A Tripod ‘Worth Seeing’ in the Olympieion at Athens (Paus. 1.18.8)

Abstract This study proposes a new reconstruction of the tripod that Pausanias (1.18.8) recorded in the Olympieion at Athens. According to his brief description, the bronze tripod was supported by Persians made from Phrygian marble. A sculptor's sketch found during the excavations of the Athenian Agora is identified as a representation of that monument. The sketch, carved from poros limestone, depicts a standing male figure dressed in eastern attire supporting the foot of a tripod. The figural type finds its closest parallels among the colossal statues from the Forum of Trajan in Rome, suggesting a new date and context for the monument in the Olympieion. The scenario favoured here is that the tripod was dedicated following Trajan's victories in Parthia, perhaps completed or commissioned by Hadrian. Cassius Dio (68.17.2) records that Trajan departed for his Parthian campaigns from Athens, where memories of Persian defeat were actively curated.

There are also dedicated Persians of Phrygian stone supporting a bronze tripod; both they and the tripod are worth seeing.
In the passage above, Pausanias (1.18.8) describes a tripod that stood in the precinct (ἐν τῷ περιβόλῳ) of Zeus Olympios at Athens. 1 His text is our only historical source for the lost Since the date in the early Principate rests on the now re-identied kneeling statues from Rome, the evidence is due for a careful reappraisal.
The inclusion of the tripod in the text of Pausanias xes a terminus ante quem in the early A.D. 160s.The date is inferred from Pausanias' description of the odeion at Patras (7.20.6).While visiting that building, Pausanias issues an apology for not mentioning the Odeion of Herodes Atticus in his account of Athens.He explains that construction of the Athenian concert hall had not yet commenced by the time he nished his rst book, adding that it was commissioned by Herodes in memory of his late wife Regilla.Pausanias therefore wrote his book on Athens and Attica sometime before or shortly after the death of Regilla, which probably occurred in A.D. 160 or 161.8 FIG. 1. Over-life-size pavonazzetto statue of a kneeling male gure in eastern attire.The hands and head are early modern restorations.Naples, Museo Archeologico Nazionale di Napoli 6117.(Photo: © Vanni Archive/Art Resource, NY) A terminus post quem for the monument is established by the identication of the material used for the statues of the Persians as 'Phrygian stone' (λίθου Φρυγίου).Here, Pausanias is referring to a prestigious type of white marble with deep red and purple veins (e.g.Fig. 1) that was quarried near Dokimeion in Phrygia (Fig. 2). 9The coloured stone was also known in antiquity as marmor synnadicum (after the placename of its administration and distribution: Synnada) and marmor phrygium.Today, it is commonly called pavonazzetto.
Systematic exploitation of the pavonazzetto-producing quarries near Dokimeion began in the late rst century B.C., during the reign of Augustus, to satisfy the needs of imperial building projects in Rome. 10 The Temple of Mars Ultor in the Forum of Augustus, dedicated in 2 B.C., supplies a securely dated example of pavonazzetto in the city for architectural purposes. 11The temple's monolithic columns and paving slabs of pavonazzetto created an impressive space for displaying the Roman military standards, which had arrived in the capital from Parthia in October of 19 B.C., following a negotiated surrender in the preceding year.
For an early use of pavonazzetto for gural sculpture, we turn to the nearby Basilica Aemilia/ Paulli on the Forum Romanum. 12Fragments of over-life-size statues of eastern gures carved from pavonazzetto were discovered in the Augustan building. 13The statues, standing with an arm raised in a gesture of structural support, were presumably positioned in mirrored pairs below an architrave in the interior of the basilica.The general concept might have derived from the famed Persian Stoa in Sparta, which was erected in the fth century B.C. from spoils of the Battle of Plataea. 14According to Vitruvius (1.1.6)and Pausanias (3.11.3), the Spartan stoa employed representations of Persians to support the roof, a metaphor for the everlasting servitude of the enemy.The statues in the Basilica Aemilia/Paulli similarly alluded to the subjugation of an eastern foe, namely the Parthians, heirs to the Persian empire. 15The difcult-to-acquire stone from Phrygia was selected for these gures in order to exoticise their eastern origin.The striking patterns created by the veined stone conveyed the stereotyped luxuriousness of eastern garments. 16In fact, pavonazzetto was used exclusively in this context to depict fabric and attire.The separately attached faces and hands of the statues were carved from white marble, which was subsequently painted. 17hile pavonazzetto was used extensively for public building projects in Augustan Rome, it was, as far as we know, absent in contemporary Athens.One telling non-appearance occurs at the Odeion of Agrippa, a concert hall constructed in the years around 15 B.C., by the son-in-law and general of Augustus.The stage oor of the odeion was paved with slabs of white and coloured marbles, which were sourced from local and regional quarries. 18The oor anticipates a wider pattern of use: marbles under imperial control, such as pavonazzetto, tend to be scarce outside Rome until the late rst century A.D., 19 particularly in the eastern Mediterranean basin.It was around this time that the quarrying of pavonazzetto expanded.Consular dates inscribed on architectural products from the quarries witness an intense period of extraction and shipping beginning during the reign of Domitian (the earliest inscriptions provide the date A.D. 92), with increasing demand under Trajan and Hadrian. 20The quarries continued to be exploited in the third century and later. 21The Prices Edict of Diocletian, issued in A.D. 301, lists pavonazzetto (Δοκιμηνοῦ) among the most expensive stones in the empire. 22he largest recorded deployment of pavonazzetto at Athens was for the Library of Hadrian, a building complex that was presumably completed in advance of the emperor's nal visit to the city in A.D. 131/2. 23According to Pausanias (1.18.9), the structure incorporated 100 columns of 'Phrygian stone' (Φρυγίου λίθου) that Hadrian Rose 2005: 62-3;Bitterer 2007;Freyberger et al. 2007: 535-50 (T. Bitterer);Schneider 2007: 72-4, gs 21-2;2016;Claridge andSiwicki 2019: 311-12. 14 Schneider 1986: 108-15;2007: 74-5.For an effort to re-date the architectural gures of the Persian Stoa to the Augustan period, see Spawforth 2012: 118-21. 15 Schneider (2016: 416-21, 424-6) advocates a more positive reading of the statues, arguing that the gures represent Trojans linking Rome to its mythical past. 16The imagined appearance of these garments was likely mediated through earlier contacts with Persian clothing; for the fth-century B.C. interactions, see Miller 1997: 153-87. 17 For traces of paint, see Freyberger et al. 2007: 543 (T.Bitterer); Schneider 2016: 406, gs 17.6, 17.7.In the case of the kneeling statues in Naples and Copenhagen, the black-stone heads and hands are early modern restorations; see Schneider 1986: 20; De Nuccio 2002: 428-9; Gasparri 2010: 137 (E.Dodero); Lipps 2016: 207. 18Bruno and Vitti 2018: 292, 294, identifying the stones as Pentelic, Hymettian, and Karystian, in addition to an unidentied green and violet marble.Agrippa's involvement in the quarrying and trade of Phrygian marble has been proposed on the basis of a restored quarry inscription, but the evidence is not compelling; see Fant 1989: 8-9. 19Fant 1993.Pavonazzetto is attested in a number of theatres in the western Mediterranean basin that were built in the late republican or early imperial period, but in many cases, the uses of the stone are not certainly dated or phased; see Beck 2022: 101-4, with table 1. 20 Fant 1989: 9-10, 29;Hirt 2010: 291-303, with the appendix on 370-402, nos 1-459. 21Niewöhner 2013. 22For the relevant Greek text of the edict, see Giacchero 1974: 211.The stones listed seem to have been priced not as squared blocks, but as revetment; see Corcoran and DeLaine 1994. 23 For the library, see Travlos 1971: 244-52, gs 314-24; Shear, Jr 1981: 374-6; Spetsieri-Choremi 1995;  Boatwright 2000: 153-7; Camp 2001: 202-3, gs 197-8; Choremi-Spetsieri and Tigginaga 2008;  Lagogianni-Georgakarakos and Papi 2018: 110-13 (G.Sarcone); Kanellopoulos 2020; Sourlas 2021.had donated to the city.Several fragmentary shafts of these columns survive today. 24The same stone also decorated the walls of the library, as Pausanias further describes.Chrysanthos Kanellopoulos has identied pavonazzetto mouldings and revetment slabs that were once attached to the interior walls. 25These colourful architectural elements created an impressive visual statement of empire in Athens.

