feisal was a fine, hot workman, whole-heartedly doing a thing when he had agreed to it. he had pledged his word that he would go at once to wejh; so he and i sat down together on new-year’s day for consideration of what this move meant to us and to the turks. around us, stretching up and down the wadi yenbo for miles, in little groups round palm-gardens, under the thicker trees, and in all the side tributaries, wherever there was shelter from the sun and rain, or good grazing for the camels, were the soldiers of our army. the mountaineers, half-naked footmen, had grown few. most of the six thousand present were mounted men of substance. their coffee hearths were outlined from afar by the camel saddles, pitched in circles round the fire as elbow-rests for men reclining between meals. the arabs’ physical perfection let them lie relaxed to the stony ground like lizards, moulding themselves to its roughness in corpse-like abandon.

they were quiet but confident. some, who had been serving feisal for six months or more, had lost that pristine heat of eagerness which had so thrilled me in hamra; but they had gained experience in compensation; and staying-power in the ideal was fatter and more important for us than an early fierceness. their patriotism was now conscious; and their attendance grew more regular as the distance from their tomes increased. tribal independence of orders was still maintained; but they had achieved a mild routine in camp life and on the march. when the sherif came near they fell into a ragged line, and together made the bow and sweep of the arm to the lips, which was the official salute. they did not oil their guns: they said lest the sand clog them; also they had no oil, and it was better rubbed in to soften wind-chaps on their skin; but the guns were decently kept, and some of the owners could shoot at long range.

in mass they were not formidable, since they had no corporate spirit, nor discipline nor mutual confidence. the smaller the unit the better its performance. a thousand were a mob, ineffective against a company of trained turks: but three or four arabs in their hills would stop a dozen turks. napoleon remarked this of the mamelukes. we were yet too breathless to turn our hasty practice into principle: our tactics were empirical snatchings of the first means to escape difficulty. but we were learning like our men.

from the battle of nakhl mubarak we abandoned the brigading of egyptian troops with irregulars. we embarked the egyptian officers and men, after turning over their complete equipment to rasim, feisal’s gunner, and abdulla el deleimi, his machine-gun officer. they built up arab companies out of local material, with a stiffening of turk-trained syrian and mesopotamian deserters. maulud, the fire-eating a.d.c., begged fifty mules off me, put across them fifty of his trained infantrymen, and told them they were cavalry. he was a martinet, and a born mounted officer, and by his spartan exercises the much-beaten mule-riders grew painfully into excellent soldiers, instantly obedient and capable of formal attack. they were prodigies in the arab ranks. we telegraphed for another fifty mules, to double the dose of mounted infantry, since the value of so tough a unit for reconnaissance was obvious.

feisal suggested taking nearly all the juheina to wejh with him and adding to them enough of the harb and billi, ateiba and ageyl to give the mass a many-tribed character. we wanted this march, which would be in its way a closing act of the war in northern hejaz, to send a rumour through the length and breadth of western arabia. it was to be the biggest operation of the arabs in their memory; dismissing those who saw it to their homes, with a sense that their world had changed indeed; so that there would be no more silly defections and jealousies of clans behind us in future, to cripple us with family politics in the middle of our fighting.

not that we expected immediate opposition. we bothered to take this unwieldy mob with us to wejh, in the teeth of efficiency and experience, just because there was no fighting in the bill. we had intangible assets on our side. in the first place, the turks had now engaged their surplus strength in attacking rabegh, or rather in prolonging their occupied area so as to attack rabegh. it would take them days to transfer back north. then the turks were stupid, and we reckoned on their not hearing all at once of our move, and on their not believing its first tale, and not seeing till later what chances it had given them. if we did our march in three weeks we should probably take wejh by surprise. lastly, we might develop the sporadic raiding activity of the harb into conscious operations, to take booty, if possible, in order to be self-supporting; but primarily to lock up large numbers of turks in defence positions. zeid agreed to go down to rabegh to organize similar pin-pricks in the turks’ rear. i gave him letters to the captain of the dufferin, the yenbo guardship, which would ensure him a quick passage down: for all who knew of the wejh scheme were agog to help it.

to exercise my own hand in the raiding genre i took a test party of thirty-five mahamid with me from nakhl mubarak, on the second day of 1917, to the old blockhouse-well of my first journey from rabegh to yenbo. when dark came we dismounted, and left our camels with ten men to guard them against possible turkish patrols. the rest of us climbed up dhifran: a painful climb, for the hills were of knife-sharp strata turned on edge and running in oblique lines from crest to foot. they gave abundance of broken surface, but no sure grip, for the stone was so minutely cracked that any segment would come away from its matrix, in the hand.

the head of dhifran was cold and misty, and time dragged till dawn. we disposed ourselves in crevices of the rock, and at last saw the tips of bell-tents three hundred yards away beneath us to the right, behind a spur. we could not get a full view, so contented ourselves with putting bullets through their tops. a crowd of turks turned out and leaped like stags into their trenches. they were very fast targets, and probably suffered little. in return they opened rapid fire in every direction, and made a terrific row; as if signalling the hamra force to turn out in their help. as the enemy were already more than ten to one, the reinforcements might have prevented our retreat: so we crawled gently back till we could rush down into the first valley, where we fell over two scared turks, unbuttoned, at their morning exercise. they were ragged, but something to show, and we dragged them homeward, where their news proved useful.

feisal was still nervous over abandoning yenbo, hitherto his indispensable base, and the second sea-port of hejaz: and when casting about for further expedients to distract the turks from its occupation we suddenly remembered sidi abdulla in henakiyeh. he had some five thousand irregulars, and a few guns and machine-guns, and the reputation of his successful (if too slow) siege of taif. it seemed a shame to leave him wasting in the middle of the wilderness. a first idea was that he might come to kheibar, to threaten the railway north of medina: but feisal improved my plan vastly, by remembering wadi ais, the historic valley of springs and palm-villages flowing through the impregnable juheina hills from behind rudhwa eastward to the hamdh valley near hedia. it lay just one hundred kilometres north of medina, a direct threat on fakhri’s railway communications with damascus. from it abdulla could keep up his arranged blockade of medina from the east, against caravans from the persian gulf. also it was near yenbo, which could easily feed him there with munitions and supplies.

the proposal was obviously an inspiration and we sent off raja el khuluwi at once to put it to abdulla. so sure were we of his adopting it that we urged feisal to move away from wadi yenbo northward on the first stage to wejh, without waiting a reply.