M AR MO RA a n i n t e r nat i o na l j o u r na l f o r a r c h a e o l o g y, h i s t o ry a n d a r c h a e o m e t ry o f marbles and stones 10 · 2014 offprint p i s a · ro m a fa b r i z i o s e r r a e d i to r e m m xv Direttore · Editor Lorenzo Lazzarini · Università i.u.a.v. (Venezia) * Comitato scientiico internazionale · International Scientiic Committee Archeologia e Storia dell’Arte · Archaeology and History of Art Clayton J. Fant · Università di Akron (oh, u.s.a.) Anna Maria Giusti · Opiicio delle Pietre Dure (Firenze) Olga Palagia · Università di Atene Patrizio Pensabene · «Sapienza» Università di Roma Isabel Rodà · Università Autonoma di Barcellona Carmelo G. Malacrino · Museo Archeologico Nazionale di Reggio Calabria Archeometria · Archaeometry Aurelio Álvarez Perez · Università Autonoma di Barcellona Vincent Barbin · Università Reims-Champagne-Ardenne (f) Claudio D’Amico · Università di Bologna James A. Harrell · Università di Toledo (oh, u.s.a.) Marino Maggetti · Università di Friburgo (ch) Myrsini Varti-Matarangas · i.g.m.e. (Atene) * «Marmora» is an International Peer-Reviewed Journal. The eContent is Archived with Clockss and Portico. THE QUARRIES OF THE MAGARA DERESI AND THE MARBLE OF THE TEMPLE OF ARTEMIS AT SARDIS Nicholas Cahill* · Lorenzo Lazzarini** Abstract Samples of marble from the Magara Deresi quarries near Sardis and from Lydian, Hellenistic, and Roman buildings from the site were analyzed using minero-petrographic and isotopic analysis. The samples from the quarries provide the irst detailed petrographic descriptions of this marble, while the isotopic analysis enlarges the known isotopic ield, which overlaps the ield for Ephesian marble, requiring petrographic examination to distinguish between the two. Results of analysis from the sanctuary of Artemis at Sardis are reported here, and show that the Hellenistic and Roman portions of the temple were both built from marble from the Magara Deresi. A block re-used as spolia, however, is isotopically similar to samples of Lydian marble buildings analyzed previously, and may come from quarries north of the Gygaean Lake. keywords: Turkey, Lydia, Sardis, marbles, ancient quarries, Artemision, archaeometry. Introduction T he city of Sardis, capital of the Lydian Empire and a metropolis throughout the Roman period, was rich in marble buildings and sculptures dating from the Archaic to the Byzantine periods. A number of these monuments have been recently archaeometrically investigated to identify the origin of the marble used in their construction. Hellenistic and Roman marble sculptures and buildings from the site had already been attributed to the Magara Deresi quarries in the foothills of the Tmolus south of Sardis on the basis of visual, petrographic, and X-ray difraction analysis (Butler 1922, 19-20, 176; Hanfmann and Waldbaum 1975, 17, 21; Hanfmann and Ramage 1978, 4-7; Whitmore 1978). More recent analysis of marble from Lydian architectural elements from the sixth and ifth centuries bc, on the other hand, indicates a possible origin from quarries north of the Gygaean Lake, although a diferent origin and possible import from Ephesus were not excluded. In spite of the importance of determining the precise sources of marble used for mon- uments in Sardis and other ancient cities, systematic investigation of the marble quarries around Sardis is still incomplete (Monna and Pensabene 1977, Ramage and Tykot 2002 and 2011). Analysis of carbon and oxygen isotope ratios of marble from Sardis was irst undertaken by Norman Herz (1987), but without naming the speciic quarries of the 18 samples he analyzed. More recently, isotopic analysis and basic petrographic examination (mgs, Maximum Grain Size) were employed on marble samples taken from two quarry sites, the Magara Deresi and ancient quarries near Gölmarmara (Ramage and Tykot 2002 and 2011). The small number of samples considered (8 and 12, respectively) and the incomplete analytical techniques applied to them make the proposed identiications of the marble sources of important Sardian monuments such as the tumulus of Alyattes inconclusive. We have therefore initiated a project to collect more information on the quarries and their marbles in order to extend the reference database for archaeometric studies, and to sample selected ancient monuments to understand the use of diferent marble * Addresses for correspondence: University of Wisconsin at Madison (wi, usa). ndcahill@wisc.edu ** Laboratorio di Analisi dei Materiali Antichi, Sistema dei Laboratori, Università iuav di Venezia, San Polo 2468/b, i 30125 Venezia (Italy). lorenzo@iuav.it «marmora» · 10 · 2014 28 Nicholas Cahill · Lorenzo Lazzarini sources through the city’s long history. In this article we present the results of analysis of samples from the Magara Deresi quarries, and from the sanctuary of Artemis. The quarries of the Magara Deresi The Magara Deresi is a steep ravine located about three kilometers south of the Temple of Artemis, feeding into the Pactolus stream (Fig. 1). Now largely dry in