NATURAL FORTRESSES STRENGTHENED

Man is essentially a pugilistic animal and experiences a keen sense of delight in hunting all objects of the chase, ferocious or otherwise, but the keenest undoubtedly when upon the track of the grandest of all game—man. But at the same time though willing to inflict injury he invariably does so at the minimum of risk to himself, deeming the preservation of his own life, the greatest of the gifts that Nature has bestowed upon him, of the first importance. Thus it is conceivable that after the selection of a stone or the fabrication of a club by primitive man he naturally proceeded to make a protection for himself to counteract the effect of those weapons when wielded by others, and the shield would follow as a logical sequence. The shield was to all intents and purposes a movable castle, since it afforded him the means of causing the greatest amount of annoyance to his enemy, while at the same time furnishing the maximum means of protection to himself; a definition which is appropriate to the first and latest type of feudal castle. As a non-movable protection he would soon recognise the advantages afforded by a tree, a rock, a fold in the ground; and the efficacy of these natural defences would suggest artificial examples where they were non-existent.

Hence the earthwork and the parapet of rock, singly or combined, may be regarded as the first of all castellation, with an origin so remote as to be practically coeval with man's first appearance upon earth. These simple means of defence are found in every country occupied by primitive races; in America they are numerous and undoubtedly point to a high antiquity, and the same holds good in many parts of Asia and Europe. In the British Isles we have a richer collection probably than can be found in any other portion of the globe, for in the habitable districts hardly a square mile exists without some indication of disturbance of the soil due in the majority of cases to some work of a defensive character.

Earthworks are of such a varied nature, with so many differences of contrast alike as regards shape, elevation and area, that to the ordinary observer any classification seems impossible, and practically it is only when descriptions and plans of the whole are aggregated for selection that they fall under different headings by presenting essential features common to a class. Hence in late years a system of differentiation has been evolved, and the allocation of an earthwork to a definite class is now possible. To the antiquary this is a source of keen satisfaction, and it is hoped that to the ordinary observer it may prove one of equal interest.

It should be borne in mind that earthworks of great antiquity are found only in those districts and localities where man could delve with his primitive appliances, and thus a classification presents itself at once in a contradistinction between the Western and Central parts of England compared with the Southern and Eastern. It is obvious that no primitive race, with their crude appliances, could dig into Cambrian, Silurian, or Carboniferous rock in order to entrench themselves, and that in those localities the breastwork would necessarily be paramount; and that entrenching would only be possible where an accumulation of detritus or alluvium existed, that is to say, in the valleys. So that, broadly speaking, the parapet prevails in Wales and the Midland counties and the ditch in the remaining portions. Those districts, reaching approximately from Dorsetshire to Yorkshire and belonging to the Cretaceous formation, would therefore roughly divide the country into two portions—the fosse prevailing to the east of it, and the breastwork to the west.

Another fact is apparent when dealing with this subject: the earthwork is much more durable than any other form of castrametation, in fact it is almost indestructible so far as meteoric agencies are concerned, whereas the parapet suffers not only from disintegration by the weathering influences of rain, frost, wind, and heat, but also from the tendency to lose its original shape through having no natural or artificial coherence between the separate parts. Thus undoubted examples of prehistoric ramparts are comparatively rare when compared with the wealth of existent earthworks.

It must be borne in mind that the study of the earthwork is the alphabet to that of castellation, and that the evolution of the latter cannot be efficiently comprehended without an intelligent appreciation of the former. So far as classification of earthworks has been made to the present time, the following table represents the general mode of procedure, and under one or other of its separate headings the whole of the earthworks, so far as our knowledge extends at the time of writing, may be allocated.

Classification of Earthworks

1. Natural Fortresses strengthened. This refers to fortresses partly inaccessible by reason of precipices, cliffs, or water, additionally defended by artificial banks or walls.

2. Fortified Hill-Tops strengthened. This includes fortresses situated on hill-tops, with artificial defences adapted to the natural configuration of the ground, or to those which are less dependent on the natural slopes.

3. Simple Artificial Enclosures, including rectangular or other forms, and all the fortifications and towns of the Romano-British period.

4. The Mount and Fosse.

5. The Mount and Bailey, consisting of natural or artificial mounds with one or more courts attached.

6. Homestead Moats.

7. Homestead Moats developed, referring to enclosures similar to No. 6 but augmented by supplementary defences.

8. Protected Village Sites.

Class I.—Natural Fortresses strengthened.

This division may very readily be subdivided into three parts dealing with natural fortresses according to the topographical characteristics as follows:

(a) Promontory forts, or cliff castles both upon the coasts and inland.

(b) Those depending upon rivers, woods, marshes, etc. for efficiency.

(c) Plateau forts.

(a) Promontory Forts.—This type of fort is prehistoric as a rule and not characterised by an excess of variation. No distinctive uniformity can be traced, it is true, but special features may be discovered in almost every example of the class. It is only natural that primitive man should seize upon any spot which promised the minimum of labour to adapt it for his purpose of protection, hence distinguishing features may be discerned in almost every case, depending upon the presence of a precipice, slope, bog, wood, chasm, marsh, etc. The description of a few of these fortresses will sufficiently illustrate the point.

