It has often been remarked with surprise that cities and strongholds once densely inhabited and heavily built can disappear and leave no material trace to posterity.  That they do so disappear should give pause to those historians who are perpetually using the negative argument, and pretending that the lack of material evidence is sufficient to disturb a strong and early tradition.  Those who have watched the process by which abandoned buildings become a quarry will easily understand how all traces of habitation disappear.  Three-quarters of what was once Orford, much of what once was Worsted, has gone, and up and down the country-sides to-day one could witness, even in our strictly disciplined civilisation, the removal, by purchase or theft, of abandoned material.

The whole of Wallingford has suffered this fate—­the mound, presumably artificial, upon which the first keep stood, and which was, probably, a palisade mound of Anglo-Saxon times, remains, but there is upon it no remaining masonry.

Next down stream of the points with a strategic importance in English history comes Reading.  But the strategic importance of Reading was not produced by the town’s possessing a site of national moment:  it was produced only by local topography.  Reading was never (to use a modern term) a “nodal point” in the communications of England.

It may be generally laid down that mere strength of position is noted and greedily seized in barbaric times alone.  For mere strength of position is a mere refuge.  A strong position (I do not speak, of course, of tactical and temporary, but of permanent, positions), chosen only because it is strong, will save you during a critical short period from the attack of a fierce, unthoughtful, and easily wearied enemy—­such as are all barbarians; but it cannot of itself fall into a general scheme of defence, nor, simply because it is strong, intercept the advance of an adversary or support a line of opposition and resistance.  Position is always of advantage to a fortress, and, in all but highly civilised times, a necessity—­as we shall see when we come to discuss Windsor—­but it is not sufficient.  A fortress, when society is organised, and when the feud of one small tribe or family against another is not to be feared, derives its principal value from a command of established communications, and established aggregations of power—­especially of economic power.  Towns alone can feed and house armies; by roads and railways alone can armies proceed.

There are, indeed, examples of a chain of positions so striking that, from their strength alone, a strategic line imposes itself; but these are very rare.  Another, and much commoner, exception to the rule I have stated is the growth of what was once a barbaric stronghold, chosen merely for its position, into a larger centre of population, through which communications necessarily lead, and in which stores and other opportunities for armies can be provided.  Such places often preserve a continuity of strategic importance, from civilised, through barbaric, to civilised times again.  Laon is an excellent instance of this, and so is Constantine another, and so is Luxembourg a third—­indeed they are numerous.