* +

HESPERIA 72 (2003) T H E D U R R E S R E G I O NA L Pages4I-II9 A RC H A E O LO G I CA L P ROJ E CT ARCHAEOLOGICALSURVEY IN THETERRITORY OF EPIDAMNUS/ DYRRACHIUMIN ALBANIA

ABSTRACT

In the springof 2001the hillyuplands immediately northwest ofthe modern cityof Durreswere for the firsttime investigated using the techniquesof in- tensivesurface survey. In total,an areaof six squarekilometers was explored and twenty-ninesites were defined, most of them new.Remains of Greek antiquitywereplentiful and include unpublished inscriptions and graves. One sitemaybe the location of a previouslyunknown Archaic temple. Included in thisarticle are descriptions ofthe areasinvestigated, alist of sites,and a cata- logue of the most diagnosticartifacts recovered. Patterns of settlementand landuse arediscussed and comparedto those recordedby othersurveys in Albania.

ModernDurres in centralAlbania is todaythe largestseaport in the coun- try and is linked by a superhighwayto the capital,Tirana, 35 km to the east (Figs. 1-2).1 Durres is the location of Epidamnus/Dyrrachium,a Greekcolony that was foundedin the later 7th centuryB.C. by Corcyra with supportfrom Corinth and other Dorian cities.2The ancientpolis centerhas long been the objectof archaeologicalresearch, and is partic- ularlywell known to ancienthistorians both as the site of a revolution that constitutedone of the proximatecauses for the outbreakof the Pelo- ponnesianWar (Thuc. 1.24-29) and as the site of a majorbattle between

. . . . 1 1. Davis,Hoti, Pojani,Stocker, and especiallyto MariaGrazia Amore and ourmlsslon ln everypOSSl Dleway. Wolpertare jointly responsible for the LorencBejko for their assistance in 2. Thuc.1.24. On the dateof the overallauthorship of thispaper. Ache- planningfieldwork. We alsothank the colony,see Eusebius(Schoene 1866, son contributeda reporton the results Universityof Cincinnatifor its support p. 89, line 1392).Jerome's continuation of surveybyTeam B. Appendix2 and of ourresearch. We thankOls Lafe of Eusebiusplaces the foundationin mostof the catalogueddescriptions of andMaria Grazia Amore for help in 627 B.C., slightlyearlier than the textof artifactsare the workof Hayes.Unless proofreadingthis report and Lida Eusebiusitself (Helm 1956, p. 97b).In otherwisenoted, all photographsare Mirajfor her hospitality in Durreson the remainderof this article,Epidam- fromthe projectarchives. manyoccasions. Fieldwork was con- nus/Dyrrachiumis referred to more We aregrateful to RichardHodges ductedwith permission from the simplyby the nameof the moderncity, (Universityof EastAnglia), Henry Instituteof Archaeologyin Tirana; Durres. Hurst(Cambridge University), and MuzaferKorkuti, its director,facilitated

American School of Classical Studies at Athens is collaborating with JSTOR to digitize, preserve, and extend access to Hesperia ® www.jstor.org 42 JACK L. DAVIS ET AL.

t: ;: f:'

: 0 L9h \ T: s ) i E

Apoll

Figure1. Map of Albaniashowing principalcenters of Greekculture. R. J. Robertson

JuliusCaesar and Pompey (Caes. B Civ. 3.35-61)at Petrain 48 B.C.3 A steepridge forms a backdropfor the modernport. The top of the south- ernmosthill is todayoccupied by the palaceof the formermonarch of Albania,Ahmed Zogu (Fig. 3). It is clearthat much of thepresent city has 3. Heuzey1886, pp. 35-88. An- beenbuilt over ancient remains. cientsources for the historyof Durres arereproduced, translated, and dis- A historyof modernresearch in Durresmight properly begin in 1861 cussedin Cabanesand Drini 1995, withthe explorationsof LeonHeuzey, one componentin a FrenchMis- pp. 19-33, andare listed in Myrto sionto Macedonia,and with the description of the LateRoman circuit of 1998,pp. 76-86. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 43

Figure2. City and port of the mod- ern city of Durres.From Tower F, lookingsoutheast.

Figure3. Palaceof Ahmed Zogu. Lookingnorth-northwest from the new ArchaeologicalMuseum of Durres.

fortificationspublished by Heuzeyand Daumet in 1876 (Fig.4).4 The nextsystematic attempt to inventorythe antiquities of theancient city was thatof the AustriansCamillo Praschniker and Arnold Schober.s Soon after,a Frenchexpedition in 1925conducted soundings in variousloca-

4. Heuzey1886, chap. 2, pp.43-55; fortificationshave been identified logicalresearch in Durres(complete Heuzeyand Daumet 1876, pp. 349- (Gutteridge,Hoti, andHurst 2001, through1993), see the annotatedbib- 392. It remainsodd that as yet no pp.395, 406, n. 23). Forthe most liographyof Myrto(1998, pp. 76-103). tracesof Hellenisticor earlierRoman comprehensivediscussion of archaeo- 5. Praschnikerand Schober 1919. 44 JACK L. DAVIS ET AL.

Figure4. Detail frommap of the Durresarea. After Heuzey l886, plan4

. -fuk

Figure5. Architecturalterracotta foundin the moderncemetery of Durres.Courtesy Institute of Archaeology, Tirana tionsand determined that, in thecentral zone of themodern city, levels of 6. Rey1925. Greekdate lie some5 m beneaththe currentstreet.6 7. See Hammond1967, pp. 425- SinceWorld War II, remainsof a Romanbath, a macellum,and an 426, 469-470,for the foundationof amphitheaterhave been excavated, and the course of anaqueduct of Trajanic the originalcolony, the uncertaintyof datehas been documented. Evidence for the Greekcity is moreexiguous its location,and what little is known andvirtually nothing is knownabout the sizeand precise location of the of its earlyhistory. originalcolony.7 Pre-Roman remains include only a Hellenistichouse; an 8. See,with earlierreferences, Miraj 1994(bath); Hoti 1989b,1996 (macel- altarof the 4th centuryB.C. discoverednortheast of the portin drained lum);To,ci 1971 and Miraj 1990 (am- marshland;and architectural terracottas of WesternGreek Archaic style phitheater);Miraj and Myrto 1982 foundin thearea of themodern cemetery of thecity (Fig. 5).8 Large num- (aqueduct);Tartari1988 (Hellenistic DURRES REGIONAL ARCHAEOLOGICAL PROJECT 45

Figure6. Generalmap of theDurres area.R. J. Robertson house);Myrto 1989 (altar); and Zeqo bersof gravesand grave markers of Archaicthrough Early Roman date 1989(architectural terracottas). havebeen located in the valleysnorth and northwest of the cityon the 9. Villais the hill of the palaceof AhmedZogu. slopesof the hillsof Dautaj,Kokoman, and Villa (Fig. 6).9 Other graves 10.To jci 1976; Hidri 1983, 1997; havebeen found closer to themarsh in thearea between Spitalla and Porto Tartari1987; Myrto 1998, p. 99. Romano.l° 46 JACK L. DAVIS ET AL.

Figure7. Viewof thecity of Durres fromtract A077. Petra at far right in TH E ARC HAE O L O G I CAL SU RVEY thedistance. Looking southeast.

To the northwestof theport of Durres,between the city and the modern villageof PortoRomano, are uplands, ca. 10 km2 in extent.These formed theimmediate hinterland ofthe ancient polis and include the ancient cem- eterieson the hillsof Dautaj,Kokoman, and Villa (Fig. 7). Untilrecently the areawas sparsely settled and, although frequently the targetof visits, extensiveinvestigation, and excavation by personnelbased at the Durres museum,it hadnever been systematically explored. In 1999it wasclear, however,that antiquitiesin theseuplands were being destroyed at an acceleratedrate as the resultof illegalsettlement of immigrantsfrom northernAlbania. The ideaof organizingan intensive archaeological sur- veyhere first began to be discussedamong members of theInternational Centerof AlbanianArchaeology in Tiranain the fall of thatyear, and preliminaryreconnaissance was carried out in a smallarea near the mod- erncemetery of the city.llSeveral concentrations of pottery and tile were foundthere by LorencBejko and Maria Grazia Amore of the Albanian RescueArchaeology Unit, as well as a much-disturbedburial containing the neckof a red-figuredamphora and two smallblack-glaze oinochoai (Fig.8).12 Fromthis auspicious beginning arose the DurresRegional Archaeo- logicalProject (DRAP), a rescuemission intended to serveas a preludeto

11. Bejkoand Amore 2000, pp. 45- below,A048-02 from S007). Added The scenesare framed above and below 46. miltosseems to havebeen employed to by a bandof egg motifs.On the shoul- 12. RobertF. Suttonhas examined imitatethe colorof Atticred-figure dera bandof tonguemotifs is pre- photographsof thesefinds and com- pottery,and a thin,dull black glaze was served.The amphoraappears to date mentsthat the fabricof the amphora usedfor the partlyfugitive background. to the lastthird of the 5th centuryB.C. appearsto be Corinthian(or a related Erosis illustratedon bothsides of the The dateof the oinochoaifound by fabric),like that of a smallchous-type neck:on one sidehe is shootinga bow, Bejkoand Amore is consistentwith oinochoefound during our survey (see andon the otherhe is seatedon a chair. thatof the amphora. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 47

Figure8. Findsfrom the vicinity of thecemetery of Durres:red-figured amphoraneck (two views); small oinochoe.Courtesy L. Bejko,M. G. Amore,and International Center of AlbanianArchaeology in Tirana

a long-termprogram of researchin the Durresarea. The projectis spon- soredby the International Center of AlbanianArchaeology in Tirana,an instituteestablished in 1999and funded by the Packard Humanities Insti- tute(PHI).13 In August2000, Davis, Stocker, and Pojani, together with AdamGutteridge of CambridgeUniversity, visited the areato be exam- ined.Fieldwork was conducted between March 10 andApril 5, 2001.14 A totalarea of ca. 600 ha wasinvestigated by two teams,generally con- sistingof fivefieldwalkers and a teamleader, in thirteenand a halfdays of fieldwork.l5

13. PHI is particularlycon- AfrimHoti (codirector,Institute of cinnati),and Aaron D. Wolpert(field cernedto ensurethe preservationof Archaeology),Genci Kotepano (driver, director,University of Cincinnati). Albania'ssignificant archaeological InternationalCenter of Albanian 15.Team A wasled byWolpert, heritage. Archaeology),Rovena Kurti (field- TeamB byAcheson. The totalarea 14.Members of the projectin- walker,University of Tirana),Ols Lafe investigatedwas calculated in two cludedPhoebe E. Acheson(team (fieldwalker,University of Cincinnati), differentways with nearlyidentical leader,University of Cincinnati), ElvanaMetalla (fieldwalker, Institute results.Using the firstmethod, we Valbona,Coko (fieldwalker, University of Archaeology),Iris Pojani (codirector, calculatedthat 598.75 ha weresur- of Tirana),Siriol Davies (fieldwalker, InternationalCenter of Albanian veyed:the areawas computed as ([the Universityof Cincinnati),Jack L. Archaeology),Eduard Shehi (field- totallength walked in the courseof Davis(codirector, University of walker,Institute of Archaeology), the projectby all teammembers] x 15 Cincinnati),Rodney D. Fitzsimons BrikenaShkodra (fieldwalker, Insti- [spacebetween fieldwalkers])/10,000. (fieldwalker,University of Cincinnati), tuteof Archaeology),Sharon R. The secondmethod, which summed AdamGutteridge (fieldwalker, Cam- Stocker(codirector, University of the areasof all polygonsdefined as bridgeUniversity), John W. Hayes Cincinnati),John L. Wallrodt(com- tracts,yields a totalsurveyed area of (ceramicanalyst, Oxford University), puteroperations, University of Cin- 600 ha. 48 JACK L. DAVIS ET AL.

In the followingsections of thisarticle we firstdescribe the methods employedby the surveyand, in retrospect,briefly evaluate their effective- ness.l6We discussin geographicalorder from north to souththe water- shedsor catchment areas investigated by the project and our discoveries in each,particularly the majorconcentrations of artifacts(sites) that have beenidentified. Next, the natureof the artifactscollected by ourteams is examined,as are overall patterns in thedistribution of materialsof specific dates.Finally, we attachtwo lengthy appendices in whichsites and other findspotsare described in detailand where a selectionof artifactsfrom themis presented.Extensive deposits of ceramicsfrom excavations in the cityof Durresand from its cemeterieshave already been published in A1- banianjournals, but these are not readilyavailable to foreignscholars. A relativelyfull description of findsfrom the hinterland of Durreswill, there- fore,be of valueboth to ceramicspecialists and to thosewho mayin the futurepursue the studyof landscapearchaeology along the eastern littoral of theAdriatic Sea.

FIELD METHODS AND SURVEY STRATEGY

In formulatingthe research strategy for DRAP, we didnot make any prior assumptionsabout the locationof ancientremains; our goal was to cover theentirety of thetargeted area so thatlarge-scale patterns in thedistribu- tionof artifactswould be mostintelligible. Although it wasour intention to collectremains of allperiods of thepast, we hopedin particularthat our resultswould supplement the archaeologicalrecord for the Graeco-Ro- manperiod, and in particularthe historyof the Greekcity. The highest priorityfor fieldwork was the construction of a GeographicalInformation System(GIS) that would permit a comprehensivemap of surfaceremains 16. Fieldand museum procedures of pasthuman activity to be drawn.17 usedwere broadly similar to thoseem- A highcoastal ridge (187 maslat its highestelevation) runs north- ployedby the NemeaValley Archaeo- southfrom the moderncity of Durresto PortoRomano, paralleled by an logicalProject (Wright et al. 1990),the earthentrack high on its easternslopes (Fig. 9). Today the southernparts PylosRegional Archaeological Project of this areaare called Spitalla, the northernparts, Porto Romano. There (Daviset al. 1997),and the Mallakastra aretwo modern reservoirs in theSpitalla area. The present cemetery of the RegionalArchaeological Project (Kor- kutiet al. 1998). city is alsolocated here and, farther to the north,there is a largernew 17.The DigitalElevation Model cemeterythat will soonbegin operation. The surveyarea was defined by (DEM)used for the project'sGIS was the seaon thewest, by the Durres-PortoRomano road and the Durres- producedby ChristopherDore of PortoRomano irrigation canal on the east,by the palace of AhmedZogu ArchaeologicalMapping Specialists on the south,and by the endof the coastalridge at PortoRomano on the (Berkeley)at a scaleof 1:10,000from Albanianarmy maps. Wallrodt north.This areais, of course,only a fractionof the totalhinterland of managedthe computersystems Durres,other parts of whichmay profitably be investigatedin thefuture.18 employedby the project. The surveyarea is underlainby sedimentaryrocks of Plioceneand 18. See the discussionat the endof Mioceneage andby alluvialfan depositsdating to the LatePleistocene this reportwhere possibilities for future and EarlyHolocene. Most extensiveare Pliocene-aged Helmesi Suite researchare discussed. In 2001we did depositsthat consist of mudrocks,siltstones, and sandstones. The steeply not surveyareas of densesettlement nearthe Durres-PortoRomano road or slopingridge in the westernpart of the surveyarea marks a geological the drainedmarsh east of the road.For contactbetween the youngerPliocene sediments and an olderMiocene- antiquitiesfrom the marsh,see Myrto ageddeposit that consists of reduced,gray, deep-water marine sediments 1989. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 49

Figure9. Westend of zone8 with eastside of coastalridge in the background.Looking southwest.

composedof mudrocks,siltstones, and sandstones.Deep outcropsof clayare commonly found in theMiocene deposit. The gypsumalso con- tainedin thisolder Miocene-aged deposit was formed in a periodof low sea levelwhen seawater in the Mediterraneanbecame sufficiently con- centratedto precipitatesuch a mineral.The smallvalleys on the eastern side of the surveyarea are underlain by LatePleistocene to EarlyHo- locenealluvial fan deposits,largely formed from local reworking of the olderPliocene bedrock. The fansconsist of coarsecolluvium and finer slope-washdeposits.l9 Beforethe startof fieldworktwelve topographic zones were defined (Fig.6). The definitionof the zonesdid not reflectancient settlement patternsor modern land use. Zones 1-11 arecatchments that for the most partrun from the high coastal ridge down to thedrained marsh east of the Durres-PortoRomano road. Zone 12 comprisesthe entiretyof the steep westernslopes of the coastalridge. The basicunit of analysiswithin each zonewas the "tract,"a polygon defined on the spotby the leaderof the 19.We thankMichael E. Timpson surveyteam, who took into account natural topography, man-made bound- forthe geologicaldescription included aries,soil conditions, and current vegetation. Vegetation and modern land in thisparagraph. usewere held uniform, and standardized categorical values were used to so JACK L. DAVIS ET AL.

lS x )k Q3

a b Figure10. a) Vegetationin the measureattributes within each tract (Fig. 10). Tracts were also kept con- surveyarea; b) landuse in the sistentlysmall so thatsubsequently it would be possibleto producea fine- surveyarea. R. J. Robertson grainedpicture of therelationship between vegetation, land use, and arti- factdensity, and to mapthe findspots of specificartifacts with considerable precision.20 Onceboundaries were defined, each tract was traversed by five to six fieldwalkersspaced 15 m apart.Close spacing proved particularly impor- tantsince the landscape is verystable and often only small erosional expo- suresoffered windows of visibilitythrough which the existenceof large artifactscatters could be surmised.Team members counted artifacts that theyobserved and then reported those counts to the teamleader at the endof eachtract, when they were recorded in a commonnotebook; this informationwas entered in a database each evening. Team members were instructedto collectall sherds of pottery that preserved part of a rim,handle, 20.Tracts were almost never larger base,traces of decoration,or thathad an otherwisedistinctive feature or than2.0 ha (99%< 2.0 ha;96% < 1.5 fabric;and all smallfinds, including all examplesof chippedand ground ha;and 84% < 1.0 ha). DURRES REGIONAL ARCHAEOLOGICAL PROJECT 5 I

Figure11. Overall premodern artifactdensity in thesurvey area. R. J. Robertson

stone.Only those artifacts that were obviously modern (i.e., post-World WarII) were to be leftin the field.The exactposition of notableartifacts andfeatures such as graveswas documented in fieldnotes, for example, "walker04 observeda tilescatter at 25 m."The teamleader also sketched eachtract and recorded the directionthat walkers walked, as well as their order.The outlinesof tractswere digitized at theend of eachday of field- work.Artifact densities were then mapped for each tract by dividingthe 21. Artifactdensity was calculated numberof premodernartifacts reported by fieldwalkersby the distance as (thesum of allpremodern artifacts observedin a tract)/(thesum of the thatthey walked (Fig. 11). In total,938 tractswere defined. Twenty-nine distancescovered by allwalkers in the anomalouslydense concentrations of premodern artifacts were labeled as tract).In thisreport sites are identified sites.2l by a three-digitnumber preceded by Fieldwalkingbegan in zones6 and8, the largestcatchments in the the letter"S."In mostinstances sites surveyarea; these now feed the two reservoirs in Spitalla.By beginning in wererevisited, but severe time con- straintsmade it impossiblein 2001 to thesebroad valleys it waspossible for team leaders to estimatehow long it collectsurface artifacts according to any wouldtake to surveythe entirearea and accordingly to set subsequent systematicprocedure. dailywork schedules. These zones were also relatively easy to surveyand 52 JAC K L. DAVI S ET AL. allowedinexperienced fieldwalkers to be trainedadequately before teams tackledthe moredifficult urban fringes and the areaof the ancientcem- eteriesknown to existin zone9 andfarther south. Aftersurveying zones 6-8 andthose parts of zones9 and10 notcov- eredby modernconstructions, the teamsmoved into zones 1-5, working fromthe passof Xhamadhanorth to PortoRomano. The two teamsfor the mostpart walked alternating zones and worked in tandemin zones5 and7, thusensuring that any differences in aptitudesand abilities between fieldwalkerswould not bias results. Survey ofthe primary catchments (zones 1-10) wasfinished in the secondweek of fieldwork,and sufficient time remainedto examinethe entiretyof the seawardside of the coastalridge (zone12), with the exceptionof dangerouslysteep slopes at the highest elevations.22

