Most islands are attached portions of the nearest mainland, severed in prehistoric times by subsidence of the intervening soil and the action of strong currents. Thus even England is a portion of the Continent, as its fauna and flora proclaim, while Ireland was severed earlier still. Thus also the Isle of Wight is Hampshire. So our Kentish islands, now only two and neither now to be effectually circumnavigated, are practically absorbed in the mainland. “Sheppey, Thanet—what else?” most would say. Yet the early geographer Nennius, writing in the eighth century, has the following quaint passage:—“The first marvel is the Lommon Marsh” (i.e., Limen, now Romney), “for in it are 340 islands with men and women living on them. It is girt by 340 rocks, and in every rock is an eagle’s nest, and 340 rivers run into it, and there goes out of it into the sea but one river, which is called the Lemn.” Truly a picturesque account of the numerous spots, where dry land first appeared in the shallow bay, and the countless sluices which intersected them.
“Romney” is probably formed by the addition of the Saxon Ige, By, or Ea—island, to the earlier Celtic Ruim—marsh, and so gives an idea of what the district was before the Romans reclaimed much by building their great Rhee Wall. Certain names in Romney Marsh preserved the same history. Oxeney (still we have the Island and the Hundred of Oxney, containing Wittersham and Stone parishes) is even now insulated by two branches of the Rother, and here, in the ancient and now diverted channel, was found in 1824 an oaken ship buried deep in sand and mud. Its name is said to mean the isle of the fat beeves. On pagan altars discovered there oxen were carved, and still it is a great cattle-raising district. We should look now in vain for the three ferries by which it was once entered. In its centre is Ebony, no doubt originally a sort of island, once called “Ebeney in Oxney,” and in an early document it appears as Hibbene. It has been suggested that the first part of the word is the old Celtic Avon, i.e., water or river, and I find that a Saxon charter of 793 A.D. calls the Gloucester Avon by the name of Aben. The third syllable is, of course, the Saxon word for island. Scotney Court, Lydd, and Scotney Castle, Lamberhurst, no doubt preserve the name of the Barons de Scotini, who came from Scotigny in North France, and possessed, in the 12th century, much land, which they held until the reign of Edward the Third, while Scot’s Hall, near Smeeth, was the seat of Knights of that name down to the time of the Armada. But for this history one might have classed Scotney with Oxney and Ebony.
Coming now to Sheppey, still an island, washed on the north by the estuary of the Thames, on the west by that of the Medway, and insulated by the Swale (Saxon Suala), crossed by a railway bridge and two ferries, the Celtic name is said to have been Malata, from Mohlt, a sheep, which the Saxon conquerors translated into Sceap-ige, Sheep Isle. It includes Elmley and Harty, once its little islands, and now peninsulas. An old name for Harty was Hertai, in which we may perhaps find the Saxon Heorat—stag, hart—as in Hartlip and Hartbourne, and ea—island. And Elmley, which island would simply denote the ley or glade in the forest in which elms were frequent, might here be Elm-isle. And another islet was Graven-ea, the grain island, on the opposite side of the Swale, now in the Faversham marshes.
Thanet, the best known and most important island, was in the 5th century separated from the mainland by the Wantsum and the estuary of the Stour, which gave an expanse of water mainly two miles broad, so that vessels from or to London sailed from Reculver to Richborough, and avoided the longer and rougher route round the North Foreland. To guard this sea-way the Romans had placed their forts of Regulbium (Reculver) and Rutupia (Richborough) at its northern and eastern entrance. Now, and for long, by the dwindling of the Stour and the practical extinction of the Wantsum, Thanet is but in name an island. Even writing in 1570, Lambarde calls it a peninsula. It is said that its Celtic name was Rimn—a headland, and that its later Saxon name, Tenet, means a beacon. Solinus, who wrote about A.D. 80, calls it Ad-Tanatos (Thanatos, Greek for Death, because its soil killed snakes even when exported for that purpose!). Nennius, in the 8th century, says Guorthigern handed to Hengist and Horsa the island, “which in their language is called Tenet, but in British speech Ruoihm,” or Ru-oichim. In a charter of 679 it appears as Tenid. The name, says McClure, may involve the Celtic Tann for oaks, as in Glastenec, an early form of Glastonbury, surviving in the English, Tanner, Tanyard, etc. From this root is derived the Breton place-name Tannouet. Tenid is its oldest Saxon form, and the Saxon Chronicle records that in 969 Eadgar ordered Tenetland to be pillaged.
Variations in the Spelling of Place-Names.
