It’s 150 years this month since Lord Bute instructed John McConnochie to excavate the site at Castell Coch. The results were investigated by William Burges, who was able to build on G. T. Clark’s 1850 report and present Bute with the option of either restoration or reconstruction.
Burges submitted the “Castell Coch Report” to the Marquess of Bute in December 1872. The report is stored in the Bute Archive at Mount Stuart. It was published by “The Architect” in April 1874.
The Architect – Volume XI
January to June 1874
“A weekly illustrated journal of art, civil engineering and building.”
Published in London by Gilbert Wood, 175 Strand, W.C.
Printed in London by Spottiswoode and Co., New-Street Square and Parliament Street
The Architect was a weekly journal founded in 1869 by Thomas Roger Smith. Smith was an English architect and academic.
Collections of weekly journals were often collated in larger volumes and published as books. The Architect has been digitised by Google and is available on Google Books.
The quality of the scanned illustrations are quite poor so I’ve attempted to clean them up and present them in a similar style. There are two further images in the centre, which are incomplete due to the crease in the book.
There are some notable differences in the design of the Well Tower and Keep to the castle today. Burges was planning to build a watch tower on the Keep to provide views of the surrounding area. There are more detailed drawings in the full report, some of which you can find in the official Cadw guidebook.
These are some highlights from the report that I found particularly interesting.
Up to the excavations of last year the castle presented simply a mass of ruins to the ordinary spectator. The moat was filled up and the entrance path went over the foundations of the gateway, of which no traces were visible. Since then the moat has been cleared out and all the other parts of the building fully excavated. The foundations and lower parts have been repaired and rendered fit for any future operations.
I now propose to go through the various parts of the building, describing their present condition, and giving my data for the conjectural restoration.
William Burges
Above the kitchen were the battlements, now utterly destroyed. And here my design becomes utterly conjectural; it is quite possible, as Mr. Clark suggests, that the tower finished with a leaden flat roof and plain battlements, or it may have finished as shown in the drawing with a high roof. I have selected the high roof as being more picturesque, and as affording much more accommodation in case any project of future restoration should be entertained.
William Burges
I must plead guilty to the fact that the rest of the tower is a pure conjectural restoration. As it is evident that what remains is of later date than the kitchen tower, I have ventured to indulge in a little more ornament. Should any actual restoration ever take place, it is this tower that I should propose for the occupation of the lord and lady of the castle.
William Burges
The complete article was published over two issues in April 1874. I’ve cleaned up the whole text and included it below if you’d like to read all 6,000 words!
April 11, 1874.
DESIGN FOR RECONSTRUCTION OF CASTLE COCH
The prosecution of the works at Castle Coch, about five miles from Cardiff, is a patent instance of the local benefits to be derived from a great landed proprietor. The complete reconstruction of this fortress is suggested by the noble owner of the ruins on purely sentimental ground, since no alteration of the original plan is proposed. The example thus set by the Marquis might be conveniently followed by others who have more extensive national ruins in their possession – ruins particularly like those of Castle Coch, which are more useful to the antiquary than the archæologist, and the entire destruction of which might be effected without serious loss to art, while a reconstruction of a Medieval castle as it is now supposed to have existed, even with the slight information at an architect’s disposal – eminently slight in the case of Castle Coch – cannot fail to be an element of progress in the career of modern architecture.
