The History of Apollo's Temple at Didyma, as told by Marble Analyses and Historical Sources

Originalveroffentlichung in: L. Lazzarini (Hrsg.), Interdisciplinary studies on ancient stone. ASMOSIA VI, Venice 2003, S. 271-278

THE HISTORY OF APOLLO'S TEMPLE AT D I D Y M A , AS T O L D B Y MARBLE ANALYSES A N D HISTORICAL SOURCES

B . E . B o r g1 a n d G . B o r g2

1 Archaeological Institute, Ruprecht-Karls-University, Marstallhof 4, D - 6 9 1 1 7 Heidelberg, Germany, barbara.borg@urz.uni-heidelberg.de

2 Institute for Geological Sciences, Martin-Luther-University Halle-Wittenberg, Domstr. 5, D-06108 Halle, Germany, borg@geologie.uni-halle.de

A b s t r a c t

T h e temple o f A p o l l o at D i d y m a in western A n a t o l i a has a par­

ticularly l o n g b u i l d i n g history f r o m ca. 3 0 0 B . C . until well into the R o m a n era. D u r i n g this period, various types o f w h i t e m a r ­ ble were used. T o determine their provenance, petrology and iso­

tope analyses o f carbon and oxygen were c o m b i n e d w i t h consid­

erations o f practicability and historical probability. Four different marble sources were identified: Milesian and Herakleian marbles, both quarried around Lake Bafa about 2 6 k m to the northeast o f D i d y m a , marble f r o m A l i k i o n the Island o f Thasos, and Pro- connesian marble f r o m the Island o f Marmara. M a p p i n g o f these marbles on the t e m p l e revealed that they were employed during different periods. W h e r e a s Milesian marble was used most wide­

ly d u r i n g the Hellenistic and R o m a n periods, Thasian marble was used o n l y for a short t i m e d u r i n g the 2nd c. B.C.; Herakleian marble was e m p l o y e d d u r i n g the later Hellenistic and early R o m a n periods and Proconnesian marble was used in the R o m a n era. I d e n t i f y i n g the marble types o f various origins, thus, c o n ­ tributes considerably to the discrimination o f individual b u i l d ­ ing phases and, especially, to the attribution o f particular parts o f the architecture to these phases.

K e y w o r d s : D I D Y M A , A P O L L O T E M P L E , M A R B L E , I S O ­ T O P E A N A L Y S I S , P E T R O L O G Y , R A R E E A R T H E L E ­ M E N T S ( R E E ) , L A K E B A F A , M I L E T U S , H E R A K L E I A , T H A S O S , P R O C O N N E S O S

I n t r o d u c t i o n

T h e f a m o u s oracle t e m p l e o f A p o l l o at D i d y m a , b u i l t over t w o predecessors f r o m the archaic p e r i o d (TUCHELT, 1991;

SCHNEIDER, 1 9 9 6 ) , c o u n t s a m o n g the largest t e m p l e s o f the G r e c o - R o m a n w o r l d . Erected o n a p l a t f o r m o f seven steps a n d a base o f a p p r o x i m a t e l y 118 b y 6 0 m , a d o u b l e c o l o n n a d e sur­

r o u n d s a b u i l d i n g c o m p l e x c o n s i s t i n g o f an open courtyard (the a d y t o n w i t h a s m a l l naiskos c o v e r i n g the sacred spring), a front hall w i t h 12 c o l u m n s (called the d o d e k a s t y l o s ) , and a hall w i t h t w o c o l u m n s between these, that is approached f r o m the a d y t o n by a w i d e f l i g h t o f stairs a n d opens i n t o the d o d e k a s t y l o s t h r o u g h a m o n u m e n t a l portal (fig. I). T h i s p o r ­ tal's t h r e s h o l d is a b o u t 1.5 m above the level o f the floor in the d o d e k a s t y l o s a n d m a y have served as a s t a g i n g area for the prophetes w h e n he emerged f r o m the h o l y space o f the a d y ­ ton. T h e t e m p l e has a l o n g b u i l d i n g history, b e g i n n i n g s o m e t i m e a r o u n d 3 0 0 B . C . and lasting w e l l i n t o the R o m a n era

( K N A C K F U S S , 1 9 4 1 ; T U C H E L T , 1 9 7 3 ; V O I G T L A N D E R , 1 9 7 5 ; R U M - SCHEID, 1994). A set o f 2 8 inscriptions d a t i n g f r o m the 3rd a n d 2nd cc. B . C . provides information about a m a j o r part o f its b u i l d ­ ing history a n d a b o u t details o f the sequence, organisation, a n d costs o f the b u i l d i n g activities themselves (REHM, 1 9 5 8 ; G U N T H E R , 1 9 6 9 / 7 0 ) .

Three o f these inscriptions also g i v e the names o f t w o sites s u p p l y i n g the marble, Marathe and Ionia Polis, the latter b e i n g the harbour from where the marble has been shipped (I. D i d y m a 39, 31; 4 0 , 16; 4 1 , 23. 2 8 ; cf. REHM, 1958). T h u s , not surpris-

Fig. 1 - Didyma, Temple of Apollo, distribution of marble types.

