1It is remarkable that, despite the immense importance and popularity of equestrian events in the ancient Greek world, the current state of knowledge concerning their venues –the hippodromes– is still somewhat limited. This is even more surprising given that during the last few decades research has actually flourished in the field of ancient sports history.¹

2There are, in fact, a considerable number of archaeologically well-attested horse racecourses in the Greek eastern Mediterranean from Roman times.² But all of them seem to have been constructed in the specific Roman way³ and to have staged chariot races of the Roman type.⁴ For the sake of terminological clarity those structures from Roman times and of Roman style should always be called “circuses”.

3Whereas many circuses are known about or have been –at least partly– excavated throughout the Roman Empire and dealt with in numerous publications, the same cannot be said for the genuine Greek venues for equestrian competitions: the hippodromes. Many of those that became known from other sources have not even been located.⁵ Also, until this conference, scholarly work focusing on the hippodrome has been rare. A quick check of the annual bibliography on ancient sports in the journal Nikephoros will easily reveal the enormous imbalance between works dealing with the circus on the one hand and the hippodrome on the other.

4The question arises whether the lack of scholarly work on the hippodrome is due to the scarcity of remains, or whether the opposite is true: namely that not much archaeological evidence of the hippodrome has been discovered because it has been somewhat neglected by scholarly interest. I’d say the latter proposition is not as implausible as it might seem at first glance.

5Anyway, taking account of the current state of research, it is the prevailing opinion that the Greek hippodrome did not have any permanent features that were comparable to the monumental circuses of Roman times, nor in many cases any permanent features whatsoever.⁶ In this context the improvised track in the horse race of the Patroclus Games is quite often referred to,⁷ but this certainly does not give us any insight into the possible appearance of the horse racecourses of the Greek big Panhellenic sports festivals. The fact that the facilities of these sports festivals were not in use very often, e.g. only every four years in Olympia, has also been noted.⁸ It is, however, hard to see this constituting a reason for a lack of architectonical expression, since the same would also be true for ancient Greek stadia, and there is clearly an abundance of archaeological findings in relation to the latter.

6Even if it is beyond doubt that the hippodrome never had the impressive structures of the circus, it would still be legitimate to question whether excessive weight has been attached to the idea of the hippodrome being not much more than a vast field.

7This very volume contains quite a few articles presenting promising new archaeological initiatives and results that shed new light on our knowledge of the Greek hippodromes.⁹ To those I just want to add observations made at two other famous sites: the first is Olympia, where an initially promising geophysical investigation unfortunately turned out to be unsuccessful. Likewise, excavations carried out in the area assumed to be in the location of the start of the horse racecourse have so far remained disappointing.¹⁰ The second is Nemea, where interesting discoveries concerning the Nemean hippodrome were uncovered a few years ago.¹¹

8My article tries to approach that topic from another perspective. It will start by focusing on a text preserved in a codex kept in the Old Seraglio in Constantinople that provides information on the length of the Olympic hippodrome. It was J. Ebert¹² who brought this codex back into discussion providing a new and very sensible understanding of the text that has since been widely accepted. What I want to do here is to present a suggestion to modify J. Ebert’s understanding of the text and to come to a different conclusion about the total length of the hippodrome in Olympia and to add some thoughts on the dimensions of hippodromes in general.

The Codex Seragliensis and the hippodrome of Olympia

9The hippodrome of Olympia can probably be considered the most important facility of its kind in the ancient world. It is mentioned a few times in the works of ancient authors.¹³ The most informative one is, of course, the well-known lengthy passage in the 6^th book of Pausanias.¹⁴ Nevertheless, even there, many questions remain open as to the actual appearance of this hippodrome. The passage in Pausanias contains, for instance, no information on the length of the whole facility. This issue, however, is addressed by a text that is preserved in a 11^th or 12^th century codex from the Old Seraglio in Constantinople. Though the text was written in medieval times it most probably originated in ancient times or at least provides ancient information.

10Codex Seragliensis Graecus 1, Fol.28^r, Z.18-20¹⁵

ὁ ὀλυµπιακὸς ‹ἀγὼν› ἔχει ἱπποδρόµιον ἔχον / σταδίους η´·
καὶ τούτου ἡ µία πλευρὰ ἔχει σταδίους γ´ (καὶ) πλέθ(ρον) α´, /
τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν στάδιον α´ καὶ πλέθρα δ´, /
πό(δας), δω´.¹⁶

The Olympic (contest) has a horse-racecourse, having 8 stadia
and each¹⁷ long side is 3 stadia and 1 plethron
the space¹⁸ next to the start is 1 stadion and 4 plethra
(all in all) 4,800 feet.¹⁹

11Some attempts have already been undertaken to make sense of this text. In 1978 Fr. Brein²⁰ came up with the idea that the words “each long side is 3 stadia and 1 plethron” refer to the distance between the turning posts of the horse racecourse. And the phrase “τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν στάδιον α´ καὶ πλέθρα δ´” would indicate the dimensions of the shorter side of the hippodrome. He believes that the 8 stadia mentioned in the first line of our text as well as the 4,800 feet in the last line (which is actually the same measurement as 8 stadia) indicate the length of one lap which was calculated as an average distance by adding up the measurements of two long sides and half of two shorter sides.²¹ I am not convinced by this explanation.

