In any case the campaign leaves the impression that Ludendorff had neither his former clearness as to the goal, nor quite the same grip on the changing situations. But in the organization of his attacks his powers were at their highest level. Surprise was to be the key by which a gate in the long-locked front was opened. In forging the key gas-shell was to be the main constituent, for Ludendorff had failed to grasp the significance of the tank and neglected to develop it in time. Only in Aug. 1918, when it was used to strike him a mortal blow did he put it in the "urgent" class of war material. The troops were trained in the new in filtration tactics already tested at Riga, and the most thorough arrangements were made for concealing and for exploiting the attacks. The assaulting divisions were to be brought up over night, the masses of artillery brought close to the front line in concealment, and their ranges obtained by methods which did away with preliminary "registration." The bombardment was to be brief but intense, and its surprise effect to be increased by lavish use of gas and smoke shell. Further, while Ludendorff had settled to strike first on the Somme sector, to which blow the code-name "Michael" was given, he also made preparations for successive attacks at other points, which besides being in readiness for the future helped to mystify the enemy. Two were on the British front and one on the French—"St. George I." against the Lys sector, "St. George II." against the Ypres sector and "Bliicher" in Champagne.
The "Michael" attack was to be made by the German 17th, 2nd, and i8th Armies (62 divisions in all) on the 47 m. front Arras-St. Quentin-La Fere, but its main force was intended to be exerted north of the Somme, and after breaking through, the I7th and and Armies were to wheel north-west and press the British army against the coast, while the river and the i8th Army guarded their flank. This plan was radically changed in execu tion because Ludendorff gained rapid success where he desired it little and failed to gain success where he wanted it most. The attack was launched on March 21 (see ST. QUENTIN, BATTLE OF 1918), and the surprise was helped by an early morning mist. But while the thrust broke through completely south of the Somme, where the defence--but also the attacking force—was thinnest, it was held up near Arras, a check which reacted on all the attack north of the river. Ludendorff, violating his new principle, spent the following days in trying to revive his attack against the strong, and strongly held bastion of Arras, main taining this direction as his principal line of effort. Meantime he kept a tight rein on the i8th Army which was advancing in the south without serious check from its opponents. As late as March 26 he issued orders which restrained it from crossing the Avre and tied it to the pace of its neighbour, the 2nd, which in turn was held back by the very limited success of the 17th Army near Arras.
Thus we see that in reality Ludendorff was bent on breaking the British army by breaking down its strongest sector of resist ance in a direct assault. And because of this obsession he failed, until too late, to throw the weight of his reserves along the line of least resistance south of the Somme. The intended wheel to the north-west might have come to pass if it had been made after passing the flank, and thus being directed against the rear, of the Arras bastion. On March 26 the attack north of the Somme (by the left wing of the 17th Army and the right of the and Army) was visibly weakening as the price of its hard-earned gains. South of the Somme the left of the and Army reached, and was now to be embarrassed by, the desert of the old Somme battlefields—a brake on progress and supply. The i8th Army alone was advancing with unslackened impetus. This situation led Ludendorff to adopt a new plan, but without relinquishing his old. He ordered for March 28 a fresh and direct attack on the high ground near Arras—by the right of the 17th Army and to be followed by a 6th Army attack just to the north between Vimy and La Bassee. But the promising situation south of the Somme led him to indicate Amiens as an additional main objec tive. Even so, he restrained the i8th Army from pushing across the Avre without further orders! On March 28 the fresh Arras attack was launched, unshielded by mist or surprise, and failed completely in face of the well prepared resistance of Byng's 3rd Army. Only then did Ludendorff abandon his original idea, and direct his main effort, and some of his remaining reserves, to wards Amiens. By March 27 the advance had penetrated nearly m. and reached Montdidier, cutting one railway to Paris; by March 3o the German flood was almost lapping the outworks of Amiens. Once the crust was broken, the very elaboration of the methods of control communication built up during three years of static warfare caused the greater flux behind the front. The extent of the retreat was primarily the measure of the loss of control by the British commanders.
Disaster had driven the Allies to an overdue step, and on Haig's appeal and Lord Milner's intervention Foch had been appointed on March 26 to "co-ordinate" the operations of the Allied armies. If he had fallen into disfavour owing to the heavy cost of his attacks in Artois during 1915 and the barren fruit of the Somme in 1916, his will-power and energy earned and created confidence. On April 14 he was definitely made commander-in-chief of the Allied armies. But before this a fresh German menace had devel oped—though not intended as such.