Our examples of shields show very few varieties, nearly all seem to follow the long kite shape or "Norman pear" Nos. 3 and 4. The great seals of William I., William Rufus, and Henry I. all show this shield, also that of Ilbert de Laci [_Archæological Journal_, vol. iv, p. 249], while the Bayeux tapestry, worked at the end of this century, represents them exactly. Seeing, as we do in the seals of this date, which mostly represented mounted horsemen, the inside only, it may be noticed that all seem to be strengthened round their edges. In those of William I. and Ilbert de Laci a metal rim is shown; in that of William Rufus two of such strips, while in that of Henry I. rivets appear to fasten a similar rim on the other side. In the Bayeux tapestry some round shields with pointed bosses appear, and one of a square shape rounded at the corners. [Illustration] In seals the lettering of the inscriptions is in plain Roman capitals, while longobardic letters for G E A and D appear sometimes, but in several seals preserved from this early date the inscriptions have unfortunately decayed away. [Illustration: Gilbert de Gant, ob. 1156.] THE TWELFTH CENTURY exhibits several varieties of shape, Norman and the Norman convex at the top (Nos. 1, 2, 3, and 4); and many of these are curved, so as to partially surround the body. The great seal of Stephen represents a pointed boss and a shield almost pear-shaped. One very curious seal, that of Richard Basset, about 1145, merits a full description. He appears bearing a kite-shaped or very elongated pear-shaped shield (No. 3), about four feet long, and carried down to a sharp point. The top is bi-lobed, like a heart, and it is strengthened all round by a metal rim, while further strength is imparted by a boss or figure like a simple escarboucle, carried out to the sides. He wields a very powerful sword, with an extraordinarily heavy hilt, and with it he cuts off the large duck-like beak of a formidable rampant-winged animal like a dragon, and, apparently, thus frees a human figure held in its beak. This curious seal is attached to a charter printed in Blomefield's _History of Norfolk_ (see No. 44). The seal of Gilbert de Gant, who died 1156, presents a shield more triangular than No. 2, and in the centre is a sharp pointed boss [_Top. and Gen._, i, 317]. The same boss appears in the seal of Ralph son of William, of Dimsdale-on-Tees, 1174-80 (_Herald and Genealogist_, i, 227); also in a good many contemporaneous seals, and it was evidently meant for use as an offensive weapon. It is noteworthy that on the first seal of Richard I. a spike is shown on the shield; his first coronation was in 1189. On the second seal there is none; his second coronation took place on 17th April, 1194. This may show the date when this fashion was discontinued. In the seal of Sewal de Ethindon, about 1167, the curved shield of Norman form, No. 2, runs down into a long point, somewhat twisted