ARAM, 23 (2011) 211-251. doi: 10.2143/ARAM.23.0.2959658

ROMAN HIPPO-STADIA: THE “” OF RECONSIDERED IN LIGHT OF THE HERODIAN HIPPO- OF MARITIMA

Prof. JOSEPH PATRICH (The Hebrew University of )

On terminology: The composite term hippo-stadium was coined by John Humphrey to denote a small, stadium-size hippodrome, like the Herodian multi-purpose structure of (Fig. 1).1 The U-shaped entertainment structure of Gerasa (Fig. 2), dubbed “hippodrome” when published,2 is even smaller, and likewise it was multi-purpose, serving for both gymnika and hippika. Hence it was classified by Humphrey in one group, together with the hippodrome of (Fig. 3).3 The Herodian stadia (Table 1),4 where horse and races were also held as well, preceded these structures by a century and a half. The most elaborate among them was

1 The full term coined is: ‘Amphitheatrical’ Hippo-Stadia: J. Humphrey, “Amphitheatrical’ Hippo-Stadia,” in: A. Raban and K. G. Holum (eds.), Caesarea Maritima. A retrospective after two millennia (Leiden–New York–Köln 1996), pp. 121-129. 2 E.B. Müller, II. The Hippodrome, in: C.H. Kraeling, Gerasa. City of the , New Haven 1938, pp. 85-100. In his magnum: Roman , for Chariot Races, Lon- don 1986, pp. 495-504, Humphrey noted (p. 498) that it is the smallest hippodrome on record in the empire. 3 So far this is the only stadium, apart from those in Caesarea Maritima and Gerasa, in which starting gates (carceres) for chariot races were found. The Roman stadia of (early third c. CE, but seemingly never completed) and Scythopolis (second c. CE), both cities of the Decapolis, might have also belonged to the hippo-stadia group, but so far no starting gates have been found there, hence their inclusion in the group of hippo-stadia rather than being simply stadia is uncertain. However, arenas of regular stadia, serving athletics, were just 30 m (ca. 100 feet) wide or less; while hippo-stadia had an ca. 50 m wide. The arena at Scythopolis was 48 m wide, and blocks that may be attributed to the far turning post (meta prima) suggest that the structure there was indeed a hippo-stadium (see below). In addition, make-shift mechanism (mangenon in Gr.) for starting gates could be easily installed in a stadium for an occasion (Humphrey, supra note 1, p. 124). A stadium, not yet explored, also existed in (Shahba) in Arabia. Agonistic inscriptions and literary sources also record athletic contests in , Gaza, Askalon, and Caesarea Philippi (in addition to Caesarea Maritima and Scythopolis). See Aupert and Callot, infra note 56, pp. 186-89. For Gaza see: Ch. Wallner, “ Zur Agonistik von Gaza,” Zeitschrift des Papyrologie und Epigraphik 135 (2001), pp. 125-135; Z. Weiss, “Games and Spectacles in Ancient Gaza: Performances for the Masses held in Buildings now Lost,” in: B. Bitton-Ashkelony and A. Kofsky (eds.), Christian Gaza in Late Antiquity, Leiden and 2004, pp. 23-40. 4 J. Patrich, “Herodian Entertainment Structures”, in: D.M. Jacobson and N. Kokkinos (eds.), Herod and . [Papers Presented at the IJS Conference, 21st–23rd June 2005, Leiden – Boston 2009, pp. 181-213, 455-467.

94097_ARAM_23_11_Patrich.indd 211 11/02/13 08:21 212 ROMAN HIPPO-STADIA

that of Caesarea Maritima, U-shaped as well.5 A more detailed comparative study of the group as a whole will permit us to trace the architectural evolution in time within this typological group since Herodian times into the second and third centuries. In fact, such a study is a desideratum, since more than sixty-five years have elapsed since archaeological works had started in the hippodrome of Gerasa, being the first of its sort to be uncovered in our region. Since then other hippo-stadia have been excavated, increasing our data-base for such a study. That of Gerasa is the best preserved; that of Caesarea is the most thoroughly excavated. One may elucidate points absent in the other. A reconsideration of the Gerasa hippo-stadium in the framework of such a comparative study can also solve the enigma of the underground structure in the arena there, and also the question of whether a solid barrier existed in the small . A small hippodrome is not a Roman , in spite of the fact that both have starting gates. The had an arena of 580 ≈ 79 m. It was narrower than the hippodrome of Olympia,6 the length of which was ca. 3 stadia (540 m), affording 48 or even 60 to compete abreast.7 Both structures, and many other Roman circuses inspired by the Circus Maximus (see Table 2), were much larger than the Hellenistic and Roman stadia of the East (see appended Tables 3-7). The dimensions determined the sort and scale of the races; in a (or in the hippodrome of Olympia) the racecourse was much longer per lap and more chariots could be accommodated in a more spacious venue. The greater distance separating the actual events and the spec- tators was compensated for by other means, such as more opulent decorations and more crowded events. The moderate dimensions set the small hippodromes of the East as a group apart, different from the large Roman circuses on the one hand, and from other stadia types (see below), where no chariot races were

5 Y. Porath, “Herod’s ‘’ at Caesarea: a multipurpose entertainment building,” in J. H. Humphrey (ed.), The Roman and Byzantine Near East: some recent archaeological research [Journal of Roman archaeology supplement series 14], Ann Arbor, MI 1995, 15–27; idem., “Herod’s ‘amphitheater’ at Caesarea (preliminary notice),” ‘Atiqot 25 (1995) 11*–19*; idem., “Herod’s ‘amphitheater’ at Caesarea,” Qadmoniot 29/112 (1996) 93–99 (Hebrew); idem, “Theatre, Racing and Athletic Installations in Caesarea,” Qadmoniot 36/125 (2003), pp. 25-42 (Hebrew); J. Patrich, “The Carceres of the Herodian Hippodrome/Stadium at Caesarea Maritima and con- nections with the Circus Maximus,” Journal of Roman Archaeology 14 (2001), pp. 269-283; idem, “Herod’s Hippodrome / Stadium at Caesarea and the Games Conducted Therein”, in L.V. Rutgers (ed.), What has to Do with Jerusalem. Essays in Honor of Gideon Foerster, Leuven 2002, pp. 29-68; idem., “More on the Hippodrome-Stadium of Caesarea Maritima: a response to the comments of Y. Porath,” JRA 16 (2003), pp. 456-459; idem, “On Circus Carceres and a third farfetched hypothesis. Comments on Y. Porath’s article in Qadmoniot 125 ‘Theatre, Racing and Athletic Installations in Caesarea’,” Qadmoniot 36, no. 126 (2003), pp. 119-120 (Hebrew). See also idem, Studies in the Archaeology and History of Caesarea Maritima: Caput Judaeae, Palaestinae, Leiden-Boston 2011, pp. 177-204. 6 Humphrey, supra note 2 (1986), pp. 132-294. 7 Humphrey, supra note 2 (1986), pp. 6-11; H.A. Harris, “The starting gates for chariots at Olympia,” and 15 (1968) 113–26.

94097_ARAM_23_11_Patrich.indd 212 11/02/13 08:21 J. PATRICH 213

held, on the other. The study of ancient entertainment structures should not remain a stagnant field. Applying the term hippo-stadium reflects the advance- ment in the study of sports and entertainment structures in their cultural setting at large, resulting from new archaeological discoveries like the one in Caesarea Maritima. Such terminology permits a more precise discussion and gives expres- sion to more refined argumentation. The general terms ‘hippodrome’ and ‘stadium’ are too simplistic and general; they became imprecise and anach- ronistic. The composite term ‘hippodrome/stadium’, or in short hippo-stadium acknowledges the appropriate architectural affiliation between both, and its dual function as both hippodrome and stadium. It also acknowledges the differences between the composite structure and a fully fledged hippodrome / circus on the one hand, and a regular stadium on the other. Naming such structures ‘circuses’8 fails to recognize all these features. We are not dealing with a barren terminology here; it conveys different cultural affiliations – either to the agones in the stadia of the Greek speaking Hellenistic east, or to the imperial Circus Maximus of Rome. The cultural settings were decisively different.9

