Wikisource

Sussex Archaeological Collections/Volume 6/An Inquiry after the Site of Anderida or Andredesceaster

< Sussex Archaeological Collections‎ | Volume 6
←
Sussex Archaeological Collections (1853)
An Inquiry after the Site of Anderida or Andredesceaster by Arthur Hussey
→
4468546Sussex Archaeological Collections — An Inquiry after the Site of Anderida or AndredesceasterArthur Hussey


​

AN INQUIRY AFTER THE SITE OF ANDERIDA
OR ANDREDESCEASTER.

BY THE REV. ARTHUR HUSSEY.

READ AT THE QUARTERLY MEETING, JANUARY, 1853.

At a Meeting of the Sussex Archæological Society in May, 1847, a paper on the above subject was read, and subsequently printed in the Archaeological Journal for that year (vol. iv); but new matter bearing upon the debated point having since arisen, it appeared desirable here to revise the entire question, rather than merely continue the former discussion, which could not be conveniently referred to by many persons.

Among the various topics which have excited, and more or less baffled the curiosity of antiquaries, one is the position of the ancient city, Anderida or Andredesceaster. But since the name even may be only partially known, it seems expedient to premise that the little information we possess respecting it amounts only to this: that the former appellation was borne by one of the fortresses, erected and maintained by the Romans, till just after a.d. 400, upon the southern shores of this island; and that, under the second title, many of our early historical records mention the siege, capture, and destruction of it, some time after the Romans had relinquished their British dominions, by the Saxon invaders. From that period the above names appear to have been totally disused: the natural consequence of which is, that now, after the expiration of nearly 1400 years, with numerous intervals of confusion and obscurity, it is left to conjecture and probability alone, whereby to identify the spot, which that long-forgotten settlement may have occupied. In our endeavours to ascertain this fact, we will begin by reviewing the circumstances attending the final extinction of the city by the Saxons, from which principally its notoriety has arisen.

​The Saxon Chronicle[1] and others of different dates allude to the fall of Andredesceaster, but, since they merely narrate its total overthrow, they are too concise to suit our present purpose so well as the copious statement of Henry of Huntingdon, which therefore we will quote:

"The kingdom of Sussex begins, which Ella governed long and most ably; but auxiliaries had joined him from his own country. . . . Belying, therefore, upon (his) large forces, he besieged Andredecester, a strongly fortified city. The Britons then collected like bees, and beat the besiegers in the day by stratagems, and in the night by attacks. No day, no night occurred wherein unfavourable and fresh tidings would not exasperate the minds of the Saxons; but, rendered thereby more ardent, they beset the city with continual assaults. Always, however, as they might assail, the Britons pressed them behind with archers, and with darts thrown with thongs; wherefore quitting the walls, the pagans directed their steps and arms against them. Then the Britons, excelling them in fleetness, ran into the woods, and again came upon them from behind, when they approached the walls. By this artifice the Saxons were long annoyed, and an immense slaughter of them was made, until they divided the army into two parts, so that while one part should storm the walls, they might have behind a line of warriors arrayed against the charges of the Britons. But then the citizens, worn down by long want of food, when they could no longer sustain the multitude of assailants, were all devoured by the sword, with the women and little ones, so that not an individual escaped. And because the foreigners had suffered such losses there, they so (utterly) destroyed the city, that it was never afterwards rebuilt. Only the desolate site as of a very noble city is pointed out to those who pass."[2]

Such is the relation of Henry of Huntingdon, a writer of the twelfth century. He does not give the precise date of the event, but places it somewhat after A.D. 490. The Saxon Chronicle assigns the year 490; others of our ancient authorities vary in some degree, though slightly. It may, and ​probably will, be objected, that the above lengthened and circumstantial account, in those respects differing greatly from the more nearly contemporary histories, being penned many hundred years after the occurrence it narrates, must at the least have been largely indebted for its details to the imagination of the writer. This certainly may be, though it by no means necessarily follows that it is, the fact; because we have good reason to believe, that many historical records formerly existed, even at very early periods, which have long ago perished. And thus it is beyond a possibility, that Henry of Huntingdon may have derived the statements he has preserved from documents which have not descended to us. However that may be, neither his diffuse description, nor the much more concise ones of other English chroniclers contain any clue whereby to ascertain the situation of Andredesceaster, except that it must have been contiguous to a forest and within easy reach of Ella's kingdom. This forest we may safely infer, even from the preceding long quotation, to have been the immense one of Anderida or Andredesweald, which extended completely through Sussex; for it is immediately after mentioning the commencement of the south Saxon kingdom, that our author recounts the fall of Andredesceaster, as if there was a close connection between the city and the district. General consent admits the locality to be either on the southern coast of Kent or on that of Sussex, though various opinions have prevailed as to the precise spot, in favour of which the probabilities preponderate. In the discussion of this question no less than eight places have been named, but of six the claims do not seem sufficiently important to require our present attention, which may be confined to the other two, namely, Newenden in Kent, and Pevensey in Sussex.[3]

