Mode of Construction. 477 five square centimetres — played the part of antae, and stood on well-wrought stone plinths, into which cuttings had previously been made for their reception. They masked the whole of the external face of the wall, one metre forty-five centimetres thick, up to the roof, whose joists rested on them. The employment of wood at Tiryns, Mycenat, and Troy in this situation may have been counselled on the one hand by the bold salience these square timbers furnished, and the firmer support they gave to the wood framing of the loft, and on the other because it was a more expeditious way of going to work. That the working of antai and door-sills of limestone has been carried out in Argolis with the stone saw is proved by the marks left by the tool on the particular blocks. The saw of the Mycenian workman can only have been a bronze knife with a sharp point which a single workman held by the handle as he cut through the stone, using extremely sharp sand or loose moist emery. The size of the saw-cuts, whether at Tiryns or Mycenae, is only two milli- metres. When the stone was sawn two centimetres deep, the piece to be removed was struck off as far as the incision reached, and the sawing was begun anew. The curves of these successive incisions still remain on the stones.^ This primitive mode of going to work was very slow and not quite satisfactory, for it had the effect of leaving rugosities on the surface, often very marked (see Fig. 84), which the workman did not trouble himself to smooth away, being satisfied with polishing a narrow band around the edge of the block, trusting to the timbers to hide all blemishes. Nevertheless, stone plays here a more important part than it does in the Troad, where very thin calcareous slabs, torn away from the parent rock, alone interpose between the wood beams and the ground (Fig. 189). In Argolis, on the contrary, they reposed, now on a base composed of several well-jointed slabs, now on a single block from fifty to sixty centimetres high.- Two different instruments seem to have been used in working the stone. Thus, whilst the soft sandstone has square holes cut with the chisel, the dowel-holes of the hard limestone and the breccia were made with the drill-auger, the lower end of which must have been a hollow cylinder or stout reed. The rapid twirling of the drill bored a round hole, in the midst of which a stone ^ ScHLiEMANN, Tirytis, ' Ibid.