EVALUATION OF FIELD DATA

Therewere very few areasthat could not be examinedowing to precipi- tousslopes or impassablevegetation. Even the steepestslopes were ter- raced,and only cliffs near the crestof the coastalridge and rockslides on its westface showed no evidenceof recentmodification for agricultural purposes.We wereconcerned, however, about the extentto whichgrass andclover in meadowsand on abandonedterraces would obscure surface artifactconcentrations. In comparisonwith more arid Aegean landscapes for whichthe surveytechniques applied to the Durreshinterland were developed,conditions were indeed much worse, and no doubtaggravated bythe fact that the survey took place when spring growth was at its height. Averageground visibility was just 33% in thesurvey area on thewhole and was30% or less in 73%of alldefined tracts (683 of 938 tracts)(Fig. 12). Despitethe generally poor visibility, premodern artifacts were recorded in 645tracts. There is, however, a strong positive correlation between good visibilityand elevated artifact densities, both between and within tracts. We concludethat variability in the visibilityof artifactsis largelydeter- minedby currentpatterns of landuse and by patterns of erosion.Almost everyplowed field yielded at leasta modestpeak in premodernartifact density.On the otherhand, erosion only infrequently occurs on a scale largeenough to exposeextensive scatters of premodernartifacts. Indeed, thereis relativelylittle active erosion in the Durresregion compared to southernGreece, probably because the facesof mostagricultural terraces havebeen stabilized by plants and the humidenvironment supports rela- tivelylush vegetation. It is thusunlikely that observed artifacts have been transportedlong distances. Continuousdisturbance of sedimentsis foundonly on the highest partsof the coastalridge, where, because of the overallsteepness of the terrain,even terraced slopes are heavily eroded. Elsewhere, active erosion is generallylocalized, resulting from several activities: recent bulldozing, extensivein thevalleys of Spitalla;bunker construction in the1970s, which triggeredsubstantial downslope erosion (Fig. 13);the buildingof field roads;and the diggingof drainagegullies, some of whichare incised to a 22. Zone 11 is entirelyurban and no depthof nearlya meter.In areasheavily browsed by livestock, erosion also surveywas possible in it. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 53

)i >D

Figure12. Overallground visibility in the surveyarea. R. J. Robertson

occursalong animal paths, many of which,like the gullies,have become deeplyincised and serve as channels for rainwater. On stableterraces often the onlyartifacts collected came from gullies, roads, or paths (Fig. 14). Theseconditions of the contemporarylandscape have serious conse- quencesfor the interpretationof the resultsof survey.In particular,it is difficultto be certainthat a lackof surfaceartifacts reflects the absenceof subsurfaceremains. It seemslikely that we havenot beensuccessful in findingartifacts in manyplaces because they are covered by stablesedi- ments,and gaps in surfacedistributions may not reflectany underlying patternin the locationof premodernremains. Artifact concentrations at 23. On artifact"halos"in the south- Durresdo not havea distinctcenter surrounded by "halos"of decreasing ernAegean and their interpretation, seeAlcock, Cherry, and Davis 1994, artifactdensity. Densities in mosttracts are either very high or very low; pp.141-142; Snodgrass 1987, pp. 113- the median"halo" tracts recognized by manysurveys in the Aegeanare 117. mlsslng.* * 23 54 JACK L. DAVIS ET AL.

Figure13. Collapsingbunkers at the east edge of tractA228. Lookingwest.

Figure14. Terracescovered with heavyvegetation. Tracts B036 and B037, lookingnortheast.

Sincecounts reported by fieldwalkerswere recorded separately it is alsopossible to assessthe extentto whichan individual's skill and experi- encemay have affected our perceptions of patterningin thedistribution of artifacts.Such "walker effects" have been studied by othersurveys.24 Our ownconclusion is thatthey have had a negligibleeffect on ourinterpreta- tionof results.For each team member in eachtract raw counts of premodern artifactswere standardized by dividingthem by the distancewalked. The artifactdensity thus calculated for eachindividual was compared to the averageddensity for all those who walked that tract. Finally for each indi- vidualthe sumtotal of suchdifferences in alltracts was averaged to pro- ducea statisticthat represents the extentto whichany individual was out of stepwith other team members. The statisticfor five of a totalof fifteen individualsdeviated from the averageby morethan one standarddevia- tion,but only in one instancediffered by morethan two standarddevia- tions.Two of thesefive individuals also recorded fewer post-World War I artifactsthan their colleagues, perhaps in partexplaining why their counts of premodernartifacts were higher than those of theirteammates. A secondproportional statistic, calculated by dividing the average den- sityfor each team member in eachtract by the averagedensity calculated 24. See Cherry,Davis, and forthe entireteam that examined that tract, can be employedto correct Mantzourani1991, pp. 37-53; forsituations where a walker worked only in placeswhere more premodern Shennan1985, p. 35. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT ss

artifactsthan the averagewere found. In suchcases simple subtraction of the numberof premodernartifacts counted by thewalker from the aver- agerecorded by his teammatestends to yielda higherabsolute difference thanin areasthat arelower in density,although the percentageof the differencemay be lesssignificant. The range of variationwhen individuals arescored on thisstatistic leaves only three of the fiveoutliers. The samplesize for two of theseteam members is small(80 and73 tractswalked, compared to an averageof 271 tractsper walker). These individualsappear to havedeviated from the meanto thisdegree because theywalked more tracts in areaswhere premodern artifact densities were genuinelyhigher or lower than the average. Counts of onlyone walker ap- peargenuinely aberrant. The factthat this individual found more artifacts thanother walkers might be a concernexcept that his artifactcounts are generallycorrelated with those of his teammatesand there is no instance in whicha sitewas defined on thebasis of his observationsalone.

DESCRIPTION OF ZONES AND THE RESULTS OF FIELDWORK

ZONE I Two distinctareas in zone 1 wereintensively surveyed (Fig. 15:a),one adjacentto zone2, the otherconsisting of the northernmostparts of the coastalridge (83.3 masl maximum height).25 Houses, many of themre- centlybuilt, now cover the lower half of theslope of theridge and much of the remainderof zone 1 is occupiedby a navalbase to whichit wasnot possibleto gainaccess. Terraces on the eastern slopes of theridge are mostly stableand grassy. The cresthas been heavily bulldozed and scrub brush is dominanton erodingoutcrops. At its northernmostedge the ridgeis ex- posedto strongwinds and storms, and steep cliffs containing thick depos- its of fineclay fall to thesea. A segmentof a mortaredbrick wall of Roman stylewas foundhere (S006) (Fig. 16).26Elsewhere finds were concen- tratedin andnear bottomland, especially in plowedfields.

25. Figures15:a and 15:bshould be 26. S006:Periods represented: MR?, beginningof the lastcentury and was consultedthoughout this section.The R;ranges represented: A-HL, A-R, describedand illustrated by Praschniker followingabbreviations are used to CL-HL, R-MED, MR-LR,LR-DA, andSchober (1919, p. 46 andfig. 58). describeperiods: Prehistoric (PH); LR-EB.This wallsegment is high Accordingto them,the wallentirely Neolithic(N); BronzeAge (BA); abovethe remainsof the well-known blockedentry to the uplandsnorth of Archaic(A) 700-480 B.C.; Classical long Romanwall at PortoRomano Durres.It rannorth from the crestof (CL) 48s323 B.C.; Hellenistic(HL) (Karaiskajand Ba jce 1975; see Gut- the coastalridge downhill to a small 323-31 B.C.; Roman(R) 31 B.C.- teridge,Hoti, andHurst 2001, p. 394, bay,then continued east for another A.D. 610;Dark Ages (DA) A.D. 600- whereit is assigneda datein the 4th 200 m beforedisappearing into the 850;Byzantine (B) 850-1450;Medi- centuryA.C.), andboth must have been marsh.Heuzey (1886, p. 46) described eval(MED) 600-1450;Ottoman (O) partof the samedefensive system. the wallas "encorepercee d'une grande 1450-1912;Modern (M) 1912-2000; Hayessuggests that the systemis likely porteen arcade."Underwater explora- andunknown (UNKN). All of these to dateto the 3rd-4thcentury, probably tion in the 1980sin the smallbay re- periodsmay be furthersubdivided into nearthe beginningof thisperiod; he vealedextensive remains of the 4th- Early(E), Middle(M), andLate (L) alsoobserved that part of an original 2nd centuryB.C. (Cekaand Zeqo phases:e.g., LCL is the abbreviation gatewaythrough the wall is preserved. 1984).See alsoHammond 1967, forLate Classical. This gatewaywas complete at the p. 470, n. 2. \ t > / g9g--z(g t\ northernpartofthesurveyarea.

56 JACK L. DAVIS ET AL.

R >,,',.'--, ,,,,i > *-B ->_5 M+6 ''t,B2'43ti233,, '>,,,c,,---"'^'''X$',B24i''' '- >- - \ FiFre15:a. Tracts and sites in the , o , ...... 1 ,, \ , / 1 i ' . ''\x1i ' B234, B242 '.'/' ' S ---- \

/ / / > \ . X \ > R.J. Robertson

ZONE2 Gardens,olive orchards, and fruit groves lie at theeastern ends of thelow ridges(77 masl and 93 masl)that define zone 2 on thenorth and south. A springin tractB293 supplies water to residentsof themore densely popu- latedzone 1. The flat,scrub-covered bottomland slopes gently from west to eastand is usedfor grazing.Higher slopes on the southernridge are mostlyterraced, with some olive and fig trees;terraces are largely covered bygrass and some brush grows on theirfaces. The northernridge has few trees,less well definedterraces, and more scrub brush. This sideof the fE\\\ ) ...... 352tt A552 B3550/B356;- B357 / [S...... A26') K .... ;B155 .2 1BS))A561. .-. E. .BR -A24 A...... A111...... - \A°63 B079B194B1618162;a ...... i4A112,', &_ . BtX)51,iB ,S: ZA27B3199;/', \ /1, ' ='...... }1iH'A164. \ o A183,° F/A279A272S :.AOS9 B107A165A29 *--E' B130A158....., -; . .B18a9/647CAOZ3A284W. E A156...... A°1SO /AO48AOS2 - A267. AO27AO14: A1

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 57

a l \ )\!!- RS;fefN0¢f;w>:gHEWd>:;:0f0\ff00A \\Da:f: X . X ; B064 ,0 . B074 . ' , ' B09S - B0697 B046 B047 -. B101 . B024 \ \; A560 \ i-- 0 B077 q l;°; \

g . A562 '-0 B061 . B080. B088 iu . B125 Brog;\B1270 3 0 2 t

f --. A563 . 3X4S6 80 'B119 02t; B133290 ) - \ (3 B085. B084. - . 1;s B126, /186 B184;' B1B8

0 A549 ' @ B11S ' .. B118 .' B 1 4 5 Br83 B191 ...... , mr9

; B3S4 LL A2a9 A263 . B156 "/ B1 - v BU8 i S B199- B197, R A270 . A231:

B350 . -. A249 A248X.: // A077 AO68 AO1 AO66 X AO61 ,--

( . , 0 .. YW;X 3, ASAOO9, AOO8. C4°ASA AC

\ .B358 S. AV9 A1OO A°79 AO80 A AO8t20 A123 A.0040N D A1210:

{RRA207; AO84 SZA127 .- /

) 0 A129 'A 092 fs A , A119,- , A166A160 ...... \

{ B348 B359 - | A147 A105: A6095z A094, "A113' ; - A117 ' A129,-- A159 A155 A1>X<

- -- A1OX . A2.02.- :: . A128)-' A208 : A149 A1

i3347 i B3610

( B346 : A570 0. A172 89 A167 kv A192 P

| ' 114' 'A12 A191

u 4 A1@8 t

\ ( . A177 A182

\ ) A179 ( . A179 .

\ A181 \ ' ' ESE 1 Figure15:b. Tracts and sites in the 0 100 500 \ southernpart of the surveyarea. i \ R. J. Robertson

27. S028:Ranges represented: valleyappears to be grazedmore extensively but thereis littleevidence N-BA?,A-CL, A-HL, A-MED, thatthat activity is promotingerosion. Many artifacts are, however, clearly CL-HL, R-O, MED-M, O-M, associatedwith disturbed and eroded areas. These places include the main MO-M, LO-M. east-westaccess road into the valley(S028) and various locations at the S020:Period represented: M; eastend of the northernridge, in particulara dirtroad between tracts rangesrepresented: A-CL, A-HL, A-R, CL-HL, CL-R, O-M, MO-M. B307and B308 (SO20).27 Finally on a low knollnortheast of the passin 28. S014:Range represented: zone3 thatleads over the coastalridge (see below), another low-density A-HL. scatterwas noted where steep terraces had been disturbed (S014).28 __ rB l ] = =$ | rb | | Si ig_ J l Wlii - IZAN + - _ [ - Romano.[w!|iET | FigureLooklng 16. Bulldozedwest. scarp and wall

58 JACK L. DAVIS ET AL.

.L _tl ilNii i: z i . ' | ilew __ fragrnentin tractB342 at Porto

= I *= !!|1F$1s jrS- 9--g<-=11 _|

ZONE3 Zone3 is boundedon the south by a steep-sidedterraced ridge (98.3 masl) thatbranches into the crescent-shaped ridge enclosing zone 4. The north- ernarm of thiscrescent runs northeast toward the northern limit of Spitalla, just southof a largeHoxha-era munitions plant at PortoRomano; it is deeplyincised by erosion.The ridgethat borders zone 3 on the northis largelyunterraced, with broad lower slopes descending gradually into the valleybottom. These grassy slopes are divided into separate parcels by barbed wireand appear to havebeen plowed in the recentpast; they are lightly eroded.The slopeof this ridgebecomes steeper near the northwestern cornerof the zone,where a passleads over the coastalridge. South of a grassytrack leading to this passa few newfarmsteads have been estab- lished.The areaabove these is terraced.In the southwesternpart of the zoneanother farmstead is nestled in a foldand is largelyhidden from view. Aboveit is the northernmostsignificant peak of the coastalridge (142 masl),a terracedtriangular eminence, the top of whichhas been disturbed by the constructionof bunkers. Artifactswere observed on theeroding slopes of thecoastal ridge and in plowedfields around the southwesternfarmstead (S012); finds from this sitewere also found in the northwesternquadrant of zone5 where theyhad been redeposited through erosion.29 A concentratedscatter far- therto the east (SOll) was exposedin an erodinganimal track near a bunker.30A diffuse distribution of artifactswas found in the lowerfields on the northernside of the catchment(SO13).31 SOl9 is locatedin the northeasternmostpart of thezone near the Durres-Porto Romano road.32

29. S012:Period represented: M; 31. S013:Periods represented: rangesrepresented: A-CL, A-HL, MHL, M; rangesrepresented: A-HL, A-R. CL-HL, R-M. 30. SOll: Rangerepresented: 32. S019:Period represented: M; A-HL. rangerepresented: A-HL. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 59

ZONE4 Thecrescent formed by low ridges branching northeast and southeast from the higherridge that constitutesthe boundarybetween zones 3 and 5 enclosesbottomland west of the Durres-PortoRomano road. This is the mostextensive unoccupied lowland in thesurvey area. Alfalfa fields stretch forover 300 m to thewest of canal-sidehousing, and there is a largegrassy meadowat the southwest. The ridges defining the crescent are completely terracedand largely stable except in severalplaces near their crests. Local- izederosion there has exposed scatters in tractsA324 andA325 andhas possiblytransported artifacts a shortdistance into tracts A403 andA404 (S010).33The terraces were otherwise devoid of findsexcept in tractA317, wherethere was a highdensity of artifactson the surfaceof a fieldroad andin the scarpabove it.

ZONE5 A roadpaved with limestone cobbles runs east-west through zone 5. There areabandoned chicken- and fish-processing plants near its easternend andderelict military barracks farther west where the valley broadens. Scat- teredhouses have been built along the road and squatters are beginning to occupythe barracks. There is littleevidence for pasturage, which may ex- plainwhy terraces on the southernridge bordering the zoneare so over- grown.A terracedinterfluve divides the western end of thevalley into two parts.North of it thelowest terraces are shallow and grassy; farther uphill theyare a meterhigh, fairly broad, and covered with maquis, especially on theirfaces. At the highestelevations (74-97 masl)there is activeerosion, andthe soilis verysandy. South of theinterfluve slopes are less steep, and hilltopstend to be grassyand unterraced. Lower slopes are for the most partovergrown with brambles, except to the eastwhere there are olives andopen grassy fields. S009covers terraces in zone5 to the southwestof the newcemetery of Durres,as well as thatpart of zone4 thatincludes the cemeteryand terracesto the northwestof it.34The cemeteryis locatedin a naturalam- phitheater,a formation that funnels slumping soil toward a centralbasin. The constructionin the amphitheaterof semicircularconcentric paved roadsand three grand concrete staircases is promotingsubstantial erosion. At thetop of theamphitheater, the digging of a trenchfor the foundations of the concreteblock wall that encloses the cemetery brought particularly largequantities of artifactsto the surface(Fig. 17). Artifacts are eroding 33. S010:Ranges represented: frombulldozed terraces in tractA399 andare embedded as muchas a A-CL, A-HL, A-R, CL-HL, O-M, MO-LO. meterwithin their matrix, suggesting that intact subsurface deposits re- 34. S009:Periods represented: PH?, mainand that these were buried by soilseroded when a massivetunnel CL, O, M; rangesrepresented: A-CL, bunkerwas built higheron the slope.In fieldwalkingand subsequent A-HL, A-R, CL-HL, CL-R, CL-B, revisitation,we discernedtwo principal components of S009that appear LCL-HL,HL-B, B-O, B-M, LB-O, to be spatiallydistinct: Archaic-Hellenistic artifacts are more plentiful in LB-M?O-M, LO-M. 35. S005:Periods represented: HL, the northand Medieval-Ottoman finds are concentrated in the south. MHL, LR?,M; rangesrepresented: Fartherwest, there is anotherconcentration of artifacts(S005) among CL-R?MHL-LHL, LHL or LR. farmbuildings next to the cobbledroad.35 Parts of thefarmyard that were wall__S___ _Q_i9+ modern9*!v!1background. lookingeast.cemetery From tractA331,in the

60 JAC K L. DAVI S ET AL.

_ _ Figure17. Site 009, with enclosure

v Figure18. Mortared brick wall at plowedproduced artifact densities that were higher than normal. Two seg- mentsof an ancientmortared brick wall were found in tractsB267 and B268(Fig. 18). Immediately east of these,in tractB270, large pieces of threeamphoras and tile were collected.

