IN THE EARLIER STYLES OF ENGLISH ARCHITECTURE. 287

which is accounted the leading characteristic of Anglo-Saxon architecture. It is not my intention to disprove (for that woukl be a difficult matter) the title to great antiquity those churches may claim, where long and short coignings are used, Imt I wish to throw out a caution to enquirers, lest this appearance should lead them to assign all these buildings to the same age. That they are for the most part early structures there can be no doubt, and this epithet may be even extended above the Norman Conquest, if we are justified in applying the words lapidei iahulafus, as used by William of iMalmesbury in his description of Benedict Bishop's churches, to those towers rising in stages from the perpent blocks of stone that run transversely on their four sides. For instance, at Earl's Barton and Barnack tliis system occm-s, at both of which places the towers rise in stages, dimi- nishing as they rise, and forming separate divisions or stories, marked also by the horizontal bands of perpent stone, from which the superior portions of the building alternately spring. This mode of construction was clearly borrowed from the Romans, who, as is sufficiently known, employed bonding courses of brick, rmuiing parafiel with the ground, to strengthen their walls, so that the inferior materials used in the intervening space might become more effectually tied together. The Romans, as may be observed in all their military build- ings now remaining in England, used their bonding courses horizontaUy ; the Anglo-Saxons used them perpendicularly. At Pevenscy there are courses of tile laid flat, at fixed intervals ; at Earl's Barton there are perpent stones placed upright, also at fixed intervals. The object of both was the same, namely, to supply the want of good building materials by such mate- rials as would hold them best together, and the English masons, placing these large blocks of Shelly oolite or Barnack Rag (for EarFs Barton is supplied with this Shelly oolite from that distance), had merely to fiU in the rubble between them, much in the same manner as brick-work is used in timber- framed houses. The talus table of Colchester castle is geologically of this formation, and, owing to the want of native materials, the architect used the Roman bricks he found in such abun- dance on the spot, both for coigns and bonds, in the same