of Bergidum, a positionwell described in Orosius' phrase 'sub moenibusBergidae' (ii, 33, 49). Cf. C. Oman, A History of thePeninsular War i (1902), 567-9; C. Hibbert, op. cit.(n. 23), 119 ff.

44 C. H. V. Sutherland, The Romans in Spain (1939),

P1. IV, g; III 399, XIII 5975, 6853, 6854, 6865, VI35I8, IX 3649, AE 1909, 58.

45 Garcia y Bellido 1956, I84 f. For a map of thestones' distribution see Garcfa y Bellido I96I, fig. I,

II8. For the base at Aguilar and illustrations ofstamps of L. Terentius, ibid. 119-22.

46 A. Garcia y Bellido, A. Fernaindez de Avilds andM. A. Garcia Guinea, Excavaciones y ExploracionesArqueologicas en Cantabria (Anejos de ArchivoEspafiol de Arqueologfa 4: Madrid, 1970), 3-24,36-43, including more stamps of L. Terentius.



50 R. F. J. JONES

Thus the fortress of IV Macedonica should perhaps be sought at Herrera rather thanAguilar, although there could have been bases in both places. Garcia y Bellido proposedfurther that the legionary base suggested by Orosius and Florus at Segisama (perhaps themodern Sasamon) was occupied by IV Macedonica before moving forward to Aguilar.47Once more only through extensive work in the field can these various possibilities be

examined and the position clarified. Outside the region of the Pisuerga, the presence of atleast a vexillation of the legion seems to be denoted by some stamped tiles found at a bath-house nearGijonon the north coast(34). The legion left Spainforthe Rhine, replacingXIVGeminaat Mogontiacum,when that was taken to Britain in A.D. 43.48

X Gemina. WhereverIV Macedonica's ortressprecisely was, it must be identifiedasthat described by Strabo as set againstthe Cantabri.49The two other Spanish legions aregiven as both being disposed againstthe Asturesand Callaeciunder a single legate, perhapssuggestingajoint base. This idea may be reinforcedby the careerof the centurionSabidius,who is expresslystated to have servedsimultaneously n VI Victrix and X Gemina: ' ita utin [leg.] X primum pilum duceret eodem[que e]mporeprinceps esset leg. VI '.50 X Geminais well attestedn theareaofAstorgaio, I2, I3, I4) andto thesouth 32, 70, 7I), the mostobvious strategic position to control both the Asturesand the Callaeci,but there is little to

associateVI Victrix with the region. Nevertheless,a double campfor both legions at sometime must remain ikely. Sabidius' servicein V Alaudae n Spain suggeststhat the arrange-ment had an earlyorigin, as might be expectedfrom parallelselsewhere.51

X Gemina'sstay in Spain is generallywell documented. Earlycoins from the coloniaeat Emeritaand Caesaraugustahow it to havebeenthereat the conquest.52The inscriptionsmentioning t fall into the two broadgeographicalgroups: at Caldasde Reyesnear the west

t_.LEO- p

C '

LUGO --0 SO o M.

FIG. 2. PLANS OF WALLS

AfterA. Garciay Bellido. Drawnby theauthor. Copyright eserved

coast (24, 25) and at Astorga (io, 12, 13, 14) and to the south at Fuente Encallada 32) andRosinos de Vidriales 70, 7I). This gives in all seven servingsoldiersand a veteran. Else-wherein the norththerearesingle inscriptions rom San Romande Cervantes 74) andfromClunia(29). Althoughthis material s quitesparse, t is muchsuperior o thatfor most of the

47 I96I, II9. Roldain HervAissuggests a base evenfurther north than Aguilar, at or near Castrillo delHaya, Zephyrus 23-4 (I972-3), 229.

48 XIII 6853-4, 6856, 6858, 6865, 6869.49Strabo iii, 3, 8.50ILS 2644 (= IX 4122).

51Fortresses suitable for two legions in Germanybefore A.D. 9: Vetera I (Camp A/C), Neuss (Camp B),Mainz, Dangstetten. See H. Schonberger, ' TheRoman frontier in Germany: an archaeologicalsurvey,' RS 59 (I969), I45.

52 Cohen I2, i5o, 604-5.



THE ROMAN MILITARY OCCUPATION OF NORTH-WEST SPAIN 5I

Spanish units and permits the proposition of separate bases, one near Caldas de Reyes, oneat Astorga, and one possibly to the south. Although Astorga seems the best candidate for thedouble camp with VI Victrix, neither there nor anywhere else is any site known large enoughto house two legions. No site is known at Caldas de Reyes and that at Rosinos de Vidrialesis far too small; 53 the surviving walls of Astorga enclose more than 2o hectares, enough for

only a single legion. The shape of the circuit is irregular, though that need raise no obstacleto an Augustan-Tiberian base, and is probably dictated by the shape of the small plateau onwhich the town stands. The surviving walls apparently date to the third century,54but theycould have preserved the lines of earlier defences (Fig. 2; pl. V). Finally we have Florus'record that Augustus gave his camp to the Astures for their capital, which can only apply toAstorga.55 It certainly appears that Astorga had extended links with the army, which we canonly partially understand at the moment. Perhaps the following sequence of occupationmight be suggested very tentatively. In the first place there was at Astorga an Augustan basefor one legion, the defences of which were maintained when the site was turned into a civiltown. X Gemina was then transferred from a base near Caldas de Reyes to share a fortresswith VI Victrix, which may perhaps also have come from Galicia at this time. This doublefortress was established either near Astorga itself or to the south, at an unknown site. The

double arrangement may have ended when IV Macedonica was removed from Cantabriaabout A.D. 43, with the possible movement of VI Victrix eastwards into Cantabria, where amiles is recorded at Calahorra (22). X Gemina itself was withdrawn to Carnuntum inPannonia, replacing XV Apollinaris when that left for Corbulo's campaigns in the east in 63,although it was to return to Spain soon during the civil wars.56

VI Victrix. VI Victrix was the last survivor of the conquest legions in Spain.57 It hadappeared with IV Macedonica and X Gemina on the coinage of Caesaraugusta underAugustus, Tiberius and Claudius.58 However, the only serving soldiers recorded in Spainare the miles at Calahorra (22), and another at Vila Nova de Famil9ao to the south ofBraga (87). More information comes from two stones from Italy. One from Bracciano wasset up by ' centuriones leg. VI ex Hispania ' in 5 B.C.59 The other is that from Ariminum in

66 giving the career of M. Vettius Valens, who had become that colonia's patron, but who hadpreviously been primus pilus of VI Victrix in Spain.60 There he had been decorated forsuccessful operations against the Astures, probably in the late fifties, shortly before thelegionary garrison was to be reduced again by the departure of X Gemina. There is no firm

evidence about the bases of VI Victrix apart from what has been given above in associationwith X Gemina. However it seems unlikely that the strategic triangle of Astorga-Lancia-Benevente, which controlled routes both north-south and east-west, would have beenabandoned when VI Victrix was the only legion in the province. It appears to have beenconsistently held throughout the occupation until that time, and was to be again when VIIGemina was installed at Leon. Its importance would presumably have been re-emphasized bythe recent troubles with the Astures. Garcla y Bellido did make the tentative suggestion that,as Astorga had by the sixties become a flourishing town, the last base of VI Victrix in Spainmay have already been at Leon itself.6' Indeed it could almost be taken further to wonderwhether Leon might even have accommodated X Gemina before that. Unfortunately thereis very little archaeological information about the earliest periods at Leon.

The events of 68-9. VI Victrix was instrumental in the beginning of the upheavals of the

civil wars. In 68 it proclaimed Galba, the governor of Tarraconensis, as Emperor.62 He

promptly raised some auxiliary units and a new legion, VII, then titled Galbiana but laterrenamed Gemina, and marched with them to Rome.63 VI Victrix did not remain alone forlong. Galba returned X Gemina to Spain from Carnuntum, after an absence of no more

53 5 hectares, see below p. 57, and Fig. 4."4 I. A. Richmond, op. cit. (n. 22).

55 Florus ii, 33, 6o.5"Tacitus, Ann. XV, 25, 5; III I435813&, 4358111,

I435811, I43591.

67 Josephus,By ii, i6, 4.58 A. Vives, La Moneda Hispdnica (I926) IV, 71 f.;

Hill, Notes on the Ancient Coinage of Hispania Citerior

(I93I), 9O and 95; Gil Farr6s, Ampurias 8 (I95I), 65;

Beltran, Numisma 6 (I956), 9; A. Garcia y Bellido,

Anuario de Historia Derecho de Espaiia 29 (959), 484.59 XI 3312.60 XI 395 (= ILS 2648).61 I96I, 125.62 Tacitus, Hist. v, i6.63 ibid. ii, 86; iii, 6, IO, 2I.



