In speaking of the walled town, however, as the castle of the community, it must not be forgotten that the castle is, in its origin, the stronghold of a single owner. That origin is still to some extent a vexed question; for the well-known theory of Mr G. T. Clark, that the castle of Norman times was identical with the _burh_ of the Saxon Chronicle, was accepted as a dogma by the antiquaries of twenty-five to fifty years ago, and a theory thus established, however precipitately, is not easily shaken. The patient and thorough work of Mrs Armitage, which deserves the admiration of every scholar, has done much to disturb the foundations on which Mr Clark built his hypothesis; and Mr Neilson, Dr Round, Mr St John Hope, and others, have contributed their share to the discovery of the real character of the evidence, and the formulation of a sounder theory. The present writer has devoted much time to the study of the original authorities for Saxon and Norman military history, and it is his conviction that the weight of documentary evidence is entirely upon the side of the views upheld with so much ability and originality by these recent investigators. At the same time, the earthworks of early castles still present several difficult problems; and the discredit into which Mr Clark’s theory has fallen is a warning against the too confident acceptation of the conclusions of a more critical age, and against the danger of forcing exceptions into the service of the rule. In the earlier part of this book, some allusion is made to methods of Roman warfare; and the main points of two of the sieges conducted by Cæsar and his lieutenants are summarily described. It need hardly be said, in view of what follows, that the methods of military architecture in the middle ages have, for the most part, their exact prototypes in Roman and Byzantine history. The student of the siege-campaigns of Philip Augustus will be constantly reminded, for example, of the relation by Ammianus of the exploits of Julian the Apostate. Too much emphasis cannot be laid on the importance, first, of the contact of the Northmen who overran England and France with the traditional expedients of Roman siegecraft, as they existed in the eastern empire, and secondly, of the influence of the Crusades upon the development of medieval fortification. The conditions of our military architecture in the middle ages were naturally governed by the methods of attack employed by a besieging force. As these had been brought to a high state of perfection in the east, an advance upon which was hardly possible, the history of English fortification, from the Norman conquest to the general adoption of fire-arms in warfare, is that of a progress towards a system of defence in which western Europe lagged far behind the older centres of civilisation.