Architecture on the Roman Coins of Alexandria Author(s): Susan Handler Source: American Journal of Archaeology, Vol. 75, No. 1 (Jan., 1971), pp. 57-74 Published by: Archaeological Institute of America Stable URL: http://www.jstor.org/stable/503682 . Accessed: 22/06/2011 07:14 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=aia. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Archaeology. http://www.jstor.org Architecture on the Roman Coins of Alexandria* SUSAN HANDLER PLATES 11-12 Depictions of architecture on coins have been noticed for over a hundred years, from the appearance of the pioneering work of T. L. Donaldson,1 but little careful work has been done with the subject.2 The only two major studies, both dealing with temples on coins, were done in the 1940's. More recent investigations have generally been shorter pieces on specific problems.4 Despite the limitations imposed by the small size of numismatic representations, and resulting lack of detail, coins if properly used can yield useful information sometimes obtainable from no other source. This is the case with the architecturalcoins of Alexandria in Egypt, which provide a valuable record of that city's ancient architecture. Alexandria's well-known Hellenistic and Roman monuments are now virtually inaccessible to excavation, since the modern city and the rising sea level have combined to cover the ancient remains.5 Alexandrian buildings of both the Hellenistic and Roman periods are, however, depicted in surprising variety on the large bronze drachma and half-drachma coins minted at Alexandria during the period of Roman rule; they appear on the coinage from the reign of Galba through that of Marcus Aurelius.6 Although the coins of many cities in the Roman empire portray buildings, the architectural types of Alexandria are more varied and numerous than those of almost any other provincial mint.' An extensive study of some thousand extant examples' has produced new and interesting information on individual buildings and on Alexandrian architectural style in general. Before considering specific buildings, something should be said of the problems of dealing with * Begun at the American Numismatic Society summer seminar in 1963, this paper was read in a later version at the 1966 annual meetings of the AIA. The present article forms part of a Ph.D. dissertationaccepted by Bryn Mawr College. I wish to thank the staff of the American Numismatic Society, particularly Miss Thompson and Miss Fagerlie, for their assistance. A dissertation fellowship from the AAUW in 1965-66 greatly aided the research. I am very grateful to my Bryn Mawr professors for their help and advice, especially Professor Ridgway. I acknowledge with thanks permission to publish coins from the following: Graeco-Roman Museum, Alexandria (figs. 13, 22); the American Numismatic Society (fig. 2-4, II, 12, 14, 15, 19, 29); Staatliche Museen Berlin (fig. 27); Koninklijk Penningkabinet, The Hague (fig. 20); Manchester University Museum (fig. 9); Bibl. Nationale, Cabinet des Medailles, Paris (figs. 28, 30); Mr. John J. Slocum (fig. I6); Bundessamm. v. Medaillen, Miinzen u. Geldzeichen, Vienna (figs. 8, Io); numbers appearing after coins are those assigned to them by the coin cabinets, or where available, published catalogues. 1 ArchitecturaNumismatica, or architecturalmedals in classical antiquity (London I859, repr. Argonaut Press, Chicago 1966). 2Works dealing in part with architectural coins, such as F. Imhoof-Blumer and P. Gardner, A Numismatic Commentary on Pausanias, JHS 6 (I885) 50-191; 7 (I886) 57-II3; 8 (1887) 6-6o; reprinted by Argonaut Press (Chicago 1966), are erratic in method. The articles of H. Dressel, written early in this century, show the first use of a systematic methodology. 3 Donald Brown, Temples of Rome as Coin Types. ANSNNM 90 (1940). Bluma Trell, The Temple of Artemis at Ephesos. ANSNNM 10o7 (I945). 4Among articles written after the monographs of Brown and Trell should be noted Marilyn Stokstad, "Architecture on the coins of Nero," NumCirc 62 (I954) col. 389-436; D. Woods, "A numismatic chapter in the Romanization of Hispania," Essays in Memory of Karl Lehmann, Marsyas, supp. I (New York 1964), and Th. Drew-Bear, "Some temple types on Greek imperial coinage," unpublished summer seminar paper, American Numismatic Society, Aug. 1967 (hereafter ANS). 5 An exception is the work done by Polish excavators on K8m-ed-Dik, the hill very near the original central square of ancient Alexandria. Levels down to Ptolemaic have been found, but the buildings uncovered are of the Roman period. References to excavation reports in A. Adriani, Repertorio d'Arte dell'Egitto Greco-Romano ser. C. vol. I (Palermo 1966) 89. Hereafter Adriani, Rep. C. 6 Augustus' Alexandrian coins follow his practice elsewhere in provincial coinage of depicting buildings at Rome-the temthan ple of Mars Ultor and the Parthian victory arch-rather local structures. 7 This is particularly true by comparison with the coins of the cities of Asia, which confine their representations almost exclusively to their Roman Neocorate temples in preference to older monuments or secular buildings. 8 These coins were studied in the following collections and museums: Alexandria, Graeco-Roman Museum; ANS, New York; coin cabinets of Athens, Demetrio Collection; Berlin (Staatliche Museen); Brussels (Bibl. Royale); Cairo; Frankfurt; University of Michigan (Kelsey Museum); Oxford (Ashmolean Museum); Paris (Bibl. Nationale); Rome (Terme); Vatican; Vienna; collection of Mr. John Slocum. Coins examined from casts or photographs: Dattari, Numi Augg. Alexandrini. Catalogo della Collezione G. Dattari (Cairo 90oI) hereafter Dattari; Dresden; Fitzwilliam Museum, Cambridge (with kind help of Bluma Trell); Fogg Museum; Hannover; Karlsruhe; Leningrad; Fabretti, Langone, Rossi, Regio Museo di Torino, Monete Greche (Torino 1883); Toronto. INTRODUCTION 58 SUSAN HANDLER [AJA 75 architectural representations on coins. Skepticism has been a feature from the beginning, for example as early as Imhoof-Blumer and Gardner's Numismatic Commentary on Pausanias of I885-I887.' The "trustworthiness"of coin evidence was suspected here because of the apparent variations, particularly in the number of facade columns,"1in individual coins depicting the same building. However, Dressel in 19o6 was the first to observe that abbreviation of columns was a standard feature of numismatic representations."1On this point, then, it is misleading to expect the coins to yield precise information. At Alexandria, for example, columns are generally reduced to a minimum of two or four, to make room for a clearer representation of cult statue and architectural order. Careful methods are needed to utilize the evidence of these abbreviated representations. Early work was based on only one or two examples of each coin type. Brown and Trell, however,12realized that maximum information could only be obtained from systematic comparison of a series of coins depicting a single building. This is particularly true for Alexandria, where the coins are dated by the regnal year of the emperor. Since the architectural representationappears along with the year on the reverse, the whole design, numerals and architectural depiction, was changed annually."3 Aspects of a building left out or abbreviatedin the older design were often emphasized in the new one. The longer the series of coins available, the greater chance there is for additional architectural features to appear. In order to indicate the amount of evidence available in each instance, the extant years of issue of the coin types will be noted in discussing specific Alexandrian monuments. The structures seen on this coinage can in almost all cases be identified as local monuments by the deities which they house and by their architectural style. The coin representations of most cities can be checkedagainstexistingremains.14 Such is not the case at Alexandria,save for the Pharos,which can be reconstructedfrom ancient and mediaevaldeIn this inscriptionsand architecturalparallels.15 stance the Pharoscoins illustratethe kind of informationavailablefromnumismaticevidence.The threestoreysof the lighthousestructureare clearly depictedin many examples,althoughtheir shape is oftenblurredby wear.Otherexamplesreducethe stagesto two in orderto depictthe sculpturaldecorationin more detail.The coins here, as in other instances,reproducethe generalappearanceof the monument,but not its proportions,material,or ornamentaldetail. Clearly,then, these representationson coinscannotstandaloneas reconstructions of lost monuments.Checkedand amplifiedagainst ancient descriptions,architecturalremains and small art objects,they suggest the appearanceof specific buildings and of an architecturalstyle which would otherwiseremainunknown. Some of the most importantAlexandrianarchithose depicting tecturalcoins will be discussed,"6 the Pharos, the Isis-Harpocratesshrine, the Isis pylon, the Canopi shrine, the Sarapaeum,the arches. AgathosDaimon altarand commemorative 9 Supra n. 2. Imhoof-Blumer (reprint supra n. 2) 27. 11 H. Dressel, "Der Matidia Tempel auf einem Medaillon des Hadrianus," Corolla Numismatica. Essays in honor of B. V. Head (Oxford 1906) 21-22. 12Supra n. 3. 13 The individual dies used in minting the coins were of course replaced during the year as they wore out. A similar design scheme was, however, generally followed throughout the year. 14 The coins of Rome, for example. Brown (supra n. 3); E. Nash, A Pictorial Dictionary of Ancient Rome 2 vols. (Berlin 1962). 15 See below, and nn. 19-26. 16 For a complete list of Alexandrianbuildings shown on the coins see Conclusion below. 17 Some think that the original Ptolemaic building may not have contained a warning night fire, since no fire is mentioned in descriptionsuntil the middle of the first century A.D. (R. J. Forbes, Studies in Ancient Technology VI [Leiden 19581 I80-181). 10 PHAROS It is not surprisingthat the Pharosis the only Alexandrianmonumentshownon the coins which from othersources.This first can be reconstructed true lighthouse17 was one of the seven wondersof the ancientworldl--the subjectof comment and descriptionby both ancient19and mediaevalwriters.20It stood, in a somewhatalteredstate, until the fourteenthcentury,and the line of its original foundationscan still be seen on the rockypromontory of Pharosislandfrom which its name derives (ill. I). One of the most helpful of the later authorsde- 18 Epiphanius, Adv. Haeres. Migne, PG 120, 265B. 19 Strabo, Geog. 17.1; Pliny NH 38-83; Josephus, Bell. Jud. 4.612-13. Poseidippos epigram, RhM 35 (1880) 258; REA I (1899) 261ff. 20 Collected and translated by Thiersch, Pharos Antike Islam und Occident (Leipzig and Berlin 1909) 35-64. Hereafter Thiersch, Pharos. See also n. 21 infra. ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA 1971] 0[(Fort •PiREIsTRI HARBOUR(Rassel TNirj TEMPLE OF POSEIDONM , 0& CATACOMBS jJ O 59 PHAROS LIGHTHOUSE KaitBet TEMPLE OF ISIS ONPHARDS PI4AI TIMONIUM ofArrfouchYi• OURn' HAR 0 UE GAREAT HARIB TEUNOSTOS ~ q$1ISLA GATEOF MOON7 L- .E m J(S• cK MASHRINE A EWoS RASites COURSES elChgafal ost CN.OF HE AR. J - in CAPTake NUVJEH .331 A D TEMPLE OFISIS DST• WOMPEY$S'<--.N--,,.THEAREit'oMCOFY" LAKE MAREOTI S SLE X ORLPAWE SANTIIRRHODOS A E0 ouScr CAcietBS HISTOICL tAND AH WILE AiicientSitesin CAPIITALS Modern Sites bracketed u ILL. I. From E. M. Forster, Alexandria, a History and a Guide (Anchor repr. 1961) 2-3 scribingthe Pharoswas a twelfth centurytraveler, Ibn al Shaik.21His careful measurementshave formed the basis for precise reconstructionof a building with tall foundation,high square first stage, octagonal middle stage and round upper storeywith the light.22Smallobjectsin glass23and terracotta24 also add to knowledgeof the structure. Lighthousesmodeledon the Pharosincludeone at TaposirisMagna along the coast west of Alexandria,25and Roman examples at Ostia, Ravenna, Syriaand Dalmatia.26 The coins corroboratewhat is known of the 21 M. Palacios di Asin, "The Pharos of Alexandria; summary of an essay in Spanish," ProcBritAc 19 (1933) 277-292. 22 Ibn al Shaik included in his measurementsan added fourth storey mosque. His measurementsfor the first three storeys agree with ancient accounts, and also give a logical system of proportions which is probably close to the original. (M. Palacios di Asin, "The Pharos," 290-291). 23 J. Hackin et al., Nouvelles Richerches Arche'ologiquesa no. 203, figs. Begram. MDAFA vol. 2 (Paris 1954) 259-262. zo0-1o2, 24 Lamp in Pharos form. Ev. Breccia,Monumentsde L'Egypte Pharos from other sources.One of the most popular types of the Alexandrian mint, Pharos coins were issued through six imperial reigns from Domitian through Marcus Aurelius,27and reappearedin year 29 of Commodus.28 The lighthouse is shown either alone (pl. i :i) or accompaniedby a striding figure of Isis Pharia, Isis as the protectressof sailors (pl. 11:2). The most careful of the Pharos coins ac- curately reproduce the three stages of the structure -tall square first stage with rows of windows, octagonal middle section, and round top storey with warning flame inside (pl. I1, fig. 3). HornGreco-Roman, vol. L, La Rovine e i Monumenti di Canopo (Alexandria 1925) pl. 40, 5. 25 Thiersch, Pharos fig. 49. 26 Thiersch, Pharos 22ff. 27Examples extant, Pharos alone: Domitian years 11-15,17, 18; Trajan years 11,14,15,16; Hadrian years 2,3,6,7,Io,II,I517,19-21; Antoninus Pius years 4-8,10-12,15-18; Marcus Aurelius year 16. Pharos and Isis: Trajan years 15,16, Hadrian years 16-18,21; Antoninus Pius years 2,3,5,8,10-12. 28 These are not bronze coins but small billion tetradrachms representingthe Pharos with a ship of the grain fleet. 60 SUSAN HANDLER blowing tritons lean outward from the top of the second stage, and a standing figure crowns the summit. Many of the coins omit the middle storey, in order to reproducethis sculpturalornament in some detail. The crowning statue as seen on the coins is of special interest: it appears to be a nude male figure holding a spear in his left hand, and a small round object, perhaps a phiale, in his right.29 This of course represents the summit decoration visible during the period of Roman rule when the coins were minted. A late Hellenistic representation of the Pharos has a variant depiction of this figure. Molded in relief on an Alexandrian glass beaker found at Begram in Afghanistan,30the summit figure stands in the same position as that on the coins, but holds in his left hand not a spear, but a steering oar. The variance can probably be explained by the earlier date of manufacture of the glass beaker, which may depict an original Pharos-crowning statue of a deified Ptolemy.1 It is not unlikely that this prominent symbol of Greek rule was changed when the Romans came to power, and the attributes of a more neutral figure, perhaps a Poseidon, substituted. A yet later change may be reflected in a sixth century A.D.North African mosaic of the Pharos, in which the summit figure is shown with the radiate headdress of Helios.32 The Isis seen alongside the Pharos on the coins (pl. I1:2) is identified by her characteristic tall headdress and sistrum. The striding goddess steadies with her upraised right hand a sail which billows in the wind. Folds of her windswept drapery blow out behind. The sail-steadying gesture is appropriate for the goddess credited with the invention of sails,33 and worshipped as protectress of ships. The epithet "Pharia" often applied to this form of Isis, probably refers to a cult place in her honor on Pharos island.34 29 In some examples, such as Hague 787, pl. II:i, the Pharos is shown in a mirror image reversal of its usual door-left position, and in these cases the statue is also generally reversed. 30 Supra n. 23. 31 Thiersch, Pharos 13. 32 F. Goodchild, "Helios on the Pharos,"AntJ 41 (1961) 218223. 33Referred to in Hyginus, Fabula 277 (p. 153, ed. M. Schmidt), Cassiodorus, Var. 5.17 (Knack, Hermes 16 [1881] 586). Roscher, Lexikon II,i, col. 474-475. 34Epithet in Statius, Silvae 3.2.202. With geographical connotation: Ovid, Amores 2.13.8. Ovid, Met. 9.773. Also Adriani, Rep. C, I 252. Calderini, Dizionario dei Nomi Geographici dell'Egitto Greco-Romano I, fasc. I (Cairo 1935) 118. Here- [AJA 75 A number of monuments make it clear that this figure of Isis often seen beside the Pharos represented an actual statue. The representationson the coins show Isis, seen from the right side, with left leg forward and bent, right leg braced and back, right arm raised to steady the front corner of the sail, and left arm down and holding the sail's farther corner. The head is in profile to the right. An overmantle covering the sleeveless chiton sweeps back behind the shoulder of the figure in a circle, and an end flutters below. The most important of the comparable representations is a fragmentarymarble statue, now 1.45 m. high, brought to Budapest from a villa near Naples (pl. 12, fig. 27).3 It is the only large-scale Isis Pharia figure which seems to have survived. The piece was tentatively identified by Heklar as a fleeing Niobid, and characterized as a Roman copy of a work of about 300 B.C.36 Closer inspection reveals a pose and drapery exactly comparable to the Isis seen on the coins. The figure is not fleeing but has her feet firmly braced on the ground, the left leg forward and bent, the right leg back. The mantle is draped acrossthe figure in sweeping horizontal folds which contrast with the vertical folds of the chiton beneath, a detail also clearly seen on the coin figures. A fragment of the back-sweeping circle of drapery remains on the Budapest figure,"3 as do attachment holes for a raised right arm and a lowered left one. The fringed mantle edges of the garment are characteristicof the dress of Isis. Unfortunately, the head has been lost. Isis Pharia reappears on many coins, on gems, and on two small works from Delos, a stone relief of the first century B.c., and a lamp from the second century A.D.38The pose of the figure on the Delos objects is identical to that of the Budapest statue, and includes the sail steadied by the goddess. The basic pose seen in the Budapest statue is also repeatedon the coins of a number of maritime after Calderini, Dizionario. 35I am grateful to Dr. J. G. Szilagyi of the Mus6e des Beaux Arts in Budapest for bringing to my attention this sculpture in his museum, and for sending the photograph reproduced in pl. 12, fig. 27. 36 A. Heklar, Museum der Bildenden Kiinste in Budapest, Die Sammlung Antiker Skulpturen (Vienna 1929) 63 pl. 51. Also reproduced in Brunn-Bruckmann's Denkmdler no. 640. Richter, Sculpture and Sculptors (1950) pl. 376, fig. 93, p. 64, sees the piece as a Roman copy of a 4th century B.c. work. 37 This is visible in photographs of the, left side of the figure. 8 Ph. Bruneau, "Isis P6lagia 'a Delos," BCH 85 (1961) relief: 437 fig. 3; lamp: 436 fig. i. See also Bruneau's "Isis ' Delos (Complements) BCH 87 (1963) 301-308. P6lagia 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA 61 cities," though it should be noted that some of these coins depict a ship with the goddess standing on the prow, a detail not seen in the Alexandrian examples. Variants in dress and pose can be seen in some Alexandrian Pharos coins with the Isis figure:40 the goddess wears only a chiton without a mantle, and stands with her head turned directly backward. The drapery flying out behind the figure seems illogical as compared with the other version, where the superfluous material of the outer garment is blown by the wind, and the backward twist of the head would be anatomically impossible. I believe that the die engraver in this case was not thinking literally of the statue, but rather of Isis Pharia "symbolically" turning back toward home as she guides a ship out of Alexandria harbor. The juxtaposition of the goddess' statue and the Pharos on the coins cannot be taken literally. Whether the Isis is a cult statue or an open-air monument such as the Nike of Samothrace cannot be known. Should the statue have originally stood out of doors, it would not have been visible standing on the Pharos rock, overshadowed by the lighthouse; it is more likely that it was placed at one end of the Heptastadion bridge which joined Pharos island to the mainland. Thiersch suggests that it stood on the end of the Heptastadion nearest to Eunostos harbor, where it could be clearly seen by departing ships."4 Two Roman terracotta lamps made at Alexandria42 allow for this possibility. Both have background designs of a curved harbor lined with monumental buildings, clearly the eastern harbor of Alexandria itself. In the foreground of one lamp are fishermen in small rowboats; on the other can be seen an arched causeway of seven arches, over The Alexandrian temples of Isis provide good examples of the Egyptian style of architectureseen in this city of mixed culture. Two types appear on the coins: an Isis temple pylon entrance and an Isis-Harpocrates shrine with an arched roof. Pylon of Isis. Coins depicting a pylon of Isis44 were issued under Trajan and Hadrian45(pl. 11:4). The pylon corresponds in its architectural form, and particularlyin the late feature of window openings,46 to temple entrances of Egyptian style built during the Ptolemaic era. Comparison with such a structure,the pylon of the temple of Horus at Edfu, explains the vertical lines flanking the central doorway of the coin representationsas flagstaves.47The figure of Isis which appears on the roof is perhaps the cult statue seen by participants in sacred rites conducted in roof chapels,"4or it could be explained as the epiphany or miraculous appearance of the goddess to her worshippers.49The Isis temple, of which this pylon was the entrance, would seem to have been a cult place of considerablesize, of traditional Egyptian architectural style. Isis-Harpocrates Shrine. Some examples of this 39 These cities are: Byblos, Kyme in Aeolis, Corinth, Nicomedia, Phocea, Anchialos in Thrace, Ephesos, Aspendos and Amastria (Bruneau, 440-441). 40 All extant Pharos and Isis coins of Trajan year 15. Also some of the Isis and Pharos examples of Hadrian year 21 (two in ANS collection). In these latter examples, Isis wears both chiton and mantle. 41 Thiersch, Pharos 13. 42 H. B. Walters, Catalogue of the Greek and Roman Lamps in the British Museum (London 1914) no. 522 with fishermen; no. 758 with Heptastadion. 43 There is a similar representation of the Heptastadion on an intaglio illustrated by Alfoldi, "Die Alex. Gotter und die Vota Publica am Jahresbeginn," Jahrbuch f. Antike und Christentum 8/9 (1965-66) pl. 2, fig. 6. 