ESSEX BRICKS 253 buildings up to, perhaps, the late seventeenth century. It is often used in the eighteenth century and the nineteenth century to pro- duce initials and dates in brickwork and, although effective and permanent, may sometimes be clumsy. Another device of arranging for three or four courses to project about 1in. from the rest of the brickwork face—the "string course", as it is called—is a most pleasing relief to an otherwise blank wall. It is met with from sixteenth century down to the end of the eighteenth century, after which it is unusual. Possibly in imitation of wooden rafter-ends, an arrangement of alternate headers projecting from the course immediately under the eaves, usually suggests a late eighteenth century or first half of the nineteenth century date, and 10in. to 101/2in. bricks were specially made for this purpose. One other form of ornament should, I think, be mentioned; the use of light red bricks to outline door and window openings in a wall built otherwise of yellow or dark red bricks—so-called "eye- brows"—made its appearance just at the turn of the seventeenth century and became general in Anne's reign. It usually dates the wall between 1695 and 1720, although country examples of much later date are known. To recapitulate briefly: I have touched upon the manufacture of bricks by hand, the limitations and consequences of the process. I have illustrated Roman bricks and the earliest certain post- Conquest bricks in England. I have suggested that the art of brickmaking may have been dormant, but not entirely extinguished in the Dark Ages; that is to say, it did not have to be re-invented in East Anglia, at least. I have described the later brick developments, including special shapes and the use of moulded brick ornament, and the changes in size, colour, and texture. How far can this knowledge of bricks be used and relied upon as evidence of date of a building? It can be used to the extent I have already indicated, provided one is alive to the pitfalls, for instance: (i) There is no observable difference between Roman Bricks of the first, second, third, and fourth centuries; (ii) The presence of exceptional sizes and colours in all periods may lead one astray: (a) 6in. x 3in. x 11/2in. bricks are found in the Roman Forum site, and even smaller ones elsewhere at Colchester; (b) 10in. x 3in. thick bricks are found in sixteenth century work, although I said they are characteristic of late eighteenth century ; (c) Small bricks (6in. x 3in. x l1/2in.) are found in Wren's work (it is true, for a special purpose) ; (d) Dutch Bricks were nearly always smaller than English counter- parts and their importation for an eighteenth century building may suggest a much earlier date in error.