I. It has been shown that the Romans, in obtaining full mastery of a powerful engine in building--the arch, were at first bold, subsequently lavish, and, it will appear, finally barbarous, in its application. From the reigns even of the early Cæsars, a tendency may be traced in their architecture to become great in mass, but little in parts--to lessen, in the first instance, the vertical or supporting members; and in the second, to load the superstructure, or supported parts. The progress of corruption might be traced, by regular steps, from vast arches, with groins planted on a single Corinthian column, to the arcades of the palace of Dioclesian at Spalatro. These still are left, exhibiting external and internal ranges of arches, springing directly from the capitals of the columns, without any intervening entablature. What more, we ask, is wanting here, to one of the most decided characteristics of one species, at least, of Gothic architecture, and an elemental principle in all kinds? nothing, save a little less elegance of workmanship in the supports, a pier substituted for the column, and the soffit of the arch bevelled instead of being square; steps successively apparent in posterior remains. Surely, then, it is carrying theory beyond all moderate limits, to contend for a separate origin of the system, when the principles of Gothic building are thus distinctly recognizable in a corruption of classic modes, at an era while yet vigorous practice prevailed, with resources undiminished for its support. This corruption, indeed, evidently proceeded, not so much from inefficiency, as from too eager pursuit of novelty--this too, unrestrained by the immediate presence of more simple forms; for, in the baths of the same emperor, appears a less licentious taste. For the exterior, indeed, such Gothic arcades do not seem to have been soon imitated; but for the interior, their adoption was almost immediate. These intermediate steps it is unnecessary farther to pursue at present.