Hill-forts.
When a clan had succeeded in establishing itself, it had to provide for its protection against cattle-lifters and slave-hunters; and gradually and by immense labour great strongholds were constructed on suitable sites. Comparatively rare in the south-east, they are conspicuous on nearly all the hilly districts of England, Wales, and Scotland;[520] but it is in the western and south-western counties that they most abound. Devonshire and the adjacent parts of Somersetshire contain not less than eighty; and almost every spur on Salisbury Plain is fortified.[521] The multiplicity of these camps bears witness not only to density of population and constant warfare, but also to the utter disunion which existed at the time when they were constructed. Supposing that the majority of the forts in Dorsetshire, for instance, were built in the Late Celtic Period, we should have to conclude that the Durotriges, who then inhabited that district, were merely a loose aggregate of scores of clans, ever ready to prey upon one another; for if the forts had been destined only to repel the attacks of some other tribe, they would hardly have been so numerous and so widely scattered. It is true that the Gallic Morini in Caesar’s time had not become welded into one state, and that the Kentish clans were under four petty kings; but in the period when the older earthworks were thrown up it would seem that far less progress had been made towards union. But even supposing that most of the prehistoric forts were later than the Bronze Age, their purpose accorded with the methods of primitive warfare. A chain of modern fortresses impedes an invader because, while they remain uncaptured, he cannot pass between them without exposing his line of communication. But in ancient times, when one tribe attacked another, it had no communications to guard: the invaders carried their food with them, and when it was spent trusted for support to the enemy’s country.[522] If a tribe had desired merely to protect its frontier, it would not have erected hill-forts but a continuous entrenchment.
Amongst those which were occupied in the Bronze Age or before may be mentioned Badbury Rings in Dorsetshire;[523] the stone fort on Whit-Tor in Dartmoor[524] and another in the Rhonddha valley in Glamorganshire;[525] Small Down camp near Evercreech in Somersetshire;[526] the fort of Carn Brea in Cornwall;[527] the series of entrenchments which mark the spurs of the hills that command the valley of the Esk from Guisborough to Whitby;[528] those which line the western border of Worcestershire;[529] Oldbury, some three miles east of Sevenoaks;[530] Hollingbury on the Sussex Downs;[531] Lutcombe Castle on the Berkshire Downs, overlooking the Vale of White Horse;[532] and the greatest of all—the Maiden Castle, whose stupendous ramparts are the pride of Dorchester.[533] But it is probable that the greater number may ultimately be referred to the Age of Bronze.[534]
The form, construction, and materials of British forts are naturally diverse. In Cornwall, Devonshire, Wales, and other places they were of course built largely or wholly of stone, the masonry being always uncemented: elsewhere they were true earthworks. Leaving out of sight the question of their date, they may be grouped in three classes.[535] The first comprises those that were erected on promontories or other heights which on one or more sides were fortified by precipice, river, or sea. Such was the fort of Carl’s Wark in Derbyshire, which, on three sides, rises almost sheer above the swamps of Hathersage Moor. On the west, where the ground slopes towards the plain, a huge earthen rampart, faced with dry masonry, afforded secure protection; and the slopes below the eastern and southern sides are strewn with great stones which must have fallen from the walls above.[536] The ‘cliff-castles’ on the coasts of Kirkcudbright and of Wales and on the headlands between the Land’s End and Cape Cornwall belong to the same group.[537] In the second class the entrenchments, traced upon commanding sites, which, however, were nowhere so steep as to dispense with artificial aid, followed the tactical line of defence which the nature of the hill indicated. Most of the heights on which they stand are covered with soil so thin that they never could have been thickly wooded, and if trees had encumbered their sides they would have been cut down; for the object of the engineers was to leave no ‘dead ground’ on which an assailant could conceal himself. If he felt strong enough to lead his clansmen to the assault, he knew that they could not avoid being exposed from the moment when they penetrated within the range of a bow or a sling. General Pitt-Rivers, who did so much to illuminate the study of prehistoric fortifications, was never weary of calling attention to the skill with which they had been designed. Once only, when he was exploring the camp at Seaford, he thought that he could detect evidence of neglect. As he stood upon the rampart he noticed that an advancing force would be able to conceal itself for a while. Presently, however, it flashed across his mind that time had done its work upon rampart and ditch; and soon excavation proved that the latter had lost by silting seven feet of its original depth. The general saw with delight that the designer had been as vigilant as any of his contemporaries. The rampart in ancient times must have been at least five feet higher; and then the garrison who manned it would have been able instantly to detect the first enemy who ventured within range. ‘How carefully,’ he wrote, ‘the defenders economized their interior space, drawing their rampart just far enough down the hill to obtain a command of view, but not one yard further.’[538]
