Page:Archaeological Journal, Volume 7.djvu/226

 150 ANGLO-SAXON REMAINS IN IVER CHURCH, BUCKS. date as the latter. No certain information as to the original termination of the chancel could be obtained without disturb- ing the foundations, which I had neither time nor authority to do. The clear extent then of Saxon masonry consists of the north and south walls of the nave, through w^hich the arcades have been cut. It now remains to examine the only frag- ment of detail wdiich they contain, namely, the Avindow interrupted by the eastern arch on the north side. This is visible on both sides of the wall, but is more perfect in its originally external face, that, namely, which is now in the aisle. On the other side the arch is less perfect, having been mutilated at some little distance above the pier-arch, while in the aisle the label of the latter cuts immediately through it. And, what is much more important, a larger portion of the jamb is visible ; it is no hard matter to complete a semicircle, but we cannot so easily guess at mouldings. Those of this window are very remarkable, being quite dif- ferent from what is usual in Norman work ; and hardly less so from other Saxon remains. But in estimating this latter diversity we must remember both that this window is clearly a specimen of much more finished and artistic work than w^e often find among our existing Saxon fragments ; and secondly, that the Saxon windows with which we are best acquainted are found in belfries ; and we know how windows in that position retained a totally distinct type of their own long after, that is, till the use of tracery was fully developed. The fragment at Ivor stands almost unique as an example of a Saxon window of any degree of finish in the body of a church. It appeal's to have had the double splay characteristic of the ruder Anglo-Saxon windows, but the intej'iial opening was considerably the wader of the two. The mouldings, like all other mouldings, must be studied in a section and not in a description ; but we may remark that, notmthstanding they are rather complicated, with a label, two chamfers on different planes, and a roll lietwcen them, there is nothing like the familiar Norman division into orders, a division excessively rare in Anglo-Saxon work. And the exterior chamfer, between the label and the roll, is even more opposed to Norman precedent. On the external side a portion of the inner chamfer, forming the actual splay, is discernible ;