Character
The use of 'Phrygian stone' also provides information regarding the character of the tripod monument.The extraction of pavonazzetto near Dokimeion was under the control of imperial administrators, and as a result, acquisition of the stone was highly restricted. 26ven though limited quantities of pavonazzetto might have been available through non-imperial channels (e.g. for making revetment), 27 large blocks for carving gural sculpture are unlikely to have been available to private citizens.Pausanias (1.18.9) makes clear that the Athenians were able to obtain pavonazzetto for the Library of Hadrian only through imperial benefaction. 28Coloured stones of any kind were utilised sparingly at Athens and are especially rare in the city for gural sculpture.The tripod is therefore overwhelmingly likely to have been an imperial dedication. 29
κολοσσοί, τὰ λοιπὰ ἀγάλματα ὁμοίως ἀπολείπεται) (1.18.6). 31In the same passage, Pausanias refers to the colossal statue (τὸν κολοσσόν) of Hadrian that the Athenians erected inside or near the precinct as similarly 'worth seeing'.Given these uses of the phrase, particularly in the context of the Olympieion, it is reasonable to conclude that the tripod caught the attention of Pausanias in part because of its large size.* * * To summarise, the tripod and its supporting gures, both presumably of colossal scale, were surely erected through imperial agency.The use of pavonazzetto establishes a terminus post quem of the Augustan period, but this early date seems unlikely for the monument.In Athens, the rst archaeologically attested use of the stone occurs in the second century A.D. Given the patterns of use of pavonazzetto in the city, and throughout the empire, a date after c. A.D. 100 seems likely.The tripod was certainly standing by the early A.D. 160s, since it was recorded in the rst book of Pausanias' travels.Therefore, the potential donors are Trajan, Hadrian and Antoninus Pius.Marcus Aurelius and Lucius Verus are improbable candidates because the main military achievements of their co-reign did not occur until the mid-160s. 32