The date: The Herodian hippo-stadium of Caesarea was inaugurated in 10/9 BCE. E.B. Müller, who was in charge of the publishing of the Anglo- American excavations at the hippodrome of Gerasa in the late twenties and early thirties of the last century, had suggested that the structure was built in the late second or early third c., and might never have been completed, and that in any case it was later than the triumphal arch (erected in 129-130),10 while Horsefield preferred a date antedating the triumphal arch by sixty or seventy years.11 After further excavations from the 1960’s to the 1990’s, Ostrasz dated the to the second half of the second c., being completed before 212 CE – a terminus ante quem given by four inscriptions on incense altars that stood on top of the carceres, one of which is datable to 209-212 AD.12 Ostrasz’s date is based on a more refined pottery reading, hence it should be

8 As Y. Porath is doing in some of his publications, such as: Herod’s Circus at Caesarea: a response to J. Patrich (JRA 14, 269-83), JRA 16 (2003), pp. 451-55, and in the title of his forth- coming Final Report: Y. Porath et al., Herod’s Circus at Caesarea Maritima. 9 On the differences between chariot races à la Olympia and à la Circus Maximus see Patrich, supra note 5 (2002), pp. 41-46. 10 Müller, supra note 2, pp. 99-100. 11 G. Horsefield, Appended Note on the Hippodrome, in Kraeling, supra note 2, pp. 100-102. 12 A.A. Ostrasz, Excavations and Research at Gerasa Hippodrome 1982-1987, 66 (1989) [ Archaeological Project 1984-1988, II, ed. F. Zayadine], 51-77, p. 71; idem, Exca- vations and Restoration of the Hippodrome at Jerash. A Synopsis, ADAJ 35 (1991), pp. 237-250; idem, “The Hippodrome of Gerasa: a case of the dichotomy of art and building technology,” in: Kh. ‘Amr, F. Zayadine and M. Zaghloul (eds.), Studies in the History and Archaeology of V, Amman 1995, 183-192, p. 184. On the dated inscription see: Z. Borowski, “Inscriptions on Altars from the Hippodrome of Gerasa,” Syria 66 (1989), pp. 81-82. After the initial excavations in 1931 and 1933, excavations were held at the hippodrome in 1963, 66, 82-87, and 89-92.

94097_ARAM_23_11_Patrich.indd 213 11/02/13 08:21 214 ROMAN HIPPO-STADIA

preferred.13 It might have been preceded by the hippodrome of Neapolis, which was built in the first half of the second c. CE, though its plan had perhaps started already under the Flavians, at the end of the first c. CE.14

The sort of games: While having one or more theatres was common in every Roman city, a hippodrome was rarer. However, Gerasa was not a regular city; it was the seat of the financial procurator of Provincia Arabia and his .15 Many of its officiales were soldiers; many of its ruling class had .16 The arena of its hippo-stadium was of small dimensions, which befitted a stadium (see below). The starting stalls (carceres), set in the southern end, indicate that horse and chariot races (hippika) were held there. However, three agonistic inscriptions17 indicate that athletic contests (gymnika) such as running, pankration, and pentathlon, normal contests in a stadium, were also held in Gerasa, and since the small hippodrome is the only sports arena in Gerasa, the gymnika were held in that structure. Hence, the structure under discussion was a multipurpose entertainment structure – a stadium with starting gates for chariot races; a hippo-stadium.

13 I. Kehrberg, “Selected Lamps and Pottery from the Hippodrome at Jerash, Syria 66 (1989), pp. 85-97. 14 Y. Magen, “Shechem,” in E. Stern (ed.), New encyclopaedia of archaeological excavations in the Holy Land, 4, Jerusalem 1993, 1357–58; idem, Flavia Neapolis. Shechem in the Roman Period, Jerusalem 2005, 143-176, p. 171 (Hebrew). Neapolis, located in , not far from Caesarea, had become a Roman city by the time of Vespasian. 15 B. Isaac, The Limits of Empire. The in the East (Rev. ed.), Oxford 1992, pp. 345-347; R. Haensch, Capita Provinciarum. Statthaltersitze und Provinzialverwaltung in der römischen Kaiserzeit, Mainz 1997, p. 244. One of the procurators of Arabia – Gaius Vibius Celer, in office in ca. 160 CE, served earlier in his career as the procurator of the troupes of Gaul, and Britain (procurator ad familias gladiatorias per Gallias et Hispanias et Britta- niam). See Gatier, next note (Syria 1996), inscr. no. 6, pp. 53-55. Bostra, the capital of Provincia Arabia and the seat of the , had a Roman circus. 16 P.-L. Gatier, Onomastique et présence romaine à Gérasa,” in: A. Rizakis (éd.), Roman Onomastics in the Greek East. Social and Political Aspects. Athens 1996, pp. 251-259; idem., Gouverneurs et procurateurs à Gérasa, Syria 73 (1996), pp. 47-56. 17 These are inscriptions nos. 192-194. No. 192 mentions a president of the games (agono- thetes), who also held the office of gymnasiarch; it is dated to the years 105-114 CE, before the structure was built (according to Ostrasz’s chronology). Hence Gerasa must have had a provisional, makeshift racecourse before the more sophisticated hippo-stadium was constructed. It is a decree of the sacred guild of crowned artists in the service of Dionysus to erect a statue in honor of Titus Flavius Gerrenus, the first agonothetes of the games established in honor of Trajan (seemingly associated with the institution of the Province of Arabia in 106 CE), as a token of gratitude for his benefactions for the guild members and their associates, victors and vanquished alike, distributing prizes and holding royal banquets. He had already served as a gymnasiarch – a six-months office in Gerasa according to Jones – earlier than the event. Inscription no. 193, dated to the early 3rd c., mentions a pakratiast, and no. 194, of the early third c. as well, is also a victor list, mentioning victories in pentathlon, hippotrophoi in Hadriana and , and games in Amman and Sidon. The occurrence of Jippotroƒoi is interesting, attesting to chariot-racing. On these inscriptions see: A.H.M. Jones, Journal of Roman Studies 18 (1928), inscr. nos. 14, 44 and 45; pp. 153-56, 173-4; C.B. Welles, The inscriptions, in Kraeling, supra note 2, pp. 442-445.

94097_ARAM_23_11_Patrich.indd 214 11/02/13 08:21 J. PATRICH 215

An inscription from , from the time of Marcus Aurelius, men- tions athletic contests held in Neapolis; hence, its small hippodrome served also gymnika. It was a hippo-stadium as well. Similarly, in Caesarea Maritima, the “amphitheatron” of Flavius , located in the south of the harbour overlooking the sea, was a multipurpose entertainment structure. According to Josephus it served for athletic competitions (gymnika), chariot and horse races (hippika), and for the staging of Roman spectacles: gladiatorial combats (munera), and hunting scenes (venationes).18 Locally it was known as the “great stadium,” or just the “stadium” of Caesarea.19

Location and Orientation: Generally the hippo-stadium was located at the fringes of the city, or outside the wall. Such is the case in Neapolis, Gerasa, Gadara and Scythopolis. The case of Caesarea is exceptional, being located within the city. Here it was Herod’s decision to locate the stadium (and the theatre) next to his palace. The orientation of the hippo-stadium in each city was determined by the local topography. In Caesarea and Gerasa it was built approximately along a N-S axis. In Caesarea the sphendone was in the south; in Gerasa, in the north. The hippo-stadium of Neapolis followed an approximate E-W axiality, and so did the stadium of Scythopolis. The spendone in Neapolis was in the east, while in Scythopolis it was in the west.

The Arena (Table 5): The arena length in Gerasa was 244.05 m; its width 51.28 m in the north, and 49.49 m in the south. Hence, the longitudinal walls did not run parallel. The total external length was 261.42 m, and the external width 76.08 m. The natural terrain slopes from east to west and from north to south; the eastern , built against the slope, was founded on bedrock, which is almost at ground level, while most of the arena, and the eastern cavea was founded on a mound of earth (Fig. 4). The south-west corner – a wall fifteen meters high – eventually tilted and collapsed, perhaps due to the 363 CE earth- quake.20 As a result, erosion of the arena had started. At present only the northern third is leveled. The external dimensions of the hippo-stadium of Neapolis are 330 m ≈ 76 m; the internal dimensions of the arena 270 m ≈ 48.60 m.21 The arena of the Gadara stadium might have been more than 238 m long; the width is unknown.22 The

18 Flavius Josephus, War 1.415; Antiquities 15.341; 16.137-41. 19 War 2.172; in the parallel paragraph Antiquities 18.57 – simply stadium is mentioned. 20 Ostrasz, supra note 12 (1989), fn. 23, p. 73. 21 Magen, supra note 14, p. 145. 22 S. Kerner and A. Hoffmann, “Gadara – Umm Qeis:preliminary report on the 1991 and 1992 seasons,” Annual of the Department of Antiquities of Jordan 37 (1993), 359-394, pp. 360-61, fig. 1, Pls. I.2 and II.1 (the length of the preserved northern long wing given in the text (283 m)

94097_ARAM_23_11_Patrich.indd 215 11/02/13 08:21 216 ROMAN HIPPO-STADIA

arena of the hippo-stadium of Caesarea was the largest: 303 m long and 50.5 m wide (in the south, where the western wing is still preserved; further north it was entirely eroded by the sea, but seemingly it did not ran exactly parallel to the eastern wing, as is the case everywhere). The original arena level was 2.20 m. a.s.l. Later it was elevated by up to 1.20 m in the section adjacent to the starting gates, and by up to 0.40 m further away.