But before proceeding further, a few remarks may be offered upon the character of that ancient settlement, of which the position is thus disputed. From the very name, then, whereby it is usually spoken of, we are assured that it was, if not originally founded, yet certainly adopted and retained ​by the Romans, since the termination, cester, from the Latin castrum, a camp or fort, is deemed always to imply such a fact with regard to any locality thus distinguished. And if it was a stationary Roman garrison, which is proved by the record of the troops maintained there, we may be sure it possessed marks of Roman occupation, in the existence there of walls composed with stone and lime. We should likewise advert to the chronicler's observation, that Andredesceaster was "a strongly fortified city—urbem munitissimam," which, indeed, is to be gathered from the obstinate resistance of the inhabitants to their Saxon assailants. And lastly, that the city was extensive is a conclusion equally clear from the statement of the numbers, which collected for the defence, of whom it may be noted, that the historian expressly styles them "citizens," as if they consisted mainly of parties fighting for their own families and homes, not simply of men assembled from the surrounding country to repel a foreign enemy. Wherefore, though positive information fails us, we perceive there is ground for believing, that Andredesceaster, was a large and regularly constructed Roman fortress; consequently, that the spot where it stood is quite as likely to contain at the present day some signs of Roman domination, as any of those numerous places in this kingdom, where that such traces remain is uncontested.

The idea that Newenden possesses the site we are inquiring after, appears to have originated with Camden, who writes thus: "Newenden, which, I am almost persuaded, was the haven so long sought for, called by the Notitia Anderida, by the Britons Caer Andred, and by the Saxons Andredeceaster: first, because the inhabitants affirm it to have been a town and harbour of very great antiquity; next, from its situation by the wood Andredswald, to which it gave the name; and lastly, because the Saxons seem to have called it Brittenden, that is, 'a valley of the Britons;' from whence Selbrittenden is the name of the whole hundred adjoining." After an account of the destruction of the place, "as Huntingdon tells us," he adds, that for many ages after only ruins were visible, "till under Edward I the Friars Carmelites had a little monastery built here, at the charge of Thomas Albuger, knight; upon which a town presently sprung up, and, with respect to the old one that had been demolished, began to be ​called Newenden, that is, 'a new town in a valley.' "[4] The opinion of Camden deservedly carries great weight; still he was not infallible, and since he is decidedly in error in attributing to Sir Thomas Albuger's monastery the rising of a town at Newenden and the derivation of the name the parish bears, he may have been mistaken in other points likewise. The notion just alluded to is clearly erroneous, because Domesday Book, nearly 160 years previous to the period assigned by Camden for the above foundation, not merely designates the place under the name "Newedene," but also declares it to have then possessed a market of such value ("yielding forty shillings less by five pence—mercatus de xl solidis, v denariis minus,") as proves it to have already attained considerable importance in the reign of King William I, very much greater in fact than it now enjoys. Moreover, the expressions of Camden imply, that the positions of the supposed site of Andredesceaster, of the " little monastery," and of the " new town in the valley," are identical, or nearly so; whereas, the truth is, that they are all separate and distinct spots, the first being above half a mile (speaking from recollection) from the second, and the latter nearly as far from the church, where are congregated the few houses now composing the village. It should be remembered, too, that after all Camden speaks doubtfully, his words being, "I am almost persuaded," so that he had not arrived at a positive conviction; wherefore we may be permitted to conjecture, that a dispassionate consideration of the objections to his theory might have altered his view of the question.

