In the meantime a report had been prepared for the annual volume of _Proceedings of the Somerset Arch. Soc._, and as by this time F.B.B. was fully persuaded that Hearne and Hollar are giving us fact when they state that the total length of the Abbey was 580 feet, he boldly drew a plan of the excavated chapel with an addition in dotted lines, making out the requisite dimension and _showing a polygonal annexe or apse_, and this he caused to be published in the 1908 _Proceedings_.[27] [27] A facsimile of this plan was published in the _Treasury_ for Christmas, 1908. [Illustration: FIG. 6.] During the Christmas holidays F.B.B. again visited the site, and on a bright sunny day looked at the clay bank, and for the first time detected a small squarish patch of pale brown discoloration in the clay at a level much higher than the footings of the chapel. He caused the face to be cut into at this point, which was nearly opposite the southern return buttress, and at a point about 2 feet within the clay face the stump of a wall appeared. This was traced and proved to run north-east at an angle of 67 degrees from the north and south axis of the chapel. The ground was opened in the corresponding position on the north side, and here a trench was found well and truly cut in the clay, and filled with debris of sixteenth-century freestone, tile, and glass fragments, chiefly of painted work, and largely of azure blue.[28] The footing-trench was inclined at a similar angle, but instead of meeting a cross wall at east, it stopped short at the point where such a return wall should be, and this indicated that in all probability the annexe had a door in the east wall,[29] behind the reredos, suggestive of the use of this part as a feretory or relic-chamber, and this recalls almost exactly the plan of the church (collegiate) of Westbury-on-Trym, which is also furnished with a semi-hexagonal apse. [28] _Cf._ p. 37, line 3. [29] _Cf._ p. 35, last line, "Portus introitus post reredos," etc. Calculating that the inner faces of the three walls of this annexe would have been designed to be more or less equal in breadth, for sake of symmetry, a line was struck across the apse at the required point, a stake put in, and a very careful chain-measure made from end to end of the whole range of Abbey buildings--_i.e._, from the internal face of the Lady Chapel or Church of St. Mary at the west, to this point, and the measure was then found to be _precisely 580 feet, as given by Hearne and Hollar_. The true plan of the chapel was published in the 1909 volume of _Proceedings_, and makes an interesting comparison with the conjectural plan (see Fig. 7, p. 64).