Castle Coch, or the Red Castle, is charmingly situated upon a steep hill of limestone, and it owes its name to the colour of the stone with which it is built. It defended the pass where now runs the Taff Vale railway, upon which vast loads of mineral wealth are daily carried into Cardif – the same valley along which OWEN GLENDOWER passed to burn Llandaff Palace and ravage the country around it. It commanded the Pass and Taff Vale for a considerable distance, besides being a signal station for the upper country. G. C., in MURRAY’S “Handbook,” says:- “A beacon fire upon the headland of Penarth, answered here and on the opposite Garth, would be repeated from the summits of the distant mountains of Brecon and Caermarthen, and would at once spread the tidings of invasion over the whole of the southern coast.” The shape of the fortress, as will be seen from the plan published this week, was triangular, consisting of three round towers, a curved curtain wall, and a gateway defended from above and from the flanking towers. Before the escarpments and the lower portions of some of the walls were repaired, the whole presented an appearance of ruin a confused as the most inveterate votaries of pic-nic could desire; and probably until very lately the castle has served rustic builders of the neighbourhood for a stone quarry, like many another more valuable relic in all parts of the world. Indeed, nothing remains above the first storey, starting from the level of the castle yard, to assist an architect in making what is known to modera eyes and ears as a conscientious restoration. Of one tower – the Keep – there exist traces of the vaulting of one storey, with bits of the vaults and ribs. Of the hall, the external wall fronting the valley exists, together with the vaulted chamber below. Of the well tower there is the well, and a cellar, which people doubtless call a dungeon: and the stairs leading to it in the thickness of the wall are in good preservation. Of the kitchen tower more remains than of the other two, and the vault of its lowest storey is complete. Enough exists of the curtain wall to reconstruct the chemin-de-ronde, and to restore the embrasures in their integrity. A bit of the drawbridge has been found, and an angle stone in which it was worked still occupies its original place. The most, which was always dry, is almost perfect. But there is nothing to show what was the shape of the upper parts of the three towers and the hall. It is therefore incorrect to call such work as that just placed in the hands of Mr. BURGES a restoration; it is really the reconstruction of a Medieval castle upon original foundations. No attempt, however, will be made to adapt old arrangements to modern habits. There will still be a drawbridge and a moat; the keep will contain apartments for the castellan and his lady, and these will be vaulted as of old. The whole of the buildings will be roofed – even the chemin-de-ronde; and upon the well tower hourds will be fixed. Mr. BURGES, in the illustrations to his report, has shown diagrams from old manuscripts as evidence of the original existence of pointed roofs upon towers of the same description, and in the same locality: and it must be owned that the general effect of the reconstructed castle, as seen in the perspective sketch, is admirably artistic and romantic.
A MEDIÆVAL FORTRESS: CASTLE COCH
We subjoin Mr. Burges’s Report to the Marquis of Bute upon Castle Coch, which we have the privilege to publish almost in extenso:-
The volume of the Archaeologia Cambrensis for 1850 contains a Paper upon the castle in question from the able pen of Mr. G. T. Clark, in which that gentleman gives the following reasons for the erection of this fortress:-
“The pass whereby the River Taff escapes through the Glamorganshire mountains, on its way to the sea, was anciently guarded on its left bank by the Cymric camp. This served both as a fortress and post of observatíon against the incursion of the Danes. When, however, the Normans had occupied the land between the sea and the mountains, it became necessary to guard the inland or mountain side, which was still held by the natives. The Cymric camp was found to be too difficult of access for munitions of war and communication with the head-quarters at Cardiff. Consequently, a site was fixed upon lower down the mountain, and sufficiently distant from the main hill to be out of reach of such military engines as the Welsh might procure, if indeed they could procure any at all.”
Upon this platform the castle was built, and a deep ditch dug all round it, except to the south, where the natural precipice rendered it useless. Mr. Clark then proceeds to describe the plan of the castle as far as could be ascertained at the time he wrote. Since August, 1871, excavations have been carried on by order of the Marquis of Bute, under the direction of Mr. McConnochie, Mr. A. Bedford having charge of the works; and, accordingly, a far better idea can now be formed of the ancient arrangement than in 1850.
The castle appears to be of two distinct dates, which, as far as can be ascertained, were not very far apart. Some assert that Ivor Bach, who took the Earl of Gloucester prisoner in Cardiff Castle in 1158, was the founder of Castle Coch, but if so, he must have erected it in the Norman interest, as Giraldus Cambrensis distinctly tells us that he was one of the earl’s dependants. I am afraid that no plan of the present building, with its pointed arches, can be referred to the time of Ivor Bach. On the contrary, I quite agree with Mr. Clark, that the earlier portion may be of the latter period of Henry Ill, and that the second portion is contemporaneos with the earlier portions of Caerphilly Castle – say the latter part of the reign of Edward I.