44 A recent article in Studia Hellenistica 16 (1968), "Le pyline Egyptien sur les monnaies imperiales d'Alexandrie," by Paul Naster, identifies the central figure as Isis save in one example (ANS Trajan year 12) where it is said to be Harpocrates (p. 187). Comparison with representationsof both Isis and Harpocrateson coins and minor art objects show that this identification cannot be correct. Harpocrates is almost always depicted clothed, with hand to mouth in characteristic pose (Dattari 3032), and with attributes such as a ram (Demetrio I619a). Also nothing in the, coin architecturesuggests that the single example cited portrays a different building from the other examples. 45Years 12,14,15 of Trajan, and years 19,20,21,24 of Hadrian. 46 S. Clarke and R. Engelbach, Ancient Egyptian Masonry (Oxford 1930) IIoff. 47 I. Noshy, The Arts in Ptolemaic Egypt (Oxford 1937) 68. 48 As at the temple of Hathor at Denderah. A. Erman, Die Aegyptische Religion 2nd ed. (Berlin 1909) 23449 M. P. Nilsson, Geschichte der Griec2ischen Religion II, 214ff. which a country man drives a donkey. At the left end of the causeway is a triumphal arch, its top adorned with a marine design of hippocamps. The causeway must represent the Heptastadion, the seven arches standing for its seven-stade length.43 The harbor is seen looking through the Heptastadion, thus from the west. Therefore the triumphal arch shown at its left end would be at the Pharos island end of the Heptastadion, and, if the depiction is correct,the mainland end of the Heptastadion would be free for the placement of the Isis Pharia as suggested by Thiersch. ISIS SHRINE AND PYLON 62 SUSAN HANDLER [AJA 75 type were issued under Trajan, but the majority of extant specimens come from the reigns of Hadrian, Antoninus Pius and Marcus Aurelius.50Represented is the facade of a shrine or temple on a two-stepped podium, with two bulbous columns crowned by Egyptian papyrus-flowercapitals. The pediment is in the form of a shallow, rounded arch. The columns are set on base moldings the form of which cannot be clearly seen, and in one example a sheathing of leaves is visible around the lower portion of the columns"1 (pl. 11:5). Below the capitals is a necking band, similar to the rope-like bindings of columns of Middle and New Kingdom Egyptian architecture52(pl. 11:5). The architrave and frieze are shown in some examples by double purled lines, in others by a series of closely spaced lines which could depict triglyphs and metopes, or more probably dentils5"(pl. 11:6). The pedimental decoration consists of a sun disk between uraeii. Within the building sits the enthroned figure of Isis, wearing an Egyptian headdress and long robe. She suckles the infant Harpocrates who wears the crown of Upper and Lower Egypt and sometimes holds a lotus flower (pl. 11:5). Two facing hawks sometimes perch atop the throne,"5and in some examples an upright palm frond or a vase on a stand is placed in front of Isis (pl. II1:7). Although derived from a long series of Egyptian statues and statuettes, this image, apart from the headdressesof Isis and Harpocrates, is completely Hellenic in style. The architecturalstyle of this temple is very interesting. It derives ultimately from simple reed and matting huts, with a roof stretched over an arched frame, which were common in Egypt from predynastic times.55The popularity of this type of structure with its characteristicshallow-archedroof continued into the Ptolemaic and Roman periods, as Egyptian scenes in Nilotic paintings and on the Palestrina mosaic demonstrate. The mosaic shows three examples of this shallow-arched building type,56one of which provides a particularly close parallel to the numismatic Isis temple. It is a small Doric-columned shrine, open front and back, with side screen walls. Through it priests, in procession, bear a cult image. An even better comparison can be found in a fragment of Alexandrian gold-glass, depicting a shallow shrine with tetrastyle Egyptian columned facade, and a niche in the back wall presumably for the cult image."5Such shrines of small size, freestanding and set within colonnaded enclosures, appear on Pompeian wall paintings with Egyptian motifs. Two examples come from the house of P. Cornelius Teges at Pompeii.58Another can be seen in the monochrome frieze of the house of Livia on the Palatine.59 Evidence for this building style may also come from an important later art work with an Egyptian ambience-the newly discovered glass mosaic panels from the Isis temple at Kenchreai.60 These opus sectile panels, dated to the early fourth century A.D.,61consist of Nilotic plant and animal friezes, a number of harbor landscapes, and two monumental figures of Plato and Homer. In the buildings of the harbor landscapes are a number of features which could be interpreted as "Egypand in one panel there appears to be a tian,""62 building with shallow-vaultedroof and the late feature of a curved apse.63Should this visual reading of the scene be correct, the appearance at such a 50 Trajan year 12, Hadrian years 18,19,20,21,23; Antoninus Pius years 2,5,7,8,10o,II,I2,24; Marcus Aurelius year II. ing on right. 60 Portions of the mosaic published in: John G. Hawthorne, "Cenchreae, Port of Corinth," Archaeology 18 (1965) 1912oo; Miriam Ervin, "News Letter from Greece," AJA 71:3 (1967) 298-299; Robert L. Scranton, "Glass Pictures from the Sea," Archaeology 20:3 (1967) 163-I73. I was fortunate to have seen photographs and drawings of other panels of the mosaic through the kindness of Prof. Scranton. References in nn. 62-63 infra are to these panels. 61Archaeology 20:3 (1967) 171. 62While paralleled in the Alexandrian coin architecture, not all of these features may be unique to Egypt. Panel VI-4-A (supra n. 60 Scranton), high front steps and disk-decorated pediment of far right foreground building. Panel VI-4-B (unpublished), hexagonal, open-sided domed structure very like the Herakles kiosk of the Alexandrian coins (infra, Conclusion). Panel VI-4-B, high, curtained doorway, like that of the Helios-Sarapistemple on the coins (infra n. 99). Compare also left-hand temple of the Palestrinamosaic (supra n. 56). 63 Panel VI-5-A, building in upper right corner. 51Comparable incised detail appears on columns of the Ptolemaic period from Edfu. Lange and Hirmer, Egyptian Architecture, Sculpture, Painting (London 1937) pl. 259. 52 W. S. Smith, The Art and Architecture of Ancient Egypt, Pelican History of Art (Baltimore 1958) pls. 62b, 76b, I62a. 53 Dentils are clearly shown in the comparable buildings of the Palestrina mosaic, M. Rostovtzeff, An Economic and Social History of the Roman World (Oxford 1957) pl. 51, and on many of the miniature shrines of this style. 54 Oxford no. 6oi. 55 Smith (supra n. 52) fig. 4, lower register. 56 Rostovtzeff (supra n. 53) 277, pl. 51. 57 M. Rostovtzeff, "Die Hellenistisch-R6mische Architekturlandschaft,"RdimMitt 26 (1911) 65, fig. 38. 58 A. Maiuri, Monumenti della Pittura Antica Scoperti in Italia. La Pittura Ellenistico-Romano. Pompeii fasc. II, Casa di "P. Cornelius Teges" 23-24, pls. III-Iv. 9 G. E. Rizzo, Monumenti della Pittura Scoperti in Italia. Sezione II Roman fasc. III. La "Casa di Livia" pl. ix 2, build- 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA late period of this longlived Egyptian building type is of considerable interest. At Alexandria itself, unfortunately, no buildings of this type remain above ground, although their appearance can be imagined from the chapels of the underground tombs,"6 the naos-shaped stelai from the cemeteries,65and even from molded terracotta lamps.66A miniature stone naos in the Alexandria Museum67 (pl. 12, fig. 28) represents an Isis-Harpocratesshrine of a form slightly different from that on the coins. An Isis dressed in a closefitting Egyptian garment and flanked by guardian animals sits within double doorways inside the shallow-arched building. The naos could be a shorthand representationfor a shrine with a porch and two interior rooms, in the innermost of which is the Isis statue. The ornate funerary chapel in the hypogeum of Kom-el-Shugafa at Alexandria68is an underground version in this same Egyptian style. Differing in details from the Isis-Harpocrates coin building, it is nonetheless a structure of the same type. Although the architectureof the Isis-Harpocrates temple shown on coins thus finds many parallels in small objects and in tombs, in the absence of excavated structures there is no way of knowing its size, and of determining whether it is a small shrine in the center of an open sacred precinct, a fullsized temple, or an interior statue naos. Location of the Isis Temples. Ancient literary sources list a number of Isis temples in and around the city of Alexandria,"6but the location of the two temples shown on the coins is not entirely clear from this list. The pylon entrance, however, may belong to a temple of Isis said to have been the first 64 Among these tombs are: Anfouchy-Breccia, Alexandria (Bergamo 1922) 329-334, figs. 246-247; Gabbari 63 built at Alexandria after its foundation."7Ausfeld locates this temple to the east of the city on a canal originally within the walls and later in an extramural cemetery area.7 The Isis-Harpocratestemple could perhaps house the Isis Mater to whom an inscription on Rhacotis hill was dedicated." If the shrine was placed on Rhacotis hill, it was probably a part of the Sarapeum complex, perhaps the building adjoining the temple of Sarapis on the west.73However, the difference in architectural styles between the IsisHarpocrates structure and the Sarapis temple on the coins may suggest that it was an inner naos for the Isis statue rather than the outer temple building. In the case of the Isis-Harpocratesshrine, it is especially disappointing that excavation has not provided more definite evidence for its location and surrounding arrangements. CANOPI PYLON AND PORTABLE SHRINE Osiris and Isis, in the form of deity-headed jars called Canopi appear on the Alexandrian coinage within a pylon and a shrine--both at first sight similar to those of Isis.74 Some examples depict only one Canopus.75 The pylon here, although much less carefully representedthan that on the Isis coins, has the same features of windows and flagstaves in the side towers (pl. 11:8). An eagle sometimes perches over the central lintel.7"The pylon shown here may, indeed, be identical with that of Isis. In that case, the presence of the Canopi probably indicates a cult place to them within the main Isis temple." A variation of the type occurs in two unique coins in the collection of Mr. John Slocum.7" ed. (Berlin 1926, reprinted 1958) I, 29, 4. This may be identical with a temple of Egyptian Isis built by Alexander at the --BSRAA 3 (I9oo) H. Thiersch, "Zwei Griiberder R6mischen time of the city's foundation. Arrian, Anabasis 3.1.5. Arrian's Kaiserzeit in Gabbari," 27, fig. 6; also BSRAA 2 (1899) 52-53, Anabasis with an English translation by E. I. Rabson (Loeb pl. K. Classical Library 1929) 292-293. 06Examples of these naiskoi in E. Breccia, La Musee Greco71 Ad. Ausfeld, "Zur Topographie von Alexandrie und Romain (Municipalite d'Alexandria 1925-31) 35, pl. 23. E. Pseudo-CallisthenesI, 31-33," RhM 55 (I900) 369. 