The principal or only real argument, by which the opinion in favour of Newenden is supported, is the certainty that there has been in the parish an ancient settlement, which, though faint traces of it only now remain, must formerly have been both extensive and of much account; and that this settlement had a direct communication with the sea,—was a port, in short.[5] The following is the description of these relics in Harris's ​History of Kent.[6] "Castle Toll. This is a raised piece of land, containing, I guess, about eighteen or twenty acres, and situated on a point of land between the river Rother and Haydon "[Hexden]" sewer; it lies about one mile and a quarter nearly east-north-east from Newenden street, and about two miles south-west from Rolvenden. On the east side it hath the remains of a deep ditch and bank, which seems to have gone quite round it. Near this Toll towards the north-north-east lies a piece of ground raised much higher than the Toll is; this was encompassed with a double ditch, the tracks of which are still to be seen in some places; and within the line is, I believe, about five or six acres of land." It being expressly declared, that the above account was given after a personal inspection, we must either admit the existence about 150 years ago of much which has now disappeared, or else pronounce the historian unworthy of credit. Dr. Harris prepares his readers to expect the gradual obliteration of the vestiges he describes through the usual effects of agricultural operations. But an intimate acquaintance on my part with both the spot in question and the late proprietor having commenced about forty years ago, I believe that, instead of the marks of early occupation becoming effaced by slow degrees, they were purposely levelled with the object of improving the general aspect and convenience of the farm. The only portion of the works noticed in the above quotation which is now traceable, or has been within my recollection, is what is there stated to have been "raised much higher than the Toll," and "encompassed with a double ditch." This is a high mound of earth of only moderate diameter, manifestly, at least in part, if not wholly, artificial, the depressions of the ditches having been much diminished through cultivation of the land during my own familiarity with the spot.

The only ostensible reason why this has been deemed the site of Andredesceaster is, that it once contained, as we have just seen, a considerable and elaborate fortification; but neither does it now exhibit, nor has it ever displayed, so far as our intelligence reaches, any evidence of that extraordinary constructive skill and care, for which the Romans were remarkable, and of which evidence examples even now remain, it must be ​observed, at others of those stations, whereof Anderida was one. Dr. Harris himself, who adopted the suggestion of Camden, remarked the entire absence of stones in and around the embankments of the Castle Toll, which circumstance he attempts to account for by supposing the place to have been resorted to as a quarry for building materials, to supply the natural deficiency of stone in the district. But innumerable proofs exist throughout the country, that in such cases the ashlar or hewn stone alone is commonly appropriated for working-up again; and although the whole should have been designedly swept away, yet most assuredly walls surrounding a far smaller space than "eighteen or twenty acres" could not possibly have been removed so completely, as not to leave in the soil some fragments of masonry, whether stone or brick, together with numerous vestiges of mortar; of which testimony we may safely challenge the production by the advocates of the Newenden theory.

In inquiries of this nature the support of our early historical writers is of course much to be desired, and accordingly for this purpose we sometimes find the first words, but no more, of a passage from Gildas adduced, when the context, as will presently appear, would give a very different aspect to the authority. Dr. Harris asserts distinctly (ut sup.), "Gildas places Andreds Chester in litore oceani ad meridiem," the real fact being, that that historian makes no allusion to Anderida or any particular place, but, speaking generally of the proceedings of the Romans preparatory to withdrawing finally from Britain, says, "In littore quoque oceani ad meridianam plagam, qua naves eorum habebantur, et inde barbariæ feræ bestiæ timebantur, turres per intervalla ad prospectum maris collocant.—And on the sea-coast southward, where their vessels were kept, and thence the barbarous wild beasts were feared, they place forts at intervals in view of the sea."[7] It has been argued that the expression "on the sea-coast south-ward" is sufficiently indeterminate to admit of being applied to Newenden; which place, it is granted, stood on the border of an estuary during the existence of Anderida. At that period however the mouth of the river flowing through the estuary was at some distance eastward from Newenden, ​originally under or very near where the town of Hithe now stands, though then most probably the channel had been diverted from the neighbourhood of Apledore toward Romney. But no weight can be laid upon this consideration, because the array of Roman coast-guard towers extended certainly from Norfolk to Hampshire.[8] A vastly more formidable objection is couched in the concluding clause of Gildas's statement, wherein he declares the Roman fortresses to have been erected, not simply "on the sea coast southwards," but even "ad prospectum maris—in view of the sea." Now the spot whereupon it is proposed to fix one of those coast towers is the very extremity and lowest edge of what alone could have been dry, sound land some 1300 or 1400 years ago; and I affirm, without fear of contradiction, that no possibility exists of obtaining from thence a view of the sea, which, at the nearest point, the modern mouth of the Rother beyond Rye, must be ten or twelve miles off, while the outlet of the ancient estuary, which was coeval with Anderida, must have been distant not less than from twenty to twenty-five miles. Moreover, a position here commanding a view of the estuary can by no means be equivalent to commanding a view of the sea, because the position would not overlook even the estuary, the ancient course of the river making a sharp turn to the left, or more eastward, immediately after passing Newenden, so that, from the elevation of intervening ground, three miles will be a very liberal allowance for the extent of the prospect down the water, and very probably nothing could have been seen beyond less than half that space. And if a navigable channel should have existed (which is very likely) in the valley under the Isle of Oxney, through which the Rother now flows seaward, the prospect in that direction would have reached but little further than on the other side, because the course of the valley speedily ​inclines to the right, so as to be completely concealed behind the uplands on the southern, or Sussex, bank. It is therefore submitted to any unprejudiced judgment, whether the passage we have been discussing is not a tolerably conclusive argument against the soundness of the opinion, which has been maintained by Camden and his followers.[9]