In certain cases, however, the hill was so extensive that if the tactical line of defence had been slavishly followed, the defenders would have been too few. Then the chief engineer modified the accepted principle. Selecting a spot at which he might safely abandon the natural line, he made his sappers build a cross rampart at right angles to it straight across the hill-top until it joined the works on the further side. An example of this device may be seen in the camp of Puttenham in Surrey.[539]
Among the more famous strongholds of the second class are Cissbury on the South Downs, which, as we have seen, was almost certainly erected in the Neolithic Age,[540] Badbury Rings, and the Maiden Castle. This noble fortress must surely have deserved its modern name. No British force could ever have taken it: no other country can show its match. Three lines of ramparts defend the northern and four the southern side: gaining the summit of the road from Weymouth, you see them outlined against the sky; and as you mount the hill-side, they rise, one behind another, like veritable cliffs. Worn by the rains of five-and-twenty centuries or more, they still stand sixty feet[541] above their fosses; and their entrances, on the east and the west, are guarded by overlapping works so intricate that if a column had succeeded in forcing its way across the abatis, it would have found itself helplessly winding in and out as through a labyrinth, pounded on either flank and enfiladed by stones and arrows discharged at point-blank range.
The strongholds of the third class were erected on lower hills or on high ground little elevated above the surrounding country, and therefore depended less for their protection upon natural features.[542] Those that have been explored belong to the Late Celtic Period.[543] It may be doubted, however, whether such forts were generally later than those whose sites were more commanding; for the inhabitants of every district could only choose the best positions which they could find.[544] Cherbury camp indeed, about four miles south-east of Fyfield in Berkshire, was built on a lowland plain.
Some of the Gallic forts which Caesar saw, and of our own, were in his time inhabited by large industrial communities; but although many of the British strongholds which belonged to the Bronze Age contain the foundations of huts and broken pottery,[545] it is doubtful whether they had more than a few occupants except in time of war.[546]
Every explorer who has tried to imagine the conditions of life in ancient British forts has noticed that many of them have no apparent source from which water can be obtained. It has indeed been suggested that where there was neither a spring nor running water within reach the garrison had recourse to dew-ponds, which are still used for watering cattle on the Hampshire downs.[547] But even these reservoirs were generally lacking. Pitt-Rivers, however, argued that in the chalk districts many sites which are now remote from water may have possessed springs. At the village of Woodcuts in Cranborne Chase, after cleaning out a Roman well, one hundred and eighty-eight feet deep, he found no water, but the iron-work of a bucket.[548] But even where there was no spring it is easy to understand how the garrison supplied themselves. None of these camps was ever subjected to a prolonged siege. No army can undertake such an operation unless it can ensure a continuous supply of food; and to do this requires forethought and organization of which barbarous clans are incapable. Again and again the Gauls with whom Caesar contended, whose civilization was far more advanced than that of the Britons of the Bronze Age, were obliged to abandon movements that might otherwise have succeeded, simply because their commissariat had been neglected.[549] When ancient Britons were obliged to take refuge in their stronghold, they knew that the danger would pass if they could hold out for a little while. Women and children who failed to reach the entrenchment in time were doubtless slain or enslaved. But otherwise the worst that was to be dreaded was the loss of crops or stock and the destruction of dwellings. We may suppose that while the cattle were being driven into the fort the women carried up in vessels of skin or earthenware as much water as would suffice for a few days. Such was the practice of the Maoris at a recent time.[550]
Primitive metallurgy.