ZONE 6 The northernmostof the tworeservoirs at Spitallalies at the eastend of zone 6. The entirearea south of the reservoirand west of the Durres- PortoRomano road is nowfilled with recently built houses. In thevalley bottomare farms and plowed fields, although the soil is poorlydrained. Tunnelbunkers in tractsB112 and B113 render slopes impassable in places andearth from their construction appears to havebeen dumped in tracts B090and B091. Buildings are beginning to be constructedat the eastern end of the ridgethat constitutes the northernborder of the valley,and thereare several farms higher on thisridge. Terraces, even at the highest elevations(95-104 masl), are generally in goodrepair. The southernridge DURRES REGIONAL ARCHAEOLOGICAL PROJECT 6 I

Figure19. Site 004, with possible Archaictemple on the borderof zones 5 and 6. Fromtract B053, lookingwest.

(86.4-117.6masl) is moreovergrown than the northern,which has scat- teredgrassy fields on top. At thewestern end ofthe valley below the coastal ridgeroad are heavily grazed terraces on whichsome olives are planted. S004is locatedon a low knollthat projects east from a highpeak in the coastalridge (Fig. 19).36Artifacts found in tractsB213, B214, and B223 mayhave been eroded downslope from the majorconcentration. Findsfrom the site included cover tiles and large pan tiles (many of them red-slipped),ridge tiles, a few Classical-Hellenisticblack-glaze sherds, andtwo fragments of anashlar block. We suspectthat an Archaic temple maybe locatedat thissite. Some 600 m eastof S004a singlefragment of an Archaicarchitectural terracotta was collected. Many terracottas in an identicalstyle and fabric were found some years ago in the areaof the modernDurres cemetery in zone8, andmay have been carried there from S004.37 S021is welldefined: artifacts are most plentiful in a scarpof a terrace in tractB036 andon the uppermostterraces of tractB037.38 Between S021and S004 a less densescatter (S023) was recognizedon two hill- tops.39At S029,high on the coastalridge to the south,a fewlithics were collected;these may be Neolithicin dateand may represent the earliest humanactivity yet identified in theimmediate environs of Durres.40Lower on thecoastal ridge to theeast is a low-densityscatter of pottery(S024).41 Nearthe Durres-Porto Romano road at theeastern end of thezone there is a densescatter of artifactsin plowedfields (S022).42 Two additional sites(S025 and S026) were found on the ridgethat divides zone 6 from zone7.43

36. S004:Period represented: CL; rangesrepresented: A-HL, A-R, A-HL, CL-HL, LCL-HL,MO-LO, rangesrepresented: A-ECL, A-CL, CL-HL, R-M. LO-M. A-HL, A-R, CL-HL. 40. S029:Period represented: N? 43. S025:Period represented: CL; 37. Seebelow, n. 88. 41. S024:Periods represented: A, rangesrepresented: A-HL, A-R, 38. S021:Periods represented: CL, HL, M; rangesrepresented: A-CL, CL-HL. M; rangesrepresented: A-CL, A-HL, A-HL, A-R, LO-M. S026:Periods represented: CL, HL, A-R, CL-HL. 42. S022:Periods represented: CL, M; rangesrepresented: CL-HL, 39. S023:Period represented: HL?; HL, M; rangesrepresented: A-CL, CL-R. JAC K L. DAVI S ET AL. 62 ZONE7 Zone7 consists of a symmetricaland steep-sided valley that ascends to the second-highestpeak in thecoastal ridge (182 masl). Modern housing con- tinuesa long wayup the valley floor from the Durres-Porto Romano road, andthere are severalfarms on the lowestslopes of the ridgebordering zone7 on the south.Other isolated farmsteads are located at thewestern endofthe valleyat higher elevations in tractsB153-B155 and B174. Most terracesin this zoneare used only for pasturage, with the result that many arereverting to scruband maquis; exceptions include a well-tendedolive grovein tract A233.Tunnel bunkers cut into the easternslopes of the coastalridge haveprecipitated some erosion but no artifactswere found there. However,localized erosion has exposed artifact scatters in a seriesof tractson a steeplysloping knoll beneath the coastalridge road (S027).44 Farthereast, S017is a denseand well-bounded scatter of tiles.45Another significant concentrationof artifactsexists at the easternend of the ridge thatforms the zone'ssouthern boundary (S008).46 Byzantine and Medi- eval-Ottoman potteryis densethere in a partlyplowed field and on slopes immediately northwestof the moderncemetery of Durres.This site ap- pearsto be spatiallydistinct from the Classical-Hellenistic remains (S007) foundon the southside of the sameridge in zone 8, althoughthe two componentsoverlap in tractson the crestof the ridge(e.g., A050).

ZONE8 Zone8 is the second-largestcatchment in the surveyarea (Fig. 20). It extendswest fromthe southernmost of thetwo reservoirs at Spitallaand is dividedbya lowspur that projects from the coastal ridge. The ridge (63.9- 116masl) that boundsthis catchmenton the northis onlysporadically terraced. Terracesbulldozed to controlerosion in tractA048 cut through atleast one intactClassical-Late Classical grave, where human cranial fragmentsand a nearlycomplete oinochoewere found (S007).47 During revisitationa largearchitectural block was observed not farto thewest. Artifactshave been erodedfrom S007 and redeposited downslope in 44. fieldsthat are currently S027:Ranges represented: plantedin alfalfa.There are also artifacts eroding A-CL,A-EHL, A-HL, fromthe scarps CL-HL, of terracesin tractA030. Other finds that are probably CL-R,LCL-EHL. fromgraves wereretrieved from a shallowdrainage ditch in tractA025. 45. S017:Ranges represented: Onthe north A-HL,CL-HL. sideof the streamthat runs through the valley bottom is a notable 46. S008:Periods represented: O, concentrationof artifactsin tractsA009, A010, A084-A086, and A092(S002).48 M;ranges represented: A-HL, A-R, Thehighest densities were associated with a recently B-O, field plowed B-M, LB-O, MED-M, andthe intersection of agriculturalroads. O-M,MO-LO. In the northwesterncorner of zone8 olivesgrow on well-maintained 47. S007:Periods represented: A, terraceson CL,O, the lowerslopes of the coastalridge. S003 is locatedat the M; rangesrepresented: A-CL, southernedge A-HL,A-R, A-B, CL-LCL, of theolive groves, centered on a lowknoll cut by the coastal ridgeroad.49 CL-HL,CL-R, CL-EM, O-M, Farmersknew that there were the remnants of anold church hereand the areais LO-M. called"Kisha e Kallmit,"i.e., "Church of the Reeds" 48. S002:Periods (Fig.21). At the represented:CL, westernend of the knollare foundations of wallsand a M;ranges represented: A-CL, A-HL, marbleslab in situ.Tiles andbricks were recovered from a broaderarea A-R,LO-M. thatincluded 49.S003: Periods tractsA079, A096, and A097. Immediately above the coastal represented: ridgeroad a substantialnumber of tiles MED,M; rangesrepresented: A-R, arescattered in tractA101 and R-M,MR-LR. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 63

Figure20. Coastalridge and southernreservoir at Spitalla.From A152, lookingnorthwest.

Figure21. Zone 8. Lookingsouth- west towardhead of valleyand site 003 fromtract A070.

severaldiagnostic sherds dating to the4th-6th century A.C. werefound in thescarp ofthe road itself. A columnof paleo-Christiantype that now lies nearbyin tractA095 is likelyalso to be of thatdate. AboveS003 is thesteep eastern face of thecoastal ridge. On its high- est peak(187 masl),at the northernend of the hill of Currila,is a naval reconnaissancestation. High and sometimes bramble-covered terraces on theridge are primarily used for grazing. At lowerelevations there are mili- tarytrenches. Theridge (65.1-98 masl) that forms the southern boundary of zone8 is completelyterraced but only sporadically cultivated. Two significant ar- tifactscatters were detected, one in a placewhere a narrowstrip of terraces (40 m wide)has been converted into a fencedgarden (S016), the other at the easternend of the ridge(S001) on the outskirtsof the heavilyurban- izedarea of Durres.50Adem Roda, the owner of thegarden, showed us an 50. S016:Period represented: LCL; inscribedgrave stele that he supposedlyfound in the areaof the hill of rangesrepresented: A-CL, A-HL, Dautaj(Figs. 22-23). He alsopointed out an ashlarblock of uncertain LCL-EHL,LCL-MHL. functionthat he discoveredwhile digging a pit forwater in tractA143 SOO1:Periods represented: CL, HL, R?,M; rangesrepresented: A-CL, (Fig.24). On the plowed terraces within the garden we founda substantial A-HL, A-R, CL-HL, LCL-EHL, quantityof diagnosticGreek artifacts, but the tracts on either side of Roda's LHL-R,MO-LO, LO-M. fencedid not yield comparable material. 64 JACK L. DAVIS ET AL.

Figure22. Gravestele S016-SF01. Obverse.

Figure23. Gravestele S016-SF01. Reverse.

Figure24. Cut blockfrom site 016 .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 6S

SOO1extends from the crestof the ridgedown to a bulldozedroad southof the reservoir.Artifacts are concentrated in placeswhere erosion hasbeen accelerated byrecent construction projects. In tract A193, a plowed fieldon the crestof the ridge,a significantconcentration of artifactsis likelyto be in situ,but in fieldsbelow it andon slopes farther west artifacts havebeen redeposited from higher upslope. A long centralspur (96 masl)extends from the hill of Currilaeast- wardto the edgeof thereservoir, dividing zone 8 in half.Under commu- nismthis area was dedicated to olivecultivation, but the northernface of the ridgeis nowso overgrownthat in placesit is unwalkableand terraces arecollapsing. The moregentle southern face is grassier;olives appear healthyand well tended.

ZONE 9 Zone9 is a narrowhairpin-shaped valley with a modernclay quarry at its westernend. It is likelythat zone 9 layoutside the limitsof the ancient Greekcity. Many ancient graves have been excavated there. Today newly builthouses occupy the entire valley bottom, the lower terraces around it, andthe topsof the easternends of the northern(45 masl)and southern (33masl) ridges that border the zone. Only a smallarea could be surveyed: 15 tractswere defined in 4.7 ha, arrayedin a crescent-shapedband that includedrubbish-strewn slopes between houses at the headof the valley andterraces recently cleared of olives.

Figure25. Tile grave in tractA185, lookingwest

At leastone intactHellenistic tile grave(S018) is erodingout of a terracescarp (tract A185) in thesoutheastern part of thehill of Dautaj,on a low knollthat has been partly destroyed by bunker construction and is 51. S018:Period represented: HL; surroundedon threesides by houses (Fig. 25).51 Other graves (M00) were rangesrepresented: A-HL, A-R, shownto membersof the projectbut lay outside the partof zone 9 that LB-MO. wassystematically surveyed (see Appendix 2) (Fig.26). i!l!LS2|l l | ?? l?6ml SA ... >S.?. . M.?:.S.7. 0,, Si :.: :Xi?;?,7 ? ?RE,i, ? 6i i ' it??ii a/2s .?il '?i ig. ?*':.;/. t ;, 8 ;.;'?^y t;i if t it'*\ ;; .?lt0 i ?0i li0t7 { ?:tZ0d?t ? XX:i i ' <0i;;;; t+ i '6it,000 '.0;; '%?.iBUt ?-ji? ii? iX i':000; i s;' ??i?02;ta.|; l ro;t.;07 i>i? 0 ?6i;?:;, Ei i;0:; i ?BE'? iXii? 2 ,5.iz.;;p,. i?i2???,'0g'';?.?: gE? i,?i / rnemna 1 r 1

66 JACK L. DAVIS ET AL.

ZONEIO Zone 10 is evenmore heavily urbanized than zone 9 andstill less of it couldbe systematicallyexamined. Apartment blocks line city streets that leadto a brickand tile factory exploiting the clay quarry in zone9. A band of tractswas walked on thewestern side of thenarrow ridge that separates the cityfrom the factoryand quarry to the west.It is impossibleto be certainif the few artifactsfound in tractsA177-A181 are in situor to knowthe originalcontext of a largecortical flake of chertfound in tract A180.

ZONEII Zone11 wasnot surveyed.52

ZONEI2 Zone12 constitutesthe entirety of thewestern seaward side of thecoastal ridgeand is thelargest of ourcatchments. Steep rockslides make the ter- rainimpassable at higherelevations, but there are more gentle slopes by theshore (Fig. 27). Much of thesouthern part of thezone is terracedand wasonce planted with olives, almost all of which have been cut down since 1991.Most of the terraceswere built by communistagricultural collec- tives.There is littleevidence for settlement, ancient or modern, in anypart of zone12. The landat present does not appear to be extensivelyused for agriculturalpurposes; the terracesare covered with tall grasses and some- timesalso with brambles.Some evidence of cultivationexists, however, includinga well-tendedgrove of matureolive trees. In the north,slopes are more gentle. Goat paths are extensive, sug- gestingthat the areais frequentlygrazed. Terraces at the highesteleva- tionsare covered by sparse vegetation and are eroding. It is clearthat sheep andgoats enter the coastalzone from the valleys to the eastand are pas- 52.See above, n. 22. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 67

Figure27. Erodedwest face of coastalridge. Looking south from west of tractB342.

turedhere. Flocks move through several passes, most commonly one in the northwesternpart of zone 3. In the northernmostpart of zone 12 therehas been massive recent disturbance from the quarrying of clay-rich slopesin thevicinity of PortoRomano. In a flatsaddle south of anextensive house complex on theborder of zones8 and12 is a significantartifact concentration (S015).53

ANALYSIS OF DISTRIBUTION PATTERNS OF ARTIFACTS

Approximately2,100 artifactswere collected from 430 of the 938 tracts definedby the two survey teams.54 All findsbrought in fromthe field were washed,counted, and describeddaily.55 Nonceramic objects as well as loomweightsand architectural terracottas were entered in a databaseof smallfinds.56 Ceramics from each tract were sorted by date, and the quan- tityin eachchronological group was recorded on preprintedforms. Diag- nosticartifacts with preserved profiles or noteworthy decoration were pho- tographedand drawn. Amongthe lithics there is no obviousPalaeolithic or Mesolithic ma- terial.57Several pieces could be Neolithicor of laterprehistoric date, but it is notimpossible that others are more recent (e.g., from relatively modern

53. S015:Period represented: R; A048-02,including the ink andwater- architecturalterracottas, one terra- rangesrepresented: A-HL, CL-HL. colorreproduction of its decoration cottaloomweight, fifteen lithics, and 54.Team A collected1,122 artifacts, (Fig.48). Drawingsof otherfinds are twostone architectural members. andTeam B collected962 artifacts. by Hayes. Othersmall finds include glass, slag, 55. Findswere processed in the 56. Fifty-eightartifacts were col- andgeological samples. newmuseum of Durresand analyzed lectedand catalogued as smallfinds, 57.We aregrateful to CurtisN. by Hayesand Stocker. Davis photo- includinga cointhat was cleaned in Runnelsand Muzafer Korkuti for graphedartifacts. Myrvete Dajlani Tiranaunder the supervisionof noteson lithics.These have in part drewB065-01, B234-SFO1, and ShpresaGjongeca;, a bronzering, two beenincorporated into ourtext. JACK L . DAVI S E T A L . 68 findsand the threshingsleds). Notable are the overallscarcity of lithic rarityof prehistoric artifacts in general.58 of the Overhalf (1,166) of the artifactsare tiles. Eighty-eight (7.5%) hasbeen tileshave at leastone slippedsurface, a higherpercentage than if this recognizedin the hinterland of Apollonia.59We areas yet uncertain preservation. variationreflects differing regional styles orvariable states of closely Potsherds(860) were generally small and relatively few could be on atleast one dated.Sixty-one sherds (7%) preserved traces of blackglaze pyxis, surface.Shapes represented in this wareinclude the skyphos, sherds oinochoe,jug, amphora, and krater. Forty (65%) of theblack-glaze Plainwares weretoo fragmentaryto be associatedwith a specificshape. fora aremore plentiful, although a specificform can only be determined butthey con- smallpercentage of thetotal. Amphoras are most common, percentage stituteonly 7% of the totalnumber of sherds,a muchlower than thanin the hinterlandof Apollonia.Pithoi are less well represented amphoras(2.5%). aremost Artifactsfrom the Archaic, Classical, and Hellenistic periods the Roman prevalent(608; 30% of all sherdsand tiles).60 Material from 1%).Another periodis noticeably underrepresented (21 sherdsand tiles; wouldex- periodthat is poorlyrepresented is the LateHellenistic. One wares,and pectto findstandard Italian and Aegean fine wares, cooking In addition, amphorasfrom these periods, but these were rarely found.61 wares Medievalpottery is not plentiful,even though 12th-14th-century (11%of all arepresent in the townof Durresitself.62 Some 226 artifacts artifacts)may artifacts)could not be datedat all.Another 783 (38%of all in date(e.g., bemodern, but at leastsome of theseare probably earlier Byzantineor Ottoman).

TOWARD A HISTORY OF THE HINTERLAND . . OF DURRES beginningof Surveyin the uplandsnorthwest of Durres,as notedat the article,is onlya firstcontribution toward a comprehensiveexamina- this about tionof theentire territory of thecity. In drawinggeneral conclusions this fact thehistory of Durresor its relationshipwith its hinterland, what shouldbe borne in mind.Nevertheless, it is worthwhilesummarizing

Archaic,Classical, and Hellenistic mirrorsthe 59.The resultsare as yet unpub- 58.This observation pottery),1988 (Megarianbowls), 1990 surveyand exca- lished.On thisproject see J. L. Davis, resultsof extensive (Archaicwares), and 1994 (Attic prehistoricfinds M. Korkuti,L. Bejko,M. L. Galaty, vation.Virtually no importsof the 6th-5thcentury B.C.). thansherds S. Mu,caj,and S. R. Stocker,"The havebeen reported other 61.Wares of theseperiods are found the newport MallakastraRegional Archaeological anda hammerstonefrom in the cityof Durresand in its ceme- to be of Late Project,"http://river.blg.uc.edu/mrap/ of Durresthat are said teries.For example, see Hidri1988; 1976,p. 301). MRAP.html.;see alsoKorkuti et al. BronzeAge date(To jci Tartari1987, 1991. Forthe scarcityof lithicfinds in the 1998. an extensive Materialof theseperiods is well 62. See Hoti 1989afor hinterlandof Durresand differences 60. Middleand the moderncity of reviewof ceramicsof the betweenour results and those of other representedin hasbeen extensively LateByzantine periods. surveysin Albaniaand Greece, see Durresand Hidri1986a (local Davis,in press. published.See DURRES.. REGIONAL ARCHAEOLOGICAL PROJECT 69

we thinkour work has contributed to an understandingof the historyof the area. Both Strabo(7.5.8) and Eratosthenes (in Stephanusof Byzantium s.v. lioppaxtov)described Epidamnus as a city on a peninsulanamed Dyrrachium.Thucydides (1.25) called it anisthmus.63 It seems likely that thepeninsula of Straboand Eratosthenes consisted of theuplands that we haveexamined, an area that has until recently been almost completely sepa- ratedfrom the mainlandby a marsh.64Therefore, we concludethat we haveexamined a centralpart of the hinterlandof the ancientpolis. But antiquitieshave also been foundin the areaof the marsh(kenete in Albanian)and it is likelythat in Graeco-Romantimes tongues of dry landprojected into it fromthe uplands,as wasstill the casein the 19th century(Fig. 4).65 Forthe firsttime we havea comprehensiveoverview of settlement andland use in theuplands. Whereas Hammond walked the extent of the coastalridge to PortoRomano and back without finding pre-Roman re- mains,we nowsuggest that the entire area was fully within the orbit of the Greekpolis (Figs. 28-29). Although few artifacts that can definitively be assignedto theArchaic period were found, Classical and Archaic-Classi- calfinds were more widespread.66 Moreover, it seemslikely that by Early Hellenistictimes all of the areaeast of the coastalridge between Durres andPorto Romano was being used for agricultural purposes. At thistime a substantialsettlement also appears to haveexisted at PortoRomano.67 Butsince we havenot been able to studysites in detailit is asyet impos- sibleto knowwhattypes of agriculturalexploitation (e.g., permanent farm- steads,temporary shelters) were characteristic of this area in theArchaic, Classical,and Hellenistic periods, particularly in lightof the substantial problemsin visibilityfaced by ourteams. We hopethat future research willallow us to definewith greater accuracy the natureof thesesites. We mayoffer a fewtentative observations about the earlyhistory of the cityon the basisof ourfieldwork. Did the earlycity of Durresorigi- nallyconsist of twoseparate settlements? Appian (B Civ.2.39) described Dyrrachiumas a portand Epidamnus as locatedon a height.We have foundno evidencethat any pre-E{ellenistic village or townexisted in the areathat we examined.If thetradition of twodistinct foci of settlementis true,we wonderif theywere not so closetogether that both lay within the urbanarea of modernDurres. We mightthen imagine that the clusterof cemeteriesthat has been excavated at Kokomanand Dautaj our zones 9 and10 markedthe edgeof the ancienturbanized area. Finally,if we arecorrect in supposingthat there was a templeat S004, the distributionof Archaicpottery found by ourteams (Fig. 28) suggests

63. See Heuzey1886, p. 44, (Sufflay1924, pp. 2>21). See also the Greekcolony, despite the factthat regardingthe interpretationof this Ducellier1981, pp. 500, 507, 525; Thucydidesis clearthat the colonywas passageof Thucydidesin lightof the Schmitt2001, pp. 537-542. establishedin Illyrianterritory (1.24) geographyof the Durresarea. 65. See alsoPraschniker and andAppian (B Civ. 2.39) speaksof a 64. In 1393 Venicedecided to turn Schober1919, fig. 43. minglingin the colonyof Corcyraeans Durresinto an islandby diggingcanals 66. We foundno evidencethat andnative Taulantians. throughsandbars at boththe northern the uplandshad been settled by Illy- 67. Its remainsare now submerged. andsouthern ends of the marsh riantribes prior to the foundationof See above,n. 26. 7o JACK L. DAVIS ET AL.

