Part 4
There is this also to be said in favour of Mr. Warne’s conjecture. An attacking force must find out and capture the strongholds of the defenders, which would naturally be made more strongly, and therefore last longer than the camps of the invaders. And this is what we see in the suggested line of the Roman advance. First, on the east, Flowers, or Florus Bury Camp, and Bindun, then Mai-dun (Maiden Castle), after that Eggardun, and finally, at the western limit of the county, Conig’s Castle and Pylsdun. All these are (as far as can be seen now) British camps of refuge; all of them must have been captured before the Roman generals could feel secure in their own isolated position on a foreign shore. That they were one and all occupied by the conquerors is also most probable, and would account for the discovery of Roman relics within their areas. No Roman camps can be seen at all approaching in strength or size these magnificent hill fortresses. It is, of course, well known that the armies of Rome never halted for a night without forming an entrenchment of sufficient size to include not only the fighting men, but the baggage train, and though traces of these still remain on the hills of Dorset, the majority have long ago disappeared.
Perhaps the most interesting example of the military occupation of the two races is to be seen at Hod Hill, near Blandford, where a well-defined Roman Camp is constructed within the area of a previously occupied British fortress, and here have been found spear heads, arrow heads, spurs and portions of harness, rings and fibulæ, and fragments of pottery, all indicating the Roman occupation; iron was found more generally than bronze, and the coins are those of the earlier emperors, including Claudius, in whose reign Vespasian made his conquests. Badbury, four miles north-west of Wimborne, Woodbury, near Bere Regis, and Hambledon, five miles north of Blandford, may be referred to as memorials of the time of the Roman occupation, though not of Roman construction.
Poundbury Camp, with its Saxon appellation, deserves special mention, for, being situated on the outskirts of Dorchester, it has been studied more frequently perhaps than any other earthwork in the county. It has the form of an irregular square, with a single vallum, except on the more exposed west side, where it is doubled, and traces have been discovered of other ramparts now obliterated. On the north the camp overhangs the river and valley, once probably a lake or morass, and here the defences are slight. The area within the vallum is about 330 yards from east to west, and 180 yards from north to south. Some authorities affirm that it was raised by the Danes about A.D. 1002, when they attacked Dorchester. Stukeley regards it as one of Vespasian’s camps when engaged in his conquest of the Durotriges, while other antiquarians claim for it a British origin, prior to the Roman invasion. Mr. Warne, whose opinions are always worthy of most careful consideration, “holds it to be a safer speculation to regard it as a Roman earthwork,” and, no doubt, in form and general outline and size it is very similar to other Roman camps, and altogether different to the magnificent British fortress Maiden Castle, not two miles away. Many Roman relics have been found, including coins ranging from the times of Claudius to Constantine, and a tumulus is still to be seen within the vallum, which alone would be an argument against its Celtic origin.
Poundbury is insignificant indeed when compared with Mai-dun, and it is impossible by mere description to convey an adequate impression of this great earth fortress, singled out by many as the finest work of its kind. It certainly surpasses all others in the land of the Durotriges, and probably nowhere in the world can entrenchments be seen of such stupendous strength. This camp, which is said to occupy 120 acres, is in form an irregular oval, embracing the whole of the hill on which it stands; its length is nearly 800 yards, and width 275 yards. On the north, facing the plain, there are three lines of ramparts, with intervening ditches, the slopes being exceedingly steep, and measuring over 60 feet from apex to base. On the south the number of ramparts is increased, but they are not so grand, and, indeed, as Mr. Warne remarks, they appear to have been left in an unfinished condition. At the east and west ends are the two principal entrances, and here the ingenuity of the designer is manifested in a surprising manner. At one end five or six ramparts, at the other as many as seven or eight are built, so as to cover or overlap one another; vallum and fossa, arranged with consummate skill, to complete the intricacies of entrance, and to compel an enemy to undertake a task of the utmost difficulty and danger.
In later times this camp was, no doubt, occupied by Roman troops as summer quarters, its healthy position rendering it very suitable for the purpose. Perhaps, still later, it became the residence of some Roman magnate, who selected that fine eminence for his country villa; at any rate, there should be no difficulty in accounting for the discovery of Roman coins and implements, or even of villas, on the sites of the camps and castles of the British. Many a hard fought battle must have raged around their earthen walls.
Ever and anon, with host to host, Shocks, and the splintering spear, the hard mail hewn, Shield breakings, and the clash of brands, the crash Of battle axes on shattered helms.