72G. Botti, Plan d'Alexandrie a l'Apoque Ptolemaique (AlexSieglin and Th. Schreiber,Die Nekropole von Komesch-Shugafa I (Leipzig 1928) 12o fig. 70. Drawings of them in France andria 1898) 136. le Corsu, "Quelques motifs Egyptiens survivant dans l'architec7a Rowe, Discovery of the Famous Temple and Enclosure of ture religieuse Alexandrine," Rd'E 18 (1966) figs. 3,4,40,41. Sarapis at Alexandria. ASAE supp. 2 (Cairo 1946) pl. 17. 66 S. Loeschke, "Antike Laternen und Lichthiiuschen," Hereafter Rowe, Discovery. Bonnlbb 118 (1909) 136, 1. 74Pylon type issued in years 12,14,16 of Trajan, under 67 Unpublished. Inv. no. 3212. From Kom el Kazui. Hadrian (year effaced on extant examples), and under Marcus 68W. Sieglin, Expedition Ernst Sieglin, Ausgrabungen in Aurelius years 5,6. Alexandria I (Leipzig 1908). Breccia, Alexandria ad Aegyptum 7 Some examples from Trajan year 12; Marcus Aurelius 164 fig. 71. years 5,6. 69 Sources collected in Adriani, Rep. C 251-252. 76 For example in Dattari 1167. These four77 Adriani, Rep. C, 249. teen allusions probably refer to perhaps half that number of 78I am grateful to Mr. Slocum for permission to study Isis temples. his collection and include the photograph pl. I:9 from a cast 70 Pseudo-Callisthenes,Historia Alexandri Magni, I. G. Kroll of his coin. ad Aegyptum 64 SUSAN HANDLER These coins represent a priest, within the pylon, carrying a Canopic image (pl. ii:9). The shrine of the Canopi is not a full-sized building, but its appearanceon the coins is of methodological interest." An example such as BM 877 (pl. ii:io) depicts two Canopi within what seems to be a carefully detailed temple with shallow-vaulted roof and two papyrus columns. Another variety, more sketchily depicted, shows the same "temple" with an illogical swag of drapery beneath the temple floor"8 (pl. II:Ii). A third coin explains the type8"(pl. II:I2). It is in reality a miniature shrine on a high-legged draped base, flanked by two sphinxes on pedestals. Were it not for the second and third shrine types, these coins would seem to depict a temple of ordinary size. In fact both Milne"2and Poole"3describe the type as a "temple." However, two comparisons suggest a more plausible explanation: a terracotta group from Egypt represents two priests carrying a similar portable Harpocrates shrine on their shoulders,"4and portable shrines are seen on coins of Tyre."8The coins of the Slocum collection may depict the transferralof the Canopi images to such a portable shrine which, like the Harpocrates and Tyrian shrines, could be a small structure carried in religious processions. The Canopi type provides a particularly clear example of the necessity of studying all coins of a series for architectural information. SARAPEUM The temple and precinct of Sarapis is unusual in being one of the few excavated monuments of Alexandria.86 The excavations carried out in the of Trajan, years 5,8, 7gType issued in years 13,15,18,20 18 of Hadrian and years 5,6 of Marcus Aurelius and Lucius Verus. 81ANS Trajan year 13. 8oDattari 1132. 82 J. G. Milne, The Ashmolean Museum, Catalogue of the Alexandrian Coins (Oxford 1933) 148. Hereafter, Milne, Oxford Cat. 83 R. S. Poole, The British Museum, Department of Coins and Medals. Catalogue of the Coins of Alexandria and the [AJA 75 1940's were never properly completed, and most of the architectural remains above foundation level had long since been plundered, starting with the columns carried off in the Middle Ages to bolster the city's sea wall."8It is unfortunate that the excavation cannot provide either confirmation or denial of the architectural information of the coins for this major Alexandrian monument. Thus the conclusions reached in the following discussion on the basis of numismatic and literary evidence should be taken as provisional, subject to modification by further knowledge. In the excavations at the Sarapeum the most important finds were the foundation plaques deposited beneath the corners of both temenos and temple.88 These contain dedications of the temple and precinct by Ptolemy III to Sarapis,89thus providing confirmation of the location of the Sarapeum, and giving a building date in the second half of the third century B.c. The temple and its precinct is located on the hill of the Rhacotis quarter of the city (ill. i), where according to tradition a temple to Isis and Osiris existed before the building of the Ptolemaic sanctuary. Since these deities were the Egyptian forms of the later Isis and Sarapis,it was logical for the Ptolemies to build their temple to Sarapis on the same site. The building included adjoining subsidiary structures-a shrine on the west dedicated to Isis, consort of Sarapis, and another on the east to their son Harpocrates.9oFrom the vast Sarapis precinct the excavators found only one Ionic capital and two Corinthian ones of Ptolemaic type.9' The temple, which faced southeast, was approached by a monumental staircase along the short side of the surrounding colonnade.92 This which will be cited. 87 Reference in Botti, Fouilles a la Colonne Thdodosien. 88 Rowe, Discovery io. Ten beneath each corner, one of each set of tablets was of gold, silver, bronze, Nile mud and faience, and five were of glass. 89 The parallel inscriptions in hieroglyphic script and Greek read as follows: King Ptolemy and Arsin6e the brother gods (dedicate) to Sarapis the temple and the sacred enclosure. 90 Plan of the temple complex: Rowe, Discovery pl. 17. Nomes (London 1892) o02. Hereafter, Poole, BM Cat. Foundation plaques show that the Harpocrates shrine was 84 W. Weber, Die Aegyptisch-GriechischenTerrakotten (Beradded on by Ptolemy IV (Rowe, Discovery 55). The earlier lin 1914) pl. 12, no. 127. Sarapistemple and Isis shrine are centered within the enclosure, 85 G. F. Hill, Catalogue of the Greek Coins of Phoenicia as is the later Roman temple. 91 Ionic capital-Rowe, Discovery I9. Corinthian capitals(London British Museum 1910) pl. 34, 16. 86 First excavated by G. Botti, L'Acropole d'Alexandrie et le Botti, Fouilles a la Colonne The'odosien124. ComparablePtoleSirapeum apres Aphthonius et les Fouilles. Memoire present6 maic Corinthian capitals exist in the Alexandria museum, for a la Societd Arch6ologiqued'Alexandrie (Alexandria 1895). G. example that illustrated in K. Ronczweski, "Descriptions des Botti, Fouilles a la Colonne Theodosienne. Memoire etc. (I897). chapiteaux Corinthienset varies du mus6e greco-romaind'AlexEv. Breccia, Les Fouilles dans le Serapeum D'Alexandrie en andrie," BSRAA supp. 22 (1927) pl. I. 92 Rowe, Discovery 61, notes beddings for this stair on the 1905-o6, ASAE 8, 62-76. Rowe, Discovery. A. J. B. Wace, whose name does not appearas author, assisted in interpretations SEside of the colonnade. 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA original rectangular colonnade with its above and below-ground rooms surrounding the temple, was enlarged some time in the Roman period. The temple of Sarapis itself was also enlarged;" the lengthening and thickening of its foundations suggests an increase both in the number of columns and facade in the temple's height. The cause and probable date for the Roman rebuilding of the temple are not certain. Ancient authorities do speak of a fire, probably in the reign of Commodus,9"but the only evidence from the site itself comes from coins found in the four foundation corners of the piscina built to the east of the enlarged Roman temple. These coins range in date from Trajan through Septimius Severus and Geta, about A.D.2II."9 If the rebuilding of the temple was contemporarywith building of the piscina, it could possibly have occurred at the time of Caracalla's visit to Alexandria in A.D.215.96 Wace, and Rowe, the excavator, think that the Sarapeum might have been destroyed at an earlier period, during the Jewish riots of Trajan's reign, 65 of representationson Alexandrian coinage."99 There is the usual difference in the types through time,100 but three main ones were used. Type I. In the first coin type the god is shown standing in a temple with a facade of two Corinthian columns and a gabled roof (pl. 11:13). Sometimes, in a variation consistent with numismatic conventions,0"'the facade is shown as tetrastyle'02 (pl. II:14). Various stylizations are used for the capitals, the most careful being that of Trajan's year 15 with tightly curled volutes above a ring of three leaves.'03The Hadrianic examples of this type display a slightly different three-tieredconvention. The gabled pediment contains a decoration of two Nikai holding up a wreath. Architrave and frieze (or cornice) are sometimes indicated by a double line along the base of the pediment. One example clumsily indicates dentils. Roof decorationis shown by small dashed lines along the roof ridge; several examples depict standing lateral acroterial figures. A vexillum flying from a corner column provides additional decoration in some examples of Trajan's A.D. I I4-Ii5.97 They believethat the known destrucyear 2. tion of the Nemeseion "near the Sarapeum" (in Within the temple a figure of Sarapis wearing reality in an entirely different quarter of the city)"s chiton and himation, a modius on his head, grasps a (ill. I) provides evidence that the Sarapeum was in his left hand, and rests his right on a small destroyed at the same time and rebuilt by Hadrian. spear stele, which is dotted in rows to indipedimented The information of the coins for the problem of cate letters the rebuilding date will be discussed below. (pl. 11:I5). Photographs of one examDattari ple, 1142, seem to show actual writing Sarapeum Coin Representations (tuc. . .trai?), although it has not been possible As a monument second only to the Pharos in for me to verify this detail from the coin itself. A importance, the Sarapeum appears in a long series few examples of Antoninus Pius year 12104 depict 93 Rowe, Discovery 60, gives the width of the Roman temple as 21.10 m. The temple could not be excavated to its full length. 94 Clement of Alexandria, Protrepticus (Loeb ed. G. W. Butterworth trans. London, New York 1919), 47. "Nor did it (fire) spare even the temple of Sarapis in the city of Alexandria." 95 Rowe, Discovery 62. 96 Wace in Rowe, Discovery 64. 97 Wace and Rowe, Discovery 62-64. 98 The Nemeseion was in the ancient Delta or Jewish quarter of ancient Alexandria, in the NE near the sea, rather than in the sw away from the sea where the Sarapeum was located. An ancient inscription confirms the general site of the Nemeseion. G. Botti, "Inscriptions Grecques et Latines trouvis en tgypte en 1897-8," BSRAA (1898) 64-65. " Examples extant from year 2 of Domitian; years 2,6,7,11of Hadrian; years 17, and 20 of Trajan; 2-8,Io,II,17,19-21 2,7,8,12 of Antoninus Pius and years 6,Io,II of Marcus Aurelius. It should be noted that other temples to Sarapis also appear on the Alexandrian coins. Sarapis shares a temple with Tyche and Demeter on coins of Trajan (for example: Dattari I152). He is also seen in a Helios-Sarapis temple which may represent that at the resort of Canopus east of Alexandria (Walter Otto, Priester und Tempel im Hellenistischen Aegypten [Leipzig and Berlin 1908] I, 400 n. 2 refers to a statue dedication by a Syrian in the Helios-Sarapistemple at Canopus. Other dedications cited by M. P. Nilsson, Geschichteder griechischen Religion II, 491 n. 3). Ancient references confirm this proliferation of temples to different forms of Sarapis at Alexandria (Calderini, Dizionario 146). 