Another curious phenomenon in our Cathedral can only be accounted for, I believe, on the transformation hypothesis. I mean the tiny pair of arches (they cannot otherwise be characterized), forming a very minute bay, and flanking the screen at the west end of the Choir. Some have supposed (Winkle's Catliedrals, vol. ii, p. 100) that the space between the towers and altar had not been accurately calculated, and that this was the space left over; others, that the object was to have some feature of importance here, at the junction of Choir and Nave, to make up for the absence of a central lantern; others, that it was a mere setting for rood-screen. These conjectures are all alike seen to be gratuitous, when we consider what would of necessity happen in the endeavour to transform a solid Norman structure into a light and open Decorated one. There would naturally be, at the termination of the arcade on either side, broad wall spaces, serving by their solidity as an abutment for the long line of arches. These would in Norman work either be left plain, or be treated with solid engaged shafts. But the Decorated architect had to do something with the space: and in this instance he determined to introduce another pair of his magnificent shafted pillars, though at so small a distance (only 2ft. 6in.) from the pair next eastward And this was no mere "pomp and prodigality" of columnar work: since the pressure of the arcades from the east had to be provided for now, no less than in the Norman period. The expedient involved, of course, a very diminutive pair of arches; and the bay had to be eked out, above, with a single triforium arch, and above that, again, with a recess, in lieu of a clerestory window: all in a somewhat irregular and extemporized fashion. But the device was on the whole very successful; and being masked, to some extent, by the screen, is not felt to be an incongruity, but harmonizes fairly with the general design.

A curious confirmation of the view here taken of these arches is supplied by Ely Cathedral. There, at the junction of Bishop Northwold's Early English presbytery (c. 1237), with the latter (14th century) chorus cantorum of Alan Walsingham, exactly such a device was restored to, and evidently to meet the same difficulty. Northwold had left the Norman chorus intact (exactly as Marshall did at Exeter forty years before), including odd spaces left in this case at the east end of the Norman arcade. These exhibited internally Norman pilasters: and there he left the pilasters to this day. But out in the aisles he applied a new treatment to these spaces; introducing a very small pair of arches, forming a miniature bay in the aide, exactly as ours do in the Choir itself. It will be shown hereafter that the difficulty of the corresponding odd space at the east end of our Nave was got over by another device.

I content myself for the present with these proofs of the fact of transformation: only adding that it is quite certain, from documents, that there was a Lady Chapel as early as 1237 (Note 13); that there was also, then or earlier still (as the windows prove), a retro-choir; that the upper windows in the transeptal Chapels, the crypt under St. James's, and also (as it seems to me) the buttresses all round the Choir and Lady Chapel, are of late 12th century date: to which may be added the corbel tables outside the east part of the Lady Chapel, and at the east end of the Choir. It seems impossible therefore to resist the evidence of the Norman and Transition structure having covered the same ground as the present one does. And the retention of its walls, in the main, is, I conceive, proved by the facts above given.

It was, then, into a Choir thus developed and transformed, - developed by Marshall, and transformed by Bitton - that Stapledon proceeded to introduce those splendid and costly equipments, which, when in their original state, and as yet neither tarnished by time and neglect, nor rifled by inconoclastic zeal, must have been almost dazzling in their magnificence. The loftiest and most elaborate episcopal throne or canopy (57 feet in height), the most exquisitely carved sedilia (27 feet high), the most costly altar, probably, (as an altar, and apart from jewels and gifts,) in the world; - these were but a part of its furniture. Of the richness, again, of the "tablature," or reredos, we can only form a conjecture from the sums lavished on it. And it seems a probable inference, as we shall presently see, that the throne was matched by stalls and other woodwork of corresponding costliness and beauty.

But let us take these various features in the order of their construction.

The first step, then, of which we have any record, was the removal of the stalls from their former position to the new one. For it should be well understood, that the new building involved new arrangements in this respect. In this, as in all Norman Cathedrals, the chorus comtorum extended orignally, no doubt, across the transepts. And though we might have supposed that this arrangement would come to an end when the building was lengthened eastward by Marshall, there is every reason for saying that it was not so. In the parallel cases of Ely and Chichester (see above), the additional arches were not thrown into the Choir, but formed a retro-choir: as they still do at Chichester, and in part at Ely also. And in both those cases the Choir remained under the lantern(1), as it still does at Chichester, and did at Ely, until Essex's alterations in the last century. The same was evidently the case at Exeter until Stapeldon's time. It is true that of Marshall's four newly added bays, two, it is probable (from the position of his tomb), were thrown into the sacrarium; the altar being removed much further eastward: while two, as at Chichester, remained behind the altar. This removal of the altar must have involved re-consecration; and accordingly, several consecration crosses, of Marshall's date, are still visible outside the south wall of the Nave, proving the correctness of the tradition, preserved by Hoker, that the first Cathedral was finished by him. The position of his tomb, in the third arch (counting from the east), on the north side, is exactly what would be assigned to him as quasi-Founder, if, as. I have supposed above, the altar of his completed Church was just there. But all this would not necessitate any alteration of the position of the stalls, but only an enlargement of the sacrarium. The stalls may hitherto have occupied the transept and first bay of the Nave, not extending into the eastern limb of the cross at all; as is still the case at Chichester. At any rate they were moved now, - I mean in Stapeldon's time. For we find in the Fabric Roll of 1309-10, a payment of 52s. 6d to John de Glaston, for "removing the stalls" (44) (stallos); not the walls, as Dr. Oliver read it, taking it to refer to the substitution of new walls for the old: a work which, I have endeavoured to shew, never took place at all. The entry is valuable, as confirming the view that the stalls originally occupied a different position from the present one. It also falls in with the unquestionably early date of the subsettia or " misereres," still remaining. These, from the stiffness of their foliage, cannot be placed later than the middle of the thirteenth century: and I have accordingly ascribed them above to Bishop Bruere (1224-1244). The origin of the name of "miserere," or "misericorde," is curious. Originally, (as the term "stall," from stare " to stand," implies), the rule was for the clergy to stand during the service, when not kneeling. By the eleventh century, however, sitting had come partially into use. Afterwards the device was hit upon of making the seat move upon hinges or pivots, so that it could at times be turned up and present a smaller seat, giving less support. By this compromise the Monks or Canons (as the case might be), were enabled to rest in some degree during an unusually long service, without altogether abandoning the standing position. As early as 1121, Peter of Clugny speaks of "the raising up of the seats," at a particular part of the service; and about the same time the word misericorde (or miserere) is applied to them, signifying the indulgence conceded by the use of them: whence the seat was also called a "patience" (i.e., sufferance, or permission) a word used in that sense by Hooker and Shak-speare. (46)