2. The capture of an important magazine under the same condition.
3. The capture of a fortress not covered. No doubt a siege is more or less a great operation, often requiring great labour; but it is an undertaking which does not contain the elements of a catastrophe. If it comes to the worst, the siege can be raised without thereby suffering a great positive loss.
4. Lastly, a successful combat of some importance, but in which there is not much risked, and consequently not much to be gained; a combat which takes place not as the cardinal knot of a whole strategic bond, but on its own account for the sake of trophies or honour of the troops. For such an object, of course, a combat is not fought at any price; we either wait for the chance of a favourable opportunity, or seek to bring one about by skill.
These four objects of attack give rise to the following efforts on the part of the defence:—
1. To cover the fortresses by keeping them behind us.
2. To cover the country by extending the troops over it.
3. Where the extension is not sufficient, to throw the army rapidly in front of the enemy by a flank march.
4. To guard against disadvantageous combats.
It is clear that the object of the first three measures is to force on the enemy the initiative, and to derive the utmost advantage from the state of expectation, and this object is so deeply rooted in the nature of the thing that it would be great folly to despise it prima facie. It must necessarily occupy a higher place the less a decision is expected, and it is the ruling principle in all such campaigns, even although, apparently, a considerable degree of activity may be manifested in small actions of an indecisive character.
Hannibal as well as Fabius, and both Frederick the Great and Daun, have done homage to this principle whenever they did not either seek for or expect a decision. The fourth effort serves as a corrective to the three others, it is their conditio sine quâ non.