52 R. F. J. JONES

than seven years, sending the new legion VII to the Danube in its place.64 However, eventwo legions were not enough in Spain for Vitellius, who added I Adiutrix after the firstbattle of Bedriacum,where it hadfought for Otho.65X Geminawas sent to southernSpainin case Otho's supportersshould cross from Mauretania.66These moves seem to have beenill-consideredby Vitellius, sinceonce in Spain I Adiutrix declared or Vespasianand carried

with it both VI Victrix and X Gemina.67 In contrastto the apparentconcern for Spanishsecurity shown by Galba and Vitellius, the new Emperor, Vespasian, seems to have beenquite content with the loyalty and peacefulnessof the Spanish provinces, for he withdrewall three legions to help deal with the threats in Germany, where they next appearin thearmy of Cerialis.68None of them returned to Spain.

A surprisingpattern may be seen in these movements. Firstthe garrison n Spain wassteadily reduced over many years. Then suddenly it received massive reinforcements,attimes of crisis when presumably roops were in great demand; but shortly afterwards hewhole army was removed, eavingSpain with no legions. After this it was not re-garrisonedfor some years, for the earliestthat VII Gemina could have arrivedwas 74, and it may nothave come until 78 or 79. The posting of I Adiutrixto Spain by Vitellius could be seen asan attempt to keep a potentially subversive legion out of more sensitive areas,but such an

idea cannot explain why GalbareturnedX Geminathere. It mayhave been becausehe nowhad an extra legion, his new legion VII, which had to be stationed somewhere after Galbahad actually taken power, and so he chose to strengthen his own former province. Nogeneral anti-Romanunrest seems likely at this time, as it was possible so soon afterwards oremove all the legions. A positive reason should be sought to make Spain apparentlysoimportantat such a criticaltime. Although X Gemina was posted to the south to safeguardthe strategicroute from the Africanprovinces,there is no reasonto doubtthat the other twolegionsremained n the militaryzone, the north. Withlittle threatof a localuprising known,the only justificationfor the concentration n the north would seem to have been the needto ensure control of the gold mines. This massive source of wealth appears to have com-manded a high priority with the various contendersof 69. Vespasian was forced to leavethem without legionaryguardwhen his greatest threatcamefrom the north of the Empire,in Germany.

3. VII GEMINA AND THE AUXILIARY UNITS

LegioVII Gemina. The eventual arrivalof VII Geminain Spain cannot be preciselydated.During the civil wars it seems to have spent much of its time marching back and forthbetween Rome andPannonia, inallyfightingforVespasianatthesecondbattleof Bedriacum,beforereturningagain to Pannonia.69It next appears n Germany n the period 724 in thecampaignsof Clemens, probablygainingthen the title of Felix.70 At some time after that itmoved to Spain, although the first definite record of it there is not until 79 (4). It was tomaintain its permanent base there throughoutthe rest of the Roman period. As far as isknown, its fortress was at Le6n from the start. The delay in posting a legion to Spain

suggests that, when it did finally happen, it was a deliberateact, not merely following anestablished traditionalpattern. Vespasian undertook a thorough review of his legionaryforcesandraisedtwo new legions, havingdisbanded our disloyalones.7' This surelyshowsthat he had a clear picture of what strength he required. In the early seventies circum-stances may have changed so that there were establishedrather more legions than werestrictly necessary, n which case it could well have been convenientto stationone in Spainagain. It avoided disturbing the Imperial administrative system by having to makeTarraconensis a civil province. Furthermore,a legion there was well-placed to supplyreinforcementswhen needed to the northernfrontier or to Mauretaniaand Africa.

However, it would have been easy to decide that Hispania Tarraconensiscould havebeen left without a legion, and we may look for a more positive role than that implied in apolicy of profound caution and conservatism. What did VII Gemina actuallydo with its

"4ibid. ii, Ii.

"6ibid. ii, 67.*6 bid. ii, 58."7 ibid. iii, 44.

68 ibid. iv, 68.69 ibid. iii, 22; iv, 39.70 ILS 2729.71 Dio lv, 24, 3.




time in Spain?It certainlyprovided troopsto act as an occasionalurbanpolice force and towork on the governor'sstaff in Tarraco.72Yet other provincesmanagedwithout a legionfor such purposes. Vespasian accompanied he posting of the legion with the grant of iusLatinum o Spain.73 As this was preceded by ' procellae of unspecifiedform,74t mayhavebeen felt that a legion was requiredto keep any agitationin check, in case magnanimity

failed. However, no more is heard of these disturbancesand they may have been entirelycivil in character;nor areany other troublesrecorded. Such threats alone can hardlyhavejustified the continuedpresenceof VII Gemina n Spaineventhroughall the strainson theEmpire'smilitaryresources mposedby the northernwarsof the second century. Againtheanswermay lie in the area'smineral wealth.

The concentrationof material at Leon leaves no doubt that it was the legion's mainbase(37, 38, 39' 53, 55), and aspecialemphasiscan be givento the largenumbersof stamped

N 50 100 200 KM.

[ LAND OVER

FIG. 3. NORTH-WEST SPAIN: THE DISTRIBUTION OF MILITARY EVIDENCE

KEY: I. Abedes (i); II. Abia de las Torres (2); III. Aguilar de Camp6o; IV. Aldea Nova (3); V. AquaeFlaviae (4,5); VI. Asturica (6-i8); VII. Babe (i9); VIII. Bracara (20,2I); IX. Caldas de Reyes (24-6);X. Castrelo del Valle (27); XI. Castrocalb6n (28); XII. Clunia (29); XIII. Durat6n (30); XIV. Fardaes (3I);

XV. Fuente Encallada (32-3); XVI. Gij6n (34); XVII. Herrera del Pisuerga (35-6); XVIII. Le6n (37-57);XIX. Near Le6n (58); XX. Lucus (59-64); XXI. Luyego (65-7); XXII. Quintana del Marco (68); XXIII.Rairiz de Veiga (69); XXIV. Rosinos de Vidriales (70-2); XXV. San Pedro de la Vifia (73); XXVI. San Romainde Cervantes (74); XXVII. Sobrado; XXVIII. Soto de la Vega (75-82); XXIX. Ujo (83); XXX. Valen9a doMinho (84); XXXI. Vila Pouca d'Aguilar (85-6); XXXII. Vila Nova de Famili9ao (87); XXXIII. Villalis(88-94); XXIV. Segisama; XXXV. Hormiguera (95); Henestrosas (96, I03); Cuena (97-8); San Vitores (99);Las Quintanas (Io4); Las Quintanillas (Ios); Reinosilla (io6); Castrillo del Haya (Io7); El Haya (io8);XXXVI. Villasidro (ioo).

tiles found there, giving the legion'svarioustitles (40-52). It has alsoleft evidence at manyother sites in the north-west: Abedes (i), Braga(20), Caldas de Reyes (26), outside Leon(58), Lugo (59), Luyego (65, 66), Quintanadel Marco(68), Ujo (83), Vila Pouca d'Aguilar(85) and Villalis (88, 89, 90, 93). Veterans are recorded at Astorga (8, 9, ii) and Lugo

(63, 64). This wide distributionseems likely to reflect not so much purely military duties,

72 II 4083, 4III, 4I22, 4I42-4, 4I47, 4I49-50,4I52-7, 4I6I-2, 4165, 4I67-8, 4I70-I, 6o88.

73 See now A. B. Bosworth, 'Vespasian and the

provinces: some problems of the early 70s A.D.',Athenaeum 5I (I973), 49, esp. 5I-5.

