Version classiqueVersion mobile

Les hippodromes et les concours hippiques dans la grèce antique

 | 
Jean-Charles Moretti
, 
Panos Valavanis

Les hippodromes d’Olympie et de Delphes

The aphesis of the Olympic hippodrome: Dimensions, design, technology

Barbara Dimde et Catharina Flämig

Résumé

The appearance and location of the ancient hippodrome of Olympia poses many unsolved riddles to this day. Even the most recent and extensive geomagnetic and archaeological investigations have not managed to locate the racecourse, once the largest monument of ancient Olympia. Theories about the starting-gate (aphesis) of the Olympic hippodrome have mainly been based on the description of Pausanias who visited the site around the middle of the 2nd century AD. A reconstruction of the mechanism at work has never, however, been undertaken, due to both the alleged lack of detail in Pausanias’ description and the lack of archaeological data with regard to hippodromes in general. This article proposes a reconstruction of the entire apparatus with components that may have degraded through exposure to the elements (wood, ropes/chains) and semi-precious parts (bronze dolphin/eagle).

Texte intégral

Dimensions and Design (C. Flämig)

  • 1 After an elaborate campaign in 2008, a geophysical survey of the area to the east of the sanctuary (...)
  • 2 Unfortunately, the anticipation caused by the geophysical survey findings in 2009 was short lived a (...)

1The appearance and location of the ancient hippodrome of Olympia poses many unsolved riddles to this day. Even the most recent and extensive geomagnetic and archaeological investigations have not managed to locate the racecourse, the once biggest monument of ancient Olympia.1 The assumption remains that the medieval floods of the River Alpheios were responsible for the destruction of this once so famous structure.2

  • 3 For summaries of earlier hippodrome reconstruction attempts see Ebert 1989, pp. 89-91 (republished (...)
  • 4 Pausanias, 5.15.5-7; 6.20.10-21.1.
  • 5 Codex Seragliensis GR1, foil 28r, ll. 16-23.

2In contrast to most other ancient competition sites in Greece, we are fortunate in Olympia to be able to rely on literary sources to give us an approximate insight into the former appearance of the ancient structure.3 We owe much credit to the well cited description of Pausanias from the 2nd century AD,4 supported by a metric text, written in Greek, dated somewhere between the 2nd and 3rd centuries AD, recorded in a manuscript of the 11th century AD, in the old Seraglio in Istanbul.5 Undoubtedly, both sources leave plenty of room for varied reading and interpretation. Needless to say, it is to our great advantage that both texts stem from a similar period in time, allowing us to draw interesting conceptions as to the appearance of the hippodrome of Olympia, if only for the time of the Roman Imperial Period.

  • 6 See above, n. 3.

3The famous starting gates (aphesis) of the Olympic hippodrome, a key feature in our research, are impossible to reconstruct without regarding them as an integrated part of the entire facility itself. A key source in regard to the appearance of the Olympic racecourse is the detailed documentation of Pausanias.6 It is documented that access to the starting point for the horses was gained by climbing the south wall of the stadium at the Hellanodikai terraces. This presents a very valuable clue as to the location of the former racecourse, namely the close proximity to the stadium, stretching in length to the south-east, with the Sanctuary of Zeus taking its position in the west. Furthermore, Pausanias’ description mentions one side of the hippodrome as longer than the other with the longer side bordering onto an artificial earth embankment. There were two turning poles and the starting field resembled the shape of a ship’s prow. The front of the prow, at which a dolphin atop a pole was planted, was positioned towards the race track, with the bow (the rear) bordering onto the Stoa of Agnaptos. The longitudinal sides of the starting gates each measured 400 ft. along which the individual starting boxes were situated. In accordance to Pausanias’ description the starting stalls were staggered in pairs parallel to each other and opened from back to front. The aim of such a sophisticated starting arrangement was to ensure an equally good starting position for all charioteers and horses across the starting field. The allocation of the starting positions was determined by a draw. In accordance with Pausanias’ documentation, the construction of the starting mechanism involved the collaboration of at least two engineers.

  • 7 See above, n. 3.
  • 8 See above, n. 4. In order to understand the measurements contained within the codex in Istanbul, th (...)

4Many reconstruction efforts, based on the detailed descriptions of Pausanias, have been attempted, with each individual attempt varying significantly from its predecessor.7 Noticeable in all reconstruction attempts however is the absence of scale, making it impossible to calculate either the length or the width of the complex in its entirety, nor the length of the approach, from the apex of the starting formation to the first turning pole, at which point the lap count started. The previously mentioned codex in Istanbul8 was the first document to shed light on the expanse of the ancient race site, providing the most detailed account to date of measurements for the hippodrome. Furthermore, the text within contains log entries of the distances covered during the horse and chariot races.

  • 9 Ebert 1989. Cf. Ebert 1991b.
  • 10 The shortest completed race distance was the foal race, covering 6 stadia, exactly 2 lengths of the (...)

5In accordance with J. Ebert’s reading, the codex in Istanbul suggests the hippodrome in Olympia covered 8 stadia, a total of 4,800 ft. each longitudinal side measuring 3 stadia and 1 plethron, and the space up to the starting point, 1 stade and 4 plethra.9 According to J. Ebert the 4,800 ft. draw reference to the total extent of the structure (i.e. the bidirectional distance) as the text explicitly refers to the longitudinal extension of the hippodrome. The remaining measurements mentioned in the text are distributed, by J. Ebert, across the interior of the racecourse –the approach (acceleration stretch), the distance between both turning poles and the width of the lane at the turning zones themselves. The approach is clearly suggested within the phrase: “the area to the starting point (has) 1 stade and 4 plethra”. A total of 1,000 ft. J. Ebert defines this run-up as the area between the prow (apex) of the starting gates to the nearest turning pole (fig. 1). This leaves the remaining text to cater for the distances between both turning poles and the length of the turning zones themselves: “thereof each longitudinal side has 3 stadia and 1 plethron”. J. Ebert defines the 3 stadia as the distance between the two turning poles, his placement being supported by the race distances recorded in the log entries contained within the same text.10 In accordance with J. Ebert’s interpretation, the remaining 1 plethron = 100 feet, make up the width of the track at the turning zones. By positioning both turning poles along the horizontal centre line of the hippodrome, the turning zones take on the shape of semicircle with a radius of 100 feet. This determines not only the width of each lane (100 ft.) but also the width of the entire hippodrome (200 ft.). The codex in Istanbul does not give reference to the starting gates. Based on Pausanias description of the starting gates, J. Ebert applies a triangle across the entire width of the hippodrome in his reconstruction: the side lengths of which correspond with the 400 ft. mentioned by Pausanias; and the base equally as wide as the hippodrome itself (200 ft.).

Fig. 1 — Reconstruction of the Olympic hippodrome (not to scale).

Fig. 1 — Reconstruction of the Olympic hippodrome (not to scale).

After Ebert 1989.

  • 11 As adopted by Decker 1995, p. 178. Sinn 2004, p. 62. Accepted largely also by Petermandl 2011, p. 4 (...)
  • 12 A third point was addressed only marginally: J. Ebertʼs reconstruction plans depict modern metres, (...)

6J. Ebert’s reconstruction was accepted almost unanimously in subsequent analysis.11 Nevertheless, two points in J. Ebert’s reconstruction attempt, which to us play a significant role in the form and function of the entire complex and how we understand it, should be reconsidered: namely the length of the approach as well as the position and design of the starting gates in relation to an equally good starting position for all horses and charioteers.12

The Length of the Approach

7As mentioned above, J. Ebert defines the total run up as being 1,000 ft. (1 stade and 4 plethra) long area between the apex of the starting gates up to the nearest turning pole. This measurement is placed by J. Ebert in his plans (fig. 1) however, not to scale. If drawn to scale (fig. 2) the extremely long approach, in comparison with the actual racecourse itself, immediately catches the eye. It is understood that all race participants were restricted to their position on the racecourse during the approach until they had reached the nearest turning pole. It was first here that the lap count started and that the individual participants were able to leave their starting lanes to compete for a more advantageous position on the course. The functionality of the elaborately calculated starting positions is questionable if the following acceleration stretch was so long that the wheat would have been separated from the chaff long before the first turning pole was reached –making for a seemingly obvious outcome and unimpressive race.

Fig. 2 — Reconstruction of the Olympic hippodrome (to scale).