Our next object is to inquire upon what grounds Pevensey is considered the site of Andredesceaster; to which the first and most obvious answer is, that the place displays strong proof of having been formerly an important Roman fortress. Of the walls not "merely detached fragments exist, but throughout the circumference the greater portion is yet standing, generally ​speaking of the original height, and in unusually perfect condition, the construction of the walls being by almost universal consent acknowledged to exhibit the distinguishing characteristics of Roman masonry. The internal area, though much less than that of the earthworks in Newenden, as estimated by Dr. Harris, comprising nearly ten acres, would have contained a sufficiently large body of men to account for the obstinate resistance, which Andredesceaster is stated to have offered to its Saxon besiegers. Moreover the situation of the Roman station at Pevensey answers Gildas's description of that selected for the Roman coast towers. Now indeed it is remote from the shore, but originally the sea closely approached, if it did not actually flow up to, the very walls, the place continuing to be a port for many centuries after all connection of the Romans with Britain had ceased.[10] That these reasons are a powerful presumption in favour of Pevensey is shown by the many persons of repute in such inquiries, who have espoused that side of the question; which, however, cannot be decided by authority merely. We will therefore only mention, Camden being cited in behalf of Newenden, that a modern antiquary of superior information on similar subjects, because he had profited by the researches of his predecessors, the late Mr. Petrie, keeper of the records in the Tower, urged an argument, which has always seemed to me highly deserving of attention; that, inasmuch as every Roman station on our coast from Burgh Castle in Norfolk southward and westward to Portchester in Hampshire can be and is identified by both the ancient and the modem name, with only a single exception in each case, it necessarily becomes a strong probability, that the ancient name of the missing Roman station ought to belong to the Roman remains, which we know only by their modem appellation, the consequent conclusion being, that Pevensey alone can be the site of the long lost Romano-British city.

Against this conclusion two objections are raised, of which a brief notice must be taken: they are, first, that the district around Pevensey does not agree with Henry of Huntingdon's account of Andredesceaster being closely surrounded by forest; secondly, that the same writer expressly declares the site of ​the ancient city to have been "desolate" in his time, whereas Pevensey was a town even previous to and at that period, and long subsequently a castle of the lords of that place was occupied within the Roman walls.

1. It is perfectly true, that nothing resembling forest now exists within several miles of Pevensey, but how greatly the condition of the country may have altered in nearly 1400 years we are able only to conjecture. History informs us, that during the irruptions of the Danes the native inhabitants sought refuge in the woods, which then covered the fens of Cambridgeshire and of the neighbouring portions of our island, the date of those events being not earlier than the ninth century, perhaps later.[11] But at this day those parts are more perfectly denuded of natural timber than the Sussex district, while the latter has vastly the advantage as to solidity of soil. Moreover the actual situation of Pevensey, (that is, of the old Roman station, which, it must be remembered, though adjoining to, is entirely distinct from, the town of that name, intervening between it and the contiguous village, westward, of Westham): the actual situation of the ancient settlement is by no means so remote from (comparatively) elevated and absolutely firm land, as may be imagined by those, who are not thoroughly acquainted with the locality. Most probably Westham was always on the same level with Pevensey, which seems to have been the case with ground northward of the former place, and certainly at no great distance to the west we reach an ascent, where, though the soil may be a wet clay, marsh and floods could not possibly extend, whatever might be the condition of the Level below. Here, therefore, we may safely conclude, would be woods of more or less density, which must have spread in every direction, if permitted to do so undisturbed, which assuredly would be ​the case at the present time. Besides, a small amount only of actual timber and copse would support the character of "Forest," since a wide extent of reeds (indigenous, and alone an effectual cover), with thickets of alder, willow, or other bushes, and an occasional large tree, dispersed at intervals over the swampy plain, would have afforded ample concealment to the fugitive Britons. Wherefore the chronicler's statement, that the Britons, when repulsed by the Saxon besiegers of Andredesceaster, "ran into the woods," speedily to return to the attack, is far from being so unsuitable to the circumstances of this district, as may on a cursory consideration be imagined.

