Tkov. ig; side of four metres. Semi-circular foundations are found along the foot of the right wall. What did they carry ? A row of pillars ? The destruction of this side of the building is so complete that to hazard a guess, even conjecturally, would be vain. By the side of this erection, and separated from it by a narrow passage, is a smaller one (b) ; like the former it is oblong in shape, and has a vestibule that opens outwardly. The resemblance ends here ; its inner economy is widely different, having two rooms of unequal size instead of a single one. There seems to have been a fellow to this structure on the left of the main edifice or hall, where, on the thither side of the great trench, are two pieces of wall that cross each other at right angles (PI. I. e). They are all that the great trench of 1872 has spared ; such as they are they suffice to give us the width of the building, which is that of the smaller chamber b. The obvious induction to be drawn from this coincidence is that it extended to the length as well. We are thus enabled, by a very specious hypothesis, to reconstruct a unit composed of a central pavilion, flanked by minor and lower ones. Though close to each other, these erections were quite detached, a narrow space intervening between their external walls. In order to go from one pavilion to another, one had to go through the common courtyard. Was this isolation demanded by social exigencies, or should it not rather be set down to the inexperience of the builder ? It was a much easier plan to set up three independent blocks of building, than to construct a whole sufficiently large so as to provide apartments for carrying on public and private life, and special ones for keeping the sexes apart, for which a whole system of covered passages and entrances had to be contrived in order to connect them the one with the other. It is inexpedient to say much of the remaining structures marked on M. Dorpfeld's plan (PL I.), for little else than the foundations remain, and the ground has kept no trace of doors having stood there. All we need remark in regard to the build- ing D, situate to the southward of pavilion e, is that it consisted of several apartments, and that in all likelihood it had its front turned to the courtyard of the palace, of which it formed perhaps an annex. The ground-plan described above, made up of a great hall preceded by a vestibule (h, k), is repeated in the erections found north-east of b. The whole has been rebuilt throughout.
