richer, but its proportions yield in beauty to those of Westminster. The richness of the whole is also vastly increased by the wall surfaces between the arches being enriched with a square diaper. The wall arcading is of exquisite design, (see Drawing,) and the spaces over it were filled with most beautiful foliage, with figures interspersed, while the spandrels of the cusping were filled with ornamental painting. When, to the richness of architectural detail, we add that of material,-the entire columns and all the subordinate shafts being of marble, and the remainder of stone of several different shades of colour, the magnificence of the internal design must have greatly exceeded that of its French prototypes. The only one point which strikes the eye as looking less rich, is the use of merely moulded capitals to the main pillars. This, however, arose from their being of Purbeck marble. It is true that at Ely and elsewhere, as in our own chapterhouse, the carved capitals are of this stubborn material; but its use may, nevertheless, be accepted as a fair excuse for moderating the workmanship. The internal designs of the transept ends are truly magnificent, indeed, I doubt whether their equals can be found elsewhere. The manner

in which they continue the lines of the general design, and yet add diversity to the forms, is truly artistic.

It is most unfortunate that the great rose windows have lost their original character; I have, however, a strong impression that the old ones may have, in their leading subdivisions, resembled that now existing in the south transept, and that the design has been simply translated from that of the thirteenth to that of the fifteenth century. I have attempted in the accompanying drawing (see next page) to translate it back again, and you will see that it makes a very fine window, in perfect accordance with the character of the church, and very much like several existing specimens. You may say that this is pure conjecture, and so it is—but it is a conjecture not devoid of some collateral corroboration, for, singularly enough, there exist in the chapter-house some encaustic tiles of a pattern evidently copied from a rose window, and agreeing precisely in its divisions with that under consideration, representing even the shafts with their caps and bases. It will be seen that my translation of the existing window into Early English almost precisely resembles the pattern given on those tiles. The square form in which the circle is inscribed seems to be original from the systematic way in which the vaulting is accommodated to it, but it must be admitted, on the other hand, that there are in the eastern jamb of the south window some indications of the design having been altered from the original intention; though, as I think, this was an alteration made during the progress of the work, as neither the opposite jamb of the same window, nor either jamb of the opposite window, show any such indications. The south window was, I believe, renewed in the fifteenth century, and again in the seventeenth ; Sir Christopher Wren informs us that it had been renewed about forty years before the date of his report.

The north window received its present form in the eighteenth century, and in no degree resembles its predecessor. Whether that which Sir Christopher Wren reports to be in a dangerous state was the original one, we have no means of telling.

The works undertaken by Henry III., and completed in 1269, terminated immediately to the west of the crossing; the line of junction can be readily traced. I think the older work may have included one bay of the great arcade and aisles, or, to say the least, some of its details were continued in that bay; but in the first clerestory window of the western arm the change is clearly seen in the diversity of its eastern from its western jambs. (See Drawing.)

The five bays west of the crossing are the work of Edward I.

They differ chiefly from the work of his father in the plan of the columns, which have four attached and four detached shafts, (the latter in most instances secured by fillets of brass,) in the greater number of the ribs of the vaulting, and in the substitution of shields for carved enrichments in the spandrels of the wall-arcading. The rib-moulds of the vaulting are