12Another way of understanding the text of the codex was presented by J. Humphrey²² in his book on Roman circuses, which is still an –or even: the– authoritative work in this field. In his sketch (fig. 1) information from Pausanias’ description²³ is also included; i.e. especially the starting mechanism on the left with its “ship-prow” shape and the starting gates as well as the so-called Taraxippos. Leaving this additional information aside, I would rather draw attention to J. Humphrey’s understanding of the passage of the codex: “τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν στάδιον α´ καὶ πλέθρα δ´.” Again πλάτος is understood as “shorter side”. The sketch clearly shows how he makes sense of this information. In this respect J. Humphrey’s view is probably influenced by the assumption that the numbers of competing chariots were as high as 40 or more. This assumption, however, is only based on a passage in Pindar²⁴ mentioning a crash of 40 competitors. J. Ebert, however, was able to show that there is most probably a mistake in the Pindar text originally referring to a crash of just four chariots.²⁵

Fig. 1 — J. Humphrey’s suggestion.

13The distance between the two turning posts in J. Humphrey’s reconstruction is 2 stadia. This is based on the fact that a foot-race that was two diauloi (i.e. 4 stadia) long was called hippios dromos.²⁶ J. Humphrey believes, like some other scholars, that the name of the foot-race, hippios dromos, was derived from the length of a lap in the hippodrome.²⁷ Therefore the lap in the hippodrome must have had 4 stadia, i.e. the distance between the turning post (2 stadia) multiplied by 2.

14As mentioned before, J. Ebert²⁸ re-introduced the codex into discussion and provided a new and sensible understanding. For him the passage “each long side is 3 stadia and 1 plethron” refers to the distance between the two turning points plus the width of the track (fig. 2). He states that it can be discerned that the distance between the two turning points in the hippodrome was 3 stadia, which means that a circuit was 6 stadia long. This view is based on another part of our text to be found a bit further down in the codex, where the running distances of the various equestrian events are listed.

Fig. 2 — J. Ebert’s suggestion.

15Codex Seragliensis Graecus 1, Fol.28^r, Z.21-22²⁹

‹.... καὶ κέλητες› τρέχουσιν οἱ µὲν ἡλικιῶται πωλικοὶ³⁰ σταδίους ς′,
οἱ ‹δὲ ἡλικιῶται› τέλειοι ‹ιβ´›,
συνωρίδες αἱ µὲν {ἡλικιῶται} πωλικαὶ κύκλους γ´,
αἱ δὲ τέλειαι η´,
ἅρµατα ‹τὰ› µὲν πωλικὰ κύκλους η´,
τὰ δὲ τέλεια κύκλους ιβ.´

‹race horses› being colts run 6 stadia
full-grown horses run ‹12›
two horse chariots of colts run 3 laps
the full-grown 8
four horse chariots of colts run 8 laps
the full-grown 12.

16The text as provided here was amended by J. Ebert as indicated by the brackets.³¹ All of these corrections seem to have been carefully made and to be completely justified. The lower parts of the passage are not helpful for our question because here the distance is indicated in laps. The first line, however, indicates the distances in stadia. And that is informative. The text says: “race horses being colts run 6 stadia”, which is, indeed, a strong indication that the distance between the turning points was 3 stadia long.

17How else could the text make sense? A distance of 2 stadia between the turning posts would mean that this race was one and a half laps long. That is not very likely. 4 or 5 stadia would not work out at all. 6 stadia between the turning posts would probably be too long. A distance of 3 stadia between the turning posts, however, would mean that 6 stadia form one lap and this would indeed appear to make sense. The number of stadia for the adult horse race was completed by J. Ebert, but 12 stadia would be two laps and that seems perfectly reasonable.

18According to J. Ebert 1 plethron: therefore must indicate the width of the track (see fig. 2).

19But what does “τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν στάδιον α´ καὶ πλέθρα δ´” mean? For J. Ebert it describes the part left from the left (= western) turning point. He translated it into German as “die breite Fläche aber bis zur Ablaufstelle”; in English that would be: “the space next to the start is 1 stadion and 4 plethra”.

