The number of the slain (as of captured ships, ch. 115) is probably trustworthy. The names of the Athenians would be recorded on the stelae which once adorned the great mound (Soros) over the tomb of the fallen heroes (Paus. i. 32. 3). That the Soros is the grave of the victors has been proved by recent excavations (Frazer, ii. 433-4). The barbarian dead may well have been counted on the field. The moderation of the estimate contrasts most favourably with later exaggerations--the 200,000 of the inscription in the Stoa Poikile (Suidas), 300,000 of Pausanias (iv. 25. 5), or the innumerable multitude of Xenophon (Anab. iii. 2. 12) and Plutarch (de Malign. 26; 862 B).

It is noticeable that H. gives no figures for the total number engaged on either side. Both Justin (ii. 9) and Nepos (Milt. 5) give 1,000 Plataeans, a good round exaggeration (ix. 28. 6 n.); but the former puts the Athenians at 10,000, the latter at 9,000 (cf. Paus. iv. 25. 5; x. 20. 2). These numbers are probably derived from Ephorus, and rest on a calculation of 1,000 men to a tribe. Yet they may well be near the truth, allowing for the total omission of light-armed troops. On the other hand, even the lowest ancient estimate of the Persians--200,000 foot (of whom only 100,000 fought in the battle) and 10,000 horse (Nep. Milt. 4, 5)--is greatly exaggerated, not to speak of the 500,000 of Plato (Menex. 240 A) and Lysias (Epitaph. 21), or the 600,000 of Justin (ii. 9). Modern estimates rest on conjecture or on the number of the Persian ships (vi. 95 n.), an insecure foundation. Duncker's 60,000 is an outside estimate; perhaps 40,000 would be nearer the mark. (How and Wells, A Commentary on Herodotus, 1928).