The first portion probably consisted of the kitchen tower, which may have been used as a keep; the hall; and the circular curtain, with a gateway. It is doubtful whether the well tower belongs to this period, as the loopholes are different from those in the kitchen tower, and there are some traces of a double order of arches in the entrance doorway.
As far as can be ascertained there was little ashlar, if any, in this earlier portion. All the doorways and arrow-slits are formed of rubble, and even the vaulting ribs of the two lower storeys of the kitchen tower are of the very rudest description, being simply flat square stones superposed. It is, however, not improbable that the windows of the ball may have had their outside jambs formed of dressed stone. The whole building was doubtless plastered on the inside, and it is not unlikely that the same process was continued on the outside, as was not unfrequently the case in contemporary buildings. Thus we know that the white tower of London was whitewashed, and I have frequently discovered old plaster on the outside of old churches. However, the castle is said to have derived its name from the colour of the stone, just as Caerphilly was called “the blue castle” from a similar cause. This, if true, would rather tell against the external plastering.
The above fortifications were evidently not considered sufficiently strong, for some time in the reign of Edward I. very considerable additions were made.
Thus the curtain was strengthened by a thick inner wall, which doubtless took the place of the wooden platform whereby the battlements were served. Access was obtained to the lower loopholes by means of a series of arches pierced in the thin back wall.
I am uncertain whether the well tower is of this period, but for the reasons given above I am inclined to think it was.
There can be, however, but little doubt about the tower I have ventured to call the keep, for it evidently blocks up the end window of the old hall. The latter building also probably received an addition in the shape of an upper storey, and a hatch-way provided at the kitchen end, so as to facilitate the serving of the meals.
The last accession of strength was the gateway, of which the foundations only remain; but fortunately there is quite sufficient to show that the arrangement of the drawbridge, portcullis, &c., bore very considerable resemblance to those of the western gateway of the inner ward at Caerphilly.
From specimens discovered in the excavations, the roofs appear to have been covered with thin flat stones, and among Mr. Bedford’s drawings is represented a fragment of an earthen glazed ridge tile.
The new work differed essentially from the old in presenting a very fair amount of ashlar, in fact it was used very much in the same manner as at Caerphilly.
In spite of the constant dilapidations of the last two centuries, there still remain a very considerable number of cut stones; several of the voussoirs have evidently belonged to the arches of the gateways. The trefoil window-heads probably come from the upper hall. The remains of the arrow-slits show us that they ended with a rounded termination, and some corbels were doubtless part of the false parapet which surmounted the gateway.
It will be remarked that nearly the whole of the work of the second period has been destroyed, probably because it was the strongest and most important portion of the building.
I have been able to ascertain the period of this final destruction. All that is evident is the employment of the miner and of fire. That the latter was one of the agents, is proved by the charred timbers found in the well during the late excavations.
The inhabitants of the neighbouring village doubtless found the loose ashlar very useful in the construction of their dwellings. Hence we have only those pieces which have been preserved by being buried in the rubbish.
Up to the excavations of last year the castle presented simply a mass of ruins to the ordinary spectator. The moat was filled up and the entrance path went over the foundations of the gateway, of which no traces were visible. Since then the moat has been cleared out and all the other parts of the building fully excavated. The foundations and lower parts have been repaired and rendered fit for any future operations.
I now propose to go through the various parts of the building, describing their present condition, and giving my data for the conjectural restoration.
The Curtain Wall
Whenever it was determined in the Middle Ages to render a castle useless the first thing was to destroy the battlements: when, however, in later times, gunpowder was used, the battlements took their chance with the rest of the buildings. In the present instance the old curtain wall and its battlement remain, the latter resembling those of Edward I’s time at Conway, Caerphilly, and elsewhere. This old wall is 3 feet 3 inches thick and rises from the top of the ditch, the inner side of which is a sloping continuation of the wall covered with a revealment of masonry. The lower portion of the curtain is pierced with sundry loopholes, some of which retain traces of the internal plaster.