20The sketch (see fig. 2) makes it clear where J. Ebert wants to locate this very “space” (πλάτος). For him it is the part of the hippodrome between the aphesis described by Pausanias³² and the western turning point. J. Ebert’s understanding of the codex has since been generally accepted by recent scholarship.³³

21My problem with this understanding is that the codex twice refers to a total of 8 stadia (or the equivalent of 4,800 feet). It is beyond doubt that these 8 stadia (or 4,800 feet) are the total of the distances stated in the text. According to J. Ebert’s understanding we would get a total length of 8 stadia only as indicated in fig. 3.

Fig. 3 — Total of 8 stadia (4,800 feet) according to J. Ebert’s suggestion.

22Why should the codex sum up these particular distances: from the end of the aphesis to the western/left turning point, from there to the eastern/right turning point plus the width of the track and then back to the eastern/right turning point and from there to the western turning point?

23I would like to suggest a slight shift in our understanding of what the text actually means and that would allow us to get a more consistent picture.

24Let’s start again with the statement and each long side is 3 stadia and 1 plethron. This seems quite clear – there is not much scope to argue about that. I completely agree with J. Ebert’s understanding that the distance between the turning posts must be 3 stadia.

25Let’s take the next step. The distance “8 stadia” referred to in the first line of the passage (as well as “4,800 feet” in the last line) informs us about a total length of the hippodrome. But what does it mean? I believe it can only mean the perimeter of the hippodrome or the distance both ways –back and forth again. I suggest: back and forth again.

26But how does the statement “the space next to the start is 1 stadion and 4 plethra” fit in? This question seems to be at the very core of a correct understanding of the text. We can see J. Ebert’s answer in his sketch (see fig. 2). Now, I believe that also this statement has to be understood as the distance back and forth again. If we do that all the numbers fit together and a consistent picture of the text emerges (fig. 4).

27The question remains how the aphesis fits into the overall picture. It is crucial here how we understand τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν. J. Ebert translates πρὸς τὴν ἄφεσιν as “towards” or “next to the start”. Furthermore he adds the aphesis to this stretch. To me it seems preferable to translate the passage not as “the space next to the start” (like J. Ebert) but as: “the space for the start”. Then the aphesis would have been inside the space of “1 stadion and 4 plethra” –back and forth (fig. 5). Pausanias (6.20.11) referring to the aphesis says: “each side of the starting-place is more than 400 feet in length”. It would fit into the 500 feet long space that the codex indicates.

Fig. 5 — The location of the aphesis according to Petermandl’s suggestion.

28If this understanding of the text of the codex were correct then that would mean it is telling us that the hippodrome in Olympia had a total length of 4 stadia.³⁴ That would be about 900 feet shorter than the reconstruction of J. Ebert. Of course, we cannot be sure that what the codex describes actually corresponds to reality. In any case, it is an ancient statement on this topic.

29I am very pleased that the earlier publication of my ideas was helpful for C. Flämig and B. Dimde, who at this conference presented a different understanding of τὸ δὲ πλάτος πρὸς τὴν ἄφεσιν. Their article with this appealing suggestion is included in this very volume.³⁵ I will come back to it a bit further down.

Four Stadia – the length of the Greek Hippodrome?

30As just pointed out the information we can glean from the Codex Seragliensis appears to be that the hippodrome in Olympia was in total 4 stadia long. Whether we can believe what this text tells us or not, remains an open question. Unfortunately, there are no other ancient sources providing the measurements of the hippodrome in Olympia to prove that information. However, there is at least some evidence suggesting that 4 stadia were a common length for hippodromes generally.

311) Fr. Brein in his above mentioned article, refers to a passage in Plutarch, stating that the measurement of 1 hippikon is 4 stadia long.³⁶ Based on the well-known fact that the measurement of one stadion is as long as a stadium³⁷ this can be considered as a strong argument to understand the measurement hippikon as the length of a hippodrome.³⁸

322) It was also mentioned above that J. Humphrey and others believed that the name of the footrace ἵππιος δρόµος (being two diauloi –i.e. 4 stadia– long) was derived from the fact that it was as long as a lap in the hippodrome.³⁹ Fr. Brein, on the other hand, thinks that this footrace took its name from the total length of the hippodrome.⁴⁰ J. Ebert, however, surmises that the footrace was denominated ἵππιος δρόµος because it consisted of two laps, exactly like the race for mature race horses.⁴¹

333) Another hint can be found in the Etymologicum Magnum, where the following entry can be found s.v. ἐν Ἐχελιδῶν· τόπος Ἀθήνῃσι σταδίων ὀκτώ. ἐν ᾧ αἱ ἱπποδροµίαι. It attests a place of 8 stadia in which horse races were held. For D.G. Kyle this source makes it possible to locate the hippodrome of Athens and he states: “Apparently the Athenian hippodrome was eight stades long and located in Echelidai, near the city”.⁴² Bearing in mind our new understanding of the Codex Seragliensis, I would, however, suggest that we can interpret these 8 stadia not only as the distance one way but rather as the distance back and forth.