Fig. 2 - Map of the eastern Mediterranean.

W/ffli Milesian Quames W Mikhail (.marries i [^^j Milesian (juames N Kftffi I icracleinn Quarries

Fig. 3 - Map of Lake Bafa region with ancient quarry areas (modified after WIEGAND, 1 9 1 3 , a n d PESCHLOW-BINDOKAT, 1 9 8 1 : fig. 1).

ingly, the search for these sites and their associated marble q u a r - ries started at the b e g i n n i n g o f the 20t h c. soon after the inscrip- tions were discovered. It was o n l y in the 1 9 7 0 s , however, that P e s c h l o w - B i n d o k a t ( 1 9 7 7 / 7 8 and 1 9 8 1 ) could finally establish the locality o f Ionia Polis at the south shore o f Lake Bafa, a for- mer e m b a y m e n t o f the Aegean Sea a p p r o x i m a t e l y 2 6 k m to the northeast o f D i d y m a , near the m o d e r n village o f Mersinet Iske- lesi (Mersinet in fig. 3). Surveying the south and east shores o f Lake Bafa, she also located and m a p p e d a large n u m b e r o f m a r - ble quarries. She convincingly argued that those stretching along the south shore from Ionia Polis to the west belonged to the city o f M i l e t u s , also owner o f the sanctuary at D i d y m a , wheteas the quarries on the east shore belonged to the nearby independent city o f Herakleia (fig. 3). T h a t the Milesian south shote quarries d i d indeed supply the marble for the t e m p l e o f A p o l l o can still be seen from the large n u m b e r of unfinished and broken c o l u m n

d r u m s l y i n g a r o u n d , w h i c h , according t o their diameters o f about 2 m , can o n l y have been intended for this very b u i l d i n g ( P E S C H L O W - B I N D O K A T , 1 9 7 7 / 7 8 a n d 1981; fig. 4).

GERMANN ( 1 9 8 1 ) studied the petrology of both the Hera- kleian and Milesian marbles. These are situated w i t h i n the contact metamorphic aureole o f the granitic gneiss core of the Menderes Massif to the northeast o f Lake Bafa. T h e most recent tectono- metamorphic overprint of the gneisses and surrounding sediments occurred in the Tertiary (SENGOR etai, 1984; Loos and REISCHMANN,

1999). T h e marbles o n the east and north shores of Lake Bafa, to w h i c h a Palaeozoic age has been attributed (DORR, 1975), occur as well-defined lenses in muscovite schists w i t h steep southerly to ver- tical dips. T h e quarries o n the south shore are o f Mesozoic age (DURR, 1975) and are hosted by moderately southerly d i p p i n g mar- ble lenses that are also enclosed in muscovite schists and phyllites.

These lenses are usually smaller and far more irregular in shape than the ones on the east shore. A marked macroscopic petrological dis- crepancy exists between the east and the south shore quarries in terms o f the textural occurrences o f dolomite within these predo- m i n a n t l y calcitic marbles. T h e south shore marbles display d o l o m i t i c layers o f u p to 4 0 c m in thickness w h i c h have been b o u - dinaged, with individual boudins showing ductile wrap-around textures along their outer margins and sets o f sharp, parallel, brit- tle fractures w i t h i n the individual boudins (figs. 5 and 6). D o - lomitic portions w i t h i n the east shore marble are less abundant and occur as millimetre-thin, parallel, laterally discontinuous laminae, or as small-scale, complex, ptygmatic folds w i t h i n the calcitic mar- ble. These different textural styles vary systematically w i t h the distance to the gneiss contact and thus represent the contact meta- m o r p h i c gradient. Germann's preliminary petrological study of the marble used at the temple also confirmed that large quantities c o m e f r o m the south shore o f Lake Bafa. However, he also identi- fied some o f the marble as Herakleian from the east shore and even recognized a third type o f marble, a particularly coarse crystalline one, w h i c h he could not identify the source for (GERMANN, 1981).

Fig. 4 — Lefka-Bur-Dag quarry with broken column drum, south shore of Lake Bafa.

T h e objectives o f o u r p r o j e c t arose f r o m the above obser- v a t i o n s and consisted (1) in the i d e n t i f i c a t i o n o f the u n k n o w n source o f the coarse crystalline m a r b l e , a n d (2) i n the m a p p i n g o f the various marbles at the t e m p l e for the purpose o f e s t i m a - t i n g the q u a n t i t i e s o f i m p o r t e d stone, d a t i n g these i m p o r t s , a n d i d e n t i f y i n g possible reasons for the i m p o r t o f m a r b l e .

M e t h o d o l o g y

A s is w e l l k n o w n today, there is still n o single m e t h o d to d e t e r m i n e the provenance o f w h i t e m a r b l e s w i t h certainty and t h u s , a c o m b i n a t i o n o f m e t h o d s is necessary. O n the other h a n d , a m u l t i - m e t h o d approach is l i m i t e d by the s m a l l size and n u m - ber o f samples that can be taken f r o m archaeological artefacts, a n d b y the a v a i l a b i l i t y o f data for possible m a r b l e sources a r o u n d the Mediterranean.