Fig. 2 — Reconstruction of the Olympic hippodrome (to scale).

After Ebert 1989.

  • 13 Cf. W. Petermandl in this volume.

8The length of the run up can be calculated very differently in context with the phrase found in the codex in Istanbul, τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν as the “area up to the starting position” including the starting gates, if interpreting the 1 stade and 4 plethra as the total length of the run up, including the length of the starting arrangement itself (fig. 3). This seems a feasible calculation as the 4 plethra (400 ft.) mentioned in the codex in Istanbul correspond to the 400 ft. documented by Pausanias as being the measure of each longitudinal length of the starting gates. The length of the approach (distance between the staring position and the nearest turning pole) is therefore reduced to 1 stade. Together with the 3 stadia, between the two turning poles, 4 stadia are left to represent the length of the hippodrome race track itself.13 Interestingly enough this distance corresponds exactly to the ancient Greek measure, hippikon –a measure which would most likely not have existed without it fulfilling a defined practical purpose. A complete circuit of the racecourse would equate to an exact 4,800 ft./8 stadia, as mentioned in the codex.

Fig. 3 — Reconstruction of the Olympic hippodrome (to scale)

Fig. 3 — Reconstruction of the Olympic hippodrome (to scale)

C. Flämig.

  • 14 Kyle 1987b, pp. 96-98.
  • 15 Petermandl 2011, pp. 39-41. and W. Petermandl in this volume.

9The measure of 8 stadia is applied not only in Olympia but also in the hippodrome in Athens.14 A unification of racecourse distances amongst all Greek hippodromes was proposed most recently by W. Petermandl, who based his theory on the fact that several of the same contestants had participated in various Panhellenic contests. A number of literary sources confirm this and provide evidence of matching race distances in equestrian based competitions. Moreover, competing horses could only have been trained for a specific race distance if the various racecourses themselves were of comparable lengths.15

The Position and Design of the Starting Gates in Relation to an Equal Race Start

10All ancient depictions of equestrian competitions share one point in common: the races took place in a counter clockwise direction. This suggests that, in Olympia, the southern length of the course was covered first. J. Ebert suggests the starting gates in Olympia covered the entire width of the hippodrome (fig. 1). This would however mean a much longer, diagonal path towards the southern lane for horses starting in the north, with a difficult turning angle, leading to a heightened potential for collisions. In consideration to the above, an equally fair starting position for all contestants could only have been reached if all horses accelerated in a straight line towards the track side which was to be covered first (fig. 4). This hypothesis coincides with Pausanias’ description of the hippodrome having one side longer than the other, with the longer side bordering onto an earthen wall. If the starting gates and the approach were to have been situated solely along the southern length of the course, this length would indeed have covered a longer distance. Furthermore, the landscape towards the south is plain, offering a reasonable explanation for the artificial earth embankment. On one hand it would have given the spectators a better view of the starting procedure, and on the other hand, it would have protected the hippodrome from floods caused by the Alpheios River.

Fig. 4 — The starting mechanism of the Olympic hippodrome.

Fig. 4 — The starting mechanism of the Olympic hippodrome.

C. Flämig.

11The construction of the starting gates was decisive for the number of contestants taking part in each race. In the attempt to reconstruct the exact starting gates of the hippodrome in Olympia we encounter many variables, some of which remain unknown. Principally there are a number of factors which influenced the appearance of the starting formation. The illustration presented here (fig. 4) is by no means a statement of the definite appearance of the starting formation, but rather a demonstration of the difficulties we are faced with in the process of its reconstruction. The design is based on the following parameters:

  • 16 This assumption is not supported by any evidence, almost all Roman circus venues during the Roman I (...)

12– A) Our reconstruction plans are based on the hypothesis that both the northern and southern race tracks were of equal width, namely 100 ft. each. This defines the starting gates as having a width of 100 ft.16

13– B) One can assume that the design of the starting gates and the division of the starting boxes within followed precise mathematical calculations, based on ancient linear footage.

14– C) Pausanias mentions a staggered starting formation with boxes opening in horizontal pairs at the start of the race. As the last boxes opened, all participants would have found themselves level with each other at the apex of the starting gates, where a dolphin atop a pole was positioned. At this point, the width of the track would have had to accommodate all horses side by side along the approach.

  • 17 Humphrey 1986, pp. 21, 44 with fig. 28.

15– D) Moreover, it should be kept in mind that the starting gates were utilized for all equestrian competitions held in the hippodrome. In order to gain approximate accuracy in the reconstruction of the starting boxes, the maximum required width must be taken into consideration, for example that of a quadriga. Based on the circumstances found at the best preserved Roman circus in Lepcis Magna, J. Humphrey concluded a maximum radius of action of 3 m for a single quadriga17 which, in ancient measure, converts almost precisely to 10 ft.

  • 18 The field of participants could be expanded to 10 if we were to base our reconstruction on the info (...)

16Mathematically, based on these parameters, sufficient space would have been available for 10 starting boxes, 5 a side. A practical interpretation however, suggests that the width of 100 ft. provided sufficient space for only 8 boxes18 as, according to Pausanias, the starting formation was staggered from the rear to the front. This means that the first boxes to open simultaneously would have been 1A and 1B (fig. 4). It was only once these chariots reached the height of box 2A and 2B, that the riders within would be released from their starting position etc. In order to insure equal opportunity amongst all contestants, an adequate run up after leaving the starting box had to be granted for all competitors. Our diagram depicts the run up as an equal distance of 90 ft. from the apex of the starting gates to the front-most two starting boxes, followed by an equal difference between box to box from front to back. The measurement of 90 ft. is relative to approximately 25-30 m, which is indeed the average distance a horse must travel to reach a competitive speed on acceleration.

17A reconstruction based on these parameters is however not without fault. It is understood that an equal distance between each starting box would pose an unfair speed advantage to the primary starters. Although they would have to lay back a greater distance, the early start would have enabled the horses to reach speeds much greater to those who started later, when reaching the apex of the starting gates. This reconstruction indicates that the formation of horses, when meeting one another at the start of the approach, would more likely have resembled a concave rather than a straight line.

18It should not be forgotten that, whilst on his visit to the hippodrome, Pausanias did not actually witness the race grounds in action, leaving him to rely on the recollection of others to describe the function of the complex starting mechanism for his documentation. It is therefore possible that his interpretation of the starting field was not perfectly accurate, as equality amongst starting positions would have been impossible to achieve if the boxes were first opened when horses were adjacent to each other. An equally good start for all contestants could only have been reached if the boxes were opened when the previous horses were halfway between their own box and the next, and if the distances between the starting boxes gradually grew from the rear to the front.

19In conclusion, without the support of archaeological evidence of the former structure, even with the support of mathematical calculations, a true reconstruction of the complex Olympic starting gates is a feat which we will struggle to achieve or even comprehend. We are however convinced that an empirical attempt, including a replicated reconstruction and an experimental starting sequence, including horses and chariots, would help to bring us closer to reclaiming a structure unique to antiquity.

Technology (B. Dimde)

  • 19 Pausanias, 5.15.5-7; 6.20.10-21.1. Most recently Petermandl 2011, pp. 37-39. For a summary of older (...)
  • 20 Kyrieleis 2011, p. 123. Petermandl 2011, pp. 38-39, 41-46. Decker 1992a, pp. 134-135, pls. 64-65. E (...)
  • 21 Cf. Bell D. J. 1990, p. 315.

20Theories about the starting-place (aphesis) of the Olympic hippodrome have mainly been based on the description of Pausanias who visited the site around the middle of the 2nd century AD.19 An actual reconstruction of the mechanism at work, though, has never been undertaken, due to both the alleged lack of detail in Pausanias’ description and the lack of archaeological data with regard to hippodromes in general.20 With respect to Pausanias’ description, however, it is important to keep in mind that he refers to the aphesis of the Olympic hippodrome of his times, i.e. the apparatus of the 2nd century AD.21

  • 22 Pausanias, 6.20.14.
  • 23 In this article, all English translations of Pausanias’ Description of Greece are given according t (...)
  • 24 Pausanias, 5.10.8, where, for example, he mentions his exegetes at Olympia.