Figures28-29. Distributionsof thatthis sanctuary,when it was firstbuilt, stood near the limitsof the Archaic-Hellenistic(left) and territoryexploited by the colony.68 It is possible,therefore, that its position pre-Romanartifacts. R. J. Robertson markedthe border of theArchaic city-state. In anycase, such an extraurban sanctuaryat a highelevation in themidst of agriculturallands would have beena visiblesymbol of the secularand religious authority of the polis overthe uplands northwest of the city. The historyof Durresin the lastthree centuries B.C. iS complicated

68. Potentiallyrelevant here are "deliberatestatement of possession" expansionat a laterdate, e.g., the Polignac'sarguments (1994) concern- (Polignac1995, p. 103).A border sanctuariesof SantaAnna and Apollo ing extraurbantemples as markersof sanctuarycould mark the frontierof the Aleosat Croton(1995, p. 100).Edlund sovereigntyin the countryside(see originalterritory claimed by a colonyat (1987)also discusses the extraurban alsoPolignac 1995, chap. 3). Temples its foundation,e.g., the sanctuaryof sanctuariesof a numberof coloniesin situatedin the chorawere often a SanBiagio at Metapontum,or markits MagnaGraecia. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 7I

Figure30. Distributionof Roman artltacts.. ,% R.J. Robertson

anda reviewof thescanty information preserved in ancienttexts does not explainwhy later Hellenistic and Roman remains are so rarein mostof the areathat we investigated(Fig. 30). In 317 B.C. Glaucias,king of the IllyrianTaulantii tribe that inhabited the area around Durres, offered asy- lumto youngPyrrhus of Epirusafter his father, Aeacides, was expelled as kingof theMolossians (Plut. Pyrrh. 3).69 Glaucias had marital ties to the Epirotethrone through his wife,Boroea, who herselfwas a Molossian

69. On the areainhabited by the detaileddiscussions of the Hellenistic Taulantii,see Wilkes 1992, map 3. historyof coastalAlbania, see Ham- This is the sameGlaucias who, allied mond1966; Cabanes 1976; Hammond with Cleitus,son of Bardylis,fought andWalbank 1988, pp. 154-155; againstAlexander the Greatat Pelion Hammond1989, p. 264;and Wilkes in 335 B.C. (Arr.Anab. 1.5-6). For 1992,pp. 122-125. 72 JAC K L. DAVI S ET AL. princess.70In 314 B.C. Cassandermarched north from Aetolia into Illyria anddefeated Glaucias and his allied cities, Durres and Apollonia. Cassander installeda garrisonat Durresand made a treatywith Glauciasin which the Illyrianking pledged not to attackany allies of Macedonia;he then returnedhome (Diod. Sic. 19.67.6-7).Soon afterward, in 313/312B.C., the Corcyraeanswent to the aidof bothApollonia and Durres. They dis- missedCassander's soldiers, freed Apollonia, and gave Durres to Glaucias (Diod.Sic.19.78.1). Severalyears later, in 307B.C., Glauciassecured Pyrrhus on theEpirote throne. It is likelythat Pyrrhus later held the city of Durres aspart of his Epiroteempire and perhaps even extended his rule in Illyria asfar north as Shkodra.Anna Comnena records that Pyrrhus lived in and usedDurres as the pointof departurefor his expeditionto aidTarentum (Alexiad3.12.8).71 In the mid-3rdcentury B.C. a kingMonunius issued coinage with an ethnicthat includes various abbreviations for the name of thecity, but it is not clearif Durreswas actually part of his kingdom.72In 229 B.C. Durres soughtthe supportof Romewhen it wasbesieged by the armiesof the Illyrianqueen Teuta (Polyb. 2.9-10) andwas awardedthe statusof am- iCUs.73 In 148B.C. it wasformally incorporated into the newlycreated Ro- manprovince of Macedonia.Soon afterward the Via Egnatiawas con- structedto linkthe citywith Thessaloniki.74 By the end of the reignof Augustusit hadbecome a Romancolony.75 At the timeof Diocletianit becamethe capitalof EpirusNova. Elsewhere,in Greece,it hasbeen suggested that radical changes in the countrysidefollowing the Romanconquest reflected a redistribution of land.76The onlysignificant Roman remains in the uplandsnorth of Durresare at PortoRomano where it is clearthat the well-knownwall hadbeen built by the 4th centuryA.C. The areasimmediately adjacent to it, nowpart of a militaryinstallation, could not be examined,but surface distributionsin zone 1 whereit waspossible to surveysuggest that there wasa focusof Romanactivity there. Althoughthe city clearly remained a valuable possession for Medieval empires,as in the Romanperiod we foundfew archaeologicalremains of this phasein thoseparts of the countrysidethat we examined.The city

70. Plutarchtells how the infant 73. On relationsbetween Rome and figs. 1-3; see alsoCIL III 1 709; Pyrrhusgrabbed Glaucias's knees and Durresin the Republicand under the AEpigr 1984,nos. 811-813). Brunt won his supportin spiteof his fearof Empire,see Cabanesand Drini 1995, (1971,p. 236) wouldlike to consider angeringCassander, who viewed pp.33-47. On the siegeof Durresand Durresa Caesariancolony. Octavian Aeacidesas an enemy. Rome'searly involvement in Illyria,see certainlypermitted displaced support- 71. See Cabanes1988, pp. 147-153, Errington1989, pp. 81-94. ersof Antonyto settlethere (Dio Cass. andHammond 1966, pp. 246-247 and 74.Wilkes 1992, p. 212;Fasolo 51.4.6).See Grant1946, pp. 275-280, n. 30, on the northernextent of the 2003. wherethe statusof Durresunder Ro- kingdomof Pyrrhus. 75.The nameof the Romancolony manrule is consideredat length.His 72. Cabanes1988, pp. 147-152; Pi- is nowknown to be IuliaAugusta from treatmentis nowsuperseded by thatof card1986. Generally, the widedistribu- inscriptionson leadpipes belonging to Burnett,Amandry, and Ripolles (1992, tion of the Hellenisticcoins of Durres an aqueductof the Hadrianicperiod p. 289),who discusscoinage incorrectly in southeasternEurope is worthnoting; andrestored by AlexanderSeverus attributedby Grantto Durres. see Crawford1985, pp. 224-245. (Mirajand Myrto 1982, pp. 132-133, 76. Alcock1993a, chap. 2. DURRES.. REGIONAL ARCHAEOLOGICAL PROJECT wasbesieged by Theodoricthe Greatin A.D. 478.77Anastasius I (491- 73 518), the Byzantineemperor and a nativeof Durres,is knownto have supportedbuilding projects in thecity, but perhaps not the construction of the fortificationsgenerally attributed to him.78Under his successor,Jus- tin I (518-527),the city was destroyed by anearthquake. In the 7th-12th centuriesDurres was the principal Byzantine stronghold on theAdriatic, but it hasbeen calculated that the city underwenta totalof thirty-two changesin overlordshipbetween 992 and1392. In the 10thand 11th centuries Durres was the sceneof strugglesbe- tweenthe tsars of Ohridand the Byzantines.Samuel took the cityin 989, it wasretaken by the Byzantines, then fell again to John Vladislav, the last tsarof Ohrid,in 1017.The NormanRobert Guiscard captured Durres in 1081and held it briefly.The citywas besieged by the Norman Bohemond in 1107-1108,then captured by WilliamII of Sicilyin 1185.After the FourthCrusade in 1204,Durres was occupied for a shorttime by Venice, butsubsequently fell under the jurisdiction of the Despotateof Epirus.It wasreceived as a dowryby Manfredof Sicilyin 1258,but recoveredfor theDespotate after his death in 1266.The citysurrendered to Charles I of Anjouin 1272,who declared himselfking of Albania. A devastatingearth- quakewrecked havoc in 1273.Durres was capturedand briefly held by KingStefan Urost II Milutinof Serbiain 1296.Charles Thopia took the cityin 1368,probably losing it in 1376to Louisof Evreuxand regaining it in 1383.Finally the citybecame a possessionof Venicein 1392,which heldit until1501. Thereis littleevidence for occupation in theuplands between Durres and PortoRomano until sometime after the Turkishconquest in 1501 (Figs.31-32). Artifacts of the Ottomanperiod are concentrated on the easternedges of theuplands near the Durres-PortoRomano road, and it is notuntil the 20th century that they are found in the entirearea. At the end of the 15thcentury Durres could be describedas "alarge destroyed city"bythepilgrimArnold Harff.The entire population in 1610consisted of 300houses, and fifty years later Evliya Nelebi found only 150 huts.79 In the firstmodern census of Albania,conducted by the Austro-Hungarian armyin the seconddecade of the 20thcentury, only 147 individuals were reportedas residentin theuplands (recorded under the entrySpitalli and includingKazanaj, Mahalla e portes[sic], andXhamallaj), although the populationofthe townof Durreshad increased to 4,175individuals living in 955 dwellings.80

77. On the laterhistory of Durres, clearlySpitalla, Xhamallaj may be see Ducellier1981; Fine 1983;Jirecek modernXhamadha (see Fig. 6), and 1916;Kiel 1990; Schmitt 2001; and Mahallae portes,"the neighborhood Stephenson2000. of the gate,"is certainlyPorto Romano, 78. See Gutteridge,Hoti, andHurst whichwas called Porta in the l9th 2001, pp. 397-398. century,presumably after the arched 79. See Harff1860, p. 65. ForEv- gatein the Romanwall; see Heuzey liyagelebi's description of Durres,see 1886,p. 46, and,later, Praschniker and Dankoffand Elsie 2000, pp. 148-151. Schober1919, p. 46, wherethe placeis 80. Seiner1922, p. 33. Spitalliis called"Porthes." 74 JAC K L. DAVI S ET AL.

Figures31-32. Distributionsof Ottoman (left) and modernartifacts.

. . R. J. Robertson THE DURRES HINTERLAND IN CONTEXT

In additionto examiningpurely local trends in settlementand land use, it is importantto comparethe results of surveyswithin Albania with each otheras well as with thoseof projectsin adjacentparts of the Adriatic to see if largerpatterns can be discerned.Such comparative survey will no doubtprove useful in the studyof ancientIllyria and Epirus,as it has elsewhere.8lAt the sametime the problemsin visibilitydiscussed above,together with the fact that we haveexamined only part of thehin- terlandof Durres,should warn us not to pushinterpretation of the data 81. E.g.,Alcock 1993a, 1993b; too far. Bintliff1997; Cherry and Davis 1998. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 75

Ourresults at Durresdiffer strikingly from those in thehinterland of Apollonia,another Corcyraean colony only 60 krnto the south,and from thosearound Butrint near the modernGreek-Albanian border. Most ob- viousis thelack of evidencefor extensive early prehistoric land use, both at Durresand Butrint. At Apolloniaintensive survey has producedthou- sandsof lithicsthat represent a nearly unbroken sequence dating from the LowerPalaeolithic through the Mesolithic.82 Such a picturemay prove to be moretypical than exceptional in southernAlbania.83 At Durres,Apol- lonia,and Butrint there is onlyslight evidence for settlement in the Neo- lithic,Bronze Age, and Early Iron Age, suggesting that the human impact on thelandscape of coastalAlbania at the timeof Greekcolonization was slight. At Apolloniathere is muchless evidence for activities in thecountry- sidethan at Durresbefore the Hellenisticperiod, except in its necropolis. Is it possiblethat the peninsular setting of Durresafforded security to the inhabitantsof the colonyand that this senseof safetypermitted more intensiveexploitation of thecountryside at anearlier date? In bothplaces, thereis a surprisinglack of Romanand Byzantine finds outside the imme- diatearea of the citycenter. Such a pictureis radicallydifferent from that revealedby intensivesurvey around Butrint, where Early Roman, Late Roman,and Byzantine ceramics were the dominantwares collected.84 It remainsfor us to determinethe extent to whichdifferences between Durres andApollonia, on theone hand, and Butrint, on theother, reflect the vari- ablehistories of thesethree major centers of Greekculture in theterritory of modernAlbania.85 It is not impossiblethat much of the earlierhistory of ruralsettlement at Butrintis obscuredby deepalluvial deposition.

THE FUTURE

Fieldworkin2001 accomplished itspreliminaryobjectives. Allopen ground in theuplands between the city of Durresand Porto Romano was system- aticallyexamined. Much remains that could be donein a secondcam- paign.Obviously a moresystematic definition of sitesand their gridded 82. See Korkutiet al. 1998.A fuller collectionare needed. There also remain significant areas within the urban reportby Runnelsis availableat http:// sprawlthat have not yet been investigated, especially in zonessouth of the river.blg.uc.edu/mrap/lithicsOl.html. moderncemetery and in placesnear the Durres-Porto Romano road. We 83. Compare,for example,the re- didnot examine these densely settled areas and, in anycase, our field pro- centreexamination of the resultsof cedureswould not have been effective. More suitable tactics would be those surveyby LuigiCardini in the 1930s: Francis2001a,2001b. employedin similarcircumstances in the village of Heraklionin theNemea 84. Hodgeset al. 1997;Pluciennick Valley(1984-1986, 1989).86 There teams of twoindividuals (one a native 1996. speaker)visited each house in thevillage, asking for permission to inspect 85. Fordifferences between the its yardand gardens, obviously a verylabor-intensive procedure, but one historyof Durresand that of Apol- thatcould and should be followedat Durres.In additionthere are loca- lonia,see Wilkes 1992, pp. 111-113; tionseast of theDurres-Porto Romano road that might be profitablysur- forButrint, see Hammond1967, pp.385, 474,499;Cabanes 1976, veyed,especially if, priorto fieldwork,geomorphological investigations s.v.Buthrotos. andremote sensing can pinpoint areas that may have been above water at 86.Wright et al. 1990. var1ous. t1mes. 1n. the - past. 76 JAC K L. DAVI S ET AL.

APPENDIX 1 GAZETTEEROF SITES

In thefollowing gazetteer, we listfor each site the zone in whichits center is located,all tracts that fall within its borders,and the numberof sherds thatcan be assignedto a singleperiod or, less specifically,to a rangeof periods.All measurements of catalogued artifacts are given in meters.Under the heading"Associated Ceramics" we list onlythose artifacts for which Hayesprovided a fullcatalogue description.

SITE 001

Area:Zone 8 Associatedtracts: A148, A149, A150, A151, A152, A153, A154, A155, A156, A157,A158, A164, A193 Periodsrepresented: CL (3);HL (2);R? (1); M (12) Rangesrepresented: A-CL (3);A-HL (17);A-R (3);CL-HL (1);LCL- EHL (1);LHL-R (1);MO-LO (1);LO-M (1) Description:Located on the northern face ofthe ridge that separates zones 8 and9, immediatelyto the southof the southernreservoir of Spitalla. Thesite covers two knolls, and the heaviest concentration of artifactsis on topof theeasternmost one. There are also artifacts downslope to thenorth in grassypasture and around new houses. Downslope erosion is activeand substantial.A revisitto the siteyielded more Archaic-Classical and Me- dieval-Ottomanpottery.

AssociafedCeramics A148-01 Basinor pithos rim, Corinthian coarse ware Figs.33, 44 P.Diam.est. 0.43 (originallyca. 0.05 more?).Edge of hanginglip wornaway. Core fired light gray, surface layer fired orange-brown, wet- smoothed.Normal brown mudstone grits, medium to small.Flat rim withhigh hanging lip, tilted outward. Archaic-Classical. lg ::_ :: l .:: t :::::Diam. ::::::::: est.::: . 0.20. "Local" oranger_ ware. Thin glaze covering, inside

.. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 77

Figure33. Basin or pithos rim A148-01(aboqwe);tilefragment A149-01(right)

F I A193-01 Bowlrim Figs.34, 44 s I _l _

i | 3ilt "l - app e ari ng lightyellow ove r the slip Rem ai n s of light yell ow was h it =s coveringexterior. Medium-thin. Sllghtly incurved rim. w------u . = E r Middle-LateOttoman. .

Ofher Cafalogged Ceramics A148-02 CorinthianTypeAamphorahandle Figure34. Bowl rim A193-01 Plaincoarse fabric. Corinthian import. Archaic-Classical.

A149-01 Tile fragment Fig.33, 44 Threeletters preserved in a rectangularframe: EIII[. Compare Hidri 1986b,pp. 110, 127, pl. XIV; Tartari and Hidri 1992; Myrto 1998, p. 87; andCabanes and Drini 1995, pp. 159-162, for tiles from Durres stampedwith the nameof aneponymous magistrate in thegenitive case followingthe preposition ss. Hellenistic.

A151-01 Black-glazeamphora neck with red-figure decoration PossibleAttic import. 5th centuryB.C., Archaic-Classical.

A153-01 Plaincoarse amphora sherd Fabricwith volcanic grits. Italian import. LateHellenistic-Roman. 78 JACK L. DAVIS ET AL.

Figure35. Kioniskos A164-SF01 A154-01 Plainmedium coarse amphora handle PossibleCorinthian Type B, Corinthianimport? Classical-Hellenistic.

A193-02 Mediumcoarse cooking ware rim LateClassical-Early Hellenistic.

UncataloguedCeramics Threefine black-glaze sherds, two of whichare possible Attic imports. Classical.

