The pentathlon was a combined competition in five events, running, jumping, throwing the diskos, throwing the javelin, and wrestling. This is one of the few facts regarding the pentathlon which may be regarded as absolutely certain. These five events are vouched for by three epigrams, one of them assigned to Simonides, and by the repeated testimony of Philostratos in his Gymnastike.^[615] Nothing proves more conclusively the utter unreliability of the statements on athletics made by late scholiasts and lexicographers, than the mistakes which they contrive to make on a matter so clearly established. The lexicon of Phavorinus, following certain late scholia, substitutes boxing for throwing the javelin; and Photius quotes certain writers as substituting the pankration for the jump. Stranger still, such mistakes survive in the present day; and our own standard Greek Lexicon by Liddell and Scott contains, in the latest edition, the appalling statement that the five exercises were the jump, the diskos, running, wrestling, boxing, the last being afterwards exchanged for javelin throwing. After this we are not surprised to find quoted the antiquated theory of BÖckh, that “no one received a prize unless he was winner in all five events,” a theory that was disproved by Philip, years before the first edition of Liddell and Scott was published. The introduction of boxing into the pentathlon is due to the mischievous habit of using such inaccurate expressions as “the Homeric pentathlon.”^[616] In heroic days, as Pindar tells us, there was no pentathlon, “but for each several feat there was a prize.”^[617]

Of these five events, three—the jump, the diskos, and the javelin—were peculiar to the pentathlon, and formed its characteristic feature. These three events were regarded as typical of the whole competition; on the Panathenaic vases given as prizes for the competition one or more of these three events, on two vases all three of them, are represented^[618] (Figs. 107, 108). The same events are among the commonest on other vases, especially red-figured vases; but we are not justified in connecting these with the pentathlon, or using them as evidence in discussing the pentathlon. These scenes for the most part represent the daily life of the gymnasium, and all that they prove is the important part which these sports played in that life. They were the only three events which required any form of apparatus; the exercises seem to have been taught in classes, and were performed both in practice and in competition to the accompaniment of the flute. If any of the three was regarded as more representative than another, it was the jump, which perhaps owed its importance partly to the extensive use of halteres in the gymnasium. The halteres were the special symbol of the pentathlon, and were frequently represented on statues of victorious pentathletes.^[619]

These three events, together with running and wrestling, were representative of the whole physical training of the Greeks, and the pentathlete was the typical product of that training. Inferior to the specialised athletes in his special events he was superior to him in general development, in that harmonious union of strength and activity which produces perfect physical beauty; and this beauty of the pentathlete won him the special commendation of thinkers such as Aristotle, who condemned all exaggerated or one-sided development.^[620]

A combined competition like the pentathlon is obviously later than any of the individual events of which it is composed, and implies a considerable development in athletics and physical education. Not that we are to regard it with certain German writers as an elaborate scheme based on abstract physiological principles evolved with much expenditure of midnight oil out of the brain of some athletic student. The pentathlon was the natural product of a number of exercises which had been familiar for centuries. But before the idea could originate of combining these exercises into a single competition to find the best all-round athlete, these exercises must have become part of the national education. The combination implies a certain amount of thought and conscious reflexion. There is in it an artificiality of which we find no trace in the Homeric sports. In view of this it is remarkable that, according to Greek tradition, the pentathlon was introduced at Olympia as early as the 18th Olympiad.

No importance need be attached to the statement of Philostratus that the pentathlon was invented by Jason. The Greeks always loved to trace their institutions back to heroic times. As, however, the passage which contains the statement is of considerable importance in discussing the method of deciding the pentathlon, it will be useful to quote it in full:—

“Before the time of Jason there were separate crowns for the jump, the diskos, and the spear. At the time of the Argo’s voyage Telamon was the best at throwing the diskos, Lynceus with the javelin, the sons of Boreas were best at running and jumping, and Peleus was second in these events but was superior to all in wrestling. Accordingly, when they were holding sports in Lemnos, Jason, they say, wishing to please Peleus combined the five events, and thus Peleus secured the victory on the whole.”^[621]

