At Durham the monastic buildings stood to the south of the church, and the library-walk of the cloister was that walk, or alley, or pane, or syde (for all these words are used), which had the church to the north of it. The library was placed there partly for the sake of warmth, partly to secure greater privacy. At Canterbury and at Gloucester, where the church was to the south of the conventual buildings, the library-walk of the cloister was still the walk next to the church, the other walks, as Mr Hope has pointed out to me, being apparently kept clear for the Sunday procession. [Illustration: Fig. 25. The cloister, Westminster Abbey. From Mr Micklethwaite's plan of the buildings.] I propose to explain the system indicated in the above quotation by reference to a plan of the cloister at Westminster Abbey, drawn by my friend Mr J. T. Micklethwaite (fig. 25)[190], and by quotations from his notes upon it. At Durham every vestige of ancient arrangement has been so completely destroyed that it is better to go to another House, where less mischief has been done, and it happens fortunately that, so far as the position of the cloister with reference to the church is concerned, Westminster is the exact counterpart of Durham. I will consider first the last paragraph of my quotation from the _Rites of Durham_, that namely which deals with the presses for books, there called "almeries or cupbords." Mr Micklethwaite shews that the two bays at the north end of the west walk of the cloister, and the second bay from the west in the north walk (fig. 25, nos. 1, 2, 4), were appropriated to the novices, by the existence of several sets of nine holes, evidently cut by boys in their idle moods for the playing of some game. Similar holes have been found at Canterbury, Gloucester, and elsewhere. Next he points out that "the nosing of the wall-bench for six feet of the third bay from the west in the north walk, and in the whole of the fourth and fifth bays, and nearly all the sixth, has been cut away flush with the riser, as if some large pieces of furniture had been placed there (_ibid._ nos. 5, 5, 5, 5). These were evidently bookcases." Eastward of these indications of bookcases "the bases of the vaulting-shafts are cut in a way which seems to shew that there was a double screen there (_ibid._ nos. 6, 6), or perhaps there were bookcases arranged so as to form a screen, which is, I think, very likely. Beyond this screen to the right are appearances in the wall [next the cloister-garth] which seem to indicate a blocked-up locker, but they are rather doubtful. And on the left is a large double locker blocked (_ibid._ 7), and the blocking appears to be ancient. This locker is of the date of the wall (Edw. I.), and may have been an additional book-closet provided, because that on the other side of the church-door [to be described presently] had become too small, and [was] blocked up when the larger bookcases were made opposite the carrells[191]."