Cavea, vomitoria and scalaria: In Gerasa the cavea was retained on the inside and outside by two parallel longitudinal U-shaped walls, set 9.6 m apart. The internal wall was a podium 2.66 m high, above which the seats were set. The seats rested on arches forming graded transversal vaults springing from transversal walls, uniformly dispersed and forming rooms with rectangular openings equipped with doors to the outside. They served as shops, and later for dwelling places. The short transversal walls were not bonded into the longitudinal ones. The sixteen or seventeen rows of seats were divided into seventeen blocks, with a capacity of ca. 15.000 spectators. At the southern end of the long wings there were stair-towers leading to the carceres roof; there the cavea was bonded to the carceres. Access to the arena was via eight vomitoria and hence to the seats by means of staircases on the left and right. Access to the upper row of seats from the outside was via nine more vomitoria. The main wide gate was on the north, under the sphendone; a ramp of a moderate slope descended there into the arena. In Gerasa the first course of the podium wall has a torus molding. Below, at the bottom of the podium wall, there is a projecting socle. The workmanship of its outer face suggests that the arena was slanted slightly upwards along the chariot’s (counter-clockwise) run; the level of the right (eastern) side of the arena was 35cm lower than the western side, slowing the run of the chariots in the turn at the meta prima.23 The structure of the cavea at Gadara (Fig. 5), seems to be similar to that at Gerasa. Only the northern long wing of the cavea was preserved there for a length of 238 m. In Neapolis, where the stadium was converted to an amphitheatre, the cavea, 13.65 m wide, was retained by three parallel U-shaped walls. On the inner side the walls delineated a long vaulted corridor 3 m wide, and beyond there were cells arranged transversally, each having an opening to the outside. In Scytopolis (Fig. 6), the sphendone in the west and only short sections of the northern and southern wings of the cavea were incorporated in the amphitheatre into which the truncated stadium was converted (like in Neapolis). The state of preservation of the outer parts of the cavea of the

is wrong, as is evident from the scale in Fig. 1; it should read 238 m). The structure is dated early third c., but seems to be never to have been completed. See also plan of Gadara Stadium in C. Bührigs article, page 286 here (ARAM 23). 23 Ostrasz, supra note 12 (1995), pp. 186-187, Fig. 2.

94097_ARAM_23_11_Patrich.indd 216 11/02/13 08:21 J. PATRICH 217

stadium phase is poor. The long wings were only slightly preserved beyond the truncated arena. The cavea is reconstructed as being ca. 16 m wide, retained by three enveloping walls; however, the outer wall is conjectural. In any case, there are no indications of the existence of a longitudinal corridor under the stadium’s cavea; only cross walls running between the longitudinal walls. In the truncated amphitheatre phase the outer peripheral wall went out of use and the cavea was resting only on the two inner walls. The cavea width in this phase was ca. 9.5 m. Under the seats there were vomitoria and other spaces.24 The sphendone was on the west, and the arena widened further east. The arched opening below the sphendone was blocked, since this side was partially rock- cut against the hill side. Hence it did not served as an aditus maximus. This indicates that the passage under the sphendone here, and presumably every- where, had no function in the ceremonial pomp at the opening of the shows or races. The horses and chariots were entering the arena from the carceres side. The podium wall, with a cornice at its top, was 3.2 m high, much higher than in all other hippo-stadia. Only three rows of seats were partially preserved in situ, but it seems that originally there were sixteen-seventeen rows of seats; the upper rows might have been made of wood. In Caesarea the cavea, with a seating capacity of ca. 10,000 spectators, had twelve or thirteen rows divided into eighteen segments (twelve of which are still preserved). The seats were set on top of a U-shaped external wall and an internal podium wall, 1.7 m high, which had projected 1.45 m (and later less) above the arena level. In between, in the east and south, there was a fill of soil, crashed kurkar and stones of different sizes bonded with lime cement, on top of which the seats were set. On the west a longitudinal vaulted corridor ran under the seats. The sphendone, on the south, had an arched entranceway (aditus maximus), 3 m wide, in its centre. The long sides terminated c. 65 m from the starting gates. Originally, access to the seats was by scalaria from the arena level. In a later stage, under the Roman procura- tors, when the ditch which separated the hippo-stadium from the city to the east was filled in, access was also possible from above, along the scalaria. A single vomitorium was located about 115 m away from the southern gate. It ran across the entire width of the eastern gallery. Splitting near the far side into two inclined vaults, it led after two successive 90° turns to a platform for the dignitaries (pulvinar), that was set above the vomitorium. This loggia was usually located opposite the finish line of the races, which was marked in lime across the right side of the arena.

24 Y. Tsafrir and G. Foerster, “The excavations of the Hebrew University of Jerusalem expedi- tion in Beth Shean, 1980–1994,” Qadmoniot 27/107–8 (1994) 113–15 (Hebrew); idem, “Urbanism at Scythopolis-Bet Shean in the fourth to seventh Centuries,” Dumbarton Oaks Papers 51 (1997) Figs. E, 17, and p. 99.

94097_ARAM_23_11_Patrich.indd 217 11/02/13 08:21 218 ROMAN HIPPO-STADIA

The evolution in the cavea substructure of hippo-stadia is illuminating. While in the earlier Herodian stadia (Table 1) there were no seating arrange- ments for the spectators, or they might have been built partially or entirely of wood, the cavea of the hippo-stadium of Caesarea was entirely stone built.25 It was retained there by two U-shaped walls. In the eastern wing and the sphen- done there was a fill between the two longitudinal walls, while in the western wing there ran a long vaulted corridor. In the hippo-stadium of Neapolis, Flavian or slightly later, the seats were set above three U-shaped walls, hold- ing a double envelope: an internal vaulted corridor and an external envelope of transversal cells set radial or perpendicular to the internal envelope, with openings to the outside. In Gerasa, still later, there were two parallel U-shaped walls and transversal cells with openings to the outside were set in between. The partition walls between the cells were not bonded to the longitudinal walls, while in Neapolis some of them were. The cavea of Gadara seems to have been similar to that in Gerasa.

The carceres: These were preserved in Caesarea, Neapolis and Gerasa. In each case the stalls were set along a curved line, with an axis of symmetry deviating slightly relative to the longitudinal axis of the arena. In Gerasa the stalls collapsed in an earthquake and the debris remained in situ, permitting a fairly accurate restoration.26 Five stalls of the double-pilasters type were installed on either side of a higher and wider pavilion. Only the eastern five were in use. On the outside, to the south and west the ground level drops abruptly down; access to the central pavilion was possible only along a ramp ascending from the east. This was the way the chariots entered the arena, circling it in the open- ing ceremonial procession (pompa). The carceres was built on top of a massive foundation wall. Each stall was 2.65 m wide and 4 m deep; the ratio width / hight was 2:3. It was roofed by a barrel vault, delineated by an arch to the front and rear. The dividing pilasters had attached piers with capitals to the front. In Neapolis the carceres had five stalls on either side of a central passage; each 3-3.10 m wide, and of the double-pilasters type. The story of the starting gates of the hippo-stadium of Caesarea, though less preserved, is much more intricate (Fig. 7). Three main phases have been discerned. In phase I (with three sub-phases), the stalls, twelve in number and with no central passage, were of the 4-piers type.27 The three sub-phases can

