The Greek colonnade was thus an exterior feature, surrounding the solid cella-wall instead of being enclosed by it as in Egypt. The temple was a public, not a royal monument; and its builders aimed, not as in Egypt at size and overwhelming sombre majesty, but rather at sunny beauty and the highest perfection of proportion, execution, and detail (Fig. 34). There were of course many variations of the general type just described. Each of these has received a special name, which is given below with explanations and is illustrated in Fig. 31. _In antis_; with a porch having two or more columns enclosed between the projecting side-walls of the cella. _Prostylar_ (or prostyle); with a columnar porch in front and no peristyle. _Amphiprostylar_ (or -style); with columnar porches at both ends but no peristyle. _Peripteral_; surrounded by columns. _Pseudoperipteral_; with false or engaged columns built into the walls of the cella, leaving no pteroma. _Dipteral_; with double lateral ranges of columns (see Fig. 39). _Pseudodipteral_; with a single row of columns on each side, whose distance from the wall is equal to two intercolumniations of the front. _Tetrastyle_, _hexastyle_, _octastyle_, _decastyle_, etc.; with four, six, eight, or ten columns in the end rows. +CONSTRUCTION.+ All the temples known to us are of stone, though it is evident from allusions in the ancient writers that wood was sometimes used in early times. (See p. 62.) The finest temples, especially those of Attica, Olympia, and Asia Minor, were of marble. In Magna Græcia, at Assos, and in other places where marble was wanting, limestone, sandstone, or lava was employed and finished with a thin, fine stucco. The roof was almost invariably of wood and gabled, forming at the ends pediments decorated in most cases with sculpture. The disappearance of these inflammable and perishable roofs has given rise to endless speculations as to the lighting of the cellas, which in all known ruins, except one at Agrigentum, are destitute of windows. It has been conjectured that light was admitted through openings in the roof, and even that the central part of the cella was wholly open to the sky. Such an arrangement is termed _hypæthral_, from an expression used in a description by Vitruvius;[9] but this description corresponds to no known structure, and the weight of opinion now inclines against the use of the hypæthral opening, except possibly in one or two of the largest temples, in which a part of the cella in front of the statue may have been thus left open. But even this partial _hypæthros_ is not substantiated by direct evidence. It hardly seems probable that the magnificent chryselephantine statues of such temples were ever thus left exposed to the extremes of the climate, which are often severe even in Greece. In the model of the Parthenon designed by Ch. Chipiez for the Metropolitan Museum in New York, a small clerestory opening through the