The battlement wall is 1 foot 10 inches thick, leaving only 1 foot 5 inches for the covered way. This 1 foot 5 inches is manifestly insufficient for the purpose, and must certainly have been enlarged by a wooden platform, as we see at the present day in the castle of Chillon, on the Lake of Geneva. When, however, the castle was strengthened in the time of Edward 1., a thick wall was substituted for this platform, access being provided to the lower loopholes by means of arches. These arches, in times of peace, would doubtless serve as storerooms.
The top of the walls seemed, in all probability, two distinct constructions. In the first place, the covered way was protected by a permanent roof. In rainy climates this would be necessary for the use of the sentinels, who would thus be protected in bad weather. Thus, to refer again to the Castle of Chillon, we find every external wall has its roof, and the same thing occurs at l’Aigle, another castle not far distant. In the present instance there are traces of an interior parapet wall. This was probably only carried up far enough to prevent people falling over, and the top of it took the interior supports of the roof.
The other addition to the battlements was the exterior wooden galleries, the object being to protect the foot of the wall, and to render an escalade more difficult. These exterior galleries, known as bretèches or hourds, were for the most part only employed in times of war, the timbers being movable and stowed away in times of peace; some hourds, however, such as those over drawbridges, were permanent, and even covered with lead, as in the Tower of London, where the keepers of the works were directed “to make on the same tower on the south side at the top deep allures of good and strong timber entirely, and well covered with lead, through which people may look even unto the foot of the same tower, and ascend and better defend it if need should be.”
In my attempt at restoration I have shown both the roofed covered way and the hourding, and represented the wall as it would appear in time of war. The hourd holes still remain, and they are at some distance from the floor of the covered-way, of course in this case planks were placed on the inner portions of the projecting beams, thus forming a wooden floor to the covered-way. The small portion of curtain which intervenes between the well tower and the gateway may have been provided with a wooden platform behind it, like the great curtain before the additional wall was built. It will be seen that there is a regular communication all round the building at a level with the top of the curtain beginning at the stairs of the hall, and continuing first by means of the covered-way of the curtain and then by wooden allures or galleries past the gate tower to the keep. By the removal of a few planks, the well tower, the gateway tower, and the keep could easily be isolated from the rest, and thus separately defended if the others should fall into the possession of the enemy.
The Kitchen Tower
At present there are remains of three storeys. The lower one, which is below the area level, has evidently been a storehouse, and is simply lighted by arrow-slits high up in the wall. It is vaulted with quadripartite groining of a very rude and simple character. The chamber above is also lighted by arrow-slits, but possesses a chimney. Most probably it was the hall of the garrison. At one side of the entrance passage another passage branches off, leading to the guardroom, and, like most passages devoted to that purpose, is bent at an angle.
The upper storey was devoted to the purposes of a kitchen, as is apparent from the existence of three fireplaces, a cupboard and drain; it was lighted by two narrow windows. I may observe, with regard to openings such as these, that when the outer jambs have disappeared a couple of rods held against the sides of the reveals will easily determine the question as to whether the opening has been a loophole or a window. In the present case the openings were clearly windows, although rather narrow ones. Above the kitchen were the battlements, now utterly destroyed. And here my design becomes utterly conjectural; it is quite possible, as Mr. Clark suggests, that the tower finished with a leaden flat roof and plain battlements, or it may have finished as shown in the drawing with a high roof. I have selected the high roof as being more picturesque, and as affording much more accommodation in case any project of future restoration should be entertained.