25 Patrich, supra note 4, pp. 198-210. 26 Ostrasz, supra note 12 (1989), pp. 66-70; idem, “Excavations and Restoration of the Hippodrome at Jerash. A Synopsis,” ADAJ 35 (1991), 237-250, pp. 243-46; idem, supra note 12 (1995), pp. 187-191. 27 There is a series of foundations in a radial layout, designated as phase 0, with top elevations of only 1.50-1.71 m above sea level, and a tight layer of beaten kurkar at an elevation of 1.55- 1.62 m a.s.l. (an intermediary layer in the arena fill), which reached up to the foundations. These

94097_ARAM_23_11_Patrich.indd 218 11/02/13 08:21 J. PATRICH 219

be distinguished from each other by the structure and the ornamentation of the piers, preserved up to one-two courses. In the first sub-phase (Ia; arena level at 2.20 m a.s.l.), built by Herod, the stalls were separated from each other by wooden partitions that ran between the flanking piers.28 In the subsequent sub- phases the piers became more and more monumental. In phase Ic (arena level at 2.50 m a.s.l.), datable to the reign of Agrippa I, the partitions between the stalls were of stone. The ceilings in all three sub-phases were seemingly flat and light, made of wooden beams. In phase I the starting gates were set along a curved line; they were parallel to each other and to the longitudinal axis of the arena. This arrangement indi- cates that the chariots started their course in parallel lanes, as at Olympia. The destination of the chariots was the far turning post – the meta prima. Herod instituted Greek-style races, according to the Olympic style, prevalent in the Hellenistic world, rather than to the Roman style, à la Circus Maximus. Such was their style throughout phase I. Phase I came to an end when a thick wall (W100) was built on top of the stalls, during the Jewish revolt (66-70 CE). Races according to the Roman style, in radial lanes, were introduced to Caesarea only early in the second century CE when, in phase II (dated to Hadrian), the layout of the stalls became radial, rather than parallel. Since in this phase the converging line of the lanes was at the meta secunda, nearer to the carceres, the elevation of the arena at the carceres was raised by ca. 0.7 m (from 2.50 m in phase Ic to 3.20 m a.s.l.).29 This arrangement enhanced the chariots’ run along the shortened distance from the starting line to the converging “white line” (marked in lime, extending from the meta secunda westward, across the right half of the arena). The esti- mated location of this meta is ca. 83 m from the central gate.30 Remains of only five stalls were preserved in situ, but the presumed width of the arena suggests that there were five stalls on either side of a central arch, like in Gerasa. The stalls, of the double-pilasters type, had a semi-column with an Attic base attached to the front. Each pilaster was 1.2 m wide, and its length

lower foundations seem to support a casual wooden carceres, set for a particular, pre-inaugural event. It is not clear how much of the cavea was already standing by that stage. Elsewhere I have suggested that the event in question was the visit of Marcus Agrippa to Herod in Caesarea, in year 15 BCE, when the hippodrome was still under construction, five years prior to its inauguration. See Patrich, supra note 5 (2003), and idem, “Herod’s Hippodrome/Stadium at Caesarea in the Context of Greek and Roman Contests and Spectacles,” in: Y. Ben-Arieh and E. Reiner (eds.), Studies in the History of Eretz Presented to Yehudah Ben Porat, Jerusalem 2003, 119-166, pp. 166-67 (Hebrew). 28 The carceres of the Circus Maximus were of wood and tufa until the time of Claudius, and individual stalls were not properly roofed under Julius and Augustus, becoming large- scale only under Claudius. See: Humphrey, supra note 2, pp. 133-34. 29 Here as well the fill has two layers, the lower one – a packed bedding for stability, reached an elevation of 2.85 m a.s.l. 30 According to Porath, supra note 8 the distance was only 68 m.

94097_ARAM_23_11_Patrich.indd 219 11/02/13 08:21 220 ROMAN HIPPO-STADIA

(without the attached semi-column) was ca. 3 m. Each stall was 3 m wide. The structure, arrangement, and dimensions are similar to the better restorable stalls at Gerasa. Extrapolation indicates that the axis of symmetry of the carceres deviates slightly westward in relation to the axis of symmetry of the arena. The calculation suggests that the pilaster on the western end was located ca. 3 m further north relative to the corresponding one on the eastern end. In phase III (dated to ), also of the double-pilasters type, the pilasters were more massive – 1.4–1.5 m wide and 3 m long, as before. There were four stalls on either side of a central gate; only the four eastern pilasters were preserved. The eastern pilaster was attached to a massive base, seemingly for holding a staircase leading to the roof of the carceres, like in Gerasa. Such might have been the case also on the western end. The width of each stall was 3.5 m. The façade had a vertical band, 0.5 m wide, protruding 0.1 m to the front. Niche-like recesses, 0.65 m wide and 0.45 m deep, were installed in the side walls. The elevation of the arena at the carceres at this phase was 3.40 m a.s.l. The piers were shifted ca. 3 m to the south relative to those of phase II, making the arena shorter accordingly. The carceres at Neapolis was of the double-pilasters type as well. There were five stalls 3-3.1 m wide on either side of a wider central gate. All were seemingly vaulted. The southern pilaster of the central gate is longer than the rest, projecting in and out of the curved strip marked by stalls. The next pilas- ter to its south projects only outwards slightly.

Barrier: In the Circus Maximus a continuous barrier, comprising of a series of elongated pools (euripi) projecting above arena level was built under Trajan.31 All kinds of decorative monuments, such as , sculptures, columns, pavil- ions, fountains and other installations, were added. Earlier the Circus Maximus had a temporary wooden barrier that was set up when races were held. As for Caesarea, the fact that chariots run in parallel lanes, as is suggested by phase I carceres, precludes any possibility for the existence of a continuous solid barrier along the middle of the arena in this phase there. The case is different when the stalls had a radial layout, like in phases II and III of the carceres. However, in

31 A roofed channel ran along the center of the arena of the Circus Maximus for hundreds of years. Permanent monuments began to be erected on this line only in the early Empire. Augustus was the first to erect an Egyptian in the middle of the barrier line, and in the course of the first c. CE more altars and pavilions were added. The euripus is mentioned explicitly for the first time only by Tertullian at the end of the second c. (Humphrey, supra note 2, pp. 11, 292–93). The term spina, mentioned for the first time only by Cassiodorus – a sixth c. author – is widespread in modern research. However, Humphrey (ibid., p. 175) has shown that in the Roman period the term euripus was actually widespread for describing the continuous barrier of the Circus Maximus. The term euripus faithfully reflects the shape of the barrier in the Roman circuses, as preserved in Lepcis Magna, , and in the eastern circus at Caesarea (see next note).

94097_ARAM_23_11_Patrich.indd 220 11/02/13 08:21 J. PATRICH 221

these phases, as in all other hippo-stadia on which we are focussing, there was no solid barrier extending between the two turning posts, since in all of them the arena is quite narrow, barely 50 m wide, too narrow to permit a full-fledged spina like in the Roman circuses of , Carthage, , or on the east of Caesarea Maritima.32 The evidence for a continuous barrier in the hippo-stadium of Neapolis is unambiguous (Fig. 3). It consists of two parallel lines ca. 20 m long of single stones set 3 m apart, some laid as headers, others as stretchers. They were found at arena level near the far end of the arena, starting at a distance of ca. 210 m from the central gate of the carceres. These single course walls supported a barrier, seemingly of wood. In spite of the deliberate efforts of the excavators, no other remains that could be attributed to a barrier were found. The inevi- table conclusion should be that this was the actual length of the barrier; there is no reason to extrapolate it any further, as was suggested in the publication.33 Elsewhere the barrier must have been marked by simple posts, not by a solid structure. In Gerasa no remains of a barrier like those at Neapolis were found. The underground three-chambered structure (Fig. 8) interpreted as a spina is a mis- take (see below). In Caesarea Maritima segments of two single-course rows of stones set 1.9 m apart were found at a depth of more than 1 m below arena level. There is no reason to interpret them as remains of a barrier 205 m long (with the gaps) as was suggested; they are rather the remains of a sub-arena drainage channel.34 In a long circus the chariots could attain high velocity until reaching the break line at the meta secunda: in the Circus Maximus of Trajan, with an area 580 m long, this section was 175-180 m long; in Merida, of 404 m-long arena, it was 127 m, in (492.5 m-long arena), 145 m, in Vienne (441 m-long arena), 138 m, and in the (with an arena 503 m long), it was 160 m long.35 Hence, it cannot be concluded that in Neapolis, where the distance of the solid barrier from the central gate was ca. 210 m, the break line was at the beginning of this barrier, and that up to this line each chariot was

32 The best preserved example is the barrier of the Roman circus at Leptis Magna. See: Humphrey, supra note 2, pp. 25-55. On Anazarbus see: M. Gough, “Anazarbus. General Description of the Site,” Anatolian Studies 2 (1952), pp. 99-101. In Caesarea as well remains of an elongated pool (euripus) with marble sides and a thick layer of plaster floor were found near the meta prima, at the southern end of the local circus. See Humphrey, ibid., pp. 477-491; Y. Porath, “Caesarea – 1994–1999”, Hadashot Arkheologiyot 112 (2000), pp. 38-40 (Hebrew section), 34*–35* (English section); idem, supra note 5 (2003) (Hebrew). 33 Magen, supra note 14, p. 146, Fig. D5. 34 Compare Porath, supra note 8 (2003), with Patrich, supra note 5 (2003). 35 Ostrasz (supra note 26, p. 239) found out that the distance from the carceres to the meta seconda in Roman circuses was ca. 30% of the arena length, the shorter distance being in – just 80 m.

