Excavations at Stobi, 1971 Author(s): James Wiseman and Djordje Mano-Zissi Source: American Journal of Archaeology, Vol. 76, No. 4 (Oct., 1972), pp. 407-424 Published by: Archaeological Institute of America Stable URL: http://www.jstor.org/stable/502874 Accessed: 02-06-2015 16:12 UTC

REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/502874?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Archaeology. http://www.jstor.org

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions Excavations at Stobi, 1971

JAMES WISEMAN AND DJORDJE MANO-ZISSI

PLATES 85-90

I. INTRODUCTION The plan presentedhere includes most of the build- discoveredin excavationsas well as The second campaign of the new excavations at ings previous those in the new areas,but in both cases in Stobi in by the University of Texas at only Austin and the National Museum of Titov Veles the west and centralparts of the city.4 The basisof the is a 2 m. contributed to the of the history survey grid generated significantly study from a datum in the CentralFountain of the ancient Moreover, excavation of the point (ill. city.' a brass set in concreteis indicated a of the EpiscopalBasilica (infra, Sections 1:17; pin by Baptistery crossed The was A. G. indicates a somewhat unexpectedgrandeur circle). grid developedby 8-9) in and he establisheda True at Stobi in the Christian it also raises Grulich 197o North early period; at about in the churchesof Mace- line, as well as a perpendicular the datumpoint. questions liturgy The areasare enclosed lines donia. grid 2 m. squares by to those the cross. Each The is the Smithsonian In- parallel forming original project sponsored by areais an stitution most of the funds were grid designatedby abbreviation,N, E, w, and, again, pro- the directionfrom the datum vided its s, indicating point, through Foreign Currency Program.2 which is the numberof the Excavation in and ended in accompaniedby grid I97I1 began 24 May area from the datum most areas on Work in the east of counting sequentially point. 31 July. parodos The areas the datum the continued until 6 grid surrounding point, then, Theater, however, August are To cite one and the Baptisterywas not completelycleared until NI/Wi, Ni/Ei, SI/EI, SI/WI. more example,the entranceto the lower Via Axia 13 August. (ill. i:I9) E of the CentralFountain lies in grid 2. THE SITE area NI/E2. The designationsthus give an im- in Mappingof the ancientcity and its environshas mediateand precisearea locationboth distance progressed enough to publish a portion of it (ill. I). and directionfrom the CentralFountain.

1 A preliminary report on the 1970 season appeared in AJA moor a leave of absence to join us at Stobi. Photographers: cited hereafter as W-MZ 1970. Richard Trimble, University of Texas at Austin, who also 75 (971I) 395-41I, 2 We are grateful to the Smithsonian Institution for its con- measured and drew the late Roman complex in the theater tinued support of the project and to the Ford Foundation for (infra, Section 6); Mrs. Marilyn Huffman, Denver, Colorado. once again providing funds for the American student members They were assisted in the dark room by Mrs. Sharon Parkey, of the staff. Numerous items of equipment were made availa- who also worked as draftsman. Mr. Apostol Keramidriev, ble for the excavations by the University of Texas at Austin, Archaeological Museum at Skopje, made transcriptions of the the National Museums at Titov Veles and Prilep, and the inscriptions on the theater seats. Dr. Robert Folk, Professor of Archaeological Museum at Skopje; our sincere thanks go to Geology at the University of Texas at Austin, joined us briefly all these institutions and their administrative officials. The in August. Mr. 2ika Radolevi', a student at the University of directors of the museums are Mr. Todor Gruev, Dr. Bo'ko Belgrade, was archivist and interpreter. Babi6 and Dr. Blaga Aleksova, respectively. The following graduate students served as field supervisors: 3 The authors of this report continued to serve as Co- Miss Harriett Blitzer, Indiana University; Mr. John Cherry, Directors along with the Administrative Director, Mr. Todor Mrs. Virginia McKeen and Miss Carolyn Snively, University Gruev. Other staff members in 1971 were: Senior Field Super- of Texas at Austin; Mrs. Susan Schaffner, Bryn Mawr College; visors: Dr. E. Mott Davis, University of Texas at Austin; Dr. Miss Nada Proeva, University of Belgrade; Miss Mariya Zoyeva, Elizabeth Gebhard, University of Illinois at Chicago Circle; University of Skopje. Museum technicians: Mrs. Nina Dimleva, Mr. Blagoye Kitanovski, National Museum at Prilep; Dr. Ivan Archaeological Museum of Skopje; Mr. Djordje Georgievski, Conserva- Conservation Institute of Mrs. Radmilla Mikuc'id, University of Skopje; Mr. Sar'o Sardovski, Macedonia; Ivani'evid, tion Institute of Macedonia; Mr. Al B. Wesolowsky, University National Museum of Prilep. The foreman was Mr. Kiro Krstev- of Texas at Austin. Museum staff: Miss Geraldine Gilligan, ski of the Archaeological Museum at Skopje. Mrs. Lucy Wise- Tufts University; Mrs. Viktoria Sokolovska, Archaeological man was especially helpful in all phases of the preparation of Museum at Skopje; Mrs. Lucy Wiseman, Austin. Architects: this manuscript. Paul Huffman, Denver, Colorado; James Parkey, University of 4The chief omissions are: the line of the outer and inner Texas at Austin. William B. Dinsmoor, Jr., Architect of the city walls (unexcavated), the "Prison" area between the Epis- Athenian Agora Excavations, also served with us for two copal Residence and the House of Parthenius; a portion of the weeks in August. We are indebted to Dr. T. Leslie Shear, Episcopal Residence; houses w of the Civil and North . Jr., Director of the Agora Excavations, for granting Mr. Dins-

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 408 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76

I WEST CEMETERY 2 PORTAHERACLEA 3 VIA SACRA 4 SEMICIRCULAR COURT 5 EPISCOPAL 6 BAPTISTERY 7 VIA PRINCIPALIS SUPERIOR 8 EPISCOPAL RESIDENCE 9 HOUSE OF THE FULLER 15 10 THEODOSIAN . PALACE II HOUSE OF PARTHENIUS 23 12 CASINO 13 THEATER 21 14 VIA THEODOSIA 7 15 VIA PRINCIPALIS INFERIOR 16 HOUSE OF PERISTERIAS 4 222 17 CENTRAL FOUNTAIN 20 18 LARGE BATH 14. 1 19 VIA AXIA 20 HOUSE OF PSALMS - 21 SYNAGOGUE BASILICA 10 16 22 SMALL BATH 23 CIVILBASILICA 24 NORTH BASILICA 74 15

3

-12 ) • -S

STOBI

0 50 100 ISO

ILL. I

In 1971 Paul Huffman set a series of steel pins in be abandoned. Among the latter are Polycharmus' concrete at 50 m. grid intersections throughout Palace and the Summer Palace for the House of much of the area enclosed by the city wall. The Psalms;5 the Winter Palace for the Large Bath; buildings were then related to those points, and the Basilica of the Quatrefoil Baptistery for the their plans photographically reduced to a scale of North Basilica; and various directional names for i:5oo and placed on the site plan. the Porta Heraclea. A number of new names have been applied to THE various streets and monuments and appear on ill. I 3. SYNAGOGUE-BASILICA for ease of reference. Other names have long been A trench 4.9 x 2.2 m. was dug in the SE corner in use, while still others have been cited in pub- of the narthex of the Synagogue-Basilicato test the lications under a variety of titles that should now pre-basilican stratigraphy and to determine the ex- 5 "Summer Palace" is misleading and we do not know for was tied to structures successively Jewish and Christian, a sure that it ever formed a part of the residence of Poly- name suitable for both religions seemed needed; thus the charmus, the Father of the Synagogue. Since the building House of Psalms.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONSAT STOBI,1971 409 tentof the Synagogue.6The remarkablythick (max. was plentiful between the two walls and in the th. 0.4 m.) concretebedding for the floor of the narrow space N of Wall 7 and included, among narthexwas preservedover most of the area.Pot- the earliertypes, a cache of nine fusiformunguen- tery and coins found below the floor date to the taria and severalpieces of Megarianbowls.'" late 4th century A.D.,supporting the construction i. Fusiform (pl. 86, fig. 2) C-71- date suggested by the context material reported 88. H. 0.115m. D. of mouth 0.028m. D. of body last yearbelow the floorof the nave.A hard-packed 0.035 m. Intact.Buff clay, self-slip.Fusiform body, earthenfloor was encounteredat 137-77/137.82m.;7 cylindricalneck with flaringrim, flat on top; foot it belongedto a room the Nentrance of which, and spreadingto circularbase. Differs little from the the room itself, had been cut throughby the foun- eight othersin cache.Slender profile typical of late dation for the nave-narthexwall. The room be- 2nd to earlyIst centuryB.c.; cf. PhilippeBruneau, longs to the period of Synagogue2, as identified "Tombesd'Argos," BCH 94 (1970) 437-531,figs. last year,8from which a large section of mosaic 210-211, esp. the unguentariafrom tombs 185 and found below the nave of the basilica had been 188. removed by earlier excavators.A small piece of 2. Megarianbowl (pl. 86, fig. 3). C-7I-83.P.H. mosaic with a chessboardpattern was found in 0.062m. Incomplete;five piecesmended. Clay light placejust N of the thresholdof the room below the red (Hue 2.5YR 6/6)," slightly micaceous,fired narthexin anothersmall trench dug in 1971;this hard.Dull red paint coversvessel, fired brownish- fragment,and anothersmall mosaic sectionbelow red on upper exterior.Hemispherical bowl with the entranceto the atriumcourt, probablybelong everted,rounded rim; two groovedlines terminate to the complexof Synagogue2. The earthenfloor a plain zone below rim. Body coveredwith rows just cited containedmaterial of the early 4th cen- of small ivy leaves. tury A.D. The rooms may form part of a potter'sshop. Immediatelybelow the floorappeared two paral- A numberof artifactsfound in the roomsare clear lel brick walls, oriented ca. E-w and ca. 1.25 m. indicatorsof a kiln in the vicinity: several frag- apart.The s wall (Wall 6) had been cut through ments of Megarianbowl molds, molds for two by the nave-narthexwall but the second (Wall 7) terracottalions' heads of a size suitablefor the feet terminatedon the E in a roundedface (pl. 85, fig. of molded bowls; small ceramic kiln separators I); the w extremitiesof both are still hidden in the and, from a depositof the same periodbelow the scarpof the trench.The preservedtop of Wall 6 nave, several cube-shapedterracotta pieces that is nearly level with the sterile gravel of the pre- probablyserved as dividersfor large vesselswhen historicbed of the river Crna which is flush with they were being fired.'2 but the N face of Wall 6 bowl mold the s face of the wall;" 3. Megarian (pl. 85, fig. 4) C-71-73- and both faces of its companionare coatedwith a H. 0.054m. D. 0.127 m. Incomplete;three pieces coarselime plasterdown to a plasterfloor at ca. mended; part of rim and base missing.Fine pink 136.69m., i.e., ca. o.85 m. below the level of the clay (Hue 7.5YR 7/4); self-slip. Hemispherical riverbed. The roomsclearly were sunk below their mold with raised flat base, flanged rim. Exterior contemporaryground level and may have beenpart plain. Decorative pattern incised on interior is of the basementfor a building the superstructure terminatedbelow the rim by a grooved line; pat- of which was removedin later leveling and build- tern of long slenderleaves, larger and roundedat ing operations. the top, rising from a plain floor and alternating The building was destroyedin a fire in the with verticalrows of puncheddots. A similarpat- late 2nd or the early Ist century B.c. Pottery tern is found on a Megarianbowl of the late 2nd

