The strength of the first deployment of the firing-line must depend on circumstances. In any case the whole breadth of that part of the enemy's front must be simultaneously engaged which is able to direct its fire against the attack, as otherwise the latter will be exposed to flanking or cross-fire.
If the ground allows of a covered approach to within effective range, a sufficiently thick firing-line should be established at once. On the other hand, where cover is scarce, it will sometimes be advisable, when sufficient time can be spared, to deploy at first only a loose, irregular firing-line, which will offer a difficult target to the enemy, or to remain at first under cover and gradually to strengthen the firing-line in the same way until it attains the strength necessary to commence the attack. In the further advance, supports must follow all the firing-lines, to make good losses, keep up the ammunition supply, and to put new moral strength into the fighting-line.
Open country in such cases demands increased depth in order to minimise loss, while close country requires that depth should be diminished, and it will be found advisable to act accordingly. Close formation can and must be maintained as long as the terrain and the hostile fire admit. It will of course be impossible to expose such formations within range of the enemy's fire. The advance will then consist of rushes, with pauses for breath, of distribution into small units, and the adoption of extended order. The nearer the decision of the fight, the closer must the supports follow, ready to co-operate.
In difficult country the order and cohesion of units must not be lost. It is especially important for the supports as well as the firing-line to make use of any available cover, so that units may be re-formed, ammunition supply regulated, and that the officers may regain their influence over the troops. This latter moral element deserves especial consideration.
When feeling its way forward the firing-line should avoid regular formation, and its rushes should not be made in too small bodies, as such a proceeding is apt to mask the fire of neighbouring detachments. It will likewise generally happen that some portions of the line are able to advance under cover of the ground more rapidly than others. These must then ask themselves the question whether their isolated advance might not lead to a reverse which would imperil the success of the attack. On the other hand, the unceasing pressing forward of all the various detachments is conditional to success. If localities exist in the foreground which might serve as supporting points for the further development of the attack, it should be considered whether they should not be occupied as rapidly as possible, and, where necessary, secured against counter-attack by being placed in a state of hasty defence. The advance of neighbouring detachments may then be facilitated from such points by energetic fire action. Such points at the same time secure the possession of the ground won and, under certain circumstances, protect those portions of the artillery and the machine-gun detachments which may accompany the attack, according as circumstances dictate. In any case such a measure (the co-operation of artillery and machine-guns) brings moral support to the attacking-line, and may at times prevent a defeat.
As soon as the firing-line has arrived within assaulting distance its fire must be increased to the utmost. All detachments in rear press forward regardless of loss as soon as they become aware that the first line is preparing for the assault. At the trumpet-call "Rapid advance"[38] the whole hurl themselves with the greatest determination and with loud cheers upon the enemy. Any reserves which may be still to the rear strain every nerve to reach the advanced firing-line. The assault should, if possible, be simultaneous. But the effort to secure this should never lead to detachments waiting for each other. Where a possibility is offered to single portions of the force to penetrate into the enemy's position they must advance independently, and all other detachments must conform.
It is important that the assault should not be begun too early, but that the firing-line should work its way forward to the closest range before rising for the final charge. The Infantry Drill Regulations lay down 150 paces as a guide for this distance in peace. That is a great deal too much. I do not think that the firing-line, especially in deep ground or uphill, can "rush" forward in full marching order 150 paces after having already carried out a lengthy attack. These instructions of the Infantry Drill should certainly be modified. Cavalry, however, should not fall into this error. A premature assault may imperil the success of the attack, because physical force may fail, and the enemy's fire is given a chance of regaining its full strength. The fire weapon should rather be used up to the last moment. The assault should take place only from the closest possible distance, and this will ensure success. For the rest, I may draw attention to the instructions laid down in the Cavalry Drill for the dispositions for the attack and for its conduct. On these points it coincides with my view, and it would be superfluous here to repeat what lies therein. If I have gone into the question of the conduct of the attack rather more closely, it is to emphasise certain points that are not prominent enough in the compressed instructions of the Regulations, and which do not generally receive sufficient consideration.
As to surprise fire action, also, the Regulations contain all the essentials (471-473).[39]
I may, however, touch here upon another matter which is connected with dismounted action, i.e. the question of armament. It has often been proposed, and from influential quarters, to replace the cavalry sword by some kind of a bayonet. If, it is said, with a certain appearance of justification, cavalry are to assault hostile positions on foot, they must have some arme blanche for the hand-to-hand fight, and this can only be a bayonet. The war in Manchuria, where such mêlées repeatedly took place, is cited as a proof of the necessity of such an armament. I cannot ally myself with such proposals. As to the repeated hand-to-hand fights in the Russo-Japanese war, these took place principally because the Russians found it impossible to evacuate their entrenchments in time, and that they took such full advantage of cover that they were in some measure surprised by the attacker. These examples can certainly not be adapted therefore to the dismounted action of our cavalry. With them it will never be a question of prepared positions—which cavalry will, as a rule, neither attack nor defend—but of actions resulting from a battle of encounter. In such cases, however, as the experience of military history teaches us, a hand-to-hand fight is quite exceptional. Even the struggle for localities is fought out almost entirely with the firearm. If the defence should consist of dismounted cavalry and cyclists, it cannot as a rule be assumed that such troops will allow the attack to come to close quarters. They will more generally, as soon as the decision of the fire fight has become clear, endeavour to reach their horses or cycles and to escape the fire of the pursuit. As a matter of fact, therefore, dismounted cavalry would really only use the bayonet on foot in quite exceptional cases, and it would only be justifiable to introduce it if there were no attendant disadvantages. Such, however, is not the case. I would regard the abolition of the present sword as a great danger, calculated to seriously injure the morale of the cavalry.