his army could never even have put forth, unless he had inflicted upon our covering fleet such a defeat as would have given him command of the sea, and with absolute control of the sea the passage of an army presents no difficulties. Of the working of these principles under modern conditions we have no example. The acquisition of free movement must necessarily modify their application, and since the advent of steam there have been only two invasions over uncommanded seas--that of the Crimea in 1854, and that of Manchuria in 1904--and neither of these cases is in point, for in neither was there any attempt at naval defence. Still there seems no reason to believe that such defence applied in the old manner would be less effective than formerly. The flotilla was its basis, and since the introduction of the torpedo the power of the flotilla has greatly increased. Its real and moral effect against transports must certainly be greater than ever, and the power of squadrons to break a flotilla blockade is more restricted. Mines, again, tell almost entirely in favour of defence, so much so indeed as to render a rapid _coup de main_ against any important port almost an impossibility. In the absence of all experience it is to such theoretical considerations we must turn for light. Theoretically stated, the success of our old system of defence depended on four relations. Firstly, there is the relation between the rapidity with which an invasion force could be mobilised and embarked, and the rapidity with which restlessness in foreign ports and _places d'armes_ could be reported; that is to say, the chance of surprise and evasion are as the speed of preparation to the speed of intelligence. Secondly, there is the relation of the speed of convoys to the speed of cruisers and flotilla; that is to say, our ability to get contact with a convoy after it has put to sea and before the expedition can be disembarked is as the speed of our cruisers and flotilla to the speed of the convoy. Thirdly, there is the relation between the destructive power of modern cruisers and flotillas against a convoy unescorted or weakly escorted and the corresponding power in sailing days. Fourthly, there is the relation between the speed of convoys and the speed of battle-squadrons, which is of importance where the enemy's transports are likely to be strongly escorted. On this relation depends the facility with which the battle-squadron covering our mobile defence can secure an interior position from which it may strike either the enemy's battle-squadron if it moves or his convoy before it can complete its passage and effect the landing.