6 See W-MZ 1970, ill. 6. 9 We are indebtedto ProfessorRobert Folk for this and 7 All levels given are the elevations above sea level; that of numerousother geologicalobservations at Stobi. the later narthex floor was 139.371 m. The latest relevant 10Only one exampleof eachis publishedhere. coin, which is from destruction debris below the deposit just 11Color designationsbeginning with "Hue" refer to hue, described, is no. 71-115, a SALVS REI PUBLICAE issue of value and chromaas illustratedin MunsellSoil ColorCharts Valentinian II (A.D. 383-392) minted in Thessalonica. Two (Munsell Color Company,Inc., Baltimore,1971 ed.). other coins from the same deposit (7I-II4, 7I-151) must 12The latter are not illustratedhere; inventoriedexamples also belong to the period between the accession of Valentinian are MF-71-17o(P.L. 0.076 m., H. 0.023 m., Th. 0.044 m.) II and the death of Theodosius I (A.D. 395). and MF-7I-17I (L. 0.08 m., H. 0.024 m., Th. 0.02 m.). 8 W-MZ 1970, Section 8.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 410 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 centuryB.c. from Kertschnow in Mainz: Theodor two terracottavessels (pl. 85, fig. 7) containinga Kraus, Megarische Becher im Romischen-Ger- hoard of ca. 500 silver denarii of the 2nd cen- manischenZentralmuseum zu Mainz (Mainz 1951) tury B.C. 9, no. 9. The largerof the two vessels(infra, no. 5) was a 4. Terracottakiln divider (pl. 86, fig. 5) MF-7i- coarse,one-handled globular jug standingupright 104. L. 0.021m. Intact.Fine pink clay (Hue 7.5YR in the pit. The smaller(infra, no. 6) was a brown 7/4). Small, nearly circular piece with shallow, glazed lekythosfound lying on its side near the flat groove acrossdiameter. Other examplesfrom top of the pit at 137-37m. Loose earth had been the same deposit,MF-71-Io5 to 107, vary in diam- packed aroundthe vessels.The smallercontained eter from o.oi8 to 0.02 m. 67 denarii,the largera partlycongealed mass that Depositsrepresenting the periodbetween the Ist yieldedover 421 denarii and one Attic tetradrachma. centuryB.C. and the early 4th centuryA.D. in this Only a small portion of the hoard has as yet areamust have been removedat least by the time been cleaned,but most of the coins are in excellent of the constructionof Synagogue2. A new trench, condition.Most denarii show on the obversethe both to L 3.3 x 3.8 m., opened below the w end of the nave head of Roma, with winged helmet, in 1971revealed that the buildershad not needed and R. This seriesseems to begin ca. 211 B.c. and to cut quite so deeplyinto the underlyingdeposits a preliminaryexamination suggests that no coin in here.The floorlevel of SynagogueI, at ca. 137-5m., the hoardwas issued after the third quarterof the lay below a destructiondeposit of the late 3rd 2nd century.15There is a great varietyof reverse centurythat included additionalfragments of the types, four of which are illustratedhere (pl. 86, fresco found last year with the repeatedlegend fig. 8).16 The tetradrachma(pl. 86, fig. 9) belongs 6 Trarp A thresholdblock to the middle of the same century." HIokXiapptoo EvX•Yy.v. lying immediatelyabove Wall 414 and two paving 5. Jug (pl. 85, fig. 10) C-7I-I3I.H. 0.125 m. D. slabs belong to this structure,Synagogue i, the of mouth 0.085m. D. of body 0.112m. Complete building of Polycharmus(pl. 85, fig. 6). except for piece of rim and small body fragment. Polycharmus anticipated later builders and in Clay red, very coarse and micaceous; self-slip; leveling part of the area for the Synagogue had cut lower part of vessel blackened from burning. down to a gravelly deposit, 2nd-Ist century B.C. Globular with flaring lip and rounded bottom, An irregular,oval pit (Pit I) intruded into this originally used for cooking. Small vertical strap deposit ca. I m. s of the basilica N stylobate, just w handle from lip to body. Narrowly spacedwheel- of Column N-I. The maximum dimension N-s of ridging coversbottom and lower part of body. Pit I was ca. 0.7 m. and was separatedfrom a small, 6. Lekythos(pl. 85, fig. II) C-7I-132. H. 0.092 m. circular pit, ca. 0.2 m. in diameter,on the sw by D. of mouth 0.051m. D. of body 0.064m. D. of a low ridge. In the latter on 16 July were found base 0.046 m. Complete.Fine pink clay (Hue 5YR

13 W-MZ 1970, Section 8. head of Roma to 1, below chin X, to r TAMPIL(VS); rev.: 14 This is the w portion of the E-w wall found in 1970 and in quadriga r, M BAEBI(VS) Q.F., in field ROMA described in ibid., 408, and ill. 6. (Sydenham [supra n. 15] coin 489; Crawford [supra n. 151 15 See Edward A. Sydenham, The Coinage of the Roman Table x where M. Baebius is listed ca. midway in a list of Republic (rev. ed., London 1952) 14-32, and Michael H. moneyers, 150-125 B.C.). No. 919 obv.: head of Roma r, Crawford, Roman Republican Coin Hoards (London 1969) Nike standing behind holding a wreath; below chin, X; rev.: Table iii. Dioscuri on horses to r with a star in field above each hat; 16L. to r, top row: nos. 7I-II65, 71-1184; bottom row: C. TER(ENTIVS) LVC(ANVS); in ex. ROMA (Sydenham * 71-1197, 71-919. No. 1165 obv.: head of Roma to r, to 1; [supra n. 15] coin 425, which he dates to 135-126 B.c.; Craw- Macedonian hel- rev.: horseman in armor with spear; behind, ford [supra n. I5] Table in where he is placed near the end in ROMA. met; beneath, Q. (MARCIVS) PILIPVS; ex., (The of a list of moneyers, 165-150 B.c.). issue of Pilipus was the latest in the Agrinion Hoard, ca. 17 No. 71-I208. Obv.: helmeted head of Athena r; helmet 130 B.C. See Rudi Thomsen and Michael H. Crawford, "Ap- has embossed floral patterns; row of globules forms border. pendix: The Denarii of the Agrinion Hoard," in Margaret Rev.: owl in center perched on amphora; prow: AGE KAPAIX Thompson, The Agrinion Hoard, NNM 159 [New York EPFOKAE AlOME; on amphora M; control letters illeg. 1968] 127 and n. 32; see also Sydenham [supra n. I5] coin Margaret Thompson, The New Style Coinage of Athens, 477 and Crawford [supra n. I5] Table x.) No. 1184 obv.: ANSNS io (New York I96I) 200-205, 312; date of coin in r be- head of Roma to r; rev.: Nike biga to holding whip; is the archon year 153-152 B.c. Thompson coin 540 D has neath L. SAVF(EIVS); in ex. ROMA (Sydenham [supra n. the same obv., but the illus. of the rev. on pl. 54 has as i5] coin 384; Crawford [supra n. i5] lists Saufeius sixth in the third name DIONY; the control letters are ME. a list of 20 moneyers, I65-I50 B.C., Table iii). No. 1197 obv.:

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 411 7/5) with small flecks of mica. Reddish-brown as "theancient city""22 of the Paeoniansand bronzes paint with bright metallic luster. Small, graceful from earlier excavationshave been dated to the vessel, ovoid body on flaringbase, deeply concave early 5th centuryB.c.23 Ceramicmaterial of even on bottom. Narrow neck and flaring rim; small the Halstattphase is reportedto have been found double-rolledhandle from lip to upper part of in a test trenchbelow the apseof the Civil Basilica body; two wheel grooves below handle. (ill. 1:23)."2It was, therefore,surprising when de- If we may associatethe earlystructure below the positsof the late 2nd to Ist centuryB.c. below the narthexwith the coin hoard,and presumablyWall nave and narthexof the Synagogue-Basilicawere 4, we might conjecturea building complex of found to lie on the prehistoricbed of the Crna house, shop and kiln to the w with an enclosed River.Of course,the buildingdiscussed in Section courtyardto the E. The court would have been 3 may have been constructedin the early2nd cen- boundedby Wall 4 on the s, by a wall indicated tury B.c.But to find the earliercity, clearlyit was by a pillaging trench on the E, and perhapsby necessaryto test areasother than the Synagogue- Wall 2 on the N.1 The presenceof such a large Basilica. hoard near rooms destroyedin a fire points to a A large test trench,20 x 3 m., on the Acropolis suddencatastrophe and, at leastlocally, a total one. (ill. I, unnumberedrectangle top center)25pro- Furtherevidence both for the hypothesisthat the duced similar results regarding the earliest city. hoard was in a courtyardand for the catastrophe The rubblewalls of a numberof small houses or is the discoveryof the skeletonsof two young ani- shops were found not far below the present sur- mals of the genus Equus in a depressionor pit face level. Habitationhere came to an end in a Nw of the hoard.Several large stones, perhaps from destructionthat can be datedto someperiod shortly a collapsingwall, were found below and on top afterA.D. 569/570, the date of the latest coin in the of the skeletons,which, although they indicated destructiondebris of one of the rooms.26A large some disturbancein antiquity,were partiallyarticu- bronze cross was found on the thresholdof an lated and, therefore,must have been placedin the adjacentroom, with the wire hangerthat had once depressionwhile some flesh still remained.A red attachedit to the door beside it. clay and earth depositat least partiallywithin the 7. Bronze cross (pl. 86, fig. 12) MF-71-129a-b. same depressioncontained part of a left human H. o.162m. L. of crossarm o.I27 m. Th. o.ooI m. tibia."9 Nearly complete;only small chips missing. Cross The animals, of course, may have died from formed of two flat strips of bronze united by a naturalor unnaturalcauses and their corpsesin- singlebronze rivet, hammered flat on both surfaces terred,but the othersure evidence of disaster,along and solderedin back.Each arm of crosshas flaring with the presenceof the human tibia and the gen- tips.There is a smallhole for a hookedwire (MF- eral contemporaneityof all deposits,suggest asso- 71-I29b). ciation.2oConfirmation must await further study These houses or shops, made of field stones set of the coins and the context pottery. in earthmortar, show at least two buildingperiods and overliesimilar small buildings, probably of the THE CITY 4. EARLIEST 4th century.Ample pottery,but few structuralele- Stobi certainlyexisted as a city in the early 2nd ments (a single rubblewall, a nearbydrain) attest centuryB.c. when Philip V was involved in mili- more or less continuouscultural activity on the taryoperations in the vicinity.21Livy refersto Stobi Acropolisfrom the 2nd century B.c. One of the

18 In a later Wall 2 became the N wall of period Synagogue in II9/II8 B.c.: SIG3, 700 and 710. 2; W-MZ 1970, Section 8. 21 33-19.3; 40.21.1; 45.29.13. 19 This deposit was disturbed by the constructionof the nave- 22 "Vetere urbe," 39.53.I6. narthex wall and other parts of the human skeleton could have 23 For references see W-MZ 1970, n. 3. been removed at that time. 24 Some of the documentation and most of the material of 20 The deposits of the depressions and pits referred to here, those excavations, conducted in 1955 by D. Kotso, were de- including Pit I and the context associated with the animal stroyed in the Skopje earthquake of 1963. skeletons, contained some material that has been dated provi- 25 Grid areas W39-4o/N53-62. sionally to the late Ist century B.C. or the early Ist century 26 71-135, found on the threshold of Room 2. Obv.: Justin A.D. This may be the result of a later disturbance. We might II and Sophia;rev.: K and, in smallerletters, ANNO vertically note that Marcus Minucius, son of Quintus, defeated an army to 1, V to r. of Gauls and Thracians near Macedonian Argos (near Stobi)

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 412 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 earliest deposits excavated included a coin of Philip halted some 30 cm. below water level without V.27 But below the late Hellenistic deposits the reaching either the bottom of the wall or a floor partially calcified gravel of the early river bed was level. The other test revealed three walls faced on encountered at 147.07-147.27 m. For comparison, the one side with molded stucco. Digging was aban- late Roman wall in the test doned in the room at a level of m. highest preserved I31.93 trench was at 151.41 m. The trenches were narrow, considering the depth Investigation on a smaller scale below the nave of the cultural remains, and must be enlarged be- of the Civil Basilica (ill. 1i:23) also failed to reveal fore the structures can be examined in detail. But a pre-Hellenistic deposit. Efforts here were concen- the presence of over 4 m. of sand above the ruins trated in a single room which had been partially strongly suggests that the lower city near the river, excavated in 1955, when a figured fresco from one even though it lay inside the city wall, was aban- of its walls had been removed to the Archaeologi- doned to the floods of the Crna at some time dur- cal Museum at Skopje.28Below the earthen floor of ing the early empire. Further investigation is the room considerable ceramic material of the 2nd planned for 1972. to early Ist century B.C. was found in association The Crna has frequently overflowed even in with a roughly circular clay pit. One complete and the 20othcentury and spread over the lower areas one nearly complete terracotta lamp with black of the ancient site, according to local reports. In- glaze were among the datable artifacts (L-71-24 deed, it was brought under control only in 1969 and 27). Below the clay pit was sterile clay and when a dam was constructed near Vozartsi, some gravel (elev. 135-95m.). There was no time at the 20 km. upriver from Stobi. Clear evidence of the end of the season to test below the apse where the rapidity with which the Crna can create deep sand Iron Age material had reportedly been found in deposits was encountered in one of the major test 1955. areas dug in 1971, begun as a test across the The earliest city, then, if it lay at all within the presumed line of the road leading to a bridge (two later confines of Stobi, was presumably located piers in the water and the road on the right bank closer to the present channel of the Crna or fur- of the Crna were visible before excavation; pl. 85, ther N along the . Excavations in three test fig. 13). The E face of the ramp on the left bank, areas have shown that the city during early Roman which was completely covered by sandy river silts times, at least, extended nearly to the left bank of when excavation began, was cleared to a depth of the Crna where short sections of the city wall have 3.1 m. where several fallen blocks indicated partial always been visible. Two of these trenches were destruction of the abutment. At the same level dug in August-September 197o by Mr. Zivojin were found several rifle cartridges dating from the Vincid of the Conservation Institute of Macedonia First World War.30 at the request of the Project directors;29they are The bridge was already in ruins when Leon not shown in ill. i, but are located ca. 200 m. E of Heuzey visited the site in 1872, but may have been the apse of the Synagogue-Basilica near the inner in use not long before that time.31 The cobble- face of the city wall. River sand was encountered stone pavement of the ramp on the left bank is immediately below the shallow surface earth (elev. characteristic of the Turkish kalderma found fre- 136.38/136.44m.). The sand continued in a series quently in the s Balkans; cross-ties of wooden of alluvial deposits to 132.31 m. in one trench; at beams are still partially preserved in the pavement this point the top of a wall appeared,its face coated (pl. 85, fig. 14). Not only the pavement but the with painted stucco. Deposits containing early bridge itself may have been constructed during the Roman pottery, Ist-3rd century A.D.,continued be- period of Turkish domination. side the wall to 129.13 m., where digging was The ramp is built over the outer wall of the 27 71-529. 30 The military front line passed through Stobi after the 28 The fresco was destroyed in the 1963 earthquake but a collapse of the front at Thessalonica; it was held by the drawing of a water bird on a red backgroundis extant. A report Serbians and French until taken by the Germans and Bul- on the earlier excavations of the Civil Basilica is being garians. prepared for publication elsewhere. 31 Leon des ruines de Stobi," RA 29 Heuzey, "D&couverte One of the tests was begun during the regular digging (1873) 2, 34. He makes it clear that the site was unoccupied season in 1970. The area of the trenches had been considered at the time of his visit. at that time as a possible site for a museum and work building.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 413 ancient city (pl. 85, fig. i5). The wall here is 2.3 m. ble walls. In addition to a small jug (C-7I-io) wide and preserved on the exterior to a height of burial gifts included five coins of the early 4th A.D. to 3.75 m.32 A later wall, only 0.63 m. thick, bedded century (coin nos. 71-66 70)."3 on earth and built up against the inner face of the Grave 8, another tile grave of about the same city wall probably in the 5th century, was also built date,is of specialinterest on two counts.The skele- over by the bridge road. A test against the inner ton, which had its head to the E,was extendedand face of the city wall at this point revealed a floor of constricted,apparently the result of burial in a large, rectangular stone plaques at a level of 130.04 tightly-wrappedshroud (pl. 87, fig. i6). A bronze m. The deposit immediately above the floor dates statuetteof Mercuryhad been placed at the feet to the 3rd century A.D. of the corpse;it is the latestevidence of cult in the Another test trench, 12 x 1.5 m., was dug across cemeteryother than that providedby the fact and the line of the Via Axia not far E of the Synagogue- manner of the burial itself.3" Basilica (see ill. i). This test lies w of the meander- 8. Bronze statuetteof Mercury(pl. 86, fig. 17) ing ridge that may mark the line of the inner city MF-71-22.H. 0.048m. D. of base o.o09 m. Intact. wall, perhaps constructed only after the lower city Solid-castbronze figure of Mercury,wearing only a and its wall had been abandoned to the flood- petasosand a shortchlamys over his left shoulder. waters of the river.33Here late Roman construc- Highly stylizedfigure on shortpedestal; two dots tions overlay walls of the Ist century A.D.and no for eyes and a shortslash for the mouth; threein- road metal was discovered. cised lines on right arm and some cross-hatching The earlier Hellenistic and pre-Hellenistic city, on chest. Winglike projectionover the chlamys, then, remains elusive. It is hoped that more exten- draperyfolds indicatedby a few incisions.He is sive and deeper soundings below the sand deposits clutching somethingin his right hand, doubtless along the Crna in 1972 will provide some clues to the traditionalmoney bag. His legs are slightly the location of earliest Stobi. bowed.There are numeroussimilar figures of Mer- cury (unpublished)in the museum at Prilep and THE CEMETERY 5. WEST one was found at Stobi in a grave excavatedin The area of excavation in the West Cemetery, 1966 (MF-66-I). begun in 1970,4 was enlarged on the N and E to a During the 3rd centurythe level of the cemetery rectangle, 9 x II.7 m., in grid area S81-87/WI34-I38 had been raised,apparently by intentionalfilling, (ill. i:i). A total of 73 graves was found (nos. and few graves can be assignedto the 3rd or the 22-86, 88-96), of which 63 were excavated. latterpart of the 2nd century.An interestinggroup The part of the cemetery investigated clearly of six cremationburials was found in the s part went out of use in the 4th century A.D.since the re- of the excavationarea below this fill.37These graves mains of several rubble walls, probably built before were coveredby a compactdeposit of gravel and the middle of that century, were found in all parts greenish earth, 0.2-0.3 m. thick, with a curved of the trench area immediately above the level of perimeterwithin the trenchwhich may originally the latest graves. Indeed, a few of the interments have been a circle.If so, aboutthree-quarters of the had been disturbed by the wall construction. circle lies within the unexcavatedarea to the s The latest burials were inhumations in shallow and E. pits. Laconian roof tiles had frequently been placed A feature common to four of the six burials in tent fashion over the body; orientation of pit covered by the gravel was a spout, formed by and body was variable. Grave 25 was such a tile placing two Laconianroof tiles edge-on,projecting grave and lay immediately beneath one of the rub- abovethe grave structure.38The spoutspresumably