100 Type I is seen from Domitian through Hadrian year 17, type 2 only in Hadrian year 17, and type 3 in Trajan year 13, Hadrian years 19-21 and in all extant examples under Antoninus Pius and Marcus Aurelius. 101Supra, Introduction. 102In some examples of Trajan, years II,I2,13. From the scale plan in Rowe, Discovery pl. 17, the foundations revealed by the excavations look large enough for 6 columns across the facade--too many to have put on the coins. 103These might at first glance resemble Egyptian composite capitals, were it not for examples from succeeding years, which clearly represent the Corinthian order. 104ANS, Antoninus Pius year 12; Athens, Demetrio, Antoninus Pius year 12; Dresden 2242. 66 SUSAN HANDLER Sarapis standing within the temple, between two stelai. Type 2. The second type repeats in somewhat simplified form the architecturaldetails of Type i. It appearsonly in year 17 of Hadrian. The capitals of the distyle facade repeat the three-tiered stylization of other Hadrianic types, or degenerate into a series of blobs. Within, Sarapis stands left1'5 and holds out his hand in greeting to the toga-clad emperor on the right, who places his hand on a pedimented stele in the center inscribed with his name (pl. i1:16). Sarapis in a few examples presents the emperor with a wreath or orb. Both hold scepters. Type 3. The third type is issued primarily by Hadrian, Antoninus Pius and Marcus Aurelius.106 The architecturalfeatures of the building are recognizably the same as before, but considerablymore careless in execution. This is particularly noticeable in the stylization of the capitals. In a few examples these have degenerated into two downcurving volutes over three dots, or into two lumps above a molding (pl. 11:17). The pediment for the most part displays a plain central disk in lieu of decoration. Only one coin depicts dentils beneath the pediment. A few examples are tetrastyle.107 Within the temple is a representationwhich must be the famous Bryaxian cult statue of Sarapis108 as it appears alone on many Alexandrian coins1'9 and in the description of Rufinus, "In this temple was a statue of Sarapis so large that it touched one wall with its right hand, the other with its left. This huge statue is said to have been made of all kinds of metals and of woods."110The god is seated 105 Right in only one extant example: Dattari 1947. 10 Trajan year 13; Hadrian years 19-21; Antoninus Pius Marcus Aurelius years 6,Io,Ii. years 2,7,8,12; 107 Dattari 30602; ANS Marcus Aurelius year 6 (2 examples); Berlin, Antoninus Pius year effaced. 108 It is now generally thought that the deity Sarapis was created under the supervision of Ptolemy Soter, and represented a Hellenized form of the Egyptian deity of Memphis, Osarapis. Recent bibliography and discussion in Peter Frazer, "Two studies on the cult of Sarapis in the Hellenistic world," OpusAth 3 (i96o) 1-3, n. I. Copies in various sizes and materials in A. Adriani, Repertorio d'Arte dell'Egitto Greco-Romano series A, I and II (Palermo 1961) I, 64; II, 154-185. 109 Poole, BM Cat. pls. 13-15. 110 Rufinus, Hist.Eccles. II. 23. 10-15. Text in Th. Hopfner, Fontes Historiae Religionis Aegyptiacae II (Bonn 1922-25) 630. 111 Rowe, Discovery 62. 112This concept is discussed by A. D. Nock, "Sunnaos 1-62. Nock, however, thinks that Theos," HSCP 41 (1930) this coin does not provide an example of the sunnaos theos idea, but shows rather a greeting of the emperor by the god. 113 Examples from Selinus in Cilicia, a temple to Trajan: [AJA 75 on a throne with elaboratelyturned legs and back. His left hand holds a scepter,his right rests on one of the heads of the Cerberus seated beside him. He wears chiton, himation and modius (pl. 11:17). The Sarapeumcoins are consistent within numismatic conventions in their depiction of the architecture of the temple. The variety of representations of the god Sarapiswithin it are, however, puzzling. I would suggest that these variations may be of help in the problem of the rebuilding date of the Sarapeum, left uncertain by the excavations. The most important coins in this respect are those of Hadrian year 17. These coins have often been taken to represent some sort of building activity at the Sarapeum. Wace and Rowe, for example, thought that the coins depicted the model of a pedimented building between Sarapis and Hadrian.11 The temple of Sarapis, as they thought, was rebuilt by Hadrian, and here the god in gratitude accepts Hadrian into his temple as a sunnaos theos,112associatedgod. However, temples in which the emperor is known to have been accepted as sunnaos theos are portrayed differently on coins:113the emperor is shown as a cult-statue within the temple, and on the pediment is a dedication to the emperor in the dative case,114 not as here on the pedimented object, in the accusative.115 Another view has it that the pedimented object is again a building, but this time the Hadrianeion,"6 the temple to Hadrian. Its appearanceon the coin in connection with the temple of Sarapis should suggest a close connection with the Sarapis precinct. The Hadrianeion, however, is known to have H. R. Hill, Catalogue of the Greek Coins of Lycaonia, Isauria and Cilicia (London, British Museum 1900) 124, 9. Pergamon, Rome and Augustus temple: R. S. Poole, Catalogue of the Greek Coins of Mysia (London, British Museum 1892) pl. 28, IO. 114Examples of such inscriptions found on actual buildings are listed by Kirchner, IG II, III, 3, I (3171-3199). All are from Athens. The Alexandrian foundation-plaquesfollow this same usage. 115Inscriptions to the emperor in the accusative appear in various parts of the empire. On altars: A. Benjamin, "The Altars of Hadrian in Athens, and Hadrian's Panhellenic Program," Hesperia 32 (1962) 73. Statue bases: J. Kirchner, IG II, III, 3, I (3287-3321). Revetment slabs (from concrete bases?): J. H. Kent, Corinth 8:3, The Inscriptions 1926-195o (Princeton 1966) I02-103. 116 J. G. Milne, A History of Egypt under Roman Rule (3d ed. London 1924) 219 n. 12. The Hadrianeion is known from an inscription of A.D. I70, which mentions an archiereus hadrianeiou kai sebaston (Archiv f. Papyruskunde II [1903] 444 n. 66). Reference from Calderini, Dizionario 89. 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA been very close to, or connected with, the Kaisareion, which was far from the Rhacotis temple of Sarapis."'1The "structure"on the coins thus could have had no physical connection with the Kaisareion. Nor does the spelling Hadrianon on the pedimented object justify an interpretation as Hadrianeion. More specifically,neither of these views are valid because they fail to explain the pedimented object on the earlier coins of Trajan and the first years of Hadrian's reign. A pedimented object is depicted on these coins, so similar to that on the coins of Hadrian's year 17 that it must be the same thing. In fact, the Trajanic coins provide the clue: the object is an inscribed, pedimented stele, identified as such by the lines of dots which represent writing, and by the apparent trace of an inscription in one example."18The same pedimented form appears again on a coin of Diocletian from Alexandria,119which shows a Nike writing an inscription on a stele identical with the one under discussion. The Trajanic coins depict Sarapis laying his hand on the stele in a gesture of protection or possession. But in the coins of Hadrian year 17 it is the other figure who places his hand on the stele. Thus Sarapis seems to transfer guardianship over the stele to the other figure, which is recognizable as Hadrian not only by his toga and scepter, but by his distinctive beard. This portrayal of him is very similar to that on an adventus coin celebrating his visit five years earlier.20 Sarapis hands over the guardianship of a stele to Hadrian in the seventeenth year of his reign. What does the stele represent? The answer comes from the governmental system of Egypt, in which public records and important private documents were witnessed and deposited in temples in other parts of the country, at Memphis and Oxyrhynchos for example.'21 Undoubtedly government records were also kept in the major Sarapeum at Alexandria, center of the government. The archives build117 The Kaisareion can be located from the existence in situ until 1879 of the two obelisks (one, now in New York, the other in London), which marked its entrance. It was very close to the present Coptic cathedral, slightly back from the ancient harbor at its midpoint. Calderini, Dizionario 118-119. See also map in E. M. Forster, Alexandria, a History and a Guide (1922, rev. 1938. Anchor Books repr. 1961) 2-3 (II.I). 118Supra, Sarapeum Coin Representations.The stele reading comes from the catalogue photograph, and cannot be taken as firm evidence, until I can check it against the original coin. 119Poole, BM Cat. no. 2523. 67 ing must have formed a part of the "daughter library" connected with the Sarapeum. We know that a change in record-keeping procedure was under way in the reign of Hadrian. An edict of T. Flavius Titianus, prefect of Egypt, issued in A.D.127, decreed that documents hitherto sent for deposit to the Naneum at Alexandria must now also have copies deposited in the new library of Hadrian.122 Furthermore, the director of the Naneum was obliged to get permission from the directorof the library of Hadrian in order to allow anyone to examine the archives under his care. A further decree of the prefect, written some five months after the first one, shows that his first decree had not been well followed. For the location of the Naneum, known only from this decree, we have no clear evidence. The coins suggest that it was somewhere in the precinct of Sarapis.Nor is the exact location of the library of Hadrian certain, although it was probably connected with the Hadrianeion. The decree indicates that five years before the issuing of this coin of Hadrian a beginning was made in shifting government records to the library of Hadrian. The year 17 coin must show the completion of this process, turning over to the library of Hadrian the documents formerly under the jurisdiction of Sarapis. The coins of Hadrian year 17, then, taken as they must be in connection with earlier ones of the same series, provide no evidence for building activity at the temple of Sarapis in the reign of Hadrian, or specificallyfor any rebuilding of the original structure. This interpretationis borne out when we look at depictions of the Sarapeum in the years following year 17 of Hadrian. It is striking that practically exclusively from this period on the cult-statue of the deity is shown in the Sarapeum series.123 Sarapis in the earlier coins was shown in an active role, suggested particularlyby his standing position. Here at last can be seen the image of Sarapis which must have been familiar to his many worshippers. 120 Milne, Oxford Cat. no. 1380. 121U. Wilcken, Urkunden der ProlemaFrzeitI, Papyri aus Unterdgyten (Berlin and Leipzig 1927) 6o6. Deposit of documents in the Memnonion at Memphis. Also Otto (supra n. Ioo). From papyrus references to scribes attached to the temples (p. 296), and contracts and documents notarized at them (p. 294), Otto concludes (p. 297) that the temples in Egypt, as in Greece, served as archive buildings. 