A Turkish retreat from Medina at that moment would have been a great victory for the Arabs, but for the British a disaster. In December 1916 Murray’s Egyptian Expeditionary Force had crossed Sinai, and Murray was now preparing for a massive push against the Turkish defences in Palestine, which ran from Gaza on the coast to Beersheba inland. To have the entire Medina garrison – 12,000 Ottoman troops with full artillery – arrive suddenly on his right flank would have been most unwelcome. In a letter to Lawrence, Clayton stressed that Fakhri Pasha’s force must be attacked and destroyed before it could reach Palestine. Under no circumstances should it be allowed to get through. Newcombe and Garland were still up-country on demolition-raids, and since Lawrence was the senior British officer in Wejh, he had no choice but to take charge of the situation. His first task required all his diplomacy, and called on the relationship of trust he had built up with Feisal. He had to explain to the Sharif that Arab priorities must be sacrificed in this case for British ones. Feisal, to whom Lawrence had recently revealed the terms of the Sykes–Picot agreement, had been anxious to push north into Syria to consolidate the Arab position there before the French could claim it. It was a great tribute to Lawrence’s persuasive power that he brought Feisal round within the space of a few hours to accept the priority of British requirements. Once done, he and Feisal sat down to decide on the distribution of their forces. Messengers were rushed off to ‘Ali and Abdallah to alert them to the new situation. ‘Ali was to move north-east with the Juhayna and Bani Salem, ‘Abdallah was to be sent dynamite and instructed to hit the railway at any point and at any cost. Maulud al-Mukhlis and Rasim Sardast were to go to Faqir with the Mule Mounted Infantry and a mountain-gun, Sharif ‘Ali ibn Hussain of the Harith – the ‘young lord’ Lawrence had encountered on his first ride to Hamra the previous October – was to go to Jayala to harass the line. Sections of the Billi and Wuld Mohammad were to take machine-guns and menace the station at al -‘Ula. The plan, made on the spur of the moment, was to contain the Turks south of that line. If they managed to pass al-‘Ula then they would be in the protection of the large Ottoman garrison at Tebuk. A key point in their retreat would be the station at Hediyya, the only permanent water supply for 200 miles, and Lawrence earmarked this as his own target. Even before arriving in Wejh he had talked about the possibility of visiting ‘Abdallah in Wadi Ais. Now, he decided, he would travel to ‘Abdallah’s camp, explain the new strategy, have a look at the railway, mine a train, and if possible capture a station. He set off for Wadi Ais with an escort the same night.

This was Lawrence’s first major operation in the field. He was not a trained field-officer, as he had always insisted, but a political officer, and his proper place was by Feisal’s side. Garland and Newcombe, both Sappers, had preceded him in action against the railway. However, this was an emergency, and Lawrence felt that ‘Abdallah’s force, sitting in Wadi ‘Ais, had done little to justify its existence over the previous months. He would now act as a spur to Abdallah, as he had done so effectively with Feisal, and in the process would strike at the key point on the railway himself. As he rode out of Wejh with his escort, it must almost have seemed to Lawrence that the entire future of the Near East campaign rested on his shoulders. As soon as he was on the road, however, the old terror of being injured reasserted itself with a vengeance. Almost every moment of profound stress in Lawrence’s life is marked with illness: his journeys in Syria had ended in malaria and dysentery: at the crucial point on his mission to Kut in 1916 an attack of fever had laid him out. The malaria was genuine, but it returned at intervals when his fear brought his psychological defences down. His ride to Wadi Ais, the terrible responsibility of his position, the necessity of doing right both by the Arabs and by his British masters, and the prospect of standing between the Turkish wolf and its home ground, took their psychical toll. By the second evening, when his party camped at the pool of Abu Zeraybat, where Feisal had lingered on his march to Wejh, Lawrence was suffering from fever, boils and dysentery. The following day – 12 March – the party set out early. Lawrence’s companions – a Moroccan named Hamed, a Syrian cook, some ‘Agayl, Rifa’a, a Merawi and an ‘Utaybi – spent the day arguing continuously. The going was difficult, and for Lawrence agonizing. After a short break at mid-morning they rode up a narrow water-course towards the Sukhur – vast striated masses of cracked and faulted volcanic rock – where they were obliged to dismount and drag their camels up over rocky shelves and a knife-blade ridge. They descended into a valley, which opened into another, and another, until they strayed into an area of black basalt boulders known as a harra, where the camels tripped and stumbled. The sun came out with a vengeance, and twice during the day Lawrence fainted. It was all he could do to stay in the saddle. In the afternoon they were obliged to make two difficult and steep descents which only added to his fatigue. Finally, at 4.15 they halted for the night in a stony water-course called Wadi Khitan, where Lawrence unhitched his saddle-bags and threw himself into the shade of some rocks, exhausted, with a headache and raging fever. What happened next forms one of those mysterious hiatuses which feature so commonly in Lawrence’s life, when we seem to be passing from the realms of solid fact – times, distances, numbers and dates, with which he crammed his diaries and letters – into a subterranean world of nightmares and shadows.

According to Lawrence’s description in Seven Pillars, his feverish reveries were suddenly disturbed by a crisp gunshot at close quarters. This did not bother him, as he imagined one of his Bedu was shooting a hare for supper. Not so. He was shortly roused by Sulayman, the ‘Utaybi, who led him over to another gully to show him the dead body of an ‘Agayli called Salem, with a gaping bullet wound in the temple. Lawrence clearly saw that the skin was burned at the edges, suggesting a shot from close quarters. At once he suspected Sulayman himself, remembering the feud between the ‘Utayba and the ‘Agayl, but ‘Ali – the senior ‘Agayli – assured him that the murderer had been none other than Lawrence’s own servant, Hamed the Moor. Lawrence sent the men to search for Hamed, and had just crawled back under his cloak when he heard a rustle. Opening his eyes wide, he saw the Moor lifting some saddle-bags nearby, evidently intending to load his camel and make off. Lawrence drew his pistol and stopped him in his tracks. The other Arabs came rushing back, and at once held court. Hamed confessed that he and Salem had argued, and he had lost his temper and shot the ‘Agayli at point-blank range. ‘Ali and the other ‘Agayl demanded an eye for an eye. Lawrence knew that this was the ancient desert law, and was anyway too shattered to argue the case for clemency – he agreed that the murder of Salem had been an unforgivable crime. It was clear that Hamed had to die, but who would perform the execution? If he died at the hands of an ‘Agayli, this would start another blood-feud between the Agayl and the Moroccans, of whom there were many in Feisal’s army. Only Lawrence, who stood in the role of a Sharif, and above tribal feuds, could safely execute the condemned man. He made the Moor enter a sandy gully which shrank to a crack a few inches wide, and allowed him a brief pause to come to terms with himself. Hamed crouched sobbing on the ground. Then Lawrence ordered him to stand up, and shot him in the chest with a trembling hand. The Moor collapsed, coughing blood, and Lawrence shot him once more but only fractured his wrist. Hamed lay in the sand, screaming, and Lawrence stepped close to him, laid the muzzle of his pistol under his jaw, and shot him for the third time. The body shivered slightly. Lawrence called the ‘Agayl to bury him, staggered over to his baggage and collapsed in his sleeping space. His diary entry for 12 March consists of a rough sketch-map of the Wadi Khitan with an arrow pointing to a place labelled ‘deathcrack’, and – in very spidery writing – the words: ‘Slept here. Terrible night. Shot.’⁹