32 The bottom of the wall was not reached because it lies finger ring (MF-71-21) and a small bronze decorative attach- below the present water table. ment (MF-71-23). The ring is a signet bearing a running 33 The presumed inner city wall has not yet been tested by lion with a flowing mane and tail raised high; there is a star excavation. in the field above. S4W-MZ 1970, Section 7. 37Graves 38-40, 44, 47, 48. 35The floor of Grave 25 was at 149.I06 m. but had been 38 Three other graves with such a spout were found outside dug from level 149.656 m. the perimeter of the gravel deposit. The graves were at ap- 36The burial could have been as early as the late 3rd proximately the same elevation, but one, no. 46, was several century. Other items found with the skeleton were a bronze meters N of other similar burials and the other two, Graves

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 414 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 shoulder. were for offerings, probably liquid, to the spirits of club in right hand and lionskin over left the dead-at the time of burial, or later, or both. Rim: plain, sloping; two small dots near U-shaped in This group of burials may appropriatelybe termed nozzle. Pierced handle, grooved in front, plain the Libation Spout Burials. All were tile graves back. Base: incised signature AVF PAR within constructed of Corinthian pan tiles and Laconian circular base ring. The type is similar to several cover tiles variously arranged. One burial, Grave lamps with U-shaped nozzles in Judith Perlzweig, Roman 38, was covered also, except for the spout, by a rec- The Athenian Agora VII: Lamps of the tangular stone cairn (pl. 87, fig. 18), 2.2 x 0.7 m. Period (Princeton I96I) 87-88, where a date of the and set within a pit 1.05 m. wide and 0.5 m. deep. late Ist to 2nd century is suggested. The roof tiles beneath the stones, however, were Grave 39 is of further interest not only because placed only over a narrower pit, 0.3 m. wide, 0.15 a considerable amount of carbon was preserved, as m. deep, cut into the center of the floor of the larger is frequent in cremation burials, but one partially pit. Such a double-pit arrangement was frequent consumed faggot was found lying partly in and among cremation burials of the West Cemetery. partly out of the lower pit. This clear indication No vessels or other burial gifts were found beneath that the burial ceremony included a fire at the the tiles,"3but two large bowls, both broken, were grave site itself was paralleled in other graves. found at the s edge of the cairn within the grave The function of the funeral fire is debatable. Al pit.o4 Wesolowsky, who supervised excavations in the Grave 39, also one of the Libation Spout Burials, West Cemetery in 1970 and 1971,42is of the opinion was set within a much smaller double pit, 0.62 x that cremation of the body invariably took place o.8 m. Laconian tiles supported a single Corinthian elsewhere since the quantity of carbon found in pan tile, as a grave cover, and the fragmentary each grave seems far too small for a fire of suffi- libation spout. The cremated skeletal remains were cient magnitude for cremation. Two hypotheses, partly within a medium-sized bowl (infra no. 9) not mutually exclusive, are here suggested. i) The and partly dispersed in the lower pit (pl. 87, fig. cremated remains may have been burned again 19). This arrangement is quite unlike that of during the ceremony at the grave; 2) the fire may Grave 38, in which the skeletal remains were have served to purify the grave site.43 A far less simply scattered throughout the extent of the lower likely explanation is that the fire served to cook pit. The grave contained, in addition to the cinerary the funeral banquet. There are ample traces of vessel, a small bowl, a terracottalamp and 36 small graveside meals in the West Cemetery during the iron nails.4' Ist century, but the chief evidence of the meal, 9. Bowl (pl. 87, fig. 20) C-71-32. H. o.I65 m. which was shared with the dead, consisted of nuts, D. of body 0.195 m. D. of mouth 0o.I27m. Intact. grapes, peaches and olives; no animal bones have Gritty, micaceous red-brown clay with self-slip. been recovered. Ovoid body on round base, concave on bottom. Flat The Libation Spout Burials all belong to the late everted rim. Two vertical strap handles. Ist or early 2nd century A.D. The gravel deposit io. Moldmade lamp (pl. 87, fig. 21) L-7I-4. L. with its distinctive green color, which was probably o0.o8 m. H. 0.039 m. W. 0.076 m. Complete ex- intended as a visible marker of the group, con- cept for small piece from base and disk. Coarse tained no recognizably later material. A badly- reddish-brown clay (Hue 5YR 6/4), slightly mica- weathered coin of Vitellius (A.D.69) was found in ceous. Worn red glaze covers body. Traces of burn- Grave 40 (coin no. 71-133) and a much better pre- ing at nozzle. Disk: Hercules standing full front, served example of the same type was found in con-

was inventoried: 55 and 65, were nearer, to the w. On the use of spouts in trated here. Only one of the nails, MF-71-59, of the graves see J. M. C. Toynbee, Death and Burial in the Roman the rest are in context storage LOT 391. The presence the West Ceme- World (London 1971) 51. nails here and in several other early burials in that a small wooden 39 Lamp fragment L-7I-I7 was found among the bones and tery suggests chest, perhaps containing may have been burned with the body. a perishable memento or offering, was burned with the corpse. 42 Mrs. Susan Schaffner shared the during part 40 C-7I-I7 is a hemispherical bowl on a small ring base responsibility and has red-brown matt paint on its interior and upper exterior; of the 1971 season. is a oval bowl with out-turned rim. 4 Toynbee (supra n. 38) 48-50o, discusses graveside fires C-7I-3I coarse, deep 41 The second bowl, C-71-38, is fragmentary and not illus- only in connection with actual cremation.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 415 tiguous Grave 44 (coin no. 71-134; pl. 86, fig. 22).44" 14. Bulbous unguentarium (pl. 87, fig. 27) G-71- Among the earliest burials excavated was Grave 35 found in Grave 74. H. 0.065 m. D. of body 79 in grid areas WI39/S85-86. The grave consisted 0.041 m. D. of neck o.oii m. Intact. Pale blue glass. of a cremation (disturbed) in a small pit adjoining Free-blown. Squat, bulbous body with flat bottom. a larger pit in which had been placed the entire Cylindrical neck; folded lip. body of an adult Equus, presumably a horse (pl. Six of the graves of the late Ist century B.c. to 87, fig. 23). The animal appears to have been de- early Ist century A.D. were double-pit cremation capitated and the entire body interred at the time burials with outer edges rimmed by red and black of the cremation burial. A few bronze rings, prob- bands.4"The bands seem to have resulted from the ably from a harness, and an illegible bronze coin fire built in the grave pit and probably remained were found with the skeleton. Pottery sherds from visible after the burial; they may even have con- the fill suggest an Augustan date. stituted a grave marker. One such burial, Grave 81, A tile grave with cremation burial (no. 8o) near contained a large number of walnuts deposited the central w scarp in grid area WI39/S84 was whole in the ashes of the funeral fire. The same roughly contemporarywith the horse burial. Burial grave contained two iron strigils and an iron arm gifts included three vessels,45one incised with the band. owner'sname. 15. Iron strigil (pl. 87, fig. 28) MF-7I-I59. L. II. Skyphos (pl. 87, fig. 24) C-71-I29. H. o.I m. 0.33 m. Ten fragments mended; complete except D. of body 0.14 m. Incomplete; mended from sev- for end of handle. Flat rectangular handle o.115 m. eral sherds found in ashy earth both outside and long in one piece with the sharply curved blade. under the tiles of Grave 80. Clay light brown (Hue Vine motif on decorative strip attached by two 7.5YR 6/4), small flecks of mica. Dull brown paint. rivets to the back of the blade. The other strigil Deep, two-handled cup on low base ring. Vertical (MF-7I-I6o) is nearly identical but has a lanceolate rim with two shallow grooves below rounded lip. pattern on the decorative attachment. The arm Graffito on body presumably names the owner, the band (MF-71-I6I) has a D. of 0.I2 m. occupant of the grave: MaKE80oLKOi:"[the cup of] Two of the earliest graves, nos. 6o and 95, were Makedonikos." cremations covered by roof tiles, but were further Numerous bulbous and piriform terracotta un- elaborated by a built rectangular perimeter at the guentaria were found frequently in association level of the top of the grave pit. The perimeter wall with the earlier graves (late Ist B.c. to early Ist of Grave 95 was built entirely of large field stones A.D.) excavated in 1971. Glass unguentaria were and a spur projects 0.5 m. to the N at the NW also numerous: three are presented here as ex- corner." Part of a similar structure with two spurs amples. to the E was found near the center of the West 12. Slender unguentarium (pl. 87, fig. 25) G-7I- Cemetery excavation area. Later burials had in- ii found in Grave 50. H. 0.13 m. D. of body 0.03 m. volved the removal of its s, w and much of its N D. of neck 0.02 m. Intact. Greenish-blue glass with wall so that the associated burial is uncertain. The milky pitting. Free-blown. Cylindrical body with most likely candidate is Grave 21, where numerous flat bottom. Long cylindrical neck with flaring rim. tomb furnishings included four figurines of the 13. Bulbous unguentarium (pl. 87, fig. 26) G- god Telesphoros.48 71-30 found in Grave 52. H. o.i m. D. of body 0.04 The richest burial in terms of funeral offerings m. D. of neck 0.02 m. Intact. Pale bluish-green found in 1971 was Grave 76, probably Augustan in glass; bulbous body with rounded bottom. date. The cremation burial of a young child in a