21According to Pausanias the Olympic aphesis was invented by Kleoitas, son of Aristokles, and was later enhanced by Aristeides.22 The ancient author adds that Kleoitas, taking much pride in his achievement, erected a statue of himself in Athens, the inscription of which reads: ὅς τὴν ἱππάφεσιν ἐν ὀλυµπίᾳ ἕυρατο πρῶτος / τεῦξέ µε Κλεοίτας υἱὸς ἀριστοκλέυς.23 Through this, several things become intelligible: firstly, an invention worth the erection of a statue must relate to something extraordinary, which means with respect to horse and chariot races that Kleoitas’ hippaphesis most probably was comprised of a mechanism to facilitate the starting procedure. It is easy to imagine that skittish horses at a starting line are hard to assemble –even if held back by assistants– so that a fair starting procedure was undoubtedly difficult yet critical to guarantee. In Greek stadia, too, it has always been the start of a race which was technically supported, never any other phase of a running competition. Secondly, it becomes clear that Kleoitas’ hippaphesis was the first and thus oldest apparatus for mechanically starting horse races at Olympia. And finally, it shows that the Olympic aphesis saw at least two different construction phases, the posterior of which Pausanias saw and described when he visited the site of Olympian Zeus, whereas he learned about the earlier one from hearsay.24

  • 25 Vollkommer 2001, p. 414.
  • 26 FD III 5, pp. 199-203, no. 50 II, l. 63; no. 74 I, l. 52.

22In order to date this earlier construction phase of the Olympic hippaphesis invented by Kleoitas, scholars came up with a homonymous sculptor who is attested in the first half of the 5th century BC and worked at Athens and Olympia.25 Although very tantalizing by name and filiation, two problems arise from the assumption that this same sculptor Kleoitas was responsible for the construction of the earlier version of the Olympic aphesis: first, one has to explain the fact that an aphesis for horses was invented by a sculptor and not an engineer or architect who by profession would seem to be more apt. More so, since at Delphi, epigraphical evidence of an architect named Kallinos has been interpreted as the person responsible for the construction and maintenance of the hippaphesis of the Delphian hippodrome.26 And secondly, the hippaphesis designed by Kleoitas would have been invented between 500 and 450 BC which is fairly hard to correlate with general developments in ancient technology. Even though the first problem might be negated, since in antiquity professions covered a wider scope than they do today, the chronological problem remains.

  • 27 On the etymology of the word see Valavanis 1999, p. 3 with n. 12, 5.
  • 28 Valavanis 1999, pp. 51, 65-67. The earliest known stadium-hysplex, however, dates to ca. 430 BC and (...)
  • 29 Valavanis 1999, p. 3: “The word hysplex makes its earliest appearance in the Lysistrata (line 1000) (...)
  • 30 Petermandl 2013, pp. 137-138. Bell D. J. 1990, pp. 313-314. Ebert 1972, pp. 176-181, no. 59. IPerga (...)
  • 31 IPergamon, no. 10, l. 3.
  • 32 Petermandl 2013, pp. 137-138, with a German translation: “Die Startschranke spannte sich und hielt (...)
  • 33 IPergamon, no. 10, l. 5. On the sound produced by hyspleges see Valavanis 1999, p. 5, and Bell D. J (...)
  • 34 SEG XXIX 951. Petermandl 2013, p. 149.
  • 35 SEG XXIX 951, l. 4.

23Yet, with respect to the date of the earliest starting-mechanism of the Olympic hippodrome some observations can be offered, although absolute proof as to its exact design cannot be provided on the basis of current source material. In terms of a mechanism having been invented to facilitate the starting procedure of a horse race, a barrier comes to mind which was placed in front of all participants and which was removed when the starting signal was given. In ancient stadia this method is well-known from archaeological traces and was called hysplex by the ancients.27 Although no archaeological remains of a hippodrome-hysplex are preserved, it is telling that in stadia this technology was in general use not before the first third of the 3rd century BC.28 The very first appearance of the hysplex-technology for hippodromes in literary sources is made during the last 3rd of the 5th century BC.29 In contrast, the earliest evidence for hysplex-technology in the Olympic hippodrome stems from an inscription found at Pergamon (Asia Minor) and dates to between 280 and 272 BC.30 The inscription’s text celebrates the victory of Attalos, adopted son of Philetairos, in the four-horse chariot race of the fouls at Olympia. Here, a hysplex-mechanism to start the races in the Olympic hippodrome is explicitly mentioned.31 As to the mechanism proper, the Pergamenian inscription refers to a barrier behind which all horses were assembled and which was tautened by using a twisted rope: ἁθτρόα δ΄ ὕσπληξ πάντα διὰ στρεπτοῦ τείνατ΄ ἔχουσα κάλω.32 Also, the loud crack which was generated by the hysplex is touched upon: [ἣ] µέγ΄ ὑπαχήσασα.33 Furthermore, an Olympic hippodrome-hysplex is also attested by an honorary epigram dated to the 4th or 3rd century BC devoted to Dexandros from Lokroi Epizephyrioi who won a hippic contest at Olympia.34 In regard to the starting procedure an almost identical expression to the aforementioned Pergamenian inscription is used: [ἁθτρ]όα τῆς ὕσπλ[ηγος...].35

  • 36 Valavanis 1999, p. 51 with n. 157. Hall 1957, pp. 698-699; cf. 707-715. Schürmann 1991, pp. 62-72. (...)
  • 37 For the start of running events in the stadium this technology was used at several sites and might (...)
  • 38 Cf. Bell D. J. 1990, p. 135, who on the basis of a philological examination comes to similar conclu (...)

24The diction in both texts, however, is evocative of a hysplex-technology which used kinetic power stored in twisted ropes or sinews and which originated in the invention of military machinery during the 4th century BC.36 Adapted to the hippodrome and by analogy to ancient stadium-technology,37 this (reconstructed) mechanism consisted of two upright wooden posts (ankones) between which a rope was stretched as a barrier and the lower part of which was inserted into a loop of pre-twisted ropes or sinews.38 These ropes were held under tension by being fastened to an anchor and thus stored the necessary power for the sudden removal of the barrier.

  • 39 Livy, 8.20.2. Humphrey 1986, pp. 132-133. Harris 1968, pp. 116, 121.
  • 40 FD III 5, pp. 199-203, no. 50 II, l. 63; no. 74 I, l. 52.
  • 41 Also, advanced “sportsˮ-technology in the stadium of Olympia did not emerge before the 1st century (...)
  • 42 Bell D. J. 1990, p. 319. Wiegartz 1984, p. 45. Harris 1968, p. 113. Petermandl 2013, pp. 137-138 (= (...)

25Summing up the above, we see that technology in hippodromes facilitating the starting procedure occurs in literary sources in the last 3rd of the 5th century BC, whereas epigraphical sources attesting to the existence of such technology at Olympia date to the late 4th/early 3rd centuries BC. In this connection it is interesting to note that Livy fixes the date for the introduction of the carceres in the Circus Maximus, which also worked on the basis of twisted ropes or sinews and therefore might be considered a Roman reflex on Greek hippodrome-technology to 329 BC.39 Also, the aforementioned inscription mentioning the architect Kallinos and his works on the hippaphesis of the Delphian hippodrome is dated to the 4th century BC –although it is by no means certain that an hippaphesis automatically implied the use of a hysplex-mechanism.40 Yet, it seems more plausible for the invention of the Olympic hippodrome-hysplex to favour a date later in the 5th or early 4th century BC rather than as early as the first half of the 5th century BC.41 Thus, with respect to the earlier hippodrome-hysplex at Olympia we might infer from the above that it was constructed by Kleoitas, son of Aristokles, who is not compellingly to be equated with the sculptor of the same name, and was invented during the end of the 5th or the beginning of the 4th century BC as a mechanism to facilitate the starting procedure of the hippic contests at Olympia. This mechanism most likely made use of a torsion-technology based on twisted ropes to remove a barrier in front of the horses. Since literary sources on the starting procedure of hippic contests in general, as well as on the one at Olympia in particular, elucidate that by the use of a hysplex-system a simultaneous start was usually effectuated, it seems highly probable that Kleoitas’ mechanism, too, provided a simultaneous starting procedure for all participants.42

  • 43 Vollkommer 2001, p. 83, s.v. “Aristeides (IV)ˮ.
  • 44 Pausanias, 6.20.14.
  • 45 Pausanias, 6.20.10-13.

