§ VIII. This being the simplest treatment of the pyramid, fig. 10, from the refectory of Wenlock Abbey, is an example of the simplest decoration of the recesses or inward angles between the pyramids; that is to say, of a simple hacked edge like one of those in fig. 2, the _cuts_ being taken up and decorated instead of the _points_. Each is worked into a small trefoiled arch, with an incision round it to mark its outline, and another slight incision above, expressing the angle of the first cutting. I said that the teeth in fig. 7 had in distance the effect of a zigzag: in fig. 10 this zigzag effect is seized upon and developed, but with the easiest and roughest work; the angular incision being a mere limiting line, like that described in § IX. of the last chapter. But hence the farther steps to every condition of Norman ornament are self evident. I do not say that all of them arose from development of the dogtooth in this manner, many being quite independent inventions and uses of zigzag lines; still, they may all be referred to this simple type as their root and representative, that is to say, the mere hack of the Venetian gunwale, with a limiting line following the resultant zigzag. § IX. Fig. 11 is a singular and much more artificial condition, cast in brick, from the church of the Frari, and given here only for future reference. Fig. 12, resulting from a fillet with the cuts on each of its edges interrupted by a bar, is a frequent Venetian moulding, and of great value; but the plain or leaved dogteeth have been the favorites, and that to such a degree, that even the Renaissance architects took them up; and the best bit of Renaissance design in Venice, the side of the Ducal Palace next the Bridge of Sighs, owes great part of its splendor to its foundation, faced with large flat dogteeth, each about a foot wide in the base, with their points truncated, and alternating with cavities which are their own negatives or casts. § X. One other form of the dogtooth is of great importance in northern architecture, that produced by oblique cuts slightly curved, as in the margin, Fig. LVI. It is susceptible of the most fantastic and endless decoration; each of the resulting leaves being, in the early porches of Rouen and Lisieux, hollowed out and worked into branching tracery: and at Bourges, for distant effect, worked into plain leaves, or bold bony processes with knobs at the points, and near the spectator, into crouching demons and broad winged owls, and other fancies and intricacies, innumerable and inexpressible. [Illustration: Fig. LVI.] § XI. Thus much is enough to be noted respecting edge decoration. We were next to consider the fillet. Professor Willis has noticed an ornament, which he has called the Venetian dentil, "as the most universal ornament in its own district that ever I met with;" but has not noticed the reason for its frequency. It is nevertheless highly interesting.