7 Pliny, NH iii, 30.



54 R. F. J. JONES

involving widespread garrisoning, as more general policing and administrative tasks.Detachments were also sent outside the province from time to time. Vexillations arerecorded in the early second century in Britain, in Lambaesis at about the same time, andabout I70 troops were sent to Italica in Baetica to guard against Moorish attack.75

Auxilia. By the time of the arrival of VII Gemina in Spain the auxiliary troops had begunto take on a much more important role, but our knowledge of their actual strength in Spainis far slighter than that of the legions. Suetonius noted that Galba had two alae and threecohorts in the province in 68, in addition to his legion, VI Victrix, and that he raised moreauxiliary troops at the same time as the new seventh legion.76 Garcla y Bellido suggested thatfive or six units were involved here.77 Other such general references are rare. A group ofinscriptions from Noricum describing the career of T. Varius Clemens record that he was'praefectus auxiliorum in Mauretaniam Tingitanam ex Hispania missorum '.78 A recentlypublished inscription from Italy describes Sextus Flavius Quietus, perhaps from Britain,as 'missus cum exerc. in exp. Maur. ab Imp. Antonino Aug.'.79 Both these must refer toarmies sent against the rising in Mauretania between I45 and I50, but tell us no more aboutSpain than that there were auxiliaries stationed there at the time. Another stone, from

Pisaurum, gives a' praefectus cohortium civium Romanorum quattuor in Hispania', but itsdate is uncertain.80 Garcla y Bellido suggested the third century, but anyway it only gives aminimum figure for the Spanish provinces as a whole.81 Rather beyond our main period,the Notitia Dignitatum records five cohorts: II Flavia Pacatiana, II Gallica, III Lucensiumand Celtibera, all in Hispania Callaecia, with I Gallica in Tarraconensis.82 As no dischargediplomas have yet been found in Spain, this material is all we have for the general composi-tion of the auxiliary garrison, apart from the evidence of individual units on gravestones anddedications.

CohorsIV Gallorum. Although altogether eight alae and seven cohorts are attested, it is muchmore difficult to fit them into a chronological framework than it was for the legions. Theearliest evidence is for cohors IV Gallorum. It has left a series of boundary stones, all dated

to the time of Claudius. One from Castrocalbon, south-west of La Bafieza, marked theboundary between the cohort's territory and that of the civitas of Bedunia (28). The othereight stones were all found at Soto de la Vega, north of La Banieza. Four of them give theboundary with Bedunia (75, 76, 8o, 82), another two with the Luggones (77, 79), and theremaining two are only fragments (78, 8i). Such a grouping of boundary markers can hardlyrepresent their original location and strongly suggests their collection and abandonmentbefore the unit's departure elsewhere, perhaps even before they were properly set up, asproposed by Garcla y Bellido.83 If so, the unit would have to have left by 54. The cohort'slater history is confused, chiefly because there may have been three or even four differentunits all with the same title, but it does seem likely that the former Spanish unit was servingin Mauretania from some date later in the first century.84 Garcla y Bellido may have beencorrect in suggesting that it first moved elsewhere in Spain, but there is no evidence to

confirm or deny this. However, their early fort must have been near the find-spots of themarker-stones, perhaps on the Rio Orbigo at La Banieza or at Soto de la Vega itself.85 Thisfirm evidence for a defined territory for an auxiliary unit in the Claudian period has importantimplications for the understanding of the administration of the region.

75Britain: ILS 2726. Lambaesis: VIII 10474,

I2; R. Cagnat, L'armee romaine d'Afrique2 (19I3)Iz 2; A. Garcia y Bellido, ' La legio VII Gemina PiaFelix y las origenes de la ciudad de Le6n,' Boletin dela Real Academia de la Historia 127 (I950), 463.Italica: II 1125-6. For various studies on aspects ofthe legion, see Legio VII Gemina, especially Garcia yBellido I970, A. Garcia y Bellido, ' Nacimiento de lalegi6n VII Gemina,' G. Alf6ldy, 'Die senatorischenKommandeure der Legio VII Gemina,' and H.-G.Pflaum, 'Les officiers equestres de la 1lgion VIIGemina.'

76 Suetonius, Galbaio.7 T96I, I34.78 III 52Ii-6. For T. Varius Clemens, see

G. Alf6ldy, Noricum (I'974), I24 f., 274, 277;

H.-G. Pflaum, Les carrieresprocuratoriennes equestressous le Haut-Empire Romain i (i96o), 368 f., no. I56.

79 H. Comfort, American Yournal of Archaeology64 (I 960), 274.

80 XI 6344.81 I96I, 147.82 Not. Dig. Occ. (ed. 0. Seeck, I876), 42, 27;

42, 28; 42, 29; 42, 30; 42, 32.83 I96I, I55.

84ibid. I57; G. L. Cheesman, The Auxilia of theRoman Imperial Army (I9I4), I74, n. i. On a diplomaof 88 from Mauretania Tingitana: AE I953, 74.

85 Garcia y Bellido I96I, I58.




Ala II Gallorum.The only other auxiliaryunit securely known to have been in Spainbeforethe civil wars is ala II Gallorum. The link comes from a stone from Italy giving the careerofQ. Atatinus Modestus, who, having served as tribunusmilitumof X Gemina in Spain forsixteen years, was made praefectusof ala II Gallorum in eadem provincia .86 His transfermust have come before the legion's withdrawal n the early sixties, or perhaps at that time.

Garclay Bellido did suggest that the ala was actuallyattached o X Gemina, but not enoughis known about the relationshipsbetween auxilia and legions for any degree of certainty,although the idea is attractive.87Ala II Gallorum s otherwise not known in Spain.

Other Units. Unfortunately most of the auxiliary units are undated. The following arebriefly attested: ala I Gigurrorum t Castrelo del Valle (27); ala Sabinianaat Aldea Nova(3); ala I Singularium n Lusitania88; ala Sulpicia at Astorga (i6); ala TautorumVictrixat Calahorra28); ala II Thracumn Asturias i iO) and in Lusitania89; cohors V Thracumat Astorga (I7), with former praefectiat Tarraco and Malaga.90 Cohors II Lucensiumsrecorded at Lugo and in the Notitia (60, 6i), whilst cohors I Gallica is only mentioned inthe Notitia.91 It can safely be assumedthat these two were in Spain late, but we have noindication of when they arrivedor when they were raised. The decurioof ala I Gigurrorum

has a patently native name and, as the Gigurri are known as a local tribe, it has been sug-gested that the ala was raised n the areaand may only have served in Spainfor a short timesoon after the conquest.92 Since many of these undated units are only recordedon singleinscriptions, it is difficultto know whether these locations always represent where the unitwas permanently stationed. The geographicalrange is wide, but it is hardly possible toconstruct any theory about a garrisoning ystem from this evidence alone.

The Villali's nd Luyego nscriptions.Bythe second half of the second centurytherearemanymoresecurely attested and dated units. They appearchieflyon the importantseries of nineinscriptionsfrom the Astorga region, from Luyego and from Villalis. All are dedicationsmade to celebratethe diesnatalisof a unit. Three units are so honoured: VII GeminaonIOJune, cohors Gallica on 22 April, and cohors Celtiberorumn I3 October. They also

recordother units and officials n the areaat the time. Moreover only two of these stonescannotbe dated to a specific year. Seven come from Villalis. One of I63 mentionsa vexilla-tion and signiferf legion VII and a decuriof cohors Celtiberorum88); one of I65 givescohorsI Gallica and a centurion, with the same signifer of legion VII (89); one of I67,I Celtiberorumnd a tesserarius,he same centurion of I Gallica as in i65, and an imaginiferof legion VII (go); one dated simply to the joint reign of Marcus Aurelius and LuciusVerus, gives cohors Gallica and decurioof I Celtiberorum9I); one of I75 gives cohorsI Gallica, but with a new centurion (92); one of I84, a vexillation of legion VII and adecurioof ala II Flavia (93); one of I9I, cohors Gallica, with a differentdecurioof ala IIFlavia (94). Both Luyego stones celebratethe dies natalis of VII Gemina. One was set upby a vexillationof it under a centurionand a decurio,n the reignof AntoninusPius (65); theother was erected in i8i by M. Aurelius Eutyches, Aug. lib. proc., a centurion of cohors

I Gallica,and one Avitus Paternusof unknownrank (66). A gravestoneof a milesof cohorsI Gallica has also been found at Luyego (67).

The provenanceof the Villalls stoneshas in the past been the source of some confusion.The five published by Hiibner in CIL were first recorded by Muratori in the eighteenthcentury as coming from ' Castro S. Christophori n Gallaecia , and this was followed byHubner. These stones were later lost, but four were rediscoveredwith two new ones of theseries at Villalls by Gomez-Moreno, who published them with corrected readings in I909

and I926. Photographsof the surviving stones have recently been published, and the sixVillalls stones are now in the Museo de San Isidoro at Leon,the two from Luyego in AstorgaMuseum. Despite these corrections, even in I967 their provenance could be given as'S. Critobal . The associationwith Villalis is confirmedby the fragmentof anotherstone

86 IX 36I0.87 Garciay Bellido I96I, 134.88 At Idanha-a-Velha (NE. of Castrelo Branco):

' Les inscriptions latines inedites du Musee Leite deVasconcelos,' 0 Archeologo Portugues, N.S. 3 (I960),

25, no. 5; F. de Almeida, Egitania (I956), IS5, no. 2.

