The first branch is said to have passed into Britain about 800 B.C., and the second about 630 B.C. Thenceforward, but for a few place-names, chiefly of rivers and heights, and still fewer words which have survived in our tongue, we know little until the visit of Julius Cæsar in B.C. 54, from whose _Gallic War_ we learn of some of the Celtic tribe-names and place-names. Otherwise we know little apart from the river roots which we find all over N. Europe (and hardly any in England are non-Celtic), especially the five main words for river or water—Afon—Dur—Esk—Rhe—and Don. Kent itself in the earliest records is found as Ceant from the Celtic Cenn—a head or headland, which again appears on the other side of our land as the Mull of Cantire. We have also our Chevening, which, like Chevenage, embodies the Celtic Cefn—ridge (still Cefn in Welsh). And “Kits Coty House” on our neighbouring Down gives us Ked—a hollow, and Coit—a wood, i.e., the hollow dolmen in the wood. Mote Park sounds modern enough to some; but our “park” is the Celtic parwyg, an enclosed place, while the much later Anglo-Saxon Mote denotes a place of local assembly. Dun was their word for a hill-fort, and so we have Croydon (with a Saxon prefix) for the fortress on the chalk range, though most of the old British fortresses which preserved the name when occupied by Romans or Saxons are in other counties. Penshurst, on the other hand, has a Saxon suffix to the Celtic Pen, still unchanged in Welsh as meaning a head or hill, perhaps only a dialectic form of the Gaelic Ceann, or Ken, which we have already noted in “Kent.”