1. quarterly, i. and iiii., argent, on a bend azure, three bucks' heads

caboshed or (Stanley); ii. and iii., or, on a chief indented azure, three bezants (Lathom); 2 and 3, gules, three legs in armour conjoined at the thigh and flexed at the knee proper, garnished and spurred or (for the Lordship of Man); 4. quarterly, i. and iiii., gules, two lions passant in pale argent (for Strange); ii. and iii., argent, a fess and a canton gules (for Wydeville). The arms on the escutcheon of pretence are not those of his wife (Anne Hastings), who was not an heiress, and they seem difficult to account for unless they are a coat for Rivers or some other territorial lordship inherited from the Wydeville family. The full identification of the quarterings borne by Anthony, Lord Rivers, would probably help in determining the point. But it should not be imagined that the definite rules which exist at the moment had any such unalterable character in early times. Husbands are found to have quartered the arms of their wives if they were heiresses, and if important lordships devolved through the marriage. Territorial arms of dominion were quartered with personal arms (Fig. 755), quarterings of augmentation were granted, and the present system is the endeavour to reconcile all the varying circumstances and precedents which exist. One point, however, stands out clearly from all ancient examples, viz. that quartering meant quartering, and a shield was supposed to have but four quarters upon it. Consequently we find that instead of the elaborate schemes now in vogue showing {544} 10, 20, 50, or 100 quarterings, the shield had but four; and this being admitted and recognised, it became essential that the four most important should be shown, and consequently we find that quarterings were selected in a manner which would seem to us haphazard. Paternal quarterings were dropped and the result has been that many coats of arms are now known as the arms of a family with quite a different surname from that of the family with which they originated. The matter was of little consequence in the days when the "upper-class" and arms-bearing families were few in number. Every one knew how Stafford derived his Royal descent, and that it was not male upon male, so no confusion resulted from the Earls of Buckingham giving the Royal coat precedence before their paternal quartering of Stafford (see Fig. 756), or from their using only the Woodstock version of the Royal Arms; but as time went on the upper classes became more numerous, arms-bearing ancestors by the succession of generations increased in number, and while in the thirteenth and fourteenth centuries it would be a physical impossibility for any man to have represented one hundred different heiresses of arms-bearing families, in later days such became the case. The result has been the necessity to formulate those strict and rigid rules which for modern purposes must be conformed to, and it is futile and childish to deduce a set of rules from ancient and possibly isolated examples originating in and suitable for the simpler genealogical circumstances of an earlier day, and assert that it is equally permissible to adopt them at the moment, or to marshal a modern shield accordingly. [Illustration] FIG. 756.--Arms of Edward Stafford, Duke of Buckingham (d. 1521): Quarterly, 1 and 4, quarterly, i. and iiii., France; ii. and iii., England, within the bordure argent of Thomas of Woodstock; 2 and 3, or, a chevron gules (for Stafford). (From MS. Add. 22, 306.) The first attempt to break away from the four quarters of a shield was the initiation of the system of grand quarters (see Figs. 755 and 756). By this means the relative importance could roughly be shown. Supposing a man had inherited a shield of four quarters and then married a wife in whom was vested a peerage, he naturally wished to display the arms connected with that peerage, for these were of greater importance than his own four quarterings. The problem was how to introduce the fifth. In some cases we find it borne in pretence, but in other cases, particularly in a later generation, we find that important quarter given the whole of a quarter of the shield to itself, the other four being conjoined together and displayed so as to occupy a similar space. These, therefore, became sub-quarters. The system also had advantages, because it permitted coats which by constant quartering had become {545} indivisible to be perpetuated in this form. So definite was this rule, that in only one of the series of Garter plates anterior to the Tudor period is any shield found containing more than four quarters, though many of these are grand quarters containing other coats borne sub-quarterly. The one instance which I refer to as an exception is the shield of the Duke D'Urbino, and it is quite possible that this should not be quoted as an instance in point. He appears to have borne in the ordinary way four quarters, but he subsequently added thereto two quarterings which may or may not have been one and the same coat of arms by way of augmentation. These he placed in pale in the centre of the others, thus making the shield apparently one of six quarters. [Illustration] FIG. 757.--Arms of George Nevill, Baron Abergavenny (d. 1535): Quarterly, 1. gules, on a saltire argent, a rose of the field (Nevill);