pieces, because if these were gilt, for example, the son would naturally be no more inclined to perpetuate a particular form of strengthening for his shield, which might not need it, than any particular artistic division with which it was involved, so that the absence of arms composed of ordinaries from the early rolls of arms may not amount to so very much. Still further, it may well be concluded that the compilers of early rolls {110} of arms, or the collectors of the details from which early rolls were made at a later date, may have been tempted to ignore, and may have been justified in discarding from their lists of arms, those patterns and designs which palpably were then no more than a meaningless colouring of the strengthening pieces, but which patterns and designs by subsequent continuous usage and perpetuation became accepted later by certain families as the "arms" their ancestors had worn. It is easy to see that such meaningless patterns would have less chance of survival by continuity of usage, and at the same time would require a longer continuity of usage, before attaining to fixity as a definite design. The undoubted symbolism of the cross in so many early coats of arms has been urged strongly by those who argue either for a symbolism for all these rectilinear figures or for an origin in articles of dress. But the figure of the cross preceded Christianity and organised armory, and it had an obvious decorative value which existed before, and which exists now outside any attribute it may have of a symbolical nature. That it is an utterly fallacious argument must be admitted when it is remembered that two lines at right angles make a cross--probably the earliest of all forms of decoration--and that the cross existed before its symbolism. Herein it differs from other forms of decoration (e.g. the Masonic emblems) which cannot be traced beyond their symbolical existence. The cross, like the other heraldic rectilinear figures, came into existence, meaningless as a decoration for a shield, before armory as such existed, and probably before Christianity began. Then being in existence the Crusading instinct doubtless caused its frequent selection with an added symbolical meaning. But the argument can truthfully be pushed no farther. THE BEND The bend is a broad band going from the dexter chief corner to the sinister base (Fig. 65). According to the old theorists this should contain the third part of the field. As a matter of fact it hardly ever does, and seldom did even in the oldest examples. Great latitude is allowed to the artist on this point, in accordance with whether the bend be plain or charged, and more particularly according to the charges which accompany it in the shield and their disposition thereupon.