launched there was no longer any doubt as to the exact point or points at which the German reserves were required, and as the attack began to lose its energy and vigour the defence grew stronger. The assaulting infantry found itself in a compara- tively narrow salient, exposed to a cross fire of artillery from both flanks, and checked in front by machine-guns which could not be hastily located. It was just at this moment that the infan- try found that its own artillery support was weakened, for guns were trying to move forward to new positions and were at the same time struggling against perhaps their greatest difficulty, namely, ignorance as to the exact position of their own infantry. As the days passed the defence steadily grew stronger, until at last the attack withered away and finally ceased altogether towards the end of the first week in October. When the time came to reckon up gains and losses it was found that against a gain of some 40 square kilometres of ground, 30,000 prisoners and more than 100 guns, the French army had put a loss of 120,000 killed or missing and 260,000 wounded. The German loss was probably not much less.
The stories of the various subsidiary attacks were not very different from that of the principal French effort. There had been no break-through, but much had been learnt. In the first place it had been proved that the attack was still stronger than the defence, for the leading infantry had passed through the German lines with surprising ease, and there was a general feel- ing that at several points, notably at Loos, the Allies might well have won greater success than that which actually came to them. There was, therefore, reasonable ground for the feel- ing that, notwithstanding bitter disappointment, the advocates of the " break-through " had been justified, and that, with the experience that had been gained, success would certainly be achieved " next time."
New Tactical Lessons. The causes of failure may be summed up thus: (a) the attacks were made on too narrow fronts; (b) there was no surprise; (c) communication between infantry and artillery was not satisfactory. So far as the first of these causes of failure was concerned, it was evident that it could be remedied by allotting more men to the attack, and that in turn developed itself into the more or less mechanical question of the provision of more guns and ammunition. The second cause was more difficult, since, as has already been seen, surprise and prolonged bombardment cannot well be reconciled. One or the other must be sacrificed. The third cause of failure led to a very definite step in infantry and artillery tactics, and to the evolution of ideas which were put into practice for the first time in the fight- ing on the Somme in 1916.
As has been seen, it had already been found necessary to abandon the idea, formed in S. Africa, that " artillery does not prepare the infantry attack," and it had been realized that an adequate artillery would have had no difficulty in dealing with the Boer defences. The fault, after all, had not been in the older theories of attack, but in the British army having tried to carry them into effect with insufficient material. Had, for instance, Lord Roberts's army in Feb. 1900 been provided with heavy and light guns on the 1915 scale, Cronje's defence at 'l Paardeberg could not have lasted six hours, and the costly | infantry attacks would have been entirely superfluous. Clearly , the fault had not been in the theory of an artillery preparation, which had now been proved to be so completely indispensable that it was necessary not only against the first system of defences but also against the less elaborate lines in rear. How to secure , this cooperation against the second and third lines of defence was in Sept. 1915 the latest of tactical problems. The solution then adopted was that the guns must lengthen range or attack new targets at the request of the infantry. Such was the training and instruction of 1915, and it was only when put to the test of actual fighting that it was discovered that no system of com- munication which the wit of man had yet devised was sufficiently good and sufficiently reliable to keep artillery fully informed of the requirements of the infantry; nor was the artillery always able, for various reasons, to do exactly what the infantry wanted, even when fully informed. It was out of the confusion and
disappointments of Champagne and Loos that the theory of methodical progress with limited objectives was eventually evolved and put into practice in the Somme fighting of 1916.
The experience of the first two years, or, more accurately, twenty-two months of the war was brilliantly summarized in a memorandum on the attack, dated June 20 1916, 1 over the signature of Marshal Foch, then general commanding the northern group of the French armies. Later experience, no doubt, caused further modification of views, but this document is invaluable as showing the stage which had been reached by the best military thought of the day; for it is in every way worthy of the great reputation which the author had already won for himself as a writer and as a student of war no less than as a great commander in the field. No excuse is needed for quoting some of the more important passages.
In the first place it is assumed as a reasonable basis for con- sideration that the German defensive organization will consist in a series of prepared positions, of which two will be complete, while the others will be more or less skeleton, but sufficiently advanced to be quickly made tenable.
" Against such defensive organizations and against an adversary who is perfectly protected, infantry is incapable of preparing an attack with rifles and machine guns, just as fire power prohibits all action by 'cold steel.' For these reasons infantry has lost its offensive power in the actual battle. Guns alone can destroy the enemy's defences, upset his organization, kill or annihilate the defenders and master his artillery.
" In consequence, the r61e of the infantry is limited to seizing and occupying the ground over which the artillery has wrought effective and complete destruction."
In another part of the same paper the marshal says:
" In such conditions the attack of every position entails two phases: (i) the destruction of the defences, or the artillery prepara- tion; (2) effective occupation by the infantry (de vive force). Experi- ence has shown that when the destruction by the artillery has been effective the advance of the infantry is easy and cheap in human life. When the destruction has been but partial or not thorough, infantry is completely checked "...
" Artillery preparation is clearly the measure of infantry possi- bility. It must be resumed the moment the advance is checked. The depth of ground shelled by the artillery decides the area which can be allotted for conquest by the infantry. It is not great two, three or four kilometres." 2
From these hypotheses it follows logically that each successive position requires fresh reconnaissance, fresh bombardment and a fresh infantry assault. Infantry, however, by virtue of its own fire-power, is more than ever able to resist counter-attack.
These new theories may be summed up in the words of another high authority: " Artillery conquers defences, infantry occupies them." What a complete change two years of war had brought about from that other phrase: " The artillery does not prepare the infantry attack, but supports it"! Yet be it noted that the latter was also the direct result of war experience.
A further example of the leading position which artillery had by this time attained is seen in the rapid elaboration of the " creeping barrage." It will be remembered that during the Champagne fighting, artillery altered range or changed its targets " at the request of the infantry." This system had proved unworkable in actual practice, and for it was gradually substituted a new plan by which infantry and artillery worked together in accordance with a carefully prepared time-table. It is only necessary here to record the fact that for the first time infantry movements conformed to artillery fire, although it is also true to say that the artillery " lifts " were based upon the speed at which it was calculated that infantry could move. It is also noteworthy that, as the idea of moving close up to the wall of bursting shell, whether shrapnel or high-explosive, became more familiar to the infantry, the artillery " lifts " were reduced and the rate of progress became slower. Such tactics required no mean degree of confidence in the infantry and a high standard of accuracy in guns and ammunition as well as of training in gun-crews. Some casualties from the barrage fire
1 It will be noted that this memorandum was issued only a few days before the opening of the Somme offensive. 1 The italics are ours.