Isotope analyses o f carbon and o x y g e n were carried out on 58 quarry samples and 4 7 samples f r o m the t e m p l e at D i d y m a b y H a r a l d Strauss at the Institut fur G e o l o g i e , R u h r - U n i v e r s i r a t - B o c h u m , G e r m a n y . These are g i v e n as delta-values i n per m i l re- lative t o the P D B standard (CRAIG, 1 9 5 7 ; CRAIG and CRAIG 1972).

T h e quarry samples f o r m a representative set for the quarries on the n o r t h , south and east shores o f Lake Bafa. T h e y were collec- ted f r o m all parts o f the quarries but o n l y f r o m excavation walls and large blocks s h o w i n g tool m a r k s in order to m a k e sure that the locations sampled were indeed quarried in antiquity. N o sam- p l i n g was d o n e i n the m a r g i n a l areas w h e r e the m a r b l e m i x e s w i t h the n e i g h b o u r i n g schists even w h e r e these showed tool marks. T e m p l e samples were taken both as loose fragments from fractured blocks and c o l u m n d r u m s and as drill cores ( 1 0 m m diameter) obtained d u r i n g restoration w o r k . O n l y those samples taken w e l l b e l o w the exposed surfaces were used for isotopic analyses in order t o avoid c o n t a m i n a t i o n t h r o u g h alterations by weathering. T h e petrographic s t u d y included all the temple's c o l u m n d r u m s (except the upper ones o f the c o l u m n s still stan- d i n g t o their full h e i g h t ) and all the floor slabs o f the front hall.

Fig. 5 - Lefka-Bur-Dag, detail of dolomitic boudin, south shore of Lake Bafa.

W a l l s and stairs were studied as w e l l , b u t in less detail. T h e p e - trographic analyses o f b o t h the quarry samples and the temple's architectural m e m b e r s has been l i m i t e d to the macroscopic (hand lens) scale since the m e t h o d had to be applicable even where s a m - ples for microscopic studies could n o t be taken.

A s a third scientific m e t h o d , rare earth elements ( R E E s ) were analysed on samples from b o t h the A r c h a i c and the H e l l e n i s t i c - R o m a n temples, w i t h the hope that this m e t h o d w o u l d h e l p to solve some r e m a i n i n g uncertainties (on these see below). T h e s e analyses were carried o u t b y Peter M o l l e r and Peter D u l s k i at the G e o f o r s c h u n g s z e n t r u m ( G F Z ) , P o t s d a m , G e r m a n y . Unfortunately, the C I - n o r m a l i z e d R E E d i s t r i b u t i o n patterns for marbles k n o w n to be o f different origins (e.g., Lake Bafa m a r b l e from the H e l l e n i s t i c - R o m a n t e m p l e and marble f r o m the A e g e a n island o f N a x o s from the Archaic t e m p l e ) matched q u i t e closely in some cases whereas in other cases marble samples f r o m a k n o w n single source (e.g., marble from L e f k a - B u r - D a g quarries w i t h its d o l o m i t i c b o u d i n s ) produced a wide range o f patterns.

T h i s result can probably be explained by the fact that, in gener- al, rather pure marbles from the innermost parts o f marble lens- es o f any origin have very similar R E E contents. T h e closer to the contact w i t h the surrounding schists and gneisses the samples are taken, the m o r e 'contaminated' they are and the m o r e varied the R E E patterns become. Since the t e m p l e marbles were quarried from b o t h the central and marginal portions o f marble lenses o f various provenances, R E E analysis turned out not to be helpful for the d e t e r m i n a t i o n o f their provenance.

Interpretation o f data for architectural m o n u m e n t s m u s t , on account o f the l i m i t s to scientific analyses noted above, opera- te w i t h suppositions o f practicability and historical probability, w h i c h , if accepted, can at the same t i m e support the interpreta- tion. A c c o r d i n g l y , interpretations in this study were based o n the f o l l o w i n g considerations.

(1) B o t h inscriptions and the quarries w i t h their r e m a i n i n g archi- tectural pieces and blank forms c o n f i r m that large a m o u n t s o f

Fig. 6 - Didyma, Temple of Apollo, threshold of the great portal show- ing dolomitic boudin typical of Milesian marble from the Lefka-Bur- Dag quarries.

O

i

t o

1

0 I i 1 1 1 1 1 1 r — i i — i 1

- 8 - 6 - 4 - 2 0

5 18 0

Fig. 7 — Lake Bafa quarry samples (outlined Herakleia field after HERZ, 1987): east shore quarries: • = quarry I, H = quarry II, • = quar- ry IV, A = quarry; south shore quarries: D = western part; o = Lefka- Bur-Dag, A = north shore quarry.

