§ XIV. It is, however, carefully to be noted, that decorations are admissible on colossal and on diminutive shafts, which are wrong upon those of middle size. For, when the shaft is enormous, incisions or sculpture on its sides (unless colossal also), do not materially interfere with the sweep of its curve, nor diminish the efficiency of its sustaining mass. And if it be diminutive, its sustaining function is comparatively of so small importance, the injurious results of failure so much less, and the relative strength and cohesion of its mass so much greater, that it may be suffered in the extravagance of ornament or outline which would be unendurable in a shaft of middle size, and impossible in one of colossal. Thus, the shafts drawn in Plate XIII., of the "Seven Lamps," though given as examples of extravagance, are yet pleasing in the general effect of the arcade they support; being each some six or seven feet high. But they would have been monstrous, as well as unsafe, if they had been sixty or seventy. § XV. Therefore, to determine the general rule for shaft decoration, we must ascertain the proportions representative of the mean bulk of shafts: they might easily be calculated from a sufficient number of examples, but it may perhaps be assumed, for our present general purpose, that the mean standard would be of some twenty feet in height, by eight or nine in circumference: then this will be the size on which decoration is most difficult and dangerous: and shafts become more and more fit subjects for decoration, as they rise farther above, or fall farther beneath it, until very small and very vast shafts will both be found to look blank unless they receive some chasing or imagery; blank, whether they support a chair or table on the one side, or sustain a village on the ridge of an Egyptian architrave on the other. § XVI. Of the various ornamentation of colossal shafts, there are no examples so noble as the Egyptian; these the reader can study in Mr. Roberts' work on Egypt nearly as well, I imagine, as if he were beneath their shadow, one of their chief merits, as examples of method, being the perfect decision and visibility of their designs at the necessary distance: contrast with these the incrustations of bas-relief on the Trajan pillar, much interfering with the smooth lines of the shaft, and yet themselves untraceable, if not invisible.