94097_ARAM_23_11_Patrich.indd 221 11/02/13 08:21 222 ROMAN HIPPO-STADIA

restricted to its lane. The break line must have been nearer to the carceres, but has left no physical remains. The exact location must have permitted a reason- able accelerating length up to the break line. But this line was not necessarily set at the meta secunda in Neapolis, but rather nearer to the 20 m long barrier, perhaps coinciding with the “finish line”, located regularly midway between the meta prima and the meta secunda, opposite the judges’ box (tribunal). A simple post set on the line connecting both metae could have marked these coinciding lines. In such a scheme the chariots could attain enough velocity up to the break line, and then the seven laps around the meta prima and the meta secunda were counted. In Caesarea the estimated location of the meta secunda was only 68 m from the phase II starting gates. But the slanting arena between the carceres and the break line enhanced the acceleration rate, thus compensat- ing for the undersized distance.36

Turning posts: The shape and structure of a meta prima is best illustrated in Caesarea. At the centre of the southern end of the arena, about 25 m north of the southern entrance, remains of the meta prima were preserved. Four suc- cessive architectural phases were discerned (matching perhaps phases of the carceres). Each is shifted slightly to the north relative to its predecessor (thus slightly decreasing the arena length). The first was marked by three walls in a somewhat trapezoidal layout, 3 m long, while the successive ones were marked by stone blocks set along a curved outline. Each block had a depression in the center of its upper face, most probably for the insertion of a wooden post. Similar blocks, either circular, like column drums, or rectangular, with depres- sions, were also found in Neapolis (one, square, near the far end of the arena), Gerasa (five, in the underground chambers located in the northern third of the arena; two found 11 m to the south of the northern entrance),37 and Scythopolis (five, near the western end of the arena). Their occurrence in Scythopolis, where no carceres was preserved, suggests that this stadium might also have been a hippo-stadium. No other remains that can be attributed to the meta prima were found in Gerasa.

Shortening of the arena and underground chambers: At present, only the northern third of the arena in Gerasa is leveled (Fig. 9). When recorded it was delineated on the south by a curvilinear wall built of fieldstones. Müller

36 In Caesarea three large blocks of stone each with a central deep depression, perhaps to hold a wooden post, were found near the N end of the actual 25 m long segment of Porath’s barrier. Interestingly, these posts are located opposite the platform for the dignitaries (pulvinar), at a distance of ca. 180 m from the starting gates. Hence they might have marked the end line. Its location in the other hippo-stadia is not known. 37 The two blocks were recorded by Müller, supra note 2, Pl. XVII.b. One of these blocks, found in a secondary use, carried an inscription. Ostrasz (supra note 26, p. 240), attributed the stones to the shortened arena, not to the meta prima.

94097_ARAM_23_11_Patrich.indd 222 11/02/13 08:21 J. PATRICH 223

was of the opinion that the shortened arena should be attributed to the Sassanian occupation (614-627 CE), and that it served as polo (a dubious proposal).38 But when the curvilinear wall was carefully dismantled it turned out that it had been set there much later, in the Ottoman period, and that its purpose was to delineate and retain the oval plot in the deserted arena, for agri- cultural purposes, rather than to demarcate a truncated arena.39 A subterranean structure, 35 m long, holding three compartments 2.5 m deep and of different shape and size, two corridors and a staircase of five steep steps was uncovered in 1982 in the truncated arena (Fig. 8). The workmanship is poor, there is no flooring, and stones from an earlier building were incorporated in the walls in a secondary use. Its northern end is 58 m distant from the sphendone center. Ostrasz interpreted the compartments as three basins (euripi) forming the under- ground part of the barrier.40 But this is a mistake; as was indicated above, euripi were elongated pools constructed above arena level, with no underground spaces underneath. Such euripi set above arena level were found in Leptis Magna, Caesarea (the Roman circus on the east of the city), Carthage and elsewhere. A somewhat similar underground structure, dubbed “sub-arena gallery C1900” comprising of three interconnected elongated compartments set perpendicular to the arena axis, was found under the centre of the truncated arena in Caesarea Maritima, belonging to the amphitheatre phase (Figs. 10, 11).41 Such under- ground spaces, known as amphitheatre carceres, are common in many amphi- theatres.42 Cages for animals and equipment for dramatic effects during the spectacles were regularly stored there. They were roofed by horizontal wooden beams. Such a structure was not a component of a circus spina. Accordingly, the subterranean structure at Gerasa should also be interpreted as an amphitheatre carceres. The truncated arena must have been at least 85 m long, and perhaps some 60 m longer if the underground structure marks the approximate centre of the contracted arena. If so, Gerasa’s would have been the longest truncated amphitheatre in the group under consideration, but not much longer than the longest shortened arena at Caesarea (see Table 8). The concave wall that had delineated it on the south must have been thin, and easily eroded. The conversion occurred in the second half of the fourth c. Epigraphical evidence suggests that this northern part of the Gerasa hippo-stadium was still

38 Müller, supra note 2, p. 99. 39 I. Kehrberg-Ostrasz, email correspondence, November 03, 2008. I am grateful to her for this information, which is going to be published in Vol. I of the forthcoming final report: A.A. Ostrasz (posthumously, with contributions, edited and completed by I. Kehrberg), The excavation, archi- tecture and architectural history of the hippodrome of Gerasa as compared with other Roman circuses. 40 Ostrasz, supra note 12 (1989), p. 62 and Fig. 3; supra note 26, pp. 237-240. 41 Porath, supra note 8. 42 J.-C. Golvin, L’amphithéâtre romain, Paris 1988, Pl. V; VIII.5; XXI.1, 2, 5; XXVII.2 etc.

94097_ARAM_23_11_Patrich.indd 223 11/02/13 08:21 224 ROMAN HIPPO-STADIA

in use for games ca. two centuries after the southern part went out of agonistic use, being occupied by pottery kilns or giving place for domestic use. It was in service as an amphitheater until about the mid sixth c.43 The truncated arenas of the amphitheaters of Neapolis (57 ≈ 39 m) and Scythopolis (95 ≈ 48 m) were delineated by massive cavea. In Neapolis (Fig. 12), the arena was made narrower, and the conversion was dated to the third c;44 in Scythopolis (Fig. 6), it was dated to the fourth c.45 In Caesarea the truncated arena was delineated on the north by four thin and curved walls, 91.5–136.5 m distant from the southern gate (Fig. 10). These walls never attained a monumental character, as was the case in Scythopolis and Neapolis. Therefore, it seems that each wall was set for a one-time event, held for a particular occasion, rather than being a permanent construction. Each of these walls was just one stone (ca. 50 cm) thick, laid as a stretcher, rendering it impossible to stand too high (the higher the wall, the poorer its stability). Such repetitive fragile construction suggests that each wall could easily be dismantled, enabling a re-use of the entire length of the arena. The latest among the walls should be ascribed a date later than phase III carceres; hence – post-Septimius Severus. In the shortened arena – still called a stadium – the persecution of some Christians martyrs took place in the early fourth century.46 The conversion of stadia to by the truncation of the arena is known in several stadia of Asia Minor and Greece. The modification was dated to the Late Roman period (mid 4th and 5th c.), when the athletics declined. with a podium wall and U-shaped hippodromes were shortened, so that part of the arena bordered by the curved end (sphendone) and the nearby galleries of seats were cut off from the rest of the arena by a concave wall and converted into oval amphitheatres.47

