A body of cavalry, in the approach formation, is advancing against the enemy, with blind ground on a flank, which would allow of the enemy's covered approach, and which perhaps it has been impossible to reconnoitre. Attack or fire surprise is feared from this quarter. How will the cavalry protect itself? The modern tactician would in most cases reply: "By an échelon to the rear." I do not think this would be suitable. How is such a formation to give safety from fire surprise, and to locate the enemy's advance and arrest it until the main body can take counter-measures? The forward échelon can here alone avail. It comes to close quarters with the enemy, attacks him before he can reach the flank of the main body, and thus gains time for defensive measures or retirement. It is quite obvious that the cases for employment of the forward échelon do not allow of being formulated. I think, however, that we should make much more use of this formation than is at present the fashion. Properly applied, such methods will ensure to us considerable superiority over our opponents. If we turn from this narrower tactical point of view to the formations on a large scale where échelon is to be found--namely, the divisions--here too the examination leads to no more favourable conclusions. I ask myself, when and under what circumstances will such a formation be advisable? During the approach to the battle of encounter it is, as I have endeavoured to prove, quite superfluous, and may even operate to our disadvantage. In this case, when total uncertainty reigns as to whether the combat will be carried out mounted or dismounted, or both, there can be no question of any stereotyped tactical formation, either of units as a whole or of smaller bodies within them. Here, as we have seen, the battle will generally develop gradually, and the fighting-line be fed from depth, until the necessary information as to the enemy has been gained and the decisive attack can be embarked upon. Under such circumstances the brigades advance according to the tasks allotted them, and make their dispositions as circumstances dictate. The depth that will be necessary can obviously not be laid down, and can be attained by échelon neither in the division nor in its subdivisions. If the division should advance under such circumstances in close formation écheloned within itself, the unnecessary danger would be run of offering an ideal target to the enemy's artillery (which must always be taken into consideration), and at the same time of hampering movement where circumstances demand the greatest freedom in all directions. If, on the other hand, the enemy's dispositions are known before a collision occurs, the Regulations themselves (426, see p. 172) allow that échelon formation is superfluous, and that the advance may be made in the deployed formation desired, brigades being on the same frontage, if the country and the character of the adversary offer the probability of a charge.