THE "BELGIC DITCHES," AND THE PROBABLE DATE OF
STONEHENGE.

HE lines of ancient earth-work, which in various parts of land intersect the country, seem to admit of a division three classes,-British roads, Roman roads, and Boundary 3. When tolerably well preserved, these different kinds arth-work may, in most cases, be distinguished from

other without much difficulty, and the British road ears as a ditch, with a low mound on each side of it, Roman road as a mound simply, and the Boundary-line ditch, with a mound on one side only. As we have no son to believe that the Britons constructed artificial roads ore the arrival of the Romans, and as we know from sar that the country was densely peopled, we might bect to find their lines of communication worn into hollows. e accumulations of filth and refuse, which would necesily result from a large traffic, when thrown aside for the eater convenience of passage, would soon form continuous unds, and perhaps the more readily, inasmuch as such ounds might, in certain localities, be usefully employed as aces. There are many bye-ways in the west of England, hich, if turfed over, would be no unfair representatives of e British roads that still exist upon the downs of Wiltshire. Our ancient boundary-lines seem also to admit of a threed division. There are, first, the boundary-lines, which efined the territories of the British tribes before the Roman onquest; secondly, those which were made by the Romanised ritons; and thirdly, the march-dikes thrown up by our ncestors, after the English colonisation of the island. The st of these three classes has sometimes attracted the attenon of the historian; but the second, though for several easons particularly interesting, has not, I believe, been itherto noticed; and, if we except the speculations of Stukeley and Warton with respect to the "Belgic ditches," am not aware that even the ancient British boundary-lines have as yet been made the subject of critical investigation. According to Stukeley, the Belgæ, as they gradually expelled the British tribes, who preceded them, constructed four

successive lines of defence-Combe-bank, Bokerly-ditch, the ditch immediately north of Old Sarum, and Wansditch. Warton supposes there were no less than seven of these ditches. He does not enumerate them, but he probably added to Stukeley's four, the Grims-ditch south of Salisbury, the ditches on Gussage Cow-down, which really appertained to the British post of Vindo-gladia, and the ditch which runs over Salisbury plain to the north of Heytesbury. Neither Warton nor Stukeley point out the districts which they suppose to have been marked out by means of these boundary-lines, and the proximity of the lines to each other, is adduced as a proof of the desperate resistance which the Belgæ had to surmount before they could effect their conquest. The resistance must have been desperate indeed, which contested the possession of a few miles of worthless down-land; and the love of property equally strong, which could think such an acquisition worthy of being secured at the expense of so much labour. There can be little doubt, that the number of these boundary-lines has been exaggerated not only by Warton, but even by Stukeley.

It may be asked, what right have we to assume that the Belgæ overspread the south of Britain, in successive waves of conquest, such as are pre-supposed in the hypothesis we are considering? The only ground for such a hypothesis that I am aware of, is contained in Cæsar's statement, "maritima pars ab iis (incolitur) qui prædæ ac belli causâ ex Belgio transierunt, qui omnes fere iis nominibus civitatum adpellantur quibus orti ex civitatibus eo pervenerunt, et bello inlato ibi remanserunt atque agros colere cœperunt."-B. G. 1. 4. It may, perhaps, be inferred from this passage, that there was a succession of predatory inroads, some of which were followed by Belgic settlements; and when, in the district which we know to have been colonised by the Belgæ, we find successive lines of boundary evidently made by a people inhabiting the sea-board, to separate themselves from the tribes of the interior, it may, I think, be admitted that the

1 That these ditches might occasionally throw impediments in the way of a party of freebooters is very possible, but that they were military lines of defence, like the Roman Walls in North Britain, or the Great Wall of China, is to the last degree improbable. Such lines of defence

would require an organised body of men to guard them, and the maintenance of such a force would be beyond the means of races only imperfectly civilised. The proper character of these ditches is clearly that of boundary-lines.

hypothesis advanced by Stukeley, and accepted by Warton, is, to say the least, not an unreasonable one.

If we attempt to trace the progress of Belgic conquest by the light of Welsh tradition, we shall be disappointed. The all but utter silence of the Triads, with respect to a people who fill such a place in history, is one of the most puzzling circumstances connected with these mysterious records. The Triad, which mentions the three "refuge-seeking tribes," tells us, that the first of these tribes came from Galedin, and had lands allotted to them in the Isle of Wight. Welsh scholars consider Galedin to mean the Netherlands; and, perhaps, we may conclude, that, according to Welsh tradition, the Belgæ came as refugees to this country, and were first located in the Isle of Wight-driven, it may be, from their own country by some inundation of the sea, an accident which appears to have been the moving cause of several of those great migrations we read of in Roman history. It is clear from Cæsar, that for some centuries before Christ, the Belgæ were the most energetic and powerful--and among halfcivilised races, this means the most aggressive-of the Gaulish tribes; and we can have little difficulty in supposing, that the fugitive Belgæ, with the aid probably of their continental brethren, might soon change their character of refugees into that of assailants. Of the inlets, opposite the Isle of Wight, by which the mainland could be assailed, Tweon-ea (now Christchurch), at the mouth of the Stour and Avon, appears to have been one of the most important in the earlier periods of our history. Here, it would seem, the Belgæ landed. The uplands in the neighbourhood are barren, but the vallies rich, and the Belgæ, we may presume, were soon in possession of the pastures along the Stour as far as the neighbourhood of Blandford. This town lies in a kind of defile, over which, at that period, the woodlands of Cranbourne Chase in all probability extended. At this wooded gorge the Britons seem to have held their own, and the course of Belgic conquest to have been diverted-in the direction afterwards followed by the Roman road and the modern railway-into the vallies of the Piddle and the Frome. We may now ask,

