across what then was by far the most thickly populated and the most important part of the island.

A third section in all such rivers (and, from what we have said above, a short and insignificant one in the case of the Thames) may be called the head-waters of the river: where the stream is so shallow or so uncertain as to be no longer navigable. In the case of the Thames these head-waters cover no more than ten to fifteen miles of country. With the exception of rivers that run through mountain districts this section of a river's course is nearly always small in proportion to the rest; but the Thames, just as it has the longest proportion of navigable water, has also by far the shortest proportion of useless head-water of all the shorter European rivers.

There is a further discussion as to what is the true source of the Thames, and which streams may properly be regarded as its head-waters: the Churn, especially since the digging of the canal, having a larger flow than the stream from Thames head; but whichever is chosen, the non-navigable portion starts at the same point, and is the third of the divisions into which the valley ranges itself when it is considered in its length, as a highway from the west to the east of England. The two limits, then, are at London Bridge and at Cricklade, or rather at some point between Lechlade and Cricklade, and nearer to the latter.

But a river has a second topographical and historic function. It cannot only be considered longitudinally as a highway, it can also be considered in relation to transverse forces and regarded as an obstacle, a defence, and a boundary.

This function has, of course, been of the highest importance in the history of all great rivers, not perhaps so much so in the case of the Thames as in the case of swifter or deeper streams, but, still, more than has been the case with so considerable and so rapid a river as the Po in Lombardy or the uncertain but dangerous Loire in its passage through the centre of France. For the Thames Valley was that which divided the vague Mercian land from which we get our weights, our measures, and the worst of our national accent, and cut it off from that belt of the south country which was the head and the heart of England until the last industrial revolution of our history.

The Thames also has entered to a large, though hardly to a determining, extent into the military history of the country; to an extent which is greater in earlier than in later times, because with every new bridge the military obstacle afforded by the stream diminished. And finally, the Thames, regarded as an obstacle, was the cause that London Bridge concentrated upon itself so much of the life of the nation, and that the town which that bridge served, always the largest commercial city, became at last the capital of the island.

We have already said that the establishment of the site of London Bridge was a capital point in the history of the river and the principal line of division in its course. What were the topographical conditions which caused the river to be crossed at this point rather than at another?

It is always of the greatest moment to men to find some crossing for a great river as low down as may be towards the mouth. For the higher the bridge the longer the detour between, at the least, two provinces of the country which the river traverses. It is especially important to find such a crossing as low down as possible when the river is tidal and when it is flanked upon either side by great flooded marshes, as was and is the Thames. For under such conditions it is difficult, especially in primitive times, to cross habitually from one side to the other in boats.

Now it is a universal rule of early topography, and one which can be proved upon twenty of the old trackways of England, that the wild path which the earliest men used, when it approaches a river, seeks out a spur of higher and drier land, and if possible one directly facing another similar spur upon the far side of the water. It is a feature which the present writer continually observed in the exploration of the old British trackway between Winchester and Canterbury; it is similarly observable in the presumably British track between Chester and Manchester; and it is the feature which determined the site of London Bridge.

From the sea for sixty miles is a succession of what was once entirely, and is now still in great part, marshy land; or at least if there are no marshes upon one bank there will be marshes upon the other. In the rare places down stream where there is a fairly rapid rise upon either side of the river the stream is far too wide for bridging; and these marshes were to be found right up the valley until one struck the gravel at Chelsea: even here there were bad marshes on the farther shore.

There is in the whole or the upper stretch of the tidal water but one place where a bluff of high and dry land faces, not indeed land equally dry immediately upon the farther bank, but at least a spur of dry land which approaches fairly near to the main stream. If the modern contour lines be taken and laid out upon a map of London this spur will be found to project from Southwark northward directly towards the river, and immediately opposite it is the dry hill, surrounded upon three sides by river or by marsh, upon which grew up the settlement of London. Here, then, the first crossing of the Thames was certain to be made.

It is not known whether a permanent bridge existed before the Roman Conquest. It may be urged in favour of the negative argument that Cæsar had no knowledge of such a bridge, or at least did not march towards it, but crossed the river with difficulty in the higher reaches by a ford. And it may also be urged that a bridge across the Rhine was equally unknown in that time. But, the bridge once established, it could not fail to become the main point of convergence for the commerce of Southern England, and indeed for much of that which proceeded from the North upon its way to the Continent. Such an obstacle would oppose itself to every invasion, and did, in fact, oppose itself to more than one historical invasion from the North Sea. It would further prevent sea-going vessels whose masts were securely stepped and could not lower from proceeding farther up stream, and would thereupon become the boundary of the seaport of the Thames. Such a bridge would, again, concentrate upon itself the traffic of all that important and formerly wealthy part of the island which bulges out to the east between the estuary of the Thames and the Wash, and which must necessarily have desired communication both with the still wealthier southern portion and with the Continent. But, more important than this, London Bridge also concentrated upon itself all the up-country traffic in men and in goods which came in by the natural gate of the country at the Straits of Dover, except that small portion which happened to be proceeding to the south-west of England: and this exception to the early commerce of England was the smaller from the comparative ease with which the Channel could be crossed between Brittany and Cornwall.

Finally, the Bridge, as it formed the limit for sea-going vessels, formed also if not the limit at least a convenient terminus for craft coming from inland down the stream. It would form the place of transhipment between the sea-going and the inland trade.

Everything then conspired to make this first crossing of the Thames the chief commercial point in Britain; and, since we are considering in particular the history of the river, it must be noted that these conditions also made of London Bridge what we have remarked it to be, the chief division in the whole course of the stream. This character it still maintains, and the life of the river from the bridge to the Nore is a totally different thing, with a different literature and a