43 Ostrasz, supra note 12 (1995), pp. 73-74; I. Kehrberg and A.A. Ostrasz, “A History of Occupational Changes at the Site of the Hippodrome of Gerasa,” in: Studies in the History and Archaeology of Jordan VI, Amman 1997, pp. 167-174. 44 Supra note 14. The case of Neapolis is an enigma, since according to Malalas (Chron- ographia, ed. L. Dindorf, Bonn 1831, p. 446), the Samaritan rebel Julian attended a chariot race in the local hippodrome in year 529, in which Samaritan, Jewish, and a Christian charioteers competed. How could this be, if the hippodrome was already converted into an amphitheater in the third c.? 45 Supra note 24. 46 J. Patrich, “The Martyrs of Caesarea: the urban context,” Liber Annuus 52 (2002), pp. 321- 346. See also idem, Studies in the Archaeology and History of Caesarea Maritima. Caput Judaeae, Metropolis Palaestinae, Leiden-Boston 2011, pp. 259-281. 47 This was the situation in the following stadia: Aphrodisias, , Perge, , Laodicaea, and the Panathenaean stadium in Athens. See: K. Welch, “The stadium at Aphrodisias,” American Journal of Archaeology 102 (1998), 547–69, pp. 565–69. Welch excludes the proposal that the chariot races referred to in few inscriptions from Aphrodisias were held in the stadium; horse and colt races are possible.

94097_ARAM_23_11_Patrich.indd 224 11/02/13 08:21 J. PATRICH 225

CONCLUSIONS

Hippostadia were generally located in the fringes of a city, outside its wall. In Caesarea it constituted a component of Herod’s entertainment compound (together with the theatre), adjacent to his palace, and segregated by a ditch from the city.48 The orientation and the position of the sphendone was dictated in each case by local topography and restrains. The Herodian composite hippo-stadium of Caesarea Maritima – a provincial capital and an early centre of Greco-Roman culture in the east – set the pattern reproduced a century and more later in all other U-shaped Roman stadia of the provinces of Judaea-Palaestina and Arabia. The Roman stadia of Neapolis, Gerasa, Scythopolis, were likewise of the hippo-stadium type. Such was seem- ingly also the stadium of Gadara, though no starting gates have been found there so far. Up to the Jewish revolt the chariot races in Caesarea followed the style of Olympia, with starting tracks set parallel to the longitudinal axis of the arena. In the early second c., under excessive Roman influence, a new carceres was built at Caesarea (carceres II), to serve chariot races in the style of the Circus Maximus, with starting tracks set in a “radial” layout. The carceres of all other hippo-stadia followed this Roman layout. However, there were five gates on either side of a central one. Such an arrangement was not fitted to serve the traditional Roman Circus Faction (which, theoretically, the twelve gates of phase I carceres, and the eight stalls of phase III carceres at Caesarea could accommodate). However, the Circus Factions are not attested in the east before the early fourth c. CE. Like the Herodian hippo-stadium of Caesarea, all were multipurpose entertain- ment structures. Hippika were an important component of the games (as is attested by the carceres of each), but gymnika were held therein as well (as is attested by the literary and epigraphical evidence), maintaining the Hellenic spirit in the East. A stadium was a common component of a Greek city, still preserved in many of them (Tables 3-4, 7); together with the gymnasium, it was an institu- tion for proper training, education, and public life.49 During the Roman period, from Augustan times onwards, this Greek institution, originally of rectangular layout with casual wooden installations and no permanent seats, underwent an architectural evolution and “petrification”, mainly in the eastern provinces out- side Greece (Athens and , rebuilt by Herod Atticus, being the exceptions).50

48 The location on the sea shore is actually also an edge. 49 N.M. Kennell, Ephebeia. A register of Greek cities with citizen training systems in the Hel- lenistic and Roman periods (Nikephoros Beihefte Band 12; Weidmannsche Verlagsbuchhandlung, Hildesheim 2006). 50 P. Aupert, “Evolution et avatars d’une forme architecturale,” in: C. Landes et. al. (edd.), Catalogue de l’exposition: Le stade romain et ses spectacles, Lattes, Cedex, pp. 95-105. K.Welch, “Greek stadia and Roman spectacles. Asia, Athens, and the tomb of Herodes Atticus,” JRA 11 (1998), pp. 117-145. A sphendone first occurred in the late fourth c. BCE, like in Isthmia (the late stadium) and Nemea. Stone seats were gradually added since the second c. BCE (Epidaurus). There

94097_ARAM_23_11_Patrich.indd 225 11/02/13 08:21 226 ROMAN HIPPO-STADIA

The composite hippo-stadium type is characteristic only of the provinces of and Arabia, perhaps due to the fact that Arab horses were at hand, having been bred.51 The impact of the Herodian hippo-stadia, especially the one at Caesarea, might have been a factor as well, since in eastern provinces remote from Herod’s kingdom three other stadium types were built, none with carceres; none a hippo-stadium: – U-shaped, like those of in Epirus (phase I),52 Cibyra,53 Ephesus,54 Athens (the Roman ),55 Delphi (Fig. 13),56 ,57 Perge, Aspendos,58 Kourion59 and elsewhere (see Table 3). – Amphitheatrical, with sphendonai on both ends (amphi-stadia), like those of Aphrodisias (Fig. 14),60 Laodicea ad Lycum,61 Nysa on the Mehander,62 Nicopolis in Epirus (phase II), and elsewhere (see Table 6).

were also Hellenistic stadiums with seats only on one long side, the other being open to the land- scape, like in (lower gymnasium), , Delphi I-II, Halieis, Laubraunda, Cos (original state), Egine (as described by Pausanias II.29.11). See Aupert and Callot, infra note 56, p. 172. 51 In the small hippodrome of Bovillea is associated with the Julia; it was not an urban structure. See Humphrey, supra note 2 (1986), pp. 561-66. 52 F. Krinziger, “Nikopolis in der augusteischen Reichspropaganda,” in: E. Chrysos (ed.), Nikopolis I, 1987, pp. 109-120; K.L. Zachos, Aktia. Athletikoi agones twn autokratorikwn chronwn sth Nikopolh ths ‘Epirou, Athens 2008, pp. 38-46 (in Greek). See also K.L. Zachos and E.A. Pavlidis, “Die frühen Bauten von Nikopolis,” in: R. Asskamp and T. Esch (eds.), – Varus und seine Zeit, Münster 2010, pp. 141-142. 53 The U-shaped stadium of Cibyra, built under Tiberius, has a straight elongated gate house with five vaulted passages resting on six pilasters, resembling a carceres. See Aupert, supra note 48, p. 100; R.G. Chase, Ancient Hellenistic and Roman Amphitheatres, Stadiums, and Theatres. The way they look now, Portsmouth, New Hamphshire 2002, p. 66. 54 The U-shaped stadium of Ephesus, built under Neron, has a straight elongated gatehouse with seven passages, resembling a 4-piers type carceres. See: P. Roos, “Wiederverwendete Start- blöcke vom in Ephesos,” Jahreshefte des Österreichischen Archeologischen Institutes in Wien 52 (1978-80), pp. 109-113; St. Karwiese, ibid. 63 (1994), Grabungen, pp. 21-24; idem, ibid., 64 (1995) Grabungen, pp. 22-23; 65 (1996), Grabungen, pp. 18-20; 66 (1997), Grabungen, pp. 19-22, 67 (1998), Grabungen, pp. 20-23. 55 C. Gasparri, “Lo stadio panatenaico. Documenti e testimonianze per una riconsiderazione dell’edificio di Erode Attico,” Annuarion della Scuola archeologica di Atene e delle missioni italiane in Oriente 52-53 (1974-75), pp. 313-392; K. Welch, “Greek stadia and Roman spectacles. Asia, Athens, and the tomb of Herodes Atticus,” Journal of Roman Archaeology 11 (1998), pp. 117-145. 56 P. Aupert and O. Callot, Fouilles de Delphes, II. 11: Topographie et . Le stade, Paris 1979; P. Aupert, Le cadre des jeux Pythiques, in: W. Coulson and H. Kyrielis (eds.), Pro- ceedings of an International Symposium on the Olympic Games, [Athens] 5-9 1988, Athens 1992, pp. 67-37, Pl. 28-35. 57 E. Akurgal, Ancient Civilizations and Ruins of , 1978, p. 70, Fig. 23. 58 Akurgal, supra note 57, pp. 330 and 334-35, respectively. 59 D. Christou, Kourion. A Complete Guide to its Monuments and Local Museum,3 Nicosia 1986, pp. 46-48. 60 K. Welch, “The stadium at Aphrodisias,” American Journal of Archaeology 102 (1998), pp. 547-569. 61 IG Rom 4.845 and 861 refer to this stadium, of two sphendone, as amphitheatron stadion. Hence amphi-stadium is the correct term to denote such a stadium type, as was already noted by Humphrey, supra note 1, p. 123. For its location on the city plan see: Akurgal, supra note 57, pp. 237. 62 Akurgal, supra note 57, pp. 234-35.

