Such weirs did certainly sufficiently regulate the stream as to make its banks regularly habitable.  If no local order, at least the interest of villagers in their mills sufficed to the watching of the stream.

We have in the place names upon the Thames a further evidence of the antiquity of its regulation, for, as will be seen in a moment, none give proof of any important settlement later than the eleventh century.

These place names not only indicate a continuous and early settlement of the banks, but also form in themselves a very interesting series, whose etymology is a little section of the history of England.

Of purely Celtic names very few survive in the sites of human habitation, though the names of the waterways are almost universally Celtic, as is the name of Thames itself.  But it is probable that in the Saxon names which line the river there are many corruptions of Celtic words made to sound in the Saxon fashion.  We cannot prove such origins.  We can surmise with justice that the “tons” and “dons” all up and down England are Celtic terminations; they are almost unknown in Germany.  There is a somewhat pedantic guess, drawn (it is said) from Iceland, that we got this national name ending from Scandinavia; so universal a habit would hardly have arisen from an admixture of Scandinavian blood received at the very close of the Dark Ages and affecting but small patches of North England.  Moreover, as against this theory, there is the fact that quite half the Celtic place names mentioned in our early history and in that of Gaul had a similar termination.  London itself is the best example.

If, however, we neglect this termination, and consider the first part of the words in which it occurs (as in Abing-don, Bensing-ton, Ea-ton, etc.), we shall find that most of the place names are Saxon in form, and some certainly Saxon in derivation.

Thus Ea-ton, a name scattered all along the Thames, from its very source to the last reaches, is the “tun” by the water or stream.  Clif-ton (as in Clifton-Hampden) is the “ton” on the cliff, a very marked feature of the left bank of the river at this place.  Of Bensing-ton, now Benson, we know nothing, nor do we of the origin of the word Abing-don.

The names terminating in “ham” are, in their termination at least, certainly Teutonic; and the same may be true of most of those—­but not all of those—­ending in “ford.”  Ford may just as well be a Celtic as a Teutonic ending, and in either case means a “passage,” a “going.”  It does not even in all cases indicate a shallow passage, though in the great majority of cases on the Thames it does indicate a place where one could cross the river on foot.  Thus Wallingford was probably the walled or embattled ford, and Oxford almost certainly the “ford of the droves”—­droves going north from Berkshire.  One may say roughly that all the “hams” were Teutonic save where one can put one’s finger on a probable Celtic derivation such as one has, for instance, in the case of Witham, which should mean the settlement upon the “bend” or curve of the river, a Celtic name with a Teutonic ending.