In 1916 the ideas of a night advance and of such a brief bom bardment were so unorthodox as to be a shock and appear a gamble to cautious opinion, especially as it would have to be carried out by troops of the new armies. The commander-in chief preferred a more limited alternative but Rawlinson per sisted, his confidence reinforced by that of the actual troop leaders in their ability to carry out the night operation. He gained his way, but the debate caused a day's delay which had serious consequences. The attack was delivered by the 9th and 3rd di visions of the XIII. corps on the right and by the 7th and 21st divisions of the XV. corps on the left. Cavalry divisions were brought up close and placed under the two corps commanders.
The hazardous and difficult approach march on the right was successfully carried out and at 3.25 A.M. on July 14, five minutes of ter the barrage fell, the whole line advanced to the assault. Surprise was achieved, originality vindicated. The whole of the German second system was rapidly overrun. On the right the resistance soon hardened and the 9th division only fought its way with difficulty through Longueval and to the outskirts of Delville Wood. On the left, opportunity—and the open country— stretched out its arms. Soon after midday the German resistance was obviously disintegrating in front of High Wood, and an effort was made to exploit the opportunity. But delay occurred and not until after 6 P.M. did the 7th division move forward afresh, with two squadrons of cavalry working on their flank—the first mounted cavalry seen on a British battlefield since 1914. The second advance, however, was less vigorous than the first and although most of High Wood was cleared, the northern corner and the flanking trenches held out. Worst of all the 24 hours' postponement had enabled fresh German reserves to arrive on the battlefield and as their strength steadily swelled, the German hold tightened, the British relaxed. Late on July 15 the Wood was evacuated under pressure of counter-attacks, and two months were to pass before it was regained. On July 14 the British offensive came within reach of open country and within sight of a strategic decision; thenceforward it degenerated into a cam paign of attrition. Two months of hard fighting followed during which the British were unable to make more than very gradual progress at disproportionate cost.
Progress was especially slow on the right, where Ginchy, Guille mont and Falfemont Farm formed a barrier against which many attacks were shattered even after Delville Wood had been won. It was important to extend on this side to get touch with the French who were gaining ground north of the Somme : High Wood in the centre, Pozieres on the left, were equal obstacles, and be hind Pozieres were all the formidable defences of which Thiepval was the centre. Division after division was thrown into the fight, fought desperately, lost heavily and apparently achieved little, though Pozieres was taken by the Australians before the end of July. The strain was at last beginning to tell heavily on the
Germans : they checked the Allied progress but their resources in men, guns and ammunition were, by Ludendorff's admission, severely taxed and they had to relinquish entirely their attacks on Verdun.
On Sept. 3 a renewed attack astride the Ancre by the V. Army formed from troops taken out of Rawlinson's IV. Army and placed under Gough, was unsuccessful, but the IV. Army on the right at last mastered Guillemont and in the next few days added Falf e mont Farm, Ginchy (Sept. 9), Leuze Wood and Bouleaux Wood to its gains, while the French made substantial progress north of the Somme and gained a big success south of it, taking 7,000 prisoners and storming three m. of the old German front line as far as Chaulnes. These successes removed the main obstacles to the advance of the British centre, freeing it from the menace of being enfiladed.
Haig thereupon planned a big spring forward for Sept. 15, second only in scale and ambition to July 1. The attack was to pivot on the left wing—Gough's Army. The object of the main blow, by Rawlinson's IV. Army, was to break through the Ger mans' original last line between Morval and Le Sars, in co-opera tion with a French thrust to the south between Combles and the Somme—"pinching out" Combles. If the opening success war ranted the attempt the attack was to be extended on the left to seize Courcelette and Martinpuich. Eight divisions were deployed for the original attack, and two detailed for the "extension." A special feature was the employment for the first time of tanks (q.v.), the armoured cross-country machines which had been invented as an antidote to the defensive obstacle of machine-guns and barbed wire. In disregard of the opinions of the tank's pro genitors, and of their own expressed agreement with these opinions, the British Higher Command had decided to utilise such machines as were available, to redeem the fading prospects of the Somme offensive. When this decision was taken only 6o of the initial 15o machines had been transported to France. Forty-nine were actu ally employed, to work in tiny detachments of two or three machines—another breach of the principles laid down by Colonel Swinton (q.v.). The rushed preparation combined with the mechanical defects of these early models to reduce the total, so that only 32 reached the starting point. Of these, nine pushed ahead with the infantry, nine failed to catch the infantry but helped in clearing the captured ground, nine broke down and five were "ditched" in the craters of the battlefield. The first nine rendered useful aid, especially in capturing Flers, but the greater prize of a great surprise stroke thus lost was a heavy forfeit to pay for redeeming in a limited degree the failure of the Somme offensive.