The order of the events and the method of deciding the pentathlon have given rise to a literature equally extensive and inconclusive.^[622] Almost every combination of events has been tried, and every conceivable method has been devised. Many of the systems proposed are so utterly unpractical that they have only to be stated to be rejected by any one with a rudimentary knowledge of practical athletics. None can be regarded as established. The evidence is too scanty and too contradictory. It consists largely in extracts from scholiasts and lexicographers, and we have seen in considering the constitution of the pentathlon the untrustworthiness of this class of evidence. It is well, therefore, to recognise from the outset that whatever solutions we may accept are only provisional, and that it is therefore in the highest degree unsafe to use such theories as evidence in the interpretation of Pindar or other poets.

First, as to the order of events, it must be premised that we are not certain that the order was fixed, and did not vary at different times and places. Still, the conservatism of the Greeks in such matters certainly makes it probable that there was a fixed order at Olympia, and that this order was generally adopted elsewhere. At all events we shall assume that this was so. The one fact which we know for certain about the order is that wrestling came last. Bacchylides definitely describes it as last, and the evidence of Bacchylides is confirmed by Herodotus and Xenophon.^[623] Describing the attack on Olympia by the Eleans in Ol. 104, when the Arcadians had usurped the presidency of the games, Xenophon says: “They had already finished the horse-race and the events of the pentathlon held in the dromos (t? d????? t?? pe?t?????) and those who had reached the wrestling were no longer in the dromos but were wrestling between the dromos and the altar.” It is generally agreed that t? d????? are the first four events, which were held in the stadium, whereas according to the view set forth in a previous chapter wrestling took place in the open space in front of the treasury steps.^[624] At all events, it is clear from Xenophon’s words that wrestling came last, and common sense tells us that this was the only possible position for it consistent with fairness. After several hard bouts of wrestling no competitor could do himself justice in the other events.

For the order of the first four events we have to fall back on the uncertain and contradictory evidence of various passages in which the events of the pentathlon are enumerated. Now in none of these passages is the order of events of any importance to the writer; in the case of an epigram it is obvious that the order is likely to be modified by metrical considerations. Still, the probability remains that such passages will in spite of metre and carelessness reflect more or less the actual order.^[625] Thus we find that in five passages wrestling comes last, in two passages it comes first, and in both of these the order of events is merely reversed, in one passage it comes second. The epigram of Simonides gives the following order: Jump, foot-race, diskos, javelin, wrestling. The epigram quoted by Eustathius gives the same order except that the foot-race comes fourth instead of second. Now, except in the epigram of Simonides, the three events peculiar to the pentathlon are always grouped together. It is probable, therefore, that they were grouped together in practice, and that the foot-race cannot have occupied the second place. Why Simonides put it after the jump is obvious, neither d???? nor p?d??e?? could possibly begin a hexameter. The foot-race, therefore, came either first or fourth. Once more, if we examine the lists we find the foot-race first in two lists, last in the two reversed lists, while two scholia follow the epigram and place it fourth. As the order in these scholia is identical with that of the epigram, it is doubtful whether they have any independent authority. The evidence, therefore, is slightly in favour of first place for the foot-race, and this order receives some slight support from the passage in Philostratus already quoted concerning the pentathlon of Peleus, and the passage of Herodotus discussed below about Tisamenus and Hieronymus.

