Wounded and repulsed, Alexander took thought for his machines. A mound was ordered, as at Gaza, from which the catapults and siege-towers could bring their barrage within range, but the natives had the pleasure of seeing their first drawbridge collapse beneath the weight of its Macedonians. The catapults were more formidable, and when one of the arrow-shooters killed the Indian chieftain, the tribesmen gave in to superior engineering. Their hired troops surrendered and were taken into Alexander's ranks, only to be massacred when they tried to escape on the following night. The rest were spared, including the chieftain's mother, who made up for the loss of her son to a catapult by sleeping with Alexander and conceiving a replacement.

It was now midwinter, and the storming of two more highland citadels kept the army busy well into the new year; the righting continued to be strenuous and the weather chilly but food was no distress: only three months' earlier, a herd of 230,000 cattle had been captured, the finest of which had been chosen by Alexander who 'wished to send them back to Macedonia to work the land', a tribute to his keen eye for the agriculture on which all ancient economies depended: the rest sufficed to feed the army for several months on milk or meat, two rare luxuries in the diet of the classical world. Meanwhile the drift of the campaign was ever eastwards. Beyond the Swat hills, on the eastern side of the river Indus, lived a rajah whose people throughout history had always supported Alexander's present enemies on the near bank. Agitators and hired troops had already come to cause trouble and as Alexander sacked and resettled one city after another on the edge of the Indus, the survivors kept retreating towards this one source of help. By early March they had been penned up into the north-east highlands overlooking the Indus itself and had fled to a steep spur north of Attock which even the hero Heracles, men said in Alexander's army, had never been able to capture. At last, the scene was set for the climax to Alexander's career as the greatest besieger in history.

In 1926 the site of this spur called Aornos was fixed by the explorations of Sir Aurel Stein, and his remarkable search has since been confirmed by archaeologists. Aornos is indeed as impressive as Ptolemy's history suggested. Where the river Indus crooks westwards above the Nandihar valley, a complex of spurs and ridges are enclosed inside its bend. Among them is Pir-Sar, the 'peak of the holy man', a long flat-topped cliff which stands over a height of 7,000 feet, guarded on the east by the broad river Indus to which it descends in a series of slippery gorges. Due north, the even higher spur of Bar Sar rises to a sharply conical point and meets Pir-Sar first by grass slopes, then by a particularly treacherous ravine, while to the west, sheer cliffs drop 2,000 feet to a strip of valley and thence rise straight to the highest peak of the whole range. To the south, Pir-Sar's terminal hillock breaks into three narrow branches, each more inaccessible than the next. As for Pir-Sar itself, its own flat summit commands a view dramatic enough to stir the most hardened Macedonian bodyguard, whether he looked away to the icebound caves of the Upper Swat headwaters or south beyond the Indus to the metallic green of the plains round Peshawar. It was a site for mountaineers, but emphatically not for warriors.

From such a vantage point Alexander's looping march up the bank of the Indus was sure to be detected. On no side of the rock could his army hope for easy access, and they had to choose between a ridgeway or a ravine. As for starving out the enemy, that was impossible because Pir-Sar had its own water-springs and a summit wide enough to grow crops for its occupiers, as it still does for the local Gujars. The cliffs and gorges were far too sheer for the catapults to come within range: when the natives talked of their Hindu god Krishna, worshipped by men dressed in lion skins, it was only natural for the Macedonians to equate him with their own royal ancestor Heracles and spread the word that not even Heracles had been able to storm Pir-Sar. To Alexander, that was another reason for attempting it himself.

From the pass, he briefly visited the Indus to check on Hephaistion's bridge-building; he then approached Pir-Sar up river from the south. At the nearest base the heaviest troops and most of the cavalry were left to prepare supplies against a long siege, while the Horse-archers and all the crack skirmishing troops continued for a day and a half up the western bank until they met with a decisive stroke of luck: nearby tribesmen surrendered, offering to lead the army to the easiest point of assault. As usual, they were believed for lack of any alternative. Ptolemy and Alexander's secretary Eumenes were sent on a reconnaissance, and striking due north from the river they seized the spur of Little Una, due west of Pir-Sar itself, helped by its covering of pine-trees and wild rhododendrons. After putting up a stockade on the hilltop, they lit a fire signal as prearranged. Alexander saw it, but so did the defenders, and it took two days' skirmishing and the despatch of a native Indian messenger before king, secretary and historian were safely united on their advance ridge.

From Little Una, the outlook was very much more favourable. Pir-Sar, said the guides, was vulnerable on its hidden north face, so Alexander

scrambled towards it along the Burimar plateau until he came to rest at the great natural fosse of the Burimar-Kandao ravine, some 800 feet deep. Here he was vulnerable from above and quite unable to fight back, as the ravine was too wide and too deep for his catapults' range. Undeterred, he once more ordered the landscape to be changed to suit him. Stakes were to be cut from the many nearby fir trees, some conveniently fallen, and a mound was to be built across a ravine as broad as a Punjab river until the catapults could find their mark. These amazing earthworks began at dawn, Alexander moving the first heap himself and then standing by 'to watch and praise those who worked eagerly but to punish those who gave up for a moment'. After one day, sixty yards were already finished, but as the sides of the ravine dropped away, the work slowed down and it was three more days before men were within fighting range of the nearest tip of Pir-Sar itself. Presumably the mound was more a platform than a complete filling; its crisscross pattern of stakes and brushwood was one more credit to the troops' skill in carpentry.

A foothold on Pir-Sar was not to be easily won. Alexander chose thirty advance guards, and 'at the sound of the trumpet, he turned to them and ordered them to follow him, for he would be the first up the rock'. Nearing the top ledge, he changed his mind and sent the guards ahead, only to see them crushed by boulders rolled down from above: he was lucky to escape with his life, and so he withdrew for the next two days. Day and night the Indians beat drums to celebrate the repulse, but the third night brought silence, a blaze of torchlight and their attempt at a surprise retreat. Hauling himself up the rockface by a rope, Alexander this time led his Shield Bearers up to the attack, killing several fugitives and clearing the summit for the building of altars to Athena goddess of victory, possible traces of which were discovered by Aurel Stein. The rock was measured by the surveyors, whose geometry was remarkably accurate, and it was then time to climb down, the last threat to the lines from Balkh to the Indus having been forcibly dispersed. 'I could only wonder,' wrote Stein, with the evidence of the landscape before him, 'that the story of Aomos should have escaped being treated altogether as a mythos. ... I had no victory to give thanks for, and yet I too felt tempted to offer a libation to Pallas Athene for the fulfilment of a scholar's hope, long cherished and long delayed.'