Almost every public and private edifice, with the exception of picturesque cottages, and buildings erected for the purposes of defence, has a regular form, consisting of two halves, one of which is the inverted image of the other. The inferior parts of the building, such as the doors and windows, have the same regular character; and hence it necessarily follows, that all the decorations, whether in the form of rectilineal borders, circular patterns, or groups of figures, should not only have the same symmetry, but should also be symmetrically related to the bisecting line which separates the building into two halves. If, for example, a rectilineal border, surrounding a building like a belt, consists of a pattern, or of lines inclined in one direction, such a border is not symmetrically related to the vertical and horizontal lines of which the building consists. Hence it will follow, that sculptures, representing an action of any kind, or statues representing living objects, when they are sufficiently large to be seen at the same time with the whole building, can never connect themselves with its regular outline. If these sculptures, or statues, are inverted so as to form a Kaleidoscope pattern, like the beautiful sculpture in the door of the temple of Jun-wassa,[12] they may then be employed without the risk of destroying the general symmetry of the edifice. These remarks are equally applicable to every object which derives its beauty from symmetry; and it is curious to observe the numerous deviations from this principle, and the bad effects they produce on some of the finest vases and ornaments of the Romans.[13] If Mr. Cockerell’s ingenious theory[14] of the original composition of the statues of Niobe and her children be correct, the mode of grouping the figures will show how much the artist was disposed to sacrifice every other kind of effect, to obtain something like a symmetrical group within the pediment. The gradual increase in the height of the statues towards the middle of the tympanum, and their inclination on both sides towards the same point, form strong proofs of Mr. Cockerell’s hypothesis, and afford a singular example, the only one with which we are acquainted, of an attempt to reconcile the apparently incongruous effect of a real picture and a symmetrical group. Had the statues been confined to one-half of the tympanum, while the other half was a reflected image of the first, we are persuaded, that though the effect, as a picture, would have been diminished, yet the effect, as a part of the temple, would have been greatly increased. A very remarkable example of this species of symmetry is shown in the fine painting of the Four Sibyls, by Raphael, which is now in the church of Santa Maria della Pace, at Rome. [12] See _Edinburgh Encyclopædia_, Art. Civil Architecture, plate CLI. [13] See Hope’s _Costume of the Ancients_, vol. ii. plates 230, 264, 278, figs 1 and 298, fig. 2.