354 History of Art in Antiquity. Not only is there a whole column, with its bull capital still in place, but the bases that once belonged to eight other pillars, including a number of jambs of doorways and nicheSi are extant, whilst the ground is literally covered with drums (Fig. 23) still attached. If this is quite enough to prove the importance of the edifice, it ' is not enough to re- move the difficulties towards restoring the plan. Door- frames and pillars appear to have been surrounded by a wall, the marks of which are visible, and which encom- passed an irregular polygon. Semi'cir- cular saliences, seem- ingly the foundations of towers, occur at regular intervals on one face, in length 7 5 metres (Fig. 168). What was the. pur- pose of these coun- terforts in such a situation? Had they extended on the other sides we should look upon them as defensive works. The remains of the inner dispositions, however, have nothing about them to recall a fortress. It is just possible that the wall did not belong to the original plan, for its foundations consist of small stones bedded in good mortar, whilst on the north face are courses of laige units joined together without cement These fragments are 2 m. 35 c. and 2 m. 75 c. thick, but the masonry wall is barely i m. 70 c Hence the presumption that it is younger than the pillars and the massive open- ings, which belonged to a hypostyle hall fronted by a porch, a hall Digitized by Google Flu. 168. — IsUkhr. Principal palace. General plan of ruins. yLANDm aod CosTi, An» mwAmimv Fhte LVIII.