III NEW EVIDENCE FROM THE ATHENIAN AGORA
A fragmentary sculptor's sketch or model, carved from poros limestone, was excavated in 1939, from a footing trench of the Post-Herulian Wall (Fig. 3; H. 0.15; W. 0.11; D. 0.06 m). 33The nd-spot, discussed further below, establishes a terminus ante quem of the last quarter of the third century A.D. for the object.The sketch represents a standing male gure that supports, on his head, the foot of a tripod (Figs 4-5).Despite the fragmentary condition of the sculpture, the identication of the carried object as a tripod is veried by the stabilising hoop, rendered as a curved horizontal band, about 3 cm above the head of the gure. 34The foot of the tripod takes the shape of a lion's paw, typical of ritual furniture.
The gure in the sketch stands with the left leg engaged.The gure wears at least three garments: (1) trousers, rendered at the lower left leg; (2) a sleeved tunic, which is belted loosely at the waist and terminates below the knees; and (3) a ground-length cloak draped over the back and fastened below the centre of the neck by a large, disc-shaped bula.The preserved left foot is a raised dome, without carved footwear.The head is turned slightly to the left side and bowed.The blank face is framed by medium-length hair.The facial features were not rendered, so it cannot be determined whether the gure had a beard and/or moustache or was clean-shaven.The back of the sketch is worked at; presumably the two additional legs of the tripod were not carved because the gural type was repeated for each of the corresponding sides.
Andrew Stewart, who provided the ofcial publication of the sketch, suggested that the gure carries something at chest level: 'an offering tray?'. 35Stewart mistook the pose of the gure for a non-existent object, thereby missing a critical detail: the left hand is, in fact, 31 Trans. Jones 1918. 32 In this unlikely scenario, the tripod monument would commemorate Lucius Verus' war in Parthia, which concluded in A.D. 166, too late for inclusion in the text of Pausanias. 33Agora S 1170; see Stevens 1949: 269 n. 3 (mentioned and identied as a sculptor's model) ;Stewart 2013: 621-2, no. 6, g. 7. 34 The sketch cannot be for a bronze folding table (a genre which often has feet resembling the paws of a lion) because the hoop would prevent its collapse.Moreover, there is no instance known to me in which the foot of a folding table rests upon a gure; cf.Klatt 1995. 35 Stewart 2013: 621. held over the right wrist at the level of the waista pose used in Roman art to represent submission and captivity, rarely deployed before the early second century A.D. 36 Stewart concluded that the gure 'looks somewhat like a Telesphoros, but what he (or anyone else) would be doing supporting a tripod is a mystery'. 37The distinctive pose, taken together with the costume and the function of the gure as a support, conrm that the sketch portrays a stereotyped image of a captive man.It is the only representation of a tripod-captive group known to me that survives from Greco-Roman antiquity; its importance, therefore, cannot be overstated.Given the otherwise unattested subject and the Athenian provenance, it is reasonable to propose that the creator of the sketch imitated a well-known local monument: the tripod described by Pausanias in the Olympieion.
To evaluate this claim, it is necessary to understand the sketch within its own context.Why was a sketch of the tripod created?Sculptors used three-dimensional sketches and models for the planning of gures and compositions. 38Athenian carvers frequently employed poros limestone for this purpose because it was inexpensive and easily carved. 39Our gure was carved almost exclusively with chisels, an approach characteristic of sketches in poros limestone.The aim was not to carve a product in detail, but to work out the overall contours of the gure and its relationship to the larger composition.While the circumstances of the related commission are lost to us today, it is possible that a request for a reduced-scale version of the monument in the Olympieion necessitated the creation of the sketch.Special meaning had accrued to the local landmark, which earlier had aroused the interest of Pausanias.Reduced-scale versions of monuments were traded in antiquity as votive offerings and as souvenirs, providing two potential uses. 40Another possibility, although less likely, is that the sketch survived in the third century A.D., as one of the original models used in the construction of the tripod monument.Whatever its specic purpose, the archaeological nd-spot of the sketch connects it to a local marble-carving atelier.The sketch was excavated from a footing trench of the Post-Herulian Wall, where the fortication passes in front of the two southernmost rooms of the west stoa of the Library of Pantainos (Fig. 3). 41Marble chippings and unnished works demonstrate that sculptors worked in those rooms in the third century A.D., until the building was destroyed during the Herulian raid in A.D. 267. 42The workshop specialised in small-format works and portraiture, and the sculptors who laboured there were skilled practitioners of mechanical copying.The sketch demonstrates that the monuments of Roman-period Athens inuenced local artists.