O

i

CO C O

0 I i i i 1 1 1 i l 1 1 r — — r — — t — i — — i 1

- 8 - 6 - 4 - 2 0

Fig. 8 - Temple marbles from Lake Bafa: • = temple samples; quarry samples: o = south shore; D = east shore; A = north shore.

marble were taken f r o m the quarries around Lake Bafa, w h i c h was nearby and accessible by boat. In a d d i t i o n , those on the south shore belonged to the owner o f the sanctuary, the city o f M i l e t u s . Hence, it is appropriate to attribute any marble used at the t e m - ple w i t h corresponding petrographic and isotopic characteristics to this source.

(2) For large m o n u m e n t s like a temple, it is h i g h l y unlikely that single blocks o f marble were ordered f r o m a source otherwise not used. T h u s , an additional marble source can o n l y be assumed i f a large n u m b e r o f architectural elements share a u n i q u e set o f p e t - rographical and geochemical criteria.

(3) For the large quantities o f marble needed for a temple, any quarry accessible by boat is more likely to be the marble source

than one that required l o n g overland transport, because such transport is m u c h m o r e expensive than transport by boat.

A n a l y t i c a l r e s u l t s

Quarry samples from Lake Bafa display isotope signatures that mostly plot into a distinct area covering and slightly exten- d i n g the isotopic field published by Herz ( 1 9 8 7 ; fig. 6; data from Table 1). Samples from the largest east shore quarries, nos. I and V o f P e s c h l o w - B i n d o k a t ( 1 9 8 1 : fig. 6 4 ) show l i m i t e d variation and have the highest 6l sO values. Ratios for the easternmost south shore quarries o n L e f k a - B u r - D a g (PESCHLOW-BINDOKAT, 1981: fig. 5) similarly plot into a small field i m m e d i a t e l y adjacent to the former one. However, so far, the marbles from the north, east, and south shore quarries are not generally distinguishable by isotopic analysis, but are, at least in part, by petrographic criteria.

M o s t conspicuous is the marble f r o m the easternmost south shore quarries o n L e f k a - B u r - D a g . In m a n y places, it displays the characteristic d o l o m i t i c b o u d i n s described above (figs. 5 and 6;

s h o w n in dark b l u e in fig. 1). A s stated already, d o l o m i t i c por- tions w i t h i n the east shore marbles are less abundant and show completely different features. M i d d l e to dark grey marble is par- ticularly characteristic o f the westernmost south shore quarries ( s h o w n in grey in fig. 1) whereas on L e f k a - B u r - D a g and in the east shore quarries only banded varieties occur, w h i c h are also m u c h lighter in colour. W h i t e marbles w i t h o u t d o l o m i t i c b o u d i n s , s h o w i n g m e d i u m crystal sizes o f 0.3 t o 0.5 m m and m a x i m u m crystal sizes o f about 1.0 m m , occur in all Lake Bafa quarry areas (shown in dark green and bluish green in fig. 1).

H o w e v e r , the east shore quarries also feature a more coarse crys- talline variety not found on the south or north shores. Here, crys- tal sizes often vary w i d e l y w i t h i n a single sample. T h e m e d i u m crystal sizes are generally 0 . 8 to 1.0 m m (in rare cases u p to 1.5 m m ) w i t h m a x i m u m crystal sizes o f 2.0 m m and locally 2.5 m m or, rarely, larger (shown in l i g h t green in fig. 1). Figure 7 shows the isotopic signatures o f t e m p l e samples that are attributed to Lake Bafa quarries, on both isotopic and petrographical grounds, p l o t t e d against the quarry samples (data f r o m Table 1 and 2).

T h e extremely coarse crystalline marble, noticed previously by Peschlow-Bindokat ( 1 9 8 1 ) and G e r m a n n ( 1 9 8 1 ) is generally w h i t e w i t h grey 'clouds' and comes from A l i k i on the southeast shore o f the Aegean island o f Thasos {fig. 2). O n l y three temple samples could be taken but their signatures all plot distinctly away from the Lake Bafa field and well into the field of Thasos A l i k i (fig.

9, data from Table 3 , shown in yellow in fig. 1). Calcite crystals vary w i d e l y in size but range mostly between 1.5 and 3-0 m m w i t h conspicuous single crystals o f u p to 7 . 0 m m , a characteristic typical o f A l i k i marble (SODINI et al., 1980; HERZ, 1988) but u n - k n o w n to any other marbles w i t h overlapping isotopic fields.

Surprisingly, another previously unrecognised marble source was found. Petrographically, this marble is most o b v i o u s l y dis- tinguishable by its b a n d i n g in various shades o f w h i t e and grey.

