Save perhaps by a bridge of boats, a permanent crossing could not have been attempted at Erith or at Greenhithe, though the narrowing of the stream at Gravesend might well have allowed a more stable structure to be established. At any rate, a crossing even so broad as that opposite Erith has nowhere in Europe interfered with the passage of commerce, or of arms where both sides of a great stream lent themselves to such a passage. But it is here that each one of the points I have mentioned is at fault. Opposite Gravesend as opposite Greenhithe, opposite Erith as opposite Woolwich, Greenwich, and Deptford, there lies upon the northern bank a belt of marsh which forbids traffic. Tilbury Fort, opposite Gravesend, stands upon a tiny circle of harder land, but all around it are the Chadwell and the Tilbury marshes. Greenhithe has right against it the projecting expanse of West Thurrock Marsh, Erith the whole breadth of Wellington and Rainham marshes; and, as one approaches London, and the river narrows, matters seem only to get worse. Woolwich faces the expanse of originally flooded soil between the Lea and the Roding, with the most of which even the economic forces making for the expansion of London have been able to do nothing, and of which so unpleasant a relic of its original condition is left in Plaistow Marshes to-day; while opposite Greenwich and Deptford lay that perfectly impossible morass, which, though turned into water meadows by a river wall many centuries ago, is still perhaps the worst building ground within the London area. We call it the Isle of Dogs. The reader must not imagine this lack of any two opposing hards upon London River below the City to be due to some coincidence. It has a fairly obvious geographical cause. Those points where the gravels and the chalk were touched by the scouring of the stream were naturally the outer cusp of its curves. The river having first determined a bend to the south or to the north, would eat away more and more on the outer edge of those bends (which the stream always follows both in flood and in ebb), and scour away the bank until it struck harder soils and was there checked. Deptford and Greenwich lie on the outer edge of the first great southern bend; Woolwich on that of the second; Erith on that of the third; Greenhithe the fourth; Gravesend the fifth; while both Grays and Purfleet represent similar checks to the bends towards the north. Now it is evident that the same process which makes a river extend its curves outward more and more by the scouring the stream along the exterior edge leaves on the inside of the curve an increasing tongue or wedge of alluvial deposit. What we have, therefore, on the Lower Thames, the continual opposition of marsh to the rare hards, is only what we should expect from the geological history of the river, and the crossing place which was at last found is much more of a coincidence and accident than the absence of a crossing place below.