If the fighting strength in each detachment depends on its composition and training, there are other elements besides the tactical value of the troops which determine the effectiveness of their combined efforts in action; these are first the leadership, which, however, depends on conditions which are beyond calculation, and secondly the numerical proportion of the arms to each other.  Disregarding provisionally the cavalry, who play a special role in battle, we must define the proportion which artillery must bear to infantry.

With regard to machine guns, the idea that they can to some extent replace infantry is quite erroneous.  Machine guns are primarily weapons of defence.  In attack they can only be employed under very favourable conditions, and then strengthen only one factor of a successful attack—­the fire-strength—­while they may sometimes hinder that impetuous forward rush which is the soul of every attack.  Hence, this auxiliary weapon should be given to the infantry in limited numbers, and employed mainly on the defensive fronts, and should be often massed into large units.  Machine-gun detachments should not overburden the marching columns.

The relation of infantry to artillery is of more importance.

Infantry is the decisive arm.  Other arms are exclusively there to smooth their road to victory, and support their action directly or indirectly.  This relation must not be merely theoretical; the needs of the infantry must ultimately determine the importance of all other fighting instruments in the whole army.

If we make this idea the basis of our argument, the following is the result.  Infantry has gained enormously in defensive power owing to modern weapons.  The attack requires, therefore, a far greater superiority than ever before.  In addition to this, the breadth of front in action has greatly increased in consequence of the former close tactical formations having been broken up through the increase of fire.  This refers only to the separate detachment, and does not justify the conclusion that in the future fewer troops will cover the same spaces as before.  This assumption applies at the most to defence, and then only in a limited sense.  In attack the opposite will probably be the case.  The troops must therefore be placed more deeply en echelon than in the last wars.  Now, the average breadth of the front in attack must regulate the allotment of artillery to infantry.  No definite proportion can be settled; but if the theoretical calculation be compared with the experiences of the last wars, conclusions may be obtained which will most probably prove appropriate.  No more than this can be expected in the domain of military science.

If we agree to the above-mentioned proportion of breadth and depth in an infantry attack, we shall be driven to insist on a reduction of artillery as compared with the past; but should we think that modern artillery helps the attack, especially by indirect fire, we must advocate, from the standpoint of offensive warfare, an increase of the artillery.  Actual war experiences alone can find the true middle path between these two extremes.