IN HACCOMBE CHURCH. 153

and his grandang1it(n" P]nli[)])a brought it to Sir Nirhohis Carew, who deceased in 1 104, aged G9, as appears from llie elegant brass to his memory in the chancel. It is clear that the tiles, even if they were laid down by Sir Warren Arcli- deacon, could not have been designed much later than 1370, but they were probably twenty years earlier. 7. Within a circle, two bii'ds seated back to l)ack, looking at each other : between them a })lant, possibly intended for some sort of dead nettle. The corners arc filled, apparently with a quarter of a circle, and another figure ada])ted in form to the circles on each side of it. This is not an imconnrion ornament. An example occui-s in the vestry of Bristol cathedral ; and I have seen a drawing of a similar tile at Tintern abl)ey. The more common arrange- ment, however, is with the whole })laced diagonally, and the birds seated on branches of the ])lant, which has usually tre- foils at its upper extremity. This occm-s at Winchester, Exeter, Bristol, and Salisbmy. And a similar but simpler form exists at Hereford. 8. A circle, the corners being filled with foliage growing from it, and having a sort of diamond formed within it by circles sprung from the corners as centres. This diamond is filled by a cross and four scpiares, and the fom^ spaces formed by the intersecting circles have fish in them. On the authority of this tile we might arrange those con- taining fish {e,v. gr. Nichols, p. vi.) in squares, head to head, and tail to tail, as w^ell as heads inwards, of which exam])les exist in the Exchequer chamber, Exeter cathedral ; in which case they resemble another not uncommon tile. Indeed some faint traces of the arrangement here suggested, exist in St. James' chapel, in the same cathedral. 9. A diamond, formed similarly to that in the centre of the last described tile, and is filled with a row of sjiots and a flower of eiglit petals, both adapted to the s[)ace. The foui- corners contain coarsely designed fiem-s-de-lis, pointing out- wards. A somewhat different tile, to which the above description would ai)ply, occurs in St. Michaers cha])el, Exeter cathe- dral. 10. This tile is divided into nine spaces by narrow strips Nichols' No. 98 seems to be a fratinicnt of a tile oftliis sort. VOL. III. X