It appears that, just as the forms and proportions of a building and of all its details were determined by precedent, yet not so absolutely as to leave no scope for the exercise of individual genius, so there was an established system in the coloring of a building, yet a system which varied somewhat according to time and place and the taste of the architect.  The frontispiece attempts to suggest what the coloring of the Parthenon was like, and thus to illustrate the general scheme of Doric polychromy.  The colors used were chiefly dark blue, sometimes almost black, and red; green and yellow also occur, and some details were gilded.  The coloration of the building was far from total.  Plain surfaces, as walls, were unpainted.  So too were the columns, including, probably, their capitals, except between the annulets.  Thus color was confined to the upper members—­the triglyphs, the under surface (soffit) of the cornice, the sima, the anta-capitals (cf.  Fig. 54), the ornamental details generally, the coffers of the ceiling, and the backgrounds of sculpture. [Footnote:  Our frontispiece gives the backgrounds of the metopes as plain, but this is probably an error] The triglyphs, regulae, and mutules were blue; the taenia of the architrave and the soffit of the cornice between the mutules with the adjacent narrow bands were red; the backgrounds of sculpture, either blue or red; the hawk’s-beak molding, alternating blue and red; and so on.  The principal uncertainty regards the treatment of the unpainted members.  Were these left of a glittering white, or were they toned down, in the case of marble buildings, by some application or other, so as to contrast less glaringly with the painted portions?  The latter supposition receives some confirmation from Vitruvius, a Roman writer on architecture of the age of Augustus, and seems to some modern writers to be demanded by aesthetic considerations.  On the other hand, the evidence of the Olympia buildings points the other way.  Perhaps the actual practice varied.  As for the coloring of Ionic architecture, we know that the capital of the column was painted, but otherwise our information is very scanty.

If it be asked what led the Greeks to a use of color so strange to us and, on first acquaintance, so little to our taste, it may be answered that possibly the example of their neighbors had something to do with it.  The architecture of Egypt, of Mesopotamia, of Persia, was polychromatic.  But probably the practice of the Greeks was in the main an inheritance from the early days of their own civilization.  According to a well-supported theory, the Doric temple of the historical period is a translation into stone or marble of a primitive edifice whose walls were of sun-dried bricks and whose columns and entablature were of wood.  Now it is natural and appropriate to paint wood; and we may suppose that the taste for a partially colored architecture was thus formed.  This theory does not indeed explain everything.  It does not, for example, explain why the columns or the architrave should be uncolored.  In short, the Greek system of polychromy presents itself to us as a largely arbitrary system.