summer but often looded in winter, the stream has cut deeply through beds of marble, exposing the stone in the sides and base of the gorge. A number of individual quarries are visible on both sides of the ravine, and tool marks in the marble of the streambed show that this too was extensively quarried in antiquity. The quarries are spread over an area of at least 0.5 km n-s along the bed of the stream, and extend well up the slopes of the ravine, about 0.3 km e-w. They range in elevation from about 390 m a.s.l. to well over 450 m, hundreds of meters higher than the buildings of Sardis which they supplied, which lie at roughly 120-180 m a.s.l. A thorough survey and study of the quarries is not the intention of this article. The precise number of individual quarries in the valley has not been tallied. Many are now dificult of access, high on the sheer face of the ravine, and have never been intensively surveyed. In many places the sides of the ravine are deeply buried in discarded marble chips from ancient quarrying, obscuring the original contours. Modern looters’ tunnels have dug through 5-10 m of solid quarry debris, without reaching the bottom of the deposits. The volume of debris and size of the quarry faces give a sense of the huge amount of stone quarried from the ravine. In 2009 Leah Long estimated that more than 20,000 cubic meters of marble was quarried from the Magara Deresi. Her dissertation (2012) is the most complete recent study; but as she could not include many of the ancient quarry sites in the Magara Dere in her survey, she herself suggests that the volume is signiicantly too low. She did not sample or analyze the marbles themselves. Ancient monuments in the ravine and vicinity include a relief of two frontal ig- ures, perhaps a god and goddess, in an aedicula cut into one of the quarried areas on the west bank of the ravine (Hanfmann and Ramage 1978, no. 156). An inscription reading OPO Ϲ , «boundary», is cut into a quarry face on the east bank (our quarry 1, Greenewalt and Rautman 1998). Stretches of an ancient road used to haul marble from the quarries are still visible in the east slope of the ravine (Fig. 2). Two artiicial, deeply quarried caves, one on the east and one on the west side of the ravine, give the modern name to the ravine, ‘Valley of the Caves’, but the bulk of the stone was probably quarried from open vertical faces and from the streambed. East of the ravine is a rectangular stone structure, perhaps a Hellenistic watchtower, now known as ‘Kresus Hastanesi’, or ‘Croesus’ Hospital’, perhaps associated with the road that ran high along the east bank of the ravine and up into the Tmolus Mountains. Marble sampling in the quarries and in the sanctuary of Artemis For this study we sampled six areas that had been worked in antiquity; other quarried areas will be sampled in future seasons. The six quarries (Fig. 1, Tab. 1) were: 1. Two quarried clif faces at roughly right angles, facing west and south, relatively high on the east slope of the ravine, near its northern end (Fig. 3). This area has been the subject of recent looting, and is also the site of the boundary inscription. 2. A cave on the east slope of the ravine. Long estimates the volume of the cave at 1,200 m3. A number of worked blocks around the cave show modern drill holes, and are said to be from a mid-20th-century attempt to re-open the quarries. The interior shows pick and wedge marks of ancient quarrying (Fig. 4). 3. In the bed and sides of the stream below this cave are pick marks and troughs showing that blocks were removed here in antiquity. Modern looting has dug a pit 15-20 m deep here, through chips and ill. 4. South of this in the bed of the stream are a number of uninished marble column drums, about 0.8 m in diameter. The MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS 29 SARDIS AND VICINITY Contour interval 20 m 0 500 1000 m Maǧara Deresi quarries 1-6 NDC 2014 Izm ir-A nk ara Hi ghw ay 200 100 Sardis ea m 300 4260000 Sanctuary of Artemis Acropolis 200 to lu s S tr 300 c Pa 300 200 4258000 300 200 400 road 2 3 500 1 6 4 Hellenistic watchtower? 300 5 Ma ǧ ara Deresi quarries 600 4256000 400 500 700 590000 800 588000 600 Fig. 1. Map of Sardis showing location of the Magara Deresi quarries. streambed here is solid marble, which seems generally iner-grained and whiter than that of quarries 1-3; and is covered with tool marks, sometimes almost invis- ible after centuries of erosion in the stream. Two uninished column drums, belonging to unidentiied buildings, are preserved here (Fig. 5). 30 Nicholas Cahill · Lorenzo Lazzarini Fig. 2. Ancient road leading to the Magara Deresi quarries. Fig. 3. View of Magara Deresi quarry 1, showing the deep pit dug by looters exposing the ancient quarry face buried by solid marble chips. The total height of the quarry face is more than 15 m. 5. Relatively high on the east bank of the stream, a tall quarried face (Fig. 6). The area around is deep in marble chips, and a modern looter has dug a tunnel 5-10 m into the chips without reaching their bottom. 