Page:History of Art in Primitive Greece - Mycenian Art Vol 2.djvu/168

 The House and the Palace. 131 peculiarities of the plan we are considering, has absolutely no exception.^ We do not by any means pretend to lead back to the beginnings of this architecture, but in the presentation of the palace we may assume that it has already passed through several stages ; though still primitive, our image is well ordered, and not destitute of the fairness which comes of ornament ; it is simpler, nevertheless, than it will ultimately be when the art which brought it into being shall hold in its grasp all its resources. Fig. 298 shows the palace in this intermediate phase of its growth. We begin by setting up the columns on the bases that are still fixed to the soil. Their intercolumnation, and the space from each to the anta which covers the head of the wall on either side, are ascertained by the existing stone blocks, on which rested the antae (Fig. 84). On a plane with these are traces of other build- ings. Accordingly, we may reasonably suppose that the middle part of the fagade, comprised within the antae, behind which stood the reception or gala hall, was given greater height than the lateral and subordinate apartments, destined for private uses. Then, too, drainage had been duly considered, and a satisfactory result obtained by a system of terraces built at different levels. In this fashion the gutters of the upper esplanades discharged the waters on those underneath until they reached the soil below. The slimness we have given to our column is in strict accordance with the proportion of the diameter to the height observable in the semi-columns of the sepulchral fa9ades. Here, as in Tomb I., the shaft has about eleven diameters and a half. The pillar, which is wooden, consists of a single or several pieces joined together ; and being timbered, is naturally more slender than a stone support. Timbered antae rested, either directly on stone bases as at Troy (Fig. 187), or as here, in wooden sleepers fixed by pegs to the supporting blocks. Reference has already been made to the house covering, and why it should be a flat roof.^ The simplest mode of building a loft is to place the architrave, made up, horizontally, of two or three parallel beams, on the columns and the heads of the wall which it traverses (Fig. 299). The archi- trave has a bold salience at either extremity of the fa9ade, and supports a kind of cornice which discharged the waters at some 2 We think that Prof. Middleton was mistaken in giving to his restored megaron a double sloping roof.
 * See ante, Vol. I. ch. iv, § 3, a,