In the search for the meaning of a place-name it is necessary to go back as far as possible and discover, if we can, its earliest form. The Anglo Saxon Chronicle, and the later Domesday Book of 1086; the gradual blending of Saxon and Norman into the English tongue; and then the invention of printing; all may have had an effect on the pronunciation, and so the spelling, of a word. Also there is the tendency to shorten a long word, as when Pepingeberia becomes Pembury, or Godwinston, through Gusseton, becomes Guston. And before the standardization of spelling which printing to a great extent effected, and in written documents such as charters or wills, the most remarkable variations will occur according to personal varieties of pronunciation. Even now, though every one reads “I am going,” in one county you may hear “I’m a gowin’,” in another “I be gooin’,” in another “I be gwaine,” and so forth, and so one wonders less at the various forms in which a name appears in writing in Saxon, in Norman, or in Early English days.
As a general rule the earliest form will be best and most likely to indicate why a name was given. To illustrate this source, both of information and of error, let me take two Kentish place-names—Westenhanger and Tenterden—giving the dates at which I find the various forms.
Westenhanger.—There is a Teutonic stem hanh which means to hang, with the Anglo-Saxon later forms, Hôn, Hêng, from which we get our place names of hanger, Ongar, etc. A hanger is a wood or copse hanging on the side of a hill, and in Kent we find Betteshanger, Hangherst, and Ackhanger, as well as Westenhanger, concerning which Leland in his Itinerary writes of “Ostinhaungre ... of sum now corruptly called Westenanger.” I find it spelled Ostrynghangre in 1274 and 1291, Westynghangre in 1343, Westingangre in 1346, Ostrynhangre in 1376 and 1381, Estynghangre in 1383, Westynganger in 1385, Ostynhangre in 1409, Ostrynghanger in 1468, Westinganger in 1472, Ostrynhanger in 1478, Westynghanger in 1511, Westhanger in 1519, and Oystenhanger in 1541. The changes of the first syllable illustrate the continuance of the Saxon Wœst and the Norman Ouest until there is the reversion to the Saxon form in our West.
Tenterden, again, has a long list of variants. Probably its Saxon name indicated the place where the Theinwarden, or Thane’s Warden or Guardian, looked after the rights and dues of various other dens where his swine had pannage and his tenants tended them. It is not mentioned in Domesday, as not of sufficient importance or taxability; but in 1190 I find it as Tentwarden, in 1252 as Thendwardenne and Tentwardenn, in 1255 as Tentwardene—this early and probably original form cropping up at intervals for another three hundred years. But in 1259 we get nearer to the extant form, as Tendyrdenn. In 1300 there is Tenterdenne, and in 1311 Tentredenne. From this point I take the spellings from the Archbishop’s register of the institutions of its parish priests, and here the earliest record is Tenterdenne in 1311. Thenceforward Tent’denne and Tant’denn in 1322, Tentrdenn in 1327, Tenterdenne in 1333, Tentwardene in 1342, Tenterdenne in 1346, Tentwardyn in 1390, Tynterden in 1394, Tent’den in 1404, Tenterden (for the first time exactly in its present form) in 1407, Tendirden in 1436, Tentwarden at various dates from 1464 to 1531, Tentreden in 1501 and 1525, Tenterden in 1511, 1523, 1539, and 1546, “Tentwarden alias Tenterden” in 1541, Tynterden in 1546, Tenterden in 1556, Tentwarden in 1560, Tenterden in 1571 and 1615, Tentarden in 1619, Tendarden in 1626, Tentarden in 1627 and 1636. Henceforth it is always Tenterden in the Lambeth Registers. These variations are the more noticeable as all occurring in one office, where one would have expected a settled and continuously adopted form, whereas in such documents as wills the testator, or even the scrivener who wrote the will, would have only the current or the personal idea as to the right spelling of a name.
Elsewhere I have given variations of the places we now know as Edenbridge, Cuxton, Shepherdswell, Bethersden, Eastry, Throwley, etc. One might add the cases of Freondesbyry (Saxon), Frandesberie (Domesday), Frenesbery, and Frendesbury, for our Frindsbury; of Estbarbrenge, Barmyage, Barmling, Barmelinge, and Berblinge, for our Barming; Æpledure (Saxon), Apeldres (Domesday), Apoldre (1381), Apeltre, Appledrau, and Appuldre, for our Appledore; of Pœdlewrtha (Saxon), Pellesorde (Domesday), and Pallesford, for our three Paddlesworths; Hertlepeshille (temp. Edw. II.), Herclepe, Hertelepe, and Harclypp (1534), for our Hartlip; and Ok’olte (i.e., Oak Wood), Ocholte, Sud-Acholt, Scottesocholt, and Nokeholde, for our Knockholt. The etymological advantage of going back is seen in the case of Ringwould, which becomes more intelligible when down to the time of Henry 3rd it was known as Ridelinwalde or Rydelynewelde (i.e., the settlement of the clan of Ryddeling by the wood), whereas not till 1476 do I find Ringeweld, and Ringewold in 1502.