344) More support is provided by the observations of St.V. Tracy and Chr. Habicht. Discussing Panathenaic horse races that were not held in the hippodrome they state that these races probably took place between the Dipylon gate and the Eleusinion. The distance between those locations is about 700 m.⁴³ That would almost perfectly correspond to a length of 4 stadia: 4 stadia = 2,400 feet. With the attic foot being 296 mm long⁴⁴ 2,400 feet would be 710.4 m.

35All this evidence does at least correspond with my suggestion made above that the text of the Codex Seragliensis seems to inform us that the hippodrome of Olympia (i.e. the whole venue) was 4 stadia long.

36As far as the length of the Olympian hippodrome is concerned C. Flämig and B. Dimde come to a similar result. Although, their understanding of the text has the advantage that the 4 stadia would be the distance between the actual starting line and the opposite turning point; that would far better correspond to the comparison with the stadium and the measure of length stadion (being the distance between starting line and finish line/turning point).

37In any case, as there is no other clear evidence on hippodrome dimensions we might ask whether 4 stadia could have been a standard length of hippodromes. The question is, whether standardised horse racecourses can generally be expected or not. Actually there are a few pros and cons for the presumption of such a standardisation.

38J. Humphrey believes that Greek hippodromes varied greatly in length and width.⁴⁵ However, taking into account what was said at the beginning of this article, it should be recalled that we hardly possess the information to be able to state that. Based on current knowledge it cannot be categorically asserted that there was no standardisation of measurements whatsoever; especially for the big Panhellenic Games. Of course, there might actually have been a number of horse racecourses that had their own individual dimensions, like, for example, the hippodrome on the Mount Lykaion.⁴⁶

39An argument supporting the view that there was no standardisation can be based on the fact that Roman circuses do indeed vary significantly in length and width.⁴⁷ The question is just whether we can compare the Greek horse races and those of the Romans.⁴⁸ I think it is legitimate to doubt that the Roman horse races organised by the factiones under different conditions from say the equestrian events of the periodos-games would have required standardise racecourse lengths.

40There are, on the other hand, a few clues that standardisation of hippodromes might well have existed. First of all we should remember that the length of the stadion,⁴⁹ where footraces were held, was indeed standardised. It was 600 feet everywhere. Deviations in the metric system are only due to the different lengths of the foot in different places.

41Another argument could focus on Plutarch⁵⁰ mentioning a hippikon as a term for a measurement of length. This would only make sense if it referred to a certain degree of standardisation. If this term really were somehow linked to the hippodrome, this would also indicate a standardisation of length of hippodromes.

42The same is true if we are willing to believe that the hippios dromos acquired its name from the total length of the hippodrome.⁵¹ This would not have worked if different hippodromes had different lengths.

43Actually standardisation was also common in other aspects of horse races. The number of laps to run seems to some extent to have been standardised. At least 12 laps appear to be quite well attested for four-horse chariots in Olympia as well as in Delphi and Isthmia.⁵² Moreover, a defined number of laps would have made even more sense if the laps were of the same length.

44A further hint could also be seen in another kind of standardisation that appears to have existed. Certainly, at least at the Panhellenic Games a standardisation of the equestrian programme can be observed comprising races for four-horse chariots, two-horse chariots and horses; and each of them in two age classes for colts and grown up horses.⁵³

45Given that the very specially prepared race horses might have been trained to run specific distances,⁵⁴ all cases where there is evidence of horses having won at different places –like the horse Pherenikos winning in Delphi and Olympia⁵⁵– would support the assumption of the same dimensions of horse racecourses.

46As a final point, we shouldn’t necessarily assume that the distances changed over the centuries. The footraces demonstrate that distances actually remained the same for hundreds of years.

47Putting all that together, we reach the conclusion that there are a few hints suggesting that the dimensions of Greek hippodromes could have been standardised, at least at Panhellenic Games. Assuming that this was true, is seems sensible to consider 4 stadia as the standard length. In any case 4 stadia is certainly a measurement that should be given serious consideration when we reflect the length of the Greek hippodrome in general or the hippodrome of Olympia in particular.