The evidence on which any attempt to fix the site of the battle of Ilipa must be based is briefly as follows. Polybius records that Hasdrubal and Scipio advanced to contest the final mastery of Spain in 206 B.C. (or less probably 207). ‘Hasdrubal encamped not far from a town called Ilipa, entrenching himself just under the hills (πρὸς ταĩς ὑπωρείαις) with a plain in front well suited for a contest and battle’ (xi, 20, 1). ‘When Scipio got near the Carthaginians and was in full sight of them he encamped on some low hills (περί τινας γεωλόφονς) opposite the enemy“ (§9). When the Carthaginian cavalry attempted to take Scipio off his guard, the Roman cavalry, which had been stationed under a hill (ὑπό τινα βουνόν), charged the enemy so suddenly that many of the Carthaginians were unhorsed and the attack was repulsed. After several days of waiting the final battle was fought on the plain, the Carthaginian infantry being ‘at no great distance from the foot of the hill’ (οὐ πολύ τῆς παρωρείας xi, 22, 8), i.e. the hill on which their camp was placed (cf. xi, 24, 7). Polybius' account breaks off after the description of Scipio's victory, but Livy, who followed it very closely, adds two facts of importance :—(1) that after the battle the Carthaginians strengthened their rampart with large stones (xxviii, 15 ; which suggests that their original camp was an earthwork), and (2) that Hasdrubal did not have to cross the Baetis (Guadalquivir) on his retreat to the Atlantic, i.e. the battle was fought on the right bank of the river. Appian adds little of value : his statement that the two camps were ten stades apart is clearly false, since the space (less than 2 km.) is too small to permit the free deployment of two armies as large as these must have been.