26Nonetheless, Pausanias attributes the enhancement of the hippodrome-hysplex to Aristeides, who is commonly identified with a Greek architect of the same name and who acted in the time of Trajan at the beginning of the 2nd century AD.43 For chronological and professional reasons it is tempting to equate the known architect of Trajan’s time with the one Pausanias talks about. However, if the Olympian aphesis of the hippodrome was indeed first constructed by Kleoitas it is most probable that –due to both the huge lapse of time and technical progress– Aristeides’ hysplex was a completely new construction instead of a simple enhancement of the pre-existing mechanism. Yet, Pausanias relates to an addition Aristeides made to the hysplex of Kleoitas: Κλεοίτα δέ φασιν ὕστερον ἀριστείδεν σοφίαν τινὰ καὶ αὐτὸν ἐς τὸ µηχάνηµα ἐσενέγκασθαι.44 Hence, it seems plausible that this addition corresponds to the mechanism that allowed the staggered opening of the starting-gates as opposed to the simultaneous one invented by Kleoitas, because this is what Pausanias accentuates and describes in detail.45 Thus, Aristeides probably changed the arrangement of the boxes from a straight line to an inverted V-shape design and added a mechanism which allowed a subsequent opening of the gates. The gates themselves consisted of individual barriers and were operated by the “old” hysplex-system with twisted ropes or sinews as a power source. So in fact, the aphesis of the Olympic hippodrome actually consisted of two hyspleges, one of which was supposedly constructed in classical times and offered a simultaneous starting procedure, whereas the other one was probably built in imperial times, not only incorporating elements of the older mechanism but also featuring new astonishing effects by providing a staggered starting procedure.

  • 46 Pausanias, 6.20.11; 20.13. Cf. Bell D. J. 1990, pp. 313-314, 315.
  • 47 Fundamental for hysplex mechanisms is Valavanis 1999. For the Roman circus see Humphrey 1986, pp. 1 (...)
  • 48 Pausanias, 6.20.12.
  • 49 On the chain reaction in a Roman circus see Humphrey 1986, pp. 161-169, and in a stadium Miller 199 (...)
  • 50 Pausanias, 6.20.12.
  • 51 Bell D. J. 1990, pp. 316-317. Cf. Humphrey 1986, p. 158: “The best way to reduce the role of chance (...)
  • 52 In this respect it is interesting to note that the word τεταγµένος, which Pausanias uses for the st (...)

27However, with respect to the Olympic hysplex-mechanism described by Pausanias we can infer at least some general aspects of the mechanism in operation. In his account Pausanias explicitly mentions the word hysplex46 which implies the universal idea of starting the races with a maximum of equality of opportunity by reducing human agency to a minimum. Also, this hysplex-mechanism probably made use of a torsion-technology on the one hand, and a release-system to set the torsion free on the other.47 In addition, Pausanias mentions a mechanical device which signaled the beginning of the race to the spectators.48 So, when attempting to reconstruct the Olympian aphesis we have to expect at least three different mechanisms: one giving the starting signal of the race, a second one triggering the release of the barriers in front of the horses, and a third one forcing the barriers in front of the horses to be withdrawn.49 Furthermore, by Pausanias’ phrasing ἀνακινεῖ µὲν δὴ τὸ ἐν τῶ̣ βωµῶ̣ µηχάνηµα ὁ τεταγµένος ἐπὶ τῶ̣ δρόµω̣ («The man appointed to start the racing sets in motion the mechanism in the altar.»)50 it is evident that all aforementioned mechanisms were set in motion by one single movement of the hand executed by the tetagmenos, which in turn leads to the conclusion that the entire mechanism was constructed as a chain reaction.51 Once set in motion, this chain reaction could not be stopped thus reducing the possibilities of human manipulation to a minimum.52

  • 53 Bell D. J. 1990, pp. 316-317.
  • 54 Galen, 18(1).438.
  • 55 Bell D. J. 1990, p. 317, argues convincingly that trigger mechanisms were also in use in hippodrome (...)

28In order to gain more information on the release system used at Olympia it is worth looking at ancient mechanical treatises dating to the 4th to the 2nd centuries BC which offer a technical term for the trigger-mechanism of certain machines and engines (σχαστηρία).53 These were obviously still in use in later times when Galen –though referring to a footrace– claims that a σχαστηρία in his time (2nd century AD) was a device to be stepped on by the starter.54 How exactly this mechanism operated is unknown, but it is conceivable that at Olympia the τεταγµένος mentioned by Pausanias stepped on a lever protruding from the altar at the centre of the aphesis thereby setting in motion the whole starting procedure.55

  • 56 Valavanis 1999, p. 5. See also Bell D. J. 1990, p. 313. Humphrey 1986, p. 8. Harris 1968, p. 116.
  • 57 On the date of the construction phases of Stadium III (A-F) see Schilbach 1992, p. 37. Rieger 2004, (...)
  • 58 AA 2012/1, p. 99, figs. 18-19. Two coins were found lying in the shafts: in the southern one a coin (...)
  • 59 AA 2012/1, p. 99: “Wahrscheinlich gehört sie zu einem Mechanismus, der durch das herabfallende Gewi (...)

29P. Valavanis, renowned scholar on ancient starting mechanisms, already pointed out that “it is reasonable to hypothesize that the original concept of a kind of barrier at the start began with horses and was later adopted for runners.”56 Hence, it is legitimate to look at the stadium-hysplex of Olympia, archaeological evidence of which comes from the so-called Stadium III E, being in use from around 25 BC until around the middle of the 3rd century AD (ca. 225-250 AD).57 Here, in 2012, German archaeologists unearthed a lead-weight of 235 kg lying in a shaft which together with its northern counterpart was constructed at each end of the western starting line.58 By the lead-weight’s find-spot, layout and dimensions it became clear that this pear-shaped weight was intended to slide along a pole piercing its entire length, and to pull with it –when being released from an elevated position– the barrier in front of the runners down to the ground (fig. 5).59

Fig. 5 — Olympia: schematic reconstruction of unilateral hysplex at western end of Stadium III.

Fig. 5 — Olympia: schematic reconstruction of unilateral hysplex at western end of Stadium III.

B. Dimde.

  • 60 Pausanias, 6.20.10.
  • 61 I owe this information to J. Schilbach, excavator of the stadium-hysplex at Olympia, who thankfully (...)
  • 62 Pausanias, 6.20.12. Humphrey 1986, pp. 8-9, equates the signal system of the Olympian hippodrome wi (...)
  • 63 One of the two lap-counting devices in the Circus Maximus at Rome consisted of seven plunging dolph (...)
  • 64 Pausanias, 6.20.11.
  • 65 Pausanias, 6.20.10-11; 20.15. Kyrieleis 2011, p. 123. Ebert 1991b, pp. 99-100. Ebert 1989, p. 95.

30If we adopt this technique to Pausanias’ description who says “… and a bronze- dolphin on a rod has been made at the very point of the ram”,60 it seems compelling to assume that this bronze dolphin was a weight, too, which was –like the one in the stadium– pierced at its entire length to render possible a downward and upward movement along a rod (fig. 6).61 Thus, we can argue that the fall of the dolphin marked the beginning of the chain-reaction, which in Pausanias’ words continues as follows: “The man appointed to start the racing sets in motion the mechanism in the altar, and then the eagle has been made to jump upwards, so as to become visible to the spectators, while the dolphin falls to the ground  ”.62 From this passage it becomes obvious that the falling of the dolphin is connected to the sudden rise of the eagle which apparently forms the next step in the chain-reaction.63 The eagle, however, is according to Pausanias positioned on an altar close to the centre of the aphesis.64 Corresponding to the measurements Pausanias gives with respect to the overall dimensions of the aphesis,65 this means that the eagle and the dolphin are some 200 feet apart (fig. 7). However, regarding the connection of eagle and dolphin, several constraints have to be made:

  1. a constructional connection between eagle and dolphin can only be effectuated by ropes or chains, not by wooden or metal elements because of their huge distance;
  2. the eagle needs to be at least slightly lighter in weight than the dolphin;
  3. the eagle can only rise as high as the dolphin falls down; and
  4. the positioning of the eagle on the altar causes –in comparison to the dolphin– an increase in height when being pulled up.

Fig. 6 — Olympia: schematic reconstruction of “dolphin planted on a rod” as trigger of the hippodrome hysplex.

Fig. 6 — Olympia: schematic reconstruction of “dolphin planted on a rod” as trigger of the hippodrome hysplex.

