Page:Archaeological Journal, Volume 1.djvu/334

This page has been proofread, but needs to be validated.
316
ON ANCIENT MIXED MASONRY OF BRICK AND STONE.

But in the castle at Colchester, which also appears to be a late Norman structure, we may perceive an attempt made to imitate the appearance of Roman work in the regular and horizontal layers or courses of Roman brick throughout the walls at intervals, and this is perhaps the nearest approximation to Roman work in external appearance we have, but when examined closely, the number of intervening courses of stone and brick greatly differ, and do not present the same degree of proportion generally observable in Roman work; for instance, in examining the courses upwards, from the Norman set-off, of plain stone cut sloping, of the basement, to a certain height, we find most of the courses of brick and stone to be in single and alternate layers, though sometimes we meet with two courses of brick and sometimes with two of stone, and here and there we find a row of bricks set edgewise. The stone with which the walls are externally cased is cut, but the inner portions of the walls are rubble. The basement up to the set-off exhibits fragments of brick irregularly disposed in the masonry, but no regular layers or bonding-courses, as above the set-off. The pilaster-like buttresses are constructed with cut stone at the angles of the lower portion, and with Roman brick at the angles of the upper. The walls are twelve feet in thickness. In the interior we find arches of doorways, windows, and fire-places, formed of single rows of Roman brick, with brick-work disposed in herring-bone fashion at the back of the fire-places, and circular and twisted funnels for the emission of the smoke. In a lofty partition-wall, we find at a considerable height eight rows of Roman brick set edgewise, and disposed in herring-bone fashion without any admixture of stone. These bricks if procured, as they probably were, from the ruins of some old Roman structure, do not appear, from a cursory examination, to have retained any traces of the ancient mortar adhering to them, which we frequently find to be the case where Roman materials have been worked up in structures of a much later date. Not unfrequently the Roman mortar was partly composed of pounded brick.

The windows in the castle at Colchester are small and plain semicircular-headed Norman lights, with external casings of cut stone flush with the wall, whilst the portal on the south side exhibits features of late Norman work in the facing of the architrave, which has bold round mouldings with a projecting hood-moulding.