44 7I-134. Obv.: laureatehead of emperor 1: AYTOKRATMR coin (no. 71-398) issued in the late 2nd century B.c., fragments OYITEAAIO[---]. Rev.: circular shield with globules on of a glass unguentarium, several iron tacks and four iron border; another row of globules on edge of coin; IEBAITOI nails were also found. MAKEAONfN. See Hugo Gaebler, Die Antiken Miinzen 46Graves 81, 82, 85, and 86 (excavated); Graves 83 and Nord-griechenlands 3 (Makedonia und Paionia), pt. i (Berlin 84 (unexcavated). 19o6) 78-79, coins 244-246. The issue is merely referred to 4 There is probably a corresponding projection on the E, in Henry Cohen, Midailles Imperiales I (Leipzig 1930) 365 but that part of the wall still lies within the scarp of the fin. We are grateful to Miss Joan Fisher for her helpful com- trench. It is of interest that inhumation burials of infants, ments on this rare issue. roughly contemporary in date, were found in the immediate 45In addition to the skyphos (no. ii), they were C-7-I30o, vicinity in shallow, unmarked pits. a small jug, and C-7I-I55, a small Samian bowl. A bronze 48 Grave 21 was excavated in 1970; W-MZ 1970, Section 7.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 416 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 globular bowl had been covered originally by sides of base and edge of pedestal on which Aphro- Laconian roof tiles that were disturbed in antiquity. dite rests her left hand. Drapery loosely knotted in Below the cinerary urn and an adjacent Corinthian at mid-thigh but pulled up in back and held pan tile were found a few terracotta unguentaria right hand. Hair parted in the middle, curls hang in back. Low x and a jumbled mass of nearly 60 terracottafigurines down from chignon base, 0o.o0510.043 (pl. 87, fig. 29). Although many of the figurines m. Circular vent hole. were unbroken, a great many fragments of others 19. Seated female. Type 2 (pl. 88, fig. 33) TF- been mended. H. m. W. of chair at m. Th. were collected and not all have yet 71-11. 0.I142 top 0.05 Grave 77, a tile grave a few meters w of Grave 76, 0.064 m. Intact. Pale brown clay (Hue IoYR 6/3). in produced six."4The figurines inventoried thus far Brownish-red to orange paint preserved patches Woman from Graves 76 and 77 represent an interesting on sides and front of chair and on hair. chiton. variety of types: seated females, 5 draped types and wears closely wrapped himation over Ionic inside the left I nude type; Aphrodite Anadyomene, 4 types; Right arm raised to chest drapery, the Knidian Aphrodite, 3 types; semi-nude Aphrodite; hand rests on left thigh. Hair parted in middle, nude, male lyre player; youth and dog; standing rolled at the sides, and in the back curls hang above her draped male; standing draped female, 2 types; down from a chignon. Chair, rising at Europa and the bull; Muse with lyre. The ex- shoulders, is flat on sides but slopes outward vent hole amples below were all found in Grave 76, except bottom on back. Below roughly circular incised no. 17 from Grave 77. in back is signature: ELo. 16. Aphrodite Anadyomene. Type 3 (pl. 88, fig. 20. Standing draped youth (pl. 88, fig. 34) TF- brown 30) TF-7I-9. H. 0.173 m. Mended from several frag- 71-29. H. o0.32 m. Intact. Light clay (Hue ments; left hand, part of back of head missing. 7.5YR 6/4). White primer paint; hair painted red. over his tunic and holds Pale brown clay (Hue IoYR 6/3). Patches of Youth wears a pallium white primer paint preserved. Aphrodite, nude, tall heavy fall of drapery in his left hand; the right and slender, has both hands raised to tie a ribbon rests on an oblique fold across his chest. Head set in her hair. Hair parted in the middle. Position well forward of shoulders as if deformed; curly relaxed, right knee bent and right elbow resting hair; facial features strongly modeled. The seam on a pedestal. Base nearly square (0.057 x 0.052 was not smoothed over and can be traced around m.); front face slopes outward. Circular vent hole entire figure. No vent hole. Base 0.041 x 0.035 m. in back. 21. Youth with dog (pl. 88, fig. 35) TF-71-26. 17. Knidian Aphrodite. Type (pl. 88, fig. 31) H. 0.136 m. Intact. Pale brown clay (Hue IoYR I on hair TF-7i-36. H. 0.174 m. Intact. Light brown clay 6/3). White primer paint. Orange paint with (Hue 7.5YR 6/4). White primer paint preserved and on back above hips. Partially nude youth over most of figurine; reddish-orange paint on right hand, clenched or holding some object, rest- amphora. Aphrodite prepares to bathe. She is nude ing on head. He is drawing back from a dog that except for drapery across her right arm, and reaches leaps at his right and rests his front paws on the to the amphora standing on a small pedestal beside youth's leg. Small, irregularly rounded object at her. Left arm tight against body, hand covers his left side, suspended from his neck. Cloak hangs pubes; left knee bent forward. Hair peaked in behind, visible from the front between his legs as front and drawn back into chignon. Figure also vertical fall of drapery with serpentine hem. Oval modeled in the back from head to buttocks. Low base 0.067 x 0.036 m. No vent hole. Incised signa- rectangular base, 0.058 x 0.038 m. ture on back above base: aUo. 18. Standing Aphrodite (pl. 88, fig. 32) TF-71-52. 22. Lyre player (pl. 88, fig. 36) TF-71-22. H. H. o.I83 m. Fragmentary; mended from several 0o.17 m. Intact. Pale brown clay (Hue IoYR fragments: right hand, right shoulder and part of 6/3). Patches of white primer paint. Partially nude back missing. Pale brown clay (Hue IoYR 6/3), youth seated on rusticated boulder. Slightly pot- discolored over much of body by burning. Patches bellied with fat protruding lips, probably a young of white primer paint on body and drapery. Hori- faun. Short cloak hangs down his back. Right leg zontal band of reddish-brown paint on front and crossed over left foot which rests on boulder. On

49It is of interest that Grave 21, in which several figurines somewhat s of Graves 76 and 77. were found in 1970, lay at approximately the same level

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 417 left thigh is a lyre, supported from behind by his 6. THE THEATER left hand; right hand rests on lyre. Rectangular Excavation in the annular corridor below the base m. x m. Worn 0.058 0.046 moldmade signa- cavea51 was begun at its sw terminus where an ture on base in back: ANTYAAO(Y). access radial corridor is located. Free space between the Compacted gravel representing prehistoric the top of the corridor fill and the crown of the bed of the River was encountered wandering Crna vault varied from a few cms. to nearly a meter at m. The earliest burials in the West Ceme- 148.2 except where an occasional break in the wall had all of the late to tery, Ist century B.c. early Ist allowed earth to fill up the area completely. A had been into the river or century A.D., dug gravel linear distance of some I.5 m. was cleared to a an sterile soil with caliche above equally it.50 level ca. 2.2 m. below the vault. The upper deposits The rich of and burial fur- variety grave types here date to the 9th to IIth centuries and suggest in the West examined nishings Cemetery, only that the upper part of the Ist tier of the theater the evolution of burial cus- briefly here, illustrate was still visible when Basil II destroyed a garrison toms in Stobi over the entire early imperial period. at Stobi in A.D. 1014."2 A test trench 4 x 2.5 m. was also evidence for a number of They provide quite dug into the lower deposits of the corridor but but burial rituals. Further different, contemporary work was halted at an arbitrarylevel 4.8 m. below material from the West analysis of the Cemetery the vault without finding the ancient floor, which can be to considerable information expected yield presumably lies not much higher than the level of on the pagan population of Stobi during the first the orchestra. The lowest deposit excavated in the three and a half centuries of the Roman Empire. test trench dates to the late 4th to 5th century A.D. The Hellenistic and Early Christian cemeteries A lateRoman building complex overlay the fill must now claim most of our attention in the field. of theE parodos(ill. 2)." A terracewall was built 50 Three tests were dug into the gravel deposit. In one, the enclosed by stone and brick perimeters, were among those river bed. gravel was found to be ca. I.I m. deep and lay on a weathered burials partly cut into the gravel deposit of the early clay-bed. A summary of elevations will perhaps help the reader 51 Light and ventilation holes near the top of the vaulted to envision the stratigraphy of this part of the cemetery. The corridor are cut into the risers of the sixteenth row of seats latest graves (early 4th century A.D.) were dug at ca. 150 m.; in the cavea. gravel fill above Libation Spout Group was 149.165/149.541 m.; 52 Cedrenus CSHB 2, 709A. top of grave 77 was 148.6 m. Graves 95 and 6o, which were 5 Plan by Richard Trimble.