Many a shout of victory must have been heard as the conquering legions forced their way over the ramparts and planted their eagles on the summit of the captured fortress. And once captured they must have been retained, at first perhaps by a fairly large garrison sufficient to prevent re-capture, then as the tide of battle ebbed from the neighbourhood the numbers might have been reduced; but the sites, always in some commanding position, would have been long utilised as points of observation and centres of control over the conquered tribes.
No revolt is recorded as taking place in the west of Britain such as that led by Queen Boadicea in the east, in the year 61; so in looking back to the Roman occupation, it is reasonable to suppose that before the end of the first century it was reduced to the condition of a Roman province. Trade would soon commence with this, the latest, addition to the Empire, and the soldiers, no longer necessary except as garrisons and guardians of the peace, would be employed in improving the means of communication. The warlike Briton (in these parts at any rate) was transformed into a peaceful husbandman, who sowed and reaped, and paid his taxes, grumbling perhaps, but on the whole contented with his lot.
Roads, or trackways, of some kind there certainly were in use by the British, linking tribe to tribe, or camp to camp, and, judging by the line of what we now term Roman roads, it is most probable that to a very great extent the ancient routes were taken as the foundation of the new system developed by the Romans. The details of this system are given by an authority of contemporary date in _The Itinerary of Antonine_, which is believed to have been compiled in the third century, and possibly corrected and added to later. In this work we find, as regards the County of Dorset, a description of roads which are easily recognised to-day, roads which are still in use throughout a considerable portion of their length. It must not be lost sight of that these roads are in very close connection with some of the principal British hill-fortresses, which fact would stamp them as being originally constructed by the British race, though to all appearance they are grand examples of Roman skill and energy. The main road, the Via Principalis, of the third and fourth centuries, comes to Dorset from Old Sarum, in Wiltshire, one of the grandest of British camps; it passes close to Badbury Camp, and then makes for Maiden Castle, and onwards to Eggardun, all of earlier date than the Roman invasion. But notwithstanding this obvious connection, the roads as we see them now bear witness to the power of Rome, and are, perhaps, some of the most obvious of the memorials of the past. They are described in the _XV. Iter._ of Antoninus, with the names of the Roman stations and the distances between them along the road from Silchester (Calleva) to Exeter (Isca Dumnoni), which forms a portion of the great Via Iceniana.
After passing Old Sarum, this road crosses the north-east border of the modern county of Dorset at a small hamlet called Woodyates (near Cranborne), taking a south-westerly course; it passes over Woodyates and Handley Down, and is described by Sir R. Colt Hoare as being at that point “the finest specimen of a Roman road I know.” It runs by Badbury Camp, and thence to Dorchester, where the direction changes to due west, parallel with the coast line, and after leaving the county near Lyme Regis proceeds towards Exeter.
It must not be supposed that this, or any other Roman road, can be traced exactly throughout its whole course. Far from it: the hand of the destroyer has been heavy indeed on these relics of the past, built with a prodigal expenditure of time and material. It is often the case that the modern “turnpike,” or county, road has been made on the very site of the ancient road, the old embankment being levelled to gain additional width. In other places cultivation and the demands of agriculture may have proved the cause of its obliteration. Here and there, especially on waste land and heather-clad downs the true Roman work may still be seen, though covered, perhaps, by grass; but with the aid of the Antonine _Itinerary_ much may be learnt and many a portion be recognised. He names, however, only two stations within the county, and the mileage is short by nearly twenty miles, so in all probability, through error in copying, one other station has been omitted. The two stations mentioned are Vindogladia and Durnovaria. Authorities differ greatly as to the true position of the former of these places; the other station, however, is remarkable for the proofs of its former importance. Here we find no less than four roads meeting, from north, south, east, and west, the east and west roads being the Via Iceniana; the others are roads of less importance, that to the south leading to the sea and towards the Roman settlement, Clavinio (or Jordan Hill), near Weymouth, and that to the north passing Stratton and on to Ilchester.
Other branch roads were made as necessity required in different parts of the county. Thus we find traces of a road leaving the Via Principalis, near Badbury, to connect with the harbour of Poole, and another, starting from the same point, running northwards. Mr. Warne is considered to have made a very interesting and clever discovery of a station which is missing in the _XV. Iter._ of Antoninus. The distance there given between Vindogladia and Durnovaria is quite obviously too short by some fourteen miles. But on Kingston Down, near Bere Regis, the cultured eyes of the learned Dorset antiquarian discovered traces of a Roman settlement, and on due investigation being made, it was considered that there was sufficient proof to establish at this point a station called Ibernium, referred to by other writers as existing in the county.
