at some later period this arch was blocked with a solid wall. On the west face of this wall remain considerable traces of mural paintings; the date, subject, and merit of them I must leave to others better versed in that branch of archæology; but I may be allowed to express a hope that somewhat better care than at present seems to be the case may be taken both of them and generally of this curious portion of the church, which is now blocked off as a coalhole. This bay is the only part of the aisle, or indeed of the whole church, which is at present vaulted; consequently over it the triforium is a real passage, now leading to the ringers' gallery, and therefore abandoned to a disgusting state of filth. The vault is very plain, irregular, and rough quadripartite; there are no signs of any intention to continue it in the rest of the aisle, and indeed vaulting could hardly have co-existed with the singular projections in the second bay from the tower. This bay contains the only Norman window, now blocked, remaining in the aisle. How much of the aisle wall is part of the original work is hard to discover. It appears to have undergone much patching, both in early and later times, and we must especially remember that the cloister stood against this wall of the church, and its erection and demolition could hardly have been effected without a great disturbance of the masonry. Three windows under dormer gables, as at Brecon and Malmsbury, appear in this aisle; but I am afraid that the stern laws of architectural science will not allow us to assign to them the primitive antiquity so fondly claimed by the author of the Leominster Guide. I fear that they never shed their light on the ministrations of any aboriginal contemporary of St. Teilo or St. David, or indeed on those of any one earlier than a possibly Norman intruder of the thirteenth century. The earliest and largest contains a plain Early English quintuplet, doubtless from the same hand as the similar triplet inserted in the west end of the aisle; the other two are very late and bungling insertions; though their tracery recalls a form, unsightly enough certainly, but

usual in that district at an earlier period. In these two last the whole of the upper part is of wood; and even in the other, part of the gable is, at present at least, of that material. The cause of this arrangement, as I have explained with regard to the similar case at Malmsbury1 is doubtless to be found in the presence of the cloister, which required the windows to be set high in the wall, so that there was not sufficient room for anything more than the small lights with which the Norman architects were contented, unless the head of the window were carried up into the roof. The Norman string breaks off within at a point just east of the vaulted bay, and does not reappear until just east of the quintuplet. Without, the string may be faintly traced, and the basement moulding more distinctly, to a point more to the east, somewhat west of the quintuplet; but part of the wall above the string has been rebuilt to introduce a very unsightly window, and there seems a certain amount of patching also below. At this point the string and basement break off, this marking probably the western limit of the cloister. I did not perceive any distinct marks of its western wall, but it is quite possible that it may have been of some irregular form, like those at Hereford and St. David's. At any rate there is a marked change in strings and masonry from this point. I even flattered myself that I could discern some slight traces of the vaulting of the cloister, and the string below the dormer window seems more certainly from its underside to have rested upon a roof. Near the point of junction is a small Perpendicular doorway which must have led to the cloister; it is however placed so low, that it could hardly have been accessible except by steps from the church. The ground has probably been raised, as the floor certainly has within, where it conceals the bases of the piers; but hardly enough to account for so marked a difference.

The eastern bay of this aisle appears to have been destroyed with the choir and transepts at the Dissolution,

1 Ecclesiologist, u. s.

and much further patching is the result. A wall was necessarily built across the end of the aisle against the wall continued from the arcade. It is perhaps too much to expect to be able to account for every botch and seam in a building which has undergone so much change and mutilation as this, but the circumstance of the destruction having extended to this bay may possibly suggest the idea that the transept had a western aisle, a notion however which I do not start with any confidence. I may mention that a roof-line against this cross wall is merely that of a shed which has been recently removed, at which time the ground was cleared, and the basement of the wall continued from the arcade was brought to light. The consequence is that the doorway which I before mentioned in the eastern bay of the aisle is now external; above it is what appears to be the way to the rood-loft, but I could find no sign of it within. At the same time a Debased window appears to have been inserted in the triforium range.

