requires that the abacus should be adjusted to and conform to the upper part of the capital, so that they fit each other. Now a square abacus would not at all suit a capital whose foliage spreads out so widely at its angles. We have spoken of the capitals of the Tower of the Winds, as, according to our opinion, showing the earliest formation of what we now call the Corinthian capital. In that example the abacus is square, and the upper row of leaves are of the kind called _water-leaves_, from their resemblance to those of water-plants, being broad and flat, and merely carved upon the _vase_ or body of the capital. The next stage of progress or transition was to add larger curling leaves and volutes to support the angles of the abacus; but then if the abacus remained square as before, it would either overhang the capital too much in the centre of each face of it, or would not cover the enlarged sweep of the leaves and volutes at the angles. Consequently, it was necessary to effect both extension and reduction for the abacus,—extension at its angles, and reduction as regards its general bulk, which beforehand apparently very difficult, if not impossible feat, was accomplished in the simplest manner possible, by merely curving the sides of the abacus. Thus not only is the abacus expressly shaped in conformity with the great projection of the leaves, &c., at the four angles of the capital,[3] but a beautiful contrast, contributing to general harmony, ensues in consequence,—the concave sides of the abacus corresponding antithetically—in other words, contrasting with the general convexity of the capital. The general structure and configuration of the Corinthian capital are now described, and, it is to be hoped, understood also. It may be as well, however, just to recapitulate: the body of the capital is surrounded by two rows of leaves, eight in each row; besides which there are four leaves, which, with the volutes over them, serve to support the four angles of the abacus, which is fashioned as we have just been explaining. Simple and limited as these elementary and _constitutional_ forms may be thought to be,—insufficient for any great variety or range of character, the Corinthian capital admits of almost infinite modification. Putting entirely aside all details and their minutiæ, the capital of this Order is susceptible of very great diversity of character in regard to its general proportions alone, as may be seen by a comparison of a few—and they but a very _few_—examples given in a note below.[4] [3] For similar reason, the same concavity in the sides of the abacus takes place in the four-faced Ionic capital, the abacus being so shaped in order that it may subtend over and cover the diagonally turned volutes.