THE GREATEST NAVAL battle of antiquity took place from dawn to dusk on 25 September 480 BC (has anyone ever noticed, by the way, that the great sea battles of history always seem to occur in autumn - Actium, Lepanto, Quiberon, Trafalgar, Leyte Gulf, etc?). The battle of Salarnis was the climax of King Xerxes' attempt to extend the dominion of the Persian Empire to the city states of the Greek mainland. According to Barry Strauss, 200,000 souls took part in the battle, with another 100,000 in the environs. These figures seem absurdly high, especially given the huge logistical problems involved (eighty-four supply ships would need to have been plying between the Persian base in Macedonia and the site of the battle at any one time just to keep the invasion forces fed and equipped), but Strauss's careful research is as persuasive on statistics as it is on all other aspects of this grim struggle. Let it be said straight away that this is a superb example of the use of scholarship to underpin a fine piece of narrative writing.

It is well known that ancient historians routinely added several noughts on the end when estimating any numbers, so we can safely discount Herodotus' 'millions' when discussing statistics in the Persian wars. Strauss's estimate that Xerxes' army probably amounted to 200,000, of whom 150,000 were actual combatants, seems about right