2. Secondly, we are to answer the objection that Pevensey does not fulfil the condition of "desolation," ascribed to the site of Andredesceaster. And here it is necessary to repeat, what seems to be generally forgotten, what perhaps is quite unknown to many, that by the two names, Anderida and Pevensey, we do not designate one and the same spot. To arrive at positive certainty with regard to events and circumstances of remote ages, whereof no records survive, is impossible; but if the Caer Pensavel Coit of the Britons[12] be the place now called Pevensey, it is a presumption, if not quite a deduction, that the appellations, Pensavel and Anderida, were in contemporaneous use. But, whether this was actually the case or not, it is undeniable, that the existing little town of Pevensey, though it stands side by side with, is perfectly distinct from, does not even encroach upon, the enclosed space formerly occupied, we contend, by the Romans as a fortified station, which they styled Anderida. The only English historian, it is believed, who affirms the continued desolation of Andredesceaster is Henry of Huntingdon, who flourished in the twelfth century, and his statement that "it was never ​afterwards rebuilt," combined with the acknowledged fact of a castle erected within the Roman walls having been long inhabited by the lords of the place, is urged as a clear confutation of the claim for such walls to have been those of Anderida. But, while the repairs of the outer defences do exhibit some traces of Norman masonry, the visible ruins of the castle seem to be not earlier than the thirteenth century, consequently that portion would have been constructed after Henry of Huntingdon wrote. And although the result of recent excavations appears to afford good reason for believing that a baronial residence had previously existed here, yet the reported Norman character of the remains thus brought to light, being of late date, may well imply even that the erection took place only in the very era of the historian, consequently perhaps not within his knowledge, if during his life. Moreover, notwithstanding the town of Pevensey, as already observed, closely adjoins the Roman position on the eastern side, and the village of Westham is equally near on the western, not a single habitation stands within the central area, nor are there perceptible indications that houses have ever stood there within any conceivable period. Wherefore we may, with much justice, apply to this locality now the expressions of the old chronicler respecting the city destroyed by Ella and his barbarian army, and say, that "only the site is pointed out desolate to those who pass."

With the view of easily and amicably terminating all debate upon the Anderida question, in 1851 the hypothesis was published,[13] that there were two Anderidas, one British, the other Roman! Of this idea the sole foundation is the imagination of the author, who, without advancing the shadow of a proof in its support, gratuitously "assumes" the fact. For such a theory therefore, the mere allusion to it will suffice.

No direct evidence upon the matter before us is supplied by those early records, which enumerate the Roman possessions in this island. Though more than one Roman road certainly existed in that portion of Sussex, the Itinerary of Antoninus describes no route through the district between Regnum (Chichester) and Kent, which omission will sufficiently account for its not containing the name of Anderida. ​The name is likewise absent from the small surviving fragment of the Peutingerian Tables. The Notitia simply states that the commander of the cohort of the Abulci was stationed at Anderida, without further intimation of the locality of the place, than that it belonged to the defensive works of the "Limes Saxonicus per Britanniam." The anonymous geographer of Ravenna mentions Anderesio, or Andereliomiba, by which it is supposed that Anderida is intended, and this probably is the fact; but these scanty notices comprise all the authentic information, which those documents afford on the subject. There is another professed authority, which I would gladly believe to be deserving of confidence, because its testimony would go far towards definitively settling the point. I allude to the work first made known and printed about a hundred years ago as the composition of Richard of Cirencester, a monk of Westminster in the fourteenth century. But such grave suspicion has long attached to this production, that I would on no account place any reliance upon it, greatly preferring to leave the cause now advocated with some appearance of weakness, rather than try to strengthen it by knowingly adducing any evidence, the reputation of which is not, like that of "the wife of Cæsar," untainted. Still I readily own myself indebted to Bertram's fabrication (which I sincerely believe the above-named work to be) for a suggestion, which I will proceed to offer as merely a conjecture.

That Limme in Kent is the site of a Roman station will be undisputed, more especially after the excavations so successfully carried on there in 1850. At Pevensey again we have in Sussex the remains of a fortress, which very few deny to be also Roman. Now between these two important garrisons there must of necessity have been some regularly established line of communication, unless the Romans departed both from every principle of military science, and from what we know to have been their constant practice in other instances. The direct distance between the extreme points, as a bird might fly, appears to be thirty-two or thirty-three miles, while, allowing for the very numerous and often great inequalities of the surface, actually it must be almost forty miles. Where such a line of communication passed has never yet been ascertained, perhaps not even sought after, but still existing traces of a Roman road ​running from the ruins at Limme in the direction of Pevensey have recently been observed by a competent judge.[14]