89 At Caparra (N of Caceres): II 8I2.90 II 4138, 4212, 1970.91Somewhere in Galicia: Not. Dig. Occ. 42, 28.92 Garcia y Bellido I96I, 142.

93 R. Saxer, Epigraphische Studien i (I967), 63.



56 R. F. J. JONES

of the series now built into the wall of a house in the village(P1.IV).94 This is most probablythe remainsof the missing stone published as CIL II 2555 (9i), but it may be a new one.Only the line giving the dedicationto the Emperorsnow survives from the inscription.The provenancegiven in CIL for all these stones has now been widely rejectedin favour ofVillalis. No suitable site for ' Castro S. Christophori has ever been identified,whereasthe

locationat Villalls fits verywell in an area of proven militaryand mining activity,especiallywith the parallelsfrom nearbyLuyego.

No full study of the whole series and its implicationshas yet been made, nor is this theplace for that. They are important not only for naming several units operating closelytogether in the later second century, but also for the officialsof the Imperialservice whoappear with them, often explicitly in command of them. An impressive assembly ofAugustanfreedmenand beneficiariian be found, in additionto Aurelius Firmus,Aug. lib.met. (94). Althoughhe only appearson the latest of the series, his title is probably he clueto what the othervariousnon-militaryofficialswere doing here. It is paralleled lsewhere nthe north-west by that of M. Ulpius Eutyches, proc. metall. Alboc.95 Their task wasspecificallythe administrationof the mines; yet here Aurelius Firmus was recorded withtroopsunder his authority. Of the units themselves,cohors Gallicaappearsmost frequently

both at Villal's and Luyego. Legion VII is also representedat both places, but cohorsCeltiberorumnly at Villalis. Ala II Flavia is well attestednot only at Villalls but at Rosinosde Vidrialesand San Pedro de la Vifia to the south also. Cohors Gallica and ala II Flaviahave both recentlyreceivedfull treatmentson the various evidence for their activities.96

Cohors Gallica. Cohors Gallica equitataciviumRomanorums clearlywell attestedin theDuerna valley at Villalis and Luyego, but good evidence for its presence in southernCallaecias also to be foundin the dedicationsby a milesat VilaPoucad'Aguilar,zo km northof Vila Real in northern Portugal,and by a signiferat Rairizde Veiga, some 6o km to thenorth in Spain, south of Orense (86, 69). Unfortunatelythese inscriptionsare undatedandit is unclearwhen the cohort was here, althoughit is more probablethat it was before itsmove to the Duerna. A praefectus f the cohort is also known from Herreradel Pisuerga n

Cantabria 36).

Ala II Flavia. The unit's full title, ala II Flavia HispanorumiviumRomanorum,eems tosuggestthat it was founded after70, andperhapsafter the grantof iusLatii to Spain. Apartfrom Villalis (93, 94), Rosinos de Vidriales (72) and San Pedro de la Vini'a73), it is knownin the 'largecentres at Astorga (I 5) and Leon (54), as well as from an unspecified site inGalicia (IO9). There is some more equivocalepigraphicevidence from Spain which mayrefer to it, and it is recorded n north Africa.97 It was probablyone of the units sent thereinthe I5os, but it must have returnedto Spain by I84 at the latest. It probably correspondsto the ' cohors II Flavia Pacatiana of the Notitia.98

Cohors Celtiberorum.Apartfrom the Villalisevidence,this cohortis known in Spain only

by the tombstoneof apraefectus tTarracoandby the Notitia reference o ' tribunuscohortisCeltiberae .99 The unit's nameis included in diplomasof the earlier second centuryfromBritainand from Mauretania, uggestingthat it arrived n Spainin the middleof the century,unless againthere was a duplicationof title.100

94A. Garcia y Bellido, ' Parerga de arqueologia yepigraffa hispanorromanas (2),' Archivo Espanlol deArqueologia 36 (I963), 205-6.

95 II 2598. The mine mentioned is probably thegold mine in Callaecia, 'metallum Albucrarense,'noted by Pliny, NH xxxiii, 8o. M. Ulpius Eutychesmay be the same man whose epitaph comes fromNarbonne (XII 4490), and who had the title ' Aug.lib. mesor', probably 'mensor', surveyor orengineer.

96 Garcia y Bellido I959; M. Vigil, 'Ala II FlaviaHispanorum civium Romanorum,' Archivo Espanfolde Arqueologza34 (I96I), I04.

97 From Spain: II 56Io, 2637, 255 I; EE IX 277.From Africa: AE I935, 2I7, no. 35. From

Mauretania:VIII 2I050 (= EE V I004)." Not. Dig. Occ.42, 27. 'Cohors I Gallica 'is also

recorded, bid. 42, 32.99 II 4141; Not. Dig. Occ. 42, 27.

100 Cohors I Celtiberorumis attested at CaersOs inWales on tile-stamps (VII I243 = EE IX I285; cf.E. Birley, 'Roman garrisons in Wales,' ArchaeologiaCambrensis 102 (I952-3), I4); on British diplomas inI05, I22, I46 (XVI 5I, 69, 93); on diplomas from

Mauretania Tingitana in I09 and II4-I7 (XVI i62,

I65). The discrepancy in the sequence of dates mayindicate two units with the same name; otherwise atransfer from Britain between I05 and I09 must bepostulated, with a return between II 7 and I22.




Site evidence. Both Luyego and Villalis are very close to major gold mines in the Duernavalley. A few possible Roman settlements are known from surfacetraces near Luyego, butnone have definitely military characteristics.101At Villalis no actual site is known either.Gomez-Moreno found some Roman material nearby at El Castrillon, and the road fromAsturicato Bracara rosses the Duerna close by at Villamontan.102The Antonine Itinerary

gives a place named Argentiolum somewhere here, which is presumably the same site asthat which originally produced the Villalis stones.103 Its precise location must be eagerlyawaited. Site information n general is very poor. Leon of course stands securely as thelegionary ortress, but little is known of the interior. The same is true to varying extents ofthe other walled towns.104

Rosinosde Vidriales. Only the site at Rosinos de Vidriales can be accepted as an auxiliarybase. It has been recognized for some time: a plan was first published by Schulten andLammerer in 1927 and later reproduced by Garcia y Bellido.105 Loewinsohn in I966claimedthat this planwas inaccurate n detail andthat his air photographsrevealed hat thefort was in fact ' un perfecto rectangulo .106 Unfortunatelyhe failedto provide a correctedplan. That given here (Fig. 4) is based upon fresh observationsmade in the autumn of I972.

ROSINOS D E VIDRIALES

SITUATION FR

San Pedrod

FOTRosinos de

Vidriales

lJea 0 1KM.

SantlbafiezdeVidriales

Tardemdza

FIG. 4. ROSINOS DE VIDRIALES

After A. Garcia y Bellido. Drawn by the author. Copyright reserved

This gives a rather argerarea than previously recorded,of approximately 50 X 200 M.107

It hashardlybeen remarkedupon thatthe fort is enclosedby a substantialstone wall, nearly2 m wide. The lower coursessurviveintact, exceptfor a short stretch ost atthe west corner.For the most part what remainsis a rubble core, heavilymortared,but some facing blocksare still in position along the north-eastern side (P1. IV). There is no sign of any earthrampartbehind the wall, but on the south-eastern side is what may perhaps have been a

101 R. F. J. Jones and D. G.Bird, op.

cit.(n.

I),

esp. 66 and fig. 3.102 CMLeo'n 69; E. Loewinsohn, op. cit. (n. i6),

fig. 27.103 It. Ant. 422, 2.104 See n. 2I.105ArchdologischerAnzeiger (1927), 202, figs. 3 and

4; Garcia y Bellido I96I, 137, figs. I2 and 13.106 op. cit. (n. i6), 29, n. 7.107 These figures are still based upon a rather

inadequate survey, but were the best obtainable onmy visit to the site in November 1972. A full-scale,accurate survey is required.