SmallFinds A164-SF01White marble kioniskos Fig.35 Oblong,0.108 x 0.094.P.H. 0.21. Unbroken at top;broken at bottom.Surfaces very badly damaged by plow. On formand type, see Cabanesand Drini 1995, pp. 50-52. Letterforms include a broken-bar alpha.The characteristicformula in threelines gives name of the deceased,patronymic, xatp£. Three lines inscribed:

],uNvE

]NvE Xat[p]£

3rd-lst centuryB.C. Hellenistic. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 79

SITE 002

Area:Zone 8 Associatedfrarts: A009, A010, A084, A085, A086, A092 Periodsrepresented: CL (3);M (33) Rangesrepresented: A-CL (1);A-HL (16);A-R (1);LO-M (1) Descritfion:Located on a lowrise in thevalley bottom in thenorthwestern partof zone 8. Findswere concentrated on the northside of a streamin twodiscrete locations; visibilitywas very low in a fieldof leafygroundcover thatseparated them (tract A086). The siteoverlooks extensive arable bot- tomlandand seems a suitablelocation for an ancientfarmstead; gentle slopesnearby could sustain olives without terracing.

CafaloguedCeramics A084-01 Black-glazefoot of closedshape Imported,Attic? Archaic-Classical.

SmallFinds A010-SF01Metalgunshell Modern.

SITE 003 jqrea:Zone 8 Sssociatedfracts: A079, A095, A096, A097, A098, A101 Periodsrepresented: MED (1);M (4) Rangesrepresented: A-R (2);R-M (10);MR-LR (3); UNKN (5) Descritfion:Situated on a spurthat descends from the coastalridge in the northwesternpart of zone 8.The area is knownas Kisha e Kallmit("Church of the Reeds").It doesnot, however, appear that the buildingwas stand- ing at anytime in the 20thcentury. On the otherhand, it doesseem that architecturalblocks have been recently removed from the site.At present the foundationsof wallsare visible, and bricks and a pieceof marbleve- neerwere found. Nearby in tractA095 is a marblecolumn. Potsherds of the 4th-6thcentury A.C. werenoted next to the sitein a roadscarp in the courseot r revls1tatzon.* . .

Sma#Finds S003-SF01Marble column Fig.36 Diam.at top 0.29;Diam. at brokenend 0.27. Not foundin situ. Middle-LateRoman. 80 JAC K L. DAVI S ET AL.

Figure36. Marblecolumn S003-SF01

A095R-SFO1Bronze finger ring Fig.37 W. band0.0043; W. bezel0.0075. Made from a thinbronze sheet, ca.0.007 thick. Flat inside, convex outside. Lightly incised lines on bezel. Komanior Arberor culture, 6th-7th century A.c.87Medieval.

SITE 004 Figure37. Bronzefinger ring A095R-SFO1.Scale 2:1 Area:Zone 6 Associatedtracts: B056, B065, B201, B202, B203, B213, B214, B223 Periodrepresented: CL (1) Rangesrepresented: A-ECL (1);A-CL (6);A-HL (21);A-R (15);CL- HL (19) Description:Located on a lowknoll that projects east from a highpeak in thecoastal ridge between zones 5 and6. Findsinclude cover tiles and large pantiles (many of themred-slipped), ridge tiles, and a fewCL-HL black- glazefragments. We suspectthat this place may have been the location of an Archaictemple, from which derive architectural terracottas found in thevicinity of the modernDurres cemetery.88

CataloguedCeramics

B065-01Tile withdark red painted band Fig. 38 Archaic-EarlyClassical.

87. Forthe significanceof these on inspectionof photographs. culturesas transitionalphases in 88. Zeqo1986, pp. 181-182. Cf. Albaniancivilization between antiquity Winter1993, pl. 48, fromthe "Blass- andmodern times, see Wilkes 1992, gelbesDach" at Calydonand fig. 131 pp.273-278. We thankEtleva Nallbani fromthe Treasury of Epidamnusat forher assessment of thisartifact based Olympia. A

DURRES.. REGIONAL ARCHAEOLOGICAL PROJECT 8 I

B

A

B B

A . B065-01

B234-SF01

Figure38.Tile B065-01and cornice orpediment fragment B234-SFO1. 0 5 DrawingM. Dajlani 2 5 cm I SmallFinds B201-SF01Architecturalfragment Pieceof cutlimestone block. B203-SF01 Architecturalfragment Fig.39 Largelimestone block.

B213-SF01Slag

Some600 m farthereast in an isolatedtract a singlefragment of anAr- chaicarchitectural terracotta was found: B234-SF01Partof cornice orpediment of a templeroof Fig. Figure39. 38 Architecturalfragment Finishedsurfaces B203-SF01 slippedyellow with decoration of blackbands. 6th-5thcentury B.C. JAC K L. DAVI S ET AL. 82 SITE 005

Area:Zone 5 Associatedtracts: B267, B268, B270 Periodsrepresented: HL (1);MHL (2);LR? (2); M (1) Ranges represented:CL-R (2);MHL-LHL (2); LHL or LR (1); UNKN(4) Description:Situated at thewestern end of thevalley, near the main lime- stone-pavedroad. Twoancient mortared brick wall fragments (probably of LR date),each abouta meterwide, were observed in tractsB267 and B268.Bricks are ca. 0.07 m in thickness.Tract B270 nearby may also be associatedwith the site.Several large amphora fragments, datable to the Hellenisticperiod, were recovered there, as was a 1 * . fragmentof an ancient aralnplpe.

r

_

::-.-:... ::.:.::..:-:::....::::::.::.::.:.::.:-.::.:: Figure40. Cookingpot rim B267-01 AssociafedCeramics B267-01Cooking pot rim,with imprint of a Figs.40,44 missinghandle Diam.est. rim 0.162; p.H.0.04. Ware probably not local: fairly darkgray, fine-textured;fine lime inclusions, traces of mica.Fairly thin- walled.Oblique rim withsmall rising lip anda smallledge at loweredge oninside. Traces of a handleon undersideof rim. Treatment of therim suggests a LateHellenistic or LateRoman (4th-6th centuryA.C.) date.Missing handle presumably looped cally verti- downto thebody; this is morelikely to be a Roman/Late feature. Roman

B270-01Amphora, Graeco-Italic Fig.44 Diam.est. of outerlip 0.20;p.H.0.10. Two large rim fragments (not joining),each witha handle-top,apparently from opposite sides of the samevessel. Pinkish-red,hard-fired, clean-breaking, fired buff at surfaces. Scatterof smallto medium-sizedbrownish grits (traces of nomica). lime, Handlesof oval-lentoidsection (W. at topca.0.06? Full IF l ^ X § , -§- ,lf.

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 83

- l Hl$A$b

#b , { Wti.iti, - # ; t w 1tl L:L t

v 10

Figure41. Amphora toe B270-02 widthnot preserved). Source uncertain (South Italian? Italian Adriatic (left)and amphora shoulder B270-03 coast?Even Sicilian?). (right) Ca.250-150 B.C.

B270-02 Amphoratoe, perhaps Lamboglia type 2 Figs. 41, 44 P.H.ca. 0.155; Diam. of toe at bottom0.06. Close-grained buff ware,clean-breaking; some fine brown-gray specks. Thick-walled. Short, thick,solid toe (tipdamaged, perhaps rounded). Shallow lower body- curve. 2nd-lst centuryB.C. (presumably).

B270-03 Amphorashoulder, Graeco-Italic or relatedtype Figs.41, 44 Diam.est. of shoulderangle 0.25; p.H. ca 0.115?Single fragment. Lightorange, rather clean-breaking; small red-brown grits. Thick- walled.Sloping shoulder with small ridge at edgeand slightly bulging belly.Ware uncertain, possibly from an Adriatic source. Around3rd century B.C.

SmallFinds B267-SF01Ceramic brick from wall Ca.0.071 thick, mortared. LateRoman?

B267-SF02Mortar sample from wall LateRoman? 84 JAC K L. DAVI S ET AL.

SITE 006

Area:Zone 1 Associatedtracts: B338, B341, B342, B343, B344 Periodsrepresented: MR? (1); R (4);R? (1) Rangesrepresented: A-HL (7);A-R (2);CL-HL (2);R-MED (4); MR-LR(4); LR-DA (1);LR-EB (1); LR-EB? (1); UNKN (2) Description:Located at the northernend of the coastalridge at Porto Romano.The ridgetop here consists of bareearth, and it is clearthat the sitehas been disturbed both by long-term erosion and recent bulldozing. Foundationsof a mortarand brick wall are preserved, similar in technique to thewell-known wall at Porto Romano. CL-HL sherds are primarily in tractsB338 and B344.

AssociatedCeramics

B338-01 Amphorarim, with a handle, Figs. 42, 44 EarlyChristian-Early Byzantine(?) type Diam.est. rim 0.039. Second handle not preserved (restored on drawing).Clean light yellow-brown ware (slight brown specks). Local fineware? Small narrow neck, lightly ribbed; rather high-arched oval- sectionedhandle(s), with slight median arris. Remains of a graffitoon neck,to left of thepreserved handle (transposed in drawingto opposite - :::: :: - : ::: ::- ::::::.:- faceof vessel). A datearound the 6th-8thcentury A.C. maybe suggestedon Figure42. Amphorarim and handle generalform. Late Roman to DarkAges. B338-01

- - - - Figure43. Tile fragmentB338-02

B338-02Tile fragment Figs. 43, 44 H. at edge0.061. One corner preserved. Medium orange-brown fabric:fine white and brown grits, a fewlarger white lumps, sparse glintingspecks (calcite?); fine grits on surface(especially upper face). Fairlysmooth breaks. Plain rounded-off end face. Tilted straight rim, roundedof£ Undersideuneven. Not a normaltype from the surveyarea. RomanImperial(?) to LateRoman? f \-_

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 85

L - L

w-- z

B267-01 A1 93-01

EHDb......

. B338-01 A149-01

P A1 48-01 - lWEo

s ev ff le lif

\ l B274 0-01

|n ___ __ ,0

B270-02 B270-03

\\ <

-\s \\

\\

\ -

\ -- *'

- - B341 -02

0 5 2 5 | I _ I cm Figure44. Potteryand tile fragments from sites 001, 005, and 006. J. W. Hayes 86 JAC K L. DAVI S ET AL.

_ _ _ _ Figure45. Ribbedamphora fragmentB342-01

B341-02 Flattile Fig.44 H. at edge0.057. One corner. Thickish, upturned at edge,with beveled(tilted) lip. Edge at siderounded off; a slightthickening at the end.Light buffware (more or lessthe "local" fabric), rather clean (slight brownspecks). Type uncommon here, and not treated in the normal manner. LateRoman or EarlyByzantine?

B342-01 Amphora,later Roman ribbed type (unclassified) Figs. 45, 52 Diam.est. belly 0.215; Diam. est. neck (at break) 0.07; p.H. ca. 0.085.One side of shoulder,with lower stump of a handle;mended from sevenpieces (one or two more, loose, may belong). Fairly deep brown ware,smooth-textured; clean breaks with fine lime specks and mica traces.Slender ribbed body; medium thickness. Straplike handle, sharply splayedat lowerattachment (max. W. 0.075).Type perhaps two- handled.From an unknown Aegean(?) source (Argolid? Asia Minor coast?Or similar). Apparentlya copy of Mid-LateRoman Amphora 3 (c£ Robinson's "micaceouswateriars" in Agora V, pl. 41), with similar handle treatment butthicker walls, which would suggest a 2nd-5th/6thcentury A.C. date. Middle-LateRoman.

OtherCatalogued Geramics B341-01Tile with mortar on surface Roman-EarlyByzantine.

SmallFinds B342-SF01Mortar sample taken from Roman wall MiddleRoman? X__r r - w 5 - ! - BX | - _s I'=-I .. _X | .... | _ | .. *S .. S _XS _ F>

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 87

SITE 007

Area:Zone 8 Associatedtracts: A024, A026, A028, A029, A030, A031, A047, A048, A049, A050,A051, A052 Periodsrepresented: A (1); CL (1);O (4);M (14) Rangesrepresented: A-CL (4);A-HL (11);A-R (20);A-B (1);CL-LCL (1);CL-HL (11);CL-R (1);CL-EM (6); O-M (36);LO-M (4); UNKN(31) Description:A densesurface scatter immediately west of themodern cem- eteryof Durreson high,recently bulldozed terraces. Human bone frag- mentsassociated with an oinochoe, A048-02, probably represent the rem- nantsof a disturbedgrave. The Ottomancomponent is concentratedin tractsA050, A051, andA052. A monumentalcut block(0.46 x 0.42 x 1.15m) was found in tractA047 in the courseof revisitation.

_ - _w M l ! | | Bfi W _w _ 1 s | | E g F _ [1>0- lS ! I | . -'

:

Figure46. Slyphosrim A024-01 (aboqwe)and basin rim A047-01 (right) AssociafedCeramics

A024-01 Black-glazeminiature skyphos rim Figs.46, 52 Diam.est. rim 0.065. Smooth light brown with polished surface; remainsof dullblack all over. "Local" fabric? Votive size? 6th centuryB.C.

A047-01 Rimof largebasin Figs.46, 52 Diam.est. rim 0.43; p.H. 0.064. Very thick buff fabric with slight brown-blackspecks. Thickened rim, rounded on top;series of grooves below(three preserved). Under rim, an arched lug-handle applied (right endmissing; presumably one of a pairon oppositesides of thevessel). A moreregional counterpart of the Corinthiancoarse-ware fabric, less gritty. Archaic-Classical. 88 JACK L. DAVIS ET AL.

1|415|R ,l_ gl^ i i9=-',_ - sj ,iiti .oJ A048-02Figure47.Red-figureoinochoe

A048-02 Red-figuresmall oinochoe (chous type) Figs.47, 48, 52 P.H.0.132 (originally ca. 0.16?); Diam. body 0.119; Diam. foot 0.088;Diam. min. of neckca. 0.055; H. of mainfrieze ca. 0.105. Unbrokenuntil recently. Corinthian or relatedfabric: smooth light yellowclay (slight tan-brown tint at surface)with a dullred wash. Slight tracesof redwash on unpaintedbottom. Munsell: on break7.5YR 8/4, polishedsurface to 5YR7.5/6; red wash ca. lOR 6.5/8. Good-qualityblack gloss (somewhat dull), fired dark brown on insideof neck.Most of innersurface of bodynot visible because of lime deposit.Gloss covers all visible parts of innersurface. Gloss ends at outeredge of foot,extending partway over resting surface on oneside of thevessel. Small triangular unpainted patch behind lower handle attachment.Red-figure frieze bordered below by a continuousnarrow reservedband (undecorated), and above by a stripon neck(opposite handle)bearing black painted ornament. Added cream/gold paint details on the figure,poorly preserved. Near-globularbody, wide-based, with wide neck (originally trefoil- mouthed,markedly pinched). Evidence of rim-pinchingvisible on preservedpart of neck.Lower stump of a round-sectioned(?)handle preservedon shoulder(originally ending on topof rim,which would havebeen pinched in at thispoint, with a knobbedterminal). Low squarishfootring. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 89

AO48-02

Figure48. Red-figure oinochoe 0 5 10 A048-02-M.Dajlani l l ' l cm

Red-figurefrieze in twopanels, separated at sidesby vertical reservedbands, "columnar," with slight cross-stripes marking offtops andbottoms. All reservedareas of the decorationbear red wash. Front panel:seated woman, winged, draped, to left,holding a patera/large phiale.Some reserved blobs under the figureindicating rocks. Woman: finedrapery lines (some thicker hems); top part of wingonce overpaintedin gold,pair of white/goldbangles on eachwrist, white/gold dot-necklaceon neck.At back,large upright palmette (nine-lobed, fleshy,not recurved), flanked by volute tendrils (horizontal pair and verticalpair flanking handle; no white/goldadditions). On reserved neck-strip,remains of a blackegg/ovolo band (eggs with thick and thin outlines);a blackdot below and between each element. Remains of two egg-motifspreserved. Vesseltype (chous, not the taller, late version) and decorative style basicallyAttic of ca.420-370 B.C., butthe fabric is Corinthianor colonialCorinthian. Probably not anearly South Italian product of similardate. Middle-LateClassical. 9o JACK L. DAVIS ET AL.

OtherCatalogued Ceramics A050-01 Plainbase fragment with drill hole in bottom Ottoman-Modern.

SmallFinds A026-SF01Possible stone pounder

A029-SF01Glass fragment

A048-SF01,SF02 Glassfragments

A048-SF03Sample of earthfrom inside oinochoe (A048-02)

A048-SF04Bones associated with oinochoe (A048-02)

A049-SF01 Stone

A050-SF01,SF02, SF03 Glassfragments

SITE 008

Area:Zone 7 Associatedtracts: A210, A211, A212, A213, A214, A215, A216, A217, A223, A224,A226 Periodsrepresented: O (4);M (1) Rangesrepresented: A-HL (1);A-R (1);B-O (16);B-M (13);LB-O (2); MED-M (9);O-M (6);MO-LO (2) Description:High-density scatters of artifactson the crestand northern faceof theridge that divides zone 7 fromzone 8. This sprawling scatter of sherdsis nearlyadjacent to S007and extends almost to the edgeof the urbanizedarea of the city.Artifacts are most dense immediately north of the moderncemetery of Durres.

AssociatedCeramics A215-01 Bowlrim Figs.49, 52 Diam.est. rim 0.21. "Local" ware, light yellow-brown. Medium thickness.Brush-smoothed interior with yellowish wash, extending over rippledexterior of rim.Originally glazed on inside?No glazeremains. Upcurvedrim with slight rippling on exterior.A laterversion of A193- 01 (fromS001)? See B283-01 (unassociated with a site)for possible laterdevelopment. 18th-beginningof the 20thcentury A.C. l l- |- ilocal facingBS 1|stvariant amphitheater of the . earlier that is "Corinthianbeing* constructed B" (i.e.,s i to Corfu)serve M series?as the Inew Definitelyt!i ww cemetery

.. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 9 I

. < .

Nl0s4. _ | s _ t

Figure49. Bowl rim A215-01 (lefi) A215-02 Jarrim Figs.49, 52 andjar rim A215-02 (righf) Diam.est. rim 0.19-0.20. "Local" ware (as A215-01), light brown. Glassylight yellow glaze (over slip) on inside,covering top of the rim. Partof stumpof a straplike(?)handle, attached at rim.Top of rim flattened,with a slightgroove below lip. Middle-LateOttoman.

SmallFinds A210-SF01Lithic

SITE 009

Area:Zone S Associatedtracts: A297, A298, A317, A329, A330, A331, A332, A333, A334, A336,A339, A340, A341, A342, A343, A344, A355, A356, A359, A360,A362, A363, A364, A372, A397, A398, A399, A400 Periodsrepresented: PH? (1); CL (2);O (8);M (16) Rangesrepresented: A-CL (14);A-HL (14);A-R (8);CL-HL (9);CL-R (3);CL-B (1);LCL-HL (1); HL-B (2);B-O (1);B-M (1S);LB-O (1);LB-M (3);O-M (92);LO-M (4) Description:Dense scatter of artifactsin andaround the large natural east- 1_

| | of Durres.The buildingof pavedaccess roads has greatly disturbed sub- | t surfacedeposits.

Sssociafed Ceramics A362-01 Amphorarim Figs.50, 52 Diam.est. rim 0.15-0.16. "Local" ware, apparently; yellow-buff, firedorange-buff at core.Fairly clean breaks; fine brownish specks. A

differentfrom the rim B135-02 (from S026). Figure50. Amphora rim A362-01 PresumablySth-4th century B.C. .L w I . | . . . _ , L5m>>_-ir . C&g-e_ | . | Figure(lef ) 51.and bowlRim ofrim waterA403-Ol jugA325-01 (rzght).