* * *
The identication of the poros limestone sketch from the Athenian Agora allows me to propose a new reconstruction for the tripod seen by Pausanias in the Olympieion (Fig. 6).In the drawing presented here, it is assumed that the statues were attached to piers that actually performed the role of supporting the bronze tripod.This structural format accords with other uses of supporting gures in Roman Athens, as for example the giants and tritons from the north façade of the Odeion of Agrippa in the Agora (Fig. 7). 43Those colossal gures, six in total, were added during renovations to the concert hall in the mid-second century A.D. Standing with one arm raised in a gesture of structural support, they emerge from an integral pier that carried the weight of the architrave.Finally, the colossal size is consistent with the description of Pausanias (Section II). 44The height of the supporting statues in the illustration, c. 3 m, or about twice life-size, is hypothetical, based on the dimensions of the pavonazzetto statues of Dacian prisoners from the Forum of Trajan, to which we now turn.
IV PARTHIANS SUPPORTING A TRIPOD The captive gure on our sketch displays similarities with the colossal (H.c. 3 m) pavonazzetto statues of Dacian prisoners from the Forum of Trajan in Rome. 45Several of these statues, later transferred to the attic of the Arch of Constantine, echo the posture and composition of the tripod-supporting gure with particular closeness (e.g. the statue on the left in Figure 8).realistic elements in their works, such as single items of dress, and less often, weapons or other attributes.For example, Schneider has identied a small handful of Roman images dated to the rst and second centuries A.D. that represent Parthians wearing a distinctive V-neck tunic, which recalls the sleeved jacket actually worn by men in that society. 50hile this type of garment is not rendered on the sketch from the Agora, its absence does not preclude a Parthian identity. 51In Athens, the Parthian-Persian equation probably resulted in more generalised imagery that drew on pre-existing representations of Persians in local art.
Over the course of the second and third centuries A.D., representations of conquered peoples became increasingly more typecast, and the Dacian captives in the Forum of Trajan provided a leading model. 52For example, two colossal pavonazzetto statues (original H. c. 3.20 m) from Ephesos adopt the gural type. 53The statues were incorporated into the façade of the East Gymnasium, a complex constructed during the Severan period. 54The better preserved of the two statues, now in I ̇zmir, includes a hexagonal shield resting against the left leg, and next to it, a bow and quiver.Given the architectural context, the statues have been plausibly identied as prisoners commissioned to celebrate the victories of Septimius Severus in Parthia. 55The presence of the bow would support this interpretation because Roman authors describe it as the weapon of choice for Parthians. 56The Arch of Septimius Severus in Rome (dedicated in A.D. 203) drew on similar visual models.The Parthian prisoners on the arch are dressed in a manner comparable to the Agora sculpture.They wear trousers and long-sleeved tunics that terminate below the knees.One prisoner, being led in chains, wears the cloak over both shoulders, with the clasp arranged below the sternal notch.These gures from Ephesos and Rome echo the general remarks of Roman authors, who describe Parthians as wearing loose-tting garments with long robes that cover their legs. 57The main intention of the sculptors of these monuments was not to depict reality, but to create a readily identiable image of an eastern foe, and in particular, to associate the Parthians with the Persians.Nowhere could this equation be more salient than in Athens.

V A TRAJANIC VICTORY MONUMENT IN ATHENS
The gural type represented on the sculptor's sketch, together with the legacy of Athens as a memorial setting for Persian defeat, open up the possibility that the tripod monument commemorated Trajan's victories in Armenia and Parthia.A historical outline of Trajan's Parthian war can be reconstructed from the histories of Cassius Dio (68.17-33), whose text was epitomised by the historian Xiphilinos in the eleventh century. 58e are told that, some time after dedicating the column in his imperial forum in May A.D. 113, Trajan departed Rome to conduct a campaign against Armenia and Parthia on the grounds that the Parthian king Osroes had violated an agreement with Rome by independently installing a new king in Armenia.On his way east, Trajan stopped in Athens, where he received an embassy from Osroes (Cass.Dio 68.17.2). 59 The Parthian delegation pleaded for peace, but Trajan reserved judgment and proceeded to Syria.
By the autumn of A.D. 114, Trajan had entered Armenia and declared the region a province (Fig. 2).In recognition of the annexation, the senate honoured Trajan with the title of Optimus.Trajan then invaded northern Mesopotamia, making it a province too.Despite a disastrous earthquake in Antioch in the winter of A.D. 115/116, Trajan continued his campaign, marching south to the Parthian capital Ctesiphon and capturing it.The senate subsequently bestowed on Trajan the title of Parthicus in February A.D. 116, and the conquest was commemorated on Roman coinage (Fig. 10).Trajan later travelled further south to the Persian Gulf.According to Cassius Dio (68.29.1), the emperor, while standing on the seashore, recalled the achievements of Alexander, remarking 'I should certainly have crossed over to the Indian people, too, if I were still young' (πάντως ἂν καὶ ἐπὶ τοὺς Ἰνδούς, εἰ νέος ἔτι ἦν, ἐπεραιώθην).Although he had not reached India, Trajan had brought the Roman empire to its greatest geographical extent (Fig. 2).Yet Roman control of these newly acquired regions was short lived.A series of revolts followed, and in A.D. 117, Trajan was forced to depart for Italy due to an illness, dying en route in Cilicia.His successor and adopted son Hadrian relinquished Armenia and Mesopotamia, re-establishing the Euphrates River as the eastern boundary of the empire.