58 R. F. J. JONES

tower (P1. V). Here the foundations and lowest courses of a rectangular tructure about6 x 5 m are attached to the back of the fort wall. It is built of small stones set in thickmortar. On the visible evidence of the two rubble structures, it is not at all clear whetherthe wall and this building are contemporaryor not. The walls of the subsidiary structuresurvive as about 50 cm thick. They can only be seen joining the main wall at one side, as

the south-westernwall stops about I -50 m from the perimeterwall. This probably repre-sents the space for a doorwayrather han just robbing, as there is here a stone suitable for adoor. What exactly was the function of this building must remaina matterfor conjectureatpresent. Its walls are certainly much less substantial than the main wall and they seemrather too slight to carry the full weight of an intervaltower; nor is there any sign of asimilar structureelsewherein the fort. It can hardlyhave been a gate tower, for the wallcontinuesunbrokenpast it. The suggestionof a local villagerwas that it was a kitchen,andthis may be as good as any, at leastuntil excavation. No gates to the fort arenow visible, butthe modern road runs through the centre of the site and may well follow the line of anancient predecessor; in which case the remains of two gates have probably been removedby the modern road and its ditches, althoughno signs of substantialmasonrywere visiblein the sides of the ditches. Almost all the interior is under cultivation, where the plough

has turned up large quantitiesof tile throughout. In itself this provides no proof of date,since modern local tile is virtually indistinguishable rom Roman. However there is verylittle tile to be found on the ploughed fieldsoutside the wallsof the presumedfort. Amongstthe tile there were also several small sherds of Roman pottery, much abraded,althoughtherewererather ewer pieces than mighthave been expectedin view of the profusionof tile.Unfortunatelynone of the sherds found was datable. The amount of occupationmaterialand the strength of the walls leave little doubt that this was a permanentestablishment,probablywith quite a long life.

Its militarycharacterwill not be proven until excavation,but the rounded cornersareapowerful argument. It also lies in a strong position, dominating the valley. It is perhapsodd that no later settlement has taken over the site, but four villages cluster around it(Fig. 4). Three of these are credited with having produced military inscriptions: Rosinositself two gravestonesof militesof X Gemina and a dedicationby the praefectusof ala IIFlavia (70, 71, 72), San Pedro de laViniaanotherdedicationby ala II Flavia (73), andFuenteEncalladaanothergravestoneof a milesof legion X andthat of a veteran of an unnamedunit(32, 33)4It seems reasonable o assignall these provisionally o the knownsite, althoughthepresenceof the two units suggests a more complex history than the visible remainsreveal.As it stands the base would have been far too small for a whole legion; so the more likelygarrison or this fort must be ala II Flavia, even though the size would still be unusualforsuch a unit.108

Castrocalbodn.ess than i 5 km northof Rosinosde Vidriales ies the site called Castrocalbonby Loewinsohn (Fig. 5).109 This is slightly misleading,since the actual site is some 3 kmsouth-west of that village. The villagelies on the Rio Eria, near where the Romanroadfrom

108 Vigil, op. cit. (n. 96), iII, claims that, as nomention is ever made of its being an ala milliaria, itmust only have been an ala quingenaria. Neverthelessclose parallels in Britain suggest that Rosinos was theright size for an ala milliaria, but rather too big for anala quingenaria. The ala quingenaria fort at BreconGaer in Wales was 3 *1I4 ha (7 *8 acres), but the fortat Newstead in Scotland in the second century had anarea of 5 4 ha (13 5 acres) and was definitely occu-pied by an ala milliaria. The fort at LeintwardineVillage in Herefordshire of the same period was 4 v55ha (II 3 acres). It was either garrisoned by an alamilliaria or was a stores depot. The milliary alaPetriana held Stanwix near Carlisle, with an area of3 * 2 ha (9 *3 acres). These comparisons tend to showeither that an ala milliaria, presumably ala II Flavia,occupied Rosinos as an orthodox fort, or that a largepart of the site was used for such purposes as stores.(For Brecon Gaer and Leintwardine, see V. E. Nash-Williams, The Roman Frontier in Wales2, ed. M. G.

Jarrett, (I969), 48-5 I, 94-5. Newstead: J. Curle,A RomanFrontierPost and ts People I9I I), 29; I. A.Richmond, Proc. Soc. Antiquaries of Scotland 84(1950), 1-38. Stanwix: excavation by F. G. Simpsonand I. A. Richmond, JRS 31 (I94I), 129-30, andpl. xii. Cf. E. Birley, ' Alae and cohortes milliariae ',in Corolla memoriae Erich Swoboda dedicata(= Romische Forschungenin Niederosterreich v, I966),54-67). Another possibility is that the visible remainsat Rosinos are of a later period, since another parallelis the fort at Piercebridge, Co. Durham, built about300, at 4.35 ha (i0o 8 acres); cf. G. S. Keeney, Trans.Durham and Northumberland Archit. and Arch. Soc. 9(1939) fig. i.

109 op. cit. (n. i6), 34 ff. The approximate dimen-sions of the three camps are: camp A, 250 X 170 m(4-25 ha = io6 acres); camp B, 150 X 115 m(I 73 ha = 4 3 acres); camp C, 125 X C. 200 m(2s5 ha = 6 25 acres).




Asturicacrossed the river. It also fits convenientlyinto a pattern Asturica-Villalis-Castro-calbon-Rosinos de Vidriales. All these are about i5 km apart and the last three dominatevalleys. However, the actual site known as Castrocalbondisrupts any such pattern ofRoman occupation,as it lies too close to Rosinos de Vidrialesand too far from Villalis. Noris it situated in a river valley at all, but on the high land that separatesthe Eria itself from

the Arroyodel Real, on a plateau below the Sierrade San Feliz, although still beside theRomanroadfrom Asturica to Bracara. The sites found by Loewinsohn provide in them-selves a remarkable ontrastwith Rosinos de Vidriales. None of the three noted shows anysign of havingbeen occupied with any intensity or for any length of time. Their perimeters

Castrocalb6n 8 -

, -

\ ,-

I ~~~~~~~~~~~~~~~40

E 1 f *f .? SIGNAL

DiTCFIH

TRACK

0SOO

M, 3~~~~~~~~~~~~~~50M

FIG. 5. CASTROCALBON

After E. Loewinsohn. Drawn by the author. Copyright reserved

areonly visiblefromthe surfaceas low earthbanks,whilst insidethereis acompleteabsenceof any occupationmaterial. Although most of the area s now open grasswith a few shrubs,there are some ploughed parts, which are just as barren as everywhereelse. By far theclearest feature is Camp B, where the rampartssurvive highest, even though it is where

recent agriculturalactivity has been greatest. Yet no tile or pottery could be found on thesurfacehere. All the evidencesuggests that these enclosureswerenot permanentestablish-ments at all, but aremorelikely to have been practicecamps. This is reinforcedby the lackof any epigraphic inds in the immediatearea, apartfrom the boundarystone of cohors VGallorum28). Eventhat does not support the idea of a Claudian ort at Castrocalbontself,but at some othersite, eitherbetween thereand Soto de la Vega,wherethe rest of the stoneswerefound, or near to Soto de la Vega itself. A later fort might have been situatedactuallyin the Eria valley, controllingthe river crossing, with the camps on the higher groundin asimilarrelationto it as those on LlandrindodCommon in Waleswere to the fort at CastellCollen.110Unfortunately he existence of such a fort remainsspeculative.

Sobrado.Anothermilitarysite, of aboutz hectares,wasnoted by Schultenon a visit in I930

at Sobrado,probablyto the west of Lugo, near the source of the Rio Tambre in Galicia.

110 The eighteen camps at Llandrindod lie z 4 kmsouth of the fort at Castell Collen, Nash-Williams,op. cit. (n. I08), 126-30. The fourteen surviving

camps have recently been surveyed: C. M. Danielsand G. D. B. Jones, Archaeologia Cambrensis iI8

(I969), 124-34.



6o R. F. J. JONES

However, he gave neither a plan nor any more precise location."' With no supporting

epigraphyeither, the site must remainquestionableuntil more informationcan be found.

4. CONCLUSIONS

The unassailableminimumof definite evidence provides a picture that is by no meanscomplete, but whichcan serve asa framework or discussion. The invasionarmywas quicklyreduced,so that after half a centurythere remaineda force of three legions, which was stillformidable. This was in its turnreducedto one legion. The civil wars of 69 brought drasticreinforcement, heir aftermatha complete legionary withdrawal. A little later VII Geminawas installedat Leon. The developmentof the auxilia's strength s hardlyknown, and theiroverall disposition scarcely at all. What is known of their distributiondoes seem to relatequite closely to that of the gold mines (Figs. i and 3). The only important areaof dis-crepancy s in the Sil valley, where no sign of military activity has been found. Otherwiseboth the armyand mining are well attestedin a general way in northernPortugal,near thenorth coast and to the south and west of Astorga. If it is now apparent that the minesbecame the most importantactivity n the north-west,the date whenthey were firstexploited

remains controversial. Syme's suggestion that the productionfigures quoted by Pliny wereAugustan supportsthe idea, once summarilydismissed by him, that a strong motivefor theoriginal conquest was to win direct control of the mineral wealth."2 If zo,ooo pounds ofgold a year was alreadycomingfrom the mines in the Augustan period,workingsmust havebegun verysoon after the conquestandhavehad a high priority or the administrators f thenewly-won region. As Blazqueznoted, in the same period Rome conquered he Val d'Aosta,where there were mines richer than those of the Alps, and sent an expedition againsttheSabaeansof Arabia Felix.113 Considered together, these actions suggest that ' economic'factorswere of some importance.

