In the battle of all arms, cavalry must be handled according to principles which are quite different and almost diametrically opposed to those which characterise its independent action as army cavalry. For, in the latter case, not only is the defeat of the enemy kept in view, but another definite object has also to be pursued. This object can only be attained if successful in the fight, while an unsuccessful battle will paralyse the activity of the cavalry, and may cost the army the loss of its organs of reconnaissance. However daring its conduct then, it should never be engaged in hopeless enterprises, and should only undertake a fight where success can be reckoned upon with a certain measure of probability.
If, therefore, the strength and intentions of the enemy are not fully known, it will be better, as we have seen, to guard against engaging the whole force in such an uncertain enterprise. Efforts should rather be made, as I have endeavoured to show, to clear up the situation by a careful feeling of the enemy and a gradual engagement of force. Once possessed of this knowledge of the situation, it will be possible either to seek a decision or to break off the fight in time to avoid the risk of incurring too considerable a loss.
Quite different is the case in the main battle. Here the objective is contained in the battle itself. It is the destruction of the enemy that is sought. It is not expected that each single detachment engaged should be victorious, but that the net result of the battle should be a victory. The task of the various detachments is only to engage and to destroy so much of the enemy's force as lies within their power. This naturally holds good for the cavalry. It is not now demanded that each single action of the cavalry should of itself be successful, but that the general engagement of the cavalry should have the greatest possible effect. A considerable result may often be obtained by the attacking cavalry drawing the enemy's fire upon itself for a time, and thus affording the infantry the possibility of gaining ground to the front, or of re-forming and receiving reinforcements.
To break off the main battle is generally quite out of the question. The very fact that the battle has been begun betokens the intention of carrying it through to a final decision, even where the enemy has shown himself to be in superior force. The various troops which advance to the conflict need not therefore reflect whether _they_ have any special prospect of success, but must strive for this success with all their power. This means for the cavalry, in by far the greater number of cases, always at least where a charge is in prospect, _the simultaneous engaging of its whole fighting strength_, naturally in that tactical formation which the conditions of weapons demand. If in its _independent_ operations cavalry must be dealt with as a strategical body, and thus employed in the fight, it is _in the main battle_ a purely tactical body, which must be engaged _en masse_, and not in detail. This contrast appears, at least to me, to be an obvious one. There is another that is equally clear.