one thinks how far back Eynsham stands from the river it is somewhat remarkable that it should have claimed the right to get at the stream. Below Oxford there is another most interesting instance of the same thing in the case of Littlemore. Littlemore stands on high and dry land up above the river somewhat set back from it. Sandford evidently interfered with its access to the water, and Littlemore has therefore claimed an obviously artificial extension for all the world like a great foot added on to the bulk of the parish. This "foot" includes Kennington Island, and runs up the meadows to the foot of that eyot.

The long and narrow parishes in the reaches below Benson, Nuneham Morren, Mongewell, and Ipsden and South Stoke are not, however, examples of this tendency.

They owe their construction to the same causes as have produced the similar long parishes of the Surrey and the Sussex Weald. The life of the parish was in each case right on the river or very close to it, and the extension is not the attempt of the parish to reach the river, but the claim of the parish upon the hunting lands which lay up behind it upon the Chiltern Hills. The truth of this will be apparent to anyone who notes upon the map the way in which parishes are thus lengthened, not only on the western side of the hills, but also upon the farther eastern side, where there was no connection with the river.

There are many other proofs remaining of the chief function which the Thames fulfilled in the early part of our history as a means of communication.

We shall see later in these pages how united all that line of the stream has been; how the great monasteries founded upon the Thames were supported by possessions stretched all along the valleys; how much of it, and what important parts, were held by the Crown; and how strong was the architectural influence of towns upon one another up and down the water, as also how the place names upon the banks are everywhere drawn from the river; but before dealing with these it is best to establish the main portions into which the Thames falls and to see what would naturally be their limits.

It may be said, generally, that every river which is tidal, and whose stream is so slow as to be easily navigable in either direction, divides itself naturally, when one is regarding it as a means of communication, into three main divisions.

There will first of all be the tidal portion which the tide usually scours into an estuary. As a general rule, this portion is not considerably inhabited in the early periods of history, for it is not until a large international commerce arises that vessels have much occasion to stop as they pass up and down the maritime part of the stream; and even so, settlements upon its banks must come comparatively late in the development of the history of the river, because a landing upon such flooded banks is not easily to be effected.

This is true of the Dutch marshes at the mouths of the Rhine, whose civilisation (one exclusively due to the energy of man) came centuries after the establishment of the great Roman towns of the Rhine; it is true of the estuary of the Seine, whose principal harbour of Havre is almost modern, and whose difficulties are still formidable for ocean-going craft; and it is true of the Thames.

The estuary of the Thames plays little or no part in the very early history of England. Invaders, when they landed, landed on the sea-coast at the very mouth, or appear to have sailed right up into the heart of the country.

It is, nevertheless, true that the last few miles of tidal water, in Western Europe at least, are not to be included in this first division of a great river.

The swish of the tide continues up beyond the broad estuary, the sand-banks, and the marshes, and there are reaches more or less long (rather less than twenty miles perhaps originally in the case of the Thames, rather more perhaps originally in the case of the lower Seine) which for the purposes of habitation are inland reaches. They have the advantage of a current moving in either direction twice a day and yet not the disadvantage of greatly varying levels of water. Thus one may say of the Seine in the old days that from about Caudebec to Point de L'Arche it enjoyed such inland tidal conditions; and of the Thames from Greenwich to Teddington that similar advantages existed.

The true point of division which separates, so far as human history is concerned, the lower from the upper part of such rivers is the first bridge, and, what almost always accompanies the first bridge, the first great town. To repeat the obvious parallel, Rouen was this point upon the Seine; upon the Thames this point was the Bridge of London. It is with the habitable and historic Thames Valley above the bridge that this book has to deal, and it will later be to the reader's purpose to consider why London Bridge crossed the stream just where it did, and of what moment that site has been in the history of the Thames and of England.

The second division in a great European tidal river, considered as a means of communication, is the navigable but non-tidal portion.

The word navigable is so vague that it requires some definition before we can apply it to any particular stream. It does not, of course, mean in this connection "navigable by sea-going boats." One may take a constant depth of so little as three feet to be sufficient for the purpose of carrying merchandise even in considerable bulk.

The legislatures of various countries have established varying gauges to determine where the navigability of a river may be said to cease. In practice these gauges have always been arbitrary. The upper reaches of a river may present sufficient depth but too fast a current, or they may be too narrow, or the curves may be too rapid, or the obstruction of rocks too common, for any sort of navigation, although the depth of water be sufficient.

Conversely, in some streams of peculiar breadth and constancy very shallow upper reaches may have early been converted to the use of man. The matter is only to be determined by the experience of what the inhabitants of a river valley have actually been able to do under the local circumstances, and when we examine this we shall usually be astonished to see how far inland a river was used until the history of internal navigation was transformed by the development of canals or partially destroyed by the development of railways. Thus it is certain that so small a stream as the Adur in Sussex floated barges up to the boundaries of Shipley Parish; that the Stour was habitually used beyond Canterbury; that so tiny a tributary as the Ant in Norfolk was followed up from its parent Bure to the neighbourhood of Worsted.

In this connection the Thames is of an especial interest, for it had, in proportion to its length, the greatest section of navigable non-tidal water of any of the shorter rivers in Europe. Until the digging of the Thames and Severn Canal at the end of last century it was possible, and even common, for boats to reach Cricklade, or at any rate the mouth of the Churn. And even now, in spite of the pumping that is necessary at Thames head and the consequent diminution of the volume of water in the upper reaches, the Thames, were water carriage to come again into general use, would be a busy commercial stream as high up as Lechlade.

This exceptional sector of non-tidal navigable water cutting right across England from east to west, and that in what used to be the most productive and is still the most fertile portion of the island, is the chief factor in the historic importance of the Thames.

From Cricklade to the navigable waters of the Severn Valley is but a long day's walk; and one may say that even in the earliest times there was thus provided a great highway right