Memorials of Trajan's visit to Athens
Athens was an appropriate location for a memorial that celebrated Trajan's accomplishments in Parthia, for several reasons.First, as discussed above, Trajan departed for Parthia from Athens, where he had received an embassy from king Osroesit was Trajan's only known visit to Athens, 60 and the rst recorded imperial visit to the city since Augustus' nal stay, over 130 years earlier. 61On these grounds alone, the emperor's presence must have drawn great attention. 62The most visible impact of the visit was probably the substantial imperial entourage and the infrastructure required to support it.James Oliver suggested that Trajan was accompanied by a large military force, evidenced by epitaphs that marked the graves of  Moreover, it is possible that the designation ἀνίκητος, or invincible, recalled honours for Alexander the Great, who had conquered Persia over four hundred years earlier.
According to a fragmentary speech of Hypereides (5.32) against Demosthenes, a bronze statue of Alexander was proposed in Athens naming him 'king [and] invincible god' (εἰκό[να Ἀλεξάν]δρου Βασιλ̣ [έως τοῦ ἀνι]κήτου θε[οῦ). 70We have no evidence that such a statue of Alexander was actually erected, 71 but if it was, Trajan's new statue was surely in dialogue with it. 72Plutarch (Alex.14.4) records that, before Alexander departed for Persia, the oracle at Delphi proclaimed to him, 'You are invincible, O child!' (Ἀνίκητος εἶ, ὦ παῖ) (see also Diod.Sic.17.93.4).Whatever the historicity of the oracle, 73 Plutarch shows that the story was current in Trajan's day.Indeed, Trajan seems to have admired Alexander and cultivated his memory. 74During his campaign in Parthia, Trajan stopped at Babylon and sacriced to Alexander in the room where he died (Cass.Dio 68.30.1).
Another statue of Trajan may have been erected around the same time in the lower city.Fragments of an over-life-size Pentelic marble statue of an emperor (original H. c. 2.30 m) (Figs 11-12) were excavated from the north stoa of the Library of Pantainos, in a room that opened onto the street joining the Agora with the Roman market (Fig. 13, no. 1). 75The emperor is represented as a victorious general.An imprisoned male gure crouches at his side, with one knee on the ground, looking sharply upward.The emperor wears ceremonial military costume, including the cuirass and the paludamentum that hangs freely from the left shoulder.The breastplate depicts Athena being crowned by winged nikai, and below this main scene, a cosmic personication spreads his arms in a supporting gesture, referring to the breadth and stability of Roman rule.Sheila Dillon recently assigned additional fragments to the statue, including the right shoulder and separately attached arm. 76Dillon's research demonstrates that the arm was outstretched, with the hand grasping a small orb, further communicating the authority of the emperor.The quality of workmanship is exceptional.Great care was expended on the surface textures of the garments, in particular.The statue was a magnicently carved agent of imperial power in Athens.
T. Leslie Shear, Jr, identied the emperor as Trajan and the kneeling gure as a Dacian.Shear reasoned that the group had been displaced from an imperial shrine, which he proposed to locate in the next room to the east (Fig. 13, yellow). 77The adjacent room was set off architecturally from the rest of the complex (Fig. 14).Its entrance had a temple-like façade, with more elaborate column bases and wider intercolumnations than the stoa from which it projected.The spaces between the columns were occupied by statue bases, as demonstrated by the lack of wear in the places they were once positioned (Fig. 14, bottom).One of these footprints matches the dimensions of a base for a statue of Trajan that was found nearby in a re-used context (Fig. 13, no.2); probably the base was originally positioned in the colonnade, at the entrance of the shrine. 78The base records the dedication of a statue, c. (Fig. 13, no.3). 79A fragmentary plaque for attachment to a base of a statue of Trajan (c.A.D. 98-117) was found in the northwest corner of the Library of Pantainos (Fig. 13, no.4). 80Dillon has presented pieces of a second marble statue of an emperor from the area, but its poor preservation thwarts an identicationthe armoured gure presumably represents Trajan or Hadrian (Fig. 13, no.5). 81In all, no fewer than three, and perhaps as many as ve statues of Trajan are witnessed along the street leading from the Agora to the Roman market.We recall that the Library of Pantainos complex itself was dedicated to Trajan, c. A.D. 98-102, together with Athena Polias and the city of the Athenians (IG II/III 3 4,2 1405). 82et us return to the kneeling prisoner as a local expression of Trajanic victory iconography (Fig. 15).The ethnic identity of the gure is difcult to pin down.Representations of kneeling captives appear on coins of Trajan only after he was engaged in war in Dacia in A.D. 101-102, and the motif is deployed most frequently on coins after A.D. 102, following his rst campaign in that region. 83This evidence suggests 79 Verdélis 1947-1948: 42-6, no.2; Shear, Jr 1973a: 176; Højte 2005: 386-7, no.Trajan 100; Camia 2011: 31,  n.58.The current whereabouts of this base are unknown.that the statue group dates to after A.D. 102. 84Like the statue on the Acropolis, it is probable that Trajan's visit prompted its erection.The very high quality of the statue certainly supports the hypothesis.The surface of the statue was painstakingly nished, perhaps with the intention that it would be seen by the emperor himself.In this scenario, the captive emphasised the recent triumphs of the emperor in Daciaa victory not referenced in the dedicatory inscription on the Library of Pantainos, which excludes the title Dacicus.Another interpretation is that the captive is a Parthian.In this respect, it is worth pointing out that the mantle, secured at the sternal notch by a large disc bula, matches the conguration on our sculptor's sketch.Trajan, more than for any other previous or subsequent reign, the gure of the defeated barbarian, in this case the male Dacian, was to become virtually a representative symbol of the reign itself, particularly on coinage.' 84 If so, it is not possible to associate the gure with either of the statues donated by Tiberius Claudius Atticus, which are dated earlier on the basis of the imperial nomenclature.The improbability that the statue and the base belong together was already pointed out by Shear, Jr 1981: 371, n. 63.Cf.Dillon 2022: 80-1, who hypothesises the association of the statue with one of these bases.
and Parthian.That said, at Athens, where there existed a long tradition of Persian defeat, local audiences may have preferred, consciously or not, the latter identication.