As it presumablywas in the first phase after the conquest also, the armywas heavilyinvolved with the mines in the later second century. Yet much had changedin the regionover that period. There was a good level of urbanization,and the road system provided

efficientcommunication. The natureof the terrain orbids the movementof anysubstantialforces except by the limited number of unchanging routes through the mountains, whichcould be controlledby a few key sites. Any troubleswere more likely to arisefrombrigandagethan from open confrontation. What was the army doing afterthe first decadesof occupa-tion? At first a difficultarea and warlikepeople would have demandeda strong controllingforce, to be reduced as the natives were pacified. The withdrawalof legions was thuspredictable, especially as the auxilia became more reliable, in spite of the military conser-vatismwhich only moved a legion when it was urgentlyneeded elsewhere. The strategyofholding key points was obvious. An importantelementin such a system mayhave been thesouthernroadfrom Asturicato Bracara,with its militarysites and earlydate.1'4It certainlyprovided a clear link between various sites with military connections in the later period.Were a fort to be discoverednear the villageof Castrocalbon,a fortifiedroute with a series

of regularly-spaced osts couldbe envisaged: Asturica-Villalis-Castrocalbon-Rosinos. herest of the road to Bracaracould have been similarly garrisoned. A triangle completed byfortifiedroutes to Lucus Augusti would have been an excellent way to hold the whole ofAsturia-Callaecia, or it would have controlled the main channels of movement, withoutattempting to penetrate the mountain fastnesses which had tormented the conqueringarmy.115 n Cantabria,penetrationseems to have followed the Rio Pisuerga, perhapswiththe legionarybasemovingnorthwardsup the valleyfrom Herreradel Pisuerga o AguilardeCampoo. In view of the apparentlycontinuing process of Romanization, the suddenreinforcementof the single remaining egion in 69 by two more must have been somethingof a surprise, as it restored the strength of a quarterof a century earlier. Clearly thecommandof Spainwas thought importantby the rivalsin the civil wars. The most evident

A. Schulten, op. cit. (n. 4), i86.112 See n. 28, above. Cf. R. Syme, op. cit. (n. 4),

295.'B Blazquez i962, 117 n. 4; S. Jameson, ' Chrono-

logy of the campaigns of Aelius Gallus and C. Petro-

nius,' JRS 58 (I968), 71.114 See nn. 17 and i8.115 Dio lii, 25 6; liv, II, 5. Florus, ii, 33, 59. It

might also have allowed some small-scale localbrigandage in the hills.



THE ROMAN MILITARY OCCUPATION OF NORTH-WEST SPAIN 6i

reasonfor this was its wealth: the legions were to protect the revenuesfrom the mines. Thearrivalof VII Gemina shortlyafterwards aises many more questions, not only in that it wassent thereat all, but in that it was kept there through all the warsand crises of the succeedingcenturies.

Exactly how many troopsaltogetherwere in Spain canhardly be calculated. We know

the legionaryestablishment,which must alwayshave formed a majorpart of the force,at thevariousperiods, but the number of supportingauxilia at any one time remainsa conjecture.If their strength were still the same as that assigned to Galbain Spain a century earlier,more than half of the total auxiliary forces would have been mentioned on the Villalisstones; but such a high concentrationof troops seems unlikely. More units may haveto beassumedelsewhere. This may perhapscast some doubt on the figures of two alae andthreecohortsfor the full strength of Galba's army of auxilia."6 The alternative nterpretation sthat the greaterpart of the later-second-century rmy was serving in the area of VillalisandLuyego. It is also interesting o note the high proportionof mounted troops attested: eightalae to seven cohorts, of which cohors Gallica was certainly part-mounted and cohorsICeltiberorummust have been also, as a decurio s attestedserving in it (88, 9I). This highproportionof cavalrywas probably a response to the rough nature of the country, but it isworthremembering hat the Asturesthemselveswere famed ashorsemen. It is probable hatthe units known from inscriptionsrepresent only a part of the actual strength. Certainlythis was the case in Britain. Excluding those units attestedfrom dischargediplomasalone,none of which have been found in Spain, in Wales only six units are known from inscrip-tions but thirty-four bases fromfield-work.117 n northernSpain there are at least fourteenunits known, but only three bases at most. Clearly the parallel cannot be pressed toofar, as many other factors may have been involved. The army's role may well have beendifferent. Nevertheless this well illustrates the shortcomings in the record of fieldmonuments in Spain, and the need for new information rom surveys on the ground.

The administrative ystem of the north-west is not entirely clear. The evidenceof theboundary stones of IV Macedonicaand of cohorsIV Gallorum n Cantabriaand to the

south of Astorga seems to suggest that there was no absolute, overall military control, atleast in the first century, but that military territoriaexisted alongside the civitates. Howwidespread his patternwas, or how long it lasted, cannot be said, but it is worth notingthatthe cohorsV Gallorum tones do not come from a fringe of the militaryarea,but fromone ofthe main centres of activity.

An aid to understandingmany problems may lie in the mineral wealth of northernSpain. Many of the mines must have been Imperial property. At Vipasca, near modernAljustrelin southern Portugal, they were leased out to conductores.Procuratorswere thechief Imperial officialsresponsiblefor such operations,assisted by various others, tabularii,commentarienses,nd subprocuratores.L18he Villalis stones, which are probably he fullestdocuments on mining administration n Spain apart from the Aljustrel tables, mentionseveralprocurators, ncludingAurelius Firmus, Aug. lib. met.(94). It seems likely that the

other procurators and beneficiariimentioned at Villalis and Luyego were also miningofficials like Firmus and M. Ulpius Eutyches, proc. metall. Alboc., from Galicia. TheAugustanfreedmenprocuratorshere were apparently n directauthorityover both legionaryand auxiliary detachments. It remains questionable how far back we may retrojectthisevidencefromthe later second century. It would be difficultto believe thatthis role in themines was newly thrust upon the army at that time, although in Noricum propertywhichhad been let out to conductores as taken under direct Imperial control about then.119

ll Suetonius, Galba io.117Nash-Williams, op. cit. (n. io8), I4-I8, 45-I22.118 Procurators controlling Sierra Morena mines in

southern Spain: II 956, II79; Procurator aurariarumin Dacia: III I3II-I2; tabularius aurariarum

Dacicarum: III I297, I3I3; commentariensis inDalmatian gold mines: III I997; subprocuratoraurariarum in Dacia: III io88.

""G. Alf6ldy, Noricum (I974), 79, and'Patrimonium regni Norici ', Bonner Jahrbiicher 170(1970), I63. More procurators are now attested fromAstorga (AE I968, 227-34. Cf. Daniel Nony,

' A propos des nouveaux procurateurs d'Astorga,'ArchivoEspanfol e Arqueologia 3 (1970), I95-20I).

It is worth noting that the ducenary procurator inTarraconensis was responsible only for Asturia-Callaecia, probably again reflecting the importance

of the mines. Normally only such procurators of highrank might be expected to command troops, but atVillalfs detachments are stated explicitly to have beenunder various Imperial freedmen procurators. I amgrateful to Professor Dr. G6za Alf6ldy for commentshere.