122Document in Grenfell and Hunt, The Oxyrhynchus Papyri part I (London, EEF 1898) 68-74, no. 34. 123 Supra n. IoI. 68 SUSAN HANDLER The Sarapis coins depict no change in the building itself during or after the reign of Hadrian; in fact no architecturalchange is evident throughout the entire coin series. After Hadrian's reign, the temple is still shown as Corinthian, with a facade which was at least tetrastyle, and a decorated pediment and roof. The Sarapeum coins, then, should show us the Hellenistic state of the temple before its Roman rebuilding; they are thus of considerable importance in providing evidence that the Ptolemaic Sarapeum was not rebuilt until after year ii of Marcus Aurelius (172-3 A.D.), when the issuing of the Sarapeum coins ceases. The temple may have been rebuilt during the reign of Caracalla when the piscina was erected in the Sarapeum precinct. If this interpretation is correct, the numismatic evidence gives a picture of one of the most famous temples of Alexandria, essentially in its original third century B.C. Ptolemaic form. Although the Sarapeum is one of the few Alexandrian buildings to have been excavated, previous depredation of the site made architectural reconstruction impossible. Hence the coins not only confirm but amplify knowledge of the building, particularly of its architectural order-though details which one would like to know for such an early example of the Corinthian order are not, of course, shown in the miniature coin representations. But the fact that such an early and important Corinthian temple at Alexandria existed unchanged through the second century of the Roman era is of considerable interest. [AJA 75 (pl. I2:I8). The architrave and cornice are indicated by horizontal lines, the frieze by a broader space between. The frieze is ornamented in some examples by dots, perhaps decorative shields (pl. I2:I9), in others by swags of garland (pl. 12:20). Over the center of the entablature are seen lumps of material from which flames appear to rise.125 Between the central columns stands a draped female figure, probably personifying Eusebeia, piety,126 who holds out her left hand to drop incense over a portable altar. Between the columns of some examples are shown crossed lines (pl. 2 :I8); others depict only four facade columns, but show the central figure more clearly (pl. 12:20). One coin (Dattari 3006)12 depicts an agathos daimon serpent wearing a skhent on one side of the altar, with a uraeus serpent wearing a two-horned sun disk on the other (pl. 12:21). This altar must have been an imposing structure. It was at first identified as a temple by E. Breccia,'28 who thought it a temple of Eusebeia, the central figure, with flames rising from the entablature.The absence of corner antae on the building could suggest either a prostyle plan or the surrounding colonnade of a peristyle temple. However, this theory necessitates a hypaethral roof construction which, to my knowledge, is used nowhere else in Alexandrian temple architecture.R. S. Poole,129on the basis of BM coin 882, proposed that the building shown was an "altarof the Kaisareion,"the temple to the deified Caesars. The crossed lines between the columns suggested to him curule chairs set out for the emperor and magistrates as deities. Poole's interpretation, however, falls before coins such as AGATHOS DAIMON ALTAR that illustrated in pl. i2:I9, which make clear that One of the most interesting of the Alexandrian the intercolumnar objects are not chairs, but openseries of architecturalcoins is that which depicts a work grilles such as those found on many other monumental, colonnade-enclosed altar. The type ancient buildings."30These grilles must have been appearson coins of Hadrian and Antoninus Pius.124 movable to allow access to the interior. The representation of the altar varies slightly, as A third identification is that of Vogt,131 who usual, according to the year of issue. The most believed that the altar was dedicated not to the complete examples depict a porticoed structure emperors, but to the agathos daimon of the city itwith six, probably Ionic, columns across the facade, self. This interpretation is persuasive when coma two- or three-step podium, and a flat entablature pared both with the literary evidence and with 124 Issued in year 21 of Hadrian, and years 2,5,7,15,17, 23,24 of Antoninus Pius. 125 These must be cakes of pitch burnt as incense. Forbes, Technology (supra n. 17) III 25. 126 The figure of Eusebeia in the same pose is also seen alone on the Alexandrian coins. For example: Dattari 1964, pl. 12. 127 Dattari 3006. Poor illustration in J. Vogt, Die Alexan- drinischen Miinzen (Stuttgart 1924) pl. 6 (hereafter Vogt). 128 Breccia, Alex. ad Aegyptum 311-312. 129 Poole, BM Cat. 93. 13o For example, on the Parthenon. G. P. Stevens, "The Setting of the Periclean Parthenon," Hesperia supp. 3 (1940) 67-73. 131 Vogt io6. ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA 1971] 69 existing monumental altars. Architectural remains of such an altar as that on the coins no longer exist at Alexandria, nor is such an altar mentioned in descriptions of sanctuaries such as the Sarapeum or the Kaisareion. The altar depicted on the coins may, however, be connected with a sanctuary mentioned in Pseudo-Callisthenes' Alexander romance, which records the founding of Alexandria. "They began to build Alexandria from Central Square, and the place first gained its name because the building of the city started from there. To those moving about the city a serpent used to appear, frightening the workmen, and they put a stop to the work to be free of the beast. The matter was turned over to Alexander. On the following day he ordered it to be subdued where it was found. And taking charge over the intruding beast at the place now called the Stoa, they subdued and killed it. And Alexander ordered that there be a temenos to it on the spot, and burying it, they deposited the remains. And nearby he ordered wreaths to be made in memory of the appearance of the agathos daimon."'32 I believe that the coins depict an altar associated with this agathos daimon temenos. A confirmation for this identification comes from coin 3006 of Dattari's collection (pl. 12:21): the agathos daimon serpent seen on the left side serves to identify the deity to whom the altar is dedicated. A festival to the agathos daimon took place on the 25th of the month of Tybi, foundation date of Alexandria, and sacrifices were also made in private houses to the heroized Alexander, "the serpent born."'33 Thus it seems that the good genius of the city and that of its founder were worshipped as one. The temenos described by Pseudo-Callisthenes- hence also the altar-was locatedat the centerof the city. It was near the tetrapylonat the intersection of Canopic Street and the Street of the Alexandria'smain thoroughfares(ill. I). Soma,"13 Not only the deity and the approximatelocationof but also the altarare given by Pseudo-Callisthenes, its probablebuilding date in the last third of the fourth centuryB.C. The altaris a remarkablemonumentwhen one considersthat it antedatesby at least a hundred yearsthe other major Hellenisticmonumentalaltars.These altarsat Pergamon,Magnesia,Kos and differ from the Alexandrianexamplein Priene"35 their podia and steps, higher than the shallow base of the earlierbuilding. A contemporary parallelto the Alexandrianaltar is foundin the altarcourtat Samothrace, alsodated 132Pseudo-Callisthenes, Historia Alexandri Magni I, G. 96-98. See also the rebuttal to Broneer, P. W. Lehmann, "Letter to the Editor," AJA 71:4 (1967) 429-432, with mention of parallel in the Alexandrianaltar coins. I am in substantialagreement with the excavators on the form of the altar-court; as they point out (54ff) not all elements in the reconstructionare certain. 138The Altar Court fig. I8. 139 A similarly "progressive"structure may be the fountain seen on only three extant Alexandrian coins (Paris 1255, London 546; Demetrio IOI2). It resembles a Ptolemaic fountain eulogized in an epigram perhaps written by Poseidippos. (0. Gu&raudand P. Jouguet, Un Livre d'colier du I1e Sidcle avant J.C. Publicationsde la Societe Royale tgyptienne de Papyrologie. Textes et Documents II [Inst. Fr. D'Arch. Orientale, Cairo 1938] 20-23.) This structure may be the ancestor of the elaborate display fountains of the Roman period. Kroll ed. (Berlin I926, repr. 1958) 32. 133 L. R. Taylor, "Alexanderand the serpent of Alexandria," CP 25 (1930) 134A. 377. Ausfeld, "Zur Topographie von Alexandrie und Pseudo-Callisthenes I, 31-33," RhM 55 (1900) 380. 135References to illustrations and publications of these altars are collected in C. G. Yavis, Greek Altars, Origin and Typology (St. Louis 1949) 197-199. K. Lehmann and D. Spittle, Samothrace 4, II, The Altar Court. Bollingen Series 60 (New York 1964) hereafter The Altar Court. 137The reconstructionand purpose of the Samothrace altar court have been questioned. H. Seyrig, "Un edifice et un rite 136 de Samothrace," CRAI (March 1965) 105-10. O. Broneer, review, "Samothrace 4, II, The Altar Court," AJA 71:1 (1967) to the thirdquarterof the fourthcenturyB.C.136 This structure,as restoredby the excavators,'37 is similar to the Alexandrianbuilding in having a continuousfrontscreenof columnsabovetwo shallow steps. The Samothracian however,is facade, enclosedbetweenantae,and becauseof the limited accessfrom the sides and back, has a colonnade only along the front.'3*Another similarityis an interiorplatformwhichwouldhave raisedthe altar properabove the stylobatelevel-a featurewhich helps to explain the pictorialconventionof the Alexandriancoins,in whichflamesrisehigh above the entablatureof the altarenclosure. The Alexandrianagathosdaimonaltaris of the greatestinterestas one of the earliestbuildingsin the city. The coins show an architecturewhich is wholly Greek with no Egyptianadmixture.Furthermore,it was a progressivestructurefor its period,'39the precursorof a long line of Hellenistic monumentalaltars. 