94097_ARAM_23_11_Patrich.indd 226 11/02/13 08:21 J. PATRICH 227

– Stadia with straight or concave longitudinal sides, and open on both ends, with no sphendone, like those of (Fig. 15a) (Hellenistic in its origin) and (Fig. 15b) (with a separate on one end), respectively63 (see Table 7).

The hippo-stadia of Caesarea, Neapolis and Scythopolis were each converted into an amphithater. The comparative study enabled us to solve an enigma of the hippo-stadium of Gerasa: The three underground compartments in the northern part were an amphitheatre-carceres of the truncated arena; they had nothing to do with a barrier (spina) of the elongated arena of the original hippo-stadium. Like the others, the hippo-stadium of Geresa was also converted to an amphitheater. The shortened arena was delineated by a simple wall of a concave layout, like the four in Caesarea, not by a thick cavea as in Scythopolis and Neapolis. A solid barrier, if occurring, was restricted to a short segment at the far end, 20-25 m long, like in Neapolis. The break line might have coincided with the end line, located further away from the carceres than the meta secunda. Evolution in the substructure of the cavea of hippo-stadia was traced in two aspects. While in Caesarea the long sides terminated 65 m from the starting gates, in Gerasa the cavea is integrated into the caeceres by two stair-towers at the ends of the long sides. All components form one structure (like in a Roman theater, where the auditorium and the scaena form one structure, while in a Greek and Hellenistic theaters these components are not bonded together). The second aspect of evolution is expressed in the substructure of the cavea. The foundation of the seats on two parallel walls, having a fill or vaulted pas- sageway in between, like in Caesarea, corresponds with Hellenistic tradition. It appears in the earliest hippodromes and circuses, such as the Hellenistic Lageion of Alexandria and the hippodrome of Cyrene. The substructure of the seats in the Circus Maximus in the days of and Augustus followed a similar arrangement. Later on the arrangement had changed. The typical layout of the seats in the later Roman circuses and stadia (as well as in Roman theatres and amphitheatres) was a network of parallel vaults set per- pendicular to the arena’s long axis. However, the evolution was gradual. In Neapolis there is a longitudinal corridor and transversal cells, while in Gerasa, as well as in Gadara and Scythopolis, only transversal compartments.

63 The stadium of Miletus had a third c. Roman propylon of eight pairs of Corinthian columns supporting arcades, on the east, facing the Hellenistic di-stylos in antis propylon on the other end. See: A. von Gerkan, Das Stadion. [Milet. Ergebnisse der Ausgrabungen und Untersuchungen seit dem Jahre 1899 (ed. Th. Wiegand), II. 1], Berlin 1921; G. Kleiner, Die Ruinen von Milet, Berlin 1968, pp. 110-113 and folded map. A. Hoffmann, “Aizanoi. Erster Vorbericht über die Arbeiten im Stadion 1982-1984”, Jahrbuch des Deutschen Archaeologischen Instituts, Archaeologischer Ainzeiger (1986), pp. 683-698; idem, C. Meyer-Schlichtmann, and H.C.v.Mosch, Aizanoi. Zweiter Vorbericht über die Arbeiten im Stadion 1987, 1988 und 1990, ibid. (1993), pp. 437-473; A. Hoff- mann, “Aizanoi: Zweiter Vorbericht ueber die Arbeiten im Stadion 1987, 1988 und 1990,” ibid. (1993), pp. 437-473; Corinna Rohn, “Die Arbeiten am Theater-Stadion-Komplex von Aizanoi,” ibid. (2001), pp. 303-317.

94097_ARAM_23_11_Patrich.indd 227 11/02/13 08:21 228 ROMAN HIPPO-STADIA

The new arrangement permitted the introduction of vomitoria leading into the arena from the outside. This required the podium wall to be elevated higher. In Caesarea it was just 1.2 m above arena level; in Gerasa 2.66 m, and in Scythopolis 3.2 m. (In Neapolis the podium wall was preserved to a maximal height of 1.5 m; no vomitorium leading into the arena is indicated there on the plan or on the graphical reconstruction (other than the aditus maximus). In Caesarea there were only scalaria leading up to the seats from the arena level, or down from outside, and just a single vomitorium in the middle of the eastern wing. The new arrangement of the cavea substructure, with transversal and radial cells, introduced vomitaria as a common feature. Thus in Gerasa there were seventeen vomitaria, eight leading to staircases at ground level, and nine holding staircases leading up.

Table 1: Arena Dimensions of Herodian Stadia

Type 1: Type 2: rectangular U-shaped – amphitheatrical

Herodium course* 315 ≈ 83 “amphiteatre”**/”hippodrome”** Samaria/ 194.5 ≈ 58.5 Caesarea “amphitheatre”** / 303 ≈ 50.5 stadium* “great stadium”** Tiberias “stadium”**

* Term given by archaeologists ** term employed by Josephus

94097_ARAM_23_11_Patrich.indd 228 11/02/13 08:21 J. PATRICH 229

Table 2: The Circus Maximus and some Roman Circuses inspired by it (arena dimensions and dates)

Site Arena Arena Date Comments length (m) width (m)

Circus Maximus 580 79 Ca. 600 BCE; Monumentalized in Dimensions stone under Trajan under Caesar and Augustus Merida 404 96 1st half of 1st c. CE Antioch 492.5 70-75 TAQ 1st half of 2nd c. CE Anazarbus 410 64 Hadrianic? Cement barrier 200 m long; sphendone on one end Caesarea Maritima 450 90 Hadrianic (circus) Tyre 450 86-92 2nd c. CE Carthage 496 77-78 Early 2nd c. CE Leptis Magna 450 65 161-62 CE Constantinople 411 ~80? Septimius Severus Bostra 420 83-97 3rd c. CE?

94097_ARAM_23_11_Patrich.indd 229 11/02/13 08:21 230 ROMAN HIPPO-STADIA

Table 3: Greek and Roman Stadia (arena dimensions and dates; arranged chronologically and alphabetically)64

Site Arena Arena Date Comments length (m) width (m)

Olympia 192.28 28.26 ~470 BCE Concave long sides (stadium III) Isthmia (early) ? 20.42 6th-5th c. BCE Delphi 177.41 29.7-25 Phases I-III: U-shaped, Hellenistic 4 phases; (~ 275 BCE); stone benches and Phase IV: 4 piers gate only in 167-77 CE phase IV (Herodes Atticus) Epidaurus 181.30 23 Late 4th c. BCE Isthmia (late) 181.20 26.60 4th c. BCE U-shaped (Alexander) Nemea 178 23.5-26.6 330-320 BCE U-shaped, Balbis and hysplex Priene, 190 130 BCE Stadium, starting lower gymnasium line; rectangular Halieis, Argolid 166.5 15.5-16.42 Nicopolis, Epirus 177.60 20 Augustan U-Shaped, later a (phase I) second sphendone was added Cibyra Tiberian U-shaped, gatehouse of five vaulted passages Ephesus Neronian U-shaped, gatehouse of seven passages Rome, 192 53 Before 86 CE U-shaped

64 The list is far from being exhaustive. Aupert, supra note 57, pp. 179-189, lists 208 sites where Greek or Roman stadiums had been recorded in literary sources, inscriptions, or architec- tural remains; 75 of them in Greece and 116 in the Orient (including Asia Minor, Syria, Arabia, Palestine and ). Of these, nothing is known about 44 stadiums, and on 24 structures only the location is known. Since 1979 the numbers can be slightly modified, but still, only few under- went a proper stratigraphical exploration (Olympia, Delphi, Epidauros, Nemea, Aizanoi). The hippo-stadia considered here are not included in Aupert’s studies.