ANALEMMA t RMOM7ROOM 3 R 4

S10 15

ILL. 2. Late Roman buildings in theater parodos

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 418 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 in connection with a series of adjoining rooms Gray wares, probably Macedonian in origin, are 6 (i, 2, and 7) on the s. At the w end of the com- still common in the late Roman deposits of the plex, part of the massive sandstone wall that closed parodos. An especially characteristic rim is no. 23 the parodos on the s was incorporated into the late below. Another type of ware, of far greater im- rubble structures as the s wall of Room 5. The portance in the history of ceramics, is also well heavier wall in its lower parts was partly built over represented: no. 24 below is part of the rim of a by walls that served as retaining walls for the deep bowl covered with a metallic (probably lead) rubbish dump underlying the late Roman build- glaze characteristicof numerous fragments of both ing (w walls of Rooms 4 and 5). open and closed vessels found in late 4th and early A final destruction date in the 5th or, more likely, 5th century deposits at Stobi. Such an early use early 6th century is indicated for the complex s of of glaze has not been reported elsewhere in the the terrace wall. The earliest construction belongs Byzantine world.54The question whether local pot- to the closing years of the 4th century (Room 3) ters manufactured the ware or imported it from with at least one major rebuilding phase before the Pannonia will have to await further information destruction and abandoning of the area. The de- on the late Roman ceramic industry at Stobi. The posit immediately below the terrace wall shows relationship, if any, between this ware and later that the latter was constructed no earlier than the Byzantine glazed pottery deserves investigation. last decade of the 4th century. 23. Bowl rim (pl. 87, fig. 38) C-7I-67. Max. P. The soft, dumped fill on which the rubble struc- Dim. 0.09 m. Thickness of rim 0.006 m. Dark tures were built sloped E-w and apparently ac- gray micaceous clay. Black lustrous paint covers cumulated over a short period of time, perhaps body. Ca. one-fourth of rim preserved.Fragment is After 50 years. the removal of most of the terrace from shallow bowl with horizontal rim. Scalloped wall and the walls of the late Roman complex that edge outlined by two curved, concentric grooves; over the lay parodos at the w end, a series of these circular groove below point of each scallop. Both sloping deposits was excavated in an area i i x 4.2 designs and scalloping were impressed. Two addi- m. The and lowest highest elevations during work tional concentric grooves bisect rim. This type of here in 1971 were 141.35and 138.13m. The original bowl is known from Greek Macedonia in contem- floor of the parodos was not reached (pl. 89, fig. porary deposits (personal communication from Mr. 37). John W. Hayes, who refers to this ware in his The differed deposits little from each other in forthcoming book on Roman pottery). This dis- color or texture: almost all were soft, mixed gray- tinctive decoration and rim type also occur at brown soils with much ash and charcoal. All con- Stobi with a lustrous red paint or occasionally tained great quantities of animal bones and pot- unpainted. sherds. The number of bronze coins recovered tery 24. Glazed rim fragment P.H. 0.057 m. from these some was C-7i-200. deposits, 250, remarkable, W. of rim 0.038 m. Th. of rim o.oii m. Coarse clay but most were All the identifia- poorly preserved. fired gray at core, red near surface (latter, Hue ble coins dated from the time of Constantius II 2.5 YR 5/8). Red to greenish-brownglaze. Outward to TheodosiusI (A.D. 337-361) (A.D. 379-395)and thickened bowl rim set off from bowl by raised all legible mint names were located in the Eastern ridge and two grooves. Rim decorated with incised The date of the coins and that of empire. the bulk wavy line and two grooves at edge. of the context pottery leave no doubt that most Open bowls with a broad rim and a similar of the fill of the accumulated dur- great parodos glaze are known from Germany during the late the second half of the ing 4th century. 4th century: Erich Gose, Gefdsstypender rdmischen hundred sherds are now known in a 54Several glazed 7th for the existence of glazed wares in the Byzantine world from recent excavations in century deposit Istanbul; see John at this period is minimal; no parallels can be cited earlier than W. "A ac- Hayes, Seventh-Century Pottery Group," 203-216, about the middle of the seventh century." He goes on to sug- R. Martin Harrison and Nezih companying Firathl, "Excava- gest a date "around or slightly after A.D. 600" for the earliest tions at in Istanbul: Fifth DOP Saraqhane Preliminary Report," appearanceof the ware. We may now add that several examples 22 (1968) 195-203. Hayes' comments on p. 216 indicate the of a glazed, thick-rimmed bowl were found in certain mid- state of previous knowledge concerning glazed wares as well 6th century context in the Gymnasium Area of Ancient Corinth as the new contribution the material from concerning Saraqhane: during the 1971 excavations directed by James Wiseman. We "The of in presence glazed pottery appreciable quantities in a are much indebted to John Hayes for his helpful comments re- of ca. A.D. is group 650-670 of considerable interest. Evidence garding catalogue nos. 23-24.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 419 Keramik im Rheinland, BonnJbb Beih. i (Bonn in excavations by the Conservation Institute of 1950) 20, no. 235. The closed vessels from Stobi are Macedonia during the I96o's."5The following com- similar to a group from Pannonian graves dated by ments on the theater are based on measurements associated coins to A.D. 350-361; even the variations and computations by William B. Dinsmoor, Jr. in glaze colors are paralleled: A. Sz. Burger, "The The lower tier of the cavea has 8 stairways and the Late Roman Cemetery at Saigv~r,"ActaArch 18 upper presumably had 15. The two tiers were (Budapest 1966) 140-141.Pannonia, indeed, seems separated by a diazoma which would have been to have preserved the lead glaze process from the located directly over the outer wall of the central Ist through the 4th century; for evidence of both annular corridor, and a second horizontal passage closed and open vessels from an early to mid-4th must have existed at the top of the theater at the century ceramic workshop in Pannonia, see E. B. back of the upper tier. The theater then allows for "Die Villa von Thomas, r6merzeitliche Tic-F6Iveny- 17 rows of seats in the lower tier and 19 in the puszta," ActaArch (Budapest 1955) 120-123. upper tier. If, as in the Theater of at Athens, 0.41 m. is allowed for each seat, there were A new state plan of the theater (ill. 3) includes 2,597 seats in the lower tier and 5,o04 in the upper the portions of the cavea and substructurerevealed tier, or a total seating capacity of 7,638. 55 Cf. Ejnar Dyggve, "Le theatre mixte du bas empire d'apres plan of the theater (his fig. 6) by Dyggve is remarkably ac- le theatre de Stobi et les diptyques consulaires,"RA (1958) 2, curate considering how little of the theater was visible at the 137-157, and especially his state plan, fig. x. The restored time.

4, l O?)RCUIR CORROtOR t

x %

-/•! , ",,%," , II \ \ . 4? 4

r ::j 4? , r0 / // III:l O\ \ L , 0,1i ",,% ,' -

---X ------

It I IC :i~ -- * / r ... ..n.n . . .- i'-.' ...u --• 197...... L, .. .I. ..r . ..x ... .' -r 20 40~ EtIG AAC

------7.63...... - -.. . r .:.J. - ...... Y------• ......

3,

ILL.3. Plan of theater

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 420 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 7. HOUSEOF THE FULLER 23-25points to industrialfunctions for this area. Excavations in the House of the Fuller were ex- A large,roughly rectangular pit lay to the N and when the structure was suc- tended to the area Nw of the large apsidal hall.56 early 5th century ceeded about the middle of the another Four main periods of habitation, each with sub- century6'by humble a oval succeeded the phases of building activity, were identified. Virgin one, small, pit larger one. The floor of the was at soil has not yet been reached. stone-paved dwelling an elevation m. A single test trench in the NE corner of Room of 150.314/150.36o I. The final ancient of habitation 21 (see below) revealed a portion of a mortar and 4. period lay just rubble wall that antedates the House of the Fuller. below the present ground surface. The chief struc- tural features include two stone rubble The small amount of contextual material suggests preserved walls and of a floor of a date in the 3rd century for the earliest floor of part (elev. 151.716m.) large, Room 21 built over the wall. rectangular stone slabs. The construction date would be late or 6th 2. This is the chief period of the House of the 5th early century, certainly but sure context for its destruction Fuller. Room 21 was entered from the apsidal post-Theodoric, hall57 and provided access on the left to Room 22 is lacking. and on the to a room or court right (unexcavated). 8. THE EPISCOPAL BASILICA The floor of Room 22 was supported on circular The main floor of the Basilica lies at terracotta bricks (pl. 89, fig. 39). Hot air entered Episcopal the level of the Via Sacra the hypocaust area beneath a brick arch from a approximately (ill. i:3 and this and furnace in the sw corner of Room 21. The hypo- 5). Worshippers passed along paved colonnaded avenue to reach the main entrance of caust was abandoned early in the 4th century58and the Basilica the Semicircular Court four floors were successively superimposed,the final opposite (ill. But the residence of the one lying 30-40 cm. higher than the first-thus I:4). bishop (ill. 1:8), to the basilica on the at a much necessitating raising the level of the threshold con- adjacent N, lay lower level and was reached from the w a monu- necting it to Room 21. The last floor was built near by mental the end of the century, clearly not long before the stairway.62 We can now see that the Basilica stood destruction of the building." The destruction Episcopal in on an artificial terrace E from debris was a deep mass of adobe brick, wooden part extending timbers and roof tiles. near the Via Sacra to the ruined outer wall of the abandoned theater. The area at least 3. No structural remains were found above the immediately s, as far w as the was at a lower level. The destruction deposits of Rooms 21 and 22, but a narthex, floor of the on the s lies ca. rubbish dump of the 5th century is to be associated principal building m. below the level of the s of the nave. with a relatively humble residential complex some- 4.4 stylobate The basilica must indeed have towered over the what to the w. This complex of small rooms with earthen floors and narrow rubble wall socles lies surrounding buildings and dominated the view from direction. above three rooms, 23-25, which had as their E any wall the w wall of Rooms 21-22. There is no reason The area to the s could be reached directly from the church A in s at present to suppose that Rooms 23-25 belonged to (ill. 4). doorway the wall of the the same residence as the first two rooms dis- narthex led to a sandstone stairway bounded on cussed."6The discovery of 30 large, rough hemis- the s by a wall of roughly-shaped stones set in a pheres of iron slag in 4th century deposits of Rooms coarse and crumbly lime mortar. In an earlier 56 The building was referred to in W-MZ 1970 as the 128) date to the reign of Valentenian I (A.D. 364-375). But North Palace, but it is far too humble to be called a palace coin 71-382 from the next earlier floor was minted during the and the directional adjective is inappropriate.Its present name reign of Theodosius I (A.D. 379-395). is suggested by a quantity of murex shells found N of the 6oThe lower deposits of these rooms, where tested, also apsidal hall. show occupation from the 3rd through the 4th century. 57 A bronze lamp was found in front of the threshold: W-MZ 61 The latest coin (71-286) belongs to the reign of Valen- 1970, Section 6. tinian III (A.D. 425-455). 58Nine coins 62 were found, the earliest of which (71-3II) The Episcopal Residence was partially excavated in the belongs to the reign of Diocletian (A.D. 284-305), and the 1930's. A brief discussion of the building and a restored plan latest (71-375) to that of Constantius II (A.D. 337-361). have been published: Dj. Mano-Zissi, "Stukatura u Stobima," 59Ceramic contextual material was plentiful in the floor Zbornik Narodnog Muzeja 3 (Belgrade 1962) 101-107. deposit. There were eight coins, three of which (71-I26, I27,

