As to whether the names of Romney and Romney Marsh have a Celtic element, opinions differ. Isaac Taylor, in his _Words and Places_, has little doubt that they come “from the Gaelic ruimne,” a marsh, and instances Ramsey, in the Fens, as coming from the same source, and finds it also in Ramsgate, i.e., the passage through an opening in the cliffs to the marshes behind. But he wrote in 1864, and in some respects is considered too imaginative by modern philologists. Ruim is undoubtedly the British name of Thanet—Ruoihm, or Ruoichim—preceding Tenit, Tenitland, Thanet—so perhaps the situation of Ramsgate in Thanet is all we have to consider. McClure ignores “ruimne” as a derivation; but does not explain the Rumin as a name of the district. The oldest English form is in a charter of 697 A.D. Rumining—seta, i.e., the dwellings of the people of Rumin, and he inclines (though admitting it may be far-fetched) to derive from “Roman,” since the whole region is full of Roman associations. Our common suffix “den,” for a deep wooded valley, gives us probably a Celtic word adopted by the conquering Jutes. Perhaps the explanation for so few Celtic names of places having survived is accounted for by the thoroughness with which the invading Jutes either slew or drove far westwards the Celts, and so re-named whatever settlements they made. Thus, in 452 A.D., according to the _Saxon Chronicle_, Hengist slew 4,000 Britons at Crayford, and these must have formed a large proportion of the population, and this was only one of a series of victories which drove the Celts backward into the far west. Purely Celtic Kent was prehistoric; Romano-Celtic it was from B.C. 55 to A.D. 413, and yet marvellously little remains of either element. This mighty race has left us little record, though its language survives in Scotland, Ireland, and Wales. In Maidstone Museum we can study its weapons, its ornaments, and its methods of sepulture; but in our walks in Kent we are rarely reminded of its long, as well as ancient, occupation of the soil. Yet in what we might consider the purest English some undoubtedly Celtic words survive, such as basket, crook, kiln, fleam, barrow, ashlar, rasher, mattock, bran, gown, flannel. Availing myself of what has been written by Celtic or Saxon scholars, I turn to the river names of Kent, of which some are obviously Celtic and others as obviously Saxon. Ash.—The western branch of the Stour is so named, and Ashford was anciently Esshetsford. Rivers have sometimes been named from the trees on their banks, and besides our Ash-ford, we find elsewhere Ashbrook and Ashbourne; though the common Celtic esk for water or river may also be considered. In this connection I note that in a direct line we have near Detling, Boxley, Thornham, Hollingbourne (Anglo-Saxon Holeyn is holly), possibly Bearsted from the Saxon Berc for birch, and Ashford from Aesc, our ash.