For the remaining events the lists appear to support the order of the two epigrams—jump, diskos, javelin, though there is not much to show whether the diskos or the javelin came first. Certain passages in Bacchylides and Pindar have been quoted to prove that the diskos preceded the javelin.^[626] On the two Panathenaic vases reproduced above, the javelin comes between the jump and the diskos. This is the position assigned to it by Philostratus when he enumerates the events of the pentathlon. Unfortunately the value of this passage is lessened by the distinction which he introduces between light events and heavy events. The heavy events, he says, are wrestling and throwing the diskos; the light events, the javelin, the jump, and the foot-race. The order is obviously reversed, but whether all three light events preceded both heavy events or not cannot be decided from this passage. Such distinctions give us no clue to the actual order, and all attempts to discover the system on which the order of events depended are absolutely futile. It is easy enough to argue that all the exercises were arranged in an ascending scale, or that easy exercises alternated with difficult, that similar exercises were grouped together, or that leg exercises alternated with arm exercises, and if we were constructing an ideal pentathlon such arguments might be of some use. As it is, we are not concerned with an ideal pentathlon but with that of the Greeks, and there is not a particle of evidence to prove that the Greeks arranged their pentathlon on any abstract principle however plausible. All we can do is to confine ourselves to the actual evidence, and the order which this evidence renders probable is foot-race, jump, diskos, javelin, wrestling.

It is unnecessary to discuss in full the various systems that have been suggested for deciding the pentathlon. These systems for the most part fall into certain well-defined groups based on certain hypotheses, and it will be sufficient briefly to examine these hypotheses.

The old hypothesis perpetuated by Liddell and Scott, that victory in all five events^[627] was necessary, may be briefly dismissed as not only unpractical but contrary to the little evidence which we possess. On such a system a victory in the pentathlon must have been an extremely rare event; for it can seldom have happened that one competitor won all five events. The idea seems to have arisen from the epigram of Simonides, and from a misunderstanding of an important passage in Herodotus (ix. 33), which is in reality a conclusive proof against it.

“Tisamenus,” says Herodotus, “came within a single contest or fall (p??a?sa) of victory, being matched against Hieronymus of Andros.” Pausanias confirms the victory of Hieronymus (vi. 14), and says of Tisamenus (iii. 11, 6), “In two events he was first, for he was superior to Hieronymus in running and jumping, but he was defeated by him in wrestling and so failed to win the victory.” The true interpretation of the passage is obvious. “Tisamenus came within a single contest of victory,” i.e. he won two events but lost the odd; or perhaps we may go farther still and give to p??a?sa its literal meaning, “a fall in wrestling.” He came within a “single fall” of winning.

Each had won two events, each had scored two falls in wrestling, and the whole contest depended on the last fall!^[628] just as we talk of winning a golf match by a single putt, or winning a rubber by the odd trick.

Yet obvious as this interpretation is, Hermann and other more recent German writers have asserted that, according to Herodotus, Tisamenus won the first four events, and only missed the victory because he was defeated in wrestling. It is more than doubtful whether the words of Herodotus can bear the meaning “he missed victory by wrestling only”; but apart from this, Hermann’s theory is absolutely contradicted by the very circumstantial statements of Pausanias. If Tisamenus won all four events, why should Pausanias expressly state that he won two? If victory in all five events was necessary, how can Hieronymus have won the pentathlon, seeing that on Hermann’s showing he only won one event? If victory in five events was not necessary, is it not ridiculous to suppose that a solitary victory in wrestling should have not only cancelled the four victories of Tisamenus, but secured the prize for Hieronymus?

