This is no idle question, and these are no hair-splitting criticisms, because the rules are held to differ in important respects in these various types of combats. In the battle of encounter there are some exceedingly dim indications of an opening for the steel, but an attack upon a "locality" "_can obviously only be carried out dismounted_" (p. 165). Pass by the old fallacious antagonism between mounted action and rifle action, and regard the essence of this proposition. Once again you have the refutation of the steel theory. The sentence means "fire governs combat." He who fires compels his enemy to accept combat on terms of fire. But "Where am I?" the harassed student may exclaim. "What of these steel-charges against extended Infantry (and, by inference, against dismounted Cavalry), whose fire enforced extension in the attacking Cavalry?" Well, let him read on. There is hope yet. For immediately after saying that an attack upon an enemy who takes up a defensive attitude can obviously only be carried out dismounted, he adds the sinister words: "It must be a matter, therefore, for serious consideration, whether such an operation shall be undertaken or not." The truth is that he has suddenly remembered those tiresome led horses. "There must be considerable numerical superiority to insure success." There must be a dismounted reserve for fire purposes, and a mounted reserve to secure the safety of the led horses, and "for reconnaissance and for operating against the enemy's flank and rear"; and then follows an acrimonious wrangle with the Regulations on the question of making one reserve, and that mounted, perform incompatible and contradictory functions. But, as usual, our sympathies are with the Regulations.