B. Dimde.

Fig. 7 — Olympia: schematic reconstruction of connection between eagle and dolphin during the starting procedure.

Fig. 7 — Olympia: schematic reconstruction of connection between eagle and dolphin during the starting procedure.

B. Dimde.

  • 66 By analogy it seems plausible that the eagle, too, was a weight in the form of an eagle being pierc (...)

31In practice, this means that there had to be a mechanism which stopped the eagle from rising because of the aforementioned difference in weight. When the starter (tetagmenos) released this mechanism, kinetic energy was set free and the heavier dolphin pulled the lighter eagle up to a height where it could be seen by the spectators, maybe some 5-6 (?) m.66

  • 67 Pausanias, 6.20.13.
  • 68 On the number of competing chariots at Olympia see above. It is interesting to note that the width (...)
  • 69 Pausanias, 6.20.12.
  • 70 If we assume that the aphesis was built in one half of the racetrack thus spanning a width of 100 a (...)
  • 71 Otherwise the mechanism would have to pierce the altar which, given the vertical movement of the ea (...)
  • 72 Harris 1968, p. 123. Cf. Humphrey 1986, p. 8.

32According to Pausanias67 the soaring of the eagle is followed by the graded opening of the chariots’ starting-boxes. These boxes open in pairs, beginning with the most rear ones at each side of the aphesis, and subsequently opening the following ones.68 Thus, the rising of the eagle is to be considered as the trigger mechanism for the staggered opening of the starting boxes. In this respect it is remarkable that Pausanias explicitly mentions the eagle’s widely stretched wings, although the bird is clearly not in flying mode: “… and a bronze eagle stands on the altar with his wings stretched to the fullest extent.”69 On these grounds it seems conceivable that the tips of the eagle’s wings initiated the next step in the chain-reaction, i.e. the graded opening of the starting boxes.70 For the mechanism proper, it has to be stated as essential that the layout of the altar in the centre of the aphesis had to be constructed two-fold or mirror-inverted as to allow the same action on each side of the aphesis.71 Furthermore, this two-fold mechanism needed to be installed at the inside of the aphesis in order not to form any obstacles to the travelling chariots (figs. 4, 8). Also, the movement which was triggered by the wings’ tips must have proceeded fairly slowly thereby allowing each chariot to gain some speed and reach the next starting box (flying start).72

Fig. 8 — Olympia: schematic reconstruction of the hippodrome aphesis and course of the cords of the different release-systems: green indicated chords are “cutˮ by the eagle’s wings and trigger the rolling of the metal balls; the blue indicated cords are triggered by the metal ball rolling down the channeled planks (planks indicated in red). The yellow line indicates the rope-/chain-connection between eagle and dolphin.

Fig. 8 — Olympia: schematic reconstruction of the hippodrome aphesis and course of the cords of the different release-systems: green indicated chords are “cutˮ by the eagle’s wings and trigger the rolling of the metal balls; the blue indicated cords are triggered by the metal ball rolling down the channeled planks (planks indicated in red). The yellow line indicates the rope-/chain-connection between eagle and dolphin.

B. Dimde.

  • 73 Cf. the mechanism described by Heron, Περί αὐτοµατοποιτηκῆς 2.1-9, where the object to be moved is (...)
  • 74 Pausanias, 6.20.11.
  • 75 Miller 1999, p. 159. It is equally possible that the same mechanism which was supposedly installed (...)
  • 76 Lycophron, Alexandra 13. Bell D. J. 1990, p. 316.

33One possible solution might be offered by the system of converting potential energy into kinetic, a principle we have seen in operation already in the case of the falling dolphin: a wooden plank equipped with a shallow channel, running inside of each side of the aphesis and slightly declining towards the aphesis’ tip could have supported a heavy bronze- or lead-ball, rolling down the channel and passing the triggers of the hyspleges of the starting boxes (fig. 9).73 This metal ball, however, could have been set loose by means of ropes stretching from the altar to the rear of the channelled planks which were cut by the eagle’s wings when he was made to rise (fig. 8). This, in any case would explain why the altar –as Pausanias tells us74– had to be erected close to the centre of the aphesis because then the distance of the cords connecting eagle and dolphin as well as those running from the altar to the rear of the channelled planks were cut in half and thus covered an almost equal distance –maybe the maximum distance over which cords could be operated smoothly. Since no physical remains of this mechanism are left we can only conjecture on how exactly this trigger-system operated. Considering release-systems from ancient stadia, it is conceivable that the eagle’s wings lifted a small bronze ring off a hook to which a rope –being slightly under tension– was attached and which at its far end held a small movable lever in an upright position (fig. 10). By lifting the ring off the hook this rope would suddenly slacken, thereby allowing the small lever to move and, in consequence, to be rolled over by the metal ball. A similar system –i.e. the pulling of a bronze ring off a hook– was, for example, reconstructed for triggering the hysplex of the Early-Hellenistic Stadium at Nemea.75 Also, literary evidence –though from the 2nd century BC– suggests that in a stadium a rope was “cut” to trigger the hysplex.76

Fig. 9 — Olympia: schematic reconstruction of the channeled plank with metal ball and hyspleges trigger (left), and metal ball in channeled plank in section (right).

Fig. 9 — Olympia: schematic reconstruction of the channeled plank with metal ball and hyspleges trigger (left), and metal ball in channeled plank in section (right).

B. Dimde.

Fig. 10 — Olympia: schematic reconstruction of the eagle triggering the staggered opening of the starting boxes (left) and movable lever held in upright position before being triggered (right).

Fig. 10 — Olympia: schematic reconstruction of the eagle triggering the staggered opening of the starting boxes (left) and movable lever held in upright position before being triggered (right).

B. Dimde.

  • 77 I am much obliged to the physicist Dr. T. Wider who thankfully supported me with his maths skills. (...)

34The necessary speed –or better slowness– of the metal ball could have been adjusted by the inclination of the wooden plank.77 Thus, once being set in motion the metal balls were able to set gradually loose the triggers of the boxes’ barriers. These triggers might have followed a similar mechanics like the release-system triggered by the eagle’s wings so that a cord stretching from the movable trigger in the channeled plank was connected at its far end to a lever or small bronze ring (?) which when being pulled released the hysplex of the starting gate (fig. 11). However, it is most likely that the “heart” of the entire aphesis- mechanism was not visible to the spectators as to increase the spectacular impression the mechanism must have had on them and was therefore set up as low as possible above ground level.

Fig. 11 — Schematic reconstruction of individual hysplex seen from above operated by torsion-based mechanism from one side only (left) and possible release-system with trigger-cord seen in section (right).

Fig. 11 — Schematic reconstruction of individual hysplex seen from above operated by torsion-based mechanism from one side only (left) and possible release-system with trigger-cord seen in section (right).

B. Dimde.

  • 78 Pausanias, 6.20.13.

35When it comes to the staggered opening of the starting boxes Pausanias describes the procedure as follows:78

Πρῶται µὲν δὴ ἑκατέρωθεν αἱ πρὸς τῆ̣     στοᾶ̣ γνάπτου χαλουσιν ὕσπληγες καὶ οἱ κατὰ ταύτας ἑστηκοτες ἐκθέουσιν ἵπποι πρῶτοι. θέοντές τε δὴ γίνονται κατὰ τοὺς εἱληχότας ἑστάναι τὴν δευτέραν τάξιν καὶ τηνικαῦτα χαλῶσιν αἱ ὕσπληγες αἱ ἐν τῆ̣ δευτέρα̣ τάξει. διὰ πάντων τε κατὰ τὸν αὐτὸν λόγον συµβαίνει τῶν ἵππων

First on either side the barriers are withdrawn by the porch of Agnaptus, and the horses standing thereby run off first. As they run they reach those to whom the second station is allotted, and then are withdrawn the barriers at the second station. The same thing happens to all the horses in turn

  • 79 Cf. Bell D. J. 1990, pp. 314-315. On stadium-hyspleges using this technology from Hellenistic throu (...)
  • 80 Humphrey 1986, pp. 21, 49-51.
  • 81 Pausanias, 6.20.11; cf. 6.20.13.
  • 82 Bell D. J. 1990, p. 313 with no. 1.
  • 83 Humphrey 1986, pp. 161-169. Junkelmann 1998, p. 111, figs. 106-109. Cf. Petermandl 2013, p. 58.
  • 84 Pausanias, 6.20.13.