For the most part these bands are m u c h m o r e distinct from one another than those seen in the streaky variants o f the Lake Bafa marbles and c o n t i n u e laterally w i t h a constant thickness, a fea-

ture w h i c h is most noticeable in the d o l o m i t i c bands ( s h o w n in red in fig. 1). D u e to the interlocking micro-texture, fractures in this 'new' marble also result in m u c h sharper edges. Crystal size varies m a i n l y between 0.5 and 0.8 m m w i t h m a x i m u m crystal size rarely exceeding 2.0 m m . T h u s , the 'new' marble tends to be slightly coarser grained than most Lake Bafa marbles but this cannot serve as a distinctive criterion. Isotopic signatures o f the

o

CO

t o

Thasos Vathi Thasos Aliki

Thasos Phaneri Lake Bafa

5 ¹⁸ 0

Fig. 9 — Temple marbles from Aliki on the Island of Thasos (Thasos fields after Herz, 1987, Lake Bafa field from this study): • = three temple samples with the rightmost symbol representing two identical ratios.

16 samples taken f r o m architectural elements displaying these petrographic characteristics p l o t into a field partly overlapping the Lake Bafa field but also extending well beyond it (fig. 10, data from Table 4). Provided that these banded marbles have a c o m m o n origin — an assumption h i g h l y probable on b o t h practi- cal and historical grounds as well - the signatures suggest a provenance from either D e n i z l i or the Island o f Marmara (ancient Proconnesus) (fig. 10). However, according to M o n n a and Pensabene ( 1 9 7 7 ) , none of the D e n i z l i varieties o f marble show the characteristic b a n d i n g displayed b y the temple samples whereas this feature is well k n o w n from Proconnesian marble.

O n practical and historical g r o u n d s as well, Marmara is by far the more probable source. T h e quarries near the modern t o w n of D e n i z l i lie about 2 0 0 k m inland from the Aegean coast and east o f D i d y m a (fig. 2). Its marble could only have been shipped on the Lykos and Menderes rivers, w h i c h w o u l d have been naviga- ble only during late winter s n o w m e l t and the peak o f the rainy season. H e n c e , it is generally assumed that, except for sarcopha- gi and perhaps some sculpture, the D e n i z l i quarries were only exploited for local use. In contrast, the quarries on the Island of Proconnesos in the Sea o f M a r m a r a (fig. 2) lie directly at the seashore. T h e y were imperial property and worked by slaves and state prisoners. T h e marble's abundance and g o o d quality made ancient Proconnesus the most productive exporter o f marble in the R o m a n E m p i r e (ASGARI, 1978).

H i s t o r i c a l i m p l i c a t i o n s

T h e m a p p i n g o f marble variants on the t e m p l e at D i d y m a itself (fig. 1) has m a j o r implications for both the sanctuary's e c o n o m y and b u i l d i n g history, w h i c h w e can o n l y s u m m a r i z e briefly in this paper (for a m u c h fuller discussion see BORG, 2001). Q u a n t i t i e s o f i m p o r t e d marbles are m u c h higher and were taken f r o m sources m u c h m o r e distant than previously assumed. T h e s e importations were o b v i o u s l y restricted to cer- tain periods and, except for rhe R o m a n period, were used to c o m p l e m e n t (as opposed to substitute) the Milesian Lake Bafa south shore marble. F r o m the b e g i n n i n g o f b u i l d i n g activities until 2 2 4 / 2 3 B . C . , w h e n marble from the L e f k a - B u r - D a g q u a r - ries was used for the threshold of the great portal (fig. 6), the sanctuary seems to have e m p l o y e d o n l y Milesian marble.

A f t e r a hiatus in the epigraphical evidence, w h i c h seems to represent a hiatus in b u i l d i n g activities as well, w o r k continued after 190/89 B . C . w i t h the erection o f the great portal's rivets o f Thasian m a r b l e and o f the dodekastylos columns. These c o l u m n s clearly s h o w a rather unsystematic alteration o f d r u m s o f Milesian and Thasian marbles, suggesting that the Thasian m a r - ble was i m p o r t e d to speed u p b u i l d i n g activities. O n l y one to t w o groups o f labourers could work at the same t i m e in most o f the small Milesian quarries. Furthermore, the quality o f the Milesian m a r b l e m a y not have been considered sufficiently h i g h for large pieces like c o l u m n d r u m s and rivets. T h e i n h o m o g e - neous calcitic-dolomitic marble from L e f k a - B u r - D a g is hard to w o r k and m u c h prone to fracturing whereas the grey varieties display a less satisfactory colour. J u s t before the dodekastylos was finished the sanctuary stopped i m p o r t i n g marble from Thasos, and in its place i m p o r t e d marble from nearby Herakleia.

A f t e r a second hiatus in the later Hellenistic period, b u i l - d i n g activities started again in the early R o m a n era and were p r o - bably sponsored by the R o m a n emperor(s). B e g i n n i n g w i t h the

o

CO

"to ₄

3 ^ 2

1

Denizli

Lake Bafa

Marmara

-2

5 ¹⁸ 0

Fig. 10- Temple marbles form Marmara (Marmara & Denizli fields after Herz, 1987; Marmara quarry samples after MANFRA et al., 1975, and AsGARi and MATTHEWS, 1995; Lake Bafa field from this study):

A = temple samples; X = Marmara quarry samples.

outer front colonnade o n the east side, the workers set o u t t o c o m p l e t e t h e outer colonnades o n all four sides w i t h o u t b o t h e r - i n g a b o u t t h e still unfinished inner colonnades. Starting o f f b y u s i n g M i l e s i a n m a r b l e for the front c o l u m n s in t h e m i d d l e , they soon switched to Proconnesian m a r b l e , w h i c h h a d the advantages o f easy a n d relatively economical accessibility b u t also o f h i g h quality. O n l y some o f t h e p l i n t h s seem t o consist o f Herakleian marble. I f this identification is correct, t h e p l i n t h s m a y have been laid o u t at the b e g i n n i n g o f the b u i l d i n g activities together w i t h t h e centre-front c o l u m n s , b u t the possibility cannot yet be ruled o u t c o m p l e t e l y that these p l i n t h s consist o f a different variety o f Proconnesian marble.

