We should come to the same conclusion, if we realized that no matter what their comparative values might be, a base at one place would not meet our necessities, and a base at the other place would.  If a base at Guantanamo would not meet our necessities in case of an invasion of the Caribbean by a naval fleet superior to ours, then it seems idle to discuss the value of Guantanamo relative to some other place, no matter how good the position of Guantanamo may be, and no matter how nearly it may approximate to adequacy.  There is no real usefulness in having a naval base anywhere, unless that naval base can accomplish the purpose for which it is desired.  A naval base is desired for purposes of war, and for no other purpose whatever; and to decide on a position for a base without keeping this fact clearly in view, is to act on an underestimate of the situation, the folly of which has been pointed out in previous pages.

We may conclude, then, that in deciding on the place for a distant permanent naval base, on which the operations of a whole fleet are to base for war, we should select the best site available, even if military strength and resources may have to be added to it artificially—­unless in the case of any site considered the difficulties of adding them are insuperable.

The last sentence may seem like shirking the whole question, because it does not state what “insuperable” means; so it may be well to add that in modern days few engineering difficulties are insuperable, as the existence of the fortress at Heligoland shows.  If the submarine and the mine did not exist, the difficulties would be greater than they actually are; because guns alone, no matter how carefully mounted and protected, could hardly be expected to keep off indefinitely the attack of a heavy fleet, or even to save from injury the fighting and auxiliary vessels anchored in its waters.  But the submarine and mine combine to keep fighting ships at distances greater than those over which ship’s guns can fire, and reduce the amount of fortification required on shore.

One of the principal sources of expense in establishing bases at some points would be that of dredging out harbors sufficiently extensive, while harbors sufficiently extensive are provided already by nature in such localities as Samana.  But, as pointed out before, harbors on large islands can be taken from the land side, as was Port Arthur; and adequate protection from land attack is, in many cases, almost impossible if the enemy has command of the sea, as a superior hostile fleet would have in the Caribbean; while the hills and waters of Culebra and Vieques Sound could long defy not only actual invasion, but any fleet attack.