The Persian Wars tradition
Athens curated and promoted the memory of Greek triumph over eastern foes, most especially the Persians. 85The Acropolis became a favoured setting for celebrating these victories, both mythical and historical.Alexander the Great dedicated Persian armour to Athena, after defeating Darius' forces at the Granicus River in 334 B.C. (Arr., Anab.1.16.7;Plut., Alex. 16.8).Some of the captured shields, with their origins triumphantly inscribed, were possibly hung on the architrave of the Parthenona building that was itself a testament to Persian defeat. 86Later, an Attalid king dedicated sculptural groups on the Acropolis that linked a series of famous victories: a gigantomachy, an amazonomachy, the battle against the Persians at Marathon and a battle against the Gauls in Mysia (Paus.1.25.2). 87 Some time between 27 and 18/17 B.C., the Athenian demos erected a round Ionic building, or monopteros, to Roma and Augustus on the Acropolis (IG II 2 3173). 88The reasons for the dedication are not well established, but its ultimate effect seems clear.The monopteros was very probably nished by 19 B.C., when Augustus, having regained the Roman military standards from Parthia, passed through Athens on his return journey to Rome. 89The location of the monopteros, in front of the east end of the Parthenon, integrated the recent accomplishments of Augustus into the wider memory landscape of Persian defeat.Beyond the Acropolis, in eastern Attica, the cult of Livia was installed in the Temple of Nemesis at Rhamnous, most probably in connection to vengeance over eastern foes. 90Nemesis, the goddess of divine retribution, had helped to deliver a decisive victory over the Persians at the Battle of Marathon.
In conception, the tripod in the Olympieion recalled the golden (or gilded) tripod erected at Delphi from the spoils of the Persian defeat at Plataea in 479 B.C. (Hdt.9.81.1;Thuc.1.132.2;Dem.59.97; Diod.Sic.11.33; Paus.10.13.9). 91The quotation of this monument, one of the most celebrated war memorials in Greece, may further explain why the Athenian tripod drew the interest of Pausanias.The inclusion of a Trajanic victory monument in the text certainly ts with his interests in the juxtaposition of the Greek past with more recent events.Pausanias' description of the tripod immediately follows extended comment on a statue of the fourth-century B.C. speech-writer Isocrates, who had advocated intensely for an Athenian-led campaign to liberate the Greek cities of Asia Minor from Persian rule.Pausanias uses the monuments to draw a contrast: the conquered easterners are placed in a position of perpetual architectural servitude, bearing an offering, while Isocrates is elevated on a high column.The inclusion of the tripod monument in Pausanias' text is, however, more than just carefully crafted allusion: it reects a built landscape that propagated living memories of Persian defeat. 85Spawforth 1994. 86Stewart 2004. 88 The circumstances of the dedication and the function of the building are not well established; for a range of interpretations, see e.g.Hoff 1989: 5-6;Morales 2017: 145-7. 90 Kajava 2000: 48-60;Lozano Gómez 2002: 60-3;Stafford 2013.On the date of the dedicatory inscription on the architrave of the temple (IG II 2 3242), see Schmalz 2009: 103-5, no.132 (late Augustan), with discussion of earlier research. 91Stephenson 2016.Note also the fourth-century B.C. acanthus column at Delphi, which comprised three female gures supporting a tripod: Martinez 2021.
But, cloud-wrapped Lord, entrust to him, too, the glorious accomplishment of this Achaemenid war, that your heart's joy may be doubled as you look on the spoils of both foes, the Getae [i.e., Dacians] and the Arsacids [i.e., the Parthian dynasty]. 92e could readily imagine that Trajan had made a similar vow to Zeus Olympios in Athens, before departing on his expedition.