It is true that some antiquaries deny the existence of high roofs in English Medieval military architecture, and ask objectors to point out examples. As nearly every castle in the country has been ruined for more than two centuries, and as the few that remain have been converted to modern uses (like Cardiff), it is not surprising that no examples are to be found. But we may form a very fair idea of the case if we consult contemporary MSS., and if we do so we find nearly as many towers with flat roofs, and as many with pointed roofs. The case appears to me to be this: if a tower presented a good situation for military engines it had a flat top; if the contrary, it had a high roof, to guarantee the defenders from the rain and the fighter sorts of missiles. Thus an arrow could not pierce the roof, but if the latter were absent, and the arrow was fired upright, in its downward fight it might occasion the same accident to the defenders as happened to Harold at Hastings.
For high roofs there are such authorities as the Benedictional of St. Ethelwold (963-984), the Bodleian copy of Caedmon’s poems (about 1000), the painted chamber at Westminster (temp. Henry III. and Edward I.), and the celebrated MS. in the British Museum – Bible Reg. 2 B. vii.
All these are known to be English work, and it is very improbable that artists would have gone out of their way to put high roofs to castles unless they were in the habit of seeing them. At the same time where there was no space to represent the high roof, or where they had to represent people speaking from gateways or towers, they made no scruple of leaving out the roof.
There were various ways of roofing towers; where the tower wall was carried up in its entire thickness the top of the tower was in three distinct zones. There were – the battlements outside them; the covered way; and then the inner wall which supported the roof. Conway Castle towers were thus finished; frequently the covered way was also protected by a roof, and beyond the battlements there were the hourds in time of war. I have ventured to terminate the well tower in this manner.
When the wall was thinned, as in the case of the kitchen tower, the battlements were strengthened with internal arches, so as to make sufficient thickness for the feet of the rafters. The gateway of Chepstow is arranged in this manner, only the wall of the battlements is of great thickness, and the embrasures are splayed both outside and inside; sometimes the embrasies, as in the last mentioned instance, have a few courses of masonry on the top and thus become windows. In most cases I believe they were protected by wooden shutters, like those employed in the portholes of ships, which shutters were either hung to the wall plate of the roof above, or into staples fixed into the stone lintels, if they were used in connection with the window variety of embrasures. According to an illumination published in the “Archaeological Journal,” these shutters appear to have been pierced with arrow-slits.
In fact, it is rather difficult to imagine how an archer could have served an open embasure without being picked off by the enemy. There is one difficulty about the kitchen tower which I have hardly been able to resolve to my own satisfaction, and that is the thickness of the walls on the flat sides, viz., that looking towards the curtain and that looking towards the hall. There may have been stone passages, as shown on the restored plan, or there may have been wooden screens. In the latter case the wall would have been comparatively thin, while the objection to the former alternative is that the plain face of the masonry looks very much as if it had been an external wall. We must, however, remember that the hall and exterior staircase might have been, and probably were, additions, and that they may have been built up against a former exterior wall; the most curious thing is that in the wall which faces the curtain there is a large square recess of some 4 inches deep, evidently formed to take the thickness of a door, and thus give more passage room.
April 18, 1874.
A MEDIÆVAL FORTRESS: CASTLE COCH
By W. Burges
The Hall
The lower hall evidently belongs to the first period, and is vaulted with a pointed rubble arch which springs low down, indeed far below the heads of the windows, the window arches consequently cut into it. There are faint traces of a flight of stairs in the jamb of the entrance doorway which load into the upper hall, terminating about where the hatch occurs. Mr. Clark supposes that this opening, instead of being a hatch, was a doorway communicating with the flat roof of lower hall, and that the latter finished with battlements. From the size of this archway and the inutility of battlements on this side of the building, as well as for other reasons, I am more inclined to believe that it was simply a hatchway for the handing out the dishes from the kitchen to the upper hall. I have, therefore, ventured to restore this portion of the building with a high roof, chimney, and other accessories of a hall. As regards the windows, a reference to the drawings will show that there are two types, one of which has a rebate on the outside, and the other has simply a hollow chamber; but none of them have any grooves for glass. Now at the period in question we very frequently find windows of cathedrals with the glass contained in wooden frames, which frames are placed on the outside (chapter-house, Salisbury). I should, therefore, be inclined to apportion the windows without the rebate to the lower hall, and those with it to the upper hall. I suspect that the lower hall was used by the garrison, and that the flight of stairs was a means of private communication with the upper hall, which was probably the residence of the governor. The ramparts would be reached by the external flight of steps, but persons ascending them would be obliged to pass the last window of the upper hall, so that they would always be under the eye of the governor. The lower hall exhibits no traces of fire-place, and was paved with thick rough paving; the upper has been utterly destroyed, and its existence, in default of documentary evidence, can only be proved by induction. Should the rectangular walls of the kitchen prove to be thin, then the division between the hatch and the hall would be made by a wooden screen, with a minstrels’ gallery above.
