It has been already observed, that in the genuine Doric the column consists of only shaft and capital, which latter is composed of merely an _echinus_ and _abacus_, the first being a circular convex moulding, spreading out beneath the other member, which, although a very important one, is no more than a plain and shallow square block upon which the architrave rests, not only firmly and safely, but so that the utmost expression of security is obtained, and pronounced emphatically to the eye. Such expression arises from the abacus being larger than the _soffit_ or under surface of the architrave itself; and as the former corresponds, or nearly so, with the lower diameter of the shaft, it serves to make evident at a glance that the foot of the column is greater than the soffit of the architrave placed upon the columns. Thus, as measured at either extremity, the column is greater than the depth or thickness of the architrave, and projects beyond the architrave and general plane of the entablature. Now this would produce a most unsightly effect were the columns of the same, or nearly the same diameter throughout. In such case they would appear not only too large, but most clumsily so, and the entablature would have the look of being set back in the most awkward and most unaccountable manner. Instead of which, the architrave, and consequently the general plane of the whole entablature, actually overhangs the upper part of the shaft, in a plane about midway between the smallest diameter of the column, just below the capital and the face of the abacus. Even this, the overhanging of the entablature, would be not a little offensive to the eye, were the abacus no larger than the architrave is deep; whereas, being larger, it projects forwarder than the face of the architrave, thereby producing a powerful degree of one species of æsthetic effect, namely, contrast,—and if contrast, of course variety also; for though there may be variety without contrast, there cannot be contrast without variety. Another circumstance to be considered is, that were not such projection beyond the face of the architrave given to the abacus, that and the rest of the capital could not correspond with the foot of the shaft, and thus equalize the two extremities of the entire column. As now managed, all contradictions are reconciled, and the different sorts of contrast are made to contribute to and greatly enhance general harmony. In the outline of the column we perceive, first, contraction,—then expansion, and that in both directions,—for in like manner as the column diminishes upwards and the capital expands from it, its shaft may be said to expand and increase in bulk downwards, so as to agree with the abacus or upper extremity.