Naturally, they fostered the notion of an intended repetition. That promoted their real design. The design itself, however, was based not merely on the war of 1870-1, but on the invasion of 1814, which led to the abdication of Napoleon, and the primary idea of it was to have _only one main line of advance_. The reason was that if an assailant takes two main lines of advance simultaneously and has to advance along the valleys of rivers converging to a point, as the Oise, the Marne, and the Seine converge towards Paris, his advance may be effectively disputed by a much smaller defending force than if he adopts only one line of advance, provided always, of course, that he can safeguard his flanks and his communications. Bear in mind the calculation that the main French army would never in any event be strong enough successfully to resist an invasion so planned. Bear in mind, too, that an echelon formation is not only supple and difficult to attack along its length on either side, but that it can be stretched out or closed up like a concertina. To maintain a formation of that kind with smaller bodies of troops is fairly easy. To maintain it with the enormous masses forming the German armies would be difficult. But the Germans were so confident of being able to compel the French to conform to all the German movements, to stand, that is to say, as the weaker side, always on the defensive, leaving the invaders a practically unchallenged initiative, that they believed they could co-ordinate all their movements with exactitude. This was taking a risk, but they took it. It is a mistake to suppose that they entered on the campaign with every movement mapped out from start to finish. No plan of any campaign was ever laid down on such lines, and none ever will be. The plan of a campaign has to be built on broad ideas. Those ideas, by taking all the essentials into consideration, the strategist seeks to convert into realised events. In this instance, there can be very little doubt that certain assumptions were treated as so probable as almost to be certainties. The first was that such forces as France could mobilise in the time would be mainly drafted to defend the fortified frontier. The next was that such forces as could be massed in time along the boundary of Belgium would be too weak seriously to impede the invasion. The third was that in any subsequent attempt to transfer forces from the fortified frontier to the Belgian boundary the French would be met and defeated by the advancing echelon of German masses. The fourth was that such an attempted transfer, followed by its defeat, would leave the fortified frontier so readily seizable, that German armies advancing swiftly into the valley of the Marne would fall upon these defeated French forces on the flank and rear. Besides, that attempted transfer would be the very thing that would promote the German design of envelopment.