This method need not be as completely dogmatic as the Roman. Our battle is too varying an affair. But some clearly defined rules, established by experience, would prevent the gross errors of inefficients. (Such as causing skirmishers to fall back when the formed rank fires, and consequently allowing them to carry with them in their retreat, the rank itself.) They would be useful aids to men of coolness and decision.

The laying down of such tactics would answer the many who hold that everything is improvised on the battle field and who find no better improvisation than to leave the soldier to himself. (See above.)

We should try to exercise some control over our soldiers, who advance by flight (note the Vendeans) or escape by advancing, as you like. But if something unexpected surprises them, they flee as precipitately.

Invention is less needed than verification, demonstration and organization of proper methods. To verify; observe better. To demonstrate; try out and describe better. To organize, distribute better, bearing in mind that cohesion means discipline. I do not know who put things that way; but it is truer than ever in this day of invention.

With us very few reason or understand reason, very few are cool. Their effect is negligible in the disorder of the mass; it is lost in numbers. It follows that we above all need a method of combat, sanely thought out in advance. It must be based on the fact that we are not passively obedient instruments, but very nervous and restless people, who wish to finish things quickly and to know in advance where we are going. It must be based on the fact that we are very proud people, but people who would all skulk if we were not seen, and who consequently must always be seen, and act in the presence of our comrades and of the officers who supervise us. From this comes the necessity for organizing the infantry company solidly. It is the infantryman on whom the battle has the most violent effect, for he is always most exposed; it is he therefore who must be the most solidly supported. Unity must be secured by a mutual acquaintanceship of long standing between all elements.

If you only use combat methods that require leaders without fear, of high intelligence, full of good sense, of esprit, you will always make mistakes. Bugeaud's method was the best for him. But it is evident, in his fight at the Hospital bridge that his battalion commanders were useless. If he had not been there, all would have been lost. He alone, omnipresent, was capable of resolute blows that the others could not execute. His system can be summed up in two phrases; always attack even when on the defensive; fire and take cover only when not attacked. His method was rational, considering his mentality and the existing conditions, but in carrying it into execution he judged his officers and soldiers by himself and was deceived. No dogmatic principles can be drawn from his method, nor from any other. Man is always man. He does not always possess ability and resolution. The commander must make his choice of methods, depending on his troops and on himself.

The essential of tactics is: the science of making men fight with their maximum energy. This alone can give an organization with which to fight fear. This has always been true.

We must start here and figure mathematically. Mathematics is the dominant science in war, just as battle is its only purpose. Pride generally causes refusal to acknowledge the truth that fear of being vanquished is basic in war. In the mass, pride, vanity, is responsible for this dissimulation. With the tiny number of absolutely fearless men, what is responsible is their ignorance of a thing they do not feel. There is however, no real basis but this, and all real tactics are based on it. Discipline is a part of tactics, is absolutely at the base of tactics, as the Romans showed. They excelled the Gauls in intelligence, but not in bravery.

To start with: take battalions of four companies, four platoons each, in line or in column. The order of battle may be: two platoons deployed as skirmishers, two companies in reserve, under command of the battalion commander. In obtaining a decision destructive action will come from skirmishers. This action should be directed by battalion commanders, but such direction is not customary. No effect will be secured from skirmishers at six hundred paces. They will never, never, never, be nicely aligned in front of their battalions, calm and collected, after an advance. They will not, even at maneuvers. The battalion commander ought to be advanced enough to direct his skirmishers. The whole battalion, one-half engaged, one-half ready for any effort, ought to remain under his command, under his personal direction as far as possible. In the advance the officers, the soldiers, are content if they are merely directed; but, when the battle becomes hot, they must see their commander, know him to be near. It does not matter even if he is without initiative, incapable of giving an order. His presence creates a belief that direction exists, that orders exist, and that is enough.

When the skirmishers meet with resistance, they fall back to the ranks. It is the rôle of reserves to support and reinforce the line, and above all, by a swift charge to cut the enemy's line. This then falls back and the skirmishers go forward again, if the advance is resumed. The second line should be in the formation, battalions in line or in column, that hides it best. Cover the infantry troops before their entry into action; cover them as much as possible and by any means; take advantage of the terrain; make them lie down. This is the English method in defense of heights, instanced in Spain and at Waterloo. Only one bugle to each battalion should sound calls. What else is there to be provided for?