6. On the west bank of the stream near its bed is a cluster of quarries including a deep quarried cave surrounded by quarried clif faces (Fig. 7). Long estimates the total stone removed here at 6,650 m3. The area between these quarries is illed with chips to an uncertain depth, and a deep lime kiln has been built within this spoil heap. In addition, samples were taken from a number of monuments at Sardis to determine their origin. Reported in this article are the samples from the sanctuary of Artemis. Blocks from both the Hellenistic and Roman elements of the building were sampled (Tab. 2 - Gruben 1961, Howe 1999, Yegül 2012, Cahill and Greenewalt forthcoming). Two samples were taken from the column of the Artemis Temple now in the Metropolitan Museum of Art, New York. This Hellenistic column probably originally be- MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS Fig. 4. View of quarry 2. Fig. 5. View of quarry 4 (modern streambed) and ancient column. Fig. 6. View of quarry 5. 31 K-mica Apatite He Mosaic, strongly strained 4.16 Sutured + + ± He Strongly lineated, mortar 4.42 Sutured ± ± He Mosaic strongly strained with kinks 1.88 Sutured He Mosaic, strongly strained with kinks 3.68 Sutured Q2 S1 He Mortar, weakly strained 2.80 Sutured Q2 S2 He Mosaic with intercryst. inegrained crystals 3.20 Sutured Q2 S3 He Mosaic with intercryst. inegrained crystals 3.60 Sutured He With granulation along boundaries 3.22 Sutured He Mosaic, weakly lineated 3.44 Sutured Q1 S1 Q1 S2 8.5 Q1 S3 Q1 S4 Quarry 2 Q2 S4 Quarry 3 Q3 S1 8.5 9.5 ± ± ++ ± + ± Carb. Opaque matter/ Minerals Graphite Dolomite (xrd) Quartz Quarry 1 Serpentine Calcite/ Dolom. Crystal’s boundaries Plagioclase Fabric mgs mm Color Index Munsell nvs ‰ 13C (+) ‰ 18O (–) +++ ++ P, Hm – 4.59 7.93 +++ + – 4.51 8.43 ++ ± – 4.51 7.18 +++ ± – 4.52 8.26 +++ + P. L – 3.00 12.65 ++ + P, L – 3.79 12.67 ++ ±P – 4.00 11.19 +++ ++ P, Hm – 3.71 10.91 +++ ++ P – 5.01 7.57 (continued) Nicholas Cahill · Lorenzo Lazzarini Fabric Type. In general: strained, with bent polysynthetic twins Magara Deresi Sample no. quarry no. 32 Table 1. Summary of the minero-petrographic and isotopic results (nvs Neutral Value Scale; He, heteroblastic; Ho, homeoblastic; P, Pyrite; Hm, Hematite; L, Limonite; +++, very abundant; ++, abundant; +, present; ±, traces; –, absent). Quarry 5 Quarry 6 4.02 Sutured ± + + He Mosaic, strongly strained 2.56 Sutured ± ± Q4 S1 9.0 He Mosaic, weakly strained 2.40 Sutured Q4 S2 8.5 Ho Purely mosaic 1.16 Curved Q5 S1 He Mosaic, weakly strained 3.58 Sutured Q5 S2 He Mortar, strongly 4.40 Sutured Q6 S1 He Mortar 3.60 Sutured ± Q6 S3 He Mosaic, lineated and strained 4.02 Sutured ± Q6 S4 He Mosaic strongly strained 4.00 Sutured He Mosaic, weakly strained 3.84 Sutured Pit no. 5 8.5 + ‰ 13C (+) ‰ 18O (–) +++ +P – 5.06 7.34 ± +++ ++ – 5.04 7.58 + +++ + Hm – 3.90 4.54 ± ± +++ + – 3.47 7.45 ± ± ++ + P, Hm – 4.27 4.93 ± +++ +P – 3.77 12.92 +++ + P, Hm, L – 4.68 5.29 ++ + P, Hm – 5.05 5.34 +++ +P – 4.54 7.04 +++ + P, Hm – + ± Carb. Opaque matter/ Minerals Graphite Dolomite (xrd) Mosaic, weakly lineated Serpentine He Plagioclase Calcite/ Dolom. Crystal’s boundaries Apatite Quarry 4 mgs mm K-mica Q3 S3 Fabric Type. In general: strained, with bent polysynthetic twins Quartz 8.5 Fabric Q3 S2 Color Index Munsell nvs ± + ± ± ± ± MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS Magara Deresi Sample no. quarry no. 33 Sutured 2.80 S4 Fragment of another Hellenistic lute from the same context he Mosaic, weakly Sutured 2.72 ± S5 Fragment of one of the ‘Dutchman’ repairs rom Roman column no. 7 he Mosaic, strongly Sutured 2,96 + S6 Fragment of rooftile, of Roman date as shown by inscriptions on the tiles he Mosaic, mortar, weakly Sutured 3.98 S7 From a block of the Hellenistic wall between cella and pronaos he Mortar, strongly Sutured 2.96 S8 From a block of the Hellenistic foundation of cella column no. 74 he Mosaic, strongly Sutured S9 From a block of the Roman foundation of peristyle column no. 50 he Mosaic, strongly Intergran. cry. Sutured ‰ 18O pdb (–) Mosaic, strongly ‰ 13C pdb (+) he + Dolomite (xrd) Fragment of the same or another Hellenistic capital from the same context S3 Op.Min. 2.88 Carb.m./ Graphite Sutured Titanite Mosaic, weakly Fragment of capital F, from the same context and also probably from one of the Hellenistic columns in antis Apatite he S2 Plagioclase 2.40 he K-mica Embayed S1 Fragment of luted column recovered from the late antique pit in the east porch, probably belonging to one of the Hellenistic columns in antis of the east porch Quartz Mosaic, weakly Object Probable provenance – 4,26 12,93 Magara Deresi quarry +++ ++ P – 4,42 8,87 Magara Deresi quarry +++ ++ P. L – 4,93 6,53 Magara Deresi quarry +++ ++ P – 4,74 7,81 Magara Deresi quarry ± +++ ++ E P – 4,71 8,74 Magara Deresi quarry ± ± ++ ±E – 4,76 6,91 Magara Deresi quarry ± ± +++ +P Hm – 4,79 5,39 Magara Deresi quarry 2.48 ± ± +++ ++ E P – 4,75 8,71 Magara Deresi quarry 1.68 ± +++ ++ E P – 2,77 9,66 Magara Deresi quarry ± ++ ± (continued) Nicholas Cahill · Lorenzo