36It is obvious that the boxes were barred by individual ὕσπληγες which due to the staggered starting procedure must have been operated not simultaneously but individually. If the element of using hysplex-technology was actually transferred from the former aphesis-apparatus invented by Kleoitas (see above) this mechanism most probably was powered by torsion stored in twisted sinew loops or metal springs held under tension.79 Since a starting box for a chariot drawn by four horses needed to have an approximate width of 3 m,80 it must have been sufficient to have the torsion mechanism of the boxes’ hyspleges work from one side only, presumably the inside of the aphesis in order not to hinder the travelling chariots coming from behind (cf. figs. 4, 8, 11). This hypothesis is strengthened by the fact that Pausanias mentions a cord stretched in front of the starting gates which reduces the weight of the barrier considerably, and with it the power needed to drive it down.81 It is noteworthy, though, that a cord stretched in front of four horses is not a very sturdy barrier given the power skittish horses are able to produce.82 In the Circus Maximus, in contrast, two-winged wooden doors (repagulae) barred the twelve boxes of the carceres.83 However, by the metal balls rolling down the channelled planks and pushing the trigger-levers for the release of the individual hyspleges, the chariots’ boxes were unblocked in a step-like arrangement, yet giving the impression that the arrival of the rear chariot effectuated the opening of its front opponent’s starting box.84 As C. Flämig already pointed out above it is most likely that the time between the staggered openings of the boxes was gradually reduced towards the front of the apex so that the barriers were in fact removed before the chariots from behind reached their front box (cf. fig. 8).

  • 85 Schürmann 1991, pp. 61-92. Hall 1957, pp. 695-715.
  • 86 One might think, for example, of an adopted version of the crane which was invented by Heron, the A (...)

37This relatively simple system was also relatively easy to handle when it came to getting it ready for the following race: the two metal balls having rolled down the wooden planks could have been brought back to their initial point of departure by assistants transporting them by hand –if one agrees on a relatively low construction above ground level. Since dolphin and eagle must have been connected by ropes or chains, it was enough to pull up the dolphin thereby bringing the eagle down to the top of the altar which could have been effectuated by means of cranes or winches (or a combination of both), commonly used with catapult machinery.85 If the supposition is correct that most of the mechanics of the aphesis-mechanism was invisible to the audience it is more likely that winches were used because those could have been placed close to the ground and maybe –to hide them completely from sight– even inside the altar.86

  • 87 As was the case of the stadium-hysplex at Delos. Valavanis 1999, p. 31 with n. 90.
  • 88 Decker 1992, p. 135. Wiegartz 1984, p. 44.

38Since the whole apparatus, the reconstruction of which has been proposed here, would have been comprised of components both delicate to exposure to the elements (wood, ropes/chains) and fairly precious (bronze dolphin/eagle), it is not unlikely that these elements were stored when not in use.87 Therefore, when Pausanias visited the site of Olympian Zeus, he most probably did not observe the mechanism in operation but saw only parts of the system, the working of which was explained to him by his exegetes.88 And yet, the aphesis of the Olympic hippodrome was obviously thrilling enough to leave him and posterity alike stupefied.

Notes

1 After an elaborate campaign in 2008, a geophysical survey of the area to the east of the sanctuary in Olympia depicted two lines running parallel to each other for 200 m. This initially sparked great anticipation in the attempt to locate the ancient hippodrome. See Wacker 2008, pp. 14-16.

2 Unfortunately, the anticipation caused by the geophysical survey findings in 2009 was short lived as further investigations yielded no results, and the conclusion of a complete destruction of the race course was reinstated. Kyrieleis 2011, p. 123. Senff 2012, p. 14. Similarly stated by Kunze 1964 several decades earlier. Wiegartz 1984, p. 41. Ebert 1989, p. 89.

3 For summaries of earlier hippodrome reconstruction attempts see Ebert 1989, pp. 89-91 (republished in J. Ebert, Agonismata. Kleine philologische Schriften zur Literatur, Geschichte und Kultur der Antike [1997], pp. 336-338). Sinn 2004, pp. 134-136. Petermandl 2011, pp. 37-39.

4 Pausanias, 5.15.5-7; 6.20.10-21.1.

5 Codex Seragliensis GR1, foil 28r, ll. 16-23.

6 See above, n. 3.

7 See above, n. 3.

8 See above, n. 4. In order to understand the measurements contained within the codex in Istanbul, the following measure description was included: the smallest unit of length was 1 foot. 1 plethron = 100 ft. 1 stade = 6 plethra or 600 ft. 1 hippikon = 6 stadia or 2,400 ft.

9 Ebert 1989. Cf. Ebert 1991b.

10 The shortest completed race distance was the foal race, covering 6 stadia, exactly 2 lengths of the racecourse (once around the hippodrome) provided that the finish line was equal to the nearest turning pole at which the lap count started. Summarized by Petermandl 2011, pp. 45-47.

11 As adopted by Decker 1995, p. 178. Sinn 2004, p. 62. Accepted largely also by Petermandl 2011, p. 47, with the crucial difference being that W. Petermandl calculates only half of the approach distance, as he wishes this to be interpreted as part of the calculated race distance. For a revised version of his earlier views see W. Petermandl in this volume. We are much obliged to W. Petermandl whose earlier publication on the topic was of great help in developing our own thoughts and ideas, and who generously shared his comprehensive knowledge with us in person.

12 A third point was addressed only marginally: J. Ebertʼs reconstruction plans depict modern metres, converted from the ancient lengths. This calls for caution in the interpretation, as foot measurements varied greatly throughout antiquity and there is no knowing on which foot measure Pasusanias and the author of the codex in Istanbul based their documentations. Without further reasoning, J. Ebert declares 1 ft. equal to 0.32 m. and therefore calculates the distance between the two turning poles as 576 m. as well as the total width of the hippodrome as 64 m. The measure of 0.32 m. per foot is most likely based on the distance measured between the two recovered starting block lines found in the stadium, which according to Mallwitz 1967, p. 37, measured 192.24 m. Dividing this number by the measure of a stade/600 ft. you are left with an exact 0.32 m. A. Mallwitz admits in his annotation 35, that the exact position of the western starting line is unknown. Furthermore, more recent investigations have suggested that the eastern and western starting lines were never used in combination with each other, making any calculation based on their corresponding positions irrelevant in determining the length of the Olympic stadium. Rieger 2004, pp. 143-145.

13 Cf. W. Petermandl in this volume.

14 Kyle 1987b, pp. 96-98.

15 Petermandl 2011, pp. 39-41. and W. Petermandl in this volume.

16 This assumption is not supported by any evidence, almost all Roman circus venues during the Roman Imperial Reign had a starting lane which was wider than the returning race lane. Humphrey 1986, p. 26, fig. 7 (Lepcis Magna); p. 120, fig. 54 (Rome, Circus Maximus); p. 302, fig. 139 (Karthago); p. 346, fig. 154 (Sagunto).

17 Humphrey 1986, pp. 21, 44 with fig. 28.

18 The field of participants could be expanded to 10 if we were to base our reconstruction on the information provided, which is that most Roman circus venues during the Roman Imperial Period had a starting lane which was wider than the returning race lane. More than 10 participants could not have been possible, as the northern lane would become too narrow, provided that all riders started in head-on position to the primary lane. Unfortunately there are no reliable records relating to the size of the participant field during the Roman Imperial Period. If you take into account the cost and efforts invested into getting horses, chariots and riders to the venue in Olympia, then the number of 8-10 participants per race is one of realistic proportions.

19 Pausanias, 5.15.5-7; 6.20.10-21.1. Most recently Petermandl 2011, pp. 37-39. For a summary of older research results see Decker 1992a, pp. 134-135 with “Korrekturzusatzˮ on page 138. Cf. Humphrey 1986, pp. 7-8, figs. 2a-b. Habicht 1985, p. 182.

20 Kyrieleis 2011, p. 123. Petermandl 2011, pp. 38-39, 41-46. Decker 1992a, pp. 134-135, pls. 64-65. Ebert 1991b, p. 99. Wiegartz 1984, pp. 41-42. Unfortunately, the article by Ν. Pierros on the mechanism of the Olympic aphesis, in Πρακτικά του Ζ΄ Διεθνούς Συνεδρίου Πελοποννησιακών Σπουδών, Πύργος, Γαστούνη Αµαλιάδα, 11-17 Σεπτ. 2005 (2006), pp. 353-362, was not available in any library in Germany and could therefore not be included in this article.