62 R. F. J. JONES

Military links with mining are now quite well established from elsewhere in theEmpire. In Britain a prime example is the fort now proven at Pumpsaintbelow the Dolau-cothi gold mine.120The Mendip lead, and perhaps silver, mines are known to have been inproduction by 49, within six years of the conquest, and there are other lead pigs bearinglegionarystamps.121An unexcavatedearthworkat the Charterhouseead mine is likely to be

a fort.122In the north of Britainthere is explicit evidence linking the army and the miningindustry.123Elsewhere Tacitus recorded legionaries nvolved in silver mining in Germanyunder Claudius,124 hilst under Trajan a centurion of XV Apollinaris was in charge ofquarryingat Mons Claudianus n Egypt.125Even at Vipasca, in the south of the Iberianpeninsula, soldiers were sometimes present.126Troops would have been needed to controlthe slaves and convicts who made up the labourforce, as well as to guardthe productsof themines, even when the actual mining was in the hands of conductores.127et there must bea strong possibility that in the north-west, especially at the largest and most complexoperationssuch as las Medulas,the mines were never leased out, but were worked directlyby the Imperial administration.128The military would then have been called upon toprovide skilled labour, not sheer muscle power, and in particularthe high standards ofengineering technique in surveying and construction required in the massive aqueduct

systems.129 It could be argued that minerals are often found in the mountainous regionswhich also provided a natural home for the rebellious tribes that commanded the army'sfirst attention, and thus the minerals were always likely to be in an area under the army'scontrol. It has also been arguedthat Spain had ' no hostile frontiersto be defendedand sono permanent military zone '.130 The paradox remains that a substantial garrison wasmaintained,where otherwise the impression given is of growing prosperity and stability.Would the army have been withdrawn, or more severely reduced, had it not been for themines? Whereversecurity depended upon the army's presence, it was to be expected thattheir duties would have included control of the mines, or at least their supervision. How-ever, if the local populationgave no compelling reasons for keeping the armyin the region,as appears o have been the case in Spain, a more direct ink withthe mines maybe proposed.Such a policy may have had parallelselsewhere n the Empire; the army'srelationshipwithmining may well repay further attention, both in Spain and in other provinces.

APPENDIX: THE EPIGRAPHICEVIDENCE

NO. PLACE UNIT RANK DATE REFERENCE

i Abedes SWof legioVII Gemina Centurio - IRG IV IOO( II 2522)Verin)

2 Abiade lasTorres Miles mmunis - EEIX 302(N.ofPalencia)

3 AldeaNova nr. alaSabiniana Signifer EEVIII I28Miranda oDouro)

4 AquaeFlaviae legioVII Gemina Legate 79 II2477

(Chaves)5 legioII Augusta Signifer - EE IVp. i6 ( II 248o)

120 Excavated by G. D. B. Jones and J. H. Little.Noted with photograph in Britannia 4 (1973), 272

and pl. xxx, and again with plan, Britannia 5 (I974),398-9.

121 A.D. 49: VII 1202, cf. VII I20I (Claudian);XIII 3491 (Neronian), stamped L II; legio XX:VIII 1209, 6, I2I8; legio VI: perhapsXIII 2612aand b.

122 Plan by G. D. B. Jones and P. R. Lewis,Britannia 2 (1971), 277.

123The best evidence comes from the fort atBrough-under-Stainmore, where lead seals link

cohors II Nerviorum, stationed at WVhitleyCastle,with a mine, probably the lead mines at Alston (I. A.Richmond, ' Roman lead sealings from Brough-under-Stainmore,' Trans. Cumberland and Westmor-land Ant. and Arch. Soc. (2nd series) 36 (1936),I04-25, esp. I09). For a discussion of the evidence

from north Wales, see G. Webster, ' The lead-miningindustry in north Wales in Roman times,' FlintshireHist. Soc. Publications3 (1952-3), 5-33.

124 Ann. xi, 20.

125 III 25.126 II 5I8I, line 23.127 Condemnation to the mines as a capital punish-

ment, along with death, exile and deportation:Justinian, Inst. iv, I8, I-2. For conductores,see FIRA2i 104; Iura 2 (I95I), 127; II 5I8I.

128 For las Medulas, see works cited n. I: Lewisand Jones (1970), and Domergue, ' Introduction. . '.

129 For aqueduct systems, see Bird, and Lewis andJones, op. cit. (n. i).

180A. L. F. Rivet, ' Social and economic aspects,'in A. L. F. Rivet (ed.), The Roman Villa in Britain

(I969), 1I95.





6 AsturicaAugusta legio I Italica Trib. mil. II 2638(Astorga) (Veteran)

7 ,, legio II Adiutrix Veteran& Miles - II 26398 ,, legio VII Gemina Veteran 222-35 II 2640

Alexandriana .F.9 ,, ? egioVII Gemina Miles (Veteran) II 264I

Io ,, legio X Gemina Miles CMLeodn. i8, 7(= AE I928, I63)

II ,, legio VII Gemina Veteran CMLeodn. 20, I2

P.F. (= AE I928, i66)I2 ,, legio X Gemina Miles - CMLeodn. I6, I

(- AE I904, i6o)I3 ,, legioX Gemina Miles - CMLeon. I7, 6I4 ,, legioX Gemina Miles CMLonp. I9,8I5 ,, alaII Flavia .R. Eques CMLeon . ig, 9

(= AE I928, i64)I6 ,, ala Sulpiciac.R. Praefectus quitum- Alf6ldy i965 (- II 2637)

I7 ,, cohors V Miles CMLeon . 20,I 0Thracum (= AE I928, I65)

I8 ,, ?Miles CMLeonp. 20,II

I9 Babe (E. of ala II? Miles EE IX 277

Bragansa)20 Bracaraugusta legioVII Gemina Miles II 2425

(Braga) F.2I ,, ala? Eques II 56I022 Calahorra legioVI Victrix Miles II 298323 ,, alaTautorum Eques II 2984

Victrixc.R.24 Caldasde Reyes legioX Gemina Miles II 2545

(N. of Pontevedra)

25 ,, legioX Gemina Centurio IRG III 3826 ,, legioVII Gemina - IRGIII 3I

27 CastrelodelValle alaI Gigurrorum Decurio Garciay Bellido I961, I39(N.ofVerfn) (= AE I963, 27)

28 Castrocalb6n cohorsIV - EEVIII I3I; GarciayGallorum Bellido96I, I47 f.

29 Clunia legio X Gemina Optio centuriae - de PalolI969; GarciayBellido 97I, I5I (= AEi969/70, 274; ILS 9239)

30 Durat6n(nr. legio II Adiutrix Miles - II2763Sepulveda)

3I Fardaes W. of - ? Miles - EE IX 265Braga)

32 Fuente Encallada legio X Gemina Miles II 263033 ,, - Veteran - 11263I34 Gij6n legio IV Garciay Bellido I96I,

Macedonica I20; Fita I904, 8of.;Blizquez I962, I20

35 Herreradel legio IV Garciay Bellido I96I,

Pisuerga Macedonica figs. 3-8; idem I960

36 ,, cohorsI Gallica Praefectus - II 29I3

equitatac.R.37 Le6n legioVII Gemina Legatus II 266o

F.38 ,, legio VII Gemina Actarius 2i6 II 2663

Ant. P.F.

39 ,, legioVII Gemina Miles 234 II 2664Sever. AlexP.F.

40 ,, legioVII Gemina - Garciay Bellido 1970,

589-9I (25 tile stamps)



64 R. F. J. JONES


4I Le6n legioVII Felix - ibid.59I (i tile stamp)42 ,, legioVII Gemina - ibid59I-3 (26 tilestamps)

F. (_ II 6252 b, c, d, e)43 ,, legioVII - ibid.593 (3 tilestamps)

GeminaP.44 ,, legioVII - - ibid.593-4 (2 tile stamps)GeminaP.F. (- II 6252 a)

45 ,, legioVII Gemina - ibid. 594 (7 tile stamps)AntoninianaP. F. (= II 6252 p)

46 ,, legioVII Gemina - 235-8 ibid.595 (i tilestamp)MaximinianaP.F. (= II 6252 r, s. t, u)

47 ,, legioVII Gemina - 238-44 ibid.595-6 (9 tilestamps)GordianaP.F. (= II 6252 V,W)

48 ,, legioVII Gemina - 244-9 ibid.596(3tilestamps)PhilippianaP. F. (= 6252 X)

49 ,, legioVII Gemina - 249-5 I ibid.596-7 (I tile stamp)Deciana Traiana (= II 6252 y, Z)

50 ,, legioVII ibid.597 (I tilestamnp)Cl(audiana?)SI ,, legioVIIGemina - ibid.597 (2 tilestamps)

Qui(ntilliana?)52 ,, legioVII - II 5676

Gemina F.53 ,, legio ) Miles - 11568I

GeminaF54 ,, alaII Flavia Optio - II 5682; CMLeodn. 3I

55 ,, legioVIIGemina Miles - CMLeonp. 34 ( AEI928, i68)

56 ,, Custos rmorum - II 266857 ,, - Miles - II 2669

58 NearLe6n legioVII Imaginifer& - HAE 494GeminaF. Vexillatio

59 LucusAugusti legioVII Gemina Miles - II 2583(Lugo)

6o ,, cohors II ?Miles - II 2584Lucensium

6I ,, cohorsIII Not. Dig.Occ.42, 29

Lucensium62 ,, cohors - - 11258563 ,, legioVII Gemina Veteran II 257264 legioVIIGemina Veteran II 2582

