Page:History of Art in Sardinia, Judæa, Syria and Asia Minor Vol 2.djvu/68

 52 A History of Art in Sardinia and Judaea. lated at this point leads to the supposition that there were here covered halls adjoining the principal building.^ Behind the last wall is a great heap of earth, rounding off in a semicircle as it approaches the ground. At the base of the large enclosure, as well as of the blocks within it, was found a great quantity of fine-grained white marble crumbled to pieces. Some of these fragments, pounded very small or pulverised, cover the upper part of the wall. Here and there are seen shallow cavities, apparently to receive tablets fitted into the wall, which they divided into panels. Now the whole building looks like a huge rock cut with the chisel ; but this cannot have been the primitive aspect of the monument, and we may reasonably suppose that every part was richly revHed in order to disguise meanness of material. If the excavations made here in 1836 by M. Gillert, then French consul at Tarsus, yielded very poor results, they were of service in demonstrating the structural nature of the twin blocks within the court. A gallery, i m. 80 c. high by i m. 50 c. wide, was with some difficulty excavated more than half-way through the depth of the larger cube, as far as Q V, but no hollows representing chambers were encountered, it being full throughout. A shaft at Q V was then sunk, without bringing aught of any interest. A trench was next attempted between the blocks at Y ; but, beyond chips of white marble and fragments of so-called red Samian pottery, the only object met with was a huge finger of white marble, supposed to have belonged to a colossal statue, the re- mains of which are strewing the ground in great masses, and which formerly graced the monument. Such an hypothesis is borne out by countless imperial coins struck at Tarsus, bearing on one side a rude structure, doubtless one of the chief monuments of the city (Fig. 273). As will be ^ Langlois mentions " vaults ; " but he uses architectural nomenclature in such haphazard fashion, that it would be unsafe to rely upon a chance word, followed by no passing allusion to his having lit upon keystones or fragments of the actual vaults built of concrete. Fig. 273. — Bronze Coin, Tarsus. Lajard, Plate IV.