The second factor in the establishment of such an obstacle is the type of soil over which the water works. It is evident that a tidal river and estuary in which sand and gravel or even chalk form the riparian soils will less produce marsh than clays will. A river which washes the silt of clays up and down with its tides will be defended by worse belts of bad land than one which runs through the other types of rock. Now in this respect also the Thames has a bad pre-eminence over its rivals. Chalk only comes near the stream once or twice, and for a very short distance, in its lower course, and though gravel and sand, as we shall see, approach the banks in more than one place (and ultimately determine the site of London), this kind of soil is nowhere that which the mass of the stream churns up or carries down in its course. The Thames deals for many miles of its upper tidal waters with clays, bringing them down towards the mouth, and has settled them for centuries upon either side of its channel in the shape of deep alluvial marshes. But there is more than this. The third factor in the problem depends upon the contours of the land upon either side of the river. A river with steep banks, fairly narrow, bounded by hills, even though it be a highly varying tidal stream, and even though it scour through a great part of its tidal course a soil of clays, cannot form wide belts of marsh upon either bank. It so happens that the Lower Thames--until the site of London is reached--nowhere enjoys even a short stretch of steep-on shore upon both banks at once. The few spots where higher land comes down to the water’s edge upon the southern or right bank are faced in every case by great level stretches opposite which, until modern works were undertaken, were regularly flooded with every return of the tide and were impassable; while the lower and smaller patches of land on the north bank (as at Purfleet or Grays) have marsh opposite them also to the south. In this respect the Scheldt appears under primitive conditions even worse off than the Thames; but the Seine continually enjoys steep banks after the first twenty miles or so up from its estuary, and in at least half a dozen places from Caudebec (which roughly corresponds to Gravesend) and Rouen, the first bridge (which we have seen to correspond to London), there are opportunities for crossing the tidal Seine, even under primitive conditions, with no considerable obstacle of marsh upon either bank.