EXISTING DATA. Azure blue does not predominate in glass of the sixteenth century. The dominant tones are white and gold, the field often being almost entirely white. The subjects are in glass of various colours. The blues tend to a steely grey. RESULT. PROVED BY THE DISCOVERY OF FRAGMENTS, RELATIVELY NUMEROUS, OF BLUE GLASS IN THE TRENCHES. THIS GLASS WAS PROBABLY REFITTED FROM THE WINDOWS OF THE EARLIER WORK ALTERED OR REMOVED BY SUCCESSIVE ABBOTS, AND APPEARS TO BE OF THE THIRTEENTH CENTURY. VI. AS TO THE VAULTING OF "FANS." SCRIPT November 7, 1907. "Abbas Beere ... fecit voltam petriam quod vocatur quadripartus, sed Abbas Whitting ... destruxit ... et restoravit eam cum nov ... multipart ... nescimus eam quod vocatur. November 13, 1907. "Wee saide that ye volte was multipartite, yt was fannes old-style in ye este ende of ye choire, and ye newe volt in Edgare's chappel ... Glost'er fannes." EXISTING DATA It is a fair inference that a chapel of this nature and period (Henry VII. to Henry VIII.) would be vaulted in "fans," and we should have thought Abbot Bere's original scheme would have provided for this. "Fannes old style" would apply to those built on a half-hexagonal section. The real Tudor fan has a circular sweep. RESULT THE EDGAR CHAPEL HAD THE LATER (MULTIPARTITE) FORM OF FAN, AS IS PROVED BY THE NATURE OF THE FRAGMENTS FOUND. ONE OF THE MAIN BOSSES IS EXTANT, AND ON ITS BACK HAS THE DIRECTION (IN SCORED LINES) FOR THE CORRECT SETTING OF THE BLOCK, AND THIS SHOWS TWELVE RIBS, IMPLYING A WHEEL OF TRACERY BETWEEN FANS. VII. AS TO THE POLYGONAL EAST END. SCRIPT November 19, 1907. "The direction of the walls ... was at an angle." November 19, 1907. "Forty and two feete was the hight of ye newe chapelle, and yt was ybuttressed with faire buttresses, and walls slantwise at ye cornere." November 26, 1907. "The east end. Seek for the pillars, and the walls at an angle." November 30, 1908. "The ending of the chappel was at an angle, the sides makyng as it were a baye in the east wall there." EXISTING DATA No existing data from which such might be inferred. The nearest example of a polygonal east end seems to be at Westbury-on-Trym, near Bristol, unless we regard the Lady Chapel at Wells as a parallel instance. This chapel, however, is in reality an elongated octagon, with a domical roof. It was erected A.D. 1326. RESULT PROVED BY DISCOVERY IN JANUARY, 1909, AFTER THE TENTATIVE PLAN OF AN ANGULAR APSE HAD BEEN IN THE PRESS FOR PUBLICATION BY THE SOMERSET ARCHÆOLOGICAL SOCIETY; AND AFTER ACTUAL PUBLICATION IN THE CHRISTMAS NUMBER OF THE _Treasury_ FOR 1908. VIII. AS TO THE DIFFERENCE IN THE FOUNDATIONS. February 19, 1908. "Thin walls and poore foundations in the new work." Nothing known. PROVED BY EXCAVATION. THE FOOTING WALLS OF THE APSE WERE THIN AND POOR, IN MARKED CONTRAST TO THOSE OF THE RECTANGULAR PART OF THE CHAPEL, WHICH WERE EXCEEDINGLY BROAD.