65 ,, legioVII Gemina Vexillatiounder I38-6I GarcfayBellido 966, I

centurio&decurio (- AE I967, 229)

66 Luyego (i) legioVII io June Garcfay Bellido I966, 2Gemina i8i (A AE I967,230)(ii) cohorsI CenturioGallica

67 ,, cohors Gallica Miles Garcfa Bellido 96I, I48

(==AE i963,28)68 Quintanael legioVII Gemina EEIX 422, ia (i tile

Marco stamp)69 Rairfz e Veiga cohors Gallica Signifer Garcia Bellido959, 34,

(S.of Orense) no.9 (=- AE I960, I50)70 Rosinos e legioX Gemina Miles - AEI928, I79;

Vidriales CMZamorap. 507I legioX Gemina Miles - AEI928, i8o;

CMZamora. 5I72 alaII FlaviaHisp. Praefectusquitum VigilI96I, p. I04, I

c. R. (- AE I963, i6)73 SanPedro de la alaII Flaviac. R. - I97 Garciay BellidoI966, 4

Vinfa (= AE I967,237)





74 San Roman de legio X Gemina Miles - Garcia y Bellido I966, 3Cervantes prov.de Lugo)

75 Soto de la Vega cohorsIV 4I-54 HAE I035 (-AE I96I,

(N. of la Bafneza) Gallorum 345)76 cohorsIV 4I-54 HAE I036 (AE i96i,

Gallorum 345)77 cohors IV - 4I-54 HAE I037 ( AE I96I,

Gallorum 345)78 cohors IV - 4I-54 HAE I038 (= AE I96I,

Gallorum 345)79 cohors V - 4I-54 HAE I039 (=AE I96I,

Gallorum 345)8o cohors V - 4I-54 HAEIo4o(= AE I96I,

Gallorum 345)8i cohors IV - 4I-54 HAE I04I (=AE I96I,

Gallorum 345)

82 cohorsIV 4I-54 HAE I042 (=AE I96I,Gallorum 345)

83 Ujo (nr.Mieres, legio VII Gemina Miles - II 5733; EE III 29S. of Oviedo)

84 Valen9ado Minho legio VI Victrix ?Veteran II 2465 add.706

85 Vila Pouca legio VII Gemina Miles EEVIII io8 ( AE I907,

d'Aguilar N. of I50)Vila Real)

86 cohorsI Gallica Miles EE VIII I09(= AE I907,

equitatac.R. I5I)87 Near Vila Nova de legio VI Victrix Miles - II 2374 555I

Famil9ao88 Villalis (i) legioVII Vexillatiounder Io June II 2552; ILS 9I25;

Gemina centurio,signifer I63 CMLeonp. 70; AE I9I0,3(ii) cohorsI DecurioCeltiberorum

89 ,, (i) legio VII Signifer 22April II 2556; ILS 9I29;Gemina i65 CMLeonp. 7I; AE I9Io,6(ii) cohorsI GallicaMilites under

centurio90 ,, (i) legioVII Imaginifer I5 Oc- II 2553; ILS9I27;

Gemina tober CMLe6np. 72; AE I9I0,4

(ii) cohorsI Celti- Vexillatioand I67

berorum Tesserarius(iii) cohorsI CenturioGallica

9I ,, (i) cohors I Gallica Milites 22 April II 2555; ILS 9I28(ii) cohorsI Celti- Decurio I6I-9

berorum92 ,, cohorsI Gallica Milites & centurio 22 April ILS 9I30; AE i9io, I;

I75 CMLeodn.7393 ,, (i) legio VII Vexillatio Io June II 2554; ILS 9I26; AE

Gemina I84 I9Io, 5; CMLen p. 74(ii) ala II Flavia Decurio

94 ,, (i)cohorsI Milites 22April ILS9I3I; AEI9IO,2;

Gallica i9i CMLeodn. 75(ii) ala II Flavia Decurio

BOUNDARY TONESOF LEGIOIV MACEDONICA95 Hormiguera legio IV 23 B.C.- Garciay Bellido I956, I

Macedonica A.D. I4

96 Henestrosade las legio IV - 23 B.C.- Garciay Bellido I956, 2

Quintanas Macedonica A.D. I4



66 THE ROMAN MILITARY OCCUPATION OF NORTH-WEST SPAIN


97 Cuena legioIV 23 B.C.- Garcia Bellido 956, 3Macedonica A.D. I4

98 Cuena legioIV 23 B.C.- Garcfay Bellido I956,4Macedonica A.D. I4

99 San Vitores legion IV 23 B.C.- Garciay Bellido I956, 5Macedonica A.D. I4

IOO Villasidro legio IV 23 B.C.- Garciay Bellido I956, 6Macedonica A.D. I4

ioi Unknown legio IV - 23 B.C.- Garciay Bellido I956, 7Macedonica A.D. 14

I02 Unknown legio IV - 23 B.C.- Garciay Bellido I956, 8Macedonica A.D. I4

103 Henestrosade las legio IV 23 B.C.- Garciay Bellido I956, 9Quintanas Macedonica A.D. I4 (nowlost)

I04 Las Quintanas legio IV 23 B.C.- Garciay Bellido I956, IO

Macedonica A.D. I4 (- II 29i6 c; nowlost)I05 Las Quintanillas legio IV 23 B.C.- Garciay Bellido I956, ii

Macedonica A.D. I4 (now lost)io6 Reinosilla legio IV 23 B.C.- Garciay Bellido I956, I2

Macedonica A.D. I4 (now lost)I07 CastrillodelHaya legio IV 23 B.C.- Garciay Bellido I956, I3

Macedonica A.D. I4 (now lost)io8 El Haya legio IV 23 B.C.- Garciay Bellido I956, 14

Macedonica A.D. I4 (now lost)

STONESNOT PRECISELY OCATED

I09 Galicia ala II Flavia Praefectus II26ooHisp. c.R. equitum

i io Asturias ala II Thracum HAE904 (= AE I96I,

338)Iii EasternAsturias ... Asturumet Diego Santos, I63, no. 62

Luggonum (-AE I963,27)

The BowesMuseum,BarnardCastle,Co. Durham



JRS vol. LXVI (1976) PLATE IV

I~~~~~~~~~~~~~A4X, i'.'. Z

OF# HEEFESIV WALL

Ptogr b the t o. t r

(T) VILLTATIS:DFAGNIE4NT' OF TNNSCRIPTtION, BUILTTr ITOT A WXATL(e p 556) (2)3 ROS-ITTNOSn- 7%DTEIDIAES DEqTTI



JRS vol. LXVI (1976) PLATE V

$: - : t- ip t

(i) ROSINOS DEVIDRIALES~~~~~:: TH BUIDIN AGIS THE DEESV WAL; (z THE WAL OF ASTRG

Photographs by the aut.ho. Copyig r ve N ^ ?

; ; ? > f X~~~~~~~~~~~~~~~

C,*Da$: \ HF2d ~KS w~ f; *,4tt 0 A,:

W1og_Sf ~~~~~7 A'*f

@* :s: + tf^: < #:4 < .tt? :-

~~~~~~~~~~0 {, r - S>

Xh ; X

X~~~~~s fS at a V x f f9 tfl . . S t 5 t z - + f ^ J f t X f

?s; , f f S f .Xv+ 9, M J\ f . vy r ' - ;

M ' > z

w>'es^ MSSltiM o f;> >eyA> WS ,, a~~~~~~~~~~~~~~~~~~~~~~~~~~~~

r g :'f Cs i | t} a: t s : - e > : ^ i s g'4 0*:4' -W, Z -,

sttt -$g 'Ms'S ':' ,''t ' , . ..v;

*tsA;;4s,tf>

f s> w sF .X s s > s*452 yPsfSR,.> s 3s;;-*-8~33tA . ^ &

$ 4: . ,> > :3W 'fe ; sa: 0; . Ss S f

Phtorah by the auhr.Cprgteere



JRS vol. LXVI (1976) PLATE VI

1~~~~~~~~iHE WALLS OF: LN (. THE WALS OFiaLLGO)

Phtgrps byteautho. Coyigt reserve

E""'.,

-0 ; 0,

_i:~~~~~~~~~W

.71 - tt

SEt-w~~~~~~~~~~~~~~~~~~~~~~~~~~~~4. 'w' v = g~~~~~~~~~~~~~~~~~N11

jones roman millitary nw spain

Documents