Keep
This tower has undergone far worse usage than any other part of the building, except the entrance-gateway. Only a portion of the lowest storey remains. This chamber – which, by the way, is octagonal – had a rib springing from each angle. The section of these ribs is precisely similar to that of those in the kitchen tower. There is, however, a slight attempt at ornament in the corbels from which the springers start. The vault is really a domical one, the filling in being concentric with the backs of the arches. Only two compartments are treated in the usual manner – viz., those where the loopholes occur, and where height was required. This lower storey was evidently used for stores, of which a considerable quantity must always have been required, especially in time of war. Of the ground-storey, only one angle remains – viz, the one which abutted against the outside of the hall. There we find the remains of a loophole and part of the passage to the latrine. In my restoration I make this passage communicate with the side of the entrance-door. This involves a very long passage, and it is quite possible that the latrine may have been entered from the apartment itself; but, when we consider the plans of the other and more perfect towers, we see how very careful the original builders were to keep those accessories apart from the living-rooms. I must plead guilty to the fact that the rest of the tower is a pure conjectural restoration. As it is evident that what remains is of later date than the kitchen tower, I have ventured to indulge in a little more ornament. Should any actual restoration ever take place, it is this tower that I should propose for the occupation of the lord and lady of the castle. The roof is similar to that of the kitchen tower, only there is added a wooden ceiling. The embrasures might be easily converted into windows by omitting the outside shutters, and putting casements in the inside. I have also ventured to add a watch tower from whence an extensive view might be obtained. This watch tower would be constructed upon the principle of many eastern minarets, which are nothing more than a corkscrew stairway, the stone steps coming right through to the outside. These minarets are so strong that they may frequently be seen standing where the adjacent mosque has long perished.
Gateway
The gateway has been utterly destroyed above the level of the inside area. Below that level it is quite perfect, and shows accurately the position of the drawbridge, the pole for the counterpoise, and the groove for the portcullis.
Several of the stones of the transverse arches have been found, and deposited in the lowest storey of the keep, and a series of corbels in the same place point to the existence of a parapet projecting some 6 inches, and carried on them.
The first thing was to get the shape of the arches by producing the course of the Voussoirs. The next was to consult the contemporary work at Caerphilly.
Of the western gatehouse of the inner ward, it will be seen how very similar one is to the other. Of course I am unable to tell the exact arrangement of the Caerphilly drawbridge, as excavations would be necessary for that purpose, but the drawbridge arrangements which have recently been brought to light at Cardiff reproduce almost exactly those at Castle Coch.
In both cases the drawbridge worked with a weight or weights attached to the inner end, the pivot being placed nearly in the middle. When the bridge was drawn up the weighted part went into a hollow in the foundations formed for the purpose, and the part above ground entirely covered the portcullis.
