"In the long run," said Napoleon, "no position whatever can be defended if it does not threaten the enemy."[15] Consequently, the force that for any, or several, of the reasons above given cannot safely keep the field must establish itself solidly in some place where, for whatsoever advantages, it is as far as possible itself secure; but whence at the same time--and this is the more important of the two considerations--it most effectually menaces the enemy.

This it does by applying again, but in another manner, the flanking, or turning, idea--by placing itself across or to one side of the line of communications upon which the enemy will depend, if he ventures to advance in the direction which the defendant has not felt himself strong enough otherwise to contest.

Of {p.187} such offensive-defensive positions there are many historical examples.
Among these the most recently conspicuous was the occupation of Plevna by the Turks in 1877, and the long consequent arrest of the Russian progress; but Mantua, in 1796, in like manner and for the same reasons, effectually stopped Bonaparte for eight months, and Genoa, in 1800, so long delayed the Austrians as to reverse the issues of the campaign signalised by the name of Marengo.
[Footnote 15: I should greatly like here to take up my parable against those who base their calculations for the numbers and kinds of naval vessels upon the idea of "a navy for defence only"; but space and relevancy both forbid.] From the simply defensive point of view, a line of works arranged consecutively around such a stationary centre has no flanks to be turned, but resembles a circle or other continuous curve which returns into itself.

Like a straight line, such a curve may be broken by superior force; but until that is done the weakness of flanks does not exist.

Moreover, succour can be more quickly sent from a centre to every threatened point of the circumference than from the middle of a line of equal length to its extremes.