70 SUSAN HANDLER [AJA 75 of a central cult statue, and the general similarities of form with the other arch representationsmake it Buildings of the Roman period also appear on the Alexandrian coins, among them triumphal or clear that this unusual issue was also a structure commemorative arches. These are rare on the coin- of the same type. 3. A third arch type is known from only one ages of the Roman provinces, especially in the east coin minted in the second year of Galba's brief to where, my knowledge, no local arches are shown on the coins."14Thus it is unusual that arches of reign,"' the earliest apparentlylocal arch to appear four different types appear with some frequency on the Alexandrian coinage47 (pl. 11:25). It is a very tall structure of three entrances, with superon the coins of Alexandria.4' imposed pilasters, the higher of which continue I. This most common type is issued under Domiabove the vaults of the arches up to the frieze. tian, Trajan and Hadrian. The arch has three The gabled pediment supports a small attic, astride barrel-vaultedopenings, the middle archwayslightly the roof and crowned by a barely visible ridge higher than those on the sides (pl. 11:22). A niche chariot. This is an architectural imarrangement above each side opening contains a figure, seated The of the die engraver possibility. inexperience sideways in some examples,'42in others standing here him to led his design, leavcrowd apparently frontally."'3 Spear-bearing figures just above the at the of the coin for both ing inadequate space top socle of each arch provide further sculptural decoand full-sized attic. ration. A pediment decorated with Nikai and disk pediment 4. A fourth arch type was issued under Trajan. is topped by an attic, which in turn is surmounted two examples survive.'48 Unlike the other by a figure in a chariot, with trophies at each Only arch coins this one seems to copy a sestertius of corner. Some examples show a Doric frieze below the Roman mint of A.D. 64-68, which portrays the pediment."44 Nero's Parthian arch."49The Alexandrian coin An apparent variant of this arch is similar, but an arch with a single opening, seen from of more elongated dimensions. In place of the pedi- depicts an in a angle three-quarter view otherwise unment the attic is ornamented with two narrow known to the Alexandrian coinage (pl. 11:26). bands of decoration, similar to the Doric frieze Pilasters flank the entrance, and in a niche on the of the other coin arches145(pl. I1:23). Possibly here as in type 3 the elongation of the lower part of the side of the arch stands a nude male figure carrying and shield. The architectural details and design did not allow enough room on the coins spear numismatic depiction here correspond very closely to include the whole upper portion. Apart from those on the to sestertius. the absence of a pediment, this representation is The copied type makes this coin unique in the so similar to the common type that I would assume arch series and indeed in the coinage of Alexandria it to be the same monument. after Augustus, which otherwise depicts only local 2. A second type of arch is seen in coins issued It is curious that such an old type of the under Trajan. The entablature is supported by buildings. Roman mint should have been copied at Alexfour Corinthian columns, which seem to rise diandria, where Roman coins did not circulate rectly to it, with no indication of arches opening freely.'50 between them (pl. 11:24). The disk-decoratedpediThe Alexandrian coin arches, particularly the ment is surmounted by an attic with the usual first century examples of Galba and Domitian, are decoration of chariot and trophies. The absence unusual for their period in having triple entrances. COMMEMORATIVE ARCHES 140None can be seen among the published coins of the standard catalogues. 141 Type r: Domitian years 5,13-15; Trajan years 10,I2,I3,I4, 16,I8,20,2I; Hadrian years 4-6,8,9. Type i variant: Trajan years 13,14. Type 2: Trajan years 12,16. Type 3: Galba year 2. Type 4: Trajan year 13 and year effaced. 142 Dattari Io83. 143Poole, BM Cat. no. 342, Pl. 29. BM Cat. no. 545; Paris 1292, 1494. 144Poole, 145For example: Oxford 625; Athens, Demetrio I022. 146ANS collection. 147Augustus' Parthian victory arch at Rome is shown earlier. See n. 6 supra. 148 ANS Trajan year I3; Paris 1439 Trajan year effaced. 149Kihler, "Triumphbogen," RE n.s. 7 (Stuttgart 1939) col. 385. Tacitus, Annales I5.I8, describes how the arch, voted earlier by the senate, was not actually built until A.D. 62, although the Romans were far from winning the Parthian war at the time. 150 M. Grant, Roman Anniversary Issues (Cambridge 1959) I6o-i68, gives examples of structures and events commemorated on the coinage at intervals of 50, Ioo or even 500 years. Unfortunately, the Alexandrian coin, minted at the awkward interval of 49 years after the erection of Nero's arch, cannot fall into this category. 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA 71 Apart from the arch of Orange,whose inscription throughconstructingfountainsor statelyportalsis Tiberian,5'thereare veryfew triplearchesuntil for you have not the wealthto squanderon things well into the secondcentury."'2 The archesof Galba like that,nor couldyou ever,methinks,surpassthe and Domitian have the featureof a pedimentin emperor'smagnificence ...",55 The Trajaniccoin the attic.This also is a fairlyunusualarchitectural arch,type2, couldbe one of the "statelyportals." detail,althoughit can be paralleledin two arches, G. Botti'56thoughtthat the archfirstseen on the one from North Africa at Mactaris,'l5the other coinage of Domitian was originally set up in from the Near East at Gerasa.154 The Gerasaarch honor of Titus; this monument,completeby the also has niches over the side entranceslike the fifth yearof Domitian'sreign,was erectedoutside Alexandrianexamples.The architectural peculiari- the ancientcity wall to the east of the city. At a ties of the Alexandriancoin arches suggest that point on the ancientroad to the Roman military Romanstructuresas localas the HellenisticPharos camp Botti believed that he could see fragments or Sarapeumare representedhere. of this arch. If these observationsare correct,the A closeparallelcan be seen in anotherEgyptian arch of Domitian may have served to mark the arch, set up by Hadrianin his new city of Anti- territoriallimits of the city outside its walls. There are no indicationsof possiblelocationsfor noopolis.This archwas still standing,althoughin somewhat ruinous condition,at the time of the the other archesseen on the coins. It is, however, French expeditionto Egypt, and its appearance tempting to connect them with the colonnade was recordedin a handsomeengraving (pl. 12, which was still visiblealong the line of the ancient fig. 29). It had threeentrances,like the Alexandrian CanopicStreetat the timeof the Frenchexpedition arches,the side ones considerablylower than the to Egypt (1798-I801)."'7 Triple arches spanning center one. Small Corinthianpilastersflank the colonnadedstreetswere a featureof cities in the side entrances,and there is the interestingfeature neighboringregionsof North Africaand Syria,but of a relievingopening running the depth of the at Alexandria,unfortunately,neither literarynor building above the vault of each side entrance. archaeologicalevidenceis decisivein determining Such an opening seems to have existedalso above their exact positionand use. the centralentrance,but it must originallyhave been concealedfrom view by a screenwall. Above CONCLUSION these openings runs a Doric entablature surI began to study the Alexandrianarchitectural mounted by an undecoratedpediment.The attic coinsfor the informationthey might provideabout abovethe pedimentis brokenaway. The similari- the Hellenisticand Roman buildingsof the city. ties of this arch to those depicted on the Alex- Clearly, coin representationsare limited in the andriancoinageof Galba,Domitian, and later of amountand kind of architecturalevidencewhich Trajan are striking, and may suggest a local can be derivedfrom them, a negativeaspectwhich traditionof arch design. will be discussedfirst, followed by the positive A numberof ancientreferencesto Alexandrian knowledge gained. The precedingdiscussionhas archesand city gates exist, but that most closely dealt with only a few of the architecturalcoin linkedto the coin representations comesin a speech types,but my conclusionsare based on all of the of Dio Chrysostom,"To the Alexandrians." Speak- types which include the following: shrines or of the Dio "In heaven's ing emperorTrajan, says templesof Athena,Hermes (2), Tyche (2), Elpis, do not see how name, you greatis the consideration Nemesis, Roma, Nilus, Euthenia, Hermanubis, that your emperorhas displayedtowardyour city? Harpocrates(2), Sarapis Tyche and Demeter, snakedeity,sacredboat,and HeraWell, then,you also must matchthe zeal he shows Helios-Sarapis, and make your country better, not, by Zeus, kles. The templesof Sarapis,Isis, Isis-Harpocrates, 151 P. M. Duval in Amy et al., L'Arc d'Orange supp. 15 to Gallia (Paris 1962) I56. Although the dedication on the arch dates to A.D.26-27, Duval thinks that the, arch was probably erected some five years earlier. 152 Kihler, "Triumphbogen," col. 480-482. 153 Illustrated and described in Anderson, Spiers and Ashby. The Architecture of Ancient Rome (London 1927) 117, pl. 59. 154 C. H. Kraeling, Gerasa, City of the Decapolis. American Schools of Oriental Research (New Haven 1938) plan 19. 155Translation from J. W. Cohoon, Dio Chrysostom III (Loeb Classical Library, New York 1940) 265. 156G. Botti, Plan de la Ville d'Alexandrie a l'Apoque Ptolemaique (Alexandria 1898) 62. 157 Description de L'Agypte pl. vol. 5, pl. 35, map, pl. 31. Text vol. 5, 298-299. 72 SUSAN HANDLER and the Agathos Daimon altar have been discussed above. Secular monuments include a monumental fountain, the arches and the Pharos. The first limiting factor in studying numismatic representations of architecture is the selection of buildings chosen for portrayal. At Alexandria, for example, the mint of the Roman government in Egypt omitted such well-known local buildings as the Museum, the Gymnasium and the Caesareum. Most frequently represented is the Pharos, architectural symbol of the city, with the Sarapeum, temple of Alexandria's official deity, next. The shrines of all other deities are seen far less often on the coinage, although the popular appeal of Isis and Harpocrates,for example, is attested by numerous statuettes.'58 It may seem that an extraordinarynumber of religious buildings are shown on the coins-not only full-sized temples, but even portable processional shrines. This profusion of religious buildings is in fact corroboratedby a Roman "census list" of the late third-early fourth century, which counts no less than 2478 public and private temples in the city of Alexandria."59Although the coins show a wide selection of both religious and secular buildings, the number of monuments which appear frequently on the coinage is quite limited. This factor, the extant number, variety and state of preservation of an architecturaltype, is of great importance in determining the amount of information which can be derived about the building depicted. The arch of Galba, for example, exists in only one worn and inconsistently engraved coin (pl. 11:25); the Pharos, on the other hand, is seen in a long series during six successive reigns.160 Since the architectural depictions change slightly from year to year, a long coin series often allows for the inclusion of new and significant structural details,161 but some information is not available even from the most complete and well-preserved series. Numismatic evidence cannot establish the number of faqade columns, precise details of architectural orders and entablatures, or the size of a building. Nor do coins give an indication of construction date. Although all of the Alexandrian architectural 158 Examples illustrated in S. Donadoni, "Iside," EAA 4, 235. Breccia, Alex. ad Aegyptum fig. 71. W. Weber, Die Aegyptisch-Griechischen Terrakotten (Berlin 1914) II, pl. 2, 7. 159 P. M. Fraser, "A Syriac Notitia Urbis Alexandrinae," JEA 37 (1951) 105. 