ceive, even from its ruins." It is, therefore, probable that an improvement in the art of masonry was introduced to such of the most costly structures of these periods, as were erected in years least exposed to factious trouble.

We have a specimen of the works of Alexander, Bishop of Lincoln, in the castle of Newark, already described.

In the present stage of our work, whilst viewing the castellated structures of this country chiefly as fortifications, it may be desirable to present an explanation of the terms used in describing several component parts of the keep and outworks. Such a section, although superfluous to many readers, may yet be acceptable to others, and can scarcely prove uninteresting to

any.

The keep (in some instances called the dungeon, and, in others, emphatically styled the tower) was the strongest part of the castle; and, consequently, formed the great dependance of the garrison in time of close siege. It was, indeed, the citadel of the fortress. Here were constructed the apartments in which the lord and his family resided; and, in early times, all the rooms intended for purposes of state and hospitality were, likewise, contained in the same isolated and limited part of the fortress.+ Although these rooms maintained a superior dignity in the esteem of

Observations on the Western Counties, as quoted in the Beauties for Dorsetshire, p. 502.

In Mr. Dallaway's "Observations on English Architecture," is the following remark, which it may be amusing to quote, in illustration of a term sometimes bestowed on the chief tower of an ancient castellated structure :"Amid the ruins of castles, we are frequently shown those of one called the "Maiden Tower," as in Lord Surrey's sonnet, at Windsor castle :

"With eyes cast up into the mayden's tower,"

Warton, in a note on this word, very satisfactorily proves that it did not refer to the habitation of the fair sex, or to the tower's having never been taken, but simply a corruption of the old French "magne," or "mayne," great. Hist. Engl. Poet. Vol. III. p. 13.

of later ages, additional halls, (as has been previously observed) unconnected with the massy outlines of the keep, were erected, as society attained a greater polish and more enlarged notions of enjoyment.

The keep was commonly situated near the centre of the fortified works; but not invariably so. Instances in which this custom

placed in a line with the exte

was neglected, and the keep was rior walls, occur in several pages of the Beauties of England.* The outward form chiefly prevailing in this part of the castle, at different eras, is noticed in the respective sections of these remarks on the progress of military architecture.

In the improved state of the science of fortification, the entrance to the citadel, or last retreat of the garrison, was guarded by portcullises,† as impervious to assault as were the ponderous walls of the structure; or by machicolations, from which, destructive weights,

* It is observable that the keep at Portchester, Goodrich, Castleton, and several other fortresses ascribed by some antiquaries to an ante-Norman date, stands close to the outward wall of the castle.

The portcullis is believed to have been first introduced to the military architecture of this country, in the instance of early Norman castles. The nature of this machine is almost too well known for repetition; but it may be observed that the herse, or portcullis, was a strong grating of timber, fenced with iron, and made to slide up and down in a groove of solid stone work, within the arch of the portal. The bottom was furnished with sharp iron spikes, designed to strike into the ground, for the sake of greater firmness and solidity, and also to break or destroy whatever should be under it, when it was let fall. The groove in which it rested was always contrived so deep in the stone work, that it could not be removed by assailants without pulling down the whole wall.-See Archæol. Vol. IV. p. 370.

Machicolations "over gates, are small projections, supported by brackets, having open intervals at the bottom, through which melted lead and stones were thrown down on the heads of the assailants; and, likewise, large weights fastened to ropes or chains, by which, after they had taken effect, they were retracted by the besieged." Grose, preface to the Antiq. of England and Wales. It must be added that machicolations were not always projecting works, but sometimes consisted of rows of square holes in the vaulting of

portals,

weights, or heated fluids, were precipitated on the heads of those who endeavoured to force a passage.

The walls of the keep were chiefly designed for protection, through a massiveness of character which derided assault before the use of artillery. Their few embrasures, or loops, for the discharge of arrows, were calculated for the annoyance, rather than the discomfiture of an enemy. The great theatre of active defence was situated on the top of the castle, where a platform was generally constructed, with an embattled parapet; and, from this elevated spot, the defendants discharged swarms of darts, or loads of weighty stones, by means of various engines.