From the intervention of marshy vallies of greater or less extent, it is hardly possible that this road could long have preserved the usual straight course, but may have inclined to the west, through what is now the main street of the town of Tenterden. From the probably moist condition of Pevensey Level during the original occupation of the Roman fort there, it is uncertain what means of egress that garrison possessed toward other posts eastward and north-eastward. In different parts of this tract are numerous eyes or islands, that is, plots of sounder soil rising somewhat above the surface of the vicinity (of which islands indeed Pevens-ey is one); and if the juxtaposition of some of these elevated spots should have enabled the Romans to form a causeway, even with the addition of a ferry, northward to Wartling, it would have rendered unnecessary a most inconveniently wide circuit westward, and thence northward through Hailsham perhaps. And when Wartling was reached, it became possible to take the most direct line for meeting the road, which pointed from Limme south-westward. For this purpose, I will venture to assert, from my own long acquaintance, more or less intimate, with the intermediate country, that the Romans, from the necessity of, in military language, "turning" the many intervening ravines, could scarcely have selected a more practicable route than that of the roads now in use, which may represent, nearly if not precisely, the ancient trackways, by Boreham Street, Ninfield, Battle, Watlington, Cripp's Comer, Staple Cross, passing the Rother at Bodiam, and thence to Sandhurst in Kent. Strangers to the district may inquire, why the supposed route might not follow the coast? To which question the reply is, that, even admitting that to have been feasible from Pevensey by Hastings to Rye, yet from Rye the then state of Romney Marsh[15] and the adjacent parts would unavoidably have brought the road again into the direct line, with the ​additional objection of crossing the Rother at a more unfavourable spot.

The entire length of road between Pevensey and Limme being thus forty miles, and, allowing for inevitable digressions, more probably forty-five, that would be beyond a single day's journey for any body of soldiers under ordinary circumstances, in the regular routine of change from one station to another. Some halting-place therefore between the post in Kent and that in Sussex must have been requisite in order to complete the chain of communications, essential to maintaining the military occupation of the country; and for such an outpost the supposed site of Anderida at Newenden, previously described, is not unlikely to have been selected; with the manifest advantage of finding there defensive (British) works already provided, together with roads, such as they were, diverging thence on every side, counterbalanced only by the inconvenience of adding a few miles to the march. Possibly Newenden might have been reached from Staple Cross by Northiam, or if, as conjectured above, the Roman intercourse from Pevensey with Limme passed through Sandhurst, it would thence (most directly) proceed toward Tenterden either partially or entirely along still-existing bye-lanes from Ringle Crouch Green at the eastern extremity of the village to the western side of that of Rolvenden, or else by some other line long since obliterated. Again, from Sandhurst the distance is about three miles, very nearly straight, and as nearly level, to Newenden Castle Toll, whence however the road must have retrograded in some degree through Rolvenden in order to reach Tenterden, unless it should have been practicable, which is doubtful, to cross the adjoining marshy valley north-eastward, and pursue the left bank of the old channel of the Rother by Small Hithe and Reading Street to Apledore, or rather to Apledore Heath, the remainder of the route being then free from serious obstructions by Ham Street, Aldington, and Court at Street to Limme.

And here the question for consideration occurs, whether the names Reading Street, Ham Street, Court at Street in Kent, and Boreham Street in Sussex justify the inference, that the Roman road passed through those places? That this is actually the fact with regard to Court at Street is affirmed by ​Mr. Wright.[16] Beside those just enumerated, there is also Brook Street between Ham Street and Tenterden, as well as Gardiner Street between Hailsham and Boreham Street.

It might be an inquiry of much interest, and certainly the field of research is hitherto untrodden, to endeavour to ascertain the course of the Roman road between Kent and Sussex. The vestiges still observable in the neighbourhood of Limme would afford a most favourable starting point, and, should the clue anywhere fail, it is confidently presumed that the recovery of it must be sought in one or the other of the two lines now indicated. If Newenden Castle is not recognised as the probable intermediate post between Pevensey and Limme, or if two such posts are deemed more likely than one, the site or sites are yet to be discovered. Beside that at Newenden, and two others a little northward in Rolvenden, (of the character of which, having never visited them, I cannot speak), neither my own knowledge, nor acquired information, whether public or private, lead me to imagine the existence of any vestiges of military works in that district.[17] If such vestiges should ever be found, it can hardly fail to be solely in woodland, which still covers a very large proportion of the Wealds of Kent and Sussex, the undisturbed remnant of the ancient Grand Forest of Anderida.


