But obstacles do intervene, and these obstacles were of the most serious kind for men who had not yet passed the early stages of civilisation. A broad river with a swift tidal current, flanked here and there (as tidal rivers always are before their embankment) by marshes; a valley floor of clay, the crossing of which must prove far more lengthy than that of any they had hitherto encountered, made the negotiation of this gap a difficult matter. Moreover, the direct line would have led them by the marshiest way of all: the fields of Snodland brook. Oddly enough the difficulty of rediscovering the original track by which the road forded the Medway, does not lie in the paucity of evidence, but rather in the confusion arising from its nature and amount. So great is this confusion that some authorities have been content to accept alternative routes at this point. Savage trails, however, never present alternatives so widely separate, and least of all will they present any alternative, even one neighbouring the main road, where a formidable obstacle has to be overcome: to do so would be to forfeit the whole value which a primitive road possesses as a guide (for this value depends upon custom and memory), and when a tidal river had to be traversed, a further and very cogent reason for a single track was to be found in the labour which its construction upon a marshy soil involved. If some one place of crossing had held a monopoly or even a pre-eminence within the limits of recorded history, the evidence afforded by it would be of the utmost value. But an indication of this simplicity is lacking. It is certain that within historic times and for many centuries continuously, the valley and the river were passed at four places, each of which now may lay a claim to be the original passage. The modern names of these places are, in their order from the sea, Cuxton, Lower Halling, Snodland, and Aylesford. Before proceeding I must repeat what was said above, that two tracks of great antiquity continue the Old Road northward on each side of the Medway far beyond any point where it would have crossed; these tracks (I have called them elsewhere 'feeders') are not only clearly defined, but have each received the traditional name of the Pilgrim's Way, and their presence adds a considerable complexity to the search for the original passage. So much of the elements of the problem being laid down, let us now recapitulate certain features which we have discovered to be true of the road in the earlier part of its course, where it had to cross a river, and certain other features which one knows to be common to other British track-ways over valleys broader than those of the Mole or the Wey. To these features we may add a few others, which are conjecturally those that such a road would possess although we might have no direct evidence of them. A list of these features will run very much as follows:--