The method is always the same; the ground having been covered with an even coat of enamel, the pattern is scratched out clear with the point of a stick or a brush handle. Plates XVI., XX., XXI. are typical examples, and show in detail the kind of pattern that was used. It is very rarely that we find anything of the kind in the previous period. There is, as we have seen, an isolated and early example of it at Canterbury, where a rather paler, poorer blue has been used than in the other windows, but there it is more delicate than in the fourteenth century, the pattern being scratched out of a very thin semi-transparent mat of enamel; and it is found too in some late thirteenth century windows in St. Urbain at Troyes, but in fourteenth century work it is frequently met with even at the beginning of the period, and by the end of the first quarter of the century it is the rule, and remains so throughout the succeeding century as well. [Illustration: PLATE XXVIII BORDERS, FROM PLATE XXV] IX EARLY FOURTEENTH CENTURY GLASS IN ENGLAND (MERTON COLLEGE AND EXETER) The windows in the chapel of Merton College, Oxford, are perhaps the earliest in which the design of the Second Period has taken a definite and typical form. Antony à Wood, in his catalogue of Fellows, says that the donor, Henry de Mamesfield or Mannesfield, whose portrait is in the windows, caused them to be made in 1283, but in view of an order in the Bursar's Rolls of 1292 for stone for building these windows, this date must be rejected. Antony à Wood's statement elsewhere that the _whole_ chapel was pulled down and rebuilt in 1424 shows he is not altogether to be relied on. The presence of the fleur-de-lis with the castle of Castile in some of the borders makes it probable[12] that they were done after Edward I.'s second marriage with Margaret of France in 1299, while the arms of an Heir-Apparent as well as of a King of England in the east window makes it certain that they were executed while Edward I. and his son were both alive, _i.e._ before 1307. On the whole, and by comparison with the York glass, I should think 1303-1305 a not improbable date for them. [Sidenote: The east window.] There are seven windows on each side and a great east window, and, with the exception of the latter, they are still fairly perfect. Of the east window it is only the beautiful "wheel" tracery which retains its original glass, the lower lights, alas! having been destroyed in 1702 to make room for a monstrosity by one Pryce--a horrible blare of yellow. What remains in the tracery has a more transitional character than the other windows, and was probably executed first, and if only the lower lights had remained they might have thrown an interesting light on the development of the style.