92 JACK L. DAVIS ET AL.

OfherCafalogued Ceramics A362-02 Amphora toe Classical-Hellenistic.

SmallFinds A359-SF01 Retouched flake Probably prehistoric.

A399-SF01 Marble fragment Possible architecturalfragment.

SITE 010

Area:Zone 4 Associatedtracts: A324, A325, A403, A404 Rangesrepresented:A-CL (2); A-HL (19); A-R (1); CL-HL (1); O-M (2); MO-LO (2) Description:A moderately dense scatter of artifacts in a low saddle just southeast of the point where the boundaries of zones 3, 4, and 5 meet. Artifacts were also found in minor drainages beneath the saddle on the eastern face of the ridge. Finds were most widespread in zone 5.

. _

_ __ ,. l _ - _

Sssociafed Ceramics A325-01 Rim of narrow-mouthed water jug(?) Figs. 51, 52 Diam. est. rim 0.039. Only one handle preserved.Not much eroded. "Local"ware (plain counterpart of the glazed items above): brown, clean-textured, with red tint at handle-core. Handle of flattened section, rising a little above rim level, slightly ridged. Middle-Late Ottoman.

A403-01 Bowl rim Figs. 51, 52 Diam. est. rim 0.27. "Local"ware, light orange-brown. Cream- yellow glaze over slip on interior, continued as a thinner (light yellow) glaze over exterior; standard added colors. Wide tilted rim with slightly thickened lip, grooved on outside. Decoration on top of rim: dark green and dark yellow splashes. 18th-beginning of the 20th century A.C. Middle-Late Ottoman. o f. _ M | --a ) cm-> t \ '

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 93

( \: \ - - 'RA-- _- r'' A024-01 l lEtt A325-01

- sz

.__

A2 15-02

A403-01

''' %**s ,:-R l )

- X /4-- W, ' -/c r B342-01

A048-02

4r n A047-01 ts-wss s%F=

0 5 2 -

Figure52. Potteryfrom sites 006410. J.W. Hayes 94 JAC K L. DAVI S ET AL.

SITE 011

Area:Zone 3 Associatedtract: A410 Rangerepresented: A-HL (5) Description:Moderate quantities of artifactsin a saddleof the ridgethat divideszone 3 fromzone 5. Artifactswere mostly found in an eroding animaltrack.

SITE 012

Area:Zone 3 Associatedtracts:A390, A391, A413,A414, A416,A432, A433,A434, B261 Periodrepresented: M (1) Rangesrepresented: A-CL (8); A-HL (30);A-R (1) Description:Near the junction of thecoastal ridge and the ridge that forms the boundarybetween zone 3 andzone 5. The mostsubstantial concen- trationswere on the faceof the coastalridge in the vicinityof a fenced houseand garden complex. Plowed fields there yielded a moderatequan- tityof Archaic-Hellenistictile.Tile was also recovered from a grassyslope betweenthe fencedproperty and brambles under the coastalridge road.

SmallFinds B261-SF01Marble fragment

SITE 013

Area:Zone 3 Associatedtracts: A445, A467, A471, A476, A477, A479, A480, A498 Periodsrepresented: MHL (1);M (4) Rangesrepresented: A-HL (11);CL-HL (1);R-M (1);UNKN (3) Description:Associated tracts on the northernside of the mainroad that runsthrough zone 3. Thereis a low-densityscatter of artifactsover a large areaand it is notobvious that they are concentrated in anyparticular tract.

CataloguedCeramics A445R-01Medium-coarse plain amphora rim Fabricwith reddish brown grits. 3rd-2ndcentury B.C.

SITE 014

Area:Zone 2 Associatedtracts: A493, A495 Rangerepresented: A-HL (2) .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 95

Description:Low-density scatter on a low knollimmediately north of a passin thecoastal ridge in thenorthwestern corner of zone3. A roadleads throughthe pass, then down into a sandand clay quarry.

SITE 015

Area:Zone 12 Associatedtracts: A569, A204, A205 Periodrepresented: R (1) Rangesrepresented: A-HL (18);CL-HL (4) Description:High-density concentrations of artifacts eroding from terrace scarpsat the southern side of a fieldroad at the edge of a sheerdrop into a modernclay quarry.

SITE 016

Area:Zone 8 Associatedtracts: A142, A143, A144, A145 Periodrepresented: LCL (1) Rangesrepresented: A-CL (1);A-HL (13);LCL-EHL (2); LCL-MHL (1);UNKN (2) Description:A narrowfenced garden (recently established in a formerolive grove)on the northernface of the ridgethat divides zone 8 fromzone 9. The sitestretches from the crestof the ridgeto thevalley bottom. Finds seemconcentrated on the upperterraces where the owner,Adem Roda, unearthedan ashlarblock in the courseof digginga pit. He showedus variousartifacts he hadrecently collected, including a pithosrim said to havebeen found nearby and an inscribedgrave stele said to havebeen foundca.300 m to the south.

AlssociatedCeramics A142-01 Red-figure(or patterned) lekanis lid Figs. 53, 60 Diam.of stem0.036; p.H. 0.023. Whole of stemof knoband part of top of lid proper(two joining pieces). Smooth clean light brown ware withorange tint, nonmicaceous (as "local" fine ware); red wash on outer (top)surface, with decoration added over this in black.Lower surface brush-smoothed,with traces of a thindark slip. Standardlate red-figure type (as in Apulianred-figure, etc.): shallow domedtop (downturnedvertical rim at bottomnot preserved), short thickstem, recessed top, tilted, around a smallcentral well (the squarish projectingrim of theknob is lost).On top,sepia wash inside well, probablytwo black circles on tiltedtop (poorlypreserved); row of black blobs,continuing (traces of radialstripes?) on bodyof lid proper.Below/

- - - - aroundthese, remains of blackbands/blobs (could be a red-figurefrieze, poorlypreserved). Derivative of anAttic type; could be eitherApulian Figure53. Red-figurelekanis lid or a local/regionalproduct. A142-01 4th centuryB.C. LateClassical. 96 JACK L. DAVIS ET AL.

A143R-O1Lid knob (banded?) Figs. 54, 60 Diam.knob 0.037. Smooth yellowish-tan "local" fine ware. Red slip coveringthe exterior,polished. Discoid top, raised a little.Traces of possibleblack banding on the slippedsurface outside, below the disc, andon bodyat thelower break. Technique perhaps to be comparedwith thatof the lekythoiM00-16 and M00-17 (see Appendix 2). Type unknown. Archaic-Classical.

OtherCatalogued Ceramics Figure54. Lid knobA143R-O1. Scale 1:1 A145-01 Fineblack-glaze body sherd Shallowvertical ribbing on exterior. 4th-3rdcentury B.C. LateClassical-Hellenistic.

SmallFinds S016-SFO1Fragment of inscribedstele Figs. 22, 23 Finewhite crystalline marble. Flat top; quarter-round molding beneath.W. 0.30.Still in possessionof Roda.On the form,see Cabanes andDrini 1995, no. 12 (5th-4thcentury B.C.). On frontnear top: ENQI OIAOKPATETE The normalformula for epitaphs at Durresemploys the nominative orvocative case for the deceasedfollowed by a patronymicin thegeni- tive.In the caseof thisinscription the sequenceappears to be nomina- tivefollowed by genitive, with ENQI beingfeminine nominative and OIAOKPATETEbeing masculine genitive. The femininename £VX is attestedat Durres(Cabanes and Drini 1995, p. 125,no.336) and the masculinename oCX0%pTN5 appearsin aninscription from Odessa honoringan Epidamnian (Cabanes and Drini 1995, p. 154,no.516). A nominativeform £VCt) wouldbe paralleledin theliterary Doric dia- lect(Thumb 1932, p. 140,no. 4; Sihler1995, p.333, n. 1).There are no masculinenouns known at Durresthat end in -£05, norany example anywherein the Greekworld of the nominativePhilokrateus (LGPN I). Forthe masculinegenitive form oCX0%pT£05, compareCabanes and Drini1995, p. 82, no.62 (forgenitive -£05 > -£05 byraising, see Buck 1955,pp. 40, 92;for -£0 > -£D in Corinthian,p. 164;in Megarian, p. 165).Reading ENQI as a masculinedative from the nominative Xenosseems less likely; this name is possiblyattested as a propername at Butrint(SEGXXXVIII 476; Cabanes 1974, p. 124,no. VI, 9), but notat Durres.Forms of letters:E withparallel bars; E withsplaying bars; withoutvertical bar. Missing are any of thelate forms of letters 89.We aregrateful to Merle characteristicof Epidamnian epitaphs of the3rd-lst centuryB.C. (e.g., Langdonand to Holt Parkerfor lunatesigma and epsilon; minuscule omega; and broken-bar alpha); see readingand commenting on a Cabanesand Drini 1995, p. 51.89 preliminaryversion of this catalogue Classical-EarlyHellenistic. entry. 4 _l w , W X i -- _ 3 _ _

.. DURRES REGIONAL ARCHAEOLOGICAL PROJECT 97

SITE 017

Area:Zone 7 Associatedtract: A277 Rangesrepresented: A-HL (1);CL-HL (3) Description:Denseand highlylocalized concentration oftiles on the north- ernface of a knollin the ridgethat divides zone 7 fromzone 8.

SITE 018

Area:Zone 9 Associatedtract: A185 Periodrepresented: HL (5) Rangesrepresented: A-HL (1);A-R (2);LB-MO (1) Description:A partiallyintact tile gravein the scarpof a terraceon the northernslope of the hill of Dautaj.Cover tiles protrude from the scarp and form an archover several diagnostic Hellenistic sherds not collected. Moretiles were found nearby.

SssociatedCeramics A185-01 Handleand body sherd of a fineclosed shape Fig.55 Inciseddecoration on exterior. 15th-17thcentury A.C. LateByzantine-Middle Ottoman.

,<

; i t - :

'i i i i Mix' u

D _ _ 0 0 0 i 1 h z

^;S1^1i _

__i1i=_=__

Figure55. Handleand body sherd A185-01

SITE 019

Area:Zone 3 Associatedtracts: B297, B298 Periodrepresented: M (1) Rangesrepresented: A-HL (4);UNKN (1) Description:Low-density scatter, mostly of tile.Much of the site is ter- racedand the landscape appears to be stable. 98 JACK L. DAVIS ET AL.

SITE 020

Area:Zone 2 Associatedtracts: B304, B305, B307, B308, B309, B310, B311, B313 Periodrepresented: M (1) Rangesrepresented: A-CL (5);A-HL (16);A-R (2);CL-HL (1);CL-R (1);O-M (5);MO-M (1);UNKN (5) Description:A moderatelydense scatter of artifactson slopesat the eastern endof theridge that borders zone 2 to thenorth. These slopes are mostly terracedand are not very steep. There does not appearto be a significant amountof erosionat present and the landscape seems to berelatively stable. Findsfrom the siteinclude those typical of an ancienthousehold assem- blage:several ancient amphora fragments, fine black-glazefragments, a fewplain wares, a cookingware fragment, and tiles.

AssociafedCeramics B313-01 Basefragment of smallskyphos(?), decorated Figs.56, 60 Diam.est. foot 0.042. One side restored. "Local" fine ware, light yellow-tan.Sepia slip on inside.Small splayed foot, offset from center of bottom.Bottom reserved (plain?), red-brown painted band on rounded undersideof foot,dull black band on edgeof foot,possible remains of a redband on topof foot.A derivativeof Corinthianand similar skyphoi. 6th-5thcentury B.C.

SITE 021 Figure56. Slyphos basefragment Area:Zone 6 B313-01 Associatedtracts: B032, B034, B036, B037, B070 Periodsrepresented: CL (24);M (2) Rangesrepresented:A-CL (7); A-HL (2);A-R (6);CL-HL (11); UNKN(1) Description:Located on terracedslopes just below the flattenedtop of a hill in the ridgethat divides zone 5 fromzone 6. Highestconcentrations of artifactswere observed in erodingscarps of terracesand in thebank of a fieldroad.

Associafed Ceramics

B036-01 Black-glazepedestal(?) foot fragment, Figs. 57, 60 probablyfrom a closedform Diam.est. base 0.095. Part of one sideonly. "Local" fine ware: light tan-orange;worn black-glaze on exterior(down to edgeof resting surface).Bottom plain. Low splayed foot with moldings on exteriorand a smallstep on underside.Broken away at junction with narrow lower body. Classical. fragment. B036-01 (lefi) and krater -- -

DURRES REGIONAL ARCHAEOLOGICAL PROJECT 99

Figure57. Black-glazepedestal(?) fragmentsB037-01, B036-02 (right)

B037-01,B036-02 Black-glazekrater, shoulder and Figs.57, 60 handle-end Probablya column-krater. Four joining pieces, plus one additional joiningsherd (B036-02). P.W. shoulder 0.155; Diam. est. at belly0.37. Smoothyellow- to orange-brownware; no obviousinclusions. Polished blackgloss ("worn" appearance) on exterior,over a smoothsurface with slightred tint (cf.appearance of M00-01;see Appendix 2). No slipon interior.Wide-bellied with rounded shoulder. A rectangular-sectioned appliedstrip, rising at steepangle, with outcurved lower end, preserved abovethe belly, presumably part of lowerterminal of a handle(e.g., a transversearched element from which rises a verticalhandle-strap, as on Laconianblack-glaze kraters). A variantof late6th-5th century B.C. Corinthian/Laconian/Attictypes, with distinctive treatment of thehandle- ends(based on a metaltype?). (Alternatively, if inverted, this could be interpretedas a dinosbase with attached feet.) Archaic-Classical.

SmallFinds B032-SF01Metalfragment

SITE 022

Area:Zone 6 Associatedtracts: B011, B016, B019, B021, B022, B101 Periodsrepresented: CL (1);HL (3);M (19) Rangesrepresented: A-CL (2);A-HL (26);CL-HL (14);LCL-HL (3); MO-LO (1);LO-M (38);UNKN (3) Description:Dense scatters of artifactsin tractsat the easternend of the ridge borderingzone 6 on the north.Several fields had been plowed when this area wasexamined. There is littleactive erosion. The modernartifacts appear to be associatedwith contemporary farms and new houses. IOO JAC K L. DAVI S ET AL.

UncataloguedCeramics Fine,black-glaze body sherd from a closedshape. Possible red- figuredecoration. Classical.

SITE 023

Area:Zone 6 Associatedtracts: B053?, B054, B055, B220 Periodrepresented: HL? Rangesrepresented: A-HL (4);A-R (10);CL-HL (4);R-M (3); UNKN(1) Description:A low-densityscatter of artifactson twohilltops in the ridge dividingzones 5 and6. Artifactswere most dense in a gullynear the crest of oneof the hillsand in a plowedfield nearby.

AssociatedCeramics BOSS-O1Tile fragment Fig. 60 H. of edge0.065. Rectangular chamfered inset under corner (0.11 x 0.08-0.09).Normal "local" light yellow-brown fabric (like that of"local" finewares). Hellenistic?

SITE 024

Area:Zone 6 Associatedtracts: B060, B061, B062, B064, B080, B081, B082, B084, B115, B117 Periodsrepresented:A (1); HL (1);HL? (1); M (3) Rangesrepresented: A-CL (1);A-HL (12);A-R (8);LO-M (2); UNKN(7) Description:Low-densityscatterof artifacts on the lower slopes ofthe coastal ridgeat thewestern end of thevalley. The sitehas no clearcenter.

AssociatedCeramics B084-01Bowl with horizontal handles below rim Figs.58, 60 Shallow"skyphos"? Diam. est. rim ca. 0.23; p.H. 0.039. "Local"(?) ware:light (brownish?) yellow ware, medium-soft, with fine specks only. Tracesof dullsepia paint preserved in the rimgrooves, possible sepia bandson thehandle. Bowl upcurved toward rim; top of rimflattened. Partof a round-sectionedhandle (presumably one of a pair)on upper wall,set horizontally,rather wide-looped (W. est.ca. 0.075). Three(?) fineshallow grooves below the rim.Possible pattern of thinvertical stripes(four per cm) around outside of handle. Figure58. Bowlwith horizontal Shapeand decorative treatment suggest a dateof ca.650-550 B.C. handlesB084-01 DURRES REGIONAL ARCHAEOLOGICAL PROJECT IOI

B060-01Tile fragment Fig.60 H. of edge0.062. Piece of rim;no cornerspreserved. "Local"(?) ware:light (brownish?) yellow ware, medium-soft, with fine specks only. Somediagonal scratches on rimslope, possibly ancient? Hellenistic?

SmallFinds B061-SF01Bitumen?

B084-SF01Bitumen?

SITE 025

Area:Zone 6 Associatedtracts: B121, B122, B136, B138 Periodrepresented: CL (1) Rangesrepresented: A-HL (7);A-R (3);CL-HL (4);UNKN (2) Description:Located on a hilltopin the ridgethat divides zone 6 from zone7. Terraces on thehill are stable and planted with well-tended olives.

SITE 026

Area:Zone 6 Associatedtracts: B128, B129, B134, B135 Periodsrepresented: CL (1);HL (6);M (10) Rangesrepresented: CL-HL (5);CL-R (1);UNKN (1) Description:Perhaps two separate concentrations. All of thetracts included in the site consistof relativelystable terraced fields, and artifacts were observedonly in areaswhere the soil had been disturbed, e.g., in ditches,a gully,and an eroding terrace scarp. S026 and S025 may be parts of a single largesite.

AssociatedCeramics B135-02 Amphorarim Figs.59, 60 Diam.est. rim 0.15. Brown ware (yellowish tint), rather rough- textured.Fine calcite lumps, fine brown specks. Not local:from a karstic environmentof a WesternGreek or Illyrian region? Tapered rim, tilted upward;a slight groove and offset at junction with neck. Probably5th-4th century B.C.

OtherCatalogued Ceramics B135-01Tile, burnished on interior Figure59. Amphora rim B135-02 Hellenistic. - - - 1 -

JACKL. DAVIS ET AL.

I02

-)--1 s--I ss B313-01 B036-01 A143R-01

t-8 /)

-

- A142-01 B135-02

B285-01

m

B084-01

B060-01

.'fo ''"" , F /} #th W'-W f t '.-;j,¢F l 4 \ :0 \ - -

B037-01 /B036-02

2 5 o 5 cmI _ _

Figure60. Potteryfrom sites 016, 020, 021, 023, 024, 026, and028. J.W. Hayes \1I| _ ^_ |__ li 11 l| I-I Il E | l I | l E | - g ht | I_ _

DURRES REGIONAL ARCHAEOLOGICAL PROJECT Io3

SITE 027

Area:Zone 7 Associatedtracts: B166, B167, B169, B170,B171 Rangesrepresented: A-CL (2);A-EHL (1);A-HL (12);CL-HL (3); CL- R (5); LCL-EHL(2) Description:A low-densityscatter of artifactsat a high elevationat the westernend of thevalley, below the coastalridge road. Slopes are in part steepand prone to erosion.The extent and focus of subsurfacedeposits are unclear.

SITE 028

Area:Zone 2 Associatedtracts: B282, B284, B285, B286 Rangesrepresented:N-BA? (1); A-CL (1);A-HL (11);A-MED (1);CL- HL (2); R-O (1);MED-M (8); O-M (11);MO-M (3);LO-M (1) Description:Moderately dense scatter of artifactsin a modernroadbed and adjacentfields. The sourcefrom which sherds are being eroded has not beendetermined.

