The above few and simple arrangements of plan are nearly all the varieties that the Greek temple style offers; and some of them are little better than distinctions without differences, inasmuch as the differences do not affect general external appearance. Peripteral, dipteral, and pseudo-dipteral, all agree in the main point, and the two latter answer to the name of peripteral as well as the first, being merely modifications of it. Great as were its æsthetic beauties, Greek Architecture was—why should we scruple to confess it?—exceedingly limited in its compass and power of expression: what it did, it did admirably, but it confined itself too much to one idea. “When you have seen one green field,” says Johnson, “you have seen all green fields;” and so we may say of Greek temples,—when you have seen one or two of them, you have seen all of them. However they may differ from one another as to the treatment of the Order adopted for them, the number of their columns, and mere particulars of that kind, they resemble each other very nearly in all leading points. Not only were their plans invariably parallelograms, but alike also as to proportion, forming a double square, or being about twice as much in length as in breadth; for so exceedingly _methodical_ was the Greek system, that the number of columns on the flanks or sides of a peripteral temple was regulated and determined by the number of those in front. The number of the columns in front was invariably an even one, as otherwise there would be no middle intercolumn; but on the flanks of the edifice, where there was no entrance, the number of the intercolumns was an even, and that of the columns an uneven one, so that a column came in the centre of these side elevations. As to the mode in which the front influenced the sides by determining the number of columns for them, the established rule seems to have been to give the flanks twice as many intercolumns as there were columns at each end: thus the Parthenon, which is octastyle, has _sixteen_ intercolumns, consequently seventeen columns, on each flank. In like manner, a hexastyle temple would have _twelve_ intercolumns and thirteen columns on its sides. There are, however, exceptions; for instance, the temple at Selinus, which has been mentioned as an example of the pseudo-dipteral mode of columniation, is an octastyle, with sixteen, or just twice as many columns on its sides as in front; consequently the intercolumns are only fifteen, and being uneven in number, there is a middle one, as in the front itself. After all, the difference caused by there being an intercolumn more or less than usual is but a very slight one, such as is to be ascertained only by counting the columns, and such as not to occasion any perceptible difference in the general physiognomy of the building.[6]