T h e s e observations also t h r o w n e w l i g h t o n the history o f the Lake Bafa quarries. I t has been supposed that these quarries had been a b a n d o n e d a b r u p t l y w h e n the G o t h s sacked the area i n 2 6 2 A . D . a n d were hardly ever used afterwards (PBSCHLOW- BINDOKAT, 1981). B u t three considerations argue against this view. (1) I t has been supposed that t h e u n f i n i s h e d c o l u m n d r u m s had cracked a n d broken apart as a result o f weathering, b u t the d a m a g e happened as a result f r o m inadequate h a n d l i n g o f t h e Milesian m a r b l e w i t h its internal i n h o m o g e n e i t i e s and natural j o i n t i n g . (2) T h e substitution o f M i l e s i a n m a r b l e b y P r o c o n - nesian m a r b l e as b u i l d i n g material for the t e m p l e o f A p o l l o , in the I m p e r i a l period, m u s t be due t o dissatisfaction w i t h the q u a l - i t y o f t h e local m a r b l e since t h e L a k e B a f a quarries were n o t all exhausted. ( 3 ) W i t h i n t h e earlier concept, it seems h a r d t o e x p l a i n w h y , after the G o t h s had left the area, w o r k i n the H e r a - kleian quarries w e n t o n (cf. the price edict b y Diocletian: MONNA and PENSABBNE, 1 9 7 7 ) whereas it was n o t taken u p again i n the south shore quarries. T h u s , it seems h i g h l y likely, that t h e Milesian south shore quarries were m o r e or less g i v e n u p i n t h e early Imperial period because o f the b a d q u a l i t y o f their marble.

A c k n o w l e d g e m e n t s

T h e authors are m u c h i n d e b t e d t o the director o f t h e exca- v a t i o n at D i d y m a , K l a u s T u c h e l t , for i n v i t i n g a n d s u p p o r t i n g their project; Lothar Haselberger, i n charge o f t h e H e l l e n i s t i c - R o m a n t e m p l e , for his e n c o u r a g e m e n t a n d advice; and C h r i s - t o p h K r o n e w i r t h for t a k i n g t h e samples d u r i n g restoration. O u r thanks also g o t o H a r a l d Strauss w h o carried o u t the isotope analyses, a n d Peter M o l l e r a n d Peter D u l s k y for carrying o u t a n d h e l p i n g t o interpret t h e R E E analyses. T h e paper benefited m u c h f r o m c o m m e n t s b y the reviewers a n d b y J a m e s A . Harrell i n particular. Finally, t h e authors w o u l d l i k e t o express their gratitude t o K l a u s T u c h e l t a n d t h e G e r m a n Archaeological Institute for their m a j o r financial s u p p o r t for the field c a m - paigns in D i d y m a a n d for t h e scientific analyses.

R e f e r e n c e s

ASGARI N . , 1978, R o m a n and early Byzantine marble quarries o f Proconnesos. Proceedings of the X International Congress of Classical Archaeology A n k a r a - I z m i r 1973, T u r k Tarih K u r u m u Basimevi, Ankara, 4 6 7 - 4 8 0 .

ASGARI N . and MATTHEWS K . J . , 1995, T h e stable isotope analysis of marble from Proconnesus, in: MANIATIS Y , HERZ N . and BASIAKOS Y.

(eds.), T h e study o f marble and other stones used in antiquity.

A s m o s i a I I I A t h e n s : Transactions o f the 3 rd International S y m p o s i u m o f the Association for the Study of Marble and Other Stones used in A n t i q u i t y , London, 1 2 3 - 1 2 9 .

BORG B.E., 2 0 0 1 , Marmor fur A p o l l - Ein Beitrag zur Baugeschichte des jiingeren D i d y m a i o n und der historischen Topographie seiner U m g e b u n g , BERGEMANN J . (ed.), Wissenschaft mit Enthusiasmus. Bei- trage zu antiken Bildnissen und zur historischen Landeskunde, Klaus Fittschen gewidmet, Marie Leidorf, Rahden/Westfahlen, 7 9 - 1 0 1 . CRAIG H . , 1957: Isotopic standards for carbon and oxygen and cor- rection factors for mass spectrometric analysis of carbon dioxide, Geochimica et Cosmochimica Acta, Vol. 12, 1 3 3 - 1 4 9 .

