In such circumstances a fresh and intact Cavalry, which has not allowed itself to be dispirited or demoralized, may render invaluable services to the beaten Infantry, and enable them, together with the columns and the whole supply organization of the Army, to effect an orderly withdrawal without being molested by the enemy, thus affording to its commander an opportunity of restoring the lost cohesion and tactical order. It will only be the rifle of the Cavalry which will gain for our hard-pressed comrades what is most needed--_i.e._, time. The idea is often put forward that Cavalry should not involve itself in an obstinate fight on foot, that it should act only by short surprise attacks, and that it possesses in its mobility the infallible means of circumventing points of resistance. This conception, which restricts the functions of Cavalry within their narrowest limits, seems to me entirely untenable. It is essentially indefensible when it is a matter of gaining time or of carrying a given position, such as a defensible post on the lines of communication, a convoy marching under escort, the destruction of a defended railway, or in any similar undertaking. It will be impossible to work round such positions--at least when engaged on the greater operations of War--both on account of the widths of the fronts occupied and the possibility of interfering with the sphere of operations of the neighbouring detachments. Often, indeed, the nature of the ground will preclude such attempts, and the experience of 1870-1871 should suffice to show how rarely such a circumvention can hope to succeed. But even where neither the strategic situation nor our immediate purpose compels us to fight, it is not always advisable or expedient to attempt to evade the opportunity. For every evasion leaves the front of our own Army clear, gives the enemy the very opportunity he is looking for to reconnoitre the position of our main bodies, uncovers our own communications (_i.e._, our own Cavalry trains and baggage), exposes our flank to the enemy, and thus offers him many chances of obtaining tactical results. Again, these attempts at circumvention lead easily to a loss of direction, which in certain cases might endanger the success of our whole operations; and, finally, the amount of time required by such outflanking attempts may not be justified by the possible results. In such cases it might on the whole be better to attack direct, for turning movements mean always a postponement of the tactical decision, and hence are always a disadvantage to the attacking side. The theory that Cavalry, thanks to its mobility, can always ride round and turn the positions it encounters breaks down in practice before the tactical and strategical demands upon the Arm, partly by reason of the local conditions, and partly because of the consideration which has to be given to time, to the endurance of the horses, and the position of the following columns.