[76] The original passage in which this is mentioned is the scholion on Aristophanes, Knights, 778: ἐν Μαραθῶνι: τόπος τῆς Ἀττικῆς εἰς ὃν ἐνώρμησαν Δᾶτις καὶ Ἀρτάβαζος Μηδικοὶ σατράπαι, etc.
Suidas, a very late author in the 10th and 11th centuries of our era, reproduces the information, drawing it evidently from the above-mentioned source.
[77] The fact that the Persians were in battle array seems to invalidate the theory which has been put forward, that the Persians, when the Greeks rushed upon them, were marching towards the lower road, with intent to reach Athens that way. If that had been so, it is not possible to imagine that the battle could have been the set battle which it appears to have been according to Herodotus’ description. His description of the main incidents of the fight seems the most absolutely reliable part of his narrative.
[78] In the late author Suidas there is a note which gives a positive support to this negative fact of Herodotus’ silence. He says that the expression χωρίς ἱππεῖς had become a proverbial expression, originating from the fact that when Datis invaded Attica, the Ionians who were with him informed the Athenians that the cavalry were away, and so Miltiades attacked and won the victory.
[79] The total loss of the Persians is given by Herodotus at 6400. This is not likely to be an understatement in that 6400 must have been included the greater part of the Persian centre, which seems to have been all but annihilated, if the circumstances of the battle be taken into consideration. It is thus against probability that more than 20,000 Persian troops took part in the engagement.
The comparative smallness of the Persian numbers is further indicated by the words which tradition attributed to Miltiades. He is said in the course of his appeal to Kallimachos (ch. 109) to have spoken confidently of success in case the Greeks took the offensive. There is no reason to insist on such words having been actually used by Miltiades. The important historical point is that in the tradition of the battle which Herodotus followed such language was attributed to him, language which could not conceivably have been used had the Persians very greatly outnumbered the Greeks.
[80] Numbers varying from 100,000 to 500,000 are given by various later authors.
[81] Cornelius Nepos, Justin, Suidas, give estimates of from 9000 to 10,000 Athenians, and 1000 Platæans. Herodotus (ix. 28, 29) gives the Athenian numbers at Platæa as 8000 hoplites and 8000 light-armed, and this at a time when many citizens were serving on the fleet at Mykale. The number 10,000 at Marathon is probably an understatement, though not one of a gross character.
[82] The Persian position is indicated by three circumstances:—
(1) The position of the “Soros,” which would presumably be situated where the majority of the Athenians must have fallen, i.e. in the centre of the line; and where, too, the decisive blow of the battle was struck.