VI COMPLETION OR COMMISSION UNDER HADRIAN?
The period between the presentation of the title of Parthicus to Trajan in February A.D. 116 and the emperor's death in August A.D. 117 leaves little time to organise the construction of an elaborate victory monument in Athens.There is, in fact, no surviving memorial anywhere for Trajan's Parthian War. 93According to Cassius Dio (68.29.3), commemorations had been planned in Rome.Following the capture of Ctesiphon, a triumphal arch (ἁψῖδα … τροπαιοφόρον) was commissioned in the capital in honour of Trajan, and 'many other' tributes were planned in his forum; how those efforts materialised is unknown.Trajan was commemorated in Rome after his death with celebration of games called the Parthica (Cass.Dio 69.1.3).
A scenario deserves consideration with regard to the compressed timeline: did Hadrian complete, or even commission, the tripod monument after the death of Trajan?The emperor's unexpected death fuelled suspicion about Hadrian's legitimacy as successor.Cassius Dio (69.1.1-4)presents unease regarding the circumstances of Hadrian's adoption by Trajan, insisting that it was Trajan's wife Plotina who had made the arrangement (see also SHA, Hadr.4.10).Some degree of controversy would help to explain Hadrian's special attention to his predecessor during the early years of his reign.In Rome, for example, Hadrian enlarged the Forum of Trajan with the construction of the temple for his adoptive father (SHA, Hadr.19.9). 94Amanda Claridge argued that Hadrian also commissioned the narrative frieze carved on the Column of Trajan as a modication to transform the structure into the emperor's tomb. 95In Pergamon, the porticoes framing the Temple of Zeus Philios and Trajan were completed during the reign of Hadrian. 96Furthermore, the statue group inside the temple, at rst comprising gures of Zeus and Trajan, was recongured with the addition of a statue of Hadrian.The group presented, in acrolithic form, divine and familial ties. 92Trans.Paton 1916, with modications. 93Lightfoot 1990: 115. 94Schowalter 1998: 247-8;Burrell 2004: 315.In Athens, too, Hadrian actively promoted the memory of his adopted father.A remarkable example of the carefully curated dynastic relationship was on display on the Athenian Acropolis.An inscribed plaque, once afxed to a base for a statue of Hadrian, declares the emperor the 'son of god Trajan Parthicus Zeus Eleutherios' (IG II 2 3312 + 3321 + 3322).Antony Raubitschek, who restored the inscription, argued that the plaque could belong to the base of the statue (εἰκόνα) of Hadrian that Pausanias (1.24.7) recorded inside the cella of the Parthenonthe only portrait statue that the ancient traveller remembered seeing in that space. 97The identication of Trajan with Zeus Eleutherios ('of freedom')the god who helped the Athenians defeat the Persians at Plataealikely derived from the emperor's victories in Parthia, as suggested by the title Parthicus, the only part of the imperial nomenclature included in the text. 98If it were displayed inside the Parthenon, then the inscription would, in effect, have elevated Hadrian as the brother of Athena, both being children of Zeus, as Raubitschek observed. 99The inscribed text thus performed double duty: to emphasise the legitimacy of Hadrian and to exploit a ctitious genealogy that promoted sacred bonds between Athens and Rome.The relationship was evoked in the lower city as well.Pausanias (1.3.2) records a statue of Hadrian that stood in front of the Stoa of Zeus Eleutherios alongside a statue of the god. 100 The large marble torso of Hadrian found nearby is surely the same statue, given the nd-spot and scale (original H. c. 2.75 m). 101On the cuirass of that gure, Athena is crowned by nikai while standing on the she-wolf of Rome.
There are further indications that the tripod monument may have been dedicated after the death of Trajan.First, Pausanias (1.18.6) associated the precinct of the Olympieion with the interventions of Hadrian, not least because of the many statues of the emperor he saw there.He and other authors record that Hadrian oversaw the dedication of the Olympieion (Paus.1.18.6;Philostr., V S 533; Cass.Dio 69.16.1;SHA, Hadr.13.6), even though the superstructure of the temple seems to have been largely completed before his reign.Second, Pausanias (1.18.9) indicates that the use of pavonazzetto was extensive at Athens under Hadrian, who had gifted 100 columns of Phrygian stone to Athens for the construction of a library complex (Section II).It is plausible, therefore, that the monument was completed early in the reign of Hadrian.Whether it was commissioned initially by Trajan or posthumously in his honour cannot be determined on the present evidence.The war in Parthia was, at any rate, of special signicance to Hadrian, who had accompanied Trajan on the campaign.

