The Division was given every opportunity of practising the attack with tanks, the platoons carrying out their training with the actual tanks and crews which were to accompany them in the operations. It will be seen that, as usual, liaison was synonymous with success, and that on the day of the attack the co-operation between the tanks and the infantry was admirable. An additional difficulty lay in the fact that somehow or another the Division had to be concentrated in the battle area some thirty-six to forty-eight hours before the battle, so as to allow the troops adequate rest before proceeding to the assembly trenches. On this account the C.R.E. and the three field companies, R.E., with the 8th Royal Scots, moved to the IVth Corps area early in November to prepare hidden shelters in which the troops might be accommodated during this period. The sappers and pioneers worked with such effect that between the 2nd and 19th November camouflaged accommodation had been provided for 5500 men in the ruined village of Metz and for 4000 men in Havrincourt Wood. The necessary dumps of material were also formed by the C.R.E., tracks for moving guns forward prepared, and, as cavalry were also detailed to take part in the operations, six water points with a capacity for watering 7000 horses per hour were constructed. Forward routes for infantry were also laid out and advanced dressing stations completed; no mean record for a little over a fortnight's work. All undertakings of this nature were made more difficult than usual by the fact that no increase in the normal amount of lorry traffic on the roads during the hours of daylight was allowed. Further, no new work was carried out in the forward area, in case it might be noticed by enemy airmen. As it turned out, the weather was fortunately cloudy and misty, so that observation from the air became practically impossible throughout the period of preparation. The area through which the Division was destined to advance was traversed by three separate trench systems, each forming integral parts of the Hindenburg system. Of these, the first, known as the Hindenburg front system, was composed of a maze of wide, heavily-wired trenches, supplemented by numerous saps, the whole presenting such a tangle of excavations that it was impossible to foretell accurately where the principle points of resistance were most likely to be found. The main framework of this front system was, however, composed of a lightly-held outpost line, a front line, and a support line. In rear of the front system, and just south of the village of Flesquières, lay the Hindenburg support system, composed of two lines of heavily-wired deep trenches, connected with each other and with the front system by numerous communication trenches.