72 Primitive Greece : Mycenian Art. precluded their carrying a heavy burden, and if our restoration exhibits a projecting member over these pillars, which they seem to support, it is a false appearance, and not likely to deceive anybody. Flush with the fifth line of sealing-holes, above the letter v of section (PI. IV.), appears a punctated shape, which below shows a block cut to an acute angle on one side, to fit the adjoining wedge-like stone, whilst the other end is inserted into the wall. The roughness of the block and its prominent situation indicate that it was never intended to be seen, but must have been concealed by a wide, projecting slab. We get a hint of what there was here from the fa9ade of the neighbour- ing Tomb II., where, notably on the right side, between the doorway and the relieving triangle, we have part of a slab which projected far out beyond the lintel, and formed a ledge or pent-house. On the lower portion of this slab are chiselled, in low-relief, discs or roundels, in which we recognize the end beams of the roof (Fig. ii8). We are quite sure that no such bas-relief ever existed in this situation on the frontispiece of the principal tomb. On the other hand, the semi-columns would have been incapable of bearing a heavy stone beam ; as to a lining, which might have been fastened with clamps to the lintel, no trace of it has been found in the intervening space between the triangle and the doorway. The only possible solution, there- fore, is that the pent-house was hollow, and entirely covered with sheets of metal nailed or riveted to one another, so as to bring this portion in harmony with the rest of the edifice. The hollow piece in question did not rest on the false columns, but on the corbelled stones of the wall right and left, where PI. IV. shows a couple of sealing-holes. In this way not only was the pent-house married to the wall, but its extremities reposed on the widely-projecting slabs of the abacus, the ends of which were also embedded in the masonry. The columns carried nothing. With regard to the ornament of the brazen beam, we have selected such forms as are most affected by the Mycenian artificer — rosettes, spirals, and discs or beam-ends. We have been particularly mindful, in the rendering of our image, to show how much greater was the finish of workman- ship, and how infinitely richer the aspect of bronze work, as compared with work executed in stone. From the ellipsoidal shape of fine holes left by nails towards