  1. ↑ Gibson's ed. p. 15.
  2. ↑ Regnum Sudsexe incipit, quod Ella diu et potentissime tenuit; venerant autem ei auxiliares a patria sua. . . Fretus igitur copiis ingentibus obsedit Andredecester, urbem munitissimam. Congregati sunt igitur Britanni quasi apes, et die expugnabant obsidentes insidus, et nocte incursibus. Nullus dies erat, nulla nox erat, quibus sinistri et recentes nuntii Saxonum animos non acerbarent; inde tamen ardentiores effecti, continuis insultibus urbem infestabant. Semper vero dum assilirent, instabant eis Brittones a tergo cum viris sagittariis et amentatis telorum missilibus. Dimissis igitur mœnibus, gressus et arma dirigebant in eos pagani. Tunc Brittones, eis celeritate præestantiores, silvas cursu petebant: tendentibusque ad mœnia rursum a tergo aderant. Hac arte Saxones diu fatigati sunt, et innumera strages eorum fiebat, donec in duas partes exercitum diviserunt, ut dum una pars urbem expugnaret, esset eis a tergo contra Brittonum excursus bellatorum acies ordinata. Tunc vero cives diuturna fame contriti, cum jam pondus infestantium perferre nequirent, omnes ore gladii devorati sunt cum mulieribus et parvulis, ita quod nec unus solus evasit. Et quia tot ibi damna toleraverant extranei, ita urbem destruxerunt quod nunquam postea reædificata est. Locus tantum, quasi nobilissime urbis, transeuntibus ostenditur desolatus." —Hen. Hunt. Hist. Angl. lib. 2.
  3. ↑ The other places are Arundel, East Bourne, Chichester, Hastings, Newhaven, and Seaford, all in Sussex ; and for a brief consideration of the pretensions advanced in their behalf respectively, the Arch. Journal, iv, 208, may be consulted.
  4. ↑ Gibson's ed., i, 274.
  5. ↑ That Newenden, in very early times, the communicated directly with the sea is indisputable, but it was a port, to compare small things with great, very much of the same kind as London is now, namely, standing upon a river several miles from open ocean. This matter will be noticed below in the discussion of a frequently quoted passage from Gildas's History.
  6. ↑ Folio, 1719, p. 215.
  7. ↑ Hist., c. xiv. Mon. Hist. Brit. 11. B.
  8. ↑ It may be and is doubted by some, that the above cited passage of Gildas refers to the Roman fortresses, of which such extensive ruins yet remain, from the improbability that such considerable works should have been undertaken by the Romans when the abdication of their dominion was decided upon, and because some of the forts, which we can still trace, seem to have been built before the departure of that people. But we now behold no other vestiges of Roman military structures on our coast beside those above-mentioned, so that if they are not included under the term "turres" used by Gildas, his constructions must have vanished entirely. Wherefore I conceive that Gildas's expression is designed to embrace Pevensey, Limme, Richborough, and others already in existence at the period spoken of, together with some perhaps of smaller size and less durable materials.
  9. ↑ Since the above was written I have met with a suggestion, which, if admitted, and it appears entitled to at least some consideration, will decide the question now before us. The subjoined quotation attributes to Newenden an ancient settlement, which has been supposed (merely from conjecture) to belong to Sussex. Very possibly Baxter may be correct, since in the original record the name follows Anderesio (Anderida) and immediately precedes Lemanis and other Kentish towns, this connection being the only clue to its situation. Moreover we have, I contend, sufficient evidence, that defensive works of importance have existed within the parish of Newenden. "Mantantonis: Levi mendo in Anonymo libro Mutuantonis legitur, et in Vaticano Mantuantonis, nullo cum Etymo; cum Mant et Ment Britannis sit Os vel Ostium, et An ton isc Tenti fluminis. Ista civitas olim de Britannorum nomine Brittenden sive Britannodunum appellata est; postero vero tempore de novis incolis Newenden sive Noviodunum. Sita est ad Odaram flumen, quod Britannico vocabulo Rother dicitur, verum vitiose pro er odar, vel hodiernâ loquelâ Yr odr quod est Limes; cum veteribus diceretur Antona. Hic fluvius Regnos sive Rencos a Cantiis dividebat. Oportuit hæc omnia nescisse Camdenum.