Trevalgue Head, one mile north-east of New Quay, is practically an island, being cut off from the mainland by a chasm through which the tide flows, thus presenting a formidable obstacle 20 feet wide in places. In order to strengthen this natural obstruction many lines of entrenchments have been thrown up, both upon the island and the mainland. The presence of quantities of flint chippings sufficiently proves that this fort was the residence of Neolithic man, probably the descendant of local PalÆolithic ancestors.

As the terms "Stone Age," "Bronze Age," "Iron Age" do not convey any idea of date to the great majority of people, it may be advisable to mention that the Stone Age approximately terminated about 3000 B.C. upon the Continent, and 1500 B.C. in the British Isles, when the Bronze Age is supposed to have commenced. These dates are of course entirely conjectural. The Iron Age commenced in Britain about 400 B.C.

The general idea of a cliff castle may be gathered from the foregoing description of Trevalgue; there are many examples to be found in our Islands, and similar ones occur in Brittany. That they are of ancient British origin is suggested by the fact that they invariably occur in a district where cromlechs, stone circles, menhirs, and other Celtic remains are to be found.

Treryn Castle, about three miles from St. Buryan, contains the famous Logan stone. The fort is a gigantic mass of granite, nearly 250 feet in height, separated from the mainland by a triple row of formidable entrenchments, still 4 or 5 yards in height. This fort is probably the finest to be found in Cornwall.

At St. David's Head is a cliff castle called Clawll y Milwyr, where a small peninsula has been converted into a formidable fortress by the erection of a great stone wall about 12 feet in thickness and still some 15 or more feet in height. The only method of approaching the enclosed space is by a narrow entrance at the end of the wall. A fosse is associated with the defence in question, and several other subsidiary walls and fosses are found. Excavation has proved that the formation of the castle occurred in the early Iron Age.

Old Castle Head, Manorbier, in Pembrokeshire, may be cited as a good example of a cliff castle, and

Dinas, four miles from Fishguard, affords another, where a natural crevasse has been carefully scarped in order to separate a headland from the mainland. The examples given have been taken from South Wales and the Cornish peninsula, where for obvious reasons less probability of disturbance during later periods has occurred. Ideal spots like Portland are to be found in the British Isles, but the operations of man in quarrying, building, etc. have probably destroyed all traces of defences erected by the primitive inhabitants.

Clifton Camps, three in number, lying on either side of the Avon, afford us examples of cliff castles remote from the sea. The projecting land jutting out into the loops of the winding river has in each case been protected by lines of trenches.

It can hardly be supposed that cliff castles generally were continuously occupied, because as a rule the area is limited, and could not afford sustenance for flocks and herds. Neither do they boast the possession of the indispensable well or spring in the majority of cases. Simplicity in plan is their chief feature, and generally the fosse defending them is single, rarely double, and practically never treble. They probably afforded the last resort when hard pressed by the enemy; abandoning flocks and herds and thinking only of life and limb, the refugees could make a last stand within them, and, if fortune still proved adverse, could lower themselves down the steep faces of the cliffs, and trust to the mercy of the waters.

(b) Another class of fortresses falling under the same heading are those which depended upon woods, marshes, rivers, and similar natural defences for their efficiency.

The Dyke Hills at Dorchester, in Oxfordshire, undoubtedly formed at one time a safe haven of refuge, being almost surrounded by swamps forming a most effective defence. At the present time, however, these have disappeared owing to the general lowering of the water-level throughout England, by drainage, locks, weirs, etc., and they consequently give no indication of former efficiency. Two great fosses may be traced reaching from the Thame to the Thames, thus cutting off a piece of land and entirely defending it by means of water.

The Isle of Avalon, near Glastonbury, is essentially a peninsula, rising from the midst of a marsh with a series of aggers and accompanying dykes carried across the isthmus.

(c) Plateau Forts.—Comb Moss. One of the finest examples of this division is Comb Moss, which is situated near Chapel-en-le-Frith in the vicinity of Derby, and at about 1600 feet above the level of the sea. Its mission is so obvious that the name of "The Castle" is applied to it locally. It is roughly triangular in shape, and upon two sides precipitous slopes occur, which descend for nearly 500 feet and offer magnificent protection. The third side leads out upon a fairly level plateau, and here a double rampart and fosse has been made, completely closing the entrance with the exception of a narrow portion at the north-east side upon the very edge of the precipice, forming a most dangerous entry and consequently could be easily defended by a small number. There is an opening in the centre of the ramparts which is probably of later date, conjecturally Roman. An ancient plan shows a spring in the open space, but it does not appear at the present time. A rough wall was constructed round the edges of the precipices to confine sheep, but the original fortress was doubtless defended by a thick and massive rampart, there being no lack of material for such a protection, while the usual timber and stone breastwork would crown it.