CRAIG H . and CRAIG V., 1972, Greek marbles: determination of provenance by isotopic analysis, Science, V o l . 176, 4 0 1 - 4 0 3 . DORR S., 1975, Ober das Alter und geotektonische Stellung des Menderes- Kristallins/S W-A natolien und seine Aquivalente in der mittleren Agais (Habilitationsschrift Marburg, Lahn).

GERMANN K . , 1981, Lagerstatteneigenschaften und herkunftstypis- che M e r k m a l s m u s t e r v o n M a r m o r e n a m Siidwestrand des Menderes-Massifs (Siidwestanatolien), Jahrbuch des Deutschen Archaologischen Instituts, Vol. 96, 2 1 4 - 2 3 5 .

GONTHER W . , 1969/70, Eine neue didymeische Bauinschrift, Istanbuler Mitteilungen, Vol. 19/20, 237-247.

HERZ N . , 1987, Carbon and oxygen isotopic ratios: A data base for classical Greek and Roman marble, Archaeometry, Vol. 29(1), 3 5 - 4 3 . HERZ N . , 1988, Classical Marble Quarries o f Thasos, WAGNER G . A . and WEISGERBER G . (eds.), Antike Edel- und Buntmetallgewinnung auf Thasos. Der Anschnitt, Beiheft 6, Deutsches Bergbaumuseum, Bochum, 2 3 2 - 2 4 0 .

KNACKFUB H . , 1941, Die Baubeschreibung in drei Banden, TH.

WIEGAND (ed.), D i d y m a I., Berlin,

Loos S. and REISCHMANN TH., 1999, T h e evolution of the southern Menderes Massif in S W Turkey as revealed by zircon dating, Journal of the Geological Society, V o l . 156, 1 0 2 1 - 1 0 3 0 .

MANFRA L., MASI U . and TURI B . , 1975, Carbon and oxygen isotope ratios o f marbles from some ancient quarries o f Western Anatolia and their archaeological significance, Archaeometry, 17.2, 2 1 5 - 2 2 1 . MONNA D . and PENSABENB P., 1977, Marmi dell'Asia Minore,

Consiglio Nazionale delle Ricerche, R o m a .

PESCHLOW-BINDOKAT A . , 1911 IIS, Ioniapolis. Z u r Topographie einer milesischen Hafenstadt am latmischen G o l f , Istanbuler Mitteilun- gen, V o l . 27/28, 131-136.

PESCHLOW-BINDOKAT A . , 1981, D i e Steinbriiche von Milet und Hera- kleia am Latmos. Jahrbuch des Deutschen Archaologischen Instituts, Vol.

96, 1 5 9 - 2 1 3 .

R E H M A . , 1 9 5 8 , Die Inschriften, T H . W I E G A N D - A . R E H M - R . HARDER (eds.), D i d y m a II, Berlin.

RUMSCHEID E , 1994, Untersuchungen zur kleinasiatischen Bauornamentik des Hellenismus, Mainz.

SCHNEIDER P., 1996, Z u m alten Sekos von D i d y m a , Istanbuler Mitteilungen, Vol. 4 6 , 1 4 7 - 1 5 2 .

SENGOR A.M.C., SATIR M . and AKKOK R., 1984, T i m i n g of tectonic events in the Menderes Massif, western Turkey: implications for tec- tonic evolution and evidence for Pan-African basement in Turkey, Tectonics, Vol. 3.7, 693-707.

SODINI J . - P . , L A M B R A K I A . , a n d K O Z E L J T . , 1 9 8 0 , L e s c a r r i e r e s d e marbre a l'epoque paleochretienne, Htudes Thasiennes. Vol. 9- Aliki. I, Ecole Francaise d ' A t h e n e s , Athenes, 7 9 - 1 3 7 .

TUCHBLT K . , 1973, Vorarbeiten zu einer Topographie von Didyma. Eine U n t e r s u c h u n g der inschriftlichen u n d archaologischen Zeugnisse, Beiheft Istanbuler M i t t e i l u n g e n V o l . 9, T u b i n g e n .

TUCHELT K . , 1 9 9 1 , Branchidai - Didyma. Geschichte, Ausgrabung und Wiederentdeckung eines antiken Heiligtums, 1765 bis 1990, A n t i k e W e l t , Sondernummer, M a i n z .

VOIGTLANDER W . , 1975, Der jiingste Apollontempel von Didyma. Geschichte seines Baudekors, Beiheft Istanbuler Mitteilungen V o l . 14, Tubingen.

WIEGAND TH., 1913, DerLatmos, M i l e t I I I 1, Berlin.

T a b l e 1 — Quarry samples f r o m Lake Bafa region (quarry numbers

Provenance Sample

no.

