Page:History of Art in Primitive Greece - Mycenian Art Vol 2.djvu/190

The Origin of Doric Architecture. 149 plain surface of the latter. This canonical rule was strictly- obeyed, for we observe that the architect of the C. Selinous temple has cut the metopes in such a way as to prevent the figures on them jutting out, or infringing on the plan of the architrave. This, when plain, is sometimes, though rarely, adorned by a frieze ; but in that case it is always associated with civil architecture, such as porches and Roman buildings. The peg, moreover, is much too slender to have effectually prevented the contrjiJtfon and expansion of a stout timber beam. To us the solution of the problem appears to have been solved by the alabaster frieze which Schliemann brought out of the ruins of the palace atJTjr^ns (Figs. 226, 227). The pattern in its main lines is one which is frequently found in the Mycenian ornamental scheme. It consists of a double design, two semi-circles front to front, divided by horizontal strokes, and a narrower rectangular band whereon are traced vertical striae, with due accompaniment of squares and elegant rosettes. This rhythmic arrangement brings to the mind metopes and triglyphs. The analogy is marked enough to have caused the learned architect who discovered the frieze to use the above terms in describing the antiquity. This composite design had been met before on glass- pastes from Spata and Menidi (Fig. 225), as well as on green or red bands that have been collected in the acropolis of Mycenae (Fig. 224), and near to the bee-hive tombs of the lower city where they formed part of the fa9ade decoration (Figs. 272, 273). The elements of the pattern are alike in both, but from the older bands to the alabaster frieze there occurs a notable difference. Every- where else, except in the Tirynthian frieze, the component parts always appear in the same order, and without breaks, answering to what may be called the organic divisions of the artificial unit ; the series only stops short at either end of the slab, to be similarly repeated on the adjacent plaque. At first, these joints can scarcely have been visible. The length of each of the bands does not correspond with the natural articulations of the decor- ation, but was regulated by the dimensions of the block from which it had been derived, and the convenience of the stone- cutter. The thickness of these slabs is uniform ; their juxta- position resulted in a continuous ornament which might be applied to any surface, whilst they would keep inviolate the