We have said that the 1st and 3rd Armies were to strike first. Tanks belonging to the 7th and 11th Battalions of the 1st Brigade were to fight with the Canadians and the 4th Corps opposite Bourlon and Gouzeaucourt, and the 2nd Brigade was to contribute the 15th Battalion, which was to co-operate with the 17th Corps opposite Graincourt and Flesquières. Altogether fifty-three fighting Tanks were to be employed. As in the 4th Army sector, the peculiar lie of the country was the chief influence which shaped our battle tactics, as in the 4th Army area a canal was the central feature of the attack. In the First Battle of Cambrai the Tanks had all attacked from south of the northward bend of the Canal du Nord near Havrincourt, and so worked up the enemy’s side of this great obstacle. Now we were in a better position to force a direct crossing, both strategically and mechanically, and the hazardous venture was to be attempted. Direct ground reconnaissance of the Canal itself was impossible, as the enemy held the hither bank in strength, but every conceivable source of information was exhaustively exploited in the endeavour to find crossing-places for the Tanks, that might offer at least a possibility of success. Daring flights were made by special observers in low-flying aeroplanes, and a wonderful mosaic was pieced together from successive sets of air-photographs. This was annotated, re-photographed, enlarged, and circulated to all concerned for further amplification and annotation as additional information was collected; Major Macavity of the Canadian Corps Intelligence, and Captain Oswald Birly of 1st Army Headquarters, being largely responsible for the thoroughness of this, as well as of several previous “over-the-line” surveys. In addition, the _Garde Champêtre_, the _Ponts et Chaussées_ service, and the engineers’ working drawings for the Canal, were all laid under contribution, as well as the evidence of a number of prisoners, refugees and _repatriés_. From such sources and on such evidence the requisite number of crossings were at length determined on, and the Tanks definitely and severally allotted to them, for good or ill. But when all had been done, there were one or two points about which there still remained a disquieting element of doubt. At one of these, where aerial photographs showed a breach through the retaining banks of the dry Canal that just might, or that just might not, allow sufficient width for Tanks to cross, a crossing was imperative for the local success of the attack. Somehow, a passage had to be positively assured--and there seemed but one sure way of keeping our contract with the infantry, who were to storm the Canal at that place.