Lazzarini Calcite Crystals Boundaries mgs Fabric. All samples show weakly/ strong strain Sample no. 34 Table 2. Results of the minero-petrographic and isotopic analyses of the marble samples of the Artemis Temple (see Table 1 for symbols). Embayed 2.48 + ± S12 From a block in the north end of the east wall of Building Q he Mortar, weakly Sutured 2.64 +++ ± ± M1 Metropolitan Museum (ny), fr. column inv. 26.59.1 he Mosaic, lineated weakly Sutured 2.42 ++ + M2 Metropolitan Museum (ny), fr. capital 26.199.283 he Mosaic, lineated Sutured 3.60 ± ‰ 18O pdb (–) Mosaic, weakly ‰ 13C pdb (+) he Mosaic, strongly Intergran. cry. Dolomite (xrd) S11 From a pre-Hellenistic block reused in the Hellenistic north wall of the cella he Op.Min. From the jamb of the Roman east door Carb.m./ Graphite S10 Titanite Apatite Plagioclase 2.96 K-mica mgs Sutured Quartz Calcite Crystals Boundaries Object Probable provenance ++ ++ E P – 5,07 4,84 Magara Deresi quarry ++ +P – 0,61 5,08 Uncertain quarry, probably north of the Gygaean Lake ++ + P, Hm – 4,53 5,60 Magara Deresi quarry ++ P, Hm – 7,90 4,70 Magara Deresi quarry ++ P – 6,50 4,50 Magara Deresi quarry +++ MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS Fabric. All samples show weakly/ strong strain Sample no. 35 36 Nicholas Cahill · Lorenzo Lazzarini Fig. 7. View of quarry 6. longed to the cella of the temple (Lazzarini and Marconi 2014). Two of the samples presented here do not belong to the Hellenistic or Roman phases of the Artemis Temple. Sample AT S11 comes from a reused block built into the north wall of the Hellenistic cella. This bears butterly/staple clamp cuttings and square dowel cuttings, and is similar to blocks built into Building Q and found elsewhere in the sanctuary. The masonry techniques of these blocks, including their carefully picked raised centers, drafted edges, and exclusive use of the lat chisel, suggests a date earlier than the Hellenistic period, perhaps in the second half of the sixth or ifth century bce; it certainly predates the Hellenistic temple. Sample AT S12 comes from one of the roughly-worked blocks in the north end of the east wall of Building Q. A similarly worked block in the southwest corner of the building bears a Lydian inscription. The blocks may well be pre-Hellenistic in date, although perhaps in reuse in this building (Cahill and Greenewalt forthcoming). An unusual feature of the Temple of Artemis is the neatly carved verse inscription around the lower drum of column no. 4, one of the Roman peristyle columns on the east façade (Buckler and Robinson 1932, no. 181; Yegül 2014). As read by Buckler and Robinson, the column proclaims that its torus and foundation block are cut from one stone; that this is the irst of all (the columns) to rise; and (in a passage that is diicult to understand) that it is (built) of stones «not wrought by the people» but «from their own stone(s)». ì Û[]ÂÖÚ· ¯ó [®]È˙·ÖÔ˜ Âx˜ âÛÙÈÓ Ï›ıÔ˜, ÚáÙÔ˜ ‰b ¿ÓÙˆÓ âÍ ¬ÏˆÓ àÓ›ÛÙ·Ì·È Ôé ‰ËÌÔÙ‡ÎÙˆÓ, àÏÏ\ à\ ÔåΛˆÓ Ï›ıˆÓ. The language is poetic and obscure. The opposition in the inal phrase would seem to be MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS between ‰ËÌfiÙ¢ÎÙÔ˜, a unique word meaning ‘wrought by the demos’, and ÔåÎÂÖÔ˜, literally ‘of the household’ or ‘one’s own’ stones. It perhaps contrasts quarries owned or controlled by the people of Sardis to those owned by the sanctuary of Artemis. However the verse is to be understood, the implication seems to be that the source of this marble was local, and that this was a source of pride. Experimental methods Minero-petrographic analyses Minero-petrographic analyses are of fundamental importance in characterizing marbles to determine their petrogenesis, i.e. the type of metamorphism (burial, contact, regional) and grade (low, medium, high). While a more or less complete characterization is possible for all marbles from monuments and quarries, the petrological determination of their genesis is generally possible only for quarry samples, and when sampling is made by geologists that combine ield studies of the marble outcrops and quarries with laboratory analyses. The limitation in the characterization is obviously due to the small size of the marble sample that may be taken from monuments. For the present study all determinations, including when possible a color evaluation on some of the quarry and monument samples, were made on a single small fragment with a maximum size of ca. 2 × 3 × 0.5 cm. Color was described using a Munsell Colour Chart (Neutral Scale). Part of the sample was then inely ground and the powder subjected to difractometric (X-radiation CuK·/Ni, at 40 Kv, 20 mA) and isotopic analyses (see below). The remaining part was used for the preparation of a thin section for a detailed minero-petrographic study of the marble under a polarizing microscope. The purpose of this examination was to determine the fabric, accessory and secondary minerals, in addition to the calcite and dolomite crystal-characteristics which are usually the principal constituents of all types