The position of Vindogladia, though a subject of long and frequent debate, and though stated by some to have been at Badbury, by others at Wimborne Minster, has now been accepted as on Gussage Down, not far from the north-west border of the county where crossed by the Via Iceniana. This is due to the researches of Sir R. C. Hoare, and stands on a par with Mr. Warne’s discovery of the other Roman station on the great military road.
* * * * *
We come now to a very interesting period of the Roman occupation, when we may imagine the military operations at an end, a firm and beneficial government established, and the colonists (at any rate), who usually obtained a third part of the conquered territory, becoming rich and enabled to build those houses that must have been the envy and admiration of the native population, with their decorative floors and walls, and ample comforts for seasons of heat or cold.
Still, as we have said before, it is not to any printed records that we can turn for its history, but rather to the result of careful excavation and the relics unearthed after fifteen centuries’ burial in the soil: in a word, we trust to the use of the spade for bringing before our minds the life of the past and restoring the memorials of ancient Dorset.
In Warne’s map of the county, prepared in the year 1865 after most patient research and personal investigation, there are more than fifty sites given where relics of the Roman colonisation have been found, exclusive of Durnovaria. Mr. Moule, writing in 1893, says: “Roman work of one kind and another has been found here in Dorset in eighty places, and that for the most part casually.” But year after year this number is increased, and, truth to tell, so frequent are the discoveries that in Dorchester the ordinary labourer, when excavating in the streets, or elsewhere, is ever on the alert, and many a treasure rewards his watchful care; and even children whose eyes have been trained aright will find, when digging in some neglected corner of garden or field, a bit of common pottery, a fragment of Samian ware, or perhaps a coin bearing the image of an Emperor of Rome. And thus our history is written: a word discovered here, a sentence there, until the story of the life of those days may be once more told afresh. The frequency of these discoveries is so far interesting that it draws attention to the large area over which the Roman settlers were distributed. No doubt they found this land of the Durotriges a pleasant land to dwell in, as we do now in this twentieth century. But here may be said, in passing, that Roman colonists were partly at least a Roman garrison. They were frequently old soldiers intended to keep in check the conquered nation, and liable to be called back to active service. But if there was no fear of a hostile rising, the military character of the colony would gradually be lost. And that, no doubt, soon happened here, for the very great majority of the relics of the Roman occupation are signs of its peaceful character.
The discovery of the sites of Roman villas scattered in more or less isolated positions throughout the county tend also to prove this, and especially when the villa is shown to have possessed one of those beautiful mosaic floors which can only have belonged to a prosperous and wealthy colonist or to a British landowner left undisturbed in his possessions, and who employed the Roman craftsmen to build him a house. These tessellated floors have been frequently exposed to view in various parts of Dorset, and too frequently, alas! through ignorance or carelessness, been neglected or destroyed; others, again, have been examined, plans or drawings made, and been covered up once more. Among those which have been described may be mentioned: Thornford and Lenthay Green, near Sherborne; Halstock, six miles south of Yeovil; Rampisham, twelve miles north of Dorchester; Frampton, six miles north of Dorchester; Preston, near Weymouth; Creech, near Wareham; Fifehead Neville, north-west of Blandford; Hemsworth, five miles north of Wimborne; and in Dorchester itself no less than twenty different pavements, either complete or in portions, besides one on the upper area of Maiden Castle. It is difficult to assign a date, even approximately, to these villas, for the coins found amidst the _débris_ cover practically the whole period of Roman occupation, and the other objects generally discovered on the site are not of much assistance. There are no records of inscribed stones being found, which might have helped; and, as a rule, the more valuable part of the building materials, such as cut stones, roof slabs, and timbers, must have been taken away when the houses were left; but the wells and refuse pits are the happy and profitable hunting-ground of the antiquary.
The tessellated pavements are so interesting and attractive that it is worth while to describe them in detail. The system adopted in their construction seems to have been as follows:—If no provision was made for heating the rooms by means of a hypocaust or hot-air flues, the ground was prepared by rough levelling, and 6 to 8 inches thick of flints rammed, or coarse, gravelly mortar or concrete laid; on this 3 or 4 inches of better class white cement, and above some fine cement to take the tessellæ; and after these were laid a liquid cement would be run into the interstices before the final polishing was commenced. The system of laying is well shewn in the annexed illustration, taken of a pavement _in situ_, before removal to the Dorset County Museum.
The tessellæ themselves, as generally found in Dorset, consist of small cubes of stone or brick, but vary in size from about ⅜ or ½ inch to 1½ inch; the smaller are used for the decorative portions; the larger for the borders, or for passages, or for the floors of houses of a humbler character. The colours are for the most part only four—namely: white, dark slate (or blue-black), red, and a sort of drab or grey; occasionally yellow is found, but not often.