In both cases there is a projection along the side of the wall, as if to allow a passage for the purpose of applying some arrangement for finally securing the bridge when it had been raised by the chains. At Castle Coch this passage – or rather side projection is only 1 foot 2 inches, but at Cardiff it is very much larger, and appears to be an addition. Beyond the drawbridge was a framework of timber supporting the road to the middle of the ditch, and which could be easily burnt if the outworks were taken. I have put the windlasses for both the drawbridge and portcullis on the upper storey, as it was from the highest storey that the captain of the tower issued his commands, it being a position where he could see how matters might be progressing. Behind the drawbridge was the portcullis, which, from the narrowness of the groove, might have been formed of metal, or at all events of wood, or by a considerable amount of iron. It is singular that both in this case and in that of Caerphilly the portcullis must have risen up in the centre of a room. This is by no means an unusual arrangement; indeed, at Chepstow the portcullis of the covered way must have gone up into the middle of the little oratory. In this case the portcullis was very small, and might have been raised by manual labour, but in the other instances it was a weighty machine, and required a windlass to move it. Two under-pieces of wood very much decayed, and about 6 inches diameter, were found in the moat; under the gateway they ended in iron ferules, and had a spike of iron inverted in the middle to act as an axle. These pieces of wood are generally supposed to be a part of the windlass. They, however, appear to me to be too small for that purpose and to be rather the last timber of the drawbridge which contained the pivots. The pivot holes are unfortunately wanting at Castle Coch, but they occur several times in Caerphilly.
Some time ago Mr. Clark very kindly informed me of a windlass still to be found in the Tower of London. I have had it measured, and from this the windlass appears to be 1 foot in diameter, and in all M. Viollet-le-Duc’s restorations in his Dictionary a much larger diameter is given. Immediately behind the portcullis was the door, secured by strong bars; and here the defences appear to have ended, as far as Castle Coch was concerned; but at Caerphilly there was another portcullis at the inner archway. The room over the gateway was devoted to the frame for hanging up the portcullis, while both at the Tower of London and at Carlisle is a light wooden erection, strengthened only by strips of iron and a multitude of nails. The floor of this chamber is pierced with sundry holes, and is also provided with a fireplace, in order to heat water and other substances to pour down upon the enemy, who might have so far destroyed the defences as to have entered the gateway.
The upper storey is devoted to the windlasses which respectively raise the drawbridge and portcullis. It is true they might have been placed in the storey below, but then they would have interfered with the working of the machicolations, i.e. the hot water holes, and might have been easily damaged if the enemy should have lit a fire below (as was often the case) to burn the wooden portcullis and the great door. Projecting from this upper storey is a hourding, or as it would in this case be called a bretèche.
This bretèche was a usual accompaniment to a gateway and drawbridge. Thus Henry III, commands the keepers of the bishopric of Winchester to make in “our castle of Winchester a drawbridge with a bretèche above it at the entry of the great tower.” It will be observed that the gateway communicates constructively only with the keep – the removal of a few planks assuring its isolation from the main curtain if necessary.
The Well Tower
Of the well tower two storeys and a part of a thin still remain. The lowest story is an oblong vault, to which access is obtained by a flight of steps leading from a lateral passage in the entrance doorway of ground storey. It is ventilated by two shafts, one at either end. These terminate outside the tower in little square holes, situated some considerable distance above the ground floor. This vault was doubtless used as a prison.
The ground floor presents us with another lateral passage on the other side of the doorway leading as usual to the guard-room. The outer doorway appears to have been of two orders, and perhaps the conjecture may be hazarded that the chapel might have been situated in this tower. This is the more probable as there is an internal staircase from the ground to the first storey. This staircase starts at the side of the well, then stops at a landing, and then turns abruptly at right angles in the thickness of the wall and enters the first story in the splay of window jamb. There are also traces of a circular staircase on the first storey, which doubtless led to the battlements.
This completes the survey of the ruins and my conjectural restoration. As to the latter I must claim indulgence, for the knowledge of the military architecture of the Middle Ages is a long way from being as advanced as the knowledge of either domestic or ecclesiastical architecture. It is true that M. Viollet-le-Duc and Mr. G. T. Clark have taught us a great deal, but we are still very far behindhand, and the restoration I have attempted will I hope be judged according to the measure of what is known, not of what ought to be known.
As regards the outworks (a most important portion of a Mediæval castle, for as long as they could be held the castle was safe from the miner) I have said nothing, for no excavations have as yet been made, and I know from experience how very difficult it is to seize the points of a case where be knowledge derived from this source is deficient.
