There was no moon, and we found it difficult to drive our tanks into the wood without knocking down trees that made valuable cover. It was none too easy without lights, which we did not wish to use, to fasten the camouflage nets above the tanks on to the branches. The track of the tanks from the ramp to the wood was strewn with branches and straw.

By the time we had finished the night had fled, and it was in the fresh greyness of dawn that we marched the weary miles to the camp at La Lovie. The men were dog-tired, my guide was not certain of the road, though he never missed it, and I had never realised the distance. After an interminable tramp we staggered into camp. The men were given some hot breakfast, and then, as the sun rose, you would have heard nothing but snores. For our sins we had arrived in a "back area" of the Salient.

That was on the 11th of July: the next twenty days were crammed full of preparations.

Every morning the men marched down to the wood, wondering a little if the shelling during the night had done any damage​—​and Oosthoek Wood was shelled every night. Gradually the tanks were "tuned" to the last note of perfection, the new Lewis guns were fired, and finally the tanks were taken out on a cloudy day to a field close by and the compasses adjusted by "swinging." Names and numbers were painted. Experiments were made with the new and not very satisfactory form of "unditching gear." Supplies of water, petrol, and ammunition were taken on board. Everything that the crews could do was done.

We were told soon after we had arrived in the Salient that during the first stages of the great battle "D" Battalion would remain in reserve. There was, in consequence, no need for us to make any elaborate reconnaissances of our own trench system, because by the time that we were likely to come into action it was probable that we should be beyond trenches and operating in the open country.

If a tank company is ordered to attack with the infantry on the first day of a battle, no reconnaissance can be too detailed and patient, for on the night before the attack a tank can do untold mischief. There are wires, light railways, emplacements, communication trenches, dug-outs to be avoided, and a specific spot to be reached at a given time. Tanks unfortunately are not allowed to roam wildly over the battlefield either before or during a battle. The route that a tank will take from the moment it starts to move up on the night before the battle to the moment it rallies after the battle is only a few yards wide. It is chosen after the most painstaking examination of aeroplane photographs and the daily reconnaissance of the enemy country. To our own front line the route is taped, and forward it should be taped​—​in the mind's eye of the tank commander.

Nor was it necessary for us to "liaise" with the infantry. Immediately a tank company commander learns that he is "going over" with a certain battalion of infantry, he begins at once to establish the closest possible "liaison." The infantry officers are entertained and shown over the tanks. A demonstration is arranged, and if time permits a dress rehearsal of the attack is carried out in order that there may be a thorough understanding between the tanks and the infantry. At the beginning of the Ypres battle combined tactics scarcely existed. The infantry attacked, the tanks helped, and the only question to be decided was whether the tanks went in front of the infantry or the infantry in front of the tanks. But even in July 1917 it was just as well to know personally the officers and men of the battalion concerned, although as late as September 1918 one Divisional Commander refused to tell his men that they would be attacking with tanks, in case they should be disappointed if the tanks broke down before the battle.

We had only to reconnoitre the routes to the canal, and make a general study of the sector in which we might be engaged.

Nothing, I suppose, sounds more elementary than to take a marked map and follow a tank route from a large wood to a canal which cannot be avoided. In practice there are not a few little difficulties. First, it is necessary to extricate the tanks from the wood without knocking down the trees, which may later be required to shelter others from aeroplanes. This requires care and skill. Then the tanks proceed along a cart-track until the route crosses a main road by a camp, where it is necessary to swing sharply to avoid important wires and some huts. Beyond the main road we trek across a field or two until the track divides, and it is easy enough in the dark to bear to the right instead of to the left. Then there is a ditch to cross, with marshy banks​—​a good crossing in dry weather, but doubtful after rain​—​and we mark an alternative. We come to a light railway, and this under no circumstances must be damaged. We arrange for it to be "ramped" carefully with sleepers, but it is just as well to carry a few spare sleepers in the tanks, because some heavy gunners live near by. The track, which by this time is two feet deep in mud, again divides, and bearing to the right we find that an ammunition column has camped across it. So we suggest that tanks through horse lines at night may produce dire results, and a narrow passage is cleared. Another main-road crossing and a bridge​—​we are doubtful about that bridge, and walk down the stream until we come to something more suitable to our weight. Along the route we look for woods, copses, or ruins, so that, if a tank breaks down, we may know the best cover for the night: you cannot afford to leave a tank lying about in the open, however skilfully you may camouflage it.

I shall never forget those hot arduous days when we tramped in the moist heat over all the possible routes, plunging, after it had rained, through sticky mud often up to our knees, setting up little sign-posts wherever it was possible to make a mistake, and wondering whether the car would meet us at the other end....