—Mantantonis. By an easy blunder it is read Mutuantonis in the anonymous book [of Ravenna], and in the Vatican [copy] Mantuantonis, with no derivation; since Mant and Ment is to the Britons mouth or entrance, and An ton isc of a river stretching outwards. This city formerly was called Brittenden or Britannodunum, from the name of the Britons; but subsequently Newenden or Noviodunum from new inhabitants. It is situated on the river Odara, which in British is called Rother, but corruptly for Er odar, or in modem language Yr odr, which is Boundary; when the ancients would have said Antona. This river divided the Regni or Renci [men of Sussex] from the Cantii [men of Kent]. Camden must have been ignorant of all this." (Glossarium Antiquitatum Britannicarum, Autore Willielmo Baxter, Cornavio, 8vo, London, 1719, p. 168). Of Baxter's etymology others must judge, but the meaning he assigns to the name of Mantantonis agrees most precisely with the locality of the old site at Newenden, and it is not very probable that a native of Shropshire could have been quite aware of the accuracy of the correspondence. The Rother does indeed divide the counties of Kent and Sussex, but for a short distance only. When in very early times the river turned to the left, immediately eastward of Newenden, toward Apledore and Limme, it could have formed the boundary only for a very few miles, unless the limit of Sussex was different then (as probably it was) from what it has now been for many centuries. Since the river deserted its old channel to run on the southern side of the Isle of Oxney, it has been deemed to separate the two counties till it enters the Marsh, but before reaching the town of Rye, and thence to the sea^ it is entirely within the county of Sussex, which is likewise the case with regard to its upper course from two miles, or less, above Newenden to its source.
  10. ↑ This assertion is proved by Domesday Book, which mentions "port dues" as accruing in Pevensey about a.d. 1086.
  11. ↑ "The Danes now proceed to Croyland, . . . the abbot Theodore . . . sent his monks to seek refuge in the marshes." About a.d. 870. Lappenberg's Hist. of England under the Anglo-Saxon Kings, by Thorpe, ii, 37. "Est autem palus illa . . . multis etiam silvis et insulis florida: intra quam sunt ecclesia Heliensis, abbatia Ramesiensis, abbatia Chateric, abbatia Thorneie, abbatia Crulande.—But that fen flourishes with even many woods and islands: within which (fen) are the church of Ely, the abbey of Ramsey, the abbey of Chateric, the abbey of Crulande." (Hen. Hunt., lib. v, anno 964. Mon. Hist. Brit., 747, E). Utter inability to consult distant authorities prevents any addition to these two quotations, which, it is confidently believed, might be largely increased, especially by the documents in the Monasticon relating to the above-named religious establishments.
  12. ↑ Baxter however understands this name to intend Ilchester in Somersetshire, interpreting the meaning as "At the head of a river mouth in a wood (Ad caput fluminei oris in silvâ)," that wood being the Forest of Selwood. (Gloss. Antiq. Britan., p. 141). The same authority also (ut sup., p. 176) considers Pevensey to be the "Miba" of the Ravenna geographer, wherein he may not improbably be correct, as the word immediately follows Anderesio or Anderelio, which is supposed to imply Anderida, and if Miba be a separate name, the fact will remove some perplexity as to the reading of the original. But in this instance Baxter is less happy in arguing from etymology, than is often the case. Elsewhere he is clearly wrong, it is conceived, in suggesting Chichester as the site of Anderida, though he too takes Anderesio to mean that place. He explains the word Anderida as signifying "The two passages," (ut sup., p. 17).
  13. ↑ In Consuetudines Kanciæ, by Charles Sandys, f.s.a.
  14. ↑ Mr. T. Wright, "Rambles of an Antiquary" in Gentleman's Magazine for 1852.
  15. ↑ I gladly acknowledge my obligation to and concurrence in Mr. James Elliot's ingenious suggestion respecting the "Ancient State of the Romney Marshes," and the alterations therein during the connection of the Romans with Britain, appended to Mr. C. R. Smith's "Report on the Excavations at Limpne in 1850," 4to, London, 1852.
  16. ↑ Gentleman's Magazine, already alluded to.
  17. ↑ After this dissertation was completed I have learned the positive existence, twenty or thirty years ago, of a (rather vague) rumour of ancient works still visible in a wood upon Burg Hill, very high ground in the parish of Etchingham on the border of Sussex, though my informant had no opportunity of searching for the reported remains. Arguing merely from a knowledge of the name "Burg Hill," Professor Airy suggests this locality for the storming of the British fortress by the Romans, shortly after Cæsar's landing, on his second expedition, in the neighbourhood, as the astronomer-royal supposes, of Pevensey, whence the direct route to London would pass very near Burg Hill along the present Hastings and London road. The same spot might have been included in the Roman line of communication between Pevensey and Limme, though I consider it would have been too far to the west, and that the natural features of the country render the course already proposed much the more probable of the two.
Retrieved from "https://en.wikisource.org/w/index.php?title=Sussex_Archaeological_Collections/Volume_6/An_Inquiry_after_the_Site_of_Anderida_or_Andredesceaster&oldid=14016399"