The secondary base in time of war is conditioned as follows.  If the enemy’s sphere of activity is distant from the base which you have equipped with store-houses and fortifications, the place cannot be of any use to you.  It can, and probably will, be a cause of additional anxiety to you, because the communications of its garrison must still be kept open.  If it is used, freedom of movement for the ships must be given up, because they cannot go so far from it as to be obliged to consume a considerable fraction of their coal in reaching it and returning to their station.  The line along which your colliers proceed to it must be effectively guarded.

Contrast this with the system of direct supply to the ships-of-war.  You choose for your flying base a position which will be as near to the enemy’s sphere of action as you choose to make it.  You can change its position in accordance with circumstances.  If you cease to use the position first chosen you need trouble yourself no more about its special communications.  You leave nothing at it which will make it worth the enemy’s while to try a dash at it.  The power of changing the flying base from one place to another gives almost perfect freedom of movement to the fighting ships.  Moreover, the defence of the line communicating with the position selected is not more difficult than that of the line to a fixed base.

The defence of a line of communication ought to be arranged on the same plan as that adopted for the defence of a trade route, viz. making unceasing efforts to attack the intending assailant.  Within the last few years a good deal has been written about the employment of cruisers.  The favourite idea seems to be that peace-time preparation for the cruiser operations of war ought to take the form of scouting and attendance on fleets.  The history of naval warfare does not corroborate this view.  We need not forget Nelson’s complaint of paucity of frigates:  but had the number attached to his fleet been doubled, the general disposition of vessels of the class then in commission would have been virtually unaltered.  At the beginning of 1805, the year of Trafalgar, we had—­besides other classes—­232 frigates and sloops in commission; at the beginning of 1806 we had 264.  It is doubtful if forty of these were attached to fleets.

It is sometimes contended that supply-carriers ought to be vessels of great speed, apparently in order that they may always keep up with the fighting ships when at sea.  This, perhaps, is due to a mistaken application of the conditions of a land force on the march to those of a fleet or squadron making a voyage.  In practice a land army cannot separate itself—­except for a very short time—­from its supplies.  Its movements depend on those of its supply-train.  The corresponding ‘supply-train’ of a fleet or squadron is in the holds and bunkers of its ships.  As long as these are fairly well furnished, the ships might be hampered, and could not be assisted, by the presence of the carriers.  All that is necessary is that these carriers should be at the right place at the right time, which is merely another way of saying that proper provision should be made for ’the stream of supplies and reinforcements which in terms of modern war is called communications’—­the phrase being Mahan’s.