of marble. More speciically, the following parameters were determined: 37 - type of fabric (homeoblastic = with roughly isodiametric grains, heteroblastic = with grains of various dimensions), in direct relationship with the metamorphic type (equilibrium, non equilibrium, retrograde metamorphism, polymetamorphism, etc.); - boundary-shapes of the calcite/dolomite grains, also directly connected to the type of metamorphic event/s that generated the marble; - maximum grain size, a parameter of signiicant diagnostic importance since it is linked to the grade of metamorphism achieved by the marble; - qualitative and semi-quantitative presence of accessory minerals, sometimes of diagnostic value. For the petrographic description, previous speciic studies of the most important ‘major’ ancient marbles, as well as other archaeometric studies of ‘minor’ marbles and classical treatises on petrotectonics were taken into consideration (Lazzarini et alii 1980, Spry 1969). Isotopic Analyses Isotopic characterization has proved to be very useful in the identiication of ancient marble artifacts. Its use is becoming more and more widespread owing to its outstanding sensitivity, the small quantity of material necessary for analysis, and the availability of an extensive database, continuously growing and often associated with other laboratory methodologies which permit increasingly trustworthy comparisons, especially if the isotopic data are evaluated together with the minero-petrographic results from the same samples, as in the present study (Gorgoni et alii 2002, Attanasio et alii 2006). The isotopic analyses were carried out on the carbon dioxide derived from small portions (20-30 mg) of the powdered sample subjected to a chemical attack with 100% phosphoric acid at 25° in a special vacuum line, according to the procedure suggested by McCrea and Craig. The resulting CO2 was then analyzed by mass spectrometry. The instrument is equipped with a triple collector and permits the measurement of both isotopic ratios (13C/12C and 18O/16O) at the same time. 38 Nicholas Cahill · Lorenzo Lazzarini Fig. 8. a) Photomicrograph of sample from quarry 1 (Q1S4) showing a mosaic of large calcite crystals with some strain evidenced by bent polysynthetic twinning, N+, long side = 2.55 mm; b) Sample from quarry 2 (Q2S3), as a) but with strongly strained fabric and deformed crystals; c) Sample from quarry 1 (Q1S2), as a) but showing a foliated (e-w) fabric with deformed crystals; d) Sample from quarry 2 (Q2S4), as a), but showing a typical mortar fabric with slightly strained crystals. The analytical results are conventionally expressed in ‰ units, in parts per thousand: and from those published in literature (Ramage and Tykot 2011). ‰ sample = (Rsample / Rstd - 1) × 1000 Results and discussion in which Rsample and Rstd represent the isotopic ratio of oxygen and carbon in the sample and in the reference standard, respectively. The standard adopted is pdb for both oxygen and carbon (the pdb standard is the rostrum of the Belemnitella americana of the Cretaceous Pee Dee Formation of South Carolina). The identiication of the provenance of the marbles of the sanctuary of Artemis was obtained by comparing their minero-petrographic and isotopic data with those collected from the analysis of the quarry samples The marble of the Magara Deresi quarries The results of the minero-petrographic and isotopic analysis of the Magara Deresi marble quarries are reported in Table 1. The color measurement was made only on a few suiciently large marble samples: it varied from a more frequent 8.5 (white to light grey) to 9.5 (white). All marbles were analyzed by xrd: in addition to calcite, they revealed only slight traces of muscovite and pyrite. The microscopic study in general showed quite homogeneous features. The MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS 39 Fig. 9. Ramage and Tykot’s 2011 reference isotopic diagrams (in grey) and the new, larger, diagram for the Magara Deresi quarries (in black). fabric detected under the polarizing microscope in 22 thin sections was that of a mosaic of interlocked calcite crystals showing clear signs of a metamorphic strain (from weak to strong - Fig. 8a, b) evidenced by deformed (sometimes kinking) crystals and bent polysynthetic traces of twinning. An associated foliation was less frequently observed (Fig. 8c), or a typical mortar fabric (Fig. 8d). All samples (with the exception of Q4S2 with curved boundaries) showed sutured boundaries and thus well interlocked crystals and very abundant ine particles of carbonaceous matter/graphite dispersed inside the calcite crystals or concentrated along their boundaries or in small aggregates. The mgs varies from 1.16 mm to 4.42 mm, with the highest frequency around 3.50 mm. Common accessory minerals are K-Mica, quartz, apatite and iron ores: of the latter, pyrite is quite abundant; hematite is often present and associated with pyrite, of which is often an alteration; limonite is much more rare. Very rare (and perhaps characteristic of these