94097_ARAM_23_11_Patrich.indd 230 11/02/13 08:21 J. PATRICH 231

Site Arena Arena Date Comments length (m) width (m)

Athens, ~200 ~30 Herodes Atticus, U-shaped; Panathenaic 144 CE Corinthian propylon Aspendus 2nd c. U-shaped Perge 234 34 2nd c. U-shaped 2nd c. Stadium Stadium; site of Christian Martyrs persecution Kourion, Cyprus 217 27 2nd c. CE U-shaped Caesarea of Equestrian events Cappadocia in the Stadium mentioned by St. Basil (4th c.) Pergamon >280 58*? Roman

Table 4: Greek and Hellenistic Hippodromes (Two stadia and more long, chronologically arranged)

Site Arena Arena Date Comments length (m) width (m) Olympia 3 stades? Place for 680 BCE Exact shape of (hippodrome) 48 quadriga quadriga; arena unknown; 648 ridden horses; elaborate starting 408 biga gates in 5th c. Alexandria 450-560 51.6/65 Ptolemaic, Referred to as 2nd c. BCE Lageion, stadium, or hippodrome; barrier 7 m wide Cyrene 351 58.4 Hellenistic “Stadium” by scholars; barrier Gortyn ? 51

94097_ARAM_23_11_Patrich.indd 231 11/02/13 08:21 232 ROMAN HIPPO-STADIA

Table 5: U-shaped Hippo-Stadia in Syria-Palaestina and Arabia

Site Arena Arena Date Comments length (m) width (m) Caesarea 303 50.5 Herodian Three carceres phases; podium wall 1.45 m high; shortened to amphitheatre Neapolis 270 48.6 Flavian, or early Carceres, wooden or 2nd c. CE barrier 20 m long on the far end; shortened to amphitheatre Gerasa 244 51.28-49.5 2nd half of Carceres, 2nd c. CE athletics, podium wall 2.66 m high; shortened to amphitheatre Gadara >238 unknown End of 2nd c. CE Only N cavea built Scythopolis ? 48 2nd c. Podium wall (amphitheater) 1.45 m high; shortened to amphitheatre

Table 6: Roman Stadia with Sphendonai on both ends (amphi-stadia) (arena dimensions and dates)

Site Arena Arena Date Comments length (m) width (m) Nysa on 192 44 Augustan amphiteatron by the Mehander Strabo 14.1.43. Aphrodisias 238 31-40 1st c CE Starting mechanism (CIG 2758); podium 1.6 m high. Laudicea ad Lycum 270 79 CE amphitheatron Phrygia (Dedicated to Titus) stadion (IG Rom 4.845 and 861)

94097_ARAM_23_11_Patrich.indd 232 11/02/13 08:21 J. PATRICH 233

Site Arena Arena Date Comments length (m) width (m) Alexandria not extant amphitheatron Taurus (?) and stadion by Strabo 17.1.10 Nicopolis, 247* 59* (Augustan in Originally Epirus (phase II) phase I) U-Shaped, second sphendone of bricks

Table 7: Stadia with two open ends (no sphendone) (arena dimensions and dates)

Site Arena Arena Date Comments length (m) width (m) Miletus 192.27 29.56 Not after Two starting lines; 197-159 BCE two long straight Aizanoi 212.3 29-42 2nd half of Two long concave 2nd c. CE tribunes

Table 8: Dimension of truncated-stadia amphitheaters

Site Arena Arena length (m) width (m) Caesarea 136.5-91.5 Ca. 50 Gerasa 145* Ca. 50 Neapolis 57 39 Scythopolis 95 48

* Conjectural.

94097_ARAM_23_11_Patrich.indd 233 11/02/13 08:21 234 ROMAN HIPPO-STADIA of Caesarea: General plan and cross-section note 5 [1995], p. 18, Fig. 5, drawn by A. Iamim). hippo-stadium Porath, (after Fig. 1a. The

94097_ARAM_23_11_Patrich.indd 234 11/02/13 08:21 J. PATRICH 235 epicting various Reconstruction d of Caesarea: hippo-stadium arena phases and amphitheatrical (A. Iamim; courtesy Y. Porath). Fig. 1b. The

94097_ARAM_23_11_Patrich.indd 235 11/02/13 08:21 236 ROMAN HIPPO-STADIA p. 57, Fig. 3). note 12 [1989], of Gerasa: General plan (after Ostrasz, hippo-stadium The Fig. 2a.

94097_ARAM_23_11_Patrich.indd 236 11/02/13 08:21 J. PATRICH 237 Reconstruction with shortened arena and amphitheatrical tunnels (E. Cohen). of Gerasa: hippo-stadium The Fig. 2b.

94097_ARAM_23_11_Patrich.indd 237 11/02/13 08:21 238 ROMAN HIPPO-STADIA of Neapolis, general plan (after Magen, note 14 [2005], p. 146. Fig. D.5, hippo-stadium Fig. 3. The indicating only the actual remains found of a barrier wall and presumed locations turning points).

94097_ARAM_23_11_Patrich.indd 238 11/02/13 08:21 J. PATRICH 239 of Gerasa, schematic cross-section (Ostrasz, note 12 [1995], p. 189, Fig. 4). hippo-stadium Fig. 4. The

94097_ARAM_23_11_Patrich.indd 239 11/02/13 08:21 240 ROMAN HIPPO-STADIA Fig. 5. Remains of the northern cavea Gadara stadium (Kerner and Hoffmann, note 22).

94097_ARAM_23_11_Patrich.indd 240 11/02/13 08:21 J. PATRICH 241 amphitheater superimposed of the (Tsafrir and Foerster, note 24). of Scythopolis, general plan hippo-stadium The Fig. 6.

94097_ARAM_23_11_Patrich.indd 241 11/02/13 08:21 242 ROMAN HIPPO-STADIA

Fig. 7. The hippo-stadium of Caesarea, carceres phases (drawing E. Levit and A. Iamim).

94097_ARAM_23_11_Patrich.indd 242 11/02/13 08:21 J. PATRICH 243 section cross- Plan and of Gerasa, underground chambers. (after Ostrasz, note 12 [1991], p. 239, Fig. 2). hippo-stadium The Fig. 8.

94097_ARAM_23_11_Patrich.indd 243 11/02/13 08:21 244 ROMAN HIPPO-STADIA

Fig. 9. The hippo-stadium of Gerasa, general plan of the northern part (Ostrasz, note 12 [1991], p. 238, Fig. 1).

94097_ARAM_23_11_Patrich.indd 244 11/02/13 08:21 J. PATRICH 245

Fig. 10. The hippo-stadium of Caesarea, general plan of the northern part, depicting “sub-arena gallery C1900” and the four walls that truncated the arena, converting it into an amphitheater (after Porath, note 5 [2003], p. 38).

94097_ARAM_23_11_Patrich.indd 245 11/02/13 08:21 246 ROMAN HIPPO-STADIA (after Porath, note 8 [2003], p. 452, Fig. 2). of Caesarea, general layout “sub-arena gallery C1900” the amphitheater phase hippo-stadium The Fig. 11.

94097_ARAM_23_11_Patrich.indd 246 11/02/13 08:21 J. PATRICH 247 of Neapolis, general plan with the superimposed amphitheatre (Magen, note 14 [2005], p. 181. Fig. E.3). hippo-stadium Fig. 12. The

94097_ARAM_23_11_Patrich.indd 247 11/02/13 08:21 248 ROMAN HIPPO-STADIA Fig. 13. U-shaped stadium: Delphi IV (the stadium of Herodes Atticus), general plan. Note the triumphal arch on open end (Aupert and Callot, note 56, Pl. IV).

94097_ARAM_23_11_Patrich.indd 248 11/02/13 08:21 J. PATRICH 249 general plan : Aphrodisias, sphendone ) with two amphi-stadium (Welch, note 60, p. 549, Fig. 2). Amphitheatrical stadium ( Fig. 14.

94097_ARAM_23_11_Patrich.indd 249 11/02/13 08:21 250 ROMAN HIPPO-STADIA Note the gates on both ends (von Gerkan, note 63, Pl. III). Fig. 15a. Stadiums with two open ends, general plans: Miletus – parallel longitudinal sides.

94097_ARAM_23_11_Patrich.indd 250 11/02/13 08:21 J. PATRICH 251 ., note 63 [1993], p. 439, Fig. 3). al

et (Hoffman Fig. 15b. Stadiums with two open ends, general plans: Aizanoi – concave longitudinal sides

94097_ARAM_23_11_Patrich.indd 251 11/02/13 08:21