6l sO (%o. P D B )

6I 3C (%o. P D B )

east shore, quarry I 96540 -2.60 2.52

96541 -2.09 2.35

96542 -2.26 3.00

96544 -2.73 2.14

96548 -1.74 2.61

96558 -1.98 2.61

96559 -2.55 2.33

96564 • -3.34 1.60

east shore, quarry II 96543 -2.43 2.82

96545 -3.10 3.04

96546 -2.36 2.90

96551 -2.79 2.80

96553 -2.93 2.71

96555 -2.37 2.83

96562 -2.23 2.79

96563 -1.81 2.79

east shore, quarry I V 98507 -3.7 1.5

98508 -3.5 1.5

99519 -2.77 1.50

east shore, quarry V 99528 -1.80 2.68

99529 -1.63 2.57

99530 -2.43 2.85

south shore, east 97501 -2.73 2.43

97502 -2.62 2.61

97503 -1.96 2.58

97504 -2.11 2.27

97504 -1.73 2.27

97505 -2.09 2.29

9750 -2.30 2.18

after Peschlow-^Bindokat, 1981)

Provenance Sample

no.

6I SO (%o. P D B )

5I 3C (%o. P D B )

south shore, west 97507 -2.51 2.32

98501 -2.6 2.7

98502 -2.0 2.2

98503 -2.0 2.5

98504 -2.6 2.4

98505 -2.2 2.2

98506 -2.7 2.2

98542 -3.0 2.5

98543 -3.1 2.7

south shore, west 97508 -2.83 2.29

97509 -2.82 2.41

97510 -2.75 3.83

97511 -6.62 1.04

97512 -5.45 1.66

97513 -3.52 2.19

97514 -3.59 2.25

97515 -1.99 1.74

97516 -3.97 2.09

98510 -3.4 2.1

98511 -3.3 2.4

98512 -3.6 1.8

98513 -6.7 2.0

98514 -2.8 1.3

98514 -2.8 1.3

98515 -2.5 3.1

north shore 99520 -3.09 1.86

99522 -2.25 1.92

99523 -2.78 2.00

99524 -5.15 2.09

T a b l e 3 — T e m p l e samples attributed to Thasos. A l i k i

Sample no.

5I 80 (%o. P D B )

81 3C

(%c. P D B ) Provenance f r o m T e m p l e

L H 2 - 0 . 7 3.2 great Portal, south rivet

9 8 5 2 2 - 0 . 2 3.1 c o l u m n C 4 , second d r u m

9 8 5 2 3 - 0 . 2 3.1 great Portal, north rivet

T a b l e 2 — T e m p l e samples attributed to Lake Bafa Quarries

Sample no. 6I 80 (%o. P D B ) 6I 3C ( f c . P D B ) Provenance f r o m T e m p l e

L H 3 -1.8 2.8 west wall, 2-column-hall

L H 4 -3.5 1.7 south anta

L H 6 -1.9 1.9 column C 9, first drum

L H 7 -1.6 2 stepped Platform, north-west corner

tm -2.7 2.2 stepped Platform, front

96547 -2.62 2.44 column E 7, third drum

96549 -1.87 2.58 column E 7, fourth drum

96550 -2.14 2.27 column E 7, third or fourth drum

96552 -2.64 2.06 column E 7, third or fourth drum

98509 -3.4 1.7 north anta

98516 -2.9 2.7 column C 9, second drum

98517 -2.5 2.0 column C 9, first drum

98519 -3.6 1.6 column A 4, second drum

98521 -2.6 1.0 north anta

98527 -2.8 1.8 adyton, north wall

98528 -1.6 2.7 adyton, west wall, first pilaster

98529 -3.8 1.6 naiskos

98530 3.0 2.1 naiskos

98540 -2.5 2.6 column D 4, second drum

98544 -2.3 2.6 column D 4, second drum

98545 -2.1 2.5 column D 4, third drum

99539 -3.44 3.51 column F 1, plinth

99547 -2.44 1.32 column B 6, base

99548 -3.07 2.13 naiskos

99549 -2.50 2.44 column B 7, second drum

99550 -2.93 2.33 adyton, north wall

99551 -2.32 1.50 column B 6, first drum

99552 -1.98 2.01 column B 7, second drum

T a b l e 4 — T e m p l e samples attributed to Marmara

Sample no. d^{l s}O (%o. P D B ) 813C (%o. P D B ) Provenance f r o m T e m p l e

LH1 -1.3 3.3 column F 1, base

LH5 -2.1 3.6 column B 10, first drum

98518 -2.6 2.0 column A 2, first drum

98520 -2.0 3.6 column A 9, first drum

98524 -1.6 3.4 column D 1, plinth and base blok

98525 -2.1 2.8 column G 1, third drum

98526 -1.8 3.0 column I 1, plinth and base blok

99536 -1.64 3.79 column B 1, first drum

99537 -1.58 3.81 column D 1, first drum

99538 -2.05 2.47 column E 1, plinth and base blok

99540 -11.19 3.28 column F 1, spira

99541 -1.24 3.07 column H 1, base

99542 -2.53 2.70 column H 1, first drum

99543 -2.06 2.84 column Q 1, first drum

99544 -3.27 3.20 column Q 1, plinth and base blok

99545 -1.63 2.92 column C 1, first drumr