The materials of which the tessellæ are composed has given rise to much discussion and, indeed, much difference of opinion; but, as a general principle, it may be assumed that, wherever possible, local stone was used. The red tessellæ are merely brick or tile of a fine description; but, as a means of obtaining a scientific opinion of the other stones, microscopic sections have been cut from the tessellæ and submitted to an expert mineralogist, who has given them the following names. The very dark stone is a fine-grained ferruginous limestone; the grey is also a fine-grained limestone; the drab or yellow is an oolitic limestone; and the white is a hard chalk, showing foraminifera very well. It is believed that the colour may be altered by submitting the stone to heat, an opinion held by Professor Buckman, and explained in a very interesting chapter of his book, _Roman Art in Cirencester_.
The mosaic floors found in Dorchester are, as a rule, of very simple but effective design, consisting of geometrical arrangements of the single guilloche, the twist or plait, the double guilloche (which is extremely handsome in mosaic work), and the ordinary fret. These, being arranged as outlines of intersecting squares and circles, leave spaces of varying dimensions, spandrels, or trefoils, which are utilised for the introduction of many diverse emblems, such as the fylfot or swastika, the duplex, sprays of foliage, urns, and interlacing knots. In the County Museum may be seen laid on the floor (in which position alone can full justice be done to the skill of the Italian artist) two nearly complete mosaic pavements. One of these shows the area of three adjoining rooms, with entrances or vestibules; the other pavement, found in 1905, is in excellent preservation, measuring 21 feet by 12 feet 6 inches, and is remarkable for two ornamental vases, with two serpents issuing from each.
The pavement at Preston, near Weymouth, still _in situ_, was discovered in 1852, the coins found near the villa dating from the middle of the third century. At Jordan Hill, close by, is the Roman settlement, Clavinio, which has been productive of a large number of very interesting relics.
At the entrance to the village of Preston, coming from Weymouth, may be seen an arch spanning a small stream. The form and masonry of the arch, as well as its proximity to the other remains here noticed, point to the probability of Roman construction, and is of special interest, as examples of Roman masonry are but rarely found still existing in the county.
The pavement at Lenthay Green was discovered in 1836, and was carefully removed to the dairy of Sherborne Castle. It contains a representation of a sitting figure playing on a lyre, and a second figure dancing and playing a pipe.
The villa on Maiden Castle was discovered by Mr. Cunnington in 1882, and as a result of his excavations he sent to the County Museum many interesting objects: fragments of mosaic floor, wall-plaster, and roof tiles, a curious bronze plate (repoussé work) representing a helmeted figure holding a spear, and coins from Helena, A.D. 290, to Arcadius, A.D. 408.
A mosaic floor at Frampton is remarkable for the introduction into the design of the Christian monogram ☧, known as the Labarum. Extensive excavations were made here at the end of the eighteenth century, and four different pavements were found. They contain numerous representations of heathen deities, Neptune being especially favoured; a motto worked into one of the borders runs: “_Neptuni vertex regmen sortiti mobile ventis_,” and some other words partly lost. The introduction of the monogram of Christ is probably of a later date than the original work. The pavement is now covered up, but Mr. Lysons, who superintended the excavations in 1797, obtained accurate drawings of the whole site, the mosaic work being shown in correct colours on seven large plates which were published, together with an accurate description.
The pavements uncovered at Fifehead Neville are also of great interest. The first was discovered in 1881, measuring about 14 feet by 12 feet, the design consisting of an urn, or vase, in the centre, around which seven small fish are depicted, and outside them are four sea-monsters, like dolphins. Coins found here date from A.D. 270 to 340. The second pavement, found in 1903, requires no description, as we are permitted, by the Editor of the Dorset Field Club, to reproduce an illustration which appears in the Club’s _Proceedings_. The general plan of the design is almost identical with a pavement found in Dyer Street, Cirencester, though the details are altogether different. They may well have been designed by the same artist.
* * * * *
Very little has been said, so far, of Dorchester itself, and yet the modern town is full of memories of the Roman Durnovaria. It lies within the boundaries of the ancient walls, their position, still plainly discernible in many places, being marked by broad walks and avenues of trees. One small portion of the masonry of the wall itself may still be seen in the West Walk. The position of the gates is also fairly easy to decide, though no vestige of them remains. The roads which issued from them have been referred to at a previous page.
It has been asserted recently—and, indeed, proved to the satisfaction of many local authorities—that the course of a Roman aqueduct can be traced here and there to the west of Dorchester skirting the adjacent valleys and hills. It is believed to have been an open water-course, obtaining its supply from the source of a small stream some twelve miles distant.