The dungeon, or prison, of the castle, was a comfortless subterranean cell, usually, but not uniformly constructed immediately beneath the keep-tower. At Rochester we find it placed under a smaller tower, which adjoins the keep; and in the castles of Warkworth, Northumberland; and Spofford, Yorkshire; two former seats of the noble family of Percy, the repulsive cell designed for the incarceration of offenders, is situated beneath a tower entirely detached from the main body of the structure, of that inhabited by the baron and his family.

Whilst mentioning the dungeons of ancient castles, (which have by some persons been confounded with the whole keep) it is desirable to remind the reader, that, although grants for cas. tles to become state-prisons were usual in the early Norman ages, we are not to understand that, in consequence of such a grant, the whole castle became a prison. The fact appears to be, that, by virtue of this permission from the crown," the usual dungeon of the castle was, by royal authority, appointed to be a public and privileged prison at all times; whereas the dungeons of other castles were permitted to be used as such only in time of war, and it was unlawful at other times to confine any persons therein. But the upper apartments of these keep towers, in which the dungeons

portals, used, as is stated above, for pouring down heated sand, melted lead, and other destructive articles.

dungeons were, continued, in both cases, to be constantly used as state apartments, for the residence of the lord of the mansion, notwithstanding the prison underneath. And hence, perhaps, arose the practice, in early times, of committing state prisoners to the custody of different lords at pleasure; which custom was continued even to the time of Elizabeth, when the origin of it was forgotten."*

The outworks, however formidable, being the weaker parts of a castle, and those, from many causes, most subject to demolition, they in few, if in any, instances retain to the present day the precise features of their original construction. The great varieties of form observable in the ground-plans of ancient castles, will be obvious on an inspection of those pages of the "Beauties" which treat of such structures. Natural circumstances, and the excursions of caprice, often operated so largely on the architect's design, that it is, indeed, impossible to present any single example, as a satisfactory illustration of the mode used in the distribution of the outline and attendant works.

The following remarks on this head may not be unacceptable. It would appear that the Anglo-Saxons constantly affected the circular form, in regard to exterior lines of defence, where such a method was not denied by imperative natural circumstances; and encompassed the keep with concentric walls. The Anglo-Normans were more variable, and introduced many bold novelties of style in the disposal of their outworks.-The fortified area attendant on the keep of most castles, of a date not earlier than the advent of the Normans, may, however, in general terms, be stated as consisting of two divisions, named the outer and inner ballia. On the extremity of the works was a circumambient ditch,f 2 A uniformly

Archæol. Vol. IV. p. 403; and Munimenta Antiqua, Vol. III. p. 251. + Ditch, moat, fosse, or vallum. These various terms are used to express a hollow space on the outside of walls, or ramparts. Archæol. Vol. XII. p. 146. When the ditch was dry, there were, sometimes, subterraneous passages, through which the cavalry could sally. Grose's Preface to Antiq. of England and Wales.

uniformly filled with water, when such a circumstance was attainable; but a ditch, or fosse, was still formed, even if it remained dry.

The most prominent part of the architectural fortification was termed the Barbican, or Barbacan; which may be succinctly described as a "small tower, for the station of an advanced guard, placed just before the outward gate of the castle-yard, or ballium." Mr. Grose, in the preface to his Antiquities of England and Wales, quotes" diverse authors," in regard to the meaning of the word Barbican, and the use to which this part of a castle was assigned; who "all agree that it was a watch-tower, for the purpose of descrying an enemy at a greater distance."+

But such an opinion appears liable to this objection:—the barbican, as usually described, was a small tower, of much less altitude than the keep; and, therefore, was not nearly so well calculated for the discovery of an enemy approaching in the distance. If we reject the probability of it being designed as a tower of ob. servation, we can scarcely believe that it was intended as a serious addition to the strong defensible character of the fortress; for it appears to have been of an inconsiderable size, and, as it was often protruded beyond the ditch, must be more easy of assault than the towers on the mural line protected by that wide and deep vallum. Possibly it was, in most instances, rather an appendage of honour to the castle; the spot for receiving stately announcements, and returning answers, by voice of herald. But, at the same time, it, assuredly, acted as a protecting cover to the entrance; although, if its customary situation, and comparative strength, be accurately described, it must have been of little avail on the occurrence of a regular siege.

The barbican, if placed beyond the outward ditch, was united to the main parts of the fortress by a bridge of stone, in early

Archeol. Vol. VI. p. 308.

Antiq. of England and Wales, 4to. edit. p. 9.

ages;