160 Supra n. 22. [AJA 75 coins were issued in the first two centuries of the Roman empire, not all of the buildings depicted are of the Roman period. The Pharos, the Sarapeum, and the Agathos Daimon altar were all built in the Ptolemaic period. Structures such as the arches are clearly Roman. The construction date of other buildings is not known, and the coins in these cases can only provide a terminus ante quem. Supplementary material is clearly needed to interpret and verify the numismatic information. Building dates have been supplied by excavation for the Sarapeum, and by ancient literary accounts for the Pharos, but in some cases the coins provide the only evidence for the existence of a building.'62 Related structures can sometimes provide useful information, as in the case of the arch at Antinoopolis. Failing this, there are comparable representations in other media such as painting, mosaic and sculpture. Despite the limitations just noted, however, the coins have provided considerable information on Alexandrian architecture of both the Hellenistic and Roman periods. Ptolemaic Buildings The Ptolemaic buildings shown on the Alexandrian coinage include the Agathos Daimon altar (331 B.c.), the Isis pylon (331 B.c. or slightly later), the Pharos lighthouse (ca. 260 B.C.), and the Sarapeum (ca. 246-221 B.c.). The earliest Alexandrian structure on the coins is the Agathos Daimon altar, contemporary with the city's foundation in the last third of the fourth century B.c., and a building striking in its originality of design. As noted above,163monumental altars do not appear in other Hellenistic centers, with the exception of Samothrace, until about a hundred years later. The small scale of the coin precludes analysis of the details of the Ionic order. Was it, for example, similar to the earlier Ionic of the temple of Athena Polias at Priene, also dedicated by Alexander?164 The Sarapeumof Alexandria is a rare example of a large-scale Hellenistic temple of the Corinthian order-the use of Corinthian here seems to be an Alexandrian preference.CharacteristicallyHellenistic, however, is the symmetricalsetting of the temple 161 Supra, Introduction. 162 For example the fountain, supra n. 140. 163 Supra, Agathos Daimon altar. 164 W. B. Dinsmoor, The Architecture of Ancient (3d ed. rev. London 1950) 221-223. Greece 1971] ARCHITECTURE ON THE ROMAN COINS OF ALEXANDRIA within a colonnade approached by a monumental stairway, an arrangement closely paralleled at Hermopolis Magna in Egypt"'• and in other Hellenistic sanctuaries.'6 Save for the possibilityof an Isis-Harpocrates shrine of Egyptian style within the temenos, there is no trace of Egyptian architecturaltradition in this building complex. The pylon of Isis, on the other hand, represents the Egyptian aspect of Alexandrian architecture. Although adding new features such as windows, it essentially carries on old traditions of Egyptian temple architecture. The Pharos is as much a triumph of engineering as of architecture. The use of successive square, octagonal and round stages to achieve its great height, and the interior arrangements for working the beacon light have both an artistic and a technological interest. It has been suggested that the Pharos architectreceived his inspiration from Egyptian pyramid and pylon structures,'67but it is more probable that he was influenced by Hellenistic military and harbor engineering, and perhaps by the general atmosphere of scientific research encouraged by the Museum at Alexandria. The influence of the Pharos was immense; it dictated the form not only of all subsequent ancient lighthouses, but of the Islamic minaret.'68The arched underground cisterns and the Heptastadion bridge-aqueduct are further examples of Alexandrian engineering skill in the Ptolemaic period. Roman Buildings Some of the buildings seen on the coins, for example the arches, can be dated to the Roman period.'69The Alexandrian monumental arches compare to many built within the Roman empire. They are unusual in using the triple-archform at an early date, and in including the feature of a pediment in the attic of the arch. 165Wace, Megaw and Skeat, Hermopolis Magna, Ashmunein. The Ptolemaic Sanctuary and Basilica (Alexandria 1959) pl. 3. Hereafter, Hermopolis Magna. This arrangement may have been a forerunner of Roman Kaisareia. E. Sj6qvist, "Kaisareia," OpusRom I (1954) 95ff. 166 Many are discussed and illustrated in P. W. Lehmann's "The Setting of Hellenistic temples," JASAH (Dec. 1954) 15-20. 167Th. Fyfe, Hellenistic Architecture. An introductory Study (Cambridge 1936) 13. 168 Thie.rsch, Pharos, figs. 115, 129-135. 169 Another building, the Helios-Sarapis temple, not dis- cussed above, should probably also be dated to the Roman period (Dattari 3803). The three-door faqade with niche above 73 Buildings Not Precisely Dated Although many of the buildings depicted on the coins cannot be precisely dated, they show characteristic features of Alexandrian architecture. The most interesting type is that represented by the Isis-Harpocratestemple, and by a number of other shrines and temples which appear on the coinage.'70 These buildings share the features of a shallowarched roof and Egyptian papyrus stalk and lotusbud columns. Their depiction provides evidence for the existence in religious structures of a building type hitherto known only from the Alexandrian underground tombs. Some examples of the Isis-Harpocrates temple have the Hellenic motif of Ionic dentils beneath the pediment,7• a mixture of Egyptian and Greek elements which can also be seen in the Alexandrian tombs. The tomb architecture suggests that there is a chronological factor in this mixture. The earliest of the tombs, such as the late third century B.C. MustafaPaschatomb,172 are completelyHel- lenic in style. In a later tomb, that of Anfushy, the decoration is almost entirely Egyptian, with the use of dentils as an added Hellenic feature in the shallow-vaulted pediment."7 The tomb of Kom-elShugafa, dated to the late first-earlysecond century A.D.,174 presentsa much greaterfusion of Hellenic and Egyptian elements, with Egypto-Greekcapitals and shallow-vaulted pediment ornamented with dentils. Thus there seems to have been a separation of Greek and Egyptian architecturalstyles in the early period of Alexandria, and an almost complete fusion of styles by the end of the first century A.D.We do not have sufficient evidence to date the beginning of this trend, but it is likely that it started gradually, and did not reach its height until the Roman period. the central door appears to have affinities with the Roman architecture of Syria. It may represent the Helios-Sarapis temple at Canopus near Alexandria (Otto, Priester [supra n. Ioo] I, 400oo n. 2. M. P. Nillson, Geschichte der Griechischen Religion IIa, 491 n. 3). 170 Temples or shrines of the sacred boat, Harpocrates (2), Hermanubis, snake deity and Nemesis. 171 Pl. II:II. 172 A. Adriani, La Necropole de Moustafa Pascha (Alexandria 1936). Good illustration in Lawrence, Greek Architecture pl. 137. 173 Breccia, Alex. ad Aegyptum 331, fig. 247. 174 Th. Schreiber, Die Nekropole von Kom-esch-Shugafa I (Leipzig 1928); Breccia, Alex. ad Aegyptum fig. 243. 74 SUSAN HANDLER [AJA 75 mopolis Magna, contemporary with the SaraThe general form of the capitals is simipeum.s78 Another striking characteristic of Alexandrian to those of the tholos of Epidauros, although lar architecture of the Greek and Roman style in all differ-the acanthus-leafband and the proportions periods is the apparent predominance of the Corinof the abacus examples are lower than Egyptian thian order. Of the temples seen on the coins, all These local at Egyptian forms of the Epidauros.79 but one are Corinthian.'75For the Ptolemaic pebe to those used on must close Corinthian capital riod our information is far from complete, and it the Sarapeum. would be impossible to assess the exact degree of A fair number of Ionic architectural fragments predominance of the Corinthian order. Of some have been found in soundings at Alexandria. A twenty-five published Corinthian capitals in the number of large matched Ionic capitals were found museum at Alexandria, four are certainly of Ptolethe eastern harbor of Alexandria, and are now in maic date.•76It should be remembered in this conthe Museum.8so A second type of in on display nection that architectural evidence of all kinds is even larger, is also on display.81 These relatively sparser for the Ptolemaic than for the capital, Ionic capitals are all of later date than the altar Roman period, due to the rising water level of the seen the coins. on city, and the re-use and destruction of earlier archiDoric No buildings appearon the coins, although tecture. The majority of the architecturalfragments is for at least some Doric at Alexanthere evidence in the Alexandria Museum are Corinthian of Rosuch as that of Mustafa Pascha, tombs dria in the man date, enhancing the impression that CorinDoric temple, now dea suburban and in small thian continued to be favored through the Roman I seen no have Doric fragments in the stroyed.182 period. Alexandria. museum at The Sarapeum is perhaps the earliest example of use of the Corinthian order for the exterior of a Save for the Pharos, the appearance of none of major shrine, and it is noteworthy that this order the buildings depicted on the coins was previously was used for the temple of Alexandria's official This is true even for the excavated Saraknown. deity. In other parts of the Hellenistic world the the existence of some buildings, such And only Corinthian temple of comparable date is the peum. the of arch Galba, can be known only from the remote sanctuary of Zeus Olbios in Cilicia."7 The as coins. Many architecturalfeatures original to AlexOlympeion at Athens, the other major Hellenistic Corinthian temple, was not begun until the early andria have also been discovered. The corpus of coins, however, gives only a bare outline of the second century B.C. The coins are not of much help in illustrating the development of Alexandrian architecture-much details of the Alexandrian Corinthian order. How- remains to be known, and it is hoped that future ever we are fortunate in possessing examples of excavation may bring to light new information to Ptolemaic Corinthian capitals in the Alexandria add to that already disclosed by the coins. Museum and at the Egyptian sanctuary of HerRUTGERS UNIVERSITY The Orders 175 176 The exception is one temple of Tyche (Dattari 3061). C. Ronczweski, "Les Chapiteaux Corinthiens et varies du Mus6e greco-romain d'Alexandrie. BSRAA supp. 22 (1927) nos. 3, 8, 12, 13. 177 Monumenta Asiae Minores Antiqua, III. J. Keil and A. Wilhelm, Denkmiiler aus dem rauben Kilikien (Manchester 1931); Lawrence, Greek Architecture206-207. 178 Hermopolis Magna pl. I, pl. 15, fig. 2. 179 Ronczweski, Les Chapiteaux Corinthiens 4, fig. I. 180 Breccia, Alex. ad Aegyptum 208, 28. 181 Alexandria Museum no. 17833. 182Mustafa Pascha tomb, n. 172 supra. At Cape Zephyrion near the ancient city there used to exist remains of a small Doric temple dedicated to Arsin6e, and dating to the second quarter of the 3rd century B.c. (C. Ceccaldi, "Le temple de Venus Arsinoe au Cap Zephyrion," REA I9 [1860] 268-272; Breccia,Alex. ad Aegyptum fig. 34. The monopteral plan, with corner columns heart-shaped in section, is unusual. I have seen no Doric fragments in the museum at Alexandria. HANDLER 2 5 9 13 PLATE 4 7 8 10 II 12 14 15 16 22 7 23 25 24 26 II PLATE 12 HANDLER 19 20 21 I8 FIG. 29. France, Commission des Monuments d'Egypte. Description de l'gypte (2nd ed. Paris 1829) pl. vol. 4, pl. 57 FIG. 27. Courtesy Muske des Beaux Arts, Budapest FIG. 28. Courtesy Alexandria Museum