Page:History of Art in Primitive Greece - Mycenian Art Vol 2.djvu/191

I50 Primitive Greece: Mvcenian Art. internal arrangement of the masonry or wood-frame of which they formed the Hning.^^^ The case is quite-'dTfierent with the Tirynthian frieze. Here we have a separate piece for each of the double elements of the system, one for the pair of tangent semi-circles, and another for the two vertical rows of ornaments. If the single pieces of the first series are thinner and longer, those of the second are narrower, and tail deeper into the wall ; they also jut out in front beyond their fellows, and overlap the facade at the sides by four centi- metres. Hence the frieze presents itself in the form of a series of slabs placed edgewise, and of pillars exercising a downward pressure on these and keeping them in place. The dimensions of the pillars, forty-three centimetres broad by fifty-seven centi- metres deep, are those of a great beam. On their internal face appears a cutting, a kind of rebate which shows that the frieze was intended to be fixed to some flat surface of the edifice. Whatever may have been the mode of joining, it served its purpose, well, and endowed the pillars with the requisite stability, and enabled them to fulfil their function. There is a last peculiarity to be noticed about this frieze: the central groove and the rosettes seen on what we have called "pillars" are not carried up to the top, they stop short within ten centi- metres of the upper rim.^ As to the frieze, we explained why, contrary to the situation it occupied at the time of its discovery, we have introduced it as principal ornament in the palace entablature, where it has the great value of con- trasting and of being set off by the timbers that surround the entablature. Let us look at our alabaster frieze now that it is fixed to the wood architrave. Thanks to the weight of the pieces composing it, thanks above all to the overlapping of the lateral slabs, and the salience of the pillars in elevation on each side, we need fear no displacement of the single pieces, for they make up a coherent and solid whole. The architrave, however, was liable to contract and expand under the action of damp and heat ; the warped surface at the back of the frieze would have caused the pieces to loosen and slip down, with the result that in a short time the edges would have become irregular and the surface rough. The annexed drawings (Fig. 305) will help the reader ^ See ante, p. 140.