Mode of Construction. 479 a saw, and two different drills. The decoration of the domed- tombs is entirely made up of applied pieces, large marble slabs, cut very thin into a variety of shapes. To have endowed these with a very accentuated and diverse relief, notwithstanding their slimness, together with extreme fineness of joints, implies no mean skill on the workman's part, whose dexterity is no less remarkable in the spacious domed-chambers hollowed out behind these frontages. The external decoration has almost entirely disappeared, but the inner building is sufficiently well preserved to enable us to grasp the methods of the Mycenian constructor. The cupola is circular on plan. To obtain a curve approaching the circle, the mason must have used a line fastened at one end to a post sunk in the ground, in the middle of the building then in process of construction ; whilst he used the other end, which he slowly moved round the central pole, to exactly measure with the eye the distance from it to the stone rings as they rose in succession. The parabolic shape which the dome describes in elevation was obtained by corbel courses or off-sets until the top was reached ; the angles were then knocked off, and at the same time the requisite convexity was worked on the surface of each block. ^ Thus, with the simple superimposition of horizontal courses, the architect succeeded in obtaining results which we now demand of the vault. But the system, though ingenious, is not free from grave faults. Here, as in the shaft-graves, the body had to be hidden underground, and the cupola covered with loose stones and earth ; their downward pressure, however, was unevenly distributed, and the thrust greater on certain points than at others ; hence the rings, which have not the same coherence as construction which makes use of voussoirs, in time slipped and overlapped one another with dire effect, in the Treasury of Atreus for example. The stones of that part of the cupola which faces the dromos have been moved towards the interior by the superimposed weight ; although the pressure was not strong enough to tear away and hurl the units into space, it none the less has brought about a very marked depression in ^ Abel Blouet and Thiersch think that the polishing of the stone surface was gone through after the units were fixed. The work could of course have been previously executed in the stone-yard; but here we may reasonably supix)se the simplest processes to have been used.
