a series of such ogees; that is, the tracery bars are described from two series of centres, one on either side of them; and each reticulation is bounded by parts of circles, alternately struck from a point within and a point without it. The more varied patterns of late decorated tracery are formed on the same principle, so far as the position of the centres and the meeting of the curves are concerned. Tangents and secants no longer give the character to the whole, and indeed secants scarcely if ever appear at all.
And where is the point of junction between the two styles? We believe it will be found in certain eccentric forms, which differ most materially to the eye from either class which we have described, but which are formed with a strange combination of the principles of both.
The tracery of Whitby, Tintern, and several other windows, contains certain figures made up of two similar figures interlaced,—as, for instance, two triangles or two squares; but the one is formed of parts of circles struck from centres within, the other of parts of circles struck from centres without the figure. The former shape themselves into rounded or foliated, the latter into acutely pointed figures; and these last carry out the pattern, which, in pure early geometrical, seems complete in itself, into the rest of the window. Henceforward there is a tendency to fusion of several parts, and when that tendency is carried out even to excess, it is still by the same means,—i. e., by combining curves struck from centres, some within and some without the resultant figure. A great difference, however, remains between the geometrical, even in its latest types, and the flowing tracery; the former still brings its circles together in angles, as tangents or secants; the latter always, where it is possible, carries on the same line in an unbroken though a complex curve.
This eccentric and extravagant tracery (and we use the words rather in their strict sense, for the character which we would express consists in a constant struggle to avoid a single centre, and to pass over certain confined limits); this eccentric and extravagant tracery never became common. It rather indicated a tendency than achieved an object; and that probably from its great complexity. Denude them of all accessories, such as cusping and foliation, and still the interlaced triangles from Canterbury (p. 91) form a figure struck from six centres, three within and three without the figure; the interlaced squares from Whitby and from Great Bedwyn (pp. 89, 90) form a figure struck from eight centres, four within and four without its own limits. So complex a system of tracery could hardly be employed very frequently or very long. Its real office was performed when it had led to the introduction of a new kind of tracery, formed by the interfusion of circles, struck alternately from centres within and without the main design or its subordinate parts.
But we have said enough, if we have vindicated our assertion, that we require the separation of at least one style from the two with which it is at present confounded—the Geometrical, that is, from the Early English and the Decorated. As for the change in nomenclature which may thus be justified and even demanded, we leave it to other persons, or at least to another occasion. As regards windows alone, Mr. Sharpe's names, Lancet, Geometrical, Curvilinear, and Rectilinear, are very expressive and very con-