The only inference which we are justified in drawing from the story of Tisamenus is that victory in three out of the five events was sufficient. This is expressly stated by a scholiast to Aristides, and is implied in a highly metaphorical passage in Plutarch describing the different points in which the letter A is superior to all the other letters of the alphabet.^[629] It has been further inferred that victory in three events was not only sufficient but necessary. The writers who have taken this view generally assume that with several competitors competing against one another it would be unusual for any individual to win three events, and various elaborate theories have been devised to get over this difficulty. Of these theories by far the most reasonable was that suggested by Professor Percy Gardner in the first volume of the Journal of Hellenic Studies. He supposed that the pentathlon was treated as a single event, and the competition was conducted as a tournament, the competitors being arranged in pairs, and each pair competing against each other in all five contests. The winner of each pair, and therefore the final winner, must necessarily have won three out of the five events. This plan has the conspicuous merit of fairness and simplicity, but it is open to several serious objections. In particular, the passage of Xenophon quoted above seems decisive against it, for Xenophon’s words naturally mean that all the events in the dromos took place before any of the wrestling. There are many practical objections. The length of such a competition would have made it tedious to spectators and competitors alike, and it must have degenerated into a mere test of endurance, in which the elements of skill, activity, and grace which made the pentathlon so popular would have been lost. I need not dwell on the hopelessly unpractical modifications of this theory proposed by Dr. Marquardt, nor on the ludicrously unfair systems suggested by Fedde, and more recently by Legrand in Daremberg and Saglio, the principle of which is the arrangement of all competitors in groups of three. It will be sufficient to examine the two assumptions on which these theories rest, viz. that in an open competition it would be unusual for any competitor to win three events, and that victory in three events was necessary. If these assumptions prove to be unfounded, the raison d’Être of all these theories disappears at once; for they have no merit whatsoever except that they satisfy these supposed conditions.

In considering the first point we must remember that the pentathlete was not a specialist in any one exercise, but an all-round athlete who combined strength and activity. Among competitors of this sort it is not unusual to find one or two men surpassing their fellows not in one event but in several, especially if most of the events require much the same qualities and physique. This was undoubtedly the case with the pentathlon. It is obvious that the same man might often win the foot-race and the long jump, or the diskos and the spear. Though less obvious it is equally probable that the diskos and the long jump might fall to the same man. It is not uncommon to find a hammer-thrower who is also a good long-jumper. The reason is that weight-throwing and jumping both require a harmonious well-timed effort of every part of the body. The use of jumping weights increased the resemblance between the two exercises; for the swing of the weights was not unlike the swing of the diskos. The general development and complete control of the muscles necessary for these events would give an equal advantage in wrestling, especially with men of the same weight, for the heavy-weight wrestler would be excluded by the very nature of the pentathlon. These considerations make it probable that the five events would commonly be divided between two or at most three competitors, and the few details which we know of actual winners confirms this view. Phayllus of Croton must have won the jump, the diskos, and the foot-race, for he won the stade-race at Delphi. Hieronymus won the diskos, spear, and wrestling. So apparently did Automedes of Phlius.^[630] Diophon, the subject of Simonides’ epigram, apparently won all five events. The only example to the contrary is the mythical pentathlon of Peleus, in which none of the heroes won more than one event.

The pentathlon of Peleus is fatal to the second assumption that victory in three events was necessary. We must either reject the evidence of the story, or abandon the assumption. And inasmuch as there is absolutely no proof of the assumption, the latter is the only course. The principal evidence on which the assumption is based has already been stated. The utmost that we can infer is that victory in three events was sufficient, and was by no means an unfamiliar result. We may further add the statement of Pollux that the term used for victory in the pentathlon was ?p?t????a?, “to win a treble,” a statement confirmed by a quite unintelligible scholion on the Agamemnon. The word t???sse?? is properly a wrestling term, meaning “to win three falls,” “to win in wrestling,” and so generally “to win a victory” or “conquer.” The cognate words t????t?? and ?t??a?t?? mean no more than “conqueror,” “unconquered.” There is no evidence of the connexion of the word in early times with the pentathlon; but the fact that wrestling was the last event in the pentathlon is itself sufficient explanation of the late use of the word ?p?t????a? to denote victory in the pentathlon, especially if, as was frequently the case, the final victory was decided by the wrestling. It is, of course, possible that the word contained some allusion to a victory in three events, but this supposition is unproved and unnecessary, and certainly does not warrant the assumption that victory in three events was necessary.^[631] Such being the case we may reject all theories based upon this assumption. Above all, there is no longer any necessity for dividing competitors into heats of two or three.