21 Cf. Bell D. J. 1990, p. 315.

22 Pausanias, 6.20.14.

23 In this article, all English translations of Pausanias’ Description of Greece are given according to the Loeb Classical Library edition, vol. II and III, 1960 and 1961, respectively. Pausanias, 6.20.14: “It was Cleoetas who originally devised the method of starting, and he appears to have been proud of the discovery, as on the statue at Athens he wrote the inscription: ‘Who first invented the method of starting the horses at Olympia, He made me, Cleoetas the son of Aristocles.’ It is said that after Cleoetas some further device was added to the mechanism by Aristeides.” See also Harris 1968, p. 121.

24 Pausanias, 5.10.8, where, for example, he mentions his exegetes at Olympia.

25 Vollkommer 2001, p. 414.

26 FD III 5, pp. 199-203, no. 50 II, l. 63; no. 74 I, l. 52.

27 On the etymology of the word see Valavanis 1999, p. 3 with n. 12, 5.

28 Valavanis 1999, pp. 51, 65-67. The earliest known stadium-hysplex, however, dates to ca. 430 BC and was unearthed in the so-called Earlier Stadium at Isthmia, but its mechanism differed considerably from hysplex-mechanisms based on a twisted rope torsion-systems: the Isthmian hysplex was most probably operated by a heavy weight which fell down a circular shaft and pulled strings by which the individual barriers in front of the runners were removed. Rieger 2004, pp. 130-158. Cf. Humphrey 1986, p. 133.

29 Valavanis 1999, p. 3: “The word hysplex makes its earliest appearance in the Lysistrata (line 1000), which was produced in 411 BC, as a metaphor for the characteristic of unanimity, or accord, which was needed in a specific project of the women. In other words, we see in the 5th century BC use of the two terms [i.e. hysplex and balbis] without any definition. This suggests that both terms were already well established in popular parlance.” On the depiction of a stadium-hysplex on a Panathenaic vase dating to 344/343 BC see Valavanis 1999, pp. 20-31. The earliest stadium-hysplex was found at the so-called Earlier Stadium of Isthmia which was in use from ca. 430-390 BC. Rieger 2004, pp. 262-264, 282-296. Harris 1968, p. 113.

30 Petermandl 2013, pp. 137-138. Bell D. J. 1990, pp. 313-314. Ebert 1972, pp. 176-181, no. 59. IPergamon, no. 10.

31 IPergamon, no. 10, l. 3.

32 Petermandl 2013, pp. 137-138, with a German translation: “Die Startschranke spannte sich und hielt sie alle zusammen mit ihrem gedrehten Seil. Mit lautem Klatschen trieb sie (dann) heraus die schnellen Fohlen.ˮ Bell D. J. 1990, p. 314, with an English translation: “The hysplex was stretched and kept them all in check with its twisted rope.ˮ IPergamon, no. 10, l. 3-4.

33 IPergamon, no. 10, l. 5. On the sound produced by hyspleges see Valavanis 1999, p. 5, and Bell D. J. 1990, pp. 317-319.

34 SEG XXIX 951. Petermandl 2013, p. 149.

35 SEG XXIX 951, l. 4.

36 Valavanis 1999, p. 51 with n. 157. Hall 1957, pp. 698-699; cf. 707-715. Schürmann 1991, pp. 62-72. Puppet-like dolls which functioned with a similar technology and are described by Aristotle see D. J. Bell 1990, p. 314.

37 For the start of running events in the stadium this technology was used at several sites and might suggest an analogue mechanism for hippodromes. Cf. Valavanis 1999, pp. 31-49, figs. 26-35.

38 Cf. Bell D. J. 1990, p. 135, who on the basis of a philological examination comes to similar conclusions.

39 Livy, 8.20.2. Humphrey 1986, pp. 132-133. Harris 1968, pp. 116, 121.

40 FD III 5, pp. 199-203, no. 50 II, l. 63; no. 74 I, l. 52.

41 Also, advanced “sportsˮ-technology in the stadium of Olympia did not emerge before the 1st century BC/AD. Cf. Rieger 2004, pp. 130-158.

42 Bell D. J. 1990, p. 319. Wiegartz 1984, p. 45. Harris 1968, p. 113. Petermandl 2013, pp. 137-138 (= Ebert 1972, no. 59), 149 (= SEG XXIX 951). Cf. Petermandl 2013, pp. 124-125, on a papyrus dating to the late 3rd/early 2nd century BC where for Olympia the starting place for the horses is called grammè (line) indicating a simultaneous start. Cf. Sophocles, Electra 709-710. On the etymology of the word hysplex see Valavanis 1999, p. 3 with n. 12, 5. Cf. Humphrey 1986, p. 132, who presumes that in the Circus Maximus prior to the introduction of the carceres in 329 BC all races were started from a straight line.

43 Vollkommer 2001, p. 83, s.v. “Aristeides (IV)ˮ.

44 Pausanias, 6.20.14.

45 Pausanias, 6.20.10-13.

46 Pausanias, 6.20.11; 20.13. Cf. Bell D. J. 1990, pp. 313-314, 315.

47 Fundamental for hysplex mechanisms is Valavanis 1999. For the Roman circus see Humphrey 1986, pp. 157-170. Cf. Junkelmann 1998, p. 111, figs. 106-109.

48 Pausanias, 6.20.12.

49 On the chain reaction in a Roman circus see Humphrey 1986, pp. 161-169, and in a stadium Miller 1999, p. 159.

50 Pausanias, 6.20.12.

51 Bell D. J. 1990, pp. 316-317. Cf. Humphrey 1986, p. 158: “The best way to reduce the role of chance, then as now, was to eliminate human agency as far as possible. […] The best way to accomplish simultaneous opening must be through some kind of mechanism, operated by just one person, a mechanism which, once activated, caused all the gates to open simultaneously.”

52 In this respect it is interesting to note that the word τεταγµένος, which Pausanias uses for the starter of the aphesis-mechanism at the end of the 5th century BC, denoted an umpire or rather an assistant to an umpire (cf. Sophocles, Electra 709-719: στάντες. δ` ἵν αὐτοῦς. οἱ τεταγµένοι βραβῆς κλήροις ἔπηλαν καὶ κατέστησαν δίφρους.). In case the word tetagmenos did not undergo a shift in meaning from the 5th century BC to the 2nd century AD it would show that during the starting procedure impartial objectiveness was aimed at by endowing an official with the release of the entire mechanism.

53 Bell D. J. 1990, pp. 316-317.

54 Galen, 18(1).438.

55 Bell D. J. 1990, p. 317, argues convincingly that trigger mechanisms were also in use in hippodromes as early as the 3rd century BC.

56 Valavanis 1999, p. 5. See also Bell D. J. 1990, p. 313. Humphrey 1986, p. 8. Harris 1968, p. 116.

57 On the date of the construction phases of Stadium III (A-F) see Schilbach 1992, p. 37. Rieger 2004, pp. 130-158, with a detailed description of the constructional alterations of the western starting line in connection with the installation of a unilateral hysplex and its relative chronological sequence within the Olympian starting devices. See also Valavanis 1999, pp. 57-59, partly refuted by Rieger 2004. On the hysplex-parts recently discovered by German archaeologists see AA 2012/1, pp. 99-100. On a relative chronology of starting devices for runners in general see Rieger 2004, pp. 375-410.

58 AA 2012/1, p. 99, figs. 18-19. Two coins were found lying in the shafts: in the southern one a coin fresh from the mint showing the Roman emperor Domitian, in the northern one a coin dating to the reign of Caracalla as Caesar, maybe a hint to repair works in that period of time. AA 2012/1, pp. 99-100.

59 AA 2012/1, p. 99: “Wahrscheinlich gehört sie zu einem Mechanismus, der durch das herabfallende Gewicht eine Startvorrichtung in Gang setzte.ˮ Cf. Hero, an Alexandrinian engineer working in the 1st century AD, who in his Περί αὐτοµατοποιτηκῆς 2.6-8, talks about automatically moving machines which were powered either by a hysplex or a lead weight. A translation of the relevant passage gives Bell D. J. 1990, p. 315.