VII THE LEGACY OF THE TRIPOD MONUMENT
The existence of the sculptor's sketch in a mid-third-century A.D. context suggests that the tripod drew admiration in its later life, evidently enough to be desirable in small format for private consumption (Section III).The sketch reveals that it was not only the sculptured monuments of Classical and Hellenistic Athens that were copied.The motif of the standing captive easterner was redeployed elsewhere in Achaia, asserting the wide resonance of the visual model.The colossal male architectural supports of the so-called Captives' Façade at Corinth share general similarities with the gure represented on the sculptor's sketch.The colossal gures have been assigned to the south side and main entrance of the basilica 97 Raubitschek 1945: 130. 98 Raubitschek 1945: 131.See also Geagan 1984: 77. 99 Raubitschek 1945: 130. 100 For imperial cult in the annex of the Stoa of Zeus Eleutherios, see Thompson 1966;Thompson and Wycherley 1972: 103;Price 1984: 141-2;Camp 2010: 74-5. 101 Athens, Agora S 166; see Shear 1933b: 178-83, no. 5, gs 8-10, pl.VI; Harrison 1953: 71-4, no.56, pls 36-7; Camp 2010: 63-4, no.16, g.33; Karanastasi 2012-2013: 358, no. 1, pl. 1.1-3.located along the Lechaion Road, which opened onto the Forum. 102The upper colonnade of the façade seems to have comprised four to eight gures engaged to rectangular piers, each with its own gural base and Corinthian capital.The gures represent both male and female subjects, presumably alternating in the composition.
The most complete gure stands 2.57 m high (Fig. 16).The gure is dressed in eastern attire: a thin, sleeved tunic with trousers; a heavier, loose garment tied over the waist; and a back-mantle, clasped in a central position below the neck.One arm is crossed over the torso; the other was probably raised toward the chin.The captive has curly, shoulder-length hair and wears a pointed cap made from soft fabric.A relief gure on a base for one of the engaged statues shows a captive male gure with hands in a different position: crossed over the waist.The date of the façade has not been resolved, but there is general agreement that it was erected in the mid-second to early third centuries A.D.
Researchers have associated the façade with the Parthian victories of Lucius Verus or Septimius Severus. 103

VIII CONCLUSION
This article has corrected the identication of a limestone sculptor's sketch from the Athenian Agora as a captive male gure supporting the leg of a tripod.Specically, it has proposed that the sketch represents the tripod recorded by Pausanias in the Olympieion.The newly revealed iconography of the supporting gures, coupled with Pausanias' identication of their material as Phrygian stone, has led to the conclusion that the tripod monument was dedicated following Trajan's victories in Parthia.The emperor's military achievements in the region, however eeting, warranted commemoration: Trajan brought the Roman empire to its largest extent, and he was the rst Roman to take the Parthian capital of Ctesiphon.The tripod monument was probably completed after the death of the emperor.Whether the dedication was commissioned by Trajan or Hadrian cannot be answered on the present evidence; arguments have been presented above for both scenarios.The monument sheds new light on Trajanic Athens, and in particular on the effect that the emperor's visit in A.D. 113 had on the city.Trajan's decision to meet the Parthian embassy in Athens, where victory over the Persians was celebrated widely, was deliberate.The erection of the tripod in the Olympieion positioned the accomplishments of Trajan in this centuries-old tradition.

FIG. 2 .
FIG. 2. Map of the Mediterranean basin showing the approximate extent of the Roman Empire in A.D. 117 and locations discussed in the text.(Drawing: T. Ross)

FIG. 3 .
FIG. 3. Plan of the Athenian Agora, with the ndspot of the sculptor's sketch indicated by the red arrow in grid square Q 14. (Plan: ASCSA, Agora Excavations)

FIG. 5 .
FIG. 5. Detail of the gure on the sculptor's sketch from the Athenian Agora.(Photo: C. Mauzy.Ephorate of Antiquities of Athens City, Ancient Agora, ASCSA: Agora Excavations.© Hellenic Ministry of Culture and Sports/Hellenic Organization of Cultural Resources Development (H.O.C.RE.D.)) FIG. 6. Proposed reconstruction of the tripod monument.(Drawing: T. Ross)

FIG. 10 .
FIG. 10.Aureus of Trajan representing, on the reverse, Parthian captives seated beneath a trophy.London, British Museum R.7740.(Photos: © The Trustees of the British Museum) FIG. 11.Reconstructed statue of Trajan with a kneeling captive, probably from an imperial shrine located between the Agora and the Roman market, Athens; position of the left arm unknown.(Drawing: B. Martens and T. Ross.Photos: C. Mauzy.Agora Excavations; Ephorate of Antiquities of Athens City, Ancient Agora, ASCSA: Agora Excavations © Hellenic Ministry of Culture and Sports/Hellenic Organization of Cultural Resources Development (H.O.C.RE.D.))

FIG. 14 .
FIG. 14. Façade of room 3 of the south street stoa, probably used as an imperial shine; lower drawing shows footprints of statue bases on the stylobate.(Drawing: W. B. Dinsmoor, Jr, ASCSA, Agora Excavations) FIG. 15.Details of the kneeling captive from the statue of Trajan.(Photos: C. Mauzy.Ephorate of Antiquities of Athens City, Ancient Agora, ASCSA: Agora Excavations, © Hellenic Ministry of Culture and Sports/Hellenic Organization of Cultural Resources Development (H.O.C.RE.D.))

59
Graindor 1931: 25.We do not know where in Athens the meeting between Trajan and the Parthian embassy occurred, but I wonder if the parties convened in the Olympieion.For a reconstruction of the route taken by Trajan from Rome, seeStrocka 2017: 399.