2 This hypothesis would receive strong confirmation if we were justified in giving to the Belgic settlers of the south-east of Dorsetshire the name of Morini. But I believe our only authority for so doing is

a dictum of "Richard of Cirencester," and I will not insult the reader by quoting a patent forgery. I allude to Bertram's clever fabrication, merely to show the reader that I have not overlooked it.

whether there be any earthworks, which might serve as boundaries to the district we have thus marked out. In the first place, we observe between Holt-Forest and Cranbourne Chase, the well-known earthwork, called Bokerly-ditch, shutting in from the northward the rich valley drained by the Wymburne-brook. From Bokerly-ditch the boundary may have followed the outline of Cranbourne Chase, have crossed the Stour south of Blandford, and then run to the north-westward along Combe-bank. There was also, some years back, "in the road from Bindon to Weymouth, a great ditch, like Wansdike, for several miles."-Hutchin's Dorset, i., 217. No such ditch is now visible on this line of road, but after a long day's search, I succeeded by an accident in finding3 its mutilated remains between the Frome and Owre-brook. The bank was to the eastward, and I have little hesitation in regarding this dike as a portion of the western boundary of the first Belgic conquest. What course it took to join Combe-bank is, at present, only matter for conjecture; but there are reasons for believing, that fragments of it still exist in the neighbourhood of the Piddle river and its tributaries.

The second Belgic conquest may have included the downs of Hants and South Wiltshire. The narrow valleys that intersect the latter meet in the neighbourhood of Old Sarum (Sorbiodunum), which must always have been, what in military language might be termed, the key of the district. The Hampshire downs appear to have been called by the Britons the Gwent, or champaign. No natural frontier separates these two tracts of down, but their northern boundary is indented,

3 The dike ran nearly parallel to, and about one or two hundred yards west of "the bounds" which separated Owre from Galton. For nearly a mile it had been fashioned into shape, and formed a clayfence some eight feet thick. A wide stretch of arable land succeeded, on which it had been levelled within the last two years by an improving landlord. Its traces, however, were sufficiently obvious, and by following them, and clambering over some terrible fences, I again lighted on the object of my search, and found it running over the common for nearly a quarter of a mile, in very fair preservation. It terminated before it again reached cultivated land. I presume there must formerly have been a tract of woodland in the neighbourhood.

As these boundary-lines are often difficult to find, it may save future investigators trouble, and prevent mistakes, to learn that there are some other curious earth-works a little to the westward, round Woodford Castle. The agger runs from the Frome due south for about a mile, then turns at right angles, and after running half-a-mile eastwards, returns to the river. The agger was thrown outwards from the ditch. I suppose this work to have been the boundary of a very ancient park. A slight fence on the top of the mound, with the aid of the interior ditch, would have effectually prevented the deer from escaping. I have seen instances of similar earth-works in Berkshire and elsewhere, which seem to admit of the same explanation.

as it were, by the highlands around "Scots Poor," from which the greater part of their extent is visible. To this point the country rises from the east and south, and also, though more slowly, from the west. On the southern and eastern slopes we still find large masses of woodland-Collingbourn-wood, Dole-wood, &c.-and there can be little doubt that these high and barren downs were once encircled with a belt of forest. This description may serve to show the importance of these heights as a landmark, and in some measure to explain the fact, that at the present day three counties, and some seven or eight parishes, meet in the neighbourhood.

During a fortnight of rather inclement weather, I examined the country lying between Westbury and Ludgershall, and succeeded in finding most of the ditches described in the "Ancient Wiltshire." It is to be regretted, that Sir R. C. Hoare was not more alive to the importance of distinguishing between the trackway and the boundary-dike. His usual phrase "a bank and a ditch," more than once made me waste a day in searching for what proved, on examination, to be a mere drift-road. North of Heytesbury, however, I found an ancient boundary-line-one clearly of British origin, and perhaps anterior to the Roman conquest. I traced it from the west of "Knook Castle" to within a couple of miles of Tilshead, when it gradually died away in cultivated land. Ancient roads occasionally entered its ditch, more particularly at the salient angles, and its mound was broken and pierced in all directions by the trackways leading to the two British villages north of Knook Castle; but still, amid all the changes of two thousand years, its crest was seen stretching over the plain, and could be followed without the chance of a mistake. The next day I found "the Tilshead ditch," within little more than a mile from the spot where I had lost the

4 There are the sites of two British villages near the boundary line; and in a straggling portion of one of them, which lies beyond the dike, and which, therefore, must have been built after the boundaryline was slighted (to use a phrase of Cromwell's time), Sir R. C. Hoare found a stone-celt beside a skeleton. It is not probable that a primitive utensil like this was used after the arrival of the Romans; but the grave may have been there before the village extended itself beyond the agger.

Coins of Arcadius have been dug up among the ruins, but, I believe, no Saxon remains. We may conclude that the villages were burnt by the Saxon invaders, and never afterwards inhabited.

5 It may be worth observing that, just at the angle where the boundary line turns suddenly to the eastward, there lay a large stone on the top of the agger. I had not time to examine it minutely, nor even to chip off a fragment to ascertain the nature of the stone.