Army, 19 mobile divisions and 5 divisions in the line. The Supreme Command had kept 3 mobile divisions for the time being at its own direct disposal. These were brought up to the region of Douai shortly before the beginning of the offensive so as to be more readily available for carrying out the Mars attack, for which they were originally intended.
The bringing up of the attack formations that were destined at once for the offensive began at the end of Feb. and was com- pleted according to plan. On March 10 the bringing up of muni- tions began. During the last few nights the artillery, flame- throwers and divisions were formed up for deployment. On the morning of March 21 came the attack, delivered simultaneously by all three armies on the whole front over 47 m. wide from Croisilles to La Fere. It came as a surprise to the British III. and V. Armies. Contrary to the expectations cherished, the offensive made less progress in the first days in the case of the XVII. Army and the right wing of the II. Army than on the left wing of the II. Army and particularly in the case of the XVIII. Army, on account of the powerful British resistance. The cutting-off of the Cambrai salient failed because of the enemy's timely evac- uation. In consequence of this the Supreme Command on the afternoon of March 22 ordered the XVII. Army to extend the success of the II. Army by an attack aimed particularly in the direction of Bapaume, and to prepare for the attack on both sides of the Scarpe (Mars) with strong forces to the N. of it.
When it became evident, on the morning of March 23, that the XVIII. Army and the left wing of the II. Army were ad- vancing unchecked towards the W. and would in all probability reach their nearest objective, the Somme, on that very day, an order was issued for continuing the operation, as soon as the line Bapaume-Peronne-Ham should have been won. " The
XVII. Army will attack with strong pressure in the directions Arras-St. Pol, the left wing in the direction of Miraumont. The II. Army will take the direction Miraumont-Lihons; the
XVIII. Army will take the direction Chaulnes-Noyon and will send strong advanced troops through Ham." The three divisions held in reserve were now given to the XVII. Army.
Through this order the whole operation was pushed a long way to the left. The XVIII. Army, which was originally to have extended its front northward to Peronne on reaching the Somme, thereby releasing forces from the II. Army to carry on the attack N. of the river, had now instead to cross the Somme and advance its right wing in a slightly S.W. direction toward Chaulnes. There were thus portions of the II. Army left S. of the Somme as well as the XVIII. Army. The Somme was therefore no longer used as a support against a French flank attack, for the offensive now took a N. direction along the whole front, N. and S. of the river, with the operative aim of separating British and French.
This most fateful decision of March 23 arose from the tactical consideration of exploiting the XVIII. Army's comparatively easy success to the benefit of the general battle situation, by making a rapid forward push. Had the XVIII. Army stopped at the Somme and the Crozat canal, as was originally^ intended, extending with its right wing only northward to Peronne, its fine initial success would have had no effect on the advance of the attack farther to the N., which had up till then not quite come up to expectation. If, on the other hand, it had carried its attack across the Somme and the canal toward the W., in conjunction with the S. wing of the II. Army, the enemy, who was still holding out against the XVII. Army and the N. wing of the II. Army, would have been threatened on his right flank. From the strate- gical standpoint the decision was even more difficult and more vital. The fundamental idea of the Michael operation had from the first always been that of beating the British and the British only. The French were only to be held off from intervention by flank action. To this end the whole of the XVII. and II. Armies were to find a field for operations to the N. of the Somme. The reinforcements sent by the Supreme Command were also to follow in this direction, being mainly disposed in echelon in rear of the II. Army's left wing for the purpose of taking over the flank protection down the Somme from Peronne. At a later stage there were probably some portions of the XVIII. Army similarly
engaged N. of the Somme. This whole strategical idea would have fallen to pieces if the strong natural obstacle of the Somme had been relinquished as a support at this point. It was clear that the left wing of the armies, entrusted with the offensive solution of its task on the far side of the Somme and the canal, would very soon not only draw upon itself considerable enemy forces, but would gradually have to prepare for a counter-offen- sive steadily increasing in strength. It had therefore to be rein- forced from the reserves, which thus, as well as a portion of the II. Army, were no longer available for use in the direction fol- lowed by the main operation. But despite these apparent disadvantages the decision must be approved from the strategical standpoint also. The position of the XVII. Army and the north wing of the II. Army, as it was on the morning of March 23, made it doubtful whether the strong enemy resistance here could be broken in time to arrive at operations in the open field at all before the arrival of enemy reinforcements. There was a danger that not only the British but the French might throw strong forces on to the battlefield N. of the Somme, and so block the break-through in or close behind the British trench system. This was made easier by the contact with the positions E. of Arras, which had been maintained. The Somme at Peronne or farther to the W. would then serve the defender just as well for a safe support as the attacker had hoped it would serve him. It would also form an excellent obstacle for the front farther on up to Ham. As the direct intervention of the French, according to the way in which things turned out, had to be reckoned with, everything depended on hindering them from carrying it out systematically. The French must be caught up into the whirlpool of destruction. But this could only be done by forcing their Bray-Noyon front and attacking impet- uously in the open.
In the days that followed Geri. Ludendorff held stubbornly to his operative aim of separating the British and French. The distribution of the reserves that had been brought up later was organized accordingly. The centre of gravity of the XVII. Army's advance, originally directed toward St. Pol, was now, on March 24, shifted more to the S. toward Doullens. The course of the battle on the whole front, up to March 25 inclusive, justified the expectation of achieving its ambitious aim. As the XVII. Army had pushed its way through the whole system of enemy positions and had advanced with its S.wing to beyond the Ancre, it too began to operate in the open field. The situation was now such as to warrant the attempt to dislodge the enemy front, both at Arras and farther to the N., by frontal attacks; and the Mars attack on both sides of the Scarpe was fixed for March 28.
The first faint doubts as to the possibility of carrying out the main operation to its full extent might have been aroused by the experiences of the XVII. Army on March 26. Its S. wing, on which everything depended, gained very little ground beyond the Ancre. As, however, in the meantime the II. Army's right wing had achieved the difficult crossing at Albert, there was hope that the XVII. Army's advance would also quickly get into its stride again. On all the rest of the front the brilliant progress of the offensive so far, particularly the impetuous forward push of the II. Army in the direction of Amiens and that of the XVIII. Army toward Montdidier, raised expectations of a continuation full of promise. Ludendorff proposed to bring about the separa- tion of his opponents by a gradual concentration of the II. and XVIII. Armies against the French, and to this end the Somme below and at Amiens had to be reached and also the Avre. But the far-reaching aims of the Supreme Command were not to be realized. On March 27 the XVII. Army's offensive came to a standstill, and the next day brought the failure of the Mars attack on both sides of the Scarpe. At this point, therefore, the operation against the British was finally abandoned. Ludendorff derided to attack their front as soon as possible in a different place, and ordered the immediate preparation of an attack on the VI. Army's right wing on the Lys front in the direction of Hazebrouck. It would, however, probably be 8 or 10 days before this could begin. Otherwise the continuation of the Michael operation, as it had turned out, seemed to promise success only