,sa

_

h ! | | X | | | T^ _

z _ | I _ l I l * | _ _ . _l l - l l l | l _ _ 11111I I I | l _JF _ Figure61. Bowl orjar base B285-01 (lept) and corewith end scraper B284-SFO1(right)

AssociatedCeramics B285-01 Bowlor jar base Figs.60, 61 Diam.est. base 0.16. Ware and treatment as B283-01(possibly sameform, but that sherd is unassociatedwith a site).White glaze on interior.Flat base, slightly hollowed; slight external molding. Chattering markson bottom. LateOttoman-Modern.

SmallFinds B284-SF01Core with end scraper Fig.61 Neolithic-BronzeAge? __ and B081-SF04_ l (notchedFigure (notched62.flake), Lithics B081-SF02 B081-SFOlflake). (flake),

Io4 JACK L. DAVIS ET AL.

_.

Scale 1:1

SITE 029

Area:Zone 6 Associatedtract: B081 Periodrepresented: N? (8) Description:Localized cluster of chippedstone (chert) artifacts, all found withinabout 20 m of oneanother on thesteep slopes of a hill.The soilat the siteis sandyand there is activeerosion.

SmallFinds B081-SF01-SF08Lithics Fig.62 Fiveflakes and three notched flakes. Neolithic.

CATALO GUED ARTIFACTS UNAS S O CIATED WITH SITES

POTTERY A192-01 Rimof largejar or basin (type uncertain) Figs.63, 64 Diam.est. rim 0.36. Dull brown, fairly hard-fired; white and brown/ blackspecks, trace of glintingspecks. More or less cooking fabric. Flat rimwith faint groove above edge, vertical lip bearingtwo grooves; near- vertical"neck." Dateand source unknown (perhaps Hellenistic?).

A428-01 Baseof largebowl(?) Fig.64 Diam.est. of base-molding0.155. Smooth-textured light yellow clay;no paintor slipvisible (but could perhaps have flaked off). Flat- based,with heavy external molding and somewhat hollowed bottom. Innersurface very smooth, so presumablyan open form. Ware close to Corinthian,but the form is not a veryobvious Corinthian one, so perhapsof colonialorigin. Looksearly (Archaic-Classical).

B001-01 Black-glazebowl base fragment Figs.63, 64 Diam.est. foot 0.049. "Local" fine ware: light brownish-yellow; poor thinworn black slip/gloss, attested down to loweredge of foot(and B001-01,B005-01. black-glazeBottom bowlrow: rim I lidt ! ' or 1 5 foot2 1 1 1 __ 1 -

DURRES.. REGIONAL ARCHAEOLOGICAL PROJECT Io5

hIf ' ' ' - - XA

Figure63.Toprozv:jarorbasinrim j b b U | | ^ e A192-01,black-glazebasefragment i| t ^ fragmentB074-01, jar rim B205-01. Scaleas shown for BOOl-Ol.

probablyending there; not on bottom).Tilted footring, rounded floor withslight cone under center. Probably from common Late Classical- Hellenisticbowl with incurved rim. 4th-2ndcentury B.C.

B005-01Black-glaze rim of largekraterlike bowl Figs.63, 64 Diam.est. rim 0.29. "Local" smooth fine ware: orange-red break, orange-brownsurfaces. Remains of thinblack slip inside and outside. Shortoutcurved rim with triangular lip. 4th-3rdcentury s.c.?

B074-01 Lid(or pedestal foot?) fragment Figs.63, 64 Diam.est. at base0.055; p.H. 0.016. One side (of rim?).One side of drawingrestored. "Local" ware, light brown. Paint banding on exterior/ top;interior plain, smoothed. Part of a flattishtop, steep side, thin splayedlip. On wall,dull black painted band; possible traces of fugitive redpainted band above. Remains of twoblack lines preserved on thetop (surroundinga missing knob?). Variant of Corinthianpyxis lid shape? LateArchaic/Classical?

B205-01Jar rim, Apulian Geometric Figs.63, 64 Diam.est. rim ca. 0.22; p.H. 0.034. Light yellow, rather soft and powdery;slightly creamier surface. Some fine brown/gray specks, rare faintglinting specks. Rim possibly polished on bothfaces, interior of bodynot polished. Rim of unevenwidth, perhaps finished on a turn- table.Technique of bodyuncertain; handmade? Wide-flaring rim, sharplyoutturned, tapered at lip;top of a rounded(?)body. Possible remainsof blackpaint on body:band at junction with rim, slanting lines below?More likely Apulian Geometric than local equivalent of Iapygian Geometricof the Salento.Jar related to Apulian/LucanianEarly Iron Agevessels with "a tenda" decoration. 8th-6th/5thcentury B.C. Io6 JAC K L. DAVI S ET AL.

:w: " X

*'||--M ss: f = + S A428-01 B001 -01 B074-01

... _. . .

_ . .- _ __-_ \

B283-01 B005-01

l l f ______=-

_ -- -| B205-01 A1 92-01

0 5 2 5 I l cm Figure64. Potteryunassociated with sites.J.W. Hayes B283-01Bowl rim Figs.64, 65 Diam.est. body 0.236."Local" ware, light brown, smooth-textured. Whiteglaze on interior.A fewpatches of slipare visible under the (otherwise glaze notpreserved). Thick-walled. Rounded form, slightly incurvedat rim. Roundedrim, rippled exterior (four shallow hollows andthree shallowridges preserved). Possibly the finalversion of the seriesA193-01 (from S001),A215-01 (from S008). For a possiblebase- form,see B285-01 (from S028). 20thcentury? Late Ottoman-Modern?

METAL A496-SF01Bronze coin90

Dyrrachiummint. 2.78 g. Diam.0.14-0.16. Obverse: head of Herakles. Reverse:arms of Herakles.Legend: ATP. See Ceka 1967, pp.38-45, 132-134. 90.We 3rdcentury B.C. Hellenistic. aregratefill to S. Gjongecaj fortheidentification of this coin. DURRES REGIONAL ARCHAEOLOGICAL PROJECT Io7

I

Figure65. Bowl rim B283-01 (lefi); :::::: ::: :: :: :: ::: : conicalloomweight A307-SF01 (right) I

TERRACOTTA A307-SF01Conical loomweight Fig.65 P.H.0.065. Very badly worn. Pierced with single suspension hole at top.C£ CorinthXII, pp.148-161, 163-170. Toobadly damaged to be datedtypologically.

Figure66. LithicsB175-01, B244-SFO1,A483-SFO1, L I T H I C S B294-SFO1.Scale 1:1 B175-SFO1Notched flake Fig. 66

B244-SFO1Blade segment, retouched? Fig. 66

A483-SFO1Unidirectionalcore Fig. 66 Formedon a pebbleor a reducedcore.

B294-SFO1Denticulate Fig. 66 Probablylate prehistoric. JACK L. DAVI S ET AL. Io8

APPENDIX2 TWOGRAV EGROUPS FROM THE *+ DURRESCE METERl ES? byJohn W;Hayes

of Thepottery described in thisappendix was presented to the Museum shown Durresby AnastasArrse (formerly of Dibra),who had previously thepots thecollection to Hotiand Davis in hishouse.91 Arrse said that all camefrom a singlegrave and pointed to whatappeared to be the remains He ofa tilegrave in thehigh earthen scarp to theside (north) of hishouse. the alsoindicated an areaof burnedsoil in the earthensteps cut from streetdown to his house. I canimagine that the vessels were associated with two distinct buri- 3rdcentury. als,one of about300 B.C., the otherdating to the mid-late the Thepostulated early group would include the smallbowl M00-04, The lelythoiM00-16 and M00-17, and possibly the smalljug M00-18. rim cut "early"lelythoi M00-16 andM00-17, however, both havethe which downin thesame way, suggesting ancient secondary use (reburial?), mightargue in favorof a singlegrave group. Is the fragmentaryskyphos Corin- M00-13part ofthe grave group or a chancefind nearby? Presumed M00- thianHellenistic products include M00-02, M00-04, the skyphos local 12,and the ring-handledskyphos M00-13. The restare presumably surface (orregional) products. Wares with a polished/burnishedslipped in andpoorly adhering added paint (as M00-16, M00-17) may be earlier black- thelocal sequence, followed (not before the 3rd century?) by proper glossvessels.

CATALO GUE 71 M00-01 Black-glazeplate with "local" stamped decoration Figs.67, Diam.0.158; H. ca.0.01. Nearly complete, in onepiece (rim neatly chipped).Clean-textured light orange ware (faint red tinge) with to smoothedsurface; black gloss all over, flaking. Double-dipping streak 8/6 oneside, visible on bothupper and lower surfaces. Munsell 7.5YR "reddishyellow" (tending to 6.25YR7/8 whendamp). on Flatfloor, thick rounded raised rim with broad groove and offset underside.Simple square-cut foot. Four unlinked palmette stamps 91. Arrse'shouse is located of tractA195 in an radiatingfrom a groove/circleat center;pair of encirclinggrooves immediatelysouth areathat was not intensivelysurveyed. (Diam.0.05). Copy of Attic4th-3rd century B.C. type,later version . . DURRES REGIONAL ARCHAEOLOGICAL PROJECT IO9

- : ::: - : - : - - :::: ;E- :::00:E- ::: :::-

Figure67. Black-glazeplate M00-01 (withoutincised arcs linking the stamps,and without scraped grooves). (left)and black-glaze dish M00-02 (right) Oneof the earliestexamples recorded of the double-dippingtreatment. Thisis notedon Corinthianrelief bowls of the earlier2nd century B.C., butis generallya Roman(Imperial) feature. For another early example, seea specimenfrom Western Greece(?) now in the BritishMuseum (Hayes1997, p. 23, pl. 6). Later3rd century B.C., presumably(the shape seems not to survive anylater).

M00-02 Black-glazedish, Corinthian(?), Figs. 67, 71 white-paintedand rouletted Diam.est. rim 0.216; Diam. base 0.08; H. 0.035.Whole of base, withone side of wideflaring rim (a singlepiece). Light grayish-buff, clean-textured.Munsell ca. 2.5YR 7/5-8/4, fired10YR on exterior. Dullgrayish gloss-slip, ending at loweredge of foot.Black dot and reservedcircle at center(crossed by a thinred arc), broad reserved line halfwayout on floor(enclosing tondo); two single lines of roulettingon outerpart of floor.Tondo: in addedwhite, poorly preserved, a large eight-pointed("Macedonian") star pattern, with a three-stroke"arrow" motifbetween each ray (set halfway out from center, pointing outward, leavinga subsidiaryunpainted disc around the centraldot andcircle). On topof rim,two or three lines of roulettingbordered by scraped/ reservedbroad lines. Lip hooked upward. Shallow groove around lower partof foot.Rouletting consists of shortneat strokes (spacing 3-4 mm): cf. CorinthVII.3, pp. 39-40. Mid-late3rd century B.C.

M00-03 Black-glazeplate/dish, stamped and Figs. 68, 71 rouletteddecoration, ware unclassified Diam.0.195; Diam. foot 0.086; H. 0.031-0.035(i.e., rim tilted). Nearlycomplete, mostly in onepiece (one rim fragment mended, two moremissing). Pale tan ware, smooth-textured; dull sepia to blackslip, IIO JACK L. DAVIS ET AL.

_ _ _ _ Figure68. Black-glaze plate/dish M00-03 endinghalfway down outside (applied by dipping,with one or two dribbles).Medium thickness. Narrowround-topped raised rim (top faceted, forming a sharpangle at inneredge). Heavy squarish foot. Two shallow grooves (Diam. 0.071) anda doubleband of rouletting(Diam. 0.099; rather long strokes) surroundingstamped tondo. Stamps: single large six-petal rosette (Diam.0.016) at center,surrounded by four large radial palmettes (0.0017x 0.0012).Stamps have internal relief details. Ware uncertain (probablynot local, but hardly Corinthian). Stamps treated in anItalian/ AsiaMinor manner. Mid 3rd-early2nd century B.C., on formand style.

M00-04 Black-glazesmall bowl with incurved rim Figs.69, 71 ("salt-cellar"),Corinthian _ _ Diam.0.064; Diam. est. foot 3.1; H. est.0.034. Unbroken, but edge of footchipped, and slip worn. Clay light yellow-tan (Munsell lOYR 8/ 0 5), withslight surface wash; dull black slip, ending on outsideof foot. | Fairlythick-walled; low foot, only partly hollowed underneath. | Undecorated.Fairly early in the series,from postulated "early" grave | group. About300 s.c.?Cf. CorinthVII.3, pp. 32-33.

The following"saucers"(M00-05 to M00-11)are all in a smooth"local" fabric,with the black gloss mostly semilustrous or rather dull They prob- Figure69. Black-glaze small bowl ablydate to the mid-late3rd century B.C. Theirsequence (if theyare not M00-04Scale 2 3 allthe samedate) is uncertain.

M00-05 Black-glaze"saucer" (small plate), "local" Figs.70, 71 Diam.0.113; H. 0.02-0.021.Complete. Mended from two pieces. Claylight yellow-brown (Munsell ca. 5YR 8/4). Blackglaze (partial) ends at restingsurface. Almost flat; low upturnedrim, footring. Incised circle at center. Mid-late3rd century B.C. .. DURRES REGIONAL ARCHAEOLOGICAL PROJECT III

Figure70. Black-glaze"saucers" _ M00-05(left), M00-06 (right), andM00-07 (bottom). Scale as shown forM00-07 M00-06 Black-glaze"saucer" Fig.70 Diam.0.113; H. 0.027-0.029.Complete apart from abrasions at rim.Mended from three pieces. Dullblack glaze, tending brownish in parts;brown stacking disc on floor(off-center). Black glaze ends as on M00-05.Rather more rounded floorthan on M00-05;one groove around center. Related to M00-05. Mid-late3rd century B.C.

M00-07 Black-glaze"saucer" Fig.70 Diam.0.111; H. ca.0.026. Complete. One side mended. Black glazeall over, dull, with various brown patches (finger marks) on exterior.Form as M00-05; incised circle at center.Cf. M00-05. Mid-late3rd century B.C.

M00-08 Black-glaze"saucer" Fig.71 Diam.0.116; H. 0.03-0.032.Base and greater part in a singlepiece. Clayfired light yellow-tan, gloss strealy sepia-black. Black glaze ends aroundtop of foot.Flattened conical form; incised circle at center.Cf. M00-05. Mid-late3rd century B.C. II2 JACK L. DAVIS ET AL.

M00-09 Black-glaze"saucer" Fig.71 Diam.est. 0.112; H. ca.0.03. Base and most of oneside (four fragments,three joining). Sloping floor, rim inturned with internal lip; coneunder base; incised circle at center.Cf. M00-05. Mid-late3rd century B.C.

M00-10 Black-glaze"saucer" Diam.est. ca. 0.115; H. est.0.024. About half preserved: whole of base,joining rim sherd, loose rim sherd. Partial slip/black glaze, ending underouter part of floor.Flattish form, upturned rim; incised circle at center.Cf. M00-05. Mid-late3rd century B.C.

M00-11 Fragmentof black-glaze"saucer" Diam.base ca. 0.048. Single piece: base and part of floor.Slip ends at restingsurface. Small spike below center. Groove (Diam. 0.015) aroundcenter of floor.Type apparently the sameas M00-05to M00-10. Mid-late3rd century B.C.

M00-12 Black-glazesmall skyphos, Corinthian Fig.71 Diam.est. rim 0.072; Diam. body 0.072; H. 0.075.A rimfragment withhandle broken, now mended. Parts of rimand one handlemissing. Corinthianware: light yellow, clean-textured; smooth black gloss, rather dull,ending on lowerexterior. Some dilute black finger smears on unpaintedarea above foot. Bodymarkedly contracted in lowerpart, with small foot; outcurved lip;thin, nearly horizontal handles, only slightly contracted toward wall (W. max.0.034, close to body0.028). Standard Corinthian 4th-3rd centuryB.C. type,later version (see Corinth VII.3, pp. 66-71). Clay is normal,with black gloss better preserved than it wouldbe at Corinth. Mid-late3rd century B.C.

M00-13 Ring-handledskyphos, probably Corinthian Fig.71 (fragmentary) Diam.est. base 0.048; p.H. 0.074 (originally ca. 0.09-0.095?). Singlefragment: about half of vesselpreserved, with lower part of one thinstraplike handle; base chipped around the edge.Restored as two- handled.Clean drab light gray ware (i.e., misfired?); remains of dull brownishgray slip all over, badly worn on innersurface. Bodymore or less hemispherical; to be restoredwith two thin verticalring handles, probably with flat spurs (finger rests) on tops. Pedestalfoot with solid stem and hollow stepped-out base. Ware similar in appearanceto thatof M00-02.Cf. CorinthVII.3, pp. 74-76. Later3rd century s.c.? - _H - - - - adS-/J) I l - : '] --

DURRES REGIONAL ARCHAEOLOGICAL PROJECT II3

M00-04

l M00-01

t_ - _ ) _" i M00-09

M00-03 - M00-08

J - ) _ - - T ;' M00-05 M00-02

_1<- (! - ,<,,}- - - <_ ?

M00-13 M00-12

o 5 2 5

cm Figure71. Potteryfrom the house of AnastasArrse. J. w. Hayes theground-down upper edge (forstem. secondary use?). "Local"(?) fabric: j_ light g

II4 JACK L. DAVIS ET AL.

M00-14 Miniatureskyphos, "local" ware Figs.72, 75 Diam.rim 0.051; H. 0.051.Body unbroken. Handles missing. Buff l ware,rather clean, fired light orange-brown under slip. Thin flaking ]g sepia-blackslip (dipped; partial coating). Finger marks visible around ffi r i i _l

Solidflat base (wire marks). Circular scars of twosmall horizontal q t handlesvisible immediately below rim. Presumably a miniature "local" _ copyofthetypeof M00-12. gi,,! 1,= Mid-late3rd century B.C. orlater. j a

M00-15 M1nzatureskyphos, "local" ware Fig.75 Diam.0.049; H. 0.048.Handles missing (one lost recently). Clay orange,fairly fine; thin slip, red to brownto black,partly lost. Slip Figure72. Miniature slyphos appliedby dipping,ending on outsideof toe. M00-14 Variantof typeof M00-14,with solid peglike toe (wire-drawn).A smallclay knob (accidental?) preserved flanking one handle. Mid 3rd-early2nd century s.c.?

M00-16 Talllekythos (copying late Attic/Apulian shape) Figs.73, 75 __ Diam.body 0.103; Diam. foot 0.07; p.H. 0.18. Body, handle, and _ % mostof neckin onepiece. Rim cut off deliberately,with a slanting, = g yellow-orange,with polished (orange-brown) outer surface bearing _ painteddecoration in fugitivedark purplish-red (iron oxide). j Lowfootring. Original rim outcurved, perhaps partly closed at top i (i.e.,mushroom-shaped, perhaps quite wide: see profile drawing). l l Painteddecoration: outside of rimand most of handlecoated, vertical stripesdown handle; irregular pattern covering body, possibly to be | restoredas a largefemale head, facing left, with hair-covering (small . l chignonat upperright?), or conceivablytwo heads (confronted?). A l paintband encircling the foot. Motifat thisdate, on the analogyof lateAttic or Apulian red- figure,most likely to be as described,spread out to coverthe whole surface,or a largeupright palmette (which cannot be restoredfrom what remainson thisvessel). This vessel type and the decorativemotifs are hardlyto be expectedlater than ca.300-280 B.C. _ _ __ Late4th-early 3rd century B.C. Figure73. Tall lekflhos M00-16 M00-17 Talllekythos Diam.body 0.11; Diam. foot 0.07; p.H. 0.195. In onepiece, with rimcut down in antiquityas on M00-16.Painted motifs perhaps similar (notyet cleaned).Same type and ware as M00-16, a littlemore slender. Late4th-early 3rd century B.C.