Sardian quarries?) is the presence of twinned albitic plagioclase, serpentine and titanite. The isotopic values of the 18 samples analyzed vary from +3.00 to +5.05 for ‰ 13C, and from -4.54 to -12.92 for ‰ 18O. In the isotopic diagram they fall only partially within the isotopic ield built from the data provided for these Sardian quarries by Ramage and Tykot, and so signiicantly enlarge the known ield (Fig. 9). Both petrographic and isotopic results allow Sardian marble from the Magara Deresi quarries to be clearly distinguished from most of the other medium/coarse-grained marbles used in antiquity. Petrographically the fabric is very characteristic, and does not match other marbles; its mgs overlaps with that of Naxian and Thasian marbles, and is well separated from other important ancient marbles including those with a ine grain size (Fig. 10). Isotopically, this marble only partially overlaps with Naxian and Ephesian Marbles (Fig. 11), the latter a marble so far resulted of only local importance, and easily distinguished by its petrographic fabric characterized by a non-strained, mostly inegrained fabric and curved calcite boundaries. The marble of the sanctuary of Artemis The results of the minero-petrographic and isotopic analyses of the marbles of the sanc- 40 Nicholas Cahill · Lorenzo Lazzarini Fig. 10. The mgs marble from the Magara Deresi quarries compared with the other most important marbles used in antiquity. Fig. 11. Plot of the new reference diagram of the Magara Deresi quarry in the general reference diagram of the most important ancient marbles with mgs > 2 mm. N = Naxian; Pr 1-2 = Proconnesian; T1 = Thasian-Phanari; T2 = Thasian-Aliki; T3 = Thasian Vathy-Saliara; Pa-2 = Parian-Lakkoi; Pa-4 = Parian Karavos; Aph = Aphrodisian. MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS 41 Fig. 12. Photomicrograph of sample AT S5, from the Roman column of the east façade of the Temple of Artemis, similar to Figure 8a, but showing a slightly strained mosaic fabric. tuary of Artemis are presented in Table 2. Comparing these results with those of table 1 and the related discussion, it appears that there is a good match between the petrographic characteristics of these samples and those of the Magara Deresi quarries (Fig. 12), with the exception of sample 11 (Fig. 13). The plotting of the isotopic data in the new isotopic reference diagram (Fig. 14) show that all the samples taken from the Hellenistic and Roman architectural elements of the temple fall within or very near the ield of the Magara Deresi quarries. It should be noted that samples AT S1, AT S9, and AT S10 plot slightly outside this isotopic ield, thus indicating that more sampling in those quarries is needed to establish a more reliable reference ield in the isotopic diagram. The samples of the capital in the Metropolitan Museum (samples M1 and M2) also proved to be made of marble from the Magara Deresi quarries. It is also signiicant that sam- ples from both the Hellenistic and Roman phases of the temple are similar, in one instance isotopically almost identical (samples AT S5, from one of the Roman columns of the east porch, and AT S8, from the foundation of one of the Hellenistic cella columns). The marble of sample AT S11, from a reused block built into the north wall of the temple, whose architectural techniques date the block to the later sixth or ifth century bce, is totally diferent from the others with respect to the boundary shape (embayed, not sutured as the other samples) and isotopic results. This sample is isotopically similar, however, to samples of Lydian architectural elements analyzed by Ramage and Tykot (2011), including the stone chamber of the tumulus of Alyattes, a piece of worked marble from the Lydian tumulus of Karnıyarık Tepe, architectural elements from Lydian building(s) of the early sixth century bce at sector ByzFort, and a kline leg from a 42 Nicholas Cahill · Lorenzo Lazzarini Fig. 13. Photomicrograph of sample AT S11, from an early block re-used in the north wall of the cella of the Temple of Artemis, similar to Figure 8a, but showing a mosaic, slightly heteroblastic fabric distinct from that of the marble from the Magara Deresi quarries. Fig. 14. Plot of the results of the isotopic analysis of the samples from the sanctuary of Artemis at Sardis. Samples M1 and M2 refer to the column and capital preserved in the Metropolitan Museum of Art (ny). MAgARA DERESI ’ s quarries AND ARTEMIS temple ’ s marble AT SARDIS tumulus near Kendirlik. Assuming that all these pre-Hellenistic marbles derive from the same quarry, the source of this marble remains to be deinitively determined. The analyzed samples are at the edge of or outside the existing reference ields of the Gölmarmara quarries, but within the ield of quarries at Yeniköy north of the Gygaean Lake, analyzed by Sekedat (2012). It is possible that further sampling of the Gölmarmara quarries would extend its isotopic ield, however; and secure attribution of these architectural blocks to any of the quarries north of the lake would depend