The interest of these details, for our present purpose, lies in the confirmation afforded by them of the choir having been treated in two portions. It is at the same time always interesting to trace the growth, in the minds of our great builders, of the conceptions which resulted in forms of so much grace and beauty. Nor can we fail to perceive how great a loss it would have been to our Choir, had the triforium arcade been partially omitted; as in the original treatment of the presbytery it was. For nothing, it may safely be affirmed, is more cathedralesque, so to speak, - more essential, that is, to a Church of the very first order, - than a distinct and well developed triforium arcade. Our Norman-Transition Church, we cannot doubt, had a regular triforium throughout. And it must ever be regretted that the absence of this feature, in the otherwise perfect Minster of York, should reduce it, in this respect, in the judgment of architectural connoisseur-ship, to a Church of the second order.

Another instance, which may be studied in our choir, of a great conception gradually arriving at perfection in the mind of an architect, is that of our marble pillars, one of the peculiar glories of our Cathedral. The compact diamondwise arrangement of their many shafts, so happily combining solidity with lightness of effect, and giving the utmost possible view from the aisles into the centre, and vice versa, was not attained at one bound, or in a moment. The germ of it is found in those Transition pillars (already referred to, p. 9) which carry the arches between the Lady Chapel and the side Chapels. These are probably Marshall's work (c. 1200). And their quatrefoil plan is one step in advance, in point of lightness, on the circular Norman pillar. The next step is to be discerned in the unique pair of pillars which occupy the north-east and southeast angles of the choir. Here the plan has become octofoil; a slender shaft being introduced at each reentering angle of the quatre-foil. Now these pillars are in the immediate neighbourhood of the pair - also unique - of retrochoir windows, before referred to (p. 12) as intermediate in character between Marshall's and Bronescombe's windows. The resemblance in style of these retrochoir windows to those of the choir in Westminster Abbey gives us c. 1230 as their probable date. And we cannot be far wrong in placing these pillars at the same date with them, as features in an Early English modification of the retrochoir. We also observe that these pillars are of Purbeck; probably the first appearance of that marble, for large columns, in the Church; though, for arcading and shafts, we have it in the Chapter House, (c. 1225); and for tombs, e.g., in the tomb of John the Chanter (1191), Marshall's (1210), and the early ones, whoever they belong to, in the Lady Chapel.

It only remained now to impart to this pillar, together with the nobler material, the more perfect diamond plan. This was carried out in the single pillar at the east end of the choir - midway between the two just described - by the introduction of two more shafts in each face of the pillar; the chief or cardinal shafts being greatly reduced in size, yet still sufficiently predominating. This, then (probably Quivil's earliest feat in columnar work, somewhere about the year 1285), became the typical pillar for the whole Cathedral. Quivil himself followed up this lead, in the transepts, and, as we shall presently see reasons for believing, in the first bay of the nave. The mouldings of this single pillar, and of the pillars of that bay of the nave are in accordance with those of the Lady Chapel; while they differ, though slightly, from those of the choir and rest of the nave. We are thus confirmed in the view that they all alike belong to Quivil: and that to him is due the idea, afterwards so magnificently carried out, of the great avenue of uniform columns, through which the eye is led on, or was intended to be, first to the ostium chori, and then to the glories of the Altar, Reredos, and Lady Chapel.

But it is time that we should return from this digression respecting the marble columns, and consider in detail the various works of embellishment and equipment carried out by Stapledon in the choir which Bitton had bequeathed to him.

His first care was, as I have already said, to fix the glass provided, (apparently) by his predecessor, for the only pair of windows which remained unglazed in the eastern part of the choir. For in his first year we find three glaziers wages for "fixing the glass forms in the aisles of the new work, 13s. 4d."(37) And he proceeded without loss of time to complete the glazing, both in clerestory and aisles, and in the northern transeptal Chapel of St. Andrew. We have to regret, indeed, that we are not able to extract from the Fabric Rolls, from this point onward, so distinct and thoroughly satisfactory an account of the stained glass operations as heretofore. A change evidently took place, at this juncture, in the mode of measuring the glass; and this introduces some obscurity into the records. It will be remembered that, in Bitton's time, we found the most perfect correspondence between the quantities of glass ordered, and the area, by measurement, of the windows. Thus, if in the Rolls, we found 1271 feet of glass ordered for four windows, two in the east end and two in the clerestory; the measurement of them, according to a method still usual, {see Note 30), gives us 1269 1/2 feet, or within a foot and a half of the amount ordered. For each of the remaining clerestory windows of the presbytery, again, 270 feet 9 inches were ordered; and the measurement gives us just 270 feet 1 1/2 inches: while, for aisle windows, 190 feet, or exactly the area by measurement, were purchased. (Notes 31, 32.) In all these cases too, the whole of the glass for one or more pairs of windows was ordered at a time. Nor is any distinction made between the glass for the lights, and that for the tracery. But on Stapledon's accession, the entries assume a different form. The glass is bought in smaller quantities at a time than an entire pair of windows required; and apparently in three portions. We have, e.g., more than once, 182 feet for a pair of clerestory windows (38) each of which, we know from former entries, required 270 feet. And then we find separate entries, of a kind unexampled hitherto, for glass "ad hernesium" This cannot well mean anything else than the tracery. In the Middle Age Latin, (see Ducange, who however did not know of this sense of the word), hernesium meant any kind of "harness " or "equipment," and might well be used for tracery, as "furnishing " the head of a window. From 53 to 56 feet of glass are ordered in the entries for the hernesium. And it is carefully specified in one of them, that the amount of 182 feet, for other parts of the window, was arrived at by full "measurement in length and breadth," implying that the hernesium was not so measured(39) And the truth would seem to be, that they began to think that the old way of measurement, (reckoning as if the lights ran all the way up to the top), was too expensive, and determined in future to buy their glass for the lights and for the tracery separately. And the 53 or 56 square feet is a fair allowance for the actual amount of glass in the heads or tracery of the windows in question. But even then the 182 feet of glass for the lights must have been supplemented by another order, of which we have no record. By the next year, however, we find that they had resorted, (apparently), to the old system of measurement, or even consented to a more expensive one. For we have no less than 615 feet of glass ordered for a pair of clerestory windows: whereas the old quantity was 540. It is called this time, "perfect or finished glass:" and it had risen a penny a foot, from 5 1/2d to 6 1/2d.(40) Perhaps there was a "strike" at Rouen in the year 1310-11.