M00-18 Smalljug/olpe (type with high-swung handle) Figs.74, 75 Diam.body 0.076; H. to rim0.115. Body complete (in one piece), exceptfor rear part of rim;handle missing (restored on profiledrawing). Low-belliedform with short narrow neck, splayed hooked rim. Two ------. -

DURRES REGIONAL ARCHAEOLOGICAL PROJECT II5

- ::::::::::::- ::::::- :::::::::::-

Figure74. Smalljug/olpe M00-18 (left),and black-glaze amphoras appliedknobs on topof rimat sides,presumably flanking a handlerising M00-19(center) and M00-20 (right) fromtop of rimand curled downward. "Local" series; poor traces of a dullblack slip. Later4th century(derived from a standardGreek type) or possibly 3rdcentury B.C., undefined

M00-19 Black-glazesmall amphora Figs.74, 75 Diam.body 0.123; Diam. rim flange 0.06; Diam. foot 0.066; H. 0.148.Complete (whole of rimdetached, recent damage?). Fabric presumed"local": clean light orange ware; dull black glaze (applied by dipping),partly flaked off, ending on lowerbody (above bottom of the fluting).Rim flange all black. No evidenceof addedcolors. Inner surface of bodycompletely coated with lime deposit. Bottom undecorated. Taperingneck, with triangular rim flange (to supporta lid not present).Flat-sectioned handles bearing two grooves each. Pair of fine grooveson upperneck, slight molding at junction of neckand body. Broad-belliedbody, low molded footring. Body covered with fine, close- setvertical fluting, continuous below handles (i.e., applied first), slightly skewedto left at bottoms;a finegroove forms lower border. Mid-late3rd century B.C. (probably).Fluting on bodyparallels that on laterGnathia ware (ca.300-240 B.C. orlater); may or maynot be earlierthan M00-20 below.

M00-20 Black-glazesmall amphora Figs.74, 75 Diam.body 0.102; Diam. rim flange 0.082; Diam. foot 0.052; H. 0.146.Body preserved in onepiece. New breaks at top;one handle mended."Local" ware; clay as M00-19, thin dull sepia-black slip (appliedby dipping), ending on lowerbody at levelof groove.Interior of bodycoated with lime deposit. No evidenceof addedcolors. Top of rim flangeand outside of rimreserved, coated with thin red-brown wash. Horizontalbrush marks from finishing. K \ I __

II6 JACK L. DAVIS ET AL.

MOO-14

T:y' f'' - y ::

M00- 18 M00- 19

\ 4 j08:g3,

M00-20 M00- 16

0 5 2 5 Figure75. Pottery from the house of - { cm AnastasArrse.J.w.Hayes

Typerelated to M00-19,but less baggy: wide, nearly cylindrical neckwith thin rim flange; high-bellied rounded body, footring of small diameter.Handles from below rim flange to abovebelly, each bearing twogrooves/median arris. Body undecorated (no fluting). Pair of fine grooveson neck,slight ridge at junction with body, groove on lower body. Mid-late3rd century B.C. (perhapsthe latter).Later in seriesthan M00-19. DURRES REGIONAL ARCHAEOLOGICAL PROJECT II7

REFERENCES

AgoraV = H. S. Robinson,Pottery of the Ceka,H. 1967.Probleme te numis- of Sanctuariesin the Countryside RomanPeriod: Chronology (The matikesilire [Problemsin Illyrian of Etruria and Magna Graecia _ r m. AthenianAgora V), Princeton1959. numlsmatcs , . lrana. (700w00B.C.), Stockholm. Alcock,S. E. 1993a.Graecia Capta: Ceka,N., andM. Zeqo.1984. "Kerk- Errington,R. M. 1989. "Romeand TheLandscapes of RomanGreece, ime nenujorene vijenbregdetare Greeceto 205 B.C.," in CAH2 VIII: Cambridge. dhe ujrate brendeshmete vendit Romeand theMediterranean to 133 . 1993b."Breaking Up the tone"[Underwater investigations B.C., A. E. Astin,F. W. Walbank, HellenisticWorld: Survey and alongthe coastlinesand inland M. W. Frederiksen,and R. M. Ol- Society,"in ClassicalGreece:Ancient watersof ourcountry], Monu- givie,eds., Cambridge, pp. 81-106. Historiesand ModernArchaeologies, mentet28, pp. 127-140. Fasolo,M. 2003. La ViaEgnatia I: I. Morris,ed., Cambridge, pp. 171- Cherry,J.F., andJ. L. Davis.1998. Da Apolloniae Dyrrachiumad Hera- 192. "NorthernKeos in Context,"in kleiaLynkestidos, Rome. Alcock,S. E.,J. F. Cherry,andJ. L. Kea-Kythnos:History and Archaeol- Fine,J. V. 1983. TheEarly Medieval Davis.1994."Intensive Survey, ogy,L. G. Mendoniand A. Maza- Balkans:ACritical Surveyprom the AgriculturalPractice, and the Clas- rakisAinian, eds., Paris, pp. 217- Sixth to theLate TwelthCentury, sicalLandscape of Greece,"in Clas- 226. Ann Arbor. sical Greece:AncientHistories and Cherry,J. F.,J. L. Davis,and E. Man- Francis,K. 2001a. Explorations in Alba- ModernArchaeologies,I. Morris, ed., tzourani.1991. LandscapeArchaeol- nia. Relocatingthe Cavesof Luigi Cambridge,pp. 137-170. ogyas Long-TermHistory. Nlorthern Cardini,Institute of WorldArchae- Bejko,L., andM. G. Amore.2000. Keosin the CycladicIslands, Los ology,University of EastAnglia, AlbanianRescueArchaeological Unit, Angeles. privatelycirculated. AnnualReport2000, Tirana. CorinthVII.3 = G. R. Edwards,Corin- . 2001b."The Lost Cavesof Bintliff,J. 1997."Regional Survey, thian HellenisticPottery (Corinth LuigiCardini: Explorations in Demography,and the Riseof VII.3),Princeton 1975. Albania1930-2001," CaveArchae- ComplexSocieties in the Ancient CorinthXII = G. R. Davidson,The ologyand PalaeontologyResearch Aegean:Core-Periphery, Neo- Minor Objects(Corinth XII), Archive3, http://www.shef.ac.uk/ Malthusian,and Other Interpretive Princeton1952. capra/3/cardini.html. Mod els,"JFA 24, pp. 1-3 8. Crawford,M. H. 1985.Coinage and Grant,M. 1946. FromImperium to Brunt,P. A. 1971.Italian Manpower Moneyunder the RomanRepublic: Auctoritas:AHistorical Study of 225B.C.-A.D. 14, Oxford. Italy and theMediterranean Economy, Aes Coinagein theRoman Empire, Buck,C. D. 1955. The GreekDialects: Berkeley. 49 B. C.-A.D. 14, Cambridge. Grammar,Selected Inscriptions, Dankoff,R., andR. Elsie.2000. Gutteridge,A., A. Hoti, andH. R. Glossary,Chicago. Evliya (elebi in Albaniaand Adja- Hurst.2001. "The Walled Town Burnett,A., M. Amandry,and P. P. centRegions (Kosovo, Montenegro, of Dyrrachium(Durres): Settle- Ripolles.1992. Roman Provincial Ohrid),Leiden. mentand Dynamics," JRA 14, CoinageI: Fromthe Death of Caesar Davis,J. L. In press."Are the Land- pp. 391-410. to the Death of Vitellius(44 B. C.-A.D. scapesof GreekPrehistory Hidden? Hammond,N. G. L. 1966. "TheKing- 69).Pt. I:Introduction and Catalogue, A ComparativeApproach," in domsin Illyriacirca 400-167 B.C.," London. Side-by-SideSurvey. Comparative BSA 61, pp.239-253. Cabanes,P. 1974."Lesinscriptions du RegionalStudies in theMediterranean .1967. Epirus:The Geography, Theatrede Bouthrotos,"in Actesdu World,S. E. Alcockand J. F. Cherry, theAncientRemains, the History, Colloque1972 sur l'esclavage, Paris, eds.,Oxford. and the Topographyof Epirusand pp. 105-209. Davis,J.L., S. E. Alcock,J.Bennet, AdjacentAreas,Oxford. . 1976.L'Epire de la mortde Y. G. Lolos,and C. W. Shelmer- .1989. TheMacedonianState: Pyrrhosa la conqueAteromaine (272- dine.1997."The Pylos Regional The Origins,Institutions and History, 167 av.; C.), Paris. ArchaeologicalProject, Part I: Oxford. . 1988.Les Illyriens de Bardylisa Overviewand the Archaeological Hammond,N. G. L., andF. W. Wal- Genthios(IVe-IIe sieclesavantg-C.), Survey,"Hesperia 66, pp.391-494. bank.1988. A Historyof Macedonia Paris. Ducellier,A. 1981.LafaJcade maritime 3: 336-167 B.C., Oxford. Cabanes,P., and F. Drini.1995. Corpus de lAlbanie au MoyenAge: Durazzo Harff,A. von.1860. Die Pilgerfahrdes desinscriptions gresques d 'Illyriemeri- et Valonadu XI e au XVe siecle, RittersHrnoldvon HarJf E. V. dionaleet d 'Epire1: Inscriptions Thessaloniki. Groote,ed., Cologne. d'Epidamne-Dyrrhachionet d'Apol- Edlund,I. E. M. 1987. The Godsand Hayes,J. W. 1997. Handbookof Medi- lonia (Etudesepigraphiques 2), Paris. the Place.The Location and Function terraneanRoman Pottery, London. JACK L. DAVIS ET AL. II8

Helm,R. 1956.Die Chronikdes Hiero- Hoti, A. 1989a."Ene me glazurenga Picard,O. 1986."Illyriens, nymous(Eusebius Werke 7), Berlin. Thraces,et qytetii Durresit(shek. X-XV)" Grecs:La monnaiedans les Heuzey,L. 1886.Les operationsmili- rapports [Glazedvessels from the cityof entrepopulations tairesde Jules Cesar, Paris. grecqueset non Durres(lOth-15th centuries)], grecques," Heuzey,L., andH. Iliria 1986,1, pp. 137- Daumet.1876. Iliria 1989,1, pp.213-240. 144. Missionarcheologique de Macedoine, . 1989b."Durres:Qytet" Paris. Pluciennik,M. 1996."Interim Report [Durres:The city],Iliria 1989,2, on Surveyin the Areaof Hidri,H. 1983."Germimene nekro- Butrint, pp.293-294. SouthernAlbania (1996)," poline Dyrrahut:Kodrat e Dautes unpub- . 1996."Tedhena arkeologike lishedmanuscript. (viti1977)" [Excavations in the perkrishterimin e hershem ne Polignac,F. de. necropolisof Dyrrachium:The 1994."Mediation, hill Dyrrah(shek. IV-VII)" [Archaeo- Competition, of Dautaj(1977)], andSovereignty: The Iliria 1983,1, logicalinformation concerning Evolution pp. 137-180. of RuralSanctuaries earlyChristianity in Dyrrachium in GeometricGreece," in . 1986a."Prodhimii qeramikes Placing (4th-7thcenturies)], Iliria 1996, the Gods:Sanctuaries and Sacred vendesete Dyrrahutne shek.Vl-II 1-2, pp. 173-181. Spacein AncientGreece, S. E. p. e. sone"[The production of local Alcock Jirecek,K. 1916."DieLage und Ver- andR. Osborne,eds., potteryat Dyrrachiumin Oxford, the 6th- gangenheitder Stadt Durazzo in pp.3-18. 2nd centuriesB.C.], Iliria 1986,1, Albanien,"in Illyrische-Albanische . pp.187-195. 1995.Cults, Territory, and the ForschungenI, L. vonThalloczy, ed., Originsof the Greek . 1986b."Nekropoli antik i City-State, Munich,pp.152-167. J. Lloyd,trans., Chicago. Dyrrahut:Rezultatet e germimeve Karaiskaj,G., andA. Bajce. 1975. Praschniker,C., andA. Schober. te viteve1973 dhe 1980"[The 1919. an- "Kalajae Durresit dhe sistemii for- Archaologische cientnecropolis of Forschungenin Alba- Dyrrachium: tifikimetperreth ne kohene vone nien und Resultsof excavationsin Montenegro,Vienna. the years antike"[The fortressof Durresand Rey,L. 1925. 1973and 1980], "Fouillesde la mission Iliria 1986,2, the systemof surroundingfortifi- franSaisea pp.99-128. Apollonied'Illyrie et a cationsin lateantiquity], Monu- Durazzo . 1988."Kupame dekor (1923-1924),"Albania 1, ne re- mentet9, pp.5-33. pp.26-48. lievte Dyrrahut"[A cupwith relief Kiel,M. 1990.OttomanArchitecture in Schmitt,O. J. 2001. decorationfrom Das venezianische Dyrrachium],Iliria Albania,1385-1912, Istanbul. Albanien 1988,1, pp.75-89. (1392-1479), Munich. Korkuti,M., J. L. Davis,L. Bejko, Schoene,A. 1866.Eusebi . l990."Qeramikaarkaike chronicorum e M.L. Galaty,S.Mujcaj,and S. R. libri duoII: Dyrrahut"[The Chronicorumcanonum, Archaicpottery Stocker.1998. "The Mallakastra Berlin. of Dyrrachium],Iliria 1990,2, Regional ArchaeologicalProject: Seiner,F., ed. 1922.Ergebnisse der Volks- pp. 161-206. First Season,1998,"Iliria 1998, zahlungin Albanienin demvon . 1994."Qeramika e stilit atik e den 1-2,pp.253-273. osterr.-ungar.Truppen zbuluarne Dyrrah,shek. Vl-V 1916-1918 LGP1XI = P. M. Fraserand E. Mat- besetztenGebiete (Schriften p.e.s."[Pottery of Atticstyles der thews,A Lexiconof GreekPersonal Balkankommission, foundat Dyrrachium,6th-5th Linguistische lEames:TheAegean Islands, Cyprus, Abteilung13), Vienna. centuriesB.C.], Iliria 1994, 1-2, and Cyrenaica,Oxford 1987. Shennan,S.J. 1985. pp. 151-163. Experimentsin the Miraj,L. 1990."Durres (Amfiteater e Collectionand Analysis of . 1997."Disa te dhenambi Archaeologi- Terma)"[Durres (The amphitheater cal SurveyData: TheEast topografinee nekropolitantik Hampshire andthe baths)],Iliria 1990,2, Survey,Sheffield. te Dyrrahut,shek. Vl-I" [Some pp.258-259. Sihler,A. L. 1995.lVew informationconcerning the Comparative . 1994."Termate Dyrrahut" Grammarof topographyof the Latin and Greek, ancientnecro- [Thebaths of Durres],Iliria 1994, Oxford. polisof Dyrrachium,6th-lst 1-2, pp.207-224. Snodgrass,A. M. 1987.An centuries],Iliria 1997, Archaeology 1-2, Miraj,F., andH. Myrto.1982. "Ujes- of Greece:The pp. 121-126. PresentState and jellesii Dyrrahut"[The aqueduct FutureScope of a Discipline,Berkeley. Hodges,R., G. SaraSi,W. Bowden, of Dyrrachium],Iliria 1982,1, Stephenson,P.2000. Byzantium's Bal- P.Chiles, O. Gilkes,K. Lako, pp. 131-156. kan Frontiers:APolitical Study A. Lane,S. Martin,J. Mitchell, of Myrto,H. 1989."Nje sanktuar antik ne thelMorthern Balkans, Cambridge. J. Moreland,S. O'Hara,M. Plu- periferite Durresit"[An ancient Sufflay,M. von. 1924.Stadte und Bur- ciennik,and L. Watson.1997. sanctuaryin the vicinityof Durres], genAlbaniens hauptsachlich "LateAntique and Byzantine wahrend Iliria 1989,1, pp.87-108. desMittelalters, Butrint:Interim Vienna. Reporton the . 1998.Albania archeologica: Tartari,F. 1987. Portand Its "Njegrup varresh Hinterland,"JRA10, Bibliografiasistematica dei centri te shek.II-IV pp.207-234. te e. sonene zonen antichi,Bari. kodrinorete nekropolitte DURRES REGIONAL ARCHAEOLOGICAL PROJECT II9

Dyrrahut"[A groupof gravesof the perdorurane Dyrrahgate shekujve Winter,N. A. 1993.GreekArchitec- 2nd-4thcentury A.C. in the hilly III-I p. e. sone"[Ancient sealings turalTerracottasfrom Je Prehistoric areaof the necropolisof Dyrra- on buildingmaterials used at Dyr- to theEnd of theArchaic Period, chium],Iliria 1987,1, pp. 153-166. rachiumduring the 3rd-lst cen- Oxford. . 1988."Godine e lashtebanimi turiesB.C.], Iliria 1992,1-2, Wright,J.C.,J. F. Cherry,J.L. Davis, ne sheshine parkut'Rinia,' Durres" pp. 149-175. E. Mantzourani,and S. B. Sutton. [A latedomestic structure in the Thumb,A. 1932.Handbuch dergrie- 1990."The Nemea Valley Archaeo- squareof the park"Rinia" at Dur- chischenDialekte I, Heidelberg. logicalProject: A Preliminary res],Iliria 1988,1, pp.91-117. Tojci, V. 1971."Amfiteatri i Durresit" Report,"Hesperia 59, pp.579-659. . 1991."Rite dhe formavarresh [Theamphitheater of Durres], Zeqo,M. 1986.a-Deshmite artitte ngavarreza e Dyrrahute shek.I-V Monumentet2, pp.37-41. lashtene Durres"[Evidence for te e. sone"[Rituals and types of . 1976."La population illyri- ancientart at Durres],Iliria 1986, gravesfrom the cemeteryof Dyrra- ennede Dyrrhachiona la lu- 1, pp. 179-185. chiumof the lst-5th centuriesA.C.], mierede s don nees hi s tori que s et . 1989."Aspekte te artitte Iliria 1991,1-2, pp. 187-201. archeologiques,"Iliria 4, pp.301- lashtene Durres"[Aspects of an- Tartari,F., and H. Hidri.1992."Vula 306. cientart in Durres],Ilirza 1989,2, antikembi materialit e ndertimitte Wilkes,J.1992. TheIllyrians, Oxford. pp.87-102.

JackL. Davis AaronD. Wolpert UNIVERSITYOF CINCINNATI UNIVERSITYOF CINCINNATI DEPARTMENTOF CLASSICS DEPARTMENTOF CEASSICS CINCINNATI,OHIO 4522I-0226 CINCINNATI,OHIO 4522I-0226 j [email protected] [email protected]

Afrim Hoti PhoebeE. Acheson IN STITUTIARKEOLOGJI K UNIVERSITYOF CINCINNATI DEPARTAMENTIDURRES DEPARTMENTOF CLASSICS SHETITORJA TAULANTIA CINCINNATI,OHIO 4522I-0226 DURRES,. . ALBANIA [email protected] [email protected] John W Hayes Iris Pojani INSTITUTEOF ARCHAEOLOGY QENDRANDERKOMBETARE PER 36 BEAUMONTSTREET ARKEOLOGJINESHQIPTARE OXFORDOXI 2PG RRUGALEK DUKAGJINI 9 ENGLAND TI RANE,.. ALBANIA ipoj ani@albaniaonline . net

SharonR. Stocker UNIVERSITYOF CINCINNATI DE PA RT M E N T OF CLA S S I CS CINCINNATI,OHIO 4522I-0226 [email protected]