on mineropetrographic analysis of those marbles to supplement the isotopic data. Sample AT S12, taken from Building Q whose Lydian inscription and masonry style suggest that the blocks, even if reused, derive originally from a pre-Hellenistic monument, might have been expected to plot with the other known pre-Hellenistic marbles of Sardis. However, it clearly derives from the Magara Deresi quarries. In conclusion, we may state with some conidence that the temple of Artemis at Sardis was built with marble from local quarries in the Magara Deresi. With further sampling and analysis, it might be possible to identify exactly which quarries were used for diferent parts of the temple; but both Hellenistic and Roman builders seem to have used the same general sources. A pre-Hellenistic block used as spolia, on the other hand, seems to derive from the same quarry or quarries used for other known Lydian architectural marble, including the tumulus of Alyattes and the elite complex on ByzFort. We hope that further analysis of dated monuments from Sardis and quarries in the region will elucidate the changing use of marble sources over time, and the opening of these locally important quarries. References Attanasio D., Brilli M. and Ogle N. 2006, The isotopic signature of classical marbles, Todi («Studia Archaeologica», 145). Buckler W. H. and Robinson D. M. 1932, Sardis vii : Greek and Latin Inscriptions, Part 1, Leiden. Butler H. C. 1922, Sardis i : The Excavations, Part 1: 1910-1914, Leiden. 43 Cahill N. D. and Greenewalt C. H. jr. forthcoming, The Sanctuary of Artemis at Sardis: Preliminary Report, 2002-2012, «aja». Foss C. 1978, Explorations in Mt. Tmolus, «California Studies in Classical Antiquity», 11, 21-61. Gorgoni C., Lazzarini L., Pallante P. and Turi B. 2002, An Updated and Detailed Mineropetrographic and C-O Stable Isotopic Reference Database for the Main Mediterranean Marbles Used in Antiquity, in J. J. Herrmann, N. Herz, R. Newman (eds.), asmosia v , Interdisciplinary Studies on Ancient Stones, Proceedings of the Fifth International Conference of the Association for the Study of Marble and Other Stones in Antiquity (Boston, June 11-15, 1998), London, 115-131. Greenewalt C. H. Jr. and Rautman M. L. 1998, The Sardis Campaigns of 1994 and 1995, «aja», 102, 469-505. Gruben G. 1961, Beobachtungen zum Artemis-Tempel von Sardis, «am», 76, 155-196. Hanfmann G. M. A. and Ramage N. H. (eds.) 1978, Sculpture from Sardis: the Finds through 1975, Cambridge (ma)-London («Archaeological Exploration of Sardis», Report 2). Hanfmann G. M. A. and Waldbaum J. C. 1975, A Survey of Sardis and the Major Monuments Outside the City Walls, Cambridge (ma)-London («Archaeological Exploration of Sardis», Report 1). Herz N. 1987, Carbon and oxygen isotope ratios: a database for Classical Greek and Roman marble, «Archaeometry», 29, 35-45. Howe Th. N. 1999, The Toichobate Curvature of the Artemis Temple at Sardis and the End of the Late Hellenistic Tradition of Temple Design, in L. Haselberger (ed.), Appearance and Essence: Reinements of Classical Architecture: Curvature. Proceedings of the Second Williams Symposium on Classical Architecture Held at the University of Pennsylvania, Philadelphia, April 2-4, 1993, Philadelphia, 199-210. Lazzarini L. and Marconi C. 2014, A New Analysis of Major Greek Sculptures in the Metropolitan Museum: Petrological and Stylistic, «The Metropolitan Museum of Art Journal», 49, 119-142. Lazzarini L., Moschini G. and Stievano A. 1980, A Contribution to the Identiication of Italian, Greek, and Anatolian Marbles through a Petrological Study and the Evaluation of Ca/Sr Ratio, «Archaeometry», 22, 173-183. Long L. 2012, Urbanism, Art, and Economy: The Marble Quarrying Industries of Aphrodisias and Roman Asia Minor, Ph.D. dissertation, University of Michigan. Monna D. and Pensabene P. 1977, Sart-Sardis, in Marmi dell’Asia Minore, Roma, 177-183. 44 Nicholas Cahill · Lorenzo Lazzarini Ramage M. H. and Tykot R. H. 2002, On the Importation of Monumental Marble to Sardis, in J. J. Herrmann, N. Herz, R. Newman (eds.), asmosia v , Interdisciplinary Studies on Ancient Stones, Proceedings of the Fifth International Conference of the Association for the Study of Marble and Other Stones in Antiquity (Boston, June 11-15, 1998), London, 335-339. Ramage M. H. and Tykot R. H. 2011, Geological Analysis of Lydian Building Stones and their Quarry Sources, in Chr. Ratté, Lydian Architecture. Ashlar Masonry Structures at Sardis, Cambridge (ma)-London («Archaeological Exploration of Sardis», Report 5), 127-132. Ratté Chr. 2011, Lydian Architecture. Ashlar Masonry Structures at Sardis, Cambridge (ma)London («Archaeological Exploration of Sardis», Report 5). Sekedat B. 2012, Large Polities and Small Quarries: Local Resources and Imperial Governance in Roman Asia Minor, Ph.D. dissertation, Brown University. Spry A. 1969, Metamorphic Textures, Oxford. Whitmore F. E. 1978, Preliminary Analysis of Marble Samples, in G. M. A. Hanfmann and N. H. Ramage (eds.), Sculpture from Sardis: the Finds through 1975, Cambridge (ma)-London («Archaeological Exploration of Sardis», Report 2), 6-7. Yegül F. K. 2012, The Temple of Artemis at Sardis, in Th. Schulz (hrsg. von), Dipteros und Pseudodipteros, Istanbul («Byzas», 12), 95-111. Yegül F. K. 2014, A Victor’s Message: The Talking Column of the Temple of Artemis at Sardis, «Journal of the Society of Architectural Historians», 73, 204-225. composto in car attere dante monotype dalla fabrizio serr a editore, pisa · roma. stampato e rilegato nella tipo gr afia di agnano, agnano pisano (pisa). * Dicembre 2015 (cz 3 · fg 22) Tutte le riviste Online e le pubblicazioni delle nostre case editrici (riviste, collane, varia, ecc.) possono essere ricercate bibliograicamente e richieste (sottoscrizioni di abbonamenti, ordini di volumi, ecc.) presso il sito Internet: www.libraweb.net Per ricevere, tramite E-mail, periodicamente, la nostra newsletter/alert con l’elenco delle novità e delle opere in preparazione, Vi invitiamo a sottoscriverla presso il nostro sito Internet o a trasmettere i Vostri dati (Nominativo e indirizzo E-mail) all’indirizzo: newsletter@libraweb.net * Computerized search operations allow bibliographical retrieval of the Publishers’ works (Online journals, journals subscriptions, orders for individual issues, series, books, etc.) through the Internet website: www.libraweb.net If you wish to receive, by E-mail, our newsletter/alert with periodic information on the list of new and forthcoming publications, you are kindly invited to subscribe it at our web-site or to send your details (Name and E-mail address) to the following address: newsletter@libraweb.net Rivista annuale · A yearly Journal * Indirizzo redazione scientiica · Scientiic Committee Address Lorenzo Lazzarini · l.a.m.a. (Dip. di Storia dell’Architettura), Università i.u.a.v., San Polo 2468, i 30125 Venezia, tel. + 39 041 2571413, -459, fax +39 041 2571434, lorenzo@iuav.it * Amministrazione e abbonamenti · Administration & Subscriptions Fabrizio Serra editore ®, Pisa · Roma Casella postale n. 1, Succursale n. 8, i 56123 Pisa, tel. +39 050 542332, fax +39 050 574888, fse@libraweb.net Uici di Pisa: Via Santa Bibbiana 28, i 56127 Pisa, fse@libraweb.net Uici di Roma: Via Carlo Emanuele I 48, i 00185 Roma, tel. +39 06 70493456, fax +39 06 70476605, fse.roma@libraweb.net I prezzi uiciali di abbonamento cartaceo e Online sono consultabili presso il sito Internet della casa editrice www.libraweb.net. Print and Online oicial rates are available at Publisher’s website www.libraweb.net. I pagamenti possono essere efettuati tramite versamento su c.c.p. n. 17154550 o tramite carta di credito (American Express, Visa, Eurocard, Mastercard). * Autorizzazione del Tribunale di Pisa n. 11 del 15 · 04 · 2005 Direttore responsabile: Fabrizio Serra A norma del codice civile italiano, è vietata la riproduzione, totale o parziale (compresi estratti, ecc.), di questa pubblicazione in qualsiasi forma e versione (comprese bozze, ecc.), originale o derivata, e con qualsiasi mezzo a stampa o internet (compresi siti web personali e istituzionali, academia.edu, ecc.), elettronico, digitale, meccanico, per mezzo di fotocopie, pdf, microilm, ilm, scanner o altro, senza il permesso scritto della casa editrice. Under Italian civil law this publication cannot be reproduced, wholly or in part (included ofprints, etc.), in any form (included proofs, etc.), original or derived, or by any means: print, internet (included personal and institutional web sites, academia.edu, etc.), electronic, digital, mechanical, including photocopy, pdf, microilm, ilm, scanner or any other medium, without permission in writing from the publisher. Proprietà riservata · All rights reserved © Copyright 2015 by Fabrizio Serra editore®, Pisa · Roma. Fabrizio Serra editore incorporates the Imprints Accademia editoriale, Edizioni dell’Ateneo, Fabrizio Serra editore, Giardini editori e stampatori in Pisa, Gruppo editoriale internazionale and Istituti editoriali e poligraici internazionali. Stampato in Italia · Printed in Italy www.libraweb.net issn 1824-6214 issn elettronico 1826-8072 SOMMARIO Editoriale · Editorial 9 saggi Francesco Tognoni, Gian Battista Brocchi e la scoperta di antiche cave di pietre ornamentali nel deserto orientale egiziano 15 Nicholas Cahill, Lorenzo Lazzarini, The quarries of the Magara Deresi and the marble of the Temple of Artemis at Sardis 27 Fede Berti, Diego Peirano, Il marmo iasio e alcuni tra i manufatti a esso collegati 45 Caterina Previato, Arturo Zara, Il trasporto della pietra di Vicenza in età romana. Il relitto del iume Bacchiglione 59 Myriam Pilutti Namer, Tra spolia e imitazioni: i capitelli della basilica di Santa Maria 79 Assunta a Torcello (ve ) Marco Da Villa, Maria Da Villa Urbani, Venezia: la Mariegola de l’arte nostra de tajapiera 101 recensioni Patrizio Pensabene, I marmi nella Roma antica (Lazzarini) 117 Jacopo Bonetto, Stefano Camporeale, Antonio Pizzo (eds.), Arqueología de la construcción iv. Las canteras en el mundo antiguo: sistemas de explotación y procesos productivos (Rodríguez Gutiérrez) 119