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 421

ru ?L~l*~ILIFLOOR CRYPT c :?~Rno to; oC-1 -, r - 4w W ? .• ... . •, ..... • <30_.S P...... _ . ... " . •.. o...... 7. _.2,c-•: . .. il ?

ILL. 4. Area s of EpiscopalBasilica period the landing at the top of the stairway was floor at elevation 144.1 m. A hard, pink cement, paved with sandstone slabs at an elevation of typical of mosaic beddings both in the Episcopal 147.63m., but at the time of the destruction of the Basilica and in the Synagogue, lay ca. 30 cm. above basilica a marble step had been added at the top Floor I and may have carried a mosaic during the and a second floor surface of square, baked tiles early days of the Floor church."6 3 (elev. 144.87m.) laid ca. 19 cm. higher, nearly at the level of the was again an earthen surface but includes a tiled sandstone threshold. drain, covered by stone slabs and part of a marble During the final period of use an earthen floor column, which runs roughly N-s. The drain, 0.35 covered the lower three steps and extended through m. wide on the interior, descends from an area ap- a doorless passage into a room to the sw (pl. 89, parently below the narthex. Its construction in- fig. 40). The reuse of architectural fragments to volved the removal of a part of the lowest sand- adorn this entrance seems a melancholy attempt to stone step; the stairway, then, had been constructed repair a crumbling elegance. Part of a marble at least by the time of Floor 2. column, badly fitted on an Ionic base and crowned Ceramic material at all levels was scanty and the with a nearly square block of green marble, stood absolute chronology of Floors 1-3 is uncertain. on one side of the passageway facing a tall mullion Coin 71-89, found in the deposit of Floor 4, dates column which was found where it had fallen for- to sometime between A.D.518 and 565.64 ward from its rectangular base. The column stood A deep destruction debris reaching nearly to the at the w end of Wall 2, a wall identical in struc- present ground surface overlay the stairway and the ture to the wall of the stairway to which it may be rooms to the SE.Most of it consisted of bricks and connected: its lower part extends below the column mortar from the collapsed s wall of the basilica. and may originally have served as a bed for a A large number of fragments of fresco were found, threshold. some preserving parts of faces, hands and drapery, The room reached by the passage just described especially in the first room. is bounded on the w by a wall which abuts the The function of that room and the next two to stairway wall at a right angle, and on the E by a the E is as yet undetermined. The wedge-shaped curved wall. Both walls are dry rubble (mortar- space enclosed chiefly by Wall 2 and the terrace less) structures; the latter is pierced by a doorway. wall of the basilica contained a sloping tile floor Tests below the earthen floor revealed three earlier that must have taken much of the rainwater from living surfaces. The earliest (Floor i) was reached the guttered roof of the basilica and channeled it only in a small test trench; it is a hard-packedearth into the drain of Floor 3. The blocking wall on 63 Numerous large chunks of mosaic were found in the room, 64 An orb and cross are visible on the obverse; the reverse but above Floor 4. preserved a K and perhaps the letters AN.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 422 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 the w was built at the time of Floor4 when the The walls of the Baptistery are constructed of drain was no longer in use. moderately large blocks set in a coarse concrete that has badly deteriorated. The N wall is pre- BAPTISTERY OF THE 9. EPISCOPAL BASILICA served to a height of nearly 4 m. above the mosaic The most significant architectural discovery of floor, the s wall only to ca. 2 m. Fresco is pre- the i971 season is the large baptistery s of the served on the lower parts of all walls and one large basilica, the existence of which was suspected last patch is still in place high on the central N apse. year (ill. 1:6; pl. 90o, fig. 41).65 It was destroyed A quantity of window glass and two mullion col- in the same 6th century catastrophe"6that over- umns found in the debris show that there were at whelmed the basilica and, presumably, the rest of least two windows in the building. the city. At the time of its final destruction the A large circular piscina (interior diam. 2.4 m.) Baptisteryhad alreadyundergone some considerable is centered in the room and set off from it by a modification of form. The earlier figured frescoes brick parapet, 0o.85/1.o0m. wide, which rises ca. had been covered by coarse yellow plaster; six un- 0.63 m. above the mosaic floor. The piscina is faced fluted columns had been spaced around the parapet with slabs, on the exterior alternating marble and of the central piscina, their Ionic bases secured to slate and on the interior white and red marble. The the mosaic floor by a thin layer of lime mortar; a floor of the piscina, also of marble, lies 1.33 m. few holes in the mosaic, perhaps damaged in some below the preserved top of the parapet.In its center earlier disaster, had been filled with cement flush rose a hollow marble cylinder, spirally fluted on its with the floor itself. exterior surface; its function is as yet unclear. The Baptistery (ill. 4) is nearly square on the The piscina was apparently entered by stepping exterior, ca. 9.4 m. to a side, but the interior space up onto the parapet and then descending one of is marked by a number of apses. The principal en- three steep marble stairways. A large marble kan- trance was from the s where a broad doorway, ca. tharos rose from about the level of the mosaic floor 1.5 m. wide, is flanked by shallow apses. The N within a space made for it in the parapet (pl. 9o, wall, built against the thick terrace wall of the fig. The kantharos is 0.94 m. high and the 42).6' basilica, has a broad central apse with a smaller diameter at the top is o.6 m.; carefully carved dou- one on each side to balance the arrangement of the ble volutes adorn two sides in imitation of handles. s wall. The E and w walls each have a single, shal- The entire vessel, including volutes, was carved low apse. In each corner there is an entrance, at from a single piece of white marble. Six small least two of which (NE and NW) were spanned by marble columns, with painted bands of red, yellow brick and concrete arches. and black, seem to have stood on small Ionic bases The Nw archway gave access to a narrow space on the parapet and may have carried the roof of in (i.6 m.) separated from the complex described a baldacchino. Section 8 by a wall or parapet, only the low, plas- Six larger columns, some with figured composite tered, stone socle of which is preserved. At the N capitals, stood on the mosaic floor close against the end of the hallway, and filling its entire width, is parapet wall during the final phase of the Bap- a doorway with a flat lintel. The doorway, which tistery. Only one base was still in place, but all the was partially blocked at some period with a brick bases, or fragments of them, were found near the wall, leads into the center of a long side of a large mortared squares marking their location. At least vaulted crypt, oriented E-w and built within the one was cut from an inscribed monument (pl. 89, substructureof the basilica's terrace wall. It is rea- fig. 43).68 sonable to suppose that the crypt once held the relics The mosaic (pl. 90, fig. 47) has a band of running of a martyr or high official of the church. spirals along the walls, around the piscina, and 65 W-MZ 1970, Section 3 fin. 2nd line, o.o9 m.; 3rd line 0.07 m.; 4th line 0.05 m. Inter- 66 Coin 71-708, an issue of the 6th century, was found lying linear space I-2, 0.03 m., 3-4, 0.025 m. on the mosaic. DEO C[AESARI 67 Part of the kantharos base was found in place and other AVGV[STO fragments were among the debris within the piscina. IOVIS LIBERA[TORIS - - 6 The principal fragment is 1-71-10 (= A-71-I21). Part SECUNDI.L.BA[ --- of the inscription was reused, face down, as the Ionic column -traces- base nearest the s entrance of the Baptistery. The base is Line 3: Juppiter Liberator was worshipped along with Minerva square, 0.37 m. on a side; H. 0.245 m. H. of letters: Ist and and Juno Regina in the time of ; see Thulin, "Iuppiter