A common feature in the systems proposed is the gradual reduction of the number of competitors at each stage of the competition, so that in the final wrestling only two or three competitors were left. The only evidence for the theory in this form is the rhetorical passage in Plutarch already noticed—evidence as untrustworthy as it is possible to conceive. There is, however, more evidence for a modified form of the theory, viz. that only those who had qualified in the first four competitions were allowed to compete in the wrestling. This appears to me now the only possible conclusion from the words of Xenophon already quoted:^[632] “The events in the dromos were already finished, and those who had reached the wrestling were no longer in the dromos, etc.” Such a system would give an advantage to the all-round athlete, and exclude the specialised wrestler. But what constituted qualification? It certainly was not confined to the winners in the first four events, otherwise Peleus would have been excluded; nor does it seem to me probable that only the two or three who had obtained the best averages in the first four competitions were permitted to wrestle. Speculation is useless; we must be content for the present to accept Xenophon’s words, and hope that some inscription or papyrus may be discovered to enlighten us.

Much has been written by archaeologists about the bye (?fed???) in the pentathlon. It is not a little curious that there is absolutely no evidence for a bye in the pentathlon at all. We hear of a bye in wrestling, in boxing, and in the pankration, but in no other competitions. Of course, if all competitors competed in wrestling a bye was unavoidable. But a bye necessarily introduces an element of luck, especially in a long competition, and we may be sure that the Greeks avoided it as far as possible. If only a certain number of competitors were admitted to the wrestling, the necessity for a bye could be easily avoided. German archaeologists, with a strange perverseness, seem to delight in introducing compulsory byes at every turn.

So far, then, we have established the principle that victory in three events was sufficient but not necessary. If no competitor won three events, or two won two events, how was the victory decided? The pentathlon of Peleus supplies the answer. Each of the heroes won one event. Peleus, besides winning the wrestling, was second in the other four events. Only two explanations of the victory of Peleus are possible. Either wrestling counted more than other events, an assumption adopted by various writers, but contrary to the whole spirit of the pentathlon, or in case of a tie at least, account was taken of second or third places, i.e. the result was decided by marks. These two principles, that the result was decided in the first place by victories in the separate events, and in the case of a tie by some system of marks, are sufficient to explain all possible cases, though the details of their application are uncertain. Let us try to see how the competition would work out on these lines.

The pentathlon began with the foot-race. The distance was a stade. The race might be run in heats if necessary; but there is no evidence for them in the pentathlon. The starting lines at Olympia could accommodate twenty starters, and it does not seem probable that there were often so many entries. The competitions in jumping, throwing the diskos and the javelin, were conducted as in the present day, all competing against all. The jump was a long jump; the diskos and the javelin were thrown for distance, not at a mark. Wrestling was conducted on the tournament principle. “Upright wrestling” only was allowed, and three falls were required for victory. Only those who had qualified in the first four events took part in the wrestling. If there were only two competitors, one of them must have won three events. Suppose there were more, at least five, A, B, C, D, E; there is no evidence that it was possible to win the pentathlon without being first in at least one event, and, therefore, what holds good of five will hold good of any smaller or larger number. There are only four possible cases.

(1) A 3, B 2, or B 1, C 1.—A wins by the first principle.

(2) A 2, B 2, C 1.—The victory would depend on the result of the fifth event which C won. If this event were wrestling, it would be reasonable to suppose that other competitors would drop out, and A and B would be matched together. If the event won by C was one of the earlier events, the issue must have been decided by the performances of A and B in that event, or perhaps by marks, i.e. by their performances in all the events.

(3) A 2, B 1, C 1, D 1.—This is a very doubtful case: the victory might be awarded to A as having won more firsts than any of the others, or it might be decided by marks.

(4) A 1, B 1, C 1, D 1, E 1.—In this highly improbable case victory can only have been decided by marks.

Complications may have been introduced by dead heats or ties: all such cases would, no doubt, have been settled by the same common-sense principles. This scheme, which I stated more fully in vol. xxiii. of the Journal of Hellenic Studies, is not affected by the modification which I have since adopted about admission to the wrestling. It is in entire accordance with modern athletic experience, and there is no passage in any ancient author which contradicts it.