One of the best definitions of a coat of arms that I know, though this is not perfect, requires the twofold qualification that the design must be hereditary and must be connected with armour. And there can be no doubt that the theory of armory as we now know it is governed by those two ideas. The shields and the crests, if any decoration of a helmet is to be called a crest, of the Greeks and the Romans undoubtedly come within the one requirement. Also were they indicative of and perhaps intended to be symbolical of the owner. They lacked, however, heredity, and we have no proof that the badges we read of, or the decorations of shield and helmet, were continuous even during a single lifetime. Certainly as we now understand the term there must be both continuity of use, if the arms be impersonal, or heredity if the arms be personal. Likewise must there be their use as decorations of the implements of warfare. If we exact these qualifications as essential, armory as a fact and as a science is a product of later days, and is the evolution from the idea of tribal badges and tribal means and methods of honour applied to the decoration of implements of warfare. It is the conjunction and association of these two distinct ideas to which is added the no less important idea of heredity. The civilisation of England before the Conquest has left us no trace of any sort or kind that the Saxons, the Danes, or the Celts either knew or practised armory. So that if armory as we know it is to be traced to the period of the Norman Conquest, we must look for it as an adjunct of the altered civilisation and the altered law which Duke William brought into this country. Such evidence as exists is to the contrary, and there is nothing that can be truly termed armorial in that marvellous piece of cotemporaneous workmanship known as the Bayeux tapestry.