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 1972] EXCAVATIONS AT STOBI, 1971 423 separating the four main panels. Each panel dips The form of this sumptuous Baptistery, essen- towards one of the corner doorways and has as a tially a quatrefoil with small corner apses inscribed centerpiece a gadrooned kantharos rising from a within a quadrangle, may have its origin in Pales- stylized floral design. Water spouts from a foun- tine; e.g., the Baptistery of St. John the Baptist at tain within the kantharos. A peacock stands on Gerasa. The plan became well known in Greece either side of the kantharos in the NE and sw and elsewhere in the 5th and 6th centuries.70Other panels and waterfowl are scattered among foliage aspects raise a number of interesting questions. The in the field. In the Nw panel two does drink from piscina, with its sealed joints and a drain under the kantharos in the presence of birds and in the the floor, seems to have been intended to hold SE panel, the least well preserved, a doe and an water. May we suppose, then, that baptism here antlered buck are the main figures. Each doorway was by immersion? But baptism by infusion was has a scale pattern and in two semicircular niches customary inside formal structures (i.e., baptis- of the parapet there is a ray design. The mosaic teries) of the 4th to 6th centuries.7' And what is formed of small stone tesserae of a variety of was the function of the marble kantharos? Surely colors. The technical quality is high and the mo- the similarity between the vessel itself and the saicist, with his varicolored pieces, succeeded in representationon the adjacent mosaic (and numer- imparting a feeling of roundness to his figures. ous other mosaics as well) is significant. In the The original interior of the Baptistery was re- mosaics the kantharos either supplies the "fluid of markably colorful. In addition to the bright mosaic life" or flows with an abundance of foliage.72 But and the varicolored marbles of the piscina, the although the representationon mosaics is common, arcuated space was enlivened by wall frescoes. Nu- there is, so far as we are aware, no vessel of com- merous faces have been found among the thousands parable size known from any other baptistery. If of fresco fragments; several with the "sideways the vessel served any function other than a sym- glance" (pl. 89, fig. 44) characteristicof some por- bolic one, it is likely to have held holy water. How traits known earlier from the narthex of the Epis- that water was utilized is another question.73 copal Basilica.69Faces are shown both frontally and There is little point in attempting a stylistic in profile, and in a variety of sizes. The varying dating of frescoes not yet mended or of the mosaic, scale of the heads suggests that the artist was trying especially since the excavation of 1972 may well to indicate depth. Each portrait and each pose so provide us with more reliable chronological evi- far noted seem individualized. The painter (or dence. It does seem most likely, however, that the group of painters) was surely an artist of excep- original construction, with mosaic and frescoes, is tional skill and imagination-see the young man no later than the mid-5th century. A detailed com- in the blue chiton (pl. 89, fig. 44) and the delicate mentary on the architecture must also await fur- features of the youth in pl. 89, fig. 45. ther study. Fortunately, the Baptistery had not (9)," RE io (Stuttgart 1919) 432. Line 4: Secundus is com- Quatrefoil Baptistery): Hoddinott (supra n. 69) 168-169. mon as a cognomen on inscriptions from Stobi. The second 72On representations of peacocks drinking the "fluid of name may be Bassus; a veteran's inscription from Stobi life" from such a vessel, see esp. E. R. Goodenough, Jewish honors C. Iulius Bassus of the tribe Maecia, a veteran of the Symbols in the Graeco-Roman World VIII: Pagan Symbols in 8th Augustan legion; CIL III 630. Line 5: the last 3 letters on Judaism (New York 1958) 55-57. There is a deer beside a r seem to be DED and suggest dedicavit. More of the text may kantharos in the mosaic of the triclinium in the House of be preserved on the bottoms of the two bases not yet removed Psalms; Dj. Mano-Zissi, "Mosaiken in Stobi," BIABulg io from the Baptistery. (1936) fig. 184. Two deer flank a similar kantharos in a 69 R. E. Hoddinott, Early Byzantine Churches in Macedonia mosaic discovered recently in a basilica at Longos near Edessa; and Southern Serbia (London and New York 1963) pl. 43: Georges Daux, "Chroniquesdes fouilles, 1967," BCH 91 (1967) Iovanka Maximovi6 "Contribution a l'6tude des fresques de 898, figs. 1-2. The style and technique of the mosaic at Longos Stobi," CahArch io (1959) 207-216. closely resemble those of the Stobi mosaics. The Longos mosaic 70 It is larger than the octagonal baptistery room (inner and several others in Greece depicting the kantharos are illus- dimensions 7.3-7.4 m.) adjoining the enormous basilica at trated in Jean-Pierre Sodini, "Mosaiques paleochretiennes de Corinthian Lechaion: Demetrios I. Pallas, 'AvaoKaq r?7?rj Grkce," BCH 94 (1970) 699-753. Ba-LXLrK~sjroi Aexaiov, Praktika (1961) 137-154. On the plan 3 Large stone vessels for holy water must usually have stood see, e.g., A. Khatchatrian,Les baptisterespalMochretiens (Paris in the atrium; cf. the large marble phiale (possibly 5th cen- 1962) figs. 63 a (Gerasa), 185 (Cos), 358 (Riva San Vitale tury) from the courtyardof the Basilicaof St. Demetrius in Thes- in Switzerland), 369 (Aquileia). salonica in A. K. Orlandos, 'H vX6o-reyos7r-aXatoXptoaravrLKl 71 At least in the Greek world most of the fonts are too BaCo-kXIKM(Athens 1952) 120, fig. 70. On the use of basins in small to allow immersion. An example close at hand is the baptisteries (e.g., baptism of children and foot-washing), see baptistery of the North Basilica (formerly Basilica of the Pallas (supra n. 70) 145-146.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions 424 JAMES WISEMAN AND DJORDJE MANO-ZISSI [AJA 76 been disturbed since its destruction and the archi- finger in the clay before the brick was fired (pl. tectural embellishments were all found where they 89,fig. 46) :5 0E'EJXl'o-oVXpLo-rE . There may have had fallen in the collapse of the building. been a fourth line in which Christ was called upon Bricks from the Episcopal Basilica frequently separately. The plea for God to have pity seems carried a molded cross in one of several varieties." at first a mournful sentiment to discover etched in Just as the brick maker in that instance knew that clay amid such extensive ruins as those of the his product was intended for a church, so, too, Baptistery. But the invocation, after all, belongs to did the manufacturer of bricks for the Baptistery. the time of construction, surely a more prosperous Many of the bricks were incised with birds or fish and happier period for the people of Stobi than and occasionally a more complex scene involving the final days of the city. a human. The most interesting brick is a fragment UNIVERSITY OF TEXAS AT AUSTIN preserving part of an invocation written with a UNIVERSITY OF BELGRADE

74 Rudolf Egger, "Die stiidtische Kirche von Stobi," JOAI 24 0.025 m. H. of letters 0o.o04/0.o6m. Coarse, reddish yellow (1929) 46. clay (Hue 5YR 6/6) with inclusions. 75TA-7I-30 (= I-7I-II). P.L. 0o.18 m. W. 0.26 m. Th.

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions WISEMAN AND MANO-ZISSI PLATE 85

;:1C;ux. ri~i ~9 r. r :-1 :e ik~:l~cF;:?: c t, ,-??;; i" h I i C hfS i t, ?x. ~ I ~i~i~- r *"?9~- ~" "; ni~?~-l ~- ,~a~-~~la;as~."~i~a~~?~B

P

FIG. I FIG. 6

-r .. :ii% k~- ;-U r ?r \i ~ r?

:l??r~-

"-? :-~;:"~:1?$ -:??i: ?-a?~ 'i 1-i::--?--ss ii ?:?:?_i. gie:??:

cE:~*... ,aY-::l ~Cs --i: FIc. 7 i?-Wii

-?*r u -- f + ?r P as~~li i,;_?: 1 ;v :.i??i: ?; ~ FIG. 13 i rC

e

WEPOWT

I

x '"'CAPApe104P-; ..

FIG. 14

FIG. 15

I I .....i i•? i!!"i•!•!i~il;$ I •i• !i4 |II I o ... I .. I U

FIG. 4 FIG. 10 FIG. II This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions PLATE 86 WISEMAN AND. MANO-ZISSI

FIG. 5

FIG. 2F

FIG. 9 FIc. 9

FIG. 8 FIG. 8

-ow

FG-FI22I-

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions WISEMAN AND MANO-ZISSI PLATE 87

FIG. 19

FIG. I8

I61 FI. FIG. 21

FIG. 38 FIG. 20

FIG. 23 I I I I

24 FIG.

F IG.29

FIG. 26 FIG. FIG. 25 28 27 FIG. This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions PLATE 88 WISEMAN AND MANO-ZISSI

- ; -: :::.--::.i L.; 3 8 N

it; 9 4 s~E f:

j

~x~~ i

-:-i ;" : cl. i ";~ :i?? PP-~ -- J CF--=: 1;; ~II:_--- _i ~di;_,-:" ;- -: - ;r: ;I__? 3

?_-i? e r I ;-, n .. :";" r -- -~*?r ?I 1~C*r

ci D ~-f~a%1R~li~?l?ri. ": a .: i:iiii r;l::B? ;r;rr;;i; o P ??i

FIG. 30 FIG. 31 FIG. 32

:i';"'; ??-~n;--:- ?

:i•:!!i• ::|

F• 3F . 4F .35F .3

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions WISEMAN AND MANO-ZISSI PLATE 89

37 FIG. i~I

FIG. 39

Ii.40

At

,~~ .....

Fic. 44 Fic FIG.45 46

FIG. 43

Z o

G 4

... .4F

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions tro

CRYPT 0

CiT

3* z z

z 0 FIG. 41 1 N CI CI

S2 4 DIWMOR,JR.

FIG. 47

FIG. 42

This content downloaded from 128.83.205.78 on Tue, 02 Jun 2015 16:12:41 UTC All use subject to JSTOR Terms and Conditions