60 Pausanias, 6.20.10.

61 I owe this information to J. Schilbach, excavator of the stadium-hysplex at Olympia, who thankfully shared results from his current research with me.

62 Pausanias, 6.20.12. Humphrey 1986, pp. 8-9, equates the signal system of the Olympian hippodrome with the throwing of the mappa executed by a Roman magistrate to start the races in a Roman circus. Harris 1968, p. 123, states that the necessity for a starting signal means that the spectators must have sat “more than a quarter of a mile awayˮ from the aphesis of the Olympian hippodrome.

63 One of the two lap-counting devices in the Circus Maximus at Rome consisted of seven plunging dolphins, although the nature of a possible connection between the Olympian dolphin and the Roman dolphins is uncertain. Harris 1968, p. 122. Cf. Humphrey 1986, pp. 262-265.

64 Pausanias, 6.20.11.

65 Pausanias, 6.20.10-11; 20.15. Kyrieleis 2011, p. 123. Ebert 1991b, pp. 99-100. Ebert 1989, p. 95.

66 By analogy it seems plausible that the eagle, too, was a weight in the form of an eagle being pierced at its entire height and sliding up and down a rod, as shown in figs. 6, 7.

67 Pausanias, 6.20.13.

68 On the number of competing chariots at Olympia see above. It is interesting to note that the width of the Olympian racetrack of 100 ancient feet per track corresponds approximately to the one of the Roman circus at Lepcis Magna. Humphrey 1986, p. 38, fig. 17. Junkelmann 1998, pp. 108-109, fig. 105. Cf. Ebert 1991b, p. 101. Harris 1968, p. 125, observed correctly: “It must have been an essential feature of the start at Olympia that until all chariots were out of the traps every charioteer must drive straight forward and keep to his lane.”

69 Pausanias, 6.20.12.

70 If we assume that the aphesis was built in one half of the racetrack thus spanning a width of 100 ancient feet, a maximum number of 8 chariots per race can be reconstructed given the fact that a four-horse chariot needed ca. 3.00 m in width (and was about 2.30 m in height). Humphrey 1986, pp. 49-51, fig. 28. Cf. Romano 2014, pp. 186-187, Map 11.5, with the dimensions of the hippodrome-stadium on mount Lykaion: 260 × 102 m. This gives some free space probably at the centre “laneˮ of the aphesis so that the two front chariots were not positioned in direct vicinity to each other as to prevent their colliding already at the start. This space could have been easily bridged by the eagle’s wings. It is interesting to note that in a Roman circus in the middle of the carceres an archway was located (Humphrey 1986, pp. 97-100) near to which the trigger mechanism of the starting gates was operated, probably by means of a catapult. Humphrey 1986, pp. 164-166, figs. 75-76. Harris 1968, p. 122.

71 Otherwise the mechanism would have to pierce the altar which, given the vertical movement of the eagle’s strings and their connection to the dolphin, seems less convincing.

72 Harris 1968, p. 123. Cf. Humphrey 1986, p. 8.

73 Cf. the mechanism described by Heron, Περί αὐτοµατοποιτηκῆς 2.1-9, where the object to be moved is running on wheels. An object on wheels triggering the hyspleges of the Olympian starting gates, however, would probably gain too much speed thus allowing the chariots not enough time to reach the next starting gate.

74 Pausanias, 6.20.11.

75 Miller 1999, p. 159. It is equally possible that the same mechanism which was supposedly installed in the aphesis’ altar and set loose the falling of the dolphin was used for releasing the rolling of the metal ball: in both cases quite a huge distance had to be bridged by cords, chains or ropes, and only one single action was sufficient to activate it.

76 Lycophron, Alexandra 13. Bell D. J. 1990, p. 316.

77 I am much obliged to the physicist Dr. T. Wider who thankfully supported me with his maths skills. On the water-pipeline constructed in the 6th century BC by the engineer Eupalinus on the island of Samos see Forbes 1957, pp. 667-668, and on Roman aqueducts Forbes 1957, pp. 670-674. It is also possible that the mechanism was operated by means of water being poured down the channelled plank, the power of which would be able to trigger the levers for the individual hyspleges of the boxes, too. It is interesting to note that the nymphaeum of Herodes Atticus and his wife Regilla brought plenty of fresh water to Olympia during the Games of 153 AD. Sinn 2004, pp. 202-206. Bol 1984, pp. 98-100.

78 Pausanias, 6.20.13.

79 Cf. Bell D. J. 1990, pp. 314-315. On stadium-hyspleges using this technology from Hellenistic through Roman imperial times see Valavanis 1999.

80 Humphrey 1986, pp. 21, 49-51.

81 Pausanias, 6.20.11; cf. 6.20.13.

82 Bell D. J. 1990, p. 313 with no. 1.

83 Humphrey 1986, pp. 161-169. Junkelmann 1998, p. 111, figs. 106-109. Cf. Petermandl 2013, p. 58.

84 Pausanias, 6.20.13.

85 Schürmann 1991, pp. 61-92. Hall 1957, pp. 695-715.

86 One might think, for example, of an adopted version of the crane which was invented by Heron, the Alexandrinian engineer, during the 1st century AD for the raising of collapsed walls: a hoist of two bigger and two smaller pulleys could have been attached at one side to the dolphin by using an additional role for the directional change of the rope, and at the other side attached to a horizontal winch with which the dolphin could be pulled up in his initial position on top of the rod. Schürmann 1991, pp. 149-150, fig. 29.

87 As was the case of the stadium-hysplex at Delos. Valavanis 1999, p. 31 with n. 90.

88 Decker 1992, p. 135. Wiegartz 1984, p. 44.

Table des illustrations

Titre Fig. 1 — Reconstruction of the Olympic hippodrome (not to scale).
Crédits After Ebert 1989.
URL http://books.openedition.org/efa/docannexe/image/6497/img-1.png
Fichier image/png, 61k
Titre Fig. 2 — Reconstruction of the Olympic hippodrome (to scale).
Crédits After Ebert 1989.
URL http://books.openedition.org/efa/docannexe/image/6497/img-2.png
Fichier image/png, 26k
Titre Fig. 3 — Reconstruction of the Olympic hippodrome (to scale)
Crédits C. Flämig.
URL http://books.openedition.org/efa/docannexe/image/6497/img-3.png
Fichier image/png, 44k
Titre Fig. 4 — The starting mechanism of the Olympic hippodrome.
Crédits C. Flämig.
URL http://books.openedition.org/efa/docannexe/image/6497/img-4.png
Fichier image/png, 58k
Titre Fig. 5 — Olympia: schematic reconstruction of unilateral hysplex at western end of Stadium III.
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-5.png
Fichier image/png, 126k
Titre Fig. 6 — Olympia: schematic reconstruction of “dolphin planted on a rod” as trigger of the hippodrome hysplex.
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-6.png
Fichier image/png, 66k
Titre Fig. 7 — Olympia: schematic reconstruction of connection between eagle and dolphin during the starting procedure.
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-7.png
Fichier image/png, 48k
Titre Fig. 8 — Olympia: schematic reconstruction of the hippodrome aphesis and course of the cords of the different release-systems: green indicated chords are “cutˮ by the eagle’s wings and trigger the rolling of the metal balls; the blue indicated cords are triggered by the metal ball rolling down the channeled planks (planks indicated in red). The yellow line indicates the rope-/chain-connection between eagle and dolphin.
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-8.png
Fichier image/png, 23k
Titre Fig. 9 — Olympia: schematic reconstruction of the channeled plank with metal ball and hyspleges trigger (left), and metal ball in channeled plank in section (right).
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-9.png
Fichier image/png, 84k
Titre Fig. 10 — Olympia: schematic reconstruction of the eagle triggering the staggered opening of the starting boxes (left) and movable lever held in upright position before being triggered (right).
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-10.png
Fichier image/png, 84k
Titre Fig. 11 — Schematic reconstruction of individual hysplex seen from above operated by torsion-based mechanism from one side only (left) and possible release-system with trigger-cord seen in section (right).
Crédits B. Dimde.
URL http://books.openedition.org/efa/docannexe/image/6497/img-11.png
Fichier image/png, 99k

Auteurs

Chercheuse indépendante

Chercheuse indépendante

© École française d’Athènes, 2019

Conditions d’utilisation : http://www.openedition.org/6540

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search