CHOIR AND TRANSEPTS.-I rest my belief that the Norman church was furnished with transepts, besides the a priori probability, amounting nearly to moral certainty, that such would be the case in a Norman minster, on certain appearances at the east end, small indeed, but sufficient, I think, for that purpose. We find there, now converted into buttresses, some projections which seem to me to have formed a pier of the west and north arches of the lantern; they are of the same square-edged character which might be expected from the general character of the church; a single impost ranging with them may also be discerned within. The effect of arches of this scale of such extreme severity of style must have been singularly striking. There is of course a similar example of still grander dimensions at St. Albans, but there they are of far greater, even proportionate, height, while at Leominster the prevailing dimension must certainly have been breadth. The remarkable thing is that, in this case, the destroyers were not, as at Malmsbury for instance, satisfied with building up the western arch of the lantern,

still less did they, as at Usk and Chepstow, leave the original central tower at the east end of the reduced building; they entered upon what seems rather an unnecessary piece of demolition, inasmuch as the whole east end, except the small fragments I have mentioned, was pulled completely down, and a new east end erected with buttresses and a Debased east window. Advancing along the present east end for about fifty feet north of the lantern pier, some appearances are found which seem to mark the extent of the transept in this direction. Close to the ground there still remains the stump of a rectangular pilaster, such as enter into the formation of the responds throughout the church. If the transepts had western aisles, this may be a portion of one of their responds; if not, it may be in some way connected with the south end of the transept; an exactly similar projection appears at the west end of the nave. Anyhow it seems to be a genuine Norman relic, and to mark the southern limit of the transept. Again, the wall turning eastward from this point, and forming the boundary between the churchyard and the workhouse premises which now cover the site of the choir, transepts and cloister, contains, rude and patched as it is, some portions which look very much as if they were in situ. A Decorated tomb with ball-flower can hardly be otherwise, and above it are pieces of a Norman

2 This difference would seem to imply that at Usk and Chepstow the choir had been moved eastward of the tower, so that the area of the latter was no longer considered as belonging to the monastic portion of the church. Besides, in these instances the destruction of the central tower would have left the church altogether towerless, while Malmsbury and Leominster had also western towers. Perhaps however this latter motive would not alone have been sufficient to account for their preservation, seeing that at Tewkesbury the central tower-the only one-formed part of the purchase of the parishioners, having been previously destined to destruction, while at Waltham the central tower was actually destroyed, and a western one subsequently added by the parishioners as its substitute. In speaking of Chepstow, it must be remembered that there a very similar change to this last took place at a later period from other causes. I ought to have mentioned Chepstow, while treating of Malmsbury, as another instance of a western tower built within an earlier nave.

string, shaken and shattered indeed, but still, I think, too regularly placed to be mere accidental fragments. I cannot help suspecting that we have here a genuine portion of the south wall of the transept. At its eastern finish we must be content to guess.

Such was the Norman priory church of Leominster, a noble example of the grandeur and simplicity of its magnificent style. What remains of it, we have seen, is comparatively little altered; but we have now to consider the remarkable additions made to it at a later period, which have so completely altered the general character of the building.

§ III. ADDITIONS TO THE NORMAN CHURCH.

EARLY ENGLISH ADDITIONS.-In 1239 the church was reconsecrated. The alterations of which we may safely presume this ceremony to have been the consummation were more extensive than might at first sight appear. There is now very little Early English work in the church. I have already mentioned two insertions of windows in that style in the Norman portion; besides this nothing remains except the inner and outer doorways of the porch, a piscina in the extreme south aisle, and the part of the west front in which the great west window is inserted; this last the pilaster and its strings claim as belonging to this period, while the extreme south portion is clearly wholly Decorated. But for the porch and piscina, one might conclude that the Early English work extended no farther south than the pilaster, and that the Decorated work was strictly an addition. But those portions appear sufficient proof that it was during the Early English period that the church was enlarged to its present extent southwards.

That is to say, the external wall of the south aisle of the Norman church was taken down, without disturbing the south arcade of the nave, and, instead of the narrow Norman aisle, a large fabric was erected of the same length and height, and greater width, than the remaining portion of the Norman church, divided by an arcade into two