of psychology, were beginning already to count. The repeated losses of fortified positions, culminating in the Wonder Work and Redoubts which had resisted the assaults of July 1st, were disastrous not only on their own account but also as indicating a weakness which might conceivably spread to the Rhine. If the theory of defence proved unsound, no degree of valour in practice would ever avail to put it right. We must not prejudge this question. We are not writing the German history of the war. But it is legitimate to say that, apart from the general retirement which the Germans ordered in March, 1917, and which reached a rate of ten miles a day, our troops gradually discovered a change in the enemy’s system of defences. He began, first on the British and afterwards on the French front, to abandon the formal lines of trenches, and to employ the natural features of the soil, when and where these might occur, as the basis of his defences. The crater, or shell-hole cavity, was brought into use in this way, and no outward mark was allowed to distinguish a fortified group of craters, subterraneously connected with one another and otherwise rendered formidable, from harmless groups in its immediate neighbourhood. Thus, the cession by the Germans of ‘only our foremost crater-positions,’ or of a ‘craterfield’ _tout court_, began to figure in their reports for the edification or delusion of German readers. An integral part of the crater-system, as worked out more elaborately at a later date, was the ‘pill-box,’ or sunk blockhouse, which was strengthened towards the foe and left more thinly built on the home side, so as to render it useless as a weapon should its fire be directed by its captors. We may conclude that the blows which had been dealt at the continuous lines of trenches in the battles of the Ancre and the Somme had alarmed the German High Command; and that a part of the motive for the retirement (and a very effective part it proved) was to prepare those fortified groups and concrete nests of deadly machine-gun fire at all kinds of irregular distances. The intention was partly to deceive the airman’s eye, and to stop that preparation of exact trench-maps to which the Germans had borne testimony on the Somme. But partly, too, the modification of the defence-system implied that our offensive had not been vain. Its immediate effect, accordingly, however serious and impeding it was to prove, was not without good hope. The vaunted theory of ‘impregnability’ had been shaken, and, though the end of the war was still out of sight, yet Thiepval, like Jutland, bore a message which the rest of the war was to expound.